The economics of lithic-resource use in South-Central Florida

MISSING IMAGE

Material Information

Title:
The economics of lithic-resource use in South-Central Florida
Physical Description:
xxviii, 702 leaves : ill. ; 29 cm.
Language:
English
Creator:
Austin, Robert James, 1949-
Publication Date:

Subjects

Subjects / Keywords:
Anthropology thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Anthropology -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1997.
Bibliography:
Includes bibliographical references (leaves 641-702).
Statement of Responsibility:
by Robert James Austin.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 028623040
oclc - 39476601
System ID:
AA00013597:00001

Full Text












THE ECONOMICS OF LITHIC-RESOURCE USE
IN SOUTH-CENTRAL FLORIDA















By

ROBERT JAMES AUSTIN















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

1997




















THIS TITLE IS BOUND IN TWO PARTS.

THIS IS PT. 1.















ACKNOWLEDGMENTS


I am indebted to many people for helping me see this dissertation through to completion. First

and foremost, I would like to thank the members of my supervisory committee: Dr. Jerald T. Milanich

(Chair), Dr. Steven A. Brandt, Dr. William F. Keegan, Dr. William H. Marquardt, Dr. Douglas S. Jones,

and Dr. Sam B. Upchurch. All of these individuals were exceptionally generous with their time and

never failed to offer me encouragement, advice, and constructive criticism. Two members deserve

special recognition: Jerald T. Milanich, who served as committee chair, and Sam B. Upchurch, who

served as an outside member on my committee. When I originally approached Jerry Milanich about

entering the graduate program at the University of Florida, I did so with a certain amount of trepidation

since I had been out of the academic circle for nearly nine years. To my pleasant surprise he not only

expressed immediate interest and enthusiasm about my proposed research topic, he agreed to serve as

my sponsor and committee chair. I could not have asked for a better advisor. His constant

encouragement and support, his not-so-subtle prodding to stay focused and finish guickdy, and his

exceptional ability to cut through the seemingly endless tangle of university red tape, have made the past

six years a much easier and rewarding experience than it might otherwise have been. Sam Upchurch

also deserves my heartfelt thanks for agreeing to serve as my geological mentor. When I began this

academic journey, Sam was on the faculty of the University of South Florida's Geology Department;

however, he eventually left academia for the world of private consulting. While this change in careers

increased his personal work load significantly, he continued to take time from his busy schedule to

answer my many questions, offer much needed advice, and look at and talk about rocks.



ii









I would also like to thank Dr. E. C. Pirkle for serving briefly on my committee before his

retirement; Dr. Barbara Purdy for sharing with me her insights on lithics and for an especially enjoyable

day gathering chert samples from the CCA quarry site: Dr. Albert C. Goodyear for many hours of

conversation which have helped shape my own ideas about the economics of lithic industries: Dr.

Lorena Madrigal for statistical advice; and Richard Estabrook, who has served as a sounding board for

many of the ideas presented in this dissertation and often has provided intelligent and well-reasoned

critiques. Archaeologists Jerry Westphal and Jay Hardman helped me during the field work portion of

this project, often with little or no monetary reimbursement. Scott Mitchell drew all of the artifact

illustrations that appear in this dissertation as well as the fossils that appear in Figure 18. Dawn Van

DePutte drew all of the maps. Members of the Kissimmee Valley Archaeological and Historical

Conservancy assisted me in the field during the testing of several sites in Highlands County. Many

individuals also allowed me to record and analyze their lithic artifact collections. I would especially like

to thank Anne Reynolds, April Felt, Don McClure, Mike Mosier, and Ruby Stevens. Anne and Charles

Reynolds, Gary Montsdeoca, Terry Mock, and Mike Wells graciously allowed me access to their

properties for the purpose of conducting test excavations. Without the help of these generous people

I would have had a much harder time obtaining the data necessary to conduct this research. I would also

like to thank Anne and Charles Reynolds, Jim and Anne Fitch, and Creighton and Bette Northrop for

sharing their homes with me. Meeting and working with these people has been a singularly rewarding

experience. I am proud to call them my friends.

I would be terribly remiss if I did not acknowledge the debt of gratitude I owe to my colleagues

and coworkers at Janus Research, and particularly Ken Hardin, president and owner of the company,

for putting up with my erratic schedule and occasional leaves of absence during the past six years.

When Inotified Ken that I had decided to return to school to pursue a Ph.D., I fully expected to be out




iii









of a job. Instead, he let me arrange a flexible work schedule that allowed me to attend classes and

devote time to conducting research and writng my dissertation.

Funding assistance was provided by a College of Liberal Arts and Sciences Dissertation

Fellowship and a Charles H. Fairbanks Scholarship from the University of Florida. The Kissimmee

Valley Archaeological and Historical Conservancy also provided funds to pay for some of my own

expenses during field trips to Highlands County, for special analyses such as radiocarbon dates and soils

analysis, and for wages and expenses for assistants Westphal and Hardman. Some of the data used in

this research came from cultural resource management projects conducted under my direction over the

past several years. In particular I would like to acknowledge the efforts of Paul Ebersbach, Natural

Resources Manager on the Avon Park Air Force Range, who has been instrumental in managing and

preserving the large number of prehistoric and historic archaeological sites on the Range.

Finally, I would like to express my greatest appreciation to my wife, Karen, for her love and

support during the past six years, and to my parents, Jack and Jane, and my brother Lenny, for always

believing in me.























iv
















TABLE OF CONTENTS


Pase

ACKNOWLEDGMENTS ............................. ... ii

LIST OF TABLES .................. .............. .... viii

LIST OF FIGURES ....................................... xiv

KEY TO ABBREVIATIONS ............................. .. xxvi

ABSTRACT ................. ....... ......... .. xxvii

CHAPTERS

1 INTRODUCTION ...... ............ ......... ... 1

2 THEORETICAL FRAMEWORK ................... ... 8

Pattern, Scale, and Culture Change ...................... 10
The Conceptual Basis ......... ................ .. .. 18
Economics of Lithic-Resource Use .......................... 47

3 DESCRIPTION OF THE STUDY AREA ..... ... .............. 58

Physical Environment ...................... . . 60
Paleoenvironmental Conditions ..... ........... ...... 79

4 GEOLOGY AND CHERT RESOURCES ..... . ........ 91

Geologic History . . . . . . . . . . . . . .. 91
Lithostratigraphy ....... .............. ... .... 93
Chert Origins and Distribution .... ............ ..... .. 102

5 PREHISTORY OF THE STUDY AREA .. ................. 112

Integrative Concepts: Tradition and Horizon .................. 112
Prehistoric Occupation of the Study Area ............. .... .. 115

6 HYPOTHESES AND TEST IMPLICATIONS .. . ......... 149


v










Strategies for Coping with Procurement Risk . .. .. .. ..... . 150
Strategies for Coping with Procurement Costs . .. .. ... . 156
Strategies for Coping with Opportunity Costs . .... ..... . 161
Summary.... . . . . . . ..... 163

7 METHODS ..... ... .. ........ ..... . . .... 166

Database ........................... . . . . .... 166
Data Collection . . . . ................ .. 175
Control Variables ..................... .. . . . ...... 177
Analysis M ethods . .. ............. . . . 185
Experimental Methods . . .... ................ . 191
Raw-Material Provenience Identification . . . ......... . . 194
Statistical Analysis of Data . . ............... . . ...... 223

8 DIFFERENTIATING BETWEEN TOOL PRODUCTION
AND CORE REDUCTION IN LITHIC WASTE-FLAKE
ASSEMBLAGES: ANALYSIS OF EXPERIMENTAL DATA ........... 230

Sullivan-Rozen Analysis ....... ........ ...... ... 231
Flake-Size Distribution ................................ 259
Conclusion .................................... .. 282

9 STRATEGIES FOR COPING WITH PROCUREMENT RISK .......... 284

Raw-Material Acquisition ........ ......... ............. 284
Raw-Material Diversification .. .. .................... 317
Scavenging, Reuse, and Recycling .......... ...... ......... 346
Storage ...................... ... ........ ... 359
Conclusion ......... ................... ........ 364

10 STRATEGIES FOR COPING WITH PROCUREMENT COSTS ......... 368

Minimization of Transport Costs ................. ..... ... 368
Lithic Economizing ................................ 405
Minimization of Production Costs .............. ....... 444
Reduction of Lithic Demand ............................. 448
Conclusion . . . . . .. . . . . . . . . . . . 459

11 STRATEGIES FOR COPING WITH OPPORTUNITY COSTS .. .... 464

Tool Diversity ................ ............. .. 464
Functional Diversity . ... ........ ........ ...... 469
Organizational Structure: Tool Assemblages ................... 477
Organizational Structure: Debitage Assemblages ................. 483
Conclusion .... ....... .. ..... .... ... ........ 565

12 SUMMARY AND CONCLUSIONS ............. ........ 570

vi









Overview of the Study. . . . . . . 570
Summary and Discussion of Research Results . .... .. . . 574
Economics and the Evolutionary Process ................... . . 592
Conclusions .......... .. ... . . . . . 600

APPENDICES

A INVENTORIES OF ARTIFACT CLASSES AND RAW-MATERIAL
CATEGORIES FOR ARCHAIC PERIOD ASSEMBLAGES .......... 604
B INVENTORIES OF ARTIFACT CLASSES AND RAW-MATERIAL
CATEGORIES FOR POST-ARCHAIC PERIOD ASSEMBLAGES ....... 611
C INVENTORIES OF ARTIFACT CLASSES AND RAW-MATERIAL
CATEGORIES BY INTRASITE FEATURE AT FORT CENTER (8GL13) ... 618
D RAW-MATIERAL IDENTIFICATIONS FOR HAFTED BIFACES ....... 625
E DIMENSIONAL DATA FOR COMPLETE HAFTED BIFACES ......... 633

REFERENCES CITED . ......... .... ....... ... . . 641

BIOGRAPHICAL SKETCH ................................... 702































vii















LIST OF TABLES


Labk Pags

1 Marine terraces in Florida as identified by various geologists. . . . . . 64

2 Paleoenvironmental history of south-central Florida . . . . . ....... 89

3 Lithostratigraphic units present in the study area (modified from Scott 1992:Table 1). 94

4 Curve-fitting models used to identify different forms of direct and indirect
procurement systems (after Findlow and Bolognese 1982) . . . . ..... 153

5 Model of expected organizational responses for coping with the problems of
procurement cost, procurement risk, and opportunity costs for mobile and sedentary
populations occupying a region of limited lithic resources . . . . 164

6 Summary data for sites used in this study . . . ....... . ... 169

7 Comparison of excavated sites in the study sample using Rafferty's (1985) criteria of
sedentariness .. . . .. . . . . . 182

8 Chert types and associated quarry clusters in Florida . . . . 217

9 Summary of experimental replications ................... . .. 232

10 Comparison of flake breakage patterns between different reduction strategies. Except
as noted, all experimental assemblages were sifted through 6.4 mm hardware cloth
prior to analysis .................. .................. .. 239

11 Results of chi-square tests of independence between Mauldin and Amick's
(1989:Table 3) composite biface waste-flake assemblage and the core-reduction flake
assemblages of Prentiss and Romanski (1989:Table 1) and Kuijt et al. (1995:Table
2) . . . . . . .. . . . . . ... .. . . 246

12 Comparison of flake-breakage patterns between different strategies for 15
experimentally produced flake assemblages. All experimental assemblages were
sifted through 6.4 mm hardware cloth prior to analysis ... ......... 247

13 Summary Sullivan-Rozen data for experimental biface-, uniface-, and core-reduction
waste-flake assemblages ......... . . . . 248

viii









14 Kolomogorov-Smimofftest of cumulative flake frequencies between biface-, uniface-,
and core-reduction strategies. Significant differences (p .05) are indicated by plus
signs; minus signs indicate no significant differences . . . . . 249

15 Three-way classification of experimental patterned-tool- and core-reduction debitage
assemblages . . .. . . . . . . 252

16 Two-way classification of experimental tool and core debitage assemblages . 252

17 Three-way classification of experimental patterned-tool- and core-reduction debitage
assem blages . . . . . . . . . . . . . . . . . . . 254

18 Summary Sullivan-Rozen data for experimental patterned tool (biface and uniface),
experimental core (large and small) reduction, and simulated mixed (patterned tool
and core) waste-flake assemblages ................... ...... . 255

19 Discriminant-function coefficients for two-group and three-group separations of
experimental debitage .......... . . . . . . . 258

20 Flake-size data for random-fracture experiments ...... . . . 266

21 Flake-size data for experimental tool and core replications. All assemblages sifted
through 6.4 mm screens . . ... .. ........ . . 269

22 Flake-size data for simulated assemblages . . . .......... . . 274

23 Raw data used in the calculation of regression lines and Analysis of Variance
(A N O V A ) . . . . . . . . . . . . . . . . . 280

24 Results of Analysis of Variance testing the equality of slopes between experimental
core-reduction and biface-reduction flake-size distributions .. . . .... 280

25 Results of Analysis of Variance testing the equality of slopes between experimental
and simulated flake assemblages. Bold entries indicate rejection of the null
hypothesis ofno difference between slopes .. . . . . . .. 281

26 Data on distance in kilometers to nearest source location (D) and abundance of raw-
material types (quarry clusters) in Archaic and post-Archaic assemblages. Except
where noted, abundance is calculated as a percent of all lithics (tools, cores, and
debitage) present in an assemblage .... . . .. .. . .. 286

27 Summary of distance-decay patterns for lithic raw materials by cultural period . 302

28 Flake-to-core (F:C) ratios by raw-material categories for Archaic and post-Archaic
assemblages in the Peace River region . . . .......... .. . 305

29 Flake-to-core (F:C) ratios by raw-material categories for Archaic and post-Archaic
assemblages in the Kissimmee region . ..... . . . 306

ix









30 Flake-to-core (F:C) ratios by raw-material categories for Archaic and post-Archaic
assemblages in the Hillsborough River region . . . . ... 307

31 Flake-to-core (F:C) ratios by raw-material categories for lithic assemblages from
Pineland (8LL33, 36, 37). Fort Center (8GL13), and Yat Kitischee (8P11753) 309

32 Cross-tabulation of hafted-biface types by intrasite features at Fort Center (8GL13). 312

33 Cross-tabulation of thermally altered bifaces between intrasite features at Fort Center
(8G L 13 ) . .. . .. . . . . . . . .. . . . . . . 3 13

34 Cross-tabulation of thermally altered silicified coral bifaces between intrasite features
atFortCenter(8GL13) ................... . . . . ..... 313

35 Cross-tabulation of raw-material types (quarry clusters) by cultural components and
geographic regions ..... . ..... ................ ..... 318

36 Lithic raw materials (quarry clusters) present at Yat Kitischee (8PI1753) . ... 323

37 Cross-tabulation of temporally diagnostic hafted bifaces by raw-material types (quarry
clusters) ......... ...... .... . .. . 326

38 Richness (R) and evenness (E) values for temporally diagnostic hafted bifaces . 328

39 Raw-material types (quarry clusters) represented in regional Archaic and post-Archaic
lithic assemblages . . . . . .... . . . ...... 331

40 Richness (R) and evenness (E) values for regional Archaic and post-Archaic lithic
assemblages . . . . ..... . ........ . . .... 333

41 Raw-material types (quarry clusters) represented by the Archaic-style hafted bifaces
at Fort Center (8GL13) and Pineland (8LL33, 36, 37) . . . . . 334

42 Richness (R) and evenness (E) values for Archaic and post-Archaic assemblages with
non-chert materials included .............. ............... .. 335

43 Presence of thermal alteration in Archaic and post-Archaic assemblages . . . 337

44 Estimates of biface production costs . .. .... .. . .. 338

45 Effect of increasing transportation costs on the average, total, and marginal returns
for biface-production strategies that employ unaltered and thermally altered lithic
materials .............. . ........... . ... 342

46 Data on the numbers of reused and recycled bifaces in Archaic, post-Archaic, and
surface-collected assemblages ... .. . . . 350



x









47 Condition and size data for complete and broken bifaces from Archaic and post-
Archaic assemblages . . .................. .. . . 352

48 Summary condition and size data and ratios for composite Archaic and post-Archaic
assemblages .... ............ . . . . . 354

49 Results of pair-wise chi-square tests of independence comparing the frequencies of
broken and complete bifaces among Archaic and post-Archaic assemblages. Bold
entries indicate a significant difference (p s .05) . . . . . 355

50 Results of pair-wise chi-square tests of independence comparing the frequencies of
small and large biface fragments among Archaic and post-Archaic assemblages. Bold
entries indicate a significant difference (p s .05) ..... . . . . . 357

51 Core size and weight data ............... . . . . . . 371

52 Size-index data for hafted bifaces . . ... . . . . 384

53 Comparison of mean size indices of hafted bifaces at Fort Center (8GL 13) with the
mean size indices of all other hafted bifaces in the study collection . . . ... 389

54 Raw data on cortex (C) versus noncortex (NC) flakes for Archaic and post-Archaic
assem blages . . . . . . . . . . . . . . . . . . 395

55 Mean size indices, mid-blade width:basal-blade width (MBW:BBW) ratios, and blade
length:basal-blade width (BL:BBW) ratios for hafted-biface types represented in the
study collection ........ . . . . . . . . 409

56 Ratios of retouched to unretouched tools (R:U) for Archaic and post-Archaic
assemblages ........... ... . .. . . ........... 429

57 Comparison of ratios of retouched (R) to unretouched (U) tools for composite Archaic
and post-Archaic assemblages by geographic region . . . . . .. 431

58 Comparison of ratios of retouched (R) to unretouched (U) tools by intrasite
proveniences at Fort Center (8GL13) ......... . . . .. . 431

59 Results of pair-wise chi-square tests of independence comparing the frequencies of
retouched and unretouched tools among Archaic and post-Archaic assemblages. Bold
entries indicate significant differences (p .05) . .... ... . . 433

60 Mean data on use intensity of non-biface tools (utilized flakes, modified flakes,
unifaces, and microliths) from Archaic and post-Archaic assemblages . . . 435

61 Results of Student's t tests (two-tailed) comparing mean numbers of utilized
functional units between assemblages. Bold entries indicate significant differences
(p .05). . . .... . . . . . .. ........ . ..... . 436


xi









62 Ratios of flakes (F) to flake tools (FT) for Archaic and post-Archaic assemblages. 439

63 Comparison of flakes (F) to flake tool (FT) ratios for composite Archaic and post
Archaic assemblages by geographic region . . . . . ... 441

64 Results of pair-wise chi-square tests of independence comparing the frequencies of
flakes and flake tools between regional Archaic and post-Archaic assemblages. Bold
entries indicate significant differences (p .05) . . . .... . . . 442

65 Core types in Archaic and post-Archaic assemblages . . . . . 445

66 Lithic density calculations for Archaic and post-Archaic components in the
Kissimmee region ........ ... ........ ...... ...... 450

67 Lithic artifact density by region and cultural components . . . . . .... 451

68 Lithic density calculations for Archaic and post-Archaic components in the Peace
River region ....... ....... .. ...... .. ...... 452

69 Summary of tool classes represented in Archaic and post-Archaic assemblages
separated by region . . ............ . ... .. 465

70 Richness (R) and evenness (E) measures for Archaic and post-Archaic tool
assemblages . . . . ..... .............. . . . . 467

71 Cross-tabulation of functional tasks by tool classes for Archaic and post-Archaic
assemblages.. ........... . . . . . 470

72 Summary data on functional variability between Archaic and post-Archaic tool
assemblages ................... ......... . . . .. 475

73 Comparison of curated (C) to expedient (E) tools for Archaic and post-Archaic
assemblages ................. ............. ...... 478

74 Comparison of curated (C) to expedient (E) tool ratios for Archaic and post-Archaic
assemblages by geographic region ................... ....... .. 480

75 Results of pair-wise chi-square tests of independence comparing the frequencies of
curated and expedient tools between regional Archaic and post-Archaic assemblages.
Bold entries indicate significant differences (p s .05) . . . . . ...... 481

76 Results of Sullivan-Rozen debitage analysis for Archaic and post-Archaic
assemblages . . ..... . ...... . . . 484

77 Multivariate classification of Archaic debitage assemblages . ..... . 486

78 Multivariate classification of composite debitage assemblages from Archaic and post-
Archaic components at 8HG 18, 8HG20, 8HG27, and 8HG34 ... . . 494

xii










79 Results of Sullivan-Rozen debitage analysis for selected Archaic assemblages
controlled for raw-material type .. . ....... . .... 496

80 Multivariate classification of raw materials from selected Archaic debitage
assemblages .... .... ...... ... . . .... 498

81 Multivariate classification of post-Archaic debitage assemblages . . . . .. 506

82 Results of Sullivan-Rozen debitage analysis for selected post-Archaic assemblages
controlled for raw-material type . . . .. . . . . . 517

83 Multivariate classification of raw materials from selected post-Archaic debitage
assemblages .. ........ .. . . ...... . . 518

84 Flake-size data for Archaic and post-Archaic assemblages. Totals include utilized and
modified flakes . . ....... ...... . ................ 527

85 Results of ANOVA tests for equality of slopes between experimental and Archaic
flake assemblages ............. .. . . . . . 529

86 Results of ANOVA tests for equality of slopes between experimental flake
assemblages and composite flake assemblages from Archaic and post-Archaic
components at 8HG18, 8HG20, 8HG27, and 8HG34 . . . . ...... 536

87 Distribution of debitage between flake-size categories for selected Archaic
assemblages controlled for raw-material type . . . ..... . .. 538

88 Results of ANOVA tests for equality of slopes between experimental flake
assemblages and flake assemblages from selected Archaic assemblages controlled for
raw-material type ................... . . . . . ........ 540

89 Results of ANOVA tests for equality of slopes between experimental and post-
Archaic flake assemblages .................. ........... .. 548

90 Distribution of debitage between flake-size categories for selected post-Archaic
assemblages controlled for raw-material type .. . . . . ...... 555

91 Results of ANOVA tests for equality of slopes between experimental flake
assemblages and flake assemblages from selected post-Archaic assemblages
controlled for raw-material type ........ ... . . . . . 556

92 Flake attributes for selected raw materials from Fort Center (8GL 13) . . 563






xiii















LIST OF FIGURES


iElan page

I Map of the study area in relation to Florida . ...... . .. 5

2 Generalized supply-and-demand curve ........ ... .. .25

3 Supply-and-demand curves showing the effect of changes in demand (a) and supply
(b ) . . . . . . . . . . . . . . . . . . . . . 2 6

4 Equilibrium analysis of Chayanov's model of work versus leisure time in peasant
societies . . . .. . . . . 28

5 Elasticity of demand between two resources . ...... ....... . 30

6 Effect of price or availability on the quantity consumed of two substitutable resources.
An increase in the cost of resource A, as indicated by the steeper slope (2), results in
a large shift in the mixture of A and B consumed (from X to Y); in other words, less
of A is consumed in preference to the cheaper B .. . . . 30

7 Graph of diminishing marginal returns. As the average return per labor hour (AR)
decreases, so too does the marginal return (MR), even though total returns (TR)
continue to rise. The optimal labor effort, indicated by (a), is the point at which
average and marginal returns are at their highest .. . . ... . . 33

8 Graphic representation of diet-breadth model (after MacArthur and Pianka
1966:Figure 1): a) As search time (AS) decreases, handling time (aH) increases. The
optimal foraging strategy (A) is achieved at the point where the addition of an
additional prey species would increase handling time more than it would decrease
search time; b) Effect of decreasing handling time on the equillibrium point. Here
changes in the forager's ability to pursue, capture, and process resources more
efficiently increases diet breadth and decreases time spent per unit of energy . 35

9 Graphic illustration of the patch-choice model (after MacArthur and Pianka
1966:Figure 2): a) Resource patches are added to the foraging inventory as long as the
decrease in between-patch travel time (T) is greater than the increase in within-patch
foraging time (H). The optimal strategy (A) is achieved when the addition of an
additional patch would increase search time within patches more than it would
decrease travel time between patches . .. . . 37


xiv









10 Application of the marginal-value theorem to the patch-choice problem (after Chamov
1976). For a given patch, the rate of energy return is a function of time spent forag-
ing within the patch. As resources within a patch become increasingly rare, within-
patch foraging time increases and the rate of energy return decreases. The optimal
point at which the forager should leave the patch occurs when the within-patch rate
of return drops to the overall rate of return obtainable from the environment as a
whole when all travel, search, and handling times are considered . . . .... 38

11 Graphic representation of different resource-procurement zones .. . . 49

12 Map of the study area showing the Kissimmee River, Lake Okeechobee, and
surrounding features . .... .. ..... . . . 59

13 Geomorphic features in the study area (after White 1970) . . . . . 63

14 Hydrological relationships in the study area (modified from Parker et al. 1955:Figure
30) . . . . . . . .. . . .. . . . . . . . . . 74

15 Hypothetical size-distance relationships resulting from different types ofraw-material
acquisition and use strategies: a) direct supply-zone access by mobile groups (return
trip is immediate with no use or deposit of exhausted tools on the return) or down-the-
line exchange; b) direct supply-zone access by mobile groups in conjunction with
planned use of raw materials resulting in greater variability in discarded tool sizes; c)
direct supply zone access by logistical groups or directed exchange to distribution
centers; d) prestige exchange . . . . . . . 159

16 Locations of sites included in this study ......... . . . . ... 168

17 Polar coordinate method for describing use-wear location . . . . . ... 191

18 Examples of diagnostic fossils used in the identification of chert provenience: a)
Lepidocclina sp., Family Orbitoididae; b) Opeclinida sp., Family Orbitoididae;
c) Nummoloculina sp., Family Miliolidae; d) Dictoconus sp., Family Valvulinidae;
e) Stensioin sp., Family Peneroplidae; f) Penoplis sp., Family Peneroplidae; g)
Archaias sp., Family Peneroplidae; h) Soites sp., Family Peneroplidae; i) charophyte
oogonia. Redrawn from the following sources: Cole (1945:Plates 1, 2, 5), Cushman
(1948:Plates 14, 24, 52). Taylor (1981:Figure 4.6), and Upchurch et al.
(1982a:Figures 16c, 19a) . ....... . . . . . . 210

19 Map of quarry-cluster locations in Florida (modified from Upchurch et al.
1982a:Figure I and Goodyear et al. 1983:Figure 8) . . . . . . 216

20 Comparison of cumulative percentages of Sullivan-Rozen flake categories among
experimental assemblages from several different studies: a)Prentiss and Romanski
(1989) and Kuijt et al. (1995): b) Tomka (1989) . . . . ...... 241




xv









21 Comparison of cumulative percentages of Sullivan-Rozen flake categories: a)
patterned-tool assemblages: b) core-reduction assemblages . . ... 243

22 Comparison of cumulative percentages of Sullivan-Rozen flake categories for
experimentally produced biface, uniface, and core assemblages . . . . 249

23 Plot of discriminant-function scores for experimental flake assemblages resulting
from patterned-tool, conventional-core, and bipolar-core reduction . . . . 254

24 Plot of discriminant-function scores for experimental and simulated waste-flake
assemblages .............. . . . . . . . ..... 256

25 The "Patterson Curve" demonstrating a) the exponential form of flake-size
distributions resulting from biface reduction, and b) the straight-line distribution that
results from using logarithmic transforms (data from Patterson 1990:Table 1) . 262

26 Comparison of experimental core-reduction flake-size distributions among
experimenters: a) Behm's (1983) combined biface and core data; b) Behm's (1983)
Stage 2 through Stage 4 biface data; c) Tomka's (1989) core, biface, and dart point
data .................... .................. . 263

27 Plot of flake-size distributions for randomly fractured chert . . . . .... 267

28 Comparison of flake-size distributions for bifaces and unifaces . . . . . 270

29 Comparison of flake-size distributions for core-reduction assemblages ....... 270

30 Comparison of flake-size distributions for composite biface and core-reduction
assemblages: a) exponential; b) log-transformed . . . . . 271

31 Comparison of flake-size distributions for early-stage biface and uniface-production
assemblages with composite biface and core-reduction assemblages: a) early-stage
biface assemblage; b) uniface assemblage . ....... . . . . . 275

32 Comparison of flake-size distributions for simulated mixed biface and core-reduction
assemblages with composite experimental biface and core-reduction assemblages: a)
75:25; b)50:50; c) 25:75 ................ . . . . . 276

33 Comparison of flake-size distributions for simulated mixed biface and small-core-
reduction assemblages with composite experimental biface and core-reduction
assemblages: a) 75:25; b)50:50 .......... . . . . . . 277

34 Regression lines fitted to log-transformed experimental flake assemblage data: a)
bifaces; b) cores ......... ....... ............. ... 278

35 Distance-decay relationships for silicified coral: a) Archaic assemblages; b) post-
Archaic assemblages. Both data sets exhibit hyperbolic fall-off patterns . . 290


xvi









36 Plots of standardized residual values versus distance for silicified coral: a) Archaic
assemblages: b) post-Archaic assemblages . . ... .. 292

37 Distance-decay relationships for Hillsborough River Quarry Cluster cherts: a) Archaic
assemblages: b) Archaic assemblages separated by region: c) post-Archaic assem-
blages: d) post-Archaic assemblages separated by region .. . . . . .. 293

38 Plots of standardized residual values versus distance for Hillsborough River Quarry
Cluster cherts: a) Archaic assemblages: b) post-Archaic assemblages . .... 294

39 Distance-decay relationships for Withlacoochee Quarry Cluster cherts: a) Archaic
assemblages; b) Archaic assemblages with Hillsborough and Osceola County sites
removed; c) Archaic assemblages separated by region; d) post-Archaic assemblages. 296

40 Distance-decay relationships for Peace River Quarry Cluster cherts: a) Archaic
assemblages; b) Archaic assemblages from Peace River and Kissimmee regions only;
c) south Florida Archaic assemblages; sites with zero representation of Peace River
chert removed ................. .. ... ... ..... .... 298

41 Distance-decay relationships for Peace River Quarry Cluster cherts: a) post-Archaic
assemblages; b) post-Archaic assemblages separated by region . . . .. 299

42 Distance-decay relationships for Ocala and Lower Suwannee/Lake Panasofikee
Quarry Cluster cherts: a) Archaic assemblages; b) post-Archaic assemblages . 301

43 Hafted bifaces from Fort Center's mound-pond complex: a) Marion projectile point
with radial fracture and abraded margins recovered from the charnel pond; b) Marion
projectile point with heavily abraded edges, provenience unknown; c-d) Columbia
projectile points from Mound A . ................ . . 314

44 Bivariate plot of richness (R) and evenness (E) measures for Archaic and post-Archaic
hafted-biface assemblages ............. . . . ... 329

45 Relationship between number of raw-material types (richness) and sample size for
Archaic and post-Archaic hafted-biface assemblages. Sample size has been plotted
on a logarithmic scale to produce a more interpretable plot . . . . .... 329

46 Bivariate plot of richness (R) and evenness (E) measures for regional Archaic and
post-Archaic lithic assemblages . ........ . . ..... ... 332

47 Relationship between number of raw-material types (richness) and sample size for
regional Archaic and post-Archaic lithic assemblages. Sample size has been plotted
on a logarithmic scale to produce a more interpretable plot ........... . 332

48 Bivariate plot of richness (R) and evenness (E) measures for Archaic and post-Archaic
assemblages with non-chert materials included .. ... . . . . 335



xvii









49 Effect of transportation costs on total, average, and marginal returns employing a
production strategy without the use of thermal alteration. The optimal labor effort is
indicated by (a) where average and marginal returns are at their highest . . 344

50 Comparison of marginal and average return curves for competing biface-production
strategies. A switch in strategies should occur at the point at which thermal alteration
of lithic raw materials in close geographic proximity to the forager results in an
average return that is equal to or greater than the marginal return of a strategy
employing unaltered lithic materials obtained from a distant source . . ... 345

51 Comparison of marginal and average cost curves for competing biface-production
strategies ...................................... .. 345

52 Biface scavenging and reuse: a) reduction of proximal and distal fragments into
functional implements; b) differential patination on scavenged and reworked biface. 347

53 Archaeological examples of the reuse of scavenged bifaces: a-e) reworked distal
fragments; f-h) reworked proximal fragments. Proveniences: a-c, f-h) 8GL13; d)
8HG688; e) 8HG767 . .. ... ...... . . . 348

54 Examples of recycled bifaces: a) distal fragment retouched along broken margin and
used as a scraper; b) stem fragment retouched along broken margin and used as an
end scraper. Proveniences: a) Lake Livingston; b) State Road 70 . . . ... 349

55 Chert cobbles from Fort Center's Mound A ......... .......... . 360

56 Cores used in the production ofmicroliths at Yat Kitischee: a) unmodified rod-shaped
piece of chert; b-c) bipolar cores showing flake removals from opposing ends; d)
exhausted bipolar core with tapered, wedge-shaped striking platform ...... . 363

57 Relationship between core size and distance to chert resources for a) Archaic period
sites and b) with 8HG51 removed . . . . .... 375

58 Relationship between core weight and distance to chert resources for a) Archaic
period sites and b) with 8HG51 removed ..... .... . . . 377

59 Relationship between core size and distance to chert resources for post-Archaic period
sites . . . . . . . . .... .. . . .. . . . ... 378

60 Relationship between core weight and distance to chert resources for post-Archaic
period sites ....... ... ..... ....... . . . . . . 378

61 Relationship between core size and distance to chert resources for post-Archaic period
sites, one outlying data point removed . . .... ... . 380

62 Relationship between core weight and distance to chert resources for post-Archaic
period sites, one outlying data point removed . ... .......... 380


xviii









63 Relationship between core size and distance to chert resources for post-Archaic period
sites: a) core size minus hammerstones; b) core weight minus hammerstones. One
outlying data point has been removed . ... . . 381

64 Relationship between distance to chert resources and a) hammerstone size and b)
hammerstone weight ............... . . . 382

65 Relationship between Archaic hafted-biface size and distance to chert resources: a)
8GL13 and 8LL33, 36, 37 specimens not included; b) 8GL13 and 8LL33, 36, 37
specimens only .... ................... ........ .. 390

66 Relationship between selected post-Archaic hafted-biface size and distance to chert
resources ....................... . . . . . 392

67 Relationship between cortex-to-non-cortex ratios and distance to chert resources for
a) Archaic period sites and b) with 8H1393c/uw removed . . . . .. 401

68 Relationship between cortex-to-non-cortex ratios and distance to chert resources for
a) post-Archaic sites, b) outlying data points removed, and c) data points above the
regression line removed . . . ....... . . ...... .... 403

69 Examples ofdistal and lateral margin resharpening of hafted bifaces from sites in the
study area: a) lateral-margin resharpening, State Road 70); b) distal resharpening,
Lake Livingston . . . . . ... ... . . 406

70 Expectations of blade resharpening model: a) size index versus ratio of midblade-
width to basal-blade width; b) size index versus ratio of blade length to basal-blade
width ............ .... . ............. .......... 412

71 Bivariate relationship between tool size and resharpening for Levy hafted bifaces: a)
midblade-width to basal-blade width ratio; b) blade length to basal-blade-width ratio. 414

72 Bivariate relationship between tool size and resharpening for Marion hafted bifaces:
a) midblade-width to basal-blade-width ratio; b) blade length to basal-blade-width
ratio ............................... ........ ... 415

73 Bivariate relationship between tool size and resharpening for Newnan hafted bifaces:
a) midblade-width to basal-blade-width ratio; b) blade length to basal-blade-width
ratio .................. ... ........ ........ ...... 416

74 Bivariate relationship between tool size and resharpening for Putnam hafted bifaces:
a) midblade-width to basal-blade-width ratio; b) blade length to basal-blade-width
ratio . . . . . . . . . . . . . . . . . . . 417

75 Bivariate relationship between tool size and resharpening for Culbreath hafted
bifaces: a) midblade-width to basal-blade-width ratio: b) blade length to basal-blade-
width ratio . . ..... .. .... . . . ... 418


xix









76 Bivariate relationship between tool size and resharpening for Hillsborough and
Sumter hafted bifaces: a) midblade-width to basal-blade-width ratio; b) blade length
to basal-blade-width ratio .. .. ......... . . 419

77 Bivariate relationship between tool size and resharpening for Bolen and Greenbriar
hafted bifaces: a) midblade-width to basal-blade-width ratio; b) blade length to basal-
blade-width ratio .... .. . . . . ................ 420

78 Bivariate relationship between tool size and resharpening for Florida Archaic
Stemmed hafted bifaces: a) midblade-width to basal-blade-width ratio; b) blade length
to basal-blade-width ratio . ... .. ............. . ... 422

79 Bivariate relationship between tool size and resharpening for Bradford, Columbia, and
Taylor hafted bifaces: a) midblade-width to basal-blade-width ratio; b) blade length
to basal-blade-width ratio..... . . . . .... . . 423

80 Bivariate relationship between tool size and resharpening for Columbia and Taylor
hafted bifaces using combined group means: a) midblade-width to basal-blade-width
ratio; b) blade length to basal-blade-width ratio . . . . . . ....... 424

81 Bivariate relationship between tool size and resharpening for Duval, Sarasota,
Jackson, and Broward hafted bifaces: a) midblade-width to basal-blade-width ratio;
b) blade length to basal-blade-width ratio .... ................ 425

82 Bivariate relationship between tool size and resharpening for Hernando hafted
bifaces: a) midblade-width to basal-blade-width ratio; b) blade length to basal-blade-
width ratio ...................... . . . . . . 426

83 Bivariate relationship between tool size and midblade-width to basal-blade-width ratio
for Pinellas hafted bifaces ....... . ..... 427

84 Bivariate plot of mean number of utilized tool edges versus mean number of
functional units for Archaic and post-Archaic assemblages . . . . . .. 438

85 Examples of cores: a-d) single-platform, pyramidal cores; e-f) amorphous cores; g-h)
bipolar cores; i) bifacial core. Proveniences: a) 8HR48; b) 8HG679; c-d) 8HG49; e)
8HG18; f) 8P11753; g) 8GL13; i) 8HR68 .................. ..... 446

86 Variation in lithic artifact density by cultural-temporal component, Kissimmee region. 451

87 Relative change in the mixture of shell and stone in the technological inventory as the
cost of stone increases. At Pineland, where stone is expensive to acquire and shell is
inexpensive, the relative abundance of shell is high. At Yat Kitischee, where the cost
of acquiring stone is much lower, the relative mix of the two resources is more even.
The shallow indifference curve reflects the substitutability of the two resources. The
lineXindicates the change in the resource mix that occurs as a result of the increase
in the cost of stone ......... . ... . . .. .... . .. 454


xx









88 Comparison of the relative change in the mixture of shell and stone in the
technological inventories of Archaic (Groves' Orange Midden) and post-Archaic
(Hontoon Island) components in the St. Johns region of central Florida. A reduction
in demand for stone is indicated during the post-Archaic period with a corresponding
increase in the use of shell. The shallow indifference curve reflects the substitutability
of the two resources. The line X indicates the change in the resource mix between
culture periods ................................... .. 457

89 Relative change in the mixture of sharks' teeth and stone in the technological
inventories at Pineland and Fort Center's mound-pond complex as the cost of stone
incases. The stronger arc ofthe indifference curve indicates that the two resources
were not as substitutable for one another as were shell and stone. This is further
indicated by the small difference in the resource mix (X) that occurs as a result of the
increase in the cost of stone ............. .. . . . . . ...... 459

90 Bivariate plot of richness and evenness values for Archaic and post-Archaic tool
assemblages ................... . . . . . . 467

91 Relationship between number of tool classes (richness) and sample size for Archaic
and post-Archaic assemblages. Sample size has been plotted on a logarithmic scale
to produce a more interpretable plot ................... . . 468

92 Bivariate plot of discriminant-function scores comparing Archaic debitage
assemblages with experimentally pro-duced debitage assemblages ........ . 486

93 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories
from 8HG51 with experimentally produced debitage assemblages . . . 488

94 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories
from 8HR68 with experimentally produced debitage assemblages . . . 488

95 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories
from 8HR71 with experimentally produced debitage assemblages .. . .... 489

96 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories
from 8HR92 with experimentally produced debitage assemblages . . . ... 489

97 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories
from 8HG35 with experimentally produced debitage assemblages . . . 490

98 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories in
the Archaic assemblage from 8HG678 with experimentally produced debitage
assemblages . .... . . . . . . 490

99 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories in
the Archaic assemblage from 8HG18 with experimentally produced debitage
assem blages . . . . . . ... . . . . . . . . . 492


xxi









100 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories in
the Archaic assemblage from 8HG20 with experimentally produced debitage
assemblages .... . .... .. .. ..... . . . ....... 492

101 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories
from 8HG34 with experimentally produced debitage assemblages . . . ... 493

102 Comparison ofthe cumulative proportions of the combined Sullivan-Rozen debitage
categories in the Archaic assemblages from 8HG18, 8HG20. 8HG27, and 8HG34
with experimentally produced debitage assemblages . . . . . ...... 493

103 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories
from 8HG767 with experimentally produced debitage assemblages ....... . 494

104 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories for
selected raw materials from 8HG51 with experimentally produced debitage assem-
blages: a) Hillsborough River Quarry Cluster; b) Withlacoochee Quarry Cluster; c)
silicified coral; d) Peace River Quany Cluster . . . . . . 499

105 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories for
Hillsborough River Quarry Cluster chert with experimentally produced debitage
assemblages: a) 8HR68; b) 8HR71; c) 8HR92 . . . . . . ....... 501

106 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories for
Cow House Creek chert with experimentally produced debitage assemblages: a)
8HR68; b) 8HR92 ............ . . . . . . .... 502

107 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories for
Bay Bottom chert from 8HR92 . ... ....... . . . . . 503

108 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories for
Withlacoochee Quarry Cluster chert with experimentally produced debitage
assemblages: a) 8HR71; b) 8HR92 . . . . . . . . ... . ...... 504

109 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories for
silicified coral with experimentally produced debitage assemblages: a) 8HR71; b)
8HR92 ............. ................ .......... .. 505

110 Bivariate plot of discriminant-function scores comparing post-Archaic debitage
assemblages with experimentally produced debitage assemblages . . .... 506

111 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories in
the post-Archaic assemblage from 8HG18 with experimentally produced debitage
assemblages ................. . . . . . . . .. 507

112 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories in
the post-Archaic assemblage from 8HG20 with experimentally produced debitage
assemblages . . ... . ..... . 507

xxii









113 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories in
the post-Archaic assemblage from 8HG27 with experimentally produced debitage
assemblages . . ... .............. . . ... 508

114 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories
from 8HR48 with experimentally produced debitage assemblages . . . .. 508

115 Comparison of the cumulative proportions of the combined Sullivan-Rozen debitage
categories in the post-Archaic assemblages from 8HG18, 8HG20, 8HG27, and
8HG34 with experimentally produced debitage assemblages . . ..... ..... 509

116 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories
from 8GL13 with experimentally produced debitage assemblages ........ 511

117 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories
from 8P 1753 with experimentally produced debitage assemblages . . ... 511

118 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories
from 8LL33, 36, 37 with experimentally produced debitage assemblages . . 512

119 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories in
the post-Archaic assemblage from 8HG678 with experimentally produced debitage
assemblages ................. ....... ............... 513

120 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories in
the post-Archaic assemblage from 8HR44 with experimentally produced debitage
assemblages .................. .................. .. 513

121 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories for
selected raw materials from 8P11753 with experimentally produced debitage
assemblages: a) Turtlecrawl Point Quarry Cluster, Type 1; b) Turtlecrawl Point
Quarry Cluster, Type 2; c) Hillsborough River Quarry Cluster Type 4; d) other
Hillsborough Quarry Cluster cherts .. ..... . . . . . 518

122 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories for
Peace River Quarry Cluster cherts with experimentally produced debitage
assem blages . . . . . . . . . . . . .. 521

123 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories for
Hillsborough River Quarry Cluster cherts with experimentally produced debitage
assemblages: a) 8HG678; b) 8HR48 . . . ... ...... ........ 522

124 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories for
selected raw materials from 8GL13 with experimentally produced debitage
assemblages: a) chert cobbles; b) Peace River Quarry Cluster; c) silicified coral; d)
combined dolomite-quartz-fossil bone ..... .... . ... 524



xxiii









125 Comparison of the cumulative proportions of Sullivan-Rozen debitage categories for
Hillsborough Quarry Cluster chert from 8GLI3 with experimentally produced
debitage assemblages . . . . .......... . . . 526

126 Comparison ofexperimental tool- and core-reduction flake-size distributions with the
flake-size distribution from the Archaic assemblage at 8HG20 .. . . .. 530

127 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distribution from 8HG35 ................ .. . 530

128 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distribution from 8HG51 ... . .. ...... ..... .. . . 531

129 Comparison ofexperimental tool- and core-reduction flake-size distributions with the
flake-size distribution from 8HR68 . . ............ ..... 531

130 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distribution from 8HR71 .... ........... . . 532

131 Comparison ofexperimental tool- and core-reduction flake-size distributions with the
flake-size distribution from 8HR92 . . ............. ..... 532

132 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distribution from the Archaic assemblage at 8HG34 . . . . .. 533

133 Comparison ofexperimental tool- and core-reduction flake-size distributions with the
flake-size distribution from 8HG767 ................... ...... 533

134 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distribution from the Archaic assemblage at 8HG18 .. . . . 535

135 Comparison ofexperimental tool- and core-reduction flake-size distributions with the
flake-size distribution from the composite Archaic assemblage from 8HG18, 8HG20,
8HG27, and 8HG34 . . . . ............. .. 535

136 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distribution from the Archaic assemblage at 8HG678 . . . ... 537

137 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distributions of Hillsborough River Quarry Cluster cherts: a) 8HG51: b)
8HR68; c) 8HR71; d) 8HR92 ............ . . . . . .... 542

138 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distributions of Cow House Creek chert: a) 8HR68; b) 8HR92 . . 543

139 Comparison ofexperimental tool- and core-reduction flake-size distributions with the
flake-size distributions of Withlacoochee Quarry Cluster cherts: a) 8HG51; b)
8HR71; c) 8HR92 .. . .... .... ....... .... 544

xxiv








140 Comparison ofexperimental tool- and core-reduction flake-size distributions with the
flake-size distributions of Bay Bottom chert from 8HR92 ... . . 545

141 Comparison ofexpeimental tool- and core-reduction flake-size distributions with the
flake-size distributions of Peace River Quarry Cluster chert from 8HG51 ..... 545

142 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distributions of silicified coral: a) 8HG5 I; b) 8HR92 . . . . . 546

143 Comparson ofexperimental tool- and core-reduction flake-size distributions with the
flake-size distribution from 8LL33, 36,37. .. ..... . ..... 549

144 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distribution from 8PI1753 ................... ...... .. 549

145 Comparison of eerimental tool- and core-reduction flake-size distributions with the
flake-size distribution from the post-Archaic assemblage at 8HG18 . . ... 551

146 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distribution from the post-Archaic assemblage at 8HG20 . . . 551

147 Comparison of expermental tool- and core-reduction flake-size distributions with the
flake-size distribution from the post-Archaic assemblage at 8HG27 . . . .. 552

148 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distribution from 8HR44 ...... ......... . . 552

149 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distribution from 8HR48 ........... . . . . . 553

150 Comparison ofexperimental tool- and core-reduction flake-size distributions with the
flake-size distribution from the composite post-Archaic assemblage from 8HG18,
8HG20, 8HG27, and 8HG34 . . . .. .. .. .. 553

151 Comparison of expeimental tool- and core-reduction flake-size distributions with the
flake-size distribution from the post-Archaic assemblage at 8HG678 ...... . 554

152 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distributions of selected raw materials from 8P11753: a) Turtlecrawl Point,
Type I; b) Turtlecrawl Point, Type 2; c) Hillsborough River Quarry Cluster, Type 4:
d) all other Hillsborough River Quarry Cluster cherts combined . . . .... 557

153 Comparison of experimental tool- and core-reduction flake-size distributions with the
flake-size distributions of Hillsborough River Quary Cluster cherts from selected
post-Archaic sites: a) 8HG678; b) 8HR48. ................ ... .. 560

154 Comparison ofexperimental tool- and core-reduction flake-size distributions with the
flake-size distributions of Peace River Quarry Cluster cherts from selected post-
Archaic sites: a) 8HG678; b) 8HR48 ........... .... . .. 561

xxv















KEY TO ABBREVIATIONS


Arch = Archaic NM = Not Measured
B = Brooksville O = Ocala
BB= Bay Bottom P= Pinellas
BBW = Basal-blade Width PArch Post-Archaic
BL = Blade Length PR = Peace River
C = Caladesi SC = Silicified Coral
CHC = Cow House Creek SD = Standard Deviation
COB = Cobbles TCP = Turtlecrawl Point
CV = Coefficient of Variation TI = Type 1
DO = Dolomite T2= Type2
FAS = Florida Archaic Stemmed T3 = Type 3
FB = Fossil Bone T4 = Type 4
GL = Glass WR = Withlacoochee River
H or HR = Hillsborough River QC = Quarry Cluster
KR = Kissimmee Region QZ = Quartz
L or LS/LP = Lower Suwannee/Lake Pana- UID = Unidentified
soffkee UIDCH = Unidentified Chert
MBW = Mid-blade Width W = Width
N = Number






















xxvi















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 ECONOMICS OF LITHIC-RESOURCE USE
IN SOUTH-CENTRAL FLORIDA

By

Robert James Austin

December 1997


Chairman: Jerald T. Milanich
Major Department: Anthropology


This is a study of the ways in which hunter-gatherers adapt to conditions of limited resources;

in this case, the resource of concern is lithic raw materials. The research was designed to examine how

prehistoric hunter-gatherers in south-central Florida coped with the costs and risks associated with lithic

resource acquisition. It also addresses the effect of limited access to raw materials on organizational

strategies, and more specifically, the use ofcurated versus expedient technologies among residentially

mobile foragers and sedentary collectors.

Economic concepts (supply and demand, marginal value, optimization) are used as the

theoretical basis for the development of testable hypotheses related to these questions. Lithic artifacts

from Archaic and post-Archaic contexts within the Kissimmee Region of south-central Florida form the

primary data set for this research. Comparative data were obtained from the analysis of contemporane-

ous assemblages in the Peace River drainage, southwest Florida coast, and the Tampa Bay region of

west-central Florida. Analysis focuses on assemblage structure and composition rather than on


xxvii










functional or technological analyses of individual artifacts. Analysis of waste-flake assemblages derived

from the replication of a variety of tool and core forms is included in the study to provide a basis for

making valid inferences regarding the technological origins of archaeological debitage assemblages.

The results indicate that the organizational strategies employed by Archaic and post-Archaic

populations within the study area were influenced to a large degree by the availability and abundance

of lithic raw materials regardless of differences in settlement mobility. That is, when compared with

assemblage data from sites located in chert-rich areas, Archaic and post-Archaic groups inhabiting the

study area practiced cost-reduction strategies to a greater degree than did their contemporaries farther

north. However, intraregional variation existed in the specific cost- and risk-reduction strategies that

were employed These strategies were influenced by several interdependent factors, including settlement

mobility and procurement strategies, as well as access to alternative raw materials.





























xxviii
















CHAPTER I
INTRODUCTION


This dissertation is a study of the ways in which hunter-gatherers adapt to conditions of limited

resources; in this case, the resource of concern is lithic raw materials. More specifically, my goal is to

examine how prehistoric hunter-gatherers in south-central Florida organized their technological systems

to cope with the problem of raw-material scarcity.

Lithic studies in Florida have traditionally focused on questions of chronology, technology, site

function, and intrasite spatial patterning (e.g., Austin 1983; Austin and Ste. Claire 1982; Ballo 1985;

Bullen 1975; Chance 1981, 1983a; Chance and Misner 1984; Daniel and Wisenbaker 1987; Estabrook

and Newman 1984; Hemmings and Kohler 1974; Purdy 1981a, 1981b; Purdy and Beach 1980; Ste.

Claire 1996). Although these studies have provided valuable information, they have been conducted

for the most part as if technology functioned separately from the rest of the cultural system. As a

consequence, they have contributed little to an understanding of the dynamics and organizational

variability in hunter-gather adaptations that are observable archaeologically in Florida (Prentiss and

Romanski 1986). In this study, I attempt to rectify this situation by focusing explicitly on technological

organization as an adaptive response to environmental structure.

As defined by Nelson (1991:57), technological organization refers to a set of "strategies for

making, using, transporting, and discarding tools and the materials needed for their manufacture and

maintenance." Therefore, when I speak of technological organization I am referring to the strategies that

are developed by human beings in response to the necessities of everyday life. The goal of this research

is to understand the factors that influenced the decisions to choose particular sets of organizational

strategies.










2

Several factors have been suggested as potentially influencing the organizational character of

prehistoric lithic industries. Mobility strategies, time stress, functional efficiency, and access to raw

materials are but a few of the more commonly cited. Of these, mobility strategies and access to raw-

materials have engendered the greatest interest and debate (e.g., Andrefsky 1994a, 1994b; Bamforth

1986; Binford 1979, 1980; Binford and Stone 1985; Daniel 1996; Gould 1980; Odell 1996; Parry and

Kelly 1987; Thacker 1996). South Florida's interior is one of the few areas in the state where

prehistoric native peoples did not have direct access to durable raw materials to manufacture tools,

potentially limiting their ability to exploit biotic resources. The archaeological evidence, limited as it

is, suggests that the region was inhabited by highly mobile hunter-gatherers during the preceramic

Archaic period and by more sedentary populations during the post-Archaic. Thus, I believed the region

had the potential to serve as a test case for evaluating the role that access to raw materials had on the

organizational strategies of both mobile and sedentary hunter-gatherers.

The specific research questions that are addressed by this study can be stated simply:

1) What were the means (procurement strategies) by which siliceous lithic materials were

obtained from their source locations and moved into south-central Florida?

2) What were the various strategies adopted by prehistoric native peoples to cope with the costs

and risks associated with procuring these resources from distant locations, and did these change through

time?

To address these questions, I have structured my analyses to move from a lower-order

perspective, in other words, that which is concerned with what is observed archaeologically in the

material record, to a higher-order perspective that examines the relationship between environmental

structure and organizational strategies. The lower-order perspective of inferring specific behaviors

based on observation of the archaeological record is strengthened through the use of experimentally

derived data on lithic waste-flake assemblages. Experiments were conducted that were designed to










3

replicate debitage assemblages resulting from different lithic-reduction strategies: patterned-tool (i.e.,

bifaces, unifaces) and core reduction (large and small). Analysis of the experimental assemblages

focused on refining existing techniques for use with archaeological assemblages. The techniques

include the Sullivan-Rozen flake typology (Sullivan and Rozen 1985; Rozen 1984) and the flake-size

distribution analysis pioneered by Patterson (1981, 1982, 1990; Patterson and Sollberger 1978).

Analyses of experimentally derived waste-flake assemblages indicates that the two techniques can be

used successfully to distinguish between patterned-tool and core-reduction strategies. The use of these

techniques is refined through the application of statistical procedures for distinguishing between the

two major reduction strategies. Discriminant-function analysis is used with the Sullivan-Rozen

typology while regression analysis and statistical comparison of regression slopes are used to distinguish

between flake-size distributions resulting from different reduction strategies.

Moving up the hierarchy of inference, specific hypotheses and test implications are developed

regarding the types of problem-solving strategies that prehistoric hunter-gatherers may have practiced

in the study area given the condition of limited access to lithic resources. These hypotheses are

grounded in microeconomic theory which views scarcity as a central element of economic decision-

making (Glahe and Lee 1989:3-5). Economics is the study of how people choose to allocate limited

resources that may have alternative uses to achieve specific goals or acquire desired goods or services.

Microeconomics focuses on individual decision makers and how they attempt to solve this problem most

efficiently. A basic assumption of microeconomics, and of this study, is that the principles of economic

decision-making as defined above, are applicable to all forms of human behavior regardless of cultural

context. Every decision that people make involves a cost -- in time, in labor, in goods or services.

Moreover, a decision to allocate resources to achieve one goal necessarily requires that resources be

diverted from achieving some other desirable goal. Thus, every decision has a direct cost and an

opportunity cost, and both must be considered before an intelligent economic decision can be made.









4

Much of the data used in this research were derived from sites in the Kissimmee region of

south-central Florida (Figure 1) with comparative data coming from sites in the Peace River valley, the

southwest Florida coast, and the Tampa Bay region. A total of 5256 artifacts from 39 sites was

analyzed. Because much of the study area has experienced little in the way of systematic archaeological

excavation, the artifacts analyzed for this research came from a variety of collections. Some were

derived from test excavations conducted specifically for this study. Many came from test excavations

and surveys conducted as part of cultural resource management (CRM) projects. Still other artifacts

came from museum collections or private collections. The temporal range of the sites spans the early

Archaic period (ca. 9500 B.P.) through the early historic period (ca. 200 B.P.), with most sites, or

components of sites, dating to the middle Archaic through late prehistoric periods (ca. 6-7000 B.P.

through 400 B.P.).

The basic limitation of this research is the small sizes of the lithic artifact assemblages at sites

in the region, particularly those dating to the post-Archaic period. This is partly due to the geological

character of the region; that is, the absence of exploitable chert outcrops. The problem is made even

more acute, however, by the limited archaeological work that has been conducted in the region,

consisting primarily of surveys and test excavations. In anticipation of potential criticism I offer the

following quote by Steve Hale regarding the problem of inadequate samples:


The archaeological record, and our samples, are always incomplete. Rather than avoid

analyzing samples that are less than ideal, I prefer to minimize the biases, and make

use of imperfect samples that are, in some cases, perhaps the only samples we will

ever have [Hale 1995:284].


Thus, rather than ignore or discount the value of the lithic data that do exist, I have chosen to forge

ahead in the firm belief that analysis of even small and incomplete samples can offer insights into

prehistoric human behavior if it is structured within a theoretical context.










5

















Penacola Tallahassee Jacksonville









Kissimmee

Tampa

PROJECT







i.
Sebring t AREA



-\Ft. Myers



Miami




















Figure 1. Map of the study area in relation to Florida.










6

This dissertation is organized as follows. In Chapter 2 I discuss in depth the theoretical

framework that guides this study. This framework is explicitly materialist and evolutionary. Economic

and ecological concepts are introduced and are followed by a discussion of their application to the study

of lithic economies. Extended treatments of the environment and geology of the study area are presented

in Chapters 3 and 4 to provide an ecological context for the study. A basic assumption that underlies

the temporal analysis is that prehistoric people during the preceramic Archaic period practiced a more

mobile settlement strategy than did later, post-Archaic groups. This is based, in part, on archaeological

evidence, but the nature of the paleoenvironment as it has been reconstructed from palynological,

sedimentological, geological, and faunal data also is crucial to this assumption. A separate chapter (4)

on geology and chert resources provides the reader with a background on regional geology which is

necessary to understand the formation and geographic distribution of chert in Florida.

Chapter 5 provides an overview of the prehistory of the study area. This lays the foundation

for much of the assumptions regarding differential mobility strategies and restricted territories that

underlie the economic analysis that follows. Chapter 6 outlines a series of hypotheses that have been

developed to address the questions posed at the beginning of this chapter. Test implications derived

from these hypotheses are specified as are the data necessary to confirm or disconfirm the hypotheses.

Chapter 7 describes the methods used in the data collection and analysis of archaeological collections,

as well as the methods employed during the experimental replications. The chapter also includes an

extended discussion on the method used to identify chert provenience which was a critical aspect of this

research.

Research results are presented in Chapters 9-12. Chapter 9 focuses on strategies for coping

with procurement risk in a chert-poor area. Procurement strategies (direct versus indirect) are identified

as are auxiliary risk-abatement strategies such as resource diversification, scavenging, and storage.

Chapter 10 addresses the costs associated with procurement and how these were minimized. Specific










7

labor-reduction strategies that are addressed include minimization of transport costs, minimization of

production costs, economizing, and reduction of demand. In Chapter 11 examine the use of curated

versus expedient technologies in the study area. The results of this analysis lead directly to an

examination of time-minimizing and resource-maximizing strategies.

The final chapter, Chapter 12, summarzes the results and conclusions presented in the previous

three chapters. I also examine the implications of the research results for understanding the evolution

of social complexity in south Florida, and more specifically the role that exchange systems may have

played in the emergence of social inequalities.

I do not believe that the conclusions presented in this dissertation necessarily represent the

"truth" about the segment of prehistoric behavior I have chosen to study. I accept the possibility, even

the inevitability, that my conclusions will be modified or discarded entirely with future research. My

hope is that this research, however imperfect, will contribute in its way to the process of leaning about

the past and will stimulate others to continue in this quest.
















CHAPTER 2
THEORETICAL FRAMEWORK


The theoretical perspective that guides this study is ecological, materialist, and evolutionary.

It takes as given the fact that human beings are biological organisms who exist as part of an ecological

system (Odum 1971:8). Like other biological organisms, humans must adapt to the constraints and

opportunities afforded them by their external environment. Within this framework, the hunting and

gathering lifestyle is viewed as an adaptive response to variation in environmental structure. This

structure is composed of a set of interrelated geological, biological, and climatological features with

structural variation occurring at varying scales of space and time. Human adaptive responses to

structural variation may include differential residential mobility, seasonal scheduling of subsistence

activities, the establishment of intra and interregional alliances, the development of exchange systems,

complexification of social structure, or the elaboration of ritual.

Yet another way in which human beings respond to variation in the environment is through

technology. This study adopts the materialist paradigm that views technology as part of the economic

infrastructure on which the structural components of society and ideology rest (Harris 1979:55-56). At

the same time, it acknowledges that the relationship between structure, infrastructure, and environment

is complex and multidimensional. No single research strategy, and certainly no single study, can hope

to explore such complexity in its entirety. Instead, I have chosen to focus on a single aspect of this

complex relationship, access to raw materials, and examine how this influenced the organizational

structure of one element of prehistoric hunter-gatherer economics, technology.




8










9

This study also attempts to understand the selective factors that shaped the organizational

structure of technology in a particular geographic region through time. Thus. it will be an evolutionary

study in the sense that evolution is viewed as a process by which changes in the frequency distribution

of behavioral traits can occur over time as a result of selection acting on existing behavioral variation.

The stimulus for selection is often change in an organism's external environment, which includes both

the physical and social environments. The study concentrates on the selection process rather than

causal processes as the two concepts are analytically distinct. Following Hannon and Freeman (1989),

causal analysis focuses on whether one phenomenon or set of phenomena affects the likelihood that a

second phenomenon or set of phenomena will occur. Selection analysis focuses on whether the rise to

prevalence ofa specific phenomenon, for example, a certain organizational configuration, is governed

by a specified set of changes in the social or natural environment. The question in selection analysis

is whether environmental conditions control the distribution of a phenomenon at different points in time

(Hannon and Freeman 1989:16). The cause or origin of the phenomenon is not at issue, and is, in fact,

the result of a separate process, that which results in the variation on which selection acts. The rise or

prevalence of organizational configurations (i.e., adaptations) is the result of the selection process and,

therefore, it is the selection process that governs the dynamics of organizational change (cf Harms

1979:59-62).

To understand why a certain form of technological organization was selected for requires going

beyond a mere demonstration of a relationship between environmental conditions and organizational

configuration. It requires an understanding of the economic logic that underlies selection. Microeco-

nomic models of human decision-making provide a methodology for studying the process by which

humans make choices in the economic sphere. More importantly, it provides a means of evaluating the

adaptive value of these choices, for if these choices are adaptive given a unified set of environmental

(natural and social) conditions, then they should be selected for.









10

In anthropology, the principles and methods of microeconomics have most often been applied

to studies of subsistence economies (e.g., Bettinger and Baumhoff 1982, 1983; Earle and Christianson

1980; Gremillion 1996; Keegan 1986, Keegan and Butler 1987), and have contributed to the

development of optimal-foraging theory and its application to the study of living hunter-gatherers (e.g.,

Hawkes 1993; Hawkes et al. 1985; Hill 1988; Hill et al. 1985, 1987; Hurtado et al. 1985; O'Connell

and Hawkes 1981: Winterhalder 1986; see Smith and Winterhalder 1992a and Winterhalder and Smith

1981 for other applications). While efforts to extend the principles of microeconomics to prehistoric

lithic studies are a relatively recent phenomenon (e.g., Bamforth 1986; Bouseman 1993; Boydston

1989; Jeske 1989; Kuhn 1994; Morrow 1996; Torrence 1983, 1989a), such an application seems to

be a potentially fruitful area of research. Stone was the principal durable raw material in many

prehistoric technologies in Florida, and stone tools were used directly in the procurement of many

subsistence resources, as well as to manufacture other tools made of wood, bone, and antler. The

procurement and use of lithic materials were, therefore, important components of the economic

adaptations of prehistoric hunter-gatherers.


Pattern. Scale. and Culture Change


Before examining the various elements of this study's theoretical framework in more detail, it

is worth discussing the concept of scale, how it affects observations of both patterning and variability

in the archaeological record, and the implications of scale for understanding culture change. The

concept of scale is perhaps one of the most fundamental issues in anthropology, as it is in science

generally (cf. Butzer 1982; Dethlefsen 1992; Gleick 1987; Keegan 1991; Levin 1992; Levins 1966;

Marquardt 1992; Schaafsma 1991). As described by Marquardt (1992:107), the concept of scale refers

simply to "the amount of space and time under consideration. Out of an infinite multiplicity of scales,

individuals comprehend patterns, recognize homogeneity, plan for the future and operate in the present










I 1

at different scales." Thus, the observation and perception of patterning is dependent on scale, with

different patterns emerging as the scales of reference change. The scale at which pattern is observed

is the effective scale (Crumley 1979:166).

To this concept of scale, Marquardt has added the concepts of "structure" and "agency"

(Marquardt 1992:104-105). Structures may be physical, such as topography, climate, and natural

resources, or sociohistorical, which refers to social, political, and economic institutions and the social

relations associated with their functioning. Agncy refers to the purposeful behavior of human actors.

Marquardt borows the notion ofontradiction from Marxist theory to refer to the dynamic tension that

comes about when "human activities take place in constantly changing sociohistorical contexts"

(Marquardt 1992:105). These concepts are similar to the concepts of structure, event, and conjuncture

used by Braudel in his model of historical process (1972 cited in Cobb 1991:170-171), and are merely

different ways of conceptualizing scale. In fact, Cobb (1991:174) observes that Braudel's cyclical model

of slow-moving, recurrent "structures" and short, rapid oscillations, or "events," recognizes the

"overlapping dimensions of variability" (i.e., scales) that characterize the historical process. Braudel's

"conjuncture" is the dynamic intersection of structure and event resulting in "new points of equilibrium

within the longer lasting structural cycle" (Cobb 1991:171), and is analogous to the process of resolving

contradictions through dialectical critique as advocated by Marquardt (1992:109-110).


Individual Variation and Cultural Pattern

Much of the recent post-processual critique of archaeology revolves around the apparent failure

of processual archaeologists to derive general laws of culture change (cf. Hodder 1985; Patterson 1990).

This is attributed, in part, to the lack of emphasis in the processualist research strategy on the role of

the individual as an agent of change (Patterson 1990:194-195). This criticism is simply the most recent

manifestation of the long-standing debate between those who view culture as the sum of things done

or thought by individuals who are active participants in its formation and transformation, the so-called










12

"normative" view, and those who view culture as a superorganic entity that exists apart from

individuals, operating and evolving according to its own set of rules, deterministically shaping and

molding individual behavior.

The debate centers around what Paul (1987:80) has referred to as "autonomy of agency." In

other words, if culture is an entity in and of itself, and this entity influences the behavior and thoughts

of its members, then culture constitutes the autonomous agent that individuals respond to. On the other

hand, if culture emerges from and is maintained by the collective actions of its members, then

individuals are the autonomous agents and culture is a consequence of their actions. The history of

anthropological theory can be viewed as a recurring struggle between these two points of view for

disciplinary dominance -- the primacy of the individual and the particularistic versus the primacy of

culture as a system and the search for general laws.

Within this context of theoretical debate, the recognition that nature, including human nature,

consists of a multiplicity of scales provides insight in terms of understanding and explaining cultural

behavior. The goal of science is to abstract, from the infinite range of variability that exists in nature,

generalities that enable us to make predictions about cause-and-effect relationships and, ultimately, to

understand and explain how and why these relationships occur. However, the degree to which pattern

and variability are observed, perceived, or analyzed is conditioned by the scales at which we choose to

make our observations. Moreover, the mechanisms that result in observed patterns may operate at

scales that are different from those at which the patterns are observed (Levin 1992:1943). In a recent

article, Edwin Dethlefsen (1992) addressed this analytical contradiction and the implications it has for

archaeology. According to Dethlefsen, the difficulty that we encounter in trying to describe or

characterize cultural phenomena in abstract general terms is due to the inherent variability of their

component parts. "Scholars who wish to transcend specificity quickly find themselves lost in a sea of

exceptions" (Dethlefsen 1992:149). Dethlefsen draws on chaos theory and mathematical topology in










13

an attempt to conceptualize complex patterning in culture. The entire range of possible relationships

between variables in a system can be conceived of topographically as the sum total of their intersections

- the entire phase sce of the system (Gleick 1987:47). Any single point in this phase space represents

a state in the system frozen in time. All the information about that state is contained in the intersection

of all variables at that point in phase space. As the system changes, that point moves to a new position

and as it continues to move it traces a line or trajectory. This trajectory of random, multidimensional

points through time become attracted to a bounded region of phase space, the strange attracor. The

lines never intersect one another or travel the same path, yet they form a distinctive pattern of

continuous looping that is predictable. The computer-generated images resulting from chaos-inspired

research (the fractal geometry of Mandelbrot sets) enable the multivariate relationships of a dynamic

system to be visualized in a way that does justice to its complexity (Gleick 1987:83-118).

Battleship-shaped curves are an example of a simple way to visualize dynamics in archaeology.

However, this graphic method considers only a single variable usually style -- measured against a time

scale. Multivariate statistics are another way of comparing large numbers of variables. But as

Dethlefsen points out, the ability of computers to manipulate large amounts of data and consider the

multiple relationships between them all does not compensate for the difficulty human beings have in

representing, thinking, talking, or writing about multidimensional reality. Hence, most of our models

of change tend to be limited to the study of one or at best a few variables at a time. While Dethlefsen

feels this obfuscates the true dynamic relationships between the large numbers of variables that actually

constitute the reality of any population, others (e.g., Levins 1966; Keegan 1991; Winterhalder and

Smith 1992) accept that the very complexity of real-world phenomena makes it impossible to construct

models that encompass both specificity and generality. Even the Mandelbrot sets of chaos theorists

distort and misrepresent the process of dynamic change by reducing it to a static state.










14

Throughout much of his paper. Dethlefsen is clearly grappling with the contradictions of

specificity versus generalization, of variation versus pattern.


Words force us to set aside the notion that every observation lies somewhere in a

distribution curve which in turn exists in a universe of intersecting, mutually

influential distribution curves. Everything has to do, more or less, with everything

else, but one can neither talk about or illustrate everything at once. Still more difficult

to visualize is the concept that every curve of distribution, even though it is itself a

generalization, represents some particulate statement about some other, greater curve

within which it is merely a point or a transect. Every degree of magnification applied

to a concept enhances its systemic ambiguity; thus, if we allow our attention to focus

altogether on its fine points, we risk losing track of the line that, when seen at a wider

angle, those points constitute [Dethlefsen 1992:152].


In other words, if we focus too closely on the particular, we risk missing the forest for the trees. His

reference to the Marcel Duchamp painting Nude Descending a Staircase nicely illustrates his point

(Dethlefsen 1992:153). If one focuses on trying to isolate any single image within this graphic

representation of motion versus time, then one loses site of the real phenomenon that is occurring --

change. For Dethlefsen it is the generality that offers the greatest potential for understanding

(Dethlefsen 1992:149), yet the generalization cannot simply be considered as a larger, more complex

representation of individual entities; it is an abstraction of a unique multidimensional reality that

consists of dynamic relationships between innumerable entities, traits, and phenomena.


Adopting a Scalar Perspective

Adopting a scalar perspective results in several important realizations that affect the ways in

which we conduct archaeological research. First, to study and understand culture and the factors that










15

result in culture change, we must accept the universal coexistence of pattern and variation. Therefore,

culture, like nature, must be viewed as a combination of unique, individual events and composite

patterns. A second important insight is that variability is an inherent condition of the natural world.

Indeed, from an evolutionary point of view, variation is essential for survival. Consequently, in

examining the material residues of human cultural behavior we should not expect strict uniformity, but

instead should expect to encounter variation in both the material record and in the behaviors that they

reflect. From this it follows that there is no single, correct scale at which to conduct archaeological

analysis. Different scales of analysis provide different kinds of answers, and connecting the answers

that explain patterns at one scale requires models that may, of necessity, mask or inhibit the study of

variability inherent at finer scales. Thus, many different types of analysis conducted at variable scales

of space and time must be pursued to approach complete understanding.

Despite the inherent difficulty of incorporating specificity and generality into a single

explanatory model of culture change, I believe that it is possible to integrate a number of different

models (or, if you will, scalar perspectives) within a single, unified theoretical framework. This

framework that I refer to is neo-Darwinian evolution because only a selectionist theory of culture can

explain how variability leads to pattern, and so provides a means for integrating the various scales of

patterning and change that we observe empirically. Timothy Earle (1991a) has put forth a strategy for

accomplishing this by integrating three levels of rationality into his research (Earle 199la:85-89).

Economic rationality is related to how people attempt to satisfy their basic needs -- food, shelter, and

reproduction. Cultural rationality refers to the ability of people to model their behavior based on the

success ofothers. These successful behaviors are often "traditional" ways of doing things that may not

always appear rational in terms of immediate costs and benefits, but because of their continued use and

success, they reduce the personal cost of experimentation. Finally, evolutionir rationality refers to










16

decisions that are based on maximizing reproductive success; for example, decisions that increase the

success of child rearing.

Although Earle does not specifically discuss scales, it is clear that his various kinds of

"rationality" represent levels of increasing scalar complexity. He recognizes, for example, both the

autonomous nature of the individual as well as the necessity for group association; he identifies

individuals as rational decision-makers who also are subject to the values of the cultural systems that

they participate in; and he relates all of these to the Darwinian notion of reproductive success. Thus,

by recognizing the overlapping nature of decision-making, which occurs on a number of different scales,

the model attempts to address the multicausal nature of human behavior which is the result of a complex

interaction between genes, environment, learning, and individual choice (cf. Winterhalder and Smith

1992:15).


Identification of Effective Scale

As the above discussion implies, specifying the effective scale at which one is working is critical

to the development and evaluation of a specific research strategy. In this study, I was concerned

primarily with identifying organizational responses to limited resource availability, determining how

these varied geographically and temporally, and understanding the factors that resulted in these

strategies being selected for. Consequently, the scales of analysis with which I chose to work are

relatively large. Geographically, I have chosen a regional scale of analysis, focusing on south Florida

and, more specifically, the Kissimmee region, which encompasses both the Kissimmee River valley and

the Lake Wales Ridge. This region was chosen because it is one of the few archaeological regions in

Florida where the prehistoric inhabitants did not have direct access to durable raw materials such as

chert, marine shell, or sharks' teeth. To provide a means of comparison, I have also included data from

archaeological sites outside of my primary study area, specifically the Peace River valley, the southwest

coast, and the chert-rich regions of west-central and north-central Florida.










17

Temporally, I have chosen to focus on two, broadly defined cultural periods -- the Archaic and

post-Archaic. The rationale for choosing such large units of time are partly methodological and partly

theoretical. Methodological issues are discussed more completely in Chapter 7; however, it is sufficient

to state here that shorter temporal scales were difficult to identify archaeologically at many of the sites

that provided lithic data for this project. This made the rather coarse-grained temporal distinction used

in this study the only viable alternative for examining diachronic trends. This two-period separation

does possess validity beyond the limitations of method in that existing paleoenvironmental and

archaeological data from south Florida indicate substantive differences between the two periods in terms

of environmental and climatic conditions, as well as settlement and subsistence strategies (see Chapters

3 and 5). Specifically, the availability of extensive marsh and wetland habitats after about 3500 B.P.

appears to have contributed to the adoption of a more sedentary lifestyle in the interior of south-central

Florida. This shift from a relatively mobile settlement strategy to one of sedentary or semi-sedentary

existence provided an opportunity to include differential residential mobility as a variable in the

temporal analysis.

These changes in climate and environment, and the corresponding settlement and subsistence

strategies that were employed in response to these changes, are important factors in the consideration

of changing organizational strategies since the geographic distribution of lithic raw materials has

remained unchanged since before humans first arrived in Florida nearly 12,000 years ago. Differential

access to lithic resources through time is a function of 1) physical factors such as erosion, exposure,

and in some cases, inundation by rising sea levels, and 2) cultural factors such as over-exploitation, the

establishment of political territories, the development of exchange systems, and settlement mobility.

Because they act at temporal scales that exceed the life spans of individual humans, the effects of

physical processes on human procurement strategies are relatively minor in comparison to the effects

of cultural practices.









18

The mobility strategies that were employed prehistorically by humans moving across the

landscape to exploit the natural environment are perhaps the most important of these cultural factors.

Because subsistence and raw-material resources are distributed unevenly in the environment, any single

settlement location will provide direct access to some resources while at the same time reducing access

to others. Highly mobile populations could structure their settlement patterns to provide access to

necessary resources, including lithics, on a periodic basis. Less mobile populations, on the other hand,

were required to situate their settlements in locations that maximized access to critical resources such

as permanent water or abundant and dependable food sources. Consequently, access to other, less

critical but equally important resources may have been restricted. The development of exchange

systems and political territories can be explained, in part, as attempts to gain access to resources not

available in the immediate environment, or alternatively, to restrict access by others to locally available

resources. Thus, the examination of the organizational strategies employed in chert-poor areas becomes

closely entwined with settlement variability and aspects of the regional ecology.


The Conceptual Basis


Environmental Structure

The focus on lithic resources might ordinarily place a study such as this in the general category

of "technological studies." I view it, however, as intrinsically ecological and economic in nature

because I emphasize the importance of environmental structure in terms of its influence on changing

strategies of resource procurement and technological organization (cf Bouseman 1993; Collins

1975:20). The physical properties of a raw material, its geographic distribution, as well as abundance

and availability are the relevant environmental factors that affect procurement strategies, assemblage

content, and variability in the organizational properties of a technology. Understanding the geological

and geographic parameters of the resource base is thus critical to any study of lithic resource use (May









19

1980). However, in order to integrate studies of technological organization into an understanding of

evolutionary change, it also is necessary to incorporate information on the abundance and availability

of subsistence and other resources since these influence the larger sphere of economic strategies of

which technology constitutes one part.

[n a general sense, structure refers to a complex system which by definition is composed of

many parts, arranged together in some organized way. Thus, environmental structure consists of the

interrelationships between climate, topography, hydrology, flora and fauna, and raw-material

distribution. Variability in environmental structure occurs at a number of different scales of space and

time, and so affect human populations differently. Higher-order variability, such as long-term

fluctuations in climate and precipitation within a geographic region that occur over several millennia

may not be observable within the lifetimes of individuals, but may be important for studying

evolutionary processes of human populations over time. Lower-order variability, such as the day-to-day

or year-to-year patterns of rainfall, is more likely to affect the choices individuals make (Butzer

1982:24). These individual decisions provide the behavioral variability on which selection acts. The

aggregate effect of human choices within the context of environmental structure results in adaptation

via the mechanism of selection.

Jochim (1981:8-9) has observed that critiques of ecological studies in anthropology have tended

to operate under two misconceptions. The first is that "environment" is restricted to the physical,

natural environment, i.e., climate, weather, landscape, and exploitable natural resources. The second

is that an ecological study is restricted to descriptions of food-getting behavior. Both of these

assumptions are erroneous. The definition of environment that is used here follows that proposed by

Winterhalder and Smith (1992:8): "..environment is defined as everything external to an organism that

impinges upon its probability of survival and reproduction. The effects can bear on development,

physiology, or behavior, and their sources can be physical, biological, or social." Thus, the environment









20

consists of both the natural and social environment, and any ecological study must account for the

various interactions between individuals in a population as well as the interactions between the

population and the physical environment in which it exists. Furthermore, as Jochim (1981:8) notes,

"ecology is the study of dynamics." It focuses on how human behavior responds to an environment that

is constantly changing. Because humans must make decisions based on perceptions of their effective

environment, ecology attempts to identify those factors in the external environment that are most

important in influencing the decision-making process.

The focus on prehistoric decision-making in an ecological context forces us to confront the

problem of "objective" versus "cognized" perceptions of the environment While the cognized

perceptions of the individual participants of a social group are important elements of any ecological

study (e.g., Johnson 1982), these may be difficult to identify archaeologically. A more plausible strategy

is to attempt to identify the decision-making processes of the participants in a culture (e.g., Mithen

1990). An alternative approach is to take an objective perspective and apply general models of human

decision-making that can be compared cross-culturally. This is the strategy that is used in this study.


The Ecological Approach

The application of ecological concepts to the study of cultural behavior has a long history in

anthropology and archaeology (Bettinger 1991; Harris 1968), but its emergence as a distinct research

strategy is credited to Julian Steward who developed the field of study known as "cultural ecology"

(Steward 1955). Cultural ecologists assume that environmental adaptation is dependent on the nature

of the available technology, the needs and structure of the society, and the specifics of the local

environment. The goal is to determine whether these adaptations initiate concomitant changes, or

"transformations," in the social structure and whether these changes are evolutionary in nature (Steward

1977:43). The cultural-ecological paradigm was embraced by anthropologists and archaeologists of the









21

1960s and 1970s who incorporated it into a systemic view of culture and cultural evolution (e.g.,

Binford 1962: Clarke 1978: Flannery 1968: Lee 1968: Plog 1974; Vayda 1969).

While the application of a cultural-ecological perspective resulted in many important insights

into human cultural behavior, it suffered from the fact that it was essentially a functionalist paradigm.

In classic functionalism, the role that social behaviors play is to maintain the structure and viability of

the social system. In cultural ecology, function is replaced by adaptation, and the role that certain

behaviors play is to maintain the viability of the social group within a particular environment. Thus,

cultural-ecological arguments were often circular; if behaviors existed and persisted, they must have

done so eas they were adaptive. There were no criteria for evaluating whether or not (or the degree

to which) certain behaviors are, in fact, adaptive; nor was there any way to evaluate whether behaviors

are nonadaptive or maladaptive (Bettinger 1991:57-58; Smith and Winterhalder 1981:3).

Cultural-ecological research strategies also tended to assume that culture is a steady-state

system and that all sources of change are external to the system (e.g., climatic change, population

growth). However, as recent reviews of evolutionary theory have pointed out, a systems approach that

assumes equilibrium is inappropriate to study evolutionary change (Bettinger 1991; Dunnell 1980;

Wenke 1981). Not only does it fail to identify the mechanisms of change, that is, how change actually

occurs within a sociocultural system, it ignores the role of the individual as an agent of change. This

last complaint is often accompanied by charges of ecological or technological determinism and

biological reductionism (e.g., Conrad and Demarest 1984:207: Robarchek 1989).

The most recent utilization of an ecological research strategy is the field of evolutionary

ecology. According to Winterhalder and Smith (1992:3), evolutionary ecology is "the study of evolution

and adaptive design in an ecological context." The field has its origin in biology and has only recently

been applied to human cultural behavior. Its practitioners apply natural selection theory in an effort to

explain the evolutionary development of human behaviors, and so attempt to utilize the same










22

explanatory principles that underlie biological evolution. According to Bettinger (1991:154),

evolutionary ecology incorporates the contradictory concepts of change and stability into a single,

coherent theory. Unlike sociobiology, evolutionary ecology does not require that behaviors be

genetically determined. It requires only that behaviors that affect individual fitness must be guided by

the principles of natural selection (Bettinger 1991:154; Winterhalder and Smith 1981:6). It does

assume a capacity for adaptive decision-making and it further assumes behavioral variability, which

is necessary for evolution and adaptation to occur. Significantly, Winterhalder and Smith (1992)

emphasize that:


..most behavior is a result of evolutionary mechanisms and processes operating

at several scales of space and time. Behavioral responses, especially those

involving learning, are most likely when the relevant environment has qualities of

high variance, novelty, and unpredictability...The spatial and temporal patten of

the relevant environmental features must be characterized with appropriate

variables and in sufficient detail to capture these qualities...[Evolutionary

ecologists] predict diverse and flexible behavior, contingent on localized and often

changing conditions. The great variety of states possible in an organism's

immediate environment leads behavioral ecologists to expect corresponding

variety in the expression of behavior [Winterhalder and Smith 1992:9; emphasis

in originall.


Economic Decision-Making

Burling (1962) defines economic behavior as that which allocates limited means (resources)

to achieve unlimited wants, needs, or desires. "Economics defined in this way has no necessary

connection with the use of money or material objects. Since we are disposing of scarce means in









23

virtually everything we do, economics in this view focuses on a particular apt of behavior and not

on certain kinds of behavior" (Burling 1962:810-811. emphasis in original). The aspect of behavior

that Burling refers to is "choice." Since it is not possible for people to utilize their limited resources to

satisfy all of their desires, choices must be made regarding the differential allocation of resources to

obtain specific wants or needs. This makes scarcity a fundamental concept in economics (Glahe and

Lee 1989:2-3) since it affects perceptions of value as well as allocation choices.

According to neo-classical theories of economics, value is created by supply and demand, and

is market-driven; in other words, the ability of a consumer to freely choose among purchasing

alternatives affects the exchange value of the commodities being offered (Baber 1987:48). On the other

hand, Marxism claims that the value of a good is defined, in part, by the amount of labor time invested

in its production (Baber 1987:54). The position taken here is that both theories present models of

economic behavior that are approximations of reality, and both are useful for understanding economic

decision-making. More to the point, a theory of labor value can be accommodated easily within a

supply-and-demand model if Burling's definition of economics is accepted. Labor time is a resource

that is limited in its availability and must be allocated among competing tasks (e.g., Smith 1979;

Winterhalder 1983). If we choose to use our resources to conduct a certain task, we are simultaneously

choosing to forgo the benefits of performing another task. The value of this "sacrificed consumption"

is the opportunity cost of the choice or decision we have made (Glahe and Lee 1989:4). The concept

of opportunity cost is a critical component in understanding non-Western economic systems, specifically

those of hunter-gatherers where labor time and energy are the primary forms of "currency." As several

recent studies have demonstrated, hunter-gatherers must evaluate the net benefits of allocating time to

activities not directly related to the acquisition of energy, such as reproduction, child-rearing, social

activities, and so on, with those of foraging (Hawkes 1987: Hawkes et al. 1985; Smith 1988:









24

Winterhalder 1983). These decisions, in turn, affect how the products obtained through the labor effort

are distributed and consumed within the society (Winterhalder 1990:10).

Models of economic behavior. According to Hirschleifer (1980), all economic problems can

be reduced to two questions: 1) what factors cause or explain fluctuations in the economic environment,

and 2) which of several alternative strategies will provide the best outcome? The first question is an

eulibrium problem and is solved through the analysis of supply and demand. The second question

is an g niaiuln problem and is addressed through the method of marginal magnitude. Although these

methods are not utilized in a formal sense in this study, the principles that underlie them are used to

develop hypotheses regarding lithic resource procurement and use. Therefore, it is necessary to review

their basic components and provide some illustrative examples.

Fuilaibrium analsis utilizes the relationship between two variables: price and quantity. Price

represents the amount of one commodity that is given in return for receiving another commodity. Price,

then, is equivalent to cost since it consists of that which must be given up, done, or sacrificed to obtain

something else. The factors that affect the price (or cost) of a commodity and the quantities that are

consumed are supply and demand. The relationships between these are shown in Figure 2. The slope

of the demand curve (DD) is negative reflecting the general principle that, other things being equal, the

quantity that is consumed will increase as the price or cost of the commodity decreases. On the other

hand, the slope of the supply curve (SS) is positive reflecting the principle that the quantity of a

commodity that is supplied will tend to increase as the price increases. Equilibrium is achieved at the

intersection ofDD and SS, which reflects the point at which all individuals are able to acquire the good

at their desired rate and all suppliers are able to distribute their good at their desired rate. Changes in

the position of the equilibrium point occur as a result of changes that occur within the economic

environment, in other words, changes in supply or demand.










25





0 s
















0 0 100
Quanuty Consumed


Figure 2. Generalized supply-and-demand curve.


It is possible to modify these variables (i.e., price and quantity) to consider the relationships

between value, probability of use, time, energy, or distance versus quantity, utility, or preference. It also

is possible to utilize equilibrium analysis to examine the effects of changes in supply and demand on

the position of the equilibrium point for an economic environment in which supply is not under the

control of individuals offering goods for purchase in a market-place but, instead, is dictated by the

differential distribution of a desired resource in the natural environment. In Figure 3, price is

represented by the labor cost (which could be measured in terms of time or energy) that must be

expended to obtain a resource. Figure 3a illustrates what happens when the supply curve is shifted to

reflect what occurs when a group ofhunter-gatherers moves to an area where the desired resource is not

available in the local environment. The new equilibrium point (P,) defines a point at which the cost of

acquisition has increased and the amount demanded has decreased as a result of increased distance to

the resource.











26



A


100 --------------------

0 0 S














S D




o Q 100
Quantity Consumed







B


100 --------------------- -*

D s S







s P.




S



S D

Q Q 100
Quantity Consumed


Figure 3. Supply-and-demand curves showing the effect of changes in demand (a) and supply (b).









27

When interpreting this curve it is important to recognize the distinction between a decrease in

the amount demanded and a decrease in demand (Glahe and Lee 1989:49-50). The former is

represented by a shift along the demand curve while the latter is represented by a shift in the demand

curve. There has been no shift in the demand curve in Figure 3a; rather, supply has decreased as a result

of moving to an area where the resource is not available. This has caused a disequilibrium in the

economic environment where demand at the lower cost exceeds supply under these new settlement

conditions. The increase in acquisition cost resulting from this settlement shift results in a decrease in

the amount or quantity of the amount consumed, thus establishing a new equilibrium point.

The alternative case is shown in Figure 3b. Here supply remains constant. Instead, demand

changes, perhaps as a consequence of an increase in population which results in an increase in

consumption. Consequently, the demand curve moves to the right, along the supply curve, and a new

equilibrium point is established. Disequilibrium at the initial labor cost occurred because the increased

rate of consumption exceeded the supply. The greater cost at the new equilibrium point is a result of

the intensification of labor effort necessary to meet the greater demand of a larger population (cf. Earle

1980:19).

Equilibrium models have wide applicability in economic analysis and they have been used to

address a number of resource problems in anthropology. The supply-and-demand relationship is often

implied rather than stated formally (e.g., Brumfiel and Earle 1987; Rathje 1971), perhaps because the

model is conventionally presented by economists within the context of market-oriented economies.

Binford (1978) uses a modified form of the classic supply-and-demand relationship to model butchering

strategies among the Nunamiut (see discussion in Bettinger 1991:107-108). Similarly, Chayonov's

(1966) argument that peasant households attempt to strike a balance between work and drudgery can

be adapted to a supply-and-demand format (Figure 4). As the time devoted to work increases, the desire

for leisure time also increases, while alternatively, as earnings increase, the desire for additional income










28




















LL

E E a
Earnings


Figure 4. Equilibrium analysis of Chayanov's model of work versus leisure time in peasant societies.


diminishes. The goal of Chayanov's model is to strike a balance between work and leisure time; that

is, to achieve a form of economic equilibrium. It also demonstrates the effect of opportunity cost on

economic decision-making. From this model it is possible to predict under what conditions individuals

will make choices to forgo leisure to increase earning or, alternatively, to forgo greater earnings to

achieve more leisure time. It also is possible to examine what the effects will be on the variables of

work time and earnings if there is a change in either the desire for income or the desire for leisure. The

first could come about as a result of more dependents in the family, while the second could occur as a

result of having to work more hours each day to support these dependents.

Demand curves also can be used to analyze and predict consumer choice between different types

of goods (Glahe and Lee 1989:72-90). The variables used in this type of analysis include utility,

income, and price. Utility is the level of satisfaction that an individual derives from a particular

commodity. Choices are assumed to maximize utility subject to the constraints of income and price.









29

One of the basic assumptions of this type of analysis is that the utility derived from additional quantities

of the good decreases as the amount of the good that is obtained increases. In other words, the more one

has of something, the less value or utility is gained from additional quantities that are acquired by the

consumer. This is known as the law of diminishing marginal utility.

Demand curves highlight two points that are important to this study: elasticity of demand

subject to the constraint of price or cost and substitutability. Elasticity of demand is the change in

demand that occurs as a result of a corresponding change in price (Glahe and Lee 1989:73). Elasticity

of demand may change as a result of the amount of total income available, preference, or the availability

of suitable alternatives. If the first two remain constant, then it is the availability of alternatives that

determines elasticity. This is illustrated in Figure 5. If the price of one resource (A) increases and the

quantity consumed of an alternative resource (B) also increases, the two resources are said to be

substitutable. On the other hand, if the quantity of resource B that is consumed decreases as the price

of A increases, then the two resources are said to be complementary. Croquet mallets and balls are an

example of complementary resources while beef and chicken might be considered substitutable. In

terms of lithic procurement and use, it is possible to use demand curves to observe how changes in the

availability of one type of raw material affect demand for a second raw material; for example, local

versus non-local stone or stone versus marine shell. This is illustrated in Figure 6 which shows the

effect on the amount consumed of two substitutable commodities (resources) when the price (or

availability) of one increases. As indicated, the availability of the cheaper alternative resource that can

perform the same task diminishes the utility, and hence the amount consumed, of the more expensive

resource.

Optimization analysis is a method for determining which alternative among a range of

alternatives will supply the best outcome (Hirschleifer 1980:44). This range of alternatives can be

reduced to one of three strategies: maximization, minimization, or stability (Stephens and Krebs











30








Complements









Subsitutes
0











0 100
Quantity Consumed of B



Figure 5. Elasticity of demand between two resources.












I

Y







0 o
Quantity Consumed of A



Figure 6. Effect of price or availability on the quantity consumed of two substitutable resources. An
increase in the cost of resource A, as indicated by the steeper slope (2), results in a large shift in the
mixture of A and B consumed (from X to Y); in other words, less of A is consumed in preference to the
cheaper B.









31

1986:7). If the best outcome is gained by choosing an alternative that is dependent on decisions made

by other individuals, then stability is the correct strategy to choose (Maynard-Smith 1982). If, however.

a choice that results in the best outcome is not dependent on the decisions of others, as is the case when

a forager must make a decision regarding the procurement of resources in the physical environment, then

maximization or minimization principles are appropriate (Stephens and Krebs 1986:7).

These strategies assume that the individual is aware of all the alternative options that are

available, knows their potential outcomes, and can accurately calculate which will prove to be the

optimal choice. A fourth strategy, referred to as "satisficing," assumes that the individual does not have

complete access to all necessary information and so will not act optimally. Instead, the individual will

choose any alternative that will provide an outcome that is perceived as "good enough" (Smith and

Winterhalder 1992b:54). However, as Smith and Winterhalder (1992b:54-55) argue, there are no

objective means of determining what criteria the individual uses to determine what is "good enough."

This determination can only be made by observing the individual make the decision. Therefore, all

analyses that utilize a satisficing model tend to employ circular reasoning. Moreover, for evolutionary

ecologists, satisficing has no meaning in terms of natural selection.

Central to the analysis of optimizing behavior is the concept of marginal value (or more

generally, marginal magnitude) which is based on the relationship between Total, Average, and

Marginal Returns. Averae return is simply the ratio of input to output using whatever currencies are

relevant to the analysis. For example, input may be measured in terms of tune or energy expended while

output may be measured in terms of protein, calories, or volume of a procured resource. Thus, the

amount of currency (output) divided by a specified unit of input (e.g., I hour or 500 calories of energy

expended) is the Average Return. This also is a measure of economic efficiency (e.g., 500 calories

procured per hour), and the inverse is the unit cost (e.g., 500 calories "costs" one hour of labor time:

250 calories "costs" one-half hour of labor time) Tota Return is simply the Average Return multiplied








32

by the total amount of input. Marginal Return is the change in total return per unit change in quantity.

or in this case, labor input. Expressed mathematically, this relationship is ATR ALT = MR. So, if

the average return for one hour of labor is 500 calories, but during the second hour the average return

is only 300 calories, the total return is 600 calories (300 x 2 hrs), but the marginal return is only 100

calories (600-500=100; 2 hrs I hr= 1: 100/1=100). In other words, the net return for the additional

hour of labor is only 100 extra calories.

Because total returns are a function of average returns, as the average return rate decreases so

too does the marginal rate of return, even though total returns may continue to rise. Another way of

saying this is that decreasing marginal returns reflect a slower rate of growth in total returns due to a

reduction in labor efficiency. The point at which average and marginal returns begin to decrease is

called "the point of diminishing returns" and it is illustrated graphically in Figure 7. Beyond this point

it is no longer beneficial to increase labor costs because for each additional unit of time expended at the

increased cost, the rate at which the total net return increases begins to diminish rapidly. If labor costs

continue to increase without a corresponding increase in total output (i.e., if the average return rate

continues to decrease), marginal returns will eventually equal 0 and total returns will begin to decrease.

Optimization analysis uses this method to evaluate when an individual should switch economic

strategies in order to maximize his or her net rate of return. Typically, a switch should be made when

the marginal return for the existing strategy is equal to or less than the average return for the alternative

strategy. This is based on the assumption that maximization of returns and economic efficiency are the

goals; however, the method can also be used to evaluate the costs associated with different strategies

rather than returns. In this type of application, the inverse of Average Returns (i.e., unit cost) is used

to calculate Total and Marginal Costs.

An example of this type of application is the work of Earle (1980). Operating under the

assumption that foragers and primitive horticulturists, when confronted with economic choices, will

choose those strategies that tend to minimize costs, Earle compares the marginal costs associated with









33



















AR

MR
-200
Labor Time




Figure 7. Graph of diminishing marginal returns. As the average return per labor hour (AR) decreases,
so too does the marginal return (MR), even though total returns (TR) continue to rise. The optimal
labor effort, indicated by (a), is the point at which average and marginal returns are at their highest.


three different subsistence strategies (hunting, collecting, and horticulture) to predict what mix of

strategies will provide the necessary nutritional requirements at the lowest cost. Given changing

conditions, such as an increase in population, the optimal mix of strategies may also change in response

to new levels of nutritional demand. Earle's analysis suggests that when this occurs, the strategy mix

that will be chosen will be the one in which the rate of increasing average costs is lowest.

Evolutionary ecologists employ optimization analysis in optimal-foraging theory under the

assumption that the goal of individual decision-making is to increase fitness in an evolutionary sense.

There are a number of different models that have been developed using optimization principles to

examine specific problems associated with human and non-human foraging. Excellent overviews are

presented in Bettinger (1991). Smith (1983), Smith and Winterhalder (1981), Stephens and Krebs









34

(1986). and Pyke et al. (1977). I will focus here only on those models that contain elements applicable

to lithic procurement and use.

The two most widely used models are the diet-breadth and patch-use models. The diet-breadth

modl assumes that resources are evenly distributed in the environment and that the forager encounters

these resources at random. It also assumes that foragers rank their prey in terms of the handling costs

necessary to procure them versus the net energy return yielded. This relationship is usually stated

formally as E Th, where E = gross energy acquired minus energy expended and Th = the time required

to pursue, capture, and consume the prey (Bettinger 1991:86). Time devoted to searching for the prey

is not included in determining the rank of the resource, but it is included in the decision to capture an

encountered prey. The model predicts that a prey will be chosen to be in the optimal diet if the marginal

energy return is equal to or greater than the average return for all higher-ranked prey. In other words,

if the ratio of energy return to handling time for a prey that has just been encountered (i.e., the forager's

marginal return) is greater than the ratio of energy return to handling time plus the additional search

time necessary to locate another prey, then the encountered prey will be included in the optimal diet.

The more prey types that are included in the diet, the more likely it is that suitable prey will be

encountered, and so search time associated with foraging tends to decrease. On the other hand, since

prey are ranked by the amount of handling time associated with their capture, as more lower-ranked prey

are included in the diet, the amount of time that needs to be devoted to pursuit and capture tends to

increase. So there is a trade-off between search time and handling time: as search time decreases,

handling time increases. This relationship is shown in Figure 8a. Note that what is plotted is the

change in these two variables (i.e., the reduction in search time versus the increase in handling time) that

is realized with the addition of an additional prey item (cf MacArthur and Pianka 1966:Figure 1). Note

also that the form of this relationship between handling time and search time is essentially that of a

supply-and-demand curve (compare with Figure 3). In Figure 8b a decrease in handling time results in









35


A




AH














0 10
Diet Breadth




B






AA












0 10
Diet Breadth


Figure 8. Graphic representation of the diet-breadth model (after MacArthur and Pianka 1966:Figure
I): a) As search time (AS) decreases, handling time (AH) increases. The optimal foraging strategy (A)
is achieved at the point where the addition of an additional prey species would increase handling time
more than it would decrease search time; b) Effect of decreasing handling time on the optimal strategy.
Here changes in the forager's ability to pursue, capture, and process resources more efficiently increases
diet breadth from A to B and decreases time spent per unit of energy.









36

a change in the equilibrium point such that the time per unit of energy obtained (i.e., cost) decreases and

diet breadth increases even though search time remains the same. This is an important consideration

for human foragers who are able develop more efficient techniques for decreasing the amount of

handling time associated with specific resources.

The patchchoice model (McArthur and Pianka 1966) is slightly different in that it assumes that

resources are not distributed evenly in the environment, in other words, they are clumped together in

resource "patches." Nor are they encountered randomly. The resource patches are known to the forager,

who is able to calculate the amount of time it will take to travel between patches to exploit the resources

there. Unlike the diet-breadth model differential depletion of resources in a patch can cause that patch

to be excluded from the foraging itinerary because increasing exploitation results in a decreasing rate

of return. The patch-choice model is a more realistic characterization of human foraging for lithic

resources since these are immobile, geographically localized, and subject to depletion. Figure 9

graphically illustrates the patch-choice problem. The basic elements are similar to the diet-breadth

model: resource patches are ranked in terms of handling time, and there is a trade-off between declines

in the forager's marginal return (returns per unit of handling time) in patches of lower rank versus the

time necessary to travel between patches. Charov (1976) refined the patch-choice model further by

applying the marginal-value theorem to determine when the forager should decide to leave one patch

and move to another. This should occur when the rate of return in the patch drops to the average overall

rate that is obtainable from the environment as a whole when travel between patches, search within

patches, and handling costs are included (Figure 10).

The central-place-foraging model (Orians and Pearson 1979) represents a further refinement

of the basic diet-breadth model wherein foraging is modeled as a trip from a specified point (central

place) to a resource patch and return. The round-trip travel time is included in the analysis of costs and

benefits. The model predicts that as round-trip travel time increases, the time spent foraging in the









37













A



'T


0 10
Patch Breadth


Figure 9. Graphic illustration of the patch-choice model (after MacArthur and Pianka 1966:Figure 2).
Resource patches are added to the foraging inventory as long as the decrease in between-patch travel
time (T) is greater than the increase in within-patch foraging time (H). The optimal strategy (A) is
achieved when the inclusion of an additional patch would increase search time within patches more than
it would decrease travel time between patches.


patch and the minimum acceptable prey size should all increase accordingly in order to offset the extra

travel cost. If traveling time is reduced to zero, a forager should take any prey that is encountered.

These predictions assume that there is no handling time involved nor do they include calculations of

transport load of the return trip as a cost. Additional refinements add these variables into the analysis.

For the first of these modification, the authors conclude that for short travel times, the forager

should choose those prey that provide a higher return in terms of energy per unit of handling time, while

for long travel times, higher-energy prey should be taken regardless of travel time. This conclusion is

based on the inclusion of travel time into the energy equation which increases total foraging time and

so necessitates that higher-energy resources be taken. Prey are ranked according to the mathematical

relationship E = C -T, + T, where E = energy return rate, C = net calorie content, T, = travel time, and










38






Optimal Depatumr Overa Rate of Retun
Time (Travel Search Handing)








,- I Im-
1 Within-pafth Rate
S.1 ^ofReturm





i Accapta Opmal

Travel <- > Foraging
TIME


Figure 10. Application of the marginal-value theorem to the patch-choice problem (after Charov
1976). For a given patch, the rate of energy return is a function of time spent foraging within the patch.
As resources within a patch become increasingly rare, within-patch foraging time increases and the rate
ofenergy return decreases. The optimal point at which the forager should leave the patch occurs when
the within-patch rate of return drops to the overall rate of return obtainable from the environment as a
whole when all travel, search, and handling times are considered.


Th= handling time (Orians and Pearson 1979:162). As the authors note, this formula is essentially the

same as that for ranking prey under non-central-place foraging theory except for the addition of travel

time. Therefore, the same prey type may be ranked differently or considered more or less profitable,

depending on distance from the central place (Stephens and Krebs 1986:54-55). This is illustrated by

a simple example of two foragers exploiting the same resource type at different distances from the

central place. For Forager I with a long travel time, C, = 500, T, = .5, T,, = 1.5, and E, = 250, while

for Forager 2, with a short travel time, C, = 500, T = .5, T, = .5, and E, = 500. Because of shorter

travel time, Forager 2 can capture a prey with C. < C, and still maintain a high net rate of energy return

relative to handling costs: C, = 300, T|, = .5, T,~ = .5, and E, = 300. But, if handling time increases









39

substantially, then E, < E, for example: C = 300. T:= .75, T,,= .5. and E,= 240. In the extreme

case, where T, = 0, the ranking ofprey should be based solely on the ratio of net return to handling time.

In terms of optimal load. and assuming that patch quality (i.e., abundance of highly ranked

resources) is held constant, Orians and Pearson (1979:164) suggest that both the size of the load and

the time spent foraging in a patch should increase with increasing distance to the central place.

However, more distant patches should not be utilized until the load increases to a point sufficient to

compensate for the greater travel time (Orians and Pearson 1979:166). If distance is held constant, and

patch quality is variable, then the time spent foraging in patches with high resource abundance should

decrease, although the optimal load may increase slightly.

These theoretical explorations indicate that for central-place foragers, decisions regarding

resources are dependent on the net return of each resource relative to both travel and handling time

(Bettinger 1991:96). Where travel time is low, handling time is more important because travel costs are

not great enough to affect the ratio between net return and total acquisition time. On the other hand,

when travel time is relatively great, handling time becomes increasingly less important and package size

increases in importance.

Finally, Orians and Pearson (1979:170) predict "if maximization of rate of energy delivery to

a central place is the sole criterion influencing selection of a central place, then the optimal central place

is one that minimizes travel time, that is, the site that lies at the 'center of gravity' of the food

distribution." In other words, cost minimization, in terms of travel time, is an optimal strategy for

maximizing the rate of return, a conclusion that mirrors those of anthropologists who have studied

hunter-gatherer settlement and subsistence strategies (e.g., Hayden 1981; Lee 1979; Yellen 1976).

Tme alloction. Yet another application of economic models which is applicable to this study

is the concept of time allocation. This has been referred to briefly above with regard to the concept of

"opportunity cost." Schoener (1971) introduced the notion of time allocation to optimal-foraging









40

studies when he considered two limiting cases to the general model of stationary, as opposed to mobile.

foragers maximizing their rates of net energy returns: time minimization and energy maximization

(Schoener 1971:376). Tim minimizers are animals that attempt to maximize fitness by limiting the

amount of time spent feeding to acquire a given energy requirement. Alternatively, ener maximizers

attempt to maximize fitness by maximizing the net energy acquired over a given time period. Smith

(1979) rephrased these concepts in terms of energetic efficiency and applied them to human foragers.

According to Smith (1979:62), in environments where the net energy available to the forager is limited,

fitness can be increased by increasing the net rate at which energy is captured over a given time period

(i.e., energy maximization). In situations where time, but not energy, is a limiting factor, his model

predicts that fitness can be increased by reducing the amount of time necessary to capture a given

amount of energy (i.e., time minimization). By minimizing the time allocated to energy capture, more

time is freed to devote to other activities that may serve to increase the adaptiveness of the forager

(Smith 1979:63).

Smith (1979:61) demonstrates mathematically that regardless of which strategy is employed,

the result is the same: an increase in the net rate of energy return. For example, where E, = net energy

gain (i.e., energy acquired minus energy expended, or E- E,), T, = time required to acquire E., and R,

- net acquisition rate (or E, T), then let E, = 10, T, = 2, and so R = 5. For a forager practicing a

strategy of time minimization, decreasing T, results in an increase in R,: 10 I = 10. Similarly, for a

forager practicing a strategy of energy maximization, increasing E, results in an increase in R,: 20 +2

= 10. By transposition, the equation R = E, + T. becomes E, = (R,)(T). Smith (1979:62) relates this

relationship to fitness under the assumption that an increase in E, will lead to an increase in fitness and,

therefore, fitness can be increased by increasing either R, (the net rate of acquisition) or T. (the time

devoted to acquiring energy). However, because more time spent acquiring energy necessarily results









41

in less time spent conducting other activities (T,). there reaches a point where fitness begins to decrease

as T. increases and To decreases.

This last point focuses on the opportunity costs and benefits of adopting the two different

strategies. Winterhalder (1983) expanded on this by using indifference curves to demonstrate that time

minimization and energy maximization are simply two poles at either end of a continuum of behavioral

options that balance the trade-off between time and energy. This enabled him to predict under what

conditions foragers should adopt one or the other strategy (Winterhalder 1983:78-81). When only a

small increase in energy gain is required to compensate for a large decrease in non-foraging time, then

the forager is said to be energy-limited. Conversely, when a large increase in energy gain is required to

compensate for a small reduction in non-foraging time, the forager is said to be time-limited. For

foragers that are time-limited, the optimal strategy is to decrease the amount of time spent foraging as

high-ranking resources become more abundant in order to obtain the same or greater intake. This leaves

more time available for non-foraging activities. For foragers that are energy-limited, the optimal

strategy is to increase the time spent foraging and increase diet breadth as high-ranking resources

become more abundant. In other words, energy-limited foragers should forage longer, while time-

limited foragers should dispense with foraging chores more quickly.

Finally, Hames (1992) introduced the effect of technology into the time-allocation problem by

observing that tools that reduce the amount of time devoted to activities or increase total production

would increase the rate of energy return. He predicted that if a new technology resulted in a reduction

in time spent in energy acquisition without an increase in production, the group could be considered

time-limited, but if the new technology increased production without a corresponding reduction in the

allocation of time, then the group could be considered resource-limited (Hames 1992:218-219).

Summay. Based on the preceding review, the following economic principles have been

extracted. I consider these principles to be the most relevant for a study of lithic-resource procurement









42

and use, and they are used in subsequent sections to develop hypotheses for testing against the

archaeological record of south Florida.

1) As the cost of acquiring a resource increases, the quantity demanded or consumed decreases.

2) The elasticity of demand for two resources is based on available income, preference, and

differential availability. If the quantity consumed of one increases as the price, preference, or

availability of the other increases, then the two resources are substitutable. If the quantity of one

decreases as the price, preference, or availability of the other decreases, then the two are complimentary.

3) Value can be conceived of as the result of a combination of supply, demand, and availability

of alternative resources. Where demand for a commodity is high but supply is low, the commodity is

likely to be more highly valued. However, if a suitable substitute is available, then the value of the

commodity in limited supply may diminish.

4) The cost of procuring a resource from the natural environment consists of a combination of

travel and handling costs. The latter includes search, procurement, and consumption costs.

5) As the abundance of highly ranked resources in an environment increases, resource breadth

should decrease resulting in a specialized resource mix. Alternatively, as the abundance of highly

ranked resources decreases, resource breadth should increase resulting in a generalized resource mix.

6) An increase in the use of lower-ranked resources decreases search and/or travel time, but

increases handling time.

7) A reduction in handling time can offset the costs of traveling between resource patches.

8) For central-place foragers, as distance to a resource increases, package size is more important

than handling time. Alternatively, for central-place foragers who live in close proximity to a resource,

handling time is more important. For both, the optimal strategy is to choose the resource that provides

the highest return per unit of time expended in procurement.









43

9) Where resources are limited but time is not, the forager should spend more time in the

procurement effort in order to maximize the net rate of return. Where time is limited, but resources are

not, attempts should be made to maximize net return over a limited amount of time in order to maximize

the net rate of return.

10) The introduction of a new technology that either reduces time spent in energy acquisition

or increases energy returns will result in more efficient energy extraction (i.e., increase the rate of energy

return).

Critiues of economic modeling. Criticism of the use of neo-classical principles to study non-

market economies abound in the literature. The substantive argument that Western formalist theories

of economic behavior are inappropriate for studying non-Western economic systems (e.g., Chayanov

1966; Halperin 1994; Meillassoux 1978; Sahlins 1972) is based on the notion that economic rationality

is culturally relative. However, as discussed above, the definition of economics as the allocation of

limited means to achieve multiple goals or objectives requires only that: I) individuals make choices,

and 2) that these choices be based on a rational evaluation of the costs and benefits associated with

different courses of action. It does not require that economies be defined in terms of money or currency,

by material objects, or even by exchange relations. Time and energy are valid forms of "limited

resources" that must be allocated; the attainment of status and prestige are valid goals that require

decisions regarding the allocation of resources that must be made in order to achieve them. The ways

in which people manipulate their limited resources to achieve desired ends or goals is what constitutes

economic behavior. If this definition is accepted, then economic behavior (and attempts to economize)

are likely to be found in all societies and in all facets of social life. Thus, economic behavior is not

divorced or separated from other aspects of social life, it is integral to it.

A second major criticism is that neo-classical economic models contain unrealistic assumptions

about how humans think and behave and assume conditions that are not often found in the real world.









44

Thus, they cannot hope to explain the diversity of behavior that is present. As Bettinger (1991:105)

is quick to point out, this is tantamount to arguing that it is impossible to develop general theories that

can be used to explain natural phenomena. All models are approximations of reality. Attempts to

increase specificity to provide a model that approaches real-world situations necessarily suffer from an

inability to abstract to more general applications. Thus, different models may be more appropriate for

some questions than for others. Baber (1987:56-57), for example, has argued that Marx's labor theory

of value is based on value perceived from the perspective of society as a whole, while the neo-classical

theory of value "shifts the level of analysis from society to the behavior of the individual" (Baber

1987:57). He concludes by advocating the use of both concepts depending on the scale at which one

is working (Baber 1987:67-69).

Yet a third criticism is that optimization models as they are used in optimal-foraging theory are

tautological in structure (e.g, Keene 1983:145-146); in other words, they assume that optimal behavior

is adaptive and that all organisms practice optimizing behavior. Therefore, the goal of optimization

analysis is not to determine whether or not behaviors are optimal, but to demonstrate that optimal

behavior exists. Practitioners of optimal-foraging theory counter this claim by arguing that optimization

is not a hypothesis that is being tested. It is a method that is used to develop hypotheses regarding how

hunter-gatherers ought to act given specified conditions. These expectations are then compared with

actual behavior in order to determine whether a group is adapted or not (Bettinger 1991:105; Smith and

Winterhalder 1992b:50-5 I Stephens and Krebs 1986:207). By using this approach, it is possible to

avoid the problem of assuming that a group is well adapted initially, and then designing a research

strategy to demonstrate that this is so.

A criticism that is closer to the subject of this study is Jochim's (1989) comments on the use

of optimization principles in lithic studies. In particular, he argues that the role oflithic resources as

a currency is not well understood, due primarily to a lack of ethnographic studies of lithic-using peoples









45

and the economic role of lithics in these societies (Jochim 1989:107). This is simply the cultural

relativity argument all over again. For societies dependent on the physical environment to provide the

raw materials necessary to extract energy from that environment, economic decisions related to

acquisition of these non-subsistence raw materials constitute an important component of their economic

systems. Moreover, lithic raw materials are rarely conceived of as a currency in economic analyses.

They represent a resource that must be acquired from the physical environment, and that acquisition

extracts a cost in terms of direct labor time as well as opportunity costs. The currencies that are most

relevant for studying the economics of lithic procurement are time and energy, the same as are used in

most other optimal-foraging analyses.

Jochim (1989:107-108) also argues that the selective penalties of differential lithic-resource

use are not well understood and, therefore, the assumptions of evolutionary ecology may not be

appropriate for lithic studies. Torrence (1989b:2) echoes this concern. However, it is not necessary to

operate under the assumptions of evolutionary ecology in order to use optimization to model economic

behavior, although it is true that most economic studies in archaeology (including this one) assume that

adaptive processes were operative even if the assumptions are not explicit. More directly, as the studies

cited above demonstrate, any economic strategy that increases cost efficiency (i.e., results in a greater

output at a lower cost over a specified period of time) will tend to maximize a forager's net return rate.

Since the overall time and energy budgets of human foragers are assumed to be limited, individuals who

successfully employ strategies that maximize energy return or minimize the time expended to acquire

energy will have a greater propensity to survive and reproduce. Thus, prehistoric hunter-gatherers

confronted by changing environmental conditions would have been under strong selective pressures to

choose such strategies in deference to other, less efficient ones (Smith 1983:626). If the procurement

and use of lithic materials were important components of the energy-capturing economies of native

peoples, and so contributed in their own way to differential fitness, and if the procurement and use of









46

these materials entailed their own sets of costs in terms of time and energy, then human decisions to

choose between alternative strategies related to raw-material procurement, tool production, and tool use

should also have been subject to selective pressures.

A more germane criticism is Jochim's observation that there may be several technological

responses to a specific problem. For example, tool size might decrease in response to the needs for a

portable technology among highly mobile foragers, or it might increase if access to raw materials is

increased Optimization alone is not sufficient to predict which of these factors was, in fact, operable;

additional information is needed to make this determination (Jochim 1989:108). Consequently, Jochim

advocates a research strategy in which multiple lines of evidence are examined to arrive at a complete

understanding and explanation of lithic-resource use (Jochim 1989:108; see also 1983:167).

Certainly, a multidimensional approach to studying the past is most desirable, as the extended

discussion of scale has emphasized. However, Jochim's cautionary argument is based on what he

perceives as the uncritical use of economic principles in lithic studies. What is needed is the

development of theories of lithic-resource use that are derived from general economic principles, but

which do not apply these blindly without considering the special properties and characteristics of the

commodity itself and its use in prehistoric societies. An example of an attempt to develop such a

theoretical framework that is specific to lithic resources is Odell's (1996) discussion of the appropriate

archaeological indicators for determining the effects of differential mobility strategies versus raw-

material availability on lithic economies. Another is Bouseman's (1993) attempt to examine the fitness

benefits of different technological strategies under the assumption that the use of technology by human

foragers can affect search and handling times associated with subsistence pursuits. Furthermore, the

time allocated to tool production and maintenance must be balanced against the costs of foraging for

food. If two technological strategies have different production and maintenance costs, then assuming









47

that the rate of energy return remains constant, the technological strategy that is less costly in terms of

time allocation should offer the greatest fitness benefit (Bouseman 1993:63).


Economics of Lithic-Resource Use


Cost and Risk

Recent research indicates that variation in the organizational structure of prehistoric lithic

assemblages was influenced to a large degree by the availability and abundance of lithic raw materials

(e.g., Andrefsky 1994a, 1994b; Bamforth 1986; Dibble 1991; Francis 1983; Henry 1989; Kuhn 1991;

Marks et al. 1991; Newman 1994; Odell 1989a; Thacker 1996). These studies assume that when

resources are perceived to be limited, people will exploit and utilize them differently than if they are

perceived as abundant. They further assume that, in general, hunter-gatherers will adopt least-cost,

least-risk solutions to problems of resource procurement. Both of these assumptions appear to be valid

ones based on theoretical considerations of optimal foraging (see discussion above), as well as

ethnographic data of modem foragers and horticulturists (e.g., Cashdan 1990; Halstead and O'Shea

1989a; Hayden 1981; Johnson 1982, 1990; Johnson and Earle 1987; Kelly 1995; Lee 1979; O'Connell

and Hawkes 1981). Furthermore, we should expect that since time and energy budgets are limited,

additional expenditures of these currencies resulting from an increase in demand for lithic materials or

a decrease in supply resulting from limited availability would result in alterations to various aspects of

raw-material acquisition, tool production, use, and maintenance in order to contribute to a more efficient

processing of the available time and energy (Torrence 1989b:3).

In non-market economies, the principal cost is labor or energy expenditure, for which time or

distance can be used as proxy measures. For example, the costs associated with direct procurement of

lithic resources include the search time associated with locating outcrops containing suitable lithic

materials, time spent traveling to and from lithic outcrops once they have been identified, and handling









48

time associated with the actual procurement and processing of the raw material which eventually results

in a functional implement.

Binford (1979; Binford and Stone 1985) has argued that the costs associated with acquiring

these materials should be minimal because their procurement would have been embedded in other

activities. In the terminology of economics, the opportunity costs associated with lithic resource

procurement would be reduced by such a strategy. In a similar vein, Torrence (1983:12) has proposed

that time stress resulting from the need to conduct other, non-technological activities can be reduced by

embedding lithic procurement in subsistence activities. However, embedding lithic procurement in other

activities only minimizes costs, it does not eliminate them (Francis 1983:22-23). Transport costs also

are a function of load; that is, the energy cost of transporting a load of specific size and weight (Jones

and Madsen 1989; Schoener 1971). The ability of individuals to transport lithics would have been

constrained by the amount that could be carried by hand or in some sort of container. If lithic

procurement was embedded in other subsistence activities, then the amount of stone transported would

have been further constrained by the need to transport other subsistence items. The amount of

subsistence goods transported also would have been limited by the amount of lithic material that was

required to be transported, and this would have been dependent on technological needs, or in other

words, demand For technological systems heavily dependent on stone as a source of raw material for

tool production, the amount of stone required would necessarily be greater than for a technological

system that utilized other, substitutable materials. This combination of factors suggests that there

would have been a maximum transport distance beyond which the costs of procuring lithic raw materials

directly would have outweighed the benefits (Jones and Madsen 1989:530).

The degree to which embedded procurement was practiced also is related to the availability and

abundance oflithic raw material. If a group resides in or visits an area of abundant raw materials, then

transport costs may well be minimized by embedding procurement in other tasks: however, in an area









49







Lithice


Game













Figure 11. Graphic representation of different resource-procurement zones.


of few or no lithic raw materials, transport costs can become significant (Bamforth 1986). Furthermore,

embedded procurement of lithics is dependent on the subsistence organization of the society, and is a

viable strategy only when the distribution of lithic resources overlaps the effective range of exploitation

of at least one other resource (cf. Kuhn 1989:39). This relationship is illustrated in Figure 11 where

hypothetical collection zones for three different resources -- plants, wild game, and lithic raw materials

-- are represented by three concentric circles. Where the distribution of lithic resources overlaps the

effective range of collection for plants, game, or both, embedded procurement of stone is likely to occur

since travel costs can be averaged out between all collected resources. When lithic resources are located

beyond the collection range of other desired resources, special trips must be made to obtain only stone

and the advantages of embedded procurement diminish. Thus, the planning necessary to acquire lithic

raw materials, and the costs associated with their transport, increase as resources become less available,

that is. as distance to lithic source areas increases.









50

Risk also is a major consideration among hunter-gatherers (Smith 1988: Winterhalder 1986).

With regard to lithic resources, risk can be conceptualized as the potential for failing to obtain necessary

raw materials for tool production, while uncertainty refers to the absence of information regarding the

distribution and availability of raw materials. Bouseman (1993:64) notes that distinguishing

archaeologically between the effects of risk and uncertainty may be difficult, and he has collapsed the

two under the general term "risk."

Risk and uncertainty are present when a population expands into a new environment where the

abundance and predictability of necessary resources are unknown. Risk extends beyond the

technological sphere of economics to impact the subsistence sphere as well since failure to obtain

durable raw materials for tool production may result in a failure to efficiently exploit available biotic

resources. Under these conditions we might expect that measures will be taken to lessen procurement

risk and increase certainty.

According to Colwell (1974), predictability consists of two components: constancy and

contingency. A resource with a high degree of constancy is one that is available year-round in known

amounts and in a specific location. Furthermore, these conditions do not vary from year to year. A

resource that exhibits variation in quantity and location on a seasonal basis is said to exhibit a high

degree of contingency. Among subsistence resources, many aquatic species exhibit a high degree of

constancy while some terrestrial mammals and plants exhibit a high degree of contingency. Lithic

resources exhibit a high degree of temporal constancy since they are immobile and are not subject to

seasonal or yearly variation in abundance or location. Their availability, however, is limited to the

extent that their geographic distribution is restricted to specific geomorphic locales. Thus, they exhibit

a moderate degree of spatial contingency.

Halstead and O'Shea (1989b:3-4) suggest that cultural responses to variability in the temporal

and geographical distribution of resources can be grouped into four main categories: mobility,









51

diversification, storage, and exchange. Binford's (1980) distinction between residential mobility and

logistical mobility defines two extremes of possible hunter-gatherer settlement strategies that are highly

influenced by environmental structure and risk-abatement concerns (cf. Kelly 1983). Residential

mobiliy enables hunter-gatherers to simply move away from an area of scarcity toward an area of area

of resource abundance. Logistical mobility enables less residentially mobile populations to send special

task groups to exploit resources and return them to the residential camp. Diversification refers to a

broadening of the resource base, either by incorporating a wider range of resource types into the

inventory or by exploiting larger and more varied geographic areas. Optimal-foraging models (e.g.,

MacArthur and Pianka 1966; Orians and Pearson 1979; Stephens and Charnov 1982) likewise suggest

that when resources are abundant (i.e., when supply is high and demand is low), foragers can afford to

be selective and will choose only high-utility resources. Under these conditions, variability in the

resource mix will be minimized. Alternatively, when resources are less abundant (i.e., when supply is

low and demand is high), a generalizing strategy will be practiced and resource variability can be

expected to increase as more and more low-utility resources are included in the procurement mix. Thus,

as the potential for a lithic-resource shortfall increases, resource variability also should increase.

Stge is a means of dealing primarily with temporal variability by stockpiling a resource for

use at a future time. Echange on the other hand, is used to buffer against spatial variability. Storage

of perishable items is a high-cost strategy, particularly for mobile foragers, since labor effort must be

expended to build and maintain storage facilities, and to prepare perishable foods to be stored (Binford

1980). Both of these factors may vary depending on the type of environment. In Florida, for example,

heat and humidity would be major factors influencing the kinds of preservation techniques and storage

practices that were utilized, and these would affect the overall cost of storage. Other costs include

defending stores from seizure by others and the costs of reduced mobility which limits the ability of

foragers to practice potentially more efficient forms of energy capture. However. for non-perishable









52

materials like stone, storage may not entail as great a cost if raw materials, blanks, or finished tools are

stockpiled or cached. Thus, in terms of lithic resources, storage is a risk-abatement strategy that can be

practiced by either mobile or sedentary populations (Odell 1996:55). Exchange also has its share of

costs including those related to acquiring or producing items to be used in exchange transactions, travel

to and from the point of exchange, and the social and political costs of establishing and maintaining

exchange relations. The bulk of these costs must be invested prior to acquisition. These "pre-

acquisition" or "preproduction" costs can be offset by limiting the amount of waste material transported

and/or the amount of processing time necessary to modify the raw material into a finished implement.

An efficient way of achieving both of these cost-reduction goals is to acquire the raw material in a

finished or nearly finished form (Morrow and Jeffries 1989:30).


Procurement Strategies

In terms of lithic procurement, there are only two principal options: direct procurement at the

source or indirect procurement through exchange. Luedtke (1976:25-27) distinguishes between casual

and deliberate direct procurement. Casual procurement occurs when raw materials are obtained while

performing other activities, and is most common in areas where chert is relatively abundant and widely

dispersed. Casual procurement tends to result in a high incidence of expedient tool use with little or no

modification or maintenance of tool edges. These expectations are consistent with central-place-

foraging and opportunity-cost models that suggest that when resource abundance is high, attempts will

be made to minimize the time spent in procurement and handling of the resource. When resources are

located nearby, travel cost is low, so handling costs are more important. Thus, attempts would be made

to decrease these costs through the use of expedient processing of raw materials. Deliberate

procurement involves making a special trip to a source location in order to obtain materials for

performing a specific task or for manufacturing tools for future use. Since the distances traveled to

obtain the necessary material can be quite great, there are limitations on the amount of material that can









53

be transported. For this reason, only prepared cores and bifacial blanks or preforms would likely be

carried back to a campsite or residence. The labor cost involved in production and transport implies that

tools manufactured from deliberately procured materials should exhibit a high rate of curation. Again.

these expectations conform to those of central-place-foraging and opportunity-cost models which

suggest that net return rates can be maximized by procuring as much of a resource as possible over a

given period oftime. Processing lithic material into blanks, preforms, or finished implements increases

the amount of usable raw material that can be transported (i.e., package size) by decreasing the load

cost. These theoretical expectations are supported archaeologically by the studies of Luedtke (1976)

in Michigan and Francis (1983) in north-central Wyoming.

Another form of direct procurement is the scavenging and recycling of previously discarded

tools and production debris (Camilli and Ebert 1992; Schiffer 1987:29-30, 106-114). The degree to

which scavenging will occur is also a function of supply and demand. In areas where direct access to

resources is limited, scavenging behavior is expected to increase (cf. Ascher 1968). On the other hand,

settlement strategies or exchange relations that increase access to lithic materials should result in a

decrease in scavenging behavior. A decrease in scavenging also may be brought about by a reduced

demand for lithic materials through the use of alternative raw materials such as shell, bone, or wood.

According to Schiffer (1987:109), the artifacts most likely to be recycled would be those with the

greatest remnant use lives; for example, intact or partially intact tool forms and larger pieces of

production debris.

Indirect rocurement involves the process of exchange of which Luedtke (1976:27) defines

three types: exchange of valuables, exchange of utilitarian goods, and gift exchange. Exchange of

valuables usually occurs within a political context and is limited to non-utilitarian or prestige goods.

Exchange of utilitarian goods occurs for the purpose of obtaming materials (food or raw materials)

necessary to conduct daily activities. According to Luedtke, this type of exchange is best characterized









54

by systematic and regular trade relationships. Gift exchange occurs in the context of social relationships

and is common among hunter-gatherers (eg., Weissner 1982). While not as intensive or regular as

utilitarian exchange, gift exchange can result in wide geographic distributions of materials (Luedtke

1976:42). Furthermore, as Winterhalder (1986) has shown, cooperative alliances, which may include

sharing and gift exchange, are an effective means of reducing variability in the return rates of small

groups of hunter-gatherers. As group size increases, however, the costs of adding new members to the

group outweighs the benefits of reduced variability achieved through cooperative exchange.

As discussed above, the cost associated with exchange can be offset by importing the raw

material in a finished form. While this may effectively reduce transport and handling costs associated

with exchange, it can also result in inflated exchange values as producers attempt to offset their own

labor costs. An alternative to the consumer is to import raw material in an unfinished form. Transport

costs can be reduced by acquiring the material in small packages such as nodules, fragments, or dressed

cores. Under these conditions, core-reduction strategies should be designed to either minimize the time

expended in tool production or maximize the number of usable tools that can be obtained from the

available material (Goodyear 1993; Johnson 1987:202).


Technological Organization

Most studies of technological organization have focused on two primary strategies -- curation

and expediency. The difference between them is related to the degree of planning that each strategy

represents. Generally speaking, technologies based on cuation consist of assemblages of tools that are

relatively sophisticated, formally distinct, and possess specialized functions. Labor time is invested in

advance of the activity to be performed in order to maximize the efficient use of time during actual task

performance (Binford 1979b:269; Nelson 1991:63; Torrence 1983:12). The utility of curated tools is

maximized by transporting them between settlements, expending effort in maintenance and

resharpeing, and by recycling broken or exhausted tools (Bamforth 1986:79; Binford 1979b:263; Odell









55

1996). E dient technologies, on the other hand, consist of assemblages that are relatively simple with

less formal patterning and minimal alteration to the raw material. Two types of expedient strategies

have been suggested by Nelson (1991:64-65). Optunistic exediency is a response to a need that

was unanticipated using whatever materials are at hand. Planned e ncy is a planned response that

attempts to minimize the investment of technological effort. The latter strategy is only viable when time

and place of use are predictable (Bleed 1986; Parry and Kelly 1987). Both forms of expediency require

that there be sufficient raw materials available so that time does not have to be spent acquiring them;

however, it is especially critical for a strategy of planned expediency since the goal is to reduce the time

spent in procuring, transporting, and producing tools. This can be accomplished by locating settlements

near a lithic source, or by caching or stockpiling raw materials at a residential location.

Advance preparation of tools for future use involves high initial costs in terms of procurement

and manufacture time. If time is invested in procurement and manufacture, then maintenance will occur

in order to extend the use life of the implement and even out the initial costs of production, regardless

of material availability (Binford 1979; Ebert 1979; Odell 1996; Parry and Kelly 1987). On the other

hand, Bamforth (1986:40) argues that curation and maintenance should occur only when raw materials

are not readily available.

If access to durable raw materials is a limiting factor in the settlement of a region, then it may

be necessary to adopt differential strategies of residential mobility in order to gain access to these

materials. Several researchers have argued that different types of mobility strategies will result in

differences in tool diversity and specialization (Binford 1980; Ebert 1979; Shott 1986; Torrence 1983).

This is due primarily to the differential need for portable assemblages and activity planning. Groups

that practice a relatively mobile lifestyle (i.e., residential mobility in Binford's [1980] terms) should

possess a tool assemblage that emphasizes portability and can be used to conduct a number of different

tasks. Groups practicing a strategy of logistical mobility are assumed to move their residential base









56

camps less often, thus the need for a portable tool assemblage is less. Further, the use of specialized

task groups and storage facilities implies an increase in activity planning and a more structured strategy

for managing time (Torrence 1983). Assuming that specialized tools are more efficient and reliable for

specific tasks than generalized tools, a settlement strategy of logistical mobility should be evidenced by

an increase in the use of specialized tools and, hence, an increase in tool diversity (Shott 1986).

Settlement systems characterized by a high degree of residential mobility should be evidenced by tool

assemblages that are more generalized and useful for a variety of functions. Furthermore, Kuhn

(1989:35-36) has argued that residentially mobile foragers tend to manufacture and maintain their tools

in short periods throughout the course of the year and across all settlement locations, while logistically

mobile hunter-gatherers tend to "gear-up" for anticipated use of their tools during scheduled periods of

time at the residential location (cf Binford 1979). While the manufacture of specialized tools increases

reliability, another strategy to increase reliability and decrease the risk of failure is to replace tools before

they wear out (Kuhn 1989:36-37). Thus, the archaeological assemblages resulting from logistically

mobile hunter-gatherers should exhibit worn but still functional tools that were abandoned or cached

at residential locations, while those of residentially mobile hunter-gatherers should possess a larger

number of exhausted tools with little residual utility remaining.

Parry and Kelly (1987) offer an alternative hypothesis for the use of lithic raw materials by

groups that practice a more sedentary existence. They argue that reduced mobility results in reduced

access to lithic raw materials, and that this often is compensated for by acquiring and stockpiling lithic

materials at permanent habitation sites. Having a ready supply of stone on hand would encourage a

casual, expedient use of the material. Consequently, unsystematic core reduction and the use of

informal, expedient tools rather than curation and recycling would be expected. Expedient use of raw

materials is a relatively wasteful strategy, however, and as they point out, having access to lithic

materials, either directly or through exchange, is an important factor in determining whether an









57

expedient technology will be practiced (Pany and Kelly 1987:300-301). The practice of stockpiling also

implies that the future use of stone has been planned for to ensure that a sufficient supply of material

will be available to meet tool-using needs until more can be acquired. As McAnany (1988:9) points out,

in those technological systems where stone is acquired indirectly through exchange, the rate of

acquisition may not be under the direct control of those who are most involved with tool manufacture

and use, a situation that further encourages the use of economizing strategies. Thus, regardless of the

practice of stockpiling, some degree of material conservation would be expected in an area with limited

direct access to stone.

In light of the theoretical concepts discussed above, several hypotheses regarding lithic-resource

procurement and use through time are presented in Chapter 6. First, however, it is necessary to describe

the environmental, geological, and cultural contexts of the study area.















CHAPTER 3
DESCRIPTION OF THE STUDY AREA


The geographic area that is the primary focus of this study is known by archaeologists as the

Kissimmee region. It is defined as that part of south-central Florida that is drained by the Kissimmee

River. For the purpose ofidentifying specific geographic boundaries I have used the drainage basin map

of Florida prepared by the U.S. Geological Survey (Conover and Leach 1975). The area includes the

rivers head-waters in southwestern Orange and northern Osceola counties, and extends southward to

include eastern Polk County, all of Highlands County, most of Okeechobee County, a portion of Glades

County including Fisheating Creek, and a small part of western Martin County, ending where the river

enters Lake Okeechobee. The sites used in this study are located in the Kissimmee River valley and on

the adjacent Lake Wales Ridge (Figure 12). Although primarily a geographic definition, recent

archaeological research (Austin 1987, 1992a, 1993, 1996a; Austin et al. n.d.; Johnson 1991, 1994:

Mitchell 1996) indicates a high degree of archaeological uniformity throughout most of the region, at

least during post-Archaic times.

In addition, data from sites located in the adjacent Peace River valley, along the southwest coast,

and in the chert-bearing regions farther north have been included to provide a comparison with sites in

the chert-poor, south-central interior. However, since the primary focus of this study is on the

adaptations of prehistoric peoples in the Kissimmee region, the environment, geology, and prehistory

of this area are emphasized in this and the following chapters.







58







59










S00




'A-
N 6 c-, ;- v


NN
-- \ -- "








/) \
ER
I, ., .-,














Figure 12 Map of the study area showing the Kissimmee River, Lake Okeechobee, and surrounding
features.
V ^i / /*t6' St
v^^^ Jf^ y









features.









60

Physical Environment


The Kissimmee region can be divided into three geographic subareas -- northern, central, and

southern. The northern Kissimmee region extends from the headwaters of the river south to Lake

Kissimmee; the central area extends from Lake Kissimmee to the boundary between Highlands and

Glades Counties; the southern area extends from the Highlands-Glades county line to Lake Okeechobee.

This geographic division follows that used by the U. S. Army Corps of Engineers (USA COE) in its

recent Kissimmee River restoration environmental impact study (USA COE 1991). These subareas

encompass a wide range of natural environments which would have affected the types of prehistoric

adaptations practiced in each.


Clmate

Water is the basic agent that shapes Florida's landscape through the processes of deposition and

erosion. It also affects the type, geographic location, and extent of a region's flora and fauna. Therefore,

climate and weather are important factors in understanding the environment and geology of the study

area.

Florida is located at a latitude of approximately 30 degrees north (Winsberg 1990:Figure 1:3).

Over most of the world, continental land masses that are located between 10 and 30 degrees north

latitude are subhumid or anid Florida's climate is different because of its maritime environment and the

influence of the Gulf Stream. as well as prevailing wind patterns and the peninsula's proximity to a

subtropical high-pressure zone known as the Bermuda-Azores high (Winsberg 1990:1-21). As a result,

south Florida's climate is defined as subtropical with long, hot, humid summers and mild winters. The

prevailing winds are southerly during the spring and summer, and northerly during the fall and winter

(Carter et al. 1989:2: Jordan 1973:IIA-2). During the winter months, however, warm air from the

Caribbean and lower latitudes of the Atlantic Ocean invades south Florida and prevents cold air from









61

the continent from penetrating the southern half of the peninsula as often as it does in north Florida

(Winsberg 1990:29)

Wet and dry seasons are well defined Approximately 55 percent of the rainfall occurs between

the months of June and September, and only about 10 percent occurs between December and February

(Florida Department ofNatural Resources [FDNRI 1974:5; Parker et al. 1955:Table 57; VanArman et

al. 1984:144). The rainy season begins and ends abruptly, with twice as much rain falling in the first

and last months of the season than in the preceding and succeeding months, respectively (FDNR

1974:78). Summer rains are primarily the result of convectional thunderstorms that occur as a result

of the heating of the land surface and increased evaporation. During the fall months, as the land surface

cools and convective energy dissipates, most rainfall occurs as a result of tropical storms. During the

winter months, rainfall occurs as a result of the southward invasion of cold, polar air masses which force

the warm, moist air over Florida to rise resulting, when cooled, in precipitation.

Interestingly, the Kissimmee River-Lake Okeechobee watershed experiences significantly less

annual rainfall than any other area of the state except the Florida Keys, averaging only 91 cm (36 in)

per year (VanArman et al. 1984:145). This may be due to reduced convection resulting from slightly

cooler surface temperatures over Lake Okeechobee and up the river valley (Brooks 1974:258; Gleason

et al. 1974:309; cf. Thomas 1974:88, Figure 17). Despite the low, annual average precipitation, rainfall

records beginning in 1871 indicate numerous incidents of flooding and extended inundation of the

basin, particularly when hurricanes followed a wet summer (USA COE 1991:A-6). This is due to two

factors: 1) the large storage capacity of the lakes in the upper basin which overflow during periods of

heavy rainfall and discharge into the lakes, swamps, and streams of the lower basin, and 2) the

accumulation of ground water in the adjacent Central Highlands and Lake Wales Ridge which eventu-

ally discharges into the Kissimmee River and some of the larger lakes which are at a lower elevation

(Kohout and Meyer 1959:17-25; Parker et al. 1955:138, 301-305)









62

Geomorpholoa

The study area encompasses several major geomorphic features (Figure 13) as defined by White

(1970). The northern Kissimmee region is situated at the interface between the Central Highlands (more

specifically the Orlando. Mt. Dora. and Lake Wales ridges) and the Osceola Plain. The central Kissim-

mee region consists of two major features -- the Osceola Plain, which contains the valley and floodplain

of the Kissimmee River, and a southward extension of the Central Highlands known as the Lake Wales

Ridge. The southern Kissimmee region is contained almost entirely within the low-lying Okeechobee

Plain. The narrow Caloosahatchee Incline borders the Lake Wales Ridge on its eastern and southern

margins. To the west of the study area is the Polk Uplands and DeSoto Plain, geomorphic features that

contain the Peace River valley.

The genesis of these features is largely the result of marine transgressions and regressions

during the Pliocene and Pleistocene. Their modem appearance has been affected by subsequent erosion,

aeolian deposition, and solution of the underling limestone. The five principal marine terraces that have

been identified in the study area are listed in Table 1. Cooke (1945) and Healy (1975) identified two

additional terraces: the Penholoway at an elevation of 21 m (70 ft) and the Talbot at about 12 m (40 ft).

However, not all geologists recognize these features as true terraces, considering them to be standstills

during either Wicomico or Pamlico times. In fact, Healy (1975) notes that the Talbot is not well

represented in south Florida. They are included in Table 1 because reference to both the Penholoway

and Talbot terraces do appear occasionally in the geological literature of the region (e.g., Kohout and

Meyer 1959:41; Parker et al. 1955:137-140).

The Silver Bluff sea stand is based on a shoreline terrace in southeast Florida that was initially

interpreted as being the result of a late Holocene transgression and subsequently was extended to other

parts of coastal Florida and Georgia (MacNeil 1950); however, recent radiocarbon dates indicate that

the Silver Bluff terrace is Pleistocene in age (Cronin et al 1981) While the Silver Bluff may not be









Lake
Kissimmee



POLK o
UPLAND ;







r STUDY- \ PLAIN
\ < VLake

STUDY OKEECHOBEE L1
o AREA



Approx. Sea% TCHEE Il

o0 10 15 20miN.. 0 4 S EE

IMMOKALEE EVERGLADES
0 10 20 30 km RISE EVER



Figure 13. Geomorphic features in the study area (after White 1970).









64

Table 1. Marine terraces in Florida as identified by various geologists.

Terrace Elevation Age

After MacNeil (1950)

Okefenokee 46 m (150 ft) Yarmouth
Wicomico 30.5m (100 ft) Sangamon
Pamlico 8-11 m (25-35 ft) Wisconsin
Silver Bluff 2.4-3 m (8-10 ft) Recent
After Puri and Vernon (1964)

Coharie 67 m (220 ft) Aftonian Interglacial
Okefenokee 46 m (150 ft) Yarmouth Interglacial
Wicomico 30.5 m (100 ft) Sangamon Interglacial
Pamlico 9 m (30 ft) Peorian Interglacial
Silver Bluff 2.4 m (8 ft) Late Wisconsin
Interstadial
After Alt and Brooks (1965)

65.5-76 m (215-250 ft) Upper Miocene
27-30.5 m (90-100 ft) Pliocene

Insignificant Stand 21-24 m (70-80 ft) Pliocene or Pleistocene
Insignificant Stand 14-17 m (45-55 ft) Pliocene or Pleistocene
8-11 m (25-35 ft) Pleistocene

After Healv 1975
Hazelhurst 65.5-97.5 m (215-320 ft) Miocene or Pliocene
Coharie 52-65.5 m (170-215 ft) Pleistocene
Sunniland and 30.5-52 m (100-170 ft) Pleistocene
Okefenokee
Wicomico 21-30.5 m (70-100 ft) Pleistocene
Penholoway 13-21 m (42-70 ft) Pleistocene
Talbot 8-13 m (25-42 ft) Pleistocene
Pamlico 2.5-8 m (8-25 ft) Pleistocene
Silver Bluff <.3-3 m (<1-10 ft) Recent









65
Holocene in age, evidence for higher-than-present sea levels during the Holocene may be present in

beach ridges on off-shore islands along Florida's west coast. Several researchers have argued that these

represent a series of higher sea levels ranging from less than I m up to 3 m above present levels since

about 5000 B.P. (e.g., Missimer 1973; Stapor and Tanner 1977; Stapor et al. 1991; Tanner 1991;

Walker 1992; Walker et al. 1994).

CentralHighlands. The Central Highlands comprises a series of localized high areas (ridges)

separated by low-lying valleys. Both the ridges and the valleys tend to be linear in form and oriented

roughly parallel to modem coastlines (White 1970:111-112). White (1970:113) suggests that the

differences in elevation between ridges and valleys are the result of differential subsidence of the

underlying limestone. He also suggests that this has occurred since Wicomico times, reasoning that if

the Wicomico sea had inundated the highlands, ridge features that are at elevations lower than the

Wicomico sea (i.e., less than 30 m) would have been destroyed (White 1970:113).

The Lake Wales Ridge represents a southward extension of the Central Highlands. It is a

narrow finger of land that begins near the common comer of Polk, Osceola, Orange, and Lake counties

and extends south for a distance of about 137 km (85 mi) ending a few miles north of Venus in

Highlands County. The maximum elevation is approximately 91 m (300 ft) AMSL near the town of

Lake Wales. The linear structure of the ridge appears to be the result of the deposition of coarse

siliciclastics along a prograding delta that subsequently was straightened along its flanks by coastal

erosion during Okefenokee times (White 1970:112-113. 120-123). A major feature is a chain of lakes

that extends down the center of the ridge in what is known as the Intraridge Valley. This narrow,

steep-walled lowland runs from the north shore of Lake Livingston near Frostproof south to Lake

Placid Most of these lakes were formed by solution of the underlying limestone which resulted in the

formation of sinkholes (Bishop 1956:32; Cooke 1939:102; White 1970:120). According to White









66

(1970:119), the valley itself may be the result of reduction by solution of the underlying limestone along

a structural lineament or fault.

The east side of the Lake Wales Ridge is a rather abrupt scarp that may have been the western

shore ofa large embayment that once occupied the Kissimmee River valley (White 1958:42-43). The

east side of the Ridge is also characterized by a series of undulating dunes that are oriented at right

angles to one another. The topography is very irregular with numerous hills separated by swales and

depressions, some of which hold water. The most likely explanation for this topography is aeolian

reworking of surface sands which formed ridges and blowouts (White 1958:39-41).

Osceola Plain. The Osceola Plain is a flat, marine terrace that is believed to have formed when

sea level was 12 m or more above present levels (Lane et al. 1980). It attains a maximum elevation of

about 24 m (80 ft) at its northern end where it rises towards the higher Orlando Ridge. The terrace

slopes gradually to the south with an elevation of about 15 m (50 ft) along the shoreline of the next

lower terrace (Lane et al, 1980). Bombing Range Ridge provides the only conspicuous topographic

relief with an elevation of 38-44 m (125-145 ft). It forms the divide between the Kissimmee River to

the east and Arbuckle Creek to the west. White (1970:138) considers Bombing Range Ridge to be a

relict marine sand bar that was created during the preceding Okefenokee high stand.

Okeechbee Plain. The Okeechobee Plain is a broad, flat terrace that contains the southern part

of the Kissimmee River, all of Lake Okeechobee, and a large part of the Everglades. It formed during

the advance of the Pamlico sea during the late Pleistocene (Brooks 1974). At the toe of the scarp that

forms the boundary between the Okeechobee and Osceola Plains, the elevation ranges between 9 and

12 m (30-40 ft). The terrain slopes very gradually to the south reaching elevations of only 6 m (20 ft)

above mean sea level at the north end of Lake Okeechobee (White 1970:142). Cooke (1939:101-102)

believed that the Lake Okeechobee basin originated as a slight depression in the Pamlico sea. He felt

that Lake Istokpoga and Lake Kissimmee, as well as other lakes in the Kissimmee basin, all originated









67

as similar types of depressions. However, if he is correct, then they must be older than Lake Okee-

chobee because they lie on higher and older marine terraces.

Polk Upland. The Polk Upland is a mature terrace that has been dissected by several well-

defined drainage systems. The Pleistocene sediments that cover it were deposited during the

Okefenokee marine transgression (Clark 1972:29; Healy 1975). According to White (1970:133), the

southern scarp of the Polk Upland is the result of a marine standstill during Wicomico times which

eroded the earlier Okefenokee terrace.

DeSoto Plain. The DeSoto Plain is a submarine terrace that developed during a standstill in the

Wicomico transgression (White 1970:141). It is extremely flat with an elevation difference of only

about 4.5-7.6 m (15-25 ft) from north to south (White 1970:140). The Peace River is fairly deeply

entrenched into this feature, particularly at its northern end. The DeSoto Plain also contains the

headwaters of Fisheating Creek.

Caloosahatchee Incline The Caloosahatchee Incline is a long, narrow, sloping surface that

borders the east flank of the Lake Wales Ridge and the southern edge of the DeSoto Plain. It is regarded

by White (1970:141) to have been deposited at the down-current end of a submarine shoal. However,

Clark (1972:31) believes that this feature represents an erosional marine shoreline that developed during

a standstill at about 18 m (60 ft), although he is not certain if this occurred during a temporary pause

in the regression of the Wicomico sea, or during a similar pause in the advance of the succeeding

Pamlico sea.


Environment

Environmentally, it is possible to divide the study area into three major zones: the higher, more

rolling Central Lakes District separates the low lying Eastern and Western Flatlands (Brooks 1982; cf.

Davis 1973).









68

Central Lakes District. The Central Lakes District includes the Lake Wales Ridge. It is

characterized by rolling hills of well-drained sand interspersed with lakes and ponds. The prevailing

vegetation is sand pine, longleaf pine, turkey oak, and scrub live oak (Abrahamson et al. 1984: Carter

et al. 1989; Davis 1943:156-160: Ford et al. 1990; Harper 1921:124, 1927:82). Pine flatwoods or

scrubby flatwoods are present in areas oflower elevation and poorer drainage (Abrahamson et al. 1984;

Carter et al. 1989; Ford et al. 1990; Harper 1921:124). Hardwood hammocks are sometimes found

along creeks and streams or along lakeshores.

Lakes are numerous and of varying types. Some are shallow basins that hold water seasonally

while others are larger, deeper, and contain water year-round. On the Lake Wales Ridge, major lakes

include Reedy Lake, Lake Livingston, Lake Jackson, Lake Josephine, Lake June-in-Winter, and Lake

Placid. The elevation of the potentiometric surface of the Floridan Aquifer in this area averages 18-27

m (60-90 ft) above sea level with the gradient to the south (Tibbals 1990:Figure 21; Yobbi 1983). Most

of these lakes lie at elevations that are above the potentiometric surface; thus there is no upward leakage

of water into the lakes from the artesian aquifer and the water levels are primarily rainfall dependent

(Adams and Stoker 1985; Belles and Martin 1984; Hammett 1981; Stewart 1966:72-76). Downward

leakage does occur, however, and the ridge serves as a major recharge area to both the artesian

(Floridan) and non-artesian aquifers (Tibbals 1990).

Freshwater seeps are common along the flanks of the ridge as a result of the lateral movement

of ground water from areas of higher to lower elevation (Bishop 1956:47-48; Carter et al. 1989:12).

These seeps emerge to contribute to the large swamp systems that are present on either side of the ridge

-- Lake Istokpoga and its associated outflow to the east, and Little Charley Bowlegs and Fisheating

Creeks to the west. The seeps are characterized by the presence of distinctive "cutthroat grass"

vegetation, and so are known locally as "cutthroat seeps" (Carter et al. 1989:12; Harper 1921:208.

1927:160-161)









69

Development is encroaching on the area today. The northern Kissinmee region lies to the south

of the rapidly expanding Orlando/Disney World urban, residential, and entertainment complex. Farther

south, near Sebring, the many large lakes attract tourists and seasonal residents who enjoy boating and

fishing. Those areas that remain undeveloped contain citrus groves.

Easten Flalands. The large expanse of relatively flat terrain and somewhat-poorly to poorly

drained soils located to the east of the Lake Wales Ridge is known as the Eastern Flatlands. It

encompasses most of the northern and central Kissaimee regions and all of the southern region. Harper

(1921:136, 137-138) described the area as "... for the most part monotonously level...Near the lake

region the topography is often a little undulating, and the transition from one region to the other is

gradual, though there are also places where it is abrupt. Shallow depressions abound, ranging in size

from lakes covering several square miles (most of these near the lake region) to small wet prairies and

cypress ponds. Streams are few and sluggish, and the rivers have extremely shallow valleys."

The principal lakes in the northern portion of this region include Lake Tohopekaliga, East Lake

Tohopekaliga, Alligator Lake, Lake Hatchinea, and Cypress Lake. These are part of the chain of lakes

that form the headwaters of the Kissimmee River (Lichtler 1972:8; Parker et al. 1955:302; USA COE

1991:A-4). Reedy Creek is a major waterway that drains the Central Highlands and provides outflow

to this lake system. Major surface-water features in the central Kissinmee region include Lake

Kissimmee, Lake Weohyakapka, Lake Arbuckle, Lake Istokpoga, the Kissimmee River, and Arbuckle

Creek.

Vegetation consists primarily of pine and saw palmetto flatwoods interspersed with grassy

ponds and cypress heads. Small knolls and ridges of moderately well-drained soil support pine and

scrub oak vegetation. The largest and most prominent of these scrubby ridges is Bombing Range Ridge,

described above. Hardwood hammocks are common along the river and major streams, and around

some of the larger lakes.









70

A unique ecological feature of the area is the treeless, palmetto prairie which is especially

common along the Kissimmee River south of Lake Kissimmee (Harper 1921:203-204, 1927:87-88,

163). Davis (1943:28) relates that the origin of these dry prairies is believed by some to have been

caused by numerous fires set by settlers and earlier by Indians. Davis doubts this because the earliest

accounts of the region refer to these prairies as natural communities extending over the same amount

of area as they presently occupy. Instead, he suggests that the absence of trees is due to the fluctuating

water table which favors vegetation, such as grasses and low bushes, with shallow root systems (Davis

1943:128).

Two major water features dominate the central portion of the Eastern Flatlands: the Kissimmee

River and Lake Istokpoga. In its natural state, the Kissimmee was originally a slow-moving, meandering

river with a floodplain that varied from 1.5-3 km in width (FDNR 1974:89: USA COE 1991:7). Much

of this floodplain consisted of pickerelweed flag marsh, maidencane flag marsh, and beakrush marsh

(Kushlan 1990:344-345; USA COE 1991:9, Table 1). Runoff from the surrounding poorly drained flat-

lands, and the rise of lake levels in the upper basin at the end of the summer rainy season, caused the

river regularly to overflow its banks. Indeed, early accounts of the region's natural environment are

nearly unanimous in commenting on the overwhelming presence of water throughout the valley (e.g.,

Davis 1943:34; Hancock 1979:19; McCaffrey et al. 1976; Parker et al. 1955:301-302: Will 1977:9-10).

In some years the flooding was so severe that the river resembled a wide lake, and low-lying areas were

often under water for several months at a time (USA COE 1991:15. A-6). As a result, the river has

experienced numerous attempts to control it (cf McCaffrey et al. 1976; Tebeau 1974; VanArman et al.

1984:138-139). The most recent, in the 1960s, was a massive flood-control project by the U.S. Army

Corps of Engineers which involved channelizing the river and constructing a series of levees and water-

control structures along its course. While successful in controlling flood waters, the negative effect of

this project on the natural ecosystem has been dramatic. Seasonally fluctuating water levels in









71

conjunction with undulating floodplain topography, a meandering river channel, oxbows. and natural,

discontinuous levees, were essential components for maintaining the river's diverse mosaic of natural

wetland habitats. By eliminating seasonal inundation, channelization has caused many marshes to dry

up allowing terrestrial vegetation to invade (VanArman et al. 1984:154). Of an estimated 14,000

hectares of wetlands that once existed in the floodplain, only about 5.700 remain today (USA COE

1991:26). Most of the drained land is used for cattle ranching.

Lake Istokpoga is the fifth largest lake in Florida. It is fed by Arbuckle Creek, which enters the

lake from the north. The lake occupies a natural depression that intersects the water table so ground

water moving down-gradient from the higher Lake Wales Ridge, located to the west, provides base flow

to the lake (Kohout and Meyer 1959:15, 17-25). At one time a large, wet savannah existed to the south

of the lake. Known as Indian Prairie, it channeled sheet flow from the lake south towards Lake Okee-

chobee (Kohout and Meyer 1959:13; Parker et al. 1955:311). According to early accounts, this broad

expanse of herbaceous grasses and cabbage palm hammocks rivaled the Everglades in its beauty and

abundance of wildlife (Davis 1943:49; Harper 1927:100-103, 166; Small 1921:57). Harper (1927:166),

for example, described Indian Prairie as "a region of beautiful palm savannahs, covering apparently a

few hundred square miles...the abundance of palmettoes and tall grass are a wonderful sight, hardly

matched anywhere else in the United States." Unfortunately, flood-control measures and drainage

canals have essentially destroyed this wetland. Today most of the area formerly occupied by Indian

Prairie is used for agriculture. It also contains significant deposits of peat (Davis 1946:129-132).

The southern portion of the Eastern Flatlands extends into the marsh and savannah

environment that characterizes the Lake Okeechobee basin. Poorly drained pine flatwoods are present

in the upland portions of this subarea with hardwood hammocks along rivers and creeks, and on natural

levees around the lake margin (Carter et al. 1989; VanArman 1984:140-141). Historic drainage

practices and the construction of massive dikes and levees have altered drastically the size and extent