THE FLORIDA ANTHROPOLOGIST
Published by the FLORIDA ANTHROPOLOGICAL SOCIETY, INC. VOLUME 65 (1-2) MARCH-JUNE 2012
THE FLORIDA ANTHROPOLOGIST is published by the Florida Anthropological Society, Inc., P.O. Box 357605, Gainesville, FL 32635. Subscription is by membership in the Society. Membership is NOT restricted to residents of the State of Florida nor to the United States of America. Membership may be initiated at any time during the year, and covers the ensuing twelve month period. Dues shall be payable on the anniversary of the initial dues payment. Members shall receive copies of all publications distributed by the Society during the 12 months of their membership year. Annual dues are as follows: student $15, individual $30, family $35, institutional $30, sustaining $100 or more, patron $1000 or more, and benefactor $2500. Foreign subscriptions are an additional $25 U.S. to cover added postage and handling costs for individual, family, or institutional membership categories. Copies of the journal will only be sent to members with current paid dues. Please contact the Editors for information on recent back issues.
Requests for information on the Society, membership application forms, and notifications of changes of address should be sent to the Membership Secretary. Donations should be sent to the Treasurer or may be routed through the Editors to facilitate acknowledgment in subsequent issues of the journal (unless anonymity is requested). Submissions of manuscripts should be sent to the Editors. Publications for review should be submitted to the Book Review Editor. Authors please follow The Florida Anthropologist style guide (on-line at www.fasweb.org) in preparing manuscripts for submission to the journal and contact the Editors with specific questions. Submit four (4) copies for use in peer review. Only one set of original graphics need be submitted. The journal is formatted using Adobe In Design. All manuscripts must be submitted in final form on CD in Microsoft format. Address changes should be made AT LEAST 30 DAYS prior to the mailing of the next issue. The post office will not forward bulk mail nor retain such mail when "temporary hold" orders exist. Such mail is returned to the Society postage due. The journal is published quarterly in March, June, September, and December of each year.
OFFICERS OF THE SOCIETY
President: Patty Flynn, P.O. Box 11052, Ft. Lauderdale, FL. 33339 (email@example.com) First Vice President: Jeffrey T. Moates, FPAN West Central Regional Center, 4202 E. Fowler Ave., NEC 16, Tampa FL 33620
Second Vice President: Theresa Schrober, 1902 Florried Court, N. Fort Myers, 33917 (firstname.lastname@example.org) Corresponding Secretary: Jon-Simon Suarez, 1710 NW 7th St, #304, Gainesville, FL 32609 (email@example.com) Membership Secretary: Pat Balanzategui, P 0 Box 1434, Fort Walton Beach, FL 32549-1434 (firstname.lastname@example.org) Treasurer and RegisteredAgent: Joanne Talley, P.O.Box 788, Hobe Sound, FL 33475 (email@example.com) Directors at Large: Chris Hardy, 1668 Nantucket Ct., Palm Harbor 34683 (firstname.lastname@example.org); Sherry Svekis, 406 Woodland
Dr., Sarasota, FL 34234 (email@example.com); Tommy Abood, 3857 Indian Trail, Suite 403, Destin, FL 32541 (lost.horizon@
earthlink.net); and Nick McAuliff, 115 Fernandina Ave., St. Augustine, FL 32080, (firstname.lastname@example.org). Immediate Past President: Robert J. Austin, P. O. Box 2818, Riverview, FL 33657-2818 (email@example.com) Newsletter Editor: David Burns, 15128 Springview St., Tampa, FL 33624 (firstname.lastname@example.org)
JOURNAL EDITORIAL STAFF
Co-Editors: Keith H. Ashley Department of Anthropology, Building 51, University of North Florida, 1 UNF Drive, Jacksonville, FL 32224-2659 (email@example.com). Vicki L. Rolland, Department of Anthropology, Building 51, University of North Florida, 1 UNF Drive, Jacksonville, FL 32224-2659 (firstname.lastname@example.org)
Book Review Editor: Jeffrey T. Moates, FPAN West Central Regional Center, 4202 E. Fowler Ave NEC 116, Tampa FL 33620
Editorial Assistant: George M. Luer, 3222 Old Oak Drive, Sarasota, FL 34239-5019 (email@example.com) Technical Assistant: Michael Boyles Center for Instruction and Research, UNF, 1 UNF Drive, Jacksonville, FL 32224-2659) Printer: Durra-Print, 717 South Woodward Ave., Tallahassee, FL 32304 Bulk Mail: TCB Marketing, 2818 South Monroe Street, Tallahassee, FL 32301
EDITORIAL REVIEW BOARD
Albert C. Goodyear, Institute of Archaeology and Anthropology, University of South Carolina, Columbia, SC 29208 (firstname.lastname@example.org)
Jerald T. Milanich, Florida Museum of Natural History, University of Florida, Gainesville, FL 32611 (email@example.com) Jeffrey M. Mitchem, Arkansas Archeological Survey, P.O. Box 241, Parkin, AR 72373 (firstname.lastname@example.org) Nancy Marie White, Department of Anthropology, University of South Florida, Tampa, FL 33620-8100 (nwhite@chumal .cas.usf.edu)
Robert J. Austin, P.O. Box 2818, Riverview, FL 33568-2818 (email@example.com)
NOTE: In addition to the above Editorial Review Board members, the review comments of others knowledgeable in a manuscript's subject matter are solicited as part of our peer review process.
VISIT FAS ON THE WEB: www.fasweb.org
THE FLORIDA 4zoot
Volume 65 Numbers 1-2
March-June 2012 TABLE OF CONTENTS19A1
FROM THE EDITOR 3
THE WAKULLA SPRINGS LODGE SITE (8Wa329): 2008 EXCAVATIONS AND NEW OSL DATING EVIDENCE 5
W. JACK RINK, JAMES S. DUNBAR, AND K. E. BURDETTE
BIOTURBATION AND THE WAKULLA SPRINGS LODGE SITE ARTIFACT DISTRIBUTION 25
DAVID K. THULMAN
GROUND-PENETRATING RADAR (GPR) SURVEY, WAKULLA SPRINGS STATE PARK, WAKULLA COUNTY, FLORIDA: USE OF GPR IN SUPPORT OF ARCHAEOLOGY 35
DAVID ANDY SMITH
GRANULOMETRIC ANALYSIS OF SEDIMENT SAMPLES FROM THE WAKULLA SPRINGS LODGE SITE 43
WAKULLA COUNTY, FLORIDA
Guy H. MEANS
THE FIRST FLORIDA "BLING": PALEOINDIAN BEADS 49
ANOTHER WAKULLA SPRINGS MASTODON? 53
KEVIN M. PORTER
WAKULLA SPRINGS ARCHAEOLOGICAL RESEARCH THE VOLUNTEER EXPERIENCE 59
MARY STERNER LAWSON
THE PURDY UNIFACE, A RECENTLY RECOGNIZED EARLY STNE TOOL TYPE IN FLORIDA 67
ScoTT E. MITCHELL AND L. H. "MONTY" PHARMER
PALEOINDIAN OCCUPATIONS ALONG THE ST. JOHNS RIVER, FLORIDA 79
DAVID K. THULMAN
GEOARCHAEOLOGICAL INVESTIGATIONS AND OSL DATING EVIDENCE IN AN ARCHAIC AND PALEOINDIAN CONTEXT AT THE HELEN BLAZES SITE (8Br27), BREVARD COUNTY, FLORIDA 87
W.JACK RINK, JAMES S. DUNBAR, GLEN H. DORAN, CHARLES FREDERICK, AND BRITTNEY GREGORY
ABOUT THE AUTHORS 110
TYPOLOGICAL, FUNCTIONAL, AND COMPARATIVE CONTEXTUAL ANALYSES OF WOODLAND HASFTED BIFACES FROM KOLOMOKI (9ER1), THOMAS J. PLUCKHAHN AND SEAN P. NORMAN 112
Cover: Artist's depiction of a mastodon kill site and the volunteers of the Wakulla excavations.
Published by the
FLORIDA ANTHROPOLOGICAL SOCIETY, INC.
.......@ iiii!i ii iiii~ ...... i~iiii ............ ~i................................i .......................................... 'W i ............................ i i '"i~'"'i '!............................ '""iii WAFT....,.........
...................................................................................... .............................. i~iii~~iiiii~iiiiiiiiiiiiiii~iiiiiii~ iiiiii~i~iii~iiiiiii~iiiiii i!! :i!.......... ii ~iii~~~is siiiiii~~i!il!!i !..........
ii~~~~~~~~~~~~ii~-- r p t ci!!iiiiii i a-.. bF C"iiiii~i!i!iil~ii iil iiiiiiii!!ii iii I i i i
FROM THE EDITOR
It has been several years since the FA produced a volume at various depths. While Andy was not able to identify other dedicated to Paleoindian archaeology in Florida, and given the archaeological deposits, his work does show the potential for amount of recent work done on Paleoindian sites and issues, it GPR's use in archaeology. seems long overdue (especially to a Paleoindian archaeologist The interpretation of OSL results requires a detailed unlike me). Seven of the articles in this double volume focus on derstanding of the sediments and their depositional history. the excavations of the Wakulla Springs Lodge site (8WA329) That background granulometric work was provided by Harley in the spring of 2008 under the direction of Jim Dunbar. This Means, also from the FGS. In the fourth article, Harley reports interdisciplinary effort, funded by the National Geographic on the results of his analyses and relates them to a fascinatSociety, drew in professionals who usually work outside of ing and unexpected geologic history of the site. Most people archaeology and a large team of dedicated volunteers. The vol- assumed the sand ridge where the site is located was created ume also includes three articles concerning other Paleoindian by eolian (wind-blown) processes, but Harley argues the sand issues: two reporting on Paleoindian sites on the St. Johns Riv- likely was deposited through fluvial processes when Wakulla er and one describing a newly identified formal Paleoindian was draining to the aquifer rather than discharging as it does tool. today. He presents a convincing LIDAR image of the area that
The first article by Jack Rink, Jim Dunbar, and Kevin demonstrates how that counter-intuitive series of events may Burdette presents an overview of the Wakulla Springs project, have occurred. the previous excavations of the site led by Calvin Jones, and The excavations uncovered several interesting artifacts, the optically stimulated luminescence (OSL) dating of the de- but the tiny seed bead only 4.3 mm in diameter described posits. OSL dating is our best opportunity for identifying early by Mary Glowacki in the fifth article may be the most intrigusites when material suitable for radiocarbon dating is lacking, ing. Mary relates the long history of personal adornment in but, like radiocarbon dating, the process of relating OSL ages prehistory and puts this small bead in a broader North Amenwith archaeological deposits is not always straightforward. can Paleoindian context. This is not the only seed bead excaOSL is a highly technical endeavor, but the authors do a good vated from a Florida Paleoindian site. Another was found at job of describing the science, technical challenges, and con- the nearby Ryan-Harley site (8JE 1004), a Suwannee point site siderations required for interpreting the results. on the Wacissa River.
The second article examines the problems encountered The discovery by park ranger Jason Vickery of a mastat the site in trying to correlate levels among the excavation odon in 2007 adjacent to the swimming area at Wakulla Spring units. My initial intent was to describe the archaeological and just down the slope from the site presents the potential for strata, debitage, and tools, but it soon was apparent that the finding a Paleoindian kill site. Preliminary investigation of the stratigraphic relationships among the units were ambiguous. bones indicates more of the animal is present, but the remains The article morphed into an examination of how the artifact have not yet been excavated. In the sixth article, Kevin Porter distributions could be explained. Ants, gophers, and other bio- reports on the discovery and reviews the long history of paturbators are not usually discussed in an archaeology journal, leontological and archaeological investigations at the spring, but in sandy sites in Florida and elsewhere, their effects on site which started in 1835 with the recovery a mastodon skull and integrity should always be considered. My conclusion is that tusks. Hopefully, the Vickery mastodon will be revisited soon. the artifact distribution at the site is likely the result of thou- In the last article concerning the Wakulla Springs Lodge sands of years of bioturbation (with some eolian deposition) site, Mary Lawson describes and illustrates her experiences that caused the artifacts to "sink" to their present locations, as a volunteer on the site, mainly working with Louis Tesar This has important implications for the OSL dating, which is cataloging artifacts. Mary interviewed several of the volundiscussed in this and the preceding article. teers many of whom had never been on a dig before and
The Wakulla Springs Lodge sits on a sandy ridge over- relates their experiences, impressions, and enthusiasm for arlooking the spring, but what underlies that ridge was unknown. chaeology. The Wakulla project depended heavily on the nonFurther, was there an easier way to find archaeological depos- professional volunteers, but their stories are seldom told. Mary its without excavating numerous test units in lawn in front of also sketched the operations, several of which are included, the lodge? In the third article Andy Smith of the Florida Geo- capturing the activities. logical Survey (FGS) writes about his ground penetrating ra- In the eighth article, Scott Mitchell and Monty Pharmer dar (GPR) survey of the area surrounding the site. He and col- of the Silver River Museum in Silver River State Park near leagues dragged the GPR device for about 15 km of transects Ocala describe a previously unrecognized formal Paleoindian to gather the data! The data can be translated into images that tool, which they name the Purdy Uniface in honor of Barbara "slice" deposits and create maps of what is below the surface Purdy, long-time Paleoindian archaeologist and an institution
VOL.65(1-2) THE FLORIDA ANTHROPOLOGIST MARCH-JUNE 2012
4 THE FLORIDA ANTHROPOLOGIST 2012 VOL.65 (1-2)
in Florida archaeology. Scott and Monty reviewed several pri- and future excavations will continue to plumb Florida's potenvate and public collections, including the large Alvin Hendrix tial for contributing to our understanding of this time period. collection in the Museum, and provide detailed data on over 50 specimens. They also discuss the manufacture and possible David Thulman uses of the tools. Their work provides a welcome addition to Guest Editor understanding the Paleoindian toolkit beyond the projectile points and demonstrates the research potential of our museum collections in the state. It will be interesting to see whether the Purdy Uniface is present outside of Florida. Keith H. Ashley
In the ninth article, I report on the Lake George Point site Vicki L. Rolland (8PU1470), a newly identified Suwannee Paleoindian site in Lake George on the St. Johns River. The site was found by local collectors and reported to the state. Over 40 Suwannee points and broken bases have been recovered from the site, along with a variety of tools and late Pleistocene-age fossils. The site is on the eastern edge of the lake on a submerged sand spit that would have overlooked a wide basin when the water tables were lower during the late Pleistocene Epoch. The article describes a survey of the site in an unsuccessful attempt to find intact Paleoindian deposits. Regardless, the presence of this large site changes our previous assumption that Paleoindians were solely focused on the western, Gulf side of the Florida peninsula.
The final article by Rink, Dunbar, Glen Doran, Charles Fredrick, and Brittany Gregory describes the recent re-evaluation of the Helen Blazes site (8BR27), first excavated by William Edwards in 1949-1951. The site sits on the eastern edge of the St. Johns River floodplain in a location that is remarkably reminiscent of the Lake George Point site. The authors found Edwards' original excavations, put in their own units, reanalyzed the stratigraphy, and collected samples for OSL dating. The Helen Blazes work is part of a larger effort to revisit and collect OSL ages from Paleoindian sites in Florida. The article describes the stratigraphy and relates the new interpretation to Edwards' earlier work. The site is intriguing because of its location (the most southeastern of known Suwannee sites and far from suitable lithic raw material) and the diversity of projectile point forms illustrated by Edwards.
The cover of the volume includes a depiction of early Floridians at Wakulla Spring (or Wakulla Sinkhole, as it likely was at that time) by Barbara Taillefer. Under that image is a group photograph, taken by Charles Montfort. of the Wakulla excavation volunteers.These ten articles do not capture all of the work on Paleoindian issues presently going on in Florida. The Rink and Dunbar team visited several other Paleoindian sites in Florida, and hopefully reports from those efforts will be forthcoming. Doctoral student Jessie Halligan from Texas A&M has completed field work on several submerged Paleoindian sites on the Aucilla River, and C. Andrew Hemmings and James Adovasio continue their deep water exploration for Paleoindian sites in the Gulf of Mexico. Ongoing excavations in Little Salt Spring (8S0 18) will undoubtedly continue to uncover new Paleoindian artifacts. In my opinion, Florida presents opportunities for the preservation of organic Paleoindian artifacts that are unique in North America. Hopefully, these
THE WAKULLA SPRINGS LODGE SITE (8WA329): 2008 EXCAVATIONS AND NEW OSL DATING EVIDENCE
W. JACK RINK', JAMES S. DUNBAR2 AND KEVIN E. BURDETTE3
'School of Geography and Earth Sciences, McMaster University, 1280 Main St.. W, Hamilton, Ontario, Canada L8S 4K]
2 Department ofAnthropology, Florida State University, Tallahassee, FL USA 32306
3 School of Geography and Earth Sciences, McMaster University, 1280 Main St.. W., Hamilton, Ontario, Canada L8S 4K]
Introduction In August 2007, a park ranger discovered the buried
remains of a mastodon in the Wakulla River below the
The Wakulla Springs Lodge site (8WA329), located in headspring (Dunbar et al. 2007; Porter 2012, this volume). The north Florida's panhandle (Figure 1), has been known for many well-preserved bones vividly demonstrate the difference in years as one of the state's early sites. Recent reassessment of preservation between upland and submerged sites in Florida. research conducted by the Florida Bureau of Archaeological The mastodon lies beneath 2 m of water about 130 m north of Research in the mid-i 990s suggests that this site was occupied the land excavation. well before the Clovis people who were initially thought to In December 2007, members of the Bureau of have been the first Americans. This report summarizes the Archaeological Research (BAR), Florida Geological Survey, results of field investigations carried out in 2008 at the Wakulla and the Geology Department and Coastal & Marine Laboratory Springs Lodge site as well as the ensuing research. at Florida State University gathered to vibra-core near the
In 1994, excavations at the Wakulla Springs Lodge site mastodon site on its northern, eastern, and western sides. The (Jones and Tesar 2000; Tesar and Jones 2004) generated an in results of that effort confirmed that the Wakulla River just situ projectile point identical to pre-Clovis points recovered below the springhead has channel-fill sediment sequences elsewhere in the region (Dunbar 2006a, 2008). This find over a meter in thickness in which the mastodon remains are suggested that the site was occupied prior to the Paleoindian buried. Clovis interval (ca. 11,200 to 10,800 BP). The Wakulla Springs
Lodge site shares a physical similarity with other early sites Previous Archaeological Investigations of in Florida, such as the Page-Ladson (8JE591) and Sloth Hole Paleoindian Age in the Region (8JE121) sites (Figure 1); that is, it is adjacent to a deep karst
feature (one of the nation's largest first magnitude springs). If The Page-Ladson site is located in the Aucilla River in that was a common feature of pre-Clovis sites in Florida, then Jefferson County, about 30 km east-southeast from Wakulla Wakulla may also possess a pre-Clovis component. In order Springs Lodge (Figure 1). The deposits there are characterized to test this hypothesis, the 2008 project involved excavations by levels of still-water deposited peat, fluvial channel-fill adjacent to the location of the 1994 discovery with the goal calcareous silt, colluvium, smectite, and a few hiatuses formed of identifying. additional Paleoindian artifacts and generating by episodes of channel bottom subaerial exposure with little to material to date radiometrically the earliest strata in the site. no deposition and reduction by oxidation of exposed organicrich levels. The stratigraphic column is over 7 m thick, but the
Previous Archaeological Investigations at total thickness of the sinkhole fill deposit remains unknown.
Wakulla Springs The upper 7 m plus section of the column is very well dated
and ranges from ca. 22,300 cal B.P. to ca. 10,700 cal B.P.
In 1994, when Calvin Jones (Tesar and Jones 2004) (Pleistocene late glacial maximum into the early Holocene). excavated a long trench in the right-of-way for a new sewer line The earliest Paleoindian evidence is in the colluvial level of at the Wakulla Springs Lodge site, he found four Paleoindian the site, which also yielded butcher-marked mastodon bones artifacts adjacent to our 2008 test units within trench segment as well as non-diagnostic stone artifacts. This level dated TS 10. The Paleoindian artifacts recovered from TS 10 occurred to ca. 14,425 cal B.P. (average of seven statistically related at depths of 105 to 115 cm below surface. They included a 14C dates from Unit 3, the pre-Clovis level at Page-Ladson) large Simpson-like preform, a Page-Ladson point, a flake (Dunbar 2006a, 2006b; Webb and Dunbar 2006). Though not preform, and a unifacial scraper. Twenty to twenty-five meters recovered from in-place contexts, several (n=5) non-Clovis, west of that location the Paleoindian levels in trench segment Paleoindian lanceolate points were identified that eventually TS 11 were found more deeply buried at 135 to 165 cm below became known as Page-Ladson points (Dunbar and Hemmings surface. Bolen artifacts in the TS 10 area were recovered from 2004). These points were made on bifacially reduced flakes levels 6 and 7, which ranged from 75 to 105 cm below surface and were not fluted, though many displayed overshot flakes with one Bolen point fragment being recovered near the base or were manufactured on thin flakes. Specimens made on thin of level 7. flakes have remnants of the original preform flake surface left
VOL. 65 (1-2) THE FLORIDA ANTHROPOLOGIST MARCH-JUNE 2012
6 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
Figure 1: Digital elevation map of the Big Bend area of Florida depicting the location of the Wakulla Springs Lodge
site and other early, Paleoindian sites in the Coastal Lowlands of North Florida.
intact in the basal area. These flat basal surfaces form flute-like aids in some cases the interpretation of technical details of the features. OSL dating (equivalent dose distributions). The geomorphic
The Sloth Hole site is located on the southern reaches setting of Wakulla Springs is the Woodville Karst Plain (WKP) of the Aucilla River, below Page-Ladson, about 27 km east- (Means 2012, this volume: the remainder of this section is southeast of Wakulla Springs Lodge (Figure 1). The deposits drawn largely from this reference). The karst nature of the there are characterized by levels of silt and shelly silt. Artifacts WKP is reflected in the number and diversity of karst features from an upper Paleoindian level included a Clovis ivory point in the area that includes sinkholes, swallets, springs, and caves or shaft that yielded an age of ca. 12,900 cal B.R along with (both air and water filled). These fluvial systems may have other artifacts. The lower Paleoindian level, which was 60 been active to different extents and at different times, but eolian cm below the Clovis level, yielded artifacts and mastodon deposits are also present. Approximately 9 km northwest of remains with botanical specimens dating to ca. 14,300 cal B.P. Wakulla Spring is a geomorphic region known as the Munson (Hemmings 1999). Sandhills, which represent a relict dune set and demonstrate
The Half Mile Rise Sink site (8TA98) is yet another site in that these features exist in the area. In addition, recent LIDAR the Aucilla River's Half Mile Rise section with Page-Ladson data used to construct a digital elevation model (DEM) points. Though the site has not been radiometrically dated, it (Means, Figure 4, 2012, this volume) show a number of other is but one of several with this distinctive artifact type (Dunbar features of interest in the geomorphology of the area. Depicted 2008). on the DEM is a linear area of higher elevation adjacent to
Wakulla Spring that extends to the south and southwest. This
Local and Regional Geological Context topographic high continues farther southwest and possibly
represents a relict barrier island or other coastal near-shore
The depositional history of the sediment at the Wakulla landform (beach ridge or bay mouth sand bar). These areas Springs Lodge site, because of its importance to the of higher topography are oriented somewhat parallel to the
interpretation of optically stimulated luminescence (OSL) modern coastline, suggesting that they were created by nearages, is the main focus of our interest in the regional geological shore marine processes. If this is the case, the sediment at the context. A detailed view of the regional geological context is Wakulla Springs Lodge site could have been transported to the essential to interpret the granulometry discussed below and near-shore marine environment by a paleo-river system. This
RINK, DUNBAR AND BURDETTE OSL DATING AT WAKULLA SPRINGS LODGE SITE 7
would have had to occur during the Sangamon Interglacial exhibited multi-modal, very poor to poor sorting of grain-size Stage, or earlier, as that was the last time sea levels reached distribution, while Ivester and Leigh (2003) found in riverine this elevation. dunes of the Georgia coastal plain "moderately to moderately
Based upon the existence of relict drainage features that well sorted (0.52t phi < r < 0.87 phi), similar to values for clearly show past, overland flow toward Wakulla Spring the fluvial sands.
(Means, Figure 4, 2012, this volume) it is plausible that Figure 2 depicts the DEM for the area very close to the
sometime during the past glacial stage a paleo-creek/river Wakulla Springs Lodge site. What appear to be crescentsystem flowed to Wakulla Spring, which, at the time, may have shaped features may in fact be eolian dune features that occur functioned as a swallet. Numerous swallets occur in the area at higher elevations than the surrounding terrain. If this is the immediately to the west and north ofWakulla Spring. A number case, it may be yet another set of eolian features for which of swallets in the Wakulla Spring basin have been shown, there is regional evidence in the Southern Coastal Plain dating via dye tracing studies to be connected to the cave systems from about 50,000 to 11,000 years ago (Ivester et al. 2001). that deliver water to Wakulla Spring. If Wakulla Spring was We further note that Pleistocene Coastal Plain dunes typically once a swallet it would have been the termination point for a have crescent shapes. Such dune fields have been identified in stream where sediment laden water would have periodically the Suwannee River area east of the site, along the Flint River flooded into and perhaps overwhelmed the basin such that basin north of the site, and from the Apalachicola River west sediment could have been deposited several meters above the to the Mississippi River. current level of the Wakulla River. This would account for the
current elevation of the sediment and for fluvial signature of Optical Luminescence Dating the sediment samples analyzed by Means (2012, this volume).
This water could have also reworked any sand deposits that Optical luminescence dating utilizes the effects of light
might have been deposited previously at Wakulla Spring exposure on radiation-sensitive defects in the lattice of the
during the Sangamon Interglacial (ca. 75,000 to 125,000 B.P.). quartz or feldspar mineral. Time dependence arises from a From the present authors' perspective, this leaves open two two-stage process in the history of a quartz grain. The first possibilities: 1) Sangamon sands were reworked exclusively stage involves the fact that during deposition of quartz, by fluvial processes even at times as late as 10,000 years ago, sunlight erases previous radiation effects and sets the quartz and 2) Sangamon sands were reworked by fluvial and eolian grain into a zero-age state. The second stage encompasses processes. the burial of the quartz to prevent further light exposure,
It is also likely that eolian processes played a role in during which natural uranium, thorium, and potassium in the the deposition or reworking of the sediment at the Wakulla surrounding minerals, as well as cosmic radiation from the Springs Lodge site. Some of the topography in the vicinity sky, introduce energy into the quartz. A small proportion of is reminiscent of crescent-shaped dune features. Although this energy is trapped inside the quartz grains. This gives rise the granulometric analysis (discussed below) from Wakulla to a time-dependent accumulation of radiation effects that can Spring suggests that the sediment is fluvial in origin it is be measured in the laboratory. plausible that paleo-river systems delivered sediment to a The total dose of radiation experienced in the burial
former nearshore environment (during a previous sea-level interval (which is delivered at a steady rate) is measured in high stand) and after the marine regression these sediments the laboratory by using light to release the stored information were isolated and available for transport via eolian processes. during which the quartz is reset to a zero age state, as happened Ivester et al. (2001 ) discuss the formation of relict dune sets at the time the quartz was exposed to light in nature. Effectively, along river systems in the southeastern United States. the quartz acts as a radiation-sensitive photographic film,
Although the results from Means (2012, this volume) which must remain in the dark to record an image that can suggest that the sediment at the Wakulla Springs Lodge site be converted to time (as opposed to photographic film, which was deposited by fluvial processes, he states "it is likely that the must be exposed to light to record the image). origin of the sediment samples is related to fluvial transport to The time interval of dark storage (burial time) is obtained a nearshore, marine environment, which then was acted upon by establishing the total dose since the last light exposure by other transport mechanisms which might include eolian." (called the equivalent dose in laboratory language), and He also hypothesizes that a paleo-river system, possibly calculating how much dose was acquired in each year of dark nearby disappearing streams, like Lost Creek or Jump Creek, burial. The latter quantity is known as the total dose rate and could have transported sediment from siliciclastic-rich areas is the amount of dose each year delivered by the uranium, to the north and west. thorium, and potassium in the environment and the amount
Previously recognized eolian dune features sourced from delivered downward from the sky due to cosmic rays. The fluvial contexts have been recognized in the southeastern burial time is given simply by the ratio of the dose acquired United States. Transported by eolian processes, dunes in during dark storage (equivalent dose) to the total dose rate and adjacent to river basins in the Southeast were formed (dose per year). Equivalent dose divided by dose per year by sediments of fluvial origin (Ivester et al. 2001; Ivester gives the number of years since burial. and Leigh 2003; Leigh 2008). Otvos and Price (2001 :152) In order for OSL to give an accurate burial age in the
found granulometric results from Louisiana dunes that archaeological context, the buried grains must remain in their
8 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
Figure 2: Processed LIDAR elevation model of the Wakulla Springs area showing multiple crescent-shaped features
that are reminiscent of crescent-shaped dune features.
burial location relative to artifacts that are buried with them. issue). Alternatively, incomplete zeroing at burial may be the If the grains that are buried move downward relative to the cause when an increasing proportion of older ages is found artifacts, then grains that were buried at shallower levels with decreasing mask size. may be found with the artifacts, and the grain-age will be too
young. Conversely, if grains that were buried move upward Archaeological Methodology relative to their original burial location, they may arrive into
locations with artifacts that were buried at a more recent time Preparations for the April to May 2008 field season in the past. In the latter case, the grain-age will be older than included establishing control points. They were recorded in the age of artifact burial. the UTM Zone 16 metric grid by using nearby existing datum
Also of importance is the use of variable aliquot size points for elevation (NAD88) and conducting a survey-grade (i.e., the number of grains tested) as a tool to identify the GPS satellite session (with the assistance of the Bureau of possibility of incomplete zeroing at burial or post-depositional Beaches and Coastal Systems of the Florida Department of mixing of grains from higher or lower levels relative to a Environmental Protection) to establish horizontal control dated sample. The spectrum of doses (or ages) obtained in an (NAD83). From the control points, a survey utilizing an individual sample is plotted for each aliquot size, and these electronic total station was used to establish the site grid in plots are compared among aliquot sizes. An interpretation the excavation area and to gather topographic data for the area of the true burial age of a level can then be obtained by use of the Wakulla Lodge overlooking the springs. The collected of statistical methods combined with information from the data were then electronically mapped (Figure 3) using geological context. An increase in the proportion of older AutoDesk Civil 3D engineering software (Dunbar et al. 2008). ages or doses as the aliquot size is reduced indicates the It should be mentioned that we were unable to calibrate our presence of a significant component of older age grains is depth measurements to the vertical datum that Calvin Jones present, whereas if the increasing proportion of younger ages used, because his benchmark was altered and moved by the is found, the presence of a significant component of younger subsequent sewer line installation. age grains is present. In each case the cause may be related Though an initial ground penetrating radar (GPR) survey to bioturbation of grains after burial (see Thulman 2012, this was conducted before the fieldwork, its results never became
RINK, DUNBAR AND BURDETTE OSL DATING AT WAKULLA SPRINGS LODGE SITE 9
available. After the fieldwork, a second GPR survey was A study to determine thermally altered versus non-thermally carried out by staff from the Florida Geological Survey with altered rock is worth doing, however the use of a digital assistance from the Public Lands Archaeology program of microscope, such as a Keyence VHX with optics capable of BAR. Closely spaced passes of the GPR unit allowed creation 1000X magnification or higher is needed to determine if the of a three-dimensional GPR tomography, one that identified patinated chert specimens display the characteristic change the lodge's pipeline networks and the top of limestone bedrock caused by thermal alteration where the individual crystals below the overlying sediments (Smith 2012, this volume), become smother and no longer resist fracture along crystal
Sampling during the field project included artifacts, boundaries as a result (Purdy 1974:51-52, Figures 11-12). bulk sediments, OSL sediments, and sediment monoliths.
Sediment was sifted through 1/4-inch (6.4 mm) screens for Dating Methodology the first 90 cm then through window screen (1.4 mm) in the
lower levels. Samples of oxidized wood, most likely the An important field objective and justification for including
decomposing remains of tree roots, were also recovered and both geoarchaeological and geological consultants on the sampled; however, nothing suggesting charcoal from a fire pit project was to insure, to the extent possible, the sediment wa identified, and none of these samples are deemed worthy column proposed for OSL sampling represented apparently of C dating. Paleoindian and Early Archaic Bolen people undisturbed areas of the stratigraphic column (but see Thulman did not thermally alter rock for tool making, but some Early [2012, this volume] for a discussion of bioturbation at the site). Archaic peoples did purposefully heat treat chert. Thus, a Both OSL sampling loci were determined to represent unified, comparison of thermally altered versus unaltered artifacts was undisturbed locations based only on visual inspection and undertaken to investigate whether artifact migration through excavation records of zones removed to expose profiles. Jack the sediment might have taken place and to estimate the Rink and Kevin Burdette conducted the OSL sampling, taking Middle/Early Archaic boundary. Howeverthis study yielded four samples from Unit B and three samples from Unit C. equivocal results, because determinations of the presence of After each sample was taken the gamma dose rates from each thermal alternation were ambiguous and difficult to confidently sample location were calculated using a gamma spectrometer identify without a reference collection and better protocols. employed in situ, and neutron activation analysis was also
M eter .......... :: :I@ ...... W 'ei
Figure 3. Digital, one-foot pixel, 2010 aerial photograph of the Wakulla Springs Lodge site depicting 10 cm topographic contour intervals as well as the locations of the survey control points, WAl and WA2. The topographic high point is 5.8 m above sea level on the contour line passing through WAl and the topographic low is about 1.9 m above sea level at the shoreline.
10 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
used to check the in situ results. Gamma spectrometric gamma sample were optically bleached by blue light illumination for dose rates were used as they better represent the full volume 40 seconds, followed by a 10,000 second pause and another of sediment irradiating a given sample. Initial studies utilized 40-second illumination. For the dose recovery test, the aliquots large several thousand-grain aliquots (8 mm diameter), and were given a known dose. Both tests continued with the initial ages were calculated using the central age model. standard SAR protocol except the preheat temperatures varied Those results were presented at the 2009 Society for American (160, 200, 240, 2800 C), with 3 aliquots from each sample Archaeology annual meeting (Rink et al. 2009). However, we receiving a different preheat temperature. For each sample continued the age evaluation by reducing the aliquot size on the dose recovery test was used to determine which preheat all samples to approximately 20-50 grains (1 mm diameter aliquot size) and used the minimum age model (Galbraith et lest b 5outh Wall Protile, Wakulla prings Lodge bite 81WA32 al. 1999) to establish the burial age. .
Sampling locations for the OSL samples were taken in Units B and C (Figures 4 and 5) at the elevations that the Paleoindian blade and endscraper, respectively. Samples were also taken above and below these elevations in order to develop a chronology within the Paleoindian levels.
All optically-dated samples were processed at the School of Geography and Earth Sciences at McMaster University under UV-filtered subdued orange light. Pure quartz grains were obtained using standard OSL preparation methods that
include HCl and 11202 digestions to remove carbonates and organics, respectively, sieving to obtain desired grain size, heavy liquid separation using lithium polytungstate to remove heavy minerals and feldspars, HF digestion to remove the outer alpha affected layer, a second H202 digestion to remove any Figure 4: Unit B south wall profile taken after the test remaining feldspars and any fluorides that may have formed unit had been expanded a half meter east and south. The during the HF digestion, and finally resieving to remove any OSL sample locations shown here represent the relative grains that no longer fall in the desired size range (90-150 positions of the sample location before this test unit was microns). expanded. The sediment profile shown here was drawn
Calculated gamma and beta dose rates were based on based on post-processed sediment sample desiccation to
neutron activation analysis (NAA) of 232Th and 40K and detect boundaries in sediment color, and no interpretadelayed neutron counting (DNC) analysis of 238U (conducted tion of the cause of the coloration changes is available. at the McMaster University Nuclear Reactor). Untreated The blade tool was recovered at OSL sample location 3B.
subsamples of the original samples were used to determine Scale bar is for vertical and horizontal dimensions. the elemental concentrations of radioactive 218U, 232Th, and 40K (Table 1). NAA/DNC-based dose rates were calculated assuming radioactive equilibrium in the 238U and 212Th decay Test C East Wall Profile Wakulta Springs Lodge Site WA329 chains.
Luminescence measurements were conducted on a RISOeEE OSL/TL-DA-I 15 reader using blue light LED stimulation (470 nim) and a 7 mm-thick Hoya U-340 filter (270-400 nim). A i "" ...........................
calibrated 9Sr beta source was used to perform laboratory .............. ..,: ..,. ....
irradiations. The single aliquot regeneration (SAR) protocol ... ,,. ........ ., .... i
(Murray and Wintle 2000) was conducted on a minimum of 24 aliquots to determine a final equivalent dose (DE). Quartz iiii! grains, between 90-150 microns, were mounted with silicone spray on aluminum discs using a 3mm and a 1mm mask and were illuminated for 100 seconds at 1250 C. The background (the last 4 s) of the OSL decay curve was subtracted from
the "fast" component (first 0.4 s) to determine the samples luminescence signal. Only aliquots whose recycling ratios Figure 5: Unit C east wall profile depiction the locations of were within 10 percent were accepted for equivalent dose DEB OSL sampling. The endscraper was recovered at OSL samdeterminations. Linear plus exponential fits were made to the ple location 2C. The sediment profile shown here was drawn dose response data to determine individual D B values, based on post-processed sediment sample desiccation to deA thermal transfer test and a dose recovery test were tect boundaries in sediment color, and no interpretation of performed to determine the final DE preheat temperature the cause of the coloration changes is available. Scale bar is (Madsen et al. 2005). For both tests, twelve aliquots from each for vertical and horizontal dimensions.
RINK, DUNBAR AND BURDETTE OSL DATING AT WAKULLA SPRINGS LODGE SITE 11
Sample #2 Frequency Plot
16-c~ ---- __- _____ 0___16
12 I .---------- ...-*.--.-..- .. .....
-3.0 -2.0 -1.0 0.0 1 .0 2.0 3.0 4.0 5.0
Grain Size (Phi)
Figure 6. Granulometry of Sample 2 (Means 2012, this volume). This sample was recovered in close proximity and elevation to our OSL B2 sample.
temperature produced a D E closest to the given dose. Once this significant feldspar contamination. Finally, moisture contents preheat temperature was determined, the thermal transfer test were measured in the lab from the recovered sediment and was analyzed to insure there was no induced charge transfer at used for the dose rate calculations. that given temperature.
A feldspar contamination check was also performed on Geological Methodology each sample to insure purity of the quartz grain separates. An initial DE was estimated by comparing the natural OSL The Florida Geological Survey collected five sand
signal (preheat T -200C) of 3 aliquots to the regenerated samples from the top to the bottom of the sediment column in OSL given by a single dose. A second identical regeneration Unit B to determine sand-size granulometry (see Means 2012, dose was applied to the same aliquots and the IRSL signal was this volume). measured. If a ratio of IRSL to regenerated OSL signal was less than 1 percent for all aliquots, it is assumed there is no
WA329 Test B SE, Total Artifacts by Level
70-80 cm A.....n.Yonger.eve.
1-3-11 cm. -1.0 0. 20 302.
E o + 11 -1 0 m 1 I!!!!!!!! .. .iI+++++++++++++++++++++++++u n1++++litil
LGrain O...... (P
130-140. m of Sample 2 (Means 2012, this v Ti sample wa
140-o1v0rc i a e a to ou O B2 sample.
1empea0cm p u a D E cloes to th g dose. O t s f c n F moisture
1Figu gv b a .. d Areg oaltifc Unts B 10 cmleel i U nit B
doewsaple o h ae lqos n h I sina ws t ilm)
80-40 cm t 2n+ E ++
12 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
Geological and Archaeological Results of accumulating in wind-blown dunes and fluvial sediments.
Granulometry, though not conclusive, can be a good indicator
Sedimentological Analysis of how sediment accumulates. Notwithstanding this, we
believe that the granulometrics do not categorically identify
We report here on the sedimentological analysis of the sediment as being of purely fluvial origin. In our opinion, sediments from the site (Means 2012, this volume). This analysis though fluvial signatures are present here, there is a high is essential to best interpret equivalent dose determinations in likelihood that eolian processes participated in the deposition OSL dating. It can help classify some equivalent dose signatures of the sediment. Fluvial deposition at Wakulla is admitted as the type resulting from incomplete zeroing during fluvial as a possibility (Means 2012, this volume). This is because deposition. All samples studied showed a similar bimodal proxy and geologic evidence indicates that inland water tables distribution of grain size and were negatively skewed. Figure in Florida were lower, sometimes considerably lower than 6 (sample 2 of Means 2012, this volume) is representative of present (Grimm et al. 1993; Grimm et al. 2006; Dunbar 2006b; all five granulometric samples, whose average mean grain Willard et al. 2007) during marine isotope stages 2 and 3, from size had a standard deviation of 0.9493, indicating that they about 59,000 cal B.P. to 12,000 cal B.P. (Wright 1999). In sum, were moderately sorted. This sample had a mean grain size of we wish to leave open the possibility that the sediments we 2.2755 phi, grain size standard deviation of 0.9546, skewness date here were deposited initially by fluvial deposition and of -0.4352 and a kurtosis of 2.6735. Negative skewness can later reworked by eolian processes. No sedimentary features represent winnowing effects commonly seen in beach and consistent with fluvial deposition were found in the sections, nearshore marine environments. The bi-modal distribution though the appearance of the sediment in profile was entirely suggests that two different current transport regimes acted consistent with eolian deposition. upon the sediment. A plot of skewness versus standard
deviation of the grain size data was constructed (Means 2012, Archaeology Figure 3, this volume) in an attempt to determine whether the
sediment transport and depositional mechanism was wave
action (beach) or fluvial processes. The five samples fell We excavated approximately 46 m3 of sediment fromseven within the river sand (fluvial) portion of the graph, suggesting test units covering a horizontal area of approximately 35.5 i2. that the sediment, at some point in time, was acted upon by About one third of the total volume excavated consisted of fluvial processes. disturbed sediment resulting from lodge construction or from
Our OSL analysis was performed on the 90-150 micrometer subsequent maintenance such as the replacement sewer line grain size fraction (phi 3.47 to 2.74). This corresponds to the that Calvin Jones' crew excavated in 1994 to mitigate the large peak in Figure 6 near 3.0 phi. This is a sand-size capable impact on the archaeological resource (Jones and Tesar 2000).
WA329 Test C NE, Total Artifacts by Level
70oo 0 !Palso !ininLvs
L o c m! !!!! !!! !!!!! !!!!!! !.!! ,
1010 30 40 50 60 70809
Figure 8. Total artifact counts by 10 cm level in Unit C
RINK, DUNBAR AND BURDETTE OSL DATING AT WAKULLA SPRINGS LODGE SITE 13
I'l ... Depending on location, artifacts from approximately
75 to 100 cm below surface represent the Paleoindian component(s) of the site. In Unit B, an Early Archaic Bolen point was recovered about 20 cm below a Clovis-like blade knife, but it clearly had been displaced. The Bolen component of the Wakulla Springs Lodge site had been previously dated to about 10,500 cal B.P., and was unequivocally found above the Paleoindian levels (Tesar and Jones 2004:155-156). The Bolen point from Unit B in 2008 came from a deeper level than the Bolen-age artifacts Calvin Jones's crew identified, also indicating that it was displaced.
A We encountered numerous dark-stained areas classified
as non-cultural features, most of which were the remains of Paleoindian unifacially retouched flake knife 2- 4 organic-rich, sediment-filled tracks of stumps, roots, or animal
with thinning flute on dorsal face CM CM
8WA329.08A.30.160.01 burrows. One particularly large feature appeared to r(
Scanned images of rotated object prepared by Louis D. Tesar 16Apr08 the work of a bulldozer that had successfully uprooted a tree and then backfilled the excavation hole in Unit A. These types Figure 9. Paleo blade recovered from Unit B. of features predictably yielded younger artifacts from deeper
levels of the site than they should. The Bolen point in Unit B is judged to represent the same type of displaced object, The most substantial ground disturbance was in Unit A and the fallen from its original level of deposition to a much deeper western half of Unit D where pipeline ditches were more than one through some type of opening, albeit temporary, in the 1 to 1.5 m deep. Shallower cast-iron storm water drains were ground. The Bolen point's angle of repose was not flat; rather encountered in Units B and E with ditches extending about 70 it was on edge with its distal end facing downward. It is why cm deep. A number of abandoned sprinkler and water lines we believe the Bolen point represents an artifact displaced by were encountered in Units B, F, and 1. These shallow pipes bioturbation (Thulman 2012, this volume). Calvin Jones also were buried about 20 to 40 cm deep. Ground disturbance of the found a downward displaced Bolen point in his TS 10 test area upper levels of the site was also detected and varied in depth in a dark-stained feature. from about 15 to 40 cm. This more generalized disturbance Our investigation adjacent to Jones' TS 10 area identified
appeared related to landscaping of the park in 1937 when the Paleoindian stone tools at 97 cm and 105 cm below surface Wakulla Lodge was built. To date, few historic records or in Unit B (Figure 4) and Unit C (Figure 5), respectively, as photographs have been found that provide details about lodge well as other artifacts to depths of 140 cm below surface. construction as it relates to ground disturbance and the types Differences in coloration of the sediment are shown here, of excavation machinery used. A bulldozer is suspected to and in this context have not been attributed to stratigraphic have been used for landscaping and tree removal, levels. The source of the color variation is not ascribed to any
Vii ii !ii I LR hiiii
Figure 10. Endscraper recovered from Unit C.
14 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
particular process because studies of this were not undertaken. The lower levels of the site yielded three notable artifacts The shallowest of the Paleoindian artifacts, a formalized blade and a number of tools of lesser importance such as biface tool, has no counterpart in the Early Archaic Bolen assemblage, fragments and flake tools. The more noteworthy artifacts but appears to be very much like a Clovis blade or Clovis-like include the blade from Unit B and an endscraper from Unit blade tool, a tool type found elsewhere in Clovis sites in North C. Perhaps the most unusual stone artifact is an item of America (Bradley et al. 2010). Other than this and one other adornment. It is a stone seed bead recovered from Unit D, 1.35 Clovis-like blade tool recovered from controlled excavations, m below surface (Glowacki 2012, this volume). there is no other evidence for a Clovis component occurring at The blade knife from Unit B is similar to or is of Clovis Wakulla Springs. manufacture (Bradley et al. 2010: Figure 9) made from a
The debitage distributions from Unit B (Figure 7) show large prepared blade struck from a blade core. Though we are that the most intensive occupation occurred within the top uncertain what tool assemblage it belongs to, its recovery from approximately 90 cm of sediments found here (referred to here the Wakulla Springs Lodge site in the area where Calvin Jones as the upper strata). These strata also had higher percentages recovered Simpson and Page-Ladson artifacts suggests that it of heat-treated lithics, a typical signature of lithics from the either belongs with that assemblage or another site component Middle Archaic and more recent periods. The distributions yet to be identified, most likely Clovis. The tool appears below about 90 cm show a dramatic drop in the total number of to have been a hafted knife on its proximal end, backed on artifacts in one unit (B), which is to be expected if we presume one lateral side and sharpened on the other. High-resolution smaller populations in the Early Archaic and Paleoindian magnification shows the cutting edge was smoothed during periods and thus less material culture in those levels. The usage, which suggests use as a knife on soft tissue. distributions in Unit C (Figure 8) do not show the clear trend The endscraper manufactured on a flake from Unit C of decreasing debitage with depth. In fact, three 5-cm intervals (Figure 10) could comfortably fit into any Paleoindian toolkit in the lower portion of the site (95-100 cm, 110-115 cm, and from Florida (Daniel and Wisenbaker 1987). It is manufactured 125-130 cm) show more debitage than the 20-40 cm interval, from a medium-sized flake. High-resolution magnification of We acknowledge that effects of pedoturbation could lead to the worked edge in two places displays a smoothness caused some of the characteristics in the distributions of debitage during usage and indicative of activities such as hide working. presented here. Suspected heated lithics, all heavily patinated, Considerable coloration differences in the sand were were also present in the lower levels but at lower percentages. found. Faint coloration differences can more easily be
Figure 11. 1970 Photograph of the Simpson point lying on the upper maxilla of Mammut americanum (mastodon) in the lobby of the Wakulla Springs Lodge. Photograph provided by Dan F. Morris.
RINK, DUNBAR AND BURDETTE OSL DATING AT WAKULLA SPRINGS LODGE SITE 15
detected once the samples from each level are desiccated and With the exception of the Clovis-like blade knife from high-resolution images acquired through a flat bed scanner Unit B, the stone tools recovered from the Paleoindian context with samples placed in optically clear sample bags. Changes at the Wakulla Springs Lodge site are not Clovis, yet the site in sediment coloration might be due to decomposing organics has yielded Page-Ladson and Simpson types. The Simpson and differential leaching and therefore represent differences from Wakulla (Figure 11), once on display in the Wakulla in post-depositional development. We also considered that Lodge, disappeared just before the State completed purchase the color differences of each level represent distinct episodes of the property. A photograph taken by Dan Morse remains as of deposition through time but were unable to reach a firm the only documentation of its existence. Neither Page-Ladson conclusion on the sources of the color variations. However, nor Simpson types have ever been found in a Clovis context, we note that the level of the Paleoindian blade knife in but both occur in the deepest levels at Wakulla found by Jones Unit B originates in dark-colored sediment compared to the (Jones and Tesar 2000). Images of several of the Paleoindian endscraper in Unit C, which was in light-colored sediment. artifacts found at Wakulla and other locales are shown in Figure The Paleoindian endscraper may therefore represent a second 12. We therefore propose that the lithics shown in Figure 12 Paleoindian site component. are part of a Simpson toolkit with the possible exception of the
The Simpson type as first proposed by Ripley Bullen Clovis-like blade tool. However, if the blade tool is related to (1968, 1975) has been applied to Paleoindian recurvate, the Page-Ladson/Simpson assemblage, it is the only artifact waisted points in Florida by many researchers (see for showing unequivocal correspondence to Clovis example Daniel and Wisenbaker 1987). However, Bullen's
Simpson type is represented by a heterogeneous assemblage Results of the OSL Dating of the of similar-shaped points that are often inseparable from the Wakulla Springs Lodge Site waisted Suwannee type in his type case collection housed at
the Florida Museum of Natural History in Gainesville (Dunbar Luminescence measurements were made on single and Hemmings 2004). In this paper, we restrict the Simpson aliquots of three different diameters: 8 mm, 3 mm, and 1 type to a single, unified category of recurvate, waisted points mm. Figures 13 and 14 show the results for two samples: B3 dominated by percussion flaking and having extreme width to and C2. This comparison is often made as a basic check for thickness ratios. incomplete zeroing at burial but can also be used to identify
Figure 12. The proposed Simpson tool kit including 1) Simpson reform also used as a core to yield flakes for small tool manufacture, recovered during Calvin Jones dig Wakulla Lodge site; 2) Simpson bifacial knife from the Chason Cache (8CA185)(a similar point was on display at Wakulla Lodge but disappeared prior to state's purchase of the park); 3) flake preform taken from a biface tool core such as depicted in 9.1 recovered during Calvin Jones dig, Wakulla Lodge site; 4) a Clovis-like blade from Unit B Wakulla Lodge site (also shown in Figure 8); 5) a Page-Ladson projectile point recovered during Calvin Jones dig Wakulla Lodge; and 6) a seed bead manufactured from a small quartz pebble recovered from Unit F.
16 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
effects of disturbance of grains after burial. Sample B3 not see any evidence for incomplete zeroing at burial, which (Figure 13) exhibits a shift to larger proportions of aliquots generally would be evidenced by a strong spreading to higher with smaller equivalent dose (DE) as the aliquot size decreases. doses with decreasing mask size. While the 8 mm mask (diameter) distribution shows no doses From these observations we believe that the mean of the lower than about 6.5 Gray (Gy), we see that the 1 mm mask equivalent dose distributions are not a good indicator of burial aliquots show 5 of 7 aliquots with doses lower than 6.5 Gy. age, and that there appears to be a mixing of older (higher We see a similar trend for sample C2 (Figure 14), where all equivalent) dose grains with younger grains. We believe that aliquots at 8 mm mask have doses larger than 7 Gy, while a our observation of increasing proportion of younger grains large proportion is less than 7 Gy in the 1 mm mask size. We with decreasing mask size is a result of dilution of the older also observe one aliquot at around 14 Gy, which is a dose that grain component by increasing contributions to the light was not observed in either of the other two mask sizes. Since sum by the younger grains on the smaller aliquots. From this the trend is to lower doses with decreasing mask size, we do analysis, we have chosen to calculate the burial age based on a statistical analysis of the 1 mm distribution that seeks to Wakulla B3 find the minimum possible burial age of the sample, called
Final DE Distributions the minimum age model (Galbraith et al. 1999). The computer
programs to calculate the central and minimum age model 8 mm Mask Size equivalent doses and other characteristics were provided by S.
(2008 Data) Huot (University of Quebec at Montreal).
N-24124 We compare the results of the central age model and the
minimum age model in Tables 2 and 3. It is seen that ages based on the minimum age model are considerably younger than that of the central age model throughout the deposit. The OSL ages are basically comparable to calibrated 14C ages (calendar years S[cal] BP), except that all 14C ages are calculated based on years before 1950 (B.P.), while the ages reported here are calculated in years before 2010. We see that Unit B has a very old basal 2 .. . . age below the archaeology of about 25 to 29,000 years ago,
3 4 5 6 7 9 10 while samples from within the upper dated levels range from
about 14,900 + 1400 years ago to 10,600 +1000 years ago.
3 mm Mask Size
5 (2008 Data) Unit C shows a lower sample within the archaeology dating
N 24/24 to 18,000 + 4500 years ago to an uppermost sample of 15,100
4 + 1600 years ago. Further support for using the minimum age
model is found in the over-dispersion parameter (Table 3), which in nearly every case exceeds a threshold of 20 percent,
which is considered to be evidence by most workers of a mixed age population of grains.
The minimum age model results are in stratigraphic order
1in both Units Band C or are statistically indistinguishable with depth within the unit. Overall they exhibit a clear trend 2 3 4 5 5 toward younger dates with decreasing elevation (Figure 15).
25 We acknowledge that this type of age-depth trend has been
1 mm Mask Size interpreted as a result of biomantling, where upwardly mobile
20 (2010 Data) sediment porsieyburies atfcsand moiisage
N- 7/48profiles (Wilder et al. 2007). Indeed, we strongly suspect that 15 biological agents such as ants may be moving sediment grains
~upward (without zeroing them) to produce older than expected 8 aliquots observed in our distributions (Thulman 2012, this
10 volume). In this case, however, not all grains are moved to the
surface as might be suspected in a full biomantling scenario. 05 The use of the minimum age model to isolate the grains that
represent the best burial age (youngest grains) may not be 0.0 ..,able to remove all the effects of upwardly mobile older grains,
2 8 9 10 though it is currently the best method available to do this.
DE (Gy) Sample C3 (Table 2) appears to be relatively older than
the trend; however, our confidence in the age of this sample Figure 13. Wakulla B3 Histogram for 8 mm, 3 mm, and is relatively low because of the small number of aliquots that
1mm masks. could be used from the analysis at the 1 mm size (4 of 48)
RINK, DUNBAR AND BURDETTE OSL DATING AT WAKULLA SPRINGS LODGE SITE 17
(this resulted from the restriction that we used only aliquots 4.9 that pass a criterion that the palaeodose error is less than 10 84 03
percent). If we restrict our analysis to the other samples but E 4.8 also exclude B 1, which was below most of the debitage, a -B83 line fitted to the data produces a slope that gives an average, 4. -i--I i_ Palaeoblade Knife FS 160
sediment accumulation rate for this portion of the deposit of Paer a82,
CO Palaeo endscraper FS27
0.046 meters/thousand years, or about 5 cm per thousand years. 4.6 C2
This represents an average accumulation of about 1 mm per 20 0-0 years. The position of two significant artifacts found in the 4.5 i
units are also shown on Figure 15 and correspond to sample C1 Bolen FS695
locations B3 and C2, which date to 13,700 + 1100 years ago 4.4 w 4 B1
and 15,600 + 1900 years ago, respectively. At this two sigma error, the levels from which the artifacts were excavated are 4.3 statistically indistinguishable in age. The Bolen point found in 5 10 15 20 25 30 35
a non-horizontally disturbed context comes from near the level Minimum OSL Age (ka)
of C 1, which happens to also show the largest age uncertainty.
Figure 15. Minimum OSL age Wakulla Springs Lodge Wakulla C2 Unit B and C combined with artifact depths on chart
Final DE Distributions
8 mm Mask Size There may be a factor of disturbance in this portion of the
(2008 Data) deposit that has also increased the uncertainy of the OSL age.
N=24/24 Its age uncertainty stems directly from the relatively large
spread in the 1 mm aliquot DE distribution, which may be due to a larger amount of mixing there than in other samples.
0 The chemical analyses and radiation dose rate data are
given in Table 3. We note the extremely low concentrations of
2 i U, Th, and K present, which are not surprising in view of the
lack of any material other than quartz. This leads to very low gamma and beta dose rates. We see for the most part extremely 1 2 good agreement between the in situ measurements of the
3 mm Mask Size gamma dose rate and those calculated by use of U, Th, and K
N 5/24 values with moisture correction. The discrepancies pr
0,8 arise from small variations in the dose rate over the 30 cm
radius sphere of influence of the gamma rays. Finally, we can 0. :3 see that the cosmic dose rate is by far the most important of all,
in most cases accounting for 50-60 percent of the total dose rate. This is important because it mitigates against uncertainty 0.2 .......in the ages that arise from moisture content variations through
i time, because cosmic rays are not significantly affected by
2 8 4 0 moisture, while beta and gamma dose rates are.
8 Overall, the minimum age model age ranges (last column
1 mm Mask Size of Table 1) constrain the archaeological levels to be as young
5(2O100ata) as 13,500-13,700 years ago in all but two samples, which
N 18 /48 could be as young as 12,600 and 9600 years ago, respectively.
Considering the older end of the uncertainty ranges of the ages, we can state that the minimum model ages constrain the o archaeological levels to be as old as 22,500, but as young as
2 11,600 years ago.
* j0 Final Discussion and Conclusions
2 4 6o 10 12 14 1 Our observations that smaller aliquots yielded dose
D (Gy) distributions with larger numbers of smaller equivalent doses
has also been observed in other sandy, shallowly-buried upland archaeological sites in central Florida at the Avon Park Figure 14. Wakulla C2 Histogram for 8 mm, 3 mm, and Air Force Range (Wilder et al. 2007). Their studies compared
1 mm masks. single grain aliquots to 8 mm diameter aliquots. They attributed
18 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
Table 1. Wakulla Springs Lodge Chemical Analyses and Dose Rate Data.
Total In-situ NAA
238 232 Water Cosmic Dose Beta Dose Total In-situ NAA Total Dose
U Th Internal Gamma Gamma
K (%) Content Rate Rate Rate
Sample (ppm) (ppm) [a] (%) (tGy/a) tGy/a) Dose Dose Rate Dose Rate
Name [a] [a] [a] (%) (Gy/a) (Gy/a) Rate (ptGy/a) (pGy/a) (Gy/a)
[b] [c] el (tGy/a) [d] [e]
OSL- 0.24 0.72 0.0512 2.69 204.61+20.4 84.5+12.9 10.1+2.3 72.5+11.4 371.7+17.4
OSLB-P OSLB- 79.5+4.0 378.7+ 13.7
OSLB- 0.42 0.92 0.0563 2.59 200.79+20.1 114.612.8 10.1+2.3 102.111.4 427.7+17.3
OSLB- 78.37+3.9 403.9-13.6
OSLB- 0.19 0.48 0.0291 2.23 198.27+19.8 56.4+12.8 10.1+2.3 50.4+11.3 315.3+17.3
OSLB- 0.28 0.76 0.0522 2.69 192.04+19.2 90.3+12.8 10.1+2.3 78.3+11.4 3
OSLC- 0.22 0.53 0.0292 2.75 205.4820.5 60.9+12.7 10.1+2.3 55.6+11.2 3.
S 0.21 0.50 0.0307 3.02 197.7119.7 60.0+12.7 10.1+2.3 53.5+11.2
0.28 1.02 0.0403 2.50 194.43+19.4 87.5+12.8 10.1+2.3 87.5+11.6
Results shown in bold are our best gamma and total dose rate estimates. They utilize the in-situ gamma dose rate
rather than the NAA-derived gamma dose rate. pGy/a is 1 x 10-6 Grays per year.
[a] U, Th, and K values were determined by NAA on sub-samples derived from the OSL samples prior to chemical
[b] Water content was calculated as a fraction of dry weight determined from laboratory measurements.
[c] Cosmic dose rate value calculated using a linear accumulation model for an accrual of the sediment above the
sample and an overburden density of 2 g/cm3.
[d] These gamma dose rates were calculated based on use of a gamma spectrometer employed in-situ.
RINK, DUNBAR AND BURDETTE OSL DATING AT WAKULLA SPRINGS LODGE SITE 19
Table 2. Wakulla Springs Lodge Optically Stimulated Dating Results (1mm Diameter Aliquots).
Sample Depth Elevation DE (Gy) DE (Gy) OSL Age (ka) OSL Age OSL Age
(cm) NAVD Central Age Minimum Age Central Age (ka) Range (ka)
88 (m) Model +/- Model +/- Two Model Minimum Minimum
Standard Sigma Age Model Age Model
B4 84.9 4.838 5.7 +/-0.3 4.0+/-0.3 15.0+7-1.1 10.6+/-1.0 9.6 to 11.6
B3 97.3 4.714 6.0 +/- 0.6 5.5 +/- 0.4 14.9 +/- 1.6 13.7 +/- 1.1 12.6 to 14.8
B2 104.1 4.646 8.5 +/- 0.6 5.1 +/- 0.4 24.7 +/- 2.1 14.9 +/- 1.4 13.5 to 16.3
B1 134.7 4.340 14.8 +/- 0.7 9.9 +/- 0.6 40.3 +/- 2.8 27.0 +/- 2.1 24.9 to 29.1
C3 83.5 4.852 7.3 +/- 1.2 5.4 +/- 0.5 20.5 +/- 3.4 15.1 +/- 1.6 13.5 to 16.7
C2 105.3 4.634 7.7 +/- 0.5 5.4 +/- 0.6 22.2 +/- 1.8 15.6 +/- 1.9 13.7 to 17.5
C1 119.9 4.488 11.8 +/-0.7 6.8+/-1.7 31.3 +/-2.4 18.0+/-4.5 13.5 to 22.5
their observations to upward mixing of older grains into levels. And three of the sites (Page-Ladson, Sloth Hole, and younger units, which we also believe to be the case at Wakulla Wakulla Springs Lodge) have Page-Ladson points. (Thulman 2012, this volume). For example, at both the Burnt Before World War II, based on geologic observations Hammock (8HG887) and Arbuckle Terrace (8P06091) sites before the advent of radiometric dating technology, Clovis in Avon Park, large aliquot age estimates were consistently (then lumped with Folsom) was originally estimated to be a factor of two times larger or older than the single aliquot 13,000 years old and younger (Antevs 1936). After the war estimates. This was true for a time range of 300 to 7400 years radiocarbon dating yielded age estimates for Clovis of 11,500 ago. In addition, in the more deeply-buried levels at the same to 11,000 cal B.P. (Haynes 1971) apparently showing the site, the larger aliquots yielded ages a factor of 3 to 7 times earlier 13,000 cal B.P. age evaluation of Antevs to be too older than single aliquot results over a time range of 6600 to old. However, it soon became apparent that a fundamental 19,000 years ago. Our results at Wakulla are consistent with assumption of radiocarbon dating, that the amount of '4C has upward mixing of grains from older layers, benchmarked been constant in the earth's atmosphere, was false, a problem primarily on the observation that unless diagnostic artifacts that resulted in the development calibration programs (Stuiver are downwardly mobile, the sediments are much too old and Reimer 1993). With the ability to calibrate radiocarbon (based on the central age model) to host artifacts. The use age to calendar years, the age of Clovis was reevaluate once of the minimum age model to isolate the true burial age may more to be 13,500 to 13,000 cal B.P. Yet another problem suffer in that it might not be able to remove all the effects was recently detected, that of inaccurate age evaluations on of upwardly mobile older grains. As such, the minimum age Pleistocene and early Holocene animal bone (Stafford et al. model may not be as young as the true burial age, but it is the 1991) and poor choices of the types of samples selected for best estimate we can make at present. This is because further dating (Lowe et at. 2001). Because the age of many Clovis analysis of more grains might reveal even a higher proportion sites were based on radiocarbon samples taken from bone and of younger grains than found thus far. other less desirable samples such as bulk sediments, the age
In the past several decades, there has been a number of of Clovis sites has been rigorously reevaluated with the oldest recalculations for the age of the Clovis tool-making tradition, Clovis sites, one of which is in Florida, now estimated to be no and now it is generally accepted to have begun around older than about 13,000 cal B.P. (Waters and Stafford 2007). 13,000 cal B.P. While there is growing consensus that there There are now three North Florida sites with age was an earlier pre-Clovis occupation in North America, the evaluations on archaeological components that are 500 to nature of the assemblage and the dates for that occupation 1,500 calendar years older than Clovis: the Page-Ladson are not yet settled. The Wakulla Springs Lodge excavations site (ca. 14,400 cal B.P.) (Webb and Dunbar 2006; Dunbar go some distance toward providing greater resolution on 2006a) and the Sloth Hole site (14,300 cal B.P.) (Hemmings those questions. The dated Page-Ladson and Sloth Hole sites 1999, 2004) in the Aucilla River, determined by calibrated indicated a pre-Clovis occupation in North Florida and the radiocarbon dating, and the results of OSL dating presented undated Wakulla Springs Lodge and Half Mile Rise Sink sites here using the most conservative young end of the two became suspected pre-Clovis sites based on the occurrence of sigma error range for the Wakulla Springs Lodge site (ca. Page-Ladson points at both of them. Three of the sites (Page- 13,500 years ago) age (see Table 2). There may also be a Ladson, Sloth Hole, and Half Mile Rise Sink) have stone Clovis component at the Wakulla Lodge site that has yet to artifacts and Mammut americanum (American mastodon) as be positively identified. Though the bulk of the Paleoindian well as other late Pleistocene faunal remains in associated stratigraphy at the site dates no younger than ca. 13,500 years
20 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
ago, the Clovis-like blade level OSL B3 (Figures 4 and 15) has Springs Lodge, Page-Ladson, and Sloth Hole) sharing dated an OSL age (minimum age model) of 13,700 + 1100 years ago, artifact levels having a time continuum older than Clovis, our which incorporating the uncertainties might be as young as ca. conclusion here is that pre-Clovis in Florida is very nearly 12,600 years ago, a Clovis culture time, or as old as 14,800 confirmed. This is based, however, on the hypothesis that years ago. Therefore, the older levels at Sloth Hole (14,300 cal Clovis age is really no older than about 13,000 years ago B.P.) and the younger Clovis levels at Sloth Hole (12,900 cal (Waters and Stafford 2007). Additional sites or reinvestigation B.P.) agree with this age range, whereas the Page-Ladson age of the ones we already know about promise to make this of 14,400 cal B.P. agrees with the older end of the age range of determination. the Clovis-like blade level.
Three of the sites, Page-Ladson, Sloth Hole, and Wakulla Acknowledgements Springs Lodge, have yielded Page-Ladson points. The PageLadson point at the Wakulla Springs Lodge site came from We thank the National Geographic Society and the Natural
the 1994 Calvin Jones excavations at a level similar to that of Sciences and Engineering Council of Canada for financial the endscraper found in Unit C that was determined to have support. We also thank over thirty volunteers on the project a minimum age of 13,700 years ago. The OSL assay taken who graciously went through our screening and enrollment below the endscraper has a younger minimum age due to its process. Volunteers participated in various activities of the field large 4500 years ago uncertainty. In addition, although not work or acted as docents informing the visiting public about radiometrically dated, the Half Mile Rise Sink site (8TA98) in our project so that the working crew remained uninterrupted. the Aucilla River also has Page-Ladson type points (recently Two groups supplied volunteers, the Panhandle Archaeological donated by the late Don Serbousek). Society at Tallahassee (PAST) and The Friends of Wakulla
Clovis points are primarily flaked by percussion flaking Springs State Park. We also thank more than 20 professionals that includes both fluting and overshot flaking. The lateral participating on the project, many of whom led the pit crews edges of Clovis points do have pressure flaking, but pressure and assisted in other ways. The Friends of Wakulla Springs flaking does not play a part as a primary means of preform State Park donated the money to rent an on-site construction reduction and thinning (Bradley et al. 2010). In fact, Ken trailer as a mobile artifact-processing lab. Project participants Tankersley (1994) has shown that Clovis technology was less also donated funds to host a project cookout for the crew and effective in its ability to thin certain types of tough, percussion- guests (see Lawson, 2012, this volume). Long-time river resistant rock compared to the technological achievements diver Don Serbousek donated a major collection to BAR that made by subsequent Folsom and Goshen knappers in the includes a handful of Page-Ladson projectile points, from an
western United States who employed pressure flaking as a underwater site (8TA98, see Figure 1) (Serbousek 1983:88more effective means of thinning biface hunting weapons. 97) during the cookout. With this in mind, Simpson points, although not overshot
flaked or fluted, employ three-fourth shot percussion flaking References
as a primary means of reduction and thinning to achieve Antevs, Ernst Valdemar remarkable width to thickness ratios of up to 22:1 (Dunbar 1936 Dating records of early man in the Southwest. The and Hemmings 2000). Page-Ladson points, the specimens American Naturalist 70:331-336.
made on thin flakes, often have a flute-like scar on one side of
their base because the flake blank was left unknapped. In other Bradley, Bruce, Michael B. Collins, and C. Andrew Hemmings words, the flute-like feature represents an unflaked segment of 2010 Clovis Technology. International Monographs in the original flake preform's surface. Other Page-Ladson points Prehistory, Ann Arbor, Mi. are reduced from thicker biface blanks and are not fluted but
often show overshot flakes as a means of thinning (Dunbar Bullen, Ripley P. 2006a). These traits suggest that the "types", Simpson and 1968 A Guide to the Identification of Florida Projectile Page-Ladson, have a likelihood of not only being pre-Clovis, Points. First edition. Florida State zA Museum,
but also Clovis'ancestor. Gainesville.
From these cumulative findings, although not absolute, 1975 A Guide to the Jdentification of Florida Projectile it appears the pre-Clovis toolkit differed from Clovis in its Points. Revised edition. Kendall Books, Gainesville. biface knife (Simpson) and projectile point (Page-Ladson)
assemblage. However, it may show some direct affinity to Daniel, I. Randolph, and Michael Wisenbaker Clovis if the large, percussion-struck uniface blade tool found in 1987 Harney Flats." A Florida Paleo-Indian Site. Baywood Unit B at Wakulla is related to the Simpson toolkit. Publishing Company, New York.
The predominance of percussion flaking as a means of
biface reduction is also Clovis-like, yet enough traits differ Dunbar, James S. so that whatever the Simpson/Page-Ladson technology is, 2006a Paleoindian archaeology. In First Floridians and it is not Clovis. With three sites in North Florida (Wakulla Last Mastodons." the Page-Ladson Site on the Aucilla
Springs Lodge, Page-Ladson, and Half Mile Rise Sink) River, pp. 403-438. Springer Press, Dordrecht, The
sharing this artifact assemblage as well as three sites (Wakulla Netherlands.
RINK, DUNBAR AND BURDETTE OSL DATING AT WAKULLA SPRINGS LODGE SITE 21
2006b Pleistocene-Holocene Climate Change: University of Florida, Gainesville.
Chronostratigraphy and Geoclimate of the Southeast 2004 The Organic Clovis: A Single Continent-wide United States. InFirst Floridians andLast Mastodons.: CulturalAdaptation. Unpublished Ph.D. dissertation, the Page-Ladson Site on the Aucilla River, edited Department of Anthropology, University of Florida,
by S. David Webb, pp. 103-158. Springer Press, Gainesville.
Dordrecht, The Netherlands.
2008 Is The Wakulla Lodge Site A Pre-Clovis Paleoindian Ivester, Andrew H. and David S. Leigh
Habitation Or Temporal Tempest? How About 2003 Riverine dunes on the coastal plain of Georgia, USA.
Both! Unpublished report sent to the National Geomorphology 51:289-311.
Geographic Society in partial fulfillment of Grant
8404-08. Ivester, Andrew H., David S. Leigh, and D. I. Godfrey-Smith
2001 Chronology of Inland Eolian Dunes on the Coastal Dunbar, James S., and C. Andrew Hemmings Plain of Georgia, USA. Quaternary Research
2004 Florida Paleoindian Points and Knives. In New 55:293- 302.
Perspectives on the First Americans, edited by
Bradley Lepper, pp. 65-72. Center for the Study of Jones, B. Calvin, and Louis D. Tesar
the First Americans, College Station, Tx. 2000 The Wakulla Springs Lodge site (8WA329): a
preliminary report on a stratified Paleoindian
Dunbar, James S., Joseph Latvis, and Kevin Porter through Archaic site, Wakulla County, Florida. The
2008 Wakulla Springs Lodge Site (WA329) Grid Coordinate Florida Anthropologist 53:98-115.
Information. Copies available from Florida Bureau
of Archaeological Research, Tallahassee. Leigh, David S.
2008 Late Quaternary climates and river channels of Dunbar, James S., Kevin Porter, and Debra G. Shefi the Atlantic Coastal Plain, Southeastern USA.
2007 Inspection of the Vickery Mastodon Site in Wakulla Geomorphology 101:90-108.
Springs State Park. Copies available from Florida
Bureau of Archaeological Research, Tallahassee. Lowe, John J., Wim Z. Hoek, and the INTIMATE Group
2001 Inter-regional correlation of plaeoclimate records for Galbraith R. F., R. G. Roberts, G. M. Laslett, H. Yoshida, and the last Glacial-Interglacial transition: a protocol J. M. Olley for improved precision recommended by the
1999 Optical dating of single and multiple grains of quartz INTIMATE project group. Quaternary Science
from Jinmium rock northernAustralia: part I Reviews 20:1175-1187.
IExperimental design and statistical models.
Archaeometry 41:339-364. Madsen, A.T., A.S. Murray., T.J. Anderson; M. Pejrup, and H.
Grimm, Eric C., George L. Jacobson, William A. Watts, 2005 Optically stimulated luminescence dating of young Barbara C. S. Hansen, and Kirk A. Maasch estuarine sediments: a comparison with 210Pb
1993 A 50,000-year record of climate oscillations from and 137Cs dating. Marine Geology 214:251-268.
Florida and its temporal correlation with the Heinrich
events. Science 261:198-200. Murray, A.S., and A.G. Wintle
2000 Luminescence dating of quartz using an improved Grimm, Eric C., William A. Watts, George L. Jacobson, Jr., single-aliquot regenerative-dose procedure.
Barbara C. S. Hansen, Heather R. Almquist, and Ann C. Radiation Measurements 32:57-73.
2006 Evidence for warm wet Heinrich events in Florida. Otvos, Ervin G., and David M. Price
Quaternary Science Reviews 25:2197-2211. 2001 Late Quaternary Inland Dunes of Southern Louisiana
and Arid Climate Phases in the Gulf Coast Region.
Haynes, C. Vance Quaternary Research 55: 150-158.
1971 Time, Environment, and Early Man. Arctic
Anthropology 8:3-14. Purdy, Barbara A.
1974 Investigations concerning the thermal alteration of Hemmings, C. Andrew silica minerals: an archaeological approach. Tebiwa
1999 The Paleoindian and Early Archaic tools of Sloth 17: 37-66.
Hole (8Je]21): an inundated site in the lower Aucilla
River Jefferson County, Florida. Unpublished Rink, W.J., Kevin Burdette, and James Dunbar
Master's Thesis, Department of Anthropology, 2009 Optical Luminescence Dating of Quartz at Wakulla
22 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
Springs: A newly dated Early North American
Site. Paper presented at the 74th annual meeting of Wilder, M., Frederick, C.D., Bateman, M.D. and D.E. Peter
the Society for American Archaeology, Atlanta, 2007 Geoarchaeological investigations in the flats of the
Georgia. Osceola Plain, Highlands and Polk Counties,
Florida. The Florida Anthropologist 60:97-116. Serbousek, Don
1983 Explorations ofaPaleo-Indian site ontheAucillaRiver. Wright, James D.
The Florida Anthropologist 36:88-97. 1999 2.8.1 Global climate change in marine stable isotope
records. NUREG/CR 5562 (Global Climate Change Stafford, Thomas W., P. E. Hare, Lloyd Currie, A. J. T. Jull, Marine Stable Isotope Records): 2-671-2-682. and Douglas J. Donahue
1991 Accelerator radiocarbon dating at the molecular
level. Journal ofArchaeological Science 18: 35-72.
Stuiver, Minze, and Paula J. Reimer 1993 Extended 14C data base and revised CALIB 3.0 14C
age calibration program. Radiocarbon 35:215-230.
Tankersley, Kenneth B.
1994 The Effects of Stone and Technology on Fluted-Point
Morphometry. American Antiquity 59:498-510.
Tesar, Louis D., and Calvin B. Jones 2004 Wakulla Springs Lodge site (8Wa329) in Edward Ball
Wakulla Springs State Park Wakulla County,
Florida. Florida Bureau of Archaeological Research,
Waters, Michael R., and Thomas W. Stafford 2007 Redefining the Age of Clovis: Implications for the
Peopling of the Americas. Science 315:1122-1126.
Watts, W. A., B. C. S. Hansen, and E. C. Grimm 1992 Camel Lake: A 40,000-Year Record of Vegetational
and Forest History From Northwest Florida.
Ecology Society ofAmerica 73:1056-1066.
Webb, S. David, and James S. Dunbar 2006 Carbon Dates. In First Floridians and Last
Mastodons. the Page-Ladson Site on the Aucilla River, edited by S. David Webb, pp. 83-102. Springer
Press, Dordrecht, The Netherlands.
Willard, Debra A., Christopher E. Bernhardt, Gregg R. Brooks, Thomas M. Cronin, Terence Edgar, and Rebekka Larson 2007 Deglacial climate variability in central Florida,
USA. Palaeogeography, Palaeoclimatology,
BIOTURBATION AND THE WAKULLA SPRINGS LODGE SITE ARTIFACE DISTRIBUTIONS
DAVID K. THULMAN
Department ofAnthropology, George Washington University, Washington D.C 20052 Email: firstname.lastname@example.org
Paleoindian sites in Florida, such as Paradise Park region (Thulman 2009). The configuration of the large spring (8MR92) at Silver Spring (Neill 1958; Hemmings 1975) and vent would have provided easy access to groundwater for both Harney Flats (8HI507) (Daniel and Wisenbaker 1987) are animals and people, and divers in the spring have recovered frequently found in sandy deposits deeper than one meter, remains of extinct fauna and Paleoindian artifacts from some which present challenges for site interpretation, because a of its deeper levels (Porter 2012, this volume). sandy living surface is easily disturbed by people, animals, In 1994, excavations adjacent to the lodge produced an and plants. The Wakulla Springs Lodge site (8WA329) is no Early Archaic Bolen-age occupation and deeper Paleoindian exception (Figure 1); the putative Pleistocene-age occupation artifacts (Tesar and Jones 2004). The Paleoindian artifacts were is about 1 m below the present land surface, and the vertical found at depths ranging from 105 to 165 cm below datum, but distribution of the artifacts clearly indicates that some post- precisely correlating those levels with the 2008 excavations depositional disturbance occurred. The spatial analysis of the was impossible, because Jones's datum was removed (Tesar artifacts (chipped stone tools, debitage, and ceramics) at the and Jones 2004:134; Rink et al. 2012, this volume). One of site takes on added significance because of the association of those Paleoindian artifacts was similar to several found at early optically stimulated luminescence (OSL) ages (Rink and the Page-Ladson site, about 16 km to the east in the Aucilla Burdette 2009) for two apparently early, but non-diagnostic, River, which Rink et al. (2012, this volume) argue are preunifacial chipped stone tools (referred to here as the paleo- Clovis in age. James Dunbar received funding from the blade and the paleo-scraper) and the unambiguous evidence National Geographic Society and tested several areas near of bioturbation (Rink et al. 2012, this volume). OSL works the 1994 excavations (Figures 1 and 2). The excavations took by testing the last time individual sand grains were exposed place during the month of April 2008 and led to the recovery to sunlight, and its use in archaeology requires that an artifact of the paleo-blade and paleo-scraper. No material suitable on the surface remains in proper association with the sand for radiocarbon dating was recovered, but several OSL dates grains as each are covered and become incorporated in the were obtained, and ages were ascribed to the two artifacts. The archaeological record. details of the OSL dating effort are reported by Rink et al.
This article examines what the statigraphic positions of (2012, this volume). artifacts at the Wakulla Springs Lodge site, especially stone
tools and debitage, reveal about site formation processes, the The 2008 Excavations
vertical integrity of the site, and the strength of the association
of the OSL ages with the two artifacts of apparent Paleoindian Unlike Jones' earlier excavation (Tesar and Jones age. The potential effects of bioturbation by the Florida 2004), which involved a single trench, the 2008 excavations harvester ant (Pogonomyrmex badius) and Southeastern consisted of seven units (Tests A-F and 1), six of which
pocket gopher (Geomys pinetis) on the site are examined in were horizontally separated by at least 2 m (Figure 2). It was some detail. The article concludes that the distribution of assumed in the project planning and during the early stages of artifacts is best explained by bioturbation-human, faunal, excavation that the levels could be chronologically correlated and floral-and one or more episodes of eolian deposition. The later among the units by identifying diagnostic artifacts, bioturbation does not automatically invalidate the association levels of artifact concentrations, clearly defined strata, or a of the OSL ages and tools, and several lines of evidence combination of the these criteria. However, as excavations indicate the association may be secure. More work needs to be proceeded, it became clear that correlating strata among the done to understand fully the site formation processes and their units would be problematic, as no clearly defined strata and effects on OSL dating of early sites. no obvious concentrations of artifacts in strata could be traced
Wakulla Springs and the surrounding area have long among units. The lack of defined strata probably resulted from
produced evidence of Paleoindian occupation (Rink et al. the homogenization of sediments through bioturbation. No 2012, this volume). Climatic reconstructions for the late unambiguous, diagnostic artifacts older than Middle Archaic Pleistocene Epoch indicate that the spring, which presently is Levy points were recovered in what appeared to be relatively one of the largest in Florida, was likely not flowing at that time undisturbed chronological position, making the correlation of and was one of several scattered sources of freshwater in the the deeper levels difficult. Complicating the correlation further
VOL. 65 (1-2) THE FLORIDA ANTHROPOLOGIST MARCH-JUNE 2012
24 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
igure. 1.... 1 in relation t theloaSprings .
was the unknown topography of the original land surface account for the vertical distribution of the artifacts and do not before the lodge was constructed, much less during the last necessarily invalidate the proper association of the OSL ages 25,000 years. The underlying, undulating limestone surface and the paleo-artifacts. may also be a complicating factor (Smith, this volume, 2012). Seven units were excavated (A-F and 1), and all but Thus, correlating strata among units by either meters above one (Unit C) encountered at least one subsurface drainage mean sea level (asl) or meters below land surface (bls) was or irrigation pipe, many of which had been abandoned and based on questionable assumptions about the topography of the forgotten. The smaller irrigation pipes were typically near the prehistoric surface. Finally, it was clear that some artifacts had surface, but some of the larger drainage pipes were deeply been displaced downward, but how many and how far they had buried. The outlines of the pipe trenches were relatively moved from their original position after discard was unknown. easy to discern in the units, and it appeared that the area of Some of the debitage was recovered far below the oldest dated disturbance due to the shallow and moderately buried pipes levels; so deep that the only reasonable explanation was that was limited. Units B, C, and F showed the least amount of they had been displaced from their original positions. Thus, the obvious disturbance. effort to correlate excavation levels among the units was set Since the focus of the project was on the deeper aside, and analysis focused on understanding how the artifacts Paleoindian component, the upper levels in each unit were dug became vertically distributed at the site, with an emphasis on in 10-20 cm levels, and the lower levels starting about 90 cm understanding whether the OSL ages could accurately date the bls were dug in 5-10 cm levels. After excavation of a level, paleo-tools under these conditions. the height of each corner of the unit was measured with a total
The positions of artifacts in all the units indicated that station and correlated to height in meters asl. Initially, 1/8 inch one or more post-depositional mechanisms had displaced hardware cloth was used for all units, but several days into the
some of the artifacts vertically. After eliminating several project we switched to window screen in an effort to capture alternative possibilities, bioturbation (human, faunal, and and evaluate the distribution of micro-debitage (less than floral) seemed the most likely cause, and analysis specifically .125 inches). The timing of this switch meant that only Unit focused on a combination of human trampling, scavenging, F used window screen through the entire column, allowing pit-digging, ants, pocket gophers, and plant roots as the most comparison of debitage concentrations through all levels only likely culprits. Along with possible eolian deposition of sand in that unit. at the site (Means, this volume, 2012), these mechanisms can
THULMAN BIOTURBATION AT THE WAKULLA SITE 25
Figure 2.The test units from the 2008 excavations at the Wakulla Springs Lodge site.
The Artifact Assemblage recovered at 97.3 cm bis in Unit B. It could be a unifacial
hafted knife or scraper. The edges of the possible haft end The excavation recovered 85 chipped stone tools and are crushed and duller than the working edge, the bulb of fragments, one possible shell tool, and 24 pottery sherds (Table percussion was apparently removed, and the proximal end was 1). Unambiguously diagnostic artifacts included pottery and squared off. The form has not been reported from Paleoindian stone points. Larger artifacts were piece-plotted, sites in Florida.
Eight points or point fragments were recovered, and all Ceramics but one (a Bolen point) date to the Middle or Late Archaic
periods. Two fragments of a Middle Archaic point recovered Pottery was recovered that dated to most time periods, in the 20-40 cm bls in Unit C and a Middle Archaic Levy including early Norwood fiber-tempered ceramics, which point recovered in the 20-40 cm bls in Unit B were the deepest have been dated to ca. 4000 B.P. (Milanich 1994). In the diagnostic Middle Archaic points found in the least disturbed undisturbed Unit C and minimally disturbed Units B and F, units. The beveled side-notched Bolen point was recovered on the deepest ceramics (Norwood) were found at approximately edge, below 140 cm bls in Unit B, more than 43 cm below 60 cm bls. the paleo-blade. Because Bolen points date securely in
Florida to the Early Archaic period (ca. 10,000-9300 B.P.; ca. Lithic Tools 111,600-10,400 cal. B.P., [CALIB 4.42]; Carter and Dunbar
2006; Webb and Dunbar 2006), it most likely was displaced The paleo-scraper (FS 275) (Figure 3) was recovered downward through a rodent burrow or root track. at 105.3 cm bls in Unit C. The distal end of the scraper was likely snapped during use. A large flake was removed from the Debitage center of the tool. Its dimensions (width 39 mam, maximum thickness 10 mam) fall near the mean width and thickness of Debitage (n-4754) was recovered from the surface to endscrapers found at Hamey Flats (8Hi507), a Suwannee! the deepest levels in all units. Debitage was sorted by size Bolen site near Tampa, Florida (Daniel and Wisenbaker (1 cm increments) and categorized as flake or blocky shatter
1987:65). The paleo-blade (FS 160) (Figure 4) was and as heated or unheated. "Heated" was determined by the
Table 1: Numbers of artifacts recovered in the 2008 excavations.
Diagnostic Biface Cores & Unifacial Expedient Abrader Paleo- Shell Ceramics Debitage
bifaces & fragments core scrapers! tools knife groover
fragments fragments fragments
8 10 6 9 50 1 1 124 4754
26 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
(larger than 1 cm) debitage.
In sum, debitage distributions show that the most
intensive occupation occurred within the upper strata, which
4 had higher percentages of heated lithics, a typical signature of
the Middle Archaic and more recent periods. The distributions
below approximately 90-100 cm show a dramatic drop in the
total number of artifacts, which it to be expected result
of presumed lower population levels in the Early Archaic and
The most likely explanation for the chipped stone artifact
(tools and debitage) distribution at the Wakulla site is a
combination of bioturbation and one or more episodes of eolian
deposition. Two of the most vigorous animal bioturbators-the
southeastern pocket gopher and several ant species-are found
in longleaf pine habitat, which likely was present at the site
2at different times in the past 25,000 years. Other bioturbators common in Florida, such as gopher tortoises, moles, mice,
burrowing owls, scarab beetles, termites, and trees surely
Figure 3.The paleo-blade, played parts in disturbing the site, but this article will focus
on humans, ants, and pocket gophers to investigate the effects
of biological disturbance on the relationship between the OSL
presence of red discoloration or sheen, but as Rink et al. ages and artifacts. (2012, this volume) point out, the analysis was cursory and Faunal bioturbation can have several effects on an determination of heat alteration needs corroboration. No archaeological site (Johnson 1990; Balek 2002). Larger determination was made as to whether any of the debitage was artifacts tend to become covered when animals move soil deliberately heat-treated. to the surface, which both covers an artifact and displaces
Debitage was grouped in 5, 10, and 20 cm levels in each it downward through gravity as the subsurface tunnels and unit in an attempt to discern concentrations that would indicate burrows collapse. Eventually, subsurface burrowing and occupation levels across units. This effort was unsuccessful tunneling combine to completely cover artifacts (Balek 2002). as no apparent vertical patterns were seen, other than a clear While large artifacts sink, artifacts smaller than the diameter decrease at approximately 90 cm bls in total number of of a tunnel or burrow can be translocated to the surface during
debitage and the percentage of heated debitage. The debitage excavation or to deeper levels through nest or burrow collapse distributions indicate that the most intensive occupation or by falling down a tunnel or burrow. Bioturbation tends to occurred within the upper approximately 100 cm of the site homogenize the soil profile, eradicating occupation levels and (referred to here as the upper strata). The debitage distributions soil horizons (Leigh 1998, 2001). Over enough time, artifacts in Units B and C are plotted in Figures 7 and 8 of Rink et al. can collect in "stone zones" at the deepest level of biological (2012, this volume). It seems reasonable to place the Early activity through repeated excavation and collapse (Johnson Archaic and earlier levels about 90 cm bls, based on the clear 1990). Floral bioturbation can affect a site through two basic decrease in debitage concentrations below 90 cm bls in Unit mechanisms: the creation of cavities through which artifacts F and concurrent decreases in Units B and C at approximately can fall once roots rot and tree uprooting, which displaces 80 cm bls and approximately 90 cm bls, respectively (but see buried artifacts entangled in the tree roots to the surface Rink et al. 2012, this volume, for a different interpretation). (Johnson 1990). This inference is supported by the distribution of debitage
showing evidence of heating, which is more likely to be found Human Bioturbation after the Early Archaic period. Between 40 and 90 cm bls,
the average number of debitage showing evidence of heating Human trampling is a form of bioturbation, which occurs
ranges from 23-41 percent heated, but below approximately when people walk across a site, step on artifacts and push 90 cm, the average is about 10 percent heated. them below the surface, or kick up soil or sand and bury
It is clear that some downward displacement of debitage the artifacts.' Experiments in sandy sites demonstrate that occurred, which is reflected in the generally even distribution artifacts can be displaced at least 10 cm below the surface of artifacts throughout the lower strata (below approximately after two hours of constant human traffic (Gifford-Gonzalez 90 cm). For example, total debitage counts in Unit C tapered et al. 1985), although the maximum depth for displacement by off from a high of 147 at 110-115 cm bls to 15 at 145-150 human trampling alone is unclear. In the same experiments, cm bls. The same general trend can also be seen in the larger walkers saw artifacts disappear only to reappear later in
THULMAN BIOTURBATION AT THE WAKULLA SITE 27
different areas of the test site, indicating the artifacts moved sites in Florida, such as Wakulla Springs, ants were likely to up, down, and horizontally in unpredictable ways in the top have caused the most profound effects, and they are discussed 10 cm. It is unclear whether trampling creates a size-sorting here. of artifacts (Stockton 1973; Gifford-Gonzalez et al. 1985), but Ants as Bioturbators at Wakulla. it appears that larger artifacts are less likely to move (Brooks
and Sassaman 1990), although that is not a certainty (Gifford- Like much of the Big Bend region of Florida, the Gonzalez et al 1985). The depth of displacement is directly area around the Wakulla Lodge was likely a longleaf related to the intensity and length of trampling and soil pine community, which occupies undulating sandy plains characteristics, including density. In sandy soils, a saturated with a high water table (Abrahamson and Hartnett 1990). zone or hardpan can impede further downward displacement. Prehistorically, a longleaf pine community likely had widely spaced trees and grass-dominated ground cover, but changing
Faunal Bioturbation climate conditions in northern Florida over the millennia
(Dunbar 2006) and the proximity of the site to the river, makes
Every animal that burrows or tunnels in sandy soils in habitat reconstruction difficult, and the type of understory is Florida could have had some effect on the Wakulla Springs a critical determining factor for some species that inhabit the site, and the mechanisms for disturbance would be similar, subsurface. In a survey of secondary-growth longleaf pine although the specifics would differ depending on the size of habitat with varying understory in the Apalachicola National the animal and its burrowing behavior. Larger animals, such as Forest, Lubertazzi and Tschinkel (2003) found a remarkable gopher tortoises, would tend to bring buried older artifacts to variety of ants: 64 arboreal, subterranean, and ground foraging the surface and coincidentally cover younger artifacts laying species. on the surface. Artifacts too large to bring to the surface would It is impossible to know which ant species inhabited tend to be displaced downward when burrows collapsed. Wakulla, so this analysis models the potential effects of the
Over time, continual burrowing by larger animals could bury Florida harvester ant (Pogonomyrmex badius), a common artifacts originally on the surface, displace larger artifacts in a ground foraging seed-harvesting ant that prefers longleaf pine single deeper stone zone, and mix smaller artifacts (regardless habitat (Harrison and Gentry 1981; Tschinkel 2004). This of age) in a layer overlying the stone zone, creating what species was selected because it is native to northern Florida, Johnson (1990) calls a two-layeredfaunalmantle. The artifact creates deep nests, moves frequently, and has been welldistributions do not indicate that larger animals that dig deep studied. Thus, it represents a probable worst-case scenario for burrows, like the gopher tortoise, had a significant impact at site disturbance by ant bioturbation. Wakulla. Ants that excavate nests in sandy soils will remove
Smaller animals, such as ants, termites, locusts, scarab one or more sand grains in their mandibles from the tunnel beetles, worms, or wasps, work the same way as larger animals, or chamber wall and transport the grain(s) to the surface for although the size of the artifacts displaced up or down via their disposal, creating an apron around the nest entrance. All ants tunnels would be much smaller. Unlike larger animals, smaller form repeating structures of shafts and chambers, but the animals rarely leave traces of their burrows at a site. In sandy arrangement of these two elements is species-specific. Some
*.. i~ii flAE.....
28 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
species excavate simple nests descending less than a meter Surface
with an upper chamber and single shaft with a few small chambers, while others, like P badius create nests that may contain 150 chambers, 10 m of tunnels, and extend 2-3.5 m deep (Tschinkel 2003). P badius colonies can construct a new nest in 4-5 days, and a mature colony can excavate 20 kg of sand in that time (Tschinkel 2004). The mounds tend to be raised 2-3 cm above the surface and form an apron up to 80 cm in diameter (Harrison and Gentry 1981). Colonies in a sandy field in South Carolina had an average foraging territory of approximately 136 m2 (a circle of approximately 6.6 m radius). Their nearest neighbor was approximately 12.5 m away, and _colonies tended to abandon their nest once or twice a year and move an average of 2.2 m to establish a new nest. New nests are typically not established at old nest locations (Harrison and Gentry 1981).
The nest of a mature colony of P badius usually has a single entrance and a "top-heavy" structure, with the majority of chambers located in the top 25-35 cm of the nest; some of the upper chambers may be constructed in the nest apron (Tschnikel 2004: Figure 6c). The diameter of the nest is about 50 cm, and there are from 6-10 levels of chambers. Chambers, which can appear to wrap around a shaft, vary in area but are about 1 cm high and are larger and more densely constructed Figure 5. The model of bioturbation by the Florida (both vertically and horizontally) in the upper part of the nest. harvester ant (Pogonomyrmex badius) at Ti A large nest will have four or five descending shafts with years.
widely spaced attached chambers. Shafts in the upper nest can be 2 cm in diameter, decreasing to 1 cm or less in the deeper cm (assuming the artifact is displaced half the height of a parts. The shafts, which tend to spiral, are inclined from 200 to chamber) and the maximum amount will be approximately 700, but they are steeper in deeper parts of the nest (Tschinkel 10 cm (assuming the artifact moves the entire height of 10 2004). underlying, collapsing chambers). Third, smaller artifacts (less
than 1 cm in diameter) can be displaced downward through Models of the Effects of Ant Bioturbation at Wakulla Springs both chamber collapse and by falling down shafts. Larger artifacts, however, will only be displaced downward through Several inferences can be drawn about the effect of ants on chamber collapse. Fourth, once the ground surface is 25-35 the Wakulla site from nest architecture and P badius behavior. cm above a larger artifact, ant bioturbation will cease to be an First, a colony will cover 136 m2 with 2-3 cm of soil every important mechanism for downward displacement. However, 272-680 years, assuming colonies do not reuse old nest sites smaller artifacts will continue to be displaced by falling down (Figure 5). Thus, relative to the original surface, an artifact tunnels, although less often, until the ground surface is over 2 on the surface will "sink" 2-3 cm every 272-680 years simply m above the artifact. from the excavated soil piled on the surface. This process, referred to as biomantling, can be a primary agent of artifact Southeastern Pocket Gophers (Geomys pinetis) burial. The variability in time comes from assumptions about the apron size: an apron of 50 cm diameter covers 100 m2 Among the larger bioturbators, the southeastern pocket
more slowly and requires 680 years, while an apron of 80 cm gopher (Geomys pinetis) has the potential for widespread diameter covers 100 m2 in 200 years. (For these calculations, disturbance (Simkin and Michener 2005). G. pinetis is I used 500 years, which assumes an apron of approximately a prodigious digger in a sandy, well-drained, frequently 59 cm in diameter.) Second, artifacts located in the upper buried, longleaf pine ecosystem (Kialisz and Stone 1984). It 25-35 cm will be most affected by the nest excavation and most actively burrows in the top 20 cm, but up to 60 cm in collapse. Depending on where the artifact is located in the depth, and creates mounds that average 47 cm long by 32 cm 25-35 cm layer, it will be displaced downward from 3 to 10 wide by 12 cm high in dimension. Mounds can cover up to cm in that layer), but more likely something less than 10 cm. approximately 2 percent of a suitable habitat per year (Simkin For the analysis below, 5 cm in downward displacement was and Michener 2005), but tunnels may cover over 7 percent of used as the average. The amount of displacement is figured by the area (Anderson 1987). Unlike the harvester ant, G. pinetis assuming a single chamber collapse will lower an overlying may disturb the same area in subsequent years, creating a artifact at most 1 cm but more likely something less than 1 patchwork of disturbed and undisturbed areas. Tunnels are cm. Thus, the minimum amount will be approximately 3 the width of a gopher, about 7.5 cm, and once excavated, the
THULMAN BIOTURBATION AT THE WAKULLA SITE 29
- --- 5 Original Surfac -- OdtginalSurface
Figure 6. The model of bioturbation by the Florida bar- Figure 7. The model of bioturbation by the Florida harvester ant (Pogonomyrmex badius) at Time 500 years. vester ant (Pogonomyrmex badius) at Time 7000 years.
the larger artifacts are significantly affected by nest collapse.
gopher may plug or fill a tunnel and later re-excavate it. Smaller artifacts that intersect shafts continue to be displaced
The southeastern pocket gopher is less active and less downward. It is possible that a few small artifacts could be 2 destructive to sites than western species of pocket gopher. m below the majority of the artifacts from the site. Several studies in California have shown pocket gophers Thus, for a single component site at least 7000 years
creating bimodal distributions of artifacts, smaller artifacts old and only subject to R badius turbation, we could see an near the surface and larger artifacts concentrated in zone 30 artifact distribution in which large and small artifacts are to 60 cm below the surface (Bocek 1986; Erlandson 1984; distributed throughout a 35 cm layer and smaller artifacts Johnson 1989). are distributed up to 2+ m below that layer, with decreasing
concentrations at depth. There may be size sorting, with larger
2.0 mn 2.0 mn
2.5Discussion artifacts concentrated nearer the top of the 35 cm layer. When
occasional C. pinetis occupations are considered, the artifacts
Figure 6. These model of human and faunal bioturbation predict will be moved up and down inby the Fltop 60 cm, and over about vsthat the surface of a site in longleaf pinat Time habitat in Florida 100 yearster ant (Pogonomyrmex badius) at Time 7000of original surface
thwill not likely remain undisturbed through time. In a simple could be buried byfacts are significantly affected by nest ccm of soil.
gopher may plug or fill a tunnel and later re-excavate it. Smaller artifacts that intersect shafts continue to be dispae
Thexample of a single component site less active and less downward. It is possible that a few del of vertical artifact distribution is dstructivable, with a sitable climax wte arn species of pocket supports complicated with multiple human occupations, especially Several studies and pinetis, human trampling will distribute tgophes those occurring at less than 7000-year intervals, additional aretingfacts throughout the upper 10 cm of thes, surface (Figure bioturbators, and periodic olian deposi turbation. Larger animals
5). Ant nests are constructed about once a year, and a uniform could have dug burrows at the site, moving larger artifacts succession of nests will have completely covered the site in down or up. Tree fall, root rot, and burned out stump holes 500 year thes. After 500 years, theand larger artifacts concentrated in zone 30 artifact distributioned culd have had the same effect. large addition, people could through a layer from 7 to 17 cm (2 cm of surface disposal + 5 have excavated hearths or f35 cm lattened living surfaces. For Johnson 1989). are distributed up to 2+ m below that layer, with decrasn
cm of displacement from nest collapse), and a small number non-diagnostic artifacts, such asr most debitagei parsing the
of smaller artifacts have displaced downward by falling down distribution to specific occupations may be impossible. nest shafts (Figure 6). After 1000 years, the artifacts are no
longer disturbed through human trampling (since they are Association of the OSL dates and palco-art facts more than 10 cm below the surface), and after 7000 years
(the amount of time it takes for artifacts on the surface of Artifacts are not the only things translocated through the original site to be displaced 35 cm; Figure 7), the last of bioturbation; sand grains will also move down through shafts
30 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
and as chambers collapse or be deposited on the surface through independent of the surrounding sand. Through human excavation. There are four scenarios that could explain how the trampling, the artifacts could have been pushed down about 10 paleo-artifacts could have ended up adjacent to "older" sand. cm and come into contact with "older" sand grains. However, ("Younger" sand grains are more recently exposed to sunlight trampling tends to move artifacts-and sand-up and down, than "older" sand grains.) First, the strata were thoroughly but it is presumed that all of the sand would have been mixed mixed, and by chance the artifacts and "older" sands came to within that 10 cm zone, and some of the sand would be "zeroed be adjacent. Second, "older" sand was displaced upward and out" at the time the artifact was initially buried. Thus, the age came to be adjacent to the artifacts without being re-exposed of the sands presumably would be the same as or close to the to sunlight. Third, the artifacts were displaced downward, but date of initial burial for the artifact. However, here the OSL the surrounding sand was not, so the artifacts were recovered samples were taken some distance from the artifacts, which in older sediments. Fourth, some combination of artifacts increase the potential for dating sands of different depositional moving down and older sand moving up in the sediments. ages than the artifacts. Tunnel and chamber collapse would
Several lines of evidence indicate that these alternatives are tend to move both the artifacts and the sand grains surrounding unlikely, but at this time we cannot conclude the dates are them in concert. It is possible that a pocket gopher or some unequivocally associated with the artifacts. other larger bioturbator could have pushed the paleo-artifacts
Two facts indicate the paleo-artifacts were minimally to lower levels, but their horizontal orientations found during disturbed. First, both were found lying flat, unlike the Bolen excavation tend against this interpretation, i.e., they would point, which was both on edge and pointed downward. Second, not be expected to consistently be lying flat if they had been the tools were at the expected depth for Pleistocene-age strata, disturbed by large bioturbators. based on Jones' earlier excavations. The evidence as to the In sum, the most likely scenario to explain a poor mixing of sand grains at the site is more equivocal. While the association of the artifacts and OSL central age model ages OSL ages are in proper chronological order with depth, it does would include a combination of tunnel collapse that dropped appear that some vertical disturbance of the sand occurred the artifacts to a new deeper level in a way that maintained (Rink et al. 2012, this volume). the orientation of the artifacts and preserved the OSL integrity
Considering these conditions of the artifacts and sand at of the deeper surrounding sands. However, a combination of the site, the first scenario seems so highly unlikely that it can 1) upward movement of older grains of sand within the site be discarded without much consideration. Sites thoroughly but without transport to the ground surface (where the grains mixed by bioturbation tend to produce OSL ages that are not in would be "re-set"), and 2) downward movement of artifacts proper chronological order (e.g., Bateman et al. 2006), skewed can also explain the distribution of the minimum OSL ages DE distributions (Bateman et al. 2003), and artifacts that are and the artifacts (Rink et al. 2012, this volume). not lying flat. (Skewed DE distributions indicate mixing of the
sediments, but see Rink et al. [2012, this volume] for a more Conclusion
thorough explanation of the significance of the DE distributions
at the site.) Based on the current state of understanding about The Wakulla Springs Lodge site has presented an ant behavior, the second scenario also seems unlikely. The interpretative challenge related to the relationship between older sand would have to have been brought up by ants, bioturbation and OSL dating. The integrity of all sandy
gophers, or other bioturbators, and deposited in a subsurface sites in Florida should be regarded critically because of the chamber alongside the artifacts, in a sequence that preserved potential for significant subsurface disturbance by a variety of the chronological sequence in each unit, or deposited on the biological actors, some of which can thoroughly alter a site surface without being "reset" by being exposed to sunlight. It (e.g., Bateman et al. 2007). Further, studies of the relationship is presumed among ant behavioralists that all ants that excavate between bioturbation and datable samples might be able to nests appear to deposit all excavated sand grains on the surface resolve the relationship between artifacts and age, but the (Sudd 1969), but this has yet to be tested. In contrast, pocket problems are undoubtedly site-dependent. It is noted here gophers tend to deposit most of their material on the surface, that it is not only OSL ages that may suffer from the effects but they will plug burrows with subsurface soils (Hickman of bioturbation; radiocarbon-datable materials are equally and Brown 1973), which will not "reset" the OSL clock. Even susceptible to displacement, albeit by slightly different soil deposited on the surface may not be fully reset, but in mechanisms. that event the DE would be skewed (Bateman et al. 2003). It is
more likely that younger sand grains would fall down shafts Notes
and chambers, which would have made the artifacts appear 1. At Wakulla, Pleistocene megafauna such as mastodons or younger than the OSL ages indicated. mammoths could have had a significant trampling impact, far
The third hypothesis would require downward exceeding the impacts of relatively small humans
displacement of the large paleo-blade and paleo-scraper
THULMAN BIOTURBATION AT THE WAKULLA SITE 31
Acknowledgements Brooks, Mark J., and Kenneth E. Sassaman
1990 Point bar geoarchaeology in the Upper Coastal Plain I thank Jim Dunbar and Jack Rink for a number of the Savannah River Valley, South Carolina; a case
of stimulating discussions that ultimately led to a more study. In Archaeological Geology of North America,
sophisticated model of the site formation processes at the edited by N. P. Lasca and J. Donahue, pp. 183-197.
Wakulla Springs Lodge site and their possible effect on the Geological Society of America, Boulder CO.
OSL dates. It has given me a greater appreciation for the nuances and potential pitfalls in accepting such data at face Carter, Brinnen, and James Dunbar value. I also want to thank Michael Faught for commenting on 2006 Early Archaic Archaeology. In First Floridians and an early version of the paper, and all the authors of this volume Last Mastodons: the Page-Ladson Site in the Aucilla and an anonymous reviewer who provided comments. River, edited by S.D. Webb, pp. 493-516. Springer,
Dordrecht, the Netherlands.
.References Daniel, R. I., & Wisenbaker, M.
1987 Harney Flats: A Florida Paleo-Indian Site. Baywood Abrahamson, Warren G., and David C. Hartnett Publishing, Farmingdale, NY.
1990 Pine Flatwoods and Dry Prairies. In Ecosystems of
Florida, edited by Ronald L. Meyers and John J. Dunbar, James S.
Ewel, pp. 130-149. University of Central Florida 2006 Paleoindian Archaeology. In First Floridians and Press, Orlando. Last Mastodons: the Page-Ladson Site in the Aucilla
River, edited by S. David Webb, pp. 403-435.
Anderson, D. C. Springer, Dordrecht, the Netherlands.
1987 Geomys bursarius burrowing patterns: Influence of
season and food patch structure. Ecology 68:1306- Erlandson, J.M.
1318. 1984 A Case Study in Faunalturbation: Delineating the
Effects of the Burrowing Pocket Gopher on the
Balek, Cynthia Distribution of Archaeological Materials. American
2002 Buried Artifacts in Stable Upland Sites and the Role Antiquity 49:785-790.
of Bioturbation: A Review. Geoarchaeology 17: 4151. Faught, Michael K., Michael B. Hornum, R. Christopher
Goodwin, Brinnen Carter, and S. David Webb
Bateman, Mark, Claire Boulter, Andrew Carr, Charles 2003 Earliest-Holocene Tool Assemblages from
Fredrick, Duane Peter, and Michael Wilder Northern Florida with Stratigraphically Controlled
2006 Preserving the palaeoenvironmental record in Radiocarbon Estimates (Sites 8LE2105 and 8JE591).
drylands: Bioturbation and its significance f o r Current Research in the Pleistocene 20:16-18.
luminescence-derived chronologies. Sedimentary
Geology 195:2-19. Gifford-Gonzalez, Diane, David B. Damrosch, Debra R.
2007 Detecting Post-depositional sediment disturbance Damrosch, John Pryor, and Robert L. Thunen
in sandy deposits using optical luminescence. 1985 The Third Dimension in Site Structure: an Experiment
Quaternary Geochronology 2: 57-4. in Trampling and Vertical Dispersal American
Bateman, Mark, Charles Fredrick, Manoj Jaiswal, and Ashok Singhvi Harrison, Janet, and John Gentry
2003 Investigations into the potential effects of 1981 Foraging Pattern, Colony Distribution, and Foraging
pedoturbation on luminescence dating. Quaternary Range of the Florida Harvester Ant, Pogonomyrmex
Science Reviews 22: 1169-1176. Badius. Ecology 62: 1467-1473.
Bocek, B. Johnson, Donald L.
1986 Rodent Ecology and Burrowing Behavior: Predicted 1989 Subsurface Stone Lines, Stone Zones, ArtifactEffects on Archaeological Site Formation. American Manuport Layers, and Biomantles Produced by
Antiquity 51:589-603. Bioturbation via Pocket Gophers (Thomomys Bottae).
American Antiquity 54:370-389.
32 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
1990 Biomantle Evolution and the Redistribution of Earth Materials and Artifacts. Soil Science 149:84-102. Tschinkel, Walter R.
2002 Subterranean ant nests: trace fossils past and Kalisz, P.J., and E. L. Stone future? Palaeogeography, Palaeoclimatology
1984 Soil Mixing by Scarab Beetles and Pocket Gophers Palaeoecology 192:321-333.
in North-Central Florida. Soil Science Society of 2004 The nest architecture of the Florida harvester ant, America Journal 48:169-172. Pogonomyrmex badius. Journal of Insect Science
Leigh, David S.
1998 Evaluating Artifact Burial by Eolian verses Webb, S. David, and James Dunbar
Bioturbation Processes, South Carolina Sandhills, 2006 Carbon Dates. In First Floridians and Last USA. Geoarchaeology 3:309-330 Mastodons. the Page-Ladson Site in the Aucilla
2001 Buried Artifacts in Sandy Soils: Techniques for River, edited by S.D. Webb, pp. 83-102. Springer,
Evaluating Pedoturbation versus Sedimentation. Dordrecht, the Netherlands.
In Earth Sciences and Archaeology, edited by Paul Goldberg, Vance T. Holliday, and C. Reid Ferring,
pp. 269-293. Kluwer Academic, NewYork.
Lubertazzi, David, and Walter Tschinkel 2003 Ant community change across a ground vegetation
gradient in north Florida's longleaf pine flatwoods.
Journal of Insect Science 3:21. http:// insectscience.
org/3.21, accessed September 30, 2010.
Rink, W. Jack, and Kevin Burdette 2009 Optical Luminescence Dating at Wakulla Springs
Lodge. Paper presented at the 74th Annual Meeting of the Society for American Archaeology,
Simkin, Samuel, and William Michener 2005 Faunal Soil Disturbance Regime of a Longleaf Pine
Ecosystem. Southeastern Naturalist 4:133-152.
Stockton, E. D.
1973 Shaw's Creek Shelter; Human displacement of
artefacts and its significance. Mankind 9:112-117.
Sudd, J. H.
1969 The Excavation of Soil by Ants. Zeitschrift fur
Tesar and Jones 2004
2004 Wakulla Springs Lodge Site (8WA329) in EdwardBall
Wakulla Springs State Park, Wakulla County, Florida.
a Summary of Eleven Projects and Management Recommendations. Bureau of Archaeological
Research, Florida Department of State. Tallahassee.
Thulman, David K.
2009 Freshwater Availability as the Constraining Factor in
the Middle Paleoindian Occupation of North-Central
Florida. Geoarchaeology 24:243-276.
GROUND PENETRATING RADAR (GPR) SURVEY, WAKULLA SPRINGS STATE PARK, WAKULLA COUNTY, FLORIDA: USE OF GPR IN SUPPORT OF ARCHAEOLOGY
DAVID ANDY SMITH
Forida Geological Survey, Department of Environmental Protection, 903 West Tennessee Street, Tallahassee, Florida 32304 email: email@example.com fl us
Introduction part of the pulse is reflected back to the surface. The GPR
receiving antenna detects the reflected waves and the GPR
Figure 1 shows the location of Edward Ball Wakulla records the arrival time and amplitude of the reflected pulse. Springs State Park in Wakulla County, Florida. For more than The depth of the boundary is determined by knowing the 12,000 years humans have lived in the vicinity of Wakulla speed of radar waves in various earth materials and the travel Spring owing to the spring being a reliable source of fresh time from the transmitter to the reflector and back to the water (Rink et al. 2012, this volume). Numerous archaeological receiver. The character (amplitude) of the reflected pulse is an discoveries (Yates et al. 2001) and paleontological discoveries indication of the difference between the materials across the (Rupert and Spencer 1988) have been found in the spring area. reflection boundary. A higher amplitude reflection indicates a During renovations at Wakulla Springs Lodge in 1995, the late larger difference in electrical properties. For a more complete State Archaeologist Calvin Jones unearthed prehistoric stone explanation of GPR theory and practice, specifically intended artifacts (Jones and Tesar 2000). At the springs, archaeological for an archaeological audience, readers are referred to Conyers discoveries have been usually limited to small stone points or (2004). tools, typically a few centimeters in size and within two meters Typically GPR surveys are collected in a rectangular grid of surface. No structural archaeological discoveries have been pattern. The grid for the GPR survey at Wakulla Springs is found. Currently, efforts are underway to refine the ages of the shown in Figure 2. The edges of the grid were fixed in the earliest occupation of the area (Rink et al. 2012, this volume), field by driving stakes into the ground at 5 m intervals. Each This ground penetrating radar (GPR) survey was conducted by "five-meter" stake was located using a TOPCON GTS-230W staff of the Florida Geological Survey and the State's Public Total Station referenced to two permanent local benchmarks. Lands Archaeology Program to identify additional areas for The "five-meter" stakes were numbered sequentially 1 to 112 future excavation in the vicinity of the earlier discoveries by as shown on Figure 2. Jones. The GPR equipment utilized for this project was a
RAMAC radar control unit with a shielded 500 MHz antenna.
Methods The 500 MHz antenna was the highest frequency antenna
available and was selected to allow both the detection of the
For decades, GPR has been used in geologic and smallest size objects and penetration to depths of at least two
engineering exploration to identify subsurface geologic and meters. The measurement settings were 470 samples and buried engineering features. More recently, GPR has been 93 nanoseconds sample time, resulting in radar reflections used to aid archaeologists and anthropologists in identifying detected to approximately a 4 m depth. The sampling interval and planning subsurface excavations (Conyers 2004). was set to 0.05 m. Thus, a radar measurement or trace was
GPR works like conventional radar at an airport or weather collected every 5 cm along the profile. The data files for each station. Conventional radar has transmitting and receiving transect were collected on a field laptop computer (Panasonic antennas. The conventional transmitter sends electromagnetic ToughBook 30C). The raw data profiles were stored in radar waves through the air. The radar waves are reflected RAMAC rd3 format. by objects such as airplanes, clouds, or thunderstorms that GPR profiles, or transects, in both the north-south and the waves encounter. The radar receiver detects the reflected east-west directions were collected at 1 m spacing. Each waves, and knowing the speed of the radar wave in the air, the transect line was fixed using a tape stretched between pin flags distance to the object can be determined, located with a tape measure at 1 m spacing between the "fiveA GPR also has transmitting and receiving antennas. As meter" staked grid points. The transect lines were named based the GPR is dragged across the ground, the GPR transmitting on the numbered stakes at each end point. For example, the antenna emits a train of electromagnetic "radar" pulses into the transect beginning at stake 8 and ending at stake 31 is named earth. Materials in the earth, both natural and manmade, have transect 8_31. Because our transect endpoints were located characteristic electrical properties (dielectric permittivity) that at 1 m spacing between the "five-meter" stakes, we named determine how the materials respond to the electromagnetic the pin flags based on their distance from the previous "fiveradar pulses. When a radar pulse encounters the boundary meter" wooden stake in sequence. The pin flags between 'fivebetween two materials having different dielectric permittivity, meter' stakes 8 and 9 were labeled 8_1, 8_2, 8_3, and 8_4. The
VOL. 65 (1-2) THE FLORIDA ANTHROPOLOGIST MARCH-JUNE 2012
34 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 1Figure 1. Map of Florida, showing the location of the2, Edward Ball Wakulla Springs State Park.
flags between 'five-meter' stakes 30 and 31 were labeled 30_1, 30_2, 30 3 and 304. Thus the transect named 8_1 30_4 began at the first pin flag after 'five-meter' stakes 8, (1 m south of stake 8 and 4 meters north of stake 9) and ended at the 4th Figure 2. Aerial photograph of the GPR study area pin flag after 'five-meter' stake 30 (4 m north of stake 30 and adjacent to Wakulla Springs Lodge, showing 5 m grid
1 meter south of stake 31). In a few cases, because of heavy locations, permanent benchmarks, and grid numbers. underbrush or buildings, transects ended at pin flag locations where no "five-meter" stakes were located. In those instances, the distance from the pin flag was measured to a stake in the north/south or east/west direction. Thus, 70s 14_9_4 is the designation for a transect from a pin flag 14 m south of stake 70 to a pin flag 4 m south of stake 9. Approximately 15 km of transects were collected within the gridded area.
In order to analyze GPR results, the recorded data are processed by filtering to enhance the radar image. Table 1 lists the filters used for processing the radar data for this project and the filtering parameter values associated with each filter. In raw GPR data, reflections can be obscured by electrical noise or interference from the radar unit itself. Sometimes, extraneous reflections associated with nearby structures or electrical fields also cause noise. As the radar unit is dragged .............
across uneven ground, variability in how close the antenna is to the ground surface can cause extraneous reflections. The filtering procedures remove the noise or interference that obscures the reflections. For this project, ReflexW software by Sandmeier Scientific Software was used to process the data files.
Commonly, displaying the radar images in color will allow better delineation of subsurface features. Color figures for the radar images discussed in this article may be found on the Florida Geological Survey web page http://www.dep.state. fl.us/geology/ by searching for Wakulla Spring GPR results.
Typically, GPR survey data are processed in transects that
show a cross section view of the recorded radar reflections Figure 3. Aerial photograph of the east grid showing the along a single profile. The horizontal distance along the profile location of west and eaast profile 8-31, and south to northis plotted on the x axis and travel time is plotted on the y axis. profiles 14-54 and 17-50.
SMITH GPR AT WAKULLA SPRINGS STATE PARK
Because the travel time can be converted into depth, the y axis shows similar depressions at distance marker 21 and 45 m. is also scaled in depth below ground surface. Figure 7 shows a map view of the east grid area at time 6.735
In order to visualize the radar reflections in a map view, nanoseconds, corresponding to 0.34 m depth. In the map view, the radar amplitude data, recorded at all transects in an area, the radar reflections from a network of shallow irrigation pipes can be separated into horizontal slices. Each horizontal slice are clearly visible. Figure 8 shows a map view of the east shows the amplitude of the reflected radar wave associated grid area at time 19.02 nanoseconds, corresponding to 0.95 with reflections recorded at a single time. If the single-travel- m depth. At this depth, the radar reflections show two deeper time amplitude data from all transects are plotted on a map, we water supply pipes. Figure 9 shows a map view of the east grid can show a map view of the subsurface at a specified depth. area at time 40.01 nanoseconds, corresponding to 2.0 m depth.
At this depth, the scattered reflections indicate reflections from
Results the top of the limestone bedrock.
Figure 10 is an aerial photograph of the west grid showing
Figure 3 shows an aerial photograph of the eastern portion the location of west to east profiles 70s14_9_4 and 85_97n5 of the study area grid and shows the location of three transects; and south to north profile 100 2 68_3. Figure 10 also shows west to east profile 8 31, and south to north profiles 14_54 and the location of a rectangular area where the horizontal slice 17_50. Figure 3 also shows a rectangle that represents the area mapping procedure was performed. Figure 11 shows a cross within which the horizontal slice map view processing was sectional view of west to east profile 70s14_9_4. In this cross performed. Figure 4 shows the cross sectional view of west section the top of the limestone also varies from approximately to east profile 8_31. The cross section runs from stake 8 on 1 to 2 m depth with relief between pinnacles of as much as 1 m. the west to stake 31 on the east. The surface of the limestone Figure 12 (west to east profile 85 97n5) and Figure 13 (south appears to be pinnacled with relief of approximately 0.5 to north profile 100 2 68_3) also show top of limestone at 1 to 1.0 m. The depth to the top of the limestone varies from to 2 m depth. Figure 14 shows a map view of the west grid approximately 2 m depth to 1 m depth. Figure 5 shows a area at 0.5 m depth (10.10 nanoseconds time), showing radar
similar cross sectional view of south to north profile 14 54. reflections from two water supply pipes. In this profile the top of limestone surface shows more relief, Figure 15 shows the map view of the west grid area at ranging from a depth of approximately 1 m to deeper than 2 m. 1.5 m depth (30.11 nanoseconds time). At this depth, scattered At distance marker 29 m, there is a depression in the limestone radar reflections from pinnacles of limestone are visible over surface, with the bottom deeper than the depth of radar signal most of the area. There is an area across the center of the grid penetration at 2.5 m. Figure 6 depicts a cross sectional view where there are no reflections. In Figure 16 (2.0 m depth/40.01 of south to north profile 17 50. The top of limestone along nanoseconds) this area does show reflections from limestone this profile also varies from approximately 1 to 2 m depth and pinnacles. The absence of reflections at 1.5 m depth and
0 0 20 30 40 so 60 70
Figure 5. Cross-sectional view of south to north profile 14_54, showing radar reflections from the top of
limestone and shallow irrigation pipes.
36 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 1-2
10 20 30 40 50 60
Ilks~n surface Shmesonlow.-i
Figure 6. Cross-sectional view of south to north profile 17_50, showing radar reflections from the top of limestone.
Table 1. GPR data processing filters used to process the Wakulla Spring survey data and associated filter
Filter Used Objective Filter Parameter Value(s)
Subtract DC Shift Remove constant offset caused by On
interference from Direct Current
GPR power source
Automatic Static Correction Adjust for changes caused by On
differences in antenna coupling
Background Removal Remove horizontal banding created On
by system noise and frequency
Bandpass-Butterworth Remove high frequency noise Lower 62.5 mHz Upper
Gain Function Enhance low amplitude reflections Linear- 1
Timed-5 ns Maximum Gain
Note: The objective associated with each filter lists the kind of noise or interference that the filter is designed to remove. Where the filter used has different settings or parameters, the filter parameter values associated with each filter are listed in the table. Eor a complete discussion of the application of GPR data processing, please refer to
SMITH GPR AT WAKULLA SPRINGS STATE PARK
... Shadow of sidewalk
Figure 7. Map view of the east grid area at
6.735 nanoseconds, corresponding to 0.34.............
m depth, showing radar reflections from a
network of shallow irrigation pipes.
Figure 8. Map view of the east
grid area at 19.02 nanoseconds,
corresponding to 0.95 m depth, ":"showing radar reflections from /on,
water supply pipes. /
Figure 9. Map view of the east ::; .../
grid area at 40.01 nanoseconds, ..... ................... i
corresponding to 2.0 m depth, ....+:.?:ii .; ..... .......... i" '.,,.
showing radar reflections from the,/
top of lim estone. ...... .... .. : .7 .: .. ............ ... <: .. .
....... .... '':i,.. :.%," .. NorthE
38 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 1-2
Figure 10. Aerial photograph of the west grid,
showing the location of west to east profiles 70s14 9 4 and 85 97n5 and south to north
profile 100 2 68_3.
0 t10 20 30 40 50 60
0. -N %
wt V A Bnchmarks L tLine of Profile
Figure 11. Cross sectional view of west to east profile 70s14-9-4, showing radar reflections of the top limestone.
0 50 20
0 10 20 30 40 50 60
0 Rltnr 0
-20 ... ..t .
0 me surface
Figure 11. Cross sectional view of west to east profile 70s4-9-4, showing radar reflections of r the top limestone.
0 10 20
7 0... ... ... . ...... ....
Figure 12. Cross sectional view of wsth to east profile 50 97n5_, showing radar reflections from the top of msoe
SMITH GPR AT WAKULLA SPRINGS STATE PARK
Figure 14. Map view of the west grid area at K 10.10 nanoseconds, corresponding to 0.5 m depth, showing radar reflections from water supply pipes lines.
Depression in the surace of the limestone (area where no reflections were seen to depth of 1.5 meters) / Limestone
Figure 15. Map view of the west grid area at 30.11 nanoseconds, corresponding to 1.5 m 4
depth, showing radar reflections from the top i. of limestone.
of th topof th limoron
were seen at 1.5 meters depth .
Fi,,gure 16. Map.,a view of the west :grid area at. .....- .. ..... :,...;:
40.11 nanoseconds, corresponding to 2.0 m :ii
depth, showing radar reflection from the top ;i!: ...
of limestone...... iJ
40 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
presence of reflections at 2.0 m depth indicates a depression in the top of the limestone. Thanks to Jim Dunbar, now retired from the Stateof
Florida Public Lands Archaeology Program (PLA) for the Conclusions idea of using GPR at the Wakulla Springs Lodge site and for
Jim and Mary Glowacki, also from PLA, for operating the total The purposes of this GPR survey were two fold. The station to locate the survey grid. Also thanks to Jim for helping principal purpose was to identify sites in the vicinity of the me review the radar profiles to try to identify archaeological earlier Jones investigation where additional excavation signatures. Thanks to Dave Paul, Clint Kromhout, Jesse Hurd, could possibly uncover artifacts of interest. Of secondary and Guy Richardson for assistance in the field. Frank Ruper importance, we wished to use GPR to learn about the geologic of the Florida Geological Survey (FGS) helped prepare the nature of the surface of the limestone underlying the area. figures and Frank and Guy Means, also from FGS, reviewed
Use of GPR to identify sites for future archaeological the manuscript. excavation: GPR can be expected to resolve individual subsurface features no smaller than approximately 10 References Cited
cm at depths of 1 to 2 meters (500 MHz antenna, MALA Conyers, Lawrence B.
GeoScience USA, Inc 2010). Although we did not expect to 2004 Ground-Penetrating Radar for Archaeology. locate artifacts as large as 10 cm, we were optimistic that GPR AltaMira Press, Walnut Creek, CA. could identify anomalous geometric shapes or features in the shallow subsurface that could be an accumulation of debitage. Jones, B. Calvin. and Louis D. Tesar Using the 500 MHz antenna, however, this did not prove to 2000 The Wakulla Springs Lodge site (8WA329) be the case. We did not observe any anomalous features that a preliminary report on a stratified Paleoindian we could identify or attribute to ancient human habitation. Through Archaic site, Wakulla County, Florida. The
Modem manmade features crisscross the subsurface in the Florida Anthropologist 53: 98-115. study area. Since the construction of the first lodge in 1937, numerous pipes were installed for water supply from the Mala GeoScience USA, Inc. spring, landscape irrigation, and sewage disposal. Our GPR 2010 MALA GPR Training Presentation. Unpublished survey clearly shows the location of these pipes (Figures 7 and commercial brochure, MALA GeoScience USA, Inc. 8). Charleston, SC.
Use of GPR to delineate shallow geological structures:
in addition to manmade structures, GPR clearly shows the Rupert, Frank, and Steve Spencer undulating surface of the limestone in the study area. The 1988 Geology of Wakulla County, Florida, BulletinNo. 60. limestone surface appears to be an erosional surface of Florida Geological Survey, Tallahassee. limestone pinnacles approximately 1 to 2 m below ground surface. By looking at horizontal slices of radar data, the top Yates, William Brian, John P. Kilgo, and Melanie Damour of the limestone can be shown in map view. Areas where the 2001 The Wakulla Springs Sand Relocation Project, 1998 surface of the limestone is deeper may correspond to areas where the land surface was a topographical depression and No. 5, November 2001. Program in Underwater Archaeology, may be areas where future excavation should be focused. Department of Anthropology, Florida State University. For example, see Figure 15, which shows a large area of the western grid where the surface of the limestone is lower than surrounding areas.
The successful application of GPR in support of archaeology must take into account the likely targets that the radar is expected to delineate. For this project, it was understood that the Paleoindians of the age of interest likely did not build structures that would be large or durable enough to be detected by the radar. For this project, our best hope was to detect an undefined assemblage of tools and debitage or a paleo-topographic surface that could indicate an area for future excavation. GPR can provide the archaeologist with hard evidence for planning excavations when the target can be detected by the radar. GPR can still provide the archaeologist with soft evidence or guidance for planning excavations even when the target is too small to be reasonably detected by the radar.
GRANULOMETRIC ANALYSIS OF SEDIMENT SAMPLES FROM THE WAKULLA SPRINGS LODGE SITE, WAKULLA COUNTY, FLORIDA
Guy H. MEANS
Office of the Florida Geological Survey, 903 West Tennesee Street, Tallahassee, Florida 32304-7716
The 2008 excavations at Wakulla Springs have expanded when it was flowing or when it was a water-filled sinkhole! our understanding of the Paleoindian and possible pre-Clovis swallet, no doubt attracted Florida's first human inhabitants. occupations of Florida. The Wakulla Springs Lodge site
(8WA349) appears to sit on a sand hill or dune about 5 m Methods
above the surface of the Wakulla River, but understanding
the geologic history of the site provides an important context Depositional history of sediments can be inferred from for interpreting the archaeology and optically stimulated the distribution of grain sizes, because different transport luminescence (OSL) dating of the site stratigraphy (Rink et mechanisms (e.g., wind, river flow, beach waves) move al. 2012, this volume). To better understand the depositional sediments differently. For example, the wind typically moves history and origin of the sediments, a granulometric analysis smaller grains than a river can, which creates a different (i.e., analysis of sediment grain sizes and distributions) was distribution of grain sizes over time. The transport mechanism done. The results were somewhat unexpected and indicate the can be inferred by plotting the average standard deviation of deposition of sand at the site was more complex than initially each sample against the skewness of the sample's distribution assumed. Instead of straightforward, and expected, eolian and comparing the result to plots from sites with known deposition, the analysis showed that fluvial deposition was depositional histories (Balsillie 1995). also a major factor. Analysis of relict geomorphic features in Five sediment samples were taken from Test Unit B at the area indicates that most of the sand was probably initially the site (Thulman 2012, Figure 2, this volume). The samples deposited during a previous sea-level high stand possibly were taken as follows: sample 1 at 93 cm, sample 2 at 70 cm, during the Sangamon Interglacial, from about 75,000 to sample 3 at 50 cm, sample 4 at 28 cm, and sample 5 at 13 cm. 120,000 B.P. All depths were measured from below the site datum using the
Wakulla Spring is located in central Wakulla County, sampling guidelines of Balsillie (1995). Figure 2 shows the Florida, within the Woodville Karst Plain (WKP) geomorphic sample locations. The samples were analyzed at the Florida province (Scott et al. 2012), which extends from just south Geological Survey in Tallahassee. of Tallahassee to the modem Gulf of Mexico coast line and Each of the sediment samples were dried, weighed, and
east to the Steinhatchee River (Figure 1). Karst features, like then run through a set of metal sieves. The opening in the sieve Wakulla Spring, are abundant in the WKP because Tertiary mesh ranged in size from -2.25 phi (4.75 mm) to 4 phi (0.0625 limestones of the Lower Miocene St. Marks Formation mm). Sieve nests (stacks of individual sieves) were shaken for
and Lower Oligocene Suwannee Limestone are at or very 30 minutes on a Meinzer Sieve Shaker, and then each fraction
near the surface. These limestone units have undergone left on the individual sieves was weighed and recorded. The karstification (dissolution of limestone) through geologic time weight data were then entered into the GRANPLOTS program due to Florida's semi-tropical climate, abundant rainfall, and (Balsillie et al. 2002), which generated grain size distribution proximity of the limestone to the surface. The karst nature of the plots. Additional analysis showed the sediment is medium to WKP is reflected in the number and diversity of karst features fine quartz sand (Wentworth grain size classification) that is in the area, which includes sinkholes, swallets (features that moderately sorted. funnel surface water directly to the aquifer), springs, and caves
(both air and water filled). Results
Wakulla Spring is connected to the longest underwater
mapped cave system in North America at 51.5 km. It is world- The average standard deviation for all samples, which is renowned for the volume and quality of water it discharges. a measure of the sorting of the sediment, was 0.9493 (Table Wakulla is a first magnitude spring (i.e., discharge in excess 1). The plot of skewness and standard deviation shows that all of 2.83 m3 per second [m3/s]) with an average discharge of 11 five samples fell within the river sand (fluvial) portion of the m3/s) (Scott et al. 2004). The water that emerges at Wakulla graph (Figure 3). This distribution indicates that the sediment, Spring comes from the upper Floridan aquifer system, which at some point in time, was transported by fluvial processes is one of the most productive aquifer systems in the world. The (Friedman 1961). However, further examination of the results large volume of water potentially available at this spring, either shows that the depositional history was not straightforward.
VOL. 65 (1-2) THE FLORIDA ANTHROPOLOGIST MARCH-JUNE 2012
42 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65(1-2)
Stopographic high continues farther southwest and looks like a Ad !ii relict barrier island, beach ridge, or bay mouth sand bar. These
c areas of higher topography are oriented somewhat parallel to
the modem coastline, suggesting that they were created by near-shore marine currents and wave action. If this was the case, the sediment at the site could have been transported to the near-shore marine environment by a paleo-river system, where it was reworked through near-shore processes. This could have occurred during the Sangamon Interglacial Stage (ca. 75,000-125,000 years ago), or earlier, which was the last a pn vde time sea levels reached an elevation 6.6-8.3 m above modem levels (Muhs et al. 2010).
It may be difficult to envision how the sand around the Wakulla site was fluvially deposited, because today the site Ap j is 5 m above the level of the spring; however, given the
topography revealed by LiDAR, a plausible explanation is Gulf of Mexico" possible. Figure 9 shows several paleo-drainage systems that
empty into or near the spring. If Wakulla Spring functioned Legend as a swallet at some point during the Sangamon Interglacial,
4 8 16 these drainages would have flowed to the spring entrance
_ Jhune carrying sediments with the flow and down into the aquifer.
Undifferentiated At times of high river volume, which could have carried large
JaksooluonFon sediment loads, that exceeded the capacity of the swallet to
Toreya Fonation accept water to the aquifer, river water would have pooled
St, Mar s o nation around the swallet entrance and dropped its sediment load
* Springs around the opening. This commonly occurs in the area today,
gemphy and several swallets are located immediately to the west and
north of Wakulla Spring, such as Lost Creek and Jump Creek Figure 1: Surficial geology in the vicinity of Wakulla Spring (Figure 1). Further, a number of swallets in the Wakulla Spring (after Scott et al. 2001). basin have been shown via dye tracing studies (Kincaid et al.
2009) to be connected to the cave systems that deliver water to Wakulla Spring. Flooding frequently occurs today, especially The grain size distribution plots showed the samples were negatively skewed with bi-modal distributions (Figures 4-8). Negative skewness can represent winnowing effects commonly seen in beach and near-shore marine environments (Friedman 1961), and bimodal distribution suggests that two transport mechanisms acted upon the sediment (although both mechanisms likely involved water).
The apparent conflict in the results could not be resolved without further investigation. Based on the plots, it seemed most likely that the sediment was transported by a river to a near-shore marine environment where wave action and/or eolian transport may have taken place. The modern topography of the site makes either transport mechanism seem unlikely, so to resolve the apparent conflict, I examined additional geologic and geomorphic data.
LiDAR (Light Detection and Ranging) is a sophisticated imaging technique that allows the topography of the ground surface to be imaged even though covered by dense forest. Here LiDAR was used to construct a digital elevation model (DEM) for the area surrounding Wakulla Spring (Figure 9). The DEM shows a linear feature of higher elevation adjacent Figure 2: Sediment sample sites from Test B. to Wakulla Spring that extends to the south and southwest. This
MEANS SEDIMENT ANALYSIS FROM THE WAKULLA SPRING LODGE SITE 43
Table 1: Granulometric analysis output data.
Sample Mean (phi) Median (phi) Standard Deviation (phi) Skewness Kurtosis
1 2.2480 2 1405 0.9449 -0.4527 2.6953
2 2.2755 2.2077 0.9546 -04352 2.6735
3 2.2757 2.213 0.9557 -0.437 2.7254
4 2.2695 2.1781 0.9564 -0.37 2.5927
5 2.2755 2.2059 0.9347 -0.4247 2.623
Skewness Versus Standard Deviation
04 4 2
0 00 010 020 030 0.40 0.50 060 070 0 80 0.90 1.0 0 1.10 1.20 130 140
Standard Deviation (sorting) phi
Figure 3: Skewness versus standard deviation showing Figure 6: Sample #3 grain size distribution plot.
probable depositional environment (after Friedman,
Sample #1 FrequencyPlot Sample #4 FrequencyPlot
122 ~ 10
-3.0 -2.0 -1.0 0.0 1.0 2.0 3 .0 4.0 5.0 0 'StndrDation (sorize phi
Grain Size (Phi) -3.0 -2.0 -1.0 0.0 1.0 2.0 3.0 4.0 5.0
Figure 4: Sample #1 grain size distribution plot. Figure 7: Sample #4 grain size distribution plot.
Sample#2 FrequencyPlot Sample#5 FrequencyPlot
16 14.. . ..
~ 12 __12
CL 1 4 )1
.3. -20 .. 00.0Z. .
.3.0 .2.0 -1.0 0.0 1.0 2.0 3.0 4.0 5.00L 0a0100
Grain Size (Phi) -3.0 -2.0 -1.0 0.0 1.0 2.0 3.0 4.0 5.0
Grain Size (Phi)
Figure 5: Sample #2 grain size distribution plot. Figure 8: Sample #5 grain size distribution plot.
44 THEFLORIDAANTHROPOLOGIST 2
- n" 1 I
11 : : ;: ; O- ,-..-:- 7 ' v F -. -, ---- -- ,-
. .. n, -7 '?! 'k. *7-- 'T : .
- .. .. r g;01 R. '... I :....- ... '. -- '.
..... f l. ........ .. I ..... . 11 7
... W- 74 '.: "i.........A. 11
, '. 11 , .!. :;..... I' ........ ..- I
.. ','; II.-I :. ....... I .. ':5 .." :,', : 1 .
.. -.. .. .. .- A
. ... 1.
.. ....... .. :4 !','-;' ;'...' ,:Ir -.1% -1:.: '.":.: -:- ,-..''..
-. ... '.....-..... .- W : -11 ..:: ...... ... ," Iel "..
,'.. ..- .... .1- :... : .'- ....... .. k" ...... .j -- '
N w,:':. I W 4 -' v.g....:...
. ." .: '..:: .
.-.. Ll :: .. :. ..
,- .J i ,...:': '0
. ": :: !! i .' !:!i !o ..
.....-I.... I ,............, ::: -.- ... .1,11.
. 1 .. I.I.....
' :: :.-. ,nx...- Y.::. ::? -- ", "; : :' ........... .... ,
. .... ...
-' O IX W r. I.. .
:W --. -1. ......
, P ,' .K '' : :p ...
.. ..... ... ... .. ..." :' A ,"
" ::.. W : 1 .1. Id :::
..i %...: '. .'I !:! !!p":::r '? -%- %"...V.....," -- ,
.. "" : % : X I ..
M h A la Sp rin g ': X :.... ',: . v:;z ... ...-- ... :: I -- "'/%.
. ::....: ... .. ...'.X' W ,' '!: ::: ... j : '.. .' '.1 ,
11 ,.. ,'. '. i .:'s .:" ,. lr ..... ... .' ", I :; : n
.1. .. ',. : -... ..... ........ ..' .. .. ., C '1:1;0 '. j W.
- ..... -". .' '- .- :...... ..'.... ,:...:x I....I..... M .:.'.
. ... -.11-1 .1
. N .. :: X. .. .. .. '. C "' .... I ... 1. ..- ::I
... -.... ::r ".. '-":':' I ..: ....... .:; .: I ...... .1 ... :R ' .. .1 .. ..' i -- x --- .. '-' .... . ....
: -... ....' 1 '- ...... '.' .. .. .; ::!'-' I.A :.
.1,4 :' .... ::. .. W ..::r' '
' .. .... .... .... I .... .. : I.. .. ".
- 1 1,... ..
. :.!: i .,
- 11 .. .... ... ,.,.
:" Q .. : 'V I ...
I- I I
.. ., ... !!: :. ,V :: ?... ::.:, X .;;. ':t' '. ... t ........::... I-: ..
: :::'z 4 :." '::..:. !: n :. ,I ; :. 11 .. v- ..'.. ::. :';e '7
11 .. .... .. ... :::.'.. ...- ... : -.. ............. 11': .: !5-'!. :: ...'.. -.. .
'..... ..... ..... ... "'... ......- V :' : :x .'!- ........ ... .... --;. .........
..... -.1 I ... '
... :. I .. ;: .:. . .... .
. : .' :::.....','.... .- ...., .. .'.. .... .LDEke %W :- w 7w .- -: : -' :- M... ': ., z:' '. ., .. ...;- :
. 1 M ......... ..... ......,.,., : z .. .... .,. .:.::. 7.:.:. ... .. 1.
:: .. :, e- ...-" ..:.. : O i !:.:! S ... .1 ... ..- I .
. .....,....,.... !: :. :::. :! .:i.:::::4: '.' X:: ;" .. .:.',: r..' ." ... '- ; '
.. -- : :. -.-, : '
.1 1 11 : '. ......... :. :: :: ,, '. :: : !i ".............
:.. ';. -- ". 4 z' .....", .. ... ... ., .. :'N
. .... I ;% ...:: :::: : :. oll. .., .; ::: ... . ....
-..." .. .... ... .. .:.... ..'..' 1.
. v -.-- ....... .. ..- ...- :. .. :'. .:
'. .. ..'....98 :. I 4.
.... ... Z. ,
." ... 1 5---- ,' .. .... ....... .- '- ,' ,,. : ... .,:, :: .:.. .I.:.. ,,, .- '
:11 . .. : .- ... .
. ..... "I .- .- : .. .... ..... <1 ..
. 11 .... .. ...::.. .. ., % .: .". .:" .... .
11 ........ '-- ..'..- %::: x .:- .:: : :. .. .. : .' .I::'I.':.I:.: ...
... ::: c':: ... .. : :, : :: :: ....:: ,' :: :: ...... .. .. ::
. '. -- :::: .' ::: .. .. ':- ........,.,........... ....... ,. .. % J..* ;,::::..::::. :O '::i '.::: ..
- .. -.. . ... .... "
, ........ ... ...- ....' .. ... :%: :::. I.... ::.%.:."
.. ... ... I .. ....... :- : ..: i : -'
. :1 ... :. ..- : : .- .
.X.- ::- : ' ..:::::' 1 :Q .:!: T ,,,? ': : :: ..: ; .:,, : ....:: : :. : l:: ....' .
. ::: '; .:. % .. ...'...:- ... .::..';:: ..... I
.. '. I .:: : :::: : ... ... .... ...... . X. ....." .I.... .
-- ... . .... ....I. ..-.-.- ::....:'. :..::: ... .. ... ,,, .: .: ,-:, ....... ,,:..::. K ::. ...:: :::. ..,. ::. ... IV- I :: ,:::.. : .. :::: .. ." : : : ., ", :,. ..::: .. '..." '.... ::, .. .... f '. ..'.
.... It., .. ... .. ." ... : .,.: .: . ,, :_ ::: : : ....
....I. .......'....... '.. .......,- : .:... I
...I.- I.. I : . .. :;,-'...
1. : :i: : ............. ... '
... : ..- .". ..... ... .... .I..... ...... .. .:. ;I ...- ...'. :: x . : ::. .
. ... I ::: :' : . .. ..... ... 11 '.
.1.1 :;:::...- .:.- :.- .... .. ... ..: :'.... .-.I.-......:..:..:: .!'.:.: ':: '. '. ..'
_ :.;: j .. '. I ':: -: .-.
. I ,, .,., .. ,% : >k: ::: ; ... .
= 1 -- .. I ? .. :1..... ............. .. X- .. ..-.::. ,- ::: ::: z : :' iW A .. z. ...... '. ..::" : ..... '.% :: : ..,- ::.::'.:._ ... ... 1. 'n .. :.g'. j
1 11 4 :: I ::: ..C ::. ". .- I... ; -:.-.-.:: X: ..... .1. _- :: X .
- : 4 : :::X ... .., :.' ...:. "; .. ,- .. .-I
... .. ..I. .3 ..... .... I..... .... .. .. .. I .. .. :.: ,.. .:::.
I ..... ....'...'.. : ... I .. ..:: ... :.: !..:::: ,. .:. : ij : .: :.: k .!. 04C ':.:....: :..
..::: I.. W -.. .. .... ... .1
. .': I : ,-...:. .I::..:.' x;.::.. .. .. .... :. IL I
. 11 1. ",::' : .'- '. .. ..:...:.:... .: .... I .
... ... 1 .... : : :.:.:! .: i :.: j: ? : .... .: ::.
L o s t '. ..x T : l::: ::: .. .:i: .......... ... I-- ... ...... ...X .. .... .. ;... .:: ....:.:...::,
11 ... 1- DII ,. : N '.:i:!:' : ..: < .-: :11'.- ..- : : '-.'..:: : ::: .:::'. : : ':g :i nl:i!:. I-: :::,%. .:::: :::: .::.:: :.. :. .U .....-'.. : .. ...
.. I ..- :.-:::..:.::: :....; 2.:... ...
:.::.:: : .. .... %- '.' ..... I.......-I.................... I ,' : ...... .w ... : .. ;, .. .
-.. I 7 "I- J:!: : : :.::: ::%I.r .. i 4- ::- ..
- : : ... .... ::: ..' .. 1 : ........ ... ...... ..:.
'.'. .. : .. ..' ....... ., .. ...- ::. .' .... .
%', ':-. X ...- ........- .... : ::!.:: :: :: ....3. ... I-.. .., ::.::: .. .::: :.:.:: .: I .... .... .. 1 11 :.::..
k-::"'- ......... :: .... : f X ., :.' .:.. 1. :.U...... ., ... .............. 4.
' .... ..... ..... ... c l
.. '.; ... I .... ... -- .:.. ...' '.x 1 .
.1 ....... .. .. :: :'
' .... .. ....... ........... ....... ...... :: : i: : ::: ..: !.... ... ;. : :.- I
.1 ... ... .e- .....::- .. ....... .... ::- .. I... .:. ;:... .:::: .:.....- ..' '
: '. M'.1- :: :X, :.: .: 11 .. ... 1 .. -1 ,
C r o o k ,,, :.:.. 3 '.. :: ..- .: ... 11 .. ., .... <.O .Z.:::
- ....... ..- ::: ..
- .. I:, :: ..
' : ;';: N .. ... .:.:..x:.X...:...r ... .... -' i ;.....:' .
- .:- %.:. :..... ..... 1- :
. ... ..:. .. ...
. .:.- : :. % .... .: :';.:':...:.. .... ... :: R e fi c t i: '! i! :! !: :: ,. ,. i] N ''. ....... .
.. : ::: ......, : j: .'. .. .. ..:. ... .
z %'... ..... ... ..... I .." ... .: s :' :.' ....:.:, ,: ... 1 :.1 I .... .
... .....- '..:::: .:W ,: .:::.. .. :.."C .... .' .... .
. 1. .. ........ ... .'-.:
.. ... '.. v_ : ::. ... .I.. :
I 0- ... .. ......!!. :'X:: : ::::.: ..' ,,!. .. .... .. C .....-:.& .::
.. .:. ... I .... .. ... . .. ......... .I x . ..: ? . . .... ....
:.. ::%...::: .: ::::: :. X '- 0 .-' 4 % .. .... ... ..... .. .. .. .
I., f .-I.. .:. ...., zx : .... ...
,,:. "'. .. .:F ... .'.:: i .... ... I ...... ." .. -11. ,:- ; : : .:.: x .:: : z ..!. :. .:.111 I
..... I .1. .... ... : :.::,..
.. .. ... ... .... ...... .....
... .......... .... I.I .. .....- ... I .... .. ...... ... .. .
........ .....- ... .. .. .... :, :.::. ..-I x
- .. ' : :..'.' ...:... .' .:: :....'.... .. .. .. I .
-.';. I ., %.. :. .' I -: ....:....e .... .. .
. ..:.: .::: :: I .. .... .... --.. .. .'
. :4. ; .: -. .:..R.: ..... B e a ch rid g e/ . . --' ......
:M :6 :% ,- J L ....'..:'...' ... ..:... '. .11 4 : .......:.: :! : : : :,
... .1 ... . .
- ....I .X .:'O '. : : .' ::! :..:: :: .... ..'
....' ? j : : : :' : ..:... :: 1. .. % X.:.
1 ,:. 5 : .:. .: :.:: :., ..... ; .....
. :' I .... ... .-.: "'e
.. .: .. IX.:: ., .-.....- '- '. ..... .: :::... 1. 5 i! !! ,, 17::: .::::: : : ;!.
.. Palm drainage t: ...:..::. ..I. : .
. T! L .'- .:; ..... : : :.:: : ,:.: :: ...... ;. .
: : .... .. .......... .. .. I .... .... 11 .. ... ,
' .. .. ... '-i' ....... ....I. .. -:-"..- ::....::.:. .... i i
-:%.'. .:::: .. ... .. .. ...
... ... .1. I... I............ ........ I I. .. ... . ." .
.I.... I .. ',a'::' ....: W .-.: .,.!! ,. '.- ..., ... ... ... :..............; ..... .. I..- ...... I ........ .. .' .:....':: 1. :1 .:
: ..: F ..,:....': .. ... '. : .- ;., :P::: .: ,.<" :..... ,, ..,. .' ..
: .. ... .... ... ..' ,.
4 1 '. .. : :...... I.... ::....
. .. '. ..:.:. s a n d b a r I :::" : ..:: ...... ...".. .. ...:. ... :: ... :R .:. .: : .
.. i .- ... ....''. i- ;. .;, ::..:.:::.,
........:;.: ....... : .: .. ....... : 4W W .' .
:i : .0 ... ... ..:! v .... ... ... .. .. ..... ..... . .. ..' ... I ".
... .. ::: .... T .. 1. :: D : : : :: !; :x L :: .. ... .. ;'-..':
..... ... .I. .. ..... .. 1 ... ..... O : '.:
.... 1. ... ... .- ..... I 6: .- ...
X :: ::! .,.,." ..... ..,. .. ....... .. :..:.: ; .. ..
.... .... .4 I.. . ... 1. ... :.:: ..... i .. : ::: :. :. .:'.'. : :.
,::: : jt'Y .!: : .I. .'i .1 4 :. ::.:. ... 1.
.. ...... -.... .. ........".. .:: ,.- : ::- ... p : .-:.:::x -' .............
b a s i n s m ',, :! ,. : .: : ...,.! .:.. llill!7:;iiili!,B!!! I.. .., ..... . : : .. :' .!.:.-: .
': : : :.:.- :: :: ::: : ......:::: :..:.: :::!:. ., .. .1 "I : i .....; .. .::..- .... ........:
:.::,... '- a ...:.:...: .. .. :1. X.:. ...... ' ..-' -.1 'W. i. .... :...: :... .' ... .. .:. : :: ; ....: '.,.......,..
...... .." .. .....I. .- P R "J : : .I... ..... .. :. .: : .....:- -'. ....
'i.- ..., .... .. ...... x .................. "-- .. .. .. .:.... '. : ..... .... ., .:.::
. ... .. ..... .... :.. .. ..
- .* .. ... . .. .; '. :: :. .
' .. .. ........- ,.- : .:. .. X :. ..... :-..:.: .. ...Y
.'' .' ..,.'%.... ... .. .. >.NOW .:.:. ,....:..& ':I- .: ...i?.., z ..........- ..:...
- X....'. .. ... ..... ........ ... :. .: : :- :: %. .. i: :::%::::: . *: :!. ::.' :, .... ...,- :
::... .,..:,-.-,,,,,,,,,,"'::,; ,r ,..,,. I % : 1. ., :
. ,.. .. ... i .........' .........'
I.... . .. ....... .jx: .. 1:- ::.- .... .. ... .. ... ..
. 'n .. ....- .:: O:X:.: !.:.: I :!
- .... r ..
. IP .... '- ....". ....- ................ ..:.-::-.-- I .
- .. .. .. ... .....
.. ...... .... ...... . ... ... ... "
:,.. :::-::: : :: .::..;:.:f .: I ... .... r ... ..... ..... .......: ':.x: :: --' : : s : i l :'.:-:11:.:...;
- I- Y ::: W W .... . : ': I I : ,' .. ...-.... .'
,4 :. 1 .... ..... ... ........ .. .. .
t ... ::::: M : ... ... ..: I: : .
-'r '-.': -Ej :.'
: ;V : ......... .. --... .' ::: % :. : .: A
...... I ..... ... ::" ::':':: r : :.: I ..: ,'; ,. .. ., .:..R. r
. ..:. ." .. i::... I ... .:....: . ," : .. '. ': : xx:::::::. :r: I... .. .: ',::.. : : : n::.: i :: ::!- : :: '. .
, .:... 1. I.. ::::,:.... ... . Y- .:: r :: ....
.:: .1 : ::..: .. : ... .... I ..... .. . ; r I ..' . .. ... I ...... :. .. . : ,r I I,
. :. :." % 1: ...: ....... I ... .. ..
1 .r A ... : .. ., . . .... .... ..... ... .. ... .. ... .... I" % : .:: :r; ... .. ...
:, .... ..... %, :: I r I 11 11 e : :: : :..
. : .. .... : : .. ..::. ... I : J :. : ". . I ..
.:.: : . .. . :' .. : '.1 ... "
. : ::. : r .. . . i : J :::: !: -,"! :i .::..: JO : : 'X ..
'. .. .... ...... : .. .. .:::: : .. .
.. ... I ... .. .. ..' ... r .. ...
.... .. i? ....... .. ........ ... I I .. .: :. : : :::..:. : .:. 2 : .. : : . : :: : : . : : .. v: ..::.: .. 11 .
- ': . ::: .. ... . ..
: : :: ... : r : . '... '. . Ir ... : W . i :: . . ;:, : :x: : .... .: . . ..... '111. .r: ': .:'O..... .. ... .
. .1. .. r .'. . .... .'.... x . r .1 : . I : : .* ; . I : :. : '. . .:
.1 .. . I ... 'f :: : ': : 1. : : : . . . . r . . . I : . ; .
. .::: .:..::: .. : .
... ' : ,. ,. .: .1 ... ... I. .. ... .. . . . r :- , . . ..:.:.:: . .. .
... . ...... . I . ... : .:er ,.: .:! :: :: v .: . .... r . . .... .. .'. . r . . I I .. .. ,., -:.
. .1 ... :1 i d r I : ::::.:: .::: '.: : ::: ..; .. :. .I...... I.. aj .:.
... .r :. : r. I . . .% : .:. I ... : .. : ... r .
. W : I .. :: : ..: .. ... ..
..... < :- : :: ... .. I .: r :.:. : . . ... . ..I.. ....... : : .
<:.., .. .. 11 .. .. ... ..... :. .. :1 .... .
. ... .. .. .... : ::.:.: : : .. 11 ... ... .:::: ::' ..:.:. ...z .
11: N :;A .. Z' ... V :.... :. ... .1 . ... .. . . :::. r I.: ..... ... : : : X.:., I
1! ... ... .................... ........ I I .. .... I ...'..... :V !.:. ::. : ... : ......
, '. .. I : ... : '.. ... ..! :7 :i .. : .... ... .. :::: .... .. "I .1 1. ..:::: .::. .: ...' I
.... .. .1 .. .. ... .. ..' .
. .. ,:r .:.:: .:.. I.....
1. .. ... I ....% .::::: I I :: .... : :1.
. .. :- : : .. .- ::I:.:' ..:. :: ",I .: I.. .1 : : .. j' .
I... .. ... . I ... .. .. ...', :.:f :::: ..::: :s' '.::! :.. '.... ..... .... ..- I .. ::.%:: : : ::::j. :.:: :::: : r:: ... ..::
. .. x I .. : ... ..;..'... .'........ : 1 ::.::,.::, .
r: r .. : : r : !:. 1:' r r .. .. .. .. .. .. ... .. .1 % ... ... ..... ... ... .. : .' . :
'!!: ',4 .. ,: ..". .. .. .... ... .. ... ... .... .... . .. : : : %' '. .. .... ........ g !.: !.i- :i. w: : : !
. I I :: .. .. ::: :. :::: :::: : : :::. ,:' ... : .:. .. .. .. :: :: : ... I- .... ... .... .. ..... : . '.. .? .:.. : .. .. ..... .... : I X ... .
.,:: ... ,. .. .1 .. 'r r v ... ..-. .. .. .. ... .. ... ,. 'r .. '.
., 2 .. .:.... .' : ?,, . Y .. .... ..r, .... .... I- ..... ::: ... :... : : .::: .:. < :
. I 1 ,. 1. i.. .. e : .....
-- .. .... ... .:. '.. < :-. ... :.:.:. 1::' ;
14 ,; .. :.. : 4 . '. .... % r . ... I -..... : : 1 ..:. % :. ::. ;; .'.-: : : : : : .1 I: ".
... : P .:.::' :..::; ....,.:' :.r. :. :. : ,.:,Irl,',' 1, .. ... ...... .... .. .1 .....
. I 11 ... '. .. .. ..... .. 11 :' : : :.. : ... . .:.:, : ..
... ::,z:,- : : .:'. .' 111 .. : :..: ' . .x. ..' '.' .: r :::::. A .I. ,:::::". .: !:. : .. .
.. .. ....:f 1:.. :::... :.:' : :.!:. .. .... : ::
:" ., k ::!' !1:1 f .. A I : & . : .. ........ ..- .... .. : f ::: .. ........ .. ..
. ....... 11 .. .- .1. ... .... ...... .... .... ... ... %....... .. .. .. %::.. '.: ... '. ::' ..: .:: .. ... .. li%
. ... .. .. ..... ... ..... : % : A 'x:x .. I .% :. ., "'r ,-,.. %,'.'
. .. .. I . ... . ..:: r ... .. :: .
-1 .. .... ,:.:.: : 7 : : 11.- .i 1. : : : ..... % ..... .. :% .1 ..- : ... .. ';: .. : :. '
:: : : : .. ... .. .. ..... I .. ... ... I V ::::: ::: X .. : : . ..... .... . .
, ".- . 'r ... . . > ... ..: .:';
. .: .. . . : .. : . X I..... ......... : : .
.: .. :.: .: : I %. . . .:3 ..... .... .... r .' : : i: : : '.! : : i: k ... ...... XI :x. :: .. .. : .. ..... .1 .. ... ... ... .. P .. .1 ... ... . . .. .. .. 11 . .... . ::: :. r ... : :: : ... ... .
I :: ::: : i. : : : : : : ': .... . I .. .. .1
.. : : :: : .. X, .. .... ., .. .... 4.. ...... : .: : : : : :. 1. .. . . I .. .. I .
. .. .. . : .1 '; .. X, ..: : : 'r : : '- . ..
.. : I I .. .... I ... .. :. : : : :::: .. : :: .!. .:.: : .. : .
.: r : : :: W ', ... ... :::: ., : . .. ...... ..
.1 1 . X : .: .. ... ...
... .: .. . .:" . . r : . ' < 'f . ... . .::. ::.. ...
11 ., . .: .: :: !.f i.:.: : : : -.'::. . :.: : : : I.. . .
:1 X. - : : .. . .. : :.: :.: : : 1. .- ". I r . .. .. : ' : . I .......... ..
,; :. . . 'i. . t J.. ':W . .. . :: : .. . ..: : :.. : .. '. ,
1. I ;. :: 1, .. ., ...... .1 ... :: ... ... ... ,c .:.
.. ... ... ..... ... ... I X I I ., .. ...... ...... ... .... .. 11 . .. .:, :.: : ::
I . .. .. ... .... .. ::: .... .. . . I .1 ... r ... . : : "V x : . 1. K ""
. e ..... ..... .. .. .1 .1. . .. . . .' . ... e . .
1: : : : . .:. : x .: .. A , I . :: :
. . .. I .'..
% 5 ; :.;- '.V, : . ::- : . . I .. . .. ........ . .. I .. .. ,
.. I r r r r .. e. : : ... :. : ' X . : . I '. : : :: 4 . : : : ... . I .. :. .. . I % i.i .. ... :.. . ... '. : .: :' :. ..::: .::: : : : : %:X.. : :. .
I :. . . .... ... .. F :
, I .I..... ... .. . ... .1. .. . ... ... . ....... - .... . . . . : : : : : .. ..: '. .. .1 Z .. .... .
.. ... ... ;"': '.: .1. I r ... ... . . .. .. :. .: ..' ..... :. . ....
I I I r : '- .- .. : : y ... I . . ... .:: . ::: ': ::::: :: :.-. %- :': I .- '.'.' .. .. ... ...... .. : .. .
. : P ::. : ::.:. :.. ,; . : : :.. .., V : .. : : .. .. ... ...... I I .. .. .. . : X. ; ..
'. . . I .. .
1. .. .... .X ... .. . ". . . .4... ... ', : :N... .. : '. .. ... .: .. : . : : ..... : .X. .... ..... .. ... .. I .. . .
W 1 :.4 :: . ..
:. .:. :: : '.'.:.. . .. : : .:: I.. : .. : I, : .:::: . .. .... .: :. .
. : ... . 11, r : i : r .:. .
::: r I : :. ... : : :: : :: : : .; : ", .. ....... .. .... .
: *' .. ; : : r '. : : .r .. :. e .. ..... .... : :! :: ,6 : ... .... .. .... !- %::!!! ,,, !: : : .. .. .... f.
:: .. . . r 11 .. x" .. .
XA . . .' .. ... . .. .. : ., . ... .1 ... . ... , .. .. .. . ... ...... . . ... .1 ..... r . .., I.: .: I : v .:.% : ..:: r 1 :: : .:.: .::.. '. : .. : : ..:.: . x .
... ...... '. . .... I, I .... I -.-I ... I .... . . .1 .... : ::::: ..:X .. : ..:.' r .... .... . . ., : : :.: ..... ... ., . . I . :.,
......... r . V .. ...... I
.... ..... . 1. 11 . : ... . .. .. < .... ... .. 1. ... ... ... ..
; . ; .. : : : . I : : :.... .: : : : : : : : ::: .: ::. :.: . . .. ... .... ... .... :.: : :.: : .I. : :.:... . : :'. .- :. I .
, . . : : I....: .1 ..: :
- .' . x . .. : . : . . .. % . .. ... : e I I ... :,. I . . . .. .. ... ... I . . . .. .. ...... .. .., ... ;I ...... . .. : . :."
...... r .......... . : : : : : :. '& .1 .. ,
..... I . .... .... ... ...
: ... .1 ...... x : : ... .....
. .... : 1 .. :.% : .. I ... ... ." .: :1 ': : : ..:. ,', :: : j .. r ...... .. ... . ... . ... .. I I '; .... . . . '
O . .- ;: .:.. ; : : : : r r .:., : : .::.- :: : : i .:.:! : : :T! : ; i ... 'r: : . .'. .. : :::.: .::." . ..... .'. . : . .. ':..... :. x . .. . , I I .. :1
.. .. : . S Pl :1 1. I . . I . ... ... .. ..
..... .. I 1. I .. ... I . : ...::.:: ..:::.. ,:. ., ..
.. ... I . .11 I .1 .. 1. ... I I .1 I v- . : r ..... . ,. . ..... .... .... : S .9
, .... : : : . .. 1 .:. : X.. . :, P ..:.. . . . .. . : : : : . .. .
:.r x 11. ... : : x ,, ..: .: . .. .... .. .. . ., '.. .. : i .: ....
: : : :. : .... ....
11 : :: : .: : . : '; ....... .... .... ,
. . le r .. .. :X' :.
... ., .:: :. . . I ..... .
I I . .. .... . .... ..... ... : .. ... .1, j. . . . .... ... ... : . .. .. :. .. . : ': X ."
r' :: : ::. . :: '. -I . .. ... ..... : ... ': :::i r ... : : :.; :i: I j : : : : .. :, : e : ::: : . .. : '. ...
' .... ... .... .:.: :: .. : ...
: : : ... .. : xx, : -:- . ' : ..: : : .. .:.: 1 .. : .....
. . .. . '. ..... .. .... .. . .1 1. .. > .
. .1 I .. I I : ..... .. .... .. :- . . ,
.... . .
I % I .i : : : ::. :N : : : : : : : : .:. .... . . :.:: : .. :. .... . .... :
. : . : :;.: . : :: :!: : : :: : : .... : .
.... .... ..: :: % I . : : : : : . .:.
..:.4, ; .. .. 1 1 :: : : ; r : . : ..' : . :. : : : :::: : .: : : : ... .... 1. I .. : :.
. .. % : ... I" : ::::.. . . ...
.. : : .. .... ..' .: ::: : : : : ... :.
.. .. . . I '. .. .. ..... ... .. ..... ........ ; .:; .: ::. :::: 1 . :: i : :!.: '; : ::: : : :- : :: : . .. .. .; : .: :::: : : !: i : : : ::. :.::. .: ; . : '. Q :.:. : ..... .... .": .. .. ., .... ... .
. 1--:...; : 'X Iz . : : :.: : 1:1 .. :1 ..: ... %. .. .. % ... ... . .: .. ..:.: : : e : ii! !! :::: !! ,! . . : ::: : : I
I :: : r 5' I.. ". : :..: : ::..' I . < : . .
:.V .. : r 1. :.: : : xxx .. :.'.' : .: !: - - - : : < .. . :...A ...
. .. I, .: . '. .. : . ... '. ..' > . . ...... .. : : ": :;.! : A n : : : F : :! :: : :: ; 4
. :: .: ... : :: : :! : :! : :: : .. ..: : . I .
..Ie ': : : : : : ... ". :1 A. ... .: .: : < K : :.: : ;: :.:: : ., : ... x .::. ki: : Ii : : !: ".' :i
..... I I r, : :...: : :. : I .... ... .. .... ... .1 .. .... . .- .
. .1 .. .. ... ...... : :. .. .. :. . ::. ..: : .. :: I ::: : : : r 1 :
., I .. r. .. .. ..:x xx : .:: :::::::. :.: : : : :: .. : ..
... . ... 1. ... .. .- ... .. :: ... . ': ;'
.. . I . r : :. : : .:..::. r ., 1
.. 1. ... .. -.... . I x : : i 1 : !2 ' : : :b". .. ... .. ...... V .,
.. . 1. .. ...... . .... .... .1 I.. ... ... .'. ..
I" .. ... 1. ; : :, : ... .... ...... .;5 .: ... ..
% . . ... I .... : .... I.. I .... .. ...
. ... ... ...... .
0 I.: .. !. .!: r : : ': :::: I .. .... ......; .. ... I I %, .
.. .1. : : ': ........ ...... ... ... I ... ... '. ..... : ... I ....... ....... X .... : : :..: : ,. :: :. .. .. .. .... : . : W : : :: z 5 !!! . .:..: ::
. : .. ... :%:: ...... .. .. .. .. ..... ........ . ,i : .
' ... . I .. : : .r.' -.. .. ...... : : .: : : .: r:: ' : : !:: : .: . ... .
. .. 1. .. I .. .. : .. ::: .
. .... ... I : .. :: : ::
.. ... : :. :. . .:: : -.- : ::. :.:.:.. .%; : ... .. : .. :: % :: ::. :: : : :.:: ... X. ... X I .. . ," .
... . ..
"P, .. ., ...... . ... '%. .: .
. I ;. . . .. ... .. . .. ....... ...
" I I r,, ,. W z ; .:. .1 .... ... ...:: ': ........ .... : .. .:.:::. : Y:'. .. .'., :.! !. ::::: xx: :: : : ..: :' : .: : ... .. ... :: :.: .1 1.
. .1 I .: : % . '; ; ': ... .
.. ... I.,. .. .... .. : : ... .. V : .. ::::::: : ::.:.: .. ,j : .:.:
, I ... I .... ... ... 11 .... . .:. : :: ..: : : . .:.:r ... .: ... : ,
I 1. I .: .. ., y r :.. ... : .. . . ..... .. ... v :.
,. : ; r .. . ........... %... ...... . "
. '. ............. I ... .. I .. .... ......
' .... .1. : : :i; .. .. .1 -: : ::.: : '.. ': ... ::::. ... ': ... :. .. : : :::.. :.:: : :. .. .. .
. . ..... .
: .. ...... :.: k ... . .. :z ..
. f I . : .: : . . : : ; : .:. :1 . .. .. I .:.' .. : 1.
, 4 .... .... .. ... : ... 1. .. r
. x::.: :::::: : .. 7 X : .. . ; .
.. ... I .. .... I e . : : :' : .. : I .
' I ... ....... ... ..... 1. I % . : . ; ;. I r 'I, ... . ... . . ... ... "
.% ... ;' : ".'. : . :: ... r. .1 r ... . .. ... ... .. ... .... ...:... ...: :, :.::.r:. .: ... .
. .... . . . .... . .... . .... .. .
: ; ', 'r ... ..' .: ,, : : :,% :.: :: . .. . '. . . ..:.:. ... .1 ..
,. ..:.. r r I :: : .. .... ,;:.:
.. .. '. . .. ..... . . J . .. .. . ... ... .
... X .x .... ... .. .. I .: . ...
.. .. I I .. .... .. . ... .. ....... f :11. .. .... : ... '
... .. I ... .... .. :: :.' r... .:: .: I :: .. : ... I .. .1 .... ... . :. .1 .
. I .... .. 1 .. ., : .... I ':' :. :. .. ... .. ... ......, ::.:::. : .. .. I
". .... r... 4 .: X. : : : :: .. .. :::.: :.. j :: . : : :: : .. ... .. '... : : .1 . .. '
: ... ... .. ... ... : .: :.::::. .. 'x.:.. ..:. o : . : :x i..Z :. .... : .
. 1. .. ..... I I 11 I .. . ... I., ... ....... ... . '. .. .: ... . '
... .. : ...i :. W '. : !: .: 4 ::: fl . 4" .:: . : . ... .... ,:.:. .1 ...... . . ...... . . .... ...... .. :. : : V ... : .: i :: ] : : : i ... . .
... .. I .Z : r . '. . !! : ::
. I ': : : : !: : : ,::: ... : :.1
I .. ... ..... I < .::.. . ........
.; ... r .... .. .... ............ 1. . . ... ... .. '... .. I :N ..::. .:.. .. : : : ; : :: : : : :'I .. ..... . .. ... ..... .. . . .. r .. .. .
MEANS SEDIMENT ANALYSIS FROM THE WAKULLA SPRING LODGE SITE 45
identify conclusively the origin of the sediment at the site and, Kincaid, T., G. Davies, C. Werner, and R. DeHan additionally, skewness versus standard deviation plots are not 2009 Demonstrating interconnection between a wastewater a definitive test for determining the environment of deposition application facility and a first magnitude spring in a of sediment. However, based upon the current evidence, the karstic watershed: Tracer study of the Tallahassee,
most likely scenario for the origin of the sediment at the site is Florida Treated Effluent Spray Field, 2006-2007, that a paleo fluvial system delivered sand to the area, probably Florida Geological Survey Report of Investigation from the west/northwest, and deposited it at the site during a Number 111 (DRAFT). Florida Geological Survey, time when Wakulla Spring functioned as a swallet or during Tallahassee.
a time when sea level was a few meters higher than present.
Muhs, D.R., K.R. Simmons, R.R. Schumann, and R.B Halley. Acknowledgments 2011 Sea-level history of the past two interglacial periods:
new evidence from U-series dating of reef corals Many thanks go to Michelle Ladle of the Florida from south Florida. Quaternary Science Reviews
Geological Survey for processing the sediment samples. 30:570-590.
Thanks also go to Dr. Thomas Scott for his assistance in the field. Frank Rupert, Dr. Kevin Burdette, Jim Dunbar, and Dr. Rupert, Frank, and Steve Spencer Christopher Williams provided editorial comments. Finally, 1988 Geology of Wakulla County, Florida. Bulletin 60. Dr. David Thulman provided numerous editorial comments Florida Geological Survey, Tallahassee.
and much appreciated guidance, which led to the somewhat timely completion of this paper. Scott, T. M., K. M. Campbell, F. R. Rupert, F.R., Arthur, J.D.,
R. C. Green, G. H. Means, T. M. Missimer, J. M. Lloyd, W. J. References Cited Yon, and J. G. Duncan
2001 Geologic Map of the State of Florida. Florida Balsillie, James H. (compiler) Geological Survey Map Series 146. Florida
1995 William F. Tanner on Environmental Clastic Geological Survey, Tallahassee.
Granulometry. Florida Geological Survey Special
Publication 40. Florida Geological Survey, Scott, Thomas M., David Paul, Guy H. Means, and Christopher
Tallahassee. P. Williams
2012 Geomorphic Map of Florida. Florida Geological Balsillie, James H., Joseph F. Donoghue, Kenji M. Butler, and Survey Map Series. Florida Geological Survey, Jennifer L. Koch Tallahassee. (in preparation.)
2002 Plotting equation for Gaussian percentiles and a
spreadsheet program for generating probability plots. Scott, T.M., G. H. Means, R. P. Meegan, R. C. Means, S. B.
Journal of Sedimentary Research 72:929-933. Upchurch, R. E. Copeland, J. Jones, T. Roberts, and A. Willet
2004 The Springs of Florida. Florida Geological Survey Dunbar, James S. Bulletin 66. Florida Geological Survey, Tallahassee.
2009 Wakulla Springs Lodge Site, Project Ice Age Florida.
Report to the National Geographic Wright, E. E., A. C. Hine, S. L. Goodbreed, and S. D. Locker
Society Committee for Exploration, Partial 2005 The effect of sea-level and climate change on the Fulfillment of Grant #8404-08. Bureau of development of a mixed siliciclastic-carbonate,
Archaeological Research, Florida Division of deltaic coastline: Suwannee River, Florida, U.S.A.
Historical Resources, Tallahassee. Journal of Sedimentary Research 75:621-635.
Friedman, Gerald M.
1961 Distinction between dune, beach, and river sands
from their textural characteristics. Journal of
Sedimentary Petrology 31:514-529.
Grimm, E.C., W.A. Watts, G. L. Jacobson, B. C. S. Hansen, H. R. Almquist, and A. C. Diffenbacher-Krall
2006 Evidence for warm wet Heinrich events in Florida.
Quaternary Science Reviews 25:2197-2211.
Ivester, A.H., D. S. Leigh, and D. I. Godfrey-Smith 2001 Chronology of inland eolian dunes on the coastal
plain of Georgia, USA. Quaternary Research 55:293302.
K >r E
Florida Aflthrop)010giCaL Society K.. ANNUALM MEETIN
Tallahassee, May 11-131-2O12
. .~ ......... I
mis. .a i i~ . .........."-m W
K t dm A-K K ~
THE FIRST FLORIDA "BLING": PALEOLITHIC BEADS
State Archaeologist, B. Calvin Jones Center for Archeology, Governor Martin House, 1001 de Soto Park Drive, Tallahassee, Florida 32301
email.: MGlowacki@dos.state.fl. us
The following discussion documents the discovery ofPaleodian
Introduction beads from two Florida sites, the Wakulla Springs Lodge
site (8WA329) and the Ryan-Harley site (8JE1004), located
Well before swash-buckling pirates and modern rappers, in North Florida (Figure 1). These finds are summarized in humans have sought out "bling," shiny little bobbles and other the context of beads reported from other Paleoindian sites in tantalizing items of body adornment. When we consider the North America toward an understanding of the nature and role relationship between humans and the aesthetic, it is clear that of this object of embellishment among the earliest Floridians. peoples' desire to possess, and, especially, to wear "pretty
things" is deeply rooted in the past. Scholars have recorded Florida Paleoindian Beads
evidence for trinkets such as beads, small decorative objects
pierced for threading or stringing, used to adorn the body and As part of the Florida Bureau ofArchaeological Research's clothing, dating as far back as the Middle Paleolithic (ca. National Geographic Institute-funded Project Ice Age Florida, 300,000- 40,000 years ago), in the Old World. Some of the the Wakulla Springs Lodge Site (8WA329) was revisited in earliest demonstrable examples date between 100,000 and 2008 to record further evidence of its Paleoindian component. 135,000 years ago. For example, seashell beads (Nassarius The site previously was tested on several occasions by B. gibbosulus, nassa mud snail) were excavated from Es-Skhul Calvin Jones (Tesar and Jones 2004), who identified a hearth in Israel and Oued Djebbane in Algeria. Their recovery, more with associated Paleoindian tools and debitage. During the than 200 km from the coast where this gastropod shell is found, 2008 project, directed by Dunbar, fine screening (with window indicates that a broad trade network likely circulated the beads screen mesh of 1.4 mm) of material from an excavation depth throughout the region (Vanhaeren et al. 2006), although the of 90 cm revealed the presence of a stone "seed" bead (Figure beads could have been directly procured. Somewhat later, but 2). As the term is used here, seed beads are small spheroid or much more definitive in nature, is a cache of 41 perforated discoid beads ranging in size from under a millimeter to several estuarine tick shell (Nassarius kraussian) beads excavated millimeters in diameter. The Wakulla stone bead is flat and from Blombos Cave in South Africa, dating to 76,000 years disk-shaped, measuring approximately 4.3 mm in diameter. ago (d'Errico et al. 2005). By the Upper Paleolithic (40,000- Only one other bead associated with a Florida Paleoindian site 10,000 years ago) bead manufacture took off, and bone and has been reported. It, likewise, was a stone seed bead (Figure ivory beads became a common artifact in archaeological sites 3), recovered from the Ryan-Harley site (8JE1004), located in of that time period (e.g., Knecht and Pike-Tay 1993; Krotova the Wacissa River about 35 kilometers to the east of Wakulla and Belan 1993; Shimkin 1978; Soffer 1985). Also at that Spring (see Figure 1). This bead had been excavated from the time the continent of Australia was inhabited, and along with Suwannee (Paleoindian) stratigraphic level along with other its human occupation came shell beads. Their manufacture micro-debitage and artifacts. It is also circular-shaped but persisted well into the historic period (Balme and Morse smaller than the Wakulla bead, measuring 2.3 mm in diameter. 2006), attesting to the longevity of these objects' value and Florida Geological Survey geologist Harley Means, their enduring desirability. who discovered and reported the Ryan-Harley site with his
Surely, then, it is no surprise that by the time humans brother Ryan in 1999, conducted an analysis of these two reached the New World, the production of beads and other beads using a scanning electron microscope (Harley Means, decorative items was part and parcel of their "cultural" toolkit. personal communication, 2010). This instrument employs However, there are relatively few examples of Paleoindian a high-energy beam of electrons in a raster scan pattern to adornment reported in the archaeological record. This is generate a topographic picture of a specimen's surface and particularly true for areas such as semi-tropical Florida, where an assessment of its composition. The beads also were ionterrestrial preservation does not compare to the dry, desert probed to determine their elemental composition. The results rock shelter and cave and, to some degree, the open air sites of of the analysis indicated that both beads were made of silicon the Southwest, from which the best examples of early people dioxide, with the Ryan-Harley bead likely sourced to local in the Americas have thus far been found. Nevertheless, on- Suwannee chert and the Wakulla bead fabricated from a going research continues to reveal more and more details quartz pebble. Additionally, both beads were bi-directionally about Paleoindians, even their taste for the aesthetic.
VOL. 65 (1-2) THE FLORIDA ANTHROPOLOGIST MARCH-JUNE 2012
48 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
dc ac ao f dl
.. ... ......
Figure 1. Map showing the location of Wakulla Springs Lodge Site (8WA329), Ryan-Htarley Site (8JE1004), andSlt Hl (8JE121); the latter is a Florida Paleoindian site where possible bead preforms were reported.
perforated through their centers, and both displayed the bidirectional, conical holes from drilling.
While these two Paleoindian stone beads seem like isolated examples, their size and posited use suggest that beads during that time were manufactured in a relatively large quantity. One would guess that enough were produced to fabricate a necklace or stitch a design into an article of clothing. It should be noted that two large ivory bead preforms were identified at Sloth Hole (8JE 121), another Paleoindian site in North Florida. Excavated by C. Andrew Hemmings (Bradley et al. 2010) in the 1990s, this site generated other carved pieces of ivory indicative of an ivory/bone workshop. The preforms support the notion that beads were being manufactured at this site and ........ S
comprise part of the Clovis component site assemblages.
Other Paleoindian Beads Figure 2. Obverse and reverse of the stone seed bead,
Wakulla Springs Lodge Site.
In the context of other Paleoindian sites elsewhere in
North America, a pattern is certainly seen. Small beads have ,
been reported as far north as British Columbia, Canada, at the "
Charlie Lake Cave site. There, a larger pentagonal-shaped .....iI:: %ii
schist bead, measuring 13.5 mm x 11.6 x mm 1.7 mm was '
recovered in association with a fluted point and other lithic ...!i} 0,
remains (Fladmark 1988:376-378). To the south, in the !: i
Northwestern Plains region, Clovis-age sites in Washington, ..."
Montana, and Wyoming have produced flat disk beads, similar.... to the Wakulla and Ryan-Harley specimens, and perforated rods ____or tubular beads made from bone and stone (see Lahren and '....=; ,,,, .......
Bonnichsen 1974; Frison and Stanford 1982; Gramly 1993). In western New York far to the east, at the Hiscock Site, a diskshaped bead made from sandstone was recorded from the site's Figure 3. Observe and reverse of the stone seed bead, Late Pleistocene level (Ellis et al. 2003:229) and another was RynHreSi.
GLOWACKI PALEOINDIAN BEADS 49
Figure 4. Left, bone seed bead from the Shifting Sands site in Texas; right, chert seed bead from the Ryan-Harley
site in Florida.
recovered from the Paleoindian component of the Sugarloaf multiplied by fifty, would not weigh enough to slow a group's Site in Massachusetts (Gramly 1999:30-31). Even farther south migration or be difficult to store discretely. Yet Paleoindian and to the west, the Folsom site of Shifting Sands (Hofman et peoples and their predecessors sometimes favored them over al. 2000) in Texas has generated a small bone bead (Figure 4). larger style beads. Did this have something to do with their Finally, on the Pacific coast of California, a Pleistocene age technology or their access to material sources? One thought site, CA-SMI-608, produced numerous shell beads (Olivella is that the small seed bead held an aesthetic quality that was biplicata, purple olive snail) found with chipped stone artifacts reflected in the high level of skill and patience required to and food remains, primarily fish and shellfish (Erlandson et make it, which carried more value than other crafted items. al. 2005:679-680). The investigations of this island site point This is quite a dichotomy from hunting megafauna, which to a seafaring lifestyle of Paleoindian peoples of the West also required great skill and patience, but of a different order. Coast that included not only essential cultural items, but also An appreciation of the small and wondrous in the face of that luxury ones as well. As with Middle Paleolithic communities which is behemoth may have struck a cord. of the Middle East and North Africa, shell beads likewise may While this is a cursory discussion of one "small" artifact have served as items of exchange (Vanhaeren 2006) for these category of the Paleoindian period, it reminds us of the depth mobile groups of the West Coast. of complexity of prehistoric culture that archaeology at times
is able to glean. But for now, it can only be said that true
Conclusion "bling" is in the eyes of the beholder.
The number of examples of beads and other small crafted
objects from Palcoindian sites demonstrates that the early References Cited
occupants of the New World appreciated the aesthetics and
invested time in making things of beauty (see for example Balme, Jane, and Kate Morse Bradley et al. 2010, Hemmings 2004, Dunbar and Webb 1996, 2006 Shell Beads and Social Behavior in Pleistocene and, http://www.texasbeyondhistory.net/gault/images/incised. Australia. Antiquity 80:799-811.
html). The seed bead raises some interesting questions about
the nature of its value by early societies. There is not much Bradley, Bruce A., Michael B. Collins, and Andrew Hemmings "bling" for the "buck" with seed beads, even if dozens could be 2010 Clovis Technology. International Monographs in applied to a single thread or stitched across an entire headband. Prehistory, Archaeological Series 17. Ann Arbor, So, why make beads so small? Yes, they are easy to transport, Michigan but even an item twice the average size of a seed bead, and
50 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
d'Errico, Francesco, Christopher Henshilwood, M. Vanharen, Knecht, Heidi, and Anne Pike-Tay and K. van Niekerk 1993 Technological and Social Dimensions o
2005 Nassariuis karassianus Shell Beads from Blombos 'Aurignacian-Age' Body Ornaments Across Europe.
Cave: Evidence for Symbolic Evidence in the Middle In Before Lascaux: The Complex Record of the Early
Stone Age. Journal ofHuman Evolution 48:3-24. Upper Paleolithic, edited by Randall White, pp. 277799. CRC Press, Boca Raton.
Dunbar, James S., and S. David Webb
1996 Bone and Ivory Tools from Submerged Paleoindian Krotova, A.A., and N.G. Belan
Sites in Florida. In The Paleoindian and Early Archaic 1993 Amvrosievka, A Unique Upper Paleolithic Site in Southeast, edited by David G. Anderson and Kenneth Eastern Europe. In From Kostenki to Clovis. Upper
E. Sassaman, pp. 331-353. University of Alabama Paleolithic-Paleo-Indian Adaptations, edited by
Press, Tuscaloosa. Olga Soffer and N. D. Praslov, pp. 125-142. Plenum
Press, New York.
Elis, Christopher J., John Tomenchuk, and John D. Holland 2003 Typology, Use, and Sourcing of the Late Pleistocene Lahren, Lawrence, and Robson Bonnichsen
Lithic Artifacts from the Hiscock Site. In The Hiscock 1974 Bone Foreshafts from a Clovis Burial in Southwestern
Site.: Late Pleistocene and Holocene Paleoecology Montana. Science 186:147-150.
and Archaeology of Western New York State, edited
by R. S. Laub, pp. 221-237. Bulletin of the Buffalo Shimkin, E. M.
Society of Natural Sciences Volume 7. Buffalo, NY. 1978 The Upper Paleolithic in North-Central Eurasia: Evidence and Problems. In Views of the Past, edited Erlandson, Jon M., Todd J. Braje, Torben C. Rick, and Jenna by L.G. Freeman, pp. 193-315. Mouton, Paris.
2005 Beads, Bifaces, and Boats: An Early Maritime Soffer, Olga
Adaptation on the South Coast of San Miguel Island, 1985 The Upper Paleolithic of the Central Russian Plain.
California. AmericanAnthropologist 107:677-683. Academic Press, Orlando.
Fladmark, Knut R., Jonathan C. Driver, and Diana Alexander Tesar, Louis D., and Calvin B. Jones 1988 The Paleoindian Component at Charlie Lake Cave 2004 Wakulla Springs Lodge Site (8WA329) inEdwardBall
(HbRf 39), British Columbia. American Antiquity Wakulla Springs State Park, Wakulla County, Florida:
53:371-384. a Summary of Eleven Projects and Management
Recommendations. Florida Master Site File Survey Frison, George C., and Dennis J. Stanford (editors) Report #6602, Division of Historical Resources,
1982 The Agate Basin Site.: A Record of the Paleoindian Bureau of Historic Preservation, Tallahassee.
Occupation of the Northwestern High Plains.
Academic Press, New York. Vanhaeren, Marian E., Francesco d'Errico, Chris Stringer,
Sarah L. James, Jonathan A. Todd, and Henk K. Mienis Gramly, Michael 2006 Middle Paleolithic Shell Beads in Israel and Algeria.
1993 The Richey Clovis Cache. Persimmon Press, Science 312:1785-1788.
1999 The Sugarloaf Site. Paleo-Americans on the
Connecticut River. Persimmon Press, Buffalo, NY.
Hemmings, C. Andrew
2004 The Organic Clovis: A Single Continent-wide
Cultural Adaptation. Unpublished Ph.D. dissertation, Department of Anthropology, University of Florida,
Hoffman, J.L., R.O. Rose, L.D. Martin, and D.S. Amick. 2000 Folsom Adornment and Bone Technology. Current
Research in the Pleistocene 17:42-45.
ANOTHER WAKULLA SPRINGS MASTODON?
KEVIN M. PORTER
Bureau qfArchaeological Research, Division of Historical Resources, Florida Department of State, 1001 de Soto Park Drive, Tallahassee, Florida 32301
email: firstname.lastname@example.org. us
In August of 2007, the Bureau ofArchaeological Research (Mammut americanum) and while the observed skeletal (BAR) received a report from the Florida Park Service (FPS) fragment did not yield reliable evidence of skeletal maturity, that Park Ranger Jason Vickery had discovered an unusual the size of the alveolar cavity is suggestive of a young adult. In skeletal fragment(s) while planting eel grass in the spring basin hopes of dating the material, a bone fragment, which appears of Wakulla Springs State Park, located in the Wakulla Springs to be a broken piece of the maxillo-zygomatic area of the face Archaeological and Historic District, Wakulla County, Florida. (presumably the same side as the inverted alveolar fragment), In response to the report, BAR and FPS staff investigated the was recovered for radiocarbon (14C) dating but was found skeletal find, now known as the Vickery Mastodon, which later to contain insufficient amounts of bone collagen to test. lies approximately 130 m north of Wakulla Springs Lodge To determine if there was enough depth in the sediment
site (8WA329) and shown in Figure 1. Currently this site is column to accommodate more of the animal's cranial and! considered part of the Wakulla Springs Archaeological and or post-cranial skeleton, a 1 m probe rod was inserted to its Historical District, however, if found to have archaeological full length into the sediments, approximately 5-8 m north of association, it will be classified as part of the Wakulla Springs the exposed bone. This indicated that there was at least 1 m underwater site (8WA24) that surrounds the springhead. This of loose sediment overlying the limestone formation, which article addresses the significance of what may be the latest in is sufficient to contain more of the skeleton. Following a long line of Pleistocene megafauna discoveries that have documentation, sample collection, and ground probing, the helped to shape our current knowledge and understanding of bone was covered by a light layer of sediment in order to the environmental and subsistence patterns of Paleoindians of protect it until it could again be investigated. the southeastern United States. In December of 2007 the site was investigated a second
time to more accurately determine the depth of the sediment
Site Investigation column in the area around the skeletal remains. At that time
staff from BAR and FPS partnered with the Florida Geological
Initial investigation of the Vickery Mastodon revealed Survey (FGS) and Florida State University (FSU) to conduct a skeletal element, resting in approximately 1.5 to 2.5 m of a vibracore operation of the site. This involved guiding an water, located just north of the spring swimming area (Dunbar upright aluminum coring tube, powered by engine vibrations, et al. 2007). Upon closer examination of the visible portion, into the sediment to extract a soil sample. Three cores were it was tentatively determined to be a fragmented portion of a extracted this way from an extended perimeter around the bone proboscidean's maxillary alveolus, or tooth socket of the upper fragment; each being approximately 6-8 m to the east, west, dentition, from which the tusk was encased and once emerged and north (a lithological analysis of the cores is forthcoming). (tusk absent; Figures 2 and 3). Orientation of the alveolus These established that the sediment column overlying the indicates that it rests in an inverted position and in a southwest limestone is at least 1.5 m in depth. to northeast direction, with the distal or northeastern end more In April of 2008, during the National Geographic Society deeply buried. It is unknown if this position is a result of funded project at the Wakulla Springs Lodge site, Jack Rink natural and/or cultural processes, or if the element is isolated and Kevin Burdette of McMasters University conducted from other parts of the skeleton. Many different taphonomic a Ground Penetrating Radar (GPR) scan over the Vickery processes (formation and transformation) could be involved in Mastodon from a glass bottom boat to test for subsurface its deposition, such as being scattered by naturally occurring anomalies. It was unclear whether such an approach was at animal activity (perhaps another mastodon) or human activity all feasible. However, using a south to north transect, the site such as prehistoric butchery, use of the area as a movie set in area was successfully scanned revealing a bulbous subsurface the last century, or modern vegetation removal and recreational anomaly in the site area (Figure 4). Although it is currently use. Nevertheless, additional fragments of the skull may be unknown if this anomalous mass is anything more than the present based on Park Ranger Vickery's report of feeling a large limestone substrate, the depth of the surrounding sediment crowned tooth, approximately 10-20 cm deeper, at the time column, along with the skull fragments, whose position is of initial discovery. Based on these findings, the remains are thought to be at the apex of the anomaly, is compelling enough tentatively thought to represent a single American mastodon to warrant further investigation of the area. Whether the find
VOL. 65 (1-2) THE FLORIDA ANTHROPOLOGIST MARCH-JUNE 2012
52 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
aeL 60 m
Figure 1. Location of the Vickery Mastodon in the spring basin, Wakulla Springs State Park.
Figure 2. Close up of alveolar fragment (5 cm scale arrow in background pointing north).
!,I' ML: :
PORTER WAKULLA MASTODON 53
is paleontological (natural death site) or archaeological (kill/ lateral aspect
butchery site) is also unknown, but its close proximity to the Wakulla Springs Lodge site combined with the presence of Paleoindian projectile points and knives in the spring basin and around the spring suggest a potential for human association.
Mastodon Remains in Wakulla Spring
The presence of an American mastodon skeleton in the for tusk
Wakulla Spring basin would not be unusual as many have been reported over the past 150 years. Exactly how many individuals have been recovered is unclear. The oldest known reports of mastodon remains in Wakulla Spring date to the early nineteenth century. In 1835 George King, a Philadelphia naturalist, is said to have collected interlocking tusks and a skull, but these were lost off Cape Hatteras or the Keys during transit (Revels 2002:97). Later in 1850, King and his assistant, palatal aspect
G. L. Brockanbrough, carried out one of the earliest known underwater recoveries of mastodon bones from the spring, noted in the June 18th issue of the Tallahassee Florida Sentinel Figure 3. Palatal and lateral aspects of a mastodon skull (Smith 1850). His detailed account explains the system of indicating the alveolar portion of the cranium found above grappling hooks and ropes used to recover the remains from the ground surface; image modified from Olsen 1979:4. the spring. The bones were reportedly collected from the northeastern side of the spring and were scattered over an area bones were discovered in shallow water during construction of of 9.14 by 24.38 m (30 by 80 feet) and at a depth of 9.14 to a swimming area at the spring (Revel 2002:98). Investigation 15.24 m (30 to 50 feet). In 1868, James Galbreath recovered by Gunter (1931) soon followed and included early use numerous bones from approximately 21.34 m (70 feet) from of the scaphander or sponge diver helmet. Figure 5 shows somewhere near the springhead, but their whereabouts today Gunter and his team recovering bones from the spring using are unknown (Revels 2002:97). In 1895, John L. Thomas this early underwater breathing technique. This recovery
collected mastodon bones from the spring, which were operation yielded an almost complete mastodon skeleton,
displayed in the Walker Library of Tallahassee and later in the which for some time, was housed at FGS (Gunter 1931), but shop window of T.B. Byrd until they were deposited in FGS eventually was displayed in the Florida Museum of History storage (Revels 2002:97). in Tallahassee, where it is currently housed (Revels 2002:97;
In 1902, A.C. Thomas reported numerous mastodon Rupert and Spencer 1988:13).
bones three miles downstream from the spring (Gunter 1931), After Ed Ball purchased Wakulla Springs and surrounding but none was known to have been collected. In the 1920s 1618.8 ha (4000 acres) in 1934 no further underwater real estate developer George T. Christie purchased Wakulla recoveries were reported to have occurred until the mid-1950s Springs (Revels 2002:37). Later in the 1930s, he coordinated (Revels 2002:41). Usage of the spring during this period was the first major underwater recovery of mastodon remains after primarily recreational, including use as a set for cinema, both
ater Surface i} "
Figure 4. GPR scan of the site area showing bulbous anomaly in the sediment column (bottom)
and a close up of the anomaly (top).
54 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
Figure 5. Herman Gunter and team recovering mastodon bones from the spring in 1931.
(Courtesy of Florida State Archives: Florida Memory Project)
above and below the surface. At the time of one particular Deep Cave Diving Team, indicates the differentloainad filming in 1956, Andy Harrold, an editor of the Florida State depths at which the remains of mastodons and othel rhsoi News Bureau, and several undergraduates from FSU (Gary animals were observed.
Salsman, Wallace Jenkins, Henry Doll, Gordon Whitney, and
Lamar Trott) made over 100 scuba descents into the spring Recommendations for Future Researc
basin and cave (Burgess 1999:86; Jenkins 1957). Although the dives were as much about early scuba use as exploration of The different megafauna observed in Wakull pin n
the spring and cave, many faunal remains were observed and the adjacent upland area is a good indication that arun, 2,0 recovered. The divers used pillowcases lined with plastic bags years ago, the area was used by late Pleistocene anml.[hi inflated by air from their scuba tanks to lift the heavy bones to remains, along with Paleoindian tools observed esweei the urfce rom epts a muc as60 (19.8 eet (Buges th sprng nd ave ystm, ave elpd t advncetherie
Buea o Golgyiniigu e d ferauna runean of reern mastodon eor es asop oset thoe thpromulate atual9frce
groudsothanCderts wel Foias chared woodves andovr eingy Presonsbefoeoemoeetdonitttecv
(199:11 tondentif the different skeletal oremainsaindicated DthatCthereiisnmuchamorentocbeelearned.fTorbettercundersandh thatin the assemblagey aolontanedimamoth, tapriSat nature and siiicae ofmanot onl heVckrmastodonsndandreitoi armdilo cme, hoseerand bin.gurae6ssf o ne ofU ther akula Spring obutlaendboaevtuisefdaeini. uasnderg aae mmerks Hexamining, otho mammoth, and an mefuaintesthserUiedStsuue Lmrot mstodonr teeth.ba researchersthe might R condethe followinrecomendatons
a fedcaes lteres the"6 Staenokoida's5 plhurchs of Dttasluewehosilaymsodnbnsfo
Wakull Srins opene aot dor toub ne sii exploration. Wakulla Sifrnto mestablsfan thseyrverl in timell wpith the
itnlloed aifmtheiS Deep a Caes Divig thea y onetteisanwt Paleoindian peid;ateo50yas oevey weslsedor n and murae som oftsa uha 0m 168fe)(ugs the Waulspringan cave system,(evl have anylrecie agoranes.asses thouglayseies
2002:102;. Romepert a nd ncer 91; Stn 1989fe).bc Figuetreoveed fpromg thve sprin fore butenpriy evdence.rinvtetory
7,Baedu on diesloby, thendiferdunae frmn f adtheoUS theoclralsores ad analyzseweposb, thet rocultntural re
PORTER WAKULLA MASTODON 55
Figure 6. Undergraduate diver from Florida State University examining both mammoth (left) and mastodon (right) teeth in 1956. (Courtesy of Florida State Archives: Florida Memory Project)
and paleontological associations of any bone and ivory tools meeting of the Society of American Archaeology Conference recovered from the spring and surrounding uplands. is to justify further investigation of the Vickery Mastodon and
to expand our understanding of the relationship of Pleistocene Conclusion megafauna and Paleoindians at Wakulla Spring and perhaps
the wider north Florida region. We are still left with a series of The primary drive behind writing this article, while in questions: is more of the mastodon skeleton present beneath part for a symposium on Wakulla Spring at the 2008 annual the spring floor; what is the nature of the deposition and the
N~O RT4W EST$OJlA$
S rn PolPlatform... .... .....
10 ~Exposed Exposd /* %
- Coral Reef Cora
:.; ...... f : rGrand \
150 :" "'[Canyon ..
-2 0 A iztdo/gb2eC Mammoth' Jaw Bone ___..____"_... .. ..
2 D Mastodon Jaw Bone Area of Fossil Finds J Large Scatterd .. =
250' E Charred Wood Sharp turn Limetone Coral
F Large Tusk to Southwest Boulders Fragments
H Leg Bone
FEET 0 100 200 300 400 500 600 700 ao0 900 00 1100
Figure 7. Wakulla Springs profile illustrating locations of megafauna remains; image modified from Olsen 1958; Rupert and Spencer 1988: 15.
56 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
extent of disturbance; was it a natural death or did humans kill this animal; does evidence of butchery remain to be analyzed, and if so is it clearly related to a kill or scavenging episode? Jones, B. Calvin, and Louis D. Tesar Based on sediment depth, the presence of skeletal fragments, 2000 The Wakulla Springs Lodge Site (8WA329): a and subsurface anomalies, the site may contain additional Preliminary Report on a Stratified Paleoindian
skeletal materials that can help answer these questions. Given through Archaic Site, Wakulla County, Florida. The the existing skeletal remains' close proximity to the upland Florida Anthropologist 2:98-114. Paleoindian site 8WA329 and the recovery of Paleoindian tools in the surrounding area (Jones and Tesar 2000), the Olsen, Stanley J. Vickery Mastodon could contain considerable archaeological 1958 The Wakulla Cave: Amateur Paleontologists and paleontological content. Investigate an Underwater Site in Florida. Natural
Acknowledgements 1979 Osteology for the Archaeologist. Papers of the
Peabody Museum of Archaeology and Ethnology v. There were many individuals involved with this site's 56 no. 3. Harvard University, Cambridge, Ma.
investigation, without which this effort would not have been possible; in particular, a debt of gratitude is due to Jim Dunbar, Revels, Tracy J. Mary Glowacki, Dave Thulman, Andy Hemmings, and the 2002 Watery Eden: A History of Wakulla Springs. Sentry
internal and external reviewers whose suggestions help bring Press, Tallahassee. this to publication. Also, thank you to those contributing significantly to the site investigation: Jason Vickery, Bonnie Rich, Lou Allen, Sandy Cook, Brian Fugate, and Phil Gerrell (Wakulla 1964 Wakulla Spring: Its Setting and Literary Visitors. The Springs State Park-Florida Park Service); Jim Dunbar, Debra Florida Historical Quarterly 42:352-363
Sheffi, Dan McClarnon, Roger Smith, and Mary Glowacki (Bureau of Archaeological Research-Florida Division of Rupert, Frank Historical Resources); Tom Scott and Harley Means (Florida 1991 The Wakulla Springs Mastodon. Florida Geological Survey), Joe Donahue and Mike Lavender (Florida Paleontological Newsletter 8:10-12.
State University); Jack Rink and Kevin Burdette (McMaster University); and Ed Green (Panhandle Archaeological Society Rupert, Frank, and S. Spencer at Tallahassee), all of whom helped to make this project and 1988 Geology of Wakulla County, Florida. Bulletin No. 60. research possible. Finally, I would like thank my family for Florida Geological Survey, Tallahassee. their continued and unconditional support.
References Cited 1850 The Bones of a Supposed Mastodon. The Florida
Sentinel, Tallahassee, Florida, June 18.
Burgess, Robert, F.
1999 The Cave Divers. Aqua Quest Publications, Inc., Stone, W. C.
New York. 1989 The Wakulla Springs Project. U.S. Deep Caving
Dunbar, James S., Kevin M. Porter, and Debra Sheffi 2007 Inspection of the Vickery Mastodon Site in Wakulla Thulman, David
Springs State Park. Report on file, Division of 2009 Freshwater Availability as the Constraining FactorinHistorical Resources, Tallahassee. the Middle Paleoindian Occupation of North-Central
Florida. Geoarchaeology 24:243-276.
1931 The Mastodon from Wakulla Springs Wakulla Upchurch, S.B., and A.F. Randazzo
County Florida. Florida Woods and Waters, 1997 Environmental Geology of Florida. In The Geology Florida Department of Game and Freshwater Fish of Florida, edited by A.F. Randazzo and D.S. Jones,
(Spring): 14-16. pp. 2 17-250. University Press of Florida, Gainesville.
Jenkins, Wallace T.
1957 The Exploration of Wakulla Springs Cave. Field
Notes Related to Each Cave Dive in 1957, with a 1962 addendum. Report on file, Bureau of Archaeological
WAKULLA SPRINGS ARCHAEOLOGICAL RESEARCH: THE VOLUNTEER EXPERIENCE
MARY STERNER LAWSON
Retired Professor, Department of English, Albany State University email. MaryLll@aol.com
For the armchair dreamer whose images of participation Responses came quickly from a variety of potential in an archaeological dig came straight from exotic movies, volunteers. A few who signed on for the project had volunteered fascinating articles in the Smithsonian, or National Geographic at other digs and generally knew what to expect. Among those, shows on High Definition television, the ultimate experience some had had worked at the Paige-Ladson site on the Aucilla would be the opportunity to be part of an archaeological River in Florida. Madeleine Carr's work on the Windover site exploration. In April of2008, the dream had a chance to become "came in handy" in her role at Wakulla. She knew what was a reality for a number of these armchair dreamers when they needed to ensure the project ran smoothly, and who to call on found out they could apply to join forces with other volunteers for assistance. and professionals at the Wakulla Springs archaeological
research project. As one volunteer put it, "An archaeological
dig was something from an Indiana Jones movie for most of Two months before the project was to begin, I us, experienced only vicariously via the magic of film. But coordinated with the Friends of Wakulla to rent a suddenly, it was right there in front of us at Wakulla Springs. construction trailer for the project. Situating the And best of all, we volunteers had been invited to be a part of trailer near the project yet out of site of the regular it. I was ecstatic!" park visitors was coordinated with park management.
This indeed would be an extraordinary opportunity, Brian Fugate, the park manager, and Lee Pyles,
for we would be connected with National Geographic, the maintenance supervisor, both offered material and
organization funding this exploration into the extent of labor to erect barriers. Aesthetically, these fit in with
Paleoindian presence at Wakulla Springs. And as Constance the park and allowed for close interaction between Clineman put it, we would be "'getting dirty together" in the park visitors and volunteer interpreters. pits alongside "celebrities from the world of archaeology,"
including archaeologists as well as anthropologists from the
Bureau of Archaeological Research, McMaster University, Carr also worked out other details including coordinating University of Texas, and the Smithsonian. And while most acquisition and storage of tarpaulins, and the clean-up of a of us were only familiar with radiocarbon dating, we learned small room in an old bathhouse where shovels, wheelbarrows, we would have an opportunity to see a specialist employ transits, and screens were stored nightly and on weekends. an intriguing new research procedure, optically stimulated Volunteers like Charles Conway, who often arrived at 8 luminescence. Then, to top it off, at some point National am to help set up the site, found the process easier because Geographic television photographers would be on hand to of Carr's organized system. He comments, "I was there at 8 film a part of the exploration for its TV show, Wild Chronicles each morning and was nearly always very useful in opening (Figure 1). the gates, de-tarping, marshalling sawhorses, screens,
To obtain a sufficient number of capable volunteers, word wheelbarrows, shovels, trowels, tapes, transit, rods, etc. for needed to be distributed widely. This task fell to interpreter and the day." Conway had volunteered on sites before, knew what historian Madeleine Carr, a founding member of The Friends to do, and was eager to get to work. of Wakulla Springs State Park who "agreed to coordinate the We neophytes soon would be learning site procedures, so human resources that would be needed" for the project. Her our inexperience was no drawback, but a few of us related experience as site administrator and media coordinator on the more to the experience of Wakulla dig volunteer John Roberts,
1 980s Windover archaeological site near Titusville, Florida
enabled her to utilize an efficient approach in organizing the
Wakulla project, where she served as field office and volunteer My grandmother subscribed to National Geographic coordinator. Working initially through the Wakulla Friends magazine when I was a kid, and, of course, I looked group, Carr circulated word of the project to the Green Guides at all the pictures and would sometimes notice and members of the local historical society, then sent out "an people digging a hole with a teaspoon and toothbrush e-mail 'blast' and 'call to action' in local media." somewhere in some far off place that I would probably
VOL. 65 (1-2) THE FLORIDA ANTHROPOLOGIST MARCH-JUNE 2012
58 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
Figure 1: National Geographic film crew.
never go to, but I thought if I ever got the chance to would additionally make our photos and recordings available
do something like that, I would. The Wakulla Springs to our project leader and to the National Geographic Society.
dig was my chance. We also signed an agreement to comply with Department
of State policies, a Bureau of Archaeology Volunteer Form,
A core group of about fifteen volunteers, myself included, also and a form indicating compliance with non-discrimination and jumped at the chance. sexual harassment policies. Moreover, the application process
My own interest in becoming a part of a dig came through included a background check, which some of us had not two sources. One involved a visit to Illinois in the early 1970s expected, though seasoned volunteers spoke of its helping to to the Koster site, which dates to about 8500 B.P. At that insure that no people on the site had been convicted of looting. time, the vast site was abuzz with university and professional It may seem a considerable amount of paperwork, but it no researchers, meticulously working in squared-off sections. As doubt served to impress in our minds that we were going to I watched, a college student down in the pits uncovered a large be part of a very serious enterprise. Eagerness to participate grinding stone. To this day I recall the vicarious thrill as she was not the sole decisive factor: clearly we had to be legally fixedly and admiringly gazed at her treasured find. As it has and ethically appropriate, and we had to understand what was done for other site visitors, my first-hand opportunity to observe involved in the project. an active exploration whetted my appetite for archaeology, What really helped in the initial stages was information and, in my case, led to my reading books and articles about the we received from our volunteer coordinator and subsequent decades of research Louis and Mary Leakey spent in Kenya. attendance at a volunteer orientation session. Madeleine Gaff s Whatever the spur to our interest--books, television, movies, helpful "read me first" document provided us with tentative or first-hand observation--the Wakulla opportunity brought us information about participation, emphasized the importance of volunteers a chance to participate locally, without a lengthy the Project Member Agreement, explained the tasks involved plane flight to a remote, torrid site. and types of teams needed, and included a project calendar for
We did have some preliminaries, however. Those of us the April 7 to May 4 endeavor. We learned in the letter as well new to volunteering at a dig were a little surprised by all the as through conversations with Madeleine that we could sign preparatory steps involved in the application process. We filled up to work as excavators, screeners, docents, photographers, in a detailed data form asking for background information site illustrators, or data entry volunteers. Eventually, as the about previous volunteer work and particular aptitudes, and need or interest arose, it turned out that quite a number of us signed a Member Agreement for the Wakulla Springs Lodge. would work in more than one capacity during the various fourNot that we would have thought of putting a video of the hour-long shifts at the site. One vital factor that we learned project on YouTube, but our signing the Wakulla project was that "no matter what our abilities or capabilities" we could agreement meant we would not publicize details of the dig and participate. As Carr commented, "Above all, no one involved
LAWSON THE VOLUNTEER EXPERIENCE AT WAKULLA SPRINGS 59
with an archaeological volunteer opportunity should overlook surface or with a trowel...with pit floors and walls that have that there are plenty of 'jobs' for those who cannot dig or carry to be preserved perfectly straight." The pit work did take heavy weights." flexibility and a level of strength, once buckets of earth needed
We learned more about our on-site opportunities at to be hefted higher and higher up to ground level. However,
a volunteer orientation led by Louis Tesar of the Florida there was never a problem finding ways to be of assistance. A Bureau of Archaeological Research. At the April 3rd meeting, couple of volunteers who discovered incipient back or other held at the B. Calvin Jones Center for Archaeology, we problems quickly shifted to different tasks, like screening, received additional substantive information, watched a visual labeling artifact bags, and documenting findings (Figure 2). presentation, introduced ourselves to others who had signed My own discovery of a limitation involved attempts to up for the project, and were briefed by Jim Dunbar. Meeting sketch the findings. Desirous of doing lithic illustration of professionals and other participants helped us to get a feel for some of the items found, I soon discovered the challenges the cast of characters we would see on a regular basis. This of that technique. Jim Dunbar kindly cued me in on the issue meeting, following our absorbing much of the written material of drawing precisely to scale when he saw my early efforts, and hearing the discussion of the project, caused us to realize and made some other suggestions, and a more-experienced more fully, as volunteer Larry Benson did, that "working on volunteer loaned me some books on lithic drawings. A few the dig was certainly an honor and privilege, not to mention days later, after some practice at home, I shared an attempt the archaeological experience of a lifetime." of mine with the University of Texas lithic pro. He glared at
From the start, Wakulla Springs welcomed us; the my attempt, then soberly pointed out a well-done illustration
enterprise was volunteer-friendly. We merely drove up to the on the t-shirt of a fellow worker, noting that it took him six entrance, greeted the genial gatekeeper, stated our project months to teach her to draw that well. Still, I kept at it, and status, and proceeded on through the main visitor entrance to enjoyed the effort, but felt far more at home doing sketches of the lodge parking area. Volunteers paid no entrance fee. Food, people at work or at rest. drinks, and restroom facilities were available in the lodge At times work was more restfully slow-paced, but adjacent to the work site. These handy site conditions were any web information on archaeological efforts stresses the rather posh compared to conditions at other digs, according to need for patience, which we learned was an important part Madeleine Carr. Of Wakulla, she said of being painstakingly accurate. When there was a lull,
some of us gathered around colleagues at another pit who
were actively engaged in measuring or perhaps discovering
Certainly the location in a state park helped. There something. "With finding a 'piece' and deciding if it needed was electricity. There was park material available to to be measured and plotted individually, work would stop,"
erect a pleasant wood fence to keep the public away Conaway noted. At other times work slowed when "finishing a from danger. There was a citizens support group [the set of squares," or when "the comers had to be measured." The Friends of Wakulla] with hundreds of members and a pace simply was a part of the learning process. And as Angret
willingness to pay to rent a construction trailer where Piasecki commented, "It was so interesting to learn about the specimens could be cataloged. methods employed in conducting a dig; i.e., the measuring,
careful handling, and logging. Everything was so carefully
Partial pit area coverage from the sun and rain was provided and meticulously done." not long into the effort when two large tents were erected. Speculation and stories about Wakulla and other This and other assistance from The Friends of Wakulla and the archaeological sites were the regular topics of discussion as we employees at the springs all supported the project enormously. dug, hefted dirt to screens, pushed wheelbarrows, documented
Our working area was set up close to the area where finds, and screened. As we chatted, we worked and learned, as Calvin Jones found the tantalizingly inspirational Paleoindian Constance Clineman notes stone blade and Clovis spear points. Actually, seeing the story
of Calvin and his astonishing find on a display placard beside
our excavation pits served as motivation. Would we have the "There were no formalities. And every 'expert' luck that he did? How much more about the age of the site was eager to gently teach the volunteers how to do
would be discovered during this project? the many curious tasks associated with the effort,
We had other questions as well, as Constance Clineman continually nurturing the novices. The work was
did when she wondered, before we began working, "Could I meticulous... Every volunteer grew in both curiosity
lift a bucket of dirt that weighed 50 pounds? Could I push and confidence with each new day on the project. It
a wheelbarrow heavy with soil?" And sometimes we did was a personal epiphany of sorts for all of us. We
learn something about the limitations of our abilities. As learned to use flat-bottomed shovels to gingerly
Charles Conaway commented of the pit work, "it was not remove layer upon layer of soil, ten centimeters. We
very physically challenging, but enough to learn what I'd measured angles, took GPS coordinates, got down
suspected--I'm not configurationally optimized to do gross on our knees in holes that were deeper than we were
or fine work with a long-handled shovel very far below the tall. We took turns sifting through each bucket of
60 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
Figure 2: Jim Dunbar shovels as volunteers screen.
dirt, using our eyes and fingertips to sort through questions if curious, was instrumental in our taking pleasure lithic material, rocks, fulgurites, calcited [sic] bone, in the project and in our encouraging others to take part in artifacts (both modem ancient!). We reeled with similar programs and/or be in support of public archaeological
excitement when we discovered a tiny bead, a point, programs. As Angret Piasecki commented, "What I most a pottery shard. We were an honest-to-goodness enjoyed was listening to Jim [Dunbar] and some of the other
'team' struggling to discover pieces of a puzzle that more knowledgeable participants talking about the findings would reveal our past, link us to those who had stood and other previous digs they were on." Often, she noted, on the very same ground thousands of years before there was "talk about what the area looked like back then and
us. It was a sacred experience. how people lived," something that could only be discovered
through backing and support for public projects of this type
A volunteer so profoundly influenced--as Constance was This support has an additional chance of occurring if the
becomes a spokesperson for the field of archaeology and for public has a greater awareness of archaeological research like personal involvement as a volunteer, the Wakulla project, and there was no problem in obtaining
At least two of us, myself and Constance, had an additional public interest at Wakulla Springs! Whatever we happened opportunity to learn more about archaeology from Louis Tesar to be doing was of interest to the passersby and provided as we worked with him, first on artifact documentation in the handy subject matter for the docents or guides. Some of the on-site trailer, and then, later, at the B. Calvin Jones Center volunteers were certified as Florida Green Guides, like MarieFor Archaeology. Cline comments about being "in the mobile Anne Luber, who "acted as a docent, showing visitors the lab, counting, bagging, and labeling each set of materials! model preforms and artifacts, explaining how the Indians may artifacts uncovered. I found this to be fascinating, no, almost have lived during that time, what types of tools they used, intoxicating. It was there in the silence of that mobile lab that what kinds of animals may have lived at that particular water I realized I was 'hooked' on archaeology. Me...an old lady with hole." a Dramatic Arts degree. Imagine that." And, as Madeleine Carr notes, during the weeks we were
Imagine, too, being able to accelerate our learning process working at the site "there would be opportunities for public by listening in on accounts of the experienced archaeologists interaction at almost every step," for there are thousands of and paleontologists on the site, including Jim Dunbar, Louis visitors who enjoy Wakulla Springs each month. Tesar, Harley Means, Mary Glowacki, Michael Collins, Jack
Rink, Pegi Jodry and Dennis Stanford (from the Smithsonian),
Andrew Hemmings, David Thulman, and others. To her, this meant that public interpretation
The chance to be the mouse in the corner, who could ask opportunities presented themselves hourly. Small
LAWSON THE VOLUNTEER EXPERIENCE AT WAKULLA SPRINGS 61
Figure 3: George Weymouth takes it down a level.
artifact cases were displayed near the site and visitors display. It was such a joy to see how both the experts from around the world learned about Florida's archaic and the volunteers joined together in an educational
past. Artificial mastodon bones helped them visualize mission to pass on their knowledge, their many gifts.
the size of megafauna while docents answered It spoke to the future. As the days passed, countless
questions from 9 a.m. to 5 or 6 p.m. each day. tiny elementary school groups stood along the ropes
lining the dig area, eyes wide, watching silence. You
could almost 'hear' their minds softy repeating, 'I'm
Madeleine observed, "'the location of the excavation adjacent gonna do this when I grow up.' to a lodge and public footpath to a waterfront ticket office
made it imperative" to be prepared for "public contact and Clearly it was an excellent time for public education on many interaction." Because of the nature of the place, she notes, levels. "Wakulla Springs lends itself well to simply 'hanging out' on Madeleine and the other docents, as well as those of us balmy spring days, eating ice creams next to the former home who stepped in when the occasion arose, were sometimes kept of mastodons, and answering questions about those holes in quite busy answering the visitors' eager questions. What are the ground." During the 290 hours she was on the site, she you doing? What do you hope to find? Carr, who helped prepare regularly answered questions from the many park visitors the guides at the site to respond, reflects, "Those who wanted whose curiosity was aroused by all the unusual activity they to act as docents learned about the time frame involved, and were observing. how to make clear to people who believed the dig was about
The curiosity of the visitors often was infectious, dinosaurs (always a challenge but fun to overcome) that it was Clineman tells of her observations, not." Among those asking us questions were bikers, hikers,
families, groups of children, and visitors from several different
countries, including Germany, Albania, Denmark, Poland, and
Guests at the Park stopped and stared. They Switzerland.
wanted to 'look' into the deep pits. They fed on our The visitors' anticipation matched ours, for the heart of
enthusiasm, became 'part' of the project. They were the matter for us all was what we might find next. Many of 'vicariously' living the event by watching us work.., the volunteers logged an astounding number of work hours in just like watching those Indiana Jones movies! But anticipation of a find: six volunteers spent over 150 hours at this time, it was 'live." They could ask questions, the site, and one totaled up 366. Most of the time we worked in talk to notable archaeologists, view the artifacts, pairs, removing soil to 60-70 cm below surface, then changing read the many interpretive panels that surrounded to shallower 10 cm layers until it was time for trowel work.
the dig area, touch the huge replicas of bones on At first, the more seasoned ones did the fine-tuning work,
62 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
but others of us were offered the chance so that all of us who /
wanted to do so were able to experience various stages of the excavation. Learning how to determine the precise elevation of found objects involved a learning curve, but being an assistant in that process was not difficult. -iii
As we delved deeper in the pits, our conversation < '"
sometimes delved more deeply into one of the more shadowy A
causes for concern in archaeology. We neophytes spoke of the meticulous steps involved in documenting a site, which \" occasionally led to conversations about sites we knew of that were not documented, or projectile-point-predators, wholly /
indifferent to site contexts, who persistently hunted artifacts. J f \"
One former neighbor of mine, a surveyor, bragged that -. .
he had found a Native American village site but would never reveal its location. He regularly ravaged it, gluing his stunning finds, his dispossessed artifacts on plywood! He is gone now, as is the history of the village. Surely our working at Wakulla would encourage all of us to think in terms of the ->
public good of archaeological documentation of places and things and warrant against our being like my neighbor, who was so indifferent to documented archaeological research of our past.
And, as well, visitors seeing our efforts to be precise about
what may have occurred at various levels, and hearing us speak of the importance of reverencing our past, could be caused to think of the importance, of judiciously caring for our land and L ids:Af its history. Judging from actions of the volunteers after the dig .
closed down, our experience certainly affected us. Many of us either found additional sites at which to volunteer, are doing Figure 4: John Roberts and the Big Toenail. further reading and research on archaeology, or are hoping for another nearby volunteer option in the near future. Through anticipation. It is rather like the difference between those who our work at Wakulla, our existing awareness of artifact context like to go fishing whether or not they catch a fish-and those and site documentation was significantly reinforced, who must catch a fish or be disgruntled. The anticipatory types
Of the effort to be precise in this documentation, Carr reigned on the site, or else the more disgruntled variety kept noted, nothing is left to chance. There are thousands of forms their counsel. The truth is that one never knew what might to be filled in, and volunteers used Jim Dunbar's method to fill be found, and that, for us, provided the anticipatory adrenalin in forms and copy more forms in the lodge office. Meticulous- that sustained us through hours of digging and troweling. And minded types made sure the forms corresponded to the even if we did not find something that day--perhaps tomorrow
specimen bags. If one bag did not match a form there was would produce the special find. always Louis Tesar who would spot the error. We knew we Volunteer John Roberts (Figure 4), who spent 224 hours
were in trouble when he left the trailer across the lawn with a at the site, captures well what it means to be a part of the certain quizzical look on his face. We certainly had a quizzical discovery process. look on our faces at times, when we chanced upon a new type of object in the pit or on the screen, for we could not determine Probably the most exciting thing that happened the significance of a given small object--like the minute bead to me was the day I dug up what I called 'the big that was found at one point. toenail.' That's what it looked like anyway. Ed Green,
We carefully bagged and labeled it and many a minute another amateur archeologist, kept telling me that he
bit of material, as well as some odd things like bottle caps listened for the 'chink,' the sound of the shovel hitting that were signs of non-virgin soil. Charles Conaway notes or scraping across something other than dirt. I was some of his findings in one pit: "fulgurites, diagnostic pottery down in one of the pits scraping dirt off the bottom shards, and lots of chips and a few chunks of chert, some with when I heard the 'chink.' I sorta knew I had something evidence of having been worked, exposed to heat, and many different because Ed perked up and was looking down smaller ones." As to the value of these findings, that was for in the hole to see what it was I had uncovered and the experts to determine after the site closed down, when the he called Jim Dunbar over. It was our turn to have analysis could begin in earnest. everyone come see what we had, and I was looking
There are people who must find something of great up at all these archeologists. Up until then, every time
value, and there are those who thrive equally as handily on something significant had been found, it had been a
LAWSON THE VOLUNTEER EXPERIENCE AT WAKULLA SPRINGS 63
seasoned archeologist who happened to be in the pit because its meaning has many elements for each of us, as one digging. I thought surely they didn't want someone can see in Larry Benson's response to his Wakulla experience
who measured his archeology experience in days to actually finish unearthing this thing. How wrong I
was! After we measured depth and distance from the I worked 3 days a week during the entire dig and
walls of the pit, they handed me a broom and a trowel enjoyed every minute of it. Shoveling our Ice
and told me to be careful. It was later identified as a Age history from the earth, working shoulder-toscraper. So I was the first person in a very long time shoulder with professionals, and learning some of
to hold the scraper in my hand. The person before me the techniques of archaeology was educational as
was most likely using it to scrape some animal hide, well as rewarding. In my opinion, volunteers are
several thousand years ago. an important asset which the BAR is calling upon
and I am pleased to see that we are being used and
Even if our finds were nowhere near as dramatic as a hide appreciated. There was a day when volunteers were scraper or projectile point, we felt as Angret Piasecki did, "I viewed as "pot hunters" and were not welcome will never forget the anticipation we felt each time another anywhere around digs. I am pleased to see this is not shovel hit the screens. We touched objects that we know the climate now and, in fact, I think we are viewed as
were handled by someone over 10,000 years ago-really "partners" rather than "potters." It really felt good to
awesome!" This sense of our connection to the past, which we be a partner.... I am ready to sign up for future digs, shared on the site with the professionals, other volunteers, and and I would encourage others who have an interest with the many visitors, was an extremely important part of the in history, archaeology and the preservation and volunteer experience, conservation of our heritage to sign up and become a
When volunteers were asked to summarize how they felt volunteer participant with Bureau of Archaeological
about the Wakulla experience, their replies ran along similar Research. lines. Frequently the responses were variations of MarieAnne Luber's comment that she "loved every second of it." Larry Benson's 56 hours on the site were well spent, and his However, because Wakulla Springs is highly valued as a involvement in the future is assured.
regional treasure, the dig had an additional allure for locals, as At 88 volunteer hours (and counting), volunteer Constance Clineman observed, Constance Clineman is already back at work, and related that
she "currently is helping with an experiment that will help
... for those of us who live in the area, we felt an substantiate the dating of artifacts found at the Wakulla site."
additional sense of identification, for this particular Her appraisal of her own personal experience captures the project explored the deep history of a place dear to spirit of what the Wakulla Springs archaeological effort meant
many of us for its unparalleled natural beauty. That to all of the volunteers.
we could volunteer to be a part of it, to feel that we
were helping on a very important archaeological I have worked many years as a Park Ranger, helping to
project was an extraordinary thing. preserve and protect our natural and cultural heritage
here in the United States. I know that volunteerism
is essential to the success of any conservation or
The context of her response reminds us that both archaeologists preservation attempt. It not only provides much and the aims of the profession are well served when regional needed additional "muscle" for each project, but resources and local residents are involved in the investigative it also is the perfect vehicle for spreading public process. Additional beneficial links include parents, schools, awareness of the critical need to preserve our past in teachers, and grassroots organizations, whose involvement order to protect our future. Volunteerism is the best
with and visitation to a site could lead to a greater way to ensure that future generations will have the
understanding of the need for personal and legislative support opportunity to enjoy the same gifts from the world for archaeological funding, around us that we have been blessed to experience
The volunteers, though, are a key to this process, as they in our own lifetimes. Volunteerism teaches selfless demonstrate their eagerness, become involved in the procedure, giving by example. Volunteerism brings joy to the operate as docents, explain to visitors, and ultimately carry on volunteer. Volunteerism is uniquely American. And, their knowledge in other venues and avenues. Additionally, in the case of the Wakulla dig, volunteerism was the
greater understanding arising from volunteers' personal underlying "glue" that filled in the gaps so that the
experience and site visitors' keen observation can lead directly project could be undertaken. When someone asks me to an increased awareness of the need to safeguard, preserve, if I am a volunteer, I smile. I take pride in being a part and document artifacts and sites. of such a wonderful group of people.
In the final analysis, it could be said simply that the
Wakulla Springs archaeological exploration of 2008 meant
much to many. It is impossible to sum up in a few words
THE FLORIDA ANTHROPOLOGIST FUND
An Endowment to
Support production of
The Flora Anth ropolgt,
the scholarly joumal published quarry by
The Florida Anthropological Society since 1948.
onations are now being accepted from
individuals, corporations, and folbndations.
inquirS and gifts may be directed to:
The Florida An thrp st
O. Box 6356
Taflabassee, FL 32314
The Flirda AntloIl Siety IS i a nonprofit org action
under section 50(c)(3) of Inrl Revnu Coe.
Contrib are euctibe as provI by
section 170 of th codes
THE PURDY UNIFACE, A RECENTLY RECOGNIZED EARLY STONE TOOL TYPE IN FLORIDA
SCOTT E. MITCHELL' AND L. H. "MONTY" PHARMER2
'Silver River Museum and Environmental Education Center, 1445 N. E. 58th Avenue, Ocala, Florida 34470 email: Scott.Mitchell@Marion.K12.fl. us
2 Silver River Museum and Environmental Education Center, 1445 N. E. 58th Avenue, Ocala, Florida 34470 email. email@example.com
Introduction Methods and Materials
In this article we introduce a newly identified unifacial Compiling the Sample stone tool type from Florida, which we have named the
Purdy Uniface in recognition of Dr. Barbara A. Purdy and her Using the sample of 38 Purdy Unifaces identified within significant contributions to Florida archaeology (Figures 1 and the Hendrix collection as a starting point, the authors sought 2). We provide background information about the recovery input from others familiar with prehistoric stone tools in and recognition of this implement, its physical attributes, Florida. Archaeological collections curated by the Bureau of explain the similarities and differences between this type and Archaeological Research (BAR) were researched for additional other unifaces, and discuss possible temporal affiliation and specimens. Several individuals with private collections were function. In this article we also highlight the importance of also contacted and asked to help identify possible examples of responsibly recovered and curated private artifact collections, this new tool type in their collections. These efforts yielded a such as the material amassed by Mr. Alvin Hendrix. total of 55 Purdy Unifaces from the multiple sites used in this
The Purdy Uniface was first identified as a unique stone analysis (Table 1).
tool type by Alvin Hendrix, who recovered large numbers and The sample used in this study includes Purdy Unifaces varieties of prehistoric and historic artifacts and fossils from from 7 Florida rivers, one well-documented terrestrial the rivers of north-central Florida during the 1960s and 1970s. archaeological site, one well-documented inundated riverine Unlike many river divers and collectors, Alvin made a decision site, and two surface finds. The Purdy Unifaces in this study early on to someday donate his collection for scientific and were recovered from areas of Florida between the Apalachicola educational use. In doing so he realized the importance of River and the Tampa Bay area (Figure 3). While specimens provenience, and in addition to keeping accurate records of all were recovered from the Apalachicola, Aucilla, Wacissa, of his finds, Alvin saved everything from a site including tool Suwannee, Santa Fe, Wacasassa, and Hillsborough rivers, and fragments and debitage. The true value of his assemblage of several terrestrial sites, the bulk of the sample was found in artifacts from various sites is that they do not consist primarily a two-mile stretch of the Santa Fe River (Figure 3). We do of beautiful projectile points but also include the mundane, yet not propose that this represents a complete sample; however, important, everyday tools such as the Purdy Uniface. the number of artifacts and spatial distribution of recovery
Throughout the years Alvin generously shared his locations are clearly sufficient to document the Purdy Uniface
collection and data with many professional archaeologists as a newly defined Florida tool. such as Barbara Purdy, Ripley Bullen, Jerald Milanich, Albert In addition, a random sample of 30 Hendrix Scrapers Goodyear, Andrew Hemmings, James Dunbar, and others was selected from the Hendrix collection to obtain "typical"
involved in Florida archaeology. During the early 1 990s he measurements for comparison to the Purdy Uniface. These began donating his collection to the Silver River Museum artifacts were selected from the same area of the Santa Fe River (SRM) in Ocala. To date, he has donated more than 16,000 where 37 of the Purdy Unifaces in this study were found. With items for research, education, and display. the exception of distal offset, we measured the same features
During one of Alvin's deliveries to the SRM he on both tool types (length, width, thickness, weight, and edge
mentioned that there was one type of tool in the group that angle). he could not identify. Upon closer inspection, it was clear that Distal offset angle of the Hendrix Scrapers is not included the tool was distinctly different from any other previously in Table 4. Unlike the Purdy Uniface, the offset angle of the described unifacial tool. The authors decided to document Hendrix Scraper was found to be so variable in our sample the Purdy Uniface as a new formally recognized tool and that we felt it was not easily compared. Hendrix scrapers began researching existing literature and artifact collections. were found to curve to the left and to the right, while others Examples of this uniface were found in several collections exhibited no distal offset at all. Additionally, several examples curated by both private individuals and professional institutions curve from the center of the artifact and then reverse the curve and a new tool type was born. at the tip (Figure 2B; Purdy 1981:19).
VOL. 65 (1-2)) THE FLORIDA ANTHROPOLOGIST MARCH-JUNE 2012
66 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
Figure 1. Dr. Barbara Purdy and Alvin Hendrix display a Purdy Uniface Uniface.
Analysis the tool maker was left handed as suggested in the making of
Bolen Beveled points (Milanich 1994:54).
Each Purdy Uniface in our sample was carefully In order to calculate this angle of offset, we first established
measured, weighed, sketched, and photographed. Maximum a longitudinal proximal (basal) centerline by measuring the length, width, and thickness in centimeters were recorded maximum widths of the tool at the base and at a point equaling using an electronic digital caliper (Figure 4 and Appendix). 25 percent of the total tool length (Figure 5). The distal The weight of each specimen was recorded in grams using centerline was then drawn using the same method but working an electronic digital scale (Appendix). When laid flat with the from the distal end. The distal offset angle is the difference unifacial surface down and viewed from above, the distal tip in degrees between proximal (basal) and distal centerlines. of the Purdy Uniface typically curves to the left; however, the Edge angles for both the right and left lateral margins were tip of one tool curves to the right, possibly an indication that taken at a point half the length of the artifact and recorded in
Table 1. Sites yielding Purdy Unifaces included in this study.
Location of Collection Site No. Collected Source
Santa Fe River 36 SRM Hendrix Collection
Santa Fe River 2 Collection Don Monroe
Suwannee River 1 SRM Hendrix Collection
Suwannee River 2 Collection Robert Allen
Harney Flats (8111507) 4 BAR
Serbousek/Cotrill, Aucilla River 3 BAR
Hillsborough River 2 Collection Paul Lien
Apalachicola River 1 Collection Harley/Ryan Means
Wacissa River 1 Collection Harley/Ryan Means
Hernando County 1 Collection Robert Allen
Marion Count 1 Collection Robert Allen
Wacasassa River 1 SRM Hendrix Collection
S...... .. ......... ...ii)! i~i i .i!!~ .. i ii
Figure 1. Dr. Barbara Purdy and Alvin Hendrix display a Purdy Uniface Uniface.
Analysis the tool maker was left handed as suggested in Ih aigo
Bolen Beveled points (Milanich 1994:54).
Each Purdy Uniface in our sample was carefully In order to calculate this angle of offset, we firsisalse
measured, weighed, sketched, and photographed. Maximum a longitudinal proximal (basal) centerline by resrn h length, width, and thickness in centimeters were recorded maximum widths of the tool at the base and at a pInteuln using an electronic digital caliper (Figure 4 and Appendix). 25 percent of the total tool length (Figure 5)" h dsa The weight of each specimen was recorded in grams using centerline was then drawn using the same method ul okn an electronic digital scale (Appendix). When laid flat with the from the distal end. The distal offset angle is th Ifenc unifacial surface down and viewed from above, the distal tip in degrees between proximal (basal) and distalcenelns of the Purdy Uniface typically curves to the left-- however- the Edg~e ngles, for both the right qnd left lateralmrgn wr
MITCHELL AND PHARMER THE PURDY BIFACE 67
degrees using a goniometer (Butler 1980; Movius et al. 1968) assemblages, which include temporally diagnostic tools other (Appendix and Figure 5). than the Purdy Unifaces, allows us to make educated guesses
In a cursory inspection of the 38 Purdy Unifaces in the about the age of the Purdy Uniface until more data become Hendrix Collection, Barbara Purdy (personal communication, available. 2010) found multiple layers of step fractures on the steep edge of many of the tools she examined (this inspection was during an initial meeting to discuss the tool as a new type and was not a formal analysis). Purdy noted that this pattern possibly indicates that at least some of these tools were subjected to intense pressure that produced this type of fracture.
We then selected a sub-sample of 20 Purdy Unifaces from the Hendrix Collection for more thorough and formal microscopic analysis by Rich Estabrook, Regional Director for the Florida Public Archaeology Network, Crystal River, Florida. Estabrook used a Meiji EMZ-8TR 0.7-4.5 stereozoom microscope to look for use and hafting wear.
Limitations of this Study
Research limitations of this study stem from the nature of DFsaV v ,w
artifact assemblages recovered from river bottoms. River finds A
often do not have the best provenience or temporal information because of site erosion and deflation, mixture of items from varying time periods, and movement of artifacts from their primary context (Purdy 1981:80-81). With the exception of the Purdy Unifaces recovered from Harney Flats (8-11507) and the Serbousek/Cotrill site (8TA98), none of the artifacts in this study was recovered in-situ or in a primary context.
Limitations also exist in the analysis of wear on these items. Because most of these artifacts were found in river bottoms where the tool edges can be eroded or damaged as the artifacts move, evidence of use-wear and hafting may be compromised (Purdy 1981:80-81). While private collectors showed the authors multiple Purdy Unifaces, we did not have access to their entire collections (i.e., we did not see all the temporally diagnostic items recovered in association with these Purdy Unifaces), so we are not certain of the temporal affiliation of the Suwannee River material collected by Robert Allen, material from the Hillsborough River collected by Paul Lien, the collections from the Apalachicola and Wacissa rivers... collected by Ryan and Harley Means, and two terrestrial sites in Hernando and Marion Counties collected by Robert Allen. LaC D k
It should be noted that these collections were not analyzed fully, not because of a lack of cooperation on the part of the collectors, but due to time and travel limitations experienced by the authors. In these cases, we were limited in developing ......
hypotheses about function and temporal affiliation due to limited data. Nevertheless, several lines of evidence support the inference that these tools were made during the Paleoindian and/or Early Archaic periods. Figure 2. A. Dorsal view Purdy Uniface (No.HC9443). B.
The strength of this study stems from the large overall Dorsal view of a typical Hendrix Scraper. C. Lateral view sample size and significant spatial distribution where the of Purdy Uniface (No. HC9443). D. Lateral view of the items were recovered. In addition, many temporally diagnostic same Hendrix Scraper. In the lateral views, the lateral artifacts were recovered along with the Purdy Unifaces from edges are both facing to the left. Both artifacts are in the the Santa Fe River by Alvin Hendrix, Hamey Flats, and the SRM Hendrix Collection. Serbousek/Cotrill site in the Aucilla River. Access to these
68 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
lateral margins. On each of the 20 specimens, dulled wear of the working edge, associated with use on a hard material such as wood, was found along both lateral margins. The lateral margins also exhibit differences in how they were used. On all AE ...... ...... ... !examples, a steep scraping edge (35to 800 with a median of
, ......- .....,550) exists on the left margin and a lower-angled cutting edge
Sant re Rive ,(300 to 590 with a median of 340) is found on the right margin.
On 19 of the 20 specimens examined by Estabrook for ,%,anea v edge wear, the steeper scraping edge extended three-fourths of
ath, k R a v (3)
aro "ut /,the way down the left margin from the distal end toward the I 1 \ \proximal end. The one exception had a steep scraping edge
Wummsa River that was localized near the distal tip and did not extend beyond
/ (1) that area. All of the tools also had a cutting edge that ran the
NU,,leou,Riet entire length of the right margin and in one case wrapped
TA around the entire proximal (basal) end of the tool. No evidence
of haft wear or grinding was found on the proximal or basal end of the tools.
Average dimensions of the Purdy Unifaces are a length of 8.5 cm, width of 4.7 cm, thickness of 1.6 cm, and weight of 60.7 g. There were 54 specimens with left offset and one with Figure 3. Map of Florida source areas of Purdy Unifaces a right offset (Appendix). and the number of the artifacts found in each area. Specimen weight exhibits the largest standard deviation
while all other measurements are more tightly clustered around an average (Appendix). While length, width, and thickness Results deviate little from the average, artifact weight differs by
almost half of the average (Appendix). This is an interesting The Purdy Uniface, like most other artifact types, is pattern that we believe is a product of varying densities variable in shape and dimensions so we have emphasized for the different cherts used to manufacture the individual unique features and average measurements of the tool that specimens we studied. The median and range for the various separate it from other formal types. To give a description of a typical Purdy Uniface we have complied average measurements from the 55 tools analyzed in this study. These measurements include maximum length (cm), width (cm), thickness (cm Dorsal View Lateral Vie
measured dorsal to ventral), distal offset in degrees, direction B
of distal offset, weight (g), and left and right edge angles in degrees (Appendix). The average (mean), median, range and standard deviation for each of the measurement categories are also provided (appendix).. 1!
Morphologically the tool is primarily unifacial, with a broad proximal base and a pointed distal end that usually curves to the left. We consider "unifacial" to mean that there ..... r
are few, if any, flakes removed from the ventral surface of the tool. In cases where several small flakes were struck from the ventral surface, they appear to be the result of thinning the bulb of percussion, and we consider these specimens unifacial.A Specifically, of the 55 specimens we examined, all but two (4 percent) were unifacial. The two specimens considered non- iJ .
unifacial, and in fact bifacial, were heavily worked on both the i
dorsal and ventral surfaces of the tool. IIa view
Fifty-four (98 percent) of the Purdy Unifaces we examined A-M e rm= lea I
have a pointed distal end that curves to the left (Figure 2 and c aximuenwidth
E- F Maimuethlckes Appendix). Furthermore, all the specimens included in this study exhibit a broad rounded proximal area and the lateral margin on the side that the distal end leans toward has a steeper Figure 4. Illustration of the Purdy tUniface showing how edge angle than the opposite side. maximum length, width, and thickness were measured.
All 20 of the sub-sample of Purdy Unifaces examined Proximal and distal edge views are also shown. by Richard Estabrook for use-wear exhibited wear on their
MITCHELL AND PHARMER THE PURDY BIFACE 69
measurements compiled from our sample are also presented in of cortex. Because of the existence of tabular chert with the Appendix. several cortex surfaces, and the occurrence of voids within
For the purpose of future artifact identification we have solid masses of chert that are filled with cortex-like material, selected specimen # HC9443 as the type specimen for the this percentage must be viewed with caution as an indicator Purdy Uniface (Figures 2, 4, and 5). The dimensions of this of the proximity of the stone used to make the tool (Richard specimen most closely resemble the average measurements Estabrook, personal communication, 2011). for the entire sample listed above. This specimen is curated at Sources of raw material were not established due to lack the SRM. of expertise and time constraints of the authors, however we
do feel that this is a worthwhile future endeavor. None of Discussion the specimens analyzed in this study exhibited evidence of
Manufacture thermal alteration, however dark patina from river staining
may obscure evidence of thermal alteration of the stone.
It is clear from our study that the Purdy Uniface was
typically made from a large flake that was struck from a Possible Uses core using percussion. While the area of the core would have presumably been prepared by the removal of several flakes Based on the results of our preliminary use-wear analysis,
to create a suitable striking platform, most of the flaking on it seems that variable edge angles on the two margins of the the dorsal surface of the tool appears to have been performed tool suggest a dual-use scraping and cutting tool (Figure 5). after the flake was struck from the core (Richard Estabrook The Purdy Uniface exhibits a steep scraping edge on one and Barbara Purdy, personal communications, 2011). After side and a lower-angled cutting edge on the other side. The removal, the flake was unifacially shaped to form an expedient proximal or basal area shows no evidence of heavy wear or (yet formally produced) tool (Richard Estabrook, personal grinding, a pattern that suggests the implement was not hated communication, 2011). in a handle but rather hand-held. Interestingly, once the tool
Use-wear and re-sharpening almost certainly further margins are dulled from use, there is very little evidence of altered the overall shape of the tool. However, it is interesting re-sharpening. to note that Purdy Unifaces of varying sizes exhibit the same Estabrook did find evidence of use-wear on each of the overall morphology. Two specimens in this sample exhibited tools he analyzed. Edge wear associated with use on a hard biracial flake removals; however, this is the exception and all material such as wood extends along both tool margins. other specimens are unifacial. The distal end on all specimens When viewed under magnification, the edge wear is smooth but one "lean" toward the left margin of the tool. This trait indicating that the cutting edge was used against fairly hard is a defining characteristic of the Purdy Uniface and appears material (Ahler 1971; Ballo 1975; Brink 1978). Further to have been created during the manufacture process. Of the research is needed in this area to establish patterns of wear and sample of 55, nine (16 percent) specimens exhibit the remains to infer tool function.
The presence of both a scraping and a cutting edge on
angle to the left a typical Purdy Uniface suggest that the tool may not have
X I been intended for a single "function" like a drill or burin but
Center line i 6 Center line of
of distalend 16 artifact may have been used in a more generalized activity such as
' I woodworking. Because there is an indication of wear on both
\ I lateral margins and an absence of hafting wear, we suggest that
k the Purdy Uniface was most likely a multi-purpose tool that
! was likely hand-held during use. Given the curve at the distal rpnrgedge Cuttng edge: end of a typical Purdy Uniface, use for stripping or shaping
ageakn%-aen etopoieo shafts of wood (much the same as a large spokeshave) is but
2 lef deedgeang .one example of a task that could have been completed with the
Kd~ c -:} implement.
t' ,o onaoc, l.h Temporal Associations
.... .............. ........ J asat end
A x Our research suggests that the Purdy Uniface most likely cenrhooo dates to the Paleoindian and/or Early Archaic periods in
A-B C-DFlorida. This tool type has been recovered in association with l temporally diagnostic tool forms dating to both the Paleoindian
and Farly Archaic periods. These include biracial projectile Figure 5. Illustration of the Purdy Uniface showing how points or knives, adzes, scrapers, knives, and the "dimple" distal offset and edge angle were calculated. The scraping or "bola" stone (Table 2). A summary of the percentages edge and cutting edge are also annotated. of temporally diagnostic forms recovered at the two-mile
section of the Santa Fe River where the largest number (38
70 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
Table 2. Temporal association of type artifacts from the Santa Fe River Hendrix Collection
Type Period Count Percentage of Total
Clovis Paleo 14 1.03%
Suwannee Paleo 48 3.56%
Simpson Paleo 36 2.67%
Simpson Mustache Paleo 2 .14%
Misc. Paleo Basal Frags Paleo 321 23.83%
Misc. Paleo Paleo 4 .29%
Dalton Paleo 1 .07%
Greenbriar Paleo 66 4.89%
Hardaway Side Notched Paleo 2 .14%
Hendrix Scraper Early Archaic 141 10.46%
Edgefield Scraper Early Archaic 33 2.44%
Waller Knife Early Archaic 18 1.33%
Dalton Adze Early Archaic 20 1.48%
Aucilla Adze Early Archaic 26 1.93%
Bola Stone Early Archaic 10 .74%
Bolen (all) Early Archaic 212 15.73%
Stanfield Early Archaic 1 .07%
Kirk (stemmed) Early Archaic 59 4.38%
Gilchrist Early Archaic 4 .29%
Union Side Notched Early Archaic 4 .29%
Arredondo Mid Archaic 1 .07%
FAS Mid Archaic 181 13.43%
Hardee Beveled Mid Archaic 2 .14%
Thonotosassa Mid Archaic 4 .29%
Wacissa Mid Archaic 12 .89%
Clay Late Archaic 4 .29%
Hamilton Late Archaic 23 1.70%
Lafayette Late Archaic 9 .66%
Levy Late Archaic 12 .89%
Putnam Late Archaic 3 .22%
Savannah River Late Archaic 3 .22%
Citrus Post Archaic 4 .29%
Culbreath Post Archaic 1 .07%
Hernando Post Archaic 8 .59%
Santa Fe Post Archaic 10 .74%o
Tallahassee Post Archaic 3 .22%
Columbia Post Archaic 7 .51%
Duval Post Archaic 2 .14%
Bradford Post Archaic 1 .07%
Jackson/Duval Post Archaic 5 .37%
Ocala Post Archaic 1 .07%
Taylor Post Archaic 1 .07%
Sarasota Post Archaic 2 .14%
Weeden Island Post Archaic 3 .22%
Pinellas Post Archaic 8 .59%
Tampa Post Archaic 2 .14%
Westo Post Archaic 1 .07%
St. Johns Series Pottery Post Archaic 12 .89%
Total: 1347 100.00%
MITCHELL AND PHARMER THE PURDY BiFACE 7
Table 3. Summary of temporal associations for diagnostic artifacts from the Santa Fe
River Hendrix Collection.
Time Period Count Percent of Total
Palco 494 36.70%o
Early Archaic 528 39.19%o
Middle Archaic 200 14.840%
Late Archaic 54 4.00%o
Post Archaic 71 5.27%o
Total 1347 100.00%
Table 4. Morphological data for Hendrix Scrapers in this study. All items are from the
Santa Fe River and are in the Alvin Hendrix Collection at SRM.
Max. Max. Max. Weight Left edge Right edge Hendrix
Length Width Thick Angle Angle Collection
cm cm cm grams degrees _degrees Number
1 8.88 2.87 1.76 45.80 45 60 HC9537
2 11.40 3.58 2.55 87.60 60 76 HC9543
3 8.28 3.24 1.92 39.50 67 64 HC9544
4 8.17 2.68 2.05 40.50 83 85 HC9552
5 8.96 3.79 2.62 66.20 73 53 HC9559
6 9.39 2.74 2.33 64.20 74 76 HC9569
7 8.94 3.02 2.75 60.60 60 84 HC9579
8 11.30 3.79 1.90 68.30 60 85 HC9580
9 8.98 3.59 2.31 68.50 64 46 HC9585
10 8.94 3.48 2.80 58.00 53 66 HC9587
11 11.80 4.23 2.01 104.70 68 65 HC9590
12 8.07 3.27 1.84 44.00 84 73 H-C9591
13 7.27 2.62 1.57 27.30 67 57 HC9596
14 8.20 2.97 2.07 43.80 68 72 HC9597
15 9.60 3.73 2.52 73.60 80 55 H1C9598
16 9.36 3.87 2.19 72.10 72 54 HC9603
17 7.91 3.31 3.01 57.50 65 66 HC9604
18 9.82 3.48 1.65 65.50 55 48 HC9607
19 11.00 3.51 2.71 101.60 70 84 HC9616
24 8.7 3.43 1t_ 1.1 4.20 71 76I 7 HC9639
72 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
or 69 percent) of the specimens included in the study were Unless broken or worn, the distal end of the Purdy Uniface recovered is presented in Table 1. This trend is also evident in is typically sharper and not as thickly formed as that of the another collection assembled by Don Munroe from the same Hendrix Scraper (Figure 2). section of river. David Thulman (2006) documented Munroe's Often found together in archaeological and riverine collection and notes that with the exception of three Archaic assemblages, it is likely that Hendrix Scrapers and Purdy stemmed points, a Hernando point, a probable Culbreath Unifaces are contemporary; however, the differences in their point, and a single prehistoric pottery sherd, the rest of the features suggest that they may have been used for different artifacts collected by Munroe all date to the Paleoindian or tasks. Figure 2 shows the considerable difference in the Early Archaic periods, morphology of the two tools.
Given the limitations of the current sample, we cannot say
for sure whether the Purdy Uniface dates to the Paleoindian, Conclusion
Early Archaic, or spans both periods. However, based on
the sample used in this study, it is very likely that the tool The Purdy Uniface is a newly defined formal stone was not manufactured after the Early Archaic period. This tool type from north-central Florida. While the recovery of hypothesis is supported by the stone tool assemblage from the this tool type with temporally diagnostic artifacts indicates section of the Santa Fe River that yielded the Purdy Unifaces that this uniface likely dates to the Paleoindian and or Early first recognized by Alvin Hendrix as a unique type. Of the Archaic periods, a larger sample of specimens is needed from temporally diagnostic tools recovered from this area, 36.7 stratified and professionally excavated sites in order to refine percent date to the Paleoindian period and 39.2 percent date the temporal association of the Purdy Uniface. Time and to the Early Archaic period. All other later periods account for additional data will tell whether this tool dates to one or both only 24.1 percent (Table 4). periods.
The Purdy Uniface is best described as a medium-sized
A Potentially Related Tool uniface made from a large flake with a pointed distal end and
broad rounded base. The distal end curves slightly to the left (in
The Purdy Uniface appears to be closely related to 98 percent of our sample). The tool exhibits steep edge angles
the Hendrix Scraper (Figure 2). The Hendrix Scraper was on one margin (usually the one that the tip leans toward) for identified by Purdy and named in recognition ofAlvin Hendrix. scraping purposes and a lower edge angle suitable for cutting However, when comparing "typical" measurements for the tasks on the opposite margin. Preliminary analysis of a subtwo types, the Hendrix Scraper differs considerably from the sample of Purdy Unifaces found use-wear on all specimens in Purdy Uniface in numerous ways (Figure 2, Appendix and the form of edge wear associated with the scraping and cutting
Table 3). of hard materials. None of the 20 Purdy Unifaces analyzed
The distal offset of the Hendrix Scraper is highly variable for use-wear in this study exhibited evidence of being hated. as mentioned above. The distal offset of the Purdy Uniface is None of these tools showed evidence of being re-sharpened more uniform and appears to be one of the defining traits of the after heavy use, and they appear to have been discarded after tool. The Hendrix Scraper is also typically thicker along most the task at hand had been completed. of its length when compared to a Purdy Uniface. Simply put, as A more comprehensive formal study of edge wear is a type, the Hendrix Scraper is much too thick when measured needed, as is an analysis of stone sources, tool failure, and the from the dorsal to ventral surfaces, to be an exhausted Purdy various "life stages" of the Purdy Uniface in order to better Uniface. The basal end of the Hendrix Scraper is also relatively understand this new tool type. Purdy wisely notes that it is narrow when compared to the Purdy Uniface which exhibits a difficult to understand fully the use of the tool from the current characteristically flat, wide proximal end or base. sample because we are only looking at the end product of
The Hendrix Scraper exhibits a smaller difference (50) in manufacture and use (Barbara Purdy, personal communication, the median edge angles of the two margins, while the Purdy 2011). Uniface exhibits a significantly greater difference (210) in the The Purdy Uniface also appears to be related to, and median angles of the two margins (Appendix and Table 4). contemporaneous with, the Hendrix Scraper. However, Furthermore, the median edge angle of both margins on the consistent differences in morphology indicate that the two Hendrix Scraper are steeper (700 and 650 respectively) than forms are separate but related types (much the same way a those found on the Purdy Uniface (550 and 34respectively) Bolen point and an Edgefield Scraper, or a Dalton Adze and (Appendix and Table 4). The typical Hendrix Scraper lacks a an Aucilla Adze, are related). Generally, when compared to a cutting edge and based on steep edge angles, appears to have Hendrix Scraper, the "typical" Purdy Uniface exhibits a more been used mainly as a scraping tool (Table 4) (Purdy 1981:18). uniform distal offset, is not as thick along the body of the tool, Conversely the Purdy Uniface typically exhibits one margin has a wider proximal end (or base), has a steep scraping edge with a steeper edge angle suitable for scraping and the other on one margin and a slightly "thinner" cutting edge on the margin with a lesser edge angle suitable for cutting (Appendix other margin, and has a more pointed proximal end (or tip). and Table 4). As more examples are identified, and we hope this article
The distal end or tip of the Hendrix Scraper generally will facilitate that, the spatial distribution of the Purdy Uniface appears as a steep blunted point, some sharper than others. will become better understood. What is clear at this time is that
MITCHELL AND PHARMER THE PURDY BIFACE 73
we have a new unifacial tool type that was made in the region information from anyone who has similar item(s) in their of Florida spanning from the panhandle to the Tampa Bay collections. area. The authors are hopeful that the Purdy Uniface, named in recognition of Dr. Barbara Purdy, will soon become another References Cited
formal tool type that archaeologists can use to better assign temporal affiliation and the type of activities that took place Ahler, Stanley A. thousands of years ago at sites across Florida. 1971 Projectile Point Form and Function at Rodgers Shelter,
Missouri. Missouri Archaeological Research Series,
Acknowledgements no. 8. College of Arts and Science, University of
Missouri-Columbia and the Missouri Archaeological
We are naming this artifact after Dr. Barbara Purdy Society. Columbia, Mo.
in recognition of her significant contributions to Florida archaeology and our understanding of prehistory. Dr. Purdy Ballo, George R. is well known for working closely with both professional and 1985 Experiments in Use-Wear Formation on Stone Tools avocational archaeologists, collectors, and the general public. Made from Florida Chert: A Study Supporting a If not for this approach, many unique artifacts types recovered Microwear Analysis of Pa leo-Indian Lithic Artifacts by private individuals, including the Purdy Uniface, would not from the Hamey Flats Site (8-Hi-507), Tampa, be known today. Dr. Purdy also provided insightful review of Florida. Unpublished Master's thesis, Department of
our manuscript and informally inspected a sample of Purdy Anthropology, University of South Florida, Tampa.
Unifaces to help us better understand what we were seeing.
Alvin Hendrix deserves recognition for his thirst for Brink, John W.
knowledge and policy of always sharing his collection in the 1978 An Experimental Study of Microwear Formation name of science. Alvin's generosity is evident in the many on Endscrapers. Mercury Series, no. 83. National
scholarly works that reference his collection and in the artifacts Museums of Canada, Ottawa. he has donated to institutions such as the Florida Museum of Natural History, the Matheson Museum, the University of Butler, William B. South Florida Anthropology Department, and the Silver River 1980 An Inexpensive and Accurate Goniometer. Lithic Museum. Technology 9.65.
Richard W. "Rich" Estabrook, of the Florida Public
Archaeology Network, contributed many hours to this project Daniel, I. Randolph, Jr., and Michael Wisenbaker and also analyzed a sub-sample of Purdy Unifaces from the 1987 Harney Flats. A Florida Paleo-Indian Site. Baywood Hendrix collection for use-wear. Rich also assisted us in Publishing, Farmingdale, NY.
tracking down other specimens such as those found in the Harney Flats assemblage. Dunbar, James S., Michael K. Faught, and S. David Webb
We appreciate the assistance Dr. David N. Dickel of the 1988 Page/Ladson (8Je591): An Underwater PaleoBAR in identifying additional specimens from state collections Indian Site in Northwestern Florida. The Florida that were included in this study. Dr. Dickel also helped with Anthropologist 41:442-452.
providing data and photographs for these artifacts. James "Jim" Dunbar of the BAR also provided many useful comments and Movius, H.L., Jr., N.C. David, H.M. Brikler, and R.B. Clay references for archaeological reports that may hold clues to the 1968 The Analysis of Certain Classes of Upper Paleolithic temporal affiliation of the Purdy Uniface. Dr. David Thulman Tools. The Peabody Museum, Cambridge, Ma.
of George Washington University shared his expertise of lithic technology and provided background information on research Milanich, Jerald T. conducted on unifaces from the Santa Fe River. Dr. Thulman 1994 Archaeology of Precolumbian Florida. University also coordinated peer review of our manuscript and offered Press of Florida, Gainesville.
important suggestions of his own.
Harley and Ryan Means, Don Munroe, Paul Lien, and Purdy, Barbara A.
Robert Allen all helped to identify Purdy Unifaces found in 1981 Florida's Prehistoric Stone Technology. University their private collections and made those specimens available Press of Florida, Gainesville.
for study. The authors would like to note that the private collections referenced in this study were legally obtained Serbousek, Don under the guidelines of the Isolated Finds Policy (IFP) of 1983 Fxploration of a Paleo-Indian Site on the Aucilla the Division of Historical Resources, Florida Department of River. The Florida Anthropologist 36:88-97.
State prior to cancelation of the policy. Finally, Martha G. Pharmer, wife of co-author Monty Pharmer, provided many Thulman, David hours of valuable assistance with research, editing, measuring 2006 A Suwannee/Bolen Artifact Assemblage from the specimens, recording data, and computer entry. We welcome Santa Fe River. The Florida Anthropologist 59:21-34.
comments pertaining to the Purdy Uniface and kindly solicit
74 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
Appendix. Morphological data for Purdy Unifaces included in this study.
Max. Max. Max. Distal Left Weight Left Edge Right Edge Provenience Note:
Length Width Thick Offset Right Angle
cm cm cm degrees None grams degrees Angle degrees
1 8.51 5.17 2.36 20 Left 86.0 50 38 Suw. River HC7059
2 9.15 5.00 1.39 38 Left 69.7 55 27 SF River HC9423
3 7.46 3.68 1.40 22 Left 34.2 64 32 SF River HC9424
4 10.00 4.37 1.00 13 Left 48.3 48 30 SF River HC9425
5 8.20 4.75 1.55 20 Left 57.3 73 32 SF River HC9426
6 8.55 4.84 2.08 29 Left 60.3 44 36 SF River HC9427
7 7.03 4.08 1.20 7 Left 30.1 44 28 SF River HC9428
8 7.82 4.47 1.65 16 Left 45.8 55 34 SF River HC9429
9 9.16 5.51 1.92 19 Left 87.8 56 36 SF River HC9430
10 8.28 4.38 1.32 41 Left 51.8 53 36 SF River HC9431
11 7.25 5.02 1.45 18 Left 46.7 40 27 SF River HC9432
12 8.18 4.40 2.06 30 Left 67.5 54 35 SF River HC9433
13 8.62 4.51 1.87 12 Left 69.7 63 35 SF River HC9434
14 8.71 6.00 1.78 33 Left 81.1 50 44 SF River HC9435
15 9.79 5.45 2.67 17 Left 106.6 60 46 SF River HC9437
16 7.91 4.58 1.76 15 Left 49.8 50 34 SF River HC9438
*17 8.58 4.91 1.60 16 Left 57.8 56 30 SF River HC9443
18 8.82 4.40 1.58 33 Left 63.0 63 32 SF River HC9444
19 8.84 5.11 1.93 23 Left 76.7 57 46 SF River HC9446
20 7.99 4.26 1.16 32 Left 46.5 47 25 SF River HC9448
21 9.50 4.70 1.49 13 Left 63.0 45 33 SF River HC9449
22 7.40 3.94 1.24 25 Left 41.2 39 45 SF River HC9451
23 12.14 7.76 2.62 30 Left 191.2 35 28 SF River HC9454
24 8.94 4.65 1.99 20 Left 72.4 46 35 SF River HC9455
25 9.00 4.51 1.90 23 Left 61.9 48 35 SF River HC9457
26 7.40 4.19 1.62 18 Left 45.9 45 28 SF River HC9458
27 7.61 4.37 1.44 30 Left 47.6 71 28 SF River HC9462
28 7.36 4.63 1.53 18 Left 46.8 60 31 SF River 11C9463
29 8.18 4.95 1.44 32 Left 55.6 62 29 SF River 11C9464
30 6.88 4.73 1.95 20 Left 57.5 63 42 SF River 11C9465
31 7.54 4.18 1.91 23 Left 50.7 62 30 SF River 11C9466
32 9.97 5.76 1.42 36 Left 76.4 29 24 SF River HC9468
33 9.36 4.47 1.87 23 Left 66.2 60 33 SF River 11C9554
34 10.08 4.20 1.66 16 Left 59.8 69 30 SF River HC9593
35 10.18 4.41 1.63 17 Left 60.4 63 31 SF River HC9606
36 8.87 4.47 1.45 19 Left 57.7 56 42 SF River 11C9611
MITCHELL AND PHARMER THE PURDY BIFACE 75
Max. Max Max Distal Left Weight Left Edge Right Edge Provenience Note:
Length Width Thick Offset Right Angle
cm cm cm degrees None grams degrees Angle degrees
37 8.71 4.23 1.39 16 Right 55.7 35 54 SF River HC15185
38 9.22 4.36 1.25 27 Left 46.9 58 30 Wacc. River HC16450
39 7.20 3.55 1.71 24 Left 39.5 67 59 Apal. Riv. Means- 244
40 6.90 4.00 1.30 29 Left 39.6 64 55 Wac. Riv. B Means -NN
41 7.50 4.80 1.27 15 Left 45.2 58 33 8TA98 BAR-07-195-242.1
42 7.47 4.13 1.53 18 Left 43.2 50 37 8TA98 BAR-07-195-249.1
43 9.00 5.35 1.84 18 Left 84.8 54 36 8TA98 BAR-07-195-250.1
44 11.6 5.59 1.88 46 Left 112.2 57 35 8H1507 BAR-81-163-3967
45 7.60 5.12 1.60 34 Left 57.5 68 36 8H1507 BAR-81-163-184
46 6.80 4.12 1.27 38 Left 36.0 55 36 8H1507 BAR-81-163-33
47 7.60 3.90 1.40 14 Left 36.8 50 35 8H1507 BAR-81-163-404
48 8.67 5.26 1.23 16 Left 54.3 51 33 SF River Monroe 9942
49 6.98 3.95 .82 11 Left 35.0 60 50 SF River Monroe NN
50 9.91 4.68 1.47 20 Left 69.6 57 32 Hills. Riv. Lien NN
51 8.45 5.05 1.47 41 Left 59.3 54 32 Hills. Riv. Lien NN
52 8.50 4.27 1.29 21 Left 52.3 70 30 Suw. Riv. Allen NN
53 9.25 4.87 1.75 30 Left 72.1 52 35 Suw. Riv. Allen NN
54 8.50 4.41 1.75 28 Left 57.4 80 36 Hem. Co. Allen NN
55 8.40 4.44 1.41 20 Left 49.7 55 38 Mar. Co. Allen NN
Avg. 8.50 4.67 1.61 23.33 Left 60.69 55.09 35.2 NA NA
S Dev. 1.13 .67 .36 8.25 NA 24.63 9.99 7.37 NA NA
Med 8.50 4.51 1.55 20 NA 57.4 55 34 NA NA
Range 8.18 4.47 1.39 20 NA 69.7 50 36 NA NA
*=Purdy Uniface Type Specimen (HC # 9443).
SF River Santa Fe River Suw. River Suwannee River
Wacc. River Waccasassa River Apal. Riv. Apalachicola River
Auc. Riv. Aucilla River 8111507 Hamney Flats Paleo Site
Wac. Riv. B Wacissa River Below Goose Pasture & Slave Canal
8TA98 Serbousek/Catrill Site (Aucilla River Half Mile Rise Paleo Site)
Hills. Riv. Hillsborough River near Harney Flats
Hemn. Co. Hemnando County Bailey Hill Mrs. Loan's property
Mar. Co. Marion County Not specific Unknown sinkhole
Abbreviations: HGC- Hendrix Collection Means Harley & Ryan Means
BAR Bureau of Archaeological Research
Monroe Don Lein Paul
Allen Robert NN No Number
Avg. Average S Dev. Standard Deviation
a#+ Aq p ap Aw, d i v 04*- 05 OMS, 10
4 UO U4,JQ C$OIJ wP j 0 tJ N 4 Aq FOJ M, 10:1 P 9 UP U,
ueS uMsMijo spu" sip al apd:vd rpstp avWw asEaW
w, ct j n c-A amoK cx ueAp v L vot w om,,, is ped a q Isnw saa M
'9009-S-Pm o lie-I aeeald 'dO:491 ta lfo dn ub is ol. j4-jaj," 5 pa ',,Wcj aup sa-10 S
j r .....
A041C-1 4 1- >e uo Pel A041 Sa!Dt -ul u I
e'-41 P' P-, -s( IPP-4 SAPMP004 z dv Se
:u s i u re d Q e LRJ ri q:,"p. u, j e, &1 5 1 PR I 1 -9 W, 6utun.3 pue 6u4jew Ip Ijim
P: P" 4 weI ,q
lei UO 10:3 6 utjpdaj d a 9m Sro"D 41 no), "'63U. L 'Piq o uoi i --,r
Mdl X) p 0j,
441 D, F U a, ipunoje jaL J04 aDjO S UPS. uot ;^.
. .... ... ..
N IYOD' L I Ic GO '6
XV04 OAVOWS HIC L AM Xv, a NI- n, S
a : 5 ....
11"I a M, '-M
... ..... JE
.......... ........... ...... .....
PALEOINDIAN OCCUPATION ALONG THE ST. JOHNS RIVER, FLORIDA
DAVID K. THULMAN
Department ofAnthropology, George Washiongton University, Washington D.C. 20052 email. firstname.lastname@example.org
The Paleoindian occupation of Florida during the late create a defined channel until just north of Lake Helen Blazes Pleistocene Epoch (approximately 11,500 to 10,000 B.P.) (White 1970:103). The upper river, between its headwaters and
- whether measured by the number of artifacts or sites- its confluence with the Econlockhatchee River, follows the St. has conventionally been perceived as concentrated along Johns Valley, connecting a series of elongated lakes aligned river basins that empty to the Big Bend region of the Gulf with the river's channel (White 1970:93-204). The outline of of Mexico, from the Apalachicola River to Tampa Bay. The the valley has been obscured in places by agricultural fields, geology of this region provided more accessible water and but Figure 2, based on the present extent of mucky soils, shows chert resources during that time (Dunbar 1991; Thulman an approximation of the likely original configuration. Today, 2009), while deep overlying sands tended to cover those the St. Johns Valley is filled with muck and peat, except for resources on the Atlantic coast. Other than the sites in and near the channel and lakes (Davis 1946). The precise basal depth Silver Springs in Marion County (Hemmings 1975; Hoffman of this valley is not known but is probably about 3.0 to 3.6 m 1983; Neil 1958, 1964) and the Helen Blazes site (8BR27) at its deepest point; several of the lakes in this area are about on the upper St. Johns River in Brevard County (Edwards 3.05 m deep with about 0.3 m of organic sediments overlying 1954), Paleoindian presence along the St. Johns River and marine clays (Brenner 1997). North of Lake George, the river its tributaries appeared to be sparse. However, recent work channel was deeply incised during the late Pleistocene (Miller indicates the Paleoindian occupation was more extensive than 1998). earlier recognized (Thulman 2006, 2009). This article reports From Sanford to Palatka, the St. Johns River is offset to on the Lake George Point site (8PU 1470), a newly identified the west and apparently follows a valley of different geological Suwannee-age site submerged in Lake George in the St. history. In this stretch, the river follows a narrower valley that
Johns River and places it in a larger context of Paleoindian connects larger lakes, including Lakes Monroe, Woodruff, and sites in the St. Johns River watershed. Suwannee-aged sites George (Florida Department of Natural Resources 1989; White have not been dated, but they likely date to the post-Clovis, 1970). Given its width and location, Crescent Lake may have Middle Paleoindian period (approximately 10,500 to 10,000 been part of the original river channel before it was captured B.P.). The Lake George Point site fits a pattern of larger by the offset (Miller 1998). Lake George is the last and largest Middle Paleoindian sites in Florida located on a vantage point lake intersected by the river before it enters its broad, highimmediately adjacent to and overlooking a broad plain, such banked channel north of its confluence with the Oklawaha as Helen Blazes, Bolen Bluff (8AL439; Bullen 1958), and River. Lake George is 186 km2 (71.8 square miles) inareaand Harney Flats (8H1507; Daniel and Wisenbaker 1987). Figure occupies a broad shallow basin that is rarely more than 4.6 m 1 shows the location of these and other sites discussed in this deep. Two large springs empty into the lake from the west: paper. Silver Glen Spring and Salt Spring (Scott et al. 2004).
During the late Pleistocene Epoch when sea levels were
The St. Johns River Watershed during the Late lower, the St. Johns was significantly different, likely dry or
Pleistocene intermittent in its lower reaches (Miller 1998; Thulman 2009).
With lower sea levels, the Floridan Aquifer, which is the
The St. Johns River is Florida's longest river, but unlike source for the largest springs in Florida, would probably not most of the other northern rivers in the state, it receives less of have been discharging to the surface in the St. Johns during its flow from large springs or groundwater infiltration; most of drier episodes as it does today (Thulman 2009), although water its flow comes directly from precipitation (Florida Department may have been available in non-flowing spring vents. Coupled of Natural Resources 1989; Miller 1998). The north-flowing with the lower precipitation at that time, the river may not river has a shallow slope and only drops 8.2 m (27 feet) in have flowed and may only have held water intermittently in elevation from its headwaters in St. Lucie County to its mouth scattered locations, such as perched water sources. Likewise, near Jacksonville, about 0.19 m per 10 km (1 foot/lO mile). peat and muck, which form in anaerobic subaerial conditions The river originates in a broad swampy valley and does not (Davis 1946), would have been spotty at best and possibly
VOL. 65 (1-2) THE FLORIDA ANTHROPOLOGIST MARCH-JUNE 2012
78 THE FLORIDA ANTHROPOLOGIST 2012 VOL.65 (1-2)
,;'! Suwa~nee P aynes ,i .... f Rive' -_ ririe
iN ..... River" + "
------ ." .............. ....... ........ t o n ,R
Figure 1. North-central Florida. Paleoindian sites mentioned in the text: 1 Goose Pasture, 2 Bolen Bluff, 3 Lake George Point Site, 4 Lake Helen Blazes, 5 Harney Flats, 6 Silver Springs.
absent in the St. Johns Valley. Thus, the modem muck and of Georgetown, Putnam County, at the northern end of Lake peat-filled headwaters of the St. Johns would also have been a George (Figure 3). The Lake George Point site maintains the wide shallow basin. potential to be an important Paleoindian site, although a recent
The northern end of Lake George was probably an area survey of the site showed that at least the easily accessible of reliable surface water during the late Pleistocene Epoch part of the site was likely completely collected by amateurs. (Thulman 2009). Highly fractured geological formations I examined more than 40 Suwannee-type points and broken underlie an area from the north end of the lake to the lanceolate point bases, scores of unifacial chipped stone tools, confluence of the Oklawaha River, which could have provided and fossils of extinct fauna recovered by the collector. (Some an outlet for water under artesian pressure from the Floridan of these artifacts would probably meet one of the definitions of Aquifer system during periods of higher precipitation. Thus, a Simpson point [see, Bullen 1975:56; Daniel and Wisenbaker people at Lake George could have had access to a variety of 1987:53], buttI adopt Dunbar and Hemming's  argument resources: water, open savannahs, swamp margins, relatively that the differences between Simpsons and Suwannees--other high vantage points, and perhaps access to raw materials for than the rare bull-tongued Simpsons are not defined well stone tools in the spring vents or in now-drowned exposures enough to make a rigorous distinction.) Figure 4 shows a in the bottom of the lake and river. A Paleoindian standing small sample of the complete Suwannee points from the site. on the edge of the St. Johns Valley or Lake George at the end Based on this assemblage and an estimate of site dimensions of the Pleistocene would have seen a dramatically different (approximately 28,000 in2), the site appeared to have been vista than a modern Floridian. Rather than the modern view substantial in size, comparable to the Suwannee-Bolen Harney of a broad flat expanse of water or muck, a Paleoindian would Flats site in Temple Terrace (Daniel and Wisenbaker 1987). look out over a broad plain. The vista would have been similar The approximate boundaries of the site are outlined in Figure to that seen from the Bolen Bluff site (Bullen 1958) at the 5. The western boundary runs along the edge of a submerged, southern edge of Paynes Prairie in Alachua County. shallow spit, which drops off to depths exceeding 2 m. Because
the private artifact collection contained relatively few Early
The Lake George Point Site (8PUJ1470) Archaic-period Bolen and Middle Archaic-period stemmed
points, it appeared likely that the Lake George Point site may
The Lake George Point Site (8PU1470) was found by have had a distinct Suwannee component, which could have
local artifact collectors in Lake George and reported on a State provided a means to distinguish between the Suwannee and of Florida Isolated Finds form1. It is located in the vicinity Bolen assemblages. of Lake George Point, which is near the unincorporated town Faunal and botanical material is preserved in the lake
THULMAN A PALEOINDIAN SITE IN LAKE GEORGE 79
identify sites in Florida that may preserve Paleoindian-age organic artifacts. The purposes of the survey were to identify the dimensions of the site, determine whether any portion of the site remained intact after years of intensive collection and still contained an undisturbed Suwannee-age component, and evaluate the potential for more intensive survey and excavation.
The survey was done during March 2009 when water levels were low and the warm weather algae blooms had not started. Nevertheless, visibility in the lake was very poor to non-existent, and work at depths deeper than 1.5 m was by feel. Three 60 m transects were done in the shallower areas within the site boundary. Due to the large size of the site and our limited time, transects and random loci were positioned to assess the potential of the site to preserve undisturbed archeological deposits. Transect 1 extended beyond the site boundary; Transect 3 was along what appeared to be an exposed section of hardpan; and Transect 4 was set to the west of the exposed hardpan. Sediment cores were taken with a hand auger along Transect 5 to assess the potential for preservationnearer to the shore where sand deposits were deeper; no material was collected. Transect 2 was taken at an offsite location in the Figure 2. A portion of the upper St. Johns River Valley area. The initial protocol was to have a team of divers swim showing the approximate boundaries of the relict lagoon along the transect, fan the bottom, and collect artifacts, but system and the Helen Blazes site (8BR27). it was soon apparent that fanning the bottom stirred up the
bottom sediments and produced zero visibility. This protocol sediments. In addition to the extant and mineralized extinct was changed midway along the first transect to sample every faunal material collected at the site, cores taken by the United 10 m by fanning the bottom for 10 minutes as a control. States Geological Survey (USGS) and the St. Johns Water Transects 3 and 4 were sampled every 10 m by establishing a Management District in the northern part of the lake revealed I m square and fanning until the hardpan was reached. Three wood and peat preservation, although the ages of these Random Loci (5, 6, and 7) were also collected in deeper water materials are unknown (Kindinger et al. 2000; Nancy DeWitt, outside the approximate site boundaries (Figure 5). personal communication 2008). Because the site presented While 109 fragments of extinct faunal material were
the opportunity for significant organic preservation, the recovered, indicating parts of the site date to the late Pleistocene Archaeological Research Cooperative, Inc., agreed to sponsor Epoch, no unambiguously diagnostic Paleoindian artifacts an underwater survey of the site as part of a larger effort to were recovered. The faunal material consisted of both extinct
Figure 3. Northern portion of Lake George showing location of the Lake George Point site. The white line bisecting the lake is the Putnam County/Volusia County boundary.
80 THE FLORIDA ANTHROPOLOGIST 2012 VOL.6 12
Figure 4. Four of the complete Suwannee points collected from the Lake George Point site.
These are representative of the majority of the Palcoindian points found.
(21 percent) and extant (79 percent) species (percentages based stained material, we can infer that the Pleistocene surae a on counts of terrestrial species; fish were not included). Table covered by sand before it was inundated. We do ntko 1 lists the number of bones of extinct and extant terrestrial how fast it takes bone and ivory to absorb the distintvedr species collected from each transect and locus. None of the stain that is common for bone recovered in manyFord faunal material showed clear signs of human modification. rivers, although it does not appear to take long (JameDubr A large number of highly water worn St. Johns Plain sherds personal communication, 2011). However, it is Ilkel( hah were recovered along with stained faunal remains of extant Pleistocene surface at the Lake George Point site wav eea species, which suggests the presence of a Woodland-period meters above the bottom of the channel of the paleo-t( on midden. Extant and extinct faunal species (Table 1), including River and would only have been inundated in themidet mammoth, giant tortoise, and horse, were identified, but of late Holocene Epoch (approximately 5000 B.P topeet more interest is that 96 percent (101 of 105) of the extinct when sea-levels rose, forcing levels in the St. Johns Rvras faunal material was unstained, while 49 percent (204 of 413) to rise (Miller 1998). of the extant material was unstained. Thus, it appears the There is no obvious indication that the site wasusd, o Pleistocene age material was covered and segregated from tool production, as was the case at Harney Flats, althuhti the absorption of minerals such as magnesium or tannins in is not easy to evaluate because the collectors did ntgte the water. The surface much of the Pleistocene-age material flaking debris. Several lines of evidence, however, sugetta
THULMAN A PALEOINDIAN SITE IN LAKE GEORGE 81
the Chipola River below Marianna, and in the Suwannee
River between the Withlacoochee River and Santa Fe River
Paleoindian site distributions generally follow the pattern
Random 0 of occurrence of individual diagnostic points, although
*= dom \ \ ",larger sites have not been found in locations with the highest
omi ,6 concentrations of individual artifacts. Three large Paleoindian
sites have been recorded: Harney Flats in Temple Terrace
near the Hillsborough River (Daniel and Wisenbaker 1987),
6dm,< Goose Pasture near the Wacissa River, and Lake George Point
lo #5,60Tm (Figure 1). The excavated portion of Harney Flats is the only
site not significantly disturbed by private collectors. Smaller
sites discussed below include Helen Blazes (Edwards 1954)
on the upper St. Johns River and Bolen Bluff (Bullen 1958)
Figure 5. The site, showing approximate site boundaries, adjacent to Paynes Prairie (Figure 1). transects and random loci. Based on a reconstruction of late Pleistocene water
availability, Thulman (2009) found a correlation between
high probability sources of available water and numbers of
collection, although these may have been stored elsewhere. Middle Paleoindian (Suwannee-age) points. Although the The collection included 13 bases that were snapped at or near highest probability for water availability and the greatest the haft. number of points was in the Santa Fe River, the segment of
the St. Johns River from the northern part of Lake George
Discussion through the Oklawaha River confluence also was determined
to be a high probability locale. The Lake George Point site is
Previous analyses of the distribution of Paleoindian within this area. Other high probability locations for surface artifacts in Florida found a robust correlation between the water in the St. Johns watershed include the lower reaches number of artifacts and water bodies, i.e., significantly more of the Oklawaha River (much of which is submerged below Paleoindian artifacts are found in or adjacent to rivers and lakes the Rodman Reservoir) and Silver Springs. Locations with a than are found in uplands (Dunbar 1991; Neil 1964; Thulman moderate potential for surface water include the southern end 2006; Waller 1970). Dunbar's (1991) work demonstrated that of Crescent Lake and portions of the St. Johns upriver from most of the artifacts were found in or near water bodies west Alexander Springs (Thulman 2006, 2009). Thulman (2009) of the central Florida ridge, but later work by Thulman (2006), did not analyze the surface water potential in the area around using somewhat different criteria2, identified a previously Lake Helen Blazes, because no points from this area were unrecognized Paleoindian presence in the Atlantic watershed. include in his data. However, based on the reconstruction of Thulman collected data on 744 Middle Paleoindian points surface water availability downriver in the St. Johns River (unfluted lanceolates with ground bases) from public and in and around Lake George, it is likely that the area around private collections from the north-central Florida peninsula, Lake Helen Blazes followed the general pattern in the river of a region generally bounded by the Chipola River, Tampa Bay, reliable surface water in scattered locations. Atlantic Coast, and Florida-Georgia border. Eighty-one (11 The Lake George Point and Helen Blazes sites are on the
percent) of the points were from the St. Johns River (n=55), its edges of wide, fiat-bottomed basins (Figures 2 and 3). While tributaries (Oklawaha [n=m6] and Silver [n=n5] rivers) or nearby the vegetation in these basins during the Middle Paleoindian waterbodies (Crescent Lake [n=m19]). While this represents period is unknown, given the much drier conditions during only about one-tenth of the entire assemblage, the percentage that time, they could have been dry savannahs that provided an compares favorably with that of Middle Paleoindian points in unobstructed vista. This choice of site location is not particular the Aucilla-Wacissa River basin (9 percent), Chipola River to the St. Johns; Bolen Bluff and Harney Flats also occupy (8 percent), and Suwannee River (11 percent); the bulk (30 similar landforms. Moreover, none of these sites occupies the percent) of the river-associated points examined by Thulman highest vantage point in the area, rather they are located at a were from the Santa Fe River (30 percent). Thus, while a clear lower and closer vantage point to the basin proper. It seems concentration of Middle Paleoindian activity was in the Santa likely that the Lake George Point and Helen Blazes sites were Fe River (downriver from River Rise to the confluence with chosen to facilitate hunting animals foraging in the wide, flat the Ichnetucknee River), there seems to have been a relatively landscapes in Lake George and the St. Johns Valley. These broad but even distribution of Palcoindians across the region, locations on the St. Johns must have been a powerful draw to concentrated in several stretches of these rivers: northern Lake Palcoindians, since their chert resources were distant, unlike at George in the St. Johns, the lower Aucilla and Wacissa rivers, Bolen Bluff and Harney Flats, where chert crops out locally.
82 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
Conclusion 3. The details of the artifact distribution, including the
possible biases, are discussed at length in Thulman (2006, Previously, the absence of a robust Paleoindian presence 2009). on the eastern side of Florida was explained by the lack of accessible chert outcrops (e.g., Dunbar 1991), but it now Acknowledgments
appears that the region was attractive to Paleoindians despite the absence of chert. The Lake George Point site and isolated I thank Dellwood Nelson for identifying the location of the finds in Crescent Lake indicate this prior interpretation was site and generously sharing his collection from the area. I also based on incomplete data. Perhaps the St. Johns River sites thank the crew of volunteer divers who worked on the survey were part of a larger Paleoindian territory that included the of the Lake George Point site, Jessie Halligan, Bob Knight, quarries in Marion and Alachua counties, or perhaps there are Micah Mones, and Don Munroe; Barbara Purdy for letting us quarry sites in now-inundated sections of the lakes and the St. stay at her home; and the U.S. Forest Service, Bob, and Don Johns River. Resolution of that question will require further for allowing us to use their boats. Thanks also to Jim Dunbar, work. Michael Faught, Keith Ashley, and an anonymous reviewer for
Paleoindian archaeology in Florida holds great promise valuable comments and encouragement in this study. Finally, for supplementing our understanding of the organic artifacts thanks to the Archaeological Research Cooperative, Inc. for from that time period. The USGS cores show that organic partially funding the research. preservation in several areas of Lake George is widespread and the remains from the Page-Ladson site (8JE591; Webb References
2006), Windover (8BR246; Doran 2002), and other Middle Archaic burial sites in Florida (Beriault et al. 1981; Wharton Austin, Robert J., and Richard W. Estabrook et al. 1981) demonstrate that wood, fabric, and other delicate 2000 Chert Distribution and Exploitation in Peninsular artifacts can be preserved in Florida waters for millennia. Lake Florida. The Florida Anthropologist 53:116-130. George and other water bodies in the area may present the best opportunity for finding organic remains of Paleoindian age. Behrensmeyer, Anna K.
The Lake George Point site itself may be a lost opportunity 1978 Taphonomic and ecological information from bone for Paleoindian studies in Florida and the broader Southeast. weathering. Paleobiology 4:150-162.
Hidden from sight and mind, the St. Johns Paleoindian sites have eluded archaeologists until recently because they Beriault John, Robert Carr, Jerry Stipp, Richard Johnson, and are totally inundated and located in places few people have Jack Meeder bothered to explore. One reason may be the potential threat 1981 The Archaeological Salvage of the Bay West Site, of animals such as alligators and alligator snapping turtles. Collier County, Florida. The Florida Anthropologist
At least one collector was bitten by a large alligator while 34:39-58.
collecting in Lake George. It is possible that portions of the site are preserved in deeper water, which is not easily accessible Brenner, Mark by collectors, and other unrecorded sites exist in the lake, but 1997 Bulk Sedimentation andNutrient Accumulation Rates exploring those areas will take a more intensive effort and in Lakes of the Upper St. Johns River Basin, Final
much better lighting. Report. SJWMD Special Publication SJ97-SP20,
Bullen, Ripley P.
1. The Isolated Finds program was instituted by the Bureau 1958 The Bolen Bluff Site on Paynes Prairie, Florida. of Archaeological Research in an effort to gather information Contributions of the Florida State Museum Social
on the private collection of artifacts in freshwater, submerged Sciences 4:1-51. state-owned lands, like navigable rivers. Collectors were 1975 A Guide to the Identification of Florida Prqjectile supposed to report artifacts on a form that was submitted to Points. Kendall Books, Gainesville.
the Bureau. The program ended in 2007, and presently it is illegal to collect artifacts on all state-owned lands. Daniel, I. R., and M. Wisenbaker
2. Dunbar found 70 percent of his Paleoindian artifacts in 1987 Harney Flats.'A Florida Paleo-Indian Site. Baywood these water bodies, but his totals included fluted and unfluted Publishing, Farmingdale, NY. lanceolate points, diagnostic non-point artifacts (such as ivory shafts), and published site reports where the artifacts could not Doran, Glen H., editor be examined. Thulman used only unfluted lanceolate points 2002 Windover." Multidisciplinary Investigations of and that he could examine. Early Archaic Florida Cemetery. University Press of
THULMAN A PALEOINDIAN SITE IN LAKE GEORGE 83
Davis, John H. Scott, Thomas M., Guy H. Means, Rebecca P. Meegan, Ryan
1946 The Peat Deposits of Florida. Their Occurrence, C. Means, Sam B. Upchurch, R.E. Copeland, James Jones, Development, and Uses, Report of Investigation Tina Roberts, and Alan Willet
No. 32. State of Florida, Florida Geological Survey, 2004 Springs of Florida, Bulletin No. 66. Florida
Tallahassee. Geological Survey, Tallahassee.
Dunbar, James S. Smallwood, Ashley M., and Albert Goodyear
1991 Resource Orientation of Clovis and Suwannee Age 2009 Reworked Clovis Biface Distal Fragments from the Paleoindian Sites in Florida. In Clovis Origins and Topper Site, 38AL23: Implications for Clovis
Adaptations, edited by R. Bonnichsen and K. L. Technological Organization in the Central Savannah
Turnmire, pp. 185-213. Center for the Study of the River Region. Current Research in the Pleistocene
First Americans, Oregon State University, Corvallis. 26:118-120.
Dunbar, James, and C. Andrew Hemmings Thulman, David K.
2006 Florida Paleoindian Points and Knives. In New 2006 A.Reconstruction ofPaleoindian Social Organization
Perspectives on the First Americans, edited by in North Central Florida. Ph.D. dissertation,
Bradley T. Lepper and Robson Bonnichsen, pp. 65- Department ofAnthropology, Florida State University,
72. Center for the First Americans, College Station, Tallahassee.
Tx. 2009 Freshwater Availability as the Constraining Factor in
the Middle Paleoindian Occupation of North-Central Edwards, W. E. Florida. Geoarchaeology 24:243-276.
1954 The Helen Blazes site of central eastern Florida:
a Study in Method Utilizing the Disciplines of Waller, Benjamin
Archaeology, Geology, and Pedology. Ph.D. 1970 Some Occurrences of Paleo-Indian Projectile Points
dissertation, Department of Geology, Columbia in Florida Waters. The Florida Anthropologist
University, New York City. 23:129- 134.
Florida Department of Natural Resources Webb, S. David, editor
1989 Florida Rivers Assessment. Florida Department of 2006 First Floridians and Last Mastodons: the PageNatural Resources, Tallahassee. Ladson Site in the Aucilla River. Springer, Dordrecht,
Hemmings, E. Thomas
1975 The Silver Springs Site, Prehistory in the Silver Wharton, Barry R., George R. Ballo, and Mitchell E. Hope
Springs Valley, Florida. The Florida Anthropologist 1981 The Republic Groves Site, Hardee County, Florida.
28:141-158. The Florida Anthropologist 34:59-80.
Hoffman, Charles A. White, William A.
1983 A Mammoth Kill Site in the Silver Springs Run. The 1970 The Geomorphology of the Florida Peninsula,
Florida Anthropologist 36:83-86. Geological Bulletin No. 51. Bureau of Geology,
Florida Department of Natural Resources, Kindinger, J. L., J. B. Davis, and J. G. Flocks Tallahassee.
2000 Subsurface characterization of selected water bodies
in the St. Johns River Water Management District,
Northeast Florida. USGS Open-File Report 00-180.
Miller, James J.
1998 An Environmental History of Northeast Florida.
University Press of Florida, Gainesville.
Neill, Wilfred T.
1958 A Stratified Early Site at Silver Springs, Florida. The
Florida Anthropologist 11:32-52.
1964 The Association of Suwannee Points and Extinct
Animals in Florida. The Florida Anthropologist
of The Flori da A n thropologist
are available from the
Pal m Beach Museu:m of
GEOARCHAEOLOGICAL INVESTIGATIONS AND OSL DATING EVIDENCE IN AN ARCHAIC AND PALEOINDIAN CONTEXT AT THE HELEN BLAZES SITE (8BR27), BREVARD COUNTY, FLORIDA
W. J. RINKl, J.S. DUNBAR2, GLEN H. DORAN3, CHARLES FREDERICK4, AND BRITTNEY GREGORY5
SSchool of Geography and Earth Sciences, McMaster University, 1280 Main St.. W, Hamilton, Ontario, Canada L8S 4K] 2Department ofAnthropology, Florida State University, Tallahassee, FL USA 32306 3 Department ofAnthropology, Florida State University, Tallahassee FL USA, 32306
4 Department of Geography and Environment, The University of Texas at Austin. 1 University Station A3100, Austin, TX 78712 58704 Wagon Trail, Cross Road, TX 76227
Introduction prairies and forests extending to the eastern margin of the St.
Johns River floodplain. The terrain west of the Atlantic Coastal
The excavation of the Helen Blazes site (8BR27) between Ridge undergoes an imperceptible descent to the Helen Blazes 1949 and 1951 by William Edwards was remarkable for its site and continues to descend even more gradually to Lake time, but the assemblage of Archaic and Paleoindian artifacts Helen Blazes, a chain-lake that is part of the St. Johns River yielded no materials that could be scientifically dated. This (Edwards 1954). Lake Helen Blazes is part of the upper St. article reports on new excavations at Helen Blazes that were Johns River drainage located about 30 km north of the river's designed to incorporate optically stimulated luminescence headwater at Blue Cypress Lake. Flat, poorly-drained prairies (OSL) dating of quartz sand grains and sedimentological and forests continue west of the St. Johns river basin for 25 studies of the sediment aimed at a better understanding of to 30 km. Just west of this lies the higher (greater than 30 m the depositional context. The new excavations yielded lithic above sea level) Osceola Plain (White 1970). materials with typologies consistent with those found by The area is now a flat prairie that still seasonally floods for
Edwards (1954). Samples for sedimentological study were short intervals. Prior to construction of the Melbourne-Tillman analyzed to better understand the depositional context of drainage system in 1923 and additional ditching near the site, cultural materials by means of geoarchaeological analysis of it was probably a shallow marsh inundated at least part of the site deposits. The OSL ages ranging from about 5,400 to 12,000 year with rainwater and overflow from Lake Helen Blazes. years before present are in line with expectations for Archaic Edwards (1954) noted that with heavy rain the area could be to Paleoindian period contexts. However, the sedimentological flooded with as much as 1 m of water for short periods. This is and geoarchaeological interpretation together with the fine still an accurate description and short term flooding is common. detail of the OSL results suggest that the cultural material at The site is positioned on a slight, almost imperceptible, rise on the site was buried by bioturbation and internal reorganization what may be a relic dune or ancient terrace of the St. Johns of the deposits rather than normal geological sedimentation River. In this region of Florida even slightly elevated areas following occupation. would have provided some protection during episodes of high
The Helen Blazes site is located in peninsular Florida water. (Figure 1). It is in Brevard County, approximately 27 km westsouthwest of the city of Melbourne and about 9 km north- Previous Work at Helen Blazes
northeast of the headwaters of the St. Johns River. The site lies
in the southeast corner of Section 31, Township 29S, R36E, Earlier work in the immediate vicinity of the site in very near Latitude 28.500 N, Longitude 80.750 W (GPS the 1940s was prompted by the landowner's recovery of
coordinates given in the methods section). megafauna skeletal material from the South Indian Fields
Physiographically, the Helen Blazes site lies 19 km west site (8BR23; Rouse 1951), which is immediately adjacent to of the Indian River on the eastern margin of the St. Johns the Helen Blazes site (Figure 2). Excavations peripheral to River, 4 km east-southeast of Lake Helen Blazes. Here the the main excavations at 8BR23 identified seemingly discrete St. Johns River floodway is some 6.5 km wide with the lake archaeological deposits that would ultimately be designated as located in the approximate center of the floodway. The Indian 8BR27, the Helen Blazes site. Edwards, a Columbia University River is a long brackish water lagoon, separating barrier Ph.D. student who had been assisting Rouse and Clarence islands and peninsulas (which rarely attain elevations greater Simpson (Florida Geological Survey) in their Paleoindian than 4 m) from the Atlantic Ocean to its east. The area between investigations, used the excavations of 1949, 1950, and 1951 the Indian River and Lake Helen Blazes is occupied by the as the corpus of his dissertation (Edwards 1954). Edwards' Atlantic Coastal Ridge, which consists of flat, poorly-drained research was conducted with significant attention to the
VOL. 65 (1-2) THE FLORIDA ANTHROPOLOGIST MARCH-JUNE 2012
86 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
Figure 1. Location Map showing the Helen Blazes Site in Florida
geological nature of the deposits and the region's geological The sedimentary sequence is characterized by marine units history. Though his attempts to correlate the depositional comprised of shells, clay, and sand at the base (Strata I and II), history with ancient sea level histories relied mostly on now clayey to sandy clay units in the central section (Strata IIIA to discarded terminology, his stratigraphic and sedimentological IJIC), and sandy units at the top (Strata IV to yIIIC). documentation was excellent. Figure 3 provides a diagrammatic Edwards reported 7 lanceolate points in his preceramic explanation of the general stratigraphic situation and correlates Strata V and VI and 32 utilized or modified flakes. Four descriptions of the sedimentary units to both the major strata fragments of non-local sandstone 'abraders' were also in (Roman numerals) and subdivisions (Roman letters) and Strata V and VI with three possibly attributed to Strata VIII.
also shows the relative stratigraphic position of diagnostic The unutilized flakes show a similar distribution with all artifacts recovered by Edwards, as well as the OSL samples coming from Strata V and VI (Edwards 1954:Table 4). Thus, and select artifacts recovered from our recent excavations, the majority of the lithics clearly come from the preceramic levels with a heavier concentration in the lower of those levels. I All the lanceolate points exhibited grinding of concave basal
___ ____edges (Edwards 1954). In the 1950s, Paleoindian studies
4, were in their infancy, and many lanceolates were compared
. : ; 'J favorably with Folsom and Clovis artifacts, which were the
N first to be described in detail. Thus, Edwards described the
lanceolates from Helen Blazes as "Folsom-like" or "Suwanee"
FPrior to the finds at Helen Blazes most of the similar a n e of looking lanceolates in the state were from the higher elevations inr-nist in nort-entral Florida or in the tam e (Sis
discarded. terminology, his. s i and ....... 1948:13). Edwards reported looking at two hundred complete
do wa ex Fie 3and fragmentary lanceolate points in the Simpson and other
............ of the private collections and most of those had concave bases with
dsrpinoo the m r s a fbasallyground edgesthatwereasstraightasspecimensD and (Roman num ) .. sn ( n l ) aF (Figure 4; Edwards 1954). He further remarked that "FishK.______ _______1 tail" bases, as in point E of Figure 4, were abundant, while the
also shows .. r e marked "ears" ofpointG of Figure 4wouldnotappeartoo
Figure 2. Edwards' original site map of Helen Blazes much out of place in the private collection assemblage. He also
excavation squares (Edwards 1954:21) noted that true fluting was extremely rare among the central
Florida specimens (he saw only two really good examples); a high proportion even lack the large basal thinning spall scars
RINK, DUNBAR, DORAN, FREDERICK, AND GREGORY THE HELEN BLAZES SITE 87
OSL Samples Edwards' Stratigraphic Strata Approximate Location
and Select Artifacts Description of Projectile Points
From Our Excavations Found by Edwards
Org anic Matter 7
1Gr ayBlack BIII
Light Gray Sand A
Mottled Gray-Brown Sand ViI
HB1-3 -Heat Treated *
Flake Very Light Brown Sand B
HB1-2 O .....S.nd
Light Brown Sand A FS3 ark rn Snd V
C umnar Dark Blue Gray C
irregularly Prismatic Bue-Gray .....
Light Blue-Green Sandy Clay
with Orange-Yellow Inclusions A
Marine Shells, Clay and Sand
Compact Blue Clay
with Some Marine Shells
Figure 4. Edwards' original photographs of stone tools found at Helen Blazes (Edwards 1954:63B). Figure 3. Edwards' original schematic stratigraphy of the Helen Blazes Site (central part of figure; redrawn from 1987; Dunbar 1991; Dunbar and Webb 1996; Goodyear and
Edwards 1954: 25) shown with respect to the approximate Warren 1972; Knight 2004; Means and Means 2004; Thulman
stratigraphic position of diagnostic artifacts he recovered. 2006, 2008, 2009). Even so, Thulman (2009:244) proposes Letters on point icons refer to Edwards (1954:63B) see Middle Paleoindian materials date between 10,800-10,200 also photograph of Figure 4, stratigraphic position of Ed- 14C B.P. In adjacent areas, Dust Cave in Alabama has ward's artifacts derived from Edwards (1954:76B), and yielded an important chronology showing a transition from
the stratigraphic position of the OSL samples and select late Paleoindian lanceolate points and pseudo-notched points artifacts collected during the recent excavations, beginning around 10,500 14C B.P. (about 12,540 cal B.P.) to
Early Archaic side-notched (Bolen) points after 10,000 14C present in some of the Helen Blazes examples. He also noted B.P. (11,470 cal B.P.) (Sherwood et al. 2004). This chronology, the workmanship manifest in the eastern Florida specimens as will be discussed, is equivalent to the OSL date we obtained he viewed fell well within the broad range of the "Folsom- at 37 cm below datum at Helen Blazes, which is the equivalent like" forms. Finally, he reported that "examples almost of Edward's Level VI that contained the lanceolate and sideduplicating points E, F and J are common in central Florida notched points (Figure 4). This is also supported by the OSL collections" (Edwards 1954:88) and the materials from Helen date from Wakulla Springs (Rink et al. 2012, this volume). Blazes showed a good bit of variability. For his time, Edwards' observation that the point assemblage from Helen Blazes was Optically Stimulated Luminescence Dating:
varied is significant and carries important implications that Physical Basis and Application
will be discussed below.
Optical luminescence dating (OSL) utilizes the effects of Regional Archaeological Context light exposure on radiation-sensitive defects in the lattice of
the quartz mineral. Time dependence arises from a two-stage Many have noted problems with the Florida Paleoindian process in the history of a quartz grain. The first stage is based record: isolated finds, out-of-context materials, few on the observation that during deposition of quartz sunlight
professional excavations, deflated contexts, burial beneath erases previous radiation effects and sets the quartz grain into deep sand strata, and materials found in submerged and a zero-age state. The second step involves the burial of the
saturated contexts (Bullen 1969; Daniel and Wisenbaker quartz to prevent further light exposure. Once buried natural
88 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
by using light to release the stored information during which the quartz is reset to a zero age state, as happened at the time the quartz was exposed to light in nature. Effectively, the quartz acts as a radiation-sensitive photographic film, 60 which must remain in the dark to record an image that can be converted to time (as opposed to photographic film that must
... .....be exposed to light to record the image). The time interval of dark storage (burial time) is obtained by establishing the total dose since the last light exposure (called the equivalent dose in laboratory language) and calculating how much dose was acquired in each year of dark burial. The latter quantity is known as the total dose rate and is the amount of dose each year delivered by the uranium, thorium, and potassium in the environment and the amount delivered downward from the Figure 5. Orthophotograph showing locations of our sky due to cosmic rays. The burial time is simply the ratio of
Shovel Test Units in relation to the geography shown in the dose acquired during dark storage (equivalent dose) to the Figure 4. total dose rate (dose per year). Dose divided by dose per year
gives years since burial.
uranium, thorium, and potassium in the surrounding minerals, In order for OSL to give an accurate burial age in the as well as cosmic radiation from the sky, introduce energy archaeological context, the buried grains must remain in into the quartz. A small proportion of this energy is trapped their burial location relative to artifacts buried with them. If inside the quartz grains. This gives rise to a time-dependent the buried grains move downward relative to the artifacts, accumulation of radiation effects that can be measured in the then grains buried at shallower levels may be found with the laboratory. artifacts, and the grain-age will be too young. Conversely, if
The total dose of radiation (delivered at a steady rate) buried grains move upward relative to their original burial experienced in the burial interval is measured in the laboratory location, they may arrive into locations with artifacts that were
Figure 6. Georeferenced orthophotograph showing our Shovel Test Unit Locations (HB-1, HB-2)
and the location of our Core 5. Also shown are the survey control points GPS-1 and GPS 2
RINK, DUNBAR, DORAN, FREDERICK, AND GREGORY THE HELEN BLAZES SITE 89
buried at a more recent time in the past. In the latter case, the grain-age will be older than the age of artifact burial.
Methods and Sample Locations A &A
Archaeological Methods I 2
Edwards' (1954) fieldwork was well documented, and his site map included two control points, one at the southeast corner of T28S, R36E, Section 31 and the other on the South Indian Fields site. These points and the road and canal features illustrated in Edwards' original site map are shown in Figure 2 . E
and can be seen in Figure 5. These details provided a means to relocate precisely the site and previous excavation units. Prior to going to the field, the Edwards site map was georeferenced and superimposed on modem aerial photographs (Figure 6), which allowed scaling distances and directions for the 4 h
placement of our test pits and tests.We wanted to be close to
the Edwards' units so we could confirm the strata he observed and put a series of Geoprobe cores parallel to the line of 5 mtg
Edwards' test unit blocks. We not only confirmed Edwards' stratigraphic observations, but we recovered a number of artifacts from both test units at levels that matched those of Edwards. N
Site relocation was conducted using the georeferenced map set, 100-m tapes, a compass, and a handheld GPS unit j
with an accuracy of 3 to 10 m depending on the time of day and overhead GPS satellite geometry. The site was relocated in this manner. To gain a higher level of survey accuracy the Figure 7. Photograph of Test Unit 2 showing zone designaSurveying & Mapping Division of Brevard County Public tions, soil Horizons, and boxes i(HB2) showing the locaWorks Department lent the services of survey crew using a tions where samples for soil micromorphology were taken high accuracy GPS receiver in real-time kinetic mode to as blocks.
pinpoint more precisely our work. Two embossed brass survey plates (Control Point GPS 1 and 2) were set into cement to while the geological data were collected in nearby cores and establish two permanent control points (a baseline) for the site Test Unit 2 (HB2). The elevation difference of the land surface (Figure 6). In turn the Brevard County survey crew set both between the two test units was less than 1 cm. The location control points, establishing their horizontal (x, y) and altitude of Geoprobe core 5 is shown on Figure 6. HB 1 was located
(z) coordinates. The survey crew shot in additional flagged about 8 m northwest of Edwards' square 176, and 11B2 was points marking test units and core locations and collected located about 3 m north of his Square K72 (Figure 6). HB1 their data in State Plane coordinates, which we converted to was excavated using a 14" screen mesh from 0 to 25 cm, and UTM, Zone 17 metric coordinate system. The survey data are deeper levels were screened using window screen mesh. HB2 presented in Table 1. was entirely screened using window screen mesh. A small
Our OSL samples were collected in Test Unit 1 (HB 1), water pump was used to facilitate screening with the window
Table 1. GPS coordinates (UTM, Zone 17) and elevations (NAD83-HARN). NE Corner means of the
northeast corner of the Shovel Test. GS means ground surface.
Point Name Easting Northing Elevation (in) Comments
Control Point GPS1 524124.772 3097171.8 14 4.790 DS-70-09-BR27
Control Point GPS2 524056.63 8 3097 169.992 4.735 DS-70-09-BR27
HB1 Shovel Test 524091.204 3097126.239 4.616 NE Corner GS
HB2 Shovel Test 524102.689 3097126.854 4.618 NE Comer GS
Core 1 524 108.545 3097 127.685 4.675 Ground Surface
Core 2 524102.852 3097127.355 4.633 Ground Surface
Core 3 524096.567 3097126.972 4.598 Ground Surface
Core 4 524090.593 3097126.804 4.566 Ground Surface
Core 5 524084.377 3097126.617 4.552 Ground Surface
90 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
screen mesh. Artifacts were collected by arbitrary levels in a second was collected from near the top of Zone 5 (Edwards' HB 1 and natural levels in HB2. All artifacts were recovered Stratum III [the columnar dark blue-gray sandy clay]). from the sand levels of the test units. Bulk samples collected from the monolith and core were
The stratigraphic profiles of these tests (Figure 3) analyzed for sediment texture, loss-on-ignition, and calcium confirmed Edwards' original findings with brown sand carbonate content, and the results are presented on Table 2.
levels (Edwards' Strata IV-VIII) resting on top of blue-gray Sediment texture was determined on a Beckman-Coulter LS sandy clay (Edwards' Stratum III). The clay level impeded 13-320 laser sizer. Samples were first boiled in concentrated excavations, because it perched rainwater, which seeped into (30 percent) hydrogen peroxide and sodium pyrophospate the pits through the porous sand levels and caused the lowest in order to remove organic matter and disaggregate the clay dark brown sand level to collapse into the pit, destroying the minerals after which they were subsequently sonicated for 30 lower portion of the profile wall. This prevented successful seconds prior to analysis. Organic matter was estimated by OSL sampling of the deepest artifact horizon above the blue- loss-on-ignition at 450C for 2 hours, and calcium carbonate gray, sandy clay level. was determined for 17 samples by means of a Chittick
apparatus (a device used to measure the calcium carbonate
Geological Methods content of soil or sediment by reacting it with hydrochloric
acid). The loss-on-ignition values exhibit a slight correlation
The deposits at the site were examined in the field within with the clay content (r-0.60), which suggests that some of HB2 (Figure 7), which exposed Edwards' Strata III through these values may reflect structural water loss and therefore VIII. The perched water table precluded clear examination of may be a poor index of organic matter in clayier deposits. The the lower deposits (Strata I through III), but a more complete micromorphological samples were dried at low temperature stratigraphic section was available by examination of a suite in an oven and then embedded in polyester resin, after which of Geoprobe cores. A 50-cm long monolith was collected they were slabbed on a rock saw and then 2 inch by 3 inch from HB2, and one of the Geoprobe cores was logged, large format thin sections were prepared. This sample was
described, and sampled for physical characterizaton. Two examined under plane and polarized light on a Leica DMEP small oriented samples (i.e., tightly wrapped with 'top' and petrographic microscope. 'bottom' noted) were carved from the profile for petrographic
(soil micromorphological) analysis. One was collected from Optically Stimulated Luminescence Methods the interface between zones 3 and 4 (Edwards' Stratum VI [the
light brown sand] and Stratum V [the dark brown sand]), and OSL samples were collected using steel tubes (5 cm
Figure 8. Photograph showing Shovel Test Unit 1 (HB-1) profiles from which the OSL samples were recovered. HB1-1 was not recovered due to flooding at the time of sampling.
RINK, DUNBAR, DORAN, FREDERICK, AND GREGORY THE HELEN BLAZES SITE 91
EAST WALI SOUTH WALL
.. ... .... ....
.~el en Blazs Test Unit HF3- East & ... South
~Wall Proff|es Showing OSL Sampling Locations
Figure 9. Schematic stratigraphy of Test Unit 2, showing units of varying sediment coloration (light vs. dark).
diameter, 30 cm long) via slide hammer into the walls of HB 1. 24 aliquots to determine a final equivalent dose (DE). Quartz The locations of the two dated samples HB 1-2 and HB 1-3 are grains, between 90-150 microns, were mounted with silicon seen in Figure 8 and in the drawn section (Figure 9). HB 1-1 spray on aluminum discs using a 3 mm and a 1 mm mask and was not collected due to flooding at the time of sampling. The were illuminated for 100 seconds at 125o C. The background depths of the OSL samples are given in Table 3. These samples (the last 4 s) of the OSL decay curve was subtracted from correspond to an upper (HB 1-3) and lower sample (HB 1-2) the "fast" component (first 0.4 s) to determine the samples' in Edwards' Unit VI (Figure 3), though it is possible that the luminescence signal. Only aliquots whose recycling ratios upper sample (HB 1-3) might be within Edwards' Unit VII. were within 10 percent were accepted for equivalent dose DE
All optically-dated samples were processed at the School determinations. Linear plus exponential fits were made to the of Geography and Earth Sciences at McMaster University dose response data to determine individual DE values. under UV-filtered subdued orange light. Pure quartz grains Thermal transfer test and dose recovery test were were obtained using standard OSL preparation methods that performed to determine the final DE preheat temperature include HCl and H202 digestions to remove carbonates and (Madsen et al. 2005). For both tests, twelve aliquots from each organics, respectively, sieving to obtain desired grain size; sample were optically bleached by blue light illumination for heavy liquid separation using lithium polytungstate to remove 40 seconds, followed by a 10,000 second pause and another heavy minerals and feldspars; HF digestion to remove the outer 40-second illumination. For the dose recovery test, the alpha affected layer; a second H202 digestion to remove any aliquots were given a known dose. Both tests continued with remaining feldspars and any fluorides that may have formed the standard SAR protocol, except the preheat temperatures during the HF digestion; and finally resieving to remove any varied (160, 200, 240, 2800 C), with 3 aliquots from each grains that no longer fall in the desired size range (90-150 sample receiving a different preheat temperature. For each microns). sample, the dose recovery test was used to determine which
Dose rates were based on neutron activation analysis preheat temperature produced a D B closest to the given dose. (NAA) of 232Th and 40K and delayed neutron counting Once this preheat temperature was determined the thermal
(DNC) analysis of 23 8U (conducted at the McMaster transfer test was analyzed to insure there was no induced
University Nuclear Reactor). Conversion of radioisotope charge transfer at that given temperature. concentrations to dose rates was done using the data of A feldspar contamination check was also performed on
Adamiec and Aitken (1998) with the addition of 235U dose each sample to insure purity of the quartz grain separates. An rates based on its expected natural abundance in relation to initial DE was estimated by comparing the natural OSL signal 238U (1 part 235U to 137 parts 238U). Untreated subsamples (preheat T =2000 C) of three aliquots to the regenerated OSL of the original samples were used to determine the elemental given by a single dose. A second identical regeneration dose concentrations of radioactive 238U, 232Th and 40K (Table was applied to the same aliquots and the infrared stimulated 4). NAA/DNC-based dose rates were calculated assuming luminescence signal (IRSL) signal was measured. If a ratio
radioactive equilibrium in the 238U and 232Th decay chains, of IRSL to regenerated OSL signal was less than 100 for
Luminescence measurements were conducted on a RISO all aliquots, it is assumed there is no significant feldspar
OSL/TL-DA- 15 reader using blue light LED stimulation (470 contamination. nm) and a 7 mm-thick Hoya U-340 filter (270-400 nm). A Moisture contents were measured in the lab from the
calibrated 90Sr beta source was used to perform laboratory recovered sediment and used for the dose rate calculations. irradiations. The single aliquot regeneration (SAR) protocol The moisture content of the sample can vary during the burial (Murray and Wintle 2000) was conducted on a minimum of history, thus two estimates were used in the age calculations.
Table 2. Sedimentological Data for Helen Blazes
Sample Depth Sand Silt Clay Mean Median Sorting Skewnes Krtis LTexua
Number (cm) () %) %) (phi) (phi) (Phi) (phi)(p)(%Cls
Monolith 1 1 92.7 6.21 1.09 2.24 2.20 1.04 0.181.250Sn
2 3 95.1. 3.97 0.93 2.01 1.98 0.91 0.14 12 .Sn
3 5 94 4.99 1.01. 2.00 1.96 1.01 0.19 13 .4Sn
4 7 92.5 6.37 1.13 2.12 2.05 1.07 0.26 14 .1Sn
5 9 94.4 4.56 1.04 2.00 1.97 0.95 0.17 12 .0Sn
6 11 95.3 3.77 0.931 2.02 2.00 0.90 0.13 11 .7Sn
7 13 95.5 3.5 1 1.99 1.96 0.89 0.13 11 .9Sn
8 15 96.4 2.56 1.04 2.02 2.01. 0.83 0.08 10 .5Sn
9 17 96.2 2.81 0.99 2.02 2.02 0.85 0.07 10 .6Sn
10 19 96.2 2.73 1.07 2.04 2.03' 0.84 0.09 10 .4Sn
11 21 96.1 2.8 1.1 2.06 2.05 0.86 0.07 10 .5Sn
12 23 96.4 2.57 1.03 2.08 2.07 0.83006 11 .0Sn
13 25 96.3 2.63 1.07 2.07 2.06 0.84 0.07 10 .7Sn
14 27 95.4 3.4 1.2 2.12 2.1.1 0.89 0.08l1'0.2Sn
15 29 952.3 3.46 1.24 2.08 2.07 0.90 0.09 11 .3Sn
16 31 95.4 3.39 1.21 2.09 2.09 0.88 0.08 11 .4Sn
17 33 96.2 2.72 1.08 1.97 1.96 0.86 0.09 10 .9Sn
18 35 95 3.73 1.27 2.06 2.04 0.92 0.12 11 .4Sn
19 37 93.4 4.93 1.67 2.15 2.13 1.09 0.23 16 .0Sn
20 39 90.4 7.21 2.39 2.33~ 2.27 1.35 0.3 4 21.2Sn
21 41 86.5 10.44 3.06 2.43 2.28 1.55 0.422.030 LaySn
22 43 84.9 13 .71 1.39 2.63) 2.46 1.33 0.30181.7 LaySn
23 45 89.2 9.71 1.09 2.36 2.26 1.19 0.23 12 .1Sn
24 47 89.4 9.57 1.03 2.39 2.28 16 0.26 13 .6Sn
25 49 89.1 9.88 1.02 2.45 2.34 1.14 0.25 13 .2Sn
26 50.5 88.7 10.31 0.99 2.45 2.34 1.18 0.25 13 .4Sn
Table 2Sedimentological Data for He1len Blazes, continued.
Sample Depth Sand Silt Clay Mean Median Sorting Skewnes Kuoi LO USATxra
Number (cm) (0%) (0%) (% ) h) (phi) (phi) (phi) (phi)(0)Cls
Core 1 2.5 93.4 5.44 1.16 2.13 2.11 1.02 0.19 14 .0Sn
2 7.5 94.5 4.38 1.12 2.07 2.05 0.94 0.16 lA .4Sn
3 12 96.8 2.26 0.94 1.95 1.93 0.82 0.09 10 .1Sn
4 17 96.1 2.82 1.08 2.03 2.02 0.86 0.07 10 .1Sn
5 22.5 9 4. '3 4.18 1.52 2.1 2.09 0.97 0.16 13 .5Sn
6 27.5 93.8 4.59 1.61 2.14 2.12 1.02 0.19 14 .0Sn
7 31.5 94.8 3.83~ 1.37 2.05 2.04 0.94 0.12 12 .6Sn
8 36 93 5.76 1.24 2.11 2.08 1.04 0.20 14 .6Sn
9 41 84.3 10.93 4.77 2.65 2.36 1.80 0.49 25 .8 LaySn
10 44.5 88.6 6.28 5.12 2.49 2.38 1. 63 0.4429716'Sn
11 48 91.4 3.59 5.01, 2.47 2.42 1.60 0.37 32 .8Sn
12 52.5 93.3 3.46 3.24 2.18 2.13 1.33 0.35 22 .7Sn
13 57.5 89.9 5.47 4.63 2.48 2.41 1.60 0.38 25.7Sn
14 6'.3.5 92.6 3.99 3.41 2.39 2.45 1.46 0.06 21 .4Sn
15 70 86.7 7.69 5.61 2.531 2.44 1.72 0.41 30 .4 LaySn
16 75 86.8 7.84 5.36 2.55 2.43 1.65 0.44 31 .6 LaySn
17 92.5 80.4 14.48 5.12 3.20 2.74 2.03 0.50 25 .6 LaySn
18 110.5 33. 60.21 6.49 4.86 4.99 2.57 0.00 09 .0Sl~a
19 115.5 86.6 6.76 6.64 2.17 1.91 1.86 0.583.510Lom an
20 120.5 88.4 5.43~ 6.17 2.06 1.93 1.70 0.50 42 .1Sn
21 125.5 92.9 3.08 4.02 1.80 1.79 1.38 0.37 35 .2Sn
22 130.5 -32.9 40.6 26.5 6.43 6.70 3.17 -0.05 06 .9La
23 135 69.8 18.6 11.6 4.00 2.46 2.95 0.740.813Sadom
24 137.5 8.5 10.88 5.62 2.84 2.42. 1.68 0.71 37 .1LaySn
25 138.5 69.5 19.4 11.1 4.17 2.81 2.88 0.661.109Sadom
26 139.5 83.5 10.88 5.62 3.14 2.77 1.54 0.45140.5 LaySn
27 140.25 69.5 19.4 11.1 4.60 4.02 2.42 0235 07.4 LaySn
28 140.75 83.5 10.88 5.62 3.38 3.02 1.55 0.46 13 .2 LaySn
29 141.5 69.5 19.4 11.1. 3.30 2.81 1.84 0.471.0.9 Sadom
30 143.5 74 17.03 8.97 3.95 2.74 2.58 0.70 12 .5 SnyLa
31 146 0 56.8 43.2 8.66 8.62 1.90 0.0309 66 it~a
32 147.5 80.5 12.51 6.99 3.28 2.65 1.95 0.72 871.6 LaySn
94 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
............... ..... ................ .... i
i,:,: .......................................................:: : ,: i,: : ,, . .. ................ ..... ..: i
Conia Cor F3 O 20
FS2 FS6 FS7
M e t rs Figure 11. Photographs of flakes recovered in our excavations (compare with Table 5). FS1 is the heat-treated flake.
RINK, DUNBAR, DORAN, FREDERICK, AND GREGORY THE HELEN BLAZES SITE 95
Table 3. Dose Rate and Optically Stimulated Luminescence (OSL) Age Results for Helen Blazes.
Sample Depth Moisture Minimum Cosmic Dose Beta Gamma Total OSL Age
(cm) (%) Age Model Rate (1 X 10-6 Dose Dose Rate Dose Minimum
DE (Gray) Gray per Rate (1 x 10-6 Rate Age Model
1 sigma year) (1 X 10-6 Gray per (1 X 10-6 (ka)
max value Gray per year) Gray
HB1-3 17.5- 7 3.5 + 0.2 258 1 10 163 +12 167 + 12 599+ 18 5.8 0.4
HB1-3 30 25810 134+10 14010 53214 6.50.5
HB1-3 5.4 to 7.0
HB1-2 34.5- 7 5.6 0.8 238 10 145 12 205 21 598 24 9.4 1.4
HB1-2 30 23810 11910 171 18 53920 10.41.6
HB1-2 9.0 to 12.0
Table 3. Footnotes
1. Beta and Gamma Dose rates determined using 238U, 232 Th and K concentrations determined using delayed neutron
counting (238U) and instrumental neutron activation analysis (232Th and K) at the McMaster University Nuclear
Reactor. Conversion to dose rates were done assuming equilibrium in the 238U and 232Th decay chains. Conversions
were done assuming that 235U was present in its expected natural ratio of I part 235U to 137 parts 238U. Dose rate
conversion data were obtained from Adamiec and Aitken (1998). Total dose rates include assumed quartz internal
238U and 232Th values from Rink and Odom (1991) of 238U 0.0665 ppm and 232Th 0.1135 ppm. These values
yielded dose rates of 8.6 +/- 2.2 and 1.9 +/- 0.4 x 10-6 Gray per year respectively, using an alpha efficiency factor of 0.04. Concentrations of 238U, 232Th and K used to calculate the beta and gamma dose rates were HB 1-3 0.6 +/- 0.1 ppm, 1.92 /- 0.13 ppm and 0.0773 wt.% respectively. 238U, 232Th and K concentrations for HB 1-2 were 0.5 +/- 0.1 ppm, 1.92 -0.13 and 0.0675 wt% respectively. For HB 1-2 we were able to measure the in-situ gamma dose rate, rather than using sediment radioactivity data; the in-situ dose rate is reported in the table and used to calculate the
ages for that sample. Age results and dose rate data were obtained using the software ANATOL provided by N.
Mercier, which uses cosmic dose rate calculations based on Prescott and Hutton (1988). As-found moisture contents
were between the 7 and 300% estimates used here to bracket time averaged moisture.
Table 4. Concentrations of radionuclides in Htelen Blazes OSL samples
Sample 28Uranium (ppm) 22Thorium (ppm) Potassium (wt. %)
HBi -3 0.6 +/- 0.1 1.92 +/- 0.13 0.0773 +7- 0.0030
HBi -2 0.5 +7- 0.1 1.92 +/- 0.13 0.0675 7- 0.0003
96 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
Higher average moisture contents yield older ages, while to Early Archaic tools (Daniel and Wisenbaker 1987; Dunbar lower average moisture contents yield younger ages. This and Vojnovski 2007). The debitage from the lower component results from changes in the beta and gamma dose rate that are (25 cm to 52 cm) are small pressure flakes consistent with affected by moisture content. both Suwannee and Bolen toolkit manufacture and, in this
Results instance, with careful stone tool maintenance. Generally
larger conical core tools, conical cores, and gravers have been
Archaeological Results recovered at Suwannee point sites such as the Norden site
The two one-meter square test units at the Helen Blazes (8GI40) in northern Florida (Dunbar and Vojnovski 2007) and (HB 1 and HB2) yielded 14 lithic artifacts, specifically 3 tools the Suwannee-Bolen point component at the Harney Flats site and 11 tool maintenance debitage flakes. Their characteristics (8H1507) near Tampa (Daniel and Wisenbaker 1987). Both the are given in Table 5 and illustrated in Figures 10 and 11. This Norden and Harney Flats sites are in areas of abundant chert is not a large number of artifacts, though they are informative, outcrops, and chert conservation measures are absent. The relative position of two of the artifacts we collected, Generally, when chert sources are located near a site, Edwards' major finds, and the locations of the OSL samples debitage tends to be more abundant and composed of larger are shown in Figure 3. and bulkier pieces than when chert sources are more distant
The stratigraphic positions of the artifacts recovered from a site. In the latter case, the debitage tends to be less indicate two lithic components are present at the site. In the abundant and of smaller size. An accepted inference garnered upper component we recovered a single heat-treated flake from the presence of fewer and smaller sized debitage is from the shallowest level of the test units (19 to 20 cm below that people were more careful and less wasteful in their use surface), which is consistent with Middle Archaic and later of chert as they had to walk a long distance to obtain it. The period sites and corresponds to Edwards' Stratum VIII where most striking aspect of the debitage assemblage recovered he recovered stemmed-Archaic points. The lower component during our investigations is its small size. Had regular 1 produced a conical core tool (FS 3.1 at 52 cm below surface) screen been used, almost nothing would have been recovered. and a graver (FS 10.1 at 30-40 cm below surface). Although Window screen was essential for recovering the smaller range they are not diagnostic, these fit nicely with similar Paleoindian of material. Given the exhausted condition of the points
.:A_ :. .
Figure 12. Plot showing th reut ofk the pyiapretIofCr5(uprat)ndam olhcletdfomTs
2 (lwerpar) a th Heen lazs Ste. he elaivestrtigaphc psiton f th tw OS saple coleced romTes
1 ar project ont th tairpi rwn fTs .Nt httesae o h etr igasaenttesm
on.the.two drawings w ith:::i:::;:;;+;: Test.. '.................. 2 ha iga:x a de cl oso v rainpee t ihnteu c nsld tdsn s
RINK, DUNBAR, DORAN, FREDERICK, AND GREGORY THE HELEN BLAZES SITE 97
Table 5. Finds List from 2009 Excavation at Helen Blazes, giving depths of finds and type of material. ST
identifies the shovel test that object was recovered from. HT heat treated lithic, NHT-not heat treated
FS No. Count Item(s) Comments
1.1 1 Debitage HT, Ocala chert, 19-20cm, ST 1
2.1 1 Debitage NHT, Chert?*, 35-40 cm, ST-1
3.1 1 Conical Core tool NHT, Chert?, 52 cm, ST-1 on top of clay
4.1 1 Plastic shotgun wad NHT, 0-5 cm
5.1 1 Utilized flake, micro tool NHT, Ocala chert, 25-30, ST-1
6.1 2 Debitage NHT, Chert ?, 35-40, ST-1
7.1 1 Debitage NHT, Chert?, 40-45, ST-1
8.1 NO Artifacts *1
9.1 1 Debitage NTH, Chert?, ~30cm, ST-2
10.1 1 Graver, micro tool NHT, Chert?, 30-40 cm, ST-2
10.2 2 Debitage NHT, Peace River Opaline 30-40 cm, ST-2
10.3 3 Debitage NHT, Chert?, 30-40 cm, ST-2
*Chert? chert, unidentified source
* 1FS numbers assigned in field based on strata prior to excavation and screening
Table 6. Characteristics of Equivalent Dose (DE) Measurements at Helen Blazes
Sample Aliquot No. of aliquots Minimum Age p-value Overdispersion
Diameter accepted Model DE (Gray) (%)
(mm) +/- 1 sigma
HB1-3 1 22/48 3.5 + 0.2, 0.2 0.045 200
HB 1-2 1 30/48 5.6 +0.5, 0.8 0.033 50
recovered by Edwards and the small size of the debitage we here adds little to his description, but our work does provide recovered, it appears that the occupants of Helen Blazes used the basis for an alternative interpretation of the geological and resharpened tools in a way that conserved raw material, events that led to the formation of these deposits. In general which is in agreement with our expectations for a site that is terms, we believe that the sediments preserved at Helen Blazes distantly located from chert resources. represent a single marine deposit that has been significantly
As shown in Table 5 the origin of identified chert sources altered by post-depositional weathering and pedogenesis. were from quarry clusters well over 100 km to the northwest Many of the attributes Edwards used to delineate geologic and southeast. Ocala chert outcrops in the Brooksville area events are post-depositional pedogenic features and not true about 150 km to the west-northwest of Helen Blazes, although, sedimentary strata. When the profile is closely examined for if one were to assume the most logical movement corridor was soil features, it becomes apparent that this profile developed the St. Johns River basin, the chert source near Ocala about under two sequential climatic phases that were significantly 175 km to the northwest may have been the nearest source. different from each other. The occurrence of Peace River opaline chert (about 135 km Primary deposition of this marine sediment most likely
southwest) is interesting because its practicality as a lithic occurred during the last interglacial (approximately 80,000 tool-making resource is marginal at best. Peace River opaline to 125,000 years ago). Falling sea levels during the ensuing chert is glassy, brittle, and prone to easy breakage compared glaciation and the drier climate and depressed water tables to the Ocala and other Florida Tertiary cherts. But the real resulted in a freely draining sandy deposit that favored the significance of opaline chert is that it is an indication of an development of a prominent alfisol within the sediments. early mobility route from the Peace River basin of southwest During this period of time clay was stripped from the upper Florida, where early sites such as Little Salt Springs and Warm part of the profile and concentrated in Zones 5 and 6 (Figure Mineral Springs are located, on the south-central Florida East 12) at depth to form an argillic (Bt) horizon (Edwards' Stratum Coast. III). Here, "at depth" means this was an internal reorganization
at the depths of Zones 5 and 6 (70 to 135 cm below surface) Geological Results rather than a "normal" geologic sedimentation where sediment
is added to the ground surface. Thus, clay was stripped from Edwards' thorough and detailed description of the Helen the top of the profile (Zones 1 to 4) and translocated into what Blazes profile is remarkable for its time. The work performed we recognized as Zones 5 and 6. At the same time, detrital
98 THE FLORIDA ANTHROPOLOGIST 2012 VOL. 65 (1-2)
Figure 13. Photomicrographs of selected features of the deposits at the Helen Blazes Site. A. Photomicrograph of the Ehorizon (Zone 3) showing the subrounded quartz grains and a lack of grain coats. B. Photomicrograph of the Bh horizon (Zone 4), showing the coats and bridges between grains of illuvial organic matter. Square shows area enlarged in the next photo. C. Same slide as B enlarged to show the coats of organic matter. D. Photomicrograph of the Btg horizon (Zone 5) showing the coats of illuvial clay, light brown color (plane polarized light). E. Same photo as D but shown in cross-polarized light. Note the birefringent clay coats around grains. F. View of superimposed coats along the margin of a channel (CH), where a clay argillan lies adjacent to the quartz (Q) sand grain and is in turn coated by illuvial organic mater (OM) showing the relative sequence in which the coats were deposited.
calcium carbonate was leached from the upper part of the that aggressively attack both the framework minerals as well profile and deposited at depth (i.e., reorganization in Zones 7, as grain coats leading to sandy epipedons that are generally 8, and 9 [135 to 240+ cm below surface in core 5], Edwards' very light in color and subsoils with illuvial concentrations of Strata I and II) in the form of small calcium carbonate nodules organic matter, iron, and aluminum. It was during this period and calcium phosphate nodules. It is important to note that (most likely the latter half of the Holocene) that Zone 4 (or the depths of these deposits vary across the site. The results Edwards' Stratum V; a Bh horizon; see Figure 12) was created, of these two processes were the formation of a profile with a which Edwards correctly identified as a Spodic B horizon sandy upper part (Edwards' Strata V to VIII) and a structured consisting of illuvial organic matter. At the same time, the subsoil with increased amounts of clay and silt (Edwards' prolonged saturation gleyed the argillic horizon and all of the Stratum III) on top of relatively unweathered sediment to which underlying deposits. secondary (pedogenic) calcium carbonate and phosphate were The cultural material within the unstructured sands in added (Edwards' Strata I and II). The latter retain primary the top half meter or so, above Edwards' Stratum III, was stratification and detrital marine shell and unlike the upper part most likely deposited on the ground surface and buried by of the profile still bear a strong resemblance to their appearance pedoturbation. Prior to the formation of the argillic horizon, immediately following deposition in a marine setting. pedoturbation by soil fauna (especially insects) would most
The onset of more mesic conditions, rise of the water likely have been in excess of 1 m, but as the argillic horizon table and periodic inundation that accompanied deglaciation became established and started to form an impediment to water significantly shifted pedogenesis here in favor of Spodosol movement, their actions would have become concentrated development. Spodosols in Florida are created by periods in the loose sands above it. This process was most likely of saturation that submerge acidic leaf litter and lead to the well underway by the time the site was inhabited during the development of strongly acidic organic compounds (chelates) Paleoindian period.