The zooarchaeology of Charlotte Harbor's prehistoric maritime adaptation


Material Information

The zooarchaeology of Charlotte Harbor's prehistoric maritime adaptation
Physical Description:
Walker, Karen Jo
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 29668226
oclc - 29176651
System ID:

Full Text







Copyright 1992


Karen Jo Walker


Much appreciation is extended to William Marquardt,

Elizabeth Wing, and Stephen Hale for the opportunity to

become involved in the southwest Florida research. The

collection and analysis of the Cash Mound, Useppa Island,

Josslyn Island, and Buck Key samples were funded by the

National Science Foundation. The Big Mound Key samples were

collected and provided by George Luer. Elizabeth Wing

provided much guidance and use of her lab, curation space,

comparative collection, and computers. Thanks go to Nina

Borremans, Laura Kozuch, and Guy Prentice for the initial

analysis of samples A-2-4 from Useppa, B-2-9 from Buck Key,

and A-l-4 from Josslyn, respectively. Laura Kozuch and

Cherry Fitzgerald also assisted with other samples.

I am indebted to reviewers of an early draft of this

dissertation. They include Kurt Auffenburg, Robert Edic,

Barbara Hoffman, Robert Knight, Elise LeCompte-Baer, George

Luer, William Marquardt, Jerald Milanich, Claudine Payne,

Irv Quitmyer, Elizabeth Reitz, Randal Walker, and Randolph

Widmer. The dissertation has further benefitted from

discussions or correspondence with Nina Borremans, Joel

Gunn, William Marquardt, Irv Quitmyer, Donna Ruhl, Frank

Stapor, and William Tanner.


Corbett Torrence drafted Figure 1 (also used in Figures

7 and 16). Merald Clark produced the final versions of

Figures 2, 3, 6, 8, 9, and 10. He also illustrated the

fishing net in Figure 19 and put together the pie chart

figures, 7 and 16. Jim Wagner drafted Figures 4, 5, 11, 12,

14, and 15. In addition, he illustrated the fish in Figures

17 and 18. Scott Swan produced Figure 13. Claudine Payne,

Irv Quitmyer, Becky Saunders, and Sam Chapman contributed

computer expertise in the final production (i.e.,

translation of software programs) of Appendices A and B.

It is with great gratitude that I acknowledge the

members of my supervisory committee. Michael Moseley,

Elizabeth Wing, Jerald Milanich, Clay Montague, and Rhodes

Fairbridge have given me much guidance during the

dissertation process. I further thank Professor Fairbridge

of Columbia University and NASA-Goddard Institute for Space

Studies for going out of his way to attend my final

examination and afterwards to visit the Pineland Site in Lee

County. Carole Mclvor of the Department of Forest Resources

and Conservation generously came to my rescue when it became

clear that Clay Montague could not attend the final

examination. In addition to my committee, I have had the

good fortune to benefit from the advice, perspective, and

encouragement of William Marquardt, director of the

Southwest Florida Project.



ACKNOWLEDGMENTS........................................ iii

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

LIST OF FIGURES............................. ............ .. .. xi

ABSTRACT..... ................................ ...... ...... . .xiii


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

The Maritime Calusa of Charlotte Harbor..............1
The Lagging Maritime Perspective.................... 5
Research Goal and Objectives........................ 8
Sample Context and Excavation...................... 12
Big Mound Key ..................................... 14
Cash Mound............... ........... .......... ... 15
Useppa Island........................... ......... 16
Josslyn Island................................... 16
Buck Key Shell Midden............................ 18
Zooarchaeological Methods.......................... 19
Sample Processing................................ 19
Identification and Quantification................. 21
Sample Size....................... ............... 23
Food vs. Commensals............................... 25
Sources of Bias.................................... 25
Comparative Dietary Contribution.................. 28


The Present-day Charlotte Harbor Estuarine
Ecosystem............ ........................ . 53
The Present-day Estuarine Gradient.................. 57
Distribution of Vertebrates........................ 60
Distribution of Invertebrates...................... 64
Inferred Local Distribution of Resources in
Prehistory................................. ...... 68
Big Mound Key................................... 70
Cash Mound............._ ..... ... .. ...................... 71
Useppa Island............... ... ................. .72
Josslyn Island...................................73

Buck Key Shell Midden.......................... 74
Inferred Regional Distribution of Resources in
Prehistory ....................................... 75


Environmental Continuity and Change................ 88
Potential Short-term Environmental Change..........90
Freezes, Red Tides, and Storms ................... 90
Seasonal Variability............................. 92
Potential Medium-term Environmental Change.........94
Climatic and Sea Level Variability............... 94
Inlet Dynamics.................................... 95
Potential Long-term Environmental Change.......... 96
Climatic Variability .............................96
Sea Level Variability ............................ 98
Inlet Dynamics .................................. 104
Estuarine-Marine Zooarchaeological Fauna as Proxy
Data.................................... ........ 105
Interpretive Potential and Time Resolution......105
Temporal Zooarchaeological Assemblages.......... 113
Effective Scale and Zooarchaeological
Potential.................................... 134


Zooarchaeological Patterns at the Local Scale.....152
Big Mound Key................................... 152
Cash Mound................................... 154
Useppa Island............................ 156
Josslyn Island .................................. 157
Buck Key Shell Midden....................... 158
Exploitation Patterns at the Regional Scale.......161
An Aquatic Exploitation .........................161
Fishing, Gathering, and Hunting Technology......165
Hypotheses for Variation in Fishing Artifacts...177

5 SUMMARY AND CONCLUSIONS ........................... 180


A ZOOARCHAEOLOGICAL DATA TABLES ..................... 205

Big Mound Key...................................... 206
Cash Mound.........................................217
Useppa Island...................................... 225
Josslyn Island..................................... 228
Buck Key Shell Midden.............................. 240


REFERENCES .............................................. 256

BIOGRAPHICAL SKETCH.................................... 276



table page

1 Generalized Cultural Chronology for the
Caloosahatchee Area.................................. 31

2 Zooarchaeological Samples Included in the
Charlotte Harbor Study............................. 33

3 Summary of Zooarchaeological Data Included in
the Charlotte Harbor Study .......................... 35

4 Regression Values for Minimum Meat Weight
Estimations ......................................... 37

5 Regression Values for Maximum Meat Weight
Estimations........................................... 39

6 Non-regression Values for Maximum Meat
Weight Estimations.................................. 41

7 Oscillating Holocene Sea Level Curves Based on
Beach Ridge Data for the Charlotte Harbor and Gulf
of Mexico Regions................................... 138

8 Relative MNI Percentages of Eastern Oyster
(EO), Crested Oyster (CO), Crown Conch (CC), and
Ribbed Mussel (RM) for Cash Mound Samples A-l-4,
A-l-8, A-l-17, and A-l-20......................... 139

9 Intersite Comparison of Hardhead Catfish
Totals............................................. 189

10 Comparison of Terrestrial and Aquatic Animal
Food Resources by Percentage ....................... 190

11 Ranking of Bony Fishes by Maximum Meat
Weight ............................................. 191

12 Archaeological Remains of Sharks by Minimum
Number of Individuals (MNI) ........................ 192

13 Distribution of Archaeological Pinfish and
Associates by Minimum Number of Individuals (MNI)..193


14 Mesh Sizes of Key Marco Net Cordage................ 194

15 Distribution of Archaeological Mullet
(Mugil spp.) ...................................... 195

16 Archaeological Terrestrial Fauna by Minimum
Number of Individuals (MNI) ....................... 196

A-i Faunal Analysis, Big Mound Key, 8CH10,
Charlotte County, Florida, May 1982 Sample, U.2/S.3,
NW Quad. Layer 11................................. 206

A-2 Faunal Analysis, Big Mound Key, 8CH10,
Charlotte County, Florida, August 1982 Sample,
U.1/S.4, Layer 8b................................. 209

A-3 Faunal Analysis, Big Mound Key, 8CH10,
Charlotte County, Florida, August 1982 Sample,
U.1/S.4, Layer 7....................................212

A-4 Faunal Analysis, Big Mound Key, 8CH10,
Charlotte County, Florida, November 1982 Sample,
U.1/S.4, NW Quad. Layer 2.......................... 215

A-5 Faunal Analysis, Cash Mound, 8CH38,
Charlotte County, Florida, June 1985 Sample, Test
A-l, Level 4 ....................................... 217

A-6 Faunal Analysis, Cash Mound, 8CH38,
Charlotte County, Florida, June 1985 Sample, Test
A-l, Level 8...................................... 219

A-7 Faunal Analysis, Cash Mound, 8CH38,
Charlotte County, Florida, June 1985 Sample, Test
A-l, Level 17 ...................................... 221

A-8 Faunal Analysis, Cash Mound, 8CH38,
Charlotte County, Florida, June 1985 Sample, Test
A-l, Level 20 ......................................223

A-9 Faunal Analysis, Useppa Island, 8LL51, Lee
County, Florida, Aug./Sept. 1985 Sample, Test A-4,
Level 2............................................ 225

A-10 Faunal Analysis, Josslyn Island, 8LL32,
Lee County, Florida, March 1985 Sample, Test A-l,
Level 4 (1/2 volume sample)........................ 228

A-11 Faunal Analysis, Josslyn Island, 8LL32,
Lee County, Florida, March 1985 Sample, Test A-l,
Level 12 ........................... ......... ..... .. 231

A-12 Faunal Analysis, Josslyn Island, 8LL32,
Lee County, Florida, March 1985 Sample, Test A-l,
Level 22........................................... 234

A-13 Faunal Analysis, Josslyn Island, 8LL32,
Lee County, Florida, March 1985 Sample, Test A-I,
Level 32....... ......... ......................... 237

A-14 Faunal Analysis, Buck Key Shell Midden,
8LL722, Lee County, Florida, March 1986 Sample, Test
B-2, Level 5...................................... 240

A-15 Faunal Analysis, Buck Key Shell Midden,
8LL722, Lee County, Florida, March 1986 Sample, Test
B-2, Level 9................... .... ... ........... 243

A-16 Faunal Analysis, Buck Key Shell Midden,
8LL722, Lee County, Florida, March 1986 Sample, Test
A-2, Level 6/7............ ........................ 246

A-17 Faunal Analysis, Buck Key Shell Midden,
8LL722, Lee County, Florida, March 1986 Sample, Test
A-2, Level 11 ............................. ......... 249

B-1 Aquatic Vertebrates by Archaeological Site
and Modern Habitat................................. ...252

B-2 Aquatic Invertebrates by Archaeological
Site and Modern Habitat............................. 254


figure page

1 Map of the Charlotte Harbor Study Area with
Geographical Features and Archaeological Site
Locations Mentioned in the Text: (1) Solana Site;
(2) Big Mound Key; (3) Cash Mound; (4) Useppa
Island; (5) Pineland Site; (6) Josslyn Island; (7)
Buck Key Shell Midden; and (8) Wightman Site........44

2 The Distribution of Charlotte Harbor
Zooarchaeological Vertebrate Samples by Number of
Taxa and Minimum Number of Individuals (MNI).
Numbers Refer to the List Given in Table 2..........46

3 The Distribution of Charlotte Harbor
Zooarchaeological Invertebrate Samples by Number of
Taxa and Minimum Number of Individuals (MNI).
Numbers Refer to the List Given in Table 2..........48

4 Comparative Percentages of Zooarchaeological
Estimated Minimum Edible Meat Weights by Site and
Animal Group (Based on Data Presented in Appendix
A) .................................................. 50

5 Comparative Percentages of Zooarchaeological
Estimated Maximum Edible Meat Weights by Site and
Animal Group (Based on Data Presented in Appendix
A) ................................ .... .......... 52

6 Monthly Salinity Profiles of Four Aquatic
Locations in the Northern Part of the Charlotte
Harbor Estuarine Complex Illustrating the Fresh to
Salt Water Gradient (Data are after Wang and Raney
1971:18) ........................................... 81

7 Comparative Percentages of Zooarchaeological
MNI by Site Representing Exploited Habitats (Based
On Data Presented in Appendix A) .................... 83

8 Thoracic Vertebrae Widths of Bony Fishes as
an Indicator of Overall Fish Size for Cash Mound,
Josslyn Island, and Buck Key ........................ 85

9 A Schematic Illustration of the Relationship
between Aquatic Vertebrates and Invertebrates
Recovered from the Five Study Sites and the
Estuarine Gradient (Based on the Detailed Data
Presented in Appendix B)........................... 87

10 Mean Sea-Level Curve for Southwest Florida
Proposed by Stapor et al. Based on Geochronology,
Geomorphology, and the Elevation of Beach Ridge
Sets Making Up the Barrier Islands (After Stapor
et al. 1991:Figure 14) ............................. 141

11 Comparative Percentages of Zooarchaeological
Food MNI, Minimum Meat Weight, and Maximum Meat
Weight by Provenience for the Josslyn Island Faunal
Samples (Based on Data Presented in Appendix A) .... 143

12 Comparative Percentages of Zooarchaeological
Food MNI, Minimum Meat Weight, and Maximum Meat
Weight by Provenience for the Cash Mound Faunal
Samples (Based on Data Presented in Appendix A) .... 145

13 The Variation Based on Percentage of MNI of
Selected Species -- Eastern Oyster, Crested Oyster,
Ribbed Mussel, and Crown Conch -- from Cash Mound
Samples A-l-20, A-l-17, and A-l-8 Dating to A.D.
150 to A.D. 270 and A-l-4 Dating to A.D. 680.......147

14 Comparative Percentages of Zooarchaeological
Food MNI, Minimum Meat Weight, and Maximum Meat
Weight by Provenience for the Big Mound Key Faunal
Samples (Based on Data Presented in Appendix A) .... 149

15 Comparative Percentages of Zooarchaeological
Food MNI, Minimum Meat Weight, and Maximum Meat
Weight by Provenience for the Buck Key Faunal
Samples (Based on Data Presented in Appendix A) .... 151

16 Intersite Variability of Approximated Subsistence
Activity Based on MNI of Exploited Animals......... 198

17 Adult Pinfish, Lagodon rhomboides, and its
Atlas and Premaxilla Bones........................ 200

18 Adult Pigfish, Orthopristis chrysoptera, and
its Atlas and Premaxilla Bones.................... 202

19 Artist's Conception of a Prehistoric Gill Net for
Nearshore Shallow-Water Fishing in the Charlotte
Harbor Area Based on Archaeological Net Remains
from the Key Marco Site............................ 204


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



Karen Jo Walker

December, 1992

Chairman: Michael E. Moseley
Major Department: Anthropology

Much discussion involving prehistoric, complex,

maritime fisher-gatherer-hunter societies centers on whether

or not the natural environments inhabited by these peoples

were productive and stable enough to account for the

cultural complexity. In the case of southwest Florida's

maritime Calusa and their predecessors, addressing this

issue first requires a multi-scalar understanding of spatial

and temporal environmental context. This is because, like

any environment, coastal southwest Florida (specifically the

Charlotte Harbor estuarine system) is characterized by

habitat heterogeneity in space and geophysical dynamism


through time. This dissertation establishes the needed

contextual framework essential for properly addressing the

broader question of productivity, stability, and complexity.

Zooarchaeological remains provide an important proxy

data set for the purpose of modeling Charlotte Harbor's

spatial and temporal estuarine paleoenvironments at multiple

scales. It is argued that estuarine archaeofauna can serve

as paleoecological data, albeit with limitations. The

analysis of fine-screened bulk samples from five sites

variously located within the Charlotte Harbor estuarine

system forms the data base of the modeling exercise. These

midden materials date from 600 B.C. to A.D. 1400, spanning

the Caloosahatchee I through III archaeological periods.

Central to the model is an estuarine gradient analysis

using salinity as the primary organizing variable for

understanding the distribution of living fauna. The spatial

focus is on both local (site) and regional (Charlotte

Harbor) scales. From a temporal perspective, short-,

medium-, and long-term forms of environmental variation are

defined in terms of potential alteration of the estuarine

gradient. Proposed zooarchaeological signatures of such

multi-scalar alteration that were explored among the

Charlotte Harbor data lead to the conclusion that medium-

and long-term sea-level fluctuations and inlet dynamics are

most likely to have affected human subsistence. For the

Charlotte Harbor samples presented here, sea-level


fluctuations of .9 to 1.8 m above (50 B.C. to A.D. 450) and

below (A.D. 550 to A.D. 850) present sea level produce

signatures of an altered estuarine gradient but more

supportive evidence is necessary to resolve the temporal


The integration of spatial and temporal perspectives at

both local and regional scales demonstrates the potential of

zooarchaeological inference and advances the hypothesis that

exploitation technology reflects the modeled environmental




The Maritime Calusa of Charlotte Harbor

The center of the world for much of the Calusa

population in the sixteenth century was southwest Florida's

highly productive Charlotte Harbor estuarine system (Figure

1). It was reported in 1564 "that the [Calusa] king was

held in great reverence by his subjects and that he made

them believe that his sorceries and spells were the reason

why the earth brought forth her fruit" (Laudonnibre

1975:110). The quote implies that the continued

productivity and stability of the natural world (i.e.,

Charlotte Harbor) were integral to the maintenance of the

Calusa paramount chief's authority.

Environmental productivity and stability may have been

particularly crucial factors for the culturally complex

Calusa because they apparently did not rely on agricultural

products (Goggin and Sturtevant 1964:183-184; Marquardt

1986:63, 1987:100, 1988:162-169; Milanich and Fairbanks

1980:243-244; Widmer 1988:224-250). Instead, as suggested

by various Spanish reports and the existence of enormous

shell middens, estuarine/marine foods appear to have been


the primary subsistence focus of these sedentary coastal


The term "maritime" is used in this dissertation to

describe a situation adjacent to the sea (i.e., marine

waters). Although Charlotte Harbor is technically an

estuarine environment rather than one of strictly marine

waters (i.e., 35 ppt salinity), the estuarine adaptation in

prehistory is viewed here as a specialized type of the

broader maritime cultural pattern (see Yesner 1980:728).

Furthermore, much of the inshore waters, such as those of

Pine Island Sound (Figure 1), maintain high salinities of

28.5 to 32.8 ppt (Alberts et al. 1969:1). Additionally,

most species of "estuarine" fish exploited by Charlotte

Harbor's prehistoric inhabitants at some point in their life

cycles migrate to marine waters. These estuarine/marine

fishes composed the bulk of the aboriginal protein intake as

indicated by zooarchaeological data (Fradkin 1976; Massaro

n.d.; Milanich et al. 1984).

Several modern researchers (Goggin and Sturtevant 1964;

Hann 1991; Lewis 1978; Marquardt 1986, 1987, 1988; Widmer

1988) have drawn on the Spanish writings of Fontaneda

(1945), Solis de Merds (1923), Rogel (Vargas Ugarte 1935),

and others, to synthesize the ethnohistory of the sixteenth-

century Calusa. Marquardt (1987:99-100) points out that

although the elite-dominated, tributary Calusa are usually

referred to as a chiefdomm" by archaeologists, it could be


argued that the society was an "early state" (Claessen

1978:538-580) or a "weak tribute-based state" (Gailey and

Patterson 1988:79) based on Spanish accounts. The

ethnohistoric sources specifically indicate that the Calusa

were nonagricultural and to date no evidence to the contrary

has been produced (Milanich 1987; Scarry and Newsom 1992;

for one opposing view see Dobyns 1983:126-130).

Wild plant foods, mostly in the form of fleshy fruits,

have been identified archaeologically. They include

hackberry, cocoplum, seagrape, mastic, prickly pear, cabbage

palm, saw palmetto, and hog plum (Scarry and Newsom 1992).

Scarry and Newsom (1992) document the virtual year-round

availability of the various fruits. The recently excavated

waterlogged samples (A.D. 200) from the Pineland Site

Complex on Pine Island are already adding to the list of

archaeological fruits (Newsom, personal communication,


Scarry and Newsom (1992) argue against the likelihood

of grain crops (maize and starchy seeds) being important in

prehistoric southwest Florida. Their experience has shown

that wherever maize is a subsistence base, cob remains have

been recovered in some number. Maize cob remains have never

been found in south Florida. Although corn pollen has been

identified at the Fort Center Site near Lake Okeechobee

(Sears 1982:120), Johnson (1990:210) argues that the soils

in question could not have supported maize cultivation. The


starchy seeds that are identified in the Charlotte Harbor

samples are not of the cultivated varieties that are

important in the prehistoric Midwest and Midsouth of the

United States (Scarry and Newsom 1992).

The Spanish chroniclers relate that the Calusa obtained

a wild plant root from interior south Florida for the

purpose of making a bread. In light of the non-

preservability of root remains and the absence of a

byproduct (unlike the case of maize cobs), the importance of

this food category remains open to debate. It has been

suggested that cut shark teeth found at the Fort Center and

Granada sites may have been used to create grater boards for

processing edible roots (Hale 1984:184; Kozuch 1991). To

date, the bread root described by the Spanish has not been

satisfactorily identified (Hann 1986).

The cultural history of the Calusa as an ethnic entity

remains unclear--whether they developed in the Charlotte

Harbor area (Widmer 1988) or originated from the Okeechobee

Basin of interior south Florida (Milanich and Fairbanks

1980:181). Furthermore, the emergence of Calusa complexity

may well have been a late phenomenon triggered by the

influx of sixteenth-century material goods into the

aboriginal economic system (Marquardt 1991).

Thus, we do not know if the ethnohistoric record is

appropriate for Charlotte Harbor temporal contexts other

than the protohistoric and historic Calusa. Conversely, nor


do we know if archaeological data from prehistoric sites are

appropriate for application to the protohistoric/historic

Calusa of the Spanish documents. Until population centers

such as the Pineland Site Complex are excavated to produce

ample data, both artifactual and subsistence, from

stratified contexts, archaeologists will not be able to

determine Calusa origins or the mechanisms that led to the

emergence of their complexity. And without an understanding

of the spatial and temporal paleoenvironmental context, we

cannot adequately evaluate the role of any perceived

subsistence change in the Charlotte Harbor region.

The LaQgging Maritime Perspective

The notion that maritime societies could develop

complex social and political formations without the benefit

of crop agriculture has long been debated, especially in the

case of coastal Peru (e.g., Moseley 1975; Moseley and

Feldman 1988; Osborn 1977; Wilson 1981). Prehistoric,

nonagricultural, complex peoples are indeed associated with

maritime settings in various locales of the world (e.g.,

North American Northwest Coast, southern California, coastal

Peru, southwest Florida, Norway and Sweden). This

association is increasingly being acknowledged by

researchers as theoretical biases inherent in unilinear

evolutionist schemes are broken down (Moseley and Feldman

1988). Unilinear evolutionists exaggerate the role of crop

agriculture as the primary cultural mechanism in the


emergence of complexity. As a result, evolutionary models

are colored by a terrestrial perspective, even when the

focus is on coastal cultures.

In Florida archaeology, symptoms of this terrestrial

bias include inappropriate recovery methods, untested

seasonal settlement models, and uncritical artifact

interpretation (see Russo 1991; Walker 1991; Walker and

Marquardt 1992). Until the 1960s, Florida archaeologists

believed that prehistoric coastal peoples subsisted

primarily on deer and small mammals, supplemented in times

of dietary stress by shellfish and fish. The application of

fine-screen recovery techniques by zooarchaeologists (e.g.,

Milanich et al. 1984) has revealed instead that fish, often

relatively small ones caught in nets, were the main

component of the native diet and were far more important

than terrestrial mammals.

Untested settlement models that depict coastal peoples

solely as seasonal residents also have been challenged

recently. Russo (1991) demonstrates that as early as the

Middle Archaic, people in southwest Florida lived year-round

on the coast and built purposeful mounds. A third symptom

of the terrestrial bias is a failure to recognize maritime-

related artifacts despite the obvious coastal association

and an available body of pertinent evidence (Walker 1991,

1992; Walker and Marquardt 1992).


There is little that argues against an estuarine/marine

food base for the Calusa and much that argues for it.

Widmer (1988) convincingly calls for an unusually high

productivity in Charlotte Harbor's estuarine/marine

environment--a year-round productivity capable of supporting

a large, sedentary, prehistoric human population. However,

as with any economic system, we cannot assume that these

estuarine/marine food resources relied upon in prehistory

remained uniformly productive and stable through space and

time for any given region. Limited by the available data,

Widmer's environmental context for the Calusa primarily

operates at broad regional (all of the southwest Florida

coastline) and temporal scales (e.g., he chooses the

traditional sea-level models). We must understand the

spatial variability of estuarine/marine resources at smaller

scales of analysis, as well as the impact of high-intensity

storms, freezes, and longer-term inlet and sea level-

dynamics in order to evaluate prehistoric human-environment


The question then becomes how to investigate Charlotte

Harbor's environment, its fluctuations, and its relationship

to prehistoric human inhabitants through space and time.

Zooarchaeological evidence (i.e., vertebrate and

invertebrate skeletal remains) represents an analytic medium

of great relevance to this question. Archaeofauna can serve

as a paleoecological data set if we make the assumption that


animal foods were procured within close proximity to the

site where the skeletal remains are recovered by


Research Goal and Objectives

The goal of this research is to employ regional

baseline zooarchaeological data to initiate a spatial and

temporal study of human-environment relationships in

prehistoric Charlotte Harbor. Such understanding is also

the goal of environmental archaeology (Butzer 1982; Evans

1978), a pursuit for which zooarchaeology is only one avenue

of inquiry. Zooarchaeological remains associated with

sedentary, coastal fisher-gatherer-hunter groups such as the

Charlotte Harbor people constitute a valid proxy data base

from which to begin to model paleoenvironments and the human

responses to them through space and time.

Independent, supportive data are essential to such

model-building (Dincauze 1987:318; King and Graham

1981:136-137; Rhoads and Lutz 1980:7, 11-12). Consequently,

data from estuarine ecological, climatic, and geological

research are drawn on. The Charlotte Harbor study,

nonetheless, is preliminary and hypotheses remain to be

tested and modified with new data sets.

Logically, one cannot truly "reconstruct" a

paleoenvironment (Dincauze 1987:292), but one can construct

a model of a past environment at a specified spatial and

temporal scale. Because of the interactive and


interdependent nature of environments, circularity in

reasoning, at times, becomes practically unavoidable in this

endeavor (Dincauze 1987:291-292). Despite this drawback,

the present study holds promise for research in the

Charlotte Harbor region.

The nature of model-building is to generalize (Levins

1966:421-422) for heuristic or operative purposes. In the

Charlotte Harbor model, it is necessary to simplify

environmental variation so that archaeologists can ask and

answer questions at a scale of, say, 100- to 200-year

increments (congruent with radiocarbon dating). In other

words, the goal, where possible, is to obtain data sets that

mediate potential short- and medium-term variation due to

intraannual and year-to-year change; for example, the

spatial perspective is based, with one exception, on

"averaged" site samples. However, intraannual and

year-to-year discontinuities do require careful

consideration when temporal interpretations are inferred.

Awareness of environmental continuity and change in

space and time at multiple scales should eventually allow

Charlotte Harbor archaeologists to focus on hypotheses more

specific to cultural change. In other words, we cannot make

valid inferences about cultural change based on faunal

patterns if we fail to recognize operative environmental

parameters at specific spatial and temporal scales. This is

because Charlotte Harbor is characterized by habitat


heterogeneity in its spatial distribution and by geophysical

dynamism through time; both attributes are typical of most

environmental systems. It is these operative factors that

dictate the comparability of intersite and intrasite

zooarchaeological samples and provide context for

human-environment relationships.

To illustrate, within a region such as Charlotte Harbor

a zooarchaeological assemblage from one site may be very

different from that of another site due to differences

(i.e., qualitative or quantitative) in the habitats that

surround each site. Therefore, between sites an assemblage

from one time period may be different from that of another

period because of a difference in location rather than a

diachronic cultural change. Within a site, an assemblage

from one time period may be different from one of another

time period due to a geophysical change such as a

fluctuation in sea level or the creation/closing of a nearby

ocean inlet rather than a diachronic cultural change.

This is not meant to imply that sociohistorical factors

were absent from Charlotte Harbor's prehistoric trajectory

of fauna use. For example, an apparent diachronic,

intrasite variation could be simply explained by variation

in site deposits (e.g., midden versus domestic floor) based

on patterning of artifacts, post holes, etc. Clearly, human

agency introduces a complex web of variables that interact

with the biotic and physical environments. We can begin to


identify this complexity only through familiarity with

environmental context.

In this dissertation, it is proposed that Charlotte

Harbor's recent estuarine paleoenvironment can be modeled

from perspectives of both space and time at local and

regional scales. Such a model, with continued adjustments,

can serve as a comparative base by which to measure

human-environment interaction. The approach used here

hinges on the existence of a prehistoric faunal exploitation

pattern that focuses on nearby resources. This pattern is

typical of maritime populations (Yesner 1980:730), and it is

established that the Charlotte Harbor zooarchaeological data

also reflect this strategy.

The Charlotte Harbor model-building exercise consists

of the following five objectives: (1) the spatial modeling

of modern estuarine heterogeneity via a gradient analysis;

(2) the spatial modeling of prehistoric estuarine

heterogeneity (using independent zooarchaeological data)

also via a gradient analysis, which serves as a test of

environmental comparability between present and past; (3 and

4) the overlay of potential short-, medium-, and long-term

temporal variation onto each of these two gradient models;

(5) the integration of the spatial and temporal dimensions

at both local and regional scales.


Sample Context and Excavation

The study has as its research universe the Charlotte

Harbor estuarine ecosystem, called here simply "Charlotte

Harbor." It is broadly defined as the subtropical coastal

area extending from Charlotte Harbor proper in the north to

Estero Bay in the south (Figure 1). For the purposes of

this study, then, the greater Charlotte Harbor area

constitutes a "region" (south Florida is also a region,

although broader in scale). The Charlotte Harbor region is

an arbitrary delineation based on a coastal ecosystem and

thus serves only as a starting point toward the

understanding of human-environment relationships in a

"dynamic region" (Marquardt and Crumley 1987:7-9). For

example, the rough chop of waters separating the Pine Island

Sound and Charlotte Harbor-Cape Haze areas (Figure 1) may

have represented a more realistic cultural boundary in the

prehistoric past. Point locations (e.g., archaeological

sites) within the region constitute "localities." The study

focuses on these two spatial scales, designated by Dincauze

(1987:261-262) as "mesoscale" (regional) and "microscale"


Although the cultural history of southwestern Florida

extends to the Early Paleoindian period, the time frame

under study in this dissertation is limited to approximately

600 B.C. to A.D. 1400, encompassing Caloosahatchee I through

IV periods (Table 1). The 2000-year span falls within


Dincauze's (1987:262) "mesoscale" temporal classification

and Butzer's (1982:24) "third order" scale of climatic

variability. Within these temporal scales, others of a

finer resolution also are recognized from which meaning is

inferred; any such scale is termed an "effective scale"

(Marquardt and Crumley 1987:2, 16; Marquardt 1985:69-70).

The use of effective scale as an organizing concept is

essential to a temporal study of the Charlotte Harbor

region. Short-term (i.e., from one day to one year),

medium-term (i.e., year-to-year), and long-term (i.e., one

hundred to several hundreds of years) effective scales in

the dynamics of the region's paleoenvironmental variation

are recognized in this study.

Excavation, volumetric, and chronological data for the

samples used in this study are presented in Table 2. All

samples exhibit good preservation owing to the predominant

calcium carbonate matrix of shell and bone. Samples were

selected on the basis of stratigraphic context. The four

Big Mound Key (8CH10) samples, excavated by George Luer in

1982 (Luer 1986:143), are from a large stratified pit

located at the summit of West Mound. Thirteen additional

samples are from column levels measuring 50 cm x 50 cm x 10

cm, excavated under the direction of William Marquardt and

the author in 1985 and 1986. For these samples from Cash

Mound (8CH38), Useppa Island (8LL51), Josslyn Island

(8LL32), and Buck Key Shell Midden (8LL722), designations


such as "A-l," "A-2," etc. indicate the excavation unit, and

the third number refers to the vertical level (e.g., "A-l-4"

is the fourth vertical level of Test Unit A-l).

A total of 206,474 bone and shell specimens were

identified in the seventeen samples. A total of 22,557

minimum number of individuals (hereafter "MNI") were

calculated. Table 3 presents a summary of these data,

broken down by sample and vertebrates versus invertebrates.

Species-specific data for all seventeen samples are

presented in Appendix A.

Big Mound Key. 8CH10

Located on the southwestern shoreline of the Cape Haze

Peninsula in Charlotte County (Figure 1), Big Mound Key is a

7 m-high shell mound complex that extends over a 15 ha area.

It is possible that the mound complex was constructed in a

spider-like effigy form. Archaeological work at Big Mound

Key has been limited but the site seems to have been

occupied since ca. A.D. 200 and possibly earlier (Luer

1986:105-106). During a site visit in the 1950s, the

Bullens (Bullen and Bullen 1956:50-51) collected Leon-

Jefferson and olive jar sherds indicating a seventeenth-

century occupation.

A large portion of Big Mound Key was intensively

bulldozed in the 1970s by treasure hunters. Along one of

the linear cuts, George Luer documented and excavated a

large pit containing stratified midden (see Marquardt


1992b:Figure 29), dating to ca. A.D. 800 (Table 2). He

collected several bulk samples for archaeobiological

analyses. Four of these were selected for inclusion in this

dissertation (Tables 2 and 3; Appendix A). Other associated

research includes Cordell (1992), Marquardt (1992b), Scarry

and Newsom (1992), and Upchurch et al. (1992).

Cash Mound. 8CH38

Cash Mound, situated in Turtle Bay (Figure 1), was

probably first inhabited during a low sea-level stand and

later became surrounded by water when sea level rose. It is

a large midden/mound site rising to more than 6 m in height

and measuring 200 m long by 125 m wide. Portions of the

site have been damaged by treasure hunters, "shell-

borrowing" activities, and storms. The Bullens excavated at

Cash Mound in 1954 (Bullen and Bullen 1956), representing

the first and only professional work here until recently

(see Marquardt 1992b).

Marquardt profiled an eroded face of a portion of

midden/mound and removed twenty-two 50 x 50 x 10 column

level samples for study. Four levels were radiocarbon-dated

to A.D. 270 60, A.D. 190 80, A.D. 150 90 (these three

are Caloosahatchee I period), and A.D. 680 70

(Caloosahatchee II period) (Table 2). These four samples

were chosen for zooarchaeological analysis (Tables 2 and 3;

Appendix A). Associated research includes that of Cordell


(1992), Marquardt (1992b, 1992c), Scarry and Newsom (1992),

and Walker (1992).

Useppa Island. 8LL51

Useppa Island is located on the estuarine side of Cayo

Costa, south of Boca Grande Pass (Figure 1). Useppa

Island's eastern edge exists as a roughly 6 m-high

Pleistocene dune remnant (Stapor et al. 1991; Upchurch et

al. 1992). Archaeological deposits on Useppa are extensive

and date as far back as 3675 B.C. Sites on Useppa were

first tested by J. T. Milanich and J. Chapman (Milanich et

al. 1984); they excavated in several locations on the island

in 1979 and 1980, demonstrating occupations from the Archaic

through the 19th century.

Marquardt's (1992b) more recent excavation in the

Collier Inn locality has produced a similar timespan of

shell midden and burial deposits. A single column level

sample, A-4-2, from this work was chosen for

zooarchaeological study (Tables 2 and 3; Appendix A). It

radiocarbon dates to 570 60 B.C. (Terminal

Archaic/Caloosahatchee I). Associated studies include those

of Cordell (1992), Hansinger (1992), Marquardt (1992b,

1992c), Quitmyer and Jones (1992), and Scarry and Newsom


Josslvn Island. 8LL32

Three of Josslyn's 19.4 hectares (Figure 1) are

comprised of shell midden/mounds that reach a maximum


elevation of 6.02 meters above sea level and, according to

Frank Hamilton Cushing (1897:337), courts and waterways.

Except for Cushing's brief 1895 investigation of one of the

"courts," Josslyn's archaeological deposits have received

little attention until the Florida Museum of Natural

History's recent involvement (Marquardt 1984, 1992a).

Josslyn's dense growths of red, black, and white

mangroves, trees of buttonwood, stopper, strangler fig, and

gumbo limbo are typical of coastal southwest Florida's

native subtropical vegetation. Geological coring has

demonstrated that the archaeological portion of Josslyn is

the oldest part of the island (Upchurch et al. 1992).

Futhermore, the lowest 65 cm (more or less, depending on the

tide) of midden is today submerged under water (Marquardt

1992b). These two pieces of information suggest a lower sea

level at the time of Josslyn's earliest occupation at circa

130 B.C., if not earlier.

In 1985, an extensive vertical profile was cleaned in a

deep looter's trench (Marquardt 1992b). From this area,

designated as operation A-l, thirty-eight 50 x 50 x 10 cm

column levels were removed for intensive analyses. Four of

these levels, dating to 130 90 B.C., 120 70 B.C. (both

Caloosahatchee I), A.D. 820 70 (Caloosahatchee IIB), and

A.D. 1200 60 (Caloosahatchee IIB/III), were chosen for

zooarchaeological study (Tables 2 and 3; Appendix A). The

inundated midden/mound base (described above) was not dated,


but a radiocarbon date was obtained from just above the

water line (see Table 2, #13). Associated studies of this

Josslyn context appear in Cordell (1992), Marquardt (1992b,

1992c), Quitmyer and Jones (1992), Scarry and Newsom (1992),

and Walker (1992).

Buck Key Shell Midden. 8LL722

Located along the northeastern shoreline of the island

of Buck Key, the Buck Key Shell Midden consists of low

"mounds" (no higher than 3 m) of shell, bone, and

artifactual debris, surrounded by red and black mangroves

(Figure 1). The middens appear to be undisturbed and have

not been investigated professionally until the 1985 work

(see Marquardt 1992b). Buck Key is today nestled behind

Captiva Island in bay waters, but originally was formed as a

barrier island between about 1,200 and 1,500 years ago

(Stapor et al. 1987:167, 169).

The Buck Key Shell Midden and its associated sand

burial mound, 8LL55, have been radiocarbon-dated to A.D.

1040-1350 (Table 2). Test A-l, placed in the shell midden

site, was excavated to 140 cm below surface (Marquardt

1992b). Test A-2, adjacent to A-l, was a 50 x 50 column

sample, excavated in ten levels. Two samples from this

column were selected for zooarchaeological analysis (Tables

2 and 3; Appendix A). Two samples originated from Test B in

the same manner (Tables 2 and 3; Appendix A). Associated

studies include Cordell (1992), Hutchinson (1992), Marquardt


(1992b, 1992c), Scarry and Newsom (1992), Upchurch et al.

(1992), and Walker (1992).

Zooarchaeolocical Methods

Sample Processing

Initially, entire levels were processed and analyzed,

but as our study progressed we found that in some cases

lesser volumes produced just as representative a data set

based on Wing and Brown's (1979:118-119) technique of

comparing number of species with minimum number of

individuals. Volumetric variation among other samples

(Table 2) is due to the varying quantity of large gastropod

shells which, once excavated, do not pack as tightly as

other midden remains.

The midden samples were water-floated in a 1.60 mm

(1/16") mesh box screen to recover botanical remains. After

slow air drying, the heavy fraction was sorted through a

series of geological sieves corresponding to 6.35 mm (1/4"),

2.00 mm (1/13"), and 1.60 mm (1/16") mesh sizes. The 6.35

mm vertebrate and invertebrate fragments were sorted,

identified, and quantified. The 2.00 mm vertebrate material

was sorted, identified, and quantified whereas the

invertebrate remains were subsampled by weight to determine

proportions only for the major classes (e.g., Gastropoda,

Bivalvia). This method allowed the inclusion of a minimum

meat weight estimate for unidentified 2.00 mm-screened

molluscan remains (category "Mollusca" in Appendix B).


Wing and Quitmyer (1985:49-58) dramatically demonstrate

the importance of fine-screen data recovery when dealing

with estuarine environments. The present study suggests

that a 2.00 mm mesh is an efficient screen size for the

objectives of the Charlotte Harbor study and that relatively

little diagnostic (below Class) material is found in the

1.60 mm-screened sample. Nevertheless, weight of the 1.60

mm-screened remains is essential if minimum meat estimates

are to be calculated. This has been done, again, by method

of proportion, sorting a 5% (by weight) subsample to

determine its major components. Thus, the unidentified

"Vertebrata (predominantly fish)" and "Mollusca" categories

include 6.35 mm, 2.00 mm, and 1.60 mm bone and shell

weights, respectively. Only 6.35 mm and 2.00 mm fauna

appear in all other quantifications.

Although I chose 6.35 mm and 2.00 mm screen sizes for

complete identification and MNI quantification of

invertebrates and vertebrates, respectively, I emphasize

that smaller screen sizes may be necessary to address

questions that I have not included in the present study.

Valuable invertebrate seasonality information, for example,

can be lost even through screen mesh as small as 1.60 mm.

For any particular environmental locale and set of research

questions, selection of screen size must be considered



Identification and Quantification

Specimens were identified using the comparative

collections of Zooarchaeology and Malacology, both located

at the Florida Museum of Natural History, Gainesville,

Florida. Scientific nomenclature and common names follow

general laboratory usage in 1986 for mammals, birds,

reptiles, and crustacea: Robins et al. (1980) for fishes,

Abbott (1974) for molluscs. Results of identification and

quantification for each of the seventeen samples are

presented in Appendix B. Fragment count and description,

fragment weight, and linear measurements are the three types

of primary data recorded in this study. Fragments of all

taxa were counted except for unidentified "Vertebrata" and

"Mollusca" (Appendix B, footnote b). In addition, counts of

unsided oyster and mussel valve fragments for Cash Mound

were of such magnitude that quantification other than shell

weight was impractical and would have served no purpose

(Appendix B, footnote e). Fragment weight was recorded,

providing the basis for minimum edible meat weight

estimates. Along with descriptive data concerning the

identification of specimens, linear measurements (in mm)

were taken for maximum meat estimations and other specific

purposes. Measurements followed the guidelines illustrated

in Quitmyer (1985:42-48) for vertebrates and invertebrates.

Secondary data include MNI, minimum edible meat

estimates, and maximum edible meat estimates. Standard


procedure was followed for calculating MNI by cultural unit,

comparing element side with age, size, and sex (Grayson

1984:27-48; Wing and Brown 1979:123). Edible meat weight

represented by bone and shell remains is presented as a

range, using a "minimum" and a "maximum" prediction

(Quitmyer 1985:38). Edible meat weight is here defined as

only the muscle tissue, with skin, viscera, and bone

subtracted. These predictions are made by establishing

allometric correlations between skeletal weight and meat

weight and between linear dimension and meat weight by using

least-squares regression (Casteel 1974; Hale et al. 1987;

Quitmyer 1985:37-38; Reitz et al. 1987:305; Wing and Brown

1979:127). These scaling methods are referred to as

skeletal mass allometry (using skeletal weight) and

dimensional allometry (using linear measurements) and employ

the allometric equation (Schmidt-Nielsen 1985:15; Simpson et

al. 1960:397):

y = aXb

log Y = log a + b (log X).

Tables A-l and A-2 list regression values for the

y-intercept and slope, based on data recorded at the Florida

Museum of Natural History. Table A-3 presents methods by

which maximum meat weights were estimated when regression

values were not available. These estimates were made by a

one-to-one size comparison with a modern specimen having a

known meat weight, or by using an average of known weights


if an archaeological specimen could not be matched to a

modern one. All values used in this study date to 1987 or

earlier and are subject to constant updating.

Throughout the Charlotte Harbor study, interpretive

emphasis is placed on the technique of Minimum Number of

Individuals (MNI). The primary quantitative objective of

the zooarchaeologist is to measure relative abundance of

species, but all methods used to do so are inherently flawed

to some degree. There are no perfect sampling or

quantitative procedures by which to analyze faunal remains

(Grayson 1984; Jackson 1989; Wing and Brown 1979). However,

it is my opinion that much of the current critical

assessment of the MNI technique (see Grayson 1984) is not

applicable to the study of maritime settings. Because of

the nature of estuarine/marine fauna and the technology used

for their exploitation, I believe that MNI units are very

appropriate measurements for Charlotte Harbor's faunal


Sample Size

Adequacy of sample size can be assessed by determining

the point of diminishing returns, that is, when few new

species are added to the faunal list (Wing and Brown

1979:118-119). I have attempted such an assessment for the

southwest Florida study area by comparing number of taxa to

MNI for each 2.00 mm (1/13") screened sample (Figures A-l

and A-2).


Figure A-I illustrates the distribution of vertebrate

samples. An MNI of 150 is the point of diminishing returns.

In other words, few new species are added once one has

identified ca. 150 individual animals. Most of the samples

show a relatively high species diversity. Following this

guide, ten samples may be considered less than

representative. However, because my methodology emphasizes

the combined treatment of vertebrates and invertebrates, a

second graph has been constructed to convey sample size

adequacy (Figure A-2).

Two curves emerge. The higher curve represents highly

diversified samples, all but one meeting the criterion of

600 MNI. The lower curve represents distinct types of

samples, all from the Cash Mound column (samples #5, #6, #7,

and #8). The Cash Mound pattern may suggest a specialized

area of the site (see text for discussion). The position of

sample #4 (Big Mound Key, Layer 2) on the graph also

suggests a specialized assemblage. Although the point of

diminishing returns is not known for this lower curve

(broken line), samples are probably well beyond where it

would occur. The sizes of these five samples, then, are

more than adequate within their own contextual realm. Data

of this kind are important for any region of study, for they

can be used as a guideline for future faunal analyses.

Food vs. Commensals

General criteria for deciding what was or was not eaten

are based on species, size, quantity of individuals, and

archaeological context. All vertebrate species identified

are assumed to have been eaten. Small barnacles and many

forms of small gastropod and bivalve animals were surely not

consumed, at least in the middens sampled. There was little

archaeological evidence for dense deposits of small

gastropod or bivalve shells. It is clear from experimental

midden research (Wing and Quitmyer 1992) that many creatures

make their way to the middens attached to larger host

species. Often small bivalve specimens were found with both

valves intact, or shells were water worn. Thus, certain

species were not included in the dietary analysis (Appendix

B, footnote c). However, occasional distinct assemblages

warranted inclusion. For example, the cross-barred venus

(Chione cancellata) at Useppa Island (Table B-9) and spotted

slipper shell (Crepidula maculosa) at Buck Key (Table B-16)

were of such size and quantity to suggest purposeful


Sources of Bias

Preservation problems relating to fragment counts and

weights are numerous and uncontrollable (Grayson 1984:21-22;

Wing and Brown 1979:121-123). The effects of scavengers and

differential preservation due to depositional conditions,

bone/shell condition, or bone/shell structure are difficult


and often impossible to assess. Results of a midden

experiment by Wing and Quitmyer (1992) suggest that

post-deposition losses of fish bones occur in shell middens.

Equally disconcerting are the endless undetectable

socio-cultural activities that determine archaeological

faunal patterns. One problem associated with massive,

complex shell mound sites is the taphonomic distinction

between primary and secondary midden deposits. Based on

test unit location, stratigraphy, and radiocarbon dates, I

believe that the midden samples in the present study

represent primary deposits.

As is often the case, it is the absence or infrequent

occurrence of expected species that puzzles the

zooarchaeologist. There are four examples of fish that are

today abundant in the Charlotte Harbor area but are either

missing or infrequent in the shell midden samples of the

present study. The significance of mullet (Mugil spp.) in

southwest Florida prehistory is a matter of concern (Goggin

and Sturtevant 1964:185; Marquardt 1986:66). Although a

mullet fishery is reported in the ethnohistoric literature

(L6pez de Velasco 1894:163; Weddle 1985:22) and the fish are

abundant today, relatively few bones are recovered from

sites, often only thoracic vertebrae. Whether the

explanation is one of sampling, preservation, environmental

change, or cultural practice should be investigated.


In addition to mullet, three more species are

conspicuously rare or absent from the faunal samples, based

on Wang and Raney's modern survey (1971:54): the bay

anchovy (Anchoa mitchilli); the silver jenny (Eucinostomus

gula); and the spadefish (Chaetodipterous faber). The first

two are fishes in the same small size class as the

killifishes (Fundulus spp.), a genus identified among the

midden remains. Perhaps these were eaten whole, and perhaps

the fibrous structure of spadefish bones prevented

preservation of this species. Hypotheses such as these

should be tested.

The nature of column sampling has inherent problems

related to intrasite (horizontal) representativeness. An

additional concern is the comparability of the Big Mound Key

feature, a large midden-filled pit, to the general midden

samples taken from all other sites. The validity of

comparison may or may not depend on the unknown function of

the large pit. I postulate that the pit's primary purpose

was something other than garbage disposal and that the food

remains were deposited secondarily, representing a sample

similar to general midden areas. This should be tested with

future excavation at Big Mound Key.

Several basic problems that plague scaling techniques

when applied to archaeofauna are discussed elsewhere

(Grayson 1984; Jackson 1989; Wing and Brown 1979).

Additionally, many species-specific regression values are


not yet available for both minimum and maximum estimates.

Recently, Grayson (1984:172-174) has argued that only the

dimensional allometric method of meat weight prediction is

valid for zooarchaeological purposes. Despite this

controversy, allometric scaling, used as a method for

predicting animal body weights (extended to meat weight for

this study), has been tested and shown to produce the most

accurate results of currently employed techniques to

estimate biomass (Casteel 1978:71-77; Wing and Brown


Another example of bias in the meat-weight estimation

method stems from frequent low MNI counts for invertebrates

in relation to fragment weight. For certain species (e.g.,

eastern oyster, lightning whelk, banded tulip, Florida horse

conch), this is seemingly due to a high degree of

fragmentation, shell structure, and density, or perhaps to a

limited size range used in scaling modern specimens.

Sometimes the resulting maximum estimate for these animals

is lower than the minimum estimate (Appendix B, footnote f).

Comparative Dietary Contribution

Minimum and maximum edible-meat weights were estimated

(discussed above) for all 17 faunal samples to provide a

range of meat potential for each animal (Appendix B).

Figures A-3 and A-4 summarize these results by site,

combining intrasite data. Bony fishes (Osteichthyes) stand

out as the primary contributors to the aboriginal diet based


on both minimum and maximum meat estimates. Although the I

importance of gathering shellfish is dramatically evidenced

by massive shell mounds dotting the landscape and

quantitatively supported by MNI figures, its role is

considerably diminished when viewed from a dietary

perspective (Figures A-3 and A-4).

Nutritional analysis has shown that, gram for gram,

shellfish contains substantially less protein and fat and

fewer calories than fish and mammals (Parmalee and Klippel

1974:431). Cash Mound's 84% MNI and 58% minimum meat of

oysters and mussels (Figure A-3), respectively, are reduced

to a paltry 8% when maximum meat is estimated (Figure A-4).

The predominance of meat contribution derived from fishing

activities is underscored when the meat of sharks and rays

(Chondrichthyes) is added to the bony fish category. This

is most evident in the Buck Key samples where 81% of the

minimum meat estimate results from fishing (Figure A-3).

Sharks and rays are represented in all site samples,

with the Useppa Island sample showing a high minimum meat

weight estimate of 25% (Figure A-3). The work of Milanich

et al. (1984) at Useppa also showed an abundance of shark

remains. As do the remains of white-tailed deer, the

appearance of adult sharks in midden samples implies

butchering and village distribution of meat. However, most

shark individuals in the study samples are juveniles.


Whereas reptiles and mammals generally represent a

negligible portion of the diet based on estimates of minimum

meat, they can be significant contributors if the maximum

meat estimates of large individuals are considered. When

the maximum meat of one sea turtle is estimated, its dietary

importance in the Big Mound Key samples becomes 19% and for

the Buck Key samples, 13% (Figure A-4). However, the high

mammal maximum meat percentage of 30% for the Big Mound Key

samples (Figure A-4) may be misleading. The deer bones

recovered from the four sampled strata in the short-lived

refuse pit possibly represent a single deer--1 MNI instead

of 4--which would substantially reduce the meat percentage.

Table 1. Generalized Cultural Chronology for
the Caloosahatchee Area (adapted from
Marquardt 1992b and Cordell 1992).

Date Period Some Diagnostic Artifacts

A.D. 1500-

A.D. 1350-

A.D. 1200-

A.D. 800(?)-

A.D. 650-

500 B.C.-
A.D. 650

1200 B.C.-

2000 B.C.-
1200 B.C.

5000 B.C.-
2000 B.C.

Caloosahatchee V

Caloosahatchee IV

Caloosahatchee III

Caloosahatchee IIB

Caloosahatchee IIA

Caloosahatchee I

Terminal Archaic

Late Archaic

Middle Archaic

European artifacts (e.g.,
metal, beads, olive jar

Safety Harbor, Glades
Tooled, and Pinellas
Plain pottery; Belle
Glade Plain diminishes

St. Johns Check Stamped,
Englewood ceramics; Belle
Glade Plain prominent

Belle Glade Red present;
Belle Glade Plain

Beginning of Belle Glade
Plain and SPCB ceramics;
Glades Red; thinner

Thick sand-tempered plain
pottery with round and
chamfered lips

Fiber-tempered pottery;

Orange Plain, Orange
Incised, Perico Incised,
Perico Plain, St. Johns
Plain; steatite

Coastal sites, but no
ceramics; broad-stemmed
bifaces, e.g., Newnan;
mortuary ponds

Table 1--continued.

Date Period Some Diagnostic Artifacts

6500 B.C.-
5000 B.C.

8500 B.C.-
6500 B.C.

11500 B.C.-
8500 B.C.

Early Archaic

Late Paleoindian

Early Paleoindian

Sites on coastal dune
ridges ca. 5000 B.C.;
earlier coastal sites
probably inundated by
rising sea level

Dalton and Bolen bifaces,
bone points, non-
returning boomerang,
socketed wooden point,
oak mortar, atlati spur

Only wooden tools known

Table 2. Zooarchaeological Samples Included in
the Charlotte Harbor Study.

Site Name Sample Sample Vol. C-14 Date
and Number Provenience Type (m3) (uncalib.)

1. Big Mound Key

2. Big Mound Key

3. Big Mound Key

4. Big Mound Key

5. Cash Mound

6. Cash Mound

7. Cash Mound

8. Cash Mound

9. Useppa Island

10. Josslyn Island

11. Josslyn Island

Layer 11

Layer 8b

Layer 7

Layer 2(1)b






















.014 A.D.880+140







column .028








Table 2--continued.

Site Name Sample Sample Vol. C-14 Date
and Number Provenience Type (m3) (uncalib.)

12. Josslyn Island A-1-22(23)b column .028 12070B.C.
(8LL32) level Beta-17334

13. Josslyn Island A-1-32(33)b column .028 13090B.C.
(8LL32) level Beta-17335

14. Buck Key Mid. B-2-5 column .018 A.D.135080
(8LL722) level Beta-16283

15. Buck Key Mid. B-2-9 column .025 A.D.125060
(8LL722) level Beta-16282

16. Buck Key Mid. A-2-6/7 column .023 A.D.133070
(8LL722) level Beta-16285

17. Buck Key Mid. A-2-11(9)b column .023 A.D.104080
(8LL722) level Beta-16287

a Luer 1986b.
b Level in parentheses is source of radiocarbon date.

Table 3. Summary of Zooarchaeological Data Included
in the Charlotte Harbor Study.

Number of Minimum
Identifiable Number of
Sample Fragments Individuals

Big Mound Key Layer 11
Vertebrates 8501 150
Invertebrates 4083 595

Big Mound Key Layer 8b
Vertebrates 8200 168
Invertebrates 5560 820

Big Mound Key Layer 7
Vertebrates 6000 98
Invertebrates 5832 1094

Big Mound Key Layer 2
Vertebrates 705 32
Invertebrates 4628 1564

Cash Mound A-l-4
Vertebrates 5953 246
Invertebrates 4864 1032

Cash Mound A-l-8
Vertebrates 2239 75
Invertebrates 5682 2821

Cash Mound A-1-17
Vertebrates 1995 69
Invertebrates 3112 1429

Cash Mound A-l-20
Vertebrates 2022 42
Invertebrates 4993 2531

Useppa Island A-4-2
Vertebrata 4937 208
Invertebrata 5685 895

Josslyn Island A-1-4 (1/2 vol.)
Vertebrata 9374 231
Invertebrata 5225 596

Table 3--continued.

Josslyn Island A-1-12
Vertebrata 14732 451
Invertebrata 13062 1021

Josslyn Island A-1-22
Vertebrata 6541 257
Invertebrata 6156 1209

Josslyn Island A-l-32
Vertebrata 13928 384
Invertebrata 7656 1110

Buck Key Shell Midden B-2-5
Vertebrata 11865 220
Invertebrata 5430 467

Buck Key Shell Midden B-2-9
Vertebrata 9793 153
Invertebrata 2809 241

Buck Key Shell Midden A-2-6/7
Vertebrata 3332 66
Invertebrata 5511 1476

Buck Key Shell Midden A-2-11
Vertebrata 1801 46
Invertebrata 4268 760

Total 206474 22557




Table 4. Regression Values for Minimum Meat
Weight Estimations.

Taxon N Log Slope
a b r2

Mammaliaa 40 1.41 0.81 0.91
Aves' 39 1.24 0.84 0.98
Serpentes" 14 1.06 0.94 0.98
Testudines' 9 1.65 0.53 0.74
Siren spp." 15 2.50 0.52 0.82
Carcharhinidae (vertebra wt.)b 48 2.35 0.88 0.98
Carcharhinidae (total wt.)c 11 0.94 1.38 0.98
Sphyrnidae (vertebra wt.) 18 1.91 0.99 0.96
Sphyrnidae (total wt.)c 20 1.88 1.03 0.98
Lamniformes (vertebra wt.)b 68 2.27 0.89 0.95
Lamniformes (total wt.)c 80 2.27 0.93 0.96
Rajiformes (total wt.)" 12 2.61 0.89 0.95
Osteichthyesd 80 1.34 0.90 0.96
Crustacea (Callinectes)a 11 0.99 0.82 0.58
Strombus alatusc 26 -0.68 0.88 0.86
Polinices duplicatusc 16 0.38 0.55 0.81
Melongena corona 100 -0.43 0.88 0.79
Busycon contrariumc 100 -0.75 1.14 0.91
Fasciolaria hunteriac 21 -0.86 1.35 0.98
Fasciolaria tulipac 26 0.11 1.00 0.94
Pleuroploca giganteac 42 -0.71 1.15 0.99
Gastropodac 135 -0.16 0.92 0.89
Geukensia demissac 100 -0.22 0.80 0.86
Crassostrea virginica (left)c 100 -0.59 0.97 0.96
Crassostrea virginica (right)c100 -0.31 0.96 0.97
Crassostrea virginica (total)c100 -0.77 0.97 0.97
Polymesoda carolinianac 40 0.01 0.83 0.85
Mercenaria campechiensisc 30 -0.51 0.86 0.96
Bivalviac 80 0.02 0.68 0.83

Table 4--continued.

Note: Regression formula:

a Source:
b Source:
c Source:
d Source:

transformed log

Quitmyer 1985
Fitzgerald 1986
Hale et al. n.d
Hale and Walker

Y = aXb
Y = log a + b(log X)
y = weight of meat in grams
x = bone, shell, or exoskeleton
in grams
a = y intercept
b = slope


Table 5. Regression Values for Maximum Meat
Weight Estimations.

Taxon N Log Slope r2
a b

Lepisosteus spp.c
Strombus alatusd
Polinices duplicatusd

Melongena corona

Busycon contrariumd

Fasciolaria hunteriad

Fasciolaria tulipad

Pleuroploca gigantead
Geukensia demissad
Crassostrea virginicad
Polymesoda carolinianad

Mercenaria campechiensi


48 0.93
18 0.39
68 0.84
12 1.40
9 0.91
8 0.98
17 0.68
13 0.75
35 0.74
14 0.53
99 0.70
26 -5.09
16 -1.47
16 -2.97
100 -4.83
100 -3.98
100 -5.84
100 -5.40
21 -5.23
21 -5.29
26 -2.97
26 -1.15
42 -5.62
80 -3.19
100 -3.62
100 -3.80
40 -3.26
0d30 -4.02
30 -1.04
135 -2.16



Vertebra wd
Vertebra wd
Vertebra wd
Vertebra wd
Vertebra wd
Vertebra wd
Atlas wd
Atlas wd
Atlas wd
Atlas wd
Alas/vert. i
Shell ht
Shell ht
Aperture ht
Shell ht
Aperture ht
Shell ht
Aperture ht
Shell ht
Aperture ht
Shell ht
Aperture ht
Shell ht
Shell ht
Valve lg
Lt valve lg
Valve lg
Valve lg
Hinge wd
Valve lg


Table 5--continued.

Note: Regression formula:
Y = aXb
transformed log Y = log a + b(log X)
where Y = meat weight in grams
X = linear measurement (mm)
a = y intercept
b = slope
a Measurements follow those described and illustrated in
Quitmyer 1985 and Hale et al. n.d.
b Source: Fitzgerald 1986
c Source: Quitmyer 1985
d Source: Hale et al. n.d.

Table 6. Nonregression Values for Maximum Meat
Weight Estimations.

Taxon N Weight Estimate (gm)

Odocoileus virginianusa
Procyon lotorb
Sigmodon hispidusb
Casmerodius albusb
Chelydra serpentinab
Kinosternon spp.b
Terrapene carolinab
Pseudemys sp.c
Gopherus polyphemusb
Siren lacertinab
Rana spp.b
Lepisosteus spp.b
Bagre marinus"
Ariopsis felisb
Opsanus spp.b
Fundulus spp.b
Mycteroperca microlepisb
Lutjanus spp.b
Eucinostomus spp.b
Haemulon spp.1
A. probatocephalusb
Sparisoma spp.b
Paralichthys albiguttab
Sphoeroides spengilerib
Chilomycterus schoepfib
G. demissa granosissimab
Other molluscad

1 or
1 or
1 or
1 or


14 comparative or 2164.22(x)
7 comparative or 380.68(x)
15 comparative or 89.06(x)
comparative var.
comparative var.

11 comparative or







Table 6--continued.

a Quitmyer 1985
b Florida Museum of Natural History Collections
c Nietschmann 1973:165
d Most other molluscan species required the comparative
method when fragmentation precluded measurements.

Figure 1. Map of the Charlotte Harbor Study Area
with Geographical Features and Archaeological
Site Locations Mentioned in the Text: (1) Solana
Site; (2) Big Mound Key; (3) Cash Mound; (4) Useppa
Island; (5) Pineland; (6) Josslyn Island; (7) Buck Key
Shell Midden; and (8) Wightman Site.

I :

, Q





0 10 KM


Figure 2. The Distribution of Charlotte Harbor
Zooarchaeological Vertebrate Samples by Number of
Taxa and Minimum Number of Individuals (MNI).
Numbers Refer to the List Given in Table 2.



.l 2 .1211
30- 03 01. 91.12oi

o,, 0l. .5 *13

0 *86
0 20


0 100 200 300 400 500 600

Minimum Number of Individuals

Figure 3. The Distribution of Charlotte Harbor
Zooarchaeological Invertebrate Samples by Number of
Taxa and Minimum Number of Individuals (MNI).
Numbers Refer to the List Given in Table 2.

) 1200 1600 2000 2400
Minimum Number of Individuals

Figure 4. Comparative Percentages of Zooarchaeological
Estimated Minimum Edible Meat Weights by Site and
Animal Group (Based on Data Presented in Appendix A).

Big Mound Key

Josslyn Island


Cash Mound

Buck Key


Useppa Island

MI Bony fishes
RL: Marine snails
I !Sharks, rays,etc.
^ Marine bivalves
= Mammals
E Crabs
-l Other

Figure 5. Comparative Percentages of Zooarchaeological
Estimated Maximum Edible Meat Weights by Site and
Animal Group (Based on Data Presented in Appendix A).

Big Mound Key

U3 Bony fishes
I3 Marine snails
El Sharks, rays, etc.
IMM Marine bivalves
iI Mammals
rl Turtles
j Amphibians
E Crabs
ii Birds

Cash Mound



The Present-day Charlotte Harbor Estuarine Ecosystem

An examination of the present-day natural environment

of the Charlotte Harbor area, along with western society's

recent impact on that environment, is a necessary first step

toward understanding past situations. A spatial model of

today's environment serves as a beginning standard for

paleoenvironmental modeling.

In the Charlotte Harbor region, three major rivers,

extensive inshore lagoons, salt marshes, mangrove forests,

and a series of barrier islands compose a complex and

dynamic estuarine ecosystem of an unusually high level of

biological production (Taylor 1974:207). Comp and Seaman

(1985:337) generally define estuaries as "semi-enclosed

bodies of water that (1) have a free connection with the

sea, (2) receive freshwater inflow through both overland

runoff and defined sources such as rivers, creeks, and

springs, and (3) contain a measurable salinity gradient."

The Peace and Myakka rivers converge to form the Charlotte

Harbor estuary proper, while to the south, the

Caloosahatchee River empties into San Carlos Bay, forming

the second major estuary (Figure 1). To the west, barrier


islands "enclose" these bodies of water, thus defining the

greater estuarine system at a regional scale. The two major

openings to the Gulf are Boca Grande Pass and San Carlos

Bay; secondary inlets are Blind, Redfish, Captiva, and

Gasparilla Passes.

Terrestrial ecological communities in the region

include mangrove forest, salt marsh, coastal strand, salt

barren, sabal-juniper hammock, oak-persea hammock, and pine

woods (Taylor 1974:210). Of these, the mangrove community

is most closely associated with the estuarine complex.

Mangrove forests extend over 22,927 ha in the study

area and are largely structured by zones of red (Rhizophora

mangle), black (Avicennia germinans), and white

(Laguncularia racemosa) mangrove varieties (Harris et al.

1983:129; Taylor 1974:210). The salt-tolerant red mangrove

dominates the water's edge throughout the estuarine system.

As one moves inland, the black and white varieties become

mixed with buttonwood (Conocarpus erectus) and other plant

species (Odum et al. 1982:2). Mammals using mangrove

forests feed on fruits, berries, insects, small reptiles,

seeds, mast, crabs, grasses, fish, bird eggs, mussels, and

other mammals. The white-tailed deer is the only mammal

known to include mangrove leaves in its diet (Odum et al.


The mangrove fringe (primarily red) and inshore

seagrass (primarily turtle grasses) meadow are the two most


productive habitats in the estuarine complex. Of lesser

productivity are the oyster bar, the littoral zone, and the

open Gulf water. The distribution and interrelationships of

all these habitats and their animal components largely

define the ecological structure of the estuarine complex.

Mangrove and seagrass ecosystems are among the most

productive biological systems in the world, even rivaling

agriculture (Odum et al. 1982:19; Zieman 1982:1). These two

estuarine plant groups produce enormous amounts of

leaf/blade detritus, supporting extensive aquatic food webs.

In addition, they provide protection from predators for many

fish and invertebrate species, particularly while in their

juvenile stages. They are closely interrelated, the

seagrass areas often extending right up to mangrove

shorelines (Odum et al. 1982:50).

Seagrasses of the Charlotte Harbor area have not been

adequately studied (Estevez et al. 1984:S-22) even though

today they account for 23,682 ha (Harris et al. 1983:133).

Primarily occurring in broad shallow-water "meadows," turtle

grasses (Thalassia testudinum and Halophila engelmanni),

shoal grass (Halodule wrightii), widgeon grass (Ruppia

maritima), and manatee grass (Syringodium filiforme) are the

five common seagrass varieties (Taylor 1974:210; Zieman

1982:8). These have slightly different salinity

requirements, with the turtle grasses being the most

abundant and forming the most expansive meadows. These


meadows support great densities of sessile and migratory

molluscs, as well as fishes and crabs that spend all or part

of their life cycles there. In addition, large predator

species frequent the inshore grasses in search of food.

If the present-day environment of Charlotte Harbor is

to be used as an analogy for the interpretation of its past

environment, we must be aware of change resulting from

modern human activity. The widely cited loss of scallop

populations due to reduced salinities and increased

turbidity as a result of causeway construction is a familiar

example (Estevez et al. 1984:PM90-PM91). More important for

our purposes, 9,904 ha of seagrasses (29%) and 128 ha of

oyster bar communities (39%) have disappeared from the

region since 1945 (Haddad and Harris 1985:668; Haddad and

Hoffman 1986:175). Researchers (Haddad and Hoffman

1986:184; Harris et al. 1983:134) attribute the startling

losses to the combined effects of dredging for the

Intracoastal Waterway, construction of the Sanibel causeway,

channeling of the Caloosahatchee River, and upland

pollution. Finally, as early as 1884, the natural flow of

the Caloosahatchee River was changed when the waterway was

linked to Lake Okeechobee (Gunter and Hall 1965:4).

A modern reduction in habitats translates into a

reduction of biological productivity. Thus, acknowledgement

of the existence of a prehistoric biotic productivity that

was much greater than that of today (at least at certain


times) is critical to our understanding of that past


The Present-day Estuarine Gradient

The ecological concept of an environmental gradient

(King and Graham 1981:129) is useful when applied to

estuarine situations for the purpose of determining faunal

distribution and abundance (Boesch 1977; Wells 1961).

Analysis of the estuarine gradient involves the recognition

of different ecological zones or communities ranging from

fresh to oceanic water and the extent to which different

aquatic species inhabit these areas (Boesch 1977:245; Odum

et al. 1982:51, 57). The zones and their associated faunal

assemblages have no sharp boundaries in space. Rather, they

exist as a graded continuum at the regional scale. The

habitat categories nonetheless allow description, and thus

an operative understanding of the heterogeneous distribution

of fauna along the estuarine gradient.

Although numerous limiting factors are involved in

gradient distributions of estuarine fauna, average salinity

is prominent among them (Boesch 1977:246; Wells 1961:239)

and provides a useful organizational tool at one or more

effective scales. For example, the oyster bed or bar

community (i.e., oysters and all associate fauna) exhibits a

certain range along the estuarine (regional) gradient;

within that range (also a continuum, in reality), point

locations can be classified as low-, mid-, or high-salinity


oyster bars. The distribution of associate fauna will vary

according to the designation (Wells 1961).

Unfortunately, there is no published gradient study of

Charlotte Harbor aquatic biota. However, based upon what is

available for the area and comparative data from other

estuarine environments, an informal model of Charlotte

Harbor's modern distribution of aquatic vertebrate and

invertebrate fauna for archaeological purposes is attempted

here. Because mobility patterns of vertebrates and

invertebrates are dramatically different, these two groups

are described separately. The descriptions are not meant to

be comprehensive; instead they emphasize animals whose

remains are commonly found in prehistoric human middens.

Figure 6 illustrates monthly salinity profiles that

closely typify the salinity gradient of the northern project

area. These data are from Wang and Raney (1971:18) and

represent readings taken at four of their collection

stations (numbers 33, 29, 13, and 15), chosen to illustrate

the general gradient moving from near-freshwater (Peace

River) to oceanic (Boca Grande Pass) conditions. Salinity

readings for Pine Island Sound show variance depending on

proximity to Captiva Pass, Redfish Pass, or Blind Pass (Wang

and Raney 1971:18). Additionally, Alberts and his

colleagues (1969:1) report that the Gasparilla Sound and

Pine Island Sound waters maintain marine salinities ranging

from 28.5 to 32.8 ppt. Similarly, the San Carlos Bay area


generally ranges from 25.0 to 35.0 ppt (Gunter and Hall


Tidal stages and thus vertical stratification should be

considered for any given estuarine location because there

can be great salinity differences between ebb and flood

position, and bottom and top waters (Estevez et al.

1984:CH113-CH118). However, tides in the Charlotte Harbor

area are of a mixed diurnal and semi-diurnal type with an

average amplitude of only about 0.60 m (Estevez et al.

1984:CH96). The implication of such a microtidal pattern is

that daily fluctuations in salinity are minor compared to

most of the world's estuaries. This is advantageous for

gradient modeling at a scale useful to archaeologists. The

fact that most sites are associated with very shallow waters

(0.3 to 1.2 meters) further mediates vertical salinity

differences for the archaeologist.

Figure 6 exhibits a pattern of seasonal salinity

fluctuation for the 1968-1969 period. Based on rainfall

data for 1965 through 1969 (Joyner and Sutcliffe 1976, cited

in Estevez et al. 1984:CH17), the only departure from an

average yearly pattern is the heavy March rain, shown in

Figure 6 as lowered salinities at all four stations.

In addition, wind shifts can result in substantial

salinity fluctuations along the gradient for intervals of

hours up to days. Periodic deviations from the average

pattern, whether daily, seasonal, or of several years


duration, imply short- or medium-term alterations in faunal

distribution and/or productivity.

Distribution of Vertebrates

Literature concerning aquatic vertebrate communities

(primarily fishes) in mangrove environments is readily

available (e.g., Odumn et al. 1982) and several systematic

fish studies exist for Charlotte Harbor (see Estevez et al.

1984; Taylor 1974:213). In particular, Gunter and Hall

(1969) and Wang and Raney (1971) present a data base useful

for archaeological research.

To describe the distribution of vertebrates, four

mangrove/fish community designations are borrowed from Odum

et al. (1982:50-51) and a fifth classification is added to

complete the salinity gradient. These are: (1) mangrove

basin; (2) mangrove-fringed streams; (3) mangrove-fringed

estuarine bays and lagoons; (4) mangrove-fringed oceanic

bays and lagoons; and, (5) the littoral zone and Gulf

waters. Types 3 and 4 are associated with seagrass meadows.

Type 1 is a backwater area, largely of freshwater

content, supporting species such as killifishes (Family

Cyprinodontidae), the greater siren (Siren lacertina), frogs

(Rana spp.), and freshwater turtles. The area immediately

to the north of Big Mound Key (8CH10) is an example of a

mangrove basin (Figure 1). These areas are known generally

to exhibit low species diversity, but sometimes high

densities of fishes do occur (Odum et al. 1982:50).


Type 2 includes major tributaries (e.g., Myakka, Peace,

and Caloosahatchee rivers), small streams (e.g., Whidden

Creek, Alligator Creek), and associated pools. These

streams are tidal-influenced, have sparse grass beds, and

show seasonal variance in terms of salinity and, thus,

species composition (Odum et al. 1982:52; Wang and Raney

1971). During rainy months, such as March and July (see

Figure 6, Peace River line), freshwater fishes sometimes

move into the estuary. Examples include Florida gar

(Lepisosteus platyrhincos), sunfishes (Lepomis spp.),

freshwater catfishes (Family Ictaluridae), and the

largemouth bass (Micropterus salmoides) (Estevez et al.

1984:PR342-PR354; Gunter and Hall 1969:20, 23, 31).

Conversely, marine predatory fishes such as needlefishes

(Family Belonidae), jacks (Family Carangidae), and stingrays

(Family Dasyatidae) invade the tidal streams in search of

food during dry periods (Odum et al. 1982:52). Fishes such

as the black mullet (Mugil cephalus), gray snapper (Lutjanus

griseus), sheepshead (Archosargus probatocephaIus), spotted

seatrout (Cynoscion nebulosus), red drum or "redfish"

(Sciaenops ocellatus), and silver perch (Bairdiella

chrysoura) use tidal streams and pools only as juveniles

(Gunter and Hall 1969; Odum et al. 1982:52). Most of these

species are represented in Charlotte Harbor's streams during

some part of the year (Wang and Raney 1971).


Environment Types 3 and 4 combine extensive mangrove

shorelines and seagrass meadows. The relationship between

these two habitats in terms of faunal use is unclear (Zieman

1982:75), perhaps due to the proximity of the two. Types 3

and 4 range higher in salinity than 1 and 2 (see Figure 6,

Charlotte Harbor and Bokeelia lines), exhibit coarser,

sandier sediments, and support a greater abundance and

diversity of fish assemblages.

In the Charlotte Harbor system, examples of Type 3,

estuarine bays and lagoons, are Turtle Bay, Bull Bay,

Matlacha Pass, and the eastern part of Pine Island Sound.

The oceanic bays, Type 4, tend to have the clearest water,

the highest salinities of inshore waters, and perhaps the

highest species diversity. Examples include San Carlos Bay,

Gasparilla Sound, and the inshore lagoons of Pine Island

Sound behind Blind, Redfish, and Captiva Passes. Because of

the configuration of the Charlotte Harbor system, it is

difficult to separate the species of Types 3 and 4 and so I

describe them as one unit.

Species such as pipefishes and seahorses (Family

Syngnathidae), gobies (Gobiidae), and the inshore lizard

fish (Synodus foetens) spend their entire life cycles within

the grassbeds. A second group of fishes largely uses the

meadows as a nursery ground, spending their juvenile life

stages in the nurturing grass habitat. Spotted seatrout,

spot (Leiostomus xanthurus), silver perch, red drum, pigfish


(Orthopristis chrysoptera), pinfish (Lagodon rhomboides),

sheepshead, and gag grouper (Mycteroperca microlepis) are

all common to abundant fishes among the grassbeds (Zieman

1982:50). Adults commonly inhabit the mangrove fringe. In

addition, anchovies are known to concentrate in seagrasses,

especially while juveniles (Carr and Adams 1973:515).

Wang and Raney (1971:22-23; 24) report that three

species of anchovy (Anchoa mitchilli, Anchoa hepsetus, and

Anchoa cubana) and the hardhead catfish (Ariopsis fells)

frequent grass flats but are abundant in all parts of the

Charlotte Harbor system. Pinfish, although densely

associated with seagrasses in juvenile and adult forms, have

a variable habitat distribution (Darcy 1985:3-6). Mullet

aggregate on a seasonal basis in grass areas, feeding

directly on grass blades (Zieman 1982:64) among other plant

and animal materials. Larger, predatory fishes such as

sharks (Lamniformes), barracudas (Family Sphyraenidae), and

jacks occasionally migrate inshore to feed in the

mangrove/grass bays.

Type 5 includes littoral zones of the barrier islands

(e.g., Sanibel, Captiva, Cayo Costa, and Gasparilla),

oceanic passes (e.g., Gasparilla, Boca Grande, Captiva,

Redfish, Blind), and open Gulf waters. Most fishes that are

primarily associated with these areas also frequent the

oceanic and estuarine bays. Examples are numerous sharks

(Hoese and Moore 1977:107-116; Larson 1980:81-95), jewfish


(Epinephelus itajara), sawfish (Pristis spp.), Florida

pompano (Trachinotus carolinus), large jacks, Spanish

mackerel (Scomberomorus maculatus), barracuda, and whiting

(Menticirrhus spp.) (Hoese and Moore 1977; Wang and Raney


Distribution of Invertebrates

Little systematic survey of aquatic invertebrates has

been undertaken in the Charlotte Harbor study area (see

Virnstein 1987:Figure 1). Based on comparative literature

and my own field observations, five zones were chosen to

examine invertebrates (primarily molluscs) along the

salinity gradient. These are (1) tidal stream; (2)

estuarine mangrove edge; (3) oyster bed; (4) seagrass

meadow; and, (5) littoral/Gulf. The classifications are

largely related to the limited mobility of aquatic molluscs.

As with the vertebrate categories, all types overlap,

creating a continuum of distribution.

Few marine invertebrates are known to venture far into

the tidal streams (Wells 1961:262) and these are highly

mobile animals that spend a small percentage of their life

cycle there. The blue crab (Callinectes sapidus), for

instance, travels upstream to the tidal-influenced marshes

where mating occurs, and returns to the estuarine bays and

later to the Gulf (Durako et al. 1985:250-251). Beds of the

marsh clam, Rangia cuneata, are associated with the Myakka

and Peace rivers (Woodburn 1965:6), as well as the


Caloosahatchee (Gunter and Hall 1969:63-64). Other than the

blue crab, the marsh clam, and mangrove prop root/mud flat

communities of small gastropods and bivalves, little is

known about invertebrates in upper tidal streams (Estevez et

al. 1984:CH160-CH163).

Type 2, for present purposes, is limited to areas of

the mangrove-fringed lower tidal streams and estuarine

locations. Molluscs commonly associated with mangrove prop

roots and adjacent intertidal muds include the eastern

oyster (Crassostrea virginica), Atlantic ribbed mussel

(Geukensia demissa granosissima), eastern white

slipper-shell (Crepidula plana), Gulf oyster drill

(Urosalpinx perrugata), scorched mussel (Brachidontes

exustus), worm-shell (Turritella spp.), crown conch

(Melongena corona), semiplicate dove-shell (Anachis

semiplicata), Atlantic bubble (Bulla striata), broad-ribbed

cardita (Carditamera floridana), coffee melampus (Melampus

coffeus), and several ceriths (Cerithium spp.) (Abbott 1974;

Odum et al. 1982:48-49). The mangrove tree crab (Aratus

pisonii) is an abundant resident.

Oyster bed communities (Type 3) are important and

frequent features in some parts of the Charlotte Harbor

estuarine system (Woodburn 1965). Turtle Bay, Bull Bay,

Matlacha Pass, and San Carlos Bay are examples of such

areas. The eastern oyster is well adapted to estuarine

situations, tolerating constant salinity fluctuations


(Butler 1954:479). It is most productive in mid- to

low-salinity estuarine waters because predators, such as

oyster drills (Urosalpinx spp.) and odostomes (e.g., Boonea

impressa), require somewhat saltier waters (Wells 1961:239,


Oyster bars support a large variety of fauna in and

among both live and dead shells by providing a hard and

protective substrate as well as a food resource. Community

profiles constructed in a North Carolina study by Wells

(1961:252) demonstrate varying species composition and a

decrease in diversity as one approaches fresh water. In

southwest Florida, the common crown conch is abundantly

associated with oyster bars. Experiments have shown that

this animal prefers salinities of 20 ppt and above but

tolerates 15.2 to 12.8 ppt for short periods (Hathaway and

Woodburn 1961:49). Other common organisms of the bar

community are the crested oyster (Ostrea equestris),

barnacles (Balanus spp.), scorched mussel, odostomes, boring

sponges (Cliona spp.), oyster drills, common jingle shell

(Anomia simplex), and slipper shells (Crepidula spp.)

(Butler 1954:486; Wells 1961:249-250; Southwest Florida

Project field observations). Migratory predators other than

the crown conch include whelks (Busycon spp.), black drum

(Pogonias cromis), stingrays, and blue crabs (Butler

1954:486; Carriker 1951; Galtsoff 1964:435, 439).


Shallow-water seagrass meadows, the fourth type,

provide extensive habitat areas for numerous mobile and

sessile molluscs. The abundance of invertebrates surpasses

even the fishes in areas of heavy shoal and turtle grasses

(Zieman 1982:49). Common gastropods include the lightning

whelk (Busycon contrarium), Say's pear whelk (Busycon

spiratum pyruloides), true tulip (Fasciolaria tulipa),

Florida horse conch (Pleuroploca gigantea), crown conch

(especially juveniles), fly-specked cerith (Cerithium

muscarum), dove shells (Anachis spp.), Atlantic modulus

(Modulus modulus), and lunar dove-shell (Mitrella lunata).

Bivalves such as southern quahog clam (Mercenaria

campechiensis), rigid pen shell (Atrina rigida), and

cross-barred venus are often embedded in large numbers in

the grass bottoms. Other organisms include pink shrimp

(Penaeus duorarum), corals (e.g., Manicimia areolata,

Porites furcata), hermit crabs (Pagurus spp.), and sea

urchins (e.g., Lytechinus variegatus, Tripneustes

ventricosus) (Zieman 1982:45-49). Although no studies are

known, it is presumed that seagrass invertebrate composition

varies along the salinity gradient in much the same manner

as the oyster bed community.

A number of species of invertebrates appear to be

restricted to the beach zone and Gulf waters. Others,

although preferring habitats in these areas, are also found

in the oceanic and estuarine bays. These two groups


comprise the fifth invertebrate category. Representative

animals include sunray venus (Macrocallista nimbosa),

southern surf clam (Spisula solidissima similis), stone crab

(Menippe mercenaria), sand dollars (Family Scutellidae), and

many small gastropods and bivalves (Abbott 1974; Wang and

Raney 1971:21).

It is emphasized that the foregoing vertebrate and

invertebrate divisions can usefully illustrate rough

segments of the salinity gradient. In reality, no species

restricts itself to these artificial types. Nonetheless,

the types allow an operable description of the continuum.

Future biological studies in Charlotte Harbor will improve

this brief descriptive distribution model.

Inferred Local Distribution of Resources in Prehistory

That present-day Charlotte Harbor is heterogeneous has

been established and its spatial variability conceptualized

in terms of abstract habitat categories. However, this

model cannot be projected directly into the past without

independent confirmation. If one assumes that the

prehistoric people targeted resources near their habitations

and that faunal evidence found at a site represents animals

processed or consumed at that site, then zooarchaeological

data can be used to test whether the spatial variability of

the present was also characteristic of the past.

Seventeen samples of archaeological fauna from five

sites Big Mound Key, Cash Mound, Useppa Island, Josslyn


Island, and Buck Key Shell Midden (Figure 1) were selected

for zooarchaeological study. The sites are located in

various parts of the greater Charlotte Harbor estuarine

system representing the area of greatest site density and

therefore do not cover the entire range of local

environmental settings. Valuable additions, for example,

would be faunal assemblages from sites located in Estero Bay

and at the mouths of and along the Caloosahatchee, Peace,

and Myakka rivers.

For modeling purposes, it is assumed that the samples

generally are from primary deposits and that they are

representative of site middens. The lack of intrasite

horizontal sampling need not be viewed as debilitating to a

study that serves as a regional baseline, one that is

subject to continual modification with each addition of new


Composite site data (i.e., combined level data within a

site, thus combining time periods), with the exception of

Useppa Island, provide the basis for zooarchaeological

spatial interpretation. The composite data sets are

presented in the text only in summarized form. However,

they are generated from raw data, all of which are presented

in Appendix A. Composite data for Big Mound Key, Cash

Mound, and Josslyn Island consist of four levels each. The

Useppa fauna is from only one level, A-4-2. The Buck Key

composite includes two Test B levels, B-2-5 and B-2-9.


Each present-day local setting of the five

archaeological sites is described below in concert with

summarized results of zooarchaeological analyses. It is

assumed that archaeofaunal data (Appendix A) represent the

faunal exploitation by prehistoric human occupants of each

site. For composite data sets, then, these

zooarchaeological data can be translated into inferred local

distribution of resources, summarized in Figure 7. The data

for Useppa are only tentatively offered as representative of

that site due to the availability of only one sample.

Big Mound Key. 8CH10

Big Mound Key today is situated at the mouth of Whidden

Creek, a stream that drains parts of the Cape Haze wetlands

(Figure 1). Patches of shallow seagrass (0.3 to 0.9 m depth

at mean low tide) occur among small mangrove islands to the

south and west in Gasparilla Sound. Oysters concentrate

around the many small mangrove islands in the sound and

adjacent bays (Woodburn 1965:24-25). Directly north of the

site is Boggess Hole, a large estuarine "pond." Farther

west are the barrier islands, Gasparilla and Little

Gasparilla, separated from each other by the shallow (0.3 m

deep at mean low water) Gasparilla Pass.

Fauna from each of these areas are represented in the

Big Mound Key archaeological samples (Appendix A, Tables A-l

through A-4; Figure 7). Cotton rat, raccoon, white-tailed

deer, and box turtle are all common to mangrove forests and


palmetto/pine forests. The presence of the greater siren,

snapping turtle (Chelydra serpentina), mud turtle

(Kinosternon spp.), and frogs suggests exploitation of a

freshwater environment. The ribbed mussel is the primary

mangrove edge mollusc, representing 15% of total Minimum

Number of Individuals (MNI) (Figure 7). These bivalves are

found today imbedded in swampy areas of black mangrove such

as on the western side of Big Mound Key. The

mangrove/seagrass habitat category is represented by 36% MNI

(Figure 7), largely consisting of three fishes pinfishh;

toadfish, Opsanus spp.; and killifish) and a host of

invertebrates (Tables A-l through A-4). Unlike other site

archaeofaunas, Big Mound Key contains a significant number

of the shark eye snail, Polinices duplicatus. The oyster

bed community contributes approximately 34% of the sample

(Figure 7). Finally, several vertebrate and invertebrate

species preferring oceanic waters (11%) are included in the


Cash Mound. 8CH38

Cash Mound is situated in Turtle Bay (Figure 1), an

area with water depths of 0.6 to 1.8 m at mean low tide and

rich in productive oyster beds (Woodburn 1965:23-24).

Seagrasses occur in the immediate vicinity and freshwater

marshes exist inland to the north.

The only terrestrial fauna recorded in the

archaeological samples (Tables A-5 through A-8) is raccoon


and an unidentified large mammal (presumably deer). The

oyster bed community constitutes 38% of the sample (Figure

7). Ribbed mussels, probably collected from intertidal

mangrove swamps, follow with 36%. Other mangrove edge

invertebrates occur, but in small numbers.

Hardhead catfish and pinfish, both common to

mangrove/seagrass areas, are the only fishes that occur in

abundance in the samples. The mangrove/seagrass habitat is

represented by only 9% MNI (Figure 7) of the faunal samples.

Cash Mound's faunal assemblage reflects a limited

exploitation strategy compared to the other four study


Useppa Island. 8LL51

Estuarine waters surrounding Useppa Island today vary

from 0.3 to 3.9 m deep at mean low tide. The area is

influenced to some degree by Boca Grande Pass (10 m at mean

low tide) but more by Captiva Pass (5.7 m at mean low tide)

due to the northward movement of currents in Pine Island

Sound (Figure 1). Seagrass and oyster habitats in the

vicinity have decreased in area due to modern human impact,

particularly the dredging of the Intracoastal Waterway.

Mangrove/seagrass and oyster habitats were heavily

exploited by Useppa's inhabitants of ca. 570 B.C. (Figure

7). The five most abundant fishes in the sample are

hardhead catfish, pinfish, pigfish, spotted seatrout, and

striped burrfish (Chilomycterus schoepfi), all common to the


mangrove/seagrass habitat (Table A-9). Oysters and their

associates are prominently represented with 46% MNI (Figure

7). The cross-barred venus is present in high numbers

compared to other site samples (Table A-9). The cotton rat,

white-tailed deer, and gopher tortoise also are present in

the archaeological sample.

Josslyn Island. 8LL32

Josslyn Island is located a short distance west of Pine

Island (Figure 1) and is surrounded by extensive and

extremely shallow beds of seagrass. Water depths are 0.3 to

0.6 m at mean low tide in all directions. This situation is

reflected in the faunal samples (Tables A-10 through A-13),

as these mangrove-fringed grass meadows are represented by

68% of the total MNI (Figure 7). Nine fishes are abundant

(more than 20 MNI). The top four fishes are pinfish,

pigfish, silver perch, and hardhead catfish. Josslyn

exhibits the greatest invertebrate diversity of all the

sites (composite total of 67 taxa). These results attest to

the high productivity of the seagrass habitat (Zieman


Although oysters and their associates comprise 19% of

the samples (Figure 7), today only one small oyster

community is observed in the Josslyn environs. Aquatic

birds such as red-breasted merganser (Mergus serrator), bay

ducks (Aythya spp.), and other ducks (Family Anatidae) favor

shallow seagrass meadows and also appear in the midden


fauna. Terrestrial areas such as mangrove forests,

marshlands, and palmetto/pine flatlands are represented by

the cotton rat, raccoon, white-tailed deer, warbler, box

turtle, and skink.

Buck Key Shell Midden. 8LL722

Buck Key is located to the east of and adjacent to

Captiva Island (Figure 1). Buck Key Shell Midden is on the

eastern shore of the island. Of the five study sites, it is

the one closest to the open Gulf, the southern portion of

the island presently bordering shallow Blind Pass (0.0 m at

mean low tide). Surrounding the island, water depths vary

from 0.1 to 2.1 m at mean low tide, and seagrass meadows lie

to the east and north. Also to the north are the deeper

ocean-influenced waters of Redfish Pass (2.1 to 10.0 m at

mean low tide).

Mangrove/seagrass fauna are predominant (62%) in the

archaeological samples (Tables A-14 through A-17; Figure 7).

The littoral/Gulf areas follow with 17% (Figure 7).

Hardhead catfish, sheepshead, silver perch, pinfish, and

striped burrfish, all common seagrass fishes, are abundant

in the midden samples. A random sample of Buck Key fish

vertebrae, relative to samples from Cash Mound and Josslyn

Island (Figure 8), reflects the proximity of Buck Key to an

ocean inlet during prehistoric occupation. There is broad

overlap in the three samples, however, a larger proportion

of the Buck Key measurements are over 3.5 mm. Because of


the geographic constriction of inlet waters, tidal cycling,

and daily movements of fish, the density of larger,

predatory fishes is greater at inlet locations.

Whelks, conchs, and tulips are abundant. Fishes and

molluscs with high-salinity preferences identified from the

midden include a host of sharks, gag grouper, red snapper

(Lutjanus campechanus), sea robin (Prionotus spp.),

barracuda, whiting, lettered olive (Oliva sayana), tellin

(Tellina spp.), coquina (Donax variabilis), stone crab, and

southern surf clam. Stone crabs favor the estuarine side of

oceanic passes as well as Gulf waters. Productive oyster

beds are not known in the immediate area today but small,

scattered beds have been noted on the east side of Buck Key

close to the mangrove shoreline.

Inferred Regional Distribution of Resources in Prehistory

Just as the present-day estuarine faunal distribution

can be modeled in terms of a gradient, so can the

heterogeneity observed in the zooarchaeological assemblages

described above. Establishing such a "zooarchaeological

gradient" involves two procedures. First, complete lists of

aquatic vertebrates (Appendix B, Table B-l) and

invertebrates (Table B-2) represented in the archaeofaunal

assemblages of Appendix A are compiled. The species are

then roughly seriated by their known preference for the

established habitat categories so that the listings in


Appendix B follow a salinity progression, or gradient, from

freshwater to oceanic water.

It is stressed again that the gradient concept treats

faunal distribution as a continuum in that it recognizes

great overlap in use of a variety of habitats by aquatic

fauna. An appropriate system of graphic symbols

representing known "preference" illustrates this point

(Appendix B). For example, sharks are depicted as generally

occurring in the inshore mangrove/seagrass habitats

("estuarine and oceanic mangrove" areas) as well as on the

Gulf shelf, but "prefer" the latter environment (Table B-l).

Figure 9 is a schematic illustration of the gradient

distribution based on the procedure just discussed and

presented (in detail) in Appendix B. The pattern is

informative. It clearly indicates (by the great overlap in

bars) that the estuarine and oceanic vertebrates

(predominantly fishes) represented by zooarchaeological

remains are highly mobile compared to the invertebrate fauna

(predominantly molluscs). The high mobility of fishes is

due to numerous factors including their free-swimming

nature, life-cycle behavior, daily salinity tolerances, and

feeding habits (Comp and Seaman 1985:359; Day et al.

1989:400-417; Lewis et al. 1985:307-309). Invertebrate

remains, as suggested by Figure 9, are even more

environmentally informative than fish because the animals


generally are not as mobile and often are restricted to very

specific salinity ranges along the gradient.

The second procedure of the gradient analysis presents

a zooarchaeological species distribution by site and

abundance. Appropriate symbols are used for variation in

abundance based on MNI (Appendix B). For this exercise

only, adjustments were made to the Useppa MNI (representing

only one column level as opposed to four levels for other

sites) to make them more comparable to the other site MNI

counts. The site distributions are illustrated along with

the habitat preference seriation (discussed above) for the

vertebrates (Table B-l) and the invertebrates (Table B-2).

From resulting site patterns it can be inferred from

the species distribution that all five study sites were

primarily associated with mangrove-fringed estuarine and

oceanic bay (including seagrass meadows) environments.

However, within this generalized pattern, intersite

differences based on habitat proximity (e.g., of marshes and

ocean inlets) and abundance (e.g., of seagrass meadow) can

be detected in the distribution patterns.

For example, some Big Mound Key fauna appear at the

low-salinity end of the gradient (see Table B-l), perhaps

due to the proximity of Cape Haze's freshwater marshes

rather than a riverine situation. This site sample also

contains fauna that suggest a high-salinity range and high

fish diversity (thirty-one species from the four samples),


reflecting the proximity of Gasparilla Pass (Figure 1; see

Wang and Raney 1971:54). The Cash Mound sample reflects a

setting of low- to mid-salinity based on the high level of

oyster exploitation and low diversity of fishes

(twenty-three species from the four samples).

Useppa and Josslyn islands fall into the mid- to

high-salinity range, with decreasing oyster beds and

increasing densities of seagrass meadow (Appendix B). These

site midden samples, particularly those of Josslyn, produced

the greatest abundance of seagrass fishes. A total of

thirty-one fish species was identified from the four Josslyn


The Buck Key faunal remains indicate the highest MNI of

animals from littoral/Gulf areas, placing Buck Key nearest

the high-salinity end of the estuarine scale. The

prehistoric ecological setting for Buck Key may have been

similar to the oceanic bay situation of Odum and colleagues

(1982:56), supporting a greater diversity of fishes than do

other environments. Of the five study sites, indeed, the

Buck Key faunal remains (B-2-5, Table A-14) produced the

highest number of taxa for both vertebrate (37) and

invertebrate (49) groups for any single sample. Looking at

Buck Key's four samples as a unit, a total of 40 fish

species was identified.

As a descriptive tool, a gradient analysis breaks down

a complex environment such as Charlotte Harbor into


understandable segments that archaeologists can relate to

prehistoric human adaptation. Salinity, used here roughly

to define those segments for Charlotte Harbor, is of course

only one of many variables determining faunal distribution

along an estuarine gradient. It is, however, perhaps the

most appropriate analytic factor for archaeological work

because for any given point location, the salinity regime is

reflected in zooarchaeological assemblages (particularly

true of molluscan remains).

Figure 6. Monthly Salinity Profiles of Four Aquatic

Locations in the Northern Part of the Charlotte Harbor

Estuarine Complex Illustrating the Fresh to Salt Water

Gradient (Data are after Wang and Raney 1971:18).

t */\.,, ./*
// \../
\ /i

\. /

/"'. /"

o.... ............... ....

./-- ." \ / '
/ \/"
/ \ .i
.\ /

-- Boca Grande Pass
-- Bokeelia
........... Charlotte Harbor
..-. Peace River


401 1--


~._ 20







Figure 7. Comparative Percentages of Zooarchaeological

MNI by Site Representing Exploited Habitats (Based

on Data Presented in Appendix A).




C, (

Big Mound Key ,
-- r 0 e

" ; / ^

Cash Mound S I

~"r "'" ":; -'5^


199 vD. S$ bl

Usoppa Island 6 \S ^ 6 r

//' "'X ^

S%0' 7-,

Josslyn Island aU



1 Mangrove/Seagrass Oyster Bed
Barnacle I Litloral/Gulf

0 I o .



Mangrove Edge

Figure 8. Thoracic Vertebrae Widths of Bony Fishes as an

Indicator of Overall Fish Size for Cash Mound, Josslyn

Island, and Buck Key.


: i ll~lllllll .. .. .._________


00 . IIlJ.,III I ui... .. ._.. I
10 -

ollI m.
0[ .l.lI III II IIIII. .... -- .... .. ..-..
1 2 3 4 5 6 7 8