Human Biological Variation and Biological Distance in Pre-Contact Florida

MISSING IMAGE

Material Information

Title:
Human Biological Variation and Biological Distance in Pre-Contact Florida a Morphometric Examination of Biological and Cultural Continuity and Change
Physical Description:
1 online resource (280 p.)
Language:
english
Creator:
Kles, Maranda Almy
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Anthropology
Committee Chair:
Warren, Michael Wade
Committee Co-Chair:
Krigbaum, John
Committee Members:
Sassaman, Kenneth Edward, Jr
Burt, Martha

Subjects

Subjects / Keywords:
archaic -- bioarchaeology -- biodistance -- craniometrics -- manasota -- morphometrics
Anthropology -- Dissertations, Academic -- UF
Genre:
Anthropology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
The archaeological relationships of cultures have long been analyzed and debated; however, the interpretation of the biological relationship of the associated populations labeled is relatively new. Cranial morphology has been used to understand origins and relationships of races or ancestry groups. More recently these analyses have been applied to archaeological populations. However, this type of bioarchaeological approach has been rarely used in the Southeast. The analysis of cranial variation within samples from sites associated with known cultural labels has the potential to provide new insight into the biological relationship or interaction of the individuals buried at these archaeological sites. This study utilizes 20 craniometric variables from 404 individuals associated with 27 sites across present-day Florida. These sites are associated with a variety of cultural labels and varied geographic and environmental settings. To statistically analyze the craniometric variables, I conducted Principle Component Analyses, Discriminant Function Analyses, and R-matrix Analyses to investigate the patterns of biological variation found within these populations and to determine if those patterns correlate with cultural, temporal, or geographic differences. It was determined that there were at least two migration events into the peninsula after the Paleoindian period. Secondly, patterns were found that suggest parent-descendant group relationships for many of the Archaic populations assessed. Additionally, there was a biological distinction between the populations associated with the “classic” Weeden Island and Manasota cultures. Lastly, it appears that the Weeden Island and Mississippian cultures had culturally recognized barriers to biological interaction that resulted in greater genetic variation within these populations.  These findings indicate that there are discernible patterns to human biological variation found within the pre-Contact populations of Florida and that this variation often corresponds to the cultural variation interpreted from the archaeological record. However, a few notable exceptions in the Archaic and Weeden Island populations show that additional work is needed both on the biological and archaeological fronts. This research has broad implications with regard to assigning unprovenienced skeletal remains to cultures or sites, which contributes to the analysis of cultural affiliation relevant to Native American Graves Protection and Repatriation Act and fosters constructive dialogue between all interested parties.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Maranda Almy Kles.
Thesis:
Thesis (Ph.D.)--University of Florida, 2013.
Local:
Adviser: Warren, Michael Wade.
Local:
Co-adviser: Krigbaum, John.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-05-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2013
System ID:
UFE0044952:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 HUMAN BIOLOGI C AL VARIATION AND BIO LOGICA L DISTANCE IN PRE CONTACT FLORIDA: A MORPHOMET RIC EXAMINATION OF BIOLO GICAL AND CULTURAL CONTINUITY AND CHANG E By MARANDA ALMY KLES A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013

PAGE 2

2 2013 Maranda Almy Kles

PAGE 3

3 T o the two people in my life who made this possible : my husband, Joseph Kles, and my mother, Marion Almy.

PAGE 4

4 ACKNOWLEDGMENTS This dissertation would not have been possible without the support and guidance of many pe ople, probably more than I list here. Firstly I would like to acknowledge my co c hairs Dr. Michael Warren and Dr. John Krigbaum who have helped me become the biological anthropolog ist I am through their mentorship from my undergraduate years onward. Also, many thanks to Dr. Kenneth Sassaman for piquing my interest in anthropology in hi s General Anthropology course and second for helping me go from the daughter of an archaeologist to a biological anthropologist who better understands the complexities of the archaeolog ical record Secondly, I want to acknowledge the people who have provi ded insight and input on my research. Dr. George Luer h is input and support ha ve helped me immensely. I appreciate the invaluable information provided by Donna Ruhl, Louis Tesar, Dr. Robert Austin, and Dr. Dav id Dickel about sites and collections Dr. Jerald Milanich who has been a mentor in various ways and who showed me many years ago that you can blend archaeology, history, and biological anthropology I am also indebted to Katie Miyar for traveling with me to collections, allowing me to stay in her house while working, and providing me with a supportive ear through this process. In a ddition, Dr. Sherry Hutt for discussions that led to this topic, Dr. Jan Matthews for being a friend, Dr. Margo Schwadron for her support, guidance, and friendship, and many others whom I am forgetting for input on various topics. Thirdly, I must acknowledge those institutions and the people who c are for : The Florida Museum of Natural History, Florida Department o f Historic Resources, Florida State University, Florida Gulf Coast University, Florida Atlantic University, University of Miami, Sarasota County

PAGE 5

5 History Center, the Smithsonian Institution, Donna Ruhl, Dr. Neill Wallis, Elise LeCompte, Dr. Dav id Dickel Dr Glen Doran, Dr. Heather Walsh Haney, Dr. Arlene Fradkin, Dr. John Gifford, Monica Faraldo, Jodi Prachett, and Dr. David Hunt. No acknowledgement would be complete without thanking my family for their love and support. In particular I want to thank my m om despite her many efforts and best intentions, I thank her for starting me in this field (even before I was born) and helping me find my true passion (that fateful trip to the cemetery in Fort Myers). It is my hope that your granddaughter follows in your footsteps and becomes an anthropologist! Also, thank you to Grandpa Sheldon and Grandpa Dick for the many tiring hours put in at the end entertaining their granddaughter so I could work I t would have taken ten times longer without you. Lastly, I want to thank Joseph Kles, my husband, without whom I never could have done this. When I told you I wanted to come back to school and you said yes I knew you were the one for me, even if you had no idea what you were getting yourself into. Five years later her e we are and you have been everyt hing to me whenever I needed it -words can never express my gratitude.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURE S ................................ ................................ ................................ ........ 11 ABSTRACT ................................ ................................ ................................ ................... 14 CHAPTER 1 INTRODUCTION AND STATEMENT OF QUESTIONS ................................ ......... 16 2 CULTURE HISTORY AND ARCHAEOLOGY OF FLORIDA ................................ .. 28 Paleoindian Period ................................ ................................ ................................ .. 28 Archaic Period ................................ ................................ ................................ ........ 29 Woodland Period ................................ ................................ ................................ .... 32 North Florida ................................ ................................ ................................ ..... 34 North and Nort hwest ................................ ................................ .................. 35 North Central ................................ ................................ .............................. 37 Northeast ................................ ................................ ................................ ... 39 Southern Peninsular Gulf Coast and South Florida ................................ .......... 40 Southern Peninsular Gulf Coast ................................ ................................ 40 South Florida ................................ ................................ .............................. 42 Mississippian Period ................................ ................................ ............................... 45 Summary ................................ ................................ ................................ ................ 46 3 BIOLOGICAL DISTANCE ANALYSIS: HISTORY, THEORY, AND METHODS ..... 51 Biological Distance Studies on Living Populations ................................ .................. 52 Biological Distance Studie s on Archaeological Populations ................................ .... 53 Metric Versus Non Metric Traits ................................ ................................ ............. 59 Biological Distance Theory, Assumptions, and Models ................................ .......... 61 Theory ................................ ................................ ................................ .............. 61 Assumptions ................................ ................................ ................................ ..... 62 Assumptions analyzed ................................ ................................ ..................... 62 Model derivations ................................ ................................ ............................. 67 Methodologies ................................ ................................ ................................ .. 70 4 METHODS AND MATERIALS ................................ ................................ ................ 72 Data Collected ................................ ................................ ................................ ........ 72 Missing Data ................................ ................................ ................................ ........... 74 Statistical Methods ................................ ................................ ................................ .. 77

PAGE 7

7 Materials ................................ ................................ ................................ ................. 80 Archaic Sites ................................ ................................ ................................ .... 80 Warm Mineral Springs (8SO19) ................................ ................................ 80 Windover Pond (8BR246) ................................ ................................ .......... 81 Little Salt Spring (8SO18) ................................ ................................ .......... 81 Bay West (8CR200) ................................ ................................ ................... 82 Republic Groves (8HR4) ................................ ................................ ............ 82 Gauthier (8BR193) ................................ ................................ ..................... 83 Bird Island (8DI52) ................................ ................................ ..................... 83 Weeden Island/Manasota Sites ................................ ................................ ........ 84 McKeithan (8CO17) ................................ ................................ ................... 84 Crystal River (8CI1) ................................ ................................ ................... 84 Hughes Island Mound (8DI45) ................................ ................................ ... 85 Yellow Bluffs Mound (8SO4) ................................ ................................ ...... 86 Dunwody (8CH61) ................................ ................................ ..................... 86 Casey Key (8SO17) ................................ ................................ ................... 87 Palmer Bur ial Mound (8SO2A) ................................ ................................ ... 87 Manasota Key Cemetery (8SO1292) ................................ ......................... 88 Venice Beach Complex (8SO26) ................................ ............................... 88 Bayshore Homes (8PI41) ................................ ................................ ........... 89 Bay Pines Site (8PI64) ................................ ................................ ............... 89 South Florida Sites ................................ ................................ ........................... 90 Captiva Mound (8LL57) ................................ ................................ ............. 90 Hutchinson Island Burial Mound (8MT37) ................................ .................. 90 Highlands Beach Mound or Boca Raton Beach Burial Mound (8PB11) ..... 91 Margate Blount Mound (8BD41) ................................ ................................ 91 Safety Harbor Sites ................................ ................................ .......................... 92 Sarasota Bay Mound (8S O44) ................................ ................................ ... 92 Safety Harbor (8PI2) ................................ ................................ .................. 92 Alachua/St. Johns/Fort Walton Sites ................................ ................................ 92 Henderson Mound (8AL463) ................................ ................................ ...... 92 Brow ne Site 5 (8DU62) ................................ ................................ .............. 93 Santa Rosa Mound or the Fort Walton Temple Mound (8OK6) ................. 93 5 RESULTS AND DISCUSSION ................................ ................................ ............... 97 All Sites ................................ ................................ ................................ ................... 98 Results ................................ ................................ ................................ ............. 98 Discussion ................................ ................................ ................................ ...... 101 Archaic ................................ ................................ ................................ .................. 102 Results ................................ ................................ ................................ ........... 102 Discussion ................................ ................................ ................................ ...... 106 Weeden Island/Ma nasota ................................ ................................ ..................... 111 Weeden Island/Manasota results ................................ ................................ ... 111 Weeden Island results ................................ ................................ .................... 113 Manasota results ................................ ................................ ............................ 115 Discussion ................................ ................................ ................................ ...... 118

PAGE 8

8 South Florida ................................ ................................ ................................ ........ 126 Results ................................ ................................ ................................ ........... 126 Discussion ................................ ................................ ................................ ...... 129 Weeden Island/Alachua/St. Johns/Fort Walton ................................ ..................... 130 Results ................................ ................................ ................................ ........... 130 Discussion ................................ ................................ ................................ ...... 132 Manasota/Safety Harbor ................................ ................................ ....................... 133 Results ................................ ................................ ................................ ........... 133 Discussion ................................ ................................ ................................ ...... 136 Biological Varia tion And Gene Flow ................................ ................................ ...... 137 Archaic Period ................................ ................................ ................................ 138 Woodland Period ................................ ................................ ............................ 138 Discussion ................................ ................................ ................................ ...... 139 Biodistance Assumptions ................................ ................................ ...................... 142 6 CO NCLUSIONS ................................ ................................ ................................ ... 198 Summary ................................ ................................ ................................ .............. 198 Limitations to This Research ................................ ................................ ................. 204 Future Research ................................ ................................ ................................ ... 206 Broader Impact ................................ ................................ ................................ ..... 207 APPENDIX A DEFINITION OF MEASU REMENTS ................................ ................................ .... 208 B RAW DATA ................................ ................................ ................................ ........... 211 C TIME AND GEOGRAPHY DISTANCE MATRIX ................................ ................... 225 D RESULTS ................................ ................................ ................................ ............. 22 8 LIST OF REFERENCES ................................ ................................ ............................. 268 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 280

PAGE 9

9 LIST OF TABLES Table page 2 1 General Scheme of Florida Culture Periods. ................................ ...................... 50 4 1 Site Information ................................ ................................ ................................ .. 95 4 2 Meas urement taken ................................ ................................ ............................ 96 4 3 Measurements used. ................................ ................................ .......................... 96 5 1 The biological distances and standard errors for those distances between all populations. ................................ ................................ ................................ ...... 149 5 2 The biological distances and standard errors for those distances all the Archaic populations. ................................ ................................ ......................... 158 5 3 The biological distances and standard errors for those distances for the third analysis of the Archaic populations. ................................ ................................ 162 5 4 The biological distances and standard errors for those distances between the Weeden Island/Manasota populations. ................................ ............................. 168 5 5 The biological distances and standard errors for those distances between the Weeden Island populations. ................................ ................................ ............. 172 5 6 The biological distances and standard errors for those distances between for the second analysis of the Weeden Island populations. ................................ ... 173 5 7 The biological distances and standard errors for those distances between the Manasota populations. ................................ ................................ ..................... 176 5 8 The biological distances and standard errors for those distances for the fourth analysis of the Manasota populations. ................................ ................... 180 5 9 The biological distances and standard errors for those distances between the South Florida populations. ................................ ................................ ................ 184 5 10 The biological distances and standard errors for those distances between the Weeden Island/Alachua/St. Johns populations. ................................ ............... 188 5 11 The biological distances and standard errors for those distances between the Weeden Island/Fort Walton populations. ................................ .......................... 191 5 12 The biological distances and standard errors for those distances between the Weeden Island/Fort Walton/Manasota/Safety Harbor populations .................. 193

PAGE 10

10 5 13 The biological distances and standard errors for those distances for the third analysis of the Manasota/Safety Harbor pop ulations ................................ ....... 195 5 14 The biological distances and standard errors for those distances for the fourth analysis of the Manasota/Safety Harbor populations ............................ 197 D 1 Independent sample t test for equality of means ................................ .............. 228 D 2 Principle Component Results ................................ ................................ ........... 231 D 3 Discriminant Function Results ................................ ................................ .......... 234 D 4 Cross validation percentage results for Discriminant Function Analysis ........... 237 D 5 R Matrix Results ................................ ................................ ............................... 250 D 6 Th e biological distances and standard errors for those distances between the sites. ................................ ................................ ................................ ................. 261 D 7 Correlation Matrix: Biodistance to Tempo ral and Geographic distances .......... 267

PAGE 11

11 LIST OF FIGURES Figure page 1 1 Map of the location and cultural association of the sites used in this study. ....... 27 5 1 Graph of the first and second factors for the PCA analysis of all the populations identified by cultural label. ................................ ............................. 146 5 2 Graph of the first and second discriminant functions for all populations. The cluster of Archaic populations is circled in blue. ................................ ............... 147 5 3 Graph of the first and second eigenvectors for all p opulations. ........................ 148 5 4 Graph of the first and second factors for the PCA analysis for the Woodland and Mississippian populations. The cluster of Manasota populations is circled in green and the Weeden Island populations in red. ................................ ......... 153 5 5 Graph of the first and second discriminant functions for the Woodland and Mississippian populations. The cluster of Manasota populations is circled in green and the Weeden Island populations in red. ................................ ............ 154 5 6 Graph of the first and second factors for the PCA analysis for all Archaic populations. ................................ ................................ ................................ ...... 155 5 7 Graph of the first and second discriminant functions for all Archaic populations. ................................ ................................ ................................ ...... 156 5 8 Graph of the first and second eigenvalue for all Archaic populations. .............. 157 5 9 Graph of the first and second discriminant functions for the second an alysis of the Archaic populations. ................................ ................................ ............... 159 5 10 Graph of the first and second eigenvectors for the second analysis of the Archaic pop ulations. ................................ ................................ ......................... 160 5 11 Graph of the first and second factors for the PCA analysis for the third analysis of the Archaic populations. The Republic Groves population is circled in light blue. ................................ ................................ ........................... 161 5 12 Graph of the first and second eigenvectors for the third analys is of the Archaic populations. ................................ ................................ ......................... 162 5 13 Graph of the first and second factors for the PCA analysis for the fourth analysis of the Archaic populations. ................................ ................................ 163 5 14 Graph of the first and second discriminant functions for the fourth analysis of the Archaic populations. ................................ ................................ ................... 164

PAGE 12

12 5 15 Graph of the first and second factors for the PCA analysis for the first analysis of the Weeden Island/Mana sota populations. ................................ ..... 165 5 16 Graph of the first and second discriminant functions for first analysis of the Weeden Island/Manasota p opulations. ................................ ............................. 166 5 17 Graph of the first and second eigenvectors for the first analysis of the Weeden Island/Manasota populations. ................................ ............................. 167 5 18 Graph of the first and second eigenvectors for the second analysis of the Weeden Island/Manasota populations. ................................ ............................. 169 5 19 Graph of the first and second discriminant functions for the first analysis of the Weeden Island populations. ................................ ................................ ....... 170 5 20 Graph of the first and second eigenvectors for the first analysis of the Weeden Island populations. ................................ ................................ ............. 171 5 21 Graph of the first and second eigenvectors for the second analysis of the Weeden Island populations. ................................ ................................ ............. 172 5 22 Graph of the first and second factors for the PCA analysis for the first analysis of the Manasota populations. The Dunwody population is circled in yello w and the Yellow Bluffs populations in light blue. ................................ ...... 173 5 23 Graph of the first and second discriminant functions for first analys is of the Manasota populations. ................................ ................................ ..................... 174 5 24 Graph of the first and second eigenvectors for the first analysis of the Manasota populations. ................................ ................................ ..................... 175 5 25 Graph of the first and second discriminant functions for the third analysis of the Manasota populations. ................................ ................................ ............... 177 5 26 Graph of the first and second eigenvectors for the third analysis of the Manasota populations. ................................ ................................ ..................... 178 5 27 Graph of the first and second factors for the PCA analysis for the fourth analysis of the Manasota populations. ................................ .............................. 179 5 28 Graph of the first and second eigenvectors for the fourth analysis of the Manasota populations. ................................ ................................ ..................... 180 5 29 Graph of the first and second factors for the PCA analysis for the first analysis of the South Florida populations. The Dunwody population is circled in yellow. ................................ ................................ ................................ ........... 181

PAGE 13

13 5 30 Graph of the first and second discriminant functions for first analysis of the South Florida populations. ................................ ................................ ................ 182 5 31 Graph of the first and second eigenvectors for the first analysis of the South Florida populations. ................................ ................................ .......................... 183 5 32 Graph of the first and second discriminant functions for second analysis of the South Florida populations. ................................ ................................ .......... 184 5 33 Graph of the first and second factors for the PCA analysis for the analysis of the Weeden Island/Alachua/St. Johns populations. ................................ .......... 185 5 34 Graph of the first and second discriminant functions for analysis of the Weeden Island/Alachua/St. Johns populations. ................................ ............... 186 5 35 Graph of the first and second eigenvectors for the analysis of the Weeden Island/Alachua/St. Johns populations. ................................ .............................. 187 5 36 Graph of the first and second factors for the PCA analysis for the analysis of the Weeden Island/Fort Walton populations. ................................ .................... 188 5 37 Graph of the first and second functions for the analysis of the Weeden Island/Fort Walton populations. ................................ ................................ ........ 189 5 38 Graph of the first and second eigenvectors for the analysis of the Weeden Island/Fort Walton populations. ................................ ................................ ........ 190 5 39 Graph of the first and second functions for the first analysis of the Weeden Island/Fort Walton/Manasota/Safety Harbor populations. ................................ 191 5 40 Graph of the first and second eigenvectors for the first analysis of the Weeden Island/Fort Walton/Manasota/Safety Harbor populations. .................. 192 5 41 Graph of the first and second functions for the third analysis of the Manasota/Safety Harbor populations. ................................ .............................. 194 5 42 Graph of the first and second eigenvectors for the third analysis of the Manasota/Safety Harbor populations. ................................ .............................. 195 5 43 Graph of the first and second functions for the fourth analysis of the Manasota/Safety Harbor populations. ................................ .............................. 196 5 44 Graph of the first and second eigenvectors for the fourth analysis of the Manasota/Safety Harbor populations. ................................ .............................. 197

PAGE 14

14 Abstract of Dissertation Presented to the Graduate School of th e University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy HUMAN BIOLOGICAL VARIATION AND BIOLOGICAL DISTA NCE IN PRE CONTACT FLORIDA: A MORPHOMET RIC EXAMINATION OF BIOLO GICAL AND CULTURAL CONTINUITY AND CHANG E By Maranda Almy Kles May 2013 Chair: Michael Warren Co chair: John Krigbaum Major: Anthropology The archae ological relationships of cultures have long been analyzed and debated ; however the interpretation of the biological relationship of the associated populations labeled is relatively new. Cranial morphology has been used to understand origins and relations hips of races or ancestry groups M ore recently these analyses have been applied to archaeological populations. However, this type of bioarchaeological approach has been rarely used in the Southeast. The analysis of cranial variation within samples from si tes associated with known cultural labels has the potential to provide new insight into the biological relationship or interaction of the individuals buried at these archaeological sites. This study utilizes 20 craniometric variables from 404 individuals associated with 27 sites across present day Florida. These sites are associated with a variety of cultural labels and varied geographic and environmental settings. To statistically analyze the craniometric variables, I conducted Principle Component Analys e s, Discriminant Function Analys e s, and R matrix Analys e s to investigate the patterns of biological

PAGE 15

15 variation found within these populations and to determine if those patterns correlate with cultural, temporal, or geographic differences. It was determined t hat there were at least two migration events into the peninsula after the Paleoindian period. Secondly, patterns were found that suggest parent descendant group relationships for many of the Archaic populations assessed Additionally, there was a biologica l distinction between the populations associated with the s Lastly it appears that the Weeden Island and Mississippian cultures had culturally recognized barriers to biological interaction that resulted in grea ter genetic variation within these populations. These findings indicate that there are discernible patterns to human biological variation found within the pre Contact populations of Florida and that this variation often corresponds to the cultural variation interpreted from the archaeological record. However, a few notable exceptions in the Archaic and Weeden Island pop ulations show that additional work is needed both on the biological and archaeological fronts. This research has broad implications with regard to assigning unprovenienced skeletal remains to cultures or sites which contributes to the analysis of cultural affiliation relevant to Native American Graves Protection and Repatriation Act and fosters constructive dialogue between all interested parties

PAGE 16

16 CHAPTER 1 INTRODUCTION AND STATEMENT OF QUESTIONS Florida has a long history of study in archaeology and physical anthropology; however, this research has rarely addressed the role of biological variation with respect to interpretation s of culture history. This is in contrast to intersecting biological and cultural studies conducted in the Southwest, Mid West, and northern Mid West w here, for example s cholars have studied migration (Steadman 2001; Watson 2010) marriage patterns (Schillaci and Stojanowski 2005; Stojanowski 2005; Stojanowski and Schillaci 2006) trade networks (Reichs 1984) language divi sions (Killgrove 2009) intrasite mound relationships (Konigsberg 1990b) and culture change (Buikstra 1976; Ousley et al. 2005; Sciulli et al. 1984; Steadman 1998) Following the example of these and other studies, this dissertation examines the range of cranial morphological variation within selected Florida populations dating from the Archaic to the pre Contact period (ca 10,000 700 years before present [ BP ] ) and secondly seeks correlation s between the biological variation and cultural variation found in those populations Specifically, this research focuses on the Archaic, the Weeden Island / Manasota culture Complex and the Glades / Caloosahatchee cultures of South Florida. These three foci were selected primarily because of the broad temporal and geogr aphic distribution of the sites across the Florida peninsula and the availability of skeletal material for analysis (Figure 1 1). The use of morphometry for the purpose of understanding the biological variation of a population has a long history in physic al anthropology. What began as a typological approach to understanding morphological differences between populations (Hooton has changed through time to an attempt to understand and appreciate the full range of human biol ogical variation. However,

PAGE 17

17 integrating morphological approaches with archaeology to understand cultural variation and population histories is relatively new. Research has shown that the analysis of biological variation within a skeletal population can help evaluate microevolutionary forces that impacted the population and therefore that biological history (Konigsberg 1990a; Rel ethford and Blangero 1990; Relethford and Lees 1982; Steadman 1998) The first published effort using biodistance with specific foc us on the archaeological record is attributable to a 1972 article by Lane and Sublett (1972) who analyzed residency patterns of individuals buried in Native American cemetery sites in New York State T he field has since expanded, and population genetics has increasingly been used to ground theoretical and methodological advances adding The most thorough study of biological variation in Florida is that of Hrdli ka (1922) Hrdli ka used a cranial mean height index ( h eight / (mean of length and breadth)) to differentiate between populations and found that a population of long headed and less robust in dividuals migrated predominantly into east and south Florida replacing or intermixing with the older more robust round headed populations. He stated that overall, the skulls he examined were rather uniform in appearance due to the admixture which crea ted numerous intermediary forms; however, he could discern two types on closer inspection T he first type, the earlier, round headed population, con sisted of individuals from the n orthern two thirds of the peninsula, including the Indians of North Florida, th Charlotte Harbor. The second type, the newer long headed population, could be found along the East coast and from the region of Lake Okeechobee south

PAGE 18

18 Though extensive and thorough for its time there are several vexing issues with this study. First, Hrdli ka used a single cranial index to differentiate the populations, which as Howells (1973) Hrdli ka had very little knowledge of the true temporal relationship of the remains; he assessed the age of the remains based on taphonomic generalizations and not based on other indicators of temporal relationship s such as associated pottery styles or burial practices, thus calling into question his assessment of the temporal sequence of his study popul ations. Third, his samples from the East coast and South Florida were limited to only 26 individuals as compared to 134 individuals from the West Coast and St. Johns regions The low number of individuals from this group rais es questions about how represe ntative his samples were of the true biological variability of the regions sampled A more recent study by Seasons (2010) addresses several of the issues found in Hrdli large r sample sizes, and more rigorous statistical analyses. Seasons (2010) conducted an analysis of seven South Florida populations spanning 8120 2 47 BP and found that there was a significant amount of variation among the populations studied and this variation was greatest between sites that were separated by the largest temporal gap. Her study also found relative homogeneity, or biological similarity, among the populations dated from 2510 247 BP. Although very valuable, t his study has a few limitations includi ng its use of only seven sites and t he temporal gap between s ome of

PAGE 19

19 the sites is quite large limiting interpretations of cultural continuity and change, due to potential secular trends within the samples. Other researchers have examined biological variati on in Florida; however, their focus has been more on Contact period populations with infrequent mention of pre Contact populations For example, Stojanowski ( 2003; 2004; 2005a; 2005b; Stojanowski et al. 2007) has conducted numerous studies of the Apalachee and Guale populations in North Florida. In doing so he has compared the Mission Period populations to several pre Contact population predominately from the Mississippian period, however he has included a few earlier populations, such as Irene Mound, Tathum Mound, and McKeithan. Stojanowski found that pre Contact populations were more homogenous than mission populations; however, there is still biological variation evident among those pre Contact populations (Stojanowski 2004; Stojanowski 2005a) In addition, Stojanowski has stated that there are biological distinctions between populations in northern and southern Florida (Stojanowski 2005b) similar to findings Stojanowski also conducted an intras ite analysis on the Ar chaic Windover Pond population and found that there were two genetically distinguishable groups using the pond for burial and that those biological divisions corresponded to differing burial artifacts (Stojanowski and Schillaci 2006) All of these studies are informative and add to the data available for the interpretation of cultural change and population structure, h owever, the results about pre Contact biological variation, particular for the earlier populations, o ften only receive cursory remark s and the interesting data for Windover Pond is only minimally presented in a n in depth paper focused on methodology for intracemetery biodistance analysis.

PAGE 20

20 In total, t hese studies show that the analysis of biological variation and biologica l distance of populations found within Florida has potential to yield valuable information. This idea is further supported by research utilizing biological distance analysis on prehistoric Native Americans These studies indicate that biological continuity or change can be observed in the bioarchaeological record as morphological variation. As mentioned above, these data can contribute to long standing archaeological debates. Biological distance studies operate with the theoretical assumption that all populations within a region exchange mates from an outside source at an equal rate (Relethford and Blangero 1990) Therefore, when a population has an increased or decreased rate of exchange with an external population it will alter the expected linear relationship between the average within group variation and genetic distance to the centroid (the average heterozygosity of all subpopulations) (Steadman 1998) It is important to note that biodistance analysis conducted on archaeological sample pop ulations incorporates several assumptions that allow for analysis and interpretation of data These assumptions include : 1) patterns of biological distance and variability are an indirect measure of the degree of social exchange be tween the groups they rep resent; 2) gene flow and genetic drift affect allele frequencies between the populations under study and that those changes are reflected in skeletal changes ; 3) a cemetery or sample population is representative of th e original breeding population; 4) envi ronmental effects are minimal and randomly distributed among the populations (Droessler 1981; Steadman 1998; Stojanowski and Schillaci 2006) Several of these assumptions are based on living populations, however they may not all apply in archaeological se tting s For example, it has been found that some cemeteries were

PAGE 21

21 places where disparate groups came together during the Archaic (Randall and Sassaman 2010) and therefore may not be representative of the breeding populations. Therefore, this research will e xamine whether or not some of these assumptions apply This study utilizes craniometric data from 27 sites to explore the biological variation of pre Contact Floridians. A craniometric approach was adopted in conce rt with multivariate statistics, so that results could facilitate the interpretation of skulls and populations as a whole not just parts, as non metric traits or simple univariate indices are prone to do ( Corruccini 1974 ; Howells 1973; van Vark and Schaaf sma 1992 ) C raniometric data are explored using several multivariate analysis methods including Principle Component Analysis (PCA), Discriminant Function Analysis (DFA), and R matrix Analysis. Principle Component Analysis provides a method to explore the v ariation with in the data showing underlying patterns without making assumptions about the relationships of the individuals being analyz ed. Discriminant Function Analysis permits test ing of the variation associated with predetermined groups (i.e. sites, ti me periods, geographic regions culture labels ) to examine if observed variation corresponds with those predetermined groups or not L astly, R matrix Analysis is a tool that permits an examination of those predetermined groups to assess the level of variation associated with each group and its relationship to the variation of other groups thus providing biological distance s and gene flow estimates The o verall objectives of this study are threefold : 1) To examine the range of variation within the populations of Florida from the Archaic to pre Contact periods as represented in the populations used in this study; 2) To understand how the range of

PAGE 22

22 variation correlates to cultural, geographic or temporal variation; 3) To test the assumption s of biological distance analysis with regard to the populations in Florida. The null hypotheses related to these objectives are: 1) There is no biological variation noted within these populations; 2) There is no correlation between biological and cultura l variation or biological variation and geographic or temporal distance ; 3) The assumptions of biological distance analysis do not hold true in these archaeological samples. These objectives will be addressed using the craniometric methods noted above and details culled from the archaeological record. This research examine s the overall variation found within sampled study sites to determine if there are any general patterns to the biological variation I f patterns are found further analysis will discern t he nature of that variation, such as correlations with culture groups, geography, or time. Fo r example, research in the Midw est ern US has found that in certain instances biological variation correlates with cultural divisions or that biological variation w as impacted by culturally defined boundaries (Reichs 1984; Steadman 1998; 2001) I n other instances there was little variation between different cultures or that cultural differences were not a boundary to gene flow (Buikstra 1976; Droessler 1981; Sciulli and Schneider 1985) Therefore this research seeks to determine if there were any cultural boundaries that impacted gene flow within the sites sampled which encompass several archaeologically defined cultural divisions and if there are any changes to thes e patterns through time. In addition, variation is assessed in relation to geography to determine if there were any geographic barriers to gene flow. For instance, South Florida has long been regarded as separate and distinct from the rest of the peninsula both environmentally and

PAGE 23

23 culturally (Clausen et al. 1979; Widmer 2002) Researchers frequently point out the differences between North and South Florida through time to continue or reinforce the view that South Florida is its own entity (Carr 2012; Goggi n 1949; Widmer 2002) As mentioned above, Hrdli ka (1922) stated that the individuals found in the mounds of Florida along the West Coast and in North Florida were distinguishable from the individuals along the East Coast and in South Florida, from Lake Ok eechobee southward. Almost a century later Stojanowski (Stojanowski 2005 b ) also suggested a division between North and South Florida. Also, it has been noted archaeologically that there are differences between the cultures of the East Coast and West Coast -particularly in their pottery styles, burial practices, and cultural tr ajectories This research examines whether there is any pattern to the biological variation found between site samples that may correlates with previous interpretations or the apparent patterns as represented by differences in material culture. Within the archaeological narrative of Florida changes in culture or cultural variation ha ve been attributed to migration events, cultural boundaries or regionalization of cultures, or hybridization of cultures through time ( Ashley and White 2012; Hrdli ka 1922; Fa ught and Waggoner 2012, L uer and Almy 1982; Milanich 1994 ) The samples and sites accessed for this research allow for three instances of cultural change/variation to be focused on in detail: t he Archaic, the Weeden Island / Manasota culture Complex and Sou th Florida. During the Florida Archaic (10000 2 500 BP) the unusual and perhaps unique practice of water burials has been documented at several sites. This practice is limited to t he Early and Middle Archaic (10000 5 000 BP) and at geographically dispersed sites in the southern half of Florida (Doran 2002) In this practice the dead were interred in, or

PAGE 24

24 very near, a freshwater source, typically a pond or spring. Understanding how, and if, these populations were related to each other will augment our understa nd ing of Archaic migration patterns a nd generate data for comparison with later populations to develop new hypotheses about cultural continuity and change The We eden Island Complex (ca. 1800 1000 BP) encompassed portions of central and northern Florida as well as portions of southern Georgia and Alabama. Previous studies suggest that this c omplex first developed out of the Woodland Swift Creek culture in most areas; however it a ppears rather suddenly in other areas such as in North Central Florida at the McKeithan s ite (Milanich et al. 1997) The c omplex is marked by the appearance of regional variants that can be differentiated by secular pottery styles, habitation sites, and subsistence patterns (Jefferies 2004; Milanich 2002; Milanich 2004) All the Weeden Island sites however, share a similar sacred/secular pottery division, in which only certain pottery is used in ceremonial aspects, particularly burial rituals. Some of these sites exhibit a special burial pat tern that includes the use of charnal houses and east side pottery caches (Milanich 1994; Sears 1958; Sears 1973; Willey 1949) However, the Cades Pond and Manasota cultures are two notable exceptions li associated archaeological assemblages appear to include only some aspects of the Weeden Island Complex and these changes in practice appear very late in the culture period (Luer and Almy 1982; Milanich 2002) For example, these sites, whi ch occur primarily in west and west central Florida, do not exhibit the east side pottery caches, although they begin including the ornamental pottery associated with the burial rituals (Luer and Almy 1982; Milanich et al. 1997) Understanding how Weeden I sland sites

PAGE 25

25 and their respective populations relate biologically may add significantly to our understanding of how the Weeden Island culture Complex developed in various areas and how the population maintained the cultural complex across an environmentally diverse land area Further, these data will contribute towards an understanding of how the Weeden Island like cultures, such as Cades Pond and Manasota, may or may not The cultures of South Fl orida are often regarded as unique from the other cultures of Florida (Carr 2012; Widmer 2002) The archaeological record of this region has been variously divided into what are believed to be relatively autonomous geographic/culture areas (Carr 2012; Gogg in 1949; Griffin 1988) defined by pottery and settlement patterns, spanning a time period from 2500 BP through Contact (Widmer 2002) However, at the time of Contact it was said that the Calusa had control over the entire region (Carr 2012; Widmer 2002) suggesting a possible shift to a dominant cultural and biological influence. Therefore, understanding the biological relationships of the populations associated with the Glades and Caloosahatchee cultures through time will provide new information to resear chers interested in this area. Understanding how biology and culture vary in Florida will allow researchers to better address many question history In addition, this research has broader implications in studies of cultural affiliation particularly with regard to the Native American Graves Protection and Repatriation Act (NAGPRA). An understanding of how biology and culture have varied (or co varied) through time may provide additional lines of evidence for contemporary Native American s and archaeologists attempting to

PAGE 26

26 Following this Introduction, Chapter 2 reviews the archaeology of Florida with a particular focus on the archaeology surrounding the biolo gical questions asked in this research. In Chapter 3 the theory of biological distance analysis and the history of its use are discussed, particularly with regard to analyzing Native American populations. Chapter 4 documents methods and materials includin g the number of individuals and the relevant archaeological background of each site used. The craniometric variables and statistical methods utilized are also presented There is also a review of the sites analyzed. The results of these statistical analyse s are outlined and discussed in Chapter 5. Chapter 6 highlights the conclusions drawn from this research and offers recommendations for future investigations.

PAGE 27

27 Figure 1 1 Map of the location and cultural association of the sites used in this study.

PAGE 28

28 CHAPTER 2 CULTURE HISTORY AND ARCHAEOLOGY OF FLORIDA This chapter provides a review of Florida archaeology, which will serve as a framework against which the biological distanc e data acquired for this research may be interpreted The following culture history focuses primarily on material aspects of culture and associated burial practices that help to identify and distinguish sample sites included in this study C ultures for whi ch no skeletal remains were used for analysis are mentioned in brief. More detailed discussion of the archaeology of the Southeast and Florida can be found in the work of Anderson and Mainfort (2002b) Anderson and Sassaman ( 1 996) Milanich (1994; 1998) Willey (1949) and Ashley and White (2012) Table 2 1 shows an overview of culture periods by geographic area within the state. Paleoindian Period T he arrival of Native Americans in the Southeast United States and the Florida peninsula occurred approxim ately between 15,000 and 11,000 BP (Anderson 1996b; Anderson et al. 1996; Milanich 1994) The landscape they encountered was quite different than that of today; the most obvious distinction was a broader peninsula and a drier environment due to a sea level 70+ meters lower than present at the end of the Pleistocene (Anderson et al. 1996) T hese hunters and gatherers had a settlement and subsistence strategy that consisted of central or c ore habitation areas near fresh water with surrounding special use sites that encompassed a generalized foraging strategy (Anderson 1996a) M any of those core habitation sites were probably located close to the existing coastlines and are therefore now underwater (Faught 2004; Faught and Donoghue 1997) Nonetheless, a few Pale oindian sites discovered in karstic areas, such as the Aucilla River, Warm Mineral Springs and Little Salt Spring have been

PAGE 29

29 investigated (Clausen et al. 1975; Clausen et al. 1979; Dunbar and Waller 1983) T hese sites provide a small window into the subsist earliest populations, however, the lack of human remains securely dated to this time period preclude s biodistance analysis as part of this research. Archaic Period With the change in climate to a wetter environment and the rise in sea level at the start of the Holocen e (ca. 9000 BP) (Anderson et al. 1996) various new environmental niches were created in Florida and the Southeast. Archaic populations began to settle into these new areas, particularly alon g lakes and rivers, while other groups continued to inhabit the same areas as did Paleoindian group s before them (Milanich 1994) T he significance of freshwater is clear throughout the Archaic, as it is seen in the choice of sites in close proximity to a w ater source, the appearance of water burials ( a Florida phenomenon ) and the building of shell mounds situated in proximity to either fresh or salt water Relevant to this research are freshwater pond burials which have produced some of the earliest secur ely dated human skeletal remains from Florida. T he se burials are unique both for their mode of internment and their preservation (Doran 2002) One of these sites, Windover Pond has received extensive attention with respect to other Archaic sites in Florid a Archaeological evidence, in particular textile patterns associated with the burials, suggests that two groups utiliz ed this pond for burial; Stojanowski (in Stojanowski and Schillaci 2006) performed a biological distance analysis of the individuals buri ed in the pond and confirmed that there were at least two groups, which he refers to as bands, who utilize d the p ond for bur ial. This shared burial practice is suggest ive that the two groups shared a belief system An additional finding

PAGE 30

30 from Windover Pond is that most of the individuals were genetically determined to be members of Haplogroup X, which is not found in modern Native American groups of the Southeast (Smith et al. 2002) These results suggest that the Windover Pond population may not be related to later populations in the region and may represent an extinct population that once inhabited present day Florida. The unique mortuary practice of staked pond burials seems to have spanned ca. 5 000 years (Doran 2002) suggesting a continued shared ideolog y was maintained for generations H owever, relatively little is known about the relationship of the various sites that exhibit this practice in particular, the biological relationships of the individuals that buried their dead in this manner Archaic pop ulations are generally characterized as relatively mobile bands or groups, suggesting they interacted with other groups on a regular basis (Anderson 1996a; Milanich 1994) Anderson (1996b) propose s macroband and band level organization based on analysis of point typologies to explain the population structure and settlement patterns of the Southeast, in particular the South Atlantic Slope. F ocused on the Carolina and Georgia South Atlantic Slope area, Anderson defines macrobands by regional variations in art ifacts and density of artifacts around proposed regional centers (Anderson 1996b) He hypothesizes that the bands within the South Atlantic Slope area occupied river drainage systems and would have participated in seasonal movement within their area to max imize use of the resources available. He also proposed a gathering of bands in the Fall Line areas (at the transition from the upper hilly country to the flat low country) during particular seasons for resource or mate exchange and/or for information and o ther social interaction (Anderson 1996b)

PAGE 31

31 Anderson (1996b) also suggests that macroband s encompass ed most of Florida, which is characterized by numerous river drainage systems and water sources including sinks and springs. Thulman (2006) s Archaic point typologies and found similar band level organization during this period along river drainage s in Florida. The broad similarit y in artifacts and general subsistence patterns suggest that these bands share d a common origin and most likely had similar methods and techniques of tool production and resource procurement. But as each settled into a different drainage area, practices changed as new raw material for tools w as found, new methods of manufacture were introduced or invented, or novel hunting strategies were developed (Thulman 2006) This Thulman argues, would have resulted in subtle differences between tool kits, but continued interaction at the macroband level likely kept groups close socially and possibly biologically. At approxima tely 5000 BP, there appears to have been an influx of new people into the peninsula based on changes in burial patterns, a shift from water burials to land burials, and new tool technologies including the arrival of soap stone vessels in some areas (Sassam an 2006) There is also a change in trade networks from the early and middle to late Archaic time periods. Early and middle Archaic trade networks encompassed large areas with populations mov ing through the region seasonally resulting in more direct socia l interactions between groups. However, by the late Archaic there appears to have been a marked shift with more sedentary groups and less direct social interaction and more communication through trade partnerships (Waggoner, 2009).

PAGE 32

32 In the archaeological record, the terminal or late Archaic is marked by the manufacture of pottery throughout the Southeast and in the peninsula of Florida (Milanich 1994) As a result, archaeologists are confident that they can identify geographically distin ct cultures based o n ceramic types and styles recovered from sites (Cordell 2004; Jefferies 2004; Milanich 1994) Woodland Period The Woodland period began ca. 2500 BP and is marked by a relatively stable environment with sea levels near modern levels and a warm humid clima te similar to the present day (Schuldenrein 1996; Watts et al. 1996) Interestingly, there is an apparent decrease in the occurrence of long distance exotic trade goods and a n increase in local or s em i local materials and goods, which occurs coevally with a greater regionalization of pottery styles and techniques (Anderson and Mainfort 2002a) Archaeologists have long held that regionalized pottery can be used as a marker for distinct archaeological cultures (Anderson and Mainfort 2002a; Willey 1949) Ofte n similarities in the pottery styles or assemblages are used to link distant groups to each other, while other times geography is used to divide somewhat similar assemblages into distinct groups (Anderson and Mainfort 2002a) The Florida Woodland period i s marked by the appearance of several distinct cultures, including Deptford in the north ern half of the peninsula and Manasota or early Glades in the southern half. These cultures share similar settlement and subsistence patterns, a coastal orientation, an d a band level society with small communities that may have traveled inland seasonally to harvest non coastal resources (Milanich 1994; Stephenson et al. 2002) The observed regionalization of pottery is associated with a rise in ceremonialism, particularl y evident in the development of the Weeden Island

PAGE 33

33 Complex Anderson and Mainfort (2002a) suggest that this may have been due to a cultural power shift from the Midwest to the Southeast, particularly in the Lower Mississippi and Gulf Coast areas, where the Coles Creek and Weeden Island cultures are present This has led archaeologists to speculate that there was interaction between the areas (Milanich 2002; Stephenson et al. 2002) but whether it was an exchange of ideology or a movement of people or both is not clearly understood. The Weeden Island Complex spanned most of the central and n orthern portions of Florida, as well as portions of southern Georgia and Alabama This broad geographic range suggests there was an extensive interaction sphere involving the exchange of objects and ideology (Anderson and Mainfort 2002a). The cultures associated with the Weeden Island Complex are marked by a distinct sacred/secular dichotomy found in the pottery associated with the village and burial mound sites (Milanich 2 002; Cordell in Milanich et al. 1997) The cultures associated with the Weeden Island Complex are also marked by the construction of burial mounds that provide evidence of elaborate ceremonial practices associated with single or multiple burial events. The se burial events involved the removal of numerous individuals from charnel houses or similar storage / processing areas and placing them on prepared surfaces for mass burial with sacred pottery or other objects of importance (Milanich 1994; 2002; Willey 1949 ) Some sites apparently developed into significant ceremonial centers with one or more large mounds indicating elaborate and repeated burial events. These sites appear to correspond with trade hubs making them key in a widespread ceremonial and exchange n etwork (Anderson and Mainfort 2002a) Mounds, such as McKeithan, a large mound site in North Central Florida, were most likely constructed in or near the village of a

PAGE 34

34 lineage that was in power. As resources in an area were exhausted or there were shifts in trade patterns, power transferred to a new lineage and mound construction began in a new area (Milanich 1994; 2002) Another significant feature of Weeden Island Complex is that despite apparently sharing various aspects of sociopolitical structure and r eligion, particularly the sacred/secular dichotomy, each culture developed distinct environmental adaptations. These adaptations are found within specific physiographic parameters, which then help to define the regional complexes based on differences in apparent settlement and economic patterns ( Milanich 2002) The thorough command of the local environment by these peoples suggests sedentism or at least a localized mobility allowing for a specialized tool kit and subsistence strategy. In the L ate Woodlan d of Florida, there is a shift to a hierarchical social system, such as that seen in the Wakulla and Suwannee culture and in some areas this corresponds to an increased reliance on maize agriculture (Anderson and Mainfort 2002a) At the end of the Woodlan d period much of the Southeast transitions into the Mississippian period, with its ranked society and associated iconography, as seen in the Fort Wa lton and Safety Harbor cultures. H owever, many of the cultures in the peninsula continue in their late Woodl and form through Contact A more detailed discussion of the Woodland period in Florida follows and is presented by geographical region in order to follow the cultural developments and transitions of each culture. North Florida The Deptford culture, which appears in the Early Woodland, is found in the southern Southeast and is divided into three regional variants : Gulf Coast, Atlantic

PAGE 35

35 Coast, and Coastal Plain Interior Riverine (Milanich 1994; Stephenson et al. 2002) Two of the three variations appea r in Florida, the Gulf Coa st and Atlantic Coast varieties. Al though they share similar subsistence and settlement patterns, these cultures differ i n subsequent sequences allowing them to be considered separate taxonomically (Stephenson et al. 2002) The G ulf Coast Deptford culture is marked by primarily coastal habitation sites with interior special use sites (Jefferies 2004; Milanich 1994) The pottery is characterized by simple stamp, linear check stamp, or check stamp pottery with a sand or grit temper (Milanich 1994; Willey 1949) In the northern Gulf Coast region late Gulf Coast Deptford develops into the Yent and Green Point complexes (Milanich 1994) which exhibit an early form of the sacred/secular dichotomy that is evident in the later Weeden Islan d Complex (Stephenson et al. 2002) The Crystal River site complex, which is one of the most studied Yent site complexes appears to begin its use in the Early Woodland (Milanich 1994; Stephenson et al. 2002) Crystal River eventually becomes a major cerem onial center with in the Weeden Island Complex and may have been use d into the Safety Harbor period, although this late use/habitation i s still speculative (Pluckhahn et al. 2010) North and Northwest In North Florida and the Panhandle, Deptford transitions into Swift Creek and Santa Rosa Swift Creek respectively (Milanich 1994; Stephenson et al. 2002) The noted differences between Swift Creek and Santa Rosa Swift Creek are differing settlement patterns and variations on pottery design (Stephens on et al. 2002) Swift Creek sites are found primarily adjacent to the salt marsh tidal stream systems, much

PAGE 36

36 like the Deptford sites, and occasionally habitation sites are found in the hardwood hammocks away from marshes (Milanich 1994) Santa Rosa Swift C reek sites however, are found along the coast, with some special use interior sites (Milanich 1994; Willey 1949) Swift Creek pottery is marked by complicated stamp patterns and a continued use of sand and grit tempers, like those found in Deptford pottery (Jefferies 2004) Santa Rosa Swift Creek also has ceramic variants very similar to Marksville pottery, a culture linked to the later Cole Creek Complex of the Lower Mississippi area (Stephenson et al. 2002) These pottery styles include Alligator Bayou St amped, Basin Bayou Stamped, Marksville Incised and Marksville Stamped pottery (Milanich 1994; Willey 1949) The Swift Creek culture also appears to have participated in a periodic aggregation of peoples for ritual purposes, as evidenced by various burial a nd ceremonial mounds (Milanich 1994; Willey 1949) Some of these mounds also contain Hopew ellian grave goods, often found in cache s with in the mound (Milanich 1994) suggesting long range exchange and association with this important interaction sphere (Bul len 1951; Greenman 1938) Evidence found at some mounds suggest elaborate mortuary customs and a cult of the dead (Willey 1949) most likely the beginnings of the later Weeden Island C omplex. In Northwest Florida, Swift Creek and Santa Rosa Swift Creek tr ansitions into Weeden Island around ca. 1700 BP (Jefferies 2004; Milanich 2002) and is characterized by an overlap in pottery types; the Swift Creek and Santa Rosa Swift Creek pottery is ultimately replaced by Weeden Island ceramics, possibly suggesting a continuity of people with a changing ideology. Again, there are noted similarities between the pottery

PAGE 37

37 of the Florida Weeden Island and the Mississippi Coles Creek cultures (Milanich 1994) suggesting continued interaction between these two areas. In the L ate Woodland period, the Swift Creek / early Weeden Island culture Wakulla culture is marked by the appearance of maize agriculture. There is a shift to smaller settlements, possibly small family groups, which appear to move around the area as the soil is exhausted from agriculture (Milanich 2002) The Wakulla culture continues its agricultural lifestyle through the late Woodland and transitions into the Fort Walton culture the only truly Mississippian culture in Florida (Milanich 2004) Though the origins of the Suwannee Valley culture differ slightly between east (Swift Creek) and west (Weeden Island McKeithan) North Florida, this culture shares many similar ities with the Wakulla culture to the west including smaller hamlet sized sites. However, p ottery styles differ and there is currently no evidence for maize agriculture by the Suwannee Valley groups (Milanich 1994; 2002; 2004) Variants of the Suwannee Valley culture co ntinue into the period of European contact. North Central In North C entral Florida there is little discussion of the Early Woodland period because there are only a few known Deptford sites and these are small There is also a conspicuous absence of Swift Creek sites, so it appears that the sudden appearance of the Weeden Island McKeithan culture is not an evolution from Swift Creek, as many of the other e arly Weeden Island cultures are but is the result of either rapid population growth in the Deptford Mc Keithan transition or a possible movement of Weeden Island people into the area (Milanich 2002) This leads to the question: Is the rise of Weeden Island in north central Florida due to new people coming into the area or is there overall

PAGE 38

38 continuity between cultures with an increase in population and the adoption of a new ideology? In southern, north central Florida, the Cades Pond culture is established around 1800 BP, coinciding with an apparent population explosion based on the rapid rise in the number o f sites in the area (Milanich 2002) The Cades Pond culture may have evolved from the preceding Deptford groups in the area. Cades Pond is labeled Weeden Island like due to the presence of Weeden Island pottery in late Cades Pond burial mounds; although such ceramics do not seem to be found at earlier sites (Milanich 2002) The late introduction of Weeden Island pottery may suggest new influences in the culture. Also, according to Milanich (1994) the Cades Pond culture is a unique adaptation of the Weede n Island Complex to an extensive freshwater, wetland environment. Around 1400 BP, archaeologists describe a significant change in the settlement and assemblage patterns of the population inhabiting the north central Florida area (Austin 2001; Rolland 2012) This is seen as a transition into the Alachua culture. Similarities in pottery type and settlement pattern suggest that the Alachua culture was related to the agricultural peoples from the south central Georgia area (Milanich 1994) Some have argued that there was a migration of peoples into the area bringing the Alachua culture, while others have argued based on radiocarbon dates that the north central Florida region s erved as the homeland of the culture change and the people subsequently migrated to the Ocmulgee region of Georgia (Milanich 2002; 2004; Schofield 2003) In either instance there are some clear distinctions between the Cades Pond and Alachua cultures, such as Cades Pond sites are near wetlands and the archaeological evidence suggests a heavy reliance on aquatic resources while the

PAGE 39

39 Alachua sites are more often distant from wetlands on high ridges and their subsistence pattern appears to rely heavily on deer and agricultural crops (Austin 2001; Milanich 1994) Additionally, the Cades Pond tradition used primarily undecorated wares in villages, but included Deptford pottery in burial mounds with a later transition to Weeden Island ceramic vessels in burial sites. However, the Alachua tradition is marked by cord and fabric marked pottery with a subsequent transition to corn cob marked vessels (Milanich 1994; Rolland 2012) Mortuary practices include pre burial mound preparation, east side burial placement, primary burials, and a lack of burial goods; the last bei ng a significant change from the former Cades Pond culture (Rolland 2012) Analysis of the biological variation associated with the Alachua culture and the earlier cultures in the area could provide a better understand of the biological aspects of this tra nsition Northeast In the early Woodland, the Atlantic coast variant of Deptford is documented north (Stephenson et al. 2002) and the Orange culture is found southeast of the river (Saunders 2004) Russo (1996) and Milanich (1994) suggest continuity between pre ceramic and the later ceramic cultures through Contact along the East Coast of Florida based on site location, tool technology and pottery productio n, supporting the view that the culture sequence in this area was an independ ent development from the Archaic Mount Taylor culture into the Orange culture and then the St. Johns culture (Milanich 1994) The St. Johns tradition ultimately encompassed most of the east coast of Florida. St. Johns pottery is marked by a distinct chalky texture (Milanich 1994) the result of the sponge speculates that are incorporated into the paste. The appearance of Deptford, Swift Creek, Weeden Island and Hopewellian items is rare in this area, however when fo und, testing has shown that the these item s are t he result

PAGE 40

40 of mimicry on local clay and not the product of trade with distant group s (Milanich 1994; 2004) The late St. Johns culture built large ceremonial centers in the later Woodland period. These ceremonial centers suggest a development of poli tical complexity similar to that see n in the Fort Walton and Safety Harbor cultures Southern Peninsular Gulf Coast and South Florida The coastal Weeden Island cultures identified southeast of the Aucilla River along the Gulf Coast are difficult to define as specific cultural entities. The shoreline is extensive with varied microenvironments along the coastline from north to south, including the salt marshes, tidal streams, inshore bays of the Gulf of Mexico, and nearby freshwater and forested environments Each habitat provides a unique ecosystem allowing for the potential development of distinct cultures with each having localized subsistence strategies and tool kits (Milanich 2002) Overall, a rchaeological work in this area has been limited; h owever, the well studied Crystal Riv er site complex appears to continue to be an active cultural center during the Weeden Island period of the Gulf Coast H ow much of the surrounding area was involv ed in the ceremonial activities that took place at Crystal River rema ins poorly understood (Pluckhahn et al. 2010) Southern Peninsular Gulf Coast One area with a fairly well defined culture along the west coast of Florida is the Manasota culture centered south of Tampa Bay, generally in the Manatee and Sarasota county are a (Luer and Almy 1982; Milanich 2002) Manasota, much like Cades Pond, appears to be a continuation of the local Archaic culture, but with the addition of Weeden Island like pottery types late in the culture period (Luer and Almy 1982; Milanich 2002) resu lting in it sometimes being labeled Weeden Island like. The

PAGE 41

41 Manasota culture dates from ca. 2500 1200 BP; however, it is not a static entity and Luer and Almy (1982) note changes in material culture through time, which may hold keys to the culture chronolo gy and external influences. Overall, the Manasota culture is marked by a settlement pattern of elevated, linear sites parallel to the shoreline, with some sites located inland from the shore; the latter typically located on elevated ground, relative to the surrounding landscape. The primary subsistence strategy focused on fishing and shellfish gathering, supplemented with inland hunting and gathering of seasonally available plants and animals. The dominant pottery was undecorated, sand tempered; thick walle d vessels transitioned to thinner walled open bowls through time. The same pottery is used in both village and burial mound sites, but in late Manasota there is an increase in Weeden Island like decorated pottery in the burial mounds (Luer and Almy 1982) Interestingly, burials occur in a variety of contexts including cemeteries, mounds, and middens (Luer 2011) however the common pattern appears to be primary, flexed burials (2500 1600 BP) later accompanied by the inclusion of thick sand tempered plain sh erds, followed by a later transition to the inclusion of an occasional Weeden Island vessel in the burial mound. Also, toward the end of the Manasota Period (1400 1200 BP) secondary bundle burials became the preferred mode of interment, and thinner, sand t empered plain pottery occurs with an increased use of Weeden Island vessels, most of which appear to be imported based on paste analysis (Luer and Almy 1982) Of note is the fact that the pottery is found scattered throughout the mound, apparently broken j ust prior to burial, and it is not found in east side caches like in other site associated with the Weeden Island Complex (Luer and Almy 1982)

PAGE 42

42 The end of the Weeden Island / Manasota culture in west central Florida is partially marked by the absence of Wee den Island pottery in burial mounds, and the introduction of Safety Harbor pottery variants. However, the people continued to live on shell middens near burial mounds and fish the local waters (Luer and Almy 1982) After 1000 BP settlement patterns seem to shift to more nucleated villages with corresponding larger populations, first in the Tampa Bay area and then in the surround Gulf Coast areas. This marks the transition to the simple chiefdoms of the Safety Harbor culture (Milanich 2002) The Manasota cul ture offers a unique opportunity for biodistance analysis to examine an apparently longstanding localized culture with transitions that may have been the result of contributions from outside influences South Florida South Florida is often discussed as a standalone entity both environmentally and culturally. It shares the appearance of Woodland ceramics and village life with the rest of the Southeast, but the archaeological record reveals radically different cultural patterns (Widmer 2002) Pottery analysis suggests the region was relatively culturally isolated from the rest of the peninsula (Marquardt and Walker 2012) However, based on ceramic traditions, archaeologists have identified at least three regional divisions of the Woodland per iod in South Florida: Glades, Okeechobee, and Caloosahatchee (Milanich 1994; Widmer 2002) Population size and political complexity appear to decrease from west to east across South Florida (Widmer 2002) perhaps supporting the varied cultural groups defin ed by pottery types. Throughout South Florida however, there is archaeological evidence of Hopewellian interaction at some sites (Widmer 2002) and through time there is an apparent rise in outside contact and/or trade as evidenced by an increase in Missis sippian influence d artifacts, sand burial mound

PAGE 43

43 construction and changes in site plans or midden/mound construction (Carr 2012; Marquardt and Walker 2012) The Glades cultures, particularly in the Ten Thousand Islands region, are marked by large shell mou nds that appear to be intentionally constructed rather than the result of refuse piling. Evidence indicates these structures were most likely used for ceremonial purposes and / or for elite residences (Widmer 2002) suggesting a well developed ceremonial and political ideology. On the Southeast coast, along the Atlantic Ocean, sites are locates on hammocks and ridges, but there is little evidence of shell mound construction. The archaeological record also suggests that many of these sites may have been used s easonally (Widmer 2002) Glades pottery is marked by a dominance of incised and punctated patterns that are distinct from other patterns throughout the peninsula (Carr 2012) This apparently self contained Glades culture appears to accept more outside infl uences, as seen in changes in pottery assemblages, during the Mississippian period when there are increased influences from the St. Johns region and the Calusa culture on the Southwest Coast (Carr 2012) The Okeechobee region is marked by large sites with characteristic linear mounds and circular ditches (Widmer 2002) The pottery assemblages are characterized by distinct plain wares (Carr 2012) Fort Center (8GL13), one of the largest recorded sites, with evidence of a charnel house, a charnel pond, and m ound structures suggest s elaborate ceremonial use (Sears 1982) Many of the mounds at Fort Center were constructed of black dirt, as opposed to shell. These mounds may have been used for ceremonial purposes, but most were probably associated with house st ructures as the area is prone to flooding and there is little natural high ground

PAGE 44

44 (Carr 2012; Widmer 2002) To the east, Carr (2012) has identified the East Okeechobee Region, which he believes is distinct from the Okeechobee and Glades regions, primarily based on the fact that there is a distinct lack of local pottery; assemblages consistently contain pottery from the St. Johns and Indian River Regions to the north, suggesting a continuity of contact along the Atlantic Coast from North to South (Carr 2012) The settlement pattern is hard to define due to modern impacts on sites; however, it appears this region was dominated by coastal settlements, particularly where rivers drained onto the coast (Carr 2012; Milanich 1994) The Caloosahatchee region is locate d along the Southwest coast of Florida and is marked by coastal settlements with large midden/mound and waterworks sites (Marquardt and Walker 2012; Widmer 2002) Caloosahatchee pottery is described as relatively conservative, plain, incurved bowls made wi th local clay (Marquardt and Walker 2012) The Caloosahatchee culture is most often associated with the later Calusa culture, which was in power at the time of European contact. Around 1200 BP, the Calusa chiefdom developed a more pronounced presence thro ughout South Florida as evidenced by changes in pottery in the Caloosahatchee region. Through time there is a rise in Belle Glade pottery, originating in the Okeechobee region, followed by an increase in St. Johns check stamped pottery which has its origin s in Northeast Florida; then a shift to Glades pottery from the Southeast coast or Ten Thousand Islands area of Florida (Marquardt and Walker 2012) These shifting pottery styles suggest shifting trade and alliance patterns and the increased large scale co nstruction projects paired with changing environmental stability provide the need for these shifting alliances (Marquardt and Walker 2012) By the time of

PAGE 45

45 Contact the Calusa were a large scale tributary patronage clientage system encompassing nearly all of South Florida. Historical accounts detail that Calusa chiefs would often take wives from other villages and the wives would bring their pottery making skills with them (Marquardt and Walker 2012; Widmer 2002) So the archaeological and historical evidence suggest s there was a rise through time of political and social interaction between the South Florida cultures, which most likely included intermarriage, but there was also an apparent rise in outside contact as w ell. E xamination of the biological variation in South Florida will allow for the analysis of the effects of outside influences, as suggested by imported Hopewellian and Mississippian artifacts, and the effect of the increas ed influence of the Calusa throug hout the region. Mississippian Period The Mississippian Period, which begins in Florida ca. 1000 BP, witnessed little overall environmental change; however, evidence suggests that there were periodic fluctuations in sea level that could have had a signifi cant impact on coastal cultures (Marquardt and Walker 2012) Many of the cultures discussed in the Woodland period continued without significant changes in structure until European Contact. However, there are two cultural changes that should be noted, prim arily because they are the only cultures that are labeled Mississippian in Florida. These include the Fort Walton culture in the Florida Panhandle and the Safety Harbor culture in the West Central Gulf Coast The Fort Walton culture, which is documented in the Panhandle between the Aucilla River and Mobile Bay, developed from the preceding Weeden Island Wakulla culture of the area, as evidenced by a continuity of pottery types. The Mississippian influences, tho ugh, are seen in Mississippian iconography, the hierarchical system of site types, and an increased reliance on agriculture, as found in other Mississippian

PAGE 46

46 groups throughout the Southeast. There are regional variants of the culture that are differentiated by pottery types and differences in settlement and burial pattern, which evidence suggests correspond to regional chiefdoms (Milanich 1994) Safety Harbor (ca. 1100 300 BP) appears to develop from the Weeden Island and Manasota cultures found along the c entral peninsular Gulf Coast, from north of Tampa Bay to Charlotte Harbor. Just as with Fort Walton there were regional variants of this culture based on variations in subsistence and settlement patterns; however, there is a common ideology expressed in th e burial ceremonialism, which includes intermittent use burial mounds containing secondary burials, and the frequent reuse of Weeden Island burial mounds (Mitchem 1989) Of note is the absence of grave goods often found in other Mississippian burial mounds (Luer and Almy 1982) Many aspects of the Safety Harbor culture suggest a continuum from the Weeden Island and Manasota cultures, including the overlap in site locations and Weeden Island type incised and punctated wares found in mortuary contexts and the plain wares found in village sites, though the sacred / secular dichotomy is not as pronounced (Luer and Almy 1982; Milanich 1994; Mitchem 1989; 2012) However, other features, like the platform mound village complexes, the iconography found on vessels, and changes in vessel forms are traced to Mississippian influence and are quite similar to features found in the Fort Walton culture (Luer and Almy 1982; Milanich 1994; 2004; Mitchem 1989; Willey 1949) Summary In summary, the focus areas for the biological research are the Florida Archaic, the Weeden Island / Manasota Complex, and South Florida cultures. Few sites from the Archaic period have produced adequate human remains, but the available populations

PAGE 47

47 are suitable for biodistance analysis, which will help d evelop a better understanding of (2006) river drainage band divisions, it can be hypothesized that bands came together at various times of the year for burial ceremonies at specific water sites. Similar to the activities discussed by Sassaman (2008) at the Tick Island / Harris Creek (8VO24) site and suggested by the studies at Windover Pond ( Stojanowski and Schillaci 2006 ) Despite the extensive studies conducted on the Windover population, little research has focused on the biological relationships of other Archaic sites. Therefore, this research examines the biological variation within the following sites: Warm Mineral Spri ngs (8SO19), Windover (8BR246), Bay West (8CR200), Little Salt Spring (8SO18), Republic Groves (8HR4), Gauthier (8BR193), and Bird Island (8DI62) to find evidence of biological and cultural relationships. Homogeneity would suggest an ancestor descendent re lationship between the sites or a single, continued biological population inhabiting the peninsula. Conversely, heterogeneity would suggest t he introduction of new populations to the peninsula or discrete divisions of peoples with little or no gene flow be tween them. Fortunately a number of Woodland period sites have yielded human remains, allowing for the examination of the relationship between social structure and cultural interactions via the biolog ical variation observed within and between samples. Si tes such as Browne Site 5 (8DU62), Palmer Mound (8SO2A), Hughes Island Mound (8DI45), Bay Pines (8PI64), Bayshore Homes (8PI41), McKeithan (8CO17), Venice Beach Complex (8SO26), Yellow Bluffs (8SO4), Crystal River (8CI1), and Manasota Key Cemetery (8SO1292 ) are analyzed to address questions about the Deptford Weeden Island/Manasota Complex. Hutchinson Island (8MT37), Highland Beach

PAGE 48

48 (8PB11), Margate Blount Mound (8BD41), and Captiva Island (8LL57) are examined to investigate the variation of South Florida, a nd in particular the Glades and Caloosahatchee cultures. Henderson Mound (8AL463) will be used to test questions about the relationship of the Alachua culture to the Weeden Island Complex Examination of each of these sites to determine how these sites rel ated to each other biologically could provide new details about the social structure of these populations. Homogeneity would suggest a shared relationship between these sites, while heterogeneity between sites may suggest underlying social divisions that s tructured the populations or migration events that resulted in cultural changes Only three sites associated with the Mississippian period were sampled for biodistance analysis: Fort Walton Temple Mound (8OK6), Safety Harbor Mound (8PI2) and Sarasota Bay Mound (8SO44) The samples from these sites are examined for evidence of biological continuity or variation in relation to the transition from Weeden Island to Fort Walton and Weeden Island/Manasota to Safety Harbor with the purpose of providing insight in to whether these transitions were the result of an influx of people, increasing heterogeneity, or due to an exchange of ideology with very little biological interaction with the rest of the Mississippian world. Also the relationship of these sites compared to each will be tested to examine the possibly interactions between these two cultures. The skeletal samples from these sites also allow for the analysis of biological variability indicative of North South or East West divisions within the peninsula. The North South division is particularly interesting, as it is often mentioned by researchers and appears to be supported archaeologically;

PAGE 49

49 to be in question. For example, the University of South Florida is located in Tam pa, while Stojanowski (2005) location north of Tampa. McGoun (1993) demarcates South Florida as anything below an invisible line between Cocoa Beach and Bradenton, while Widmer (2002) defines it as an ything below 27 Latitude or an imaginary line between Jupiter and Venice. Therefore, the questions that can be addressed in this research are: is there a biological (1922) and Stojanowski (2005b) ha ve suggested, and if so where and when does that distinction appear in the Florida peninsula? In addition, (1922) skeletal analysis appears to indicate that there is biological variation between the East and West Coasts of Florida and Carr (2012 ) has stated that archaeological evidence supports a distinction between the East and West coasts of Florida. This research will also try to examine this supposition. To reiterate, this review of the archaeology of Florida serve s as a framework against wh ich the biodistance data can be examined and has identified several areas in which biological distance analysis may prove a useful tool in introducing additional lines of evidence and inquiry concerning continuity, relatedness, migration, integration, repl acement, or mate exchange The following chapter details the theory and methods of biological distance analysis.

PAGE 50

50 Table 2 1 General Sch eme of Florida Culture Periods. Northwest North North Central Northeast North peninsula Gulf Coast Central peninsula Gulf Coast Glades Okeechobee Basin East Okeechobee Caloosahatchee 700 BP Caloosahatchee V Caloosahatchee IV 1000 BP Fort Walton Safety Harbor Glades III Caloosahatchee III Belle Glade IV Caloosahatchee IIB Wakulla Weeden Island Alachua St. Johns II Glades II Caloosahatchee IIA 1500 BP Suwannee Valley Cades Pond Weeden Island Belle Glade III Weeden Island Weeden Island Weeden Island McKeithan Weeden Island Santa Rosa Swift Creek Swift Creek Belle Glade II 1900 BP Yent Complex Caloosahatchee I Deptford Manasota Glades I Belle Glade I 2500 BP Deptford Deptford Deptford St. Johns I Transitional Transitional Transitional Transitional Transitional Transitional Archaic Terminal Archaic 9000 BP Archaic Archaic Archaic Archaic Archaic Archaic Archaic Archaic 12000 BP Paleoindian Paleoindian Paleoindian Paleoindian Paleoindian Paleoindian Paleoindian B ased on Milanich and Fairbanks (1980) ; Marquardt and Walker (2012) ; Luer and Hughes (2011) ; Callsen (2008) ; Griffin ( 1988)

PAGE 51

51 CHAPTER 3 BIOLOGICAL DISTANCE ANALYSIS : HISTORY, THEORY, AND METHODS This dissertation reevaluates human skeletal remains and archa eological evidence from select Florida archaeological sites to provide new insight into population migration into the peninsula and pre Contact culture change. There are various methods to analyze biological variation and biological distance. Some researchers have utilized ancient DNA to evaluate questions of biological relationships and biological distance, however, this particular approach is not practical in Florida where human remains suitable for DNA analysis can be limited Therefore other methods to a ssess biological variation and biological distance such as the statistical analysis of metric data are necessary. It has been shown that metric and nonmetric data provide biological distance estimates that broadly correspond to genetic distances and known population histories. Biological distance is the statistical representation of morphological variation, which is an expression of genetically controlled traits (Griffin et al. 2001) These methods have been used successfully to address various questions ab out the population history of both living and archaeological samples; however, as mentioned previously, very little work has been focused on populations in the Southeastern United States and particularly in Florida. Therefore this chapter will review the t heory behind biological distance analysis, beginning with research on living populations with known biological histories, moving to archaeological studies that have provided new insight regarding archaeological events and, lastly, reviewing the assumptions used in biodistance studies of archaeological populations and the mathematical models involved.

PAGE 52

52 B iological Distance Studies on Living Populations Previous studies of living populations have shown that metric and nonmetric data provide biological distance estimates that broadly correspond to genetic distances and known population histories. Spielman (1973), for example, examined nineteen Yanomama villages in South America and found a correspondence between the allele frequencies and anthropometric data col lected. The genetic relationships that were found between villages could be explained by the known historical and geographic relationships of the villages. In another study, Friedlaender and colleagues (1971) analyzed several populations in South Central B ougainville in New Guinea. Their analysis of blood polymorphism gene frequencies and anthropometric measurements produced very similar evolutionary tree structures, and the relationships evidenced in those trees were explained by linguistic, geographic, an d migrational distances between populations. Additional studies on living populations of varying sizes have shown that morphological data correspond with genetic data and the results can be explained through a number of factors, including geography, lingui stics, migration events, or social divisions (Basu et al. 1976; Pollitzer 1958; Sanghvi 1953) Relethford (1994) examined evolutionary relationships and population migration by comparing metric variation to genetic variation within the major geographical g (1994) results also showed that the genetic and craniometric data were in agreement, both qualitatively and quantitatively These data further support the utility of craniometrics as a proxy for genetic relationships. These studies confirm that genetic and metric analysis can be used to effectively differentiate between populations Additionally, the studies show that metric analysis yields distance patterns that are consistent with genetic distances and correspond to

PAGE 53

53 p atterns predicated by language relationships, migration, and geographic distances in living populations. Therefore, it can be assumed that the statistical analysis of metric data can produce similar results in archaeological populations that may be interpr eted effectively based on cultural/archaeological differences and geographic and temporal distances. However, confirming biological distance results in archaeological populations is considerably more difficult This is because there is rarely any genetic data to compare to the results and population histories are often unknown. However, as will be shown below, there are still many questions that biological distance analysis can address with regard to the biological and cultural relationships of past popula tions. B iological Distance Studie s on Archaeological Populations Previous research utilizing biological distance analysis on prehistoric Native Americans has examined broad issues like migration into the Americas or more narrow topics such as regional culture changes, interaction spheres, the impact of Spanish contact, language divisions, and site structure. The timing and route of migration into the Americas has been a long debated topic. Researchers have utilized archaeological, genetic, and morphological evidence to support a wide variety of theories. Despite differi ng interpretations of the results, these studies show the utility of craniometric data in the analysis of population history. For example, Powell and Neves (1999) examined Paleoindian, Archaic, and modern Native American populations from across the United States asking questions about genetic variation and migration into the Americas. Their study showed that there is phenotypic variation, within and between the Native American populations is most likely the result of genetic drift, selection and gene flow a cting on a population that resulted from a

PAGE 54

54 single migration event. In contrast, Chatters (2010) examined a collection of samples from the Northwestern United States and Canada and found that morphological variation found within the sample populations was t he result of multiple migrations out of Beringia from a stock population and that genetic drift and gene flow did not have as (2001) analysis of remains associated with several Mid west tribes suggested that there were multiple migrations by different populations into the America s resulting in the biological diversity found among the populations studied On a smaller scale, there have been several studies utilizing biological distance to anal yze specific or localized archaeological questions about culture change and continuity. For example, archaeologists have long argued two theories for the development of the Hopewell culture in the Illinois Valley; the first is that the people associated wi th the Hopewell culture in the Illinois Valley were the result of a migration event, while others argued that the Hopewell culture developed in situ as the result of a change in ideology without any major migration event or minimal genetic influx. Buikstra (1976) examined non metric features in human skeletal remains recovered from four sites dated between 2050 suggest that the Middle and Late Woodland populations were relatively stable and had occupi ed the area for a significant amount of time, with no evidence of a migration into the area. Further, Buikstra was able to demonstrate that sites exhibited differentiation or isolation by distance, meaning those sites closer together were more related than sites that were more distant from each other. Additional work was conducted by Droessler ( 1981) utilizing craniometric data to examine Late Woodland and Mississippian groups.

PAGE 55

55 iddle to Late Woodland periods and demonstrated biological continuity into the Mississippian period. Biodistance analysis suggested that the populations were moderately homogenous, but there were differences based on geographic distance between the sites a nd that these differences varied based on sex, suggesting variations in post marital residence patterns. Steadman (1998) looked at the Mississippian transition in the Central Illinois Valley, focusing specifically on the amount of external gene flow evide nt in the populations, exploring the theory that a movement of peoples from the American Bottom, the region around Cahokia in southern Illinois along the Mississippi River, brought the Hopewell traditions with them. Questions about the origin of the later Oneota culture and its relation to the Mississippian groups were also addressed in this study. Steadman confirmed the earlier results of biological continuity for Woodland to Mississippian. Further, using model bound methods, Steadman (1998) was able to sh ow that there was very little external gene flow into the populations, further supporting an in situ ideology shift to Mississippian culture. The study also revealed that the Oneota population was unrelated to the Mississippian populations of the Central I llinois Valley, confirming that this new culture was the result of a population migration into the area Similar work was conducted by Sciulli et al. (1984) analyzing the populations associated with the transition from the Archaic Glacial Kame culture to the Early Woodland Adena culture. Sciulli et al. (1984) examined non metric traits and found that the two populations were not biologically homogenous. However, they were still quite

PAGE 56

56 closely related and this could either be explained by an ancestor descend ant relationship between the two populations or, if the samples were temporally close, a shared recent ancestor (Sciulli et al. 1984) (1985) analysis of the Late Archaic populations in Ohio examined three closely related, but sligh tly different burial complexes, the Red Ocher, Western Basin Late Archaic, and Glacial Kame complexes. Their results found that all of the samples were relatively homogenous, or similar, and that sites that were closer together geographicall y were more clo sely related than more distant sites were (Sciulli and Schneider 1985) Since no significant differences were found between the burial complexes it suggests that the apparent archaeological/cultural differences did not appear to affect the biological inter actions of the sample populations studied. Tatarek et al. (2000) compared these Archaic samples to later more sedentary agriculturalist populations in Ohio. They found that the late Prehistoric populations were rather heterogeneous and that this heterogene ity did not appear to be correlated with geographic distances or variations in material culture; the authors suggest that there was marked population instability and that various other cultural historical factors may have contributed to the biological vari ation found within the sample populations The examination of interaction spheres can also be addressed with the analysis of biological variation. In these spheres, much like that found in Florida in the Weeden Island Complex, the archaeological evidence often suggests that populations were interacting as evidenced by the exchange of goods and it is usually assumed that they were also exchanging mates along trade networks. However, research by Reichs (1984) which examined metric and non metric data, sugge sts that this may not always

PAGE 57

57 be the case. Her analysis of populations in Illinois and Ohio associated with the Hopewell Interaction Sphere found that there was little to no biological exchange between these two areas. This research showed that despite shar ed mortuary activities and an exchange of material goods, differing morphological patterns found among the population samples corresponded to patterns of regionally distinct adaptions in the non mortuary aspects of material culture, such as pottery (Reichs 1984) In a similar vein, researchers have analyzed language groups and their relationship to biological variation, as some assume that language differences would be a barrier to biological interaction. Killgrove (2002; 2009) has examined language groups in North Carolina where archaeologists have long held that the three language groups, Iroquoian, Algonkian, and Siouan, are compartmentalized or isolated cultural and biological entities. These bounded areas can be projected back to the pre Contact period as static entities and used to label and analyze pre Contact populations. Killgrove analyzed non metric traits of individuals from the North Carolina coastal plain to evaluate the assumption of biological separation between the language groups and she fou nd that there was little correlation between the linguistic/cultural divisions and the biological divisions found within the populations, suggesting that this long help assumption needed to be re evaluated (Killgrove 2009) h of the work in the Southeast has focused on the impact of European Contact on Native American populations. For example, Griffin et al. (2001) exa mined the pre and post Contact Guale populations in Tennessee, North Carolina, Georgia, and Florida for biological variation. The study found that the pre Contact Guale populations were rather heterogeneous, with differing amount of diversity

PAGE 58

58 at various time periods, suggesting that these populations were not as continuous as previously thought Stojanow ski has also done numerous studies of the Mission Period population in North Florida (Stojanowski 2003; 2004; 2005a; 2005b; Stojanowski et al. 2007) His research has shown that the pre Contact Apalachee of North Florida was a relatively homogenous entity, both culturally and biologically, most likely due to language differences with their neighbors and historical evidence of constant warfare (Stojanowski 2005b) They also managed to maintain aspects of their social structure well into the Contact period, i ncluding the maintenance of four biological subgroups within the population (Stojanowski 2005a) H owever, as Spanish contact increased, and disparate groups were forced together as Missions coalesced, the heterogeneity of the Apalachee population increased (Stojanowski 2005b) In his work he has also analyzed coastal and inland pre Contact populations in northern Florida and southern Georgia. His findings suggested that pre Contact populations were more homogenous than mission populations, but when these pr e Contact populations were examined alone, there was still variation evident among the populations (Stojanowski 2004; Stojanowski 2005a) late Irene Mound populations, suggesting ch anges in the biological composition of the population through time (Stojanowski 2004) Stojanowski also showed that the pre Contact inland populations of Georgia were more closely related to each other than to the coastal populations (Stojanowski 2004) and that there were biological distinctions between northern and southern sites within Florida (Stojanowski 2005b) structure at the various missions, providing details of ea rly burial practices following

PAGE 59

59 familial or cultural patterns that deteriorate as the Mission Period progresses (Stojanowski 2004; 2005b) Stojanowski (in Stojanowski and Schillaci 2006) used a similar approach in his intrasite analysis o f the Archaic Windo ver Pond population, examining the relatedness of individuals buried in the pond. He theorized that there were distinct social groups utilizing the pond for burial based on the differences in the woven shrouds found with the burials. The results indicated that there were two genetically distinct groups, or bands, using the pond for burial and that these genetic divisions coincided with the differences in observed textile s at the site. Studies conducted by Konigsberg (1990) and Key and Jantz (1981) have show n intrasite temporal trends in mounds or burial areas used by the populations that buried at particular sites. For example, analysis revealed subtle differences in morphological variation between the populations that buried individuals in Mounds 1, 5, 7, 8 and 11 at the Pete Klunk Mound Group in Illinois and that the temporal seriation of the mounds based on biological variation corresponded to the archeological seriation of the mounds (Konigsberg 1990) These studies indicate that biological continuity o r change can be observed in the bioarchaeological record as morphological variation, and that these data may be used to better understand culture change, interaction spheres, and site structure. Metric Versus Non Metric Traits One thing of note in the var ious studies m entioned above is that some relied on non metric characteristics, while other used metric data, and others utilized both; there have been many debates concerning the utility of metrics vs. non metrics with regard to biological distance analys is. It has been argued that non metric traits are more useful for biodistance studies because they are less impacted by sex and age factors, have

PAGE 60

60 less environmental variability, are not correlated to each other, and are easily defined and standardized (Cor ruccini 1974) Corruccini (1974) tested all of these assumptions, except the environmental impact aspect, by analyzing sex and age dependence, intercorrelation of variants, and the correspondence of results between non metric and metric analyses. As a result, Corruccini observed the foll owing patterns: significant sex and age related variation that correlated with traits presentation; some correlation between traits; and metric and non metric analyses produced similar results, however, metric traits produced a greater separation between p opulations than non metric traits. However, Corruccini argues that both non metric and metric analyses have a legitimate place in biodistance research. Another issue is the statistics used to analyze non metric features are still rather simplistic, often r elying on comparisons of averages or other simple statistical c alculations. In addition, some have argued that non metric traits are only slightly correlated to biological distance, they are not normally distributed, and their significance in different pop ulations varies (van Vark and Schaafsma 1992) limiting their utility in biodistance studies. On the other hand, Pietrusewsky (2000) argued that metric traits are better suited for studies of biological variation due to the precision and repeatability of m easurements, the conservative nature of the continuous variation, and the known heritability of metric traits, which is discussed below. T his project focuses on metric analysis, because as both Pietrusewsky (2000) and Howells (1973) have discussed, metric data are ideally suited to powerful multivariate statistical analys e s, allowing for the interpretations of skulls and populations as a whole not just their respective parts This was best explained by Howells (1973 :3 4 ) :

PAGE 61

61 [M]ethods of multivariate analysi as a configuration of the information contained in all its measurements. Next, they allow populations to be treated as configurations of such units, taking account of their variation in shape because they in t urn are handled as whole configurations of individual dimensions. Finally, the relations and differences between all the populations being considered are set forth in terms o f their several individual multivariate ranges of variation. Thus it is possible t o see the range of the whole species in such complete and objective informational terms. That is the importance of multivariate statistics: they fit the model of populations looked on not as centroids of means, but as swarms of the varying individuals who compose them; and the differentiation of these swarms from one another constitutes a statement of the degree and nature of the difference between the populations. Although the information is ultimately limited by the measurements selected to describe the s kull, their relationships and their relative taxonomic significance are not otherwise biased by the worker. With a review of how the study of biological variation and biological distance analysis can contribute to our understanding of archaeological varia tion and populations histories, we turn to a discussion of the theory that underlies this type of analysis. Biological D istance Theory, Assumptions, and Models Theory Biological distance studies operate on the theoretical model that all populations within a region exchange mates from an outside source at an equal rate, resulting in a linear relationship between the average within group variation and genetic distance to the centroid (average heterozygosity of all subpopulations ) for each population (the nul l hypothesis) (Relethford and Blangero 1990) However, when a population has an increased rate of exchange with an external population, their within group heterogeneity will increase due to the influx of new genes and the linear relationship will be violat ed. Therefore, groups that receive greater external gene flow will have greater within group variation than expected by the null hypothesis, while populations with less than average external gene flow will have lower heterogeneity than expected. On the oth er hand

PAGE 62

62 when looking at populations within a region and only considering gene flow within that region and not from external sources, populations that exchange mates will become more phenotypically similar, or homogeneous, over time, while populations that do not exchange mates with each other will become more dissimilar, or heterogeneous, over time (Steadman 2001; Stojanowski and Schillaci 2006) Assumptions Biodistance analysis conducted on archaeological sample populations must incorporate a few assumptions; several of which are based on living studies, while other assumptions are necessary to allow for conservative analysis and interpretation of the data These assumptions are discussed in detail below and this research will examine populations; particularly if biodistance as an indirect measure of social exchange is supported by the archaeological evidence of social interaction, if cemeteries are representative of breeding populations based on levels of homogeneity, and if environmental effects are minimal or should have a stronger consideration in some circumstances. Assumptions an alyzed The first of these assumptions, which has been supported in living populations, is that patterns of biological distance and variability can be used as an indirect measure of the degree of social exchange between the groups they represent (Basu et al. 1976; Droessler 1981; Friedlaender et al. 1971; Pollitzer 1958; Steadman 1998) This assumption can be examined in the archaeological record through the comparison of the archaeological findings for each site, in particular pottery styles, burial pract ices, and presence of trade goods. If groups are interacting socially, they are likely exchanging

PAGE 63

63 ideas and material culture. Therefore, social exchange should be evidenced in the archaeological record as the presence of trade goods or shared patterns of p ottery style, ritual activity, or burial pattern. Shared burial practices, particularly over a large distance or extended time frame, may suggest cultural continuity and/or social exchange. It may be assumed that in this process mates are exchanged, which would impact the biological distance between the populations, bringing them closer together. However, this may not always be the case, if populations are found to be either archaeologically similar but biological distant or biological similar but archaeolo gically dissimilar, such relationships would be of particular interest. The second assumption is that gene flow and genetic drift affect allele frequencies between the populations under study and that those changes are reflected in skeletal changes (Stoja nowski and Schillaci 2006) This assumption has been studied extensively both to determine the heritability of craniometrics and the impacts of gene flow and genetic drift on skeletal morphology. Heritabilities can range from 0 to 1 and studies have found that craniometrics have a high heritability, often between 0.55 and 1 (Devor 1987; Droessler 1981; Relethford and Harpending 1994). It has also been found that osteometrics may react more slowly to the effects of gene flow making them more useful in the st udy of long term migration patterns (Relethford and Lees 1982) However, genetic drift may have less of an impact than previously thought, because studies have found that the random fluctuations caused by genetic drift should cancel each other out over the large number of alleles involved in metric traits (Livingston 1972; Spielman 1973) Therefore, it can be assumed that variation observed in metric data is due to underlying genetic variation and that variation is more likely due to an

PAGE 64

64 increase or decrease in gene flow between populations than due to random genetic drift. The third assumption is that a cemetery or sample population is representative of the original breeding population, understanding that it is a temporal aggregation or lineage, not an actu al contemporaneous biological population (Cadien et al. 1974; Droessler 1981; Stojanowski and Schillaci 2006) This assumption raises a number of issues when dealing with the archaeological record as discussed, however, first it can be argued that individu als buried in cemeteries are representative of the living populations, as the living population is the one who chose where to place the dead and how they were interred (Pearson 1999) It can then be assumed that, in most instances, individuals buried in a cemetery, and subsequently included in a sample collection, participated in a culture in some form and were considered members of the group as representatives of a popul realistic or idealized (Pearson 1982; Sears 1958) Therefore a population found in a population was tight knit and homogenous, one would expect the cemetery population to also be genetically homogenous. However, if the group had a loose definition of membership and incorporated many outlying groups in various activities one could also expect this to be true in the cemetery population, many individuals of varied biological history are interred in a single cemetery. Therefore analysis of the biological variation of Additionally, i sotopic research conducted by Quinn et al (2008) on the Middle Archaic

PAGE 65

65 site of Tick Island / Harris Creek in Florida found that the majority of the individuals were Harris Creek, or were brought to the site after their death, were isotopically distinct and local individuals. These findings suggest that a population found within a Flori da burial site will most likely be representative of the individuals who lived in the area and therefore would be considered the breeding population. Despite these positive results, the analysis of sample collections recovered from burial sites and interpr eted as populations and/or lineages must be approached with conservatism and caution. Aside from the potential cultural bias of who was placed in a burial site verses who was placed elsewhere, there are additional biases related to excavation strategies (s ections verses whole cemetery excavation), and curation methods (selection of remains to be kept or discarded), all of which may alter the extent of variation observed in the sample. This bias can be accounted for, to an extent, by reviewing collection his tories; however, this information is not always recorded. The potential for bia s is considered in the analytical process. The fourth assumption is that environmental effects are minimal and randomly distributed among the populations ( Stojanowski and Schil laci 2006) In the study of biodistance it is understood that environmental, sexual, and age related factors can affect results, but statistical adjustments, selection of minimally affected traits, and regional approaches can lessen the impact of environme ntal variables and selective effects (Relethford and Lees 1982; Stojanowski and Schillaci 2006) A few studies have focused on craniometrics and environmental factors. Devor (1987) found that

PAGE 66

66 environmental effects were minimal on craniofacial measurements affecting soft tissue measurements more than skeletal measurements. Lundstr m (1955) found in a twin study that genetics played a stronger role than environment in craniofacial form, even under the extreme condition of thumb sucking Relethford (2004) foun d that climate and developmental plasticity can influence craniometrics, however, the factors do not erase the underlying population structure. Relethford (2002) has also shown that, with regard to craniometrics, approximately 13% of variation is found amo ng geographically / environmentally distinct regions, 6% of the variation is found within a region, and 81% is found within the populations in a region. These findings suggest that, due to the multivariate pattern of among group variation, genetic distances found are in agreement with selectively neutral genetic markers. In other words, natural selection does act on cranial morphology, but by using multiple craniometrics, those measurements impacted by selective pressures will have less of an impact on the ou tcome and therefore results can be viewed as the result of gene flow between the populations understudy. Further, a skeletal study by Rothhammer and Silva (1990) analyzed South American prehistoric populations from a moderately tight temporal frame with va ried geography, climate, and altitude, found that geographic isolation had the greatest impact on biological distance, whereas climate and altitude played only a small role in the biological distances observed. Therefore, in this study, the environmental influence on variation is most likely minimal, particularly among archaeological populations from similar temporal frames and location. The environment in Florida is not as variable as that demonstrated in global population studies, or that seen in the wid ely varying environments of South

PAGE 67

67 America The environment has changed through time in Florida, particularly when comparing the Early Archaic to the Early Woodland or the Mississippian time periods; therefore if significant differences are found among the populations associated with these time frames, environmental changes will have to be considered as a factor. In summary, environmental selection has been shown to impact craniometrics, but it does not obscure the underlying population relationships. Theref ore, the results of this study can still be used to interpret cultural and biological continuity and change from the Archaic to pre Contact period. Model derivations The study of biological variation is based on population genetic models; however, these models are of limited use when studying skeletal samples that no longer contain intact DNA. Therefore, Relethford and Blangero (1990) modified the Harpending and Ward model (1982) of heterogeneity for allele frequencies to be used with continuous traits, s uch as craniometrics. This modification has allowed for the magnitude and direction of gene flow across space and time to be modeled in skeletal samples. The modification of these formulae for continuous traits is fully described by Relethford and Blangero (1990) and Relethford et al. (1997) relationship (R) matrix, which is a variance covariance relationship matrix of population similarities, which provides genetic distance estimates and allows for extra local ge ne flow patterns to be assessed (Steadman 1998; Stojanowski and Schillaci 2006) The Harpending Ward model states that the expected heterozygosity of populations i[E(H i )] is a function of the heterozygosity of the total region ( H t ) and the genetic distance between population i and the regional centroid ( r ij ). E(H i )= H t (i r ij )

PAGE 68

68 The genetic distance of a population to the regional centroid is computed as the diagonal of the R matrix of scaled variances and covariances about the regio nal mean allele frequencies. For each allele the elements of the R matrix for populations i and j are computed as: where p i and p j are the allele frequencies of populations i and j respectively, and p is the average allele f requency for all populations in the analysis weighted by population size. The elements of the R matrix are then averaged overall alleles (Relethford and Blangero 1990) The multivariate equation for calculating expected within group heterogeneity for quan titative traits is a function of the pooled mean within group phenotypic variation across all subpopulations within a region (v w ), the genetic distance of subpopulations to the centroid (r ji ), and the average genetic distance to the centroid across all reg ional subpopulations (F ST ), such that: E(v i )= v w (1 r ji ) 1 F ST The residual is the difference between the expected and observed heterozygosity, the sign and magnitude of which indicates the relative extent of gene flow from external sources into each population; so positive residuals indicate greater external gene flo w, while negative residuals indicate lower than expected rates of external gene flow. The total genetic variation among populations, F ST is given as the average weighted diagonal of the R matrix:

PAGE 69

69 F ST = w i r ji Where w i is the relative population size of populations i, and g is the number of populations (Relethford and Blangero 1990) F ST represents the minimum genetic differentiation between the regional populations under the assumption that all heritabilities are equal to 1 (Williams Blangero and Blan gero 1989). As mentioned previously, heritability for a trait can range from 0 to 1 (Devor 1987; Droessler 1981; Relethford and Harpending 1994), and a heritability of 0 indicates that none of the phenotypic variation of that trait results from additive ge netic variation While a heritability of 1 indicates that all phenotypic variation results from the underlying genetic variation. Heritabilities close to 0 indicate a lack of genetic variation not a lack of genetic control, while heritabilities close to 1 i mply little selection, little environmental variation, and maximum additive genetic variation (Stojanowski 2005a) Therefore using a heritability of 1 in these calculations results in the minimum genetic distance and minimum F ST being calculated (Stojanows ki 2005b) ; therefore any results found to be significant can be safely interpreted as significant. Computations of R require that certain quantities be weighted by population size producing an unbiased R matrix. This accounts for differences in effective populations size, statistically removing the potential effect of genetic drift in smaller populations (Relethford and Blangero 1990), however for these analyses the population sizes are unknown, so all populations were weighted equally, using a value of 1, which again gives the minimum results (Relethford 1996). The genetic distances between the populations are calculated by converting the elements of the R matrix into a distance matrix (Relethford and Harpending 1994)

PAGE 70

70 d 2 ij = r ii r jj 2r ij Methodologies There are two types of methodologies for determining biological distances: model free and model bound (Relethford and Lees 1982) Model free procedures indirectly apply models to measure biological similarity, which are then used to assess overall phenotyp ic similarity. Model free procedures can be broken down into differentiation studies and comparative studies. Differentiation studies seek to determine the extent of variation among groups and the biological distance between groups without evaluating the p attern of that variation. Comparative studies seek to determine the pattern of among group variation and then relate that pattern to other biological, demographic, and/or historical patterns (Relethford and Lees 1982) The comparative studies use correlati on matrices to directly compare geographic and biological distances (Howells 1973; Relethford and Lees 1982; Steadman 1998) or other cultural and historical factors relating to population isolation and migration (Steadman 2001) The results of these correl ation matrices would suggest if there was a geographic, linguistic, and/or social relationship between the clustered populations or would indicate what factors may have barred gene flow (Relethford and Lees 1982) Model bound methods directly incorporate measures of population similarity into the models ultimately seeking to estimate a specific parameter, such as admixture, genetic drift or gene flow (Pietrusewsky 2000) These methods often require more assumptions than model fr ee methods, but they can be used to evaluate kinship and isolation by distance (Relethford and Lees 1982) Model bound methods require population estimates, heritability estimates, and tight temporal frames.

PAGE 71

71 In summation, this discussion has shown the utility of the analysis of biological variation of skeletal populations, the assumptions used in that analysis, and the models used in these types of analyses. After this review it should be clear how these metho ds will provide valuable insight in the biological and cultural variation of pre Contact Florida and particularly the questions highlighted in the previous discussion of Florida archaeology. The following chapter details the methods used and the population samples studied in this analysis.

PAGE 72

72 CHAPTER 4 METHODS AND MATERIALS It was determined that metric data would be utilized for this study because sex and age factors can be corrected for statistically, measurement definitions are open to less interpretatio n than non metric trait definitions, and there is no need for a threshold value with metric assessments as there are with non metric data. Corruccini (1974) stated that non metric traits are more susceptible to drift due to their underlying genetic nature, whereas metric variation is more impacted by gene flow, which is the focus of this study. Furthermore, metric data allows for multivariate analysis, which allows a skull to be handled as a unit or a configuration of measurements. Multivariate analysis all ows the comparison of all the points as a whole for an individual and for a population. It allows to full range of variation within a population to be visualized and assessed (Howells 1973). Data Collected Craniometric data were collect on from 404 skulls from 27 pre Contact archaeological site s throughout peninsular Florida (Table 4 1). These samples are housed at the Florida Museum of Natural History, Florida State University, Florida Gulf Coast University, University of Miami, Florida Atlantic Universit y, the Florida Department of Historic Resources, the Sarasota County History Cente r, and the Smithsonian Institution. A suite of 40 measurements was originally recorded for each individual based on a set of standard measurements (Table 4 2) (Buikstra and Ubelaker 1994; Droessler 1981 ) (Rhine 1990) were developed by the author to permit the inclusion of more individuals. These new

PAGE 73

73 measurements were 1) orbital width at nasion, which is the nasal widt h taken at the height of nasion, and 2) nasion to the external point on the frontomalar suture (fmt), which is the distance between these two landmarks in a horizontal plane. The second measurement was taken as a potential proxy for upper facial breadth (f mt fmt) so that when one of the sides was damaged the nasion to fmt measurement could be doubled and used as a replacement for the upper facial breadth measurement. However, the author conducted a paired sample t test, utilizing SPSS 19 (Inc. 2010) and fo und that the upper facial breadth measurement and the nasion to fmt measurement doubled do not correspond well, therefore this new measurement cannot be used as a proxy for the other All measurements were taken to the nearest millimeter using spreading an d sliding calipers and a mandib u lometer following standard anthropometric techniques. Measurements were taken preferentially on the left side, when the left was missing the right was substituted. A definition of all measurements can be found in Appendix A. The author followed standard anthropological techniques for the estimation of skeletal sex and age (Buikstra and Ubelaker 1994) Sex was labeled in the following categories: male, male?, unknown, female?, or female. Sex determination was assessed principally on cranial morphology (Buikstra and Ubelaker 1994) O nly adults were used in this analysis Age was labeled in these categor ies: unknown, young, young middle, mid dle, middle old or old Age estimation was based primarily on dental attrition of the molars, young was assessed as little to no wear, particularly on the 2 nd and 3 rd molars, middle as moderate wear on the molars, and old as extreme wear on the molar, particularly the 1 st and 2 nd molars; these methods followed the Murphy and Miles methods (Miles 2001; Murphy 1959a; Murphy 1959b) Analysis of postcranial

PAGE 74

74 indicators of sex and age were abandoned early in the study when it became apparent that individuals were often commingled and parsing out matching cranial and post cranial elements was not possible when working with bones that are out of context. To test for intraobserver error, 22 individuals from Palmer Mound and Dunw ody were re measured approximately six months after the initial data collection. SPSS 19 (Inc. 2010) was used to perform an independent t test for significant differences between measurement one and measurement two (Table D 1) A 95% confidence interval wa s used. Mastoid length [2] was found to have a high rate of intraobserver error; therefore it was removed from further analysis. Missing Data Many individuals observed were fragmentary and yielded incomplete data. Some variables were missing more often th an others. Analysis could not continue without a complete data set, therefore the issue of how to fill in missing data had to be addressed, as removing all variables with missing data would have resulted in all variables being removed or removing individua ls with incomplete data would also have resulted in nearly all individuals being removed from the analysis. Therefore the missing data needed to be filled in for at least some of the variables allowing for more of the individuals to be included in the anal ysis. Missing data can be completed in one of several ways including substitution of group means, substitution of grand means, or prediction of the missing value by means of multiple regressions. The first two methods result in a loss of variability either within the group, which inflate inter group distances or between groups, which diminish or obscure inter group differences. Multiple linear regression uses the patterns of variation within the known variables to fill in the missing variables resulting in size consistency being maintained and as a statistical method is

PAGE 75

75 less likely to affect within group variance (Droessler 1981) Thus, multiple linear regression analysis was chosen to fill in the missing data. However, this method bases the imputation on th e other variables, both within the individual and within the populations that are available so if there are very few measurements available to produce regression formulae the results are limited. Therefore, it was determined that measurements missing in > 60% of the individuals sampled would be removed from further analysis. This decision resulted in the following measurements being removed from analysis: bizygomatic breadth, basion prosthion, alveolar breath and length, nasion prosthion, nasal breadth, bio rbital breadth, interorbital breadth, mastoid length, chin height, bigonial breath, and bicondylar breadth. In addition, after review it was determined that three of the Archaic P eriod sites did not have mandible measurements. Therefore, either these measu rements would have to be filled in based on variability from other sites, potentially altering the true variability of the populations sampled or these measurements would have to be removed for the entire study. It was determined that at this time these me asurements would be removed. In summary, 20 measurements were used in this analysis that follows (Table 4 3). Appendix B contains the raw data for the 20 measurements used in analysis. To complete the missing data, multiple linear regression analysis was conducted on the data collected using SPSS 19.0.0 (Inc. 2010) Regression by site, and when possible by sex within site, was done to generate values for the missing data. In all cases, the error terms were chosen randomly from a normal distribution since m ore than half of the cases were missing data.

PAGE 76

76 The missing data for Gauthier, Windover, Palmer, Bayshore Homes, Manasota Key Cemetery Crystal River, Highland Beach, Margate Blount, Santa Rosa, Safety Harbor, and Captive Island were completed using regress ion analysis by site and by sex. Warm Mineral Springs, Bay West, Republic Groves, Bird Island, Bay P ines, Browne Tract, Casey Key, and Hutchinson Island were completed by site. The remaining sites were too incomplete to fill in at the site level. The remai ning sites represent several cultures and are from diverse temporal frames and geographic areas; therefore filling in the missing data had to be done selectively. Research suggests that sites within similar geographic regions or similar time frames are mor e likely to be similar to each other than sites from more distant locations or times (Konigsberg 1990a; Konigsberg 1990b) Therefore filling in the incomplete data by analyzing sites from the same cultural time frame or geographic region would most likely be more representative of the true variation within the populations than trying to complete the missing data by comparing disparate samples. The completed data sets from Bay West, Republic Groves, and Gauthier were analyzed together to fill in the missing data for Little Salt Springs, as these sites are closest to Little Salt Springs chronologically and culturally. Henderson Mound, Hughes Island, and McKeithan were analyzed with Browne Tract and Crystal River to fill in the missing data because these sites have a Deptford and/or a Weeden Island component. Dunwody was analyzed with Captive Island to fill in missing data because of geographic proximity. Yellow Bluffs was analyzed with Palmer Mound, Venice Beach Complex, Casey Key, and Manasota Key Cemetery bec ause these sites are considered to belong to the Manasota culture. Sarasota Bay Mound was

PAGE 77

77 analyzed with Safety Harbor Mound to fill in missing data as both of these sites are identified as belonging to the Safety Harbor culture. Statistical Methods All va riables were converted to Z scores, which removed the size differences or sexual dimorphism influences, while maintaining the shape differences (Relethford 1994) Once the data set was complete, SPSS 19 (Inc. 2010) was utilized to conduct Principle Compon ent Analysis (PCA) and Discriminant Function Analysis to explore the variation in the data. The preliminary analysis of these data focused on the variation within the populations as a whole, first to determine if there were any patterns to the variation a nd then to compare the variation within populations to the centroid, or the average variation of all the populations being investigated. The initial analysis was performed using principal component analysis (PCA), which focuses on the interrelationship amo ng a large number of variables and seeks to identify underlying patterns of variation (Pietrusewsky 2000) PCA is a statistical method used for data reduction that seeks to maximize group differences, so it does not utilize any a priori knowledge, such as site PCA results may be viewed as an unbiased projection of the variability patterns within a sample. PCA essentially converts the variables into eigenvectors, which represent patterns in the variations, usually the product of several variables acting tog ether to differentiate the groups (Pietrusewsky 2000) To test the suitability of this analysis for the data the Kaiser Meyer Test of Sphericity were examined. KMO indicates if there are sufficient it ems for each factor or whether or not using this analysis with these variables is appropriate; a value

PAGE 78

78 greater than 0.5 indicates that this is a method is suitable for the data being analyzed. e variables in the population correlation matrix are not correlated ultimately indicating whether or not the correlation matrix is significantly different than the identification matrix. After PCA, discriminant function analysis (DFA) was conducted, where the independent variables are the predictors and the dependent variables (i.e. site, culture, geographic region, etc.) are the groups Unlike PCA, DFA does seek to maximize group differences; therefore it does take a priori knowledge into account. DFA is usually used to predict membership in naturally occurring groups. DFA analysis is broken into a 2 step process: (1) testing significance of a set of discriminant functions, and (2) classification. The first step is a step wi se process that evaluates which variables maximize intergroup distances to develop discriminant function coefficients and formulae. The second step uses these formulae to classify individuals into groups and the cross validation step checks the classificat ion results giving an evaluation of the strength of the results (Pietrusewsky 2000). This is done because the estimates of accuracy in the classification step are often artificially high as the same individuals that were used to create the discriminant fun ctions are used to test its performance, inflating its accuracy. Whereas, the cross validated results are a more accurate representation of the utility of the functions as they operate with a leave one out method in which one individual is held out of the analysis and the discriminant function is generated from n 1 individuals in the samples then the held out individual is classified using this new function. The process is repeated for all individuals creating an unbiased estimate of the functions performan

PAGE 79

79 provide significant discrimination between the groups, it is also an indication of the total, unexplained variance, for the discriminant function. Therefore, a smaller Wilks' Lambda value is desired as this indicates that there is little unexplained variance and a greater proportion of the total variance is explained by the discriminant function. If groups were identified through PCA and DFA methods then the next phase of analysis, R matrix and biological distance estimation were undertaken to estimate how closely related each sample population is. These calculations were done utilizing the RMET5.0 program, provided by John Relethford (Relethford and Blangero 1990; Relethford et al. 1997) Fo r these analyses a heritability of 1 was used to obtain the minimum genetic distance and minimum F ST Also, since the population sizes are unknown, all populations were weighted equally, using a value of 1, again giving the minimum genetic distance (Releth ford 1996) This phase of the analysis will help us understand how the populations being studied are related to each other; small biological distances would suggest that the populations are related, while large distances would suggest that the populations are unrelated (Steadman 2001) The unbiased F ST or the total genetic variation among the populations, is examined to give an indication of the amount of variab ility within the populations. The F ST provides an easily interpreted assessment of the biologica l variation of the populations involved in an analysis, which can be compared to other researchers F ST results for cross cultural comparison of genetic variation regardless of data type (Steadman 2001) The F ST is the sum of the product of a populations si ze multiplied by the populations distance from the regional centroid (r ii ), therefore the number of populations used and the variation found within those populations impacts the F ST value. In addition, analysis of the residuals, the

PAGE 80

80 observed variation minu s the expected variation, indicates size and direction of gene flow, positive values indicate flow into the population while, negative values indicate lower than expected levels of flow into the population or even gene flow out of the population. In each a nalysis, populations were removed as analysis proceeded based on variability or archaeological interpretations to facilitate further analysis. Each of these situations is discussed in the results section. Pearson product moment and Spearman rho correlatio ns were used to determine if there were any significant correlations between biological distance and time or geography. Pearson is used for values that are measured on a continuous scale and are approximately normally distributed, while Spearman is used fo r data that are ordinal or violate the assumptions of Pearson correlation, such as a skewed distribution. The statistic r and r s respectively, estimates the correlation between the variables and the P value is reported as the 2 tailed sigma. Materials Ar chaic Sites Warm Mineral Springs (8SO19) Warm Mineral Springs is a now inundated Early Archaic or possibly late Paleoindian burial site in southern Sarasota County. The site was a sinkhole when water levels were considerably lower than today, however as sea level rose groundwater levels rose and the sink became a spr ing. Some evidence suggests that the burials found in the spring were above the water at the time of their interment on the ledges or cave like recesses in the spring opening (Cockrell and Murphy 1978) While other research indicates that the remains may n ot be the result of direct interment and

PAGE 81

81 are either the result of individuals falling into the spring and dying or bodies being thrown into the spring and decomposing (Clausen et al. 1975) Radiocarbon dates place human activity from 1 0,630 8920 BP (Clause n et al. 1975) There were three crania available for this study, one was excavated from the 35 40 foot ledge and is well provenience, another cranium came from the debris cone at the bottom of the spring and therefore its original burial location is unkno wn, and the third cranium has no provenience (Skow 1986). Windover Pond (8BR246) Windover is an Early Archaic pond burial site located in Brevard County. The burials appear to be primary flexed burials and were placed predominately along the north and west edges of the pond. The burials may have been placed in band and/or family groupings with in the pond. Evidence of wooden stakes and woven mats suggest the burials were intentionally staked down into the peat of the pond. The burials date from 9,000 7,900 BP ( Doran 2002). The vast majority of the remains were collected during controlled excavat ions, with only the earliest discovered materials were collected from dredge piles. Little Salt Spring (8SO18) Little Salt Spring is a Middle Archaic pond and slough burial site in southern Sarasota County. Much like Warm Mineral Springs, Little Salt Sprin g is a former cenote turned spring. The earliest human activity at this site dates to the Paleoindian P eriod however, the largest amount of activity including the interment of the dead dates to the Middle Archaic period (Clausen et al. 1979) The remains collected in controlled excavations from two locations, the spring basin and the slough. The remains excavated from the slough were in an extended position on a pier of green wax myrtles

PAGE 82

82 and evidence suggests that there may have been stakes or structures a round the burials. Radiocarbon dating places one of the slough burials at 5850 70 BP. The spring burials appear to be primarily from the eastern side of the basin, though a child burial was found on the western edge. One of the buri als in the basin was d ated to 5 220 90 BP ( Clausen et al. 1979). Unfortunately, the exact provenience of the remains used in this study, whether they are from the slough or the basin, is unknown at this time, so the spatial relationship of the remains is currently unknown Ba y West (8CR200) Bay West is a Middle Archaic pond burial site located in Collier County. Unfortunately, due to the salvage nature of the excavation, orientation and provenience of the burials was not recorded; the remains were collected as the pond was dre dged in transects (Beriault et al. 1981) The presence of burned and shaped wooden stakes suggests that the burials were staked down into the peat like Windover (Beriault et al. 1981) The burials date from 7,550 5,500 BP Of note, one stone bead manufactu red from non Florida green stone was found with the burials, suggesting an extensive trade network or extra local individuals (Purdy 1991a) Republic Groves (8HR4) Republic Groves is a Middle Archaic pond burial site that is located in Hardee County. The site was discovered when a ditch was dug through the site and other heavy machinery activity in the area resulted in extensive disturbance to the burials, therefore many of the remains have little provenience information. Further excavation was done in a c ontrolled fashion allowing for the notation of burial information including the fact that many of the artifacts or burial goods were found with the child and infant burials, similar to the pattern seen at Windover (Purdy 1991b) The remains appear to have been

PAGE 83

83 primary, flexed burials and one set of six burials appears to have been an intentional group burial possibly suggesting a family (Wharton et al. 1981) Tapered wooden stakes, showing evidence of hammering, were also found in association with the burials (Wharton et al. 1981) The stakes were dated between 6520 65 BP and 5745 105 BP. A pendant made a non local serpentine like material was found with the remains again suggesting an extensive trade network or non local indi viduals (Purdy 1991b) It is estimated that only about 1% of the potential cemetery area was actually sampled (Wharton et al. 1981) Gauthier (8BR193) Gauthier is a Middle Archaic cemetery that was situated along the edge of a pond located in Brevard Coun ty. The burials appear to be primary flexed burials and may have been placed in clan and/or family groups within the site. However, these burials were not pond burials as seen at the Windover and Bay West sites. The projectile points found wit h the burials suggests a Middle to Late Archaic date, which is suppo rted by a radiocarbon date of 4 340 170 BP (Carr 1981; Maples 1987) Bird Island (8DI52) Bird Island is a Late Archaic cemetery site located along the coast in Dixie County. The burials from this si te were collected or excavated at various times and little is known about their orientation or provenience, but it appears that all the remains came from one small area on the south side of the island. Also, one observer indicated that the burials appeared to have been discrete, flexed burials (McFadden 2011) This site is on the northwest coast of Florida and it not adjacent to any large fresh water body, such as a pond. One radiocarbon date fo r a burial suggests an age of 4 570 110 BP (Stojanowski and Do ran 1998) This site has also yielded the largest soapstone sample

PAGE 84

84 in Florida dating to ca. 3800 BP. The presence of soapstone suggests that these individuals either traveled from or were in frequent contact with groups outside Florida, since the closest s ource of soapstone is in north central Georgia (Lovejoy 1985; McFadden 2011; Stojanowski and Doran 1998; Yates 2000) Weeden Island / Manasota Sites McKeithan (8CO17) McKeithan is three mound complex located in Columbia County excavated during various field sessions in the mid to late 1970s by Dr. Jerald Milanich. Human remains were found in all three mounds; however the two individuals used in this study come from Mound B and Mound C. The burials were primarily secondary burials resulting from the use of a charnel house, with the notable exception of the individual buried in the center of Mound B. A pottery cache was found in Mound C. The site appears to be a ceremonial center associated with the Weeden Island Complex Radiocarbon dates indicate that the mo unds were used approximately 1700 BP (Milanich et al. 1997) Crystal River (8CI1) Crystal River is a large midden / mound complex located in Citrus County that contains two flat topped temple mounds, two burial mounds, several middens and a stela. The remains used in this study were excavated by Ripley Bullen in 1960 from the main burial mound and a smaller burial mound called Mound G. However, despite an extensive study by Ka t zmarzyk (1998) no provenience information is available about the remains beyond the level of mound association. The main mound contained predominately primary extended burials and there was evidenc e of pottery caches (Willey 1949) while Mound G does not have any evidence of a pottery cache and many

PAGE 85

85 of the burials appear to have been primary, flexed (Katzmarzyk 1998) Various artifacts found during the Moore and Bullen excavations suggest an extende d trade network (Weisman 1995) Pottery analysis and radiocarbon dates suggest that the site was occupied from the Deptford period through the Weeden Island Complex or 2200 BP 1400 BP and it appears that the two burial mounds may have been contemporaneous (Katzmarzyk 1998; Pluckhahn et al. 2010) Radiocarbon dates from the burial mound (2490 60 BP) and Mound G (2520 60 and 1990 40) suggest that at least some of the individuals in both mounds were buried during the Early Woodland Period (Pluckhahn et al. 2010) which corresponds to the Deptford period Artifacts from this site have also suggested a link to the Yent Complex (Sears 1962) that predates, but is believed to be associated with, the later Weeden Island Complex (Milanich 1994; Stephenson et al 2002) Site usage, and most likely burial, continued into the late Weeden Island period based on pottery analysis and additional radiocarbon dates from the various structures at the site (Pluckhahn et al. 2010) Hughes Island Mound (8DI45) Hughes Island is a burial mound located in along the coast in Dixie County. The mound was excavated during a field school conducted by John Goggin in 1954. A 20 foot trench was excavated in five foot squares. The burials were concentrated near the center of the mound. T he base of the mound appeared to be midden refuse that was then capped with tan sand and secondary burials were placed on the surface and covered (Silbereisen 1954) Pottery analysis suggests that the mound dates to the early Weeden Island period and a rar e embossed copper disc suggests long distance trade with the Great Lakes region (Christman 1954)

PAGE 86

86 Yellow Bluffs Mound (8SO4) Yellow Bluffs is a soil burial mound located along Sarasota Bay in Sarasota County. The mound was excavated in 10x10 squares by t he Sarasota County Historical Society with the assistance of Dr. Jerald Milanich in 1969. The mound was originally dated to the Safety Harbor period (Milanich 1972) however reanalysis of the pottery and radiocarbon dating indicate that the mound dates to the early Manasota culture at the Deptford horizon (2115 2040 BP) (Luer 2011; Luer and Hughes 2011) Pottery sherds include Deptford pottery, Orange and fiber tempered pottery, and St. Johns pottery and paste analysis indicates that several of the sherds originated in north Florida, suggesting an extended trade network (Luer 2011) Ten burials were exca vated on the west side of the mound, most of which were primary, flexed (Milanich 1972) and it appears to be a continuous use mound (Luer 2011) The provenience information for the two individuals examined in this study has been confused through time; how ever, in reading Milanich (1972) and interpreting notes on the remains it can be suggested that one individual is from Burial 10 and the other is either from Burials 5, 6, 7, or 8. Dunwody (8CH61) Dunwody is a shell midden site located on the shore of Le mon Bay in Charlotte County. The burials were collected during a salvage project conducted by Ripley and Adelaide Bullen in 1965. The burials were collected after being exposed by a bulldozer from the northwestern edge of the midden along the shoreline; ho wever, no other provenience information is available (Gold 2006; Luer 1999) The burials appeared to have been primary flexed burials (Gold 2006) Based on artifacts collected at various times, the midden evidences multiple occupations dating from the Arch aic through the Safety Harbor period (Luer 1999) Two radiocarbon dates on bone provided dates of

PAGE 87

87 1150 40 BP and 1290 40 BP (Gold 2006) suggesting the burials are associated with the Manasota period. Casey Key (8SO17) Casey Key is a sand burial mound located on the north end of Casey Key in Sarasota County. The collections used in this analysis were the result of donations by various individuals who excavated in the mound, therefore little is known about the provenience of the remains; the mound has s the site in 1959 and made a small collection of pottery. This collection and a donated collection from the site contain numerous plain sherds and a high percentage of Weeden Island decorated sherds, (Bullen and Bul len 1976) suggesting a late Manasota period site. Palmer Burial Mound (8SO2A) Palmer is a sand burial mound located in Sarasota County. The burials were collected during controlled excavations conducted by Ripley and Adelaide Bullen in 1959, 1960, and 1962. Approximately 400 burials were excavated; the majority of the burials, 75%, were pr imary, flexed burials, the remaining 25% were mostly from the upper layers of the mound and consisted primarily of secondary or bundle burials. The pottery associated with the mound is predominately sand tempered plain with an increased incidence of decora ted or Weeden Island type pottery in the upper layers. published analysis of the site suggests that there were three phases of building / burial in the mound that coincided with changes in pottery type frequencies and burial patterns; an early phase possibly associated with the dedication of the mound, a middle phase associated with most of the burials, and a later phase that appears to have Weeden Island influences. Two radiocarbon dates were performed providing dates

PAGE 88

88 of 1100 105 BP from the upper layer of the mound and 2350 110 BP from the lowest layer of the mound. Bullen indicated that the first date was accurate and consistent with the pottery chronologies, but that the second date was too early for the mound and therefore was the result of f ill brought in by the later populations from other previously inhabited locations (Bullen and Bullen 1976) The burial patterns, pottery, and radiocarbon dates suggest that this is a Manasota period burial mound with a late Weeden Island influence evident in the upper most layers. Manasota Key Cemetery (8SO1292) Manasota Key is a cemetery site located on a natural beach ridge in Sarasota County. There is no published report on this site at this time. Field notes housed at the Florida Department of Historica l Resources indicate that the burials were primary, flexed burials. Sand tempered plain pottery found within the site is consistent with the Manasota culture. Radiocarbon dates taken from a bone and a shell sample date to 1800 80 BP and 1830 90 BP, respectively, coinciding with the late Manasota period. Venice Beach Complex (8SO26) Venice Beach Complex is a midden/mound complex located along the coast in Sarasota County. The site consists of several shell middens and possibly burial mounds s ome of which are now underwater (apparently due to coastal erosion) ; however the site has been extensively damaged by modern construction. It is unknown at this time when or where the remains were excavated. The site has been associated with the Weeden Isl and or Manasota cultures based on pottery analysis (Luer and Almy 1982; Milanich and Fairbanks 1980)

PAGE 89

89 Bayshore Homes (8PI41) Bayshore Homes is a part of a mound complex located on Boca Ciega Bay in Pinellas County in the Tampa Bay area. This complex appe ars to have been occupied during two phases, first during the late Manasota period (approximately 1600 1200 BP) and then again during the late Weeden Island / early Safety Harbor period (approximately 1000 800 BP) (Austin 2012) Burials have been excavat ed in both Mound B and Mound C. It is believed that Mound C dates to the earlier phase and Mound B dates to the later phase based on pottery analysis. The burials appear to be a combination of flexed and bundle burials. Mound B has two apparent phases sepa rated by a thin layer of white sand. The upper layer appears to be older midden that is re deposited over an earlier layer, based on radiocarbon dates, including some rather anomalous dates for two upper burials that date to the early Archaic (3728 4065 BP and 2153 2337 BP) (Austin et al. 2012) The burials appear to have been concentrated primarily on the southwest side of the mound with a smaller grouping of burials on the northeast side of the mound (Austin et al. 2008; Sears 1960) The burials used in t his study were primarily excavated by William Sears with the Florida Museum of Natural History from Mound B between 1957 and 1958. The rest of the remains were collected by non professionals; a few of which are believed to have been excavated from Mound C, while others were collected from unknown locations within the site; most likely Mound B or Mound C. Bay Pines Site (8PI64) Bay Pines is a sand and shell burial mound located on the grounds of the Veterans Administration Hospital along the shore of Boca Ci ega Bay in Pinellas County. The site consists of four shell ridges or mounds, one of which is the Nursing Home

PAGE 90

90 Mound, which yielded the skeletal remains used in this study. The remains were excavated by Gallagher and Warren in 1971. Most of the burials app eared to be primary, flexed burials. Pottery analysis suggests an early Weeden Island component to the burials, however some Deptford pottery was found in association with other burial and there was no apparent pottery cache found at the site. There may ha ve been two burial phases though as the burials are found on two sides of the ridge, one dominated by sand and the other by shell (Gallagher and Warren 1975) Luer and Almy (1982) associate this site with the Manasota culture. Little to no provenience info rmation is available for the remains used in this study. South Florida Sites Captiva Mound (8LL57) Captiva was a sand burial mound located on a barrier island in Lee County or the Caloosahatchee region. The site had been partially looted prior to excavati on by Henry Collins, Jr. from the Smithsonian Institution in 1928. This mound is now destroyed. Collins (1929) described two apparent burial phases, an earlier phase with primary, flexed burials and a later phase with bundle burials. The latter overlay the earlier phase (Collins 1929) Unfortunately, there is no additional information available about the provenience of the remains examined in this study, such as their general spatial relationships or if they came from the earlier or later phase. A radiocarb on date of 1590 40 BP places this mound in the early Caloosahatchee culture phase (Stojanowski and Johnson 2011) Hutchinson Island Burial Mound (8MT37) Hutchinson Island is a sand burial mound located along the Atlantic Coast in Martin County. The buri als were excavated in 1971 by an avocational archaeologist,

PAGE 91

91 Robert Holman. He described the burials as primarily consisting of skulls and suggested that the scattered and damaged bones throughout the mound were the result of damage from subsequent burials. The only diagnostic artifacts found with the burials were Glades plain sherds (Holman 1975) It has been suggested that this mound may have an Archaic component, but most likely dates to the Glades I time period ( Carr and Steele 1993; Wheeler et al. 2002) Highlands Beach Mound or Boca Raton Beach Burial Mound (8PB11) Highlands Beach is sand burial mound that is part of the Spanish River Complex located on the Atlantic Coast in Palm Beach County. Approximately 128 burials were excavated in 1980 with assis tance from then at Florida Atlantic University. The excavation was done in 2x2 squares and was a somewhat hurried process due to the impending condominium construction. Detailed information about the provenience of the remains is lacking beyond the 2x2 square level, however, the burials are described as being primary, flexed burials (Winland 2002) Pottery analysis suggests that the mound dates to the Glades II and III periods or 1400 800 BP (Wheeler et al. 2002; Winland 2002) Margate Bl ount Mound (8BD41) Margate Blount is a black muck burial mound located in Broward County. The site, which consists of a habitation site and burial mound, was excavated over a three year period by the Broward County Archaeological Society. Extensive distur bances to the site destroyed the integrity of some of the burial mound and minimal provenience information is available. Both primary and secondary burials were found, some were found in association with wooden slabs, in addition several burials were appar ently found in a log crypt (Williams 1983) Pottery analysis suggests that the site is

PAGE 92

92 associated with the Glades I I and II I periods, but it is believed that the burial mound was constructed during the Glades II period (Williams 1983) Safety Harbor Sites Sarasota Bay Mound (8SO44) Sarasota Bay Mound is a sand burial mound located in Sarasota County. The mound was excavated by Ripley Bullen in 1968. There is no written report available regarding the excavation. The site is listed as either Weeden Island or Safety Harbor in the Florida Master Site File. Safety Harbor (8PI2) Safety Harbor is a sand burial mound located on the northwest shore of Tampa Bay in Pinellas County and is associated with a temple mound and a midden component. The burial mound is now considered destroyed. The remains used in this analys is were excavated by William Ste rling in 1929. The burials were described as secondary burials laid down in stages as evidenced by layers in the sand. Evidence at the base of the mound. Pottery found within the mound suggests a Safety Harbor period burial mound and the pottery is described as quite similar to the Fort Walton pottery series (Willey 1949) This time frame is supported by a radiocarbon date on huma n bone from the mound dating to 1150 40 BP, which is the late Weeden Island or early Safety Harbor period (Stojanowski and Johnson 2011) Alachua/St. Johns/Fort Walton Sites Henderson Mound (8AL463) Henderson is a low sand burial mound located in Alachua County. The mound is associated with the Hickory Pond Phase of the Alachua culture ( AD 800 1250). This site

PAGE 93

93 was excavated in 1975 by Charles Fairbanks. There were approximately 41 burials; however they were in very poor condition so many of the remains were reinterred and only a few individuals were collected for analysis. The majority of the burials were primary, extended burials, with a few in a flexed position and a few secondary burials (Loucks 1976) The earliest burials appear to have been placed in shallow graves, and then the mound was constructed around the burials with subsequent burials added to the margins, particularly along the western edge of the mound (Loucks 1976) It has been sugge sted that the burials were laid out in a radial pattern (Luer and Almy 1987) Sherds found in and near the mound were primarily sand tempered plain wares suggesting a Deptford occupation prior to the Alachua phase (Loucks 1976) Radiocarbon dates taken fro m the mound date to 1360 80 BP, 1210 60 BP, and 1020 100 BP support the Hickory Pond Phase of the Alachua tradition (Schofield 2003) Browne Site 5 (8DU62) Brown Site is a sand burial mound located in Duval County. The mound was excavated by William Sears in 1955 and 1957 and the more than 40 burials were excavated from two trenches placed in the mound. The burials include several apparent mass burials and some cremati ons along with apparent discrete burials. No other provenience information is known about the burials. Pottery analysis suggests that the mound was used during the Weeden Island I St. Johns Ib time periods (Sears 1957) A copper disc suggests long distance trade (Florida Master Site File, Browne Site 5). Santa Rosa Mound or the Fort Walton Temple Mound (8OK6) Fort Walton Temple Mound was a mound built in alternating layers of soil and shell located along Santa Rosa Sound in Okaloosa County in the Panhandle region of

PAGE 94

94 Florida. The site was excavated by George Sternberg, a surgeon with the United States Army, prior to 1875. Artifacts found are described as rude pott ery, bone awls, a single shell tool, along with a hemispherical piece of granite and a biconcave disc of chalcedony (Sternberg 1875) Sternberg (1875) describes the burials as complete, suggesting primary burials in the upper most layers of the mound. However, there is no additional information available about the provenience of the burials examined in this and soil layers and he described primary extended and secondary burials in the upper levels of the mound. The pottery that Moore collected suggests the burial s are associated with the Fort Walton culture (Willey 1949) Excavations by Willey (1949) in the nearby village site evidenced occupation from the Deptford period through the Fort Walton phase.

PAGE 95

95 Table 4 1 Site Information Site Site # Cultural association Sample Size Radiocarbon Dates (BP) Source Henderson AL463 Alachua 6 1360 80 Schofield 2003 1020 100 Gauthier BR193 Archaic 25 4340 170 Maples 1987 1600 190 Windover BR246 Archaic 63 9,530 110 Doran 2002 6,990 70 Dunwody CH61 Caloosahatchee 5 1290 40 Gold 2006 1150 40 Crystal River CI1 Weeden Island 29 2520 60 Katzmarzyk 1998 1620 40 McKeithan CO17 Weeden Island 2 1660 80 Milanich et al. 1997 1470 70 Bay West CR200 Archaic 7 6,780 135 Beriault et al. 1981 5,500 80 Hughes Island DI45 Weeden Island 7 None Bird Island DI52 Archaic 6 4570 110 Stojanowski and Doran 1998; Yates 2000 3630 70 Browne Site 5 DU62 Weeden Island / St. Johns Ib 6 None Hutchinson Island MT37 Glades 12 None Bayshore Homes PI41 Weeden Island or Manasota 27 3380 40 Austin et al. 2012 2230 30 Bay Pines PI64 Weeden Island or Manasota 5 None Manasota SO1292 Manasota 21 1830 90 Florida Master Site File 1800 80 Casey Key SO17 Manasota 8 None Little Salt Springs SO18 Archaic 6 6,180 95 Clausen et al. 1979 5,220 90 Warm Mineral SO19 Archaic 3 10,000 200 Clausen et al. 1975 8,920 190 Palmer SO2 Manasota 44 2350 110 Bullen and Bullen 1976 1100 105 Venice Beach Complex SO26 Manasota 1 None Yellow Bluffs SO4 Manasota 2 2010 40 Luer and Hughes 2011 1770 40 Sarasota Bay Mound SO44 Manasota or Safety Harbor 2 None Fort Walton Temple Mound OK6 Fort Walton 9 None

PAGE 96

96 Table 4 1. Continued Site Site # Cultural association Sample Size Radiocarbon Dates (BP) Source Safety Harbor Mound PI2 Safety Harbor 16 1150 40 Stojanowski and Johnson 2011 Highland Beach PB11 Glades 30 None Margate Blount Mound BD41 Glades 6 None Captiva Island LL57 Caloosahatchee 46 1590 40 Stojanowski and Johnson 2011 Republic Groves HR4 Archaic 9 6,520 65 Wharton et al. 1981 5,745 105 Table 4 2 Measurement taken Maximum length (g op) Nasal height (n ns) Chin height (id gn) Maximum breadth (eu eu) Nasal breadth (al al) Body height at Mental foramen Bizygomatic breadth (zy zy) Orbital breadth (d ec) Body thickness at Mental foramen Basion Bregma (ba b) Orbital height Bigonial width (go go) Cranial base length (ba n) Biorbital breadth (ec ec) Bicondylar breadth (cdl cdl) Basion Prosthion length (ba pr) Interorbital breadth (d d) Minimum Ramus breadth Max. Alveolar breadth (ecm ecm) Frontal chord (n b) Maximum Ramus breadth Max. Alveolar length (pr alv) Parietal chord (b l) Maximum Ramus height Biauricular breadth (au au) Occipital chord (l o) Mandibular length Upper facial height (n pr) Foramen magnum length (ba o) Mandibular angle Minimum frontal breadth (ft ft) Foramen magnu m breadth Mastoid length [2] (Droessler) Upper facial breadth (fmt fmt) Mastoid length (Moore Jansen et al) Mastoid breadth Occipital condyle length Occipital condyle breadth Orbital breadth at Nasion Nasion to fmt Based on Buikstra and Ubelaker 1994 and Droessler 1981. Definitions can be found in Appendix A. Table 4 3 Measurements used. Maximum length (g op) Nasal height (n ns) Foramen magnum breadth Maximum breadth (eu eu) Orbital breadth (d ec) Occipital condyle length Basion Bregma (ba b) Orbital height Occipital condyle breadth Cranial base length (ba n) Frontal chord (n b) Nasion to fmt Biauricular breadth (au au) Parietal chord (b l) Orbital breadth at Nasion Minimum frontal breadth (ft ft) Occipital chord (l o) Mastoid breadth Upper facial breadth (fmt fmt) Foramen magnum length (ba o)

PAGE 97

97 CHAPTER 5 RESULTS AND DISCUSSION The multivariate statistical analyses discussed in the previous chapter were performed to explore the patterns of morphological variatio n between the samples from each site and associated culture within pre Contact Florida. These analyses were performed in an iterative fashion to ascertain the source of variation within each group and the craniometric variables associated with that observe d variation. The initial analyses focused on the overall patterns of variation within all of the samples followed by the analysis of patterns of variation within certain sub samples including the sample populations from sites labeled Archaic, Weeden Island Manasota, cultures linked to the Weeden Island or Manasota cultures, and South Florida. First Principal Component Analysis (PCA) was used to explore the variability of the samples being analyzed. The initial analysis of the patterns of variation is important because in Florida, particularly during the Archaic, evidence suggests that cemeteries wer e places for different groups to come together for burial purposes and therefore may not represent a single population (Randall and Sassaman 2010) High levels of variation within a site would suggest that the individuals sampled were from fluid population s in which mates were coming from various groups, or that those populations buried at the cemeteries represented individuals from various discrete populations that had little biological interaction, but shared common burial locations and customs. L ower lev els of variation would suggest that the populations were relatively homogenous with mates coming from selected populations and that those buried at a cemetery shared both cultural and biological relationships.

PAGE 98

98 Second, when patterns of variation were found using PCA Discriminant Function Analysis (DFA) was used to differentiate the populations followed by R matrix analysis to calculate the biological distance and gene flow. Pearson and Spearman tests were used to test for possible temporal and geographic co rrelations Appendix C contains the geographic and temporal distances between sites used for the correlation analyses. Each analysis began with all samples associated with the sites und er investigation. A s patterns emerged populations were removed from the analysis to further examine structure and to elucidate causal factors accounting for variation and distances between sample populations For each analysis, the amount of variability explained by the test, what craniometric variables contribute d to that va ria bility, and gene flow direction were examined to discern patterns present within the data Observed patterns guided subsequent testing of samples within each set of analyses. The results for each test conducted are included in Appendix D O verall resul ts for each analysis are discussed below with a focus on results that were either statistically significant or biologically informative. All Sites Results The first set of analyses included all sample si tes (N=404) to determine if there were underlying patterns to overall variation. All PCA tests conducted ( KMO 0.804) were s ignificant suggesting that PCA wa s an appropriate method for analyzing these data. In all tests the total variance explained by the first two principal components was L ambda test was significant. The percentage of variance explained was between 42.4% and 47.7% The cross validati on results were approximately 41% for all tests. For the R matrix tests the

PAGE 99

99 unbiased F ST decreased from the first to the third test and the percent variance explained by the first two eigenvectors was low between 50% and 58 % for each test. In the first PC A test with all the sites included, the craniometric variables with the highest loading (> 0.7) in the first component were minimum frontal breadth, upper facial bread th, and nasion to fmt. The variable with the highest loading in the second component was parietal length. Figure 5 1 depicts a graph of the first two components that the Archaic populations group together at the top of the graph. In the DFA, the variables that contributed to the first and second functions were maximum breadth and orbital bread th respectively. Figure 5 2 s h ows a graph of the first two functions with Warm Mineral Springs clearly distinct from the other populations All other Archaic populations except for Bird Island, group together. Lastly, the Woodland and Mississippian popul ations and the Archaic Bird Island population group together. In the R Matrix analysis of all the sites the unbiased F ST is 0.161536, suggesting a moderate level of variability within these populations Again Warm Mineral Springs is clearly distinct from the all of the populations and the Archaic populations group more closely (Figure 5 3). The largest biological distances are found between Warm Mineral Springs and the other populations (Table 5 1 ). In the comparison of the biological distances determined by R matrix analysis and time a significant positive correlation was found between the two variables (i.e. with increased time there is increased biological distance ) However, when the bio logical distances associated with Warm Mineral Spring s were remov ed, the correlation disappeared. Additionally, when the distances between Warm Mineral Springs and the

PAGE 100

100 other sites were examined independently no significant correlation was found (Table D 7) For the second test all Archaic sites, except for Bird Island, were removed and no significant pattern was noted. For the third test Bird Island was removed. The variables with the highest loading (> 0.7) in the first component of the PCA analysis were minimum frontal breadth, u pper facial breadth, and nasion to fmt. The variable with the highest loading in the second component was maximum length. A separation of the Weeden Island and Manasota populations became apparent in the graph of the first two principle components. It als o became apparent that the Safety Harbor population is more closely related to the Manasota populations, while the St. Johns, Alachua, and Fort Walton populations were more closely related to the Weeden Island populations; finally no apparent pattern was e vident for the South Florida populations analyzed (Figure 5 4). In the DFA, no particular variables contributed to the first or second functions. Again, the patterning between populations is similar to that found using PCA (Figure 5 5 ). In the R Matrix an alysis of all the sites, the unbiased F ST is 0.104228, suggesting a moderate level of variability within these sites, however, decreased variation was observed compared to the previous analys e s, suggesting a greater similarity between these populations tha n those populations involved in the previous analyses. The biological distances and eigenvalues further support a separation between the Weeden Island and Manasota populations. No significant correlations were found between biological distance and time.

PAGE 101

101 D iscussion Overall, these findings suggest that there were at least two migrations into the peninsula after the Paleoindian period, the first populating the Archaic sites, except Bird Island, and the second producing the Woodland/Mississippian populations. In addition, these findings suggest that there are patterns to the biological variation found within these populations. This two migration hypothesis is supported by the biological relationships/distances found between Warm Mineral Springs and the other p opulations. These findings indicate that there was most likely a migration event into the peninsula in th e Early to Middle Archaic. This hypothesis is supported by the archaeological evidence presented by Faught and Waggoner (2012) suggesting that there w as an absence of people in Florida between the Paleoindian/Early Archaic and the Middle Archaic based on radiocarbon dates and changes in stone tool technology. They suggest that the individuals associated with the Middle Archaic are most likely the result of a mid Holocene migration into the peninsula. The first analysis also suggests that the Archaic populations, except for Bird Island, are closely related to each other and are biologically distinct from later Woodland and Mississippian populations Thes e results also provide evidence of a second migration of new peoples into the peninsula in the late Archaic. Bird Island, a L ate Archaic site, appears to be part of this second migration based on its close relationship to the later populations. This infere nce is supported by c hanges in burial patterns and material culture from the Middle Archaic to the Woodland periods i.e. a shift away from freshwater burials (Milanich 1994) and the introduction of soapstone vessels and pottery ( Jeffries 2004; Sassaman 2006 ). The finding of Haplogroup X in the genetic analysis on the Windover

PAGE 102

102 Pond population (Smith et al. 2002) also appear to support this conclusion, as this h aplogroup is not found among later Native American populations. Temporal trends must be conside red given the large amount of time covered by these sampled populations, in particular one would expect that if these individuals were from a single population the samples associated with the various sites/cultures would become increasingly similar throug h time, or have increasingly smaller biological distances, due to the homogenizing effects of gene flow over time (Konigsberg 1990) This is clearly not the case as the biological distances between the Warm Mineral Springs sample and the other sites are no t correlated with time. Second, if the variation observed was due to secular trends one would expect to see more of a gradation over time in the relationship of sites, with sites that are similar temporally being closer. This does not appear to be the cas e in these data as there were clear separations noted between the Weeden Island and Manasota populations and between the Fort Walton and Safety Harbor populations despite their broadly similar time frames. Therefore both the biological and archaeological evidence appear s to support the interpretation that that these individuals were not all from the same stock population and that there were cultural ly defined barriers that prevented gene flow. Archaic Results The populations sampled from the Archaic are f rom: Warm Mineral Springs, Windover Pond, Little Salt Spring, Bay West, Republic Groves, Gauthier, and Bird Island, with a total of 119 individuals. A ll PCA tests conducted ( KMO > 0.807 ) were significant suggesting that PCA is an appropriate method for ana lyzing these data. In all cases the total variance explained by the first two principal components was low,

PAGE 103

103 between 42% and 46%. For all DFA test s Lambda test was significant. With each test the percentage of variance explained by the first two functions improved from 59.6% to 86.7%. The percent age of samples correctly classified in cross validation was low, between 50% and 60%. For the R matrix tests the unbiased F ST decreased steadily with each test and the percent variance explaine d by the first two eigenvectors was between 80% and 88% for each test. In the first PCA test with all Archaic sites included, the variables with the highest loading (> 0.7) in the first component were minimum frontal breadth, u pper facial breadth, and nasion to fmt. The variable with the highest loading in the second component was occipital height. Figure 5 6 depicts a graph of the first two components that Warm Mineral Spring s is separated from the rest of the Archaic samples, Little Salt Spring and Bay West fall to the left side of the cluster, and Bird Island is situated to the right side of the cluster, suggesting a pattern to the variation observed between these populations. In the DFA, the variables that contributed to the first function were basion bregma, basion nasion, nasion nasospinal e and foramen magnum length, and those contributing to the second function were biauricular breadth, foramen magnum breadth, and occipital height It was also noted that Warm Mineral Sprin gs was correctly classified 100% of the time in the cross validation, followed by Bird Island (83.3%) and Republic Groves (77.8%) supporting the observation that these populations are biologically quite distinct from the other populations Figure 5 7 shows the graph of the first two functions which shows a clear separation of Warm Mineral Springs and Bird Island from the other populations, and Republic Groves is also separated, albeit slightly from Windover, Bay West, Little Salt Spring, and Gauthier. In t he R Matrix analysis the

PAGE 104

104 unbiased F ST is 0.2987, suggesting a high level of variability within the Archaic samples. Bird Island exhibits the highest level of variance at 1.169. Bird Island and Warm Mineral Springs both had a high positive residual of 0.476 suggesting higher than average external gene flow. Figure 5 8 depicts a graph of the first two eigenvalues show ing a clear distance between of Warm Mineral Springs, Bird Island, and Republic Groves from all other Archaic populations Warm Mineral Springs shows the largest biological distances between it and the other populations and Bird Island and Republic Groves have the n ext highest distances (Table 5 2 ). The Pearson correlation test found a significant correlation between biological distance and time when the Warm Mineral Springs distances were included; however the correlation disappeared when those distances were removed. No significant correlations related to geography were observed (Table D 7) In PCA, if one population is significantly different or separated from the other populations it can alter the PCA analysis, therefore the most distinct populations can be removed to allow analysis of the other populations. Therefore Warm Mineral Springs was removed for the second analysis In this analysis the variables with the highest loading (>0.7) in the first component were minimum frontal breadth, upper facial breadth, and nasion to fmt The variable with the highest loading in the second component was basion to bregma. In the DFA, the variables that contributed to the first function were foramen magnum breadth and length, and the second function were basion bregma, basion nasion, biauricular breadth, and nasion to fmt Also, Bird Island (83.3%) and Republic Groves (77.8%) again classified correctly th e most often in cross validation. Figure 5 9 portrays a graph of the first two functions with a clear separation

PAGE 105

105 of Bird Island from the other populations, and Republic Groves offset slightly from Windover, Bay West, Little Salt Spring, and Gauthier. T he s econd R Matrix analysis indicates there is less variation in these samples than in the previous test ( unbiased F ST is 0.1503 ) Bird Island exhibits both the highest level of variance (1.218) and the highest positive residual of 0.519. Figure 5 10 shows a graph of the first two eigenvalues evidencing a clear distancing of Bird Island and Republic Groves from the other sites. The Bird Island and Republic Groves samples have the largest biological distances between them and the other sites. Gauthier is approximately equidistant, biologically, from Bay West, Little Salt Spring, and Windover; while Little Salt Spring is closer to Windover than Bay West In the next analysis Bird Island was removed as it was the most distinct from the other sites. The variables with the highest loading in the first component of PCA were minimum frontal breadth, upper facial breadth, orbital breadth, and nasion to fmt The variable with the highest loading in the second component was basion to bregma. It can be seen in F igure 5 11 that Republic Groves falls to one side of the cluster. In the DFA, the variables contributing to the first function were basion bregma and foramen magnum breadth and the variables for the second function were nasion nasospinale, orbital height, and maximum breadth. Republic Groves (77.8%) was classified correctly in cross validation most often suggesting a distinction between this population and the others. T he third R Matrix analysis ( unbiased F ST is 0.0986) indicates that the overall variabilit y has decreased Gauthier exhibits the highest level of variance at 1.041 and a high positive residual of 0.174. Figure 5 12 depicts the first two eigenvalues show ing a

PAGE 106

106 clear separation of Repu blic Groves from the other samples Republic Groves has the larg est biological distances between it and the other sites (Table 5 3 ). The final analysis was conducted with Republic Groves removed. The variables with the highest loading (> 0.7) in the first component were minimum frontal breadth, upper facial breadth, o rbital breath, and nasion to fmt The variable with the highest loading in the second component was basion bregma. Figure 5 13 portrays significant overlap between the samples in the first two components In the DFA, the variables contributing to the first component were foramen magnum breadth, orbital height, maximum breadth, and basion bregma, while the second function was influenced primarily by orbital breadth, minimum frontal breadth, upper facial breadth, orbital breadth at nasion, basion nasion, pari etal length, frontal height, nasion to fmt mastoid breadth, and biauricular breadth. Figure 5 14 graphically depicts the first two functions showing that the populations from these four sites are separated with Windover and Gauthier grouping closer togeth er and Bay West and Little Salt Spring closer together, biologically. In the final R Matrix analysis the unbiased F ST is 0.0460, which is indicative of a low level of variability. All of the sites are approximately equidistant from the regional centroid b ased on r ii values. Discussion These results suggest that when all of the Archaic samples are included in the analysis there is a high level of variability and there are some underlying pattern s to that variability. Warm Mineral Springs and Bird Island are most distinct suggesting that the individuals buried at these sites represent different biological populations from those represented at the other sites. Republic Groves appears similar to the other four sites, but on closer analysis appears distinct. I n the PCA analyses the variables that

PAGE 107

107 contribute d the most to differentiating sample populations include upper facial breadth, minimum frontal breath, and nasion to fmt all of which are measures of the width of the upper facial region. In the discriminant function analysis the height of the skull and face (basion bregma and basion nasion) contribute to the first two functions most often, followed by the shape of the foramen magnum (foramen magnum length and width), and the width of the skull (maximum breadth and biauricular breadth). The most obvious result of these analyses is that Warm M ineral Springs is significantly biologically distinct from the other Archaic populations T his can be seen in its placement on th e graphs of the PCA and DFA results and its biological distance determined by R matrix analysis (Figures 5 6 and 5 7; Table 5 2 ). The rationale for this is suggested in both the archaeology of the burials and the associated radiocarbon dates. Evidence sugg ests that the burials were placed on the ledge within the spring when the water levels were lower (Cockrell and Murphy 1978) so these were not true water burials like those found at Windover, Little Salt Spring, Republic Groves, and Bay West. Radiocarbon dates, conducted on associated materials and not the bone itself, indicate that the Warm Mineral Springs remains may date to the late Paleoindian period (Clausen et al. 1975) The statistical analysis results reported here suggest that these remains are biologically distinct from the other Archaic remains, further support ing the inference that these remains were associated with the Paleoindian period The variation / distance noted between these Paleoindian remains and the Archaic samples suggests that thes e two temporal groups were biologically distinct lineages isolated from each other for many generations This suggests that the Archaic individuals migrated into Florida in the early to middle Archaic period. This hypothesis is further

PAGE 108

108 supported by archae ological evidence presented by Faught and Waggoner (2012) suggesting that there was an absence of people in Florida between the Paleoindian/Early Archaic and the Middle Archaic and that those individuals associated with the Middle Archaic are most likely eninsula. The interpretation of Warm Mineral Springs as Paleoindian remains however is supported by craniometric analysis of only three individuals Thus, this supposition is less robust. The second significant result of these findings is that the Bird Island population also appears to represent a group of individuals distinct from the other Archaic populations included in these samples (Figures 5 9). Bird Island is a late Archaic burial s ite that differs from the other Archaic sites in that it is neither a pond burial site nor is there a n associated large body of freshwater. Bird Island is also associated with a large collection of soapstone vessel s (Lovejoy 1985; Yates 2000) which to date have not been found at any of the other Archaic sites in this sample. In addition, the raw material for these vessels is not found in Florida, the closest source being in Georgia or Alabama (Yates 2000) This suggest s that the individuals associated with the burials at Bird Island eit her migrated from present day Georgia or Alabama area bringing their vessels with them or they were in contact with individuals from that area through trade networks. Given the distinct biological signature of this sample in comparison to the other Archaic samples it s uggest s that these individuals again represented a different biological lineage than the other Archaic samples This supports an additional migration into the peninsula during the late Archaic or that there was a significant shift in mate exc hange networks resulting in a new biological signature. Overall, the biological analysis and the changes in mortuary practice and material culture appear to offer strong evidence that

PAGE 109

109 there was a migration of new peoples into the peninsula in the late Arch aic Although these results are promising, the small sample size (N=6) means that these results are preliminary and must be interpreted with caution The individuals interred at the remaining Archaic sites ( Windover, Little Salt Spring, Bay West, Republic Groves, and Gauthier ) appear to represent a sample of somewhat closely related populations (Figure 5 11). Given the temporal gaps between these sites (none were occupied at the same time as determined by 14 C dating ) this pattern most likely reflects broad ly inclusive ancestor descendent relationship s All of these sites share very similar burial practices, and aside from Gauthier, all of the burials were interred in the water and were most likely staked down with some type of burial shroud (Beriault et al. 1981; Clausen et al. 1979; Doran 2002; Wharton et al. 1981) Gauthier is similar in that it was located next to a pond (Carr 1981) so association with freshwater seems significant to the groups burying the dead in this location. Republic Groves, however, may have a slightly different biological influence than the other populations it has the largest biological distances from the oth er sites (Figure 5 12; Table 5 3 ). This could be due to variations in trade networks between the sites because some of the sites a re on the East coast (Windover and Gauthier) and others are on the West Coast (Little Salt Spring Republic Groves, and Bay West). However, this does not seem to account for all of the variation since Republic Groves is most closely related to Windo ver despite the largest geographic and temporal distances, and Republic Groves is more distinct from Little Salt Spring and Bay West despite being both geographically and temporally closer. In addition, Gauthier is closer biologically to Windover, Bay West and Little Salt Spring than Republic Groves, despite larger temporal and geographic

PAGE 110

110 distances These results suggest t hat there is some, as of yet, unknown social interaction impacting the Republic Groves population differently than the other Archaic po pulations. The individuals at Republic Groves may have been interacting with a new group resulting in a new mate pool that the earlier and later populations did not interact with or the Republic Groves population insulated itself from other groups resultin g in genetic drift, shifting its biological signature. Nonetheless, sampling may have also affected the results because it is estimated only 1% of the Republic Groves burials were excavated (Wharton et al. 1981) The last analysis ( Windover, Bay West, Lit tle Salt Spring, and Gauthier ) suggests an East West division of these Archaic sites. G raph s of the discriminant function and R matrix analyses (Figures 5 13 and 5 14) show Windover and Gauthier closer and Bay West and Little Salt Spring closer as well Th ese results suggest some type of gene flow pattern that was consistent between East and West Coast sites. Perhaps this is due to variation in the trade routes on the East and West Coasts because trade would have resulted in contact with distant groups and may have supplied a likely source for mates. Another possibility is mate exchange during ceremonial gatherings, if these sites were places to come together to bury the dead; however based on mortuary practices these groups may have had very similar ceremo nial activities. Thus it could logically be assumed that participants would have been somewhat consistent through time, which accounts for the observed equivalent biological distances but not the apparent East/West distinction. In summary, the results fo r the analysis of the Archaic samples represented by individuals from Warm Mineral Springs, Windover, Little Salt Spring, Bay West,

PAGE 111

111 Republic Groves, Gauthier, and Bird Island indicate that there were at least two migration events associated with this overa ll time period ( 1) between the Paleoindian / Early Archa ic and Middle Archaic periods and (2) during the Late Archaic. The individuals from Windover, Little Salt Spring, Bay West, and Gauthier seem to be closely related, while the individuals from Republic Groves are loosely related, suggesting a parent descendent group relationship and a stable population present and widely dispersed across the peninsula during the Archaic. However, there were variations between these sites th at resulted in slight biological differences that need to be investigated further by incorporating larger sample sizes and additional analysis of the associated archaeology. Weeden Island / Manasota Weeden Island/Manasota r esults The sampled populations (N= 152) Complex were McKeithan, Crystal River, and Hughes Island and those from Weeden Island like or Manasota sites were Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Venice Be ach Complex, and Yellow Bluffs. M ost of the PCA tests conducted ( KMO > 0.728 ) were significant suggesting that PCA was an appropriate method to analyze the data, however the KMO was low ( < 0.5) for two of the tests. In all cases the total variance explain ed by the first two principal Lambda test was significant. For each test the percentage of variance explained by the first two functions was between 54% and 100%. Cross validat ion results were between 45% and 65%. For the R matrix tests the unbiased F ST decreased with each test and

PAGE 112

112 the percent variance explained by the first two eigenvectors was between 65% and 100%. In the first PCA test with all samples included, the variabl e with the highest loading (> 0.7) in the first component was maximum length. The variable with the highest loading in the second component was maximum breadth Figure 5 15 shows the graph of the first two components with the populations from the classic W eeden Island sites, Crystal River, Hughes Island, and McKeithan clustering to one side; with respect to Weeden Island like or Manasota populations, Yellow Bluffs clusters with Weeden Island populations while, Palmer Mound and Manasota Key Cem etery form a separate cluster on the other side. In the DFA, no variables were found to contribute significantly to the first function, however, maximum length contributed to the second function It was also noted that the Crystal River sample (71.4%) clas sified correctly in cross validation most often suggesting that this population is distinct fr om the other groups. Figure 5 16 depicts the graph of the first two functions with Venice Beach Complex, Bayshore Homes, Manasota Key Cemetery, Casey Key, and Palmer, forming a distinct cluster from a Crystal River, McKeithan, Yellow Bluffs, and Hughes Island cluster The Dunwody population is separated somewhat from the other groups, b ut it is closer to the Weeden Island like/ Manasota culture cluster. In the R Matrix analysis, Venice Beach Complex was not used in the analysis as only one individual was sampled from this site. T he unbiased F ST ( 0.137490 ) indicated moderate variability w ithin these populations Yellow Bluffs (1.690), McKeithan (1.430), and Hughes Island (1.111) exhibit the highest levels of variance. Yellow Bluffs also had a high positive residual of 0.766, suggesting higher than average external gene flow. Figure 5 17 is a

PAGE 113

113 graph of the first two eigenvalues again show ing a cluster of the classic Weeden Island sites with Yellow Bluffs and a cluster of Manasota sites. B iological distances are presented in Table 5 4 A significant correlation was found between biological d istance and geography for these populations. In the second set of analyses McKeithan, Venice Beach Complex, and Yellow Bluffs were removed due to their small sample size one or two individuals, to determine if their variability significantly impacted th e analysis. T he variable with the highest loading (> 0.7) in the first component was maximum length and the variable with the highest loading in the second component was maximum breadth. In the DFA, the variable that contributed to the first function was f rontal height and the variable contributing to the second function was maximum length It was noted that Crystal River (78.6%) again classified correctly in cross validation most often. There was little change in the PCA or DFA graphs from the first to the second analysis. In the R Matrix analysis the unbiased F ST ( 0.098010 ) suggests a decrease in the level of variability within these sites. The variance is approximately equal in all the sites, though Hughes Island has the highest (1.138) The Hughes Island sample also had a high positive residual of 0.194, while the Bay Pines sample ha d a high negative residual ( 0.238), suggesting respectively higher and lower than average external gene flow for these two populations Figure 5 18 shows a graph of the first two eigenvalues with the distinct clustering o f the classic Weeden Island populations and the Manasota populations Weeden Island r esults Island samples separately from the Manasota samples might provide clarify patterns

PAGE 114

114 regarding the internal structure between and within these two cultures. The classic Weeden Island results will be discussed first. Based on site associations fo und in the PCA and DFA a nalyses, it was determin ed that the sites used for this analysi s (N=44) Weeden Island sites ( Crystal River, McKeithan, and Hughes Island ) and the Weeden Island like/Manasota sites of Yellow Bluffs and Bay Pines However, it was determined that PCA was not a suitable test for these data because the KMO score was 0.433 and even when McKeithan and Yellow Bluffs were removed due to their small sample size s the KMO score was only 0.482. Therefore the PCA analysis is not discussed fur ther for these two tests However, i t was determined that DFA and R matrix were suitable for Lambda test was significant. For the two tests the percentage of variance explained by the first two functio ns rose from 79% to 100%. Cross validation results were between 59.1% and 65%. For the R matrix tests the unbiased F ST decreased and the percent variance explained by the first two eigenvectors rose from 88.1% to 100%. In the first analysis, the variables that contributed to the first function in the DFA were frontal height and occipital condyle length and the second function were nasion to fmt, maximum length, upper facial breadth, and mastoid breadth. In both the DFA and R matrix analyses (Figures 5 19 a nd 5 20 ), individuals from McKeithan and Yellow Bluffs separated from those of Crystal River, Hughes Island, and Bay Pines. In the R Matrix analysis of all samples produced an unbiased F ST of 0.364732, suggesting high variability within these populations. Hughes Island, Crystal River, and Bay Pines all had high positive residuals, while McKeithan and Yellow Bluffs had high negative residual s

PAGE 115

115 suggesting higher and lower than average external gene flow for these two sites respectively. The largest biological distances were observed in the comparisons of the Yellow Bluffs and McKeithan populations The smallest biological distance was between Bay Pines and Hughes Island (Table 5 5 ). A significant correlation between biological distance and geography was found (Table D 6) In the second analysis with McKeithan and Yellow Bluffs removed the variables that contributed to the firs t function of DFA were nasion to fmt maximum length, upper facial breadth, occipital condyle length, and frontal height and the second function were all of the other variables It was also noted that Crystal River (82.1%) classified correctly in cross validation most often. In the R Matrix analysis for these populations, the unbiased F ST ( 0.115085 ) indicated a decrease in the amount of biological variation within these populations. The variance is approximately equal in all the sites, al though Hughes Island has the highest level (1.199) In addition, Hughes Island had a high positive residual of 0.227, while Bay Pines had a high negative residual ( 0.295), intima ting higher and lower than average external gene flow for these two sites respectively. Figure 5 2 1 is a graph of the first two eigenvalues showing a separation for the populations in this analysis The l argest biological distances are found between the Crystal River and the Hughes Island / Bay Pines samples, with Hughes Island and Bay Pines appear ing close biological ly (Table 5 6 ). Manasota r esults T he samples (N=115) from the sites labeled as Weeden Islan d like or Manasota culture were compared for patterns of variability. These sites included Palmer Mound, Casey Key, Manasota Key Cemetery, Dunwody, Venice Beach Complex, Bayshore Homes, Bay Pines, and Yellow Bluffs. Bay Pines and Yellow Bluffs were include d in

PAGE 116

116 these analyses because the se two sites or components of the se sites have been culturally identified as Manasota. T heir relationship to other Manasota sites is of interest, just as their relationship against populations is o f interest All PCA tests conducted ( KMO > 0.728 ) were s ignificant suggesting that PCA wa s an appropriate method for analyzing these data. In all cases the total variance explained by the first two principal components was low, between 39% and 41%. For all DFA test s Lambda test was significant. With each test the percentage of variance explained by the first two functions improved from 54.6% to 81.7%. The cross validation results were approximately 50% for all tests. For the R matrix t ests the unbiased F ST decreased steadily with each test and the percent variance explained by the first two eigenvectors was b etween 80% and 91% In the first PCA test, the variables with the highest loading (> 0.7) in the first component were minimum frontal breadth, upper facial breadth, and nasion to fmt The variables with the highest loading in the second component were maximum length, nasal height, and orbital height. Figure 5 2 2 depicts th e graph of the f irst two components with Dunwody clustered at the bottom of the group as is Yellow Bluffs, while the other populations are more generally dispersed In the DFA, the variables that contributed to the first function were nasal height, maximum length, and par ietal width and orbital height contributed to the second function. Figure 5 23 portrays the graph of the first two functions with Dunwody separat e at the top of the graph and Yellow Bluffs separated at the bottom but Yellow Bluffs is close to Bay Pines. I n the R Matrix analysis the unbiased F ST ( 0.150160 ) suggests moderate variability within these populations Yellow Bluffs exhibits the highest level of variance at 1.791, but this is most

PAGE 117

117 likely due to small sample size (N=2) The other variances are roug hly equal. All of the sites, except Yellow Bluffs and Dunwody, exhibit negative residuals. This indicated lower than expected external gene flow Figure 5 24, a graphical representation of the first two eigenvalues shows a tight grouping of Palmer Mound, Casey Key, Bayshore Homes, and Manasota Key Cemetery, with Dunwody, Yellow Bluffs, and Bay Pines separating. The largest biological distance is noted between Dunwody and Yellow Bluffs, followed by the distances between Yellow Bluffs and the other sites. Th e biological distances between Palmer and Manasota and C asey Key are smallest (Table 5 7 ). No significant correlations between biological distance and time or geography were found. In the second set of tests, Venice Beach Complex and Yellow Bluffs were rem oved, because these are the smallest samples; subsequently, D unwody was removed after tests showed that it did not significantly change the analysis results (Appendix D). I n the analysis of Palmer Mound, Casey Key, Manasota Key Cemetery, Bayshore Homes, a nd Bay Pines, the variables with the highest loading (> 0.7) in the first component were upper facial breadth and nasion to fmt The variable with the highest loading in the second component was occipital height In the DFA, the variables that contributed to the first function were nasal height, maximum length, parietal width, frontal height, orbital height, and occipital condyle breath, while occipital height and basion bregma contributed to the second function. Fig ure 5 25 depicts the graph of the first t wo functions with Bay Pines separate from the other sites and Palmer and Casey Key grouping closely with Manasota Bayshore Homes is also close to this latter group. In the R Matrix analysis of all sites the unbiased F ST ( 0.054757 ) indicated a significant

PAGE 118

118 drop in variability within these sites. The variances within the sites are roughly equal as are the residuals. Figure 5 26 shows the graph of the first two eigenvalues show ing a tight grouping of Palmer Mound, Casey Key, and Manasota Key Cemetery, with Ba y Pines and Bayshore Homes distinct The largest biological distance is noted between Bay Pines and the other sites and again the smallest biological distances are observed between Palmer and Manasota and Casey Key (Table D 6 ). For the next test, Bay Pines was removed from the analysis. In this analysis, the variables with the highest loading (> 0.7) in the first component were upper facial breadth and nasion to fmt The variables with the highest loading in the second component were nasal h eight and orbital height. It can be seen in the graph (Figure 5 27 ) of the first two components that there is significant overlap between all samples In the DFA, the variables that contributed to the first function were maximum length, foramen magnum leng th, parietal length, biauricular width, occipital condyle breath, and orbital height, while nasal height occipital condyle length, mastoid breadth, frontal height, and maximum breadth contributed to the second function. In the R Matrix analysis for all pop ulations, the unbiased F ST is 0.025861, suggesting a decrease d level of variability within these populations Again, the variances and residuals are roughly equal. Figure 5 28 shows the graph of the first two eigenvalues with Palmer and Casey Key close, an d Manasota closer to these sites than Bayshore Homes. The largest biological distance is noted between Bayshore Ho mes and the other sites (Table 5 8 ). Discussion A moderate level of variation was seen in the analysis with all of the Weeden Island and Weede n Island like/ Manasota samples and variability decreased as samples were systematically removed from analysis Co mpared to the Archaic results the

PAGE 119

119 samples in this analysis exhibited more variation than the Archaic group of samples from Windover, Little Sa lt Spring, Bay West, and Gauthier. The only sub sample with less variation than the se core Archaic sites is the sub sample from Palmer, Manasota Key Cemetery, Casey Key, and Bayshore Homes. Based on the previously discussed interpretation for migration eve nts into the peninsula of Florida during the Archaic these results suggest that there were no significant migration events into the peninsula (bases on the sites in this study) ; however the biological variation was higher across the peninsula suggesting either higher levels of gene flow into the populations than that observed for the ancestor descendent population s discussed in the Archaic or that there were more barriers to gene flow between these populations resulting differentiation of the populations and elevated levels of variation The variation seen within the total population samples relates to overall skull shape as represented by the maximum length and breadth of the skull, in the sub samples variation relates to facial dimensions typically represented by minimum frontal breadth, frontal height, nasion to fmt and upper facial breadth. These analyses indicate Island and the Weeden Island like/ Manasota populations (Figures 5 16 and 5 17), perhaps indicating that the populations represented by the samples from these sites had very little, if any, interaction with each other. Some of this may be due to temporal separation, but many of these sites were broadly contemporaneous meaning these populations would have had the opportunit y to exchange mates. Furthermore, there is no correlation between biological distance and time. However, geography most likely played a role as there is a correlation between geography and biological distance. It is

PAGE 120

120 also clear that the Crystal River popula tion is di stinct biologically as compared to these populations This is unlikely due to its large sample size because Palmer also has a large sample size and it not correctly identified as often as the Crystal River population. Therefore it can be said t hat the individuals buried at Crystal River are more internally homogenous despite being buried in two locations (Mound G and the burial mound) than other sites. Perhaps this site was used by a particular group of individuals for an extended period of time for burial, while the other sites may have had more population variability resulting in less homogeneity observed in the individuals sampled Another finding of interest in these analyses is the grouping of Bay Pines and 17), as both Bay Pines and Yellow Bluffs have been labeled archaeologically as belonging to the Manasota culture, most likely early Manasota on the Deptford horizon (Luer 2011; Luer and Almy 1982) However, subsequen t analysis demonstrates that Yellow Bluffs is not as closely associated with the Weeden Island samples as initially indicated and Bay Pines straddles the Weeden Island and Manasota samples The initial proximity of Yellow Bluffs and Bay Pines to the Weeden Island sites may be explained by the fact that they both appear to contain a Deptford component based on radiocarbon dates and pottery analysis (Gallagher and Warren 1975; Luer 2011 ; Luer and Hughes 2011 ) which in many areas appears to be the predecessor of the Weeden Island Complex based on similarities between the two cultures in pottery and mortuary practices. In the also appears to contain a Deptford component based on radiocarbon dates (Ka tzmarzyk 1998; Pluckhahn et al. 2010) Therefore, these results may indicate that these populations either interacted

PAGE 121

121 with each other or shared a common mate source during the Deptford period, or they have a common origin prior to the Deptford period, most likely in the Late Archaic migration population. Intriguingly, this close proximity is lost in the subsequent analyses, perhaps due to the fact that in the earlier analyses the biological signature of the core Manasota sample (Bayshore Homes, Palmer, Man asota Key, and Casey Key) influenced the assessment of the Yellow Bluffs and Bay Pines samples, by pushing them towards Crystal River the more similar population, both temporally and biologically However, once the Manasota signal was removed it becomes clear that these two sites were not that closely related to the Crystal River population. Rather it appears that Yellow Bluffs is distinct from both the Weeden Island and Manasota populations, while Bay Pines falls between them. Despite this apparent com mon origin, further analysis shows that these populations deviated and became the subsequent Manasota and Weeden Island populations associated with differing material culture and mortuary practices. Ultimately, Yellow Bluffs and Bay Pines are geographicall y located at the northern limits of the Manasota region, particularly Bay Pines, in an area where contact with outside groups may have been more common. Further, the high negative residua ls found for the Bay Pines sample suggest that lower than expected gene flow, or gene flow from a source no sampled in this analysis. Interpreted another way, Bay Pines was most likely receiving gene flow from one o r more of the populations in this analysis, and the closer biological relationship of Yellow Bluffs and Bay Pines to Crystal River (Figure 5 17 and 5 18), with their shared Deptford component, may support a relationship between these sites or a relationship with a common mate source during the Deptford period. Overall

PAGE 122

122 though, based on the available data and the tests conducted as part of this research, it appears that the Bay Pines population falls and Weeden Island like/ Manasota cultures both geographically and biologically suggesting that this population was either on th e cusp of the subsequent cultural split or it maintained contact with both culture populations through time Unfortunately, the analysis of these results and inferences based on these results should be viewed with caution due to the small sample sizes of t he Yellow Bluffs and Bay Pines samples and due to the poorly understood archaeology of Bay Pines and inadequate information about the provenience of the remains from Crystal River. The relationships of these sites needs to be further explored with addition al samples from these sites and from other temporal and geographic locations. It should also be noted that Bayshore Homes, which is very close geographically to Bay Pines and has Deptford period radiocarbon dates, consistently groups with the Manasota sit es. Austin (2012) and Austin and Weisman ( in press ) have suggested that the upper layer of the Bayshore Homes mound is re deposited from another location based on radiocarbon dates, therefore the remains found in the mound may have been re deposited from a nother location associated with a Manasota culture resulting in the closer relationship of this site to the other Manasota populations. However, nothing in the excavation conducted by Sears (1960) suggests this and he indicated that the burial pattern was consistent with a charnal house. Therefore at this time the inferred close relationship between the Bayshore Homes population and the Manasota samples cannot be adequately explained. A dditional work at Bayshore Homes will clarify its

PAGE 123

123 archaeology and relat ionship to other sites in the Tampa Bay area and the Manasota culture. level of variability among these samples, implying that the Weeden Island Complex as a cultural label does not encompass a single biological population The first explanation for this may be that the Crystal River sample consists of a group from the earlier Deptford period. The alternative explanation is that these sites represent separate mound centers us ed by different groups for burial purposes as proposed by Milanich (2002) Either explanation could be supported by the separation of the McKeithan sample from the other sites, because if a portion of the sample from Crystal River consists of the earlier Deptford population component this component is absent from the McKeithan population resulting in the separation found Alternatively, perhaps both the McKeithan and Crystal River sites were large mound centers that may have been inhabited coevally and the refore different populations were using the se mound centers to bury their dead. The distances found between Crystal River, McKeithan, and Hughes Island lend support to this second theory (Table 5 5) The correlation between biological distance and geography for these populations is weak (p=0.046), but this finding may indicate that the geographic distance between these sites may have played a role in the biological separation or lack of gene flow/excha nge between these populations. In addition, this research suggests that Bay Pines and Hughes Island are more closely related to each other than they are to the Crystal River population. Both of the former sites have small sample sizes, however, which may h ave led to their being forced together by the Crystal River signal. On the other hand, Crystal River is a unique

PAGE 124

124 archaeological site evidenced by its material remains and site structure (Bullen 1951; Pluc khahn et al. 2010; Weisman 1995). Findings in this s tudy suggest that the biological composition of the Crystal River sample population is also distinct from other populations with broadly similar temporality and cultural labels. Crystal River is a multicomponent site and as previously noted in Chapter 5, the skeletal remains sampled were excavated from two different mounds within the site, Mound G and the Burial Mound. Therefore the biological and archaeological relationship of the site itself requires further research before these findings can be meaningf ully compared to other sites. In the analysis of the Manasota populations there was less variation overall than the Weeden Island populations Bay Pines and Yellow Bluffs are the two sites that appear to be most biologically distinct, followed by Dunwody, while the populations sampled from Palmer and Casey Key appear to be closely related (Figure 5 22 ). The association of the Dunwody sample with the other Manasota culture sites sampled is open to interpretation. A nalyses suggest that the Dunwody populatio n may not be biologically as closely related as other Manasota populations in this study. As noted below, the Dunwody population sample seems somewhat related to the South Florida populations examined in this research. However this may be an artifact for t he method used to fill in the missing data. That is, the missing data for Dunwody was filled in through regression analysis with the Captiva Island sample another South Florida site. However, if this did not influence the results and Dunwody is indeed anom alous, it may be due to i t s geographic location and time period. Dunwody is considered to be at the southern limits of the Manasota area and has been dated very late in the Manasota

PAGE 125

125 period (Gold 2006) Considering the geographic location and chronological position of the site, Dunwody shows evidence for increased external gene flow If true, this could support the idea that this area and west central Florida more generally, experienced cultural and biological influences from several spheres. In support, it is noted that by early Spanish contact, Carlos, the Calusa leader reportedly asked Menendez to take him, by ship north along the coast, to Tampa Bay, to the chiefdom of Tocaboga so he could attack Pe rhaps this was the result of a long standing feud between /within two powerful chiefdoms (Milanich 1995) Thus, west central Florida, particularly within present day Manatee, Sarasota and Charlotte counties lying between two warring power s (at least by the very late Woodland and into the Mississippian periods) may and changes in material culture (e.g. introduction of decorated/mortuary ceramics) may be found in the archaeological record (Almy 2012 ; Worth 2013 ). Lastly, the populations of the Palmer, Manasota, and Casey Key sites were found to be closely related (Table 5 8 ). Based on radiocarbon dates and pottery analysis (Bullen and Bullen 1976; Florid a Master Sites File), these sites were either used coevally or sequentially and it appears that the populations utilizing these sites were biologically very close. Another possibility, if the sites were used coevally, is that each site may have represent ed a specific segment of the population, or social divisions, that buried their dead in designated locations based on societal norms (Luer 2011) The variation found within samples from these sites (0.030) is lower than the core Archaic samples (0.045) and t his may be due to the fact that these Manasota sites are closer both temporally and geogr aphically than the Archaic sites. Alternatively, it

PAGE 126

126 could indicate that there was more homogeneity within these populations than the Archaic samples, meaning they had a smaller or more select pool of mates than in the Archaic populations. Overall, these analyses show that there is a distinction between the Weeden Island like/ land site samples used here. This research also shows that the Weeden Island samples exhibited more biological variation perhaps because the Weeden Island Complex encompasses multiple biological groups possibly due to social and geographic barriers to gene flow, whereas the Manasota label seems to encompass a more hom ogenous population suggesting fewer barriers to gene flow There also appears to be evidence that there is a shared Deptford population between these cultures suggesting a shared biological origin, most likely via the late Archaic migration into the penins ula. In summation, it appears that (1982) hypothesis that the Manasota culture is an outgrowth of the local population most likely settling in the area sometime in the late Archaic. The Manasota culture t hen flourished as a culture complex separate from surrounding cultures to the north and south for some time as evidenced by the pottery styles, mortuary practices, and biological separation. South Florida Results The South Florida populations (N=99) consist of samples from Dunwody, Hutchinson Island, Highland Beach, Margate Blount, and Captiva Island. A ll PCA tests ( KMO > 0.759 ) were significant suggesting that PCA is an appropriate method for analyzing these data. In all cases the total variance expl ained by the first two principal

PAGE 127

127 Lambda test was significant. For each test the percentage of variance explained by the first two functions was between 73% and 84%. Cross valid ation produced percentages between 72% and 79%. For the R matrix tests the unbiased F ST decreased between the two tests and the percent variance explained by the first two eigenvectors was 72% and 75% for the two tests. In the first PCA test with all the sites included, the variables with the highest loading (> 0.7) in the first component were upper facial breadth and nasion to fmt The variable with the highest loading in the second compone nt was maximum length, basion bregma, and frontal height. Figure 5 29 depicts the graph of the first two components show Dunwody separating from the rest of the samples towards the bottom of the graph. In the DFA, the variables that contributed to the first function were orbital height and breadth, occipital height, mini mum frontal breadth, occipital condyle breadth, and orbital breadth at nasion, and the variables contributing to the second function were basion bregma, occipital condyle length, and parietal breadth It was also noted that the Captiva Island (78.3%) and H ighland Beach (73.3%) samples classified correctly in cross validation most often suggesting that these populations were distinct from the other groups Figure 5 30 depicts the graph o f the first two functions with the Margate Blount and Hutchinson Island populations close together, while Dunwody falls nearer Captiva Island and Highlands Beach separates somewhat. In the R Matrix analysis the unbiased F ST ( 0.150619 ) indicates moderate variability within these sites. Hutchinson Island and Margate Blount exhi bit the highest levels of variance at 1.039 and 1.070, respectively. Margate Blount also had a high positi ve residual of 0.201, indicative of

PAGE 128

128 higher than average external g ene flow. The graph (Figure 5 31 ) of the first two eigenvalues shows that t he sites are somewhat distant from each other, al though Captiva Island and Dunwody are closer together, while Hutchinson Island, Margate Bloun t, and Highland Beach are close Captiva and Dunwody have the smallest biological d istance between them (Table 5 9 ). A positive correlation between biological distance and geography is noted in these data (Table D 7) The Dunwody sample was assigned to South Florida for these analyses based on its geographic location, however as discussed above, it has been labeled a s a M anasota culture site (Gold 2006) Therefore the analysis of South Florida was repeated without the Dunwody sample In the PCA test the variables with the highest loading (> 0.7) in the first component were upper facial breadth and nasion to fmt The variab le with the highest loading in the sec ond component were nasal height and mastoid breadth. No strong pattern emerged in the graph of the first two components. In the DFA, the variables that contributed to the first function were orbital height, frontal hei ght, minimum frontal breadth, occipital condyle breadth, and orbital breadth at nasion, and those contributing to the second function were fo ramen magnum breadth, basion bregma, maximum length, parietal breadth, orbital breadth, and occipital condyle lengt h It was also noted that Captiva Island (87%) and Highland Beach (73.3%) classified correctly in cross validation most often suggesting that these populations were somewhat distinct from the other samples Figure 5 32 depicts a graph of the first two func tions with that Margate Blount and Hutchinson Island populations closely grouped and Captiva Island and Highlands Beach separate In the R Matrix analysis of all the sites the unbiased F ST fell to 0.119637, indicating decrease d variability within

PAGE 129

129 these sit es. Hutchinson Island and Margate Blount exhibit the highest levels of variance at 1.041 and 1.067, respectively. None of the residuals were particularly high and the biological distance between the Captiva Island and Margate Blount populations was the lar gest (Table D 6) Discussion These results indicate that the level of variability is low within the South Florida sample whether Dunwody is included or not This level of variability is similar to that seen in the Manasota sample. T he samples from the Ma rgate Blount, Highland Beach, and Hutchinson Island sites group together while the Captiva and Dunwody samples cluster ( Figures 5 30 and 5 31 ) It is also noted that in the PCA analyses the variables that contribute the most to differentiating the populations are upper facial breadth and nasion to fmt both are measures of the width of the upper facial region. In the DFA, the shape of the eyes and nose (orbital height and orbital breadth at nasion) contribute to the first two functions most often followed by the height and width of the skull (basion to bregma and parietal width). Dunwody may be somewhat different based on the PCA graph from the f irst analys is (Figure 5 29 ), which s upport s its assignment to the Manasota culture. Nonetheless, the somewhat close biological relationship with the other South Florida samples may indicate a closer biological relationship to these populations than previou sly thought All of the analyses of Dunwoody appear to support the notion that could indicate that is interacted with or received gene flow from both at various times. I t is important to note that the Dunwody sample is th e result of a salvage operation, therefore, it is unclear how representative the sample is of the overall population it is

PAGE 130

130 and the cultural assignment of these remains is limited due to the sparse nature of the archaeological materials associated with these remains (Gold 2005) Finally, it can be noted that overall there is an east west division in the South Florida samples (Figures 5 30 and 5 32 ), with Margate Blount, Highland Beach, and Hutchinson Island on the East Coast and Captiva Island and Dunwody on the West Coast. This pattern is supported by the correlation between bio logical and geographic distance, and all of these findings suggest that there may have been cultural and geographic barrier s to gene flow between the sample populations however the lack of samples from the central portion of the southern peninsula limit this interpretation. Weeden Island / Alachua / St. Johns / Fort Walton Results The first analysis in this series (N 52) compares the Weed en Island samples from Crystal River, Hughes Island, and Bay Pines to the Alachua culture sample from Henderson Mound and the St. Johns culture sample from Browne Site 5. The second analysis (N=49) compares the samples from the first three sites to the For t Walton Temple Mound sample. McKeithan and Yellow Bluffs were not used in these analyses due to their small samples sizes and the previous analyses indicated that their inclusion had little impact on the overall results In the first PCA test, the variab les with the highest loading (> 0.7) in the first component were minimum frontal width and upper facial breadth. The variables with the highest loading in the second component were frontal height and mastoid breadth. Figure 5 33 shows a graph of the first two components with Browne Site 5 separating slightly to one side from the other sites In the DFA, the variable that contributed the most to the first function was orbital height, while nasion to fmt contributed most to the

PAGE 131

131 second function Browne Site 5 (66.7%) and Crystal River (64.3%) classified correctly most often in cross validation suggesting that these population s were somewhat distinct from other sampled populations The graph in Figure 5 34 portrays th e first two functions shows a grouping of the samples from the Hughes Island, Bay Pines, and Henderson Mound sites. Crystal River separates from the group somewhat and Browne Site 5 is clearly separate from the other sites In the R Matrix analysis, the unbiased F ST ( 0.170943 ) suggests moderate level s of variability within these sites Browne Site 5 had the highest level of variance at 1.106 and the highest positiv e residuals (0.572). Figure 5 35 shows a graph of the first two eigenvalues with Browne Site 5 sample clearly distant from the other site s amples Browne Site 5 also exhibits the highest biological distances between it and the other samples (Table 5 1 0 ). In the second analysi s the variables with the highest loading (> 0.7) in the first component were maximum length and upper facial breadth. The variable with the highest loading in the second component was orbital breadth. In Figure 5 36 which is a graph of the first two components it can be seen that the Hughes Island sample clusters to one side of the graph and while the Fort Walton Templ e Mound sample clusters toward the other. In the DFA, the variable that contributed the most to the first function nasion to fmt and occipital condyle length. The variables that contributed to the second function were maximum length, upper facial breadth, orbital breadth at nasion, basion b regma, mastoid breadth, and frontal height It was also noted that the Fort Walton Temple Mound (66.7%) and Crystal River (75%) frequently classified correctly in cross validation suggesting that these population s were so mewhat distinct from the other sampled populations Figure 5 37 depicts the graph of the first two functions with

PAGE 132

132 a grouping of the Hughes Island and Bay Pines samples While, Crystal River is somewhat separated and Fort Walton Temple Mound is clearly sepa rate d In the R Matrix analysis, t he unbiased F ST ( 0.103819 ) suggested a moderate level of variability within these sites Fort Walton Temple Mound had the highest level of variance at 1.345 and the highest positive residuals (0.430), while Bay Pines has a high negative residual of 0.402. Figure 5 38 depicts the graph of the first two eigenvalues with Bay Pines and Hughes Isl and clustering together with Crystal River some distance from them, but all the populations are separated from the Fort Walton Temple Mound population Fort Walton Temple Mound and Crystal River show the largest biological distance, while Hughes Island and Bay Pines have the smallest ( Table 5 11 ). Discussion The first analysis shows that the population sample from the Browne Site 5 appears to be distinct from those of the Weeden Island and Alachua culture populations; however, having only one site to compar e makes these results open to interpretation. The grouping of the Henderson Site with the Weeden Island samples (2003) argument that the Alachua culture represented an in situ culture change rather than a migra tion of people into the area as first suggested by Milanich (1998) The second analysis may imply that the Mississippian influences noted in the Fort Walton culture could have been due to an actual migration of peoples into the area because this site is cl early separated from the Weeden Island cultures. Yet, the level of variation based on F ST values does not change significantly between these analyses and the analyses of the Weeden Island sites alone This finding argues against an influx of people; sugges ting that the Fort Walton culture change could have been due to the

PAGE 133

133 adoption of an ideology without a large migration of peoples. Alternatively, t hese results could suggest that the populations inhabiting the panha ndle were biologically separate from the W eeden Island populations to the east These results could suggest there was a long term biological separation of the populations in these two areas that correlates with the variation seen in the cultural developments of these two areas starting with the d istinction between the Swift Creek and Santa Rosa Swift Creek cultures continuing on in the development of the later cultures in these two areas However, t hese interpretations are preliminary because no other populations from the panhandle area, particula rly sites predating the Fort Walto n Temple Mound, were used in this analysis. Manasota / Safety Harbor Results This analysis (N=179) was run to determine the relationships of the Safety Harbor samples ( Sarasota Bay Mound and Safety Harbor ), the Fort Walton Temple Mound sample, the Weeden Island samples ( Hughes Island, Crystal River, and McKeithan ) and the Manasota samples ( Palmer Mound, Casey Key, Manasota Key Cemetery, Dunwody, Venice Beach Complex, Bayshore Homes, Bay Pines, and Yellow Bl uffs ) All PCA tests conducted ( KMO > 0.717 ) were significant suggesting that PCA was an appropriate s ta tistical method for analyzing these data. In all cases the total variance explained by the first two principal components was low, between 36% and 42%. Lambda test was significant. With each test the percentage of variance explained by the first two functions improved from 76.0% to 82.4%. Cross validation results for these analyses improved from 44% to 58%. For the R matrix tests the unbiased F ST was relatively constant with each test and the percent variance explained by the first two eigenvectors improved from 64% to 92.8%.

PAGE 134

134 In the first PCA test with all the populations included, the variable with the highest loadi ng (> 0.7) in the first component was upper facial breadth. The variable with the highest loading in the second component were maximum length and parietal length. In the DFA, no variables contributed to the first function and the variables that contributed to the second function were occipital condyle breadth and length. Figure 5 39 a graph of the first two functions shows the separation of Manasota and Weeden Island ; the Fort Walton Temple M ound sample clustered in with the Weeden Island sites and the Sa fety Harbor sites fell in with the Manasota sites. In the R Matrix analysis of all the sites the unbiased F ST was 0.136718, suggesting a moderate level of variability within these sites. Yellow Bluffs exhibits the highest level of variance at 1.620 and Sar asota Bay Mound exhibits the lowest level of variance at 0.447. The other variances are roughly equal. Yellow Bluffs has the highest positive residuals and Sarasota Bay Mound has the highest negative residuals. Figure 5 40 shows the grap h of the first two eigenvalues and is similar to Figure 5 39 with the Manasota and Safety Harbor sites grouping together and the Weeden Island sites grouping with Fort Walton Temple Mound. T he largest biological distances were noted between the Manasota/Safety Harbor sites a nd the Weeden Island sites (Table 5 1 2 ). In the second test Venice Beach Complex, Yellow Bluffs, McKeithan, and Fort Walton Temple Mound samples were removed, the first three due to small samp le size and the last because of its distinction from the Safety Harbor sites. The overall observed pattern was similar to the previous analysis. Therefore the next test removed Crystal River, Hughes Island, and Bay Pines leaving the Manasota and Safety Harbor populations For this analysis, the variables with the highest loading (> 0.7) in the first

PAGE 135

135 component of the PCA analysis were upper facial breadth and n asion to fmt. The variables with the highest loading in the second component were maximum length and parietal leng th. In the DFA, the variable that contributed to the first function was occipital condyle breadth, and the second function was influenc ed by maximum length, basion bregma, and parietal length. Figure 5 41 shows the graph of the first two functions with the Manasota sites of Palmer, Manasota Key Cemetery, Casey Key and Bayshore Homes grouping together, and Safety Harbor and Sarasota Bay Mound together and Dunwody separate from both groups In the R Matrix analysis the unbiased F ST ( 0.120718 ) indicated moder ate variability within these sites. Figure 5 42 a graph of the of the first two eigenvalues show s a tight grouping of Palmer Mound, Casey Key, Bayshore Homes, and Manasota Key Cemetery, while Safety Harbor and Sarasota Bay Mound group together, and Dunwo dy separates from both groups T he largest biological distances were noted between Dunwody and the Safety Harbor populations (Table 5 13 ). In the final analysis Bayshore Homes and Dunwody were removed The variables with the highest loading (> 0.7) in the first component of the PCA analysis were minimum frontal breadth, u pper facial breadth, and nasion to fmt. The variable with the highest loading in the second component was basion to nasion. No distinct pattern can be seen in the graph of the first two co mponents. In the DFA, the variable that contributed to the first function was occipital condyle breadth and length, and the second function was influenced by nasion to fmt. Figure 5 43 shows a graph of the first two functions evidencing a separation betwee n the Manasota sites and the Safety Harbor sites. In the R Matrix analysis the unbiased F ST decreased slightly ( 0.119704 )

PAGE 136

136 Sarasota Bay Mound exhibits the lowest level of variance at 0.470 and high negative residuals. The other variances are roughly equal. Figure 5 44 depicts the graph of the first two eigenvalues and shows a cluster of the Manasota site samples ( Palmer Mound, Casey Key, and Manasota Key Cemetery ) and a separation of Safety Harbor and Sarasota Bay Mound this cluster and each other T he larg est biological distances are noted between the Safety Harbor site sample and the three Manasota populations (Table 5 14 ). Discussion T he first analysis suggests several facts about th e Fort Walton Temple Mound site. F irst the Fort Walton Temple Mound sample is more closely related to the Weeden Island samples than the Manasota samples ; though its tight grouping with the Weeden Island sites is most likely an artifact of the testing methods as a previous analysis suggested that the Fort Walton Temple Mound is not that closely related to the Weeden Island populations Second, the Fort Walton Temple Mound is not related to the Safety Harbor sites (Figu res 5 39 and 5 40), seeming to discount the theory that there was contact be tween these locations as suggested by similarities in mound structure and iconography (Milanich 1994) The subsequent analyses indicate that the Safety Harbor sites are more closely related to the Manasota sites than to the Weeden Island sites; however, o n close inspection they are distinguished from the Manasota sites of Palmer Mound, Manasota Key Cemetery, Casey Key, and Bayshore Homes. Sarasota Bay Mound does appear to be more closely related to Palmer Mound, Manasota Key Cemetery, and Casey Key, and Sa fety Harbor is more closely related to Bayshore Homes (Figures 5 41 and 5 42). This is probably due to geographic proximity, the first sites are in the Sarasota Bay area

PAGE 137

137 and the second set of sites is in the Tampa Bay area, which suggests a likely biologic al continuity between these cultures in these locations. Comparison of the unbiased F ST values between these analyses and the Manasota site analyses show a higher level of variation when the Safety Harbor sites are included, suggesting that there was incre ase in biological variation between these two cultures, possibly the result of an influx of people into the area bringing the new Mississippian ideas that helped shaped the Safety Harbor culture However, it appears that the Manasota cultural influence was still strong as there are continued similarities archaeologically between the two cultures in site selection and lack of burial goods in burial mounds (Luer and Almy 1987; Mitchem 2012) Overall, these findings support the theory that the Safety Harbor cu lture was a local development with some outside influences (Luer and Almy 1987; Mitchem 2012) I f these are the same biological influences that may have impacted the Fort Walton culture is an open question and hopefully further investigation will help clar ify this line of inquiry. Biological Variation And Gene Flow Based on the variation found overall and within the various sub samples discussed above, it was determined that there were patterns to the level of biological variation as represented by F ST is the R matrix analysis. Therefore an analysis of the various temporal/cultural frames, as outlined in Chapter 2, was undertaken to examine gene flow patterns as represented by the residuals. In the analysis of all populations the F ST is 0.161536 and Warm M ineral Springs, McKeithan, and Yellow Bluffs have large positive residuals and Bay Pines and Sarasota Bay Mound have large negative residuals. The F ST for the Archaic populations is 0.298673, for the Woodland and

PAGE 138

138 Mississippian populations it is 0.102450, f or the Woodland populations it is 0.099426, and for the Mississippian populations (Fort Walton Temple Mound, Safety Harbor, Sarasota Bay Mound) it is 0.549114. In addition, Fort Walton Temple Mound has a large positive residual and Sarasota Bay Mound has a large negative residual. Archai c Period The F ST for all the Archaic populations (Warm Mineral Springs, Windover, Little Salt Spring, Republic Grove, Bay West, Gauthier, and Bird Island) is 0.298673 The F ST for the archaeologically similar sites (Windover, Little Salt Spring, Republic Groves, Bay West, and Gauthier) is 0.098585, and for the biologically similar populations (Windover, Little Salt Spring, Bay West, and Gauthier) is 0.045971. Warm Mineral Sprin gs and Bird Island have large positive residuals and Windover and Bay West have large negative residuals. Woodland Period The F ST for all the Woodland populations (Browne Site 5, Henderson Mound, Palmer Mound, Hughes Island, Bay Pines, Bayshore Homes, McK eithan, Yellow Bluffs, Crystal River, and Manasota Key Cemetery, Hutchinson Island, Highland Beach, Margate Blount Mound, and Captiva Island) is 0.099426. Within the Woodland populations are several subgroups: Weeden Island, Manasota, and South Florida. Th e F ST for the Weeden Island populations (Crystal River, Hughes Island, McKeithan, and Bay Pines) is 0.263833. McKeithan has a large positive residual while Bay Pines has a large negative residual. The F ST for the Manasota populations (Palmer Mound, Casey K ey, Bayshore Homes, Manasota, Yellow Bluffs, and Dunwody) is 0.158604, for the archaeologically similar sites (Palmer Mound, Manasota, Casey Key, and Bayshore Homes) it is 0.054757, and for the archaeologically and geographically similar site

PAGE 139

139 (Palmer Mound Manasota, and Casey Key) it is 0.030003. Overall, Yellow Bluffs has a high positive residual and Palmer and Casey Key have moderately high negative residuals. For the South Florida populations (Highland Beach, Margate Blount Mound, Hutchinson Island, Cap tiva Island) the F ST is 0.119637. The residuals are all close to zero with no obviously large positive or negative residuals. Discussion These results indicate that the populations associated with the sites labeled Archaic are moderately variable, but clo ser inspection of those results reveals that Warm Mineral Springs and Bird Island influence that variability and that the archaeologically similar sites are rather homogenous. This is supported by the large positive residuals associated with these two site s In other words, the populations at these two sites were receiving larger than average levels of external gene flow or gene flow from a source different from the other sites. It may also indicate that the sites associated with the Woodland and Mississipp ian periods were less variable than the Archaic samples suggesting that these populations most likely share a common origin and may not have had diverse gene flow compared to that in the Archaic. When the Mississippian period populations were removed, the populations associated with the Woodland period appeared more homogenous indicating that there was increase d variability in the Mississippian period possibly suggesting increased or new sources of gene flow. The homogeneity noted in the Woodland population s is misleading, and closer inspection reveals several patterns. The Weeden Island populations were clearly more variable than the other populations, possib ly due to biological separation between mound centers (i.e. Crystal River, McKeithan, Hughes Island) resulting in increased biological variability. It is noted that McKeithan has a large positive residual suggesting

PAGE 140

140 it was receiving above average gene flow from an outside source; however the small sample size makes this result suspect. The large negativ e residual associated with Bay Pines is more reliable and indicates that this site was receiving less than average gene flow from an outside source or it was receiving gene flow from within the populations used in the analysis. These results support the co ncept that populations associated with the Weeden Island Complex had culturally recognized boundaries that prevented gene flow, possibly villages associated with mound centers. The Manasota populations were much less variable than the Weeden Island populat ions, particularly when Yellow Bluffs is removed and the sites that are closest geographically are also very homogenous. The large positive residual for Yellow Bluffs may be explained by its Deptford period association as it seems to have received gene flo w from another source, or it could be an artifact of small sample size. When the geographically similar sites are examined, only Manasota has a slightly elevated positive residual perhaps because it was receiving more external gene flow than either Palmer or Casey Key. This supports the supposition that Palmer and Casey Key were most likely ancestor descendent populations based on temporality Overall it can be suggested that the Manasota population s had more open gene flow or there were fewer cultural bo undaries to gene flow than in the populations associated with Weeden Island Complex that seems to have restricted gene flow into or between the populations associated with those sites. Lastly, the South Florida populations were found to be less variable than the Weeden Island populations, but more variable than the geographically similar Manasota populations and the residuals support the concept that the sites were receiving simil ar levels of external gene flow. This may be indicative of the fact that there were some

PAGE 141

141 cultural divisions, along with geographic distance, limiting gene flow between populations that are not as well defined as in the Weeden Island Complex but more defin ed than the Manasota cultures. The populations associated with the Mississippian period are highly variable and this is most likely due in part to the geographic separation of the sites associated with the sampled populations I t is also likely that to th e two culture groups had distinct mate sources and biological histories. Further, it appears highly unlikely that these populations were exchanging mates. The residuals suggest that Fort Walton Temple Mound received high levels of external gene flow perha ps from other Mississippian populations; however, this cannot be adequately assessed at this time Sarasota Bay Mounds large negative residual seems to imply that this population was isolated compared to the other two sites, but again small sample size cou ld be artificially inflating the se results. Overall the Mississippian period is marked by increased regionalization, resulting in significant distinctions between cultures archaeologically (Ashley and White 2012) ; this separation also appears to have had a significant impact on the biological diversity of the period. Another way that these results could be interpreted is that sites with higher positive residuals h ad more open cemetery structure meaning that more diverse populations were coming together to bury their dead sites with negative residuals had more insular populations or practices with respect to who could be buried in the cemeteries. Future detailed analyses of the structure of individual cemeteries will help with evaluating these interpretation s.

PAGE 142

142 Biodistance Assumptions All of the results discussed above were examined to determine whether or not the assumptions of biological distance analysis hold true in these Florida populations. In particular, if the patterns of biological distance and varia bility are an indirect measure of the degree of social exchange between the groups they represent, if a cemetery or sample population is representative of the original breeding population, and if environmental effects are minimal and randomly distributed a mong the populations. Examination of the first assumptions that biodistance is an indirect measure of social exchange as evidenced by archaeological similarities or dissimilarities is predominately supported by this research. As can be seen in the analysi s of the Archaic populations there is a core population represented by the individuals from Windover, Bay West, Little Salt Spring, and Gauthier, who were close biologically. These individuals were also associated archaeologically they share similar burial patterns, site structure ( village site and burial site proximity) and artifacts (particularly those found in the cemeteries). There are subtle distinctions between these sites that are most likely related to variations in mate sources, but overall these s ites are consistent biologically. Given the temporal frame covered by these sites they are most likely ancestor descendant groups, but may be evidence that social exc hange occurred in transition from one site to the next supporting the assumptions that bi ological distance is an indirect measure of social exchange. In addition, the results from the Weeden Island / Manasota analyses further support this assumption. Based on archaeological evidence, differences in pottery styles and burial patterns, it is sugge sted that the Weeden Island and Manasota populations were distinct or had very little interaction, at least in the early periods. The biological evidence supports this idea of separation.

PAGE 143

143 The Weeden Island and Manasota populations repeatedly separate from each other in all the tests, therefore the archaeological and biological evidence points to very little social exchange between these groups. Late in the Manasota period Weeden Island pott ery and burial patterns begin to appear in the archaeological record (Luer and Almy 1982) perhaps indicating social interaction, however the Weeden Island sites used in this study are associated with the early Weeden Island period and only the Palmer Moun d site appears to have Weeden Island like burials. The other late sites continue with the flexed burial pattern typical of the early and middle Manasota periods (Luer and Almy 1982) Therefore the ability to test this late period biological / social interact ion was limited. There is one site sample where it does not appear that biological distance is a good measure of social interaction, the Republic Groves sample. Analysis of this sample implies that there are many archaeological similarities with the other Archaic sites, particularly Little Salt Spring and Bay West, and it falls temporally between these two sites. Given the pattern seen in the other Archaic sites it would be expected that Republic Groves would fall neatly into line with the other population s, however this is not the case. Despite appearing archaeologically similar to the other sites, the population is biologically distinct This relationship needs further investigation to understand this distinction. In summary, it can be concluded that, th is research supports the assumption that in Florida biological distance is an indirect measure of social interaction. Analysis of the second assumption: cemeteries are representative of breeding populations based on levels of homogeneity is supported by t his dissertation research.

PAGE 144

144 Most of the studied sites have a similar level of phenotypic variance represented by the observed phenotypic variance in the R matrix analysis. The sites that have obviously higher and lower levels of variance than others, McKeit han, Yellow Bluffs, Warm Mineral Springs, and Sarasota Bay Mound samples, are most likely artifacts of sampling since these samples were limited to two or three individuals L arger samples may have produced homogeneity levels more consistent with the other samples. Overall the variance levels of the other sites are similar implying that the remaining populations were internally consistent suggesting breeding population s. H igher levels of variation would have indicated multiple or diverse mate sources. Anal ysis of the residuals also suggest that most of the sites received average levels of external gene flow, or levels consistent with the null hypothesis, supporting the view that the cemeteries represent breeding populations. However a few sites had higher t han expected positive residuals ( Warm Mineral Springs, McKeithan, and Yellow Bluffs ) which suggest s that they received elevated levels of external gene flow or that they had more open cemetery structures possibly meaning they were less likely to be breed ing populations H owever as mentioned previously, small sample sizes significantly limit interpretation. The third assumption that environmental effects are minimal appears to hold true within specific time frames; however environment may play a role i n the stock populations resulting in the drastic biological differences seen in the migration events. The environmental differences between the Paleoindian and Archaic periods and the Archaic and Woodland periods may have impacted the populations throughou t the Southeast resulting in the apparent differences noted in the populations migrating into the peninsula. H owever this supp osition would need to be tested through the analysis

PAGE 145

145 of various populations throughout the Southeast. It can be said though that environment does not appear to significantly impact the populations used in this study. For example, despite geogr aphic/ environmental differences between the Archaic populations they are relatively biologically consistent, particularly the core populations mentioned previously. Similarly, all of the Woodland and Mississippian populations appear overall consistent despite widely varied temporal and geographic distances which would have had environmental differences.

PAGE 146

146 Figure 5 1 Graph of the first and second factors for the PCA analysis of all the populations identified by cultural label

PAGE 147

147 Figure 5 2 Graph of the first and second discriminant functions for all populations. The cluster of Archa ic populations is circled in blue.

PAGE 148

148 Figure 5 3 Graph of the first and second eigenvectors for all populations.

PAGE 149

149 Table 5 1 The biological distances and standard errors for those distances between all populations Warm Mineral Springs Gauthier Little Salt Spring Windover Bay West Palmer Henderson Mound Warm Mineral Springs 0.193 0.239 0.191 0.225 0.208 0.289 Gauthier 1.131 0.051 0.019 0.046 0.026 0.071 Little Salt Spring 1.260 0.062 0.040 0.059 0.073 0.096 Windover 1.194 0.061 0.022 0.046 0.027 0.075 Bay West 1.166 0.061 0.000 0.093 0.066 0.086 Palmer 1.423 0.109 0.280 0.232 0.259 0.070 Henderson Mound 1.945 0.218 0.194 0.320 0.163 0.247 Hughes Island 1.443 0.145 0.183 0.228 0.126 0.147 0.014 Browne Site 5 1.478 0.192 0.237 0.300 0.157 0.142 0.205 Bay Pines 1.296 0.083 0.141 0.196 0.120 0.168 0.102 Dunwody 1.262 0.218 0.368 0.413 0.281 0.228 0.271 McKeithan 1.942 0.528 0.530 0.674 0.507 0.501 0.209 Venice Beach Complex N/A N/A N/A N/A N/A N/A N/A Sarasota Bay Mound 1.594 0.225 0.295 0.166 0.177 0.127 0.344 Yellow Bluffs 2.302 0.611 0.917 0.795 0.632 0.255 0.271 Casey Key 1.274 0.102 0.187 0.203 0.179 0.021 0.159 Crystal River 1.677 0.240 0.358 0.397 0.244 0.241 0.157 Bayshore Homes 1.789 0.208 0.360 0.350 0.334 0.042 0.198 Hutchinson Island 1.818 0.234 0.261 0.447 0.306 0.182 0.154 Manasota 1.271 0.050 0.178 0.157 0.111 0.016 0.189 Bird Island 1.690 0.155 0.189 0.196 0.251 0.098 0.250 Republic Groves 1.942 0.271 0.178 0.159 0.232 0.425 0.380 Highland Beach 1.552 0.153 0.157 0.297 0.148 0.237 0.062 Margate Blount Mound 1.836 0.077 0.183 0.201 0.103 0.179 0.129 Fort Walton Temple Mound 2.229 0.371 0.336 0.392 0.311 0.280 0.000 Safety Harbor 1.641 0.339 0.598 0.452 0.427 0.150 0.380 Captiva Island 1.272 0.108 0.226 0.226 0.158 0.096 0.224 The biological distances are on the bottom half of the table and the s tandard errors for each analysis are on the top half

PAGE 150

150 Table 5 1 Continued Hughes Island Browne Site 5 Bay Pines Dunwody McKeithan Venice Beach Complex Sarasota Bay Mound Warm Mineral Springs 0.247 0.256 0.252 0.249 0.386 N/A 0.357 Gauthier 0.057 0.068 0.061 0.079 0.181 N/A 0.142 Little Salt Spring 0.089 0.101 0.095 0.124 0.209 N/A 0.178 Windover 0.061 0.074 0.071 0.092 0.191 N/A 0.129 Bay West 0.076 0.086 0.087 0.109 0.201 N/A 0.156 Palmer 0.054 0.058 0.069 0.076 0.174 N/A 0.124 Henderson Mound 0.063 0.097 0.089 0.113 0.166 N/A 0.185 Hughes Island 0.094 0.058 0.109 0.149 N/A 0.144 Browne Site 5 0.221 0.108 0.082 0.195 N/A 0.164 Bay Pines 0.000 0.236 0.110 0.175 N/A 0.178 Dunwody 0.288 0.058 0.196 0.234 N/A 0.218 McKeithan 0.132 0.414 0.225 0.681 N/A 0.284 Venice Beach Complex N/A N/A N/A N/A N/A N/A Sarasota Bay Mound 0.103 0.197 0.244 0.544 0.564 N/A Yellow Bluffs 0.157 0.358 0.339 0.505 0.371 N/A 0.231 Casey Key 0.090 0.143 0.190 0.269 0.318 N/A 0.107 Crystal River 0.065 0.199 0.084 0.269 0.187 N/A 0.400 Bayshore Homes 0.139 0.132 0.134 0.192 0.357 N/A 0.232 Hutchinson Island 0.191 0.163 0.128 0.246 0.450 N/A 0.389 Manasota 0.060 0.112 0.052 0.150 0.498 N/A 0.108 Bird Island 0.051 0.175 0.113 0.392 0.302 N/A 0.148 Republic Groves 0.361 0.551 0.304 0.726 0.663 N/A 0.140 Highland Beach 0.035 0.206 0.019 0.232 0.190 N/A 0.280 Margate Blount Mound 0.187 0.213 0.128 0.304 0.388 N/A 0.315 Fort Walton Temple Mound 0.068 0.250 0.111 0.392 0.118 N/A 0.269 Safety Harbor 0.210 0.210 0.240 0.311 0.603 N/A 0.083 Captiva Island 0.083 0.115 0.061 0.124 0.344 N/A 0.093

PAGE 151

151 Table 5 1 Continued Yellow Bluffs Casey Key Crystal River Bayshore Homes Hutchinson Island Manasota Bird Island Republic Groves Warm Mineral Springs 0.415 0.229 0.228 0.235 0.253 0.206 0.271 0.270 Gauthier 0.190 0.048 0.040 0.037 0.053 0.025 0.064 0.063 Little Salt Spring 0.252 0.086 0.084 0.084 0.086 0.068 0.095 0.081 Windover 0.203 0.054 0.041 0.038 0.061 0.030 0.063 0.046 Bay West 0.216 0.080 0.067 0.075 0.085 0.055 0.098 0.083 Palmer 0.143 0.033 0.035 0.017 0.043 0.016 0.053 0.071 Henderson Mound 0.175 0.082 0.063 0.067 0.073 0.069 0.103 0.105 Hughes Island 0.153 0.067 0.046 0.055 0.072 0.047 0.070 0.097 Browne Site 5 0.187 0.080 0.068 0.060 0.074 0.060 0.093 0.121 Bay Pines 0.191 0.093 0.060 0.067 0.076 0.058 0.091 0.104 Dunwody 0.213 0.103 0.083 0.074 0.091 0.072 0.127 0.144 McKeithan 0.257 0.173 0.136 0.159 0.181 0.179 0.179 0.212 Venice Beach Complex N/A N/A N/A N/A N/A N/A N/A N/A Sarasota Bay Mound 0.236 0.141 0.165 0.142 0.174 0.126 0.156 0.144 Yellow Bluffs 0.177 0.134 0.142 0.183 0.153 0.191 0.229 Casey Key 0.349 0.064 0.045 0.062 0.040 0.073 0.101 Crystal River 0.173 0.254 0.032 0.055 0.035 0.068 0.088 Bayshore Homes 0.227 0.084 0.160 0.041 0.024 0.055 0.082 Hutchinson Island 0.465 0.136 0.271 0.129 0.046 0.085 0.104 Manasota 0.292 0.036 0.155 0.050 0.151 0.058 0.075 Bird Island 0.384 0.093 0.198 0.100 0.256 0.098 0.106 Republic Groves 0.820 0.452 0.614 0.534 0.637 0.389 0.395 Highland Beach 0.390 0.124 0.160 0.209 0.127 0.149 0.196 0.413 Margate Blount Mound 0.425 0.120 0.131 0.161 0.177 0.121 0.171 0.348 Fort Walton Temple Mound 0.120 0.259 0.165 0.179 0.295 0.264 0.180 0.328 Safety Harbor 0.166 0.195 0.235 0.123 0.337 0.133 0.230 0.725 Captiva Island 0.239 0.112 0.158 0.134 0.213 0.058 0.172 0.355

PAGE 152

152 Table 5 1 Continued Highland Beach Margate Blount Mound Fort Walton Temple Mound Safety Harbor Captiva Island Warm Mineral Springs 0.220 0.281 0.288 0.235 0.198 Gauthier 0.033 0.053 0.072 0.055 0.025 Little Salt Spring 0.062 0.094 0.100 0.111 0.068 Windover 0.035 0.064 0.066 0.054 0.026 Bay West 0.056 0.078 0.092 0.091 0.055 Palmer 0.034 0.063 0.059 0.035 0.020 Henderson Mound 0.050 0.087 0.054 0.093 0.067 Hughes Island 0.041 0.090 0.061 0.070 0.045 Browne Site 5 0.068 0.098 0.091 0.075 0.055 Bay Pines 0.049 0.093 0.079 0.086 0.054 Dunwody 0.079 0.117 0.114 0.094 0.063 McKeithan 0.136 0.191 0.140 0.194 0.155 Venice Beach Complex N/A N/A N/A N/A N/A Sarasota Bay Mound 0.149 0.181 0.163 0.125 0.118 Yellow Bluffs 0.163 0.196 0.141 0.138 0.141 Casey Key 0.049 0.077 0.082 0.064 0.045 Crystal River 0.032 0.060 0.052 0.046 0.029 Bayshore Homes 0.035 0.063 0.053 0.035 0.026 Hutchinson Island 0.041 0.076 0.076 0.067 0.046 Manasota 0.034 0.061 0.064 0.040 0.022 Bird Island 0.067 0.092 0.082 0.078 0.062 Republic Groves 0.073 0.101 0.086 0.103 0.065 Highland Beach 0.058 0.051 0.056 0.026 Margate Blount Mound 0.119 0.083 0.083 0.059 Fort Walton Temple Mound 0.161 0.191 0.074 0.054 Safety Harbor 0.389 0.283 0.327 0.037 Captiva Island 0.130 0.150 0.223 0.177

PAGE 153

153 Figure 5 4 Graph of the first and second factors for the PCA analysis for the Woodland and Mississippian populations. The cluster of Manasota populations is circled in green and the Weeden Island populations in red.

PAGE 154

154 Figure 5 5 Graph of the first and second discriminant functions for the Woodland and Mississippi an populations. The cluster of Manasota populations is circled in green and the Weeden Island populations in red.

PAGE 155

155 Figure 5 6 Graph of the first and second factors for the PCA analysis for all Archaic populations.

PAGE 156

156 Figure 5 7 Graph of the first and second discriminant functions for all Archaic populations.

PAGE 157

157 Figure 5 8 Graph of the first and second eigenvalue for all Archaic populations.

PAGE 158

158 Table 5 2 The biological distances and standard errors for those distances all the Archaic populations Warm Mineral Springs Gauthier Little Salt Spring Windover Bay West Bird Island Republic Groves Warm Mineral Springs 0.193 0.237 0.186 0.229 0.273 0.261 Gauthier 1.398 0.055 0.024 0.046 0.084 0.074 Little Salt Spring 1.521 0.127 0.044 0.069 0.104 0.099 Windover 1.399 0.137 0.079 0.043 0.077 0.056 Bay West 1.502 0.089 0.088 0.107 0.113 0.092 Bird Island 2.116 0.44 0.351 0.423 0.495 0.141 Republic Groves 2.207 0.494 0.418 0.333 0.406 0.992 The biological distances are on the bottom half of the table and the s tandard errors for each analysis are on the top half

PAGE 159

159 Figure 5 9 Graph of the first and second discriminant functions for the second analysis of the Archaic populations.

PAGE 160

160 Figure 5 10 Graph of the first and second eigenvectors for the second analysis of the Archaic populations.

PAGE 161

161 Figure 5 11 Graph of the first and second factors for the PCA analysis for the third analysis of the Archaic populations. The Republic Groves population is circled in light blue.

PAGE 162

162 Figure 5 12 Graph of th e first and second eigenvectors for the third analysis of the Archaic populations. Table 5 3 The biological distances and standard errors for those distances for the third analysis of the Archaic populations. Gauthier Little Salt Spring Windover Bay West Republic Groves Gauthier 0.072 0.029 0.058 0.091 Little Salt Spring 0.188 0.057 0.086 0.121 Windover 0.149 0.115 0.055 0.071 Bay West 0.128 0.128 0.147 0.108 Republic Groves 0.561 0.478 0.396 0.412 The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half.

PAGE 163

163 Figure 5 13 Graph of the first and second factors for the PCA analysis for the fourth analysis of the Archaic populations.

PAGE 164

164 Figure 5 14 Graph of the first and second discriminant functions for the fourth analysis of the Archaic populations.

PAGE 165

165 Figure 5 15 Graph of the first and second factors for the PCA analysis for the first analysis of the Weeden Island/Manasota populations.

PAGE 166

166 Figure 5 16 Graph of the first and second discriminant functions for first analysis of the Weede n Island/Manasota populations.

PAGE 167

167 Figure 5 17 Graph of the first and second eigenvectors for the first analysis of the Weeden Island/Manasota populations.

PAGE 168

168 Table 5 4 The biological distances and standard errors for those distances between the Weeden Island/Manasota populations Palmer Mound Hughes Island Bay Pines Dunwody McKeithan Yellow Bluffs Casey Key Crystal River Bayshore Homes Manasota Key Cemetery Venice Beach Complex Palmer Mound 0.067 0.083 0.094 0.179 0.167 0.035 0.042 0.018 0.021 N/A Hughes Island 0.268 0.069 0.142 0.174 0.165 0.094 0.054 0.068 0.058 N/A Bay Pines 0.287 0.010 0.140 0.178 0.207 0.119 0.062 0.076 0.070 N/A Dunwody 0.402 0.586 0.438 0.253 0.262 0.128 0.105 0.090 0.093 N/A McKeithan 0.529 0.285 0.239 0.832 0.270 0.196 0.149 0.163 0.189 N/A Yellow Bluffs 0.427 0.224 0.444 0.919 0.442 0.213 0.147 0.160 0.171 N/A Casey Key 0.035 0.290 0.406 0.486 0.485 0.630 0.084 0.052 0.050 N/A Crystal River 0.356 0.125 0.093 0.483 0.269 0.253 0.492 0.039 0.039 N/A Bayshore Homes 0.048 0.251 0.203 0.326 0.377 0.355 0.142 0.238 0.025 N/A Manasota Key Cemetery 0.043 0.131 0.129 0.332 0.573 0.416 0.100 0.199 0.059 N/A Venice Beach Complex N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half.

PAGE 169

169 Figure 5 18 Graph of the first and second eigenvectors for the second analysis of the Weeden Island/Manasota populations.

PAGE 170

170 Figure 5 19 Graph of the first and second discriminant functions for the first analysis of the Weeden Island populations.

PAGE 171

171 Figure 5 20 Graph of the first and second eigenvectors for the first analysis of the Weeden Island populations.

PAGE 172

172 Table 5 5 The biological distances and standard errors for those distances between the Weeden Island populations Hughes Island Bay Pines McKeithan Yellow Bluffs Crystal River Hughes Island 0.076 0.228 0.218 0.080 Bay Pines 0.170 0.207 0.253 0.063 McKeithan 1.305 0.886 0.291 0.187 Yellow Bluffs 1.163 1.489 1.202 0.228 Crystal River 0.633 0.222 1.021 1.689 The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half. Figure 5 21 Graph of the first and second eigenvectors for the second analysis of the Weeden Island populations.

PAGE 173

173 Table 5 6 The biological di stances and standard errors for those distances between for the second analysis of the Weeden Island populations Hughes Island Bay Pines Crystal River Hughes Island 0.102 0.106 Bay Pines 0.196 0.095 Crystal River 0.682 0.345 The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half. Figure 5 22 Graph of the first and second factors for the PCA analysis for the first analysis of the Manasota populations. The Dunwody population is circled in yel low and the Yellow Bluffs populations in light blue.

PAGE 174

174 Figure 5 23 Graph of the first and second discriminant functions for first analysis of the Manasota populations.

PAGE 175

175 Figure 5 24 Graph of th e first and second eigenvectors for the first analysis of the Manasota populations.

PAGE 176

176 Table 5 7 The biological distances and standard errors for those distances between the Manasota populations Palmer Bay Pines Dunwody Yellow Bluffs Casey Key Bayshore Homes Manasota Key Cemetery Venice Beach Complex Palmer 0.084 0.102 0.175 0.037 0.021 0.020 N/A Bay Pines 0.297 0.143 0.229 0.121 0.081 0.073 N/A Dunwody 0.497 0.461 0.293 0.135 0.098 0.099 N/A Yellow Bluffs 0.488 0.627 1.241 0.215 0.181 0.180 N/A Casey Key 0.049 0.423 0.568 0.653 0.050 0.048 N/A Bayshore Homes 0.070 0.246 0.411 0.530 0.127 0.025 N/A Manasota Key Cemetery 0.035 0.158 0.395 0.498 0.090 0.060 N/A Venice Beach Complex N/A N/A N/A N/A N/A N/A N/A The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half.

PAGE 177

177 Figure 5 25 Graph of the first and second discriminant functions for the third analysis of the Manasota populations.

PAGE 178

178 Figure 5 26 Graph of the first and second eigenvectors for the third analysis of the Manasota populations.

PAGE 179

179 Figure 5 27 Graph of the first and second factors for the PCA analysis for the fourth analysis of the Manasota populations.

PAGE 180

180 Figure 5 28 Graph of the first and second eigenvectors for the fourth analysis of the Manasota populations. Tab le 5 8 The biological distances and standard errors for those distances for the fourth analysis of the Manasota populations. Palmer Casey Key Bayshore Homes Manasota Key Cemetery Palmer 0.039 0.027 0.024 Casey Key 0.034 0.056 0.054 Bayshore Homes 0.109 0.127 0.034 Manasota Key Cemetery 0.054 0.093 0.110 The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half.

PAGE 181

181 Figure 5 29 Graph of the first and second factors for the PCA analysis for the first analysis of the South Florida populations. The Dunwody population is circled in yellow.

PAGE 182

182 Figure 5 30 Graph of the first and second discriminant f unctions for first analysis of the South Florida populations.

PAGE 183

183 Figure 5 31 Graph of the first and second eigenvectors for the first analysis of the South Florida populations.

PAGE 184

184 Table 5 9 The biol ogical distances and standard errors for those distances between the South Florida populations. Dunwody Hutchinson Island Highland Beach Margate Blount Captiva Dunwody 0.121 0.101 0.163 0.080 Hutchinson Island 0.503 0.057 0.094 0.051 Highland Beach 0.430 0.285 0.095 0.390 Margate Blount 0.724 0.317 0.463 0.920 Captiva 0.248 0.263 0.318 0.468 The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half. Figure 5 32 Graph of the first and second discriminant functions for second analysis of the South Florida populations.

PAGE 185

185 Figure 5 33 Graph of the first and second factors for the PCA analysis for the anal ysis of the Weeden Island/Alachua/St. Johns populations.

PAGE 186

186 Figure 5 34 Graph of the first and second discriminant functions for analysis of the Weeden Island/Alachua/St. Johns populations.

PAGE 187

187 Figure 5 35 Graph of the first and second eigenvectors for the analysis of the Weeden Island/Alachua/St. Johns populations.

PAGE 188

188 Table 5 10 The biological distances and standard errors for those distances between the Weeden Island/Alachua/St. Johns populations. Henderson Mound Hughes Island Browne Site 5 Bay Pines Crystal River Henderson Mound 0.073 0.162 0.100 0.078 Hughes Island 0.072 0.172 0.068 0.065 Browne Site 5 0.878 1.102 0.184 0.107 Bay Pines 0.176 0.005 1.042 0.076 Crystal River 0.292 0.218 0.645 0.202 The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half. Figure 5 36 Graph of the first and second factors for the PCA analysis for the analysis of the Weeden Island/Fort Walton populations.

PAGE 189

189 Figure 5 37 Graph of the first and second functions for the analysis of the Weeden Island/Fort Walton populations.

PAGE 190

190 Figure 5 38 Graph of the first and second eigenvectors for the analysis of the Weeden Island/Fort Walton populations.

PAGE 191

191 Table 5 11 The biological distances and standard errors for those distances between th e Weeden Island/Fort Walton populations. Hughes Island Bay Pines Crystal River Fort Walton Temple Mound Hughes Island 0.074 0.087 0.106 Bay Pines 0.026 0.080 0.115 Crystal River 0.416 0.217 0.082 Fort Walton Temple Mound 0.404 0.358 0.478 The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half. Figure 5 39 Graph of the first and second functions for the first analysis of the Weeden Island/Fort Wa lton/Manasota/Safety Harbor populations.

PAGE 192

192 Figure 5 40 Graph of the first and second eigenvectors for the first analysis of the Weeden Island/Fort Walton/Manasota/Safety Harbor populations.

PAGE 193

193 Table 5 12 The biological distances and standard errors for those distances between the Weeden Island/Fort Walton/Manasota/Safety Harbor populations Palmer Hughes Island Bay Pines Dunwody McKeithan Sarasota Bay Mound Yellow Bluffs Casey Key Crystal River Bayshore Homes Manasota Fort Walton Temple Mound Safety Harbor Palmer 0.062 0.081 0.088 0.182 0.142 0.161 0.035 0.041 0.019 0.018 0.065 0.038 Hughes Island 0.219 0.067 0.130 0.165 0.158 0.164 0.088 0.054 0.067 0.055 0.066 0.076 Bay Pines 0.268 0.000 0.123 0.178 0.208 0.201 0.116 0.063 0.076 0.068 0.081 0.093 Dunwody 0.343 0.468 0.291 0.248 0.260 0.249 0.124 0.095 0.083 0.082 0.125 0.105 McKeithan 0.560 0.229 0.239 0.796 0.311 0.271 0.198 0.146 0.167 0.187 0.146 0.193 Sarasota Bay Mound 0.242 0.182 0.457 0.904 0.762 0.256 0.168 0.190 0.167 0.150 0.176 0.150 Yellow Bluffs 0.387 0.225 0.402 0.800 0.455 0.352 0.212 0.143 0.155 0.165 0.153 0.149 Casey Key 0.035 0.243 0.381 0.458 0.504 0.279 0.627 0.085 0.053 0.048 0.101 0.075 Crystal River 0.352 0.126 0.101 0.376 0.248 0.613 0.227 0.507 0.038 0.040 0.055 0.051 Bayshore Homes 0.051 0.238 0.207 0.269 0.411 0.412 0.316 0.154 0.239 0.024 0.059 0.039 Manasota 0.024 0.114 0.117 0.226 0.556 0.261 0.374 0.088 0.214 0.052 0.067 0.043 Fort Walton Temple Mound 0.343 0.096 0.120 0.494 0.149 0.358 0.190 0.442 0.198 0.239 0.288 0.075 Safety Harbor 0.179 0.263 0.296 0.417 0.579 0.241 0.232 0.301 0.292 0.152 0.162 0.340 The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half.

PAGE 194

194 Figure 5 41 Graph of the first and second functions for the third analysis of the Manasota/Safety Harbor populations.

PAGE 195

195 Figure 5 42 Graph of the first and second eigenvectors for the third analysis of the Manasota/Safety Harbor populations. Table 5 13 The biological distances and standard errors for those distances for the third analysis of the Manasota/Safety Harbor populations Palmer Dunwody Sarasota Bay Mound Casey Key Bayshore Homes Manasota Safety Harbor Palmer 0.107 0.150 0.036 0.024 0.021 0.047 Dunwody 0.527 0.279 0.146 0.095 0.109 0.116 Sarasota Bay Mound 0.269 1.029 0.171 0.164 0.142 0.159 Casey Key 0.033 0.650 0.267 0.055 0.047 0.083 Bayshore Homes 0.094 0.354 0.352 0.155 0.030 0.039 Manasota 0.042 0.469 0.183 0.071 0.095 0.053 Safety Harbor 0.276 0.503 0.272 0.364 0.146 0.259 The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half.

PAGE 196

196 Figure 5 43 Graph of the first and second functions for the fourth analysis of the Manasota/Safety Harbor populations.

PAGE 197

197 Figure 5 44 Graph of the first and second eigenvectors for the fourth analysis of the Manasota/Safety Harbor populations. Table 5 14 The bio logical distances and standard errors for those distances for the fourth analysis of the Manasota/Safety Harbor populations Palmer Sarasota Bay Mound Casey Key Manasota Safety Harbor Palmer 0.168 0.034 0.022 0.063 Sarasota Bay Mound 0.414 0.187 0.168 0.181 Casey Key 0.023 0.385 0.048 0.100 Manasota 0.048 0.367 0.083 0.070 Safety Harbor 0.535 0.443 0.571 0.501 The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half.

PAGE 198

198 CHAPTER 6 CONCLUSIONS Summary This study utilized craniometric data to examine morphological variation as expressed in 20 craniometric variables, measured on 404 individuals from 27 sites the pre Contact po pulations. Previous research in Florida employed simple indices to interpret biological relationship s utilized only a few sites from a very broad temporal range (Seasons 2010) or focused on Contact period remains with occasional mention o f pre Contact results (Stojanowski 2003; 2004; 2005b) A c raniometric approach was chosen in concert with multivariate statistical methods to allow for the interpretation of individuals and populations as a whole, not just parts, as simple indices are pro ne to do. In addition, a larger selection of sites result ed in a better representation of various cultural and temporal frames. These craniometric data are explored using Principle Component Analysis (PCA), Discriminant Function Analysis (DFA), and R matri x Analysis. These particular methods were chosen as Principle Component Analysis seeks to explore the underlying patterns of variation without using a priori assumptions about the relationships of the individuals being analyzed Discriminant Function Analy sis tests the variation associated with predetermined groups (i.e. sites, time p eriods, geographic regions ) to examine if the variation corresponds with those groups or not L astly, R matrix Analysis determines the level of variation associated with each predetermined group, also providing biological distance and gene flow estimates This approach led to three overall objectives: 1) Examine the range of variation within the populations of Florida from the Archaic to pre Contact as represented in the

PAGE 199

199 popu l ations used in this study; 2) u nderstand if, or how this variation correlates to cultural, geographic or temporal variation; 3 ) t est the assumptions of biological distance analysis with regard to the populations in Florida. The null hypotheses were stated as: 1) There is no biological variation noted within these populations; 2) There is no correlation between biological and cultural variation or biological variation and geographic or temporal distance ; 3) The assumptions of biological distance analysis do not hold true in these archaeological samples. With regard to overall variations, there is biological variation within the sampled populations and that variation correlates with cultural, geographic, and temporal divisions. In addition, the assumptions o f biological distance analysi s hold true in the samples used biological distance is an indirect measure of social exchange, cemeteries are representative of breeding populations, and environmental affects appear to be minimal and randomly distributed. The refore all null hypotheses were rejected. In total, the results of the biological analyses combined with archaeological evidence suggested that there were at least two significant migration events into Florida after the Paleoindian period. A migration occ urred in the Early to Middle Archaic as evidenced by the distinction between Warm Mineral Springs and the other Archaic populations (Figure 5 3). This finding is supported by changes in stone tool technology (Faught and Waggoner Jr. 2012) and burial patter ns, including the shift to water burials (Cockrell and Murphy 1978; Doran 2002) The second migration occurred in the Late Archaic as represented by the distance between the Archaic populations and the Bird

PAGE 200

200 Island sample, the other sites used in the study (Figure 5 2). The archaeological evidence supports these findings, particular the cessation of water burials and changes in material culture; these changes are exemplified in the Bird Island population as they utilized soa pstone vessels which are not found at the other Archaic sites In addition, t he burials of Bird Island were not associated with a large freshwater source (Lovejoy 1985; Stojanowski and Doran 1998; Yates 2000) With regard to geographic variation, this re search did not find any evidence to support the previous North/South distinction hypothesized by (1922) and Stojanowski (2004; 2005 b ) However, there is some evidence to support East/West distinctions in certain Archaic samples (Bay West, Little S alt Spring, Gauthier, and Windover) and in the South Florida samples (Margate Blount, Highland Beach, Hutchinson Island Captiva and Dunwody ) (Figures 5 14 and 5 30 respectively). This is most likely due to variations in mate selection possibly related to variations in trade patterns. The other samples included in this study do not provide evidence for East/West patterns, suggesting that those populations were possibly utilizing different trade routes or others factors resulted in differing mate selection methodologies It should be noted that West Coast samples outnumber East Coast samples and North Florida samples outnumber South Florida samples, thereby limiting the strength of these results and neither proving nor disproving in terpretations. Larger samples and inclusion of would better address this question. After the above findings established that there are patterns to the biological variation within the population samples, a more detailed analysis of the Archaic,

PAGE 201

201 Weeden Island, Manasota, and South Florida samples was conducted. The analysis of the Archaic s amples suggested that there was a core population that inhabited the peninsula for approximately 5000 years and buried their dead at Windover, Little Salt Spring, Bay West, Republic Groves, and Gauthier. However, biodistance analysis also suggested that gene flow for Republic Groves was different from the other sites (Figures 5 11 and 5 12; Table 5 4) as this sample appear s to be distinct based on PCA and DFA analyses, from the others a finding also supported by the levels of variation and biological distance determined by R matrix analysis. The biological variation noted in the Weeden Island and Manasota samples indicate biological distinction s between the populations buried at these sites (i.e. t he populations associated with these cultures did not exchange mates suggesting less social interaction ) Evidence suggest s that they had a common biological origin during, or prior to the Deptford period as seen in the close relationship of the individuals buried at Crystal Rive r and Yellow Bluffs (Figure 5 16) B y the later periods (after ca. 1800 BP, based on radiocarbon dates associated with the sites) these populations were distinct as seen by the bio Manasota p opulations (Figure 5 18). In addition, Crystal River correctly classifies using DFA more often than any other population indicating that it is a distinct and closely related group It is of particular interest that the Weeden Island sites show some biolo gical distancing between mound centers as represented by the Crystal River, McKeithan, and Hughes Island populations (Figure 5 21). T he Manasota populations are all closely related biologically (Figures 5 25 and 5 27). In total these findings support Luer (1982) statements that the Manasota culture was the result of

PAGE 202

202 a localized cultural florescence unrelated to the early Weeden Island Complex Though the Weeden Island Complex and the Manasota culture m ay have shared a common ancestral population they quickly distinguished themselves both biologically and culturally. It is apparent that the pottery and mortuary practices of the late Manasota culture were influenced by the Weeden Island Complex H owever the potential biological influences cannot be fully tested at this time due to a lack of biological material from late Weeden Island sites. Further analysis is needed on additional biological samples to fully understand the biological and cultural complexities of the Weeden Island Complex and it in fluence on various other cultures. As mentioned above the pattern noted in the South Florida samples is consistent with East/West divisions and geographic distance appears to impact the biological relationships of these groups. This appears to correlate w ith the cultural divisions between the East and West Coast of South Florida. The impact of outside influences and the effect s of the Calusa on the region in later periods could not be effectively tested with the samples utilized Additional samples from the region and from more diverse time periods, including the Mississippian period, would allow for further analysis of these questions. The analyses of the Alachua, St. Johns, Fort Walton, and Safety Harbor populations are limited due to single sample populations from these culture labels. N onetheless, the results suggest ed that further analyses are warranted. A necessarily limited interpretation suggest s the following. The Alachua culture may be a local culture change event as thi s sample appears to be closely related to the other Weeden Island samples from the area (Figure 5 34), despite the drastic change in material culture and

PAGE 203

203 site s ettlement patterns that have le d other s to hypothesize a migration event establishing this cultu re in North Central Florida (Milanich 1994) Safety Harbor also appears to be closely related biologically to the Manasota samples from the area, which is supported by similarities in site selection, mortuary practices, and pottery styles (Luer and Almy 19 87) However an increase in biological variation and a slight biological separation of the Manasota and Safety Harbor populations suggested an influx of people or new mate sources (Figure 5 43). The Fort Walton population appears to be distinct from the We eden Island and Safety Harbor populations, while the St. Johns population i s also distinct from its Weeden Island neighbors suggesting biological separation between these populations (Figures 5 39 and 5 34 respectively ). W hether this is due to differing a ncestor populations, isolation through time by cultural divisions, or differing mate sources cannot be elucidated at this time. Analysis of the biological variability within various time periods (Archaic, Woodland, and Mississippian) and cultural labels ( Weeden Island, Manasota, Alachua, etc.) provided a model of population variability and structure through time in Florida. The Archaic is a moderately variable population suggesting more open mate sources or more open cemetery structures. In the transition from the Archaic to the Woodland, the overall level of variability increased, most likely due to more strictly defined cultural barriers to gene flow related to an increase in cultural diversity. This variation appears to adhere to distinct culturally defi ned patterns. The populations associated with the Weeden Island Complex exhibited a high level of biological variation that is most likely due to either the limitation of gene flow between sites or each mound center having different external sources of gen e flow. In either case, this demonstrates recognition of

PAGE 204

204 socially or culturally defined barriers to mate selection. T he Manasota culture appears to have a more open strategy for mate selection because these populations were more homogenous suggesting gene flow between sites. South Florida populations appear to have been more heavily impacted by geographic barriers to gene flow between the East and West Coasts, however the small number of sites and lack of data from the central region of South Florida limi t this interpretation. Lastly, the Mississippian period had another increase in variability, though interpretation is limited. Nonetheless, it is suggested that this increased variability was due to an increase in regionalization and a shift to more insul ar cultural/social units resulting in very little gene flow between cultures thereby increasing morphological diversity between subpopulations Migration into the area from external sources could have also played a role in the increase variability in the Mississippian period, but further research is needed to evaluate this possibility. Limitations t o This Research Several issues limited the interpretation of this research and if addressed in later studies, significant strides can be made in understanding the correlations between biology and culture in pre Contact Florida. First, small sample sizes decrease the probative value of statistical analyses and introduce sampling error. S ome will argue that such samples should not be included in analys e s. Certainl y, large samples would strengthen the statistical results; however by limiting samples particular when examining poorly preserved skeletal remains we also limit the types of questions we may ask of the archaeological record and bioarchaeological research thereby narrowing our understanding of biological and cultural variation in Florida and the Southeast. Therefore, the results of this research may be used to add to the interpretation and

PAGE 205

205 understanding of cultural and biological change in pre Contact Florida. Indeed, further efforts to locate additional collections may increase sample sizes associated with particular cultures, geographic regions, or temporal frames thus strengthening results and subsequent interpretation s A se cond limi t at i on to this study is the low level of variance explained by the first two components/factors/vectors in most of the analyses. It is desired that most of the variance, i.e. 80% or more, be explained in the first two components/factors/vectors if the variables used are suitable for explaining the variation found in the populations. However, this was not the case in this study; this is probably not because the variance is poorly explain by the variable used but is most likely because there is a va riable missing from the analysis culture. However, quantifying cultural variation is a difficult task, making its inclusion as a variable problematic thereby limiting our ability to test its impact on the total variance explained. However, when considerin g cultural labels the analysis of the results suggest that culture is the missing variable accounting for at least some of the variation seen in these populations. Future testing with culture as a variable or cultural components as variables m a y bear out t his supposition. A third limitation to this study is the methodology used to fill in missing data. It could be suggested that the choice of sites for regression analysis may have created false biological proximities, such as in the Dunwody and Captiva Isla nd populations. Yet, the results of the Yellow Bluffs analysis suggest that this may not be a significant concern, as the missing data were filled in using the other Manasota populations and subsequent analyses showed that the Yellow Bluffs population is s till quite distinct from the other Manasota samples. This issue, like the first, cannot be overcome until there

PAGE 206

206 are sufficiently large samples to allow for completion of missing data through intrasite analysis as opposed to intersite analysis. The fourth limitation is the lack of archaeological data for some sites; this of course limits the comparison of the archaeological relationships of sites and can make it difficult to interpret the intrasite relationships of the individuals, such as the lack of prove nience data for the individuals excavated at Crystal River. Fortunately, this can be corrected, to an extent, through continued research of records and new archaeological investigation. However, some of these data are simply gone and cannot be reproduced. Despite these limitations, the data presented in this dissertation are the most comprehensive in Southeast ern North America utilizing 27 sites and 402 individuals from a wide range of cultural labels and temporal frames. Further these data have confirme d that there is patterned pre Contact populations Future Research There are a number of avenues for future research. The first is obtaining additional samples from within the peninsula to further test some of the fin dings presented here, particularly samples from sites related to the Weeden Island, Alachua, St Johns, and Fort Walton cultures. Also sites from South Florida and the Panhandle region are conspicuous in their absence Future research may also include sites from the broader Southeast ern United States In particular, populations from the Louisiana area could be used to test the possible relationship of the Weeden Island Complex to the Marksville culture. T he study of populations from Alabam a or Georgia could help determine the origins of the Bird Island population. Additionally, comparisons with other

PAGE 207

207 Southeast population s could enhance the understanding of migration origins and reveal other relationsh ips that could indicate smaller scale mi grations, which could have influenced the biological variation of some populations (e.g., those associated with the Fort Walton or Safety Harbor cultures ) Broader I mpact The broader impact of this research unproven ienced remains. Currently, the use of these remains in biological and archaeological studies is avoided. The results presented here suggest that it may be possible to determine the associations of unprovenienced individuals to a particular cultur al group o r in some instances sites; resulting in more samples being available for analysis and providing new details for the investigation of questions in the archaeological record. This research also impacts efforts to repatri ate human remains to legally designated tribal groups. A key question in the analysis of remains that fall under NAGPRA is cultural affiliation. There are tens of thousands of remains in the United States listed as culturally unaffiliated T his is often be cause there is no modern Federally recognized tribe, but in other instances it is because there is no provenience for the remains However, this research has shown that with the right comparative samples it may be possible to link unprovenienced remains to specific culture labels, time periods, or geographic zones; thereby providing additional information to the institutions and tribes involved in cultural affiliation assessments conducted under the auspices of NAGPRA. Additionally the study of modern tribe s may help elucidate tribal origins, relationships to other tribes, or relationships to archaeological material providing new insight into archaeological questions and historical queries.

PAGE 208

208 APPENDIX A DEFINITION OF MEASUREMENTS 1. Maximum Length (g op): distan ce between glabella (g) and opisthocranion (op) in the midsagittal plane, measured in a straight line. 2. Maximum Breadth (eu eu): maximum width of skull perpendicular to midsagittal plane 3. Bizygomatic breadth (zy zy): direct distance between most lateral po ints on the zygomatic arches (zy zy). 4. Basion Bregma Height (ba b): direct distance from the lowest point on the anterior margin of foramen magnum (ba), to bregma (b). 5. Cranial Base Length (ba n): direct distance from nasion (n) to basion (ba). 6. Basion Prosthion Length (ba pr): direct distance from basion (ba) to prosthion (pr). 7. Maxillo Alveolar Breadth (ecm ecm): maximum breadth across the alveolar borders of the maxilla measured on the lateral surfaces at the location of the second maxillary mo lars (ecm). 8. Maxillo Alveolar Length (pr alv): direct distance from prosthion (pr) to alveolon (alv). 9. Biauricular Breadth (au au): least exterior breadth across the roots of the zygomatic processes (au). 10. Upper Facial Height (n pr): direct distance from na sion (n) to prosthion (pr) 11. Minimum Frontal Breadth (ft ft): direct distance between the two frontotemporale (ft). 12. Upper Facial Breadth (fmt fmt): direct distance between the two external points on the frontomalar suture (fmt). 13. Nasal Height (n ns): direct distance from nasion (n) to the midpoint of a line connecting the lowest points of the inferior margin of the nasal notches (ns). 14. Nasal Breadth (al al): maximum breadth of the nasal aperture (al al). 15. Orbital Breadth (d ec): laterally sloping distance fr om dacryon (d) to ectoconchion (ec). 16. Orbital Height: direct distance between the superior and inferior orbital margins. Instrument: sliding caliper.

PAGE 209

209 17. Biorbital Breadth (ec ec): direct distance between right and left ectoconchion (ec). 18. Interorbital Breadt h (d d): direct distance between right and left dacryon (d). 19. Frontal Chord (n b): direct distance from nasion (n) to bregma (b) taken in the midsagittal plane. 20. Parietal Chord (b l): direct distance from bregma (b) to lambda (1) taken in the midsagittal p lane. 21. Occipital Chord (l o): direct distance from lambda (l) to opisthion (o) taken in the midsagittal plane. 22. Foramen Magnum Length (ba o): direct distance from basion (ba) to opisthion (o). 23. Foramen Magnum Breadth: distance between the lateral margins o f foramen magnum at the points of greatest lateral curvature. 24. Mastoid Length: vertical projection of the mastoid process below and perpendicular to the eye ear (Frankfort) plane. 25. Chin Height (id gn): direct distance from infradentale (id) to gnathion (gn ). 26. Height of the Mandibular Body: direct distance from the alveolar process to the inferior border of the mandible perpendicular to the base at the level of the mental foramen. 27. Breadth of the Mandibular Body: maximum breadth measured in the region of the mental foramen perpendicular to the long axis of the mandibular body. 28. Bigonial Width (go go): direct distance between right and left gonion (go). 29. Bicondylar Breadth (cdl cdl): direct distance between the most lateral points on the two condyles (cdl). 30. M inimum Ramus Breadth: least breadth of the mandibular ramus measured perpendicular to the height of the ramus. 31. Maximum Ramus Breadth: distance between the most anterior point on the mandibular ramus and a line connecting the most posterior point on the co ndyle and the angle of the jaw. 32. Maximum Ramus Height: direct distance from the highest point on the mandibular condyle to gonion (go).

PAGE 210

210 33. Mandibular Length: distance of the anterior margin of the chin from a center point on the projected straight line placed along the posterior border of the two mandibular angles. 34. Mandibular Angle: angle formed by the inferior border of the corpus and the p osterior border of the ramus. 35. Mastoid length (2): distance between inferior anterior margin of the parietal notch and the most inferior point of the mastoid process. 36. Mastoid breadth: distance between porion (p) and the most anterior edge of the superior terminus of the digastric groove. 37. Occipital condyle length: maximum length of the occipital condyle along the long axis of the condyle, measured from the edges of the auricular surface. 38. Occipital condyle breadth: maximum breadth of the occipital condyle, m easured perpendicular to the length axis from the edges of the auricular surface. 39. Orbital breadth at nasion: nasal width taken at the level of nasion (n). 40. Nasion to fmt: distance from nasion (n) to the external point on the frontomalar suture (fmt), measur ed in a horizontal plane.

PAGE 211

211 APPENDIX B RAW DATA Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. WMS 5 5 178 126 114 90 100 44 37 30 107 120 23 50 28 WMS 1 1 175 128 107 92 114 88 99 44 35 31 103 109 85 32 29 24 49 25 24 13 WMS 5 5 188 134 44 31 113 117 47 35 Gauthier 1 1 178 140 127 97 91 106 53 39 35 104 118 93 37 31 27 53 29 26 15 Gauthier 5 1 184 132 133 103 127 90 103 112 103 100 37 30 26 51 30 28 14 Gauthier 4 5 181 114 112 26 50 29 Gauthier 1 1 171 135 125 99 118 93 100 48 41 29 101 103 96 36 29 29 28 26 16 Gauthier 5 5 181 137 132 100 123 94 103 107 110 100 34 28 28 52 28 26 16 Gauthier 5 1 184 138 136 101 130 56 42 38 114 113 100 37 28 27 54 34 25 15 Gauthier 1 1 170 133 132 95 119 88 100 54 37 36 105 110 96 36 30 24 49 25 26 12 Gauthier 3 6 182 137 132 94 124 84 101 107 120 96 34 29 24 50 26 26 15 Gauthier 5 6 163 115 107 93 28 53 28 Gauthier 1 6 172 130 94 107 104 110 28 50 27 12 Gauthier 5 6 181 142 131 101 130 98 104 110 118 95 35 30 26 51 33 26 13 Gauthier 1 1 182 136 133 102 122 89 102 52 39 108 105 104 36 29 25 50 26 29 14 Gauthier 5 3 182 140 137 103 124 90 99 116 111 101 36 32 29 51 30 26 14 Gauthier 1 6 177 136 126 96 106 117 104 109 28 52 26 25 16 Gauthier 1 6 178 131 134 97 116 92 105 108 120 100 36 28 24 51 27 22 16 Gauthier 5 5 181 114 124 158 97 30 58 Gauthier 1 5 188 137 135 101 125 96 106 114 124 94 36 31 28 54 29 26 13 Gauthier 4 5 183 135 122 96 107 113 113 105 29 51 30 Gauthier 5 5 187 139 143 103 100 109 118 119 104 40 33 30 53 30 28 15 Gauthier 5 5 189 138 147 111 101 114 121 121 103 35 29 31 58 29 25 15 Gauthier 1 1 170 135 134 100 127 93 104 50 107 105 91 37 30 25 50 24 26 14 Gauthier 5 5 184 136 134 104 93 107 52 40 34 114 109 32 25 53 30 26 14 Gauthier 5 6 187 141 145 107 130 97 105 116 108 102 41 29 52 34 15 Gauthier 1 3 186 141 142 98 120 87 104 116 112 111 35 29 25 51 30 28 15 Gauthier 1 6 182 115 108 107 30 29 LSS 2 1 86 102 51 37 31 107 24 49 LSS 5 3 92 103 51 39 33 25 51 LSS 1 1 83 99 54 39 32 111 23 48 Key: Sex: 1= Female, 2= Female?, 3= unknown, 4= Male?, 5=Male; Age: 1= Young; 2= Mid young, 3= Middle Age; 4= Mid old 5= Old, 6= Unknown

PAGE 212

212 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. LSS 4 5 180 139 103 131 92 106 52 39 33 106 112 102 37 25 54 32 LSS 5 3 189 135 143 103 127 95 107 119 123 98 40 24 55 32 LSS 2 5 90 102 50 38 33 109 22 52 Windover 5 3 183 135 138 100 122 98 108 52 41 34 114 114 99 34 24 54 29 Windover 1 3 178 129 144 103 124 96 107 53 39 33 112 114 96 36 29 24 53 30 27 14 Windover 5 1 188 136 123 96 106 56 39 33 118 110 31 54 26 Windover 5 3 188 142 146 108 137 95 103 55 39 34 115 109 107 39 32 27 51 33 32 15 Windover 1 3 174 131 134 101 118 91 104 52 39 35 110 110 94 36 35 24 25 25 13 Windover 5 5 185 140 138 98 119 92 106 40 34 115 27 54 28 30 13 Windover 5 5 184 132 130 102 123 87 102 55 37 33 106 103 99 37 30 25 49 33 25 14 Windover 1 1 175 126 130 96 122 91 102 48 37 33 110 111 94 35 32 24 50 28 24 13 Windover 2 6 182 128 114 93 103 115 115 50 28 Windover 1 5 193 142 125 99 106 52 40 35 114 118 112 22 33 Windover 1 5 175 134 119 88 97 45 38 32 113 120 24 48 25 Windover 5 3 187 132 127 94 50 106 113 28 33 Windover 5 5 171 135 124 92 102 112 112 96 23 51 26 Windover 5 3 181 142 146 106 128 92 107 51 40 34 118 115 103 37 34 24 33 28 17 Windover 5 1 189 133 139 109 123 96 109 54 42 34 112 117 31 54 28 26 16 Windover 1 5 184 136 118 91 113 116 98 30 24 51 24 24 13 Windover 5 3 180 138 145 106 124 86 101 114 112 98 39 35 28 52 28 Windover 1 1 177 127 49 40 31 104 125 24 49 23 25 12 Windover 1 1 177 125 115 87 100 100 102 98 24 51 23 22 10 Windover 1 1 180 131 133 101 120 91 100 47 40 32 108 112 97 34 30 24 48 30 13 Windover 5 1 168 124 130 97 117 84 99 50 37 35 104 111 92 32 22 47 27 24 13 Windover 5 3 178 137 91 107 106 117 25 51 23 Windover 5 5 185 129 146 105 129 96 109 55 114 110 104 38 29 55 28 30 13 Windover 5 5 192 144 97 106 109 119 101 54 34 Windover 1 1 178 140 95 109 119 54 23 Windover 5 1 188 136 140 102 126 91 103 54 40 31 112 119 105 38 25 25 Windover 4 5 175 129 136 101 115 96 104 50 39 33 113 116 92 35 24 51 28 25 13 Windover 5 1 183 136 152 105 120 94 105 116 120 31 52 27 14 Windover 5 5 196 135 135 100 114 54 44 33 115 112 37 28 58 32 Windover 5 5 185 137 117 99 105 113 117 96 28 53 30 Windover 5 3 186 132 137 108 128 95 106 52 40 33 116 114 29 26 55 28 26 15 Windover 4 1 187 136 126 105 109 53 45 37 116 120 94 57 26

PAGE 213

213 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. Windover 5 3 181 141 128 94 110 57 39 35 112 110 102 27 55 33 Windover 1 1 184 130 125 95 103 108 106 106 24 54 23 Windover 4 5 177 127 114 83 96 109 113 96 26 48 25 Windover 2 5 177 136 126 92 101 107 113 25 51 30 Windover 2 1 185 135 134 104 127 88 101 112 120 30 50 30 25 14 Windover 5 3 187 143 137 109 132 97 108 53 42 29 113 116 29 54 34 Windover 2 3 171 126 112 88 103 54 38 36 105 111 94 26 51 28 Windover 5 1 195 128 97 107 41 34 115 115 26 54 28 25 14 Windover 5 2 190 138 137 106 127 90 107 50 38 32 112 121 98 38 32 26 55 32 29 15 Windover 1 1 178 135 118 94 100 48 39 34 104 112 108 26 49 29 Windover 2 1 175 128 136 103 114 92 103 51 39 30 111 113 94 35 32 27 52 26 26 15 Windover 5 3 195 135 119 90 104 54 109 126 99 27 50 33 Windover 5 5 185 137 101 117 105 102 25 52 28 Windover 5 1 183 135 127 92 106 54 108 117 93 25 53 34 Windover 1 1 181 134 96 103 110 113 99 53 27 Windover 5 5 182 135 123 95 104 110 118 25 52 32 Windover 5 5 190 136 121 102 110 112 119 102 31 54 31 Windover 5 1 187 147 145 106 129 90 112 54 43 33 119 103 109 36 31 28 57 26 28 16 Windover 5 1 187 130 130 93 109 55 41 35 110 115 26 54 33 Windover 2 3 183 142 141 106 129 98 107 57 40 34 114 113 104 38 31 28 53 31 29 14 Windover 5 5 185 137 129 98 105 52 39 34 116 119 98 27 53 33 Windover 1 1 173 137 140 99 122 84 97 53 108 110 103 38 27 49 26 27 12 Windover 5 5 182 125 92 105 51 39 35 105 108 102 27 52 27 Windover 4 1 170 129 122 95 120 91 101 47 38 31 101 107 91 35 31 25 51 25 26 Windover 2 6 179 131 128 99 121 85 100 103 107 103 36 31 23 50 28 Windover 2 1 175 127 130 100 117 88 102 102 109 97 35 29 26 52 31 25 15 Windover 5 3 179 136 136 105 132 92 109 113 109 98 39 35 27 54 29 12 Windover 5 3 186 131 138 106 131 100 110 58 41 33 111 106 105 39 35 28 54 28 27 12 Windover 5 1 190 137 142 106 124 91 107 49 40 31 113 121 101 36 33 27 52 33 30 16 Windover 5 1 180 130 140 103 121 93 104 49 39 33 109 108 103 36 30 25 52 29 26 12 Windover 2 1 180 145 140 101 131 90 104 54 38 30 112 117 99 36 33 29 51 27 28 14 Bay West 5 3 176 139 86 100 106 112 106 33 26 51 33 Bay West 1 1 172 130 120 86 98 105 104 91 30 26 48 30 Bay West 4 1 173 132 131 95 82 99 111 103 95 37 31 25 50 28 14 Bay West 4 1 171 131 136 98 130 81 97 103 111 92 38 24 50 25 25 15

PAGE 214

214 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. Bay West 2 3 177 136 129 96 84 98 103 100 106 36 29 20 49 25 14 Bay West 2 3 182 142 133 102 92 102 53 36 34 110 108 101 34 27 27 51 14 Bay West 5 1 180 141 142 105 97 106 54 41 36 112 111 96 38 33 29 52 27 13 Palmer 5 3 176 143 136 101 132 96 109 51 41 34 115 114 90 37 29 52 30 27 13 Palmer 5 1 174 122 96 106 54 107 110 26 50 25 Palmer 2 1 175 134 95 105 36 109 112 25 52 27 Palmer 1 3 178 140 134 100 122 96 107 53 41 36 108 112 96 38 30 26 53 27 24 12 Palmer 1 4 175 132 134 96 122 96 105 114 111 92 37 28 26 53 25 22 14 Palmer 1 4 135 90 105 105 28 52 26 Palmer 5 4 103 116 131 32 57 30 Palmer 5 4 186 95 107 110 30 Palmer 5 3 190 146 133 100 110 33 122 114 98 28 32 53 35 Palmer 1 1 172 136 111 102 24 49 28 Palmer 5 4 100 110 110 114 29 55 34 Palmer 1 1 92 108 48 35 108 101 27 54 29 25 12 Palmer 4 3 176 96 110 120 110 30 56 32 Palmer 1 4 173 94 102 117 105 92 52 29 Palmer 2 1 182 149 129 97 122 118 28 52 29 Palmer 3 6 145 35 104 107 30 Palmer 5 5 183 140 143 101 128 95 108 52 41 34 119 113 100 36 27 27 56 33 27 12 Palmer 2 4 191 141 124 124 120 54 26 Palmer 1 4 165 135 130 94 126 96 107 43 109 103 92 39 31 55 29 24 13 Palmer 5 4 193 142 142 105 131 96 109 50 44 34 117 120 96 39 33 24 54 33 28 12 Palmer 5 4 175 90 115 115 104 58 33 Palmer 4 4 172 141 135 96 109 111 118 90 56 30 Palmer 1 1 178 134 138 100 126 88 106 51 42 35 116 108 96 36 30 28 53 31 26 14 Palmer 5 4 179 100 112 114 97 59 31 Palmer 1 4 174 131 134 103 126 92 105 51 42 36 107 111 94 35 28 28 51 29 26 15 Palmer 2 1 176 134 140 103 125 94 55 41 36 110 110 97 36 29 24 52 32 25 14 Palmer 1 5 180 136 130 100 122 94 105 50 40 36 112 111 98 32 28 25 50 30 24 15 Palmer 5 5 190 141 148 108 131 97 109 116 126 96 39 29 30 55 33 28 10 Palmer 2 3 178 150 141 103 136 101 112 55 44 36 113 106 109 39 31 26 56 33 28 16 Palmer 4 1 186 143 146 107 138 95 105 52 41 35 116 117 97 36 30 54 30 29 12 Palmer 1 5 172 131 131 97 120 86 99 48 39 35 110 106 97 36 30 23 49 26 27 16 Palmer 2 3 114 103 99 27 54

PAGE 215

215 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. Palmer 1 1 185 145 137 101 129 101 107 46 39 32 116 111 100 36 29 30 27 14 Palmer 4 3 173 146 95 103 112 102 54 30 Palmer 5 4 184 143 143 110 130 100 112 109 116 97 38 34 27 54 36 27 12 Palmer 5 3 189 146 129 94 106 122 106 108 27 52 30 Palmer 1 1 171 145 136 98 126 93 102 49 39 35 109 108 98 37 32 23 52 30 Palmer 5 4 184 148 139 105 135 93 110 52 43 36 115 107 94 39 30 27 56 34 29 15 Palmer 3 5 172 140 122 90 106 114 110 100 28 54 29 Palmer 1 2 174 133 85 103 50 40 35 111 112 95 32 28 14 Palmer 5 5 177 145 132 90 104 38 119 111 98 25 27 Palmer 5 5 186 146 109 133 97 108 122 118 105 36 30 28 54 33 31 15 Palmer 5 1 168 138 92 101 106 94 105 24 51 30 Palmer 5 2 184 147 135 100 109 50 43 35 115 118 98 26 55 34 Henderson 4 3 145 101 53 28 26 15 Henderson 5 1 175 110 108 101 26 54 28 Henderson 2 6 100 96 34 29 52 24 23 14 Henderson 2 6 Henderson 2 6 144 125 97 34 34 25 27 13 Henderson 3 6 165 134 119 95 113 107 53 23 Hughes Island 5 5 107 28 53 29 Hughes Island 5 3 115 102 29 53 35 26 14 Hughes Island 2 4 101 31 Hughes Island 4 5 167 107 114 51 26 Hughes Island 5 5 176 92 112 100 56 29 Hughes Island 2 6 137 110 96 100 38 27 51 25 25 14 Hughes Island 5 5 108 119 111 95 33 29 60 30 Browne Site 5 1 4 161 147 131 94 131 93 102 50 39 33 106 97 92 33 31 27 50 28 28 15 Browne Site 5 5 5 180 144 145 106 113 134 29 54 32 27 12 Browne Site 5 1 1 140 104 31 27 50 30 23 12 Browne Site 5 5 6 133 106 38 31 41 Browne Site 5 5 3 179 142 142 107 134 111 115 99 36 30 23 53 34 23 17 Browne Site 5 2 1 88 102 105 95 49 27 Bay Pines 2 6 174 139 130 91 110 107 103 30 55 32 Bay Pines 1 2 173 148 135 101 130 96 107 57 42 36 111 112 88 34 30 26 54 35 28 13 Bay Pines 2 6 184 137 144 104 93 102 115 115 104 36 28 23 50 28 25 14 Bay Pines 5 6 183 114 109 26 53 28

PAGE 216

216 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. Bay Pines 1 1 169 131 98 125 96 108 54 39 36 105 106 95 32 30 25 52 26 Dunwody 5 5 170 123 94 109 107 102 98 58 32 Dunwody 1 2 170 139 126 52 41 112 110 95 33 Dunwody 2 5 165 147 130 94 103 104 99 25 52 29 27 14 Dunwody 1 1 158 142 127 82 129 91 99 111 100 98 36 22 50 30 25 13 Dunwody 5 2 175 139 95 105 112 110 28 54 33 28 13 McKeithan 5 1 165 152 137 102 133 52 37 36 112 92 107 27 36 50 30 25 12 McKeithan 4 4 112 104 90 49 23 Venice Beach Complex 5 3 177 138 140 108 130 97 111 56 41 35 117 105 98 33 28 28 54 33 23 14 Sarasota Bay 5 5 187 148 145 107 138 92 105 53 40 34 118 109 104 38 35 28 55 32 29 Sarasota Bay 5 4 190 145 147 106 134 88 106 51 41 35 115 120 101 37 33 29 52 29 Yellow Bluffs 5 3 181 160 143 104 111 117 108 87 29 55 35 Yellow Bluffs 1 1 160 151 130 94 56 110 92 93 51 22 Casey Key 1 1 171 141 132 95 117 95 105 48 39 35 110 104 95 37 34 24 28 24 13 Casey Key 4 4 175 144 144 104 126 90 104 50 39 36 118 110 101 34 30 24 53 29 29 15 Casey Key 2 2 174 138 140 99 126 90 50 39 35 116 109 96 35 29 27 28 25 13 Casey Key 4 4 181 137 135 95 122 88 99 50 39 37 117 101 109 36 28 21 48 29 27 13 Casey Key 2 3 180 149 139 100 139 100 112 117 107 114 34 28 27 56 29 23 12 Casey Key 5 3 182 138 134 96 128 96 107 50 42 35 116 112 98 42 31 27 56 33 28 13 Casey Key 5 5 172 141 135 98 126 96 101 49 40 35 112 107 94 37 30 25 51 28 27 15 Casey Key 5 3 184 144 136 103 128 97 106 50 118 109 99 35 38 26 55 32 25 12 Crystal River 2 3 163 148 130 94 127 90 101 56 39 36 109 105 92 34 24 50 30 12 Crystal River 5 1 175 142 140 104 125 89 105 53 39 36 104 110 105 33 29 28 26 12 Crystal River 5 5 35 10 Crystal River 1 2 164 137 129 100 85 103 105 105 89 29 28 25 24 23 13 Crystal River 2 5 163 136 104 91 27 Crystal River 4 5 150 103 108 27 22 13 Crystal River 5 6 168 149 136 107 107 107 29 Crystal River 5 3 92 37 32 30 24 Crystal River 2 5 144 102 28 Crystal River 5 4 160 137 141 100 130 89 103 108 103 92 35 21 30 23 14 Crystal River 4 3 165 143 95 106 97 105 29 28 Crystal River 5 3 172 136 136 100 133 91 105 54 43 35 105 97 97 34 27 28 28 27 11 Crystal River 5 5 110 26

PAGE 217

217 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. Crystal River 2 5 121 99 36 30 28 24 14 Crystal River 2 5 161 148 140 96 129 92 104 110 104 98 32 27 23 27 26 15 Crystal River 5 5 97 30 Crystal River 5 3 110 112 28 29 Crystal River 5 3 174 141 127 100 131 97 104 55 40 34 112 92 97 38 30 26 29 29 Crystal River 4 6 181 152 140 104 138 106 113 34 112 110 102 40 28 29 28 14 Crystal River 1 3 166 136 130 97 121 88 100 110 101 91 33 30 25 26 28 14 Crystal River 4 5 160 148 141 94 127 90 112 109 95 36 28 24 28 24 15 Crystal River 5 1 183 116 106 24 32 29 12 Crystal River 5 4 139 101 29 Crystal River 4 5 173 141 134 107 31 Crystal River 5 3 176 150 136 99 137 103 108 39 37 118 106 92 38 33 27 34 Crystal River 1 1 160 137 132 97 125 86 102 55 38 33 103 97 94 37 32 26 27 27 Crystal River 5 4 170 90 106 117 109 33 35 Crystal River 1 6 92 105 111 27 Bayshore Homes 5 3 173 148 141 105 141 97 114 57 113 112 96 35 31 57 32 27 12 Bayshore Homes 5 3 179 149 146 107 136 96 110 55 42 38 119 103 103 41 32 28 55 32 33 13 Bayshore Homes 2 1 166 136 140 96 122 88 103 50 38 35 112 106 101 36 30 25 51 33 28 11 Bayshore Homes 5 5 178 148 139 100 106 53 42 38 112 113 24 53 33 Bayshore Homes 5 1 175 139 143 105 131 92 107 53 43 37 111 111 99 34 30 28 52 32 25 12 Bayshore Homes 5 1 190 148 100 112 52 44 35 123 100 111 36 26 30 56 36 Bayshore Homes 5 3 173 139 137 100 130 97 112 51 42 35 111 104 94 34 29 27 55 31 26 15 Bayshore Homes 5 4 171 145 130 101 134 96 107 115 98 91 32 25 54 29 24 Bayshore Homes 5 4 144 141 137 98 112 109 98 39 34 23 15 Bayshore Homes 5 6 175 146 147 101 103 111 113 113 96 35 31 28 56 37

PAGE 218

218 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. Bayshore Homes 5 4 173 148 109 136 99 112 58 44 36 115 113 101 35 28 56 32 25 14 Bayshore Homes 5 1 176 144 88 110 115 103 56 28 Bayshore Homes 2 6 170 141 128 92 49 34 106 98 26 55 23 Bayshore Homes 2 1 166 134 97 108 110 90 29 53 32 Bayshore Homes 5 6 177 135 124 112 102 27 56 34 Bayshore Homes 5 4 167 144 137 99 130 92 107 51 38 33 116 102 98 30 30 29 53 29 29 13 Bayshore Homes 1 5 168 142 131 92 108 52 41 36 108 105 100 26 53 25 Bayshore Homes 5 3 181 142 139 106 133 99 111 55 43 35 116 104 100 38 32 30 55 31 30 11 Bayshore Homes 2 4 53 42 55 28 27 14 Bayshore Homes 1 4 176 146 134 98 53 41 37 113 103 105 24 55 30 Bayshore Homes 5 3 95 57 41 38 113 28 55 27 Bayshore Homes 1 3 164 138 141 101 129 83 98 49 36 108 100 103 36 30 23 48 23 29 13 Bayshore Homes 1 1 166 135 131 104 125 92 106 58 40 36 107 103 93 36 30 25 54 28 24 14 Bayshore Homes 2 3 141 91 104 50 41 34 52 29 Bayshore Homes 1 1 164 134 90 101 50 86 122 102 25 50 29 Bayshore Homes 1 5 169 135 93 103 108 99 50 24 Bayshore Homes 5 3 174 140 127 88 108 113 108 97 25 53 31

PAGE 219

219 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. Bayshore Homes 2 2 164 137 137 97 127 94 108 52 41 35 110 94 98 36 29 27 53 32 24 15 Bayshore Homes 2 5 179 141 152 101 48 42 35 122 114 110 33 26 26 29 27 11 Hutchinson 5 6 191 148 141 101 131 91 109 120 108 110 36 29 29 54 29 30 12 Hutchinson 1 6 171 140 107 50 31 Hutchinson 4 6 173 140 134 128 92 105 98 103 32 29 25 23 12 Hutchinson 2 6 141 131 83 100 40 34 108 23 52 Hutchinson 1 6 161 85 93 102 102 95 29 Hutchinson 5 5 180 121 51 30 Hutchinson 4 5 170 140 136 94 130 115 105 96 35 54 30 23 16 Hutchinson 4 5 168 147 135 88 111 55 41 37 112 95 105 57 31 Hutchinson 5 6 184 146 130 103 111 120 117 53 33 Hutchinson 1 6 170 151 90 104 111 97 101 51 32 Hutchinson 2 3 169 142 94 101 114 101 95 24 50 32 Hutchinson 5 5 183 138 98 111 50 42 38 108 112 98 31 56 40 Manasota 5 5 188 140 146 111 131 100 110 118 116 102 35 27 55 32 26 14 Manasota 1 6 179 139 123 114 114 96 28 54 30 Manasota 5 1 182 140 139 99 137 97 106 115 110 99 38 28 55 29 24 11 Manasota 5 5 190 143 125 101 111 105 56 32 Manasota 1 6 177 140 137 98 95 107 110 110 99 39 31 24 54 28 Manasota 5 1 183 146 138 101 112 116 104 108 55 32 Manasota 2 6 179 94 105 108 97 29 Manasota 5 1 187 137 139 107 134 101 113 57 44 37 113 118 88 38 30 30 54 31 30 14 Manasota 2 3 185 141 123 108 109 126 85 29 55 32 Manasota 4 1 180 138 138 96 135 95 108 110 107 112 34 28 28 55 34 25 13 Manasota 1 5 185 131 93 48 38 32 115 117 98 30 53 32 22 14 Manasota 5 5 179 140 134 103 130 99 110 51 40 37 108 110 97 35 29 27 52 34 28 13 Manasota 5 6 174 143 136 98 112 114 99 100 35 30 28 56 30 Manasota 1 1 171 122 99 131 90 103 52 37 36 106 95 95 38 23 51 31 25 13 Manasota 5 1 175 139 139 98 128 92 103 115 119 88 37 28 53 31 24 14 Manasota 5 4 185 138 135 103 125 92 103 111 107 101 37 29 26 52 35 Manasota 2 1 174 131 129 99 120 84 99 49 37 31 107 104 96 35 27 27 50 32 26 12 Manasota 5 5 181 141 145 107 135 96 112 115 105 105 41 32 58 35 Manasota 5 5 168 140 131 88 102 42 112 97 26 50 30

PAGE 220

220 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. Manasota 5 1 168 133 135 101 127 96 108 54 40 36 112 109 89 34 28 25 53 32 27 15 Manasota 1 1 174 141 131 99 128 94 108 42 34 112 106 92 36 35 28 55 27 Bird Island 5 1 196 143 147 105 133 103 108 52 40 35 116 118 109 41 30 33 52 32 30 12 Bird Island 5 3 182 147 133 99 116 117 104 102 27 58 30 Bird Island 4 1 166 137 92 100 108 103 22 51 28 Bird Island 2 1 104 111 115 119 31 56 30 Bird Island 5 1 199 145 137 102 116 117 120 106 58 35 Bird Island 5 3 182 145 144 107 128 101 109 117 113 95 31 28 26 54 28 26 13 Republic Groves 5 1 175 137 90 101 107 117 23 52 29 Republic Groves 5 4 178 141 126 91 110 110 27 50 30 Republic Groves 5 4 189 131 88 119 108 55 30 Republic Groves 5 5 137 119 83 118 28 Republic Groves 5 4 187 138 88 106 117 119 95 53 27 Republic Groves 5 3 192 138 146 112 127 88 56 42 34 122 118 99 34 32 26 53 28 29 15 Republic Groves 5 1 184 137 129 108 58 40 35 114 113 29 54 33 Republic Groves 4 3 188 141 140 101 123 95 105 54 38 35 119 115 101 40 34 26 54 28 31 14 Republic Groves 1 4 178 137 121 84 103 57 112 111 25 51 27 Highlands Beach 1 2 179 146 136 127 95 106 109 105 34 31 24 Highlands Beach 5 3 181 138 93 113 109 113 36 31 27 52 34 28 15 Highlands Beach 5 5 177 149 138 98 134 95 105 114 120 91 37 29 28 53 29 27 13 Highlands Beach 1 3 174 51 38 33 115 112 100 50 28

PAGE 221

221 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. Highlands Beach 5 4 182 135 91 104 59 41 37 109 114 103 34 27 52 31 29 Highlands Beach 5 3 184 116 110 115 99 60 32 Highlands Beach 1 1 174 144 125 104 110 112 105 33 Highlands Beach 1 5 176 141 139 99 127 87 104 53 39 36 104 35 32 26 51 27 23 16 Highlands Beach 5 2 176 146 132 99 87 58 38 38 109 98 106 40 30 24 48 30 29 11 Highlands Beach 5 4 184 156 144 100 143 97 110 56 40 40 116 113 109 37 33 28 55 30 Highlands Beach 5 3 190 147 134 90 106 120 106 108 23 54 36 Highlands Beach 5 6 172 131 96 103 110 109 106 97 38 32 27 55 32 Highlands Beach 5 6 174 100 111 112 109 94 55 34 Highlands Beach 5 3 177 149 100 108 110 101 106 27 56 32 Highlands Beach 2 5 174 113 116 25 25 Highlands Beach 1 1 176 138 119 100 105 109 104 103 27 52 27 Highlands Beach 1 1 177 148 132 102 132 92 106 111 106 103 37 32 53 29 Highlands Beach 1 2 177 146 137 96 125 84 98 48 37 34 115 108 105 34 32 22 48 29 26 15 Highlands Beach 4 2 177 153 139 56 41 40 119 115 103 25 54 32 Highlands Beach 5 1 178 146 140 105 137 98 108 55 42 36 115 107 95 34 31 26 54 26 27 10 Highlands Beach 1 1 151 132 127 86 103 54 39 36 109 100 107 33 32 24 51 30 25 13

PAGE 222

222 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. Highlands Beach 5 3 192 134 100 106 121 122 25 53 30 33 15 Highlands Beach 5 3 177 145 134 97 107 118 108 101 29 54 29 Highlands Beach 1 1 166 142 106 110 114 100 29 54 28 Highlands Beach 5 2 179 138 100 120 90 103 115 108 102 35 30 24 52 33 25 15 Highlands Beach 5 6 168 134 123 85 100 108 113 89 23 51 31 Highlands Beach 1 3 170 139 126 90 102 100 101 102 25 52 29 Highlands Beach 1 3 163 137 125 87 96 97 22 Highlands Beach 5 6 179 144 136 102 97 109 113 109 103 34 55 30 Highlands Beach 2 5 173 140 129 93 123 94 106 111 105 98 32 26 52 29 27 14 Margate Blount 5 4 188 141 151 105 131 97 106 58 41 38 119 111 114 40 32 27 53 30 31 14 Margate Blount 1 1 171 134 141 101 120 94 100 48 39 33 112 99 108 35 29 25 48 23 28 13 Margate Blount 5 1 175 141 143 104 123 100 106 112 107 101 37 29 24 55 30 20 15 Margate Blount 5 3 171 135 125 97 112 103 105 29 50 32 Margate Blount 5 5 169 135 136 98 122 90 100 110 102 102 36 31 24 47 29 Margate Blount 2 1 137 141 103 122 85 96 53 39 34 113 107 98 23 47 22 25 13 Fort Walton Temple Mound 1 1 151 99 108 51 42 33 106 95 30 54 20 Fort Walton Temple Mound 5 3 176 143 146 105 128 57 40 35 113 95 93 37 30 53 29 Fort Walton Temple Mound 5 3 166 153 146 102 128 90 101 60 40 38 113 100 98 38 33 24 50 30 27 14 Fort Walton Temple Mound 4 3 185 137 148 112 126 99 104 114 109 105 38 33 52 30 Fort Walton Temple Mound 1 1 159 143 130 96 125 91 100 53 37 35 101 95 97 41 34 23 50 24 31 14

PAGE 223

223 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. Fort Walton Temple Mound 5 4 155 141 101 124 94 102 53 40 34 112 96 93 31 30 24 52 24 27 12 Fort Walton Temple Mound 5 6 177 156 142 117 111 111 101 58 33 Fort Walton Temple Mound 2 5 159 144 140 97 88 99 52 37 34 111 103 91 29 27 21 50 29 26 14 Fort Walton Temple Mound 5 4 171 142 92 104 54 42 35 105 109 22 51 Safety Harbor 5 5 180 151 146 106 138 97 111 119 110 101 37 29 28 54 33 Safety Harbor 5 3 181 148 133 88 104 51 41 34 121 117 27 51 32 Safety Harbor 5 3 172 142 90 103 57 40 37 111 104 27 50 33 Safety Harbor 5 3 185 153 138 95 105 123 115 27 52 32 Safety Harbor 1 2 170 142 137 98 126 93 104 116 102 97 36 27 22 51 28 Safety Harbor 1 3 165 130 95 119 93 102 50 41 32 108 96 96 35 32 26 49 28 Safety Harbor 1 1 165 142 147 100 132 96 102 55 41 36 114 107 101 37 37 23 50 29 Safety Harbor 3 6 139 138 102 128 84 97 51 39 33 109 27 49 25 26 Safety Harbor 2 6 175 141 135 89 128 97 106 115 109 100 35 33 26 53 27 Safety Harbor 5 1 175 140 143 104 126 49 42 32 111 110 103 35 29 28 54 27 22 11 Safety Harbor 5 4 188 144 139 103 95 110 52 42 35 117 108 29 55 34 Safety Harbor 5 3 171 145 129 94 105 115 106 93 26 53 31 Safety Harbor 5 5 184 156 53 45 36 123 109 105 27 55 34 Safety Harbor 5 5 185 140 142 108 96 108 58 42 35 111 109 104 38 30 28 54 33 Safety Harbor 1 2 171 139 137 98 93 50 41 34 105 112 92 32 26 26 51 34 24 Safety Harbor 5 1 149 98 112 54 43 33 111 109 31 57 34 Captiva 1 6 176 145 131 100 129 85 101 111 105 95 34 32 24 50 29 Captiva 2 6 174 134 119 95 106 116 111 24 53 28 Captiva 2 3 171 143 137 98 124 87 105 51 41 35 115 103 103 32 30 23 53 25 27 13 Captiva 5 5 184 146 136 108 138 96 110 60 42 35 114 106 104 39 30 30 55 37 30 14 Captiva 2 4 176 140 125 99 129 95 106 103 109 93 33 28 26 52 26 Captiva 2 3 175 136 129 99 127 87 102 51 39 34 112 96 102 37 29 27 49 24 Captiva 5 5 178 149 135 101 120 56 26 Captiva 5 2 182 148 137 98 108 124 107 106 55 30 Captiva 5 4 177 141 128 105 132 91 105 55 41 35 112 104 84 39 38 25 53 30 28 13 Captiva 1 3 173 140 130 93 105 111 107 105 27 53 26 Captiva 5 2 172 140 134 106 129 94 105 53 40 32 113 106 85 35 27 28 52 29 29 12

PAGE 224

224 Site Sex Age g op eu eu ba b ba n au au ft ft fmt fmt n ns d ec Orb. Ht n b b l l o ba o F. Mag br. Orb. br. at n n fmt Mast. br. Occ. cond. leng. Occ. cond. br. Captiva 2 2 173 149 131 90 101 113 107 100 25 50 25 Captiva 5 6 183 156 140 97 105 117 103 109 52 37 Captiva 5 6 176 144 133 102 135 92 105 113 102 94 40 25 52 34 Captiva 2 6 176 143 126 93 99 112 98 102 22 50 30 Captiva 5 6 180 152 97 107 113 96 106 28 53 31 Captiva 1 2 178 133 121 90 103 110 111 95 27 52 28 24 15 Captiva 5 6 183 132 126 95 106 113 115 101 27 54 35 Captiva 5 4 180 138 132 92 110 118 105 101 28 55 34 Captiva 5 5 183 148 135 101 112 120 107 28 56 34 Captiva 5 3 188 146 133 102 111 124 116 107 56 31 Captiva 1 2 176 136 134 102 121 89 101 54 38 34 112 101 94 34 23 48 29 26 15 Captiva 1 1 176 146 130 99 88 100 52 39 33 105 103 105 34 31 26 51 34 25 12 Captiva 1 5 169 141 126 87 101 110 91 102 51 26 Captiva 1 1 168 93 103 110 100 99 26 52 27 Captiva 5 4 180 149 144 109 137 96 110 121 112 99 34 31 27 55 26 26 14 Captiva 5 3 187 136 134 106 134 88 104 110 123 91 38 27 52 35 Captiva 4 3 188 155 151 108 135 95 41 34 124 118 105 34 29 30 53 30 Captiva 1 1 175 141 134 96 124 87 101 55 40 32 115 97 100 37 30 22 51 30 Captiva 1 5 188 128 97 104 123 118 27 51 31 Captiva 1 1 170 139 126 93 103 109 102 97 26 50 30 Captiva 5 6 181 148 130 85 100 116 98 113 27 50 33 Captiva 5 4 193 148 142 108 140 100 113 60 43 37 121 103 116 40 31 28 56 36 Captiva 2 2 170 135 131 100 128 86 100 50 38 31 110 107 87 33 28 26 50 32 Captiva 1 1 171 147 130 94 126 95 103 49 40 34 110 104 97 33 31 25 51 22 22 12 Captiva 5 3 186 150 140 105 137 95 108 41 37 118 112 96 38 33 28 54 26 Captiva 5 1 179 136 131 102 133 84 103 110 105 96 36 31 25 50 31 Captiva 2 1 172 139 131 101 123 93 103 50 42 36 108 102 94 35 30 26 50 24 28 14 Captiva 5 2 177 144 133 101 130 93 103 53 41 36 112 104 95 39 32 25 51 32 32 12 Captiva 1 5 176 147 125 90 117 106 100 23 52 27 Captiva 5 6 193 150 149 108 130 93 106 129 112 105 37 31 53 31 Captiva 2 6 176 139 138 104 128 89 102 116 102 101 36 28 52 28 27 16 Captiva 2 4 176 143 130 91 106 110 103 104 25 53 32 Captiva 5 6 177 153 141 97 120 110 97 27 Captiva 1 6 181 141 123 92 100 115 110 97 24 51 24 Captiva 1 1 171 139 131 95 124 92 100 57 40 37 110 108 96 38 29 26 50 23 25 14

PAGE 225

225 APPENDIX C TIME AND GEOGRAPHY DISTANCE MATRIX Warm Mineral Springs Gauthier Little Salt Spring Windover Bay West Palmer Henderson Mound Hughes Island Browne Site 5 Bay Pines Dunwody Warm Mineral Springs 5120 3760 1200 3320 7735 1187 7660 8085 7410 8240 Gauthier 1.947 1360 3920 1800 2615 3150 2540 2965 2290 3120 Little Salt Spring 0.031 1.916 2560 440 3975 4510 3900 4325 3650 4480 Windover 2.050 0.195 2.021 2120 6535 7070 6460 6885 6210 7040 Bay West 0.962 2.263 0.959 2.424 4415 4950 4340 4765 4090 4920 Palmer 0.278 2.047 0.295 2.128 1.222 535 75 350 325 505 Henderson Mound 2.585 2.050 2.572 1.902 3.436 2.440 610 185 860 30 Hughes Island 2.516 2.615 2.513 2.507 3.459 2.300 0.843 425 250 580 Browne Site 5 3.408 2.136 3.387 1.942 4.108 3.333 1.175 1.983 675 155 Bay Pines 0.912 2.051 0.916 2.072 1.874 0.668 1.872 1.639 2.879 830 Dunwody 0.159 2.099 0.186 2.205 0.909 0.321 2.719 2.615 3.560 0.989 McKeithan 3.209 2.721 3.199 2.562 4.094 3.036 0.697 0.937 1.303 2.418 3.332 Venice Beach Complex 0.205 2.075 0.230 2.165 1.122 0.108 2.539 2.407 3.418 0.775 0.215 Sarasota Bay Mound 0.388 2.015 0.398 2.085 1.345 0.132 2.317 2.168 3.227 0.537 0.451 Yellow Bluffs 0.409 2.012 0.419 2.080 1.367 0.155 2.297 2.145 3.210 0.515 0.474 Casey Key 0.287 2.054 0.304 2.135 1.229 0.010 2.439 2.296 3.335 0.663 0.327 Crystal River 1.885 1.908 1.876 1.821 2.791 1.710 0.769 0.751 1.864 1.110 2.006 Bayshore Homes 0.898 2.026 0.901 2.047 1.860 0.658 1.867 1.646 2.867 0.026 0.978 Hutchinson Island 2.094 1.306 2.066 1.494 1.794 2.327 3.297 3.736 3.429 2.678 2.189 Manasota 0.158 2.102 0.189 2.203 0.987 0.247 2.663 2.546 3.520 0.915 0.078 Bird Island 2.582 2.706 2.579 2.598 3.528 2.361 0.914 0.091 2.040 1.697 2.678 Republic Groves 0.664 1.287 0.633 1.400 1.179 0.812 2.288 2.437 2.933 1.114 0.813 Highland Beach 2.293 2.091 2.271 2.284 1.642 2.563 3.996 4.339 4.226 3.062 2.336 Margate Blount Mound 2.121 2.099 2.100 2.294 1.432 2.394 3.942 4.245 4.232 2.917 2.155 Fort Walton Temple Mound 5.484 6.156 5.497 6.056 6.410 5.208 4.280 3.552 5.128 4.623 5.505 Safety Harbor 1.037 1.905 1.034 1.909 1.988 0.825 1.657 1.481 2.654 0.225 1.136 Captiva Island 0.462 2.252 0.476 2.380 0.609 0.666 3.047 2.962 3.849 1.334 0.347 Temporal distance (based on average dates) between sites on the top half and geographic distance between sites on the bottom half.

PAGE 226

226 McKeithan Venice Beach Complex Sarasota Bay Mound Yellow Bluffs Casey Key Crystal River Bayshore Homes Hutchinson Island Manasota Bird Island Warm Mineral Springs 7895 7610 8760 7570 8160 7390 6655 7610 7645 5360 Gauthier 2775 2490 3640 2450 3040 2270 1535 2490 2525 240 Little Salt Spring 4135 3850 5000 3810 4400 3630 2895 3850 3885 1600 Windover 6695 6410 7560 6370 6960 6190 5455 6410 6445 4160 Bay West 4575 4290 5440 4250 4840 4070 3335 4290 4325 2040 Palmer 160 125 1025 165 425 345 1080 125 90 2375 Henderson Mound 375 660 490 700 110 880 1615 660 625 2910 Hughes Island 235 50 1100 90 500 270 1005 50 15 2300 Browne Site 5 190 475 675 515 75 695 1430 475 440 2725 Bay Pines 485 200 1350 160 750 20 755 200 235 2050 Dunwody 345 630 520 670 80 850 1585 630 595 2880 McKeithan 285 865 325 265 505 1240 285 250 2535 Venice Beach Complex 3.140 1150 40 550 220 955 0 35 2250 Sarasota Bay Mound 2.908 0.239 1190 600 1370 2105 1150 1115 3400 Yellow Bluffs 2.886 0.262 0.023 590 180 915 40 75 2210 Casey Key 3.034 0.112 0.128 0.151 770 1505 550 515 2800 Crystal River 1.327 1.813 1.582 1.561 1.708 735 220 255 2030 Bayshore Homes 2.418 0.765 0.527 0.504 0.654 1.106 955 990 1295 Hutchinson Island 3.986 2.292 2.372 2.382 2.336 2.988 2.653 35 2250 Manasota 3.270 0.140 0.378 0.401 0.252 1.944 0.905 2.236 2285 Bird Island 0.948 2.468 2.229 2.206 2.357 0.839 1.705 3.825 2.607 Republic Groves 2.964 0.815 0.824 0.830 0.821 1.712 1.089 1.574 0.822 2.517 Highland Beach 4.693 2.497 2.644 2.660 2.573 3.589 3.039 0.820 2.402 4.424 Margate Blount Mound 4.639 2.324 2.481 2.498 2.404 3.499 2.895 0.893 2.223 4.329 Fort Walton Temple Mound 3.838 5.300 5.099 5.079 5.200 4.251 4.644 7.186 5.428 3.462 Safety Harbor 2.220 0.932 0.694 0.671 0.822 0.901 0.214 2.631 1.068 1.545 Captiva Island 3.670 0.559 0.797 0.820 0.671 2.345 1.324 2.136 0.419 3.024

PAGE 227

227 Republic Groves Highland Beach Margate Blount Mound Fort Walton Temple Mound Safety Harbor Captiva Island Warm Mineral Springs 3328 8360 8485 8760 8310 7870 Gauthier 1792 3240 3365 3640 3190 2750 Little Salt Spring 432 4600 4725 5000 4550 4110 Windover 2128 7160 7285 7560 7110 6670 Bay West 8 5040 5165 5440 4990 4550 Palmer 4407 625 750 1025 575 135 Henderson Mound 4942 90 215 490 40 400 Hughes Island 4332 700 825 1100 650 210 Browne Site 5 4757 275 400 675 225 215 Bay Pines 4082 950 1075 1350 900 460 Dunwody 4912 120 245 520 70 370 McKeithan 4567 465 590 865 415 25 Venice Beach Complex 4282 750 875 1150 700 260 Sarasota Bay Mound 5432 400 275 0 450 890 Yellow Bluffs 4242 790 915 1190 740 300 Casey Key 4832 200 325 600 150 290 Crystal River 4062 970 1095 1370 920 480 Bayshore Homes 3327 1705 1830 2105 1655 1215 Hutchinson Island 4282 750 875 1150 700 260 Manasota 4317 715 840 1115 665 225 Bird Island 2032 3000 3125 3400 2950 2510 Republic Groves 5032 5157 5432 4982 4542 Highland Beach 1.969 125 400 50 490 Margate Blount Mound 1.843 0.219 275 175 615 Fort Walton Temple Mound 5.704 7.672 7.537 450 890 Safety Harbor 1.101 3.070 2.938 4.602 440 Captiva Island 0.987 2.162 1.966 5.808 1.482

PAGE 228

228 APPENDIX D RESULTS Table D 1 Independent sample t test for equality of means t df Sig. (2 tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper Sex Equal variances assumed 0.42 42.00 0.68 0.23 0.54 0.87 1.32 Equal variances not assumed 0.42 41.98 0.68 0.23 0.54 0.87 1.32 Age Equal variances assumed 0.43 42.00 0.67 0.18 0.43 1.04 0.68 Equal variances not assumed 0.43 41.70 0.67 0.18 0.43 1.04 0.68 Max l ength Equal variances assumed 0.10 40.00 0.92 0.24 2.27 4.35 4.83 Equal variances not assumed 0.10 40.00 0.92 0.24 2.27 4.35 4.83 Max br eadth Equal variances assumed 0.27 38.00 0.79 0.50 1.88 4.31 3.31 Equal variances not assumed 0.27 37.97 0.79 0.50 1.88 4.31 3.31 Bizy gomatic br eadth Equal variances assumed 0.20 4.00 0.85 0.33 1.67 4.96 4.29 Equal variances not assumed 0.20 3.99 0.85 0.33 1.67 4.96 4.30 Ba B Equal variances assumed 0.13 20.00 0.90 0.27 2.16 4.24 4.79 Equal variances not assumed 0.13 19.97 0.90 0.27 2.16 4.24 4.79 Ba N Equal variances assumed 0.00 20.00 1.00 0.00 1.80 3.76 3.76 Equal variances not assumed 0.00 20.00 1.00 0.00 1.80 3.76 3.76 Ba Pr Equal variances assumed 0.29 16.00 0.77 0.33 1.14 2.08 2.74 Equal variances not assumed 0.29 15.71 0.77 0.33 1.14 2.08 2.75 Max. alv eolar br eadth Equal variances assumed 0.40 16.00 0.70 0.44 1.12 1.94 2.83 Equal variances not assumed 0.40 15.91 0.70 0.44 1.12 1.94 2.83 Max alv eolar l ength Equal variances assumed 0.72 24.00 0.48 0.62 0.85 2.38 1.15 Equal variances not assumed 0.72 23.89 0.48 0.62 0.85 2.38 1.15 Biaur icular Equal variances assumed 0.28 34.00 0.78 0.44 1.57 2.75 3.64 Equal variances not assumed 0.28 33.94 0.78 0.44 1.57 2.75 3.64 N Pr Equal variances assumed 0.15 18.00 0.88 0.30 1.98 4.47 3.87 Equal variances not assumed 0.15 18.00 0.88 0.30 1.98 4.47 3.87 Min Front al breadth Equal variances assumed 0.28 38.00 0.78 0.40 1.44 2.51 3.31 Equal variances not assumed 0.28 37.94 0.78 0.40 1.44 2.51 3.31 Up per fac ial br eadth Equal variances assumed 0.14 37.00 0.89 0.19 1.30 2.81 2.44 Equal variances not assumed 0.14 36.93 0.89 0.19 1.29 2.81 2.44

PAGE 229

229 Table D 1. Continued t df Sig. (2 tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper N Ns Equal variances assumed 0.24 22.00 0.81 0.33 1.40 2.57 3.24 Equal variances not assumed 0.24 22.00 0.81 0.33 1.40 2.57 3.24 Nasal br eadth Equal variances assumed 0.33 22.00 0.74 0.25 0.75 1.31 1.81 Equal variances not assumed 0.33 20.75 0.74 0.25 0.75 1.31 1.81 Or bital breadth Equal variances assumed 0.23 19.00 0.82 0.17 0.76 1.76 1.41 Equal variances not assumed 0.23 18.99 0.82 0.17 0.75 1.75 1.41 Orb ital height Equal variances assumed 0.23 21.00 0.82 0.15 0.65 1.50 1.19 Equal variances not assumed 0.23 20.49 0.82 0.15 0.65 1.50 1.20 Biorb ital breadth Equal variances assumed 0.20 12.00 0.84 0.43 2.14 4.23 5.08 Equal variances not assumed 0.20 11.98 0.84 0.43 2.14 4.23 5.08 Interor b ital breadth Equal variances assumed 1.02 16.00 0.32 0.89 0.87 0.95 2.73 Equal variances not assumed 1.02 15.98 0.32 0.89 0.87 0.95 2.73 Frontal chord Equal variances assumed 0.21 42.00 0.84 0.36 1.76 3.92 3.19 Equal variances not assumed 0.21 42.00 0.84 0.36 1.76 3.92 3.19 Parietal chord Equal variances assumed 0.00 40.00 1.00 0.00 1.67 3.37 3.37 Equal variances not assumed 0.00 39.68 1.00 0.00 1.67 3.37 3.37 Occipital chord Equal variances assumed 0.36 34.00 0.72 0.61 1.70 4.06 2.84 Equal variances not assumed 0.36 33.91 0.72 0.61 1.70 4.06 2.84 Foramen magnum length Equal variances assumed 0.18 20.00 0.86 0.18 1.01 1.92 2.28 Equal variances not assumed 0.18 19.76 0.86 0.18 1.01 1.92 2.28 F oramen magnum breadth Equal variances assumed 0.65 22.00 0.53 0.58 0.90 2.46 1.29 Equal variances not assumed 0.65 21.63 0.53 0.58 0.90 2.46 1.29 Mast oid length Equal variances assumed 0.90 19.00 0.38 1.15 1.28 1.52 3.83 Equal variances not assumed 0.90 18.92 0.38 1.15 1.28 1.52 3.83 O rbital breadth at n Equal variances assumed 0.32 34.00 0.75 0.22 0.70 1.20 1.65 Equal variances not assumed 0.32 33.74 0.75 0.22 0.70 1.20 1.65 n fmt Equal variances assumed 0.84 40.00 0.41 0.62 0.74 0.87 2.11 Equal variances not assumed 0.84 39.82 0.41 0.62 0.74 0.87 2.11 Mast oid l ength 2 Equal variances assumed 4.16 40.00 0.00 4.71 1.13 7.00 2.43 Equal variances not assumed 4.16 39.99 0.00 4.71 1.13 7.00 2.43

PAGE 230

230 Table D 1. Continued t df Sig. (2 tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper Mast oid br eadth Equal variances assumed 0.29 42.00 0.78 0.23 0.79 1.37 1.83 Equal variances not assumed 0.29 41.81 0.78 0.23 0.79 1.37 1.83 Occipital condyle length Equal variances assumed 0.36 24.00 0.72 0.23 0.64 1.09 1.55 Equal variances not assumed 0.36 23.99 0.72 0.23 0.64 1.09 1.55 Occ ipital cond yle br eadth Equal variances assumed 0.74 24.00 0.47 0.38 0.52 1.46 0.69 Equal variances not assumed 0.74 24.00 0.47 0.38 0.52 1.46 0.69 Chin Equal variances assumed 0.13 16.00 0.90 0.11 0.87 1.95 1.73 Equal variances not assumed 0.13 16.00 0.90 0.11 0.87 1.95 1.73 Body height Equal variances assumed 0.14 28.00 0.89 0.13 0.94 1.80 2.07 Equal variances not assumed 0.14 27.98 0.89 0.13 0.94 1.80 2.07 Body thick l Equal variances assumed 0.12 30.00 0.91 0.06 0.54 1.17 1.04 Equal variances not assumed 0.12 29.96 0.91 0.06 0.54 1.17 1.04 Bigonial Equal variances assumed 0.36 24.00 0.72 0.69 1.93 4.67 3.28 Equal variances not assumed 0.36 23.99 0.72 0.69 1.93 4.67 3.28 Bicondylar Equal variances assumed 0.06 8.00 0.95 0.20 3.17 7.11 7.51 Equal variances not assumed 0.06 7.99 0.95 0.20 3.17 7.11 7.51 Min ramus Equal variances assumed 0.25 33.00 0.81 0.31 1.25 2.85 2.23 Equal variances not assumed 0.25 32.52 0.81 0.31 1.25 2.86 2.24 Max ramus Equal variances assumed 0.17 26.00 0.86 0.36 2.06 3.88 4.59 Equal variances not assumed 0.17 25.92 0.86 0.36 2.06 3.88 4.59 Ramus height Equal variances assumed 0.05 26.00 0.96 0.07 1.45 3.05 2.91 Equal variances not assumed 0.05 25.86 0.96 0.07 1.45 3.05 2.91 Mand l Equal variances assumed 0.03 26.00 0.97 0.07 2.10 4.39 4.24 Equal variances not assumed 0.03 25.96 0.97 0.07 2.10 4.39 4.24 Mand angle Equal variances assumed 0.11 26.00 0.91 0.21 1.89 3.67 4.10 Equal variances not assumed 0.11 25.99 0.91 0.21 1.89 3.67 4.10

PAGE 231

231 Table D 2 Principle Component Results Question Sites % variance accounted for by first two components KMO Sphericity Approx. Chi squared df sig. All sites Warm Mineral Springs, Gauthier, Little Salt Spring, Windover, Bay West, Palmer, Henderson Mound, Hughes Island, Browne Site 5, Bay Pines, Dunwody, McKeithan, Venice Beach Complex, Sarasota Bay Mound, Yellow Bluffs, Casey Key, Crystal River, Bayshore Homes, Hutchinson Island, Manasota, Bird Island, Republic Groves, Highland Beach, Margate Blount Mound, Fort Walton Temple Mound, Safety Harbor, Captiva Island 39.12 0.82 0 2934.6 190 0.000 Palmer, Henderson Mound, Hughes Island, Browne Site 5, Bay Pines, Dunwody, McKeithan, Venice Beach Complex, Sarasota Bay Mound, Yellow Bluffs, Casey Key, Crystal River, Bayshore Homes, Hutchinson Island, Manasota, Bird Island, Highland Beach, Margate Blount Mound, Fort Walton Temple Mound, Safety Harbor, Captiva Island 36.902 0.807 1961.42 190 0.000 Palmer, Henderson Mound, Hughes Island, Browne Site 5, Bay Pines, Dunwody, McKeithan, Venice Beach Complex, Sarasota Bay Mound, Yellow Bluffs, Casey Key, Crystal River, Bayshore Homes, Hutchinson Island, Manasota, Highland Beach, Margate Blount Mound, Fort Walton Temple Mound, Safety Harbor, Captiva Island 36.507 0.804 1889.37 190 0.000 Archaic Warm Mineral Spring, Windover, Little Salt Spring, Bay West, Republic Groves, Gauthier, Bird Island 45.059 0.819 1182.61 190 0.000 Windover, Little Salt Spring, Bay West, Republic Groves, Gauthier, Bird Island 42.79 0.808 1072.50 190 0.000 Windover, Little Salt Spring, Bay West, Republic Groves, Gauthier 42.644 0.807 986.737 190 0.000 Windover, Little Salt Spring, Bay West, Gauthier 44.474 0.811 955.321 190 0.000

PAGE 232

232 Table D 2. Continued Question Sites % variance accounted for by first two components KMO Sphericity Approx. Chi squared df sig. Weeden Island/ Manasota Crystal River, McKeithan, Hughes Island, Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Venice Beach Complex, Yellow Bluffs 36.126 0.744 999.1 190 0.000 Crystal River, Hughes Island, Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound 35.971 0.741 956.424 190 0.000 Weeden Island Crystal River, McKeithan, Hughes Island, Yellow Bluffs, Bay Pines 35.507 0.433 389.758 190 0.000 Crystal River, Hughes Island, Bay Pines 36.008 0.482 378.054 190 0.000 Manasota Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Venice Beach Complex, Yellow Bluffs 39.001 0.744 929.845 190 0.000 Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound 39.112 0.745 907.661 190 0.000 Bay Pines, Bayshore Homes, Casey Key, Manasota Key Cemetery, Palmer Mound 39.969 0.74 906.058 190 0.000 Bayshore Homes, Casey Key, Manasota Key Cemetery, Palmer Mound 40.984 0.728 895.983 190 0.000 South Florida Dunwody, Captiva, Hutchinson Island, Highland Beach, Margate Blount 41.855 0.764 973.7 190 0.000 Captiva, Hutchinson Island, Highland Beach, Margate Blount 42.586 0.759 954.813 190 0.000 Weeden Island/ Alachua/ St. Johns/ Fort Walton Crystal River, Hughes Island, Bay Pines, Henderson Mound, Browne Site 5 35.547 0.605 385.269 190 0.000 Crystal River, Hughes Island, Bay Pines, Fort Walton Temple Mound 34.782 0.559 398.62 190 0.000

PAGE 233

233 Table D 2. Continued Question Sites % variance accounted for by first two components KMO Sphericity Approx. Chi squared df sig. Manasota/ Safety Harbor Crystal River, McKeithan, Hughes Island, Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Venice Beach Complex, Yellow Bluffs, Sarasota Bay Mound, Safety Harbor, Fort Walton Temple Mound 36.134 0.767 1168.48 190 0.000 Crystal River, Hughes Island, Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Sarasota Bay Mound, Safety Harbor 36.031 0.762 1077.44 190 0.000 Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Sarasota Bay Mound, Safety Harbor 39.331 0.743 1032.64 190 0.000 Casey Key, Manasota Key Cemetery, Palmer Mound, Sarasota Bay Mound, Safety Harbor 41.35 0.717 823.456 190 0.000

PAGE 234

234 Table D 3 Discriminant Function Results Question Sites % variance accounted for by first two functions Wilk's Lambda test of all functions Original classificati on % Cross validation classification % Wilk's Lambda Chi squared df Sig. All Warm Mineral Springs, Gauthier, Little Salt Spring, Windover, Bay West, Palmer, Henderson Mound, Hughes Island, Browne Site 5, Bay Pines, Dunwody, McKeithan, Venice Beach Complex, Sarasota Bay Mound, Yellow Bluffs, Casey Key, Crystal River, Bayshore Homes, Hutchinson Island, Manasota, Bird Island, Republic Groves, Highland Beach, Margate Blount Mound, Fort Walton Temple Mound, Safety Harbor, Captiva Island 47.7 0.006 1924.8 4 520 0.000 63.4 41.6 Palmer, Henderson Mound, Hughes Island, Browne Site 5, Bay Pines, Dunwody, McKeithan, Venice Beach Complex, Sarasota Bay Mound, Yellow Bluffs, Casey Key, Crystal River, Bayshore Homes, Hutchinson Island, Manasota, Bird Island, Highland Beach, Margate Blount Mound, Fort Walton Temple Mound, Safety Harbor, Captiva Island 42.4 0.012 1187.34 400 0.000 63.9 41.2 Palmer, Henderson Mound, Hughes Island, Browne Site 5, Bay Pines, Dunwody, McKeithan, Venice Beach Complex, Sarasota Bay Mound, Yellow Bluffs, Casey Key, Crystal River, Bayshore Homes, Hutchinson Island, Manasota, Highland Beach, Margate Blount Mound, Fort Walton Temple Mound, Safety Harbor, Captiva Island 43.6 0.013 1140.22 380 0.000 64.6 40.4 Archaic Warm Mineral Spring, Windover, Little Salt Spring, Bay West, Republic Groves, Gauthier, Bird Island 59.6 0.230 394.035 120 0.000 80.2 58.0 Windover, Little Salt Spring, Bay West, Republic Groves, Gauthier, Bird Island 65.6 0.570 293.016 100 0.000 80.2 56.9 Windover, Little Salt Spring, Bay West, Republic Groves, Gauthier 76.9 0.117 206.941 80 0.000 79.1 54.5 Windover, Little Salt Spring, Bay West, Gauthier 86.7 0.239 126.126 60 0.000 48.2 53.5

PAGE 235

235 Table D 3 Continued Question Sites % variance accounted for by first two functions Wilk's Lambda test of all functions Original classification % Cross validation classification % Wilk's Lambda Chi squared df Sig. Weeden Island/ Manasota Crystal River, McKeithan, Hughes Island, Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Venice Beach Complex, Yellow Bluffs 66.4 0.024 503.619 200 0.000 76.3 45.4 Crystal River, Hughes Island, Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound 74.1 0.043 414.858 140 0.000 77.6 54.4 Weeden Island Crystal River, McKeithan, Hughes Island, Yellow Bluffs, Bay Pines 79 0.008 147.412 80 0.000 95.5 59.1 Crystal River, Hughes Island, Bay Pines 100 0.095 64.805 40 0.008 95.0 65.0 Manasota Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Venice Beach Complex, Yellow Bluffs 54.6 0.056 288.385 140 0.000 75.7 48.7 Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound 62.5 0.099 226.418 100 0.000 75.0 50.0 Bay Pines, Bayshore Homes, Casey Key, Manasota Key Cemetery, Palmer Mound 72.8 0.174 163.497 80 0.000 73.8 52.3 Bayshore Homes, Casey Key, Manasota Key Cemetery, Palmer Mound 81.7 0.248 124.027 60 0.000 71.6 50.0 South Florida Dunwody, Captiva, Hutchinson Island, Highland Beach, Margate Blount 73.3 0.042 270.127 80 0.000 93.6 78.7 Captiva, Hutchinson Island, Highland Beach, Margate Blount 83.1 0.062 225.781 60 0.000 92.9 72.7

PAGE 236

236 Table D 3 Continued Question Sites % variance accounted for by first two functions Wilk's Lambda test of all functions Original classification % Cross validation classification % Wilk's Lambda Chi squared df Sig. Weeden Island/ Alachua/ St. Johns/ Fort Walton Crystal River, Hughes Island, Bay Pines, Henderson Mound, Browne Site 5 86.6 0.023 145.903 80 0.000 90.4 46.2 Crystal River, Hughes Island, Bay Pines, Fort Walton Temple Mound 92.4 0.048 109.63 60 0.000 93.9 55.1 Manasota/ Safety Harbor Crystal River, McKeithan, Hughes Island, Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Venice Beach Complex, Yellow Bluffs, Sarasota Bay Mound, Safety Harbor, Fort Walton Temple Mound 56.9 0.013 700.175 260 0.000 76.0 44.7 Crystal River, Hughes Island, Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Sarasota Bay Mound, Safety Harbor 64.9 0.024 555.884 180 0.000 77.0 50.3 Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Sarasota Bay Mound, Safety Harbor 70.3 0.059 311.947 120 0.000 76.8 47.2 Casey Key, Manasota Key Cemetery, Palmer Mound, Sarasota Bay Mound, Safety Harbor 84.4 0.068 208.692 80 0.000 82.4 57.1

PAGE 237

237 Table D 4 Cross validation percentage results for Discriminant Function Analysis Question Warm Mineral Springs Gauthier Little Salt Spring Windover Bay West Palmer Henderson Mound Hughes Island Browne Site 5 Bay Pines All sites Warm Mineral Springs 100.0 .0 .0 .0 .0 .0 .0 .0 .0 .0 Gauthier .0 28.0 12.0 12.0 4.0 .0 .0 .0 .0 4.0 Little Salt Spring .0 .0 33.3 50.0 .0 .0 .0 16.7 .0 .0 Windover .0 12.7 9.5 58.7 7.9 .0 .0 .0 .0 .0 Bay West .0 14.3 28.6 .0 28.6 .0 .0 .0 .0 .0 Palmer .0 11.4 .0 .0 .0 40.9 .0 .0 .0 .0 Henderson Mound .0 .0 .0 .0 .0 .0 16.7 33.3 .0 .0 Hughes Island .0 .0 .0 .0 .0 .0 28.6 .0 .0 28.6 Browne Site 5 .0 .0 .0 .0 .0 .0 .0 .0 16.7 .0 Bay Pines .0 20.0 .0 .0 .0 .0 .0 40.0 .0 .0 Dunwody .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 McKeithan .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 Venice Beach Complex .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 Sarasota Bay Mound .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 Yellow Bluffs .0 .0 .0 .0 .0 .0 .0 50.0 .0 .0 Casey Key .0 .0 .0 .0 .0 50.0 .0 .0 .0 .0 Crystal River .0 .0 .0 .0 .0 .0 10.7 .0 .0 14.3 Bayshore Homes .0 .0 .0 3.4 .0 10.3 .0 6.9 .0 .0 Hutchinson Island .0 .0 .0 .0 .0 .0 .0 .0 16.7 8.3 Manasota .0 .0 .0 .0 9.5 9.5 .0 .0 .0 9.5 Bird Island .0 16.7 .0 .0 .0 .0 .0 16.7 16.7 16.7 Republic Groves .0 .0 .0 11.1 11.1 .0 .0 .0 .0 .0 Highland Beach .0 .0 .0 3.3 3.3 .0 .0 3.3 3.3 13.3 Margate Blount Mound .0 .0 16.7 .0 .0 .0 16.7 .0 .0 .0 Fort Walton Temple Mound .0 .0 .0 .0 .0 .0 22.2 .0 .0 .0 Safety Harbor .0 .0 .0 .0 .0 6.3 .0 .0 .0 .0 Captiva Island .0 8.7 .0 2.2 4.3 13.0 .0 2.2 4.3 2.2

PAGE 238

238 Table D 4. Continued Question Dunwody McKeithan Venice Beach Complex Sarasota Bay Mound Yellow Bluffs Casey Key Crystal River Bayshore Homes Hutchinson Island All sites Warm Mineral Springs .0 .0 .0 .0 .0 .0 .0 .0 .0 Gauthier 4.0 .0 4.0 .0 .0 8.0 4.0 .0 .0 Little Salt Spring .0 .0 .0 .0 .0 .0 .0 .0 .0 Windover .0 .0 .0 .0 .0 .0 .0 .0 .0 Bay West .0 .0 .0 .0 .0 .0 .0 .0 .0 Palmer .0 .0 .0 .0 .0 9.1 2.3 11.4 .0 Henderson Mound .0 .0 .0 .0 .0 .0 .0 .0 .0 Hughes Island .0 .0 .0 14.3 .0 .0 28.6 .0 .0 Browne Site 5 .0 .0 .0 .0 .0 16.7 33.3 .0 .0 Bay Pines .0 .0 .0 .0 .0 .0 .0 .0 .0 Dunwody 60.0 .0 .0 .0 .0 .0 .0 .0 20.0 McKeithan .0 .0 .0 .0 .0 .0 .0 .0 .0 Venice Beach Complex .0 .0 .0 .0 .0 .0 .0 100.0 .0 Sarasota Bay Mound .0 .0 .0 .0 .0 .0 .0 .0 .0 Yellow Bluffs .0 .0 .0 .0 .0 .0 .0 .0 .0 Casey Key .0 .0 .0 12.5 .0 12.5 .0 .0 .0 Crystal River .0 .0 .0 .0 .0 .0 46.4 7.1 3.6 Bayshore Homes 6.9 .0 .0 3.4 .0 3.4 3.4 37.9 6.9 Hutchinson Island .0 .0 .0 .0 .0 8.3 .0 8.3 41.7 Manasota .0 .0 .0 4.8 .0 9.5 .0 19.0 4.8 Bird Island .0 .0 .0 .0 .0 .0 .0 .0 .0 Republic Groves .0 .0 .0 .0 .0 .0 .0 .0 .0 Highland Beach .0 .0 .0 .0 .0 3.3 6.7 .0 3.3 Margate Blount Mound 16.7 .0 .0 .0 .0 .0 16.7 .0 .0 Fort Walton Temple Mound .0 11.1 .0 .0 .0 .0 .0 11.1 .0 Safety Harbor .0 .0 .0 .0 .0 .0 .0 .0 6.3 Captiva Island 2.2 .0 .0 .0 .0 6.5 2.2 .0 .0

PAGE 239

239 Table D 4. Continued Question Manasota Bird Island Republic Groves Highland Beach Margate Blount Mound Fort Walton Temple Mound Safety Harbor Captiva Island Total All sites Warm Mineral Springs .0 .0 .0 .0 .0 .0 .0 .0 100.0 Gauthier 4.0 .0 .0 .0 8.0 .0 .0 8.0 100.0 Little Salt Spring .0 .0 .0 .0 .0 .0 .0 .0 100.0 Windover 3.2 1.6 6.3 .0 .0 .0 .0 .0 100.0 Bay West 14.3 .0 .0 .0 .0 .0 .0 14.3 100.0 Palmer 11.4 6.8 .0 4.5 .0 .0 .0 2.3 100.0 Henderson Mound .0 .0 .0 .0 16.7 33.3 .0 .0 100.0 Hughes Island .0 .0 .0 .0 .0 .0 .0 .0 100.0 Browne Site 5 .0 16.7 .0 .0 .0 .0 16.7 .0 100.0 Bay Pines .0 .0 .0 .0 20.0 20.0 .0 .0 100.0 Dunwody .0 .0 .0 .0 .0 .0 .0 20.0 100.0 McKeithan .0 .0 .0 50.0 .0 50.0 .0 .0 100.0 Venice Beach Complex .0 .0 .0 .0 .0 .0 .0 .0 100.0 Sarasota Bay Mound .0 .0 .0 .0 .0 .0 50.0 50.0 100.0 Yellow Bluffs .0 .0 .0 .0 .0 50.0 .0 .0 100.0 Casey Key .0 12.5 .0 .0 12.5 .0 .0 .0 100.0 Crystal River .0 3.6 .0 .0 7.1 3.6 .0 3.6 100.0 Bayshore Homes 3.4 .0 .0 .0 3.4 .0 6.9 3.4 100.0 Hutchinson Island .0 .0 .0 .0 .0 .0 .0 16.7 100.0 Manasota 14.3 9.5 .0 .0 .0 .0 .0 9.5 100.0 Bird Island 16.7 16.7 .0 .0 .0 .0 .0 .0 100.0 Republic Groves .0 .0 77.8 .0 .0 .0 .0 .0 100.0 Highland Beach .0 .0 .0 53.3 .0 .0 .0 6.7 100.0 Margate Blount Mound .0 16.7 .0 .0 .0 .0 .0 16.7 100.0 Fort Walton Temple Mound .0 .0 .0 .0 11.1 44.4 .0 .0 100.0 Safety Harbor .0 .0 .0 .0 .0 6.3 81.3 .0 100.0 Captiva Island .0 2.2 .0 .0 2.2 .0 4.3 43.5 100.0

PAGE 240

240 Table D 4. Continued Question Palmer Henderson Mound Hughes Island Browne Site 5 Bay Pines Dunwody McKeithan Venice Beach Complex Sarasota Bay Mound Yellow Bluffs Palmer 43.2 .0 .0 .0 2.3 .0 .0 .0 .0 .0 Henderson Mound .0 33.3 16.7 .0 .0 .0 .0 .0 .0 .0 Hughes Island .0 28.6 .0 .0 28.6 .0 .0 .0 14.3 .0 Browne Site 5 .0 .0 .0 16.7 .0 .0 .0 .0 .0 .0 Bay Pines .0 .0 40.0 .0 20.0 .0 .0 .0 .0 .0 Dunwody 20.0 .0 .0 .0 .0 40.0 .0 .0 .0 .0 McKeithan .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 Venice Beach Complex .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 Sarasota Bay Mound .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 Yellow Bluffs .0 .0 50.0 .0 .0 .0 .0 .0 .0 .0 Casey Key 37.5 .0 .0 .0 .0 .0 .0 .0 12.5 .0 Crystal River .0 7.1 3.6 .0 7.1 .0 .0 .0 .0 .0 Bayshore Homes 10.3 .0 3.4 3.4 .0 3.4 .0 .0 3.4 .0 Hutchinson Island .0 .0 .0 16.7 8.3 .0 .0 .0 .0 .0 Manasota 9.5 .0 .0 .0 9.5 .0 .0 .0 9.5 .0 Bird Island .0 .0 .0 16.7 .0 .0 .0 .0 .0 .0 Highland Beach .0 3.3 3.3 3.3 10.0 .0 .0 .0 .0 .0 Margate Blount Mound .0 .0 .0 .0 .0 16.7 .0 .0 .0 .0 Fort Walton Temple Mound .0 33.3 .0 11.1 .0 .0 .0 .0 .0 .0 Safety Harbor .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 Captiva Island 15.2 .0 2.2 4.3 4.3 4.3 .0 .0 .0 .0

PAGE 241

241 Table D 4. Continued Question Casey Key Crystal River Bayshore Homes Hutchinson Island Manasota Bird Island Highland Beach Margate Blount Mound Fort Walton Temple Mound Safety Harbor Captiva Island Total Palmer 11.4 2.3 11.4 .0 15.9 6.8 4.5 .0 .0 .0 2.3 100.0 Henderson Mound .0 .0 .0 .0 .0 .0 .0 16.7 33.3 .0 .0 100.0 Hughes Island .0 28.6 .0 .0 .0 .0 .0 .0 .0 .0 .0 100.0 Browne Site 5 16.7 33.3 .0 .0 .0 16.7 .0 .0 .0 16.7 .0 100.0 Bay Pines .0 .0 .0 .0 .0 .0 .0 20.0 .0 .0 20.0 100.0 Dunwody .0 .0 .0 20.0 .0 .0 .0 .0 .0 .0 20.0 100.0 McKeithan .0 .0 .0 .0 .0 .0 50.0 .0 50.0 .0 .0 100.0 Venice Beach Complex .0 .0 100.0 .0 .0 .0 .0 .0 .0 .0 .0 100.0 Sarasota Bay Mound .0 .0 .0 .0 .0 .0 .0 .0 .0 50.0 50.0 100.0 Yellow Bluffs .0 .0 .0 .0 .0 .0 .0 .0 50.0 .0 .0 100.0 Casey Key 25.0 .0 12.5 .0 .0 12.5 .0 .0 .0 .0 .0 100.0 Crystal River .0 57.1 7.1 3.6 .0 .0 3.6 3.6 3.6 .0 3.6 100.0 Bayshore Homes 3.4 .0 37.9 6.9 10.3 .0 .0 3.4 .0 6.9 6.9 100.0 Hutchinson Island 8.3 .0 8.3 41.7 .0 .0 8.3 .0 .0 .0 8.3 100.0 Manasota 9.5 .0 19.0 4.8 14.3 9.5 .0 .0 .0 .0 14.3 100.0 Bird Island .0 .0 .0 .0 33.3 50.0 .0 .0 .0 .0 .0 100.0 Highland Beach 3.3 6.7 3.3 3.3 .0 .0 53.3 3.3 .0 3.3 3.3 100.0 Margate Blount Mound .0 16.7 .0 .0 16.7 16.7 .0 16.7 .0 .0 16.7 100.0 Fort Walton Temple Mound .0 .0 11.1 .0 .0 .0 .0 11.1 33.3 .0 .0 100.0 Safety Harbor .0 6.3 .0 6.3 .0 .0 .0 .0 6.3 75.0 6.3 100.0 Captiva Island 4.3 2.2 .0 .0 2.2 2.2 2.2 2.2 .0 4.3 50.0 100.0

PAGE 242

242 Table D 4. Continued Question Palmer Henderson Mound Hughes Island Browne Site 5 Bay Pines Dunwody McKeithan Venice Beach Complex Sarasota Bay Mound Yellow Bluffs Palmer 43.2 2.3 .0 .0 2.3 .0 .0 .0 2.3 .0 Henderson Mound .0 16.7 33.3 .0 .0 .0 .0 .0 .0 .0 Hughes Island .0 28.6 .0 .0 28.6 .0 .0 .0 14.3 .0 Browne Site 5 .0 .0 .0 16.7 .0 16.7 .0 .0 .0 .0 Bay Pines .0 .0 40.0 .0 20.0 .0 .0 .0 .0 .0 Dunwody 20.0 .0 .0 .0 .0 60.0 .0 .0 .0 .0 McKeithan .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 Venice Beach Complex .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 Sarasota Bay Mound .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 Yellow Bluffs .0 .0 50.0 .0 .0 .0 .0 .0 .0 .0 Casey Key 37.5 .0 .0 .0 .0 .0 .0 .0 12.5 .0 Crystal River .0 7.1 3.6 .0 7.1 .0 .0 .0 .0 .0 Bayshore Homes 10.3 .0 .0 6.9 6.9 3.4 .0 3.4 3.4 .0 Hutchinson Island .0 .0 8.3 16.7 8.3 .0 .0 .0 .0 .0 Manasota 14.3 .0 .0 .0 14.3 .0 .0 .0 9.5 .0 Highland Beach .0 3.3 6.7 3.3 13.3 .0 .0 .0 .0 .0 Margate Blount Mound .0 16.7 .0 .0 .0 16.7 .0 .0 .0 .0 Fort Walton Temple Mound .0 33.3 .0 .0 .0 .0 11.1 .0 .0 .0 Safety Harbor .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 Captiva Island 13.0 .0 2.2 4.3 8.7 4.3 .0 .0 .0 .0

PAGE 243

243 Table D 4. Continued Question Casey Key Crystal River Bayshore Homes Hutchinson Island Manasota Highland Beach Margate Blount Mound Fort Walton Temple Mound Safety Harbor Captiva Island Total Palmer 13.6 2.3 13.6 .0 15.9 2.3 .0 .0 .0 2.3 100.0 Henderson Mound .0 .0 .0 .0 .0 16.7 .0 33.3 .0 .0 100.0 Hughes Island .0 14.3 .0 .0 .0 14.3 .0 .0 .0 .0 100.0 Browne Site 5 16.7 33.3 .0 .0 .0 .0 .0 .0 16.7 .0 100.0 Bay Pines .0 .0 .0 .0 .0 .0 20.0 .0 .0 20.0 100.0 Dunwody .0 .0 .0 20.0 .0 .0 .0 .0 .0 .0 100.0 McKeithan .0 .0 .0 .0 .0 50.0 .0 50.0 .0 .0 100.0 Venice Beach Complex .0 .0 100.0 .0 .0 .0 .0 .0 .0 .0 100.0 Sarasota Bay Mound .0 .0 .0 .0 .0 .0 .0 .0 50.0 50.0 100.0 Yellow Bluffs .0 .0 .0 .0 .0 .0 .0 50.0 .0 .0 100.0 Casey Key 25.0 .0 25.0 .0 .0 .0 .0 .0 .0 .0 100.0 Crystal River .0 57.1 7.1 3.6 .0 3.6 3.6 3.6 .0 3.6 100.0 Bayshore Homes 3.4 .0 34.5 6.9 6.9 .0 3.4 .0 6.9 3.4 100.0 Hutchinson Island 8.3 .0 8.3 41.7 .0 .0 .0 8.3 .0 .0 100.0 Manasota 9.5 .0 19.0 4.8 14.3 .0 .0 .0 .0 14.3 100.0 Highland Beach 3.3 3.3 3.3 3.3 .0 53.3 3.3 .0 .0 3.3 100.0 Margate Blount Mound 16.7 16.7 .0 .0 .0 .0 16.7 .0 .0 16.7 100.0 Fort Walton Temple Mound .0 .0 11.1 .0 .0 .0 11.1 33.3 .0 .0 100.0 Safety Harbor .0 6.3 .0 6.3 .0 .0 .0 6.3 75.0 6.3 100.0 Captiva Island 4.3 2.2 2.2 .0 2.2 2.2 2.2 .0 4.3 47.8 100.0

PAGE 244

244 Table D 4. Continued Question Warm Mineral Springs Gauthier Little Salt Spring Windover Bay West Bird Island Republic Groves Total Archaic Warm Mineral Springs 100.0 .0 .0 .0 .0 .0 .0 100.0 Gauthier .0 64.0 12.0 8.0 12.0 4.0 .0 100.0 Little Salt Spring .0 .0 16.7 50.0 16.7 .0 16.7 100.0 Windover .0 15.9 14.3 57.1 9.5 .0 3.2 100.0 Bay West .0 14.3 28.6 28.6 14.3 14.3 .0 100.0 Bird Island .0 16.7 .0 .0 .0 83.3 .0 100.0 Republic Groves .0 .0 .0 11.1 11.1 .0 77.8 100.0 Gauthier Little Salt Spring Windover Bay West Bird Island Republic Groves Total Gauthier 64.0 12.0 8.0 12.0 4.0 .0 100.0 Little Salt Spring .0 16.7 50.0 16.7 .0 16.7 100.0 Windover 15.9 14.3 57.1 9.5 .0 3.2 100.0 Bay West 14.3 28.6 28.6 14.3 14.3 .0 100.0 Bird Island 16.7 .0 .0 .0 83.3 .0 100.0 Republic Groves .0 .0 11.1 11.1 .0 77.8 100.0 Gauthier Little Salt Spring Windover Bay West Republic Groves Total Gauthier 60.0 12.0 12.0 16.0 .0 100.0 Little Salt Spring .0 16.7 50.0 33.3 .0 100.0 Windover 17.5 14.3 57.1 9.5 1.6 100.0 Bay West 28.6 28.6 28.6 14.3 .0 100.0 Republic Groves .0 .0 11.1 11.1 77.8 100.0

PAGE 245

2 45 Table D 4. Continued Question Gauthier Little Salt Spring Windover Bay West Total Gauthier 60.0 12.0 12.0 16.0 100.0 Little Salt Spring .0 16.7 66.7 16.7 100.0 Windover 14.3 15.9 58.7 11.1 100.0 Bay West 42.9 28.6 14.3 14.3 100.0 Weeden Island/ Manasota Palmer Mound Hughes Island Bay Pines Dunwody McKeithan Venice Beach Complex Yellow Bluffs Casey Key Crystal River Bayshore Homes Manasota Total Palmer Mound 52.3 4.5 .0 .0 .0 .0 .0 22.7 2.3 6.8 11.4 100.0 Hughes Island .0 .0 42.9 .0 .0 .0 14.3 .0 28.6 .0 14.3 100.0 Bay Pines .0 40.0 20.0 .0 .0 .0 .0 .0 .0 .0 40.0 100.0 Dunwody 20.0 .0 .0 60.0 .0 .0 .0 .0 20.0 .0 .0 100.0 McKeithan .0 .0 50.0 .0 .0 .0 .0 .0 50.0 .0 .0 100.0 Venice Beach Complex .0 .0 .0 .0 .0 .0 .0 .0 .0 100.0 .0 100.0 Yellow Bluffs .0 50.0 .0 .0 .0 .0 .0 .0 50.0 .0 .0 100.0 Casey Key 37.5 .0 .0 .0 .0 .0 .0 37.5 .0 25.0 .0 100.0 Crystal River .0 3.6 10.7 .0 .0 .0 3.6 .0 71.4 10.7 .0 100.0 Bayshore Homes 13.8 .0 3.4 3.4 3.4 6.9 .0 10.3 3.4 44.8 10.3 100.0 Manasota 9.5 4.8 9.5 4.8 .0 4.8 4.8 9.5 .0 23.8 28.6 100.0

PAGE 246

246 Table D 4. Continued Question Palmer Mound Hughes Island Bay Pines Dunwody Casey Key Crystal River Bayshore Homes Manasota Total Palmer Mound 65.9 4.5 .0 .0 15.9 2.3 4.5 6.8 100.0 Hughes Island .0 .0 42.9 .0 .0 42.9 .0 14.3 100.0 Bay Pines .0 40.0 20.0 .0 .0 .0 .0 40.0 100.0 Dunwody 20.0 .0 .0 60.0 .0 20.0 .0 .0 100.0 Casey Key 50.0 .0 .0 .0 37.5 .0 12.5 .0 100.0 Crystal River .0 3.6 10.7 .0 .0 78.6 7.1 .0 100.0 Bayshore Homes 17.2 .0 3.4 6.9 6.9 3.4 51.7 10.3 100.0 Manasota 14.3 4.8 9.5 4.8 9.5 .0 23.8 33.3 100.0 Weeden Island Hughes Island Bay Pines McKeithan Yellow Bluffs Crystal River Total Hughes Island 14.3 42.9 14.3 14.3 14.3 100.0 Bay Pines 20.0 20.0 .0 20.0 40.0 100.0 McKeithan .0 50.0 50.0 .0 .0 100.0 Yellow Bluffs 50.0 .0 .0 50.0 .0 100.0 Crystal River 3.6 14.3 3.6 .0 78.6 100.0 Hughes Island Bay Pines Crystal River Total Hughes Island 14.3 57.1 28.6 100.0 Bay Pines 40.0 40.0 20.0 100.0 Crystal River 7.1 10.7 82.1 100.0

PAGE 247

247 Table D 4. Continued Question Palmer Mound Bay Pines Dunwody Venice Beach Complex Yellow Bluffs Casey Key Bayshore Homes Manasota Total Manasota Palmer Mound 59.1 2.3 .0 .0 .0 15.9 11.4 11.4 100.0 Bay Pines .0 60.0 .0 .0 .0 .0 .0 40.0 100.0 Dunwody 20.0 .0 60.0 .0 .0 .0 20.0 .0 100.0 Venice Beach Complex .0 .0 .0 .0 .0 .0 100.0 .0 100.0 Yellow Bluffs .0 .0 .0 .0 .0 .0 50.0 50.0 100.0 Casey Key 37.5 .0 .0 .0 .0 37.5 25.0 .0 100.0 Bayshore Homes 10.3 3.4 6.9 10.3 .0 6.9 51.7 10.3 100.0 Manasota 9.5 14.3 4.8 .0 4.8 14.3 23.8 28.6 100.0 Palmer Mound Bay Pines Dunwody Casey Key Bayshore Homes Manasota Total Palmer Mound 56.8 .0 .0 18.2 11.4 13.6 100.0 Bay Pines .0 60.0 .0 .0 .0 40.0 100.0 Dunwody 20.0 .0 60.0 .0 20.0 .0 100.0 Casey Key 37.5 .0 .0 37.5 25.0 .0 100.0 Bayshore Homes 13.8 3.4 6.9 6.9 55.2 13.8 100.0 Manasota 14.3 14.3 4.8 14.3 23.8 28.6 100.0 Palmer Mound Bay Pines Casey Key Bayshore Homes Manasota Total Palmer Mound 61.4 .0 15.9 11.4 11.4 100.0 Bay Pines .0 60.0 .0 20.0 20.0 100.0 Casey Key 37.5 .0 37.5 25.0 .0 100.0 Bayshore Homes 17.2 3.4 10.3 55.2 13.8 100.0 Manasota 23.8 4.8 9.5 28.6 33.3 100.0 Palmer Mound Casey Key Bayshore Homes Manasota Total Palmer Mound 56.8 18.2 9.1 15.9 100.0 Casey Key 37.5 37.5 25.0 .0 100.0 Bayshore Homes 17.2 10.3 55.2 17.2 100.0 Manasota 28.6 9.5 28.6 33.3 100.0

PAGE 248

248 Table D 4. Continued Question Dunwody Hutchinson Island Highland Beach Margate Blount Mound Captiva Total South Dunwody 60.0 20.0 .0 .0 20.0 100.0 Hutchinson Island 8.3 58.3 .0 16.7 16.7 100.0 Highland Beach 3.3 3.3 73.3 6.7 13.3 100.0 Margate Blount Mound 16.7 .0 16.7 66.7 .0 100.0 Captiva 8.7 2.2 6.5 4.3 78.3 100.0 Hutchinson Island Highland Beach Margate Blount Mound Captiva Total Hutchinson Island 66.7 .0 16.7 16.7 100.0 Highland Beach 3.3 73.3 6.7 16.7 100.0 Margate Blount Mound .0 16.7 66.7 16.7 100.0 Captiva .0 8.7 4.3 87.0 100.0 Henderson Mound Hughes Island Browne Site 5 Bay Pines Crystal River Total Weeden Island/ Alachua/ St. Johns/ Fort Walton Henderson Mound 16.7 33.3 .0 16.7 33.3 100.0 Hughes Island 42.9 .0 14.3 42.9 .0 100.0 Browne Site 5 16.7 .0 66.7 .0 16.7 100.0 Bay Pines .0 40.0 .0 40.0 20.0 100.0 Crystal River 10.7 3.6 3.6 21.4 60.7 100.0 Hughes Island Bay Pines Crystal River Fort Walton Temple Mound Total Hughes Island .0 28.6 42.9 28.6 100.0 Bay Pines 40.0 .0 40.0 20.0 100.0 Crystal River 10.7 10.7 75.0 3.6 100.0 Fort Walton Temple Mound 11.1 .0 22.2 66.7 100.0

PAGE 249

249 Table D 4. Continued Question Palmer Hughes Island Bay Pines Dunwody McKeithan Venice Beach Complex Sarasota Bay Mound Manasota/ Safety Harbor Palmer 47.7 2.3 .0 .0 .0 .0 2.3 Hughes Island .0 .0 28.6 .0 .0 .0 14.3 Bay Pines .0 40.0 40.0 .0 .0 .0 .0 Dunwody .0 .0 .0 60.0 .0 .0 .0 McKeithan .0 .0 .0 .0 .0 .0 .0 Venice Beach Complex .0 .0 .0 .0 .0 .0 .0 Sarasota Bay Mound .0 .0 .0 .0 .0 .0 .0 Yellow Bluffs .0 50.0 .0 .0 .0 .0 .0 Casey Key 50.0 .0 .0 .0 .0 .0 12.5 Crystal River .0 7.1 14.3 .0 .0 .0 .0 Bayshore Homes 10.3 .0 .0 10.3 .0 3.4 3.4 Manasota 19.0 .0 9.5 .0 .0 .0 9.5 Fort Walton Temple Mound .0 22.2 .0 .0 11.1 .0 .0 Safety Harbor .0 .0 6.3 .0 .0 .0 .0 Table D 4. Continued Question Yellow Bluffs Casey Key Crystal River Bayshore Homes Manasota Fort Walton Temple Mound Safety Harbor Total Palmer .0 20.5 .0 4.5 18.2 2.3 2.3 100.0 Hughes Island 14.3 .0 14.3 .0 14.3 14.3 .0 100.0 Bay Pines .0 .0 .0 .0 20.0 .0 .0 100.0 Dunwody .0 .0 .0 .0 20.0 .0 20.0 100.0 McKeithan .0 .0 50.0 .0 .0 50.0 .0 100.0 Venice Beach Complex .0 .0 .0 100.0 .0 .0 .0 100.0 Sarasota Bay Mound .0 .0 .0 .0 50.0 .0 50.0 100.0 Yellow Bluffs .0 .0 .0 .0 .0 50.0 .0 100.0 Casey Key .0 25.0 .0 12.5 .0 .0 .0 100.0 Crystal River 3.6 .0 60.7 10.7 .0 3.6 .0 100.0 Bayshore Homes .0 10.3 .0 48.3 6.9 .0 6.9 100.0 Manasota .0 14.3 .0 23.8 23.8 .0 .0 100.0 Fort Walton Temple Mound .0 .0 .0 11.1 11.1 44.4 .0 100.0 Safety Harbor .0 6.3 6.3 .0 .0 6.3 75.0 100.0

PAGE 250

250 Table D 4. Continued Question Palmer Hughes Island Bay Pines Dunwody Sarasota Bay Mound Casey Key Crystal River Bayshore Homes Manasota Safety Harbor Total Palmer 56.8 4.5 .0 .0 2.3 20.5 .0 4.5 9.1 2.3 100.0 Hughes Island .0 .0 42.9 .0 14.3 .0 28.6 14.3 .0 .0 100.0 Bay Pines .0 40.0 40.0 .0 .0 .0 .0 .0 20.0 .0 100.0 Dunwody 20.0 .0 .0 60.0 .0 .0 20.0 .0 .0 .0 100.0 Sarasota Bay Mound .0 .0 .0 .0 .0 .0 .0 .0 50.0 50.0 100.0 Casey Key 50.0 .0 .0 .0 12.5 25.0 .0 12.5 .0 .0 100.0 Crystal River .0 7.1 10.7 .0 .0 .0 71.4 10.7 .0 .0 100.0 Bayshore Homes 10.3 3.4 3.4 10.3 3.4 6.9 .0 48.3 6.9 6.9 100.0 Manasota 23.8 .0 9.5 .0 9.5 9.5 .0 23.8 23.8 .0 100.0 Safety Harbor .0 .0 6.3 .0 .0 .0 12.5 6.3 .0 75.0 100.0 Palmer Dunwody Sarasota Bay Mound Casey Key Bayshore Homes Manasota Safety Harbor Total Palmer 54.5 .0 2.3 18.2 11.4 11.4 2.3 100.0 Dunwody 20.0 60.0 .0 .0 .0 .0 20.0 100.0 Sarasota Bay Mound .0 .0 .0 .0 .0 50.0 50.0 100.0 Casey Key 50.0 .0 12.5 25.0 12.5 .0 .0 100.0 Bayshore Homes 17.2 10.3 3.4 3.4 44.8 10.3 10.3 100.0 Manasota 23.8 4.8 9.5 14.3 28.6 19.0 .0 100.0 Safety Harbor .0 .0 .0 .0 18.8 .0 81.3 100.0 Palmer Sarasota Bay Mound Casey Key Manasota Safety Harbor Total Palmer 59.1 2.3 20.5 15.9 2.3 100.0 Sarasota Bay Mound .0 .0 50.0 .0 50.0 100.0 Casey Key 50.0 12.5 25.0 .0 12.5 100.0 Manasota 28.6 4.8 19.0 42.9 4.8 100.0 Safety Harbor .0 .0 .0 6.3 93.8 100.0 Table D 5 R Matrix Results Question Sites Unbiased F ST % variance Within group Phenotypic Variance

PAGE 251

251 accounted for by first two eigenvectors r(ii) Observed Residual All Warm Mineral Springs, Gauthier, Little Salt Spring, Windover, Bay West, Palmer, Henderson Mound, Hughes Island, Browne Site 5, Bay Pines, Dunwody, McKeithan, Sarasota Bay Mound, Yellow Bluffs, Casey Key, Crystal River, Bayshore Homes, Hutchinson Island, Manasota, Bird Island, Republic Groves, Highland Beach, Margate Blount Mound, Fort Walton Temple Mound, Safety Harbor, Captiva Island 0.162 57.1 Warm Mineral Springs 1.342 0.302 0.646 Gauthier 0.066 0.896 0.043 Little Salt Spring 0.128 0.680 0.197 Windover 0.152 0.760 0.093 Bay West 0.085 0.680 0.241 Palmer 0.072 0.712 0.221 Henderson Mound 0.093 0.745 0.167 Hughes Island 0.013 0.974 0.018 Browne Site 5 0.088 1.026 0.110 Bay Pines 0.018 0.633 0.354 Dunwody 0.174 0.737 0.093 McKeithan 0.269 1.291 0.556 Sarasota Bay Mound 0.103 0.410 0.492 Yellow Bluffs 0.273 1.452 0.721 Casey Key 0.044 0.721 0.240 Crystal River 0.114 0.789 0.102 Bayshore Homes 0.089 0.860 0.056 Hutchinson Island 0.149 1.031 0.176 Manasota 0.023 0.847 0.136 Bird Island 0.081 1.048 0.124 Republic Groves 0.309 0.539 0.156 Highland Beach 0.073 0.849 0.083 Margate Blount Mound 0.087 1.136 0.219 Fort Walton Temple Mound 0.135 1.168 0.298 Safety Harbor 0.181 0.737 0.086 Captiva Island 0.040 0.894 0.072

PAGE 252

252 Table D 5 Continued Question Sites Unbiased F ST % variance accounted for by first two eigenvectors Within group Phenotypic Variance r(ii) Observed Residual Palmer, Henderson Mound, Hughes Island, Browne Site 5, Bay Pines, Dunwody, McKeithan, Sarasota Bay Mound, Yellow Bluffs, Casey Key, Crystal River, Bayshore Homes, Hutchinson Island, Manasota, Bird Island, Highland Beach, Margate Blount Mound, Fort Walton Temple Mound, Safety Harbor, Captiva Island 0.102 50.7 Palmer 0.077 0.743 0.216 Henderson Mound 0.091 0.786 0.159 Hughes Island 0.022 1.023 0.007 Browne Site 5 0.111 1.053 0.129 Bay Pines 0.040 0.653 0.345 Dunwody 0.211 0.751 0.068 McKeithan 0.224 1.333 0.526 Sarasota Bay Mound 0.109 0.405 0.520 Yellow Bluffs 0.211 1.509 0.689 Casey Key 0.066 0.745 0.226 Crystal River 0.098 0.822 0.114 Bayshore Homes 0.069 0.894 0.073 Hutchinson Island 0.128 1.042 0.135 Manasota 0.039 0.904 0.094 Bird Island 0.076 1.024 0.064 Highland Beach 0.095 0.898 0.042 Margate Blount Mound 0.084 1.141 0.189 Fort Walton Temple Mound 0.095 1.208 0.268 Safety Harbor 0.150 0.784 0.099 Captiva Island 0.054 0.930 0.052 Palmer, Henderson Mound, Hughes Island, Browne Site 5, Bay Pines, Dunwody, McKeithan, Sarasota Bay Mound, Yellow Bluffs, Casey Key, Crystal River, Bayshore Homes, Hutchinson Island, Manasota, Highland Beach, Margate Blount Mound, Fort Walton Temple Mound, Safety Harbor, Captiva Island 0.104 52.6 Palmer 0.080 0.749 0.214 Henderson Mound 0.082 0.795 0.166 Hughes Island 0.031 1.038 0.023 Browne Site 5 0.113 1.071 0.142 Bay Pines 0.042 0.661 0.342 Dunwody 0.197 0.759 0.081 McKeithan 0.268 1.341 0.575 Sarasota Bay Mound 0.121 0.408 0.512

PAGE 253

253 Table D 5 Continued Question Sites Unbiased F ST % variance accounted for by first two eigenvectors Within group Phenotypic Variance r(ii) Observed Residual Yellow Bluffs 0.187 1.522 0.671 Casey Key 0.070 0.755 0.218 Crystal River 0.094 0.832 0.116 Bayshore Homes 0.071 0.901 0.071 Hutchinson Island 0.122 1.055 0.136 Manasota 0.043 0.910 0.092 Highland Beach 0.089 0.904 0.049 Margate Blount Mound 0.085 1.150 0.192 Fort Walton Temple Mound 0.096 1.233 0.286 Safety Harbor 0.141 0.790 0.109 Captiva Island 0.050 0.938 0.057 Archaic Warm Mineral Spring, Windover, Little Salt Spring, Bay West, Republic Groves, Gauthier, Bird Island 0.299 87.2 WMS 1.103 0.364 0.476 Gauthier 0.079 0.937 0.070 LSS 0.047 0.743 0.299 Windover 0.053 0.831 0.205 Bay West 0.648 0.738 0.285 Bird Island 0.366 1.169 0.476 Republic Groves 0.378 0.587 0.093 Windover, Little Salt Spring, Bay West, Republic Groves, Gauthier, Bird Island 0.150 81.1 Gauthier 0.891 0.991 0.046 LSS 0.032 0.786 0.218 Windover 0.061 0.883 0.092 Bay West 0.063 0.786 0.019 Bird Island 0.327 1.218 0.519 Republic Groves 0.330 0.627 0.068

PAGE 254

254 Table D 5 Continued Question Sites Unbiased F ST % variance accounted for by first two eigenvectors Within group Phenotypic Variance r(ii) Observed Residual Windover, Little Salt Spring, Bay West, Republic Groves, Gauthier 0.099 79.9 Gauthier 0.099 1.041 0.174 LSS 0.050 0.844 0.069 Windover 0.060 0.935 0.032 Bay West 0.036 0.839 0.088 Republic Groves 0.249 0.673 0.049 Windover, Little Salt Spring, Bay West, Gauthier 0.046 82.1 Gauthier 0.063 1.047 0.141 LSS 0.018 0.854 0.094 Windover 0.070 0.947 0.048 Bay West 0.033 0.839 0.095 Weeden Island/ Manasota Crystal River, McKeithan, Hughes Island, Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Yellow Bluffs 0.137 65.0 Palmer Mound 0.100 0.815 0.258 Hughes Island 0.065 1.111 0.002 Bay Pines 0.067 0.699 0.412 Dunwody 0.323 0.845 0.039 McKeithan 0.216 1.430 0.495 Yellow Bluffs 0.224 1.690 0.766 Casey Key 0.157 0.803 0.202 Crystal River 0.110 0.904 0.157 Bayshore Homes 0.059 0.983 0.138 Manasota Key Cemetery 0.056 0.993 0.131

PAGE 255

255 Table D 5 Continued Question Sites Unbiased F ST % variance accounted for by first two eigenvectors Within group Phenotypic Variance r(ii) Observed Residual Crystal River, Hughes Island, Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound 0.098 85.4 Palmer Mound 0.092 0.921 0.088 Hughes Island 0.069 0.944 0.194 Bay Pines 0.050 0.963 0.238 Dunwody 0.253 0.757 0.094 Casey Key 0.116 0.897 0.068 Crystal River 0.138 0.874 0.052 Bayshore Homes 0.045 0.968 0.036 Manasota Key Cemetery 0.022 0.992 0.019 Weeden Island Crystal River, McKeithan, Hughes Island, Yellow Bluffs, Bay Pines 0.365 88.1 Hughes Island 0.261 1.139 0.233 Bay Pines 0.149 0.758 0.822 McKeithan 0.424 1.301 0.232 Yellow Bluffs 0.644 1.799 1.137 Crystal River 0.347 0.897 0.314 Crystal River, Hughes Island, Bay Pines 0.115 100.0 Hughes Island 0.130 1.199 0.227 Bay Pines 0.000 0.823 0.295 Crystal River 0.215 0.945 0.068 Crystal River, Hughes Island, McKeithan 0.384 100.0 Hughes Island 0. 207 1.161 0.280 McKeithan 0.528 1.289 0.432 Crystal River 0.416 0.908 0.152 Manasota Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Yellow Bluffs 0.150 80.6 Palmer 0.052 0.864 0.287 Bay Pines 0.137 0.748 0.300 Dunwody 0.332 0.869 0.057

PAGE 256

256 Table D 5 Continued Question Sites Unbiased F ST % variance accounted for by first two eigenvectors Within group Phenotypic Variance r(ii) Observed Residual Yellow Bluffs 0.355 1.791 1.008 Casey Key 0.105 0.872 0.215 Bayshore Homes 0.051 1.038 0.114 Manasota Key Cemetery 0.196 1.041 0.149 Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound 0.159 90.4 Palmer 0.059 0.858 0.341 Bay Pines 0.332 0.861 0.009 Dunwody 0.371 1.776 0.974 Casey Key 0.095 0.870 0.283 Bayshore Homes 0.042 1.034 0.187 Manasota Key Cemetery 0.052 1.036 0.172 Bay Pines, Bayshore Homes, Casey Key, Manasota Key Cemetery, Palmer Mound 0.055 90.0 Palmer 0.041 0.891 0.057 Bay Pines 0.115 0.768 0.107 Casey Key 0.051 0.882 0.056 Bayshore Homes 0.051 1.063 0.124 Manasota Key Cemetery 0.016 1.068 0.095 Bayshore Homes, Casey Key, Manasota Key Cemetery, Palmer Mound 0.026 90.2 Palmer 0.018 0.886 0.094 Casey Key 0.007 0.880 0.111 Bayshore Homes 0.052 1.059 0.114 Manasota Key Cemetery 0.027 1.063 0.092 South Florida Dunwody, Captiva, Hutchinson Island, Highland Beach, Margate Blount 0.106 75.0 Dunwody 0.190 0.727 0.017 Hutchinson Island 0.106 1.039 0.055 Highland Beach 0.142 0.899 0.046 Margate Blount 0.211 1.070 0.201 Captiva 0.104 0.942 0.045

PAGE 257

257 Table D 5 Continued Question Sites Unbiased F ST % variance accounted for by first two eigenvectors Within group Phenotypic Variance r(ii) Observed Residual Captiva, Hutchinson Island, Highland Beach, Margate Blount 0.120 72.7 Hutchinson Island 0.097 1.041 0.032 Highland Beach 0.126 0.890 0.087 Margate Blount 0.128 1.067 0.093 Captiva 0.128 0.937 0.038 Weeden Island/ Alachua/ St. Johns/ Fort Walton Crystal River, Hughes Island, Bay Pines, Henderson Mound, Browne Site 5 0.171 94.1 Henderson Mound 0.079 0.865 0.181 Hughes Island 0.080 1.090 0.045 Browne Site 5 0.52 9 1.106 0.572 Bay Pines 0.074 0.766 0.285 Crystal River 0.093 0.878 0.152 Crystal River, Hughes Island, Bay Pines, Fort Walton Temple Mound 0.104 100.0 Hughes Island 0.072 1.072 0.039 Bay Pines 0.000 0.712 0.402 Crystal River 0.165 0.863 0.067 Fort Walton Temple Mound 0.178 1.345 0.430 Manasota/ Safety Harbor Crystal River, McKeithan, Hughes Island, Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Yellow Bluffs, Sarasota Bay Mound, Safety Harbor, Fort Walton Temple Mound 0.137 64.0 Palmer 0.093 0.782 0.221 Hughes Island 0.037 1.073 0.008 Bay Pines 0.069 0.678 0.351 Dunwody 0.297 0.812 0.035 McKeithan 0.247 1.356 0.524 Sarasota Bay Mound 0.214 0.447 0.422 Yellow Bluffs 0.178 1.620 0.711 Casey Key 0.163 0.767 0.158 Crystal River 0.129 0.872 0.091

PAGE 258

258 Table D 5 Continued Question Sites Unbiased F ST % variance accounted for by first two eigenvectors Within group Phenotypic Variance r(ii) Observed Residual Bayshore Homes 0.071 0.943 0.083 Manasota 0.050 0.962 0.088 Fort Walton Temple Mound 0.105 1.275 0.286 Safety Harbor 0.124 0.819 0.149 Crystal River, Hughes Island, Bay Pines, Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Sarasota Bay Mound, Safety Harbor 0.111 70.7 Palmer 0.086 0.819 0.060 Hughes Island 0.074 1.125 0.234 Bay Pines 0.074 0.728 0.162 Dunwody 0.226 0.839 0.095 Sarasota Bay Mound 0.160 0.459 0.349 Casey Key 0.116 0.812 0.037 Crystal River 0.149 0.916 0.097 Bayshore Homes 0.055 0.989 0.081 Manasota 0.023 1.002 0.062 Safety Harbor 0.146 0.860 0.039 Bayshore Homes, Casey Key, Dunwody, Manasota Key Cemetery, Palmer Mound, Sarasota Bay Mound, Safety Harbor 0.121 86.8 Palmer 0.053 0.854 0.066 Dunwody 0.355 0.853 0.226 Sarasota Bay Mound 0.147 0.458 0.372 Casey Key 0.081 0.856 0.037 Bayshore Homes 0.045 1.030 0.102 Manasota 0.032 1.036 0.095 Safety Harbor 0.130 0.896 0.051 Casey Key, Manasota Key Cemetery, Palmer Mound, Sarasota Bay Mound, Safety Harbor 0.120 92.8 Palmer 0.080 0.898 0.008 Sarasota Bay Mound 0.102 0.470 0.399 Casey Key 0.068 0.872 0.030 Manasota 0.071 1.077 0.178 Safety Harbor 0.278 0.941 0.242

PAGE 259

259 Table D 5. Continued Question Sites Unbiased F ST % variance accounted for by first two eigenvectors Within group Phenotypic Variance r(ii) Observed Residual Woodland/ Mississippian Palmer, Henderson Mound, Hughes Island, Browne Site 5, Bay Pines, Dunwody, McKeithan, Sarasota Bay Mound, Yellow Bluffs, Casey Key, Crystal River, Bayshore Homes, Hutchinson Island, Manasota, Highland Beach, Margate Blount Mound, Fort Walton Temple Mound, Safety Harbor, Captiva Island 0.104 52.6 Palmer 0.080 0.749 0.214 Henderson Mound 0.082 0.795 0.166 Hughes Island 0.031 1.038 0.023 Browne Site 5 0.113 1.071 0.142 Bay Pines 0.042 0.661 0.342 Dunwody 0.197 0.759 0.081 McKeithan 0.268 1.341 0.575 Sarasota Bay Mound 0.121 0.408 0.512 Yellow Bluffs 0.187 1.522 0.671 Casey Key 0.070 0.755 0.218 Crystal River 0.094 0.832 0.116 Bayshore Homes 0.071 0.901 0.071 Hutchinson Island 0.122 1.055 0.136 Manasota 0.043 0.910 0.092 Highland Beach 0.089 0.904 0.049 Margate Blount Mound 0.085 1.150 0.192 Fort Walton Temple Mound 0.096 1.233 0.286 Safety Harbor 0.141 0.790 0.109 Captiva Island 0.050 0.938 0.057 Woodland Palmer, Henderson Mound, Hughes Island, Browne Site 5, Bay Pines, Dunwody, McKeithan, Yellow Bluffs, Casey Key, Crystal River, Bayshore Homes, Hutchinson Island, Manasota, Highland Beach, Margate Blount Mound, Captiva Island 0.099 51.9 Palmer 0.087 0.764 0.230 Henderson Mound 0.073 0.813 0.196 Hughes Island 0.032 1.058 0.004 Browne Site 5 0.112 1.097 0.130 Bay Pines 0.035 0.670 0.380 Dunwody 0.194 0.772 0.105 McKeithan 0.247 1.386 0.566 Yellow Bluffs 0.209 1.548 0.687

PAGE 260

260 Table D 5. Continued Question Sites Unbiased F ST % variance accounted for by first two eigenvectors Within group Phenotypic Variance r(ii) Observed Residual Casey Key 0.061 0.775 0.247 Crystal River 0.085 0.846 0.150 Bayshore Homes 0.072 0.919 0.092 Hutchinson Island 0.106 1.073 0.100 Manasota 0.053 0.923 0.108 Highland Beach 0.077 0.916 0.089 Margate Blount Mound 0.095 1.174 0.188 Captiva Island 0.053 0.954 0.077 Mississippian Sarasota Bay Mound, Safety Harbor, Fort Walton Temple Mound 0.549 100.0 Sarasota Bay Mound 0.092 0.384 1.141 Safety Harbor 0.793 1.130 0.783 Fort Walton Temple Mound 0.763 0.757 0.358

PAGE 261

261 Table D 6 The biological distances and standard errors for those distances between the sites. Palmer Henderson Mound Hughes Island Browne Site 5 Bay Pines Dunwody McKeithan Sarasota Bay Mound Yellow Bluffs Casey Key Palmer 0.077 0.060 0.064 0.074 0.082 0.182 0.126 0.153 0.036 Henderson Mound 0.286 0.065 0.106 0.093 0.122 0.171 0.192 0.196 0.089 Hughes Island 0.184 0.012 0.104 0.061 0.121 0.151 0.152 0.170 0.073 Browne Site 5 0.168 0.245 0.274 0.119 0.084 0.208 0.173 0.204 0.091 Bay Pines 0.186 0.105 0.000 0.291 0.122 0.177 0.184 0.208 0.097 Dunwody 0.261 0.318 0.352 0.054 0.255 0.246 0.229 0.235 0.113 McKeithan 0.514 0.204 0.114 0.461 0.197 0.706 0.294 0.283 0.178 Sarasota Bay Mound 0.111 0.348 0.118 0.219 0.243 0.569 0.560 0.244 0.146 Yellow Bluffs 0.286 0.374 0.231 0.429 0.409 0.614 0.483 0.233 0.194 Casey Key 0.035 0.188 0.111 0.203 0.195 0.321 0.311 0.106 0.424 Crystal River 0.263 0.183 0.080 0.221 0.110 0.312 0.207 0.393 0.206 0.292 Bayshore Homes 0.042 0.256 0.195 0.164 0.172 0.241 0.407 0.215 0.250 0.115 Hutchinson Island 0.185 0.174 0.200 0.220 0.119 0.305 0.447 0.356 0.503 0.153 Manasota 0.017 0.213 0.084 0.133 0.070 0.189 0.495 0.094 0.315 0.049 Bird Island 0.106 0.241 0.057 0.201 0.125 0.428 0.288 0.143 0.408 0.101 Highland Beach 0.271 0.066 0.042 0.276 0.021 0.296 0.177 0.290 0.478 0.138 Margate Blount Mound 0.173 0.133 0.177 0.245 0.120 0.349 0.358 0.295 0.449 0.116 Fort Walton Temple Mound 0.288 0.007 0.072 0.277 0.118 0.434 0.138 0.249 0.169 0.282 Safety Harbor 0.141 0.435 0.260 0.229 0.270 0.325 0.614 0.097 0.206 0.213 Captiva Island 0.104 0.244 0.105 0.124 0.075 0.139 0.344 0.093 0.296 0.130 The biological distances are on the bottom half of the table and the standard errors for each analysis are on the top half.

PAGE 262

262 Table D 6. Continued Crystal River Bayshore Homes Hutchinson Island Manasota Bird Island Highland Beach Margate Blount Mound Fort Walton Temple Mound Safety Harbor Captiva Island Palmer 0.037 0.018 0.045 0.017 0.056 0.037 0.064 0.062 0.036 0.021 Henderson Mound 0.068 0.076 0.078 0.075 0.105 0.052 0.090 0.057 0.101 0.072 Hughes Island 0.050 0.064 0.076 0.053 0.073 0.043 0.092 0.064 0.078 0.050 Browne Site 5 0.073 0.066 0.084 0.065 0.100 0.078 0.106 0.097 0.080 0.058 Bay Pines 0.066 0.074 0.078 0.063 0.096 0.051 0.095 0.083 0.093 0.058 Dunwody 0.091 0.083 0.102 0.080 0.135 0.089 0.126 0.122 0.099 0.067 McKeithan 0.144 0.171 0.187 0.185 0.184 0.139 0.194 0.149 0.203 0.160 Sarasota Bay Mound 0.170 0.145 0.176 0.128 0.161 0.156 0.185 0.166 0.132 0.123 Yellow Bluffs 0.144 0.150 0.194 0.162 0.201 0.180 0.206 0.154 0.149 0.154 Casey Key 0.070 0.050 0.066 0.044 0.077 0.053 0.079 0.088 0.068 0.049 Crystal River 0.036 0.059 0.037 0.071 0.037 0.063 0.054 0.049 0.031 Bayshore Homes 0.188 0.045 0.025 0.060 0.041 0.066 0.057 0.036 0.029 Hutchinson Island 0.297 0.149 0.048 0.089 0.044 0.077 0.079 0.071 0.048 Manasota 0.163 0.052 0.153 0.060 0.038 0.062 0.066 0.041 0.024 Bird Island 0.204 0.117 0.262 0.099 0.072 0.092 0.083 0.082 0.065 Highland Beach 0.209 0.271 0.142 0.180 0.216 0.062 0.054 0.062 0.029 Margate Blount Mound 0.141 0.161 0.167 0.112 0.148 0.134 0.085 0.086 0.062 Fort Walton Temple Mound 0.167 0.198 0.294 0.262 0.168 0.179 0.177 0.078 0.057 Safety Harbor 0.245 0.117 0.351 0.132 0.244 0.439 0.282 0.341 0.039 Captiva Island 0.173 0.152 0.211 0.066 0.179 0.158 0.157 0.233 0.178

PAGE 263

263 Table D 6. Continued Palmer Henderson Mound Hughes Island Browne Site 5 Bay Pines Dunwody McKeithan Sarasota Bay Mound Yellow Bluffs Casey Key Palmer 0.075 0.059 0.065 0.075 0.082 0.189 0.125 0.150 0.036 Henderson Mound 0.270 0.064 0.106 0.093 0.121 0.178 0.191 0.192 0.087 Hughes Island 0.174 0.009 0.107 0.066 0.121 0.165 0.150 0.167 0.071 Browne Site 5 0.178 0.248 0.296 0.118 0.083 0.212 0.178 0.202 0.093 Bay Pines 0.196 0.108 0.000 0.282 0.121 0.178 0.189 0.205 0.099 Dunwody 0.256 0.304 0.355 0.048 0.253 0.248 0.229 0.232 0.112 McKeithan 0.576 0.249 0.197 0.494 0.205 0.729 0.310 0.287 0.188 Sarasota Bay Mound 0.104 0.339 0.109 0.249 0.275 0.569 0.681 0.241 0.145 Yellow Bluffs 0.265 0.349 0.213 0.417 0.386 0.593 0.515 0.216 0.190 Casey Key 0.032 0.174 0.100 0.221 0.209 0.315 0.382 0.099 0.400 Crystal River 0.254 0.179 0.087 0.216 0.111 0.305 0.267 0.384 0.186 0.288 Bayshore Homes 0.051 0.246 0.205 0.160 0.168 0.232 0.427 0.236 0.237 0.124 Hutchinson Island 0.190 0.173 0.215 0.208 0.111 0.292 0.459 0.375 0.475 0.166 Manasota 0.015 0.204 0.081 0.144 0.088 0.191 0.565 0.084 0.291 0.046 Highland Beach 0.260 0.060 0.044 0.273 0.027 0.283 0.230 0.288 0.451 0.132 Margate Blount Mound 0.171 0.135 0.184 0.235 0.122 0.336 0.418 0.294 0.424 0.125 Fort Walton Temple Mound 0.293 0.014 0.094 0.271 0.106 0.421 0.147 0.283 0.160 0.290 Safety Harbor 0.135 0.411 0.250 0.226 0.261 0.314 0.647 0.099 0.184 0.204 Captiva Island 0.101 0.231 0.105 0.126 0.077 0.138 0.389 0.098 0.271 0.128

PAGE 264

264 Table D 6. Continued Crystal River Bayshore Homes Hutchinson Island Manasota Highland Beach Margate Blount Mound Fort Walton Temple Mound Safety Harbor Captiva Island Palmer 0.037 0.019 0.046 0.017 0.036 0.064 0.063 0.035 0.021 Henderson Mound 0.068 0.075 0.078 0.074 0.051 0.090 0.059 0.099 0.071 Hughes Island 0.051 0.065 0.078 0.052 0.044 0.092 0.068 0.077 0.050 Browne Site 5 0.072 0.065 0.082 0.066 0.078 0.104 0.096 0.080 0.058 Bay Pines 0.066 0.074 0.076 0.065 0.052 0.095 0.081 0.092 0.058 Dunwody 0.090 0.082 0.100 0.080 0.087 0.124 0.121 0.097 0.067 McKeithan 0.153 0.174 0.189 0.193 0.147 0.202 0.150 0.206 0.166 Sarasota Bay Mound 0.169 0.148 0.178 0.126 0.155 0.184 0.171 0.132 0.123 Yellow Bluffs 0.141 0.148 0.191 0.159 0.176 0.203 0.152 0.146 0.150 Casey Key 0.069 0.051 0.068 0.043 0.052 0.080 0.089 0.067 0.049 Crystal River 0.036 0.058 0.037 0.036 0.062 0.054 0.047 0.030 Bayshore Homes 0.187 0.044 0.026 0.041 0.066 0.056 0.036 0.029 Hutchinson Island 0.285 0.141 0.049 0.044 0.077 0.077 0.070 0.047 Manasota 0.158 0.062 0.160 0.038 0.062 0.067 0.040 0.024 Highland Beach 0.202 0.265 0.143 0.176 0.061 0.055 0.060 0.028 Margate Blount Mound 0.131 0.168 0.164 0.113 0.130 0.086 0.085 0.061 Fort Walton Temple Mound 0.170 0.188 0.278 0.270 0.181 0.185 0.077 0.057 Safety Harbor 0.231 0.116 0.336 0.126 0.416 0.269 0.330 0.038 Captiva Island 0.164 0.150 0.205 0.065 0.151 0.150 0.229 0.167 Gauthier Little Salt Spring Windover Bay West Bird Island Republic Groves Gauthier 0.064 0.029 0.055 0.099 0.087 Little Salt Spring 0.147 0.052 0.080 0.122 0.117 Windover 0.154 0.094 0.055 0.091 0.068 Bay West 0.120 0.131 0.156 0.133 0.108 Bird Island 0.494 0.401 0.476 0.562 0.165 Republic Groves 0.560 0.489 0.401 0.450 1.094

PAGE 265

265 Table D 6. Continued Gauthier Little Salt Spring Windover Bay West Gauthier 0.073 0.032 0.060 Little Salt Spring 0.177 0.047 0.091 Windover 0.175 0.106 0.063 Bay West 0.122 0.141 0.196 Palmer Mound Hughes Island Bay Pines Dunwody Casey Key Crystal River Bayshore Homes Manasota Palmer Mound 0.069 0.085 0.105 0.037 0.046 0.023 0.022 Hughes Island 0.252 0.069 0.144 0.094 0.056 0.068 0.058 Bay Pines 0.266 0.000 0.143 0.120 0.064 0.076 0.071 Dunwody 0.472 0.540 0.403 0.136 0.104 0.091 0.102 Casey Key 0.033 0.251 0.356 0.505 0.087 0.054 0.049 Crystal River 0.394 0.116 0.086 0.420 0.475 0.038 0.043 Bayshore Homes 0.083 0.214 0.169 0.299 0.138 0.212 0.028 Manasota 0.042 0.114 0.113 0.370 0.078 0.219 0.068 Palmer Dunwody Yellow Bluffs Casey Key Bayshore Homes Manasota Palmer 0.103 0.188 0.036 0.022 0.023 Dunwody 0.517 0.282 0.138 0.097 0.106 Yellow Bluffs 0.627 1.139 0.225 0.183 0.192 Casey Key 0.042 0.608 0.751 0.051 0.051 Bayshore Homes 0.081 0.407 0.551 0.136 0.025 Manasota 0.060 0.470 0.609 0.112 0.062 Palmer Bay Pines Casey Key Bayshore Homes Manasota Palmer 0.090 0.040 0.028 0.021 Bay Pines 0.294 0.126 0.081 0.082 Casey Key 0.047 0.382 0.057 0.051 Bayshore Homes 0.122 0.194 0.144 0.033 Manasota 0.037 0.179 0.084 0.104

PAGE 266

266 Table D 6. Continued Palmer Hughes Island Bay Pines Dunwody Sarasota Bay Mound Casey Key Crystal River Bayshore Homes Manasota Safety Harbor Palmer 0.068 0.084 0.093 0.151 0.036 0.044 0.023 0.020 0.048 Hughes Island 0.261 0.070 0.135 0.154 0.093 0.057 0.070 0.060 0.087 Bay Pines 0.272 0.003 0.125 0.207 0.119 0.065 0.076 0.073 0.099 Dunwody 0.367 0.478 0.274 0.259 0.126 0.093 0.082 0.087 0.106 Sarasota Bay Mound 0.271 0.126 0.393 0.820 0.172 0.184 0.166 0.149 0.157 Casey Key 0.035 0.263 0.366 0.430 0.263 0.085 0.055 0.047 0.083 Crystal River 0.377 0.134 0.099 0.325 0.506 0.470 0.038 0.042 0.052 Bayshore Homes 0.085 0.244 0.181 0.230 0.359 0.151 0.214 0.028 0.041 Manasota 0.037 0.138 0.135 0.252 0.220 0.069 0.222 0.073 0.051 Safety Harbor 0.288 0.344 0.322 0.392 0.255 0.350 0.286 0.162 0.237 Hutchinson Island Highland Beach Margate Blount Captiva Hutchinson Island 0.061 0.100 0.054 Highland Beach 0.325 0.090 0.039 Margate Blount 0.346 0.384 0.091 Captiva 0.278 0.307 0.428

PAGE 267

267 Table D 7 Correlation Matrix: Biodistance to Temporal and Geographic distances Question Correlation Temporal distance Geographic distance All Pearson Correlation 0 .457 ** 0.067 Sig. (2 tailed) 0.000 0.231 N 325 325 All without Warm Mineral Springs Pearson Correlation 0.059 0.041 Sig. (2 tailed) 0.306 0.475 N 300 300 Archaic Pearson Correlation .66 7** 0.206 Sig. (2 tailed) 0.001 0.371 N 21 21 Archaic without Warm Mineral Springs Pearson Correlation 0.202 0.308 Sig. (2 tailed) 0.471 0.265 N 15 15 Weeden Island/Manasota Pearson Correlation 0.031 0 .43 3** Sig. (2 tailed) 0.835 0.003 N 46 46 Weeden Island Pearson Correlation 0.018 0 .76 2* Sig. (2 tailed) 0.970 0.046 N 7 7 Manasota Pearson Correlation 0.281 0.370 Sig. (2 tailed) 0.310 0.175 N 15 15 South Spearman's rho Correlation 0.055 .79 4** Sig. (2 tailed) 0.881 0.006 N 10 10 Woodland Pearson Correlation 0.050 .192 Sig. (2 tailed) 0.587 0.036 N 120 120 Mississippian Spearman's rho Correlation 0.332 0.193 Sig. (2 tailed) 0.226 0.491 N 15 15 *. Correlation is significant at the 0.05 level (2 tailed). **. Correlation is significant at the 0.01 level (2 tailed).

PAGE 268

268 LIST OF REFERENCES Anderson DG. 1996a. Approaches to modeling regional settlement in the Archaic period Southeast. In: Sassaman KE, and Anderson D G, editors. Archaeology of the M id Holocene Southeast. Gainesville: University Press of Florida. p 157 176. Anderson DG. 1996b. Models of Paleoindian and early Archaic settlement in the lower Southeast. In: Anderson DG, and Sassaman KE editors. The Paleoindian and E arly Archaic Southeast. Tuscaloosa: The University of Alabama Press. p 27 57. Anderson DG, and Mainfort J r RC. 2002a. An Introduction to Woodland ar chaeology in the Southeast. In: Anderson DG, and Mainfort J r R C, editors. The Woodland Southeast. Tuscaloosa: The University of Alabama Press. p 1 19. Anderson DG, and Mainfort J r RC, editors. 2002b. The Woodland Southeast. Tuscaloosa: The Uni versity of Al abama Press. Anderson DG, O'Steen LD, and Sassaman KE. 1996. Environmental and chronological considerations. In: Anderson DG, and Sassaman KE editors. The Paleoindian and E arly Archaic Southeast. Tuscaloosa: The University of Alabama Press. p 3 15. Anders on DG, and Sassaman KE, editors. 1996. The Paleoindian and Early Archaic Southeast. Tuscaloosa: The Uni versity of Alabama Press. Ashley K, and White NM. 2012. Late Prehistoric Florida: Archaeology at the edge of the Mississippian world. Gainesville: Univer sity Press of Florida. Austin R. 2001. An archaeological survey of unincorporated Alachua County, Florida (Phases 1 and 2). Report on file at Florida Division of Historical Resources, Tallahassee. Austin R. 2012. Bayshor e Homes. Personal communication Gai nesville. Austin R, Mitchem J, Fradkin A, Foss JE, Drwiega S, and Allred L. 2008. Bayshore Homes archaeological survey and National Register Evaluation. Report o n file at Central Gulf Coast Archaeological Society, Pinellas Park. Austin R, Mitchem J, and We isman BR. In press Radiocarbon dates and late prehistory of Tampa Bay. In: Wallis N, and Randall A, editors. New histories of Pre Columbian Florida. Gainesville: University Press of Florida. Basu A, Namboodiri K, Weltkamp LR, Brown WH, Pollitzer WS, and Spivey MA. 1976. Morphology, serology, dermatoglyphics, and microevolution of some village populations in Haiti, West Indies. Hum Biol 48(2):245 269. Beriault J, Carr RS, Stipp JJ, Johnson R, and Meeder J. 1981. The archaeological salvage of the Bay West s ite, Collier County, Florida. Fla Anthropol 34(2):39 58.

PAGE 269

269 Buikstra J E 1976. Hopewell in Lower Illinois Valley: A regional study of human biological variability and prehistoric mortu ary behavior. Evanston : Northwest University, Archaeological Program Scient ific Papers, no. 2. Buikstra JE, and Ubelaker DH. 1994. Standards for data collection from human skeletal remains: Proceedings of a Seminar at the Field Museum of Natural History. Fayetteville: Arkansas Archaeological Survey Press. Bullen RP. 1951. The eni gmatic Crystal River site. Am Antiq 17(2):142 143. Bullen RP, and Bullen AK. 1976. The Palmer site. Florida Anthrop ological Society Publications 8: 1 53. Cadien JD, Harris EF, Jones WP, and Mandarion LJ. 1974. Biological lineages, skeletal populations, an d microevolution. Yrbk Phy s Anthropol 18:194 201. Callsen P. 2008. Glades period settlement patterns in the Everglades culture area. Masters Thesis, Florida Atlantic University. Carr R S 1981. Florida Anthropologist inter view with Calvin Jones, Part II. E x cavations of an Archaic cemetery in Cocoa Beach, Florida. Fla Anthropol 34(2):81 89. Carr RS. 2012. Mississippian influence in the Glades, Belle Glade, and East Okeechobee areas of South Florida. In: Ashley K, and White NM, editors. Late p rehistor ic Florid a: Archaeology at the edge of the Mississippian w orld. Gainesville: University Press of Florida. p 62 80. Carr RS, and Steele W. 1993. An archaeological assessment of the Hutchinson Island Site (8MT37), Martin County, Florida. AHC technical report #68. Mia mi: Archaeological and Historical Conservancy. Chatters JC. 2010. Peopling the Americas via multiple migrations from Beringia: Evidence from the early Holocene of the Columbia Plateau. In: Auerbach BM, editor. Human variation in the Americas: The integration of archaeology and biological an thropology. Carbondale : Southern Illinois University, Center for Archaeological Investigations. p 51 76. Ch ristman H. 1954. Hugh's Island M ound site. Unpublished paper on file at the Florida Museum of Natural History, Gainesville. Clausen CJ, Brooks HK, and Wesolowsky AB. 1975. The early man site at Warm Mineral Springs, Florida. J Field Archaeol 2(3):191 213. Clausen CJ, Cohen AD, Emiliani C, Holman JA, and Stipp JJ. 1979. Little Salt Spring, Florida: A unique underwater site. Science 203(4381):609 614. Cockrell WA, and Murphy L. 1978. Pleistocene man in Florida. Archaeol East N Am 6:1 13.

PAGE 270

270 Collins HB. 1929. The "Lost" Calusa Indians of Southwestern Florida. Explor ations and field work of the Smithsonian Institution in 1928. Washington DC: The Smithsonian Institution. p 151 156. Cordell AS. 2004. Past variability and possible manufacturing origins of late Archaic fiber tempered pottery from selected sites in peninsu lar Florida. In: Sanders R, and Hayes CT, editors. Early pottery: technology, function, style, and interaction in the lower Southeast. Tuscaloosa: The University of Alabama Press. p 63 104. Corruccini RS. 1974. An examination of the meaning of cranial disc rete traits for human skeletal biological studies. Am J Phy s Anthropol 40 (3) :425 446. Devor EJ. 1987. Transmission of human craniofacial dimensions. J Craniofac Genet Dev Biol 7 (2) :95 106. Doran GH. 2002a. Introduction to wet sites and Windover (8DR246) in vestigations. In: Doran GH, editor. Windover: Multidisciplinary investigations of an early Archaic Florida cemetery. Gainesville: University Press of Florida. Doran GH, editor. 2002b. Windover: Multidis ciplinary investigations of an E arly Archaic Florida c emetery. Gainesville: University Press of Florida. Droessler J. 1981. Craniometry and biological distance: Biocultural continuity and change at the late Woodland Mississippi an interface. Evanston : Center for American Archaeology at Northwestern University Dunbar JS, and Waller BI. 1983. A distribution analysis of the Clovis/Suwanee Paleo Indian site of Florida a geographic approach. Fla Anthropol 36(1 2):18 30. Faught MK. 2004. The underwater archaeology of paleolandscapes, Apalachee Bay, Florida. Am Anti q 69(2):275 289. Faught MK, and Donoghue JF. 1997. Marine inundated archaeological sites and paleofluvial systems: examples from a karst controlled continental shelf setting in Apalachee Bay, Northeastern Gulf of Mexico. Geoarchaeology 12(5):417 458. Faugh t MK, and Waggoner, Jr JC. 2012. The early Archaic to middle Archaic transition in Florida: An argument for discontinuity. Fla Anthropol 65(3):153 176. Friedlaender JS, Sgarmella Zonta LA, Kidd KK, Lai LYC, Clark P, and Walsh RJ. 1971. Biological diverge nces in South Central Bougainville: An analysis of blood polymorphism gene frequencies and anthropometric measurements utilizing tree models, and a comparison of these variables with linguistic geographic, and migrational "d istances." Am J Hum Genet 23(3) :253 270. Gallagher JC, and Warren LO. 1975. The Bay Pines site, Pinellas County, Florida. Fla Anthropol 28(3):96 116.

PAGE 271

271 Goggin JM. 1949. The archaeology of the Glades area, southern Florida. New Haven: Yale Peabody Museum. Gold M. 2005. Osteological Analysis of the Manasota Period Dunwody Site. Masters Thesis, University of Florida. Gold M. 2006. Osteological analysis of the Manasota period Dunwody site. Fla Anthropol 59(1):35 51. Greenman EF. 1938. Hopewellian traits in Florida. Am Antiq 3(4):327 33 2. Griffin JW. 1988. The archeology of Everglades National Park: A synthesis. Tallahassee: National Park Service, Southeast Archeological Center. Griffin MC, Lambert PM, and Driscoll EM. 2001. Biological relations hips and population history of n ative peopl es in Spanish Florida and the American Southeast. In: Larsen CS, editor. Bioarchaeology of Spanish Florida: The impact of colonialism. Gainesville: University Press of Florida. p 226 273. Harpending HC, and Ward R. 1982. Chemical systematics and human evol ution. In: Nitecki M, editor. Biochemical aspects of evolutionary biology. Chicago: The University of Chicago Press. p 213 226. Holman R. 1975. A l etter report to Mr. Dennis English on the excavations of the Hutchinson Island Burial Mound. On file at Flori da Division of Historical Resources, Tallahassee. Hooton EA. 1973. Racial types in America and their relations to Old W orld types. In: Jenness D, editor. The American Aborigines: Their origin and antiquity Fifth Pacific Science Congress, Canada, 1933. New York: Cooper Square Publishers. Howells WW. 1973. Cranial variation in man: A study by multivariate analysis of patterns of difference among recent human populations. Cambridge: Ha rvard University Deland : T he Florida Historical Society. Academy of Medicine 4(7):802 816. IBM Corp. Released 2010. IBM SPSS Statistics for Windows, Version 19.0. Armonk, NY: IBM Corp Jantz RL, a nd Owsley DW. 2001. Variation among early North American crania. Am J Phys Anthropol 114(2):146 155.

PAGE 272

272 Jefferies RW. 2004. Regional cultures, 700 B.C. A.D. 1000. In: Fogelson R, and Sturtevant W, editors. The handbook of the North American Indian: Southeast. Washington D C: Smithsonian Institution. p 115 127. Katzmarzyk C. 1998. Evidence of stress in a Precolumbian population from Mound G at the Crystal River site, Florida. Masters Thesis, University of Florida. Key P, and Jantz RL. 1981. A multivari ate analysis of temporal change in Arikara craniometrics: A methodological approach. Am J Phys Anthropol 55(2):247 259. Killgrove K. 2002. Defining relationships between Native American groups: A biodistance study of the North Carolina coastal plain. Maste rs Thesis, East Carolina University. Killgrove K. 2009. Rethinking taxonomies: Skeletal variation on the North Carolina coastal plain. Southeast Archaeol 28(1):87 100. Konigsberg LW. 1990a. Analysis of prehistoric biological variation under a model of isol ation by geographic and temporal distance. Hum Biol 62(1):49 70. Konigsberg LW. 1990b. Temporal aspects of biological distance: Serial correlation and trend in a prehistoric skeletal lineage. Am J Phys Anthropol 82(1):31 44. Lane RA, and Sublett AJ. 1972. Osteology of social organization: Residence pattern. Am Antiq 37(2):186 201. Livingston FB. 1972. Genetic drift and polygenic inheritance. Am J Phys Anthropol 37(1):117 126. Loucks LJ. 1976. Early Alachua tradition burial ceremonialism: The Henderson Mound Alachua County, Florida. Masters Thesis. Gainesville: University of Florida. Lovejoy DA. 1985. Salvage archaeology on Bird Island (8DI52). Report on file with University of South Florida, Tampa. Luer G. 1999. Cedar Point: A late Archaic thro ugh Safety Ha rbor o ccupation on Lemon Bay, Charlotte County, Florida. Maritime Archaeology of Lemon Bay, Florida Anthropological Society Publications 14:43 56. Luer G. 2011. The Yellow Bluf fs Mound revisited: A Manasota p eriod burial mound in Sarasota. Fla Anthropol 64 (1):5 32. Luer G, and Almy M. 1982. A definition of the Manasota culture. Fla Anthropol 35:34 58. Luer G, and Almy M 1987. The Laurel Mound (8SO98) and radial burials with comments on the Safety Harbor period. Fla Anthropol 40(4):301 320.

PAGE 273

273 Luer G, and Hughes D. 2011. Radiocarbon dating the Yellow Bluffs Mound (8SO4), Sarasota, Florida. Fla Anthropol 64(1):33 46. Lundstrom A. 1955. The significance of genetic and non genetic factors in the profile of the facial skeleton. Am J Orthod 41 (12) :91 0 916. Maples WR. 1987. Analysis of skeletal remains recovered at the Gauthier site, Brevard County, Florida. Miscellaneous Project Report Series No 31. Gainesville: Florida State Mus eum, Department of Anthropology. Report on file at the Florida Museum of Natural History, Gainesville. Marquardt W, and Walker KJ. 2012. Southwest Florida during the Mississippi period. In: Ashley K, and White NM, editors. Late Prehistoric Florida: Archaeology at the Edge of the Mississippian World. Gainesville: University Pres s of Florida. p 29 61. McFadden P. 2011. Discussions about Bird Island. Personal communication Kles M. Gainesville. McGoun WE. 1993. Prehistoric peoples of South Florida. Tuscaloosa: The University of Alabama Press. Milanich J T 1972. Excavations at the Ye llow Bluffs Whitaker Mound, Sarasota, Florida. Fla Anthropol 25(1):21 41. Milanich J T 1994. Archaeology of Precolumbian Florida. Gainesville: University Press of Florida. Milanich J T 1995. Florida Indians and the invasion from Europe. Gainesville: Univer sity Press of Florida. Milanich J T 1998. Florida Indians from ancient times to the present. Gainesville: University Press of Florida. Milanich J T 2002. Weeden Island cultures. In: Anderson DG, and Mainfort J, Robert C, editors. The Woodland Southeast. T uscaloosa: The University of Alabama Press. p 352 372. Milanich J T 2004. Prehistory of Florida After 500 B.C. In: Fogelson R, and Sturtevant W, editors. H andbook of North American Indians: Southeast. Washington DC.: Smithsonian Institution. p 191 203. Mil anich JT, Cordell AS, Knight J, Vernon J, Kohler TA, and Sigler Lavelle BJ. 1997. Archaeology of northern Florida, A.D. 200 900: The McKeithan Weeden Island culture. Gainesville: University Press of Florida. Milanich J T and Fairbanks C. 1980. Florida arch aeology. Orlando: Academic Press.

PAGE 274

274 Miles AEW. 2001. The Miles method of assessing age from tooth wear revisited. J Archaeol Sci 28(9):973 982. Mitchem J. 1989. Redefining S afety Harbor: Late Prehistoric/ Protohistoric archaeology in west peninsular Florida. PhD dissertation. Gainesville: University of Florida. Mitchem J. 2012. Safety Harbor: Mississippian influence in the Circum Tampa Bay Region. In: Ashley K, and White NM, editors. Late prehistoric Florida: Archaeology at the edge of the Mississippian world. Gainesville: Universi ty Press of Florida. p 172 185. Murphy T. 1959a. The changing pattern of dentine exposure in human tooth attrition. Am J Phys Anthropol 17(3):167 178. Murphy T. 1959b. Gradients of dentine exposure in human molar tooth attrition. Am J Phys Anthropol 17(3):179 186. Neumann GK. 1954. Archaeology and race in the American Indian. Yrbk Phys Anthropol 8:213 242. Ousley SD, Billeck WT, and Hollinger RE. 2005. Federal repatriation legislation and the role of physical anthropology in repatriati on. Yrbk Phys Anthropol 48:2 32. Owsley DW, Bennett SM, and Jantz RL. 1982. Intercemetery morphological variation in Arikara crania from the Mobridge S ite (39WW1). Am J Phys Anthropol 58(2):179 185. Pearson MP. 1982. Mortuary practices, society, and ideolo gy; an ethnoarchaeological study. In: Hodder I, editor. Symbolic and structural archaeology. Cambridge: Cambridge University Press. p 99 113. Pearson MP. 1999. Learning from the dead. In: Pearson MP, editor. The archaeology of death and burial. College Sta tion: Texas A&M University Press. p 1 20. Pietrusewsky M. 2000. Metric analysis of skeletal remains: Methods and applications. In: Katzenberg MA, and Saunders SR, editors. Biological anthropology of the human skeleton. New York: Wiley Liss. p 375 415. Pluc khahn TJ, Thompson VD, and Weisman BR. 2010. Toward a new view of history and process at Crystal River (8CI1). Southeast Archaeol 29(1):164 181. Pollitzer WS. 1958. T he Negroes of Charleston (S.C.): A study of hemoglobin types, serology, and morphology. Am J Phys Anthropol 16(2):241 264. Powell JF, and Neves W. 1999. Craniofacial morphology of the first Americans: Pattern and process in the peopling of the New World. Yrbk Phys Anthropol 42:154 188.

PAGE 275

275 Purdy BA. 1991a. Bay West. The art and archaeology of Florida's wetl ands. Boca Raton: CRC Press p 54 65. Purdy BA. 1991b. Republic Groves. The art and archaeology of Florida's wetl ands. Boca Raton: CRC Press p 167 178. Quinn RL, Tucker BD, and Krigbaum J. 2008. Diet and mobility in Middle Archaic Florida: S table isotopic and faunal evidence from the Harris Creek archaeological site (8Vo24), Tick Island. J Archaeol Sci 35 (8) :2346 2356. Randall AR, and Sassaman KE. 2010. (E)mergent complexities during the Archaic Period in Northeast Florida. In: Alt S, editor. Ancient complexities: New perspectives in Pre Columbian North America. Salt Lake City: The University of Utah Press. p 8 31. Reichs KJ. 1984. Pearls or people: A biometric ana lysis of interregional exchange during Hopewell times. Cent Issues Anthropol 5(2):47 65. Relethford JH. 1994. Craniometric variation among modern human populations. Am J Phys Anthropol 95(1):53 62. Relethford JH. 1996. Genetic drift can obscure population history: Problem and solution. Hum Biol 68(1):29 44. Relethford JH. 2002. Apportionment of global human genetic diversity based on craniometrics and skin color. Am J Phys Anthropol 118(4):393 398. Relethford JH. 2004. Boas and beyond: Migration and craniom etric variation. Am J Hum Biol 16(4):379 386. Relethford JH, and Blangero J. 1990. Detection of differential gene flow from patterns of quantitative variation. Hum Biol 62 (1) :5 25. Relethford JH, Crawford MH, and Blangero J. 1997. Genetic drift and gene fl ow in post famine Ireland. Hum Biol 69(4):443 465. Relethford JH, and Harpending HC. 1994. Craniometric variation, genetic theory, and modern human origins. Am J Phys Anthropol 95 (3) :249 270. Relethford JH, and Lees FC. 1982. The use of quantitative traits in the study of human population structure. Yrbk Phys Anthropol 25:113 132. Rhine S. 1990. Non metric skull racing. In: Gill R, editor. Skeletal attribution of race: Methods for forensi c anthropology. Albuquerque: Max well Museum Anthropological Papers. p 9 20.

PAGE 276

276 Rolland V. 2012. The Alachua of North Central Florida. In: Ashley K, and White NM, editors. Late Prehistoric Florida: Archaeology at the edge of the Mississippian world. Gainesville: University Press of Florida. p 126 148. Rothhammer F, and Silva C. 1990. Craniometrical variation among South American prehistoric populations: Climatic, altitudinal, chronological, and geographic contributions. Am J Phys Anthropol 82(1):9 18. Sanghvi LD. 1953. Comparison of genetical and morphological methods for a study of biological differences. Am J Phys Anthropol 11(3):385 404. Sassaman KE. 2006. Dating and explaining soapstone vessels: A comment on Truncer. Am Antiq 71(1):141 156. Sassaman KE. 2008. The new Archaic, it ain't what is used to be. The SAA Archaeological Record. Washington D.C.: Society for American Archaeology. p 6 12. Saunders R. 2004. Spatial variation in Orange culture pottery: Interaction and function. In: Saunders R, and Hays C, editors. Early pottery: Technology, function, style, and interaction in the lower Southeast. Tuscaloosa: The University of Alabama Press. p 40 62. Schillaci MA, and Stojanowski CM. 2005. Craniometric variation and population history of the prehistoric Tewa. Am J Phys Anthropol 12 6 (4) :404 412. Schofield S. 2003. A model of mig ration: The Alachua culture. Unpublished paper on file at the Florida Museum of Natural History, Gainesville. Schuldenrein J. 1996. Geoarchaeology and the mid Holocene landscape history of the greater Southeast. In: Sassaman KE, and Anderson DG, editors. A rchaeology of the Mid Holocene Southeast. Gainesville: University Press of Florida. p 3 27. Sciulli PW, Lozanoff S, and Schneider KN. 1984. An analysis of diversity in Glacial Kame and Adena skeletal samples. Hum Biol 55(4):603 616. Sciulli PW, and Schneid er KN. 1985. Cranial variation in the terminal Archaic of Ohio. Am J Phys Anthropol 66 (4) :429 443. Sears WH. 1957. Excavations on lower St. Johns River, Florida. Contributions of the Florida State Museum: Social Sciences no. 2. Sears WH. 1958. Burial mound s on the Gulf coastal plain. Am Antiq 23(3):274 284. Sears WH. 1960. The Bayshore Homes Site, St. Petersburg, Florida. Contributions of the Florida State Museum, Social Sciences no. 6.

PAGE 277

277 Sears WH. 1973. The sacred and the secular in prehistoric ceramics. In: Lathrap DW, and Douglas J, editors. Variation in anthropology: Essays in honor of John C McGregor. Urbana: Illinois Archaeological Survey. p 31 42. Sears WH. 1982. Fort Center: An archaeological site in the Lake Oke echobee basin. Gainesville: University of Florida Press. Seasons SM. 2010. An assessment of microevolutionary change among prehistoric Florida populations through the analysis of craniometric data. Masters Thesis, University of South Florida. Silbereisen A. 1954. Hugh's Island Mound (DI45). Unp ublished paper on file at the Florida Museum of Natural History, Gainesville. Skow J. 1986. Surprising finds from a Florida sinkhole. Smithsonian 17(9):72 85. Smith DG, Rolfs BK, Kaestle F, Malhi RS, and Doran GH. 2002. Serum albumin phenotypes and a preli minary study of the Windover mtDNA haplogroups and their anthropological significance. In: Doran GH, editor. Windover: Multidisciplinary inves tigations of an E arly Archaic Florida cemetery. Gainesville: University Press of Florida. Spielman RS. 1973. Diffe rences among Yanomama Indian villages: Do the patterns of allele frequencies, anthropometrics, and map location correspond? Am J Phys Anthropol 39(3):461 480. Steadman DW. 1998. The population shuffle in the Central Illinois Valley: A diachronic model of M ississippian biocultural interaction. World Archaeol 30(2):306 326. Steadman DW. 2001. Mississippians in motion? A population genetic analysis of i nterregional gene flow in West Central Illinois. Am J Phys Anthropol 114 (1) :61 73. Stephenson K, Bense JA, an d Snow F. 2002. Aspects of Deptford and Swift Creek of the South Atlantic and Gulf coastal plains. In: Anderson DG, and Mainfort J, Robert C, editors. The Woodland Southeast. Tuscaloosa: The University of Alabama Press. p 318 351. Sternberg GM. 1875. India n burial mounds and shellheaps near Pensacola, Florida. Proceedings from The Am Assoc Advance Sci. Detroit, Michigan. p 282 292. Stojanowski CM. 2003. Differential phenotypic variability among the Apalachee mission populations of La Florida: A Diachronic P erspective. Am J Phys Anthropol 120 (4) :352 363. Stojanowski CM. 2004. Population history of native groups in pre and postcontact Spanish Florida: Aggregation, gene flow, and genetic drift of the southeastern United States Atlantic Coast. Am J Phys Anthrop ol 123 (4) :316 332.

PAGE 278

278 Stojanowski CM. 2005a. Biocultural histories in La Florida: A bioarchaeological perspective. Tuscaloosa : The University of Alabama Press. Stojanowski CM. 2005b. Spanish colonial effects on Native American mating structure and genetic var iability in Northern and Central Florida: Evidence from Apalachee and Western Timucua. Am J Phys Anthropol 128 (2) :273 286. Stojanowski CM, and Doran GH. 1998. Osteology of the Late Archaic Bird Island site (8DI52), Dixie County, Florida. Fla Anthropol 51(3 ) :139 146 Stojanowski CM, and Johnson KM. 2011. Brief communication: Preliminary radiocarbon Phys Anthropol 145(1):163 167. Stojanowski CM, Larsen CS, Tung TA, and McEwan BG. 2007. Biologica l structure and health implications from tooth size at Mission San Luis de Apalachee. Am J Phys Anthropol 132 (2) :207 222. Stojanowski CM, and Schillaci MA. 2006. Phenotypic approaches for understanding patterns of intracemetery biological variation. Yrbk P hys Anthropol 49:49 88. Tatarek NE, and Sciulli PW. 2000. Comparison of population structure in Ohio's late Archaic and late prehistoric periods. Am J Phys Anthropol 112 (3) :363 376. Thulman DK. 2006. The reconstruction of Paleoindian social organization in North Central Florida. Ph.D. Dissertation, Florida State University. van Vark GN, and Schaafsma W. 1992. Advances in the quantitative analysis of skeletal morphology. In: Saunders SR, and Katzenberg MA, editors. Skeletal biology of past peoples: Research methods. New York: Wiley Liss. p 225 257. Watson JT. 2010. The introduction of agriculture and the foundation of biological variation in the southern Southwest. In: Auerbach BM, editor. Human variation in the Americas: The integration of archaeology and bi ological an thropology. Carbondale : Southern Illinois University, Center for Archaeological Investigations. p 135 171. Watts WA, Grimm EC, and Hussey TC. 1996. Mid Holocene forest history of Florida and the coastal plain of Georgia and South Carolina. In: S assaman KE, and Anderson D G, editors. Archaeology of the M id Holocene S outheast. Gainesville: University Press of Florida. p 28 38. Weisman BR. 1995. Crystal River: A ceremonial mound center of the Florida Gulf Coast. Tallahassee: Florida Bureau of Archaeological Research. Wharton BR, Ballo GR, and Hope ME. 1981. The Republic Groves site, Hardee County, Florida. Fla Anthropol 34(2):59 80.

PAGE 279

279 Wheeler RJ, Kennedy WJ, and Pepe JP. 2002. The archaeology of coastal Palm Beach County. Fla Anthropol 55(3 4):11 9 156. Widmer RJ. 2002. The Woodland archaeology of South Florida. In: Anderson DG, and Mainfort J, Robert C, edi tors. The Woodland S outheast. Tuscaloosa: The University of Alabama Press. p 373 397. Willey GR. 1949. Archaeology of the Florida Gulf Coast. Baltimore: The Lord Baltimore Press. Williams Blangero S, and Blangero J. 1989. Anthropometric variation and the genetic structure of the Jirels of Nepal. Hum Biol 61(1):1 12. Williams WB. 1983. Bridge to the past: Excavations at the Margate Blount site. Fla Anthropol 36(3 4):142 153. Winland KJ. 2002. Disease and population ecology in the East Okeechobee area. Fla Anthropol 55(3 4):199 220. W orth J. 2013. Pineland during the Spanish Period. In: Marquardt W, and Walker KJ, editors. The archaeology of Pinel and: A coastal Southwest Florida site complex, AD 50 1710. Gainesville: Institute of Archaeology and Paleoenvironmental Studies. p 767 792. Yates WB. 2000. Implications to Late Archaic exchange networks in the southeast as indicated by the archaeological e vidence of prehistoric soapstone vessels throughout Florida. Ph.D. Dissertation, Florida State University.

PAGE 280

280 BIOGRAPHICAL SKETCH Maranda Almy Kles began her anthropological career at the foot of her mother going to archaeological sites. She later ear ned a Bachelor of Arts and Master of Arts in a nthropology from the University o f Florida followed by a Master of Science in t oxicology from the University of Florida. While pursuing her degrees Ms. Kles has had the opportunity to conduct research in the N ational Archives, various local museums and historical societies, the Smithsonian Institution, and various universities and state facilities. Ms. Kles also worked as a Death Investigator for the District Twelve Medical ore returning for her PhD. During that time she earned Diplomat status with the American Board of Medicolegal Death Investigators. Upon returning to the University of Florida, Ms. Kles has worked at the C.A. Pound Human Identification Lab as a graduate ana lyst. While working at the lab she earned Member status with the Physical Anthropology section of the American Academy of Forensic Sciences. During summer semesters, Ms. Kles worked for the National Park Service at the Southeastern Archeological Center as a NAGPRA intern. She has continued her interest in NAGPRA related research by working with museum s to assist with their inventories and notices. In 2012, Ms. Kles received the John W. Griffin Award from the Florida Anthropological Council.