PRE CONTACT MARINE RESOURCES IN THE MARIANA ARCHIPELAGO By PATRICK M. Oâ€™DAY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2015
2015 Patrick M. Oâ€™Day
To Nicole, Penelope, and Otto
4 ACKNOWLEDGMENTS I am greatly indebted to numerous people who supported me in the completion of this project. First and foremost I would like to acknowledge my wife Nicole who provided invaluable assistance throughout all stages of this project an d Dr. Geoff Clark from ANU who supported the fieldwork conducted in Saipan. Dr. Mike Carson and Dr. John Peterson from the MARC also supported this research and graciously let me participate in their ongoing work at Ritidian, Guam. I am also grateful to nu merous friends and colleagues in the Pacific Archaeology community. They include Dr. Jolie Liston, Lon Bulgrin, Dr. Boyd Dixon, Dr. Hiro Kurishina, Darlene Moore, Judith Amesbury, Dr. Stephen Athens, Tim Reith, Mike Desilets, Jen Robbins, Amanda Sims, Dr. Sean Conaughton, Dr. Ethan Cochrane, and Dr. Nicolette Parr, who have all been long time friends and mentors. I would also like to thank the Guam and the CNMI Departments of Historic Preservation. Lastly, I would like to thank my committee members and coll eagues from the University of Florida.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES ...........................................................................................................................8 LIST OF FIGURES .......................................................................................................................11 LIST OF ABBREVIATIONS ........................................................................................................16 ABSTRACT ...................................................................................................................................17 CHAPTER 1 OVERVIEW ...........................................................................................................................19 Introduction.............................................................................................................................19 Pre Contact Fisheries and Their Significance ........................................................................23 The Data Set ...........................................................................................................................25 Theoretical and Analytical Methods .......................................................................................27 Ecological Approaches in Archaeology ..........................................................................29 The Ecological Approach in Marianas Archaeology ......................................................31 Historical Ecology, the Longue Dure , and Historically Related Cultures .....................33 Goals, Objectives, and the Problem ................................................................................37 Goals .........................................................................................................................37 Objectives .................................................................................................................37 Research Problem .....................................................................................................38 Questions .........................................................................................................................39 Hypothesis .......................................................................................................................40 Analytical Methods .........................................................................................................41 Thesis Organization .........................................................................................................43 2 PHYSICAL, HISTORICAL, AND CULTURAL CONTEXTS ............................................47 Introduction.............................................................................................................................47 Physical Setting ......................................................................................................................47 The Mariana Islands ........................................................................................................54 Saipan .......................................................................................................................57 Guam ........................................................................................................................58 Cultural Background ...............................................................................................................60 Archaeological Background ...................................................................................................62 Pre Latte Period (1000â€“3500 cal a BP) ...........................................................................64 Early Unai Phase (3000â€“3500 cal a BP) ..................................................................64 Middle Unai Phase (2500â€“3000 cal a BP) ...............................................................65 ....................................................................65 .......................................................................66
6 ............................................................................67 Previous Archaeology .....................................................................................................67 Historical Background ............................................................................................................68 The Spanish Colonial Period (AD 1521â€“1898) ...............................................................69 The First U.S. American Period in Guam (AD 1898â€“1941) ...........................................70 .....................70 .............................................72 The Japanese World War II Period (AD 1941â€“1944) .....................................................73 U.S. American World War II Period (AD 1945â€“1948) ...................................................73 The Second U.S. American Period (AD 1948â€“Present) ..................................................74 Ethnohistory of Marine Resource Exploitation ......................................................................74 Southe ast Asia and the Pacific ........................................................................................75 The Marianas ...................................................................................................................80 3 ARCHAEOLOGICAL TESTING AND LABORATORY ANALYSIS ...............................94 Introduction.............................................................................................................................94 Methods ..................................................................................................................................94 Excavation .......................................................................................................................94 Laboratory .......................................................................................................................96 Results .....................................................................................................................................99 ............................................................................................................................99 Stratigraphy ..............................................................................................................99 Laboratory results ...................................................................................................102 ...............................................................122 Ritidian Grotto ...............................................................................................................126 Stratigraphy ............................................................................................................127 Laboratory Results .................................................................................................128 Summary and Discussion of Ritidian Grotto Results ....................................................141 4 FISHERIES SCIENCE ANALYSIS AND RESULTS ........................................................195 Introduction...........................................................................................................................195 Methods ................................................................................................................................196 Results ...................................................................................................................................201 Morphometric Analyse s ( Anadara cf. antiquata) .........................................................201 Morphometric Analyses ( Gafrarium pectinatum ) .........................................................204 Age -Based Analyses ( Anadara cf. antiquata) ...............................................................206 Summary and Discussion ..............................................................................................208 5 GE NERAL DISCUSSION AND CONCLUSIONS ............................................................224 Introduction...........................................................................................................................224 Empirical Findings ................................................................................................................232 Ethnohis tory and History ...............................................................................................232 Laboratory Analyses ......................................................................................................233 Diversity ........................................................................................................................234
7 Length and Age .............................................................................................................238 Comparative Analysis, Bapot 1 and Ritidian Grotto Remains ......................................239 Implications ..........................................................................................................................240 Limitations and Future Research ..........................................................................................242 LIST OF REFERENCES .............................................................................................................245 BIOGRAPHICAL SKETCH .......................................................................................................266
8 LIS T OF TABLES Table page 31 Descriptions of soils from stratigraphic layers sectioned in the north wall of Block A, Bapot 1, Block A. Soils descriptions were derived from the authorâ€™s field notes and Clark et al. (2010:25 26). .................................................................................................146 32 Bapot 1, Block A radiocarbon dates (*Denotes rejected date). .......................................147 33 Total counts and weights for invertebrate remains from Bapot 1, Block A. ...................148 34 List of identified gastropod species from Bapot 1, Block A including common names and habitat preferences. ...................................................................................................148 35 Summary of total NISP and % NISP for gastropods from Bapot 1, Block A. ................151 36 List of identified bivalve species with common names and habitat preferences from Bapot 1, Block A. ............................................................................................................153 37 Summary of total NISP and %NISP for bivalve shells from Bapot 1, Block A. ............154 38 Totals for identified fish bone from Bapot 1, Block A. ...................................................155 39 Summary table of identified reptile bone from Bapot 1, Block A. .................................155 310 Summary table of identified bird bone from Bapot 1, Block A. .....................................155 311 Summary table of identified mammal bone from Bapot 1, Block A. ..............................156 312 Summary table of all analyzed faunal remains recovered from Bapot 1, Block A (*denotes unavailable data). .............................................................................................156 313 Soil descri ptions for TU 2, TU 3, and TU -4, Ritidian Grotto. ........................................157 314 Ritidian Grotto Ritidian Grotto, TU 1 radiocarbon dates (*Denotes rejected date).. ......157 315 Proposed sequence for TU 3 and TU 4, Ritidian Grotto. *=Layer IV was natural undisturbed sand and contain ed no cultural materials. ....................................................158 316 Summary pottery table for TU 3, Ritidian Grotto. ..........................................................158 317 Summary pottery table for TU 4, Ritidian Grotto. ..........................................................159 318 Summary artifact table for TU 3, Ritidian Grotto. ..........................................................159 319 Summary artifact table for TU 4, Ritidi an Grotto. ..........................................................159
9 320 Summary table of the invertebrate assemblage recovered from TU 3 and TU 4, Ritidian Grotto. ................................................................................................................160 321 List of identified gastropod shells from TU 3 and TU 4, Ritidian Grotto.......................160 322 Summary table of identified gastropod species from TU 3 and TU 4, Ritidian Grotto. .161 323 List of bivalve species recovered from TU 3 and TU 4, Ritidian Grotto. .......................163 324 Summary of bivalves recovered from TU -3 and TU 4, Ritidian Grotto. ........................163 325 Summary table of fish bone recovered from TU 3 and TU 4, Ritidian Grotto. ..............164 326 Summary of bird bone from TU 3 and TU 4, Ritidian Grotto. .......................................164 327 Summary of mammal bone from TU 3 and TU 4, Ritidian Grotto. ................................164 328 Summary of NISP of human remains recovered from TU 3 and TU 4, Ritidian Grotto. ..............................................................................................................................165 329 Summary of faunal assemblage for TU 3 and TU 4, Ritidian Grotto. ............................165 41 Summary of Anadara cf. antiquata valve length data for singlefactor ANOVA by cmbd, Bapot 1. .................................................................................................................210 42 Results of single factor ANOVA for Anadara cf. antiquata for mean valve lengths by cmbd, Bapot 1. ............................................................................................................210 43 Results of Tukey test with calculated q values for pairs of Anadara cf. antiquata valve groups with significantly different mean lengths by cmbd, Bapot 1. The critical q value, or â€œStudentized rangeâ€ for the comparisons = 4.974. ...................................211 44 Summary of Anadara cf. anadara mean valve length data by date before present ( cal a BP ), Bapot 1. .................................................................................................................211 45 Results of single factor ANOVA for Anadara cf. antiquata mean valve lengths by date before present, Bapot 1. ...........................................................................................211 46 Results of Tukey test with calculated q values for all pairs of Anadara cf. antiquata valve length groups with significantly different means by date before present, Bapot 1. The critical q value, or â€œStudentized rangeâ€ for the comparisons = 3.366. ...........212 47 Summary of Gafrarium pectinatum mean valve length data (mm) for singlefactor ANOVA by cmbd, Bapot 1. ............................................................................................212 48 Results of single factor ANOVA for mean Gafrarium pectinatum valve lengths by cmbd, Bapot 1. .................................................................................................................212
10 49 Results of Tukey tests showing calculated q values for all pairs of Gafrarium. pectinatum valve length groups with significantly different means by cmbd, Bapot 1. The critical q value, or â€œStudentized rangeâ€ for the comparisons = 4.974. ...............213 410 Summary of data for single factor ANOVA of mean valve lengths for Gafrarium pectinatum grouped by date before present, Bapot 1. .....................................................213 411 Results of Single factor ANOVA for mean Gafrarium pectinatum valve lengths by date before present, Bapot 1. ...........................................................................................213 412 Results of Tukey tests showing calculated q values for all pairs of Gafra rium pectinatum valve length (mm) groups with significantly different means by date before present, Bapot 1. The critical q value, or â€œStudentized rangeâ€ for the comparisons = 3.633. .......................................................................................................214
11 LIST OF FIGURES Figure page 11 Map of the islands of Micronesia, Melanesia, and Polynesia and their location relative to the Andesite Line. .............................................................................................45 12 Methods Relevant to Study. ...............................................................................................46 21 Map of the Mariana Islands. Map inset shows the location of the Marianas relative to east Asia, southeast Asia, and New Guinea. ......................................................................84 22 Map of Saipan Island showing the location of Lau Lau Bay and the Unai Bapot archaeological site. .............................................................................................................85 23 Photograph of latte elements located on the surface of Bapot 1 approximately 15 m south of Block A, facing west (photograph taken in 2006 by the author). ........................86 24 Photograph of toppled latte elements located in the eastern portion of the Bapot 1 site, facing east (photograph taken in 2006 by the author). ...............................................86 25 Map of the northern tip of Guam showing the location of the Ritidian Grotto site. Map inset shows the Island of the Guam w ith the Pago -Adelup Fault Line. ....................87 26 Excavations of the Ritidian Grotto rockshelter looking north out from cave. Photograph courtesy of Dr. Mike Carson and Dr. Hiro Kurashina. ..................................88 27 Entrance to cave just south of the Ritidian Grotto rockshelter. .........................................89 28 View of the limestone promontory above the Ritidian Grotto site, facing northeast (photograph taken in 2011 by author). ...............................................................................90 29 Chart showing all of the historic periods and phases for the Mariana Archipelago. .........91 210 Decorated pottery (From Butler 1994: Fig. 9 and Fig. 11). A) Achugao Incised sherds. B) San Roque Incised sherd. ..................................................................................92 211 Pottery surface treatments. A) Limefilled and impressed Ipao Stamped sherd (From Ray 1996:86, Fig. 45). B) Lime filled and impressed Ipao Stamped sherd (From Spoehr 1957:121, Fig. 56). C) Lime filled and impressed Ipao Stamped sherd (From Spoehr 1957:121, Fig. 56). D) Impressed Tarague Striated sherd (From Ray et al. 1996:85, Fig. 25). E) Striated Tarague Striated sherd (From Moore 2002:71, Fig. 3). F) Sherd with fing ernail impressions (From Thompson 1979:85, Fig. 14). ......................92 212 Examples of rim impressions (Solheim 2003:8, Fig. 1 5). ................................................92
12 213 Illustrations of Latte Period surface treatments (From Moore 2002:71, Fig. 3). A) Lightly combed surface. B) Boldly combed surface. C) Trailed surface. D) Wiped surface. ...............................................................................................................................93 31 Close up map of the Bapot 1 site showing the locations of excavation units from previous and current archaeological investigations. ........................................................166 32 Plan view map showing layout of the nine excavation units within Block A, Bapot 1 site, Saipan. See Figure 3 1 for the location of Block A within the Bapot 1 site. ...........167 33 Stratigraphic profile and photog raph of the north wall of Block A, Bapot 1. .................168 34 Stratigraphic profile and photograph of the south wall of Block A, Bapot 1. .................169 35 Stratigraphic profile and photograph of the east wall of Block A, Bapot 1. ...................170 36 Stratigraphic profile and photograph of the west wall of Block A, Bapot 1. ..................171 37 Bapot 1, Block A, calibrated radiocarbon dates plotted versus depth below surface. .....172 38 Deposition model of radiocarbon dates from Bapot 1, Block A created using OxCal v4.2.2. ...............................................................................................................................173 39 Bapot 1 site pottery. A) Decorated (San Roque type) Pre Latte sherd from TU 2, 250â€“260 cmbd. B) Decorated (Achugau type) Pre -Latte sherd from TU 5, 230â€“240 cmbd, Bapot 1 (Courtesy of Olaf Winter, ANU). ...........................................................174 310 Bapot 1 site pottery. A) Decorated Pre-Latte sherd from TU 5, 210â€“210 cmbd. B) Decorated Pre-Latte sherd from TU 3, 190â€“200cmbd, Bapot 1 (Courtesy of Olaf Winter, ANU). .................................................................................................................174 311 Bapot 1 site pottery. A) Decorated Pre-Latte rim sherd from TU 6 (Catalog No. 126), 180â€“190 cmbd. B) Decorated Pre -Latte sherd from TU 6, 130â€“140 cmbd, Bapot 1 (Courtesy of Olaf Winter, ANU). ......................................................................175 312 Frequency of ceramics (NISP) versus depth below datum, Bapot 1 (Courtesy of Olaf Winter, ANU). .................................................................................................................175 313 Ground Cypraea moneta shell beads from 180â€“250 cmbd, Bapot 1 (Courtesy of Olaf Winter, AN U). .................................................................................................................176 314 Small shell beads from Catalog Number 119, 170â€“180 cmbd, Bapot 1. ........................176 315 Pendants. A) Possible Cypraea tigris shell pendant, Catalog Number 156, TU 1, 230â€“240 cmbd. B) A pendant made from the ground apex of a conical gastropod shell, possibly a Conus sp. Catalog Number 157, TU 3, 220â€“230 cmbd, Bapot 1 (Courtesy of Olaf Winter, ANU). ....................................................................................176
13 316 Fragment of a shell band, or bracelet. Catalog Number 151, TU 1, 210â€“220 cmbd, Bapot 1 (Courtesy of Olaf Winter, ANU). ......................................................................177 317 Bivalve shell scraper. Catalog Number 67, TU 4, 100â€“110 cmbd, Bapot 1. ..................177 318 Pencil urchin tools. A) Pencil urchin tool, Catalog 164, TU 1, 240â€“250 cmbd. B) Close up of worked surface of pencil urchin tool, Catalog Number 164, Bapot 1. ........177 319 Tridacna sp. shell tools. A) Tridacna sp. shell adze, Catalog 8, TU 8, 20â€“35 cmbd. B) Possible adze or scraper, Catalog 122, TU 2, 180â€“190 cmbd, Bapot 1 (Courtesy of Olaf Winter, ANU). .....................................................................................................178 320 Worked shell. A) Pearl shell fishhook, Catalog 171b, TU 3, 230â€“240 cmbd. B) A possible fishhook blank, Catalog Number 89, TU 1, 140â€“150 cmbd (Courtesy of Olaf Winter, ANU). .........................................................................................................178 321 Worked shell implement, or possible compound fishhook point. Catalog 143, TU 6, 200â€“210 cmbd, Bapot 1. ..................................................................................................179 322 Frequency of shell artifacts versus depth below datum from Bapot 1 (Courtesy of Olaf Winter, ANU). .........................................................................................................180 323 Frequency of lithic artifacts versus centimeters below datum from Bapot 1. .................180 324 Bapot 1 frequency of fish bone. A) versus cmbd. B) Versus cal a BP. ...........................181 325 NISP frequencies of gastropods and bivalves. A) Plotted versus centimeters below datum. B) Plotted versus cal a BP, Bapot -1. ....................................................................182 326 NISP of identified species. A) Plotted versus cmbd. B) Plotted versus cal a BP, Bapot 1. ............................................................................................................................183 327 Shannon-Wiener Indexes for all Bapot 1 invertebrates, gastropods, and bivalves. A) Plotted versus cmbd. B) Plotted versus cal a BP. ............................................................184 328 Plan view map of the Ritidian Grotto cave and rockshelter with locations of TU 1 through TU 6. ..................................................................................................................185 329 Stratigraphic profiles of the west walls of TU 2, TU 3, and TU 4 and the north wall of TU 4, Ritidian Grotto. .................................................................................................186 331 Ritidian Grotto radiocarbon dates plotted versus cmbd. ..................................................188 332 Deposition model based on radiocarbon dates from Ritidian created using OxCal v4.2.2 (Bronk Ramsey 2013). ..........................................................................................189 333 Shell ornaments. A) Cypraea sp. bead, or ornament, from TU 4, Layer I, Level 1. B) Worked Conus sp. shell from TU 3, Layer I, Level 3, Ritidian Grotto. ..........................190
14 334 Shell beads and bead blank. A) Shell beads from TU 4, Layer III, level 10. B) Conus sp. bead blank and bead from TU 4, Layer I, Level 3, Ritidian Grotto. ..........................190 335 Shell and coral tools. A) Bivalve shell scraper from TU 4, Layer II, Level 5. B) Coral file, or abrader, from TU 3, Lay er II, Level 5, Ritidian Grotto. ............................190 336 Fishhook fragments from TU 4, Layer I, Level 3, Ritidian Grotto. ................................191 337 Frequency of bird bone recovered from TU 3 and TU 4 plotted by stratigraphic layer and excavation level, Ritidian Grotto. .............................................................................191 338 Frequency of fish bone NISP. A) Plotted versus stratigraphic layer and excavation level. B) Plotted versus cal a BP, Riti dian Grotto. ...........................................................192 339 NISP of identified invertebrate species. A) Plotted versus cmbd. B) Plotted versus cal a BP, Ritidian Grott o. .................................................................................................193 340 Shannon-Wiener diversity index values for all invertebrate species, gastropods, and bivalves. A) Plotted versus centimeters below datum. B) Plotted versus cal a BP, Ritidian. ............................................................................................................................194 41 Valve measurements and terms. Dashed line marks where valv e was sectioned for aging and solid line marks the total length of the shell. ...................................................214 42 Acetate peels from three A nadara valves. Whi te dots mark opaque growth increments. A) Catalog No. 102, Specimen No. 10, RMA=14. B) Catalog No. 160, Specimen No. 24, RMA=7. C) Catalog No. 160, Specimen No. 6, RMA=4. ................214 43 Mean valve length for Anadara cf. antiquata from Bapot 1 with one standard deviation above and below the mean. A) plotted by cmbd. B) Plotted by date before present (cal a BP ). ............................................................................................................215 44 Frequencies of Anadara cf. antiquata shell lengths based on 5mm length bins, Bapot 1. ............................................................................................................................216 45 Cumulative frequencies of Anadara cf. antiquata valve lengths based on 5 mm length bins, Bapot 1. ........................................................................................................217 46 Mean valve length for Gafrarium pectinatum from Bapot 1 with one standard deviation above and below the mean. A) Plotted by cmbd. B) Plotted by date before present (cal a BP ). ............................................................................................................218 47 Frequencies of Gafrarium pectinatum valve lengths based on 5 mm bins, Bapot 1. ......219 48 Cumulative frequencies of Gafrarium pectinatum valve lengths base on 5 mm bins. ....220 49 Estimated length at age according to date before present. Lines represent datespecific linear growth models. .........................................................................................221
15 410 Frequency distributions for Anadara cf. antiquata valves based on RMA grouped by early and recent periods, Bapot 1. ...................................................................................222 411 Cumulative frequencies of Anadara cf. antiquata based on RMA and grouped by older and recent periods, Bapot 1. ...................................................................................223
16 LIST OF ABBREVIATIONS ANOVA Analysis of variance ANU Australian National University BP Before present. cmbd Centimeters below datum. CNMI Commonwealth of the Northern Marianas Islands. K S Kolomogorov Smirnov (K S) goodness of fit test. MARC Micronesian Area Research Center, University of Guam. MNI NISP Minimum number of individuals. Number of individual specimens or number of identified specimens. RMA Relative Minimum Age.
17 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PRE CONTACT MARINE RESOURCES IN THE MARIANA ARCHIPELAGO By Patrick M. Oâ€™Day August 2015 Chair: Kenneth Sassaman Cochair: Debra Murie Major: Anthropology I examine preContact marine resource exploitation in the Mariana Islands by studying variations in archaeological residues of ancient fisheries. These residues included invertebrate shells and fish bones recovered from the Unai Bapot 1 site on Saipan and the Ritidian Grotto site on Guam. Evaluating pre Contact marine resources relied upon the precepts of historical ecology. This involved studying prehistoric marine environmental human interactions through the use of basic archaeological and zooarchaeological techniques, applying methods from fisheries biology, considering longterm regional and local histories through ethnographic and his torical data, and considering past and present environmental conditions. The primary goal of this project was to evaluate the human impacts upon prehistoric marine fisheries by studying variations in archaeological residues from two archaeological sites in the Marianas Islands. Marine resources have been the focus of several previous archaeological studies in the Marianas. Those described numerous trends and variations in marine faunal remains through time. However, those studies primarily consisted of cult ural resource management projects conducted in support of development. Budgets, especially for the analysis of faunal materials, are usually constrained and analyses of marine shell and fish bone are often inconsistently reported. Analyses of marine faunal materials in previous work mainly relied upon number of individual
18 specimens (NISP) and weight frequencies of identified species through time. Variation in NISP or weight frequencies is usually attributed to overexploitation and past environmental changes , such as mid -Holocene sea level fluctuations. The present study builds upon this previous work through considering regional similarities in traditional marine resource exploitation practices on different island groups in the Pacific, changes in species fr equencies, species diversity, size frequency fluctuations of two bivalve species in the Bapot 1 site assemblage, and changes in the age structure of Anadara cf. antiquata shell from the Bapot 1 site through time. This project shows that pre Contact marine fisheries were not subjected to intensive levels of exploitation that led to overexploitation, and that marine resources, especially marine invertebrates, represented a stable resource base throughout the prehistoric period in the Marianas. I concluded tha t variations in archaeological residues of marine resources were slight and most likely the result of complex social processes.
19 CHAPTER 1 DISSERTATION OVERVIEW Introduction This dissertation is the outcome of an interdisciplinary research project that combined methods and theories from the fields of fisheries science and archaeology. These methods and theories were directed at studying prehistoric marine resource exploitation practices in the Mariana Islands and determining their impacts upon ancient fisheries. The cumulative effects of fishing to support the increasing demands of growing human populations over more than 2,000 years of occupation may have been negatively impacted ancient fisheries in the Marianas. The Marianas consist of a 500 milelong arc of islands located in the western Pacific Ocean ( Figure 1 1). Recent events in history divided the Marianas into two different political entities. These include the U.S. Territory of Guam and the U.S. Commonwealth of the Northern Mariana Islands (CNMI). CNMI consists of the islands of Saipan, Tinian, Aguijan, and Rota and a group of ten small and mostly un inhabited volcanic islands in the northern portion of the chain. The Prehistoric Period in the Marianas Archipelago spans approximately 3,500 years, beginning with human settlement of the island chain by 3500 cal a BP (Butler 1994:16) and concluding with the arrival of Ferdinand Magellan in AD 1521 (Olive y Garcia 2006). T he primary goal of this project was to evaluate the human impacts upon prehistoric marine fisheries by studying variations in archaeological residues from two archaeological sites in the Mariana Islands. Analyses of these residues employed archaeological and zooarchaeological methods and considered historical, ethnographic, and environmental data related to the region. The present study also incorporated methods from fisheries science to evaluate archaeological residues. More specifically, population dynami cs and age and growth parameters were evaluated through the study of the hard structures of marine invertebrate shells
20 recovered from two archaeological sites. Results of these analyses address various questions regarding the role of marine resources in human settlement and the establishment of traditional Chamorro1 culture in the Mariana Archipelago. Primary among these questions are: to what extent did the prehistoric inhabitants of these islands impact their marine resource base and were these impacts si gnificant enough to cause changes in resource procurement strategies for exploiting prehistoric fisheries? Currently, much of the variation in faunal remains recovered from archaeological sites located throughout the Mariana Islands has been ascribed to ch anges in prehistoric environmental conditions. More specifically, changes in the relative abundance of different species with dissimilar habitat preferences are attributed to a fall in sea level that coincided with the arrival of the first settlers in the Marianas (Amesbury 1999, 2007; Amesbury et al.1996; Carson 2012). This drop in sea level followed a midHolocene high -stand that reached a maximum height of approximately 2 meters (m) above modern sea levels roughly 4,200 years ago. It is argued that human settlement and population growth was facilitated by falling sea levels, as productive coastal habitats expanded as the oceans receded (Dickinson 2000:744). Implementation of fisheries science methods aimed at understanding the impacts of prehistoric human exploitation of marine resources will enable a better understanding of the complex interactions between past environmental conditions and culture. These methods may also highlight the specific roles these interactions played in coping with environmental adversity, initial settlement of the Marianas, development of local marine exploitation practices, and growth of complexity in Chamorro society. It is generally accepted by archaeologists working in different regions throughout the world that aquatic resour ces were an important factor in the development of prehistoric societies 1 Chamorro are the indigenous people of the Marianas Islands .
21 (Ambrose 1967; Binford 1968; Claassen 1998; Erlandson 1988; Mannino and Thomas 2002). The first voyagers to reach the Marianas from points west would have found a familiar suite of marine resources to which their existing exploitation methods and technologies would have been aptly suited. Fisheries utilized by prehistoric peoples, like modern fisheries, consisted of a large range of economically important species of aquatic organisms t hat were utilized in numerous ways. This is especially true, both in the past and in the present, among the remotely spaced island chains of the tropical Pacific, which are incredibly rich and possess diverse varieties of edible fish, mollusks, and other m arine organisms (Kirch and Dye 1979). Marine resources were undoubtedly important to the earliest settlers in the region because tropical island landscapes, previously untouched by humans, were poor in endemic terrestrial plant and animal species and may h ave represented poor habitats for human settlers without a retinue of domesticated plants and animals (Bellwood 1979:21). This is because the number of genera of terrestrial species sharply decreases on Pacific islands the farther east they are from mainland and insular Southeast Asia and New Guinea. For example, there are almost 200 species of indigenous land mammals in New Guinea, whereas the islands of southeast Polynesia have only one or two terrestrial mammals. The situation is different with respect t o marine fish and shellfish. Although the number of marine species also decreases as one moves farther east and away from Southeast Asia, this decline in marine species is much less severe than that of terrestrial species (Oliver 1989:248). Therefore, mari ne resources were likely vital in sustaining early island settlers during early periods of domesticated plant and animal introduction and the development of naturally depauperate island lands for agriculture. Materials and organisms extracted from the sea are currently and historically the most significant resources available to the inhabitants of the Pacific Island nations. Not only do Pacific
22 Islanders eat more fish than the rest of the worldâ€™s population, these nations pin much of their hopes for future economic development upon exploiting marine resources. It is also the smallest island groups, with the least amount of dry land, that depend upon marine resources the most (Adams et al. 1999:366). At present, within the coastal waters of the Mariana Archip elago, there are 895 known species of gastropods (Smith 2003:245), 339 known species of bivalves (Paulay 2003b:221), and 1,106 known species of fish, the vast majority of which (1,020 species) inhabit coral reefs (Myers and Donaldson 2003). The bones and s hells of some of these species have been identified from archaeological sites throughout the Marianas and were exploited for a number of purposes. For the two sites considered in the present study, these reasons include subsistence (as evidenced by burnt a nd fragmented shell and bone representing food refuse) and as a source of raw material for the manufacture of tools, fishhooks, and beads. All of the species of fish that have been identified in archaeological assemblages in the Marianas are still consumed and can be caught in the inshore habitats surrounding the Marianas or purchased in local markets. This, however, is not the case with regard to shellfish. During numerous fishing trips on Guam to gather modern comparative specimens, the collection of only two species of large gastropods were regularly observed (pers. Obs., P. Oâ€™Day). These included the spider conch (Lambis lambis ) and the large species of topshell ( Trochus niloticus ), which was introduced to the Marianas an in attempt to establish a commer cial fishery following World War II (Smith 2003:245). Although it appears that collecting shellfish for subsistence purposes is currently uncommon in the Marianas, many of the species of gastropods and bivalves identified during this study are common throughout Asia and the Pacific. People can routinely be seen collecting shellfish from different coastal environments on various islands in the tropical Pacific and ethnographic research showed that this activity is conducted almost exclusively by women
23 (Chapm an 1987; Meehan 1982). Mollusk species like the arc clam, Anadara cf. antiquata (of particular interest for the present study), currently represent substantial parts of important subsistence level, artisanal, and small scale commercial fisheries for women in coastal communities throughout the Pacific (MacKenzie 2001). In some Fijian villages, womenâ€™s daily fishing activities provide an important source of protein for their respective households. Also, many women on Pacific Islands with limited employment op portunities have been able to provide themselves with incomes and gain status within the household through selling portions of their catches (Fay -Sauni et al. 2008). Pre-Contact Fisheries and Their Significance Pre Contact fisheries were integral to human history throughout the world. These resources include fish, mollusks, invertebrates, mammals, reptiles, and plants from both freshwater and marine environments. Among Pacific Island archaeologists, marinebased subsistence practices have been a source of debate in the study of the initial settlement and the histories of social development on islands throughout the region (Davidson and Leach 2001:115). In the present study, prehistoric fisheries were evaluated using the archaeological residues of exploited marine species. These residues include fish bones and invertebrate shells recovered from archaeological deposits. These hard structures represent the basic units of analysis through which prehistoric fisheries populations can be measured, including species composition, relative abundance, size, and estimates of the age of individuals. From these data, size frequency distributions and the age structure of a population can be extrapolated. Human harvesting, or fishing, kills fish and intensive harvests elimi nate old individuals from a population (Cushing 1981:4). When individuals are removed from a population by fishing, the population may respond in a multitude of ways to fill these gaps. This may include fluctuations
24 in the total number of individuals, chan ges in total biomass of the population, shifts in size frequencies, changes to age structure, and spatial distributions. Understanding these responses to fishing is vital and is studied by fisheries scientists in an effort to manage modern fisheries stocks (Haddon 2001:1). Characterizing prehistoric fisheries, especially those based on archaeological residues from early sites, can provide novel information to fisheries scientists who study heavily exploited modern stocks. This is because early sites, particularly those related to early human settlement, may represent an important opportunity to evaluate populations subjected to extremely low levels of human exploitation. According to some researchers, the lack of data from pre industrialized time periods has negatively affected the study of heavily exploited fisheries (Bolle et al. 2004:314). The lack of a historically informed perspective within fisheries science has also led to what is referred to as the â€˜shifting baseline syndromeâ€™ (Jackson et al. 2001; Pauly 1995). Archaeologists working within various cultural contexts throughout the world have attributed numerous landmark developments in the evolution of human societies to the exploitation of abundant stocks of fish and shellfish. Examples of these devel opments include the establishment of permanent settlements, substantial increases in human populations, and a rise in social complexity, especially in regard to chiefdoms (Arnold 1992; Binford 1968; Claassen 1996, 1998; Erlandson 1988; Meehan 1982; Moseley 1975; Moss 1993). In remote Oceania, the exploitation of marine resources, for various reasons, fueled human exploration and the development of numerous and unique societies for at least three millennia. The settlement of the Mariana Archipelago by the an cestors of the present day Chamorro around 3500 cal a BP represents one of the earliest examples in the remote islands of the Pacific of what would develop into a complex society with a maritime economy. Although the indigenous culture of the
25 Marianas expe rienced European contact and permanent colonization prior to any other island group in the Pacific, vestiges of this society still remain. The indigenous language, or Chamorro, is still spoken, especially by older members of the community. This includes a traditional taxonomy of plants, animals, and procurement strategies. To some extent, the traditional fishing knowledge and strategies that were developed over the last three millennia are still employed and have been well documented (Amesbury and Hunter -An derson 2003a, 2008). This is likely true for all island groups in Oceania that were settled by Austronesian speakers, who modified and built upon their existing repertoire of knowledge and strategies for exploiting marine fisheries in tropical insular envi ronments where terrestrial resources were relatively impoverished (Kirch and Dye 1979). The Data Set The primary data for this study consists of archaeological faunal remains and artifacts derived from two archaeological sites located on two islands in th e Mariana Archipelago. These sites include the Unai Bapot 1 site on Saipan Island in the CNMI and the Ritidian Grotto site, on the U.S. Territory of Guam. Excavations of the Unai Bapot 1 site were conducted by Dr. Geoffrey Clark, Olaf Winter, and Patrick O â€™Day in April 2008 and excavations of the Ritidian Grotto site for this study were conducted by Nicole I. Vernon and Patrick M. Oâ€™Day in June 2009. These sites were selected based on several factors. First, previous investigations at both sites identified deep, intact, stratified cultural deposits that contained rich assemblages of marine faunal remains and artifacts. This was essential for obtaining samples large enough to be subjected to statistical analyses. Second, the chronological occupations, or use of the sites were relatively long and overlapped with one another. Lastly, the sites are located in different
26 environmental settings. This was viewed as an important consideration with regard to evaluating the role environmental parameters played in prehis toric marine resource exploitation. The archaeological residues of fisheries from current excavations at the Ritidian Grotto and Bapot 1 sites consist of similar species of marine mollusks and fishes. These include species of gastropods, bivalves, bones from various fishes, and a shark tooth. Although remains of some of the same species were recovered from both contexts, each exhibits different patterns of site use and resource exploitation. The Bapot 1 site is an open site situated on a coastal sand plain at the northern end of sheltered Laulau Bay on the east coast of Saipan and he Ritidian Grotto site is a rockshelter cave complex located at the base of a large limestone escarpment just above the coast plain at the exposed northern tip of Guam. Both quantitative data (utilized in both fisheries science and archaeology) and qualitative observations (based on anthropological, ethnohistorical, ethnographic, and historical research) were employed in the present study. Quantitative measurements were derived from multiple sources. The first are measurements gathered from archaeological specimens of bone and mollusk shells to track changes in prehistoricperiod fisheries populations and evaluate how changing human behaviors relate to marine resource exploitat ion. This primarily involved characterizing the population dynamics of prehistoric fisheries at various points within the cultural sequence in the Marianas through the archaeological record, utilizing methods and models employed in fisheries science. The s econd involved measuring changes in the prehistoric material culture, including artifacts, architectural features, and archaeological site type and function. Lastly, I used information related to past environmental transformations, namely the late Holocene drop in sea level (Amesbury 2007; Amesbury et al. 1996; Athens and Ward 2004; Butler 1994; Dickinson 2000; Carson 2011).
27 Qualitative observations were derived from anthropological, ethnohistoric, ethnographic, and historic research to formulate a framewor k from which prehistoric fisheries and related aspects of Chamorro society can be better contextualized. This involved highlighting beliefs, gender roles, and techniques drawn from historical accounts and ethnographies to create appropriate analogies to better understand prehistoric fisheries. Theoretical and Analytical Methods The broad orientation of this research is ecology, which is the study of interactions among humans, other organisms, and the physical environment (Crumley 1994:3; Reitz and Wing 2008:88). The specific focus of this study concerns the prehistoric ecology of the Austronesian speakers who settled the Mariana Islands approximately 3,500 years ago, the fish and shellfish they exploited, and their local marine environments. This study also follows the basic precepts of historical ecology and views human environme nt relationships as interactive rather than deterministic and views humans as the primary agents of change that can both increase (through management or conservation) or decrease diversity (through overexploitation). Through the implementation of multidisc iplinary research that engages a broad range of data sources, the history of this relationship can be deciphered (Bale and Erickson 2006:1â€“2; Crumley 1994:5â€“7). Because this study focuses on tropical Pacific islands, it is important to consider the unique aspects of island ecosystems and how they are similar to tropical ecosystems in Near Oceania and the islands of Southeast Asia. This large region is the point of origin for all species of plants and animals that colonized the Mariana Archipelago prior to the arrival of humans. This area was also the locus from which the Austronesian speaking agriculturalists who ultimately voyaged to the Marianas acquired their material, cultural, and ideological traditions. The key ecological difference between lands of origin and lands of colonization is that in the remote
28 islands of the Pacific, flora and fauna are less diverse than their paleotropical counterparts in the islands of Southeast Asia. Pacific islands, prior to human settlement, were characterized by a smal l, less diverse suite of terrestrial biota that lacked natural defenses against large predators. Although the coral reefs and marine environments surrounding the Marianas and other islands in the western Pacific are seemingly rich, they are much less diver se than those in Near Oceania and Southeast Asia. These naturally isolated and stable environments had developed long before the arrival of humans and the consequences of human settlement on islands are therefore often viewed as dramatic and negative (Athens 1997:248; Kirch 1997:2; Steadman 1997:51, 2006:41). The long standing natural and cultural connections between Southeast Asia and the Marianas are an important consideration in this study. Whereas initial settlement of the Marianas Archipelago marked an important historical transformation, it should not be considered as a point from which the environmental histories of these islands and island Southeast Asia completely diverged. A less frequent mode of interaction between island Southeast Asia occurred following initial settlement. Settlement of the Marianas about 3,500 years ago and the subsequent development of indigenous society prior to European colonization about 500 years ago, represent only a portion of a larger ongoing interaction sphere. Any meaningful reconstruction of this specific historical interaction requires an understanding of its spatial and temporal orientation within the larger regional historical context. This is best achieved through consideration of largescale and/or long term views of history, what Braudel (1980:27) calls the longue dure . The theoretical perspectives behind much of the current archaeological research conducted in the Marianas is ecologically based; however, these perspectives are more aligned with cultural ecology , cultural materialism, and evolutionary ecology in that these do not view the human environment interaction as dialectical (Bale 1998:4). The present study owes much
29 to this current body of work, which has made significant leaps forward in establishing detailed occupational sequences for the various islands of the Marianas and proposed many productive hypotheses related to human environment relationships within these sequences. To better establish a theoretical framework for the present study, brief descr iptions of ecology in archaeology and the use of an ecological perspective in Marianas archaeology are presented below. Ecological Approaches in Archaeology During the late nineteenth and early twentieth centuries, concepts of evolution, natural selection, and adaptation were being increasingly developed and codified by new research. This led to a transition in biology from a descriptive, taxonomically oriented field to a field dominated by a process oriented perspective. Also during this period, ecology (first conceived of by Ernst Haekel in 1858) emerged as a recognized field of study. During the first half of the twentieth century new concepts were incorporated into this burgeoning field, leading to a shift from the normative, taxonomically driven practice of biology into a science more focused on interactive processes and the application of quantitative methods (Yesner 2008:39). Concerns with interactive processes and ecological models were also adopted by the social sciences in the early twentieth cent ury. The early use of ecological models varied according to different schools of thought. In the U.S., ecological models were first employed by anthropologists and archaeologists, who sought alternatives to Boasian particularism and the formulation of expl anatory frameworks for cultural areas, populations, technologies, economies, and even sociopolitical concerns. Chief among them were the former student of Boas, Alfred Kroeber and his student Julian Steward. Whereas they were both influenced by British functionalism, their adaptation of ecological models was largely driven by a unique American perspective based upon the perceived lack of social complexity of preColumbian peoples.
30 Native peoples in this context were viewed as intrinsically linked to or even part of, the natural environment (Yesner 2008:40). Also important to the advent of ecological frameworks in American archaeology was the fact that early twentieth century anthropologists utilized archaeology to study the historical development of Native A merican cultures. Steward, who conducted archaeological research like Boas and Kroeber, suggested ecological frameworks for both archaeology and anthropology (Steward and Setzler 1938). In 1955, Steward (1955:30) introduced the term â€œcultural ecologyâ€ and described it as a method for studying relationships between people and their environments, stressing that certain aspects of culture factored more significantly in relationships than others. These constituted the â€œcultural core,â€ which includes features th at are most closely related to subsistence economic activities (Steward 1955:37). The cultural core also considered political or religious features to be important if proven to be adaptive (Trigger 1989:291). Less important in governing humanenvironmental interaction were aspects of culture, based on social structure, political organization, or ideology (Yesner 2008:40). Following Steward and the shift from descriptive to process oriented perspectives, ecological approaches in archaeology were further codi fied by methodological and theoretical advances, including new techniques for analyzing faunal materials and artifacts, and applying traditional concepts of ecology, including limiting factors, niches, and biogeographical distributions to human populations . More recently, ecologically based approaches expanded beyond focusing upon environmental impacts on human populations to address human impacts upon environments. Currently, there is a broad spectrum of ecologically based approaches that range from those focused on adaptation or environmentally dominated perspectives to those more focused on human agency (Crumley 1994:3; Yesner 2008:49). The former perspectives include behavioral ecology, systems ecology, sociobiology, cultural materialism, evolutionary ecology,
31 and Julian Stewardâ€™s cultural ecology, which is the perspective from which the others are derived (Bale 1998:2; 2006:3â€“4). The Ecological Approach in Marianas Archaeology Systematic archaeological inquiry in the Marianas began in the early 1920s wi th Hans Hornbostelâ€™s work for the Bernice P. Bishop Museum in Honolulu, Hawai â€˜i (Thompson 1932). Since then, numerous archaeological projects have been completed on the islands of Guam and Saipan. Because of Americaâ€™s strategic military interests in the region, Guamâ€™s current political status as a U.S. Territory, CNMIâ€™s status as a Commonwealth of the United States, and a large U.S. military presence in the archipelago, most archaeological work in the region is related to cultural resource management and federal compliance issues. This has led to the establishment of a detailed prehistoric sequence based upon a significant catalog of archaeological and paleoenvironmental data. Much of the effort in analyzing zooarchaeological remains from prehistoric fisher ies in the Pacific has been focused on the relationship between environmental variations, especially climate fluctuations and differences in prehistoric sea levels, and changes in archaeological material culture (Allen 2006; Morrison and Addison 2008; Morrison and Cochrane 2008). These studies have attempted to explain these variations through applying optimal foraging or encountercontingent prey choice models from behavioral ecology (Broughton et al. 2011; Broughton and Oâ€™Connell 1999; Stephens and Krebs 1986). Others have also applied these models on modern peoples with subsistencebased economies to formulate analogies to study variability in archaeological remains (Bird et al. 2002:458). Many of these have highlighted general trends in archaeological as semblages and others have attempted to quantify the effectiveness of women and children as foragers and how their efforts contributed to the archaeological record (Bird and Bird 2000). This is an important implication as gender and age -
32 specific subsistence practices in Oceania contribute to individual households in different ways (Chapman 1987). In general terms, the techniques and gear utilized in prehistoric fisheries in the Pacific, including the Marianas, resulted in relatively low levels of exploitati on compared to modern fisheries that utilize a much broader array of techniques and gear, and are subject to different economic and political considerations (Kirch and Dye 1979; Kirch and Green 2001). The primary focus of most recent substantive research h as been the human environment interaction as the focal point for explaining key cultural developments in Marianas prehistory. Major trends in the research included the timing of initial human settlement of the archipelago, changes in settlement patterns, evolving artifact types (most noticeably in ceramic vessel forms and surface treatments), shifting resource exploitation practices, agricultural intensification, the appearance of indigenous architecture, and the development of hierarchy (Amesbury 2007; Ame sbury et al. 1996; Carson 2008, 2011; Moore and Hunter-Anderson 1999). Most of this work was heavily influenced by cultural ecological or neo evolutionary perspectives. According to this recent work, variations in prehistoric human demography, pottery and artifact types, marine resource exploitation, and the expansion of terrestrial resource production (Moore 2005) are viewed as adaptive strategies devised to deal with changing environmental conditions or to overcome limiting environmental constraints. In t he early portion of the archaeological sequence this hinges on a high regional, isostatic sea level stand of approximately 2 m above current levels. Sea level began to decline to current levels at approximately the time of initial settlement of the Mariana Islands, around 3500 cal a BP (Dickenson 2000). Variation in the later portions of the prehistoric sequence, such as the development of hierarchy, megalithic indigenous architecture, and increased diversity in mortuary practices and pottery production, have largely been attributed to stabilizing environmental conditions, increasing population, and consequent
33 competition over resources (Graves 1986; 1991, Graves et al. 1991; Hunter-Anderson 2012). Whereas human environment interactions were primary concerns in these important studies, the role of humans as active agents is subjugated by the natural environment. Historical Ecology, the Longue Dure , and Historically Related Cultures The theoretical framework of the current project is intended to explain the specific interaction between prehistoric human inhabitants and marine ecosystems of the Mariana Islands. The supposition underlying this explanation is that humans are keystone species in this dialectical interaction, armed with unique cogitative abilities and highly developed indigenous knowledge concerning tropical island marine habitats. This knowledge was derived from prolonged interactions and mixing between human populations, dating to the initial Pleistocene colonization of the islands of Southeast A sia and the subsequent early to mid -Holocene movements of Austronesian peoples across the region (Donahue and Denham 2010:225). Archaeological residues formed through repetitive social practices within this interaction through time. Some of these residues are detectable across various scales of space and time (Heckenberger 2005:41 42) and likely represent the blueprints from which settlers in the Marianas began to construct the local environment and modify established social structures related to subsistenc e economies and technology. Certain residues, or elements, of these social structures display strong historical continuity in Micronesia with both the distant past and the present. These include basic marine subsistence economies and technology, for which there is a relatively large body of archaeological, linguistic, historic, and ethnohistorical evidence supporting the continuity of these structures over very large scales of time and space. Following Kirch (2010:9), the Marianas, like Hawai â€˜i represents a â€œmodel systemâ€ for studying certain stages in the development of indigenous island culture.
34 There are differing views regarding the development, introduction, and spread of agricultural economies into and across the islands of Southeast Asia. One widely accepted view is that Neolithic agriculturalists first settled in Taiwan from southern China before roughly 7000 cal a BP. From Taiwan, farming and pottery were introduced to the Philippines between about 4000 and 3500 cal a BP, and then spread south into I ndonesia and east into the Pacific (Bellwood 2004; Diamond and Bellwood 2003). The other has been previously mentioned and involves prolonged interaction and mixing between human populations (Donahue and Denham 2010). Although these views clearly differ, b oth highlight the fact that substantial complexity had developed among people who ultimately sailed from the islands of Southeast Asia into the remote of islands of the Pacific. This included a highly developed Neolithic agricultural economy based primaril y upon rice, a retinue of domesticated animals (pigs, dogs, and chickens), and other plant domesticates. A suite of Neolithic artifacts also accompanied early voyagers and included pottery (particularly red slipped wares with impressions, incising, or stam ping), polished stone adzes, a variety of shell artifacts (bracelets, beads, fishhooks, and tattooing chisels), and stone net sinkers (Bellwood 1995:99, 1997:219â€“230, 2002:26; Diamond and Bellwood 2003:601; Donahue and Denham 2010:225; Szab and Oâ€™Conner 2004:621â€“622). Introductions of these artifacts and materials have usually been attributed to the early expansion of Austronesian speakers out of Taiwan. The same material and artifact types, however, could also have been introduced from non-Austronesian sp eaking peoples from mainland Southeast Asia, who were also skilled in pottery and polished adze production, agriculture, and fishing (Higham 2002; Higham and Thosarat 1994; Szab and Oâ€™Conner 2004). Although the exact timing and point of origin for the se ttlement of the Mariana Islands is still debated, human populations that were producing pottery and intensively collecting terrestrial
35 and marine resources in Southeast Asia were firmly established by 3500 cal a BP (Carson and Kurashina 2012; Clark et al. 2010). This timeline illustrates that the study of Marianas prehistory involves considerations of both regional and local scales of interaction. Collectively these spheres of interaction encompass millennia of human environment relationships and the develo pment of vast amounts of indigenous knowledge regarding terrestrial and marine environments of tropical Pacific Islands. The first voyagers to reach the Marianas found unique natural environments, but many of the physical characteristics and species that o ccupied those new environments were likely familiar. This is because the flora and fauna endemic to the Marianas dispersed from Southeast Asia (Steadman 2006). It is therefore probable that they arrived in the Marianas with a well developed understanding of many aspects of local ecological relationships and had developed applicable systems of managing some of the local resources (Berkes 2008:3). This probably included species with large natural distributions that were economically important to various groups of prehistoric peoples like the arc clam, Anadara cf. antiquata. This clam is currently found throughout the western Indo Pacific, northern Australia, eastern Polynesia, Hawai i, southern Japan, east Africa, Madagascar, and the Red Sea. It represents an important subsistence resource in many of these localities and is also commercially harvested in Indonesia (Carpenter and Niem 1998:146147). It is frequently recovered in large numbers from archaeological sites throughout the Pacific (Amesbury 1999, 2007; Clark et al. 2010). It is important to remember that the early settlement of the Marianas was a small part of a much larger episode of human history. This episode encompassed the rapid colonization of the remote islands of the Pacific by sophisticated Neolithic maritime cultures from insular Southeast
36 Asia. The first remote islands to be reached by these Neolithic voyagers were in Melanesia and in western Micronesia around 3500 cal a BP. Voyaging was a complex enterprise with a highly developed system of navigation. It was also a deliberate enterprise and was likely conducted for different reasons, including trade, maintaining social ties, exploration, and settlement of new territories. Settlement, in contrast to other reasons, probably required a more extensive list of cargo including useful plants, animals, other raw materials, tools, and a founding population of women and men (Irwin 1992). This highlights the fact that the primary settlers of the Marianas, as well as other parts of the Remote Pacific, l eft behind long established, historical human environment interactions in island and mainland Southeast Asia that had developed over millennia (Donahue and Denham 2010; Higham and Thosarat 1994). Initial human settlement and the development of indigenous society in the Mariana Archipelago, as reflected in the archaeological record, ethnohistoric, and historical accounts, sprouted from a seed of complex interactions between numerous prehistoric societies in the islands of Southeast Asia. The development of an indigenous culture and alteration of the landscape in the Mariana Islands resulted directly from an elaboration of knowledge, beliefs, and material cargo that early Austronesian voyagers brought with them, coupled with the unique set of environmental ci rcumstances in which the initial settlers found themselves. Within this long trajectory, operating on both regional and local scales, are visible cultural features formed by repetitive social practices. For this study the most pertinent traces are the mat erial remains of the prehistoric marine subsistence economy (e.g., marine shell and fish bone), technology (e.g., ceramics and artifacts), and settlement patterns (e.g., site distribution).
37 Goals, Objectives, and the Problem Goals The overall goals of the present research were to evaluate both long term and short -term trends in prehistoric fisheries populations represented by archaeological residues and identify the human behaviors behind these trends. Expected trends include changes in species composition, species diversity, and changes in the population dynamics of certain species over time. There are probably many prehistoric human behaviors, or cultural practices, that impacted and created trends in marine resource residues. Thes e may have included ancient belief systems and management practices related to fisheries and age, gender, and class roles related to fishing and collecting marine resources. Objectives The primary objective of this project was to develop a complete unders tanding of the roles that humans played in prehistoric interactions between humans and marine environments in the Mariana Islands and establish analogies or homologies between the past and present (Heckenberger 2005:34; Kirch and Green 2001:19) within the context of a Southeast Asian Mariana Archipelago interaction sphere. Secondary objectives included: 1) determining how this interaction factored into the development of prehistoric indigenous culture in the Mariana Archipelago through time, 2) effectively adopting principles and methods from fisheries science to analyze archaeological faunal remains, 3) identifying fine, small scale, short term changes in the archaeological evidence related to prehistoric fisheries (e.g., identifying the initial effects of human predation upon previously unexploited populations), and 4) advancing avenues of research that address more complex issues regarding human environment interactions related to the prehistoric subsistence economies, beyond approaches that necessarily pr ivilege the environment.
38 Research Problem How did the prehistoric human population of the Marianas impact their ancient fisheries? This simple question represents the primary problem this investigation labors to address. Generally in the Pacific, research on how human behaviors relate to variation in archaeological residues of marine resource exploitation is limited. Rather, conspicuous trends in the archaeological residues of marine resources in the Pacific are commonly seen as adaptations to changing or l imited local environmental conditions (Allen 2006; Hunter -Anderson 2012; Morrison and Addison 2008; Morrison and Cochrane 2008). In the Marianas for example, one notable trend is a major shift in the relative abundance of gastropods vs. bivalve species tha t appears in archaeological assemblages dating to the later part of the prehistoric period. This has been attributed to changing environmental conditions specifically related to falling sea level that coincided with settlement of the Marianas (Amesbury 200 7, 1999; Amesbury and HunterAnderson 2003). Another notable trend in some early period faunal assemblages in the Mariana Islands is low frequencies of fish bone (Amesbury and Hunter -Anderson 2003). It appears systematic consideration of this trend has suffered from the nature of contract cultural resource management projects conducted in support of construction and development that currently dominate archaeological research in the Marianas. Despite the fact that the governments of Guam and the CNMI recogni ze the importance of cultural resources and have enacted various laws to protect them, these laws have not been effectively enforced and enjoy only minimal standards of compliance. This has severely limited the collection of data related to past subsistence patterns. Additionally, the limited data that have been collected were often inconsistently reported (Amesbury and Hunter-Anderson 2003:35).
39 In addressing this problem, it was anticipated that the present research would build upon previous work and expl ore alternative explanations that the first trend, theories that invoke processes of optimization and the ranking of resources or are prompted by demographically or environmentally triggered resource depression, cannot account for (Zeder 2012:246). It w as important to begin to address the latter trend, which is related to fish bone frequencies, through both the use of sampling techniques designed specifically for recovering faunal materials and through considering ethnographic and historical data from re gional and local contexts regarding traditional modes of marine resource exploitation. Questions This research seeks to describe the role of fisheries in prehistoric society in the Marianas. Specifically, I integrated multiple levels (broadly defined domains) of ethnohistoric, archaeological, and environmental data derived from different spatial (regi onal, local, and site level) and temporal (prehistoric, historic, and modern) scales (Heckenberger 2005:41). This project therefore, investigated the following research questions in the context of the Bapot 1 site on Saipan and the Ritidian site on Guam: 1. Can humaninduced vs. environmentally induced change be detected through the analysis of archaeological marine fauna samples? 2. Can the initial effects of human predation upon previously unexploited marine resources be detected? 3. Can thresholds of human inter action and related changes in subsistence strategies be detected in prehistoric fisheries? 4. What role (managers, engineers, or adaptive organisms) did indigenous peoples play in the human environment interaction throughout the prehistoric periods of the Mar iana Islands? 5. How does traditional marine exploitation in the Mariana Islands compare to related cultures in Oceania?
40 Hypothesis As previously stated, indigenous society in the Mariana Archipelago sprouted from the cultural, ideological, and technological seeds planted by complex Neolithic Austronesian speaking voyagers who originated somewhere in islands of Southeast Asia by approximately 3500 cal a BP. This is inferred by strong archaeological (Carson and Kurashina 2012; Clark et al. 2010), linguistic (B lust 1995; Pawley and Greene 1984; Pawley and Ross 1993), ethnohistoric (Peterson 2009), and human biological evidence (Vilar et al. 2012) and supports the idea that the Marianas and Southeast Asia represent historically related cultures (Kirch 2010:14). As a starting point for further investigation, the research described herein assumes for the sake of argument that the first settlers to land upon the beaches of the Mariana Archipelago were well equipped with a formative array of technology, traditional k nowledge, a complement of useful plants and animals, and a complex understanding of paleotropical ecosystems with traditional knowledge related to managing them (Anderson 2005; Berkes 2008). They were also skilled voyagers who could have periodically retur ned to their point of origin. Because of this established traditional knowledge, elements of which may have included resource management and propensities to modify environments to suit specific human needs and the relatively low intensity of prehistoric marine resource exploitation, it is presumed that prehistoric human marineenvironment interactions were inherently more complex than is currently depicted. It is also presumed that through the careful application of specific methods from fisheries science i n the analysis of the archaeological remains (e.g., fishbone and shell), remnants of this interaction can be teased out of the archaeological materials. Also, the impacts of this interaction may be discernable from archaeological residues. Hypothetically, prehistoric management practices minimized negative impacts on ancient fisheries, leading to minor changes in the marine resource base over long periods of time. In contrast, environmentally driven exploitation
41 practices were characterized by punctuated ch anges in resource procurement strategies over shorter periods of time. In the former case, I hypothesize that analysis of archaeological residues of ancient marine fisheries would reveal minor, or statistically insignificant, differences between measures o f species abundance, species diversity, and length and age parameters of some species throughout the occupational sequence. In the latter case, I hypothesize that one or more statistically significant changes in species abundance, diversity, and length and age parameters of some species would be prevalent in the occupational sequence. Analytical Methods A methodological framework was devised for the project to integrate data derived from multiple independent sources and fields, that span large periods of ti me, and covers both site level and regional spatial scales. This methodological framework was based on direct historical (Heckenberger 2005:41; Schmidt 1997:24) approach, and sought to establish links between evidence derived from archaeology, ethnohistory , history, linguistics, and fisheries science. Figure1 2 contains a general list of methods employed in this study. These were used to evaluate cultural elements related to marine resource economies among Southeast Asia, other Pacific Island culture histor ies, and the Marianas and explain variations in the archaeological residues. Fisheries Science: Because of the economic and political importance of fisheries and the ecological effects fishing can have upon fish populations (Smith 1994:1â€“5), management of fisheries stocks have long been the primary focus of fisheries science. A population is a group of individual organisms belonging to a single species that have properties which are not possessed by individual organisms (Haddon 2001:19). These properties in clude size, growth rates, a geographical distribution, immigration and emigration, and the groupâ€™s age structure. Fisheries management requires a strong understanding of the dynamics of a population, which include mechanisms of stock production and regulat ion (Ricker 1977:2). The behavioral dynamics of a
42 population are related to changes in the properties of a population through time and it is the aim of the fisheries science methods employed in this study to describe the aspects of this dynamic behavior in the context of prehistoric fisheries of the Mariana Islands (Haddon 2001:19). The fisheries science methods incorporated into this study required adjustments because of the nature of archaeological contexts and because fisheries science and knowledge concerning population dynamics was developed in conjunction with industrialized fisheries. This has led to a lack of data concerning historic or prehistoric fisheries (Bolle et. al. 2004; Smith 1994:3). Because of the fragmentary and transformed nature of arch aeological deposits, appropriate materials and data derived from archaeological contexts for fisheries based analyses are limited. These limitations include reliance upon shells from animals that died thousands of years ago, insufficient samples of remains to perform statistical analyses, lack of direct knowledge of the human behaviors related to the prehistoric fishery (Smith 1983:431), and lack of supplementary life history and environmental data directly related to the sampled archaeological populations (Ambrose 1983:312). Because of the nature of the data, as mentioned above, this study employed density dependent population models to identify changes in mean trophic levels in the archaeological record, related to either direct human action or environmental action. According to these models, all populations are limited in size or live in limited environments and population size is regulated through natural controls. Fishing changes the abundance of mature organisms within a population (Cushing1981:1; Ricker 1954:560). Density dependent responses to changes in abundance may include changes in growth rates or the age structure of a population (Haddon 2001:37). The main goal in this analysis was to identify changes in the dynamics of prehistoric fisheries populations and determine if they were caused by changes in human exploitation
43 patterns. Prehistoric marine subsistence economies were likely small in scale and the negative impacts related to human predation may have been relatively small. Human induced changes in prehistoric fisheries populations were probably brief punctuated moments within the stable flow of history. These punctuated moments were directly related to changes in new fishing strategies or technologies, which may have been correlated with chan ges in other forms of material culture and social structures at various points through time. Thesis Organization From preceding sections of this chapter it can be seen that previous archaeological work in the Marianas has added much to the understanding of prehistoric marine resource exploitation in the Pacific. While this body of previous work has accomplished much and provided a solid foundation for the present research, the consideration of faunal remains in the Marianas is dominated by environmentally b ased interpretations and, generally, has not been subject to systematic testing that produces comparable data sets. Thus, the human impacts upon prehistoric marine fisheries in the Mariana Islands remain unknown. For these reasons a major part of this rese arch involved the use of basic archaeological and zooarchaeological techniques, and applying methods from fisheries biology. The next four chapters contain the description and analysis of the archaeological residues of marine resource exploitation obtained during fieldwork. Chapter 2 provides background information regarding the physical, cultural and historical contexts of the Mariana Archipelago. Chapter 3 describes the archaeological field and laboratory methods employed, results of excavations with descriptions site stratigraphy and recovered materials, and results of all laboratory analyses. These include radiocarbon dating, artifact and faunal analyses, marine resources analyses, and human remains analysis presented by site under separate headings (Bapot 1 and Ritidian Grotto). Chapter 4 covers the methods and results of the fisheries science-
44 based analysis of marine invertebrate shells for the present project. Methods utilized for the study include shell measurements, aging samples of shell, and statistical analyses. Chapter 5 includes a discussion of the results and presents the empirical findings and conclusions of the res earch. This final chapter also discusses the implications, limitations, and the future of the research.
45 Figure 1 1. Map of the islands of Micronesia, Melanesia, and Polynesia and their location relative to the Andes ite Line.
46 Figure 1 2. Methods Relevant to Study. Site Level Local Scale Regional Scale Review of ethnohistoric, historic, indigenous knowledge, and ecological and fisheries science data related to marine resources in the western Pacific Review of ethnographic, historical, linguistic, and archaeological data related to the Marianas Site formation processes Identification and descriptions of culturally deposited fish and shellfish remains including associated habitat preferences and collecting behaviors Artifact types and function Review of marine environments and species in the Marianas Fisheriesscience based analyses
47 CHAPTER 2 PHYSICAL, HISTORICAL, AND CULTURAL CONTEXTS Introduction The following chapter provides background information regarding the physical, cultural and historical contexts of the Mariana Archipelago. Ethnohistoric information related to marine resource exploitation in the Pacific is also discussed. This includes reviews of ethnohistoric data related to the western Pacific region and data specific to the Mariana Archipelago and Cham orro culture. The first section of this chapter provides information regarding the physical or environmental properties of the remote islands of the tropical Pacific, the Mariana Islands, Guam, and Saipan. The second section provides information regarding the cultural context related to the project. This will familiarize the reader with the geography and history of the region and provide useful contextual information for the analysis and interpretation of the archaeological data. Physical Setting Numerous fields of research have proposed significant boundaries for the Pacific Basin, based upon the study of formative geological properties of islands, the movement and colonization of flora and fauna, and the consideration of ethnographic and linguistic data re lated to the distribution of human populations in this region. Respectively, these include the boundaries known as the Andesite Line, the cultural areas of the Pacific (including Melanesia, Micronesia, and Polynesia), the dividing line between Near and Rem ote Oceania, and the distribution of the Austronesian language family, which encompasses most of the indigenous languages of Southeast Asia and all of the indigenous languages currently spoken in the Pacific (Tryon 2006). Within this broader context, a dis cussion of how the Mariana Islands fit into
48 current archaeological concerns in this region and how the sites and materials of the current study relate to current research trends will be presented. The various island groups of the tropical Pacific occupy a vast expanse of open and almost featureless ocean. Throughout this vast expanse there are numerous island groups that display many striking cultural and environmental similarities that are immediately apparent to anyone familiar with the region. Even to th e seasoned traveler these similarities seem to significantly outweigh the dissimilarities between the island groups. This has led to the proposition of various regions, or boundaries, in an effort to describe, or classify, environmental conditions, the dis persal of plants and animals, and the distribution of human populations in the Pacific. Regional distinctions, like â€œMicronesia,â€ have been in use since the mid nineteenth century, however no consensus regarding these regional distinctions has been reached among modern researchers and it has become clear that these distinctions are problematic in the context of current anthropological research in the Pacific islands (Kirch and Green 2001:63). As Peterson (2009:14) points out, modern researchers are currentl y faced with some of the same problems scholars were faced with in the nineteenth century on topics that consider the distribution of human populations in the Pacific. These problems are related to deciding when to emphasize the differences and when to emphasize the similarities. The many similarities observable in all of the island groups throughout the tropical Pacific are a consequence of the fact that the different islands were formed by the same evolutionary processes and colonized by plants and animals in much the same manner (Kirch 2000; Steadman 2006). Thus, the various island groups consist of essentially the same environmental constituents (i.e., similar geological formations, flora, and fauna). Between island
49 groups and between different islands within the same group, how ever, these environmental constituents exist in different frequencies or densities based upon their evolutionary history. For example, all of the archipelagos in the Pacific are comprised of islands at different stages of geologic development, and range fr om younger, less weathered islands that lack extensive fringing reefs, streams and soil development, to older, highly weathered islands with stream eroded valleys that contain rich soils and large, protective fringing reefs. This environmental gradient bet ween islands within an archipelago determined the amounts and availability of vital resources and was important to early settlers and ultimately played an important role in the development of more complex societies late in the prehistoric period (Kirch 200 0:46). Island groups such as Palau ( Figure 11) are comprised of both geologically young, high volcanic islands and relatively old and small, low lying atolls. Environmental similarities are also evident in regard to the rich marine habitats that surround many Pacific islands and how both prehistoric and modern indigenous people exploit them. Currently, many of the same species of fish and shellfish can be caught in different island groups throughout the Pacific. Numerous species of highly prized reef fish and shellfish are in the catches of local fishermen in the Marianas, the Philippines, Palau, Fiji, Hawaiâ€™i, and the Marshall Islands (pers. Obs., P. Oâ€™Day). These include species such as the unicorn tang ( Naso unicornus ), rabbit fish (Singanus sp.), various species of parrotfish (Scarus spp.), turban shells (Turbo spp.), species of conch (Strombus gibberulus gibbosus and Strombus mutabulis mutabulis ), species of Venus clams ( Gafrarium pectinatum and G. tumidum ), numerous species of giant clam (Tridacna spp.), and ark clam, particularly Anadara cf. antiquata. These fish and mollusks appear to be taken by similar means and cooked in much the same way throughout the island groups mentioned above. The remains of many of the same species of fish and shellfish hav e also been observed in faunal assemblages from
50 archaeological sites in Palau, the Marianas, Fiji, Hawai i, and the Marshall Islands (pers. Obs., P. Oâ€™Day). Mangroves are a particularly important coastal habitat in the Pacific Islands. They are a source of organic carbon and nutrients for surrounding marine environments, contain several types of important raw materials for human populations, and protect coastlines from erosion. This habitat is a significant source of detritus and nutrients that support man y species of fish, crabs, and shellfish that are exploited by human populations. Mangroves also provide fuel, construction materials, and compounds used in dyes and medicines, among other resources important for human subsistence (Woodroffe 1987:180). Addi tionally, this habitat supports nurseries for several species of juvenile reef fish (Nagelkerken et al. 2000) and is used as breeding ground for various species of birds; mammals; and reptiles. For these reasons, the current destruction of mangroves throug hout the world is of major economic and environmental concern (Alongi 2002). The most observable dissimilarities among the island groups, aside from obvious location differences, are products of the relatively recent history of colonization by numerous Eur opean and Asian entities in the region. Foreign colonization of the Pacific is characterized by a succession of different people and organizations with diverse goals, representing numerous nations including Spain, Britain, France, The Netherlands, Portugal , Germany, Russia, the United States of America, China, and Japan. Although, these powers asserted their influences in different locations and at different times in history, the general process of colonization has been constant for at least the last three centuries (Torrence and Clarke 2000:4). Because the Mariana Islands are located along historic sailing lanes between Mexico and the Philippines, the Spanish colonized these islands centuries before other European nations made major inroads into the
51 island groups of the Pacific (Hezel 1983). The Spanish imposed their beliefs and customs in the Marianas with the same zeal that they had exercised in the Philippines and the Americas. This resulted in extensive social and religious changes in the indigenous cult ure before the traditional aspects of life in the Marianas could be recorded by the more scientifically minded colonists of the late nineteenth and early twentieth centuries (Peterson 2009:33). Therefore, much of what is known of traditional culture in the Marianas is derived from a few, often biased, historical texts (Barratt 2003; Thompson 1945) and extrapolated from multiple fields of inquiry including linguistics (Blust 1995; Ross 1989), ethnography (Fritz 2001), archaeology, physical anthropology (Hans on and Pietrusewsky 1997), and the study of modern island peoples (Peterson 2009). The Marianas are the northernmost islands within the vast expanse of islands known as Micronesia (Kirch 2000; Petersen 2009). Micronesia refers to one of three geographic a reas defined by scholars in the nineteenth century for the purpose of mapping the human populations of the Pacific. The other two include Melanesia and Polynesia ( Figure1 1) and were proposed by M. G. L. Domeny de Rienzi at a meeting of the Paris Socit d e Gographie in 1831. During a views regarding these classifications, which were subsequently published (Petersen 2009:13 14). Although both used the same terms (M conception of these geographic areas influenced all subsequent Pacific ethnologies and is cited Micronesia consists of ap proximately 2,100 islands spread out over a portion of the Pacific Ocean covering roughly 7.4 million km2. If the landmasses of all of the islands in Micronesia were combined they would account for only 2,700 km2 of dry land (Kirch 2000:65). The total
52 comb ined area of Guam and Saipan is approximately 672 km2, which accounts for about one fourth of the total area of dry land in Micronesia. The Andesite Line is viewed as an important factor in the distribution of both natural biotas and human settlement. The Andesite Line marks the points at which the Pacific Plate, the Indo -Australian Plate, and the Philippine Sea Plate meet and divide Oceania into two geologic regions (Bellwood 1979:20). It also marks the boundary between Near and Remote Oceania (Wickler 2002:30). West of the Andesite Line and situated on the Indo-Australian Plate, are the island groups of New Zealand, Tonga, Fiji, the Solomon Islands, the Bismarck Archipelago, and New Guinea. Also west of the Andesite Line, but located on the Philippine Sea Plate, are Palau and the Marianas (Intoh 1997) ( Figure 11 ). Islands in this portion of the Pacific Ocean are typically more geologically diverse, with lighter â€œandesiticâ€ minerals characteristic of continental crust. On the Pacific Plate and east of the Andesite Line are the volcanic islands like the Hawaiian Archipelago, comprised of dark, heavy rock, characteristic of thin oceanic crust. These islands are geologically young, less than 5 million years old, in comparison to islands such as New Zealand and New Caledonia, which are Gondwanaland fragments, originally associated with the continents of Antarctica and Australia (Steadman 2006:7). This boundary is significant in regard to the spread of human populations throughout the Pacific because the islands west of the Andesite Line are also generally larger, more biologically diverse, and are naturally imbued with more marine and terrestrial resources. These islands also have fertile land for cultivation and better raw materials, such as superior mineral res ources for lithic tool manufacture (Kirch and Green 2001). Ancient Near and Remote Oceania are regions that were originally conceived of by Roger Green (1991) in an effort to deal with some of the inadequacies imposed upon research by the
53 early categories of Melanesia, Micronesia, and Polynesia (Kirch and Green 2001:63). Near Oceania is comprised of New Guinea, the Bismarck Archipelago, and the Solomon Islands. Human settlement of Near Oceania began approximately 50,000 years ago from the islands of Southea st Asia (Irwin1992:3). Remote Oceania, which includes Melanesian, Polynesian, and Micronesian island groups, is enormous and consists of all of the Pacific islands east of this boundary ( Figure 11). In stark contrast to Near Oceania, human settlement in Remote Oceania began between 4,300 and 3,500 years ago, possibly in the Marianas (Athens et al. 2004; Athens and Ward 2004). This boundary (like the Andesite line) also marks a gradient in levels of biodiversity. Near Oceania is markedly richer than the islands of Remote Oceania and the eastern edge of Near Oceania also marks the extent of Pleistocene human settlement (Wickler 2002:30) . Though these regional boundaries have been proposed out of the perceived inadequacies of the throughout this dissertation as both are useful for discussion of human settlement, resource exploitation, and social development in the Marianas. In terms of environment and initial settlement of the western archipelagos of Remote Oceania, cultural areas such as Micronesia and Polynesia are problematic because island environ ments throughout Remote Oceania are comprised of similar elements. Also, much of what is known about early settlement of the islands groups located immediately east and north of the Near Oceania boundary line (i.e., the Marianas) is known from archaeologic al research that highlighted similar traits, including early site distributions, pottery styles, and artifact types. Therefore, categories such as Micronesia and Polynesia are useful for detailed considerations of cultures related to specific island groups , linguistics, and recent histories. The following sections of this chapter will discuss the
54 environmental contexts of the Marianas Archipelago, the Islands of Guam and Saipan, and the environmental variables specific to the Bapot 1 and Ritidian Grotto sit es. The Mariana Islands Embedded in the chronicles of violent events surrounding Magellanâ€™s brief and ill fated stay in the Marianas are the first descriptions of fishing. This is seen in the accounts of Antonio Lambardo, or Pigafetta, who wrote the most substantial and factual description of this first instance of European contact with the Chamorro people (Barratt 2003:5). Unfortunately, like many of the descriptions of Chamorro fishing recorded by subsequent early visitors, Pigafettaâ€™s is brief and only describes the catching of flying fish ( Exocoetus spp.) using fishhooks made of fishbone. The account of Juan Pobre de Zamora, a Franciscan Friar who stayed on Rota for seven months in 1602, further describes fishing for flying fish and other pelagic specie s, such as mahi mahi ( Coryphaena equiselis ), marlin (Istiophoridae), and sharks (Elasmobranchii) (Driver 1993:15). Whereas this is the most detailed account of fishing from the early Spanish period in the Marianas, inshore fishing strategies are not mentio ned (Amesbury et al. 1986:5 7). These very general descriptions persisted, such as Friar Peter Cooman's passage written in the late 1600s regarding night fishing from canoes, shoreline hook and line fishing performed by women, and superstitions related to fishing (Coomans 2000:14). Perhaps one of the earliest, most detailed descriptions of fishing and shellfish gathering is by Freycinet, who visited the Marianas in 1819. Freycinet describes the technologies used at the time, including hooks, spears, the ach o (baiting device, used to catch atchuman, or Decapterus spp.), nets of varying type, and fish pools and enclosures built from mud barriers and reed dikes, analogous to the stone enclosures built during ancient times. He also describes the use of some of t hese technologies in the capture of various species of fish, including magnahak , or manahoc (rabbitfish, or Percoidei), laguas (parrotfish, or Scaridae), gaaga (flying fish), anaho (mahi mahi), and eels (Anguilidae), and provides short
55 descriptions of fishing for crabs and turtles and the collection of shellfish by women along the coastal rocks and sands (Freycinet 2003). Spanish accounts following Freycinetâ€™s descriptions continue to be brief and undetailed, usually describing the abundance and flavor of fish and emphasizing the exclusive use of inshore habitats by the Chamorros (Sanz 1991:7). Some accounts mention the taking of other species of fish, such as atulay (horse mackerel) and tiao (goatfish), and provide more detail on the use of the acho, a prac tice that has been passed down from ancient times (Ibez y Garca 1992; Olive y Garca 2006). Others, like Marche (1982) provide biased descriptions of Chamorro fishing. During his 1887 visit to the Marianas, Marche (1982:8) reported that because of their â€œlazy lifestyle,â€ little is taken from the sea and fishing is only done within the reef. In 1899, after Germany had purchased the Northern Marianas from the waning Spanish Empire, George Fritz (2001) assumed the post of District Administrator and took a k een interest in Chamorro language and culture. In his writings on Chamorro marine resource exploitation he elaborated in greater detail on the traditional technologies and techniques still in use during the late nineteenth and early twentieth centuries. He provides a fairly comprehensive description of targeted species, including clams and snails collected by women and children on the reef at low tide. He also describes two preserved types of ancient fishing that were still used during his tenure in the Mar ianas: 1) atcho poco, a chumming device used for catching atchuman (Decapterus spp.); and 2) use of a lure fish. This latter fishing involves tying a thread to the pierced jaw of a parrotfish. The tethered fish was allowed to swim over the reef and when th e fish was pulled back in with the thread, other fish followed and were then taken with a net. This tethered fish was then tied to a stone or confined in a trap for future use (Fritz 2001:71).
56 Between 1900 and 1940, under the administration of the U.S. Navy, political and economic conditions on Guam improved. Basic civil liberties, however, did not improve during this period, and until 1941 Chamorros still relied on a subsistence economy (Rodgers 1995). Everyday life for Chamorros in the Northern Marianas, following Spanish rule, first under German governance between 1899 and 1914 and then under the Japanese Mandate from 1914 until 1944, also witnessed large economic, political, and social changes. However, throughout both colonial periods, Chamorro society maintained a core body of distinct customs centered on the family, church, and farm and the people were able to maintain their language along with many other customs (Spoehr 1954). This appears to include some traditional methods of marine resource exploit ation. Since World War II, marine resource exploitation has shifted from traditional subsistence to more modern modes of subsistence, commercial, and recreational forms of fishing. This included changes in techniques and materials. Gear composed of natural materials was replaced by modern manufactured materials, including monofilament lines and nets, metal hooks, poles, reels, spear guns, and SCUBA gear, and those changes led to large declines in all inshore fisheries (Hensley and Sherwood 1993:130). Micronesians, like all Pacific Island societies, rely heavily upon the sea as a source of food and raw materials. This was likely more so for pre Contact inhabitants of the Mariana Islands than it is for its current inhabitants. In the different coastal environm ents (i.e. lagoons, mangroves, flats, and reefs) adjacent to both the Bapot 1 site on Saipan and the Ritidian site on Guam one can observe individuals engaged in different marine resource procurement activities and one can easily find many species of fish and shellfish that have been collected at the same locations for thousands of years. Unlike today, however, marine resources, as seen through the archaeological evidence, appear to have been much richer on the reefs adjacent to the Bapot 1
57 and Ritidian sit es. This is because those areas were likely subjected to lower levels of exploitation (Kirch and Dye 1979) and possibly strict management policies (Johannes 1978:351). It appears that some aspects of Traditional Chamorro culture survived approximately 500 years of colonial rule. Some surviving aspects of traditional culture appear to be related to fishing and shellfish gathering practices that have been handed down from prehistoric ancestors and were likely practiced until relatively recently, in the mid tw entieth century. One of these included gender divisions related to the collection of shellfish. The prevalence of this relationship between women and shellfish gathering in island cultures throughout Oceania (Chapman 1987) attests to the antiquity of this practice. Saipan Saipan consists of approximately 123 km2 of dry land. It is the second largest island in the Mariana Archipelago and the largest land mass within the CNMI (Athens and Ward 2005; Moore et al. 1992:3). Like the other limestone islands of the group, Saipan can be topographically described as a system of superimposed limestone terraces fringed by cliffs that mark past sea levels (Spoehr 1954:XXII). The island can be divided into six geomorphic units including the axial uplands, low limestone pl atforms, low terraced beaches, Donni clay hills, the western coastal plain, and the southeastern coastal fault ridges (Cloud et al. 1956). Because of high porosity of its soils, Saipan has no perennial streams, lakes, or ponds aside from Lake Susupe, which is located in a wetland on the southwestern coastal plain. The lake has no surface outlet to the sea and it is believed this lake formed in sinkholes in the underlying limestone. Freshwater is found mainly in springs located on the flanks of the uplands ( Athens and Ward 2005). Unai Bapot -1 (CNMI Site No. SP -1-0013). The Unai Bapot 1 site (CNMI Site No. SP 10013) is located at the eastern end of Laulau (Magicine) Bay on the east coast of the Island of
58 Saipan ( Figure 2 2). Bapot 1 is one of three locations containing the remnants of latte2 structures at the north end of Laulau Bay. These are located east of the large Laulau site and were first identified by Spoehr (1957) ( Figure 22). The sites are situated on a sandy coastal plain bordered on the north by limestone terraces and outcrops dating to the Pleistocene and Miocene (Clark et al. 2010:22). Bapot 1 contains the remnants of at least two latte sets ( Figure 2 3 and Figure 2 4). Similar to previous investigations, the current project encountered deep deposits of cultural materials mixed with coastal sands which abruptly gave way to sterile l ightly colored beach sands at the base of excavations. Behind the site, there is a limestone rock bluff which partitions the beach sands from upland red clays soils. The limestone bluffs and outcrops are intermittently distributed along the bay and in some areas the sandy coastal soils rise gradually to meet the outcrops and red clay soils. The soils at Bapot 1 were primarily wind deposited and were derived from beach sands blown inland from the prevailing east and northeasterly winds. Storm deposited layer s were also noted in previous excavations in other locations of the site. While these events may have temporarily accelerated the deposition of soils at portions of the site (Marck 1978:5), no storm deposits that normally consist of a layer of sterile sand were sectioned during excavations for the current project. It appears that the soils accumulated at a relatively steady pace through the prehistoric occupation of the site. Guam Guam is the southernmost and the largest island in the Marianas, comprised o f 549 km2 of dry land (Taboroi et al. 2005). It is also the most populated and commercially developed island in Micronesia (Steadman 2006). Guam is divided into two distinct physiographic provinces by the Pago -Adelup Fault line ( Figure 2 5). The northern region consists of a low 2 L atte sets are large pillars with capstones placed on top. These are made of quarried limestone, coral, or basalt and may have been foundations for large wooden structures (Carson 2012a:1).
59 relief limestone plateau and the southern region is comprised mostly of long, low ridges with steep faces and an uplifted lim estone component located on the eastern coast (Taboroi et al. 2005). The Orote Peninsula, although south of the fault line, is composed largely of uplifted limestone, similar in character to the northern region. Rainfall on Guam averages 216 â€“292 cm per ye ar (Gingerich 2003). Vegetation has been affected by frequent typhoons and widespread human disturbance, including military use of the island and devastation during World War II (Liston 1996; Oâ€™Day and Vernon 2011). Ritidian Grotto Site. The Ritidian Grott o site is located at the northern tip of Guam ( Figure 2 5), an area characterized by limestone plateaus and terraces (Kurashina 1990). The Ritidian Grotto Site is a rockshelter ( Figure 26) and cave complex ( Figure 27) located at the base of a large limestone escarpment ju st above a sandy coastal plain ( Figure 2 8). Several natural processes have contributed to the formation of the site. The rockshelter formed as an undercut or notched portion of the escarpment which was likely situated in the intertidal zone when sea levels were much higher during the Mid -Holocene. These intertidal notches were dissolved by biochemical and biomechanical actions of seawater and encrustin g organisms during tidal changes (Emory 1962:B64). As sea levels fell and the sandy coastal plain formed in front of the Ritidian Grotto site by wave deposited sediments. The carbonate soils, or sand, of the coastal plain are derived from carbonate detritu s from the coral reef in front of the site (Short 2005:220). The cave and rock shelter features of the Ritidian Grotto site were further eroded through dissolution of the bedrock by water (T brosi et al. 2005). Ritidian is the northernmost point and forms the northwest corner of the Island of Guam. The shoreline near the Ritidian Grotto Site is picturesque, with a narrow white sand beach and fringing coral reef. Ritidian Point is exposed to the open ocean from the west and north and is
60 regularly pounded by high surf. It is also often windy, which conspires with the large waves to create strong inshore currents throughout much of the year. These factors create relatively dangerous conditions that can limit access to portions of the local inshore marine envir onment. Cultural Background The Marianas and the islands of Remote Oceania were settled by people who spoke a language that belongs to the Austronesian Language family. This language family is thought to be comprised of approximately 1,000 to 1,200 distinct languages (Belwood et al. 1995:1; Kirch 2000:7) and is the most widely dispersed language group in the world, covering half the circumference of the globe, extending from Madagascar to Easter Island and from Taiwan in the north to New Zealand in the south (Belwood et al. 1995:1; Pawley and Ross 1993:1 2). Austronesian languages are believed to have derived from a single parent language that was probably spoken on Taiwan approximately 5,000 to 6,000 years ago (Belwood et al. 1995:4; Blust 1995:458459). A ustronesian speakers were likely the first proficient voyagers and colonized islands beyond Near Oceania that were previously unreachable to the ancient Pleistocene inhabitants of Melanesia (Pawley and Ross 1993:2). Also, this diaspora was achieved primari ly through the colonization of new uninhabited lands such as the remote islands of the Pacific and of inhabited islands of Near Oceania, where older indigenous peoples with different languages were replaced by Austronesian -speaking colonists (Bellwood et al. 1995:3). The widespread distribution of Austronesian languages was first noticed by early Europeans, who, by the beginning of the seventeenth century had commented on the similarities between languages spoken in Madagascar and Indonesia. With the second voyage of Captain James Cook at the end of the eighteenth century, it had been established that Austronesian extended all the way to the islands of eastern Polynesia. Because of a relatively long history of research in the Pacific region, comparative ling uistics has made major contributions to the study of prehistoric
61 archaeology. Primarily, this includes lexical reconstructions of prehistoric vocabularies related to numerous aspects of Austronesian cultural traits. These are fishing, food production, voya ging, and kinship. Aside from serving as a supplementary tool for the interpretation of material culture, linguistics have also yielded important data related to possible points of origin for different groups of prehistoric settlers of the different archip elagoes of Remote Oceania, and have been a productive source of guiding hypotheses for research concerning Pacific prehistory (Blust 1995:454, 2002; Pawley and Green 1984). Chamorro, the indigenous language still spoken in the Marinas, has been of particul ar interest to linguists for more than a century (Reid 2002). All of the inhabitants of the islands of Remote Oceania speak languages that belong to the Oceanic subgroup of Austronesian languages, except for the inhabitants of Palau and the Mariana Islands (Blust 2000:83). These are generally seen as outliers in the Remote Pacific, belonging to their own Austronesian subgroup (Malayo Polynesian), and are considered to be more strongly related to the languages of the Philippines and Indonesia (Zobel 2002). S ome specific points of origin have been suggested for the Chamorro language, including the island of Sulawesi in Indonesia (Zobel 2002:432) and the Philippines (Blust 2000:109; Reid 2002:87). Even if a specific place of origin for the Chamorro and Palauan languages ultimately proves elusive, there is a clear consensus that these Malayo Polynesian languages originated somewhere in the islands of Southeast Asia (Kirch 2000). Similarities in design motifs on red slipped pottery from sites in the eastern islan ds of Southeast Asia and the Marianas have long been noted by archaeologists, suggesting close ties between these areas (Hung et al. 2011:910; Solheim 2003:3). Furthermore, based on similarities in pottery decorations and material culture, Hung et al. (2011) and Carson and Kurashina
62 (2012:431) recently identified northern Luzon as a possible point of origin for early Pre Latte peoples in the Marianas. This research focused upon archaeological sites and materials that date to the preContact period of human occupation in the Marianas Archipelago. Although all the islands in the Marianas currently share a common indigenous ancestry and language, the islands of Guam and Saipan have developed unique local histories that are different from one another. This is e specially true during later historic periods that involved successive instances of European, U.S., and Japanese colonization. More recent events related to World War II also played a substantial role. During these historic periods, there were massive chang es in indigenous populations, settlement patterns, language, religion, politics, economics, and subsistence. The environmental conditions on all of the larger southern islands were extensively altered by industrial agricultural practices, widespread destru ction caused by World War II (Oâ€™Day and Vernon 2011), commercial development, and invasive species of plants and animals. Modern historic processes have also led to the establishment separate political entities, namely the U.S. Territory of Guam and the CN MI, which includes Rota, Tinian, Aguiguan, Saipan, and the northern islands of the archipelago (Tomonari Tuggle et al. 2005). These historic periods factored into the preservation of the archaeological sites considered in this research and impacted previou s archaeological research conducted in the Marianas. Because of this, a brief description of each historic period directly related to the current investigation is presented below. These historic periods are also shown in Figure 2 9. Archaeological Background The Prehistoric Period in the Marianas spans at least 2,000 years, from about 3,500 years to 500 years ago, as indicated by radiocarbon dates from pottery bearing sites, including the Achugao, Chalan Piao, and Unai Bapot sites on Saipan, the Unai Chulu and House of Taga sites
63 on Tinian, and the Ritidian site on Guam (Butler 1994; Carson and Kurashina 2012; Craib 1999). Whereas the earliest radiocarbon dates from archaeological contexts cluster around 3500 cal a BP, the timing of initial settlement in the Marianas has not been completely resolved. Data derived from paleoenvironmental investigations on Guam suggest that humans were present by 4300 cal a BP. Because environmental evidence for settlement prior to 3500 cal a BP is indirect, it is not widely accepted by archaeologists working in the region. Basically, charcoal particles begin to appear in sections of paleoenvironmental cores dating to about 4 325 cal a BP. Because there is no charcoal in core sections during the 5,000 years preceding this date, Athens and Ward (1999:148) argue that the appearance of charcoal at 4325 cal a BP marks the introduction of fire into the Marianas by humans. Paleoenvir onmental data collected by Athens et al. (2004:16) from other islands displays a similar pattern, or time lag, between indirect but substantial evidence for the presence of people by 4300 cal a BP and full blown settlement evidenced by the accumulation of food refuse, locally made pottery, and other artifacts by 3500 cal a BP. Critics of settlement prior to 3500 cal a BP largely cite environmental limitations brought on by high mid -Holocene sea levels (Dickinson 2000). Some argue that the high sea levels pr ior to 3500 cal a BP rendered the Marianas unapproachable and uninhabitable to seafaring peoples from island Southeast Asia (Hunter -Anderson 2011:22). Early archaeological work in the Marianas first proposed two prehistoric periods based on pottery types a nd the presence of latte structure foundations. These were described by Spoehr (1957) as the Pre -Latte and Latte periods of prehistoric Marianas culture. Although the particulars of these prehistoric periods have been modified and expanded (Moore 2002), current archaeological research in the Marianas still relies on Spoehrâ€™s basic chronology.
64 Pre-Latte Period (1000â€“3500 cal a BP) The number of known Pre -Latte sites on Guam remains low, possibly because the deposits are deeply buried and have remained undisturbed during commercial development. Sites with Pre -Latte Period components are more numerous in Saipan. The Pre-Latte period b egins at approximately 3500 cal a BP and concludes with the onset of lattebased construction at around 1000 cal a BP. Based on pottery, Moore (2002) proposed four phases for the Pre -Latte period ( Figure 2 9). These are discussed in detail below. Early Unai Phase (3000â€“3500 cal a BP) The first 500 years following the earliest dated archaeological sites constitute the Early Unai Phase of the Pre -Latte Period. Settlements during that period were focused on coastal locations with access to marine resources. Cultural materials associated with this phase include a high proportion of bivalve remains (Amesbury 1999; 2007), thinwalled pottery, stone and shell tools, fishing equipment, and shell ornaments (Bath 1986; Graves and Moore 1995). The dominant pottery type during this period was nondecorated Marianas Redware (Moore 2002:7). The name for this pottery type is taken from the red or reddish brown exterior coloring of the vessels. The Marianas Redware vessels are characterized by thin walls, restricted openings, everted rims, and slipped exteriors. Styles of decorated pottery have been recovered from various sites throughout the archipelago in very small n umbers. These include Achugao Incised ( Figure 210 A ) sherds decorated with complex rectilinear motifs with zones filled with small punctations and San Roque Incised ( Figure 210 B ) sherds with curvilinear garland designs made by linking incised arches with stamped circles or punctations (Butler 1994:2 7). Decorated sherds with these design types are often small and do not decisively display vessel morphology. They are often dark gray in color and average 5 to 6 mm thick. Tempering typically consists of calcareous sand, though some volcanic sand mixing h as been observed (Moore 2002:7).
65 Middle Unai Phase (2500 â€“3000 cal a BP) The Middle Unai Phase spans the next 500 years following the Early Unai Phase of the Pre Latte Period. Archaeological assemblages appear very similar to the Early Unai Phase, with the exception of pottery. The incised Achugao and San Roque pottery are replaced by limefilled or bold line designs called Ipao Stamped ( Figure 2 11 A , Figure 2 11 B , and Figure 2 11C) and Tarague Str iated pottery ( Figure 2 11D and Figure 2 11E ). Tarague Striated consists of scrape marks or impressions arranged parallel to the rim on the inside or outside of the vessel (Moore 2002:89). Other surface treatments have also been observed on pottery dating to this phase including fingernail impressi ons ( Figure 211 F ), dots, and ridges ( Figure 212) (Moore 2002:8â€“ 9). Marianas Redware is also present in assemblages dating to this phase. The commonly unthickened rims are often everted or flared and some vessels may have been carinated. Calcareous tempering is typical, though some volcanic sand tem pering has also been observed. As in the earlier phase, the surfaces may be plain or slipped. A new surface treatment during this period is polishing or burnishing. Wall thickness ranges from 6 to 20 mm, with a mean thickness of 9 to 10 mm (Moore 2002:9). The number of sites increased during this phase and they were usually located along the coastline and in caves and rock shelters not far from the shore (Fulmer et al. 1999). The next 900 years constituted the Late Unai Phase of the Pre -Latte Period. Again, the primary characterization for this phase is derived from pottery attributes. During this phase, vessel design forms became less complex (Moore 2002:10). A small percentage of Ipao Stamped ware is still present, alth ough, by the end of this phase designs were only applied to the rim ( Figure 2 12). Flat bottomed ceramic pans are characteristic of this phase, with w all thickness
66 ranging from 20 to 40 mm. The vessel diameters average about 64 cm, with an average height of 16 cm. Matt impression is also observed, which is possibly an indicator of the vessel being molded in a woven mat until dried, to support the thick, heavy walls. Surface treatments during this phase include plain, polishing, burnishing, and smoothing. In addition to pans, oval -shaped vessels and spouts are also attributes associated with this phase (Moore 2002:10). The use of calcareous sand temperin g continued to dominate as before, with some mixing with volcanic sand, and in a few instances, quartz inclusions. The quartz probably originated in Saipan (Dickinson et al. 2001, Moore 2002:10), which would indicate movement of raw materials or completed vessels between islands. The interior regions of Guam were settled at this point in time, although coastal settlement remained the norm (Hunter -Anderson and Moore 1994). The Royal Palm (Dilli et al. 1993) and Naton Beach sites (Hunter -Anderson et al. 1998) are two examples that have acceptable radiocarbon dates for the Late Unai Phase. Beginning with this phase, there is a decline in the presence of flat bottomed pans. Walls typically are thinner, bases are rounder, and rim s slightly incurvate. The rims are unthickened or slightly thickened. Multiple tempers are used during this period, including calcareous sand, mixed sands, and volcanic sands. Surface treatments are either plain or polished/burnished. According to Moore (2002), a typical vessel during this period was 15 cm in height and 32 cm in diameter, had a 9 mm unthickened rim, 11 mm thick walls, volcanic sand tempering, and was finished with a plain smooth exterior. The pottery transitions observed during this period were gradual and did not occur at every site concurrently (Moore 2002:11). Site types during this phase include rockshelters and openair sites along rivers and the coast. Sites with light deposits are also found in inland areas (Carucci 1993; Fulmer et al . 1999).
67 The Latte Period is defined mainly by the appearance of latte sets, or large pillars with capstones made of limestone, coral, or basalt placed on top. They are believed to represent foundations for large wooden structures (Carson 2012a:1). The latte sets are arranged in two rows, 3 to 4 m apart and consisted of from 6 to 12 pillars with caps (Bodner 1997:90). Pottery is characterized by Marianas Plain ware, a thick walled ware with no slip and little decoration. Surfaces of vessels were roughened or combed ( Figure 2 13 A and Figure 2 13 B ), trailed ( Figure 213C) and wiped ( Figure 2 13D ) (Moore 2002:71). Settlements expanded inland during the Lat te Period to more marginal environments, such as the interior uplands with no direct access to marine resources. At European contact, settlements consisted of groups of latte houses nucleated into villages (Tomonari Tuggle et al. 2005). Latte structures were used for habitation, storage, or for rituals. Archaeological evidence points to a stratified society with chiefs able to organize the labor needed to construct monumental architecture. It has been hypothesized that latte were symbols of corporate landholdings (Graves 1986; Hunter -Anderson 1989). Subsistence consisted mainly of fishing and farming. Crops included taro, rice, breadfruit, yams, bananas, and sugarcane. Site types dating to this period include latte structures, artifact scatters, isolated mortars, quarries, and caves/rockshelters (Tomonari Tuggle et al. 2005). Previous Archaeology As is the case for most archaeological sites in the Marianas, the first work at the Bapot 1 site was conducted by Hans Hornbostel in the 1920s (Thompson 1932). The first extensive surface survey was performed by Alexander Spoehr (1957:31), who recorded three latte set clusters with associated artifacts scatters. Spoehr also noted recent disturbances to the ground surface of the Bapot 1 site, consisting of World War I I defensive trenches excavated by the Japanese military. Following Spoehrâ€™s survey, numerous excavation projects were conducted
68 that highlighted the siteâ€™s potential for understanding early settlement (Carson 2008; Marck 1978; Ward and Craib 1985). In 2005, Dr. Mike Carson (Carson and Welch 2005) conducted testing at the Bapot 1 site to collect information for assessing the siteâ€™s eligibility for listing on the National Register of Historic Places. That work consisted of the excavation of two 1 x2 m test u nits (TU) and sectioned rich cultural deposits that extended to 2.2 m below the ground surface. An Anadara cf. antiquata shell was submitted for radiocarbon analysis and produced a date of 3560 3270 cal a BP. That date and late prehistoric period surface f eatures indicate that the Bapot 1 site was occupied, or used, for approximately 2,500 years. This encompasses the entire prehistoric period of the Marianas. At the Ritidian Grotto site, various episodes of archaeological research have been conducted since the 1920s, also beginning with the work of Hans Hornbostel (Thompson 1932). Numerous latte sets were identified (Reed 1952; Reinman 1977) along with extensive surface artifact scatters (Kurashina 1990), Spanish period resources (Jalandoni 2011), and Pre latte Period deposits dating to 3500 cal a BP (Carson 2010, 2012). More recent excavations at the Ritidian Grotto site were conducted by Dr. John Peterson from the Micronesian Area Research Center (MARC), University of Guam. During these excavations, materia ls were recovered that dated from approximately 3000 to 500 cal a BP, encompassing portions of both the Pre Latte and Latte Period. Historical Background European contact in the Marianas Islands occurred in 1521, with the arrival of Magellanâ€™s expedition. This marked the beginning of a succession of distinctive historic periods that differed on each island in the archipelago. The historic period is commonly divided into the Spanish Colonial, First U.S. American (on Guam), the German Colonial (in the CNMI), the Japanese
69 Colonial (in the CNMI), The Japanese World War II (on Guam), the U.S. American World War II, and Second U.S. American periods. Each of these periods marks significant changes in the everyday lives of the Chamorro peoples. The Spanish Colonial Period (AD 1521â€“ 1898) After the arrival of Magellan in 1521 and the establishment of the Manila Galleon trade in 1565, contact between the Spanish and Chamorro was sporadic and usually consisted of yearly provisioning stopovers for Manila bound Spanish ships from Mexico (Russell 1998). A permanent Spanish presence in the Marianas was not established until 1668 with the arrival of the first officially sponsored mission. This heralded the complete subjugation of the Chamorro population, a violent process tha t lasted 30 years (Spoehr 1954:9). Spanish missionary descriptions from the seventeenth century characterize Guam as a thriving island, with some 180 villages, the largest consisting of 150 residences (Lvesque 1995). It was estimated that the population a t that time was 20,000 (Tomonari Tuggle et al. 2005). Spanish control and epidemics associated with foreign contact dramatically changed the character of the islands and by the end of the seventeenth century the Chamorro population was reduced to approxima tely 1,600 survivors (Russell and Flemming 2005, in Tomonari Tuggle et al. 2005). Many Chamorros fled coastal settlements as Spanish forces burned villages. The Spanish Reduccin during this period consolidated the Chamorro population of the Marianas into seven mission stations. Six of these were located on Guam and included Pago, Merizo, Hagta, Agat, Umatac, and Inarajan and one was located on the island of Rota (Rogers 1995). Foreign contact also impacted the landscape thorough the introduction of new s pecies such as cattle, horses, deer, pigs, goats, and carabao. New materials were incorporated into Chamorro sites during the Spanish Period, including the Mexican metate and mano, porcelain and stoneware dishes and jars, and metal
70 tools such as hoes. Also, metal nails that were shaped into fishhooks have been recovered (Moore 2007). Spanish sites exhibit westernstyle structures and artifacts, mostly in the urban areas such as Agaa and Umatac. On the Orote Peninsula, a new village replaced the Chamorro vi llage of Sumay, and three Spanish forts were constructed in the Apra Harbor area. Fort Santiago was constructed in 1721 on Orote Point to protect the entrance to Apra Harbor. Then, in 1737, Fort San Luis was built on San Luis Point. In 1801 Fort Santa Cruz was built on a shoal within the Inner Harbor (Carucci 1993:70). The First U.S. American Period in Guam (AD 1898 â€“1941) In 1898, the U.S. took control of Guam during the Spanish-American War, although U.S. rule was not firmly established until the turn of t he century. During this period the U.S. occupied Guam as a base for the Pacific cable and, along with Samoa, Luzon, and Hawai â€˜i, as a coaling station for the navy to guard trade interests with China and guard against future war in the Pacific (Gaily 1988:1 213). Urban and military activity was mainly concentrated in the villages of Sumay, Agaa, and Piti. Much of the development and prosperity of this period was centered on government facilities that served the Piti Navy Yard and the Marine Barracks at Suma y. These included schools, a hospital, courts, police, and augmentation of the water supply system (Lodge 1954:67). A seaplane ramp was built in 1921 at the U.S. Marine Corps air station. This was in operation for approximately ten years, until 1935, when Pan American World Airways used the facilities for its clippers to land in Apra Harbor on the San Francisco Manila-Hong Kong route (Fulmer et al. 1999). This period marks a significant ch ange in the history of the Marianas, as the islands of the archipelago were split up between the U.S. and Germany. Also, a fundamental shift in the CNMI occurred, from Spanish control that focused primarily on missionary work, to Germanyâ€™s focus
71 on economi c development (Hoffman 1950:2). Immediately following the U.S. victory over Spain during the Spanish-American War and the signing of the Paris Peace Treaty on 10 December 1898 (von Bennigsen 2003), Guam was taken over by the U.S. (Gailey 1988). Without the Philippines, Spain had little interest in maintaining a colonial presence in the region. In a move to expand their colonial interests and extend their military capabilities in the western Pacific, Germany bought Spanish possessions in Micronesia, including the Caroline Islands (consisting of Yap, Kosrae, Pohnpei, and Chuuk), Palau, and the Marianas (excluding Guam). The German Emperor formally declared a protectorate over the Northern Marianas on 18 July 1899 (von Bennigsen 2003). During the German Period, the population of Saipan began to increase, a portion of which included Chamorro families returning from Guam. In 1907, after several years of severe storms, 400 Carolinians were evacuated from the Mortlock Islands south of Chuuk and from Satawal and Loss ap in the Carolines and were given homesteads on Saipan (Craddock 1982:56). Saipan was the administrative center of the Northern Mariana Islands under German rule. Tanapang Harbor consequently became the most important port for the territory. German, Japa nese, British, and U.S. ships anchored in the harbor to trade food, tobacco, textiles, iron, wood, and alcohol (for nonnative consumption) in exchange for copra, coffee, lemons, oranges, dried trepang, tortoiseshell, shark fins, and shells (Farrell 1991:272). As a result of this increase in trade, infrastructural improvements were made to the harbor and major roads. Although the German Colonial period was short, some policies enacted during that time left positive and lasting legacies, including an emphasi s on Chamorro customs and keeping Chamorro as the primary language taught in a compulsory public education system. Unoccupied lands were also tightly controlled and foreigners were prohibited from acquiring land, which prevented the establishment of planta tion monopolies (Spennemann 2007:327). These factors, in part, account
72 for some of the current differences in Chamorro cultural practices and language use between Guam and the CNMI. In September 1914, with the beginning of World War I, the South Seas Squadron of the Japanese Navy steamed out of Yokosuka, Japan, bound for the German Colonies in Micronesia, presumably to pursue the German East Asiatic Squadron. On 14 October 1914, the Japanese battleship Katori dropped anchor off Garapan Town on Saipan, effectively seizing control of all of Germanyâ€™s colonies in Micronesia. Although this alarmed Britain and America, who viewed Japanâ€™s occupation of Germanyâ€™s Micronesian colonies as flagrant expansionism (P eattie 1988), control of the Northern Mariana Islands was transferred to Japan under the military administration of the Provisional Japanese South Seas Defense Force (Rinji Nanyo Gunto th Seas Development Company) began large scale sugar production on Saipan. Soon after, improvements were made to Tanapag Harbor, including construction of two 1,500ft long concrete jetties, dredging of the channel and harbor to support 28ft draft ships, and alteration of the harbor to allow access by Japanese amphibious aircraft. By 1936, it was possible for passengers to disembark directly onto a government pier (Farrell 1991:321, Russell 1984:61). During this period, large sugarcane plantations were dev eloped and the land was divided into rectangular plots, covering more than 70 percent of the islands at their peak. Roads followed the regular arrangements of the fields, and a narrow -gauge railroad that connected fields, processing plants and warehouses w as also built. At the beginning of World War II, Saipan served primarily as a rear area supply base. As the war intensified, however, the island became integral to the defense of the Japanese Empire.
73 Much effort was expended to reinforce the military pres ence, which involved construction of new fortifications, upgrading airfields, and increasing the number of troops (Hoffman 1950). The Japanese World War II Period (AD 1941â€“1944) On 8 December 1941, the Japanese launched an attack on Guam consisting of the Japanese Army South Seas Detachment assembled in the Bonin Islands and the 5th Defense Force based in Saipan. The fighting consisted of two days of air strikes followed by the landing of ground forces at Tumon Beach, Agaa Bay, and Merizo. Because U.S.held Guam was unfortified, the island quickly fell to the Japanese forces. Follow ing the invasion of Guam, the Japanese immediately set up an economic system to provide crops for the local and Japanese populace on Guam and for export. Heavy handed policies enacted by the Japanese, including the confiscation of land and reducing local f armers to slaves for work in rice fields, quickly alienated the local population (Gailey 1988:34). Many Chamorros fled the urban areas to escape the new Japanese regime, which also included a forced language change in schools and drafting residents to buil d defensive fortifications. U.S. American World War II Period (AD 1945â€“1948) U.S. air strikes on Saipan began on 11 June 1944, followed by an amphibious assault on 15 June at beaches that included Tanapag, Chalan Kanoa, and Garapan. The 2nd Marines destro yed Garapan on 3 July and moved northward. On 7 July, a massive counterattack against American forces occurred at Tanapag that resulted in 461 American dead and up to 4,000 Japanese dead after only one day of fighting (Adams et al. 1996:8). The battle for Saipan ended officially on 9 July 1944, although scattered resistance continued for some time (Farrell 1991:352, 367â€“369, Russell 1984:86â€“97). On 8 July 1944, U.S. forces attacked Guam. This began with 13 days of continual bombardment of coastal defense gu ns, anti aircraft weapons, airfields, portable defensive
74 features, high ground, and various other forms of military infrastructure on Guam. The U.S. assault came in the form of air strikes and ground forces landing at the villages of Asan and Agat on the s outhwestern portion of the island. The battle of Orote Peninsula consisted of four days of intense fighting and the Sumay Village area was one of the last areas to be cleared of Japanese defenders during this battle. This is because the cliffs west of the village had been heavily fortified to defend Apra Harbor and the town itself was extensively mined by the Japanese, making the streets impassable to tanks. Once the island was secure, the U.S. began preparations for developing Guam into a staging area for an attack on Japan, and Apra Harbor became a major naval base. The Orote Peninsula was transformed into a large port, replete with airstrips and facilities for ship repair, housing, and storage (Lodge 1954). The Second U.S. American Period (AD 1948 â€“Present ) In 1948 the U.S. Navy transferred control of Guam to the U.S. Department of the Interior. The Organic Act of 1950 provided locals with U.S. citizenship and some degree of self rule. In 1970 the people of Guam elected their first governor and today Guam i s a territory of the U.S. and retains a substantial military presence. This period is also characterized by the highest level of commercial development for both the private sector and U.S. military enterprises. Ethnohistory of Marine Resource Exploitation Marine resources have been an essential part of the everyday lives of peoples of the western Pacific. These include the inhabitants of Southeast Asia, Near and remote Oceania, and the Chamorro. Because of this, the following sections of this chapter discus s ethnohistoric information regarding marine resource exploitation within this broad region and information specific to the Mariana Islands. The historical and ethnographic literature pertaining to fishing and shellfish gathering in Micronesia is relatively restricted. However, this information provides an important contextual
75 foundation from which appropriate analogies can be formulated for the interpretation of analyses presented in this dissertation. Because specific information regarding traditional marine resource exploitation in the Marianas is limited, information from the Pacific Basin that includes Southeast Asia ( Figure 1 1) was also considered. The following discussion focuses upon the general aspects of marine resource exploitation most pertinent to the archaeological data recovered from both the Bapot 1 and Ritidian Grotto sites. This includes traditional modes of resource management, descriptions of traditional techniques and gear, and gender or age specific divisions in exploiting marine resources. First, general information regarding marine resources in the broader context of Oceania and Southeast Asia is discussed, followed by a narrative about the limited historical and ethnographic information related to the Marina Islands. Southeast Asia and the Pacific The modern states of Southeast Asia, includi ng southern China, Vietnam, Cambodia, Laos, and Thailand contain some of the richest aquatic habitats in the world. These include broad continental shelves, river deltas, mangrovelined coasts, and inland lakes. Along with intensive use of the rich aquatic environments, other key Neolithic developments, including rice cultivation, factored into the movement of peoples into the islands of Southeast Asia and remote Oceania (Higham and Thosarat 1994). The fishing techniques and technologies developed by Proto-Austronesians began around 6,000 years ago in the relatively calm waters of this region that contains numerous islands and extensive continental shorelines. Ultimately, this led to the development of cultures highly adapted to the rich maritime environment s of Southeast Asia with technologies that eventually enabled settlement of remote Oceania (Kirch and Dye 1979:5354). In Southeast Asia, sea nomads such as the Sama Bajau, a traditional, boat dwelling seafaring nomadic people, still live scattered through out a vast maritime area of 3.25 million km2 from the Philippine Islands of Palawan, Samar, Mindanao, and the Sulu Archipelago to the
76 Indonesian Islands of Borneo, Sulawesi, and other small dispersed islands of Eastern Indonesia ( Figu re 1 1) (Sather 1997:2). The Sama Bajau language is a sub -group of Austronesian in the Malayo -Polynesian family. Currently, the SamaBajau people engage in a wide range of fishing and mari ne resource exploitation activities and collect numerous species from inshore and pelagic habitats. They also conduct long range nomadic fishing expeditions, build a variety of craft for fishing and transport, and use multiple types of gear, from simple traps, hook and line, and spears, to long line gear and hookah dive equipment (Stacey 2007:38). As skilled maritime specialists, sailors, messengers, explorers, and traders, the SamaBajau played important roles in the development of historical maritime stat es in Southeast Asia during the sixteenth and seventeenth centuries (Stacey 2007). Evidence of complex marine resource exploitation patterns from Ancestral Polynesian archaeological sites includes various forms of fishing tackle, including fishhooks, trol ling lures, octopus lures, and net weights. Also, faunal materials from these sites are often dominated by fish bone, derived primarily from fish species found on inshore reefs, with the rare appearance of pelagic species. Bones from turtle are also common , along with shark and marine invertebrate shell. The suite of species found within archaeological contexts indicates that Ancestral Polynesians fully exploited their marine environments, especially littoral and inshore areas, and sometimes ventured into t he open ocean to catch pelagic species (Kirch and Green 2001:131 135). The act of fishing on Pacific Islands has also been imbued with complex meaning, and taboos and beliefs have been attached to all aspects of this important subsistence activity. In old Hawai â€˜i, oneâ€™s social status could be elevated by landing large and dangerous species like the swordfish (Xiphias gladius ). Superstitions regarding this species indicated that whoever caught this fish became a conqueror and was given strength. The swordfis h was reserved for chiefs and
77 was laid upon alters with a human sacrifice, as was done with tiger sharks (Titcomb 1972:68). Taboos and legends also surrounded the often dangerous activity of gathering opihi (Cellana sandwicensis ), which is a limpet usuall y found along wave pounded cliffs and shorelines. This species of limpet was reportedly the most common species of shellfish consumed in historic times and is still highly prized in Hawai â€˜i (Titcomb 1978:343). Although archaeological evidence regarding th e exploitation of marine resources is relatively rich in terms of the number and variety of species collected, it generally lacks direct evidence attesting to the variety of materials and techniques used catch and harvest marine resources. Kirch and Green (2001:132) pointed out that most of the materials used to make fishing gear consisted of perishable materials. Furthermore, the impacts of colonialism have blurred direct connections between the complexity of marine resource procurement practices observed historically and archaeological evidence. Unfortunately, ethnographic work on the subject has largely focused upon the material aspects of fishing rather than the relationships between technology, behavior, and the environment. To remedy this, Kirch and Dy e (1979:61) conducted ethnoarchaeological research on Niuatoputapu Island in Tonga ( Figure 1) and recorded 37 named traditional fishing methods in current use. These included methods that used nets, spears, traps, fishing line, diving, and poisoning. They also documented genderspecific activities in which women regularly participate or conduct s eparately from their male counterparts. Importantly, they also observed that these gender specific activities are viewed differently. Diving, for example, was practiced only by young men. Particularly accomplished divers and their exploits were often rever ed in conversation. In contrast, groping for fish and gathering shellfish on the reef at low tide was largely the domain of women and children, and was looked down upon by men. The fact that shellfish gathering is practiced largely by women
78 and children ha s also been observed in Hawaiâ€˜i (Titcomb 1978), in Australia (Meehan 1982), Papua New Guinea (Swadling and Chowning 1981), and Fiji (Sahlins 1962). In the Marshall Islands, women did not participate in fishing expeditions. They were viewed as bad omens or unclean beings subject to strict taboos during menâ€™s fishing activities. Women, however, contribute to the diet by collecting shellfish, certain reef fish, and octopus (Perosian Husa 2004:67). Gleaning, or nubba, amongst the Bajau in Indonesia, is also con ducted by women and children to satisfy domestic needs. This activity is conducted along beaches, shallow waters, sand flats, and on fringing reefs during daylight hours during which trepan, or sea cucumber (Holothuroidea), sea urchins, seaweeds, shellfish, crustaceans, hard corals, and sponges are collected (Stacey 2007:38). This was explained further by Chapman (1987:270271), who pointed out that throughout Oceania, reef gleaning is carried out by women, children, and sometimes men, but women contribute most of the protein during this venture. Chapman (1987) also noted, however, that this distinction is a general one and that women on some Pacific islands were skilled at fishing on and beyond the reef. In the instances where they were observed using spear s and diving for fish, their skills compared favorably with men. Chapman (1987) further notes that the main differences between the sexes are that women are more restricted, mainly to reef flats and lagoons, and rarely venture into deep waters, the tackle employed by women is simpler, womenâ€™s fishing is more secular than menâ€™s, and involved less ritual while menâ€™s fishing activities are surrounded by taboos and complex beliefs (Chapman 1987:270 271). Bird and Bliege Bird (1997:4950) observed among modern M eriam Islanders in the Torres Strait between Australia and Papua New Guinea that whereas men engage in intertidal gathering with children and women, womenâ€™s efforts are much more directed and accounted for significantly higher yields of edible flesh. Also observed among the Meriam Islanders were distinctive
79 differences between juvenile and adult shellfish collecting. Children on the Meriam Islands were extremely active in collecting on the reef flat, they collected independently from adults, and children learned about collecting on the reef from other children (Bird and Bliege Bird 2000:463). Aside from traditional techniques, technologies, and beliefs related to fishing in the Pacific, traditional taxonomies, knowledge of the behavior and reproductive cycl es, and the management and conservation of marine resources was highly developed in Pacific island societies. Native taxonomies were noted by early European explorers in the Pacific. During Cookâ€™s voyage it was recorded that Tahitians knew of 150 distinct species of fish, with approximately 48 described as edible (Dâ€™Arcy 2006:34). Palauan fishermen could name up to 300 different species of fish and readily differentiate between species with subtle physical differences using a naming system strikingly similar to that used by biologists (Johannes 1981:10). Hawaiians had also developed an extensive system of nomenclature based upon keen observation. The Hawaiian system relied upon more than just gross anatomical and morphological characteristics and included habitat preference, economic uses, behavioral traits, color, and lifestyle changes (Titcomb 1978:327). Because of this, the traditional systems were often more detailed than the binomial Linnaean classification system for some species (Johannes 1981:10). A t raditional ethic of conservation has also been recorded throughout the Pacific. Hawaiians constantly thought of the conservation and supply of resources, both terrestrial and aquatic. Fishing grounds were never depleted and when new rich fishing grounds we re discovered, the fish were often fed so that they would remain in that spot and increase. These grounds were then exploited carefully (Titcomb 1972:12). Current island societies throughout the
80 Pacific employ a system of reef and lagoon tenure, in which h arvest rights in certain areas are controlled by social groups like families or clans. These are also controlled by a chief, acting on behalf of the social group (Berkes 2008:87). Each group regulates their area and imposes a wide range of restrictions and regulations to maintain high sustainable yields. In other words, it is in the best interest of a given group to limit harvests in the area they control as the benefits of high sustainable yields accrue directly back to them (Johannes 1978:267). Johannes (1978:356) further argued that recent, wide spread degradation of marine resources throughout the Pacific resulted from the breakdown of traditional conservation practices. The specific, interrelated causes of this are introduced money economies, undermined traditional authority, and implementation of colonial laws. The Marianas Embedded in the chronicles of violent events surrounding Magellanâ€™s brief and ill fated stay in the Marianas are the first descriptions of fishing. This is seen in the accounts of A ntonio Lambardo, or Pigafetta, who wrote the most substantial and factual description of this first instance of European contact with the Chamorro people (Barratt 2003:5). Unfortunately, like many of the descriptions of Chamorro fishing recorded by subsequent early visitors, Pigafettaâ€™s is brief and only describes the catching of flying fish (Exocoetus spp.) using fishhooks made of fishbone. The account of Juan Pobre de Zamora, a Franciscan Friar who stayed on Rota for seven months in 1602, further describe s fishing for flying fish and other pelagic species, such as mahi mahi ( Coryphaena equiselis ), marlin (Istiophoridae), and sharks (Elasmobranchii) (Driver 1993:15). Whereas this is the most detailed account of fishing from the early Spanish period in the M arianas, inshore fishing strategies are not mentioned (Amesbury et al. 1986:5 7). These very general descriptions persisted, such as Friar Peter Cooman's passage written in the late 1600s regarding night fishing from canoes, shoreline hook and line fishing performed by women,
81 and superstitions related to fishing (Coomans 2000:14). Perhaps one of the earliest, most detailed descriptions of fishing and shellfish gathering is by Freycinet, who visited the Marianas in 1819. Freycinet describes the technologies used at the time, including hooks, spears, the acho (baiting device, used to catch atchuman, or Decapterus spp.), nets of varying type, and fish pools and enclosures built from mud barriers and reed dikes, analogous to the stone enclosures built during anc ient times. He also describes the use of some of these technologies in the capture of various species of fish, including magnahak , or manahoc (rabbitfish, or Percoidei), laguas (parrotfish, or Scaridae), gaaga (flying fish), anaho (mahi mahi), and eels (An guilidae), and provides short descriptions of fishing for crabs and turtles and the collection of shellfish by women along the coastal rocks and sands (Freycinet 2003). Spanish accounts following Freycinetâ€™s descriptions continue to be brief and undetailed, usually describing the abundance and flavor of fish and emphasizing the exclusive use of inshore habitats by the Chamorros (Sanz 1991:7). Some accounts mention the taking of other species of fish, such as atulay (horse mackerel) and tiao (goatfish), and provide more detail on the use of the acho, a practice that has been passed down from ancient times (Ibez y Garca 1992; Olive y Garca 2006). Others, like Marche (1982) provide biased descriptions of Chamorro fishing. During his 1887 visit to the Marian as, Marche (1982:8) reported that because of their â€œlazy lifestyle,â€ little is taken from the sea and fishing is only done within the reef. In 1899, after Germany had purchased the Northern Marianas from the waning Spanish Empire, George Fritz (2001) assum ed the post of District Administrator and took a keen interest in Chamorro language and culture. In his writings on Chamorro marine resource exploitation he elaborated in greater detail on the traditional technologies and techniques still in use during the late nineteenth and early twentieth centuries. He provides a fairly comprehensive description of
82 targeted species, including clams and snails collected by women and children on the reef at low tide. He also describes two preserved types of ancient fishing that were still used during his tenure in the Marianas: 1) atcho poco, a chumming device used for catching atchuman (Decapterus spp.); and 2) use of a lure fish. This latter fishing involves tying a thread to the pierced jaw of a parrotfish. The tethered fish was allowed to swim over the reef and when the fish was pulled back in with the thread, other fish followed and were then taken with a net. This tethered fish was then tied to a stone or confined in a trap for future use (Fritz 2001:71). Between 1900 and 1940, under the administration of the U.S. Navy, political and economic conditions on Guam improved. Basic civil liberties, however, did not improve during this period, and until 1941 Chamorros still relied on a subsistence economy (Rodgers 1995). Ever yday life for Chamorros in the Northern Marianas, following Spanish rule, first under German governance between 1899 and 1914 and then under the Japanese Mandate from 1914 until 1944, also witnessed large economic, political, and social changes. However, t hroughout both colonial periods, Chamorro society maintained a core body of distinct customs centered on the family, church, and farm and the people were able to maintain their language along with many other customs (Spoehr 1954). This appears to include s ome traditional methods of marine resource exploitation. Since World War II, marine resource exploitation has shifted from traditional subsistence to more modern modes of subsistence, commercial, and recreational forms of fishing. This included changes in techniques and materials. Gear composed of natural materials was replaced by modern manufactured materials, including monofilament lines and nets, metal hooks, poles, reels, spear guns, and SCUBA gear, and those changes led to large declines in all inshore fisheries (Hensley and Sherwood 1993:130).
83 Micronesians, like all Pacific Island societies, rely heavily upon the sea as a source of food and raw materials. This was likely more so for pre Contact inhabitants of the Mariana Islands than it is for it s current inhabitants. In the different coastal environments (i.e. lagoons, mangroves, flats, and reefs) adjacent to both the Bapot 1 site on Saipan and the Ritidian site on Guam one can observe individuals engaged in different marine resource procurement activities and one can easily find many species of fish and shellfish that have been collected at the same locations for thousands of years. Unlike today, however, marine resources, as seen through the archaeological evidence, appear to have been much rich er on the reefs adjacent to the Bapot 1 and Ritidian sites. This is because those areas were likely subjected to lower levels of exploitation (Kirch and Dye 1979) and possibly strict management policies (Johannes 1978:351). It appears that some aspects of Traditional Chamorro culture survived approximately 500 years of colonial rule. Some surviving aspects of traditional culture appear to be related to fishing and shellfish gathering practices that have been handed down from prehistoric ancestors and were l ikely practiced until relatively recently, in the mid twentieth century. One of these included gender divisions related to the collection of shellfish. The prevalence of this relationship between women and shellfish gathering in island cultures throughout Oceania (Chapman 1987) attests to the antiquity of this practice.
84 Figure 2 1. Map of the Mariana Islands. Map inset shows the location of the Marianas relative to east Asia, southeast Asia, and New Guinea.
85 Figure 2 2. Map of Saipan Island showing the location of Lau Lau Bay and the Unai Bapot archaeological site.
86 Figure 2 3. Photograph of latte elements located on the surface of Bapot 1 approximately 15 m south of Block A, facing west (photograph taken in 2006 by the author). Figure 2 4. Photograph of toppled latte elements located in the eastern portion of the B apot 1 site, facing east (photograph taken in 2006 by the author).
87 Figure 2 5. Map of the northern tip of Guam showing the location of the Ritidian Grotto site. Map inset shows the Island of the Guam with the Pago -Adelup Fault Line.
88 Figure 2 6. Excavations of the Ritidian Grotto rockshelter looking north out from cave. Photograph courtesy of Dr. Mike Carson and Dr. Hiro Kurashina.
89 Figure 2 7. Entrance to cave jus t south of the Ritidian Grotto rockshelter.
90 Figure 2 8. View of the limestone promontory above the Ritidian Grotto site, facing northeast (photograph taken in 2011 by author).
91 Figure 2 9. Chart showing all of the historic periods and phases for the Mariana Archipelago.
92 Figure 2 10. Decorated pottery (From Butler 1994: Fig. 9 and Fig. 11). A) Achugao Incised sherds . B) San Roque Incised sherd. Figure 2 11. Pottery surface treatments. A ) Limefilled and impressed Ipao Stamped sherd (From Ray 1996:86, Fig. 45). B ) Limefilled and impressed Ipao Stamped sherd (From Spoehr 1957:121, Fig. 56). C) Lime filled and impressed Ipao Stamped sherd (From Spoehr 1957:121, Fig. 56). D) Impressed Tarague Striated sherd (From Ray et al. 1996:85, Fig. 25). E) Striated Tarague Striated sherd (From Moore 2002:71, Fig. 3). F) Sherd with fingernail impressions (From Thompson 1979:85, Fig. 14). Figure 2 12. Examples of rim impressions (Solheim 2003:8, Fig. 1 5). A B A B C D E F
93 Figure 2 13. Illustrations of La tte Period surface treatments (From Moore 2002:71, Fig. 3). A) Lightly combed surface. B) B oldly combed surfa ce. C) T railed surface. D ) W iped surface . A B C D
94 CHAPTER 3 ARCHAEOLOGICAL TESTING AND LABORATORY ANALYSIS Introduction Styles define the appearance of various text elements in your document, such as headings, captions, and body text. When you apply a style to a paragraph or word, you can apply a whole group of character or paragraph formats or both in one simple operation. When you want to change the formatting of all the text of a particular element at once, you just change the style that's applied to that element. Styles make formatting your document easier. Additionally, they serve as building blocks for outlines and tables of contents. The followi ng chapter describes the archaeological methods employed during fieldwork and in the laboratory. Results of excavations include descriptions of sectioned stratigraphy and summaries of materials recovered during excavations. The results of laboratory analys es for radiocarbondated materials, artifacts, faunal remains, and human remains are also presented by site under separate headings (Bapot 1 and Ritidian Grotto). Methods Methods employed for the collection of field and laboratory data from Bapot 1 and the Ritidian Grotto sites were standardized. This involved use of the same excavation methods and use of the same tools and procedures in the laboratory. This facilitated comparisons between data collected from Bapot 1 and the Ritidian Grotto sites. Excavation Fieldwork for the present project was conducted in conjunction with the research of Dr. Geoffrey Clark from the Archaeology and Natural History Museum, Australian National University (ANU). The primary goal of excavations at the Bapot 1 site for this res earch was to
95 obtain materials for radiocarbon analysis and determine the age of the oldest cultural deposits (Clark et al. 2010:30). Work for the present project conducted at Ritidian was also collaborative and fieldwork coincided with research conducted by Dr. Mike Carson from the Micronesian Area Research Center (MARC) at the University of Guam. Dr. Carsonâ€™s work was conducted at Ritidian between 2006 and 2011 and focused primarily on investigating the earliest parts of the archaeological sequence. Mater ials from both the Bapot 1 and Ritidian Grotto sites are considered for the current study because their preContact occupational periods overlap and each offers data regarding local pre Contact environments, depositional processes, prehistoric resource use, and settlement patterns. Ancillary to these site characteristics, this project incorporated archaeological excavation techniques aimed at recovering appropriate samples for zooarchaeological and fisheries sciencebased analyses. To better incorporate me thods from fisheries science, excavation methods were designed to recover samples of shell and bone from archaeological contexts that were large enough to be statistically analyzed and compared. Basic archaeological and zooarchaeological methods for collecting and analyzing faunal materials were employed, including careful considerations of site context and recovery techniques. All individual test units measured 1 x 1 m and were excavated by 10 cm levels following natural stratigraphy. Excavated soils from each 10cm level were sieved through 1/8 inch and 1/16 inch -square mesh screen, and a subsample of bulk excavated soil from each level was also collected for sorting in the laboratory. This ensured that small bones, shell fragments, and other forms of cul tural and environmental data were recovered.
96 Laboratory Laboratory analyses for the present project focused upon the archaeological residues of marine resources. Artifacts and faunal materials were sorted during excavation according to type. Along with bulk sediment samples, the materials sorted according to type were further sorted and assigned to the lowest taxonomic order possible, following Reitz and Wing (2008:164167). Because extremely large amounts of shell were recovered during excavations of Block A at Bapot 1, a subsample of invertebrates was selected for analysis. This included all invertebrate remains from TU 1, TU 4, and TU 6. To aid identification of archaeological bone and shell fragments, live fish and shellfish were collected during numerous fishing trips on Guam, and bought at local fish markets, following the approach of Rietz and Wing (2008:146152). Species identification and collection of information regarding habitat preferences were facilitated with appropriate keys from numerous references (Abbott and Dance 2000; Allen 2000; Allen et al. 2003; Carpenter and Niem 1998; Cernohorsky 1972; Donaldson 1996, 2002; Gosliner et al.1996; Kay 1979; Lamprell and Whitehead 1992; Myers and Donaldson 2003; Paulay 2003b; Randall 1985, 2001; Smith 2003). Basic analyses of the primary data were performed to define the nature of the human marine environment interaction and compare species frequencies through time. These analyses included comparisons of relative frequencies of taxa through time, using specimen counts, or the number of identified specimens (NISP) of bone and shell attributed to a taxon and through comparisons of species diversity within the marine resource assemblage through time. The number of identified specimens is related to the number of identifiable bone or shell elements of an animal, siteformation, sampling techniques, and laboratory methods and therefore, are subject to various biases (Reitz and Wing 2008:203). First, different invertebrate
97 and vertebrate species have highly diver gent types and numbers of hard tissues that preserve differently in archaeological deposits. In the case of vertebrates, one animal can consist of numerous hard structures, and NISP cannot determine if individual bone fragments are independent from one another (Grayson 1973:432). Second, NISP can be influenced by cultural practices (e.g. transportation, butchering, and cooking) that can destroy and disperse bone and shell. Third, NISP can be heavily biased by recovery and laboratory methods. For instance, n ot employing fine screens during excavations will not ensure that remains from small taxa are recovered. Also, restricting identifications of elements to a limited priority list can skew NISP (Reitz and Wing 2008:203â€“204). Despite these biases, NISP was us ed for the present study because of factors related to the nature of the faunal assemblages from Bapot 1 and the Ritidian sites that inhibited calculating the minimum number of individuals (MNI) for each species. MNI is also commonly used to estimate the r elative frequencies of taxa and is defined as the minimum number of elements of an animal needed to account for all of the parts of that species within a site (Shotwell 1955:272). The most limiting factors preventing calculations of MIN were the extremely fragmented nature of bones and shells from some taxa. Also, elements of some taxa, especially fish, were nondescript and small and simply could not be identified beyond class. Therefore, NISP in this case, is an estimate of the number of bones and shells f rom certain taxa that have survived and were recovered from the faunal assemblages of the Bapot 1 and Ritidian sites (Ringrose 1993:1932). Because this involved the use of nominal scale data (i.e., the distribution of observations, or NISP, among categorie s, or species), the ShannonWiener diversity equation was employed (Zar 1999:40 41). This equation is commonly used as an ecological metric and combines two quantifiable factors. These are the number of species within a habitat, or species richness, and the evenness (relative abundance of individuals in each
98 species), or dominance patterns of different species, within a habitat (Kr ncke and Reiss 2010:60; Labrune et al. 2006:35). In the present study, the Shannon Wiener Index was used to assess possible cha nges in habitat condition or shifts in human exploitation patterns through time. This involved calculating Shannon -Wiener index values for identified marine invertebrate assemblages in samples from different excavation levels, i.e. from different time peri ods. It was assumed that the remains from each provenience (excavation level or dated time period) were collected from the same habitats. These include the reefs, lagoons, and littoral environments associated with the study sites. Comparisons of these valu es through time were used to evaluate possible negative impacts caused by preContact marine resource exploitation. I assumed that a negative impact on a taxocene would be reflected by a decline in the value of the Shannon Wiener Index, which is expressed as: = ( ) log Here, is the information content of the sample, P is the relative abundance of the i th taxon within the sample, Log Pi is the natural logarithm of Pi , and s is the number of taxonomic categories. According to this measure of diversity, for two samples with the same number of taxa (richness), the sample with a more even distribution of individuals among taxa, i.e. greater equitability, will display higher diversity. Also, for two samples with similar equitability (distribution of relative abundances of individuals in the taxonomic categories), the sample with greater species richness will possess higher diversity (Reitz and Wing 2008:111 112). All radiocarbon dates were calibrated in OxCal v4.2.2. (Bronk Ramsey 2013) using the 09 marine calibration curve from Reimer et al. (2009). The â€œU Sequence Functionâ€ was also used in OxCal v4.2.2. to create a uniform deposition model from the calibrated dates (Bronk
99 Ramsey 2008; Bronk Ramsey and Lee 2013). From this model, age depth plots were created for each site. Results The chronological sequence used to analyze materials for this project relied upon a total of 29 radiocarbon dates. These included 20 dates from Bapot 1, Block A (Table 3 2) and nine dates from the Ritidian Grotto site ( Table 3 14). The Bapot 1 radiocarbon dates were determined at the Waikato Radiocarbon Dating Laboratory, New Zealand (Clark et al. 2010). The Ritidian Grotto samples were collected by Drs. John Peterson and Mike Carson from the MARC, during preliminary excavations of the site in 2007, and run at Beta Analytic Inc., Florida. The conventional radiocarbon ages from these analyses were calibrated using the INTCAL09 atmospheric curve data (Reimer et al. 2009) with OxCal v4.2.2 (Bronk Rams ey 2013). Excavations for the current project consisted of a 3 x3 m excavation block (Block A) located approximately 10 m north of the toppled latte elements on the 7 m contour of the sandy coastal plain ( Figu re 31). Block A was divided into nine 1x1 m test units (TU) ( Figure 3 2); each unit was excavated to approximately 2.5 m below the ground surface. A portion of TU 2 along the north wall of Block A was excavated to 3.2 m below the surface to test for the presence of deeply buried materials or deposits ( Figure 33). The excavation of TU 9 was stopped at 115 cm below datum (cmbd), where a human burial was discovered ( Figure 3 4 and Figure 35). Stratigraphy A total of seven stratigraphic layers (Layers I through VII) that contained rich concentrations of cultural material, and a single layer (Layer VIII) of naturally deposited beach sand, were encountered during the excavation of Block A. Detailed descriptions of the soils from each layer were recorded ( Table 3 1) and all walls of Block A were profiled with detailed scale
100 drawings and photographed ( Figure 33 Figure 36). Soil descriptions were derived from the authorâ€™s field notes, in combination with Clark et al. (2010:2526). Generally, Layers I through IIV represent a complex stratigraphic sequence with the presence of dense artifacts, faunal materials, various subsurface features, large limestone rocks, and irregular lower boundaries. Other than a natural disturbance consisting of a decomposed tree root on the west wall of Block A ( Figure 3 6), the primary cultural contexts of the siteâ€™s stratigraphy were intact. Layer I was divided into two sublayers (Ia and Ib). This was based upon differences in color, textu re, and the presence of historic materials in the upper portions of Layer I. The upper portion, Layer Ia was approximately 20 cm thick, consisted of dark silty sand with roots and rootlets, and appears to have been historically mixed, as it contained both traditional and World War II era artifacts. These included Latte Period potsherds, small amounts of bone and shell, historic glass, pieces of vulcanized rubber, and metal shrapnel. Layer Ib was comprised of dark loose sand. This sublayer ranged from 520 cm thick and had a very irregular and uneven lower boundary. Layer Ib did not contain any historical artifacts or modern materials, suggesting that it was relatively intact. Traditional artifacts included Latte Period pot sherds, shell, and bone. Some thin red pieces of pottery were also present and may represent older pottery that was mixed with late preContact materials. Layer II ranged from approximately 20 to 40 cm thick and had a relatively level, even lower boundary. Sediments from Layer II consisted of light brown winddeposited beach sand with thick walled, red slipped pottery, lithics, marine shell, and some bone. Fragments of Turbo spp. shell were particularly common and bone from Layer II included specimens of fish, reptiles, and a small mammal, possibly a rat or mouse.
101 Layer III was 2060 cm thick with an even lower boundary that was between 90 120 cmbd. A human burial was encountered in Layer III at 115 cmbd in TU 9 in the southeast corner of Block A ( Figur e 3 2 Figure 3 5). Layer III consisted of brown silty sand, portions of which were indurated. This layer contained an increasingly dense concentration of cultural materials that included thick walled red slip ped pot sherds that were especially common in TU 6, some thin red and black pot sherds, nonhuman bone, shell, basalt andesite flakes, a Tridacna sp. adze, shell beads, and pearl shell fishhook fragments. Layer IV ranged from approximately 100 to 150 cmbd and consisted of yellowish brown, loose silty sand. The lower boundary of Layer IV was irregular and uneven. There is a portion of Layer IV that extends from 140 to 190 cmbd into Layer V on the north wall profile of Block A ( Figure 3 3). This appears to have formed from a degraded tree root. Large amounts of thickwalled pot sherds were present in the upper portions of Layer IV. These also had heavy red slip and were derived from flat based trays or platters. Below 130 cmbd the density of thin redslipped Marianas Red Ware sherds increased. Various tool stamped sherds were also present. Stone adzes, lithic flakes, Conus sp. shell rings, and laterally ground Cypraea moneta shell beads were also present. Layer V was approximately 100 cm thick and ranged between 120 and 230 cmbd. The coarse yellowish brown sandy sediments of Layer V contained thin red slipped pre Latte period pottery from small globular carin ated jars with everted rims with in situ base sherds. Shell and stone artifacts including shell rings, beads, fishhook fragments, stone adze fragments, and a large sub lenticular volcanic adze were also present. Flecks, larger chunks, and in situ charcoal were also encountered in Layer V, suggesting that shallow fire pits or hearths were also present.
102 Layer VI was situated between 160 and 240 cmbd and consisted of brown silty calcareous sand. Some of the sandy sediments in layer VI were indurated. Stone art ifacts, including a cache of three adzes in TU 7, large quantities of thin red slipped Pre -Latte period pottery, shell beads and rings, shell fishhooks, nonhuman bone (from birds and fish), and marine shells dominated by Anadara cf. antiquata were present in Layer VI. Layer VII consisted of yellowish red indurated silty sand. This layer had a very irregular lower boundary and was situated between 200 and 240 cmbd. This layer was the base of the cultural deposit and contained similar cultural materials as L ayer VI. These included basalt, andesite, tuff, chert, and quartz/calcite lithic artifacts; a large amount of bird bone derived from the flightless rail ( Gallirallus cf. philippensis ); thin redslipped pottery, and shell ornaments. Stratigraphic differences from Layer VI included the incorporation of red silts and clay into the sandy beach sediments and presence of charcoal from anthropogenic burning in Layer IIV. Layer VIII consisted of coarse, very pale brown, indurated sand. This layer did not contain cu ltural materials and consisted of naturally deposited beach sands. The north half of TU 2, along the north wall of Block A, was further excavated to a depth of 3 m below the ground surface without encountering pre Contact remains. Laboratory r esults The following section presents the results of all laboratory analyses of Bapot 1 materials. This includes summary descriptions of radiocarbon and artifact analyses conducted by Dr. Geof frey Cark and Olaf Winter, at AN U and a summary of the analysis of terrestria l fauna c onducted by Trevor Worthy, at AN U. The results of analyses of fish and shell fish are also presented. Bapot-1 r adiocarbon d ating. Twenty samples from Block A at the Bapot 1 site were submitted for radiocarbon analysis. These materials included un identified charcoal, pieces of
103 charred coconut shell, marine shell artifacts, and an Anadara cf. antiquata shell ( Table 3 2). Following Clark et al. (2010:26), two of these dates were excluded from further consideration in the current analysis. These included sample number Wk 23210, which yielded a date range of 38273703 cal a BP and Wk23753, which yielded a date range of 24382351 cal a BP. The date from Wk23210 was rejected because it may have been derived from a naturally deposited Anadara cf. antiquata shell and the date from Wk 23753 was rejected because it yielded an apparently â€œtoo oldâ€ date from a shallow context3. Charcoal for this date was collected from TU 7 near a subsurface feature. Based on the result of the radiocarbon analysis of this sample and re examination of the stratigraphic profiles, the feature in TU 7 was re interpreted as having been disturbed, possibly by a tree root, which may have displaced the sample that yielded the old date. Calibrated dates were plotted against depth ( Figure 3 7). A uniform deposition model was calculated using the â€œU Sequence Functionâ€ in OxCal v4.2.2. and these results were also plotted by depth to generate an agedepth model and identify depositional periods (Figure 38). This produced a date/depth relation similar to that described by Clark et al. (2010:26) and was used as a timeline for assessment of chan ges in pre Contact fisheries. According to this timeline, samples recovered from 40 cmbd dated to approximately 1400 cal a BP, samples from 60 cmbd dated to around 1500 cal a BP, samples from 80 cmbd dated to roughly 2000 cal a BP, samples from 110140 cmbd dated to approximately 2200 cal a BP, samples from 160210 cmbd dated to about 3100 cal a BP, and samples from 220250 cmbd dated to approximately 3300 cal a BP ( Figure 3 8). Clark et al. (2010:29) noted considerable variation in radiocarbon data that resulted R values for one shell sample (Wk 23771) ( Figure 3 7). This may be attributed to R values for shell dates. 3 These dates were primarily excluded because they conflicted with known information regarding depths and cultural materials from previous excavations (Clarke et al. 2010:26).
104 Artifact Analyses . Various types of artifacts were recovered during the excavation of Bloc k A. Artifact types included pottery, worked shell, worked stone, and worked bone. Analyses of these artifacts were conducted by Dr. Geoffrey Cark and Graduate Student Olaf Winter, at ANU, and results are summarized below. Pottery . Detailed analyses of pot tery recovered during the 2008 excavations of Block A were completed by Mr. Winter. A summary of his unpublished data is presented here, for comparative purposes. Although a partially intact Pre-Latte period pot was recovered, the assemblage generally cons isted of many (18,824) small, fragmented pre Contact pottery sherds. This total count consisted of 18,047 nondiagnostic sherds, 777 rim and carinated sherds, and 43 sherds with surface decorations. The pottery assemblage spans the entire sequence of preContact human occupation in the Marianas and included decorated sherds typical of the earliest dated contexts, which are Achugao type and San Roque type (Butler 1994:32). These two types are derived from what Moore (2002) describes as the Early Unai Phase ( 35003000 cal a BP) and feature rectilinear designs with in filled zones of small dots or dashes (Moore and Hunter-Anderson 1999: 490 492). A sherd with San Roque style surface decorations was recovered from TU 2, between 250 and 260 cmbd and had faint rec tilinear lines and small punctuation marks ( Figure 3-9A ). This sherd was recovered from excavation levels below 220250 cm. Samples from these depths dated to 31103362 cal a BP (see Table 3 2) ( Figure 3 7 and Figure 3 8) (Clark et al. 2010:26). An Achugau type sherd with distinctive limefilled curvilinear surface decoration patterns was recovered from TU 5, at 230240 cmbd ( Figure 3 9 B ). This type is indicative of pottery recovered from early contexts throughout the Marianas (Butler 1994: 32). The Achugao type
105 sherd is also from the earliest levels of B lock A and dates from between 3110 and 3362 cal a BP (Table 3 2) ( Figure 37 and Figure 38). Some sherds with deep depressions or incisions and circles were recovered during excavations. The firs t was taken from TU 5, 200210 cmbd ( Figure 3 10 A ) and the second was recovered from TU 3, 190200 cmbd ( Figure 310 B ). These appear to resemble the boldly incised, limefilled patterns typical of the Middle Unai Phase, dating to 2500 3000 cal a BP (Moore 2002). These patterns include bold lines, ci rcles, and chevrons. Also, both sherds were recovered from between 190 and 210 cmbd, and therefore likely coincide with radiocarbon dating samples from TU 1, TU 2, and TU 8 that were collected from 160 210 cmbd. These samples dated to approximately 3000 3065 cal a BP ( Table 3 2) (Clark et al. 2010:26). This is well within the date range for the proposed Middle Unai Phase (25003000 cal a BP). Sherds wi th surface decoration styles indicative of the Late Unai Phase (1600 2500 cal a BP) of the Pre -Latte Period were also recovered. Motifs from this phase include hollow circular designs, dots, and other simple lip or rim decorations ( Figure 212) (Moore 2002). An impressed rim sherd with a hollow circular pattern ( Figure 311 A ) was recovered from TU 6, at 180190 cmbd. Radiocarbon samples from this depth dated to approximately 30003065 cal a BP (Table 32) ( Figure 3 7 and Figure 3 8). The other included a small Pre-Latte Period sherd with deep limefilled punctuation marks ( Figure 3 11 B ). This sherd was collected from TU 6, at 130140 cmbd an d therefore coincides with radiocarbon samples collected from TU 2, 5, and 8 that date to between 2043 and 2189 cal a BP. The frequency of ceramics, plotted by depth below datum, indicates that the highest rates of ceramic deposition occurred in the inter with the largest percentage of sherds occurring between 220 and 230 cmbd ( Figure 312).
106 Samples collect ed from between 190 and 230 cmbd dated to approximately 3300 cal a BP and BP) ( Table 3 2) ( Figure 2 9 and Figure 3 -8). In general, the radiocarbon dates associated with various Pre -Latte decorated sherd types agree with the Pre-Latte Phases proposed by Moore (2002). Shell. The shell artifact assemblage varied considerably and included a substantial collection of shell ornaments and tools. Worked shell was by far the most common artifact, aside from pottery, and included 229 items. This comprised 43 pieces of worked shel l, 14 fishhooks, 118 shell rings, 10 Tridacna spp. adzes, 21 shell discs, and 23 worked sea urchin spines. Ornaments included various types of beads including ground Cypraea moneta shell beads with the top and bottom portions of the shell ground flat so that the interior portion of the shell is exposed ( Figure 313). Small cylindrical beads were also common ( Figure 314). These appear to have been m anufactured by grinding the apex of small conically shaped gastropod shells flat on both ends to form a disk. Holes were then drilled through the center of the small disk to form a bead. Shell ornaments from the early portion of the sequence included small ground Cypraea moneta and small cylindrical beads, large pendants made from the dorsal portions of Cypraea spp. shells ( Figure 3 15 A ), large disk sha ped pendants also made from the ground apex of a conical gastropod shell, possibly Conus sp. ( Figure 3 15 B ), and fragments of shell rings or bracelets, likely manufactured from the body whorls of large Trochus or Conus shells ( Figure 3 16). Similar shell artifacts have also been reported from other early Pre -Latte Period sites in the
107 Marianas including Achugao Point and Chalan Piao, Saipan and Unai Chulu on Tinian (Butler 1994; Craib 1999). Shell tools consisting of scrapers, chisel like implements, adzes, and fishhooks were recovered. Scrapers were m ost commonly made from bivalve shells of the genera Gafrarium, Fimbria and Tridacna. Most bivalve shell scrapers have worn posterior, ventral, and anterior margins ( Figure 3 17). These were likely formed from scraping vegetable matter while holding the valve by the dorsal margin and umbo. The modern use of bivalve shells in Fiji to scrape Pandanus spp. leaves for making mats creates similar wear pattern s on the edges of shells (pers. Obs., P. Oâ€™Day). Chisel like tools were made from large pencil urchin ( Heterocentrotus cf. mammillatus ) spines. These appear to have been manufactured by grinding down the end of a large spine to from a sharp round edge ( Figure 3 18 A ). Some spines seem to have broken, or cleaved, creating a similar pattern. However, close examination under a microscope revealed that the worked surfaces of chisels have striations from being ground ( Figure 3 18 B ). Shell adzes were mainly made from the umbo, or hinge portion of Tridacna s pp. clams ( Figure 3 19 A ). Tridacna adzes were present from later portions of the stratigraphic sequence and were entirely absent from the early Pre-La tte Period materials from the Bapot 1 site. This was also the case at Achugao Point (Butler 1994) and Chalan Piao (Moore et al. 1992), and Unai Chulu on Tinian (Craib 1993). Whereas Tridacna adzes were limited to the Latte Period, an adze like tool made fr om a large bivalve shell was recovered from approximately 180 190 cmbd ( Figure 3 19 B ). This artifact dated to 2907 cal a BP ( Table 3 2). Because it appears to be much thinner than adzes manufactured from Tridacna shell, this implement may have been used as a scraper.
108 Fishhooks were recovered from excavation levels between 100 cmbd and the deepest cultural deposits in Block A. These included â€œJâ€shaped hooks made from pearl shell ( Figure 320A ), a possible pearl shell fishhook blank ( Figure 3 20 B ), and a possible shell point for a compound fishhook ( Figure 3 21). Fishhooks and other types of fishing gear have been recovered from Latte Period and Pre -Latte Period contexts from various sites on Guam and the CMNI (Amesbury and Hunter-Anderson 2 003). Shell artifacts were recovered throughout all levels from Block A excavations. The frequency of shell artifacts plotted by cmbd, however, shows two major depositions of shell Figure 3 22). Samples recovered Lithic . Preliminary analyses of lithic materials recovered from Block A excavations identified approximately 115 stone or lithic artifacts. Lithic artifacts were recovered from all stratigraphic layers in Block A except for Layer I. Artifact types included chert and basalt flakes and debitage (n=82), sling stones (n=4), stone adzes and adze fragments (n=22), stone net sinkers (n=2), lithic cores or retouched flakes (n=5), and a single stone pounder. Most of the stone adzes were present in the deeper layers (Layers IV through VII). These included a large sub lenticu lar volcanic adze recovered from Layer V that measured 20 cm long by 10 cm wide and a cache of three adzes made from volcanic tuff, which were recovered from Layer VI (Clark et al. 2010:25 26). Lithic artifacts were recovered from all excavation depths except for 130140 cmbd. The frequencies of stone artifacts plotted by centimeters below datum indicate three major depositions of lithics at the Bapot 1 site. All of the other excavation depths contained smaller
109 numbers of lithics, ranging from two to four i ndividual artifacts ( Figure 323). The first occurred between 190 and 250 cmbd. Samples from these depths dated to between 3100 and 3300 cal a BP and therefore coincide with the Early Unai Phase of the Pre-Latte Period. The second occurred between 150 and 170 cmbd. Samples from this depth dated to approximately 3100 cal a BP and therefore coincide with the later portion of the Early Unai Phase. Samples from 3550 cmbd dated to approximately 1400 cal a BP and coincide with the Huyong Phase of the Pre-Latte Period ( Figure 29). Faunal materials . The fau nal assemblage recovered from Block A was relatively rich, consisting of bone from a comparatively limited number of vertebrate species and an extremely diverse assortment of marine invertebrates. Shells from two species of land snail were also recovered. Vertebrate remains consisted mostly of fish and bird bone. Reptile and mammal bone was also recovered, but limited to a few elements of gecko, monitor lizard, rat, and fruit bat. The remains of marine invertebrates included fragments of coral, shells of gastropods and bivalves, and sea urchin spines. Natural beach shells with waterworn surfaces and indeterminate fragments of shell were also found in the assemblage during laboratory analysis. A subsample of invertebrates was selected for analysis. This inclu ded all of the invertebrate remains from TU 1, TU 4, and TU 6. Inverte brate s . Excavations at Block A produced a particularly rich and well preserved collection of invertebrate remains. The analyzed sample of invertebrate remains totaled 21,051 identified individual specimens weighing 42.99 kg. This included coral, sea urchin remains, gastropod shell, bivalve shell, beach shell, and indeterminate shell. Only two fragments of coral were recovered, which accounted for <1% of the invertebrate assemblage. Gastr opod shell accounted for 44% of the assemblage, bivalves accounted for 53% of the assemblage, large
110 spines from the pencil urchin ( Heterocentrotus cf. mammillatus ) accounted for 1%, and beach shell accounted for 2% of the total assemblage ( Table 3 3). Many shells were whole and could be readily identified to the species level. Broken and fragmented specimens also displayed attributes that allowed easy identification. Gastropods were most diverse and included 9,238 identified shells from at least 37 marine and two terrestrial species. All of the identified gastropods are generally common on reefs, reef flats, and in lagoons on the islands of Saipan and Guam (Table 3 4). The most common species recovered included Cellana radiata orientalis (Rayed wheel limpet), which consisted of 3,270 specimens and 35% of the total NISP, Turbo spp. (turban shells), which consisted of 1,765 specimens and 19% of the total NISP, and Strombus gibberulus gibbosus (mutable conch), which consisted of 805 specimens and 9% of the total NISP. Specimens relegated to the category Class Gastropoda included shells that were fragmented and could not be accurately identified to family, genus, or species. These consisted of 1,428 specimens and accounted for 15% of the total NISP. All of the other identified shell specimens belonged to species that were minor contributors to the total NISP of the assemblage and each accounted for 13% or less of the total NISP (Table 3 5). Shell from approximately 19 identified species of marine bivalves was recovered. This included 11,128 individual identified specimens. Most of these are currently common on reefs and in lagoons in the Mariana Islands (Table 3 6). The most common species according to count was Gafrarium pectinatum (the comb Venus clam), which accounted for 30% of the assemblage. Anadara cf. antiquata (the arc clam) accounted for 25% of the assemblage and is currently rare in the Marianas (Paulay 1992). This is likely related to modern development of the coastlines of islands in the Marianas and the reduction of this speciesâ€™ preferred habitat, mangroves (Paulay
111 1992). In general, however, it seems that bivalves are not currently harvested in substantial numbers (pers. obs., P. Oâ€™Day). Tellin shells, or members of the Family Tellinidae, accounted for approximately 20% of the assemblage. Identified genus and species of te llin shells included Tellina scobinata (rasp tellin) and Quidnipagus palatam (palate tellin). Shells that could not be confidently identified beyond Class Bivalvia accounted for 20% whereas all other identified species collectively accounted for approximat ely 5% of the total bivalve assemblage ( Table 3 7). Vertebrates. Bones from fish, reptiles, birds, and small mammals were recovered during excavations of Block A. Most of the bones were small and consisted of fragmented elements. Very few specimens were identifiable beyond class (Chondrichthyes, Osteichthyes, Reptilia, Aves, or Mammalia). The remains of marine fish were the most numerous, followed by bir ds. Bones from reptiles and mammals were relatively rare in the assemblage. A total of 834 specimens of fish bone weighing 128.8 g were recovered from Block A. Fish bone was dominated by very small vertebrae, spines, and fragments. Because of this, most of the fish bones could only be definitively identified to Class Osteichthyes. This class includes all bony fishes and consists of 45 orders, 435 families, and some 28,000 species. A little more than 1,000 species within Class Osteichthyes are known to inhabit inshore areas of the Marianas (Myers and Donaldson 2003). Approximately 797, or 95%, of the fish bone assemblage was categorized as Class Osteichthyes ( Table 3 8). A single shark tooth (Order Carcharhiniformes, or requiem shark) and a total of 37 bone specimens were identified to the family level. They included parrotfish (Scaridae), a jack (Carangidae), a triggerfish (Balistidae), and surgeonfish (Acanthuridae). These four identified families of fish inhabit reefs and lagoons and can be observed in shallow inshore waters on both Saipan and Guam and are among the 13 most common families identified in archaeological assemblages from Guam (Olmo 2013:10). These
112 famili es also contain a relatively large number of species. Currently, about 16 species belonging to Order Carcharhiniformes, 54 species of parrotfish, 38 species of jack, 28 species of triggerfish, and 66 species of surgeonfish live in the Marianas (Myers and D onaldson 2003). These taxa display a wide range of specific feeding and spawning behaviors. Capturing certain species, even different species of the same family, probably required different strategies and gear types. Parrotfish were the most common family and accounted for 5% of the total NISP and 23% of the total weight of the fish bone assemblage ( Table 3 8). This is because lateral skull elements fr om parrotfish, including dentaries, premaxillas, and especially their robust dental, or pharyngeal plates, are distinctive and were well preserved. Parrotfish feed mainly upon algae from rocks and corals and their dense dental plates are used to grind cora l. It is believed that these fish create most of the carbonate sand found on beaches through this process (Dye and Longenecker 2004:61). Currently there are seven genera and 54 species known in the Marianas from the family Scaridae (Myers and Donaldson 2003). Because parrotfish remains were relatively prevalent in the assemblage, various species of parrotfish in the Marianas were collected for comparative specimens to facilitate bone identification. This proved to be difficult because modern reef fish stocks in the Marianas have been overfished. Fishing trips to collect specimens were often unsuccessful and generally only small specimens (averaging approximately 249 mm total length) from a limited number of species ( Scarus psittacus , S. gibbus , S. altipinus , and Chlorurus frontalis ) could be captured or bought at local fish markets. Terrestrial vertebrate remains accounted for a comparatively small percentage of the overall faunal assemblage at Bapot 1 (Table 3 12). Whereas specimens of reptile, bird, and mammal bone are limited, their remains may provide insight into changing preContact diets, resource exploitation strategies, site or household use patterns, and possibly, genderor age -
113 specific subsistence behaviors. Skeletal remains from reptiles, birds, and mammals from Block A were analyzed by Trevor Worthy from ANU and summaries of his analysis are presented below. Two species of reptile were identif ied from ten bone specimens. These included four bones identified as gecko and six bones identified as Varanus cf. indicus , or the mangrove monitor lizard (Table 3 9). Gecko bone occurred throughout most of the sequence at depths from 50 to 210 cmbd and date from 13501599 and 28903214 cal a BP (Table 3 2). Geckos are widespread throughout Oceania and species from an extensive fossil lizard assemblage have been recorded from the Marianas (Steadman 2006:63). Geckos are generally small lizards and do not represent an important food resource for humans. Their remains were likely introduced into the siteâ€™s matrix through natural processes. Six bones from Varanus cf. indicus were recovered from approximately 5062 cmbd. Radiocarbon samples recovered from this depth produced a date of 14151525 cal a BP (Table 3 2). Interestingly, this large lizard may have been introduced to the Marianas during the pre Contact period. This issue is currently unresolved and in need of further study. Whereas there is a Chamorro name (Hilitai ) for the monitor lizard, suggesting an ancient familiarity with the species, archaeological evidence for its preContact introduction is limited. The limited remains from the present s tudy cannot resolve this issue. Further confounding this matter is the widespread natural distribution of this species, their tolerance to saltwater environments, their ability to cross large bodies of water, and the lack of fossil remains of the lizard from pre human contexts (Cota 2008). A total of 215 bird bones were identified, including three species of seabirds and eight species of landbirds (Table 3 10). Seabirds, including a storm petrel and terns, accounted for a small percentage of the assemblage, with three identified bones. Each represented <1% of the total number of identified bird bone. Landbirds were more diverse and consisted of a wading
114 bird, flightless rails, a crake, and numerous small birds (warblers, white eyes, starlings, honey eaters, and other small sparrows or passerines). Indeterminate bird bone dominated the avian assemblage with a total of 119 specimens accounting for 55% of th e collection. Bone from most identified landbird species was generally limited. Wading birds, crakes, warblers, white eyes, honey eaters, and other passerines accounted for 1% of the total assemblage. Skeletal remains from starlings accounted for 4% and r ails were the most common after unidentified birds, accounting for 34% of the total assemblage. Remains of the flightless rail were present throughout most of the sequence, from 85 230 cmbd in stratigraphic Layers II through VII (Table 3 1) ( Figure 33 Figure 3 6). Radiocarbon dated samples from these layers ranged between 1949 2041 cal a BP at 7080 cmbd, 31613445 cal a BP at 220 cmbd, and 29803318 cal a BP at 230 cmbd (Table 3 2). A total of four mammal bones from two species were identified. These included two rat and two fruit bat bones (Table 3 11). Rat is only present in the later portion of the sequence, between 49 and 90 cmbd. Radiocarbon samples from these depths produced dates that ranged is coincides roughly with the late part of the Pre Latte Period and corroborates a relatively late preContact period introduction of rat to the Marianas (Steadman 1999). Two species of rat, including the Pacific rat ( Rattus exulans ) and the Asian house ra t ( Rattus tanezumi , sometimes referred as Rattus rattus monsorius ), were introduced to Micronesia prior to European contact. Currently only the Pacific rat has been identified in archaeological contexts in the Marianas (Wickler 2004). It is currently debated whether it was introduced intentionally or unintentionally, as a stowaway (Kirch 2000; White et al. 2000; Wickler 2004). Its limited occurrence in Latte Period contexts in the Marianas raises an interesting question regarding this debate. If the introduction of rat was unintentional, then it
115 would be reasonable to as sume that it would have found its way into earlier Pre-Latte deposits following early voyages (Wickler 2002:35). Two fruit bat (Pteropid) or flying fox bones were recovered from between 170 and 190 cmbd, roughly corresponding with Layers V and VI (see Table 3 1) ( Figure 3 3 Figure 36). A charcoal sample collected from between 180 190 cmbd produced a date range of 29463059 cal a BP ( Figure 3 7). This corresponds approximately with the end of the Early Unai and beginning of the Middle Unai Phases of the Pre -Latte Period. Fruit bats were likely hunted and eaten throughout the entire pre Contact pe riod (Steadman 2006:503). Currently, these animals are highly prized food items and are extremely rare in the Marianas, having been hunted to the point of extirpation. Marine r esource analysis. The primary objective of the present project is to gain a comp lete understanding of the roles humans played in prehistoric interactions with marine environments in the Marianas. Zooarchaeological and fisheries sciencebased analyses were essential to this primary objective. The methods and results of fisheries science based analyses are presented in Chapter 4 and the following sections focus upon analyses of the archaeological residues of marine resources, using zooarchaeological methods. These basic zooarchaeological analyses included: 1) identification of all econom ically important species exploited in the past, 2) identification of the habitat preferences of these species, 3) determination of how frequently these species were exploited through time, and 4) measurement of the diversity of the exploited marine species in the assemblage. To incorporate a time dimension, all data were analyzed with respect to both depth below the excavation datum, i.e. with respect to the 10 cm thick excavation levels, and according to the timeline determined using radiocarbon dates ( Table 3 2) ( Figure 3 7 and Figure 38).
116 For most previous archaeological work in the Marianas, measuring tre nds or changes in marine resources through time has relied largely upon comparing NISP or weight frequencies of shell and bone samples from different excavation levels and stratigraphic layers. This was because budgetary constraints of cultural resource management projects limited the number of radiocarbon dates that could be obtained and precluded consistent analyses of faunal materials. Frequencies of NISP were also plotted by excavation level, or depth, and according to the deposition model derived from radiocarbon dates ( Figure 3 8). This was done because differences between data plotted by depth vs. data plotted by age can occur and bias interpreta tions. Depth is an arbitrary measure used as a sampling control during archaeological excavation and may show changes in NISP frequencies throughout the stratigraphic sequence, but when plotted, these data will show a contiguous pattern that does not repre sent the actual sequence of events related to the deposition of anthropogenic soils within an archaeological site. Whereas, data plotted according to a sequence based on radiocarbon dates will show a more realistic pattern of deposition. This will incl ude gaps in the sequence that may represent different rates of midden deposition, alternating periods of site occupation and abandonment, or sifts in activity areas within a site. Although additional radiocarbon dates may be necessary to reconstruct a more accurate occupational sequence for the site, 18 radiocarbon assays were available for the present study. This is a considerably larger suite of dates than has been obtained by other projects in the Marianas that were mostly limited to construction related contract work. Shells dominated the faunal assemblage from Block A, with bivalves accounting for 50% and gastropods accounting for 42% of the total NISP. Other invertebrate remains, including coral, sea urchin, beach shell, and indeterminate shell accoun ted for the remaining 8% of the total NISP ( Table 3 12). Vertebrate remains from marine species consisted entirely of fish bone.
117 Although fish dominat ed the vertebrate portion of the faunal assemblage, further underscoring the importance of marine resources, it is important to reiterate how disproportionately large the invertebrate assemblage from the Bapot 1 site is with fish bone accounting for only 4% of the total NISP and <1% of the total ( Table 3 12). Fish . The fish bone assemblage from Block A was small and fragmented, precluding the application of any fisheries sciencebased analyses and severely limiting the application of basic zooarchaeological analyses. However, the general lack of fish bone in the Bapot 1 site assemblage raises an important question regarding the human marine env ironment interaction. This relates to how the faunal assemblage reflects resource exploitation behaviors, intrasite activities, and the limitations of local environmental conditions. These may have included gender, age , or status related exploitation pr actices, specialized activities relegated to specific work areas, e.g. the fringing reef, lagoon size, and exposure to high seas and strong currents. Frequencies of NISP for fish bone were plotted by depth and date following the deposition model created us ing OxCal v4.2.2 ( Figure 3 8). When plotted according to depth, fish bone was more frequent within upper levels of the stratigraphic sequence, between 49 and 80 cmbd ( Figure 3 24 A ). Following the depositional model based on radiocarbon dates ( Figure 3 8), materials recovered from these depths dated to the Huyong Phase of the PreLatte Period around 14001500 cal a BP. Spikes, or increases in the frequency of fish also occurred at 100120 cmbd and 190210 cmbd (Figure 3 24 A ). Materials from 100120 cmbd dated to 2200 cal a BP and samples from 190210 cmbd dated to 30003065 cal a BP (Table 3 2). These dates coincide with the Late Unai Phase (1600 2500 cal a BP) and the Middle Unai Phase (25003000 cal a BP), respectively. The lowest frequencies of fish bone occurred in the deepest levels between 220 and 240 cmbd, dating to 3300 cal a BP.
118 When plotted according to date, fishbone was most frequent in samples dating to 3100 cal a BP, coinciding with the Early Unai Phase (3000 3500 cal a BP) and in samples dating to 2000 cal a BP, which coincides to the Late Unai Phase (2500 1600) ( Figure 3 24 A ). Fish bone was less frequent in samples dating to 2200 cal a BP (during the Late Unai Phase) and at 1500 cal a BP (during the Huyong Phase), prior to the start of the Latte Period. Overall, fish remains were most commonly deposited at Bapot 1 during the later phases of the Pre -Latte Period ( Figure 3 24 B ). The low frequency of fish bone in general indicated that the means by which fish bones were deposited at Bapot 1 were limited. Marine invertebrates. T he following section considers changes in the frequencies of NISP, the number of identified species, and species diversity, by depth and by date before present. This included differences in the frequencies of all invertebrate species and differences in gas tropod vs. bivalve species. Differences in NISP frequencies of gastropods vs. bivalves were apparent through time and bivalves generally occurred more frequently in excavation levels between 104 and 260 cmbd (Figure 325 A ). Gastropods are only slightly more frequent in levels from and 6595, 181190, 210, and 230 cmbd with the overall NISP frequency of both gastropods and bivalves decreasing between 65 and 113 cmbd. Bivalves appear to have been harvested more frequently than gastropods throughout much of the sequence, especially between 122250 cmbd. When plotted by date, bivalves are also more frequent throughout most of the sequence, between 2000â€“3300 cal a BP ( Figure 325 B ). Gastropods were only slightly more frequent at 1500 cal a BP. Commensurate with NISP frequencies plotted according to depth, NISP frequencies plotted by date decreased later in the sequence, between 1500 and 2000 cal a BP. Given these dates, this shift from higher frequencies of bivalve species to higher frequencies of
119 gastropod species occurred late in the PreLatte Period, du ring the Late Unai Phase (1600 2500 cal a BP) at the Bapot 1 site. Measuring species diversity within the invertebrate assemblage through time was important in regard to assessing procurement strategies, species preference, and the types of habitats that were exploited. The diversity of the invertebrate assemblage was inv estigated to look more closely at the significant changes in frequencies of gastropods and bivalves and to highlight possible differences in habit preferences and shell use between the two invertebrate classes. That is, many of the more commonly identified species of bivalves, like Anadara cf. antiquata or Gafrarium pectinatum prefer soft sandy or silty bottoms, whereas the majority of the identified gastropod species prefer rocky substrates and reef. First, the number of identified species in each excavati on level and dated provenience was plotted for the total invertebrate assemblage, then for gastropods and bivalves separately. This was also done for calculated Shannon-Wiener diversity index values. Plots of species counts were also important in determini ng whether the number of species within the invertebrate assemblage was fluctuating through time or if the number of individuals with a species was fluctuating. This is because the Shannon-Wiener function produces a higher diversity index value with increasing numbers of new species and with greater equitability in the distribution of individuals among the species (Reitz and Wing 2008:111112). The total count of identified invertebrate, gastropod, and bivalve species varied between excavation levels throug hout the sequence. Generally, numbers of identified species oscillated with peaks at 230 240, 190200, 150160, 109130, and 90104 cmbd (Figure 3 26 A ). Lower numbers of species occurred at depth intervals of 240260, 170190, 140153, 100 113, and
120 4885 cmbd. Both high and low counts of identified species were relatively consistent throughout the sequence and there were no extreme high or low values. W hen plotted by date, the frequencies of all identified species of invertebrates were highest in samples dating to between 2000 and 3300 cal a BP ( Figure 326 B ). In samples dating to 3300 cal a BP, a total of 48 species were identified; at 3100 cal a BP, 54 species were identified; at 2200 cal a BP, 63 species were identified; and at 2000 cal a BP, 51 species were identified. The number of identified species declined after 2000 cal a BP and 24 species were identified in samples dating to 1500 cal a BP. The number of identified gastropod and bivalve species followed a similar pattern. When plotted by date, the numbers of both gastropod and bivalve species w ere highest in samples dating to between 3300 and 2000 cal a BP and decreased in samples dating to 1500 cal a BP ( Figure 326 B ). Calculated ShannonWie ner index ( ) values for the total invertebrate assemblage, gastropods, and bivalves were plotted by depth (cmbd) and by date before present. When plotted by depth, Shannon-Wiener values for all invertebrates oscillated around a mean value of 2.39 throughout the occupational sequence of the site. When diversity values for gastropods and bivalves were plotted separately by depth, the dominance of values oscillated back and forth between the classes through time ( F igure 327 A ). Values for gastropod species oscillated around a mean value of 1.24 and values for bivalves oscillated around a mean value of 1.15. More specifically, bivalve samples f rom between 240 and 260 cmbd had higher values than gastropod species. At t 230 cmbd this trend was reversed, with higher diversity values for gastropod species. Values for gastropods and bivalves oscillate back and forth between 230 and 200 cmbd with higher values for bivalves at 220 cmbd. At 200 cmbd values for gastropods are higher until 140 cmbd, where values for bivalves are higher until 113 cmbd. At this point values
121 for gastropods and bivalves converge at around 1.2. After 113 cmbd, diversity values for gastropod species became dominant ( Figure 3 26 A ), like species counts plotted according to depth below datum (Figure 3 27 A ). Frequencies of identified species, or species richness , and diversity index values plotted according to date illustrate similar pattern s. The number of all identified invertebrate species and the number of species for each class are roughly similar in samples dating to 2000, 3100, and 3300 cal a BP, with a peak in the number of identified species at about 2200 cal a BP. The sample dating to around 1500 cal a BP contained the lowest numbers of identified species ( Figure 3 26 B ). The distribution of values for all invertebrates plotted by date suggests that species diversity decreases from 2.51 at 3300 cal a BP to 2.43 at 3000 cal a BP then increases to 2.50 at 2000 cal a BP and to 2.64 at 2000 cal a BP. At cal a BP 1500 the value for total invertebrates decreases to 2.35 ( Figure 327 B ). When both classes were plotted separately by date, the distribution of values suggested that the species diversity of gastropods slightly decreases through time until 2200 cal a BP while the species diversity for bivalves slightly decreases from 1.23 at 3300 cal a BP to 1.16 at 3100 cal a BP and then increases to 1.24 at 2200. At 2200 cal a BP, values for gastropods and bivalves are about the same. After 2200 cal a BP the diversity of species in each class diverged and values for gastropod species increased to 1.42 at 2000 cal a BP and then decreased to 1.37 at 1500 cal a BP. Diversity index values for bivalves decrease to 1.21 at 2000 cal a BP and to .97 at 1500 cal a BP ( Figure 3 27 B) . There were distinct differences in species diversity values between gastropod and bivalve species. Gas tropods consisted of more individual taxa categories and lower NISP (44% of the total faunal assemblage), whereas bivalves consisted of less individual taxa and higher NISP
122 (50% of the total faunal assemblage) than gastropods. The possible increase in the diversity of gastropods through time was because samples from this class contained more individual taxa and the abundance of NISP of each gastropod taxon was more evenly distributed between samples. Lower diversity values of bivalves is attributable to a lower number of identified individual taxa within the invertebrate samples and disproportionately high abundance of a NISP of a few taxa, particularly high NISPs of Gafrarium pectinatum , accounting for 16% of the total invertebrate assemblage, and Anadara cf. antiquata, accounting for 13% of the total invertebrate assemblage. Only one gastropod species had a disproportionately high abundance. This was Cellana radiata orientalis , or the rayed wheel limpet which accounted for 16% of the total invertebrate as semblage. Testing at the Bapot 1 site identified eight stratigraphic layers, seven of which contained cultural materials. Materials from these layers produced dates ranging from 3385â€“3446 to 1284â€“ 1316 cal a BP (Table 3 2). These dates indicated that the Block A portion of the Bapot 1 site was occupied throughout the Pre -Latte Period, including the Early, Middle, and Late Unai Phases and the Huyong Phase. Artifact analyses conducted by Mr. Winter from AN U showed that all pottery styles indicative of the Pre-Latte Period were recovered. These included samples of the earliest decorated styles, including San Roque and Achugao type sherds. Decorated sherds typical of the middle and later portions of the Pre -Latte Period were also recovered, including incised, punctated, limeimpressed, and decorated rims. Pottery sherds were generally fragmented and small, but recovered from all cultural layers. The highest densities of sherds were recovered from the deepest stratigraphic layers, dating to between 3003 3141 and 3300 cal a BP. According to
123 the distribution of the various pottery types and associated radiocarbon dates, th e Bapot 1 site pottery assemblage generally agreed with the chronological sequence proposed by Moore (2002). Shell and lithic artifacts were also more frequent in deeper excavation levels dating to between 3003â€“3141 and 33853446 cal a BP. Shell artifact types were the most diverse and consisted of several kinds of ornaments and tools. Invertebrate species remains included coral, sea urchin, gastropod shell, bivalve shell, beach shell, and indeterminate shell. Collectively, 21,051 individual specimens wei ghing 42.99 kg were analyzed for the present study. Coral fragments and pencil urchin (Heterocentrotus cf. mammillatus ) spines were rare in the assemblage and their presence is related to their use as tools rather than important subsistence items. Gastropo ds accounted for 44% of the assemblage and were most diverse, with 9,238 individual shell specimens from about 37 species. Bivalves accounted for 53% of the assemblage with 11,128 individual specimens from approximately 19 species. Aside from two species o f land snail, all invertebrates identified in the Bapot 1 assemblage are known to inhabit inshore marine environments including the littoral zone, mangroves, lagoons, and coral reefs. Some of these are extremely rich and productive environments. The most common gastropod species included the rayed wheel limpet accounting for 35% of the total NISP. Bivalve NISPs are dominated by shells from three species. These include Gafrarium pectinatum accounting for 30% of the assemblage, Anadara cf. antiquata accountin g for 25% of the assemblage and Tellin shells accounting for 20% of the assemblage. Vertebrate faunal materials recovered from Bapot 1 comprised marine and terrestrial species including bones from fish, birds, reptiles, and mammals. Although fine screens and bulk sediment sampling were employed during excavation, a relatively small sample of fish remains was recovered. These consisted mostly of small undiagnostic elements of heavily fragmented
124 and burned bone. Therefore, most of the assemblage could not be identified beyond Class Osteichthyes. Disproportionately low amounts of identified fish remains beyond Class Osteichthyes were also reported in assemblages from other archaeological sites on Guam. Excavations at Tarague recovered 217 identifiable parrotfi sh remains from a total of 7,002 fish bones (Amesbury and Hunter-Anderson 2003:49). By weight, the Tarague remains consisted of 52.67 g of fish bone identified to the family level and 207.33 g identified only as fish (Olmo 2013:1). Olmo (2013:1) also cites results of 1992 excavations on Orote Peninsula, Guam that recovered over 1,100 fish bones, of which only 59 elements were identified to the family level. Bones from terrestrial species, including reptiles, birds, and small mammals comprise a relatively sm all portion of the overall faunal assemblage. Reptile remains included bones from geckos and monitor lizards. Geckos do not represent a food resource, and it is currently debated if the monitor lizard was a prehistoric or historic period introduction. Bird bone was the most common terrestrial fauna recovered during excavations and included bones from seabirds, landbirds, and small birds. Mammals were represented by four bones from two species. These included two rat and two fruit bat bones. Rat is only pre sent in the later portion of the sequence and corroborates a relatively late preContact period introduction of rat to the Marianas (Steadman 1999). Two fruit bat bones were recovered from Layers V VI, but were likely hunted and eaten throughout the entire pre Contact period (Steadman 2006:503). Currently, these animals are highly prized and have been nearly extirpated in the Marianas. Marine resources comprising invertebrate shell and fish bone dominated the faunal assemblage from the Bapot 1 site. Althoug h fish dominated the vertebrate assemblage, fish bones were small and fragmented and could not be subjected to statistical analyses. When plotted by
125 depth, fish bone was most frequent in the upper excavation levels dating to around 14001500 cal a BP. However, when plotted according to date frequencies of fish bone NISP fluctuate between higher and lower numbers through time with one major increase between 3100 and 3300 cal a BP. This appears to show that there were no major shifts in the frequency of fish bone NISP after 3100 cal a BP. Distributions of invertebrate frequencies show that the most intensive harvests of gastropods and bivalves occurred between 2200 and 3100 cal a BP. Also, bivalves were more frequently harvested than gastropods throughout mos t of the sequence, except at 1500 cal a BP. This late shift in the dominance of gastropod shell appears to be minor, however, and occurred earlier during the Pre -Latte Period at the Bapot 1 site than at other sites in the Marinas. These include Chalan Piao on Saipan (Amesbury et al. 1996) and at various sites on Guam (Amesbury 2007, 1999; Graves and Moore 1985; Leidamann 1980) where this shift happens later with the onset of the Latte Period at 1000 cal a BP . To further evaluate this shift in gastropod and bivalve frequencies, diversity and species richness of the invertebrate assemblage was analyzed. First, frequencies of the number of identified species were plotted by depth (cmbd) and date for all invertebrates, gastropods, and bivalves. Second, calculated ShannonWiener diversity index values were plotted by depth and date for all invertebrates, gastropods, and bivalves. When plotted by depth, frequencies of identified species for all invertebrates, gastropods, and bivalves fluctuated in the levels throu ghout the stratigraphic sequence. When plotted by date, frequencies of identified species increase from 3300 cal a BP to the highest number of species at 2200 cal a BP. After 2200 cal a BP, the numbers of species decreases to the lowest species count at 15 00 cal a BP.
126 When plotted by depth, values oscillated between higher and lower values. When plotted separately, the values for gastropod and bivalve species also oscillate between higher and lower index values through time, with bivalves becoming mo re dominant over gastropods during two portions of the sequence. When plotted by date, the distribution of values for all invertebrates, gastropods, and bivalves indicated that diversity values also oscillated between higher and lower values. It also sh owed that gastropods have higher diversity values than bivalves throughout the sequence except at 2200 cal a BP where values for gastropods and bivalves are almost equal. After 2200 cal a BP, values for each class diverge, with an increase in gastropod diversity values to its highest level at 2000 cal a BP and bivalve diversity values decreasing to its lowest level at 1500 cal a BP. The gastropod assemblage consisted of more individual taxa categories and lower NISPs, whereas bivalves consisted of les s individual taxa and higher NISPs. Increases in gastropod diversity values are likely a consequence of higher frequencies of individual species that were more evenly distributed among taxa. Lower bivalve diversity values are likely attributable to a lower number of identified individual taxa and the disproportionately high abundance of Anadara cf. antiquata, Gafrarium pectinatum , and Telin shells. Also, the use of NISP for calculating values has likely led to high estimates of diversity for bivalve shel ls and some gastropod species. This is because each individual bivalve consists of two shells, whereas, most gastropods consist of a single shell. Therefore, index values for bivalves plotted on Figure 3 27 B are likely lower than they appear. Ritidian Grotto Excavations at the Ritidian Grotto site for the present project were part of a long term collaborative research project conducted by researchers from the MARC at the University of Guam and from the University of Hawai â€˜i, M noa, in the Ritidian Unit of the Guam National
127 Wildlife Refuge (Carson 2012). This included two 1x 1 m test units (TU 3 and TU -4). Sampling included the use of finemesh screens and the collection of bulk sediment samples for sorting in the laboratory to optimize the recovery of shell and bone elements from small animals. The primary purpose of employing these methods was to collect comparable archaeological faunal materials and artifacts. TU 3 and TU 4 were excavated at the north end of the interior portion of a rockshelter. The interior of the shelter is level and covers an area approximately 15 m long by 3 m wide. There is also a small cave located at the southeast corner of t he rockshelter. The entrance of the cave is roughly 2 m above the floor of the rockshelter and there are very faint pictographs on the interior walls of the cave. There are also small lusong, or grinding basins, in the bedrock of the rockshelter, approxima tely 2 m southeast of TU 2 ( Figure 3 28). Stratigraphy Excavations of TU 2, TU 3, and TU 4 identified four stratigraphic layers (Layer I IV) (Table 3 13) ( Figure 329 and Figure 330 ). Layer I was a dark gray very fine silty calcareous sand and contained Latte Period pottery, bone, and marine shell. Shell from the giant African land snail ( Achatina fulica) was also present in the upper portion of Layer I. This species was introduced to Guam in 1945 and to the Northern Marianas Islands between 1936 and 1938 (Cowie 2000:149). Layer II consisted of light brown, very fine loose sand. This layer ra nged from approximately 3040 cm thick and contained pottery, bone, marine shell, worked stone, and shell artifacts. Latte Period pottery was present in the upper portions of Layer II and a Pre Latte rim sherd was recovered from the lower portions of Layer II, at 68 cmbd. Layer III was pale brown in color and consisted of very loose, fine sandy soil. Pottery, bone, and shell were present. Artifacts were typical of the PreLatte Period. These included thin red slipped pottery and small shell beads. Layer III consisted of pale brown very loose fine sand and contained red-
128 slipped Pre -Latte Period pottery, non human bone, shell, and fragmented burnt human bone. Small shell beads were also present. Layer IV consisted of pale brown, very loose fine sand. No cultur al materials were present and this layer is undisturbed, or unmodified, naturally deposited sand. Laboratory Results The results of laboratory analyses are discussed in the following section. These include results of radiocarbon dating analyses and summaries of the analysis of artifacts, faunal materials, and human remains. Results of analyses of marine resources are also presented, followed by a summary and discussion of the analysis of the Ritidian Grotto site materials to conclude the chapter. Rit idian Grotto Radiocarbon Dating. Materials from TU 2, TU 3 and TU 4 could not be dated for the present project. Relative dates were derived from radiocarbon analyses of materials collected from TU 1. This excavation unit was located approximately 4 m south of TU 2 (see Figure 328). Test Units 1, TU 3, and TU 4 were all excavated on the same level interior surface of the rockshelter, sectioned the same s tratigraphic layers and recovered similar types of cultural materials. For the purposes of this study, I assumed that they are chronologically similar. Profiles of the walls of TU 1 were unavailable; however, the author participated in the excavations of TU 1 in 2007. Originally, 13 samples from TU 1 were submitted for radiocarbon analysis. Four of these were rejected from further consideration for this project because they were incompatible with the stratigraphic contextual information. They consisted of excessively old dates from a shallow layer (Beta239576), too recent dates derived from deep layers (Beta 239572 and Beta 239582), or were from a level in TU 1 that was deeper than the bases of TU 2, TU 3, and TU 4 (Beta239579) (Table 3 14).
129 Using OxCal v4.2.2 (Bronk Ramsey 2013) the calibrated dates from TU 1 were plotted with a 68.2 percent probability according to depth of the sample below the ground surface ( Figure 3 31). In order to measure changes in marine faunal remains through time, a deposition model was also created using OxCal v4.2.2 (Bronk Ramsey 2013) ( Figure 3 32). According to the deposition model, materials recovered from 3040 cmbd dated to approximately 989 29 cal a BP, materi als from 4050 cmbd dated to about 1181 29 cal a BP, materials from 5060 cmbd dated to around 1370 40 cal a BP, materials from 100 cmbd dated to approximately 1790 40 cal a BP, a sample from between 88 105 cmbd dated to around 2810 40 cal a BP, ma terials from 115120 cmbd dated to about 2970 40 cal a BP, and samples from 105125 cmbd dated to 3140 40 cal a BP (Table 3 14) ( Figure 3 32). Whereas diagnostic pot sherds and artifacts recovered from TU 3 and TU 4 generally agree with the proposed sequence listed above, additional radiocarbon dates derived from TU 3 and TU 4 materials may alter this sequence. There is significant overlap be tween radiocarbon dates from 3040 and 4050 cmbd. This may be the result of recent disturbances of the upper soil layers in the rock shelter as suggested by the presence of African land snail shell. There is also a significant time gap between dates from 110120 cmbd and 120130 cmbd. This may have resulted from the wall collapse that occurred during excavations of the deeper layers of TU 1. These layers consisted of very loose dry sand that was difficult to excavate, which limited the collection of databl e samples. Artifact analysis. Artifacts recovered from TU 3 and TU 4 at the Ritidian Grotto site are described in the following sections. These consist of summary descriptions of the recovered artifact types included pottery, worked shell, and worked cora l.
130 Pottery . The pottery recovered from TU -3 and TU 4 was analyzed by Nicole I. Vernon using the methods, typologies, and attribute descriptions developed by Moore (2002). Generally, the assemblage consisted of small fragmented pieces of pottery with very f ew diagnostic sherds. No articulating sherds from whole or partial pots were recovered, precluding vessel analyses. The total assemblage consisted of 427 pot sherds, including 205 from TU 3 and 222 from TU 4. This included 18 rim sherds and 187 body sherds from TU 3 (Table 3 16) and 18 rim sherds and 204 body sherds from TU 4 (Table 3 17). Thirty seven diagnostic sherds were recovered and accounted for 9% of the assemblage. Diagnostic sherds possessed attributes indicative of time period and included one limeimpressed body sherd from TU 3 (Table 3 16) and 36 vessel rim fragments, including four redslipped everted Pre-Latte rim sherds recovered from TU 4 (Tabl e 3 17). Attributes such as redslipped thin everted rims and limeimpressions have been recovered from contexts dating to various Pre Latte Periods and their presence in the current assem blage generally conform to the pottery sequence proposed by Moore (2002). A total of 390 undiagnostic body sherds were recovered, accounting for 91% percent of the total pottery assemblage. Shell and w orked c oral . A total of 28 artifacts weighing 22.2 g were recovered from TU 3 and TU 4. TU 3 yielded 10 artifacts with a combined weight of 13.2 g ( Table 3 18) and TU 4 produced 18 artifacts weighing 9.0 g ( Table 3 19). These consisted of worked shell and a single coral tool. Artifact types included shell beads, worked shell items, scrapers, fishhook fragments, and a coral file, or abrader. Beads were present in all stratigraphic layers and consisted of various types. These included worked Cypraea sp. shell (Figure 3 33 A ) and a modified whole Conus sp. shell ( Figure 333 B ). Small Conus sp. disk beads ( Figure 3 34 A ) were the most common type found, with a
131 total of seven recovered from TU 3 and a total of 12 from TU 4. The manufacturing process of these beads has been previously described by Moore et al. (1992:6871) and is illustrated by comparing a bead blank made from the apex of a Conus sp. shell with a finished bead ( Figure 3 34B ) recovered from excavations. Note that similar sutures are visible on both the blank and the finished bead. Disk type beads were found throughout the sequence, were morphologically similar from layer to layer, and were most frequent in Layers II III, Levels 910 (se e Table 3 18 and Table 3 19). Two tools were recovered, including a bivalve shell scraper ( Figure 3 35 A ) and a coral file, or abrader (Figure 335 B ). Both were recovered from Layer II, Level 5. Materials recovered from this depth dated to sometime between 776 and 1380 cal a BP (Table 3 14 and Table 3 15). This coincides with the early portion of the Latte Period. Other non decorative items included two small shell fishhook fragments ( Figure 3 36). These were recovered from TU 3, Layer I, Level 3 at approximately 30 40 cmbd and date to the early part of the Latte Period between 1012 and 1380 cal a BP (Table 3 14 and Table 3 15). Faunal m aterials . Excavations of TU 3 and TU 4 also produced a relatively rich faunal assemblage that included vertebrate bone and a wide variety of invertebrate shell remains. These included species of ter restrial invertebrates and numerous species of marine invertebrates. Terrestrial invertebrates included a shell from a land snail and fragments of a large species of land crab ( Birgus latro ). Marine invertebrates included corals, gastropods, bivalves, and crabs. Vertebrate remains comprised mostly fish; however, bird and mammal bone was also recovered. Mammal bone consisted of pig, rat, and fruit bat.
132 Invertebrates . Generally, the same invertebrate species identified from the Bapot 1 site excavations were present in the assemblage of coral and shell recovered from TU 3 and TU 4 from the Ritidian Grotto site on Guam. Two pieces of staghorn coral (Acropora formosa) were recovered. These included a coral tool with ground surfaces ( Figure 335 B ) and an unworked branch. The worked piece was recovered from TU 3, Layer II, Level 5, and the unworked piece was recovered from TU 3, Layer II, Level 9. The unworked piece was not water worn suggesting that it was collected for use as a file, or abrading tool. This branching, tree like coral species, Acropora formosa, is relatively common throughout the tropical Pacific and forms dominant colonies in large shallow port ions of lagoons (Carpenter and Niem 1998:110). The two coral specimens accounted for a very small percentage of the total assemblage, consisting of <1% ( Table 3 20). A total NISP of 2,538 gastropod shells derived from 27 taxa was recovered from TU 3 and TU 4. The total assemblage of gastropod shell weighed 3.79 kg and accounted for 54% NISP of the total invertebrates and 61% of the total by weight ( Table 3 20). All identified species from TU 3 and TU 4 were also present in the Bapot 1, Block A gastropod assemblage (Table 3 4 and Table 3 21). These included spec ies commonly found in nearshore marine habitats. A total of nine specimens of Achatina fulica shell weighing 15.4 g were recovered. These consisted of three specimens weighing 3.7 g from TU 3 and six specimens weighing 11.7 g from TU 4. All of these were o nly present within the first 10 cm of Level 1, Layer I, in TU 3 and TU 4. Gastropod shells recovered from TU 3 and TU 4 were fragmented and 54% of the assemblage was only identifiable as gastropod. Of the gastropod species that were identified to the fami ly, genus, or species level, Turbo spp., or turban shell, and Pythia scarabaeus , or salt
133 marsh snails, were the most common. Turbo spp. shell accounted for 12% of the total NISP and 36% of the total weight, and Pythia shell accounted for 20% of the total N ISP and 14% of the total weight of the assemblage. The remainder of the identified gastropods accounted for minor percentages of the total NISP, ranging from >1 to 2% of the total NISP. Some of the larger shells accounted for greater percentages of the wei ght even though they accounted for low percentages of the total NISP. These large shells included specimens of Trochus spp., which accounted for 4% of the total weight, and Cerithium nodulosum , or the giant knobbed cerith shell, which accounted for 7% of t he total weight of gastropod shells in the assemblage ( Table 3 22). A total NISP of 927 identifiable bivalve shells weighing 2.06 kg were recovered fro m TU 3 and TU 4 at the Ritidian Grotto site. These were derived from 15 species ( Table 3 23) and accounted for 20% of the total NISP and 33% of the tot al weight of the faunal assemblage ( Table 324). Of the bivalve species identifiable to the family or genus level, shells from the genus Modiolus spp., family Tellinidae, and genus Tridacna spp. were most common. Modiolus spp., or horse mussels, accounted for 6% of the total NISP and 1% of the total by weight, Tellin shells accounted for 3% NISP and 2% by weight, and Tridacna, or giant clam shells, accounted for 12%t of the total NISP of bivalve shells and 60% of the total by weight. The rest of the identified species collectively accounted for 4% of the total NISP and approximately 12% of the total by weight ( Table 3 24). Individually these various species accounted for small percentages of the total NISP and total weight of the bivalve assemblage. Vertebrates . Results of the laboratory analyses of vertebrate remains are presented in the following section. Excavations of TU 3 and TU 4 produced a relatively rich assemblage of vertebrate remains. This included bones from fish, birds, and mammals.
134 A total of 4,673 specimens of fish bone weighing 173.8 g were recovered from TU -3 and TU 4. This included a total of 1,798 specimens from TU 3 weighing 63.2 g and a total NISP of 2,875 weighing 110.6 g from TU 4. The condition of the fish bone recovered from the Ritidian Grotto site was exceedingly poor. M ost fish bone specimens consisted of very small cranial fragments, vertebrae, and spines. Approximately 5% of the fish bone assemblage was burned. These factors severely inhibited species level identifications and 99% NISP and 95% by weight of the fish bone was categorized as Class Osteichthyes ( Table 3 25). Twenty -seven fish bone specimens were identified to the family level. These included one premax illa from the family Serranidae, or sea bass, and 26 specimens of parrotfish (Scaridae) bone. Remains from the Serranid accounted for <1% of the total NISP and <1% of the total weight. Parrotfish accounted for 1% of the total fish bone by NISP and 5% by we ight (Table 3 25). The parrotfish remains from the Ritidian Grotto site also consisted of lateral skull elements from parrotfish, including robust pha ryngeal teeth, dentaries, and premaxillas. The NISP of Ritidian Grotto fish bone was much larger than the Bapot 1 site assemblage; however, the larger bone assemblage from the Ritidian Grotto site was derived from a much smaller excavation sample. This sug gests that different resource exploitation strategies were practiced at the two sites. Also, limiting environmental factors such as the lack of fresh water, lagoons, and dangerous surf and ocean currents at Ritidian may have led to an overall emphasis on f ish rather than a more diverse pattern of exploitation that more equitably would have included marine invertebrates and birds. Other vertebrate remains included bones from birds and mammals. After fish, bird bones were the second most common class of ident ified vertebrate. A cursory examination of the bird bone assemblage was conducted by the author. These preliminary identifications of Ritidian
135 Grotto site bird bones relied upon consideration of distinctive morphological characteristics and microstructural properties unique to bird bone. These include various adaptations for flight, including the fusing and enlargement of certain elements to form a framework for muscle attachments (Reitz and Wing 2008:59 60) and the pneumatization, or replacement, of marrow with air. This adaptive process has resulted in the one of the most characteristic attributes of bird bone, namely, hollow thin-walled bones (Higgins 1999:1450). A total NISP of 195 bird bones weighing 18.8 g was recovered from TU -3 and TU 4 (Table 3 26). This included 172 specimens of hollow, pneumaticized midshaft bone fragments, two terminal phalanx fragments, an ulna fragment, and a tarsometatarsu s. Similar to the fish bone assemblage from TU 3 and TU 4, most of the bird bones were recovered from early in the stratigraphic sequence at the bottom of Layer II, 8090 cmbd and Layer III, 100110 cmbd ( Figure 337). Materials from these depths dated between 27783473 cal a BP (Table 3 14 and Table 3 15) and coincide with the Late Unai Phase (1600â€“2500 cal a BP) and the Middle Unai Phase (2500â€“3000 cal a BP). A total of 27 mammal bones weighing 18.6 g were identified. This included one bone from Sus scrofa , or pig, seven Rattus spp., or rat bones, 12 Pteropid, or fruit bat bones, and seven unidentified mammal bones. Rat was present in the later (Layer I, Level 1, 10 cmbd) and the earlier (Layer III, Level 10, 100 cmbd) portion of the stratigraphic sequence. Materials from Layer I, Level 2, date to before 1012 cal a BP (Table 3 15). Although this date falls within the Latte Period, the presence of rat bone in Layer I, within 10 cm from the surface, is likely the result of more recent noncultural processes such as burrowing. Samples collected from 100 cmbd dated to between 1559 and 3473 cal a BP (Table 3 14). This date range covers the entire Pre Latte Period (1000â€“3500 cal a BP). According to the debate regarding rat bones in Micronesia discussed in the Bapot 1 site faunal materials secti on, the presence of rat bone in
136 Layer III at 100 cmbd is significant. This is because rat bones have not been recovered from a secure context in the Marianas that date to the Pre-Latte Period (Steadman 1999). These rat bones have not been identified to the species level and considering the disturbanceprone nature of rockshelter deposits and the burrowing habits of rats, the probability that these rat remains date to the Pre-Latte Period is low. Rather, these remains were likely introduced into the deposit more recently through an animal burrow. A single bone from Sus scrofa , or pig was identified from Layer I, Level 2 (Table 3 27). This is between 10 20 cmbd (Figure 329 and Figure 330). Despite extensive archaeological excavations on Guam and the Northern Marianas, pig bone has not been identified in prehistoric contexts in this archipelago (Wickler 2002). The presence of this bone within the first 20 cmbd is also most likely the result of histo ric or modern disturbance. Twelve fruit bat bones were recovered from Layer II at 70 90 cmbd (Table 3 27). Materials from 70 90 cmbd dated to approximately cal B.P. 31412778. This dates to the Middle Unai Phase (cal B.P. 30002500) of the Pre -Latte period. As stated previously in the faunal mat erials section for the Bapot 1 site, fruit bats were likely collected throughout the Prehistoric Period (Steadman 2006:503) and have been nearly extirpated, especially on Guam. Human r emains . A total of 19 specimens of bone identified as human were recovered from TU 3 and TU 4 (Table 3 28). This included nine specimens from TU 3 and 10 from TU 4. The bone was highly fragmented and spread throughout the stratigraphic sequence and no evidence of a burial feature was encountered during excavations. Elements consisted of long bone fragments, rib fragments, a tibia fragment, cranial vault fragments, and cortical bone fragments. Whole elements included a cervical vertebrae and an incisor. The presence of adult bone fragments and a subadult rib fragment suggest that there are remains from at least two
137 individuals in the assemblage. Because some of the human bone specimens had been burned and no burial fea tures were found, and the bones were highly fragmented and scattered, the remains may have been burned and then deposited within the rockshelter. Their presence between 40 and 90 cmbd suggests that this may have taken place on multiple occasions during the early prehistoric occupation of the site. Marine r esource analysis. The Ritidian Grotto site assemblage was dominated by marine fauna, similar to the Bapot 1 assemblage. In contrast, however, fish accounted for 49% of the total NISP and only 3% of the total assemblage by weight. This is because most of the fish bones consisted of small unidentifiable spines, vertebra, and fragments. The invertebrate assemblage was dominated by gastropods, which accounted for 26% of the total NISP and 59% of the total by w eight. Bivalves were the second most common class in the assemblage and accounted for 10% of the total NISP and 32% of the total by weight. Indeterminate shell accounted for just 13% NISP and 4% of the total by weight. All of the other invertebrate categor ies each accounted for <1% of the total NISP and weight of the assemblage. This included beach shell and coconut crab ( Birgus latro ) (Table 3 29). Ot her vertebrate remains, including bird and mammal, accounted for very small percentages of the total assemblage. The 195 bird bones identified accounted for only 2% of the total NISP and the 27 mammal bones accounted for <1% of the total NISP, and both accounted for <1% of the total by weight (Table 3 29 ). Fish . When frequencies of fish bone NISP were plotted by depth (cmbd), fish bone appears to be more common during the earliest portions of the stratigraphic sequence and was most frequent between 90100 cmbd (Figure 338 A ). Materials recovered from these depths dated to between 1500 and 1700 cal a BP and coincides with the Late Unai Phase (1600 â€“2500
138 cal a BP). This is also the case when fish bone NISP samples were plotted by the proposed dating sequence based on the deposition model created using OxCal v4.2.2 (Bronk Ramsey 2013) ( Figure 3 32). The vast majority of fish bone was recovered from samples dating to 1500â€“ 1700 cal a BP ( Figure 3 38 B ). Marine Invertebrates . The marine shell assemblage recovered from TU 3 and TU -4 was relatively small when compared to the Bapot 1 site assemblage. Ther e were no particularly abundant species represented by well preserved shells that could be subjected to morphometric or age based analyses like the large numbers of Anadara cf. antiquata or Gafrarium pectinatum that were so prevalent in samples from the Bapot 1 site. The dominant species of gastropods from TU 3 and TU 4 included Turbo spp. (n=277) and Pythia scarabaeus (n=518). Turban shell remains consisted of extremely fragmented shells and numerous opercula. Although the opercula were large and well pres erved, it was difficult to identify these structures accurately to the genus and species level. Pythia scarabaeus accounted for 20% of the total NISP (Table 3 -22) is commonly found in coastal environments throughout the tropical western Pacific (Poutiers 1998:645). Shells of this species were thin, brittle, and too fragmented to conduct morphometric analyses. The most abundant bivalve species included Modiolus spp. (n=56), tellin (Tellinidae) (n=27), and Tridacna spp. shells (n=113). All of the other identified species of gastropods and bivalves each accounted for minor portions of the assemblage and were generally fragmented and could not be accurately identified to the family, genus, or species level. Therefore, shells assigned to Class Gastropoda and to Class Bivalvia accounted for most NISP in the assemblage (Table 3 20).
139 To determine if changes in exploitation patterns, preference, or habitat use were apparent within the occupational sequence, comparisons of the frequencies of identified species and species diversity between stratigraphic excavati on levels and the proposed radiocarbon date sequence were conducted. First, the number of identified species and the calculated Shannon H ) were plotted by depth (cmbd). Second, the number of identified species and the calculated Shann onWiener index values were plotted by date, as was done with shell data from the Bapot 1 site. Aside from an abrupt increase in the frequencies of species for the total invertebrate assemblage in samples from 110 100 cmbd, there was a general decrease in the frequencies of species in samples from shallower excavation levels between 10 cmbd through 80 90 cmbd ( Figure 3 39 A ). The highest frequencies of species occurred in samples from 80 90 cmbd, with a punctuated increase in samples from 40 cmbd. Frequencies of gastropod species followed the same pattern, with high frequencies in samples from 80 90 cmbd and an increase in species in the sample from 40 c mbd ( Figure 3 39 A ). Frequencies of identified bivalve species were also similar to frequencies of identified gastropod species. The number of bivalve species increased abruptly in samples from 110100 cmbd and fluctuated between five and eight species in samples from 10030 cmbd. Bivalve species abruptly decreased to two in samples from 20 10 cmbd and the highest number of identified bivalve species occurred in samples from 40 cmbd ( Figure 3 39 A ). Shell data from the excavation layers were combined according to the proposed dating sequence ( Table 3 15). This enabled species frequencies and ShannonWiener diversity values to be examined by date before present. When plotted by date, following the proposed dating sequence, invertebrate species frequencies showed an abrupt increase in samples from 1500 â€“
140 1700 to 1800 cal a BP ( Figure 339 B ). Species fr equencies then decreased in samples from 15001700 cal a BP to samples from 1000 cal a BP. There was also a marked increase in the number of identified species in the sample from <1000 cal a BP. This pattern was also similar for gastropod and bivalve speci es ( Figure 3 39 B ). Shannon-Wiener index values plotted by depth followed a similar pattern as species frequency plots. The highest diversity values fo r the total invertebrate assemblage and gastropod species were in samples from 100 70 cmbd ( Figure 3 40 A ). For bivalves, the highest values were in sam ples from 7030 cmbd. Diversity values for all invertebrates decreased in samples from 7060 cmbd and oscillated around a mean of 1.75 in samples from 6010 cmbd. Values for gastropods oscillated around a mean of 1.25 in samples from 5010 cmbd and values for bivalves oscillated around a mean of 0.75 in samples from 6010 cmbd. Although diversity values for total invertebrates and gastropods follow a similar pattern when plotted according to depth, values for bivalves are more variable. When plotted by dept h, diversity values for bivalves increased from 0.02 in samples from 110 cmbd to the highest diversity value of 1.17, in the sample from 70 cmbd. The diversity value for bivalves in the sample from 60 cmbd abruptly decreased and then increased in samples f rom between 50 30 cmbd, to a diversity value of 1.10. Diversity values decreased again in samples from 20 10 cmbd to around 0.5 ( Figure 340 A ). When S hannon-Wiener index values for combined samples were plotted by date before present, values increased from the lowest values in samples dating to 1800 cal a BP to the highest values in samples dating to 1400 cal a BP for all invertebrates, gastropods, and bivalves. Between 1700 and 1500 cal a BP and at 1000 cal a BP, diversity values from the total invertebrate and gastropod samples decreased and then slightly increased in samples younger than 1000 cal a BP. Diversity index values from bivalve samples datin g between 1200 and 1400
141 cal a BP decreased and then increased to close to the diversity value of the sample dating to 1400 cal a BP, the highest diversity value for the assemblage of identified bivalve species ( Figure 3 40B ). Although diversity values for all invertebrates, gastropods, and bivalves oscillated between higher and lower values throughout the sequence, like values for species from the Bapot 1 site assemblage, diversity values for gastropods were always higher than values for bivalves in the Ritidian Grotto site assemblage. This is because the number of species of bivalves in the Ritidian Grotto site assemblage is relatively small and less di verse than the Bapot 1 bivalve assemblage. Summary and Discussion of Ritidian Grotto Results Excavations of TU 2 through TU 4 at the Ritidian Grotto site identified four stratigraphic layers. The first three of these stratigraphic layers (Layers I III) consisted of cultural sediments and contained Latte and Pre-Latte period materials. Layer IV consisted of pale brown sand and contained no cultural materials. Radiocarbon dated materials recovered from similar stratigraphic layers in TU 1 suggests that prehis toric occupation of the Ritidian Grotto site extended from approximately 989 29 to 3140 40 cal a BP. The Ritidian Grotto site is a small component of the larger Ritidian area and likely represents a temporary habitation site that may have been periodic ally used as a fishing or resource procurement camp. The presence of fragmented and burnt human bone and the presence of pictographs in the small cave suggest that other activities also took place at the site. This site was selected for the current study because it yielded comparable types of artifacts and faunal materials as the Bapot 1 site on Saipan, it has overlapping occupational sequences with evidence of late Pre-Latte Period and Latte Period use, and because the Ritidian Grotto and Bapot 1 sites li kely represent divergent residence patterns and marine resource exploitation strategies. This is because the Ritidian Grotto site has a relatively small fringing reef and is exposed to extremely rough ocean currents and waves. These dangerous ocean conditi ons
142 may have limited marine resource exploitation activities to certain times of year when seas were relatively calm, or limited the available resources that could be collected. Furthermore, because Ritidian is located on the northern limestone portion of Guam, it lacks any significant source of fresh water. This may have affected long term settlement patterns and the exploitation of both terrestrial and marine resources. A relatively small number of artifacts were recovered during excavation of the Ritidian Grotto site. Artifact types included pottery, worked shell, and a piece of worked coral. Pottery consisted of 427 sherds. This included 205 sherds comprising 18 rim sherds and 187 body sherds from TU 3, and 222 sherds from TU 4, including 18 rim sherds a nd 204 body sherds. Excavations of TU 3 and TU 4 produced 28 worked shell and coral artifacts. TU 3 yielded 10 artifacts and TU 4 produced 18 artifacts. Shell artifacts included beads, scrapers, fishhook fragments, unspecified worked shell items, and a worked coral file or abrading tool. Beads were the most common artifact type after pottery and were present in all stratigraphic layers. Shell tools included a bivalve shell scraper and a coral file. Both were recovered from samples dating to between 776 and 1380 cal a BP. Fishhook fragments were also recovered and dated to sometime between 1012 and 1380 cal a BP. Faunal remains recovered from excavations of TU 3 and TU 4 included both terrestrial and marine species of invertebrates and vertebrates. Terrestri al invertebrates included land snail and crab shell remains and terrestrial vertebrates included pig, rat, and fruit bat bones. Marine invertebrate remains included coral, gastropods, bivalves, and crab. Aside from two species of terrestrial snail and frag ments of land crab carapace, all of the marine invertebrates included species that inhabit nearshore habitats ranging from the splash, or tidal zone, to the edge of the reef (Table 3 21 and Table 3 23).
143 Human remains, consisting of 19 specimens of bone, were also recovered from TU 3 and TU 4. No burial features were identified and bones were highly fragmented and dispersed between 40 and 90 cmbd. Collectively, the human bone assemblage represents the remains of at least two individuals, including one adult and one subadult. Some of the human bones were burned, suggesting the individuals were cremated and then deposited within the rockshelt er. Marine fauna dominates the Ritidian Grotto site assemblage. Fish bone accounts for 49%, marine gastropods account for 26%, and bivalves account for 10% of the total NISP. Bird bone and fish bone occurred more frequently during the earliest portions of the stratigraphic sequence, especially in samples from 90 100 cmbd, which dated to 15001700 cal a BP. These dates coincide with the Late Unai Phase of the Pre Latte Period. The low percentages of identifiable fish bones were also reported from previous w ork at Ritidian. Archaeological excavations conducted in 1989 resulted in the recovery of 30 elements identified to the family level from a total of 1,017 fish bones. These 30 identifiable elements were derived from six families of reef fish. Parrotfish bones were the most common, accounting for almost 50% of the total (Amesbury and Hunter-Anderson 2003:46). When compared to the Bapot 1 site assemblage, the amount of marine shell recovered from TU 3 and TU 4 was relatively small. Furthermore, none of the id entified species recovered from the Ritidian Grotto site was abundant enough to be statistically analyzed. Dominant species of gastropods from TU 3 and TU 4 include Turbo spp. (n=277) and Pythia scarabaeus (n=518). Many of the gastropod species can be easi ly observed on the small fringing reef and reef flat that lies north of the Ritidian Grotto site. Only two terrestrial gastropod shells were identified, which included Pythia scarabaeus and Achatina fulica . Pythia scarabaeus , or the saltmarsh snail, is wil dly distributed throughout the Pacific (Cernohorsky 1972:212), can live amphibiously in
144 coastal swamps, forests, and in littoral environments, and is currently collected by some villages in Indonesia for subsistence (Carpenter and Niem 1998:641). As previously mentioned, the recently introduced Giant African land snail ( Achatina fulica ) is one of the worst invasive species in the Marianas (Cowie 2000). Its presence within archaeological contexts indicates that some level of disturbance has taken place. Dom inant bivalve species included Modiolus spp. (n=56), tellin (Tellinidae) shells (n=27), and Tridacna spp. shells (n=113). Many species of bivalves identified in the Ritidian Grotto shell assemblage were the same as those identified in the Bapot 1 site assemblage, however, TU 3 and TU 4 was less diverse. The Bapot 1 site assemblage included 20 bivalve species ( Table 3 6), whereas the Ritidian Grotto site yielded 15 (Table 3 23). This may be related to differences in nearshore marine habitats or differences in sampling volume between the sites. The bivalve shells from the Ritidian Grotto site were also more fragmented, with a relatively high proportion of shell that could not be identified beyond class Bivalvia. This included 74% of the total NISP and 24% of the total assemblage by weight. The various bi valve species identified from the Ritidian Grotto site appear to reflect the local marine environment, which consists of a relatively narrow fringing reef with a coarse sand beach. There is no lagoon and with the lack of streams or fresh surface water at R itidian, no silty or muddy substrates. Because of these conditions, species adapted to living on hard substrates and in coarse littoral sands dominate the Ritidian Grotto assemblage ( Table 3 23). All of the other identified species of gastropods and bivalves were relatively minor contributors by NISP or weight to the overall assemblage. In general, most of the invertebrate shell was fragmented and difficult to identify. Comparisons of species frequencies and species diversity by depth (cmbd) and date revealed fluctuations between lower and higher species frequencies and diversity values
145 throughout the stratigraphic and chronological sequences. This is th e case for all invertebrate species, gastropods, and bivalves. The lack of clear trends in increasing or decreasing species richness or diversity values is likely related to the function of the Ritidian Grotto site. With the dominance marine resource resid ues in the faunal assemblage, the dominance of shell artifacts, and the general lack of surface and subsurface features at the site suggest the rockshelter was intermittently occupied, and that fish and shell fish were the primary targets of exploitation during site occupation. High species frequencies and diversity index values may simply reflect more intensive periods of occupation or use.
146 Table 3 1. Descriptions of soils from stratigraphic layers sectioned in the north wall of Block A, Bapot 1, Block A. Soils descriptions were derived from the authorâ€™s field notes and Clark et al. (2010:25 26). Layer Soil Description Material Content Ia Very dark brown (10YR 2/2) hardpacked silty calcareous soil with tree roots and fragments of eroded limestone. Latte Period pottery with medium thick plain sherds and abruptly thickened rims and small quantities of marine shell and bone. Modern materials included bottle glass, WWII shrapnel, and a few pieces of vulcanized rubber. Ib Dark yellowish brown (10YR 4/4) loose sandy soil. Latte style pottery with a few eroded thin red pot sherds possibly representing older ceramics that have been mixed with late prehistoric ceramics. IV Yellowish brown (10YR 5/4) loose silty calcareous sand. Upper levels of layer produced a lot of thick walled ceramics with red slip from flat based trays. Marianas Redware increased, including decorated sherds. Adzes, flakes, small Conus sp. shell rings and ground Cypraea sp. beads also present . V Yellowish brown (10YR 5/6) indurated coarse sand in upper portions of the layer and coarse sand with pockets of indurated sand in the lower portions of the layer. Thin red slipped pottery from small to medium diameter carinated jars, in situ base sherds, shell artifacts, and a large sub lenticular volcanic adze (20 cm long by 10 cm wide). VI Brown (7.5YR 4/3) silty calcareous sand with areas of indurated sand. Large quantities of thin red slipped pottery, shell artifacts, a cash of three st one adzes made of an altered tuff, and faunal remains including bird and fish bone and marine shell. VII Yellowish red (5YR 4/6) indurated silty sand. Contained similar artifacts as those found in Layer VI: red slipped pottery, shell artifacts, stone arti facts made of a variety of materials, and faunal remains with abundant bone from a rail ( Gallirallus cf. philippensis ). VIII Very pale brown (10YR 7/4) coarse indurated calcareous sand. No cultural materials present.
147 Table 3 2. Bapot 1, Block A radiocarbon dates (*Denotes r ejected date). Lab. No. CRA 13C (0.2) Cal BP Date (68.2% probability) Sample Unit, Depth (cmbd) 1386 30 22.6 0.2 ?Coconut shell 1581 35 23.4 0.2 Nut shell cf. Cocos nucifera 2043 30 24.3 0.2 Unidentified charcoal 2189 30 24.5 0.2 Unidentified charcoal 2168 32 27.9 0.2 Unidentified charcoal 2175 30 25.7 0.2 Unidentified charcoal 2907 32 25.1 0.2 Unidentified charcoal 2866 32 25.3 0.2 Unidentified charcoal 2922 30 24.6 0.2 Unidentified charcoal 2904 30 21.8 0.2 Nut Shell Unit 2, 200210 2910 30 25.1 0.2 Unidentified charcoal 2900 30 25.5 0.2 Unidentified charcoal 3192 30 2.1 0.2 Cypraea tigris artifact (Herbivore) 3013 30 25.5 0.2 Unidentified charcoal 3182 30 0.6 0.2 Conus sp. artifact (Carnivore) 3010 30 28.1 0.2 Unidentified charcoal 2908 30 24.9 0.2 Unidentified charcoal 2386 30 25.9 0.2 Unidentified charcoal 3484 35 0.7 0.2 Anadara sp. Filter feeder
148 Table 3 3. Total counts and weights for invertebrate remains from Bapot 1, Block A. Taxa NISP NISP (%) Wt (g) Wt (%) Coral 2 <1 14.4 <1 Gastropod 9328 44 18,910.8 44 Bivalve 11128 53 22,970.1 53 Heterocentrotus cf. mammillatus (Sea Urchin) 172 1 180.6 <1 Beach shell 373 2 806.4 2 Indeterminate 48 0 103.5 <1 Totals 21,051 100 42,985.8 100 Table 3 4. List of identified gastropod species from Bapot 1, Block A including common names and habitat preferences. Taxa Common Name Habitat Preference (Abbot and Dance 2000; Lamprell and Whitehead 1992) Haliotis varia Variable abalone or common ear shell Rocky shores and coral reefs littoral to shallow subtidal depths among rocks and under stones. Cellana radiata orientalis Rayed wheel limpet On rocks in the intertidal zone. Patella flexuosa Star Limpet Rocky and coral rock bottoms, on stones or large shells in the subtidal and shallow intertidal areas. Patelloida striata Striate limpet On rocks in the intertidal zone. Trochus spp. Trochus snails Littoral and shallow sublittoral zones on rocks and coral reefs and eel grass. Trochus spp. Trochus snails Li ttoral and shallow sublittoral zones on rocks and coral reefs and eel grass. Astraea rhodastoma Rose mouth star shell Found under coral rocks or exposed on reef in intertidal zone. Astralium Star snails Found under coral rocks or exposed on reef in intertidal zone. Turbo argyrostomus Silver mouthed turban Coral reefs in lagoons Turbo setosus Rough turban Exposed are as on coral reefs, sublittoral zone in shallow water. Nerita plicata Plicate nerite Upper portions of the shoreline, in cracks and crevices of rock benches, moves up and down with the tide.
149 Table 3 4. Continued Taxa Common Name Habitat Preference Nerita polita Polished nerite Intertidal rocky and coral reefs. Littorinidae Periwinkles Intertidal areas and splash zones, to well above high tide mark. Cerithium nodulosum Giant knobbed cerith Sand and rubble on intertidal reef flats near outer edge of reefs. Lambis lambis Common spider conch Reef flats, rubble bottoms, and mangroves; associated with red algae on which it feeds. Strombus gibberulus gibbosus Gibbose conch Shallow waters on clear sand and sea grass beds, in sandy patches on reef flats, in sandy lagoons and sandy mud bottoms, intertidal to sublittoral to a depth of 20 m. Vasum ceramicum Ceram Vase On shallow reefs, intertidal to shallow littoral zones to 20 m. Vasum turbinellus Top vase Rocky bottoms and reef flats, intertidal to shallow subtidal waters. Conus abraeus Hebrew cone Reef dwelling in clean or muddy sand, under rocks and corals, mostly on intertidal benches and subtidal reef flats to 3 m. Puncticulis pulicarius pulicarius Flea bite cone Intertidal to more than 75 m d eep in deep sands away from limestone outcrops and growing corals in sand channels and patches on the reef flat and in bays. Conus sponsalis Sponsal cone Intertidal to 100 m deep. Strombus mutabulis mutabulis Mutable conch Sandy and rubble bottoms of cor al and rocky reefs, in clear or turbid waters, sheltered and exposed areas, and most common just below tide marks. Cypraea caputserpentis Serpentâ€™s head cowrie Common on coral reefs and rock benches exposed to wave action, in intertidal and shallow subtidal areas. Cypraea tigris Tiger cowrie Abundant on reef areas on sand, among rocks and coral, on branch coral, in tide pools near sea weeds, intertidal to sublittoral zones to 30 m.
150 Table 3 4. Continued Taxa Common Name Habitat Preference Cypraea annulus Gold ring cowrie Common in shallow tide pool in vegetation and under stones, common in all intertidal zones. Cypraea maculifera Reticulated cowrie In coral reefs, under slabs and rocks, in gullies, and holes of the algal crests. Cypraea moneta Money cowrie Common in the open, high in the intertidal zone to shallow portions of the subtidal zone. Cypraea scurra Jester cowrie Often deeply buried in dead coral rubble, reefs, and under coral heads, over the reef ledge, in shallow sublittoral waters to a depth of 10 m. Cymatium nicobaricum Nicobar hairy triton On sand and rock bottoms associated with coral reefs, in sublittoral zones to 100 m. most common between tide marks and shallow sublittoral zones. Cymatiidae Triton shells Rocky bottoms from the intertidal zone to depths of 100 m. Naticidae Moon shells Sand and mud dwellers. Tunnidae Tun shells Sandy bottoms and sea grasses. Bursa bufonia Warty frog shell Low tide levels, shallow sublittoral zones to 20 m. Thais armigera Belligerent rock shell Rocky shores and coral reefs, intertidal and shallow subtidal zones. Morula uva Grape drupe On rocks and coral, in shallow and deeper waters. Drupa ricinus Prickly spotted drupe On rocks and coral, in shallow and deeper waters. Drupa clathrata The clathrate drupe On rocks and coral, in shallow and deeper waters. Mitra stictica Miter shell Reef platforms under rocks and in crevices, intertidal, sublittoral, and shelf zones to 200 m. Vexillum lautum Costate miters (Vexillidae) Littora l and shallow subtidal zones, in rock crevices and coral.
151 Table 3 4. Continued Taxa Common Name Habitat Preference Nassariidae Nassa mud snails Sandy or muddy soft bottoms in marine and brackish waters, in intertidal and sublittoral zones. Oliva annulata Amethyst olive Burrows in sandy bottoms in the shallow subtidal zone. Harpidae Harp shells Burrows in sandy bottoms from low tide areas to deep shelf zone. Land Snails Pythia scarabaeus Common Pythia, salt marsh snail Coastal environments especially near forests Achatina fulica African land snail Diverse habitats including agricultural areas, coastal zones, natural and planted forests, scrub and shrublands, riparian areas, wetlands, and urban areas Table 3 5. Summary of total NISP and % NISP for gastropods from Bapot 1, Block A. Taxa Total NISP NISP (%) Polyplacophora 4 <1 Haliotis varia 58 1 Cellana radiata orientalis 3270 35 Patella flexuosa 233 2 Patellidae 261 3 Patelloida striata 2 <1 Trochus spp. 173 2 Astrea rhodastoma 9 <1 Astralium 10 <1 Turbo argyrostomus 51 1 Turbo setosus 119 1 Turbo spp. 1646 18 Nerita plicata 10 <1 Nerita polita 25 <1 Nerita sp. 23 <1 Littorinidae 1 <1 Cerithium tenuifilosum 8 <1
152 Table 3 5. Continued Taxa Total NISP NISP (%) Cerithium nodulosum 1 <1 Cerithium spp. 10 <1 Cerithiidae 6 <1 Lambis lambis 2 <1 Lambis sp. 11 <1 Strombus luhuanus 22 <1 Strombus gibberulus gibbosus 116 1 Strombus mutabulis mutabulis 805 9 Strombus spp. 159 2 Cypraea caputserpentis 17 <1 Cypraea tigris 1 <1 Cypraea annulus 10 <1 Cypraea maculifera 135 1 Cypraea moneta 30 <1 Cypraea spp. 181 2 Cymatium nicobaricum 3 <1 Cymatium spp. 12 <1 Bursa sp. 4 <1 Thais armigera 4 <1 Thais sp. 4 <1 Morula uva 5 <1 Drupa morum 22 <1 Drupa ricinus 157 2 Drupa spp. 63 1 Mitridae 9 <1 Vexillum lautum 1 <1 Nassarias sp. 3 <1 Vasum ceramicum 4 <1 Vasum turbinellus 8 <1 Vasum sp. 1 <1 Oliva annulata 2 <1 Harpa major 2 <1
153 Table 3 5. Continued Taxa Total NISP NISP (%) Conus abraeus 14 <1 Puncticulis pulicarius pulicarius 2 <1 Conus sponsalis 2 <1 Conus spp. 146 2 Pythia scarabaeus 22 <1 Gastropoda (unidentified) 1428 15 Totals 9328 100 Table 3 6. List of identified bivalve species with common names and habitat preferences from Bapot 1, Block A. Taxa Common Name Habitat Preference (Abbot and Dance 2000; Lamprell and Whitehead 1992) Pinctada sp. Pearl oysters Attached to various substrates including rocks, shells, pebbles and plants, in warm shallow waters, sometimes in dense colonies. Mytilidae Sea mussels Attached to hard substrates, sometimes nestled together, also coral or rock borers. Modiolus sp . (possibly Modiolus auriculatus) Eared horse mussel Attached to rocks and in crevices. Isognomon sp. Tree oysters Attached to hard substrates in tropical shallow waters often in dense colonies. Spondylus sp. Thorny oysters Strongly cemented to hard substrates at right valve in coralline areas. Codakia tigerina Pacific tiger lucine Buried in sandy bottoms in coral reef areas, in shallow sublittoral zones to 20 m. Codakia punctata Punctate lucine Buried in sand between shale blocks and in coral reef areas in sublittoral zones to 20 m. Carditidae Carditas Attached to substrate in shallow waters. Fragum fragum White strawberry cockle In littoral to sublittoral sand to 20 m. Tellina scobinata Rasp tellin In coarse sand and gravel bottoms. Quidnipagus palatam Palate tellin In coarse sandy bottoms. Asaphis violascens Pacific asiphis Deeply buried in sand to coarse gravel in littoral and sublittoral zones to 20 m. Gafrarium pectinatum Comb Venus Intertidal and shallow sublittoral zones to 20 m.
154 Table 3 6. Continued Taxa Common Name Habitat Preference Gafrarium tumidum Tumid Venus Intertidal and sublittoral zones to 30 m. Pitar prora Prora Venus shell Moats and lagoons up to 25 m (Paulay 1987:19). Periglypta puerpera Maidens purse shell or youthful Venus Coarse sand bottoms in intertidal and sublittoral zones to 25 m. Tridacna maxima Elongate giant clam On reefs partially embedded in coral, in littoral and shallow waters to 20 m. Atactodea striata Striate beach clam Abundant in sandy beaches in the intertidal zone. Arca avellana Hazelnut ark Basally attached to rocks, corals, or under boulders on sand in low tide pools to depths of 80 m. Anadara cf. antiquata Antique ark On muddy bottoms, in intertidal and sublittoral zones to a dept h of 25 m. Table 3 7. Summary of total NISP and %NISP for bivalve shells from Bapot 1, Block A. Taxa NISP NISP (%) Mytilidae sp. 2 <1 Modiolus spp. 17 <1 Isognomon spp. 10 <1 Spondylus spp. 1 <1 Limidae 1 <1 Codakia tigerina 1 <1 Codakia punctata 4 <1 Carditidae 2 <1 Fragum fragum 4 <1 Fimbria spp. 3 <1 Tellina scobinata 11 <1 Tellinadae 1957 18 Quidnipagus palatam 118 1 Asaphis violascens 10 <1 Gafrarium pectinatum 3309 30 Dorisca sp. 1 <1 Gafrarium tumidum 75 1 Gafrarium spp. 506 5
155 Table 3 7. Continued Taxa NISP NISP (%) Pitar prora 1 <1 Periglypta puerpera 1 <1 Tridacna maxima 4 <1 Tridacna spp. 48 <1 Atactodea striata 15 <1 Anadara cf. antiquata 2751 25 Arca avellana 1 <1 Bivalvia (unidentified) 2273 20 Totals 11,128 100 Table 3 8. Totals for identified fish bone from Bapot 1, Block A. Taxa Common name NISP NISP (%) Wt (g) Wt (%) Chondrichthyes Cartilaginous fishes 1 <1 0.1 <1 Scaridae Parrotfishes 32 5 29.8 23 Carangidae Jacks 1 <1 0.1 <1 Balistidae Triggerfishes 1 <1 0.4 <1 Acanthuridae Surgeonfishes 3 <1 0.5 <1 Osteichthyes Bony fishes 797 95 98.0 76 Total 835 100 128.9 100 Table 3 9. Summary table of identified reptile bone from Bapot 1, Block A. Taxa Unit Depth (cmbd) NISP Gekkonidae Gecko 7 1 Gekkonidae Gecko 3 1 Varanus cf. indicus Mangrove monitor lizard 7 6 Gekkonidae Gecko 5 1 Gekkonidae Gecko 3 1 Table 3 10. Summary table of identified bird bone from Bapot 1, Block A. Taxa Common name NISP NISP (%) Seabirds Nesofregetta sp. Storm petral 1 <1 Puffinus cf. carneipes Flesh footed shearwater 1 <1 Anous sp. Terns 1 <1
156 Table 3 10. Continued Taxa Common name NISP NISP (%) Landbirds Wading birds 1 <1 Gallirallus sp. cf. philippensis Flightless rail 73 34 Porzana sp. Crake 1 <1 Acrocephalus sp. Warbler 2 1 Zosterops sp. White eye 1 <1 Aplonis spp. Starling 8 4 Meliphagid sp. Honey eater 3 1 Myzomela sp. Honey eater 1 <1 Passerines Sparrows and small birds 3 1 Aves Indeterminate 119 55 Total 215 100 Table 3 11. Summary table of identified mammal bone from Bapot 1, Block A. Taxa Unit # Depth (cmbd) Count Rattus sp. Rat 8 1 Rattus sp. Rat 3 1 Pteropid Fruit bat 1 1 Pteropid Fruit bat 7 1 Table 3 12. Summary table of all analyzed faunal remains recovered from Bapot 1, Block A (*denotes unavailable data). Taxa NISP NISP (%) Wt (g) Wt (%) Invertebrates Coral 2 <1 14.4 <1 Gastropods 9328 42 18,910.8 44 Bivalves 11128 50 22,970.1 53 Sea Urchin ( mammillatus 172 1 180.6 <1 Beach Shell 373 2 806.4 2 Indeterminate Shell 48 <1 103.5 <1 Vertebrates Fish 836 4 130.3 <1 Reptiles 10 <1 . -* <1 Birds 215 1 . -* <1 Mammals 4 <1 . -* <1 Total 22116 100 43116.1 100
157 Table 3 13. Soil descriptions for TU 2, TU 3, and TU 4, Ritidian Grotto. Layer Soil Description Material Content I Dark gray (10YR 4/1) loose very fine silty calcareous sand with limestone pebbles and cobbles and sparse banyan tree roots. Latte Period pottery, African land snail shell, bone, and marine shell. II Light brown (7.5YR 6/3) loose very fine sand with sparse roots and limestone pebbles and cobbles Latte and pre Latte Period potsherds, shell, non human a nd disturbed human bone, lithics, and worked marine shell. III Very pale brown (10YR 8/3) loose, fine sandy soil. Pre Latte period pot sherds, shell, nonhuman bone, burnt human bone fragments. IV Very pale brown (10YR 8/3) very loose fine sand. No cultural materials present Table 3 14. Ritidian Grotto, TU 1 radiocarbon dates (*Denotes r ejected date). Lab. No. CRA (BP) 13C (0.2) Cal a BP (68.2% probability) Sample Depth (cmbd) 1250 40 +1.2 Organic sediment 700 40 24.8 Cellana sp. Shell 1350 40 +1.6 Organic sediment 1000 40 24.8 Organic sediment 1370 40 26.2 Organic sediment 2820 40 25.4 Charcoal, Coconut Shell 1790 40 24.1 Unidentified charcoal from organic lens 100 3140 40 +1.5 Anadara sp. Shell 2970 40 2.0 Carbonate concretion 550 40 +3.4 Strombus sp. Shell 5810 40 +3.9 Cellana sp.
158 Table 3 14. Continued Lab. No. CRA (BP) 13C (0.2) Cal a BP (68.2% probability) Sample Depth (cmbd) 1550 40 27.9 Unidentified charcoal 3500 40 3.2 Foraminifera Table 3 15. Proposed sequence for TU 3 and TU 4, Ritidian Grotto. *=Layer IV was natural undisturbed sand and contained no cultural materials. below surface (cm) Modelled Dates (Cal a datum (cm) Proposed Dates Cal a 989+ 29 1000 1181+29 1200 1370+40 1400 100 1790+40 1800 2810+40 Layer IV* 2970+40 Layer IV* 3140+40 Layer IV* Table 3 16. Summary pottery table for TU 3, Ritidian Grotto. Cat. Layer/ level A Rim B Rim Indeterminate Rim Thickened Rim Body Lime Impressed Body Total 108 I/1 2 17 19 109 I/2 4 44 48 110 I/3 4 47 51 111 I/II/4 1 18 19 112 II/5 3 29 32 113 II/6 1 1 114 II/7 2 2 115 II/8 19 19 116 II/9 2 7 1 10 118 II/10 1 1 2 4 Totals 3 5 6 4 186 1 205
159 Table 3 17. Summary pottery table for TU 4, Ritidian Grotto. Cat. Layer/ level A Rim B Rim Indeterminate Rim Thickened Rim Body Lime Impressed Body Total 100 I/I 1 29 30 101 I/2 2 88 90 102 I/3 2 36 38 103 I/II/4 1 3 4 104 II/5 7 7 105 II/6 4 4 106 II/7 1 1 3 5 107 II/8 3 2 9 14 117 II/9 2 19 21 119 II/10 3 6 9 Totals 4 4 4 6 204 222 Table 3 18. Summary artifact table for TU 3, Ritidian Grotto. Layer/Level Artifact Type Count Wt. (g) Material I/3 Worked shell 1 5.0 Conus sp . II/5 Bead 1 0.9 Cypraea moneta II/5 Worked shell 1 0.6 S hell II/5 File/abrader 1 6.3 Coral II/9 Beads 3 0.1 Conus sp. III/10 Beads 3 0.3 Conus sp. Table 3 19. Summary artifact table for TU 4, Ritidian Grotto. Layer/Level Artifact Type Count Wt (g) Material I/1 Bead/ornament 1 5.2 Cypraea sp. shell I/3 Fishhook fragments 2 <0.1 S hell I/3 Bead blank and bead 2 0.8 Conus sp. shell I/II/4 Bead 1 <0.1 S hell II/5 Beads 2 0.2 Conus sp. shell II/5 Bivalve shell scraper 1 0.6 Codakia punctata II/6 Beads 2 0.2 Conus sp. shell II/8 Beads 2 0.2 Conus sp. shell II/9 Beads 2 0.3 Conus sp. shell III/10 Beads 3 0.5 Conus sp. shell
160 Table 3 20. Summary table of the invertebrate assemblage recovered from TU 3 and TU 4, Ritidian Grotto. Taxa NISP Wt (g) NISP (%) Wt (%) Coral 2 19.8 <1 <1 Gastropod 2538 3792.6 54 61 Bivalve 927 2056.4 20 33 Beach Shell 9 10.9 <1 <1 Birgus latro 10 29.3 <1 <1 Indeterminate 1231 265.1 26 4 Totals 4716 6167.8 100 100 Table 3 21. List of identified gastropod shells from TU 3 and TU 4, Ritidian Grotto. Taxa Common Name Habitat Preference (Abbot and Dance 2000; Lamprell and Whitehead 1992) Polyplacophora Chitons On rocks in the intertidal zone. Cellana radiata orientalis Rayed wheel limpet On rocks in the intertidal zone. Trochus spp. Trochus snails Littoral and shallow sublittoral zones on rocks and coral reefs and eel grass. Astralium Star snails Found under coral rocks or exposed on reef in intertidal zone. Turbo setosus Rough turban Exposed areas on coral reefs, sublittoral zone in shallow water. Turbo spp. Turban shells Inhabits reefs, found under coral rocks, and are sometimes exposed in the intertidal reef zone. Nerita plicata Plicate nerite Upper portions of the shoreline, in cracks and crevices of rock benches, moves up and down with the tide. Nerita polita Polished nerite Intertidal rocky and coral reefs. Littorinidae Periwinkles Intertidal areas and splash zones, to well above high tide mark. Cerithium nodulosum Giant kno bbed cerith Sand and rubble on intertidal reef flats near outer edge of reefs. Strombus gibberulus gibbosus Gibbose conch Shallow waters on clear sand and sea grass beds, in sandy patches on reef flats, in sandy lagoons and sandy mud bottoms, intertidal to sublittoral to a depth of 20 m. Strombus mutabulis mutabulis Mutable conch Sandy and rubble bottoms of coral and rocky reefs, in clear or turbid waters, sheltered and exposed areas, and most common just below tide marks. Cypraea caputserpentis Serpentâ€™s head cowrie Common on coral reefs and rock benches exposed to wave action, in intertidal and shallow subtidal areas.
161 Table 3 21. Continued Taxa Common Name Habitat Preference Cypraea annulus Gold ring cowrie Common in shallow tide pool in vegetation and under stones, common in all intertidal zones. Cerithiidae Horn shells Shallow waters in clean coral and weedy sand in the intertidal zone. Cypraea moneta Money cowrie Common in open in high in the intertidal zone to shallow portions of the subtidal zone. Cypraea scurra Jester cowrie Often deeply buried in dead coral rubble, reefs, and under coral heads, over the reef ledge, in shallow sublittoral waters to a depth of 10 m. Cypraea spp. Cowrie shells Generally associated with coral reefs Tunnidae Tun shells Sandy bottoms and sea grasses Thais armigera Belligerent rock shell Rocky shores and coral reefs, intertidal and shallow subtidal zones. Morula uva Grape drupe On rocks and coral, in shallow and deeper waters. Morula sp. Rock shells On rocks and coral, in shallow and deeper waters. Drupa ricinus Prickly spotted drupe On rocks and coral, in shallow and deeper waters. Conus sp. Cone shells Found in coral sand, on hard reef substrate, under coral, or in crevices. Nassarias sp. Nassa mud snails Sandy or muddy soft bottoms in marine and brackish waters, in intertidal and sublittoral zones. Land Snails Pythia scarabaeus Common Pythia, salt marsh snail Coastal environments especially near forests Achatina fulica African land snail Diverse habitats including agricultural areas, coastal zones, natural and planted forests, scrub and shrublands, riparian areas, wetlands, and urban areas Table 3 22. Summary table of identified gastropod species from TU 3 and TU 4, Ritidian Grotto. Taxa NISP NISP (%) Polyplacophora 2 <1 Cellana radiata orientalis 39 2 Trochus spp. 27 1 Astralium 13 1
162 Table 3 22. Continued Taxa NISP % NISP Turbo setosus 19 1 Turbo spp. 277 11 Nerita plicata 21 1 Nerita polita 16 1 Nerita spp. 21 1 Littorinidae 1 <1 Cerithium nodulosum 13 1 Cerithiidae 21 1 Strombus gibberulus gibbosus 17 1 Strombus mutabulis mutabulis 57 2 Strombus spp. 30 1 Cypraea caputserpentis 2 <1 Cypraea annulus 1 <1 Cypraea moneta 14 1 Cypraea scurra 2 <1 Cypraea spp. 25 1 Tunnidae 1 <1 Natica sp. 1 <1 Bursa bufonia 1 <1 Thais armigera 1 <1 Morula uva 1 <1 Morula spp. 3 <1 Drupa ricinus 10 <1 Drupa spp. 3 <1 Mitra stictica 1 <1 Nassarias spp. 2 <1 Conus sp. 6 <1 Pythia scarabaeus 518 20 Achatina fulica 9 <1 Gastropoda 1363 54 Total 2538 100
163 Table 3 23. List of bivalve species recovered from TU 3 and TU 4, Ritidian Grotto. Taxa Common Name Habitat Preference Pinctada spp. Pearl oysters Attached to various substrates including rocks, shells, pebbles and plants, in warm shallow waters, sometimes in dense colonies. Modiolus spp. Horse mussels Attached to rocks and in crevices. Isognomon spp. Tree oysters Attached to various hard substrates in tropical shallow waters often in dense colonies. Codakia tigerina Pacific tiger lucine Buried in sandy bottoms in coral reef areas, in shallow sublittoral zones to 20 m. Codakia punctata Punctate lucine Buri ed in sand between shale blocks and in coral reef areas in sublittoral zones to 20 m. Codakia spp. Lucinas shells In clean coral sand in the intertidal zone. Fragum fragum White strawberry cockle In littoral to sublittoral sand to 20 m. Tellina scobinata Rasp tellin In coarse sand and gravel bottoms. Tellinadae Sunset shells Sandy areas in the intertidal zone. Quidnipagus palatam Palate tellin In coarse sandy bottoms in the littoral zone. Gafrarium pectinatum Comb venus Intertidal and shallow sublittoral zones to 20 m. Tridacna maxima Elongate giant clam On reefs partially embedded in coral, in littoral and shallow waters to 20 m. Tridacna spp. Giant clam shells Basally attached to coral substrate. Atactodea striata Striate beach clam Abundant in sandy beaches in the intertidal zone. Anadara cf. antiquata Antique ark On muddy bottoms, in intertidal and sublittoral zones to a depth of 25 m. Table 3 24. Summary of bivalves recovered from TU 3 and TU 4, Ritidian Grotto. Taxa NISP Wt (g) NISP (%) Wt (%) Fragum fragum 2 3.3 <1 <1 Pinctada sp. 7 46.1 1 2 Modiolus sp. 56 23.1 6 1 Pinctada sp. 7 46.1 1 2 Modiolus sp. 56 23.1 6 1 Isognomon spp . 7 17.1 1 1 Codakia punctata 13 51.4 1 2 Codakia tigerina 1 3.9 <1 <1 Codakia spp . 2 11.9 <1 1
164 Table 3 24. Continued Taxa NISP Wt (g) NISP (%) Wt (%) Tellina scobinata 1 7.9 <1 <1 Tellinadae 27 35.9 3 2 Quidnipagus palatam 2 7.3 <1 <1 Gafrarium pectinatum 3 4.8 <1 <1 Tridacna maxima 2 60.5 <1 3 Tridacna spp. 113 1226 12 60 Atactodea striata 3 3.5 <1 <1 Anadara cf. antiquata 4 60.3 <1 3 Bivalvia 684 493.4 74 24 Totals 927 2056.4 100 100 Table 3 25. Summary table of fish bone recovered from TU 3 and TU 4, Ritidian Grotto. Taxa NISP Wt (g) NISP (%) Wt (%) Serranidae 1 .5 <1 <1 Scaridae 26 9.2 1 5 Osteichthyes 4646 164.1 99 94 Total 4673 173.8 100 100 Table 3 26. Summary of bird bone from TU 3 and TU 4, Ritidian Grotto. Layer/level NISP Wt (g) I/3 2 0.4 I/II/4 1 0.1 II/7 7 2.0 II/8 24 4.5 II/9 110 8.9 III/10 48 2.6 III/11 3 0.3 Total 195 18.8 Table 3 27. Summary of mammal bone from TU 3 and TU 4, Ritidian Grotto. Taxa Layer/Level NISP Wt (g) Rattus sp. I/1 1 <0.1 Sus scrofa I/2 1 1.1
165 Table 3 27. Continued. Taxa Layer/Level NISP Wt (g) Pteropid II/7 2 0.1 Mammalia II/7 1 0.9 Mammalia II/8 1 0.8 Pteropid II/9 10 0.8 Mammalia II/9 1 13.5 Rattus sp. III/10 6 0.3 Mammalia III/10 4 1.1 Total 27 18.6 Table 3 28. Summary of NISP of human remains recovered from TU 3 and TU 4, Ritidian Grotto. Layer/Level NISP I/II/4 1 II/5 1 II/6 7 II/7 6 II/8 2 II/9 2 Total 19 Table 3 29. Summary of faunal assemblage for TU 3 and TU 4, Ritidian Grotto. Taxa NISP NISP (%) Wt (g) Wt (%) Invertebrates Coral 2 <1 13.8 <1 Gastropods 2538 26 3792.6 59 Bivalves 927 10 2056.4 32 Beach Shell 9 <1 10.9 <1 Indeterminate Shell 1231 13 265.1 4 Birgus latro 10 <1 29.3 <1 Vertebrates Fish 4673 49 173.8 3 Birds 195 2 18.8 <1 Mammals 27 <1 18.6 <1 Total 9612 100 6379.3 100
166 Figure 3 1. Close up map of the Bapot 1 site showing the locations of excavation units from previous and current archaeological investigations.
167 Figure 3 2. Plan view map showing layout of the nine excavation units within Block A, Bapot 1 site, Saipan. See Figure 3 1 for the location of Block A within the Bapot 1 site.
168 Figure 3 3. Stratigraphic profile and photograph of the north wall of Block A, Bapot 1.
169 Figure 3 4. Stratigraphic profile and photograph of the south wall of Block A, Bapot 1.
170 Figure 3 5. Stratigraphic profile and photograph of the east wall of Block A, Bapot -1.
171 Figure 3 6. Stratigraphic profile and photograph of the west wall of Block A, Bapot 1.
172 Figure 3 7. Bapot 1, Block A, calibrated radiocarbon dates plotted versus depth below surface.
173 Figure 3 8. Deposition model of radiocarbon dates from Bapot 1, Block A created using OxCal v4.2.2.
174 Figure 3 9. Bapot 1 site pottery. A ) Decorated (San Roque type) Pre -Latte sherd from TU 2, 250â€“260 cmbd. B ) Decorated (Achugau type) Pre -Latte sherd from TU 5, 230â€“240 cmbd, Bapot 1 (Courtesy of Olaf Winter, ANU). Figure 3 10. Bapot 1 site pottery. A ) Decorated Pre -Latte s herd from TU 5, 210â€“210 cmbd. B ) Decorated Pre-Latte sherd from TU 3, 190â€“200cmbd, Bapot 1 (Courtesy of Olaf Winter, ANU). A B A B
175 Figure 3 11. Bapot 1 site pottery. A ) Decorated Pre -Latte rim sherd from TU 6 (Ca talog No. 126), 180â€“190 cmbd. B ) Decorated Pre -Latte sherd from TU 6, 130â€“140 cmbd, Bapot 1 (Courtesy of Olaf Winter, AN U). Figure 3 12. Frequency of ceramics (NISP) versus depth below datum, Bapot 1 (Courtesy of Olaf Winter, ANU). A B
176 Figure 3 13. Ground Cypraea moneta shell beads from 180â€“250 cmbd, Bapot 1 (Courtesy of Olaf Winter, AN U). Figure 3 14. Small shell beads from Catalog Number 119, 170â€“180 cmbd, Bapot 1. Figure 3 15. Pendants. A ) Possible Cypraea tigris shell pendant, Catalog N umber 156, TU 1, 230â€“240 cmbd. B ) A pendant made from the ground apex of a conical gastropod shell, possibly a Conus sp. Catalog Number 157, TU 3, 220â€“230 cmbd, Bapot 1 (Courtesy of Olaf Winter, AN U). A B
177 Figure 3 16. Fragment of a shell band, or bracelet. Ca talog Number 151, TU 1, 210 â€“220 cmbd, Bapot 1 (Courtesy of Olaf Winter, AN U). Figure 3 17. Bivalve shell scraper. Catalog Number 67, TU 4, 100â€“110 cmbd, Bapot 1. Figure 3 18. Pencil urchin tools. A ) Pencil urchin tool, Cat alog 164, TU 1, 240â€“250 cmbd. B ) Close up of worked surface of pencil urchin tool, Catalog Number 164, Bapot 1. A B
178 Figure 3 19. Tridacna sp. shell tools. A ) Tridacna s p. shell adze, Catalog 8, TU 8, 20â€“35 cmbd. B ) Possible adze or scraper, Catalog 122, TU 2, 180â€“190 cmbd, Bapot 1 (Courtesy of Olaf Winter, AN U). Figure 3 20. Worked shell. A ) Pearl shell fishhook, Cata log 171b, TU 3, 230â€“240 cmbd. B ) A possible fishhook blank, Catalog Number 89, TU 1, 140â€“150 cmbd (Courtesy of Olaf Winter, AN U). a b a b
179 Figure 3 21. Worked shell implement, or possible compound fishhook point. Catalog 143, TU 6, 200â€“210 cmbd, Bapot 1.
180 Figure 3 22. Frequency of shell artifacts versus depth below datum from Bapot 1 (Courtesy of Olaf Winter, AN U). Figure 3 23. Frequency of lithic art ifacts versus centimeters below datum from Bapot 1.
181 Figure 3 24. Bapot 1 frequency of fish bone . A) versus cmbd. B ) Versus cal a BP. A B
182 Figure 3 25. NISP frequencies of gastropods and bivalves . A) P lotted versus centimeters below datum. B) P lotted versus cal a BP, Bapot -1. 65 85 96 104 113 122 130 140 153 160 171 181 190 200 210 220 230 240 250 260 NISP Gastropod 198 204 297 360 244 451 353 427 563 989 797 485 488 936 743 673 501 390 111 118 NISP Bivalve 124 147 239 380 351 908 764 750 576 1072 832 401 457 966 642 1021 478 623 265 132 0 200 400 600 800 1000 1200 NISP A B
183 Figure 3 26. NISP of identified species. A) Plotted versus cmbd. B) P lotted versus cal a BP, Bapot 1. A B
184 Figure 3 27. Shannon Wiener Indexes for all Bapot 1 invertebrates, gastropods, and bivalves. A) Plotted versus cmbd. B ) P lotted versus cal a BP. A B
185 Figure 3 28. Plan view map of the Ritidian Grotto cave and rockshelter with locations of TU 1 through TU 6.
186 Figure 3 29. Stratigraphic profiles of the west walls of TU 2, TU 3, and TU 4 and the north wall of TU 4, Ritidian Grotto.
187 Figure 3 30. Stratigraphic profiles of the East walls of TU 2 through TU 4, and the south wall of TU 2, Ritidian.
188 Figure 3 31. Ritidian Grotto radiocarbon dates plotted versus cmbd.
189 Figure 3 32. Deposition model based on radiocarbon dates from Ritidian created using OxCal v4.2.2 (Bronk Ramsey 2013).
190 Figure 3 33. Shell ornaments. A ) Cypraea sp. bead, or ornament , from TU 4, Layer I, Level 1. B) Worked Conus sp. shell from TU 3, Layer I, Level 3, Ritidian Grotto. Figure 3 34. Shell beads and bead blank. A ) Shell beads from TU 4, Layer III, level 10. B ) Conus sp. bead blank and bead from TU 4, Layer I, Level 3, Ritidian Grotto. Figure 3 35. Shell and coral tools. A ) Bivalve shell scraper from TU 4, Layer II, Level 5. B ) Coral fi le, or abrader, from TU 3, Layer II, Level 5, Ritidian Grotto. A B A B A B
191 Figure 3 36. Fishhook fragments from TU 4, Layer I, Level 3, Ritidian Grotto. Figure 3 37. Frequency of bird bone recovered from TU 3 and TU 4 plotted by stratigraphic layer and excavation level, Ritidian Grotto.
192 Figure 3 38. Frequency of fish bone NISP . A) P lotted versus stratigraphic layer and excavation level. B) P lotted versus cal a BP, Ritidian Grotto. A B
193 Figure 3 39. NISP of identified invertebrate species . A ) P lotted versus cmbd. B ) P lotted versus cal a BP, Ritidian Grotto. A B
194 Figure 3 40. Shannon Wiener diversity index values for all invertebrate species, gastropods, and bivalves . A ) P lotted versus centimeters below datum. B ) P lotted versus cal a BP, Ritidian . A B
195 CHAPTER 4 FISHERIES SCIENCE ANALYSIS AND RESULTS Introduction This chapter focuses on the methods and results of the fisheries sciencebased analysis of marine invertebrate shells for the present project. Methods utilized for the study include morphometric measurements of shells, preparation of acetate peels from she ll specimens for aging, and statistical tests utilized to analyze the archaeological residues of prehistoric fisheries. Age based analyses were also performed on Anadara cf. antiquata valves to measure impacts on the fishery related to human harvesting. An adara cf. antiquata shells were used because they were common in the assemblage, relatively well preserved, and easily identifiable. Results of the fisheries sciencebased analyses are also presented. The primary focus of these analyses was to evaluate the effects of human exploitation on prehistoric fisheries. To accomplish this, length and age based characteristics of the archaeological residues of prehistoric marine invertebrate fisheries populations were compared through time. Invertebrates, in this case, are and excellent proxy for studying prehistoric fisheries and the prehistoric marine resource based economy of the Marianas. This is because they are relatively well preserved and abundant in archaeological contexts, are represented by an extensive l ist of species that were harvested from almost every aquatic habitat that exist on the islands of Guam and Saipan, were used for several reasons aside from subsistence, and have been tied to various beliefs and gender and age related subsistence practices throughout the Pacific. Some species, commonly found in archaeological contexts, are also widely distributed throughout warm regions of the Pacific and represent important modern fisheries (Broom 1985; MacKenzie 2001).
196 The primary objective of applying th ese analyses was to evaluate interactions between pre Contact humans and the marine environment and determine how these factored in the development of prehistoric indigenous culture in the Mariana Archipelago through time. These included identifying changes in the archaeological residues of prehistoric fisheries (e.g. identifying the initial effects of human predation upon previously unexploited populations or over exploitation of a population), and advancing avenues of research that address more complex is sues regarding human environment interactions related to the prehistoric subsistence economies, beyond approaches that privilege the environment. Methods Fisheries sciencebased analyses relied upon the length of Anadara cf. antiquata and Gafrarium pectin atum valves and age properties of prehistoric populations of Anadara cf. antiquata. This required recording metric data and estimating ages for subsamples of shell from the assemblage. Following identifications of shells from each provenience in the assemb lage, all of the whole left valves, or left valves from which maximum length could be recorded, were selected for analysis. Shell length was measured from the anterior to the posterior margin the left valve in a sample and was recorded to the nearest milli meter using digital calipers ( Figure 41) (Claassen 1998:21). Because of the large number of shells and the timeintensive process involved in preparing acetate peels for age determination, a sub sample was selected for aging. This included a range of sizes for each dated period (2000, 2200, 3100, and 3300 cal a BP). This was done to identify significant differences in length vs. age ranges between per iods that might represent changes in growth rates in the prehistoric Anadara populations through time. A determination of relative minimum age (RMA) of bivalves was made through the examination of growth bands in the shell structure of a subsample of left Anadara cf. antiquata
197 valves. Determining the age structure of a prehistoric population of shellfish was important because populations undergo changes with increasing fishing pressure. Most notably, the age structure becomes truncated or skewed towards smaller, and therefore younger, fish or mollusks as exploitation increases over time (Ricker 1975). The RMA of a given specimen was determined by counting opaque and translucent growth bands from the sectioned umbo of each shell ( Figure 4 1). The growth bands are similar to those found in the hard structures of fish, and consist of alternating opaque and translucent bands laid down during different time s of the year. Normally, this includes fast summer growth leaving thick translucent bands and slow winter growth leaving narrow opaque bands (Ambrose 1983:311). While, Anadara cf. antiquata inhabits tropical areas with similar temps year round and therefore may not exhibit seasonal growth patterns. However, it has been shown that the tropical species Anadara senilis exhibits seasonal variations in growth resulting in the formation of growth bands. These bands form during the rainy season when water temperat ures are lower, salinity decreases, and turbidity increases, which effects feeding and growth rates (Broom 1985:18) To test the efficacy of aging bivalve shells from archaeological contexts, acetate peels were made from a subsample of Anadara cf. antiquata shells using the method of Ropes (1984, 1987) for preparing acetate peels from sectioned valves of ocean quahog (Artica islandica ). First, valves were sectioned, or cut, through the dorsal -ventral axis of the umbo ( Fi gure 4-1) using a Buehler ISOMET Low Speed Saw or a benchtop tile saw fashioned with a rheostat and diamond blade. The sectioned surface of each valve was then hand ground with four succe ssively finer grits (600, 800, 1200, and 2000) of wetable carborundum paper. This removed saw marks and created a smooth blemish free surface. Each specimen was then soaked in 100% household bleach for a few hours. The surface of each cut and polished shel l was then etched by immersion
198 in 1% hydrochloric acid solution for one minute and rinsed with fresh water. Acetate peels were produced by flooding the etched surface of the shell with acetone and then covering the flooded surface with a sheet of clear acetate. Any excess acetone was drained and the shell was turned over with the attached acetate sheet and pressed onto a sheet of paper for one minute. After drying for one hour, the acetate sheet was peeled off and sandwiched between two clean glass slides f or aging. To determine the age of a specimen each acetate peel was examined five times through a stereomicroscope with transmitted light at 7.5X to 35X magnification. All of the visible opaque bands were recorded. The RMA was then determined by recording t he mode from the five rounds of aging for each specimen. Visibility of growth bands differed among specimens and some were easier to age than others. This may have been a consequence of leaching or other chemical taphonomic processes caused by burial in calcareous sand for long periods of time. The ease or difficulty of reading the peels was recorded using a tripartite code with the following designations: Grade A ( Figure 4 2 A ) = clearly visible growth bands that could be read without making adjustments to the light source, position of the peel on the microscope, or focal length of the microscope. Grade B ( Figure 42 B ) = most growth bands were clearly visible and some were faint. Some adjustments to either the light source, position of the peel on the microscope, or focal length of the microscope had to be made to age the specimen. Grade C ( Figure 42C) = Most growth bands were faint and difficult to read and some were clear. Multiple adjustments to either the light source, position of the peel on the microscope, or focal length of the microscope had to be made. Age and growth data from a subsample of aged Anadara cf. antiquata shells were fit to a simple linear regression model to estimate length at RMA data for the assemblage of left valves from Bapot 1. This allowed the large number of length measurements to be converted into age estimates (Cushing 1981:6). When using linear models to estimate age, the length of some valves
199 was less than the intercept value for each model. In these cases, the ages of these valves were recorded as 0. The version of the linear regression equation used to estimate leng th -at RMA was: = + Here, Yi is the estimated length at RMA and and are population parameters (Zar 1999:326). Analysis of the data employed statistical tests to measure differences in length frequencies of prehistoric populations of Anadara cf. antiquata and Gafrarium pectinatum differentiated by depth (cmbd) and dated proveniences. Differences in mean length relationships within groups were compared using a single factor analysis of variance (ANOVA) for samples by depth and date before present. All ANOVAs were calculated using Microsoft Excel 2010. To test for differences in mean length among groups, the multiple comparison Tukey test for unequal size samples was used (Zar 1999:211214). A q value in the Tukey test was derived by calculating the standard error using the following equation: = 2 1 + 1 Here, s2 is the error mean square from the ANOVA and n is the number of data in each group. The q value was then calculated from: = If the calculated q value, derived from th e equation above, was greater than or equal to the critical value, q,v,k, derived from the â€œStudentized rangeâ€ table (Zar 1999:Appendix 64, Table B.5), then the null hypothesis Ho: B = A was rejected and there were significant differences in the means of the two samples (Zar 1999:211).
200 Distributions of percent length and percent cumulative length frequencies for Anadara cf. antiquata and Gafrarium pectinatum valves were plotted by grouped, or binned, shell lengths according to early (3100â€“3300 cal a BP) and late (2000â€“2200 cal a BP) periods. Lengths were sorted into 5mm bins. The frequency is the percentage of valves in these length bins by species, based on the total number of shells measured for each species. Cumulative frequency is the cumulative percentage of the lengths of valves within each bin. Plots for frequency and cumulative frequency were created to facilitate comparisons of different length distributions (Zar 1999). To determ ine if there were differences in the ages of Anadara cf. antiquata valves from samples from different time periods, the Kolomogorov-Smirnov (K -S) goodness of fit procedure was used (Zar 1999:478). First, the observed frequencies of length for each RMA were recorded. The cumulative observed frequencies, Fi , were determined from which the cumulative relative frequencies were obtained from: = Here, n is the number of measurements, rel Fi is the proportion of the sample measurements. Then for each RMA the cumulative relative expected frequency, rel , is determined. The test statistic for the K -S is determined by calculating: = = and = = for each i . For the latter equation it is important to know that F0 = 0, so D 'i = rel i. The K S test statistic is: = [( max ) , ( ) ] ,
201 indicating that D equals the largest value for Di or D 'i. Critical values for the K -S statistic are ,n and if D ,n, the H0 is rejected at the â€“ 479). Results Although thousands of individual fish bones and shell specimens were recovered during excavations at both the B apot 1 and Ritidian Grotto sites, large numbers of NISPs from most identified species were relatively uncommon. This was especially the case for faunal remains from the Ritidian Grotto site. Only Anadara cf. antiquata and Gafrarium pectinatum shells were recovered from Bapot 1 in large enough numbers to be subjected to fisheries sciencebased analyses. These analyses were focused on measuring changes in population dynamics through measuring average shell size and relative minimum age through time. First, th e results of the morphometric analyses of Anadara cf. antiquata and Gafrarium pectinatum shells are presented. The results of the agebased analysis of Anadara cf. antiquata are then presented followed by a chapter summary. Morphometric Analyses ( Anadara c f. antiquata) Shells from Anadara cf. antiquata (antique arc clam) were ubiquitous throughout the stratigraphic sequence of Bapot 1 and accounted for 25% of the total NISP of the bivalve assemblage ( Table 3 12). The average shell length with one standard deviation above and below the mean was plotted for every Anadara cf. antiquata shell sample recovered from each 10 cm level below the excavation d atum and by date before present. When mean shell lengths for Anadara cf. antiquata were plotted by depth (cmbd), mean lengths fluctuated around 40 mm in samples collected between 130 and 260 cmbd. There was a decline in average length in samples from between 100 and 130 cmbd and an abrupt increase in Figure 43 A
202 Table 3 2). To test for significant differences in average valve length between levels, a singlefactor ANOVA was conducted. Results of the ANOVA showed a significant difference at the p<0.05 level between the 19 groups [F (18, 874)=4.709, p 0.0001] (Table 4 1 and Table 4 2). To determine which samples were significantly different, Tukey tests were used to compare each possible pair of measured shell samples by depth (cmbd). The critical q value, or â€œStudentized rangeâ€ (Zar 1999:211), of 4.974 was determined from the st atistical table for Critical Values of the q Distribution with = 0.05, v (or degrees of freedom within groups) = 874, and k (or number of groups) = 18 (Zar 1999:Appendix 64, Table B.5). Calculated q values larger than the critical q value of 4.974 represent pairs with significantly different means. The Tukey tests revealed that most of the compared sample pairs did not have significantly different mean lengths. Significant differences mbd and samples from 11 cmbd, and in samples Table 4 3). Charcoal Table 3 2) and therefore corresponds wi th the Late Unai Phase of the Pre-Latte Period. Low frequencies of ceramics ( Figure 3 12), shell artifacts ( Figu re 3 22), and lithic artifacts ( Figure 3 23) were recovered from this depth. An inc reasing trend in the frequency of fish bone ( Figure 324 A ), the beginning of a decreasing trend in the overall frequencies of marine invertebrates ( Figure 3 25 A ), and an increase in diversity index values for the overall marine invertebrate assemblage, with a
203 decreasing trend in diversity index val ues for bivalves ( Figure 3 27 A ), are also associated with materials recovered from these depths. When plotted by date before present, the mean shell l engths of Anadara cf. antiquata shells was 42 mm in samples dating to 3300 cal a BP, 40 mm in samples dating to 3100 cal a BP, 39 mm in samples dating to 2200 cal a BP, and 35 mm in samples dating to 2000 cal a BP ( Tab le 41) ( Figure 4 3 B ). Results of a single factor ANOVA showed a significant differ ence at the p<0.05 level between the four groups [F (4, 888)=8.747, p 0.0001] (Table 4 5). Based on these results, post -hoc comparisons using the Tukey test for each possible sample pair were conducted. The critical q value of 3.633 was determined from the statistical table for Critical Values of the q Distribution with = 0.05, v (or degrees of freedom within groups) = 888, and k (or number of groups) = 4 (Zar 1999:Appendix 64, Table B.5). Results of the Tukey test produced five calculated q values for pairs that were highe r than 3.366, having significantly different mean shell lengths. These included pair wise comparisons of samples dating to 2000 and 2200 cal a BP, 2000 and 3100 cal a BP, 2000 and 3300 cal a BP, and 2200 and 3300 cal a BP (Table 46). Thus, mean shell length during the early (3100 and 3300 cal a BP) and middle (3100 2200 cal a BP) portions of the dated sequence were not significantly different. Mean she ll lengths from the later part of the middle portion of the sequence (2200 cal a BP) were significantly different from those of the earliest (3300 cal a BP) portion of the sequence and mean lengths from the late portion of the sequence were significantly d ifferent from all of the other groups. When the percent frequency of valve lengths were plotted based on 5 mm bins, the lengths of valves from the more recent time period (2000â€“2200 cal a BP) were skewed to the left of the plot (Figure 44). This indicated that the valves from the later time period are generally smaller than those from the older time period (3100 â€“3300 cal a BP). Cumulative frequencies of
204 valve lengths plotted based on 5 mm bins ( Figure 4 5), also showed that valves from the later, more recent, period were generally smaller that valves from the earlier period. Morphometric Analyses ( Gafrarium pectinatum ) Shells from Gafrarium pectinatum , or the comb Venus clam, were also prevalent throughout the stratigraphic sequence and accounted for 30% of the total NISP of the bivalve assemblage (Table 3 12). Mean shell lengths for Gafrarium pectinatum were plotted by depth (cmbd) and by date before present. This highlighted similar trends in m ean shell length to those observed for Anadara cf. antiquata. When plotted by depth (cmbd), the mean shell length of Gafrarium pectinatum decreased throughout the sequence with the largest means derived from deeper and older samples and the smallest means in the shallower and younger samples. These included mean lengths of 29.1 mm â€“240 cmbd, and 29.9 Figure 4 6 A ). Unidentified charcoal samples from â€“3318 and 2980â€“3137 cal a BP, and a Conus 230 cmbd dated to 3379â€“3445 and 3161â€“ 3319 cal a BP (Table 3 2). Most of these dates fall within the Early Unai Phase of the Pre-Latte Period (3000 to 3500 cal a BP). Later in the sequence, mean shell lengths in sa to 26.1 mm in she Figure 4 6 A ). Results of the singlefactor ANOVA for mean lengths of Gafrarium pectinatum shell samples by excavation depth showed a significant difference at the p < 0.05 level between the 18 groups [F (18, 815) = 6.304, p 0.0001] (Table 4 7 and Table 4 8). Based on these data, Tukey tests were
205 performed for each possible pair of shell samples from the different depths and the cri tical q,v,k of 4.974 was determined from the statistical table for Critical Values of the q Distribution with = 0.05, v (or degrees of freedom within groups) = 815, and k (or number of groups) = 19 (Zar 1999:Appendix 64, Table B.5). Results of the Tukey test showed significant differences in mean shell lengths between paired samples from 65 85 cmbd and 220230 cmbd, 90104 cmbd and 220230, 130142 and 190200 cmbd, and 130140 cmbd and 220230 cmbd. Also, samples from 109122 cmbd were significantly dif ferent from all samples from 140153 cmbd through 240250 cmbd (Table 4 9). This was also similar to the results of pair wise comparisons of mean shell lengths for Anadara cf. antiquata samples by depth (cmbd) and emphasized a significant shift in mean length for both Anadara cf. antiquata and Gafrarium pectinatum shells from 109 122 cmbd. Materials from this depth dated to the Late Unai Phase of the Pre-Latte Period. Mean shell lengths of samples of Gafrarium pectinatum shell plotted by date before present illustrated a general decrease in mean size from the earliest to the lat est portions of the sequence. The largest means were derived from samples dating to 3300 cal a BP with a mean length of 29.2 mm, and 3100 cal a BP with a mean length of 28.32 mm. Mean length in samples dating to 2200 cal a BP decreased to 25.9 mm and furth er decreased in samples dating to 2000 cal a BP, to 25.6 mm ( Figure 4 6 B ). Results of a singlefactor ANOVA for the mean lengths of Gafrarium pectinatum shells from dated samples showed a significant difference at the p < 0.05 level between the four groups [F (4, 830) = 25.505, p 0.0001] (Table 4 10 and Table 411). Based on these data, Tukey tests were performed for each possible pair of dated groups and a critical q,v,k value of 3.633 was determined from the statistical table for Critical Values of the q Distribution with = 0.05, v (or degrees of freedom within groups) = 830, and k (or number of groups) = 4 (Zar 1999: Table B.5,
206 Appendix B). Results of the pairwise comparisons showed significant differences between paired groups dating to 2000 and 3100 cal a BP, 2200 and 3100 cal a BP, and 2000 and 3300 cal a BP (Table 412). The percent frequency of Gafrarium pectinatum valve lengths plotted based on 5mm bins, showed that the lengths of valves from the more recent time period (2000â€“2200 cal a BP) have a relatively even distribution while shell lengths from the older time period (3100 â€“3300 cal a BP) are skewed to the right of the plot ( Figure 4 7). This also indicates that the valves from the later time period are generally smaller than those from the older time period (3100 â€“ 3300 cal a BP). Cumulative frequencies of Gafrarium pectinatum valve lengths plotted based on 5 mm bins ( Figure 4 8), also showed that valves from the later, more recent , period were generally smaller that valves from the earlier period. Comparisons of the means show a significant change in mean length at around 2000 cal a BP, during the Late Unai Phase of the Pre -Latte Period. These changes in mean shell size of prehisto ric Anadara cf. antiquata shell are considered further below, using age based analyses. Age -Based Analyses ( Anadara cf. antiquata) A total of 101 left Anadara cf. antiquata valves were selected for aging. These included 22 valves from the 2000 cal a BP sample, 26 valves from the 2200 cal a BP sample, 31 valves from the 3100 cal a BP sample, and 22 valves from the 3300 cal a BP sample. Lengths ranged from 20.559.7 mm in the 2000 cal a BP sample, 18.064.2 mm in the 2200 cal a BP sample, 24.768.8 mm in t he 3100 cal a BP sample, and 27.366.4 mm in the 3300 cal a BP sample. Relative minimum ages ranged from 1 19 in the 2000 cal a BP sample, 118 in the 2200 cal a BP sample, 1 17 in the 3100 cal a BP sample, and 219 in the 3300 cal a BP sample. Aging Anadara cf. antiquata shells recovered from Bapot 1 using acetate peels was a difficult process. Only a small proportion of the sectioned shells produced acetate peels with
207 clearly visible growth bands that were coded with a grade of â€œA,â€ with clearly visible opaque bands. These were dominated by young individuals, with shells that displayed one or two clearly visible growth increments. Most specimens were scored with a grade of â€œC,â€ with faint increments requiring light source adjustments, repositioning the peel on the microscope, or adjustments to the focal length of the microscope. This was likely related to the shells having been buried in calcareous sand for thousands of years, and perhaps chemically altered. Because average shell lengths between the early and late portions of the dated sequence were significantly different and mean lengths during the middle portion of the sequence (2200 3100 cal a BP) were not significantly different, separate regressions for combined age samples from the early groups (31003300 cal a BP) and late period groups (20002200 cal a BP) were used in estimating mean length at age. Results of the K S test (Zar 1999) showed significant differences between RMA of samples from 20002200 cal a BP and 31003300 cal a BP at p<0.05 and D = 0.124. Differences in growth were evident from linear regressions based on estimated mean valve lengths at age for the two dated samples, with samples from the 2000 2200 cal a BP showing more rapid growth ( Figure 4 9). Using mean length at RMA estimated with simple linear regressions for all un aged shells in the assemblage, shell length at age structures between dated periods were compared. Age frequency distributions of shells from each dated period were plotted. A total of 34 left and 339 left Anadara cf. antiquata valve shells were measured from the 2000 and 2200 cal a BP samples respectively. The RMA from these combined samples ranged from < 1, o r 0, to 20, with the majority between 3 and 7 ( Figure 4 10 ). The sample dating to 3100 cal a BP consisted of 383 left valves and samples dating to 330 0 cal a BP equaled 137 measured left valves. The RMA from these combined samples ranged from < 1 to 20 with the highest percentages of valves in
208 RMA bins between 2â€“5 RMA ( Figure 4 10 ). Generally, samples from the later portions of the sequence have a greater proportion of older specimens while samples from the earlier periods have a higher proportion of younger specimens. This is evident in shells dating to 2000 and 2200 cal a BP in which the percent frequency of RMA is skewed towards the right of the plot and with frequency distributions for RMA in samples dating to 3100 and 3300 cal a BP in which the percent frequency of RMA is skewed towards the left of the plot ( Figure 410). When cumulative frequencies of RMA are plotted, it is apparent that older clams accumulated earlier in the frequency distribution of later period (2000â€“2200 cal a BP) and therefore grew slower than the clams from the earlier periods (3100â€“3300 cal a BP) ( Figure 4 11). Summary and Discussion Fisheries sciencebased methods and principles were employed to evaluate the archaeological residues of two prehistoric fisheries. These included shells from Anadara cf. antiquata and Gafrarium pectinatum recovered from Block A, Bapot 1. Specifically, length based analyses were used on Anadara cf. antiquata and Gafrarium pectinatum shells and age based analyses were used on Anadara cf. antiquata to measure the effects of human fishing upon the two marin e taxa. Mean shell lengths of samples of Anadara cf. antiquata and Gafrarium pectinatum from different depths and dates before present were compared using the post -hoc Tukey test. Results of the length based analyses indicated that the mean length of samp les of Anadara cf. antiquata and Gafrarium pectinatum from 109122 cmbd differed significantly from almost all other samples. Materials from this depth dated to 2147 â€“2350 cal a BP, corresponding with the Late Unai Phase of the Pre -Latte Period. This was corroborated by analyses of mean shell lengths of Anadara cf. ant iquata and Gafrarium pectinatum by date before present. Generally, samples dating to around 2200 cal a BP were significantly different from those dating to 3300 cal a BP.
209 Also, mean lengths from samples of both Anadara cf. antiquata and Gafrarium pectinatum from the later part of the sequence were significantly different from groups in the middle and early portions of the sequence. Generally, mean lengths of both Anadara cf. antiquata and Gafrarium pectinatum appear to decrease through time. The pair wise comparisons of mean Anadara cf. antiquata lengths analyzed by depth and date, and pair wise comparisons of mean Gafrarium pectinatum lengths analyzed by depth and date, appear to agree with the changing frequencies of artifacts and vertebrate and invertebrate remains. During the Late Unai Phase of the Pre-Latte Period, frequencies of artifacts recovered from Bapot 1 decreased. Frequencies of marine invertebrates also decreased while the frequency of fish bone increased. Furthermore, diversity index values for the overall marine invertebrate assemblage and gastropods increased while diversity index values for bivalves decreased. Results of age based analyses showed that growth rates between Anadara cf. antiquata shells from early in the sequence (3300 3100 ca l a BP) were significantly faster than growth rates of shells from the later parts of the sequence (22002000 cal a BP). Also, shells from the later portions of the sequence were generally older that those from the early portions of the sequence. While th e results of the K S test showed significant differences between RMA of samples from 20002200 cal a BP and 31003300 cal a BP at the p <0.05 level, these differences appear to be relatively minor and not indicative of overharvesting. Normally, overharvesti ng would cause shifts in age distributions towards younger specimens as older and bigger clams were fished out. Also, overexploitation usually causes increased growth rates in individuals. Rather, the higher proportion of older calms were recovered from sa mples from the more recent periods and the slower growth rates in clams from the more recent periods contradict the expected effects of overexploitation.
210 Table 4 1. Summary of Anadara cf. antiquata valve length data for singlefactor ANOVA by cmbd, Bapot 1. Groups Count Sum Average Variance 4 116.4 29.100 14.727 14 523.7 37.407 117.178 7 254.2 36.314 95.111 9 291.5 32.389 30.024 30 988.3 32.943 50.393 34 1252.0 36.824 68.684 68 2754.9 40.513 70.840 82 3268.5 39.860 66.544 125 5110.2 40.882 52.249 91 3864.6 42.468 50.710 36 1507.8 41.883 70.633 44 1716.9 39.020 63.237 73 2884.1 39.508 51.077 52 2113.8 40.650 79.094 87 3308.2 38.025 33.983 39 1683.8 43.174 75.019 54 2204.8 40.830 58.137 28 1202.4 42.943 68.126 16 676.3 42.269 43.148 Table 4 2. Results of single factor ANOVA for Anadara cf. antiquata for mean valve lengths by cmbd, Bapot 1. Source of Variation SS df MS F P value F crit Between Groups 4997.57 18 277.642805 4.709168839 1.615668 Within Groups 51529.22 874 58.95792112 Total 56526.79 892
211 Table 4 3. Results of Tukey test with calculated q values for pairs of Anadara cf. antiquata valve groups with significantly different mean lengths by cmbd, Bapot 1. The critical q value, or â€œStudentized rangeâ€ for the comparisons = 4.974. Depth (cmbd) 6.36 5.97 7.19 5.31 8.33 6.66 5.58 6.19 5.37 7.76 4.99 6.38 5.07 7.01 5.55 Table 4 4. Summary of Anadara cf. anadara mean valve length data by date before present (cal a BP), Bapot 1. Groups Count Sum Average Variance 2000 34 1185.800 34.876 80.959 2200 339 13373.900 39.451 65.664 3100 383 15395.400 40.197 56.224 3300 137 5767.300 42.097 63.023 Table 4 5. Results of single factor ANOVA for Anadara cf. antiquata mean valve lengths by date before present, Bapot 1. Source of Variation SS df MS F P value F crit Between Groups 2142.86744 4 535.71690 8.747375 2.381956 Within Groups 54383.92611 888 61.24316 Total 56526.79355 892
212 Table 4 6. Results of Tukey test with calculated q values for all pairs of Anadara cf. antiquata valve length groups with significantly different means by date before present, Bapot 1. The critical q value, or â€œStudentized rangeâ€ for the comparisons = 3.366. Date (cal a BP) 2000 cal a BP 2200 cal a BP 2200 4.60 3100 5.31 3300 6.45 6.01 Table 4 7. Summary of Gafrarium pectinatum mean valve length data (mm) for singlefactor ANOVA by cmbd, Bapot 1. Groups Count Sum Average Variance 5 111.4 22.28000000 5.56700000 10 276.5 27.65000000 33.23833333 21 532.3 25.34761905 27.84961905 21 549.7 26.17619048 26.93690476 99 2426.9 24.51414141 22.47000206 34 934.3 27.47941176 21.86350267 61 1587.1 26.01803279 20.47816940 45 1227.5 27.27777778 16.13676768 83 2326.4 28.02891566 15.32208052 55 1564.1 28.43818182 14.76018182 35 999.1 28.54571429 25.64608403 28 798.2 28.50714286 24.48365079 62 1793.7 28.93064516 12.28347171 38 1079.0 28.39473684 16.47889047 98 2737.9 27.93775510 14.80381759 63 1883.8 29.90158730 16.78790067 49 1380.9 28.18163265 16.18361395 19 562.2 29.58947368 24.67210526 8 232.5 29.06250000 9.294107143 Table 4 8. Results of single factor ANOVA for mean Gafrarium pectinatum valve lengths by cmbd, Bapot 1. Source of Variation SS df MS F P value F crit Between Groups 2086.77 18 115.9316534 6.304125053 1.616524002 Within Groups 14987.69 815 18.38980864 Total 17074.46 833
213 Table 4 9. Results of Tukey tests showing calculated q values for all pairs of Gafrarium. pectinatum valve length groups with significantly different means by cmbd, Bapot 1. The critical q value, or â€œStudentized rangeâ€ for the comparisons = 4.974. Depth (cmbd) 5.07 7.79 7.69 6.76 6.15 8.99 5.33 6.71 7.92 5.41 5.96 11.02 7.13 6.92 6.68 Table 4 10. Summary of data for single factor ANOVA of mean valve lengths for Gafrarium pectinatum grouped by date before present, Bapot 1. Groups Count Sum Average Variance 2000 cal a BP 36 920.2 25.56111 27.90930 2200 cal a BP 260 6725.5 25.86731 22.25310 3100 cal a BP 399 11298.4 28.31679 16.15497 3300 cal a BP 139 4059.4 29.20432 17.47592 Table 4 11. Results of Single factor ANOVA for mean Gafrarium pectinatum valve lengths by date before present, Bapot 1. Source of Variation SS df MS F P value F crit Between Groups 1492.731 3 497.57710 26.50469 2.61563 Within Groups 15581.730 830 18.77317 Total 17074.460 833
214 Table 4 12. Results of Tukey tests showing calculated q values for all pairs of Gafrarium pectinatum valve length (mm) groups with significantly different means by date before present, Bapot 1. The critical q value, or â€œStudentized rangeâ€ for the comparisons = 3.633. Date (cal a BP) 2000 cal a BP 2200 cal a BP 3100 5.169 10.031 3300 6.359 Figure 4 1. Valve measurements and terms. Dashed line marks where valve was sectioned for aging and solid line marks the total length of the shell. Figure 4 2. Acetate peels from three A nadara valves. White dot s mark opaque growth increments. A ) Catalog No. 102, Specimen No. 10, RMA=14. B ) Catalog No. 160, Specimen No. 24, RMA=7. C) Catalog No. 160, Specimen No. 6, RMA=4. A B C
215 Figure 4 3. Mean valve length for Anadara cf. antiquata from Bapot 1 wi th one standard deviation above and below the mean . A) plotted by cmbd . B ) P lotted by date before present (cal a BP). A B
216 Figure 4 4. Frequencies of Anadara cf. antiquata shell lengths based on 5mm length bins, Bapot 1.
217 Figure 4 5. Cumulative frequencies of Anadara cf. antiquata valve lengths based on 5 mm length bins, Bapot 1.
218 Figure 4 6. Mean valve length for Gafrarium pectinatum from Bapot 1 with one standard dev iation above and below the mean. A ) P lotted by cmbd . B ) P lotted by date before present (cal a BP). A B
219 Figure 4 7. Frequencies of Gafrarium pectinatum valve lengths based on 5 mm bins, Bapot 1.
220 Figure 4 8. Cumulative frequencies of Gafrarium pectinatum valve lengths base on 5 mm bins.
221 Figure 4 9. Estimated length at age according to date before present. Lines represent datespecific linear growth models.
222 Figure 4 10. Frequency distributions for Anadara cf. antiquata valves based on RMA grouped by early and recent periods, Bapot 1.
223 Figure 4 11. Cumulative frequencies of Anadara cf. antiquata based on RMA and grouped by older and recent periods, Bapot 1.
224 CHAPTER 5 GENERAL DISCUSSION AND CONCLUSIONS Introduction This study evaluated the complex interactions between prehistoric humans and mar ine environments in the Mariana Archipelago using historical, ethnographic, and environmental data. Zooarchaeological and fisheries sciencebased methods were used to analyze the remains of fish and, primarily, shellfish derived from archaeological context s. The goal of this project was to explain variations in marine faunal assemblages from two sites in the Mariana Islands and address questions regarding the role of marine resources in human settlement and the evolution of traditional Chamorro culture. The se questions included: 1. Can humaninduced vs. environmentally induced change be detected through the analysis of archaeological marine faunal samples? 2. Can the initial effects of human predation upon previously unexploited marine resources be detected? 3. Can thresholds of human interaction and related changes in subsistence strategies be detected in prehistoric fisheries? 4. What role (managers, engineers, or adaptive organisms) did indigenous peoples play in the human environment interaction throughout the prehistoric periods of the Mariana Islands? 5. How does traditional marine exploitation in the Mariana Islands compare to related cultures in Oceania? The questions above were addressed to expand upon current explanations of variation in faunal remain s recovered from archaeological sites throughout the Mariana Islands. Current explanations largely ascribe this variation to changes in prehistoric environmental conditions. More specifically, changes in the relative abundance of different species with dis similar habitat preferences are attributed to environmental changes related to a fall in sea level that coincided with the arrival of the first settlers in the Marianas (Amesbury 1999, 2007; Amesbury et al. 1996;
225 Carson 2012), and the destruction of aquati c habitats such as mangroves during later portions of the sequence. This sea level drop stems from the mid -Holocene high stand that reached its maximum level of approximately two meters above modern sea level about 4,200 years ago. It is argued that human settlement and population growth was facilitated by the expansion of productive coastal habitats created by falling sea levels (Dickinson 2000:744). Also, as populations grew, sensitive habitats such as mangroves were destroyed, leading to shifts in the ex ploitation of gastropods and bivalves. The first question concerned determining if humaninduced or environmentally induced change could be detected through the analysis of archaeological marine faunal samples. It was assumed, in the present study, that en vironmental factors influenced prehistoric marine habitats and cultural practices concurrently. It was expected that environmental factors would have had significantly different effects upon marine fish and shellfish populations than cultural practices. Sp ecifically, large-scale changes such as a two meter drop in sea level would cause significant longterm changes in mangrove, lagoon, and reef habitats and their inhabitants. These environmental changes, exerting pressures concurrent with human predation upon marine resources, would lead to a long term trend of resource degradation. This would be marked by decreases in mean length and changes in the age structure of certain fisheries, significant changes in species frequencies, and shifts in species diversit y. It was also expected that significant differences would be detected during transitional periods in the archaeological sequence, like the beginning of the Latte Period at 1000cal a BP. Whereas, lengthbased and age based analyses found statistically sign ificant differences between two species of bivalve from the Bapot 1 site shellfish assemblage, these differences were relatively small. Moreover, no evidence was found that instances of environmental change or human exploitation were intense enough to caus e
226 variations in prehistoric exploitation patterns during the occupational sequence of the Bapot 1 site. Also, comparisons of calculated invertebrate species diversity values from both sites indicated that species diversity oscillated between higher and low er values throughout the sequence . This suggests that marine habitats and the prehistoric human exploitation of these habitats were relatively stable through time. The most significant change in the archaeological residues of prehistoric marine resources was found in samples dating to around 2000 2200 cal a BP . This included differences in mean lengths of Anadara cf. antiquata and Gafrarium pectinatum shell and a significant change in growth rates of Anadara cf. antiquata shells from the Bapot 1 site, wher e shells from 2000 2200 cal a BP had faster growth than shells from 31003300 cal a BP . These changes appeared to be strongly related to cultural parameters and may reflect small scale exploitation practices, either by a small prehistoric human population or a smaller group within the population. Based on ethnographic data, it is also worth considering that early inhabitants of the Marianas may have practiced some form of resource management or conservation. It was expected in the present study that environmentally induced changes, such as sea level changes and consequent habitat loss or deg radation would have led to more substantial changes in species frequencies and mean length and growth rates of certain species as habitat degradation added stress to human harvesting pressures. The present study suggests that the drop in sea level from the mid Holocene high stand and its coincident effects on mangroves was a slow process. This allowed prehistoric peoples to adjust to slowly changing marine habitats. The second question was related to determining if the initial effects of human predation upon previously unexploited marine resources could be detected. To answer this question, fisheries sciencebased methods were employed to measure the density dependent effects of human exploitation upon prehistoric fisheries populations through the use of archaeological
227 residues. This also involved looking for significant changes in species frequencies, major changes in the mean lengths of Anadara cf. antiquata and Gafrarium pectinatum shells, and changes in growth rates in Anadara cf. antiquata shells in sam ples from the deepest stratigraphic levels, marking the beginning of intensive use of the Bapot 1 site. Changes in NISP frequencies were apparent in the earliest levels, including increases in gastropod and bivalve shell NISP in depths from 260200 cmbd and between 3300 and 3100 cal a BP . However, this was likely the result of increased deposition from more intensive use of the Bapot 1 site in the initial occupation period. This is because no differences in mean length of Anadara cf. antiquata and Gafrarium pectinatum shells or growth rates of Anadara cf. antiquata could be found between samples from the early part of the Pre-Latte Period (3100â€“3300 cal a BP). Also, a similar pattern was observed in materials recovered from the Ritidian Grotto site with incr eases in the frequencies of shell artifacts, lithics, pottery, and fish bone in the deepest stratigraphic layers. The third question asked if thresholds of human interaction and related changes in subsistence strategies could be detected in prehistoric fi sheries represented by archaeological residues. Here, thresholds are significant changes in the length and age parameters of a prehistoric fishery that can be linked to significant changes in artifact types, exploitation practices, or settlement patterns t hat coincide with the different phases of the Pre-Latte Period proposed by Moore (2002). Results of the present project found that mean lengths of both Anadara cf. antiquata and Gafrarium pectinatum from Bapot 1 from samples from 109122 cmbd were signific antly different from almost all other samples. This coincided with the middle portion of the Late Unai Phase of the Pre Latte Period (25001600 cal a BP ). Materials from the Bapot 1 site dating to this period are characterized by lower frequencies of artif acts, including pottery, lithics, and shell. Furthermore, changes in the frequencies of faunal remains were
228 apparent, including an increase in the frequency of fish bone and a general decrease in the frequency of invertebrate remains. There are also signi ficant differences in materials from the Ritidian Grotto site dating to Late Unai Phase of the Pre-Latte Period (1600â€“2500 cal a BP ). These include abrupt increases from the lowest frequencies of bird bone, fish bone, NISP of identified invertebrate specie s, and diversity index values in samples dating to 1800 cal a BP, to the highest frequencies and index values in samples dating to approximately 1500 1700 cal a BP . Lower frequencies of pottery were also recovered from samples dating to this period from bo th TU 3 and 4. The fourth question was aimed at discerning the role of prehistoric indigenous peoples within the human marine environment interaction. To answer this question, two contrasting viewpoints were considered. The first views early prehistoric peoples as subject to, or otherwise limited by, their environment. The second views prehistoric peoples as active agents, armed with extensive knowledge related to exploiting and managing marine resources. In regard to the first point of view, changes and de velopments in exploitation practices were necessitated by fluctuating environmental conditions, which might, for example, force prehistoric peoples to change exploitation patterns. This could include shifting the focus from a particularly abundant and easi ly collectable species, impacted by over harvesting, to a broader array of less abundant species. It was expected that archaeological evidence of this would include significant changes in NISP frequencies of various species, diversity of exploited species, mean length of Anadara cf. antiquata and Gafrarium pectinatum shells, and lengthat RMA parameters of prehistoric Anadara cf. antiquata populations. It was also expected that these could be linked to changes in frequencies of terrestrial vertebrate specie s and artifact types. In regard to the second viewpoint, expected results of fisheries based analyses included nonsignificant changes in mean length and
229 RMA through time. It was also expected that there would be non significant changes in species diversit y through time. Results of the present project, however, highlight the inherent complexity in the prehistoric human environment interaction. The NISP and distribution of fish bone in both the Bapot 1 and Ritidian assemblages shows that fish remains were mo re common in the later portions of the Pre -Latte Period. Significant changes in mean Anadara cf. antiquata and Gafrarium pectinatum shell length from the Bapot 1 site were found in samples dating to 2000 2200 cal a BP . Diversity indices for both the Bapot 1 and Ritidian Grotto sites were found to oscillate between higher and lower values throughout their occupational sequences. When combined, these sequences extend from the earliest phase of the Pre-Latte Period to after 1000 cal a BP, during the Latte Peri od. Whereas diversity values for gastropods are always higher than values for bivalves in the Ritidian Grotto site assemblage, values for gastropods and bivalves from the Bapot 1 site oscillate back and forth throughout the siteâ€™s occupational sequence. Wh ile results of fisheries based analyses of the two prehistoric shellfish fisheries showed significant differences changes in mean length and mean RMA frequencies, these appear to have been relatively minor changes that could not be linked to past environmental fluctuations or instances of overexploitation. Rather, these species appeared to represent a relatively stable resource that was not significantly impacted by prehistoric exploitation or land use practices. Furthermore, these data, coupled with the general lack of fish remains, especially in the Bapot 1 site assemblage, suggest that marine resource exploitation activities were not necessarily subject to environmental conditions. In terms of marine fisheries exploitation, it appears Pre -Latte peoples we re relatively unconstrained by environmental conditions and that explanations of variation in the archaeological residues in the Marianas would benefit from a closer consideration of human
230 agency and sociological issues related to marine exploitation pract ices. These might include the roles of gender, age, and class and traditional management practices. The fifth question asked â€œHow does traditional marine exploitation in the Mariana Islands compare to related cultures in Oceania?â€ This was posed to provid e important context and to aid in formulating alternative explanations regarding variation in the archaeological residues of prehistoric marine fisheries. Around 6,000 years ago, in the rich marine environments of the islands of Southeast Asia, Proto-Austr onesians began to develop fishing techniques and technologies that would be linked to the settlement of the Pacific Islands (Kirch and Dye 1979:5354). Because of this, it was expected that core similarities in marine resources exploitation practices would be found among island groups inhabited by Austronesian speakers. Based on this common ancestry, all of the traditional inhabitants of Pacific Islands share numerous similarities in regard to the types of materials they used to manufacture gear, strategies they employed to exploit fisheries, and beliefs associated with marine resource exploitation. Unfortunately, materials like plant fibers and wood that are commonly used to make traditional fishing gear are perishable. Also, colonialism blurred direct connections between the complexity of historic marine resource exploitation and archaeological evidence. Ethnographic information on this subject is also limited and has primarily focused upon material aspects rather than the relationships between technology, behavior, and the environment. This is especially true in the Marianas, where European colonialism was first established in the Pacific. Although early historical accounts of fishing and shellfish gathering in the Marianas are largely undetailed and biased , they highlight the complexity of Chamorro marine resource exploitation and certain practices that may offer more detailed explanations regarding the variability in archaeological faunal assemblages. Although Chamorro marine exploitation practices are uni que in many ways,
231 they also share basic similarities with those of other indigenous Pacific Islanders, which may stem from ancient and shared ancestral origins. One example is the harvesting of shellfish by women. This genderspecific practice was document ed relatively early in the chronicles of explorers who visited the Marianas and is prevalent throughout the Pacific. Unfortunately, ethnographic research regarding gender and age roles in marine resource exploitation is rare in Micronesia. Results of the present project suggest that a closer examination of gender or age related exploitation practices from modern, historic, and ethnographic sources would greatly enhance explanations regarding variability in the archaeological residues of prehistoric marine resource exploitation. There are various aspects of both the Bapot 1 and Ritidian Grotto site assemblages that may benefit from this. First is the low frequency of fish remains comprised of spines and vertebrae from small individuals. It is unlikely that f ish represented a less important prehistoric resource than shellfish or that the collection of fish from reefs adjacent to both sites was somehow limited by environmental conditions. Second are the relatively minor changes in mean lengths of Anadara cf. antiquata and Gafrarium pectinatum shell and growth rates of Anadara cf. antiquata from the Bapot 1 site. This indicated that prehistoric harvests were not intensive enough to cause significant changes in the population structure of these species. As these species appear to represent a relatively stable resource for almost 2,000 years, it is reasonable to assume that the impact of prehistoric harvests were constrained by small human populations, by exploitation limited to a smaller group within the main population such as women and children, or were subject to some form of traditional management. The application of zooarchaeological and fisheries sciencebased methods in this study enabled a better understanding of the complex interactions between past environmental variables
232 and cultural processes, and specifically, the roles these inter actions played in the initial settlement of the Marianas, how prehistoric peoples coped with environmental adversity, and the development of local marine exploitation practices. It is apparent that whereas local environments directed the exploitation of pr ehistoric marine resources, the environment was not a limiting factor for settlement, population growth, and the development of unique and complex socio cultural processes in the Mariana Archipelago. The general lack of significant changes in the population dynamics of at least two species of clams, significant changes in species diversity over time, and stark differences in marine resource use related to site type and location suggest that the prehistoric inhabitants of the Marianas were equally, if not mo re, influenced by cultural factors. They were not subjugated by environmental conditions that were unfamiliar, or particularly poor. Rather, settlement was initiated by voyagers with a set of cultural baggage, people who were pre adapted to the tropical marine environment of the Marianas. I summarize in this concluding chapter the empirical findings of the study and how they relate to human settlement, cultural evolution and the environment, discuss the theoretical implications of the study and the project â€™s limitations, and provide ideas regarding future avenues of research. Empirical Findings The empirical findings of the analyses related to historical, environmental, and archaeological data suggest that prehistoric marine environment/human interactions were inherently complex. This relationship began in the earliest occupation period, signaled by large accumulations of artifacts, shell, and bone on various beaches in the Mariana Archipelago. Ethnohistory and History Traditional knowledge, practices, and beliefs related to marine resource exploitation have been widely documented throughout the Pacific and it appears that similarities between island
233 groups generally outnumber the differences. This is attributable to multiple reasons, which include similari ties in marine habitats, resource types, and the fact that all Pacific Islanders stem from a common ancestral group (Kirch 2002). Cultural similarities among some archipelagoes were also reinforced by regular interactions between island groups during the prehistoric period. This was likely the norm, rather than isolation or low frequencies of interaction (Dâ€™Arcy 2006; Peterson 2009). The writings of early European explorers, historians, and anthropologists and archaeological data suggest that the techniques , technologies, knowledge, and beliefs surrounding traditional fisheries in the Marianas are very similar to those found throughout the Pacific. This is especially true among Micronesian groups. The results of the current project show that the patterns obs erved during the analysis of archaeological data from the Bapot 1 and Ritidian Grotto sites might be best explained in the context of traditional fishing and gathering practices, rather than as purely adaptive responses to changing environmental conditions . Some of these practices and beliefs, as described in historic texts by early European explorers, colonists, ethnographers and anthropologists, can still be observed today in various island groups throughout the Pacific, though to a lesser degree in the M arianas. Laboratory Analyses The results of basic analyses of the faunal assemblages and artifacts from Bapot 1 and the Ritidian Grotto sites generally agreed with previously documented patterns of prehistoric cultural development. Bone from marine vertebr ate and shell from invertebrate species confirmed the importance of marine resources in the prehistoric subsistence economy of the Marianas throughout the Pre -Latte and Latte Periods. The diversity of identified specimens of fish bone and marine invertebrate shell shows that all marine environments, especially inshore habitats, were fully exploited by the earliest inhabitants of the Marianas and were relatively stable throughout the sequences of both sites. This included 37 species of marine gastropods, two
234 species of terrestrial gastropod, 19 species of bivalve, four species of bony fish, and a shark from Bapot 1, Block A and 25 species of marine gastropod, two species of terrestrial gastropod, 20 species of bivalve, and two species of fish from the Ritidia n Grotto site. Inter and intra site variability in marine faunal remains from the two sites, such as the comparatively small assemblage of fish bone (n=834) recovered from approximately 22 cubic meters of soil from the Bapot 1, Block A excavations versus the relatively large assemblage of fish bone (n=4,673) recovered from approximately 2.2 cubic meters of soil from TU 3 and TU 4, Ritidian Grotto, suggests complex marine exploitation practices, both technological and sociological, were brought with the ini tial settlers and were well established early in the occupational sequence. The rich assemblage of shell artifacts, including beads in different stages of manufacture, fishhooks, scrapers, and adzes, also illustrates that shell was an important raw material used in the manufacture of functional items and adornments and that some species were exploited for reasons other than subsistence. Diversity An attempt was made to identify prehistoric trends utilizing frequencies of the number of identified marine invertebrate species and calculated Shannon Wiener diversity index values within the marine invertebrate assemblages from both sites through time. Previous studies equated declines and increases in different species within assemblages from other sites in the M arianas to changes in species preferences caused by the loss of natural habitat. In the Marianas this is specifically related to prehistoric shifts in the frequency of gastropod and bivalve species, notably a shift from the dominance of Anadara cf. antiquata in assemblages dating to the PreLatte Period to the dominance of gastropods in assemblages dating to the Latte Period at sites on Guam and Saipan. Because Anadara cf. antiquata in the Marianas is associated with mangrove habitats (Pauley 1992) that are sensitive to sea level change (Woodroffe et al. 1985), the decline
235 in this species has been equated with the probable loss of mangrove habitat because of declining sea level. This consequently led to increases in the frequency of gastropod species in samp le assemblages dating to the Latte Period (Amesbury 1999; Amesbury et al. 1995; Graves and Moore 1985; Pauley 1992). In general, frequencies of total identified invertebrate, gastropod, and bivalve species plotted by depth for the Bapot 1 and the Ritidian Grotto sites illustrate similar patterns. The number of identified species in each case fluctuated throughout the Pre -Latte Period in the same fashion, with two or more modes of increased frequency. However, values from the deepest excavation level were extremely low in the Ritidian Grotto assemblage and values for gastropods and bivalves from the Bapot 1 site oscillate back and forth, with the highest values for gastropods in the upper levels of the sequence. When plotted by date, frequencies of total invertebrates, gastropod, and bivalve species from the Bapot 1 site displayed similar patterns with a general increase in the frequency of species counts to a maximum at 2200 cal a BP and decline from 1500 to 2000 cal a BP. This is similar to species frequen cies plotted by cmbd that also decreased throughout the sequence. Shannon-Wiener Index values plotted by date for samples from Bapot 1 also showed that index values oscillated between higher and lower values and that values for gastropods and bivalves osci llated back and forth with higher values for gastropods and decreasing values for bivalves in the late portion of the sequence. Species frequencies and diversity index values for samples from the Ritidian Grotto site plotted by depth and date displayed si milar patterns to samples from the Bapot 1 site. Species counts and diversity index values plotted by excavation depth fluctuate throughout the stratigraphic sequence, like samples from the Bapot 1 site, however, index values for gastropods
236 from the Ritidi an Grotto site are always higher than values for bivalves. When plotted by date, species counts and diversity index values for total invertebrates, gastropods, and bivalves show the same patterns of a sharp increase in frequencies and diversity index values early in the sequence. For species frequencies, this is followed by a decrease and subsequent increase during the later portions of the sequence and for diversity values this is followed by oscillating high and low values throughout the later portions of the sequence. Within the shell samples recovered from TU 3 and TU 4, the numbers of identified species and calculated diversity index values show that gastropods are more diverse than bivalves. Fluctuating species counts suggest that habitat types and inv ertebrate species were exploited at different levels of intensity and for different purposes through time. The diversity index values oscillating around a mean suggest that gastropods and bivalves were preferentially harvested at different times in the seq uence, exploited invertebrate populations were stable through time, and invertebrates were not overexploited or impacted by environmental change. The Ritidian Grotto site shell data also agree with the bivalve/gastropod shift discussed above, that occurred around 1000 cal a BP in western Saipan (Amesbury 1999:358; Amesbury et al. 1996:56). However, while dominant bivalve species become less frequent, they were not completely replaced by a particularly dominant species of gastropod like Strombus gibberulus gibbosus , which is so prevalent in Latte Period assemblages at other sites in Guam and Saipan (Amesbury 1999). Also, results of the present study suggest this shift began earlier at the Bapot 1 site, beginning around 2000 cal a BP and was complete by around 1500 cal a BP. This coincides remarkably well with paleoenvironmental data from cores collected in Lake Susupe, located on the southwest coast of Saipan. These indicated that during the Mid Holocene high sea level stand between 1500 and 5123cal a BP, mang rove pollen was dominant. Also, frequencies
237 of mangrove pollen track the expansion and contraction of mangrove habitat, which also traces the rise and fall of the Mid Holocene high sealevel stand that reached its present level at 1500 cal a BP (Athens and Ward 2005:51). Variation in this bivalve/gastropod shift between sites on Guam and Saipan has been well documented. Although there were sharp declines in the abundance of Anadara cf. antiquata in the marine invertebrate remains from sites at Tumon Bay, in northern Guam, and on Saipan during the Latte Period, at sites in southern Guam, bivalve shells are abundant in marine invertebrate assemblages throughout the Latte Period. These variations in the abundance of Anadara shell in archaeological assemblages are related to geologic differences between northern Guam (similar to Saipan), which consists of a limestone plateau lacking surface water, and southern Guam, which consists of volcanic cuesta with river drainages (Amesbury 1999). Deltaic or estuarine areas , followed by embayments and lagoons, are the optimal environments for the establishment and growth of mangroves in the Pacific (Woodroffe 1987:181). Both of these environment types were likely common in southern Guam following the stabilization of sea lev els to their modern position at 1500 cal a BP, which would explain the abundance of bivalves like Anadara cf. antiquata in some shell assemblages from sites in southern Guam during the Latte Period. Gastropod shell is more abundant than bivalves in the mar ine invertebrate assemblage from the Ritidian Grotto site throughout the period of occupation as defined in TU 3 and 4. Also, all of the identified gastropod and bivalve species, aside from Anadara cf. antiquata, from this site prefer hard, rocky, or sandy substrates. Only four Anadara cf. antiquata shells were recovered during excavations of TU 3 and TU 4. This near absence of Anadara shell reflects the environmental conditions of the Ritidian Grotto site, which is characterized by narrow fringing reefs.
238 Length and Age Analyses of mean lengths of Anadara cf. antiquata and Gafrarium pectinatum shell from Bapot 1, Block A through time showed similar patterns. When plotted according to depth below datum, mean shell lengths of Anadara cf. antiquata and Gafrarium pectinatum appear to generally decrease throughout the sequence, first with an increase in shell length in samples from 109122 cmbd and 7896 cmbd, followed by an abrupt decrease in shell length in samples from 6585 cmbd. Tukey tests for significant differences in mean shell length for all possible pairs of Anadara cf. antiquata and Gafrarium pectinatum samples showed that the sample from 109 122 cmbd was significantly different from samples from most other excavation levels. Shells from this de pth date to approximately 2200 cal a BP. Mean shell lengths from Anadara cf. antiquata and Gafrarium pectinatum samples plotted by date also showed a general decline in mean length through time with Anadara cf. antiquata and Gafrarium pectinatum shell, reaching their minimum mean lengths after 2200 cal a BP. Tukey tests for significant differences in mean shell length for all possible pairs of mean length samples for Anadara cf. antiquata and Gafrarium pectinatum , combined according to date, generally showed that samples from 2000 cal a BP were significantly different from the rest of the dated samples. Using the simple linear regression models to estimate RMA from an aged subsample of shell showed that there was a significant difference in age f requencies and growth rates between samples from the early part of the sequence (3100â€“3300 cal a BP) and the late part of the sequence (2000â€“2200cal a BP), with shell samples from the later portion of the sequence growing faster. Also, frequencies of estim ated RMA at length values, show that there is a higher percentage of younger Anadara cf. antiquata shells in samples from the earlier periods and a higher percentage of older shells in samples from the later periods. This contradicts expected
239 outcomes of overharvesting and suggests that Anadara cf. antiquata represented a stable and abundant resource throughout the Pre -Latte Period. Comparative Analysis, Bapot 1 and Ritidian Grotto Remains Differences between marine faunal assemblages from the Bapot 1 and Ritidian Grotto sites further highlight the dynamic relationship between humans and the environment. Important differences include varying environmental conditions, site type, settlement sequences, and the frequencies of identified species within the ass emblages. The environmental factors at each site undoubtedly influenced prehistoric uses, species collected, and amounts of shell and bone recovered during excavations. The inshore marine environment associated with the Ritidian Grotto site has probably changed little since prehistoric occupation ca . 1800 cal a BP. This is evident from the identified species of fish and shellfish recovered from TU 3 and 4. The Bapot 1 site however, has likely changed significantly since prehistoric occupation between 3300 and 1400 cal a BP. This probably involved the early presence of mangrove, near or around the site, which diminished as sea level dropped. Although the environment limited the range of species that could be exploited, patterns in the proportional frequencies and distributions of shell and bone within the assemblages of these sites was likely dictated by behaviors related to exploitation. The most extreme cases for this are the disproportionately small amount of fish bone recovered from the Bapot 1 site vs. th e relatively high frequency of fish bone from the Ritidian Grotto site and the large amounts of Anadara cf. antiquata and Gafrarium pectinatum recovered from the Bapot 1 site vs. the almost complete absence of these species from the Ritidian Grotto site. These patterns may reflect differences in collection strategies, processing and transport (Bird et al. 2002), or genderand age related exploitation practices, if one were to accept historical and modern practices as appropriate analogies. Historically and ethnographically, shellfish have been, and continue to be gathered by women and children in
240 traditional Pacific Island societies. Some gastropod species, such as Cellana radiata orientalis (dominant in the assemblage of both sites), Pythia scarabaeus (the most common gastropod recovered from TU 3 and TU 4, Ritidian Grotto), and Strombus mutabulis (recovered from both the Bapot 1 and Ritidian Grotto sites), are commonly gathered by children. Shells from bivalves including Modiolus spp. and Tridacna spp. and bones from parrotfish (recovered from both sites), and bones from the flightless rail (recovered from the Bapot 1 site), are also often exploited by women and children throughout the Pacific. Implications The focus on evaluating pre Contact marine resour ce exploitation, particularly marine mollusks in the present study, was predicated upon various aspects of recent archaeological research and related work aimed at reconstructing past environmental conditions. A lot of this work was concerned with the init ial settlement of the Marianas and focused primarily upon dating deposits containing the earliest types of thin red slipped pottery with taletell decorations attributed to the initial settlers of the islands. Although that work was productive, adding much to the current understanding of the pre Contact period in the Marianas, detailed considerations of archaeological faunal materials remain limited and various issues with current interpretations of settlement and prehistoric cultural change remain to be resolved. Issues of particular concern include early radiocarbon dates and pollen data derived from sediment cores that suggest a human presence in the Marianas by 4200 cal a BP, a clear understanding of the effects of an approximate two meter drop in sea level in the Marianas to present levels at 1500 cal a BP, and the complexity of marine shell assemblages analyzed during the current study. Because these early dates from sediment cores were not directly associated with pottery and artifact bearing sites, th ey have been essentially discounted, although they provide compelling evidence of a preceramic period in Remote Oceania that may be characterized by a
241 long period of unique economic and social practices, specifically associated with the process of island colonization. This would presumably entail the manipulation of unique environmental circumstances. Similar paleoenvironmental evidence has also been collected for other island groups in the western Pacific and shows similar lag times between accumulations of charcoal and pollen regime changes, attributed to anthropogenic burning and landscape modification at around 4200 cal a BP and appearance of the earliest sites containing artifacts, shell, and bone at around 3500 cal a BP. Whereas pottery is a key element in discussions of the expansion of Austronesian speaking peoples into the Pacific, it is not necessarily a technology whose absence would inhibit voyaging, colonization, or the development of extremely stratified societies. These early paleo core dates also underscore complex humanenvironment interactions, concerning changes in inshore marine ecosystems (particularly mangroves), sea level change, and marine resource exploitation patterns. This intense focus on early pottery deposits, aside from narrowing alternative views regarding the presence of humans in the Marianas prior to the manufacture of pottery, has also narrowed considerations of early humanenvironment interaction. In the current scheme, early Pre -Latte settlers were hedged onto narrow beach strands by higher sea levels and cultural development was largely driven by changes in the natural environment and human induced degradation of certain resources. The present study has furthered current understandings regarding the complexities related t o human exploitation of marine resources. Theoretical implications of this study suggest that integrating broader considerations of available data will enable more detailed descriptions of specific economic practices during the past and whether they changed or remained static. Through understanding the complexity inherent within specific economic practices, such as collecting shellfish, a more complete understanding of events important to regional, as well as
242 local issues, including the process of initial s ettlement, interaction between island groups and communities, and the development of unique island societies, can be reached. Lastly, broader considerations of the early humanenvironment interaction that include traditional systems of resource management could also greatly expand current interpretations of island settlement, past marine resource exploitation practices, and the development of several unique and complex island cultures. Likewise, traditional management systems may prove to be more effective than western concepts in addressing modern environmental problems. This is because traditional systems are based on stores of traditional knowledge attuned to local conditions, rely upon community based processes in developing best practices and regulatio ns, are based on communal property which is controlled and regulated by an identified group, and are based upon a relationship of coexistence (Berkes 2008). Limitations and Future Research The application of aging methods and analyses used in fisheries science for the present study represents preliminary research and issues regarding the size of the aged sample of Anadara shell should be expanded. Also, conducting age based analyses on other species of bivalves such as Gafrarium pectinatum and species of abundant gastropods such as Cellana radiata orientalis could provide a more detailed understanding of the changes in exploitation between bivalves and gastropods. The consideration of ecological factors and specific methods for aging shell from each species has to be employed. Future plans include developing these methods for different species in the Bapot 1 site assemblage to address questions regarding sex and age related exploitation practices, traditional conservation practices, environmental change, an d overexploitation. Radiocarbon dates were another limitation for the study, especially for the Ritidian Grotto site. TU 3 and 4 relied upon dates derived from similar stratigraphic layers in TU 1.
243 Excavations of TU 3 and 4 were conducted to highlight the complexity of marine resource faunal assemblages between different site types and periods of occupation. Priorities for future research on materials from the Ritidian Grotto site should include radiocarbon analyses and, like the fragmented marine vertebrat e bone assemblage from the Bapot 1 site, more detailed identifications of fish bone. This would require collection of locally derived comparative specimens of fish and would represent a timeconsuming task, as modern fish stocks in the Marianas have been extensively overfished. Although many radiocarbon dates were obtained from the Bapot 1 site, there is a gap of almost 1,000 years in the sequence, between 140 and 160 cmbd. Although this may represent a long hiatus in site occupation, more samples from thes e depths should be submitted for analysis. The last limitation to the present study is the general lack of detailed historic and ethnographic information specifically pertaining to shellfish gathering in the Marianas. This is also true for other island groups in Micronesia, as well as Melanesia and Polynesia. The present study relied upon historical colonial accounts of life in the Marianas and ethnohistoric accounts from throughout Oceania, and recent research pertaining to shellfish exploitation in Austra lia (Meehan 1982) and the Mer Islands in the Torres Strait, which are between Indonesia and Australia (Bird and Bliege Bird 1997, 2000; Bird et al. 2002). Historical accounts of the Marianas provided only very general information regarding shellfish exploi tation and underscore the pervasive effects of the long history of colonial occupation, which eliminated the indigenous Chamorro population of the Marianas through warfare and mixing with immigrants from Spain, Mexico, and the Philippines. Recent ethnograp hies on shellfish gatherers have highlighted important differences between contemporary and prehistoric accumulations of shell, modern and ancient exploitation practices (Chapman 1987; Meehan 1982), and the importance of various
244 shellfish species to tradit ional diets (Haines 1979, 1982; Kinch and Bagita 2003; Rawlinson et al. 1995). Variation was shown to be the product of collection and transport strategies where procurement locations were different from where items were consumed (Bird et al. 2002:467). Al though shellfish gathering is not presently a common traditional marine resource procurement practice in the Marianas, consideration of invertebrate exploitation on neighboring island groups like Palau, where species like Anadara cf. antiquata are still gathered and consumed, may provide important data from which more appropriate analogies could be drawn.
245 LIST OF REFERENCES Abbott, Tucker R. and Peter S. Dance 2000 Compendium of Seashells . Odyssey Publishing, Carwoola, NSW, Australia. Adams, Tim, Paul Dal zell, and Esaroma Ledua 1999 Ocean Resources. In The Pacific Islands, Environment and Society , edited by Moshe Rappaport, pp. 366 381. The Bess Press, Honolulu. Adams, J. D., C. Denfeld, and M. J. Tomonari Tuggle 1996 Test Excavation of a World War II Japanese Mass Grave at Achugao, Saipan, Commonwealth of the Northern Mariana Islands . Prepared for Division of Historic Preservation, Commonwealth of the Northern Mariana Islands and Division of Health and Welfare, Japanese Government. International Archaeolog ical Research Institute, Inc., Honolulu. Allen, Melinda, S. 2006 New Ideas about Late Holocene Climate Variability in the Central Pacific. Current Anthropology . 47:521 535. Allen, Gerald, Roger Steele, Paul Humann, and Ned Deloach 2003 Reef Fish Identification, Tropical Pacific. New World Publications Inc., Jacksonville, Florida. Allen, Gerry 2000 Marine Fishes of the Great Barrier Reef and South -East Asia . Western Australian Museum, Perth, Australia. Alongi, Daniel M. 2002 Present State and Future of the Worldâ€™s Mangrove Forests. Environmental Conservation 29(3):331â€“349. Ambrose, Jearld, Jr. 1983 Age Determination. In Fisheries Techniques , edited by L. A. Neilsen, D. L. Johnson, and S. S. Lampton, pp. 301â€“324. American Fisheries Society, Beth esda Maryland. Ambrose, William 1967 Archaeology and Shell Middens. Archaeology and Physical Anthropology in Oceania 2:169 187. Amesbury, Judith R. 1999 Changes in Species Composition of Archaeological Marine Shell Assemblages in Guam. Micronesica 31(2):347 366. 2007 Mollusk Collecting and Environmental Change During the Prehistoric Period in the Mariana Islands. Coral Reef 26:947 958.
246 Amesbury, Judith R. and Rosalind. L. Hunter-Anderson. 2003 Review Of Archaeological And Historical Data Concerning Reef Fi shing In The U.S. Flag Islands Of Micronesia: Guam And Northern Mariana Islands . Final report prepared for Western Pacific Fishery Management Council, Honolulu. Micronesian Area Research Services, Guam. Amesbury, Judith R., Darlene R. Moore, and Rosalind L . Hunter -Anderson 1996 Cultural Adaptations and Late Holocene Sea Level Change in the Marianas: Recent Excavations at Chalan Piao, Saipan, Micronesia. Bulletin of the Indo -Pacific Prehistory Association 15:53 69. Amesbury, Steven S., Frank A. Cushing, and Richard Sakamoto 1986 Guide to Coastal Resources on Guam: Vol. 3, Fishing on Guam . University of Guam Marine Laboratory, Contribution Number 225. University of Guam Press, Guam. Anderson, M. Kat. 2005 Tending the Wild: Native American Knowledge and the Management of Californiaâ€™s Natural Resources . University of California Press, Oakland, California. Anderson, Richard O., and Robert M. Nueman. 1996 Length, Weight, and Associated Structural Indices. In Fisheries Techniques, Second Edition , edited by Brian R. Murphy and David W. Willis, pp. 447482. American Fisheries Society, Bethesda, Maryland. Arnold, Jeanne E. 1992 Complex Hunter-Gatherer -Fishers of Prehistoric California: Chiefs, Specialists, and Maritime Adaptations of the Channel Islands. American Antiquity 57(1):6084. Athens, Stephen J., Michael F. Dega, and Jerome V. Ward 2004 Austronesian Colonization of the Marianas Islands: The Paleoenvironmental Evidence. Indo-Pacific Prehistory Association Bulletin 24 (2): 21â€“30. Athens, Stephen J. and Jerom e V. Ward 1997 Hawaiian Native Lowland Vegitation in Prehistory. In Historical Ecology in the Pacific Islands , edited by Patrick V. Kirch and Terry L. Hunt, pp. 248â€“270. Yale University Press, New Haven, Connecticut. 1999 Paleoclimate, Vegetation, and Land scape Change on Guam: The Laguas Core. Excerpted from Archaeological Inventory Survey of the Sasa Valley and Tenjo Vista Fuel Tank Farms, Piti District, Territory of Guam, Mariana Islands , by Boyd Dixon, J. Stephen Athens, Jerome V. Ward, Tina Mangieri, Ti mothy Rieth. International Archaeological Research Institute, Inc., Honolulu. 2004 Holocene Vegetation, Savanna Origins and Human Settlement of Guam. Records of the Australian Museum 29:15 30.
247 2005 Holocene Paleoenvironment of Saipan: Analysis of a Core fr om Lake Susupe. Micronesian Archaeological Survey, Report Number 35. CNMI Division of Historic Preservation, Department of Community and Cultural Affairs, Saipan. Bale, William 1998 Advances in Historical Ecology: Introduction. In Advances in Historical E cology , edited by William Bale, pp. 1 â€“10. Columbia University Press, New York. Bale, William and Clark L. Erickson 2006 Time and Complexity in Historical Ecology. In Time and Complexity in Historical Ecology, Studies in the Neotropical Lowlands , edited by William Bale and Clark L. Erickson, pp. 1â€“17. Columbia University Press, New York. Barratt, Glynn 2003 The Chamorros of the Mariana Islands, Early European Records, 1521-1721. Occasional Historical Papers Series, No. 10, CNMI Division of Historic Prese rvation. Bath, Joyce E. 1986 The San Vitores Road Project Part 1: Final Report . Unpublished report, Pacific Studies Intitute, Guam. Bellwood, Peter 1979 Manâ€™s Conquest of the Pacific, The Prehistory of Southeast Asia and Oceania. Oxford University Press, N ew York. 1995 Austronesian Prehistory in Southeast Asia: Homeland, expansion and Transformation. In The Austronesians: Historical and Comparative Perspectives , edited by Peter Bellwood, James J. Fox, and Darrell Tyron, pp. 103 118. Australian National Uni versity, Canberra. 1997 Prehistory of the Indo -Malaysian Archipelago. University of Hawaiâ€˜i Press, Honolulu. 2002 Farmers, Foragers, Languages, Genes: the Genesis of Agricultural Societies. In Examining the Farming -Language Dispersal Hypothesis , edited by Peter Bellwood and Collin Renfrew, pp. 17â€“28. McDonald Institute for Archaeological Research, Cambridge. 2004 First Farmers, The Origins of Agricultural Societies . Blackwell Publishing, Boston. Bellwood, Peter, James J. Fox, and Darrell Tyron 1995 Chapter 1. The Austronesian in History: Common Origins and Divers Histories. In The Austronesians, Historical and Comparative Perspectives . Edited by Peter Bellwood, James J. Fox, and Darrell Tryon. ANU E Press.
248 von Bennigsen, Rudolph 2003 The German Annexation of the Caroline, Palau, and Mariana Islands . Translated by. Dirk H.R Spennemann. Division of Historic Preservation, Saipan, CNMI. Berkes, Fikret 2008 Sacred Ecology, Second Edition. Routledge, New York. Bird, Douglas W., Jennifer L. Richardson, Peter M. Veth, and Anthony J. Barham 2002 Explaining Shellfish Variability in Middens on the Meriam Islands, Torres Strait, Australia. Journal of Archaeological Science 29:457â€“469. Bird, Douglas W. and Rebecca Bliege Bird 1997 Contemporary Shellfish Gathering Strat egies Among the Meriam of the Torres Strait Islands, Australia: Testing Predictions of a Central Place Foraging Model. Journal of Archaeological Science 24:39â€“ 63. 2000 The Ethnology of Juvenile Foragers: Shellfishing Strategies Among Meriam Children. Journal of Anthropological Archaeology 19:461â€“476. Binford, Lewis 1968 Post Pleistocene Adaptations. In New Perspectives in Archaeology, edited by S. Binford and L. Binford, pp. 313â€“341. Aldine Publishing, Chicago. Blust, Robert 1995 The Prehistory of Austrones ian Speaking Peoples: A View from Language. Journal of World Prehistory 9: 453â€“510. 2000 Chamorro Historical Phonology. Oceanic Linguistics 39:83â€“122. 2002 The History of Faunal Terms in Austronesian Languages. Oceanic Linguistics 41:89â€“139. Bodner, Connie C. 1997 On Architecture and Social Power: Some Possible PhilippineOceanic Links. In Indo-Pacific Prehistory: Chiang Mai Papers , edited by Peter Bellwood and D. Tillotson, pp. 89â€“102. Indo Pacific Prehistory Association, Australian National University, Ca nberra. Bolle, Loes J., Adriaan D. Rijnsdorpa, Wim van Neerb, Richard S. Millnerc, Piet I. van Leeuwena, Anton Ervynckd, Richard Ayersc, and Ellen Ongenaee 2004 Growth Changes in Plaice, Cod, Haddock And Saithe In The North Sea: A Comparison Of (Post ) Me dieval And Present Day Growth Rates Based On Otolith Measurements. Journal of Sea Research 51:313â€“328. Braudel, Fernand 1980 On History . University of Chicago Press, Chicago.
249 Bronk Ramsey, C. 2008 Deposition Models for Chronological Records. Quaternary Science Reviews 27(12):42â€“60. 2013 OxCal radiocarbon calibration program, version 4.2.2. University of Oxford Radiocarbon Unit, Oxford. http://www/rlaha.ox.ac.uk/O/oxcal.php. Bronk Ramsey, C. and Lee, S. 2013 Recent and Planned Developments of the Program OxCal. Radiocarbon 3):720â€“730. Broom, M. J. 1985 The Biology and Culture of Marine Bivalve Mulluscs of the Genus Anadara. International Center for Living Aquatic Resources Management, Manila, Philippines. Broughton, Jack M. and James F. Oâ€™Connell 1999 On Evolutionary Ecology, Selectionist Archaeology, and Behavioral Archaeology. American Antiquity 64(1):153â€“165. Broughton, Jack M., Michael D. Cannon, Frank E. Bayham, and David A. Byers 2011 Prey Body Size and Ranking in Zooarchaeology: Theory, Empirical Evidence, and Applications from the Northern Great Basin. American Antiquity 76(3):403â€“428. Butler, Brian M. 1994 Early Prehistoric Settlement in the Marians Islands: New Evidence from Saipan. Man and Culture in Oceania 10:15â€“38. Carpenter, Kent E. and Volker H. Niem (editors) 1998 FAO Species Identification Guide for Fisheries Purposes, the Living Marine Resources of the Western Pacific, Vol. 1. Food and Agriculture Organization of the Un ited Nations, Rome. Carson, Mike T. 2008 Refining Earliest Settlement in Remote Oceania: Renewed Archaeological Investigation at Unai Bapot, Saipan. Journal of Island and Coastal Archaeology 3:115â€“139. 2010 Radiocarbon Chronology with Marine Reservoir Correction for the Ritidian Archaeological Site, Northern Guam. Radiocarbon 52:1627â€“1638. 2011 Paleohabitat of First Settlement Sites 1500 â€“1000 B.C. in Guam, Mariana Islands, Western Pacific. Journal of Archaeological Science 38:2207â€“2221. 2012 Evolution of a n Austronesian Landscape: The Ritidian Site in Guam. Journal of Austronesian Studies 3(1):55â€“86. 2012a An Overview of Latte Period Archaeology. Micronesica
250 Carson, Mike T. and David Welch 2005 Archaeological Survey, Mapping, And Testing of the Bapot Latte Site (SP-1-0013) in Laulau, Saipan, Commonwealth of the Northern Mariana Islands . International Archaeological Research Institute, Inc., Honolulu. Carson, Mike T. and Hiro Kurashina 2012 Re Envisioning Long-Distance Migration: Early Dates in the Mariana Islands. World Archaeology 44(3):409â€“435. Carucci, James. 1993 The Archaeology of the Orote Peninsula: Phase I and II Archaeological Inventory Survey of Areas Proposed for Projects to Accommodate Relocation of Navy Activities from the Philippin es to Guam, Mariana Islands . International Archaeological Research Institute, Inc., Honolulu. Cernohorsky, Walter O. 1972 Marine Shells of the Pacific, Vol. II. Pacific Publications, Sydney, Australia. Chapman, Margaret D. 1987 Womenâ€™s Fishing in Oceania. Human Ecology 15(3):267â€“288. Chernohorsky, Walter O. 1972 Marine Shells of the Pacific, Vol. II. Pacific Publications, Sydney, Australia. Clark, Geoffrey, Fiona Petchy, Olaf Winter, Mike Carson, and Patrick Oâ€™Day. 2010 New Radiocarbon Dates from the Bapot 1 Site in Saipan and Neolithic Dispersal by Stratified Diffusion. Journal of Pacific Archaeology 1(1):2135. Claassen, Cheryl 1996 A Consideration of the Social Organization of the Shell Mound Archaic. In Archaeology of the Mid -Holocene Southeast edited by Kenneth E. Sassaman and Florida. 1998 Shells . Cambridge Manuals in Archaeology, Cambridge University Press, Cambridge. Cloud, P.E. Jr., R.G. Schmidt and H.W. Burke 1956 Ge ology of Saipan Part I: General Geology. U.S. Geological Survey Professional Paper 280-A. U.S. Geological Survey, Washington D.C. Coomans, Fr. Peter 2000 History of the Mission in the Marianas Islands: 1667-1673. Translated and edited by Rodrique Lvesque. Occasional Historical Papers Series No. 4, CNMI Division of Historic Preservation, Saipan, CNMI.
251 Cota, Michael 2008 Varanus Indicus and Its Presence on the Mariana Islands: Natural Geographic Distribution vs. Introduction. Biawak (2)1:18 27. Cowie, Robert H. 2000 Nonindigenous Land and Freshwater Mollusks in the Islands of the Pacific: Conservation Impacts and Threats. In Invasive Species in the Pacific: A Technical Review and Draft Regional Strategy, edited by G. Sherley, pp. 143 172. South Pacific Regio nal Environmental Programme, Aipa, Samoa. Craddock, Elfriede 1982 Life in the Northern Mariana Islands during the German Administration. Commonwealth Council for Arts and Culture, Saipan, CNMI. Craib, John L. 1993 Early Occupation at Unai Chulu, Tinian, Commonwealth of the Northern Mariana Islands. Bulletin of the Indo -Pacific Prehistory Association 13:116 134. 1999 Colonisation of Marina Islands: New Evidence and Implications for Human Movements in the Western Pacific. In The Pacific From 5000 to 2000 BP, Colonisations and Transformations , edited by J. C. Galipaud, pp. 477â€“485. Institute de Recherche pour le Dvelopment, Paris. Craib, John L. and Ann K. Yaklavich 1996 Final Report, Cultural Resources Management Overview Survey, U.S. Naval Activities, Guam Waterfornt Annex, Mariana Islands, Terretory of Guam in Conjunction with Department of Defense Legacy Resource Management Program: Project #70. Ogden Environmental and Energy Services Co., Inc., Honolulu. Crumley, Carole L. 1994 Historical Ecology, A Mult idimensional Ecological Orientation. In Historical Ecology, Cultural Knowledge and Changing Landscapes , edited by Carol L. Crumley, pp. 1â€“16. School of American Research Press, Santa Fe, New Mexico. Cushing, David H. 1981 Fisheries Biology, A Study in Popu lation Dynamics, Second Edition. The University of Wisconsin Press, Madison, Wisconsin. Dâ€™Arcy, Paul 2006 The People of the Sea, Environment, Identity, and History in Oceania. University of Hawaiâ€˜i Press, Honolulu. Davidson, J. M. and B. F. Leach 2001 The Strandlooper Concept and Economic Naievte. In The Archaeology of Lapita Dispersal in Oceania , edited by G. Clark, A. J. Anderson, and T. Sorovi Vunidilo. Terra Australis 17:115â€“124.
252 Diamond, Jerad and Peter Bellwood 2003 Farmers and Their Languages: The First Expansions. Science 300: 597 603. Dickinson, William R. 2000 Hydro isostatic and Tectonic Influences on Emergent Holocene Paleoshorelines in the Mariana Islands, Western Pacific Ocean. Journal of Coastal Researc h 16: 735â€“746. Dickinson, William R., Brian M. Butler, Darelene R. Moore and Marilyn Swift 2001 Geologic Sources and Geographic Distribution of Sand Tempers in Prehistoric Potsherds from the Mariana Islands. Geoarchaeology 16:827â€“854. Dilli BradleyJ., E. M elanie Ryan, L.Wade Workman, and Alan E. Haun 1993 Archaeological Monitoring and Limited Data Recovery, Royal Palm Resort, Tumon, Tamuning Municipality, Territory of Guam . Paul H. Rosendahl, Inc., Agaa, Guam. Donaldson, Terry J. 2002 High Islands Versus Low Islands: A Comparison of Fish Faunal Composition of the Palau Islands. Environmental Biology of Fishes 65:241 248. Donaldson, Terry J. 1996 Fishes of the Remote Southwest Palau Islands: A Zoogeographic Perspective. Pacific Science 50:285 308. Donohue, Mark, and Tim Denham 2010 Farming and Language in Island Southeast Asia, Reframing Austronesian History. Current Anthropology 51(2):223 256. Driver, Marjorie G. 1993 The Account of Fray Juan Pobreâ€™s Residence in the Marianas, 1602. Micronesian Area Research Center, University of Guam, Mangilao, Guam. Dye, Thomas S. and Ken Longenecker 2004 Manual of Hawaiian Fish Remains Identification Based on Skeletal Reference Collection of Allan C. Ziegler and Including Otoliths . Special Publication 1., Society for H awaiian Archaeology. Emory, K. O. 1962 Marine Geology of Guam, Geology and Hydrology of Guam, Marianas Islands . Geological Survey Professional Paper 403 -B, A study of the topography and sediments of submerged terraces, fringing reefs, channels, and the l agoon, and various shore features of a tropical island. United States Government Printing Office, Washington, D.C. Erlandson, Jon M. 1988 The Role of Shellfish in Prehistoric Economies: A Protein Perspective. American Antiquity 53(1):102 109.
253 Farrell, Don A. 1991 History of the Northern Mariana Islands . Public School System, Saipan, CNMI. Fay -Sauni, Veikila Vuki, Susan Paul, and Marica Rokosawa 2008 Womenâ€™s Subsistence Fishing Supports Rural Households in Fiji: A Case Study of Nadoria, Viti Levu, Fiji. SPC Women in Fisheries Bulletin 18:2629. Freycinet, Louis Claude de 2003 An Account of the Corvette Lâ€™Uraineâ€™s Sojourn at the Mariana Islands, 1819. Supplemented with the Journal of Rose de Freycinet, Translated by Glynn Barratt. Occasional Historical Papers No. 13, CNMI Division of Historic Preservation, Saipan, CNMI. Fritz, George 2001 The Chamorro, A History and Ethnography of the Mariana Islands . Translated by Elfriede Craddock. Edited by Scott Russell. Occasional Historical Papers Series, No. 1. CNMI Division of Historic Preservation, Saipan, CNMI. Fulmer, John, Mellisa Kirkendall, David DeFant, and Alan E. Haun 1999 Archaeological Subsurface Testing, Sumay Village Site, Fort Santiago Site, and the Orote Point Cave Site, US Naval Activities (NAVACTS) W aterfront Annex, Santa Rita Municipality, Territory of Guam . Paul H. Rosendahl, Ph.D., Inc., Hilo, Hawaiâ€˜i. Gailey, Harry 1988 The Liberation of Guam, 21 July â€“ 10 August 1944. Presidio Press, Novato, California. Gingerich, S.B. 2003 Hydrologic Resources o f Guam. Water -Resources Investigation Report 03 -4126. U.S. Geological Survey, Honolulu. Graves, Michael, W. 1986 Organization and Differentiation Within Late Prehistoric Ranked Social Units, Mariana Islands, Western Pacific. Journal of Field Archaeology 13:139 154. 1991 Architectural and Mortuary Diversity in Late Prehistoric Settlements at Tumon Bay, Guam. Micronesica 24(2): 169 194. Graves, Michael, W. and Darlene Moore 1985 Tumon Bay Overview, Cultural and Historical Resources . Prepared for Department of Parks and Recreation, Agana, Guam, by Micronesian Area Research Center, Department of Anthropology, University of Guam, Mangilao, Guam. Graves, Michael W., Terry L. Hunt, and Darlene Moore 1991 Ceramic Production in the Mariana Islands: Explaining Chang e and Diversity in Prehistoric Interaction and Exchange. Asian Perspectives , 24(2):211 233.
254 Grayson, Donald K. 1973 On the Methodolgy of Faunal Analysis. American Antiquity 39(4):432â€“439. Greene, R. 1991 Near and Remote Oceania: Disestablishing â€œMelenesia.â€ in Culture History. In Man and a Half: Essays in Pacific Anthropology and Ethnology in Honour of Ralph Bulmer , edited by A. Pawley, pp. 491â€“502. The Polynesian Society, Aukland. Gosliner, Terrence M., David W. Behrens, and Gary C. Williams 1996 Coral Reef Animals of the Indo -Pacific. Sea Challengers. Monterey, California. Haddon, Malcolm. 2001 Modeling and Quantitative Methods in Fisheries . Chapman and Hall/CRC. Boca Raton, Florida. Haines, A. 1979 The Subsistence Fishery of the Purari Delta. Science in New Guinea 6(2):80â€“95. 1982 Traditional Concepts and Practices and Inland Fisheries Management. In Traditional Conservation in Papua New Guinea: Implications for Today , edited by L. Mourata, J. Pernetta, and W. Heaney. Institute of Applied Social and Economic Research, Papua New Guinea. Hanson, Douglas B. and Michael Pietrusewsky 1997 Bioarchaeological Research in the Mariana Islands of the Western Pacific: An Overview. American Journal of Physical Anthropology 104:267 269. Heckenberger, Michael J. 2005 The Ecology of Power: Culture, Place, and Personhood in the Southeastern Amazon, A.D. 1000â€“2000. Routledge, New York. Hensley, Rebecca A. and Timothy S. Sherwood 1993 An Overview of Guamâ€™s Inshore Fisheries. Marine Fisheries Review 55(2):129138. Heze l, F. X. 1983 The First Taint of Civilization: A History of the Caroline and Marshall Islands in Pre -Colonial Days, 1521â€“1885. University of Hawaiâ€˜i Press, Honolulu. Higgins, Jude 1999 Tnel: A Case Study of Avian Zooarchaeology and Taphonomy. Journal of A rchaeological Science 26:1449 1457. Higham, Charles 2002 Early Cultures of Mainland Southeast Asia . River Books, Bangkok.
255 Higham, Charles and Rachanie Thosarat 1994 Khok Phanam Di, Prehistoric Adaptation to the Worldâ€™s Richest Habitat . Case Studies in Arch aeology Series, edited by Jeffrey Quilter. Harcourt Brace College Publishers, Fort Worth, Texas. Hoffman, Major Carl W. 1950 Saipan: The Beginning of the End. Historical Division Headquarters, U.S. Marine Corps. Battery Press, Nashville, Tennessee. Hung, H., M. T, Carson, P. Bellwood, F. Z. Campos, P. J. Piper, E. Dizon, M. A. Bolunia, M. Oxenham, and Z. Chi 2011 The First Settlement of Remote Oceania: the Philippines to the Marianas. Antiquity (85):909â€“926. Hunter -Anderson, Rosalind 1989 Archaeological Investigations in the Small Boat Harbor Project Area, Agat, Guam . International Archaeological Research Institute, Inc., Honolulu. 2011 The Latte Period in Marianas Prehistory, Who is Interpreting it, Why and How? In Pacific Island Heritage, Archaeology, Ide ntity, and Community , edited by Jolie Liston, Geoffrey Clark, and Dwight Alexander. Australian National University E Press, Canberra. 2012 Running to Stay in Place: An Adaptive Escalation Model for the Latte Period. Micronesica 42(1/2): 148 182. Hunter -An derson, Rosalind, Judith R. Amesbury, and Darelene R. Moore 1998 Results of Monitoring and Data Recovery at Naton Beach, Tumon Bay, Guam . Micronesian Archaeological Research Services, Guam. Hunter -Anderson, Rosalind and Moore, Darlene R. 1994 Retreat or Expansion: Tracking Prehistoric Settlement in Guamâ€™s Interior . Paper presented at the annual meeting of the Australian Archaeological Association Ibez y Garcia, Luis de 1992 History of the M arianas, Caroline, and Palau Islands . Translated by Marjorie G. Driver. MARC Educational Series No. 12. Micronesian Area Research Center, University of Guam, Mangilao, Guam. Intoh, Michiko 1997 Human Dispersals into Micronesia. Journal of Anthropological Science 105(1):15 28. Irwin, Geoffrey 1992 The Prehistoric Exploration and Colonisation of the Pacific. Cambridge University Press, Cambrdge.
256 Jackson, Jeremy B.C., M.X. Kirby, W.H. Berger, K.A. Bjorndal, L.W. Botsford. B.J. Bourque, R.H. Bradbury, R. Cooke , J. Erlandson, J.A. Estes, T.P. Hughes, S. Kidwell, C.B. Lange, H.S.Lenihan, J.M. Pandolfi, C.H. Peterson, R.S. Steneck, M.J. Tegner, and R.R. Warner. 2001 Historical Overfishing and the Recent Collapse of Coastal Ecosystems. Science 293:629 638. Jalan doni, Andrea 2011 Casa Real or not Real? A Jesuit Mission House in Guam . Masterâ€™s thesis. University of the Philippines, Manila. Johannes, R. E. 1981 Words of the Lagoon: Fishing and Marine Lore in the Palau District of Micronesia. University of Californi a Press, Berkeley, California. 1978 Traditional Marine Conservation Methods in Oceania and there Demise. Annual Review of Ecology and Systematics 9:349_364. Kay, Allison E. 1979 Hawaiian Marine Shells, Reef and Shore Fauna of Hawaiâ€˜i, Section 4: Mollusca. Bishop Museum Press. Honolulu. Kinch, Jeff and Jane Bagita 2003 Women in Fisheries in Milne Bay Provence, Papua New Guinea: Past Initiatives, Present Situation and Future Possibilites. In SPC Women in Fisheries Bulletin 12:32â€“37. Kirch, Patrick V. 1985 Feathered Gods and Fishhooks: An Introduction To Hawaiian Archaeology and Prehistory. University of Hawaiâ€˜i Press, Honolulu. 1997 The Lapita Peoples: Ancestors of the Oceanic World . Blackwell Publishers, Oxford. 2000 On the Road of the Winds: An Archaeological History of the Pacific Islands before European Contact . University of California Press, Berkeley, California. 2010 How Chiefs Became Kings: Divine Kingship and the Rise of Archaic States in . University of California Press Berkeley, California. Kirch, Patrick V. and T.S. Dye. 1979 Ethnoarchaeology and the development of Polynesian fishing strategies. Journal of the Polynesian Society 88:53 76. Kirch, Patrick V. and R. C. Green. 2001 Hawaiki, Ancestral Polynesia: An Essay in Histo rical Anthropology . Cambridge University Press, Cambridge.
257 Krncke, Ingrid and Henning Reiss 2010 Influence of Macro Fauna LongTerm Natural Variability on Benthic Indices Used in Ecological Quality Assessment. Pollution Bulletin Kurashina, Hi ro, (editor) 1990 Archaeological Investigations at the Naval Facility (NAVFAC) Ritidian Point, Guam, Marianas . Micronesian Area Research Center, University of Guam, Mangilao, Guam. Labrune, Cline, Jean Michel Amouroux, Rafael Sarda, Eric Dutrieux, S bastien Thorin, Rutger Rosenberg, and Antoine Gr mare 2006 Characterization of the Ecological Quality of the Coastal Gulf of Lions (NW Mediterranean). A Comparative Approach Based on Three Biotic Indices. Marine Pollution Bulletin Lamprell, Ke vin and Thora Whitehead 1992 Bivalves of Australia, Volume 1 . Crawford House Press, Bathurst. Lvesque, R. (editor) 1995 History of Micronesia: A Collection of Source Documents , Vol. 4: Religious Conquest. Lvesque Publications, Gatineau, Qubec. Leach, B. Foss and Janet M. Davidson 2001 The Use of Size -Frequency Diagrams to Characterize Prehistoric Fish Catches and to Assess Human Impact on Inshore Fisheries. International Journal of Osteoarchaeology 11:150 162. Leidemann, Helen H. 1980 Intra-Site Variat ion at Ypao Beach, Guam, A Preliminary Assessment . Masterâ€™s thesis, Behavioral Science, University of Guam, Mangilao, Guam. Liston, Jolie 1996 The Legacy of Tarague Embayment and its Inhabitants, Anderson AFB, Guam . International Archaeological Research In stitute, Inc., Honolulu. Lodge, Major O. R. 1954 The Recapture of Guam . Historical Branch, G 3 Division Headquarters, U.S. Marine Corps. MacKenzie, Clyde L. 2001 The Fisheries for Mangrove Cockles, Anadara sp., From Mexico to Peru, With Descriptions of The ir Habitats and Biology, The Fishermanâ€™s Lives, and the Effects of Shrimp Farming. Marine Fisheries Review 63(1):1 39. Mannino, Marcello A. and Kenneth D. Thomas 2002 Depletion of a Resource? The Impact of Prehistoric Human Foraging on Intertidal Mollusc C ommunities and Its Significance for Human Settlement, Mobility and Dispersal. World Archaeology 33(3): 452 474.
258 Marche, Antoine -Alfred 1982 The Mariana Islands . Translated by Sylvia E. Cheng. Edited by Robert D. Craig. Micronesian Area Research Center, Uni versity of Guam, Mangilao, Guam. Marck, Jeffrey C. 1978 Interim Report on the 1977 Laulau Excavation, Saipan, Northern Mariana Islands . Commonwealth of the Northern Mariana Islands and the Office of Historic Preservation, Saipan, CNMI. Meehan, Betty 1982 Shell Bed to Shell Midden . Australian Institute of Aboriginal Studies, Canberra. Moore, Darlene R. 2002 Guamâ€™s Prehistoric Pottery and its Chronological Sequence . Micronesian Archaeological Research Services, Guam. 2005 Archaeological Evidence of a Prehistoric Farming Technique on Guam. Micronesica 38(1):93â€“120. 2007 Latte Period and Spanish Period Archaeology at Old Pago, Guam. Prepared for Mr. and Mrs. Richard Untalan and Guam Preservation Trust. Micronesian Archaeological Research Services, Guam. Moore, Darlene R. and Rosalind L. Hunter-Anderson. 1999 Pots and Pans in the intermediate Pre -Latte (25001600 BP) Mariana Islands, Micronesia. In The Pacific from 5000 to 2000 BP, Colonizations and Transformations , edited by Jean Christophe Galipaud and I . Lilley, pp. 487 503. Institut de recherche pour le dveloppement, Paris. Moore, Darlene R., Rosalind L. Hunter-Anderson, Judith R. Amesbury, and Eleanor F. Wells. 1992 Archaeology at Chalan Piao, Saipan . Micronesian Archaeological Research Services, Guam . Morrison, Alex E. and David J. Addison 2008 Assessing the Role of Climatic Change and Human Predation on Marine Resources at the Fatu ma -Futi Site, Tutuila Island, American Samoa: An Agent Based Model. Archaeology in Oceania (43):22 34. Morrison, Alex E . and Ethan E. Cochrane 2008 Investigating Shellfish Deposition and Landscape History at the Natia Beach Site, Fiji. Journal of Archaeological Science (35):2387 2399. Moseley, Michael 1975 The Maritime Foundations of Andean Civilization. Cummings, Menlo Pa rk, California.
259 Moss, Modanna L. 1993 Shellfish, Gender, and Status on the Northwest Coast: Reconciling Archaeological, Ethnographic, and Ethnohistoric Records of the Tlingit. American Anthropologist Myers, Robert F. and Terry J. Donaldson 2003 The Fishes of the Marianas. Micronesica 35/36:594 648. Nagelkerken, G. van der Velde, M. W. Gorissen, G. J. Meijer, T. vanâ€™t Hof, and C. den Hertog 2000 Importance of Mangroves, Seagrass Beds, and the Shallow Coral Reef as a Nursery for Important Coral Reef Fisheries, Using a Visual Census Technique. Eseturine, Coastal and Shelf Science 51:31â€“44. Oâ€™Day, Patrick M. 2002 Late Archaic Freshwater Mussel Exploitation Along the Middle Savannah River, Georgia. Masterâ€™s thesis. Department of Anthropology, Univ ersity of Florida, Gainesville, Florida. Oâ€™Day, Patrick M. and Nicole I. Vernon. 2011 Historic Resources of the Carolinas Heights Region, Island of Tinian, Commonwealth of the Northern Marianas . Garcia and Associates, Kailua, Hawaiâ€˜i. Olive y Garc a, Francisco 2006 The Mariana Islands, Random Notes Concerning Them, Second Edition. Translated and Annotated by Marjorie G. Driver. Spanish Documents Collection, Richard Flores Taitano. Micronesian Area Research Center, University of Guam, Mangilao, Guam. Oliver , Douglas L. 1989 Oceania: The Native Cultures of Australia and the Pacific Islands . 2 vols. Olmo, Richard K. 2013 New Flesh for Old Bones: Using Modern Reef Fish to Understand Midden Remains from Guam, Mariana Islands. In Prehistoric Marine Resource Use in the Indo Pacific Regions , edited by Rintaro Ono, Alex Morrison, and David Addison, pp. 1â€“31. Australian National University Press, Canberra. Paulay, Gustav 1992 Trends in Shellfish Assemblages in the Marianas Islands. In Archaeology at Chalan Piao, Saipan, by Moore, Darlene R., Rosalind L. Hunter -Anderson, Judith R. Amesbury, and Eleanor F. Wells., Appendix C. Micronesian Archaeological Research Services, Guam. 2003a Marine Biodiversity of Guam and the Marianas: Overview. Micronesica 35/36:3 25.
260 2003b Marine Bivalvia (Mollusca) of Guam. Micronesica 35/36:218 243. Pauly, Daniel 1995 Anecdotes and the Shifting Base Line Syndrome of Fisheries. Tree 10(10):430. Pawley, Andrew and Roger Greene 1984 The ProtoOceanic Language Com munity. The Journal of the Pacific 19 (3):123146. Pawley, Andrew and Malcolm Ross 1993 Austronesian Historical Linguistics and Culture History. Annual Review of Anthropology 22:425 59. Peattie, Mark R. 1988 Rise and Fall of the Japanese in Micronesia, 1885-1945. Pacific Islands Monograph Series, No. 4. University of Hawaiâ€˜i Press, Honolulu. Peterson, Glenn 2009 Traditional Micronesian Societies, Adaptation, Integration, and Political Organization. University of Hawaiâ€˜i Press. Honolulu. Pertosian -Husa, Carmen C.H. 2004 Traditional Fishing Techniques in the Marshall Islands . Republic of the Marshall Islands, Ministry of Internal Affairs, Historic Preservation Office, Alele Museum. Alele Report 2004/3, Majuro A toll. Poutiers, J.M. 1998 Gastropods. In FAO Species Identification Guide for Fisheries Purposes, The Living Marine Resources of the Western Pacific, Volume 1: Seaweeds, Corals, Bivalves, and Gastropods, e dited by Kent E. Carpenter and Volker H. Niem, pp. 363648. Food and Agriculture Organization of the United Nations, Rome. Rainbird, Paul. 1994 Prehistory in the Northwest Tropical Pacific: The Caroline, Mariana, and Marshal Islands. Journal of World Prehistory 8:293349. Randall, John E. 1985 Guide to Hawaiian Reef Fishes . Treasure of Nature, Kaneohe, Hawaiâ€˜i. 2001 Surgeonfishes of Hawaiâ€˜i and the World. Mutual Publishing, Bishop Museum Press. Honolulu, Hawaiâ€˜i. Rawlison, N., D. Milton, S. Blaber, A. Sesewa, and P. Sharma 1995 A Survey of the Subsistence and Artisanal Fisheries in Rural Areas of Viti Levu, Fiji . Australia Centre for International Agricultural Research, Canberra.
261 Ray, Col. Erwin, William R. Fortini, Jr., and Juan L. Babauta 1996 Archaeological Data Recovery at Akitsu Shojiâ€™s Residence in San Roque, Saipan, CNMI. AB Business Management and Consulting Services, Saipa, CNMI. Reed, Erik 1952 General Report on Archaeology and History of Guam . National Park Service,Santa Fe, New Mexico . Reid, Lawrence A. 2002 Morphosyntactic Evidence for the Position of Chamorro in the Austronesian Language Family. In Collected Papers on Southeast Asian and Pacific Languages , edited by R. S. Bauer, pp. 63â€“ 94. Pacific Linguistics, Canberra. Reimer, P. J., Baillie, M. G. L., Bard, E., Bayliss, A., Beck, J. W., Blackwell, P. G., Bronk Ramsey, C., Buck, C. E., Burr, G. S., Edwards, R. L., Friedrich, M., Grootes, P. M., Guilderson, T. P., Hajdas, I., Heaton, T. J., Hogg, A. G., Hughen, K. A., Kaiser, K. F., K romer, B., McCormac, F. G., Manning, S. W., Reimer, R. W., Richards, D. A., Southon, J. R., Talamo, S., Turney, C. S. M., van der Plicht, J., & Weyhenmeyer, C. E. 2009 IntCal09 and Marine09 radiocarbon age calibration curves, 050,000 years cal BP. Radioc arbon 51(4), 1111 1150. Reinman, Fred, M. 1977 An Archaeological Survey and Preliminary Test Excavations on the Island of Guam, Mariana Islands, 1965-66. Miscellaneous Publications 1. Micronesian Area Research Center, University of Guam, Mangilao, Guam. Re itz, Elizabeth J. and Elizabeth S. Wing. 2008 Zooarchaeology. Cambridge University Press, Cambridge. Ricker, William E. 1954 Stock and Recruitment. Journal of the Fisheries Research Board of Canada 11(5):559â€“623. 1975 Computation and Interpretation of Biological Statistics of Fish Populations . Bulletin of the Fisheries Researh Board of Canada. Bulletin 191. Department of the Environment, Fisheries and Marine Service, Ottawa. 1977 The Historical Development. In Fish Population Dynamics , edited by J. A. Gull and pp. 1â€“26. John Wiley and Sons, London. Ringrose, Trevor J. 1993 Bone Counts and Statistics: A Critique. Journal of Archaeological Science 20:121â€“ 157. Rogers, Robert F. 1995 Destinyâ€™s Landfall: A History of Guam . University of Hawaiâ€˜i Press, Honolulu.
262 Ropes, John W. 1984 Procedures for Preparing Acetate Peels and Evidence Validating the Annual Periodicity of Growth Lines Formed in Shells of Ocean Quahogs. Artica islandica . Marine Fisheries Review 46(2):27â€“35. 1987 Preparation of Acetate Peels of Valves from the Ocean Quahog for Age Determinations . NOAA Technical Report NMFS 50. U.S. Department of Commerce, National Technical Information Service, Springfield, Virginia. Ross, Malcolm D. 1989 Early Oceanic Linguistic Prehistory. The Journal of Pacific Hist ory 24(2):135 149. Russell, Scott. 1984 From Arabwal to Ashes â€” A Brief History of Garapan Village: 1818 to 1945. Micronesian Archaeological Survey Report 19. Department of Education, Saipan, CNMI. 1998 Tiempon I Manmofo na: Ancient Chamorro Culture and History of the Northern Mariana Islands . Division of Historic Preservation, Commonwealth of the Northern Mariana Islands, Saipan, CNMI. Sahlins, Marshall D. 1962 Moala, Culture and Nature on a Fijian Island. The University of Michigan Press, Ann Arbor, Mic higan. Sanz, Manuel 1991 Description of the Marianas Islands, Manuel Sanz, Manila, 1827. Translated by Marjorie G. Driver. MARC Educational Series No.10. University of Guam, Mangilao, Guam. Sather, Clifford A. 1997 The Bajau Laut, Adaptation, History, and Fate in a Maritime Fishing Society of South-Eastern Sabah. Oxford University Press, New York. Schmidt, Peter R. 1997 Iron Technology in East Africa: Symbolism, Science, and Archaeology . University of Indiana Press, Bloomington, Indiana. Short, Andrew D. 2005 Carbonate Sandy Beaches. In Encyclopedia of Coastal Science. Maurice I. Schwartz Ed. Springer, Dordrecht, The Netherlands. Shotwell, Arnold J. 1955 An Approach to Paleoecology of Mamals . Ecology 36:327â€“337. Smith, Barry D. 2003 Prosobranch Gastropods of Guam. Micronesica 35/36:244 27.
263 Smith, Courtland L. 1983 Evaluating Human Factors. In Fisheries Techniques , edited by L. A. Neilsen, D. L. Johnson, and S. S. Lampton, pp. 431â€“446. American Fisheries Society, Bethesda Maryland. Smith, Tim D. 1994 Scaling Fisheries: The Science of Measuring the Effects of Fishing , 1855â€“1955. Cambridge University Press, Cambridge. Solheim II, Wilhelm G. 2003 Southeast Asian Earthenware Pottery and Its Spread. In Earthenware in Southeast Asia , edited by John Miksic, pp. 131. Singapore University Press, Singapore. Spennemann, Dirk H.R 2007 Edge of Empire, The German Colonial Period in the Northern Mariana Islands, 1899-1914. Heritage Futures International, Albury, NSW, Australia. Spoehr, Alexander 1954 Saipan, The Ethnology of a War -Devastated Island. Chicago Natural History Museum, Chicago. 1957 Marianas Prehistory, Archaeological Survey and Excavation On Saipan, Tinian, and Rota. Chicago Natural History Museum, Chicago. Stacey, Natasha 2007 Boats to Burn: Bajo Fishing Activity in Australian Fishing Zone. The Australian National University E Press, on line at http:/epress.anu.edu.au//boats_citation.html. Steadman, David, W. 1997 Prehistoric Extinctions of Polynesian Birds: Reciprocal Impacts of Birds and People. In Hist orical Ecology in the Pacific Islands , edited by Patrick V. Kirch and Terry L. Hunt, pp. 51â€“79. Yale University Press, New Haven, Connecticut. 1999 The Prehistory of Vertebrates, Especially Birds, on Tinian, Aguigan, and Rota, Northern Mariana Islands. Mic ronesica 31(2):319 345. 2006 Extinction and Biogeography of Tropical Pacific Birds . The University of Chicago Press, Chicago. Stephens, D. W. and J. R. Krebs 1986 Foraging Theory . Princeton University Press, Princeton, New Jersey. Steward, Julian H. 1955 The Theory of Cultural Change, The Methodology of Multilinear Evolution. Board of Trustees of the University of Illinois, Champaign Urbana, Illinois.
264 Steward, Julian H. and F. M. Setzler 1938 Function and Configuration in Archaeology. American Antiquity 4:4 â€“10. Swadling, Pamela and Ann Chowning 1981 Shellfish Gathering at Nukulau Island, West Britain Province, Papua New Guinea. Journal de la Socit Ocanistes 72/73(37):19 167. Szab, Katherine and Sue Oâ€™Connor 2004 Migration and Complexity in Holocene Isl and Southeast Asia. World Archaeology 36(4): 621 628. Tabroi, D., J. W. Jensen, and J. E. Mylroie 2005 Karst Features of Guam, Mariana Islands. Micronesica 38(1):17â€“46. Thompson, Dean M. 1979 Marianas Plain Pottery from the Tanapag Site, Saipan, Marianas . Unpublished Masterâ€™s Thesis. University of Iowa, Iowa City, Iowa. Thompson, Laura, M. 1932 Archaeology of the Marianas Islands . Bernice P. Bishop Museum, Bulletin 100. Honolulu, Hawaiâ€˜i. 1945 The Native Culture of the Mariana Islands . Bernice P. Bishop Museum Bulletin 185. Honolulu, Hawaiâ€˜i. Titcomb, Margaret 1972 Native Use of Fish in Hawaiâ€˜i . University of Hawaiâ€˜i Press, Honolulu. 1978 Native Use of Marine Invertebrates in Old Hawaiâ€˜i. Pacific Science 32(4):325 385. Tomonari Tuggle, Myra J., H. David Tuggle, and David J. Welch 2005 Final â€“Regional Integrated Cultural Resources Management Plan for COMNAVREG Marianas Lands. Volume I: Guam . International Archaeological Research Institute, Inc., Honolulu. Torrence, Robin and Anne Clarke 2000 Negotiating Difference: Practice Makes Theory for Contemporary Archaeology in Oceania. In The Archaeology of Difference, Negotiating Cross -Cultural Engagements in Oceania , edited by Robin Torrence and Anne Clarke, pp. 1 31. Routledge, London. Trigger, Bruce G. 1989 A History of Archaeological Thought . Cambridge University Press, Cambridge. Tryon, Darrell 2006 Proto-Austronesian and the Major Austronesian Subgroups. In The Austronesians, Historical and Comparative Perspectives , edited by Peter Bellwood, James J. Fox, and Darrell Tryon. ANU E Press, Canberra.
265 Vilar, Miguel G., Chim W. Chan, Dana R. Santos, Daniel Lynch, Rita Spathis, Ralph M. Garruto, and J Koji Lum 2012 The Origins and Genetic Distinctiveness of the Chamorros of the Marianas Islands: An mtDNA Perspective. American Journal of Human Biology Ward, Graeme, and John Craib 1985 Preliminary Report on Archaeological Research at Unai Bapot, Saipan, duringJanuary 1985. Unpublished manuscript on file at Division of Historic Preservation , Department of Community and Cultural Affairs, Saipan, CNMI. White, Peter J., Geoffrey Clark, and Stuart Bedford 2000 Distribution, Present and Past, of Rattus praetor in the Pacific and its Implications. Pacific Science 54(2):105 117. Wickler, Stephen 2002 Modelling Colonisation and Migration from a Zooarchaeological Perspective. In Colonisation, Migration, and Marginal Areas, a Zooarchaeological Approach, edited by M. Mondinin and S. Wickler, pp. 28â€“41. Oxbow Books, Oxford. Woodroffe, Colin D. 1987 Paci fic Island Mangroves: Distribution and Environmental Settings. Pacific Science 41:166 185. Woodroffe, C. D., B. G. Thom & J. Chappell 1985. Development of widespread mangrove swamps in mid-Holocene times in northern Australia. Nature 317:711â€“713. Yesner, David R. 2008 Ecology in Archaeology. In Handbook of Archaeological Theories , edited by Alexander Bentley, Herbert D. G. Maschner, Christopher Chippindale, pp. 39 55. Altamira Press, Lanham, Maryland. Zar, Jerrold H. 1999 Biostatistical Analysis . Prentice Hill, New Jersey. Zeder, Melinda A. 2012 The Broad Spectrum Revolution at 40: Resource Diversity, Intensification, and an Alturnative to Optimal Forging Explanations. Anthropological Archaeology 31:241â€“264. Zobel, Erik 2002 The Position of Chamorro and Pal auan in the Austronesian Family Tree: Evidence from Verb Morphosyntax. In Collected Papers on Southeast Asian and Pacific Languages , edited by R. S. Bauer, pp. 405 â€“434. Pacific Linguistics, Canberra. Zohar, I., T. Dayan, E. Galili, E. Spanier. 2001 Fish Processing During the Early Holocene: A Taphonomic Case Study from Coastal Israel. Journal of Archaeological Science 28:1041â€“1053.
266 BIOGRAPHICAL SKETCH Patrick Oâ€™Day currently works for Garcia and Associates and manages cultural resource management projects for private sector and government undertakings in Hawaiâ€˜i. Patrickâ€™s international work experience includes projects in Micronesia and Japan. He has also conducted work for the State Historic Preservation Offices in Guam and the CNMI, The Republi c of Palauâ€™s Bureau of Arts and Culture, and various private sector construction firms in Guam and the CNMI. Throughout his graduate career, Patrick also worked on various interdisciplinary research projects in Florida, Fiji, Hawaiâ€˜i, the Bahamas, and Sout h America. Beginning in 1999 Patrick worked on various projects assessing vertebrate and invertebrate fisheries in the Gulf of Mexico for the University of Florida, Department of Fisheries and Aquatic Sciences as part of his interdisciplinary graduate research. Patri ck has conducted species counts on artificial and natural reefs, surveyed and mapped underwater habitats, and collected various species of fish and marine invertebrates utilizing a wide variety of gear and techniques. He has also conducted laboratory analy ses on these specimens for life history data and contributed to peerreviewed publications. In many ways, these experiences in fisheries science laid the groundwork for Patrickâ€™s research interests in prehistoric fisheries.