Lithic reduction, group mobility, and settlement along Cowhouse Creek, Hillsborough County, Florida

Lithic reduction, group mobility, and settlement along Cowhouse Creek, Hillsborough County, Florida

Material Information

Lithic reduction, group mobility, and settlement along Cowhouse Creek, Hillsborough County, Florida
Torp, Lyle Carlton
Place of Publication:
Tampa, Florida
University of South Florida
Publication Date:
Physical Description:
xi, 224 leaves : ill. ; 29 cm.


Subjects / Keywords:
Indians of North America -- Antiquities -- Florida -- Hillsborough County ( lcsh )
Quarries and quarrying, Prehistoric -- Florida -- Hillsborough County ( lcsh )
Dissertations, Academic -- Applied Anthropology -- Masters -- USF ( FTS )


General Note:
Thesis (M.A.)--University of South Florida, 1995. Includes bibliographical references (leaves 203-224).

Record Information

Source Institution:
University of South Florida
Holding Location:
Universtity of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
33939261 ( ALEPH )
021226868 ( OCLC )
F51-00120 ( USFLDC DOI )
f51.120 ( USFLDC Handle )

Postcard Information



This item is only available as the following downloads:

Full Text


LITHIC REDUCTION, GROUP MOBILITY, AND SETTLEMENT ALONG COWHOUSE CREEK, HILLSBOROUGH COUNTY, FLORIDA by Lyle Carlton Torp A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts in the Department of Anthropology in the University of South Florida April 1995 Major Professor: J Raymond Williams, Ph.D.


Graduate Council University of South Florida Tampa, Florida CERTIFICATE OF APPROVAL MASTER'S THESIS This is to certify that the Master's Thesis of Lyle Carlton Torp with a major in the Department of Anthropology has been approved by the Examining Committee on November 17, as satisfactory for the Thesis requirement for the Master of Arts degree. Thesis MaJoT PLDfessor: J,. Ravmond Williams, Ph.D. White, Ph.D.


Lyle Carlton Torp 1992 G)---------------------All Rights Reserved


various aspects of lab work. Bob Sellers shouldered the responsibility of running the lab after the completion of the SWFWMD work, and deserves praise for his assistance. Thanks are due to Brian Evensen and several other collectors for access to their collections, and for sharing their thoughts about Paleo-Indian occupations along Cowhouse Creek. A debt of gratitude must also be extended to Richard Estabrook, who patiently answered my questions about lithic analysis, and to Ed McCabe, who helped organize the computerized databases. I would like to extend special thanks to Dr. J. Raymond Williams for his encouragement and professional guidance. As the Principal Investigator, Ray has offered his guidance while still allowing me great latitude as to the course of this project. His contribution as my thesis advisor to the preparation of this manuscript through its countless revisions can not be understated, and his friendship and patience are greatly appreciated. Drs. Nancy White and Roger Grange, Jr. have also been instrumental in the development of this work as members of this thesis committee. Several others have shared their thoughts on various aspects of Hillsborough River Basin research, including Joan Deming, Dana Ste. Claire, Dr. Sam Upchurch, Randy Daniel, and Dr. Al Goodyear. Maggie Goetze, Susanne Marcus, and many others have contributed in countless ways, providing time and interest in the project, and are owed a debt of thanks. iii


ACKNOWLEDGEMENTS Countless persons have contributed time and effort during the year-and-a-half period from the excavations at 8-Hi-495 and 8-Hi-496 to the writing of this work. The Southwest Florida Water Management District provided funding for the excavations and logistical support throughout the grant period. Ken Kramer, the SWFWMD Project Manager, deserves a great deal of credit for his contribution to the success of the initial project, and his enthusiasm for our work. The field crew consisted of the 14 students enrolled in a University of South Florida Summer Field School i n archaeology: Jonathan Bloom, Douglas Hattaway, Rebecca Kane, Aziz Mammad, Kari North, Dawn Potvin, Nola Ravner, Laura Rounds, Robert Sellers, Susan Steakley, Julie Swann, Pam Vojnovski, Laurie Watson, and Anna Ruth Worten. Terrence Simpson, John Darsey, and Steve Gouldman contributed to the fieldwork. Mike Garner co-supervised the field activities. The laboratory phases of the lab work and staff assisting on various analysis included Robin Van Auken, Jay Germano, Doug Hattaway, Beverly Hungerford, Dawn Potvin, Robert Sellers, Susan Steakley, Pam Vojnovski, and Nancy Winterbauer, although many others contributed to ii


TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES ABSTRACT CHAPTER 1: INTRODUCTION CHAPTER 2: CULTURAL SEQUENCE OF THE CENTRAL PENINSULAR GULF COAST ARCHAEOLOGICAL REGION Paleo-Indian Period Archaic Period Post-Archaic Periods CHAPTER 3: ENVIRONMENTAL SETTING Prehistoric Environment Environmental Change Sea Level Change Contemporary Environment CHAPTER 4: PREVIOUS ARCHAEOLOGICAL INVESTIGATIONS vi vii ix 1 8 9 19 27 36 37 39 41 49 IN THE VICINITY 55 CHAPTER 5: RESEARCH BACKGROUND Physiography of Cowhouse Creek Geology Soils Lithic Resources Hunter-Gatherer Models CHAPTER 6 : THE COWHOUSE EAST HEAD AND WEST HEAD PROJECT Field Methodology Laboratory Methodology Lithic Analysis CHAPTER 7: RESULTS AND INTERPRETATIONS Determination of Time Determination of Function Discussion of the Relationship between 8-Hi-495 and 8-Hi-496 CHAPTER 8: PREHISTORIC QUARRYING BEHAVIOR iv 74 74 76 78 84 88 93 93 103 104 114 115 125 133 167




LIST OF TABLES TABLE 1. Raw Material Distribution, 8-Hi-495 121 TABLE 2. Site Profile, 8-Hi-495 128 TABLE 3 Site Profile, 8 -Hi-496 132 TABLE 4. Indices of Early Stage Reduction and Biface Discard, 8-Hi-495 and 8-Hi-496 158 vi


LIST OF FIGURES FIGURE 1. Detail of Hillsborough County with the Location of the Cowhouse East Head and Cowhouse West Head Project Area FIGURE 2. Early Prehistoric Chronologies in Florida FIGURE 3. Late Pleistocene/Early Holocene Sea Levels (from Daniel and Wisenbaker 1987:153) FIGURE 4. Investigated Sites in the Vicinity of the Cowhouse East Head and Cowhouse West Head Project Area FIGURE 5. Location of 8-Hi-495 and 8-Hi-496 Plotted on the U.S.G.S. Thonotosassa Quadrangle, 1987 FIGURE 6. Stratigraphic Sequence at 8-Hi-495 FIGURE 7. Quarry Clusters Throughout Florida (after Upchurch et al. 1981:11) FIGURE 8. Cowhouse East Head and Cowhouse West Head Project Area FIGURE 9. Site Map, Cowhouse East Head FIGURE 10. Site Map, Cowhouse West Head FIGURE 11. Biface Manufacturing Key (after Estabrook and Newman 1984:75) FIGURE 12. Sample of Middle Archaic Points from EUs 1, 2, and 5, 8-Hi-495 FIGURE 13. Possible Suwannee Preform from 8-Hi-495, EU 2, Level 12 FIGURE 14. Oblong Unifacial Scraper from 8-Hi-495, EU 2, Level 12 vii 2 20 42 61 75 81 87 94 97 100 110 117 118 118


FIGURE 15. Biface Distribution by Level, 8-Hi-495 FIGURE 16. Biface Distribution by Excavation Unit, EUs 1, 2 and 5, 8-Hi-495 FIGURE 17. Blanks from the Cowhouse West Head Site FIGURE 18. Blanks from the cowhouse East Head Site FIGURE 19. Preforms from the Cowhouse East Head Site FIGURE 20. Archaic Stemmed Points Recovered from EU 9, 8-Hi-496 FIGURE 21. Debitage Comparisons by Number, 8-Hi-495 and 8-Hi-496 FIGURE 22. Debitage Comparisons by Weight, 8-Hi-495 and 8-Hi-496 FIGURE 23. Debitage Comparison of Material With Cortex, by Number and Weight FIGURE 24. Debitage Comparison of Material Without Cortex, by Number and Weight FIGURE 25. Mean Flake Weight by Unit, 8-Hi-495 and 8-Hi-496 FIGURE 26. Mean Flake Weight: Material With Cortex, 8-Hi -495 and 8-Hi-496 FIGURE 27. Mean Flake Weight: Material Without Cortex, 8-Hi-495 and 8-Hi-496 FIGURE 28. Silicified Coral Blank from 8-Hi-495, EU 1, Level 3 viii 123 124 137 139 140 146 153 154 155 156 162 163 164 189


LITHIC REDUCTION, GROUP MOBILITY, AND SETTLEMENT ALONG COWHOUSE CREEK, HILLSBOROUGH COUNTY, FLORIDA by Lyle Carlton Torp An Abstract Of a thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts in the Department of Anthropology in the University of South Florida April 1995 Major Professor: J. Raymond Williams, Ph.D. ix


This thesis expands on the results of archaeological investigations conducted at the Cowhouse East Head (8-Hi-495) and Cowhouse West Head (8-Hi-496) sites, in Hillsborough County, Florida. Excavations were conducted by a University of South Florida summer field school in archaeology during the summer of 1990, and were funded by a grant from the Southwest Florida Water Management District ( SWFWMD) Initial recording of the site indicated both Paleo-Indian and Archaic period components. Analysis of recovered material supported this hypothesis, and although no diagnostic Paleo-Indian points were recovered, evidence has been slowly developing indicating that the immediate area was important for Paleo-Indian as well as Archaic period groups. It is suggested here that resources available in this area, particularly the abundant chert sources along Cowhouse Creek, allowed for the repeated visitation of the area by prehistoric populations. Results of excavation at the two sites support the original designation of site functions as a base camp (8-Hi-495) and a quarry site (8-Hi-496). It is suggested here that base camps along Cowhouse Creek during these time periods were particularly oriented toward the procurement of stone for tool production. Other critical resources such as wood, water, and faunal and floral materials would have been available from the Paleo-Indian through Archaic periods, and would have allowed for relatively long-term habitation of sites in this area. The use of quarry sites as the basis X


for settlement studies is emphasized, as lithic scatters are the dominant class of site in the area. Lithic industries therefore, when viewed in their totality, should be the most productive manner of assessing several of the larger and more complex regional problems. Abstract approved: Major Professor, J. Raymond Williams, Ph.D. Professor, Department of Anthropology Date of Approval xi


CHAPTER ONE INTRODUCTION 1 This thesis develops and expands on the results of an archaeological survey and excavation of two sites along opposite sides of Cowhouse Creek in Hillsborough County, Florida (Figure 1). The excavations were conducted as part of a University of South Florida archaeology field school during the summer of 1990. Looting at the sites was the initial impetus for these investigations, and the collection of baseline data from sites impacted by this destructive activity was viewed as the first step in long-term responsibility for archaeological sites under SWFWMD control. Many years of looting at the sites have had a significant impact on the area from an archaeological viewpoint as well as making it a safety hazard and eyesore. The primary objective of the excavations was to determine the nature and extent of the sites, which would act as a basis for future research along Cowhouse Creek and continuing research in the Hillsborough River Basin. These were discussed in the resulting report to the Southwest Florida Water Management District (Torp 1991). Several artifact collectors from the area hav e frequented the Cowhouse sites as well as several other loci


FLORIDA GULF O F 1'\EXJC O HILLSBOROUGH COUNTY ATLANTJC OCEAN 0 PASCO 1'\ANATEE I 0 K ilomtlen SCALE ,POLK I I I I I I I I g l -----------------------+---, : I 2 FIGURE 1. Detail of Hillsborough County with the Location of the Cowhouse East Head and Cowhouse West Head Project Area.


3 along Cowhouse Creek over a long period of time. Some of these collections were made available to us, and they suggest that the entire drainage basin for Cowhouse Creek contains Paleo-Indian material. While education and public awareness of the goals of professional archaeology were important aspects of the fieldwork, gaining insight into the problems of looting as well as means of archaeological site protection and the prevention of vandalism were important objectives as well. Recommendations were discussed, in terms of mitigating damage to the sites and preventing looting activity, as well as educational and law enforcement needs. Research goals focused primarily on understanding the nature of lithic procurement and quarrying activities that could be understood archaeologically, and better understanding the role of these types of sites within the context of regional systems. As there have been only limited investigations of quarry sites (see Chance 1980, 1981, 1982; Chance and Misner 1984; Purdy 1975, 1977, 1981), there is a corresponding lack of comparative data and information regarding activities at these.types of sites. Furthermore, little attention has been given to the relationship of these types of sites to other sites. The investigations were also geared toward the testing of many of the lower and wetter areas along the slough which have been traditionally ignored until recently, with an acknowledgement of the biases of conventional survey


4 methodologies that deemphasized "low probability areas" and a recognition of the importance of wet-site archaeology. Although it was hoped that the testing of these areas would allow for the recovery of materials that normally would not be preserved in the acidic soils of this region, this was not the case. Even with the flotation of larger samples from the swampy areas, little faunal and floral material was recovered. These investigations did, however, allow for a better understanding of lithic procurement along the edges of the modern wetlands to develop, as worked lithic material was recovered in areas that are normally under water for most of the year. The Cowhouse East Head site is believed to be a multicomponent site, with indications of use as a base camp during Paleo-Indian and Middle Archaic times. Although no diagnostic Paleo-Indian points were recovered during the excavations, private collections were studied. Sufficient information was recorded by these informants to allow for the vertical isolation of the Paleo-Indian component. The Archaic component is represented by several Newnan points and variant forms. there were no occupations. There were no ceramics 'at the site, and indications of other (post-Archaic) The Cowhouse West Head site also apparently contains both Paleo-Indian and Archaic period components, although too few diagnostic tools were recovered to allow for an adequate temporal assessment. Again, no ceramics were


5 recovered. Cowhouse West Head functioned primarily as a lithic procurement (quarry) area, along with the lower, wetter area of the slough between the two sites. Activities at this site were found to be quite different from activities at Cowhouse East Head, although the temporal/cultural affiliations are believed to be the same. Quarries are an important source of information concerning the temporal and spatial/geographic relationships between sites. Quarries are structurally different than other sites; the primary material obtained is a temporally and spatially fixed resource (chert), and these sites constitute the often ignored incipient stages of the lithic tool industry. This is particularly important in regions such as the Central Peninsular Gulf Coast region, where lithic remains compose the dominant artifact class from whic h interpretations may be made concerning past cultural systems. It is suggested that as an indicator of group mobility, quarry sites offer valuable information regarding collecting versus foraging systems. The most striking aspect of quarry sites in this area appears to be that the procurement of lithic material was far less an opportunistic endeavor than has been assumed, lending support to the idea that Florida Paleo-Indian groups were much less geographically mobile than their counterparts to the north (see Goodyear, in Daniel and Wisenbaker 1987:iv). Despite some other evidence suggesting an increasing tendency toward sedentism during


6 the Archaic period (such as population growth, an increase in the size and number of settlements, etc.), it appears that groups during both Paleo-Indian and Archaic periods tended to utilize a collecting strategy to a greater extent than a foraging one (though strategies may have varied dramatically on a seasonal basis). Addressing the relationships of contemporaneous sites is difficult, particularly given the inability to obtain dates for local sites and tool types in general. In earlier periods especially, phases based on typological development are comparatively long and it is difficult, if not impossible, to determine how frequently settlements moved within them. The utilization of quarry sources as the base of study, particularly as this is the class of data most often available, is the most effective method of determining behavioral trends in settlement during a given cultural period, and as such, should be afforded increased research efforts. Certainly, quarry assemblages deserve as much attention as sites containing products of later stages of lithic reduction. Despite the fact that any understanding of behavior at these sites must first be considered as a result of the site's setting (geologically and geographically) behavioral patterns at quarry sites can be discerned. These understandings may be most useful in the contemplation of relationships between sites throughout a geographical region, and it is concluded that the investigation of quarry


7 sites is most productive when the emphasis is on the larger regional context rather than a site-specific level.


8 CHAPTER TWO CULTURAL SEQUENCE OF THE CENTRAL PENINSULAR GULF COAST ARCHAEOLOGICAL REGION The term 'culture area' has been a useful way of breaking down large geographical regions into zones that exhibit similar environmental and topographic features, and which, partially as a result, contain similar cultural traits. Generally, the later the time period, the greater the difference between cultural groups as each becomes better adapted to its environmental context. Thus, in the later stages of cultural development of aboriginal groups in Florida, it is possible to differentiate more clearly between the various groups, particularly i n Contact period Florida, in which several different tribes have been ethnohistorically identified. In early prehistory however, cultural groups appear quite similar, particularly given the limited nature of the archaeological record. Stirling (1936) first classifiep. distinctive archaeological regions for Florida, although the Central Gulf Coast archaeological region was initially defined by Goggin (1947). The region was reduced in size with the classification of Milanich and Fairbanks (1980). Currently, the Central Gulf Coast region is defined as extending from Pasco County to the north, the Gulf of Mexico to the west,


9 Charlotte Harbor to the south, and the flatwoods to the east, although the interior extent of the culture region remains unclear (Milanich and Fairbanks 1980:24). With the exception of the Tampa Bay area, the Central Gulf Coast region was generally one of the less densely occupied regions of the state (Milanich and Fairbanks 1980:24). Paleo-Indian Period The cultural sequence begins with the PaleoIndian period at the end of the Pleistocene epoch, at approximately 12-14,000 years ago. Milanich and Fairbanks (1980:35) note that it is entirely possible that PaleoIndians were in Florida as early as 17,000 years ago. The earliest documented evidence however, is from Little Salt Spring, a Paleo-Indian site inundated by rising water, which dates from approximately 12,000 to 9000 years ago (Clausen et al. 1979). During this time, sea levels were lower due to the retaining of water by the ice sheets of the Wisconsin glaciation, and the Florida coastline extended well beyond the present Gulf of Mexico and Atlantic shorelines. At the time of the initial human occupation of Florida, sea levels were 35 meters below the present'level, and many Paleo-Indian sites were undoubtedly inundated by rising sea levels. As a result, our understandings about Paleo-Indian settlement patterns in Florida are limited. Paleo-Indian sites have long been associated with water sources, particularly springs and spring-fed rivers (Dunbar and Waller 1983; Neill 1964; Waller 1969), although as Milanich


10 and Fairbanks (1980:37) note, many of these Paleo-Indian specimens are isolated finds, or from rivers or collections where provenience has been lost. Studies of Paleo-Indian point distributions suggest a concentration of sites in areas where Tertiary age limestone outcrops, in and around sinkholes or karst features which expose limestone (Anderson 1989; Dunbar 1988; Dunbar et al. 1989; Dunbar and Waller 1983). "An amazing 90% of the Paleo-Indian sites containing the diagnostic artifacts Clovis, Suwannee, or Simpson projectile points and carved proboscidean ivory foreshafts are located near karst depressions that penetrate the Tertiary limestones to expose the Floridan aquifer as a source of potable water" (Dunbar et al. 1989: 25) These distributions, however, often suffer from the biases that Milanich and Fairbanks noted above, and distributions (Anderson 1989; Dunbar 1988) may reflect other things as well. Recent underwater excavations however, have begun to address the problem of site distributions in drowned areas. As more dry land Paleo-Indian locales are professionally investigated, information regarding settlement patterns will begin to be developed more fully. During the Late Pleistocene and Early Holocene many of Florida's inland rivers, lakes, and other surface water features were dry except for a few (mostly lowland) water holes that became the focus for Paleo-Indian activity (Dunbar et al. 1989:25). Other researchers have emphasized the importance of water resources, particularly springs and


11 spring-fed rivers, to Paleo-Indian settlement (Neill 1964; Dunbar and Waller 1983; Dunbar et al. 1989). In addition to the human need for water, these areas would also have been important as attractors of game animals. Based on his observations of Paleo-Indian artifacts in broad, shallow waters, Waller (1972) suggests that these areas were often game crossings which afforded Paleo-Indian hunters with the opportunity to ambush prey. He also states that this would allow for the floating of prey to ledges for butchering (Waller 1972:14). The presence of lithic resources would also have been an important determinant for prehistoric populations. Gardner (1977:260) notes that lithic deposits, as a secure resource, "served as a point where any segment of the population could readily predict the other segments would return to." It is most probable that Paleo-Indian groups entered Florida from the north as a result of natural expansion (Milanich and Fairbanks 1980:37). According to Martin (1973), the time frame involved may have been as little as 350 years from the initial movement into the New World, until populations reached the Gulf of Mexico. Although this is based on a model that suggests the primary importance of big-game hunting (and the expedited movement of human populations along a front which locally decimated the large herd animals), the time frame is based on a moderate growth rate for human populations in the New World. There is a


12 great probability that migration routes and expansion of populations at this time can not be correlated with biological models of animal expansion--that expansion would not be limited by the maximum rate of expansion, but rather, by cultural limits to migration. At present, there are little data on the lifeways of Paleo-Indian peoples, although the general consensus seems to be that they were hunting and gathering nomads, moving seasonally as the availability of plant and animal resources changed. Paleo-Indian groups presumably consisted of small migratory bands subsisting by hunting, fishing, and gathering. Evidence from Paleo-Indian sites in the north indicates the nomadic nature of Paleo-Indian bands. Goodyear, however, feels that data from the Harney Flats site in Hillsborough County indicates that "Paleo-Indian settlement behavior in peninsular Florida may be significantly different from that of systems to the north" (Goodyear, in Daniel and Wisenbaker 1987:iii). He further states that the apparent lack of exotic cherts at Harney Flats may indicate that Paleo-Indian groups in peninsular Florida were less geographically mobile because of the lack of pronounced seasonal climatic changes (Goodyear, in Daniel and Wisenbaker 1987:iv). Population numbers are thought to be low during this period, and groups were probably organized around bands with seasonal dispersal of macro and micro bands to efficiently exploit high resource areas. The estimation o f population


13 parameters however, is a notoriously difficult problem confronting archaeologists due to the nature of the data. Thus, most attempts at population estimation are based on the results of ethnographic research on similar types of present-day groups. The debate between big-game theory and non-big-game theory has played an important part in Paleo-Indian studies. Even today, though much less than in the recent past, there is a great emphasis on the importance of the hunting of megafauna, and its primary importance to the lifeways of the Paleo-Indian period. Willey (1966:38) epitomizes the traditional view of big-game hunting proponents: "There can be no question that it [the big-game hunting pursuit] was an activity of great and, probably primary importance." At the same time, other researchers rejected the primary importance of big-game hunting, and as Griffin (1964:224) notes: "The restriction of the diet of these early hunters to big-game has been by certain archaeologists not by Paleo-Indians who were present 12,000 to 10,000 years ago." While abundant evidence for the exploitation of megafauna has been documented for Paleo-Indian groups in the western United States, this type of predation pattern has not been documented as fully in the east. Since this functional association of fluted points and extinct megafauna was observed in the west, archaeologists have been concerned with this idea as the basis for Paleo-Indian subsistence (Mason 1962). One of the more provocative


14 studies at that time was that of Williams and Stoltman {1965) who found that fluted points and other Paleo-Indian lanceolate forms had a strong correlation with major river systems and the remains of extinct megafauna. Bullen, Webb, and Waller {1970) observed similar findings in north-central Florida, with the association of Suwannee points and late Pleistocene fauna of the interior rivers, including one apparent instance of butchering marks on a mastodon vertebra, while the bone was 'green.' Other researchers have espoused the latter view of Paleo-Indian subsistence however, particularly with better recovery techniques and the more recent advent of research designs specifically oriented toward the better understanding of Paleo-Indian lifeways. As Goodyear, House, and Ackerly note: Even if the "smoking" Clovis point is someday found embedded in the skull of a mastodon or any other extinct big game species, it will only contribute one form of evidence toward understanding Paleo-Indian subsistence systems. It seems highly unlikely that Early Man concentrated on any one animal species given the established array of economically useful game species in the late Pleistocene environment, especially in the temperate Southeast. It is also doubtful that there was any single Pateo-Indian subsistence pattern in the eastern United States given the extreme ecological and climatic diversity of the broad area. The formal similarities between the Upper Paleolithic and Paleo-Indian stone and bone technology suggest a strong hunting orientation in both economies, especially when compared to the expanded tool inventories of the ensuing Holocene Archaic societies. While it is not surprising that some as yet unspecifiable portion of the Paleo-Indian diet was comprised of gathered foods, this idea is only slowly coming to the fore (Goodyear, House


15 and Ackerly 1979:93). After an extensive review of Paleo-Indian literature, Daniel and Wisenbaker (1987:134) conclude as well, that ... the idea of Paleo-Indian subsistence being almost totally dependent on big-game no longer seems tenable." The typical Florida Paleo-Indian diagnostic point is the Suwannee, which Bullen describes as "a usually large and fairly heavy, lanceolate shaped, slightly waisted point with concave base, basal ears, and basal grinding of bottom and waisted parts of [the] sides" (Bullen 1968:48). The Simpson point, also described by Bullen ( 1968) is another tool diagnostic of Paleo-Indian in Florida. Generally, Simpson points are "wide bladed, relatively narrow waisted, fairly thin, concave based, medium to large sized, point with grinding on bottom and waisted edges" (Bullen 1968:49). The Clovis point, which is the most well-known horizon-marker of Paleo-Indian groups in the New World, is also occasionally found in Florida. Paleo-Indian points are widely distributed along the riverbeds of North Florida, such as the Aucilla, Withlacoochee, Suwannee, Santa Fe, Ichetucknee and Oklawaha, although this conclusion is based upon private collections or isolated finds, and thus are in poor contexts (Milanich and Fairbanks 1980:37). As Bullen (1975) states, and numerous others have archaeological materials reiterated, is limited the dating in Florida. of The accepted ranges of dates for Florida materials are based on


16 correlations with similar materials from outside the state, and may or may not accurately reflect the actual dates of these occupations in Florida (Bullen 1975:2). Based on technological and stratigraphic evidence, some temporal trends have been documented. Point length and point thickness appear to decrease through time, while waisting increases (Draft Comprehensive Plan 1990:34). Other tools have remained poorly documented as markers of Paleo-Indian activity--which is one reason that relatively few sites can be confidently assigned to this period. Thumbnail scrapers, turtleback scrapers, blade knives, flake knives, gravers, burins and bone 'points,' and sandstone abraders are part of a tool tradition that continues into the Archaic period, and as a result, do not allow for distinction by cultural period. Other tools such as bolas (oval ground stone weights), Waller knives, and bifacial knives may be diagnostic of the Paleo-Indian period, although without data from good contexts, the isolation of these tools as diagnostic is a tenuous assumption. Results of research at Harney Flats have allowed for the development of a rough typology for unifacial scrapers used during the Paleo-Indian period, including thumbnail scrapers, endscrapers, discoidal scrapers, and oblong scrapers (Daniel and Wisenbaker 1987:65-74), although little comparative work has been done to discern differences in lithic assemblages between the Paleo-Indian and Archaic periods.


17 Paleo-Indian sites may be divided into five categories, based on their general site type: base camps and villages, quarry sites, short term camps, kill sites, and isolated projectile point finds (Draft Comprehensive Plan 1990:38-39). The Silver Springs site, in Marion County, was the first systematically excavated Paleo-Indian terrestrial site in Florida. The site was originally investigated by Neill (1958), although a reinvestigation of the site (Hemmings 1975) produced no additional Paleo-Indian material. The Harney Flats site (Daniel and Wisenbaker 1987), located in Hillsborough County, has undoubtedly yielded the most data for Paleo-Indian studies in Florida. Extensive excavations at the site yielded a suwanneejBolen (Paleo-Indian) occupation, which was vertically segregated from an overlying Newnan (Archaic period) occupation. Although it was not possible to segregate the Bolen and suwannee occupations at the site, it did allow for the recovery of the associated Paleo-Indian assemblage. Warm Mineral Springs and Little Salt Springs located in Sarasota County have yielded a great amount of environmental and cultural data. Human skeletal remains and associated organic material recovered from the Warm Mineral Springs site date from approximately 10,000 years ago (Cockrell and Murphy 1978). The Page-Ladson site in Jefferson County, represents a kill site in which Paleo-Indian artifacts were found in association with animal bones.


18 The Late Paleo-Indian period begins at approximately 9500 years ago (Milanich and Fairbanks 1980). This has been the source of disagreement among researchers, some preferring to call it the Dalton period (Bullen 1975; 1976) and others using the rubric of Late Paleo-Indian or Transitional period (Purdy 1981; Milanich and Fairbanks 1980). Bullen places Dalton (which he dates from 10,000 to 8000 years ago) as the first cultural phase of the Archaic (Bullen 1976:34). Bullen recognized two traditions within this period--the Dalton tradition proper (represented by Santa Fe, Tallahassee, and Dalton points) and a notched tradition (represented by Greenbriar, Bolen and Hardaway points) as the basis for Dalton technology (Bullen 1976). Bullen (1976) and Hemmings (1972) argue that the diagnostic tools from this time period differ from Paleo-Indian technology sufficiently to necessitate separation from the Paleo-Indian, while Milanich and Fairbanks (1980:45) contend that they appear to be transitional stylistically from Paleo-Indian to Archaic point forms. Milanich and Fairbanks (1980:42) state that although populations increased, the basic lifeway remalned similar to that of the Paleo-Indian period. Although deer was an important resource, a trend which increases in later periods, evidence from Devil's Den in Levy County indicates that the disappearance of many other species of fauna associated with Paleo-Indian subsistence did not occur in Florida until sometime after 8500 years ago (Martin and Webb


19 1974). In addition to diagnostic points from this period, Bullen and Beilman (1973) have documented a large rnicrolithic tool complex based on excavations at the Nalcrest site in Polk County. Artifacts measuring from 1-4 ern in length included scrapers, stemmed points, drills, spurred gravers, and other artifacts. Other sites have yielded similar tool complexes, which appear to be in association with Bolen points of the Late Paleo-Indian period, as well as some early Archaic contexts (Milanich and Fairbanks 1980:48). These probably represent examples of early special-use sites. Archaic Period The Paleo-Indian period is followed by the Archaic period, beginning at approximately 8500 years ago, with the disappearance of many species of fauna. The Archaic has been divided into three stages, Early, Middle, and Late, primarily based on projectile point characteristics (Milanich and Fairbanks 1980). Bullen's divisions however, as noted earlier, are somewhat different for the Archaic (Figure 2). Dalton has been interpreted as Late PaleoIndian or Transitional by most recent investigators, an interpretation that will be followed here. Milanich and Fairbanks (1980) date the Early Archaic from 8500 to 7000 years ago (currently revised to 9500-7000 BP by the Florida Cornprehensi ve Plan, 1990) the Middle Archaic from 7000 to 5000 years ago, and the Late Archaic


1z w en w ti: a... w ti: 0 u.. w III en ti: c( w >3000 4000 late Preceramic Archaic 5000 Middle 6000 Preceramic Archaic 7000 Earty Preceramic Archaic 6000 9000 Dalton 10000 Paleo-Indian 11000 12000 13000 14000 Bullen ( 1975, 1976) Late Archaic Middle Archaic Earty Archaic Late PaleoIndian or Transitional Paleo-Indian Mllanich and Fairbanks (1980) Preceramic Archaic Late PaleoIndian or Transi t ional PaleoI ndian Purdy {1981a) Late Archaic Middle Archaic Earty Archaic Paleo-Indian Draft Comprehensive Plan (1990) FIGURE 2. Early Prehistoric Chronologies in Florida. 20


21 from 5000 to 3000 years ago based upon environmental shifts toward present-day conditions, the reduction in land size as sea levels rose, the changes in faunal and floral resources, and the cultural shifts which apparently took place as a result. The terminology of Milanich and Fairbanks (1980) allows for consistency with behavioral trends and cultural responses beyond Florida, although chronological dates may differ in various regions outside of Florida. Populations during the Archaic primarily subsisted on hunting, fishing, shellfish gathering, and plant collecting. Although Archaic groups were perhaps more sedentary than the Paleo-Indian groups, Archaic populations probably moved seasonally in order to efficiently exploit resources on the coast and in the interior. Population sizes, again, are difficult to determine, although it is probable that populations began to grow with the increasing sedentism and increasing technological efficiency. Generally, Archaic sites are numerous and diverse in their cultural manifestations, including large midden sites and base camp sites, and small special-use sites such as lithic reduction workshops, quarry sites, special use and extractive camps, and cemeteries. Each of these functional classes of sites has been investigated in the Central Gulf Coast area. Evidence from Little Salt Spring, the Windover site near Titusville, and the Republic Groves site in Hardee County suggests that Archaic period burial practices were much more complex than was once thought. Burials at these


22 sites were placed in peat-producing ponds or sloughs, pinned down with wooden stakes, and interred with a variety of burial goods (Clausen et al. 1979; Doran and Dickel 1988; Wharton et al. 1981). Other cemetery sites such as the Gauthier site in Brevard County, are not associated with a peat-producing pond, although the cemetery may have been in a slough at the time of interment. The burials at this site were contained within a burial pit with few associated grave goods. At the Tick Island cemetery in Volusia County, burials were placed directly in a shell midden. Jahn and Bullen (1978) have interpreted this site as a large shell midden in which a number of primary, flexed group burials were placed into pits. At some later point, a sand mound was placed over the midden and sealed with muck. The tool kit during the Early Archaic consists of a larger variety of projectile points/knives which include both stemmed and non-stemmed varieties. Prevalent point types during this period include stemmed, basally chipped types such as Arredondo, Hamilton and Kirk Serrated points (Bullen 1975:38-39). As Milanich and Fairbanks (1980:50-51) state: The larger variety of projectile point (and knife?) types present in Early Archaic times may reflect ethnicity or, more likely, different site functions. Some of the points and related tool complexes are like those found in adjoining states, suggesting that communication over large areas was occurring or that populations were still entering (and leaving?) Florida as part of the yearly subsistence round. The division between Early and Middle Archaic occurs at


23 about 7000 years ago based on evidence from three sites, all of which contain Newnan projectile points: the Newnan site, 6850 BP (Clausen 1964:28; Milanich and Fairbanks 1980:54); Little Salt Spring, 7000 BP (Clausen et al. 1979:611); and, Tick Island, 6200 BP (Jahn and Bullen 1978:22, n.1). Dates for Newnan points are based on association with burials at the Newnan site, and date to 5400 BP (Bullen 1975). At the Newnan site, 186 Archaic stemmed points were recovered, of which 95 percent were Newnan points (Clausen 1964; Clausen et al. 1975; Milanich and Fairbanks 1980:58). Another important type site for this period is Little Salt Spring, which apparently was not occupied after the Late PaleoIndian period, until a Newnan occupation at about 7000 years ago (Clausen et al. 1979). Newnan projectile points are the most distinctive and widespread of the diagnostic points from this period (Milanich and Fairbanks 1980:54). Hillsborough points appear to be a Newnan variety (Milanich and Fairbanks 1980:54). Small, special-use camp sites predominate: ... we can guess that these sites were used for hunting and foraging of certain, perhaps seasonal resources, by small groups ... such sites are characterized by lithic debitage and a smaller amount of tools, including points, knives, scrapers, and a few larger chopping or hammering tools (Milanich and Fairbanks 1980:57). Most of these sites are difficult to interpret due to the paucity of cultural material and archaeological features, with the exception of lithic artifacts (Chance 1982:29). As a result, our complete understanding of many of these types


24 of sites is limited because only one class of artifact (lithic material) is differentially preserved as a result of the highly acidic soils which have degraded most or all of the organic material of these sites. Milanich and Fairbanks state that larger sites believed to be central-base villages for larger groups may cover several acres, with thousands of pieces of debitage and tools, while other sites, apparently occupied for shorter periods of time still may extend over a relatively large area (Milanich and Fairbanks 1980:57). This makes estimations of population sizes and permanence of occupation difficult. The problem is compounded with the lack of paleobotanical and faunal remains characteristic of lithic scatters in this region. Research by Hemmings and Kohler ( 1974) at the Lake Kanapaha site in central Alachua County, attempted to provide a basis for the types of problems associated with lithic scatters. Their work was directed at addressing the archaeological distribution of stone tools and debris and if and how it reflects prehistoric activity patterns, the isolation of different tool complexes stratigraphically, and human activity patterns horizontally (Hemmings and Kohler 1974:47). Other tools from Archaic sites include scrapers, drills, choppers, knives, blades, gouges and hammerstones, as well as a number of bone, shell, and wooden tools recovered from Archaic period sites. None of these tool types can be considered as exclusively diagnostic of the


25 Archaic period, as little research has been directed at producing chronologies of these tool forms. Archaic period groups were migratory hunters and gatherers, although there is evidence that they were more sedentary than the earlier Paleo-Indian groups, with a wider range of resources utilized. Milanich and Fairbanks state that Middle Archaic people were performing the same types of activities at their villages and camps as their predecessors. If this is true, Middle Archaic groups must have been using a much greater variety of tools to do them. The implication here is that there is an increasing tendency toward sedentism, most likely on a seasonal basis, which led to the acquiring of more specialized tools (Milanich and Fairbanks 1980:57). At least some of the tools are larger, and less transportable in large numbers (Milanich and Fairbanks 1980:57). As of yet, no good comparative analysis of lithic collections from camps versus central bases has been performed (Milanich and Fairbanks 1980:58). such an analysis will allow for the developing of understandings about the uses of different sites, particularly the relationships between central bases, and special-use sites, procurement sites, etc. within a settlement system. Quarry sites are also known for the Middle Archaic. These consist of both large sites in localities of major outcroppings where chert was mined, and very small sites where smaller outcroppings were present such as along rivers, around lakes, and where erosion has cut through the


26 soil exposing chert deposits in the limestone. Quarrying activity was probably carried out in conjunction with other activities at camps (Milanich and Fairbanks 1980:58). Although there have been many quarry sites recorded, relatively few have been investigated. The Senator Edwards site (Purdy 1975) was an Early and Middle Archaic lithic workshop at which natural chert outcroppings were used, although no evidence of true quarrying activity was seen. The Container Corporation of America site (Purdy 1981) was another procurement area that was apparently utilized during every cultural period. The site was characterized by a high density of debitage, although a fairly significant amount of the recovered materials appear to have been utilized. The most intensely investigated and reported quarry site is the Wetherington Island site (Chance 1981, 1982; Chance and Misner 1984), located on Cowhouse Creek, and investigated as part of the Interstate-75 project. At Wetherington Island, there is only a small amount of late stage reduction activity, which seems to have taken place at the northern extreme of the site, or at another locale. Thus, a relatively unbiased quarry assemblage was isoiated, allowing an understanding of behavior at quarry sites (Chance and Misner 1984). Numerous other Middle Archaic sites have been excavated throughout the state, including several important sites located and investigated as part of the Interstate-75 project in Hillsborough County. Some of these are discussed


27 in greater detail in a following section of this report. The beginning of the Late Archaic period is not well documented, and the date of 5000 years ago is little more than an educated guess (Milanich and Fairbanks 1980:60). Diagnostic point types include smaller, stemmed, and corner notched types such as Culbreath, Clay, and Lafayette points. Hemmings and Kohler, in a brief review of lithic assemblages from the Lake Kanapaha site and Late Archaic assemblages from the Johnson Lake site, Dixie Lime Caves, and the Silver Springs site in Marion County, and Bolen Bluff in Alachua County, indicate the dependence on hunting and collecting, the processing of animal and plant foods, working of wood, bone, antler, hides, plant fibers, and other raw materials, and the preparation of stone tools associated with these activities (Hemmings and Kohler 1974:62). Post-Archaic Periods The end of the Archaic is marked by the appearance of fired clay pottery. This varies in different regions, but 4000-3000 years ago is a generally acceptable date for the Tampa Bay area (Milanich and Fairbanks 1980:60). This period is called the Orange period, and Bullen (1972) places the plain fiber-tempered pottery with the Orange series of East Florida. Phelps ( 1965) and Milanich and Fairbanks (1980) place this within the Norwood series particular to the Gulf Coast (Tampa Bay and north) which was originally designated as st. Simon's by Willey (1949). Fiber-tempered


28 pottery of the Norwood series consists of two types, Norwood Plain and Norwood Simple Stamped (which has characteristic parallel dowel impressions). Spatially, Norwood Plain appears to range all through the Gulf Coast, while Norwood Simple Stamped ranges from northeastern Hillsborough County to the north. The basic Archaic-like hunting-foraging subsistence pattern continued through this period, although Milanich and Fairbanks (1980:60) note that villages probably served as central bases, possibly year round, indicating increasing sedentism. Significant cultural changes, however, were taking place, including an increased degree of interaction with groups outside of Florida (Milanich and Fairbanks 1980:61). The beginnings of horticulture apparently developed during this period, although evidence is limited (Milanich and Fairbanks 1980:61). It is after the Orange phase, by about 3200 years ago, that the use of the culture area distinctions becomes significant. Prior to this period, material culture and (presumably) lifeways were similar throughout Florida, and in many cases, the southeast United States. At this point however, archaeologically documented regionalism begins to develop beyond the local adaptations to a specific area. This trend continues until the Contact period. The Florida Transitional phase ( 3200 to 2500 years ago), is a period of significant cultural change. Bullen postulates that diffusion of various cultural traits


29 resulted in the spread of several ceramic and tool traditions throughout a wide range (Bullen et al. 1978). Semi-fiber-tempered pottery at the beginning of the period is gradually replaced by sand tempered, sand and limestone tempered, and temperless chalky-ware. By the end of the Florida Transitional period, distinct regional ceramic traditions can be recognized. Bullen (1959:53) suggests that horticulture may have become important during this period, indicated by the appearance of heavy adze-like tools. Milanich and Fairbanks further this idea by suggesting that there was a move away from a hunting-collecting-gathering subsistence base into a more horticultural-based subsistence pattern (Milanich and Fairbanks 1980:61). Archaeological evidence for this, however, remains to be documented and appears unlikely in the Manasota culture areas, discussed below. The Deptford cultures, which extend through the northern Gulf Coast of Florida and the Atlantic Coast into Georgia and South Carolina appear to have had little influence in the Tampa Bay region, and southward. The cultural manifestations in this area were described by Willey (1949) as Perico Island, although this was poorly defined due to the lack of excavated sites at that time. The Manasota phase (500 BC to AD 800) was first suggested by Luer and Alroy (1979) to replace the Perico Island phase. In this area, Deptford and Early Weeden Island ceramics occur only rarely, the dominant ware being sand-tempered


30 undecorated pottery. An extensive shell and bone tool industry is also characteristic of Manasota period sites. The period has been divided into an early period (500 BC to AD 400) and a late period (AD 400 to 800). Luer and Almy (1982) suggest two subsistence patterns which can be documented for the Manasota phase--a fishing and shellfish collecting complex associated with sites in coastal estuaries, and a plant collecting complex in the pine flatwoods (Luer and Almy 1982:43). Settlements along the coastal areas are composed of large village sites and smaller extraction and special-use camps. The larger settlements allowed for access to both marine and inland flatwoods locales, which allowed for the minimum effort for the maximum gain in terms of the relationship between habitation areas and resource areas. period begins at approximately AD 400, The late Manasota with changes in village ceramics and an influx of Weeden Island influence, including the construction of burial mounds and the use of elaborate funerary pottery. In the areas where Manasota culture was dominant, Manasota influence continued until approximately AD 800, with the spread of Safety Harbor influence throughout the area. During this period of Weeden Island influence, Manasota subsistence patterns continued, population increased, and there is a population shift into the interior regions of the state. Luer and Almy (1982:47) document shifts in Manasota burial practices, suggesting a gradual influx of Weeden


31 Island influence into the area. Ceramic changes reflect this shift as well, as pottery decoration begins to reflect Weeden Island influences, although the idea that a distinct Weeden Island manifestation was present between Manasota and Safety Harbor appears to be in error (Draft Comprehensive Plan 1990:110). The culmination of this influx in areas outside of the Manasota area is the adoption of Weeden Island socio-political organization and ceremonialism. The only Central Peninsular Gulf Coast correlate to Mississippian cultural traits is Safety Harbor, from A.D. 900 to as late as A.D. 1725 (Mitchem 1989:556-560). The closing dates for this period vary throughout Florida and the Southeast with differences in degree of contact (Mitchem 1989:565-566). Mitchem (1989:557-567) defined four phases: the Englewood phase (AD 900-1000); the Pinellas phase (AD 1000-1500); the Tatham phase (AD 1500-1567); and the Bayview phase (AD 1567-1725). Mitchem (1989:567-579) also defined five regional variations of Safety Harbor culture: Northern, Circum-Tampa Bay, Manasota, Inland, and South Florida, each defined mainly on the basis of ceramic variations and settlement pattern differences. Archaeologically, the Safety Harbor culture is defined on the basis of incised and punctated ceramics (Willey 1949). The most common pottery types in Safety Harbor contexts are Englewood Incised, Sarasota Incised, Safety Harbor Incised, Port Washington Incised, and Pinellas Incised. Most of the decorated types appear to have been


32 produced for use in mortuary contexts, and are rarely found in domestic contexts. The various types of utili tar ian plain wares show a high degree of regional variation. In the area north of the Tampa Bay area, Pasco Plain is the predominant utilitarian ware, while Pinellas Plain wares are common in the Tampa Bay region. To the south of the Tampa Bay area, sand-tempered plain wares and Belle Glade Plain are the most common Safety Harbor ceramic wares. The typical Safety Harbor stone tool assemblage includes the Pinellas point, as well as Ichetucknee and Tampa points (Bullen 1975), although scrapers, blades, and various types of utilized flake tools similar in style to preceding cultural periods have been recovered from Safety Harbor contexts. Ground and polished stone celts have been recovered from Safety Harbor burial mounds, and do not appear to be utilitarian tools. Various styles of plummets and pendants are also apparently associated with mortuary practices (Draft Comprehensive Plan 1990:166; Mitchem 1989:399-402). The Safety Harbor cultural manifestation was first identified and defined on the basis of excavations at Philippe Park, in Pinellas County (Griffin and Bullen 1950; Willey 1949). Historically, this group is known in the area as the Tocobaga. Tocobaga society was stratified, and Bullen (1978:50) suggests that the Tocobaga were composed of small chiefdoms, often at war with each other. Subsistence was probably similar to preceding periods, and hunting,


33 gathering and shellfish collecting probably remained the major part of the subsistence base. Spanish accounts indicate that maize was cultivated by the Tocobaga, although archaeological evidence for this has not been documented for the Tampa Bay area (Luer and Alroy 1981:147). Settlements during this period reflect the sociopolitical development of the Safety Harbor groups. Settlements cluster around ceremonial centers, almost all of which are located in coastal areas (Luer and Almy 1981). Luer and Alroy {1981) state that there were between 15 and 20 centers in the Tampa Bay region. Goodyear {1972) views the "Tocobaga Timucuans" within a model of "central town theory" in which large towns with temple mounds were situated among smaller villages, which depended upon the temple towns for political and religious services. Despite the increase in cultural complexity, there is no evidence to suggest that populations were growing during this period, and in fact, evidence suggests that population reductions might have taken place, particularly after the arrival of Europeans to the New World. The prehistoric era was brought to a close by the Spanish expeditions into Florida. By A D. 1565 there is evidence of a good deal of contact with the Spanish military and missionaries, which altered the traditional lifeways of the aboriginal groups in Florida. Disease and cultural disruption followed the Spaniards, and by the end of the seventeenth century these aboriginal groups were virtually


34 extinct (Milanich and Fairbanks 1980; Weisman 1989; Mitchem 1989). After repeated missionaries left conflicts, the Tampa the Bay Spanish area to soldiers focus on and the conquest and missionization of tribes in northern Florida. The once thriving villages and cultures of the Tampa Bay area, however, were decimated. According to Neill (1968), remnant Tocobaga joined groups of Spanish fisherman, who moved into the rich shellfishing areas of Tampa Bay. Several sites on Old Tampa Bay were probably sites of Spanish/Indian shellfishing during the first half of the eighteenth century (Deming 1980:10). In the early eighteenth century, groups of Creek Indians moved into Florida to escape the political and population pressures of the expanding American frontier. These groups became known as the Seminoles. Most Seminole sites are small, and generally not well known, although an increasing body of recent literature is developing our understandings about Seminole lifeways (see Weisman 1989). These sites are generally characterized by the presence of the traditional Creek brushed surface pottery, as well as European trade goods. After the British took possession of Florida in 1763, numerous trading posts were established near Seminole towns. As a result of the shifting economy that this trade brought, the material culture of the Seminoles changed dramatically. Following a second period of Spanish control (1783 to 1821),


35 Florida became an American territory. During this period, the American government implemented a policy of containment and removal of the Seminoles. Beginning in 1823, Seminoles from the Gulf Coast area were moved to the Moultrie Creek reservation. Fort Brooke, located in what is now downtown Tampa, was established to monitor the southern boundary of the reservation. Its occupation spanned from 1824 to the 1880s, and played an important role in the Seminole Wars as well as the development of Tampa (Piper and Piper 1982). The result of the Seminole Wars (1818-1819; 1835-1842; 1855-1858) was the eventual removal of most of the Florida Seminoles to Indian Territory in Oklahoma. However, some groups of Seminoles remained in small pockets, particularly in the Everglades, escaping removal by government forces.


A CHAPTER THREE ENVIRONMENTAL SETTING consideration of the environmental 36 contexts associated with the cultures of a particular period is important in allowing for the understanding of resource availability and group movement throughout a geographic range. Substantial climatic and vegetational changes occurred during the Late Pleistocene and Early Holocene, and these would have had a significant impact on cultural groups, both directly and indirectly, with changes to biotic communities, and the availability of water and food resources. The modern climate of Hillsborough County may be characterized as subtropical. The average annual temperature in Tampa is 71.5 degrees F (21.9 degrees C), with an average winter temperature of 62.5 degrees F (16.9 degrees C) and an average summer temperature of 81. 4 degrees F (27.4 degrees C) (Leighty et al. 1958:2). The average annual precipitation in Tampa is approximately 51 inches (130 em), although total yearly precipitation fluctuates considerably. The majority of rainfall occurs in the summer months, as a result of thunderstorms, tropical depressions, and hurricanes (Wright 1973:17). The Hillsborough River drains approximately 1,787


square kilometers of area in Pasco, Hillsborough Counties (see Figure 1). in the Green Swamp area and flows 37 Hernando, Polk, and The river originates in a southwestern direction to Hillsborough Bay at Tampa. Upchurch suggests that Cowhouse Creek may have held the main flow of the Hillsborough River until as recently as 3000 years BP {Sam B. Upchurch, personal communication, 1991). In addition, several sinkholes and perched water ponds are scattered throughout the area. Dunbar ( 1982:63) notes that the Hillsborough River has been in existence for at least 27,000 years. Although the exact course of the river has been the subject of debate, the Hillsborough River and Cowhouse Creek have probably been continuous sources of water and faunal and floral resources for much of the period of human occupation of the area. Prehistoric Environment The Pleistocene environment of Florida may be characterized as cooler and drier during the Wisconsin maximum at about 18,500 BP {Whitehead 1973; Gates 1976). Although full glacial conditions did not exist in the southern latitudes, the circumstances influencing the advances and retreats of ice sheets have had a marked effect on all of North America. The average summer temperature was as much as 10 degrees c cooler than the present {Gates 1976), while the winter temperatures are believed to have been milder (Bryson and Wendland 1967). There was probably a more even


38 distribution of rainfall throughout the year (Coleman 1982:149), and a lack of marked seasonal fluctuations. Pollen samples from lake sediments (Watts 1969 1971 1975 I I I 1980) indicate the presence of an oak savanna and herb/grassland prairie during the period from 33,000 to 13,500 BP. Pine, which was rare at the beginning of the Pleistocene, increased in abundance toward the end of the period (Watts 1980:400). At the close of the Pleistocene, three basic faunal regions are presumed to have existed in the southeastern United States, based on the dominant flora (Webb 1981: fig. 4.1.10). The Boreal zone from the middle of South Carolina northward, was primarily a tundra habitat, and contained such species as wooly mammoth, caribou, horse, and bison (Webb 1981:76). The Temperate zone, in southern South Carolina and northern Georgia, was ecologically diverse, consisting of mixed temperate forests and grasslands. Using the ratio of mammoth to mastodon remains as an indicator of open and wooded habitats, Webb (1981:79) notes that there was a gradation into open grasslands along the coastal plain. American Mastodon, mammoth, deer, and bison w ere present in this zone. The Subtropical zone included parts of southern Georgia, Florida, and the Gulf Coastal Plain (including areas which were inundated by rising sea levels). Evidence from Devil's Den, Little Salt Spring, and Warm Mineral


39 Springs in peninsular Florida indicate the diversity of faunal species present in Florida during the late Pleistocenejearly Holocene transition. Representative fauna include extinct species such as the giant sloth, dire wolf, sabercat, horse, giant armadillo, tapir, pecarry, and mastodon (Martin and Webb 1974; Clausen et al. 1979; Scientific Applications, Inc. 1981; Webb 1981:80). Most birds, freshwater fish, and reptiles recovered from Pleistocene contexts are indistinguishable from modern faunal species in Florida, with the exception of the giant tortoise and the giant box turtle, which have become extinct within the past 10,000 years. Similarly, many extant mammalian species were also present, including white-tailed deer, panther, rabbit, black bear, raccoon, gray fox, muskrat, squirrels, and opossum. The variety of faunal species present indicates that the subtropical zone was one of the richest and most diverse environments during that time (Carbone 1983:16). Environmental Change To the north, palynological data indicate that temperate deciduous forests were replacing boreal types in the southern Appalachians by as early as 16,500 BP (Delcourt and Delcourt 1985:19). These mid-latitudes were essentially transformed by 12,500 BP, with cool, mesic temperate species the dominant cover (Delcourt and Delcourt 1979). South of the mid-latitudes during the period from 12,500 to 8500 BP however, climate and vegetation were quite different. Data


40 suggest that the southeastern evergreen forest (Braun 1950) was stable during the end of the Pleistocene, with "no major invasions of cool-temperate or boreal species" (Delcourt and Delcourt 1983:269). Evidence from Lake Annie (Watts 1975) suggests that during the period between 37,000 and 13,000 BP, south central Florida was extremely dry, with a period of maximum aridity between 18,500 and 14,600 BP (Wright 1981). The dominant vegetative cover consisted of a sand dune scrub (Delcourt and Delcourt 1981). By 14,600 BP, the cool and dry climate began to change, with the lessening of glacial influence (Watts and Struvier 1980). Watts (1980) suggests that broad-leafed mesic forests began to develop, indicating a period of greater rainfall. Pollen from Lake Annie (Watts 1975) during this time documents the disappearance of dune vegetation by about 13,000 BP, and the replacement of this vegetation with an oak scrub prairie. watts (1980) dates the beginning of the Holocene in peninsular Florida to 1213,000 BP. Pollen from Little Salt Spring and other sites indicates that Florida was still relatively dry during 12,000 to 9000 BP (Brown 1981; Wright 1981). Paleobotanical evidence from Little Salt Spring and hydrological evidence from Warm Mineral Springs suggest wetter conditions from 9000 to 8000 BP (Lazarus 1965:56). Increasingly moist conditions occurred during the Middle Archaic after about 6000 years ago, and a gradual


41 change in forest cover took place with oaks giving way to pines or mixed forests in some places. These changes may be associated with settlement pattern shifts during Middle Archaic (Milanich and Fairbanks 1980:60). The change to a warmer, moister climate resulted in successive vegetational changes, culminating with the long-leaf pine forests, cypress swamps, and bayheads which characterize peninsular Florida today (Watts 1971). Palynological (Long 1974; Watts 1980) and geomorphological (Thanz 1975) evidence suggests that this essentially modern vegetational community has dominated for the past several thousand years. Sea Level Change The rise in eustatic sea level since 18,500 years ago is an important factor in attempting to understand the prehistoric occupations of coastal areas, and the landscape during periods with different water table levels and different topography (Figure 3). Inundation of archaeological sites in the Central Gulf Coast region is particularly critical as there is a very gradual gradient to the coastal plain, and sea level changes of only a few meters would either flood several thousand square miles of land, or create new shorelines in presently inundated areas of the Gulf of Mexico (Scholl and Struvier 1967). Sea level models and archaeological investigations support the idea of fluctuations in sea level since the arrival of human populations into Florida. Sea levels were substantially lower during the Paleo-Indian and Early


/ ................ .\., ? : \ L ...... \ so.,. _. .. -' _ ____ ... N FIGURE 3. Late Pleistocene/Early Holocene Sea Levels (from Daniel and Wisenbaker 1987:153). 42


43 Archaic period. The land mass of the Florida peninsula was as much as 50 percent larger during the glacial maximum (Milanich and Fairbanks 1980:37). By the Middle Archaic, however, sea level estimates are not as well-documented. Between 16,000 and 7000 years ago, the rise in sea levels appears to have been rapid (Milliman and Emory 1968). Eustatic change over the past 6000 years remains unclear. Three models have been suggested for this period: that sea levels reached their present levels by 8000 years ago with several fluctuations (Fairbridge 1960, 1961, 1974; Morner 1969, 1974); that sea levels remained relatively constant (Fisk 1956; McFarlan 1961); and that sea level has risen at a constant rate which gradually decreased in the present (Shepard 1964; Scholl 1963, 1964; Scholl and Struvier 1967; Scholl, Craighead, and Struvier 1969; Milliman and Emory 1968; Scientific Applications 1981). The "Fairbridge Curve" (Fairbridge 1960, 1961, 1974) suggests several fluctuations in sea level through time. In this model, sea levels at approximately 6000 BP, were 2-3 meters above present levels. These dropped to approximately one meter below the current point, and then rose again to 3 meters above modern levels by about 3700 BP. By 3200 BP, sea levels again dropped to one meter below the present level. At 2300 BP, sea level was 2 meters higher than present, and by approximately 1800 BP, had dropped to 2 meters below current levels. After this point, there were several minor fluctuations of 0.5 meter or less.


44 Investigations at St. Vincent Island (Stapor and Tanner 1977) suggest a fall in sea level from 6000 to 4000 BP a I rise prior to 2100 BP, a decline between 1800 and 1500 BP, and a rise in sea levels from 1500 to 800 BP, to present levels. The prediction curves, from archaeological data and the positions of erosional scarps, closely follows the proposed model of Fairbridge (1974). The model forwarded by Milliman and Emory (1968) suggests that sea level declined rapidly between 21,000 and 16,000 BP, to a depth of approximately 130 meters below the modern level. Between 16,000 and 7000 BP, sea levels rose rapidly. From this point, they rose gradually, reaching modern levels by approximately 2000 BP (Milliman and Emory 1968: fig. 2). Dillon and Oldale ( 1978) reevaluated these curves, and determined that the maximum depth was only 80-85 meters below present levels, opposed to Milliman and Emory's (1968) determination of 130 meters. The problem of determining accurate sea level curves is compounded with data indicating that sea level has not risen as rapidly on the east coast of the North American continent, suggesting that the methods of analysis are somewhat deficient, or that there has been a subsidence of land along the eastern coast (Shepard 1963). Recent investigations have identified five problems in determining accurate sea level curves: the selection of sea level indicators, the tendency of these indicators to be


45 transported, the unreliability of current dating techniques, the relative rarity of sea level indicators older than 6000 years BP, and shallow water indicators which do not allow for accurate determinations of sea level standstills or short term reversals (Scientific Applications 1981:43-45). At Ray Hole Springs, in northwest Florida, a stratigraphic sequence of marine sediments overlying a level dominated by oyster shell was reported. Underlying the oyster level, and above the limestone bedrock, was a large chunk of live oak, suggestive of a coastal hammock (Dunbar et al. 1989:28). This material was dated to approximately 8000 BP. The presence of oyster is indicative of brackish water, while the oak suggests a fresh water environment. A comparison to dates for the oyster layers indicate that "a brackish water coastal environment had replaced terrestrial/freshwater habitats by about 7500 years ago" (Dunbar et al. 1989:28). At the Turtlecrawl Point site in Boca Ciega Bay, Early and Middle Archaic period artifacts were recovered. Goodyear, Upchurch, and Brooks (1980) have determined that the Early Archaic occupation at the site was between 13 and 28 kilometers inland from the Gulf of Mexico. By the Middle Archaic occupation at the site however, the site was located on or near the coast (Goodyear, Upchurch, and Brooks 1980). Dunbar et al. (1989) report similar findings at the Ecofina Channel site, located approximately 5 km offshore, in northwest Florida. Under recent sediments was a level of


46 articulated oyster shell (representing an extinct oyster bar which existed when the water was less salty, and there was a fresh water influx at the mouth of the river), followed by a level of disarticulated oyster shell and other brackish water species. This latter level contained abundant chert debitage and tools, including the base of an Archaic stemmed point. A cypress root recovered from at and below this level was dated to approximately 5000 BP (Dunbar et al. 1989:29). Thus, during the Archaic period, the site was near the coast, along a freshwater river, with brackish water surrounding the site as the river met the Gulf of Mexico (Dunbar et al. 1989:29). Of particular importance to sites during this period, particularly inland sites, are the effects of sea level changes on the availability of potable water. Changes in the levels of the Floridan Aquifer and water table aquifer would not only limit the flow of water from springs, but change drainage patterns as well. Water from the Floridan Aquifer is available in sinkholes and springs, where the Miocene and earlier age limestones are not covered by more recent deposits which contain the water table aquifer. Water from the water table aquifer is available from perched water systems. These are shallow bodies of water held near the surface by the impermeable clay of the weathered Tampa formation limestones (Sinclair 1973:13), and fed from both the water table aquifer and rainwater. The base level of these aquifers is


47 dependent on sea level (Dunbar 1982:77-80). Early inhabitants of the Florida peninsula generally settled around sources of water from permanent solution lakes and seasonal perched water supplies. Shallow ponds and rivers fed by the Floridan Aquifer were dry due to the depressed level of the Aquifer as a result of lower sea levels. Areas around sinkholes were settled (Clausen et al. 1975, 1979), as were other areas of the karst region in which both solution lakes and perched water sources were available (Dunbar and Waller 1983; Dunbar 1988; Dunbar et al. 1989). A synthesis of the hydrology of the area was proposed by Dunbar (1982), based on data collected during the Interstate-75 project. This study follows sea level curves as suggested by Fairbridge (1974), and proponents of similar views (e.g., Marner 1969; Stapor and Tanner 1977). By 10,000 BP, the perched water systems began to retain water for longer periods of time as rainfall increased. Wetter conditions between 9000 and 8000 BP contributed to the formation of relatively shallow, perched water table lakes and ponds, including Lake Thonotosassa (Watts and Struvier 1980). Water levels in these systems approached modern levels by 8500 BP. Water levels of the Floridan Aquifer however, remained lower due to the lower sea level of this time. Dunbar (1982:99) suggests that the Hillsborough River flowed intermittently during this period, and consisted of a


48 series of discontiguous shallow ponds for much of this time. By 6000 BP, Dunbar ( 1982:98) suggests that the Floridan Aquifer was at its modern level, resulting in fresh water discharge from springs and spring-fed rivers, including the Hillsborough River. Arid conditions dried up many of the perched water sources during this time, restricting potable water sources to springs, rivers, and sinkholes (Dunbar 1982:98). From 6000 to 5000 BP, during a period of higher sea levels, the Floridan Aquifer was approximately 1. 5 meters above its modern level (Dunbar 1982:101), resulting in abundant surficial water sources. By 4200 BP, Dunbar suggests that the Floridan Aquifer dropped to 1.5 meters below modern levels (Dunbar 1982:102). Increasing rainfall levels maintained the perched water systems, as the flow from the Floridan Aquifer decreased. By 2500 BP, and continuing until 250 BP, the level of the Floridan Aquifer began to rise to modern levels. Rainfall during this period was greater than current amounts, and lower areas were probably flooded (Dunbar 1982:102). Complicating the results of this research is the course of the Hillsborough River, which Upchurch (1982:145) suggests followed a somewhat more direct route than its present course. Although geomorphological testing of this theory was of limited success, stimulating evidence suggests that the course of the Hillsborough River was what is presently the cowhouse creek channel. Upchurch (Sam B. Upchurch, personal communication, 1991) feels that this


49 course may have lasted until as late as 3000 BP. Regardless of the flow of the Hillsborough River, Dunbar's (1982) research indicates that settlement along Cowhouse Creek must have been limited to a great extent by the availability of water resources. During the Paleo-Indian period, when the area was drier, and water resources restricted, perched ponds and sinkholes would have afforded the inhabitants potable water. The vegetational cover at the time would have consisted of mixed forests of oak and pine through much of the area, with grasses covering the drier elevated areas. By the Middle Archaic period, pine was replacing oak as the dominant cover (Watts 1975), and water resources were available from rivers and springs, as well as sinkholes. By the Late Archaic, the environment was approaching modern conditions. The swamps and wetlands characteristic of the modern setting would have developed, and the vegetational and topographic patterns would have been similar to the present. Contemporary Environment The general land area of which Cowhouse Creek is a part contains three distinct plant communities: riverine forest, pine flatwoods, and oak/pine sandhills (Burch 1974:2). Differences in vegetative cover are primarily the result of drainage patterns and the availability of water. These in turn affect the animal populations that occupy a given area, as well as the relative abundance of the fauna. The riverine forest is dominated by cypress forest


50 composed of bald cypress with mixed hardwoods and softwoods interspersed throughout. Common hardwoods include water ash and southern red maple (most common), and sweetgum, water hickory, and blue beech (frequent) (Burch 1974). Breaks in the canopy allow button brush, cornel, Florida privet, and other water-tolerant shrubs to grow. Royal fern and smartweed are also common. Epiphytic tillandsia are common in the trees. Hardwood riverine forests are also fairly common, and can be divided into two main types. Generally, these are less uniform than the cypress dominated riverine communi ties, and the boundaries are often difficult to delineate, as they merge w ith the cypress dominated areas (Burch 1974:4). Bayheads develop in areas in which there is standing water for most of the year, and are dominated by redbay, swamp bay, and loblolly bay. Associated shrubs include gallberry, fetter bush and wax myrtle. Several sedges and chain fern are also common. Mixed hardwood riverine forests contain an occasional cypress, though for the most part are dominated by southern red cedar and palmetto. Water and laurel oaks, sweetgum, blue beech, American elm, water hickory, southern red maple, black gum, and water ash are common tree species. Water locust, palmetto, southern red cedar, and southern magnolia are predominant along rivers and permanent ponds. The shrubs of the mixed hardwood riverine community are diverse, although common species include Florida privets, wax myrtle, button


51 bush, cornel, American strawberry bush, gallberry, Hawthorn, and chicksaw plum. Several ferns are also locally frequent. Tillandsia are epiphytic in this community as are the orchids (Burch 1974:5). Grapes, Virginia creeper, and other vines are common to both the wet and dryer areas. Drier areas associated with slightly higher elevations contain the larger oaks and a saw palmetto undergrowth (Burch 1974:6). The pine flatwoods are dominated by pine, generally with a shrub layer primarily composed of saw palmetto. Cypress heads and strips occur along streams and wetter areas, and bayheads occur around semi-permanent ponds. Burch (1974:7) defines the flatwoods as the area in which "the drainage pattern is altered by the presence of an organic hardpan." Along ponds and wet areas, cypress heads and bayheads similar to the riverine forest occur. The cypress heads and bayheads are "wetter for a longer part of the year than the pine/palmetto community which makes up much of the area, and differ in that the bayheads tend to occur on areas that may stay wet all year while cypress is able to tolerate periodic fluctuations in the water level" (Burch 1974:7). These communities also differ from the riverine areas, in that these have a marked "edge effect" and a particularly rich flora at their margins (Burch 1974:7). The pine flatwoods also include much of the land which has been cleared and grazed in historic times. Slash pine now largely replaces longleaf pine throughout. Shrubs are


52 dominated by saw palmetto, although grounsel-tree, wax myrtle, holly, bumelia, and chicksaw plum are locally common (Burch 1974:8). The oak/pine sandhills encompass, for the most part, the elevations surrounding Cowhouse Creek at elevations above the 40 foot contour line (Burch 1974:2). These areas are generally rich with herbaceous and shrubby plants, although modern land modifications along these elevations have disturbed this vegetational community extensively (Burch 1974:9). Turkey oak and live oak are dominant tree forms in this type of community, and pines are common. Persimmon, wiregrass, milkweeds, greeneyes, camphorweed, yucca, and many other plants are common Approximately 41 species of mammals are present in the general area, including 1 marsupial, 3 insectivores, 8 bats, 1 edentate, 9 carnivores, 15 rodents, 2 lagomorphs, and 2 ungulates (Brown 1974:68). Animals such as the common opossum, yellow bat, nine-banded armadillo, raccoon, gray squirrel, cotton mouse, and cotton rat are abundant throughout the area, although the nine-banded armadillo wa s only introduced to the area in the beginning of this century. The wild pig, another common mammal introduced by Europeans, is common to the floodplain forests and cypress swamp environments. reported throughout bobcat, black bear, Other common mammals that have been eastern Hillsborough County include cottontail rabbit, gray fox, mink, several species of squirrels, panther, opossum, river otter,


53 raccoon, striped skunk, and white-tailed deer, although as habitats are destroyed by development, many of these species are becoming less abundant. The amphibians and reptiles of the area represent a complex and diverse assemblage, with approximately 71 species present in the Lower Hillsborough Flood Detention Area (McDiarmid and Godley 1974:29). This includes 21 species of amphibians (5 salamanders and 16 frogs), and 50 species of reptiles (including alligator, 10 turtles, 1 amphisbaenian, 10 lizards, and 28 snakes) (McDiarmid and Godley 1974:33). Common species within the area are alligator, gopher tortoise, soft shell turtle, snapping turtle, box turtle, diamondback rattlesnake, cottonmouth moccasin, water snakes, the greater siren, frogs and toads. Birds are also abundant in this area, which is particularly important for breeding, winter residence, and as transient habitats for approximately 200 species, although only 100 species occur in significant numbers. Passerines, geese and ducks, vultures, crows, hawks, herons, woodpeckers, owls, and wild turkeys are common species. Most of the species require several of the habitats for survival, typically foraging over large areas (McDiarmid and Godley 1974:61). Breeding birds were dominated by five species, which comprised over 80 percent of the identified breeding species in the Lower Hillsborough FDA biological assessment: Carolina Wren, Parula Warbler, Red-eyed Vireo, Cardinal, and


54 Blue-gray Gnatcatcher. The winter population was somewhat larger, although this is affected to some degree by habitat destruction, and the resulting shifts in migration patterns and available wintering locales. Freshwater fish are also abundant, with approximately 43 species known to inhabit the Hillsborough River drainage system (Barnett 1972). Three major plant habitats in the area are important to the maintenance of the fish population: shallow runs in heavy swamp forests, areas of moderate to swift currents in less dense areas of swamp forest, and, large open areas of relatively deep water with little or no current (Cowell 1974). Common species include mudfish, gar, sunfish and largemouth bass, and catfish. Aquatic invertebrates are also common, with a total of 143 species collected during the Lower Hillsborough FDA biological assessment (Cowell 1974:87).


55 CHAPTER FOUR PREVIOUS ARCHAEOLOGICAL INVESTIGATIONS IN THE VICINITY Early research in the Tampa Bay area was mainly descriptive work with detailed accounts of excavations of shell middens and mounds. These early ventures were often poorly controlled, and the interpretations of the sites and the excavated materials largely speculative. Early excavators of sites in Hillsborough County include J. H. Allen (1846), T.A. Conrad (1846), Daniel Brinton (1859, 1867), Jeffries Wyman (1870), R.E.C. stearns (1870, 1872), A.W. Vodges (1879), and James Shepard (1886). The first comprehensive report on archaeological sites in the Tampa Bay area was made by S.T. Walker in the late nineteenth century (Walker 1880a, 1880b, 1883, 1885). At the turn of the century, C.B. Moore (1900, 1903) located and excavated a number of burial mounds in the area. Much of his work was incorporated in subsequent syntheses of Florida archaeology, particularly due to the great number of sites he excavated, as well as his propensity to keep detailed records of his work. W.H. Holmes (1894, 1903) incorporated Moore's ceramic collections in his study of ceramic type distributions in the eastern U.S., although Holmes never excavated in Florida. Many other well known figures in archaeology have


56 worked in this area: F. H. Cushing ( 1897), J. w. Fewkes (1924), M.W. Stirling (1930, 1935), Gordon Willey (1949), J. Clarence Simpson (1937, prior to World War II. 1938, 1941), and numerous others Stirling (1930) excavated at the Safety Harbor site, which became the type site for Safety Harbor culture. Fewkes (1924) and Stirling (1930, 1935) excavated at Weeden Island, which became the type site for Weeden Island culture. Simpson, as an archaeologist with the Florida State Archaeological Survey from 1935 to 1938, oversaw the excavations of 11 sites by Works Progress Administration (WPA) crews, five of which are in the Hillsborough River basin. Unfortunately, these excavations were not explained in detail by Simpson ( 19 3 7) Bullen (1952) later compiled what was available of the materials and site information. Although there are several references to archaeological materials predating ceramics, little attention was paid to Paleo-Indian and Archaic lithic scatters until relatively recently, as a result of cultural resource managementrelated investigations. One notable exception was the work of Simpson ( 1938, 1941), who made several references to quarry sites, including a reference to the Cowhouse Creek area (Simpson 1938:63). Many of the sites referenced by these early investigators were later destroyed as Tampa Bay and surrounding areas rapidly grew during and after World War II. Excavations in the Tampa Bay area during the 1950s and


57 early 1960s focused on artifact description in order to produce classifications of artifact types and determine cultural sequences, similar to previous investigative efforts. Sites investigated include Battery Point (Bullen and Bullen 1953, 1954), Johns Island (Bullen and Bullen 1950), Maximo Point (Sears 1958; Bushnell 1962), Perico Island (Bullen 1950), and Terra Ciea (Bullen 1951). By the middle 1960s, archaeological methodologies began to reflect the use of organized research designs in order to answer specific questions about cultural behavior. Archaeological investigations utilized concepts and methodologies from other disciplines in addressing archaeological questions. Among sites excavated under these new concepts in this area were Warm Mineral Springs (Cockrell and Murphy 1978; Clausen, Brooks, and Wesolowsky 1975), Little Salt Spring (Clausen et al. 1979; Clausen and Emiliani 1979), and many other recent investigations including the Interstate-75 project and Tampa Palms project archaeological investigations. There are many recorded sites within this part of the Hillsborough River Basin. Most of these sites are not the result of any one systematic survey and often lack essential data. Several recent surveys and excavations have added significantly to the understanding of the prehistory of the region, and have allowed for data essential to a synthesis of the area's prehistory to be developed. One of the early surveys in the area was undertaken


58 prior to the construction of the Tampa By-Pass canal (Seabury et al. 1975). The survey located five previously unrecorded sites, and assessed several others in the vicinity of the canal route. Unfortunately, a significant portion of the survey area was not accessible to the survey archaeologists (J. Raymond Williams, personal communication, 1990). Thus, several areas which might have yielded significant sites, including the possible earlier identification of Paleo-Indian sites in the Thonotosassa area, were not recorded. The L211 Calvin Jones Impoundment in 1978, was Area Survey, conducted by B. the project which originally located and recorded the Cowhouse East Head and Cowhouse West Head sites, as well as several adjoining sites. No report of this work was published, although Florida Master Site File forms were submitted for previously unrecorded sites. In a survey of seven proposed recreation resource areas within the Hillsborough River Basin (Daniel, Wisenbaker and Fryman 1979) 19 sites were located andjor assessed, in areas totaling approximately 3879 acres. Ten of the sites were deemed significant (Daniel, Wisenbaker, and Fryman 1979:78). Only limited subsurface testing, however, was conducted throughout the seven areas. A survey of the proposed Fletcher Avenue Park (McCullough and Fisher 1978) located and recorded five prehistoric sites. These were investigated in detail during


59 a second phase of work (McCullough 1979). The Parking Lot site ( 8-Hi-462), was initially believed to represent a Safety Harbor period habitation site based on the recovery of a Pinellas projectile point and st. Johns Plain ceramics (McCullough and Fisher 1978). The determination of cultural period was revised to the Late Archaic after sixteen 2x2 meter excavation units yielded Florida Archaic Stemmed points and a Thonotosassa point (McCullough 1979). The Canoe site (8-Hi-459) consisted of a small scatter of Wakulla Check Stamped and Belle Glade ceramics (McCullough and Fisher 1978). Water covered most of the site, and there was little additional testing to determine accurate boundaries and site function. The Lettuce Lake site (8-Hi460), was believed to contain two culturally related components, whic h were thought to be contiguous throughout the site (McCullough and Fisher 1978). Twenty-three 2x2 meter excavations yielded Hernando points and Glades Plain pottery, which were dated to the Perico Island (now Manasota) period (McCullough 1979). The Playground site (8-Hi-461) could not be assigned to any cultural period in the initial testing period (McCullough and Fisher 1978), although twelve 2x2 meter excavation units yielded several Florida Archaic Stemmed points. The site was interpreted as a Late Archaic camp (McCullough 1979). A major highway project in the early 1980s was the construction of the Interstate-75 bypass, a large portion of which was in this part of Hillsborough County. The corridor


60 itself was a 41.2 mile route, approximately 300-400 feet wide, although from Interstate 4 to the Hillsborough River the corridor was approximately 1320 feet wide (Jones and Tesar 1982:59). Thirty-one sites were located and assessed within the right-of-way, and 13 sites were selected for Phase II excavation. Nearby sites investigated as part of the Interstate-75 project are discussed below in more detail (Figure 4). The Phase II sample was chosen in order to provide a sample of cultural phases and periods within the three physiographic zones covered by the right-of-way--the Gulf Coastal Lowlands, the Polk Uplands, and the Hillsborough River-Zephyrhills Gap, as well as a sample of sites within micro-environmental zones (Jones and Tesar 1982:60). Special attention was given to sites which appeared relatively undisturbed, and which had the highest concentrations of artifacts or greatest vertical depths. Specialized activity sites such as quarry sites were also given a high priority in the planning of excavation strategies (Jones and Tesar 1982). Although the research goals at the sites took somewhat different courses, the general design was to identify the periods of occupation, locate the site boundaries and any activity areas within the sites, and to describe the functional aspects of the sites. Four of the sites were recommended for Phase III excavations: 8-Hi-472B, 8-Hi-473, 8-Hi-476A, and 8-Hi-507. As a compromise with tight road construction schedules


. POLK .. !'lANA TEE 10 0 M itts '2. BOAEI:EH:r:EADHa:J:9==:==:==:3'.0 I< i lomt 1 tr s SCALE FIGURE 4. Investigated Sites in the V icinity of the Cowhouse East Head and Cowhouse West Head Project Area. 61 j


62 however, decisions for Phase III work were made within two to four weeks after completion of Phase II, and comprehensive research questions were formulated before cultural materials from Phase II were analyzed and assessed (Jones and Tesar 1982:61). Thus, while the schedule fostered salvage work at the sites which were destroyed soon afterwards, attempts at rigorous problem-oriented research could not be performed due to the time constraints. The Landfill site (8-Hi-99) was a 31 acre site located in the eastern half of the southeast quarter of the southeast quarter of Section 12, Township 28 South, Range 19 East, the southeast quarter of the southwest quarter of the southwest quarter of Section 7 Range 20 East, and the northwest quarter of the northwest quarter of Section 18, Range 20 East. The site was originally called the Sanitary Fill site by Bushnell (FMSF 1963). Jones divided the site into four sub-areas (A-D), and assigned it a Middle Archaic period affiliation based on Newnan points recovered by local collectors (Jones n.d.:17). Phase II investigations (Hardin 1982) identified the site as a short term extractive camp, associated with the hunting and butchering of animals predominantly during the Middle and Late Archaic periods. Although a suwannee point diagnostic of the Paleo-Indian period was recovered, it was at a depth less than most of the later artifacts. The site was characterized by a significant amount of damage to the integrity from various modern activities, including increased erosion resulting


63 from land clearing and refuse dumping at the site. The Fowler Avenue site (8-Hi-393 ) is located in Township 28 South, Range 19 East, in the southwest quarter of the southeast quarter of the southwest quarter of Section 12 (FMSF 1975), and encompassed an area of approximately five acres (Jones n.d.). Jones notes the close association of this site with a chert quarry area located immediately north of Fowler Avenue along the Hillsborough River (Jones n.d.:16). The site was originally recorded as a quarry by Williams (FMSF 1975), and the quarry source noted by Jones was apparently included in the original assessment. The original site name was the Boat Launch site (FMSF 1975). Phase II investigations (Gagel 1981) revealed that the Fowler Avenue site was a short-term settlement dating to the Middle Archaic period. Underwater investigations were also conducted there (8-Hi-393cjuw), from which were recovered a substantial array of faunal remains, including mastodon, giant land tortoise, horse, and deer, in addition to a dense concentration of lithic debitage. No association could be drawn between the megafauna at the site and the lithic material due to the depositional nature of the river, and the lack of adequate excavation controls (Palmer, Dunbar, and Clayton 1981:5). The debitage from the underwater excavations was predominantly unaltered lithic material and debitage resulting from the initial stages of lithic reduction. The Fletcher Avenue site (8-Hi-450) occupies a sand


64 ridge in the southwest quarter of the southwest quarter of the southeast quarter of Section 1 of Township 28 South, Range 19 East. Originally, the site was recorded as the Power Line A or Power Line Area site (FMSF 1965, 1977). Phase II investigations at 8-Hi-450D identified an Early Archaic habitation, and a Middle Archaic occupation, during which the site apparently functioned as a camping station (Daniel and Wisenbaker 1981). The Wetherington Ridge site (8-Hi-472) is located in the southwest quarter of Section 6, Township 2 8 South, Range 20 East. Jones notes that this was the largest (approximately 59 acres) of the archaeological sites to be impacted by Interstate-75 construction (Jones n.d.). Occupational areas surround two large sinkholes on the crest of a sand ridge (Jones n.d.). Jones also identified eight habitat ion or activity areas during the initial survey (Jones n.d.:18). Data from the phase II investigations suggest that Wetherington Ridge was occupied from the Paleo-Indian period through the Transitional period (Haviser 1983). The site was described as a base camp for much of its occupational history, based on the relatively large size of the site, and the wide variety of tool types present (Haviser 1983:76). Phase III investigations focused on the Late PaleoIndian/Early Archaic, and the Transitional period occupations, and were aimed at gaining a better understanding of the functional and spatial attributes of


65 the site. These periods were best represented in the earlier investigations, particularly in sub-area B (Johnson 1985). During the Late Paleo-Indian/Early Archaic, the predominant tool forms were general utility tools, with a significant number of these indicative of use on hard wood or bone (Johnson 1985:95). Later components in this area of the site reflect less intensive use. Activity sets during the Transitional period suggest a subsistence procurement station (e.g., butchering, meat and plant processing) (Johnson 1985:96). The Wetherington Island site (8-Hi-473) is located on the north side of Cowhouse Creek, in the northeast quarter of the southeast quarter of Section 12, Township 28 South, Range 19 East, just south of the Wetherington Ridge site. Jones originally designated it the Wetherington Valley site, and noted the extensive quantities of silicified coral and chert debris (Jones n.d.). The site functioned as a quarry, although Chance (1980, 1981, 1983) later termed it a "lithic procurement station." Activities at the site were limited to early stage lithic reduction. Chance and Misner (1984) note that other activities at Wetherington Island were spatially segregated from the areas of lithic procurement. During Phase II investigations, 14 units were excavated, although only three 1x1 meter quadrants were analyzed for the Phase II report ( 1982) The site was placed in a temporal framework of Middle to Late Archaic. This determination, however, was based on a single


66 diagnostic point {Newnan) dating to the Middle Archaic, and the lack of ceramics at the site. Phase III work was aimed at adding to the understandings concerning temporality, function, activity location, spatial distribution, and site function (Chance and Misner 1984:16-21). While diagnostic points from the Middle and Late Archaic periods were recovered, evidence also indicated a Late Paleo-Indian occupation {Chance and Misner 1984:212). The original functional designation of the site as a predominantly lithic procurement and early stage reduction area was supported by the Phase III investigations. Variability, which was expected to reflect the lithic reduction continuum through the site (essentially from south to north) was documented, although it was not as apparent as was initially expected (Chance and Misner 1984:224). Other activities at the site apparently were very limited, although there was some indication that implements recovered were used to work relatively soft materials. These tools were apparently clustered in an area along the creek (Chance and Misner 1984:225). The Deerstand site ( 8-Hi-483) is located along the eastern side of a ridge that extends north-south in the eastern half of the northeastern quarter of Section 1, Township 28 south, Range 19 East. It is located on a meander loop of the Hillsborough River. Jones identified three cultural sub-areas (A-C), two of which were recommended for Phase II testing (Jones n.d.:20). Sub-area


67 A was classified as a lithic workshop and camping station, and appears to have been intermittently used during most of the Archaic, although there was evidence of an ephemeral occupation by ceramic period groups (Daniel 1982). Daniel notes that biface manufacture was the predominant activity, though ceramics and possible butchering tools were recovered. Daniel notes that these instances are probably the result of discrete events rather than the general behavior pattern at the site location (Daniel 1982:161). Sub-area B was reported separately. The lithic assemblage consisted primarily of late stage lithic material, with tools at the site almost all scrapers and knives, indicating butchering, hide preparation, and vegetal food preparation (Gagel 1981). The Harney Flats site (8-Hi-507) has undoubtedly become one of the most important in Florida. It was originally recorded by Jones as the Anderson site. No Paleo-Indian materials were recovered, although a Paleo-Indian occupation was assigned based on artifacts recovered by local collectors (Jones n.d.). Phase II investigations (Daniel and Wisenbaker 1983) were aimed at identifying the Paleo-Indian component at the site. Most of the Paleo-Indian material was encountered in a 60 em band within a sandy zone between 1 and 1. 6 meters below the surface. A Middle Archaic (Newnan) occupation overlying the Paleo-Indian (Suwannee and Bolen) occupation was identified, although only a limited effort was aimed at understanding the Archaic


68 occupation. Phase III excavations (Daniel and Wisenbaker 1987) consisted of a concerted effort at understanding the PaleoIndian occupation, particularly the analysis of a PaleoIndian tool assemblage from the site, and the assessment of spatial organization. Three large open-area excavations were conducted towards this goal. Although there was a lack of recognizable intra-site patterning, the work at Harney Flats was the first systematic analysis of Paleo-Indian materials from a professional excavation in Florida, as well as a body of data available for comparative purposes. Additionally, as its importance was realized, several other lines of research aided in the interpretation of the Harney Flats site, particularly the analysis and replicative experiments of use-wear on stone tools recovered from the site (Ballo 1985), and an assessment of raw material selection during the Paleo-Indian period (Upchurch n.d.). Other recent investigations in the area are also of note, particularly the Tampa Palms project, which consisted of several years of work at sites which were to be destroyed as a result of planned residential and commercial construction. Investigations in 1981 (Austin and Ste. Claire 1982) consisted of excavations at three sites. The Tampa Palms site (8-Hi-557) was dated to the Early to Middle Archaic periods, with several apparent functions indicative of a short-term residential camp. The Priscilla site (8-Hi-559)


69 had a Late Archaic cultural affiliation, and was probably a "special extractive camp associated with a collecting strategy" (Austin and Ste. Claire 1982:212). The Rock Hammock site (8-Hi-556), dating from the Transitional through Weeden Island periods, was a special activity site associated with lithic procurement and stone tool production (Austin and Ste. Claire 1982:212). Although the three sites represented a diverse group of temporal and functional classes, the authors were able to document trends in subsistence strategy, intensity of occupation, and lithic technology. What makes this research important, however, is not the unique nature or inherent significance of the sites investigated (a characteristic of sites chosen for Phase III excavations during the Interstate-75 project), but rather the use of explicit problem-oriented research designs, and the interpretation of the data within a broad theoretical framework. Excavations at the Marita and Ranch House sites were conducted in 1983 (Estabrook and Newman 1984). The Marita site (8-Hi-558), which was occupied from the Transitional through Historic periods, was probably an intermittently occupied extractive camp. The most intensive occupation at the site was apparently during the Weeden Island period. The Ranch House site (8-Hi-452) consisted of a series of occupations which were apparently not superimposed vertically, but could be isolated horizontally (Estabrook and Newman 1984:324). Area A is described as a limited


70 activity area, dating to the Middle or Late Archaic period. Area B is a multicomponent, multiple activity area, occupied from the Early Archaic through Weeden Island related or Safety Harbor periods, although the major occupation of the site apparently occurred during the Middle Archaic (Estabrook and Newman 1984:316). Area c consisted of a spatially isolated limited activity area, in which the dominant activity was the production of stone tools. Limited evidence suggests a date of Transitional or Early Manasota for this area (Estabrook and Newman 1984:318). The Hidden River DRI survey (Deming et al. 1984) located six previously unrecorded sites (8-Hi-2121 through 8-Hi-2126), evaluated the extent of 8-Hi-461 (discussed above) and tested three recorded sites ( 8-Hi-450, 482, 483). Each of the newly located sites was described as limited activity sites and were not deemed significant (Deming et al. 1984:95). The lack of diagnostic tools prevented the determination of a temporal span for sites 8-Hi-2121 through 8-Hi-2123. site 8-Hi-2124 was dated to the Florida Transitional based on the recovery of sand tempered plain and sand and fiber tempered pottery. Deming et al. (1984) suggest that 8-Hi-2125 was a multicomponent site, dating to the Middle Archaic based on the recovery of Newnan and Levy points, and ceramic periods based on three sand tempered plain pottery fragments. Site 8-Hi-2126 was assigned to the Safety Harbor period, based on the recovery of a single Pinellas point. This, however, appears only to


71 be an ephemeral occupation (Deming et al. 1984:91). Site 8-Hi-482, was originally recorded in 1972 by Karlis Karklins, and revisited in 1977 by John Fioroni, who named it the Powerline B site. It was visited by B. Calvin Jones during the survey of the Interstate-75 right-of-way, and was deemed not significant. Deming et al. ( 1984) identified four discrete occupation areas within the site boundaries based on artifact distributions. The depth of the cultural materials was greater than the tests, and the vertical density of artifacts appeared uniform throughout the tests. Diagnostic pottery recovered indicates an occupation dating to the Florida Transitional through the Weeden Island related periods, although it remains unknown if the lower, unexcavated portions of the site contained an earlier occupation. The Fletcher Avenue site (8-Hi-450) adjacent to the south (also called the Powerline A site), and the Deerstand site ( 8-Hi -48 3) were investigated during the Phase II investigations of the Interstate-75 project, as discussed above. Recently, a report was compiled by Weisman and Marquardt (1988) as a management tool for archaeological sites on SWFWMD properties. While the purpose of this report was to create an archeological inventory of sites on SWFWMD lands in hopes that this would facilitate management, the report lacks sufficient data on each of the sites evaluated. A more beneficial approach (which would probably


72 have been a more effective management tool) might have been to update Florida Master Site File forms based on field evaluation of the sites, and expand on the recommendations for each of the sites. Apparently there was little, if any, field assessment of these sites. There were also no provisions made for their frequent recommendation "revisit site to determine current condition and develop evaluation of significance" (Weisman and Marquardt 1988). In this compilation, both Cowhouse East Head and Cowhouse West Head sites are listed, although as stated previously, the sites were not reassessed for the report. That site inventories and relevant information on recorded sites in the area are incomplete is not debatable. Few sites on the Florida Master Site File have adequate cultural/temporal, spatial, or any sort of controlled quantitative data, and in areas which have been resurveyed, sites are often combined or resubmitted to the Florida Master Site File as new sites, with little or no consistency. Creek sites In many cases, particularly with the Cowhouse surveyed by the Interstate-75 and L-211 Impoundment Area surveys, cultural materials were often not collected from the field or in some cases not cataloged or curated. Site boundaries are often vague, and there is almost never any discussion of the rationale behind site size determinations. Although the Interstate-75 survey attempted to remedy some of these problems by designating sub-areas to sites, these problems are not likely to be


73 overcome without an increased emphasis on gathering adequate data through archaeological survey. As the archaeological resources within this area are being destroyed at a fairly rapid rate due to developmental pressure and looting, combined with the generally low level of funding for archaeological research, the time has come for a rethinking of archaeological survey methodology.


CHAPTER FIVE RESEARCH BACKGROUND Physiography o f Cowhouse Creek 74 The Cowhouse East and West Head Project area encompassed approximately 150 acres, although the spatial extent of cultural material was much greater (approximately 200 acres or more), and would have required more intensive efforts at determining accurate site boundaries. The Cowhouse East Head site is located in Township 28 South, Range 20 East, Section 4, predominantly in the northwest portion of the southeast quarter (Figure 5). The highest density of artifacts is associated with an elevated sand ridge which juts out into the lower floodplain of the Hillsborough River and Cowhouse Creek. The vegetation cover on this elevated area consists of pine and saw palmetto. On the lower parts of the site, vegetation grades into the riverine forest, dominated by cypress and other water-tolerant vegetation. The investigated portion of the Cowhouse West Head site is located in the north half of the southwest quarter (Figure 5). The great spatial extent of this site and the difficulty in determining site boundaries and differentialuse areas within the site boundaries make it difficult to characterize, although the investigated portions of the site


I ( 3 2 ) - o t r r & ::::. J...f ----_ g ' ' '--' ' ---.... --' ... / ' ---..:::--> c:::J ----/ -0 ----::-lll+, l N I -f ( I -.,-...;:..__._ ____ 0 \ Trai lev -:::} -"PJ: /--... -!J ,16 ''MORRIS -\..:::: ...... \ 0 : t o o o o /.] T o 0 o 0 o o o o 0 o ......-,' .. 7 .. . M "' / :) /' --" -"", :.. . S 8-Hi-496 l_) . ..,--=--/ L. ... _ .... _J OHIO ,c': ..... ... THONOTOSASSA QUADRANGLE FLORIDA-HILLSBOROUGH CO. 7.5 MINUTE SERI ES ( TOPOGRAPHIC) PHOTORE V ISED 1987 1 SCALE 1:2 4 000 . . .... TQWNSHP 28 SOUTH RANGE 20 EAST SECTlONS 4 NJ 5 Ml CONTOUR INTERVAL 5 FEET FIGURE 5. Location of 8-Hi-495 and 8-Hi-496 Plotted on the U.S.G.S. Thonotosassa Quadrangle, 1987. 7 5


76 consist primarily of oak/pine sandhills on the northern and western portion, which changes t o a riverine forest as the elevation grades fairly steeply to the Cowhouse Creek floodplain. Cowhouse Creek lies within the Zephyrhills physiographic province at the border of the Polk Uplands (Puri and Vernon 1964: fig. 7), and is characterized by a relatively low topographic relief. Elevations range from a minimum of approximately 8.4 meters to a maximum of about 11.6 meters above mean sea level. Other sites are the product of very similar settings, and appear to consist of habitation sites along the ridges and elevated areas overlooking the creek, while specialized activity areas extend further away from the creek, and onto the creek floodplain. This conjecture remains to be tested, however, as the available data from the majority of the recorded sites along Cowhouse Creek are minimal. Drainage of the investigated sites Creek, which runs between the two Hillsborough River, which intersects immediately adjacent (Figures 4 and 5). Geology is to Cowhouse sites, and the Cowhouse Creek During the Mesozoic and Paleozoic geologic eras (600 to 135 million years ago) sandstone, shale and igneous rock were deposited through the Florida peninsula. These deposits are overlain by several hundred meters of carbonate rocks, principally limestone and dolomite, whic h were


77 deposited when the western Florida peninsula was covered by a shallow sea, during the late Cenozoic era ( 20 to 30 million years ago) It is these Cenozoic deposits of limestone which form the principal component of the modern Floridan aquifer. Limestone formations include Cedar Key limestone, Lake city and Oldsmar limestones (Inglis, Williston and crystal River formations), Suwanee limestone, and Tampa-st. Mark's limestone. Tampa and Suwanee limestone deposits outcrop around and within the project area, and are almost continuous along Cowhouse Creek. To the south, the limestone blankets are mostly covered by younger clastic deposits (Puri and Vernon 1964). As a result, the Hillsborough county area is the southernmost area in the southeastern United states to contain limestone and chert outcrops which would have been available for aboriginal procurement (Goodyear, in Daniel and Wisenbaker 1987). Overlying the limestone deposits is a sandy clay stratum which fills karstic voids in the limestone layers (Wright 1973). Upchurch et al. ( 1981:3) suggest .that this clay is the result of carbonaceous weathering of the limestone. Pliopleistocene and recent deposits consist of undifferentiated fine-grained quartz sand of marine and eolian origin. These deposits vary from several inches to several feet in thickness. overlying these deposits is often a thin layer of organic material of recent deposition.


78 The Cowhouse Creek valley is a small swampy floodplain which supports a hydric hammock and cypress as principal vegetation. Upchurch (1982:144-145) notes that: Cowhouse Creek ... constitutes either a partly abandoned chanel of the Hillsborough River, or a flood relief chanel that is used primarily in periods of high discharge and is now in a phase of downcutting to intercept the river. Since bedrock can be found in the channel the latter case is not likely, and it is best to presume that the creek is an abandoned channel of the river. If the former case is correct, then some interesting speculation as to the relationship of the river to early man can be made. There is a small amount of circumstantial evidence to support the hypothesis that the Cowhouse Creek valley was the main channel of the Hillsborough River within the last few thousand years. Evidence for this includes the exposure of hardpan above river level in some areas, orientations of tributary creeks such as Trout Creek and Cowhouse Creek in relation to the Hillsborough River, and the abundance of archaeological sites along Cowhouse Creek (Upchurch 1982:145). Soils Most of inland Florida is characterized by podzolic soils which are characterized by a black organic horizon overlying a podzol zone from which iron, aluminum, and most other weatherable minerals have been leached out (Limbrey 1975:137-141). These minerals accumulate in an illuvial layer above the parent material. Generally, the lack of nutrients in podzolic soils lead to restricted vegetation. Thus, the plants that are able to grow can tolerate low nutrient levels. Almy (1978), Grange (1978), Deming (1980) and others


79 have discussed the relationship between soil type and site location. Soils generally control the use or type of occupation of an area due to such factors as the drainage characteristics of soils (which allow for habitation in a given area) and the productivity of soils (which affects the natural vegetation or cultivated foodstuffs, and which in turn affects the faunal resources available in an area). Increasingly, however, archaeologists are looking at other areas beyond campsites and base camps, and how these are important to prehistoric populations. This includes areas which are in poorly drained soils, particularly resource locales and areas which might have been used for interment of the dead, a trait thus far particular to the Archaic period. Soil also affects the preservation of organic materials at archaeological sites, with acidic soils increasing the decay of organic matter. Most of the soils throughout Cowhouse Creek range from acidic to very acidic. The only class of artifacts common to all of these sites is lithic material in the form of debitage and tool forms. Soils in the project area consist of several soil types (United States Department of Agriculture 1989). Immokalee fine sand, which is somewhat poorly drained, and nearly level, extended through most of the excavated area of Cowhouse East Head. In this soil series, a typical stratigraphic cross-section would consist of a very dark gray fine sand A horizon from 0 to 8 inches, a light gray


80 fine sand E horizon from 8 to 36 inches, a black fine sand Bh1 horizon from 36 to 46 inches, a dark reddish brown fine sand Bh2 horizon from 46 to 52 inches, and a dark brown fine sand Bw horizon from 52 to 80 inches (United states Department of Agriculture 1989:77). Generally, the excavation units (EUs) placed in this soil type contained six zones, based on soil color, texture, and composition. While not differing greatly from the expected soil horizons, these were important in guiding excavation methodology, and thus played an important part in the Cowhouse investigations (Figure 6). Zone 1 was a light gray fine sand from 0 to 25 em below surface. In most cases, the top part of this zone was a slightly darker gray, which faded into a lighter gray in the top 10 em, as would be expected with a greater amount of organic material and debris at or near the surface. Zone 2 consisted of white fine sand to approximately 85 em below surface, which although slightly gray, was markedly different than the gray levels above. This zone contained the Archaic period component at the site. Zone 3 was described variously, from very dark gray, very dark brown, or very dark grayish brown highly organic packed sand, and ranged from 5 to 15 em in thickness. These layers were markedly different from zone 2. Zone 4 was described as a range between dark yellowish brown, yellowish brown, pale brown and very pale brown, even though there was a significant amount of variation within the zone. This zone began at depths of 90 to 100 em below


8 1 N 1002 N 1002 N 1020 N 1020 E 1004 EU 2 E 1006 E 994 EU3 E 992 o Nort h Wall Profile + + Sout h Wall P rofile 0 10 I 1 0 GRAY SAID LIQI! GR.\! 5.1..1(!) 20 20 30 40 40 so 50 II 60 iii!JTE SMD li!!Il'ESAJD 60 e E! () 7 0 70 () Q) Q) () 80 () co co ..... ..... ::I 90 90 ::I Cll ) Cll 100 0 100 ..... Q) Q) co 11010 .s:: 110 VERY PALE BRCllll SA.I(!) Yil.LOOSH BRQIII M I V .s:: .j.J .j.J 0. 0. 120 120 Q) c lJO 130 1 4 0 140 150 GRAY ISH BRCllll rum SAID 150 V I GRAY CIJ,YE'f 5.\Jll 160 ___ ____ _________ ______ 160 170 (Not Excavated) 170 180 180 0 50 100 150 200 t 0 50 100 150 200 (CIU) (em) Soil Zones FIGURE 6 stratigraphic Sequence a t 8-Hi-495.


82 surface, and contained the Paleo-Indian component of the site. Zone 5 was similar to zone 3, and was probably evidence of a forming hardpan resulting from water table changes due to the construction of the By-Pass Canal and other projects. Generally, this zone was a compact fine sand with an average thickness of 3 to 5 ern. Recorded colors varied from very dark brown, dark brown, very dark gray brown, and dark grayish brown. Zone 6 ranged from light gray, gray, and grayish brown in color, and generally extended deeper than the excavation units, although in EUs 1 and 2, zone 4 also extended deeper than the area excavated. Basinger, Holopaw, and Sarnsula soils (depressional), cover the three depressions in the site area of Cowhouse East Head. Generally, these are the soil associations for shallow ponds with grass, although these areas were dry for the majority of the field season. Generally, Basinger soil is along the exterior of the depressions, while Holopaw and Sarnsula soils are in the interiors of the deeper depressions. Although testing was done in an area with this soil association, no excavation units were placed in these areas. Chobee sandy loam, which is covered with water for most of the year except during the dry winter season and early spring seasons before the summer rains, is deep and very poorly drained. The former classification of these mixed soils as Peace River soils was based on their not having been accurately mapped by the Soil Conservation Service


83 because these areas are generally too wet and densely vegetated for accurate mapping (United States Department of Agriculture 1950:31). These soils cover most of the lower areas along the floodplain of this portion of Cowhouse Creek, including EU 6. Winder series soils consist of deep, poorly drained soils, formed in loamy marine sediment. They are in broad sloughs, and basically cover the Cowhouse slough to the southern part of the project area, and two large pockets along the slope to Cowhouse Creek, between the sites. These soils are fine-loamy and silicious. These soils are closely associated with the Chobee series discussed above. Myakka fine sands consist of deep, poorly drained and very poorly drained soils, and cover the higher elevations above the floodplain. EU 11, which was placed on the higher, relatively flat ridge, above the slope into the Cowhouse Creek floodplain, was in this soil association. The remainder of the excavation units at Cowhouse West Head were in an area that appeared transitional to Winder fine sands and the Myakka series soils. While a soil zone classification was developed for the main area of Cowhouse East Head, it must be stressed that these divisions were noted to allow for comparisons throughout the area in which several excavation units were placed. The use of soil zones was less meaningful in other areas of the sites. As EU 6 and the excavation units at Cowhouse West Head were placed at some distance apart, these


84 areas could not be compiled as easily or as accurately into soil zone designations. Lithic Resources Stone was an important resource to prehistoric populations, and one of the attractive forces behind settlement along Cowhouse Creek. Simpson (1937, 1941) was among the first to note the importance of the chert resources in this area to the production of tools during prehistoric times. Along the south bank of the Hillsborough, in the region known as the "Cowhouse, 11 along the lower reaches of Cypress Creek, and along the entire course of Flint Creek, can be found evidence of the most intensive quarrying operations so far found in the county (Simpson 1937:63). Simpson (1941:32) suggested two reasons for the intensive exploitation of the chert resources of this area: the generally high quality of the chert; and, as the southernmost occurrence of accessible chert, it would have been particularly important to groups to the south of the Tampa Bay area. Upchurch et al. (1981:1) define 'chert' as a "fine-grained sedimentary rock composed of silica in various mineral phases ... frequently containing minor impurities." Silicified replacement limestone and agatized coral are the result of the silicification of two different host materials: the former of the limestone strata and the latter of aragonite coral heads of Miocene age (Upchurch et al. 1981). Replacement cherts are formed by the silicification


85 of limestone through the precipitation of silica-rich groundwater. Generally, the original rock fabric is poorly to moderately preserved. This type is widespread, and is associated with the Tampa, Suwanee, and to a lesser extent, the Hawthorn formations. It is often within the montmorillonitic clay referred to as the silex beds that the silicification process begins ( Cooke 19 4 5 ) Altschuler et al. (1964) suggest a sequence of alterations from montmorillonites to kaolinite, in which silica could be released into groundwater. They suggest that the cherts of the Miocene were formed from these solutions. Replacement cherts are the most common form of chert in the area, and can be seen throughout the course of Cowhouse Creek. Lithic procurement and quarry areas along Cowhouse Creek include Rock Hammock, the Fowler Avenue Bridge site, Wetherington Island, and the Cowhouse West Head site, although outcrops may be found throughout the course of the creek. Upchurch ( 1980) notes that prehistoric quarry sites tend to occur on bedrock highs, which may be formed by the differential weathering of bedrock, and the resulting high areas of weathering-resistant cherty areas. Upchurch (1980:51) suggests that the replacement cherts are the most workable cherts in terms of aboriginal tool production: The (replacement) cherts should be easily extractable and water saturated. Water saturation is necessary because it improves the ease with


which the chert can be flaked. Heat treatment reduces the need for wet cherts. The green clay beds which overly the Miocene section in this area fill both requirements. First, the cherts are generally in relatively small blocks with little limestone matrix remaining to be cobbled off. Second, the clays keep the chert moist. Therefore, only certain types of outcrops would be preferred. 86 Several recent studies have attempted to identify quarry origins of lithic material found at archaeological sites: Purdy and Blanchard (1973), Purdy (1976), Upchurch (1980), and Upchurch et al. (1981). Upchurch et al. (1981:9) emphasize the use of the quarry cluster concept in determining the provenience of chert. These 'quarry clusters' are locales within the exposure belt of a single formation which is known to contain numerous exposures of chert that are uniform in fabric, composition, and fossil content (Upchurch et al. 1981:9). Upchurch et al. (1981:8) note that: ... it has become apparent that identification of specific quarry sites is difficult, if not impossible. It is valid to use gross texture, fossil content, or trace element composition to determine the source of a chert specimen if it was found far from the chert-bearing strata. It is not valid to try to identify specific sites, when there are many within an area that have similar geologic settings. Nineteen chert quarry clusters have been identified for Florida (Figure 7) by Upchurch et al. (1981). As archaeological materials can be traced to their quarry cluster source fairly readily, quarry cluster analysis provides the means for testing settlement models of mobile populations.

PAGE 100


PAGE 101

88 Hunter-Gatherer Models Strategies for procurement, processing, and storing resources are central to the understanding of huntergatherer settlement systems. Binford (1980) suggests two basic strategies for hunting and gathering economies--foraging and collecting. Foraging strategies are based on the movement of "consumers to goods with frequent residential moves," while collecting strategies are based on the movement of "goods to consumers with generally fewer residential moves" (Binford 1980:15). Foraging strategies are generally applicable in cases where the landscape is relatively homogeneous and resources are more or less equally accessible or distributed. Distinctive to the foraging strategy is that foragers collect food and critical resources on a daily basis and do not generally store foods. Collecting strategies are adapted to differential landscapes where resources are not evenly distributed, and there is a need for logistically organized procurement of resources. In this type of settlement, groups move to the locale of a particular resource during times of abundance, and procure other necessary resources through the organization of special work groups. Archaeologically, a foraging strategy will be manifest in small residential bases, while a collecting strategy would result in larger, more permanent residential bases and a greater density and diversity of tool types. Small, special task camps would radiate from the camp, based on the

PAGE 102

89 closest distance to specialized resource areas which could be utilized from the central base. In a collecting strategy, resource areas will tend to be archaeologically invisible, with few or no tools present. site manifestations would tend to be spatially similar, given the assumption that the primary resources at a given locale are similar to those of other locales. In the case of lithic material acquisition, however, this assumption will not necessarily hold true. Goodyear (1979) suggests a model of chert use during the Paleo-Indian period based on the utilization of high quality cryptocrystalline raw materials. He suggests that the use of fine-grained cherts by Paleo-Indian groups reflects the high mobility of these groups. Durable stone tools would be preferred over lower grade materials because these materials would not be available for a greater part of the seasonal cycle. During the Archaic period, an increasing reliance on local, more coarse-grained raw material can be seen in tool inventories. Increasing population sizes and the decrease in territory favored a decreasing group mobility, and thus the exploitation of locally available lithic raw materials to a greater extent (Goodyear 1979:10). Ste. Claire (1985) suggests that this reliance on lower-grade chert created the need for a technological response, and is one reason for the increase in the use of thermal alteration of lithic materials during the Archaic periods.

PAGE 103

90 Austin and Ste. Claire (1982:29) suggest that both of these strategies were in effect during early prehistoric periods within the Hillsborough River Basin, with a gradual shifting in strategy from a mobile, foraging strategy during the Paleo-Indian period, to a more sedentary collecting strategy during the Archaic as a result of climatic changes. Daniel (1985:264) contends that the opposite is true: ... settlement patterns in the study area [Hillsborough County, along the Interstate-75 corridor) during the late Pleistocene to middle Holocene can generally be characterized as exhibiting a hypothetical trend from systems emphasizing logistical mobility in the PaleoIndian period to adaptations approaching a more foraging one sometime in the Archaic (Daniel 1985:264). Evidence for a more restricted pattern of mobility in this part of Florida comes from two sources. Environmental data (presented previously) indicates that water resources (and consequently floral and faunal resources) were somewhat more restricted during the Paleo-Indian period, satisfying the dispersed resources criterion which fosters a collecting strategy. Results from excavation at Harney Flats indicate for the most part, no exotic cherts were utilized in the production of tools, and that the only raw materials utilized were from the Hillsborough River and Upper Withlacoochee Quarry Clusters (Daniel, Wisenbaker and Ballo 1986 :50). Suwannee and Bolen points, on the other hand, appear to have been preferentially produced from pelodial grainstone-breccias (also known as 'Bay Bottom' chert) which

PAGE 104

91 is from a source now inundated by Tampa Bay (Goodyear et al. 1983:58). The portable and flexible tool kit of Paleo-Indian groups may indicate that incongruencies in resources may have necessitated increased mobility during part of the year. The locations of the sources of these lithic materials suggest an east-west pattern of movement. The restricted land mass of peninsular Florida as seas reached modern levels probably prevented Archaic populations from traveling as far, while the increase in the homogeneity of the resources resulting from climatic change probably made moves of such distances unnecessary. Thus, in this respect, Archaic populations were somewhat more sedentary. Daniel (1985:264-265) also cites the increased use of informally-made and expedient tool assemblages of the Archaic period as evidence for a decrease in the degree of residential permanence. Daniel (1985) suggests that band territories during the Archaic were also oriented along the main drainages to include both upland and lowland areas. He postulates that a subsistence scheduling system developed during the Archaic, characterized by periods of band aggregation and dispersal. This strategy allowed for the seasonal occupation of sites geared toward the utilization of specific target resources. According to Daniel: ... sites ... composed of a whole band (aggregation) were placed on upland borders during the fall and winter to principally hunt deer and gather nuts.

PAGE 105

During the spring and summer the group dispersed into small or extended family groups behaving like foragers employing a mobility strategy designed for coverage, thus seeking to maximize the "encounter" with resources. This would be in the lowlands utilizing lacustrine resources and gathering wild plants, and along the coastal rivers for fish and shellfish (Daniel 1985:266). 92 Daniel (1985) sees the climatic transition at the Middle Archaic affecting positioning tactics. He suggests that during the spring and summer, a foraging strategy was utilized, and there is a high degree of residential mobility. In the fall and winter, a collecting (logistical strategy) was utilized, characterized by aggregated base camps (Daniel 1985:266-267). According to Daniel (1985:268), these aggregated base camps would be located at the border of the Polk Upland and Zephyrhills Gap physiographic zones. Cowhouse Creek lies on this boundary.

PAGE 106

93 CHAPTER SIX THE COWHOUSE EAST HEAD AND WEST HEAD PROJECT Field Methodology After an initial walk-over of the sites (Figure 8), it was decided that the locations of concentrated looting were an adequate indicator of better areas for excavation, and to use the available time for testing of many of the lower probability areas to determine the nature of the periphery of the site proper. As the sites were so large spatially, this allowed us to concentrate on artifact densities in the lower and wetter parts of the sites in an effort to understand the utilization of these areas. By testing the lower probability portions of the sites, it was hoped that some new understandings about prehistoric utilization of areas which are now part of a wetland environment might be seen (c.f. Purdy 1988). It was also decided that this was the best strategy for dealing with an area which has been intensively developed in recent years. The topography has also been altered to a great extent, particularly with massive construction projects that leveled sites, and the construction of the Tampa By-Pass Canal which altered drainage patterns, and dried up much of the surrounding wetland environments. The incredible changes can be seen easily with a look at the USGS topographic maps

PAGE 107

FIGURE 8. \l/1,{ ____/ COWHOUSE EAST HEAO ANO COWHOUSE WEST HEAO PROJfCT AREA, If /27 feet 1990 1/ ...... 27.5 200 nl!: I in. 200 feet
PAGE 108

95 for Thonotosassa from 1943, 1974, and 1987. At the Cowhouse East Head site, tests were dug every 50 meters on an east-west transect through the highest elevation at the site. At the west side of the site, a north-south transect was tested along a 40 degree east of north line that basically followed the course of the creek. This line extended from the junction of Cowhouse Creek and the Hillsborough River south approximately 500 meters, until the land was too wet to test, and the artifact density dropped off (although it must be noted that this probably reflects the bias of the survey transect). On the east side of the site, a transect parallel to the previous line was tested, from the Hillsborough River to the powerline corridor. Although there were other areas which would have merited a closer examination, it was decided that testing would only cover the area immediately adjacent to the area we were prepared to excavate. This was done in the hope that any future work at the site would be preempted by further testing of the site's extent. All of the test units were backfilled upon completion of excavation and recording. At the Cowhouse West Head site, five test units were dug to determine concentrations of cultural material. Due to the amount of recovered materials and the long distance it had to be carried, fewer tests could be dug. The denser vegetation at the site made it difficult to move around, despite the many bike paths throughout the area. Ultimately, this made work at the site more difficult, as

PAGE 109

96 the curved bike trails were difficult to orient on a map, and because it seemed that there were new trails every time we visited the site. Areas were flagged for excavation, either because of their association with close proximity to looter pits (EUs 8, 9, 10, and 12), or their elevation and soil types (EU 11). The excavation units at this site were further separated, possible. Again, in order to cover as large an area as all of the test units were backfilled after excavation and recording. Each of the 50x50 ern test units were dug to a depth of at least one meter, determine if any and the bottoms were auger-tested to cultural material extended deeper vertically, and to document the stratigraphy below depths accessible by shovel testing. Although precise control was difficult deeper than 70 ern or so, this careful method of testing was based not on the need to recover cultural materials, but to allow for the documentation of stratigraphy throughout the site, artifact densities within the different soil zones, and in the development of the methodology used for excavation. Excavations at the sites were conducted over a seven week period, from early May to late June, 1990. Six excavation units (EUs 1 through 5, and EU 7) were placed in areas thought to be less disturbed by looting (Figure 9). EU 6 was placed in the low, swampy area, which was selected from the tested areas to contain a large amount of cultural material. The potentially interesting areas located by the

PAGE 110

COWHOUSE EAST HEAD SITE 8 -Hi495 M eters 0 4 0 8. 4 tj ..... ..,, ei ........... ........ I 8.9 FIGURE 9. HILI.SilOROOGH RIVER 8. 4 Datua Elc a v ation Unit (EU /) Lootin q Activity .. ... Site Map, Cowhouse East Head. ./ 1.0 -..J

PAGE 111

98 eastern north-south transect were left alone, as the training of students would not allow for the placement of excavation units out of supervision range. Artifacts mapped in situ were measured from the southwest stake, using a line level and a metric tape to record the depth below unit datum. Floor levels were recorded in the same manner. At Cowhouse East Head, excavation units were mapped from reference points marked by wooden stakes. These points were later tied into the permanent datum. Generally, excavation units were cleared of overburden, and dug at 10 em arbitrary levels, until soil color or texture changes were noted. All soil was screened through 1/4 inch hardware mesh screen. All of the excavation units at Cowhouse East Head, with the exception of EU 6 (which was 1x2 meters), were 2x2 meters, with 10 em columns retained at each corner to prevent the stakes from collapsing. Palmetto roots were left in place, until the soil surrounding them was cleared, and the roots were easy to clip and remove. These units were then excavated by natural stratigraphy, with divisions within these levels after each 10 em interval as noted above. Thus, new levels were initiated after changes in soil color, soil texture, or artifact densities were noted. The guiding principle in excavation was that levels excavated in this way could be lumped together in the lab if need be, while it is impossible to divide arbitrary 10 em

PAGE 112

99 levels after excavation. The impetus for the use of this excavation methodology was to stress the importance of the development of site specific excavation methodologies, rather than settling for the use of 10-20 em arbitrary levels in the excavation of deep sand lithic sites. It was hoped that excavation by natural stratigraphy would allow for greater control over artifact provenience, even though these levels were not necessarily associated with cultural levels, and may in fact have no cultural significance at all. However, the use of natural stratigraphic excavation allowed for a better understanding of natural and cultural processes affecting the site, which was an important concern given the amount of looting. It might also be argued that natural strata do in fact play a role in artifact provenience, particularly as we are able to document some degree of vertical movement of artifacts and its relationship to the archaeological stratigraphy, as well as the formation of stratigraphy. At the Cowhouse West Head site (Figure 10), all of the excavation units (EU 8 through 12) were 1x2 meters, with 10 em columns retained only at the southwest corner. Other than these exceptions, excavation methodology remained the same. The excavation units were placed in random areas along the main bike path, in order to keep them as spread apart as possible. Vehicle access at this site was limited, and we were only able to bring the truck to the top and bottom of the trail. This meant that all of the equipment

PAGE 113

100 ... ... COWHOUSE WEST HEAD SITE 8-H i 496 ll.O RIVlROO fOREST ----Ell 101 10.0 0 Meters 60 DatUJ Looting Activity Excavation U nit I EU I ) I 8.5 8 0 ( ( FIGURE 10. Site Map, Cowhouse West Head.

PAGE 114

101 had to be carried in by hand and that all of the artifacts (in most cases, bags of soil which needed to be waterscreened) had to be carried out. Screening was done using 1/4 inch hardware mesh. Soils such as the clays encountered at Cowhouse West Head which could not be screened in the field were bagged in doublelined trash bags and labeled with flagging tape inside and out. These were brought back to the lab and water-screened. Once an excavation unit was completed, crews cleared off the walls and photographed and mapped the stratigraphy on a 24 by 36 inch sheet of drafting vellum. While the major goal of this activity was to record unit profiles and to allow for the division of the site into zones based on similarities in stratigraphy throughout the site, it also permitted the assessment of the effectiveness and accuracy of excavation by natural strata. In cases where walls had varying profiles, more than one wall were profiled. After profiles were completed, column samples were taken, based on the field notes of excavation. Thus, soil samples were available for each level excavated. Special care was taken to ensure that samples were cleanly taken and labeled accurately. The size of these samples ranged from one half gallon to two gallons, depending on the amount of soil that could be removed without contamination from the other levels. The one exception to this was EU 6, from which was taken a full 50x50 ern column sample from the southeast corner of the south wall. Soil samples were used

PAGE 115

102 for soil analysis and flotation. Initial mapping was done during the excavation period, primarily to record the locations of the excavation units and to establish a permanent datum and several semi-permanent (wooden stake) reference points. Mapping activities were continued in late July, for several days in September, and again in December, when water levels were low and vegetation a little less thick. Several hundred readings were taken with the transit. The datum at 8-Hi-495 is the southwest corner of the powerline tower nearest to the excavated area (Tower number 148). At 8-Hi-496, the nearest powerline tower was quite a distance away (Tower number 146). There was no other permanent fixture to use as a datum. It was decided that the towers would be much safer for use as a permanent datum than metal rods that might be disturbed by looters or bikers, cause an accident (given the large amount of offroad biking and hiking) and which could be lost as the vegetation grew up around them. The powerline towers were the most permanent fixtures at the site. The power line corridor remained fairly clear, and the outline showed up relatively well on the aerial photographs. By using the powerline towers, it was also possible to connect the two sites, as well as map the creek more effectively. Eight semi-permanent reference points were used throughout the area at Cowhouse East Head. Twenty-one reference points were used at Cowhouse West Head, which, for the sake of

PAGE 116

103 accuracy, was mapped in two complete loops of the site by survey teams. This allowed for the accurate mapping of the site to the powerline towers, which were quite a distance away from the widely spaced excavation units. Because of the probability of error, excavation units at Cowhouse West Head were not assigned northing and easting designations. Laboratory Methodology In the lab, all artifacts were washed, and new bags labeled with the provenience. All bags were recorded and boxed separately, with each excavation unit being placed in its own box or boxes. Inventories were developed for each excavation unit, including test units, surface materials, and unprovenienced materials from excavation units (materials which were disturbed from an excavation unit that was looted). Artifact analysis was done at the Archaeology Laboratories of the University of South Florida, Tampa, where the materials are curated, along with excavation and analysis forms, field notes, photographs, maps, and other project-related information. As a good part of the material from EU 6 and the Cowhouse West Head site was bagged with the s 'urrounding clay matrix, these soils were collected in double-bagged trash bags and brought back to the lab, where they were waterscreened through 1/4 inch mesh screen. When reduced in volume, these levels were placed in a bicarbonate solution to degrade the clay matrix. Sample levels were screened through 1/8 inch mesh to determine the loss of smaller

PAGE 117

104 flakes through the 1/4 inch screen. The catalog methodology was to roughly sort all of the artifacts according to their position in a tool reduction trajectory. It was also kept in mind that there would at some point be a need to have some sort of standardization with materials from the Interstate-75 project excavations and other recent excavations in the area. Soil samples were stored in half-gallon zip-lock bags (with the exception of EU 6, which because of its larger size remained in double-layered garbage bags), each labeled with provenience. From these bags, vials of soil were removed for transport to Haines Testing Laboratory, where soil pH and phosphate could be measured. Flotation samples were then taken, approximately one gallon for each natural stratigraphic layer on the unit profiles. In some instances, further flotation samples were used. Samples were also taken for further soil analysis and for use in creating the stratigraphic zone descriptions, as well as other descriptive analyses of the soils. Lithic Analysis The debris resulting from stone tool manufacture, tool use, and tool modification is the most abundant class of archaeological remains recovered from sites in the Central Peninsular Gulf Coast archaeological region. This is due primarily to the region's abundant silicified limestone and silicified coral outcroppings. The highly acidic, eolian soils of this region contribute to this phenomenon as well,

PAGE 118

105 and as a result there is very little preservation of organic material. Debitage can also be considered the class of archaeological remains most likely to have been disposed of in the same area as it was created (de facto refuse), rather than at some other locale (primary or secondary refuse) (Schiffer 1972). While stone tools can be manufactured, used, and discarded at the same or different locations, debitage usually remains at or near the same location as it is produced. These considerations make the intensive analysis of stone tools and their manufacture waste materials of paramount concern to archaeologists working in this region. The goals of the lithic artifact analysis were three-fold: 1) to determine the origin and production stage of the stone tools manufactured, used, and discarded at the sites; 2) to provide information which would aid in identifying site use (function); and, 3) to provide information regarding the culturaljtemporal affiliation(s) of the sites. Two kinds of lithic raw material (chert) were used by the inhabitants of the sites: silicified coral and silicified limestone. Silicified coral is the replacement of the original coral aragonite skeletal material with quartz. Such replacement often preserves the fabric of the coral resulting in a distinctive "star" pattern in the stone. The fossil species most commonly silicified is Siderastraes, a common fossil found in Oligocene and Miocene

PAGE 119

106 formations in Florida and south Georgia (Upchurch et al. 1981). The distribution of this material is widespread in Florida. Artifacts made from silicified coral can not, as yet, be assigned to any of the known coral quarry areas. Upchurch et al. (1981) and Goodyear et al. (1983) state that the closest known coral quarry sources to the project area are located on the northern end of Honeymoon Island in Pinellas County and Wesley Chapel in Pasco County. Upchurch ( 1980:53-55) notes two sources of silicified coral near Cowhouse Creek along the Tampa By-Pass Canal, although there was no indication that these sources were quarried during prehistoric times. Local collectors have noted coral sources exposed in other areas, particularly an area approximately one mile south of Lake Thonotosassa. As this material is difficult to flake in its unaltered state, it is often thermally altered to enhance its flaking qualities. Silicified limestone is the replacement of limestone by quartz (Upchurch et al. 1981). This replacement usually retains both the fabric of the limestone and the fossils contained within it. Upchurch et al. (1981) have devised methods to differentiate silicified limestone materials based on their original geological contexts. They have identified 19 quarry clusters for the state, each containing any number of specific quarry locations. Each quarry cluster contains silicified limestone with similar fabric and fossils (Upchurch et al. 1981).

PAGE 120

107 The process of stone tool production has been investigated in several studies (e.g., Muto 1971; Crabtree 1972; Callahan 1979). Collins {1975) has developed a generalized model of stone tool production which outlines a series of manufacturing steps, and the end result in the archaeological record. This model includes the following steps: 1) Raw material acquisition 2) Core preparation and initial reduction 3) Primary trimming 4) Secondary trimming and shaping 5) Maintenance/modification 6) Discard Flake size distribution analysis allows for the assessment of debitage counts over a series of flake size ranges. This analysis is based on the relationship of completed tool size to debitage {flake) size. Stone tool manufacture is a subtractive process. Thus, the smaller the implement being manufactured, the smaller the flakes removed during manufacture. The debitage produced while finishing or modifying a tool should, on the average, be smaller than flakes produced during the initial shaping of the tool. This idea is central to several debitage studies such as Newcomer {1971), Ahler {1975), Raab, Cande, and Stahle (1979), and Burton {1980). While large flakes are the product of early stage reduction, small flakes are produced throughout the manufacturing process {Newcomer 1971; Patterson 1982; Patterson and Sollberger 1978). Cortex is the natural surface, or outer rind, of flint-

PAGE 121

108 like materials (Crabtree 1972:56). Fragments of cortex are associated with raw material procurement activities or the early stages of stone tool manufacture. The amount of dorsal surface cortex was measured using the primary, secondary, and non-decortication categories established by White (1963). Primary decortication flakes exhibit cortex on the entire dorsal surface of the flake. Generally, this means that these type of flakes were the first removed from a core in the lithic reduction trajectory. Secondary decortication flakes exhibit cortex on only a portion of the dorsal surface of the flake. Again, these flakes are primarily associated with early stages of tool reduction, although flakes with small amounts of cortex (i.e., those in later manufacturing stages) were still placed in this category. Non-decortication flakes were flakes with no cortex. These types of flakes are associated with late stage reduction and resharpening of tools. While primary and secondary decortication flakes are generally the products of earlier stages of manufacture, non-decortication flakes can be produced throughout the process. Shatter was distinguished from flakes in this analysis, although it must be noted that the constituents of this category may have been the result of two distinct processes. Shatter is defined as angular and irregularly shaped pieces of lithic material lacking the attributes (dorsal and ventral faces, bulb of force, striking platform) typically associated with flakes. Although shatter is generally

PAGE 122

109 considered to be associated with the earlier stages of stone tool manufacture (Binford and Quimby 1972:347; House and Wogaman 1978:59), it may also be produced by thermal alteration. Shatter was separated into those pieces with cortex and those without cortex. Thinning index values were calculated for bifacial tools by dividing the weight of the biface by its planview area (after Johnson 1981). The resulting index quantifies the degree to which the biface has been thinned and allows the placement of bifaces within a reduction trajectory (Figure 11) (e.g., Austin and Ste. Claire 1982; Estabrook and Newman 1984). Early stage reduction (ER) and biface discard (BD) indices were also calculated. These indices allow for distinguishing between assemblages representing the manufacture of bifaces from those representing the use, maintenance, and discard of bifaces (House and Ballenger 1976:96). A high value of BD would indicate a high degree of biface use, modification, and discard. A high value of ER would indicate the relative importance of tool manufacture over tool use, while a low ER calculation would suggest the opposite (House and Ballenger .1976: 96). The index of biface discard was calculated by dividing the number of bifaces, biface fragments, and points by the number of primary and secondary decortication flakes, after Daniel and Wisenbaker (1981:88). The index of early stage reduction was determined by dividing the number of shatter, primary, and secondary decortication flakes by the number of

PAGE 123

Biface Manufacturing Key Bifacial Modification is predominantly soft hammer, with the lateral edges completely worked I Yes I Lateral margins are r---No--Type 1 Blank straightened, and a regular, ----No---Type 2 collateral flaking pattern Blank has been established I Yes I Thinning index is less than 1 7 gmsjsq em, and Preform hafting modification is present I Yes I Hafted Biface FIGURE 11. Biface Manufacturing Key (after Estabrook and Newman 1984:75). 110

PAGE 124

111 non-decortication flakes. These indices should allow for comparisons between inter-site assemblages to be made, although these have been documented in few archaeological reports in this area. Edge angles have been used to infer functional aspects of tools. Experiments and archaeological investigations (Gould, Koster and Sontz 1971; Wilmsen 1970) have demonstrated the utility of edge angle analysis in understanding tool use. Wilmsen (1970:70-71) suggests that there are three general classes of edge angles: 1) acute edge angles ( < 35 degrees) are most suitable for cutting and slicing, and are primarily related to skinning, butchering, and meat preparation. Tools which fall into this range are often flakes utilized without modification of the edge of the flake. 2) medium edge angles (46 degrees to 55 degrees) are useful for a variety of tasks and are generally classified as multi-purpose tools. Inferred uses for these tools are skinning and hide scraping, sinew and plant fiber shredding, heavy cutting of wood, bone, and antler, and tool back blunting. 3) high edge angles ( > 60 degrees) are related to heavy wood or bone working activities, or heavy shredding. Edge angles were measured using a goniometer, following the methodology of Butler (1980), and were assigned to one of the three categories listed above. Several use-wear studies (e.g., Semenov 1964; Keeley 1977; Odell 1975, 1980; Hayden 1979; Ballo 1985) have demonstrated the importance of use-wear analysis in archaeological interpretation. Experimental studies (e.g.,

PAGE 125

112 Keller 1966; Tringham et al. 1974; Odell and Odell-Vereecken 1980) have been particularly revealing as to the functions of stone tools, as well as the identification of utilized tools and flakes. The purpose of the use-wear analysis was to determine whether a tool was used on hard, medium or soft materials, as well as whether it was used in a scraping, cutting, chopping, battering or boring motion. The presence of rounded edges and polish was used as an indicator of use as an abrader on soft materials. Fracture patterns are useful in determining the type of material worked and were used to back up edge angle determinations. Generally, working bone, antler, and hard woods results in characteristic step and hinge fracturing, with the sides of the scars becoming abraded or crushed with use. Small, scaliform scarring is produced on soft materials such as hides, meat or plant material. Tools utilized on medium materials would exhibit edge damage similar to working both hard and soft materials, although the damage would tend to be shallower and shorter than working hard materials, and larger than damage evidenced from working softer materials. Use-wear location is important in determining the mode of action which produced the damage. Generally, edge damage on one surface is an indication that the tool was used in scraping, while bifacial damage is an indicator of a cutting, slicing, or sawing action. Tools used in a boring action will generally exhibit rotary wear on the edge in

PAGE 126

113 contact with the worked material. Hammerstones will generally show evidence of battering on one or more surfaces. All lithic materials were microscopically examined at least three times under magnification for any evidence of use damage. Examination of suspected tools was undertaken using a 7x to 30x stereomicroscope. Generally, the use-wear determinations were conservatively made. Keeley (1974) cautions the use of edge damage as an indicator of tool function due to the resemblance of use-wear damage to edge damage caused by manufacture, handling, and other non-use activities. This was an important consideration, given the amount of edge damage which might have occurred as a result of trampling and other depositional processes, as well as from incidental edge damage occurring during the excavation and handling of such a large amount of lithic material in the laboratory. Thus, only tools which showed a significant amount of patterned edge damage were considered utilized. Notes were made concerning edge wear, and outlines of most tools were recorded with descriptions and locations of edge damage, including whether damage was evidenced on one or both surfaces, and in some instances, the types of damage. As Keeley ( 1974) suggests, the analysis of striations and polish on tools are more sensitive indicators of tool use, and were noted where applicable in this study as well.

PAGE 127

CHAPTER SEVEN RESULTS AND INTERPRETATIONS 114 There is no doubt that stone (chert and silicified coral) was an important resource for groups during the Paleo-Indian through Archaic periods. However, we cannot lose sight of the fact that lithic technology was only one aspect of the hunter-gatherer way of life. Several recent studies have demonstrated the utility of lithic analysis to the understanding of broader questions about cultural behavior (e.g., Wilmsen 1970; Odell 1980; Tringham et al. 1974; Collins 1975). Thus, although lithic materials represent only a small component of prehistoric material culture, there is much that can be gleaned from this class of data. The interpretation of site activities is based on several assumptions. Plainly stated, these assumptions are: 1) that generalizations about cultural behavior may be made from a very limited class of data; 2) that functional and typological classifications of lithic materials from archaeological sites can be made accurately, and that changes in these assemblages are the result of some impetus (i.e., environmental change) creating the need for a cultural response (that technological change reflects a functional adaptation to some degree rather than merely stylistic preference); 3) that these data can be understood despite having undergone significant change since the materials were first deposited in the

PAGE 128

archaeological record; and, 4) that the use of ethnographic knowledge about regularities (patterns) in modern hunting and gathering populations, particularly in regard to their relationship with their relative environment, is an accurate reflection of past cultural systems. 115 These assumptions must be kept in mind when viewing the inferences made, as they may skew the understanding of activities during prehistoric times if it is ultimately shown that the basis for these inferences is patently incorrect. Determination of Time The Cowhouse East Head and Cowhouse West Head sites were originally classified by B. Calvin Jones as Archaic sites with a possible Paleo-Indian component. Testing, the examination of artifacts from disturbed areas, and an intensive surface reconnaissance addressed the need for an understanding of artifact and debitage type distributions in order to determine whether both of these components were present, as well as to isolate these both vertically and horizontally, if possible. Conclusive results were not obtained prior to excavation. However, some degree of confidence was attained in the separation of components during the excavation and analysis of the recovered materials (Torp 1991). The following criteria were used in establishing the presence of two distinct components at the Cowhouse East Head site: 1) diagnostic tools/projectile points in vertical isolation;

PAGE 129

2) changes in artifact and debitage types (i.e., changes in functional classes of tools present, amount of coral tools and debitage, etc.) suggestive of a different occupation; and, 3) the isolation of discrete clusters of tools and changes in artifact densities vertically. 116 Components could not be separated vertically at the Cowhouse West Head site. Three excavation units from Cowhouse East Head were analyzed (EUs 1, 2, and 5). These were considered to be representative of the site as a whole. Diagnostic projectile points and point fragments recovered from the site all dated to the Middle Archaic period (Figure 12). None were recovered lower than level 5. In the lower component, the proximal end of a symmetrically squared-base biface was recovered (Figure 13). It is possible that this specimen is a preform fragment of a Suwannee or Suwannee-like projectile point. Johnson (1985:51) reports finding an almost identical example associated with the Paleo-Indian component of the Wetherington Ridge site. Johnson (1985:51) suggests that the Wetherington Ridge specimen was a salvaged preform fragment which was subsequently as a knife. This Cowhouse East Head fragment is manufactured from a pelodial grainstone-breccia (Bay Bottom chert), which appears to be a preferred chert type during the Paleo-Indian period for the manufacture of bifaces (Daniel et al. 1986:51; Goodyear et al. 1983). Local collectors suggest that points similar to the

PAGE 130

FIGURE 12. Sample of Middle Archaic Points from EUs 1,2 and 5 8-Hi-495. 117

PAGE 131

0 1 2 3 4 5 FIGURE 13. Possible S uwannee Preform from 8-Hi-495, EU 2 Level 12. FIGURE 14. Oblong U nifacial Scraper from 8-Hi-495, EU 2 Level 12. 118

PAGE 132

119 artifacts recovered from Wetherington Ridge and Cowhouse East Head are widespread throughout Hillsborough County contexts. Two such collectors argue that these represent a previously undefined point type, which they label "Cowhouse Slough" points (Anderson and Puckett 1989:3-4). They describe the point type as a "lanceolate blade that curves inward at the base ... [with a] convex to straight base [which is] usually not ground" (Anderson and Puckett 1989:4). Three oblong unifacial scrapers were recovered from the same level as the Suwannee-like biface which may be considered diagnostic of the Paleo-Indian period (Purdy 1981:20-21). Figure 14 shows one of these scrapers. Although this tool type represents one that continues in later time periods, Purdy (1981:20) notes that later examples of this tool type tend to be wider and smaller in size. Unfortunately, little comparative research has been aimed at discerning differences between Paleo-Indian and Archaic tool kits, and the results here must be viewed as somewhat speculative at this point. Excavation methodology was geared toward noting changes in soil texture, color, composition, and artifact densities, during fieldwork. This prevented the mixing of materials from different components if such were encountered at the sites. Artifact density dropped off significantly at various points between 80 and 95 em below surface, in which a 10-15 em interval of low debitage density was noted prior to an increase in the number of artifacts recovered.

PAGE 133

120 Although this varies between units, this was noted in each of the excavated units at Cowhouse East Head. A total of 491 pieces of coral debitage were recovered from the excavated units at Cowhouse East Head. The vast majority of this (93.28%) occurs in the upper component, while only 12 (2.44%) silicified coral flakes were recovered from the lower levels (Table 1). An almost equivalent volume of soil was excavated for each of the isolated components. Despite the fact that coral comprised less than two percent of the total debitage count (and 2.80% of the total debitage associated with the Archaic component), coral represents only 0.18 percent of the total lower level debitage. A more pronounced distinction can be seen in the general tool assemblages of EUs 1, 2, and 5. In the Archaic component, 7 of 30 (23.33% ) tools were manufactured from coral, while only 1 of 44 ( 2 27%) of the Paleo-Indian component tools is derived from this material. Of the manufacture debris (blanks and preforms) recovered from the site, 2 of 7 (28.57% ) associated with the Archaic period occupation were coral, while all 13 of the Paleo-Indian blanks and preforms were of local silicified.limestone. One blank from EU 5, Level 7 was not included due to the lack of confidence in assigning it to a particular component. Of the 10 projectile points recovered from the three complete excavation units (EUs 1, 2 and 5), six were manufactured from silicified coral, silicified limestone. while four were manufactured from All of these points were diagnostic

PAGE 134

121 TABLE 1. Raw Material Distribution, 8-Hi-495. Raw Material Distribution: Cow house East Head Component Raw Bifaces Debitage Material Count Percent Count Percent Silicified Limestone 11 28.21 15,879 63.33 ARCHAIC Silicified Coral 8 20.51 458 1. 83 Silicified Limestone 18 46.15 6,682 26.65 PALEO-INDIAN Silicified Coral 0 0.00 12 0.05 Silicified Limestone 31 79.49 24,581 98.04 SITE TOTALS Silicified Coral 8 20.51 491 1.96 TOTAL 39 100.00 25,072 100.00 Notes: 1) EUs 1, 2, and 5; 8-Hi-495. Totals include all levels, including those not assigned to a specific component. 2) Bifaces include blanks, preforms, utilized bifaces, and projectile points.

PAGE 135

122 of the Middle Archaic period, and all were within the levels associated with that occupation. These distributions seem in keeping with patterns of silicified coral use during the Paleo-Indian a n d Archaic time periods (see Daniel and Wisenbaker 1987). The distribution of bifaces (blanks, non-specialized biface forms, preforms and projectile points) by level exhibited a distinct bimodal grouping (Figure 15). When graphed by individual excavation unit, a similar pattern exists (Figure 16). Soil movement and other postdepositional processes would tend to move the heavier objects deeper within the soil. Over a long period of time these processes would tend to cluster these artifacts at some depth. It is suggested that the distribution is in fact, caused by two discrete occupations. The depth and stratigraphy of the components, and the type of materials recovered are similar to those from the Interstate-75 excavations (Randy Daniel, personal communication, 1991). Adding to the confidence of assigning a Paleo-Indian affiliation to the lower component were several factors, including an analysis of private collections from Cowhouse Creek sites (which contained several Paleo-Indian points from the Cowhouse East Head site), published assessments of Paleo-Indian occupations in the immediate area (Goodyear et al. 1983; Daniel and Wisenbaker 1987), and a recent reexamination of assemblages recovered from the Interstate-75 Project, which revealed

PAGE 136

6 5 4 3 2 1 Bitace Distribution by Level Cowhouse East Head Number 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 D Blanks Preforms Excavatlon Level -Utilized Bi1ace-:---l B Projectile 8-Hi-495; Total trom EUs 1 ,2,5 FIGURE 15. Biface Distribution by Level, 8-Hi-495. 123

PAGE 137

Biface Distribution by Excavation Unit Cowhouse East Head Number of Blfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 -EU 1 Excavation Level r::::::::::n EU 2 8-Hi-495; Blanks, Pretorms, Utilized Bitaces, Points, and Point Fragments FIGURE 16. Biface Distribution by Excavation Unit, EUs 1, 2, and 5, 8-Hi-495. 124

PAGE 138

125 that at least 8 of the 13 sites excavated contained Paleo Indian material (Daniel, Wisenbaker and Ballo 1986:54; Randy Daniel, personal communication, 1991). Determination of Function Once cultural materials were separated by cultural/temporal association, it was necessary to determine the nature of the occupations. Binford and Binford (1966) distinguish between two broad classes of data within ''technologically simple societies:" extraction and maintenance. Extraction activities are those activities associated with the procurement of subsistence items and raw materials, while maintenance activities are those related to the preparation and manipulation of the goods or materials procured during extraction activities (Binford 1983:100). Thus, the types of tools (representing different functional classes) recovered from the site would allow for an assessment of the types of activities occurring at the site. Similarly, the debitage associated with a site would allow for an understanding of what stages of the lithic reduction process were conducted at the site, and an indication of site use as well. Inferences regarding site functions allowed for a determination of site type. This includes base camps and work camps. Base camps are "selected primarily in terms of adequate life-space, protection from the elements, and central location with respect to the distribution of resources" (Binford 1983:101), and are reflective of

PAGE 139

126 maintenance tasks. Work camps, such as kill sites, quarry sites, collection stations, and lithic workshops, are generally reflective of extractive tasks. These sites were occupied for short periods of time while sub-groups carried out specific extractive tasks (Binford 1983:101). While there are many sites along Cowhouse Creek, little is known with regard to their relationship to each other. Thus, one goal of this research was to determine how these sites fit together, and to make inferences regarding their relationship with other sites along the creek. As a complete range of material throughout the lithic reduction continuum was recovered sequentially from the Cowhouse West Head and Cowhouse East Head sites, it may be confidently assumed that at a given point in time, material from Cowhouse West Head was transported from this site to Cowhouse East Head, and completed tool forms produced. Binford (1983:101) states that "the way in which these two general classes of camps [base camps and work camps) are utilized by a single society defines the settlement system." This settlement system is undoubtedly greater in extent than Cowhouse Creek, and must be approached both 'in regard to the sites' relationship to the others within the Cowhouse Creek area, as well as the broader context of seasonal movements and dispersals throughout the seasonal cycles. As Austin and Ste. Claire (1982:30) point out, it is necessary to distinguish between functionally different types of sites before it becomes possible to understand

PAGE 140

127 prehistoric settlement systems. This has proven to be a notoriously difficult problem, and although several attempts have been made to confront the lack of functional identification for lithic scatters (e.g., Hemmings 1975; Wharton 1979; Austin and Ste. Claire 1982), this remains largely unresolved. Austin and Ste. Claire (1982) suggest a shift in emphasis from the functional identification of sites as a whole, to the identification of activity sets (Austin and Ste. Claire 1982:36). Their model (Austin and Ste. Claire 1982:37-43) allows for the identification of activity sets from which site function could be inferred. The model is particularly useful in that it is geared toward sites dominated by a lithic assemblage. Recovered materials from EUs 1, 2, 5, and 7 were used in the site assessment of Cowhouse East Head. EU 7 it must be noted, was damaged by vandals prior to the excavation of the lower levels, and excavation had to be abandoned. As these represent a complete Archaic period assemblage from that excavation unit, totals for EU 7 are included in the site profile (Table 2). The levels between the two components contained few tools and are not included in the totals presented in Table 2. The excavated volume for the Archaic component is approximately 12 cubic meters ( EUs 1,2,5,7), while the Paleo-Indian materials discussed were recovered from approximately 8.2 cubic meters (EUs 1,2,5). At cowhouse West Head, artifacts from EUs 6, 8, and 12 as well as those recovered from surface contexts were used. As

PAGE 141

TABLE 2. Site Profile, 8-Hi-495. SITE PROFILE: COWHOUSE EAST HEAD (8-HI-495) CLASS General Utility Knives utilized flakes bifaces Scrapers utilized flakes modified flakes unifaces bifaces Scrapers/Knives utilized flakes modified flakes Chopping Tools/Adzes Hunting Implements Projectile Points complete bases stems tips mid-sections Fabricating or Processing Perforators/Drills Graving Tools Graver/Perforators Bur ins Abraders Chert Working Implements Cores Hammers tones Manufacture Debris Blanks (type-1) Blanks (type-2) Preforms Miscellaneous Biface Fragments TOTALS PALEO-INDIAN COUNT PERCENT 1 11 9 9 2 3 Tools 1 2 8 8 3 3 5 65 1.54 16.92 13.85 13.85 3.08 4.62 1.54 3.08 12.31 12.31 4.62 4.62 7 .69 100.03 EU 1,2,5 ARCHAIC COUNT PERCENT 2 2 30 10 6 1 4 2 2 11 2 5 1 3 3 1 1 2 2 2 4 1 97 2.06 2.06 30.93 10.31 6.19 1. 04 4.12 2.06 2.06 11.34 2.06 5.15 1.03 3.09 3.09 1.03 1. 03 2.06 2.06 2.06 4.12 1. 03 99.98 EU 1,2,5,7 Note: Some tools are included in more than one functio nal category. 128

PAGE 142

129 noted previously, materials could not be separated into discrete vertical components at Cowhouse West Head. Descriptions of other tools recovered from the site are addressed later in this chapter. Evidence from Cowhouse East Head indicates that most activity sets were present (Table 2). Although activities appear somewhat more limited during the Paleo-Indian occupation of the site, it too was considered a maintenance camp for lack of comparative data with which to judge. While activities were not as diverse as those represented at Harney Flats, the site was not as large and excavations not as intense as those performed at Harney Flats. Based on the variety of activity sets present, the spatial extent of the site, and the probable repeated occupation of the site through time, it was determined that this site represented a quarry-related base camp of the type expected in Daniel's (1985) settlement model. Tools classified as general utility tools comprised the majority of both the Paleo-Indian and Archaic assemblages. Flake tools were the most common tool type in the Cowhouse East Head assemblage. Generally, the distinction between utilized flake tools versus modified flake tools has been used to distinguish between an expedient assemblage and a curated assemblage, which may relate some information regarding length of occupation and degree of mobility. However, data regarding these categories is limited, particularly in cases where high quality lithic material is

PAGE 143

130 abundant. Within the general utility category, 11 of 22 (50. 0 % ) of the Paleo-Indian flake tool assemblage were utilized flakes, while this number increased to 36 of 48 ( 75.0%) total flake tools in the Archaic component. The numbers of unifacial and bifacial tools remained relatively high during both occupations. No projectile points were recovered from the PaleoIndian component of the site ( EUs 1, 2, and 5) In the Archaic component, 18 projectile points and point fragments were recovered ( EUs 1, 2, 5, and 7) None exhibited extensive maintenance, and evidence of use in cutting and/or scraping activities was subtle, if present. Thus, these are probably best associated with hunting activities. Three of the complete specimens had broken tips, and were not resharpened. Nine of these points were manufactured from coral, six of which were thermally altered. Five of the nine silicified limestone points were thermally altered. The number of blanks and hammerstones recovered from the Paleo-Indian component was relatively high, suggesting that activities associated with tool production was a more important site activity during that period. While fabricating and processing tools were present in the assemblages of both components, they were somewhat more important during the Archaic component. The presence of a highly polished sandstone abrader in the Archaic component also indicates the presence of hide processing. Scraping tools and blades generally had low edge angles (60.7% low,

PAGE 144

131 28.6% medium, 10.7% high) during the Archaic component, while the Paleo-Indian component was generally characterized by high edge angles (31.2% low, 25.0% medium, 43.8% high). Edge damage on the Paleo-Indian assemblage was primarily scaliform scarring, although some step and hinge fracturing was present. This indicates that these tools were probably utilized primarily in the processing of soft materials as well. As some degree of evidence on these tools suggests the working of hard materials, it seems that these were more general purpose tools than those associated with the Archaic occupation of the site. Both assemblages contained primarily moderate to heavy use-wear, although this is more of a reflection on the conservative nature of the use-wear analysis. The results of these analyses suggest that activities during the Archaic were associated with the skinning and processing of soft materials, butchering activities, and meat preparation. Both Paleo-Indian and Archaic tool groups contained relatively high percentages of tools with moderate edge angles. These tools are generally classified as multi-purpose tools, and are associated with skinning, scraping, and hide preparation as well as heavy cutting and shredding activities. Site activities at the Cowhouse West Head site (Table 3) were primarily associated with the raw material acquisition through primary trimming phases of lithic processing. Evidence of these activities was not present at Cowhouse East Head, but was present in the area between the

PAGE 145

TABLE 3. Site Profile, 8-Hi-496. SITE PROFILE: COWHOUSE WEST HEAD (8-HI-496) CLASS COUNT PERCENT General Utility Knives utilized flakes bifaces Scrapers utilized flakes modified flakes unifaces bifaces Scrapers/Knives utilized flakes modified flakes Chopping Tools/Adzes Hunting Implements Projectile Points complete bases stems tips mid-sections Fabricating or Processing Tools Perforators/Drills Graving Tools Graver/Perforators Bur ins Abraders Chert Working Implements Cores Hammers tones Manufacture Debris Blanks (type-1) Blanks (type-2) Preforms Miscellaneous Biface Fragments TOTALS 2 1 2 1 1 4 1 10 12 11 2 2 49 4.08 2.04 4.08 2.04 2.04 8.16 2.04 20.41 24.49 22.45 4.08 4.08 99.99 EU 6,8,12 Surface 132

PAGE 146

133 two sites. This area ( EU 6) was excavated as part of Cowhouse East Head (due to its location on the eastern side of the creek), and remained part of that site for reasons of site management despite the functional difference between the lowland procurement area and the habitation area on the rise (Torp 1991). Functionally, this area is similar to the Cowhouse West Head site, and is treated as part of that site for purposes of this thesis. Discussion of the Relationship Between 8-Hi-495 and 8-Hi-496 The research conducted at the Cowhouse East Head and Cowhouse West Head sites determined that tools and debitage at the sites roughly conformed to Austin and Ste. Claire's (1982) model of lithic procurement and processing. As the two sites followed that of the extractive and maintenance site dichotomy discussed previously, Austin and Ste. Claire's (1982:41-43) model for lithic procurement and processing allowed for the establishing of the relationship between the Cowhouse East Head and Cowhouse West Head sites. This model (paraphrased somewhat from Austin and Ste. Claire 1982:41-43) states that raw material acquisition would be characterized by large hammerstones, anvils, an abundance of large, primary decortication flakes, shatter material, cores, core fragments and broken chert nodules. These types of artifactual materials will be located at the source of raw material. If excavation is necessary to expose the lithic material, they suggest that tools used for such tasks may be found at the procurement area. Large scrapers,

PAGE 147

134 choppers, stone picks, and hammers could be expected for these activities. Large flakes and chunks of stone obtained from existing debris would have been utilized with a minimum of effort expended in manufacturing a functional working edge. Edge wear may be either slight or pronounced, depending on the specific activity (e.g., digging, chopping tree roots, etc.). Core preparation and reduction would be evidenced by cores, core fragments, core rejuvenation flakes, shatter material, hammerstones, anvils, and an abundance of large primary and secondary decortication flakes. Smaller flakes will occur as well, and would be characterized by a relative absence of platform and dorsal faceting, obtuse platform angle, and platform battering. Primary trimming would be evidenced by small hammerstones, aborted flakes, primary, secondary, and non-decortication flakes of all sizes. Flake characteristics would include relatively steep to moderate platform angle, minimal to moderate dorsal faceting, a relative absence of platform faceting, and high flake weight. Platform preparation will represent several techniques including battering and abrasion. Secondary trimming activities would be represented by an abundance of aborted blanks (type-2) and preforms, and secondary and non-decortication flakes which are generally medium to small in size. Although some hard hammer work may continue, it would be expected that at this stage, soft hammer use would predominate. Thus, few hammerstones would

PAGE 148

135 be expected. Flake characteristics would include relatively acute platform angle, increased dorsal and platform faceting, medium to low flake weight, and possible evidence of thermal alteration. Platform margins should exhibit evidence of abrasion. Tertiary trimming activities would be evidenced by an abundance of aborted preforms, as well as non-decortication flakes which are generally small in size. Soft hammer percussion and pressure flakingjretouch would be expected to dominate this stage of reduction. Flake characteristics would include a moderate to low platform angle, extensive dorsal and platform faceting, low flake weight, and possible evidence of thermal alteration. Platform margins should exhibit evidence of platform preparation in the form of abrasion, as well as crushing due to pressure flaking. The use and discard phases of the model would be evidenced by worn or broken tools, utilized flakes, and exhausted or re-utilized cores. Maintenance activities will be evidenced by tools displaying re-utilization. Retouch flakes should display acute platform angles with evidence of use-wear along the platform margins. Using this lithic procurement and processing segment of the model (Austin and ste. Claire 1982:41-43), the relationship of the two sites could be determined, based on the fact that lithic reduction is a continuum, and that each of the stages from procurement to tool production, reworking and discard would generally be present in all cases. The

PAGE 149

136 model could also be refined somewhat, particularly with evidence from the Cowhouse West Head site. At Cowhouse West Head, lithic reduction progressed from the acquisition through primary trimming stages. Despite the presence of blanks and some usable bifaces, little evidence of secondary trimming (soft hammer percussion) was observed at the site. Blanks from Cowhouse West Head were predominantly worked by hard hammer percussion. All exhibited bifacial modification, though few had completely worked lateral margins (Figure 17). Without extensive efforts at refitting material from the two sites, a definitive statement concerning the relationship between the sites is somewhat difficult. However, a gross lithological examination of the material from the two sites indicates that with the exception of the coral material recovered from Cowhouse East Head, the debi tage and manufacturing debris appears to be almost exclusively from Cowhouse West Head. In addition, the lack of stone sources in any of the immediate directions other than toward cowhouse West Head and the continuum of lithic reduction processes between the two sites strongly suggest that quarried material from Cowhouse West Head ultimately found its way to the Cowhouse East Head site. Of the 13 blanks recovered from the three excavation units at Cowhouse West Head, six (46.2% ) showed evidence of rejection or failure during manufacture which precluded any further use of the material. Eleven (84.6% ) were assigned

PAGE 150

--FIGURE 1 7 Blanks from the Cowhouse West Head Site. ...... w -..J

PAGE 151

138 to the type-1 blank category. Two were classified as type-2 blanks. Both of these type-2 blanks were worked by hard hammer, and resembled the type-1 blanks in manufacture, with the exception that these two specimens were manufactured from large flakes rather than cores. One of these specimens is shown in Figure 17 (at the lower right side of the photograph). As a result, the lateral margins were roughly straight and complete modification of lateral edges was made easier. Both were rejected because they could not be thinned. None of the blanks from this site were thermally altered, and none exhibited collateral flaking patterns. No preforms were recovered from these excavation units. Blanks at Cowhouse East Head consisted of ten type-1 (66.67%) and five type-2 (33.33%) blanks (Figure 18). Only one was thermally altered. Seven preforms were recovered, six of which were production failures (Figure 19). were thermally altered. None cores were also recorded, which appear to be roughedout nodules at a stage prior to bifacial reduction. This may suggest that what has been labeled by others as core fragments, are actually blanks in various stages of production and use. Many of these cores show some degree of battering which may have been incidental to manufacture. All of the Cowhouse West Head specimens contained fairly high amounts of cortex, possibly signs of rejection. One core was recovered from Cowhouse East Head, which, in contrast, exhibited almost no cortex.

PAGE 152

FIGURE 18. Blanks from the Cowhouse East Head Site. t-' w 1.0

PAGE 153

140 FIGURE 19. Preforms from the Cowhouse East Head Site.

PAGE 154

141 Austin and Ste. Claire (1982:163) note that at the Rock Hammock site, "manufacturing debris ... shows evidence of rejection and is more accurately classified as rejected blanks". Austin and Ste. Claire ( 1982:145) also suggest these may represent the production of flake tools, as the flakes recovered exhibited high striking platform angles and platform preparation characteristic of blade production. Type-1 blanks at Cowhouse West Head had a mean volume of 204.57 cubic em, while type-1 blanks at Cowhouse East Head had a mean of 229.03 cubic em (not accounting for rejection). Four of the Cowhouse West Head type-1 blanks were probably rejected. Three of these appear to have been broken, possibly during production, one due to inclusions within the stone. One other had a vein of inclusions which precluded further reduction of the blank. The two largest examples from Cowhouse East Head had some degree of battering, indicative of their use as hammerstones. One of these was silicified coral, while the other was of local silicified limestone. The elimination of obviously rejected blanks made the Cowhouse West Head sample size too low to be a significant gauge for comparison. However, it appears that with the elimination of incompletely modified type-1 blanks (those with inclusions or imperfections in the material) and manufacturing breaks (incomplete type-1 blanks), the volume of Cowhouse West Head blanks would be somewhat higher than those recovered from Cowhouse East Head.

PAGE 155

142 Type-2 blanks from Cowhouse West Head had a mean volume of 73.34 cubic ern, compared to a value of 53.50 cubic ern at Cowhouse East Head. Both of the Cowhouse West Head specimens were rejected, as noted previously, because they could not be thinned. None of the Cowhouse East Head type-2 blanks appear to have been rejected. Because there is a difference in type-1 blank size between the two sites, it is suggested that non-rejected blanks were curated for a period of time prior to manufacture into completed tool forms. This may mean that a surplus of lithic material in the form of blanks was maintained for use at the base camp or for export from the area. This remains to be quantitatively addressed, specifically with regard to factors of rejection within the blank assemblages of the two sites. It is apparent that lithic reduction ceases for the most part at the type-1 blank stage at Cowhouse West Head, and begins at this stage at Cowhouse East Head. The fact that this process is structured (evidenced by the dominance of type-1 blanks at the quarry site) also provides evidence the use of the quarry site merely for the e xpedient gathering of stone materials. Flakes suitable for utilization were probably collected during all phases of manufacture. Very few curated blades were evident at Cowhouse East Head, and it appears that utilized flakes from other modes of manufacture were their

PAGE 156

143 source rather than specialized core production for blade removal. At Wetherington Ridge a similar pattern exists. Johnson (1985:58) notes that the utilized flake assemblage consisted almost entirely of expedient flake tools with the exception of a few blade-like flakes. The importance of blanks in the process of gathering and preparing stone for use cannot be understated. Daniel and Wisenbaker (1987:171) note the flexibility of this as a tool: ... the bifacial core ... was probably the most general purpose "tool" in the assemblage in that it could fulfill a number of different purposes. When properly reduced, it would function as a "handaxe." If large enough, it would serve as a core from which flakes could be removed. Ultimately, it could even have been transformed into a projectile point. Its most important property was undoubtedly its portability. The identification of the tool kit associated with quarrying activities was another important problem addressed during the investigation of Cowhouse West Head. While it was not expected that a complete functional tool kit would be recovered or even represented at the site, several tools associated with quarrying activity were recovered. One important aspect of this tool inventory is that (with the exception of hammerstones, anvils, and manufacture debris), these tools were all recovered from surface contexts, and may represent 'site furniture.' Of these surface tools, two hammerstones were recovered. One exhibits heavy battering on the sides and end. Interestingly, this tool was thermally altered coral,

PAGE 157

144 and not the silicified limestone characteristic of the site. The second hammerstone displayed moderate battering on opposite ends of the tool. This artifact was heat altered silicified limestone, although it appears that this was a post-depositional alteration. Four chopping implements were also recovered from surface contexts, all of which were manufactured from silicified limestone. These tools were probably used for chopping roots and digging. Three were bifacially-worked, while one was classed as unifacial. The unifacial specimen was thermally altered. All exhibited extensive edge damage. Several tools were recovered from surface contexts which apparently were not directly related to quarrying activities, including an unidentified Florida Archaic Stemmed projectile point. The point is broken, and although the break is not the result of manufacture break, it is undetermined whether this is the result of use or postdepositional in nature. A thermally altered oblong scraper, and a non-altered discoidal scraper were found, both indicative of heavy use on a hard surface (probably woodworking). A biface with heavy step fracturing on one edge was also found. The specimen had a moderate edge angle and was probably suited for a variety of tasks, although it was probably utilized in heavy cutting or woodworking activities. A relatively large unifacial tool with a rounded blade appears to have been manipulated in a cutting action. The

PAGE 158

145 tool appears thermally altered. Edge angle is moderate, and edge damage is regular but fairly light on both margins of one edge. Two pieces of shatter were found which have some patterned edge damage. These may have been quarry shatter which were used expediently for scraping activities, although it is also possible that edge damage was the result of trampling or other post-depositional forces. Two preforms were recovered from surface contexts. Both are thermally altered, and neither appear to have been utilized. While some of these tools, particularly the heavy woodworking tools, may have had some importance in site activities, it is also likely that these represent forays through the site locale for purposes other than lithic raw material acquisition. recovered from closed Very few general purpose tools were excavation contexts. These were limited to a unifacial scraper from EU 6 a thermally altered coral scraper from EU 8, and three projectile points recovered from EU 9. EU 9 was located on a relatively high knoll to the south of the other excavation units. Of these three points (all of whic h are typical Archaic stemmed points), one was fashioned from coral, one from Upper Wi thlacoochee Quarry Cluster c hert (both of which also showed signs of extensive resharpening, and obvious discard), and one is apparently of local chert (Figure 20). The two points manufactured from non-local chert were recovered from level 3, while the silicified limestone point was recovered from level 4. This point appears to be a Levy

PAGE 159

i: I . I FIGURE 20. Archaic Stemmed Points Recovered from EU 9, 8-Hi-495. 146

PAGE 160

147 type point, with a broken base and tip, which may represent manufacture breaks. Several thermally altered flakes were recovered from this excavation unit. The two non-local points do not appear to have been manufactured at that locale, as no Upper Withlacoochee Quarry Cluster or coral material was noted despite a fairly intensive scan of the recovered debitage. A large sample of this unit was also screened through fine-meshed screen (1/8 in), with no retouch flakes being recovered. This excavation unit ( EU 9) was also the unit from which the vast majority of organic materials were recovered. Approximately 12.7 grams of fish vertebrae and 198.9 grams of freshwater shell were collected. It remains undetermined whether these are cultural remains or naturally deposited materials during an episode of high water. These materials extend throughout the vertical extent of the unit, and though they are not differentially associated with the clay matrix or the eroded limestone matrix of the excavation unit, may be a factor of preservation, or may be fossil remnants from the limestone formation. The latter seems unlikely, as the shells appear to represent a freshwater species common to the area. The pH levels for the excavation unit, in both matrixes, are similar to that of other excavated units. Within the three analyzed 1x2 meter excavation units at Cowhouse West Head, a total of 15,233 pieces of debitage, weighing 149,634.4 grams were recovered. As there was

PAGE 161

148 considerable variability between excavation units, each will be discussed separately. EU 6 was located on the east side of the slough, between the sites. As noted previously, the original manuscript on this project (Torp 1991) treated this unit as part of Cowhouse East Head. This facilitates the management of the two sites by using the creek as the boundary between the sites. Functionally, however, this low, wet area was characterized by quarrying activities, and for purposes of this thesis, will be considered as part of the Cowhouse West Head quarry site. other activities No evidence of tool reduction or any in the area between the main area of Cowhouse East Head and this procurement area was noted during testing and surface reconnaissance, and it appears that this area may be considered separate from that site. This area was approximately nine meters above mean sea level, and is under water during most of the year. An extremely dry field season allowed for the excavation of this unit without reaching the water table. Almost 12,000 (11,921) pieces of debitage were recovered from this 1x2 meter excavation unit, weighing a total of 129,708.7 grams. An additional 29,471 grams of non-worked material was collected from the unit. Thirteen levels were excavated to a depth of 115 em below surface. The volume excavated was approximately 3.3 cubic meters. Excavation was halted at this depth when green-blue gleyed Miocene-age clays were encountered. No culturally worked

PAGE 162

149 material was recovered from this depth. Four fragments of non-worked coral were recovered from this unit, weighing a total of 14.6 grams. Level 6 (45-55 ern below surface) contained three cores, three type-1 blanks, and four harnrnerstones. The harnrnerstones exhibited moderate to heavy damage. In level 7 (55-65 ern below surface), four cores were found. In level 8 (65-75 ern below surface) two type-1 blanks and five harnrnerstones were recovered. Three of these harnrnerstones exhibited heavy battering, while two contained moderate damage on their ends. Level 10 (75-85 ern below surface) contained two type-2 blanks and one harnrnerstone exhibiting moderate to heavy use. One core was recovered from level 11 (95-105 ern below surface). A type-1 blank was recovered from level 12 (105-115 ern below surface) which exhibited moderate battering from use as a hammer. None of the blanks or cores from this unit were thermally altered. A unifacial scraper was the only tool recovered from this unit which was not directly involved in the quarrying activities. This tool exhibits a high edge angle and a moderate amount of use. Given the depth of the artifact (level 11), this may represent a tool which was deposited in the pit during quarrying or at some point while the pit was still open. It is also possible that the artifact was manufactured at that locale. EU 8 was a 1x2 meter unit located on the west side of Cowhouse Creek, on the slope leading to the floodplain (10-

PAGE 163

150 10.5 meters above mean sea level). The total number of debitage recovered from this unit was 2,702 pieces, weighing a total of 15,647.1 grams. A total of 4,327.1 grams of nonworked lithic material was recovered. The excavated volume of this unit was approximately 2.2 cubic meters. Artifact density in this unit was approximately one-third that of EU 6. Excavation reached a maximum depth of 153 em below surface, although low artifact density allowed for the excavation unit to be down-sized to a 1x1 meter square at 70 em below surface. Only one level contained tools and manufacture debris other than debitage and raw material. Level 4 (30.5-40 em below surface) contained two type-1 blanks, and a fragment which appears to be a utilized bifacial tool (thermally altered) This level also contained a thermally altered coral scraper, with a low edge angle and moderate to heavy edge damage along one margin. Four thermally altered coral flakes were recovered from this level ( 13. 9 grams) No other coral was recovered from this unit. EU 12 was located just south of EU 8, at approximately 10 meters above mean sea level. There were 610 pieces of debi tage, weighing a total of 4, 278.6 grams in this 1x2 meter unit. Non-worked lithic material weighed a total of 1834 grams. A volume of approximately 3.6 cubic meters was excavated. Excavation reached 180 em below surface (level 16), although almost no artifacts were recovered from below level 13 ( 150 em below surface). One small ( 2. 2 grams)

PAGE 164

151 coral flake was recovered. At the bottom of level 4, a silicified limestone biface fragment was recorded at 52 em below surface. This artifact had a low edge angle and there was no evidence of utilization. In Level 6 ( 54-64 em below surface) three type-1 blanks were recovered. None of these artifacts were thermally altered. A large core (3497.8 grams) was also recovered from level 6. It is interesting to note that the weight of this core is almost that of the debitage weight for the entire unit. Efforts at refitting debitage from this unit to the core were minimal, although it was possible to articulate several of the larger fragments to this core. Despite the extreme differences in artifact density among these three excavation units, several important similarities exist within the debitage assemblages. Primary decortication flakes comprised 17.4% of the total debitage at Cowhouse West Head, which ranged from 12.42% in EU 8, 18.72 in EU 12, and 18.74 in EU 6. Secondary decortication flakes ranged from 25.61% in EU 8 to 29.81% in EU 6, and 30% in EU 12. Non-decortication flakes comprised 61.97% in EU 8, 51.45% in EU 6, and 51.28% in EU 12. overall, nondecortication flakes constituted 53.72% of the debitage at Cowhouse West Head. In contrast, the cowhouse East Head debitage was dominated (80.65% overall) by non-decortication flakes, while secondary decortication flakes ( 15.3% ) and primary decortication flakes ( 4. 05% ) comprised a far less

PAGE 165

significant portion ratios varied only 152 of the debi tage assemblage. These slightly throughout the different excavation units and between the Paleo-Indian and Archaic components of the site. A comparison of the two sites is presented in Figures 21 and 22. When compared by number and weight, there is a considerable difference between cortex and non-cortex material {Figures 23 and 24). Index of Early Stage Reduction (ER) scores were calculated using silicified limestone debitage. These scores varied between excavation units at Cowhouse West Head, with values of 0.99 (EU 8), 1.52 (EU 6), and 1.53 (EU 12). At Cowhouse East Head, the unit total scores ranged from 0.38 (EU 1), 0.39 {EU 5), and 0.42 {EU 2). Indices of Biface Discard were calculated using utilized bifaces, preforms, projectile points and fragments. Blanks were excluded from this computation due to the factor of rejection which was difficult to correlate between the two assemblages. In a temporary extractive campsite, the material used for tool production would be suspected to be confined to the local lithic material being procured. Thus, the debitage would reflect the amount of tool production. While a major portion of the debitage was in fact local material (i.e., material from Cowhouse West Head), approximately 10.9% of the tool population were manufactured from coral. Similarly, a large number of the projectile points recovered from Cowhouse East Head were manufactured from

PAGE 166

Percent of Total by Number Cowhouse East Head/Cowhouse West Head 8-HI -495 8-HI-496 Percent PRIMARY SECONDARY NON-DECORT. Flake Type SITE TOTALS ITHilliB-HI-495 -8-HI-496 FIGURE 21. Debitage Comparisons by Number, 8-Hi-495 and 8-Hi-496. 153

PAGE 167

Percent of Total by Weight Cowhouse East Head/Cowhouse West Head 8 -HI-495 8-HI-496 Percent PRIMARY SECONDARY NON-DECORT. Flake Type SITE TOTALS llifiliiliJ 8-HI-495 8-HI-496 FIGURE 22. Debitage Comparisons by Weight, 8-Hi-495 and 8-Hi-496. 154

PAGE 168

Material With Cortex Cowhouse East Head/Cowhouse West Head Eill] 8-HI-495 8-HI-496 BY NUMBER BY WEIGHT EU I EU 6 Materlal wlth Cortex EU -EU 2 8 D EU 5 EU 12 EU 1,2, 5 (8-Hi-495);EU 6 8 ,12 (8-Hi-496) FIGURE 23. Debitage Comparison of Material With Cortex, by Number and Weight. 155

PAGE 169

Material Without Cortex Cowhouse East Head/Cowhouse West Head 8-HI-495 8-HI-496 Percent BY NUMBER BY WEIGHT Material Without Cortex !illiliiliJ EU 1 R EU 6 EU 2 EU 8 EU 1,2,5 (8-Hi-495);EU 6 8 ,12 (8-Hi-496) D EU 5 EU 12 156 FIGURE 24. Debitage comparison of Material Without Cortex, by Number and Weight.

PAGE 170

157 coral. As coral sources are not located at Cowhouse West Head, this material in the form of raw material or completed tools was undoubtedly transported to the site from another locale. To test this, indices of Early stage Reduction (ER) and Biface Discard (BD) were calculated for the different raw materials (Table 4). Used in conjunction, these two indices may foster a better assessment of specific assemblages or components of assemblages, and provides collaborating or contrasting evidence for understanding extractive versus maintenance activities. This statistic has been used only infrequently in the archaeological literature, and it is unknown what constitutes adequate sample size or how these can be used as part of an assessment of specific site types. Thus, values calculated for the Cowhouse East Head assemblage should not be compared quantitatively with other sites to determine degree of 'basecamp-ness' or whether the assemblage from another site represents a base camp or extractive camp when compared to this statistic. However, when used with some restraint, these statistics may illuminate several aspects of a specific assemblage. Some degree of confidence in the use of these statistics is based on the conclusion that the Cowhouse East Head and Cowhouse West Head sites are essentially part of the same assemblage, segregated by site. An inspection of these figures for Cowhouse East Head indicates some degree of difference between the use of coral

PAGE 171

TABLE 4. Indices of Early Stage Reduction and Biface Discard, 8-Hi-495 and 8-Hi-496. Indices of Early Stage Reduction and Biface Discard Site Component Material ER BD PALEOSILICIFIED INDIAN LIMESTONE .429 .0058 SILICIFIED LIMESTONE .380 .0031 ARCHAIC 8-HI-495 SILICIFIED CORAL .272 .0897 SITE SILICIFIED TOTAL LIMESTONE .391 .0026 SITE SILICIFIED 8-HI-496 TOTAL LIMESTONE 1. 398 .0005 Shatter + Primary + Secondary decortication flakes ER = Non-decortication flakes Utilized bifaces + Projectile points BD = Primary + Secondary decortication flakes 158

PAGE 172

159 and silicified limestone. The comparable ER scores suggest a similar rate of coral tool manufacture, while the BD scores indicate that coral implements were discarded at a higher rate than their silicified limestone counterparts. The ER indices suggest for one thing, that the manufacturing stage at which lithic material entered the site was similar for both types of material. The ER index is roughly the ratio of the product of early stage reduction to that of late stage reduction, suggesting that if the technology and the desired end products of manufacture were similar, these ratios would roughly be the same for a given type of material. The difference between BD indices may reflect one of two things: either a low rate of silicified limestone tool use and resulting discard, or that the majority of implements manufactured from silicified limestone cherts were exported from the site. of coral debi tage in the Given the generally low amount site assemblage, it must be concluded that, while a certain amount of coral material was transported to the site for use in the production of tools (and that this material was treated in same manner as silicified limestone), many more of the silicified coral tools were transported to the site in the form of completed tools, used, and then discarded. During the period of seasonal occupation, silicified limestone chert was procured (although coral found its way into the production system) and the groups retooled with this material for the next part

PAGE 173

160 of the seasonal round. These indices were also calculated to portray differences between the Archaic and Paleo-Indian occupations at Cowhouse East Head. Materials recovered from the levels of low artifact density between the two isolated zones were excluded from this analysis. In each excavation unit, the ER values for the Paleo-Indian component were somewhat higher than the values for the Archaic component. This is in keeping with the notion that site activities during the Paleo-Indian occupation were somewhat more oriented toward biface production. The indices for Cowhouse West Head indicate that the site was primarily concerned with the early phases of lithic reduction. The Index of Early Stage Reduction was high for each excavation unit, and as a site statistic, indicates that activities of this nature were the major contributor of debitage at the site. Biface discard (other than rejected blanks) was low, and the assemblage approximated the hypothetical value of 0.00 (House and Ballenger 1976:96) for quarry sites. It is important to stress again, that these values are not necessarily valid in comparisons with assemblages over space, and that a BD index at or close to House and Ballenger's ( 1976:96) hypothetical quarry index does not make Cowhouse West Head more of a quarry than another comparable site. It is also important to note that this statistic does not prove this a 'perfect' quarry site, but

PAGE 174

161 rather, only indicates that the use, maintenance and discard of bifaces was not an important function of the site. As a general caution against the over-interpretation of statistical analyses, the ER indices for coral would have been relatively high and the BD index low for the site. The negligible frequency of coral debitage and bifaces precluded the calculation of indices for that material. As a stand-alone statistic, however, they may have suggested a correlation that does not exist, given other realities at the site. Mean flake weights ranged from 1.88 (EU 5), 1.90 (EU 1), and 1.96 (EU 2) at Cowhouse East Head to 5.79 (EU 8), 7.01 (EU 12), and 10.81 (EU 6) at Cowhouse West Head (Figure 25). These values were also calculated for cortex material (Figure 26) and non-cortex material (Figure 27) in the debitage assemblage. As expected, mean flake weights for cortex material were considerably higher for the quarry associated debitage, while only slightly higher for non-cortex material. Coral debi tage at Cowhouse East Head portrayed significantly lower flake weight than silicified limestone, with a mean flake weight of 0.73 grams. The mean flake scar length of blanks from the two sites also portray some differences between the sites. The mean flake scar length from Cowhouse West Head blanks was 3.98 em, compared to a value of 3.08 em in the Cowhouse East Head assemblage (for silicified limestone blanks). Silicified coral blanks had a mean flake scar length of 4.11 em,

PAGE 175

Comparison of Mean Flake Weights Cowhouse East Head/Cowhouse West Head 8-HI-495 8-HI-496 Mean Flake Weight (gms.) 1 1 10 9 8 7 6 5 4 3 2 1 0 -'----..1=== UNIT MFW (Flakes and Shatter) [!H!l] EU I B EU 6 l!il!ll E U 2 -EU 8 EU 1,2, 5 ; 8-Hi-495; EU 6 ,8,12; 8-Hi-496 D EU 5 B EU 12 FIGURE 25. Mean Flake Weight by Unit, 8-Hi-495 and 8-Hi-496. 162

PAGE 176

Comparison ot Mean Flake Weights Cowhouse East Head/Cowhouse West Head 8-HI-495 8-HI-496 Mean Flake Weight (gms. ) 18r--------------------------------------. 17 16 15 14 13 1 2 1 1 10 9 8 7 6 5 4 3 2 1 0 __.__ __ nmmw Eu 1 B EU 6 MATERIAL WITH CORTEX (Flakes and Shatter) EU 2 -EU 8 D EU 5 -EU 12 EU 1,2, 5 ; 8-Hi-495; EU 6 8 ,12; 8 -Hi-496 FIGURE 26. Mean Flake Weight: Material With Cortex, 8-Hi-495 and 8-Hi-496. 163

PAGE 177

4 3 2 1 Comparison of Mean Flake Weights Cowhouse East Head/Cowhouse West Head 8-HI-495 8 -HI-496 Mean Flake Welght (gms. ) !ITiliiJ E U 1 -EU 6 NON-CORTEX MATERIAL (Fl a kes and Shatter ) Ia EU 2 EU 8 D EU 5 EU 12 EU 1,2 5 ; 8-Hi-495; EU 6 8 1 2 ; 8-Hi-496 FIGUR E 27. Mea n Flake Weight: Material Without Cortex, 8-Hi-495 and 8-Hi-496. 164

PAGE 178

165 possibly indicating that these forms were at a similar stage of manufacture to the silicified limestone specimens from Cowhouse West Head. As noted previously, no silicified coral blanks were recovered from Cowhouse West Head. Type-1 blanks from Cowhouse West Head portrayed a mean flake scar length of 4.51 em, compared to a value of 3.16 em for type-2 blanks. At Cowhouse East Head, these values were 3.38 em for type-1 blanks, and 2.94 em for type-2 blanks. In this comparison of flake scar length by blank type, factors of rejection were not considered. What all of these data indicate is that activities which occurred at Cowhouse West Head were centered around the procurement of lithic material, and the initial shaping of this stone into rough bifacial forms which could be transported easily to the base camp. Usable flakes manufactured during this initial shaping of the stone were probably collected. While other activities undoubtedly occurred at Cowhouse West Head, the archaeological expression of these activities i s both limited and probably ephemeral to quarrying activities. At the base camp, a wider variety of tasks were performed. Although there is considerable evidence to suggest that tool production and maintenance tasks were activities whic h received considerable attention, most other tasks that would be expected by Austin and Ste. Claire (1982) are present. The fact that stone was plentiful within a short distance to the living area obscures the importance of these other tasks at

PAGE 179

166 Cowhouse East Head to a certain degree. The assertion that the bulk of the stone recovered at Cowhouse East Head originated from Cowhouse West Head is enhanced by the gross lithological examination of the material, and by the lack of such an accessible source of stone in directions other than to the west of Cowhouse East Head. Blanks recovered from the two sites show significant differences between the type-1 blanks from Cowhouse West Head and the type-2 blanks from the Cowhouse East Head assemblage. The type-1 blanks recovered from both sites show some degree of overlap in dimensions and overall size. Partially, this is the result of the inability to adequately account for factors of rejection, and the result of a relatively small sample size in cases where rejection could be accounted for. The mean flake scar lengths, however, were greater in the Cowhouse West Head sample, indicating that those blanks which were transported to Cowhouse East Head had been modified to a further extent than those left at the site of procurement. With an increased data base resulting from other investigations within the area, it may be more feasible to better define these areas which now overlap.

PAGE 180

CHAPTER EIGHT PREHISTORIC QUARRYING BEHAVIOR 167 The term "quarrying" is used to denote the location of the long-term and/or intensive activities of lithic extraction and the initial treatment of this material. Butler and May (1984:185) note that actual excavation to locate and remove the raw material from its parent geological matrix is a relatively rare phenomenon at sites in the continental United States, and thus, the argument of Chance (1981, 1982; Chance and Misner 1984) favoring the use of the term "procurement si te11 to replace "quarry si te11 on the grounds that the term is technically inaccurate, is moot. Quarry sites must be considered as an integral part of a settlement/subsistence system that includes lithic workshops, hunting and butchering sites, short-term campsites, base camps, and other types of sites. Because lithic material is a fixed resource, the initial behavioral pattern to be observed is that of utilization versus non-utilization of available material. Some areas of extensive outcroppings are apparently ignored in favor of other areas. The most obvious discriminating factor is that of the local environment (i.e., was the chert under water, or in an otherwise inaccessible or unfavorable area). A second

PAGE 181

168 factor is the quality of the raw material itself, including decisions to exploit chert resources opportunistically during the acquirement of other critical resources, or the situating of camps in close proximity to chert resources. Due to the limited nature of investigations at quarry locales and the deficient nature of the Florida Master Site File information regarding the various types of "lithic scatters" of this part of Hillsborough County, there are very few chronological data concerning the use of quarry sites. This problem is compounded by the lack of diagnostic artifacts contained in quarry site assemblages. The wide temporal range of sites in the immediate area suggests that these sites were utilized during all of the defined cultural periods in the Central Peninsular Gulf Coast region, although the most intensive use of the Cowhouse Creek quarry sites probably occurred during the Paleo-Indian through Archaic periods. Cowhouse Creek is located within an area that is dense with lithic scatters, including several sites that have been labeled 'quarry' sites. Relatively little, however, is known about cultural behavior at these types of sites. One reason for this is that relatively little research has been aimed directly at understanding lithic procurement. As Chance (1981:109) notes: ... the necessary excavation and evaluation of site 8-Hi-473 [Wetherington Island] presented a multiplicity of research and logistical of taxing complexity. The character of the l1th1c assemblage was found to be at variance with that

PAGE 182

of lithics usually collected in Florida and the overwhelming number of potential artifacts was organizationally problematic. 169 Exceptions are Purdy's investigations at the Container Corporation of America site (Purdy 1981), the York site (Purdy 1977), and the Senator Edwards site (Purdy 1975), and Chance's (1981, 1982; Chance and Misner 1984) research at Wetherington Island, which has the added importance in its proximity to the Cowhouse East Head and Cowhouse West Head project area. Quarry studies were largely initiated with the work of W. H. Holmes (1894; 1919), although Gerard Fowke had conducted several systematic studies of quarry sites while working for the Bureau of Ethnology ( Fowke 1884; 1902). Holmes noted that many of the quarry pits he studied were surrounded by workshop areas where trimming of the raw material took place. As quarrying activities progressed, these individual workshop areas were obliterated by the enlarging mass of debris (Holmes 1894:12). Holmes states through several of his writings, that the almost exclusive product of these sites (in addition to hammerstones) was "some form of blade or disk--a blank--intended, in most cases, no doubt, to be subsequently elaborated into an implement of more highly specialized form" (Holmes 1894:14). He further noted: If long, thin blades were desired, they could be made; if broad, strong, discoidal forms were needed, they could be shaped with equal But no matter what the final forms 1n v1ew were--and it is evident that such forms were

PAGE 183

generally in view--the quarry work covered only the incipient stages of shaping, i.e., the roughing out. 170 Bryan (1950) disputed two major points made by Holmes based on his observations of quarry assemblages. In this work Bryan states that: ... many of the so-called "blanks" and "rejects" are usable tools, mainly axes, and that they were actually used ... [and] that many flint quarries were not only sources of flint for export, but also industrial sites or factories to which materials such as wood and bone were brought to be worked in the presence of abundant tools. These main theses are supported and inspired by the very greatly expanded knowledge of typology developed in the past thirty years; by the recognition in the debris of quarries of numerous utilized flakes, and of shaped and utilized pieces of irregular shape; and by our present saner attitude toward the antiquity of man in America (Bryan 1950:3). These disputes over the end-product of the quarry industry and whether these sites tended to focus primarily on the initial stages of stone procurement or whether the complete range of tool production activities occurred dominates the early literature on quarry sites. What is significant about this long-standing debate is that much of the recent literature remains preoccupied with these same problems. The procurement of lithic raw material may range from an occasional, opportunistic activity to that of a highly concentrated activity. Generally two factors operate in this patterning: local geological conditions and cultural patterns (Butler and May 1984:185). It is apparent that these types of sites must be evaluated individually, as quarry sites are characterized by variability. This

PAGE 184

171 variability is also an important intra-site factor, as is demonstrated by the extreme range of debitage densities in the excavation units evaluated in the previous chapter. The procurement of lithic material is based, at least in part, on the availability of lithic raw material. Evidence of different behavior might be observed at sites where lithic raw material was easily procured (or more difficult to procure), or where quarrying was not the primary activity. In instances where other resources could be exploited within the site bounds, these types of activities undoubtedly occurred as well. This makes both the classification of quarry sites as well as interpretation more difficult, particularly at the individual site level. The earliest investigation of quarries in Florida was probably carried out by Simpson during his tenure with the Florida Geological Survey. As noted in a previous chapter of this thesis, Simpson visited several quarry sites in the area, noting the extensive nature of these sites, the generally high quality of the chert, and the probable importance of the materials to other areas with less abundant lithic resources (Simpson 1941). Simpson also notes the Muscogean meaning of Thonotosassa as "Flint Place." Simpson states: Evidence remains that fires were built to help break the stones into rough shapes after which they were further worked by hammering into blanks that were carried away to be finished. There is no evidence that any implements were finished here but the extent of operations is shown by the abundance of broken spalls and rejects found a

PAGE 185

half mile in all directions. This flint material has been so plentiful that individuals have found it profitable to recover it for use as concrete aggregate (Simpson 1941:31-32). 172 In an earlier work, Simpson provides a more specific discussion of a quarry site assemblage: In the preliminary research at the more important of these quarries, only one finished blade was found. Quantities of broken and complete blanks were found at random everywhere. The method employed in quarrying was shallow trenching and pitting. The large boulders were broken and fractured on the spot by the use of fire, and further reduced by striking with any sui table stone at hand. None of the smaller hammerstones for sui table finishing work were found at the quarry site. The material seems to have been roughed out into suitable flakes and blanks, and finished elsewhere (Anonymous 1939:63-64). The Two Egg Quarry in Jackson County (Sharon and Watson 1971) was the first modern investigation of a lithic procurement site in Florida. Six points and one point fragment were recovered, and the authors note that there was no indication that the site functioned as a living area, that raw material was mined and roughed-out, or that tools were fashioned directly at the site (Sharon and Watson 1971:80). The authors do not discuss the time periods during which the site was used, although the projectile points recovered from the site suggest an Archaic period occupation. The Senator Edwards Chipped stone Workshop site, located in Marion County was investigated by Purdy (1975). Purdy differentiates between core/blade and blank/projectile point production (Purdy 1975:180-181), both of which were

PAGE 186

apparently performed. period component. 173 The site contained only an Archaic The Container Corporation of America site in Marion County (Purdy 1981), was investigated for the purpose of determining whether it represented "a universe of chert procurement and utilization practices" including the methods used to quarry chert nodules, how the nodules were broken into workable fragments, and where these pieces were taken to be fashioned into tool forms (Purdy 1981:91). Unfortunately, many of these objectives were not investigated fully (Purdy 1981:108), and little data concerning lithic procurement was published. Initially, Purdy assumed that debitage at the site would indicate that nodules were broken into manageablesized pieces, and then removed to be converted into tool forms elsewhere {Purdy 1981:91). Purdy {1981:93) notes that the material recovered was fairly uniform throughout the deposit, and that approximately 14 percent of the assemblage was utilized. Only one biface was recovered from Area 1 (a thermally altered Pinellas point). It was not indicated if this point was manufactured at the site. Area 3 produced five bifaces, the base of a Paleo-Indian point, a thermally altered Citrus point, a Kirk Serrated point, and two unidentified bifaces. Area 4 produced two Levy points, nine Pinellas points, four Jackson or Duval points, and two unfinished points. Again, it was not discussed whether these could have been produced at that locale.

PAGE 187

Purdy notes that: ... it appears that after the stone was exposed, the block-on-block technique or the firesetting method was used to break up the huge nodules in some levels ... Most formal tools were made elsewhere at the site and these areas where final reduction occurred are evidenced by the recovery of i_ncreased numbers of points, fine chipping debr1s that is not found where procurement took place, and thermally-altered flakes (Purdy 1981:103-104). 174 Chance and Misner (1984:52) note that investigations of quarry sites tend to focus largely on activities at adjacent workshops rather than directly at the source of procurement. They cite several questions posed by Runnels (1983) as the basis for excavation methodology. These questions included determining if refinement occurs at the source, whether production stages are spatially segregated, and if finishing elsewhere is a uni versa! practice. While several works discuss lithic procurement, Wetherington Island is apparently the only published data from a quarry site in Florida aimed at understanding the initial procurement and treatment of lithic raw material, as opposed to the emphasis on secondary trimming and tool production states of the quarried material. Thus, this site represents the first published data concerning behavior at quarries in Florida (Chance and Misner 1984:3). The Wetherington Island site is characterized by spatially segregated areas of lithic raw material procurement. Chance ( 19 81, 19 8 2 ) and Chance and Misner (1984), note that the majority of primary reduction

PAGE 188

175 activities took place near the creek, where quantities of chert were available, while secondary activities took place further away from the creek: ... assemblages collected progressively farther away from the resource location could be placed progressively along the reduction continuum. Therefore, materials recovered adjacent to the creek were associated with procurement, initial testing and trimming, while those from the more distant subareas reflected secondary trimming and shaping (Chance and Misner 1984:4). This observed progression of reduction activities was not evidenced at Cowhouse West Head. Partially, this may be due to the limited nature of testing in directions other than the area between the two sites. In this area, the creek may have masked areas where other stages of lithic reduction were conducted. To the north, at the crest of the ridge (EU 11), excavated materials are similar to that of the areas along the ridge, and do not appear to represent a succession of reduction activities. A detailed analysis of debitage from this unit, however, has not been conducted. No tools were recovered from this unit. To the north of this area, houses and other disturbed areas made testing impossible. To the west, thick ( > 1 meter) deposits of wind-blown sands prevented adequate testing. In the low area between EU 6 and the ridge on which Cowhouse East Head was located, no indications of the intermediate stages of reduction were noted. Thus, while there was spatial segregation of procurement and later-stage activities, it does not appear

PAGE 189

176 to follow a progressive pattern of reduction similar to that of Wetherington Island. Only primary and secondary reduction activities occurred at Cowhouse West Head, while activities associated with the completion of tool forms were carried on at an off-site locale. As discussed previously, one of these locales was the Cowhouse East Head site. While quarrying activities through time may have been more or less evenly dispersed throughout the Cowhouse West Head site based on the availability of lithic materials, it is also possible to utilize existing data to present a preliminary model of raw material selection based on water levels and stream flow over time. Thus, changing environmental conditions may have played an important role in determining the availability of chert and ease of procurement. It is hypothesized that procurement patterns were more or less limited to areas where chert was exposed by stream flow. With a lower water table and a lower stream flow (if the river was flowing at all) during the PaleoIndian period, the soil overburden covering the chert would have been less thick in the lower area closest to the Cowhouse Creek channel. As the stream flow increased by the Middle Archaic period, these areas would have been underwater andjor covered with a greater amount of sediment. By the Middle Archaic period, these lower-lying areas would have been somewhat less accessible. During this time, as stream action and periodic episodes of higher water eroded the overburden from areas along the ridge, quarrying

PAGE 190

177 activities would have been concentrated to a greater extent on these higher areas. This does not mean that the materials from EU 6 are necessarily Paleo-Indian in origin. If Daniel (1985:266) is correct that these sites were occupied predominantly during the fall and winter, a period of the year during which precipitation is low (Wright 1973), these floodplain areas might have been accessible to Archaic period groups. The model does suggest, however, that the areas moving progressively up the ridge would tend to be Archaic in origin, as high water levels during the spring and summer months would have scoured the ridge making chert more accessible. Clearly, more local environmental data are needed for this area, as is survey or excavation results from quarry sites and areas along the Cowhouse Creek floodplain. Without any method of dating sites devoid of diagnostic tools, however, it seems unlikely that this model can be adequately tested. Chert was more easily procured from the lower areas along the slough, where the sand overburden was not as thick as the higher elevations of the ridge surrounding the creek. Certainly by the Middle Archaic, water resources were plentiful, and Cowhouse Creek probably had a significant discharge. An increasing discharge (resulting from environmental changes discussed previously), even if only seasonally, would have scoured the sand overburden from the floodplain and the areas at the base of the ridge, making chert nodules more accessible. This can be seen in modern

PAGE 191

178 conditions as well, with the lower areas frequently flooding, despite the construction of the Tampa By-Pass Canal, and several other construction projects that have lowered the water tables and decreased the flow of water through the Cowhouse Creek channel. Surficial water and stream flow, which was probably lower by a significant amount prior to the more modern conditions occurring by the Middle Archaic period (Watts 1980) may have allowed (or even necessitated, if the overburden along the ridge was too thick) earlier populations to procure lithic material from areas that are now under waters of the creek or from a wider variety of sources. The latter condition may help explain the somewhat less extensive nature of activities occurring at Cowhouse East Head during the Paleo-Indian occupation, as the availability of water (in addition to the availability of exposed chert) would have been a greater factor in site selection during the Paleo-Indian period. These lower and wetter areas where chert was easily extracted may have been unsuitable areas for carrying out tool production activities. The observed pattern at both Wetherington Island and Cowhouse West Head of later stage reduction having taken place further from the creek or at off-site locations may have been done as a result of relative ease of procurement. This pattern may also have been the result of behavior in which individuals dispersed throughout these lower areas to dig for and gather exposed

PAGE 192

179 chert nodules which were roughed-out into usable forms, and then converted into ( type-2) blanks in a more communal setting. This may be one reason for the inability of Chance and Misner (1984) and Torp (1991) to determine the precise ending stage of reduction occurring at quarry sites. While blanks appear to be the dominant early stage form represented at the associated base camps, it was difficult to determine to what extent, if any, these were shaped at the base camps as opposed to the work having been done prior to export from the quarry sites. For this reason, the distinction between type-1 and type-2 blanks was used. As discussed previously, it appears that type-1 blanks were fashioned at quarry sites and transported off-site for further modification. Chance (1981, 1982) and Chance and Misner (1984) state that while most of the material recovered from the Wetherington Island site consisted of early stage reduction debi tage, a significant number of utilized flakes were recorded. Chance observes that these utilized flakes generally were characterized by very minimal usage. Wear patterns associated with utilized flakes could not be replicated in an experimental setting (Chance 1982), and may have been the result of trampling and post-depositional factors to a greater extent than actual use-wear on the flakes. Many flakes recovered from Cowhouse West Head had some degree of edge damage, although again, most of this is probably post-depositional in nature. In addition, Chance

PAGE 193

180 notes the association of hammerstones to areas containing the greatest amount of raw material, indicating their use and importance in quarrying activities. This pattern was also noted at the Cowhouse West Head site. Austin and Ste. Claire's (1982) investigations at Rock Hammock focused on Hernando point production, and set the stage for research aimed at the quarry side of the lithic reduction continuum. At Rock Hammock, one part of the site was dominated by blank production. Despite the fact that the actual procurement locale was not located during the investigation, the authors indicate that it was within close proximity, as large amounts of cortex fragments and decortication flakes were recovered. Austin and Ste. Claire note that "Cores are absent from the site ... and core fragments are scarce suggesting that initial core reduction and the production of blanks occurred at the procurement area with the blanks transported to Rock Hammock for subsequent reduction" (Austin and Ste.Claire 1982:163). In their discussion of Rock Hammock, Austin and Ste. Claire (1982:163) note that: Possible evidence of blank production occurs in the eastern portion of the major activity area in the form of what has heretofore been labeled by others as core fragments ... The manufacturing debris recovered from this area shows evidence of rejection and is more accurately classified as rejected blanks ... The blade-like flakes recovered from this concentration probably demonstrate numerous unsuccessful attempts at the production of blanks which could be reduced into ... desired form (Austin and Ste. Claire 1982:163). Austin and ste. Claire (1982:145) suggest that these may

PAGE 194

181 represent the production of flake tools, as the flakes recovered exhibited high striking platform angles, and platform preparation characteristic of blade production. If true, the selection of flakes for utilization must have been highly selective and dependent on the task at hand, with little concern over waste of usable flakes. Very few curated blades were evident at Cowhouse East Head, and it appears that utilized flakes from other modes of manufacture were their source rather than specialized core production for blade removal. At Wetherington Ridge, a similar pattern exists, and Johnson (1985:58) notes that the utilized flake assemblage consisted almost entirely of expedient flake tools with the exception of a few blade-like flakes. Austin and Ste. Claire (1982:133) also note the inability to locate the source of the chert which composed this site assemblage, despite an intensive survey by several researchers, including Sam B. Upchurch, the geologist coordinating most of the chert studies at the nearby Interstate-75 corridor sites. If the production of blanks was in fact taking place at the site, them it might be assumed that an area nearby would contain a quarry assemblage. Austin and Ste. Claire (1982:133) observe that several eroded and exposed limestone surfaces nearby may have been the source of exploitation, although there was no evidence that the material was utilized and in one case proved to be insufficiently silicified for use.

PAGE 195

182 This may be an indication of a different manner of procurement behavior, which in contrast to the Wetherington Island and Cowhouse West Head sites, resulted in complete tool production episodes. In this case, large amounts of early stage reduction material were recovered (39% decortication material; and, 43.4% medium and 35.2% large flakes on a small, medium, large, extra large scale). This site appears similar to the sites investigated by Purdy (1975, 1977, 1981) in that the tool production areas appear to have taken place directly at the source of procurement, although Purdy never stresses whether the quarrying activities were spatially segregated from the areas of tool production, or even if the recovered tools were produced at the sites. The Austin and Ste. Claire report, however, addressed this problem and were unable to locate any area of concentrated quarrying activity. While the site contained a single Newnan point of the Middle Archaic period, the major use of Rock Hammock apparently took place during a postArchaic (Hernando) time frame. This may indicate that there are different types of quarry sites, or at least two discernible activity sets: the quarry site (with an assemblage of early stage reduction materials) and the workshop, at which raw material was procured, and the processing of these materials was continued through the production of completed tool forms. With an increased emphasis on the detection of activities at

PAGE 196

183 these types of sites it may be possible (with an increased database), to determine whether this represents opportunistic procurement of lithic resources, or whether this is somehow a function of time, where more sedentary (post-Archaic?) groups (which would not necessarily be located near a chert resource area) procured workable chert nodules while conducting other subsistence activities.

PAGE 197

184 CHAPTER NINE SETTLEMENT SYSTEM IMPLICATIONS OF QUARRY SITES There is a tendency for archaeological research to concentrate on the interpretation and discussion of habitation sites and settlements, with the result that functionally different types of sites are ignored. The Interstate-75 excavations sought to remedy that situation, at least in part, by the excavation of different types of sites. Only a limited amount of effort was made in the interpretation of these sites within the settlement system. Ironically, what the Interstate-75 project sought to do was somewhat lessened by the fact that many of these sites are best understood as functions of a settlement system and can best add information about the other types of sites when viewed as part of a system, and not as an individual site. One of the notions of site catchment analysis (c.f. Vi ta-Finzi and Higgs 1970) is that in areas where the geographical distribution of essential resources and the technology used to exploit these resources are similar, the patterns of human responses expectingly persist through time. The problem lies in what constitutes critical resources for site location. Site catchment studies tend to assume that human groups took advantage of all resources available, despite the fact that modern conditions, not past

PAGE 198

185 conditions, are generally used in these types of analyses, and that ethnographic evidence suggests that subsistence efforts were highly selective. This assumption also defies the evidence of ethnographic studies which indicate that subsistence activities are generally not of prime importance in many contexts. As Jochim (1981:153) stresses: Fixed resources are more important to settlement decisions than are mobile ones. Fixed resources are predictable in space and thus more reliable to procure ... as a result, hunter-gatherers demonstrate a hierarchical nesting of their activities in space around their camps, with procurement of the most reliable and fixed resources nearby and exploitation of more mobile and unreliable resources at greater distances. Soil type and drainage characteristics have been used as indicators of favorable habitation site location, and proximity to water would be a significant factor. Lithic resources would also be a concern, and in fact, as the primary class of data which archaeologists in this region must address, is probably the best activity to utilize in the development of settlement models. While trade and redistribution networks would most likely develop in areas where lithic material was scarce, or the material was of poor quality, these can be traced by quarry cluster analysis. Two important distinctions must be made in reference to between-site association patterns--that it should be generally true that spatially close sites would have similar assemblages, and that quarries should have distinct assemblages compared to sites where lithic procurement was

PAGE 199

186 not undertaken. That is, that quarry sites, by definition, are sites at which a particular resource was acquired, and, because lithic tool production is a subtractive process, quarries would "feed" this resource to the other sites within the settlement system. Generally, lithic material removed from quarries is done so in an archaeologically understandable (i.e., patterned) manner, the end-product being manufactured for export to other sites within the system. While the end-product may be cores or blanks, these will tend to leave the quarry in a predetermined form (as part of an industry), no matter where the material was intended to be delivered. Despite the wide range of potential uses for lithic material (and its ultimate expression within a particular functional class of site), the initial stages of lithic preparation will not be duplicated at any other site. Again, this stems from the fact that as a subtractive process, initial work was done prior to the removal of that material from the quarry. As lithic material, particularly good quality material, was valuable in terms of availability and energy requirements for transformation of the raw material to a usable form, it should be expected that some sort of behavioral standardization for the quarry industry was practiced (though it may have changed over time in reflection of the development of trade networks, changing settlement or subsistence strategies, and so on). stone, as a relatively heavy material, and a relatively

PAGE 200

187 bulky commodity in unmodified form, would probably have been thinned at the source in relation to the distance it had to be moved, and the amount of material that had to be transported. It may be assumed that during Archaic times, with larger and somewhat less geographically mobile populations some sort of lithic redistribution networks undoubtedly developed. In this scenario, it is also possible that aggregate groups of bands periodically congregated at or near the sources of lithic materials to replenish supplies which for the remainder of the seasonal cycles may have been less available. It is apparent that quarrying and primary reduction activities were occurring at the quarries, while secondary and tool production and modification stages were occurring at adjacent or possibly more distant sites. Due to the inherent bulk and weight of the unmodified material and the need to modify it to some type of end-product, utilization of lithic resources would have been of a direct access or supply zone nature (Ives 1984). Quarry sites cannot be characterized in isolation, and must necessarily be viewed within the framework of a settlement system. The only activity which can be surely documented at quarry sites is quarrying. As is the case with subsistence and other prehistoric activities, quarrying may have been practiced alone, or in combination with other tasks. This was first noted by Holmes (1919), and reiterated by Chance and Misner (1984).

PAGE 201

188 At Cowhouse East Head, a block of silicified coral was recovered (EU 1, Level 3; Figure 28). This artifact was classified as a type-1 blank, but clearly exhibited less reduction than most of the other blanks recovered from either of the two investigated sites. As there were no identified local coral outcrops, this probably represents an example of material brought back to the camp in less modified form because it was not part of the structured quarrying activity as found at Cowhouse West Head. Certainly, this may represent an anomaly, where the material was procured in an opportunistic manner rather than as part of the "industry" associated with quarry sites, and a larger, less-modified form was carried back to the camp. For one thing, this may suggest that coral was procured in a different manner than silicified limestone, at least in areas where this raw material is not abundant in quarry sites. It may also be an indication that despite the availability of good quality local stone, coral was specifically sought out for certain purposes. If it is true that chunks of this type material were carried back to the habitation site rather than completed into recognizable blanks prior to transport to base camps, it should indicate that less stone was removed from the quarry site at a given visit (less could be carried back), or that more people were involved in the removal and initial reduction of the stone. This differential treatment of

PAGE 202

I FIGURE 28. Silicified Coral Blank from 8-Hi-495, EU 1, Level 3. 189

PAGE 203

material may indicate something about the sizes groups, the permanence of occupations, 190 of these and the specialization of the persons producing stone tools. Research designs specifically aimed at generating testable implications for these aspects of behavior need to be generated, however, before any information is produced regarding these facets of procurement and tool manufacturing behavior. The real importance of Austin and Ste. Claire's (1982) model is that this manner of site functional analysis may represent the only viable method of classification of the sites in this area when the analysis moves from the site specific to the settlement system level. Fortunately, every recorded site within the general vicinity does not need to be carefully investigated to allow for some understandings about the relationship of sites to develop. As these time periods constitute a long range of time depth, the problem of contemporaneity provides an obstacle, as the landscape is almost completely covered by "preceramic lithic scatters." The method of understanding site relationships by site catchment analysis or by comparing types of associated debitage and tools becomes a difficult problem logistically, as archaeological sites in this area are more appropriately defined by modern landforms rather than discrete occupation areas. Because the Paleo-Indian and Archaic periods are temporally long and there are no ways to obtain dates for

PAGE 204

191 these sites, determining which sites are contemporaneous is difficult. Thus, when viewing sites over a geographic landscape, the number of specifically typed sites may be exaggerated in relation to lithic procurement locations, which as a fixed resource will remain a constant through time. Other special-use sites will be obliterated in the archaeological record, as larger base camps mask the expression of these smaller sites. This problem is compounded by the generally low archaeological visibility of Paleo-Indian sites, the bias toward recording lithic scatters as Archaic, and the difficulties in assessing site functions. There are several factors that should be noted concerning the apparent delinquency of spatial analysis studies in this region. First, there is the problem of site survival, due both to the great and rapid nature of urban development within the area, and to the effects of looting throughout the area. Although looting may very well allow for the quicker identification of site locations in unsurveyed areas, important relevant data is lost, and the site somehow becomes less important to intensive survey efforts in proportion to the amount of looting damage. The second major problem lies in the survey methodology utilized within this region, and in fact, most of Florida. There is a bias toward larger sites. Sites with a greater density of lithic artifacts consume more time and effort, and are somehow believed to be more important or at least

PAGE 205

192 have the potential to be more important than smaller sites. Smaller sites are archaeologically less visible, and it is more difficult to assess them given the low density of artifacts. Smaller sites are often located in unsurveyable areas or in low probability areas and are not tested with the same vigor as larger ones. There is also a problem of inadequate recording, particularly in terms of activity sets present (site function), and temporal associations (cultural classification). Previously, though still apparent today, is the prevailing notion that the typical "lithic scatter" in eastern Hillsborough County is associated with the "undetermined Archaic" period, despite the fact that one of the largest Paleo-Indian sites in Florida is located within a few miles. There is also a considerable body of evidence remaining largely untapped in the form of artifact collections which contain a considerable amount of PaleoIndian material from this region. As noted in a previous section, Chance dates the Wetherington Island site to the Archaic, citing the preponderance of sites in the local area which date to this period. The recovery of Archaic period tools may be somewhat more of an indication of the degree of site use, and it is certainly not surprising that Paleo-Indian tools are not present at a quarry site, as Paleo-Indian components are often missed at sites which should have some indication of diagnostics. Interestingly, one area of the Wetherington Island site (Area D) produced several Paleo-Indian

PAGE 206

193 artifacts. This area was located furthest from the chert source. Addressing the relationships of contemporaneous sites is difficult, particularly given the lack of a method of obtaining dates for local sites and tool types in general. In earlier periods especially, phases based on typological development are comparatively long and it is difficult, if not impossible, to determine how frequently settlements moved within them. Thus, the utilization of lithic material sources as the base of study, particularly as this is the class of data most often available, is the most effective method of determining behavioral trends in settlement during a given cultural period, and as such, should be afforded increased research efforts. The available literature indicates the existence of two distinct types of sites that exhibit some degree of lithic procurement--the quarry site and the lithic workshop. While both of these types of sites would contain a higher degree of initial reduction activities than occurs at other sites, these differ in the end-product. The perceived dichotomy may be partially due to the biases of reporting and the lack of standardization of terminology. The few sites discussed in this study seem to indicate that this pattern does exist, and should become more pronounced as more sites of this type are investigated. At quarry sites, stone was fashioned for export with few, if any, complete episodes of tool production. At

PAGE 207

194 lithic workshops, where available outcrops of stone were utilized, tool production was carried out to the completion of finished tools. This may be a result of several factors, such as the abundance of locally available stone, the distance of chert resources to the base camp (or the area where the tools produced would be used), or other things. It may also be a function of time, in which the lack of necessary resources compounded with increased sedentism, required a change in procurement behavior. Each of these site manifestations, despite the type of activity being similar (the modification of stone into usable forms) is the result of different systems, or at least, very different aspects of the same system.

PAGE 208

CHAPTER TEN SUMMARY AND CONCLUSIONS 195 Prehistoric quarries, while a relatively common site type designation, have been the focus of few archaeological investigations. Despite the fact that several large sites of this type exist within the Hillsborough River Basin, only the Wetherington Island (Chance 1981, 1982, Chance and Misner 1984), Rock Hammock (Austin and ste. Claire 1982) and Cowhouse West Head (Torp 1991) sites have addressed lithic procurement and the initial stages of lithic reduction in any detail. These investigations offer a significant body of data for understanding the often ignored initial stages of lithic technology and provide a vehicle for studying settlement patterns over a fairly wide geographical region. Lithic studies have tended to remain largely atheoretical, particularly it seems, in Florida. The major emphasis of lithic analysis has tended to focus on tools individually and as assemblages as designators of tool or site function. Little is mentioned concerning the behavioral aspect of lithic use, or in other words, the social dimensions of this technological system. "Unfortunately the degree of refinement in observation and measurement exceeds our present abilities to make sense of the observed variation or to deal effectively with the

PAGE 209

196 social context of tool production, use, and discard" (Cross 1983:88). It is suggested that an increased emphasis on quarry studies offers both an understanding of the initial stages of tool production (including behavioral aspects of this process), and a manner of viewing settlement behavior, based on the fairly easy ability to locate quarry sites on the regional landscape due to their differing archaeological expression. While tools can be traced to a quarry cluster, giving some indication of seasonal movements or exchange between groups, materials associated with tool production (blanks, cores, etc.) may also have some degree of mobility throughout the settlement system and may also be observed in sites some distance from their source. aspect of the quarry industry as completed tools throughout the system. This is an important is the tracking of It remains unknown how patterning in the quarry facet of the lithic industry ties into the workshop aspect of tool production. It is probable that two separate factors have operated in the formation of these two site types, one related to the cultural groups themselves and the other to the spatial and environmental setting. The patterned/specialized quarrying of raw materials as part of a collecting strategy versus a more opportunistic procurement of raw material and toolmaking in a foraging strategy would be evident based on factors of group size and mobility, degree of craft specialization, and other agents,

PAGE 210

197 while spatially, the proximity of base camps to lithic raw material sources and other critical resources would determine the way a site is exhibited in the archaeological record. These may not be mutually exclusive forces. For example, while procurement strategies evidenced at the Cowhouse East Head and Cowhouse West Head (and at Wetherington Ridge and Wetherington Island) sites exhibit a pattern of collecting, this may merely be the result of a settlement system that situated base camps at the juncture of the Polk Uplands, Zephyrhills Gap, and Gulf Coastal Lowland physiographic zones, in an area that happened to be lithic resource rich. It becomes apparent that both PaleoIndian and Archaic cultural periods tended to utilize a collecting strategy to a greater extent than a foraging one, at least in this specific setting. The differences between Paleo-Indian and Archaic groups in this respect may be more of a function of overall population size than a significant shift in resource procurement or settlement strategies. populations (and perhaps, smaller With smaller overall group size) the archaeological representation of 1 i thic extraction sites from the Paleo-Indian period is lower than the Archaic. Climatic change may have also been a significant factor, and probably created a certain necessity of greater mobility during the Paleo-Indian period as water (and faunal and floral) resources would presumably have been somewhat more dispersed.

PAGE 211

198 Certainly, one factor contributing to this problem is the bias against the recognition of Paleo-Indian components or sites, at which artifact density is low or there are no diagnostic tools. The resulting lack of data has probably exaggerated the differences between the Paleo-Indian and Archaic periods. Without an increased emphasis on regional research problems, survey methodology, and improved theoretical frameworks, few advancements will be made in the near future. The recovered material from Cowhouse East Head and Cowhouse West Head allow for the tracking of a lithic assemblage from procurement through use and discard phases of this technological system. These patterns seem to hold true concerning the Wetherington Island and Wetherington Ridge sites. Since the Paleo-Indian and Archaic periods constitute a long range of time depth, the problem of contemporaneity becomes an obstacle, as the prehistoric landscape is almost completely covered by what has been termed in the Florida Master Site File inventory as "preceramic lithic scatters." The methods utilized for understanding site relationships by site catchment analysis or by comparing types of associated debitage and tools become a difficult problem logistically, as archaeological sites in this area are more appropriately defined by modern landforms (for management purposes) rather than discrete occupation areas. Although there has been an

PAGE 212

increased emphasis on resolving this function problem (see Austin and ste. 199 site type and site Claire 1982:34-43; Deming 1980:31-32), little has changed in survey methodology to address this increased need for adequate data. In addition, there is s till the problem of addressing intersite relationships, which is made more difficult by the especially long time period during which these sites were formed, and the lack of chronological determinants. The need to differentiate between base camps that were associated with particular special-use sites, however, may not prove to be a barrier to the study of settlement patterning through this area. Given the almost exclusive nature of these sites as lithic scatters and their high spatial density throughout the general vicinity, it may be more appropriate to view patterning among these sites in terms of a quarry and base camp dichotomy. The analysis of the relationships between extractive (quarry) sites and base camps may be viewed as an initial step in determining the association between maintenance and extractive camps. As lithic debitage is de facto refuse, it should be possible to graph and compare the composition of debitage between sites to give an indication of whether or not at a particular period of time an association existed between the extraction (quarry) and maintenance (habitation) sites of an area. Through an analysis of the distribution of completed tool forms throughout the area, it may be possible to track the course of mobile groups.

PAGE 213

200 Quarry sites, being dissimilar to other sites within a settlement system, would have a unique archaeological expression and could be separated, particularly at the survey level. The ability to identify the quarry cluster origin of a given assemblage allows for the analysis of sites situated between distant from sources quarry clusters, and of lithic material. sites more Utilizing comparative data concerning the utilization of chert from Paleo-Indian and Archaic sites would allow for an assessment of changes in settlement systems through time, although with limitations on accurate dating methods for lithic sites, intra-period changes remain a difficult problem. Over the wider geographic range of Cowhouse Creek, relatively few of the recorded sites contain ceramic material, and it may be somewhat conclusively stated that the dominant occupations along the creek occur earlier than the Late Archaic. The fact that quarry sites of this size are present in direct association with earlier period sites indicates that Paleo-Indian and Archaic groups made significant use of these quarries, despite the fact that few diagnostic tools are represented in the quarry assemblages and there is little chance of separating components vertically. Later sites tend to be situated along the modern course of the Hillsborough River, and while these later groups may have accessed the Cowhouse Creek quarry sites, it is interesting to note that the debitage

PAGE 214

201 assemblage from a workshop such as Rock Hammock, despite being of a later time period, contains a significant amount of decortication material with no indication of adjacent quarry areas (Austin and Ste. Claire 1982:133). In this case, some degree of effort was expended toward the procurement of large blocks which were modified at a workshop or base camp, or that lithic resources were procured as a secondary result of food and other resource procurement activities. Clearly some cultural changes occurred with increased sedentism and may have created new factors of site selection, decreasing the importance of maintaining camps adjacent to quarry sites. Changes in the course of the Hillsborough River may also have made the Cowhouse Creek area less attractive for long-term settlement. Unfortunately, due to the fact that chert procurement studies and lithic material tracing in this area are a relatively new arena of study, more questions than answers are produced. Significant research questions requiring more data include: 1) When was chert quarried--continually, seasonally, or during periods when there was a low supply? 2) What effect, if any, did the availability of chert supplies in this area have on the cultural development and settlement of the area? How does the behavior encountered at sites where lithic resources were more limited compare to this type of settlement system? 3) Do lithic procurement strategies change through time?

PAGE 215

4) Is the use of coral (and thus the increased use of thermal alteration in cultural behavior) the result of reduced settlement/foraging areas? Is coral (which does not occur in large outcrops as the silicified limestone cherts) exploited expediently, or was it specifically sought out despite the availability of silicified limestone? 5) Is there any evidence to suggest control of the resource by particular groups, or was there equal access to quarry resources? 202 The study of these questions requires several distinct data sets: dates from both quarry sites and non-quarry sites, site functional data, and more developed understandings concerning the distribution of this type of chert (both silicified coral and silicified limestone) in the form of debitage and tools within and outside of the Hillsborough River quarry cluster. Preliminary interpretations for the cultural factors operating within the settlement system can and should be made. Ideally, the accurate recording of the Cowhouse Creek sites should be tested by a systematic archaeological survey of the creek given the limited nature of the recorded data from sites associated with that drainage system.

PAGE 216

203 LITERATURE CITED Allen, J.H. 1846 Some Facts Florida. 2(1):38-42. Respecting the Geology of Tampa Bay, American Journal of Science, Series Almy, Marion M. 1978 The Archaeological Potential of Soil Survey Reports. Florida Anthropologist 31:75-91. Altschuler, Z.S., J.B. Cathacart, and E.J. Young 1964 Geology and Geochemistry of the Bone Valley Formation and its Phosphate Deposits, West Central Florida. Guidebook, Field Trip No. 6, Geological Society of America Annual Meeting. Miami Geological Society, Miami. Anderson, 1989 Anderson, 1989 David G. The Paleo-Indian Colonization of Eastern North America: A View from the Southeastern United States. Paper presented at the 51st Annual Meeting of the Southeastern Archaeological Conference, Tampa. Son and Doug Puckett Field Guide to Point Types of the Florida. Ms. on file, University Florida, Tampa. State of of South Austin, 1982 Robert and Dana Ste. Claire The Deltona Project: Prehistoric Technology in the Hillsborough River Basin. University of South Florida. Department of Anthropology, Archaeological Report No. 12. Tampa. Ballo, George R. 1985 Experiments in Use-Wear Formation on Stone Tools Made from Florida Cherts: A Study Supporting a Microwear Analysis of Paleo-Indian Lithic Artifacts from the Harney Flats Site, Tampa, Florida. M.A. Thesis, Department of Anthropology, University of South Florida. Barnett, B.S. 1972 The Freshwater Fishes of the Hillsborough River

PAGE 217

Binford, 1980 1983 Binford, 1966 Binford, 1972 204 Drainage, Florida. M.A. Thesis, Department of Biology, University of South Florida. Lewis R. Willow Smoke and Dog's Tails: Hunter-gatherer Settlement Systems and Archaeological Site Formation. American Antiquity 45(1):4-20. Working at Archaeology. Academic Press: New York. Lewis R. and Sally R. Binford A Preliminary Analysis of Functional Variability in the Mousterian of Levallois Faces. American Anthropologist 68(2), pp. 238-295. Lewis R. and George I. Quimby Indian Sites and Chipped Stone Material in the Northern Lake Michigan Area. In An Anthropological Perspective, Lewis R. Binford, Ed. pp 346-372. Braun, E. Lucy 1950 Deciduous Forests of Eastern North America. The Blackiston Company, Philadelphia. Brinton, 1859 Daniel Notes on the Florida History, Indian Tribes Philadelphia. Peninsula: Its and Antiquities. Literary J. Sabin, 1867 Artificial Shell Deposits of the United States. Smithsonian Institutional Annual Report for 1866, pp. 356-358. Washington, D.C. Brown, Janice G. 1981 Palynologic and Petrographic Analyses at Bayhead Hammock and Marsh Peats at Little Salt Springs Archaeological Site ( 8S0188), Florida. Unpublished M.A. Thesis, Department of Geology, University of South Carolina, Columbia. Brown, Larry N. 1974 Mammalian Faunal Assessment. In Biological Assessment of the Lower Hillsborough Flood Detention Area for Recommendation of Ecologically Based Land Management, Bruce C. Cowell, Ed. Report to the Southwest Florida Water Management District. Department of Biology, University of South Florida. Bryan, Kirk 1950 Flint Quarries-The Source of Tools, and At the same Time, the Factories of the American Indian.

PAGE 218

205 Papers of the Peabody Museum of American Archaeology and Ethnology, vol. 17(3). Bryson, R.A. and W.M. Wendland 1967 Tentative Climatic Patterns for Some Late Glacial and Post-Glacial Episodes in Central North America. In Life, Land and Water, w. Mayer-Oaks, Ed., pp. 271-298. University of Manitoba Press, Winnipeg. Bullen, Ripley P. 1950 Perico Island: 1950. Florida Anthropologist 3:40-44. 1951 1959 1968 1972 1975 1976 1978 The Terra Ceia Site, Manatee County, Florida. Florida Anthropological Society Publications No. 3. The Transitional Period of Florida. Southeastern Archaeological Conference Newsletter 6:43-53. A Guide to the Projectile Points. Gainesville, Florida. Identification of Florida State Florida Museum, The Orange Period of Peninsular Florida. In Fiber-Tempered Pottery in Southeastern United States and Northern Columbia: Its Origins, Context, and Significance, R.P. Bullen and J.B. Stoltman, Eds., pp. 9-33. Florida Anthropological Society Publications 6. A Guide to the Projectile Points, Books, Gainesville. Identification of revised edition. Florida Kendall Some Thoughts on Florida Projectile Points. Florida Anthropologist 29(1):33-38. Tocobaga Indians and the Safety Harbor Culture. In Tacachale: Essays on the Indians of Florida and southeastern Georgia During the Historic Period. J. T. Milanich and S. Proctor, Eds., pp. 50-58. University Press of Florida: Gainesville. Bullen, Ripley P. and Lawrence Beilman 1973 The Nalcrest Site, Lake Weohyakapa, Florida. Florida Anthropologist 26:1-22. Bullen, Adelaide K. and Ripley P. Bullen 1950 The Johns Island Site, Hernando County, Florida. American Antiquity 16:23-45.

PAGE 219

1953 1954 206 The Point Site, Bayport, Hernando County, Flor1da. Florida Anthropologist 6:85-92. Further Notes on the Battery Point Site, Bayport, Hernando County, Florida. Florida Anthropologist 7:103-108. Bullen, R.P., W. Askew, L.M. Feder, and R.L. McDonnell 1978 The Canton Street Site, St. Petersburg, Florida. Florida Anthropologist 12:77-94. Bullen, Ripley P., Davids. Webb, and B.I. Waller 1970 A Worked Mammoth Bone From Florida. American Antiquity 35:203-205. Burch, D.G. 1974 Terrestrial Plant Associations. In Biological Assessments of the Lower Hillsborough Flood Detention Area for Recommendation of Ecologically Based Land Management, Bruce c. Cowell, Ed. Report to the Southwest Florida Water Management District. Department of Biology, University of South Florida. Burton, John 1980 Making Sense Out of Waste Flakes: Investigating the Technology and Chipped Stone Assemblages. Archaeological Science 7:131-148. Bushnell, Frank New Methods for Economics Behind Journal of 1962 The Maximo Point Site. 15:89-101. Florida Anthropologist Butler, Brian, and Ernest May (Eds.) 1984 Prehistoric Chert Exploitation: Studies from the Midcontinent. Center for Archaeological Investigations, Southern Illinois University at Carbondale, Occasional Paper No. 2. Butler, William B. 1980 An Inexpensive and Accurate Goniometer. Lithic Callahan, 1979 Technology 9(3):65. Errett The Basics of Biface Knapping in the Eastern Fluted Point Tradition: A Manual for Flintknappers and Lithic Analysts. Archaeology of Eastern North America 7(1):1-180. Carbone, Victor A. 1983 Late Quaternary Environments in Florida and the Southeast. Florida Anthropologist 36:3-17.

PAGE 220

207 Chance, Marsha 1980 A Preliminary Report on Phase II Investigations at Wetherington Island: A Lithic Procurement Site in Hillsborough County. Bureau of Historic Sites and Properties, Division of Archives, History and Records Management. Miscellaneous Report Series: Tallahassee. 1981 1982 Wetherington Island: An Archaic Procurement Site in Hillsborough County. Anthropologist 34:109-119. Lithic Florida Phase II Investigations at Wetherington Island: A Lithic Procurement Site in Hillsborough County, Florida. Interstate 75 Highway Phase II Archaeological Reports Number 3: Tallahassee. Chance, Marsha and Elizabeth Misner 1984 Archaic Lithic Procurement Behavior at Wetherington Island, Hillsborough county, Florida. Florida Division of Archives, History, and Records Management, Tallahassee. Carl J. Clausen, 1964 The A-356 Unpublished Anthropology, Site and the Florida Archaic. M.A Thesis, Department of University of Florida. Clausen, Carl J., H.K. Brooks and A.B. Wesolowsky 1975 Florida Spring Confirmed as 10,000-Year-Old Early Man Site. Florida Anthropologist 28:1-38. Clausen, Carl J. and Cesare Emiliani 1979 Little Salt Spring. Sea Frontiers 25(5):258-265. Clausen, Carl J A.D. Cohen, c. Emiliani, J.A. Holman, and J.J. Stipp 1979 Little Salt Spring, Florida: A Unique Underwater Site. Science 203:609-614. Cockrell, W.A. and Larry Murphy 1978 Pleistocene Man in Florida. Archaeology of Eastern North America 6:1-13. Coleman, J. M. 1982 Recent Seasonal Rainfall and Relationships in Peninsular Florida. Research 18:144-151. Collins, Michael B. 1975 Lithic Technology as Inference. In Lithic a Means of Technology: Temperature Quaternary Processual Making and

PAGE 221

208 Using Stone Tools. Mouton Publishers: Earl Swanson, Ed., pp. 15-34. The Hague. Cooke, C.W. 1945 Geology of Florida. Bulletin No. 29. Florida Geological Survey Conrad, T.A. 1846 Observations on the Geology of a Part of Eastern Florida with a Catalogue of Recent Shells of the Coast. American Journal of Science Series 2, Volume 2, pp. 36-48. Cowell, Bruce C. 1974 Aquatic Communities. In Biological Assessments of the Lower Hillsborough Flood Detention Area for Recommendation of Ecologically Based Land Management, Bruce c. Cowell, Ed. Report to the Southwest Florida Water Management District. Department of Biology, University of South Florida. Crabtree, Don E. 1972 An Introduction to Flintworking. Occasional Papers of the Idaho State Museum, Number 28. Cross, John R. 1983 Twigs, Branches, Trees, and Scale in Lithic Analysis. Hammers and Theories, James Keene, Eds., pp. 87-106. York. F.H. Forests: Problems of In Archaeological A. Moore and Arthur Academic Press: New Cushing, 1897 Exploration of Ancient Key-Dweller Remains on the Gulf Coast of Florida. Proceedings of the American Philosophical Society 35:329-448. Daniel, Randy 1982 A Phase III Proposal for Salvage Investigations of Site 8-Hi-507. Manuscript on 'file, Bureau of Historic Sites and Properties, Division of Archives, History and Records Management, Tallahassee. 1985 1991 A Preliminary Model of Hunter-Gatherer Settlement in Central Florida. Florida Anthropologist 38:261-273. Personal Communication. Daniel Randolph and Michael Wisenbaker An Inland Archaic 1981 Test Excavations at 8H1450D:

PAGE 222

1983 1987 209 Occupation in Hillsborough county, Florida. Interstate Highway Phase II Archaeological Reports . Bureau of Historic Sites and Properties, D1v1s1on of Archives History and Records Management, Tallahassee. A Preliminary Report on the Excavations at Harney Flats, Hillsborough County, Florida. Florida Anthropologist 36:67-80. Harney Flats: A Florida Paleo-Indian Site. Baywood Publishing Company: Farmingdale, New York. Daniel, Randolph, Michael Wisenbaker, and George Ballo 1986 The Organization of a Suwannee Technology: The View from Harney Flats. Florida Anthropologist 39:24-56. Daniel, Randy, Michael Wisenbaker and Mildred Fryman 1979 An Archaeological and Historical Survey of the Seven Proposed Recreation Resource sites in the Lower Hillsborough Flood Detention Area, Hillsborough County, Florida. Bureau of Historic Sites and Properties, Division of Archives, History and Records Management. Miscellaneous Project Report Series. Tallahassee. Delcourt, 1979 1983 1985 H.R. and P.A. Delcourt Late Pleistocene and Holocene Distributional History of the Deciduous Forest in the Southeastern United States. Veroffentlichungen des Geobotanischen Institutes der ETH, Stiftung Rubel (Zurich) 68:79-107. Late Quaternary Vegetational Dynamics and Community Stability Reconsidered. Quaternary Research 19:265-271. Quaternary Palynology and Vegetational History of the Southeastern United States. In Pollen Records of Late-Quaternary North American Sediments, V.M. Bryant and R.G. Holloway, Eds., pp 1-37. American Association of Stratigraphic Palynologists Foundation, Dallas. Deming, Joan 1980 The Cultural Resources of Hillsborough County: An Assessment of Prehistoric Resources. Manuscript on File, Historic Tampa/Hillsborough County Preservation Board, Tampa. Deming, J., M.M. Almy, R.W. Estabrook, and E.A. Horvath 1984 The Hidden River ( DRI) Archaeological Project:

PAGE 223

210 Cultural Resources Assessment survey and Evaluative Site Testing. Manuscript on File, Archaeological Consultants, Inc., Sarasota, Florida. Dillon, William P. and Robert N. Odale 1978 Late Quaternary Sea Level Curve: Reinterpretation Based on Glaciotectonic Influence. Geology 6(1):56-60. Doran, Glen H., and David N. Dickel 1988 Multidisciplinary Site. In Wet Site Ed., pp. 263-289. Jersey. Investigations at the Windover Archaeology. Barbara A. Purdy, Telford Press, Caldwell, New Dunbar, James 1982 The Effect of Geohydrology and Natural Resource Availability on Site Utilization at the Fowler Bridge Mastodon Site (8Hi393cj uw) in Hillsborough County, Florida. In Report on Phase II Underwater Archaeological Testing at the Fowler Bridge Mastodon Site (8Hi393cjuw), Hillsborough County, Florida Interstate 75 Highway Phase II Archaeological Report Number 5, Tallahassee. 1988 Archaeological Sites in the Drowned Tertiary Karst Region of the Eastern Gulf of Mexico. Florida Anthropologist 41:177-181. Dunbar, J.S., S.D. Webb, M. Faught, R.J. Anuskiewicz, and M.J. Stright 1989 Archaeological Sites in the Drowned Tertiary Karst Region of the Eastern Gulf of Mexico. In Underwater Archaeology Proceedings From the Society for Historical Archaeology Conference, J. Barto Arnold III, Ed. pp 25-31. Society for Historical Archaeology: Baltimore, Maryland. Dunbar, James, and Ben I. Waller 1983 A Distribution Analysis of the Clovis/Suwannee Paleo-Indian sites of Florida--A Geographic Approach. Florida Anthropologist 36:18-30. Estabrook, Richard W. and Christine Newman 1984 Archaeological Investigations at the Marita and Ranch House Sites, Hillsborough County, Florida. University of South Florida, Department of Anthropology, Archaeological Report No. 15. Fairbridge, Rhodes W. 1960 The Changing Level of the Sea. Scientific American 202(5):70-79.

PAGE 224

1961 1974 211 Eustatic Changes in Sea Level. In Physics and Chemistry of the Earth. L. H. Ahrens et al., Eds. Volume 4, pp. 99-185. The Holocene Sea Level Record of South Florida. In Environments of South Florida: Present and Past. P.J. Gleason, Ed. Miami Geological Society Memoir 2:223-229. Fewkes, J.W. 1924 Preliminary Archaeological Investigation at Weeden Island, Florida. Smithsonian Miscellaneous Collections 76:1-26. Fisk, H.N. 1956 Soil Mechanics and Foundations. 8th Texas Conference on Engineering. Gagel, Katherine 1981 Mitigative Excavations at the Fowler Avenue Site (8-Hi-393), Hillsborough County, Florida. Bureau of Historic Sites and Properties, Division of Archives, History and Records Management. Tallahassee. Gardner, 1977 William M. Flint Run Complex and Its Implications for Eastern North American Prehistory. In Amerinds and Their Paleoenvironments in Northeastern North America, WalterS. Newman and Bert Salwen, Eds., pp. 257-263. Annals of the New York Academy of Sciences, Vol. 288, New York. Gates, w.c. 1976 Modeling the Ice Age Climate. Science 191:1138-1144. Goggin, John M. 1947 A Preliminary Definition of Archaeological Areas and Periods in Florida. American Antiquity 13:114-127. Goodyear, Albert c. 1972 Political and Religious Change in the Tampa Bay Timucua: An Ethnohistorical Reconstruction. Manuscript on File, Department of Anthropology, University of South Florida, Tampa. 197 9 A Hypothesis for the Use of Cryptocrystalline Raw Materials Among Paleoindian Groups of North America. Research Manuscript Series 156. South carolina Institute of Archaeology and

PAGE 225

Goodyear, 1979 Goodyear, 1980 Anthropology, Columbia. University of 212 South Carolina, A.c., J.H. House, and N.W Ackerly Laurens-Anderson: An Archaeological Study of the Inter-Riverine Piedmont. Anthropological Studies 4 South Carolina Institute of Archaeology and Anthropology, University of South Carolina. Albert c., Sam B. Upchurch, and Mark J Brooks Turtlecrawl Point: An Inundated Early Holocene Archaeological Site on the West Coast of Florida. In Holocene Geology and Man in Pinellas and Hillsborough Counties, Florida. Compiled by Sam B Upchurch. Southeastern Geological Society. Goodyear, A.C., S B. Upchurch, M.J. Brooks, and N N. Goodyear 1983 Paleo-Indian Manifestations in the Tampa Bay Region, Florida. Florida Anthropologist 3 6 ( 1 -2):40-66. Gould, Richard A., Dorothy A Koster and Ann H .L. Sontz 1971 The Lithic Assemblage of the Western Desert Aborigines of Australia. American Antiquity 36:149-169. Grange, Roger T Jr. 1978 An Evaluation of Archaeological Resources in Hillsborough County, Florida. Manuscript on file, Historic Tampa/Hillsborough County Preservation Board, Tampa. Griffin, 1964 Griffin, 1950 James B. The Northeastern Woodland Area. In Prehistoric Man in the New World, Jesse D. Jennings and Edward Norbeck, Eds. The University of Chicago Press, Chicago, pp. 223-258. John B. and Ripley P. Bullen The Safety Harbor Site, Pinellas County, Florida. Florida Anthropological Society Publications No. 2 Hardin, Kenneth 1982 Mitigative Excavations at the Landfill Site (8Hi99), Hillsborough County, Florida. Manuscript on File, Bureau of Historic Sites and Properties, Division of Archives, History and Records Management. Tallahassee. Haviser, Jay B. 1983 Mitigative Excavations at the Wetherington Ridge

PAGE 226

213 Site (8-Hi-472), Hillsborough County, Florida. Bureau of Historic Sites and Properties. Division of Archives, History and Records Management, Tallahassee. Hayden, Brian (Ed.) 1979 Lithic Use-wear Analysis. Academic Press: New Hemmings, 1972 1975 Hemmings, 1974 York. E. Thomas Early Man in the South Atlantic States. Eastern States Archaeological Federation Bulletin 31:1011. The Silver Springs Site, Prehistory in the Silver Springs Valley, Florida. Florida Anthropologist 28:141-158. E. Thomas and Timothy A. Kohler The Lake Kanapaha Site in North Central Florida. Bureau of Historic Sites and Properties. Division of Archives, History and Records Management, Bulletin No. 4, Tallahassee. Holmes, William H. 1894 Earthenware of Florida: Collections of Clarence the Academy of Natural 1903 1919 B. Moore. Journal of Sciences 10:105-128. Aboriginal Pottery of the Eastern United States. 20th Annual Report of the Bureau of American Ethnology. pp. 1-237. Washington, D.C. Handbook of Aboriginal American Antiquities. Part I Introductory: The Lithic Industries. Bureau of American Ethnology Bulletin 60. House, John M. and David Ballenger 1976 An Archaeological Survey of the Interstate 77 Route in the South Carolina Piedmont. Institute of Archaeology and Anthropology Research Manuscript Series 104. University of South Carolina, Columbia. House, John M. and Ronald W. Wogaman 1978 Windy Ridge: A Prehistoric Site in the InterRiverine Piedmont in South Carolina. Anthropological Studies 3. Occasional Papers of the Institute of Archaeology and Anthropology, University of South Carolina, Columbia. Ives, David J. 1984 The crescent Hills Prehistoric Quarrying Area:

PAGE 227

214 More than Just Rocks. In Prehistoric Chert Exploitation: Studies from the Midcontinent. Brian Butler and Ernest May, Eds., pp.187-195. Center for Archaeological Investigations, Southern Illinois University at Carbondale, Occasional Paper No. 2. Jahn, Otto L. and Ripley P. Bullen 1978 The Tick Island Site, St. Johns River, Florida. Florida Anthropological Society Publications 10. Jochim, Michael 1976 Hunter-gatherer Subsistence and Settlement: A Predictive Model. Academic Press: New York. 1981 Johnson, 1979 Johnson, 1985 Jones, B. n.d. Strategies for Survival: Cultural Behavior in an Ecological Context. Academic Press: New York. Jay K. Archaic Biface Manufacture: Production Failures, A Chronicle of the Misbegotten. Lithic Technology 8:25-35. Robert E. Archaeological Excavations at the Wetherington Ridge Site (8Hi472B) in Hillsborough County, Florida. Manuscript on file, Florida Bureau of Archaeological Research, Tallahassee. Calvin Phase III Proposal for Salvage Investigations of Site 8Hi507 (Project No. 10075-3425 and 3416). Manuscript on file at the Division of Archives, History and Records Management, Tallahassee. Jones, B. Calvin, and Louis Tesar 1982 An Update on the Highway Salvage Program in Florida. Florida Anthropologist 35:59-62. Keeley, Lawrence H. 1974 Technique and Methodology in Microwear Studies: A Critical Review. World Archaeology 5:323-336. 1977 The Function of Paleolithic Flint Scientific American 237(5):108-128. Tools. Keller, c.M. 1966 The Development of Edge Damage Patterns on Stone Tools. Man 1:501-11. Lazarus, William C. 1965 Effects of Land Subsidence and Sea Level Changes on Elevation of Archaeological Sites on the

PAGE 228

215 Florida Coast. Florida Anthropologist 18:49-57. Ralph G., et. al. Leighty, 1958 Soil Survey of Hillsborough County, Florida. u.s. Depa7tment of Agriculture, Soil Conservation Serv1ce. Washington D.C. Limbrey, Susan 1975 Soil Science and Archaeology. Academic Press: New York. Long, Robert W. 1974 Origin of the Vascular Flora of Southern Florida. In Environments of South Florida: Present and Past. P.J. Gleason, Ed. Miami Geological Society Memoir Number 2 Luer, George and Marion Alroy 1979 Three Aboriginal Shell Middens on Longboat Key, Florida. Manasota Period Sites of Barrier Island Exploitation. Florida Anthropologist 32:34-45. 1981 Temple Mounds of the Tampa Bay Area. Florida Anthropologist 34:127-155. 1982 A Definition of the Manasota Culture. Florida Anthropologist 35(1):34-58. Martin, Paul S. 1973 The Discovery of America. Science 179:969-973. Martin, R.A., and S.D. Webb 1974 Late Pleistocene Mammals from Devil's Den Fauna, Levy County. In Pleistocene Mammals of Florida, S.D. Webb, Ed. University Presses of Florida: Gainesville. Mason, Ronald J. 1962 The Paleo-Indian Tradition in Eastern North America. Current Anthropology 3 :227-246. McCullough, David L. 1979 Cultural Resource Management at the Fletcher Avenue Park, Hillsborough County, Florida. university of south Florida, Department of Anthropology Archaeological Report Number 10. McCullough, David L., and Elizabeth Fisher 1978 An Archaeological and Historical Survey of the Proposed Fletcher Avenue Park Site, Hillsborough county Florida. University of South Florida, of Anthropology Archaeological Report Number 7

PAGE 229

216 McDiarmid, Roy W. and J. Steve Godley 1974 Amphibians and Reptiles. In Biological of the Lower Hillsborough Flood Detent1on Area for Recommendation of Ecologically Based Land Management, Bruce c Cowell, Ed. to the Southwest Florida water Management D1str1ct. Department of Biology, University of South Florida. McFarlan, E., Jr. 1961 Geological Society of America Bulletin 72:129. Milanich, Jerald T., and Charles H. Fairbanks 1980 Florida Archaeology. Academic Press, New York. Milliman, John D. and K.O. Emery 1968 Sea Levels During the Past 35,000 Years. Science 162:1121-1123. Mitchem, 1989 Jeffrey Redefining Safety Prehistoric/Protohistoric Peninsular Florida. University of Florida. Harbor: Late Archaeology in West Ph.D. Dissertation, Moore, C B 1900 Certain Anti qui ties of the Florida West Coast. 1903 Journal of the Academy of Natural Sciences. Volume 11. Certain Aboriginal Mounds of the Central Florida West Coast. Journal of the Academy of Natural Sciences. Volume 12, part 3. Morner, N.A. 1969 The Late Quaternary History of Kattegat Sea and Swedish West Coast: Deglaciation, Shoreline Displacement Chronology, Isostacy, and Eustacy. sveriges Geologiska Undersoknin 640. Muto, Guy 1971 A Technological Analysis of the Early Stages in the Manufacture of Lithic Artifacts. M.A. Thesis, Department of Anthropology, Idaho State University. Neill, Wilfred T. 1958 A Stratified Early Site at Silver Springs, Florida. Florida Anthropologist 11:33-48. 1964 Trilisa Florida. Pond, an Early Site in Marion County, Florida Anthropologist 17:187-200.

PAGE 230

217 1968 An and Spanish Site on Tampa Bay, Florida. Flor1da Anthropologist 21:106-116. Newcomer, Mark H. 1971 Some Quantitative Experiments in Manufacture. World Archaeology 3:85-93. Odell, George H. Handaxe 1975 Micro-wear in Perspective: A Sympathetic Response to Lawrence H. Keeley. World Archaeology 7:226-240. 1980 Toward a More Behavioral Approach to Archaeological Lithic Concentrations. American Antiquity 45: 405-431. Odell, George H. and Frieda Odell-Vereecken 1980 Verifying the Reliability of Lithic Use-Wear Assessments by "Blind Tests": The Low Power Approach. Journal of Field Archaeology 7:87-120. Palmer, Jill, Jim Dunbar, and Danny Clayton 1981 Phase II Underwater Archaeological Testing at the Fowler Bridge Mastodon Site (8Hi393juw). Interstate 75 Highway Phase II Archaeological Report, Number 5, Bureau of Historic Sites and Properties, Florida Division of Archives, History and Records Management, Tallahassee. Patterson, L W. 1982 The Importance of Cultural Resource Cultural Resource 21(3):10-12. Distribution. Flake Size Management Management Abstracts and and Archaeology Patterson, L.W. and J.B. Sollberger 1978 Replication and Classification of Small Size Lithic Debi tage. Plains Anthropologist 23:103-112. Phelps, David s. 1965 The Norwood Series of Fiber-tempered Ceramics. Southeastern Archaeological Conference Bulletin 2:65-69. Purdy, Barbara A. 1975 The Senator Edwards Chipped Stone Workshop Site (Mr-122), Marion County, Florida: A Preliminary Report of Investigations. Florida Anthropologist 28:178-189. 1976 The Application of Instrumental Techniques to

PAGE 231

218 Prehistoric Remains of Florida. Lithic Technology 5:5-6. 1977 The York Site (8-Al-480), Alachua County, Florida: Observations on Aboriginal Use of Chert. Florida Anthropologist 30:3-8. 1981 Florida's Prehistoric Stone Technology: A study of the Flintworking Technique of Early Florida Stone Implement Makers. University Press of Florida, Gainesville. 1988 Wet Site Archaeology. Barbara Purdy, Ed. Telford Press: New Jersey. 1990 Personal Communication. Purdy, Barbara and Frank N. Blanchard 1973 Petrograph as a Means of Tracking Stone Tools From Florida. Florida Anthropologist 26(1):121-125. Puri, Harbans S. and Robert 0. Vernon 1964 Summary of the Geology of Florida and a Guide to the Classic Exposures. Florida Geological Survey Special Publication No. 5. Raab, Mark L., Robert F. Cande, and David W. Stahle 1979 Debitage Graphs and Archaic Settlement Patterns in the Arkansas Ozarks. Mid-Continental Journal of Archaeology 4(2):167-182. Schiffer, Michael B. 1972 Archaeological Context and Systematic Context. American Antiquity 37(2):156-165. 1976 Behavioral Archaeology. Academic Press: New York. Scholl, D.W. 1963 Sedimentation in Modern Coastal swamps, 1964 Southwestern Florida. In American Association Petrol. Geological Society of America, Bulletin 47:1581-1603. Recent Sedimentary Record in Mangrove Swamps and Rise in Sea Level over the Southwestern Coast of Florida, Part 1 Marine Geology 1(4):344-366. Scholl, D.W., F.C. Craighead, Sr., and Stuiver 1969 Florida Submergence Curve Rev1sed: Its Relation to coastal Sedimentation Rates. Science 163:562-564.

PAGE 232

219 Scholl, D.R. and M. Stuiver 1967 Recent Submergence of South Florida: A Comparison with Adjacent Coasts and Other Eustatic Data. Bulletin of the Geological Society of America 78(4):437-454. Scientific Applications, Inc. 1981 A Cultural Resources Survey of the Continental Shelf from Cape Hatteras to Key West. Volume 1. Introduction and Physical Environment. Report Submitted to the Bureau of Land Management, Atlanta, Georgia. Seabury, Patricia, N. Swilley, J. Morten, and A. Derehnaucourt 1975 An Archaeological Survey of the Tampa By-Pass Canal Right-of-Way, Hillsborough County, Florida. Ms. on file, Florida Division of Archives, History and Records Management, Tallahassee. Sears, William H. 1958 The Maximo Point Site. 11(1):1-10. Semenov, S.A. 1964 Prehistoric Technology. Dart. Sharon, D.W. and T.C. Watson Florida Anthropologist Somerset: Adama and 1971 The Two Egg Quarry Site. Florida Anthropologist 24(2):77-80. Shepard, F.J. 1964 Sea Level Changes in Possible Archaeological 143:574-576. the Past 6000 Years: Significance. Science Shepard, James 1886 Shell Heaps and Mounds in Florida. Smithsonian Institution Annual Report 1885:902-906. Simpson, 1937 1938 1941 J. Clarence Report on Activities in Hillsborough County, Florida. Florida State Board of Conservation. second Biennial Report, pp. 109-116. Tallahassee. Aboriginal Stone Quarries of Hillsborough County and Sources of Abrasives and Pigment. Third Biennial Report, Florida State Board of conservation. source Material for Florida Aboriginal Artifacts.

PAGE 233

Sinclair, 1973 220 Journal of the Florida Academy of sciences 5:32-34. w.c. Hydrologic Characteristics of the surficial Aquifer in Northwest Hillsborough County, Florida. Information Circular No. 86, Florida Bureau of Geology, Tallahassee. stapor, F.W. and W.F. Tanner 1977 Late Holocene Mean Sea Level Data from St. Vincent Island and the Shape of the Late Holocene Mean Sea Level Curve. In Coastal Sedimentology, William F. Tanner, Ed. Coastal Research and the Department of Geology, Florida State University, Tallahassee, Florida. Stearns, R.E.C. 1870 Rambles in Florida. American Naturalist 3:349360; 395-405; 455-470. 1872 Remarks on Mounds and Shell Heaps of Tampa Bay, Florida. Proceedings of the California Academy of Science. Series 1 (4):214-215. Ste. Claire, Dana M. 1985 Thermal Alteration as a Cultural/Temporal Marker in Florida. Paper Presented at the Thirty-Seventh Annual Meeting of the Florida Anthropological Society, Daytona. Stirling, 1930 1935 1936 Matthew W. Prehistoric Mounds in the Vicinity of Tampa Bay, Florida. Explorations and Fieldwork of the smithsonian Institute in 1929. pp. 183-186. Washington, D.C. Archaeological Work in Florida. Explorations and Field-Work of the Smithsonian Institution 1934:5760. Florida Cultural Affiliations in Relation to Adjacent Areas. In Essays in Anthropology in Honor of Alfred Louis Kroeber, pp.351-357. University of California Press: Berkeley. Storck, Peter L. Margin and Early PaleoMid-Continent Region. In of Archaeology, Vol. 1988 The Late Wisconsin Ice Indian Occupation in the Midcontinental Journal 13(2):259-272.

PAGE 234

221 Thanz, Nina 1975 of Environmental and Cultural Change 1n Northeastern Florida During the Late Archaic. Manuscript on file, Department of Anthropology, University of Florida, Gainesville. Torp, Lyle C. 1991 Archaeological Investigations at the Cowhouse East Head and Cowhouse West Head Sites: Final Report to the Southwest Florida Water Management District. Manuscript on File, Department of Anthropology, University of South Florida, Tampa. Tringham, R.E., G. Cooper, G. Odell, B. Voytek, and A. Whitman 1974 Experimentation in the Formation of Edge Damage: A New Approach to Lithic Analysis. Journal of Field Archaeology 1:171-196. United States Department of Agriculture 1958 Soil Manual for Hillsborough County, Florida. 1989 Upchurch, 1980 1982 1991 Upchurch, 1981 Soil Manual for Hillsborough county, Florida. Sam B. Chert Origins and Availability Pinellas and Hillsborough Counties. In Holocene Geology and Man in Pinellas and Hillsborough Counties. Compiled by Sam B. Upchurch. pp. 48-62. Southeastern Geological Society: Tallahassee. Geology of Archaeological Site 8-Hi-471 on Cypress Creek, Hillsborough County, Florida. In Alroy Archaeological Excavations at the Cypress Creek Site (Appendix B). Interstate 75 Highway Phase II Archaeological Reports Number 4: Tallahassee. Personal Communication. Sam B., Richard N. Strom and Mark G. Nuckels Silicification of Miocene Rocks from Central Florida. symposium on the Miocene of the southeastern United States. Southeastern Geological Society and Florida Bureau of Geology. Vita-Finzi, c., and E.S. Higgs 1970 Prehistoric Economy in the Mount Carmel Area of Palestine: Site Catchment Analysis. Proceedings of the Prehistoric Society. Volume 36:1-37. Vodges, A.W. 1879 Notes on a Lost Race of America. American Naturalist 13:9-11.

PAGE 235

222 Walker, S.T. 1880a Report on the Shell Heaps of Tampa Bay, Florida. Annual Report of the Smithsonian Institution for 1879: 413-422. Washington, D.C. 1880b Preliminary Explorations Among the Indian Mounds in Southern Florida. Annual Report of the Smithsonian Institution for 1879:854-868. Washington, D.C. 1883 The Aborigines of Florida. Smithsonian 1885 Institution Annual Report 1879:854-868. Mounds and Shell Heaps on the West Coast of Florida. Annual Report of the Smithsonian Institution for 1883:854-868. Washington, D.C. Waller, Benjamin T. 1969 Paleo-Indian and other Artifacts from a Florida Streambed. Florida Anthropologist 22:37-39. 1972 Some Occurrences of Paleo-Indian Projectile Points in Florida Waters. Eastern States Archaeological Federation Bulletin 31:14-15. Watts, w. A. 1969 A Pollen Diagram From Mud Lake, Marion County, North-central Florida. Geological Society of American Bulletin 80:631-642. 1971 Postglacial and Interglacial Vegetation History of Southern Georgia and Central Florida. Ecology 52:676-690. 1975 A Late Quaternary Record of Vegetation From Lake Anne, South-central Florida. Geology 3:344-346. 1980 Late Quaternary Vegetation History at White Pond on the Inner Coastal Plain of South Carolina. Quaternary Research 13:187-199. watts, W.A. and M. stuiver 1980 Late Wisconsin Climate of Northern Florida and the Origin of Species Rich Deciduous Forest. Science 210:325-327. Webb, S. 1981 David A Cultural Resources Survey of the Continental Shelf from Cape Hatteras to Key West. Volume 1: Introduction and Physical Environment, pp. I-73 to I-112. Report Submitted by Science Applications, Inc., to the Bureau of Land Management.

PAGE 236

223 Weisman, Brent 1989 Like Beads on a String. University of Alabama Weisman, 1988 Wharton, 1979 Press, Tuscaloosa. Brent R. and William H. Marquardt A Comprehensive Archaeological Resource Inventory for the Southwest Florida Water Management District, Brooksville, Florida. Department of Anthropology, Florida Museum of Natural History, University of Florida, Gainesville. Barry An Archaeological Assessment Survey of the Construction Impact Areas of the Upper Hillsborough Flood Detention Area, Southeastern Pasco County, Florida. Manuscript on File, University of South Florida, Department of Anthropology. Tampa. Wharton, Barry, George Ballo, and Mitchell Hope 1981 The Republic Groves Site, Hardee County, Florida. Florida Anthropologist 34:59-80. White, Anta M. 1963 Analytic Description of the Chipped-stone Industry from Snyders Site, Calhoun County, Illinois. In Miscellaneous Studies in Typology and Classification. Anthropological Papers, Museum of Anthropology, University of Michigan, Number 19. Whitehead, P.R. 197 3 Late Wisconsin Vegetational Changes in Unglaciated Eastern North America. Quaternary Research 3:621-631. Willey, Gordon R. 1949 Archaeology of the Florida Gulf Coast. 1966 Smithsonian Miscellaneous Collections Volume 113. Washington, D.C. An Introduction to American Archaeology Part One: North and Middle America. Prentice Hall: Englewood Cliffs, New Jersey. Williams, J. Raymond 1990 Personal Communication. Williams, 1965 Stephen and James Stoltman An outline of southeastern Un1ted States Prehistory With Particular Interest on the PaleoIndian Era. In Quaternary of the United States, H.E. Wright, Jr. and David G. Frey, Eds., pp. 669-683. Princeton University Press: Princeton.

PAGE 237

224 Edwin H. Wilmsen, 1970 Lithic Analysis and Cultural Inference: A PaleoIndian Case. Anthropological Paper Number 16. University of Arizona, Tucson. Wright, Alexandra P. 1973 Environmental Geology and Hydrology, Tampa Area, Florida. Florida Bureau of Geology Special Publication No. 19, Tallahassee. Wright, H.E., Jr. 1981 Vegetation Years Ago. East of the Rocky Mountains 18,000 Quaternary Research 15:113-125. Wyman, Jeffries 1870 From Explorations in Florida. Peabody Museum of American Archaeology and Ethnology, Harvard University Reports 3:8-9. Yellen, John E. 1977 Archaeological Approaches Academic Press: New York. to the Present.


Download Options

No images are available for this item.
Cite this item close


Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.


Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.


Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.


Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.