<%BANNER%>

The Analysis of Terrestrial Snails to Infer Hunter-Gatherer Environmental Impacts in the Middle Savannah River Valley, G...

Permanent Link: http://ufdc.ufl.edu/UFE0044877/00001

Material Information

Title: The Analysis of Terrestrial Snails to Infer Hunter-Gatherer Environmental Impacts in the Middle Savannah River Valley, Georgia and South Carolina
Physical Description: 1 online resource (170 p.)
Language: english
Creator: Snyder, Christine L
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: archaeology -- malacology -- pupillidae -- snails -- southeast -- stallings -- terrestrial
Interdisciplinary Ecology -- Dissertations, Academic -- UF
Genre: Interdisciplinary Ecology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This study investigates the impact ancient hunter-gatherers had on their environment by analyzing the changes observed in terrestrial snail assemblages recovered from archaeological sites in the middle Savannah River valley of Georgia and South Carolina.  Snails are a proxy for land alteration and temporal variations in the composition of snail assemblages provide useful data on changing land use.  If hunter-gatherer populations in this region impacted their environment through the removal or alteration of ground cover, snail species sensitive to changes in ground cover would have been affected.  Changing snail assemblages have been used as a means to interpret past anthropogenic impact among agricultural societies, but rarely among hunter-gatherers, whose impact on environments is presumed to be minimal due to mobility, small populations, and simple technology.  However, some hunter-gatherer societies achieved levels of settlement permanence and land-use intensity that rival those of food-producing societies.  Among these “complex” hunter-gatherers was the Stallings Culture of the middle Savannah region, which spanned approximately 1000 years and occurred over three time periods including the Classic Stallings (ca. 4100-3800 cal B.P.), Early Stallings (ca. 5100-4100 cal B.P.), and Late Archaic preceramic periods (ca. 5300-4100 B.P.).  The Late Archaic preceramic periods specifically include the Mill Branch and Paris Island phases.  The archaeological sites in the study area include Stallings Island (9CB1), Ed Marshall (38ED5), and Mims Point (38ED9), which collectively span the Late Archaic preceramic, the Early Stallings, and the Classic Stallings periods.  Each of the sites contained large assemblages of snails in pit and midden contexts.  For comparison, modern snail samples were collected in known environmental conditions to model snail habitat preferences in the general locations of the archaeological sites.  The modern snail samples, along with previous accounts in the literature, were used to determine taxon preferences for either open or wooded habitats.  Based on the results of the modern snail collections and observed habitat tendencies, the snails identified from the samples at Stallings Island, Ed Marshall, and Mims Point were utilized to infer past habitats and environmental impacts. The samples from Stallings Island were located in the intact remnant of a looter’s pit, which was excavated in stratigraphic sequence and dated with a series of radiocarbon assays.  All of the samples were dated to the preceramic Mill Branch phase.  A total of 11,995 terrestrial snails were identified to 19 different taxa, out of which 5 taxa were considered as having open habitat tendencies (Euconulus chersinus, Gastrocopta sp., Gastrocopta contracta, Lucilla sp., and Pupillidae) and 5 taxa were considered as having wooded habitat tendencies (Discus patulus, Glyphyalinia sp., Haplotrema concavum, Helicodiscus parallelus, and Pupisoma sp.).  The results indicated the site was mainly wooded until a disturbance event occurred, which caused a shift in the snail taxa to being dominated by more snails in the taxa with open habitat tendencies.  The samples from Ed Marshall were located in several pit features, including shallow basins and large shell-filled pits associated with fired-clay floors.  Radiocarbon dating and/or diagnostic pottery dated one feature to the preceramic Mills Branch phase, four to the Early Stallings period, and three to the Classic Stallings period.  A total of 2,454 terrestrial snails were identified to 38 different taxa, out of which 3 taxa were considered as having open habitat tendencies (Gastrocopta rupicola, Lucilla sp., and Pupillidae) and 5 taxa were considered as having wooded habitat tendencies (Discus patulus, Glyphyalinia sp., Helicodiscus parallelus, Pupisoma sp., and Zonitidae).  The results indicated a shift in habitat between the preceramic Mill Branch phase and the Early Stallings period from a more open habitat to a more wooded habitat.  The samples from Mims Point were located in several features including one burial, large storage pits, shallow basins, hearths, and occasional postholes that span a circular-village complex.  Radiocarbon dating and/or diagnostic pottery dated all of the features to the Classic Stallings period.  A total of 996 terrestrial snails were identified to 32 different taxa, out of which 3 taxa were considered as having open habitat tendencies (Gastrocopta rupicola, Lucilla sp., and Pupillidae) and 5 taxa were considered as having wooded habitat tendencies (Discus patulus, Glyphyalinia sp., Glyphyalinia wheatleyi, Practicolella lawae, and Ventridens sp.).  The results supported the circular-village complex with a central plaza.  The feature located in the presumed central plaza contained more snails within the taxa with open habitat tendencies.  The other features, which were located in the surrounding areas, among the presumed household clusters and edge of the village complex contained more snails within the taxa with wooded habitat tendencies. Due to the relatively close proximity of the three sites in this study, the samples were combined and compared over all three time periods in an attempt to show general trends in habitat over time.  The results indicate a change in habitat occurred in the middle of the preceramic Mill Branch phase which resulted in a shift from more snails in the taxa with wooded habitat tendencies to more snails in the taxa with open habitat tendencies.  Another shift occurred during the Early Stallings period, which was dominated by snails in taxa with wooded habitat tendencies.  Several fluctuations in habitat were evidence during the Classic Stallings period, but the trend was a gradual decrease in the number of snails of taxa with wooded habitat tendencies.  These results support an increase in population pressures and land clearing during the end of the preceramic Mill Branch phase, a period of limited impact during the Early Stallings period, and a return of intensive habitation and environmental impact during the Classic Stallings period. This research supports the use of terrestrial snails from archaeological sites to interpret past environmental impacts and habitat changes at a small scale (within a site, i.e.- Mims Point) and over time (over the three time periods noted within and among the sites).
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 Christine L Snyder.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Sassaman, Kenneth E.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044877:00001

Permanent Link: http://ufdc.ufl.edu/UFE0044877/00001

Material Information

Title: The Analysis of Terrestrial Snails to Infer Hunter-Gatherer Environmental Impacts in the Middle Savannah River Valley, Georgia and South Carolina
Physical Description: 1 online resource (170 p.)
Language: english
Creator: Snyder, Christine L
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: archaeology -- malacology -- pupillidae -- snails -- southeast -- stallings -- terrestrial
Interdisciplinary Ecology -- Dissertations, Academic -- UF
Genre: Interdisciplinary Ecology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This study investigates the impact ancient hunter-gatherers had on their environment by analyzing the changes observed in terrestrial snail assemblages recovered from archaeological sites in the middle Savannah River valley of Georgia and South Carolina.  Snails are a proxy for land alteration and temporal variations in the composition of snail assemblages provide useful data on changing land use.  If hunter-gatherer populations in this region impacted their environment through the removal or alteration of ground cover, snail species sensitive to changes in ground cover would have been affected.  Changing snail assemblages have been used as a means to interpret past anthropogenic impact among agricultural societies, but rarely among hunter-gatherers, whose impact on environments is presumed to be minimal due to mobility, small populations, and simple technology.  However, some hunter-gatherer societies achieved levels of settlement permanence and land-use intensity that rival those of food-producing societies.  Among these “complex” hunter-gatherers was the Stallings Culture of the middle Savannah region, which spanned approximately 1000 years and occurred over three time periods including the Classic Stallings (ca. 4100-3800 cal B.P.), Early Stallings (ca. 5100-4100 cal B.P.), and Late Archaic preceramic periods (ca. 5300-4100 B.P.).  The Late Archaic preceramic periods specifically include the Mill Branch and Paris Island phases.  The archaeological sites in the study area include Stallings Island (9CB1), Ed Marshall (38ED5), and Mims Point (38ED9), which collectively span the Late Archaic preceramic, the Early Stallings, and the Classic Stallings periods.  Each of the sites contained large assemblages of snails in pit and midden contexts.  For comparison, modern snail samples were collected in known environmental conditions to model snail habitat preferences in the general locations of the archaeological sites.  The modern snail samples, along with previous accounts in the literature, were used to determine taxon preferences for either open or wooded habitats.  Based on the results of the modern snail collections and observed habitat tendencies, the snails identified from the samples at Stallings Island, Ed Marshall, and Mims Point were utilized to infer past habitats and environmental impacts. The samples from Stallings Island were located in the intact remnant of a looter’s pit, which was excavated in stratigraphic sequence and dated with a series of radiocarbon assays.  All of the samples were dated to the preceramic Mill Branch phase.  A total of 11,995 terrestrial snails were identified to 19 different taxa, out of which 5 taxa were considered as having open habitat tendencies (Euconulus chersinus, Gastrocopta sp., Gastrocopta contracta, Lucilla sp., and Pupillidae) and 5 taxa were considered as having wooded habitat tendencies (Discus patulus, Glyphyalinia sp., Haplotrema concavum, Helicodiscus parallelus, and Pupisoma sp.).  The results indicated the site was mainly wooded until a disturbance event occurred, which caused a shift in the snail taxa to being dominated by more snails in the taxa with open habitat tendencies.  The samples from Ed Marshall were located in several pit features, including shallow basins and large shell-filled pits associated with fired-clay floors.  Radiocarbon dating and/or diagnostic pottery dated one feature to the preceramic Mills Branch phase, four to the Early Stallings period, and three to the Classic Stallings period.  A total of 2,454 terrestrial snails were identified to 38 different taxa, out of which 3 taxa were considered as having open habitat tendencies (Gastrocopta rupicola, Lucilla sp., and Pupillidae) and 5 taxa were considered as having wooded habitat tendencies (Discus patulus, Glyphyalinia sp., Helicodiscus parallelus, Pupisoma sp., and Zonitidae).  The results indicated a shift in habitat between the preceramic Mill Branch phase and the Early Stallings period from a more open habitat to a more wooded habitat.  The samples from Mims Point were located in several features including one burial, large storage pits, shallow basins, hearths, and occasional postholes that span a circular-village complex.  Radiocarbon dating and/or diagnostic pottery dated all of the features to the Classic Stallings period.  A total of 996 terrestrial snails were identified to 32 different taxa, out of which 3 taxa were considered as having open habitat tendencies (Gastrocopta rupicola, Lucilla sp., and Pupillidae) and 5 taxa were considered as having wooded habitat tendencies (Discus patulus, Glyphyalinia sp., Glyphyalinia wheatleyi, Practicolella lawae, and Ventridens sp.).  The results supported the circular-village complex with a central plaza.  The feature located in the presumed central plaza contained more snails within the taxa with open habitat tendencies.  The other features, which were located in the surrounding areas, among the presumed household clusters and edge of the village complex contained more snails within the taxa with wooded habitat tendencies. Due to the relatively close proximity of the three sites in this study, the samples were combined and compared over all three time periods in an attempt to show general trends in habitat over time.  The results indicate a change in habitat occurred in the middle of the preceramic Mill Branch phase which resulted in a shift from more snails in the taxa with wooded habitat tendencies to more snails in the taxa with open habitat tendencies.  Another shift occurred during the Early Stallings period, which was dominated by snails in taxa with wooded habitat tendencies.  Several fluctuations in habitat were evidence during the Classic Stallings period, but the trend was a gradual decrease in the number of snails of taxa with wooded habitat tendencies.  These results support an increase in population pressures and land clearing during the end of the preceramic Mill Branch phase, a period of limited impact during the Early Stallings period, and a return of intensive habitation and environmental impact during the Classic Stallings period. This research supports the use of terrestrial snails from archaeological sites to interpret past environmental impacts and habitat changes at a small scale (within a site, i.e.- Mims Point) and over time (over the three time periods noted within and among the sites).
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 Christine L Snyder.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Sassaman, Kenneth E.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044877:00001


This item has the following downloads:


Full Text

PAGE 1

1 THE ANALYSIS OF TERRESTRIAL SNAILS TO INFER HUNTER GATHERER ENVIRONMENTAL IMPACTS IN THE MIDDLE SAVANNAH RIVER VALLEY, GEORGIA AND SOUTH CAROLINA By CHRISTINE L.C. SNYDER 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 2012

PAGE 2

2 2012 Christine L.C. Snyder

PAGE 3

3 To my family

PAGE 4

4 ACKNOWLEDGMENTS I would like to sincerely thank the chair of m y committee, Kenneth Sassaman, for taking me under his wing and accepting me as his student when I had nowhere else to turn. Without his assistance, research interests, and support, my degree and research project would not have been possible. His support and guidance will always be appreciated and remembered. I would also like to thank my other committee members, Kitty Emery, Mike Daniels, and Doug Jones for their guidance and assistance during this project. Kitty Emery and Mike Daniels stayed with me d uring my entire graduate school experience and I am very thankful for their willingness to assist me throughout all of my endeavors, regardless of my path, and their constant support. I would like to extend a sincere thank you to Evan Peacock with Mississ ippi State University for his assistance with the modern snail collection and analysis used in this research. Even though we have not met face to face, his support with the project and assistance with the snail identifications was essential to the success of my dissertation. I would like to thank Peter Hallman and Evan Peacock for the collection of the modern snail samples and the initial ground work associated with this research project. I would like to thank my employer, Sheriff Donald Eslinger, the Sem inole County and my coworkers for their support when I commuted to Gainesville to complete my coursework, rearranging the schedule around my coursework and research hours, and providing an environment that supports ed ucation and research. I would like to thank my family, including my Mom and Dad, for providing the basis for my education that enabled me to pursue this degree and for their never ending love and support. I would also like to sincerely thank my husband, J ared, for giving me the

PAGE 5

5 family of which I had always dreamed and for supporting me in pursuing my goals. Lastly, I would like to thank my two children, Caine and Conley, for reminding me to fulfill my goals no matter how difficult they may seem and strive to be the best mother possible. Without family, nothing else matters.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ ........ 12 ABSTRACT ................................ ................................ ................................ ................... 15 CHAPTER 1 BACKGROUND ................................ ................................ ................................ ...... 19 2 TERRESTRIAL SNAILS IN ARCHAEOLOGY ................................ ........................ 32 3 INTRODUCTION TO TERRESTRIAL SNAILS ................................ ....................... 41 Terrestrial Snail B iology ................................ ................................ .......................... 41 Terrestrial Snail Ecology ................................ ................................ ......................... 45 Terrestrial Snail Taphonomy ................................ ................................ ................... 51 4 MODERN SAMPLES ................................ ................................ .............................. 53 Introduction ................................ ................................ ................................ ............. 53 Collection Methods ................................ ................................ ................................ 54 Field Procedures ................................ ................................ .............................. 55 Site Descriptions ................................ ................................ .............................. 56 Open habitats ................................ ................................ ............................. 57 Wooded habitats ................................ ................................ ........................ 58 Laboratory Procedures ................................ ................................ ..................... 59 Modern Snail Identifications ................................ ................................ .................... 60 Snails within Habitats ................................ ................................ .............................. 68 Snail Density ................................ ................................ ................................ ........... 69 Discussion ................................ ................................ ................................ .............. 72 5 COLLECTION METHODS FOR THE ARCHAEOLOGICAL SAMPLES ................. 94 Field Procedures ................................ ................................ ................................ ..... 94 Stallings Island ................................ ................................ ................................ 94 Ed Marshall ................................ ................................ ................................ ...... 99 Mims Point ................................ ................................ ................................ ...... 101 Laboratory Procedures ................................ ................................ ......................... 103 6 RESULTS OF THE ANALYSIS OF ARCHAEOLOGICAL SNAIL SAMPLES ....... 109

PAGE 7

7 Stallings Island ................................ ................................ ................................ ...... 109 Ed Marshall ................................ ................................ ................................ ........... 113 Mims Point ................................ ................................ ................................ ............ 116 Discussion ................................ ................................ ................................ ............ 117 Stallings Island ................................ ................................ ............................... 117 Ed Marshall ................................ ................................ ................................ .... 118 Mims Point ................................ ................................ ................................ ...... 119 7 CHANGES OVER TIME ................................ ................................ ....................... 141 Results ................................ ................................ ................................ .................. 141 Discussion ................................ ................................ ................................ ............ 144 8 CONCLUSION ................................ ................................ ................................ ...... 149 Future Research ................................ ................................ ................................ ... 151 Summary ................................ ................................ ................................ .............. 157 LIST OF REFERENCES ................................ ................................ ............................. 158 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 169

PAGE 8

8 LIST OF TABLES Table page 4 1 Location and aspect of the modern samples ................................ ...................... 74 4 2 Habitat observations for the modern samples ................................ .................... 74 4 3 Descriptions of each modern sample categorized as an open habitat ................ 75 4 4 Descriptions of each modern sample categorized as a wooded habitat ............. 75 4 5 Historical data on snail habitat among taxa from Hubricht (1985) Pilsbry (1939 19 48), Nekola and Coles (2010), and La Rocque (1970) ......................... 76 4 6 Descriptive statistics of each taxon identified in the modern samples across the samples ................................ ................................ ................................ ........ 79 4 7 List of total individual snails and percentages within open and wooded habitat samples in the modern samples ................................ ................................ ......... 80 4 8 Total number of individuals of the snails wi thin taxa in each of the modern samples ................................ ................................ ................................ .............. 81 4 9 Total number and percentages of snail taxa with open and wooded habitat tendencies in the modern samples ................................ ................................ ..... 82 4 10 Chi square results comparing the association of open and wooded taxa found within open or wooded habitats in the modern samples ........................... 82 4 11 Total number and p ercentages of snails within the families Pupillidae and Zonitidae with open and wooded habitat tendencies in the modern samples ..... 82 4 12 Chi square results comparing the association of snail taxa in the families Pupillidae and Zonitidae within open or wooded habitats in the modern samples ................................ ................................ ................................ .............. 82 4 13 Total number and percentages of snails with open and wooded habitat tendencies amon g the modern samples ................................ ............................. 83 4 14 Soil volume and total number of snails in each of the modern samples ............. 83 4 15 Snail density per liter of soil in each of the modern samples and in the taxa with open and wooded habitat tendencies ................................ .......................... 84 5 1 Radiocarbon dates for the strata in LP81 at Stallings Island ............................ 105 5 2 Features from Ed Marshall and the corresponding radiocarbon dates and periods at Ed Marshall (38ED5) ................................ ................................ ........ 105

PAGE 9

9 5 3 Radiocarbon dates for the features from Mim s Point (38ED9) ......................... 105 6 1 Total number of individual snails in each taxon within the strata of LP81 at Stallings Island ................................ ................................ ................................ 120 6 2 Descriptive statistics for each taxon within the strata of LP81 at Stallings Island ................................ ................................ ................................ ................ 121 6 3 Habitat preferences for the taxa found in LP81 at Stallings Is land ................... 121 6 4 Total number and percentages of the snails in the taxa with open and wooded habitat tendencies in each stratum in LP81 at Stallings Island ........... 122 6 5 Summary statistics of the taxa with open and wooded habitat tendencies each stratum in LP81 at Stallings Island ................................ ........................... 122 6 6 Chi square resul ts comparing the association of taxa with open and wooded habitat tendencies across the strata in LP81 at Stallings Island ....................... 122 6 7 Descriptive statistics for each taxon within Stratum III (A, C and D) in LP81 at Stallings Island ................................ ................................ ................................ 123 6 8 Descriptive statistics for each taxon within Stratum I (B, C and D) in LP81 at Stallings Island ................................ ................................ ................................ 123 6 9 Total number and percentages of the snails in the taxa with open and wooded habitat tendencies among Strata VI, IV, III and I in LP81 at Stallings Island ................................ ................................ ................................ ................ 124 6 10 Summary statistics of the taxa with open and wooded habitat tendencies in Strata VI, IV, III, and I in LP81 at Stallings Island ................................ ............. 124 6 11 Chi square results comp aring the association of taxa with open and wooded habitat tendencies in Strata VI, IV, III and I in LP81 at Stallings Island ............ 124 6 12 Chi square results comparing the association of t he taxa with open and wooded habitat tendencies in Strata III and I in LP81 at Stallings Island .......... 124 6 13 Total number of individual snails identified in each feature at Ed Marshall (38 ED5) ................................ ................................ ................................ ............ 125 6 14 Descriptive statistics for the snails identified in each taxon at Ed Marshall (38ED5) ................................ ................................ ................................ ............ 126 6 15 Habitat preferences for the taxa individual snails at Ed Marshall (38ED5) ....... 126 6 16 The total number and percentages of snails within the taxa with open and wooded habitat tendencies in each feature at Ed Marshall (38ED5) ................ 127

PAGE 10

10 6 17 Summary statistics for the taxa with open and wooded habitat tendencies in each feature at Ed Marshall (38ED5) ................................ ............................... 127 6 18 Chi square results comparing the association of the taxa with open and wooded habitat tendencies across the features at Ed Marshall (38ED5) ......... 127 6 19 The total number and percentages of snails within the taxa with open and wooded habitat tendencies in Preceramic Mill Branch Early Stalli ngs and Classic Stallings periods at Ed Marshall (38ED5) ................................ ............. 128 6 20 Summary statistics for the taxa with open and wooded habitat tendencies from the Preceramic Mill Branch Early Stallings and Classic Stallings periods at Ed Marshall (38ED5 ) ................................ ................................ ........ 128 6 21 Chi square results comparing the association of taxa with open and wooded habitat tendencies found within the Preceramic Mill Branch phase and Early Stallings period at Ed Marshall (38ED5) ................................ ........................... 128 6 22 Chi square results comparing the association of the taxa with open and wooded habitat tendencies found within Early Stallings and Classic Stalli n gs periods at Ed Marshall (38ED5) ................................ ................................ ........ 128 6 23 Chi square results comparing the association of the taxa with open and wooded habitat tendencies found in the Preceramic Mill Branch ph ase Early Stallings period and Classic Stallings period at Ed Marshall (38ED5) ............. 129 6 24 Total number of individual snails identified in each feature from Mims Point (3 8ED9) ................................ ................................ ................................ ............ 130 6 25 Descriptive statistics for each taxon individual snails at Mims Point (38ED9) .. 131 6 26 Ha bitat preferences for the taxa individual snails at Mims Point (38ED9) ........ 131 6 27 The total number and percentages of snails within the taxa with open and wooded habitat tendenci es found in each feature at Mims Point (38ED9) ....... 131 6 28 Summary statistics for the total number of snails within the taxa with open and wooded habitat tendencies in the features from M ims Point (38ED9) ....... 132 6 29 Chi square results comparing the association of the taxa with open and wooded habitat tendencies within the features from Mims Point (38ED9) ........ 132 7 1 Radiocarbon dates associated with samples from Stallings Island (LP81), Ed Marshall (38ED5), and M i ms Point (38ED9) and the cor responding periods: Preceramic Early Stallings and C lassic Stallings ................................ ............ 145

PAGE 11

11 7 2 Total number and percentages of the snails within the taxa with open and wooded habitat tendencies in the samples fr om Stallings Island (LP81), Ed Marshall (38ED5), and Mims Point (38ED9) ................................ ..................... 146 7 3 Total number and percentages of the snails within the taxa with open and wooded h abitat tendencies in the samples from the Preceramic Mills Branc h Early Stallings and Classic Stallings periods ................................ ................... 146 7 4 Chi square test results comparing the associati on of the snails within the taxa with open and wooded habitat tendencies and the three time periods: Preceramic Mills Branch Early Stallings and Classic Stallings ....................... 147

PAGE 12

12 LIST OF FIGURES Figure page 1 1 Model of Late Archaic culture histor y for the Savannah River Valley with radiocarbon assays from Stallings Island ................................ ........................... 25 1 2 T opographic map of the area of archaeologic al deposits at Stallings Island ...... 26 1 3 West profile drawing of LP81 at Stallings Island showing stratigraphic unconformity between upper and lower strata, outline of pedestal, and radiocarbon assay s on samples from intact strata ................................ .............. 27 1 4 Map of Stallings Island (9CB1) with insets showing locations of other Stallings Culture Sites in the immediate are a ................................ ..................... 28 1 5 Plan distribution of excavation unit, midden, and features at Mims Point (3 8ED9) ................................ ................................ ................................ .............. 29 1 6 Plan distribution of fea tures excavated by the Cosgrove s ................................ .. 30 1 7 Plan distribution of features excavated by the Cosgroves ................................ .. 31 4 1 Google Earth image showing the collection sites for 16 of the 20 modern samples (5 to 20) ................................ ................................ ................................ 85 4 2 Contour map of the Savannah River bank to the northeast of Stallings Island showing the locations of modern samples 14, 15, 16, 17, 19 and 20 ................. 86 4 3 Google Earth image of Savannah River bank to northeast of Stallings Island showing the locations of modern samples 14, 15, 16, 17, 19 and 20 ................. 87 4 4 Contour map of Stallings Island s ite showing the locations of modern samples 7, 8, 9, 10, 11 and 12 ................................ ................................ ........... 88 4 5 Google Earth image of Stallings Island site showing the locations of modern samples 7, 8, 9, 10, 11 and 12 ................................ ................................ ........... 89 4 6 Contour map of Savannah River bank to southeast of Stallings Island showing the locations of modern samples 5 and 6 ................................ ............. 90 4 7 Google Earth image of Savannah River bank to southeast of Stallings Island showing the locations of modern samples 5 and 6 ................................ ............. 91 4 8 Average soil volume among the moder n samples within the open and wooded habitat samples ................................ ................................ ..................... 92 4 9 The average total snail density per liter of soil within the open and wooded habitat samples ................................ ................................ ................................ .. 92

PAGE 13

13 4 10 The average snail density per liter of soil between snails with open or wooded habitat tendencies within the open and wooded habitat samples ......... 93 5 1 West profile of LP81 showing the strata and corresponding radiocarbon date s at Stallings Island ................................ ................................ .................... 106 5 2 Plan distribution of excavation unit, midden, and features at Mims Point (38ED9) ................................ ................................ ................................ ............ 107 5 3 Plan distrib ution showing the pit features excavated at Mims Point (38ED9), which specifies the location of Features 1, 50, 51, 52 and 66 .......................... 108 6 1 Percentages of the snails within the taxa with open and wooded habitat tendencies across the strata within LP81 at Stallings Island ............................ 132 6 2 Percentages of the snails within the taxa with open and wooded habitat tende ncies across the strata (excluding Stratum IIIA) within LP81 at Stallings Island ................................ ................................ ................................ ................ 133 6 3 Means and standard deviations of the total number of snails in the taxa with open and wooded h abitat tendencies across the strata within LP81 at Stallings Island ................................ ................................ ................................ 133 6 4 Percentages of the snails within the taxa with open and wooded habitat tendencies across Strata VI, IV, III and I within LP81 at Stallings Island ......... 134 6 5 Percentages of the snails within the taxa with open and wooded habitat tendencies between Strata III and I within LP81 at Stallings Isl and .................. 134 6 6 Means and standard deviations of the total number of snails in the taxa with open and wooded habitat tendencies among the four main strata within LP81 at Stallings Island ................................ ................................ ............................. 135 6 7 Percentages of the snails within the taxa with open and wooded habitat tendencies among Feature 30, which dates to the preceramic Mill Branch phase at Ed Marshall (38ED5) ................................ ................................ .......... 135 6 8 Percentages of the snails within the taxa with open and wooded habitat tendencies among the features dating to the Early Stallings period at Ed Marshall (38ED5) ................................ ................................ .............................. 136 6 9 Percentages of the snails within the taxa with open and wooded habitat tendencies among the features dating to the Classic Stallings period at Ed Marshall (38ED5) ................................ ................................ .............................. 136 6 10 Means and standard deviations for the total number of snails within the taxa with open and wooded habitat tendencies among the features from Ed Marshall (38ED5) ................................ ................................ .............................. 137

PAGE 14

14 6 11 Percentages of the snails within the taxa with open and wooded habitat tendencies during the Preceramic Mill Branch Early Stallings and Classic Stallings periods at Ed Marshall (38ED5) ................................ ......................... 137 6 12 Means and standard deviations for the total number of snails within the taxa with open and wooded habitat tendencies among the Pre ceramic Mill Branch Early Stallings and Classic Stallings peri ods at Ed Marshall (38ED5) 138 6 13 Percentages of the snails within the taxa with open and wooded habitat tendencies among the features at Mims Point (38ED9) ................................ ... 138 6 1 4 Means and standard deviations of the total number of snails within the taxa with open and wooded habitat tendencies among the features from Mims Point (38ED9) ................................ ................................ ................................ ... 139 6 15 Plan distribution of features at Mims Point (38ED9), showing projected pit feature clusters of a circular domestic compound with a central area largely devoid of features ................................ ................................ ............................. 140 7 1 Percentages of snails within the taxa with open and wooded habitat tendencies in chronological order across the samples from Stallings Island (LP81), Ed Marshall (38ED5), and Mims Point (38ED9) ................................ ... 147 7 2 Percentages of snails within the taxa with open and wooded habitat tendencies in chronological order in the samples from the three time periods: Preceramic Mills Branch, Early Stalli ngs, and Classic Stallings periods ........... 148

PAGE 15

15 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE ANALYSIS OF TERRESTRIAL SNAILS TO INFER HUNTER GATHERER ENVIRONMENTAL IMPACTS IN THE MIDDLE SAVANNAH RIVER VALLEY, GEORGIA AND SOUTH CAROLINA By Christine L.C. Snyder December 2012 Chair: Kenneth Sassaman Major: Interdisciplinary Ecology This study investigates the impact ancient hunter gatherers had on the ir environment by analyzing the changes observed in terrestrial snail assemblages recovered from archaeological sites in the middle Savannah River valley of Georgia and South Carolina. Snails are a proxy for land alteration and temporal variations in the composition of snail assemblages provide useful data on changing land use. If hunter gatherer populations in this region impacted their environment through the removal or alteration of ground cov er, snail species sensitive to changes in ground cover would have been affected. Changing snail assemblages have been used as a means to interpret past anthropogenic impact among agricultural societies, but rarely among hunter gatherers, whose impact on e nvironments is presumed to be minimal due to mobility, small populations, and simple technology. However, some hunter gatherer societies achieved levels of settlement permanence and land use intensity that rival those of food producing societies. Among t gatherers was the Stallings Cultur e of the middle Savannah region, which spanned approximately 1000 years and occurred over three time periods including the Classic Stallings (ca. 4100

PAGE 16

16 3800 cal B.P.) Early Stallings (ca. 5100 4100 ca l B.P.) and Late Archaic preceramic periods (ca. 5300 4100 B.P.) The Late Archaic preceramic periods specifically include the Mill Branch and Paris Island phases. The a rchaeological sites in the study area include Stallings Island ( 9CB1 ), Ed Marshall (38ED5), and Mims Point (38ED9), which collectively span the Late Archaic preceramic, the Early Stallings and the Classic Stallings period s Each of the sites contain ed large assemblages of snails in pit and midden contexts. For comparison, modern snail samples were collected in known environmental conditions to model snail habitat preferences in the general locations of the archaeological sites The modern snail samples, along with previous accounts in the literature were used to determine taxon prefe rences for either open or wooded habitats. Based on the results of the modern snail collections and observed habitat tendencies, the snails identified from the samples at Stallings Island, Ed Marshall, and Mims Point were utilized to infer past habitats a nd environmental impacts. The samples from Stallings Island were located in the intact remnant of a l p it, which w as excavated in stratigraphic sequence and dated with a series of radiocarbon assays All of the samples were dated to the p receramic Mill Branch phase. A total of 11, 995 terrestrial snails were identified to 19 different taxa out of which 5 taxa were considered as having open habitat tendencies ( Euconulus chersinus, Gastrocopta sp., Gastrocopta contracta, Lucilla sp., and Pupillidae) and 5 taxa were considered as having wooded habitat tendencies ( Discus patulus, Glyphyalinia sp., Haplotrema concavum, Helicodiscus parallelus, and Pupisoma sp.) The results indicated the site was mainly wooded until a disturbance event occurred, which c aused

PAGE 17

17 a shift in the snail taxa to being dominated by more snails in the taxa with open habitat tendencies. The samples from Ed Marshall were located in several pit features, including shallow basins and large shell filled pits associated with fired clay floors. Radiocarbon dating and/or diagnostic pottery dated one feature to the preceramic Mills Branch phase, four to the Early Stallings period, and three to the Classic Stallings period. A total of 2,454 terr estrial snails were identified to 38 differe nt taxa, out of which 3 taxa were considered as having open habitat tendencies ( Gastrocopta rupicola, Lucilla sp., and Pupillidae) and 5 taxa were considered as having wooded habitat tendencies ( Discus patulus, Glyphyalinia sp., Helicodiscus parallelus, Pu pisoma sp., and Zonitidae) The results indicated a shift in habitat between the preceramic Mill Branch phase and the Early Stallings period from a more open habitat to a more wooded habitat. The samples from Mims Point were located in several features including one burial, large storage pits, shallow basins, hearths, and occasional postholes that span a circular village complex. Radiocarbon dating and/or diagnostic pottery dated all of the features to the Classic Stallings period. A total of 996 terre strial snails were identified to 32 different taxa, out of which 3 taxa were considered as having open habitat tendencies ( Gastrocopta rupicola, Lucilla sp., and Pupillidae) and 5 taxa were considered as having wooded habitat tendencies ( Discus patulus, Gl yphyalinia sp., Glyphyalinia wheatleyi, Practicolella lawae, and Ventridens sp.) The results supported the circular village complex with a central plaza The feature located in the presumed central plaza contained more snails within the taxa with open h abitat tendencies. T he

PAGE 18

18 other features, which were located in the surrounding areas, among the presumed household clusters and edge of the village complex contained more snails within the taxa with wooded habitat tendencies. Due to the relatively close pro ximity of the three sites in this study, the samples were combined and compared over all three time periods in an attempt to show general trends in habitat over time. The results indicate a change in habitat occurred in the middle of the preceramic Mill B ranch phase which resulted in a shift from more snails in the taxa with wooded habitat tendencies to more snails in the taxa with open habitat tendencies. Another shift occurred during the Early Stallings period, which was dominated by snails in taxa with wooded habitat tendencies. Several fluctuations in habitat were evidence during the Classic Stallings period, but the trend was a gradual decrease in the number of snails of taxa with wooded habitat tendencies. These results support an increase in popul ation pressures and land clearing during the end of the preceramic Mill Branch phase, a period of limited impact during the Early Stallings period, and a return of intensive habitation and environmental impact during the Classic Stallings period. This rese arch supports the use of terrestrial snails from archaeological sites to interpret past environmental impacts and habitat changes at a small scale (within a site, i.e Mims Point) and over time (over the three time periods noted within and among the sites ).

PAGE 19

19 CHAPTER 1 BACKGROUND The Stallings C ulture wa premier Late Archaic forager societies (Sassaman 1993a) dating from ca. 5000 3500 cal B.P. and distributed throughout the Savannah River valley and portions of the coasts of Georgia and South Carolina (Figure 1 1 ) (Sassaman 2006; Sassaman et al. 2006). Stallings Culture was a hunter gatherer population whose members collected and ate shellfish and discarded the inedible remains in middens, mounds, and pits. The Stallings Culture is best known for its namesake pottery series (Stallings fiber tempered pottery), which is among the oldest pottery in North America (Sassaman et al. 2006) It is also well known because of its general trend towards a more sedentary lifestyle, as indicated by smaller territories and more permanent villages in the later periods of the culture which marked the beginning of less nomadic lifestyle s The history of Stallings Culture extends well beyond the appearance of pottery to include ancestr al and affiliat ed societies in the region dating back to at least 5800 cal B.P. Two of the key ancestral phases in the vicinity of Stallings Island are the Paris Island phase of ca. 5300 4700 cal B.P., and the Mill Branch phase of 4700 4100 cal B.P. The Stallings Cultu re itself is divided into two phases: Early Stallings (ca. 5100 4100 cal B.P.), and Classic Stallings (ca. 4100 3800 cal B.P.). It was during the Classic Stallings phase that communities of the middle Savannah River valley achieved a level of settlement permanence and formality paralleling that of sedentary farmers. A central location of settlement for Classic Stallings Culture was Stallings Island (9CB1), a seven hectare (16 acre) island in the Savannah River, about eight miles upstream from the city of Augusta, Georgia Stallings Island was first researched by

PAGE 20

20 Charles C. Jones, a nineteenth century antiquarian, in 1861 (Sassaman 2006 ). Jones was interest ed in the human burials on Stallings Island and speculated that the site was a huge necropolis and a de dicated cemetery (Jones 1861). Jones (1861) noted items of domestic refuse throughout the midden and pit fill of burials, but provided little detail on the types of evidence he uncovered ( Jones 1861). After Jones, William Claflin conducted several exp editions from 1908 1925 and collected a large number of artifacts and several human skeletons (Sassaman 2006; Sassaman et al. 2006). Claflin (1931) suggested that habitation occurred on the island, but he did not record a sufficient body of evidence to su p port his conclusions In 1929, Charles and Harriet Cosgrove conducted stratigraphic excavations at Stallings I sland which included the use of a grid (Claflin 1931). The Cosgroves were the first to excavate features other than burials and concluded tha t the site was inhabited by one cultural group and they also noted remains of earlier people below the deposits of the main tribe (Sassaman 2006 ) (Figure 1 2) Stallings Island was investigated repeatedly since the Cosgroves expedition in the mid 20 th cen tury. Investigations include those of Charles H. Fairbanks in 1942, H. Bruce Greene in 1961, and Donald Crusoe and Chester DePratter in 1976 (Sassaman 2006; Sassaman et al. 2006). Stallings Island was purchased by the Archaeological Conservancy in 1998 a fter which detailed archaeological research was conducted by the University of Florida (UF), which included the collection of the samples for this research project. The UF expedition of 1999 map ped and test ed Stallings Island in an attempt to reconstruct the dating (Sassaman et al. 2006) (Figure 1 3). Fifty six intact pit features and a 3 meter

PAGE 21

21 deep sequence of stratified shell midden with good preservation were excavated at the site. In total, o ver 4.25 cubic meters of 1/8 inch water scr eened feature and midden matrix and nearly 100 flotation samples of at least 10 liters each were collected for analysis (Sassaman 2 001 ). Besides Stallings I sland, three other sites were i nvestigated in the m iddle Savannah River v alley by Sassaman and colleagues in the 1990s : Ed Marshall, Mims Point, and Victor Mills (Figure 1 4 ). Ed Marshall (38ED5) is located 0.5 km east of Stallings Island and 1.6 km downs tream from Mims Point between a levee and backwater slough of the Savannah River and is a multicomponent habitation site. The site was excavated three times between 1995 and 1998, which resulted in the collection of 98 m 2 of material and 67 pit features. Horizontal separation of compo nents allowed the clear delineation of Mill Branch, Early Stallings and Classic Stallings features. Domesticate architecture was observed in the Mill Branch and Early Stallings components and site wide patterning indicated a circular village plaza complex approximately 30 m in diameter (Sassaman 2001). Ed Marshall is a lower elevation riverine site, which is prone to flooding and considered a floodplain Mims Point (38ED9) is the most extensive excavation of a Stallings Culture village (Sassaman 1993b). Mims Point is located on a ridge nose at the confluence of Stevens Creek and the Savannah River about 1.2 km north of Stallings Island (Sassaman et al. 2006) It contained the remains of a 0.2 ha habitation complex which dat ed to a few decades of Classi c Stallings times ( ca. 3950 3900 cal B.P.) (Sassaman n.d.) Three field seasons resulted in the hand excavation of 364 m 2 which exposed several household clusters that represented a complete and well preserved circular village

PAGE 22

22 plaza complex dated to the Classic Stallings phase. This site is considered upland and is not prone to flooding. Victor Mills (9CB138 ) is located on a sloping ridge overlooking the Savannah River and is located approximately 1.25 km south of Stallings Island This is a small site that is dated to ca. 4450 cal B.P. and is the earliest known occupation of the middle Savannah by Early Stallings groups. Two field expeditions in 1994 resulted in 63 m 2 o f hand excavation This site is believed to have been one of the first incursions o f Coastal Plain Stallings groups into the middle Savannah when indigenous Mill Branch groups were seasonally using riverine sites, such as Stallings Island and Ed Marshall (Sassaman 2001). Each of these four sites has different microenvironments, including the slope, aspect and soil, which affected the type and proportion of local food resources (Sassaman 2001). The four sites are considered well preserved Late Archaic remains and span the Paris Island, Mill Branch, Early Stallings and Classic Stallings ph ases. The Paris Island phase dates from ca. 5350 4700 cal B.P. (Figure 1 1 ). Most known Paris Island occupations were seasonal encampments and lacked larger and more permanent settlements (Sassaman 2006). The Paris Island phase was followed by the Mill Branch phase of ca. 4700 4200 cal B.P. another preceramic culture (Figure 1 1 ). Mill Branch occupations were noted o n Ed Marshall and on Stallings Island (Sassaman 2006). The Mill Branch phase component at Stallings Island was dated to about ca. 4700 46 00 cal B.P (Figure 1 1). The Early Stallings phase was an early pottery phase, which occurred during the Paris Island, Early Stallings, and Classic Stallings phases and began in the Coastal

PAGE 23

23 Plain around ca. 5100 cal B.P. (Figure 1 1 ). During the initial first centuries of Early Stallings occupation in the Middle Savannah, Stallings Island was underutilized or abandoned and only transient use of Stallings Island was found from ca. 4450 4200 cal B.P. (Sassaman et al. 2006). A single component deposit of Ea rly Stallings phase was located at Victor Mills, which was dated to ca. 4450 cal B.P. and another component was located at Ed Marshall, which was dated to ca. 4200 cal B.P. (Sassaman et al. 2006). A program of recent analysis with materials recovered from the four middle Savannah sites involved vertebrate fauna, freshwater mussels, human skeletal remains, plant remains, and subsistence technological innovations, including pottery. In addition, spatial data from old and recent excavations have been used to infer patterns of community organization (Sassaman et al. 2006). Evidence shows that the Classic Stallings occupations at Stallings Island, Mims Point, and perhaps Ed Marshall entailed circular arrangements of 6 15 m in diameter (Figures 1 5, 1 6 and 1 7). At Stallings Island the central plaza held many of the Classic Stallings burials, a pattern that appears to be unique to this central site. Compared to settlement s of the Paris Island, Mill Branch, and Early S tallings phases, those of the Classic Stallings phase were apparently permanent and thus the most likely to entail environmental impacts attending fixed, intensive land use. However, other evidence from Stallings Island suggest that shell fishing was most intensive during the earlier occupations, a practice that may have had a measurable effect of mollusk populations possible impacts, the multi century abandonment at Stalling s Island following the Mill

PAGE 24

24 Branch occupation may have been sufficient to return the locale to its pre occupation state. Terrestrial snails in the assemblages from Stallings Island and related sites provide an independent means to monitor changes in veget ative cover. This study will determine if the archaeological sites were open habitats potentially created by cleared areas or plazas due to deforestation, increased population or consistent trampling, or forested habitats through natural or reforestation p rocesses. The hypotheses are: 1) overall, an increase in open habitat species will be found through time as the intensity of the settlement increased; 2) variations in proportions of open habitat species will co vary with periods of occupation and abandon ment; and 3) open habitat species will tend to occur with greatest frequency in central plazas of permanent villages within the Classic Stallings phase

PAGE 25

25 Figure 1 radiocarbo n assays from Stallings Island ( + 2006)

PAGE 26

26 Figure 1 showing pylons, locations of excavation units of most prior expeditions, and 1999 units of 2006)

PAGE 27

27 Figure 1 upper and lower strata, outline of pedestal, and radiocarbon assays on from Sassaman et al. 2006)

PAGE 28

28 Figure 1 4 Stallings Culture Sites in the immediate area and the greater distribution of shell )

PAGE 29

29 Figure 1 (38ED9), showing projected pit feature clusters of a circular domestic al. 2006)

PAGE 30

30 Figu re 1 feature types as published in Claflin 1931:Plate 7; and (b) showing only those

PAGE 31

31 Figure 1 ibution of features excavated by the Cosgroves (a) showing Mims Point circular village model superimposed over projections of pit features by depth into clay; and (b) showing human interments coded for sex (after Claflin 2006)

PAGE 32

32 CHAPTER 2 TERRESTRIAL SNAILS I N ARCHAEOLOGY Humans are known to have affected their surrounding environment to the extent that they and Nowacki 20 08). The activity of humans clearing land and creating more forest edges is believed to have change d the proportions of disturbance based taxa (Abrams and Nowacki 2008). Land snails have been used in the analysis of modern sites to determine current envi ronmental impacts of agriculture, including fertilizers, grazing and pastures, in locations around the world due to their sensitivity to environmental changes (Suominen 1999; Greenwood and McKenzie 2001; Labaune and Magnin 2001; Labaune and Magnin 2002; Ol iver et al. 2005; Hof and Bright 2010). Land snails cannot only be used for modern environmental studies ; they can also be used in studies involving environmental archaeology to provide robust contexts to understand past human behavior (Peacock and Gerber 2008). Terrestrial mollusks are believed to offer keys to understanding the organization of past biotas and to reconstruct the organization of past communities (Kohn and Arua 1999). Cowie and Grant Mackie (2004) stated that land snail faunal composition from archaeological excavations can indicate habitat or microhabitat preferences and environmental change. Ecology is often concerned with how environmental factors affect an organism, but the organisms can also be used as indicators of the environment (p ast or present). Matteson (1960) considers this view as applied ecology rather than ecology itself. Environmental archaeology allows the use of different types of data from different disciplines to research questions regarding long term human/nature int eractions (Peacock and Melsheimer 2003). Using gastropod remains

PAGE 33

33 is common where environmental archaeology is well established and is often used in Europe, Mesoamerica, and Polynesia (Peacock and Melsheimer 2003). Land snails are used to establish overal l environmental conditions because they are essentially site specific indicators that were rarely eaten or exploited and were not generally affected by cultural biases (Peacock and Melsheimer 200 3; Peacock et al. 2005 ). Land snails have also been found to have a great variety of regional patterns in species richness and differentiation (Stanisic et al. 2007) They are often used to determine the nature and magnitude of prehistoric human environmental disturbances/impacts and to explore past en vironmental conditions (Peacock and Melsheimer 200 3; Peacock et al. 2005 ). Bobrowsky (1984) noted four reasons why gastropods may occ ur in archaeological deposits: 1) the snails were used as a food source; 2) the snails represent natural popula tions once living at the site; 3) the snails were accidentally i ntroduced into the assemblage; 4) the snail shells served as a cultural purpose rather than diet. Most agree the gastropod assemblage should be considered a natural occurrence (Bobrowsky 1984). Some believe that la nd snails represent one of the best preservation classes known among living animals and have fairly stable and accurate death assemblages (Bobrowsky 1984). Few animals can withstand the ravages of time better than mollusk shells (Matteson 1960). Meyrick and Karrow (2007) state four reasons why land snails are better paleoecological indicators than pollen: 1) the shells can survive in a range of conditions, including oxidized deposits that rarely preserve other fossil groups such as pollen ; 2) land snail a ssemblages provide information about the local paleoenvironments, such as shade and moisture conditions, at the time of deposition ; 3) most snail fauna s can be used to detect small scale disturbance s

PAGE 34

34 resulting from anthropogenic activity ; and 4) most snail s hells can be identified to species level. M ollusk shells are often utilized in archaeological investigations (Evans 1972) because they are known to re tain the original color patterns and appear as if they had died more recently (Matteson 1959). T hey a nd are also commonly used in Europe in conjunction with other discipline s such as sedimentology, vertebrate paleontology, archaeology and palynology (Meyrick and Karrow 2007). The first known attempt to use gastropods for environmental reconstruction in th e United States occurred on Goose Island, Massachusetts in 1867 when Morse observed terrestrial snails and stated that they can exist in only hardwood growths (Bobrowsky 1984 ; Martin 2000 ). Martin (2000) reported that Morse excavated a shell midden and fo und land snail s hells approximately 0.6 m down The shells were identified as Glyphyalin i a indentat a Paravitrea multidentata and Appalachina sayan which are common in hardwood forests and rare in Maine. Based on this information, Morse concluded there were changes in vegetation (hardwoods to a spruce growth ) and resident molluscan species. Casual notations of land snails at archaeological sites became common by the late nineteenth century and early twentieth century (Bobrowsky 1984). The first comment about gastropods being used as a food source occurred in 1930 by Baker (Bobrowsky 1984). Gastropods became relatively important in archaeology during the late 1930s and early 1940s and it was often noted when snails were not found (Bobrowsky 1984). Gastr opods continued to be mentioned in archaeological sites during the 1940s and early 1950s (Bobrowsky 1984). The debate about whether or not snails were used as a food source be came common in the 1950s and continues today (Bobrowsky 1984). Morrison (1942) studied faunal changes in

PAGE 35

35 prehistoric Indian middens over time in the Pickwick Basin of the Tennessee River in northwest Alabama. Morrison concluded that the land snail fauna w as representative of those living in the flood plain or river bottom forest and indicated a partial clearing of the surrounding forest, which implied an increase in agriculture and a more continuous habitation of the site. reports appeared in the archaeological literature distribution list and ecological synthesis in 1969 ( Baerreis 1969; Bobrowsky 1984). The larger archaeological community began to take note of the use of gastropods in archaeology in the 1960s mainly due to Baerrei from 1969 to 19 80 ( Baerreis 1969; 1971; 1973; 1974; 1980 ). Baerreis (1973) reviewed research in (1942) research in Alabama and pointed out that Morri upon the assumption that the gastropods lived at the site and were fossil assemblages rather than considering that they could have been drift shells and/or comprise a thanatocoenose (assemblage brought together after death) Baerreis also reviewed of the Mississippi River in southern Illinois. Fowler concluded a change in environmental conditions based on the presence of a land snail, Allogona profunda which was found in the basal zones but not in the top of the deposit and was only found in cool, damp wooded habitats Fowler also noted a shift in proportion from a high proportion of terrestrial snails to more aquatic snails and back to a high prop ortion of terrestrial snails. Fowler interpret ed the shift as an increased consumption of gastropods for food during the middle period rather than considering a shift in habitat

PAGE 36

36 due to flooding in the valley. Baerreis concluded that gastrop ods are reflections of environmental conditions when the site was occupied; however, the sampling techniques must be sufficient in order to interpret the results correctly. There are a limited number of studies using terrestrial snails to interpret the imp act of hunter gatherers on their environment. Two studies include Klippel and Turner (1991) and Harms (2008). However, multiple studies have used terre strial snails to interpret the p aleoenvironmental history of archaeological sites, which include Barber (1988), Peacock and Melsheimer (2003), Peacock et al. (2005), Meyrick and Karrow (2007) and Peacock and Gerber ( 2008). Other research has used land snails as an indicator of environmental change due to the impact of human colonization on secluded island s in the Pacific. Some of these include research in: the Pacific Islands (Solem (Christensen and Kirch 1981), the Cook Islands (Allen and Schubel 1990 ; Kirch et al. 1991 ; Allen 1992, 1994, 1998 ), Mangaia Island in East Polynesia (Kirch and Ellison 1994), the Pitcairn Islands (Preece 1998), New Caledonia (Cowie and Grant Mackie 2004) and Easter Island (Kirch et al. 2009). The Pacific Islands have been well studied due to th e high rate of extinctions and replacement of the natural biotas by exotic species since the beginning of Europ ean colonization (Cowie 2001a, 2001b; Cowie and Robinson 2003). Klippel and Turner (1991) researched gastropods from an Archaic period shell midd en in middle Tennessee. Based on a change in land snail density and species distributions, two possibilities were considered: 1) the hunter gatherers may have modified the local environment of the shell midden to the extent that it was transformed

PAGE 37

37 into a habitat suitable for land snails that are found in open habitats today or 2) climatic conditions may have supported a more open habitat during the late Middle Archaic Period rather than the mesophytic forest habitat presently in that area. Klippel and Tur ner (1991) reject the idea that hunter gatherers impacted their environment or cleared their camp sites of trees due to their mobile lifestyles and reject the idea that open they had. They lean more towards a climatic change that has been previously suggested between 8000 and 5000 years B.P. based on pollen and vertebrate faunal data from the mid south. Harms (2008) used terrestrial and aquatic mollusk s along with carbon dat ing from archaeological sites in Wisconsin dated to the Archaic period to show environmental change during approximate dates and concluded that mollusks could be used to indicate regional environmental change, specifically from warm and moist to cooler and drier conditions. Harms (2008) found an extreme decrease in snail quantity, loss of species diversity and a decrease in snail density, which indicated a change to a harsher climate for those mollusks. Peacock and Melsheimer (2003), Peacock et al. (2005 ) and Peacock and G erber (2008) compared land snail taxa from protohistoric, historic and modern assemblages from the Black Belt Province of Mississippi and found a shift in the dominant taxa, which m ay be attributed to climate change or human induced ero sion and soil loss that caused a drop in the soil moisture. Land snails and freshwater mussels were used to interpret human transformation of the landscape in n ortheastern Mississippi. Land snails and other artifacts were also used to interpret site forma tion processes in Mississippi and

PAGE 38

38 found land snails helpful when determining pit fill episodes. Based on the land snail assemblages from the archaeological sites and the analysis of the modern land snail ecological habitats, they were able to conclude tha t the landscape had transformed from hardwood to cedar dominated mosaic due to land clearance and erosion over time. (2003) research is similar to this study because they analyzed the presence of terrestrial snail taxa among sites, concentrating on those that were present in open habitat sites compared to wooded habitat sites. The variation in snail taxa was compared to the historical habitat ranges in order to interpret environmental conditions. Peacock and Melsheimer (20 03) found the snail taxa changed with a shift from mesic to xeric conditions due to climate change and/or human induced erosion and soil loss. Evans (1972) also stated that archaeological snail faunas can be used to estimate their former abundance, which can be compared to their nearest present day analogue in order to surmise the past environment in which they lived. Barber (1988) was the first modern archaeologist to use land snails to interpret the environment in the northeastern U.S. Barber (1988) p romoted the use of land snail assemblages to interpret macroclimate and microenvironments using a variety of methods including the species spectrum approach (Evans 1972), environmental tolerance approach (Barber 1979; Barber 1980) and morph analysis (Baerr eis 1974 1980). Meyrick and Karrow (2007) also used molluscan data to reconstruct the paleoenvironmental history of southern Ontario and found a progression of species variation from exposed open habitats to closed woodland habitats.

PAGE 39

39 Solem (1964) revie wed terrestrial snails from the Pacific Islands in the 19 th century and found that endemic snails have been replaced by exotics and the majority of endemic snails are restricted to small patches of native forests, whereas the exotic snails are found in the cultivated and ecologically altered areas. Christensen and Kirch (1981) studied archaeological terrestrial snails on Tikopia, southeastern Solomon Islands and found a shift from endemic to exotic land snails that corresponded to the colonization of human s and environmental impacts that occurred on the vegetation due to land clearing for agricultural purposes. Kirch et al. (1991) studied human environmental impacts, such as deforestation and cultural stress from overpopulation and agricultural purposes, o n Cook Island by analyzing archaeological vertebrates, invertebrates (including aquatic mollusks) and vegetation. The study found relationships between the archaeological samples and the chronology of human habitation and environmental impa cts on the isla nd. Allen (1992, 1994 1998) and Allen and Schubel (1990) used an environmental archaeological approach to interpret and date human settlement on the Cook Islands in the Pacific. Land snail assemblage s were reviewed and found to include taxa associated w ith human habitation and gardens, which agreed with wood charcoal identifications and suggested an early settlement time frame for one site within the study area. The snail fauna excavated in New Caledonia showed a change in faunal composition, including a decrease or elimination of native species and the introduction of exotic species, which indicated a change in habitat and vegetation due to human intrusion into the area since European colonization (Cowie and Grant Mackie 2004). Kirch et al. (2009) stud ied terrestrial snails on Easter Island and found a shift from endemic molluscan fauna to synanthropic

PAGE 40

40 and anthropochoric (human transported) fauna over time, which may be due to human induced habitat alteration including deforestation, vegetation clearing increased erosion and the influx of human related predators, all of which began when the island was initially colonized by prehistoric peoples and they began to use the land for agricultural purposes. These studies indicate that terrestrial snails can b e used successfully to interpret anthropogenic impact on the environment and to reconstruct the paleoenvironmental history of archaeological sites, including those of hunter gatherers.

PAGE 41

41 CHAPTER 3 INTRODUCTION TO TERRESTRIAL SNAIL S Terrestrial Snail Biology Soft bodied land snails and slugs belong to the phylum Mollusca and the class Gastropoda. The Latin word molluscus means soft and the class name, Gastropoda, originates from the Greek words gaster (belly) and podos (foot) (Martin 2000). Gastropods are the most diverse group in the phylum Mollusca and have been prospering since their appearance in the middle Cambrian (Tojo and Ohno 1999). In terrestrial ecosystems, the number of terrestrial mollusk species is only outnumbered by arthropods (Alvarez and Willig 1993; Secrest et al. 1996). Terrestrial gastropods are taxonomically diverse and well described (Nekola 2005). Over 35,000 species of Gastropoda exist globally, including 3600 in Europe, 1600 in Australia, 1400 in New Zealand and 1200 in North Am erica (Nekola 2005; Pigati et al. 2010). There are at least 70,000 extant species found in terrestrial, marine and freshwater habitats (Pigati et al. 2010). Land snails are ubiquitous in the eastern United States and include both shelled animals and slug s, which lack shells (Hotopp 2002). There are at least 500 native species of snails in the eastern United States (Hotopp 2002; Hotopp et al. 2010). The land snails are usually ubiquitous and abundant and play important ecosystem roles, which make them g ood indicators of site conditions (Hotopp et al. 2010). Land snail adult sizes range from 1mm to 208mm (Vagvolgyi 1975). Approximately 50% of all land snail species and roughly 90% of all individual land snails are less than 5mm in maximum dimension (Ne kola 2009). Most of these smaller snails are in the family Pupillidae ( Vagvolgyi 1975 ; Nekola 2009). Almost 10% of the entire United States and Canadian fauna (approximately 125 species) are in the family

PAGE 42

42 Pupillidae and within the genera Bothriopupa Cha enaxis Columella Gastrocopta Pupilla Pupisoma Pupoides Sterkia and Vertigo (Nekola 2009; Nekola and Coles 2010). The Pupillidae represent approximately 12% of the total fauna east of the continental divide and are the third most diverse family; the two most diverse are Polygyridae and Zonitidae, who respectively make up 30% and 22% of the total fauna. (Nekola 2009; Nekola and Coles 2010). Pupillidae represent 25 75% of all species and individuals at a regional scale and at least 30% of the species d iversity and 33% of all individuals within a site (Nekola and Coles 2010). Pupillids may also represent 80 100% of total molluscan diversity within sites in some regions (Nekola and Coles 2010). Most terrestrial snails feed on plant matter, including li ve vegetation, rotting leaves, rotting wood, sap, fungi and some even feed upon animal scats, animal carcasses, other snails, nematodes or other invertebrates ( Boycott 1934 ; Evans 1972; Hotopp 2002 ). Many snails also ingest small soil particles and will r asp larger rocks or shells of dead mollusks to ingest the calcium carbonate essential for reproduction, shell development and other physiological needs ( Martin 2000 ; Hotopp 2002 ). During times of demand, such as egg laying, snails will mobilize calcium fr om their own internal organs and shells (Hotopp 2002). The importance of available calcium to snails leads to them having a preference for the tree species composition of the habitat in which they prosper because tree species composition influences the ca lcium content of the upper soil horizons by the leaf litter (Hotopp 2002). The nutrient concentrations in the upper 2002). Therefore, some believe that some land snails can live in areas with calcium poor soils if the local trees concentrate calcium in their leaves because the snails can

PAGE 43

43 either absorb the dissolved calcium carbonate through their moist skin or drink it from the upper layers of the soil (Martin 200 0). Snail species have been associated with soil calcium, magnesium and acidity (pH) (Hotopp 2002). The most important soil parameters are considered extractable calcium, water soluble calcium (includes calcium citrate), pH, moisture and vegetation (Ware born 1969; Martin 2000 ; Hotopp 2002 ). These soil parameters may influence snail distribution and abundance (Wareborn 1969; Martin 2000). Land snails have many natural enemies. Snails may be consumed by a variety of predators, such as the larvae of lamp yrid beetles, adult ground beetles, ants, salamanders, shrews, birds, millipedes, frogs, turtles, snakes, lizards, moles, rats, mice, chipmunks and squirrels ( Martin 2000 ; Hotopp 2002). Most species have an annual life cycle (9 15 months), but some of the larger snails may live two or three years ( Boycott 1934 ; Evans 1972). Others consider their lifespan to be 1 10 years (Martin 2000). Tojo and Ohno (1999) found that shells grow incrementally and form an alternating sequence of growth lines and growth inc rements, which is referred to as the growth pattern. The growth pattern has been used successfully in aging several intertidal gastropods and some gastropods that exhibit coiling; however, coiling can make it more difficult (Tojo and Ohno 1999). Most sn ails are hermaphrodites and some are protandrous (function as males when first sexually mature and later as females) (Evans 1972). Mating may involve courtship that can sometimes be elaborate and take many hours (Evans 1972). Most species lay 50 300 eggs ( ov iparous) (Evans 1972). Some of the larger snails lay between 100 to 200 eggs a year and the smaller ones may lay less than 50 a year

PAGE 44

44 (Boycott 1934). The eggs are deposited in loose damp soil and other locations that have suitable moisture retention p roperties (Evans 1972). Most species breed in the spring, but some breed in the fall (Evans 1972). Individually, small snails have a lower fecundity, but they have a higher reproductive output on a per v olume basis (Nekola et al. n.d. ). Even though most snails are hermaphrodites, the pupillids can self fertilize their eggs with their own sperm (Nekola 2009). The ability to self fertilize (and their small size) gives the pupillids an advantage in migration over other snails that cannot because only one i ndividual is needed to start a new population (Nekola 2009). Pupillid population densities often exceed hundreds if not thousands per square meter and are not highly endangered (Nekola 2009). Terrestrial mollusks have a coiled shell containing the visce ral organs, which can be fully withdrawn for protection (Barker and Mayhill 1999). Most gastropod locomotion is by means of a muscular foot, which contracts and facilitates movement (Bobrowsky ccurs by progressive waves of contraction and expansion of a ventral muscular foot that secretes slimy mucus, which enables the snail to glide over rough surfaces (Martin 2000). Terrestrial tin 2000) because they walk on a film of thin mucus and use water in the process, which is restored when they eat or drink (Boycott 1934). The amount of water lost from their skin depends on the humidity of the air (Boycott 1934) and the shells provide pr otection against water loss (Evans 1972). Terrestrial mollusk shells are useful in investigations of evolution and ecology and are usually species specific, yet malleable in form and size by the environment with the

PAGE 45

45 hell of the mature snail (Barker and Mayhill 1999). Snails are invertebrates that have a hard shell that is mainly composed of calcium carbonate (Evans 1972). The shells are coated with the periostracum, which is a thin proteinaceous coat (Evans 1972). T he periostracum is normally destroyed within one year of death, which leaves the calcium carbonate exposed to physical weathering and other degradation (Evans 1972). Once the periostracum is lost, ground waters have access to the crystal structure (Yates 1986). The crystalline form of calcium carbonate in the shells is pure aragonite (in most species) or calcite (in a few exceptions) (Evans 1972). Snails prefer calcareous soils because they need calcium for their shells (Boycott 1934). Terrestrial Snai l E cology Land snails are ecologically important for many reasons including their role in nutrient cycling ( Hodder 1991 ; Alvarez and Willig 1993 ). The ancient origin of terrestrial mollusks and slow evolution combined with low vagility and high rates of a llopatric speciation and ectothermal regulation make them susceptible to disturbance and modification of the microclimate, which makes them superb biogeographic models (Barker and Mayhill 1999; Bloch and Willig 2006). Most mollusks do not flourish in area s with intensive conditions of human occupation because hum an s destroy their characteristic habitats (Boycott 1934). Regardless of their ecological importance, the ecology of most land snails is not well known ( Secrest et al. 1996 ; Bloch and Willig 2006). Archaeological snail faunas can be used to estimate their former abundance and this can be compared to their nearest present day analogue to surmise the past environment in which they lived (Evans 1972). Land snails have limited dispersal

PAGE 46

46 capabilities, low mobility and small home ranges which make them an excellent study organism for evolutionary and ecological studies and can be used to infer paleoclimates and habitats ( Boycott 1934 ; Hodder 1991 ; Alvarez and Willig 1993 ; Waggoner et al. 2006). Activ e dispersal rates of individual land snails are very low and most cannot cross barriers of 102 103 m (Nekola et al. n.d. ). The landscape migration is solely related to passive dispersal methods that are aided by their adhesive mucus, which adheres to pass ing vertebrates (Nekola et al. n.d. ; Nekola 2009). Snails may be passively transported by locomotion, rafting during floods, migratory birds, aerial dispersal via wind, storms, birds, humans and large insects ( Boycott 1934 ; Vagvolgyi 1975; Martin 2000). Darwin and others suggest ed that terrestrial gastropods disperse by attaching themselves to newts, other amphibians, birds, water beetles, mammals and floating timbers (Bobrowsky 1984). Others favor chance dispersal by wind (Bobrowsky 1984). Based on thi s, smaller snails have a greater ability to be passively transported over long distances both overseas and overland (Nekola et al. n.d.; Vagvolgyi 1975 ; Nekola 2009). Environmental change may affect the snail population by changing the abundance of certa in species; some may increase, some may decrease, some may be eliminated or some may not be effected (Evans 1972). After the change occurs, other indigenous fauna may migrate into the area because they are now able to tolerate the new environment (Evans 19 72). The new colonizat ion is limited due to several factors such as the abundance of the colonizing species, the distance and type of land over which they must travel and their rate of dispersal (Evans 1972). Based on their low mobility, land snails may also be used to reflect patch dynamics (Alvarez and Willig 1993).

PAGE 47

47 Land snails lack great horizontal mobility, which makes them ideal subjects for local macrogeography studies (Barber 1988). Most land snails are more likely to move upward from the ground s urface onto vegetation than down into the base of leaf litter (Barber 1988). Evans (1972) state d that some snails may live up to 2 m below the soil surface and even though each species has an optimal habitat, they may live widely above or below it and som e may even aestivate or hibernate within the soil. However, Barber (1988) stated it is rare for snails to burrow from one level of a site to another. Therefore, archaeological snails are representative of their original location. Bobrowsky (1984) also s tated that post depositional effects on land snails are less significant than aquatic snails and the main source of transport or disturbance of the shells is due to earth worm activity. Other disturbance factors include burrowing mammals, tree throws, tra mpling and freeze/thaw cycles (Bobrowsky 1984). Peacock et al. (200 5 ) found that many of the very small species can be moved downward rather easily in the silt loam matrices, but not in the less porous, less permeable clays. Snails in the middle Savannah research area would not travel a significant depth into the soil or a significant horizontal distance along the surface due to the soil density (e.g. less permeable clays), habitat and species behavior common in the research area (personal communication, Slapcinsky 2011). snails burrowing from one site level to another is very rare and they are more likely to move above the ground onto vegetation than they are to burrow into the ground Therefore, the location of the snails in this study should be indicative of their location in the soil strata (chronology) if they remained in their natural location as in life (primary

PAGE 48

48 deposition). H owever some may be found in unnatural locations due to the depositional practices of humans (secondary deposition). Snails occupy a range of habitats above and within the soil including in the leaf litter, at the base of grasses, on the stalks of plants, beneath fallen branches or even on trees or stone walls (Evans 1972). In forest habitats, snails usually live among low plants, in the leaf litter and woody debris in the upper soil horizons (Hotopp 2002). They may also climb trees or follow crevices deeper underground (Hotopp 2002). Most snails (almos t 90%) live within 5 cm of the soil surface, mainly in the detritusphere (Nekola 2003). an herbivore, scavenger and carrion feeder that lives close to the ground surface and is preyed upon (Evans 1972). The l itter on the ground surface is a complex, three dimensional and horizontally stratified habitat which can be subdivided into several different habitats: newly fallen leaves at the top of the litter, fragmented leaves, twigs and decomposed litter farther d own and wet litter and fine humus above the ground level (Ba r ker and Mayhill 1999). The subunits of litter, bark and logs allow niche partitioning within these microhabitats, which promotes high diversity of land snails (Barker and Mayhill 1999). Litte r or top soil species may choose these habitats due to the protection against water loss (Alvarez and Willig 1993). Most land snails prefer moist and less disturbed habitats; however, some are adapted to live in arid environments (Evans 1972). The majorit y of snails move around at night when it is dark, cooler and the relative humidity is the highest or during wet weather ( Boycott 1934 ; Evans 1972). Soil moisture and sunlight levels influence land snail communities (Nekola 2003). Therefore, humidity and temperature are controlling factors on the distribution and abundance of Mollusca (Evans 1972).

PAGE 49

49 Density reduct behavioral responses to avoid desiccatio n caused by gaps in the forest, which increase the light and create c hanges in the air temperature, soil temperature, relative humidity and soil moisture (Alvarez and Willig 1993). Grasslands usually have drier, sunnier habitats compared to forests that are wet and shaded (Nekola 2003). This may be a factor; however, wood lands are more diverse environments than grasslands because they have a larger number of ecological niches (Evans 1972). Characteristic terrestrial mollusk species assemblages have been correlated to differences in vegetation, which indicates a degree of h abitat specialization (Barker and Mayhill 1999). Barker and Mayhill (1999) found that species assemblages in scrublands and forests corresponded to canopy tree species, canopy height, floristic diversity, altitude, litter mass and litter pH. Shell sizes have been found to increase with mollusk species richness and floristic diversity, which ind icated small shelled species dominate early successional and floristically poor sites and the addition of larger snail species into vacant niches that result in inc reased species richness (Barker and Mayhill 1999). Vegetation changes may also affect the temperature of the habitat, which can create micro environmental temperature increases and potentially change the snail fauna that can survive (Baur and Baur 1993). The competition among different snail species for food and living space is not considered to be significant in determining their distribution or abundance in a habitat because they do not form specific associations with one another or with other animals or plants ( Boycott 1934 ; Evans 1972). Their presence is determined by environmental conditions, such as moisture and lime (Boycott 1934).

PAGE 50

50 The occurrence of certain species may be governed by factors such as the physical and chemical properties of the soil a nd ground surface rather than the presence of grassland or woodland (Evans 1972). Hence, similar hab itats ( such as grassland ) may be occupied by different snai l faunas, or dissimilar habitats ( such as grassland and woodland ) may be occupied by the same sn ail faunas because they are occupying the same niche within different habitats (Evans 1972). This shows tha t factors, such as the moisture retention capacity of the soil, the rate of nitrogen turnover, shelter and lime, may be of higher importance than th e gross structure of the vegetation (Evans 1972). Barker and Mayhill (1999) found a weak or nonexistent correlation of mol lusk diversity and abundance with pH among single vegetation sites Correlations are normally found between the vegetation soil pH and litter pH and between the soil or litter calcium content and pH S ignificant correlations of snail abundance and low calcium content were found to be normal. Different environmental niches may also be occupied by the same species (Evans 1972). Land snail communities may be correlated with litter depth and may be impacted by the architecture of the organic litter and the underlying soil (Nekola 2003). Anthropogenic soil disturbance also influences snail community composition and will impact soil inver tebrates (Nekola 2003). Anthropogenic activity may increase soil compaction and decrease organic litter layers, which will limit the type of snail that can withstand that type of habitat (Nekola 2003). Disturbances, such as clear cutting, may change the environment of the litter layer by reducing moisture which changes the abundance and number of snails and taxa able to live in the altere d habitat (Hodder 1991 ; Hylander et al. 2004 ). Some disturbances can impact some ecosystems to the

PAGE 51

51 point that the com munities are considered disturbance mediated (Secrest et al. 1996). The survival rates of snails in these disturbed areas depend upon the variation of Clear cutting radically affects land snails by decreasing species abundance and numbers (Hodder 1991). The pH of the organic litter may also be a strong influence on species richness and individual abundance (Nekola 2009). Nekola (2009) found high and moderately acidic sites have sign ificantly lower species richness compared to neutral/calcareous habitats, but the most acidic sites still supported 5 10 species. Thus, anthropogenic and non anthropogenic impacts effecting the environment can be gauged by changes in species composition ( Waggoner et al. 2006). Terrestrial S nail Taphonomy Fossil assemblages can either be residual, transported or a mixed assemblage (Fagerstrom 1964 ). Residual fossil assemblages consist of nearly all of the specimens that belonged to the same ecological co mmunity, but they are not present in the same numbers and sizes as they were during life due to moderate alteration by pre burial factors, which ha ve removed a portion of the original community (Fagerstrom 1964 ). Fagerstrom ( 1964 ) state d that the residual fossil community is the same as the life assemblage because most of the fossils are found in their original habitats and living positions. Transported fossil assemblages occur when nearly all the specimens have been altered by preburial transportation an d almost none of the fossils are found in their original habitats or living positions (Fagerstrom 1964 ). Mixed fossil assemblages contain large numbers of specimens that belonged to the sa me ecological community, but also contain a large number of specime ns from contemporaneous communities or from the erosion of preexisting rocks (Fagerstrom 1964 ). The life assemblage is

PAGE 52

52 different than the death assemblage. The death assemblage is a representation of the specimens in the position they were at the time of death, which may be different than their original habitats and living positions (Fagerstrom 1964 ). M ost nonmarine gastropods preserve well in many different deposits however, the potential preservation of gastropod shells may vary based on the durabili ty of the periostracum (proteinaceous shell cover) (Bobrowsky 1984). Evans (1972) believed shells are readily preserved in lime rich soils and sediments because they are composed mainly of calcium carbonate and are highly calcareous and are rarely preserv ed on neutral or acidic soils. However, Coles and Walsh (1999) found that substantial land snail populations are often supported in acidic habitats in eastern North America. The thickness of the shells may also determine the resistance to environmental conditions i.e. thin shelled species will be destroyed more easily than more robust shells (Evans 1972). Alvarez and Willig (1993) found that d ifferences in shell aperture area, shell thickness and epiphragm thickness result in differences in water loss rates in xeric and mesic habitats. Based on this, Alvarez and Willig (1993) were able to correlate s maller shell aperture, greater shell thickness and greater epiphragm thickness with species in xeric habitats This research considers terrestrial snails as a proxy for land alteration and temporal variations due to the impact of hunter gatherer populations on their environment resulting from the removal or alteration of ground cover. Snails were collected from three archaeological sites and their corresp onding modern sites. The snail assemblages will be analyzed as a means of interpreting past anthropogenic impacts among hunter gatherer societies.

PAGE 53

53 CHAPTER 4 MODERN SAMPLES Introduction As discussed in Chapter 3 terrestrial snails in temperate zones acro ss the globe occupy a variety of niches, from closed canopy forests to grassy meadows to open, rocky terrain. To establish the distribution of terrestrial snail taxa across different microhabitats in the study area, in 2002, Dr. Evan Peacock (Mississippi State University) and Peter Hallman (formerly with the University of Florida) collected modern snails from 20 sites, representing distinct microhabitats. The method s and results of this aspect of the study are reported in this chapter with the goal of det ermining the expectations for terrestrial snail distributions across inferred microhabitats during the Late Archaic period in the middle Savannah River Valley. The identification of taxa tolerant to open habitat is o f particular interest in the analysis o f the modern snail taxa i n this study. The results of the analysis will be used as a basis to infer instances of land clearing among prehistoric denizens in the study area. Details of habitat preferences and tolerance levels are not well known for many terrestrial snail taxa. Even taxa that have been considerably studied are not always sufficiently sampled across multiple microhabitats or geographical areas at different times of the year, which limits the ability to determine their full distribution ran ge with certainty. Other problems with this type of analysis include accidental intrusion of a species into another habitat (Barber 1988), periodic burrowing and postmor tem displacement of shell through soil erosion. Barber (1988) suggested that these i ssues can be addressed by assessing the total number of individual s recovered, historical including their ability

PAGE 54

54 to burrow or intrude into a site, and the likelihood that the species could have been transported into the site. Similar to many types of biotic remains discovered at archaeological sites, modern land snail collections and records of the habitat cond itions under which they lived a re needed to interpret the archaeological data in this study (Peacock and Gerber 2008). As with historical records, modern analogues for biotic communities identified in the archaeological record cannot be assumed which is a constant challenge in environmental archaeology (Peacock and Gerber 2008). Fu rthermore, providing a robust context for archaeological interpretations enables proper data a nalysis and assumptions o n which to base our interpretations of the environmental conditions at the archaeological sites ( Peacock and Gerber 2008). Therefore, t h e modern snail identifications will be used along with the documented vegetation data to find nonrandom associations between taxa and habitat type. The taxa showing habitat tendencies will then be used to analyze each sample to determine if each sample is indicative of an open or wooded habitat. The results of the modern sample s will be used in the following chapters to infer the types of habitats over time and throughout the sites in an attempt to decipher population pressures and instances of land clear ing within the study area. Collection Methods Occurrence data utilizes presence only sampling data and does not consider the absence of species in the analysis of the results. Occurrence data were used in this study to determine the tendencies of snail taxa to occur in certain habitats categorized as open or wooded. Historically, occurrence data has been used to estimate species diversity and distribution within an area for ecological and conservation purposes

PAGE 55

55 ( Keating et al. 1998; Hellman and Fowler 19 99; Plotkin and Muller Landau 2002; Bailey et al. 2007). Some consider presence only data as biased due to sampling issues that may not record a species when it is in fact present and based on the inability to confirm MacKenzie 2005). Sampling bias may be increased due to small sample areas, small sample sizes, short sampling durations (Fisher et al. 1943; Gotelli and Colwell 2001; White 2004; Fridley et al. 2006) or only sampling locations that are readily accessible such as roadsides or urban areas (Reddy and Davalos 2003; Phillips et al. 2009). To reduce bias, Fridley et al. (2006) and Bailey et al. (2007) recommend sampling multiple times over a variety of locations to increase the number of habitats and time per iods sampled and to increas e sample sizes. To reduce bias i n thi s study, a large number of individuals were sampled (large sample sizes), a wid e variety of microhabitats to represent variations found within open and wooded habitats were collected the samp les were collected within the same general location as the archaeological sites, and the modern live snail assemblages and the death assemblages within each microhabitat were collected to obtain a sufficient representation of the native fauna. Despite eff orts to alleviate biases, some still exist. These biases include the inability to collect multiple samples over time due to the remote location and the comparison of modern snail fauna with archaeological fauna. Field Procedures In August 2002, s amples o f soil litter and modern snails were collected by Dr. Evan Peacock (Mississippi State University ) and Peter Hallman (formerly with the University of Florida) in an attempt to collect a representative sample of the terrestrial snail fauna composition from s ites similar in geographical location as the archaeological sites of this study Paired samples of leaf/grass litter were collected at 20 different

PAGE 56

56 locales su rrounding the Stallings Island s ite ( Table 4 1, Figure 4 1) The locales were chosen to represen t different micro habitats including mature hardwood forests, mixed hardwood floodplains, upland hardwoods, pine forests, early successional stands and human created clearing s V egetation type and density were documented at each site. The paired samples i ncluded litter samples and hand collection of larger snails. The litter sam ples were collected from 1 meter square sampling plots in an attempt to provide a constant volume among samples. V isible larger snails were collected by hand within a 10 meter dia meter circle surrounding each litter sampling locale. S ite Descriptions The modern sample s ites were chosen based on their locale in relation to the archaeological sites, soil moisture, ground coverage, vegetation type and canopy coverage Only 14 of t he 20 modern samples had sufficient sample sizes to warrant statistical comparisons of relative frequencies The remaining six samples were not considered due to the limited number of individual snails (less than 10) recovered within each sample. The sna ils collected by hand were collected to bolster the paired soil samples by potentially collecting taxa from the immediate area that were not found in the soil samples Therefore, the hand collected snails were not used in the statistical comparisons in th is study. Six of the 14 modern samples (14, 15, 16, 17, 19 and 20) sampled were collected on the northeast side of the Savannah River to the northeast of Stallings Island ( Table 4 1, Figure 4 2 ) Four of the six samples were collected in open habitats ( 15, 16, 19, and 20) and two of the six samples were collected in wooded habitats (14 and 17) ( Table 4 2, Figure 4 3 ) Two of the open habitat samples and one of the wooded habitat samples w ere st

PAGE 57

57 aspect ( Table 4 1, Figure 4 2 ) Two of the open/grassy habitat samples and one of the wooded habitat samples w with a south/southwest aspect. ( Table 4 1, Figure 4 2 ) Six of the 14 modern samples (7, 8, 9, 10, 11 and 1 2) utilized were collected on Stallings Island ( Table 4 1, Figure 4 4 ) Two of the six samples were collected in open habitats (8 and 12) and four of the six were collected in wooded habitats (7, 9, 10, and 11) ( Table 4 2, Figure 4 5 ) All of the samples ( Table 4 1, Figure 4 4 ) Both of the open samples were located in areas with a minimal slope; one of which had a northeast aspect and the other had a southwest aspect ( Table 4 1, Figure 4 4 ) Three of the wooded samp les had a southeast/east aspect and o ne of the wooded samples had a northwest aspect with a minimal slope. ( Table 4 1, Figure 4 4 ) Two of the 14 modern samples (5 and 6) were collected on the east side of the Savannah River to the southeast of Stallings Is land ( Table 4 1, Figure 4 6 ) Both of the samples were collected in wooded habitats ( Table 4 2 Figure 4 7 ). The two samples south west aspect ( Table 4 1, Figure 4 6 ) Open h abitats As shown in Figures 4 3 and 4 5 six of the fourteen samp le areas chosen for this study were categorized as open canopy habitat s (samples 8, 12, 15, 16, 19 and 20). The samples considered as open habitats supported various types of ground vegetation, includi ng grasses, weeds and briars but lacking tree cover The vegetation observed at the samples categorized as open habitats is described on Table 4 3.

PAGE 58

58 The vegetation at sample 8 was described as having very dense grass, briars and thick cane approximately 6 ft tall with one small Sumac. Sample 12 was complementary to sample 8 and was described as being cleared land densely covered with grasses, briars and cane approximately 7 ft tall with one small sumac. Sample 15 was described as being open grassland l ocated under power lines with approximately 4 ft tall dense grass and briars without any trees. The soil was described as grey compact sandy loam and the site was located on the first terrace of the river. Sample 16 was complementary to sample 15 and loc ated approximately 30 meters to the northwest of sample 15. This sample was also on the first terrace of the river and consisted of open grassland with approximately 3 ft tall dense grass and briars without any trees. The soil was also described as grey sandy loam. Sample 19 was located on a ridge top under the power lines and was described as being open grassland approximately 25 meters southeast of the woods. The site was void of trees and had been mowed recently but high grass was present around the base of the power line poles. The site was selected as a cleared upland habitat. Sample 20 was complementary to sample 19, but was not mowed as recently. It was also located on a ridge with a slight slope under the power lines and was described as gras sland with thick grass cover approximately 10 inches high and void of trees. Wooded h abitats As shown in Figures 4 3 4 5 and 4 7 e ight of the fourteen samples were collected from areas categorized as wooded (closed canopy) habitats (samples 5, 6, 7, 9, 10, 11, 14 and 17). T he samples considered as wooded habitats consisted of hardwood to mixed hardwood forests (closed canopy) with ground coverage ranging from logs and

PAGE 59

59 leaf litter to a small number of grass tufts. The vegetation observed at the samples considered wooded habitats is described on Table 4 4. Sample 5 was described as a hardwood forest with large oaks, elms, sweet gums and a couple of palmettos on a natural levee beside the Savannah River. Sample 6 was complementary to sample 5 and had the same type of vegetation and description. Sample 7 was located in a hardwood forest with thick ground cover to the north of the power lines on Stallings Island. Sample 9 was complementary to sample 7 and described as being in a hardwood forest with a lar ge number of logs and debris on the ground. Sample 10 was described as being on top of the main shell mound on Stallings Island and in a hardwood forest with a large number of logs and debris on the ground. Sample 11 was complementary to sample 10 and ha d the same type of vegetation and description. Sample 14 was located on the first terrace of the Savannah River in a hardwood forest with limited ground vegetation/cover. Sample 17 was located on a high ridge and described as upland mixed hardwood forest with ground coverage consisting of several clumps of grass approximately 10 in high. Laboratory Procedures The samples were returned to the laboratory at Mississippi State University with Dr. Peacock and weighed. The litter was then washed through a seri es of nested screens measuring 5 mm, 1 mm, 500 micrometers and 250 micrometers. Care was taken to wash all large pieces of detritus during this process. Afterwards, the snails were separated from the litter by hand, with foil tweezers or with a small pai nt brush, depending upon the size of the shell. The detritus was systematically sorted to ensure all snail shells were recovered from the samples.

PAGE 60

60 The identifications were partially completed by Dr. Peacock and Dr. Jochen Gerber (The Field Museum, Chicago Illinois) and used as a reference for the majority of the other identifications. The remaining shells were observed on a white background and separated using a small paint brush or by hand, depending upon shell size, based upon observable shell characte ristics. Some of the shells were thoroughly cleaned in order to observe the internal structures within the aperture. Identifiable shells were then classified by taxa using a strong light source and a microscope at 5X to 40X magnification, when necessary. In almost all instances, t he snails were counted using whole shells or the shell apices. If individualizing characteristics were present on the shell aperture or other fragments they were also counted If multiple fragments were present from one appar ent species, then it would only be counted as one individual; only fragments from the same area were counted as multiple individuals. Fragments that presented no individualizing characteristics were not counted and separated into container s labeled as uni dentified. The snails were identified based upon taxonomy guides and geographical ranges in Pilsbry (1939 1948), Hubricht (1985), and Nekola and Coles (2010) In addition snails previously identified by Evan Peacock and Jochen Gerber were used for refer ence Each container was clearly marked with the sample data and identified taxon information. The total number of individual shells within each taxon identified per site was recorded. Nomenclature follows that of Hubricht (1985) and Nekola and Coles (2 010). Modern Snail Identifications Snail habitat preferences may vary based up on many different factors including available moisture, acidity, shade, soil type, vegetation type predators and geographical location The variation in snail habitat prefere nces has been observed

PAGE 61

61 throughout the literature and will be used to support the open and wooded habitat categories listed above ( Table 4 5 ) Thirty eight taxa were identified from the 14 modern soil samples, which included a total of 1247 individual sna ils that represented 9 families and 19 genera (Table 4 6) Only terrestrial snail taxa found in quantities greater than or equal to 15 individuals were utilized to find nonrandom associations between snail taxa and habitat types amo ng the modern samples The taxa identified and the total number of individuals identified in the modern samples are listed on Table 4 6. A total of 52 individual snails were identified as Euconulus chersinus ( Table 4 6 ). Euconulus chersinus has been previously found in mois t leaf litter habitats (Table 4 5 ). In contrast, Archer (1939) found the taxa in a wider range of environments in Ann Arbor Michigan, including flood susceptible grass marsh habitats, swamp woods, forests, roadsides, and open fields. However, this study found more Euconulus chersinus in wooded habitats (75%) than open habitats (25%) in the modern samples (Table 4 7) A total of 93 individual snails were identified as Gastrocopta contracta ( Table 4 6 ). Gastrocopta contracta has previously been documente d in a variety of habitats ranging from open to wooded (Table 4 5 ). Several others have also recorded the taxa in habitats ranging from open to wooded in Livingston County, Michigan (Archer 1939), Ann Arbor, Michigan (Goodrich 1943), the Great Lakes regio n of North America (Nekola 2003), northeastern Wisconsin (Nekola 2004), southern Ontario (Meyrick and Karrow 2007), and throughout eastern North America (Nekola and Coles 2010). Similarly, this study found Gastrocopta contracta equally divided between ope n (50.5%) and wooded (49.5%) habitats in the modern samples (Table 4 7 ). The majority of the snails within

PAGE 62

62 this taxon were found in samples 9, (n=13), 10 (n=16), 11 (n=17), and 12 (n=45) ( Table 4 8 ). Samples 9, 10, and 11 are described as hardwood forest s with a large amount of logs and debris on the ground (Table 4 4). Sample 12 is described as cleared land that is covered with dense grasses, briars, 7 ft cane, and one Sumac (Table 4 3). These samples indicate a variety of habitats and support the resu lts recorded in previous literature (Table 4 5). A total of 65 individual snails were identified as Gastrocopta pentadon ( Table 4 6 ). Gastrocopta pentadon has been previously documented in habitats ranging from wooded to open (Table 4 5 ). Archer (1939) d ocumented the species in several different habitats including Bi rch Maple swamp woods habitats with fluctuations in water levels, on the richer slopes and upper dry zones of Oak Hickory communities, roadsides, open fields and pastures. In the modern sampl es, Gastrocopta pentadon was equally divided in open habitats (50.8%) and wooded habitats (49.2%) (Table 4 7 ) A total of 39 individual snails were identified as Gastrocopta rupicola ( Table 4 6 ). Gastrocopta rupicola has been previously observed in wooded and roadside environments (Table 4 5 ). Nekola and Coles (2010) recorded Gastrocopta rupicola in decomposed leaf litter mainly and often under dense shrub or vine thicket cover in lowland forest scrub and disturbed habitats T his study found Gastrocopta rupicola in more open habitats (94.9%) than wooded (5.1%) habitats in the modern samples (Table 4 7 ) The majority of the snails within this taxon were found in samples 8 (n=26), 19 (n=5), and 20 (n=6) (Table 4 8) Sample 8 was described as having very dense grass, briars, 6 ft tall cane and one small Sumac and samples 19 and 20 were described as being open grassland under power lines with recently mowed grass (Table 4 3).

PAGE 63

63 Sample 8 may have provided the dense cover needed for this species to survive an d samples 19 and 20 would be considered disturbed habitats as described by Nekola and Coles (2010), thus showing consistent habitat tendencies as recorded in the previous literature (Table 4 5). A total of 26 individual snails were identified as Glyphyalin ia cryptomphala and a total of 150 individual snails were identified as Glyphyalinia species ( Table 4 6 ). Glyphyalinia cryptomphala has previously been documented in moist leaf litter (Table 4 5 ) and was found in only wooded habitats (100%) in the modern samples (Table 4 7 ) Individuals that were only able to be identified as Glyphyalinia species also showed a preference for wooded habitats (73.3%) rather than open habitats (26.7%) in the modern samples (Table 4 7 ) A total of 392 individual snails were i dentified as Hawaiia minuscula ( Table 4 6 ). Historically, Hawaiia minuscula has been found in a variety of habitats including wooded, damp leaf litter, urban and open (Table 4 5 ). Harms (2008) considered the tax on an opportunistic species that can tolera te a wide range of environmental conditions in Kentucky. Leonard (1959) also found the species in a variety of habitats in Kentucky, which included under logs, sticks and stones and in clumps of grass in floodplains and upland areas. Archer (1939) presen ted similar results from Livingston County Michigan and others have stated that Hawaiia minuscula prefers moist environments, regardless of the habitat ( Solem 1978 ). Nekola (2004) found the taxa in more lowland grasslands, but also in forested sites in th e Great Lakes Region of North America. Hubricht (1985) also found Hawaiia minuscula on bare ground habitats, including floodplains, meadows, roadsides, railroad tracks and urban waste grounds,

PAGE 64

64 but never in leaf litter. T his study found more Hawaiia minus cula in wooded habitats (74.4%) than open habitats (25.3%) (Table 4 7) A total of 67 individual snails were identified as Helicodiscus p arallelus ( Table 4 6 ). Helicodiscus parallelus has been found in wooded, urban, and roadside habitats, and historicall y has been found to need some type of shade for its survival (Table 4 5 ). N ekola (2003 2004 ) found the taxon in the trash of vacant lots in Ann Arbor, Michigan and in a variety of habitats in northeastern Wisconsin and southern Michigan, including only g rassland sites in the Great Lakes region of North America. Harms (2008) considered Helicodi s cus parallelus a forest species that flourished in damp, shady places in Kentucky. Meyrick and Karrow (2007) located the species in closed woodland habitats in so uthern Ontario. In contrast, Archer (1939) reported it in a variety of habitats, including under rotten logs, in leaf mold, grass, mold, rock piles and bramble thickets in Livingston County Michigan. Overall, most suggest that Helicodi s cus parallelus pre fers rotting wood or damp leaf litter in moist shaded environments (Table 4 5 ). The modern samples showed similar preferences and found a higher percentage of Helicodi s cus parallelus in wooded habitats (61.2%) than open habitats (38.8%) (Table 4 7 ). A tot al of 54 individual snails were identified as Punctum minu t i ssimum and Punctum smithi and combined into the genus Punctum ( Table 4 6 ). The two species were combined due to their similarities and common preferences to live in deep leaf litter ( Table 4 5 ). Punctum species were located in more wooded habitats (85.2%) than open habitats (14.8%) in the modern samples (Table 4 7 )

PAGE 65

65 A total of 46 individual snails were identified to the family Pupillidae ( Table 4 6 ). Pupillidae is a family of snails that include s the genera Gastrocopta, Pupisoma and Vertigo Snails identified as Pupillidae were either fragmentary or juveniles and were unable to be identified to genus or species level. The snails identified as Pupillidae were found in more open habitats (58.7%) than wooded habitats (41.3%) in the modern samples (Table 4 7 ) A total of 43 individual snails were identified to the genus Pupisoma ( Table 4 6 ). Pupisoma species has been previously documented in wooded areas ( Table 4 5 ). However, in this study Pupiso ma species was found in more open habitats (74.4%) than wooded habitats (25.6%) (Table 4 7 ) The individuals were found in samples 9 ( n=11 ) and 12 ( n= 28) (Table 4 8) Sample 9 is described as a hardwood forest (Table 4 4) and sample 12 is described as cl eared land with dense grass, briars, 7 ft tall cane, and one small Sumac (Table 4 3). The dense ground cover in sample 12 may have created a more shaded and protected environment that permitted the snails within this taxon to survive and flourish in the h abitat, unlike what has been previously recorded. A total of 29 individual snails were identified as Vertigo rugosula and a total of 15 individual snails were identified as Vertigo tridentata ( Table 4 6 ). Vertigo rugosula and Vertigo tridentata have a his tory of being found in open, grassy and roadside habitats with V. rugosula being able to survive in dry environments ( Table 4 5 ). Vertigo tridentata has also been located in leaf litter ( Table 4 5 ). Vertigo rugosula and Vertigo tridentata were found in m ore open habitats (100%, 60% respectively) than wooded habitats (0%, 40% respectively) in the modern samples (Table 4 7 )

PAGE 66

66 A total of 100 individual snails were identified to the family Zonitidae ( Table 4 6 ). Zonitidae is a family of snails that has been observed more often in wooded habitats (Table 4 5 ), a nd includes the genera Hawaiia and Glyphyalinia Snails identified as Zonitidae were usually juveniles that were unable to be identified to genus or species due to the lack of species specific adult sh ell characteristics. The Zonitidae were only found in only wooded habitats (100%) in the modern samples (Table 4 7 ) T ax a with open habitat tendencies included Gastrocopta rupicola Pupillidae, Pupisoma species, Vertigo rugosula and Vertigo tridentata (T able 4 7 ). The taxa with wooded habitat tendencies included Euconulus chersinus, Glyphyalinia species Glyphyalinia cryptomphala, Hawaiia minuscula, Helicodiscus parallelus Punctum species and Zonitidae (Table 4 7 ) However, Hawaiia minuscula was exclu ded from the taxa with wooded habitat tendencies due to the large number of individuals and because the taxon is historically not habitat specific (Table 4 5). The total number of individuals with open habitat tendencies w as compared to the total number o f individuals with wooded habitat tendencies. In the open habitat samples, a total of 134 individuals (60.6%) with open habitat tendencies and a total of 87 individuals (39.4%) with wooded habitat tendencies were found (Table 4 9). In the wooded habitat samples, a total of 38 individuals (9.5%) with open habitat tenden cies and a total of 362 individuals (90.5%) with wooded habitat tendencies were found (Table 4 9). As expected, the results indicate that more taxa with open habitat tendencies are located in the open habitats and more taxa with wooded habitat tendencies are located in the wooded habitats.

PAGE 67

67 The C hi square test was used to determine the strength of associations between the habitat tendencies and habitat types. The dependent sampling techniq ues used in the collection of the modern samples skewed the results of the Chi square test to some degree, but the results clearly indicate a significant association. The Chi square test to evaluate nonrandom associations between habitat tendencies of the taxa and the habitat types of the samples resulted in a C hi square p value of <0.0001, which suggests an extremely significant ass ociation (Table 4 10 ). Regardless of the dependent sampling method concerns, nonrandom associations among habitat tendencies and the habitat types of the samples are evident in the numerical data alone, which is verified by the Chi square statistic even if its probability values cannot be accepted uncritically due to the biases of dependent data. To increase the number of indiv iduals within taxa and to utilize data within the taxa with less than 15 individuals, all of the snail fauna within the families Pupillidae and Zonitidae were combined Th e family Pupillidae included all of the snails within the taxa : Gastrocopta species, Gastrocopta armifera, Gastrocopta contracta, Gastrocopta pentadon, Gastrocopta rupicola, Gastrocopta tappaniana Pupillidae, Pupisoma species, Pupisoma discoricola, Vertigo species, Vertigo clappi, Vertigo oscariana, Vertigo pygmea, Vertigo rugosula and Vertigo tridentata The family Zonitidae included all of the snails within the taxa : Glyphyalinia species, Glyphyalinia cryptomphala, Hawaiia minuscula Mesomphix species, Mesomphix andrewsae, Mesomphix globosus, Striatura meridionalis, Zonitidae and Zon itoides arboreus The two families were compared within open and wooded habitat samples to determine if there was a statistically significant association between taxa and habitat The taxa within the family

PAGE 68

68 Pupillidae were more numerous in the open habit at samples (64. 9 %) than the wooded habitats (35 .1 %) and the taxa within the family Zonitidae were more numerous in the wooded habitat s amples ( 76.7 %) than open habitats ( 22.7 %) (Table 4 1 1 ). The Chi square test was used to determine the strength of assoc iations between the habitat tendencies of the snails in the families Zonitidae and Pupillidae and the habitat types of the samples, which resulted in a Chi square p value of <0.0001, which suggests an extremely significant association (Table 4 12). N onran dom associations among habitat tendencies of the families Zonitidae and Pupillidae and the habitat types of the samples are evident in the numerical data alone, which is verified by the Chi square statistic even if its probability values cannot be accepted uncritically due to the biases of dependent data. The results support a significant association between the habitat tendencies of the snail taxa and the habitat type. The change in percentages observed with the removal of Hawaiia minuscula indic ated that Hawaiia minuscula may be a generalist species or be too numerous to include in the taxa with wooded habitat tendencies E ach sample was analyzed further to determine if any site specific variations or microhabitat inclinations could explain the numbers a nd tendencies of the taxa found within each of the modern samples. Snails within Habitats The tax a indicating open habitat tendencies include : Gastrocopta rupicola Pupillidae, Pupisoma species, Vertigo rugosula and Vertigo tridentata and the taxa indica ting wooded habitat tendencies included Euconulus chersinus, Glyphyalinia species Glyphyalinia cryptomphala, Helicodiscus parallelus Punctum species and Zonitidae (Table 4 7) The samples were compared to determine if taxa with open or

PAGE 69

69 wooded habitat t endencies were found within the corresponding habitats The taxon Hawaiia minuscula was excluded from this comparison due to the reasons listed previously. Four of the six open habitat samples (8 12, 19, and 20 ) were observed to have more individuals wit h open habitat tendencies (59.1% 55.1%, 100%, and 100%, respectively) (Table 4 1 3 ). In contrast, sample 15 was observed to have individuals with only wooded habitat tendencies (100%) and sample 16, which is complementary to sample 15, was observed to hav e slightly more individuals with wooded habitat tendencies (52.2%) (Table 4 13). Samples 15 and 16 were located under power lines and had vegetation consisting of tall dense grass that was approximately 3 to 4 ft high (Table 4 3). The tall dense grass m ay have provided sufficient shade and protection from the sun and weather to support snails with more generalist or wooded habitat tendencies. Based on these results samples 15 and 16 may represent intermediate habitat s between open and wooded In contra st to the open habitat samples, all eight of the wooded habitat samples (5, 6, 7, 9, 10, 11, 14 and 17) were observed to have more individuals with wooded habitat tendencies (100%, 100%, 9 2.3 %, 71.8 %, 88.2%, 95.6%, 100%, and 100%, respectively) (Table 4 1 3 ). The results support the associations between taxa with wooded habitat tendencies and the wooded habitat samples. Snail dens ity per liter of soil may also a ffect the overall results and associations between habitat tendencies and habitat type and will be discussed below Snail Density Most studies compar ing sites or communities standardize the sampling techniques by either collecting a fixed number of sample units or collecting a fixed amount of

PAGE 70

70 sample (area or volume) (Cao et al. 2002). This study standardized the area sampled and attempted to collect a similar amount of soil litter from each location. However, without carrying a scale, the amount of soil collected varied slightly. To correct the variation of volume among samples, the data was sta ndardized through mathematical means, such as averaging (Hurlbert 1971). The data was further analyzed to determine if the total number of individuals with open or wooded habitat tendencies was dependent upon the ini tial volume of soil collected or if it was based on the actual habitat preferences of the taxa It was also analyzed to determine if a larger number of individual snails were found within open or wooded habitats due to the possibility that the general environment or microhabitats could have cr eat ed more preferable living conditions in either habitat The open habitat samples ranged from 3.0 L to 5.9 L of soil with an average of 4. 6 L of soil ( Table 4 14, Figure 4 8 ) The wooded habitat samples ranged from 4.6 L to 6.6 L of soil with an averag e of 5.5 L of soil ( Table 4 14, Figure 4 8 ) All of t he open habitat samples contained a total of 420 individual snails within all taxa, rang ing from 0 to 99 in dividual snails in each taxon (Table s 4 7 and 4 14) All of the wooded habitat samples contain ed a total of 827 individual snails within all taxa, ranging from 0 to 293 individual snails in each taxon (Table s 4 7 and 4 14) Considering all of the identified snail s the open habitat samples had an average density of 15.4 snails per liter of soil (r anging from 2.1 to 35. 4 snails per liter of soil ) and the wooded habitat samples had an average density of 19.3 snails per liter of soil (ranging from 3.4 to 82. 7 snails per liter of soil) (Table 4 1 5 Figure 4 9 ).

PAGE 71

71 In the previous sections, only five snail taxa were determined to have a pre ference to live in the open habitat samples ( Gastrocopta rupicola Pupillidae, Pupisoma species, Vertigo rugosula and Vertigo tridentata ) and only six snail taxa were determined to have a preference to live in the wooded habitat samples ( Euconulus chersinus, Glyphyalinia species Glyphyalinia cryptomphala, Helicodiscus parallelus Punctum species and Zonitidae) The density of snails with open habitat tendencies among the six open habitat samples ranged from 0.0 snails per liter of soil to 10.2 snails per liter of soil with an average of 4 .7 snails per liter of soil ( Table 4 15, Figure 4 10 ) In comparison, the density of snails with wooded habitat tendencies among the six open habitat samples ranged from 0.00 snails pe r liter of soil to 3.1 snails per liter of soil with an average of 0.8 snails per liter of soil (Table 4 15 Figure 4 10 ) The open habitat snail densities among the eight wooded habitat samples ranged from 0.0 snails per liter of soil to 8 .3 snails per l iter of soil with an average of 3.3 snails per liter of soil. (Table 4 1 5 Figure 4 10 ) In comparison the density of snails with wooded habitat tendencies among the eight wooded habitat samples ranged from 2.6 snails per liter of soil to 29.8 snails per liter of soil with an average of 8.6 snails per liter of soil (Table 4 1 5 Figure 4 10 ). As expected on average, snails with open habitat tendencies were found in greater densities in the open habitat samples than snails with wooded habitat tendencies S nails with wooded habitat tendencies were also found in greater densities in the wooded habitat samples than snails with open habitat tendencies However, a larger difference was observed between the densities of open and wooded snail fauna in the wooded h abitat samples than the open habitat samples. This may be due to

PAGE 72

72 sampling error or microhabitat variations that allowed for a wider range of habitat tolerances within the open habitat samples In general, more snails were found per l iter of soil in the wooded habitats than the open habitats, which may indicate harsher environmental conditions in open habitats than in wooded habitats This could be due to lower moisture levels, higher temperatures and /or less protection from the sun and environment. Discussio n Modern soil litter and snail samples were collected to determine the tendencies of the snail s to be found in open or wooded habitats. Fourteen different robust samples were collected from three locales on and around Stallings I sland. The fourteen samples represented six open and eight wooded habitats. Five snail taxa showed a tendenc y to be found in the open habitat s amples The taxa with open habitat tendencies included: Gastrocopta rupicola, Pupillidae Pupisoma species, Ve rtigo rugosula and Vertigo tridentata (Table 4 7 ). Seven snail taxa showed a tendenc y to be found in the wooded habitat s amples The taxa with wooded habitat t endencies included: Euconulus chersinus, Glyphyalinia species, Glyphyalinia cryptomphala, Hawa iia minuscula, Helicodiscus parallelus, Punctum species and Zonitidae (Table 4 7 ). Due to the high number of individuals within the taxon Hawaiia minuscula and previous studies on this taxon indicating they are generalists, they were removed from the gro up categorized as having wooded habitat tendencies, which allowed the taxa with open habitat tendencies to be visualized within the samples (Table 4 5). The C hi square test also supported a significant association between the taxa with open and wooded hab itat within the open an d wooded habitat samples (Table 4 10 ). The C hi square test also supported a significant association between the taxa within the family Pupillidae and the open

PAGE 73

73 habitat samples and a significant association between the taxa within the family Zonitidae and the wo oded habitat samples (Table 4 1 2 ). The five taxa with open habitat t endencies ( Gastrocopta rupicola, Pupillidae Pupisoma species, Vertigo rugosula and Vertigo tridentata ) and the six taxa with wooded habitat t endencies ( Euconu lus chersinus, Glyphyalinia species, Glyphyalinia cryptomphala, Helicodiscus parallelus, Punctum species and Zonitidae) will be used in the following chapters as a basis to infer the habitats at the Late Archaic sites and to infer change s in the habitat s over time.

PAGE 74

74 Table 4 1. Location and aspect of the modern samples Modern Sample Stallings Island Northeast of Stallings Island Southeast of Stallings Island Aspect Elevation (feet) 5 X S W <200 6 X S W <200 7 X NW 166 172 8 X NE 166 172 9 X SE/E 166 172 10 X SE/E 166 172 11 X SE/E 166 172 12 X SW 166 172 14 X SW <200 15 X SW <200 16 X SW <200 17 X SW 200 250 19 X SW 200 250 20 X SW 200 250 Table 4 2. Habitat observations for the modern samples Modern Sample Ope n Habitat Wooded Habitat 5 X 6 X 7 X 8 X 9 X 10 X 11 X 12 X 14 X 15 X 16 X 17 X 19 X 20 X

PAGE 75

75 Table 4 3. Descriptions of each modern sample categorized as an open habitat Sample Location Description 8 Stallings Island Very de nse grass, briars, 6 ft tall thick cane, one small Sumac 12 Stallings Island Cleared land, densely covered with grasses, briars and 7 ft tall cane, one small Sumac 15 NE of Stallings Island Open grassland under power lines, dense 4 ft tall grass, briars, grey compact sandy loam soil, on the first terrace of the river 16 NE of Stallings Island Open grassland, dense 3 ft tall grass, briars, grey sandy loam soil, on the first terrace of the river 19 NE of Stallings Island Open grassland, on a ridge top und er power lines, recently mowed with high grass around base of power line poles, approximately 25 meters southeast of the woods. Considered as cleared upland habitat. 20 NE of Stallings Island Open grassland, on ridge with slight slope under the power line s 10 inch high thick grass cover, not mowed as recently as sample 19. Table 4 4 Descriptions of each modern sample categorized as a wooded habitat Sample Location Description 5 SE of Stallings Island Hardwood forest with large oaks, elms, sweet gums and a couple of palmettos. Located on a natural levee beside the river. 6 SE of Stallings Island Hardwood forest with large oaks, elms, sweet gums and a couple of palmettos. Located on a natural levee beside the river. 7 Stallings Island Hardwood forest with thick ground cover. To the north of power lines. 9 Stallings Island Hardwood forest with a large number of logs and debris on the ground. 10 Stallings Island Hardwood forest with a large number of logs and debris on the ground. Located on the main shell mound. 11 Stallings Island Hardwood forest with a large number of logs and debris on the ground. Located on the main shell mound. 14 NE of Stallings Island Hardwood forest with limited ground vegetation/cover. Located on the first terrace of the r iver. 17 NE of Stallings Island Upland mixed hardwood forest with several clumps of 10 inch high grass on the ground. Located on a high ridge.

PAGE 76

76 Table 4 5 Historical d ata on s nail h abitat among taxa f rom Hubricht (1985) (H), Pilsbry (1939 1948) (P), Nekola and Coles (2010) (N), and La Rocque (1970) (L) Snail All Dry Wet Wooded Leaf Litter Urban Open/ Sunny Grassy Roadsides Anguispira alternata HNL L Anguispira fergusoni HN Carychium exile L HL Carychium exiguum HN X HN (logs ) L Columella simplex P P HP HP P Discus patulus HN (logs) L H X Euconulus chersinus H (moist) Gastrocopta armifera HP H H G. contracta H X H H HP HP H G. corticaria N G. pellucida P H P G. pentadon P HP X P HP P G. riparia G. rupicola HP H G. tappaniana H P Glyphyalinia cryptomphala H (moist) G. indentata H G. wheatleyi N H (moist) N Haplotrema concavum N H Hawaiia minuscula N (decay) N (damp) H H Helicodiscus parallelus HL L H H Inflectarius inflectus H H H H H Lobosculum pustuloides H (rocky) N H Lucilla sp. Mesodon clenchi

PAGE 77

77 Table 4 5 Con t inued Snail All Dry Wet Wooded Leaf Litter Urban Open/ Sunny Grassy Roadsides Mesodon thyroidus HN H H H H H Mesomphix andrewsae H (moist) Mesomphix globosus H H (wet) Mesomphix pilsbryi H H H Paravitrea capsella Patera appressus H N H (Rocky) Patera perigrapta H X N Pract icolella jejuna HN N H Punctum minutissimum H (deep) N Punctum smithi H (deep) Pupisoma sp. HP Pupoides albilabris P HL PL H Retinella electrina Stenotrema stenotrema H H H H H Striatura meridionalis H Strobilops labyrinthica HN HN Triodopsis dentifera T. hopetonensis H H (wet) H (wet) T. vannostrandi H (drier) H H Ventridens cerinoides H H (under) V. demissus H V. gularis HL H V. inertextus H (acidic) H Vertigo oscariana P V. rugosula H HP P

PAGE 78

78 Table 4 5 Con t inued Snail All Dry Wet Wooded Leaf Litter Urban Open/ Sunny Grassy Roadsides V. clappi H V. tridentata P H PN Xolotrema caroliniense H H Zonit idae L Zonitoides arboreus Anywhere shaded (N) N m oist (N) N N Z. lateumbilicatus H (cool) Z. patuloides H (moist)

PAGE 79

79 Table 4 6 Descriptive statistics of each tax on identified in the modern samples across the samples Taxon N MEAN MEDIAN MIN MAX SD Anguispira alternata 1 Carychium exiguum 14 14 14 14 14 Discus patulus 1 Euconulus chersinus 52 6.5 5.5 1 23 7 Gastrocopta species 9 4.5 4.5 4 5 0.7 Gastrocopta armifera 2 1 1 1 1 0.0 Gastrocopta contract a 93 18.6 16 2 45 15.9 Gastrocopta pentadon 65 9.3 6 1 25 9.3 Gastrocopta rupicola 39 9.8 5.5 2 26 11 Gastrocopta tappiania 1 Glyphyalinia species 150 12.5 10 2 39 10 Glyphyalinia cryptomphala 26 8.7 9 6 11 2.5 Haplotrema concavum 3 1 1 1 1 0 Hawaiia minuscula 392 65.3 30.5 1 257 96.3 Helicodiscus species 2 2 2 2 2 Helicodiscus parallelus 67 8.4 7 1 15 4.9 Inflectarius inflectarus 5 5 5 5 5 Mesodon thyroidus 5 5 5 5 5 Mesomphix species 3 1.5 1.5 1 2 0.7 Mesomphix andrewsae 1 Mesomphix globosus 4 1.3 1 1 2 0.6 Polygyra pustuloides 2 2 2 2 2 Practicolella jejuna 1 Punctum species 54 9 4 2 29 10.4 Pupillidae 46 9.2 11 4 13 4 Pupisoma species 43 14.3 11 4 28 12.3 Pupisoma discoricola 1 Striatur a meridionalis 9 9 9 9 9 Ventridens species 1 Ventridens demissus 2 2 2 2 2 Vertigo species 1 Vertigo clappi 4 4 4 4 4 Vertigo oscariana 1 Vertigo pygmea 1 Vertigo rugosula 29 9.7 1 1 27 15 Vertigo trid entata 15 3.8 3 1 8 3 Zonitidae 100 100 100 100 100 Zonitoides arboreus 2 1 1 1 1 0

PAGE 80

80 Table 4 7. List of total individual snails and percentages within open and wooded habitat samples in the modern samples Species Open Habitat Samples (n) Percent Wo oded Habitat Samples (n) Percent Anguispira alternata 0 0% 1 100% Carychium exiguum 0 0% 14 100% Discus patulus 0 0% 1 100% Euconulus chersinus 13 25% 39 75% Gastrocopta species 5 55.6% 4 44.4% Gastrocopta armifera 2 100% 0 0% Gastrocopta contracta 47 50.5% 46 49.5% Gastrocopta pentadon 33 50.8% 32 49.2% Gastrocopta rupicola 37 94.9% 2 5.1% Gastrocopta tappaniana 1 100% 0 0% Glyphyalinia species 40 26.7% 110 73.3% Glyphyalinia cryptomphala 0 0% 26 100% Haplotrema concavum 0 0% 3 100% Hawaiia minuscula 99 25.3% 293 74.7% Helicodiscus species 2 100% 0 0% Helicodiscus parallelus 26 38.8% 41 61.2% Inflectarius inflectus 0 0% 5 100% Mesodon thyroidus 0 0% 5 100% Mesomphix species 0 0% 9 100% Mesomphix andrewsae 0 0% 1 100% Mesomphix globosus 0 0% 4 100% Polygyra pustuloides 2 100% 0 0% Practicolella jejuna 0 0% 1 100% Punctum species 8 14.8% 46 85.2% Pupillidae 27 58.7% 19 41.3% Pupisoma species 32 74.4% 11 25.6% Pupisoma discoricola 0 0% 1 100% Striatura meridionalis 0 0% 9 100% Vent ridens species 1 100% 0 0% Ventridens demissus 0 0% 2 100% Vertigo species 0 0% 1 100% Vertigo clappi 4 100% 0 0% Vertigo oscariana 0 0% 1 100% Vertigo pygmea 1 100% 0 0% Vertigo rugosula 29 100% 0 0% Vertigo tridentata 9 60% 6 40% Zonitidae 0 0% 1 00 100% Zonitoides arboreus 2 100% 0 0%

PAGE 81

81 Table 4 8. Total number of individuals of the snail s (n= > 15) within taxa in each of the modern samples Tax on (n) Sample Euconulus chersinus Gastrocopta contracta Gastrocopta pentadon Gastrocopta rupicola Glyphyalinia cryptomphala Glyphyalinia species Hawaiia minuscula Helicodiscus parallelus Punctum species Pupillidae Pupisoma species Vertigo rugosula Vertigo tridentata Zonitidae Open Habitats: 8 6 2 6 26 0 2 0 6 4 0 0 0 0 0 12 7 45 25 0 0 14 10 15 4 12 28 1 8 0 15 0 0 0 0 0 12 0 5 0 0 0 0 0 0 16 0 0 1 0 0 12 26 0 0 11 0 0 0 0 19 0 0 0 5 0 0 0 0 0 0 0 1 0 0 20 0 0 1 6 0 0 63 0 0 4 4 27 1 0 Wooded Habitats: 5 0 0 2 0 0 5 1 11 12 0 0 0 0 0 6 1 0 0 0 0 8 0 7 0 0 0 0 0 0 7 5 0 0 0 0 14 257 7 29 13 0 0 0 100 9 2 13 17 0 11 23 35 15 0 6 11 0 3 0 10 6 16 0 2 0 6 0 0 3 0 0 0 0 0 11 23 17 13 0 0 39 0 1 2 0 0 0 3 0 14 2 0 0 0 6 7 0 0 0 0 0 0 0 0 17 0 0 0 0 9 8 0 0 0 0 0 0 0 0 Total (n): 52 93 65 39 26 150 392 67 54 46 43 2 9 15 100

PAGE 82

82 Table 4 9. Total number and percentage s of snail taxa with open and wooded habitat tendencies, excluding Hawaiia minuscula in the modern samples Open Taxa (n) Percent Wooded Taxa (n) Percent Open Habitats 134 60.6% 87 39.4% Wooded Habitat s 38 9.5% 362 90.5% Table 4 1 0 Chi square results comparing the association of open and wooded taxa found within open or wooded habitats, excluding Hawaiia minuscula in the modern samples Open Taxa (n) Wooded Taxa (n) Total DF X 2 (Yates) X 2 (Pearson) P Value Open Habitats 134 87 22 1 1 183.3 185.9 <0.0001 Wooded Habitats 38 362 400 Total 172 449 621 Table 4 1 1 Total number and percentages of s nail s within the families Pupillidae and Zonitidae with open and wooded habitat tendencies in th e modern samples Species Open Habitat Samples (n) Percent Wooded Habitat Samples (n) Percent Pupillidae 227 64. 9 % 123 35. 1 % Zonitidae 1 42 2 2.7 % 483 76.7 % Table 4 1 2 Chi square results comparing the association of snail taxa in the families Pupillidae and Zonitidae within open or wooded habitats in the modern samples Pupillidae (n) Zonitidae (n) Total DF X 2 (Yates) X 2 (Pearson) P Value Open Habitats 227 1 42 369 1 167.6 169.4 <0.0001 Wooded Habitats 123 4 83 606 Total 350 625 975

PAGE 83

83 Table 4 1 3 Total number and percentage s of s nail s with open and wooded habitat tendencies excluding Hawaiia minuscula among the modern samples Modern Samples Open Taxa ( n) Percent Wooded Taxa (n) Percent Open Habitats: 8 26 59.1 % 18 40.9 % 12 49 55.1 % 40 44.9 % 15 0 0.0% 17 100% 16 11 47.8 % 12 5 2.2 % 19 6 100% 0 0.0% 20 42 100% 0 0.0% Wooded Habitats: 5 0 0.0 % 28 100 % 6 0 0.0% 16 100% 7 13 7.7% 155 92.3% 9 20 28.2 % 51 71.8 % 10 2 11.8 % 15 88.2 % 11 3 4.4 % 65 95.6 % 14 0 0.0% 15 100% 17 0 0 .0% 17 100% Table 4 1 4 Soil volume and total number of snails in each of the modern samples Modern Samples Soil Volume (L) Total Number (n) Open Habitats: Average: 4.6 Total: 420 8 5.9 61 12 4.8 171 15 4.1 17 16 3.0 53 19 5.2 11 20 4.8 107 Wood ed Habitats: Average: 5.5 Total: 827 5 5.2 42 6 5.4 19 7 5.2 430 9 6.5 139 10 5.5 55 11 4.6 100 14 5.0 17 17 6.6 25

PAGE 84

84 Table 4 1 5 Sna il density per liter of soil in each of the modern samples and in the taxa with open and wooded habitat tendencie s Modern Samples Total Snail Density Per Liter of Soil Total Open Taxa Snail Density Per Liter of Soil Total Wooded Taxa Snail Density Per Liter of Soil Open Habitats: 8 10.3 4.4 3.1 12 35.4 10.2 8.3 15 4.2 0. 0 4. 1 16 17. 7 3.7 4.0 19 2.1 1.2 0. 0 20 2 2.3 8.8 0.0 Average: 15.4 4.7 3.3 Wooded Habitats: 5 8.1 0.0 5. 4 6 3.5 0.0 3.0 7 82.7 2.5 29.8 9 21.4 3.1 7.8 10 10.0 0.4 2.7 11 21.7 0.7 14.1 14 3.4 0.0 3.0 17 3.8 0.0 2.6 Average: 19.3 0.8 8.6

PAGE 85

85 Figure 4 1. Google Earth i mage showing the collection sites for 16 of the 20 modern samples ( 5 to 20 )

PAGE 86

86 Figure 4 2 Contour map of the Savannah River bank to the northeast of Stallings Island showing the locations of modern samples 14, 15, 16, 17, 19 and 20

PAGE 87

87 Figure 4 3 Googl e E arth image of Savannah River bank to northeast of Stallings Island show ing the locations of modern samples 14, 15, 16, 17, 19 and 20

PAGE 88

88 Figure 4 4 Contour map of Stallings Island site showing the location s of modern samples 7, 8, 9, 10, 11 and 12

PAGE 89

89 F igure 4 5 Google E arth image of Stallings Island site showing the location s of modern samples 7, 8, 9, 10, 11 and 12

PAGE 90

90 Figure 4 6 Contour map of Savannah River bank to southeast of Stallings Island showing the location s of modern samples 5 and 6

PAGE 91

91 Figure 4 7 Google E arth image of Savannah River bank to southeast of Stallings Island showing the location s of modern sample s 5 and 6

PAGE 92

92 Figure 4 8. Average soil volume among the modern s amples within the open and wooded habitat samples Figure 4 9 The average total snail density per l iter of soil with in the open and wooded habitat samples

PAGE 93

93 Figure 4 10 The average snail densi ty per l iter of soil between snails with open or wooded hab itat tendencies within the open and wooded habitat samples

PAGE 94

94 CHAPTER 5 COLLECTION METHODS FOR THE ARCHAEOLOGICAL SAMPL ES Field Procedures Stallings Island As discussed in Chapter 1, Stallings Island has been excavated on numerous occasions beginning in the 1850s. In addition to previous scientific excavations, the site was damaged by pot hunting and illicit digging until the Archaeological Conservancy acquired it in 1999 and began to protect the site from further damage (Sassaman 2002). In 1999, funding from the National Geographic Society along with permission from the Archaeological Conservancy provided the opportunity for a team from the University of Florida to clear, map and test Stallings Islan occupational history through radiometric dating (Sassaman 200 2 ). A total of 195 mapped on Stallings Island. T est excavations were conducted in several of the which were mostly lo cated in the area of occupation at the crest of the island, as well as along the northern sideslope of the site, where secondary midden accumulated As a result, over 50 square meters of test excavations revealed 56 intact pit features and stratified shel l midden with a three meter deep sequence (Sassaman 2002). The latter deposit was exposed in the excavation of Looter Pit 81 (LP81), a 2 x 2 m unit sited to maximize the vertical profile of the midden. The Cosgroves dug a trench through this area during their 1929 expedition, but concluded that shell midden in this downslope portion of the site was redeposited by histori c era flood events (Claflin 1931 ). However, recent looting had exposed what appeared to be intact, stratified deposits. In 1991, testi ng in LP81 revealed that the Cosgroves were only partially

PAGE 95

95 correct; the upper 0.7 1.3 m of deposition in this location was indeed redeposited by alluvial forces, but the lower 2+ meters was fully intact and all dated to the prepottery era (late Paris Islan d and early Mill Bra n ch phases, ca. 4400 4200 B.P., or ca. 4900 4700 cal B.P.) (Sassaman et al. 2006) The lower 2 m of shell midden exposed in LP81 provides the best opportunity to examine changes in ground cover via changing frequencies of terrestrial s nails. Bulk soil samples taken from the lower meter of intact midden contained an abundance of snail shells with a wide range of diversity. The grid west profile of LP81 is illustrated in Figure 5 1. The upper 0.7 1.3 m of this profile is characterized b y cross bedded and contorted strata with alternating sandy clay, silty loam, and shell. Evidently, this was the result of recent erosion and redeposition, which contrasts sharply with the strata below. At approximately 70 cm below the surface at the sout h corner, the upper unit transitions to alternating layers of crushed and whole shell with occasional clay stringers and charcoal lenses, all with a consistent dip and strike that follows the surface contours of the side slope. It appeared the lower 2.3 m of stratified midden was unaffected by the erosional forces ( flood ing ) that removed an unknown amount of near surface midden and left a scoured surface at 70 cm, upon which midden eroded from upstream was redeposited. Once it was established that the low er portion of the profile was unaffected by erosion, a one meter wide pedestal was left unexcavated. Upon completion of the excavation, sediment from this pedestal was removed entirely by stratigraphic units and transported to the University of Florida fo r water screening and flotation processing. Material suited to radiometric dating (shell and charcoal), artifacts, plant remains, vertebrate fauna, and invertebrates, including massive quantities of terrestrial snail shells, were recovered and

PAGE 96

96 comprise on e of the best assemblages of prepottery era materials at Stallings Island. Because the deposits of LP81 accumulated on the margins of habitation space at Stallings Island, they are assumed to consist of secondary midden, that is, refuse that was collected terrestrial snails in the strata of LP81 include not only those that colonized an ever accumulating midden of shell and other material with other ref use and deposited along the sideslope. In short, the terrestrial snail shells of LP81 arguably consist of an assemblage indicative of ambient conditions across the site. Excavations in the habitation area atop Stallings Island provided additional contexts for the pre pottery components of the site and all of the contexts of the Classical Stallings phase (ca. 4100 3800 cal B.P.) (Sassaman et al. 2006) All contexts of recovery in the habitation area consist of pit features, some with complex stratified sequ ences. Like the deposits of LP81, the fill of pits likely includes secondary refuse that was collected from the general habitation area and buried, but also primary deposits that were emplaced directly from nearby activities, such as food processing and c onsumption. In either case, terrestrial snails also likely entered pits incidentally from the immediate area. Pits were typically cross sectioned in the field, with fill from the first half processed by 1/8 inch water screening, and fill from the second h alf retrieved for flotation processing. Most features contained abundant freshwater shell, wood charcoal, vertebrate fauna, artifacts, and the shells of a variety of terrestrial snails.

PAGE 97

97 Seven radiocarbon assays were obtained from the lower strata of LP81 and published by Sassaman et al. (2006) These confirm that the intact midden of LP81 predates the local inception of pottery by several centuries An additional 15 assays from pit features enable a reasonably sound reconstruction of the occupational seq uence for Stallings Island through the mid fourth millennium B.P. Thirteen of the assays were obtained from freshwater shell, one from unidentified charred material and the remainder from wood charcoal and/or charred hickory nutshell. All assays, except o ne, fall in the Late Archaic period and all but two other assays cluster into two discrete groups: 1) thirteen assays ranging ca. 4600 4000 B.P. or 5300 4500 cal B.P.; and 2) six assays ranging ca. 3800 3400 B.P. or 4200 3600 cal B.P. (Sassaman et al. 2006 ). The dates coincide with Classic Stallings, Early Stallings, and Pre Ceramic cultures and are listed with the corresponding strata on Table 5 1. The location of the strata and the corresponding radiocarbon dates within LP81 are illustrated on Figure 5 1. The assays ranging from 4600 4000 B.P. (5300 4500 cal B.P.) reflect a relatively continuous deposition of the materials sampled and are evenly divided between pit features and the strata of LP81 (Sassaman et al. 2006) Stratum VI and the overlying St ratum V were located at the base of the LP81 column and the assays provided a date of 4550 4450 B.P., which establishes the onset of shell midden formation shortly after 4500 B.P. (Sassaman et al. 2006). A second assay from Stratum VI suggests the midden formation may have started a few centuries earlier. However, the layer was a saprolitic clay of sterile substrate, which is typically exposed in the slopes of landforms eroded by the river and its tributaries. This suggests the surface of

PAGE 98

98 this stratum ma y have been exposed to a number of organic inputs before the shell began to accumulate. Four assays collected from the overlying strata of LP81 center on ca. 4200 B.P. and represent the rapid accumulation of 1.5 m of shell rich midden (Sassaman et al. 2006 ) No pottery was observed in this undisturbed context. A series of four pit features in the habitation area have assays ranging from ca. 4300 to 4000 B.P. (Sassaman et al. 2006). All but one of the features lacked pottery. The presence of the pottery in the one pit is indicative of an intrusion from an older occupation, if the radiometric assay is taken at face value. Six as says from five pit features of the habitation area range from 3800 to 3400 B.P., which occurred during the well dated interval of Classic Stallings times (Sassaman et al. 2006) The dates correspond to numerous examples of drag and jab fiber tempered pottery and food remains. The Clas sic Stallings features reflect a more intensive use of large and deep p its over the earlier occupat ions. An approximately 200 year occupational hiatus occurred between ca. 4000 and 3800 B.P. and the early and late components at Stallings Island (Sassaman et al. 2006) The hiatus corresponds with the latter half of the Mill Branch and the Early Stalling s. Components of the latter phase were found at Victor Mills and Ed Marshall in the middle Savannah area and at numerous sites downstream from Stallings Island in the Coastal Plain and on the coast. Sassaman (2008) has argued that the presence of Early S tallings components in the middle Savannah area signifies a movement of some Coastal Plain communities into the middle Savannah region after ca. 4200 B.P.

PAGE 99

99 Ed Marshall Ed Marshall (38ED5) is a small shell bearing site located to the east of Stallings Isla nd on the South Carolina side of the Savannah River floodplain. Of the three sites used in this study, Ed Marshall is the only one to contain components of each of the three time periods in question (preceramic Late Archaic, specifically the Mill Branch phase; Early Stallings; and Classic Stallings) (Sassaman n.d.). It is also the only site to be situated at low elevation in the floodplain and is only one to two meters above the high water level of the impounded river. Water levels before the Stevens Cr eek dam was constructed were considerably lower than today and the site itself is elevated about anthropogenic deposits. Judging from the large number of artifacts, features, and food remains that accumulated, Ed Marshall was clearly a locus of intensive habitation throughout the Late Archaic period, but it was also likely to be abandoned during periods of excessive rainfall that would have caused the river to spill out over its r elatively low lying, natural levees. Even with damming and modern water controls, Ed Marshall is subject to periodic flooding; a flood chute, which cut through the site sometime in the middle part of the last century, is still visible on the surface in th e southeastern end of the site. In addition to flood damage, Ed Marshall has been subjected to moderate looting in recent decades. In 1994, archaeologists of the Savannah River Archaeological Research Program (SRARP) excavated two small test units in resp onse to looting at Ed Marshall (Sassaman 1996). The excavation unearthed features including post holes, shell filled pits and a hearth. Additional funds were secured in August 1995 and the SRARP team returned to excavate two stratigraphic trenches, a 5 b y 5 m block and several isolated

PAGE 100

100 tests, which totaled 41 square meters (Sassaman 1996). The following year, a field school administered through Augusta State University partnered with SRARP staff to expand operations with another small block excavation an d several additional test units. A final report of the combined three seasons of work at Ed Marshall has not yet been issued, but collections and field records curated at the Laboratory of Southeastern Archaeology, University of Florida are available for study. For the present research, several of the pit features excavated at Ed Marshall were analyzed for terrestrial snails. Features of the preceramic Mill Branch phase consisted of shallow basins filled with very dark organic soil, occasional shell, vert ebrate fauna, charcoal, and abundant artifacts. One such feature, Feature 30, provided charcoal that returned a radiocarbon assay of 4140 + 90 B.P. (Table 5 2 ) (Sassaman n.d.). Another such feature, Feature 25, returned a radiocarbon assay of 4090 + 80 B.P. (Table 5 2 ) (Sassaman n.d.). Associated post holes and other features indicative of architecture suggest that Ed Marshall was a habitation location during the Mill Branch phase. Unfortunately, only Feature 30 provided an adequate sample of terrestrial sn ails for analysis. Features of the succeeding Early Stallings phase consisted of a wider array of types compared to the preceramic period, which included one, or possibly two, fired clay floors of presumed habitat structures. Features sampled for terres trial snails consisted of relatively large shell filled pits in association with fired clay floors. Two such features were sampled for radiocarbon dating: Feature 22, which returned an assay of 3840 + 90 B.P., and Feature 24, with an age estimate of 3800 + 60 B.P. (Table 5 2 ) (Sassaman n.d.). Both of these features, along with two features lacking radiocarbon

PAGE 101

101 assays and containing diagnostic pottery, Features 8 and 17, provided adequate samples of snails to include in this analysis. Features dating to the Cla ssic Stallings period were likewise diverse, but generally smaller than those of the Early Stallings period. One such feature provided charcoal for radiocarbon dating, Feature 37b, which returned an age estimate of 3560 + 60 B.P. (Table 5 2 ) (Sassaman n.d.) This feature provided a good sample of terrestrial snails, as did two other features (Features 55 and 56), which contained diagnostic Classic Stallings pottery, but no radiocarbon assays. The site wide distribution of Classic Stallings pottery at Ed Ma rshall suggests strongly that the site was the locus of a circular village plaza complex similar to the ones observed at Mims Point ( Figure 5 2 ) and inferred at Stallings Island. Mims Point Mims Point (38ED9) is a multicomponent site on a ridge nose at the confluence of Stevens Creek and the Savannah River, approximately one kilometer north of Stallings Island (Sassaman 1993b). Although Mims Point has a Late Woodland component dating to the 14 th millennium B.P., as well as a Middle Archaic component dating to ca. 5700 B.P., its primary component dates to the Classic Stallings phase, ca. 3700 3500 B.P. (Table 5 3 ) (Sassaman n.d.). Mims Point has expressed the greatest clarity of internal site structure observed in the three sites of this study, owing in lar ge measure, to the lack of Early Stallings and preceramic Late Archaic components. Inferred from an assemblage of dozens of pit features and hearths is a pattern of feature clusters indicative of houses arranged in a circular fashion around a relative fea ture free interior zone, all ab out 30 m in diameter (Figure 5 2 ). The Classic Stallings settlement at Mims Point has been heralded as the best example of a circular village plaza complex in the

PAGE 102

102 region and has been used to infer similar arrangements at bot h Stallings Island and Ed Marshall (Sassaman 1996; Sassaman et al. 1996). Excavations at Mims Point were undertaken in three phases from 1992 1995 by staff of the Savannah River Archaeological Research Program (SRARP), in cooperation with the U.S. Forest S ervice, which owns the site. The initial field efforts in 1992 included site mapping, extensive subsurface coring, limited testing of undisturbed areas of the site, and a 68 m 2 block excavation of disturbed midden (Sassaman 1993b). Sassaman (1993b) uncov ered over two dozen prehistoric features, including human burials, refuse filled pits, postholes, rock clusters, and a possible hearth during the 1992 block excavation. Subsequent excavations expanded the initial block to open up an additional 298 m 2 to r eveal a large portion of the circular village, which was mostly undisturbed. A plan map of Mims Point showing all of the features excavated is shown in Figure 5 3 As noted, features tend to cluster into groups believed to be indicative of households arra nged in a circle around a central, open plaza, some 15 16 m in diameter. Pit clusters include large storage pits, shallow basins, hearths, and occasional postholes, although the lowest of the site precludes preservation of features less than ~30 cm in dep th. Despite the lack of definitive architectural features, the overall plan of pits, hearths, and other features provides strong indirect evidence of a circular community plan of contemporaneous households. Eight radiocarbon assays were obtained on wood c harcoal or charred hickory nutshell from six of the Classic Stallings features. With one exception, the assays fall in a two century span of 3700 3570 B.P. (Sassaman n.d.). The exception is Feature 1, a

PAGE 103

103 human burial containing Classic Stallings pottery, but dating to 4025 + 65 B.P. (Sassaman n.d.). and drop punctate, the hallmark of Classic Stallings culture. An age estimate thi s early is not unusual for the coast of South Carolina and Georgia, the ultimate heart of Stallings Culture, although not widely expressed in the middle Savannah until after ca. 3750 B.P. (Sassaman n.d.). Nonetheless, the presence of drag and jab pottery places Feature 1 in the Classic Stallings milieu, and thus, is included in this study. The other features included in this study involve pits (Features 51 and 52) and hearths (Feature 66) of the household clusters, plus one pit feature (Feature 50) locate d in the presumed central plaza. For the purpose of this study, the features from Mims Point analyzed for terrestrial snails are considered to be contemporaneous, with the exception of the burial (Feature 1) which is likely to be on the early end of the s equence and thus indicative of early or pre village configuration. Laboratory Procedures The samples were collected and transported to the University of Florida for further analysis. The samples were measured for volume and then processed through a waters creen or a flotation machine, fractionated into three size grades (>1/4 in, >1/8 in, and <1/8 in), and sorted into their constituent components. All mollusks shells were removed with the aid of magnification, strong light and forceps or small paint brushe s. The shells were then stored in either glass vials or plastic zip top bags. Each container was clearly labeled with the sample information, screening method and initial weight. As a portion of this research, the snails were observed on a white backgr ound and separated by hand or with a small paint brush based on observable characteristics.

PAGE 104

104 Some of the snails were thoroughly cleaned in order to visualize the internal structures within the aperture. Identifiable snails were then analyzed using a stron g light source and a microscope at 5X to 40X magnification, when necessary. The snails were counted using whole shells or fragments bearing individualizing characteristics. If multiple fragments were present from one apparent species, then it would only be counted as one individual; only fragments from the same area were counted as multiple individuals. Fragments that presented no individualizing characteristics were not counted and separated into container labeled as unidentified. The snails were ident ified based upon taxonomy guides and geographical ranges in Pilsbry (1939 1948), Nekola and Coles (2010) and Hubricht (1985). In addition, snails previously identified by Evan Peacock and Jochen Gerber were used for reference. Each container was clearly marked with the sample data and identified taxon information. The total number of individual shells within each taxon identified per site was recorded. Nomenclature follows that of Hubricht (1985) and Nekola and Coles (2010).

PAGE 105

105 Table 5 1 Radiocarbon dates for the strata in LP81 at Stallings Island (Sassaman et al. 2006) Stratum Measured C14 Age (B.P.) Calibrated C14 Age (cal B.P.) IA 4210 + 60 ca. 4700 IB / IC 4140 + 70 IIIB 4210 + 60 IV 4260 + 70 ca. 4900 V 4370 + 70 VI 4360 + 70; 4830 + 70 Table 5 2 Features from Ed Marshall and the corresponding radiocarbon dates and periods at Ed Marshall (38ED5) (Sassaman n.d.) Period 38ED5 Feature C14 Dates PC 30 4140 + 90 ES 8 ES 17 ES 22 3840 + 90 ES 24 3800 + 60 CS 37b 3560 + 60 CS 55 CS 56 Table 5 3 Radiocarbon dates for the features from Mims Point (38ED9) (Sassaman n.d.) Period 38ED9 Feature C14 Dates (B.P.) CS 1 4025 + 65 CS 50 CS 66 3700 + 150 CS 51 3580 + 60 CS 52

PAGE 106

106 Figure 5 1. West profile of LP81 sh owing the strata and corresponding radiocarbon dates (Sassaman n.d.)

PAGE 107

107 Figure 5 2 (38ED9), showing projected pit feature clusters of a circular domestic compound with a central are a largely devoid al. 2006)

PAGE 108

108 Figure 5 3 Plan distribution showing the pit features excavated at Mims Point (38ED9), which specifies the location of Features 1, 50, 51, 52 and 66 (Sassaman n.d.)

PAGE 109

109 CHAPTER 6 RESULTS OF THE A NALYSIS OF ARCHAEOLOGICAL SNAIL SAMPLES A total of 31,163 terrestrial snails and 3,027 aquatic snails were identified from the archaeological samples in this study. To compare snails collected from similar collection techniques all of the samples without snails collected in the light and heavy fractions were eliminated for the final analysis. These were removed in order to only consider the samples with a definitive representation of all the snails present in order to reduce the bias associated with only locating snails in one fraction. Samples with ambiguous ages were also eliminated for the final analysis. Stallings Island A total of 11,932 terrestrial snails were identified into 19 taxa from the samples recovered from LP81 (Table 6 1 ). The taxa inc lude two families (Discidae and Pupillidae), seven genera ( Anguispira sp., Discus sp., Gastrocopta sp., Glyphyalinia sp., Lucilla sp., Mesodon sp., and Pupisoma sp.), and ten species ( Anguispira alternata, Discus patulus, Euconulus chersinus, Gastrocopta c ontracta, Gastrocopta rupicola, Glyphyalinia indentata, Haplotrema concavum, Hawaiia minuscula, Helicodiscus parallelus, and Xolotrema caroliniense ). Descriptive statistics for each taxon within the strata are shown on Table 6 2 The total number of snai ls within a taxon range from 11 individuals (mean = 1.4) ( Anguispira sp.) to 2166 individuals (mean = 270.8) ( Discus patulus ). The minimum number of individuals range from 0 (multiple taxa) to 117 ( Discus patulus ) and the maximum number of individuals ran ge from 4 ( Xolotrema caroliniense ) to 595 ( Glyphyalinia sp.). The standard deviations range from 1.8 ( Mesodon sp. and Xolotrema concavum ) to 209.5 ( Glyphyalinia sp.).

PAGE 110

110 The taxa found in LP81 were compared to the results of the modern snail analyses in Chap ter 4 and previous research (Table 4 5). A total of five taxa were considered as having open habitat tendencies and a total of five taxa were considered as having wooded habitat tendencies (Table 6 3 ). The remaining taxa did not show a tendency for eithe r habitat or were considered generalists (Table 6 3 ). The taxa designated as having open habitat tendencies include: Euconulus chersinus, Gastrocopta sp., Gastrocopta contracta, Lucilla sp., and Pupillidae (Table 6 3 ). The taxa designated as having woode d habitat tendencies include: Discus patulus, Glyphyalinia sp., Haplotrema concavum, Helicodiscus parallelus, and Pupisoma sp. (Table 6 3 ). The total number of individual snails and percentages of snails within the taxa with open and wooded habitat tende ncies within each stratum are listed on Table 6 4 and illustrated on Figure 6 1 Strata VI, IV, IIID, IIIC, and IIIA were found to have more snails with wooded habitat tendencies (83.8%, 93.1%, 67.5%, 64.0%, and 98.5%, respectively) than snails with open habitat tendencies. Strat um ID had approximately the same percentage of snails with open and wooded habitat tendencies (44.1% and 55.9%, respectively). In contrast, Strata IC and IB had more snails with open habitat tendencies (69.9% and 72.1%, respectiv ely) than snails with wooded habitat tendencies. The percentages show a shift in habitat tendencies among the taxa between Strata ID and IC (Figure 6 1 ). An observable disturbance ( S tratum II) between Stratum IIIA and S tratum ID can be observed in Figure 5 1. Due to the disturbance and smaller layer S tratum IIIA was removed. Once S tratum IIIA was removed, a linear relationship was observed in the percentages of taxa with open and wooded habitat tendencies from S tratum IV to

PAGE 111

111 S tratum IB (Figure 6 2 ). The linear relationship clearly indicates a shift from taxa with more wooded habitat tendencies to taxa with more ope n habitat tendencies over time (Figure 6 2 ). The habitat tendencies were analyzed further for the s trata in LP81. Summary statistics (total n umber, means, and standard deviations) for the taxa with open and wooded habitat tendencies for each stratum are listed on Table 6 5 and illustrated on Figure 6 3 The statistical assessment of nonrandom tendencies in archaeological data is intrinsically difficult due to common archaeological excavation methods in both the in field collection of snail assemblages in sample plots and the recovery of archaeological samples in bulk, which result in dependent data. Regardless of the dependent collection meth ods, the associations of particular taxa within the spatial units of collection are considered meaningful when they are distributed across sample units nonrandomly. Therefore, Chi square tests were used to determine the strength of associations between th e habitat tendencies and habitat types. The dependent sampling techniques certainly skewed the results of the Chi square test to some degree, but the results clearly indicate a significant association. The Chi square test was used to evaluate nonrandom a ssociations between habitat tendencies of the taxa and the strata in LP81 and resulted in a Chi square value of 1405.973 (df=7) an d a p value of <0.0001 (Table 6 6 ), which suggests an extremely significant association. Regardless of the dependent sampling method concerns, nonrandom associations among habitat tendencies, habitat types, and ultimately strata (a measure of time and change) are evident in the numerical data alone, which is verified by the Chi square

PAGE 112

112 statistic even if its probability values can not be accepted uncritically due to the biases of dependent data. To further analyze the samples from LP81, the taxa within the S trata IIIA, IIIC and IIID were combined into S tratum III and the taxa within the S trata IB, IC and ID were combined into S tratu m I. Strata III and I were compared to determine if larger scale changes in the taxa could be observed. Descriptive statistics for the taxa identified within S tratum III and S tratum I are listed in Tables 6 7 and 6 8 respectively. More snails were foun d in the taxa with wooded habitat tendencies in S trata VI (83.8%), IV (93.1%), and III (67%) ( Table 6 9, Figure 6 4 ). In contrast, more snails were found in the taxa with open habitat tendencies in S tra tum I (64.3%) ( Table 6 9, F igure 6 4 ). The shift in taxa from S tratum III to stratum I is also illustrated in Figure 6 5 Summary statistics (total number, means, and standard deviations) for the taxa with open and wooded habitat tendencies for S trata VI, IV, III, and I are listed on Table 6 10 and illustr ated on Figure 6 6 The Chi square test was used to evaluate nonrandom associations between habitat tendencies of the taxa and S trata VI, IV, III, and I in LP81 and resulted in a Chi square value of 1113.46 (df=3) and a p value of <0.0001 (Table 6 11 ), w hich suggests an extremely significant association. The Chi square test to evaluate nonrandom associations between habitat tendencies of the taxa and S trata III and I in LP81 resulted in a Chi square value of 745.6 (df=1) and a p value of <0.0001 (Table 6 12 ), which suggests an extremely significant association. N onrandom associations among habitat tendencies, habitat types, and ultimately strata (a measure of time and change) are evident in the numerical data alone, which is verified by the Chi square st atistic even if

PAGE 113

113 its probability values cannot be accepted uncritically due to the biases of dependent data. Ed Marshall A total of 2454 individual snail shells were identified from 8 of the 30 features (8, 17, 22, 24, 30, 37b, 55 and 56) (Table 6 13 ). Th e snails were identified as 38 different taxa, which included 3 families, 10 genera, and 25 species (Table 6 13 ). Only 15 of the 38 taxa contained 14 or more individuals. The taxa included 3 families (Discidae, Pupillidae, and Zonitidae), 5 genera ( Angui spira Gastrocopta Glyphyalinia Lucilla and Pupisoma ), and 7 species ( Anguispira alternata, Carychium exile, Discus patulus, Gastrocopta contracta, Gastrocopta pellucida, Hawaiia minuscula, and Helicodiscus parallelus ). Descriptive statistics for the t axa containing 14 or more individuals are shown on Table 6 14 The total number of snails within a taxon range from 14 ( Carychium exile and Pupisoma sp.) to 751 ( Discus patulus ) and the mean ranges from 1.8 ( Carychium exile and Pupisoma sp.) to 93.9 ( Disc us patulus ). The minimum number of individuals ranges from 0 (multiple taxa) to 8 ( Discus patulus ) and the maximum number of individuals ranges from 9 ( Gastrocopta sp.) to 298 ( Discus patulus ). The standard deviation ranges from 3.8 ( Pupisoma sp.) to 110 .7 ( Discus patulus ). Based upon the results of the modern snails in Chapter 4 and previous records (Table 4 5), three taxa from Ed Marshall were considered as having open habitat tendencies ( Gastrocopta rupicola Lucilla sp. and Pupillidae) and five taxa were considered as having wooded habitat tendencies ( Discus patulus, Glyphyalinia sp. Helicodiscus parallelus, Pupisoma sp. and Zonitidae) (Table 6 15 ). The taxa with open and wooded habitat tendencies were compared among the features within the

PAGE 114

114 Precera mic Mill Branch, Early Stallings and Classic Stallings phase s (Table 6 16 ). Even though the Preceramic Mill Branch phase only contained one feature (Feature 30), a total of 147 snails with open or wooded habitat tendencies were identified. Feature 30 had an almost equal number of snails in taxa with open and wooded habitat tendencies (55.8% and 44.3%, respectively ) ( Table 6 16, Figure 6 7 ). All of the features in the Early Stallings phase had more snails in taxa with wooded habitat tendencies than taxa w ith open habitat tendencies (100%, 8 1.5%, 97%, and 73.5%) ( Table 6 16, Figure 6 8 ). Similarly, Features 37b and 56 in the Classic Stallings phase also had more snails in taxa with wooded habitat tendencies than taxa with open habitat tendenc ies (84.2% and 62.5%) ( Table 6 16, Figure 6 9 ). However, Feature 55 in the Classic Stallings phase had more snails in taxa with open habitat tendencies than taxa with wooded habit at tendencies (63.1%) ( Table 6 16, Figure 6 9 ). Summary statistics (total number, mean, a nd standard deviation) for the taxa with open and wooded habitat tendencies in each feature from Ed Marshall are listed on Table 6 17 and shown on Figure 6 10 The Chi square test was used to evaluate nonrandom associations between habitat tendencies of the taxa and the features resulted in a Chi square value of 270.33 (df=7) and a p value of <0.0001 (Table 6 18 ), which suggests an extremely significant association. N onrandom associations among habitat tendencies, habitat types, and the features (a measu re of time and change) are evident in the numerical data alone, which is verified by the Chi square statistic even if its probability values cannot be accepted uncritically due to the biases of dependent data. The features within each period were combined and compared over time. The results of the preceramic Mill Branch phase are the same as above because it is only

PAGE 115

115 comprised of one feature, Feature 30. An increase in the taxa with wooded habitat tendencies was observed from the preceramic Mills Branch ph ase (44.2%) to the Early Stallings period (86.2%) (Table 6 19 Figure 6 11 ). A slight decrease in the taxa with wooded habitat tendencies was observed from the Early Stallings period (86.2%) to the Classic Stallings phase (74.3 %) (Table 6 19 Figure 6 11 ) Summary statistics (total number, mean, and standard deviation) for the taxa with open and wooded habitat tendencies from the three time periods from Ed Marshall are listed on Tabl e 6 20 and shown on Figure 6 12 The Chi square test was used to evalua te nonrandom associations between habitat tendencies of the taxa and the Preceramic Mill Branch and Early Stallings periods and resulted in a Yates corrected Chi square value of 220.52 (df=1) and a p value of <0.0001 (Table 6 21 ), which suggests an extreme ly significant association. The Chi square test was used to evaluate nonrandom associations between habitat tendencies of the taxa and the Early Stallings and Classic Stallings periods and resulted in a Yates corrected Chi square value of 27.59 (df=1) an d a p value of <0.0001 (Table 6 22 ), which suggests an extremely significant association. The Chi square test was used to evaluate nonrandom associations between habitat tendencies of the taxa and the three time periods and resulted in a Chi square value of 128.642 (df=2) and a p value of <0.0001 (Table 6 23 ), which suggests an extremely significant association. N onrandom associations among habitat tendencies, habitat types, and ultimately the time periods are evident in the numerical data alone, which is verified by the Chi square statistic even if its probability values cannot be accepted uncritically due to the biases of dependent data.

PAGE 116

116 Mims Point A total of 996 individual snails were identified from five features at Mims Point ( F eatures 1, 50, 51, 52 and 66) (Table 6 2 4 ) The snails were identified as 32 different taxa which included 3 families, 7 genera, and 22 species (Tabl e 6 2 4 ). Only 14 of the 32 taxa included 14 or more individual snails within each taxon. The 14 taxa included 1 family ( Pupil lidae ), 3 genera ( Glyphyalinia Lucilla and Ventridens ), and 9 species ( Anguispira alternata, Discus patulus, Gastrocopta contracta, Gastrocopta rupicola, Glyphyalinia indentata, Glyphyalinia wheatleyi, Hawaiia minuscula, Practicolella lawae and Triodops is vannostrandi ) (Table 6 2 4 ). Descriptive statistics for the taxa containing more than 14 individuals are shown on Table 6 25 The total number of snails within a taxon range from 16 (mean = 3.2) ( Gastrocopta contracta ) to 282 (mean = 56.4) ( Hawaiia min uscula ). The minimum number of individuals ranges from 0 (multiple taxa) to 7 ( Hawaiia minuscula ) and the maximum number of individuals ranges from 9 (SD = 4.4) ( Gastrocopta contracta ) to 163 (SD = 70.6) ( Discus patulus ). The taxa contai ning more than 1 4 individuals were compared to the results in Chapter 4 and previous research (Table 4 5). A total of three taxa ( Gastrocopta rupicola, Lucilla sp. and Pupillidae) were considered as having open habitat tendencies and a total of five taxa ( Discus patulus Glyphyalinia sp. Glyphyalinia wheatleyi, Practicolella lawae, and Ventridens sp. ) were considered as hav ing woode d habitat preferences (Table 6 26 ). The percentages of the individual snails within the taxa with open and wooded habitat tendencies were compar ed among the features (Table 6 27 ). Due to the limited number of snails recovered from feature 52 (n=30), no snails with either habitat tendency were identified (Table 6 27 Figure 6 1 3 ). Features 1, 51 and 66 had more

PAGE 117

117 individual snails in taxa wi th wooded habitat tendencies (92.3%, 90.5%, and 68.0%, respectively) (Table 6 27 Figure 6 1 3 ). In contrast, Feature 50 had more individual snails in taxa with open habit at tendencies (80.5%) (Table 6 27 Figure 6 1 3 ). Summary statistics (total number, m ean, and standard deviation) for the total number of snails within the taxa with open and wooded habitat tendencies in the features from Mims Point are listed on Table 6 28 and shown on Figure 6 1 4 The Chi square test was used to evaluate nonrandom asso ciations between habitat tendencies of the taxa and the features and resulted in a Chi square value of 248.31 (df=3) and a p value of <0.0001 (Table 6 29 ), which suggests an extremely significant association. N onrandom associations among habitat tendencie s, habitat types, and ultimately the features (across the site) are evident in the numerical data alone, which is verified by the Chi square statistic even if its probability values cannot be accepted uncritically due to the biases of dependent data. Discu ssion Stallings Island The terrestrial snails were compared before the disturbance event and through time in an attempt to interpret hunter gatherer environmental impact and the habitat types surrounding the midden (LP81). Based on the radiocarbon dating and stratigraphic contexts of the strata in LP81, it was concluded that the contents were deposited through time by the hunter gatherers and are representative of the ambient snail fauna. The ambient snail fauna could have accumulated by either accidental ly being transferred into the midden with other debris or through colonization of the midden through natural means. The results indicate a shift in habitat throughout time and among the s trata in LP81 (Figures 6 1, 6 2, 6 4, and 6 5). Strata VI, IV, and III

PAGE 118

118 were found to have more snails in the taxa with wooded habitat tendencies. The shift in snail preferences occurs between S trata III and I, with an intermediate habitat occurring in S tratum ID, and more snails with open habitat tendencies in S trata IC and IB (Figure s 6 1 and 6 2 ). The results of the snails collected from LP81 suggest a change occurred in the habitat in between S trata ID and IC. This corresponds to the disturbance visible in the west profile of LP81 on Figure 5 1. Whatever the distur bance was, it apparently resulted in a more open canopy environment, which allowed an increase in the number of snails that prefer or are able to withstand higher temperatures and less protected conditions. The shift in snail taxa supports evidence of a c hange in the ambient environment from more wooded conditions to open habitat conditions between S trata IIIA and IC possibly due to the disturbance evident in S tratum II. Ed Marshall The results of the archaeological snail faunas from Ed Marshall (38ED5) in dicate a shift in habitat between the Preceramic Mill Branch phase and the Early Stallings period. The shift to a more wooded habitat in the Early Stallings period supports previous evidence that the site may have had a decrease in human occupation during this time, possibly due to flooding which commonly occurred at the site due to its lower elevation. The decrease of snails in taxa with wooded habitat tendencies from the Early Stallings period to the Classic Stallings period supports a change in land us e at the site that led to more open habitat conditions. The indication of a decrease in shaded greater permanence during the Classic Stallings period (as shown by terre strial land snail evidence of increased land clearance for settlement) and thus his hypothesis of a

PAGE 119

119 formal circular village plaza complex similar to those observed at Mims Point (see Chapter 5) and inferred at Stallings Island (Sassaman 1996; Sassaman et a l. 1996). Mims Point As shown on Figure 6 15 Feature 1 was located in a burial along the edge of the block and Features 51, 52, and 66 were located along the edges of the block adjacent to houses within the circular ring. In comparison, Feature 50 was lo cated in a central pit near the center of the circular village (Figure 6 1 5 ). The archaeological snail fauna supported the interpretation of a central plaza village complex previously observed at Mims Point (Figure 6 1 5 ). A larger number of snails in tax a with open habitat tendencies were found in Feature 50 than in Feature s 1, 51, 52, and 66 (Table 6 27, Figure 6 13). The open habitat in the central plaza area may have been created by deforestation due to the creation of an open plaza for cultural pract ices or living structures and further cleared by trampling and use. The results support more wooded/covered habitat conditions surrounding the central plaza as observed at Features 1, 51, 52, and 66 (Figure 6 15). These results also support the circular village plaza complex concept because the areas surrounding the central plazas are projected to have a more shaded and protected environment due to an increase in the vegetation and the presence of living structures (Figure 6 15).

PAGE 120

120 Table 6 1 Total number of individual snails in each taxon within the strata of LP81 at Stallings Island Tax on IB (n) IC (n) ID (n) IIIA (n) IIIC (n) IIID (n) IV (n) VI (n) Total Anguispira alternata 88 46 78 43 67 106 100 53 581 Anguispira species 0 0 0 0 0 8 0 3 11 Disci dae 511 278 129 0 367 360 59 54 1758 Discus patulus 345 198 244 131 443 464 224 117 2166 Discus species 81 0 0 0 1 51 0 0 133 Euconulus chersinus 0 1 16 0 0 1 0 3 21 Gastrocopta contracta 420 374 228 1 394 331 11 26 1785 Gastrocopta rupicola 0 3 4 0 3 10 7 2 29 Gastrocopta species 414 2 5 0 0 0 0 2 423 Glyphyalinia indentata 1 0 1 0 0 2 8 1 13 Glyphyalinia species 172 144 315 0 496 595 75 133 1930 Haplotrema concavum 0 3 3 0 9 16 0 0 31 Hawaiia minuscula 6 58 252 0 263 123 102 258 1062 Helicodisc us parallelus 10 18 27 0 59 28 9 24 175 Lucilla species 58 0 0 0 8 0 0 0 66 Mesodon species 2 0 0 1 2 5 3 4 17 Pupillidae 468 466 216 1 200 221 12 22 1606 Pupisoma species 0 0 0 0 63 44 0 0 107 Xolotrema caroliniense 4 2 4 3 4 1 0 0 18 Total 2580 159 3 1522 180 2379 2366 610 702 11932

PAGE 121

121 Table 6 2 Descriptive statistics for each taxon within the strata of LP81 at Stallings Island Tax on N Mean Median MIN MAX SD Anguispira alternata 581 72.6 72.5 43 106 24.3 Anguispira species 11 1. 4 0 0 8 2.9 Discidae 1758 219.8 203.5 0 511 184.9 Discus patulus 2166 270.8 234 117 464 133 Discus species 133 16.6 0 0 81 31.5 Euconulus chersinus 21 2.6 0.5 0 16 5.5 Gastrocopta contracta 1785 223.1 279.5 1 420 183.4 Gastrocopta rupicola 29 3.6 3 0 10 3.4 Gastrocopta species 423 52.9 1 0 414 145.9 Glyphyalinia indentata 13 1.6 1 0 8 2.7 Glyphyalinia species 1930 241.3 158 0 595 209.5 Haplotrema concavum 31 3.9 1.5 0 16 5.5 Hawaiia minuscula 1062 132.8 112.5 0 263 111.6 Helicodiscus parallelus 175 21.9 21 0 59 17.9 Lucilla species 66 8.2 0 0 58 20.3 Mesodon species 17 2.1 2 0 5 1.8 Pupillidae 1606 200.8 208 1 468 188.9 Pupisoma species 107 13.4 0 0 63 25.3 Xolotrema caroliniense 18 2.3 2.5 0 4 1.8 Table 6 3 Habitat preferences for the t axa found in LP81 at Stallings Island Tax on Open Habitat Preferences Wooded Habitat Preferences N/A Anguispira sp. X Anguispira alternata X Discidae X Discus sp. X Discus patulus X Euconulus chersinus X Gastrocopta sp. X Gastrocopta c ontracta X Gastrocopta rupicola X Glyphyalinia sp. X Glyphyalinia indentata X Haplotrema concavum X Hawaiia minuscula X Helicodiscus parallelus X Lucilla sp. X Mesodon sp. X Pupillidae X Pupisoma sp. X Xolotrema caroliniens e X

PAGE 122

122 Table 6 4 Total number and percentage s of the snails in the taxa with open and wooded habitat tendencies in each strat um in LP81 at Stallings Island Stratum Open Habitat Taxa (n) Open Habitat Taxa (%) Wooded Habitat Taxa (n) Wooded Habitat Taxa (%) VI 53 16.2% 274 83.8% IV 23 6.9% 308 93.1% IIID 553 32.5% 1147 67.5% IIIC 602 36.0% 1070 64.0% IIIA 2 1.5% 131 98.5% ID 465 44.1% 589 55.9% IC 843 69.9% 363 30.1% IB 1360 72.1% 527 27.9% Table 6 5 Summary statistics of the taxa with open an d wooded habitat tendencies each stratum in LP81 at Stallings Island Stratum Open Habitat Taxa (n) Open Habitat Taxa (mean) Open Habitat Taxa (SD) Wooded Habitat Taxa (n) Wooded Habitat Taxa (mean) Wooded Habitat Taxa (SD) Total VI 53 10.6 12.4 274 54.8 6 5.1 327 IV 23 4.6 6.3 308 61.6 96.1 331 IIID 553 110.6 155.9 1147 229.4 278.0 1700 IIIC 602 120.4 175.2 1070 214 235.0 1672 IIIA 2 0.4 0.5 131 26.2 58.6 133 ID 465 93 118.0 589 117.8 150.1 1054 IC 843 168.6 231.8 363 72.6 92.1 1206 IB 1360 272 223.8 527 105.4 152.6 1887 Table 6 6 Chi square results comparing the association of taxa with open and wooded habitat tendencies across the strata in LP81 at Stallings Island VI IV IIID IIIC IIIA ID IC IB Total Open Habitat Taxa 53 23 553 602 2 465 843 1 360 3901 Wooded Habitat Taxa 274 308 1147 1070 131 589 363 527 4409 Total 327 331 1700 1672 133 1054 1206 1887 8310 X 2 = 1405.97 df= 7 p value= <0.0001

PAGE 123

123 Table 6 7 Descriptive statis tics for each taxon within S tratum I II (A, C and D) in LP81 at Stallings Island Tax on N Mean Median MIN MAX SD Anguispira alternata 216 72.0 67.0 43 106 31.8 Anguispira species 8 2.7 0.0 0 8 4.6 Discidae 727 242.3 360.0 0 367 209.9 Discus patulus 1038 346.0 443.0 131 464 186.5 Discus species 5 2 17.3 1.0 0 51 29.2 Euconulus chersinus 1 0.3 0 0 1 0.6 Gastrocopta contracta 726 242.0 331.0 1 394 211.1 Gastrocopta rupicola 13 4.3 3.0 0 10 5.1 Gastrocopta species 0 Glyphyalinia indentata 2 0.7 0 0 2 1.2 Glyphyalinia species 1091 363.7 496.0 0 595 318.8 Haplotrema concavum 25 8.3 9.0 0 16 8.0 Hawaiia minuscula 386 128.7 123.0 0 263 131.6 Helicodiscus parallelus 87 29.0 28.0 0 59 29.5 Lucilla species 8 2.7 0.0 0 8 4.6 Mesodon species 8 2.7 2.0 1 5 2.1 Pupillidae 422 140.7 200.0 1 2 21 121.4 Pupisoma species 107 35.7 44.0 0 63 32.3 Xolotrema caroliniense 8 2.7 3.0 1 4 1.5 Table 6 8 Descriptive statis tics for each taxon within S tratum I (B, C and D) in LP81 at Stallings Island Tax on N Mean Median MIN MAX SD Anguispira alternata 212 70.7 78.0 46 88 21.9 Anguispira species 0 Discidae 918 306.0 278.0 129 511 192.5 Discus patulus 787 262.3 244.0 198 345 75.2 Discus species 81 27.0 0 0 81 46.8 Euconulus chersinus 17 5.7 1.0 0 16 9.0 Gastrocopta contracta 1022 340.7 374 .0 228 420 100.2 Gastrocopta rupicola 7 2.3 3.0 0 4 2.1 Gastrocopta species 421 140.3 5.0 2 414 237.0 Glyphyalinia indentata 2 0.7 1.0 0 1 0.6 Glyphyalinia species 631 210.3 172.0 144 315 91.7 Haplotrema concavum 6 2.0 3.0 0 3 1.7 Hawaiia minuscula 3 16 105.3 58.0 6 252 129.7 Helicodiscus parallelus 55 18.3 18.0 10 27 8.5 Lucilla species 58 19.3 0 0 58 33.5 Mesodon species 2 0.7 0 0 2 1.2 Pupillidae 1150 383.3 466.0 216 468 144.9 Pupisoma species 0 Xolotrema caroliniense 10 3.3 4.0 2 4 1.2

PAGE 124

124 Table 6 9 Total number and percentage s of the snails in the taxa with open and wooded habitat tendencies among S trata VI, IV, III and I in LP81 at Stallings Island Stratum Open Habitat Taxa (n) Open Habitat Taxa (%) Wooded Habitat Taxa (n) Wooded H abitat Taxa (%) VI 53 16.2% 274 83.8% IV 23 6.9% 308 93.1% III 1157 33.0% 2348 67.0% I 2668 64.3% 1479 35.7% Table 6 1 0 Summary statistics of the taxa with open and wooded habitat tendencies in Strata VI, IV, III, and I in LP81 at Stallings Island Stratum Open Habitat Taxa (n) Open Habitat Taxa (mean) Open Habitat Taxa (SD) Wooded Habitat Taxa (n) Wooded Habitat Taxa (mean) Wooded Habitat Taxa (SD) Total VI 53 10.6 12.4 274 54.8 65.1 327 IV 23 4.6 6.3 308 61.6 96.1 331 III 1157 231.4 330.7 2348 469.6 544.2 3505 I 2668 533.6 530.1 1479 295.8 381.8 4147 Table 6 11 Chi square results comparing the association of taxa with open and wooded habitat tendencies in S trata VI, IV, III and I in LP81 at Stallings Island VI IV III I Total Open Habitat Taxa 53 23 1157 2668 3901 Wooded Habitat Taxa 274 308 2348 1479 4409 Total 327 331 3505 4147 8310 X 2 = 1113.46 df= 3 p value= <0.0001 Table 6 12 Chi square results comparing the association of the taxa with open and wooded habitat tend encies in S trata III and I in LP81 at Stallings Island Open Taxa Wooded Taxa Total DF X2 (Yates) P Value I 2668 1479 4147 1 745.6 <0.0001 III 1 157 2348 3 505 Total 3825 3 827 7 652

PAGE 125

125 Table 6 13 Total number of individual snails identified in eac h feature at Ed Marshall (38ED5) Tax on 30 8 17 22 24 37b 55 56 Total Anguispira sp. 0 7 0 7 22 5 13 0 54 Anguispira alternata 10 11 19 8 16 14 4 5 87 Carychium sp. 0 0 0 0 0 1 0 0 1 Carychium exile 0 0 0 0 13 1 0 0 14 Discidae 0 2 7 20 22 0 8 4 63 Di scus sp. 0 0 2 0 0 1 2 0 5 Discus patulus 8 34 64 238 68 298 25 16 751 Euconulus chersinus 1 0 0 0 1 0 0 0 2 Gastrocopta sp. 0 0 0 8 9 0 1 0 18 Gastrocopta contracta 8 3 10 40 71 2 28 5 167 Gastrocopta pellucida 0 0 0 1 0 0 6 0 7 Gastrocopta pentadon 0 0 0 0 8 0 1 0 9 Gastrocopta rupicola 0 0 7 0 32 4 10 4 57 Glyphyalinia sp. 57 11 17 19 103 0 4 9 220 Glyphyalinia indentata 4 0 0 0 0 3 4 0 11 Glyphyalinia wheatleyi 0 0 0 0 0 2 0 0 2 Haplotrema concavum 0 0 0 1 0 0 0 0 1 Hawaiia minuscula 24 62 7 1 81 132 12 82 96 560 Helicodiscus parallelus 0 3 35 33 0 0 1 0 72 Lucilla sp. 81 0 0 0 0 0 21 0 102 Mesodon sp. 1 2 1 0 1 2 2 0 9 Mesodon inflectus 0 2 0 0 0 0 0 0 2 Mesodon thyroidus 1 2 0 0 6 0 2 0 11 Mesomphix sp. 0 2 0 0 0 0 0 0 2 Mesomphix glo bosus 0 0 0 0 4 0 0 0 4 Mesomphix pilsbryi 0 0 0 0 0 6 0 0 6 Polygyra pustuloides 0 0 0 0 1 0 0 0 1 Pupillidae 1 0 20 9 35 54 22 11 152 Pupisoma sp. 0 0 0 0 2 11 1 0 14 Pupoides albilabris 0 0 0 0 4 0 1 1 6 Strobilops labyrinthica 1 0 0 0 0 0 0 0 1 Triodopsis vannostrandi 0 1 0 0 0 0 0 0 1 Ventridens sp. 0 0 0 0 2 5 0 0 7 Ventridens cerinoides 0 1 1 1 0 1 3 0 7 Ventridens gularis 0 0 0 0 0 3 0 0 3 Vertigo tridentata 1 0 0 4 0 0 3 0 8 Xolotrema caroliniense 0 1 0 0 0 0 0 0 1 Zonitidae 0 0 3 0 13 0 0 0 16 Total 198 144 257 470 565 425 244 151 2454

PAGE 126

126 Table 6 14 Descriptive statistics for the snails identified in each taxon (n= > 14) at Ed Marshall (38ED5) Tax on N Mean Median MIN MAX SD Anguispira sp. 54 6.8 6 0 22 7.7 Anguispira alternata 87 10 .9 10.5 4 19 5.2 Carychium exile 14 1.8 0 0 13 4.6 Discidae 63 7.9 5.5 0 22 8.6 Discus patulus 751 93.9 49 8 298 110.7 Gastrocopta sp. 18 2.3 0 0 9 3.9 Gastrocopta contracta 167 20.9 9 2 71 24.3 Gastrocopta pellucida 57 7.1 4 0 32 10.7 Glyphyalinia sp. 220 27.5 14 0 103 10.7 Hawaiia minuscula 560 70.0 11 0 26 8.1 Helicodiscus parallelus 72 9.0 0.5 0 35 15.5 Lucilla sp. 102 12.8 0 0 81 28.5 Pupillidae 152 19.0 15.5 0 54 18.3 Pupisoma sp. 14 1.8 0 0 11 3.8 Zonitidae 16 2.0 0 0 13 4.6 Table 6 15 Habitat preferences for the taxa (n= > 14 ) individual snails at Ed Marshall (38ED5) Tax on Open Habitat Preferences Wooded Habitat Preferences N/A Anguispira sp. X Anguispira alternata X Carychium exile X Discidae X Discus patulus X Gastroc opta sp. X Gastrocopta contracta X Gastrocopta rupicola X Glyphyalinia sp. X Hawaiia minuscula X Helicodiscus parallelus X Lucilla sp. X Pupillidae X Pupisoma sp. X Zonitidae X

PAGE 127

127 Table 6 16 The total number and percentage s of snails within the taxa with open and wooded habitat tendencies in each feature at Ed Marshall (38ED5) Period Feature Open Habitat Taxa (n) Open Habitat Taxa Wooded Habitat Taxa (n) Wooded Habitat Taxa PC 30 82 55.8% 65 44.3% ES 8 0 0% 48 100% 17 27 18.5 % 119 81.5 % 22 9 3.0% 290 97.0 % 24 67 26.5 % 186 73.5 % CS 37b 58 15.8% 309 84.2% 55 53 63.1% 31 36.9% 56 15 37.5 % 25 62.5% Table 6 17 Summary statistics for the taxa with open and wooded habitat tendencies in each feature at Ed Marshall (38ED5) Feature Open Habitat Taxa (n) Open Habitat Taxa (mean) Open Habitat Taxa (SD) Wooded Habitat Taxa (n) Wooded Habitat Taxa (mean) Wooded Habitat Taxa (SD) 30 82 27.3 46.5 65 13 24.8 8 0 0.0 0.0 48 9.6 14.4 17 27 9.0 10.1 119 23.8 26.4 22 9 3. 0 5.2 290 58 101.6 24 67 22.3 19.4 186 37.2 46.1 37b 58 19.3 30.1 309 61.8 132.1 55 53 17.7 6.7 31 6.2 10.6 56 15 5.0 5.6 25 5 7.3 Table 6 18 Chi square results comparing the association of the taxa with open and wooded habitat tendencies across th e features at Ed Marshall (38ED5) 30 8 17 22 24 37b 55 56 Total Open Habitat Taxa 82 0 27 9 67 58 53 15 311 Wooded Habitat Taxa 65 48 119 290 186 309 31 25 1073 Total 147 48 146 299 253 367 84 40 1384 X 2 = 270.33 df= 7 p value= <0.000 1

PAGE 128

128 Table 6 19 The total number and percentages of snails within the taxa with open and wooded habitat tendencies in Preceramic Mill Branch (PC), Early Stallings (ES), and Classic Stallings (CS) periods at Ed Marshall (38ED5) Period Open Taxa (n ) Open Taxa Wooded Taxa (n) Wooded Taxa PC 82 55.8% 65 44.2% ES 103 13.8 % 643 86.2% CS 126 25. 7 % 365 74. 3 % Table 6 20 Summary statistics for the taxa with open and wooded habitat tendencies from the Preceramic Mill Branch (PC), Early Stallings (ES) and Classic Stallings (CS) periods at Ed Marshall (38ED5) Period Open Habitat Taxa (n) Open Habitat Taxa (mean) Open Habitat Taxa (SD) Wooded Habitat Taxa (n) Wooded Habitat Taxa (mean) Wooded Habitat Taxa (SD) Total PC 82 27.3 46.5 65 13 24.8 147 ES 103 34.3 32.3 643 128.6 164.6 746 CS 126 42.0 39.0 365 73 148.8 491 Table 6 21 Chi square results comparing the association of taxa with open and wooded habitat tendencies found within the Preceramic Mill Branch phase (PC) and Early Stallings period (ES) at Ed Marshall (38ED5) Period Open Habitat Taxa Wooded Habitat Taxa Total DF Yates X 2 P Value PC 82 65 147 1 220.52 <0.0001 ES 103 643 746 Total 185 Table 6 22 Chi square results comparing the association of the taxa with open and woode d habitat tendencies found within Early Stallings (ES) and Classic Stallings (CS) periods at Ed Marshall (38ED5) Period Open Habitat Taxa Wooded Habitat Taxa Total DF Yates X 2 P Value ES 103 643 746 1 27.59 <0.0001 CS 126 365 491 Total 229 1008 1237

PAGE 129

129 Table 6 2 3 Chi square results comparing the association of the taxa with open and wooded habitat tendencies found in the Preceramic Mill Branch phase (PC), Early Stallings period (ES), and Classic Stallings period (CS) at Ed Marshall (38ED5) Period PC ES CS Total Open Habitat Taxa 82 103 126 311 Wooded Habitat Taxa 65 643 365 1073 Total 147 746 491 1384 X 2 = 128.642 df= 2 p value= <0.0001

PAGE 130

130 Table 6 2 4 Total number of individual snails identified in each feature from Mims Point ( 38ED 9 ) Tax on 1 50 66 51 52 Total Anguispira sp. 3 0 0 0 0 3 Anguispira alternata 29 6 5 25 2 67 Carychium exile 1 2 1 0 0 4 Discidae 4 0 8 0 0 12 Discus patulus 163 0 5 19 0 187 Gastrocopta sp. 0 4 1 0 0 5 Gastrocopta contracta 0 7 9 0 0 16 Gastrocopta pellucida 2 0 2 0 0 4 Gastrocopta pentadon 0 0 1 0 0 1 Gastrocopta rupicola 0 34 1 0 0 35 Glyphyalinia sp. 12 22 12 0 0 46 Glyphyalinia indentata 9 13 0 0 0 22 Glyphyalinia wheatleyi 19 0 0 0 0 19 Haplotrema concavum 1 0 0 0 0 1 Hawaiia minuscula 7 118 90 40 27 282 Helicodiscus parallelus 0 0 0 1 0 1 Lucilla sp. 27 35 0 2 0 64 Mesodon thyroidus 1 0 0 0 0 1 Mesomphix sp. 2 0 0 0 0 2 Polygyra pustuloides 1 0 0 0 0 1 Practicolella lawae 17 0 0 0 0 17 Pupillidae 1 22 7 0 0 30 Pupoides albilabris 0 0 1 0 0 1 Stenotrema stenotrema 1 0 0 0 0 1 Strobilops labyr inthica 0 0 2 0 0 2 Triodopsis sp. 1 1 0 0 0 2 Triodopsis hopetonensis 1 1 0 0 0 2 Triodopsis vannostrandi 23 0 0 0 0 23 Ventridens sp. 123 0 0 0 0 123 Ventridens cerinoides 10 1 0 0 0 1 1 Xolotrema caroliniense 4 0 0 0 0 4 Zonitidae 1 0 3 1 0 5 Total 465 265 148 88 30 996

PAGE 131

131 Table 6 25 Descriptive statistics for each taxon (n=> 14 ) individual snails at Mims Point (38ED9) Tax on N Mean Median MIN MAX SD Anguispira alternata 67 13.4 6 2 29 12.6 Discus patulus 187 37.4 5 0 163 70.6 Gastrocopta contracta 16 3.2 0 0 9 4.4 Gastrocopta rupicola 35 7 0 0 34 15.1 Glyphyalinia sp. 46 9.2 12 0 22 9.3 Glyphyalinia indentata 22 4.4 0 0 13 6.2 Glyphyalinia wheatleyi 19 3.8 0 0 19 8.5 Hawaiia minuscula 282 56.4 40 7 118 46.1 Lucilla sp. 64 12.8 2 0 35 16.9 Practicolella lawae 17 3.4 0 0 17 7.6 Pupillidae 30 6 1 0 22 9.4 Triodopsis vannostrandi 23 4.6 0 0 23 10.3 Ventridens sp. 123 24.6 0 0 123 55.0 Table 6 26 Habitat preferences for t he taxa (n=> 14 ) individual snails at Mims Point (38ED9) Tax on Open Habitat Preferences Wooded Habitat Preferences N/A Anguispira alternata X Discus patulus X Gastrocopta contracta X Gastrocopta rupicola X Glyphyalinia sp. X Glyphyalinia ind entata X Glyphyalinia wheatleyi X Hawaiia minuscula X Lucilla sp. X Practicolella lawae X Pupillidae X Triodopsis vannostrandi X Ventridens sp X Table 6 27 The total number and percentage s of snails within the taxa with open and wooded habitat tendencies found in each feature at Mims Point ( 38ED9 ) Feature Open Habitat Taxa (n) Open Habitat Taxa Wooded Habitat Taxa (n) Wooded Habitat Taxa 1 28 7.7 % 334 92.3 % 50 91 80.5% 22 19.5% 66 8 32.0% 17 68.0% 51 2 9.5% 19 90.5% 52 0 0.0 % 0 0.0%

PAGE 132

132 Table 6 28 Summary statistics for the total number of snails within the taxa with open and wooded habitat tendencies in the features from Mims Point (38ED9) Feature Open Habitat Taxa (n) Open Habitat Taxa (mean) Open Habitat Taxa (SD) Wooded Habitat Taxa (n) Wooded Habitat Taxa (mean) Wooded Habitat Taxa (SD) Total 1 28 9.3 15.3 334 66.8 71.0 362 50 91 30.3 7.2 22 4.4 9.8 113 66 8 2.7 3.8 17 3.4 5.3 25 51 2 0.7 1.2 19 3.8 8.5 21 52 0 0.0 0.0 0 0 0.0 0 Table 6 29 Chi square results co mparing the association of the taxa with open and wooded habitat tendencies within the features from Mims Point (38ED9) 1 50 66 51 52 Total Open Habitat Taxa 28 91 8 2 0 129 Wooded Habitat Taxa 334 22 17 19 0 392 Total 362 113 25 21 0 521 X 2 = 248.31 df= 3 p value= <0.0001 Figure 6 1 Percentages of the snails within the taxa with open and wooded habitat tendencies across the strata within LP81 at Stallings Island

PAGE 133

133 Figure 6 2 Percentages of the snails within the taxa with open and wooded habitat tendencies across the strata (excluding S tratum IIIA ) within LP81 at Stallings Island Figure 6 3 Means and standard deviations of the t otal number of snails in the taxa with open and wooded habitat tendencies across the strata wit hin LP81 at Stallings Island

PAGE 134

134 Figure 6 4 Percentages of the snails within the taxa with open and wooded habitat tendencies across S trata VI, IV, III, and I within LP81 at Stallings Island Figure 6 5 Percentages of the snails within the taxa with o pen and wooded habitat tendencies between S trata III and I within LP81 at Stallings Island

PAGE 135

135 Figure 6 6 Means and standard deviations of the t otal number of snails in the taxa with open and wooded habitat tendencies among the four main strata within LP8 1 at Stallings Island Figure 6 7 Percentages of the snails within the taxa with open and wooded habitat tendencies among Feature 30, which dates to the preceramic Mill Branch phase at Ed Marshall (38ED5)

PAGE 136

136 Figure 6 8 Percentages of the snails within the taxa with open and wooded habitat tendencies among the features dating to the Early Stallings period at Ed Marshall (38ED5) Figure 6 9 Percentages of the snails within the taxa wit h open and wooded habitat tendencies among the features dating to the Classic Stallings period at Ed Marshall (38ED5)

PAGE 137

137 Figure 6 10 Means and standard deviations for the total number of snails within the taxa with open and wooded habitat tendencies amon g the features from Ed Marshall (38ED5) Figure 6 11 Percentages of the snails within the taxa with open and wooded habitat tendencies during the Preceramic Mill Branch (PC), Early Stallings (ES), and Classic Stallings (CS) periods at Ed Marshall (38ED 5)

PAGE 138

138 Figure 6 12 Means and standard deviations for the total number of snails within the taxa with open and wooded habitat tendencies among the Preceramic Mill Branch (PC), Early Stallings (ES), and Classic Stallings (CS) periods at Ed Marshall (38ED5) Figure 6 13 Percentages of the snails within the taxa with open and wooded habitat tendencies among the features at Mims Point (38ED9)

PAGE 139

139 Figure 6 1 4. Means and standard deviations of the total number of snails within the taxa with open and wooded habitat tendencies among the features from Mims Point (38ED9)

PAGE 140

140 Figure 6 1 5 Plan distribution of features at Mims Point (38ED9), showing projected pit feature clusters of a circular domestic compound with a central area largel y devoid of features (Sassaman n.d.)

PAGE 141

141 CHAPTER 7 CHANGES OVER TIME During the analysis, several samples within each site were dated utilizing pottery artifacts and radiocarbon assays from mainly wood charcoal, some hickory nutshell, and diagnostic. The radi ocarbon assays resulted in the establishment of three main time periods represented within the three sites. The time periods include the Preceramic phase (including Mill Branch and Paris Island), the Early Stallings period, and the C lassic Stallings perio d (Table 7 Stallings Island and one of the features, Feature 30, from Ed Marshall (38ED5) were dated to the Preceramic Late Archaic, specifically the Mill Branch and Paris Island phases. Radiocarbon a ssays dated LP81 from 3860 + 70 B.P. to 4840 + 70 B.P. and Feature 30 at 4140 + 90 B.P. (Sassaman et al. 2006). Four features, Features 8, 17, 22, and 24, from Ed Marshall (38ED5) were dated to the Early Stallings period (ranging from 3840 + 9 0 B.P. to 3800 + 60 B. P.) (Sassaman n.d.). Three features, Features 37b, 55, and 56, from Ed Marshall (38ED5) and all of the features from Mims Point (38ED9), Features 1, 50, 66, 51, and 52, were dated to the Classic Stallings period (ranging from 4025 + 70 B.P. to 3580 + 60 B.P.) (Sassaman n.d.). Based on the chronological sequence of the strata and features from Stallings Island, Ed Marshall, and Mims Point the taxa with open habitat tendencies and the taxa with wooded habitat tendencies were compared over time among the three t ime periods in an attempt to detect regional Results The total number and percentages of individuals within the taxa with open habitat tendencies and wooded habitat tendencies were compare d among the samples in

PAGE 142

142 chronological order and are listed on Table 7 2. The total number of individuals in the taxa with open habitat tendencies within the samples from the Preceramic Mills Branch phase range from 2 to 1360 (1.5% to 72.1%). The total num ber of individuals in the taxa with wooded habitat tendencies within the samples from the Preceramic Mills Branch phase range from 65 to 1147 (44.2% to 67.5%). The total number of individuals in the taxa with open habitat tendencies within the samples fro m the Early Stallings period range from 0 to 67 (0% to 26.5%) The total number of individuals in the taxa with wooded habitat tendencies within the samples from the Early Stallings period range from 48 to 290 (73.5% to 100%). The total number of individ uals in the taxa with open habitat tendencies within the samples from the Classic Stallings period range from 0 to 91 (0% to 80.5%). The total number of individuals in the taxa with wooded habitat tendencies within the samples from the Classic Stallings p eriod range from 0 to 334 (0% to 92.3%). A shift in the percentages throughout all of the samples is illustrated on Figure 7 1. More snails were found to have wooded habitat tendencies in the samples from Stallings Island (LP81) including Strata VI, IV IIID, IIIC, IIID, and ID. A shift occurs in the next three samples from Stallings Island (LP81) Strata IC and IB, and Ed Marshall (38ED5) F eature 30, where more snails were observed to have open habitat tendencies. The next six samples from Ed Marsha ll, Features 8, 17, 22, 24, and 37b, and Mims Point, Feature 1, were observed to have more individual snails with wooded habitat tenden cies The next three samples continue to alternate from more individual snails with open habitat tendencies (Ed Marshall F eatu re 55, and Mims Point, F eature 50) and wooded ha bitat tendencies (Ed Marshall, F eature 56) The next two samples

PAGE 143

143 were observed to have more individual snails with wooded habitat tendencies (Mims Point, Features 66 and 51). The most rece nt sample, Mims Point, F eature 52, did not have any snails with open or wooded habitat tendencies. To compare the different time periods, the samples within each period were combined and the results are listed on Table 7 3. In the preceramic Mill Branch phase, a tot al of 3983 (47.1%) individual snails were found within the taxa with open habitat tendencies and a total of 4474 (52.9%) individual snails were found within the taxa with wooded habitat tendencies. In the Early Stallings period, a total of 103 (13.8%) ind ividual snails were found within the taxa with open habitat tendencies and a total of 643 (86.2%) individual snails were found within the taxa with wooded habitat tendencies. In the Classic Stallings period, a total of 255 (25.2%) individual snails were f ound within the taxa with open habitat tendencies and a total of 757 (74.8%) individual snails were found within the taxa with wooded habitat tendencies. The percentages of snails with open and wooded habitat tendencies within the three time periods are il lustrated on Figure 7 2. It is apparent that more snails with wooded habitat tendencies were present in all three time periods; however, an increase in the percentage of snails with wooded habitat tendencies occurred during the Early Stallings period (Fig ure 7 2). T he Chi square test was used to determine the strength of associations between the habitat tendencies and habitat types among the three time periods, Preceramic, Early Stallings, and Classic Stallings and resulted in a Chi square value of 448.4 4 (df=2) a nd a p value of <0.0001 (Table 7 4), which suggests an extremely significant association. Regardless of the dependent sampling method concerns, nonrandom

PAGE 144

144 associations among habitat tendencies, habitat types, and ultimately the three time periods ( samples over time) are evident in the numerical data alone, which is verified by the Chi square statistic even if its probability values cannot be accepted uncritically due to the biases of dependent data. Discussion The results of the analysis indicate that the three time periods were comprised of more wooded habitats than open habitats based on the percentages of snails within the taxa previously identified as having open or wooded habitat tendencies. However, upon analyzing the sites separately, varia tions in the taxa with open and wooded habitat tendencies were apparent, which indicated several shifts in open and wooded habitats over time, with more variations in the Classic Stallings period possibly due to spatial variations (Figure 7 1). An increa se in the taxa with wooded habitat tendencies was observed during the Early Stallings period (Figure 7 2). However, all of the samples within the Early Stallings period were from the Ed Marshall site This may indicate a time of limited use at Ed Marshal l due to unfavorable conditions possibly due to flooding. The shift in taxa with open and wooded habitat tendencies over time also indicates more intensive occupation of the sites during the preceramic Mill Branch phase, less occupation during the Early S tallings period, and a return of habitation during the Classic Stallings period (Figure 7 2). The changes in habitation, suggested by the shift in environmental conditions apparent by the variations in the snail fauna at the sites, support previous eviden ce of the habitation fluctuations a t the sites over the three time periods present in this research.

PAGE 145

145 Table 7 1. Radiocarbon dates associated with samples from Stallings Island (LP81), Ed Marshall (38ED5), and M i ms P oi nt (38ED9) and the corresponding periods: Preceramic (PC), Early Stallings (ES), and Classic Stallings (CS) (Sassaman n.d.; Sassaman et al. 2006) Period Site Stratum/Feature C14 Dates (B.P.) PC LP81 VI 4360 + 70; 4830 + 70 PC LP81 V 4370 + 70 PC LP81 IV 4260 + 70 PC LP81 IIIB 4210 + 6 0 PC LP81 IB / IC 4140 + 70 PC LP81 IA 4210 + 60 PC 38ED5 30 4140 + 90 ES 38ED5 8 ES 38ED5 17 ES 38ED5 22 3840 + 90 ES 38ED5 24 3800 + 60 CS 38ED9 1 4025 + 65 CS 38ED5 37b 3560 + 60 CS 38ED5 55 CS 38ED5 56 CS 38ED9 50 CS 38ED9 66 3 700 + 150 CS 38ED9 51 3580 + 60 CS 38ED9 52

PAGE 146

146 T able 7 2. Total number and percentages of the snails within the taxa with open and wooded habitat tendencies in the samples from Stallings Island (LP81), Ed Marshall (38ED5) and Mims Point (38 E D9) in chronological order Period Site Stratum/ Feature Open Habitat Taxa (n) Open Habitat Taxa (%) Wooded Habitat Taxa (n) Wooded Habitat Taxa (%) PC LP81 VI 53 16.2% 274 83.8% PC LP81 IV 23 6.9% 308 93.1% PC LP81 IIID 553 32.5% 1147 67.5% PC LP81 III C 602 36.0% 1070 64.0% PC LP81 IIIA 2 1.5% 131 98.5% PC LP81 ID 465 44.1% 589 55.9% PC LP81 IC 843 69.9% 363 30.1% PC LP81 IB 1360 72.1% 527 27.9% PC 38ED5 30 82 55.8% 65 44.2% ES 38ED5 8 0 0.0% 48 100.0% ES 38ED5 17 27 18.5% 119 81.5% ES 38ED5 22 9 3.0% 290 97.0% ES 38ED5 24 67 26.5% 186 73.5% CS 38ED9 1 28 7.7% 334 92.3% CS 38ED5 37b 58 15.8% 309 84.2% CS 38ED5 55 53 63.1% 31 36.9% CS 38ED5 56 15 37.5% 25 62.5% CS 38ED9 50 91 80.5% 22 19.5% CS 38ED9 66 8 32.0% 17 68.0% CS 38ED9 51 2 9.5% 1 9 90.5% CS 38ED9 52 0 0.0% 0 0.0% Table 7 3. Total number and percentages of the snails within the taxa with open and wooded habitat tendencies in the samples from the Preceramic Mills Branch (PC), Early Stallings (ES), and Classic Stallings (CS) perio ds Period Open Habitat Taxa (n) Open Habitat Taxa (%) Wooded Habitat Taxa (n) Wooded Habitat Taxa (%) PC 3983 47.1% 4474 52.9% ES 103 13.8% 643 86.2% CS 255 25.2% 757 74.8%

PAGE 147

147 Table 7 4. Chi square test results comparing the association of the snai ls within the taxa with open and wooded habitat tendencies and the three time periods: Preceramic Mills Branch (PC), Early Stallings ( ES), and Classic Stallings (CS) Period Open Habitat Taxa (n) Wooded Habitat Taxa (n) Total df Chi square p value PC 3983 4474 8457 2 448.44 <0.0001 ES 103 643 746 CS 255 757 1012 Total 4341 5874 10215 Figure 7 1. Percentages of snails within the taxa with open and wooded habitat tendencies in chronological order across the samples from Stallings Island (L P81), Ed Marshall (38ED5), and Mims Point (38ED9)

PAGE 148

148 Figure 7 2. Percentages of snails within the taxa with open and wooded habitat tendencies in chronological order in the samples from the three time periods: Preceramic Mills Branch, Early Stallings and Classic Stallings periods

PAGE 149

149 CHAPTER 8 CONCLUSION This study was enabled by the collection of modern snails from known habitats. The interpretations of the archaeological snails were based upon the modern land snail collections and records of the ecological conditions under which they lived. In addition to the modern land snail data and habitat tendencies observed through the analysis of the modern snail samples, referencing historical records was also an essential part of this research to conclude habitat tendencies within the taxa. Previous literature has utilized terrestrial gastropods to interpret past environmental/climate conditions at archaeological and modern sites (Peacock and Gerber 2008). This research concluded that properly collected and proces sed archaeological soil samples, in combination with a modern representative collection for comparison, can provide adequate numbers of terrestrial gastropods that can ultimately be used to interpret past habitat conditions. The presumed habitat condition s of the snail taxa can then be used along with other data, including organic materials, vertebrate fauna, charcoal, and other artifacts, to suggest the location of habitat structures and anthropogenic land use. It must be noted that proper collection tec hniques, i.e. bulk soil samples, and proper soil processing techniques, such as fine screening and water screening, are essential to the proper recovery of terrestrial snails due to the variation in shell sizes from less than 1mm to several centimeters. R ecovery of minute snails is crucial to gathering a full representation of the snail fauna to use as indicators for past environmental conditions. Nekola and Coles (2010) specifically state that snails in the family Pupillidae must be adequately sampled an d correctly identified to properly quantify diversity within snail populations, which ultimately may show habitat

PAGE 150

150 preferences. Without the proper sampling techniques and correct identifications of all the snail taxa present at the site, the results may be skewed and void due to the missing data (Nekola and Coles 2010). In addition to proper sampling techniques, the collection of modern comparative samples from the same geographical area is an essential part of this type of research. Due to the variation in habitat preferences based on geographical locations, modern samples collected in the same area from known habitats is necessary for the proper variations (Peacock an d Gerber 2008). The archaeological terrestrial snails, along with the modern samples, permitted the interpretation of the habitats at the three sites within this study: Stallings Island, Ed Marshall, and Mims Point. The archaeological snails from the mi dden (LP81) at Stallings Island allowed a comparison of the ambient habitat conditions over time and indicated a disturbance event that created more open habitat conditions in the upper strata. The archaeological snails from Ed Marshall indicated more woo ded conditions during the Early Stallings period, which supported a decrease in habitation. The archaeological snails from Mims Point supported the central village plaza concept previously observed through the analysis of other artifacts and remains at th e site by indicating more of an open habitat in the central plaza of the village (Sassaman et al. 2006). Three main time periods, preceramic Mill Branch phase, Early Stallings period, and Classic Stallings period, were represented at the three sites. Th e archaeological snails were compared over the three time periods and showed fluctuations in the

PAGE 151

151 habitats over time with an overall shift in habitat type from wooded to open habitat conditions in the preceramic Mill Branch phase to an increase in wooded ha bitat conditions during the Early Stallings phase (as seen at Ed Marshall) and a decrease in wooded habitat conditions during the Classic Stalling phase. The shift over time indicates a period of land clearance (more snails in taxa with open habitat tende ncies). This is possibly the result of intensive habitation during the preceramic Mill Branch phase, to a period of forest regrowth (more snails with wooded habitat tendencies) possibly associated with inhabitation during the Early Stallings period to an increase in land clearance. The results suggest increased occupation or land use and possibly a shift to more permanently and extensively cleared central plaza village complexes throughout the Classic Stallings period. In conclusion, t he hypotheses of thi s study we re supported through the analyses of the terrestrial snails : 1) overall, an increase in open habitat species w as found through time as the intensity of the settlement increased through the Classic Stallings period, with a possible break in habita tion during the Early Stallings period ; 2) variations in proportions of open habitat species co var ied with known periods of occupation and abandonment at the sites and over time ; and 3) within the Classic Stallings phase, open habitat species occur red wit h greatest frequency in central plazas of the permanent village at Mims Point. Future Research This study could have been optimized by additional collections of the modern samples. Research has shown that collecting samples multiple times throughout diffe rent seasons result in a more holistic representation of the population. Multiple

PAGE 152

152 samples over time allow snails with seasonal or rainfall preferences to be collected and gives a better representation of the snail fauna at the sample sites. Climatic chan ges are well known and accepted in archaeology and are often the topic of research projects. Large scale disturbances have also been found to effect terrestrial snail populations (Bloch and Willig 2006). The results of this study may indicate an overall geographical or climatic change or natural large scale disturbance, which created a shift in the snail taxa present at the sites over time rather than local habitat changes due to human impacts. Further research into the climatic conditions and shifts occ urring during the time periods in question may result in an alternative explanation for the shift in snail taxa rather than site specific environmental impacts due to habitat pressures or environmental changes due to the creation of central plaza village c omplexes. The death and life assemblages of the modern snails could have been analyzed as a portion of this study to determine the actual population being sampled at the modern sites. This could have been done by documenting the number of snail shells con taining the soft body of the mollusk near the time of collection. Due to the length of time from the original collection and final identifications, this was not possible in this study; however, future studies conducted in a similar fashion may want to con sider incorporating this type of analysis to differentiate what taxa were actually alive at the time of collection compared to what taxa were a part of the death assemblage at the time of collection. The age of the shells may identify sampling bias associ ated with mixing the taxonomic data of live and dead assemblages and could be compared for additional research interests. Thurman et al. (2008) determined that dead shell surveys

PAGE 153

153 can be used as a predictor of live snail density and are a useful tool for s tudying the ecology of certain species, but are not a perfect numerical predictor due to variation differences found between the live and dead snail abundances. Depending upon the species, the life assemblages will be more indicative of the present enviro nmental conditions and comparing the two assemblages may alert the researcher that additional collections are needed in order to get a better sample for that particular habitat or environmental condition (Thurman et al. 2008). Several taxa were identified in the archaeological samples that were either not found in the modern samples or only found in very small numbers. Some of the taxa include Anguispira alternata, Discus patulus, Haplotrema concavum, Triodopsis species, and Xolotrema caroliniense Others were identified in the modern samples but not in the archaeological samples or were found in very limited numbers. Some of these taxa include Paravitrea capsella, Patera species, Polygyra pustuloides, Striatura meridionalis, Strobilops labyrinthica, and Zonitoides arboreus Some of the taxon have certain habitat preferences; however, others are considered generalists (Table 4 5). The shift in the presence/absence or increase/decrease in numbers of the snails within each taxon may indicate a shift in cli matic conditions creating temperature or rainfall changes or other environmental constraints. The change could be related to the increase of human activity and/or increases in environmental contaminants such as pesticides, herbicides or other pollutants. Trace metals and polycyclic aromatic hydrocarbons have successfully been measured in terrestrial snails to interpret the biological impact of atmospheric pollutants on the environment in urban areas (Regoli et al. 2006). Harms (2008) used snail assemblag es to study changes to the regional

PAGE 154

154 environment in combination with radiocarbon dating. In addition, analyzing the archaeological and modern soil samples for pollutants could be combined with the taxonomic data, and possibly radiocarbon dating, to interpr et the changes that may have occurred and impacted the snails that are present or absent today. Morphometric analysis could also be conducted on the snails identified in this study to interpret environmental impacts and changes over time. Pressures from p opulation density have been negatively associated with body size within species due to intraspecific competition (Perry and Arthur 1991). The associations between population density and body size have also been studied in single and mixed colonies to loca te interspecific competition effects (Cameron and Carter 1979; Perry and Arthur 1991). Intraspecific variation in shell morphology, including shell size, relative shell height, and whorl number have been used in combination with heavy metal concentrations from the soil, plants, and soft bodies of snails to interpret heavy metal bioaccumulation in an attempt to discover increased anthropogenic pressures leading to increased pollution (Nica et al. 2012). Soil acidification may also be altered from increases in pollution and other human activities; therefore, soil pH, which is correlated to the concentration of calcium, may affect some terrestrial snails and can be studied by analyzing the correlations of shell chemical composition, soil composition, and spec ies richness and density (G rdenfors et al. 1995). The change in morphological features may also be used to indicate increased population pressures in addition to environmental changes (Williamson, et al. 1976; Tattersfield 1981; Goodfriend 1986; Baur 1 988). Analyzing the variations in the sizes between the taxa with open habitat tendencies and the taxa with wooded habitat

PAGE 155

155 tendencies may also give strong indications for shell size, shape, and thickness as it relates to the habitat preferences of the tax on. Baur (1988) determined that shell height and number of whorls was not influenced by habitat and concluded it was negatively correlated with population density and positively correlated with the degree of plant cover within the habitat. Tattersfield ( 1981) also determined that shell size was not influenced by environmental factors; however, mean shell size was negatively associated with population density. Larger snails are often associated with moister conditions and smaller adult snail sizes may be related to higher population densities (Goodfriend 1986). In contrast, smaller relative aperture areas and whorl cross sectional areas are often related to drier conditions (Goodfriend 1986). Shell dimensions including the thickness and width of the aper tural lip, the shape, size and placement of the internal lamellae, and the general shell form can also be used to identify certain species (Nekola and Barthel 2002). Shell size, genetic variations, and marks of predation have also been used to study micro geographic adaptation to biotic aspects of the environment (Schilthuizen et al. 2006). Backeljau et al. (2009) attempted to determine if large clearcuts and reforested areas within a known forest taxon could be associated with morphometric variability (e. g. aperture size) in addition to altered genetic composition in neutral genetic markers. These types of analyses could also be conducted with the modern or archaeological samples or in future research from these sites. Other research has shown that past c ultural groups relied heavily on snails for human consumption. The use of the snails located in this study should also be considered; however, most snails known to be used for human consumption were larger

PAGE 156

156 sized taxa and easily provided a source of meat f or consumption. The majority of snails in this study were minute species and were approximately 1 to 5 mm long/in diameter. Only a few taxa in the genus Mesodon would have provided a large source of meat for consumption. These were identified at several of the sites, but none of the individuals had any indications on the shell that they had been used for consumption. Some indicators would have been soot from charcoal or certain portions of the shell would have been broken to release the soft body of the mollusk. Due to the smaller sizes and lack of indicators for past consumption of the snails in this study, human consumption was not considered as a reason for the snail assemblages being found in the archaeological samples. Future research involving sn ail assemblages, archaeological or modern, should also consider analyzing the surrounding soil matrix for extractable calcium, water soluble calcium and pH (Hotopp 2002; Nekola 2010). These three soil parameters have been shown to be related to land snail communities and cause changes in the ecosystems due to the role of land snails in cycling calcium to higher trophic levels (Hotopp 2002). The soil chemistry measurements can also be correlated with snail abundance and species richness to test the associa tions (Hotopp 2002; Nekola 2010). Due to the bulk soil sampling methods commonly utilized in archaeological excavations, and litter sampling methods used in modern sampling, soil analysis could easily be utilized for additional data regarding the land sna il abundance and richness at the sites. Soils collected from modern sites could also be used to determine snail preferences for calcium and pH and compared to the corresponding archaeological samples. Unfortunately, the soil samples in this study were no t available for chemical analysis.

PAGE 157

157 Summary Terrestrial snails from archaeological and modern sites have the potential to assist in a variety of research areas involving the environment, which was supported in the results of this research and the prior researc h mentioned above. The potential research areas include environmental archaeology, anthropology, paleoecology, ecology, climatology, and environmental science, including anthropogenic impacts on the environment and studies involving pollution. This study only inv olved a sites in the southeastern United States; however, terrestrial snails can be utilized in virtually any research area due to their large geographical ranges and their sensitivity to the surrounding environment. Terres trial snails are an invaluable re source to scientists and should be considered for future research projects involving past or present environmental conditions a nd environmental impacts.

PAGE 158

158 LIST OF REFERENCES Abrams, Marc D. and Gregory J. Nowacki 2008 Native Americans as Active and P assive Pr omoters of Mast and Fruit Trees in the Eastern USA. The Holocene 18:1123 1137. Allen, Melinda S. 1992 Dynamic Landscapes and Human Subsistence: Archaeological Investigations on Aitutaki Island, Southern Cook Islands. Ph.D. Di ssertation. Ann Arbor: University Microfilms International. 1994 The Chronology of Coastal Morphogenesis and Human settlement on Aitutaki, Southern Cook Islands, Polynesia. Radiocarbon 36:59 71. 1998 Holocene Sea Level Change on Aitutaki, Cook Islands: Landscape Change and Human Response. Journal of Coastal Research 14:10 22. Allen, Melinda S., and Susan E. Schubel 1990 Recent Archaeological Research on Aitutaki, Southern Cooks: The Moturakau Shelter. Journal of the Polynesian Soci ety 99:265 296. Alvarez, Javier, and Michael R. Willig 1993 Effects of Treefall Gaps on the Density of Land Snails in the Luquillo Exper imental Forest of Puerto Rico. Biotropica 25:100 110. Archer, A. F. 1939 The Ecology of the Mollusca of t he Edwin S. George Reserve, Livingston County, Michigan In Occasional Papers of the Museum of Zoology. No. 398, pp. 1 24. University of Michigan. University of Michigan Press, Ann Arbor, Michigan. Backeljau, Thierry, Frederik Hendrickx, Kirt Jordaens, He ike Kappes, Luc Lens, and Jean Pierre Maelfait. Biological Journal of the Linnean Society 98:839. Baerreis, D. A. 1969 A Preliminary Analysis of Gastropods from the Mill Creek Sites. Io wa Archaeological Society Journal 16:333 343. 1971 Environmental Reconstruction through Molluscan Remains: A Preliminary Report on the A. C. Banks Site. In Prehistoric Investigations edited by M. McKusick, pp. 95 108. Report 3, Office of State A rchaeologist, Iowa. 1973 Gastropods and Archaeology. In Variation in Anthropology, Essays in Honor of John C. McGregor, edited by Donald W. Lathrap and Jody Douglas, pp. 43 53. Illinois Archaeological Survey, Urbana, Illinois. 1974 Mollusc an Remains from 13CK405. Iowa Archaeological Society Journal 21:131 143.

PAGE 159

159 1980 Habitat and Climatic Interpretation from Terrestrial Gastropods at the Cherokee Site. In The Cherokee Excavations: Mid Holocene Ecology and Human Adaptation in No rthwestern Iowa, edited by D. C. Anderson and H. A. Semken, pp.101 122. Academic Press, New York. Bailey, Larissa L., James E. Hines, James D. Nichols, and Darryl I. MacKenzie 2007 Sampling Design Trade Offs in Occupancy Studies with Imperfect Detec tion: Examples and Software. Ecological Applications 17:281 290. Barber, Russell J. 1979 Human Ecology and the Estuarine Ecosystem Ph.D. Dissertation, Department of Anthropology, Harvard University, unpublished. 1980 Analysis of the Gastr opod Assemblage from the Great Diamond Island Site, Casco Bay, Maine M.S. on file, Department of Anthropology and Geography, University of Southern Maine, Portland, Maine. 1988 The Use of Land Snails from Prehistoric Sites for Paleoenvironmental Reconstruction. In Holocene Human Ecology in Northeastern North America Edited by George P. Nicholas. pp. 11 28. Plenum Press, New York. Barker, Gary M., and Pauline C. Mayhill 1999 Patterns of Diversity and Habitat Relationships in Terrestrial M ollusc Communities of the Pukeamaru Ecological District, Northeastern New Zealand. Journal of Biogeography 26:215 238. Baur, Bruno 1988 Microgeographical Variation in Shell Size of the Land Snail Chondrina clienta. Biological Journal of the Linn ean Society 35:247 259. Baur, Bruno, and Anette Baur 1993 Climatic Warming Due to Thermal Radiation from an Urban Area as Possible Cause for the Local Extinction of a Land Snail. Journal of Applied Ecology 30:333 340. Bloch, Christopher P., and Mic hael R. Willig 2006 Context Dependence of Long Term Responses of Terrestrial Gastropod Populations to Large Scale Disturbance. Journal of Tropical Ecology 22:111 122. Bobrowsky, Peter T. 1984 The History and Science of Gastropods in Archaeol ogy. American Antiquity 49:77 93 Boycott, A. E. 1934 The Habitats of Land Mollusca in Britain. Journal of Ecology 22:1 38. Cao, Yong, D. Dudley Williams, and David P. Larsen 2002 Comparison of Ecological Communities: The Problem of Sample Representativeness. Ecological Monographs 72:41 56.

PAGE 160

160 Cameron, R. A. D., and M. A. Carter 1979 Intra and Interspecific Effects of Population Density on Growth and Activity in Some Helicid Land Snails (Gastropoda: Pulmonata). Journal of Animal Ecolo gy 48:237 246. Christensen, C. C., and Patrick V. Kirch 1981 Nonmarine Mollusks from Archaeological Sites on Tikopia, Southeastern Soloman Islands. Pacific Science 35:75 88. 1986 ahu, Bishop Museum Occasional Papers 26:52 80. Claflin, William H., Jr. 1931 Peabody Museum of American Archaeology and Ethnology Papers 14(1). Cambridge. Coles, Brian F. and Gerald E. W alsh 1999 A Revised List of Arkansas Terrestrial Mollusks with Notes on the Geographic Distribution of Species. Journal of the Arkansas Academy of Sciences 53:32 37. Cowie, Robert H. 2001a Decline and Homogenization of Pacific Faunas: The La nd Snails of American Samoa. Biological Conservation 99:207 222. 2001b Invertebrate Invasions on Pacific Islands and the Replacement of Unique Native Faunas: A Synthesis of the Land and Freshwater Snails. Biological Invasions 3:119 136. Cowie, Ro bert H., and A. C. Robinson 2003 The Decline of Native Pacific Island Faunas: Changes in Status of the Land Snails of Samoa through the 20 th Century. Biological Conservation 110:55 65. Cowie, Robert H., and J. A. Grant Mackie 2004 Land Snail Fauna of Me Aure Cave (WMD007), Moindou, New Caledonia: Human Introductions and Faunal Change. Pacific Science 58:447 460. Dixon, B. D. Soldo and Charles C. Christen. 1997 Radiocarbon Dating Land Snails and Polynesian Land Use on the Island of Hawaii Archaeology 6:52 62. Evans, J. G. 1972 Land Snails in Archaeology: With Special Reference to the British Isles Seminar Press, London. Fagerstrom, J. A. 1964 Fossil Communities in Paleoecology: Their Recognition a nd Significance. Geological Society of America Bulletin 75:1197 1216.

PAGE 161

161 Fisher, R.A., A.S. Corbet, and C.B. Williams 1943 The Relation between the Number of Species and the Number of Individuals in a Random Sample of an Animal Population. Journal o f Animal Ecology 12:42 58. Fowler, Melvin L. 1959 Summary Report of Modoc Rock Shelter 1952, 1953, 1955, 1956. In Illinois State Museum Report of Investigations, No. 8, Springfield. Fridley, Jason D., Robert K. Peet, Eddy van der Maarel, and Jo H. Willems 2006 Integration of Local and Regional Species Area Relationships from Space Time Species Accumulation. The American Naturalist 168(2):133 143. 1995 Effects of Soil Acidifi cation on Forest Land Snails. Ecological Bulletins 44:259 270. Goodfriend, Glenn A. 1986 Variation in Land Snail Shell Form and Size and its Causes: A Review. Systematic Zoology 35:204 223. Goodrich, Calvin 1943 The Walker Beecher Paper of 1876 on the Mollusca of the Ann Arbor Area, pp. 1 26. In Occasional Papers of the Museum of Zoology, University of Michigan No. 475 University of Michigan Press. Ann Arbor, Michigan. Gotelli, Nicholas J., and Robert K. Colwell 2001 Quantifying B iodiversity: Procedures and Pitfalls in the Measurement and Comparison of Species Richness. Ecology Letters 4:379 391. Greenwood, K. L., and B. M. McKenzie 2001 Grazing Effects on Soil Physical Properties and the Consequences for Pastures: A Revi ew. Australian Journal of Experimental Agriculture 41:1231 1250. Harms, Sheena 2008 Terrestrial and Aquatic Mollusks as Environmental Indicators at the Brogley Rockshelter, Grant County, Wisconsin. Bachelors Thesis, University of Wisconsin La C rosse. Hellman, Jessica J., and Gary W. Fowler 1999 Bias, Precision and Accuracy of Four Measures of Species Richness. Ecological Applications 9:824 834. Hodder, Ian 1991 Interpretive Archaeology and Its Role. American Antiquity 56:7 18.

PAGE 162

162 Hof, Anouschka R., and Paul W. Bright 2010 The Impact of Grassy Field Margins on Macro Invertebrate Abundance in Adjacent Arable Fields. Agriculture, Ecosystems and Environment 139:280 283. Hotopp, Kenneth P. 2002 Land Snails and Soil Calc ium in Central Appalachian Mountain Forest. Southeastern Naturalist 1:27 44. Hotopp, Kenneth P., Timothy A. Pearce, Jeffrey C. Nekola and Kathleen Schmidt 2010 New Land Snail (Gastropoda: Pulmonata) Distribution Records for New York State. Pro ceedings of the Academy of Natural Sciences of Philadelphia 159:25 30. Hubricht, Leslie 1985 The Distributions of the Native Land Mollusks of the Eastern United States In Fieldiana, Zoology New Series No. 24 Field Museum of Natural History. H urlbert, S. H. 1971 The Nonconcept of Species Diversity: A Critique and Alternative Parameters. Ecology 52:577 586. Hylander, Kristoffer, Christer Nilsson, and Tove Gothner 2004 Effects of Buffer Strip Retention and Clearcutting on Land Snai ls in Boreal Riparian Forests. Conservation Biology 18:1052 1062. Jones, Charles C., Jr. 1861 Monumental Remains of Georgia. John M. Cooper and Company, Savannah, Georgia. Keating, Kim A., James F. Quinn, Michael A. Ivie, and LaDonna L. Ivie 1998 Estimating the Effectiveness of Further Sampling in Species Inventories. Ecological Applications 8:1239 1249. Kirch, Patrick V., Carl C. Christensen and David W. Steadman 2009 Subfossil Land Snails from Easter Island, Including Hotumatua anakenana new genus and species (Pulmonata: Achatinellidae). Pacific Science 63:105 118. Kirch, Patrick V., and Joanna Ellison 1994 Palaeoenvironmental Evidence for Human Colonization of Remote Oceanic Islands. Antiquity 68:310 322. Kirch, Pa trick V., J.R. Flenley, and D.W. Steadman 1991 A Radiocarbon Chronology for Human Induced Environmental Change on Mangaia, Southern Cook Islands, Polynesia. Radiocarbon 33:317 328.

PAGE 163

163 Kohn, Alan J. and Ingela Arua 1999 An Early Pleistocene Molluscan Assemblage from Fiji: Gastropod Faunal Composition, Paleoecology and Biogeography. Paleogeography, Paleoclimatol ogy, Paleoecology 146:99 145. Klippel, Walter E. and William B. Turner 1991 Terrestrial Gastropods from Glade Sere and the Hayes Shell Midden in Middle Tennessee. In Beamers, Bobwhites, and Blue Points: Tributes to the Career of Paul W. Parmalee edited by J. R. Purdue, W. E. Klippel and B. W. Styles, pp. 189 213. Springfield: Illinois State Museum Papers. La Rocque, Aurele 1970 Pleistocene Mollusca of Ohio, Part 4 Terrestrial Gastropoda. Ohio Division of Geological Survey Bulletin 246. Labaune, Corinne, and Frederic Magnin 2001 Land Snail Communities in Mediterranean Upland Grasslands: The Relative Import ance of Four Sets of Environmental and Spatial Variables. Journal of Molluscan Studies 67:463 474. 2002 Pastoral Management vs. Land Abandonment in Mediterranean Uplands: Impact on Land Snail Communities. Global Ecology & Biogeography 11:237 245. Leonard, A. Byron 1959 Handbook of Gastropods in Kansas. The University of Kansas Department of Zoology State Biological Survey. Topeka, Kansas. MacKenzie, Darryl I. 2005 What are the Issues with Presence Absence Data for Wildlife Manag ers? The Journal of Wildlife Management 69:849 860. Martin, Scott M. 2000 Terrestrial Snails and Slugs (Mollusca: Gastropoda) of Maine. Northeastern Naturalist 7:33 88. Matteson, Max R. 1959 Snails in Archaeological Sites. American Anthro pologist, New Series 61:1094 1096. 1960 Reconstruction of Prehistoric Environments through the Analysis of Molluscan Collections from Shell Middens. American Antiquity 26:117 120. Meyrick, R. A. and P. F. Karrow 2007 Three Detailed, Radio carbon Dated, Holocene Tufa and Alluvial Fan Mollusc Successions from Southern Ontario: The First in Northeastern North America. Paleogeography, Paleoclimatology, Paleoecology 243:250 271.

PAGE 164

164 Morrison, J. P. E. 1942 Preliminary Report on Mollusks F ound in the Shell Mounds of the Pickwick Landing Basin in the Tennessee River Valley. In Bureau of American Ethnology Bulletin 129 pp. 337 392. Washington. Nekola, Jeffrey C. 2003 Large Scale Terrestrial Gastropod Community Composition Patterns in the Great Lakes Region of North America. Diversity and Distributions 9:55 71. 2004 Terrestrial Gastropod Fauna of Northeastern Wisconsin and the Southern Upper Peninsula of Michigan. American Malacological Bulletin 18:21 44. 2005 Geograp hic Variation in Richness and Shell Size of Eastern North American Land Snail Communities. Records of the Western Australian Museum Supplement 68:39 51. 2009 Tiny Jewels: An Introduction to Pupillid Taxonomy, Ecology, and Collection. American Co nchologist 37:22 27. 2010 Acidophilic Terrestrial Gastropod Communities of North America. Journal of Molluscan Studies 76:144 156 Nekola, Jeffrey C., Gary M. Barker, Robert A. D. Cameron and Beata M. Pokryszko n.d. Latitudinal and Longitu dinal Variation of Body Size in Land Snail Populations and Communities. In Global Patterns of Body Sizes edited by F. Smith and K. Lyons. University of Chicago Press (in press). Nekola, Jeffrey C. and Matthew Barthel 2002 Morphometric Analysis of Carychium Exile and Carychium Exiguum in the Great Lakes Region of North America. Journal of Conchology 37:515 531 Nekola, Jeffrey C. and Brian F. Coles 2010 Pupillid Land Snails of Eastern North America. American Malacological Bulletin 28:29 57. Nica, Dragos, Marian Bura, Iosif Gergen, Monica Harmanescu, and Despina Maria Bordean 2012 Bioaccumulative and Conchological Assessment of Heavy Metal Transfer in a Soil Plant Snail Food Chain. Chemistry Central Journal 6:55. 2002 Late Archaic Freshwater Mussel Exploitation along the Middle Savannah River. M.A. Thesis. Department of Anthropology. University of Florida, Gainesville, Florida. Oliver, Ian, Denys Garden, Penelope J. Greenslade, Bronwyn Haller, Denis Rodgers Owen Seeman, and Bill Johnston 2005 Effects of Fertiliser and Grazing on the Arthropod Communities of a Native Grassland in South eastern Australia. Agriculture, Ecosystems and Environment 109:323 334.

PAGE 165

165 Peacock, Evan and Jochen Gerber 2008 Using Land Snails and Freshwater Mussels to Chart Human Transformation of the Landscape: An Example from Northeast Mississippi, U.S.A. In Case Studies in Environmental Archaeology 2 nd edition edited by Reitz, Elizabeth J., C. Mar garet Scarry and Sylvia Scudder, pp. 123 141. Springer Science Business Media. Peacock, Evan and Rebecca Melsheimer 2003 Terrestrial Gastropods from Archaeological Contexts in the Black Be lt Province of Mississippi. In Blackland Prairies of the Gulf Coastal Pla in: Nature, Culture, and Sustainability edited by Evan Peacock and Timothy Schauwecker, pp. 27 47. University of Alabama Press. Alabama. Peacock, Evan, Janet Rafferty and S. Homes Hogue 200 5 Land Snails, Artifacts and Faunal Remains: Understand ing Site Formation Processes at Prehistoric/Protohistoric Sites in the Southeastern United States. In Archaeomalacology: Molluscs in Former Environments of Human Behaviour edited by Daniella E. Bar Yosef Mayor, pp. 6 17. Durham. Perry, Richard, and W allace Arthur 1991 Shell Size and Population Density in Large Helicid Land Snails. Journal of Animal Ecology 60:409 421. Phillips, Steven J., Miroslav Dudik, Jane Elith, Catherine H. Graham, Anthony Lehmann, John Leathwick, and Simon Ferrier 2009 Sample Selection Bias and Presence Only Distribution Models: Implications for Background and Pseudo Absence Data. Ecological Applications 19:181 197. Pi gati, Jeffrey S., Jason A. Rech and Jeffrey C. Nekola 2010 Radiocarbon Dating of Smal l Terrestrial Gastropod Shells in North America. Quaternary Geochronology 5:519 532. Pilsbry, Henry A. 1939 Land Mollusca of North America (North of Mexico), Volume I Part 1 Monographs of the Academy of Natural Sciences of Philadelphia 3:1 574 1940 Land Mollusca of North America (North of Mexico), Volume I Part 2. Monographs of the Academy of Natural Sciences of Philadelphia 3:575 994. 1946 Land Mollusca of North America (North of Mexico), Volume II Part 1. Monographs of the A cademy of Natural Sciences of Philadelphia 3:1 520. 1948 Land Mollusca of North America (North of Mexico), Volume II Part 2. Monographs of the Academy of Natural Sciences of Philadelphia 3:521 1113. Plotkin Joshua B. and Helene C. Muller Landau 2002 Sampling the Species Composition of a Landscape. Ecology 83:3344 3356.

PAGE 166

166 Preece, R. C. 1998 Impact of Early Polynesian Occupation on the Land Snail Fauna of Henderson Island, Pitcairn Group (South Pacific). Philosophical Transactions of the Royal Society of London, Series B Biological Sciences 353:347 368. Reddy, Sushma, and Liliana M. Davalos 2003 Geographical Sampling Bias and its Implications for Conservation Priorities in Africa. Journal of Biogeography 30:1719 1727. Regol i, Francesco, Stefania Gorbi, Daniele Fattorini, Sara Tedesco, Alessandra Notti, Nicola Machella, Raffaella Bocchetti, Maura Benedetti, and Francesco Piva 2006 Use of the Land Snail Helix aspersa as Sentinel Organism for Monitoring Ecotoxicological Effects of Urban Pollution: An Integrated Approach. Environmental Health Perspectives 114:63 69. Sassaman, Kenneth E. n.d. Radiocarbon data on file at the Laboratory of Southeastern Archaeology, Department of Anthropology, University of Florida, Gainesville. 1993a Early Pottery in the Southeast: Tradition and Innovation in Cooking Technology. University of Alabama Press, Tuscaloosa. 19 93b Mims Point 1992: Archaeological Investigations at a Prehistoric Habitation Site in the Sumter National Forest South Carolina. Savannah River Arc haeological Research Papers 4. Occasional Papers of the Savannah River Archaeological Research Program, South Carolina Institute of Archaeology and Anthropology, University of South Carolina, Columbia. 1996 Excavations in the Stallings Midden at the Ed Marshal Site (38ED5). Paper presented at the 22 nd Annual Conference on South Carolina Archaeology, Columbia, SC 2001 Grant Proposal submitted to the National Science Foundation. 2006 People of the Shoals: Stallings Culture of the Savannah River Valley University Press of Florida. Gainesville, Florida. The Archaeological Record 8:6 6 40. Sassaman, Kenneth E., Meggan E. B lessing, and Asa R. Randall 2006 Stallings Island Revisited: New Evidence for Occupational History, Community Pattern, and Subsistence Technology. American Antiquity 71:539 565. Sassaman, Kenneth E., M.A. Cabak, and J.C. Gillam 1996 Distr ibutional Archaeology in the Aiken Plateau Paper presented at the 53 rd Annual Meeting of the Southeaster Archaeolo gical Conference, Birmingham, Alabama

PAGE 167

167 Sassaman, Kenneth E., I.R. Daniels, Jr. and C.R. Moore 2002 G. S. Lewis East: Early and Late Archaic Occupations along the Savannah River, Aiken County, South Carolina. In Savannah River Archaeological Research Papers 1 SRRP#12 Occassional Papers of the Savannah River Archaeological Research Program, South Carolina Institute of Archaeolog y and Anthropology, University of South Carolina, Columbia. Schilthuizen, Menno, Angelique van Til, Merijn Salverda, Thor Seng Liew, S. Sheena James, Berjaya bin Elahan, and Jaap J. Vermeulen 2006 Microgeographic Evolution of Snail Shell Shape and Predator Behavior. Evolution 60:1851 1858. Secrest, Michele F., Michael R. Willig, and Lottie L. Peppers 1996 The Legacy of Disturbance on Habitat Associations of Terrestrial Snails in the Luquillo Experimental Forest, Puerto Rico. Biotropica 28:5 02 514. Solem, Alan 1964 New Records of New Caledonian Nonmarine Mollusks and an Analysis of the Introduced Mollusks. Pacific Science 28:130 137. 1978 Classification of the Land Mollusca In Fretter, V. & J. Peake, Pulmonates, Volume 2A S ystematics, Evolution, and Ecology pp. 49 67. Academic Press, London. Stanisic, John, Robert A. D. Cameron, Beata M. Pokryszko and Jeffrey C. Nekola 2007 Forest Snail Faunas from S. E. Queensland and N. E. New South Wales (Australia): Patterns of Local and Regional Richness and Differentiation. Malacologia 49:445 462. Suominen, Otso 1999 Impact of Cervid Browsing and Grazing on the Terrestrial Gastropod Fauna in the Boreal Forests of Fennoscandia. Ecography 22:651 658. Tattersfield, P. 1981 Density and Environmental Effects on Shell Size in Some Sand Due Snail Populations. Biological Journal of the Linnean Society 16:71 81. Thurman, Carter F., Lane P. Shackleton, and David G. Haskell 2008 Does the Density of Dead Shells Predict the Density of Living Anquispira cumberlandiana Lea 1840 (Gastropoda: Discidae)? American Midland Naturalist 159:478 481. Tojo, Bunji and Terufumi Ohno 1999 Continuous Growth Line Sequences in Gastropod Shells. Palaeogeography, Palaeoclim atology, Palaeoecology 145:183 191.

PAGE 168

168 Tyre, Andrew J., Brigitte Tenhumberg, Scott A. Field, Darren Niejalke, Kirsten Parris, and Hugh P. Possingham 2003 Improving Precision and Reducing Bias in Biological Surveys: Estimating False Negative Erro r Rates. Ecological Applications 13:1760 1801. Vagvolgyi, Joseph 1975 Body Size, Aerial Dispersal, and Origin of the Pacific Land Snail Fauna. Systemic Zoology 24:465 488. Waggoner, Janet, Stephanie A. Clark, Kathryn E. Perez and Charles Lydea rd 2006 A Survey of Terrestrial Gastropods of the Sipsey Wilderness (Bankhead National Forest), Alabama. Southeastern Naturalist 5:57 68. Wareborn, Ingvar 1969 Land Molluscs and Their Environments in an Oligotrophic Area in Southern Sweden. Oikos 20:461 479. White, E. P. 2004 Two Phase Species Time Relationships in North American Land Birds. Ecology Letters 7:329 336. Williamson, P., R. A. D. Cameron, and M. A. Carter 1976 Population Density Affecting Adult Shell Size of Sn ail Cepaea nemoralis L. Nature 263:496 497. Yates, Timothy 1986 Studies of Non Marine Mollusks for the Selection of Shell Samples for Radiocarbon Dating. Radiocarbon 28:457 463.

PAGE 169

169 BIOGRAPHICAL SKETCH Christine Snyder was born and raised in Richmond Virginia and graduated from Varina High School in 1995. She earned a Bachelor of Science degree in biology, with minors in environmental s tudies a nd political s cience in 1995 and a Master of Science degree in b iology in 1999 from Virginia Commonwealth University. Upon graduating, Christine became employed at the Virginia D epartment of Forensic Science in the Biology Section and transferred to the Latent Print Section, where she was trained and later hired as a Forensic Scientist in footwear and tire i mpression analysis While in training, she also worked as a Per Diem Medicolegal Death Investigator for the Office of the Chief Medical Examiner for the Commonwealth of Virginia. In late 2002, Christine moved to Jacksonville, Florida to work at the Flori da Department of Law Enforcement as a Crime Laboratory Analyst. In 2003, she mov ed Office in Sanford, Florida where she was hired as a Crime Sc ene Analyst and F ootwear and T ire I mpression E xaminer. In 2005, she enrolled a t the University o f Florida to pursue a Ph.D. in interdisciplinary ecology with a concentrati o n in a nthropology. Christine is a C ertified Senior Crime Scene Analyst and C ertified Footwear Examiner through the International Association for Identification ( IAI). She also serve s as the Chair of the Standards Committee of the Scientific Working Group for Shoeprint and Tire Tread Evidence (SWGTREAD), and as a member of the Florida Emergency Mortuary Operations Response System (FEMORS), the Crime Scene Certific ation Board for the IAI the Region 2 Director for the Florida Division of the IAI (FDIAI), the Crime Scene Certification Committee for the FDIAI and the Footwear and Tire Track Identification Committ ee for the FDIAI She has also worked as a contract gr ant reviewer for the National Institute of Justice and as an adjunct instructor with the

PAGE 170

170 Institute of Police Management and Technology at the University of North Florida Christine graduated with her Ph.D. in 2012 and continues to work in the forensic sci ence field at the local, state and national level. She plans to pursue her interests in the field of malacology, determine the possibility and potential of malacology in forensics, and to utilize her ecology background to inform others about their impact on the environment.