Citation

Material Information

Title:
Research Reports in Belizean Archaeology
Creator:
Morris, John
Badillo, Melissa
Thompson, George
Publisher:
University Press of Florida
Institute of Archaeology National Institute of Culture and History Publication
Publication Date:
Language:
English
Physical Description:
Book

Notes

Acquisition:
Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Lila Sadkin.

Record Information

Source Institution:
University of Florida Institutional Repository
Holding Location:
University of Florida
Rights Management:
All rights reserved by the submitter.

Downloads

This item is only available as the following downloads:


Full Text

PAGE 2

Research Reports in Belizean Archaeology Volume 1 5 Archaeological Investigations in the Eastern Maya Lowlands: Papers of the 20 17 Belize Archaeology Symposium Edited by John Mor ris, Mel i s sa Badillo and George Thompson Institute of Archaeology National Institute of Culture and History Belmopan, Belize 201 8

PAGE 3

ii Research Reports in Belizean Archaeology Research Reports in Belizean Archaeology is an annual publication of the Institute of Archaeology, National Institute of Culture and History, Belmopan, Belize. The journal publishes original, peer reviewed papers on the archaeology, prehistory, and ethno history of Belize. Papers may also treat more general theoretical and methodological issues with relevance to Maya archaeology. NOTICE TO AUTHORS Authors submit manuscripts to the editor for consideration as ARTICLES in English or Spanish. The normal expected length of an ARTICLE should not be less than 10 pages but not exceeding 20, including all references cited. Tables and illustrations should be limited to those that are appropriate and necessary. Detailed information on policy, style, and technical matters of manuscript preparation is g iven in English in the Research Reports in Belizean Archaeology : Submission Guidelines . For additional information, updates, or clarification, contact the Institute of Archaeology, NICH, Belmopan, Cayo District, Belize Editorial Board Institute of Archaeology, NICH JOHN MORRIS MELISSA BADILLO GEORGE THOMPSON Editorial Advisory Committee for Research Reports in Belizean Archaeology JAMES AIMERS Department of Anthropology, State University of New York Geneseo, Geneseo, New York JAIME J. AWE Department of Anthropology, Northern Arizona University, Flagstaff, Arizona ARLEN F. CHASE Department of Anthropology, University of Central Florida, Orlando, Florida JAMES GARBER Department of Anthropology, Texas State University, San Marcos, Texas ELIZAB ETH GRAHAM Institute of Archaeology University College London London, United Kingdom CHRISTOPHE HELMKE Institute of Cross Cultural and Regional Studies, University of Copenhagen, Copenhagen, Denmark BRETT HOUK Department of Sociology, Anthropology, and Social Work, Texas Tech University, Lubbock, Texas GYLES IANNONE Department of Anthropology, Trent University, Peterborough, Canada LISA LECOUNT Department of Anthropology, University of Alabama, Tuscaloosa, Alabama PATRICIA MCANANY Department of Anthropology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina HEATHER MCKILLOP Department of Geography and Anthropology, Louisiana State University, Baton Rouge, Louisiana HOLLEY MOYES School of Social Sciences, Humanities and Arts University of California Merced, Merced, California FRED VALDEZ JR. Department of Anthropology, University of Texas at Austin, Austin, Texas GABRIEL WROBEL Department of Anthropology, Michigan State University East Lansing, Michigan JASON YAEGER Department of Anthropology, University of Texas at San Antonio, San Antonio, Texas Research Reports in Belizean Archaeology ( ISBN 978 976 82641 5 2 ) is published by the Institute of Archaeology National Institute of Culture and History Culvert Road B elmopan, Cayo District, Belize C.A. Telephone: +5018222106 or +5018222227. Email: research@nichbelize.org Cover design by Melissa Badillo and John Morris (Institute of Archaeology ) and Steven Richards (Print Belize, Ltd.) Front Cover is based on the 201 7 BAS poster with images from Lamanai and Cuello Back cover design is a compilation of images from Lamanai, St. Georges Caye, Cerro Maya, Cahal Pech, Marco Gonzalez and the Ulua Vase from Pacbitun Copyright 20 1 7 by the Institute of Archaeology NICH Printed by Print Belize Limited. ISBN 97897 6 82641 5 2

PAGE 4

J. Morris et al. iii CONTENTS page SECTION ONE: SETTLEMENT AND LANDS CAPE ARCHAEOLOGY 1. Sampling and Timeframes: Contextualizing The Protoclassic And Early Classic Periods At Caracol, Belize Arlen F. Chase and Diane Z. Chase .................................................................................................... 3 2. Expanding SubPlaza Explorations of Middle Preclassic Architecture at the Site of Pacbitun, Belize George J. Micheletti, Kaitlin E. Crow, and Terry G. Powis ............................................................... 17 3. Exploring Changes in Activities i n Maya E Groups: Archaeological and Geochemical Analysis of E Group Plas ter Floors at Actuncan, Belize Borislava Simova, E. Christian Wells, David W. Mixter, and Lisa LeCount ..................................... 27 4. Fidelity Tests o f Lidar Data f or t he Det ection of Ancient Maya Settlement i n t he Upper Belize River Valley, Belize Bernadette Cap, Jason Yaeger, and M. Kathryn Brown ..................................................................... 39 5. Classic Maya Household Ceramic Be longings: An Untapped Resource f or Understanding Daily Life Sherman Horn III and Anabel Ford .................................................................................................... 53 6. Reconstructing P reclassic M aya H ousehold E conom ies in the Belize River Valley Claire E. Ebert and Jaime J. Awe ....................................................................................................... 65 7. Rock Bottom: Maya Lithic Technology i n t he Early Terminal t o Late Middle Preclassic Periods at Cahal Pech and Blackm an Eddy, Cayo District, Belize W. James Stemp, Jaime J. Awe, M. Kathryn Brown, and James F. Garber ....................................... 79 8. Preclassic Animal Resource Use and t he Origins of Ancient Maya Lifeways and Society: Contributi ons from Belize Zooarchaeology Norbert Stanchly and Chrissina Coleen Burke ................................................................................... 93 9. Ritual Use of Animals i n Ancient Maya Mortuary Contexts: Results of Faunal Analysis f rom t he St ructure A9 Tomb at Xunantunich Chrissina C. Burke, Katie K. Tappan, Gavin B. Wisner, and Jaime J. Awe ................................... 105 10. Fools Make Feasts, and Wise Men Eat Them: Interpreting Problematic SmashAnd Trash Deposits at Kakabish, Belize Kerry L. Sagebiel and Helen R. Haines ........................................................................................... 115 11. The Early Late Classic Transition at Cuello: Results f rom t he 2017 Season of t he Classic Cuello Archaeological Project James L. Fitzsimmons, Natalie Figueroa, and Prasanna Vankina ................................................... 121

PAGE 5

iv 12. Long Ago but Not Forgotten: The Early Preclassic Swasey Cer amic Sphere of Northern Belize Laura J. Kosakowsky, Kerry L. Sagebiel, and Duncan C. Pring ..................................................... 131 13. The Development of Terminal Preclassic and Early Classic Royal Architecture at Chan Chich, Belize Toms Gallareta Cervera, Brett A. Houk, and Claire Novotny ....................................................... 141 14. Early Maya Civilization i n t he Three Rivers Region Fred Valdez, Jr. ................................................................................................................................ 153 15. Situating Preclassic Interments and Fire Pits at Santa Rita Corozal, Belize Adrian S.Z. Chase, Diane Z. Chase, and Arlen F. Chase ................................................................ 159 16. Chetumals Dragonglass: Postclassic Obsidian Production and Exchange at Santa Rita Corozal, Belize Max Seidita, Diane Z. Chase, and Arlen F. Chase .......................................................................... 169 SECTION TWO: GENERAL RESEARCH REPORTS 17. Tagged Walls: The Discovery of Ancient Maya Graffi ti at El Castillo, Xunantunich Leah McCurdy, M. Kathryn Brown, and Neil Dixon ...................................................................... 181 18. Two Unusual Finds f rom Courtyard 3, Pacbitun, Belize Sheldon Skaggs and Terry G. Powis ............................................................................................... 195 19. From Photogs t o Models: Digital Archaeology of Pre Hispanic Pacbitun, Belize Jon Spenard, Michael Mirro, George J. Micheletti, and Terry G. Powis ........................................ 207 20. Setting t he Stage i n Central Belize: 30,000 Years of Tropical Climate, Landscape Transformation, and Human Interaction Lisa J. Lucero, Jean T. Larmon, and Aime E. Carbaugh ............................................................... 219 21. Ceremonial Cir cuit(S) at Cara Blanca, Belize Jean T. Larmon and Aime E. Carbaugh ......................................................................................... 231 22. Preceramic Cultural History in Southern Belize and Its Environmental Context Keith M. Prufer ................................................................................................................................ 241 23. Plastered: Cave Constructions at Las Cuevas Erin E. Ray, Holley Moyes, and Linda Howie ................................................................................ 253 24. Contingent Multi Crafting, Surplus Household Production, and the Maya Quest for Salt Heather McKillop ............................................................................................................................ 265 25. Assessment of t he Shell Midden at t he E leanor Betty Salt Works, Belize Valerie Feathers and Heather McKillop .......................................................................................... 275

PAGE 6

J. Morris et al. v 26. King Kak [U Ti?] Chan Kawil of Pusilha: An Ancient Maya King Andrew D. Somerville, Christian M. Prager, and Geoffrey E. Braswell ........................................ 287 27. Some Lessons Cant Be Taught, They Simply Have To Be Learned: Experiences f rom Three Seasons of Investigations at Alabama, Stann Creek District, Belize Meaghan M. Peuramaki Brown, Shawn G. Morton, Cristina Oliveira ........................................... 297 28. Investigating Ancient Maya Settlement, Wetland Features, and Preceramic Occupation around Crooked Tree, Belize: Excavations and Aerial Mapping w ith Drones Eleanor Harrison Buck, Mark Willis, Satoru Murata, and Jessica Craig ........................................ 307 29. Community Archaeology at Aventura: Archaeology about Communities and Archaeology f or Communities, Re sults of the 2016 Field Season Cynthia Robin, Laura Kosakowsky, Kacey Grauer, and Zachary Nissen ....................................... 319 30. The Maya of Ambergris Caye and Their Neighbors Scott E. Simmons, Tracie Mayfield, James J. Aimers, and W. James Stemp ................................. 329 31. Napoleonic Shakos and t he Order of t he Garter: West India Regiments on St. Georges Caye James F. Garber, Jacob H. Bentley, and Lauren C. Springs ............................................................ 341

PAGE 7

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 3 -15. Copyright 2018 by the Institute of Archaeology, NICH, Belize. 1 SAMPLING AND TIMEFRA MES: CONTEXTUALIZING THE PROTOCLASSIC AND EAR LY CLASSIC PERIODS A T CARACOL, BELIZE Arlen F. Chase and Diane Z. Chase The era of transition between the Late Preclassic (300 B.C. A.D. 250) and the Early Classic (A.D. 250-550) Periods is one which saw great change within ancient Maya society. This change is reflected in the ceramics of this transitional era. Ceramicists have had difficulty isolating distinct ceramic c omplexes within the transitional era and have instead tended to focus on specific stylistic markers (e.g., mamiform tetrapods) that were thought to be hallmarks for this transition. These stylistic markers became known as the Protoclassic and, while eas ily identified, they were never securely anchored within broader patterns of change. To this day the Protoclassic Period remains enigmatic within Maya archaeology. There are disagreements on whether or not the term should be used in Maya archaeology and, if used, how and to what the term should refer. Much of what has been used to identify the Protoclassic falls within the realm of ceramics and, thus, that data class will be the pri mary one utilized here. This paper first examines the history of and use of the term Protoclassic in Maya archaeology; it then uses data from Caracol, Belize to assess the relevance of the term both to Maya Studies and to interpretations of ancient Maya society. Introduction A solid chronology of the ancient Maya past is key to outlining the development of the ancient Maya. This chronology is continuously undergoing review and refinement in both the highlands and the lowlands using comparative analysis of individual site chronologies based on ceramics, stratigraphy, and radiocarbon dating. Perhaps the most difficult time to assess is the transition from the Preclassic to Classic Period a time that is also of clear import in assessing the rise and develop ment of Maya civilization. Among the issues relating to the transition from the Preclassic to Classic Periods in the Maya a rea are the relative paucity of excavated Protoclassic remains and preconceptions by researchers about both ceramics and this tempor al era that are not grounded in contextual information. In the highlands there remains disagreement over exactly how the sequences of the various early sites articulate with each other (e.g. Inomata et al. 2014; Love 2017). A large part of this disagreement resides in the nature of the data being used and in how researchers constitute phases and undertake ceramic analysis. While radiocarbon dating is useful in resolving some of these issues, it still needs to be anchored in high quality archaeological dat a (Bayliss 2015). The same chronological issues that are found in the Maya highlands also reverberate in the Maya lowlands and are reflected in the kinds of samples that are used to build chronologies and phases and to model trade linkages. Given the limited hieroglyphic record for the Protoclassic and Early Classic Periods in the Southern Maya lowlands, pottery has generally been used to determine temporal occupation and often these temporal interpretations are derived from a limited sample of archaeologi cally recovered remains. While the total sample of primary deposits containing well dated pottery samples has been increasing each year, Krejci and Culbert (1995: 104) correctly pointed out a quarter century ago that Preclassic and Early Classic contexts in the Southern Lowlands provides a rather slim representation of small structure burials and caches and are far from a balanced sample. They (1995: 114) further argued that the beginning of the Early Classic does not mark a break in ritual patterns, but that the break occurs a century or so later in the mid fourth and early fifth centuries (see also Patino Contreras 2016). In contrast to Krejci and Culberts (1995) assessment, the archaeological data from Caracol, Belize instead suggest a conti nuous development in ritual patterns through the Early Classic Period and indicate that these patterns were not limited to elite contexts, but were present among various levels of society. Thus, the archaeological data from Caracol, Belize not only provide a solid chronological sequence for this transition, but

PAGE 8

Contextualizing the Protoclassic and Early Classic Periods at Caracol 4 they also significantly add to the contextually collected sample, enabling better interpretations. Although the ceramic modes that constituted the Protoclassic had a broad distribution (e.g., Pring 1977), most past assessments of the Protoclassic and Early Classic eras were often largely dependent on whole vessels from burials and tombs derived from limited excavation loci, sometimes only a single structure at any one site. Because the pottery vesse ls within these burials contained a variety of exotic ceramic forms and decorative modes that were not well represented in the sherd material from general excavations, they were often viewed as being eliterelated (e.g., Culbert 1977; Callaghan 2013: 311; Callaghan and Nievens des Estrada 2016:209210) or ritual specific (ReeseTaylor and Walker 2002:102), meaning that how they articulated with the rest of society was fairly unclear (see Lincoln 1985 for initial discussion of Preclassic ceramics in Early Classic contexts). Protoclassic ceramics first had been organized as a category by George Vaillant (1927, 1935) in relation to what he referred to as the Q Complex, which was viewed as having origins in Central or South America and as comprising the introduction of polychrome, mammiform tetrapod feet on vessels, ring and annular bases, pot stands, and spouted vessels into the Maya lowlands. The first published reference that recognized the early nature of these materials in the Maya area was in 1931 an d related to four burials excavated in several residential groups in the Mountain Cow region of Caracol (Thompson 1931), although Gann (1918: Plate 13b) had previously published a complete mammiform tetrapod from Santa Rita Corozal in 1918. A year later, a large sample of Protoclassic and Early Classic transitional deposits, originally recovered by Merwin in 1912 in Structure B of Group II at Holmul, were published (Merwin and Vaillant 1932); because Merwin had died and Vaillant wrote up the final publishe d report from notes, there were unresolved issues in the interpretation of these materials in terms of their dating, seriation, and meaning (e.g., Hammond 1984; Callagan 2013). An extensive deposit of ceramic vessels relating to this temporal era was also recovered at Nohmul, Belize, unfortunately from a single building that was devoid of real context because of bulldozing (Anderson and Cook 1944) and, again leading to questions of dating, seriation, and interpretation (Hammond 1984). An early tomb exca vated at the Belizean site of Pomona added grist to the discussion (Kidder and Eckholm 1951). Gordon Willeys excavations at Barton Ramie in the early 1950s recovered four Protoclassic burials and led to an interpretation of these ceramics as having resul ted from a migration of peoples into the Southern lowlands from the Pacific Coast of El Salvador (Willey and Gifford 1961; Sharer and Gifford 1970), something now considered unlikely (Demarest and Sharer 1986). In 1955 Robert Smith (1955: 2223) segmented the Early Classic into three parts at Uaxactun based on the presence of specific vessel forms: a z angle bowl for Tzakol 1; a basal flanged bowl for Tzakol 2; and, a tripod cylinder for Tzakol 3; he had originally defined a Protoclassic phase called Matzan el (between Chicanel and Tzakol), based on Merwin and Vaillants [1932] Homul data, but after analysis decided that the Uaxactun ceramics did not support its existence (believing that it had just not been well sampled in the Carnegie Institution excavations at the site). The University of Pennsylvania excavations at Tikal also did not recover a detailed sequence of these expected deposits (e.g. Culbert 1993), but such remains were recovered in subsequent excavations undertaken by Juan Pedro Laporte (1995; Laporte and Fialko 1987, 1995) in Tikals Lost World Complex. Ritual ceramics associated with Naj Tunich Cave in Guatemala also proved to be largely of Protoclassic and Early Classic date (Brady et al. 1998) Finally, several more recently excavated inte rments from various sites in northern Belize have provided significant ceramic associations (e.g., Guderjan et al. 2014; Houk and Valdez 2011; Houk et al. 2010; Kosakowsky et al. 2016; Sullivan and Valdez 2006); other materials have come from Nakum, Guatem ala (Zralka et al. 2014). These combined data continue to show that there are major issues in archaeological sampling for this temporal era. The history of the Early Classic Period in the Maya Southern lowlands is one of relatively small population

PAGE 9

A.F. Chase and D. Z. Chase 5 levels (see Culbert and Rice 1990), likely the result of the collapse of early Preclassic states in the northern Peten (Hansen 2015). While populations increased over time, there are far fewer primary deposits to recover when compared to the omnipresent Late Cl assic Period; additionally, many of these earlier deposits have been transposed and redeposited by later activities; thus, the smaller recovered samples have led to difficulties in characterizing the transition from the Late Preclassic into the Early Classic Period. Because of the longevity of the Caracol Archaeological Project (D. Chase and A. Chase 2015, 2017), however, a substantial sample of archaeological materials from 56 primary contexts have been collected from throughout the site. Caracol primary deposits are spatially widespread and cover the entire Late Preclassic through Early Classic Periods. Caracol Sample The Caracol Archaeological Project has recovered 19 caches (non finger bowl), 38 burials, and 2 other contexts consisting of secondary refuse that contain either Protoclassic or Early Classic ceramics. In conjunction with Thompsons (1931) Mountain Cow materials, th is sample permits a firm understanding of the sites Early Classic Period and the ceramic forms and modes that have traditionally been used to understand these temporal eras. These deposits may be dated to between A.D. 150 and A.D. 500 and reveal a fairly continuous ceramic development and one that appears not to be restricted to a single segment of Maya society. On the earlier end of this sequence are two burials that date to approximately A.D. 150 that may be characterized as a Late Preclassic expression of the Protoclassic, following the division suggested by Brady et al. (1998). One of these Late Preclassic transitional deposits (S.D. C117B 5) contained the skeletal remains of a female interred in an Ix Chel diety costume (Rich 2003) accompanied by a w ide variety of goods, including 2 pottery figurines (human whistle and armadillo), 32 ceramic vessels, and over 7000 shell and jadeite beads sewn onto a mantle fringed with dog teeth (A. Chase and D. Chase 2006). Stylistically, the vessels included within this interment included 4 incipient polychrome bowls (2 with ring bases), 2 tetrapod jars (one with Usulatanstyle decoration), 1 tetrapod bowl, 6 miniature vessels (2 with tetrapods), 1 large dish, 1 large jar, 15 labial flanged bowls, and 2 resist composite angle bowls (A. Chase and D. Chase 2006: fig.1). The composite angle bowls are similar to others known from Nohmul, Belize (Hammond 1984, vessel 17). The second Caracol deposit (S.D. C52A 1) comes from a chultun burial located approximately 3 km dis tant from the site epicenter; the chultun burial was associated with 6 vessels (A. Chase 1994: fig. 13.3). Two of these vessels were decorated with Usulutan style wavy line decoration. One of these vessels had foreshortened mammiform tetrapod supports and a grooved hook rim; one vessel had a labial flange; three were rounded bottom bowls; and, the last was an elaborately incised deep bowl with its 4 tetrapod supports removed in antiquity. The rounded bowl form seen in this deposit continues into the Early Classic era and occurs in 3 later deposits that span the Early Classic Period (C14C/2; C14C/4; and C10A/1). The early part of the Early Classic Period at Caracol is characterized by bowls or plates with large tetrapod feet and the ap pearance of orange wares and true polychromes, as well as the persistence of Sierra Red slip on these new forms. Deposits containing these materials were initially found by Thompson (1931) in a vaulted tomb and in three chultuns during his excavations in the Mountain Cow part of the site. Vessels placed within one chultun interment included a Sierra Red mammiform tetrapod, a ring base orange bowl with black pseudoUsulutan decoration on its interior, and a small decorated jar with a circumferential incisi on on its interior lip. The Cahal Cunil vaulted chamber 1 excavated by Thompson (1931) similarly contained 2 Sierra Red tetrapod bowls, 2 decorated jars with handles, a bowl with an annular base (similar to one from Holmul; Callaghan 2013: fig. 7a), a tet rapod redslipped bowl containing the modeled image of a frog (see analogous vessel in Bonnafoux 2008: fig. 6.3e), and a miniature buff color jar with incisions (similar to one illustrated in refuse of a similar date to the west of Caana; A. Chase and D. C hase 2016: fig. 106a). The presence of a Sierra Red mammiform tetrapod in an early

PAGE 10

Contextualizing the Protoclassic and Early Classic Periods at Caracol 6 Figure 1 Associated vessels from a bedrock interment (S.D. C121C-5) in Caracol Structure F24: a. Sierra Red; b. probably Corriental Appliqued; c. possibly Xtabcab Inc ised; d. Guacamallo Red -on-Orange. Early Classic burial also occurs in a residential group outside of the site epicenter ( Figure 1 ), where this form is associated with an orangeslipped potstand, a shoe pot (see Brady 1992), and a small jar with circumfer ential incision on its interior lip. A second chultun excavated by Thompson (1931) in the Mountain Cow region yielded an orange ware polychrome tetrapod bowl, and orange ware polychrome collared jar, and a potstand that was once stuccoed and painted. A t hird, and final, chultun in the Mountain Cow region also yielded a tetrapod orange ware polychrome plate, a red on orange pot stand, a large hemispherical orangeware polychrome bowl, and a large decorated jar. Two other orange ware polychrome tetrapods ( both with feet removed) are known from Caracol deposits: one is from Tulaktuhebe (C14C/2), 3.5 km southeast of Caracols epicenter, associated with a redware deep dish, both from a looted tomb; the second is associated with a handled and decorated jar with Figure 2 Associated vessels from a cache (S.D. C171C -1) located immediate south of a re-entered tomb in Caracol Structure B40: a. Ixcanrio Orange Polychrome; b. possibly Corriental Appliqued. a circumferential lip groove and was recovered in associ ation with a tomb in a residential group just southeast of Caracols epicentral C Group ( Figure 2). Isolated burials that were likely associated with this earliest expression of the Early Classic have also been widely recovered at Caracol. A burial recove red in a plaza immediately west of Caana yielded a decorated handled jar with interior incised lip associated with a decorated collared bowl. A burial southwest of the Central Acropolis yielded a ringbase collared jar that is Ixcanrio Polychrome. One ot her form

PAGE 11

A.F. Chase and D. Z. Chase 7 associated with the earliest Early Classic at other sites in the Maya area includes z angled vessels (see Smith 1955 for Uaxactun). For Caracol, this form has been recovered in a looted deposit at Tulakatuhebe as well as in a residential plaza 4 km northeast of the site epicenter ( Figure 3); however, this form occurs in isolation and is not directly associated with any of the other vessel forms. A final burial from immediately west of Caana yielded an Actuncan Polychrome basal flanged bowl in association with a polychrome pot stand (A. Chase and D. Chase 2005a: fig.3c,d), transitional to the middle facet of the Early Classic at the site. Interestingly, while they are prevalent in deposits at Holmul (Callaghan 2013) and in northern Belize at sites like Nohmul and Santa Rita Corozal (D. Chase and A. Chase 2006), no chocolate pots have been found in any of the Protoclassic or Early Classic Caracol deposits; the only one known was recovered from a Late Preclassic chultun burial that precedes this temporal era (A. Chase and D. Chase 2011a: fig. 13a). The appearance of polychrome basal flanged bowls at Caracol appears to mark the next evolution of pottery subassemblages at the site. Basal flanged bowls are present i n a wide variety of contexts at the site, having been recovered in 19 burials. The Caracol sample also attests to the lack of overlap between polychrome basal flanged bowls with tetrapod ceramic plates, something suspected but not demonstrated elsewhere. At both Homul (Callaghan 2013) and Nohmul (Hammond 1984), they are seriated as being later, but tetrapod hemispherical bowls can co occur with basal flanged bowls as documented in contexts at Uaxactun (Smith 1955: figs. 3e and 12s) and Kaxob (Berry et al 2004: 256257). Yet, it is clear that the basal flange bowl form dominates the middle of the Early Classic Period and is likely derived from the labial flanged bowl form of the Late Preclassic Period. Sierra Red basal flange bowls have been recovered i n tombs at Chanchich (Sullivan and Valdez 2006) and Pomona (Kidder and Eckholm 1951) with tetrapod plates, but polychrome basal flange bowls appear to supplant the tetrapod plate as part of the Caracol burial assemblage. There are six burials in residentia l groups in which basal flange bowls constitute the only Figure 3 Caracol z-angle bowls from a looted tomb in Caracol Structure 8F8 (C14C/15) and from a plaza excavation associated with Caracol Structure 4T17 (C129C/2): a. probably Boleto Black -on-Orange; b. eroded Aguila Orange. pottery vessel included. In other deposits basal flange bowls co occur with shoe pots (e.g., A. Chase and D. Chase 2005a: fig. 6) and in one a basal flange bowl cooccurred with cylinder tripods (e.g., A. Chase 1994: fig. 13.4). They are also associated with hour glass censers in several interments. Another vessel form that appears to be introduced at the same time as the basal flange bowl is a spouted bowl or jar; this form has been recovered from 5 Early Clas sic interments. The previously reported cremation from Caracols Northeast Acropolis, believed to represent an individual from Teotihuacan (A. Chase and D. Chase 2011b), contained vessels that are transitional between the Late Preclassic and Early Classic Period as well as two nubbinfooted tripod vases that resemble one assigned to the earlier Protoclassic at Nohmul (Hammond 1984, vessel 14). Seven flaring rim bowls were also found in association with a basal flanged bowl and a spouted jar in a tomb in St ructure D8 in the South Acropolis; three flaring rim bowls were in association with a basal flanged bowl and two spouted bowls (one potentially lidded) from an infilled tomb west of Caana. The Northeast Acropolis cremation also contained 7 basal flanged bowls (A. Chase and D. Chase 2011b: fig. 3), 4 of which portrayed a reclining,

PAGE 12

Contextualizing the Protoclassic and Early Classic Periods at Caracol 8 Figure 4 Associated vessels from a tomb (S.D. C181B -1) in Structure B33 in the Northeast Acropolis: a. Dos Arroyos Orange Polychrome; b. undesignated; c., d. Pucte Brown. possibly bound, individual on the bowl exterior. The imagery of this reclin ing figure is widespread, occurring on basal flange bowls from other sites, such as Holmul (Callaghan 2013: fig. 22a), Uaxactun (Smith 1955: fig. 3e), Dos Hombres (Houk and Valdez 2011: fig. 4) and Batsub Cave (Prufer and Dunham 2009: fig. 4). At Caracol a similar basal flanged vessel with a reclining individual occurs in a tomb in Structure A33 in the Northeast Acropolis in association with a miniature vessel, a spouted bowl, and a blackware goblet with a tubular base ( Figure 4 ); this basal flanged vessel contains an interior hummingbird image that is almost identical to one recovered at Batsub Cave (Prufer and Dunham 2009: fig. 4). The blackware goblet is similar to other ones recovered in Burial 177 at Tikal (Culbert 1993: fig. 37b1) and in Burial P2B 2 at Santa Rita Corozal (D. Chase and A. Chase 2005: fig. 5); this form may derive from the combination of an incurved bowl on a pot stand, as can be seen at both Holmul (Callaghan 2013: Fig. 24a), and at Nohmul (Hammond 1984, vessel 3). A residential tomb fleshes out some of the other vessel possibilities for the middle part of the Early Classic at Caracol; besides a miniature cup with face and a basal flanged bowl, the tomb (S.D. C95A 1; see A. Chase and D. Chase 2005a: fig. 4) also cont ained a large jar, a miniature jar with ring base, an inverted goblet with bird handle (similar to forms at Tikal [Laporte and Fialko 1995: fig. 30] and Holmul [Callaghan 2013: fig.24b]), and a tripod bowl with modeled peccary feet and three incised deity heads. Of the nine known burials with tripod cylinders at Caracol, only one is associated with a basal flange bowl (this same tomb also is associated with a spouted vessel; see A. Chase 1994: 167169). Four of the burials with cylinder tripods occur in th e epicenter of the site and the other five are associated with residential groups. Three different interments with cylinder tripods were recovered from an excavation into Structure C47, approximately 600 m south of the epicenter (A. Chase and D. Chase 2014). The earliest deposit was a tomb that contained a series of smaller artifacts (including a 16.4 cm long jadeite tube) and 8 ceramic vessels ( Figure 5), 2 of which were cylinder tripods. Perhaps the most spectacular vessel in this residential tomb was a brownware bowl with 6 incised glyphic cartouches (A. Chase and D. Chase 2014: figs. 122a and 123). The glyphs are of an early style but may refer to an early form of a

PAGE 13

A.F. Chase and D. Z. Chase 9 Figure 5 Associated vessels from a tomb (S.D. C203B 16) in Caracol Structure C 47: a. Lucha Incised; b., c. Caldero Buff -Polychrome; d. Quintal Unslipped; e., f., g. Pucte Brown; h. possibly Candelario Appliqued. primary standard sequence, iconographically signal creation mythology; they also appear to document the unknown site of Bital (see also A.Chase et al. 1991:10). Other vessels in this deposit included a polychrome jar, a tripod footed hemispherical bowl with a bird in its interior (similar in form and type to vessels in a tomb and an interment at Tikal; see Culbert 1993: f igs. 29c g and 32c), and a truncated black goblet. The tomb was also directly associated with a hidden Early Classic cache that contained a small lidded urn inside two lip to lip vessels; inside the urn were 2 shell Charlie Chaplins, 1 drilled flamingo to ngue shell, 3 beads (one each of shell, bone, and jadeite), and 2 jadeite chips. The second crypt in Structure C47 had clearly been re entered, as indicated by the inclusion of 14 finger bowls in the fill of the burial, but contained 5 Early Classic vessels: 2 cylinder tripods, 1 basal flanged bowl, 1 deep bowl, and 1 shoe pot ( Figure 6 ). An extensive caching deposit that included part of a large ceramic figurine among the finger bowls had been placed above the third Early Classic crypt, signaling a Figure 6 Associated vessels from a tomb (S.D. C203B 14) in Caracol Structure C47: a. Aguila Orange; b. eroded Dos Arroyos Orange -Polychrome; c. possibly Pucte Brown; d. eroded Saxche Orange-Polychrome; e. probably Corriental Appliqued. re entry here as well. This third Early Classic burial was associated with 1 cylinder tripod, 4 bowls, and 5 dishes; the upper portion of this infilled crypt had been used to place a Late Classic burial with two pottery vessels accompanied by 2 small cache pots. A very l ate polychrome lidded cylinder tripod with 3 bulbous feet was recovered in Structure B42 in a residential group in association with 5 polychrome bowls and 3 lateral flanged ringbase dishes ( Figure 7 ; A. Chase and D. Chase 2005b: fig. 18); these materials are transitional into the Late Classic Period. A large number of Early Classic caches have also been recovered both in the epicenter and in residential groups. This includes the one hidden in the tomb wall mentioned above; two other lidded urns of Early C lassic date were recovered above this same residential tomb.

PAGE 14

Contextualizing the Protoclassic and Early Classic Periods at Caracol 10 Figure 7 Associated vessels from a tomb (S.D. C171B -9) in Structure B42: a., c., e., f., g. Saxche Orange-Polychrome; b., d., i. Pajarito OrangePolychrome; c. Veracal Orange. Early Classic urns associated with Charlie Chaplin figures have been recovered from 3 widely spaced residential groups (e.g., A. Chase and D. Chase 2006:44; Lomitola 2012) as well as from four different structures in the site epice nter (e.g., A. Chase and D. Chase 2005a:31). It is suspected that an urn recovered in Tulaktuhebe with a painted Principle Bird deity on its interior lid and a dead corn god on its interior base also dates to the Early Classic (A. Chase and D. Chase 1987: fig. 41a,b,c). Another Early Classic deposit from Structure D1 consisted of the burnt remains of 14 ceramic vessels (10 large Aguila Orange flaring walled bowls and 4 polychrome ring based dishes) in association with 2 limestone bars, 1 partial jadeite b ead, 1 polished piece of jadeite, 16 obsidian lancets, and 25 obsidian blade fragments (A. Chase and D. Chase 2007: figs. 8183). An even more spectacular cache consisted of an Early Classic tripod cylinder with a polychrome scene of 3 figures (figure wit h feather offerings; prisoner; warrior) on the cylinder and a polychrome lid with a deity head handle ( Figure 8 ); this vessel, excavated during the 2017 field season in Structure I28, had been

PAGE 15

A.F. Chase and D. Z. Chase 11 Figure 8 Early Classic cylinder tripod used as a cache vessel (S.D. C210B 3) in Structure I28: cylinder probably Dos Arroyos Orange -Polychrome; lid a polychrome variant of Positas Modeled. re purposed in the fill of a Late Classic construction. Conclusion Three inter linked issues have hampered a full understanding of the Late Preclassic to Early Classic transition: the interpretation of exotic ceramic forms and decorations; preconceived temporal barriers; and, limited sampling. The exotic forms and decorative modes led to an early ceramic definition of the Q Complex, interpreted as a set of specific ceramic forms introduced to the Maya area from elsewhere (Vaillant 1924). The limited occurrences of this complex led to its association with Maya elite and an early consideration of these materials as being classlinked (e.g., Lincoln 1985), thus providing an easy explanation for why so few primary deposits have been excavated. The Protoclassic was a time of great experimentation in ceramic forms, decorations, and slips that crossed perceived temporal boundaries. Because of limited contextualized and stratified deposits, materials found in secondary fill contexts were often preassigned to temporal associations, thus aggravating any attempt to better understand ceramic traditions. Although the Late Preclassic type Sierra Red was recognized as extending into the Early Classic and orange wares were recognized as existing in the Late Preclassic relatively early (e.g., A. Chase and D. Chase 1983; Ciudad Ruiz 1988:95; Graham 1986; Kosakowsky 1982: 3435, 1987:82), the formal sorting of sherds from secondary fill contexts tended to reify the rigid boundary that was perceived between the Late Preclassic and the Early Classic Periods, as there was no way to be su re that the correct temporal frame was selected. Thus, specific ceramic forms were provided with inferred temporal meaning, regardless of context and associations. The true transition from Late Preclassic to Early Classic was a palimpset of ceramic forms and decorative modes. While there is some temporal faceting (Brady et al. 1998), it is largely a single temporal period with great fluidity in ceramic forms and decorations. New ceramic trends occur at the end of the Late Preclassic Period that become mo re codified in the early part of the Early Classic Period with the introduction of true polychrome and large mammiform tetrapod dishes, plates, and deep bowls. Sierra Red versions of these vessel forms also occur in the early part of the Early Classic Per iod. The middle part of the Early Classic Period is also characterized by a wide diversity of ceramic forms, but the prominent form is a basal flanged bowl. What the data do suggest is

PAGE 16

Contextualizing the Protoclassic and Early Classic Periods at Caracol 12 that mammiform tetrapod plates were replaced in burial contexts by ba sal flanged bowls. Cylinder tripods characterize the latest part of the Early Classic Period, especially in monochrome slipped form. However, more elaborately decorated cylinder tripods appear earlier and are not restricted to the latest part of the Earl y Classic Period. Thus, the latter half of the Early Classic does not break with ritual traditions as was argued for by Krejci and Culbert (1995: 114). The distribution of the Caracol Protoclassic and Early Classic ceramic forms indicates that these items were widely available to the inhabitants of the site and were not restricted in their distribution. However, there is a gradient in status and wealth that can be seen in the data. The Caracol Late Preclassic burial that contained 32 vessels is the richest known interment for this temporal era in the Southern Maya Lowlands (based on data in Krejci and Culbert 1995). The upper tomb in front of Structure A6 contained 26 vessels dating to the later part of the Early Classic Period, signaling the wealth of its occupant(s). Both of these interments indicate that the highest elite were likely associated with the Caracol epicenter. But, the presence of Protoclassic and Early Classic ceramics in special deposits throughout the site also suggests that these it ems were generally available to the rest of the population and not restricted in distribution. Significantly, residential groups that were occupied in the Early Classic Period continued to be utilized into the Late Classic Period and the mix of ceramics f rom interments placed during this later interface also evince a great fluidity in ceramic forms, similar to what occurred during the earlier conversion. Thus, the archaeological data from Caracol help to demystify the transition from the Late Preclassic t hrough the Early Classic Periods. References Anderson, A. Hamilton and H. J. Cook 1944 Archaeological Findings near Douglas, British Honduras. Notes on Middle American Archaeology and Ethnology 40. Carnegie Institution of Washington, Washington D.C. Bayliss, Alex 2015 Quality in Bayesian Chronological Models in Archaeology. World Archaeology 47: 677700. Berry, Kimberly A., Sandra L. Lopez Varela, Mary Lee Bartlett, Tamarra Martz, and Patricia A. McAnany 2004 Pottery Vessels of Kaxob. In P.A. McAnan y, Ed. Kaxob: Ritual, Work, and Family in an Ancient Maya Village pp. 193 261. Cotsen Institute of Archaeology Monumenta Archaeologica 22. University of California, Los Angeles. Bonnafoux, Patrice 2008 Etude Iconographique des Ceramiques du Classique A ncien dans les Basses Terres Mayas. Ph.D. dissertation. Universite de Paris I Pantheon Sorbonne, Paris. Brady, James E. 1992 Function and Meaning of Lowland Maya Shoe Pots. Ceramica de Cultura Maya 16: 1-10. Brady, James E., Joseph Ball, Ronald Bishop, Duncan Pring, Norman Hammond, and Rupert Housley 1998 The Lowland Maya Protoclassic: A Reconsideration of its Nature and Significance. Ancient Mesoamerica 9: 1738. Callaghan, Michael G. 2013 Politics through P ottery: A View of the Preclassic-Classic Transition from Building B, Group II, Holmul, Guatemala. Ancient Mesoamerica 24: 307-341. Callaghan, Michael G. and Nina Nievens de Estrada 2016 The Ceramic Sequence of the Holmul Region, Guatemala. Anthropological Paper 77. The University of Arizona Press, Tucson. Chase, Arlen F. 1994 A Contextual Approach to the Ceramics of Caracol, Belize. In D.Z. Chase and A.F. Chase, Eds. Studies in the Archaeology of Caracol, Belize pp. 157-182. Monograph 7. Pre -Columbian Ar t Research Institute, San Francisco. Chase, Arlen F. and Diane Z. Chase 1983 La Ceramica de la Zona Tayasal -Paxcaman, Lago Peten Itza, Guatemala. Privately distributed by The University Museum, University of Pennsylvania, Philadelphia. Available at www.caracol.org 1987 Investigations at the Classic Maya City of Caracol, Belize: 1985-1987. Monograph 3. Pre Columbian Art Research Institute, San Francisco. 2005a The Early Classic Period at Caracol, Belize: Transitions Complexity, and Methodological Issues in Maya Archaeology. Research Reports in Belizean Archaeology 2: 17 -38.

PAGE 17

A.F. Chase and D. Z. Chase 13 2005b Searching for Caracols Last Urbanites: Continued Investigation of Small Structures in and near Caracols Epicenter: 2005 Field Report of Caracol Archaeological Project. On -file with Belize Institute of Archaeology and available at www.caracol.org 2006 Before the Boom: Caracols Preclassic Era. Research Reports in Belizean Archaeology 3: 41 57. 2007 Late Classic Ritual Variation in a Maya Community: Continued Investigation of Structures in and near Caracols Epicenter: 2007 Field Report of the Caracol Archaeological Project. On -file with Belize Institute of Archaeology and available at www.caracol.org 2011a Heterogeneity in Residential Composition: Continued Investigation in and near Caracols Epicenter: Caracol Archaeological Project Investigations for 2011. Onfile with Belize Institute of Archaeology and ava ilable at www.caracol.org 2011b Status and Power: Caracol, Teotihuacan, and the Early Classic Maya World. Research Reports in Belizean Archaeology 8: 3 18. 2014 Ancient Social Integration in a Maya Neighborhood: Investigation of Adjacent Residential Complexes near Caracols Epicenter: Caracol Archaeological Project Investigations for 2014: A Continuation of the 2012 and 2013 Research Focus. On -file with Belize Institute of Archaeology and available at www.caracol.org 2016 Investigating Early Long -Distance Interaction in Caracols Epicenter: Caracol Archaeological Project Investigations for 2016. On-file with Belize Institute of Archaeology and available at www.caracol.org Chase, Arlen F., Nikolai Grube, and Diane Z. Chase 1991 Three Terminal Classic Monuments from Caracol, Belize. Research Reports on Ancient Maya Writing 36. Center for Maya Research, Washington, D.C. Chase, Diane Z. and Arlen F. Chase 2005 The Early Classic Period at Santa Rita Corozal: Issues of Hierarchy, Heterachy, and Stratification in Northern Belize. Research Reports in Belizean Archaeology 2: 111129. 2006 The Dawn of Maya Civilization: Preclassic Period Ar chaeology from Santa Rita Corozal. Research Reports in Belizean Archaeology 3: 85 100. 2015 Thirty Years of Archaeology at Caracol, Belize: Retrospective and Prospective. Research Reports in Belizean Archaeology 12: 3 -14. 2017 Caracol, Belize and Changi ng Perceptions of Ancient Maya Society. Journal of Archaeological Research 25(3): xx -xx. Ciudad Ruiz, Andres 1988 Desarrollo Ceramico en el Alto Salama, Guatemala. Ceramica de Cultura Maya 15: 93 -130. Culbert, T. Patrick 1977 Early Maya Development at Tikal, Guatemala. In R.E.W. Adams, Ed. The Origins of Maya Civilization, pp. 2743. University of New Mexico Press, Albuquerque. 1993 The Ceramics of Tikal: Vessels from the Burials, Caches, and Problematic Deposits. Tikal Report 25A. University Museum Monograph 81. University of Pennsylvania, Philadelphia. Culbert, T. Patrick and Don S. Rice 1990 Eds., Precolumbian Population History in the Maya Lowlands University of New Mexico Press, Albuquerque. Demarest, Arthur A. and Robe rt J. Sharer 1986 Late Preclassic Ceramic Spheres, Culture Areas, and Cultural Evolution in the Southeastern Highlands of Mesoamerica. In P. Urban and E. Schortman, Eds. The Southeast Maya Periphery pp. 194223. University of Texas Press, Austin. Graham Elizabeth 1986 Barton Ramie Ceramic Types at Colson Point, North Stann Creek: A Focus on the Protoclassic. Ceramica de Cultura Maya 14: 32-48. Guderjan, Thomas H., Steven Bozarth, David Glassman, Robert Lichtenstein, and Norbert Stanchly 2014 Mortuary Ritual in the Terminal Preclassic: Evidence from the Maya Site of Blue Creek in Northern Belize. Research Reports in Belizean Archaeology 11: 347359. Hammond, Norman 1984 Holmul and Nohmul: A Comparison and Assessment of Two Maya Lowland Protoclassic Sites. Ceramica de Cultura Maya 13: 117. Hansen, Richard D. 2015 Cultural and Environmental Components of the First Maya States: A Perspective from the Central and Southern Maya Lowlands. In L.P. Traxler and R.J. Sharer, Eds. The Origins of Maya States pp. 329 -416. University of Pennsylvania Museum of Archaeology and Anthropology, Philadelphia.

PAGE 18

Contextualizing the Protoclassic and Early Classic Periods at Caracol 14 Houk, Brett A. and Fred Valdez, Jr. 2011 The Precocious Dead: Status, Power, and Early Tombs in the Eastern Three Rivers Region. Research Reports in Belizean Archaeology 8: 151158. Houk, Brett A., Hubert R. Robichaux, and Fred Valdez, Jr. 2010 An Early Royal Tomb from Chan Chich, Belize. Ancient Mesoamerica 21: 229 -248. Inomata, Takeshi, Raul Ortiz, Barbara Arroyo, and Eugenia J. Robinson 2014 Chronological Revision of Preclassic Kaminaljuyu, Guatemala: Implications for Social Process in the Southern Maya Area. Latin American Antiquity 25:377 -408. Kidder, Alfred V. and Gordon F. Eckholm 1951 Some Archaeological Specimens from Pom ona, British Honduras. Notes on Middle American Archaeology and Ethnology 102. Carnegie Institution of Washington, Washington, D.C. Kosakowsky, Laura J. 1982 A Preliminary Summary of Formative Ceramic Variability at Cuello, Belize. Ceramica de Cultura May a 12: 26 -42. 1987 Preclassic Maya Pottery at Cuello, Belize. Anthropological Papers of the University of Arizona No. 47. University of Arizona Press, Tucson. Kosakowsky, Laura J., Robin Robertson, and Debra S. Walker 2016 Figures and Tables for the Ceram ics from a Terminal Preclassic Chultun Style Burial from the Site of Blue Creek, Northern Belize. Research Reports in Belizean Archaeology 13: 309-316. Krejci, Estella and T. Patrick Culbert 1995 Preclassic and Classic Burials and Caches in the Maya Lowlands. In N. Grube, Ed. The Emergence of Lowland Maya Civilization (Acta Mesoamericana 8), pp. 103-116. Verlag Anton Saurwein, Mockmuhl. Laporte, Juan Pedro 1995 Preclasico a Clasico en Tikal: Proceso de Transformacion en Mundo Perdido. In N. Grube, Ed. The Emergence of Lowland Maya Civilization (Acta Mesoamericana 8), pp. 17 -33. Verlag Anton Saurwein, Mockmuhl. Laporte, Juan Pedro and Vilma Fialko 1987 La Ceramica del Clasico Temprano desde Mundo Perdido, Tikal: Una Reevaluacion. In P.M. Rice and R.J. S harer, Eds. Maya Ceramics: Papers from the 1985 Maya Ceramic Conference, pp. 123 181. BAR International Series 345(i), Cambridge. 1995 Un Reencuentro con Mundo Perdido, Tikal, Guatemala. Ancient Mesoamerica 6: 4194. Lincoln, Charles E. 1985 Ceramics and Ceramic Chronology. In G.R. Willey and P. Mathews, Eds. A Consideration of the Early Classic Period in the Maya Lowlands pp. 5594. Institute for Mesoamerican Studies Publication 10, SUNY, Albany. L omitola, Lisa M. 2012 Ritual Use of the Human Form: A Contextual Analysis of Charlie Chaplin Figures in the Maya Lowlands. M.A. Thesis, Department of Anthropology, University of Central Florida, Orlando. Love, Michael 2017 Kaminaljuyu Chronology and Ceramic Analysis: An Alternative View. Latin American An tiquity doi:10.1017/laq.2017.70. Merwin, Raymond E. and George C. Vaillant 1932 The Ruins of Holmul, Guatemala. Memoirs of the Peabody Museum 3(2). Harvard University, Cambridge. Patino -Contreras, Alejandro 2016 Explaining Tzakol: Social Interaction During the Early Classic: A View from Naachtun, Peten, Guatemala. Estudios de Cultura Maya 48:39 -70. Pring, Duncan C. 1977 Influence or Intrusion? The Protoclassic in the Maya Lowlands. In N. Hammond, Ed. Social Process in Maya Prehistory, pp. 135-165. Academic Press, London. Prufer, Keith M. and Peter S. Dunham 2009 A Shamans Burial from an Early Classic Cave in the Maya Mountains of Belize, Central America. World Archaeology 41(2): 295320. ReeseTaylor, Kathryn and Debra S. Walker 2002 The Passage of the Late Preclassic into the Early Classic. In M.A. Masson and D.A. Freidel, Eds. Ancient Maya Political Economies pp. 87 122. AltaMira Press, Walnut Creek. Rich (Brown), Shayna 2003 An Analysis of a Female Protoclassic Costume from the Site of Caracol, Belize. M.A. Thesis, Maya Studies / Liberal Studies, University of Central Florida. Sharer, Robert J. and James C. Gifford 1970 Preclassic Ceramics from Chalchuapa, El Salvador and Their Relationships with the Lowland Maya. American Antiquity 35: 441 -462.

PAGE 19

A.F. Chase and D. Z. Chase 15 Smith, Robert E. 1955 Ceramic Sequence at Uaxactun, Guatemala. 2 vols. Publication 20, Middle American Research Institute. Tulane University, New Orleans. Sullivan, Lauren A. and Fred Valdez, Jr. 2006 The Late Preclassic to Early Classic Transiti on in the Three Rivers Region. Research Reports in Belizean Archaeology 3: 73 -84. Thompson, J. Eric S. 1931 Archaeological Investigations in the Southern Cayo District, British Honduras. Publication 301, Anthropological Series XVII(3). Field Museum of Natural History, Chicago. Vaillant, George C. 1927 The Chronological Significance of Maya Ceramics. Ph.D. Dissertation. Harvard University, Cambridge. 1935 Chronology and Stratigraphy in the Maya Area. Maya Research 2(1): 119-143. Willey, Gordon R. and James C. Gifford 1961 Pottery of Holmul I Style from Barton Ramie, British Honduras. In S.K. Lothrop, Ed. Essays in Pre -Columbian Art and Archaeology pp. 152 170. Oxford University Press, London. Zralka, Jaroslaw, Wieslaw Koszkul, Katarzyna Radnicka, Lau ra Elena Sotelo Santos, and Bernard Hermes 2014 Excavations in Nakum Structure 99: New Data on Protoclassic Rituals and Pecolumbian Maya Beekeeping. Estudios de Cultura Maya 44: 85 -117.

PAGE 20

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 1726 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 2 EXPANDING SUB -PLAZA EXPLORATIONS OF MIDDLE PRECLASSIC ARCHITECT URE AT THE SITE OF P ACBITUN, BELIZE George J. Micheletti Kaitlin E. Crow and Terry G. Powis Since the mid 1990s, large -scale excavations have focused on documenting the Middle Preclassic period at the ancient Maya center of Pacbitun. Our investigations at Pacbitun have continued to reveal supporting evidence of earlier development of social complexity than was previously conceived. Pacbituns Middle Preclassic saga begins in t he 9th century BC with the construction of apsidal structures in Plaza B. By the late Middle Preclassic (600 300 BC), these initial platforms were replaced by sturdier rectangular architecture. Through this transition, both forms of architecture are thou ght to function as workshops where an intensified and increasingly standardized production of shell beads had developed from the previous facet. Occurring simultaneous to Pacbituns successful economic enterprise, the construction of a large plastered platform built in the adjacent elevated plaza, Plaza A, reveals the development of public ritual at the site. The recent discovery of this late Middle Preclassic platform, El Quemado, has added the final sociopolitical element to support Pacbituns early adv ancement in the Middle Preclassic period. Our investigations during the 2017 field season aimed to gather more information on the installation and progression of the social, economic, and political institutions as we attempted to better understand the rel ationship between each of them. Introduction The Preclassic period for the Ancient Maya is thought to have brought about an era of drastic social development. During the Early Preclassic (1800 1000 BC), groups of people began to establish residences in the interior of the Southern Maya Lowlands. Sustained agricu ltural settlements of the Middle Preclassic (1000400 BC) period would eventually lead to population growth and the emergence of observable societies increasing in complexity. At the ancient Maya site of Pacbitun, Belize, the mechanisms of advancing compl exity in the Middle Preclassic are present in the form of utilitarian and ceremonial architecture, and the evolution of a budding economic enterprise. Nearly three decades of investigations have worked to unearth a modest Middle Preclassic community which lay hidden beneath the central plazas of Pacbituns site core. The initial discovery of this early community in the 1990s found several rudimentary platforms buried beneath Plaza B. Modifications to these structures amid the Middle Preclassic suggest tha t Pacbitun underwent a significant architectural and orientational transformation one which helps to divide this period into an early and late facet. Found amongst these structures, a myriad of shell fragments and lithic tools associated with shell bead production is indicative of an organized economic establishment at Pacbitun. Although our initial perception believed this early community to be simple in terms of social complexity, it would be forever altered after the more recent discovery of El Quemad o (Q), a large ceremonial platform found buried beneath the northern end of Plaza A. Excavations dating back to 2013 have systematically unearthed the massive platform in an effort to reveal the structures configuration and architectural design. With a better understanding of Q, we believe the structure has the potential to not only supplement what we know about the socioeconomic affairs of Pacbitun, but also introduce a previously unknown ceremonial dimension that may hint at the advent of sociopolitica l organization at the site during the Middle Preclassic period. However, before we can fully comprehend these larger social concepts, there are impending enquiries concerning both plazas, and the Middle Preclassic architecture that need s to be resolved. R egarding Plaza B, considerable evidence has been discerned about the production aspect of this early community space; yet, very little information has been found to determine the residential status of the sub plaza structures. Were these platforms solely used for manufacturing shell beads, or did they also

PAGE 21

Expanding Sub -Plaza Explorations of Middle Preclassic Pacbitun 18 function as domestic space? In Plaza A, the inimitable nature of El Quemado is justification enough to continue our intensive explorations to enhance what we know about Pacbituns Middle Preclassic peri od. Our thorough investigation of Q has enabled us to study details such as construction methods, and materials used, providing insight into the origins of this platform. Moreover, analyzing the desecration and burial of Q, and the construction of Plaza A, has also contributed information concerning the termination and abandonment of the platform. Unfortunately, this damage has also hindered our efforts to understand the architectural dimensions and plaza configuration and may have destroyed crucial evidence concerning the structures purpose. With little to no evidence indicating the ceremonial function or political meaning, the structures plaza position and orientation may prove to be vital clues to understanding its presence at Pacbitun. Thus, in see king answers for the uncertainties in Plaza A and Plaza B, the excavations of the 2017 field season set out to expand our investigations from these previously explored areas associated with Pacbituns Middle Preclassic community. In Plaza B, excavations w ere expanded to the west of the late Mai phase (600 300 BC) platform, SubStructure B 2, to further expose its contemporaneous neighbor, SubStructure B 3. Our preliminary analysis of the artifacts recovered from the 2017 excavation of B 3 suggest this pl atform was also involved in shell bead production and did not exhibit any evidence of a residential component. Beneath Sub Structure B 3, our investigations would also unearth more of the early Mai phase (900600 BC) platforms, SubStructures B 1 and B 4, as well as a third platform (Sub Structure B 16), belonging to the same phase. All three platforms have offered new information concerning the layout and function of these early structures. In Plaza A, explorations were directed towards the unexplored a reas on the west and north sides of the building. At the culmination of our fifth and final year of excavating El Quemado, we were finally able to ascertain the platforms dimensions and confirm the structures southfacing theme after locating the northw est corner and northern facade of the structure. The discovery of another feature on the northside of the structure has also added to what we know about the architectural design of Q. The Site of Pacbitun Pacbitun is a medium sized site located on the so uthern limits of the Belize River Valley. Situated between the foothills of the Maya Mountains and the lowland tropical forest, Pacbitun straddles two ecozones creating a unique contrasting environment that would have offered a multitude of diverse resour ces. Situated around the sites five main plazas (Plaza A E), 41 masonry structures are densely constructed within the 145,000 squar e meter limits of the site core (Healy 1990:250; Figure 1). Plaza A and Plaza B have been identified as the location of Pa cbituns original settlement dating as far back as 900 BC. These two plazas continued to flourish as the communal hub until the sites presumed abandonment in AD 900. Several causeways lead from Pacbituns site core out into a periphery that is laden with house mounds and minor centers found amongst countless agricultural terraces (Healy et al. 2007; Weber and Powis 2010). Karstic landmarks such as sinkholes, rockshelters, bedrock outcrops, and caves are f ound throughout the region, but are most prevalent in the southern and eastern areas of the periphery zone (Spenard 2014; Spenard et al. 2013) The Middle Preclassic Settlement in Plaza B Extensive investigations in and around Pacbituns site core have been ongoing, sporadically, since the 1980s. Early excavations, conducted by Paul Healy of Trent University, primarily studied the monumental buildings in the core that were initially constructed during Pacbituns Puc phase (300100 BC) and modified up unti l the sites fall in the Tzib phase (AD 800900) (Healy 1990; Healy et al. 2004). Though Healy (1990:256) ceramically identified a Mai phase (900 300 BC) occupation at the site, the potential scale and significance of this early inhabitance would not be r ecognized until the mid1990s when multiple subplaza platforms were discovered beneath a thick midden deposit that forms the surface of Plaza B (Hohmann 2002).

PAGE 22

Micheletti, Crow, and Powis 19 Figure 1 Map of Pacbituns core zone. (Healy et al. 2007:19). Aside from radiocarbon dates and ceramic analysis, two distinct construction patterns, and a shift in orientation, have helped to divide Pacbituns Mai phase into early (900 600 BC) and late (600300 BC) facets. Associated with the early Mai phase (600300 BC), SubStructure B 1 (B 1) and SubStructure B 4 (B 4) were found just above bedrock and are simple constructions composed of a shallow, twocourse high foundation made of roughlyshaped limestone blocks and filled with tamped marl. A meter wide alleyway containing the same tamped marl surface also appears to separate the B 1 and B 4 platforms. Interestingly, the orientation of these two platforms, running in a northeast to southwest direction, is unlike any other found at Pacbitun. O nce both early structures are abandoned however, their late Mai Figure 2 Sub -Structures B -2 and B -3 construction directly on top of earlier SubStructures B -1 and B -4. phase (600300 BC) replacements, SubStructure B 2 (B 2) and SubStructure B 3 (B 3), were constructed directly overtop and oriented slightly west of north a pattern that

PAGE 23

Expanding Sub -Plaza Explorations of Middle Preclassic Pacbitun 20 would persist throughout the sites existence ( Figure 2). Aside from the contrasting orientation, the later platforms are larger and well constructed when compared with their predecessors. B 2, initially explored in the 1990s and revisited after the inception of PRAP in 2008, measures 8.3 m (east west) by 5.5 m (northsouth). Despite the physical discrepancies between B 2 and the two early Mai phase structures, each platform appears to share one significant commonality. Associated with each structure, dozens of chert microdrills and burin spalls were found among st thousands of marine shell artifacts representing all stages of shell bead production (Hohmann 2002; Powis 2009:11; Powis 2010). Over 3000 shell beads and 1500 pieces of shell detritus were recovered during excavation of B 2 in the 1990s. Another 2000 shell beads and 1500 shell detritus were also recovered from the 2009 excavations (Powis 2009:11; Powis 2010:14). The majority of shell found during these previous field seasons were determined to be a non local marine shell which would have required exte nsive trade connections to obtain from the coast (Hohmann 2002; Powis 2009; Powis 2010). The 2017 Excavations in Plaza B Excavations in Plaza B had two main objectives. Our first goal was to expand outward from the previously excavated shell bead workshop, B 2, to expose the neighboring platform, B 3. In doing so, the artifacts associated with B 3 could be analyzed and compared with those found in B 2, helping to determine if the structures function was associated with shell bead production or if it was used for another purpose (i.e., residential). Directly beneath B 2 and B 3, excavations would also further expose the early Middle Preclassic constructions, B 1 and B 4. At the beginning of the season we removed a large portion of backfill from the previ ous excavations of B 2 in Plaza B. Once this was done, the eastern wall of B 3 (running parallel to the west wall of B 2) and its southern wall, could be uncovered in a similar manner as the previous 2008 and 2009 investigations. To reach the depth of the platforms, we would need to excavate through a Middle Preclassic midden known to extend across Plaza B into the neighboring courtyards to the south and Plaza D to the north. The midden covering B 2 measures about one meter in thickness but began to rapi dly thin out towards B 3, approximately one meter to the west, measuring around 2030 cm in thickness. This is a drastic change in a very short space suggesting that the midden may continue to taper down to the west in Plaza B. Once the level of B 3 had been reached, excavations continued to the west along the southern wall to expose the platforms southwestern corner thought to be located in 2008. Excavations would also move north along the eastern wall of B 3 in search of the structures northeastern co rner. Units were placed under the impression that the northeastern corner of B 3 would align with the northwestern corner of B 2, a common pattern in Plaza B (Hohmann 2002: 186; Powis 2009: 10). This assumption would initially lead us to believe that B 3 would be much smaller in size compared to B 2. However, after excavations extended approximately 5.6 m to the north, B 3 did not turn to the west but would continue to the north for another 3 m, bringing the total length of the western wall to 8.6 m (nor th south). Thus, B 2 and B 3 both measure about 5 m by 8 m. However, the length of B 3 runs northsouth, set at a rotational difference of 90 degrees from B 2. B 3 measures 4.7 m east to west and 8.6 m north to south. Whereas, B 2 measures 8.25 m east t o west and 5.5 m north to south (Hohmann et al. 1999:20). Similar to B 2, the interior floor of B 3 is made of a tamped marl surface. Unlike the plaster surfaces of Plaza A, the marl floors of Plaza B were more susceptible to embedded artifacts (Powis 201 0: 15). The floor of B 3 measured around 2030 cm in thickness and contained a variety of artifacts including ceramics, shell beads, chert drills, greenstone, and obsidian. Specifically, 192 shell beads, 658 shell detritus, and 19 chert drills were recov ered from B 3. The floor also contained a high density of jute Interestingly, the limestone walls of B 3 rests directly upon the walls of the early Middle Preclassic platforms, Sub Structure B 1 and B 4. Unlike the late Mai (600 300 BC) platforms,

PAGE 24

Micheletti, Crow, and Powis 21 these walls run northeast to southwest beneath B 2, B 3, and their shared alleyway. This is not where the differences stop however. Excavations would reveal that B 1 and B 4 do not have corners but are actually ovoidal or apsidal in shape ( Figure 3). Each pl atform is crudely constructed and unique when compared to other similarly shaped structures, such as those found at Cuello in northern Belize, constructed at about the same time (Hammond et al. 1991). The floors of B 1 and B 4 are also tamped marl and contain a variety of the same artifacts as B 3; however, the early floors are much thinner, measuring around 5 cm thick. Both early platforms appeared to be constructed just above a modifi ed and leveled bedrock surface. The rounded western ends of B 1 and B 4 begin to turn beneath the southeastern portion of B 3. As our excavations continued to expand to the west, another platform was discovered running beneath the western portion of B 3. Designated as Sub Structure B 16, the third early Middle Preclassic p latform also appears to be ovate or apsidal in shape ( Figure 4). However, the length of B 16 runs northsouth as opposed to the northeast to southwest orientation of the other early Middle Preclassic structures. Aside from the orientation, B 16 shares all the same characteristics as B 1 and B 4 including the floor thickness and artifact assemblage. Recovered from the early Middle Preclassic structures were 487 shell beads, 1280 pieces of shell detritus, and 16 c hert drills. Excavations into Plaza B have illustrated that, between the construction of the early and late Middle Preclassic platforms, the production of shell beads appears to be a constant theme. In the early Mai phase (900 600 BC) orientation seems to vary, while in the late Mai phase the architecture appears to be consistently oriented northsouth. The later buildings were also better constructed compared to those of the previous era. Both early and late platforms in Plaza B were, however, separated by one meter alleyways ( Figure 5 ). Moreover, there does not seem to be any apparent change in artifacts from one period to the next, nor from one platform to another. However, the shell beads recovered from B 2 demonstrate a refinement in production, wi th the beads becoming smaller Figure 3 Sub Structure B -1 and Sub Structure B -4s apsidal shape beneath Sub -Structure B -3 and Sub-Structure B4, looking southwest. Figure 4 Sub Structure B -16s eastern and western walls exposed beneath Sub-Structure B -3, looking northwest. Figure 5 Photogrammetry of Plaza B excavations from 2017.

PAGE 25

Expanding Sub -Plaza Explorations of Middle Preclassic Pacbitun 22 Figure 6 Shell beads from Sub-Structure B 3 in Plaza B. Figure 7 An artistic rendition of El Quemado. (Sketch image created by George J. Micheletti). and more standardized (Hohmann 2002:201). It should be noted that while these refined beads were found in B 3, there were fewer in number compared to those recovered in B 2 ( Figure 6). Jute was equally abundant in both the early and late Middle Preclassic structures. Over 85,000 complete or mostly complete jute were recovered during the 2017 field season. Plaza A and the Middle Preclassic Platform, El Quemado In 2013, the monumental platform El Quemado (otherwise known as Q), was discovered beneath Plaza A while investigating an anomaly previously detected by GPR (ground penetrating radar) ( Figure 7 ) (Skaggs et al. 2014). Sitting meters beneath construction fill, Qs discovery sparked several successive years of excavations to expose as much of the building as possible (Davis et al. 2015; Micheletti et al. 2016; Micheletti et al. 2017; Skaggs et al. 2014). Radiocarbon dating frames the platforms existence within the l ate Mai phase (600 300 Figure 8 A photogrammetric model of the eastern wall of El Quemado exposed in 2016. (Created by George J. Micheletti). BC) with its construction commencing around 550 BC and its burial occurring around the onset of the fourth century BC. The name El Quemado, meaning the burned one, was given due to the structures heavily burned plaster surface as a result of either long term burning practices or one single termination event before its burial. The latter is further suppor ted by additional destruction in the form of defaced stairs, armatures, and possibly even masks as well as the removal of plaster from the cor ners and sides of the building. The first three years of excavation focused on the structures south face which uncovered a central staircase flanked by a pair of upper and lower armatures protruding from the sixth step and third step respectively. With the southern side of the structure fully exposed in 2015, based on the attributes of the stairs, we were able to hy pothesize that Q was either a radial pyramid or a southfacing structure. If Q was radial, the north, east, and west sides would be roughly identical to the layout of the southern stairs. On the other hand, if Q was a southfacing structure, the southern stairs would be unique from all other sides. However, because the Maya predominantly constructed symmetrically, as a southfacing structure, the east and west ends would be similar to one another. Thus, to determine Qs architectural configuration, the east and west ends of the platform were the most logical areas to excavate in 2016 Excavations near the east end of Q were by far the most extensive and would almost completely expose this side of the structure by

PAGE 26

Micheletti, Crow, and Powis 23 Figure 9 A photogrammetric model of the western wall of El Quemado exposed in 2017. (Created by Andrew Vaughan). the end of the 2016 field season ( Figure 8 ). As the excavation progressed, it was clear that the east side was not constructed in the same manner as the southern stairs, meaning that Q was not a radial pyramid. Though much of the summit appeared to be dismantled, the structural integrity was still present to show that Qs east s ide is a steeply inclined wall standing approximately 2 m tall that was mainly composed of large cut limestone blocks. All plaster had been removed from the upper half of the structure in antiquity revealing what remained of the stone robbed wall that onc e formed the eastern summit. The lower half however, was still heavily plastered over. One distinct rectangular feature, simply referred to as an appendage due to its functional ambiguity, protrudes out from the wall about 1 m to the east and is positioned about 4.5 m from the southeastern corner of Q. The appendage, measuring 2 m in length (northsouth) at its base, had also been partially destroyed making it difficult to determine the features true height. Due to time constraints, efforts to expose t he west end of Q we were only able to unearth the upper half of the southwestern corner which had also been stripped of its plaster facade. Though it would seem, through a process of elimination, that Q could definitively be categorized as a south facing structure, excavations were not able to locate the appendage feature on the west side to confirm this hypothesis. Furthermore, excavations had yet to locate the north corners or north face of Q to allow us to determine the northsouth dimensions of the str ucture. Thus, the 2017 excavations were set to continue to explore on the west and north sides of Q. Our extensive investigation of the west end would not only search for a better preserved wall and appendage feature but would also attempt to finally loc ate a northern corner. Although the northern excavations on the centerline of Q appeared to be fruitless in 2016, a cut stone alignment found in a unit beneath Structure 3, the northern structure in Plaza A, would justify revisiting this location as well The 2017 Excavations in Plaza A After removing backfill from the 2016 excavations of the western wall of El Quemado, the 2017 units were set to follow the cut stones of the wall from the summit down in search of a preserved plaster plaza floor. Similar to the eastern wall, the plaster had been removed from the upper half of the western wall but remained intact near the base. As anticipated, soon after we had located the lower plaster facade of the wall, a western appendage mirroring the eastside appenda ge was encountered. Located approximately 4.4 m from the southwest corner of Q, the western appendage was much better preserved than its eastern counterpart. Though the top had been destroyed, much of the plaster surface still coated the sides of the appendage. Continuing to follow the wellpreserved lower facade down to the base of the structure, the plaster surface would eventually lip away from the building horizontally and continue seamlessly as the structures plaza floor.

PAGE 27

Expanding Sub -Plaza Explorations of Middle Preclassic Pacbitun 24 Moving north, though m uch of the western wall had been left intact, the upper half of the structure near the northwest corner appeared to have been stone robbed in antiquity. As excavations continued to the north, approximately 3.6 m north of the westside appendage, the northw estern corner was finally located. Thus, from the northwest corner to the southwest corner, the main platform of Q measures roughly 10.5 m ( Figure 9 ) Noteworthy, excavations near the southwestern corner at the base of Q discovered an east west alignment of nicely cut stones ( Figure 10 ). Running beneath Q, the three course high cut stone alignment was clearly below the level of the platforms plaster plaza floor suggesting that this structure predates the construction of Q. Interestingly, excavations in Plaza A in 2014 found the corner of a similar subplaza construction in front of Structure 5 at about the same depth (Micheletti 2016). Though the function of these structures is not yet known, their construction and elaboration greatly exceed the platfor ms of Plaza B. Shifting to our explorations of Qs north side, after removing backfill from our previous excavations beneath Structure 3 on the centerline of the platform, our investigations of an east west cut stone alignment was determined to be one of t he many task units thought to support the massive amounts of fill brought in to bury Q and build up Plaza A. However, while exploring on the south side of the task unit, a third appendage protruding from the northern facade of Q was discovered. Because excavations had initially exposed the east side of the appendage, we were able to easily locate and uncover a small section of what little remained of the plaster surface at the base of the north side of Q to the east of the appendage ( Figure 11). Excavati ons of the northside appendage proved to be difficult as the preservation was extremely poor. Although time impeded our attempt to further explore this feature, as the season drew to a close, we were able to determine that the appendage was much longer than those on the east and west sides of Q, approximated to be nearly 3 m in length. Finally, with all four sides of Q located, we have determined the full length and width of Pacbituns Middle Preclassic platform. Measuring east to west, from appendage to Figure 10 A south facing photo of a cut stone alignment thought to predate the construction of El Quemado. Figure 11 A southwest facing photo of the northside appendage of Q expose in 2017. appendage, Q is 31.5 m long stretching across Plaza A from the northeastern corner of Structure 2 to the base of Structure 4. North to south, from the northside appendage found beneath the base of Structure 3 to the foot of the southern staircase nearly extending to the center line of Structure 2, Q measures 20.4 m wide. The western and northern excavations of Q were also able to solidify Qs configuration as a southfacing structure. Conclusion The 2017 excavations in Plaza A and Plaza B have both confirmed and altered our interpretations of these areas, h elping to unravel more of the Middle Preclassic story of the Maya at Pacbitun. Sometime after the initial settlement of the site, at least three crudely constructed early Mai phase (900 600 BC) apsidal platforms set at two different orientations were used as work space for the production of shell bead accessories. The onset of the late Mai

PAGE 28

Micheletti, Crow, and Powis 25 phase (600300 BC) appears to issue in a new ideology expressed in the construction of several rectangular platforms oriented with the cardinal directions. Whatever th e cause, it clearly had no influence over shell bead production. In fact, the industry seems to intensify and become standardized as evinced in the late Mai platforms, B 2 and B 3. Although the later constructions are slightly improved, this space continues to be a nonresidential area designated for craft production. Where then, were these specialized craftsmen living and what was their social position at Pacbitun? It is difficult to answer one of these questions without knowledge of the other. Can we even safely assume that the economic organization at Pacbitun ha d created social divisions? Though it is still difficult to answer this question with certainty, the argument supporting a social dynamic at the site has become more conceivable after the discovery of the monumental platform, El Quemado. Due to the abundance of cultural materials associated with the platforms in Plaza B, our current understanding of the Middle Preclassic activities in this plaza greatly exceeds what we know about the happening s of the large ceremonial platform in Plaza A. However, we do know that Qs construction, a project unlike any other previously undertaken at Pacbitun, coincides or closely follows the late Mai phase (600300 BC) architectural shift in Plaza B indicating that the site was far more socially advanced than the crude production platforms had let on. The monumental platform, now known to measure 31.5 m (east west) by 20.4 m (northsouth), would have needed a large, organized labor force derived from the surrounding local community. With Q representing the first of its kind at Pacbitun, the project would have also needed skilled personnel for specialized tasks, and planners to generate and engineer an architectural design. Furthermore, something can also be sa id about the platforms southfacing configuration. Q, built by and for the community, lacks any form of superstructure suggesting that the platforms activities were meant to be visible to a public audience. Might Q have functioned as a southfacing stage for ritual/ceremonial performance? If so, with the recent architectural shift in orientation emphasizing the cardinal directions, could its northern position, as viewed by the audience, and/or the northern backdrop signify a cosmological or mythologica l significance? Regardless of its precise function, the ceremonial nature of Q adds to the list of attributes that have helped to categorize Q as monumental architecture. Thus, Pacbituns late Middle Preclassic community, equipped with an organized economy, community, and ceremonial center, appears to demonstrate the emergence of political organization as well. Acknowledgements We would like to take the time to thank Dr. John Morris, and the rest of the staff at the Institute of Archaeology for their continuous support of our project. The generous financial support of the Alphawood Foundation makes our research at Pacbitun possibl e, and we would like to extend a special thank you to Kristin Hettich. We would also like to thank the people of San Antonio for welcoming us into their village and the local workmen for their hard work and dedication. To our yearly host, Dr. Francis Mor ey, thank you again for your gracious accommodations in Santiago Juan an abode that can offer peace and quiet or a quick getaway to nearby Remos. We would like to acknowledge Cayo Auto Rentals for our reliable transportation. Lastly, we want to acknowl edge Sheldon Skaggs, Jeff Powis, Nicaela Cartagena, Jenny Weber, Mike Lawrence, Jon Spenard, Mike Mirro, Norbert Stanchly, Karen Pierce, Andrew Vaughan, and Jaime Awe. References Awe, Jaime 1992 Dawn in the Land Between the Rivers: Formative Occupation at Cahal Pech, Belize and its Implications for Preclassic Development in the Maya Lowlands. Unpublished Ph.D. dissertation, Institute of Archaeology, University of London, UK. Cheetham, David T. 1995 Excavations of Structure B 4, Cahal Pech, Belize. In Bel ize Valley Preclassic Maya Project: Report of the 1995 Field Season, edited by P.F. Healy and J.J. Awe, 17 -44. Trent University, Occasional Papers in Anthropology 10. Peterborough, Ontario.

PAGE 29

Expanding Sub -Plaza Explorations of Middle Preclassic Pacbitun 26 1996 Reconstruction of the Formative Period Site Core of Cahal Pech, Belize. In Belize Valley Preclassic Maya Project: Report of the 1995 Field Season, edited by P.F. Healy and J.J. Awe, 1 -33. Trent University, Occasional Papers in Anthropology 12. Peterborough, Ontario. David, Jeffery, and Terry G. Powis 2015 El Quemado: Preliminary Findings of Pacbituns Hidden Structure. In Pacbitun Regional Archaeological Project (PRAP): Report on the 2014 Field Season. Report Submitted to the Institute of Archaeology, National Institute of Culture and History, Belmopan. Hammo nd, Norman, Amanda Clarke, and Cynthia Robin 1991 Middle Preclassic Buildings and Burials at Cuello, Belize: 1990 Investigations. Latin American Antiquity Vol 2 No 4:352 -363 Healy, Paul F. 1990 Excavations at Pacbitun, Belize: Preliminary Report on the 1986 and 1987 Investigations. Journal of Field Archaeology 17:247 -262. Healy, Paul F., Christophe Helmke, Jaime J. Awe, and Kay S, Sunahara 2007 Survey, Settlement, and Population History at the Ancient Maya Site of Pacbitun, Belize. Journal of Field Archa eology 32:1739. Hohmann, Bobbi 2002 Preclassic Maya Shell Ornament Production in the Belize Valley, Belize Ph.D. Dissertation, University of New Mexico, Albuquerque. University Microfilms, Anna Arbor. Hohmann, Bobbi, Terry Powis, and Carmen Arendt 1999 The 1997 Investigations at Pacbitun, Belize. In Belize Valley Preclassic Maya Project: Report of the 1995 Field Season, edited by P.F. Healy and J.J. Awe, 19 -30. Trent University, Occasional Papers in Anthropol ogy, 12. Peterborough, Ontario. Micheletti, George, Terry G. Powis, and Kaitlin Crow 2016 Exploring Middle Preclassic Monumental Architecture at Pacbitun, Belize: Recent Archaeological Excavations of El Quemado. In Pacbitun Regional Archaeological Project (PRAP): Report on the 2015 Field Season. Rep ort Submitted to the Institute of Archaeology, National Institute of Culture and History, Belmopan. Micheletti, George, Kaitlin Crow, and Terry G. Powis 2017 In Search of El Quemados Architectural Edges: Expanding Excavations in Plaza A at Pacbitun, Beli ze. In Pacbitun Regional Archaeological Project (PRAP): Report on the 2016 Field Season. Report Submitted to the Institute of Archaeology, National Institute of Culture and History, Belmopan. Powis, Terry 2009 Pacbitun Preclassic Project: Report on the 2008 Field Season. Report Submitted to the Institute of Archaeology, National Institute of Culture and History, Belmopan. 2010 Pacbitun Preclassic Project: Report on the 2009 Field Season. Report Submitted to the Institute of Archa eology, National Institute of Culture and History, Belmopan. Skaggs, Sheldon, and Terry G. Powis 2014 Geophysical and Geological Explorations of Pacbitun, Belize. In Pacbitun Regional Archaeological Project (PRAP): Report on the 2013 Field Season. Report Submitted to the Institute of Archaeology, National Institute of Culture and History, Belmopan. Spenard, Jon, R. Bryan Reece, and Terry G. Powis 2012 Identifying Hinterland Borders: An Initial Report on the 2011 Archaeological Investigations at Sak Pol Pa k, Cayo District, Belize. Research Reports in Belizean Archaeology 9:107118. Spenard, Jon, Teresa Wagner, Terry G. Powis 2013 Of Shells, Soda Straws, Caves, and Kings: Crafting, Body Practices, and Identity Making among the Ancient Maya of Pacbitun, Beli ze. Research Reports in Belizean Archaeology 10:147 155. Weber, Jennifer U. and Terry G. Powis 2010 The Role of Caves at Pacbitun: Peripheral to the Site Core or Central to the Periphery? Research Reports in Belizean Archaeology 8:199207.

PAGE 30

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 2737 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 3 EXPLORING CHANGES IN ACTIVITIES IN MAYA E-GROUPS: ARCHAEOLOGICAL AND G EOCHEMICAL ANALYSIS OF E -GROUP PLASTER FLOORS AT AC TUNCAN, BELIZE Borislava Simova, E. Christian Wells, David W. Mixter, and Lisa LeCount E -Groups were among the first monumental spaces constructed in Preclassic Maya centers and served as important venues for negotiating social interactions and political integration among newly settled peoples. The activities and beliefs associated with the se ritual complexes were integral in shaping Preclassic societies and later reorganizing them in the Classic period. Because Preclassic E -Groups persisted on the landscape over long periods of time, understanding the structure of and changes in activities occurring within them becomes critical for understanding large -scale change in not only ideology, but also social and political practice. Geochemical analysis of occupation surfaces offers a means for supplementing data from punctuated archaeological rem ains with microscopic residues from recurring or cyclical ritual activity occurring within such complexes. In this chapter, we present archaeological and geochemical data from five sequential occupation surfaces from an E -Group complex at the site of Actu ncan, Belize, spanning the Late and Terminal Preclassic Periods (300 BC -AD 250). Results indicate persistent use of food and drink in conjunction with intermittent symbolic deposits, which, though showing gradual shifts ove r time, did not give way to excl usionary displays of authority. Introduction Ritual served a vital function in the construction of both public and private spaces across the Maya Lowlands. We come to understand the role of ritual in the past through architectural elaborations, burials, and monuments, which demonstrate substantial labor and material investment. However, constructions of ritual buildings, caches, and monuments were often lar ge scale and intermittent, commemorating calendrical cycles or important events. Their presence in the archaeological record not only indicate s gatherings and performances occurring in conjunction with such events, but also a myriad of other ritual practi ces known to occur from ethnographic and ethnohistoric accounts. However, these generally involve the use of perishable structures and goods, burning, prayer and processions, all of which are difficult to examine archaeologically. As ritual complexes, su ch as Preclassic E Groups, persist on the landscape over long periods of time, the shifts in activities occurring within them become critical for understanding large scale change in not only ideology, but also social and political practice. In this paper, we use geochemical analysis to supplement archaeological data on ritual activity within the E Group complex at the site of Actuncan, Belize E Groups have a long history of investigation in the Maya archaeology. Their iconic form with an eastern range st ructure and western radial pyramid, first systematically examined at the site of Uaxactun, is ubiquitous throughout the Lowlands. Recent research demonstrates their connections to the origins of social complexity in the Maya Lowlands, ca. 1000 B.C. to 800 B.C. (Estrada Belli 2011; Inomata et al. 2013; Freidel et al. 2017). Many are also linked to shifts in sociopolitical organization in the Classic period, as the ritual spaces were appropriated for political display in the form of carved monuments and roya l burials (Freidel and Schele 1988; Aimers and Rice 2006). Although many aspects of the architecture and deposits found within them point to the ritual functions of E Groups (e.g., Aveni and Hartung 1989; Aveni et al. 2003; EstradaBelli 2012; Aoyama et a l. 2017), little direct evidence of activity has been recovered from these complexes. Bridging the gap between foundational caches and royal burials using more direct proxies of recurring activities will allow researchers to more explicitly link the chang ing uses and meanings of this pervasive and important complex to sociopolitical shifts occurring at many sites across the Lowlands. Geochemical analysis offers a means for supplementing understanding of punctuated archaeological remains with that of repeti tive, or cyclical ritual activity occurring within such complexes. Over time, occupation surfaces accumulate trace amounts of chemicals from

PAGE 31

Exploring Changes in Activities in Maya E -Groups 28 activities such as food processing, burning, and even storage of certain materials. With repetitive occurrence, t hese residues are more likely to preserve at detectable levels over long periods of time (Middleton and Price 1996). Multi elemental characterization of floors allows us to examine the use of space even in the absence of artifacts. In this chapter, we pr esent results from the geochemical analysis of five plaster floors form the Actuncan E Group, in conjunction with architectural and artifact data. When taken together, these analytic techniques improve our detection and understanding of prehistoric activi ty, allowing us to address recurring performance within the complex and relate it to ideological and political shifts occurring in the Preclassic period. Actuncans E Group in Context Actuncan is a ridgetop site, located on the bank of the Mopan River, in western Belize ( Figure 1). It was originally occupied around 1000 B.C. and abandoned in the Early Postclassic period (A.D. 10001250) (McGovern 2004; LeCount and Blitz 2001, 2012; LeCount and Keller 2011; LeCount 2013). Many of its key ceremonial and ci vic structures, including the triadic group of Actuncan South and the E Group of Actuncan North, were established in the Preclassic period ( Figure 2). These architectural features, as well as large stucco masks and a carved stela, suggest the adoption of divine kingship at the site during the Terminal Preclassic Period (100 B.C. A.D. 300). In the Terminal Postclassic period (A.D. 7801000), following the decl ine of Classic kingship in the Lowlands, local power at the site was recentered through the construction of a new civic center and resignification of longlived buildings (Mixter et al. 2014, LeCount et al. 2011, Simova et al. 2014). The Preclassic E Group, located on the northwest ridge of the Actuncan site core, was among the earliest structures established at the site. Comprised of an elongated eastern platform and western radial pyramid flanking a plaza, the E Group rem a ined in use into the Early Classic per iod (A.D. 300600) and served as the site of ritual commemoration in the Late Classic (Donohue 2014). Current understandings of its construction and Figure 1 Location of Actuncan within the Eastern Maya Lowlands (LeCount 2004: Figure 1). Figure 2 Map of Actuncan site core highlighting location of the E -Group complex. occupation history closely align with interpretations of E Groups as places of communal, integrative rituals (Chase and Chase 1995; Aimers and Rice 2006).

PAGE 32

Simova et al. 29 Figure 3 Nort h Profile of Str. 26 with labeled construction phases and floors. E Groups have a long history of investigation in the Maya Lowlands. Early discussions of activities within E Group largely focused on its functions as a solar observatory (La porte and Fialko 1990, 1995; Aveni and Hartung 1989; Aveni et al. 2003). From these observations, the link was drawn to celebrations of agricultural cycles (Aimers 1993, Stanton and Fridel 2003). Recent ly, greater attention has been give n to the early em ergence of the complex, associating it with the Middle Preclassic to Late Preclassic transition (ca. 1000 800 B.C.) and the emergence of many markers of Maya social and political complexity. In this light, E Groups have been discussed in relation to pla cemaking activities, community building, and new patterns in interregional interactions and sedentism (EstradaBelli 2011, 2012; Inomata et al. 2015, 2017). Both integrative and exclusionary practices appear central to the interpretation of these spaces. Their ability to unite dispersed populations (Estrada Belli 2012) is often intertwined with the strategic manipulation of valuable materials and labor. Given Classic period patterns, it is not surprising that many scholars attribute activities of emerging elites to the early E Groups (e.g., Aoyama et al. 2017; Rice 2015). Excavations within the eastern platform of Actuncans E group lend evidence to suggest a different trajectory, one where the communal liturgical functions persist throughout the history of the E Group. In cases such as this, apparent similarities in Preclassic activities must be interrogated more closely to determine where critical differences alter the meanings and activities of the E Group complex in the Classic period. While there is a consensus E Groups formed an important ritual space, critical to the creation of Preclassic communities, a clearer understanding of the specific nature of ritual activities that occurred within them is necessary. To supplement traditional archaeological evidence of ritual, and further understand the kinds of activities that occurred at Actuncans E Group and how they might have changed through time, we conducted geochemical analysis of five sequential occupation surfaces on the summit of the eastern rang e structure. But before moving into the chemical analysis, I will provide an overview of the construction history of the E Group complex, based on two seasons of excavations into the eastern platform (Str. 26), its central shrine (Str. 27), and the wester n radial structure (Str. 23). Construction Phases Excavations in the Eastern platform revealed three distinct phases of construction, beginning with a Cunil Earthen platform (Str. 26sub 2), a Late Preclassic clay and c obble platform (Str. 26 sub 1), named Brown Jay Platform, and a Terminal Preclassic masonry platform (Str. 26), named Owl Platform ( Figure 3). A Bayesian model incorporating nine

PAGE 33

Exploring Changes in Activities in Maya E -Groups 30 radiocarbon dates, LeCounts ceramic seriation work, and the structure strat igraphy have provided a better understanding of the timing of these constructions (see LeCount et al. 2017). Mixters work with Bayesian modeling has helped narrow down the dates of construction events on the range structure and additionally provided some interesting conclusions about the timing and pace of construction. The most unusual aspect of the complex is the Earthen Mound, (Str 26sub 2) below the range structure. It consists of about a half meter of redeposited clay with artifacts. We located a foundational cache of Cunil ceramics with remnants of burning within the mound, dating to ca. 1000 B.C. Given the limited exposure of the construction, we cannot ascertain the form of the mound, whether it was circular, pyramidal, or elongated like later versions of Structure 26. What it does indicate is an early occupation and significance of this ridgetop location to Middle Preclassic populations at the site. The Cunil Earthen Mound was partially buried under a large cobble fill, which appears to have extended the ridgetop to the east prior to the next construction. The subsequent Late Preclassic Brown Jay Platform (Str. 26 sub1) has a central platform constructed out of brown clay with occasional yellow clay lenses, fronted by small cobble walls ( Figu re 4 ). The platform was raised, a series of small terraces were added to its western faade, and unusual linear cobble features were constructed on its summit in the second phase of this construction. A potential third phase was likely also present, bury ing this architecture, but later constructions appear to have cut into its terraces, leaving only the brown clay with yellow lenses. The subsequent Owl Platform (Str. 26) represents a substantial shift in architectural techniques and style occurring in the Terminal Preclassic period. It features six masonry staircase constructions and nine summit plaster floors, whose constructions span from about 200 B.C. to A.D. 260 based on constrained Bayesian modelled dates (LeCount et al. 2017). However, plaza modif ications at the base of the structure continued into the Early Classic. During this time frame, we see shifts in the style of architecture and substantial reworking of the platform, particularly in the staircase. Rather Figure 4 Central Platform of Brown Jay (Structure 26 sub -1 1st). Figure 5 Central Staircase of Owl Platform (left: Str. 26 6th, right: Str. 26 4th). than burying previous steps to expand the size of the structure and build fresh, sections of the existing staircases appear to be added on to, cut into, and even reused. In the earliest staircase, the stonework is consistent, making use of small dressed limestone blocks. Over time, stone size becomes variable and more, larger blocks begin to be incorporated into the construction ( Figure 5). We also have some indication of a shift in the pace of construction of the summit floors. Using the difference function on the Bayseian model constructed in OxCal 4.2 (Bronk Ramsey 2009), Mixter indicates that 5 to 110 years passed between the construction of the first six summit floors, for an average of 1 to 22 years between floor constructions. In contrast, the last t hree summit floors were built over a period of 215 to 441 years, for an average of one floor built every 72 to 147 years. Our chemical analysis of activity areas is currently limited to five of the first six summit floors of the

PAGE 34

Simova et al. 31 Figu re 6 Boxplot comparing concentrations (mg/kg) of 16 elements from the Str. 26 Floors. Terminal Preclassic Owl Platform, prior to the shift in pace of construction. Simovas 2017 excavations in the E Group plaza worked to augment this sam ple size and provide a basis for understanding how different parts of the complex were used. For the time being, these five floors help us target an important period during which the complex was actively used and consistently modified. Geochemical Analys is of Late Preclassic Floors Platform Floors Multi elemental analysis of inorganic residues preserved within constructed floors have been increasingly used to study activities within prehistoric settlements (Middleton and Price 1996; Wells et al. 2000; Ter ry et al. 2004; Hutson and Terry 2006; LeCount et al. 2016). Lime plaster, used to construct the platform floors of Structure 26, traps and preserves a variety of chemical compounds over very long periods, and so is ideal for studying chemical residues of ancient activities. For this study, we collected point samples along the exposed plaster surfaces in 50 cm intervals using a staggered lattice design (see Wells 2010). They were processed and analyzed using InductivelyCoupled Plasma Mass Spectrometry ( ICP MS) at the University of South Florida following protocols developed by Christian Wells. The calibrated concentrations of 21 elements were determined, and the data show less than five percent variation on the U.S. National Institute of Standards and T echnology Certified Reference Material (NIST CRM) standards and internal quality control blanks. While we

PAGE 35

Exploring Changes in Activities in Maya E -Groups 32 collected data on 21 elements, we did not consider calcium (Ca), magnesium (Mg), strontium (Sr), and aluminum (Al), as variation among these elements represent the natural limestone derived matrix. Surprisingly, the only element among the remaining 17 elements to show any anthropogenic variation was phosphorus (P) ( Figure 6 ). Geochemical and Archaeological Evidence of Prehistoric Activity The archite ctural setting of the range structure, overall, suggests a limited set of activities should be present. For instance, generalized signatures from biological debris (e.g., skin and oils) and detritus from feet and clothing (Middleton and Price 1996) should not be present to the same degree as in a domestic area, such as a house patio. Food preparation, marked by manganese (Mn), potassium (K), sodium (Na), and Mg, is also unlikely to occur on the platform centerline. Occurrence of a delimited set of activities is well supported by the lack of anthropogenically enriched element concentrations, besides P. However, this signature also limits the range of ritual practices which could have occurred within the platform. Fo r instance, lack of K, Mg, and Na, also suggest a lack of ritual burning in this location (Heidenreich et al. 1971, Middleton and Price 1996). This is not to say that fire, known to be an important component of ritual from epigraphic and ethnographic evidence is entirely absent, but could suggest that it is occurring in smaller manifestations, perhaps within censors, which are more easily contained and removed from the floors, or alternatively burning could be occurring within other areas of the complex, l ike the plaza. Additionally, the lack of iron (Fe) and other transitional metals (Ti, Ni, Cu, Zn), suggest that pigments such as hematite and ochre, which were important to ritual display and craft production, were also not present in sufficient quantitie s, if at all, to leave a signature. Figure 7 presents results from the spatial analysis of P signatures across the sampled floors. Concentrations of P are expressed as parts per million, or mg element/kg matrix (mg/kg). The P signatures suggest the deposition of organic materials, which contain Figure 7 Kriged image map of the interpolated distribution of P across five floors from Str. 26. Floors are arranged in stratigraphic order. phosphates. These findings are consistent with what we would expect to find if food and drink were present. The earliest constructed floors of the sample, Javier and Armando, had the highest concentration of P, with values as high as 32 and 65 mg/kg respectively. The last constructed floor, Lupe Fiasco, conversely, had the lowest concentration, with highest values under 3 mg/kg. The spatial distribution of P signatures

PAGE 36

Simova et al. 33 Figure 8 Artifacts recovered from features on Javier Floor (top left: Aguacate Orange Jar, top right: Old River Unslipped Jar, left: Striated dish). varied from floor to floor, suggesting that vertical contamination between superimposed floors was unlikely. The earliest plaster floor we analyzed, Javier Floor, was in use between 195 and 105 BC. During excavations we exposed a roughly constructed pavement of stones, possibly a constructed alt a r feature, in the southeast portion of the trench. It was associated with a pile of jute, or freshwater snails often consumed by the Maya of this region, and a cached, striated brown dish ( Figure 8). Additionally, a pit feature in the western portion of the floor yielded several large, fragmented jars, partially refit in the lab. The elevated P signature complements this emphasis on food and drink in the archaeological reco rd and show s a broader dispersion of foodrelated activity across this area than we see in subsequent floors. This could be an indication of longer period of use of this floor or higher intensity of activity accompanying the initial construction of the Ow l P latform. On the next constructed floor, Armando Floor, we observed few archaeological indicators activity. Features outlined in Figure 7 show one small posthole and two bases of postholes from later constructions. Artifact indicators of activity in direct association with the floor are also lacking. However, the chemical signatures reveal substantial anthropogenic enrichment of the plaster floor. Armando Floor has the highest level of phosphorous among the sample d floors, suggesting there was continued emphasis on food and/or use of other organic materials in this phase of the E Group. It can be difficult to determine spatial patterning from the limited exposure of all the floors, but activity here does appear co ncentrated to the south, aligning with the earlier stone pavement or alt a r. This suggests a persistent organization of activity within the architectural space. In the next two floors, Luciano and Santo Floor, we see a similar lack of archaeological featur es and artifacts associated with the surfaces. In these two phases, the organic signatures persist in lower concentrations (12 and 15.5 mg/kg), but appear more spatially constrained. The hotspots of activity revealed in Figure 7 suggest more discreet de position areas of organic material i n comparison to earlier floors. In the last sampled plaster floor, Lupe Fiasco Floor, we again have greater archaeolog ical evidence of ritual activitie s, particularly with the placement of a burial on the structure. Thi s floor was in use between 145 to 50 BC, with three additional floors constructed above it. Unlike Javier Floor, there were no in situ artifacts indicative of food or drink, however we exposed a number of postholes of varying sizes across its surface. Du e to the limited lateral exposure of the excavations, we were unable to discern a clear pattern in the placement of the postholes, but suggest that they represent ritual activity occurring on the platform. In the nearby site of Xunantunich, Brown (2017) has identified similar clusters of postholes with remnants of wooden beams, suggesting that perishable scaffolds or alt a rs were repeatedly erected in front of structures. Discussion Phosphorus, derived from phosphates in organic materials, has long been re cognized as an indicator of human activity, but in many studies, correlates with increased concentrations of other elements, as well. The lack of other anthropogenically enriched concentrations of elements was also unique in comparison with previous geochemical studies in other contexts within Actuncan. As a point of comparison, Fultons (2015) dissertation work examining Terminal Classic Period residential areas in Actuncan North, found a variety of signatures suggesting generalized use of open spaces be tween house groups and heightened activity

PAGE 37

Exploring Changes in Activities in Maya E -Groups 34 surrounding houses. These domestic contexts demonstrated enriched levels in P, K, Mn, and Fe, among others. LeCount and colleague s (2016) work on an elite administrative structure also found a variety of signatures, but in more spatially discrete arrangements, pointing to specialized functions of the various rooms and shifts in their functions over time. Differences in the construction of floors and degree of weathering between interior and exterior spaces could partially contribute to these differences (see Middleton and Price 1996), but the pattern rem ains highly unusual. Given existing perceptions of the E Group as a special function complex, perhaps we should be more surprised that any chemical signatures are present at all, rather than the platform presenting a clean stage for observations and periodic ritual performance or displays. However, it should be noted that while the strength of the signatures is not directly related to the duration and intensity of activity, they do tend to represent repetitive activity which is more likely to allow for the accumulation of inorganic elements in the plaster. Food and drink were and continue to be important components of many celebratory, ri tual, and political gather ings. The discrepancies in the signatures produced by food processing and consumption in domestic and public contexts observed here prompt us to more specifically examine the channels through which food was introduced into the archaeological record. In th is vein, the differences in the early Javier and later Lupe Fiasco Floor are particularly interesting. Baysian modeling suggests that the two floors were in use over similarly long spans of time, yet Javier Floor demonstrated much higher concentrations of phosphate within its matrix. Several possibilities may explain this discrepancy, among them intensity of use and changing patterns in maintenance and use of space. Javier Floor was the first floor of the Owl Platform (Str. 26), inaugurating a new, distinc t construction. As such, it may have borne greater activity, allowing for more accumulation of residues. Conversely, Lupe Fiasco Floor was the last in a relatively rapid set of modifications, after which floor constructions slowed greatly. Whereas Javie r Floor may have been associated with a revitalization of the complex, Lupe Fiasco may have been associated with its declining role in the community. Another intriguing possibility is suggested by the increased presence of postholes, possibly indicative of perishable alt a rs. Perhaps the weaker chemical signature is due to a greater reliance on constructed alt a rs for the display of food and other offerings, which could then be more fully cleared away, but not without some spills and overflows. Continued e xcavations within the complex are needed to provide support for these scenarios, but in either scenario, important shifts in ritual and sociopolitical practice appear to be reflected in the archaeological and geochemical markers of activity within the E Gr oup. Because there are many cultural and natural factors affecting site formation processes, we cannot fully rule out the possibility that any food consumption and offerings suggested by the P signatures were accompanied by other kinds of activities which left no residues or were quickly swept or washed away by rains. However, the lack of other signatures does suggest ce rtain trends in the use of the structure when examined in relation to other archaeological, epigraphic and ethnographic data on Maya ritua l. Conclusions Investigation of the eastern platform of Actuncans E Group demonstrate the utility of geochemical analysis in situating archaeological data from intermittent construction and specialpurpose deposits within ongoing, repetitive activities. Through this approach, we are better able to identify potential shifts in practice that underlie broader ideological, social, and po litical change within the site. The construction history of the platform demonstrates an early and persistent importance of this location within the site. The use of the E Group arrangement further points to certain shared practices and ideologies among Lowland populations. However, shifts in architectural styles over time make it clear that the space was amenable to change as the needs and expectations of local populations changed. The nature of excavated features and caches within the platform further support a responsiveness to local sociopolitical dynamics. Although aspects

PAGE 38

Simova et al. 35 of the deposits, such as reference to water in jar and jute deposits are broadly shared with other E Groups (Freidel et al. 2017), the overall nature of Late Preclassic ritual deposits breaks with patterns observed among other E Groups whose functions become more intimately tied to the enactment of authority and divine kingship (e.g., Uaxactun and Tikal). By examining the types and patterning of anthropogenic residues within the structure floors, we are not only learning more about the activiti es that took place within it, but also evaluating continuity and change at a different scale. In examining five floors from a single construction phase, we are able to detect subtle shifts in the use and meaning of one building, Structure 26, within the c omplex. The observed P signatures and lack of other elements suggest that food offerings, apart from consumption of food within feasts, were an important component of ritual within the E Group complex. When we consider the setting on top of the platform, with the limited available space within a broader, open plaza, it does seem unlikely to have food consumption, or preparation, occurring within Structure 26. The lack of residues from pigments and crafting also suggest that ritual paraphernalia and elabo rate costuming may not have figured in performance on the platform. However, it is possible that the nature of performance simply did not allow for these items to impact the chemistry of the plaster floors. Although these findings conform to expectations that the complex serves a limited, ritual function, they do indicate that the platforms served a broader purpose than solar and calendrical observations, prompting further awareness of the variability in activities that may have taken place within the com plex. In comparing the signatures across the five floors, we begin to see subtle changes in the way local populations engage with the platform. In the absence of in site features and artifacts, we are still able to relate earlier floors to more extensive activity, marked by higher phosphorus signatures, while later floors show a decline. Unfortunately, the soil chemistry in and of itself does not allow us to draw unequivocal conclusions about the nature of Late Preclassic ritual occurring on the eastern s tructure. The shifts we see could represent a declining importance of the complex, or merely a shift in the practical engagement with it, using perishable alt a rs for instance. However, when used in conjunction with architectural and artifact data, contin ued analysis can elucidate the broader significance of these shifts and their relationship to developments across the site. Current research suggests that the E Group complex was not significantly manipulated by elites or enlisted in political displays. The only burial found within Lupe Fiasco floor was modest, placed in an unlined pit without burial offerings. Cached vessels recovered so far are unassuming plainwares and although the chemical signatures indicate food and drink consumption, the only faunal remains recovered in substantial amounts were jute shells. This was not the place for conspicuous consumption, competitive displays, or exclusionary practice. Instead, we interpret it as a key location for the placement and display of communal offerings of food and drink, overlooking the banks of the Mopan, remaining visible and accessible to the broader community. The eastern platform of the E Group represents a distinct space, which we will continue to examine in relation to the complex as a whole. Acknowledgments Major funding for the Actuncan Archaeological Project was provided by the National Science Foundation (BCS0923747), the National Geographic Society Committee for Exploration and Research (927913), and the University of Alabama, awarded t o Lisa LeCount. We gratefully acknowledge this aid. Investigations took place through the permission and generous support of Drs. John Morris and Allan Moore, directors of the Belize Institute of Archaeology (IA). We also wish to thank the staff of the I A for their work exporting materials from Belize to the US and organizing the BAS and RRBA. We are also grateful to the the Galvez and Juan families for permitting us to excavate on their lands, and Azucena Galvez for her hospitality. We were assisted in our research by the hard work and collaboration of many dedicated individuals from San Jos de Succotz and Benque Viejo del Carmen. Don Cruz Puc, Carlos Cocom, and Rene Uck have served admirably as our foremen, and we relied heavily on their experience and that of our lab

PAGE 39

Exploring Changes in Activities in Maya E -Groups 36 crew. We also wish to thank Carolyn Freiwald for her continued analysis of osteological and faunal remains, John Blitz for his tireless support, and BrieAnna Langlie for her aid in the excavations of Structure 26. Works Cited Aimers, James J., and Prudence M. Rice 2006 Astronomy, Ritual, and the Interpretation of Maya E -Group Architectural Assemblages. Ancient Mesoamerica 17( 1): 79 96. Aoyama, Kazuo, Takeshi Inomata, Flory Pinzn, and Juan Manuel Palomo 2017 Polished Greenst one Celt Caches from Ceibal: The Development of Maya Public Rituals. Antiquity 91(357): 701 717. Aveni, Anthony F., Anne S. Dowd, and Benjamin Vining 2003 Maya Calendar Reform? Evidence from Orientations of Specialized Architectural Assemblages. Latin Ame rican Antiquity 14(2): 159 178. Aveni, Anthony F., and Horst Hartung 1989 Uaxactun, Guatemala, Group E and Similar Assemblages: An Archaeoastronomical Reconsideration. In World Archaeoastronomy edited by Anthony F. Aveni, pp. 441 461. Cambrid ge Universit y Press, Cambridge. Bronk Ramsey, Christopher 2009 Bayesian Analysis of Radiocarbon Dates. Radiocarbon 51(1):337360. Brown, M. Kathryn 2017 E Groups and Ancestors: The Sunrise of Complexity at Xunantunich, Belize. In Maya E Groups: Calendars, Astronomy, and Urbanism in the Early Lowlands edited by David A. Freidel, Arlen F. Chase, Anne S. Dowd, and Jerry Murdock, pp. 386 411. Maya Studies. University Pre ss of Florida, Gainesville, FL. Chase, Arlen F., and Diane Z. Chase 1995 External Impetus, Internal Synthesis, and Standardization: E-Group Assemblages and the Crystallization of Classic Maya Society in the Southern Lowlands. In The Emergence of Maya Civilization: The Transition from the Preclassic to the Early Classic edited by Nikolai Grube. Acta Meso amer icana 8. Verlag Anton Saurwein. Donohue, Luke 2014 Excavations at Structure 26 and 27 in Actuncans E Group. In The Actuncan Archaeological Project: Report of the 2013 Field Season, edited by Lisa J. LeCount, pp. 131 150. Report submitted to the Beliz e Institute of Archaeology, Belmopan, Belize. Estrada-Belli, Francisco 2011 The First Maya Civilization: Ritual and Power Before the Classic Period Routledge, London; New York. 2012 Early Civilization in the Maya Lowlands, Monumentality, and Place Makin g: A view from the Holmul Region. In Early New World Monumentality edited by Richard L. Burger and Robert M. Rosenswig, pp. 198 227. University Press of Florida, Gainesville. Freidel, David A., Arlen F. Chase, Anne S. Dowd and Jerry Murdock (editors). 2 017 Maya E Groups: Calendars, Astronomy, and Urbanism in the Early Lowlands Maya Studies. University Pre ss of Florida, Gainesville, FL. Freid el, David A., and Linda Schele 1988 Kingship in the Late Preclassic Maya Lowlands: The Instruments and Places of Ritual Power. American Anthropologist 90(3): 547 567. Fulton, Kara A. 2015 Community Identity and Social Practice During the Terminal Classic Period at Actuncan, Belize. Unpublished Dissertatio n, University of South Florida. Heidenreich, C. E., Hill, A. R., Lapp, D. M. & Navatril, S. 1971 Soil and Environmental Analysis at the Robitaille Site. In Paleoecology and Ontario Prehistory edited by Hurley and C. E. Heidenreich, pp. 179 237. Technical Report No. 2., Department of Anthropology, University of Toronto, Canada. Hutson, Scott R., and Richard E. Terry 2006 Recovering Social and Cultural Dynamics from Plaster Floors: Chemical Analyses at Ancient Chunchucmil, Yucatan, Mexico. Journal of Archaeological Scienc e 33(3): 391 404. Inomata, Takeshi, Jessica MacLellan, and Melissa Burham 2015 The Construction of Public and Domestic Spheres in the Preclassic Maya Lowlands: Public and Domestic Spheres in the Maya Lowlands. American Anthropologist 117(3): 519 534. Ino mata, Takeshi, Flory Pinzn, Juan Manuel Palomo, Ashley Sharpe, Ral Ortz, Mar a Beln Mndez, and Otto Romn 2017 Public Ritual and Interregional Interactions: Excavations of the Central Plaza of Group A, Ceibal. Ancient Mesoamerica 28(1): 203 232.

PAGE 40

Simova et al. 37 Inomata, Takeshi, Daniela Triadan, Kazuo Aoyama, Victor Castillo, and Hitoshi Yonenobu 2013 Early Ceremonial Constructions at Ceibal, Guatemala, and the Origins of Lowland Maya Civilization. Science 340(6131): 467 471. Laporte Juan Pedro, and Vilma Fialko 1990 New Perspectives on Old Problems: Dynastic References for the Early Classic at Tikal. In Vision and Revision in Maya Studies edited by Flora S. Clancy and Peter D. Harrison, pp. 33 66. University of New Mexico Press, Albuquerque. 1995 Un Reencuentr o con Mundo Perdido, Tikal, Guatemala. Ancient Mesoamerica 6: 41 94. LeCount, Lisa J., David W. Mixter, and Borislava S. Simova 2017 Preliminary Thoughts on Ceramic and Radiocarbon Data from Actuncans E -Group Excavations. In The Actuncan Archaeological P roject: Report of the 2016 Field Season pp. 21 42. Report submitted to the Belize Institute of Archaeology, Belmopan, Belize. LeCount, Lisa J., E. Christian Wells, Thomas R Jamison, and David W. Mixter 2016 Geochemical characterization of inorganic resi dues on plaster floors from a Maya palace complex at Actuncan, Belize. Journal of Archaeological Science: Reports 5: 453 464. Middleton, William T., and Barbara J. Price 1996 Identification of Activity Areas by MultiElemental Characterization of Sediment s from Modern and Archaeological House Floors Using Inductively Coupled Plasma Atomic Emission Spectroscopy. Journal of Archaeological Science 23: 673 687. Rice, Prudence M. 2015 Middle Preclassic Interregional Interaction and the Maya Lowlands. Journal of Archaeological Research 23(1): 1 47. Terry, Richard E., Fabian G. Fernandez, Jacob Parnell, and Takeshi In omata 2004 The Story in the Floors: Chemical Signatures of Ancient and Modern Maya Activities at Aguateca, Guatemala. Journal of Archaeological Sc ience 31: 1237 1250. Wells, E. Christian 2010 Sampling Design and Inferential Bias in Archaeological Soil Chemistry. Journal of Archaeological Method and Theory 17: 209 230. Wells, E. Christian, Richard E. Terry, J. Jacob Parnell, Perry J. Hardin, Mark W J ackson, and Stephen D. Houston 2000 Chemical Analyses of Ancient Anthrosols in Residential Areas at Piedras Negras, Guatemala. Journal of Archaeological Science 27(5): 449 462.

PAGE 41

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 3951 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 4 FIDELITY TESTS OF LI DAR DATA FOR THE DET ECTION OF ANCIENT MAYA SETTLEMENT IN THE UPPER BEL IZE RIVER VALLEY, BELIZE Bernadette Cap, Jason Yaeger and M. Kathryn Brown The study of ancient settlement patterns provides essential information about the relationship of sites of ancient human acti vity to each other and to the natural landscape. Traditional approaches of archaeological site discovery involve hours of traversi ng the landscape and manually recording finds. The application of airborne LiDAR has revolutionized the study of ancient settlement patterns because it captures information about large swaths of land quickly and at high resolution. The success of this te chnology to record mounded ancient features is variable, however o ur research explores the fidelity of LiDAR data for the identification of ancient Maya settlement in the Mopan and Macal River valleys, Belize. We present here a comparison of known ancien t features mapped by the Xunantunich Settlement Survey in the 1990s with features identified in airborne LiDAR data. We detected less than 40% of known features in the LiDAR data and found that factors such as modern urbanization and vegetation density ha d the strongest effects on the visibility of ancient features. Understanding the limitations of LiDAR is important for its application in interpretations of ancient Maya settlement patterns. Introduction One of the basic components of archaeological fieldwork is survey, identifying sites and other traces of human activity and locating them precisely in space. There are many ways to find sites, but in the Maya lowlands, systematic pedestrian survey has been the most common method for identifying ancient sites. Pedestrian survey is very time intensive, particularly in the tropical forest environment that is common in the region, and it takes many years to document settlem ent patterns on a regional scale. As a result, there are few areas for which we have a comprehensive regional understanding of ancient Maya settlement patterns. In the past decade, the application of airborne LiDAR (Light Detection and Ranging) to archaeo logical survey in the Maya lowlands, beginning with the area around Caracol (Chase et al. 2010a, 2010b), has revolutionized archaeological survey. We now possess detailed settlement pattern data for contiguous blocks as large as 1257 km2 from zones across the Maya lowlands. The analysis of this data has revealed that LiDAR survey is not equivalent to the results gained through pedestrian survey, however. In this paper, we present a detailed comparison of settlement data obtained through LiDAR and through pedestrian survey in an effort to systematically identify and quantify the differences between the two approaches. LiDAR and Archaeological Survey in the Maya Lowlands Airborne LiDAR survey has transformed archaeological survey in the Maya lowlands because it can capture information about the modern landscape quickly and over large, contiguous expanses. LiDAR data is gathered by airborne instrumentation that emits laser pulses toward the earths surface and records the return of those pulses as the plane flies systematically over a study area (Fernandez Diaz et al. 2014; White 2013). The return data for each pulse allow for the calculation of the location of the object that reflected the pulse, and collectively these comprise a point cloud of three dimensional records that can number into the millions. Surface models can then be created from the point cloud. For identifying archaeological sites, the most useful models are those that only consider points identified via computer algorithm as returns from the ground surface, which can be isolated to essentially strip away vegetation and modern buildings to reveal a bare earth landscape. In this model, ancient mounds, terraces, reservoirs, and other cultural features can be identified as patterns in topographic relief. Through examination of multiple types of LiDAR derived models of the modern surface, we gain views of ancient settlements and their relationships to the natural topography and environmental features like rivers and streams, views that are vastly superior

PAGE 42

Fidelity Tests of LiDAR Data 40 in scale, resolution, and spatial accuracy than is possible with pedestrian survey. Recent applications of LiDAR across the Maya lowlands have yielded many new insights into Maya society (e.g., Chase et al. 2011; Chase et al. 2014a, 2014b; Fernandez Diaz et al. 2014; Hare et al. 2014; Hutson 2015; Hutson et al. 2016; Inomata et al. 2017; Prufer and Thompson 2016; Prufer et al. 2015 Reese Taylor et al. 2016; Von Schwerin et al. 2016). Two concerns that archaeologists have for LiDAR data is its accuracy, that is how close the measurements are to the true or reference values (Fernandez Diaz et al. 2014: 9990), and its fidelity the degree of exactness with which a feature is represented by the point cloud or elevation raster (Fernandez Diaz 2014: 9992). Scholars have primarily examined data fidelity by studying the strengths and weaknesses of LiDAR for identifying sites in diverse environmental and cultural contexts. In all of these studies, the sites and features identified most readily using LiDAR are raised features (e.g., mounds, terraces, causeways, walls), sunken features (e.g., reservoirs, aguadas), and composite features (ditch andrampart fortifications). Features that have little or no topographic relief are genera lly not detectable with LiDAR, although they might be identified in systematic pedestrian survey. In the Maya area, scholars have employed two methods, often in tandem, to systematically evaluate the efficacy of LiDAR for detecting ancient settlement. Som e have conducted digitally based LiDAR surveys and then used pedestrian survey to groundtruth the LiDAR survey data (Hutson 2015; Inomata et al. 2017; Prufer et al. 2015; Reese Taylor et al. 2016). When features identified in the LiDAR data are targeted for survey, groundtruthing allows scholars to quantify the impact of false positives, those features identified in the LiDAR data that are not in fact cultural features. Ground truthing ideally also should include areas for which LiDAR data did not indic ate any cultural features, as this allows scholars to evaluate the impact of false negatives, those cultural features present in a survey area that are not identified in the LiDAR data (ReeseTaylor et al. 2016). Another approach to systematically evaluati ng LiDAR survey data is to compare it with data from previously collected pedestrian survey data (Hutson 2015; Hutson et al. 2016; Inomata et al. 2017; Prufer et al. 2015). This is effective for identifying false negatives, but it can leave false positives unresolved, as the researcher may not be able to discount the possibility that a feature identified in the LiDAR data was not observed during pedestrian survey. The studies referenced above demonstrate archaeological features that can be observed during pedestrian survey often are not identified in LiDAR data. The lowest identification rates are associated with areas of dense vegetation, as the vegetation impedes the ability of the laser pulses to reach the ground surface, thus reducing the number of ground returns. Furthermore, the algorithms that sort points return some of the lowest vegetation returns as ground returns, introducing significant noise into the bare earth models for areas that have dense, low brush. These issues can create significant numbers of false negatives, as we discuss below. Low vegetation can also create false positives, however, as low ancient features and clumps of brush can look very similar in LiDAR derived visualizations (Hutson et al. 2016; Prufer et al. 2015). In fact, in some areas, ground truthing has revealed more false positives than false negatives (Inomata et al. 2017; Reese Taylor et al. 2016). In these ways, vegetation density and type significantly influence the visibility of ancient features in LiDAR data. B ecause vegetation often varies between study areas, careful consideration of the limitations of a specific LiDAR data set is important prior to its analysis for settlement patterns (Fernandez Diaz et al. 2014), an issue that is even more important for broa der comparative analyses that seek to integrate LiDAR data sets from multiple regions. We draw particular attention to the fact that the features that are least visible in LiDAR derived visualizations are low mounds (Hutson et al. 2016; Prufer et al. 2015) Most of these features are the remains of houses once occupied by the commoners who made up the majority of ancient Maya society. Thus, the LiDAR data systematically skew our understanding of ancient settlement patterns and, more broadly, ancient Maya society. This fact makes tests of the fidelity of LiDAR data essential, as they

PAGE 43

Cap, Yaeger, and Brown 41 allow us to make more informed interpretations about ancient Maya settlement patterns that can be observed in LiDAR data. Here we analyze the fidelity of airborne LiDAR data c ollected in Belizes Mopan and Macal River valleys, an area with a long history of ancient Maya occupation ( Figure 1 ). Situated within our study area the major centers of Xunantunich, Buenavista del Cayo, Actuncan, Las Ruinas de Arenal, and Guacamayo. Wh ile limited research has been conducted at Las Ruinas de Arenal and Guacamayo, over a century of investigations in the Mopan River valley provides evidence of political fluctuations between the sites of Xunantunich, Buenavista del Cayo, and Actuncan from t he Middle Preclassic through the Terminal Classic periods (900 BC AD 900) (Ashmore 2010; Helmke and Awe 2013; Leventhal and Ashmore 2004). Most of what we know of this political history derives from studies of the largest sites, but the identification a nd characterization of the settlements located between these competing centers is important for understanding the reasons for and impacts of these shifting political fortunes. Our study region was once covered with the broadleaf deciduous tropical forest t ypical of the central Maya lowlands, but slash andburn milpa agriculture and mahogany logging beginning in the mid19th century have left few, if any, areas of mature forest. Instead, human activity, particularly over the last 50 years, has created a mosaic of land use and vegetative cover. There are some zones of modern urban settlement, but much more of the region consists of plowed agricultural fields, cattle pasture, and successional forest growth of various ages that have developed as milpa fields and pastures were left abandoned over the decades. To test the effectiveness of LiDAR survey in this mosaic landscape, we compared the results of pedestrian survey conducted in the 1990s by the Xunantunich Settlement Survey (XSS) and Xunantunich Archaeologi cal Project (XAP), directed by Wendy Ashmore and Richard Leventhal, with a survey using LiDAR data that we conducted for the same study area. We found a low rate of identification: only 37% of previously mapped sites appeared in the LiDAR data. We discus s several reasons for this Figure 1 LiDAR coverage of the upper Belize River valley highlighting the location of our study area and ancient Maya sites. Figure 2 The Xunantunich Settlement Survey and Xunantunich Archaeological Project survey areas. low rate, some of which are unique to the study area. Our findings create an important baseline from which to build preliminary interpretations of settlement patterns in the valley and plan for further tests of the L iDAR data through groundtruthing pedestrian surveys. Our study highlights some of the limitations of LiDAR data, which are relevant for archaeologists working in areas where comparative datasets are not available and for remote sensing experts,

PAGE 44

Fidelity Tests of LiDAR Data 42 Table 1 Site type definitions and site frequencies recorded by the Xunantunich Settlement Survey. Site Type Characteristicsa XSS Count XSS % I 1 isolated mound, less than 2 m in height 162 42% II 24 mounds, informally arranged, all less than 2 m in height 116 30% III 24 mounds, orthogonal arrangement, all less than 2 m in height 62 16% IV 5 or more mounds, informally arranged, all less than 2 m in height 7 2% V 5 or more mounds, at least 2 with orthogonal arrangement, all less than 2 m in height 15 4% VI 1 or more mounds, at least one 25 m in height 17 4% VII 2 or more mounds, at least one 5 m or higher 7 2% Totals 386 100% a From E hret 1995: Table 1 who can draw on this and other studies like it to make improvements in LiDAR data collection technologies and the algorithms used to analyze and process LiDAR data. The Xunantunich Settlement Survey From 1992 to 1996, XSS conducted intensive pedestrian surve y of 11.8 km2 along and between the Mopan and Macal rivers (Ashmore 1994, 1995; Ashmore et al. 1994; Connell 2000; Keller 1993; Neff et al. 1995; McGovern 1993; Robin 1999; Walkey 1994; Yaeger 1992, 2000, 2010; Yaeger and Connell 1993). For this study, we draw on a selection of areas mapped by the XSS survey ( Figure 2 ): three 400 m wide transects (T/A1, T/A2, and T/A3), a quadrangle centered on the Xunantunich architectural center (T/A4), a zone northeast of Xunantunich (O/A2), and the area around San Lorenzo on the opposite side of the Mopan River (SL). The large areal coverage and diversity of environments and vegetation represented in these survey zones make this a good dataset for testing the fidelity of the LiDAR d ata in this region. The XSS survey methods for transects T/A1 T/A3 involved establishing a central brecha along the transects main axis and cutting perpendicular survey lines ( picados ) 200 m from the brecha at 20 m intervals. The survey crew then walked systematically between the picados looking for evidence of human occupation. The methods employed in the T/A4, O/A2, and SL survey areas were similarly systematic and equally intensive. These methods provided as complete a view of the modern and ancient landscape as was possible at the time. All the sites XSS identified were recorded with tape and compass mapping and a total station. All survey data was digitized using AutoCAD software, which we converted to ArcGIS shapefiles and geo rectified using GPS points collected in 1995 (Neff et al. 1995) and by our project in 2009. We adjusted the projections to WGS1984 and conducted gross spatial adjustments in ArcGIS to correct the transect orientations per information derived from the LiDAR data. While there is some residual error in the conversion of the original digital files, we have confidence that the spatial data is accurate to a level that allows for faithful and direct comparison to the LiDAR data. XSS defined a site as one or more archaeological feat ures located within 25 m (Neff et al. 1995: Table1). It classified sites into one of seven types (Types I VII) according to the number of mounds present, their organization, and maximum mound height

PAGE 45

Cap, Yaeger, and Brown 43 Figure 3 Sites recorded by the (a) Xunantunich Settlement Survey pedestrian survey and (b) LiDAR survey. Figure 4 Kernel density analysis of the (a) Xunantunich Settlement Survey pedestrian survey and (b) LiDAR survey. Table 2 Mound densities recorded by the Xunantunich Settlement Survey. Survey Zone Survey Area (km 2 ) Mound Count Density of mounds per km 2 TA/1 3.30 426 129 TA/2 1.50 153 102 SL 0.95 140 147 O/A2 1.20 124 103 T/A3 1.30 35 27 T/A4 0.70 154 224 Totals 8.95 1032 a 120 a An additional 41 mounds located off transects T/A1 (n=37) and T/A3 (n=4) were recorded by XSS but are not considered in the total presented here

PAGE 46

Fidelity Tests of LiDAR Data 44 ( Table 1 ) (Ehret 1995; Neff et al. 1995). We employ the typology to maintain comparability with XSS data and because the physical attributes used to distinguish the site types can be observed in the LiDAR data. XSS identified 386 sites consisting of 1078 mounds ( Figure 3; Table 1). Most sites ha d less than 5 mounds, all under 2 m in height (Types I and II). Sites were not evenly distributed across the landscape ( Figure 4 ) A comparison of mound densities across the transects ( Table 2 ) reveals that the highest density of mounds (224 mounds/km2) occurs within the T/A4 survey zone, while the linear transects varied from 129 mounds/km2 in T/A1 to just 27 mounds/km2 in TA/3. The average density recorded by XSS was 120 mounds/km2. This is comparable to densities documented by Anabel Ford (1990) just north of the Mopan River, where densities varied from 105 to 129 mounds/km2. To compare these densities with those documented at larger centers, we can examine Tikal, where surveys recorded densities of approximately 235 structures/km2 in the 9 km2 epicenter of the site and 145 structures/km2 in peripheral areas (Carr and Hazard 1961; Rice and Puleston 1981:133). The latter are not much higher than the averages in the Mopan valley and adjacent areas. Comparison of settlement densities such as this demon strates the variation of Maya polities across the lowlands and contributes to understanding the broader complexities of Maya society. LiDAR S urvey Our ability to study ancient Maya settlement in western Belize on a regional scale was transformed in 2013 with the acquisition of 1057 km2 of airborne LiDAR data (Chase et al. 2014b). The NSF National Center for Airborne Laser Mapping (NCALM) at the University of Houston conducted the survey in April and May of 2013 with a Cessna 337 equipped with a GEMINI sy stem laser flying 600 m above ground. The timing at the end of the dry season was ideal, as we could expect few leaves to be present on deciduous forest vegetation and the agricultural fields to be cleared of most crops. The accuracy of the LiDAR points was approximately 2030 cm RMS (root means square) error horizontally and 510 cm RMS Figure 5 Example of LiDAR derived visualizations and satellite imagery used in the LiDAR survey: (a -l) bare earth hillshades; (m) multidirection hillshade ; (n) principal component hillshade; (o) local relief model; and (p) satellite imagery. error vertically. The result is a LiDAR dataset with remarkable resolution of 15 to 25 points per m2 for the entire area. We have focused our analyses on a study area of approximately 150 m2 in the Mopan and Macal River valleys (Figure 1). Our initial examination of a bare earth model derived from the LiDAR data revealed many sites that had not been previously mapped (Yaeger et al. 2016). Here we focus on more detail ed analysis, focusing on LiDAR data from the XSS survey zones shown in Figure 2. To analyze the LiDAR data systematically and identify ancient sites, we created a suite of visualizations based on the LiDAR derived digital elevation model (DEM) created by N CALM. These included: 1) 12 bare earth hillshades created by shifting the azimuth of the sun in 30degree increments; 2) a multi direction bare earth hillshade; 3) a principal components analysis of a bare earth; and 4) a local relief model ( Figure 5). We chose this grouping of

PAGE 47

Cap, Yaeger, and Brown 45 Table 3 Frequency of site types in the LiDAR survey. Site Type LiDAR Survey Count ( % ) LiDAR Sites also Identified by XSS Count ( % ) Count of LiDAR Sites Assigned to Same Site Type by XSS I 129 (63%) 36 (26%) 29 II 20 (10%) 36 (26%) 9 III 37 (18%) 38 (27%) 21 IV 0 0 2 (1%) 0 V 0 0 9 (6%) 0 VI 13 (6%) 13 (9%) 9 VII 7 (3%) 7 (5%) 7 Totals 206 (100%) 141 (100%) 75 visualizations due to their potential to identify features hidden in shadows, and to bring out visually low, raised features on the landscape. We examined the visualizations in 500 x 500 m blocks at an initial scale of 1:3000. When a possible feature appeared in any of the visualizations, we marked it with a point in an ArcGIS point shapefile. For each location, we then examined the original point cloud data, creating multiple profile views across the area to identify topographic profiles that showed a high potential to be an ancient feature. Through this process, we eliminated some modern constructions and trees. We also examined satellite imagery to help confirm modern features. Because ancient mounds can be very similar to undergrowth, particularly low, dense brus h (Hutson et al 2016; Prufer et al. 2015), we only identified features as ancient mounds if they had a clear, smooth, mounded shape. Through close analysis of the LiDAR data, we identified 206 sites, totaling 466 mounds ( Figure 3 ; Table 3 ). Of the 206 sit es identified, only 141 were mapped by XSS, and these 141 sites account for only 37% of the sites recorded by XSS. In some cases, we were confident that a site identified in the LiDAR corresponded with a site identified by XSS, but not all of the features mapped by XSS were visible in the LiDAR data. As a result, the frequencies of site types vary between the two surveys (Tables 1 and 3). Comparison of the number of mounds mapped by XSS and mounds identified in the LiDAR survey provides a more direct, one to one comparison. Of the 466 mounds identified in the LiDAR survey, XSS mapped 381 of these, which represent 35% of the mounds mapped by XSS. The slightly higher identification rate of sites likely reflects the ability to identify a site regardless of whether all features (i.e., mounds) were identified. Regardless, whether using the site or the mound as a unit of comparison, the broader picture of LiDAR fidelity remains the same: nearly two thirds of the settlement identified through pedestrian survey was not evident in the LiDAR data. Our analysis of the LiDAR data also identified 65 sites and 85 mounds that were not mapped by XSS. While XSS methods were thorough and entailed 100% systematic coverage, it is likely that some of these new sites identif ied in the LiDAR represent sites that XSS missed. It is also possible that many of these mounds are false positives. Targeted groundtruthing is required to determine the degree to which the LiDAR analysis introduces false positives. Despite these limita tions, the LiDAR data do provide important settlement information. Sites with taller mounds and a larger areal footprint are highly visible in the LiDAR data and more likely to be assigned to the same site type. Because LiDAR data can be reliably used to identify most of the Type VI and VII sites in a region, the LiDAR can be used to reconstruct the top tiers of a regional settlement hierarchy. Furthermore, the overall pattern of settlement density as revealed in kernel analysis of the LiDAR data is broadly comparable to that revealed through pedestrian survey ( Figure 4 ). Thus, the LiDAR data provides insight into regional variability of settlement density across a region. Finally, although not a focus here, the

PAGE 48

Fidelity Tests of LiDAR Data 46 LiDAR provides reliable, georeferenced da ta regarding natural topography and geomorphological features, which can then be correlated with settlement data to study questions about humanenvironment dynamics. Impediments to Visibility in the LiDAR D ata The XSS sites with the lowest rate of identification in the LiDAR survey are those that have low mounds less than 2 m in height and small footprints (Tables 1 and 3). These findings mirror those of other studies (e.g., Hutson et al. 2016; Prufer et al. 2015) and represent a significant challenge for archaeologists and LiDAR experts. If we accept the consensus that the majority of these features are Maya houses, then the LiDAR data presents us with a skewed view of Maya settlement patterns, one that introduces a bias against non elite members of Maya society, as the larger, taller constructions more likely represent elite houses and associated structures. Being attentive to this bias is critical so that we do not contribute further to the historic emphasis within Maya studies on the upper echelons of ancient Maya society. In order to improve our ability to use LiDAR data as a tool for identifying ancient settlements, we must better understand the factors that can lead to the underrepresentation of certain t ypes of sites. In our study area, the reasons are varied, and we would not expect the same suite of factors to be present in other study areas. Given that human population expansion and agricultural development are increasingly transforming landscapes ac ross the Maya lowlands into mosaics of vegetation cover, however, we expect our findings to have broad applicability. As mentioned above, the modern history of land use plays a very strong role in shaping our ability to identify ancient settlements in the LiDAR data. This is particularly clear in areas of modern urban expansion: for example, it proved very difficult to identify sites in the modern village of San Jos Succotz, west of Xunantunich. The density of the modern buildings masked the sites and mounds located in yards and empty lots between houses. Since the 1990s the village has experienced significant growth, which also has caused the destruction of some sites that had been documented by XSS. Site destruction has also been driven by agricultur e, as traditions of slash andburn handclearing of fields have been replaced by mechanized agricultural practices such as plowing. Plowing tends to lower the elevation of ancient mounds and reduces their relief, making them more challenging to identify i n the LiDAR data. Unfortunately, the effects of urbanization and mechanized agriculture are difficult to quantify without a strong, longitudinal data set that can be used to trace the destruction of sites. The other major factor impeding the identificatio n of ancient settlement in the LiDAR data from our study area is modern vegetative cover. The visibility of any feature on the ground surface in LiDAR derived visualizations is directly affected by the number of ground returns produced during the initial LiDAR data collection. The number of ground returns is impacted by the obstructions that are present above the ground surface, particularly the leaves, stems, branches, and trunks of plants. Leaves are diffuse in terms of reflectivity, which allows for s ome of the pulse to travel through them and reach the next target, but in areas of dense vegetation, they can still impede pulses from reaching the ground. One way to mitigate this problem in the tropical deciduous forests of Maya lowlands is to conduct Li DAR collection flights near the end of the dry season, when many trees have dropped their leaves. Furthermore, by conducting multiple collection flights in different directions over the study area, one increases the chances of having a laser pulse travel between vegetation to the ground. Despite these efforts, there will always be uneven distribution of ground returns across a region, a factor that is strongly shaped by vegetation. As noted above, the western Belize LiDAR data was collected to maximize gr ound returns, flying multiple flights in different directions during April and May, when leaf cover would be at its lowest. Because of the differences in vegetation and land use practices across our study area, we were particularly interested to assess ho w vegetation affected our LiDAR survey results and how they compared to other areas of the Maya lowlands. Our initial

PAGE 49

Cap, Yaeger, and Brown 47 Table 4 Effect of vegetation canopy density on identification of site types I, III, and VI in the LiDAR data. Canopy Density Type I (n=38) % identified T ype III (n=32) % identified T ype VI (n=12) % identified 0 to .39 n/a n/a 100.0% .4 to .49 57.1% n/a n/a .5 to .59 23.1% 100.0% n/a .6 to .69 26.9% 57.1% 100.0% .7 to .79 18.8% 88.9% 50.0% .8 to .89 25.6% 100.0% n/a .9 to .99 18.0% 45.5% 70.0% analysis of this factor in the study area focuses on canopy density and ground point density, both of which rel a te to the ability of the laser pulses to reach the modern surface (Salleh et al. 2015). We created a vegetation density raster using the methods described by ESRI (2016) that converts the LiDAR point cloud data first into two rasters, one based on ground returns and the other based on above ground returns. The cell size chosen in this calculation was 4 m. These two rasters were then added together to create a raster that represented the total returns per cell. The above ground return raster as then divided by the total points raster to produce a raster that represented the density of returns above ground. In this final raster, cell values ranged from 0 to 1, where 0 represents bare earth and 1 represents the densest vegetative coverage in which no laser pulses reach the ground. The vegetation densities in our study area break down roughly into three general categories. Densities of 0.7 to 1.0 are found most commonly in forested portions of the study area, while values between 0.2 and 0.69 tend to be urban areas, fields with crops, orchards, and less densely forested areas. D ensities below 0.2 are found along the waterways, active cattle pastures, and fields devoid of crops. There are few locations within the study area that have such low values. For our analysis we used sites, as defined by XSS, as the unit of study. Because a single site can occupy an area with different vegetation densities, we associated each site with the density that covered the greatest area of a sites footprint. We excluded sites for which there was too much diversity in vegetative density to identi fy a predominant density. We also excluded sites identified in the LiDAR survey that were not also mapped by XSS. We focus here on the effect of vegetation density on visibility of three site types that capture the spectrum of variation within the XSS sit e types. Sites of Type I are isolated mounds less than 2 m tall that represent the sites that we expected to be the most difficult to identify in LiDAR data. They also are the majority of sites present in the study area. Type III sites also consist of m ounds less than 2 m tall, but the orthogonal arrangement of multiple mounds presents linear visual cues that are not common in nature and are generally easier to detect by eye in LiDAR data. Finally, Site Type VI sites have mounds between 2 and 5 m in hei ght and are among the largest sites in terms of mound height and footprint. We expect that larger targets like these would be more visible on the LiDAR derived visualizations regardless of vegetative impediments. In our canopy density dataset, we found th at no matter the density of vegetation, sites with larger mounds and more mounds have a greater chance of being identified in the LiDAR data ( Table 4 ). Fewer of the smallest sites (Site Type I) were visible when under vegetative cover, and to a significan t degree (chi square of Site Type I and III at .9.99 canopy density 2= 6.433, p=0.0112; chi square of Site Type I and VI at .9 .99 canopy density 2=12.260,

PAGE 50

Fidelity Tests of LiDAR Data 48 Table 5 Effect of LiDAR ground point densities on identification of site types I, III, and VI in the LiDAR data. Poi nt Density (per 4m 2 ) Type I (n=38) % identified Type III (n=31) % identified Type VI (n=12) % identified 0 to 25 11.9% 42.9% 50.0% 25 to 50 25.7% 42.9% 75.0% 50 to 75 22.2% 81.0% 66.7% 75 to 100 21.0% 83.3% 100.0% 100 to 125 66.7% 60.0% 100.0% 125 to 150 100.0% 50.0% 100.0% p=0.0004). There was also a significant difference in the success rate for Site Type I identifications between the highest and lowest 2=5.586, p=.0181). Under the lowest vegetative density category (.4 .49), however, we still were not able to identify all Site Type I sites. This suggests that even low density vegetation affected site visibility. Unexpectedly, there is not a strict linear relationship between vegetation density and identification rate for any of the three site type categories. For example, identification rates for Type I sites between vegetation dens ities of 0.5 to 1.0 are not significantly different. For Type I sites in particular, there appears to be a quasi threshold at roughly at the 0.5 vegetation density level between better and worse identification rates. This finding highlights the general difficulty in locating small sites in LiDAR data. Determining the specific cause of this pattern, however, likely requires pedestrian survey. Overall, our results of the canopy density analysis demonstrate that vegetation does impede the visibility of sit es within the LiDAR data, but to varying degrees. We performed a second test of the LiDAR data to examine the effect of ground return density on site visibility. The density of points categorized as ground returns is, in part, a reflection of canopy densi ty. Therefore, we expected to obtain similar results to those in the analysis just described. For this analysis, we used the bare earth raster created during the process of making the canopy density raster, as described above. We again examined the visi bility of Site Types I, III, and VI, and of these, only those where we could determine a dominant density of ground returns. The results of the ground point density analysis were indeed similar to the canopy density results ( Table 5 ), but the correlation between the number of ground returns and likelihood of identifying sites was stronger than was the case for the canopy density analysis. As the number of ground returns decreased, there was a significant reduction in identification rates for smaller sites (e.g., chi square for Site Types I and III at 0 25 ground point density I and VI at 0 25 ground point density Our results also seem to indicate that there might be point density thres holds for the identification of certain types of sites. For example, our success rate in identifying Type I sites jumped significantly once the ground point density exceeded 100 points/4 m2. For Type III sites, the threshold is slightly lower at more tha n 50 points/4 m2. These thresholds may be particular to our study area and methods we used to visualize the LiDAR data, but they suggest real limits to visibility in the LiDAR data. Conclusions The sites archaeologists find and the ways they find them impact our understandings of the ancient past. Our investigations into the efficacy of LiDAR data as a tool for identifying ancient settlement in the Mopan and Macal River valleys demonstrates that it is of great utility for documenting la rger sites, but has limitations for smaller sites. Similar findings have been documented elsewhere in the Maya region and

PAGE 51

Cap, Yaeger, and Brown 49 other locations (e.g., Chase et al. 2011; Crow et al. 2007; Hutson et al. 2016; Prufer et al. 2015), suggesting some broad limitations of LiDAR in archaeological applications. Collectively, these studies demonstrate that although LiDAR is truly revolutionary in its ability to provide information on large continuous expanses of land very quickly, traditional archaeological methods of pe destrian survey are still important for settlement pattern studies. In our region, pedestrian surveys will be important for checking the information obtained from LiDAR data and our interpretations of the data. We are particularly interested in understan ding the relationship of false negatives and false positives in the LiDAR to determine if we can improve our abilities to identify ancient settlement in LiDAR data. Because natural landscapes and ancient settlement traces vary, LiDAR data collection methods may differ between projects. We are also in an exciting time of methodological development and innovation, and we expect LiDAR to improve over time as the technology itself improves, and as archaeologists experiment with a wide range of methods to visua lize LiDAR data. We argue that tests of LiDAR data fidelity should be an important part of this process, as they help us quantify and understand the limitations of LiDAR to specific study areas. The mosaic modern landscape of the upper Belize River valley presents a particularly challenging case in application of LiDAR products, which rely on single algorithms to characterize a diverse landscape. Our tests of the LiDAR in the valley indicate that different vegetation densities affected the ability of las er pulses to reach the actual modern ground surface. With this knowledge we can collaborate with LiDAR technology experts to create more nuanced approaches for visualizing LiDAR data for archaeological research. Acknowledgements We extend heartfelt thanks for the generous and continued support our project has received from the Belize Institute of Archaeology, particularly Dr. John Morris, Dr. Jaime Awe, and Ms. Melissa Badillo. We are grateful to Wendy Ashmore for permission to uti lize the pedestrian survey data collected by the Xunantunich Settlement Survey, funded by the National Science Foundation (SBR 9321503). Funding for the Western Belize LiDAR acquisition in 2013 was generously provided by the Alphawood Foundation. The Nat ional Center for Airborne Laser Mapping collected the LiDAR data and produced some of the derived products used in this study. The Worldview 2 satellite imagery used was collected in 2010 and obtained by a grant from Digital Globe. The LiDAR analysis was undertaken while Cap was a Postdoctoral Research Fellow with support from University of Texas at San Antonios College of Liberal and Fine Arts and Office of the Vice President for Research. We extend a special thank you to the staff, students, and crew of both the Mopan Valley Archaeological Project and Mopan Valley Preclassic Project for their continued collaborative support and good cheer. References Cited Ashmore, Wendy 1994 Settlement Archaeology at Xunantunich, Belize, Central America. In Xunantunich Archaeological Project: The 1994 Field Season, edited by Richard M. Leventhal and Wendy Ashmore, pp. 722. Report submitted to the Belize Institute of Archaeology, Belmopan. 1995 Settlement Archaeology at Xunantunich, 1995. In Xunantunich Arch aeological Project: The 1995 Field Season edited by Richard M. Leventhal and Wendy Ashmore, pp. 10-25. Report submitted to the Belize Institute of Archaeology, Belmopan. 2010 Antecedents, Allies, Antagonists: Xunantunich and its Neighbors. In Classic May a Provincial Politics: Xunantunich and its Hinterlands edited by Lisa J. LeCount and Jason Yaeger, pp. 4664. University of Arizona Press, Tuscon. Ashmore, Wendy, Samuel V. Connell, Jennifer J. Ehret, Clarence H. Gifford, L. Theodore Neff, and Jon C. Van denbosch 1994 The Xunantunich Settlement Survey. In Xunantunich Archaeological Project: The 1994 Field Season edited by Richard M. Leventhal and Wendy Ashmore, pp. 248-290. Report submitted to the Belize Institute of Archaeology, Belmopan. Brown, M. Kath ryn, Jennifer Cochran, Leah McCurdy, and David Mixter 2011 Preceramic to Postclassic: A Brief Synthesis of the Occupation History of Group E, Xunantunich. Research Reports in Belizean Archaeology 8: 209219.

PAGE 52

Fidelity Tests of LiDAR Data 50 Carr, Robert F. and James E. Hazard 1961 Map of the Ruins of Tikal, El Peten, Guatemala. Tikal Reports, No. 11. Museum Monograph 21. University Museum, University of Pennsylvania, Philadelphia. Chase, Arlen F., Diane Z. Chase, and John Weishampel 2010a Lasers in the Jungle: Airborne Sensors Reveal a Va st Maya Landscape. Archaeology 64(4): 2729. Chase, Arlen F., Diane Z. Chase, Christopher T. Fisher, Stephen J. Leisz, and John F. Weishampel 2010b Geospatial Revolution and Remote Sensing LiDAR in Mesoamerican Archaeology. Proceedings of the National Academy of Sciences 109: 1291612921. Chase, Arlen F., Diane Z. Chase, John F. Weishampel, Jason B. Drake, Ramesh L. Shrestha, K. Clint Slatton, Jaime J. Awe, and William E. Carter 2011 Airborne LiDAR, Archaeology, and the Ancient Maya Landscape at Caraco l, Belize. Journal of Archaeological Science 38: 387-398. Chase, Arlen F., Diane Z. Chase, Jaime J. Awe, John F. Weishampel, Gyles, Iannone, Holly Moyes, Jason Yaeger, and M. Kathryn Brown 2014a The Use of LiDAR in Understanding the Ancient Maya Landscape: Caracol and Western Belize. Advances in Archaeological Practice 2: 208 221. Chase, Arlen F., Diane Z. Chase, Jaime J. Awe, John F. Weishampel, Gyles, Iannone, Holly Moyes, Jason Yaeger, and M. Kathryn Brown, Ramesh L. Shrestha, William E. Carter, and Juan Carlos Fernandez -Diaz 2014b Ancient Maya Regional Settlement and Inter Site Analysis: The 2013 West -Central Belize LiDAR survey. Remote Sensing 2: 8671 8695. Connell, Samuel V. 2000 Were They Well Connected? An Exploration of Ancient Maya Regional Integration from the Middle -Level Perspective of Chaa Creek, Belize. Unpublished PhD dissertation, Department of Anthropology, University of California. Crow, Peter, S. Benham, B. J. Devereux, and G. S. Amable 2007 Woodland Vegetation and its Implications for Archaeological Survey Using LiDAR. Forestry 80: 241 -252. Ehret, Jennifer 1995 The Xunantunich Settlement Survey Test Pitting Program. In Xunantunich Archaeological Project: The 1995 Field Season, edited by Richard M. Leventhal and Wendy Ashmore, pp. 164-192. Report submitted to the Belize Institute of Archaeology, Belmopan. Environmental Systems Research Institute, Inc. (ESRI) 2016 Estimating Forest Canopy Density and Height. Electronic docume nt, http://desktop.arcgis.com/en/arcmap/10.3/manage data/las -dataset/lidar -solutions -estimating forest density -and -height.htm Accessed December 7, 2017. Fernandez-Diaz, Juan Carlos, William E. Carter, Ramesh L. Shrestha, and Craig L. Glennie 2014 Now You See ItNow You Don't: Understanding Airborne Mapping LiDAR Collection and Data Product Generation for Archaeological Research in Mesoamerica. Remote Sensing 6: 9951 10001. Ford, Anabel 1990 Maya Settlement in the Belize River Area: Variations in Residence Pat terns of the Central Maya Lowlands. In Precolumbian Population History in the Maya Lowlands, edited by T. Patrick Culbert and Don S. Rice, pp. 167182. University of New Mexico Press, Alburquerque. Hare, Timothy, Marilyn Masson and B. Russell 2014 High -d ensity LiDAR Mapping of the Ancient City of Mayapan. Remote Sensing 6: 9064 9085. Helmke, Christophe G. and Jaime J. Awe 2013 Ancient Maya Territorial Organization of Central Belize: Confluence of Archaeological and Epigraphic Data. Contributions in New World Archaeology 4: 59 -90. Hutson, Scott R. 2015 Adapting LiDAR data for regional variation in the tropics: A case study from the Northern Maya Lowlands. Journal of Archaeological Science: Reports 4: 252263. Huston, Scott R., Barry Kidder, Cline Lamb, Daniel Vallejo -Cliz, and Jacob Welch 2016 Small Buildings and Small Budgets: Making LiDAR Work in Northern Yucatan, Mexico. Advances in Archaeological Practice 4: 268 283. Inomata, Takeshi, F lory Pinzn, Jos Luis Ranchos, Tsuyoshi Haraguchi, Hiroo Nasu, Juan Carlos FernandezDiaz, Kazuo Aoyama, and Hitoshi Yonenobu 2017 Archaeological Application of Airborne LiDAR with Object-Based Vegetation Classification and Visualization Techniques at the Lowland Maya Site of Ceibal, Guatemala. Remote Sensing 9: 563589. Keller, Angela 1993 Vision and Revision: The Remapping of Xunantunich. In Xunantunich Archaeological Project: The 1993 Field Season, edited by Richard M. Leventhal and Wendy Ashmore, pp. 86-99. Report submitted to the Belize Ins titute of Archaeology, Belmopan.

PAGE 53

Cap, Yaeger, and Brown 51 Leventhal, Richard M. and Wendy Ashmore 2004 Xunantunich in a Belize Valley Context In The Ancient Maya of the Belize Valley: Half a Century of Archaeological Research, edited by James F. Garber, pp. 168 179. University o f Florida Press, Gainesville. McGovern, James O. 1993 Survey and Excavation at Actuncan. In Xunantunich Archaeological Project: The 1993 Field Season edited by Richard M. Leventhal and Wendy Ashmore, pp. 100-127. Report submitted to the Belize Inst itute of Archaeology, Belmopan. Neff, L. Theodore, Cynthia Robin, Kevin Schwartz, and Mary K. Morrison 1995 The Xunantunich Settlement Survey. In Xunantunich Archaeological Project: The 1995 Field Season edited by Richard M. Leventhal and Wendy Ashmore, pp. 13 9 -165. Report submitted to the Belize Institute of Archaeology, Belmopan. Prufer, Keith M. and Amy E. Thompson 2016 Lidar -Based Analyses of Anthropogenic Landscape Alterations as a Component of the Built Environment. Advances in Archaeological Practice 4: 393 409. Prufer, Keith M., Amy E. Thompson, and Douglas J. Kennett 2015 Evaluating airborne LiDAR for detecting settlements and modified landscapes in disturbed tropical environments at Uxbenka, Belize. Journal of Archaeological Science 57: 1 13. ReeseTaylor, Kathryn, Armando Anaya Hernndez, F. C. Atasta Flores Esquivel, Kelly Monteleone, Alejandro Uriarte, Christopher Carr, Helga Geovannini Acua, Juan Carlos Fernandez-Diaz, Meaghan Peuramaki -Brown, and Nicholas Dunning 2016 Boots on the Ground at Yaxnohcah: Ground Truthing LiDAR in a Complex Tropical Landscape. Advances in Archaeological Practice 4: 314 338. Rice, Don S. and Dennis E. Puleston 1981 Ancient Maya Settlement Patterns in the Peten, Guatemala, In Lowland Maya Settlement Patterns, edited by Wendy Ashmore, pp. 121156. University of New Mexico Press, Albuquerque. Robin, Cynthia 1999 Towards an Archaeology of Everyday Life: Maya Farmers of Chan Nohool and Dos Chombitos Chik'in, Belize Unpublished PhD dissertation, Department of Anth ropology, University of Pennsylvania, Philadelphia. Robin, Cynthia, Andrew R. Wyatt, Laura J. Kosakowsky, Santiago Juarez, Ethan Kalosky, and Elise Enterkin 2012 A Changing Cultural Landscape: Settlement Survey and GIS at Chan In Chan: An Ancient Maya Farming Community edited by C. Robin. University of Florida Press, Gainesville. Salleh, Mohd Radhie Mohd, Zamri Ismail, Muhammad Zulkarnain, and Abdul Rahman 2015 Accuracy Assessment of LiDAR -derived Digital Terrain Model (DTM) with Different Slope and C anopy Cover in Tropical Forest Region. ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences II 2 -W2(1):183 189. Von Schwerin, Jennifer, Heather Richards -Rissetto, Fabio Remondino, Maria Grazia Spera, Michael Auer, Nicholas Bille n, Lukas Loos, Laura Stelson, and Markus Reindel 2016 Airborne LiDAR Acquisition, Post Processing and Accuracy -Checking for a 3D WebGIS of Copan, Honduras. Journal of Archaeological Science: Reports 5: 85 104. Walkey, John 1994 Xunantunich Archaeological Reserve Survey: 1994. In Xunantunich Archaeological Project: The 1994 Field Season, edited by Richard M. Leventhal and Wendy Ashmore, pp. 93-100. Report submitted to the Belize Institute of Archaeology, Belmopan. White, Devin A. 2013 LiDAR, Point Clouds, and their Archaeological Applications. In Mapping Archaeological Landscape from Space edited by Douglas C. Comer and Michael J. Harrower, pp. 175 186. Springer, New York. Yaeger, Jason 1992 Xunantunich Settlement Survey. In Xunantunich Archaeological Project: The 1992 Field Season edited by Richard M. Leventhal and Wendy Ashmore, pp. 100-126. Report submitted to the Belize Institute of Archaeology, Belmopan. 2000 Changing Patterns of Social Organization: The Late and Terminal Class ic Communities at San Lorenzo, Cayo District, Belize. Unpublished PhD dissertation, Department of Anthropology, University of Pennsylvania, Philadelphia. 2010 Landscapes of the Xunantunich Hinterlands. In Classic Maya Provincial Politics: Xunantunich and its Hinterlands edited by Lisa J. LeCount and Jason Yaeger, pp. 233-249. The University of Arizona Press, Tuscon. Yaeger, Jason and Samuel V. Connell 1993 Xunantunich Settlement Survey. In Xunantunich Archaeological Project: The 1993 Field Season edited by Richard M. Leventhal and Wendy Ashmore, pp. 172-218. Report submitted to the Belize Institute of Archaeology, Belmopan.

PAGE 54

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 5363 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 5 CLASSIC MAYA HOUSEHO LD CERAMIC BELONGING S: AN UNTAPPED RESOURCE FO R UNDERSTANDING DAILY LIFE Sherman Horn III and Anabel Ford Archaeological research in the Maya area has relied on ceramics as chronological markers that are critical to the relative dating of Maya sites. A working ceramic chronology was established for the central Lowland Maya area with the Uaxactun project in the 1930s and expanded by subsequent work on collections from Barton Ramie in Belize. By the 1960s, the type variety system was refined and elaborated by Tikal project ceramic analysts. Type-variety has proven useful for establishing the Maya ceramic timeline, but it provides few data to use for the distribution of vessel forms and shapes. Formal characteristi cs relate direct ly to vessel function, and knowing the proportions and distributions of different vessel types can reveal fundamental aspects of ancient Maya daily life. This paper examines the distribution of Late Classic Maya vessel forms and shapes and explores vessel diversity from Maya residential units in the El Pilar area. Analysis of Late Classic vessel forms and shapes reveals common and variable forms and functions and describes the essential belongings of Maya residential units. These are the building blocks of Maya communities within the forested landscape, which we continue to map in ongoing fieldwork at El Pilar. Introduction Archaeologists have recognized the importance of studying ancient Maya household remains for many decades (Wauchope 1938), as they comprise the basic units of socioeconomic production and reproduction in communities (Arnold 1985; Netting et al. 1984). Artifacts from residential unit assem blages reveal only part of a households total possessions, but they do reflect a range of daily activities in ancient society. Descriptions and analysis of household possessions are integral to comparative studies of wealth inequality (Smith 1987), and r ecent studies have examined variability among Maya household assemblages to explore differences in wealth, status, or specialization (e.g., Blackmore 2012; Ford 2010; Ford and Olson 1989; Keller 2012; Lucero 2001; Robin 2013; Pagliaro 2011 ). Ceramics have been curiously absent from discussions of household assemblage variability. Traditional Maya ceramic analyses have focused on constructing chronologies (e.g., Gifford 1976; Smith 1955), interpreting iconographic elements (e.g., Reents Budet and Ball 1994) and identifying imported vessels and trade wares, but few studies explicitly address how pottery functioned in daily life (but see Howie 2012; Lucero 1994, 2001). Maya pottery assemblages are known for their diversity and complexity, and pottery from domestic contexts present no exception to this rule: household assemblages reveal variability that has not been fully explored. We argue that ceramics possess great potential to address fundamental issues of Maya Figure 1 El Pilar and the Maya Lowland s. (Figure by Ford). household diversity, even as pottery remains an underexploited resource because of the historical focus on chronological variability. We have examined formal, functional, and technological aspects of ceramic inventories to characteriz e Late Classic (c. AD 600900) Maya household assemblages and in this paper, initiate the exploration of household belongings. Different vessel forms correspond to different

PAGE 55

Classic Maya Household Ceramic Belongings 54 functions: cooking, serving, storage, and transport. Analysis of functional vari ability can clarify important differences in domestic practice and address the quality and quantity of ceramics used by different residential units. Describing the formal aspects of each collection is the first step. With a basic formal and functional characterization of ceramic collections, we can compare consumption patterns and investigate how households used pottery. Our research assemblages derive from household middens and construction fill in the El Pilar area of the upper Belize River, and each c ontext dates to the Late Classic ( Figure 1 ). The household materials represent the range of environmental and social contexts of the area and provide a broad baseline for comparing the socioeconomic relationships encoded in ancient pottery. The Potential of Ceramics for Understanding Maya Households Research on Maya residential patterns dates the work of Ricketson and Ricketson at Uaxactun (1937), at a time when archaeologists focused on defining chronologies and excavating temples (Taylor 1967). Work by Gordon Willey and colleagues (1965) at Barton Ramie began an important shift in focus toward the examination of Maya households and established a scholarly tradition focused on defining settlement patterns and site rank hierarchies (e.g., Ashmore 1981; Ar nold and Ford 1980; Robin 2012; Puleston 1973; Rice 1976; Webster and Gonlin 1988, among others). Settlement pattern surveys have since identified numerous small structures around monumental city centers, which are interpreted as representing comparable r esidential units; just as we are involved with at El Pilar. Maya residences are presented as single structures or architectural groups associated with open domestic spaces, and they provide opportunities to compare household wealth, access to resources, and relative social status (see Ford 1990, 1991, 1992; Robin 2012). Data collected at the household level represent the most basic social units in ancient Maya society and have the potential to answer questions of resource acquisition and social reproduction (Wilk and Rathje 1982). Households also represent the essential c ontexts of daily life for elite and general members of preindustrial communities. Late Classic Maya residences were the settings for the variety of domestic activities the storage, preparation, and serving of food important among them that are reflect ed in data gathered methodologically by archaeologists. The remnants of ancient household belongings are dominated by pottery and stone tool fragments, materials that are ubiquitous across the settled landscape of the Maya. Maya household studies have focused on relationships between the size and frequencies of the small structures and associated exotic artifacts with variable distributions (see Robin 2012). Studies suggest residential unit size relates to material consumption, with larger structure containing more material and exotic artifacts than smaller ones (Abrams 1994; Ford 1991, 1992; Ford and Fedick 1992; Ford and Olson 1989). Recent studies show that domestic artifacts were more diverse than had been assumed in the past, and that small households could access what archaeologists presumed were sumptuary items (Ford 2010; Ford and Olson 1989; LeCount 1996; Robin 2012, 2013; Wiewall and Howie 2010). There are tantalizing indications of great diversity among everyday Maya farmers. Examinations of nonlocal resources such as obsidian or green stone (e.g., Asaro et al. 1978; Ford et al. 1997; Horn 2015, among others), production of chert tools or marine shell ornaments (e.g., Hohmann 2002; Shafer and Hester 1983), and the distributions of other rare ma terials (e.g., granite, slate) reveal variability in household consumption. The excavations of multiple residences at Chan, in the upper reaches of the Belize River area, reveal this variability at the community scale (see chapters in Robin 2012). Detail ed descriptions of ceramic household inventories, however, are notably absent from these comparisons, which leaves an important segment of the material record unexplored. The archives of data on resource procurement, production, distribution, and consumpti on contained in pottery have tremendous potential to reveal different scales of socioeconomic interactions, from household

PAGE 56

Horn III and Ford 55 endeavors to regional interactions (Arnold 1985). Macroscopic fabric analyses document variability in materials and composition, aid ing in the reconstruction of socioeconomic networks that link producers and consumers (Horn 2015). Petrographic and compositional analyses can be used to identify resources accessed for production and technological traditions of potters (e.g., Ford and Spera 2007; Howie 2012; L pez Varela 2005; Shepard 1955). We can employ these macro and microscopic analyses to investigate household resource use, participation in economic systems, and knowledge of local environments for ceramic collections from El Pilar and the upper Belize River area. Ceramics provide an untapped reservoir of comparative information on everyday household practices and common activities. Collections from excavations in the El Pilar area contain 236,388 cataloged ceramic artifacts weighi ng 3,418 kilograms. These collections primarily include vessel fragments (identifiable rims, bodies, bases, jar necks, handles, and pods), but they also contain ornamental and utilitarian items (figurines, spindle whorls) and amorphous baked clay objects (briquettes, daub) in smaller numbers. We have recorded 1.8 ceramic objects for every chert artifact, and the total weights of ceramics and chert objects are nearly identical. Ceramics were as important as other materials to completing household tasks, a nd yet the ways vessels functioned in Classic Maya households have been widely overlooked. Documentation of ceramic belongings is basic to understanding the domestic activities that structured the daily lives of farming household members in Late Classic t imes. Our formal/functional assemblage descriptions allow comparisons that reflect activities prevalent at individual households. From Chronology to Possessions A New Direction for Maya Ceramic Studies Early Maya ceramic studies established chronologic al markers to facilitate dating monumental construction sequences (e.g., Thompson 1939). The Carnegie Institution of Washingtons excavations at Uaxactun, Guatemala (Smith 1955), produced a detailed sequence of stylistic horizons Mamom, Chicanel, Tzakol and Tepeu that are essential to temporal comparisons across lowland Maya sites to this day. The Uaxactun analysis insightfully classified ceramics by vessel form ( Figures 2 4) and recognized relationships between form and function (see Shepard 1963). Formal and functional categories provide a solid basis for comparing assemblages from different sites, but comparative studies developed to consider detailed stylistic features instead (Smith and Gifford 1966). James Gifford (1976) advanced applications of the type variety classification system, which continues to dominate Maya ceramic studies. Giffords application began with his analysis of pottery from Barton Ramie, a well known domestic settlement along the Belize River. Type variety was developed to analyze ceramics in the American Southwest (Colton 1953; Wheat et al. 1958), and the system was refined in the Maya area in a way that elevated stylistic attributes, such as surface treatments (e.g., slips, washes, burnishing ), over characteristics more closely related to function, such as vessel form. This method privileges the most elaborately decorated vessels and largely ignores formal and functional characteristics (but see LeCount 1996 ; Lucero 2001 ). Attention to surfa ce decoration relegates large numbers of unslipped fragments, primarily representing utilitarian ceramics that were important to everyday household activities, to simple sherd counts of varieties unspecified, if they are counted at all. Defining types by surface treatment can subsume a range of form likely representing vessels used for distinct purposes in different social contexts into a single type, as is the case with Late Classic Belize Red:Belize Red variety pottery, which is common in the eastern Maya lowlands ( Figure 5 ). Attention to surface treatment, a temporally dynamic stylistic attribute in Maya pottery, makes type variety effective for assessing occupation sequences at sites. Type variety assessments are designed to provide a common terminology for constructing chronologies at the regional scale and are a useful assessing shared styles (e.g., Ball 1977; Kosakowsky 2012). Tabulations of type frequencies are not helpful in assessing vessel function, however, a necessity for understanding

PAGE 57

Classic Maya Household Ceramic Belongings 56 Figure 2 Jars from Uaxactun (Figure by Ford, after Smith 1955: Figure 47b). Figure 3 Bowls from Uaxactun (Figure by Ford, after Smith 1955: Figure 48a, 48b). Figure 4 Plates from Uaxactun (Figure by Ford, after Smith 1955:51b). the daily domestic activities of household members. We need new frames of reference, contextualized within the human ecology of ceramic production and consumption that regard vessels as finished pro ducts that were circulated and used by ancient Maya households. We can describe the formal and functional characteristics of Late Classic Maya household ceramic assemblages to begin moving beyond chronology and into an investigation of daily life. Vessel forms and shapes provide clues to everyday activities, and we can examine variability among households by comparing archaeological data from different residential contexts. We may expect similarities among cooking vessels, for example, as cooking w ould

PAGE 58

Horn III and Ford 57 Figure 5 Belize Red:Belize Red variety vessels from Barton Ramie (after Gifford 1976: Figure 162). Figure 6 Survey Transects and Settlement in the El Pilar Area. Figure by Ford. be a common activity shared among households, while concurr ently expecting the quality and size of serving vessels to vary with wealth and status. Comparative analyses such as these, however, and even a basic definition of what comprises a typical household ceramic inventory cannot be accomplished with type var iety methods. Our study takes a first step toward unlocking the great potential of Maya ceramics to explore ancient daily routines by presenting a comparative baseline, with descriptions, of household pottery assemblages from El Pilar Upper Belize River area. Defining pottery assemblages is the foundation for comparing household belongings. These data can reveal common and unique activities among different households and how these relate to independent wealth indicators, such as exotic artifact frequencie s and house size. Were household assemblages essentially the same in kind, regardless of wealth? Were basic items, equivalent to service sets, available to all households? If we consider the frequencies and diversity of vessel forms and their associated functions, we will come closer to answering these fundamental questions. Household Belongings of the Late Classic Maya in the El Pilar area Our ceramic collections derive from survey and excavations in the El Pilar area north/northwest of the upper Beliz e River. The El Pilar area occupies an ecotone between western limestone ridges and eastern coastal plains, and we classify its topography into three environmental zones the valley, foothills, and ridgelands. These distinctive landforms relate to the g eology of the area, and differences in soil quality and agricultural potential characterize each zone and the available resources that would have impacted household subsistence and wealth (Fedick 1995; Ford and Fedick 1992). Based on research conducted by Ford and colleagues between 1983 1993, e xcavation data were based on surveys of over 600 hectares and 400 residential units conducted by Ford and colleagues between 1983 1999 (Ford 1990, 1991, 1992). One eighth of the residential units (n = 48) were sele cted for testing in a stratified random sample based on distance from the river, and eleven of these households representing a range of large, medium, and small residential units of the environmental zones along with specialized production areas were s elected for full scale excavation from the three environmental zones ( Figure 6 ). We focus on assemblages recovered at residential units that are dominated by Late Classic pottery to investigate the household belongings of the Maya. Our formal study is ong oing, and the integration of detailed microscopic and technological studies are

PAGE 59

Classic Maya Household Ceramic Belongings 58 Table 1 Diagnostic Late Classic Vessel Forms from the El Pilar Area. Table by Ford. Figure 7 Basic Vessel Forms in the El Pilar Area (Figure by Ford). planned for future research. We targeted basic vessel form, and varia tion in vessel shape and size, to achieve a functional perspective in this study. In describing the merits of functional analyses, the pioneering ceramicist Anna O. Shepard (1963:224) once said: The study of vessel shape can be approached from the standpoint of function, esthetics or taxonomy. Function has the appeal of human interest; the purpose of the vessel tells us something of the activities and customs of the people who used them. Bearing this in mind, we examine the ceramic assemblages fro m the El Pilar area and ask what these remains can tell us about household belongings and agency in daily life of Late Classic times. Vessel Forms of the Late Classic Maya Ceramic Belongings Our ceramic assemblage datasets include all sherds collected fro m residential unit excavations. We identified 9,651 diagnostic ceramic sherds1 to assess household inventories and subjected all diagnostic sherds to an attribute analysis to delineate variation among the assemblages. Distinct attributes were evaluated and measured for diagnostic sherds and were tabulated for analysis. Vessel form categories bowls, jars, plates, and vases were divided into common shapes by rim and profile characteristics whenever possible. Formal and analytical attributes provide basic descriptions for the El Pilar ceramic collections, and we have worked to ensure these data are comparable with Period Assessment (relative chronology) Temper Uniformity (evenness, spatial/size distribution) Ceramic Form (general vessel category) Temper Percent (relative abundance) Shape (rim profile characteristics) Firing Description (color; presence/absence firing horizons) Rim Diameter (cm) Surface Description (absence/ presence of slips/paints/washes) Munsell Slip Color (value, chroma, hue) Design Elements (absence/presence nature/type decorative embellishments) Munsell Paste Color (value, chroma, hue) Wall Thickness (mm) Pocking (spalling; removal of carbonate inclusions from surface) Weight (gm) HCl Reaction (identify carbonate -bearing ceramic bodies)

PAGE 60

Horn III and Ford 59 Table 2 Late Classic General Form and Common Shape Diagnostics from Full -Scal e Excavations (Table by Ford). Residential Unit Geographic Zone Household Size Form Diagnostics Shape Diagnostics 281 -21 Valley Medium 2254 1288 278 -26 Valley Small 862 397 278 -66 Foothills Small 190 104 272 -220 Foothills Medium 553 344 272 -182 Foothills Medium 512 280 272 -168 Ridgelands Small 72 35 272 -162 Ridgelands Medium 3 0 272 -145 Ridgelands Large 1098 593 272 -136 Ridgelands Large 403 230 272 -032 Ridgelands Small 1 1 272 -025 Ridgelands Large 2864 1768 Total 8812 5040 Figure 8 Common Shapes in Late Classic Assemblages in the El Pilar Area (Figure by Ford). other analytical systems by using objective and standardized attribute descriptors, including Munsell colors and Udden Wentworth grain sizes. For each catal ou ged sherd that was assigned a period assessment, f ifteen attributes were compiled for each sherd that was assigned a temporal assessment ( Table 1 ). Our comparative analyses are ongoing, and we present preliminary results of common Late Classic vessel sha pes. We aim to demonstrate the importance of this approach to characterizing Maya ceramics and its potential for comparative analyses at multiple scales. Vessel form categories relate to function and reflect the range of household activities in the El Pilar area. Table 1 and Figure 7 illustrate the relative frequencies of vessel forms. It is clear that these formal categories occur in differing proportions, suggesting variation in function and use. The low percentage of vases (5%), for example, suggests they served more restricted functions or saw more restricted use than bowls (40%). We can compare this composite picture to assemblages across the Maya lowlands to look for differences in activities, and it serves as a baseline for

PAGE 61

Classic Maya Household Ceramic Belongings 60 Table 3 Diagnostic Late Classic Vessel Shapes by Geographic Zones (Table by Ford). Residential Unit Household Size Bowls Jars Plates Vases Total N % N % N % N % 281 -021 Medium 433 34 429 33 297 23 129 10 1288 278 -026 Small 176 44 117 30 76 19 28 7 397 Valley Totals 612 36 546 33 373 22 157 9 1685 278 -066 Small 29 28 59 57 6 6 10 9 104 272 -220 Medium 174 50 77 22 43 13 50 15 344 272 -182 Medium 78 28 67 24 106 38 29 10 280 Foothills Totals 281 39 203 28 155 21 89 12 728 272 -168 Small 16 46 16 46 0 0 3 8 35 272 -162 Medium 0 0 0 0 0 0 0 0 0 272 -145 Large 142 24 303 51 78 13 70 12 593 272 -136 Large 77 33 119 52 19 8 15 7 230 272 -032 Small 0 0 1 100 0 0 0 0 1 272 -025 Large 567 32 920 52 205 12 76 4 1768 Ridgelands Totals 802 31 1359 52 302 11 164 6 2627 Total 1692 34 2108 42 830 16 410 8 5040 comparing individual household assemblages in the El Pilar area (Table 1, Figure 7 ). Using the attributes discussed above, we identified and defined common Late Classic rim shapes within our broader form categories and assigned about 60% (n = 5040) of our diagnostic sherds to a shape class ( Table 2 Figure 8 ). Preliminary Household Comparisons It is no surprise that assemblages varied in size by residential unit across the El Pilar area. Noteworthy are the residential units 272162 and 272032, which yielded almost no Late Classic sherds ( Table 2 ). Residential units in the valley, for example, have larger quantities of diagnostic sherds than comparably sized households in the foothills, indicating higher levels of pottery use in the Late Classic. The larger sample of residential units from the ridgelands shows more variability in pottery use that requires additional inv estigation ( Table 3 ). Basic vessel shape distribution reveals additional patterns of pottery use and activities. At the highest level of comparison that of the geographic zone we see similarities in the percentages of bowls that make up household assemblages. This pattern likely reflects the multiple uses bowls can play in preparing and serving food regardless of relative household wealth or environmental zone. Jars varied more by environmental zone and made up about half of ridgeland household assem blages, which may relate to the abilities of those households to acquire and store food. Plates accounted for a higher proportion of household from valley and foothills residences than those in the ridgelands. While these vessels are associated with food service and would be expected in higher proportions at larger households such as those in the ridgelands, their distribution may be related to production as well, but this is difficult assess (Ford and Lucero 2000). The presence of vases at households of all sizes and in every environmental zone suggests these vessels, often thought to be wealth indicators, were not restricted to specific segments of society. Finally, household assemblages in the valley and foothills more closely resemble each other than those from the ridgelands. Residential units in the foothills showed the greatest variability in assemblage composition. The small household 278066, for example, possessed relatively more jars and fewer plates than other households in the foothills. Hou sehold 272182 yielded majority

PAGE 62

Horn III and Ford 61 plates, which may reflect greater involvement in social events involving food service or possibly production activities. Late Classic ridgeland household assemblages varied more in scale than in kind. Three of four resident ial inventories contained high proportions of jars compared to the total collection, but the assemblage associated with residence 272 168 was less diverse than the other ridgeland households. This assemblage contained equal, but small, numbers of jars and bowls, yet no plates, suggesting a limited range of cooking and serving activities occurred there. Despite the lack of plates, 272168 yielded three vase fragments, which was unexpected at a small household with a limited ceramic inventory. More diverse assemblages from the larger residential units 272136 and 272025 contained lower percentages of vase fragments, further complicating interpretations of the role vases played in Late Classic daily life. Conclusions and Directions for Future Research Stud ies of ancient Maya pottery have focused primarily on building chronologies for much of the past century, and we are now well positioned to move forward toward more nuanced understandings of Maya household activities (see Lucero 2001; Pagliaro 2011) Our preliminary comparisons show the benefit of functional ceramic analyses to address questions of household archaeology and ceramic ecology. We suggest that detailed attribute analyses of residential pottery assemblages are vital to understanding activities in general and domestic activities in particular of the ancient Maya. Additional research on the functional variability revealed by these analyses will produce information critical to understanding resource use, economic systems, and daily activities tha t characterized life in the Late Classic period. We are working to incorporate data that reflect differences in food storage, cooking, and service practices among the different households into our continuing settlement pattern studies, which have been focu sed on mapping the social and environmental geography of the El Pilar area revealing household wealth differentiation and distribution. We have used a combination of LiDAR imagery and traditional field survey techniques over the past five years to map var iations in settlement patterns that are the initial step toward understanding household differentiation. The next phase involves examining the make up of household belongings and how they vary across socioeconomic and geographic contexts. How does household differentiation relate to environmental zone and participation in different social events and economic networks in different areas? Comparative study, incorporating both stylistic and technological attributes, will refine investigations of household c eramic assemblages and the activities responsible for the variability evident in our preliminary analyses. This line of inquiry holds great potential for explaining complexity in Late Classic Maya agrarian communities. Only by investigating variability of household belongings will we gain a full picture of the importance of ceramics to Late Classic daily life. 1We considered sherds diagnostic if they provided information on basic vessel form. References Cited AbramAbrams, Elliot M. 1994 How the Maya Built Their World: Energetics and Ancient Architecture 1 vols. University of Texas Press, Austin, Austin. Arnold, Dean E. 1985 Ceramic theory and cultural process. New Studies in Archaeology Cambridge University Press, New York. Asaro, F., H. V. Michel R. Sidrys and F. Stross 1978 High -Precision Chemical Characterization of Major Obsidian Sources in Guatemala. American Antiquity 43(3):436443. Ashmore, Wendy (editor) 1981 Lowland Maya Settlement Patterns 1st ed. University of New Mexico Press, Albuquerque, New Mexico. Ball, Joseph W. 1977 The Archaeological Ceramics of Becan, Campeche, Mexico Tulane University. Copies available from 43. Blackmore, Chelsea 2012 Recognizing Diffence un Small Scale Settings: An Examination of Social Identity Formation at the Northeast Group, Chan. In Chan: An Ancient Maya

PAGE 63

Classic Maya Household Ceramic Belongings 62 Farming Community edited by C. Robin, pp. 173191. University of Florida Press, Gainesville, Florida. Colton, Harold S. 1953 Potsherds: An Introduction to the Study of Prehistoric Southwestern Ceram ics and Their Use in Historic Reconstruction The Northern Arizona Society of Science and Art, Flagstaff, Arizona. Fedick, Scott L 1995 Land Evaluation and Ancient Maya Land Use in the Upper Belize River Area, Belize, Central America. Latin American Antiquity 6(1):1634. Ford, Anabel 1990 Maya Settlement in the Belize River Area: Variations in Residence Patterns of the Central Maya Lowlands. In Prehistoric Population History in the Maya Lowlands edited by T. P. Culbert and D. S. Rice, pp. 167 -181. University of New Mexico Press, Albuquerque, New Mexico. 1991 Economic Variation of Ancient Maya Residential Settlement in the Upper Belize River Area. Ancient Mesoamerica 2:35 46. 1992 The Ancient Maya Domestic Economy: An Examination of Settlement in the Upper Belize River Area. Paper presented at the Primer Congreso Internacional de Mayistas, San Cristobal. 2010 Conspicuous Production of Exotics Among the Maya: The Organization of Obs idian Procurement, Production, and Distribution at El Pilar. In Produccin de Bienes de Prestigio Ornamentales y Votivos de la Amrica Antigua edited by E. M. Tisoc, R. S. Ciriaco and E. G. Licn, pp. 111 -130. Syllaba Press, Doral. Ford, Anabel and Jeann e E. Arnold 1982 A Reexamination of Labor Investments at Tikal: reply to Haviland, and Folan, Kintz, Fletcher and Hyde. American Antiquity 47(2):436 440. Ford, Anabel and Scott L. Fedick 1992 Prehistoric Maya Settlement Patterns in the Upper Belize River Area: Initial Results of the Belize River Archaeological Settlement Survey. Journal of Field Archaeology 19:3549. Ford, Anabel and Lisa Lucero 2000 The Malevolent Demons of Ceramic Production: Where Have All the Failures Gone? Estudios de Cultural Maya 2 1:57-74. F ord, Anabel and Kirsten Olson 1989 Aspects of Ancient Maya Household Economy: Variation in Chipped Stone Production and Consumption. In Prehistoric Maya Economies of Belize, edited by P. A. McAnany and B. L. Isaac, pp. 185-211. Research in Econo mic Anthropology Supplement no.4. vol. 4. JAI Press, Greenwich. Ford, Anabel and Frank Spera 2007 Fresh Volcanic Glass Shards in the Pottery Shards of the Maya Lowlands. Research Reports in Belizean Archaeology 4:111 118. Ford, Anabel, Fred Stross, Frank Asaro and Helen V. Michel 1997 Obsidian Procurement and Distribution in the Tikal -Yaxha Intersite Area of the Central Maya Lowlands. Ancient Mesoamerica 8:101110. Gifford, James C. 1976 Prehistoric Pottery Analysis and the Ceramics of Barton Ramie in the Belize Valley. Memoirs of the Peabody Museum of Archaeology and Ethnology Harvard University; v. 18. Peabody Museum of Archaeology and Ethnology Harvard University, Cambridge. Hohmann, Bobbi M. 2002 Preclassic Maya Shell Ornament Production in the Belize Valley, Belize, Anthropology, University of New Mexico, University Microfilms, Ann Arbor. Horn III, Sherman W. 2015 The Web of Complexity: Socioeconomic Networks in the Middle Preclassi c Belize Valley. Ph.D dissertation, Department of Anthropology, Tulane University, New Orleans. University Microfilms, Ann Arbor. Howie, Linda A. 2012 Ceramic Change and the Maya Collapse: A Study of Pottery Technology, Manufacture and Consumption at Lama nai, Belize. BAR International Series 2373. Archaeopress, Oxford, UK. Keller, Angela 2012 Creating Community with Shell. In Chan: An Ancient Maya Farming Community edited by C. Robin, pp. 253270. University Press of Florida, Gainesville. Kosakowsky, La ura J. 2012 Ceramics and Chronology at Chan. In Chan: An Ancient Maya Farming Community edited by C. Robin, pp. 42-70. University Press of Florida, Gainesville. LeCount, Lisa Jeanne 1996 Pottery and Power: Feasting, Gifting, and Displaying Wealth among the Late and Terminal Classic Lowland Maya. Ph.D. thesis, Anthropology, University of California, Los Angeles, Los Angeles, California.

PAGE 64

Horn III and Ford 63 Lpez Varela, Sandra L. 2005 The Microcosmos of For mative Pottery from K'axob, Belize. In New Perspectives on Formative Mesoamerican Cultures edited by T. G. Powis, pp. 159 -170. vol. BAR International Series 1377. Archaeopress, Oxford. Lucero, Lisa Joyce 1994 Household and Community Integration among Hinterland Elites and Commoners: Maya Residential Ceramic Assemblages of the Belize River Area Doctoral Dissertation, Archaeology Program, Philosophy in Archaeology, University of California, Los Angeles, Los Angeles. 2001 Social Integration in the Ancie nt Maya Hinterlands: Ceramic Variability in the Belize River Area Anthropological Research Paper No. 53. Arizona State University, Tempe, Arizona. Netting, Robert McC., Richard R. Wilk and Eric J. Arnould (editors) 1984 Households: Comparative and Histor ical Studies of the Domestic Group University of California Press, Berkeley, California. Pagliaro, Jonathan B. 2011 Rethinking Thinking: An Ecological Based Consideration of Contextual Diversity in Late Classic Ceramic Assemblages f rom the Upper Belize R iver Area, Belize. Ph.D. Thesis, Anthropology, Southern Methodist University, Dallas, Texas. Puleston, Dennis Edward 1973 Ancient Maya Settlement Patterns and Environment at Tikal, Guatemala: Implications for Subsistence Models. Ph.D. thesis, Department o f Anthropology, University of Pennsylvania, Philadelphia, Pennsylvania. Reents -Budet, Dorie and Joseph W. Ball 1994 Painting the Maya Universe: Royal Ceramics of the Classic Period Duke University Press in association with Duke University Museum of Art, Durham, North Carolina. Rice, Don S. 1976 The Historical Ecology of Lakes Yaxh and Sacnab, El Petn, Guatemala. Ph.D. dissertation, Anthropology, Pennsylvania State University, University Park, State College, Pennsylvania. Ricketson, Oliver Garrison and Edith Hill Bayles Ricketson 1937 Uaxactun, Guatemala: Group E -19261931 Washington: Carnegie Institution of Washington. Robin, Cynthia (editor) 2012 Chan: An Ancient Maya Farming Community University Press of Florida, Gainesville, Florida. 2013 Everyday Life Matters: Maya Farmers at Chan. University Press of Florida, Gainesville, Florida. Shafer, Harry J. and Thomas R. Hester 1983 Ancient Maya Chert Workshops in Northern Belize, Central America. American Antiquity 48:519-543. Shepard, Anna O. 1963 Ceramics for the Archaeologist Carnegie Institution of Washington, Washington, D.C. Smith, Michael E. 1987 Household Possesion and Wealth in Agrarian States: Implications for Archaeology. Journal of Anthropological Archaeology 6:297335. Smith, Rob ert E. 1955 Ceramic Sequence at Uaxactun, Guatemala 1. 2 vols. Middle American Research Institute, Tulane University, New Orleans, Louisiana. Smith, Robert E. and James C. Gifford 1966 Maya Ceramic Varieties, Types, and Wares at Uaxactun: Supplement to "C eramic Sequence at Uaxactun, Guatemala". Tulane University. Taylor, Walter W. 1967 A Study of Archaeology Southern Illinois University Press, Carbondale, Illinois. Thompson, J. Eric S. 1939 Excavations at San Jose, British Honduras Publication 506. Carn egie Institution of Washington, Washington DC. Wauchope, Robert 1938 Modern Maya Houses: A Study of their Archaeological Significance 1st ed. Carnegie Institution of Washington, Washington, D.C. Webster, David L. and Nancy Gonlin 1988 Household Remains of the Humblest Maya. Journal of Field Archaeology 15(2):169190. Wiewall, Darcy and Linda Howie 2010 A Synthesis of Ceramic Production and Consumption at Lamanai, Belize, during the Postclassic to Spanish Colonial Periods. Research Repo rts in Belizean Archaeology 7:201217. Wheat, Joe Ben, James C. Gifford and William W. Wasley 1958 Ceramic Variety, Type Cluster, and Ceramic System in Southwestern Pottery Analysis. American Antiquity 24(1):34 -47. Wilk, Richard R. and William L. Rathje 1982 Household Archaeology. American Behavioral Scientist 25(6):617639.

PAGE 65

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 6578 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 6 RECONSTRUCTING PRECLASSIC MAYA HOUSEHOLD ECONOMIES IN THE BEL IZE RIVER VALLEY Claire E. Ebert and Jaime J. Awe Archaeologists traditionally attribute the emergence of socio-economic inequality to elite control of local craft production and regional redistribution systems. Households also employed a diverse set of economic strategies to access raw materials and fin ished craft items that formed the foundations of the domestic economy. This study uses geochemical sourcing of obsidian and ceramics to characterize the economic behaviors that structured the domestic economy at the ancient Maya community of Cahal Pech, l ocated in the Belize Valley of west central Belize, during the Preclassic Period (1200 BC AD 300). Portable X-ray systems based on the di fferential consumption of source material, which developed during the Preclassic and persisted through the Terminal Classic Period (~1200 cal BC -cal AD 900). Instrumental neutron activation analysis (INAA) of Preclassic ceramics from Cahal Pech identified contrasting provisioning strategies based on long-distance and local ceramic exchange monopolized by some households. Understanding the irregular distribution of economically important resources between households can shed light on the social and economi c contexts that led to the emergence of institutionalized hereditary inequality. INTRODUCTION Archaeologists have long focused on exploring the dynamics of prehistoric economies because production, distribution, and consumption of resources are embedded within larger social processes. Researchers examining the emergence of social and economic stra tification in Mesoamerica during the Formative (Preclassic) Period have argued that elite status was maintained through the monopolization of regional distribution of specialized crafts (e.g., Clark 1987:280; Santley 1984). The redistribution of these goods by aggrandizing elites generated economic and social debt for subordinate members of society, resulting in transgenerational social hierarchies (Clark and Blake 1994). More recently, studies of ancient Mesoamerican economies have shifted their focus to examine the structural and functional aspects of household economic organization to examine broader socioeconomic developments (e.g., Douglas and Gonlin 2012; Hirth 2009, 2016). Households were the most basic economic unit in ancient Mesoamerican societ ies (Ashmore and Wilk 1988; Wilk and Rathje 1982), and the domestic economy, which was structured for the acquisition of subsistence resources by all segments of society, formed the foundation upon which all other economic activities were based (Hirth 2012 2016). Ranked social organization developed when a household, or group of households, formalized their economic wellbeing into social status and authority through the institutional economy, which underwrote the social, political, and religious Figu re 1 Map of the Belize Valley with locations of Preclassic period sites mentioned in text. Inset map shows locations of obsidian sources identified from Preclassic contexts in the Belize Valley region. activities for the society as a whole (Hirth 2016:2 1). The irregular distribution of subsistence and nonsubsistence resources probably encouraged economic variation and may have motivated some households towards production and distribution of specific resources intended to improve not only their househol d well being, but also their status within the community. In this study, we use geochemical compositional data of obsidian and ceramic artifacts from the lowland Maya community of Cahal Pech, located in the Belize Valley of western Belize ( Figure 1 ), to examine the structure and function of the domestic and institutional economies during the Preclassic Period (~1200 BC

PAGE 66

Reconstructing Preclassic Maya Household Economies 66 Figure 2 Chronological periods for Cahal Pech with associated ceramic phases. AD 300; Figure 2 ). The Preclassic represents a critical transition in Maya prehistory, when the development of settled village life, increased reliance on maize agriculture, and the adoption of ceramic technology appeared across the lowlands. By the Late Preclassic, M aya society had become complex and hierarchical, with small village settlements developing into large centralized polities serving as the focal points of economic and political activity (Chase and Chase 2012). While evidence exists for the long distance m ovement of many different commodities into Belize Valley during this dynamic period (e.g., greenstone, Powis et al. 2016; granite, Tibbits 2016), our focus is on obsidian and ceramics because they are ubiquitous in all contexts and were essential for the d aily subsistence of the majority of Maya households through time. Documenting distributional patterns of these key items within and between households can be used to determine how access to overlapping and contrasting economic networks may have impacted household wealth and status beginning in the Preclassic Period. To understand differences in Preclassic economic networks associated with the development in socioeconomic inequality at Cahal Pech, we performed technological and portable X ray fluorescence (pXRF) geochemical sourcing of obsidian artifacts ( n =1189) from the sites civicceremonial site core and peripheral household groups. The results indicate that all households relied primarily on imported obsidian blades from sources in the southern highl ands of Guatemala. El Chayal obsidian dominated assemblages from Preclassic domestic contexts, a pattern that persisted until the abandonment of Cahal Pech in the Terminal Classic (~AD 850/900). Differential use of source materials between households, however, suggests that obsidian was obtained through decentralized domestic procurement systems from the Preclassic through Terminal Classic periods. Comparisons of our data from Cahal Pech to previously sourced assemblages from the sites of Blackman Eddy and Chan in the Belize Valley indicate this pattern was present across the region. Instrumental neutron activation analysis (INAA) of ceramics ( n =192) from radiocarbon dated Early to Late Preclassic Period deposits in the site core and two peripheral domes tic groups identified contrasting long distance and local economic networks. INAA identified seven compositional groups corresponding to changing production patterns. By the Middle and Late Preclassic, the ceramics from higher status households were compositionally distinct when compared to peripheral household settlements. Comparative analysis of ceramic assemblages from Cahal Pech and sites in the central Petn region of Guatemala suggest that Mars Orange wares were exchanged between high status groups Ceramic exchange may have been one avenue for Maya households to underwrite economic status within a developing institutional economy. The Preclassic Cahal Pech Economy Cahal Pech is a medium sized center located ~2 km south of the confluence of the Ma cal and Mopan Rivers in the Belize Valley ( Figure 3 ). Stratigraphic excavations and radiocarbon dating conducted by the Belize Valley Archaeological Reconnaissance (BVAR) Project in the Cahal Pech epicenter at Str. B4 and Plaza B indicate the site was first settled between ~1200 1100 cal BC as a

PAGE 67

Ebert and Awe 67 Figure 3 Map of Cahal Pech showing the civic ceremonial site core (top) and location of house groups (bottom) sampled for obsidian and ceramic compositional studies.

PAGE 68

Reconstructing Preclassic Maya Household Economies 68 small agrarian village composed of economically autonomous households (Awe 1992; Ebert 2017; Ebert et al. 2017). The earliest residential occupation is associated with the appearance of Cunil complex cer amics (~1200/1100900 BC), the majority of which are utilitarian wares including jars, bowls, and gourdshaped tecomates (Sullivan and Awe 2013). The Cunil assemblage also contains decorated serving vessels including slipped bowls, plates, and censers dep icting kan cross, avian serpent, and flamed eyebrow designs (Awe 1992; Garber and Awe 2009; Sullivan and Awe 2013). The presence of El Chayal obsidian flakes and nodules in the earliest levels at Cahal Pech indicate integration of the Belize Valley into broader regional economic networks (Awe 1992; Awe and Healy 1994; Ebert 2017; Stemp et al., this volume). Population expansion and economic growth at Cahal Pech and other Belize Valley sites during the Middle Preclassic were accompanied by the construction of public architecture restricted to larger house groups, signaling the emergence of higher status individuals within local communities. The appearance of increasingly standardized ceramics and evidence for the expansion of long distance exchange network s dealing in exotic items also appear at this time. Obsidian blades, decorated pottery, jade, and other valuables have been identified throughout the Belize Valley region in Middle Preclassic contexts (Awe 1992; Hohmann 2002; Kersey 2006; Powis et al. 2016). At Cahal Pech, the Middle Preclassic Kanluk ceramic complex (900350 BC) was composed primarily of coarse paste utilitarian ceramics (Jocote Orangebrown) and fine paste Mars Orange serving wares including slipped Savana Orange and Reforma Incised type s (Awe 1992; Gifford 1976; Horn 2015; Peniche May 2016). Little archaeological evidence exists for the centralized control of production or redistribution of imported items by higher status groups at Cahal Pech. While Awe and Healy (1994) documented a tr ansition in obsidian technology in the Middle Preclassic assemblage towards finished prismatic blades, obsidian coming from both the El Chayal and San Martn Jilotepeque (SMJ) sources were consumed differentially between households at the site (Awe and Hea ly 1994; Peniche May 2016). Other crafting activities, such as shell bead production (e.g., Hohmann 2002; Lee and Awe 1995; Peniche May 2016), connected some Cahal Pech households with different longdistance exchange networks. Bead production, however, i s not evenly distributed across the site, likely indicating that acquisition of shell via long distance exchange was directly regulated by individual households. The Late Preclassic Period (350 BC AD 300) saw the fluorescence of large civic ceremonial cent ers throughout the Belize Valley and evidence for the development of institutionalized elite rulership at Cahal Pech and other Beli z e Valley sites such as Blackman Eddy, Xunantunich, and Barton Ramie (Awe 1992; Brown et al. 2013; Garber et al. 2004; Healy et al. 2004; Willey et al. 1965). The construction of elaborate tombs and offerings within monumental temple architecture appeared at Cahal Pech during the corresponding Xakal ceramic phase, signaling the development of a royal lineage at the site (Awe 1992; Awe n.d.; Garber and Awe 2009; Healy et al. 2004). The presence of symbolically significant items such as high quality jade crafts within burials suggest that high status individuals were involved in exchange of exotic items that were translated into wealth and prestige. Evidence for status differentiation appears also within the settlement zone after 350 BC in the form of largerscale domestic and non domestic architectur e. Radiocarbon dates of construction phases indicate that low masonry platforms and temple structures were built at several house groups around the Cahal Pech site core (e.g., Tzutziiy Kin, Zopilote, Zubin, and Cas Pek groups), which likely functioned as public temple buildings associated with nearby domestic structures (Ebert et al. 2016, 2017). Radiocarbon dates from burials and ceramic associations from several other large house groups suggest that this pattern of social, economic, and spatial growth occurred throughout the Cahal Pech hinterlands during the Late Preclassic (Ebert 2017; Ebert et al. 2017). Obsidian pXRF Analyses Obsidian samples analyzed in this study were derived from surface collection and stratified contexts within the Cahal Pech mo numental site core, and from ten residential groups located throughout the sites periphery. A total of 1189 artifacts were subjected to pXRF geochemical sourcing analyses. Temporal assignments are based on relative ceramic associations and radiocarbon dates where possible (see Awe 1992;

PAGE 69

Ebert and Awe 69 Ebert 2017; Ebert et al. 2016, 2017). Technological analyses of the assemblage were performed by Cassana Popp and Ken Hirth in the Mesoamerican Economy and Archaeology Lab at The Pennsylvania State University. Results a re reported by Ebert (2017). General patterns in obsidian technology show that the Cahal Pech assemblage was composed primarily of finished prismatic blades. Medial segments of blades are the most common artifact from Preclassic and later Classic Period contexts, with blades becoming more common beginning in the Middle Preclassic. Obsidian blade cores and manufacturing debris are not common in the Preclassic assemblage, or the assemblage for any time period, indicating that finished blades were likely im ported to Cahal Pech in a pattern consistent with wholeblade or processed blade trade (De Len et al. 2009, Stemp et al., this volume). Geochemical characterization of obsidian artifacts was conducted at The Pennsylvania State University Ceramics Laborato ry according to standard procedures using a Bruker Tracer III V+ SD handheld XRF spectrometer with X rays emitted from a rhodium tube (see Ebert et al. 2015). Cluster analysis of pXRF data identified five obsidian source groups in the Cahal Pech assemblag e ( Figure 4). The majority of artifacts were imported from the El Chayal source (61.5%, n =732), with smaller amounts from the Ixtepeque (19.3%; n =230) and SMJ (18.5 %, n =220) sources. Ucareo (>0.01%, n =1) and Pachuca (>0.01%, n =6) blade fragments from the central Mexican Highlands are also present. The Early Preclassic obsidian assemblage was derived exclusively from excavated contexts at Str. B4, located in Plaza B of the site core ( Table 1 ). Obsidian nodules and percussion flakes compose more than half of the Early Preclassic assemblage (~68%, n =15), with only one pressure blade artifact (3rd series corner blade) present in early contexts. All the Early Preclassic artifacts were assigned to t he El Chayal source. Additional types of obsidian appear during the Middle Prelcassic, as the percussion flake tradition was replaced by prismatic pressure blades technology at Cahal Pech (Awe and Healy 1994; Ebert 2017; Stemp et al., this volume). While the inhabitants of the site core primarily consumed blades from the SMJ source, blades from El Chayal are the dominant type found in peripheral household Figure 4 Bivariate log10 transformed elemental concentrations for obsidian sample from Cahal Pe ch. Ellipses represent 90% confidence intervals for group membership based source samples with known proveniences (data courtesy of Archaeometry Lab at MURR). groups. Imported blades continued to compose the majority of the obsidian assemblage throughout the Preclassic. One obsidian blade from the Ucareo source, recovered from late Middle Preclassic levels at Str. B4, documents possible connections with the central Me xican Highlands. El Chayal blades continued to dominate the obsidian assemblage both in the site core and settlement (63%) throughout the Classic Period, with smaller but relatively even amounts of SMJ obsidian through the Terminal Classic. Ixtepeque obsidian composes ~24% of the Late Classic assemblage, and is found in higher proportions in the settlement ( n =134) compared to the site core ( n =60) during this period. Blades from the Pachuca source also enter the assemblage during the Early Classic, though t he total number in the Classic Period sample analyzed for this study is small ( n =6). By the Terminal Classic, El Chayal blade fragments make up over 83% of the assemblage. Ceramic INAA Analyses Ceramic samples subjected to INAA were chosen from common types of diagnostic ceramics (total n =192) from radiocarbon dated contexts in the sites civicceremonial core and from two peripheral settlement groups (Ebert 2017). All sherds were identified to type: variety mode

PAGE 70

Reconstructing Preclassic Maya Household Economies 70 Table 1 Comparison of obsidian sources for site core, settlement, and surface/unknown contexts analyzed for Cahal Pech by chronological period. The San Martn Jilotepeque source is abbreviated as SMJ. Time periods abbreviated as follows: EPC = Early Preclassic, MPC = Middle Preclassic, LPC = Late Preclassic, EC = Early Classic, LC = Late Classic, TC = Terminal Classic, UNK = Unknown Period. Source EPC MPC LPC EC LC TC UNK Total Source El Chayal 22 44 19 13 517 106 11 732 Site Core 22 20 18 7 210 102 9 388 Settlement 24 1 6 307 4 2 344 SMJ 82 18 8 100 10 2 220 Site Core 65 18 6 34 10 1 134 Settlement 17 2 66 1 86 Ixtepeque 10 6 7 194 11 2 230 Site Core 2 6 5 60 11 84 Settlement 8 2 134 2 146 Pachuca 1 5 6 Site Core 1 3 4 Settlement 2 2 Ucareo 1 1 Site Core 1 1 Settlement Total Period 22 137 43 29 816 127 15 1189 classification according to standard classifications for Cahal Pech and the Belize Valley (Awe 1992; Gifford 1976; Sullivan and Awe 2013). A total of 125 sherds from contexts radiocarbon dated to the Cunil and Kanluk phases (Awe 1992; Ebert et al. 2017; Peniche May 2016) were sampled from excavations in Str. B4 and Plaza B in the Cahal Pech site core. Samples were also chosen from Middle and Late Preclassic contexts at two house groups in the Cahal Pech periphery: the Tzutziiy Kin ( n =40) and Zopilote ( n =27) groups. Samples from the Zopilote Group come from late facet K anluk (750350 BC) and early/late facet Xakal phase (350 BC AD 300) contexts at Structure 1 (Ebert and Fox 2016). Samples from Tzutziiy Kin are derived from excavations of domestic buildings at Structure 2 and 3, and date to the early/late facets of the Late Preclassic Xakal ceramic phase (Ebert et al. 2016, 2017). All ceramic samples were prepared for INAA using standard procedures at MURR by Daniel Peirce and Michael Glascock (see Glascock 1992; Neff 2000). Initial identification of compositional groups was based on mean and standard deviations for concentration data for each element within the sample. Hierarchical cluster analysis and principal component analysis were then applied to elemental data to refine Figure 5 Bivariate plot of INAA samples displayed based on canonical discriminant functions #1 and #2. Ellipses represent 90% confidence of membership for identified groups in the assemblage. compositional group membership. INAA results for the Cahal Pech sample were also compared to the results of over 12,000 previous analyses by MURR using Euclidian Distance searches to identify similarities with other identified geochemical compositional groups in Mesoamerica. The Cahal Pech ceramics divide into seven groups that generally correspo nd with type: variety classifications from different time periods and contexts ( Figure 5 and Table 2 ).

PAGE 71

Ebert and Awe 71 Table 2 Ceramic compositional groups at Cahal Pech identified by INAA for each chronological period and ceramic phase, listed by conte xt. Early facet (EF) and Late Facet (LF) components of ceramic phases are listed when present. Time periods abbreviated as follows: EPC = Early Preclassic, MPC = Middle Preclassic, LPC = Late Preclassic. Compositional Group Context EPC MPC LPC % Total Cunil EF Kanluk LF Kanluk EF Xakal LF Xakal Group A 1% n=2 Site Core 2 Settlement Group B 18% n=34 Site Core 12 18 1 Settlement 3 Group C 7% n=13 Site Core 3 5 Settlement 2 3 Group D 37% n=71 Site Core 21 17 24 Settlement 1 7 1 Group E 1% n=2 Site Core 1 Settlement 1 Group F 2% n=3 Site Core 3 Settlement 2 Group G 23% n=45 Site Core 1 1 1 Settlement 1 10 24 7 Unassigned 11% n=22 Site Core 2 4 9 Settlement 1 2 3 1 Total n 37 50 58 37 10 100% Group A consists of two Cunil sherds of an unspecified white slipped type, the only two samples from excavations at Plaza B with ash temper. Group B ( n =34) contains all other sherds analyzed for this study with ash temper, as well as vessels characterized by fine texture calcite/quartzite pastes. Many of these samples are decorated with dull slips and post slip incising (e.g., Baki Red Incised, Mo Mottl ed, and Kitam Incised types). Euclidean Distance searches indicate that the Cahal Pech specimens are compositionally unique to previously analyzed samples in the MURR database from the Maya region. Groups C and D contain ceramic samples attributed primarily to the late facet Kanluk ceramic phase (750350 BC). Group C ( n =13) ceramics are primarily Mars Orange wares (Savana Orange and Reforma Incised types; Gifford 1976:7376) and were distributed between late Middle Preclassic site core (62%) and settlemen t contexts (38%). Group D is the largest compositional group ( n =71) in the INAA sample. Most specimens come from site core contexts (87%) with Cunil and Kanluk ceramic phase temporal assignments. The group is composed primarily of unslipped utilitarian pottery (57%, e.g., Sikiya and Jocote types), but also contains high frequencies of Savana Orange wares (37%). Euclidean Distance searches indicate that many of the specimens in this group are compositionally similar to previously analyzed samples of Midd le Preclassic Mars Orange ceramics from the site of Holtun, Guatemala (Callaghan et al. 2017). Group E ( n =2) and Group F ( n =3) ceramics compose only 3% of the total Cahal Pech INAA sample. While both groups are compositionally distinct, they exhibit high degrees of internal compositional variability, which indicates slightly different paste recipes for each sherd. Groups E and F are found relatively evenly between site core and settlement contexts, and are composed primarily of Joventud Red sherds from th e Kanluk

PAGE 72

Reconstructing Preclassic Maya Household Economies 72 Figure 6 Comparison of obsidian sources between Cahal Pech site core and settlement contexts for the Preclassic period. ceramic complex. Group G ( n =45) is the second most common group in the Cahal Pech INAA sample, and is composed of sherds from the Late Preclassic Xakal ceramic complex (350 BC AD 300), suggesting a preferences for this paste recipe within household groups during later time periods. Specimens in this group are found almost exclusively at pe ripheral household groups, with ~74% of sherds samples from the Tzutziiy K'in Group and ~65% of the sherds samples from the Zopilote Group assigned to this group. The most common ceramic types include Sierra Red and Joventud Red, with small numbers of uns lipped utilitarian wares (Jocote Orange brown and Sayab Daub Striated types). Discussion We used geochemical compositional methods of obsidian and ceramics to identify the economic mechanisms associated with developing inequality among Preclassic households at the site of Cahal Pech. The results indicate that, while obsidian and ceramic economies overlapped to supply households with items needed for everyday subsistence, they were structured in different ways. The pXRF data reported in this stud y document a relatively decentralized network of domestic obsidian consumption at Cahal Pech throughout the Preclassic. The inhabitants of Cahal Pech were active participants in longdistance obsidian exchange systems with the southern highlands of Guatem ala as early as 1200 cal BC (Awe and Healy 1994; Ebert 2017; Peniche May 2016). The results of pXRF analyses of obsidian indicate the presence of El Chayal percussion flakes and one pressure blade artifact within the earliest Cunil domestic contexts at th e site core. The sample size for this period is small ( n =22), however, and is derived from only one context (Str. B4) in the Cahal Pech site core, making it difficult to assess hypotheses about differential obsidian consumption between households The lon g distance procurement networks accessed by Cahal Pech expanded during the Middle Preclassic. SMJ obsidian became the most abundant source at the site, with El Chayal and Ixtepeque artifacts found less frequently. Different types of obsidian, however, we re not evenly distributed between households. While SMJ blades became prevalent in the site core (~74% of all site core artifacts for the period), El Chayal remained the primary source for blades consumed by peripheral households (~49% of all settlement a rtifact for the period; Figure 6 ). The differential procurement of obsidian types suggests a lack of centralized control over redistribution. Instead, it is more likely that obsidian moved through a network of decentralized exchange relationships operati ng at the householdlevel. By the Late Preclassic, both El Chayal and SMJ blades became more evenly distributed within the Cahal Pech site core, with Ixtepeque also composing a smaller portion of the assemblage. A small sample size for peripheral house g roups (one El Chayal blade) limits our interpretation of obsidian consumption in the settlement versus the site core for this period. Based on patterns in the geochemical data for later periods, however, it appears that decentralized

PAGE 73

Ebert and Awe 73 Ta ble 3 Comparison of Preclassic obsidian sources from Cahal Pech and other Belize Valley sites. Site El Chayal SMJ Ixtepeque Other Total n Citation Blackman Eddy Kersey 2006 Early Preclassic Middle Preclassic 1 35 1 37 Late Preclassic 3 3 Cahal Pech Ebert 2017 Early Preclassic 22 22 Middle Preclassic 44 82 10 1 137 Late Preclassic 19 18 6 43 Chan Meierhoff et al. 2012 Early Preclassic Middle Preclassic 6 6 Late Preclassic 8 22 11 41 Figure 7. Proportions of obsidian sources at Blackman Eddy, Cahal Pech, and Chan for Early through Late Preclassic periods.

PAGE 74

Reconstructing Preclassic Maya Household Economies 74 domestic procurement of blades persisted through the end of the Terminal Classic at Cahal Pech. Comparisons to obsidian provenance studies for other Preclassic sites in the Belize Valley document a similar pattern of local and regional decentralized procurement of obsidian ( Table 3 ). While published data for geochemically sourced assemblages is relatively limited, we consider obsidian source data from Blackman Eddy (Kersey 2006) and Chan (Meierfhoff et al. 2012) to interpret differences in obsidian consumption between these sites and Cahal Pech during the Preclassic ( Figure 7 ). Cahal Pech possesses the only evidence for obsidian during the Early Preclassic Period (Stemp et al., this volume). Whereas Cahal Pech relied solely on El Chayal obsidian ( n =22) at this early date, provenance data for Middle and Late Preclassic assemblages show the development of differential procurement networks at the site level at other sites across the Belize Valley. Compared to the almost complete reliance on SMJ obsidian by Blackman Eddy and Chan, Cahal Pech consumption patterns indicate the use of higher proportions of El Chayal obsidian imported as blades. By the Late Preclassic, the Cahal Pech and Chan assemblages are relatively evenly spread between obsidian types, while Blackman Eddy became reliant on El Chayal obsidian. Despite sma ll samples sizes for some periods, theses comparisons indicate that Preclassic sites developed independent procurement strategies to provision themselves with nonlocal obsidian, and that dominant sources shifted through time. Ceramic INAA data from Cahal Pech suggest the development of craft specialization and distribution beyond the householdlevel that may have contributed to status and wealth of some households. INAA identified three compositional groups (A, B, and D) that contained diagnostic Cunil ce ramic types, indicating a preference for these paste recipes during the Early Prelcassic. Both Groups A and B contained high proportions of fine paste slipped and groovedincised Cunil vessels derived exclusively from the Cahal Pech site core. Specimens in these two groups were also found to be compositionally unique compared to previously analyzed ceramics in the MURR database, suggesting that Cunil ceramics were produced and distributed locally in the Belize Valley. While vessels attributed to Groups A and B were primarily decorated types, the Cunil complex sherds in Group D are utilitarian, including unslipped jars and bowls used for daily tasks including water storage and cooking (Sullivan and Awe 2013). The differential distribution of Cunil utilita rian versus decorated serving wares between compositional groups may suggest individual (household) specialized production. The Middle Prelcassic Kanluk complex ceramic assemblage from Cahal Pech was composed primarily of Jocote Orange brown utilitarian ce ramics and fine Mars Orange Paste serving wares including undecorated and decorated types (e.g., Reforma Incised; Awe 1992; Peniche May 2016). A correlation between compositional groups for Middle Preclassic Jocote vessels and earlier Cunil utilitarian wa res suggests persistence in local production of these types for domestic consumption. Typological studies from sites in the Belize Valley have also documented high frequencies of Mars Orange ceramics (~60 50%) in Middle Preclassic ceramic assemblages, pos sibly suggesting local production within the Belize Valley region (Awe 1992; Gifford 1976; Kosakowsky 2012; Peniche May 2016). Over 77% of the Savana Orange sherds analyzed in this study were assigned to compositional Groups C ( n =27) and D (n =35). These sherds were derived primarily from site core contexts associated with high status residences and public architecture (Horn 2015; Peniche May 2016). Euclidean Distance searches for the Cahal Pech Mars Orange ceramics within the MURR database identified com positionally similar ceramics from the site of Holtun, located in the central Petn of Guatemala ( Figure 8 ) Non local Mars Orange sherds from Holtun, also associated with monumental architecture in that sites civic ceremonial epicenter, formed a distinc t compositional group (Group 1; Callaghan et al. 2017). Though Holtun and Cahal Pech assemblages possess similar paste recipes, higher frequencies of Mars Orange paste wares in the Cahal Pech assemblage (77%) versus Holtun region (~12%; Callaghan and Neiv ens de Estrada 2016), suggest the Belize Valley as the likely origin of the Holtun Mars Orange assemblage. The Late Preclassic (early/late facet Xakal ceramic phase) at Cahal Pech and sites across the lowlands saw the introduction of distinctive Chicanel s tyle ceramics, characterized by matte or waxy finish red and black slips (Awe 1992; Gifford 1976). The development of this regional ceramic

PAGE 75

Ebert and Awe 75 Figure 8 Bivariate plot of Cahal Pech ceramic compositional groups compared Middle Preclassic Group 1 ceramics at Hotlun, Guatemala (after Callaghan et al. 2017) based upon canonical discriminant functions #1 and #2. Ellipses represent 90% confidence of membership for identified groups in the assemblage. style and more tightly integrated obsidia n exchange networks corresponds to the rapid growth of major civic ceremonial centers (Ebert et al. 2017). At Cahal Pech, a program of large scale monumental construction occurred in the site epicenter (Plazas A and B; Awe 1992; Healy et al. 2004). Sever al peripheral settlements also witnessed the construction of larger scale residential buildings after ~350 cal BC (see Awe 1992:207; Ebert et al. 2016, 2017). The Xakal complex ceramics sampled for INAA in this study derive from contexts at the peripheral Tzutziiy Kin and Zopilote settlement groups (Ebert 2017). The majority (~96%) of these ceramics are restricted to compositional Group G, which includes common Xakal types (Sierra Red, Joventud Red, Sayab Daub striated) with both utilitarian (e.g., large jars, bowls, spindle whorls) as well as more specialized forms (e.g., serving dishes, spouted vessels). While most of the later samples were derived from household contexts, the correlation between time period and context may have important implications for understanding diachronic patterns of ceramic production and consumption at Cahal Pech, and more broadly within the lowland region. Because our sample from Cahal Pech is derived primarily from peripheral households, Group G ceramics may represent diffe rential production between the households and site core. The shift in paste recipe at Cahal Pech may also correspond to the adoption of Chicanel style ceramics as a result of the development of regional interaction networks. Additional INAA analyses of L ate Preclassic ceramics from the Cahal Pech site core and from other Maya sites are necessary to characterize differential production and consumption patterns that may be associated with local tradition and status. Conclusions Domestic economies were essential links in local communities to larger regional socio economic systems among early Maya societies, and household production and exchange likely shaped the function of broader institutional economies (Hirth 2012, 2016). The results of this study indica te that economic networks became increasingly complex and interconnected throughout the Preclassic, with the function of production and exchange varying by the type of goods consumed through time. Both obsidian pXRF and ceramic INAA data indicate that households were self sufficient and procured or produced most of the items necessary for daily activities. Obsidian source data connect Preclassic households at Cahal Pech to a diversity of economic networks operating between the Belize Valley, highland Guat emala, and highland Mexico. The differential procurement of blades produced from different obsidian types suggests a lack of centralized control over redistribution (e.g., Clark 1987; Santley 1984). Our data show instead that obsidian moved through a net work of decentralized exchange relationships operating at the household level. These results indicate that the exchange of finished blades likely did not contribute to unequal economic relationships between households at Cahal Pech. INAA data show that ceramics were differentially consumed through time. Local production of specialized ceramic serving vessels with ideologically significant designs first appear at Cahal Pech during the Early Preclassic Cunil phase, and were produced a nd consumed locally. The patterning of INAA data also provides evidence for the development of inter regional exchange of specialized Mars Orange pottery between high status groups at Cahal Pech and sites in the central Petn. Production and distribution of these specialized vessels may have been used as one strategy by emergent high status households at Cahal Pech to link people in other regions of the lowlands into networks of interdependency within a

PAGE 76

Reconstructing Preclassic Maya Household Economies 76 developing institutional economy organized above the level of the household. Future research focused on characterizing obsidian and ceramic assemblages from other Preclassic contexts at Cahal Pech, other Belize Valley sites, and sites across the Maya region will help us reconstruct variation in assemblages may reveal the economic strategies that shaped both local and regional economies and contributed to institutionalized social and economic differentiation. Acknowledgements Research at Cahal Pech, Belize was conducted under the auspices of the Belize Val ley Archaeological Reconnaissance (BVAR) Project directed by Drs. Jaime Awe and Julie Hoggarth. We extend our gratitude to Dr. John Morris and the Belize Institute of Archaeology for their continued support of BVAR Project field research, and permission t o analyze obsidian and ceramic samples from Cahal Pech. Access to the pXRF used for obsidian sourcing analyses was provided by the Penn State University Department of Anthropology Ceramics Analysis Laboratory, directed by Dr. Sarah McClure. INAA analyses were conducted by Drs. Daniel Peirce and Michael Glascock at the University of Missouri Research Reactor (MURR) with financial support from the NSF Archaeometry Program Grant to MURR (BCS 1415403). Additional financial support for this research was provi ded by the Penn State Delbert and Marie Welch Dissertation Research Award, an NSF Dissertation Improvement Grant (BCS 1460369, C. Ebert and D. Kennett), and the Tilden Family Foundation, San Francisco, California. Open access to obsidian and ceramic compos itional data analyses is available at ScholarSphere (https://scholarsphere.psu.edu/). References Ashmore, Wendy and Richard R. Wilk 1988 Household and Community in the Mesoamerican Past. In Household Community in the Mesoamerican Past edited by R.R. Wil k and W. Ashmore, pp. 1 28. Albuquerque: University of New Mexico Press. Awe, Jaime J. 1992 Dawn in the Land between the Rivers: Formative Occupation at Cahal Pech, Belize, and its Implications for Preclassic Development in the Central Maya Lowlands Unpublished PhD Dissertation, University of London, London. Awe, Jaime J. n.d. Archaeological Evidence for the Preclassic Origins of the Maya Creation Story and the Resurrection of the Maize God at Cahal Pech, Belize. In, Moyes, Holley and Allen Christen son (Eds), The Myths of the Popol Vuh in Cosmology and Practice University Press of Colorado, Boulder. Awe, Jaime J. and Paul F. Healy 1994 Flakes to Blades? Middle Formative Development of Obsidian Artifacts in the Upper Belize River Valley. Latin American Antiquity 5:193205. Brown, M. Kathryn, Leah McCurdy, Whitney Lytle, and Thomas Chapman 2013 Mopan Valley Preclassic Project: Results of the 2011 Field Season. Research Reports in Belizean Archaeology 10: 137 146. Callaghan, Michael G. an d Nina Nievens de Estrada 2016 The Ceramic Sequence of the Holmul Region, Guatemala Tucson: University of Arizona Press. Callaghan, Michael G., Daniuel E. Pierce, Bridget Kovacevich, and Michael D. Glascock 2017 Chemical Paste Characterization of Late Mi ddle Preclassic period Ceramics from Holtun, Guatemala and its Implications for Production and Exchange. Journal of Archaeological Science: Reports 12: 334 345. Chase, Arlen F. and Diane Z. Chase 2012 Complex Societies in the Southern Maya Lowlands: Their Development and Florescence in the Archaeological Record. In Oxford Dictionary of Mesoamerica edited by Deborah L. Nichols and Christopher A. Pool, pp. 255267. Oxford: Oxford University Press. Clark, John 1987 Politics, Prismatic Blades, and Mesoameric an Civilization. In The Organization of Core Technology edited by Jay K. Johnson and Carol A. Morrow, C.A., pp. 259 284. Boulder: Westview Press. Clark, John and Michael Blake 1994 The Power of Prestige: Competitive Generosity and the Emergence of Rank Societies in Lowland Mesoamerica. In Factional Competition and Political Development in the New World edited by Elizabeth Brumfiel and John Fox, pp. 1730. Cambridge: Cambridge Univer sity Press. Costin, Cathy Lynn 1991 Craft Specialization: Issues in Defining, Documenting, and Explaining the Organization of Production. In Archaeological Method and Theory edited by Michael B. Schiffer, pp. 1 56. Tucson: University of Arizona Press.

PAGE 77

Ebert and Awe 77 De Len, Jason P., Kenneth G. Hirth, and David M. Carballo 2009 Exploring Formative Period Obsidian Blade Trade: Three Distribution Models. Ancient Mesoamerica 20: 113 128. Douglass, John G., and Nancy Gonlin 2012 Ancient Households of the Americas: Conc eptualizing What Households Do. Boulder: University Press of Colorado. Ebert, Claire E. 2017 Preclassic Maya Social Complexity and Origins of Inequality at Cahal Pech, Belize Unpublished PhD Dissertation. The Pennsylvania State University, University Par k, PA. Ebert, Claire E., Brendan J. Culleton, Jaime J. Awe, and Douglas J. Kennett 2016 AMS 14C Dating of Preclassic to Classic Period Household Construction in the Ancient Maya Community of Cahal Pech, Belize. Radiocarbon 58: 6987. Ebert, Claire E., Ma rk Dennison, Kenneth G. Hirth, Sarah B. McClure, and Douglas J. Kennett 2015 Formative Period obsidian exchange along the Pacific Coast of Mesoamerica. Archaeometry 57(S1): 5473. Ebert, Claire E., Nancy Peniche May, Brendan J. Culleton, Jaime J. Awe, and Douglas J. Kennett 2017 Regional Response to Drought during the Formation and Decline of Preclassic Maya Societies. Quaternary Science Reviews 173:211235. Garber, James F. and Jaime J. Awe 2009 A Terminal Early Formative Symbol in the Maya Lowlands: The Iconography of the Cunil Phase (1100900 BC) at Cahal Pech. Research Reports in Belizean Archaeology 6: 151160. Garber, James F., M. Kathryn Brown, Jaime J. Awe, and Chirstopher J. Hartman 2004 Middle Formative Prehistory of the Central Belize Valley: An Examination of Architecture, Material C ulture, and Sociopolitical Change at Blackman Eddy. In The Ancient Maya of the Belize Valley: Half a Century of Archaeological Research edited by James F. Garber, pp. 25 47. Gainesville: University Press of Florida. Gifford, James C. 1976 Prehistoric Pottery Analysis and the Ceramics of Barton Ramie in the Belize Valley Cambridge: Harvard University. Glascock, Michael D. 1992 Characterization of Archaeolo gical Ceramics at MURR by Neutron Activation Analysis and MultivariateStatistics. In Chemical Characterization of Ceramic Pastes in Archaeology edited by Hector Neff, pp. 1126. Madison: Prehistory Press. Healy, Paul F., David Cheetham, Terry G. Powis, an d Jaime J. Awe 2004 Cahal Pech: The Middle Formative Period. In The Ancient Maya of the Belize Valley: Half a Century of Archaeological Research edited by James F. Garber, pp. 103 124. Gainesville: University Press of Florida. Hirth, Kennth G. 2009 Craft Production, Household Diversification, and Domestic Economy in Prehispanic Mesoamerica. In Housework: Craft Production and Domestic Economy in Ancient Mesoamerica edited by Kenneth G. Hirth, pp. 1332. Archaeological Papers of the American Anthropologica l Association no. 19. New York: Blackwell Publishing. 2012 Markets, Merchants and Systems of Exchange. In Oxford Dictionary of Mesoamerica edited by Deborah L. Nichols and Christopher A. Pool, pp. 639652. Oxford: Oxford University Press. 2016 The Azte c Economic World: Merchants and Markets in Ancient Mesoamerica Cambridge: Cambridge University Press. Hohmann, Bobbi 2002 Formative Maya Shell Ornament Production in the Belize Valley, Belize Unpublished PhD Dissertation. University of New Mexico, Albuquerque. Horn, Sherman W., III 2015 The Web of Complexity: Socioeconomic Networks in the Middle Preclassic Belize Valley Unpublished PhD Dissertation. Tulane University, New Orleans. Kersey, Kelsey M. 2006 Emerging Elite Economies: Formative Period Obsidian Distribution in the Belize River Valley. In The Belize Valley Archaeology Project: Results of the 2005 Field Season, edited by James F. Garber, pp. 3162. San Marcos: Texas State University. Kosakowsky, Laura J. 2012 Ceramics and Chronology at Chan. In Chan: An Ancient Maya Farming Community edited by Cynthia Robin, pp. 4270. Gainesville: University Press of Florida. Lee, David F. and Jaime J. Awe. 1995 Preclassic Architecture, Burials, and Craft Speci alization at the Cas Pek Group, Cahal Pech. In Belize Valley Preclassic Maya Project: Report on the 1994 Field Season, edited by Paul F. Healy and Jaime J. Awe, pp. 95 115. Trent University, Department of Anthropology Occasional Papers in Anthropology No.1 0. Peterborough: Trent University. Meierhoff, J., Mark Golitko, and James D. Morris 2012 Obsidian Acquisition, Trade, and Regional Interaction at Chan. In Chan: An Ancient Maya Farming Community edited by Cynthia Robin, pp. 271 288. Gainesville: Universi ty Press of Florida.

PAGE 78

Reconstructing Preclassic Maya Household Economies 78 Neff, Hector 2000 Neutron Activation Analysis for Provenance Determination in Archaeology. In Modern Analytical Methods in Art and Archaeology edited by E. Cilberto and G. Spoto, pp. 81 134. New York: John Wiley and Sons, Inc. 2014 P ots as Signals: Explaining the Enigma of Long distance Ceramic Exchange. In Craft and Science: International Perspectives on Archaeological Ceramics edited by Marcos Martinn Torres, pp. 1 12. Doha: Bloomsbury Qatar Foundation. Peniche May, Nancy 2016 Building Power: Political Dynamics in Cahal Pech, Belize during the Middle Preclassic, Unpublished PhD Dissertation. University of California, San Diego, San Diego. Powis, Terry G., Sherman Horn, Gyles Iannone, Paul F. Healy, James F. Garber, Jaime J. Awe Sheldon Skaggs, and Linda A. Howie 2016 Middle Preclassic Period Maya Greenstone Triangulates: Forms, Contexts, and Geology of a Unique Mesoamerican Groundstone Artifact Type. Journal of Archaeological Science Reports 10: 59 73. Santley, Robert S. 198 4 Obsidian Exchange, Economic Stratification, and the Evolution of Complex Society in the Basin of Mexico. In T rade and Exchange in Early Mesoamerica edited by Kenneth G. Hirth, pp. 4386. Albuquerque: University of New Mexico Press. Stemp, W. James, Jaime J, Awe, M. Kathryn Brown and James F. Garber. 2017 Rock Bottom: Maya Lithic Technology in the Early Terminal to Late Middle Preclassic Periods at Cahal Pech and Blackman Eddy, Cayo District, Belize. Research Reports in Belizean Archaeology, vol. 15. Sullivan, Lauren A. and Jaime J. Awe 2013 Establishing the Cunil Ceramic Complex at Cahal Pech, Belize. In Ancient Maya Pottery: Classification, Analysis, and Interpretation edited by James J. Aimers, pp. 107120. Gainesville: University Press of Florid a. Tibbits, Tawny L. B. 2016 Geochemical Sourcing of Granite Ground Stone Tools from Belize Unpublished PhD Dissertation. The University of Iowa, Iowa City. Wilk, Richard R. and William L. Rathje 1982 Household Archaeology. American Behavioral Scientis t 25: 617 639. Willey, Gordon R., William R. Bullard, John B. Glass, and James C. Gifford 1965 Prehistoric Maya Settlements in the Belize Valley Cambridge: Cambridge University Press.

PAGE 79

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 7991 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 7 ROCK BOTTOM: MAYA LI THIC TECHNOLOGY IN THE EARLY TERMINAL TO LATE MIDDLE PRECLASSIC PERIO DS AT CAHAL PECH AND BLACKMAN ED DY, CAYO DISTRICT, B ELIZE W. James Stemp, Jaime J. Awe, M. Kathryn Brown, and James F. Garber Excavations of the earliest Maya occupations at Cahal Pech and Blackman Eddy provide information to piece together the lithic technology of the first Maya in western Belize. The Cunil/Kanocha to late facet Kanluk/Jenney Creek stone tool assemblages from t hese two sites document different lithic raw material acquisition patterns and changes in tool production over time. In the Cunil/Kanocha phase, the Maya at both sites relied on hard-hammer and bipolar reduction to produce expedient flakes from locally av ailable chert; similar expedient flake production of Guatemalan obsidian is documented at Cahal Pech. In the early facet Kanluk/Jenney Creek phase, expedient chert tool production continues, but obsidian prismatic blades begin to appear as trade items and evidence of chert bifaces and unifaces made from local chert is documented. In the late facet Kanluk/Jenney Creek phase, new tool types such as burin spall drills of local chert and stemmed macroblades manufactured from northern Belize chert are recovered for the first time. These general trends in stone tool technology are compared to the lithic assemblages from other contemporaneous periods at sites in Belize, Guatemala, and Honduras in an attempt to understand changes in early Maya lifestyles in the t erminal Early to late Middle Preclassic. Introduction Recent discoveries of more preceramic (Archaic) points throughout western Belize and excavations at some of the earliest dated Maya occupations, such as those at Cahal Pech and Blackman Eddy ( Figure 1 ), are providing necessary information to piece together early lithic production techniques and patterns of raw material use. Not surprisingly, we see that Maya sites settled in the terminal Early to early Middle Preclassic were not large scale stone tool production centers, but seem to be associated with the lifestyle changes of relatively small communities of early Maya farmers. Reconstructions of stone tool sequences permit interpretations concerning the development of lithic technology from preceramic peoples to early Maya populations and shed some light o n how this technology factors into larger behavioral adaptations associated with major cultural changes in this significant period of transition in western Belize and beyond. The Preceramic Period in Western Belize (ca. 11,500 1200 B.C.) Although no fluted Palaeoindian chipped stone bifaces have been recovered from western Belize, debitage from Actun Halal may date to the Pleistocene (Lohse and Collins 2004). Yet, many stemmed and barbed preceramic chert points have been recovered in the Cayo District, mostly as surface finds (Kelly 1993; Iceland Figure 1 Locations of obsidian sources and archaeological sites mentioned in the text. 1997; Lohse et al. 2006; Stemp and Awe 2013). In all, 11 Lowe points and six Sawmill points provide the best evidenc e for hunter gatherers in the area in what is traditionally defined as the Archaic period (Stemp and Awe 2013; Stemp et

PAGE 80

Maya Lithic Technology at Cahal Pech and Blackman Eddy 80 al. 2016; Stemp et al. 2018). Some of these tools from western Belize have been identified as Northern Belize Zone (NBZ) chert, indicat ing contact with this region. Other chert bifaces were made from high quality stone that, in some cases, can be locally sourced (Horowitz 2017). To date, no obsidian has been dated to the preceramic period of western Belize; however, it has been found in southern Belize (K. Prufer, pers. comm. 2016). In terms of function, current evidence suggests that Lowe points were hafted tools used as thrusting spears/harpoons and knives; whereas, Sawmill points appear to have been affixed to atlatl darts (Stemp et al. 2016; see Kelly 1993). Another diagnostic tool form identified in the Late Archaic of western Belize is the constricted adze (Stemp and Awe 2013; see Iceland 1997; Lohse et al. 2006). The constricted adzes were hafted tools used for chopping/adzing w ood, most likely associated with forest clearance and horticulture (Gibson 1991). The Terminal Early Early Middle Preclassic Periods at Cahal Pech and Blackman Eddy Cunil/Kanocha Phase (1200 900 B.C.) In the periods associated with the earliest potter y in western Belize, all of the diagnostic forms and most of the production technologies of the preceding preceramic period are absent. Cahal Pech In the earliest levels that correspond to the Cunil phase, tool makers relied overwhelmingly on chert from limestone deposits around the site and on chert nodules retrieved near the Macal River, as well as small amounts of locally procured quartz/quartzite (Stemp 2012; Horn 2015: 337, Table 7.9; Peniche May 2016: 266). At Structure B4, the South Trench i n Group B, and in the Group B Plaza, the majority of tool types are cortical and non cortical flakes, some with simple retouch, and flake cores and core fragments (Stemp 2012; Horn 2015; Peniche May 2016) (Figure 2, Tables 1 and 2). Hard hammer percussion was the dominant reduction technique used to strike flakes from relatively small nodules; however, some bipolar percussion occurred as well (Stemp 2012). This was an expedient technology to produce flakes that could be minimally modified through Figu re 2 Hard -hammer flakes and blocky fragments of local chert from the Cunil phase at Structure B4, Cahal Pech. Figure 3 Hard -hammer flakes and blocky fragments of El Chayal obsidian from the Cunil phase at Structure B4, Cahal Pech. unifacial edge r etouch. In terms of other flake tools, Peniche May (2016: 271, Table 6.2) identified flakeblades or for tuitous blades in Cunil levels. Despite the suggestion by Peniche May (2016: 277) that biface technology was present in Cunil times at Cahal Pech the very crudely chipped triangular, general utility biface (SF 769) from Plaza B was recovered from a deposit of black dirt that contained no sherds (Horn 2015: 317), which makes a Cunil designation questionable. There is no reliable evidence for b ifaces at this site (cf. Peniche May 2016: 269, Fig. 6.3) in Cunil levels In Group B, obsidian tool types and production methods are the same as those discussed for chert. The Maya relied on the same hard hammer percussion technique to

PAGE 81

Stemp et al. 81 Table 1 Percentages of tool types from the Cunil to late facet Kanluk phases from Unit 7, Structure B4, Cahal Pech. STRUCTURE B4, UNIT 7 Late facet Kanluk (N = 71) Early facet Kanluk (N = 29) Cunil (N = 57) oval bifaces 1.4% 0 0 drills (burin spalls) 1.4% 0 0 drills (on flakes) 11.4% 0 0 scrapers 0 3.4% 0 flake cores 0 0 0 flake core frags. 5.6% 0 0 tert. ret. flakes (0% cortex) 0 0 1.7% pr. Flakes (100% cortex) 0 0 3.4% sec. flakes (>50% cortex) 5.6% 13.7% 12.1% sec. flakes (<50% cortex) 30.6% 31.0% 36.2% tert. flakes 34.7% 44.8% 36.2% sec. bif. th. flakes (>50% cortex) 0 0 0 sec. bif. th. flakes (<50% cortex) 2.8% 0 0 tert. bif. th. flakes (0% cortex) 5.6% 6.9% 0 sec. macroflakes (<50% cortex) 0 0 1.7% bl. frags. 11.1% 0 6.9% burnt frags. 0 0 1.7% Table 2 Percentages of tool types from the Cunil and late facet Kanluk phases from the South Trench, Group B, Cahal Pech. SOUTH TRENCH, GROUP B Late facet Kanluk (N = 63) Cunil (N = 987) oval bifaces 0 0 drills (burin spalls) 90.4% 0 drills (on flakes) 1.6% 0 scrapers 0 0 stemmed macroblades 3.2% 0 flake cores 0 0.4% flake core frags. 0 1.1% bifacial flake cores 0 0.1% tert. ret. flakes (0% cortex) 0 0 pr. flakes (100% cortex) 0 2.1% sec. flakes (>50% cortex) 0 12.2% sec. flakes (<50% cortex) 0 36.0% tert. flakes (0% cortex) 4.8% 43.4% sec. bif. th. flakes (>50% cortex) 0 0 sec. bif. th. flakes (<50% cortex) 0 0 tert. bif. th. flakes (0% cortex) 0 0 sec. macroflakes (<50% cortex) 0 0 sec flakes (<50% cortex) as recycl. core 0 0.1% bl. frags. 0 4.3% burnt frags. 0 0.2%

PAGE 82

Maya Lithic Technology at Cahal Pech and Blackman Eddy 82 Table 3 Percentages of tool types from the Kanocha to late facet Jenney Creek phases from lithic sub -assemblage from Structure B1, Blackman Eddy (from Yacubic 2006). STRUCTURE B1 Late facet Jenney Creek (N = 469) Early facet Jenney Creek (N = 1869) Kanocha (N = 346) bifaces 0 0.2% 0 blades 0.4% 0.5% 0 chunks 8.4% 2.5% 3.5% cores 3.9% 1.6% 0.3% hammerstones/ hammerstone flakes 0.6% 0.9% 1.4% flake blades (fortuitous blades) 1.5% 1.0% 1.2% flakes 85.2% 93.1% 93.6% bifacial thinning flakes 0 0.2% 0 produce flakes/spalls from nodules (Awe and Healy 1994: 197; Ebert 2017: 135, Table 4.2; Ebert and Awe, this volume) ( Figure 3 ). All of the sourced obsidian dated to the Cunil phase at Cahal Pech comes from El Chayal (ECH) (Kersey 2007: 5, Table 2; Ebert 2015: 214, 2017: 136, 139, Table 4.4; Ebert and Awe, this volume). Importantly, the earliest evidence for obsidian blade technology in the Cunil phase is provided by a prismatic blade fragment recovered from Structure B4 (Ebert 2017: 135, Table 4.2; Ebert and Awe, this volume). Currently, there is no evidence for blade production itself at Cahal Pech. Blackman Eddy Almost all of th e tools in the subassemblage analyzed by Yacubic (2006: 67) from the Kanocha phase in Structure B1 at Blackman Eddy were made from locally procured chert; however, all hammerstones or hammerstone flakes in this phase were quartzite. The cherts are consis tent with those known from limestone outcrops and nodules near the site (Yacubic 2006: 6667, 75 76). The Kanocha phase chipped chert sub assemblage consists exclusively of debitage and some production implements ( Table 3 ). As at Cahal Pech, hard hammer core reduction was used to produce both cortical and noncortical flakes, as well some flakeblades/fortuitous blades (Yacubic 2006). At Blackman Eddy, obsidian blades and one flake were encountered in early deposits. These artifacts were found in context s that contained both Kanocha and early facet Jenney Creek ceramics, and may date to the early Middle Preclassic period; however, a Kanocha date may also be possible (Brown 2003). The Early Middle Preclassic Period at Cahal Pech and Blackman Eddy Early facet Kanluk/Jenney Creek Phase (900650 B.C.) Cahal Pech In the early facet Kanluk phase, the general pattern of tool production remains similar to what was observed in the Cunil phase (Stemp 2012). The majority of tools are s imple flakes of local chert produced through hardhammer percussion. Quartzite hammerstones, and at least one chert hammerstone, were recovered in some Plaza B units (Horn 2015: 337, Table 7.9). Major technological changes come in the form of more unifac ially retouched flake tools (e.g., scrapers) and evidence for hardhammer production of bifaces and bifacial thinning flakes of local chert (Tables 1 and 2). The absence of macroflakes, macroblades, or macroblade/flake cores from these levels indicates th at the bifaces were made on nodules or cobbles. Peniche May (2016: 271272, Tables 6.26.3) and Horn (2015: 308309, 313, Fig. 7.1A, 317326, Fig. 7.2) noted the presence of unifacial chert celts, chert percussion blades, and the appearance of the first chert microdrills. Although few in number, the need for bifaces and celts may suggest a growing importance on land clearance and/or digging in soil that accompanies greater agricultural development or permanent settlement (see Potter 1991: 27). As in the Cunil phase, evidence from Group B demonstrates a continued reliance on hardhammer percussion to produce useable

PAGE 83

Stemp et al. 83 Figure 4 Bifaces produced from local chert nodules from the early fact Jenney Creek phase from Structure B1, Blackman Eddy (from Yacubic 2 006). obsidian flakes (Awe and Healy 1994). However, more obsidian prismatic blades appear in early facet Kanluk times (Awe 1992; Horn 2015; Peniche May 2016; Ebert 2017; Ebert and Awe, this volume). Despite the increase in blades, there is still no evi dence for blade production at Cahal Pech itself. Obsidian from San Martn Jilotepeque (SMJ) becomes the dominant source in this phase with small quantities of Ixtepeque (IXT) obsidian present as well (see Awe and Healy 1994; Kersey 2007: 5: Table 2; Ebert 2015: 214; 2017: 139, Table 4.4; Ebert and Awe, this volume). Blackman Eddy At Blackman Eddy, the overall pattern of raw material procurement in the early facet Jenney Creek phase lithic sub assemblage remains generally consistent with the earlier Kanocha phase sub assemblage. Most artifacts were made from locally obtained cherts with very few quartz or quartzite pieces. There is continued heavy reliance on hardhammer chert flakes and cores; however, four chert bifaces provide evidence for this t ool type at the site ( Figure 4). Moreover, some of the chert flakes from this period are bifacial thinning flakes (Yacubic 2006: 77, Table 7). Chert drills on flakes from Blackman Eddy were found associated with marine shell debitage in early facet Jenney Creek deposits (Table 3) suggesting early craft production. The first securely dated evidence of obsidian, both as blades and hardhammer flakes, occurs in the early facet Jenney Creek Figure 5 Heavily patinated stemmed macrobl ade medial fragment of NBZ chert from the late facet Kanluk phase from the South Trench, Group B, Cahal Pech. phase, as noted above. Based on sourcing work by Kersey (2007: 4: Table 1; see Ebert 2015: 214), some of these prismatic blade fragments origina ted from SMJ. One possible biface fragment, perhaps an eccentric, was from an unknown source (Kersey 2007: 8). The Late Middle Preclassic Period at Cahal Pech and Blackman Eddy Late facet Kanluk/Jenney Creek Phase (650350 B.C.) Cahal Pech In the late facet Kanluk phase at Cahal Pech, there continues to be heavy use of local chert; however, NBZ chert is present for the first time. Although there is still reliance on local production of chert flakes from nodules, there is an increase in evidence for bi face production using hardhammer percussion (Tables 1 and 2). Biface fragments and bifacial thinning flakes made from local chert from Structure B4 and Plaza B at Cahal Pech provide good evidence for biface technology (Stemp 2012; Horn 2015: 337, Table 7 .9; Peniche May 2016: 273, Table 6.4). More bifaces suggest the possibility of greater need for agricultural tools at this time.

PAGE 84

Maya Lithic Technology at Cahal Pech and Blackman Eddy 84 Horn (2015: 310313, Fig. 7.1B) suggested a tranchet adze was recovered from Plaza B in this phase; however, this artifact does not resemble examples from northern Belize (Shafer and Hester 1983, Shafer 1991). Instead, the first tool types at Cahal Pech from northern Belize workshops are stemmed macroblade fragments (Shafer and Hester 1983; Shafer 1991) recovered from the South Trench and Plaza B (Stemp 2012; Peniche May 2016: 270, Fig. 6.4, 273, Table 6.4; Horn 2015: 314315) ( Figure 5 ). Minimally patinated macroblade fragments can be identified as NBZ ch ert. Drills produced on chert flakes and small burin spall drills were recovered from late facet Kanluk phase levels in Structure B4, in the South Trench, and in Plaza B (Stemp 2012; Horn 2015; Peniche May 2016) ( Figure 6 ). The need to perforate hard mate rials, like shell, may indicate an emerging craft industry focusing on bead production. Other drills on burin spalls at this time come from nearby Cas Pek (Lee and Awe 1995) Based on the recovery of finished obsidian prismatic blades and a few examples o f production waste (i.e., exhausted cores or core fragments) in this phase, most blades continued to arrive at Cahal Pech in finished form (Awe and Healy 1994; Horn 2015; pers. comm., 2017; Peniche May 2016). Most obsidian still originated from SMJ and at least one blade segment was sourced to Ucareo in Mexico (Kersey 2007: 5, Table 2; Ebert 2017: 139, Table 4.4; Ebert and Awe, this volume), suggesting an expansion of trade relations further north. Blackman Eddy At Blackman Eddy, there is continued relian ce on locally available chert and quartzite (Yacubic 2006) in late facet Jenney Creek and an increase in the diversity of obsidian sources represented by prismatic blades. The chert technology primarily consisted of core reduction to produce flakes (Table 3). Although bifaces were present, no bifacial thinning flakes were recovered (Yacubic 2006) suggesting that production did not occur near Structure B1. Chert drills were also recovered, indicating the continued production of mari ne shell beads (Cochran 2009). Figure 6 Burin spall drills of local chert from the late facet Kanluk phase, South Trench, Group B, Cahal Pech. Given the absence of any obsidian debitage or polyhedral blade cores or core fragments within the sub assemblage analyzed, obsidia n prismatic blade segments from Blackman Eddy indicate acquisition of finished blades (Yacubic 2006). As in the previous phase, the dominant source at the site is still SMJ (88%) with single artifacts coming from ECH (5.5%) and IXT (5.5%), respectively (K er sey 2007: 4, Table 1). Comparisons to Other Sites: Western and Northern Belize in the Terminal Early Preclassic Late Middle Preclassic Pacbitun As at Cahal Pech, the earliest obsidian at Pacbitun occurs in the form of a hardhammer flake recovered from the Mai complex (1020820 B.C.) (Awe and Healy 1994: 198). Later in the Middle Preclassic, Pacbitun (Hohmann 2002; Powis et al. 2009) provides good evidence for bead manufacture using chert burin s pall drills and shell beads. Obsidian blades are reported in the Middle Preclassic (Awe and Healy 1994: 198) with sourced obsidian identified as

PAGE 85

Stemp et al. 85 Table 4 Trends in raw material types, tool types, and production techniques from the Prec eramic to the late facet Kanluk/Jenney Creek phase in Western Belize. Phase Tool types Raw Materials Production Technology Preceramic (Archaic?) Stemmed bifaces Constricted unifaces Simples flakes Local chert CBZ chert *Obsidian flakes in Southern Belize* Hard hammer percussion Soft hammer percussion Indirect percussion (notching) Pressure flaking (alternate beveling) Hafting Projectiles Cunil/Kanocha Simple flakes Retouched flake tools Prismatic blades Local chert Obsidian Hard hammer percussion Bipolar percussion Indirect percussion/punch (known locally)? EF Kanluk/ Jenney Creek Simple flakes Retouched flake tools Bifaces Prismatic blades Fortuitous blades Crude choppers/ unifacial celts Local chert Obsidian Hard hammer percussion Bipolar percussion Bifacial thinning Indirect percussion/punch (known locally)? Hafting? LF Kanluk/ Jenney Creek Simple flakes Retouched flake tools Bifaces Prismatic blades Burin spall drills Fortuitous blades Crude choppers/ unifacial celts Stemmed macroblades Local chert NBZ chert Obsidian Hard hammer percussion Bipolar percussion Bifacial thinning Hafting Burination Indirect percussion/punch (known locally)? originating from ECH (Kersey 2007: 13, Table 3). Cuello Analyzed together, the lithic artifacts from both the Swasey (1200900 B.C.) and Bladen (900650 B.C.) ceramic phases provide some comparative information for early chipped stone tool technology in northern Belize. The chipped chert assemblage consists of a variety of tools (McSwain 1991a: 168, Table 8.1; 1991b: 342, Table 4), in particular large bifaces of Colhalike and local chert, stemmed macroblades, some small bifaces, some drills, many blades, large numbers of scrapers, and pointed tools. Although most tools were made from local chert and chalcedony, the ea rly inhabitants of Cuello also used significant numbers of tools made from NBZ chert, which they most likely produced themselves. The chipped chert technology at Cuello is essentially unchanged in the subsequent Lopez Mamom phase (McSwain 1991a: 168, Tabl e 8.2). Chert debitage recovered at this site predominantly consisted of local chert, but NBZ chert flakes were also recovered in significant numbers, representing both flake core technology and onsite biface production (McSwain 1991b: 341, Table 2; 344345). Based on clarifications by Hammond concerning a processing error (see Awe and Healy 1994), there was no obsidian associated with Swasey phase deposits. As such, the earliest obsidian artifacts consist of three prismatic blades (see Johnson 1991: 169, 171, Table 8.6) from the Bladen phase. In the later Lopez Mamom phase, three more blades are reported (Johnson 1991: 171, Table 8.6).

PAGE 86

Maya Lithic Technology at Cahal Pech and Blackman Eddy 86 Hammond (1991: 197) states that three of the earliest obsidian artifacts from Cuello dated to the Bladen phase were from SMJ. Colha The Bolay ceramic complex at Colha is contemporaneous with Bladen at Cuello. The chipped chert artifacts from this complex were divided into two formal tool groups [a blade subassemblage and a biface sub assemblage], as well as a casual/ ad hoc tool group consisting of a trimmed flake/uniface sub assemblage (Potter 1991: 21; see Shafer and Hester 1983). The blade subassemblage primarily consisted of hardhammer macroblades and smaller blades, some of which were transformed into burins and burin spalls used to drill marine shell beads (Potter 1991: 21, 24; also see Hester and Shafer 1984 for Labpek). The biface sub assemblage consists of three tool types, including a T shaped adze, a wedgeform adze, and a celt or biface form, all of which we re likely produced using a hardhammer technique on macroblade blanks (Potter 1991: 2527). Given the regional distribution of formal stone tools produced on high quality NBZ chert at other sites, the volume of manufacture, and the production of shell bea ds, Shafer and Hester (1991: 82, 91) argued that Colha was engaged in cottage industry craft specialization in the Middle Preclassic. In terms of the obsidian from Middle Preclassic Colha, Dreiss and Brown (1989: 68, Table 2, see Dreiss 1988: 28, Table 1 6, 129 Appendix C) reported that all of the sourced obsidian was from SMJ. However, Brown et al.s (2004: 231, Table 3) sourcing of a larger sample of obsidian from the Middle Preclassic indicated that a significant proportion of the obsidian (32%) originated from ECH and that a small amount also arrived from IXT (4%). Kaxob At Kaxob, McAnany and Peterson (2004: 296299, Fig. 11.15, Tables 11.7, 11.9) noted that a substantial amount of debitage (including drill tips) and many different formal tool types, including oval bifaces, other biface forms, blades, T shaped unifaces, and both stemmed and unstemmed macroblades, were recovered from the early facet Chaakkax phase (800600 B.C.). These tool s were made from local chert, local chalcedony, and NBZ chert. Three obsidian blade segments were recovered from the early facet Chaakkax phase as well; however, there was no evidence for on site production. All sourced obsidian came from SMJ (McAnany 2004: 308; Fig. 12.2, 309 ). Comparisons to Other Sites: The Pasion River Region of Guatemala in the Terminal Early Preclassic Late Middle Preclassic Ceibal and Caobal A similar reliance on expedient flakecore technology was noted at sites in the Real Xe phase (1000700 B.C.) in the Pasion River region of Guatemala. At Ceibal, the earliest stone tools were made predominantly from chert (97.6%) with far fewer produced from obsidian (Aoyama 2017a: 281, Table 2). At Caobal, chipped stone from the earliest deposits at the site also mostly consisted of chert (96.4%) with very few obsidian artifacts (Aoyama and Munson 2012: 35, Table 2). The chipped chert was overwhelmingly represented by hardhammer percussion flakes, flake tools (e.g., denticulates and scrapers), and simple flake cores, although chert biface and blade technology was present at Ceibal in the Real Xe phase (Aoyama 2017a: 282, Table 3; Aoyama and Munson 2012: 35, 37, Table 3). Hard hammer reduction of obsidian nodules or macroflakes also occurr ed at Ceibal and Caobal. Evidence for blade technology, in the form of prismatic blades, is present as well (Aoyama 2017a: 283, Table 4; Aoyama and Munson 2012: 35). At Caobal, obsidian only came from ECH (Aoyama and Munson 2012: 35). In addition to ECH (81.7%) obsidian, small quantities of SMJ (17%) and IXT (1.3%) material were also found at Ceibal (Aoyama 2017a: 281, Table 1; but the percentages in the early Middle Preclassic in Aoyama 2017b: 217, Table 2 are ECH 74.1%, SMJ 24.5%, IXT 1.4%). No polyhedral cores, macroblades, percussion blades, or initial series blades occurred at Caobal, indicating acquisition of finished blades. However, evidence for on site blade production is found at Ceibal. This is the earliest documented blade production for SMJ

PAGE 87

Stemp et al. 87 and ECH obsidian in the Maya lowlands (Aoyama 2017a: 282283; 2017b: 224). In the Escoba Mamom phase (700 350 B.C.) at Ceibal and Caobal, there is heavy reliance on local informal chert tools in the form of flakes and retouched flake tools. At Cei bal, there is also evidence for continued local chert blade production, as well as biface technology based on the recovery of biface fragments and bifacial thinning flakes (Aoyama 2017a: 282, Table 3). An oval biface fragment and evidence for biface reduc tion in the form of two thinning flakes do not appear until the Mamom phase at Caobal (Aoyama and Munson 2012: 3637, Table 3) In the Escoba Mamom phase, there is evidence for local prismatic blade production at Ceibal and the amount of obsidian reaching the site substantially increased as evidenced by the higher percentage of obsidian (29.3%) in the chipped stone assemblage (Aoyama 2017a: 281, Table 1). In this phase, most obsidian comes from SMJ (93.4%) and much smaller amounts were procured from ECH (6 .4%) and IXT (0.2%) (Aoyama 2017a: 281, Table 1, 2017b: 217, Table 2). Blade production at Ceibal is demonstrated by the recovery of exhausted polyhedral cores, macroblades, percussion blades, crested blades, flakes from polyhedral cores, and platform r ejuvenation flakes. As such, obsidian, particularly SMJ, was imported as large polyhedral cores that were used to produce blades locally. The comparatively high percentage of cortex on obsidian artifacts (16.4%) and lower percentage of prismatic blades ( 38.6%) also indicate the continued import of large flakes and small nodules of SMJ for the production of percussion flakes as well (Aoyama 2017a: 289) At Caobal, the amount of obsidian increased slightly from the earlier phase. Most obsidian (69.1%) orig inated from SMJ, while smaller amounts of ECH (29.4%) and IXT (1.5%) obsidian were recovered (Aoyama 2017a: 289). The change in the primary source of obsidian at Caobal in this phase is similar to that observed at Ceibal. Moreover, low percentages of cortex on obsidian at Caobal suggests that Ceibal was provisioning Caobal with finished blades and semi exhausted cores (Aoyama 2017a: 289; 2017b: 225). Tikal Chert blade technology is first documented in the early Middle Preclassic Eb phase (800600 B.C.) at Tikal ( MoholyNagy 2003: Table 2.30). In the subsequent Tzec phase (600350 B.C.) deposits, chert prismatic blades, exhausted polyhedral cores, oval bifaces, and stemmed bifacial points were recovered ( Moholy Nagy 2003: Table 2.30). Like chert blades, obsidian blades first appear in the Eb phase. In the Tzec phase, more obsidian blades (17) and a polyhedral core fragment were recovered from the site, suggesting the possibility of local blade production (Moholy Nagy 2003: Tables 3.18, 3.24, 3.29; MoholyNagy et al. 2013: 87). The earliest Middle Preclassic ob sidian blades were predominantly from SMJ, although some material sourced to ECH is reported (Moholy Nagy et al. 2013: 89, Table 6). Evidence for local blade production, based on two exhausted SMJ blade core fragments, also occurs nearby in the Ah Pam pha se of the Middle Preclassic (750 550 BC) in the central Petn Lakes region (Rice e t al. 1985: 595). Comparisons to Other Sites: The Copan Valley and the La Entrada Region in Honduras in the Terminal Early Preclassic Late Middle Preclassic Copan and the La Entrada Region In Honduras, the earliest stone tools date to the Early Preclassic Rayo phase (1400 900 B.C.) at Copan and appear in the form of expedient core and flake technology (Aoyama 1999:53). Most of the material (90.5%) represents general hard hammer debitage. Unlike the sites in Belize and Guatemala, the majority of lithic artifacts (90.5%) were made from obsidian and only eight were produced from local chert. However, this may be a product of sampling (G. Braswell, pers. comm., 2017). Despi te the high proportion of obsidian, of which 97.4% was identified as originating from the IXT source, no prismatic blades or polyhedral cores were recovered. Aoyama (1999:53) notes that 27% of the obsidian artifacts possess cortex, which suggests the IXT obsidian was acquired in the form of large flake

PAGE 88

Maya Lithic Technology at Cahal Pech and Blackman Eddy 88 spalls or small nodules, possibly indicating direct acquis ition or downthe line trading. In the Middle Preclassic (900 300 B.C.), the majority of tools in the Copan Valley and the La Entrada region were mad e from obsidian (59.3% Copan; 55.8% La Entrada). The chert artifacts are mostly hard hammer flakes (94.3%), some informal flake tools (e.g., denticulates, scrapers and a drill), and flake cores, with very few chopping tools (Aoyama 1999: 59, Table 4.1) The obsidian artifacts from the Copan Valley overwhelmingly consist of hard hammer flakes (93.4%), informal flake tools, and some prismatic blades (2.7%) (Aoyama 1999: 60, Table 4.2). Most (99.5%) of the obsidian tools were made from IXT obsidian. The majority of IXT obsidian continued to be acquired as large flake spalls or small nodules. The rest of the obsidian came from La Esperanza (0.3%) and ECH (0.2%). No polyhedral cores or macroblades were recovered in the Copan Valley, indicating that there was no local prismatic blade production (Aoyama 1999: 65). At the Middle Preclassic sites in the La Entrada region, most artifacts were made from obsidian (67.6%) with fewer produced from local cherts (32.4%). As in the Copan Valley, chert tools were mostly hard hammer debitage (86.8%). Informal flake tools, flake cores, and a few crude chopping tools were also recovered (Aoyama 1999: 62, Table 4.4) Unique to La Entrada in the Middle Preclassic is the fact that San Luis/Source Y was the source of most (71.1%) of the obsidian artifacts. The remaining tools consisted of IXT obsidian. Most obsidian artifacts are general debitage (85.2%) or flake tools (5%) made using hardhammer percussion. The absence of prismatic blades made from San Luis obsidian, the recovery of flake cores, and the very high percentage of artifacts with cortex (85.8%) indicates that this obsidian was acquired directly from nearby outcrops or local streams. The eight IXT obs idian blades, however, imply downthe line exchange at the La Entrada sites in this period (Aoyama 1999: 63) Discussion and Conclusion Evidence for changes in lithic raw material use and stone tool technology can be traced from the preceramic through to the Middle Preclassic periods in western Belize (Table 4). Hafted formal tools made from fine grained stone (e.g., NBZ chert) and the complex array of manufacturing techniques that characterize preceramic lithic technology are not found in the subsequent Cunil/Kanocha phases at Cahal Pech and Blackman Eddy (Lohse 2010; Stemp et al. 2016; cf. Clark and Cheetham 2002; Iceland 2005). In Cunil/Kanocha, stone tools were manufactured from locally obtained chert that was used to make simple flake tools using har dhammer and bipolar percussion. At Cahal Pech, obsidian is introduced in the form of hardhammer flakes struck from small ECH nodules with few prismatic blades. Our current analysis of Cunil and Kanocha lithic material does not suggest technological con tinuity between the preceramic and the earliest Maya ceramicusing populations in western Belize when compared to the scattered Archaic lithic tools recovered. We recognize that more data are necessary to further support this assertion. New evidence from Early Xunantunich, however, may shed light on this issue as preceramic deposits have been found underlying Early and early Middle Preclassic occupation (Brown et al. 2011). In the early facet Jenney Creek/Kanluk phase at Cahal Pech and Blackman Eddy, flak e core reduction still dominates, but bifacial chert tools are present. Obsidian blade technology is more visible, but there is no evidence for local blade production. In the late facet Jenney Creek/Kanluk phase, local chert flakecore reduction continues as does biface production. There is also the introduction of finished tools made from NBZ chert, in particular the stemmed macroblade. There are significant numbers of burin spall drills, and associated shell beads. Prismatic blades become the dominan t obsidian tool type and hardhammer reduction of obsidian nodules dwindles Despite general knowledge about trends in lithic technology in western Belize, more work is needed to fully flesh out lithic procurement and production in the terminal Early to la te Middle Preclassic periods in this region. Lithic connections to sites further away, such as Colha, demonstrate the early integration of Maya communities into complex regional socioeconomic networks and acquisition of obsidian

PAGE 89

Stemp et al. 89 demonstrates the establish ment of long distance exchange systems to provision sites with resources that were not found locally. Lithic data from Cahal Pech and Blackman Eddy provide good foundations for understanding the roles of chipped stone tool technology in the cultural devel opment of the first lowland Maya communities and highlight the need for more effort in reconstructing the technological transition from preceramic hunter gatherers to early settled farmers in western Belize. Acknowledgments We would like to thank the Bel ize Institute of Archaeology for their continued support of our investigations in the Belize River valley, and for providing us with permission to excavate at Cahal Pech and Blackman Eddy. Funding for the Belize Valley Archaeological Reconnaissance (BVAR) Project has been generously provided by grants from the Gordon Childe Fund of University College London, the Social Sciences Research Council of Canada, the National Science Foundation, and the Tilden Family Foundation of San Francisco, California. Funding for the Belize Valley Archaeological Projects excavations at Blackman Eddy and Cahal Pech was provided by Texas State University, Foundation for the Advancement for Mesoamerican Studies, Inc., and Southern Methodist University. Funding for lithic analy sis was provided by a Faculty Development Grant from Keene State College. We are also grateful to all the staff, local archaeologists, and students who, over more than 30 years, have contributed significantly to our investigations at Cahal Pech and Blackm an Eddy. References Aoyama, Kazuo 2017a Ancient Maya Economy: Lithic Production and Exchange Around Ceibal, Guatemala. Ancient Mesoamerica 28: 279-303. Aoyama, Kazuo 2017b Preclassic and Classic Maya Interregional and Long -Distance Exchange: A Diachronic Analysis of Obsidian Artifacts from Ceibal, Guatemala. Latin American Antiquity 28: 213-231. Aoyama, Kazuo 1999 Ancient Maya state, urbanism, exchange and craft specialization. Chi pped stone evidence from the Copan Valley and the La Entrada Region, Honduras University of Pittsburgh Memoirs in Latin American Archaeology 12, University of Pittsburgh, Pittsburgh. Aoyama, Kazuo and Jessica Munson 2012 Ancient Maya Obsidian Exchange an d Chipped Stone Production at Caobal, Guatemala. Mexicon 34: 3442. Awe, Jaime J. 1992 Dawn in the Land between the Rivers: Formative Occupation at Cahal Pech, Belize and its Implication for Preclassic Developments in the Maya Lowlands. Unpublished Ph.D. dissertation, Institute of Archaeology, University of London, England. Awe, Jaime and Paul F. Healy 1994 Flakes to Blades? Middle Formative Development of Obsidian Artifacts in the Upper Belize River Valley. Latin American Antiquity 5: 193 -205. Brown, M Kathryn 2003 Emerging Complexity in the Maya Lowlands: A View from Blackman Eddy, Belize. Ph.D. dissertation, Southern Methodist University. UMI, Ann Arbor, MI. Brown, M. Kathryn, Jennifer Cochran, Leah McCurdy, and David W. Mixter 2011 Preceramic to Postclassic: A Brief Synthesis of the Occupation History of Group E, Xunantunich. Research Reports in Belizean Archaeology 8: 209220. Brown, David O., Meredith L. Dreiss, and Richard E. Hughes 2004 Preclassic Obsidian Procurement and Utilization at the Maya Site of Colha, Belize. Latin American Antiquity 15: 222 -240. Clark, John E., and David Cheetham 2002 Mesoamericas Tribal Foundations. In The Archaeology of Tribal Societies edited by W. A. Parkinson, pp. 278 339. International Monographs in Prehistory, Archaeological Series 15, University of Michigan, Ann Arbor. Cochran, Jennifer L. 2009 A Diachronic Perspective of Marine Shell Use from Structure B1, Blackman Eddy, Belize. M.A. thesis. University of Texa s, Arlington. University Microfilms, Ann Arbor, MI. Dreiss, Meredith L. 1988 Obsidian at Colha, Belize: A Technological Analysis and Distributional Study Based on Trace

PAGE 90

Maya Lithic Technology at Cahal Pech and Blackman Eddy 90 Element Data Papers of the Colha Project, Vol. 4, Texas Archaeological Research Labor atory, University of Texas at Austin and Center for Archaeological Research, University of Texas at San Antonio, Austin. Dreiss, Meredith L. and David O. Brown 1989 Obsidian Exchange Patterns in Belize. In Prehistoric Maya Economies of Belize edited by P .A. McAnany and B.L. Isaac, pp. 57 90, Research in Economic Anthropology, Supplement 4, JAI Press, London. Ebert, Claire E. C. 2017 Preclassic Maya Social Complexity and Origins of Inequality at Cahal Pech, Belize. Unpublished Ph.D. dissertation, Department of Anthropology, Pennsylvania State University, State College, PA. Ebert, Claire E. 2015 Chemical Characterization of Obsidian Artifacts from Cahal Pech and Lower Dover, Belize. In The Belize Valley Archaeological Reconnaissance Project: A Rep ort of The 2014 Field Season, edited by Julie A. Hoggarth and Jaime J. Awe, pp. 210 221. Belize Institute of Archaeology, National Institute of Culture and History, Belmopan. Ebert, Claire C. and Jaime J. Awe 2018 Reconstructing Preclassic Maya Household Economies in the Belize River Valley. Research Reports in Belizean Archaeology 15. Gibson, Eric C. 1991 A Preliminary Functional and Contextual Study of Constricted Adzes from Northern Belize. In Maya Stone Tools: Selected Papers from the Second Maya Lithic Conference edited by T.R. Hester and H.J. Shafer, pp. 229 -237, Monographs in World Archaeology No.1, Prehistory Press, Madison. Hammond, Norman 1991 Obsidian Trade. In Cuello: An Early Maya Community in Belize edited by N. Hammond, pp. 197 -198, Cambridge University Press, Cambridge. Hester, T. R. and H. J. Shafer 1984 Exploitation of Chert Resources by the Ancient Maya of Northern Belize, Central America. World Archaeology 16: 157-173. Hohmann, Bobbi M. 2002 Preclassic Maya Shell Ornament Production in the Belize Valley, Belize. Ph.D. dissertation, Department of Anthropology, University of New Mexico. Albuquerque, NM. Horn III, Sherman W. 2015 The Web of Complexity: Socioeconomic Networks in the Middle Preclassic Beliz e Valley. Unpublished PhD Dissertation, Tulane University, New Orleans, LA. Horowitz, Rachel A. 2017 Ancient Maya Economic Variability: Lithic Technological Organization in the Mopan Valley, Belize. Unpublished PhD dissertation, Tulane University, New Orleans, LA. Iceland, Harry B. 2005 The Preceramic to Early Middle Formative Transition in Northern Belize: Evidence for the Ethnic Identity of the Preceramic Inhabitants. In New Perspectives on Formative Mesoamerican Cultures edited by T. G. Powis, pp. 15 26, British Archaeological Reports, International Series 1377, Oxford. Iceland, Harry B. 1997 The Preceramic Origins of the Maya: The Resul ts of the Colha Preceramic Project in Northern Belize. Unpublished Ph.D dissertation, Department of Anthropology. University of Texas at Austin, Austin, TX. Johnson, Jay K. 1991 Obsidian: A Technological Analysis. In Cuello: An Early Maya Community in Belize edited by N. Hammond, pp. 169172. Cambridge University Press, Cambridge. Kelly, Thomas C. 1993 Preceram ic Projectile Point Typology in Belize. Ancient Mesoamerica 4: 205-227. Kersey, Kimberly M. 2007 Emerging Elite Economies: formative Period Obsidian Distribution in the Belize River Valley. IAOS Bulletin: International Association for Obsidian Studies 36: 2 -19 Lee, David F. and Jaime J. Awe. 1995 Preclassic Architecture, Burials, and Craft Specialization at the Cas Pek Group, Cahal Pech. In Belize Valley Preclassic Maya Project: Report on the 1994 Field Season, edited by P. F. Healy and J. J. Awe, pp 95 -115. Trent University, Department of Anthropology Occasional Papers in Anthropology No.10, Peterborough. Lohse, Jon C. 2010 Archaic Origins of the Lowland Maya. Latin American Antiquity 21: 312 -352. Lohse, Jon C., Jaime Awe, Cameron Griffith, Robert M. Rosenswig, and Fred Valdez, Jr. 2006 Preceramic Occupations in Belize: Updating the Paleoindian and Archaic Record. Latin American Antiquity 17: 209226.

PAGE 91

Stemp et al. 91 Lohse, Jon C., and Michael B. Collins 2004 Lithic Artifacts Recovered in association with Pleistocene Fauna from Actun Halal Cave, Western Belize. Unpublished report on file with the Western Belize Regional Cave Project, Bloomington, and the Institute of Archaeology, Belmopan, Belize. McAnany, Pat ricia A. 2004 Obsidian Blades and Source Areas. In Kaxob: Ritual, Work, and Family in an Ancient Maya Village edited by P. A. McAnany, pp. 307-315. Monumenta Archaeologica 22, Cotsen Institute of Archaeology, University of California, Los Angeles. McAna ny, Patricia A. and Polly A. Peterson 2004 Tools of the Trade: Acquisition, Use, and Recycling of Chipped Stone. In Kaxob: Ritual, Work, and Family in an Ancient Maya Village edited by P. A. McAnany, pp. 279 -305. Monumenta Archaeologica 22, Cotsen Instit ute of Archaeology, University of California, Los Angeles. McSwain, Rebecca 1991a Chert and Chalcedony Tools. In Cuello: An Early Maya Community in Belize edited by N. Hammond, pp. 160173, Cambridge University Press, Cambridge. McSwain, Rebecca 1991b A Comparative Evaluation of the Producer Consumer Model for Lithic Exchange in Northern Belize, Central America. Latin American Antiquity 2: 337351. Moholy -Nagy, Hattula 2003 The Artifacts of Tikal: Utilitarian Artifacts and Unworked Material. Tikal Rep ort No. 27, Part B. University of Pennsylvania Museum of Archaeology and Anthropology, University of Pennsylvania Press, Philadelphia. Moholy -Nagy, Hattula, James Meierhoff, Mark Golitko, and Caleb Kestle 2013 An Analysis of pXRF Obsidian Source Attributions from Tikal, Guatemala. Latin American Antiquity 24: 7297. Peniche May, Nancy 2016 Building Power: Political Dynamics in Cahal Pech, Belize during the Middle Preclassic, Unpublished PhD, Department of Anthropology, University of California -San Diego, San Diego, CA. Potter, Dennis R. 1991 A Descriptive Taxonomy of Middle Preclassic Chert Tools at Colha, Belize. In Maya Stone Tools: Selected Papers from the Second Maya Lithic Conference, edited by T.R. Hester and H.J. Shafer, pp.2129, Monographs in Wor ld Archaeology No.1, Prehistory Press, Madison. Powis, Terry G., Paul F. Healy and Bobbi Hohmann 2009 An Investigation of Middle Preclassic Structures at Pacbitun. Research Reports in Belizean Archaeology 6: 169 177. Rice, Prudence M., Helen V. Michel, F rank Asaro, and Fred Stross 1985 Provenience Analysis of Obsidians from the Central Petn Lakes Region, Guatemala. American Antiquity 50: 591604. Shafer, Harry J. 1991 Late Preclassic Formal Stone Tool Production at Colha, Belize. In Maya Stone Tools: Selected Papers from the Second Maya Lithic Conference edited by T.R. Hester and H.J. Shafer, pp. 31-44, Monographs in World Archaeology No.1, Prehistory Press, Madison. Shafer, Harry J. and Thomas R. Hester 1991 Lithic Craft Specialization and Productio n Distribution at the Maya Site of Colha. World Archaeology 23: 7997. Shafer, Harry J. and Thomas R. Hester 1983 Ancient Maya Chert Workshops in Northern Belize, Central America. American Antiquity 48: 519 -543. Stemp, W. James 2012 Changes in Stone Tool Technology from the Late Archaic to the First Maya Communities in Western Belize. Paper presented at the Sixth annual Maya at the Playa Conference, Palm Coast, FL. Stemp, W. James and Jaime J. Awe 2013 Possible Variation in the Late Archaic Period Bifa ces in Belize: New Finds from the Cayo District of Western Belize. Lithic Technology 38:1731. Stemp, W. James, Jaime J. Awe, M. Kathryn Brown, Eleanor Harrison -Buck, Christophe G. B. Helmke, Gabriel D. Wrobel, and Jason Yaeger 2018 Four Preceramic Points Newly Discovered in Belize, a Comment on Stemp et al. ( Latin American Antiquity 27(3), 2016, pp. 279299). Latin American Antiquity in press. Stemp, W. James, Jaime J. Awe, Keith M. Prufer, and Christophe G.B. Helmke 2016 Design and Function of Lowe and Sawmill Points from the Preceramic Period of Belize. Latin American Antiquity 27, 3: 279-299. Yacubic, Matthew P. 2006 The Chipped Stone Tool Industries of Blackman Eddy, Belize. M.A. thesis, Brigham Young University, Provo, UT.

PAGE 92

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 93104 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 8 PRECLASSIC ANIMAL RESOURCE USE AND THE O RIGINS OF ANCIENT MAYA LIFEWAY S AND SOCIETY: CONTR IBUTIONS FROM BELIZE ZOOARCHAEOLOGY Norbert Stanchly and Chrissina Coleen Burke This paper examines the important contributions made by zooarchaeological research in Belize to our understanding of Preclassic animal resource use and the origins of Maya lifeways and societies. The Preclassic Maya utilized animals for both dietary and nondietary purposes. Although animals played an important role as basic nutritional supplements and as raw materials for artifact production, zooarchaeologists also recognize the importance of fauna in the development of social relations. Animals figured prominently in Maya creation mythology and served to create, maintain, and solidify Preclassic political and social relationships via both public and private performance and mechanisms such as feasting. Introduction This paper summarizes contributions from zooarchaeological research in Belize to our current understanding of the u se of faunal resources by the Preclassic Lowland Maya. The Classic Maya exploited a broad and diverse array of fauna for basic nutritional requirements and as raw material for artifact production (Powis et al. 1999; Emery 2008). Animals also played an im portant social role in the dynamics of sociopolitical relationships (Pohl 1985). Differential access to fauna was based on status or authority (Emery 2002). The recovery of faunal remains from burial and cache contexts attests to their importance in Classic Maya ideology (Pohl 1983). Recent archaeological research at several ancient Maya centres (Figure 1) indicates many of the hallmarks of Classic Maya civilization had their genesis during the earlier Preclassic, or Formative period (Adams 1977; Garber 2004; Hammond 1991; Hanson 1989, 1990; Healy and Awe 1995; McAnany 2004; Pendergast 1981). These include, for example, the emergence of ranked and other forms of complex hierarchical society, public art and architecture, craft specialization, longdistance interregional trade networks, increasingly sophisticated subsistence economies, the development of writing, and the introduction of kingship (Hammond 1986; McAnany 2004). In assessing the Preclassic use of faunal resources by the ancient Lowland Maya we c ontextualize faunal exploitation with respect to Classic period patterns of animal resource use. To what extent, if any, do Classic period patterns of faunal use find their origins during the Preclassic? Zooarchaeological research in Belize has a history that can be traced back over five decades. Prior to the beginnings of excavations at Altun Ha, David Pendergast understood the value of faunal analysis to understanding Maya subsistence patterns. He assembled a multidisciplinary team from the Royal Ontar io Museum to collect as many modern mammal and bird species as possible for building a skeletal reference collection to aid in the analysis of any faunal remains expected to be collected during the following six years of excavation at the site (Pendergast 1979). This collaborative approach to understanding ancient Maya animal use continued in the 1970s with the ROMs excavations at two cave sites in Belize: Actun Polbilche and Eduardo Quiroz Cave, both excavated by Pendergast with the analysis of the animal remains undertaken by Dr. Howard Savage and Elizabeth Luther, at the University of Toronto (Pendergast 1971 1974). This represents some of the earliest faunal research of cave sites in the Maya Lowlands. Throughout the 1970s and 1980s, faunal research became an important component of analysis at several sites in the Maya Lowlands, including here in Belize. One of the most influential studies on the Preclassic Maya was by Norman Hammond at the site of Cuello in the Orange Walk District (Hammond 1991). The analyses of the large faunal assemblage from Cuello has provided important insights on Preclassic Maya animal resource use (Carr and

PAGE 93

Preclassic Animal Resource Use Figure 1. Location of s ites m entioned (Map by Claire Ebert). Fradkin 2008; CluttonBrock and Hammon 1994; Wing and Scudder 1991). In the succeeding decades, zooarchaeological analysis has become a standard facet of archaeological research. Our understanding of the relationship between the ancient Maya and their animal resources has shed important light on patterns o f dietary use, bone and shell artifact production, craft specialization, trade and exchange, differential access to faunal resources, the importance of animals in the establishment and maintenance of socio political organization, and the importance of fauna in Maya ritual and ideology. Through a comparison of these assemblages we contextualize Preclassic Maya faunal utilization within the temporal framework of the later Classic period. To what extent, if any, do Classic period patterns of faunal use find t heir origins during the Preclassic? What patterns can be discerned in the use of animals as subsistence items during the Preclassic? Were there regional differences in the procurement of fauna for food? Was there differential access to food sources based on rank or status? If so, when did this emerge? What do faunal assemblages tell us about Preclassic trade and exchange of animal resources? Was there craft specialization in the manufacture of bone and shell artifacts? What do faunal remains tell us of Preclassic Maya ideology? The Maya Preclassic Many of the hallmarks of Classic Maya civilization had their genesis during the earlier Preclassic, or Formative period. These include, for example, the emergence of ranked and other forms of complex hierar chical society, public art and architecture, craft specialization, long distance interregional trade networks, increasingly sophisticated subsistence economies, the development of writing, and the introduction of kingship. The Preclassic, or Formative peri od, spans approximately 2,250 years with a traditional, but inferred, start date of 2,000 BC and ending ca. AD 250. The Preclassic is divided into three main divisions, Early Preclassic, Middle Preclassic, and Late Preclassic. The Middle and Late Preclassic periods have been further subdivided. The Middle Preclassic has an early and late facet, while the latter three hundred years of the Late Preclassic are sometimes referred to as the Terminal Late Preclassic or Proto classic period. The Early Preclassi c period has, to date, proven to be largely invisible archaeologically, although recent investigations at Cuello in northern Belize, and the sites of Cahal Pech (Awe 1992; Healy et al. 2004) and Blackman Eddy (Garber et al. 2004) in western Belize, have radiocarbon dates suggesting initial occupation of these sites during the Terminal Early Preclassic, ca. 1,200 9000 BC. This period witnessed the beginnings of settled village life, the adoption of maize agriculture, and the introduction of ceramic technology. These early communities were likely small with egalitarian social and political organization, while ideology and religion were probably shamanic in nature. The Middle Preclassic period saw the expansion of village settlements along rivers throughout the Lowlands. During the Early Middle Preclassic, the re is first evidence of public architecture (Micheletti and Powis 2015; Garber et al. 2004) and the first indications of

PAGE 94

Stanchly and Burke 95 social and political complexity. By the Late Middle Formative we have ranked societies in the form of chiefdoms and the introduction of monumental architecture (Hansen 1989, 1990). Ideology and religious practice continues to be shamanic in origin, and early ch iefs were probably village shamans (Freidel 1992). By the Late Preclassic, Maya society underwent profound changes, many of which reached their florescence during the Classic period. Formalized monumental architecture, including vaulted tombs and stuccodecorated facades, is seen at several centres throughout the lowlands and definitive architectural styles are apparent including the tripartite arrangement of buildings atop the summit of structures as seen at sites such as El Mirador in Guatemala (Hansen 1990), and Lamanai in northern Belize (Pendergast 1981). The Late Preclassic also saw the emergence of the archaic state (Marcus 1993) and the adoption of divine kingship (Freidel and Schele 1988). In summary, the Preclassic period saw the beginnings of se ttled village life, the adoption of ceramic technology, increasing social and political complexity, an ideology grounded initially in kin based shamanism, and culminating in the appearance of large urban centres with monumental architecture, writing system s, strict status divisions, and the appearance of divine kings or lords. Preclassic Faunal Utilization Archaeological research on the Preclassic has a long history in Maya archaeology dating back to the Carnegie Institution of Washingtons Uaxactun Projec t of the 1930s (Hammond 1982:355; McAnany 2004:3). Despite this, it has been argued that our knowledge of the Preclassic Maya is a poor reflection of the wealth of knowledge obtained from research into the Classic Maya (Healy and Awe 1995:2). Although we agree this is certainly true with respect to the study of elements of ancient Maya society; however we argue this does not appear to apply to our understanding of faunal utilization. Despite the exponentially greater amounts of archaeology conducted on C lassic period centres, the amount of zooarchaeological data recovered from Preclassic contexts is at least comparable if not more substantial than that recovered from later Classic contexts. Belize contribution to our understanding of Preclassic animal ut ilization is significant, and largely reflects the concerted efforts of many researchers whose primary interest has been to document and investigate the beginnings of Maya civilization ( Table 1 ) The sites we discuss here include both inland and coastal M aya centres and include: Cerros, Cuello, Cahal Pech, Pacbitun, Colha, Kaxob, Blue Creek, Colha, Caracol, Kakabish and Blackman Eddy ( Table 2 ) The Early Preclassic Zooarchaeological data for this period come mainly from the margins of the Maya lowlands, for example, the Soconusco region of the Mexican and Guatemalan Pacific coastal plain. Even here the evidence is scant with most vertebrate faunal materials recovered from Terminal Early Preclassic contexts. At the site of Paso de la Amada, Chiapas, faunal material recovered from redeposited midden indicates a focus on estuarine resources with the vertebrate assemblage dominated by a diverse array of marine and brackish water fishes (Wake 2004). Fish species include gar, sea catfish, jacks, snapper, moja rras, shark or ray, and cichlids (Wake 2004:215, Table 1). The diversity represented among the fish taxa is also reflected in the amphibian, reptilian, and mammalian species identified. This diversity attests to the knowledge and familiarity that the inhabitants of the site had of the microhabitats within their local environments, a trend that first appears with the introduction of semi permanent and/or permanent village settlement in the Archaic period of Mesoamerica and continuing th roughout the Preclas sic period. Although some larger game such as deer and peccary are present, the vertebrate assemblage indicates a reliance on smaller species such rabbit, turtle, lizard, gopher, and fish (Wake 2004:219). This reflects an established Archaic pattern of faunal utilization in other parts of Mesoamerica such as the Tehuacan Valley (Flannery 1967). Wake (2004:211) makes additional reference to increased amounts of invertebrate remains recovered from Archaic and Early Formative period shell mounds at the site.

PAGE 95

Preclassic Animal Resource Use Evidence for animal utilization in the Early Preclassic in Belize is scant and at this point, appears restricted to materials recovered from terminal Early Preclassic levels. At Cahal Pech and Cuello, the earliest ceramic phases yielded relatively small numbers of vertebrate remains (Awe 1992; Stanchly 1994; Wing and Scudder 1991). These included mostly mammal bone at both sites and at Cuello, there is greater diversity with reptiles (i.e. freshwater turtles), amphibian and fish also represented (Wing and Scudder 1991). At Cahal Pech, Cunil phase levels have produced mainly mammal remains and freshwater shells (Stanchly 1995). Recent analysis of faunal materials from Maya Hak Cab Pek Rockshelter (Orsini 2016) in southern Belize suggest less diversity in animal taxa utilization during the Preceramic period (ca. 9,1203,348 B.C [after Orsini 2016:31]) with an emphasis on larger mammals. The Middle Preclassic Faunal assemblages from Middle Preclassic contexts at the sites of Blackman Eddy (Garber et al. 2004), Cahal Pech (Powis et al. 1999; Stanchly 1995), Cuello (Wing and Scudder 1991), Colha (Shaw 1999), Pacbitun (Healy et al. 2004; Stanchly 1999a), and Kaxob (Masson 2004a, 2004b), indicate a continued f ocus on the hunting and procurement of small terrestrial and riverine vertebrate and invertebrate species, although larger game such as deer, peccary, begin to increase in relative numbers. These larger animals would have provided significant amounts of protein, thereby supplementing a diet increasingly reliant upon maize agriculture. In addition to the continued exploitation of local microhabitats and environs, the faunal assemblages indicate increasing knowledge of more distant sources of both vertebrate and invertebrate resources. The presence of marine fish and shellfish species at distant inland sites such as Cahal Pech (Powis et al. 1999:368369) and Pacbitun (Healy et al. 2004:224) indicates the establishment of longdistance trade and exchange netw orks by the Middle Preclassic. The faunal samples recovered from these sites indicate similarities in the types of animals procured for food. Although diversity in exploited taxa is still evident, there is a common reliance on small, medium, and large gam e species such as armadillo, river turtles, agouti and paca, peccary, deer, and, with the exception of the Belize Valley sites (Stanchly and Awe 2015) and Pacbitun (Boileau 2013), domestic dog. In fact, dog becomes a dominant species at all northern Beliz e sites (e.g. Cuello, Colha, Kaxob) by the end of the Middle Preclassic, not only as a food source, but possibly also as an important food in ritual feasting (CluttonBrock and Hammond 1994; Shaw 1999). Mollusc utilization during the Middle Preclassic foc ussed primarily upon the exploitation of freshwater snails and bivalves. Beginning in the Middle Preclassic, we see a rapid increase in the use of the freshwater mollusc triad of jute ( Pachyhilus spp.), apple snail ( Pomacea flagellata ), and pearly mussel ( Nephronaias sp.). The presence of these three shellfish, often in numbers exceeding the hundreds of thousands (Healy et al. 2004:224225), in midden and construction core contexts, is almost a ubiquitous feature of Middle and Late Preclassic Maya sites. These include Cahal Pech (Stanchly 1995), Blackman Eddy (Garber et al. 2004), Pacbitun (Healy et al. 1990, 2004; Stanchly 1999a), Caracol (Cobos 1994), Minanha (Stanchly et al. 2008; Solis 2010), Kaxob (Harrigan 2004), Blue Creek (Stanchly 1999b), Kakabish (Stanchly 2013) and Lamanai (Stanchly, personal observation). The presence of large amounts of jute in these contexts is commonly interpreted as food refuse, especially where the specimens exhibit either broken or punctured spires which severs the musc le attachment of the snail from its shell (Healy et al. 1990). At some sites, the exceedingly large amounts of jute found in platform and plaza core contexts indicates their use as construction material and not necessarily solely as items of food refuse (S olis 2010; Walden and Biggie 2017). Finally, the possibility exists that the deposition of freshwater shells in such large numbers may also be ideologically based, perhaps related to some sort of water or sea related ritual (Halperin et al. 2003; Wagner e t al. 2013; Walden and Biggie 2017). The near universal occurrence of large quantities of freshwater molluscs in Middle Preclassic

PAGE 96

Stanchly and Burke 97 deposits in Maya sites requires further explanation. The presence of finished marine shell beads and debitage in Middle Prec lassic contexts at Blackman Eddy, Cahal Pech, Kakabish, Blue Creek, and Pacbitun, indicates the beginnings of craft specialization in shell bead production. Healy (1998) believes these were likely part time craft specialists engaged in cottage industry production. The presence of marine shell at distant inland sites indicates knowledge of and access to coastal resources by the end of the Early Middle Formative. The types of marine shell species used in the production of beads are informative. During th e Middle Preclassic the dominant species favoured as raw material were the conchs (Family Strombidae). Moreover, to our knowledge, there is no evidence for the use of Spondylus until the Late Preclassic. This suggests either the Maya did not possess the knowledge or skill to access the deep water habitat of the thorny oyster, or the shell held no value as a prestige item during the Middle Preclassic. The appearance of Spondylus in Late Preclassic cache and burial contexts (see McSwain et al. 1991), may be a reflection of the presence of increasingly formalized status differences and the beginnings of the institution of kingship. Our understanding of the ritual use of fauna during the Middle Preclassic is vague. Some remains have been recovered from burial or cache contexts. At Pacbitun, an Early Middle Preclassic cache contained marine shell beads (Healy et al. 2004:216). A large midden deposit associated with building construction at Blackman Eddy during the Early Middle Preclassic has been interpreted as evidence of ritual feasting (Garber et al. 2004:37). The midden contained large quantities of freshwater and marine shells, and faunal remains were dispersed throughout (Garber et al. 2004:37). In summary, data from several Middle Preclassic contexts indicate a continuation of the exploitation of locally available species. There is an increase in the relative abundance of medium to large size game. Domestic dog is important at several sites in northern Belize, but its appearance on the menu at sites within the Belize River Valley region is rare in comparison. Invertebrate remains become more common and we see the almost ubiquitous deposits of huge a mounts of freshwater shells in midden and core contexts. Marine shell bead production appears, further suggesting the beginnings of craft specialization, with conch shell the preferred raw material. The use of Spondylus as a raw material is, to our knowl edge, not known in the Maya Lowlands until the Late Preclassic. The presence of marine shell and marine fish at inland sites indicates access to coastal resources. The exact nature of this access is not known; however, we can infer the beginnings of elab orate long distance trade and exchange networks. The presence of fauna in caches and burials indicates the ir importance in Maya ideology. The Late Preclassic The faunal assemblages from the sites of Colha and Cerros have been the subject of detailed zooa rchaeological analyses (Carr 1985; Shaw 1991, 1999) and provide us with a good framework from which to examine Late Preclassic patterns of animal resource use. The Colha and Cerros faunal assemblages reflect marked differences in access to faunal resources, in both quality and quantity, the increasing importance of fauna as elite prestige items, and the further elaboration of long distance trade networks. The Colha data (Shaw 1999) point to the importance of fauna in ritual and competitive feasting as a mechanism for creating socio economic and sociopolitical relationships (see Clark and Blake 1994) Subsistence patterns at both sites indicate the increasing importance of large mammals to the diet. Marine and freshwater fish occur in the Colha assemblage and marine fish dominate the Cerros assemblage. This reflects Cerros coastal location within Chetumal Bay. Dog continues to be an important dietary contributor at Colha and also at Cerros. Their presence at Cerros is interpreted by Carr (1985:126) as r aising of dog as a supplemental food source for times when fish could not be procured to provide the s ites meat supply. Leslie Shaw (1999:9495) provided a convincing argument that the frequency of dog remains in Preclassic deposits is related to their use in ritual feasting. She stated that if they

PAGE 97

Preclassic Animal Resource Use Table 1 Subsistence and Dietary Trends d uring the Maya Preclassic. were solely a food source then they would have provided a re liable source of meat throughout Maya prehistory as the only domesticated mammal. The general lack of dog found in Classic deposits indicates they were not a preferred meat source. The high frequency of dog, especially during the terminal Late Preclassic may be indicative of their use in ritual feasting as a social mechanism to legitimate and solidify power (Shaw 1999:95). The latter part of the Late Preclassic would have represented a time of changing political fortunes for many with the transition f rom a kin based ideological system to one of kingship. Interestingly, dog remains are not a common find in sites (e.g. Cahal Pech and Pacbitun) within the greater Belize River Valley area (Boileau 2013; Stanchly and Awe 2015). At Caracol, the majority of dog remains are worked teeth found within Late Preclassic burials (Giddens Teeter 2001). High frequencies of freshwater shell continue to be documented in Late Preclassic midden contexts. Marine shell is found in Late Preclassic contexts such as burials and caches at numerous inland sites, including Tikal (MoholyNagy 1989) and Spondylus makes its d ebut as an elite prestige item. In summary, Late Preclassic faunal assemblages indicate an increasing reliance on mammals with dog and deer being the dominant species. Dog was likely reared for their use in feasts related to the emergence of [Late Preclassic] power hierarchies (Masson 2004b:397). Freshwater sh ell continues to play an important dietary role although they are less common than in the preceeding Middle Preclassic. We begin to see some of the faunal patterns most commonly attributed to the Classic period. These include the presence of shell and bone artifacts as burial and cache inclusions, with the notable appearance of Spondylus as a prestige item.as burial and cache inclusions, with the notable appearance of Spondylus as a prestige item. Discussion The preceding review of Belize contribution to our understanding of Preclassic Lowland Maya animal use indicates the importance of animal resources of food sources, as raw material for artifact production, as agents in the negotiation of emergent politica l hierarchies, and as important objects of ritual. Several patterns are apparent, some are continuations of previous patterns of Archaic period exploitation of resources, many are precursors of emergent Classic period patterns, and finally, others appear t o be distinctly Preclassic in nature. The Preclassic Maya exploited a broad and diverse array of local and distant fauna for basic nutritional requirements and as raw material for artifact production. This diversity, a pattern that persists throughout t he Classic and Postclassic periods, is testament to the intimate familiarity the Maya had with local environments. Throughout much of the Preclassic the majority of site specific faunal assemblages reflect a reliance on hunting of Early Preclassic 20001000 BC Middle Preclassic 1000400 BC Late Preclassic 400 BC AD 250 maize agriculture huntingforaging focus on riparian resources? faunal resources mirror Archaic patterns of exploitation? most of the above inferred more data needed intensified maize agriculture focus on riparian fauna but increase in fauna associated with secondary growth and maize fields; dogs important in some regions long distance trade & procurement of marine fauna freshwater shellfish important resource feasting important in establishing socio political alliances intensive agriculture systems maize is staple crop differential access to faunal resources appears increased focus on large game species feasting important in establishing socio political alliances less focus on freshwater shell

PAGE 98

Stanchly and Burke 99 Table 2 Summary of Preclassic Faunal Use of Select Belize Sites. Site Comments Reference Blackman Eddy Middle Preclassic focus on procurement of small terrestrial and riverine taxa with relative increase in abundance of larger game; large amounts of freshwater shell in Middle and Late Preclassic contexts; presence of marine shell beads during Middle Preclassic; possible evidence of feasting during Middle Preclassic Garber et al. 2004 Blue Creek large amounts of freshwater shell in Late Preclassic contexts; Middle and Late Preclassic marine shell beads Stanchly 1999b Cahal Pech Presence of freshwater shells, dog, peccary and deer in Early Preclassic Cunil phase; Middle Preclassic focus on procurement of small terrestrial and riverine taxa with relative increase in abundance of larger game; presence of marine fish and marine shell by Middle Preclassic; large amounts of freshwater shell in Middle and Late Preclassic contexts; presen ce of marine shell bead manufacturing during Middle Preclassic; dog not common Awe 1992; Powis et al. 1999; Stanchly 1995; Stanchly and Awe 2015 Caracol large amounts of freshwater shell in Middle and Late Preclassic contexts; appearance of marine shell and fish by Late Preclassic; modified dog remains in Late Preclassic burials Cobos 1994; Giddens Teeter 2001 Cerros increasing importance of large mammals to diet by Late Preclassic; high relative abundance of fish reflects coastal location; Late Preclassi c dog remains interpreted as supplemental food source Carr 1985 Colha Middle Preclassic focus on procurement of small terrestrial and riverine taxa with relative increase in abundance of larger game; dog common in Middle Preclassic; use of dogs in feastin g during Late Preclassic; differential access to faunal resources by Late Preclassic Shaw 1991, 1999 Cuello dog common in Middle Preclassic; use of dogs in feasting during Late Preclassic; increase in relative abundance of large mammals by Late Preclassic Carr and Fradkin 2008; Clutton Brock and Hammond 1994; Wing and Scudder 1991 Kakabish large amounts of freshwater shell in Middle and Late Preclassic contexts; presence of marine shell beads during Middle Preclassic; appearance of worked Spondylus by La te Preclassic Stanchly 2013 Kaxob Middle Preclassic focus on procurement of small terrestrial and riverine taxa with relative increase in abundance of larger game; dog remains common by end of Middle Preclassic; large amounts of apple snails in Middle an d Late Preclassic Harrington 2004; Masson 2004a, 2004b Pacbitun Middle Preclassic focus on procurement of small terrestrial and riverine taxa with relative increase in abundance of larger game; extremely large amounts of freshwater shell (i.e. jute ) in Mi ddle and Late Preclassic contexts; presence of marine shell bead manufacturing during Middle Preclassic; appearance of worked Spondylus by Late Preclassic Boileau 2013; Healy et al. 2004; Stanchly 1999a

PAGE 99

Preclassic Animal Resource Use what we assume to be locally availab le terrestrial and riverine food sources. The procurement of local species was supplemental to an increasing reliance on maize agriculture. The Preclassic Maya made use of freshwater fish, turtles, and molluscs, as well as small to large size mammals, su ch as armadillo, paca and agouti, peccary and deer. There is a trend toward the greater importance of larger sized mammals, such as deer and peccary, by the Late Preclassic. This increase in mammal appears to be at the expense of some of the turtle specie s. Freshwater fish appear to be a relatively consistent occurrence throughout the period. The use of domestic dog increases throughout the Middle and Late Preclassic periods but becomes relatively rare in later Classic period assemblages, suggesting to s ome researchers (e.g. Shaw 1999), that their use as a food source reflects their importance in ritual feasting rather than a dietary staple. Beginning in the Middle Preclassic, marine resources appear in both midden and ritual contexts. Marine shell bead production is evident at a number of coastal and inland sites, indicating the possible beginnings of craft specialization in the production of shell artifacts. Marine fish are also found at distant inland sites as food refuse. These may have been special ty food items (Wing and Scudder 1991). The presence of both cranial and post cranial elements at Cahal Pech indicates that the fish were transported whole some 110 kilometres from the Caribbean, suggesting that they were preserved in some manner, perhaps by salting or smoking the fish (Powis et al. 1999). The presence of marine fish and shellfish indicates that the Middle Preclassic Maya had developed extensive trade and exch ange networks at an early date. During the Middle Preclassic we also see the appe arance of vast quantities of freshwater snails and bivalves in construction core and within midden contexts. This appears to have been a ubiquitous feature of Lowland Maya sites lasting until the Late Preclassic. As with domestic dog, the use of freshwat er shellfish and invertebrates diminishes greatly during the Classic only to make a resurgence in the Postclassic. The presence of large quantities of freshwater shellfish, especially jute, often numbering in the hundreds of thousands, implies their use a s a food source although nondietary explanations are worthy of future study. With the onset of the Late Preclassic we see an increasing focus on larger game, such as deer. This appears to be especially true in deposits interpreted as representing elite c ontexts (e.g. at Cuello). Both quantity and quality increase. This is likely the beginning of a pattern noted in the Classic period marking the differential access to large game by the elite (Pohl 1976). There is also a concomitant increase in the amount of ritually deposited fauna in burial and cache contexts, and, in particular, the presence of greater amounts of marine shell, both worked and unworked. It is not until the Late Preclassic that we see the first evidence of the use of Spondylus in such deposits. During the Classic period, the thorny oyster became a common occurrence in the burial and tombs of Maya kings and their noble ranks (Moholy Nagy 1984). The Late Preclassic likely represents the adoption of Spondylus as a prestige item. The use of dogs in feasting events in conjunction with the appearance of differential access to faunal resources as food and prestige items reflects the emergence of increasingly stratified society and kingship. Excavations at the sites discussed, highlight the significant contributions made by research in Belize to our understanding of Maya Preclassic faunal utilization. Concluding Remarks This paper has addressed key questions pertaining to the use of faunal resources by the Preclassic Lowland Maya and the contributions made to such discussions by zooarchaeological research in Belize. A number of patterns have been discerned in the use of fauna as subsistence items during the Preclassic. Although there are some basic regional differen ces in the procurement of fauna for food, such as the greater frequency of marine resources at coastal sites such as Cerros, there is a great deal of uniformity in terms of the diversity of resources noted. The limited data we have on hand does appear to i ndicate that at least by the Late Preclassic and probably as early as the Late Middle Preclassic there was differential access

PAGE 100

Stanchly and Burke 101 to animal resources based on rank or status; however, more data are needed. In particular, we need to excavate representative samples of elite and nonelite deposits. In a similar vein to Classic Maya archaeology, much of our knowledge of Late Preclassic centres has focused on the excavation of elite contexts to study the emergence of institutionalized kingship. Zooarchaeological d ata make it clear that by the Middle Preclassic the Lowland Maya were engaged in interregional long distance trade and exchange. The exact nature of this remains unclear and additional investigation is required. The presence of marine fish at sites such as Cahal Pech, Caracol and Pacbitun may indicate that there were coastal sites focused on the procurement and redi stribution of marine resources. References Adams, Richard E.W. (Ed.) 1977 The Origins of Maya Civilization University of New Mexico Press, Albuquerque. Awe, Jaime J. 1992 Dawn in the Land Between the Rivers: Formative Occupation at Cahal Pech, Belize and Its Implications for Preclassic Development in the Maya Lowlands Unpublished Ph.D. Dissertation, Institute of Archaeology, University of L ondon. Boileau, Arianne 2013 Maya Exploitation o f Animal Resources During The Middle Preclassic Period: An Archeozoological Analysis From Pacbitun, Belize Unpublished Masters Thesis, Department of Anthropology, Trent University. Peterborough, Ontario. Carr, Helen S. 1985 Subsistence and Ceremony: Faunal Utilization in a Late Preclassic Community at Cerros, Belize. In Prehistoric Lowland Maya Environment and Subsistence Economy edited by M.D. Pohl, pp. 115 -132. Papers of the Peabody Museum of Archaeolog y and Ethnology, Vol. 77. Harvard University, Cambridge. Carr, Helen S., and Arlene Fradkin 2008 Animal resource use in ecological and economic context at Formative Period Cuello,Belize. Quaternary International 191:144153. Clark, John E., and Michael Blake 1994 The Power of Prestige: Competitive Generosity and the Emergence of Rank Societies in Lowland Mesoamerica. In Factional Competition and Political Development in the New World edited by E.M. Brumfiel and J.W. Fox, pp. 17 30. Cambridge University Press, Cambridge. Clutton -Brock, Juliet., and Norman Hammond 1994 Hot Dogs: Comestible Canids in Preclassic Maya Culture at Cuello, Belize. Journal of Archaeological Science 21:819-826. Cobos, Rafael 1994 Preliminary Report on the Archaeological Mollusca and Shell Ornaments of Caracol, Belize. In Studies in the Archaeology of Caracol, Belize edited by D.Z. Chase and A.F. Chase, pp. 139147. Pre -Columbian Art Research Institute Monograph 7, San Francisco. Dillon, Brian D. 1988 Meatless Maya? Ethnoarchaeo logical Implications for Ancient Subsistence. Journal of New World Archaeology 7:5970. Emery, Kitty F. 2002 The Noble Beast: Status and Differential Access to Animals in the Maya World. World Archaeology 34: 498 515. 2008 Techniques of ancient Maya bone working: evidence from a classic Maya deposit. Latin American Antiquity 19(2): 204221. Flannery, Kent V. 1969 Vertebrate Fauna and Hunting Patterns. In The Prehistory of the Tehuacan Valley Vol. 1., edited by D.S. Byers, pp. 132 177. University of Texa s Press, Austin. Freidel, David A. 1992 The Trees of Life: Ahau as Idea and Artifact in Classic Lowland Maya Civilization. In Ideology and Pre -Columbian Civilizations edited by A. Demarest and G.W. Conrad, pp. 115 -133. School of American Research Press, Sante Fe. Freidel, David A. and Linda Schele 1988 Kingship in the Late Preclassic Maya Lowlands: The Instruments and Places of Ritual Power. American Anthropologist 90:547-567. Garber, James F. (Ed.) 2004 The Ancient Maya of the Belize Valley: Half a Century of Archaeological Research. University of Florida Press, Gainesville. Garber, James F., M. Kathryn Brown, Jaime J. Awe, and Christopher J. Hartman 2004 Middle Formative Prehistory of the Central Belize Valley: An Examination of Architecture,

PAGE 101

Preclassic Animal Resource Use Mater ial Culture, and Sociopolitical Change at Blackman Eddy. In The Ancient Maya of the Belize Valley: Half a Century of Archaeological Research, edited by J.F. Garber, pp. 2547. University of Florida Press, Gainesville. Giddens Teeter, Wendy 2001 Maya Animal Utilization in a Growing City: Vertebrate Exploitation at Caracol, Belize Unpublished PhD Dissertation, Department of Anthropology, University of California, Los Angeles. Gifford, James C. 1976 Prehistoric Pottery Analysis and the Ceramics of Barton Ramie in the Belize Valley Memoirs of the Peabody Museum of Archaeology and Ethnology, Vol. 18. Harvard University, Cambridge. Halperin, Christina T., Sergio Garza, Keit h M. Prufer, and James E. Brady 2003 Caves and Ancient Maya Ritual Use of Jute. Latin A merican Antiquity 14(2): 207220. Hammond, Norman (Ed.) 1991 Cuello: An Early Maya Community in Belize Cambridge University Press, Cambridge. Hammond, Norman 1986 New Light on the Most Ancient Maya. Man 21:399-413. Hansen, Richard D. 1989 Archaeological Investigations at Nakbe, Peten, Guatemala: 1989 Field Season. Institute of Archaeology, UCLA. 1990 Excavations in the Tigre Complex, El Mirador, Peten, Guatemala. Papers of the New World Archaeological Foundation 62. New World Archaeological Foundation, Provo. Harrington, Ryan 2004 Mollusca of Kaxob: For Supper and Soul. In Kaxob: Ritual, Work, and Family in an Ancient Maya Village edited by P.A. McAnany, pp. 399412. Monumenta Archaeologica 22. Cotsen Institute of Archaeology, University of California, Los Angeles. Healy, Paul F. 1998 Preclassic Maya of the Belize Valley. Paper presented at the XIV International Congress of Anthropological and Ethnological Sciences, College of William and Mary, Williamsburg, Virginia. Healy, Paul F., an d Jaime J. Awe (Eds.) 1995 Belize Valley Preclassic Maya Project: Report on the 1994 Field Season. Occasional Papers in Anthropology No. 10. Trent University, Peterborough. Healy, Paul F., Kitty F. Emery, and Lori E. Wright 1990 Ancient and Modern Maya Ex ploitation of the Jute Snail ( Pachychilus ). Latin American Antiquity 1(2): 170-183. Healy, Paul F., David Cheetham, Terry G. Powis, and Jaime J. Awe 2004 Cahal Pech: The Middle Formative Period. In The Ancient Maya of the Belize Valley: Half a Century of Archaeological Research edited by J.F. Garber, pp. 103-124. University of Florida Press, Gainesville. Healy, Paul F., Bobbi Hohmann, and Terry G. Powis 2004 The Ancient Maya Center of Pacbitun. In The Ancient Maya of the Belize Valley: Half a Century of Archaeological Research edited by J.F. Garber, pp. 207-227. University of Florida Press, Gainesville. McAnany, Patricia A. (Ed.) 2004 Kaxob: Ritual, Work, and Family in an Ancient Maya Village Monumenta Archaeologica 22. Cotsen Institute of Archaeology, University of California, Los Angeles. McAnany, Patricia A. 2004 Situating Kaxob Within Formative Period Lowland Maya Archaeology. In Kaxob: Ritual, Work, and Family in an Ancient Maya Village edited by P.A. McAnany, pp. 1-9. Monumenta Archaeologica 22 Cotsen Institute of Archaeology, University of California, Los Angeles. McSwain, Richard, Jay K. Johnson, Laura J. Kosakowsky, and Norman Hammond 1991 Craft Technology and Production. In Cuello: An Early Maya Community in Belize edited by N. Hammond, p p. 159191. Cambridge University Press, Cambridge. Marcus, Joyce 1993 Ancient Maya Political Organization. In Lowland Maya Civilization in the Eighth Century A.D., edited by J.A. Sabloff and J.S. Henderson, pp. 111 -184. Dumbarton Oaks,Washington. Masson, Marilyn A. 2004a Faunal Exploitation from the Preclassic to Postclassic Periods at Four Maya Settlements in Northern Belize. In Maya Zooarchaeology: New Directions in Method and Theory edited by K.F. Emery, pp. 97 122. Cotsen Institute of Archaeology Mon ograph 51. University of California, Los Angeles. 2004b Contribution of Hunting and Fishing to Subsistence and Symbolic Expression. In Kaxob: Ritual, Work, and Family in an Ancient Maya

PAGE 102

Stanchly and Burke 103 Village edited by P.A. McAnany, pp. 383-397. Monumenta Archaeologica 22. Cotsen Institute of Archaeology, University of California, Los Angeles. Micheletti, George J. and Terry G. Powis 2015 Origins of the Block Party: Investigations of Preclassic Architecture Over and Under Plaza A at Pacbitun, Belize. In Research Report s in Belizean Archaeology Vol. 12, Papers of the 2014 Belize Archaeology Symposium edited by John Morris, Melissa Badillo, Sylvia Batty and George Thompson, pp. 201215. National Institute of Culture and History, Belmopan, Belize. Moholy -Nagy, Hattula 1 978 The Utilization of Pomacea Snails at Tikal, Guatemala. American Antiquity 43:65 -73. 1989 Formed Shell Beads from Tikal, Guatemala. In Proceedings of the 1986 Shell Bead Conference edited by C.F. Hayes III, pp. 139155. Rochester Museum and Science Center, Rochester. Orsini, Stephanie R. 2016 From Turkeys to Tamales: Paleoindian to Preclassic Period Faunal Use at Maya Hak Cab Pek Rockshelter in Southern Belize. Unpublished MA Thesis, Department of Sociology and Anthropology, University of Mississippi. Oxford, MS. Pendergast, David M. 1971 Excavations at Eduardo Quiroz Cave, British Honduras (Belize), Royal Ontario Museum Art and Archaeology Occasional Papers No. 21 Toronto, ON. 1974 Excavations at Actun Polbilche, Belize, Archaeology Monograph #1. Royal Ontario Museum, Toronto, ON. 1979 Excavations at Altun Ha, Belize 1964-1970, Vol. 1 Royal Ontario Museum, Toronto. 1981 Lamanai, Belize: Summary of Excavation Results, 19741980. Journal of Field Archaeology 8:29 -53. Pohl, Mary D. 1976 Ethnozoolo gy of the Maya: An Analysis of Fauna from Five Sites in Peten, Guatemala. Unpublished PhD dissertation, Harvard University, Cambridge. 1983 Maya ritual Faunas: Vertebrate remains from burials, caches, caves, and cenotes in the Maya Lowlands. In Civilizati on in the ancient Americas edited by R. Leventhal and A. Kolate, pp. 55 103. Albuquerque: University of New Mexico Press. 1985 The privileges of Maya elites: Prehistoric vertebrate fauna from Seibal. In Prehistoric Lowland Maya Environment and Subsistenc e Economy: Papers of the Peabody Museum of Archaeology and Ethnology Vol. 77, edited by M.D. Pohl, pp. 133145. Harvard University, Cambridge, MA. Powis, Terry G., Norbert Stanchly, Christine D. White, Paul F. Healy, Jaime J. Awe, and Fred Longstaffe 1999 A Reconstruction of Middle Preclassic Maya Subsistence Economy at Cahal Pech, Belize. Antiquity 73:364-376. Sharer, Robert J. 1992 The Preclassic Origin of Lowland Maya States. In New Theories on the Ancient Maya edited by E.C. Danien and R.J. Sharer, p p. 131-136. University of Pennsylvania, University Museum Monograph 77, Philadelphia. Shaw, Leslie C. 1991 The Articulation of Social Inequality and Faunal Resource Use in the Preclassic Community of Colha, Northern Belize. Unpublished PhD Dissertation. University of Massachusetts, Amherst. 1999 Social and Ecological Aspects of Preclassic Maya Meat Consumption at Colha, Belize. In Reconstructing Ancient Maya Diet edited by C.D. White, pp. 83100. University of Utah Press, Salt Lake City. Solis, Wendy 2 010 Ancient Maya Exploitation of Jute (Pachychilus Spp.) at Minanha, West Central Belize Unpublished M.A. Thesis, Department of Anthropology, Trent University. Peterborough, ON. Smith, Robert E. 1955 Ceramic Sequence at Uaxactun, Guatemala. Middle Americ an Research Institute Publication 20. Tulane University, New Orleans. Stanchly, Norbert 1995 Formative Period Maya Faunal Utilization at Cahal Pech, Belize: Preliminary Analysis of the Animal Remains from the 1994 Field Season. In Belize Valley Preclassic Maya Project: Report on the 1994 Field Season, edited by P.F. Healy and J.J. Awe, pp. 124-149. Occasional Papers in Anthropology No. 10. Department of Anthropology, Trent University, Peterborough. 1999a Preliminary Report on the Preclassic Faunal Remains from Pacbitun, Belize: 1995 and 1996 Field Seasons. In Belize Valley Preclassic Maya: Report on the 1996 and 1997 Field Seasons edited by P.F. Healy, pp. 41 -52. Occasional Papers in Anthropology No. 13. Trent University, Peterborough.

PAGE 103

Preclassic Animal Resource Use 1999b An Analysis of the Faunal Remains: The 1996 and 1997 Field Seasons. In The Blue Creek Project: Working Papers from the 1997 Season, pp. 115 131, edited by W. David Driver, Helen R. Haines, and Thomas H. Guderjan. Maya Research Program, St. Marys University, San Anton io, Texas. 2013 A Preliminary Analysis Of The Kakabish Faunal Remains. In Kakabish Archaeological Research Project (KARP) Interim Report On The 2012 Field Season edited by Cara G. Tremain and Helen R. Haines, pp. 19 24. Department of Anthropology, Trent University. Peterborough, Ontario. Stanchly, Norbert and Jaime J. Awe 2015 Ancient Maya Use of Dog ( Canis lupus familiaris ): Evidence from the Upper Belize River Valley. In Research Reports in Belizean Archaeology Vol. 12, Papers of the 2014 Belize Ar chaeology Symposium edited by John Morris, Melissa Badillo, Sylvia Batty and George Thompson, pp. 227237. National Institute of Culture and History, Belmopan, Belize. Stanchly, Norbert, Wendy Solis, and Gyles Iannone 2008 Zooarchaeological Investigation s at Minanha, Belize: Research Goals, Methods and Preliminary Results. In Archaeological Investigations in the North Vaca Plateau, Belize: Progress Report of the Tenth (2008) Field Season edited by G. Iannone and S. Macrae, pp. 112126. Department of Anth ropology, Trent University, Peterborough. Wake, Timothy A. 2004 A Vertebrate Archaeofauna from the Early Formative Period Site of Paso de la Amada, Chiapas, Mexico. In Maya Zooarchaeology: New Directions in Method and Theory edited by K.F. Emery, pp. 209 -222. Cotsen Institute of Archaeology Monograph 51. University of California, Los Angeles. Wagner, Logan, H. Box, and Susan K. Morehead 2013 Ancient Origins of the Mexican Plaza: From Primordial Sea to Public Space Univers ity of Texas Press, Austin. Walden, John and Michael Biggie 2017 Settlement Excavations at Tutu Uitz Na and Pech Na in the Lower Dover Hinterland: Results of the 2016 Field Season. In The Belize Valley Archaeological Reconnaissance Project: A Report of th e 2016 Field Season, edited by Claire E. Ebert, Chrissina C. Burke, Jaime J. Awe, and Julie A. Hoggarth, Volume 22, pp. 238 -288. Institute of Archaeology, Baylor University, Waco, TX; Department of Anthropology, Northern Arizona University, Flagstaff, AZ. Wing, Elizabeth S. and Sylvia J. Scudder 1991 The Exploitation of Animals. In Cuello: An Early Maya Community in Belize edited by N. Hammond, pp. 8497. Cambridge University Press, Cambridge.

PAGE 104

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 105 113 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 9 RITUAL USE OF ANIMAL S IN ANCIENT MAYA MO RTUARY CONTEXTS: RESULTS OF FAUNAL ANALYSIS FROM THE ST RUCTURE A9 TOMB AT XUNANTUNICH Chrissina C. Burke, Katie K. Tappan, Gavin B. Wisner, and Jaime J. Awe During the 2016 field season excavations at Xunantunich, Belize Valley Archaeological Reconnaissance (BVAR) Project archaeologists discovered the tomb of an elite male individual within Structure A9. Besides ceramic vessels, jade beads, and obsidian blades, we recovered an abundance of faunal remai ns in the burial chamber. The faunal materials associated with the burial include large cat claw bones, a locus of animal limb bones, and worked shell pieces. This paper presents a detailed discussion of the zooarchaeological assemblage from the tomb, as well as the methods used, and results identified from the faunal analysis. Additionally, the importance of animals to ancient Maya ritual mortuary practices is addressed, with specific focus on large cats. This research discusses the interplay between d eer and large cats to ancient Maya elite and commoners and reveals a possible interpretation for the identi ty of this individual based on the faunal remains present. Finally, we demonstrate how faunal analysis can provide context for understanding elite burials in the past. Introduction Zooarchaeological data analysis contextualizes many aspects of archaeological research. From contributing an understanding of diet, and bone tool or personal adornment manufacture, to the identification of species employed in ritual contexts these analyses provide a holistic understanding of human and animal connections in the past. In the summer of 2016, the Belize Valley Archaeological Reconnaissance (BVAR) Project excavated structure A9 on the west side of Plaza A at the site of Xunantunich. Excavations at the base of A9 revealed a set of hieroglyphic panels and two eccentric caches. Ex cavations in the structure itself uncovered a large vaulted tomb. With a maximum length of 444 cm, an average floor length of 353 cm and width of 214 cm this is one of the largest burial chambers excavated in the regio n to date (Tilden et al. 2016). Withi n the tomb, we recovered an anatomically athletic, adult male individual along with 36 ceramic vessels, six jade beads, 13 obsidian blades, two bone hairpins, one shell ring, and an abundance of vertebrate faunal remains (Tilden et al. 2016). The faunal m aterials were concentrated in two locales, those directly associated with the burial and a locus, or pile, of longbone faunal elements near the feet of the individual. The locus of animal long bones was approximately 30 cm to the northwest of the feet o f the individual and contained four chert cores generally placed in the four cardinal directions ( Figure 1 ). Faunal materials recovered from the tomb include multiple taxa, elements, and individuals. This faunal research highlights the importance of inclu ding animals in mortuary analysis, and more specifically, the incorporation of an understanding of taphonomic processes that influence the preservation and, therefore, the researchers ability to interpret human behaviors. Below, questions asked, methods employed, results identified, and discussion of the importance of species, bone elements, and bone sides present within the tomb to ancient Maya ideology, are undertaken. Finally, we emphasize how zooarchaeological analysis contributes to understanding e lite individuals significance wit hin burial and ritual contexts. The individual interred within the A9 tomb is of great importance to our understanding of the ancient polity of Xunantunich. Given the limited number of burials recovered in the site core t o date, and this being only the second elite burial recovered in the site core, any information contextualizing the importance of this discovery, including the fauna associated wit h the individual, is essential. Zooarchaeology Research and Questions Zooa rchaeologists focus on human animal interactions in the past via multiple lines of inquiry. First, as expected, animals are studied as a part of human diets such as questions concerning what animals people were eating; second, skeletal and shell remains can be used to assist with the reconstruction of ancient

PAGE 105

Ritual Use of Animals in Ancient Maya Mortuary Contexts 106 Figure 1 Map of Structure A9 tomb drawn by Merle Alfaro, images of tomb and fauna long bone locus. environments specific animals may be found in certain seasons or different years and climates; third, evaluating the modification of animal and shell materials into personal adornments and tools; finally, and more importantly to this research, human cultures, such as the ancient Maya, often employed fauna for ritual purposes. Therefore, these materials can represent significant symbols in everyday life or leadership, as demonstrated through ethnography and iconography. To understand the significance of the individual buried within the tomb, several questions were identified to guide research. First, as with all faunal analyses, understanding the basic demographics of the faunal material recovered is a necessary step in the process. What skeletal elements (bones), si des, portions, and taxa are present within the tomb? After identifying what is present in the faunal assemblage, determining preservation biases is next. What is known taphonomically about the fauna associated directly with the burial and those cached ne ar the feet? Are there differences in preservation and did diagenesis conceal any cultural modifications? Third, given the ritualistic nature of faunal materials associated with human burials, do the associated taxa reflect the status of the individual b uried within the tomb? Finally, as many zooarchaeologists around the world have noted, right or left sides may be important culturally to the selection of animal remains as mortuary goods, therefore, what do the elements, sides, and taxa present suggest about this individual within the context of Maya ideology at the site? Materials and Methods The burial and associated artifacts, including the faunal remains, were excavated,

PAGE 106

Burke et al. 107 mapped and removed by BVAR personnel in 2016. Most of the faunal remains were s ubsequently dry brushed in Belize and then exported to the Northern Arizona University, Department of Anthropology, Faunal Analysis Laboratory (NAUDAFAL) for further cleaning and analysis by Burke, Tappan, and Wisner. The materials used for the analysis in clude comparative faunal collections from the Stanley J. Olsen Laboratory of Zooarchaeology at the Arizona State Museum, the Charles L. Douglas Vertebrate Zoology Collection at the Museum of Northern Arizona, and the Northern Arizona University, Department of Anthropology, Faunal Analysis Laboratory. Identifications were also accomplished using osteological guides, although our taxonomic classifications remained conservative, as not to inflate results (Andrews 1969; Gilbert 1985 and 1990; McKusick 2001; Ol sen 1964, 1968, 1979, and 1982). Data collected include in field contextual information, Phylum, Class, Order, more specific taxonomic category (if possible), body portion (such as exoskeleton, cranial, appendicular, axial), element, portion of said element, side, age (generally, subadult or adult), sex, size class (small, medium, or large), and modifications either cu ltural or taphonomic in nature. All faunal material was quantified to NISP, number of identified specimens, where identified as to skeletal element or taxon/sizeclass (size class identified following Pendergast 1971:78). Also recorded was the minimum number of elements (MNE), to specific element and side to assist with identifying our MNI values or minimum number of individuals and to determ ine if right or left elements we re more prevalent (Lyman 1994). Finally, any bone and shell artifacts were analyzed for modifications into tools or other personal adornments. Analysis of these materials allowed for a more specific identification and prese ntation of details tied to the individual interred in the tomb to furthe r contextualize the importance of the individual Results As we address the results of this analysis, attention should first be called to the taphonomic context within which the analy sis was made. Taphonomy refers to natural or cultural processes, like animals chewing on bones or an eart hquake fracturing architecture that lead to an incomplete and biased archaeological record (Efremov 1940). Taphonomic agents are the immediate physical cause of modifications to animal bones or artifacts and taphonomic effects are the results or traces of these agents. By studying these processes, we can remove the taphonomic overprint, and truly observe the patterns resulting fro m natural and cul tural effects. The faunal remains recovered in the tomb were eroded, exfoliated, and generally poorly preserved. This was likely caused by the collapse of the tombs vaulted roof, and due to the destructive nature of the limestone deposits within the cham ber. Additionally, much of the wall plaster had fallen from the walls over time and as it got wet and spread out on the floor, it aggregated in clumps and coated all the cultural remains on the tomb floor (Tilden et al. 2016). This hardened plaster made removal of the materials difficult, leading to fragmentation, which further obscured natural and cultural modifications, such as cut marks. In spite of their condition, we were able to spatially separate the faunal remains into two locales within the tomb, those associated directly with the human skeletal remains, and a cache of long bones near the feet of the individual in the northeast corner of the tomb. We were also able to identify carnivore gnawing marks on several of the deer longbone elements suggesting the animal remains were available to scavenging carnivores prior to caching. Within the marrow cavity of the long bone locus elements, a very dark, organic matrix, consisting of plant roots was noted this further led to the increased fragme ntation during removal. Given the extreme humidity within the tomb as well, skeletal elements often crumbled under the pressure of excavation tools. Besides the three elements displaying carnivore modification, one longbone fragment exhibited cut marks. These marks appear more like hacking or chopping marks, which are produced by striking the bone with a stone implement at a perpendicular angle, leaving a v shaped cross section, with small fragments of bone crushed inwards (Potts and Shipman

PAGE 107

Ritual Use of Animals in Ancient Maya Mortuary Contexts 108 Table 1 Faunal Remains Recovered from A9 Tomb Taxonomic Category Location within Tomb NISP % NISP for Structure Class: Mammalia Cache 1,853 82% Order: Artiodactyla Cache 46 2.03% Order: CF Artiodactyla Cache 99 4.37% Panthera onca Cache 47 2.08% Puma concolor Cache 13 0.57% Odocoileus virginianus Cache 191 8.44% Class: Mammalia Human Associated 6 0.27% Class: Gastropoda Human Associated 3 0.13% Order: Felidae Human Associated 5 0.22% Order: Rodentia Human Associated 1 0.04% Total 2,264 100% *Location within Tomb = directly associated with the human skeletal remains (human associated) or recovered from the cache of long bones in the northeast corner of the tomb (cache). **NISP = Number of Identified Specimens, where specimens are identified to skeletal element and taxon. Note Panthera onca and Puma concolor elements were severely fragmented. 1981:577). Chopping marks are common in disarticulation of body portions, suggesting these marks could be t he results of acqu iring the limbs for this locus. The longbones located near the feet of the individual are generally all oriented southeast to northwest ( Figure 1 ). A few of the elements appeared to have been moved, likely because of taphonomic factors resulting with some having an east to west orientation. The skeletal elements in this locus are also stacked upon one another given the taphonomic issues, the best estimate is that there was approximately 3 to 4 layers of these elements. Als o, surrounding the long bones were four chert cores, positioned approximately in the four cardinal directions. Our analyses identified 2,264 specimens to either a specific taxon and/or skeletal element even if element was just general long bone. Taphonomic issues in the tomb led to this higher than expected NISP value and does not represent cultural processing of the faunal remains. During excavations, while elements were becoming more and more fragmentary, an attempt was made to record the number of lo ng bones present. The number of identified specimens (NISP), where identified reflects skeletal element or body portion (appendicular, axial, cranial) and taxon, associated directly with the human skeletal remains is 15 ( Table 1 ). Of these, 12 of the ele ments are mammal and three are mollusk. Five of the elements are the third phalanx, or the claw bone, of a felid. The size of the third phalanx suggests either jaguar ( Panthera onca ) or cougar ( Puma concolor ), but identification, even with comparative ma terials from the Museum of Northern Arizona (MNA) in Flagstaff and Arizona State Museum (ASM) in Tucson, was not possible. The claw bones were found near the hands of the human burial. One of the elements associated directly with the burial is a rodent r ight dentary from the family Muridae (mice and rats), although unidentifiable to species. This rodent is likely intrusive and taphonomic, not cultural. The remaining six mammalian skeletal elements are unidentifiabl e to skeletal element or taxon. Besides the mammal remains, three shell pieces were recovered. One of these pieces is a worked shell adornment that possibly functioned as a ring ( Figure 2 ). This ring was found below the left hand of the individual. Only half of the ring remains, but it resem bles a bow tie ring recorded at Dzibilchaltun (Taschek 1994). The remaining two shell pieces appear to be shell inlays or potentially bead spacers, although the perforations are not complete and appear more like depressions. One piece was found west of the cervical vertebrae, and the second was recovered east of the mandible. These pieces ( Figure 3) are nearly identical in size, the first one has a length of 28.5 mm, a width of 7.25 mm, and breadth of 2.00 mm. The second one,

PAGE 108

Burke et al. 109 Figure 2 Shell ring recovered near hand of individual buried in the tomb. Figure 3 Shell inlay piece. The two holes appear to be taphonomically caused, while the single depression seems to be intentionally created. which is missing a very small piece, is 28.1 mm in length, 7.00 mm in width, and has a breadth of 2.00 mm. Both were modified in a similar pattern and they appear to be shell pendants with inlays. Shell artifacts with inlays are fairly common in caches and el ite burials in the Belize Valley (Thompson 1931: Plate XLVII; Willey et al. 1965: Fig. 310). At Baking Pot, for example, Audet (2006:199, fig. 5.18) recovered a similar inlaid shell pendant, except that the latter was carved. More elaborately carved and inlaid shell discs are also reported at several cave sites, such as A ctun Neko (Morton et al. 2012). Finally, two bone hairpins were recovered from under the east side of the human remains. Again, due to the poor preservation, these materials were very fr agmentary and recovered in 17 pieces. Two, pointed tip fragments were recovered with two blunt ends, suggesting at least two separate hairpins. The NISP of skeletal elements cached near the feet of the burial is 2,249. Given their poor preservation; many elements did not retain anatomical features useful for taxon identification, although a few more specific identifications were possible with the comparative collections at the MNA and the ASM. Elements splintered and fractured during removal, therefore, NISP counts indicate the high degree of fragmentation. Individual complete element counts are not possible given this extreme fragmentation. Of these skeletal elements, 251 have been identified to genus and species (Table 1). Radii, femora, tibiae, metat arsals, one patella and one calcaneus were recovered from the longbone locus, although significantly more tibiae were observed. The few femora recorded were easily identified to species given the diagnostic nature of the proximal and distal ends. The tw o right femora represent one puma and one white tailed deer, while one of the two left femora w as successfully identified as jaguar. We think it is reasonable to assume that the remainder of the long bones in the locus are jaguar, puma, and white tailed d eer given the lengths and diameters of the elements. Because of the poor preservation of species specific diagnostic features, however, our ident ifications remain conservative. The minimum number of individual (MNI) for the long bone locus is four individ uals based on the number of remaining left tibiae. There is a possibility, however, that at least 10 individuals migh t be represented, but here again, their poor preservation precludes us from making an accurate count. The MNI for third phalanges or claw s bones is more problematic, given the difficulty of siding those elements to right or left paw and sp ecific phalanx within said paw. Finally, mammalian faunal remains associated directly with the hands of the burial include five unsided third phalanx or claw bones, representative of a jaguar or puma pelt, robe, or more likely mittens (Ballinger and Stomper 2000; Willey 1972). Minimum number of individuals for the third phalanges was not possible due to the fragmentary nature of the

PAGE 109

Ritual Use of Animals in Ancient Maya Mortuary Contexts 110 remains and the difficulties in identifying side, forelimb, and hind limb in these elements. Discussion Fauna within the Maya region are frequently associated with ritual activities and the supernatural. Often fauna in ritual contexts are nondietary in nature. For the tomb longbone locus, we discovered a higher number of tibiae, which, given the lack of significant meat utility, supports a ritual context for the locus itself. The femora associated with the large cats and deer is interesting because they are t he only femora identified within the bone pile, and gener ally, these elements would be tied to subsistence given the large muscle masses associated with the femur, but a lack of cut marks suggests these elements were included for the locus. The patella an d calcaneus recovered could indicate the longbones were fleshy when placed within the tomb, meaning the patella was a rider instead of intentionally included (Binford 1981:234). The third phalanges associated directly with the burial are common across the Maya region. Sites such as Altar de Sacrificios, Altun Ha, Calakmul, El Per Waka El Zotz, Holmul, Tikal, Piedras Negras, Uaxactun, and Yaxchilan all have claw bones associated with burials (Edmonson 1971; Pendergast 1982; Pohl 1983; Scherer 2015). The interpretation, when such elements are found, has commonly been that the interred individual was covered with a large cat pelt (Smith 1950). Another possibility is that the individual was interred wearing jaguar mitts or gloves akin to the image of t he dancing figure on the Altar de Sacrificios vase (Montgomery 2000) Symbolically, left sided ideology is tied to the heart and life giving powers and the underworld hind limb elements are offered to the gods and used to make tools (Schlesinger 2001:178183). Left tends to indicate ritual activities in elite ceremonial contexts according to Pohl (1976; 1983), while right can be correlated to male and power ideologies according to Brown (2004). It has also been suggested that male activities are often f ound on the right side of buildings, lending more credence to the argument that with better preservation we would have observed more right sided elements in the tomb (Inomata 2001). Several of the elements were confidently identified to jaguar, puma, and w hite tailed deer. From an ethnozoological perspective these species are related and represent specific ideological concepts for the Maya. First, jaguars were associated with the night, caves, underworld, hunting, and stealth activities, which could be one reason for them showing up in the bone pile locus since similar artifact loci from tombs are often very similar to those found in caves and cenotes (Miller and Taube 1993:103104; Pohl 1983:56). Second, the large cats are associated with validating an elite individuals status as a ruler (Teeter 2004:188). Since jaguars and other large cats symbolize masculinity, courage, strength, power, and destructiveness the animal itself is thought to protect Maya rulers the wayob or animal spirit protectors are often represented in tombs through jaguar claws, bones, and teeth to benefit and demonstrate the rulers rank in society (Anderson and Medina Tzuc 2005:129; Pohl 1983:73; Schlesinger 2001:163166). Given the jaguar occupies the top of the food chain th e Maya connected the animal to humans also considered the top of the food chain therefore, identifying themselves as the big cat (Miller and Taube 1993:102). Since jaguars and humans occupied the same power realm, the two also shared control over the rainforest (Miller and Taube 1993:103104). So, the ancient Maya believed their dominance in the world was reflected in the predator and prey relationship between the jaguar and deer suggesting the elite were the jaguars and the commoners or rivals were the deer (Pohl 1994:132). After major victories, the Maya elite would feast by consuming deer, metaphorically demonstrating their power to eat their defeated rivals, victims, or subordinates (Pohl 1994:132). Since the rulers were of the highest importan ce they would wear jaguar pelts at these feasts to symbolize their connection to the jaguar (Miller and Taube 1993:102). At Copn for example, 16 jaguars were sacrificed to celebrate the installation of the 16th ruler, with one whole jaguar being buried beneath a mound in the Great Plaza (Miller and Taube 1993; Pohl 1983). At Altun Ha, at least three pelts were present in

PAGE 110

Burke et al. 111 the Sun Gods tomb based on the number of third phalanges rec overed (Miller and Taube 1993). Jaguars depicted with deer were frequently tied to ruling or power and control; the jaguar deer dance demonstrates this constant struggle for power between the elite and the commoners (Pohl 1983:71). In fact, the mere presence of a jaguar in any context may validate the power and rights of an individual to rule (Pohl 1983:73). Coincidently, the word for jaguar, balam was often tied to government officials and native priests in 16th century Yucatn, and has often meant sorcerer in Central Mexico (Benson 1998:69; Pohl 1983:73; Saunders 1994:109110). Larger cats can also be tied to religious significance, specifically with medicinal and supernatural properties in their meat, blood, t eeth, an d claws (Emery and Healy 2014). Finally, deer were considered supernatural as well and often associated with the sun (Pohl 1983:62). Deer are also commonly found in elite burials, as they were the food of the rulers and not the food of the commone rs. Conclusion Based on the faunal remains recovered from the tomb, specifically the locus of jaguar, puma, and deer long bones as well as large cat third phalanges, we suggest the fauna represent the jaguar deer dichotomy in Maya ideology. With the elit e represented by the large cats and the rivals or commoners represented by the deer we propose the individual buried in this tomb is that of a significant leader. More specifically, the osteological evidence for the individual interred suggests a middle aged male, with an athletic build indicating he may have played a physically active role. Additionally, the faunal remains have been AMS radiocarbon dated, producing a date range of AD 685 to 890, placing him within the context of the Naranjo and Car acol conflict (Awe et al. this volume). Given the lack of other animal taxa as mortuary goods this individuals fauna suggests that in war, he was likely a skilled leader, potentially even an elite warrior he was not buried with any other faunal remai ns besides the adornments he was wearing. Instead, all the large cat and deer elements argue elite leadership status, especially when coupled with the discovery of the hieroglyphic panels flanking the staircase of the structure within which the tomb was f ound. The discovery of this tomb and the analysis of the fauna within, provide context for the role of Xunantunich in the regional political landscape in western Belize during the Late Classic. Continued analysis and research of faunal materials associat ed with elite burials can provide a more holistic understanding of individual identity and tie this ident ity back to regional politics. Acknowledgements We would very much like to thank Doug Tilden, Diane Slocum, Kelsey Sullivan and Hannah Zanotto for their excavation and analysis of the tomb. We are grateful to Rafael Guerra and Julie Hoggarth for assisting with the export of the faunal materials for our research analysis. We also thank Jorge Can, Arlen Chase, Shawn Morton, and Norbert Stanchly for their insightful comments. Finally, we would like to thank the Tilden Family Foundation for their continued support of the Xunantunich Archaeology and Conservation Project, and the Institute of Archaeology for providing an opportunity to submi t our wo rk to a wider audience. References Anderson, E. N. and Felix M. Tzuc 2005 Animals and the Maya in Southeast Mexico. The Unive rsity of Arizona Press, Tucson. Andrews, E. Wyllys IV 1969 The archaeological use distribution of Mollusca in the Maya lowlands. Middle American Research Institute Series No. 34. Middle American Research Institute, Tulane University of Louisiana, New Orleans. Audet, Carolyn M. 2006 Political Organization in the B elize Valley: Excavations at Baking Pot, Cahal Pech, and Xunantunich. PhD Dissertation, Department of Anthropology, Vanderbilt University, Nashville, Tennessee. Awe, Jaime J., Christophe Helmke, Doug Tilden, and Diane Slocum 2018 Lets Talk of Graves, Eccentrics and Epitaphs: The Socio Political Implications of Recent (2016) Discoveries at Xunantunich, Belize. Research Reports in Belizean Archaeology, Vol. 15.

PAGE 111

Ritual Use of Animals in Ancient Maya Mortuary Contexts 112 Ballinger, Diane A. and Jeffrey Stomper 2000 The Jaguars of Altar Q, Copan, Honduras: Faunal Anal ysis, Archaeology, and Ecology. Journal of Ethnobiology 20(2):223 -236. Benson 1998 The Lord, The Ruler: Jaguar Symbolism in the Americas. In Icons of Power: Feline Symbolism in the Americas, edited by Nicholas J. Saunders, pp. 5376. Routledge, London. Binford, Lewis R. 1981 Bones: Ancient Men and Modern Myths Academic Press, New York. Brown 2004 Dangerous Places and Wild Spaces: Creating Meaning with Materials and Space at Contemporary Maya Shrines on El Duende Mountain. Journal of Archaeological Method and Theory 11(11):31 -58. Edmonson, Munro S. 1971 The Book of Counsel: The Popol Vuh of the Quiche Maya of Guatemala. Middle American Research Institute. Publication 35. New Orleans, Tulane University. Efremov, J. A. 1940 Taphonomy: A New Branch of Paleontology. Pan -American Geologist 74(2):81. Emery, Kitty F. and Paul F. Healy 2014 A History of Animal Use at the Ancient Maya Site of Pacbitun. In Zooarchaeology of the Ancient Maya Site of Pacbitun (Belize) edited by Paul F. Healy and Kitty F. Emery, pp. 152 176. Trent University Occasional Paper, Number 16. Trent University Press, Toronto. Gilbert, B. Miles 1985 Avian Osteology Missouri Archaeological Society Inc., Columbia. 1990 Mammalian Osteology Missouri Archaeolo gical Society Inc., Columbia. Inomata, Takeshi 2001 The Power and Ideology of Artistic Creation. Current Anthropology 42(3):321349. Lyman, R. Lee 1994 Vertebrate Taphonomy Cambridge University Press, Cambridge. Miller, Mary E. and Karl A. Taube 1993 T he Gods and Symbols of the Ancient Maya: An Illustrated Dictionary of Mesoamerican Religion. Thames and Hudson, London. McKusick, Charmion R. 2001 Southwest Birds of Sacrifice. The Arizona Archaeological Society, Globe, AZ. Montgomery, John 2000 Altar de Sacrificios, Guatemala: The Altar Vase. In the Montgomery Drawing Collection, FAMSI Resources data base. Morton, Shawn, Christophe Helmke and Jaime J. Awe 2012 Investigations at Actun Neko, Caves Branch River Valley, Belize. In Heart of Earth: St udies in Maya Ritual Cave Use, edited by James E. Brady, pp. 8394. Bulletin 23, Association for Mexican Cave Studies, Austin. Olsen, Stanle y J. 1964 Mammal Remains from Archaeological Sites: No.1 Southeastern and Southwestern United States Harvard Unive rsity, Papers of the Peabody Museum of Archaeology and Ethnology, Cambridge. 1968 Fish, Amphibian and Reptile Remains from Archaeological Sites: No.2 Southeastern and Southwestern United States. Harvard University, Papers of the Peabody Museum of Archaeol ogy and Ethnology, Cambridge. 1979 Osteology for the Archaeologist: No.3 The American Mastodon and the Woolly Mammoth, No. 4 North American Birds: Skulls and Mandibles, No. 5 North American Birds: Postcranial Skeletons Harvard University, Papers of the P eabody Museum of Archaeology and Ethnology, Cambridge. 1982 An Osteology of Some Maya Mammals Harvard University, Papers of the Peabody Museum of Archaeology and Ethnology, Cambridge. Pendergast, D. M. 1971 Excavations at Eduardo Quiroz Cave, British H onduras (Belize). Royal Ontario Museum, Art and Archaeology Occasional Papers No. 21.Toronto, Ontario. 1982 Excavations at Altun Ha, Belize, 1964 1970, Volume 2. Royal Ontario Museum. Pohl, M. D. 1976 Ethnozoology of the Maya: An Analysis of Fauna from Five Sites in the Peten, Guatemala. PhD dissertation, Department of Anthropology, Harvard University 1983 Maya Ritual Faunas: Vertebrate Remains from Burials, Caches, Caves, and Cenotes in the Maya Lowl ands. In Civilization in the Ancient Americas: Essays in Honor of Gordon R. Willey edited by Richard M. Leventhal and Alan L. Kolata, pp. 55 -

PAGE 112

Burke et al. 113 103. University of New Mexico Press, Albuquerque. 1994 The Economics and Politics of Maya Meat Eating. In The Eco nomic Anthropology of the State: Monographs in Economic Anthropology, No. 11, edited by Elizabeth M. Brumfiel. pp. 119-s148. University Press of America, Inc, Lanham. Potts, Richard, and Pat Shipman 1981 Cut Marks Made by Stone Tools on Bones from Olduva i Gorge, Tanzania. Nature 291:577580. Saunders, Nicholas J. 1994 Predators of Culture: Jaguar Symbolism and Mesoamerican Elites. World Archaeology 26(1):104117. Scherer, AndrewK. 2015 Mortuary Landscapes of the Classic Maya. University of Texas Press, Austin. Schlesinger, Victoria 2001 Animals and Plants of the Ancient Maya. University of Texas Press, Austin. Smith, A. Ledyard 1950 Uaxactun, Guatemala: Excavations of 19311937. Carnegie Institu tion of Washington, Washington. Taschek, Jennifer 1994 T he Artifacts of Dzibilchaltun, Yucatan, Mexico: Shell, Polished Stone, Bone, Wood, and Ceramics New Orleans, Tulane University, Middle American Research Institute, Publication 50. Teeter, Wendy G. 2004 Animal Utilization in a Growing City: Vertebrate Exploitation at Caracol, Belize. In Maya Zooarchaeology: New Directions in Method and Theory, edited by Kitty F. Emery, pp. 177-191. Cotsen Institute of Archaeology, University of California, Los Angeles. Tilden, Douglas, Jaime J. Awe, Diane L. Slocum, Hannah H. Zanotto, Chrissina C. Burke, Ashley McKeown, Lee Meadows Jantz, Christophe Helmke, and Jorge Can 2017 The Structure A9 Tomb at Xunantunich Belize. In The Belize Valley Archaeological Reconnaissance Project: A Report of the 2016 Field Season, edited by Claire E. Ebert, Chrissina C. Burke, Julie A. Hoggarth, and Jaime J. Awe, pp. 346 -378. Belize Institute of Archaeology, National Institute of Culture and History, Belmopan. Thompson, J. Eric S. 1931 Archaeological Investigations in the Southern Cayo D istrict, British Honduras. Field Museum of Natural History Anthropological Series, Vol XVII, No. 3, Chicago, USA. Willey, Gordon R. 1972 The Artifacts of Altar de Sacrificios. Peabody Museum of Archaeology and Ethnology, Harvard University, Papers Vol. 6 4, No. 1. Cambridge, MA. Willey, Gordon, R. William R. Bullard Jr., John B. Glass and James G. Gifford 1965 Prehistoric Maya Settlement in the Belize Valley Papers of Peabody Museum, Harvard University No 54 Harvard University, Cambridge.

PAGE 113

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 115 119 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 10 FOOLS MAKE FEASTS, AND WISE MEN EAT THEM: INTERPRETING PROBLEM ATIC SMASH-AND-TRASH DEPOSITS AT KAKABIS H, BELIZE Kerry L. Sagebiel and Helen R. Haines Maya archaeologists commonly discover smash-and -trash deposits consisting of large quantities of broken sherds, lithics, faunal materials, and other remains in varying contexts on Maya sites. Interpretations of these deposits commonly fall into four gro ups: simple trash or midden deposits, remains of feasts, termination rituals, or rituals of commemoration and veneration. At Kakabish, Belize, a series of smash-and trash deposits, dating to the Formative period and consisting primarily of sherds and reconstructible vessels, were encountered in the Group D South Plaza. This paper will systematically present the expectations for these four different explanations, focusing on ceramic remains, and attempt to interpret the Group D South Plaza deposits by comparing them to the set of expectations for each. Introduction Kakabish is located in northcentral Belize about 10 km northwest of Lamanai. The site was first occupied during the Middle Formative period ca. 800 600 BC The earliest evidence of this occupati on is in the center of the Group D South Plaza in Operation 8. In this location a series of smash and trash deposits were encountered spatially associated with Formative platforms and a secondary burial placed in a hollow carved into bedrock. This burial was associated with many jade and other objects indicating the significance of the person to the early community. This burial and its contents have been discussed in detail elsewhere (Lockett Harris 2016). The focus of this paper is the smash and trash or problematical deposits located above the burial and exploring possible interpretations of them based primarily on the ceramics. Smash and trash deposits in the Maya area typically consist of large quantities of artifacts, mostly ceramic sherds, and are often associated with structures or monuments. The common interpretations of these deposits are as termination rituals, the remains of feasts, commemoration rituals, or simple accumulations of trash. This paper will explore some of the expectations for t he ceramics from these types of deposits and will then compare those expectations to the deposits found in Operation 8. Middens Middens are trash deposits and the expectations for such are that they will consist mostly of utilitarian objects that have rea ched the end of their use life (Schiffer 1972, 1987). Ceramics found in middens will consist of: a variety of domestic types (mostly monochrome slipped, unslipped, and striated) and domestic forms (cooking, food preparation, storage, and serving), few rit ual types or forms, relatively small sherds that have little to no uselife left, few reconstructible vessels, and no whole vessels. Depending on how long the midden is in use, it may contain sherds from a range of time periods. Feasting Feasting may be defined as a form of ritual activity centered on the communal consumption of food and drink. Rituals of this kind have played many important social, economic, and political roles in the lives of peoples around the world (Dietler 2011:179). Feasts are, therefore, occasional events of a special nature participated in by numerous people at discrete periods of time. Ceramics disposed of after a feast will likely have the following characteristics: a variety of large capacity utilitarian types and forms for food cooking and preparation, a high proportion of decorated serving vessels for social display, a large quantity of small serving vessels (possibly of redundant type and size), vessels for special festival foods and beverages, a lack of high capacity st orage vessels, and the presence of ritual types and forms (e.g., censers, musical instruments, figurines, etc.). Assuming any intentionally or unintentionally broken ceramics will be disposed

PAGE 114

Interpreting Problematic Smash and Trash Deposits at KaKabish 116 of immediately as part of the feasting ceremony as ritual trash (Walker 1995), sherds will likely be large, show little use wear or reuse, and many vessels will be reconstructible. The sherds should also date to a relatively short, discrete period of time. Termination Termination rituals are events that end the use life of a structure or monument. They sometimes occur with the complete abandonment of a site. The expectations for the ceramic remains from these events are more difficult to delineate as we have less ethnographic and historical knowledge of these kinds of events than we do for middens, feasting, or commemoration. In addition, termination rituals may involve feasts, or middens might be used as fill if the ev ent includes burying a structure or monument, so there is likely to be overlap with these other practices. If we assume that these events are mostly ritual in nature, then they should be rather discrete events, meaning the ceramics in them should date to a short period of time. Also, the ritual nature of termination will mean overlap with ceramics used in feasting, particularly, the remains of a large proportion of ritual vessel types and forms and many large sherds and reconstructible vessels. However, it is possible that during abandonment events, utilitarian ceramics will also be disposed of or abandoned rather than moved to new locations given their weight, bulk, and fragility (Schiffer 1987). There are also historical accounts of the destruction of all domestic pottery in every household during calendrical ceremonies in Mesoamerica and these often coincide with termination of buildings (Chase and Chase 1998; Tozzer 1941). In general, termination events likely will consist of a wider variety of potte ry than for feasting events, but will mirror feasts in terms of the deposits consisting of large sherds, reconstructible vessels, and ceramics dating to a defined period of time. Veneration Veneration is a recurring set of ritual behaviors tied to the veneration of a deity, person, place, and/or thing. Therefore, evidence of veneration will be located near burials and tombs, ritual structures, pilgrimage areas, natural features, shrines, or monuments. Veneration events tend to be recurring and often cons ist of the deposition of heirlooms, ritual objects, and unique objects as well as the removal of objects and parts of objects as mementos (Palka 2014). In some ways, veneration deposits may resemble midden deposits in that they will likely be used periodi cally over time, therefore, the ceramics in them will come from a range of time periods. The inclusion of heirloom ceramics and partial ceramic objects and the removal of pieces of ceramic objects as mementos will mean that many small sherds and non recon structible vessels will be included as in middens. Veneration deposits may resemble feasting and termination deposits in that they may include more ritual objects. It is also possible that types and forms will skew towards special types that are decorated. However, forms may be somewhat limited to vessels that can hold offerings such as serving vessels and small storage vessels. Cooking, food preparation vessels, and large storage vessels are likely to be less common. The two aspects of veneration that may distinguish the deposit type from feasting and termination is the inclusion of nonlocal ceramic types, as people often travel some distance to pay respects and make offerings, as well as the inclusion of heirlooms. The above expectations for the type s of ceramic remains found in these four kinds of deposits is quite generalized and specific quantities of each ceramic variable are not given ( Table 1 ). But they serve as a model or heuristic device against which to test particular deposits, as will be a ttempted for the Operation 8 smash andtrash deposits below. Deposits in Operation 8 in the Group D South Plaza Perhaps the most significant aspect of the Operation 8 smashand trash deposits is their location directly above an individual buried in a shallow hollow carved into bedrock. This secondary burial is surrounded by 23 small pits also carved into bedrock that contained a variety of artifacts: 47 jade objects (including an Olmec spoon pendant shaped like a bird), 2500+ shell beads, a deer ant ler, speleothems and a stalagmite, ochre and chalk balls, quartz

PAGE 115

Sagebiel and Haines 117 Table 1 General expectations for four different kinds smash-and -trash of deposits. Midden Feasting Termination Veneration Utilitarian types and forms dominant X A large proportion of ritual types and forms X X X A large proportion of decorated types X X Small sherds with no use life left predominant X Large sherds and reconstructible vessels X X X Sherds from many time periods X X Sherds from discrete time periods X X A large proportion of non local types X A large proportion of heirlooms X Location near specialized structures or monuments X X X crystals, chunks of ochre, faunal remains, obsidian, carbon, and 1,502 sherds. One pit was capped with an inverted Consejo Group modeled bowl and a Consejo Group incised bowl was inverted over the burial. The radiocarbon date from charcoal within the Consejo Red incised vessel is 799 511 BC (median 655 BC ). The Consejo vessels are part of the Mormoops Complex (Swasey Sphere) and date to 800600 BC. Directly above the burial and pits is the first smash and trash consisting of 4,409 sherds, 26 partially reconstructible vessels, faunal remains, obsidian, and carbon. If the bedrock deposits and smash andtrash deposits are considered together as a single deposit they contain: 5,911 sherds, 2 whole vessels, 26 partially reconstructible vessels, 5 spouted ve ssels, 4 worked sherds, 2 buckets, 2 cups, 2 lids, 1 censer, 1 bottle, and 1 possible figurine fragment. The preservation of the sherds in the smash and trash is good with 57% retaining their surface, 21% are rim sherds, and 75% of rims are identifiable t o form. The forms are quite significant with 89% identified as bowls and dishes and 9% as slipped jars (serving vessels). Only 1% of the identifiable forms are unslipped jars (cooking or storage). Half of the vessels in the smash and trash are types fall ing within the Noctilio Complex (Mamom Sphere) and dating to 600400 BC the other half may be curated or heirloom vessels from the earlier Mormoops Complex. Other important aspects of the deposit are the inclusion of many likely imported vessels with volc anic ash paste (Gomer 2013) and vessels with resist designs along with a large quantity of local decorated (incised and modeled) vessels. Comparing the deposit to the above expectations, the evidence is strongest that the earliest smash and trash contains the remains of a feast or other foodrelated ritual include: the large number of reconstructible vessels, the good overall preservation of the vessels, the very high proportion of serving vessels, the very low proportion of cooking and storage vessels, the high proportion of decorated ceramics, the inclusion of forms likely used for festival foods and beverages including spouted vessels for chocolate, buckets, cups, and a bottle and the inclusion of a possible figurine fragment and a censer. Evidence tha t this ritual or feast was related to the veneration of the person buried in bedrock includes the location directly above the burial and the inclusion of what are likely heirloom vessels and imported ceramics.

PAGE 116

Interpreting Problematic Smash and Trash Deposits at KaKabish 118 The burial and first smash andtrash were cove red by a floor and an uncarved limestone marker was erected that remained visible for some time as it protruded through the floor. Later, a second floor was laid down and a second smash and trash was deposited on top of it. The second smash andtrash con sisted of 1,276 sherds, 9 whole vessels, 71 reconstructible partial vessels, a lip to lip cache, 2 bottles, 1 spouted vessel, 1 worked sherd, and 1 censer. The deposit also contained faunal remains, charcoal and ash, and obsidian. The preservation of the sherds is excellent with 87% retaining their surface, 18% are rim sherds, and 78% of rims are identifiable to form. The forms are 85% bowls and dishes and 10% slipped jars (serving vessels). Again, only 1% of the identifiable forms are unslipped jars (cooking or storage). Four radiocarbon dates from the deposit range from 762388 BC (median of 575 BC), which accords with the dates of the Mamom pottery in the deposit. Twenty three percent of the pottery is earlier Mormoops pottery and could be the remains of curated or heirloom vessels. Many of the vessels are decorated, such as Muxunal Red on cream vessels that appear to be resist technique and Guitara Incised sherds with the double line break motif. Evidence that this is likely the remains of a feast or other foodrelated ritual include: the large number of whole and reconstructible vessels, the inclusion of a cache, the excellent preservation of the vessels, the very high proportion of serving vessels, the very low proportion of cooking and storage v essels, the high proportion of decorated ceramics, the inclusion of forms likely used for festival foods and beverages including a spouted vessel for chocolate and a bottle and the inclusion of a censer. The large amount of charcoal, ash, and faunal rem ains in the deposit also provides evidence of food preparation and consumption. Evidence that this ritual or feast was related to the veneration of the person buried in bedrock includes the location directly above the burial and the inclusion of what are likely heirloom vessels as well as the possible removal of portions of many vessels as mementos. At some unknown point later in the Formative, Structure (Str.) D 1 (not fully excavated) was constructed less than 2 m south of the bedrock burial and smash and trash deposits. It was then enlarged or re faced (Str. D 2) and a Chicago Orange (Formative) jar was cached in it. Later, an apron was added to the front (Str. D 3) with fill containing transitional Mamo Norm ally, building fill is thought to be recycled trash, although it is assumed that this kind of fill is unlikely to be brought from very far away. The fill of Str. D 3, however, is remarkably similar to that of the two earlier smash andtrash deposits and m ay also represent the remains of some kind of feast or ritual. The fill of Str. D 3 contains 818 sherds, 1 whole vessel, 5 buckets, 2 basins, 2 lids, 2 possible figurine fragments, 1 spouted vessel, 1 bottle, 1 censer, and 1 drum along with faunal remains, ash, and obsidian. The preservation is good with 71% of the sherds retaining their surface, 13% are rim sherds, and 57% of rims are identifiable to form. The forms are 76% bowls and dishes and 12% slipped jars, and only 3% unslipped cooking and storage jars. Interestingly, the deposit contains many small serving bowls and one vessel of unidentif iable form with a modeled face. This fill deposit is comparable in preservation and content to the two smash and trash deposits with the major difference being t hat fewer rims are identifiable to form, there is a lack of whole and reconstructible vessels, and there are fewer decorated vessels. However, there is still evidence that this is the remains of a feast or other foodrelated ritual including: the good pre servation of the vessels, the high proportion of serving vessels, the low proportion of cooking and storage vessels, the inclusion of forms likely used for festival foods and beveragesincluding a spouted vessel for chocolate and a bottle and the inclusion of a censer, a drum, two possible figurine fragments, and the vessel with the modeled face. The large amount of ash and faunal remains in the deposit also provide evidence of food preparation and consumption. Given that this fill is less well preserved and lacks the large number of whole and reconstructible vessels found in the earlier smash and trash deposits, it is likely that it was

PAGE 117

Sagebiel and Haines 119 exposed as midden for some time before being incorporated into Str. D 3. Whether it is all from a single event or wheth er it simply represents the remains of several special events located somewhere in the Group D South Plaza is unclear at this time. However, its location near the antecedent burial and smash and trash deposits is intriguing suggesting that ritual and, pos sibly, commemorative events continued to be held in that location for hundreds of years. Conclusions For many years Mayanists have interpreted smash and trash deposits as middens, termination events, the remains of feasts, and as evidence of commemorative or veneration activities. By further exploring what these events entailed in terms of behaviors and material remains and by measuring archaeological deposits against these expectations, we can begin to tease out different events, that will allow us a fir m basis from which to explore and theorize about issues such as placemaking, ancestor veneration, and political theater. References Chase, Diane Z., and Arlen F. Chase 1998 The Architectural Context of Caches, Burials, and Other Ritual Activities for the Classic Period Maya (as Reflected at Caracol, Belize). In Function and Meaning in Classic Maya Architecture edited by Stephen D. Houston, pp. 299 332. Dumbarton Oaks, Washington D.C. Dietler, Michael 2011 Feasting and Fasting. In The Oxford Handbook of the Archaeology of Ritual and Religion, edited by Timothy Insoll, pp. 179 194. Oxf ord University Press, New York. Gomer, Alice 2013 Humble Beginnings? A Study of the Origins of Settlement at the Site of Kakabish, Belize, through the Use of Macroscopic an d Petrographic Analysis of Ceramics. Unpublished M.A. Thesis, University College London. Lockett -Harris, Joshuah 2016 Sacred Space, Ancestors, and Authority: New Evidence of Developing Middle Formative Period Socio political Complexity from Kakabish, Nor thern Belize. Unpublished M.A. Thesis, Trent University, Peterborough, Ontario, Canada. Palka, Joel W. 2014 Maya Pilgrimage to Ritual Landscapes: Insights from Archaeology, History, and Ethnography University of New Mexico Press, Albuquerque. Schiffer, Michael B. 1972 Archaeological Context and Systemic Context. American Antiquity 37(2):156 165. 1987 Formation Processes of the Archaeological Record University of New Mexico Press, Albuquerque. Tozzer, Alfred M. 1941 Landas Relacion de las Cosas de Yucatn: A Translation. Papers of the Peabody Museum of Archaeology and Ethnology, Vol. 18. Harvard University, Cambridge, Massachusetts. Walker, William H. 1995 Ceremonial Trash? In Expanding Archaeology, edited by James M. Skibo, William H. Walker, and Axel E. Nielsen, pp. 67 79. University of Utah Press, Salt Lake City.

PAGE 118

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 121 130 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 11 THE EARLY -LATE CLASSIC TRANSIT ION AT CUELLO: RESULTS FROM THE 2017 SEASON OF THE CLASSIC CUELL O ARCHAEOLOGICAL PROJECT James L. Fitzsimmons, Natalie Figueroa and Prasanna Vankina The archaeological site of Cuello is one of the most recognizable Preclassic sites in the Maya area. Research began there in the mid -1970s under Norman Hammond and his team, and in the decades that followed Cuello and other early Maya sites like Nakbe, El Mirador, and San Bartolo was fundamental in changing our image of the Maya Preclassic. The majority of the work at Cuello focused on a very small, early portion of the city center, in and around an area known as Platform 34. However, much of the visible stone architecture at the site dates from the Classic Period, whe n the site transformed from being a dispersed rural community to a rather centralized, large town. Modest in size but clearly following the trend towards political complexity common in the Maya lowlands at that time, Cuello became part of a larger communi ty of Classic polities in northwestern Belize. The Classic Cuello Archaeological Project (CCAP) is exploring these later eras at Cuello and undertook its first excavations in January 2017. This paper presents the results of that first season. Introduction The archaeological site of Cuello is one of the most recognizable Preclassic sites in the Maya area ( Figure 1 ). Research began there in the mid 1970s under Norman Hammond and his team, and in the decades that followed Cuello and other early Maya sites like Nakbe, El Mirador, and San Bartolo was fundamental in changing our image of the Maya Preclassic. The early research at Cuello demonstrated that the origins of settled village life in the Peten and Yucatan began as early as 1200 BC, if not before, and shed light on the history of the earliest, identifiably Maya inhabitants of northwestern Belize. Most of the ea rly work at Cuello focused on a very small, Preclassic portion of the city center, in and around an area known as Platform 34, although mapping revealed a settlement that was much larger. One can find, for example, standing architecture approximately 1200 meters southwest, 750 meters south, 600 meters west, and 500 meters north of the heart of the ceremonial precinct, otherwise known as Platform 1 or the Acropolis (Hammond 1991: 9; see Figure 2 ). This stone architecture dates from the Classic Period, wh en the site transformed from being a dispersed rural community to a rather centralized, large town. Modest in size but clearly following the trend towards political complexity common in the Maya lowlands at that time, Cuello became part of a larger commun ity of Classic polities in northwestern Belize. The Classic Cuello Figure 1 Map of the Maya area (after Hammond 1991: Figure 1.1). Archaeological Project (CCAP), under the direction of James Fitzsimmons (Middlebury College), is exploring these late r eras at Cuello and undertook its first excavations in January 2017. This paper presents the results of that first season. Cuello The Cuello site is located at a mid point between two larger ceremonial centers, Nohmul and El Pozito, east of the River Hondo and very close to the modern town of Orange Walk,

PAGE 119

The Early -Late Classic Transition at Cuello 122 Figure 2 Map of Cuello (after Hammond 1991: Figure 2.2). Belize ( Figure 1 ). From the work done by Hammond, Kosakowsky, and others on the original Cuello project, there are indications that the general population at Cuello may have reached its peak during the Early Classic. Indeed, the Early Classic appears to have been a time of great change at the site. For example, the 60 test pits their team sunk thr oughout the site suggest that the stone architecture created during the Preclassic was actually not reused or remodeled by the Early Classic occupants, who seemed to prefer to build in new areas or atop perishable Preclassic buildings. The excavations of the 1970s and 1980s suggested, moreover, that the Maya built the northeast group in the Early Classic and that there was a local switch from kin based to community level ceremonial as well as political activity (Wilk and Wilhite, Jr. 1991: 125133). Cuell o then falls under the control of one family or group of related families

PAGE 120

Fitzsimmons, Figueroa, and Vankina 123 Figure 3 The Ceremonial Precinct or Acropolis (after Hammond 1991). based in the new, ceremonial precinct (Norman Hammond, personal communic ation 2016). According to Wilk and Wilhite, Jr. (ibid 127), the Early Classic was also a time when wealth disparities were on the rise. What Early Classic residential information we have suggests that basic, perishable houses coexisted with well plastered platforms faced with cut stone. Like other sites in the Maya lowlands, this was accompanied by a switch from building such houses on a single raised substructure, as was common in the Late Preclassic, to houses framing the familiar patios we know all too well for the Classic Period (Kurjack 1974; Ringle and Andrews 1988). Moreover, the number of lone platforms increased during the Early Classic: Wilk and Wilhite, Jr. (1991: 127128) have suggested that this heterogeneity stems from differential access to resources and is further evidence for the widening gap between the haves and the have nots at the dawn of the Classic Period.

PAGE 121

The Early -Late Classic Transition at Cuello 124 The Late Classic does not, on the basis of those same test pits, appear to have been a good time for Cuello. The results from those pits suggested that the population declined markedly during the Late Classic. Scholars have suggested that because Cuello is not in as advantageous a location for intensive farming/transportation as places like Nohmul, El Pozito, and Lamanai, at least some of the population here may have emigrated to these increasingly prosperous centers during the Late Classic (e.g., see Hammond 1974: 181; Wilk and Wilhite, Jr. 1991: 133). This is not to say that the earlier project at Cuello found no evidence of Lat e Classic construction at the site, but that it looked as if construction slowed down: the population became less visible archaeologically. According to Wilk and Wilhite, Jr., Cuello does not appear to have ever recovered from this decline, with little to no clear evidence of occupation during the Postclassic. 2017 Season at Cuello How does the above picture of Cuello accord with the results of the 2017 season in the northeast group ( Figure 3)? The short answer is that it does not. Perhaps a longer, bet ter answer would be its complicated. The vast majority of the test pits dug outside of the Preclassic epicenter in the 1970s and 1980s were nowhere near the northeast group. As a result, although we probably have a good sense of what the general popul ation was doing during the Early Classic, without major work in the northeast group we cannot hope to understand the political transformations happening at the site during the Classic Period. Over the course of the 2017 season, we excavated several test p its in the easternmost plaza fronting Structure 19, the largest temple pyramid at the site. We eventually combined three of the test pits to form a trench designed to find the front face of Structure 19. We also cleared out and cleaned two looters tunne ls/trenches in that structure. The results were rather surprising, especially in light of the enormous sequence represented by the earlier excavations in and around Platform 34. First, we were able to recover Fine Grey sherds on the surface in a few areas of the plaza. They were found in three different test pits, each 2025 meters apart from one another. We know that fine paste wares span the full Late Classic and Terminal Classic phases through Protohistoric eras in the Maya area, but that the peak of their popularity and distribution occur in the later Late Classic and Terminal Classic. As Joe Ball (2014) and others have noted, their distribution is panlowland, varying somewhat regionally by period (Bishop 2003; Forsyth 2005). As Laura Kosakowsky ha s observed, there is a trace occurrence of Late Classic Fine Gray in northwestern Belize as well (Kosakowsky et al. 2013). We also know that Fine Gray can be separated into two macrogroups, Chablekal and Tres Naciones, associated respectively with Late int o Terminal Classic timespan and with a fully Terminal Classic dating (e.,g., see Foias and Bishop 2013; Bishop and Rands 1982). Those two groups of Fine Grey pottery, divided chemically and stylistically, have been used to effectively document and map a n umber of patterns in the Maya area (e.g, see Ball 2014; Bishop and Foias 2013; Demarest 2013; Foias and Bishop 1997; Forne et al 2010; Forsyth 2005; Bishop et al. 1982). We do not know yet whether we are looking at Chablekal or Tres Naciones at Cuello although given the pattern for northwestern Belize it is probably Chablekal, which was a product of the greater Palenque region, the surrounding coastal plain of Tabasco and western Campeche, and the lower and middle Usumacinta drainage. If it is Chablekal, then it dates to no earlier than 750 or 760 this is when Chablekal first emerges in its home region. Given that Fine Grey is scattered on the surface, the northwest group may have collapsed on the same schedule that many other Classic centers did, pa rticularly those who fell in the late 8th or early 9th centuries. That being said, there is a Postclassic facet for Cuello: Hammond and his team found Postclassic refuse deposits around Structure 35, some dating as late as the 12th century. They even found a Late Postclassic effigy incensario in the vicinity of Str. 35 (Hammond 1991). But as of yet there is no evidence of actual Postclassic settlement here. As a result, refining the ceramic sequence for these end timesincluding, but not limited to, kno wing whether we are dealing with Chablekal

PAGE 122

Fitzsimmons, Figueroa, and Vankina 125 or Tres Naciones is critical for understanding how and when the site was abandoned. The Fine Grey we have is not refuse per se. It is surface scatter. Hopefully we will be able to get a larger sample of Fine Gr ey in future seasons, but for the moment it looks as if the final phase in the northeast group dates to the late eighth century. The second surprise is that the Classic epicenter appears to have been built atop a small, natural hill ( Figures 4 and 5). To the north, the hill slopes down into a rock quarry, and then back up to a flat region dotted with mounds and chultuns. Given the long occupation history for Cuello, the hill came as a bit of a shock: bedrock in this back area is perhaps one meter below the surface. By comparison, the test pits from the 1970s around Structure 35 were approximately 56 meters deep before they finished. There were a few fake out moments with the bedrock in these pits: one of the layers just 40 cm below the surface looks lik e bedrock, with a whitish paste, but does not have that familiar feel or the sterile soil one would generally associate with bedrock. We finally hit that sterile soil after about a meter, continuing downwards for a while just to make sure. This would be the third surprise. The humus layer in these pits is not quite like solid rock to excavate, but years if not decadesof being pounded down daily by cattle from the local Cuello distillery and ranch makes it a very different experience from digging normal humus. The floor below it as well as the ballast is actually a relief (Figure 5): it is much easier to excavate than the top level, which presumably would be much wider if not for the cattle. The overall sequence is not only shallow but brief: there is on ly one floor in front of Structure 19. Owing to its position just below the surface, it is in very bad shape. Presumably the floor was resurfaced many times over its history, but there is no evidence of any serious modifications and no second floor to be found, at least nowhere near Str. 19 or any of the nearby buildings. Below the floor are successive layers of light colored, loose soil that get progressively darker until one hits the white of bedrock. It was the same in every pit we dug, although admi ttedly because we did not excavate in or around Platform 1 this season, we may Figure 4 CC100 1 -4. The hill slopes downward, south to north, into a low plain dotted by mounds and chultuns. Figure 5 CC100 6. East wall profile. have hit upon a particularly late or hastily built portion of the ceremonial precinct. The final surprise would be the dates. Thus far we do not have radiocarbon for the plaza, but the ceramics seem to place the floor in the Late Classic (Kosakowsky and Robin son, personal communication 2017). The best samples we have are from floor and from the whitish layer immediately below the ballast for that floor. Much of the information we have on the ceramics from the 2017 season is preliminary; a more detailed analy sis has yet to take place and will surely refine the information presented here. The floor of the plaza has a mixture of Early Classic and Late Classic ceramic types, although it is so torn up by roots, rockfall, and other intrusive features that it is dif ficult to date with any precision. The layer below, however, is in terrific shape. Here the ceramics are a mixture of Preclassic and Terminal Preclassic types, so common Aguila Oranges and Sierra Reds as well as less common Zapote Striated and Yaxnik Thr oughthe Slip Incised. The layer

PAGE 123

The Early -Late Classic Transition at Cuello 126 Figure 6 Structure 19, west profile. below the floor fill of the plaza, in each of the test pits we excavated, contain Tzakol 3/Tepeu 1 types and forms as well (Kosakowsky and Robins on, personal communication 2017). What this means is that level below the plaza floor corresponds to sometime between 550 and 700 AD. Although the ceramic sample in the deepest, final layers is very small, the general sequence seems to be a mixed Early C lassicPreclassic layer, followed by what might be Late Preclassic. There is a somewhat similar picture with regard to Structure 19 itself ( Figure 6 ). Over the course of the season we cleaned out two looters pits. The first is directly on top of the bui lding on the east side; the looters followed the contours of the entrance to collapsed room on the top and started throwing rocks and debris down the back of the pyramid. The building appears to have had a vault on the top, despite its small size: Structure 19 is approximately 10.5 meters tall. The looters did not go below the level of the platform floor here, which extends across the entire top of the structure. In the second looters pit, they punched through the south side of the building. As can be seen in the drawing, Structure 19 was refurbished at least once, with a floor running through the looters pit. The looters disturbed and destroyed a feature of some sort: in the west wall of the pit there is a section of stone out past the retaining wal l that looks to have been disturbed in antiquity. Large boulders, either from a collapsed vault or disturbed rubble fill, proceed downwards along wall of the looters pit until one gets to a section of smaller stones. There arent many, and one can peel them off of the wall and find large stones behind them, but they are certainly odd. Our best guess is that a feature perhaps a cache was put below the floor in antiquity and then carved out by the looters, who left only bits of rubble clinging to the wall to hint at what it would have been. It does not look like the looters disturbed a burial: there is no bone in the pit, no evidence of a vault, and no burial floor in the pit. If it were a chamber, the looters would have had to destroy the ceiling, the f loor, and most of the walls to create what appears in the drawing. The feature looks more like a cache that was set into the floor of the original building; the hypothetical cache and original floor were then covered in the next construction phase. Whate ver the looters did find, they decided to keep going another meter into the building before giving up. The sherds in both looters pits were almost exclusively Late Preclassic types, including Margay Black on Red, although there are both Terminal Preclass ic and Middle Preclassic sherds as well (a Tower Hill Red onCream) in the fill (Kosakowsky and Robinson, personal communication 2017). From our own excavations on the outside of the building, it looks as if the builders laid a

PAGE 124

Fitzsimmons, Figueroa, and Vankina 127 foundation in Tzakol 3/Tepeu 1 times, so as to not have the building sink in the loose, sandy soil it sat upon, and then built the first phase of the structure. The plaza floor extends to the building, and as with the test pits we found no purely Early Classic material. Overall the building, like the plaza, gives the impression of a rather shallow time depth. Discussion We might sum up the 2017 season as follows: 1) There is an emerging LateTerminal Classic phase for Cuello, as represented by the appearance of Fine Grey (Chablekal?) in surface scatter; 2) No dedicated Early Classic layers have yet been found, although the structures surrounding Platform 1 may reveal a completely different sequence; 3) The plaza in front of Structure 19 was built at earliest at the beginning of the Late Cl assic, possibly as late as 700 AD; 4) The layers below the plaza floor are shallow and utterly unlike those found around Platform 34, which goes back to the Middle Preclassic; and 5) Structure 19 would appear to have a relatively shallow time depth, perhaps bui lt in Tzakol 3/Tepeu 1 times. The earliest phase we have atop the natural hill of the northeast group is Late Preclassic at best. We may even be looking at an Early Classic date here: many of the identifiable types we have, even for the lowest level atop the hill are transitional Late PreclassicEarly Classic ceramics. There are no floors, postholes, or any real signs of activity in the ceremonial precinct before the Late PreclassicEarly Classic transition. In other words, the Cuello we excavated in the 2017 season is thoroughly unlike the one from the 1970s and 1980s. Taken by itself, the data from portion of the precinct we excavated does not fit with the idea of a boom in population during the Early Classic or a steep decline in the Late Classic. If one were excavating Cuello without the prior history of research here, one would probably conclude that there was a boom in construction in the Late Classic, with a minor Late PreclassicEarly Classic facet. How are we supposed to reconcile these two very different results? Again, we have to remember that the original project at Cuello excavated 60 test pits. Those pits were determined at random, with almost none of them falling anywhere near the northeast ceremonial precinct. They provide an image of a Preclassic Cuello that changes suddenly in the Early Classic, with a shift to a new ceremonial center in the northeast, a population boom during the Early Classic, a decline in the Late Classic, and a sparse Postclassic component. The 2017 excavations were from 6 test pits, were not random, and were only in the east part of the ceremonial precinct. They suggest a Late Classic transition, with nothing that is purely Early Classic. Most of the material is not even Preclassic. So again, what do we do with these two results? Perhaps the two results are not mutually exclusive. We have terms like Late Preclassic, Early Classic, Late Classic, and Terminal Classic for a reason. They help us to discuss broad moments in prehistory. But it is possible for us to fall victim to our own categories. If we look at what has happened to the Preclassic over the last twenty years or so, what we see is a Preclassic whose outer boundaries are eroding. To be sure, there is a boundary between the Preclassic and Classic. B ut it is not a hard and fast line for many places. We would do well to remember that for terms like Early and Late Classic too. What we may be seeing at Cuello, with these two seemingly different results, is evidence of change happening during a transitional phase, namely the Early Late Classic transition. Tzakol 3/Tepeu 1 spans a rather large time span, ranging from about 550 to 700 AD. We found these ceramics in a band across the plaza, stretching to the base of Structure 19. There are no pur ely Early Classic layers in the northeast ceremonial precinct; Early and Late Classic ceramics blend together in each of the test pits we excavated. So what if the transition to Classicstyle ceremonial and political behavior at Cuello actually happens so metime between 550 and 650? One would still see expansion and development, in terms of settlement, for the Early Classic, as well as a gradual decline in the Late Classic, which is

PAGE 125

The Early -Late Classic Transition at Cuello 128 what the older (and much more complete) settlement data suggests. What if Cuello becomes more complex first and then one family or group of families decides to build the ceremonial precinct? In this scheme, there are several emerging religious, political, and economic hierarchies in the town (heterarchy). One family builds upon preexisting ideas and inequalities, creating a kind of bottom up (as opposed to top down) social enchantment after local conditions become suitable for it (e.g., see Geertz 1980; Demarest 1992; Houston 1998; Canuto and Fash 2004). They do not create a n ideology to suddenly sell to the populace, convincing them to become more like their larger neighbors, but rather exploit an ideology already present in the town as a result of the social, political, and religious climate of the Maya lowlands in the mi d 500s and 600s. Architecturally, the ceremonial precinct appears like a sudden change, but given the settlement data it may actually represent the outgrowth of a slow process begun in the Early Classic. What if the family (or group of families) build the ceremonial precinct and seize power during the Early Late Classic transition, a time of increasing complexity but fail in the longterm because places like Nohmul or El Posito outpace Cuello? Perhaps once the locals make the shift to a Classicstyle polity they realize that other places do it better and begin to leave. Obviously, more research is necessary to flesh out the timing and the reasons for these transitions. Conclusion In order to confirm the EarlyLate Classic transition hypothesis presented here, we will have to return to Structure 19 as well as sink further pits in Platform 1 (where there is a pre existing test pit from the 1980s). The shallow time depth, the lack of multiple floors in our excavations, and the missing, unequivocally Ear ly Classic layers in the northeast group all suggest a very short period of glory for the area. The Fine Grey here, as well as sporadic encounters with Postclassic material around the site, suggest s a long but meager presence too. The family or group of families responsible for the ceremonial precinct fell out of favor in the 8th century but the settlement itself, having a much longer history, managed to survive longterm. Why would Cuello have suddenly reinvented itself, only to fall apart? The answer t o this question, unfortunately, rests upon two other long term questions for the project. The first, simply put, is Why Cuello? So why would anybody choose to live here during the Classic Period, as opposed to anywhere else within 5 10 kilometers or so? Undoubtedly the farming was good, as it certainly was during the Preclassic, but there does not appear to be a clear reason for why people would live here, as opposed to the town that is now Orange Walk. Cuello is not on the river that would be Orange Walk and although it has chert in abundance, the same could be said of many places in the general vicinity. Perhaps the people at Cuello were avoiding floods, but there is probably more to this question than the river. The second, and rather interrelated question, is What was Cuellos relationship to the surrounding population centers? Nohmul and El Posito are the largest centers in the area and presumably drew people in from the countryside. As Hammond has suggested, Nohmul, El Posito, and distant Lam anai may indeed have been better farmland. Cuello is flanked on either side by the sites of Nohmul and El Posito. San Estevan is a little over 10 km away across the river. Is Cuello working with them? Competing with them? Subordinate to one of them? Perhaps Cuello becomes a good place to live, if for a brief moment, because it is responding to larger geopolitical or economic events. More research needs to be done at Cuello to answer both of these questions. The key is creating more of a regional framework for the Classic Period in northwestern Belize, perhaps focusing locally on the transition between the Early and the Late Classic. Acknowledgements The Classic Cuello Archaeological Project would like to thank the Belize Institute of Archaeology, p articularly the current and former Directors, Drs. John Morris and Jaime Awe, respectively. Likewise, thanks are long overdue for the efforts of IA staff members, including Melissa Badillo, Sylvia Batty, Jorge Can, and Josue Ramos, for helping

PAGE 126

Fitzsimmons, Figueroa, and Vankina 129 to get the 2017 season going. The project would also like to thank Srs. Waldir Cuello, Francisco Cuello, and the entire Cuello family for their help with the 2017 season, as well as Orlando and Cindy De la Fuente for their hospitality. The CCAP would like to thank Middlebury College, Christopher D. Gladstone, and Elise J. Rabekoff for their generosity and support of the 2017 season. Likewise, we would like to thank Laura Kosakowsky and Robin Robinson, who generously agreed to look at waves upon waves of jpegs of ce ramics at the Vancouver meetings. Finally, the project would like to thank Norman Hammond, whose support, insight and guidance over the years has been invaluable. References Ball, Joseph 2014 Rethinking the Becan ceramic sequence: continuities, disjunctions, segmentation, and chronology. Latin American Antiquity 25: 427448. Bishop, Ronald L. 2003 Five decades of Maya Fine Orange ceramic investigation by INAA. In Patterns and Process: A Festschrift in Honor of Dr. Edward V. Sayre ed. Lambertus van Zest, pp. 81-91. Smithsonian Center for Materials Research and Education, Suitland. Bishop, Ronald L., R.L. Rands, and George Holley 1982 Ceramic compositional analysis in archaeological perspective. Advances in Arch aeological Method and Theory 5: 275330. Canuto, Marcello A., and William L. Fash, Jr. 2004 The blind spot: where the elite and non-elite meet. In Continuities and Changes in Maya Archaeology: Perspectives at the Millennium, edited by Charles W. Golden an d Greg Borgstede, pp. 51 -76. Routledge, New York. Demarest, Arthur A. 1992 Ideology in ancient Maya cultural evolution. In Ideology and Pre -Columbian Civilizations edited by Arthur A. Demarest and G.W. Conrad, pp. 135157. School of American Research Press, Santa Fe. 2013 The collapse of the Classic Maya kingdoms of the southwestern Petn: Implications for the end of Classic Maya civilization. In Millenary Maya Societies: Past Crises and Resilience edited by M. Charlotte Arnauld and Alain Breton, pp. 22-48. Fitzsimmons, James L. 2015 The charismatic polity: Zapote Bobal and the birth of authority at Jaguar Hill. In Classic Maya Polities of the Southern Lowlands edited by Damien B. Marken and James L. Fitzsimmons, pp. 225 -241. University of Colorado Pre ss, Tucson. Foias, Antonia, and Ronald Bishop 1997 Changing ceramic production and exchange in the Petexbatun region, Guatemala: reconsidering the Classic Maya collapse. Ancient Mesoamerica 8(2): 275-291. Foias, Antonia, and Ronald Bishop, eds. 2013 Ceramics, Production and Exchange in the Petexbatun Region Vanderbilt University Press, Nashville. Forn, M., R. L. Bishop, A. A. Demarest, M. J. Blackman, and E. L. Sears 2010 Gris Fino, Naranja Fino: presencia temprana y fuentes de produccin, el caso de Cancun, In XXIII Simposio de Investigaciones Arqueolgicas en Guatemala 2009, edited by B.Arroyo, L. Paiz Aragn, A. Linares Palma, and A. L. Arroyave, 1163 1182. Guatemala City: Museo Nacional de Arqueologa y Etnologa. Forsyth, Don 2005 A survey of Terminal Classic ceramic complexes. In Geographies of Power: Understanding the Nature of Terminal Classic Pottery in the Maya Lowlands (eds. S.L Lpez Varela and A.E. Foias), 23 -40, BAR International Series 1447, Archaeopress, Oxford. Geertz, Clifford 1980 Negara: The Theater State in Nineteenth Century Bali. Princeton University Press, Princeton. Hammond, Norman 1980 Early Maya ceremonial at Cuello, Belize. Antiquity 54: 176190. 1991 Cuello: An Early Maya Community in Belize Cambridge, En gland: Cambridge University Press. Hammond, Norman, Amanda Clarke, and Sara Donaghey 1995 The long goodbye: Middle Preclassic Maya Archaeology at Cuello, Belize. Latin American Antiquity 6(2): 120128. Hammond, Norman, Amanda Clarke, and Francisco Estrad a Belli 1992 Middle Preclassic Maya buildings and Burials at Cuello, Belize. Antiquity 66: 955 -964. Hammond, Norman, Amanda Clarke, and Cynthia Robin 1991 Middle Preclassic buildings and burials at Cuello, Belize: 1990 investigations. Latin American Antiquity 2: 353-363. Housley, Rupert M., Norman Hammond, and Ian A. Law 1991 AMS Radiocarbon dating of Preclas sic burials at Cuello, Belize. American Antiquity 56: 514 -519.

PAGE 127

The Early -Late Classic Transition at Cuello 130 Houston, Stephen D., editor 1998 Function and Meaning in Classic Maya Architectur e Dumbarton Oaks, Washington, D.C. Houston, Stephen D., Hector Escobedo, Mark Child, Charles Golden, and A. Rene Muoz 2003 The Moral Community: Maya Settlement Transformation at Piedras Negras, Guatemala." The Social Construction of Ancient Cities Ed. Smith, M.L., pp. 212 253. Washington, D.C.: Smithsonian Institution Press. Kosakowsky, Laura J. 1993 Intra -site variability of the Formative ceramics from Cuello, Belize: an analysis of form and function. Ph.D. dissertation, University of Arizona. Univers ity Microfilms, Ann Arbor, Michigan. 1997 Preclassic Maya pottery at Cuello, Belize. Anthropological Papers of the University of Arizona No. 47. University of Arizona Press, Tucson. Kurjack, E.B. 1974 Prehistoric Lowland May Community and Social Organization Middle American Research Institute, Tulane University, Publication 38. New Orleans, LA. Law, Ian A., Rupert A. Housley, Norman Hammond, and Robert E.M. Hedges 1991 Cuello: resolving the chronology through direct dating of conserved and low -collagen bone by AMS. Radiocarbon 33(2): 1 -19. McAnany, Patricia 1995 Living with the Ancestors University of Texas Press, Austin. Pring, Duncan C. 1977 The Preclassic ceramics of northern Belize. Ph.D. dissertation, University of London. University Microfilms, Ann Arbor, Michigan. Pring, Duncan C., and Norman Hammond 1982 The stratigraphic priority of Swasey ceramics at Cuello, Belize. Ceramica de Cultural Maya 12: 4348. Temple University, Philadelphia. Ringle, William, and E. Wyllys Andrews V 1988 Formative residence at Komchen, Yucatan, Mexico I n Household and Community in the Mesoamerican Past eds. R. Wilk and W. Ashmore, pp. 171-199. University of New Mexico Press, Albuquerque. Robin, Cynthia 1989 Preclassic Maya Burials at Cuello, Belize. BAR International Series 480. British Archaeological Reports, Oxford. Robin, Cynthia, Andrew Wyatt, James Meierhoff, and Caleb Kestle 2015 Political interaction: a view from the 2000 -year history of the farming community at Chan. In Classic Maya Polities of t he Southern Lowlands edited by Damien B. Marken and James L. Fitzsimmons, pp. 99122. University of Colorado Press, Tucson. Wilk, Richard R., and Hal Wilhite, Jr. 1991 The community of Cuello: patterns of household and settlement change. In Cuello: An Ea rly Maya Community in Belize edited by Norman Hammond, pp. 118 -133. Cambridge University Press, Cambridge.

PAGE 128

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 131 140 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 1 2 LONG AGO BUT NOT FOR GOTTEN: THE EARLY PRECLASSIC SWASEY CERAMIC SPHERE OF NORTHERN BELIZE Laura J. Kosakowsky Kerry L. Sagebiel and Duncan C. Pring When Early Preclassic ceramics were encountered in the 1970s at the site of Cuello in Northern Belize there were few other sites or regions with comparable pottery. Since the original discovery of Swasey ceramics at Cuello, many sites in northern Belize and adjacent areas have been identified that share this pre -Mamom tradition. In th is paper we propose the identification of and describe the principal attributes of the ceramics of the Swasey Sphere including the Swasey and Bladen Ceramic Complexes at Cuello, and the Mormoops Ceramic Complex at Kakabish. We will highlight the geographic extent of Swasey Sphere ceramics and compare them to other contemporary Early Preclassic pottery such as Cunil in the Belize Valley, Eb in the Petn, and Xe in the Pasion River area. The Early Preclassic is a period of increased sedentism, population growth, and emerging Maya cities, however, pre-Mamom ceramics across the southern Maya Lowlands exhibit only minor shared attributes and they appear to be more regionalized than subsequent time periods. Introduction In the 1970s when pre Mamom ceramics were uncovered at the site of Cuello (Hammond 1975, 1976, 1978, 1991; Hammond et al. 1979; Pring 1977) near the modern town of Orange W alk in northern Belize ( Figure 1 ), there were few other known sites with early pottery. Middle Preclassic Mamom pottery (Smith 1955; Smith and Gifford 1966) was well documented at other Maya lowland sites, but pre Mamom ceramics were found only at Tikal (Culbert 1993), Yaxha/ Sacnab (Rice 1979), Altar de Sacrificios (Adams 1971), and Seibal/[Ceibal] (Sabloff 1975). Some types of the earliest ceramic complex, Jenney Creek, at Barton Ramie (Gifford 1976) included possible pre Mamom attributes. Since that time, pre Mamom ceramics have been found at numerous other sites in northern Belize and throughout the southern Maya Lowlands. The Swasey Ceramic Complex at Cuello The earliest preMamom ceramics at Cuello, called Swasey (Pring 1977), are simple in design, execution, and form, although technologically sophisticated (Kosakowsky and Pring 1998). The complete sample of Swasey ceramics from Cuello numbered in the tens of thousands, however a sub sample of representative Swasey pot tery was selected from chronologically unmixed deposits for the type descriptions, totaling 4,794 diagnostic rim sherds, and three whole vessels. The Swasey Ceramic Complex is defined by five ceramic groups ( Table 1 ), and is confined to the lowest Figure 1 Map of northern Belize, showing location of Cuello and Kakabish and other archaeological sites in the region. levels of excavation of Platform 34 at Cuello, identified on stratigraphic sections and Harris matrices for the site as Phases 0 (old land surface) and I (Hammond 1991). The Swasey Complex dates to between 1,000800 BCE. Only 10% of the assemblage is unslipped (Copetilla Group), while the remaining 90% is monochrome slipped, most commonly red of the Consejo Group with a generally gloss y, nonwaxy surface finish, a white underslip, and surfaces that exhibit characteristic rootlet

PAGE 129

The Early Preclassic Swasey Ceramic Sphere of Northern Belize 132 Table 1 Ceramic Types of the Swasey Complex at Cuello. Ware Group Type Variety Unspecified Copetilla Copetilla Unslipped Copetilla Variety Patchchacan Pattern Burnished Patchchacan Variety Rio Nuevo Glossy Consejo Consejo Red Consejo Variety Backlanding Incised Backlanding Variety Backlanding Incised Groovedincised Variety Pettville Red and cream Pettville Variety Pettville Red and cream Variety Unspecified Other Consejo Group Unspecified Tiger Tiger Buff Tiger Variety Cowpen Incised Cowpen Variety Cowpen Incised Groovedincised Variety Machaca Machaca Black Machaca Variety Chacalte Incised Chacalte Variety Fort George Orange Chicago Chicago Orange Chicago Variety erosion marks. Less common monochrome ceramics include the Tiger (Buff) Group, the Machaca (Black) Group, and the self slipped or washed pottery of the Chicago (Orange) Group. Decoration consists of groove incising (Backlanding Incised: Grooved incised Variety and Cowpen Incised: Grooved incised Variety) and dichrome slips that are created by leaving the white underslip uncovered and slipping only one side of the vessel red (Pettville Red andcream). Examples of fine line incising occur in the monochrome red, buff, and black groups, and fewer examples of modeling and punctation on monochrome r eds, and pattern burnishing on unslipped jars (Patchchacan Pattern Burnished) also occur (see Pring [1977] and Kosakowsky [1983, 1987] for greater details). Common vessel forms are bowls and jars with thickened rims and squared lips. Strap handles fashio ned of double and triple cylinders of clay are characteristic, attached loosely to jars of the Tiger and Chicago ceramic groups, from the rim to the neck/ body juncture. There does not appear to be a specialized set of ceramics (Kosakowsky and Pring 1998; Robin 1989), unlike early ceramic assemblages elsewhere in Mesoamerica (Clark and Blake 1994; Clark and Gosser 1995), and the repertoire of Swasey Figure 2 Ceramics of the Swasey Ceramic Complex at Cuello: a. Copetilla Group; b. Consejo Group; c. Machaca Group; d. Tiger Group; e. Chicago Group. pottery includes a complete range of forms for both utilitarian and non utilitarian functions ( Figure 2). In order to test the validity of the ceramic sequence, the analysis of the ceramics from the later 1 992 and 1993 excavations at Cuello was

PAGE 130

Kosakowsky, Sagebiel, and Pring 133 Table 2 Ceramic Types of the Bladen Complex at Cuello. Ware Group Type Variety Unspecified Copetilla Copetilla Unslipped Gallon Jug Variety Rio Nuevo Glossy Consejo Consejo Red Estrella Variety Barquedier Incised Barquedier Variety Barquedier Incised Grooved incised Variety Fireburn Red and cream Fireburn Variety Fireburn Red and cream Variety Unspecified Cudjoe Composite Cudjoe Variety Cudjoe Composite Variety Unspecified Sand Hill Gouged incised Sand Hill Variety Canquin Blackon red Canquin Variety Other Consejo Group Unspecified Ramgoat Ramgoat Red Ramgoat Variety Calcutta Incised Grooved incised Variety London Red and unslipped Variety Unspecified Tiger Tiger Buff Cut and Throw Away Variety Cowpen Incised New Home Variety Last Chance Grooved incised Last Chance Variety Peppercamp Pattern burnished Peppercamp Variety Machaca Machaca Black Wamil Variety Chacalte Incised Yo Creek Variety Quamina Quamina Cream Quamina Variety Tower Hill Red on cream Tower Hill Variety Tower Hill Red oncream Variety Unspecified (Resist) Isabella Bank Incised Isabella Bank Variety Doubloon Bank Grooved incised Doubloon Bank Variety Saltillo Orange on cream Saltillo Variety Crabcatcher Crabcatcher Red Crabcatcher Variety Gold Button Gold Button Brown Gold Button Variety Fort George Orange Chicago Chicago Orange Nago Bank Variety Cotton Tree Incised Cotton Tree Variety Willows Bank Pattern b urnished Willows Bank Variety Unspecified Honey Camp Honey Camp Orange brown Honey Camp Variety Copper Bank Incised Copper Bank Variety done by Kosakowsky and Pring (1998) without prior knowledge of the stratigraphic context numbers or Harris matrix information from the site. When chronological assessments were plotted on the Harris matrix, they produced a clear distinction between unmixed contexts with Swasey pottery, identified by vessel forms and types present only in that complex, and those mixed with the later Bladen types. The Bladen Ceramic Complex at Cuello The original definition of the Swasey Complex (Pring 1977) included Bladen pottery. The larger sample size from later seasons allowed the subsequent refinement of the sequence and separation of Bladen pottery from earlier Swasey pottery (Kosakowsky1987). The Bladen Complex lies stratigraphically above Swasey and, although clearly derivat ive from Swasey, represents a period of elaboration in

PAGE 131

The Early Preclassic Swasey Ceramic Sphere of Northern Belize 134 Figure 3 Ceramics of the Bladen Ceramic Complex at Cuello: a. Quamina Group including Tower Hill Red-oncream; b. Honey Camp Group. ceramic technology and design. The complete sample size of Bladen ceramics numbered in the tens of thousands, however a sub sample of approximately 13,000+ diagnostic rim sherds and 13 whole vessels was selected for intensive analysis and type descriptions. While some researchers have suggested that the modal differences between Swasey and Bladen argue for the faceting of a single complex (Andrews in Andrews and Hammond 1990:579; Valdez 1987), Kosakowsky and Pring (1998) feel that the introduction of new ceramic types and vessel form ch anges are significant enough to warrant the identification of separate complexes using Gifford's (1976) definition of a ceramic complex. The Bladen Complex dates to between 800600 BCE. The Bladen Complex is composed of nine ceramic groups ( Table 2 ) of wh ich eight are slipped, again predominantly monochrome red of the Consejo Group, and is marked by the first appearance of cream (Quamina) and red on cream slipped pottery (Tower Hill redoncream) that is antecedent to the red on creams in Figure 4 Specialized ceramics of the Bladen Ceramic Complex at Cuello: a. Crabcatcher Red; b. Gold Button Brown (Illustrated by Justine Hopkins). the later Middle Preclassic. The Copetilla (Unslipped), Consejo (Red), Tiger (Buff), Machaca (Black), and Chicago ( Orange) Groups continue from the Swasey Complex with new types and varieties, and vessel forms. New ceramic groups include the previously mentioned Quamina (Cream), Ramgoat (Red), which lacks the cream underslip of the Consejo Group, and Honey Camp (Orang e brown) ( Figure 3 ). Two new ceramic groups, Crabcatcher (Red) and Gold Button (Brown) were found only in a chultun, Feature 361, and consist mostly of small thin walled bowls ( Figure 4) that may represent the debris from the ceremonial drinking of chicha during a feasting event (Hammond et al. 1995). While the rest of the Bladen ceramic assemblage, like the preceding Swasey pottery, represents a full range of utilitarian and non utilitarian types and forms, the chultun sample appears specialized, althoug h there does not seem to be a specialized set of Bladen mortuary vessels (Robin 1989). Monochrome slips of the Bladen complex are similar in texture (glossy and non waxy) to the preceding Swasey slips, but are more consistent in color. Bowl and jar forms continue to predominate, with the addition of flaring sided dishes. Thickened rims and squared lips

PAGE 132

Kosakowsky, Sagebiel, and Pring 135 Table 3 Ceramic Types of the Mormoops Complex at Kakabish. Ware Group Type Variety Unspecified Copetilla Copetilla Unslipped Unspecified Rio Nuevo Glossy Consejo Consejo Red Unspecified Barquedier Incised Unspecified Barquedier Incised Groovedincised Fireburn Red and cream Unspecified Cudjoe Composite Unspecified Ramgoat Ramgoat Red Unspecified Machaca Machaca Black Unspecified Chacalte Incised Unspecified Quamina Quamina Cream Unspecified Tower Hill Red oncream Unspecified Fort George Orange Chicago Chicago Orange Unspecified tend to be replaced by direct and exteriorly folded rims and round lips. Decoration is more complex than in the preceding Swasey Complex and includes incising, groove incising, punctation, gouge incising, modeling, black line smudging, and resist. Pattern burnishing, which first appeared in the Swasey Complex on the unslipped type, is extended to the buff and orange brown slipped types. Incised loop handles predominate, replacing the strap handles of the previous complex. As stated previously, the Swasey Cer amic Complex at Cuello is confined to only the lowest levels of excavation (Hammond 1991). The succeeding Bladen Ceramic Complex corresponds to Phases II IIIA on stratigraphic sections for Cuello (Hammond 1991; Kosakowsky and Pring 1998), and thus the rat ionale for splitting the two complexes is based not only on the introduction of new types and forms, thus conforming to type: variety rules (Gifford 1976; Willey et al. 1967), but on clear stratigraphic differences. Furthermore, the decision to split the ceramics into separate complexes was intended to help highlight, rather than obscure the differences. The Mormoops Ceramic Complex at Kakabish The site of Kakabish is located in north central Belize, approximately 10 km northeast of the site of Lamanai between the New River Lagoon and the Rio Bravo Escarpment (see Figure 1). Recent excavations at the site began in 2005, with the bulk of the research occurring since 2010 defining a long occupation sequence (Sagebiel and Haines 2015). The Mormoops Comple x is the earliest complex at Kakabish, dating to the Early Preclassic 800 600 BCE and overlaps in time and principle identifying attributes with the Bladen Complex at Cuello (Kosakowsky 1983, 1987). The following description of Mormoops ceramics is based on 2,017 sherds, including 673 rims, and 37 reconstructible vessels ( Table 3 ). Mormoops ceramics have almost exclusively been excavated from the site center, primarily out of the basal deposits of the Group D Plaza (particularly Operations 3, 8, and 15) (Aimers 2012; Haines 2012; Haines et al. 2014; Lockett Harris 2013, 2014; Sagebiel and Haines 2015). The predominant Consejo Group at Kakabish has a bright, glossy, nonwaxy, red slip on a white underslip. The red slip has typical white rootlet erosion a nd fully eroded surfaces often have a white, chalky appearance, apparently, the remnants of the underslip. There are instances where the underslip was not applied and the slip was applied directly to the paste body. The slip in these cases tends to have a reddish orange rather than bright red color and lacks the white rootlet and chalky erosion. This type is known as Ramgoat Red and is part of the Ramgoat Group (Pring 1977; Valdez 1987).

PAGE 133

The Early Preclassic Swasey Ceramic Sphere of Northern Belize 136 Paste colors are light pastels of white to yellow (buff), brown (ta n), pink to orange, or gray with thick gray cores. Large angular calcite inclusions predominate, but sherd inclusions occasionally occur. Several Consejo vessels were sampled for petrographic analysis and were found to have volcanic ash inclusions (Gomer 2013). Bowls are the most prevalent form; most have flared to slightly outcurved walls and flat bases. Rims are direct or slightly everted with round or, occasionally, square lips. Jars with short necks and tecomates are also fairly common and there are several bottles as well ( Figure 5 ). The Consejo Group also includes Barquedier Incised: Unspecified, Barquedier Incised: Groovedincised, Fireburn Redand cream, and Cudjoe Composite. Other Mormoops Groups include Copetilla, Ramgoat, Machaca (including Chalcate Incised), Quamina (including Tower Hill Red on cream), and Chicago (Haines et al. 2014; Sagebiel and Haines 2015). Early Preclassic Ceramics in Northern Belize and the Southern Maya Lowlands Swasey Sphere ceramics, similar to those at Cuello and Kakabish, also have been identified in northern Belize (see Figure 1) at the following sites with full ceramic sphere membership (Ball 1976): Colha (Valdez 1987), Nohmul (Hammond et al. 1987, 1988; Pring 1977), El Pozito (Pring 1977), San Estevan (Rosensw ig 2008), Santa Rita Corozal (Chase and Chase 1987; Pring 1977), Pulltrouser Swamp and Kaxob (Fry 1989) Blue Creek (Kosakowsky and Lohse 2003), and Kichpanha (McDow 1997). Sites in northern Belize with peripheral sphere membership include: Chan Chich (V aldez and Houk 2000), Gran Cacao (Sagebiel 2005), and Dos Hombres (Sullivan and Sagebiel 2003). Additionally, Swasey Sphere ceramics have been identified at the site of Rio Azul in the northeastern Petn of Guatemala (Adams 1999) and in southeastern Campe che (Walker 2014). While several Bladen decorative modes, such as fineline incising, exist in common with the Xe ceramics of Altar de Sacrificios (Adams 1971), and the Real Xe complex at Ceibal Figure 5 Ceramics of the Mormoops Ceramic Complex at Kakabish: a. Consejo Group; b. Machaca Group; c. Chicago Group; d. Copetilla Group; e. Quamina Group including Tower Hill Red -on-cream. (Seibal) (Inomata et al. 2013; Sabloff 1975) in Guatemala, they are typologically distinct; surface colors and textures as well as pastes are very different. The paucity of white slipped sherds in the Cuello and Kakabish collections, a ceramic marker for Xe, supports this differentiation. Early Eb pottery from the earliest contexts at Tikal also exhibits modal decorati ve similarities to the Bladen Complex (Culbert 1993), as does the early Ah Pam pottery of the Lake Yaxha/Sacnab region (Rice 1979), but these are also typologically distinct from the pre Mamom in northern Belize. Modal similarities, such as red oncream d ecoration also exist with coeval ceramics from the Belize Valley including the late Cunil at Cahal Pech (Sullivan and Awe 2013), Kanocha at Blackman Eddy (Garber et al. 2004), and early Jenney Creek at Barton Ramie (Gifford 1976). However, all these preM amom complexes throughout the southern Maya Lowlands are marked by regional heterogeneity rather than the increasing homogeneity that begins in the Middle Preclassic. Conclusions The original dating of the Swasey Complex at Cuello was controversial due to

PAGE 134

Kosakowsky, Sagebiel, and Pring 137 problems with radiocarbon samples that resulted from the recycling of old wood in later contexts (see Hammond 1984; Marcus 1983, 1984; Potter et al. 1984). The re evaluation of the radiocarbon dates from the site (Andrews and Ham mond 1990; Housley et al. 1991; Law et al. 1991) produced a shortened chronology, and the extensive excavations demonstrated, without a doubt, the stratigraphic position of the Swasey ceramics beneath Bladen (Pring and Hammond 1982). The shortened chronology placed the lowest levels at Cuello between 1,200 1,000 BCE though there is still some disagreement about the assignment of 1,200 BCE for the earliest Swasey dates and whether or not the earliest Cunil dates in the Belize Valley predate those in northern Belize (Lohse 2010). The recent work by Lohse (2010) on the Preceramic in Belize, summarizes these early radiocarbon dates, and suggests that at least by 1,000 BCE there were sedentary, pottery producing villages throughout the southern Maya lowlands. U nfortunately, the dating controversy about Swasey ceramics resulted in some questioning the regional differences in the preMamom (Lopez Varela 1996). Another consequence of the shortened chronology was the renaming of the time period in the archaeologica l literature as Early Middle Preclassic rather than Early Preclassic to reflect these later dates and to differentiate the ceramics from some of the earlier pottery in the Maya highlands, Pacific coastal region, and the Gulf Coast of Mexico (Coe and Di ehl 1980; Lowe 1975, 1978; Sharer 1978). While there is currently a gap in occupation between the Archaic and the earliest ceramics in northern Belize (Lohse 2010), we have decided to return to the original designation of Swasey Sphere ceramics as Early Preclassic to highlight the important differences between this pre Mamom pottery and the subsequent Middle Preclassic Mamom, rather than linking it to the increasing ceramic homogeneity of the Middle Preclassic and very different ceramic attributes. By 1 ,000 BCE there were sedentary, pottery producing communities throughout northern Belize and the rest of the southern Maya Lowlands, and the ceramics produced in these emerging Maya cities exhibit only minor shared attributes. They certainly appear to be m ore regionalized than subsequent time periods in the Middle and Late Preclassic. Swasey may be one of four coeval Early Preclassic southern Maya Lowland ceramic spheres: Swasey in Northern Belize, Xe in the Pasion River area, Eb in the Petn, and Cunil in the Belize Valley and the southeastern Petn. The identification of these four, separate yet coeval, ceramic spheres in the Early Preclassic supports the ceramic evidence of regionalization and perhaps only loosely connected, although growing, Maya villa ges and cities. Acknowledgements We would like to express our gratitude to Dr. John Morris and everyone at the Institute of Archaeology and the National Institute of Culture and History for their decades of support for our research efforts. The KaKabish Project, under the direction of Dr. Helen Haines received funding from the National Geographic Society Committee for Research and Exploration, and the Social Science and Humanities Research Council of Canada (SSHRC). We are particularly grateful for the work of Joshua Lockett Harris, John Baker, Alice Gomer, Dr. Jim Aimers, and Dr. Cara Tremain. The Cuello Project, under the direction of Dr. Norman Hammond, received major funding from the National Geographic Society Committee for Research and Ex ploration, The Trustees of the British Museum, The Research Council of Rutgers University, and support from numerous other institutions including Cambridge University, Boston University, and The University of Arizona. We are grateful to the many students, staff, and men from Yo Creek who collaborated on the Cuello excavations for more than four decades, and the former directors of the IA during those years: Joe Palacio, Dr. Elizabeth Graham, the late Harriot Topsey, and the late Win n el Branche who supporte d our research. References Adams, Richard E.W. 1971 The Ceramics of Altar de Sacrificios Papers of the Peabody Museum of Archaeology and Ethnology Vol. 63, No. 1. Harvard University, Cambridge, MA.

PAGE 135

The Early Preclassic Swasey Ceramic Sphere of Northern Belize 138 1999 Rio Azul. An Ancient Maya City University of Oklahoma Press. Norman, Oklahoma. Aimers, James 2012 Ceramic Report of the 2011 Field Season. In KaKabish Archaeological Research Project (KARP) Interim Report, edited by Helen R. Haines, pp. 7376. On file with the Institute of Archaeology, NICH, Belmop an, Belize. Andrews, E. Wyllys V, and Norman Hammond 1990 Redefinition of the Swasey Phase at Cuello, Belize. American Antiquity 55:570 -584. Ball, Joe W. 1976 Ceramic Sphere Affiliations of the Ba rton Ramie Ceramic Complexes. In Prehistoric Pottery Analy sis and the Ceramics of Barton Ramie by James C. Gifford, pp. 323-330. Memoirs of the Peabody Museum of Archaeology and Ethnology Vol. 18. Harvard University, Cambridge, MA. Chase, Arlen F and Diane Z. Chase 1987 Putting Together the Pieces: Maya Pott ery of Central Belize and Central Peten, Guatemala. In Maya Ceramics: Papers from the 1985 Maya Ceramic Conference, edited by Prudence M. Rice and Robert J. Sharer, pp. 4772. BAR International Series 345(i). British Archaeological Reports, Oxford. Clark, John E., and Michael Blake 1994 The Power of Prestige: Competitive Generosity and the Emergence of Rank Societies in Lowland Mesoamerica. In Factional Competition and Political Development in the New World edited by Elizabeth M. Brumfiel and John W. Fox pp. 17 -30. Cambridge University Press, Cambridge. Clark, John E., and Dennis Gosser 1995 Reinventing Mesoamerica's First Pottery. In The Emergence of Pottery, edited by W. K. Barnett and J.W. Hoopes, pp. 209221. Smithsonian Institution Press, Washington, DC. Coe, Michael D., and Richard A. Diehl 1980 In the Land of the Olmec 2 vols. University of Texas Press, Austin. Culbert, T. Patrick 1993 The Ceramics of Tikal: Vessels f rom the Burials, Caches and Problematical Deposits Tikal Report No. 25, Part A. The University Museum. University of Pennsylvania, Philadelphia. Fry, Robert E. 1989 Regional Ceramic Distributional Patterning in Norther n Belize: The View from Pulltrou ser Swamp. In Prehistoric Maya Economies of Belize edited by Patricia A. McAnany and Barry L. Isaac, pp. 91 111. Research in Economic Anthropology, Supplement 4. JAI Press, Greenwich, CT. Garber, James F., M. Kathryn Brown, Jaime J. Awe, and Christopher J. Hartman 2004 Middle Formative Prehistory of th e Central Belize Valley: An Examination of Architecture, Material Culture, and Sociopolitical Change at Blackman Eddy. In The Ancient Maya of the Belize Valley, Half a Century of Archaeological Research, edited by James F. Garber, pp. 2547. University Pre ss of Florida, Gainesville. Gifford, James C. 1976 Prehistoric Pottery and the Ceramics of Barton Ramie in the Belize Valley Memoirs of the Peabody Museum of Archaeology and Ethnology Vol. 18. Harvard University, Cambridge, MA. Gomer, Alice 2013 Humble Beginnings? A Study of the Origins of Settlement at the Site of KaKabish, Belize, through the Use of Macroscopic and Petrographic Analysis of Ceramics. Unpublished Masters Thesis, University College, London, England. Haines, Helen R. 2012 KaKabish: The Research Foundations Mapping Techniques, Terminology, and Ceramic Phase Names. In KaKabish Archaeological Research Project (KARP) Interim Report on the 2011 Field Season, edited by Helen R. Haines, pp. 13 18. On file with The Institute of Archaeology, NI CH, Belmopan, Belize. Haines, Helen R., Alice Gomer, and Kerry L. Sagebiel 2014 Plant thou no roses at my head: A Discussion of the Middle Formative Period Buried Deposits at KaKabish, North -central Belize. In Research Reports in Belizean Archaeology Vol. 11:337 346, edited by John Morris, Jaime Awe, Melissa Badillo, and George Thompson. Institute of Archaeology, NICH, Belmopan, Belize. Hammond, Norman 1984 Two Roads Diverged: A Brief Comment on "Lowland Maya Archaeology at the Crossroads." American Antiquity 49:821826. Hammond, Norman (editor) 1975 Archaeology in Northern Belize: British Museum -Cambridge University Corozal Project 197475 Interim Report Centre of Latin American Studies, Cambridge University, Cambridge. 1976 Archaeology in Northe rn Belize: Cambridge University Corozal Project 1976 Interim Report. Centre of Latin American Studies, Cambridge University, Cambridge. 1978 Cuello Project 1978 Interim Report Publication No. 1. Archaeological Research Program, Douglass College, Rutgers University, New Brunswick, NJ.

PAGE 136

Kosakowsky, Sagebiel, and Pring 139 1991 Cuello: An Early Maya Community in Belize Cambridge University Press, Cambridge. Hammond, Norman, Amanda Clarke, and Sara Donaghey 1995 The Long Goodbye: Middle Preclassic Maya Archaeology at Cuello, Belize. Latin Ame rican Antiquity 6(2):120-128. Hammond, Norman, Sara Donaghey, Colleen Gleason, Justine C. Staneko, Dirk Van Tuerenhout, and Laura J. Kosakowsky 1987 Excavations at Nohmul, Belize, 1985. Journal of Field Archaeology 14:257 -281. Hammond, Norman, Laura J. K osakowsky, Anne Pyburn, John Rose, Justine C. Staneko, Sara Donaghey, Mark Horton, Catherine Clark, Colleen Gleason, Deborah Muyskens, and Thomas Addyman 1988 The Evolution of an Ancient Maya City: Nohmul. National Geographic Research 4:474 495. Hammond, Norman, Duncan Pring, Richard Wilk, Sara Donaghey, Frank P. Saul, Elizabeth S. Wing, Arlene V. Miller, and Lawrence H. Feldman 1979 The Earliest Lowland Maya? Definition of the Swasey Phase. American Antiquity 44(1):92-110. Housley, Rupert M., Norman Hamm ond, and Ian A. Law 1991 AMS Radiocarbon Dating of Preclassic Burials at Cuello, Belize. American Antiquity 56(3):514519. Inomata, Takeshi, Daniela Triadan, Kazuo Aoyama, Victor Castillo, and Hitoshi Yonenobu 2013 Early Ceremonial Constructions at Ceibal Guatemala. And the Origins of Lowland Maya Civilization. Science 340: 467-471. Kosakowsky, Laura J. 1983 Intra -site Variability of the Formative Ceramics from Cuello, Belize: An Analysis of Form and Function. Ph.D. dissertation, University of Arizona. 1987 Preclassic Maya Pottery at Cuello, Belize Anthropological Papers of the University of Arizona No. 47. University of Arizona Press, Tucson. Laura Kosakowsky, Laura J. and Jon C. Lohse 2003 Investigating Multivariate Ceramic Attributes as Clues to An cient Maya Social, Economic, and Political Organization in Blue Creek, Northwest Belize. Research Report Submitted to the Ahau Foundation. On file with the Institute of Archaeology, NICH, Belmopan, Belize. Laura J. Kosakowsky, Laura J. and Duncan C. Pri ng 1998 The Ceramics of Cuello, Belize: A New Evaluation. Ancient Mesoamerica 9(1):5566. Law, Ian A., Rupert A. Housley, Norman Hammond, and Robert E.M. Hedges 1991 Cuello: Resolving the Chronology Through Direct Dating of Conserved and Low -Collagen Bon e by AMS Radiocarbon 33(2): 1 -19. Lockett -Harris, Joshua J. 2013 Middle Formative Foundations in Group D: Buried Platform, Mortuary Offerings, and Ritual Paraphernalia. In KaKabish Archaeological Research Project (KARP): Interim Report on the 2012 Fie ld Season, edited by Cara G. Tremain and Helen R. Haines, pp. 41 63. On file with The Institute of Archaeology, NICH, Belmopan, Belize. 2014 2013 Plaza Excavations: SouthPlaza D Operation 8 Units 4, 5, 6, 7, and 8. In KaKabish Archaeological Research Project (KARP): Report on the 2013 Field Season, edited by Helen R. Haines, pp. 13 36. On file with The Institute of Archaeology, NICH, Belmopan, Belize Lohse, Jon C. 2010 Archaic Origins of the Lowland Maya. Latin American Antiqui t y 21(3):312352. Lo pez-Varela, Sandra 1996 The Kaxob Formative Ceramics: The Search for Regional Integration Through a Reappraisal of Ceramic Analysis and Classi fication in Northern Belize. Ph. D Dissertation, University College, London. Lowe, Gareth W. 1975 The Early Preclassic Barra Phase of Altamira, Chiapas: A Review with New Data Papers No. 38. New World Archaeological Foundation, Brigham Young University, Provo, UT. 1978 Eastern Mesoamerica. In Chronologies in New World Archaeology, edited by R.E. Taylor and C.W. Meighan, pp. 331 393. Academic Press, New York. Marcus, Joyce 1983 Lowland Maya Archaeology at the Crossroads. American Antiquity 48:454 -482. 1984 Reply to Hammond and Andrews. American Antiquity 49:829833. McDow, David A. 1997 An Analysis of the Whole Ceramic Vessels from Kichpanha, Northern Belize. Undergraduate Honors Thesis, Department of Anthropology, University of Texas, Austin. Potter, Daniel R., Thomas R. Hester, Stephen L. Black, and Fred Valdez 1984 Relationships Between Early Preclass ic and Early Middle Preclassic Phases in Northern Belize:

PAGE 137

The Early Preclassic Swasey Ceramic Sphere of Northern Belize 140 A Comment on "Lowland Maya Archaeology at the Crossroads." American Antiquity 49:628-631. Pring, Duncan C. 1977 The Preclassic Ceramics of Northern Belize. Ph.D. dissertation. University of London. University Microfilms, Ann Arbor, Michigan. Pring, Duncan C, and Norman Hammond 1982 The Stratigraphic Priority of Swasey Ceramics at Cuello, Belize. Ceramica de Cultura Maya 12:4348. Temple University, Phil adelphia. Rice, Prudence M. 1979 Ceramic and Nonceramic Artifacts of Lakes Yaxha -Sacnab, El Peten, Guatemala. Part I. The Ceramics. Ceramica de Cultura Maya 10:1-36. Temple University, Philadelphia. Robin, Cynthia 1989 Preclassic Maya Burials at Cuello, Belize BAR International Series 480. British Archaeological Reports, Oxford. Rosenswig, Robert M. 2008 San Estevan Archaeological Project 2005. Institute of Mesoamerican Studies Occasional Publication No. 15. State University of New York, Albany. Sabloff, Jeremy A. 1975 Excavations at Seibal Department of Peten, Guatemala: Ceramics Memoirs of the Peabody Museum of Archaeology and Ethnology Vol. 13, No. 2. Harvard University, Cambridge, MA. Sagebiel, Kerry L. 2005 Blue Creek Regional Political Ecology Project Ceramic Report: 2004 Season. In 2004 Season Summaries of the Blue Creek Regional Political Ecology Project, Upper Northwestern Belize, edited by Jon C. Lohse, and Kerry L. Sagebiel, pp. 16 42. Maya Research Program, Ft. Worth, Texas, and the University of Texas at Austin, Fort Worth, Te xas. On file with the Institute of Archaeology, NICH, Belmopan, Belize. Sagebiel, Kerry L. and Helen R. Haines 2015 Never Ending, Still Beginning: A New Examination of the Ceramics of KaKabish, Belize. In Research Reports in Belizean Archaeology, Papers of the 2014 Belize Archaeology Symposium, Vol. 12 pp. 358366. Edited by John Morris, Melissa Badillo, Sylvia Batty, and George Thompson. Institute of Archaeology, NICH, Belmopan, Belize. Sharer, Robert J. 1978 Pottery and Conclusions. In The Prehistory o f Chalchuapa, El Salvador vol. 3, edited by Robert J. Sharer, pp. 1 226. The University Museum, University of Pennsylvania, Philadelphia. Smith, Robert E. 1955 Ceramic Sequence at Uaxactun, Guatemala. Vols. I and II. Middle American Research Institute, P ublication No. 28. Tulane University, New Orleans, LA. Smith, Robert E., and James C. Gifford 1966 Maya Ceramic Varieties, Types, and Wares at Uaxactun: Supplement to Ceramic Sequence at Uaxactun, Guatemala . Middle American Research Institute Publicati on No. 28:125-74. Tulane University, New Orleans, LA. Sullivan, Lauren and Jaime J. Awe 2013 Establishing the Cunil Ceramic Complex at Cahal Pech, Belize. In Ancient Maya Pottery. Classification, Analysis and Interpretation edited by James John Aimers, pp.107-120. University Press of Florida, Gainesville. Sullivan, Lauren and Kerry L. Sagebiel 2003 Changing Political Alliances in the Three Rivers region. In Heterarchy, Political Economy, and the Ancient Maya edited by Ver non L. Scarborough, Fred Valdez Jr., and Nicholas Dunning, pp. 2536. The University of Arizona Press, Tucson. Valdez, Fred, Jr. 1987 The Ceramics of Colha, Northern Belize. Ph.D. dissertation, Harvard University. University Microfilms, Ann Arbor, Michigan. Valdez, Fred Jr., and Brett A. Houk 2000 The Chan Chich Ceramic Complexes. In The 1998 and 1999 Seasons of the Chan Chich Archaeological Project, edited by Brett A. Houk, pp. 127140. Papers of the Chan Chich Archaeological Project Number 4. Me soamerican Archaeological Research Laboratory. The University of Texas, Austin. Walker, Debra S. 2014 Notas sobre la Secuencia Cermica de Yaxnohcah. In Proyecto Arqueolgico Yaxnohcah, 2014, edited by Kathryn Reese Taylor. Informe de la Tercera Temporada de Investigaciones. University of Calgary, Alberta. Willey, Gordon, T. Patrick Culbert, and Richard E.W. Adams 1967 Maya Lowland Ceramics: A report from the 1965 Guatemala City Conference. American Antiquity 32(3): 289-315.

PAGE 138

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 141 151 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 13 THE DEVELOPMENT OF TERMINAL PRECLASSIC AND EARLY CLASSIC ROYAL ARCHIT ECTURE AT CHAN CHICH, BELIZE Toms Gallareta Cervera, Brett A. Houk and Claire Novotny The archaeological site of Chan Chich is the southernmost and the second largest Maya city in the Belizean portion of the Three Rivers adaptive region. Excavations at the Upper Plaza, located in the sites center, have yielded evidence of a continuous occupation that dates from the Middle Pr eclassic to the Late Classic period and includes a Terminal Preclassic royal Maya tomb. Excavations in this area during the 2017 season yielded evidence of different types of elite architecture, such as a long platform, dated to 400 BC, and a funerary cry pt containing a probable royal burial dated to the Early Classic period. The results of our excavations give us rich information about Chan Chichs transition from a small village to an early Maya kingd om. Introduction Chan Chich is the southernmost Maya city in the Belizean portion of the Three Rivers adaptive region, which spans portions of Belize, Mxico, and Guatemala (Dunning et al. 1998; Garrison and Dunning 2009). The Ro Bravo, Booth's River, and Ro Azul/Ro Hondo and th eir watersheds define the region and encompass over a dozen large sites including Chan Chich, Dos Hombres, and La Milpa in Belize, and San Bartolo, Xultun, La Honradez, and Ro Azul in Guatemala ( Figure 1). The Guatemalan half of the region was home to so me spectacular Preclassic developments including a royal tomb, dating to 150 BC, and elaborate polychrome murals, perhaps 50 years younger, at San Bartolo (Saturno 2006:73). Although smaller than the largest centers in the western portion of the region, by the Late Classic period Chan Chich was the second largest site in the eastern half of the Three Rivers adaptive region, trailing only La Milpa in monumental area (Houk 2015: Table 10.1). The monumental core of the site is centered on a 350m long, northsouth line of contiguous plazas on a broad hill overlooking Chan Chich Creek ( Figure 2). The architectural center of the site is arguably Structure A 1, a large tandem range building that divides the Main Plaza from the Upper Plaza, separating public space from private space. Elevated approximately 7 m above the Main Plaza, the Upper Plaza constitutes an elevated acropolis or palace group with two large temple pyramids, attac hed lateral courtyards, and a commanding view of the Main Plaza from the central landing and eight once vaulted rooms that face north from Structure A 1. Accessed by stairways on either side, the Figure 1 Map of northwestern Belize and Three Rivers adaptive region showing the locations of major Maya sites. central landing on Structure A 1 is the only formal entryway into the Upper Plaza. During the first season of excavations at Chan Chich in 1997, the Chan Chich Archaeological Project encountered a Terminal Preclassic royal Maya tomb in the Upper Plaza at the site (Houk et al. 2010). That discovery, coupled with subsequent excavations of early Middle Preclassic floors and features, promot ed additional excavations in the Upper Plaza to investigate the earliest settlement of the site and the subsequent transformation of a small village into the seat of power for an early Maya king. In this paper, we summarize our evolving understanding of t he Preclassic foundations of Chan Chich and present our newest data on the

PAGE 139

Royal Architecture at Chan Chich 142 Figure 2 Map of Chan Chich. Terminal Preclassic and Early Classic royal architecture at the site. History of Excavations During the first three seasons of CCAP excavations (19971999), Hubert Robichaux (2000) directed investigations in the Upper Plaza, documenting looters trenches, excavating chronological test pits, exposing collapsed rooms on Structure A 1, and studying the final a rchitectural phases of Structure A 13. A major focus of Robichauxs work, however, particularly during the 1997 seasons, involved documenting a collapsed royal tomb, which Robichaux discovered during the course of test pit excavations in front of Structur e A 15, the large temple pyramid on the southern side of the Upper Plaza (Houk et al. 2010). The Upper Plaza has been a primary area of interest over the past six seasons, and the 2016 and 2017 seasons in the Upper Plaza, part of a three year Alphawood Fou ndation grant, specifically set out to study the development of the royal acropolis and its dynastic architecture subsequent to the establishment of a royal dynasty at the site ca. AD 200250 and to examine how architecture reflects the evolving relations hip between political organization (i.e., divine kingship) and monumental construction (Houk 2016a:6). The investigations included additional stratigraphic excavations, broad horizontal exposures of buried architectural features, and a robust program of radiocarbon dating. While the 2016 and 2017 investigations focused on deposits in the plaza, the planned 2018 season include new excavations on the buildings bordering the plaza. In 2016 and 2017, the CCAP excavated chronological test pits in the center of the plaza, in the southeast corner of the plaza, in front of Structure A 13, at the base of Structure A 1, and in the southwestern courtyard at the base of Structure A 15 ( Figure 3 ). Combined with results from previous seasons, the data from

PAGE 140

Gallareta Cervera, Houk, and Novotny 143 Figure 3 Contour map of the Upper Plaza showing the 2017 excavations, Subop CC-15 -B from 2016, and the location of Tomb 2. these units provide a much more detailed chronology for the plaza s development. The most complicated and informative excavations, howe ver, constitute a block of units in the northern part of the plaza, which exposed the truncated platform of a buried temple and a later intrusive crypt, which contained the remains of potentially another royal individual. These discoveries are described b elow. The Middle Preclassic Community: Evidence from the North and East Upper Plaza It appears, although our excavation sample is limited in many areas of the site, that the first occupants of Chan Chich settled on the hilltop that is now buried by the Upper Plaza during the Middle Preclassic period, around 900 BC or slightly earlier. Occupation remained focused on this area for several centuries until the beginning of the Late Preclassic period when the small village expanded into areas now covered by the Main Plaza, Back Plaza, Western Plaza, and Normans Temple. Even with this expansion, however, the Upper Plaza remained the center of the village. Figure 4 Photograph of floor and surface and terrace of early version of Structure A -1 in Subop CC-17 -I. View to the north. The oldest radiocarbon dates come from the deepest floors in the north central part of the Upper Plaza and suggest the first occupants of the site settled there in the early Middle Preclassic period. The two samples, which came from floor fill above bedro ck and were collected over the course of two seasons from the same excavation unit, returned 2 sigma date ranges of cal 911 804 BC and cal 931 833 BC (Gallareta et al. 2017:Tables 2.2 and 2.3). The residents of the site gradually expanded their settlement on the hilltop throughout the Middle Preclassic period, and samples from the northcentral, northeast, and east parts of the plaza, as well as from below Structure A 1, yielded radiocarbon dates spanning cal 800 400 BC (Gallareta et al. 2017:Tables 2.2 an d 2.3). Thus far, our excavations have only documented plaster surfaces some of which are presumably plaza floors, while others may be platform surfaces that date to the Middle

PAGE 141

Royal Architecture at Chan Chich 144 Preclassic period, with one notable exception. Excavations in 2016 at the base of Structure A 1 documented an apparent buried structure (Lot CC 15 B 4), which two radiocarbon samples date to cal 766 540 BC and 749407 BC as described by Houk (2016a:11). This may be the earliest version of Structure A 1, which forms the northern edge of the Upper Plaza ( Figure 4). Hubert Robichaux (1998) encountered the same structure approximately 10 m to east and documented a thick plaster surface, which rolled down as a step or terrace. The 2016 and 2017 excavation s similarly documented a 3540 cm step or terrace, indicating the presence of a long structure that oversaw Chan Chichs Upper and Main Plaza areas from the Middle Preclassic. The lower surface exposed on this structure during the 2016 excavations is at th e approximate elevation of the modern plaza floor and was originally mistaken for the plaza floor. Below the surface, the 2016 excavations encountered 1.1 m of cobble/small boulder fill, which buried a well preserved plaster floor (Lot CC 15 B 9). Below this surface, excavations documented an additional five floors above bedrock, which lay approximately 2.5 m below the modern plaza surface in this area. An additional four radiocarbon dates, spanning the Middle Preclassic period, date this sequence, with the deepest sample from above the oldest floor, returning the oldest age range of cal 826 782 BC (Houk 2016a:Table 1.4). Excavations in 2012 in the northeast area of the Upper Plaza encountered a sequence of six floors above bedrock, which was 2.25 m below the modern plaza surface. The deepest deposits included an eroded plaster floor, which was possibly constructed to create a level surface over uneven bedrock. A single radiocarbon sample from the 1550 cm thick fill layer returned a range of cal 805569 BC (Houk 2016b:Table 7.10). Above this floor, the excavations revealed a 40 cm thick midden, which contained Swasey ceramics (Kelley 2014:56) and produced a single cal 799 to 766 BC date from charcoal (Houk 2106b:Table 7.10). Robichaux (1998) encountered this same midden near the base of Structure A 1 in 1997. Structure A 13 consists of a large mound located at the eastern section of the Upper Plaza A chronologybuilding test pit at the base of the Figure 5 East wall profile of Subop CC-15 -M. structure documented additional Middle Preclassic floors overlying bedrock ( Figure 5 ). The 2x 3 m unit yielded evidence of six plaster floors and ceramic materials that range from the Middle Preclassic period on the lower floors to the Late Classic perio d. At its lowest level, we observed evidence of a posthole ( Figure 6 ) with Mamom ceramics and two AMS dates (cal 554 411 BC and cal 644 552 BC) that bracket this ancient feature to the Middle Preclassic period. The area was later covered by a plaster flo or and a platform foundation made of carved stones, oriented east to west and dated to cal 762482 BC (Gallareta et al. 2017:Tables 2.2 and 2.3). The Late Preclassic City: Evidence of Royal Architecture in the Upper Plaza During the Late Preclassic period, the Upper Plaza expanded to south as documented in Subop CC15 Q in the southeast corner of the plaza and Subop CC15 L at the western base of Structure A 15and vertically with new floors and new structures. Subops CC15 Q and CC 15L yielded ev idence of two architectural features, possibly platforms: the oldest, in Subop CC 15 Q, dated to cal 358 278 BC and the latter to cal 236 185 BC (Gallareta et al. 2017:Tables 2.2 and 2.3). In the central part of the plaza, the

PAGE 142

Gallareta Cervera, Houk, and Novotny 145 Figure 6 Photograph of the Middle Preclassic posthole in bedrock at the bottom of Subop CC -15-M. Figure 7 Orthomosaic of the 2017 northern block excavations showing the plan of Blanca and Crypt 1 (top) and plan drawing of Blanca (bottom). Late Preclassic se quence began with a series of floors with thin layers of construction fill, which buried the Middle Preclassic floors. In the southern end of Subop CC15 A, an 11 m long trench, excavations documented six floors spanning the early Middle Preclassic into the Late Preclassic that predate the first documented structural feature in this part of the plaza an alignment of cut stone blocks constructed on an eroded plaster floor (Lot CC 15 A 7). This alignment of finely shaped and regular limestone blocks extends at least 22.5 m east west (Herndon et al. 2014:38) and has been exposed in multiple excavation units between 2012 and 2017. While we have been unable to date Lot CC 15 A 7, the floor upon which the feature rests, the next oldest floor (Lot CC15 A 8) returned a cal 767 434 BC date (Gallareta et al. 2017:Tables 2.2 and 2.3). Subsequent to the construction of the alignment, the Maya raised the plaza floor to the southcomparable floors are not found on the north side of the alignment. The first flo or was plaster, like those that preceded it, but the second floor was a compact dirt surface that apparently extended over much of the plaza area south of the alignment and elevated the plaza floor to the same elevation as the top of the alignment (Kelley 2014). This floor measured 20cm thick and was constructed during the Late Preclassic or Terminal Preclassic period based on a date of cal 204 96 BC from a sample obtained in 2014 (Houk 2016b:Table 7.10) and a date of cal AD 128236 from a sample collecte d in 2016 (Houk 2016:Table 1.5). Combined, these dates bracket the construction of the alignment, suggesting it was built near the end of the Late Preclassic period. Our current interpretation of this alignment is that it was a step or platform associated with a buried substructural platform nicknamed Blanca and described below. Blancas Construction Sequence and its Relation to the Upper Plaza Excavations of the 2017 season revealed the presence of a buried, truncated platform in the northern section of the Upper Plaza, south of Structure A 1 ( Figure 7). The structure base was made with large rectangular, white blocks of cut limestone, which were slightly inclined inwards excavators nicknamed the structure

PAGE 143

Royal Architecture at Chan Chich 146 Blanca because of the white stones. The uncove red section of the structure measured 8.75 m east west by 4.20 m northsouth and we know the structure continues to the east and north, beyond our excavation block. Blancas form is complex and its partial dismantling obscures its final Late Preclassic configuration. The portion we exposed consists of two, possibly three, tiers with a projecting front axial outset, which would have measured 4.5 m wide but was partially destroyed by subsequent construction (discussed below). The overall shape is rectan gular with rounded corners. The axial outset is battered, while the other faces are not. The two tiers are low enough to possibly function as steps. Additionally, a stone alignment located to the north of the basal body is interpreted as a possible thir d tier of the platform. The plaster floor in front of Blanca suggests that the northern plaza was repaved at least three times while the structure was in use. The three plaster floors were very close together, only separated by thin layers of fill. Ceram ics recovered above the last floor associated to Blanca are from the Mamom (600400 BC) and Chicanel (400 BCAD 150) spheres. Ceramics from the inside of Blanca were mostly Chicanel (400 BCAD 150) types. Based on architectonic style and associated ceram ic materials we suggest that Blanca was constructed around 400 BC. Before the Early Classic, and probably around the Terminal Preclassic period, Blanca was dismantled or chopped and buried under a massive renovation, which apparently elevated the Upper P lazas floor to its modern level in the northern part of the plaza. An intrusive primary burial was placed on top of Blanca after the structure was already in filled and covered by the main Upper Plaza floor, providing a terminus ante quem for Blancas de struction. Burial CCB17 consisted of a prone and extended individual oriented north south with his hands on top of his pelvis. The cranium was covered with an inverted large Society Hall Impressed bowl dated to the Late Preclassic period (Gallareta Cerv era et al. 2017). It is unclear if the individual was buried in a prepared cist or deposited as a simple burial; although we noticed three rough stones to the west of the burial, we cannot say conclusively that any funerary architecture or prepared surfaces were present. It is also unclear if the Late Classic plaza floor was broken in order to inter this burial or if the floor was constructed above this individual. However, a single radiocarbon date obtained from a piece of bone from the burial returned a 2sigma age range of cal 154 BCAD 47 (Gallareta Cervera et al., 2017, Tables 2.2 and 2.3 ). Terminal Preclassic, Tomb 2 The construction of Tomb 2 and its capping shrine is the next documented significant construction event in the Upper Plaza. Located in the southern part of the plaza, north of Structure A 15, Tomb 2 occupied a portion of the plaza south of the major construction that buried Blanca. The elliptical tomb chamber spanned 3.25 m by 0.8 m and cut 1.15 m into bedrock. Placing the tomb in be drock required cutting through a series of four older floors. Kelley (2014), based on subsequent excavations east and north of the tomb, suggested the youngest floor cut by the tombs construction was the compact dirt surface described above and documente d in multiple locations in the southern and central areas of the plaza, meaning the tomb was constructed after cal AD 128236. This assessment aligns with the ceramic data, which suggest an approximate date of AD 200350 for the vessels in the tomb (Houk et al. 2010). Twelve large limestone capstones sealed the chamber; these, in turn, were buried beneath rubble fill and an apparent low shrine platform (Houk et al. 2010:232 233). In addition to the 11 ceramic vessels, the tomb contained the deteriorated remains of an adult male, several poorly preserved organic artifacts, and four jade jewelstwo ear spools, a tubular bead, and a helmet bib head pendant (Houk et al. 2010). The latter diadem particularly when combined with the elaborateness of the tombs construction, the tombs location, and the diversity of grave goods indicates the individual buried in Tomb 2 was an early king of Chan Chich (Houk et al. 2010). The Early Classic Acropolis: Excavations of the Upper Plaza Chamber Excavations at the north of the Upper Plaza between Structure A 1 and Tomb 2

PAGE 144

Gallareta Cervera, Houk, and Novotny 147 Figure 8 Upper Plaza crypt after excavation. View to the north. discovered an intrusive, rectangular chamber, oriented northsouth that sliced through Blancas platform face and several underlying floors ( Figure 8). The chambers builders re used a Middle Preclassic floor as the chambers floor. The chambers four walls were heterogeneous in style and construction techniques. The chamber measured 1.60 m east west by 2.3 m northsouth, and the walls were preserved to a height of 1.25 m. The northern wall was made of large and nicely carved rocks covered with stucco and faint t races of red paint remaining. Additionally, the north wall had a possible step composed of two large, semi carved stones of the same size and shape. A small, carved stone directly to the south of this wall might have been used as a step into the chamber. The east and west walls were different; both were made of small, uncut, and roughly faced rocks. The southern wall is the most unusual and consists of two parts. The lower part is a layer of compact soil between the chamber surface and a plaster floor, not a formally constructed wall. The upper part of the wall consisted of roughly shaped stones placed on top of this stucco floor. The chamber floor rolled up onto the dirt fill at the south of the chamber, suggesting that this was the original construc tion technique used to build the chamber. The upper preserved courses on the western and southern walls were apparent vault stones many of the stones on the western walls upper course were inadvertently removed during excavations, before the field crew re cognized the chamber as a constructed feature but the preserved examples jut into the chamber. This suggests that the chamber was originally a sunken, vaulted room, accessed via stairs on the north. If so, the vault would have risen above the level of the Plaza Floor. However, after a period of use, the vault was destroyed and the chamber filled to the level of the plaza as discussed below. All four walls had evidence of deteriorated stucco plasteri ng. Samples recovered from under the chambers re used floor suggest it was constructed and between cal 796 to748 BC, and a charcoal sample from the floor surface yielded a date of cal AD 237 333 (Gallareta Cervera et al. 2017: Tables 2.2 and 2.3). Although we are unsure of the original function of the chamber, the Maya used it as a crypt prior to filling it, and we have designated the feature Crypt 1. Excavations in the chamber yielded the remains of at least two individuals located in the southern half of the chamber directly on the surface floor. The center and northern portions of the crypt did not yield any cultural remains. One of these individuals (Burial CCB16B) was articulated, primary, and extended, with its head to the east and its feet, cros sed at the ankle, to the west ( Figure 9; Novotny et al. 2017). A bone sample dates Burial CCB16B to cal AD 247353 (Gallareta Cervera et al. 2017: Tables 2.2 and 2.3). A funerary offering of an Ixcanrio Orange Polychrome pedestal bowl dated to the Termi nal Preclassic or Early Classic period was associated with Burial CC 16B. When buried, the individual was wearing two Spondylus shell ear flares and a serpentine helmet bib head pendant associated with rulership and a possible heirloom from the Late Precl assic period (see Houk et al. 2010) as funerary regalia. Although similar in style to the diadem from Tomb 2, the helmet bid head pendant from Crypt 1 is thinner, less well crafted, and of less exotic raw material. Ceramics in the chamber fill and surrou nding the burial consisted mostly of Tzakol sherds with some Chicanel sherds, suggesting that this context dates to the Early Classic period. Burials CCB16A, B16C and B16D consisted of clusters of disarticulated bone fragments belonging to adults locate d at the south end of the crypt. Burial CCB16A consisted of bones of the left foot, an articulated

PAGE 145

Royal Architecture at Chan Chich 148 Figure 9 Photograph of Burial CC-B16B during the 2017 season, view to the south. The remains recorded as Burial CC-B16C are visible at the right edge of the photograph, north of Burial CC -B16Bs feet, and the remains recorded as Burial CC-B16D include the cranium and long bones visible south of Burial CC-B16Bs lower legs. Burial CCB-16A is not in this photograph as it was excavated in 2016. right leg, and an articulated right wrist and hand (Novotny et al. 2016). Burial CCB16C was a cluster of bones located adjacent to the feet of Burial CCB16B, approximately 10 cm to the south of Burial CCB16A. Burial CCB16D was immediately south of the lower legs of Burial CCB16B and comprised a cranium stacked on top of a pile of long bones in the southwestern corner of the crypt. Novotny and colleagues (2017) suggest that the best explanation for the burial location within the chamber is that Burial CC B16A was interred first, perhaps in a flexed position given the posit ion of the right leg, and subsequently disturbed by the interment of Burial CCB16B before decomposition was complete. Burials CC B16A, B16C and B16D may be the displaced remains of the same person (Novotny et al 2017). At some point after Burial CC B1 6B was interred, the Maya destroyed the crypts vault and filled the chamber with large, medium, and small boulders and sediment, before covering it with the final floor of the Upper Plaza. The fill in the northern part of the chamber yielded higher artif act densities as well as evidence of burning, approximately 65 cm above the floor in the room. The nature of this event is unclear, but charcoal from the deposit yielded a date of cal 55 BCAD 211 (Gallareta Cervera et al. 2017: Tables 2.2 and 2.3). Char coal recovered from beneath possible capstones in the fill returned a date range of cal AD 87227. Ceramics from the chambers context are mixed, yielding a mix of Early Classic and Late Preclassic types. Despite these two Terminal Preclassic dates, six other samples from the crypt largely date to the Early Classic period

PAGE 146

Gallareta Cervera, Houk, and Novotny 149 (Gallareta Cervera et al. 2017: Tables 2.2 and 2.3). Discussion Middle Formative Community Our excavations indicate that around 900 BC, during the Middle Preclassic period, the community of Chan Chich occupied and gave meaning to the landscape through the construction of formal architecture in the Upper Plaza. Stratigraphic evidence suggests the construction of multiple plaza floors in the north, middle, and east portions of the Upper Plaza, as well as possible public buildings made of vernacular architecture at the plazas north and east edges. Unfortunately, most evidence of Middle Preclassic architecture comes from plaster floors and platform surfaces. However, evidence in the north of the plaza reveals evidence of a substructure that predates Structure A 1 and dates to cal 766 540 BC and 749407 BC (Houk 2016a:11). Based on previous and current excavations we estimate that this Middle Preclassic version of the structure extended at least 15 m east west along the north edge the plaza. This long structure oversaw the Middle Preclassic landscape from a perch at the edge of the Upper Plazas hilltop. An early version of Structure A 13 suggests the use of platforms made of large rectangular carved stones and postholes likely used to sustain a wood and thatch superstructure. Robichaux (1999:34) also documented a Middle Preclassic posthole south of Structure A 1, further suggesting the use of this vernacular architectural style at the Upper Plaza. Moreover, Robichaux (1999:37) also suggests that the size and location of the feature might indicate that the structure was quite large and public in function. The five to six floor renovations and the raising of the plaza surface 2.25 m above its natural level during the Middle Preclassic period suggests a high degree of community cooperation and organization. Late and Terminal Pr eclassic Village Turned Kingdom The Late Preclassic period at Chan Chichs Upper Plaza was a time of political growth and architectural expansion. The plaza expanded to the south with the construction of platforms and stone buildings, suggesting a growth of political relevance and consolidation of power by elite members. The f loor sequence in the northern Upper Plaza is complex. Excavations have yielded a sequence of eight Preclassic plaster floors associated with Structure A 1 in Subop CC 15 B. On the western corner of Blanca, we uncovered three plaster floor levels, but we did not reach bedrock, which suggests that there might be more floors underneath this area. The excavation of Blanca, an 8.75m long platform made with large rectangular, cut blocks of white limestone and with round corners constructed around 400 BC, gives us a small window to explore the stratigraphic complexity of the Late Preclassic and Termin al Preclassic periods at Chan Chich. We know that the southern face of Blancas axial outset was slightly battered, sloping inward, and consisted of cut and regular limestone blocks. At its eastern and western edges there might have been two steps or ter races in addition to the outsets platform face. We also know that this substructure continues to the north and to the east, but its final form is unclear. The floor at Blancas base was renovated at least three times during the Late Preclassic period and it is likely contemporaneous with the Late Preclassic platform buried beneath Structure A 1 (see Houk 2016). At this time, this early version of Structure A 1 had been actively used as the northern edge of the Upper Plaza for around 500 years. Excavatio ns also suggest that Blanca was chopped and buried under the plaza floor in the Late Preclassic and further dismantled to accommodate a c rypt during the Early Classic. By the end of the Terminal Preclassic period, the Upper Plaza apparently housed an ear ly divine king. This individuals tomb was placed in the southern part of the plaza sometime around AD 250, based on ceramic and stratigraphic data. This early king indicates the elite had further consolidated their power and ruled a small kingdom from t he royal architecture of the Upper Plaza. Early Classic Funerary Crypt Excavations of the intrusive chamber, which cut through part of Blanca, suggest that

PAGE 147

Royal Architecture at Chan Chich 150 the chamber: a) is in fact an improvised crypt, b) dates to the Early Classic period, c) housed multiple elite burials, and d) was probably re entered during its use. Architec tur ally, the unit consisted of four walls that formed a rectangular chamber with its longest side oriented northsouth. We also identified large cut stones which we interpret as vault stones in the south, east and center portions of the chamber, although these were floating in fill within the chamber. The four masonry walls from thi s chamber were not homogeneous, and their construction cut through Blanca, the Late Preclassic period platform. This can be observed on the east and west walls and particularly on the southern wall, which was built on top of a floor level before reaching the crypt floor. The northern wall was made of carved stones and originally was covered with a layer of stucco. On top of this wall, we found at least two steps, which sugge st an entrance to the chamber. Crypt 1 had two burials: one primary and one secondary. The primary individual (Burial CC 16B) had associated funerary items and elite markers that identify it as a high elite, or perhaps royal, individual. Burial CC B 16A was interred in the middle of t he chamber, and then disturbed sometime later when Burial CCB 16B was interred. During the Early Classic period, the chamber vault was removed and the crypt was filled with a silty matrix and medium to large, uncut stones. There are some interesting para llels between Tomb 2 and Crypt 1. Both consisted of elaborate funerary architecture protecting the remains of individuals with objects associating them to the high elite. The crypt, however, consisted of a reused funerary space and was hence less exclusi ve than Tomb 2. Funerary paraphernalia and grave goods were also less prominent in the crypt Burial CCB16B only had one vessel, a serpentine jade object, and Spondylus ear flares as grave goods. However, the inclusion of a helmet bib head pendant, a possible heirloom from the Late Preclassic period, indicates that the burials belonged to a set of powerful individuals, perhaps even a divine king from the Early Classic period. Analysis by Novotny e t al. (2017) reveals that there were a minimum of two individuals, all adults and possibly males, present at the crypt. The location of the crypt, an almost mirror of Tomb 2 but at the northern portion of the Upper Plaza, and its proximity to Structure A 1, a monumental structure used since the Middle Preclassic period, associates this burial with regal activities. We suggest that Burial CCB16 consists of members of the Chan Chich royal court who were in office at the beginning of the Early Classic perio d and that the chamber, an elite crypt, function as an exclusive burial space for important members of the royal court. Conclusions Although our excavations have not yet targeted the structures surrounding the Upper Plaza to any great degree yet, the pict ure emerging from the excavations within the plaza itself shows the transformation of a small Middle Preclassic village into a small kingdom by the Terminal Preclassic period. The rulers used the plaza as a royal necropolis for an early king and a likely successor in the Terminal Preclassic and Early Classic periods, respectively. Through time, the ruling family radically modified the Upper Plaza, truncating and burying an early temple and creating, and then filling, a royal crypt, as they enhanced the mo numentality of their small kingdom Future excavations in the Late Classic structures surrounding the plaza should add clarity to this picture. Acknowledgments The authors would like to thank the Institute of Archaeology for permission to conduct research in the CCAP/BEAST permit area, as well as Mr. Michael Bowen and Mr. Alan Jeal (Gallon Jug Ranch) and Mr. Jeff Roberson and Mr. Alex Finkral (Yalbac and Laguna S eca Ranches) for granting us permission to work on their properties. Finally, this work would not have been possible without funding from Alphawood Foundation. References Gallareta Cervera, Toms, Brett A. Houk, Paisley Palmer 2017 The 2017 Investigations in the Upper Plaza at Chan Chich, Belize. In The 2017 Season of the Chan Chich Archaeological Project edited by Brett A. Houk and Claire Novotny, pp. 33 68. Papers of

PAGE 148

Gallareta Cervera, Houk, and Novotny 151 the Chan Chich Archaeological Project, Number 12. Department of SASW, Texas Tech University, Lubbock. Herndon, Kelsey E., Gregory Zaro, Brett A. Houk, Samantha Mitchell, and Erica Gallis 2014 The 2014 Excavations of the Chan Chich Dynastic Architecture Project. In The 2014 Season of the Chan Chich Archaeological Project edited by Brett A. Houk, pp. 31 68. Papers of the Chan Chich Archaeological Project, Number 8. Department of SASW Texas Tech University, Lubbock. Houk, Brett A. 2015 Ancient Maya Cities of the Eastern Lowlands. University Press of Florida, Gainesville. 2016a An Introduction to the 2016 Season of the Chan Chich Archaeological Project and the Belize Estates Archaeological Survey Team. In The 2016 Season of the Chan Chich Archaeological Project edited by Brett A. Houk, pp. 1 38. Papers of the Chan Chich Archaeological Project, Number 11. Department of SASW, Texas Tech University, Lubbock. 2016b The Chan Chich Archaeological Project: 1996 to 2016 Project Lists. In The 2016 Season of the Chan Chich Archaeological Project edited by Brett A. Houk, pp. 171 188. Papers of the Chan Chich Archaeological Project, Number 11. Department of SASW, Texas Tech University, Lubbock. Houk, Brett A., Hubert R. Robichaux, and Fred Valdez, Jr. 2010 An Early Royal Maya Tomb from Chan Chich Belize. Ancient Mesoamerica 21:229 248. Kelley, Krystle 2014 Establishing the Acropolis: Two Seasons of Investigations in the Upper Plaza of Chan Chich, Belize. Unpublished MA thesis, Department of SASW, Texas Tech University, Lubbock. Novotny, Anna, Ashley Booher, and Valorie V. Aquino 2016 Bioarchaeological Analysis of Human Skeletons from Chan Chich, Belize: The 2016 Field Season. In The 2016 Season of the Chan Chich Archaeological Project edited by Brett A. Houk, pp. 65 78. Papers o f the Chan Chich Archaeological Project, Number 11. Department of SASW, Texas Tech University, Lubbock. Novotny, Anna, Toms Gallereta Cervera, Briana Smith, and Gertrude Kilgore 2017 Bioarchaeological Analysis of Human Skeletons from Chan Chich Belize: The 2017 Field Season. In The 2017 Season of the Chan Chich Archaeological Project edited by Brett A. Houk and Claire Novotny, pp. 143 156. Papers of the Chan Chich Archaeological Project, Number 12. Department of SASW, Texas Tech University, Lu bbock. Robichaux, Hubert R. 2000 Looking Down on the Public: The 1999 Excavations on the Upper Plaza. In The 1998 and 1999 Season of the Chan Chich Archaeological Project edited by Brett A. Houk, pp. 57 70. Papers of the Chan Chich Archaeological Project, Number 4. Mesoamerican Archaeological Research Laboratory, The University of Texas at Austin. Saturno, William A. 2006 The Dawn of Maya Gods and Kings. National Geographic Magazine 209:68 77.

PAGE 149

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 153 158 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 14 EARLY MAYA CIVILIZAT ION IN THE THREE RIV ERS REGION Fred Valdez, Jr. Several prehistoric Maya sites in the Three Rivers Region of NW Belize and NE Peten, Guatemala are considered for their developments from the Preclassic to the Early Classic. Architecture, ceramics, and burials are among the data for interpretations of changes in Maya Civilization for these early periods. In NE Peten, Rio Azul excavations revealed a rich trajectory of developments from the Preclassic (ca 400 BC) into the Early Classic including the practice of painted tombs. La Milpa, Chan Chich, and Dos Hombres, in NW Belize, provide significant data for early tombs in the general region. Broader connections and comparisons to neighboring areas as wel l as more distant contacts or influences are presented and relayed to social, political, and economic interaction(s). Introduction The transition from the Maya Late Preclassic (400 BCAD 250) to the Early Classic (AD 250600) saw a dramatic change in population settlement, material culture, and political organization. As is always the case, change for any society is a turbulent and d ifficult process. Several prehistoric Maya sites in the Three Rivers Region (TRR) of NW Belize and NE Peten, Guatemala are considered for their developments/changes from the Late Preclassic to the Early Classic (Valdez and Scar borough 2013; Figures 1 and 2 ). A few comments on the Three Rivers Middle Preclassic serve as a means of giving context to the Late Preclassic of the region. While there are indications of Middle Preclassic occupants in the TRR, it is most significantly represented on the Guatemala side of the border with a decorated structure, G 103 sub 2, at the Maya site of Rio Azul ( Figure 3 ). The decoration of the structure consists of a central panel with a U element at center and two dots/circles, one above and one below the U. This centr al panel is flanked by large Jscrolls that may represent a skyband ( Figure 4 ). The entire panel is incised onto a mattefinished thick plaster. The region generally, however, lacks a significant Middle Preclassic settlement yet discovered. The Late P reclassic, therefore, demonstrates a broad and significant occupation across the entire region. On the western portions of the property, Rio Azul and El Pedernal serve as models of major Late Preclassic activity while Dos Hombres, La Milpa, Medicinal Trai l, and Chan Chich provide equally important data along the eastern and southern areas. I will use settlement Figure 1 Map of the Three Rivers Region (TRR). data, architecture, ceramics, and burials as descriptive markers for Maya occupation and activity. These material culture aspects also help elucidate the similarities and especially the differences from the Late Pre classic into the Early Classic. The Early Classic traditions in the TRR are represented by developments at Rio Azul, Dos Hombres, La Milpa, and the Barba Group. Interestingly, there are two significant phases in the region, 1) an early Early Classic seemingly represented by Maya material culture and 2) a late Early Classic demonstrating stylistic

PAGE 150

Early Maya Civilization in the Three Rivers Region 154 Figure 2 Map of Select Site s in the TRR. Figure 3 Map of G -103 Location at Rio Azul. interests and perhaps interaction with Teotihuacan. Late Preclassic Certain features and attributes of the Late Preclassic in the Three Rivers Region are quite Figure 4 Stucco Panel of G -103 -sub 2. representative of most of the Maya. The discussion here will focus on a few ceramics (types and forms), architecture (both dom estic and ritual), and burials. Similar ceramics across the Maya region include Sierra Red and Society Hall types (among many others), while spouted jars (or chocolate pots) and basins are among familiar Late Preclassic forms. It is rather fascinating just how similar several types seem to be across broad Maya regions. The ceramic similarities are so close in many cases that pottery from one site could be easily lost within a collection from another site. An additional intriguing development is the technological changes in Maya pottery from the early part of the Late Preclassic to material of the very Late or Terminal Preclassic. By the Terminal Preclassic forms become more complex, slips are harder and leaning towards a glossy finish, and polychrome is formally introduced before the P reclassic is complete. Maya architecture of the Late Preclassic has several representative forms and diagnostic features. Architecturally, a common structure shape/form includes round (and/or keyhole shaped) platforms. This architectural platform type h as been excavated in the Belize side of the TRR at Medicinal Trail (Hyde 2015; Figure 5), at Dos Hombres (Houk 1996), and at El Pedernal (Hendon 1989) in Guatemala. It remains uncertain, however, if these platforms supported domestic structures or building of more ritualistic interests. Of significant ritual and/or community interests were monumental constructions as discovered with G 103 at Rio Azul (Valdez

PAGE 151

Valdez, Jr. 155 Figure 5 Late Preclassic Round Platform at Medicinal Trail. Figure 6 Late Preclassic Mo numental Architecture at Rio Azuls G-103. 1995; Figure 6) and Structure 27 at La Milpa (Trein 2017). These building s provide some of the best evidence for coordinated labor and the use of thick plaster painted red. The constructions also provide insights into Late Preclassic Maya architectural technology. Originally, these large Preclassic temples were likely platforms for timber temples. Eventually, timber temples were replaced with temples of thin walled masonry structures with perishable roofs. This latter form was in place by the Protoclassic/very Late Preclassic as evidenced at Rio Azuls G 103. Late Preclassic burials are known from several sites in the Three Rivers Region including Dos Hombres (Trachman and Valdez 2001) and Chan Chich (Houk and Robichaux 2003). These two sites serve as proxies for regional expressions of interment activities. Several burials dating to the Late Preclassic were excavated in the Dos Hombres area. The burials, which included a dults and subadults, were placed with similar grave offerings including greenstone, shell, and ceramic vessels. An interesting observation by Trachman (Trachman and Valdez 2001), involved a burial from the Danc er Group near Dos Hombres that had a female b uried with a shell bivalve found over the pelvic region The practice of wearing a bivalve in such a position was reported as an activity from the ethnohistoric record that may represent a practice for socializing the gender of the individual. Although i t cannot be stated as an identical or similar concern extending back to the Late Preclassic, the action is nonetheless intriguing as a possibly longlived tradition. One of the best examples of a Terminal Preclassic (or Protoclassic) tomb comes from the si te of Chan Chich (Houk 1998a, 1998b) at the southern part of the TRR. Chan Chich Tomb 2 was excavated in the Upper Plaza of the site by Houk and Robichaux (2003). The tomb chamber was elliptical in shape and had been covered by a dozen capstones. Among the burial furniture were greenstone artifacts, possible cotton paper, fragments of painted stucco, fragmentary wood, and 11 ceramic vessels. Among the pottery vessels were five mammiform tetrapod support bowls, two spout and bridge jars (often called chocolate pots), two basal flange bowls, a ring base jar, and a basal angle bowl. The vessels are all chronologically assigned to the end of the Preclassic, though several forms continue into the Early Classic. One slip type in particular, found in t he Chan Chich tomb, is Rio Bravo Red which is known to bridge the Preclassic and Early Classic ( Sagebiel 2005:247 253; Sullivan and Valdez 2006:79). Early Classic The TRR Early Classic seems to be stylistically and in content similar to many sites across the en tire Maya area (Buttles, Sullivan, and Valdez 2005; Sullivan and Valdez 2006). Data for the Early Classic is limited to ceramics, settlement impressions, and several tombs and burials. Architecture for Early Classic structures is rather limited. One int eresting adjustment of the Early Classic city is the move to monumental construction almost entirely of masonry. Stone are better cut and shaped to reduce the amount of plaster required, especially as compared to the

PAGE 152

Early Maya Civilization in the Three Rivers Region 156 thick rolling plaster of the Late Prec lassic. It may be the lack of significant timbers that led to the development of the corbel vault and more extensive masonry construction. Given the limited knowledge of Early Classic architecture in the TRR, the focus here will be settlement comments, pottery descriptions for two phases, and burial s that also reflect two phases. Population estimates and settlement have often been limited to data concerning major Maya centers. Mayanists had been left with the impression that there was a decline during th e Early Classic that accounted for an underrepresented occupation. In the Three Rivers Region, the site of Rio Azul had a significant Early Classic presence while the Belize side of the TRR seemed to promote the population decline theory. Fortunately, much survey and excavation at small sites and rural localities has demonstrated that a significant part of Early Classic populations may have moved from site centers to hinterland settlements. The reason(s) for population re settlement is uncertain, but I suspect the Maya may have fouled their cities as one possible scenario. The cities remained important and sacred, but much of the population had moved out, utilizing the centers primarily for ritual(s). Early Classic Maya pottery can be quite distinctive with a particular shade of orange rather than the red of the Late/Terminal Preclassic. Vessel forms combined with the glossy orange slip are the strong indicators of Early Classic pottery. The basal flange bowl form (Figure 7 ) is another good indicator of Early Classic Maya. It seems that two signif icant temporal phases can be seen in Maya pottery. The early phase of the Early Classic follows a particular transition out of the Terminal Pr eclassic. This early phase is what may be called an Early Classic Maya phase that includes the basal flange bowl. A later phase may be termed the Early Classic MayaTeot ihuacan phase that continues with many of the same forms, but add a slab footed t ripod cylinder, often with a cover or lid. These basic distinctions serve to validate the two phase division posited here, but there is a far more extensive array of pottery to each phase. Figure 7 Early Classic Basal -Flange Bowl From Dos Hombres To mb. Figure 8 Early Classic Painted Tomb at Rio Azul. Figure 9 Early Classic Tomb at Dos Hombres. Burials and tombs for the Early Classic are known across the TRR. Rio Azul is known for its phenomenal painted tombs ( Figure 8). Most of the excavated and documented (looted) tombs of Rio Azuls Early Classic date to the proposed second/late phase, MayaTeotihuacan. An early phase Early Classic tomb is represented in the TRR by Dos Hombres Structure B 16 (Durst

PAGE 153

Valdez, Jr. 157 1998). The Early Classic tomb found and excavated at this locale was found beneath a cut in the floor of an elite residential structure near the main plaza (Durst 1998; Robichaux and Durst 1999; Sullivan and Sagebiel 2003. See Figure 9). The tomb was capped by more than 20,000 fragments of obsidian blades, flakes, and cores. The ceramics chronologically place the tomb in the early phase of the Early Classic with a particularly diagnostic. Dos Arroyos Orange Polychrome basal flange bowl (Figure 7) and a Yaloche Cream Poly chrome scutate lid. A late phase Early Classic tomb (or cist) was excavated at the Barba Group, a residential group about 2.5 km northwest of Dos Hombres ( Hageman 2004). This group is located in the Dos Hombres periphery, on a survey transect between Dos Hombres and La Milpa. The Barba Group tomb was located below a small shrine on the east side of the plazuela. Among the burial furniture was a series of ceramic vessels including an effigy turkey pot, a jaguar effigy bowl, and a Teotihuacanstyle tripod cylinder with a lid ( Figure 10 ). The tripod cylinder has a finely executed Maya head as the lid handle. Another late phase Early Classic tomb, Burial B11.67, was uncovered at La Milpa in the Great Plaza near Structure 1 (Hammond et al. 1996). A former chultun had been utilized to construct the tomb. Among the various associated artifacts was a Paradero Flu ted Teotihuacn style tripod cylinder and a scutate lid. Summary and Comments The ancient sites of Chan Chich, Dos Hombres, the Barba Group, and La Milpa provide significant data for tombs of the Terminal Preclassic as well as early and late phases of the Early Classic (Robichaux and Durst 1999). The painted Early Classic tombs of Rio Azul, especially Tombs 19 and 23, are primarily dated to the late phase of the period. The various burials, Preclassic (such as from Dos Hombres), and Early Classic along w ith ceramics of the TRR, and limited architectural developments all indicate important changes between the two periods. While architecture, ceramics, and burials are among the data for interpretations of changes in Maya Civilization for these early periods Figure 10 Early Classic Tomb/Crypt Vessels from the Barba Group. Note Teotihuacanstyle t ripod with l id. these artifacts also serve as indicators of continuity. Generally, change or transitions likely represent significant social and political change that is always a stressful and difficult adaptation for society. The successful changes from one period to another attests to the adaptive strategies and flexibilit y of ancient Maya society. These strategies are what allowed for the development and growth of civilization. The success of early Maya Civilization may also be found in broader connections to neighboring areas as well as more distant contacts or influences. The most successful of civilizations are tho se that adapt to and develop external exchanges. The material culture of the Three Rivers Region inhabitants show connections regionally (trade items with Uaxactun) and internationally (adapting imagery from Teotihuacan). The various ideas and/or notions mentioned or hinted at in this paper are not unique to the Three Rivers Region, but the TRR may serve as a model of greater Maya developments and adaptations. References P. J. Buttles, L. A. Sullivan, an d F. Valdez, Jr. 2005 Early Classic Manifestations in Northern Belize. Research Reports in Belizean Archaeology, Institute of Archaeology, NICH, Belize, Vol. 2:99109 Belmopan. Hendon, Julia 1989 The 1986 Excavations at BA -20. In Rio Azul Reports Number 4, The 1986 Season. Edited by R.E.W. Adams. The University of Texas at San Antonio.

PAGE 154

Early Maya Civilization in the Three Rivers Region 158 Durst, Jeffrey J. 1998 Early Classic Maya Tomb in Northwestern Belize. Paper presented at t he 63rd Annual Meeting of the Society for American Archaeology. Hageman, Jon B. 2004 Late Classic Maya Social Organization: A Perspective from Northwestern Belize. Unpublished Ph.D. dissertation, Department of Anthropology, Southern Illinois University, Carbondale. Hammond, Norman, Gair Tourtellot, Sara Donaghey, and Amanda Clark 1996 Survey and Excavation at La Milpa, Belize, 1996. Mexicon XVIII:86 91. Houk, Brett A. 1996 The Archaeology of Site Planning: An Example from the Maya Site of Dos Hombres, Belize. Unpublished Ph.D. dissertation, Department of Anthropology, The University of Texas at Austin. 1998a An Introduction to the 1997 Season. In The 1997 Season of the Chan Chich Archaeological Project edited by Brett A. Houk, pp. 1 14. Papers of the Chan Chich Archaeological Project, Number 3. Center f or Maya Studies, San Antonio, Texas. 1998b Closing Remarks on the 1997 Season. In The 1997 Season of the Chan Chich Archaeological Project edited by Brett A. Houk, pp. 91 109. Papers of the Chan Chich Archaeological Project, Number 3. Center for Maya Stu dies, San Antonio, Texas. Houk, Brett A., and Hubert R. Robichaux 2003 The Protoclassic in the Three Rivers Region. Paper Presented at the 68th Annual Meeting of the Society for American Archaeology, Milwaukee. Robichaux, Hubert R., and Jeffrey J. Durst 1999 Aspects of Ancient Maya Ideology, as Derived from Two Recently Discovered Elite Tombs in Northwestern Belize. Paper presented at the 64th Annual Meeting of the Society for American Archaeology, Chicago. Robichaux, Hubert R., Jennifer Jellen, Alexandr a Miller, and Jennifer Vander Galien 2000 Preliminary Report on the 1998 Excavations on the Upper Plaza. In The 1998 and 1999 Seasons of the Chan Chich Archaeological Project edited by Brett A. Houk, pp. 49 56. Papers of the Chan Chich Archaeological Proj ect, Number 4. Mesoamerican Archaeological Research Laboratory, The University of Texas at Austin. Sullivan, Lauren A., and Fred Valdez, Jr. 2006 The Late Preclassic to Early Classic Transition in the Three Rivers Region. In Research Reports in Belizean A rchaeology, The Institute of Archaeology, NICH, Belize. Volume 3:74 83. Trachm an, Rissa and Fred Valdez, Jr. 2001 Expressions of Childhood: Life and Death Among the Ancient Maya. Paper presented at the 100th Annual Meeting of the American Anthropologica l Association. Washington, D.C. Trein, Debora 2017 Results of the 2015 Field Season at Structure 27 in Courtyard D, La Milpa. In Programme for Belize Archaeological Project: Activities from the 2015 Investigations. In Research Reports from the Programme for Belize Archaeological Project Occasional Papers, Number 22. Center for Archaeological & Tropical Studies, The University of Texas at Austin Valdez, Fred Jr. 1995 Religion and Iconography of the Preclassic Maya at Rio Azul. In Religion y Sociedad en el Area Maya edited by C. Varela T., J.L. Bonor V., Y. Fernandez M., pp. 211 218. Publicacion de la Sociedad Espanola de Estudios Mayas. Madrid, Spain. Valdez, Fred Jr. and Vernon Scarborough 2013 Past, Present, and Fut ure: Archaeology in NW Belize. R esearch Reports in Belizean Archaeology, Institute of Archaeology, NICH, Belize. Volume 10:51 55. Belmopan.

PAGE 155

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 159 167 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 15 SITUATING PRECLASSIC INTERMENTS AND FIRE -PITS AT SANTA RITA COROZAL, BELIZE Adrian S.Z. Chase, Diane Z. Chase, and Arlen F. Chase Excavations were carried out at Santa Rita Corozal from 1979 through 1985 by the Corozal Postclassic Project. Although the original project sought to excavate only Postclassic occupation, substantial Preclassic remains were also incidentally recove red and recorded during field investigations. This Preclassic data was mentioned in passing in subsequent publications, but was not fully illustrated or published because of the original focus on the Postclassic Period Maya. Approximately 31% of the burials recovered during 4 years of excavation at Santa Rita Corozal by the Corozal Postclassic Project dated to the Precl assic Period. A total of 41 Preclassic burials were recovered from Santa Rita Corozal; 5 of these dated to the early Middle Preclassic Peri od; 4 dated to the later Middle Preclassic Period; and 32 dated to the Late Preclassic Period. The majority of thes e interments were accompanied by one or more ceramic vessels that permitted the dating of the deposit. This paper presents the archaeological data relating to the Preclassic burials recovered at Santa Rita Corozal and also positions these interments in te rms of the broader Maya world. Introduction Of all the practices carried out by ancient and modern peoples, ritual acts probably comprise the most conservative, meaning that, once ingrained in the social fabric, they are the hardest to change. Almost by definition, ritual becomes a codified form of action that is completed in a prescribed way and often at prescribed times, thus permitting researchers to identify these patterns in the archaeological record. Perhaps most significant to the ancient Maya were the ritual acts associated with death and burial. The living treated the bodies of their dead in particular ways, placed them in certain areas, in certain conditions, and with certain objects. It is likely that ritual acts were carried out that related to the memory of the dead, to the transition of the dead into some other state of being, and to their lasting relationship with the living (e.g., Metcalf and Huntington 1991). The Preclassic archaeological record of the ancient Maya has yielded a series of early burials, the patterning of which pro vides insights into early Maya ritual. In addition to helping understand early ritual behavior, parallels to some of these patterns can be established in later time periods in the Maya archaeological record. Among the more interesting of these parallels is the identification of an early ritual pattern that linked the use of fire and fire pits to Preclassic burials, a linkage that we see as continuing in various modified form throughout Maya prehistory. Yet another area of interest relates to the prim ary and secondary nature of Figure 1 Map of Santa Rita Corozal, Belize, created by the Corozal Postclassic Project (after D.Z. Chase and A.F. Chase 1988), showing labeled structures associated with Preclassic materials: Burials only: Structures 37 (O p. P22), 73 (Op P6), 92 (Op. P24), 159 (Op. P19), and 182 (Op. P28); Fire -pits only: Structure 58 (Op. 3); Burials and Fire pits: Structures 35 (Op. 10), 134 (Op. P12), and 189 (Op. P30). interments and the implications of these patterns for the interpret ation of mortuary ritual, particularly in relation to the antiquity of double funerals (e.g., D. Chase and A. Chase 1996: 7677). The site of Santa Rita Corozal ( Figure 1 ) has had a long history of intensive research: initially being dug by Thomas Gann at the

PAGE 156

Preclassic Interments and Fire -pits at Santa Rita Corozal transition of the 19th to 20th centuries (1900, 1918); being subject to test pitting and limited excavations by several projects in the early 1970s (Green 1973; Pring 1973; Sidrys 1983: 124158); then being the focus of four seasons of excavation by the Corozal Postclassic Project from 1979 through 1985 (D. Chase 1981, 1985, 1986; D. Chase and A. Chase 1988); and, more recently seeing further research during the course of stabilization by the Belize Institute of Archaeology in 2013. While the site is b est known for its extensive Postclassic remains (D. Chase and A. Chase 1988, 2004a, 2008), Santa Rita Corozal has yielded spectacular Early Classic materials (D. Chase and A. Chase 2005) as well as some of the earliest evidence of settlement known from nor thern Belize (D. Chase 1981; D. Chase and A. Chase 1988, 2006; ReeseTaylor 2016). This paper builds on earlier research carried out by the Corozal Postclassic Project that was briefly presented in a two articles summarizing the Preclassic and Early Classic Period remains recovered at the site (D. Chase and A. Chase 2005, 2006). The earlier work served to establish the widespread remains at Santa Rita Corozal for both of these time periods. This paper seeks a more synthetic statement concerning the associations and implications of the 41 Preclassic burials recovered at the site ( Table 1 ). Contextualizing Santa Rita Corozals Preclassic Burials Preclassic treatment of the dead at Santa Rita Corozal is relatively consistent throughout the site (see D. Chase and A. Chase 2006). A total of 41 interments may be assigned to the Preclassic Period with 3 more assigned to the Protoclassic era; of the Preclassic interments, 5 date to the Early Middle Preclassic, 4 to the Middle Preclassic, and 32 to the Late P reclassic Period. For all of the individuals interred during the Preclassic Period, 26 were in flexed position, 1 was in extended position (see Figure 2 ), 5 were disarticulated, and 9 were of indeterminate position. The majority of the interments were of a single individual, but two interments contained more than one individual. In both cases, a fully articulated individual was buried with the partial remains of either one or two others. These interments suggest that subsequent later (Late Classic Period ) practices identified at sites such as Caracol, Belize where multiple individual interments and twopart burial practices associated with double funerals are common (e.g. D. Chase and A. Chase 1996, 2011) may have had great antiquity. No clear correlation of body orientation with age, sex, date, or burial offerings exists within the Santa Rita Corozal dataset; the heads of the deceased were nearly evenly split between those with head to the north and south, with a more limited number with heads to the northwest and east. Twenty nine burials were accompanied by one or more ceramic vessels; six burials had more than one vessel with one burial being accompanied by 8 vessels. In the Middle Preclassic a single smaller vessel was often set upright near or on the chest of the interred individual ( Figure 3 ). In the Late Preclassic, very large shallow bowls or platters were inverted over flexed bodies, minimally capping the head and sometimes the entire body ( Figure 4 ). Some aspects of the Late Preclassic bur ials show continuity with later periods; in the subsequent Classic Period at Santa Rita Corozal, inverted bowls were often used to cover the skull of both flexed and extended interments. The earliest Middle Preclassic burials recovered at Santa Rita Coroz al were associated with a Middle Preclassic raised eastern construction, but most of the other recovered Preclassic burials were associated with open spaces and not with raised constructions. Given the research focus of the the Corozal Postclassic Project excavations generally did not focus on the penetration of Preclassic constructions; thus, further excavation at the site may yield other burial patterns typical of those found at other Preclassic sites in northern Belize. When Santa Rita Corozals burial s are compared to other sites with excavated Preclassic Period burials in northern Belize (see Figure 5), the variability in burial patterns among these sites is striking ( Table 2 ). At Cerros, across the bay from Santa Rita Corozal, 32 Preclassic intermen ts were recovered (Walker 2016: Table 3.1); 11 were interred in a seated position; 8 were encased in ceramic containers (see Cliff 1982 for the context of the burial vessels and Carr 1989 for an analysis of the associated faunal material in these vessels

PAGE 157

A. S. Chase, D. Chase, and A. F. Chase 161 Table 1 Preclassic burials from Santa Rita Corozal (41 burials with 44 individuals; 3 individuals are associated with S.D. P10B 6 and 2 with S.D. P19A -10). Burial # Individuals Body Position # Vessels Phase P6E 2 1 Disarticulated 1 Late Preclassic P10B 2 1 Flexed 0 Late Preclassic P10B 3 1 Extended 1 Late Preclassic P10B 4 1 Flexed 0 Late Preclassic P10B 5 1 Flexed 0 Late Preclassic P10B 6a 3 Flexed 1 Late Preclassic P10B 6b (1) Disarticulated 0 Late Preclassic P10B 6c (1) Disarticulated 0 Late Preclassic P10B 8 1 Flexed 1 Late Preclassic P12B 10 1 Flexed 1 Middle Preclassic P12B 11 1 Flexed 1 Early Middle Preclassic P12B 13 1 Flexed 0 Late Preclassic P12B 14 1 Indeterminate 1 Late Preclassic P12B 15 1 Flexed 1 Early Middle Preclassic P12B 16 1 Indeterminate 0 Late Preclassic P12B 18 1 Flexed 1 Middle Preclassic P12B 19 1 Indeterminate 0 Middle Preclassic P12B 24 1 Flexed 1 Early Middle Preclassic P12B 25 1 Indeterminate 0 Early Middle Preclassic P12B 26 1 Flexed 1 Early Middle Preclassic P12B 27 1 Indeterminate 0 Late Preclassic P12B 28 1 Indeterminate 0 Late Preclassic P12B 5 1 Flexed 1 Late Preclassic P12B 6 1 Indeterminate 0 Late Preclassic P12B 7 1 Indeterminate 0 Late Preclassic P12B 8 1 Indeterminate 0 Late Preclassic P19A 10 2 Flexed+Skull 1 Late Preclassic P19A 4 1 Flexed 1 Late Preclassic P19A 5 1 Flexed 5 Late Preclassic P22A 4 1 Disarticulated 1 Late Preclassic P22A 5 1 Flexed 8 Late Preclassic P22A 6 1 Flexed 1 Late Preclassic P24A 1 1 Flexed 1 Middle Preclassic P24C 1 1 Flexed 1 Late Preclassic P24C 2 1 Indeterminate 1 Late Preclassic P24C 3 1 Disarticulated 1 Late Preclassic P28B 3 1 Indeterminate 1 Late Preclassic P28B 5 1 Indeterminate 1 Late Preclassic P30D 2 1 Flexed 2 Late Preclassic P30D 6 1 Flexed 1 Late Preclassic P30D 7 1 Flexed 1 Late Preclassic P30D 8 1 Flexed 2 Late Preclassic P30D 11 1 Flexed 4 Late Preclassic

PAGE 158

Preclassic Interments and Fire -pits at Santa Rita Corozal Figure 2 Preclassic burials recovered in Operation C10B in Santa Rita Corozal Structure 35, showing an extended and two flexed individual interments, as well as a spouted chocolate pot (vessel cross -section shown in A. Chase and D. Chase 1987b: fig. 2b). Figure 3 An Early Middle Preclassic burial recovered from Operation C12B in Santa Rita Corozal Structure 134 showing the placement of a shallow bowl on the chest of the individual (vessel cross -section similar to that shown in A. Chase and D. Chase 1987b: fig. 2a). [considered to be food]); 4 were in a flexed position; 2 were extended; 1 was disarticulated in a pit; and 6 burials were of indeterminate position. Only one of the seated burials appears to have used a large shallow bowl as a covering device (Walker 2016: Fig. 3.6). Walker (2016:70) noted that burials contained within ceramic vessels tended to be placed within buildings while seated burials usually derived from open spaces. For Kaxob, 101 Preclassic burials were recovered (Storey 2004: Table 6.4): 26 were in extended position, 15 were seated, 10 were flexed, 3 were indeterminate, and 47 others were disarticulated. At Cuello, 157 burials can be assigned a Preclassic date (Ro bin 1989: Table 10; Hammond et al. 1991, 1992): 27 of these were interred in a seated positon; 47 were Figure 4 An example of a Late Preclassic burial in Operation C10B in Santa Rita Structure 35 covered by an inverted red -slipped platter (vessel cro ss-section shown in A. Chase and D. Chase 1987b: fig. 2c). Figure 5 Map of Northern Belize, showing location of comparative sites (after Hammond 1975: Figure 11.1). extended; 26 were flexed; 18 were disarticulated bundles; 20 were simply disarticula ted; 15 were indeterminate as to position; and, 4 consisted only of skulls (possibly caches; see Haviland 1990). However, aside from a semi frequent placement of inverted bowls over crania, no additional correlations of body orientation with age, sex, dat e, or burial offerings emerge from the large sample of burials at Cuello (Robin 1989:152). These data show the great variability in burial patterns found at sites in northern Belize during the Preclassic Period. Kaxob also illustrates the interment of mu ltiple individuals within a single grave pit during the Late Preclassic Period, signaling the retention of bundled remains to accompany the burial of, what has been interpreted to be, an individual of

PAGE 159

A. S. Chase, D. Chase, and A. F. Chase 163 Table 2 Comparison of Preclassic b urial contexts between sites in northern Belize. Cerros data from Walker 2016: Table 3.1; Kaxob data from Storey 2004: Table 6.4; Cuello data from Robin 1989: Table 10, Hammond et al. 1991, 1992; Santa Rita Corozal (SRC) data presented in this paper. The disarticulated total for Santa Rita Corozal does not include 2 disarticulated individuals placed with a flexed burial in S.D. P10B -6 or a skull placed with a flexed individual in S.D. P19A -10). Site Seated Ceramic Container Flexed Extended Disarticulated Indeterminate Total Preclassic Cerros 11 8 4 2 1 6 32 Kaxob 15 0 10 26 47 3 101 Cuello 27 0 26 47 38 19 157 SRC 0 0 25 1 3 12 41 Figure 6 Example of a sherd lined fire -pit in the basal portion of Operation C3B in Structure 58 of Santa Rita Corozal. higher status (McAnany et al. 1999; Storey 2004). While earlier arguments were made that seated burials represented individuals of some authority (e.g., McAnany et al. 1999:133), their widespread distribution in northern Belize and t heir co existence with other patterns does not support this assertion. In fact, different segments of Preclassic Maya society may have practiced varying burial patterns. At Cuello, in contrast to the wide variety of interment patterns recovered during initial research (Robin 1989), subsequent excavations in more formal buildings of Middle Preclassic date almost exclusively recovered extended individuals in cist burials (one interment in 1992 contained 2 flexed individuals), most with vessels covering the ir skulls (Hammond et al. 1991: Fig. 5; 1992). This association is likely significant and suggests that there may have been distinct societal divisions at this early date that to some extent may be correlated with burial type; two rich child burials (with multiple grave goods) in this sample also su ggest that some degree of social ranking existed in the Maya Lowlands by the later seventh century B.C. (Hammond et al. 1991:362; 1992:961). At Santa Rita Corozal cist and simple burials both may be found to contain additional grave goods; however, as no ted, the sample resulted from research focused on the Postclassic Period instead of focusing specifically on the recovery of early remains. Thus, at least some of the variability in interment patterns seen in the Preclassic archaeological record of northe rn Belize may correlate with differing social levels recovered at the various sites. Fire pits One of the more interesting features at Santa Rita Corozal are a series of firepits associated with the Preclassic Period levels of the site ( Figure 6 ). These features have a widespread occurrence in the archaeological records of northern Belize but are consistently associated only with Preclassic activity. Fire pits generally consist of circular or hemispherical depressions lined by stones, ceramics, or both that usually show evidence of having been subjected to the high heat of a fire (the soil, rocks, and ceramics are discolored; in addition, ash may be present). The ceramics can be piece meal in the feature or, alternatively, represent the almost intact base of a plainware vessel (at Kaxob, a series of Sierra Red shallow bowls were recovered in a sherd line pit [Berry et al. 2004:211]; nothing like this was found at Santa Rita Corozal). At Cuello, Gerhardt (1988:9091) describes these features throughout her Preclassic sequence and originally assigned them a domestic function related to cooking

PAGE 160

Preclassic Interments and Fire -pits at Santa Rita Corozal and washing; they were later recognized as spanning a wider range of functions and were not simply hearths (Hammond et al. 1991:355). Some of these Preclassi c features at Cuello do not appear to have been fired, leading to subsequent speculation of their uncertain function (Hammond 1991:236). Fire pits are also known to appear at Cerros (Robertson, personal communication, 2015), Colha (Hester et al. 1981), Nohmul (Pyburn 1989), and Kaxob (Bobo 2004). The Kaxob sherd lined pits underwent intensive analysis by Victoria Bobo (2004:104), who argued that they were involved in multiple functions involving steaming, soaking, and burning. The assumption that these pits were used for domestic activities is clear from their earliest definition by Gerhardt (1988:90) and their explicit association by Pyburn (1989) to ancient Maya cuisine, although Bobo (2004:87) does say that they possibly could relate to ceremonial undertakings. Bobo notes that she had a difficult time with her analysis of sherd lined pits because these features were referred to as either hearths or firepits at other sites; however, she (2004:89, 92) provides detailed information for 5 of these features from Colha, 26 of these features from Cuello, and 30 more sherd lined pits from Kaxob. Even though the one published picture of these features for Kaxob shows them in association with a burial, Bobo (2004:93) did not engage in a discussion of their possible ceremonial nature. While Bobo (2004) argues that some of her pits occur within buildings, the majority of them are actually outside of constructions, a finding consistent with pits from Nohmul (Pyburn 1989), Cuello (Hammond and Gerhardt 1990), and Santa Rita Corozal. For Santa Rita Corozal, 12 sherdlined fire pits were recovered in four distinct locales: one from a deep cut in the axial trench through Structure 58; two from the axial trench through Structure 35; five from an axial tren ch through Structure 134; and four from an axial trench through Structure 189 (see Figure 1). At Santa Rita, fire pits are broadly located at the site, but always in association with Preclassic stratigraphic levels. Given that the excavations undertaken by the Corozal Postclassic Project were focused on the Postclassic Period and often did not strive to open earlier levels, the amount of fire pits recovered suggests that they were quite common. Intriguingly, the one areal excavation of a sizeable Preclas sic construction (Operation 24) was not associated with a fire pit (see D. Chase and A. Chase 2006:92). In only one excavation were the recovered firepits in direct association with structural remains; most were not clearly in association with Preclassic buildings. However, all recovered firepits at Santa Rita Corozal were found in areas of Preclassic interments, suggesting to us that these features may not have had a purely domestic function. The one case in which fire pits are clearly associated with a construction is in excavations of the earliest known structure at Santa Rita Corozal, dating from the early part of the Middle Preclassic Period (D. Chase and A. Chase 2006). The five fire pits are on axis to the two earliest phases of Structure 135, wh ich is associated with five Middle Preclassic interments. Three of the fire pits in Structure 135 are directly above an Early Middle Preclassic burial and the other two are on axis to the construction. In our estimation, this constitutes a strong argumen t that the interment and the fire pits were linked. Similar linkage is also seen in Structure 189 where another fire pit directly overlies a Late Preclassic burial. Thus, we believe that these fire pits are the Preclassic equivalent of the ritual burning that is seen throughout the Classic Period archaeological record in ritual contexts. Burning within architectural contexts has been recorded for most excavated sites within the Maya lowlands and has often been conjoined with other ritualized acts within the Maya archaeological record. Many researchers have noted the association between fire and ritual buildings and contexts, often ascribing these acts to termination deposits (i.e., Mock 1998; Harrison Buck 2012). While clearly dating to a different tem poral era, Late Postclassic and historic contexts also incorporate the ritual destruction of items by burning (e.g., Tozzer 1941). But, there are differences in scale in these acts, as well. William R. Coe (1990: 937) documented the extensive burning and scorching that was associated with Tikals North Acropolis and specifically commented on numerous empty burnt pits, noting that an interesting coupling

PAGE 161

A. S. Chase, D. Chase, and A. F. Chase 165 also exists between ignition and sever al interments. At Caracol, Belize burning was intimately associated with the interment of individuals; censers filled with burnt ash, probably derivative from some ritual related to interment, were sometimes placed within tombs and other burning occurred after the placement of the burial (e.g. A. Chase and D. Chase 1987a: 2627); in other cases an entire burial was incinerated before being incorporated into the archaeological record (e.g., A. Chase and D. Chase 2011). In the Postclassic era, ritualized burning is sometimes literally incorporated within the body of the dead; one of the skulls recorded from the Cenote of Sacrifice at Chichen Itza had a circular hole in its top and had been repurposed as an incense burner (Coggins 2004). Thus, the associati on of purposeful fire with ritual activity and death has a long tradition in Mesoamerica. Fire is not only associated with mortuary rituals, but also with other ceremonies, including world renewal activities that took place at regular temporal intervals th roughout Mesoamerica (Christenson 2016; Elson and Smith 2001; Stuart 1998; Vail and Looper 2015). Evidence for the episodic deposition of both interments and caches associated with temporal cycles has previously been identified for Late Classic Period Car acol, Belize (e.g., A. Chase and D. Chase 2013; D. Chase and A. Chase 2004b, 2011). A future line of research may draw similarities between the Preclassic and Classic practices. Conclusion Maya ritual has shown great continuity and persistence over time, and analysis of Preclassic burial patterns provides insight into early Maya ritual patterns. Some of these patterns, such as flexed burials and the placement of inverted vessels over the individuals head and upper body, continued into later time periods. Similarly, the few multiple individual interments with mixed articulation recovered at Santa Rita Corozal suggest that the curation of the dead and the double funeral practices common in the Late Classic Period at many sites can trace its ancestry to pr actices already under way in the Preclassic Period. Given the persistence of burial ritual, differences and similarities within and among sites are significant. As noted above, Preclassic Period interments from Santa Rita Corozal, at least those thus far sampled, are somewhat less varied than those at some other Maya sites in northern Belize. While this may be due to sampling or wealth variations, it may also be due to greater cohesion in ritual practices among some sites and not others. Initial data ana lysis and research on Preclassic burials in northern Belize attempted to identify correlations between measured categories that included age, sex, date, or burial offerings; often the results were inconclusive. However, the primary additional pattern iden tified at Santa Rita Corozal was the association between firepits and Preclassic burials. This linkage between fire and interment can be seen as continuing in various modified forms throughout Maya prehistory, although the fire pits themselves are most evident in Preclassic contexts. Thus, we believe that these fire pits are the Preclassic equivalent of the ritual burning that is seen in throughout the Classic Period archaeological record in ritual contexts. A question to be answered by later research is whether the conjoined burning and human interment activities were part of wider calendric and world renewal ceremonies that were continued in somewhat different form throughout the rest of Maya prehistory. References Berry, Kimberly A., Sandra L. Lop ez Varela, Mary Lee Bartlett, Tamarra Martz, and Patricia A. McAnany 2004 Pottery Vessels of Kaxob. In P.A. McAnany, Ed. Kaxob: Ritual, Work, and Family in an Ancient Maya Village pp. 193 261. Cotsen Institute of Archaeology Monumenta Archaeologica 22. University of California, Los Angeles. Bobo, Victoria 2004 Soaked and Steamed Sustenance: Evidence from Sherd -Lined Pits. In P.A. McAnany, Ed. Kaxob: Ritual, Work, and Family in an Ancient Maya Village pp. 87 -104. Cotsen Institute of Archaeology Monumenta Archaeologica 22. University of California, Los Angeles. Carr, Helen Soraya 1986 Faunal Utilization in a Late Preclassic Maya Community at Cerros, Belize. Doctoral Dissertation, Department of Anthropology, Tulane University.

PAGE 162

Preclassic Interments and Fire -pits at Santa Rita Corozal Chase, Arlen F. and Diane Z. Chase 1987a Investigations at the Classic Maya City of Caracol, Belize: 1985-1987. Monograph 3. Pre Columbian Art Research Institute, San Francisco. 1987b Putting Together the Pieces: Maya Pottery of Northern Belize and Central Peten, Guatemala. In P.M. Rice and R.J. Sharer, Eds. Maya Ceramics: Papers from the 1985 Maya Ceramic Conference, pp. 47 -72. BAR International Series 345(i), Oxford. 2011 Status and Power: Caracol, Teotihuacan, and the Early Classic Maya World. Research Reports in Belizean Archaeology 8:3 18. 2013 Temporal Cycles in the Archaeology of Maya Residential Groups from Caracol, Belize. Research Reports in Belizean Archaeology 10:1323. Chase, Diane Z. 1981 The Maya Postclassic at Santa Rita Corozal. Archaeology 34(1): 2 5 -33. 1985 Ganned But Not Forgotten: Late Postclassic Archaeology and Ritual at Santa Rita Corozal, Belize. In A.F. Chase and P.M. Rice, Eds. The Lowland Maya Postclassic pp. 104 125. University of Texas Press, Austin. 1986 Social and Political Organization in the Land of Cacao and Honey: Correlating The Archaeology and Ethnohistory of the Postclassic Lowland Maya. In J.A. Sabloff and E.W. AndrewsV, Eds. Late Lowland Maya Civilization: Classic to Postclassic pp. 347-377. Uni versity of New Mexico Press, Albuquerque. Chase, Diane Z. and Arlen F. Chase 1988 A Postclassic Perspective: Excavations at the Maya Site of Santa Rita Corozal, Belize. Monograph 4. Pre -Columbian Art Research Institute, San Francisco. 1996 Maya Multiples : Individuals, Entries, and Tombs in Structure A34 of Caracol, Belize. Latin American Antiquity 7: 61 -79. 2004a Santa Rita Corozal: Twenty Years Later. Research Reports in Belizean Archaeology 1: 243255. 2004b Patrones de Enterramiento y Cclos Residenciales en Caracol, Belice. In R. Cobos, Ed. Culto Funerario en la Sociedad Maya: Memoria de la Cuarta Mesa Redonda de Palenque pp. 203 230. INAH, Mexico, D.F. 2005 The Early Classic Period at Santa Rita Corozal: Issues of Hierarchy, Heterarchy, and Stratification in Northern Belize. Research Reports in Belizean Archaeology 2: 17 -38. 2006 The Dawn of Maya Civilization: Preclassic Archaeology From Santa Rita Corozal. Research Reports in Belizeain Archaeology 3: 85100. 2008 Late Postclassic Ritual at Santa Rita Corozal, Belize: Understanding the Archaeology of a Maya Capital City. Research Reports in Belizean Archaeology 5: 79 -92. 2011 Ghosts amid the Ruins: Analyzing Relationships Between the Living and the Dead Among the Ancient Maya at Caracol, Belize. In J. Fitzsimmons and I. Shimada, Eds. Living with the Dead: Mortuary Ritual in Mesoamerica, pp. 78 101, University of Arizona Press, Tucson. Christenson, Allen J. 2016 The Burden of the Ancients: Maya Ceremonies of World Renewal from the Pre columbian Period to the Present. University of Texas Press, Austin. Cliff, Maynard Burrell 1982 Lowland Maya Nucleation: A Case Study from Northe rn Belize. Ph.D., Dedman College, Southern Methodist University, Dallas, TX. Coe, William R. 1990 Excavations in the Great Plaza, North Terrace, and North Acropolis of Tikal Tikal Report 14. University Museum Monograph 61. University of Pennsylvania, Phi ladelphia. Coggins, Clemency C. 2004 Artifacts from the Cenote of Sacrifice, Chichen Itza, Yucatan. Peabody Museum Memoirs Volume 10. Peabody Museum Press, Harvard University, Cambridge. Elson, Christina M. and Michael E. Smith 2001 Archaeological Deposi ts from the Aztec New Fire Ceremony. Ancient Mesoamerica 12: 157174. Gann, Thomas 1900 Mounds in Northern Honduras. Nineteenth Annual Report, 1897-1898, Bureau of American Ethnology, Part 2, pp. 661691. Smithsonian Institution, Washington, D.C. 1918 The Maya Indians of Southern Yucatan and Northern British Honduras. Bureau of American Ethnology Bulletin No. 64. Smithsonian Institution, Washington, D.C. Gerhardt, Juliette C 1988 Preclassic Maya Architecture at Cuello, Belize. BAR International Series 464. British Archaeological Reports, Oxford. Green, Ernestine 1973 Location Analysis of Prehistoric Maya Sites in Northern British Honduras. American Antiquity 38: 279 -293.

PAGE 163

A. S. Chase, D. Chase, and A. F. Chase 167 Hammond, Norman 1975 Introduction. In N. Hammond, Ed. Archaeology in Northern Beli ze: British Museum-Cambridge University Corozal Project: 1974 -1975 Interim Report, pp. 6 -14. Centre of Latin American Studies, University of Cambridge, London. Hammond, Norman and Juliette C. Gerhardt 1990 Early Maya Architectural Innovation at Cuello, Belize. World Archaeology 21: 461 -481. Hammond, Norman, Amanda Clarke, and Francisco Estrada Belli 1992 Middle Preclassic Maya Buildings and Burials at Cuello, Belize. Antiquity 66: 955 -964. Hammond, Norman, Amanda Clarke, and Cynthia Robin 1991 Middle Preclassic Buildings and Burials at Cuello, Belize: 1990 Investigations. Latin American Antiquity 2: 352-363. Harrison -Buck, Eleanor 2012 Rituals of Death and Disempowerment among the Maya. In E. Harrison -Buck, Ed. Power and Identity in Archaeological Theory and Practice: Case Studies from Ancient Mesoamerica, pp. 103115. University of Utah Press, Salt Lake City. Haviland, William A. 1990 Review of Preclassic Maya Burials at Cuello, Belize by Cynthia Robin. Journal of Field Archaeology 17: 494-496. Hester, Thomas R., Jack D. Eaton, and Harry J. Shafer 1981 Eds., The Colha Project: Second Season: Interim Report. Center for Archaeological Research, University of Texas, San Antonio. McAnany, Patricia A., Rebecca Storey, and Angela Lockard 1999 Mortua ry Ritual and Family Politics at Formative and Early Classic Kaxob, Belize. Ancient Mesoamerica 10: 129146. Metcalf, Peter and Richard Huntington 1991 Celebrations of Death: The Anthropology of Mortuary Ritual 2nd edition. Cambridge University Press, C ambridge. Mock, Shirley B. 1998 The Sowing and the Dawning: Termination, Dedication, and Transformation in the Archaeological and Ethnographic Record of Mesoamerica. University of New Mexico Press, Albuquerque. Pring, Duncan 1973 Operation 8B, Santa Rita Corozal. In N. Hammond, Ed. British Museum-Cambridge University Corozal Project, pp. 63-67. Centre of Latin American Studies, University of Cambridge, London. Pyburn, Anne 1989 Maya Cuisine: Hearths and Lowland Economy. In P.A. McAnany and B.L. Isaac, Eds Research in Economic Anthropology: Prehistoric Maya Economies of Belize, pp. 325-346. Supplement 4. JAI Press, Inc., Greenwich (CT). ReeseTaylor, Kathryn 2016 The First Settlers on Chetumal Bay. In D.S. Walker, Ed. Perspecitives on the Ancient Maya of Chetumal Bay pp. 3355. University Press of Florida, Gainesville. Robin, Cynthia 1989 Preclassic Maya Burials at Cuello, Belize. BAR International Series 480. British Archaeological Reports, Oxford. Sidrys, Raymond 1983 Archaeological Excavations in N orthern Belize, Central America. Monograph 16. Institute of Archaeology, University of California, Los Angeles. Storey, Rebecca 2004 Ancestors: Bioarchaeology of the Human Remains of Kaxob. In P.A. McAnany, Ed. Kaxob: Ritual, Work, and Family in an Anci ent Maya Village pp. 109 -138. Cotsen Institute of Archaeology Monumenta Archaeologica 22. University of California, Los Angeles. Stuart, David 1998 The Fire Enters His House: Architecture and Ritual in Classic Maya Texts. In S.D. Houston, Ed. Function and Meaning in Classic Maya Architecture, pp. 373 425. Dumbarton Oaks, Washington, D.C. Tozzer, Alfred M. 1941 Landas Relacion de las Cosas de Yucatan. Papers of the Peabody Museum of Archaeology and Ethnology 18. Harvard University, Cambridge. Vail, Gabrielle and Matthew G. Looper 2015 World Renewal Rituals and the Postclassic Yucatec Maya and Contemporary Chorti Maya. Estudios de Cultura Maya 45: 121 -140. Walker, Debra S. 2016 Life and Afterlife at Cerro Maya, Belize. In D.S. Walker, Ed. Perspeci tives on the Ancient Maya of Chetumal Bay pp. 56 -75. University Press of Florida, Gainesville.

PAGE 164

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 169 178 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 16 CHETUMALS DRAGONGLA SS: POSTCLASSIC OBSI DIAN PRODUCTION AND EXCHA NGE AT SANTA RITA CO ROZAL, BELIZE Max Seidita, Diane Z. Chase, and Arlen F. Chase The reconstruction of patterns of obsidian production and exchange has been vital to increasing our understanding of ancient Mesoamerican economies. During the Postclassic Period (ca. 9501532 C.E.) inhabitants of the site of Santa Rita Corozal, Belize pa rticipated in multiple exchange networks to provision themselves with the materials to produce and distribute obsidian artifacts. Analyses of obsidian artifacts dating to the Postclassic Period recovered by the Corozal Postclassic Project from 19791985 h as demonstrated that Santa Rita Corozals population was engaged in the production of obsidian blades, likely from partially reduced polyhedral cores. Here we present information on 572 obsidian artifacts, including details relating to the tool production sequence and pXRF sourcing data. We further present the statistical distributions of objects in relation to hypothesized function or status of each structure. Procurement of these cores from at least six different obsidian sources likely occurred as a r esult of indirect trade carried out by non-specialist traders. The sources represented include Otumba, Mexico, a resource not previously identified in the region. Imported raw materials were transformed into finished artifacts and then distributed throug h a market system. These findings contrast with the patterns of procurement and production seen at neighboring sites thought to be members of the Chetumal polity, where prismatic blade production and number of sources being exploited were more limited. We conclude by discussing these findings in relation to neighboring sites and within our broader understandings the Postclassic Period regional economy. Introduction Obsidian is one of the most commonly recovered materials in the Maya region and some types of obsidian and obsidian artifacts, such as prismatic blades, were utilitarian goods used in everyday life and transported in bulk into sites across the lowlands (Braswell 2003; Edwards 1978; Smith 2003). A testimony to obsidians popularity and perhaps its unique characteristics, the ancient Maya imported obsidian into regions where other sources of chipped stone, such as chert, were easily accessible and commonly available to the general population. Since the early 1900s, Maya archaeologists have viewed the presence of obsidian as an indicator of interregional exchange and consumption, as well as one avenue by which to explore production (Chase and Chase 1989; Kidder 1947; Sidrys 1976). The persistent emphasis on obsidian by Maya researchers may be attributed to its ubiquity in the archaeological record, the resilient nature of both finished artifacts and production debitage, and its el emental properties which allow for the attribution of an obsidian artifact to a specific geological source (Ferguson 2012; Shackley 2011). With the recent upswell in the study of Maya market exchange, obsidian has become a vital component in any discussion of the various mechanisms of distribution employed by the ancient Maya. Obsidian has been included in market studies at sites throughout the Maya region including Caracol (Chase and Chase 2015; Chase and Chase 2014) and Xunantunich (Cap 2015; Cap, et al 2015) in Belize; Tikal, Guatemala (Masson and Freidel 2012); Mayapan, Mexico (Masson and Freidel 2012); and Ceren, El Salvador (Sheets 2000). Many of these studies, including the one presented here for Santa Rita Corozal, rely upon data derived primaril y from the excavation of households (see Cap (2015); Cap, et al. (2015) for a recent and notable exception). Less often discussed than the broader mechanisms of distribution across sites are the means by which th e market itself is provisioned. With the above in mind, here we present an updated study of the Postclassic Period obsidian economy at the site of Santa Rita Corozal in northern Belize. Through a combination of lithic analysis, a sourcing study utilizing pXRF, and statistical analysis of obsidian locations, we seek to understand the importation, production, and distribution of obsidian at Santa Rita Corozal during the Postclassic Period (ca 950 1532 C.E.). We argue that the Postclassic Period population of Santa Rita Corozal was importing partially reduced polyhedral cores from a variety of Guatemalan and Mexican obsidian sources,

PAGE 165

Postclassic Obsidian Production and Exchange at Santa Rita 170 producing blades locally, and then distributing the obsidian through market exchange. Through this research we aim to contribute information from the Postclassic provinc ial capital of Santa Rita Corozal to discussions of other obsidian economies in northern Belize. More broadly, we seek to demonstrate the value of considering the provisioning of local economies with nonlocal goods in conjunction with local methods of di stribution, the importance of large sample sizes in the elemental sourcing of obsidian, and the value of existing collections. Santa Rita Corozals Obsidian Economy Located in northern Belize, Santa Rita Corozal was situated along the coast at the western end of Chetumal Bay; the site is now largely subsumed by the sprawl of Corozal town and rising sea levels. Santa Rita Corozal and its surrounding areas have been continuously occupied since at least the early Middle Preclassic Period (Chase 1981:26) with an Early Classic prominence in the region (Chase and Chase 2005). However, Maya occupation at the site reached its peak population during the Late Postclassic Period when the population of the city itself is believed to have been approximately 7,000 inha bitants (Chase 1990), not including people in the surrounding landscape. The site was situated near the mouths of three major river systems: The New River and Freshwater Creek (which both have their head waters in central Belize) and the Rio Hondo (which reaches into the Peten of Guatemala) ( Figure 1 ). These river systems would have served as important transportation routes and would have fostered communication and exchange between the coast and inland settlements (Chase and Chase 1989). Additionally, Sa nta Rita Corozals coastal location would have facilitated its participation in both circum peninsular trade and exchange along the Belize coast (Chase 1986); seaborne trade has long been thought to be the primary means of long distance exchange amongst th e Postclassic Period Maya (Sabloff and Rathje 1975). Investigated by several projects since the early 1900s, Santa Rita Corozal was most intensively excavated by the Corozal Postclassic Project (CPP) from 1979 1985 under the direction of Diane and Arlen Chase (Chase Figure 1 Map of Chetumal Bay showing sites mentioned in text. Figure 2 Map of Santa Rita Corozal From Seidita (2015), adapted from D. Chase and A. Chase 1988. 1982; Chase and Chase 1988). Excavations took place over the span of four years and consisted of the mapping of 200 features including structures, platforms, and chultuns, as well as the excavation of 46 of these features ( Figure 2 ). The 572 Late Postclassic obsidian artifacts discussed here were recovered by the project

PAGE 166

Seidita, Chase, and Chase 171 (information on 27 obsidian points recovered during these investigations may be found in (Marino, et al. 2015)). The CPP established Santa Rita Corozal as the likely location of Chetumal, the Postclassic Period provi ncial capital and the namesake of the Chetumal province (Chase 1982, 1985; Chase and Chase 1988, 2004, 2008). Chetumal was described in Spanish accounts as an economic center which participated in extensive trade networks and was known for producing and e xporting honey and cacao (Chase 1986). The Chetumal province itself is thought to have occupied the area of modern day northern Belize and southern Quintana Roo. While the relationship between Santa Rita Corozal and smaller sites in the region has not be en fully explored, work by Marilyn Masson (2000) suggested that, during the Postclassic Period, Santa Rita Corozal sat atop a hierarchy of sites, which may have included Caye Coco, Laguna de On, Ichpaatun, and Sarteneja. Following Masson (2000), the prim ary site of Santa Rita Corozal was likely supported by smaller secondary sites such as Caye Coco, which were in turn supported by tertiary communities such as Laguna de On. Importation and Production Seidita (2015), undertook an initial study of the Santa Rita Corozal obsidian, focusing on finished artifacts and debitage in an attempt to understand the type and extent of production that had occurred amongst the Postclassic Period population of Santa Rita Corozal. Ad ditionally, this analysis was able to establish the likely form of the obsidian prepared for transport for example nodules, cores, or finished artifacts as various stages of reduction result in diagnostic and finished artifacts indicative of where in t he reduction sequence local production began (e.g., Clark and Bryant 1997). Understanding the forms that were imported allows for partial modeling of the social and economic relationships which drove the provisioning of obsidian to nonobsidian producing regions, such as the Maya Lowlands (Hirth 2008). Our analysis of the obsidian assemblage shows that, unsurprisingly, Santa Rita Corozals primary industry was the production of final series prismatic blades ( Table 1 ). Of the 572 Postclassic Period obsidia n artifacts, 498 are complete and segmented blades. Of these blades, over 95% are blade segments, most of which may be classified as purely or partially media l. The majority of these segments retain the tongue and tongue facets indicative of purposef ul segmentation of blades (Hirth et al. 2006). De Leon (2009) has argued that selecting for medial segments, the flattest portion of the blade, makes blades easier to haft. This is supported by the presence of side notching on 15% of the blade assemblage. Most commonly blades were notched with a single unilateral side notch. While we do not discuss usewear analysis here, this evidence suggests a preference for flatter blades that were more easily hafted, a claim supported by an earlier use wear study on Santa Rita Corozal materials by (Hartman 1980) that found that blade segments had been hafted for activities like those involving the repetitive cutting of a fibrous material Additionally, two initial series blades, from early in the reduction sequence, were recovered; both have extensive edge damage that likely re sulted from their use as tools. Obsidian production debitage and refuse includes cores, blade errors, and error removals. Thirteen cores and core fragments dating to the Postclassic Period wer e recovered. Of these cores, only a single complete core was recovered. This exhausted core showed evidence of bipolar blade removal and is only 3.45 cm in length. If we consider the aforementioned complete final series blades as indicative of the size of cores at Santa Rita Corozal, the average core would have been around 3.6 cm in length, with lengths ranging from 2.6 cm to 7.3 cm. The proximal section of the single complete core and the other proximal fragments indicate that cores were prepared by ha ving their platforms ground. This method of preparation, while labor intensive, facilitates easier and more predictable blade removal (Crabtree 1968; Hirth et al. 2006). In addition to these cores, two platform preparation flakes were recovered. Other debitage and production errors include plunging blades, hinge fractures, and attempts at removing these errors through rejuvenation flakes. Analysis of Santa Rita Corozals obsidian assemblage demonstrates that during the Postclassic Period the population o f Santa Rita

PAGE 167

Postclassic Obsidian Production and Exchange at Santa Rita 172 Table 1 Table of quantities and frequency of obsidian sources present in the sample. Sources El Chayal Ixtepeque San Martin Jilotepeque Otumba Pachuca Pico de Orizaba n 161 332 4 34 3 3 % 30% 62% <1% 6% <1% <1% Figure 3 Hierarchical Cluster Analysis of the elements Mn, Fe, Zn, Rb, Sr, Y, Zr, Nb for chemical groups and known sources. Figure 4 Bivariate plot of Sr/Rb by Zr/Rb PPM values, 90% confidence ellipses.

PAGE 168

Seidita, Chase, and Chase 173 Figure 5 Bivariate plot of log10 Mn x Log10 Rb PPM values, 90% confidence ellipses. Corozal was consuming blades produced locally. The assemblage does not represent any loci of production, such as workshops, which are characterized by significant amounts of debitage and refuse with little in the way of finished artifacts (Clark and Bryant 1997). The debitage and refuse which is present are indicative of the importation of prepared polyhedral cores or partially reduced polyhedral cores. The lack of percussion debitage, and the limited number of initial series blades indicate that cores were likely prepared for the pressure flaking of final series blades prior to importation (Anderson and Hirth 2008; Hirth, et al. 2006). Lastly, the small number of cores and core fragments recovered could easily account for the blades present in the assemblage. By treating core fragments as unique individual cores, the thirteen cores from Santa Rita Corozal that date to the Postclassic P eriod could potentially produce 2,600 blades (Clark 1988; Clark and Bryant 1997). This is 4.5 times the number of blades recovered; if we were to include the distal portion of three plunge blades, this climbs to 5.5 times the blades recovered. We do not suggest, however, that the full ancient sample of cores and blades has been recovered. To identify the obsidian sources and source regions present in the Postclassic Period assemblage a sample of 536 pieces of obsidian dating to the Postclassic Period were assayed via pXRF. The data presented here derives from a more recent and detailed analysis than that presented previously by Seidita (2015). Using a Bruker Tracer IIISD pXRF samples were assayed for 90 seconds, at 40kV max voltage, without a vacuum, ut ilizing Brukers 0.006 Cu, 0.001 Ti, 0.012 Al, green filter. The pXRF data was normalized via a Log10 transformation to reduce the potential for analysis to favor elements with larger values by transforming all the elemental values into the same ord er of magnitude (Millhauser, et al. 2015; Popelka Filcoff 2006). Following normalization, data was subjected to two rounds of hierarchical cluster analysis to establish the number of distinct chemical groups within the data set and to offer preliminary id entification of group sources. These preliminary assignments were then substantiated via bivariate plots of PPM data and ratios of PPM data, specifically Sr/Rb by Zr/Rb. Frahm (2016) demonstrated that bivariate plotting of ratios helps control for size i nduced spread resulting from samples of less than ideal dimensions. This type of plotting has

PAGE 169

Postclassic Obsidian Production and Exchange at Santa Rita 174 the additional benefit of adding a third or potentially fourth variable in this case an element to consider on a twodimensional plot (Frahm 2016). Thus, through a combination of element ratios and standard bivariate plots it is possible to better characterize the sources present within the assemblage. To confirm the preliminary source assignments generated through the use of cluster analysis, confidence ell ipses are generated at 90% confidence using reference material including source samples analyzed in an earlier pXRF analysis, published pXRF, and INAA data (Glascock and Cobean 2002; Millhauser, et al. 2015; Millhauser, et al. 2011). The combination of hie rarchical cluster analysis and bivariate plotting of ratios and ppm data was clearly able to distinguish between the sources present in the Postclassic Period assemblage ( Figures 3, 4, 5). The initial round of cluster analysis identified six distinct chem ical groups within the assemblage. The second round of cluster analysis preliminarily identified chemical clusters for all six groups within the Postclassic Period Santa Rita Corozal assemblage including: Pachuca (Group 1), Pico de Orizaba (G roup 2), San Martin Jilotepeque (Group 3), El Chayal (Group 5), Otumba (Group 4), and Ixtepeque (Group 6) (see Figure 3). The bivariate plots of Sr/Rb by Zr/Rb confirm the preliminary assignments generated by the hierarchical cluster analysis (Figure 4). While Otumba and El Chayal have a slight overlap in this plot they may be separated by plotting Log10 Mn by Log10 Rb (Figure 5). These are the same sources, albeit with a substantially larger sample size, as documented in the earlier study. Significantly, the presen ce of Pico de Orizaba and Otumba were confirmed through this additional analysis; and, in the case of Otumba, the amount of obsidian attributed to that source increased in the second study. Along with green obsidian from Pachuca these sources represent th e only Mexican obsidian sources documented at mainland settlements in northeastern Belize during this time. Taken together these forms of analysis allow us to draw conclusions regarding the provisioning of the local obsidian economy. While the sample size of assayed materials dating to earlier periods at Santa Rita Corozal is too small to be conclusive, it appears that there is an explosion in the diversity of sources being exploited during the Postclassic Period. In particular, the presence, diversity, and quantity of Mexican obsidian were not expected, especially given the regions pattern of source exploitation, wh ich favored Guatemalan sources (Braswell 2003; Golitko and Feinman 2015). We attribute the diversification of sources to increased interregi onal exchange during the Postclassic Period, something Mayanists have recognized since the 1970s (e.g., Sidrys 1976b). This type of exchange would have had the added benefit of aiding in the provisioning of the growing population. Given the distributions it does not appear that the provisioning was a result of the elite investing in social ties to bring obsidian into the site. In his modeling of obsidian provisioning in Central Mexico, Kenneth Hirth (1998, 2009) has described this type of provisioning a s unspecialized, meaning that the obsidian is likely being exchanged along with other goods and not on commission or as part of a system of elite reciprocity. Furthermore, while it is possible that individuals were familiar with the regions from which t he obsidian came, aside from skilled crafters it is unlikely that the average individual would have been able to distinguish between the various gray obsidian sources. These assertions are supported by the production evidence which, while not from a formal workshop, are indicative of the local production of blades from polyhedral cores. Obsidian was being imported into Santa Rita Corozal in forms ready to produce blades. Given the length of the complete blades present in the assemblage and the general lack of initial series blades, it is likely that the population of Santa Rita Corozal was receiving cores that had already been reduced by having blades removed. If true, then it is highly unlikely that obsidian was being provisioned directly from the communities engaged in the extraction of obsidian. Directly procuring obsidian would result in obsidian artifactual remains that were earlier in the reduction sequence, such as nodules, percussion cores, or larger cores (see, for instance, Demarest et al. 2014: 202 203). Instead, the Santa Rita Corozal assemblage is

PAGE 170

Seidita, Chase, and Chase 175 Table 2 Table displaying the t-test results for high and other status structures. High Status (n=13) Other Status (n=12) P Obsidian Density 0.4896 0.2529 0.0926 Table 3 Table displaying the ANOVA results for high, middle, and low status structures. High Status (n=13) Middle Status (n=4) Low Status (n=8) P Obsidian Density 0.4896 0.2832 0.1924 0.2229 characterized by the presence of final series blades and small exhausted cores. Therefore, we argue that, prior to entering local exchange systems, obsidian was being provisioned from a variety of sources and source regions as part of the general Postclassic Period in terregional exchange and not directly from sources via the involvement of elite political networks. Distribution In addition to considering how the local obsidian economy was provisioned we sought to understand how obsidian was distributed and exchanged at the site level. This was accomplished through the application of Hirths (1998) distributional approach to assess the presence of market exchange. This method is well suited for the Santa Rita Corozal data, especially as the obsidian assemblage represents the end users of obsidian and not necessarily the loci of production, such as a workshop or marketplace. The distributional approach infers the mechanisms of exchange based on the expected distribution of artifacts. Thus, by testing to see how homoge neous the distribution of a specific commodity is across a sites social statuses, it is possible to infer the method of exchange. Following this approach, a homogeneous distribution of nonlocal items is more indicative of a system where there is relativ ely equal access, whereas a heterogeneous distribution is indicative of a system where access is based on social status. Typically, these are conceived as market exchange and redistr ibution exchange, respectively. In our study, we chose to employ surface area of excavations as the standard method of comparison between units. Due to the low lying invisible nature of architecture at Santa Rita Corozal and the complicated construction stratigraphy, we only consider structures and samples for which an unamb iguous Postclassic date is available. In total, 25 domestic structures are included in our sample, and by using status designations generated within previous work (Chase 1992), we were able to consider two different scenarios. The first is a simple elite versus nonelite comparison of 13 high status structures against 12 non high status structures; these were compared via a t test ( Table 2). The second was an ANOVA based on the classification of 13 structures as high status, 4 structures as middle status and 8 structures as low status ( Table 3). The results of these tests demonstrate that no statistically significant differences exist in terms of obsidian densities in either scenario. The average obsidian density of high status structures is 0.489 pieces per square meter while nonhigh status structures contained 0.242 pieces per square meter. A similar trend is seen in the second scenario where middle and low status structures contained 0.192 and 0.283 pieces per square meter, respectively. While high status structures consumed slightly greater quantities of obsidian, they do not appear to have received privileged access to the method of distribution. We attribute these differences to either purchasing power or need. In particular, the elite residents of Platform 2 (Chase and Chase 1988:25 31), one of the most intensely investigated locations at the site, are

PAGE 171

Postclassic Obsidian Production and Exchange at Santa Rita 176 believed to have participated in a variety of activities, both domestic and religious. On Platform 2, Structure 73 may have been the location of an activity that required greater quantities of obsidian as its density is higher than that of nearby Structure 74, where use wear analysis has suggested that the structure was likely the location for specialized tasks perhaps in support of Platform 2 (Hartman 1980). The lack of a statistically significant difference in obsidian densities among social statuses suggests that obsidian was not being distributed by elites along a social hierarchy. Instead, following Hirths (1998) distributional approach, the relatively homogeneous distribution noted for Santa Rita Corozal Postclassic obsidian is more indicative of access to a market system where wealth and /or need impacted distribution. Conclusion Through this study we have been able to characterize the provisioning and distribution of obsidian at the Postclassic provincial capital of Santa Rita Corozal. The Santa Rita Corozal industry appears to be significantly different, in terms of both industry and source exploitation, from other Postclassic sites w ithin the Chetumal province. Analysis of obsidian at Caye Coco and Laguna de On indicate that those populations were not primarily engaged in the production of blades; visual sourcing and limited chemical analysis at these sites suggests that their popula tions drew exclusively upon Guatemalan sources. We suggest that the local obsidian economy for northern Belize was provisioned through participation in the increased interregional exchange of the Postclassic Period. Access to locally produced obsidian bl ades was facilitated through a market system, possibly controlled by Santa Rita Corozal. However, because some obsidian production waste and debitage wound up in archaeological contexts with the consumers of the obsidian industry, potentially some individ uals obtained prepared cores in the market and produced their own blades locally or purchased blades from itinerant crafters working in this economy. Santa Rita Corozals trade networks were also broader than other sites in the area. The center received significant amounts of obsidian form Otumba, Mexico and this occurrence constitutes the only known instance of obsidian from this source in the broader region. While this may be due to Santa Rita Corozals advantageous geographic location on Chetumal Ba y or to its political primacy as the Postclassic capital of Chetumal Province, it may also be the result of not a sufficiently large enough sample of obsidian being subjected to elemental analysis from other sites. It is clear that, with the increased availability of relatively inexpensive, rapid, portable, and archaeologically valid pXRF technology, large samples or entire assemblages of other sites should be assayed without assuming the sourcing of gray obsidian. Had the sample size for Santa Rita Coroz al not been so large, it is entirely possible that we may have missed the presence of obsidian sources from Otumba, Pico de Orizaba, and San Martin Jilotepeque. This analysis provides not only greater refinement of obsidian analysis for Late Postclassic S anta Rita Corozal but also showcases the importance of sampling strategies and sampling sizes in obsidian analysis. Acknowledgements The Corozal Postclassic Project, directed by Diane and Arlen Chase, carried out research at Santa Rita Corozal from 1979 through 1985. Funding for this research was provided by: the University Museum, the Department of Anthropology, and the College of Arts & Sciences of the University of Pennsylvania; a grant from Sigma Xi; grants from the Philadelphia and New York chapters of the Explorers Club; and 2 grants from the National Science Foundation (BNS 8318531 and BNS 8509304); other support for the 1985 field season was provided by the University of Central Florida. Seidita also thanks Brigitte Kovacevich and Michael Callag han at the University of Central Florida for the use of the pXRF to carry out the last round of analysis and Nicholas Tripcevich at the University of California Berkeleys Archaeological Research Facility for the use of the pXRF in the initial study. Publ ications relating to the research carried out by the Corozal Postclassic Project at Nohmul and Santa Rita Corozal are available at www.caracol.org

PAGE 172

Seidita, Chase, and Chase 177 References Anderson, J. Heath and Kenneth Hirth 2008 Obsidian Blade Production for Craft Consumption at Kaminaljuyu. Ancient Mesoamerica 20(1):163172. Braswell, Geoffrey 2003 Obsidian Exchange Spheres In The Postclassic Mesoamerican World, edited by M. Smith and F. Berdan, pp. 131-158. University of Utah Press, Salt Lake City. Cap, Bernadette 2015 How to Know It When We See It: Marketplace identification at the Classic Maya Site of Buenavista del Cayo, Belize. In The Ancient Maya Marketplace: The Archaeology of Transient Space edited by E. King, pp. 111137. University of Arisona Press, Tucson, AZ. Cap, Bernadette, Meaghan Peuramaki -Brown and Jason Yaeger 2015 Shopping for Household Goods at the Buenavista Del Cayo Marketplace. In Research Reports in Belizean Archaeology edited by J. A. J. Morris, pp. 25 36. vol. 12. Institue of Archaeology, NICH, Belize, Belmopan, Belize. Chase, Arlen and Diane Z. Chase 2015 The Domestic Economy of Caracol, Belize: Articulating with the Institutional Economy in an Ancient Maya Urban Setting. Research Reports in Belizea n Archaeology 12:1523. Chase, Diane and Arlen Chase 2014 Ancient Maya Markets and Economic Integration at Caracol, Belize. Ancient Mesoamerica 25:239-250. Chase, Diane Z. 1981 The Maya Postclassic at Santa Rita Corozal. Archaeology 34(1):25-33. 1982 Spatial and Temporal Variability in Postclassic Northern Belize. Dissertation, Department of Anthropology, University of Pennsylvania, Philadelphia, PA. 1985 Ganned But Not Forgotten: Late Postclassic Archaeology and Ritual at Santa Rita Corozal, Belize. In The Lowland Maya Postclassic edited by A. F. Chase and P. M. Rice, pp. 104 125. University of Texas Press, Austin, TX. 1986 Social and Political Organization in the Land of Cacao and Honey: Correlating the Archaeology and Ethnohistory of the Postclass ic Lowland Maya In Lat Lowland Maya Civilization: Classic to Postclassic edited by J. A. Sabloff and E. W. Andrews, pp. 347-377. University of New Mexico, Albuquerque, NM. 1990 The Invisible Maya: Population History and Archaeology at Santa Rita Corozal In Prehistoric Population History in the Maya Lowlands edited by T. P. C. a. D. S. Rice, pp. 199 213. University of New Mexico Press, Albuquerque. 1992 Postclassic Maya Elites: Ethnohistory and Archaeology In Mesoamerican Elites: An Archaeological Asse ssment edited by A. F. Chase and D. Z. Chase, pp. 124-135. University of Oklahoma Press, Nroman, OK. Chase, Diane Z. and Arlen F. Chase 1988 A Postclassic Perspective: Excavations at the Maya Site of Santa Rita Corozal PARI Monograph 4. Precolumbian Art Research Institute, San Francisco. 1989 Routes of Trade and Communication and the Integration of Maya Society: The Vista From Santa Rita Corozal, Belize. In Coastal Maya Trade edited by H. McKillop and P. Healy, Peterborough, Ontario, Canada. 2004 Sant a Rita Corozal: Twenty Years Later. Research Reports in Belizean Archaeology 1:243 255. 2008 Late Postclassic Ritual at Santa Rita Corozal, Belize: Understanding the Archaeology of a Maya Capital City. Research Reports in Belizean Archaeology 5:79 -92. Cl ark, John 1988 The Lithic Artifacts from La Libertad, Chiapas, Mexico: An Economic Perspective. In Papers No. 52, New World Archaeological Foundation Brigham Young University Press, Provo, UT. Clark, John and Douglas Bryant 1997 A Technological Typology of Prismatic Blades and Debitage from Ojo de Agua, Chiapas, Mexico. Ancient Mesoamerica 8(2):111-136. Edwards, Clinton R. 1978 Pre -Columbian Maritime Trade in Mesoamerica. In Mesoamerican Communication Routes and Cultural Contacts edited by C. Navarrete and T. Lee. New World Archaeological Foundation, Provo, UT. Ferguson, Jeffrey 2012 X -Ray Fluorescence of Obsidian In Handheld XRF for Art and Archaeology edited by A. N. Shugar and J. L. Mass. Leuven University Press, KU Leuven, Belgium. Frahm, Ellery 2016 Can I get chips with that? Sourcing small obsidian artifacts down to microdepitage scales with portable XRF. Journal of Archaeological Science 9:448467.

PAGE 173

Postclassic Obsidian Production and Exchange at Santa Rita 178 Glascock, Michael D. and Robert H. Cobean 2002 Summary of Elemental Conc entration Data for the Obsidian Source Groups in Mexico. In A World of Obsidian: The mining and trade of Volcanic Glass in Ancient Mexico edited by R. H. Cobean, pp. 239277. University of Pittsburgh, Pittsburgh. Golitko, Mark and Gary Feinman 2015 Procu rement and Distribution of Pre -Hispanic Mesoamerican Obsidian 900 BC AD 1520: s Social Network Analysis. Journal of Archaeological Method and Theory 22:206247. Hartman, N 1980 A Preliminary Microwear Analysis of Obsidian from Santa Rita Corozal, Belize MSCA Research Papers in Science and Archaeology 1(5):136140. Hirth, Kenneth 1998 The Distributional Approach: A New Way to Identify Market Place Exchange in the Archaeological Record. Current Anthropology 39(4):451 476. 2008 The Economy of Supply: Modeling Obsidian Procurement and Craft Provisioning at a Central Mexican Urban Center. Latin American Antiquity 19(4):435457. 2009 Craft Production, Household Diversification, and Domestic Economy in Prehispanic Mesoamerica. Archaeological Papers of the American Anthropological Association 19(1):1332. Hirth, Kenneth, Bradford Andrews and J. Jeffrey Flenniken 2006 A Technological Analysis of Xochicalco Obsidian Prismatic Blade Production In Merchants, Markets, and Exchange i n the Pre Columbian World, edited by K. Hirth and J. Pillsbury, pp. 6395. Dumbarton Oaks Pre Columbian Symposia and Colloquia, Washington D.C. Kidder, Alfred V. 1947 The Artifacts of Uaxactun, Guatemala. Carnegie Institute of Washington, Washington D.C. Marino, Marc, Lucas Martindale Johnson and Nathan Meissner 2015 Postclassic Tool Production at Santa Rita Corozal: Implications for Domestic Craft Production and Regional Exchange in Flaked Stone In Perspectives on the Ancient Maya of Chetumal Bay edite d by D. Walker, pp. 251263. University Press of Florida, Gainesville, FL. Masson, Marilyn 2000 In the Realm of Nachan Kan. University of Colorado Press, Boulder, CO. Masson, Marilyn and David Freidel 2012 An Argument for Classic -Era Maya Market Exchang e. Journal of Anthropological Archaeology 31:455-484. Millhauser, John K., Lane Fargher, Verenice Y. Heredia Espinoza and Richard Blanton 2015 The geopolitics of obsidian supply in Postclassic Tlaxcallan: A portable X -ray fluorescence study. Journal of Ar chaeological Science 58:133-146. Millhauser, John K., Enrique Rodrguez -Alegra and Michael D. Glascock 2011 Testing the accuracy of portable X -ray fluorescence to study Aztec and Colonial obsidian supply at Xaltocan, Mexico. Journal of Archaeological Sci ence 38:31413152. Popelka -Filcoff, Rachel 2006 Applications of Elemental Analysis for Archaeometric Studies: Analytical and Statistical Methods for Understanding Geochemical Trends in Ceramics, Ochre, and Obsidian, Department of Chemistry, University of Missouri -Columbia, Columbia, Mo. Sabloff, Jeremy and William Rathje 1975 The Rise of Maya Merchant C lass. Scientific American 233(4):72-82. Seidita, Max 2015 Chetumal's Dragonglass: Postclassic Obsidian Production and Exchange at Santa Rita Corozal, Belize, Department of Anthropology, University of Central Florida, Orlando, FL. Shackley, M. Steven 2011 An Introduction to X -Ray Fluorescence (XRF) Analysis in Archaeology In X Ray Fluorescence Spectrometry (XRF) in Geoarchaeology edited by M. S. Shackley, pp. 7 44. Springer, New York. Sheets, Payson 2000 Provisioning the Ceren Households. Ancient Mesoam erica 11(2):217230. Sidrys, Raymond 1976 Classic Maya Obsidian Trade. American Antiquity 41:449-464. Smith, Michael 2003 Key Commodities In The Postclassic Mesoamerican World, edited by M. Smith and F. Berdan, pp. 117-125. The University of Utah Press, Salt Lake City, UT.

PAGE 174

SECTION TWO: GENERAL RESEARCH PAPERS Xunantunich: Virtual Reconstruction of El Castillo during the Late Classic Hats Chaak phase, showing the location of Tut Building on the northeast corner. Xunantunich: Plan of southeastern room of Tut Building showing doorjambs, interior walls, bench surface, and earthquake damage. Excavated in MVPP 2016 Op 13w.

PAGE 175

Xunantunich: Elevation of north wall of southeastern room of Tut Building showing locations of graffiti encountered in the fie ld and through PTM Xunantunich: Illustration of the decapitator scene incised on the north wall of the southeastern room of Tut Building.

PAGE 176

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 181 193 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 17 TAGGED WALLS: THE DI SCOVERY OF ANCIENT MAYA GRAFFITI AT EL CASTILLO, XUNA NTUNICH Leah McCurdy, M. Kathryn Brown, and Neil Dixon Investigations at El Castillo, Xunantunich, uncovered a series of Late Classic rooms on the eastern side of the acropolis. A small private stairway provided access to these rooms suggesting a special function. Our excavations of the southeastern room revealed that it was intentionally filled with clay and stacked stones in a reverential manner. The doorjambs and walls were covered with incised images and designs, ranging from simple sketches (graffiti) to more formal renderings. In this paper we highlight the methods that we used to document the corpus of incised images as well as pres ent our preliminary interpretations of their meaning. We argue that some plastered walls served as canvases for sketching, artistic training, and learned scribal expression. Furthermore, we believe this room was a special place where an ancient Maya sage trained apprentices in the arts and sacred knowledge. Lending support to this interpretation, the walls were partitioned into sections and several images were repeated as if the designs were being practiced. This newly discovered room provides a glimpse into how ancient Maya s acred knowledge was passed on. Introduction Investigations at Xunantunichs main acropolis, El Castillo, revealed a series of Late Classic rooms that belong to a vaulted building under Structure A 5, nicknamed the Tut Building. A private stairway provided access to these rooms, suggesting it may have had a special function. Our excavations of the southeastern room revealed that it was intentionally filled with clay and stacked stones in a reverential manner. The doorjambs and walls were covered with a dense concentration of incised images and designs of the kind usually called graffiti, quite possibly the largest corpus encountered in Belize. Graffiti is a catch all term used to refer to graphic elements and designs incised into plaster surfaces. The term is used to describe a range of designs and does not necessarily impl icate intentional vandalism or destruction, nor counter cultural tagging that the term implies in modern contexts. The term has not been precisely defined. The images in the Tut Building ranged from simple sketches to more formal, detailed renderings. Through the documentation, analysis, and interpretation of this newly discovered corpus of graffiti, we hope to contribute to our understanding of Maya graffiti. In this paper we briefly discuss previous interpretations of graffiti, highlight the metho ds that we used to document the corpus of incised images, and present our preliminary interpretations of their meaning. Studies of Ancient Maya Graffiti For convenience, we use the term graffiti, despite some issues with it. Ancient Maya graffiti has oft en been defined as crude or haphazard incising on plaster surfaces, including walls and benches. Early interpretations of graffiti focused on the unrefined nature of the designs, and a perceived lack of foresight or planning in their execution. Edward H. Thompson and George Dorsey (1898) suggested that graffiti was drawn by some young idler, raising themes echoed by subsequent scholars. Along similar lines, Sir J. Eric S. Thompson (1954: 10) suggested that graffiti was the result of doodling by bored or inattentive novices, or was made by Maya novices before an initiation ceremony. These early considerations represent graffiti as un planned, isolated renderings, executed by individuals and not part of a larger program or institution. Graffiti is often set apart from more formal forms of art due to its seemingly simplistic or unplanned nature. As Kampen (1978: 167; also Trik and Kampe n 2011) observes in his study of the graffiti of Tikal, [O]n every stage of conception and execution, the graffiti appear to lack the logical and organized principles of construction characterizing other varieties of Maya art. He argues that the Tikal g raffiti were terminal acts of desecration of abandoned structures. He argues that examples found within sealed contexts dating to the Classic period represent terminal acts after the building was no longer in

PAGE 177

Ancient Maya Graffiti at El Castillo, Xunantunich 182 formal use, and that examples in final phase, unfilled rooms were the work of later people who lived at the site in the Eznab phase (A.D. 850925). We find this inference problematic, given that much of the final phase graffiti has Classicperiod iconography. Kampen (1978: 169) recognizes this, but he asserts, The fact that certain themes and motifs of Late Classic art appear in the graffiti in no way proves contemporaneity. Although it is plausible that some graffiti may represent terminal acts or desecration by post abandonment squatters, we fin d Kampens explanation overly simplistic. Furthermore, his conclusion that graffiti represent terminal acts forecloses the need for further inquiry to understand the nature of Maya graffiti. More recent interpretations have moved beyond these early explan ations. Haviland and Haviland (1995) applied a psychological lens to graffiti and its production, suggesting that the haphazard nature and inferior quality derives from the mental/psychological state of the artists. They argue that graffiti would be produced under the influence of hallucinatory trances and most likely created by high status people within secluded interiors. Building on Haviland and Haviland (1995), Jason Yaeger (2010) interprets examples of graffiti found within a specialized room in the Late Classic palace of Xunantunich as possible examples of this type of ritual behavior. A recent overarching consideration of graffiti as a subset of Maya material culture was written by Scott Hutson (2011), focusing on the graffiti documented at Tikal. Hutson outlines four previous explanations for the crudeness of graffiti. First, he discusses the possibility that graffiti are practice sketches executed by Maya artists to prepare for future works. In this case, the crudeness is a circumstance of thei r preliminary intention. Second, Hutson (2011) considers the arguments by Kampen (1978) and others that graffiti represents decadent art of post abandonment squatters, and thus that graffiti should not be considered evidence of primary occupational periods. Third, Hutson reiterates Haviland and Havilands (1995) hallucinatory hypothesis. Lastly, Hutson focuses on the possibility that the interpreted crudeness of graffiti derives from the youth of its creators and therefore, lack of training. Hutson (2011) favors this last hypothesis, and argues that many examples of graffiti were created by a ncient Maya children. Furthermore, he states that people of different ages and skill levels left their marks on the same walls and that there was a dialogue betwee n these different artists" (2011: 404). Hutsons interpretations move us towards a more holistic understanding of graffiti and a framework for more systematic analysis. Graffiti, Sages, and Sacred Knowledge Ancient Maya graffiti may have served a number of functions, one of which may have been to facilitate learning and the transfer of specialized knowledge. Building upon Hutsons work, we suggest that the newly discovered graffiti from El Castillo at Xunantunich represents specialized training of young nobles. Incised on the exterior and interior walls and bench of this room were nearly 200 graffito. This room was intentionally filled and built upon as the foundation for a Late/Terminal Classic structure (Structure A 5). All graffiti was buried by this filling episode and thus, predated it. In addition, while the majority of elements in the room were crudely executed, there are a number of elements that cannot be labeled crude by any measure and could only have been created by accomplished and lear ned artists, or sages. Furthermore, there is clear evidence of practice through repeated designs. This diversity in a single room indicates that graffiti was not just produced by untrained artists, but at least in a few examples, was incised by highly sk illed artists. These lines of evidence from Xunantunich lead us to suggest that this room was a place of learning and practice. We suggest that the graffiti found in Tut Building are practice and preparatory sketches executed by novices and sages within a scribal training and/or sage academy housed at El Castillo during the Late Classic period (Brown and McCurdy 2017). As keepers of time, history, and sacred knowledge, Maya scribes and sages required long periods of training in arts and ideological tradit ions (Brown and McCurdy 2017). We see direct evidence of such training, in a mythological sense, on an unprovenienced Late Classic ceramic vessel housed at the Kimbell Art

PAGE 178

McCurdy, Brown, and Dixon 183 Figure 1 Map of El Castillo acropolis showing the l ocation of Tut Building and Operation 13 excavations on the northeast corner. Museum in Fort Worth, Texas. The vessel depicts two separate scenes of Itzamna instructing two pairs of juveniles or young men (Rossi 2017). This vessel provides a rare example of Late Classic pedagogy (Rossi 2017). Further, recent findings by William Saturno and colleagues (2012) at the Guatemalan site of Xultun provide evidence of noble training and education practices. At Los Sabios, a high status res idence within one of Xultuns administrative areas, Saturno et al. (2012) discovered a well preserved room interior with intact wall murals. The room also features incised calendar notations projecting future dates, indicating that astronomical and calend rical training was primary in this space. This space contains evidence of specialized knowledge being generated, curated and transmitted, alongside evidence for codex book creation and inscription (Rossi 2017: 9). An informative analogy for such educati on in Mesoamerican cultures can be drawn from the late Postclassic Aztec Calmecac (Berdan 2014; Calnek 1988), or academy for children of nobles and talented commoners. In the Calmecac, young males were schooled in sacred knowledge of religion, ritual, rea ding, writing, the calendar, and military tactics

PAGE 179

Ancient Maya Graffiti at El Castillo, Xunantunich 184 Figure 2 Virtual Reconstruction of El Castillo during the Late Classic Hats Chaak phase, showing the location of Tut Building on the northeast corner. Figure 3 Plan of southeastern room of Tut Building showing doorjambs, interior walls, bench surface, and earthquake damage. Excavated in MVPP 2016 Op 13w.

PAGE 180

McCurdy, Brown, and Dixon 185 (Berdan 2014). Taking the Xultun case as a complementary example and the Aztec Calmecac as an analog y, we explore the idea that the southeastern room of Tut Building (and potentially the entire structure) was a space designated for specialized training of young nobles by learned individuals such as scribes and sages. Excavations and Findings at Tut Buil ding In 2016, The Mopan Valley Preclassic Project (MVPP) expanded earlier investigations of the northeast corner of El Castillo acropolis where Tut Building is located (Figure 1). As shown in the virtual reconstruction (Figure 2), Tut Building sits on the Medial Terrace level of the acropolis and likely dates to the early parts of the Late Classic Hats Chaak phase (approximately 670700 CE). We began excavations of Tut Building i n 2012, with a preliminary 5meter trench situated east west along the southern wall. Previous work in the area was conducted by the Tourism Development Project (TDP) directed by Jamie Awe. We exposed TDPs excavation area and expanded to the west to inv estigate the extent of the preservation. Though there was obvious loss to the structure on the east side, our excavations revealed good evidence of a well preserved and intentionally blocked doorjamb leading into a room interior. In 2013, we returned to i nvestigate the doorjamb and interior. We documented some collapse to the south of the structure above a well preserved series of platform surfaces supporting the Tut Building as earlier construction episodes (see McCurdy 2016). Our excavations centered o n the fill behind the blocked entranceway, without entering the room interior. The fill consisted of dense marl deposits with relatively few artifacts. The only major distinctions we encountered within this fill were columns of dark, organic sediments on either side of the entranceway that continued into the interior fill. This fill reached from the interior floor to the center of the partially surviving corbel vault, which we were able to quickly document before ending our season in 2013. In 2014 and 2015, our efforts were diverted to the eastern architectural platforms that underlie Tut Building and to ceramic analysis (see McCurdy 2016). We returned to the southeastern room in 2016 with the goal of entering the interior and understanding the purpose of the dark, organic sediments. We hypothesized that these were intentionally placed to protect painted elements, potentially murals, on the interior plaster walls. We began by exposing the vault and corbels from above. We removed (after fully documenti ng and marking) only those corbels that posed safety threats as we continued excavations below. We revealed that the north vault half and the western gable wall survived almost in their entirety while the eastern gable wall suffered loss on its southern s ide. We installed as system of wood braces as we removed fill beneath the surviving vault to ensure safety and stability of the structure. The room was intentionally filled through the sequential construction of eight retaining walls. Construction debris and sediments constitute the majority of fill materials. On the eastern side of the room (approximately 50 cm east of the eastern doorjamb along the southern wall), we encountered a concentration of orange/tan volcanic ash chunks with leaf and tree branc h impressions placed approximately 1 m above the original floor surface. In addition, beneath this layer of intentionally cached volcanic material, we also encountered a thick deposit of fine orange red dry clay. We believe that it is likely that these a sh chunks were acquired via trade from a volcanic region to be used as temper in ceramic production with high quality local clay. After removing all fill, we were able to fully define this southeastern room of Tut Building. The interior space measures approximately 2.9 m (N S) by 4.3 m (E W) (Figure 3). The walls stand approximately 2.3 m in height, supporting the vault with an apex of approximately 4 m prior to partial collapse. The room contains a very large masonry bench that covers almost the entire floor area, with only a 75 cm area between the southern interior wall and the northern edge of the bench (see Figure 3). As indicated in Figure 3, there is a very large disturbance extending from the northwestern corner across the bench and into the southeastern corner of the room. This is most likely damage

PAGE 181

Ancient Maya Graffiti at El Castillo, Xunantunich 186 Figure 4 Elevation of north wall of southeastern room of Tut Building showing locations of graffiti encountered in the field and through PTM. Figure 5 Examples of graffiti from the southeastern room interior of Tut Building (not to scale): (a) full-body human figure dressed like a king; (b) anthropomorphic head in profile with fabric cap and ear ornament; (c) zoomorphic/dog in profile; (d) itzaat glyp h; (e) illustration of ballplayer engaged in play with twisted serpent images.

PAGE 182

McCurdy, Brown, and Dixon 187 resulting from earthquake movement and subsequent settlement. The damage seen within the room and at the top of the vault may actually be the result of two sepa rate earthquakes (Jorge Can personal communication 2016). Despite this damage, the plaster within the room survives quite well. Our investigations of the interior plaster surfaces revealed a total of 192 incised elements and one painted element across bot h doorjambs all interior walls, as well as the face and surface of the interior bench. Additional graffiti elements were documented on the exterior walls during the field seasons of 2016 and 2017 and will be discussed in future publications. The western doorjamb contains the densest concentration of graffiti elements (40 elements over 1 x 2 m surface area). The western interior wall also contains a high concentration of designs with the concentration decreasing clockwise around the room. In order to thoroughly document the incised images on the plaster walls and benches, we handmapped approximately 75 individual elements exposed within the southeastern room. Due to time constraints, we chose a core of significant elements to document throug h mapping. Our goal was to document these finds as fully as possible and develop a detailed record for future analyses of form, tool use, quality of line, and superimposition of elements. A team of four mappers used inch screen to construct scaled draw ing grids that could be hung and/or stabilized on vertical and horizontal surfaces without damaging plaster. We mapped 1:1 renderings of each element, focusing on: (1) position, form, and proximity to nearby elements; (2) weight, quality, and differences in incised or gouged line; (3) layered sequences of superimposition; (4) contextual markings including scored panel lines; (5) damage to plaster as it affected the visibility and preservation of individual incised images and their surroundings. In addition to creating a detailed record, this mapping strategy offered the opportunity to study these elements in extreme detail. It was through this experience that many insights regarding content, style, and technique were developed. In addition to the time consuming, but necessary, hand mapping methods utilized for documentation, we completed a comprehensive photographic documentation of all plaster surfaces within the chamber. Figure 4 illustrates the elevation of the north wall showing 19 graffiti elements r ecognized in the field and the additional elements tentatively recognized in 2016 through analysis of the Polynomial Texture Mapping (PTM) models created by co author Neil Dixon. PTM is a photographic technique using multiple light sources to construct di fferentially textured composite images of surfaces and allows for high resolution digital analysis. Sequentially firing camera flashes positioned full circle around a focal point (such as a section of wall with graffiti) creates a series of images that ar e merged by Dixon into PTM models. Our PTM analysis consisted of using an open source PTM/RTI viewer to digitally inspect these composite images of each major surface. Similar to raking light photography techniques, PTM/RTI software allows the user to di gitally manipulate the light source to simulate passing a flashlight over something in real life. The changes in light direction allow for more clarity in viewing the incised graffiti designs. The PTM composites also serve as high resolution digital docu mentation of all surfaces in the room PTM analysis of the north wall after the 2016 field season resulted in recognizing over 49 potential graffiti elements that we did not recognize with the naked eye. In total, PTM analysis of all surfaces resulted in a catalog of 91 potential new elements. In 2017, we groundtruthed these potential new elements and found that 56 were indeed graffiti we had not recognized during our investigations in 2016. During the groundtruthing process, we also encountered an add itional 17 elements previously unrecognized. These results will be fully documented in future publications. The G raffiti: Content & Iconography In terms of content, graffiti designs in the southeastern room of Tut Building include (1) full body anthropomorphic figures (e.g. Figure 5a); (2) anthropomorphic heads with embellishments such as hats and ear ornaments (e.g. Figure 5b); (3) deity figures (e.g. Figure 9); (4) isolated anthropomorphic facial features such as noses; (5) isolated anthropomorphic body parts; (6) full zoomorphic figures (e.g.

PAGE 183

Ancient Maya Graffiti at El Castillo, Xunantunich 188 Figure 6 Bar graph comparing the frequency of 15 categories of graffiti imagery, including unknown imagery. Figure 7 Illustration of the decapitator scene incised on the north wall of the southeastern room of Tut Building.

PAGE 184

McCurdy, Brown, and Dixon 189 Figure 5c); (7) zoomorphic heads; (8) isolated zoomorphic body parts; (9) zoomorphic facial features; (10) hieroglyphs (e.g. Figure 5d); (11) ritual scenes (e.g. Figures 5e, 7); (12) architectural features; (13) simple or decorated shapes; (14) line based elements; (15) markings such as tick marks. Figure 6 organizes the collection of elements in terms of the frequency of each category. Anthropomorphic heads and anthropomorphic figures are the most commonly represented elements. One of the most remarkable elements is the ritual scene on the west wall representing a ballplayer actively portrayed in a playing stance with a series of twisted serpent accompaniments and a potential associated sacrificial victim (Figure 5e). More direct images of sacrifice are included on the north wall where a pair of regaliaclad and masked warriors/priests armed with curved blades have already decapitated a victim placed between them (Figure 7). These decapitators carry the heads of their victims by the hair. There are several other depictions of warriors holding spears. This may indicate that the graffiti artists/novices were also engaging in military training, as in the Calmecac analogy described above. While all room walls retain segm ents of red wall paint, the north wall features the only graphic element painted with black pigment that we have been able to identify thus far (Figure 8). While it appears that this element only partially survives, it is also possible that the image was never fully completed. This element is reminiscent of a large cartouche, however it does not appear to have enclosed a glyph. Instead, it serves as a frame for painted designs within its interior space. A zoomorphic or potentially avian figure is attach ed to the main outline in the upper left quadrant. There are additional decorative elements in the middle including potential representations of feathers, flowing water, and zoomorphic eyes. The preservation and/or incompleteness of the painting make ide ntification of these elements difficult at this time. We are exploring options for techniques that will produce distinct views of this painted element so that we might assess the interior iconography. Our interpretation of the meaning of each graffito is still on going. There are patterns that can be distinguished from element content that may reflect the training agenda or primary goals of scribal education in this special place of learning. The Los Sabios chamber from Xultun, discussed above, appears t o have focused on astronomical knowledge. The room within the Tut Building may have been a place where young nobles learned to depict anthropomorphic and zoomorphic forms as well as to render possible historic events. Many of the incised images within th e room were the size that would adorn ceramic vessels and, therefore, may represent practice for renderings on a different medium. Although tentative, the volcanic chunks and fine clay found within the filled room, may lend support to this idea. Below, w e discuss the lines of evidence that support our hypothesis that the graffiti may represent pedagogical activities within this specialized room. Indications of Training and Artistry First, several groupings of graffiti consist of repeated elements, often with slight variations between each element, and possible evidence for assessment. For example, the west wall features a series of at least four jaguar or puma heads drawn side by side within scored panels (see below). Each iteration clearly depicts a fe line head, but the formation of the ears, eyes, and mouth vary. There is differential use of a curly ear versus a rounded ear as well as a slit like eye versus a triangular eye. These differences indicate that the repetition may have been a way to practi ce or experiment with these features. Additionally, some of the heads appear to have a large X gouged through them. This may indicate that some sort of technique focused or content assessment occurred within the room whereby trainees were tasked with representing specific themes (i.e. jaguar heads) until they or their mentor w ere satisfied with the product. Secondly, t he room features a series of gouged lines that appear to form separated workspaces for incising. On the north wall (see Figure 4), the artists scored vertical lines from the level of the bench surface almost to the vault spring and horizontal lines from t he northeast corner to approximately halfway across the wall. This created a series of square segments that may have been partitioned for individual artists

PAGE 185

Ancient Maya Graffiti at El Castillo, Xunantunich 190 Figure 8 Painted frame element (likely partial) on the north wall of the southeastern room of Tut Building. Figure 9 Calligraphic Chac Xib Chac rendering incised on the east wall of the southeastern room of Tut Building.

PAGE 186

McCurdy, Brown, and Dixon 191 or may represent wor kspaces over a certain period of time. On the west wall, vertical panels were created by gouging a series of vertical lines across the wall, likely serving the same function. A further indication of training derives from the collection of isolated element s such as profile noses that appear to be abandoned for doovers nearby. On the western doorjamb, there are a series of attempts to draw a nose in profile that appear to be abandoned. It seems that if a nose in its first rendering was unsatisfactory, the artist moved slightly over on the wall and began anew, oftentimes completing a full head in profile. A high concentration of doovers is located on the west wall where artists were attempting to represent the twisted snake motif associated with the b allplayer (see Figure 5e). The overlapping snake bodies are difficult to render without crossing erroneous lines. It appears that a novice took great pains to practice this motif so as to properly detail the crisscross pattern and real istically end with snake heads. Approximately eye level on the western doorjamb, an artist incised the full figure profile of a man dressed in kingly regalia with royal headgear (see Figure 5a). This representation was rendered at a scale and in a style commensurate with im ages of nobles painted on ceramic vases. What we refer to as graffiti may have been both a training mechanism and technique used by artists to plan for future artworks. As mentioned above, our findings of materials used in ceramic production may indicate that additional stages in art production processes occurred within close proximity of the Tut Building. The technical qualities of incising into plaster may indicate that this preparation applied more to the subtractive arts, such as stone carving. Jamie Awe (personal communication 2017) reported that TDP encountered a small stone block carved with a standing figure in profile within the vicinity of Tut Building. Though we are unable to directly tie this carved stone block to the southeastern room, it is possible that it represents a discarded product from a later stage within a production sequence that began with incising into plaster. As mentioned above, the collection of graffiti in the southeastern room evidences differential skill and the presence of talented and proficient artists. The most remarkable elements have been identified by Christophe Helmke (personal communication 2017) as the work of well trained and experienced artists with specialized knowledge of iconographic and hieroglyphic form and content. For example, just above the bench surface on the east wall a deity rendering (Figure 9) was finely incised with a calligraphic quality similar to the whiplash style of ceramic decoration recognized by Coe (1973:91 cited by Doyle and Houston 2017). According to Helmke (personal communication 2017), this rendering represents the Chac Xib Chac figure known from Palenque in a highly decorative and evocative style. Just above this calligraphic rendering, a hieroglyph (see Figure 5d) was executed i n a more straightforward style but represents a title very relevant to our hypothesis that this room served as a sage academy (Brown and McCurdy 2017). Helmke (personal communication 2017) translates this glyph as [i]tza ta or itzaat meaning wiseman or sage. In addition to the knowledge of writing embedded in this element, it is a direct indication of the type of person who likely used this space and trained others to reach the status of sage. Conclusions In summary, we suggest that the southeaste rn room of Tut Building at Xunantunich may have functioned as a place of specialized training for young nobles. The practice and preparatory sketches incised on the plaster walls may have been educational tools and exercises to ensure that novices learned important skills and symbolism. While we can generally date the use of the room to the Late Classic Hats Chaak period (see McCurdy 2016), we have not yet determined how long the space was used. It is possible that the walls could have been used for months or even years. A fresh coat of limewash or plaster could fill in previous incised areas and offer a brandnew surface upon which to work. We found indications of multiple plaster layers on the exterior walls of Tut Building in the 2017 field season and plan to follow up with plaster coring to determine the number of resurfacings. In addition, MVPP investigations supervised by Kit Nelson and Jennifer Cochran

PAGE 187

Ancient Maya Graffiti at El Castillo, Xunantunich 192 in 2017 focused on several buildings of Group C located directly south of El Castillo. Building upon the Calmecac model, Brown hypothesizes that the novices being trained at Tut Building would have likely stayed for periods of time at th e city center and thus would require accommodations (Brown et al. 2017). We are currently testing Group C structures for evidence that can link them to Tut Building and indications that they served as living spaces for novices. Our recent findings of gra ffiti and additional patolli boards on benches at Group C are very promising. As an overarching conclusion to this paper and a beginning to renewed considerations of graffiti, we hope that our investigations at El Castillo add to the current dialogue surro unding the source, purpose, and significance of what we call graffiti today. Indeed, we encourage efforts to determine better and more appropriate terms for incised designs on plaster walls because we feel the term graffiti evokes negative connotations that are not evidenced in our findings at Xunantunich. The diversity, quality, and specific content in Tut Building, should motivate a revision of our understanding of this set of material culture and the behaviors that created it. Acknowledgements We wou ld like to thank the Belize Institute of Archaeology for allowing us the opportunity to conduct research in the country of Belize. We are particularly appreciative of the IA staff for their dedication and assistance. We are thankful for the assistance an d support of all members of the Mopan Valley Preclassic Project and Mopan Valley Archaeological Project staff, volunteers, and UTSA Field School students. This research is funded by the Alphawood Foundation, the University of Texas at San Antonio, Ben and Trudy Termini, and Santiago and Mary Ester Escabedo. These excavations were undertaken in collaboration with local archaeological excavators. The supervising local excavator was Ronaldo Tut, who has excavated at Xunantunich and other sites since 1996. Other excavators and assistants included Armin Puc, Santiago Chuc, Emerson Cano, and Robert Cunil. Ismael Chan, Emerson Cano, Armin Puc, and Christopher Chan also greatly contributed to the documentation of the finds within the room. References Berdan, Frances F. 2014 Aztec Archaeology and Ethnohistory. Cambri dge University Press, New York. Brown, M. Kathryn and Leah McCurdy 2017 The Chamber of Secrets at Xunantunich. Paper presented at the Santa Fe Institute Third Maya Working Group: Telling Time, Myth History and Everyday Life in the Ancient Maya World. Santa Fe, New Mexico. Brown, M. Kathryn, Leah McCurdy, and Kit Nelson 2017 Investigating Pedagogical Practices of the Ancient Maya: An Example from Xunantunich. Paper presented at the 8th Annual Sout hcentral Conference on Mesoamerica. Tulane University, New Orleans. Calnek, Edward 1988 The Calmecac and Telpochcalli in pre-conquest Tenochtitlan. In the Work of Bernardino de Sahagun: Pioneer Ethnographer of the SixteenthCentury Aztec Mexico. Edited by J.J. Klor de Alva, H. Nicholson, and E. Quiones Keber. SUNY Albany, New York. Coe, Michael D 1973 The Maya Scribe and His World. New York, Grolier Club. Doyle, James and Stephen Houston 2017 The Universe in a Maya Plate. Maya Decipherment: Ideas on Ancient Maya Writing and Iconography. Accessed November 1, 2017. https://decipherment.wordpress.com/category/vesse ls/ Haviland, William, and Anita de Laguna H aviland 1995 Glimpses of the Supernatural: Altered States of Consciousness and the Graffiti of Tikal, Guatemala. Latin American Antiquity 6:295 309. Hutson, Scott R. 2011 The art of becoming: The graffiti of Tikal, Guatemala. Latin American Antiquity 22(4 ): 403 426. Kampen, Michael 1978 The Graffiti of Tikal. Estudios de Cultura Maya 11:155 179. McCurdy, Leah 2016 Building Xunantunich: Public in an Ancient Maya Community. PhD Dissertation, Department of Anthropology, The University of Texas at San Antonio.

PAGE 188

McCurdy, Brown, and Dixon 193 Rossi, Franco D. 2017 Pedagogy and State: An Archaeological Inquiry into Classic Maya Educational Practice. Cambridge Archaeological Journal : 1 -18. DOI: 10.1017/S0959774317000580. Saturno, William A., Anthony Aveni, Da vid Stuart, and Franco D. Rossi 2012 Ancient Maya Calendrical Tables from Xultun, Guatemala. Science 33: 714 717. Thompson, Edward H., and George A. Dorsey 1898 Ruins of Xkickmook, Yucatan. Field Columbian Museum Publication 28, Anthropological Series Vol II, No. 3. Field Columbian Museum, Chicago. Thompson, J. Eric S. 1954 The Rise and Fall of Maya Civilization University of Oklahoma Press, Norman. Trik, H. and M. E. Kampen 2011 The Graffiti of Tikal: Tikal Report 31 (Vol. 31). Philadelphia: University of Pennsylvania Press. Yaeger, Jason 2010 Shifting Political Dynamics as Seen from the Xunantunich Palace. In Classic Maya Provincial Politics: Xunantunich and its Hinterlands, ed. by L.J. LeCount & J. Yaeger, pp. 145 160. University of Arizona Press, Tucson.

PAGE 189

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 195 205 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 18 TWO UNUSUAL FINDS FR O M COURTYARD 3, PACBI TUN, BELIZE Sheldon Skaggs and Terry G. Powis Recent investigations in a large, enclosed courtyard on the southwest corner of the ancient Maya site of Pacbitun, Belize, revealed evidence of successive emplacements of ritually important deposits within its plaza. Initial analysis of the stratigraphy and ceramic material suggests that the entire courtyard plaza has only one or two floors, with construction and use during the Late Classic period (AD 550 800). Graves with slate capstones were found west of the central excavation unit. Either these were initially intrusive through the plaza floor, or subsequent revisiting of the burials breached the plaza floor. A burial on top of the western -most slate capstones was particularly interesting, with associated fragments of a partially restorable Ulua Valley style carved marble vase and a pair of carved shell atlatl finger loops as grave goods. Introduction During renewed excavations into what is believed to be an elite residential and administrative area of Pacbitun, i.e. Courtyard 3, we encountered two unusual artifacts. Neither the Ulua Valley style carved marble vase, nor carved shell atlatl finger loops, have ever been found before at Pacbitun. In fact, these are rare finds for any site in the Maya Lowlands. Since these artifacts were found together in a single burial, they help identify the social status of the individual interred and stylistically date the burial to the latter half of the Late Classic Period (Coc Phase, AD 550800). Description of Pacbitun Pacbitun is a medium sized site located along the southern rim of the Belize Valley and on a limestone plateau adjacent to the Maya Mountains. The site core is located three kilometers to the east of San Antonio village in the Cayo District, Belize ( Figure 1 ). Its position intersects two ecozones with resulting access to a wide variety of economic and ritual resources including granite, slate, pine, springs, and fertile agricultural land (Healy 1990:248). The habitation zones are archaeologically defined as three separate areas based on presumed function and population density: t he Epicenter, the Core, and Periphery Zones (Healy et al. 2007:17). The Epicenter is the location of the main religious and administrative structures. It sits on an artificially leveled hill, oriented east west, and has 41 known masonry buildings, three main plazas, and an additional two plazas that are adjacent to the north side, labeled Plazas A to E ( Figure 2; Healy 1990:250). An additional architectural group located to the northeast of Plaza A was designated as the Eastern Court (Cheong 2013). A la rge reservoir is located just north of this architectural complex. The Core Zone encompasses the Epicenter and a one square kilometer buffer around it (Campbell Trithart 1990). The area beyond the Epicenter is dotted with small mounds, although a few larg er structures, courtyard groups, agricultural terraces, springs or reservoirs, and four sinkholes are also present (Healy et al. 2007:18, Figure 3; Richie 1990; Spena rd et al. 2012; Sunahara 1995). The Periphery Zone is estimated to cover nine square kilometers around the site. It is the agricultural area for the site center, consisting of several hundred small house mounds spread over the landscape, as well as several smaller (~5 m tall), hilltop pyramidal structures, plaza groups, minor centers, an d terraces (Spenard 2011; Turner et al. 2015; Ward 2013; Weber 2011; Weber and Michelet ti 2016; Weber and Powis 2014). Previous Excavations of the Pacbitun Courtyards At the southern end of the Epicenter are three enclosed courtyards, flanked by 13 struct ures ( Figure 3 ). These courtyards and surrounding structures are generally described as the palace of the Epicenter, due to the surrounding range structures, the sequential elevation differences from Courtyards 1 to 3, and the restricted access to each courtyard. Investigations into the northern structure of Courtyard 2 (Structure 23) were carried out by Cassandra Bill and Paul Healy in 1986 (Bill 1987; Healy 1990; Healy et al. 1995; Healy et al.

PAGE 190

Two Unusual Finds from Courtyard 3, Pacbitun, Belize 196 Figure 1 Map of Belize Valley showing location of Pacbitun, and all the sites mentioned in the text. Red arrow highlights location of Pacbitun, Belize. Figure 2 Map of Pacbitun Site Core including Structure 10.

PAGE 191

Skaggs and Powis 197 Figure 3 Map of Pacbitun Site around Courtyards 1 to 3 showing all excavation units (Unit numbering in zoomed portion leave off the initial 17 -B-CT3 of all unit name).

PAGE 192

Two Unusual Finds from Courtyard 3, Pacbitun, Belize 198 2007). Additional excavations of the eastern most structure (Structure 25) in 2016 (Skaggs et al 2017), the northern end of Courtyard 3 (Structures 22 and 33) (Pierce and Skaggs n.d.), and the centers of Courtyards 2 and 3 in 2017 (Skaggs and Cherico n.d.) provide additional evidence on the nature of the palace area. Excavations into the plazas of all three courtyards reveal only one or two plaster floors above bedrock. The courtyard plaza construction differs considerably through time as each courtyard required differing amounts of leveling before the first plaster floor was laid down. A 3 m by 3 m excavation unit was placed in the center of Courtyard 2, revealing a wall that measured 0.7 m high. It was six courses high and two courses wide running NE to SW across the northern side of the unit ( Figure 4). A midden like layer containing animal bone, jute ceramics, and a few pieces of jade was found covering the wall. Preliminary ceramic analysis suggests the midden is the same late Middle Preclassic (600 300 BC) depositional event that occurred in Plazas A and B. This midden was ultimately covered by the first limestone plaster floor (Skaggs and Cherico n.d.). Bills (1987) trench into Courtyard 2, under Structure 23, and into Plaza B also encountered a single course wide wall around 0.7 m high; however, in her operations the wall ran directly north to south, and the facing was east in Courtyard 2 and west under Structure 23 and in Plaza B. The proximity of the east facing wall in Bills (1987) operation and the south facing wall in the center of Courtyard 2 excav ation from 2017 suggests these may be part of the same construction. The west facing walls under Structure 23 and Plaza B suggest there is more than one structure, and covering a larger area than the earliest structures in Plaza B (Crow and Powis n.d.). A single radiocarbon date, in material Bill characterized as similar to but not midden, from next to the wall returned a date range of 40 BC AD 220 (Bill 1987:123128) suggesting the wall was covered in or before the Terminal Preclassic (Ku Phase, 100 BC AD 300). Courtyard 2 clearly had early occupation of some sort, and in buildings unlike those found in either Plaza A or B. Excavations in Courtyards 1 and 3 show little evidence of such early occupation, Figure 4 Middle Preclassic (600 -400 BC) wall excavated in northwest corner of center unit of Courtyard 2. Photo courtesy of Mike Lawrence. Figure 5 Photograph looking north into units 17-B-CT3 4 to 6, showing the two long slate slabs (right side) leaning against haphazard ci st grave one western side of excavation unit 17 -B-CT3 -1 in Courtyard 3. Smaller slate slabs which are part of another grave are see on the left side. however. In Courtyard 1, under the north end of Structure 25 and presumably under Structure 24, evidenc e of a Preclassic midden was encountered, but no charcoal useful for dating purposes was found. Just above the midden on the north end, a fill layer was found containing a Pucte Brown, Hermitage Complex ceramic dish with sooting located on its interior su rface. A sample of the sooting was submitted for radiocarbon dating. Results indicated a Late Classic (AD 550 650) date. Furthermore, charcoal found in a unit just above bedrock centered in front of Structure 25 yielded a radiocarbon date of AD 640 675. In these central units, the first plaster floor sits on top of core. Limestone boulders ranging from 0.6 m to 1.5 m in diameter were stacked nearly two meters high before ballast and plaster was laid down on top. Structure 25 was built on top of

PAGE 193

Skaggs and Powis 199 this first plaster floor surface. A fac ing found in the core suggests there may have been an earlier platform built directly on bedrock, which was subsequently covered by later core (Skaggs et al. 2017). Courtyard 3 has no evidence at all for construction before the Late Classic Period. The central excavation unit revealed only one preserved plaster floor located above a thick white marl layer that sat directly on bedrock. Charcoal from two primarily Dolphin Head Red, Spanish Lookout Complex ceramic caches cut into this plaster floor dated to AD 552648 and AD 545 645, respectively (Skaggs and Cherico n.d.). These caches were between graves in Unit 17B CT3 1, both of which will be discussed in future publications. 2017 Excavations in Courtyard 3 Over t he course of the 2017 field season at Pacbitun, we conducted operations in the plaza of Courtyard 3. From an elevation point of view, Courtyard 3 is the deepest of the three courtyards making up the palace area of Pacbitun. We placed a large 3 m by 3 m excavation unit (17B CT3 1) in the center of the plaza to get its chronological history. During our investigations, we identified five caches and two burials in this initial unit. The graves extended outside the unit, so we continually expanded to the west in order to fully expose the haphazard capstone cist graves (after Welsh 1988). During the expansion, we found additional graves above and below slate slabs. These additional burials were found in highly disturbed contexts. The two unusual artifact s, forming the basis of this paper, were found during the investigatio ns into these special deposits. To the west of the capstone cist burials in Unit 17B CT3 1, we found two long (1.3 m) slate slabs ( Figure 5 ). While defining the extent of the slabs, we encountered smaller slate slabs and a 1 m wide column of burnt limestone rocks ( Figure 6 ). These smaller slabs were laid down horizontally to form the top of a haphazard cist under the column of limestone rocks. The column itself was covering part of a burial which sat on top of the slate slabs. We are labeling this column of burnt limestone rocks a cairn capstone cist, making the grave a dual cist Figure 6 Photograph of the column of burnt limestone found in center excavation units of Courtyard 3. The limestone terminated below the humus layer, and was piled up on a burial sitting above a slate capstone cist seen in Figure 5. Photo courtesy of Jeff Powis. Figure 7 Map of Units 17 -B-CT3 -5 and 6 and burials in blue and purple with slate slabs in yellow. Left, position of skeleton and grave goods (including two unusual artifacts) above small slate slabs; right, position of skeleton without grave goods found below sm all slate slabs. arrangement ( Figure 7 and 8). Previously, slate capstones have only been associated with elite Tomb 19, found in Structure 1 at Pacbitun (see Figure 2 ). There are similarities in the presumed status of the individual found in Tomb 1 9 and our burial based on the artifacts we found. Tomb 19 was placed 5 m below the axial stair in Structure 1, and is the only vaulted tomb ever found at the site (Healy 1990; Healy et al. 2004). In the tomb, the skeleton of only one individual was found under a layer of thousands of chert flakes and cores, with the head and torso also covered in cinnabar. The burial contained 19 slipped vessels, with a quarter of them having polychrome designs. Additionally, one painted marine valve, three polished jade beads, one pyrite tube, a pair of circular shell earspools, five

PAGE 194

Two Unusual Finds from Courtyard 3, Pacbitun, Belize 200 Figure 8 Cartoon of vertical arrangement of burials and finds above and below small slate slabs along imaginary N S line running through the grave. Some finds are further east of west than others in the cartoon compared to position of the line. V represents vase fragments, J represents jade beads, Loop represents shell atlatl finger loops. The stratigraphy of units was highly disturbed, but the da sh dash -dot line represents bottom of topsoil and possible ballast floor layer found in other units, while the long dashdot line represents the plaster floor found in other units. hollow bone tubes, and a drilled circular slate mirror backing were also r ecovered (Healy et al. 2004). 2017 Unusual Finds in Courtyard 3 The first unusual artifact was found across multiple excavation units. Fragments of it started appearing about 40 cm below the ground surface in the northwestern corner of the 3 m by 3 m uni t. The artifact is an Ulua Valley style carved marble vase ( Figure 9). These vases were carved during the Late Classic Period (AD 650850) from white marble blocks located in the Ulua Valley of Northwest Honduras (Luke et al. 2006). The vase was broken and scattered across the plaza, but most fragments were found among the rocks of the cairn capstone cist. Other pieces o f the marble vase were found in association with, and under the skeletal material of the cairn capstone burial CT32, which was just above the horizontal slate slabs ( Figure 10 ). Figure 9 Photograph of largest portion of the Ulua Valley -style carved marble vase. Photo courtesy of Jeff Powis. These vases are rare finds, with only 153 fragmentary or whole vessels in museum collect ions, and only 53 were excavated by what was considered professional archaeologists at the time of discovery (DavisSalazar et al. 2007; Wells et al. 2014). Only one whole vase, from a burial in Palmarejo, Honduras (Wells et al. 2014), has been recovered and analyzed with the full suite of modern archaeological techniques. Pacbitun now joins one of only four sites in the lowland Maya Region where fragments of these vases have been found. Fragments from the other find sites (Altun Ha, Chac Balam, San Jose and Uaxactun) have stylistic elements similar to the vase found at Pacbitun. These features, such as taller cylinder height to diameter dimensions, frontal zoomorphic heads, feline handles, and borders with voussure (half moon) motifs, date to the latte r half of the production time range (Luke 2010). Through isotopic analysis of a number of these marble vases, the source of these luxury craft goods has

PAGE 195

Skaggs and Powis 201 Figure 10 Upper photograph of a fragment of the Ulua Valley -style carved marble vase fragment in contact with skull of cairn capstone cist burial (CT3 -2) above the slate slabs. In lower photograph, red circle shows human bone, blue circle is around the skull, grey circles show areas of marble vase fragments, green circle is area of the jade pieces s hown in Figure 11, and the white curves are location of atlatl finger loops (described later) at 10cm below level shown. been correlated to production in Travesia (Luke et al., 2006; Luke and Tykot, 2007). The locations of lowland Maya vase finds are associated with elite locations, such as temples and palace area caches or burials, dating to Late Classic or Terminal Classic Periods (Luke 2010). The Pacbitun vase, although not complete, was reconstructed in order to determine its dimensions. It stands at a height of 24.5 cm (although it may be up to 5 cm taller as there is evidence of a ring base that broke off), and an outside diameter of 15.2 cm. The walls of the vase are carved to 0.8 cm thickness, smoothed on the inside with a pattern of volutes (swi rling scroll like patterns around a central dot) in bas relief on the exterior. An anthropomorphic face with a headdress dominates the center of the vase with hints of profile faces along the sides. Figure 11 Photographs of a carved shell atlatl finger loop from different angles. The pair of loops were found where the head should have been in the cairn capstone cist burial (CT3 -2). Borders across the top and bottom of the vase are overlapping voussures. Th e obverse and reverse of the vase are separated by two protruding handles, which, unfortunately, were not recovered. These handles are important in stylistically dating the piece. Evidence remains around the blank area of the vase where the handles conne cted to mammalian feet carved into the bas relief pattern. Therefore, the animal portrayed was prone with all four feet resting on the vase itself, unlike earlier style vases with bat heads or animals connected by the back instead of the feet to the vase. The form of the feet appears to suggest a feline form, which correlates with the other late (AD 750 850) style elements (Luke 2010). The other unusual artifact is also a unique find for Pacbitun. It consists of a pair of carved marine shell (cf. West Ind ian chank Turbinella angulata ), crescent shaped lunates ( Figure 11 ). The lunates are sometimes mistaken for ear spools, pendants, or bracelets/adornments; we believe the artifacts were actually part of an ornate atlatl that was buried with the individua l. Ekholm first suggested that U shaped stone or shell carvings such as these were the finger loops of an atlatl after he observed an intact loop on a historically collected atlatl from Mexico housed the British Museum (1962). The pair from Pacbitun we re found close enough together ( Figure 10 ) that they certainly could have been

PAGE 196

Two Unusual Finds from Courtyard 3, Pacbitun, Belize 202 lashed to an atlatl. Unfortunately, no other parts of the atlatl were recovered. The finger loops were found where the skull of the cairn capstone cist burials head should act ually have been located. The finger loops were found in the darker burial soil at the very southern edge of the grave, along with two thin jade beads and a tooth filed to form the T symbol (Romero classification C3; Romero 1986). While there is some debate, Williams and Whites study (2006) suggests that there may be a relationship between filed teeth and elite status. They also found that the C3 modification was exclusive to males at Lamanai. The relative locations of artifacts, bones, and jade beads are shown in Figure 10 and Figure 12. There were a number of other finds with this burial which are similar to artifacts from Tomb 19 such as a fine limestone bead and a carved bone tube. Interestingly, the lower burial (CT33) below the slate slabs co ntained no grave goods at all. The finger loops from archaeological contexts in Mexico examined by Ekholm were housed in the American Natural History Museum and the Museum of the American Indian, Heye Foundation. He examined a total of 3 complete pairs an d 17 single loops (Ekholm 1962). Of all potential Maya Lowland examples, there is only one other shell atlatl finger loop, which was found at Uaxactun (Kidder 1947:66). Intact specimens of whole wood atlatls, including the wood finger loops, were found i n a cenote at Chichn Itz (Coggins and Shane 1984:108), and in situ carved bone finger loops were found in a burnt layer of Tikal palace Structure 5D 51 (Harrison 2003:105). Another archaeologically excavated Maya atlatl comes from a possible Early Classic shell atlatl hook found in a Caracol burial, Special Deposit C117F 1 (Chase and Chase 2011:11). Beyond these examples, evidence of atlatl use by the Maya comes from compari son of different projectile points to try to determine darts from arrows (Aoyama 2005; Ciofalo 2012). Maya atlatl use is also clear from evidence based on iconography, with examples ranging from AD 378 to the Postclassic period at sites like Tikal, Ucanal Naranjo, Caracol, Uaxactun, and Chichn Itz (Chase and Chase 2002:43; Hassig 1992; Schele and Freidel 1990:156157). From Figure 12 Photograph of the thin jade beads also found near the atlatl finger loops. these contexts, it is clear that elite s used atlatls, and that the atlatl might even have been a symbol of power. Conclusions These unusual finds from Courtyard 3 at Pacbitun come from the same burial (CT3 2), but the context is quite disturbed, as evidenced by the Ulua Valley style carved marble vase fragments being found in different layers and up to 2 m away from the grave itself. All the finds from this burial certainly could have come from a Late Classic period interment that was later revisited or intentionally desecrated. There are al so other artifacts, like fragments of fine ground stone bowls and a figurine head, found less than 2 m away from the grave. All of these may have once been part of the same burial. However, pending radiocarbon analysis of the bone, teeth, and charcoal fr agments found in the burial, it is uncertain exactly the age of the burial. The stratigraphy and preliminary ceramic analysis suggest Late Classic Period for the burial fill, but that only tells us that the grave was likely placed through an early Late Cl assic plaster floor (which the caches were cut into), and some of the contents removed at a later time. The Ulua Valley style carved marble vase stylistically dates the burial to the latter half (AD 750850) of the Late Classic period (Coc Phase, AD 550 800). The individual buried was elite, as shown by the elite grave goods, filed teeth, and the placement of the grave in the center of the courtyard. The burial arrangement, with one

PAGE 197

Skaggs and Powis 203 cist grave on top of another is unusual as well, and the cairn of burnt stone covering the top individual might be quite rare, since we can find no references of this burial pattern. Acknowledgements We would like to thank Dr. John Morris, Melissa Badillo, and the rest of the Institute of Archaeology for granting us permissi on to work at Pacbitun. We thank Christophe Helmke for his thoughts on designs and directions towards identification of artifacts, and Norbert Stanchly for identification of the shell. Additional thanks go out to Oscar Mai, Javier Mai, Lute Mai, John Mai, Peter Cherico, and all of the other students and local workers at the site. A special acknowledgement goes out to the Bronx Community College Geospatial Center staff. Financial support was generously provided by the Alphawood Foundation and the CUNY Research Foundation. References Aoyama, K. 2005 Classic Maya warfare and weapons: spear, dart, and arrow points of Aguateca and Copan. Ancient Mesoamerica 16(2):291304. Bill, Cassandra R. 1987 Excavation of Structure 23: A Maya "Palace" at the Site of Pacbitun, Belize. Unpublished Master's thesis, Department of Anthropology, Trent University, Peterborough, Ontario. Campbell -Trithart, Melissa J. 1990 Ancient Maya Settlement at Pacbitun, Belize, Unpublished Master's thesis, Department of Anthropology, Trent University, Peterborough, Ontario. Chase, Diane Z. and Chase, Arlen F. 2002 Classic Maya Warfare and Settlement Archaeology at Caracol, Belize. Estudios de Cultura Maya 22:3351. 201 1 Status and Power: Caracol, Teotihuacan, and the Early Classic Maya World. Research Reports in Belizean Archaeology 8:3 18. Cheong, Kong 2013 Archaeological Investigation of the North Group at Pacbitun, Belize, Unpublished Master's Thesis, Department of Anthropology, Trent University, Peterborough, Ontario. Ciofalo, Andrew J. 2012 Maya Use and Prevalence of the Atlatl: Projectile Point Classification Function Analysis from Chichn Itz, Tikal, and Caracol. Unpublished Master's Thesis, Department of Anthropology, University of Central Florida, Orlando. Coggins, Clemency Chase, and Orrin C. Shane 1984 Cenote of Sacrifice: Maya Treasures from the Sacred Well at Chichn Itz. University of Texas Press, Austin. Davis -Salazar, Karla L., E. Christian Wells and Jos E. Moreno -Corts 2007 Balancing Archaeological Responsibilities and Community Commitments: A Case from Honduras. Journal of Field Archaeology 32(2):196205. Ekholm, Gordon F. 1962 U -Shaped "Ornaments" Identified as Finger Loops from Atlatls American Antiquity 28(2):181185. Harrison, Peter D. 2003 The Atlatl from Operation 96D, Structure 5D 51, Group 5D -11, Tikal. Appendix J. In The Artifacts of Tikal: Utilitarian Artifacts and Unworked Material. Tikal Report No. 27, Part B edited by Hat tula Moholy -Nagy, pp. 105-125. University of Pennsylvania Museum of Archaeology and Anthropology, Philadelphia. Hassig, Ross 1992 War and Society in Ancient Mesoamerica. University of California Press, Berkeley. Healy, Paul F. 1990 Excavations at Pacbitu n, Belize: Preliminary Report on the 1986 and 1987 Investigations. Journal of Field Archaeology 17:247 -262. Healy, Paul F., Jaime J. Awe, Gyles Iannone, and Cassandra R. Bill 1995 Pacbitun (Belize) and Ancient Maya Use of Slate. Antiquity 69:337 348. Healy, Paul F., J aime J. Awe and Hermann Helmuth 2004 Defining Royal Maya Burials: A Case from Pacbitun. In The Ancient Maya of the Belize Valley: Half a Century of Archaeological Research, edited by James F. Garber, pp. 228237. University of Florida Pre ss, Gainesville. Healy, Paul F., Christophe Helmke, Jaime J. Awe, and Kay S. Sunahara 2007 Survey, Settlement, and Population History at the Ancient Maya Site of Pacbitun, Belize. Journal of Field Archaeology 32:1739.

PAGE 198

Two Unusual Finds from Courtyard 3, Pacbitun, Belize 204 Kidder, A. V. 1947 The Artifacts of Uaxactun, Guatemala (No. 576). Carnegie Institution of Washington. Washington DC. Luke, Christna 2010 Ulua Marble Vases Abroad: Contextualizing Social Networks Between the Maya World. In Trade and Exchange: Archaeological Studies from History and Preh istory and Lower Central America edited by Dillian, Carolyn D. and Carolyn L. White, pp. 37-58. Springer, Berlin. Luke, C., and Tykot, R.H. 2007 Celebrating place through luxury craft production: Travesa and Ulua Style Marble Vases. Ancient Mesoamerica 1 8: 315 328. Luke, C., Tykot, R.H., and Scott, R.W. 2006 Petrographic and stable isotope analyses of Late Classic Ula Marble Vases and Potential Sources. Archaeometry 48 (1): 13 29. Parry, William 1987 Chipped Stone Tools in Formative Oaxaca, Mexico: Their Procurement, Production, and Use University of Michigan Museum of Anthropology Memories No. 20. Ann Arbor. Pierce, Karen and Sheldon Skaggs n.d. Preliminary Investigations of Structures 22 and 33 at Pacbituns Triple Courtyard Palace Group. In Pacbitun Regional Archaeological Project (PRAP): Report on the 2017 Field Season, edited by Terry G. Powis. Report Submitted to Institute of Archaeology, National Institute of History and Culture, Belmopan City, Belize. Richie, Clarence J. 1990 Ancient Ma ya Settlement and Environment of the Eastern Zone of Pacbitun, Belize. Unpublished Master's thesis, Department of Anthropology, Trent University, Peterborough. Romero Molina, Javier 1986 Catlogo de la Coleccin de Dientes Mutilados Prehispnicos IV Parte Coleccin Fuentes Instituto Nacional de Antropologa e Historia, Mexico City. Schele, Linda, and David Freidel 1990 A Forest of Kings William Morrow, New York. Scott E. Simmons 1995 Maya Resistance, Maya Resolve: The tools of autonomy from Tipu, Belize. Ancient Mesoamerica 6: 135146. Simmons, S. E. 2002 Late Postclassic-Spanish Colonial Period stone tool technology in the southern Maya lowland area: the view from Lamanai and Tipu, Belize. Lithic Technology 27(1): 4772. Skaggs, Sheldon, Nica ela Cartagena, and Samuel Kontoh 2017 Exploring Courtyard 1: Excavations at Structure 25 and Courtyard 1 of Pacbitun, Belize. In Pacbitun Regional Archaeological Project (PRAP): Report on the 2016 Field Season, edited by Terry G. Powis, pp. 6 19. Report Su bmitted to the Institute of Archaeology National Institute of Culture and History, Belmopan, Belize. Skaggs, Sheldon and Peter Cherico n.d. Courtyard Excavations at Pacbitun, Belize. In Pacbitun Regional Archaeological Project (PRAP): Report on the 2017 F ield Season, edited by Terry G. Powis. Report Submitted to Institute of Archaeology, National Institute of History and Culture, Belmopan City, Belize. Spenard, Jon 2011 Heading to the Hills: A Preliminary Reconnaissance Report on Pacbitun's Regional Lands cape. In Pacbitun Regional Archaeological Project (PRAP): Report on the 2010 Field Season, edited by Terry G. Powis, pp. 3389. Report Submitted to the Institute of Archaeology National Institute of Culture and History, Belmopan, Belize. 2012 Defining Com munity Ch'een: A Report on the 2011 Archaeological Cave and Karst Landscape Investigations around Pacbitun, Cayo District, Belize. In Pacbitun Regional Archaeological Project (PRAP): Report on the 2011 Field Season, edited by Terry G. Powis, pp. 144-188. R eport Submitted to the Institute of Archaeology National Institute of Culture and History, Belmopan, Belize. Sunahara, Kay S. 1995 Ancient Maya Settlement: The Western Zone of Pacbitun, Belize, Unpublished Master's thesis, Department of Anthropology, Trent University, Peterborough. Turner, Jeffrey, Keener Smith, and Blake Bottomley 2015 From Guacamayo to Pacbitun: A Settlement Survey between Two Medium -Sized Regional Centers In Pacbitun Regional Archaeological Project (PRAP): Report on the 2014 Field Season, edited by Terry G. Powis, pp. 6-21. Report Submitted to the Institute of Archaeology, National Institute of Culture and History, Belmopan, Belize. Ward, Drew 2013 Investigations of a Ground Stone Tool Workshop at Pacbitun, Belize, Unpublished Master's Thesis, Department of Anthropology, Georgia State University, Atlanta. Weber, Jennifer U. 2011 Investigating the Ancient Maya Landscape: A Settlement Survey in the Periphery of Pacbitu n,

PAGE 199

Skaggs and Powis 205 Unpublished Master's Thesis, Department of Anthropology, Georgia State University, Atlanta. Weber, Jennifer U., and George Micheletti 2016 Implementing Airborne LiDAR Data for Archaeological Analysis and Prospection at the Ancient Maya Site of Pacbitun Belize. Pacbitun Regional Archaeological Project (PRAP): Report on the 2015 Field Season, edited by Terry G. Powis, pp. 36-44. Report Submitted to the Institute of Archaeology, National Institute of Culture and History, Belmopan, Belize. Weber, Jennifer U., and Terry G. Powis 2014 Assessing Terrestrial Laser Scanning in Complex Environments: An Approach from the Ancient Maya Site of Pacbitun, Belize. Advances in Archaeological Practice 2(2):6074. Wells, E. Christian, Karla L. Davis -Salazar, Jos E. Moreno -Corts, Glenn S. L.Stuart, and Anna C. Novotny 2014 Analysis of the Context and Contents of an Ulua style Marble Vase from the Palmarejo Valley, Honduras. Latin American Antiquity 25(1):82100. Williams, Jocelyn S. and Christine D. White 2006 Dental Modification in the Postclassic Population from Lamanai, Belize. Ancient Mesoamerica 17:139 151.

PAGE 200

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 207 217 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 19 FROM PHOTOGS TO MODE LS: DIGITAL ARCHAEOL OGY OF PRE HISPANIC PACBITUN, B ELIZE Jon Spenard, Michael Mirro George J. Micheletti and Terry G. Powis Since its inception, the Pacbitun Regional Archaeological Project (PRAP) has experimented regularly with an array of digital technologies for more comprehensive documentation and data presentation than has been possible with traditional recording methods such as hand-drawn maps, photographs, and written descriptions. In this paper, we discuss our use of photogrammetry and virtual site tours. Specifically, we focus on the benefits of this technology for archaeology in a variety of commonly encountered contexts including, architectural and landmark mapping, unit and archaeological feature modeling, rock art identification, and artifact analysis. This discussion is centered on three sites at Pacbitun Slate and Crystal Palace caves, and the El Quemado structure and archaeological materials a nd features associated with them. At Slate Cave, we employed photogrammetry for unit and feature mapping, modeling artifacts, and documenting the only known rock art panel in the Pacbitun region. At Crystal Palace, we used this technique for modeling the entrance chamber, mapping small stone wall terrace features and a vessel cache, and for modeling and illustrating a ceramic bowl from that offering. Lastly, in Pacbituns site core, we used it to document the El Quemado structure, a Middle Preclassic per iod ceremonial structure. Introduction Since its inception in 2008, the Pacbitun Regional Archaeology Project has experimented with an array of digital technologies for more comprehensive documentation and better data presentation than has been possible with traditional record ing methods, such as handdrawn maps, still photographs, and written descriptions. As well, over the last several years with the use of aerial LiDAR and terrestrial laser scanning, and more recently with great technological leaps in digital photography, archaeological methods have been undergoing a digital revolu tion. This paper details several of the ways we use photogrammetry at the preHispanic Mays site, Pacbitun, Cayo, Belize. Specifically, we discuss our use of this technology, and the benefits and limitations we experienced employing it in a variety of co mmonly encountered archaeological contexts from architectural and comprehensive landmark mapping, to artifact analysis, rock art identification, and unit and archaeological feature modeling. Our discussion is centered on two caves around Pacbitun Crystal Palace, and Slate Cave- and the recently uncovered Late Middle Preclassic structure, El Quemado, in Pacbituns site core, discussed in further detail by Crow et al. in this volume. Throughout this discussion, we attempt to demonstrate how using these tech nologies helps convey a phenomenological sense of the places we investigate, while also making them broadly accessible for interested audiences. In that regard, we favor including Internet links to the models discussed in this paper in lieu of static, cap tioned figures, although we do provide some of the latter. Thus, this paper is best read on an internet connected device where the reader can navigate to the models and interact with them as they proceed through the text. By using photogrammetry as a cart ographic tool, we achieve a higher level of detail, accuracy, and precision, increased measurement resolution and content in our cave maps while reducing field time spent on recording and measuring features, all resulting in more archaeological data collec ted in a single season. Our ultimate goals for this paper are critically analyzing and advancing newly accessible digital methods for studying and documenting archeological sites. Introduction to Photogrammetry Photogrammetry is the art and science of obtaining precise geometric measurements of an object, including position, shape, and surface attributes, without making physical contact with them. Traditionally, photogrammetric processes results in three dimensional, mathematically computed positions usi ng two or more overlapping photographs of a subject taken with cameras located in known unique locations. Early processes required location of cameras and control points within the scene for the aerial triangulation algorithms used to compute model positi ons. Structure from motion (SfM) is a recent advance in photogrammetry where no

PAGE 201

Digital Archaeology of Pre -Hispanic Pacbitun 208 reference information is required, resulting in a relatively unscaled and non oriented model. Using SfM methods and software, an object is photographed with an overlapping se ries of digital images from which the software determines the geometry of 3D positions and camera locations (Westoby et al. 2012). A 30 percent overlap in photos is ideal. Models may be subsequently referenced using 3D similarity transformation based on a small number of control points and then exported for further post processing in GIS, 3D, and graphic software programs. Photogrammetric and SfM techniques used at Pacbitun can be categorized as CloseRange Photogrammetry, which is defined by an object to camera distance less than 300 meters (Mathews 2008). In contrast, Aerial Photogrammetry is typically aviation based and requires groundcontrol survey and results in highl y accurate surface terrain models and orthoimagery. Close Range Photogrammetry in combination with SfM provides a high degree of flexibility in choice of cameras, camera mountings, photographic techniques, groundcontrol (if required), and software, making these methods ideally suited to archaeological fieldwork. Taking advantage of high resolution digital single lens reflexive cameras (dSLR) capable of producing sharp, high resolution photographs, producing models with minimal ground sample distances (GSD ), or pixel resolution of the object surface is possible. For example, models generated using a 10 megapixel dSLR with a 20 mm lens can produce a GSD less than half a millimeter with a shooting height of 1.4 meters (Mathews 2008). There is a direct corre lation between the increased resolution of photographs and GSD of resultant models. With the affordability of dSLR cameras capable of capturing images in excess of 30 or 40 megapixels, producing extremely detailed models with GSD below 0.1 millimeters or even 0.01 millimeters is possible depending on the camera height. While such resolution may exceed requirements of mapping features exceeding one meter, it is of great benefit to artifact documentation, where the resolution allows for modeling of individual temper grains in ceramics and flaking patterns in lithic artifacts allowing for hands off analysis. Close Range Photogrammetry results in a variety of products, ranging from simple 3D models to printed solid models. During the 2016 PRAP field season, w e created several photogrammetric work flows to capture 3D data, each of which were determined by specific mapping goals we created for each of the subjects. When illustrating a feature for presentation purposes, we generated unreferenced moderately highresolution 3D models of the subject. For the most part, we uploaded these models to the senior authors SketchFab website, https://sketchfab.com/jonspenard, exported them as PDFs, added them as interactive figures in reports, and used them in presentation s for conveying a sense of place to the audience (Spenard et al. 2017). With mapping grade models, we produced a low resolution spatially referenced 3D model, which were converted to orthoimagery and elevation models, and used to generate plan and profile maps of individual features. Archival and analytical 3D models are generated in highresolution and could be used to document artifacts removed during excavations, unmovable or threatened features, and/or objects with fine or complex details. Such model s allow analysts to take measurements on the digital object, out of the field, potentially alleviating the need for export permits. The high resolution of this last class of models also permits accurate 3D printing that can provide tangible representation s of artifacts for use in the classroom, in museums, or Archaeology Day celebrations, and other public venues. A second photogrammetric visualization tool, photosphere, was used to create virtual tours of several caves surrounding Pacbitun. Virtual tours, created using linked and spherically projected panoramic photographs, can provide a practical way to convey, scale, feeling, morphology of the environment, and a sense of place, both large and small. Much like Google Streetview, a virtual tour consists o f a series of interactive photospheres allowing a user to look around the cave and walk from one sphere to the next with the click of a mouse. Within each sphere, a user can select hotspots

PAGE 202

Spenard et al. 209 containing informational text, photographs, close range 3D models, maps, video, sound, and other media providing details on objects within view. We created photospheres using a tripod mounted dSLR camera with a fisheye lens. Photographs were taken every 60 degrees at three inclinations ( 45, 0, and +45 degrees) as well as nadir. Using panorama software, each set of photographs representing a sphere was stitched together into rectangular images with a spherical projection that appears highly distorted. When loaded into a viewer, the images lose their distortion and provide an accurate view of the subject cave chamber. Photographs were taken with a relatively high resolution camera resulting in the production of a very high resolution panorama allowing users to zoom in on details in the scene. In addition to viewing virtual tours on a computer monitor or television screen, tours can be viewed using the stereoscopic headsets such as Google Cardboard viewer or Oculus, creating a more immersive virtual environment. Ambient sounds recorded from th e location of the camera can also be added to these tours, increasing the phenomenological sense of place. We reiterate here, such models are comprehensive photorealistic representations of everything in the environment photographed, including potentially highly sensitive archaeological features such as burials, caches, whole ceramic vessels, rock art, etc. Therefore, great care and attention to preservation and protection must be taken when creating and, and especially distributing such models, as they do show exactly where these archaeological materials are located and how to navigate to them. Placing password protections on online models can help alleviate some of that concern, yet hosting websites often retain irrevocable perpetual rights to any models uploaded. For example, Sketchfab, the website we use for hosting models, states in its Terms of Use, Section 4.2, By using the Services, you grant Sketchfab a worldwide, nonexclusive, royalty free, perpetual, irrevocable sub licensable (through multiple tiers) right and license to use and adapt the User Content for the purposes of developing, distributing, providing, improving, and promoting the Services (emphasis added) (Sketchfab 2017). Thus, while the archaeologist may use passwords to protect such models from being openly accessible, we cannot guarantee the hosting company will follow suit. One of the benefits of photogrammetry is that it is relatively cheap, in fact, most of us now carry around with us the basic tools necessary for doing it. For starters, all that is required is a decent digital camera, preferably a dSLR, but it can be successfully executed with a smartphone. For the photogrammetric models in the caves, we used a Nexus 6p smartphone with 12.3 mega pixel camera, an iPhone 7 also with a 12 mega pixel camera and, a tripod mounted Nikon D3000 dSLR. All photogrammetry lighting was achieved with diffuse sunlight or LED video lights placed out of site yet positioned to illuminate chambers in their entirety. This latter technique requi red a significant time investment, sometimes several hours, for determining the best position for ensuring comprehensive coverage, and invisibility of the light panels. During the 2017 field season, we experimented with placing two, slightly side facing L ED light panels on a dual bracket mounted to the top of the camera as suggested by Dominic Rissolo of the Cultural Heritage Engineering Initiative of University of California, San Diego (Dominic Rissolo personal communication). Employing this lighting tec hnique proved to be a great success and recuperated several hours of the work day for photography and other archaeological endeavours. Since the lights were placed above the camera, shadows were no longer an issue, and the scene was always perfectly light ed for each image captured with only minor adjustments for brightness needed. For the models of the El Quemado structure, we used a Nikon D800 DSLR and diffuse sunlight. For software, we used AgiSoft Photoscan to make photogrammetric models, which runs $550 for an education license, and Pano2VR for the virtual tours, costing $350, a total budget of $900 dollars (US) We believe one of the common frustrations of archaeology is offering compelling and satisfying descriptions. Due to technological limitations, cave descriptions in the archaeological literature and presentations, are commonly restricted to a few lines noting total length and/or depth, number of chambers,

PAGE 203

Digital Archaeology of Pre -Hispanic Pacbitun 210 Figure 1 Plan view map of Crystal Palace cave. Figure 2 Screen capture of digital model of five-vessel ceramic cache on Crystal Palace Ledge 1. height of ceiling for each, presence of archaeological features, etc. of the landmarks studied. These narrative descriptions are bolstered with a few representative photographs and a plan view map. Yet, in our experience, we have found that such descriptions lack the ability to truly capture and convey the character of the underground places we inve stigate. But, with the digital recording technologies such as those we discuss here, we are able to provide a visual aid that bring s the caves alive. Crystal Palace Cave Traditionally we would describe Crystal Palace cave as a collapsed, ovular shaped si nkhole with six chambers, each separated by

PAGE 204

Spenard et al. 211 large collapse boulders or walls of columnar formations ( Figure 1 ). The ceiling ranges from about 1.5 m tall at the entrance, to nearly 8 m tall at its highest. The entrance slopes sharply downward, but the May a constructed several terraces and low walls, forming a path through the entrance. Beyond the entrance, the cave floor is covered in mud, and several modern stacks of Late Classic period sherds can be found along elevated ledges. Though this may be an ad equate overall description, it fails to capture the essence of the cave. We are currently working on completing the virtual tour of this cave, but we have also spent a great deal of time modeling these features using photogrammetry, which we will be incorporating into the virtual tour once the models are processed. One of those we have processed from the 2016 field season is a five piece ceramic bowl cache on a newly discovered ledge ( Figure 2 ). Not only do we see the vessels in the cache, but by exploring the model online ( https://skfb.ly/6vyYK ), the reader can move it around to observe the topography of the ledge, study the positioning of the vessels in relation to one another, zoom in or out onto specific de tails, etc. We also note here the chamber ceiling is about 0.5 m above the bowls, and is heavily populated with cave formations, but we were able to mask those out, allowing for more unfettered viewing of the entire feature than is possible in reality. S imply put, none of these actions are possible with a simple photograph, a series of them, or plan view map of the ledge. In addition to spatial data, this technique can have sub millimeter precision for reasons discussed above. The model linked above was processed at a low resolution, thus, small details are a bit obscured; nevertheless, if the reader zooms in on any one of the bowls, they can see that the contents of each are topographically modeled. But we can gather even more data from these models. B y exporting a model of one of the bowls into a program such as ArcGIS, we were able to produce true to form profile of the vessel with sub millimeter measurements ( https://skfb.ly/6vz7t ) (Figure 3). It is through this same technique, but at a different scale, that we can produce plan and profile maps of caves and other karst features, and unit profiles, as discussed below. Returning to the model of the ledge, the reader can see a few of the issues with this technique ( https://skfb.ly/6vyYK ; see also Figure 2). Notice that all the bowls have large holes in them. These could have been caused by a number of factors, either there was not enough overlapping coverage with the photos, or as is more likely the case, the computer program was unable to find any common points in those areas between the series of photographs, and thus simply could not reconstruct them. This point leads us to another downside of photogrammetry, which is that c omplex models require a significant amount of computer processing power, in some cases more so than is commonly available in field laptops. This has often meant that all but the simplest models must be processed after returning back home from the field. Thus, any issues with the photography can result in data loss if modeling an excavation or in the case of a surface feature, would need to wait until the next field season to reshoot. Slate Cave In Slate Cave, we experimented with photogrammetry in a vari ety of contexts, demonstrating its utility on multiple scales. This cave was chosen for modeling because it has the only known rock art in the Pacbitun region, a series of simple faces and geometric shapes carved into an active flowstone formation protrud ing from the ceiling near the caves entrance ( Figure 4 see also https://skfb.ly/6vyYG ). Most 3D viewers allow the user to rake light across the model, and in doing so, we discovered a here to for previously unrecognized figure carved into the formation, likely a spider monkey. As the opening of this cave received ample daylight, and we were excavating in it, we also modeled the floor and entrance area. With these data, we created an orthophoto, digital elevation model (DEM), and hillshade, all of which were used as base layers for a new trueto form map of the Entrance Area. The photography took approximately one hour, while processing the different maps and drawing the final map took a total of about three hours. A similar hand drawn field map with equal

PAGE 205

Digital Archaeology of Pre -Hispanic Pacbitun 212 Figure 3 Vessel profile produced from 3D digital photogrammetric model of one of the Crystal Palace Ledge 1 cache vessels. Figure 4 Screen capture of digital model of Slate Cave petroglyph panel. Tags 1-8 show the monkeys head, left fist, left elbow, tip of tail, right elbow, crotch, left foot, and right foot respectively.

PAGE 206

Spenard et al. 213 precision and detail of this area would hav e taken at least a full day in the field to produce. With the ability to draw the map in the lab after the field season, we were able to devote more time to our excavations. As well, a sidebyside comparison of our original handdrawn map and our new map of this area shows the level of accuracy and detail using photogrammetry is unmatched ( Figure 5 ). Photogrammetry excelled in mapping unit excavations in this cave. We excav ated a pit feature in a low alcove against the cave wall containing a human cranium surrounded by a cluster of bone tubes. The pit was covered by slate and limestone slabs with possible human remains found beneath ( ht tps://skfb.ly/6vz79 ) Each layer and mapped object was documented using photogrammetry in lieu of creating handdrawn maps. The models were then imported into ArcGIS and these features were digitized into highly precise plan and profile drawings of the fe ature. We also used the technique to make a fully navigable model of the entrance area. Such models can be uploaded to software and apps for use with virtual reality goggles, which allow the user to become fully immersed in the scene. Nevertheless, the end product was less than desirable. Many spots in the chamber walls were unable to be reconstructed by the software, leaving large, vacant holes. Moreover, due to changing light conditions as we moved from direct to indirect sunlight, the resulting photographs of the same areas varied in color, leaving the model appearing blurry. El Quemado Structure Exposed in 2013 beneath the main plaza at Pacbitun, the large ceremonial platform, El Quemado, or Q, has been the primary focus of the site core investigati ons up to the 2017 field season. Radiocarbon dates indicate the platform was constructed in the Middle Preclassic period around 600 BC and was eventually terminated around the onset of the Late Preclassic period (ca. 400 BC) (Powis et al. 2017). Sealed below marl and dirt filled task units and capped by several plaza floors, El Quemado had been entombed for over two and a half millennia. Excavations to this point have revealed an architectural layout unlike any other documented in the Belize Vall ey region (Micheletti et al. 2017; Micheletti et al. 2016; Micheletti and Powis 2015). The structures pristine state of preservation is likely owed to the severely burned plaster surface derived from either a single termination event or longterm ritual use. Because of Qs rare architectural form, nearly flawless condition, and buried state, we decided to digitally curate the structure as first exposed. A three dimensional model of Q will allow us to further investigate its construction methods, structu ral attributes and features, as well as the unusual method of deposition. A model will also aid those researching early monumental architecture of the ancient Maya and can serve as a visual aid in educational and public settings. Our first efforts at the digital preservation of El Quemado was in 2013 through the use of terrestrial laser scanning (TLS) (Weber and Powis 2014). While this was an effective method, it proved to be too costly, and thus not a practicable means of annual documentation. As a low cost alternative, yet just as effective method, we decided to use photogrammetry. In 2015, after excavations had uncovered the south face of El Quemado, project members Jeff Powis and Andrew Vaughan photographed the subplaza structure and produced the fi rst 3D photogrammetric model using Agisoft Photoscan software (Vaughan et al. 2016). As the project resumed excavations in 2016, we planned to continue producing photogrammetric models of the newly exposed areas ( Figure 6). Our ultimate goal was to add t hese newly exposed areas to the model of Q produced in 2015, which could be done by either processing photos in the Photoscan software or by manually connecting each of the models using other 3D processing software. After the 2016 excavations located and e xposed Qs east and west sides and southern plaza floor, each of these areas were photographed and modeled. To ensure that each model would join together, back dirt was removed from previously excavated areas to reexpose modeled architecture adjacent to the newly exposed 2016 units. Doing so uncovered recognizable features and attributes used to align the previous photos with the new ones. Another strategy we experimented with was creating a photogrammetric modelled path that would

PAGE 207

Digital Archaeology of Pre -Hispanic Pacbitun 214 Figure 5 Side -by -side comparison of (a) portion of hand-rendered plan view map of Slate Caves entrance area, and (b) plan view map produced from digital model. Note that the maps are oriented in different directions.

PAGE 208

Spenard et al. 215 Figure 6 Photogrammetric models of El Quemados architecture exposed in the 2016 field season (a) east side, (b) west side. Figure 7 Composite digital model of El Quemado from 2015 and 2016 field seasons.

PAGE 209

Digital Archaeology of Pre -Hispanic Pacbitun 216 serve to spatially link each of the 2016 units. This helped to preserve the spatial (orientation and scale) integrity of the separately exposed areas of architecture to aid with the processing. It would also aid in the manual connection should the alignment not work in Photoscan. In this case, any manual processing would be done using the 3D processing software, 3DReshaper. The primary purpose of the 3DReshaper software was to integrate the photogrammetric models with the TLS data acquired in 2013. After processing each years photosets i n Photoscan, we were able to successfully align the east edge of Q with the south face model. This was likely due to the re exposed cut limestone blocks of the south facing wall, photographed in both 2015 and 2016. The only visible issues were minor chan ges in lighting and soil color. The poor preservation of the summit had also resulted in gaps in the final model. On the other hand, the west edge had completely failed to align due to an insufficient number of photographs in overlapping area. Simply stated, there were not enough recognizable, overlapping points in this area in either photoset for Photoscan to properly merge the two models. Also, while the modelled path linking each exposed area was able to create an accurate spatial layout of the exposed architecture, Photoscan was unable to properly orient the west edge due to the lack of overlapping points. Although there were only minor issues with the eastern edge and the south face of Q, we experimented with manually attaching each of the models us ing 3DReshaper. After re orienting, scaling, and cropping and smoothing edges, each of the models fit together and were able to be merged into a single compound mesh. Nonetheless, minor issues with light and soil coloring, and small gaps in unpreserved a reas remained. Although the results were not as detailed and defined as with the model produced with Photoscan, the manually attached model created in 3DReshaper was more expedient and could be physically manipulated ( Figure 7 ). All in all, with our photogrammetric work of Q, we learned that models from multiple years were able to be reconnected with sufficient overlap between excavated units; however, a more effective approach would be the creation of permanent, completely immobile datum markers the softw are can use as recognizable points. Moreover, recreating similar lighting conditions over the years, (photos taken during the same time of day, in the shade, beneath tarp, etc.) would allow for the creation of cleaner models. Conclusion To sum, we have f ound that photogrammetry excels with small scale commonly encountered archaeological contexts such as surface caches, rock art, and unit excavations. Producing mapping grade models saves valuable field time, increases map accuracy, and produces moderate quality models that are easily sharable electronically. This ability to share electronically, and online, as we hope to show with virtual tours and photogrammetric models, allows us to bring our work to the public, and make it as accessible to as wide an audience as possible. Additionally, while photogrammetry worked well for modeling settlement architecture, it only fared moderately well mapping cave chambers. Even smaller rooms and chambers required many photographs to ensure proper overlapping, and lighting had also proved to be a challenge initially. Nevertheless, maps with higher detail and precision than are able to be drawn by hand could be cre ated from the models, saving significant field time for other endeavors. Acknowledgements We are grateful to the Institute of Archaeology in Belize, specifically Dr. John Morris and Melissa Badillo, for their support of the Pacbitun Regional Archaeologic al Project (PRAP). A special acknowledgement goes to the Alphawood Foundation, and California State University San Marcos for financially supporting PRAP, and to Petzl for providing caving equipment. We would also like to thank Mr. Alfonso Tzul, and Mr. Joe Mai, the landowners, who graciously allowed us to excavate on their private land, and would like to express our appreciation to all the people of San Antonio village for their constant support throughout the field seasons. This paper would not be poss ible without the dedicated Pacbitun staff who commit their time and effort to the project in and out of the field including; Nikki

PAGE 210

Spenard et al. 217 Cartegena, Kaitlin Crow, C.L. Kieffer Nail, Adam King, Mike Lawrence, Karen Pierce, Jeff Powis, Sheldon Skaggs, and Andrew Va ughan. Thanks also to Amy and Eleanor Spenard. References Matthews, N.A. 2008 Aerial and Close-Range Photogrammetric Technology: Providing Resource Documentation, Interpretation, and Preservation. Technical Note 428. U.S. Department of the Interior, Bureau of Land Management, National Operations Center, Denver, Colorado. Micheletti, George, Kaitlin Crow, and Terry G. Powis 2017 Q and A: Exposing El Quemados Architectural Configuration in Plaza A at Pacbitun, Belize. Research Reports in Belizean Archaeology 14:2330. Micheletti, George, Terry G. Powis, Sheldon Skaggs, and Norbert Stanchly 2016 Early Maya Monumental Architecture in the Belize River Valley: Recent Archaeological Investigations of El Quemado at Pacbitun. Research Reports in Belizean Arc haeology 13:4350. Micheletti, George, and Terry G. Powis 2015 Origins of the Block Party: Investigations of Preclassic Architecture Over and Under Plaza A at Pacbitun, Belize. Research Reports in Belizean Archaeology 12:205-216. Powis, Terry G., Jon Spe nard, Sheldon Skaggs, George Micheletti, and Christophe Helmke 2017 An Ancient Maya City Living on the Edge: The Culture History of Pacbitun, Belize. Research Reports in Belizean Archaeology 14:191-212. SketchFab 2017 Terms of Use. Electronic document, https://sketchfab.com/terms accessed December 29, 2017. Spenard, Jon, Michael J. Mirro, Jennifer U. Weber, and Terry G. Powis 2017 Digital Documentation of Ancient Ritual Landmarks: Modeling Senses of Place with Photogrammetry, LiDAR, and Virtual Tours. Paper presented at the Emerging from the Place of Darkness: Subterranean Archaeology in Mesoamerica symposium, organized by James Brady, 82nd Annual Meeting of the Society for American Archaeology, Vancouver. Va ughan, Andrew, Jeff Powis, and Terry G. Powis 2016 From The Application of Terrestrial Structure from Motion at Pacbitun: Investigation of a Possible Low Cost Complement to Terrestrial Laser Scanning for the Capture of 3D Archaeological Data. In Pacbitun R egional Archaeological Project (PRAP): Report on the 2015 Field Season, edited by Terry G. Powis, pp. 6 -24. Institute of Archaeology, National Institute of History and Culture, Belmopan City, Belize. Weber, Jennifer, and Terry G. Powis 2014 Assessing Terr estrial Laser Scanning in Complex Environments: An Approach from the Ancient Maya Site of Pacbitun, Belize. Advances in Archaeological Practice 2(2):123 -137. Westoby, M.J., J. Brasington, N.F. Glasser, M.J. Hambrey, and J.M. Reynolds 2012 Structure -from -Motion photogrammetry: A low -cost, effective tool for geoscience applications. Geomorphology 179 (2012 300 314.

PAGE 211

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 219 230 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 20 SETTING THE STAGE IN CENTRAL BELIZE: 30,0 00 YEARS OF TROPICAL CLIMATE, LA NDSCAPE TRANSFORMATI ON, AND HUMAN INTERACTION Lisa J. Lucero, Jean T. Larmon, and Aime E. Carbaugh The Maya engaged in a sustainable manner with a tropical environment for millennia beginning c. 12,000 years ago. This interaction endured even when the population was at its peak c. 500800 CE. Disruptions did occur, especially between c. 800900 CE whe n several prolonged droughts ultimately led to an urban diaspora. The Maya, however, adapted and continue to do so. Preliminary evidence from diverse datasets collected by the Valley of Peace Archaeology (VOPA) project in central Belize provides a window into at least 30,000 years of climate and landscape histories and human immigration. In this paper, we present varied datasets from different sites in central Belize from individual farmsteads (between Yalbac and Cara Blanca) to a civic-ceremonial center (Yalbac) to a pilgrimage destination (Cara Blanca) to show that the Maya had an enduring relationship with the forested landscape, indicated by evidence for biodiversity and a healthy forest. This project provides a unique regional perspective of a tropi cal environment with the goal of understanding the natural -human dialectic and explores adaptive responses and strategies. Introduction Human natural ecosystem relationships are always changing, reflecting human adaptation, as well as naturally oc c ur r ing regional and global climate change. However, we now live in the Anthropocene, the first epoch of our own making, with potentially detrimental consequences, especially if we do not adapt a more sustainable manner. We can learn from our forebears, especially those who have shown millennia of sustainable adaptation, like the ancient Maya. For the Maya, we posit that their immigration into the area now known as the Maya area did not bring about destruction, but co existence. The lowland Maya adapted beg inning c. 12,000 years ago (Lohse et al. 2006; Prufer et al. 2016; Rosenswig et al. 2014) in a diverse tropical environment with high but dispersed biodiversity. A sustainable way of life endured even when population was at its peak between c. 500 and 800 CE in the Late Classic period. In fact, population density at present in the southern lowlands remains about one to two orders of magnitude less than the density of the Late Classic Period (Turner and Sabloff 2012: 13912). This astounding fact not only demonstrates a long term sustainable interaction, but also potential solutions to current day issues. The long term goals of the Valley of Peace Archaeology (VOPA) project in central Belize are to provide a regional perspective on the natural human dialec tic by exploring adaptive responses and strategies from the perspective of archaeology, paleobotany, dendroecology, paleolimnology, paleontology, isotopic paleoecology, and ancient human genetics. Ideally, results will provide insights and solutions for t he present and future in tropical areas where over 40% of the current population resides and where, due to global climate change, the tropical belt is likely to expand. Here, we present preliminary findings on a subset of these datasets, namely Pleistocen e megafauna fossils, tree specimens, archaeological materials and human remains, and what they indicate about the world in which the Maya lived and adapted. Once the Maya adopted the use of domesticated plants by c. 2000 BCE, they began to settle in farmst eads and small communities (Rosenswig et al. 2015) Over time, population size increased to the extent that eventually kingship emerged to cope with increasingly complex needs by c. 100 BCE. The Late Classic period witnessed the rise of powerful kings and the highest population size; the Maya lived in one of the 100s of centers as well as in non center areas or hinterlands (Lucero 2006). Each of the 100s of centers had its own king, though some were more powerful than others, namely those at Tikal and Calakmul, largely due to their location in areas with large plots of fertile soils. These areas, however, lacked permanent surface water because much of the seasonal rain percolated through the porous limestone bedrock. To sustain year round access to wa ter,

PAGE 212

Tropical Climate, Landscape Transformation, and Human Interaction 220 early leaders built what eventually became massive reservoirs (Scarborough 1993; Scarborough et al. 2012). Urban planning and layout increasingly became interlinked with reservoir systems; the more rulers relied on progressively more complex reservoi r systems, the more vulnerable they became to any disruptions or change (i.e., path dependent). Without extensive irrigation systems, t he Maya relied on seasonal rainfall to grow cultigens, to replenish reservoirs, manufacture plaster and ceramics and for other daily needs, especially potable drinking water. In the intensive farming period during the rainy season, farmers lived in hinterland farmsteads and communities and worked their fields, relying on a mix of small scale, localized subsistence features such as terraces, dams, channels, raised fields and others to grow diverse crops in a dispersed pattern, mirroring the mosaic distribution of fertile soils and other resources (Ford and Clarke 2016). In the dry season, many farmers came to centers to use royal reservoirs and participate in markets and large public ceremonies (Lucero 2006). In exchange for access to center amenities, farmers contributed tribute in the form of labor, goods, and services. Eventually, the reservoir systems resulted in an anth ropogenic landscape of centers and interlinked farmsteads interspersed with forests Kings served as water managers, affording them the means to acquire tribute to fund the political economy because not only did they provide dry season water, but clean water through designing and maintaining constructed wetland biospheres (Lucero et al. 2011). Their role was tested in the face of changing climate, specifically, several prolonged droughts between c. 800 and 900 CE (Douglas et al. 2015; Kennett et al. 2012; Medina Elizalde et al. 2010) that exacerbated existing problems at centers, which, depending on the polity, consisted of the overuse of resources, deforestation and erosion, population growth and expansion, and disrupted trade networks. Even centers with out problems struggled as water sources dried up. Ultimately, most farmers in the southern lowlands abandoned kings, centers, and eventually hinterlands (Lucero et al. 2015). This urban diaspora, where population declined c. 90% in the interior southern lowlands (Turner and Sabloff 2012), had a lasting impact as Maya emigrated in all four directions in search of new land, water, and opportunities. In the end, the different histories of kings and farmers relate to the different constructs in which they e xisted: inflexible vs flexible strategies; a reliance on massive vs small scale diverse water systems; and entrenched and rigid vs resilient and adaptable systems (Lucero et al. 2015:1151). They moved to coastal areas and along major rivers where market towns and trade thrived, especially in the northern lowlands (e.g., Graham 2011; Masson and Freidel 2012; Sabloff 2007) Those that did not abandon the interior lived in smaller communities The Maya never re occupied southern lowlands centers the northe rn lowlands are another story for another time. Maya World Management Throughout their history, the Maya interwove their lifeways in accordance with their cosmocentric worldview, which differs from an anthropocentric one because it situates objects, humans, animals, land, water, everything in an analogous manner each plays a role in maintaining their place in the world and the world itself (Lucero 2017, in press). This melded worldview meant that they made use of and maintained the world, as well as relationships with other world parts. This is not to say that this relationship was perfect t he overuse of resources and deforestation happened and readjustments were made. While deforestation was localized, its extent is not agreed upon (Fedick 2010). Whatever the case, the millions of Maya today speak to their sustainable worldview and way of life, even in the face of the last 500 years of colonial and post colonial histories. Another means of management was through their treatment of sacred places, namely caves and pools, many of which were isolated from settlement. Their isolation promoted conservation because flora and fauna flourished, unencumbered by human habitation. This strategy, in addition to diversifying what they planted in their fields and forest management, together provided them the means to live for millennia without destroying their environment.

PAGE 213

Lucero, Larmon, and Carbaugh 221 Figure 1 Numbered Cara Blanca pools on a pre 2010 hurricane Google Earth image; much of the area south of the pools has since been cleared. Top: Buckley and divers in Pool 1 collecting tree specimens. Bottom, left to right; Pool 1 893020 cal yr BP year old wood specimen; Pool 1 26,975120 cal yr BP sloth tooth; foreman E. Vasquez holds a giant sloth humerus fossil from Pool 1. Tony Rath/VOPA. This cosmology of conservation promoted a sustainable human environment relationship a nd perhaps still can. This system a mosaic of built, managed and untouched areas noticeably differs from current practices of clear cutting and monocropping throughout the tropics (Ford and Nigh 2015). Land clearing not only provides conditions for stag nant water to collect, but also can result in the spread of pests, as well as land atmosphere interactions including increasing temperatures and decreasing precipitation ( DAlmeida et al. 2007) Further, increasing temperatures decrease photosynthesis, wh ich slows CO2 absorption ( Meineke et al. 2016) Human encroachment in the natural world is undeniable, but how we proceed in the future may mean the difference between hard choices and longterm survival vs. short term solutions and disaster. Maya World Management in the VOPA Area To assess human natural ecosystem relationships over the long term, we first need to set the stage and determine what this tropical world was like before and during Maya occupation. We attempt to do so by focusing in several areas in the 115 km2 VOPA research area: 1) Cara Blanca, a sacred landscape and pilgrimage destination ( Figure 1 ); 2) Yalbac, a medium sized center ( Figure 2 and 3) dispersed

PAGE 214

Tropical Climate, Landscape Transformation, and Human Interaction 222 Figure 2 A 3D rendering of Yalbac. Top left: cleared fi elds south of Pools 7-9 and northeast of Yalbac; white smears against the dark soils are ancient mounds and the clusters of green vegetation are unplowed mounds. VOPA. hinterland farmsteads between Yalbac and Cara Blanca (see Figure 2). C ara Blanca is comprised of 25 pools along an east west fault, the north of which consists of a steep escarpment c. 100 m high; most pools are cenotes five to 60+ m deep, while others are lakes c. 2 18 m deep (Lucero and Kinkella 2015). We find noticeable settlement near the western pools or lakes (e.g., nos. 79), but much less so near the central pools or cenotes (e.g., 1 5), which is interesting given that water levels of cenotes do not drop much, even by the end of the dry season. We posit that the absence of noticeable human interference promoted biodiversity where the Maya neither farmed nor hunted. Even though Cara Blanca has plentiful water and fertile soils, the Maya did not change much in the area at least until between c. 800 and 900 CE when the droughts struck and the Maya constructed ceremonial buildings at cenotes as a pilgrimage destination (Lucero et al. 2016) Maya consider such openings in the earth as portals to the underworld and a place through which they supplicated gods, such as Chahk the rain god, as well as ancestors for rain,

PAGE 215

Lucero, Larmon, and Carbaugh 223 Figure 3 Mound 4 near Pool 7 and Burial 1 Individual A Looking north. VOPA. critical in this rainfalldependent society. The Maya left offerings at thousands of portals, many of which were pilgrimage destinations most commonly jars, as well as exotic items, and the ultimate offering, people These s hort t erm responses were to no avail ; visitors to Cara Blanca became part of the diaspora out of the southern lowlands along with most everyone else. Pool 1, 100 x 70 m in size, is a 60 m deep cenote isolated from settlement and farming. We have presented results elsewhere on our excavations at the water temple and ceremonial platform indicating people visited from throughout the Maya lowlands (e.g., Lucero and Kinkella 2015; Lucero et al. 2016). Increasing evidence suggests that Pool 1 may have been one of several visited, likely as part of a ceremonial circuit comprised of sacred features with little or no residential settlement. But the cenotes have a history long before humans came onto the scene. In 2014, divers recovered several Eremotherium laurillardi fossils from Pools 1 and 20, an extinct giant sloth species that can reach 6 m or 20 feet in length (McDonald 2015). Human immigration c. 12,000 years ago may have contributed to the extinction of regional megafauna and the establishment of the modern biota (Roberts et al. 2017) These fossils are the first of this species found in Belize, though they are found from Brazil up into Florida. Larmon and colleagues recently conducted carbon and oxygen isotope analyses of the inner orthodentine of a giant sloth tooth recovered from a fossil bed c. 21 m below the surface of

PAGE 216

Tropical Climate, Landscape Transformation, and Human Interaction 224 Pool 1 by paleontologist H. Gregory McDonald of the Utah Bureau of Land Management that dates to 26,975360 cal. BP (Larmon et al. 2016, n.d.) I ntra tooth isotopic analysis of a 9cm long fragment of the E. lauril lardi molar tooth illustrates the potential of these cenotes as archives of paleoenvironmental information. Through the study of carbon and oxygen isotopes, and using methods that Larmon and colleagues refined to produce the most accurate data possible fr om nonenamel bioapatite, they were able to reconstruct how the diet of the extinct giant ground sloth shifted from the wet season to the dry season c. 27,000 years ago. This isotopic record of a year in the life of a giant sloth shows a tropical savanna with a long dry season that stands in stark contrast to the modern dense tropical forest with a seven month wet season. Despite the increasing aridity, Cara Blanca pools likely have been providing fresh water for tens of thousands of years In fact, what are now the deepest cenotes would have likely been the only water sources in the vicinity during this arid period; giant sloths and other megafauna climbed down into the sinkhole, drank water, could not climb out, became stuck, and became part of the ceno tes geological history; this scenario likely occurred at Pool 20 as well, a 40 m deep cenote also with a megafauna fossil bed Divers also noted masses of trees in the pools. Samples collected in 2014 by dendroecologist Brendan Buckley of the Lamont Doherty Earth Observatory of Columbia University, have all been identified so far as broadleaf species, mostly from the Meliaceae family (e.g., mahogany); none, however, have been dated (Buckley 2015). Divers also collected a wood specimen, possibl y a broadleaf species, from a fossil bed c. 25 m below the surface of Pool 1 that dates to almost 9000 years ago (8930 20 BP) (Lucero 2011) If it is a broadleaf species, it would suggest a wetter period than does the sloth tooth nearly 18,000 years prior The second type of settlement is the center of Yalbac. The Maya built this medium sized center on a plot of fertile land along Yalbac Creek that includes at least one small reservoir, three large plazas, several range structures, a ballcourt, six pyrami d temples 8 16 m high, and an acropolis over 20 m tall (Graebner 2002). Test pits in two of the plazas revealed several floors that yielded ceramics dating from c. 300 BCE through c. 900 CE (Conlon and Ehret 2002). Dispersed farmsteads are found in the i mmediate vicinity. For present purposes, we just want to note that the Maya resided here for at least 1200 years, suggesting an enduring, sustainable relationship with the local environment. The third kind of settlement is that found between Yalbac and Ca ra Blanca scattered farmsteads that mirror the distribution of fertile soils. In 2016, the Spanish Lookout Corporation permitted us to conduct a salvage archaeology program in recently cleared stretches of land where we focused our efforts in several diff erent areas (Benson 2017): 1) near Pool 7, a lake and the western most pool with residential settlement; 2) near Yalbac; and 3) the intermediate area between Cara Blanca and Yalbac. We excavated eight residential units, from small solitary mounds to a pla tform compound that yielded residential artifacts and eight burials with 14 individuals, most with grave goods. While most of the excavated materials near the surface indicate Late Classic or Terminal Classic occupation, we do not know how much history ha s been sheared off during clearing and plowing. At a small mound near Pool 7, for instance, we recovered two Postclassic projectile points dating to c. 900 and 1100 CE or later, suggesting the area might have been occupied post urban diaspora. Carbon (13C /12C, 13C) and nitrogen (15N/14N, 15N) isotope values provide information about human diet, specifically, C3 vs. C4 plant consumption and protein sources (Ambrose 1990). Maize, a C4 plant, was the main dietary staple for the Maya (Friewald 2011). Analy sis of teeth and bones from 2016 excavations (Carbaugh 2017) confirm that over half of the individuals diet consisted of C4 plants, likely maize, while nitrogen values indicate terrestrial herbivores were the predominate source of protein. All 13 individ uals sampled appear to have had similar dietary habits regardless of mound type and including two of the individuals interred in Structure 3, a ceremonial platform at the edge of Pool 1.

PAGE 217

Lucero, Larmon, and Carbaugh 225 Table 1 Strontium results of 2016 VOPA human remains. Field Mound Burial/ Human Cache Individual Age Sex Tooth 87Sr/86Sr Modified Teeth Pool 7 MF 4 1 A Young Adult (18 24 years) Male LM1 0.70804 No LM3 0.70773 B Mid Adult (35 40 years) ? RP2 0.70838 Yes C Adolescent (15 20 years) ? RM1 0.70794 Yes D Young Child (3 4 years) ? Rdm2 0.70777 No E Adolescent (12 15 years) ? RM1 0.70786 No RM2 0.70790 F Adolescent (16 20 years) ? LM1 0.70808 No LM3 0.70802 G Young Child (3 4 years) ? Rdm2 0.70801 No 4 A Infant (3 years (+/ 12 months) ? Ldm2 0.70778 No 1 2 A Adolescent (16 20 years) ? LM1 0.70800 Yes LM3 High Rb/Sr MF 4 1 East Str. 5 A Adolescent (15 years (+/ 3 years) ? LM1 0.70818 No LM3 0.70828 1 North Str. 8 A Young Adult (18 22 years) Male? LP1 0.70836 No Pool 1 Str. 3 2 Adolescent Young Adult (18 22 years) Male? RP1 0.70779 No Strontium isotope ratios (87Sr/86Sr) were also assessed, a method that examines individual and population level migration. Results for 11 of the individuals recovered from the hinterlands during the 2016 field season ranged from 0.70773 to 0.70838 ( Table 1 ). This tight range suggests that all these i ndividuals are local to the Yalbac hinterlands. In fact, the strontium ratio for several individuals fall within the local baseline established for sections of the Belize River Valley (Freiwald 2011). Specifically, VOPA individuals overlap with the Bel ize River Zone (3 teeth) and the Vaca Plateau Zone (4 teeth). For two individuals, their first molar fell outside of range for these zones, while the third molar fell within this range (Md 4 BU 1 Ind A and Md 1 BU 5). For example, the Pool 1 individual, buried in the ceremonial platform

PAGE 218

Tropical Climate, Landscape Transformation, and Human Interaction 226 Figure 4 DNA damage pattern of one of the teeth analyzed showing extreme damage and authenticating the results as ancient DNA and not contamination. R. Malhi/VOPA. (Str. 3), fits within the range of strontium ratios for individuals from the Yalbac area, perhaps indicating a local individual. We cannot say for certain if there was migration into the area or not since we do not have an established baseline for these sites as of yet. Ceramic forms and styles suggest visitors from the Petn, the northern lowlands, and eastern Belize (Ferree and Benson 2017; Kosakowsky 2017). Ripan Malhi and his PhD student Alyssa Bader at the University of Illinois at Urbana Champaign ( UIUC) have conducted preliminary ancient DNA analysis to explore population history, including migration and relatedness. Over the last decade ancient DNA analytical techniques has undergone a dramatic transformation with improvements in technology (e.g., Lindo et al. 2016) Whereas prior to these recent innovations researchers were limited to sequencing only a small segment of the mitochondrial genome, at present only a few labs worldwide are able to sequence complete mitochondrial and nuclear genomes tr ansforming the field of ancient DNA to the

PAGE 219

Lucero, Larmon, and Carbaugh 227 Figure 5 PCA plot of tooth overlaps with present -day Maya. These results are based on 184 DNA markers. R. Malhi/VOPA. field of paleogenomics. This includes Malhis lab at UIUC. Specific protocols and measures are taken to minimize contamination and detect it when it does occur. The ancient DNA lab at the Institute for Genomic Biology has positive pressure, hepa filtered air; and researchers have strict protocols to min imize and detect contamination. Dr. Malhi has begun assessing familial relationships through ancient DNA analysis using the program ANGSD or Analysis of next generation sequencing data (Korneliussen et al. 2014) Preliminary runs of two teeth and two cal culus samples from Late Classic burials from Mound 4 near Pool 7 (Figure 3) demonstrate acceptable DNA preservation with characteristic signs of DNA damage ( Figure 4 ). In addition, the genomic data were compared with populations worldwide, and one individual shows closest affinity to Maya, and the other individual exhibits the closest affinity to Pima both Native American populations ( Figure 5 ). As far as we know, this is one of the first successful DNA extractions from pre Columbian Maya human remains. It is only the beginning, and we plan further ancient DNA analysis pending funding. Concluding Remarks While these results are preliminary, they have one factor in common: they show the potential a regional interdisciplinary project has to understand huma n nature relationships. They also hint of a changing environment that ultimately gave rise to what we witness at present a healthy tropical forest with high biodiversity. The last few centuries of implementing a different interaction may soon change this fact (e.g., clear cutting large areas, mono cropping, widespread use of chemical fertilizers and pesticides, etc.). The Maya, even though immigrating into the area c. 12,000 years ago, adapted for millennia without destroying their home because of the su stainable strategies they used. They left a managerial, sustainable imprint on the forested landscape. In conclusion, some of the challenges the Maya faced at the end of the Classic period are similar to our own current struggles with accelerating global climate change. The more

PAGE 220

Tropical Climate, Landscape Transformation, and Human Interaction 228 we reveal about past decisions, the more ammunition we have to proceed in the long term to avoid destroying the place we call home. Humans, after all, as Aldo Leopold stated in 1939, are just one link of many in the biotic world (Leopold 1991[1939]:268). Acknowledgments We gratefully acknowledge the Institute of Archaeology for t heir support over the years. We also want to thank Forestland for their generous support, as well as the University of Illinois Center for Latin American and Caribbean Studies for a travel grant. Our foremen (Ernesto Vasquez and Cleofo Choc) and field as sistants, as usual, were of great help. Finally, we want to thank Yalbac Ranch, especially Jeff Roberson, for allowing us to continue exploring Cara Blanca, and Spanish Lookout Corporation for allowing us to conduct salvage archaeology on their property. Referenc es Ambrose, Stanley H. 1990 Preparatio and Characterization of Bone and Tooth Collagen for Isotopic Analysis. Journal of Archaeological Science 17:431-451. Benson, Erin M. 2017 Results of the 2016 Valley of Peace Archaeology Project Salvage Excavations In Results of the 2016 Valley of Peace Archaeology Project: Cara Blanca Pool 1 Excavations and the Yalbac Salvage Archaeology Program edited by L. J. Lucero, J. T. Larmon, and E. M. Benson, pp. 4372. Report submitted to the Institute of Arch aeology, National Institute of Culture and History, Belize. Buckley, Brendan M. 2015 Report on Belize Dendrochronology Project: May 2014 Expedition. In Results of the 2014 Valley of Peace Archaeology Project: Underwater and Surface Explorations at Cara Bl anca, edited by L. J. Lucero, pp. 125-129. Report submitted to the Institute of Archaeology, National Institute of Culture and History, Belize. Carbaugh, Aime E. 2017 Analysis Human Skeletal Remains from Cara Blanca Pool 1 and Yalbac Salvage Program. In Results of the 2016 Valley of Peace Archaeology Project: Cara Blanca Pool 1 Excavations and the Yalbac Salvage Archaeology Program edited by L. J. Lucero, J. T. Larmon, and E. M. Benson, pp. 107 -157. Report submitted to the Institute of Archaeology, Natio nal Institute of Culture and History, Belize. Conlon, James M., and Jennifer J. Ehret 2002 Time and Space: The Preliminary Ceramic Analysis for Saturday Creek and Yalbac, Cayo District, Belize, Central America. In Results of the 2001 Valley of Peace Archaeology Project: Saturday Creek and Yalbac edited by L. J. Lucero, pp. 820. Report submitted to the Department of Archaeology, Ministry of Tourism and Culture, Belize. DAlmeida, C., C. J. Vrsmarty, G. C. Hurtt, J. A. Marengo, S. L. Dingman, and B. D. Keim 2007 The Effects of Deforestation on the Hydrological Cycle in Amazonia: A Review on Scale and Resolution. International Journal of Climatology 27:633 647. Douglas, Peter, M. J., Mark Pagani, Marcello A. Canu to, Mark Brenner, David A. Hodell, Timothy I. Eglinton, and Jason H. Curtis 2015 Drought, Agricultural Adaptation, and Sociopolitical Collapse in the Maya Lowlands. Proceedings of the National Academy of Sciences 112: 5607 -5612 Fedick, Scott L. 2010 The Maya Forest: Destroyed or Cultivated by the Ancient Maya? Proceedings of the National Academy of Sciences 107:953 -954. Ferree, Tyler, and Erin M. Benson 2017 Ceramics from the 2016 VOPA Field Season. In Results of the 2016 Valley of Peace Archaeology Project: Cara Blanca Pool 1 Excavations and the Yalbac Salvage Archaeology Program edited by L. J. Lucero, J. T. Larmon, and E. M. Benson, pp. 1544. Report submitted to the Institute of Archaeology, National Institute of Culture and History, Belize. Ford, Anabel, and K. C. Clarke 2016 Linking the Past and Present of the Ancient Maya: Lowland Land Use, Population Distribution, and Density in the Late Classic Period. Oxford Handbook of Historical Ecology and Applied Archaeology edited by C. Isendahl an d D. Stump. Oxford, Oxford University Press. DOI: DOI: 10.1093/oxfordhb/9780199672691.013.33 Ford, Anabel, and Ronald Nigh 2015 The Maya Forest Garden: Eight Millennia of Sustainable Cultivation of the Tropical Woodlands Left Coast Press, Walnut Creek. Freiwald, Carolyn 2011 Maya Migration Networks: Reconstructing Population Movement in the Belize River Valley During the Late and Terminal Classic Unpublished PhD dissertation, University of Wisconsin Madison.

PAGE 221

Lucero, Larmon, and Carbaugh 229 Graham, Elizabeth 2011 Maya Christians and their Churches in Sixteenth -century Belize. University Press of Florida, Gainesville. Graebner, Sean M. 2002 Monumental Architecture and the Ancient Maya: The Royal Acropolis at Yalbac, Central Belize. Unpublished Masters thesis, Dep artment of Sociology and Anthropology, New Mexico State University, Las Cruces. Kennett, D. J., S. F. M. Breitenbach, V. V. Aquino, Y. Asmerom, J. Awe, J. U. L. Baldini, P. Bartlein, B. J. Culleton, C. Ebert, C. Jazwa, M. J. Macri, N. Marwan, V. Polyak, K M. Prufer, H. E. Ridley, H. Sodemann, B. Winterhalder, and G. H. Haug 2012 Development and Disintegration of Maya Political Systems in Response to Climate Change. Science 338, 788-91. Korneliussen, T. S., A. Albrechtsen, and R. Nielsen 2014 Angsd: Analysis of Next Generation Sequencing Data. BMC Bioinformatics 15(1):356. Kosakowsky, Laura 2017 VOPA Ceramics 2016 General Summary. In Results of the 2016 Valley of Peace Archaeology Project: Cara Blanca Pool 1 Excavations and the Yalbac Salvage Archaeology Program edited by L. J. Lucero, J. T. Larmon, and E. M. Benson, pp. 1114. Report submitted to the Institute of Archaeology, National Institute of Culture and History, Belize. Larmon, Jean T., S. Ambrose, H. G. McDonald, B. W. Fouke, H. Wang, and L. J. Lucero. 2016 What the wild things ate: isotopic analysis of an extinct giant sloth tooth from Cara Blanca, Belize. 50th Annual GSA North-Central Section Meeting, Paper No. 7 8, Geological Society of America Abstracts with Programs 48. Larmon, Jean T., Stanley H. Ambrose, H. Gregory McDonald, and Lisa J. Lucero n.d. Paleoecology of the Giant Ground Sloth, Eremotherium laurilardii (Xenarthra, Megatheriidae) from Cara Blanca, Belize. To be submitted to Science. Leopold, Aldo 1991[1939] Biotic View of the Land In The River of the Mother of God: And other Essays by Aldo Leopold, edited by J. B. Callicott and S. L. Flader pp. 266271. University of Wisconsin, Madison. Lindo John, Emilia Huerta -Sanchez, Shigeki Nakagome, Morten Rasmussen, Barbara Petzelt, Joycelynn Mitchell, Jerome S. Cybulski, Eske Willerslev, Michael DeGiorgio, and Ripan S. Malhi 2016 A Time Transect of Exomes from a Native American Population Before and After European Contact. Nature Communications 7: 13175/DOI: 10.1038/ncomms13175. Lohse, Jon C., Jaime Awe, Cameron Griffith, Robert M. Rosenswig, and Fred Valdez, Jr. 2006 Preceramic Occupations in Belize: Updating the Paleoindian and Archaic Record. Latin American Antiquity 17:209-226. Lucero, Lisa J. 2006 Water and Ritual: The Rise and Fall of Classic Maya Rulers Uni versity of Texas Press, Austin. 2011 VOPA 2011: Exploring Cara Blanca Pool 1 and Assessing Yalbac Report submitted to the Institute of Archaeology, National Institute of Culture and H istory, Belize. 2017 Ancient Maya Water Management, Droughts, and Urban Diaspora: Implications for the Present. In Tropical Forest Conservation: Long Term Processes of Human Evolution, Cultural Adaptations and Consumption Patterns pp. 162 188. UNESCO Mex ico. in press A Cosmology of Conservation in the Ancient Maya World. Journal of Anthropological Research 74(3) Fall 2018 Lucero, Lisa J., and Andrew Kinkella 2015 Pilgrimage to the Edge of the Watery Underworld: An Ancient Maya Water Temple at Cara Blanca, Belize. Cambridge Archaeological Journal 25:163 -185. Lucero, Lisa J., Roland Fletcher, and Robin Coningham 2015 From Collapse to Urban Diaspora: The Transform ation of Low -Density, Dispersed Agrarian Urbanism. Antiquity 89:11391154. Lucero, Lisa J., Joel D. Gunn, and Vernon L. Scarborough 2011 Climate Change and Classic Maya Water Management. Water 3 :479 494; doi: 10.3390/w3020479 Lucero, Lisa J., Jessica Harrison, Jean Larmon, Zachary Nissen, and Erin Benson 2016 Prolonged Droughts, Short -Term Responses and Diaspora: The Power of Water and Pilgrimage at the Sacred Cenotes of Cara Blanca, Belize. WIREs Water doi:10.1002/wat2.1148 Masson, Marilyn A., and David A. Freidel 2012 An argument for Classic era Maya market exchange. Journal of Anthropological Archaeology 31:455-484. McDonald, H. Gregory 2015 Pleistocene Faunal Remains from Cara Blanca, Belize In Results of the 2014 Valley of Peace Archaeology Project: Underwater and Surface Explorations at Cara Blanca edited by L. J.

PAGE 222

Tropical Climate, Landscape Transformation, and Human Interaction 230 Lucero, pp. 113124. Report submitted to the Institute of Archaeology, National Institute of Culture and History, Belize. Medina -E lizalde, Martn, Stephen J. Burns, David W. Lea, Yemane Asmerom, Lucien von Gunten, Victor Polyak, Mathias Vuille, and Ambarish Karmalkar 2010 High Resolution Stalagmite Climate Record from the Yucatn Peninsula Spanning the Maya Terminal Classic Period. E arth and Planetary Science Letters 298:255262. Meineke, Emily, Elsa Youngsteadt, Robert R. Dunn, and Steven D. Frank 2016 Urban Warming Reduces Aboveground Carbon Storage. Proceedings of the Royal Society B 283: 20161574, doi.org/10.1098/rspb.2016.1574 Prufer, Keith M., Clay ton R. Meredith, Asia Alsgaard, Timothy Dennehy, and Douglas Kennett 2017 The Paleoindian Chronology of Tzib Te Yux Rockshelter in the Rio Blanco Valley of Southern Belize. Research Reports in Belizean Archaeology 14:321-326. Robert s, Patrick, Chris Hunt, Manuel Arroyo -Kalin, Damian Evans, and Nicole Boivin 2017 The Deep Human Prehistory of Global Tropical Forests and its Relevance for Modern Conservation. Nature Plants 3 Article number: 17093. doi:10.1038/nplants.2017.93 Rosenswi g Robert M., Deborah M. Pearsall, Marilyn A. Masson, Brendan J. Culleton, Douglas J. Kennett 2014 Archaic Period Settlement and Subsistence in the Maya lowlands: New Starch Grain and Lithic Data from Freshwater Creek, Belize. Journal of Archaeological Science 41:308-321. Rosenswig, Robert M., Amber M. VanDerwarker, Brendan J. Culleton, and Douglas J. Kennett 2015 Is it Agriculture Yet? Intensified Maize -use at 1000 cal BC in the Soconusco and Mesoamerica. Journal of Anthropological Archaeology 40:89 108. Sabloff, Jeremy A. 2007 It depends on how you look at things: new perspectives on the Postclassic period in the northern Maya lowlands. Proceedings of the American Philosophical Society 151:1125. Scarborough, Vernon L. 1993 Water Management in the Southern Maya Lowlands: An Accretive Model for the Engineered Landscape. Research in Economic Anthropology 7:17 -69. Scarborough, V. L., N. P. Dunning, K. B. Tankersley, C. Carr, E. Weaver, L. Grizioso, B. Lane, J. G. Jones, P. Buttles, F. Valdez, and D. L. Lentz. 2012 Water and sustainable land use at the ancient tropical city of Tikal, Guatemala. Proceedings of the National Academy of Sciences 109:1240812413. Turner, B. L., and Jeremy A. Sabloff 2012 Classic Period Collapse of the Central Maya Lowlands: Insights about Human Environment Relationships for Sustainability. Proceedings of the National Academy of Sciences 109: 13908 13914

PAGE 223

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 231 240 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 21 CEREMONIAL CIRCUIT(S) AT CARA BLANCA, BELIZE Jean T. Larmon and Aime E. Carbaugh This paper presents the results from the exploration of the three previously unexplored pools in the hypothesized ceremonial circuit of Cara Blanca, central Belize. During the 2017 field season, the Valley of Peace Archaeological project surveyed the three far eastern pools of the Cara Blanca 25 (Pools 22, 23, 25) in order to explore the role they might have played in the L ate to Terminal Classic ritual landscape. D uring the Classic period (250900 CE) many Maya rulers garnered power by exploiting their followers reliance upon rain. When several prolonged and severe droughts struck the Maya area during the Terminal Classic period and the rains failed, so too did rulers power. Despite periodic droughts, the Cara Blanca pools in central Belize remained a consistent resource for freshwater. Data collected over many years of research at Cara Blanca suggest that the 25 pools are part of a ritually prescribed path, or ceremonial circuit, which developed in part as a response to rulers failures. Ceremonial circuits are paths that Maya walked connecting built architecture in a way that makes explicit their relationship to that space including both the architecture and sacred, unbuilt spaces. This paper explores how these three pools, as well as two of the other Cara Blanca pools (Pools 1 and 15), may have played a unique role in the ritual landscape. Introduction During the Classic period (250900 CE) many Maya rulers garnered power by exploiting their followers reliance upon rain. When several prolonged and severe droughts struck the Maya area during the Terminal Classic period and the rains failed, so too did rulers power (Lucero 2006). Despite periodic droughts, the Cara Blanca pools in central Belize remained a consistent resou rce for freshwater. Data collected over many years of research at Cara Blanca suggest that the 25 pools, which formed along an east west axis, are part of a ritually prescribed path, or ceremonial circuit, which developed in part as a response to rulers failures (Lucero et al. 2016, 2017). Ceremonial circuits are paths that Maya walked connecting built architecture in a way that makes explicit their relationship to that space including both the architecture and sacred, unbuilt spaces (Vogt 1969). Often a ceremonial circuit follows the path of the sun moving from east to west (Astor Aguilera 2010:131143; Ashmore 2009). During the 2017 field season, the Valley of Peace Archaeological project surveyed the far eastern pools of the 25 (Pools 22, 23, 25). B ecause these three pools are the eastern most in the system, they may have been a point of departure for the ceremonial journey. Additionally, if multiple ritual processions were undertaken at Cara Blanca, investigation of the final three pools integrati on into the landscape is essential to understanding the space. This paper presents the results from our survey and Figure 1 Map of all 25 Cara Blanca pools showing soil class. Courtesy of VOPA. the role the pools might have played in the Late to Terminal Classic ritual landscape. Cara Blanca, Belize Cara Blanca, in Central Belize, is a system of 25 pools, both shallow lakes and cenotes steep sided water filled sinkholes ( Figure 1 ). These pools line the base of a steep limestone cliff, which rises up 100 m above the pools. The blue, sometimes muddied waters stand in stark contrast to the white limestone cliff and the dense primary and secondary jungle vegetation. Each pool is unique with the cenotes ranging from 5 60 m deep and the lakes from 2 18 m. The western most pools, Pools 79, are lakes and have ancient Maya settlements on their southern sides. The central pools, Pools 15, which are all cenotes, have a noticeable dear th of residential

PAGE 224

Ceremonial Circuit(s) at Cara Blanca 232 settlement and a handful of hypothesized ceremonial structures. The clear juxtaposition of pool use, particularly the lack of residential structures near the cenotes, which would have retained water regardless of drought conditions, suggests a specialized use of the cenotes It was only when the Maya experienced several prolonged droughts c. 800900 CE (Medina Elizalde et al. 2010) that they built anything of a substantial nature in these previously untouched areas near the central pools, constructing what we argue might be pilgrimage destinations (Lucero et al. 2016, 2017). Lucero and colleagues have hypothesized that these pools were part of a ceremonial circuit, accessed by the Maya during the late Late (700 800 CE) and Terminal Classic (800950 CE) periods (Lucero and Kinkella 2015). The importance of surveying the final three Cara Blanca pools lies in both completing the survey of the Cara Blanca area, as well as, collecting further data to inform our hypothesis. Here, we present the results of the 2017 survey and previously collected data at Cara Blanca Pool 1 (Lucero et al. 2017; Larmon 2017), which, taken in concert, might offer further insight into the significance of the Cara Blanca space. We arg ue that Cara Blanca was home to a ceremonial circuit, or ritual processions. This is indicated by the synchronic distinction of residential and ceremonial spaces by isolating the sacred spaces of the pools from residential areas (Lucero et al. 2017). In this case, both the built and unbuilt spaces are essential. As visitors walked the paths connecting the constructed ceremonial spaces they passed through the jungle, an unmodified space, in effect tying together anthropogenic and natural spaces in a w ay that deconstructs the dichotomy between the two. Traditionally, the Maya walk ceremonial circuits to reaffirm their relationship with and to sacred, forested places (e.g. Vogt 1969:144, 149, 390). As community members, or perhaps members of multiple re sidential communities, processed through the Cara Blanca space, they might have followed the path of the sun, from east to west (Astor Aguilera 2010:131143; Ashmore 2009) or they might have been processing multiple ceremonial paths at Cara Blanca. Based upon ceramic chronologies, the Cara Blanca space was visited most formally during the late Late and Terminal Classic period (Kosakowsky 2017). As several, prolonged and severe droughts struck the region and political turmoil encouraged people to migrate out of centers to the hinterlands, the Cara Blanca circuit was formalized with the construction of ritual structures (Larmon 2017; Larmon and Amin 2017; Larmon and Nissen 2015; Lucero et al. 2016, 2017; Lucero and Kinkella 2015). The importance of surveying the final three Cara Blanca pools lies in both completing the survey of the Cara Blanca area, as well as collecting further data to inform our hypothesis that Cara Blanca was home to a ceremonial circuit, or set of ritual processions. Pool 1 and its asso ciated structures would have served a pivotal role in the Cara Blanca circuit, perhaps acting as a locus of ritual for those taking part in the procession. We will first present evidence for water related rituals at Pool 1 and its associated structures. Then, we will introduce the final three pools, which were surveyed this summer, as well as discuss our attempts to explore a previously unknown sinkhole near Pool 15. Finally, we consider how Cara Blanca may have been one large ceremonial circuit, as well as how each pool might have played an important role in independent ritual processions. Previous Fieldwork at Pool 1: 2014 2016 Located on the southern edge of Pool 1 is a ceremonial complex comprised of three structures (Str. 1, 2, and 3) built around a plaza with the cenote situated on the northern side of the plaza ( Figure 2 ). A number of other structures are located nearby and are likely related to the ritual events which took place at Pool 1. Archaeological excavations at Pool 1 began in 2012 when the VOPA crew excavated Str. 1, a hypothesized water temple on the edge of Pool 1, which is the deepest of the cenotes at 60+ meters. Excavations by Lucero and colleagues uncovered a collapsed corbel vaulted building measuring 20 x 7.5 m, 3.5 m tall, atop a 2m wide stepped platform (Lucero et al. 2016; Lucero and Kinkella 2015). With fill of tufa (a rock formed by the accumulation of calcium carbonate around organic materials in the water), an unusually high number of water jars (72.1% of the recovered v essels), and water laden

PAGE 225

Larmon and Carbaugh 233 Figure 2 Map of Pool 1 showing associated Structures, left to right M -186 Sweatbath, Pool 1 Str. 1, Pool 1 Str. 3. Adapted from Kinkella 2007:Fig .3.8. Str. 1 photograph by Tony Rath. Sweatbath and Str. 3 photograph courtesy of VOPA. symbolism (see Lucero and Kinkella 2015), Str. 1s connection to water and water rituals is evident. The temple appears to have restricted access, with narrow hallways and small doorways. Based upon ceramic chronologies, it was constructed rapidly, either as one event or multiple, diachronically close events. Just across the plaza and 22 meters from Str. 1, Str. 3 may have acted as the loci for water related rituals. The structure itself is roughly 7.46 x 3.65 m, though the east and south edges of the platform melt into the landscape, and 0.8 m tall (Larmon 2017). During the 2014 season, we began excavations on the structure. We exposed nearly the extent of the structure, including a layer of medium to large sized boulders that had been placed over a burned plaster surface covered in a sheet of smashed, partial ceramic vessels (Larmon and Nissen 2015). During the 2016 season, we conducted additional excavations to explore beneath the burned plaster s urface of Str. 3 and found that, based upon ceramic chronologies and like Str. 1, the structure appears to have multiple construction events within a relatively short time frame in the late Late Classic/Terminal Classic period. Several features of Str. 3 s construction suggest that this platform played an important role in the ritual process. On the center edge of the north side c. 1 m from the water, they built a step from which visitors to the structure likely made offerings into the pool. Just south of this step, and between the uppermost ballast and fill on the north end of the structure, we exposed a feature of flat stones that could have served as additional support for multiple processions to the edge of the pool and concomitant rituals. In total, we recovered 6792 ceramic sherds from Str. 3, the majority of which came from the surface, with layers of burned sherds on top of a burned plaster surface. The vessels represent styles from different regions, overlapping with those of the Belize Valley, northern Belize, and the eastern Petn. This diversity does not necessarily indicate that the ceramics were imported, but rather that sites in the region might be peripherally linked to other regions (Kosakowsky 2017). The low percentage of rims (8.3%, n= 563) suggest that the Maya either smashed vessels and removed vessel parts, or brought pieces to Pool 1 from elsewhere. Although we cannot determine ceramic origins without petrographic analysis,

PAGE 226

Ceremonial Circuit(s) at Cara Blanca 234 Table 1 Measurements of the final three pools (22, 23, 25) and Pool 21. Pool Diameter Depth 21 c. 60 x 60 m c. 13 m 22 c. 30 x 30 m c. 6 m 23 c. 30 x 30 m c. 11 m 25 c. 25 x 25 m c. 9 10 m the lack of complete vessels demonstrates the significance of the pieces themselves. Pool 1 visitors might have brought connections to and representations of their home and community in the form of vessels or sherds to tie them from this threshold to the oth erworld. We also exposed three individuals buried in Str. 3 (Carbaugh 2017). Each of the three individuals was interred in the ceremonial structure without any grave goods, suggesting that the three burials could have served as deposits or even caches i n a dedicatory ceremony, designating the space as sacred. Maya visitors would have walked over these three buried caches as they proceeded to the step from which they threw offerings into the cenote. Each individual was interred in different strata and progressively getting closer to the surface near the northern edge of the platform, pulling the visitor along the platform, towards the otherworld, as they traverse the structure. Just 400 m to the west of Pool 1 is the hypothesized sweatbath, M 186. Initi al investigations at M 186 revealed a squirclelike building that abuts the west end of a longrange structure that itself has two or three looters trenches. Measuring exactly 3.66 x 3.66 m, the western most room resembles a sweatbath. It has rounded c orners and a semi domed roof (which was nearly completely destroyed both by looters and by 2010 Hurricane Richard), characteristics that have been noted in numerous other sweatbaths throughout the Maya region (see Larmon and Amin 2017). Though the ceramic assemblage of the sweatbath dates to the same periods as Pool 1, late Late and Terminal Classic periods, the composition is different. Only seven of the 16 rims recovered (or 43.8%) were jar rims, compared to 57.7% from Str. 3 and 72.1% jar rims from Str 1. This difference suggests a use of the sweatbath distinct from the other excavated structures at Pool 1. We hypothesize that the sweatbath was used by visitors to cleanse themselves before participating in rituals at the edge of Pool 1 (Larmon and Am in 2017). The 2017 Survey: Final Unexplored Sinkholes If Cara Blanca served as a ceremonial circuit, each cenote along its path could have functioned as the focal point for the performance of water related and other ceremonies. From 1998 2014, Andrew Kinkella (2008, 2009, 2015) conducted an extensive and thorough survey of the Cara Blanca area, identifying the 25 pools and areas that were in need of further study. Though Kinkella had been able to reach 22 of the Cara Blanca pools he did not reach the three eastern most pools. Survey conditions to these three were particularly strenuous because of complete inundation with swampy waters, a prevalence of Black Poisonwood trees, and various hazards that accompany swampy jungle conditions. Pool 21, the la st pool reached by Kinkella, lies at the western most edge of the swamp. He had attempted the survey from the east, choosing the shortest path from the Yalbac road. In order to avoid some of the issues that he ran into, he suggested that we try to approac h the pools from the north, first reaching Pool 21 and then moving east ward to Pools 22, 23, and 25 ( Table 1 ). Our exploration began by cutting our way south from an overgrown side road off of the main Yalbac road. Figure 3 shows our attempted path to ea ch pool. Pool 21 ( Figure 4 ), approximately 60 x 60 m and 13 m deep, was muddied by the recent rains. Swimming around the edge of the pool we noted an outflow with a strong current on the east side leading to Pool 22, which is just c. 100 m to the east. Unfortunately, just east of Pool 21, the landscape becomes completely inundated

PAGE 227

Larmon and Carbaugh 235 Figure 3 Map of our 2017 routes to visit Pools 22, 23, and 25. Courtesy of VOPA. Figure 4 Drone shot of Pool 21, 22, 23, 25. Courtesy of VOPA. Figure 5 The inundated landscape surrounding Pools 23, 22, and 23. Photography by Tony Rath. ( Figure 5), making both survey and access to the eastern pools very difficult. The pools are surrounded by swampy conditions and thick stands of red mangrove. The waist dee p swampy jungle terrain imposed itself upon us, making clear that any who try to navigate the landscape are at once enveloped and appreciative of the journey. Because of the conditions surrounding the pools, it was difficult to get adequate photos of them while on land. Instead, we sent up the drone to get aerial images. We used a Phantom 3D Professional drone often during survey to correctly orient ourselves to the pools for which we had no GPS points. Pool 22, approximately 30 x 30 m in diameter and 6 m deep, and Pool 23 (see Figure 4), approximately 30 x 30 m in diameter and 11.2 m deep, are nestled within dense vegetation and an inundated flood plain. Dr. Ed Boles, an Aquatic ecologist and environmental consultant for various Belizean institutions (Be lize Center for Environmental Studies, Belize Audubon Society, Belize Electric Company, Belize Department of the Environment, United Nations Development Programme, The Nature Conservancy, University of Belize, CATIE), accompanied us for the Pool 22 and 23 survey, and was able to provide us with an assessment of their biodiversity. As he mentioned in an email (June 23, 2017), these pools are particularly interesting for a number of reasons. The presence of red mangroves inland suggests that these pools could be remnants of a time when the sea level was much higher mangroves are tolerant to salt and mineral rich water. The fact that these stands survived in isolation suggests that the pools are heavily fed by groundwater that is mineral rich. This is furth er evidenced by the presence of tufa in both Pool 22 and 23, and iron oxide coatings on clam shells, tufa, and so on. Both pools had water lilies, which act as ecosystem engineers, providing habitat quality control and funnel oxygen from the atmosphere into the sediments. Dr. Boles also noted that the pools had abundant amounts of wood, often whole trees that have likel y been deposited by hurricanes over the years. The trees had a thick, dense layer of periphyton (algae, fungi, bacteria, cyanobacteria, protozoans, microinvertebrates) that represent s a production system. In his preliminary survey of these two pools, D r. Boles noted a number of species of fish, crabs, and clams ( Table 2 ). We also noted

PAGE 228

Ceremonial Circuit(s) at Cara Blanca 236 Table 2 Biodiversity table constructed by Dr. Ed Boles. turtles and crocodiles. His survey highlights the resource richness of these pools. The last pool that we visited was Pool 25 (see Figure 4), the eastern most in the Cara Blanca system. This pool, c. 400 m east of Pool 23, ended up being the most difficult to access. Based on aerial images, we planned to hit an open area c. 200 m from Pool 25. Our drone flights the previous days had shown us that the expansive grasslands that encompass Pool 25 had been recently burned, except for a strip along the edge of the fores t. The burning around Pool 25 was caused by arsonpoachers coming into the area at night and advertently or inadvertently setting fire to the property. We expected to hit this open grassland and burned landscape and be able to see the pool. Instead, the opening that we had seen in the aerial images was cutting grass that rose well over our heads, the densest vegetation that we had encountered yet. In addition, the ground was completely inundated and deep unseen watery holes often caught us unawares making traversing the area difficult. Finally, we reached Pool 25, c. 25 x 25 m and 9 10 m deep. There was an input on the west side of the pool, likely connecting Pool 25 to Pools 23, 22, and so on. But, again, the area in between these two pools was completely inundated. Ultimately, we did not note any additional Maya settlement with this survey, but this would have been near impossible given the field conditions. In order to better understand how these pools are interwoven in Cara Blancas history, we need to assess their condition in the dry season. We can, however, attest to their resource richness and their magnetic qualities, the perfectly round pools that might have been tucked away in dense and harsh vegetation a sort of haven. Sinkhole near Pool 15 On our final day in the field, we attempted to revisit Pool 15 and associated caves. Kinkella had previously identified three caves just 500 m northwest of Pool 15 (Kinkella 2009). These caves are small and were used for rituals indicated by the Terminal Classic (c. 800 900 CE) jar sherds recovered. In addition, Kinkella noted on the escarpment above Pools 14 and 15 that the Maya built seven structures, a possible water shrine (Kinkella 2009:138142). During a 2016 fly over, Tony Rath noted an additional, large cave or sinkhole that sits just below the

PAGE 229

Larmon and Carbaugh 237 lookout with the water shrine and above Pool 15 ( Figure 6 ). There is a chance that the large cave extends down into the pool, though this is speculative. Unfortunately, we choose the wrong route to access the cave. Though we were able to access a road that brought us south to just 0.6 km from the cave, we ended up traversing the escarpment to the east of the cave and about 200 m out hit the escarpments steep edge which we could not descend safe ly and had to turn back. While we did not reach the cave, we did come up with a better access point for future exploration. The Pool 15 area could provide important insights regarding the presence and use of ceremonial circuits in the Cara Blanca region. As ReeseTaylor (2002:159163) mentions, one of the most common ethnographic ritual circulations is from the base to the summit of a mountain, through which participants are considered to be uniting the three worlds: the underworld (from the cenote), the human world (from the surface of the water up the cliff, perhaps through the cave), and the heavens (on the hill top shrine). Pool 15 might represent an ancient example of this contemporary ritual. In a future field season, it would be worth following the drainage ditch down to Pool 15 or walking over from Pool 2. From there, it might be possible to explore the large cave and move up to the hilltop shrine. Otherwise, it will be easiest to approach the water shrine from the west, the path of least resistan ce. Discussion and Conclusions Cara Blancas use as a ceremonial circuit has been explored previously by the VOPA team (Lucero et al. 2016, 2017; Lucero and Kinkella 2015). Cara Blanca has a number of purely ritual structures and spaces, including the Water Temple, Str. 3, and the sweatbath. As evidenced by ceramic styles, visitors to these pools appear to be coming from all over Belize and beyond (though this will have to be confirmed with petrographic analysis). The timing of the use and material footprint at Cara Blanca all suggest its sacred nature. The pools themselves, as watery portals to the underworld, signify a sacred space. Vogt (1969) writes of the importance of rituals at water holes in ethnographic examples of ceremonial Figure 6 Drone shot of Pool 15 showing the associated cave and hilltop/water shrine. Courtesy of VOPA. processions to cross shrines on caves, hilltops, and in households (at Cara Bla nca, that would be cenotes, ceremonial structures, and hilltops). Finally, no residential settlement was imposed upon the Cara Blanca space. Certainly, the treatment of the Cara Blanca landscape would corroborate its use as a space for ritual procession. What we should consider, however, is that it was not a single procession taking place at Cara Blanca. Kathryn Reese Taylor (2002) notes that while it is common for there to be more than one type of ritual (in this case procession) in a particular geogra phy, this fact is often overlooked when considering ritual circuits. She proposes that there are distinct types of ritual circuits and that the incorporation of one or more of these ritual circuits is requisite to the design of a proper Maya civic center (Reese Taylor 2002:144). And, again though it is often overlooked, it is not just primary stops within a circuit that are important but also the circuits themselves, the journey. Cara Blanca was visited most during the Terminal Classic period, as civic centers lost their power and Maya began to migrate away from cities. Therefore, in this analysis Cara Blanca plays the role of the center of power and we consider the multiple processions that may have taken place in the watery landscape. The first type o f processions discussed by Reese Taylor is ritual circumambulation. This consists of all participants moving in a counter clockwise procession from one point on the landscape to designated other points (Reese

PAGE 230

Ceremonial Circuit(s) at Cara Blanca 238 Taylor 2002:145). It is in this procession th at the three pools explored during the 2017 field season play a role as the eastern most pools they would have been the point of departure for this procession (Astor Aguilera 2010:131143; Ashmore 2009). The landscape within which Pools 21, 22, and 23 are nestled is completely inundated. Participants in the procession would have been emerging from the water, the lower world, and moving west wards to the rest of the ceremonial circuit. It is this circuit that would have encompassed the entire landscape. Ritual circumambulation has been used to define and maintain boundaries, and this procession might have been a means of further setting apart the ritual landscape from surrounding utilitarian activities, as well as a means of connecting more deeply to the landscape. It is possible, too, that additional smaller processions were carried out within the landscape. The next type of procession that Reese Taylor defines is a banner procession, which moves from the periphery to the center. These processions are a mechanism to strengthen integration and social solidarity within towns or villages comprised of dispersed settlements (Reese Taylor 2002:152). At Cara Blanca, Pool 1, the locus of ritual activity, might have acted as the central point of the circuit. As the deepest of the cenotes it is the portal that would bring visitors closest to the underworld. We have previously hypothesized that there was a procession from the sweatbath, just c. 400 m to the west of Pool 1, to the Water Temple (Larmon and Amin 2017; Lucero et al. 2017). Visitors to the space cleansed themselves at the sweatbath before engaging with the rest of the landscape. Here, the sweatbath would be considered the periphery and the processions would have been a means of integrating the lan dscapehighlighting the idea that the entire landscape, not just the ceremonial stops, is essential. Ethnographically, these processions have been a way to indicate rotating political authority and social integration perhaps the procession at Cara Blanca Pool 1, when accessed by diverse visitors, acted to disseminate authority, all of our feet share this path, and integrate a diverse community. The final procession type discussed by Reese Taylor are processions from the base to the summit of a mountain which connects the three realms of the cosmos (Reese Taylor 2002:159). In these cases, the procession moves from the south (often a depression or body of water) to the north (the hilltop). At Cara Blanca, this type of procession might have been a part of the Pool 15 space. As mentioned above, at Pool 15, there is a cenote to the south, sitting just below a cave on a hillside that opens up to a hilltop upon which Kinkella (2009) noted a possible water shrine. If the Maya emerged from the cenote perhap s through the cave, and processed to the shrine on top of the hill, they would have symbolically been connecting the watery underworld, the human realm, and the heavens. This act imbues the Cara Blanca landscape with the sacred essence of the otherworlds and integrates the landscape so as to facilitate supplication and communication. In conclusion, though we did not identify additional Maya architecture that can inform us directly of how the Maya were interacting with the final three pools at Cara Blanca, the material recovered from other pools, particularly from Pool 1 and the potential for ambulatory use near Pool 15 suggest that Cara Blanca served as an active circuit. If this is the case, and visitors to the space followed ethnographic examples (from e ast to west), then the three most easterly pools would likely have been the beginning of a ceremonial journey, with the trials of traversing the spaces between constructed features contributing to the significance of the experience. Here, we move beyond a singular consideration of the Cara Blanca space and explore the diverse ways in which the landscape was a materially experienced example of the integration of the worlds, the physical and the spiritual. The processions discussed above are not exhaustive and certainly there are additional ways that visitors to Cara Blanca may have connected to the landscape. Whether the space served as a cohesive circuit, each pool impacting the way another is experienced, or each pool and path was accessed separately and for distinct reasons, the weight of the journey resided in the process and nature of the mutually experienced human and environment interaction.

PAGE 231

Larmon and Carbaugh 239 References Astor Aguilera, Miguel Angel 2010 The Maya World of Communicating Objects: Quadripartite Crosses, Trees, and Stones. University of New Mexico Press, Albuquerque. Ashmore, Wendy 2009 Meoamerican Landscape Archaeologies. Ancient Mesoamerica 20:183-187. Carbaugh, Aime E. 2017 Analysis of Human Skeletal Remains from Cara Blanca Pool 1 and the Yalbac Salvage Program. In Results of the 2016 Valley of Peace Archaeology Project: Cara Blanca Pool 1 Excavations and the Yalbac Salvage Archaeology Program edited by L. J. Lucero, J. T. Larmon, and E. M. Benson, pp. 107 -157. Report submitted to the Ins titute of Archaeology National Institute of Culture and History, Belize. Kosakowsky, Laura 2017 VOPA Ceramics 2016 General Summary. In Results of the 2016 Valley of Peace Archaeology Project: Cara Blanca Pool 1 Excavations and the Yalbac Salvage Archaeolo gy Program edited by L. J. Lucero, J. T. Larmon, and E. M. Benson, pp. 1114. Report submitted to the Institute of Archaeology National Institute of Culture and History, Belize. Kinkella, Andrew 2008 Over the Bajo and Through the Pools: The 2007 Settleme nt Survey Transect from Yalbac to the Cara Blanca Pools. In Results of the 20 07 Valley of Peace Archaeology Project: Yalbacs Settlement, edited by L. J. Lucero, pp. 99-114. Report submitted to the Institute of Archaeology, National Institute of Culture an d History, Belize. 2009 Draw of the Sacred Water: An Archaeological Survey of the Ancient Maya Settlement at the Cara Blanca Pools, Belize Unpublished PhD dissertation, Department of Anthropology, University of California, Riverside. 2015 Chechem at the End of the Road: The 2014 Cara Blanca Settlement Survey (CBSS). In Results of the 2014 Valley of Peace Archaeology Project: Underwater and Surface Exploration at Cara Blanca, edited by L. J. Lucero, pp. 146151. Report submitted to the Institute of Archaeology, National Institute of Culture and History, Belize. Larmon, Jean T. 2017 Cara Blanca Explorations: Pool 1 Structure 3 and Related Investigations. In Results of the 2016 Valley of Peace Archaeology Project: Cara Blanca Pool 1 Excavations and the Yalbac Salvage Archaeology Program edited by L. J. Lucero, J. T. Larmon, and E. M. Benson, pp. 86-106. Report submitted to the Ins titute of Archaeology National Institute of Culture and History, Belize. Larmon, Jean T. and Anuj Amin 2017 The Cara Bla nca Sweatbath: Ritual Purification on the Ceremonial Circuit. In Results of the 2016 Valley of Peace Archaeology Project: Cara Blanca Pool 1 Excavations and the Yalbac Salvage Archaeology Program edited by L. J. Lucero, J. T. Larmon, and E. M. Benson, pp. 73 -85. Report submitted to the Institute of Archaeology National Institute of Culture and History, Belize. Larmon, Jean T. and Zach Nissen 2015 Exploratory Excavations at Pool 1: Structure 3 and the Plaza Test Pit. In Results of the 2014 Valley of Peace Archaeology Project: Underwater and Surface Exploration at Cara Blanca, edited by L. J. Lucero, pp. 60-75. Report submitted to the Institute of Archaeology, National Institute of Culture and History, Belize. Lucero, Lisa J. 2006 Water and Ritual: The Ris e and Fall of Classic Maya Rulers University of Texas Press, Austin. Lucero, Lisa J., and Andrew Kinkella 2015 Pilgrimage to the Edge of the Watery Underworld: An Ancient Maya Water Temple at Cara Blanca, Belize. Cambridge Archaeological Journal 25(1):16 3 185. Lucero, Lisa J., Jessica Harrison, Jean Larmon, Zachary Nissen, and Erin Benson 2016 Prolonged Droughts, Short -Term Responses and Diaspora: The Power of Water and Pilgrimage at the Sacred Cenotes of Cara Blanca, Belize. WIREs Water doi:10.1002/wat2.1148 Lucero, Lisa J., Jean T. Larmon, and Aime E. Carbaugh 2017 The Ancient Maya Ceremonial Circuit of Cara Blanca, Belize. Research Reports in Belizean Archaeology 14: 249-259. Medina -Elizalde, Martn, Stephen J. Burns, David W. Lea, Yemane Asmerom, Lucien von Gunten, Victor Polyak, Mathias Vuille, and Ambarish Karmalkar 2010 High Resolution Stalagmite Climate Record from the Yucatn Peninsula Spanning the Maya Terminal Classic Period. Earth and Planetary Science Letters 298: 255-262. Reese Taylor, Kathryn 2002 Ritual Circuits as Key Elements in Maya Civic Center Design. In Heart of Creation: The Mesoamerican World and the Legacy of Linda Schele, edited by Andrea Stone, pp. 143165. University of Alabama Press, Tuscaloosa.

PAGE 232

Ceremonial Circuit(s) at Cara Blanca 240 Vogt, Ev on Z. 1969 Zinacantan: A Maya Community in the Highlands of Chiapas The Belknap Press of Harvard University Press, Cambridge.

PAGE 233

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 241 251 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 22 PRECERAMIC CULTURAL HISTORY IN SOUTHERN BELIZE AND ITS ENVIRONMENTAL CONTEXT Keith M. Prufer This paper presents the environmental context for Early Holocene cultural developments in southern Belize and describes three archaeological sites that are producing evidence of human activities starting at the end of the last ice age and continuing until the advent of agriculture. It is well known that humans colonized Central America by at least 10,500 BC, and likely earlier (Chatters et al. 2014; Kennett et al. 2017). Central America formed a bottleneck for humans migrating from North to South America, and given its diverse geology, climate, and tropical resources it is not surprising that people successfully exploit ed this region throughout the Holocene We focus this discussion primarily on the context for early humans in southern Belize, but also draw broadly on well documented archaeological accounts f r om elsewhere in the region. Figure 1. Map showing the locations of the three pre ceramic sites located in southern Belize as well as better known Classic Period Maya centers. The central topographic feature in the region is the Maya Mountains, which are not as well -known as the more accessible foothills sites. Introduction Humans have been active agents in southern Belize throughout the Holocene. Their presence dates back to the initial colonization of the New World. The first Mesoamericans arrived in the reg ion 10 millennia prior to the development of urban populations, and some of them settled in southern Belize. The earliest known communities in southern Belize are Ek Xux located in the Maya Mountains, Uxbenk in the foothills ( Prufer et al 2017), and smal l coastal trading communities (McKillop 1996), with the earliest occupations in the Late Preclassic. Other centers did not develop until the end of the Early Classic period ca. AD 400 (Nimli Punit, Pusilh and Quebrada de Oro) or the Late Classic ca. AD 7 00 (Lubaantun, Xnaheb, Muklebal Tzul, and a host of smaller centers). By AD 750 there were at least 40 centers with public architecture in southern Belize, with largely independent rulers and significant populations dispersed across the agricul turally ric h hills and valleys. However, a growing body of research suggests that human presence in the region is much older and that the cultural adaptations in the Classic Period built on the strategies used by people throughout the Holocene. Below we provide back ground for the physical and climate context of the arrival of humans into the region, and preliminary results of work in several rockshelters in southern Belize that are expanding our knowledge of the first people to arrive in Central America. Geographi ca l Setting for Southern Belize Southern Belize ( Figure 1 ) is a geographically distinct region in Central America with a diverse set of geological and biotic resources that has facilitated a 13,000year history of human occupation. Physically, the region is circumscribed by the Maya Mountains to the west, a series of swampy bajos to the south along the Temash and Mojo rivers, the Caribbean Sea to the east and inhospitable pine barrens to the north. It is one of the wettest places in the Americas, receiving over 4000 mm of rainfall annually, more than double the precipitation of the Petn and seven times as much as the northern Yucatan Peninsula (Douglas et al. 2015) It is also a seasonal desert (Haug et al. 2003) where for several months each year there is littleto no rainfall and evapora tion exceeds precipitation.

PAGE 234

Preceramic Cultural History in Southern Belize 242 Geologically, the region is complex. The central topographic feature, the Maya Mountains, were formed by Devonian subaerial volcanic activity characterized by lava flo ws, pyroclastic activity and volcanoclastics, some locally altered hydrothermally, and by the Pennsylvanian Permian Santa Rosa Group of argillaceous and arenaceous sediments and carbonates. The eastern slope, bounding the Bladen River, is aproned by Terti ary and Cretaceous limestones of the Coban Formation (Petersen et al. 2012) Combined with high precipitation during the quaternary the result is a hydrologically carved network of caves and cockpit karst overlaying earlier volcanics that have been central to the lives of all people who have lived in these landscapes. The interior valleys of the Maya Mountains also have a complex geological history. The upper reaches of many tributaries have volcanic and metamorphic float, and soil pedogenesis in the all uvial valleys incorporates sedimentary as well as volcanic materials, making them a rich agricultural landscape, often surrounded by near vertical mountains hosting different biotic communities and productive potentials (Dunham and Prufer 1998) The foothi lls region, which was home to many of the Classic Period centers as well as most of the modern Maya speaking agricultural villages, has a different geological history. Known as the Toledo Formation (or the Toledo Uplands), these rolling hills are composed of Late Cretaceous Early Tertiary turbidite conglomerates with interbedded sandstones, mudstones, volcanics, and volcanoclastics, with sediments likely originating from the C oban volcanic arc migration (Cornec 1986) In some portions of the Toledo format ion, particularly near several major Classic Period centers (Uxbenk, Lubaantun, and Nim li Punit), hilltops are dominated by soft interbedded tertiary bedrock exposed through weathering and human mediated agricultural clearing. When cleared of vegetation as part of an agricultural cycle pedogenesis is rapid, with calcareous sandstone and mudstone breaking down rapidly (over a scale of weeks to months) as it is exposed to temperature and moisture differentials and rootlet activity (Culleton 2012) The res ult is an almost renewable source of high quality soils for farming and there is little need to engage in landscape intensification techniques like terracing to conserve soils (Prufer et al. 2015) Interspersed across this hilly landscape are massive karst ridges rising over 250 m above the Toledo Formation. These limestone remnants are late Tertiary early Cretaceous La Cumbre carbonate megabreccias (Cornec 1986) possibly formed during the collapse of the platform paleoscarp immediately following the KT boundary Chicxulub impact event (Bralower, Paull, and Leckie 1998) The coastline and pine forest to the north are quaternary in age and are composed of chert/quartz terraces as well as alluvial river terraces and sand bars. Pleistocene and Holocene kar stification of the CretaceousTertiary limestones have produced some of the key features used by humans as they colonized and modified these landscapes in southern Belize. These include the rockshelters occupied during the Paleoamerican and Archaic period s and the incredible subterranean cave landscape that formed a key component of the Mesoamerican worldview (Prufer and Brady 2005) Climate Setting for Southern Belize The climate context provides a critical framework for understanding past cultural adapt ations ( Figure 2). Annual rainfall in southern Maya lowlands is primarily controlled by the seasonal migration of the inter tropical convergence zone (ITCZ) with marked meridional contrast (Haug et al. 2001) whereby southern Belize receives considerable rainfall each year, often in excess of 4,000 mm (Ridley et al. 2015) Mean annual temperature is approximately 26 degrees C. During the winter dry season (February May) evaporation frequently exceeds precipitation. Given its location relative to the equator, at the northern margin of the annual ITCZ migration, southern Belize is sensitive to even small variations in the mean position of the ITCZ and its rainfall distribution (Lechleitner et al. 2017; Ridley et al. 2015) Other climate modulators that play significant roles in the precipitation variability of the region include changes in the strength of the North Atlantic high and variability in El Nino Southern Oscillation (ENSO). High sea level

PAGE 235

Prufer 243 Figure 2 Two records of climate variability during the early colonization of Central America. The Cariaco basin titanium density record records rainfall sediment runoff into an anoxic closed basic off the coast of Venezuela (Haug et al. 2001) as a proxy for the northsouth migration of the ITCZ. The Juxtalahuaca oxygen isotope record is a paleorainfall proxy from a speleothem in central Mexico (Lachniet et al. 2013). pressures (SLP) in the North Atlantic High lead to stronger trade winds. This results in cooler than normal sea surface temperatures (SST) and reduced Caribbean basin precipitation that has decadal scale variability through the North Atlantic Oscillation (NAO, (Proctor et al. 2000). This variabi lity is clearly exhibited in our records covering the past 2000 years (Lechleitner et al. 2017; Smirnov et al. 2017) At shorter timescales, ENSO exerts strong inter annual precipitation variability in the Central American tropics, establishing a zonal seesaw SLP and SST pattern across the eastern Pacific and western Atlantic region. The result is that during ENSO+ (lower SLP and higher SST) periods it is usually dryer and warmer along the Central American coastline during the rainy season (Zhu et al. 2012) resulting in drought conditions on severe but short time scales. Paleoclimate data strongly suggest that climate conditions are significantly different today than when the first humans arrived in the region. The Cariaco shallow marine record off the c oast of Venezuela (Peterson et al. 2000; Haug et al. 2001) provides a proxy for changes in the position of the ITCZ. The Cariaco reflectance and Ti concentration data suggest that the climate context for the first human movements into the neotropics was d uring a period that was dryer (Haug et al. 2001) and cooler (Grauel et al. 2016) than conditions during the Holocene. Shallow lake records from Petn (Escobar et al. 2012) also show a dry Late Pleistocene to Younger Dryas interval. This is supported by numerous studies in lower Central America and tropical South America (Piperno 2011a; Piperno and Jones 2003) Two rainfall reconstruction records provide a general paleoclimate backdrop for the Late Pleistocene, and there is one continuous record through th e Holocene, though it cannot be considered a local record but r ather a more general indication of low latitude rainfall patterns in the sub tropical Americas (Figure 2). The Cariaco Basin Ti (Haug et al. 2001) concentration data reflect hydrologic chang es in the Orinoco River drainage basin of the northern coast of South America. These result from shifts in the position of the ITCZ driven by insolation variability. The other common feature is a shift towards wetter conditions during the early Holocene and then a trend toward drier conditions, especially the Cariaco Ti record, later in the Holocene related to insolation

PAGE 236

Preceramic Cultural History in Southern Belize 244 changes in the strength of the regional monsoon. This record is relevant for understanding changes in rainfall distributions in the lat e Pleistocene through late Holocene and is the longest continuously resolved ITCZ record for the New World. The Juxtalahuaca 18O speleothem record (Lachniet et al. 2013) from Central Mexico is a bit closer to southern Belize and is one of the few records of the North American Monsoon covering parts of the late Pleistocene and Younger Dryas. Although it is discontinuous, it shows general agreement with the Cariaco ITCZ reconstruction as being drier prior to 10,000 BC, reflecting broad hemispheric trends o f low latitude paleoclimate. Mesoamerican Foragers before agriculture The initial New World colonists that arrived in Central America by at least 12,500 BC (Braje et al. 2017) encountered a very different, and far less tropical, environment than today. At that time the landscape was comprised of heterogeneous, even patchy, vegetation across small distance scales; and stretches of forest alongside water courses in regions where forests were significantly reduced (Piperno 2006:286). Pollen and macrofossil plant data suggest the structure of forests may have already been tropical, but the distribution of these was less than in the modern climate regime and vegetation was m ore diverse than simple Pleistocene grassland/Holocene forest dichotomies would suggest (Piperno 2011a) Confronted with a greater diversity of large mammals and a wider range of riparian forest and grasslands humans would have initially adapted to ecosys tems that were far different than today. By 9,000 BC conditions were becoming wetter and warmer and, in the Petn, there is evidence that closed canopy forests were undergoing anthropogenic burning (Renssen et al. 2009; Anderson and Wahl 2015) with mixed herbaceous and woody plants being represented in charcoal records. Pre agricultural burning peaks between 6,000 and 4,000 BC (Schpbach et al. 2015) during the Holocene Thermal Maximum, arguably the warmest and wettest period of the Holocene (Renssen et al. 2009), and likely reflects increased anthropogenic burning. After 8,500 BC the abundance of higher ranked plant and animal resources declined as rainforest overtook many Pleistocene open areas where game would have fed on scrub and grasses (Piperno and Pearsall 1998). The Paleoamerican Period (> 13,500 BC 7,000 BC) is the least understood period in the Maya region. Initial colonists into the New World arrived in North America prior to 13,500 BC, spreading rapidly along the Pacific coast, and reaching southern Chile by 13,000 years ago (Braje et al. 2017; Dillehay et al. 2017) This rapid southward migration was accompanied by significant eastward movements in North America, evidence of which has recently emerged in Florida and Montana (Halligan et al 2016; Rasmussen et al. 2014) The now well documented early colonization of tropical South America, perhaps as early as 12,500 BC (Brandini et al. 2017; Surez 2017) necessitated the movement of people through middle and lower Central America. Human pre sence has been well documented in Panama (Ranere and Cooke 1991) and Costa Rica (Swauger and Mayer Oakes 1952; Snarskis 1979) Nicaragua (Waters 1985) Honduras (Kennett et al. 2017; Scheffler et al. 2012a) and Highland Guatemala (Brown 1980; Gruhn et al. 1977) In Mesoamerica, drier conditions to the north of the tropical Maya lowlands have facilitated the identification of Paleoamerican surface sites, including locales in central, west, and north Mexico (Ochoa 2012; Gonzalez et al. 2015; Sanchez and Carp enter 2012) In the Maya Lowlands, with high precipitation and extensive tropical foliage, fewer surface sites have been identified and almost no stratified sites are known. One exception is El Gigante rockshelter (Kennett et al. 2017; Scheffler et al. 2012a) a large rockshelter in western Honduras on the periphery of the Maya Lowlands. There, stratified deposits document occupation from 7,0009,000 cal BC, and include well preserved macrobotanical remains as well as evidence of hunting and food prepara tion. In the northern Yucatan peninsula, a near complete human skeleton was found with extinct fauna in a submerged cave (Chatters et al. 2014) The minimum age of this skeletal material is 10,000 BC based on U series dates of small calcite florets that had precipitated on bone before the skeleton was submerged by rising sea and ground water levels. Those dates are supported

PAGE 237

Prufer 245 by an abundance of Pleistocene faunal remains also found in the submerged chamber. The Archaic Period is better documented, particu larly outside of the Maya area in central and western Mexico, where studies have examined the origins of agriculture, diet changes in coastal settlements, and the emergence of social complexity (Flannery 2002; Kennett et al. 2010; Lesure 2011; MacNeish and Nelken Terner 1983; Rosenswig 2014; Rosenswig et al. 2015; Smith 1997; Voorhies et al. 2002). In the arid regions of northern Mexico, where ground visibility and site detection is not hampered by dense tropical vegetation, there is a long history of rese arch into Archaic Period adaptations (Guadalupe and Carpenter 2012). In the tropical Maya region however, far less is known about tropical adaptations during this time (Kennett et al. 2010). Recent studies suggest a gradual adoption of domesticated plant s by 4,000 BC (Rosenswig et al. 2014), although in the Soconusco full scale maize agriculture may not have been adopted before 1,000 BC (Rosenswig et al. 2015) even though sedentary agricultural communities are present by 1,500 BC. Between 8,000 3,000 B C the gradual processes of plant domestication were underway in Central America with evidence for human cultivation of native crops including maize ( Zea ), manioc ( Manihot ), arrowroot ( Maranta), and yams ( Dioscorea ) in parallel with the exploitation of wild resources by transitional hunter gatherers (Greaves and Kramer 2014; Piperno 2011b). Recent studies have also emphasized the importance of the transition to the Archaic as a time of mixed and flexible subsistenc e economies, as evidenced by a broad range of wild plant foods and early domesticates at El Gigante rockshelter in Honduras (Scheffler et a l. 2012b; Kennett et al. 2017). As noted by Rosenswig (2014:142) the spotty nature of Archaic age archaeological dat a from Mesoamerica means that what we know is likely not representative of the range of peoples and the variety of adaptations that existed across the region particularly the broadleaf forests of the tropical Maya lowlands, and especially for the Early Ar chaic (8,0005,000 BC). A wide range of cultural changes occurred during the Archaic, with a general trend towards increased reliance on plants as a source of food and changing environmental conditions that may have favored plant tending and agriculture. Thus, social changes were likely mediated by subsistence changes, and these were driven by demographic pressure, environmental change, and socioeconomic competition (Winterhalder and Kennett 2006). The period from 8,000 to 6,000 BC is generally considere d to have been wetter and warmer, prior to a drier interval lasting until 3,500 BC (Mueller et al. 2009). This suggests that the transition to agriculture spanned several phases of significant climate and environmental change with the relationships b etween them poorly constrained. Preceramic Forag ers in Southern Belize Three sites largely define what we know about the preceramic foragers in southern Belize. All three are rockshelters. All three are located near to permanent sources of water, and all three have evidence of food processing, tool making and tool use, and mortuary activities (Figure 1). Here, work at the se shelters is summarized and in future volumes of the RRBA we will present more detailed accounts of the data analysis and interpretati on. Mayahak Cab Pek and Saki Tzul are two large rockshelters located in an interior valley of the Maya Mountains in the Bladen Nature Reserve (BNR), a protected wilderness area where there has been minimal human disturbance of archaeological sites. These rockshelters are located over 30km from the nearest modern settlements. Their remoteness along with the protected status of the BNR (access to the Nature Reserve is limited to only scientific researchers with required permits) has greatly facilitated preservatio n of the archaeological record. Both rockshelters were first investigated in 1998 by the Maya Mountains Archaeological Project (MMAP). At that time, shallow excavations at both sites produced shallow burials ( Figure 3), abundant faunal remains, evi dence of stone tool production, and were thought to be local mortuary sites used by residents of two nearby Classic Period sites, Ek Xux and Muklebal Tzul (Saul et al. 2005; Dunham and Prufer 1998) The 1998 excavations at both rockshelters were

PAGE 238

Preceramic Cultural History in Southern Belize 246 Figure 3 Photo of Prufer and archaeologist David Goldstein in excavation Unit 34/35 in 1998. This Protocolassic burial was generally the deepest level reached prior to the start of the BPAAP. constrained by a short field season and were limited to the top 100 cm of sediments and did not penetrate below ceramic levels (Prufer 2002) From 2014 to 2017 the Bladen Paleoindian and Archaic Archaeological Project has revisited both sites and conducted more exten sive excavation. Both rockshelters have deep deposits of cultural materials dating to the period prior to 10,000 BC ( Figure 4) and continuing through the Classic Maya collapse at ca. AD 800 1000. Though the two rockshelters are located 1.4km apart, both have similar stratigraphic sequences and contain similar assemblages of artifacts and biological remains dating to the late Pleistocene. Both rockshelters have dry sediments and large overhangs, reflecting that little if any direct rainfall affects their contexts. Sediment formation differs significantly from surface sites in that water transport is not a major contributing factor. Sediments are deposited as a result of microbial and physical breakdown of the limestone cliff, wind blown pollens and plant materials, animal activity, and anthropogenic activities, with the Figure 4 View of the massive cliff wall of Saki Tzul rockshelter. The site is over 1200 m2. The overhanging wall provides significant shelter from the elements. bulk of the Holoce ne deposits likely resulting from the latter. The dry aeolian nature of the rockshelters also helps to explain the excellent preservation of unburned bone other organic materials and the very minor presence of root activity in a tropical environment (Fig ure 5 ) Mayhak Cab Pek is an east facing shelter with approximately 160 m2 sheltered, while Saki Tzul has a south aspect and has approximately 890 m2 sheltered. Both have approximately 3.5 m of deposits that primarily consist of midden fill replete with faunal bone, carbonized plant

PAGE 239

Prufer 247 Figure 5 Profiles drawings from both Maya Hak Cab Pek and Saki Tzul showing similar stratigraphic sequences and across all times period. materials, chipped stone, ground stone, terrestrial snail shells, and many millions of small spirelopped jute (Pachychilus sp.) riverine snail shells, all of which appear to have been consumed as a food. Stratigraphy at both rockshelters is relatively undisturbed and consists of Classic Period Maya deposits overlaying dense jute midden fill dating to the middle and late Archaic (5k2.5k BC). Below this are early Archaic and Paleoindian contexts with a gradual decline in jute shell but abundant faunal, human, paleobotanical, and lithic remains. Human burials have a wide range of orientations, positions, and degrees of completeness, but are mostly primary interments that w ere placed in very shallow depressions and covered with either sedim ents or small river cobbles. Burials are found in all time periods and consist of males and females and range from neonates to older adults. Preliminary analysis of faunal remains suggests only small shifts in species composition over the Holocene, but a likely change in selection or preference from larger mammals in the early part of the record to smaller mammals in the later part (Orsini 2016). To date, only 8.4 m2 at MHCP and 10.6 m2 at ST have been excavated. The excavations at Mayahak Cab Pek and Sa ki Tzul complements previous work we conducted in the Rio Blanco Valley, near to the Classic Period center Uxbenk. There, excavations in the small rockshelter Tzib te Yux from 20122015 documented an Archaic to Paleoindian chronology and cultural materia ls dating to before 10,500 BC (Prufer et al. 2017) Combined with geomorphological testing we have evidence of an occupation lasting well over 10,000 years in that small interior valley Unfortunately, the record from Tzub Te Yux is truncated and the top 50006500 years were removed, likely as an effort mining the jute shell middens by residents of the Classic Period center. Combined, these three records suggest strongly that the signal of pre ceramic human activity is well preserved in some rockshelter deposits that are not affected by fluvial or erosion. It also indicates that there can be excellent preservation of some organic materials, primarily bone and carbonized plant materials. With ongoing analytical research, we will present more detailed inte rpretations of these shelter s in future RRBA publications. Acknowledgements We thank the Belize Institute of Archaeology for their continuing support of our research into the earliest Belizeans and their life histories. Funding for this project comes from the Alphawood

PAGE 240

Preceramic Cultural History in Southern Belize 248 Foundation and the National Science Foundation. Our work was a lso conducted with permits from the Belize Forest Department. We express special gratitude to our collaborators with the Yaaxche Conservation Trust. References Anderson, Lysanna, and David Wahl 2015 Two Holocene paleofire records from Peten, Guatemala: Implications for natural fire regime and prehispanic Maya land use. Global and Planetary Change 138:82 92. Braje, Todd J., Tom D. Dillehay, Jon M. Erlandson, Richard G. Klein, and Torben C. Rick 2017 Finding the first Americans. Science 358(6363):592 594. Bralower, Timothy J., Charles K. Paull, and R. Mark Leckie 1998 The Cretaceous -Tertiary boundary cocktail: Chicxulub impact triggers margin collapse and extensive sediment gravity flows. Geology 26(4):331 334. Brandini, Stefania, Paola Bergamaschi, Mar co Fernando Cerna, Francesca Gandini, Francesca Bastaroli, Emilie Bertolini, Cristina Cereda, Luca Ferretti, Alberto Gmez Carballa, Vincenza Battaglia, Antonio Salas, Ornella Semino, Alessandro Achilli, Anna Olivieri, and Antonio Torroni 2017 The Paleo -In dian Entry into South America According to Mitogenomes. Molecular Biology and Evolution Brown, Kenneth L. 1980 A Brief Report on Paleoindian-Archaic Occupation in the Quiche Basin, Guatemala. American Antiquity 45(02):313 324. Chatters, James C., Douglas J. Kennett, Yemane Asmerom, Brian M. Kemp, Victor Polyak, Alberto Nava Blank, Patricia A. Beddows, Eduard Reinhardt, Joaquin ArroyoCabrales, Deborah A. Bolnick, Ripan S. Malhi, Brendan J. Culleton, Pilar Luna Erreguerena, Domin ique Rissolo, Shanti Morell -Hart, and Thomas W. Stafford 2014 Late Pleistocene Human Skeleton and mtDNA Link Paleoamericans and Modern Native Americans. Science 344(6185):750 754. Cornec, JH 1986 Provisional Geological Map of Belize: Belmopan. Belize Min istry of Petroleum, Belmopan. Culleton, Brendan 2012 Human Ecology, Agricultural Intensification and Landscape Transformation at the Ancient Maya Polity of Uxbenk Southern Belize. Unpublished Dissertation, University of Oregon, Eugene. Dillehay, Tom D. Steve Goodbred, Mario Pino, Vctor F. Vsquez Snchez, Teresa Rosales Tham, James Adovasio, Michael B. Collins, Patricia J. Netherly, Christine A. Hastorf, Katherine L. Chiou, Dolores Piperno, Isabel Rey, and Nancy Velchoff 2017 Simple technologies and diverse food strategies of the Late Pleistocene and Early Holocene at Huaca Prieta, Coastal Peru. Science Advances 3(5):e1602778. Douglas, Peter M. J., Mark Pagani, Marcello A. Canuto, Mark Brenner, David A. Hodell, Timothy I. Eglinton, and Jason H. Curti s 2015 Drought, agricultural adaptation, and sociopolitical collapse in the Maya Lowlands. Proceedings of the National Academy of Sciences 112(18):5607 5612. Dunham, Peter S, and Keith M Prufer 1998 En la cumbre del Clsico: descubrimientos recientes en la Montana Maya en el sur de Belice. In In XI Simposio de Investigationes Arqueolgicas en Guatemala. (Juan Pedro Laporte and Hctor L. Escobedo, editors) pp. 165 170. Ministerio de Cultura y Deportes, Instituto de Antropologa e Historia y Asociacin Tikal, Ciudad de Guatemala. Escobar, Jaime, David A. Hodell, Mark Brenner, Jason H. Curtis, Adrian Gilli, Andreas D. Mueller, Flavio S. Anselmetti, Daniel Ariztegui, Dustin A. Grzesik, Liseth Prez, Antje Schwalb, and Thomas P. Guilderson 2012 A 43-ka reco rd of paleoenvironmental change in the Central American lowlands inferred from stable isotopes of lacustrine ostracods. Quaternary Science Reviews 37:92 104. Flannery, Kent V. 2002 The Origins of the Village Revisited: From Nuclear to Extended Households. American Antiquity 67(3):417 433. Gonzalez, Silvia, David Huddart, Isabel Israde -Alcntara, Gabriela Domnguez-Vzquez, James Bischoff, and Nicholas Felstead 2015 Paleoindian site s from the Basin of Mexico: Evidence from stratigraphy, tephrochronology and dating. Quaternary International 363:4 19. Grauel, Anna -Lena, David A. Hodell, and Stefano M. Bernasconi 2016 Quantitative estimates of tropical temperature change in lowland Ce ntral America during the last 42 ka. Earth and Planetary Science Letters 438:37 46. Greaves, Russell D., and Karen L. Kramer 2014 Hunter gatherer use of wild plants and domesticates: archaeological implications for mixed

PAGE 241

Prufer 249 economies before agricultural int ensification. Journal of Archaeological Science 41:263 271. Gruhn, Ruth, Alan Lyle Bryan, and Jack D Nance 1977 Los Tapiales: a Paleo -Indian campsite in the Guatemalan highlands. Proceedings of the American Philosophical Society 121(3):235 273. Guadalupe Sanchez, and John Carpenter 2012 Paleoindian and Archaic Traditions in Sona, Mexico. In In From the Pleistocene to the Holocene: Human Organization and Cultural Transformations in Prehistoric North America edited by C. Britt Bousman and Bradley Jay Vier ra, pp. 125 147. Texas A&M University Press, College Station, TX, October 22. Halligan, J. J., M. R. Waters, A. Perrotti, I. J. Owens, J. M. Feinberg, M. D. Bourne, B. Fenerty, B. Winsborough, D. Carlson, D. C. Fisher, T. W. Stafford, and J. S. Dunbar 201 6 Pre -Clovis occupation 14,550 years ago at the Page -Ladson site, Florida, and the peopling of the Americas. Science Advances 2(5):375. Haug, Gerald H., Konrad A. Hughen, Daniel M. Sigman, Larry C. Peterson, and Ursula Rhl 2001 Southward Migration of the Intertropical Convergence Zone Through the Holocene. Science 293(5533):1 304 1308. Haug, GH, D Gnther, LC Peterson, DM Sigman, KA Hughen, and B Aeschlimann 2003 Climate and the Maya. PAGES News 11(2):3. Kennett, Douglas J., Dolores R. Piperno John G. Jones, Hector Neff, Barbara Voorhies, Megan K. Walsh, and Brendan J. Culleton 2010 Pre pottery farmers on the Pacific coast of southern Mexico. Journal of Archaeological Science 37(12):3401 3411. Kennett, Douglas J., Heather B. Thakar, Amber M. VanDerwarker, David L. Webster, Brendan J. Culleton, Thomas K. Harper, Logan Kistler, Timothy E. Scheffler, and Kenneth Hirth 2017 High -precision chronology for Central American maize diversification from El Gigante rockshelter, Honduras. Proceedings of the National Academy of Sciences 114(34):9026 9031. Lachniet, Matthew S., Yemane Asmerom, Juan Pablo Bernal, Victor J. Polyak, and Lorenzo Vazquez-Selem 2013 Orbital pacing and ocean circulation -induced collapses of the Mesoamerican monsoon over the past 22,000 y. Proceedings of the National Academy of Sciences 110(23):9255 9260. Lechleitner, Franziska A., Sebastian F. M. Breitenbach, Kira Rehfeld, Harriet E. Ridley, Yemane Asmerom, Keith M. Prufer, Norbert Marwan, Bedartha Goswami, Douglas J. Kennett, Valo rie V. Aquino, Victor Polyak, Gerald H. Haug, Timothy I. Eglinton, and James U. L. Baldini 2017 Tropical rainfall over the last two millennia: evidence for a low -latitude hydrologic seesaw. Scientific Reports 7:45809. Lesure, Richard G. 2011 Early Mesoamerican social transformations: archaic and formative lifeways in the Soconusco region. Univ ersity of California Press. MacNeish, Richard S., and Antoinette Nelken -Terner 1983 The Preceramic of Mesoamerica. Journal of Field Archaeology 10(1):71 84. Mueller, Andreas D., Gerald A. Islebe, Michael B. Hillesheim, Dustin A. Grzesik, Flavio S. Anselmetti, Daniel Ariztegui, Mark Brenner, Jason H. Curtis, David A. Hodell, and Kathryn A. Venz 2009 Climate drying and associated forest decline in the lowlands of northern Guatemala during the late Holocene. Quaternary Research 71(2):133 141. Ochoa, Guillermo Acosta 2012 Ice Age Hunter Gatherers and the Colonization of Mesoamerica. The Oxford Handbook of Mesoamerican Archaeology :129. Petersen, H.I., B. Holland, H.P. Nytoft, A. Cho, S. Piasecki, J. de la Cruz, and J.H. Cornec 2012 Geochemistry of crude oils, seepage oils and source rocks from Belize and Guatemala: indications of carbonate -sourced petroleum systems. Journal of Petroleum Geology 35(2):127 163. Peterson, Larry C., Gerald H. Haug, Konrad A. Hughen, and Ursula Rhl 2000 Rapid Changes in the Hydrologic Cycle of the Tropical Atlantic During the Last Glacial. Science 290(5498):1947 1951. Piperno, Dolores 2006 Quaternary Enviro nmental History and Agricultural Impact on Vegetation In Central America. Annals of the Missouri Botanical Garden 93(2):274 296. 2011a Prehistoric human occupation and impacts on Neotropical forest landscapes during the Late Pleistocene and Early/Middle H olocene. In Tropical Rainforest Responses to Climatic Change edited by Mark Bush, John Flenley, and William Gosling, pp. 185 212. Springer Berlin Heidelberg, Berlin, Heidelberg. 2011b The Origins of Plant Cultivation and Domestication in the New World Tr opics: Patterns, Process, and New Developments. Current Anthropology 52(S4):S453 S470. Piperno, Dolores, and Deborah M. Pearsall 1998 The origins of agriculture in the lowland neotropics Academic Press, San Diego.

PAGE 242

Preceramic Cultural History in Southern Belize 250 Piperno, Dolores R., and John G. Jones 2 003 Paleoecological and archaeological implications of a late Pleistocene/Early holocene record of vegetation and climate from the pacific coastal plain of panama. Quaternary Research 59(1):79 87. Proctor, C. J., A. Baker, W. L. Barnes, and M. A. Gilmour 2000 A thousand year speleothem proxy record of North Atlantic climate from Scotland. Climate Dynamics 16(10 11):815 820. Prufer, Keith M., Clayton R Meredith, Asia Alsgaard, Timothy Dennehy, and Douglas J Kennett 2017 The Paleoindian chronology of Tzib T e Yux rockshelter in the Rio Blanco valley of southern Belize. Research Reports in Belizean Archaeology 14:309 314. Prufer, Keith M., Amy E. Thompson, and Douglas J. Kennett 2015 Evaluating airborne LiDAR for detecting settlements and modified landscapes in disturbed tropical environments at Uxbenk, Belize. Journal of Archaeological Science 57:1 13. Prufer, Keith Malcolm 2002 study of sacred space in the Maya Mountains of southern Belize. PhD Dissertation, Department of Anthropology Southern Illinois University. Prufer, Keith Malcolm, and James Edward Brady 2005 Stone houses and earth Lords: Maya religion in the cave context University Press of Colorado, Boulder. Ranere, Anthony J., and Richard G. Cooke 1991 Paleoindian occupation in the Central American tropics. Clovis: Origins and Adaptations, Center for the Study of the First Americans, Corvallis, OR :237 253. Rasmussen, Morten, Sarah L. Anzick, Michael R. Waters, Pontus Skoglund, Michael DeGiorgio, Thomas W. Stafford, Simon Rasmussen, Ida Moltke, Anders Albrechtsen, Shane M. Doyle, G. David Poznik, Valborg Gudmundsdottir, Rachita Yadav, Anna -Sapfo Malaspinas, Samuel Stockton White V, Morten E. Allentoft, Omar E. Cornejo, Kristiina Tambets, Anders Eriks son, Peter D. Heintzman, Monika Karmin, Thorfinn Sand Korneliussen, David J. Meltzer, Tracey L. Pierre, Jesper Stenderup, Lauri Saag, Vera M. Warmuth, Margarida C. Lopes, Ripan S. Malhi, Sren Brunak, Thomas Sicheritz -Ponten, Ian Barnes, Matthew Collins, L udovic Orlando, Francois Balloux, Andrea Manica, Ramneek Gupta, Mait Metspalu, Carlos D. Bustamante, Mattias Jakobsson, Rasmus Nielsen, and Eske Willerslev 2014 The genome of a Late Pleistocene human from a Clovis burial site in western Montana. Nature 506 (7487):225 229. Renssen, H., H. Sepp, O. Heiri, D. M. Roche, H. Goosse, and T. Fichefet 2009 The spatial and temporal complexity of the Holocene thermal maximum. Nature Geoscience 2(6):411 414. Ridley, Harriet E., Yemane Asmerom, James U. L. Baldini Sebastian F. M. Breitenbach, Valorie V. Aquino, Keith M. Prufer, Brendan J. Culleton, Victor Polyak, Franziska A. Lechleitner, Douglas J. Kennett, Minghua Zhang, Norbert Marwan, Colin G. Macpherson, Lisa M. Baldini, Tingyin Xiao, Joanne L. Peterkin, Jaim e Awe, and Gerald H. Haug 2015 Aerosol forcing of the position of the intertropical convergence zone since ad 1550. Nature Geoscience 8(3):195 200. Rosenswig, Robert M. 2014 A Mosaic of Adaptation: The Archaeological Record for Mesoamericas Archaic Perio d. Journal of Archaeological Research 23(2):115 162. Rosenswig, Robert M., Deborah M. Pearsall, Marilyn A. Masson, Brendan J. Culleton, and Douglas J. Kennett 2014 Archaic period settlement and subsistence in the Maya lowlands: new starch grain and lithic data from Freshwater Creek, Belize. Journal of Archaeological Science 41:308 321. Rosenswig, Robert M., Amber M. VanDerwarker, Brendan J. Culleton, and Douglas J. Kennett 2015 Is it agriculture yet? Intensified maize -use at 1000 cal BC in the Soconusco and Mesoamerica. Journal of Anthropological Archaeology 40:89 108. Sanchez, Guadalupe, and John Carpenter 2012 Paleoindian and archaic traditions in Sonora, Mexico. From the Pleistocene to the Holocene: Human Organization and Cultural Transformations in Prehistoric North America 17:125. Saul, Julier Mather, Keith M Prufer, and Frank P Saul 2005 Nearer to the gods. Rock shelter burials from the Ek Xux Valley, Belize. In Stone houses and Earth Lords: Maya R eligio n in the Cave C ontext pp. 297 323. Universi ty of Colorado Press, Boulder. Scheffler, Timothy E., Kenneth G. Hirth, and George Hasemann 2012a The El Gigante Rockshelter: Preliminary Observations on an Early to Late Holocene Occupation in Southern Honduras. Latin American Antiquity 23(04):597 610. Scheffler, Timothy, Kenneth Hirth, and George Hasemann 2012b The El Gigante Rockshelter: Preliminary Observations on an Early to Late Holocene Occupation in Southern Honduras. Latin American Antiquity 23(4):597 610.

PAGE 243

Prufer 251 Schpbach, Simon, Torben Kirchgeorg, Da niele Colombaroli, Giorgia Beffa, Marta Radaelli, Natalie M. Kehrwald, and Carlo Barbante 2015 Combining charcoal sediment and molecular markers to infer a Holocene fire history in the Maya Lowlands of Petn, Guatemala. Quaternary Science Reviews 115:123 1 31. Smirnov, DA, SFM Breitenbach, G Feulner, FA Lechleitner, KM Prufer, JUL Baldini, N Marwan, and J Kurths 2017 A regime shift in the Sun -Climate connection with the end of the Medieval Climate Anomaly. Scientific Reports 7(1):11131. Smith, Bruce D. 1997 The Initial Domestication of Cucurbita pepo in the Americas 10,000 Years Ago. Science 276(5314):932 934. Snarskis, Michael J. 1979 Turrialba: A Paleo -Indian Quarry and Workshop Site in Eastern Costa Rica. American Antiquity 44(01):125 138. Surez, Rafael 2017 The human colonization of the Southeast Plains of South America: Climatic conditions, technological innovations and the peopling of Uruguay and south of Brazil. Quaternary International 431:181 193. Swauger, James L., and William J. Mayer Oakes 1952 A Fluted Point from Costa Rica. American Antiquity 17(03):264 265. Voorhies, Barbar, Douglas J. Kennett, John G. Jones, and Thomas A. Wake 2002 A Middle Archaic Archaeological Site on the West Coast of Mexico. Latin American Antiquity 13(2):179 200 Waters, Michael R 1985 Early Man in the New World: an evaluation of the radiocarbon dated pre -Clovis sites in the Americas. Environments and extinctions: man in Late Glacial North America. Orono: Center for the Study of Early Man. p:125 44. Winterhalde r, Bruce, and Douglas J. Kennett 2006 Behavioral ecology and the transition from hunting and gathering to agriculture. In Behavioral Ecology and the Transition to Agriculture edited by Douglas J. Kennett and Bruce Winterhalder, pp. 1 21. University of California Press, January 1. Zhu, Xiaojie, R. Saravanan, and Ping Chang 2012 Influence of Mean Flow on the ENSO Vertical Wind Shear Relationship over the Northern Tropical Atlantic. Journal of Climate 25(3):858 864.

PAGE 244

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 253 263 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 23 PLASTERED: CAVE CONS TRUCTIONS AT LAS CUE VAS Erin E. Ray, Holley Moyes, and Linda Howie The ancient Maya site of Las Cuevas, in Western Belize features a cave system that runs beneath the main plaza. Investigations by the Las Cuevas Archaeological Reconnaissance project suggest that the site functioned as a Late Classic ritual pilgrimage venue and that the cave was used for large public centrally organized performances. The cathedral -like cave entrance contains monumental architecture consisting of at least 76 plastered platforms. We expect that the level of managerial oversight should be correlated with the consistency of building materials employed in their construction. Plasters from both cave and surface contexts were analyzed using geochemica l methods including XRF (pXRF), XRD, SEM -EDS and FTIR to examine their chemical composition. Results demonstrate considerable variation in plaster recipes in the cave and on the surface. Introduction Las Cuevas, a small to midsized site, is located in the Chiquibul Forest Reserve in the Cayo District of Western Belize. It lies north of the Maya Mountains and west of the Vaca Plateau 14 km southeast of the larger Caracol polity ( Figure 1). The site was first investigated in 1957 by then commissioner of the Department of Archaeology A.H. Anderson and Adrian Digby from the British Museum (Digby 1958)) and more recently by the L as Cuevas Archaeological Reconnaissance project (LCAR) under the direction of Dr. Holley Moyes. Las Cuevas consists of 26 structures that include temples, a palace, long linear constructions, and a ballcourt. The structures are organized on an east/west axis facing two plazas A and B. The layout encircles a dry sinkhole leading to the entrance of an extensive cave system ( Figure 2 ). Based on five field seasons of chronology building through excavation and test pitting (Kosakowsky et al. 2013; Moyes et al 2012; Moyes et al. 2015), 25 AMS dates suggest that the surface constructions occurred in the later part of the Late Classic period between AD 640985 at the 2sigma range. Ceramic crossdating places all building phases at the site into the later part of the Late Classic period (Tepeu 2/ Spanish Lookout 2), which agrees well with the radiocarbon dates. Temple I, the eastern structure of Plaza A was constructed in 4 phases, but all appear to date to this time period based on their ceramic chronology (Kosakowsky 2012, 2013, Kosakowsky et al. 2013). A small ceramic deposit located below the platform upon which the ballcourt sits suggests that people were present in the area as early as the Late Preclassic Figure 1 Map of Belize including location of Las Cuevas. (250 BC AD 250), though no formal constructions or evidence of cave use date to this earlier period Despite its proximity to the colossal site of Caracol (14 km to the southeast), Las Cuevas bears little resemblance to its neighbo ur and no evidence uncovered thus far demonstrates any formal relationships between the two (Kosakowsky 2013; Moyes et al. 2015). Rather, Kosakowsky argues that the ceramics foun d at the site represent a number of ceramic spheres with styles from the Peten, Belize Valley, and Southern Belize. As Moyes and her colleagues have argued elsewhere (Moyes 2012, 2015,

PAGE 245

Plastered: Cave Constructions at Las Cuevas 254 Figure 2 Map of the site core and its relation to the sinkhole and cave. Kosakowsky et al. 2013), Las Cuevas likely functioned as a Late Classic pilgrimage site. It is likely that the cave was first utilized in the early part of the Late Classic period, prior to the construction of the surface components of the site (Moyes et al. 2017). The cave system at Las Cuevas consists of ten chambers (Moyes 2011, 2012, 2015). The Entrance Chamber is massive, measuring approximately 105 meters in length and 40 meters in width. At the rear of the Entrance Chamber a constructed w all with a small doorway that occludes the opening to the tunnel system ( Figure 3 ). The 315 m tunnel system winds through a series of small constructed walls and blockages culminating at Chamber 8, which terminates at a small window overlooking the entrance chamber approximately eight meters above the ground su rface of the cave (Moyes 2012). In the center of the Entrance Chamber is a cenote lined with a cut stone block retaining wall that descends to an underground river, which surfaces at its base. The water level rises and falls with heavy rains. The chamber is heavily modified with monumental architectural constructions including terraces, retaining walls, stairs, and platforms that are topped with layers of thick plaster. The project mapped and recorded 76 plastered platforms, seven staircases and three sets of terraces thus far, though is likely that additional platforms are located beneath the chambers muddy floor in the northern areas. Terraces are at the cave entrance and within the cenote itself. Though they are in poor condition, some plaster was noted on terraces near the southwest corner of the cave mouth (Ray 2018) and thick plaster still covers the stairways as well. Platforms are arranged in clusters that are interconnected via sta ircases to facilitate navigation between individual platforms and platform complexes. Some complexes consisted of several tiers, reminiscent of a wedding cake, with the largest tier at the bottom and one to two smaller platforms on top usually built again st the cave wall. It is currently unclear if these were built in a single phase, though replastering occurred on some platforms as evidenced by visible separations between burned and stained floors such as on Platform 42 ( Figure 4 ). Preliminary radiocarb on dates from the platforms have dated constructions between AD 670 AD 950. These largescale constructions suggest that the cave was used for well organized ceremonies that could be viewed by many observers supporting a large number of participants. Ba sed on their spatial areas, 200500 people could have been easily accommodated either sitting or standing on the platforms (Moyes et al. 2015:243). Based on the size and scale of the constructions, Moyes and her colleagues argued that this massive perform ance space was created and maintained for the benefit of elites, subelites, and possibly their ret inues (Moyes et al. 2015: 246). Structure 1 is an 11 m tall eastern temple in Plaza A that is positioned directly above the cave entrance. Based on the number of plaster floors present, Structure 1 was built in at least four construction phases as indicated by four plastered floors (Carpenter 2013; Robinson 2012). The structure was constructed using dry laid fill and contained very few ceramic artifacts. Th ose that were present date to the late facet of the Late Classic (700 900AD) (Kosakowsky 2012; Kosakowsky 2013). The floors were underlain by small cobbles to create smooth surfaces to accommodate the plasters ( Carpenter 2013; Robinson 2012). Methods Of the 76 platforms located in the Entrance Chamber of the cave at Las Cuevas, 16 platforms were selected for at least one type of geochemical analysis and 8 were swept and

PAGE 246

Ray, Moyes, and Howie 255 Figure 3 Map of the Entrance Chamber. Note the walled construction at the rear of the entrance chamber. Figure 4 Evidence of multiple plastering episodes from Platform 42. cleared for systematic testing ( Table 1 ). The sediments covering the platforms consisted of bat guano, air borne particles, and calcite sloughing from the ceiling above. Platforms were chosen for analysis based on degree of completeness and location within the cave incorporate spatial variation (Ray 2014). Additionally, those with multiple plastering episodes and clear edges were targeted. Excavation and clearing (sweeping) of the platforms yielded artifacts, including jute ( Pachychilus sp., freshwater gastropod), animal bone, potsherds, and obsidian blades. Samples were also taken from each of the four construction phases from Structure 1 in Plaza A. Samples were removed using a hammer and trowel and wrapped with foil to retain shape and orientation for transport to the lab. All of the methods utilized provide different yet complementary result s. Portable X ray fluorescence or pXRF was used to identify both light and heavy elements on the plaster surface and subsurface. A Fourier Transform Infrared Spectrometer, or FTIR was used to determine the structure of the elements on the surface. This method of analysis allows researchers to obtain information about crystalline and amorphous materials as well as

PAGE 247

Plastered: Cave Constructions at Las Cuevas 256 Table 1 List of all samples taken for analysis. organic materials. FTIR was used in conjunction with the pXRF data in order to get the big picture view of the chemical composition of the plaster surface. A Bruker Tracer IIIV handheld XRF and a Bruker Alpha Diamond ATR FTIR on loan from California State University, Long Beach courtesy of Dr. Hector Neff was us ed for analysis. An X Ray Diffractometer in the Imaging and Microscopy Lab at UC Merced was used to specifically identify the crystalline compounds present in the plaster matrix. Visual identification at a larger scale (microscopical analysis) and small scale (electron microscopy) of binders, aggregates, particle size, and surface treatments round out the analysis of the plaster matrix and recipes. In situ pXRF analysis was performed on Platforms 14, 29, 59, 60, and 61 using the Bruke r Tracer Handheld pXRF. All of these samples were run on high energy and Platform 14 was also run on low energy following the same specifications listed below. Running the samples at low energy provides information about light elements (Mg to Cl). Physi cal samples were collected from each intersection from the platforms, with assistance from Andrew Neff, Daniel Neff and Hector Neff. These samples were brought back to the field lab and were homogenized and analyzed on both high and low energy. Bulk samp les were collected from platforms 4, 5, 26, 30, 32, and 48 and brought back to UC Merced for high energy analysis with a Bruker Tracer handheld pXRF. The high energy was run using a green filter at 40 kV and 26 and the low energy was run using a blue f ilter at 15 kV and 27 with a Samples pXRF XRD FTIR Petrography SEM LC S3 x x LC S5 x LC T1 x LC T2 x x LC P4 x LC P5 x LC P14 x x x x x LC P26 x x LC P29 x x LC P30 x x LC P31 x x LC P32 x x x LC P37 x LC P42 x x x LC P48 x x LC P59 x LC P60 x LC P61 x LC P62 x LC P63 x LC P64 x LC F1 x x x LC F2 x x LC F3 x LC F4 x x x LC 2013F4 x x LC F5 x x Bedrock x x x

PAGE 248

Ray, Moyes, and Howie 257 vacuum pump. Samples for FTIR from platform 14 were systematically collected in the field and were homogenized and analyzed in the lab. Preparation for XRD consisted of homogenizing sample using a gate grinder. They were t hen placed in a sample holder and compacted into the sample holder with a small pestle. This was done to try to ensure that each orientation of crystals found within the sample are represented so that the diffraction can identify each of the orientations making it easier to identify the phases present. Plaster samples from 13 unique contexts were analyzed using XRD. Seven different platforms and one surface structure were examined. The samples were run at 45 KeV and 40 using a PANalytical XPert Pro Theta. Analysis of the spectra was completed using XPert HighScore Plus software with an automated Search & Match to identify the different phases, or crystal structures, present in each sample. The Scanning Electron Microscope was utilized to provide information about the binder and aggregate particle sizes. The particle sizes of the carbonate binders are a reflection of the firing temperatures. Powdered samples from Platforms 14, 31, and 42 were prepared on carbon tape and then sputter coated with g old. Small bulk samples from Platform 14 were selected based on size in order to fit inside the sample area within the SEM. To stabilize small samples, they were attached to carbon tape on an SEM stub. Finally a bedrock sample was placed directly in the sample chamber for SEM EDS analysis. Powdered plaster samples were run in high vacuum mode using an FEI Quanta 200 ESEM with a tungsten filament. On average vacuum pressure was 105 Torr. Using secondary electrons, the surface morphology of the samples were examined a nd the particle sizes measured. Plaster and bedrock samples were sent for petrographic analysis conducted by Dr. Linda Howie. Platforms 14, 26, 29, 30, 32, 42, and 48; stairway 3; terrace 2; and Structure 1 floors 1, 2, and 4 were sent for analysis. Before preparing thin sections, samples were first examined under a stereomicroscope in cross section to document structural, textural, and compositional characteristics. The structural and textural examination looks for the overall homogeneity Figure 5 Concentrations of iron in ppm. presence or absence of voids, microstrata or layers, overall appearance, and possible presence of any surface treatments such as paint. The compositional analysis identifies presence of different carbonate, mineral and organic constituents that occur as particles and rock and mineral clasts through microscopic analysis before and after effervescence as the result of interaction with dilut e hydrochloric acid solution. The full petrographic analysis has not been completed so only data from the preliminary report is included here (Howie 2016). Results Preliminary analysis of plaster samples sent for petrographic analysis demonstrated a high amount of variability. This included differences in the color of clasts present, ranging from pink to green; the presence of yellow, amorphous or gel like areas which could indicate hydraulic properties; and differences in clast shape, size and compositi on suggesting variation in aggregates and source material (Howie 2016). Microscopic analysis also

PAGE 249

Plastered: Cave Constructions at Las Cuevas 258 Figure 6 FTIR spectra of surface and subsurface from platform 14, note the absence of some of the peaks from the subsurface. identified several painted surfaces that were not recognized macroscopically in addition to replastering events. High iron content on Platform 14s plaster surface was present on large portions of the platform. We initially suspected that this might be due to clay minerals mixed into the plaster matrix. However, analysis of surface and subsurface samples from the same location show a lack of iron subsurface. Additionally, iron content was lower underneath the overhang ( Figure 5 ), suggesting the iron was not part of the plaster fabric, but was deposited through activity on the platform FTIR data was interpreted using the comparative data published by the Kimmel Center for Archaeological Science Infrared Standards Library from the W eizmann Institute of Science (2014) As expected, the most predominate material came from the calcite mineral from the calcium carbonate plaster matrix. However, one large peak and several smaller peaks, around the 1050 cm1 wavelength, were not easily ide ntified ( Figure 6 ). This particular wave number can be related to either phosphates or silicates. Similarly, these peaks were completely absent from subsurface samples suggesting that a different source of the iron had been introduced onto the platform s urface. Though many of the samples produced the same or almost identical XRD spectra, several of the samples from the cave were not homogenous. Two samples showed peaks at 31, one from the cave (Platform 62) and one from the surface (Structure 1 Floor 2) This indicates the presence of dolomite instead of the normal calcite. Though both are forms of calcium carbonate they are often associated with different source material. Analysis of the three platforms showed a distinct difference in particle size di stribution. Platform 31 had a more varied and larger

PAGE 250

Ray, Moyes, and Howie 259 Figure 7 Carbonate particles at 500x magnification, a) Platform 14; b) Platform 42; c) Platform 31. Figure 8 EDS spectra. a) Bedrock sample from site core; b) Titanium sphere found on the bedrock surface. particle size. This is suggestive of a lower firing temperature that would not allow the limestone to slake properly. Platform 42 exhibited smaller particle sizes but some small variation. Using the bulk sample from Platform 14, crystals of many sizes were examined though again they appeared to be more homogenous in size ( Figure 7). Several points of interest were examined using the EDAX Gene sis energy dispersive xray spectrometer though in the bulk plaster sample both the surface and subsurface. On the surface, several pieces of lead and possibly iron and zinc were identified. Though the presence of iron and zinc were already identified in the bulk XRF analysis, lead was a new discovery. The origin of the lead is currently unknown. No lead was found in analysis of bat guanos from Chechem Ha, also located in Western Belize (Moyes 2006). The subsurface of the plaster appears to be homogenous ly constructed of calcium carbonate. These findings are supported by previous XRF analysis. Finally EDS was used to analyze a bedrock sample from the surface. The bedrock is likely calcite though further tests should be used to determine the phase sinc e XRD results suggest the presence of dolomite elsewhere ( Figure 8a ). On the surface of the bedrock, a small titanium sphere was identified ( Figure 8b). Discussion Results of the FTIR analysis suggest that the peaks at 1050 cm1 wavelength represent a phosphate mineral created through the interaction between phosphoric acid from bat guano and the plaster surface (Ray 2018). As explained by Hutson and Terry (2006), residues can remain imbedded in the plaster in the reactive zone just below the surface by migrating there through trampling of the plaster surface. Though in the case of the cave,

PAGE 251

Plastered: Cave Constructions at Las Cuevas 260 trampling is not the only type of post depositional modification; acidicbat guano and drip water interacts with the basic calcium carbona te plaster. Therefore, the results suggest that it is a phosphate mineral created through the interaction between phosphoric acid from bat guano and the plaster surface. According to cave research from Romania conducted by Giurgiu and (2013) and Du mitras and colleagues (2008), hydroxyapatite is the most common phosphate mineral in bat guano deposits especially in drier caves as could be considered the case with the cave at Las Cuevas. Shahack Gross (2004) and colleagues also acknowledge the prevalence of Fe rich phosphates in degraded bat guano. Preliminary petrographic analysis of the plaster surfaces suggests that many of the platforms may have been painted red. Additionally this pigment appears to be of different manufacture, as indicated by di fferences in color and morphology of particulate material relating to the colorant. So far it is evident on all of the earlier phases (when two construction phases are present). This suggests that either this was an earlier practice that was later abandoned or the presence of the pigment on exposed surfaces has since disintegrated. It is likely to be the latter and the iron on exposed surfaces associated with the phosphates derives from these pigments. The petrographic analysis proved to be the most use ful in identifying the v ariability of the plaster recipes. pXRF and FTIR data did not identify an y differences between the plasters. Chemically the plasters are all very similar and the high carbonate content masked some of the variability. However, XRD analysis demonstrated variation in carbonate phases meaning that different source material was utilized for cave and surface constructions. Although the petrographic analysis is not yet complete, preliminary results from the thin sections, suggest that mu ltiple types of carbonate rock and carbonate rock derived materials were used in the plaster recipes. While this is supported by other evidence from the XRD analysis this new data demonstrates that multiple carbonaterockderived ingredients were used wit hin one batch of plaster perhaps suggesting the use of specific kinds of carbonate rock that produce hydraulic properties. Finally, the observed variation in overall textural and compositional characteristics as well as the size, quantity and composition of clasts and aggregates reveal the differences in plaster recipes. At the site of Copan Abrams and Rue (1988) analyze deforestation in the context of the Late and Terminal Classic in the site core. They argue that the main driver of deforestation was the acquisition of wood for domestic use (i.e. cooking hearths in the home) and that the wood used for plaster construction and building domestic structures and clearing for residence and agriculture was negligible. Using estimates of labor and wood use for the construction of the extent of the Late Classic site core, and an estimate of total available wood, they argue that ultimately the Copan landscape was not sufficient to satis fy the needs of the population. What this doesnt take into account though is t he pine used for ritual contexts or plaster used to cover large plazas. At Chan, pine was most commonly found in ritual contexts (Lentz et al. 2012). Also their argument against plaster production uses a model in which they split the production of plaste r into a 50 year cycle but this may not be that simple. If they are trying to plaster one temple or large public works over a very short period this large scale event may prove to have even more detrimental effects on the overall forest. Also what this neglects to demonstrate is the change through time from the Early Classic into the Late Classic. It has been noted elsewhere that during the later construction phases plaster became thinner thus suggest ing a mitigation of resources (McNeil et al. 2009). N eithe r does this account for plaza plaster which may have been thicker than on the surface of the temple as it is repeatedly used and walked on. Other sites see a switch to sascab as a way to preser ve resources (Villaseor 2009). At Las Cuevas a change i n firing temperature could produce the variation in particle sizes present in the plaster (Ray 2018). Particle sizes can be quickly and accurately measured using a laser scattering technique and a powdered sample. Future work will employ this technique t o definitively measure

PAGE 252

Ray, Moyes, and Howie 261 differences in particle size distribution amongst the samples. The identification of many unique plaster recipes at the site was an unusual and unexpected find. The bedrock that has been examined near the cave and the calcium carbon ate that has been analyzed thus far from the cave has all been calcite rather than dolomite. This suggests that some of the source material, at least at some point in time was coming from elsewhere. It could be that these materials were found in a nearby outcrop that was used for building materials elsewhere at the site core, especially since the cut stones used in Plaza A are visually distinct from the underlying bedrock Geologic maps of the country are not highly resolved. We would expect to see homog eneity in raw material based on the assumption of local procurement. Though it is likely that procurement was still local, the results of the microscopic and XRD analyses suggest variation in the landscape. Geological survey of the area around Las Cuevas began in the 2017 field season which confirms this variability. The variation that has been identified is likely the result of small batches of plaster (Ray 2018) There are a number of competing explanations for this finding. Moyes et al. (2015) argued that the cave was constructed by and for elite use which should produce consistently similar plaster recipes, which suggests a conscripted labor force. However, these results may not support this model. Although it is obvious based on the organization of the constructions in the cave that the architectural modifications were planned and that some sort of centralized authority oversaw platform and plaster manufacture, work groups may have been cooperative rather than conscripted. This is not at odds with other examples of ritual constructions such as the construction of European cathedrals that were labors of devotion and civ ic pride using voluntary labor. However it is also possible that the sampling strategy employed for this project may have masked some of the similarities in plaster recipes if the architecture was built up in sections. Additionally, although our chronologies are quite good for archaeological timeframes, it is possible that the platforms were built as an accretionary process over time. This would easily account for differences in plaster recipes. To summarize the variations in plaster recipes and source material from these data could be the result of: 1. small workgroups bringing in different material from the local area. Plaster could be created either as a type of offering or as individual contrib ution to the religious project. 2. accretionary building practices. There could be considerable variation between clusters rather than within clusters that our sampling strategy would not have identi fied. Conclusion Las Cuevas was only occupied for a short time during the Late to Terminal Classic Period (AD 750AD 850). It was a politically tumultuous period for the ancient Maya across the lowlands and the elite were losing power and control at many Classic Period sites. The scale and organization of the constructions despite the variation in plaster recipes suggest a centrally organized design. Because there is a clear organization to these constructions it seems unlikely that it was the re sult of individuals building their own platforms. The most parsimonious explanation for the variability in plaster recipes as well as evidence of multiple construction phases for individual platforms is that the architecture was augmented over time. Stru cture 1 consisted of at least four construction phases, the project obtained only relative dates for the floors based on the stratigraphy alone as the ceramic materials could not be further refined other than to say they dated to the late facet of the Late Classic Period. Except for the dolomitic source material, none of these recipes seem to match with recipes from the cave. It is possible that the cave and the surface site were constructed at different times. Alternatively cooperative or lineage groups may have supplied the labor and construction efforts necessary to erect the massive temples and complex network of platforms in the cave contributing their own source materials from the sites periphery. This could account for variation in the source

PAGE 253

Plastered: Cave Constructions at Las Cuevas 262 mat erials. Further research should involve mapping variation of limestone sources in the local landscape and could help to sort this out. Acknowledgements This work would not have been possible without the support of many individuals and organizations. Fi rst a big thank you to Dr. John Morris and Dr. Jaime Awe and the rest of the staff at the Institute of Archaeology for their support and for giving us the opportunity to present this research. Additionally, this research would not have been possible witho ut the help of numerous individuals in the field and in the lab. Thank you to Dr. Hector Neff for getting us started on this project, and Andrew and Daniel Neff for help collecting samples. Dr. Kathleen Hull allowed us the use of her pXRF in her lab at U C Merced. Dr. Asmeret Berhe provided access to the FTIR analysis program. Dr. Valerie Leppert and Dr. Mike Dunlap from the Imaging and Microscopy Facility at UC Merced provided support and training for the use of the XRD and SEM EDS. Finally we would li ke to thank the Alphawood Foundation and UC Merced for their cont inued support of this research. References Abrams, E.M. and Rue, D.J. 1988 The Causes and Consequences of Deforestation Among the Prehistoric Maya. Human Ecology 16(4): 377395. Carpenter, Maureen 2013 Stairways to Heaven. In Mark Robinson and Holley Moyes (eds.) The Light and Dark Side of Las Cuevas: Archaeological Investigations of the 2013 Season. Report on file with the Institute of Archaeology, NICH, Belize. Digby, Adrian 1958 New Maya City Discovered in British Honduras: First Excavations at Las Cuevas, An Underground Necropolis Revealed. The Illustrated London News, Feb. 15, 1958. Georgeta, Stefan Marincea, Essaid Bilal, and Frederic Hatert 2008 Apatite -(CaOH) In the Fossil Bat Guano Deposit from the Dry Cioclovina Cave, Sureanu Mountains, Romania. Canadian Mineralogist 46(2): 431 445. 2013 Mineralogical Data on Bat Guan o Deposits from Three Romanian Caves. Studia Universitatis Babes -Bolyai, Geologia 58(2): 13 18. Howie, Linda 2016 HD Analytical Solutions Technical Report 201600 3: A Preliminary Microscopical Characterization Study of Late Classic Maya Building Materials from Las Cuevas and related Cave Sites, Belize. Hutson, Scott R, and Richard E Terry 2006 Recovering social and cultural dynamics from plaster floors: chemical analyses at ancient Chunchucmil, Yucatan, Mexico. Journal of Archaeological Science 33(3): 391 404. Kosako wsky, Laura J. 2012 LCAR 2012 Ceramic Report. In Holley Moyes and Mark Robinson (eds.) 2nd Report of the Las Cuevas Archaeological Reconnaissance Project: The 2012 Field Report. Report on file with Institute of Archaeology, NICH, Belize. 2013 Ceramic Analysis from the 2013 Excavations at Las Cuevas. In Mark Robinson and Holley Moyes (eds.) The Light and Dark Side of Las Cuevas: Archaeological Investigations of the 2013 Season. Report on file with the Institute of Archaeology, NICH, Belize. Kosakowsky, Laura J., Holley Moyes, Mark Robinson, and Barbara Voorhies 2013 Ceramics of Las Cuevas and the Chiquibul: At World's End. Research Reports in Belizean Archaeology 10:2532. Insti tute of Archaeology, NICH, Belize. Lentz, David L., Sally Woods, Angela Hood, and Marcus Murph 2012 Agroforestry and agricultural production of the ancient Maya at the Chan Site. In Chan: an ancient Maya farming community in Belize Cynthia Robin ed. p g.89 -110. University Press of Florida. McNeil, C L, D A Burney, and L P Burney 2010 Evidence Disputing Deforestation as the Cause for the Collapse of the Ancient Maya Polity of Copan, Honduras. Proceedings of the National Academy of Sciences 107(3): 1017 1022. Moyes, Holley 2006 The Sacred Landscape as a Political Resource: A Case Study of Ancient Maya Cave Use at Chechem Ha Cave, Belize, Central America Ph.D. dissertation, Department of Anthropology, State University of New York, Buffalo. 2012 Constructing the Underworld: The Built Environment in Ancient Mesoamerican Caves. In Heart of Earth: Studies in Maya Ritual Cave Use, edited by James E. Brady, pp. 95-110, Association for Mexican Cave Studies, Bulletin Series No. 23,

PAGE 254

Ray, Moyes, and Howie 263 Austin, TX. 2015 The Cave at Las Cuevas. In Dreams at Las Cuevas: Investigating a Location of High Devotional Expression, Results of the 2014 Field Season, edited by Mark Robinson and Holley Moyes, pp. 89-108. Report on file at The Institute of Archaeology, National Institute of Culture and History, Belmopan, Belize. Moyes, Holley, and James Edward Brady 2013 Caves as Sacred Space in Mesoamerica. In Sacred Darkness edited by Holley Moyes. University of Colorado Press, Boulder. Moyes, Holley, Laura Kosakowsky, Mark Robinson, and Barbara Voorhies 2012 Better Late than Never: Preliminary Investigations at Las Cuevas. Research Reports in Belizean Archaeology 9: 221231. Institute of Archaeology, NICH, Belmopan, Belize. Moyes, Holley, Laura Kosakowsky, and Erin Ray 2017 The Chronology of Ancient Maya Cave Use in Belize. Research Reports in Belizean Archaeology 14, pp. 327338, Institute of Archaeology, NICH, Belmopan, Belize. Moyes, Holley, Mark Robinson, Barbara Voorhies, Laura Kosakowsky, Marieka Arksey, Erin Ray, Nic holas Bourgeois, and Shayna Hernandez 2015 Dreams at Las Cuevas: A Location of High Devotional Expression of the Late Classic Maya Research Reports in Belizean Archaeology 12: 239 250. Institute of Archaeology, NICH, Belize. Moyes, Holley, With contributions by Mark Robinson, Laura Kosakowsky, Barbara Voorhies, Rafael Guerra, Fabrizio Galeazzi, and Josue Ramos 2011 Sleeping Next to the Giant: Preliminary In vestigations of the Las Cuevas Site, Chiquibul Reserve, Belize: A Site Report of the 2011 Field Seas on On file at The Institute of Archaeology, National Institute of Culture and History, Belmopan, Belize. Ray, Erin E. 2014 Traces of Ritual: Understanding Ancient Maya Cave Ritual using Geochemical Signatures from Platform Surfaces at Las Cuevas. In Mark Robinson and Holley Moyes (eds.) Dreams at Las Cuevas: Investigating a Location of High Devotional Expression Report of the 2014 Field Season. Report on file with Institute of Archaeol ogy, NICH, Belize. Ray, Erin E. 2018 Geochemical Analysis of Construct ion Materials from Las Cuevas, Belize. M.A. Thesis. University of California, Merced. Robinson, Mark 2012 Las Cuevas Surface Excavations. In Holley Moyes and Mark Robinson (eds.) 2nd Report of the Las Cuevas Archaeological Reconnaissance Project: The 2012 Field Report. Report on file with Institute of Archaeology, NICH, Belize. Shahack -Gross, R., Berna, F., Karkanas, P. and Weiner, S. 2004 Bat Guano and Preservation of Archaeological Remains in Cave S ites. Journal of Archaeological Science, 31 (9): 12591272. Villaseor Alonso, M.I. 2009 Lowland Maya lime plaster technology: a diachronic approach. Ph.D. Dissertation, UCL (University College London). Weizmann Institute of Science 2014 Kimmel Center for Archaeological Science Infrared Standards Library Weizmann Institu te of Science, Rehovot, Israel.

PAGE 255

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 265 273 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 24 CONTINGENT MULTI -CRAFTING, SURPLUS HOUSEHOLD PRODUCTION, AND THE MAYA QUEST FOR SALT Heather McKillop Addressing the location of residences for ancient Maya craft workers is important in order to evaluate the organization and control of production of commodities with restricted distribution but wide appeal, such as obsidian, chert, and salt. Wooden buildi ngs at the Paynes Creek Salt Works include salt kitchens, where brine was evaporated in pots over fires to make salt. Similar salt kitchens also were likely used at other salt works along the coast of Belize and Pacific coast of Guatemala that have brique tage but lack preserved wood. Stone architecture associated with some Yucatecan coastal salt works and with Salinas de los Nueve Cerros beside an inland salt spring in Guatemala suggest elite control or oversight of production. Spatial patterning of wooden architecture and artifacts at Site 7, the largest of the Paynes Creek Salt Works, suggests some of the buildings were used for multi -crafting or as residences, perhaps by elites Introduction From at least the Late Preclassic period, the Maya sought sources of salt, a commodity that was biologically necessary, but also valued for preserving fish and meat, tanning skins, enhancing the flavor of food, and for medicinal and other uses. Paleoindian and Archaic people obtained enough dietary salt from hunting wild animals. The quest for salt began with increasing reliance on domesticated plant foods, especially carbohydrate rich corn which was the dietary foundation that supported large sedentary populations throughout the Maya ar ea. This pattern is mirrored in other areas world wide where staple carbohydrates deficient in salt underwrote large populations and the rise of civilizations (McKillop n.d.). In the Maya area, salt was obtained from solar evaporation along the arid coas ts of the Yucatan and by evaporation in pots over fires along the coast of Belize, Pacific coast of Guatemala, and at inland salt springs. Household production of salt and other commodities for personal use works with communities located near sources of s alt water. The restricted distribution of salty water in the Maya area required strategies for those not near salty water to acquire salt ( Figure 1; McKillop 2005a). By the Classic period, even small communities such as Ceren, El Salvador, had household s pecialization in different commodities for marketplace trade in several other communities (Sheets et al. 2015): Householders produced surplus commodities in the course of regular subsistence and home maintenance. Surplus household production Figure 1 Map of the Maya area showing location of Ancient Salt Works. took place within the home or at varying distances from the home. In some cases, there were workshops at the source of resources with a restricted distribution for the extraction or production of the resource, including salt, highquality chert, jadeite, and obsidian. The high quality outcrops of chert in northern Belize were exploited from the Middle Preclassic to the

PAGE 256

Multi-crafting, Surplus Household Production, and the Quest for Salt 266 Early Postclassic at Colha (Shafer and Hester 1983) by surplus household production of stone tools and tool blanks. Jadeite was extracted from outcrops along the Motagua River by nonelites (Rochette 2009). Obsidian was quarried from outcrops in the volcanic highlands of Mexico, Guatemala, and Honduras and shaped into macrocor es for trade. Household production often included multicrafting production of a variety of crafts, and sometimes inclu ded contingent, multi crafting production that required additional tasks (Hirth 1998). Householders were accustomed to multi tasking in their daily lives in house construction, from extraction and transportation of wood and leaf for pole andthatch structures, to procurement of temper and clay for making pots, and production of farming implements. Contingent, multi crafting is an extensio n of subsistence activities: Inland trade of salted fish required fishing or otherwise acquiring the fish, as well as sun drying or salting the fish. Salt drying fish required making or otherwise acquiring salt. Both sundrying and salt drying fish requir ed outdoor or indoor racks. Chert, jadeite, and obsidian objects were available in marketplaces at lowland cities and towns (Chase et al. 2015). In some cases, householders brought their goods and resources for trade at marketplaces in various communities Sometimes traders transported commodities greater distances. Coastal trading ports such as Moho Cay (McKillop 2004), Wild Cane Cay (McKillop 2005b), Isla Cerritos (Andrews et al. 1989), and Xcambo (Sosa et al. 2014) as well as inland ports such as Cancu en (Demarest et al. 2014) provided sa f e harbor, accommodations, and warehousing. They also added maritime and other commodities, such as salt, stingray spines, and salted fish, or further manufacturing, such as jadeite at Cancuen (McKillop 2007, n.d.). In this paper I evaluate the conditions and significance of extraction and production of restricted resources carried out as part of household production. In some cases, extraction and production of restricted resources was carried out as part of surplus household production at home. In other cases, surplus household production was spatially segregated from homes. Independent factory or attached specialization for the state are situations in which the extraction and production of restricted resources was n ot directly tied to household production. With household and surplus household production, labor is controlled by the householders. In contrast, labor is controlled by factory owners. State control of corve labor for public works removes both the worke r and control of labor from the household. Attached specialization to produce commodities for the state at the Maya city of Aguateca consisted of elite household production where highlytrained specialists produced goods for the royal court, but also some goods they produced and traded themselves (Inomata 2001). Surplus household production is carried out in catchment zones that vary by the localized abundance of resources. Some resources are seasonally abundant, such as deer, manatee, and green sea turtles. Other resources are naturally restricted in distribution, such as obsidian, chert, and jadeite. Some resources are suitable for extraction or production in specific conditions, including chocolate, cotton, and scarlet macaws. There are limits to daily travel to extract and produce commodities, with distances significantl y greater by boat than by land. Modern salt makers live near their salt kitchens or salt pans, either year round or seasonally. At Sacapulas, Guatemala, up to 30 salt kitchens are located near a salt spring, with the homes located farther back. Production takes place year round, although most production is carried out in the dry season when the salt spr ing is accessible (Reina and Monaghen 1981). On the southeast coast of China in Shandong Province, salt works were located along the coast, with residences located farther back (Flad 2013). State sponsored overseers were located still farther back. In c entral Mexico, salt works were located a few km from the salt workers residences. Residences were suspected of being nearby Zhongba, a large salt works flooded by the Three Gorges Dam in China (Flad 2011). Northern River Lagoon was a village and trading port where salt was produced to salt fish for inland trade (Masson and Mock 2004: 372373). In west Mexico, salt makers travel during the

PAGE 257

McKillop 267 Figure 2 Map of the Paynes Creek Salt Works, with insert showing location in southern Belize. Base map by Mary Lee Eggart. dry season to the coast to make salt, setting up temporary homes (Williams 2003). The Paynes Creek Salt Works The Paynes Creek Salt Works include 110 sites over a five square km area in a salt water lagoon system on the south coast of Belize, north of the modern town of Punta Gorda ( Figure 2; McKillop 2005a, n.d.). During the dry season, the shallow lagoon system becomes a natural solar evaporation pond, concentrating the salinity, thereby reducing the time for evaporating brine in pots over fires to make salt, as compared with the open sea. Furthermore, less wood fuel is needed. As indicated by excavations at two sites w ith earthen mounds that are not underwater, the salt water was poured through salty soil to enrich the salt content before the evaporation process (Watson et al. 2013). The canoe excavated at the Eleanor Betty site was held in place by stakes and had a pottery funnel below (McKillop et al. 2014). Sediment chemistry research at Chan bi (Site 24) suggests that salty red mangrove peat may have been cut from the seafloor or land (Sills et al. 2016). Pouring brine though salty soil is documented in historic times at Sacapulas, in the highlands of Guatemala (Reina and Monaghen 1981), in West Mexico (Williams 2003), and elsewhere. With the preservation of wooden buildings, a canoe, and abundant briquetage not subject to breakage from trampling after the sites were abandoned, the Paynes Creek Salt Works are a model for salt production elsewhere along the coast of Belize. Transect excavations along the interior walls of salt kitchens at 10 sites revealed high percentages of briquetage, from 90 98%, indicating tha t salt production was the main activity (McKillop n.d.; McKillop and Sills 2016, 2017). The salt works may have been the location of surplus household production, with the possibility that some of the wooden buildings were houses. Alternatively, the salt workers lived nearby in the coastal area, perhaps at Wild Cane Cay, a short canoe paddle from the salt works. Wild Cane Cay was a 10 acre village and trading port, originally settled in the Early Classic that expanded in the Late and Terminal Classic (McKillop 2005b)

PAGE 258

Multi-crafting, Surplus Household Production, and the Quest for Salt 268 Mass production of Salt as Indicated by Standardization of Briquetage Standardization of briquetage at the Paynes Creek Salt Works suggests mass production of salt. However, the results indicate differences among individual salt works, suggesting there were work parties that may have consisted of families from nearby communities or seasonal residents from inland cities. In addition to calculating the co efficient of variation, which is commonly used in standard ization studies in archaeology I also reported the average median variation to evaluate the standardization of briquetage from three sites in the Stingray Lagoon area: Stingray Lagoon, David Westby, and Orlandos Jewfish sites (Figure 2; McKillop 2002:127132, Tables 4.14.6). The av erage median variation (AMV) statistic was used instead of the coefficient of variation (CV) because some of the samples have distributions significantly different from normal. The AMV reduces the effect of outlying values in a distribution, unlike the CV To calculate the AMV, each value is subtracted from the median. The absolute value of this number is divided by the median. The average of this number is then multiplied by 100 to produce the AMV. The diameters of jars and bowls of Punta Ycacos Unslip ped, as well as the diameters of the salt pot vessel supports solid clay cylinders were standardized, as compared to a sample of pots from Wild Cane Cay (McKillop 2002: Table 4.1). The differences were statistically significant. Interestingly, the averag e median variation of the briquetage from the three salt works was different for each site, and the results were statistically significant (McKillop 2002: Tables 4.4, 4.6). The standardization of briquetage was further supported by findings at Chan bi in the East Lagoon (McKillop and Sills 2017: Table 4.1). The statistically significant differences in standardization of Punta Ycacos bowls, jars, and vessel supports at the three main salt works in the Stingray Lagoon area indicate the salt workers at each of the sites likely made their own salt pottery. Multicrafting at the Salt Works Salt production required a variety of supporting activities, so it can be evaluated in terms of contingent multi crafting (Hirth 1998). In addition to salt production, other activities at the salt works included making the salt making pottery, collecting posts and sharpening the ends, and other woodworking. The sand and clay for making the salt production pottery is available widely wit hin the lagoon system. Clay is exposed in the cut banks where the lagoon abuts the broken pine ridge (savannah) in many locations. Quartz sand is available in thick lenses below the sea floor in multiple locations. Punta Ycacos Unslipped pottery was poorly formed and fired at low temperatures, making it friable and not viable for transporting from the place of production. Posts for building construction were obtained from nearby locations in the mangroves, from the nearby deciduous rainforest to the south, but not from the adjacent broken pine ridge to the north. Some deciduous hardwoods such as Sapodilla grow in dry patches within the mangrove ecosystem. Palmetto palms are ubiquitous on higher ground behind the mangroves, so would have taken little ef fort to obtain. Post diameters indicate that most hardwood trees were young, with diameters rarely exceeding 18 cm. The hardwood posts were cut and sharpened at one end, before they were driven into the ground. Other wood working included making canoes, paddles, and other wooden objects. Living at the Salt Works? Although most of the salt works have one or two wooden buildings, a few sites have significantly more, prompting the question whether some of the buildings may have been residential. Excavation of two buildings at Site 74 indicate both were salt kitchens, with extremely high densities of briquetage compared to other pottery (McKillop and Sills 2016). High densities of briquetage also were discovered in transect excavations at other salt works ( McKillop n.d.). Site 7 and 60 are large sites, each with multiple buildings. The possibility of residential use of Site 7 will be evaluated using the spatial patterning of wooden buildings and artifacts that were individually mapped on the surface of the sea floor.

PAGE 259

McKillop 269 Figure 3 Labeling flags that Mark Wooden Posts and Artifacts at Site 7. Photo by H. McKillop. Site 7 is an underwater site at the entrance to the West Point area of the lagoon system ( Figure 2 ). We used flotation survey in order to protect the wooden posts that protruded from the seafloor and not trample on the artifacts embedded in the sea floor (McKillop 2005a). The entire site area was systematically traversed by flotation survey, with the tea m on Research Flotation Devices (RFDs), back and forth across the site and beyond in all four directions. During survey, all wooden posts were marked with pin flags. All diagnostic pottery, all stone artifacts, and a selection of briquetage were flagged by pin flags ( Figure 3 ). Few of the wooden posts were visible on the sea floor, which is covered by a thin layer of silt. In cases where the posts protruded slightly above the sea floor, they are worm eaten and decayed. The palmetto palm posts retain the outer shell of bark. Posts were embedded in solid, red mangrove peat that was virtually indistinguishable from wooden posts. When discovered, posts were cleared around the exterior in order to define the post for a diameter measurement, and to verify th e post was indeed a post and not a mangrove root or mangrove peat. Certainly, some wooden posts were not found. The flagged posts and artifacts were labeled, with each post and artifact location mapped using a total station. The locations of buildings we re not evident during survey, but only became evident after the digital data were downloaded and added to the project GIS. Mapped posts include a line of palmetto palm Figure 4 Map of Site 7 wooden posts. posts and hardwood posts. Most of the palm etto palm posts form a closely spaced line interpreted as a land retaining wall on the east side of the site. In contrast to most of the other underwater sites that consist of one or two wooden structures, Site 7 has over 400 wooden posts that define the outlines of 10 or more wooden buildings around three plazas ( Figure 4 ). Several of the buildings have distinct rectangular footprints, including Structures B, C, and G. Most buildings are oriented slightly northeast to southwest. Most buildings are singl e room structures, similar to the salt kitchens at Site 74 (McKillop and Sills 2016). Structure 3 has at least two rooms, as well as an anteroom. None of the wooden buildings had stone foundations or plaster floors, typical of structures at other Maya sit es. The wooden structures provide form to invisible architecture described at other sites, where there are no visible remains of pole and thatch buildings on the modern ground surface (Johnston 2004; Somers and McKillop 2005). However, just as wooden s tructures built of pole and thatch form the majority of buildings in traditional Maya villages, pole and thatch buildings likely formed the majority of houses, workshops, and other structures in antiquity. The wooden structures at the Paynes Creek Salt Wo rks are the only preserved pole and thatch buildings from ancient Maya times. The temporally diagnostic Belize Red, Pantano Impressed (also called Warrie Red), and other nonbriquetage pottery support the main

PAGE 260

Multi-crafting, Surplus Household Production, and the Quest for Salt 270 use of Site 7 in the Late (600800 C.E.) to Te rminal Classic (800 900 C.E). Unit stamped decorations occur on the exterior of jars around the shoulder and sometimes extending farther down the body of the vessel, as with the single example from Site 7 ( Figure 5). The surface finish and color are eroded and the calcite temper is dissolved by the acidic peat. Other pottery corroborate the Late to Terminal Classic age of the site and point to multicrafting and perhaps residential use of some buildings. A partial vase with incised glyphs below the r im on the exterior is typical of the Terminal Classic elsewhere. The surface finish and paint are eroded, so the vase has not been assigned to a pottery type. A miniature vessel described by other researchers as a snuff bottle, also was recovered. Vases with glyphs and snuff bottles were not recovered from any of the other underwater sites. A perforated potsherd disk used as a spindle whorl is typical of the Classic period and suggests spinning cotton for clothing or fishing lines or nets. The only not ched sinker, which was made from a pebble, was found at Site 7. Sherds from large incense burners, as well as candeleros, were recovered from Site 7. The presence of incense burners, candeleros, Belize Red, snuff bottle, and the vase with glyphs, point t o residential use of one or more buildings at Site 7. Spatial patterning of artifacts suggests buildings were used for different purposes, including some dedicated to salt production, other crafts or residential use ( Figure 6 ). Most artifacts are found al ong the interior walls of buildings. This pattern was observed at the Ceren Site, which was rapidly abandoned due to an imminent volcanic eruption that covered the sixth century village with volcanic tephra, which also sealed the remains (Sheets et al. 20 15). The interpretation for Ceren was that artifacts were stored in the rafters or on shelves along the walls and dropped during the volcanic eruption. They represent evidence of household activities but are not activity areas. There is a different dist ribution of briquetage and nonbriquetage pottery, suggesting different uses for buildings. Buildings B, C, and G include Belize Red and Pantano Impressed pottery but no briquetage. The lack of briquetage in the three wooden buildings at Site 7 suggests they were not salt Figure 5 Unit -stamped decoration on Site 7 Pantano Impressed (Warrie Red) sherd. Photo by H. McKillop. Figure 6 Distribution of pottery types and wooden buildings at Site 7. kitchens, but were instead used for other purposes, including as residences for the salt workers. Structures C, H, and I include briquetage but no other pottery types. Mangrove Unslipped jars occur s in limited numbers with the briquetage. The distribution of artifacts indicates most activity was indoors instead of the plazas. The three plazas are virtually devoid of artifacts. Artifacts do not extend beyond the line of palmetto palm posts that form the east side of the site. The missing element in salt production at Site 7 is any trace of the b rine enrichment process. Only Site 50 and Killer Bee have earthen mounds that represent discarded, soil

PAGE 261

McKillop 271 leached of its salt (Figure 2; Watson et al. 2013). The wooden canoe raised by wooden stakes with a funnel below from the Eleanor Betty Site represents the brine enrichment process, in which salty water was poured through a container of salty soil, with the enriched brine collected below (McKillop et al. 2014). Earthen mounds were likely associated with each salt kitchen, as they are in modern salt works (Williams 2003). Sea level rise and wave action have deflated the earthen mounds and spread the leached soil over the sea floor at the Paynes Creek Salt Works. The Placencia Lagoon Salt Works (Sills 2016, 2017) further corroborate the hypothesis that earthen mounds were once common at the Paynes Creek Salt Works. The Placencia Lagoon Salt Works lack known wooden structures and are known by the presence o f earthen mounds. Other salt works along the coast of Belize, the Yucatan coast of Mexico, and the Pacific coast of Guatemala, also likely had perishable structures of pole and thatch. Salt works along the coast of Belize and the Pacific coast of Guatemal a had salt kitchens associated with briquetage. Excavations at the single earthen mound at the Guzman Site indicates it was a discard mound from the brine enrichment process (Nance 1992). Salt works elsewhere along the Pacific coast of Guatemala also had salt kitchens, but are only known by the presence of earthen mounds and briquetage (Coe and Flannery 1967). On the Mexican coast of the Yucatan, where salt was produced by solar evaporation due to the arid climate, temporary housing may have consisted of pole and thatch stru ctures that have not preserved. The Quest for Salt The restricted distribution of salt sources in the Maya area did not coincide with centers of population. The ancient Maya quest for salt resulted in salt works associated with inlan d salt springs and salty ocean waters. On the arid climate along the Yucatan coast, salt water was collected in salt pans where it was evaporated by solar evaporation. The resulting salt was collected for local use and transport elsewhere. The earliest known salt production in Belize is during the Late Preclassic at Cerros, using evaporation of brine in pottery vessels to make salt (Robertson 2016). Elsewhere in Belize, salt production was associated with briquetage during the Late and Terminal Classic at Wits Cah AKal (Murato 2011), Marco Gonzalez (Aimers et al. 2016), Northern River Lagoon (Masson and Mock 2004), Placencia Lagoon (Sills 2016, 2017), and at the Paynes Creek Salt Works (Figure 1). The widespread availability of salt works along the coas t and at inland salt springs during the Late and Terminal Classic underscores the likelihood that salt was produced for local use and regional distribution, but not for long distance trade. During the expansion of circum Yucatan trade in the Terminal Classic, Itza Maya traders based on the Gulf Coast of Mexico evidently took control of north coast salt works and Chichen Itza, some 100 km inland from its coastal port of Isla Cerritos (Andrew and Robles 2004). Some researchers have argued that other inland centers controlled salt produced in salt flats in coastal estuaries along the Yucatan coast. Komchen may have controlled salt production on the north coast of the Yucatan and its distribution elsewhere during the Late Preclassic. In some cases, elite resi dences at salt works suggest control or oversight of production, as at Salinas de los Nueve Cerros, Xcambo, and Emal (Figure 1). Salinas de los Nueve Cerros has elite residential stone architecture located beside salt production facilities where brine was evaporated in large jars to make salt (Woodfill et al. 2015). The site dates from the Preclassic through Classic period and may have supplied salt for nearby cities and smaller communities in the interior of the southern lowlands (Figure 1). Northern Ri ver Lagoon, Marco Gonzalez, and Moho Cay are coastal trading ports where salt was produced, so presumably the residents of the trading ports controlled salt production and trade (Figure 1). Some of the salt works likely had wooden architecture and leachin g mounds as at the Paynes Creek Salt Works. Although wooden architecture was used for salt kitchens at the Paynes Creek Salt Works, it may also have been used for residences for the salt workers, as suggested by the spatial patterning of wooden buildings and artifacts at Site 7.

PAGE 262

Multi-crafting, Surplus Household Production, and the Quest for Salt 272 Acknowledgements Field research at Site 7, Paynes Creek Salt Works, was made possible by permits from the Belize Institute of Archaeology. The field work was funded by a Faculty Research Grant from Louisiana State University for the discovery and initial survey of Site 7, National Science Foundation Grant 0513398 and an NSF Research Opportunities for Undergraduates (REU) Supplement for the s ystematic flotation survey and total station mapping of Site 7. The field team included Cory Sills, Mark Robinson, Matt Helmer, Amanda Pitcock, John Young, Orlando Usher, and the author. Friendship and encouragement of our host family at Village Farm, Tanya Russ and John Spang, are appreciated. References Andrews, Anthony P., Frank Asaro, Helen V. Michel, Fred H. Stross, and Pura Cervera Rivero 1989 The Obsidian Trade at Isla Cerritos, Yucatan, Mexico. Journal of Field Archaeology 16:355363. Aimers, James, Elizabeth Haussner, Dorothy Farthing, and Saturo Murata 2016 An Expedient Pottery Technology and its Implication for ancient Maya Trade and Interaction. In Perspectives on the Ancient Maya of Chetumal Bay edited by Debra S.Walker, pp. 149 -161. University Press of Florida, Gainesville. Chase, Arlen F., Diane Z. Chase, Richard E. Terry, Jacob M. Horlacher, and Adrian S. Z. Chase 2015 Markets Among the Ancient Maya: The Case of Caracol, Belize. In The Ancient Maya Marketplace: The Archaeology of Transient Space edited by Eleanor King, 226 250. Tucson: University of Arizona Press. Coe, Michael D. and Kent V. Flannery 1967 Early Cultures and Human Ecology in South Coastal Guatemala. Smithsonian Contributions to Anthropology 3. Washington, D.C.: S mithsonian Institution. Demarest, Arthur, Chloe Andrieu, Paola Torres, Melanie Forne, Tomas Barrientos, and Marc Wolf 2014 Economy, Exchange, and Power: New Evidence from the Late Classic Maya Port City of Cancuen. Ancient Mesoamerica 25: 187-219. Flad Rowan 2011 Salt Production and Social Hierarchy in Ancient China: An archaeological investigation of specialization in Chinas Three Gorges. Cambridge: Cambridge University Press. 2013 Salt Technology in Ancient China: Resources and Networks in Two Bronze Age Cases. Paper presented at the XXXV Coloquio de Antropologa e Historia Regionales El Pasado Tecnolgico: Cambio y Persistencia, El Colegio de Michoacan, Las Fuentes, Michoacan, Mexico. Hirth, Kenneth G. 1998 The Distributional Approach. Curre nt Anthropology 39:451 476. Inomata, Takeshi 2001 The Power and Ideology of Artistic Creation: Elite Craft Specialists in Classic Maya Society. Current Anthropology 42:321 349. Johnston, Kevin J. 2004 The Invisible Maya: Minimally Mounded Residential Settlement at Itzan, Peten, Guatemala. Latin American Antiquity 15: 145-175. Kepecs, Susan M. 2003 Chickinchel. In The Postclassic Mesoamerican World edited by Michael E. Smith and Francis F. Berdan 259 268. Salt Lake City: University of Utah Press. Masson, Marilyn A. and Shirley B. Mock 2004 Ceramics and Settlement Patterns at Terminal Classic Period Lagoon Sites in Northeastern Belize. In The Terminal Classic in the Maya Lowlands: Collapse, Transition, and Transformation, edited by Arthur A. Demarest, Prudence M. Rice, and Don S. Rice, pp. 367401. University Press of Colorado, Boulder. McKillop, Heather 2002 Salt: White Gold of the Ancient Maya. University Press of Florida, Gainesville. 2004 The Classic Maya Trading Port of Moho Cay. In The Ancient Maya of the Belize Valley, edited by James F. Garber, pp. 257272. University Press of Florida, Gainesville. 2005a Finds in Belize document Late Classic Maya Salt Making and Canoe Transport. Pr oceedings of the National Academy of Sciences 102: 56305634. 2005b In Search of Maya Sea Traders Texas A & M University Press, College Station. 2007 Ancient Mariners on the Belize Coast: Salt, Stingrays, and Seafood. Belizean Studies 29(2): 1528. McKillop, Heather, Elizabeth C. Sills, and Vincent Cellucci 2014 The Ancient Maya Canoe Paddle and the Canoe from Paynes Creek National Park, Belize. Research Reports in Belizean Archaeology 11: 297 306.

PAGE 263

McKillop 273 McKillop, Heather and E. Cory Sills 2016 Spatial Patterning of Salt Production and Wooden Buildings Evaluated by Underwater Excavations at Paynes Creek Salt Work 74. Research Reports in Belizean Archaeology 13 229 237. 2017 The Paynes Creek Salt Works: A Model for Coastal Belize. In Ancient Maya Commodities edited by Jennifer Matthews and Tom Guderjan, 67 86. Tucson: University of Arizona Press. Murata, Saturo 2011 Maya Salters, Maya Potters: The Archaeology of Multicrafting on Non-Residential Mounds at Wits Cah AkAl, Belize Ph.d. disse rtation, Boston University, Boston. Nance, C. Roger 1992 Guzman Mound: A Late Preclassic Salt Works on the South Coast of Guatemala. Ancient Mesoamerica 3:27 46. Parsons, Jeffrey R. 2001 The Last Saltmakers of Nexquipayac, Mexico: An Archaeological Et hnography Anthropological Papers 92, Museum of Anthropology. Ann Arbor: University of Michigan. Reina, Ruben E., and John Monaghan 1981 The Ways of the Maya: Salt Production in Sacapulas, Guatemala. Expedition 23:13 33. Robertson, Robin 2016 Red Wares, Zapatista, Drinking Vessels, Colonists and Exchange at Cerro Maya. In Perspectives on the Ancient Maya of Chetumal Bay edited by Debra S. Walker, pp. 125-148. Gainesville: University Press of Florida. Rochette, Erick T. 2009 Jade in Full: Prehispanic Domestic Production of Wealth Goods in the Middle Motagua Valley, Guatemala. In Housework: Craft Production and Domestic Economy in Ancient Mesoamerica, edited by Kenneth G. Hirth, 205 224. Archaeological Papers of the American Anthropological Association, Number 19. Shafer, Harry J. and Thomas R. Hester 1983 Ancient Maya Chert Workshops in Northern Belize. American Antiquity 48:519 543. Sheets, Payson, Christine Dixon, David Lentz, Rachel Egan, Alexandria Halmbacher, Venicia Slotten, Rocio Herrera, and Celine Lamb 2015 The Sociopolitical Economy of an Ancient Maya Village: Ceren and Its Sacbe. Latin American Antiquity 26: 341 -361. Sills, E. Cory 2016 Re-evaluating the ancient Maya salt works at Placencia Lagoon, Belize. Mexicon 38(3):69 74. Sills, E. Cory 2017 Extent of Brine Enrichment for Salt Production at the Ancient Maya Placencia Lagoon Salt Works, Belize. Research Reports in Belizean Archaeology 14:271 278. Sills, E. Cory, Heather McKillop, and E. Christian Wells 2016 Chemical signatures of ancient activities at Chan bi A submerged Maya salt works, Belize. Journal of Archaeological Sciences: Reports 9:654 662. Sierra Sosa, Thelma, Andrea Cucina, T. Douglas Price, James H. Burton, and Vera Tiesler 2014 Maya Coastal Production, Exchange, Life Style, and Population Mobility: A View from the Port of Xcambo, Yucatan, Mexico. Ancient Mesoamerica 25:221 238. Somers, Bretton and Heather McKillop 2005 Hidden Landscapes of the Ancient Maya on the South Coast of Belize: D iscovering Invisible Architecture at Arvins Landing. Research Reports in Belizean Archaeology 2: 291 300. Watson, Rachel, Heather McKillop, and E. Cory Sills 2013 Brine Enriching Slag Heaps or Mounded Remains of Salt Makers Homes? Earthen Mounds in the Mangroves at the Paynes Creek Salt Works. Research Reports in Belizean Archaeology 10:297 304. Williams, Eduardo. 2003 La Sal de La Tierra. Zamora, Michoacan: El Colegio de Michoacan. Woodfill, Brent K.S., Brian Dervin Dillon, Marc Wolf, Carlos Ave ndao, and Ronald Canter 2015 Salinas de Los Nueve Cerros, Guatemala: A Major Economic Center in the Southern Maya Lowlands. Latin American Antiquity 26 :162 179.

PAGE 264

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 275 285 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 25 ASSESSMENT OF THE SHELL MIDDEN AT THE ELEANOR BETTY SALT WORK, BELIZE Valerie Feathers and Heather McKillop The underwater site of Eleanor Betty is part of a network of Classic period (A.D. 300900) salt production sites located along the coast of southern Belize (Feathers et al. 2017; McKillop 2002, 2005a; McKillop and Sills 2017). Most of these sites lay unde r 0.5 to 1.5 m of water due to sea-level rise. An inundated shell deposit was discovered during the 2011 field season. The deposit was excavated by the authors during the 2013 field season to assess the nature of the deposit (cultural or natural) and to determine the use of the shell (ritual or dietary). The midden was located between1630 cm below the sea -floor and extended both inside and outside of the underwater wooden structure. The shells were exported to Louisiana State University for macroscopic and microscopic analyses. Approximately 3,979 fragments were identified as Crassostrea rhizophora A total of 198 minimum number of individuals (MNI) of Crassostrea rhizophora were present. Evidence for predation was determined using the Height -Length Ratio (HLR) for complete shells. Butcher marks were analyzed and assigned to classes based on their break patterns. Introduction Analyses of a large underwater marine shell deposit were carried out to evaluate the role of shell at the Eleanor Betty Site, one of the Paynes Creek Salt Works, in southern Belize (Feathers et al. 2017; McKillop 2005a, 2017a, 2017b; McKillop and Sills 2016, 2017). Shell remains frequently are recovered from coastal and inland Maya sites. However, only a few shell middens have been excavated, including Cancun (Andrews et al. 1975), Moho Cay (McKillop 1984, 2004), Frenchmans Cay (McKillop and Winemiller 2 004), Butterfly Wing (McKillop 1996, 2005b: 141), the Eleanor Betty site (Feathers et al. 2017), and five Paynes Creek Salt Works in deep water (McKillop 2017a). Analyses of the Eleanor Betty shell deposit focused on evaluating if the deposit was natural or cultural, whether there was evidence of modification as artifacts, and whether there were butchering marks indicating the shells were collected for food. Importance of Marine Shell in the Maya Area Lange (1971) hypothesized that the Classic period inl and population of the northern Yucatn and Belize was larger than what could be supported by agriculture. He argued that the exploitation of marine resources and the preservation of marine fish for inland trade increased the nutritional quality of the inland diet. Although lacking evidence of nutritional quality of the inland diet, Langes hypothesis of Figure 1 Map of Maya area with sites mentioned in the text. inland transport of seafood continues to be evaluated. The coastal Maya traded marine shells to inland communities for ritual and food ( Figure 1). Excavations at Isla Cancun revealed that thousands of shells had been butchered for subsistence purposes, including Strombas gigas

PAGE 265

The Shell Midden at Eleanor Betty Salt Work 276 (queen conch) and Melongena corona (crown conch; Andrews 1969: 57, 58). Abundant Strombas gigas shells from Tancah, a coastal site along the western coast of the Yucatn just north of Belize, suggest marine exploitation for export as well as local consumption (Miller 1977). Remains of Dinocardium r. vanhyningi (cock le shells) excavated at Dzibilchaltun were used for food (Andrews 1969: 59). Working Shell for Trade and Utilitarian Use Marine shells were modified to make jewelry, scrapers, and other items for coastal and inland use. Microwear analysis of stone tools and the presence of an incised shell in Structure 4A, indicates shell manufacturing at Pooks Hill, Belize (Stemp et al. 2010) and at Aguateca, Guatemala (Emery and Aoyama 2007; Inomata et al. 2002). At Aguateca, marine shell orna ments were made using chert flakes, bifacial thinning flakes, oval bifaces, bifacial points, and obsidian prismatic blades. Leaders of the Chan community were involved in the production of Strombus shell beads and other shell ornaments (Robin et al. 2014). Marine shell ornaments may have been produced by members of the royal family or attached specialists at Copan (Aoyama 1995). Evidence for the elite control of marine shell production and trade was found at Piedras Negras and Aguateca (Sharpe and Emery 2 015). Several thousand marine mollusks were recovered from the primary center of Aguateca, whereas only four shells were recovered from Punta de Chimino, a secondary center located downstream from Aguateca. Elite caches contained more marine mollusks tha n other caches at Aguateca and Piedras Negras, suggesting the elite controlled the distribution of marine items to the secondary centers. Evidence for nondietary use of marine shells also can be found at coastal sites. Shells were carved into tinklers an d discs for nonutilitarian purposes at the trading ports of Wild Cane Cay and Moho Cay (McKillop 2004:269, 2005b: Figure 6.32). Gastropods from Moho Cay were used as scrapers, sinkers, whorls, gorgets, beads, hammers, and as cutting tools (McKillop 1984: 3032). Queen conch remains, along with a large amount of worked and unworked shells from Ek Luum on Ambergris Caye, Belize, indicated that Ek Luum was a processing station where shells were prepared for inland trade (Shaw 1995). Few shells at Ek Luum exhibit breaks indicative of meat removal. The absence of breaks could indicate inland trade for consumption or trade of a nonmanufactured shell for ritual purposes. Worked shells were recovered from other locations on Ambergris Cay, including shells cups from Ek Luum, San Juan, and Chac Balam as well as tinklers, pendants, and beads (Garber 1995:126133, 135). Coastal Inland Maritime Trade: Ritual and Status The most common instances of coastal inland trade of marine shells occur in the archaeological record with the recovery of Spondylus shells. Elite households excavated at Tikal (Moholy Nagy 1963), Caracol (Teeter 2004), and Aguateca (Emery and Aoyama 2007; Inomata et al. 2002) revealed Spondylus shells. Over 30 marine shell species were recovered from Chichn Itz (Cobos 1989). The abundance of shell remains suggests a reliance on local shell. Excavations at Lubaantun uncovered the remains of a worked qu een conch used to produce shell disc blanks along with other marine shells (Wing 1975). Excavations at Altun Ha, Belize, revealed caches with a Spondylus disc and tubular shell beads, a Spondylus notched pendant and Oliva beads, as well as many other she ll adornments (Pendergast 1979). Remains of pendants made from Spondylus americanus and tinklers made from Oliva sayana were recovered from Dzibilchaltun (Andrews 1969 54:55). Spondylus shells were included as grave offerings in Structure10L 26 tomb at C opan (Beaubien 2004). Kidder et al. (1946:145) reported that shell formed part of the mortuary furniture in every tomb at Kaminaljuyu. Shell remains included Spondylus Oliva Olivella and Marginella shells in the form of trumpets, tinklers, and penda nts. Spondylus shells, a carved Strombus shell, and unidentified shell teeth from a jadeite mask were recovered from the burial of Yukom Yichak Kak at Calakmul (Vargas et al. 1999). Other sites where marine shell has been recovered from inland burial s

PAGE 266

Feathers and McKillop 277 included Pacbitun (Healy 1990), Caledonia (Healy et al. 1998), Batsub Cave (Prufer and Dunham 2009), Dos Pilas (Emery 2008), Buenavista del Cayo (Yaeger et al. 2015), Chan (Robin et al. 2014), and Lamanai (Pendergast 1981). Shells were used as jewelry, music, tools, or musical instruments (Chase 1981; McKillop 1984, 1996:59, 2004: 269; 2005b: Figure 6.32). Ancient Maya Shell Middens Marine shells were important to the coastal and inland ancient Maya for food, tools, and ornaments. Marine shells have been recovered from shell middens and construction fill at coastal sites, in burials and caches at coastal and inland sites, and in household refuse deposits (Andrews 1969; McKillop 1984, 2005b; McKillop and Winemiller 2004; Pendergast 1992). F ew ancient Maya shell middens have been excavated (Andrews et al. 1975; Feathers et al. 2017; McKillop 1984, 1996, 2005b:36, 37, 39, 141; McKillop and Winemiller 2004 ). Mollusks were an integral part of the coastal Maya diet. Isla Cancun, a Preclassic Maya sett lement in Quintana Roo, Yucatn, had a large midden containing a variety of marine remains, including 6,547 shells, from 99 species (Andrews 1969; Andrews et al. 1975). About 28.6% of the midden was comprised of queen conch (Andrews et al. 1975:186187). Shells recovered from Frenchmans Cay were analyzed to provide an indepth look at the diet and landscape of the Maya who occupied the site (McKillop and Winemiller 2004). Three mounds are present on Frenchmans Cay Great White Lucine, Crown Conch, and Spondylus Excavations indicated Frenchmans Cay was a Late Classic to Early Postclassic site (A.D. 600 1000; McKillop 2005b). Fifty eight genera of shell were recovered for a total of 2,785 shells with a weight of 13,528.46 g. Ninety eight species, inc luding Isognomon alatus and Crassostrea rhizophorae (mangrove oysters), and 1,315 minimum number of individuals (MNIs) were identified. Butchering for meat removal was indicated by the presence of a circular hole in the spire of queen conch shells. Almos t all species recovered were edible and likely contributed to the coastal diet (McKillop and Winemiller 2004). A large midden was discovered on the northern end of Moho Cay near Belize City (McKillop 1984). The midden contained predominately marine shells and manatee bones, but also pottery sherds, broken chert, obsidian, as well as the remains of deer, shark, and green turtle (McKillop 1984, 2004). Small holes for meat removal were found near the muscle attachment point of the Queen Conch shells recovered from the site. Marine shell at the trading port of Wild Cane Cay included abundant shells from midden deposits, as well as modified shell from burials (McKillop 2005b). An infant burial from Fighting Conch mound included a carved Spondylus shell disk (M cKillop 2005b: Figure 6.32a). A carved Melongena shell disk was found in a Late Classic burial in household middens (McKillop 2005b: Figure 4.8). Butterfly Wing is a Protoclassic (75 B.C A.D. 400) shell midden located by the mouth of the Deep River in southern Belize (McKillop 1996, 2002: 11 ; McKillop et al. 2004:349). Strombus pugilis, Crassostrea rhizpophora, and Isognomon alatus shells dominated the shell midden, all edible species. The recovery of a Strombus gigas celt indicates the ancient Maya at Butterfly Wing also were using the sea as a resource for utilitarian items (McKillop 1996:59). In an attempt to categorize marine resource use and trade, Andrews (1969:41) divided shell remains into two categories. The first concerns shells from coastal middens, which would have been used for local consumption and perhaps for meat extraction for inland trade, whereas the second focuses on shells at inland sites, which were not used for subsistence purposes, but rather for ritual events. He suggests that sites close to the coast, such as Dzibilchaltun, imported shell for ritual and as a food delicacy. The Eleanor Betty Shell Deposit An underwater shell deposit was unexpectedly discovered at the Eleanor Betty site during excavations in 2011(Aucoin 2012; Fe athers et al. 2017). Excavations were carried out in 2013 to map the extent of the shell

PAGE 267

The Shell Midden at Eleanor Betty Salt Work 278 deposit, to evaluate if the deposit was natural or cultural, and to analyze the shells for evidence of butchering for food or modification as artifacts. The sediment was expected to be alkaline and preserve bone, since shells have calcium carbonate (CaCo3). Four transects were added north and south of the 2011 Transect to define the extent of the shell deposit (Feathers et al. 2017). A total of 19 new units w ere exca vated in 10 cm levels to a depth of 30 cm below the sea floor. The shell deposit was mapped and excavated separately from the rest of the unit. Transect 6, Unit 34 m, was excavated in two centimeter levels from 16 cm to 30 cm depth due to the concentrat ion of shells. Two centimeter levels were possible because most of the shells were broken and small. The deposit was compact with little sediment. A wall of Transect 4 was cleaned to see the depth of the deposit ( Figure 2 ). The presence of charcoal, ce ramics, and briquetage above, within, and below the shell deposit indicates it was associated with the Eleanor Betty salt work s Shell was exported under permit to the Archaeology Lab at Louisiana State University (LSU) for further study. Laboratory Methods The minimum number of individuals (MNI) and the number of individual species present (NISP or fragments), weight, presence/absence of butcher marks, and height length ratio (HLR) for human predation were determined for the recovered shell, following me thods used by other researchers (Andrews et al. 1975; Claassen 1998; Kent 1988; McKillop and Winemiller 2004). The MNI assessment used a trait unique to the shell, in this case the umbone (the raised protuberance located posteriorly to the hinge on a bivalve shell; Figure 3 ) and counting the number of umbones on one side of the shell. All shells were included in the counts for NISP. Weights for the shell were obtained using a Delta Range Mettler PE 3600 electronic balance in the Archaeology Lab at LSU. A plastic container was used to hold the shell on the scale during the weighing process. Shell weight can vary according to the environmental setting. For example, a shell of one taxa found in less acidic soil has more calcium carbonate Figure 2 Schematic profile drawing of shell midden excavation. Drawing by V. Feathers. Figure 3 Photograph of Crassostrea rhizophora from the Eleanor Betty site. The circle indicates the umbone. Photo by V. Feathers. than that of the same taxa of shell from more acidic soil. Thus, the weight of the shells will be different. Shells were counted and sorted by species in the lab at LSU, in the field, and at the field base camp in Belize. Complete right and left umbones were used to obtain an MNI count. The umbones were sorted by side. Measurements of the shell height (maximum dorsal ventral) and shell length (anterior posterior) were converted into height length rations (HLR) to assess p redation and

PAGE 268

Feathers and McKillop 279 Figure 4 Measurements for Height -Length Ratio. Orange line represents height. Blue line represents length. Photo by V. Feathers. identify if the shell deposit at Eleanor Betty was a multiple or one time deposit (Gunter 1938; Kent 1988:28; Figure 4 ). Archaeologists use shell size to determine the impact of human predation and meat yield. Shell size is also determined by age and can be subs tituted for age. The height length ratios for both complete right and left valves were obtained by dividing the height by the length. Only shells with a complete umbo, ventral, anterior, and posterior margins were used. Measurements were obtained with s liding calipers. Butcher marks were identified and classified into six categories: 1) L shaped, 2) V shaped, 3) horizontal break, 4) vertical break, 5) slanted break, and 6) notched break based on Kents (1988) methods for opening oysters ( Figure 5 ). Ther e are several opening methods which result in the different break patter n s. The first method involves heating, steaming, or boiling the oyster to easily open the valves and cook the meat. This method does not leave a butcher mark on the shell. The secon d method is stabbing. A blade like object is forced between the valves along the posterior margin in order to cut the abductor muscle. A Ushaped notch usually forms parallel to the muscle scar on the margin of the shell as a result. Marks of this natur e are classified as notched (Category 6). The third method, hammering, involves striking the oyster. A hammer stone is used to lightly strike the shell valve above the abductor muscle. This impact stuns the oyster, allowing the harvester to open the v alves and retrieve the meat. A small abrasion mark is usually located on the valve as a result of hammer stone use. The cracking method involves breaking the ventral edge of the shell in order to remove the meat with a blade. A hammer stone is usually e mployed with this method. This method results in a straight break along the ventral margin. Categories 1 5 could be a result of the cracking method. Results Approximately 4,733 shell fragments weighing 2,304.24 g were recovered from the excavated units. Eighty three percent (n=3,933) of the recovered shell were associated with the shell midden. The most abundant species was Crassostrea rhizophora (n=3,979), red mangrove oysters, which form beds along the roots of red mangroves in intertidal, brackish waters beside the underwater site. Transect 6, Unit 3 4 m contained the most oysters (n=3,204). Assessment of the right and left umbones of C. rhizophora result ed in an MNI of 198 shells (198 left, 64 right). A total of 57 left valves and 43 right valves was measurable (complete; Table 1 ). Seventy of the 100 complete valves had complete dorsal ventral and anterior posterior margins and were used for HLR measurements. The average for the left and right valves was 1.89 cm and 1.87 cm, respectively. The maximum HLR was 2.84 cm for the left and 2.92 cm for the right. The minimum was 1.22 cm for the left and 1.35 cm for the right. The range was 1.62 cm for the le ft and 1.57 cm for the right. The standard deviation was 0.34 cm for the left and 0.33 cm for the right. Twentyfour of the 100 shells could be measured for length and not height. Six could be measured for height but not length. Butcher marks were prese nt for 57 of the 198 shells assessed for MNI. The most abundant break was a notch (n=17; Figure 5 Table 2 ). No hammer stone abrasions were observed. However, the outside layer of the shell was fragile and flaking, which may have obscured the hammer sto ne scar. The notch marks could have resulted from the use of a tool, such as a piece of chert, inserted parallel to the abductor muscle in order to open the shells and collect the meat. The cracking method appears to be the most employed method for openi ng

PAGE 269

The Shell Midden at Eleanor Betty Salt Work 280 Figure 5 Photos of break patterns on Eleanor Betty Crassostrea rhizophora shells based on Kent (1988). A) Notch; B) Horizontal; C) Vertical; D) V -Shape; E) Slanted; and F) L -shape. Table 1 Height -Length Ratio (HLR) Measurements for Shell from all Excavations. Height cm Length cm Left Right HLR left cm HLR right cm Total # 57 43 Minimum 2.66 1.19 1.22 1.35 Maximum 7.7 3.78 2.84 2.92 Average 4.69 2.47 1.89 1.87 Table 2 Classification of C. rhizophora break patterns from the Eleanor Betty Shell Midden. Break Pattern Number Present Processing Method L shaped 14 Cracking V shaped 6 Cracking Horizontal Line 8 Cracking Vertical Line 10 Cracking Notch 17 Stabbing Slanted 2 Cracking

PAGE 270

Feathers and McKillop 281 shells based on the vertical, horizontal, V shaped, and L shaped breaks. If these breaks are considered as one class, then this combination would result in the most abundant break pattern (n=38). Discussion The shell midden was determined to be a cultural midden for several reasons. An abundance of briquetage (3,721 g) and charcoal (6,248.48 g) was found intermixed with the deposit. No human remains and only two unidentified animal bones were recovered during excavations. However, the absence of skeletal material is not surprising since Eleanor Betty was a salt production workshop. There is no evidence the shells were used as jewelry, music, tools, or musical instruments as seen at other Maya sites (Chase 1981; McKillop 1 984, 1996:59, 2004a 269; 2005b: Figure 6.32). Approximately 55% of the recovered shell was located within the shell midden feature (Transect 6, Unit 34 m). Based on the height length ratio averages for the left and right valves, the oysters for Eleanor Betty are bed oysters (Kent 1988). The average salinity for C. rhizophora habitats is 7.2 to 28 practical salinity units (PSU). Growth can occur in areas with 0 PSU to 40 PSU (Galstoff 1964; Nascimento 1991). The underwater environment is ideal for grow th and harvesting of mangrove oysters. The Eleanor Betty site is in an intertidal, brackish area surrounded by red mangroves that have C rhizophora oysters growing on the prop roots in the water. Shell height and length were measured to determine if ther e was evidence of overharvesting. If the measurements showed a large difference in the average height length ratio, then this difference would suggest the salt makers were regularly harvesting, and possibly over harvested, the oysters. The shells would not have matured into adulthood, resulting in a difference of shell height and length throughout the midden. Alternatively, if the average HLR measurements do not differ, then human predation by the salt makers likely did not occur. The HLR measurements showed no major differences as the averages were 1.87 cm and 1.89 cm for the right and left, respectively, indicating a lack of predation by humans. Figure 6 Porosity of Crassostrea rhizophora shell due to predation by sponges, mud conchs, oyster bores, and/or barnacles from the Eleanor Betty 2013 excavations. However, several shells were porous in appearance, suggesting predation by animals such as sponges, mud conchs, oyster borers, or barna cles (Galstoff 1964; Kent 1988; Figure 6 ). Although the shell midden was small compared to other shell middens, several of the shells had butcher marks. The butcher marks indicate the shells were used for food. A hammer stone likely was used to open the shells. The pressure of the hammer stone would have broken the ventral edge of the shell and stunned the oyster (Kent 1988). The salt makers would have removed the meat for consumption at this point. Although butcher marks are clear evidence of modifi cation for subsistence, heat also may have been used to open the oysters. The shell midden may have been part of a ritual prior to the start of salt production season. The shell feature and associated ceramics lay 16 to 30 cm depth below the seafloor. Ab out 40 g of Warrie Red and Mangrove Unslipped jars were associated with the shell midden feature. The water jars likely would have stored brine. A few Belize Red sherds from serving vessels (14 g) were recovered. The associated briquetage is composed of sand temper and not shell, so shell clearly was not used as temper. No ceramics, shell, charcoal, or botanicals were encountered below 30 cm depth, indicating this deposit was the initial layer of the site. Ocarinas and serving vessels recovered at othe r Paynes Creek Salt Works indicate the occurrence of ritual activities (McKillop 2002). Rituals are performed prior to salt production season at Sacapulas (Reina and Monaghan 1981). Temple platforms, shrines, and large wooden crosses attest to the sacred

PAGE 271

The Shell Midden at Eleanor Betty Salt Work 282 nature of Emal, Yucatn, (Kepecs 2003:128). A central ball court in addition to burnt offerings of cacao was discovered near the epicenter of Salinas de los Nueve Cerros (Woodfill et al. 2015). Once the ritual was complete, the salt makers at Eleanor Bet ty could have built their hearth on top of the initial offering. Due to the shallow nature of the site (55 cm below the water table at its deepest depth on the western edge of the excavated units), the hearth (located in Units 2 3 m and 34 m of all trans ects) did not survive due to the slow inundation process. The changing tides would have spread the shell northto south and east to west, creating a shallow, elongated shell midden rather than a smaller, heaped deposit. Sea level rise has inundated the s ite and most of the other Paynes Creek Salt Works (McKillop et al. 2010). Conclusions Eleanor Betty was one of the Paynes Creek Salt Works located along the coast of southern Belize during the Classic period. The midden consisted of a single species of shell, C rhizophora (MNI=198). The recurrence of a single species underscores the interpretation that the shell midden was a singleuse event. In contrast, at the nearby trading port of Wild Cane Cay, there were more th an 45 species of shell (McKillop 2005b). At Frenchmans Cay, there were 98 species of shell (McKillop and Winemiller 2004). Butterfly Wing was dominated by several species, including C rhizophora Isognomon alatus and Spondylus pugilis (McKillop 2002). The Cancun shell midden also had a variety of shells (Andrews 1969). The Classic period midden on Moho Cay was dominated by Strombus gigas but also contained manatee and other animal bones (McKillop 1984, 2004). Single species reliance on nearby available shell at Eleanor Betty suggests the shell midden was a single meal or feast perhaps eaten as part of an opening ceremony prior to the salt production season. The presence of a few Belize Red sherds the remains of painted serving bowls tempered with volcanic ash also suggests that a small feasting event took place, as proposed for the nearby Stingray Lagoon site (McKillop 2002:95). Sea level rise and fall, in conjunction with the changing motion of the tides, likely spread the shell midden across t he site over time, obscuring the middens original greater depth and more concentrated location. Available data for the Cancun, Butterfly Wing, Frenchmans Cay, Moho Cay, and Wild Cane Cay middens indicate a heavy reliance on marine resources, supporting, in part, Langes (1971) hypothesis that marine fauna were an essential part of the Maya diet. The presence of butcher marks on 57 of the recovered shells from the Eleanor Betty shell midden indicate they were used as part of a meal, once again supporting, in part, Langes (1971) hypothesis that the coastal Maya had a heavy reliance on marine fauna. Unlike shells recovered from sites such as Aguateca, Piedras Negras, Wild Cane Cay, and Moho Cay, the shells at Eleanor Betty do not provide additional evidenc e for shell modification for musical instruments or utilitarian items. Although true for the coastal Maya, more evidence exists for the support of marine fauna as status indicators when incorporated into the inland Maya diet than as a stable dietary resou rce. Most marine shell was imported for ritual purposes and not for subsistence. Acknowledgements Support for this research was provided by the National Science Foundation under Grant No. 1026796 to H. McKillop, K. McKee, H. Roberts, and T. Winemiller u nder permits from the Belize government Institute of Archaeology. The Robert C. West Graduate Student Research Award from the Department of Geography and Anthropology, Louisiana State University, partially funded airfare for V. Feathers. We would like to thank our 2013 field crew, Dr. E. Cory Sills, Dr. Rachel Watson, and our boat captain, John Young, for their assistance with field excavations and analyses as well as their camaraderie. Additionally, we are grateful for the hospitality and support of our host family at Village Farm, Tanya Russ and John Spang. References Andrews IV, E. Wyllys 1969 The Archaeological Use and Distribution of Mollusca in the Maya Lowlands Middle American

PAGE 272

Feathers and McKillop 283 Research Institute Publication 34. Tulane University, New Orleans. Andrews IV, E. Wyllys, Michael P. Simmons, Elizabeth S. Wing, E. Wyllys Andrews V, and Joann M. Andrews 1975 Excavation of an Early Shell Midden on Isla Cancun, Quintana Roo, Mexico. In Middle American Research Institute Publication 31, edited by Margaret Harrison and Robert Wauchope, pp. 147 197. Tulane University, New Orleans. Aoyama, Kazuo 1995 Microwear Analysis in the Southeast Maya Lowlands: Two Case Studies at Copan, Honduras. Latin American Antiquity 6:129-144. Aucoin, Taylor 2012 Excavation of Transect 4 at the Eleanor Betty Site, A Submerged Classic Maya Salt Work, Belize Honors Thesis, Department of Geography and Anthropology, Louisiana State University, Baton Rouge, Louisiana. Beaubien, Harriet F. 2004 Ex cavation and Recovery of a Funerary Offering of Marine Materials from Copn. In Maya Zooarchaeology: New Directions in Method and Theory edited by Kitty F. Emery, pp. 45-54. Cotsen Institute of Archaeology. University of California, Los Angeles. Chase, Diane Z. 1981 The Maya Postclassic at Santa Rita Corozal. Archaeology 34:2533. Claassen, Cheryl 1998 Shells University Press, Cambridge. Cobos, Rafael 1989 Shelling In: Marine Mollusca at Chichn Itz. In Coastal Maya Trade edited by Heather McKil lop and Paul F. Healy, pp. 4958. Occasional Papers in Anthropology 8. Trent University, Peterborough. Emery, Kitty F. 2008 Techniques of Ancient Maya Bone Working: Evidence from a Classic Maya Deposit. Latin American Antiquity 19:204 -221. Emery, Kitty F. and Aoyama Kazuo 2007 Bone, Shell, and Lithic Evidence for Crafting in Elite Maya Households at Aguateca, Guatemala. Ancient Mesoamerica 18: 6989. Feathers, Valerie, Heather McKillop, E. Cory Sills, and Rachel Watson 2017 Excavating the Underwater Shell Deposit at an Ancient Maya Salt Work in Belize: The Eleanor Betty Site. Research Reports in Belizean Archaeology 14: 289-298. Galstoff, Paul S. 1964 The American Oyster Crassostrea viginica Gmelin Electronic Document, http://www.nefsc.noaa.gov/publications/classics/gal tsoff1964/ Garber, James F. 1995 The Artifacts. In Maya Maritime Trade, Settlement, and Populations on Ambergris Caye, Belize edited by Thomas Guderja n and James Garber, pp. 113137. Labyrinthos, Culver City, Ca. Gunter, Gordon 1938 Comments on the Shape, Growth, and Quality of the American Oyster. Science 88:546-547. Healy, Paul F. 1990 Excavations at Pacbitun Belize: Preliminary Report on the 1986 and 1987 Investigations. Journal of Field Archaeology 17:247 -262. Healy, Paul F., Jaime J. Awe, and Hermann Helmuth 1998 An Ancient Maya Multiple Burial at Caledonia, Cay District, Belize. Journal of Field Archa eology 25:261-274. Inomata, Takeshi, Daniela Triadan, Erick Ponciano, Estela Pinto, Richard E. Terry, and Markus Eberl 2002 Domestic and Political Lives of Classic Maya Elites: The Excavation of Rapidly Abandoned Structures at Aguateca, Guatemala. Latin American Antiquity 13:305 -330. Kent, Bretton W. 1988 Making Dead Oysters Talk: Techniques for Analyzing Oysters from Archaeological Sites. Maryland Historical Trust, Historic St. Marys City, Jefferson Patterson Park and Museum. Kepecs, Susan 2003 Salt Sources and Production. In The Postclassic Mesoamerican World edited by Michael E. Smith and Frances F. Berdan, pp. 126130. The University of Utah Press, Salt Lake City. Kidder, Alfred V., Jesse D. Jennings, and Edwin M. Shook 1946 Excavations at Kaminaljuyu, Guatemala Carnegie Institution of Washington Publication 561. Washington, D.C. Lange, Frederick W. 1971 Marine Resources: A Viable Subsistence Alternative for the Prehistoric Lowland Maya. American Anthropologist 73:619 639. McKillop, Heather 1984 Prehistoric Maya Reliance on Marine Resources: Analysis of a Midden from Moho Cay, Belize. Journal of Field Archaeology 11:25 35. 1996 Ancient Maya Trading Ports and the Integration of Long -Distance and Regional

PAGE 273

The Shell Midden at Eleanor Betty Salt Work 284 Economies: Wild Cane Cay in south-c oastal Belize. Ancient Mesoamerica 7:4962. 2002 Salt: White Gold of the Maya University Press of Florida, Gainesville. 2004 The Classic Maya Trading Port of Moho Cay. In The Ancient Maya of the Belize Valley, edited by James F. Garber, pp. 257272. University Press of Florida, Gainesville. 2005a Finds in Belize Document Late Classic Maya Salt Making and Canoe Transport. Proceedings of the National Academy of Sciences 102:5630 -5634. 2005b In Search of Maya Sea Traders Texas A&M University Press, College Station. 2017a Diving Deeper in Punta Ycacos Lagoon at the Paynes Creek Salt Works, Belize. Research Reports in Belizean Archaeology 14: 279-288. 2017b Early Maya Navigation and Maritime Connections in Mesoamerica. In The Sea in History: The Medieval World edited by Michel Balard, pp. 701715. The Boydel Press, Woodbridge, England. McKillop, Heather and E. Cory Sills 2016 Spatial Patterning of Salt Production and Wooden Buildings Evaluated by Underwater Excavations at P aynes Creek Salt Work 74. Research Reports in Belizean Archaeology 13:229 237. 2017 The Paynes Creek Salt Works, Belize: A Model for Ancient Maya Salt Production. In The Value of Things edited by Jennifer P. Matthews and Thomas H. Guderjan, pp. 67 -86. University of Arizo na Press, Tucson. McKillop, Heather, E. Cory Sills, and Jessica Harrison 2010 Sea-Level and Inundation of the Classic Maya Paynes Creek Salt Works: Evidence from Marine Sediment. Research Reports in Belizean Archaeology 7:245252. McKillop, Heather and Terance Winemiller 2004 Ancient Maya Environment, Settlement, and Diet: Quantitative and GIS Spatial Analysis of Shell from Frenchmans Cay, Belize. In Maya Zooarchaeology: New Directions in Method and Theory edited by Kitty F. Emer y, pp. 57 -80. Cotsen Instituteof Archaeology. University of California, Los Angeles. Miller, Arthur G. 1977 The Maya and the Sea: Trade and Cult at Tancah and Tulum, Quintana Roo, Mexico. In The Sea in the Pre -Columbian World edited by Elizabeth P. Ben son, pp. 97140. Trustees for Harvard University, Washington D.C. Moholy -Nagy, Hattula 1963 Shells and other marine material from Tikal. Estudios de Cultura Maya 3:65 83. Nascimento, Iracema A. 1991 Crassostrea rhizophorae (Guilding) and C. brasiliana (Lamarck) in South and Central America. In Estuarine and Marine Bivalve Mollusk Culture edited by Winston Menzel, pp. 125134. CRC Press, Boca Raton. Pendergast, David M. 1979 Excavations at Altun Ha, Belize, 1964 1970, Volume 1. Royal Ontario Museum, Toronto. 1981 Lamanai, Belize: Summary of Excavation Results, 19741980. Journal of Field Archaeology 8:29 -53. 1992 Noblesse Oblige: The Elites of Altun Ha and Lamanai, Belize. In Mesoamerican Elites: An Archaeological Assessment edited by Diane Z. Chase and Arlen F. Chase, pp. 61 -79. University of Oklahoma Press, Norman. Prufer, Keith M. and Peter S. Dunham 2009 A shamans burial from an Early Classic cave in the Maya Mountains of Belize, Central America. World Archaeology 41:295 -320. Reina, Ruben E., and John Monaghan 1981 The Ways of the Maya: Salt Production in Sacapulas, Guatemala. Expedition 23(3):1333. Robin, Cynthia, Laura Kosakowsky, Angela Keller, and James Meierhoff 2014 Leaders, farmers, and Crafters: The Relationship between Leading Households and Households across the Chan Community. Ancient Mesoamerica 25:371-387. Sharpe, Ashley E. and Kitty F. Emery 2015 Differential animal use within three Late Classic Maya states; Implications for politics and trade. Journal of Anthropological Archaeology 40:280-301. Shaw, Leslie C. 1995 Analysis of Faunal Material from Ek Luum. In Maya Maritime Trade, Settlement, and Populations on Ambergris Caye, Belize edited by Thomas Guderjan and James Garber, pp. 175181. Labyrinthos City, California. Stemp, W. James, Christophe G.B. Helmke, and Jaime J. Awe 2010 Evidence for Maya Household Subsistence and Domestic Activities: Use -Wear Analysis of the Chipped Chert Assemblage from Pooks Hill, Belize. Journal of Field Ar chaeology 35:217-234.

PAGE 274

Feathers and McKillop 285 Teeter, Wendy G. 2004 Animal Use in a Growing City: Vertebrate Exploitation at Caracol, Belize. In Maya Zooarchaeology: New Directions in Method and Theory edited by Kitty F. Emery, pp. 177-191. Cotsen Institute of Archaeology. U niversity of California, Los Angeles. Vargas, Ramn, Sylviane Boucher, Paula Alvarez Gonzlez, Vera Tiesler Blos, Valeria Garca Vierna, Renata Garca Moreno, and Javier Vzquez Negrete 1999 A Dynastic Tomb from Campeche, Mexico: New Evidence on Jaguar P aw, a Ruler of Calakmul. Latin American Antiquity 10:47 -58. Wing, Elizabeth S. 1975 Animal Remains from Lubaantun. In Lubaantun: A Classic Maya Realm by Norman Hammond, pp. 379382. Monographs of the Peabody Museum of Archaeology and Ethnology No. 2. Harvard University: Cambridge, Massachusetts. Woodfill, Brent K.S., Brian Dervin Dillon, Marc Wolf, Carlos Avendao, and Ronald Canter 2015 Salinas de los Nueve Cerros, Guatemala: A Major Economic Center in the Southern Maya Lowlands. Latin American Anti quity 26:162 179. Y aeger, Jason, M. Katherine Brown, Christophe Helmke, Marc Zender, Bernadette Cap, Christie Kokel Rodriquez, and Sylvia Batty 2015 Two early Classic Elite Burials from Buenavista del Cayo, Belize. Research Reports in Belizean Archaeology 12: 181 -191.

PAGE 275

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 287 295 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 26 KING KAK [U TI?] CHAN KAWIL OF PUSILH A: AN ANCIENT MAYA KING Andrew D. Somerville Christian M. Prager and Geoffrey E. Braswell It is often assumed that the individual is invisible in archaeological research, but it is possible to construct person -centered accounts of past lived experiences. In this paper, we adopt a developmental approach to sketch out the life history of a single individual who ruled Pusilha, Toledo District, Belize, during the middle of the eighth century. We call him Ru ler G, but he probably was named Kak [U Ti] Chan Kawil Kul Un Ajaw. We adopt a conjunctive approach using epigraphic, biological, and archaeological data. We examine his childhood, adulthood, death, and afterlife to better understand the role of Pusi lha in the ancient Maya world. Introduction Pusilha was an important city for the relatively brief period of AD 574 until around AD 751 (Prager et al. 2014). Our excavations, conducted between 2002 and 2005 (Braswell et al. 2004, 2005), uncovered the remains of 22 individuals within and among the structur es of the Gateway Hill Acropolis ( Figure 1 ), the elite administrative and ceremonial center of the city (Pitcavage and Braswell 2009). The wide range in the quality and quantity of goods included in the burials implies great differences in social status. Of all the interments, Burial 8/4 stands out as unique due to the prominent size and location of the tomb, which was found within the largest freestanding platform of the city and at its highest point (Braswell and Gibbs 2006). Moreover, the tomb is one of the richest burials excavated in southern Belize. The stylistic dating of ceramics, the recent translations of hieroglyphic texts from the site, and bioarchaeological studies together suggest that Burial 8/4 probably contained the remains of the king n icknamed Ruler G, one of the last ajaw of Pusilha. If it is his tomb, as we think, this is the first royal burial of a Maya king of Belize whose exploits are known from hieroglyphic texts. To gain a fuller picture of what life was like for the Maya eli te, we situate Ruler G within different scale contexts (Bronfenbrenner 2005), which were relevant to individual development in southern Belize during the Late Classic period. Childhood of Ruler G The only known monument erected during the reign of Ruler G is Stela E ( Figure 2 ), whose front portrays him with two seated captives. The Figure 1 The Gateway Hill Acropolis, Pusilha, showing the location of excavated burials, including the royal tomb, Burial 8/4 (mapped by Braswell and Susan Maguire). reverse of the monument is better preserved and includes a hieroglyphic account of his ancestry. Sadly, his name can only be partially read. Nonetheless, it includes both KAK and CHAN, perhaps with an infixed KAWIL All three are exceedingly common e lements in royal names and titles throughout the Maya lowlands. The first ruler of Pusilha was named Kawil Chan Kinich (Heavenly Radiant Kawil) and the second was called Kak U Ti Chan (Fire is the Mouth of Heaven), a name he shared with the cont emporary 11th Ruler of Copan. Ruler

PAGE 276

An Ancient Maya King 288 Cs name ends with that of his predecessor. Ruler Ds name also contains Chan, and Ruler Es complicated title sequence contains the phrase Kak U Kawil Thus, what we can read of Ruler Gs name and titles is cons istent with those of his predecessors, and it is entirely possible that the missing portion is U Ti We therefore speculate that his full name and title was Kak [ U Ti ?] Chan Kawil kuhul Un Ajaw Fire is the Mouth of the Heavenly Kawil Divine Lord of the Avocado [or Pusilha] (Prager et al. 2014). Stela E also names the mother of Ruler G and identifies her as an ajaw or divine ruler in her own right, whom we call Ruler F. This is a rarity in the Maya world and may indicate that she as cended to the throne during a crisis of succession. Ruler Gs father was a foreign lord who married into the ruling family. Ruler Gs paternal grandfather was called Junew Chak Muyal Chan Yopat Kak Ti Kawil, which includes the theonym Chan Yopat, bes t known from Copan and Quirigua, but also from Naranjo. Just south of the pyramid in which Ruler Gs remains were found, we excavated Late Classic Burial 3/1, which contained a man whose strontium and oxygen isotope signatures are consistent with Copan, a ccompanied by a reliquary toothcache of a second individual also from there. It is purely speculative, but this could be the burial of Ruler Gs non royal father and his own ancestor, both from the Copan region. It is unusual that Ruler Gs father had f oreign roots. Most typically, the Maya practiced female hypogamy and neolocality. In this case, a lord from the Copan region seems to have moved to Pusilha to marry a powerful local woman who would be or already was a queen in her own right. As the son o f a local female ruler and a foreign lord, Ruler G grew up in a position of privilege. This is reflected in his physical remains. Isotopic studies of strontium, oxygen, and carbon from Ruler Gs second molar provide information about his life between the ages of about two and seven years. Strontium and oxygen isotope ratios demonstrate that Ruler G spent his childhood at Pusilha. Stable carbon isotope ratios, which reflect the proportion of maize in the diet, also were obtained from the second molar. T hese results imply that the Figure 2 Stela E: (a) front, 2016 photo courtesy of Bruce Love; (b) reverse, copyright The Trustees of the British Museum. childhood diet of Ruler G included exceptional amounts of maize, with perhaps upwards of 85% of his calories coming from maize, maize fed animals, and maizederived products. This was the third highest value of the sample from Pusilha, indicating that the childhoo d diet of Ruler G was even more maize focused than many of his elite peers, who also consumed high maize diets (Somerville et al. 2016). Of the 18 analyzed skeletons from Pusilha, five (28%) display some evidence of cranial modification, a cultural practice that may be closely tied to concepts of identity. Analysis of Ruler Gs cranial bones show no evidence of modification, which indicates that this was not a requirement for a potential king. Ruler G also displays no evidence of linear enamel hypoplasia, which is typical of the population. Just two of 18 individuals, or 11% of the sample, show this sign of childhood disease or poor diet. No evidence of porotic hyperostosis, a sign of anemia, was observed on Ruler Gs skull, supporting the notion of good health in childhood (Pitcavage and Braswell 2009). In sum, epigraphic and bioarchaeological data suggest that Ruler G was the child of a queen of Pusilha and a foreign lord, and that he

PAGE 277

Somerville, Prager, and Braswell 289 spent his childhood years at the site. At the microsystem level, his family and servants fed Ruler G meals that included large amounts of maize based foods. The diet appears to have been nutritious and reliable. Exosystemic factors also influenced Ruler Gs childhood. He was born during a long period of political crisis and instability at Pusilha that began after AD 672 and continued for 40 years. In AD 711, a founder of a new dynastic line and holder of the ochkin kalomte title, became king (Prager et al. 2014). But within the next 20 years, the female Ruler F whose parentage is unknownand her son, Ruler G, were both inaugurated as ajaw of Pusilha. Thus, Ruler G assumed power during a long period characterized by multiple crises of succession. This uncertainty could have created a potentially unstable political en vironment during his earliest years. Cultural and social factors common across the Maya world and Mesoamerica more broadly shaped certain aspects of Ruler Gs development. Cranial modification was widespread in the Maya world as a means of embodiment and socialization, and the decision not to modify Ruler Gs head may have been influenced by a number of factors operating at the societal level. Such factors could have included onomastic features of the childs name, traditions associated with a particular birth date, and ritual requirements of the parents. Duncan and Hofling (2011) argue that cranial modification was closely tied to rituals designed to protect infants from harm, particularly soul loss and evil winds. The absence of modification may mean t hat a child was not viewed as in need of such special protection. It is intriguing to note that the foreigner in Burial 3/1 who grew up in the Copan region also lacks evidence of cranial modification. If our speculation that this is the burial of the father of Ruler G is correct, it may be that their lack of cranial modification reflects a shared identity. Adulthood of Ruler G We lack detailed epigraphic data concerning the adult years of Ruler G, but we can place his accession in the 20 year window before his dedication of Stela E in AD 731. The text of Stela E states that Ruler G performed a stonebinding ritual, which Stua rt (1996) interprets as a ceremony in which stelae would be wrapped in cloth. This was performed to mark the end of the katun The two sitting and bound prisoners on the front of the monument suggest that Ruler G was a warrior who took captives, another common practice of Maya kings. As the last certain ajaw of Pusilha who used the emblem and whose name we can read, Ruler G may have lived during a politically tumultuous period. Nonetheless, the depiction may be regal propaganda or evidence that a vassal from another site presented prisoners to Ruler G. Many ancient Maya people underwent dental modification as young adults, perhaps as a rite of passage. Modification would have served throughout adulthood as a symbol of identity and perhaps status. All four of Ruler Gs upper incisors were filed sometime after his childhood on both the lingual and buccal sides, creating a sharp ridge running mesiodistally along the midline of the tooth (Pitcavage and Braswell 2009). This form is Romeros Type A 4. Odd ly, Ruler G does not have inlayed teeth. Such inlays of jade or pyrite were commonly encountered at Pusilha. Additional data on Ruler Gs adult years can be derived from his teeth and oral health. Four of Ruler Gs teeth (26.7 %) had dental caries (Pitcavage and Braswell 2009). All exhibited mild to moderate amounts of dental calculus. Although multiple factors contribute to tartar formation, diets high in protein and carbohydrates often correlate with higher calculus levels. Privileged access to maize and animal products, therefore, is consistent with the calculus and isotope data obtained from the remains of Ruler G. Arthritic lipping was observed on two of his cervical vertebrae, evidenced by the presence of osteophytes around the circumference of b oth the superior and inferior aspects of the vertebral bodies. His age at death cannot be precisely determined either from the hieroglyphic texts or his remains, but arthritis and the condition of his teeth are consistent with old age. At the broadest cul tural scale, social and ideological norms influenced many of the details of Ruler Gs daily life. The stylistic choice of A 4 dental modification, for instance, shows engagement with a broader Maya trend of

PAGE 278

An Ancient Maya King 290 Figure 3 Plan of Burial 8/4, Pusilha, the tomb of Ruler G (drawing by Braswell). aesthetic modifications of the teeth. Curiously, Williams and White found that Type A 4 is a very rare form with only two out of 922 individuals (0.2%) exhibiting this type of modification. At Pusilha, Type E 1 modification, with inlays of jadeite and pyrite, was the most common form of dental modification (30%; Pitcavage and Braswell 2009). The dental modification of Ruler G, then, was exceptional both within his home city and across the Maya world. This rare form of dental modification may have signaled his privileged status as an ajaw Death of Ruler G Ruler G died sometime before the katun ending in AD 751 (Prager et al. 2014). We believe Burial 8/4 is his tomb ( Figure 3 ). This identification is based on several factors. First, the presence of royal diadems and the central location of the tomb in the highest pyramid of the royal acropoli s imply that the occupant was a king. Second, pottery dates to the middle of the eighth century AD. Third, the individual is an elderly male and not a female, and hence, cannot be Ruler F. Although we cannot completely rule out individuals who are named as kings on undated monuments, Ruler G is the best fit (Braswell and Gibbs 2006; Prager et al. 2014). The tomb of Ruler G is by far the most elaborate at Pusilha, and one of the richest known from southern Belize. It was located at the top of the tallest free standing structure at the site, itself on top a series of terraces that rise some 79 m above the Machaca River and the ancient Maya bridge. The location of the tomb at the highest point of Pusilha served to remind people of the ongoing presence and legacy of this important ajaw His monument, Stela E, was located across the river in the center of the city and in the middle of a line of stelae erected by his predecessors. This too, would have maintained his memory and situated him literally at the c enter of the dynasty. Examination of his remains tells us nothing about the circumstances surrounding Ruler Gs death. There are no other signs of disease or trauma. In contrast, Burial 8/4 tells us much about the preparation of Ruler G for the afterlife Outside of the tomb, we found several broken cache vessels suggesting the presence of dedicatory offerings placed in the pyramid fill. Nearby, we also found an anthropomorphic chert eccentric. This form of eccentric is well documented in the Maya regi on, and a second

PAGE 279

Somerville, Prager, and Braswell 291 Figure 4 Large chert eccentrics: (a) Pusilha, found outside Burial 8/4; (b) Nim li Punit, found inside Structure 7, Tomb 5 (photographs by Braswell). more elaborate version was recently recovered in Tomb 5 of Nim li Punit ( Figure 4 ; Braswell 2017; Borrero et al. 2016). The stone tomb was accessed from the south and was closed by large capstones rather than a vault. The tomb was opened before we excavated it, probably in antiquity. There is ample evidence that objects and bones wer e disturbed. Ruler G was the only body within Burial 8/4. He was placed in an extended, supine position with his head to the north. This burial position is common in southern Belize and is the opposite of the Belize Valley pattern of head in the south, prone burials. His face was turned to the east. For the ancient Maya, the north was the direction associated with the sun at its zenith and the heavens, and the east was associated with the rising sun and, hence, Figure 5 Spondylus shell found over face of Ruler G and obsidian eccentrics from Burial 8/4 (photographs by Braswell). rebirth. It may be that Ruler G as well as others at Pusilha was buried in this position to reflect a belief in resurrection and afterlife. Ruler Gs face was covered with two Spondylus shells ( Figure 5 ). Many simple burials in southern Belize have a stone slab covering the cranium; it may be that the head of the deceased was viewed as needing protection. Above the head, we found three lunate eccentrics of obsidian and chert. A large trilobe obsidian eccentric was recovered just below waist level. All three obsidian artifacts are visually consistent with Mexican sources and seem to be imports. A jade earflare was found next to the head on the east side and Ruler G wore a small necklace or choker made of dozens of small jadeite beads. Four pyrite tesserae were recovered, as was a single pierced pearl found near the chest of Ruler G. His body was not coated with cinnabar. A total of 197 jadeite or other greenstone items were found in Burial 8/4. Most of these, including 43 large to medium size jade beads, were found within a crude cache vessel to the northwest of Ruler Gs head. These probably formed a necklace. Sixty fi ve small, flat pieces of jade, two small curved pieces, and other

PAGE 280

An Ancient Maya King 292 Figure 6 Three Jade diadems from Burial 8/4. Note that the bottom example is a single, double -sided diadem (photographs by Braswell). carved jadeite elements may have formed a small mosaic, perhaps a mask. A total of eight additional fragments of jadeite earflares also may have been part of this posited mosaic, or, alternatively, portions of earflares left for Ruler G to wear in the afterlife. All were found in such a jumbled condit ion within the cache vessel that there is no hope that they can be reconstructed. Indeed, it is possible that these assorted pieces were repurposed simply as valuable raw material to be placed in the tomb. Three jade artifacts are of particular importance They are the diadems worn by a king ( Figure 6 ). Two examples, which are single sided, were found in the cache vessel, as was a small fragment of the doublesided pendant. Most of the double sided pendant, however, was found against the east wall of the tomb. This suggests that the tomb was disturbed in antiquity. Thirteen ceramic vessels were found along the eastern side of the tomb, but nothing was placed on the western wall. Interestingly, some sherds of a tripod polychrome plate were found scatter ed to the west of Ruler Gs head, but most of the vessel was placed against the eastern wall. This provides further evidence of tomb entry in antiquity. There is an interesting pattern in the arrangement of the 13 vessels. There are two clusters on the e astern wall: seven in the north and six in the south (Figure 3, shown in orange and brown). With one exception, all of those in the northeastern cluster are brightly colored polychrome plates, dishes, and cylinders. All of those in the southwestern clust er are monochrome polished black/brown cylinders. We suggest that the number and arrangement of the vessels is deliberate. It may be that the vessels represent the 13 divisions of the heavens. They are placed in the east because this is the direction of rebirth, associated with the rising sun. The arrangement of polychrome vessels in the north and dark vessels in the south suggests their association with day versus night, and the heavens versus the underworld. A total of 766 obsidian blades, flakes, and chunks were found throughout the tomb. Recent XRF analysis sourced 764 pieces to El Chayal, one to Ixtepeque, and one was too small to assign to a source. These artifacts may have been placed above the capstones, only to have fallen inwards when the tom b was opened in antiquity. Such deposits are well documented, if poorly understood, at sites such as Tikal. In s um, the nature and contents of Ruler Gs tomb place him in the local cultural system of Pusilha and, more broadly, within the Maya world. His body was positioned in a manner that is common at Pusilha and southern Belize, but different from sites in other areas, especially the Belize Valley. The use of capstones rather than a vaulted roof also links the tomb to local practices; there are no May a vaults in southern Belize. Nonetheless, the arrangement of his body and the ceramic offerings that accompanied him may have broader cosmological significances that tie him to religious practice throughout the Maya region. The composition of the tomb suggests that north was associated with the heavens and light, that the south was associated with darkness and night, and that the east was the direction of rebirth. Moreover, the practice of sealing the top of the tomb with obsidian is related to Peten burial practices. The presence of a great quantity of jade is consistent with the burial

PAGE 281

Somerville, Prager, and Braswell 293 Figure 7 Evidence of looting activities, 1992 2005: (a) Burial 8/4 during excavation, note greatly disturbed surface surrounding tomb (photograph by Braswell); (b) scoops of pottery piled on side of tomb by a looter in 2005 (photographs by Sherry A. Gibbs). practices of kings throughout the Maya region. The three obsidian eccentrics found within the tomb point even further afield. The lunates and, especially, the trilobe form are common at distant Teotihuacan, and the obsidian appears to be imported from a source in central Mexico. Finally, although not present in the tomb itself, a figurine head wearing the goggles of the Teotihuacan storm god was recovered several meters away in the fill of the structure. Ruler G, then, may have had indire ct exchange or ideological ties with distant central Mexico. Conclusion: The Afterlife o f Ruler G After Ruler G died, Pusilha again seems to have entered a political crisis. The final stela of the city (Stela F), erected in AD 751, bears the name Kak K alaw but he does not employ an emblem glyph. By the end of the eighth century, most of the city was abandoned. The outlying group called Moho Plaza was occupied during the Terminal Classic period, and a hieroglyphic stair there the only one ever discovered in Belize dates to AD 798. Burial 8/4 was opened at some point in antiquity. It was entered from the top. Giant capstones were found shifted sideways and left on the partially collapsed walls of the tomb. We are not certain why the tomb was opened. The cache vessel full of jade was left in place, but part of one of the three diadems which already may have been broken was moved to the opposite side of the tomb. Broken pottery was moved. Some skeletal items also were disturbed. Perhaps the tomb was opened to remove bones to be used as reliquaries; we found no identifiable elements from the lower half of Ruler Gs body. But if this was the case, why was the tomb left improperly sealed, and why were no new offerings left? The next disturbance of Ruler Gs tomb took place between 1992 and 2001 ( Figure 7a ). At that time, inhabitants of San Benito Poite, a village that straddles the central portion of

PAGE 282

An Ancient Maya King 294 Pusilha, began extensive looting of the site. Virtually every mou nd greater than one meter in height has been trenched. During this activity, several royal burials in the northernmost structure at the top of the Gateway Hill Acropolis were looted. Locals search only for jade and have many stories about muecas like th e three diadems discovered with Ruler G. In addition to destroying some of the most important royal burials at Pusilha, looters also came very close to discovering the tomb of Ruler G. We chose that pyramid for excavation because it was so thoroughly loo ted that our work was a salvage operation. When we finally cleared six trenches and a tunnel at the top, we realized that a major tomb had been mis sed by looters by a matter of centimeters. The discovery was not without problems. The day it became clear a royal tomb had been found, we placed two armed guards there. At 3 A.M. and after a terrific storm, one of them ran to the village to tell us that the tomb had been looted. It appears that only one person entered the tomb, and he had no hand tools. He scooped out pottery along the eastern wall and left it in a line of piles along the tomb wall, directly above from where they were taken ( Figure 7b ). We do not know what, if anything, he took from the tomb. But the discoveries made the next day including the cache vessel filled with jade, the eccentrics, the remains of Ruler G, and additional undisturbed vessels in the southeastern corner of the tombsuggest to us it was very little. This was a sad intrusion in the life history of Ruler G, and we cannot know what knowledge, if any, was lost. Many local people are now learning that the past of Toledo District is their own history. Nonetheless, a few still see only the economic potential of looting. In this regard, the Qeqchi are no different from othe r impoverished and disenfranchised people around the world. The remains of Ruler G were excavated and he currently reposes at the Belize Institute of Archaeology in Belmopan. Select jade items from his burial, including two of his royal diadems, are now displayed in Belize City. The excavation of Burial 8/4 was a major discovery in the archaeology of Belize. If our identification is correct, this is the first Maya king in the country whose tomb has been excavated and whose exploits are described in carved hieroglyphs. No such kingly tomb has ever been identified at Caracol. Although not as powerful as Kinich Janaab Pakal of Palenque, Jasaw Chan Kawil of Tikal, or Kinich Yax Kuk Mo of Copan, Kak [U Ti] Chan Kawil of Pusilha occupies a similar role in modern Belize as the first known ancient ruler of this nation whose remains have been identified. Thousands of Belizeans and foreign tourists now see objects from his royal tomb and learn about his life history each year at the Museum of Belize. We hope that these new interactions with Ruler G an ongoing dialogue with the dead will positively impact the life histories of modern Belizeans. Acknowledgements The Pusilha Archaeological Project was co directed by Geoffrey E. Brasw ell, Cassandra R. Bill, and Christian M. Prager. The project was generously funded by the National Science Foundation Archaeology and International Research Fellowship Programs (SBE 0215068 and INT 02 2581), the Wenner Gren Post PhD Program (Gr. 6848), the Foundation for the Advancement of Mesoamerican Studies, Inc. (Gr. 00029), two awards from the School of American Research, and the Faculty Senate of the University of California, San Diego. All these awards were granted to Braswell. Participants in the 2005 field season included Beniamino Volta, Pierre Robert Colas, Brittany Frazier, and Sherry Gibbs (field director). We thank Dr. Jaime Awe and Dr. John Morris of the Institute of Archaeology, NICH, for their support in 2005 and throughout all our work in Toledo. Human remains from Burial 8/4 were initially studied by Megan Pitcavage and later by Andrew Somerville. Cassandra Bill conducted the ceramic analysis and Edwin Barnes illustrated the pottery and lithic artifacts from Burial 8/4. Christian Pra ger analyzed and drew all the texts from Pusilha. References Borrero, Mario, Maya Azarova, and Geoffrey E. Braswell 2016 In the Palace of the Wind God: The Discovery of the Nim li Punit Wind Jewel. Research Reports in Belizean Archaeology 13:193 206.

PAGE 283

Somerville, Prager, and Braswell 295 Braswell, Geoffrey E. 2017 Recent Discoveries in the Classic Maya Palace Complex of Nim li Punit, Belize. Journal of Field Archaeology 42(2):69-81. Braswell, Geoffrey E., and Sherry A. Gibbs 2006 In the Land of the Avocado: Recent Archaeological Investigations at Pusilha, Toledo District, Belize. Research Reports in Belizean Archaeology 3:271286. Braswell, Geoffrey E., Christian M. Prager, and Cassandra R. Bill 2005 The Kingdom of the Avocado: Re cent Investigations at Pusilha, A Classic Maya City of Southern Belize. Anthropological Notebooks 11:5986. Braswell, Geoffrey E., Christian M. Prager, Cassandra R. Bill, Sonja A. Schwake, and Jennifer B. Braswell 2004 The Rise of Secondary States in the Southeastern Periphery of the Maya World. Ancient Mesoamerica 15:219-233. Bronfenbrenner, Urie 2005 Making Human Beings Human: Bioecological Perspectives on Human Development Sage, Thousand Oaks, California. Duncan, William N., and Charles A. Hofling 2 011 Why the Head? Cranial Modification and Ensoulment among the Maya. Ancient Mesoamerica 22:199-210. Pitcavagae, Megan R., and Geoffrey E. Braswell 2009 Entierros mltiples en el Reino del Aguacate: Evidencia indirecta del sacrificio humano durante el pe riodo clsico tardo y terminal. XXII Simposio de investigaciones arqueolgicas en Guatemala, edited by Juan Pedro Laporte, Brbara Arroyo, and Hector Meja, pp. 803 -809. Museo Nacional de Arqueologa y Etnologa, Guatemala. Prager, Christian M., Beniamino Volta and Geoffrey E. Braswell 2014 The Dynastic History and Archaeology of Pusilha, Belize. In The Maya and Their Central American Neighbors: Settlement Patterns, Architecture, Hieroglyphic Texts, and Ceramics edited by Geoffrey E. Braswell, pp. 245-307. Routledge, Oxon and New York. Somerville, Andrew D., Margaret J. Schoeninger, and Geoffrey E. Braswell 2016 Political Alliance, Residential Mobility, and Diet at the Ancient Maya City of Pusilha, Belize. Jour nal of Anthropological Archaeology 41:147 158. Stuart, David 1996 Kings of Stone: A Consideration of Stelae in Ancient Maya Ritual and Representation. RES: Anthropology and Aesthetics 29/30:148 171.

PAGE 284

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 297 306 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 27 SOME LESSONS CANT BE TAUGHT, THEY SIMP LY HAVE TO BE LEARNED: EXPERIENCES FROM THREE SEASONS OF INVESTIGATIONS AT ALABAMA, STANN CREEK DISTRICT BELIZE Meaghan M. Peuramaki Brown Shawn G. Morton, Cristina Oliveira From 2014 to 2015, the Stann Creek Regional Archaeology Project (SCRAP) completed a preliminary survey of the ancient Maya site of Alabama in the southern reaches of the Stann Creek District, and initiated a program of settlement testing in 20 16. Located in the material culture sub-r egion of East -Central Belize, Alabama appeared relatively rapidly during the late facet of the Late Classic to Terminal Classic periods (ca. 700-900 CE). The two phases of SCRAP research thus far have helped to reinforce lessons learned from previous rese arch in the region, as well as introduce new lessons regarding the nature of ancient Maya material remains in East -Central Belize and how to pursue their archaeological recovery. This paper presents the lessons SCRAP members have learned ranging from issu es dealing with the adoption of old maps and excavation notes, effaced earthen -core architecture, granite as construction materials, poor pottery preservation, etc. and how they will help to shape and direct future investigations. This paper is dedicated to the late Mr. Gonzalo Choc (Figure 1), who was a much-loved member of the Stann Creek Regional Archaeology Project from 2014 to 2016. Introduction As the title quote by Picoult (2007:74) suggests, this paper examines eight lessons learned by, or reinforced for, members of the Stann Creek Regional Archaeology Project (SCRAP) during the first three years of investigations at the ancient Maya site of Alabama in the southern end of the Stann Creek District of Belize; in particular, we are sharing those lessons we had previously been taught in one form or another but had to experience firsthand in order to truly appreciate. The version of the paper presented at the 2017 BAS was inten tionally lighthearted. Our aim, here, is to present these lessons in a somewhat more formal context, as they might prove useful in the work of other archaeologists, students, teachers, tour guides, etc. For many, these lessons are not new, but they are ones that we should be reminded of from time to time, as a form of self reflection as practitioners of Maya archaeology. Each lesson starts with the original quote(s) provided by a project member, and the lesson is discussed as it relates to our ongoing w ork in the Stann Creek District Lesson 1: There are Significant Maya Archaeological Sites in the Stann Creek District I ts interesting that there are certain areas of the country that are underrepresented and understudied. I never fully realized some of the potential bias in what we learn almost exclusively coming from major centers in the heartland, and how important it is to look at smaller settlements and surrounding habitation ( SCRAP Team Member #1, personal communication, 2017). The presence of ancient Maya archaeological sites and material culture in the Stann Creek District typically comes as a surprise to many people, despite an entire book

PAGE 285

Experiences from Three Seasons of Investigations at Alabama 298 Figure 2 Map of East -Central Belize (Stann Creek District) showing location of known major inland and coastal Maya archaeological sites. Compiled by M. Peuramaki -Brown. written on the subject by Elizabeth Graham (1994; see also Graham 1976, 1978, 1982, 1986, 1987, 1989, 2001), who is considered the progenitor of Stann Creek District archaeology. Public or professional, many would be hardpressed to name a single preColumbian site in the region. Prior to her extensive survey and testing in the 1 970s, little archaeological work had been conducted in the district, limited to the inland sites of Pomona (Kidder and Ekholm 1951; MacKie 1985), Kendal (Gann 1918; Price 1899), and Pearce (Joyce 1931), and some of the offshore cays (Mitchell Hedges 1931). The known sites of the district can be divided into two broad categories (Figure 2) : the inland sites, where the majority of settlement and civic ceremonial life was focused among the broadleaf forests of the alluvial valleys, and the coastal sites where specialized activities took place, such as limemaking in the Early Classic, saltmaking in the Late Classic, and waystations along the coastal sea trade route during the Postclassic (MacKinnon 1986, 1989a, 1990; MacKinnon and Kepecs 1989; MacKinnon and May 1990; Sills 2016; see also many of the aforementioned publications by Graham). In terms of sites open to the public, there is but one: The Mayflower Sites (Mayflower, Maintzunun, Tau Witz) located in the Mayflower Bocawina National Park (Stomper et al. 2004). Many of the sites in the northern half of the district, including Pomona and Kendal, have their origins in the Preclassic with occupation extending into the Postclassic (Graham 1985), while sites in the southern half appear limited to occupation spanning from the Late Classic to Early Postclassic, such as Pearce and Alabama (Dunham et al. 1995; MacKinnon et al. 1993). For three seasons, SCRAP has been working at the small major centre of Alabama, and todate, occupation of the site appears to have dramatically increased in the late facet of the Late Classic, perhaps even originating at this time, and extended into the e arly facet of the Early Postclassic (Peuramaki Brown 2016, 2017; Peuramaki Brown et al. 2017). From 20142015 we concentrated on resurveying the monumental core of the site and producing the first systematic settlement survey in all of the Stann Creek District. In 2016, we then initiated testing of settlement sites within a naturally

PAGE 286

Peuramaki -Brown et al. 299 bounded area of the settlement zone in order to develop a sense of the architecture outside of the monumental core, as well as to begin refining the settlement occupation chr onology and characterizing resident households. In 2016, we also extended our research focus to the north at the Pearce Sites in the Cockscomb Basin Forest Reserve (Peuramaki Brown and Morton 2016). Lesson 2: Producing Field Reports with Explicit Methods is Essential I always knew it was important to file both a preliminary and final report, as well as copies of all field documentation as soon as possible after a research season, but I never truly understood how important these documents could prove to be ( SCRAP Team Member #2, personal communication, 2017). We learned not to assume that because we've done archaeology in one region that we know exactly what were doing in another ( SCRAP Team Member #3, personal communication, 2017). Prior to our res earch at Alabama, the Point Placencia Archaeological Project (PPAP) conducted preliminary mapping, extensive testing, as well as consolidation work in the epicentre of the site in the 1980s (MacKinnon 1988a, 1988b, 1989c; MacKinnon and May 1991; Walters 1988). When we began our background research on Alabama in 2013, our first goal was to go through reports from the period to determine what exactly had been done at the site, how we could build from the results, and to ensure that we would not be redoing what had already been done. Unfortunately, the few reports on file in the Belize Institute of Archaeology (IA) archives were limited in content with regard to investigation methods and details, and no original field notes were included. Additionally, only a limited number of previous project members were available to discuss their memories of details regarding this research conducted some 30 years ago. The ability to retrace the past work of archaeologists is critical and underlines the requirements that permit holders have in Belize regarding annual documentation and report filing. This is an important fact of which all Belizeans should be made aware: the results of any archaeological research (data) are required to stay in Belize. SCRAP team members spe nd much time, effort, and money preparing both preliminary and final field reports, and copies of all of our field documentation, along with copies of all presentations and publications, are put on file at the IA each year. Additionally, much of our data is shared in open access format on our project website ( www.scraparchaeology.com ). An important part of this reporting work is simply explaining our applied methods and techniques in terms of survey, excavation, and preliminary artifact analysis. The yearly need to explain our methods, no matter how standard they may seem in archaeology, also helps us to continually reflect on our procedures; Are they actually suitable for what we are doing or are w e just doing them because that is how it has always been done? Lesson 3: Map Making is an Interpretive Act The degree of interpretation involved in map making seems to be correlated fairly heavily to the familiarity with the subject matter. The difference lies in drawing what you see vs. what you interpret. This reinforces the fact that all map making is subjective (SCRAP Team Member #4, personal communication, 2017). Due to changes in standards, technologies, and personnel over time, including the particular experiences of individuals, longterm mapping of archaeological sites can be a complicated affair. This must always be remembered and respected, especially when dealing with older project maps and drawings. In the 1980s, the PPAP mapped the monumental core at Alabama, beginning with a simple tape and compass map before moving to a transit produced coarse topographic map (1 m resolution) and rectilinear interpretation ( Figure 3 ). When we returned to the site in 2014, these maps were an essential source for aligning the work of the PPAP with our own. By measuring buildings and comparing our results to the maps, which were typically spot on, we were able to plan and

PAGE 287

Experiences from Three Seasons of Investigations at Alabama 300 Figure 3 Progression of maps of Alabama, from left to right, of the monumental core: tape and -compass (PPAP 1986), topographic transit -made (PPAP 1987, 1 m contour), rectilinear interpretation (PPAP 1988), combined topographic total station made and rectilinear interpretation (SCRAP 2016, 0.5 m contour). All images are on file at the Belize Institute of Archaeolo gy. Figure 4 Taproot and ant nest damage in ALA -047A mound. Photo SCRAP 2016. contextualize our present operations. However, there is no such thing as a definitive site map, and new technologies, opportunities, and hindsight encouraged us to remap the monumental core. When it comes to archaeological mapping, it is important to start with what i s physically visible on the groundalthough, this too can be extremely subjectivevs. what is understood or interpreted from the ground. Our new combined epicentre map was made using a total station, which includes fine topographic detail (shots taken app roximately every 2 m on a grid) along with our rectilinear interpretation. Our epicentre reconnaissance identified 20 major structures (the tallest, Str. 3, measuring 7.5 m), 4 plazas, and a sacbe as well as 14 uncarved granite monoliths, presumably monu ments of some sort. The area covers 2.48 hectares, not including the surrounding borrow pits or Strs. 19 and 20, making the monumental core of Alabama slightly larger than that of Nim Li Punit in Southern Belize following the calculation process presented in Houk (2015). O ur understanding of Alabama, represented as sequences of maps, will no doubt change over time as new visualization techniques are applied and we are able to incorporate more past and future excavation detail. Lesson 4: Effaced Earthen C ore Architecture Represents a Unique Challenge Meaghan told me we would probably be digging an empty pile of dirt. She wasnt lying (SCRAP Team Member #5, personal communication, 2017). Our 2016 season focused on test excavations at settlement sites presumed to represent commoner domestic buildings and associated spaces (Peuramaki Brown, ed. 2016). As most of the Alabama settlement is located in an active citrus orchard, vs the monumental core which is covered in broadleaf forest, we were prepared for significant disturbance related to the original preparation of the orchard and earlier banana plantation, as well as continued disturbance from various maintenance processes. What we were n ot entirely prepared for was the destructive ability of ants, spiders, and other

PAGE 288

Peuramaki -Brown et al. 301 Figure 5 From left to right, compaction test setup at ALA -047A to detect effaced earthen -core (sandy -clay) platform vs. intact granite face at ALA -047B and ALA-047C. Photos SCRAP 2016. borrowing creatures within the relatively loose sediments, in addition to the very large taproots from a particular vine that commonly appears in citrus orchards ( Figure 4). In addition, armadillo hunting as a form of illicit excavation made an appearance in the monumental core. Each of these taphonomic agents and effects played a significant role in our applied methodologies and hence our documentation strategies. Previous work in the district mentioned the presence of predominantly sandy clay for the construction core of monumental construction platforms, mined from surrounding alluvial plains, and faced with granite or other nonlimestone facing blocks and the use of crush ed granite as a flooring material at some sites (e.g. Graham 1994; MacKie 1985; Price 1899). Little to no plaster is found in the architecture of the district, likely related to the lack of significant limestone deposits in the region; rather, blocks of s late and granite are often used as paving for the surfaces of monumental platforms. The lack of plastering agents on nonmonumental platforms also means pebble/cobble ballast layers are typically absent; this makes identifying the actual surfaces of smaller platforms, if they survive, quite difficult to distinguish from overburden sediments, including colluvium. Most platforms also lack artifacts within their construction core, and were often missing their granite facings, either partially or completely r emoved in antiquity, if present to begin with. Thus, resulting in only minimal fallen or slumped stone material at some sites, making it even more difficult to distinguish the actual surfaces of platforms. At the largest platform that we tested in the Ala bama settlement, just over 2 m tall, these issues required us to develop a compaction test to approximate where exactly the platform stratigraphy started within the mound ( Figure 5 ). At the smallest platform, the careful noting of where artifacts were and were not encountered also helped to delineate the exterior from the interior of the platform. Additionally, recording even the most minute of details in stratigraphy was at times our only key to understanding a domestic platform. On the other hand, othe r platforms were very clearly discerned based on the intact granite facings easily located by probing with a machete to find intact alignments and crushed granite lenses (former surfaces). In upcoming seasons, we will better test the methodology for dea ling with effaced earthen core architecture that was recently proposed by Brouwer Burg et al. (2016). After our season of testing in the citrus orchard, it was determined that, despite difficulties, these excavations are worthwhile as there appears to be far more in the way of material culture

PAGE 289

Experiences from Three Seasons of Investigations at Alabama 302 (artifacts) associated with settlement mounds and nonmounded areas than is typically encountered with the monumental architecture of the district. Lesson 5: Pottery is Still Valuable without Type Variety Elizabet h Graham always told me that the pottery of Stann Creek was horrible. I didnt really believe her. I do now ( SCRAP Team Member #2, personal communication, 2017). Im super jealous. Probably the first time anyone has ever said that about the Stann Creek Assemblage ( SCRAP Team Member #6 commenting on recent petrographic study results, personal communication, 2017). Since we first contemplated work in the Stann Creek District, we wer e warned to be prepared for no artifacts in construction core material (as mentioned above) and for really poor pottery due to the highly erosive, acidic soils of the district. The pottery we found during our 2015 surface collection was in tolerable shape with some surviving slips and decorations; however, materials below ground surface typically have no surviving surface treatments. Pottery materials from the predominantly clay occupation horizonthe surface atop of which the domestic platforms were con structed is often reduced to red smears that have the shape of a pottery sherd, but no consistency. As a result, typevariety analysis, which depends heavily on surface treatment, is difficult if not impossible with the Alabama assemblage. The forms can provide us with an idea of time period, as can some wares/fabrics; however, the most useful information comes from source characterization and technology studies Since 2015, we have been working on preliminary petrographic studies of both clays collected from the Alabama area and pottery recovered from surface collection. This study has proven helpful in providing preliminary macrovisual, microvisual, and technological descriptions of the most common pottery fabrics recovered at Alabama, and has demonstra ted the presence of locally manufactured (within 10 km) wares and the use of local clay like sediments for construction daub; nonlocal wares include one that originates to the south in the Bladen as well as one to the north near the Hummingbird; and even grog from vessels produced in the Belize Valley, used in another possible locally manufactured ware (Peuramaki Brown and Howie 2017). Ongoing petrographic analysis will help us to create a typology for the area, and to better understand certain human reso urce relationships of the Alabama Maya to compare with resource use by the modern Maya of the Alabama area (Toledo Maya Cultural Council & Alcaldes Association 1997:112113). Lesson 6: Responsible Use of Appropriate Visualization Techniques is Essential The Care Bear drawing? More information would be nice (SCRAP Team Member #7, personal communication, 2017). In addition to more intensive artifact studies, such as our pottery petrography, basic artifact analysis and documentation is critical to any ar chaeology project. We have learned that the visualization of artifacts, particularly through illustration, photography, and 3D model generation, is critical to capturing as much data as possible. It is also in critical to our ability to convey informatio n regarding special finds to our colleagues and community members. Unfortunately, multiple visualization methods have not always been employed in the district, for a variety of reasons; previously produced illustrations, such as the Care Bear stone, l eave much to be desired and many questions unanswered (see MacKinnon 1988a: fig. 1). The SCRAP team has made a point of using multiple visualization methods for our small finds documentation and analyses, including photography, illustration, and 3D scanning, alongside standard measuring, colour designations, and other quantitative and qualitative description ( Figure 6). Starting in 2018, we will also be incorporating 3D scanning of excavations into our 3prong visualization strategy. In addition to serving analysis and

PAGE 290

Peuramaki -Brown et al. 303 Figure 6 From left to right, artifact photography, technical illustration, and 3D scan/model. Images SCRAP 2016. outreach purposes, these visual aids will be used in the creation of an online artifact catalogue, as well as an Open Education Resource called the Athabasca University Virtual Archaeology Lab that consists of basic introductory lab exercises for new archaeology s tudents. Lesson 7: Archaeology Isnt About Finding Temples in the Jungle I didnt really realize that some centers are overgrown and difficult to access. I had seen photos of ancient Maya centres and assumed they always looked that way, but seeing the centre at Alabama made me realize how much work goes into making a site 'tourist ready' ( SCRAP Team Member #1, personal communication, 2017). One of our favourite lessons todate, because it is one we often forget, was from a project member who came to the Maya area for the first time. They had been previously taught about Maya archaeology in the jungle, but the only images ever shown in class or in textbooks were of consolidated buildings within the jungle. Seeing what is actually meant by temples in the jungle completely covered platform mounds was new to them, and reminds us of what tourists to the region might expect to see vs. some of the reality of archaeological sites in Belize. As a result, we believe it is very important to convey both the nicely consolidated buildings of sites such as Cahal Pech alongside those still in the bush, such as many of the amazing platform mounds of sites such as El Pilar. Making sure tourists visit both types of settings helps to emphasize this lesson and to demonstrate how much work goes into tourism development in Belize. It is also worth noting that in a study by Ramsey and Everitt (2008) involving interviews with tourists and Belizeans regarding site development in the Cayo District, the most common obser vation was the need for more educational value when it comes to visiting sites, but that this did not necessarily go handin hand with the need or desire for more consolidation. In fact, the ability to see sites in nature as opposed to consolidated was identified as important, and that simply increased amounts of tour guide information, museums, and maps were what was actually desired for more quality, educational experiences. These possibilities could easily be extended to the use of augmented reality applications at sites, where original buildings could be seen using portable tablets or smartphones at trigger points on the landscape; thus, eliminating the need for increasingly expensive consolidation efforts. We are currently applying for funds to t est such emerging technology at Alabama, for the purpose of locally oriented education and outreach. Lesson 8: Archaeology is a Team Effort Just because you don't speak the same

PAGE 291

Experiences from Three Seasons of Investigations at Alabama 304 language as someone else doesn't mean you can't learn from each other ( SCRAP Team Member #3 commenting on working with older Maya field assistants, personal communication, 2017). Building relationships with people in the area where you are working is just as important as the actual archaeology ( SCRAP Team Member #8 commenting on our community outreach and engagement efforts, personal communication, 2017). Our final lesson is acknowledging how much of a team effort archaeology really is. Not only among the archaeologists themselves, but with our entire crew, collaborators, local community members and leaders, property owners, government representatives, cultural associations, wilderness societies, etc. The list is never ending and we wis h to make it clear that none of our research is possible without this team approach. We look forward to expanding our networks in upcoming seasons when we have our first archaeological field school at Alabama in 2018; continue expanding our focus toward t he site of Pearce and understanding its relationship with the Alabama Maya; and new incorporations of geomorphological and botanical studies in both areas. With this paper, it is not our intention to be patronizing or to diminish the work of our colleagues. Rather, we hope that these few lessons that we ourselves have had to relearn, through their specific contextualization within the aims and operations of the Stann Creek Regional Archaeology Project, can serve to remind us and others of some of the litt le discussed realities of archaeological field work. Acknowledgements We would to say thank you and botik to many individuals, institutions, associations, and businesses for their support over the first three years of fieldwork at Alabama, as well as the preliminary reconnaissance trip at Pearce. These include all SCRAP 20142016 team members; Mr. Greene and Ms. Canton of Greene Groves; Mr. Chiac and the Maya Mopan Village Council; Mr. and Ms. Teul of Miss Juanas Kitchen; Mr. Sansone and Cardies Hotel; Dr. Morris and the staff of the Belize Institute of Archaeology; Ms. Alpuche of the National Archives of Belize; Mr. Coc and Belize Audubon Society; Mr. Arana and the Stann Creek House of Culture; Drs. Graham, Dunham, and Prufer. This research was funded by grants from the Social Science and Humanities Research Council of Canada and the Research Centre at Athabasca University, as well as a crowdsourced funding initiative through Experiment.com. References Brouwer Burg, Marieka, Astrid Runggaldier, and Eleanor Harrison -Buck 2016 The afterlife of earthen -core buildings: A taphonomic study of threatened and effaced architecture in Central Belize. Journal of Field Archaeology 41(1):17-36. Dunham, Peter, Robert C. Murray, William E. Brooks, Robert P. Reynol ds, Theresa H. Cookro, and Jeremy F. Jacobs 1995 Field Report of the 1995 Season of the Maya Mountains Archaeological Project (MMAP). Report submitted to the Department of Archaeology, Forest Department, Office of Geology and Petroleum, Belize Audubon Soci ety, Belmopan, Belize; The National Geographic Society, Washington, D.C.; KeyCorp, Cleveland, OH; The Dart Foundation, Mason, MI. Gann, T.W.F. 1918 The Maya Indians of Southern Yucatan and Northern British Honduras Bureau of American Ethnology Bulletin No. 64. Smithsonian Institution, Washington, D.C. Graham, Elizabeth 1976 Archaeology of the Stann Creek District, Belize. Stann Creek Project 1975 Interim Report. Centre of Latin American Studies, University of Cambridge. 1978 Archaeological Investigati ons in the Stann Creek District. Belizean Studies 6(4):16 -26. 1982 The Highlands of the Lowlands: Environment and Archaeology in the Stann Creek District, Belize, Central America. Unpublished PhD dissertation, University of Cambridge, Great Britain. 1985 Facets of Terminal to Post -Classic Activity in the Stann Creek District, Belize. In The Lowland Maya Postclassic edited by A. F. Chase and P. M. Rice, pp. 215229. University of Texas Press, Austin.

PAGE 292

Peuramaki -Brown et al. 305 1986 Barton Ramie Ceramic Types at Colson Point, North Stann Creek: A Focus on the Protoclassic. Ceramica de Cultura Maya 14:32 -38. 1987 Resource Diversity in Belize and its Implications for Models of Lowland Trade. American Antiquity 52:753 -767. 1989 Brief Synthesis of Coastal Site Data from Colson Point, Placencia, and Marco Gonzalez, Belize. In Coastal Maya Trade edited by H. McKillop and P. F. Healy, pp. 135154. Trent University, Occasional Papers in Anthropology, No. 8. Peterborough, Ontario. 1994 The Highlands of the Lowlands: Environment and Archae ology in the Stann Creek District, Belize, Central America. Monographs in World Archaeology No. 19. Prehistory Press, Madison, Wisconsin and The Royal Ontario Museum, Toronto, Ontario. 2001 Stann Creek District. In Archaeology of Mexico and Central Americ a: An Encyclopedia edited by S. Toby Evans and D. L. Webster, pp.684686. Garland Publishing, New York. Houk, Brett A. 2015 Ancient Maya Cities of the Eastern Lowlands University of Florida Press, Gainesville. Joyce, T.A. 1931 Report of the British Mus eum Expedition to British Honduras, 1931. Unpublished manuscript. Copy on file in the Latin American Library at Tulane University (Rare Collections, RBC 972.82(910) J89 1 LAL). Kidder, A.V., and Gordon F. Ekholm 1951 Some Archaeological Specimens from Pom ona, British Honduras. Notes of Middle American Archaeology and Ethnology, Carnegie Institution of Washington, Division of Historical Research No.102, Feb. 16, 1951:429-437. MacKie, Euan W. 1985 Excavations at Xunantunich and Pomona, Belize, in 1959 -60 BAR International Series 251. British Archaeological Reports, Oxford, Great Britain. MacKinnon, J. Jefferson 1986 In Search of the Ancient Maritime Maya. Wisconsin Academy Review 32(3):111-122. 1988a Point Placencia Archaeological Project Belize, Central America: Excavations at Chacben Kax, Alabama, Stann Creek District, Belize, 1986 88. Report submitted to Department of Archaeology, Belmopan, Belize. 1988b Chacben Kax, 1985-88: Excavations at Alabama, Stann Creek District, Belize. Report submitted to Department of Archaeology, Belmopan, Belize. 1989a Coastal Maya Trade Routes in Southern Belize. In Coastal Maya Trade edited by H. McKillop and P. F. Healy, pp. 111 122. Trent University Occasional Publications in Anthropology, no. 8. Peterborough, Ontario. 1989b Spatial and Temporal Patterns of Prehistoric Maya Settlement, Procurement, and Exchange on the Coast and Cays of Southern Belize. Unpublished PhD dissertation, University of Wisconsin -Madison. 1989c Preliminary Report of the Point Placencia Archaeological Project: Consolidation of the Ballcourt at Chacben Kax and Archaeological Survey of the Northern Cays of the Stann Creek District, Belize. 1990 Tobacco Range, South Water Cay, Placencia Cay and Maya Sea Trade Routes in Belize. Mexicon 12(4):75-78. MacKinnon, J. Jefferson, and Susan M. Kepecs 1989 Prehispanic Saltmaking in Belize: New Evidence. American Antiquity 54(3):522533. MacKinnon, J. Jefferson, and E. May 1990 Small -Scale Maya Lime Making in Belize: Ancient and Modern. Ancient Mesoamerica 1(2):197-203. 1991 The Ballcourts of Chacben Kax and its Neighbors: The Maya Ballgame and the Late Classic Colonial Impulse in Southern Belize. In The Mesoamerican Ballgame: papers presented at the international colloquium The Mesoamerican Ballgame 2000 BC -AD 2000, Leiden, June 30thJuly 3rd, 1988 edited by G.W. van Bussel, P. L. F. van Dongen, and T. J. J. Layenaar, pp.7180. Rijksmuseum voor Volkenkunde, Leiden MacKinnon, J. Jefferson, Jeffrey M. Olson, and Emily M. May 1993 Megalithic Maya Architectural Features at the site of Chacben Kax, Alabama, Stann Creek District, Belize, CA. Mexicon 15(1):14. Mitchell-Hedges, Frederick W. 1931 Land of Wonder and Fear Century and Co., New York Peuramaki -Brown, Meaghan M. 2016 Settlement and Resource Development at Alabama, Belize: Past, Present, and Future Investigations. RRBA 13:239 250.

PAGE 293

Experiences from Three Seasons of Investigations at Alabama 306 2017 Revisiting the Ancient Maya of Alabama, Belize: Description, Recent Research, and Future Directions. Mexicon 39:6472. Peuramaki -Brown, Meaghan M. (editor) 2016 The Stann Creek Regional Archaeology Project: Report of the Third (2016) Field Season. Report presented to the Belize Institute of Archaeology, Belmopan, Belize. Peuramaki -Brown, Meaghan M., and Linda Howie 2017 Pre -Columbi an pottery and building materials of the East -Central Belize Maya: a petrographic characterization study. Paper presented at the Annual Meetings of the Society for American Archaeology, Vancouver, British Columbia, Friday, March 31, 2017. Peuramaki -Brown, Meaghan M., and Shawn G. Morton 2016 The 2016 Reconnaissance of the Pearce Sites of the Cockscomb Basin Reserve. In The Stann Creek Regional Archaeology Project: Report of the Third (2016) Field Season, edited by M. M. Peuramaki Brown, pp.111124 Report presented to the Belize Institute of Archaeology, Belmopan, Belize. Peuramaki -Brown, Meaghan M., Shawn G. Morton, Tawny Tibbits, Lisa Green 2017 Phase I Reconnaissance (2014 -2015) at Alabama: A Summary of SCRAP Investigations in East -Central Belize. Resea rch Reports in Belizean Archaeology 14:299-309. Picoult, Jodi 2007 Vanishing Acts Simon and Schuster, New York. Price, H. W., Esq. 1899 Account of excavations on Sittee River, British Honduras. Society of Antiquaries Feb. 16, 1899. Ramsey, Doug, and John Everitt 2008 If you dig it, they will come! Archaeology heritage sites and tourism development in Belize, Central America. Tourism Management 29:909 916. Sills, Cory 2016 Re -evaluating the ancient Maya salt works at Placencia Lagoon Belize. Mexicon 38(3):6974. Stomper, Jeffrey, Wendy Brown, and Elizabeth Pope 2004 Recent Research at Mayflower, Stann Creek District, Belize. Research Reports in Belizean Archaeology 1:323331. Toledo Maya Cultural Council and Toledo Alcaldes Associa tion 1997 Maya Atlas: The Struggle to Preserve Maya Land in Southern Belize North Atlantic Books, Berkeley, California. Walters, Gary Rex 1988 Field Report Dec. 1987-Jan. 1988, Chacben Kax, Point Placencia Archaeological Project. Report submitted to Depa rtment of Archaeology, Belmopan, Belize

PAGE 294

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 307 317 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 28 INVESTIGATING ANCIENT MAYA SETTLEMENT, W ETLAND FEATURES, AND PRECERAMIC OCCUPATION AROUND CROOKED TREE, BELIZE: EXCAVATIONS AND AERIAL MAPPING WITH DRONES Eleanor Harrison Buck, Mark Willis, Satoru Murata, and Jessica Craig Between 2014 -2017 the Belize River East Archaeology (BREA) project conducted archaeological survey in the areas between the large Maya centers of Chau Hiix and Altun Ha. Unlike the uplands, we have found that settlement in this low -lying coastal zone is situated in relativ ely isolated pockets of higher ground. These sites are all circumscribed by marginal land inadequate for farming. For this reason, we argue that these sites were heavily reliant on the adjacent wetlands for agriculture, building ditched and drained fields, while also relying on these biologically -rich environments for hunting and aquaculture. Here we report on our 2017 fieldwork, which investigated the sites of Chulub and Crawford Bank located on Crooked Tree island and its adjacent wetland features in t he Western Lagoon, which were mapped using drones. Our investigations have revealed a long history of human-wetland interaction, beginning in the pre -ceramic period and continuing through ancient Maya times. Wetland modifications are typically attributed to the ancient Maya. Yet, it appears that preceramic groups were the first to intensively manage these environments. We suggest, particularly in the context of aquaculture and the construction of fish weirs, that these later modifications may represent a continuum of preceramic activity, rather than a break from it by the Preclassic Maya. Introduction The BREA project study area encompasses the eastern Belize River watershed between Belmopan and Belize City, a roughly 6000 sq. km area ( Figure 1). Over the course of seven years (2011 2017), our investigations of the BREA study area have identified a dense occupation and a long his tory of settlement in the eastern Belize Valley, extending from Formative to Colonial times, ca. 900 BC AD 1900 (HarrisonBuck, ed. 2011, 2013, 2015a, 2015b; HarrisonBuck, Murata, and Kaeding 2012; HarrisonBuck, Kaeding, and Murata 2013; HarrisonBuck et al. 2015, 2016, 2017; Runggaldier et al. 2013). In recent years, we have extended our investigations to the easternmost part of the Belize River Watershed, which comprises a low lying coastal zone with numerous small creeks and tributaries along with siz eable tracts of perennial wetlands. Altun Ha and Chau Hiix are the two largest sites in this part of the BREA study area. The latter is situated along the Western Lagoon Wetland, the largest inland wetland in all of Belize. Between 20142017, BREA conducted archaeological survey in the areas between the centers of Chau Hiix and Altun Ha ( Harrison Buck et al. 2015, 2016, 2017; Norris et al. 2015 [ see inset on Figure 2]). Unlike the uplands, we have found that settlement in the coastal zone is situated in relatively isolated pockets of higher ground. For Figure 1 Map of BREA study area (m ap prepared by M. Brouwer Burg). instance, Jabonche one of the largest sites that we identified and mapped between Chau Hiix and Altun Ha is positioned on one of t he few areas of high ground found along Black Creek, a tributary of the Belize River (Harrison Buck, Brouwer Burg et al. 2016 [refer to Figures 1 and 2] ).

PAGE 295

Ancient Maya Settlement, Wetland Features, and Precerami c Occupation 308 Figure 2 Map of the BREA study area with inset showing sites in the lower Belize River Watershed (m ap prepared by M. Brouwer Burg). The areas around Jabonche and other neighboring sites, such as Chulub, Chakan Waxak Nikte' and Kunahmul are surrounded by marginal land inadequate for farming (Figure 2). For this reason, I have argued that these sites were heavily reliant on the wetlands for agriculture, building ditched and drained fields (visible in satellite imagery), while also relyin g on these biologically rich environments for hunting and aquaculture (HarrisonBuck 2014). The results of our fieldwork in 2017 build on a long term, interdisciplinary research project involving a human wetland study. During the 2017 field seasons, BREA focused on mapping these wetlands using unmanned aerial vehicles (UAVs), otherwise known as drones, revealing numerous ditched fields and drainage canals that we believe were constructed by the ancient Maya. BREA also conducted an excavation in one of the pond features thought to possibly function as a fish weir in the Western Lagoon Wetland that connects to one of the east west canal features. This sizeable canal feature extends from the site center of Chau Hiix eastward across the Western Lagoon, cuttin g through the southern end of Crooked Tree island just south of the site of Chulub. During the January 2017 field season, BREA mapped the site of Chulub and performed several test excavations at this site. In addition, d uring the summer 2017 field season our team identified what appears to be an extensive preceramic site running along the eastern shoreline of the Crooked Tree island and we performed one test

PAGE 296

Harrison -Buck et al. 309 excavation of this preceramic site at Crawford Bank Below, we summarize these finds. Ancient Maya Occupation and Human Wetland Interactions Drone Mapping of the Western Lagoon Wetlands Our efforts to investigate the perennial wetlands in the BREA study area continued in 2017. Examining satellite imagery publically available on Google Earth, BREA detected a large network of water features in the form of ponds or wells connected to a series of long, linear canals in the adjacent Western Lagoon wetlands, which we believe were constructed by the ancient Maya. During the summer 2017 season, we carried out an expansive mapping project of the wetlands using drones. We have shown elsewhere that drone mapping is a very efficient and cost effective means of mapping large scale archaeological landscapes (Harrison Buck, Brouwer Burg et al. 2015; HarrisonBu ck, Willis, and Walker 2016; Willis and Walker 2015). One of the largest and longest canals in the Western Lagoon wetlands extends east from the site center of Chau Hiix and stretches across the Western Lagoon wetlands and bisects the southern tip of Crook ed Tree island just south of Chulub (Figures 2 and 3). The geospatial mapping with drones offered an efficient and cost effective means of mapping a huge area of the wetlands in a short amount of time. In less than a week, our drone operator Mark Willis mapped over 10,000 acres of the Western Lagoon wetlands that is an area over 40 km2. To put this into perspective, the aerial extent mapped is equivalent to 75% of Manhattan ( Figure 3a ). A closeup of the longest canal extending from Chau Hiix across the Western Lagoon wetlands reveals the channel and other subtle topographic details that the drone was able to detect (Figure 3b ). Wetland Features in Western Lagoon (Operation 34) We carried out one small test excavation of a pond feature positi oned along the axis of this main linear channel in the Western Lagoon wetlands (see Figure 3b). Operation 34 was a small 1 x 2 m unit placed on the western edge of Figure 3 a. Area of Western Lagoon Wetlands mapped with drones; b. close -up of the southern area showing linear canal extending east from Chau Hiix bisecting a pond feature on the eastern side of the lagoon (maps prepared by M. Willis). the pond. Apparently, the pond feature still had standing water the week prior so we timed the excavation perfectly at the very end of the dry Chau Hiix pond

PAGE 297

Ancient Maya Settlement, Wetland Features, and Precerami c Occupation 310 Figure 4 Map showing portions of Ek Tok site center (map prepared by S. Murata and A. Kaed ing. season and the unit did not become inundated until we were almost a meter in depth. We hypothesized that this feature may have been used as a fish weir by the ancient Maya. However, aside from a piece of chert debitage, our excavatio ns yielded virtually no other cultural material, only mud and organic material. Mapping the Sites Adjacent to the Wetlands: Ek Tok and Chulub During the 2017 January and summer field seasons, we continued the survey, mapping, and excavation of select Maya sites located adjacent to the wetlands in the lower reaches of the BREA study area. Using a Total Station and GPS, we devoted three weeks in January 2017 to mapping the site core of Ek Tok, located on the western shores of the Western Lagoon, the perenni al wetland discussed above (Figures 2 and 4). The survey and mapping has allowed us to record detailed topographic information for the site core and more accurately tie in the site to our existing GIS map of the BREA study area. Ek Tok is located about a kilometer and a half north of Chau Hiix. Several sacbes or roads were identified during reconnaissance that radiate out from Ek T ok, including one that extends to the south and may in fact link up with the Chau Hiix site core. Further reconnaissance is n eeded to confirm this and is planned for future field seasons. Ek Tok is a sizeable satellite center, comprising several pyramidal structures and three discrete plaza groups, as wel l as numerous isolated mounds.

PAGE 298

Harrison -Buck et al. 311 Figure 5 Map of Chulub (map prepared by S. Murata and A. Kaeding). During the January 2017 season, we focused most of our attention on mapping Chulub. This modest sized Maya settlement is located on the southeastern end of the Crooked Tree island near the shore of the Crooked Tree Lagoon, a seasonal wetland (see Figure 2) Our survey team mapped with a Total Station the site center of Chulub, which consists of a main plaza group and other outlying mounds that were associated with a series of pond and cana l features that appear to link to the nearby lagoon ( Figure 5). These outlying mounds oriented toward the water features may not be residential, but rather, productionoriented (see further

PAGE 299

Ancient Maya Settlement, Wetland Features, and Precerami c Occupation 312 Figure 6 Artifacts from Chulub: a. polished axe head; b. arrow point; c. net weights (photos by E. Harrison-Buck). below). The area around the site center is a cohune ridge suitable for farming, but circumscribing the site is a fringing wetland f orest inadequate for farming. For this reason, the inhabitants of Chulub likely relied on the lagoon and nearby wetlands for additional sources of food, such as turtles and fish, and may have built the canal and pond features to facilitate aquaculture and other production and processing activities. Excavations at Chulub (Operations 31, 32, and 33) BREA conducted s everal test excavations at the Maya site of Chulub including Operations 31, 32, and 33 (Figure 5). Preliminary analysis of the artifacts from all excavations suggests that Chulub was primarily occupied during the Early Postclassic period (ca. AD 900 1200), with small amounts of material suggesting an earlier Terminal Classic component at the site. Below I briefly describe the results from each excavation. Operation 32 was placed over an all stone mound identified on the southwest side of the main plaza group and revealed a poorly preserved rectangular s haped shrine building. Although an earlier Terminal Classic phase was identified lower down in Operation 32, the final construction phase dates no earlier than the Early Postclassic (ca. AD 900 1200). The preservation was very poor, making it difficult t o reconstruct the buildings original configuration, but it may have consisted of upright slab construction typical of the Postclassic that has since collapsed. I believe the configuration of this structure may have originally been in the form of a radial shrine with a series of outset staircases, perhaps similar to one BREA investigated at Saturday Creek several years ago, which also dated to the Postclassic (HarrisonBuck and Flanagan 2015). Two other 1 x 2 m test units (Operation 33a and 33c) were place d on the sides and back of the largest mound in the main plaza in an effort to recover midden (trash) deposits a c b

PAGE 300

Harrison -Buck et al. 313 Figure 7 Crawford Bank Op. 35 (looking east) and three examples of lithic tools found in the excavation (photos by E. Harrison -Buck). associated with the main plaza group at Chulub. In both excavations, remains of faunal material, including turtle fish and other wetland taxa were present in the archaeological record. Remains of larger land animals, such as deer, were also identified in large quantities in all excavations at Chulub. Several significant artifacts were recovered, including a small Postclassic arrow head point ( Figure 6a ). While David Pendergast and others have long argued that these small points were used for hunting birds, Joel Palka has suggested to me (based on his studies of hunting and fishing practices among the contemporary Lacandon Maya in Guatemala) that these points were more likely used for spearing fish. Operation 31 was a 2 m x10 m excavation unit positioned on an outlying mound to the north of the main plaza, located adjacent to one of the water features. The goal of the excavation was to better understand the function of these pond and canal like features found in between these outlying structures. One hypothesis that we wanted to test was whether these water features, which become seasonally inundated but retain som e water throughout the year, were used by the Maya for aquaculture. The excavation revealed an intact terrace wall and yielded a high density of lithic material, including a number of specialized tools, such as polished axes which suggest wood working ( Fi gure 6b ). One possibility is that this area was used for carving dugout canoes and the canal features facilitated the movement of these vessels from the workshop to the open lagoon waters. In addition, our investigations revealed a relatively high densit y of animal bone compared to the other two excavations. Faunal material included deer, turtle, and fish remains, as well as net weights that may have been used for netting fish ( Figure 6c ). Although the water feature itself needs to be tested, the fauna

PAGE 301

Ancient Maya Settlement, Wetland Features, and Precerami c Occupation 314 combined with the net weights and lithic material suggest that multiple production activities may have occurred at this locale, including wood carving and the processing of meat and fish. Pre Ceramic Occupation at Crawford Bank The Crawford Bank site is located on the east side of the island on the property of the Crooked Tree Lodge adjacent to the Crooked Tree Lagoon (Figure 7 ). Operation 35 comprises a narrow 1 x 12 m strip trench with the long axis running east west so as to bisect a limestone feature running along the shoreline of the Crooked Tree Lagoon. We initially wondered whether the limestone was some kind of historic feature as there were sizeable concentrations of historic artifacts found in this vicinity, including fragments of glass bottles, ceramics, and clay pipes. While most of the material appeared to date to the nineteenth and early twentieth centuries, there were two intact bottles that the owner showed us that were identified as dating to the mid to late eig hteenth century and were among the earliest historic material we had seen anywhere on the island so we decided to test the feature for historic remains. The 1 x 12 m unit was initially divided into six 1x2 m squares (A F). We started to remove the thin layer of topsoil that covered the limestone surface and found only a handful of historic artifacts, but a plethora of pomacea shell and lithic debitage mixed with a few chipped stone tool fragments. With so much lithic material we started to wonder whethe r the limestone feature was an ancient Maya feature, but noticeably absent were any Maya ceramic sherds. We decided to extend the excavation unit another eight meters (G J) to the east in the direction of the shoreline to catch the eastern edge of the lim estone feature, which we did in Square J. It was here that we found some of our most exciting finds, including a lithic tool referred to as a Lowe point that is diagnostic of the preceramic period (Figure 7, bottom left) In addition, we found dense conce ntrations of freshwater pomacea shell in direct association with the lithic tools and debitage identified throughout the excavation. The barbed point noted above was found just below ground surface in the far eastern end of Square J, closest to the shore of the lagoon. It was lying on or just above a clearly defined gray sandy, occupation surface that was associated with a dense shell midden heap found in the far western end of Square J. Notably, on this gray sandy surface near the barbed point we also f ound several pieces of slate that appear to be worked. Through the course of our excavations, it became clear that the limestone feature was likely a natural outcrop, perhaps the remains of an ancient shoreline or a natural bedrock outcrop, although a dens e yellow clay matrix appears to run underneath some of the rock outcrop, as seen in Sq. J. We also found in this yellow clay a few more pieces of lithic debitage and at least one tool, which would stratigraphically predate the shell midden and barbe d poi nt also found in Square J. Pre Ceramic Occupation in Belize: Crooked Tree and Beyond No pre ceramic occupation has ever been reported from Crooked Tree, but there are numerous sites surrounding the area around northern Belize where similar points have been reported (Lohse et al. 2006:Fig. 2). Similar projectile points with barbed edges, including the so called Lowe and Sawmill points, were first defined by the Belize Archaic Archaeological Reconnaissance (BAAR) project directed by Scotty McNeish between 19791982. Into the 1980s and 90s, their work continued as part of the Colha project. In the decades since the BAAR project quite a bit of new data has been collected on the preceramic period in Belize that have been published more recently. For instance Lohse and colleagues (2006) published a comprehensive report of finds with a typological framework of diagnostic lithic forms for the preceramic period in Belize. At Crawford Bank, we seem to have most of the assemblage rendered in their typology, inclu ding the barbed point, pointed unifacial tools, macroblades and small blades, and hammer stones (see Figure 7; cf. Lohse et al. 2006: Fig. 8). Lohse and colleagues (2006:217) suggest these types date to the Earl y Archaic (ca. 3500 1900 B.C.).

PAGE 302

Harrison -Buck et al. 315 Over the las t decade since Lohse et al. published their article, more preceramic sites have been recorded in Belize, including a series of rock shelters in the Rio Blanco valley of southern Belize investigated by Keith Prufer and his team (2017). Prufer and colleague s have presented a revised preceramic chronology for southern Belize based on a series of preceramic burials and stratified contexts with jute shell middens and barbed points found in these stratified contexts dating as early as 9,000 years B.P. Based on a large series of radiocarbon dates from these stratified deposits they have convincingly argued that the barbed Lowe points date as early as the Paleoindian period, rather than the Early Archaic as has been previously suggested. Prufer and his team (2017:321) conclude that by at least 10,500 BC the exploitation of nearby stone tool resources and the processing of freshwater snail were a major part of the use of the rock shelter. Similarly, it appears that preceramic groups who visited the Crawford Bank site in Crooked Tree also exploited the local resources, which includes an abundance of pomacea shell, which are a plentiful resource in the wetland environments that characterize C rooked Tree. We suggest the procurement and processing of pomacea shell represents a major activity for the prec eramic groups at Crawford Bank. The barbed point from Crawford Bank is by far the most diagnostic piece found in our excavation and is arguably our most important find. To date, there have been less than a hundred such points reported from Belize. It bears the strongest resemblance to an example reported from a preceramic site near Ladyville, which is in the BREA study area and not too far from Crooked Tree (Stemp et al. 2016:Fig. 2c). According to James Stemp and colleagues (2016:292293), the seriated edges of the barbed points suggest they were meant to stay lodged in an animal, rather than be easily pulled out. He concludes that the barbed points would not be conducive for stabbing or thrusting big game animals, but more advantageous for hunting aquatic prey (Stemp et al. 2016:293). That the barbed Lowe point from Crawford Bank was found right at the edge of the lagoon shoreline lends su pport to this interpretation. In fact, the day we were excavating Sq. J and found the Lowe point, fisherman walked by to hunt fish near Crawford Bank and on their way back stopped by our excavation and showed us their catch. When we showed them the Lowe point, one remarked that they had found a similar point in the middle of the lagoon. This is where one would expect to lose a hafted barbed point if you were using it for hunting aquatic prey, as James Stemp and others suggest. That said, our excavations were dominated by an abundance of pomacea and we were surprised by the utter lack of fish or other small water bodied faunal remains that would require a barbed spear. It seems almost inconceivable that the early humans coming to Crooked Tree would only g ather Mollusca and not take advantage of the other rich and biologically diverse resources, namely fish that are available in the surrounding lagoon and wetlands. It may be simply a matter of preservation. This shoreline is seasonally inundated by lagoon water. That only the shell and lithic materials were found is perhaps because only materials that are more impervious to water have survived. Further excavation farther away from the waters edge may help to clarify this issue and also hopefully present deeper stratified deposits than we exposed in the Op. 35 strip trench. Concluding Thoughts Compared to the ancient Maya civilization, very little is known about the preceramic occupations in Belize. However, early use of wetland environments has been documented in northern Belize. For instance, wetland investigations in the Rio Hondo area by Mary Pohl and Kevin Pope identified early deposits of unifacial tools as well as a barbed Lowe point and suggested an Early Archaic date for these deposits. Althou gh more recent investigations by Prufer and others are bringing into question the traditionally accepted chronology for the preceramic, these data demonstrate that wetland environments have a long history of use and were attractive not only to the Maya, but also to the preceramic groups as well. Pohl and colleagues (1996) suggest that wetland modification with the building of drainage canals in the northern Belize area

PAGE 303

Ancient Maya Settlement, Wetland Features, and Precerami c Occupation 316 began as early as 1000 B.C. (cited in Lohse et al. 2006:223). While scholars have attri buted most of the wetland modification to the ancient Maya, it appears that preceramic groups were the first to intensively manage these environments and it may be worth considering these later modifications, particularly in the context of aquaculture and the construction of fish weirs, as perhaps a continuum of preceramic activity, rather than a break from it by the Preclassic Maya. Jon Lohse and colleagues (2006:221) observe that many early preceramic sites may exist in these perennially wet environments but their seasonal inundation [poses] severe logistical challenges to researchers. This is certainly the case for Crawford Bank, where the preceramic deposits we uncovered were found along a shoreline that is seasonally inundated by the rising waters of Crooked Tree Lagoon during the rainy season. Fortunately for us, we decided to conduct our 2017 investigations at the tail end of the dry season during the first week of June. The full extent of the site is unknown, but local informants indicate that the limestone outcrop extends the length of Crawford Bank. If the preceramic occupation follows the bedrock outcrop it may well extend for a kilometer or more along the eastern shoreline of Crooked Tree island. In the future, we plan to continue our research on the preceramic of Crooked Tree and the deep history of humanwetland interactions in this area. Acknowledgements The 2017 BREA field seasons were incredibly productive and this would not have been possible without the all star staff of the BREA project. I would like to take this opportunity to personally thank David Buck, Jessica Craig, Grace Dietz, Alex Gantos, Kelin Flanagan, Adam Kaeding, Satoru Murata, Lori Philips, Astrid Runggaldier, Katie Shelhamer, Sara ClarkeVivier, and Mark Willis. Their hard work and commitment to the BREA project are what made the two field seasons during January and summer such an enormous success. Together with my co authors, we also wish to thank our group of UNH field school students Caroline Aubry, Emma Berman, Riley Boss, Erin Hohorst, RoseAlaina Leone, Jeremy Kassel, Paul Lovely, Rebecca Philibert, Katherine Titus for their particip ation and contribution to the BREA project during the January 2017 season. I am especially grateful to the Alphawood Foundation for their generous support of the BREA project and I also wish to acknowledge the University of New Hampshire for providing additional financial support for the 2017 season. Finally, I would like to thank the Institute of Archaeology, particularly Dr. John Morris, Ms. Melissa Badillo, and Ms. Delsia Marsden for their continued support of our work and for granting me a permit to c onduct the investigations reported herein. References Harrison -Buck, Eleanor (editor) 2011 Surveying the Crossroads in the Middle Belize Valley: A Report of the 2011 Belize River East Archaeology Project Occasional Paper No. 5. University of New Hampshire, Durham. http://www.breaproject.org/DOWNLOADS/BREA 2011_Report.pdf Harrison -Buck, Eleanor (editor) 2013 Archaeology in the Middle Belize Valley: A Report of the 2012 Belize River East Archaeology Project Occasional Paper No. 6. University of New Hampshire, Durham. http://breaproject.org/DOWNLOADS/BREA_2012 _Report.pdf Harrison -Buck, Eleanor (editor) 2015a Investigations of the Belize River East Archaeology Project: A Report of the 2014 and 2015 Field Seasons Volume 1. Occasional Paper No. 7. Univ ersity of New Hampshire, Durham. Harrison -Buck, Eleanor (editor) 2015b Investigations of the Belize River East Archaeology Project: A Report of the 2014 and 2015 Field Seasons Volume 2. Occasional Paper No. 7. University of New Hampshire, Durham. Harriso n -Buck, Eleanor 2014 Ancient Maya Wetland Use in the Eastern Belize Watershed. Research Reports in Belizean Archaeology 11:245-258. Harrison -Buck, Eleanor, Marieka Brouwer Burg, Mark Willis, Chester Walker, Satoru Murata, Brett Houk, and Astrid Runggaldier 2015 Drones, Mapping, and Excavations in the Middle Belize Valley: Research Investigations of the Belize River East Archaeology (BREA) Project. Research Reports in Belizean Archaeology 12:295 304.

PAGE 304

Harrison -Buck et al. 317 Harrison -Buck, Eleanor, Marieka Brouwer Burg, Satoru Murata, Hugh Robinson, Adam Kaeding, and Alex Gantos 2016 Rivers, Wetlands, Creeks, and Roads: Investigating Settlement Patterns in the Middle and Lower Reaches of the Belize Watershed. Research Reports in Belizean Archaeology 13:137-148. Harrison -Buck, Eleanor, Jessica Craig, and Satoru Murata 2017 From Ancient Maya to Kriol Culture: Investigating the Deep History of the Eastern Belize Watershed. Research Reports in Belizean Archaeology 14:353-361. Harrison -Buck, Eleanor and Kelin Flanagan 2015 In vestigating a Postclassic Shrine at Saturday Creek: Excavations of Structures 10 an 11 (Operation 23). In Investigations of the Belize River East Archaeology Project: A Report of the 2014 and 2015 Field Seasons Volume 1, edited by E. Harrison -Buck, pp. 162 184. Occasional Paper No. 7. Unive rsity of New Hampshire, Durham. Harrison -Buck, Eleanor, Adam Kaeding, and Satoru Murata 2013 A Network of Waterways in the Eastern Belize Watershed: Recent Survey and Excavation of the 2012 BREA Field Season. Research Reports in Belizean Archaeology 10:7790. Harrison -Buck, Eleanor, Satoru Murata, and Adam Kaeding 2012 From Preclassic to Colonial Times in the Middle Belize Valley: Recent Archaeological Investigations of the BREA Project. Research Reports in Belizean Arch aeology 9:131140. Harrison -Buck, Eleanor, Mark Willis, and Chester Walker 2016 Using Drones in a Threatened Archaeological Landscape: Rapid Survey, Salvage, and Mapping of the Maya Site of Saturday Creek, Belize. Special Issue Drones in Archaeology. The SAA Archaeological Record 16(2):3035. Lohse, Jon C., Jaime Awe, Cameron Griffith, Robert M. Rosenswig, Fred Valdez, Jr. 2006 Preceramic Occupations in Belize: Updating the Paleoindian and Archaic Record Latin American Antiquity 17(2): 209-226. Nor ris, Brian, Joseph Nigro, Satoru Murata, and Hugh Robinson 2015 Initial Survey of the Lower Belize Watershed. In Investigations of the Belize River East Archaeology Project: A Report of the 2014 and 2015 Field Seasons Volume 1, edited by E. Harrison -Buck, pp. 48 -57 Occasional Paper No. 7. Unive rsity of New Hampshire, Durham. Pohl, Mary D., Kevin O. Pope, John G. Jones, John S. Jacob, Dolores R. Piperno, Susan D. de France, David L. Lentz, John A. Gifford, Marie E. Danforth, and J. K athryn Josserand 1996 Early Agriculture in the Maya Lowlands. Latin American Antiquity 7:355 372. Prufer, Keith M., Clayton R. Meredith, Asia Alsgaard, Timothy Dennehy and Douglas Kennett 2017 The Paleoindian Chronology of Tzib Te Yux Rockshelter in the Rio Blanco Valley of Southern Belize. Research Reports in Belizean Archaeology 14: 321-326. Runggaldier, Astrid, Marieka Brouwer Burg, and Eleanor Harrison -Buck 2013 Hats Kaab: A Newly Discovered E -Group at the Closing of the 13th Baktun. Research Reports in Belizean Archaeology 10: 6575. Stemp, W. James, Jaime J. Awe, Keith M. Pru fer, and Christophe G.B. Helmke 2016 Design and Function of Lowe and Sawmill Points from the Preceramic Period of Belize. Latin American Antiquity 27(3):279299. Stemp W. James, Jaime J. Awe, M. Kathryn Brown, Eleanor Harrison -Buck, Christophe G. B. Helmke, Gabriel D. Wrobel, and Jason R. Yaeger 2018 Four Preceramic Points Newly Discovered in Belize: A comment on Stemp et al. (2016:279 299). Latin American Antiquity 29 (1): in press. Willis, Mark and Chester Walker 2015 Drones Over Saturday Creek: Mapping the Hinterland Settlement Using Unmanned Aerial Vehicles (UAV). In Investigations of the Belize River East Archaeology Project: A Report of the 2014 and 2015 Field Seasons Volume 1, edited by E. Harrison -Buck, pp. 4857 Occasional Paper No. 7. University of New Hampshire, Durham.

PAGE 305

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 319 328 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 29 COMMUNITY ARCHAEOLOG Y AT AVENTURA: ARCHAEOLOGY ABOUT COMMUNITIES AND ARCHAEOLOGY FOR COMMUNITIES, RESULTS OF THE 2016 FIELD SE ASON Cynthia Robin, Laura Kosakowsky, Kacey Grauer, and Zachary Nissen This paper examines community archaeology at Aventura through the presentation of the results of the 2016 season of the Aventura Archaeology Project. Aventura is a community with a five millennia history spanning forager horticulturalist, Pre Columbian Maya, historic, and contemporary periods. The Aventura Archaeology Project addresses community at many levels. Research undertaken at Aventura in 2016 includes site core survey and excavations, pocket bajo excavations, and ceramic analysis. Ne w research in 2016 focused on expanding our understanding of the Pre -Columbian Maya community of Aventura. In tandem with studying the many past communities that inhabited the place Aventura, the project also works with local cultur al heritage leaders to promote archaeology and site protection today. While this paper focuses on the results of the 2016 archaeological research that further knowledge about the Pre -Columbian Maya community of Aventura, it is the theoretical thesis of this paper that the combi ned goals of studying past communities and working with contemporary communities enriches archaeological research on communities, both ancient and modern. The 2016 season of the Aventura Archaeology Project saw the inauguration of two annual community programs. Through the mutually beneficial goals of archaeological research and archaeological education, the Aventura Archaeology Projects seeks to develop richer understandings of past and present communities through archaeology. Introduction Community archaeology, archaeology about community and archaeology for community, in its broadest sense is a key goal of the Aventura Archaeology Project. The project seeks to undertake an archaeology that both investigates the many ancient communities that inhabited Aventura across its long five millennia history and promotes community collaboration today. We begin this paper by outlining our theoretical premise: the combined goals of st udying past communities and working with contemporary communities enriches archaeological research on communities, both ancient and modern. We then focus on how the results of the 2016 field season at Aventura expand our understanding of the Pre Columbian Maya community at Aventura. The 2016 archaeological research at Aventura included site core survey and excavations, pocket bajo excavations, and ceramic analysis. This research underscores the ecological and ritual underpinnings of community at Aventura and the duration of the Pre Columbian Maya community of Aventura. We conclude by highlighting the projects community programs initiated in 2016 that demonstrate the mutually beneficial goals of archaeological research and education. An archaeology abou t and for communities illustrates the unique ways that archaeology, with its access to the long term history of human societies can address issues of community. Community Archaeology: Archaeology about Communities and Archaeology for Communities The proje ct we propose of an archaeology about and for communities is indebted to a long history of research within two related fields of archaeology: the archaeology of community and community archaeology (e.g., Ashmore and Wilk 1988; Atalay 2012; Colwell Chanthaphonh and Ferguson 2007; Flannery 1976; Kolb and Snead 1997; Marshall 2002; Pyburn 2009, 2011; Yaeger and Canuto 2000). Community is a key aspect of human societies past and present that is a primary context for social, political, economic, and religious i nteraction and identity formation. While archaeologists have long discussed the ancient community, Canuto and Yaegers edited volume, Archaeology of Communities: A New World Perspective (2000) brought renewed critical thinking on defining and analyzing com munities in the archaeological record. Archaeology of Communities sought to untangle the definitions of an archaeological site (a location on the landscape where past human activities occurred) and an ancient community. An archaeological site can be many things and it may or may not be coterminous with an ancient community. Archaeologists must determine if the site they are excavating is a community. If it

PAGE 306

Community Archaeology at Aventura 320 is, understanding it as a community requires attention to how past people created community throug h agency, practice, materiality, interaction, and identity formation. A community brings together people, place, and premise (Watanabe 1992). It is both a place and a group of people who create a salient social identity based upon that place: a dynamic socially constituted institution that is contingent upon human agency for its creation and continued existence. (Yaeger and Canuto 2000: 5). Because of the dynamic and historically contingent nature of communities, interaction and identity are central t o a definition of community (Yaeger and Canuto 2000). People constitute communities through repeated interaction. Across their lives people may identify with one or more communities. In contrast with the archaeology of community, community archaeology, defined broadly as an archaeology that is engaged with local communities, is a relatively recent addition to academic archaeology (Marshall 2002). This addition has demonstrated that public involvement, heritage management, and collaboration with communit ies are now central issues in archaeology receiving scholarly attention (Atalay 2012). Scholarship in archaeology is expanding not just on the findings of archaeological research, but also on the importance of community archaeology and how engaging with local stakeholders produces a better archaeology, both for professional archaeologists and local communities alike (e.g., Atalay 2012; Colwell Chanthaphonh and Ferguson 2007; Marshall 2002; Pyburn 2009, 2011; Wylie 2002, 2008). Communityengaged archaeolo gy produces better knowledge exchanges, better archaeological preservation, and also better science, as the involvement of academic and non academic voices in all aspects of the archaeological process expands the rigor of hypothesis generation and testing (e.g., Wylie 2002, 2008). Within Maya archaeology, Ren (2006) argues that collaboration with local stakeholders offers new perspectives to framing Maya history that attends to stakeholders relationships with archaeological materials. Communitybased archaeology creates community between archaeologists and the local constituents they serve. Community Figure 1 Bay of Corozal showing the locations of Aventura, Cerro Maya, and Santa Rita. archaeology brings the collaborative aspects of doing archaeological research to the foreground through interaction and engagement. It is agenti ve practice based, and creates shared identities and knowledge around the importance of archaeological protection and education. At the heart of both the fields of the archaeology of community and community archaeology is the idea that communities are socially constituted, dynamic, and historical contingent: they are created and recreated through human agency. It is the premise of this paper that archaeologists and their collaborators, with their access to the long term history of human societies, can uniquely and fruitfully conjoin the fields of the archaeology of community and community archaeology to enrich our understanding of community both today and in the past. The archaeological site of Aventura has long been central in the lives of members of adjacent communities in northern Belize due to its visibility along the northern highway. This centrality and the activity of local cultural heritage leaders was as much an impetus for the initiation of the Aventura Archaeology Project in 2015 as was the academic goal of understanding a long lived community. Given the limited nature of research at Aventura prior to the inception of the Aventura Archaeology Project, the si te is best known as the ancient Maya site whose main temple is easily visible as you drive up the northern highway approaching Corozal Town. Aventura is situated today in an area of sugar cane farming. As the largest architecture at the site is inaccessi ble for sugar cane farming,

PAGE 307

Robin et al. 321 it remains shrouded in forest canopy standing in clear contrast with the cane fields that surround it. Most visible is Aventuras main temple, Structure 1 in Group A, which stands 20 meters (66 feet) in height. Less visible from the northern highway perspective is that this site was a community, an urban center with six central civic ceremonial groups, which form the core of a Pre Columbian city with a dense urban settlement of 206 buildings per square kilometer (Robin et al. 2017). Understanding Aventura as a community has implications for ancient studies and modern communities alike. Aventura as a Site and Community through Time The Aventura site is located 10 km southwest of contemporary Corozal Town and adjacent to the vill age of San Joaquin, 10 km southwest and 10 km west, respectively, of the better known sites of Santa Rita and Cerro Maya (Figure 1). It is part of the Bay of Chetumal region, which spans what is now the northern part of Belize and the southern part of Qui ntana Roo, a region that today is divided by the modern national boundary between Belize and Mexico (Walker 2016). Previous research at Aventura prior to the inception of the Aventura Archaeology Project in 2015 included a rough map of the site core and ei ght tests pits conducted by Raymond Sidrys of UCLA in 1974 (Sidrys 1983) and excavation of three residences by Rafael Guerra, Sherilyne Jones, and Melissa Badillo of the Institute of Archaeology in 2007. Based on this research, we were aware that Aventura had a long occupation history spanning the Middle Preclassic to the Spanish Conquest. Sidrys identified occupation and use of Aventura spanning the Middle Preclassic to the Late Postclassic and identified Aventura's peak in the Late Classic to Early Post classic periods. The 2015 research of the Aventura Archaeology Project added further chronological depth to the history of the site with the identification of Archaic artifacts and Historic Era sites (Kosakowsky 2015; Robin et al. 2017). This evidence su ggests the potential for an even longer five millennia occupation of the site potentially going back to the Late Archaic around 3400 BC. Figure 2 Results of the 2016 and 2015 Aventura settlement survey. Map by Kacey Grauer. Figure 3 Location of Archaic surface finds at Aventura. Map by Jennifer Reese.

PAGE 308

Community Archaeology at Aventura 322 Aventuras central ceremonial complex consists of six plaza groups, Groups A to F, which contain seven temples ranging in height from six to 20 meters. The combined settlement surv ey research at Aventura in 2015 and 2016 identified 119 mounds in a km2 area around the central precinct of the site (Figure 2; Nissen 2016). Aventuras long history extending over 5000 years across the Late Archaic, Pre Columbian civilization, and Historic periods indicates that the place of Aventura has always been dynamic and changing. Aventura played a role for diverse communities through time, and here we present a few snapshots of those communities and their relationship to the place of Aventura During the Late Archaic period (post 3400 BC), the community that came together at Aventura was mobile. Based on the location of archaic surface finds, forager horticulturalists may have first been attracted to Aventura due to the presence of pocket bajos (Figure 3). Pocket bajos are small non draining karstic depressions that are less than 2 km2 (Dunning et al. 2006; Grauer 2016). They are seasonal wetlands today, but may have consistently held water in the past. Mobile forager horticulturalist com munities may have come together at Aventuras pocket bajos due to their resource potential. By the Preclassic period, agriculturalists had established a permanent community at Aventura which grew and expanded through time, seeing its heyday in the Terminal Classic to Early Postclassic period (750 1100 AD). Certainly by the time of its heyday that community consisted of six civic ceremonial plazas and seven temples and provided for the political, religious, economic, and social needs of community members (see Figure 2). We do not yet know the exact timing of the abandonment of the Pre Columbian city of Aventura, but during and after the War of Castes in Yucatan (1847 1901 AD), Maya and Mestizo immigrants, British colonialists, and Africans toiled in multi ethnic communities of sugar production work at Aventura. Survey has identified the presence of two sugar m ills and one Caste War church (Figure 4; Jones 2015; Robin et al. 2017). Figure 4 Caste War Church at Aventura. Today, communities of sugar cane farmers and residents of San Joaquin work and live in and around Aventura, as well as alongside the remains of the now defunct US American commercial papaya venture of the Fruta Bomba company located 250 meters north of the PreColumbian community center 2016 Research at Aventura: Investigating the Duration, Ecological, and Ritual Underpinnings of Community Archaeological research during the 2016 field season at Aventura focused on the period of Pre Columbian Maya civilization and expanded our understanding of the Pre Columbian Maya community at Aventura. The 2016 research accomplished four research goals. (1) Survey research continued within Aventuras civic ceremonial core. This research mapped the B and C plaz as, portions of the A plaza, and patio group (Group 48) adjacent to the C plaza (Figure 5). The survey work was conducted using Total Station, GPS, and GIS technologies (Nissen 2016). (2) Operation 1 investigated two of the three pocket bajos located in Aventuras civic core, Bajos 1 and 2 that flank Aventuras main temple (Structure 1 in Group A). This research examined the use of pocket bajos at Aventura and collected environmental data (Grauer 2016). (3) Operation 2 investigated the northern most mou nd in Aventuras B plaza. This research examined chronology, construction history, and subsurface preservation (Robin et al. 2016). (4) Laboratory analysis of ceramic material was undertaken for the 2016 survey and excavation work as well as from the 200 7 salvage excavation work by Rafael Guerra, Sherilyn Jones, and Melissa Badillo the Institute of

PAGE 309

Robin et al. 323 Figure 5 Map showing the location of the 2016 survey at Aventura. Map by Zachary Nissen. Archaeology at 3 structures to the north of the city center (Kosakowsky 2016). Operation 1: Pocket Bajos as Ecological and Ritual Foci of Community The 2016 and 2015 survey at Aventura, directed by Zachary Nissen and Kacey Grauer respectively, documented the central precincts of the site that are situated around t hree pocket bajos (see Figure 2). In 2016 Kacey Grauer directed the first test and trench excavations in the two pocket bajos that flank Aventuras main temple (Structure 1 in Group A), Bajos 1 and 2 (Operation 1), with the goal of evaluating the ecological environment at Aventura, the relationship between ecology and community, and assessing preservation of paleo ethnobotanical remains for future analyses (Grauer 2016). Trench excavations at the pocket bajos were placed upon the edges of the pocket bajos to determine the form of their edges and the presence of human modification. The surface of the bedrock in the trench at Bajo 1 was undulating from erosion with cut lines from plowing machine s. The irregularity of the Figure 6 Undulating bedrock on Bajo 1 edge. Photo by Kacey Grauer. Figure 7 Drop -offs in bedrock edge of Bajo 2. Photo by Kacey Grauer. undulations and the gradual, sloping nature of the edge of Bajo 1 suggests the form of this pocket bajo was the result of natural formation processes (Figure 6; Grauer 2016: 25). In contrast, the two nearly right angled drop offs along the edge of Bajo 2 suggest residents modified the edge of this karstic depression, forming large step like terraces (Figure 7). The higher edge to the west appears that it could

PAGE 310

Community Archaeology at Aventura 324 have been carved out by water (eroded away). However, the second edge is much more square and the bedrock between the two is quite flat. This bedrock stands in stark contrast to the clearly natural bedrock encountered at Bajo 1. These drop offs are too large and spread apart to have served the function of stairs; however, their presence may have made accessing Bajo 2 easier. The modificat ion of the edges of karstic depressions for access has been documented elsewhere in the Maya region (Grauer 2016: 28 29; Munro Stasiuk et al. 2014). Kacey Grauers 2016 excavations additionally identified that the pocket bajos not only played an important role in the ecological underpinnings of Aventuras diverse communities, but they played a ritually important role as well. Excavations revealed a concentration of burned ceramic fragments along the edge of Bajo 2. Laura Kosakowsky (2016) determined that the 28 ceramic fragments were burned on the interior and came from the same incensario, suggesting that ritual practices involving the burning of incense were taking place at the edge of the pocket bajo. Water worn stone found in Bajo 2 provisionally sugg ests the presence of standing water in these karstic depressions at some point in time. Intriguingly, several marine shell fragments were also found in Bajo 2. While the marine shell fragments may have entered the pocket bajos through a variety of pathwa ys, some of which may have involved unintentional incorporation, it is possible that they were intentionally incorporated in the pocket bajo as references to watery places. Evidence from Grauers 2016 excavations suggests water rituals were potentially oc curring on the edge of Bajo 2. Thus, pocket bajos would not only have been critical for the ecological underpinnings of Aventuras diverse communities, but may have been central places for community water ritual as well (Grauer 2016: 27 28). Operation 2: Defining the Civic Nature of Aventuras Community Center Operation 2, directed by Zachary Nissen and Melissa Jones, investigated the northern most structure (Structure 1) in Aventuras B plaza (Figure 8). Structure 1 in the B plaza is a range type struct ure that stands 3.4 meters tall and is one of the smaller structures in Aventuras central precincts. Excavations consisted of a single eight meter long trench located on the western edge of the structure in a heavily damaged area of the structure. Opera tion 2 research examined chronology, construction history, and subsurface preservation (Robin et al. 2016). As expected for an excavation in the area of modern damage, the architecture of Plaza B Structure 1 was disturbed and we encountered mostly collapsed material and structure fill. Excavations uncovered the eroded dry core fill of the structure, but this structure fill was not excavated. Once excavations continued below the area of modern damage, we identified a well preserved sequence of three plaza floors. Given that dry core fil l formed the construction material for Plaza B Structure 1 few artifacts were encountered in Operation 2 excavations making chronological assessment difficult (Kosakowsky 2016). As is the case across the site of Aventura, Terminal Classic/Early Postclassi c (750 1100 AD) material predominated. Early Classic material was found in Fill 3, the fill of Plaza Floor 3, the initial plaza floor constructed in this area of Group B, but the small ceramic assemblage was too eroded to assign a time period with certainty (Kosakowsky 2016). Given the dry core construction fill and lack of domestic artifacts it is plausible that Plaza B Structure 1 was an administrative building, although additional work would be needed to test this hypothesis. The northern location of this structure initially suggested that it might have been a residential structure, as northern structures in Maya civic plazas are often residential buildings (Ashmore 1991; Ashmore and Sabloff 2002). If Plaza B Structure 1 is indeed an administrative building and since it is one of the smaller structures in Aventuras civic core, this evidence may suggest that Aventuras community center was fully dedicated to civic ceremonial activities and residential functions were located beyond Aventuras six cent ral plaza groups. Ceramic Analysis and the Duration of Community at Aventura Ceramic analysis conducted by Laura Kosakowsky (2016) continued to identify that

PAGE 311

Robin et al. 325 Figure 8 Map showing location of Operation 2 excavations. Map by Zachary Nissen. Aventuras period of largest expansion is during the Terminal Classic to Early Postclassic period (750 1100 AD). Up until 2016, Chen Mul/Kol Modeled censors of the Late Postclassic period had only been found in Aventuras main A plaza (Kosakowsky 2015; Sidrys 1983). Raymond Sidrys suggested that this evidence indicated that Aventura was abandoned as a residential community in the Late Postclassic period and became a pilgrimage dest ination. New archaeological research at Aventura does not refute the claim that Aventura became a pilgrimage center in the Late Postclassic period but it does complicate the original model. In the 2016 season, Laura Kosakowsky analyzed the ceramics from the 2007 salvage excavations by Rafael Guerra, Sherilyn Jones, and Melissa Badillo of the Institute of Archaeology at three households located 250 meters north of Aventuras civic ceremonial core. Kosakowskys (2016: 38) analysis of the Institute of Archa eologys 2007 salvage excavation ceramics identified the first evidence of Chen Mul or Kol Modeled censer fragments outsize of Aventuras A plaza and within domestic contexts, suggesting the possibility of domestic occupation at Aventura across the Postcla ssic, although further research is necessary to test this hypothesis. As Zachary Nissen (2017) notes pilgrimage and residence are not necessarily antithetical to one another. Residential communities who inhabit pilgrimage locations can provide critical s upport for visiting pilgrims and pilgrims can provide long distance connections for residents (Masson 2002, 2003; Milbrath et al. 2008). Community Collaborations The 2016 season at Aventura also saw the inauguration of the projects annual community programming in collaboration with local cultural heritage leaders. Collaborative community programming through the Aventura Archaeology Project seeks to facilitate com munities of engagement with archaeology today. All public archaeology events were developed with our partners the Institute of Archaeology, Institute of Social and Cultural Research, Corozal House of Culture, and San Joaquin Village Council. Every member of the Aventura Archaeology Project, local and foreign, is involved in the design and production of the projects public archaeology events. Two annual events were initiated in 2016.

PAGE 312

Community Archaeology at Aventura 326 In San Joaquin, the community adjacent to Aventura, in collaboration with the San Joaquin Village Council, Institute of Archaeology, and Institute of Social and Cultural Research, we initiated an event called Aventura Archaeology Day. In 2016, Aventura Ar chaeology Day attracted over 150 visitors, from children to grandparents. Geared to all age groups, Aventura Archaeology Day, includes exhibits, posters, demonstrations, arts and crafts, sports, and face painting, promoting archaeological and cultural education and preservation. Debra Wilkes, Director of the Corozal House of Culture, inaugurated Ancient World Week at the Corozal House of Culture along with the Institute of Archaeology and Aventura Archaeology Project. This week long event was geared towar ds teachers, principals, and tour guides and presented talks, workshops, demonstrations, and a site tour of ongoing research at Aventura. Building upon the success of Aventura Archaeology Day and Ancient World Week, in 2017 the Aventura Archaeology Project created a third partnership with Angelita Magana, Director of the San Joaquin R.C. Church summer school to provide an on site field experience for primary school students at Aventura. As much as Archaeology Day and Ancient World Week provided residents o f Belizes northernmost Corozal district with an opportunity to learn about and engage with archaeology, they provided Aventura Archaeology Project members an opportunity to become better archaeologists and do better research. Foreign and local project me mbers alike were afforded new ways of interacting with archaeological materials through these events. These events were venues for the active creation of communities of archaeological engagement that brought together people united by an interest in the ar chaeology of Belize. Through the mutually beneficial goals of an archaeology about and for communities the Aventura Archaeology Projects seeks to develop richer understandings of past and present communities and illustrate the unique role that archaeology can play in this process. Both today and in the past the place of Aventura has an important role to play in the creation of a diver sity of communities. Acknowledgements We would like to thank the Belize Institute of Archaeology, Director John Morris, and all of the staff for their support for the second field season of the Aventura Archaeology Project. The site of Aventura was suggested to us by Director John Morris as a key site in Belize that needed new archaeological attention. The 2016 season of th e Aventura Archaeology Project would not have been possible without the dedicated work of all project members including Andres Novelo (foreman), Rosita Mai (accommodations manager), archaeological assistants Christian Arana, Frankie Arana, Johnny Catzim, a nd Brandon Itzab, and undergraduate students Melissa Jones and Azarel Sutherland. The 2016 research was completed with the kind permission of the Aventura landowners, including Evan Cowo, Sarida Cowo, Rabey Cruz, Romeo Jimenez, Florencio Javier Martinez, Guillermo Tun, Aumelia Tzul, Adalberto Witz, and Fernando Yam. The 2016 research was made possible due to the generous funding of an Alumnae of Northwestern University Grant to Cynthia Robin and a Northwestern University Graduate Research Grant to Kacey G rauer. We thank our community collaborators who made our community outreach possible: Debra Wilkes, Director of the Corozal House of Culture, Florencio Javier Martinez, Chairman, San Joaquin Village Council, and Angelita Magana, Director of the San Joaqui n R. C. Church, summer school. We thank Allan Moore, Sylvia Batty, and Josue Ramos of the Institute of Archaeology, Selene Solis, Giovanni Pinelo, and Linette Sabido of the Institute of Social and Cultural Research, and Leah McCurdy and Jorge de Leon for their assistance and expertise in community programming. References Ashmore, W. 1991 Site -Planning Principles and Concepts of Directionality among the Ancient Maya. Latin American Antiquity 2(3):199 -226. Ashmore, W., and Sabloff, J. A. 2002 Spatial orders in Maya civic plans. Latin American Antiquity 13(2):201 -15.

PAGE 313

Robin et al. 327 Ashmo re, Wendy, and Richard R. Wilk 1988 "Household and Community in the Mesoamerican Past." In Household and community in the Mesoamerican Past edited by R. R. Wilk and W. Ashmo re, pp. 1 -27. Albuquerque: University of New Mexico Press. Atalay, S. 2012 Community -based Archaeology: Research With, By, and For Indigenous and Local Communities. Berkeley: University of California Press. Canuto, M. A. and Yaeger, J. 2000 The Archaeology of Communities: A New World Perspective. London: Routledge. Colwell Chanthaphohn, C. and Ferguson, T. J. 2007 Collaboration in Archaeological Practice: Engaging Descendant Communities New York: AltaMira Press. Dunning, N. P., Beach, T., & Luzzadder -Beach, S. 2006 Environmental Variability among Bajos in the Southern Maya Lowlands and its Implications for Ancient Maya Civilization and Archaeology. In Precolumbian Water Management: Ideology, Ritual, and Politics pp. 81-99. University of Ariz ona Press, Tucson. Flannery, K. V. 1976 The Early Mesoamerican Village New York: Academic Press. Grauer, K. 2016 Operation 1 Excavations. In Aventura Archaeology Project: 2016 Season, edited by C. Robin, pp. 22 -30. Report on file with the Institute of Archaeology, Belize. Jones, M. 2015 Survey of Historic Sites at and Near Aventura. In Aventura Archaeology Project 2015, edited by C. Robin, pp. 73 -91, report on file at the Institute of Archaeology, Belize. Kolb, M. J. and Snead, J. E. 1997 Its a Small World After All: Comparative Analyses of Community Organization in Archaeology. American Antiquity 62(4): 609 -628. Kosakowsky, L. J. 2015 Analysis of Surface Collections. In Aventura Archaeology Project: 2015 Season, edited by C. Robin, pp. 36 -72, report on file at the Institute of Archaeology, Belize. 2016 Aventura Archaeology Project 2016 Ceramic Report. In Aventura Archaeology Project: 2016 Season, edited by C. Robin, pp. 34 -55, report on file at the Institute of Archaeology, Belize. Marshall, Y. 200 2 What is Community Archaeology? World Archaeology 34(2): 211-219. Masson, M. 2002 Community Economy and the Mercantile Transformation in Postclassic Northeastern Belize. In Ancient Maya Political Economies edited by M. Masson and D. Freidel, pp. 335 364. New York: AltaMira Press, New York. 2003 Economic Patterns in Northern Belize. In The Postclassic Mesoamerican World edited by M. Smith and F. Berdan, pp. 269 -281. Salt Lake City: University of Utah Press. Milbrath S., Aimers, J., Peraza Lope, C., and Folan, L. F. 2008 Effigy Censers of the Chen Mul Modeled Ceramic System and their Implications for Late Postclassic Maya Interregional Interaction. Mexicon 20(5): 104112. Munro -Stasiuk M. J., Manahan, T. K., Stockton, T., and Ardren. T. 2014 Spatial and Physical Characteristics of Rejolladas in Northern Yucatn, Mexico: Implications for Ancient Maya Agriculture and Settlement Patterns. Geoarchaeology 29(2): 156 172. Nissen, Z. 2016 The 2016 Aventura Survey. In Aventura Archaeology Project: 2016 Season, edited by C. Robin, pp. 7 -21, report on file at the Institute of Archaeology, Belize. 2017 (Re)Creating Monumental Space: The Everyday Use of Plaza Space at Aventura, Belize. Paper Pr esented at the Society for American Archaeology 82nd Annual Meeting, Vancouver, Canada. Pyburn, K. A. 2009 Practising Archaeology As if it Really Matters. Public Archaeology 8(2 3): 161175. 2011 Engaged Archaeology: Whose Community? Which Public? In New Perspectives in Global Public Archaeology edited by K. Okamura and A. Matsuda, pp. 29-41. Springer. Ren, A. C. 2006 Maya Archaeology and the Political and Cultural Identity of Contemporary Maya in Guatemala. Archaeologies 2(1): 8 -19. Robin, C. Nissen, Z., and Jones, M. 2016 Operation 2 Excavations. In Aventura Archaeology Project: 2016 Season, edited by C. Robin, pp. 31 -33, report on file at the Institute of Archaeology, Belize.

PAGE 314

Community Archaeology at Aventura 328 Robin, C., Kosakowsky, L. J., Grauer, K., Nissen, Z., Jones, M. 2017 Urban Design, Its Antecedents and Successors at Aventura: Results of the 2015 Field Season. Research Reports in Belizean Archaeology 14: 167177. Sidrys, R. V. 1983 Archaeological excavations in Northern Belize, Central America Institute of Archaeology, University of California, Los Angeles. Walker, D. S. 2016 Perspectives on the Ancient Maya of Chetumal Bay. Gainesville: University Press of Florida. Watanabe, J. M. 1991 Maya Saints and Souls in a Changing World. Austin: University of Texas Press. Wylie, A 2002 Thinking from Things: Essays in the Philosophy of Archaeology Berkeley: University of California Press. 2008 Legacies of Collaboration: Transformative Criticism in Archaeology. Distinguished Lecture in Archaeology, Archaeology Div ision, American Anthropological Association, San Francisco, California. Yaeger, J. and Canuto, M. A. 2000 Introducing An Archaeology of Communities. In The Archaeology of Communities: A New World Perspective, edited by M. A. Canuto and J. Yaeger. London: Routledge.

PAGE 315

Research Reports in Belizean Archaeology Vol. 15, 2018, pp. 329 339 Copyright 2018 by the Institute of Archaeology, NICH, Belize. 30 THE MAYA OF AMBERGRI S CAYE AND THEIR NEI GHBORS Scott E. Simmons Tracie Mayfield James J. Aimers, and W. James Stemp Ancient Maya occupation on Ambergris Caye has been documented from Preclassic through Postclassic times. Work at the sites of San Pedro and Marco Gonzalez has concentrated on several structures in which solid evidence has been found for connections to Maya polities in northern Belize and beyond. Other sites on the caye have also yielded evidence of these connections. Nonetheless, r elationships between island communities, as w