REMOTE SENSING AND SOIL SCIENCE APPLICATIONS TO
UNDERSTANDING BELLE GLADE CULTURAL ADAPTATIONS
IN THE OKEECHOBEE BASIN
WILLIAM GRAY JOHNSON
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
William Gray Johnson
If not for the generous contributions of my friend and
confidant, Bob Ulshafer, this dissertation would not have been
accomplished. His camper provided a roof over my head while
I was in the field and his four-wheel drive vehicle got me out
of some rough country. Furthermore, his assistance in the
field was invaluable. Naming a site after him was a minor
reward for all his help. I only wish he had lived to see the
completion of this dissertation.
Jerald T. Milanich, my committee chairman, deserves
special praise as he was instrumental in focusing my attention
on the archaeology of the Lake Okeechobee Basin. His knowl-
edge of the area and his support for my project were outstand-
ing. The encouragement and support of committee members, Drs.
Mary Collins, Marvin Harris, William Keegan, William
Marquardt, and Michael Moseley, are also greatly appreciated.
Dr. Helen Armstrong, Chief Librarian of the Map and
Imagery Library on the University of Florida campus, was
especially helpful in assisting me with the aerial photograph
collections. Her generous contributions of time and knowledge
Support from Highlands Hammock State Park in the form of
free use of their facilities was greatly valued. Park
managers, Peter Anderson and Valinda Nichols, along with the
rest of the Park's personnel are thanked for their cooperation
Special thanks go to my informants for their time,
patience, and knowledge of the local geography. Jeff and
Debbie Clemens, Larry Luckey, John Pullen, Charlie Wilson, and
George and Peggy Wedgworth are among those who were especially
helpful. In addition, John Fitzpatrick and the staff at the
Archbold Biological Station are acknowledged for their role in
my documentation of the sites on Buck Island Ranch.
Finally, I want to thank the staff at the Florida
Department of State, Division of Historical Resources, for
believing in my project and the Florida Historic Preservation
Advisory Council for funding it. State Historic Preservation
Officer George Percy deserves the highest praise for his part
in securing project research funds.
TABLE OF CONTENTS
ABSTRACT . .
1 INTRODUCTION . . . . . . . . .
2 PREVIOUS RESEARCH AND THEORETICAL CONSIDERATIONS
Previous Research . . . . . . . .
Theoretical Perspectives . . . . . .
Theoretical Perspectives on Belle Glade Culture
3 THE BELLE GLADE CULTURE AREA . .
Environmental Background . . .
Culture History . . . . .
Summary of Cultural Adaptations
Settlements . . . . .
Archaeological Assemblage .
Major Sites . . . . .
Settlement Patterns . . .
Social Organization . . .
Belief Systems . . . .
Subsistence . . . . .
4 MAIZE AND SOILS IN SOUTH FLORIDA .
Precolumbian Maize in South Florida
The Antillean Route: A Not-so-Likely
The Soils at Fort Center: Not P
Maize . . . . . . .
Fort Center . . . . .
The Circular Ditches . . .
The Maize Pollen . . .
Evidence of Complexity at Fort
Testing Sears' Model: A
Application to Ar
Interpretation . . .
. . . . . viii
. . . . . . . . . 2 6
. . . . 28
Morphological Descriptions . . . .
Laboratory Analyses . . . . . .
Discussion . . . . .
Ethnohistoric Evidence: Conflicting Accounts .
Discussion . . . . . . . . .
Summary . . . . . . . . .
5 REMOTE SENSING, AND ARCHAEOLOGICAL FIELD AND
METHODS . . . . . . . .
Background and Literature Review . . .
Aerial Photographs . . . . . .. .
Pedestrian Survey and Informant Interviews .
Laboratory Analysis . . . . . .
6 RESULTS . . . . . . .
New Sites . . . . . .
Okeechobee County . . .
Highlands County . . .
Glades County . . . .
Hendry County . . . .
Updated Sites . . . . .
Highlands County . . .
Glades County . . . .
Hendry County . . . .
Negative Data . . . . .
7 DISCUSSION OF EARTHWORKS .
Mounds . . . . .
Ditches . . . .
Borrows . . . .
Embankments . . .
Summary . . . .
Mound Groups . .
Borrows . . .
8 SUMMARY . . . . . .
Belle Glade Archaeology . .
Remote Sensing . . . . .
Maize in South Florida . . .
Theoretical Considerations . .
Future Research . . . .
. . . . 87
. . . . 180
MORPHOLOGICAL CHARACTERISTICS OF SELECTED SOILS 191
REFERENCES . . . . . . . . . . 197
BIOGRAPHICAL SKETCH. . . . . . . .. 210
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
REMOTE SENSING AND SOIL SCIENCE APPLICATIONS TO
UNDERSTANDING BELLE GLADE CULTURAL ADAPTATIONS
IN THE OKEECHOBEE BASIN
William Gray Johnson
Chairperson: Jerald T. Milanich
Major Department: Anthropology
Remote sensing, in the form of early aerial photographs,
was utilized to locate and describe types of aboriginal
earthworks in the Okeechobee Basin. These data, combined with
a review of the relevant literature and the Florida Master
Site Files, provide the basis for a typological classification
of the earthworks. The typology is compared with chronologi-
cal information derived from the Fort Center site to provide
a synthesis of the culture history of the Basin. Data from
analysis of soils at the Fort Center site suggest that maize
played at best a minor role in the development of the Basin's
The archaeology of the Okeechobee Basin has been de-
scribed as "one of the most unknown and enigmatic in all of
Florida" (Milanich and Fairbanks 1980:181, see Figure 1) .
Although much has been done in the last decade (e.g., Hale
1984, 1989; Sears 1982), the nature of the aboriginal occupa-
tion of the Basin remains debated and, at least somewhat,
enigmatic. This is especially true of the extraordinary
earthworks found there.
This dissertation attempts to synthesize what is known
about the archaeological settlements in the region and provide
a framework for understanding the changes apparent in the
aboriginal earthworks associated with the Belle Glade culture,
the precolumbian and early colonial period peoples who
occupied the Basin. In addition, the question of maize
agriculture as a subsistence base for the Belle Glade people
is addressed through a study of soils at Fort Center.
Previous archaeological research in the Okeechobee Basin
has proved that precolumbian sites can be located by reviewing
aerial photographs for large earthworks, mounds and middens
(Carr 1975). However, this method can yield erroneous data if
not coupled with a pedestrian survey (cf. Clausen 1980).
The Okeechobee Basin and Key Sites (Source:
Milanich and Fairbanks 1980:168).
Thus, these methods were combined with informant interviews to
provide a comprehensive survey of this archaeologically little
known part of Florida.
Early black and white aerial photographs were reviewed in
order to find aboriginal sites within the project area.
Figure 2 provides an example. Photographs are available in
stereo pairs at the Map Library located on the University of
Florida campus. The earliest were taken in 1938 and provide
coverage of the area prior to the changes in vegetation that
resulted from extensive flood control. Large earthwork sites,
such as Fort Center, Tony's Mound, and others are easily
discernible. Mound sites, on the other hand, are not so
easily found, but can be identified with the aid of a stereo-
A ground-truthing pedestrian survey of the sites and
potential sites followed. Unfortunately, time constraints,
access to private properties, and limited manpower inhibited
my ability to ground-truth all of the sites and potential
sites found during the review of aerial photographs. Still,
the project resulted in the recording of 38 previously
unrecorded sites and provided new information for nine
previously recorded sites.
These data, combined with a review of relevant literature
and the Florida Master Site Files, provide the basis for a
discussion of the basic elements that make up the 40 known
earthworks in the Lake Okeechobee Basin, and their change from
Early Aerial Photograph (BUN-5D-133) of
Tony's Mound (8Hn3).
one form or type to another. This study provides a modifica-
tion of Belle Glade chronology and sheds light on Belle Glade
origins. Also, the application of early aerial photography to
remote sensing archaeological sites is addressed.
The question of maize agriculture in South Florida has
been debated for decades. Maize pollen from the Fort Center
site provides the most convincing evidence to date of its use
among the Belle Glade peoples. However, an examination of the
soils at Fort Center, specifically the purported planting
surface, indicates that maize could not be grown there in any
substantial quantity. Chapter 4 details the maize question
and provides evidence to support this statement.
PREVIOUS RESEARCH AND THEORETICAL CONSIDERATIONS
The foundations for our understanding of the Belle Glade
culture are attributable to John Goggin (1940, 1947, 1952,
n.d.a). His work on the Glades archaeological area incorpo-
rated research within the Okeechobee Basin as a sub-area.
Previously, Matthew Stirling (1936:355) designated the Glades
archaeological area as "the region between the Kissimmee and
Indian rivers and all of the peninsula from Lake Okeechobee to
the Florida keys." Stirling (1935, 1936) believed the entire
area was an archaeological unit that had been home for the
Calusa. Other than the exceptional quality of wood carvings
found by Cushing (1897) at the Key Marco site, he noted the
material culture of the area was "characterized by the use of
an inferior grade of pottery, perforated shell hoes, shell
plummets, antler adze sockets, and bone projectile points"
A.L. Kroeber (1939) considered South Florida to be a
distinct culture area in his study of cultural and natural
areas of North America. He too noted the quality of wood
carvings found by Cushing (1897) but, due to the lack of
additional finds, he characterized the culture as a poorer
phase of the Southeastern type.
Thus, Goggin's work in the area in the 1940s was, for all
practical purposes, a pioneering endeavor. His use of
Stirling's (1935, 1936) "Glades archaeological area" was
modified to include all of tropical Florida--with a northern
boundary stretching from Boca Grande Pass on the west coast to
just below Ft. Pierce on the east coast. Within this area, he
recognized three sub-areas: the Tekesta sub-area, the Calusa
sub-area, and the Okeechobee sub-area and three culture
periods: Glades I, II, and III (Goggin 1947:119-120).
At Fort Center, Goggin realized the distinctiveness of
the Belle Glade earthworks, noting that the site "has few
parallels" (1952:52). His analysis of the materials recovered
from Fort Center and from the nearby Platt site, coupled with
Rita Porter's (1952) analysis of Belle Glade Plain rim sherds
provided a ceramic sequence "remarkably like that from the
Belle Glade Midden" (Goggin 1952:65). In addition, he was the
first archaeologist to recognize the significance of aerial
photographs as a remote sensing technique for locating
archaeological sites in the Basin (Goggin 1952:52).
Goggin (n.d.a) also amassed a great deal of the ethno-
historic documentation on the South Florida natives. Writing
of the people living around Lake Okeechobee, Goggin (n.d.a:56)
states, "The affiliations of the people around Lake Okeechobee
are uncertain but it is possible that they were Calusa."
However, his article with William Sturtevant recognized them
as a distinct tribe called the Mayaimi (Goggin and Sturtevant
Gordon Willey's (1949) report on the excavations at the
Belle Glade type site and Big Mound City earthworks (conducted
by Matthew Stirling under the Works Projects Administration in
the 1930s) provides a cultural chronology specific to the
Basin. His scheme introduced the Belle Glade I and II
periods. Unfortunately, this scheme received little attention
from subsequent researchers.
William Sears' (1966:17-18) early work on culture areas
in South Florida distinguished the Okeechobee Basin from the
Glades archaeological area as a separate cultural region.
Later, Sears (1967) referred to the Okeechobee Area as the
Belle Glade area, a name change that Griffin et al. (1979:28)
found agreeable. Carr and Beriault (1984) prefer the term
Okeechobee Area but basically agree with the objective and are
not so concerned with the semantics. For the purposes of this
dissertation, Belle Glade area, Okeechobee Basin and Okee-
chobee Area are used interchangeably.
Outside the purely culture area designation arguments,
Sears' (1982) book on the excavations at Fort Center and his
articles (Sears 1971, 1977; Sears and Sears 1976) provide a
culture chronology distinct from previous work in the area.
His is a four-period chronology that separates cultural
developments at the Fort Center site. Sears (1982:193-201)
recognizes these developments from their relatively simple
beginnings to the later complex interactions between the Belle
Glade culture and the Calusa. Furthermore, he concludes that
terraforming and the presence of maize pollen at Fort Center
indicate that cultural development within the Basin was based
on maize agriculture.
The identification of precolumbian maize pollen at Fort
Center has led to vigorous debate among researchers interested
in the implications of maize agriculture among the South
Florida natives. Henry Dobyns (1983) uses this information in
combination with other lines of evidence to argue that
agriculture was practiced among the Calusa--a position that
enabled him to bolster his population density estimates for
the Calusa. Jerald Milanich (1987) responds to Dobyns
(principally because Dobyns argued that de Soto's landing was
actually south of Tampa Bay and in Charlotte Harbor) by taking
Dobyns' (1983) evidence to task and refuting each point.
However, Milanich (1987:180) admits that the Fort Center maize
pollen is clearly a major bit of evidence that cannot be
refuted. On the other hand, Milanich and Ruhl (1986) indicate
that maize may not have served as a dietary staple for the
Belle Glade culture.
William Keegan (1987) addresses the issue as it relates
to Sears' (1971, 1982) hypothesized route of movement for the
origins of maize at Fort Center. His research leads him to
believe that maize could not have been introduced to South
Florida via the Antilles until maize was already well estab-
lished in the Eastern Woodlands (circa A.D. 1000). However,
Donald Lathrap (1987) takes exception to Keegan's view, noting
that a preservation bias could exist in the archaeological
record especially if maize were used in its green corn form.
The question of maize agriculture among the Belle Glade
culture continues to generate much debate. In chapter 4 I
summarize current knowledge on this subject and present new
data from soil science applications to provide evidence that
favors Milanich and Ruhl's (1986) view.
Other research that addresses subsistence among the Belle
Glade culture includes Stephen Hale's (1989) dissertation and
a journal article (Hale 1984). Both focus on zooarchaeo-
logical analysis of faunal remains from Fort Center and other
sites within the Okeechobee Basin. Hale's research provides
baseline data on faunal resource utilization and addresses the
issue of maize agriculture. In addition, Hale provides an
analysis of Belle Glade settlement pattern and argues that
earthwork sites are in predictable locations.
Robert Carr's (1975) technical report on a survey of the
lake margin and a later article (Carr 1985) provide insights
on the circular earthworks of the area. Carr (1973, 1975,
1985) was instrumental in demonstrating the effectiveness of
early aerial photographs as a remote-sensing application to
study Belle Glade earthworks (although John Goggin used this
technique in the 1940s, as indicated by his notes on archaeo-
logical site file cards at the Florida Museum of Natural
History). Carr also provided some guidelines for research in
the Martin County portion of the Basin (Carr 1973).
In addition to these archaeologists, a number of re-
searchers working elsewhere have addressed issues related to
the Belle Glade culture area. John Griffin's (Griffin et al.
1982; Griffin 1988) work in Southeast Florida and the Ever-
glades and William Marquardt's (1986, 1988) and Randolph
Widmer's (1988) work in Southwest Florida have discussed the
relationships between these areas and the Belle Glade region.
Griffin et al. (1982) specifically consider the question
of maize agriculture among the Southeast Florida natives
because of Sears' (1982) evidence of maize use among the Belle
Glade peoples. They conclude that no evidence exists to
suggest maize was grown there (Griffin et al. 1982:230). They
do find evidence for trade between the peoples who inhabited
the Granada site and the Belle Glade culture in the form of
Belle Glade Plain pottery, but they cannot specify the items
that were traded from Southeast Florida to the Okeechobee
Basin (Griffin et al. 1982:392).
In his synthesis of the archaeology of Everglades
National Park, Griffin (1988:48-65) provides a comprehensive
overview of the spatial and chronological relationships
between the various culture areas of South Florida. In his
conclusions, he indicates the archaeology shows more affini-
ties with the Southeast coast than any other area (Griffin
Marquardt's (1986, 1988) examination of the Calusa social
formation indicates tributary relationships between the
various native groups of South Florida. Specifically, he
highlights the ethnohistoric documentation indicating that a
bread is made from roots in the interior of Florida and that
roots were brought to the Calusa on the Southwest Florida
coast. While the exact nature of the relationship between the
Calusa and the inhabitants of the Belle Glade area is not
clear, some form of distribution of goods from the Basin to
the coast is certainly suggested.
In The Evolution of the Calusa, Widmer (1988:274-276)
discusses the role of this bread in terms of demographic
distinctions between the Okeechobee Basin inhabitants and the
coastal Calusa population. Based on population estimates
provided by Fontaneda (1945), he determines at least a six-
fold difference in population size between the two groups. He
then argues that the demographic imbalance provided a military
advantage to the coastal Calusa that enabled them to force the
Belle Glade peoples to pay tribute for not being raided
These works provide a wide array of thought on the
archaeology of the Okeechobee Basin. In the next two sec-
tions, I examine theoretical perspectives as they have
developed in the discipline and as they have been applied to
the study of the Belle Glade culture.
The archaeologists who have worked in the Basin have
arrived at markedly different conclusions about the cultural
developments that took place. Two schools of thought are
evident: historical particularism and cultural ecology. A
short history of these theoretical orientations within a
framework of cultural evolutionism provides a backdrop for
discussing the conclusions that have been offered and suggest-
ing alternative strategies.
Cultural evolution has been a prominent topic of study
for anthropologists since the beginning of the discipline.
Probably the dominant form is one that has been labeled
cultural evolutionism. Leslie White (1959:vii) states that
the theory of cultural evolution dominated most of cultural
anthropology during the last three or four decades of the
nineteenth century. It was advanced by scientists to replace
the theory of creation as provided by Judeo-Christian theol-
ogy. At its most basic level, cultural development was seen
as a linear progression from simple to complex. Lewis Henry
Morgan's (1877) work provided the stage model for the evolu-
tion of human society. He recognized three distinct stages
that all human groups progress through: savagery, barbarism,
However, during the latter part of the nineteenth century
and the early part of the twentieth century, a reaction
against evolutionism developed in both the United States and
Germany. In the U.S., Franz Boas led the attack and instilled
the ideas of cultural relativity in his students. In Germany,
the nonevolutionist school of thought was led by Fritz
Graebner and his colleagues. Their work promoted diffusion as
an interpretation of similarities between cultures in noncon-
tiguous areas, rather than independent development.
Rathje and Schiffer (1980:302) state that "the attack
against cultural evolutionism was led by Franz Boas, who
helped establish the conceptual scheme we now call historical
particularism. This perspective identified several basic
historical processes that were used to explain specific
instances of change." Among the processes identified were
migration, diffusion, and rarely, invention.
Bower (1986:66) points out that the "historical partic-
ularist view rests on the assumption that cultural traits, or
customs, are 1) independent of one another; and, 2) present in
any given culture as a result of fortuitous contact with donor
cultures and the preferences of the recipients." Unfortunate-
ly, such assumptions require that any explanation fit within
the parameters of individual histories and do not allow for
broad generalizations. Thus, historical particularism has
largely fallen out of favor.
According to White (1959:viii), antievolutionism reached
its peak during the 1920s, but began subsiding and set the
stage for a comeback of cultural evolutionary theory in the
late 1940s. He quotes the News Bulletin of the American
Anthropological Association (1948) as stating that at the
annual meeting of the Association at Albuquerque in December,
1947, "a reawakening of interest in the problem of cultural
evolution . was noticeable." A number of articles
followed on the theory of cultural evolution (White 1959:viii
cites Jacobs 1948; Hoebel 1958; Lesser 1952; Steward 1953,
1955; and Childe 1951).
White (1959:ix) states that he "absorbed the antievo-
lutionist doctrines of the Boas school as a graduate student.
But as he began to teach, he found, first, that he could not
defend this point of view, and later that he could no longer
hold it. In due time he cultivated the theory of evolution in
a course called Evolution of Culture,. . and has attacked
the position of the antievolutionists in a number of arti-
cles." White (1959) then summarized mid-twentieth century
cultural evolutionist theory in The Evolution of Culture.
Since then, cultural evolutionism has provided the basis
for most archaeological and ethnological work on societal
complexity (Gailey and Patterson 1987:3). This model views
societies as entities that develop through a linear progres-
sion of stages. White (1959) believed societal complexity
could be viewed as a sequence of evolutionary stages that
could be documented historically and archaeologically. He
defined each stage in terms of four components--technological,
sociological, ideological, and sentimental or attitudinal--in
which the technological component was the basis and determi-
nant of cultural systems (White 1959:18).
Julian Steward (1951:380), on the other hand, suggested
the term "levels of sociocultural integration" rather than
stages of cultural evolution. He states, "the concept of
levels of integration does not presuppose any evolutionary
sequence." Rather, it is "a methodological tool for dealing
with cultures of different degrees of complexity." He
referred to his approach as a multilinear theory of evolution
that "seeks cross-cultural regularities and explanations but
presupposes no universal schemes" (Steward 1965:733).
According to Hatch (1973:114), "the key to Steward's
ideas about culture is his concept of ecology." Steward
(1955:30) states, "The principal meaning of ecology is
'adaptation to environment.'" Thus, cultural ecology is the
study of the relationship between culture and the natural
Steward's model distinguished between the culture core--
those features most closely related to subsistence and
economic activities--and the secondary features of culture
that are not strongly tied to the core but "are determined to
a greater extent by purely cultural-historical factors--by
random innovations or by diffusion" (1955:37). Thus, it was
the relationship between the core and the environment that was
the most important determinant of cultural systems.
Gailey and Patterson (1987:3) state, "Cultural evolution-
ism differentiated society from the natural world, and culture
. . mediated the relationship between them." Furthermore,
they note that parallel developments in positivist forms of
Marxist social theory began focusing on the use of resources
and technology as determinants of cultural development. Both
schools saw societies developing through a linear progression
of stages, and stages were defined in terms of variations in
the base and corresponding superstructural forms (Gailey and
However, problems in the positivist models began to
surface in the late 1950s (Gailey and Patterson 1987:3). Rowe
(1962) challenged the base-superstructure model. Leach (1954)
pointed out that linear models of development could not
account for the internal destruction of state institutions by
their subjects. And, Adams (1956) highlighted the reduction-
ist nature of state-formative models and believed they were
unable to account for the dynamics of historically specific
The results of these critiques led to the creation of two
distinct lines of thought: cultural materialism, as described
by Harris (1968, 1979), and a strand of evolutionist thought
which gave primacy to political dynamics (Fried 1967; Service
1962). The former retained the distinction between nature and
culture and continued to emphasize the primacy of the culture
core--what Harris (1979) calls the infrastructure. The
latter, on the other hand, has challenged the economic
determinism of the earlier evolutionists, but retained the
notion of linear development. The outcome of these lines of
thinking has been a slight revision in the sequence of stages,
i.e., to band, tribe, chiefdom, and state, as developed by
During this period, anthropological archaeology accepted
evolution as an integral part of its program, largely through
the efforts of Binford (1962, 1965). The so-called "new"
archaeology stressed an emphasis on explaining processual
change scientifically. Explicit goals, the identification of
cultural and natural formation processes, and a move away from
detailed descriptions of artifacts and artifact categories
were its hallmarks.
Systems theory, as promoted by Kent Flannery (1968, 1972)
was an outgrowth of this movement. Borrowing largely from
ecology, systems theorists viewed cultural change as a process
tending towards equilibrium. According to Gailey and Patter-
son (1987:4-5), it retained the idea of determination by the
political level and erased the distinction between nature and
culture. This model abstracted individuals from historical
and social conditions and conceived of them as components of
a system, interchangeable with each other and with those in
other systems. Transformation of these systems is viewed as
a product of factors other than human action--especially human
action as shaped by class, gender, ethnicity, and other
dimensions of social life. Thus, culture, as created by
humans in a social division of labor, is irrelevant to the
analysis (Gailey and Patterson 1987:5).
During the 1970s, a modified form of the systems model,
called catastrophe theory, was adopted (Gailey and Patterson
1987:5). This form was created for dealing with crises.
Colin Renfrew (1978, 1979) provided some of the best examples
of catastrophe theory arguing that crises or catastrophes
occur in the absence of continuous political and economic
expansion. Gailey and Patterson (1987:18) note that the model
appeared in Great Britain following a period of profound
At the same time, cultural materialism took on a more
important role in the development of anthropological theory.
Harris (1979) produced his seminal work on the subject leading
to vigorous debate on the primacy of the infrastructure in
determining cultural systems.
Interestingly, and almost simultaneously, two new schools
of thought emerged. Both sought to explain cultural develop-
ment from a Darwinian evolutionary perspective. These schools
of thought are sociobiology--as most profoundly stated by
Bateson (1985)--and cultural selectionism--as named by Rindos
Cultural selectionism argues for a Darwinian perspective
on the evolution of culture principally because it serves as
the basis for understanding evolutionary processes in modern
biology. Robert Dunnell (1980:36) states, "Modern evolution-
ary biology is relevant because it is in biology that the
scientific theory of evolution has been developed and because
biologists have argued its applicability to human phenomena
under the label of sociobiology" (e.g., Wilson 1975, 1978)
[italics original]. But perhaps the strongest argument for
supporting a Darwinian perspective on cultural evolution comes
from an approach that Dunnell (1980:36) has aptly named the
evolution of the "human capacity for culture." This approach,
while essentially biological, merges at its extremes with both
cultural evolution and sociobiology. Its basic tenets are to
be found in the assumption that the human capacity for culture
evolved within a biological context and, therefore, was
subjected to the same evolutionary forces that affect all
biological organisms. As such, this approach is necessarily
immersed in the Darwinian perspective.
One of the most recent advocates of the cultural selec-
tionists school of thought comes from Robert Boyd and Peter
Richerson (1985). Although trained as ecologists, they became
interested in the dearth of Darwinian models in the social
sciences. They noted the use of analogies from biology and
concepts such as adaptation, but were surprised by the lack of
any "systematic theoretical argument for cultural behavior
that paralleled the Darwinian theory of biologists" (Boyd and
Richerson 1985:vii). The result of their work, the dual
inheritance model, seeks to provide such a systematically
This model views culture as the transmission of traits
from one generation to the next via the mechanism of social
learning. In this view, culture is defined "in terms of
people's mental state (the "emic" concept of anthropology or
the "cognitive" of social psychology)" (Boyd and Richerson
1985:36). The authors believe "that it is important to
exclude behavior and the products of behavior from the
definition of culture because behavior is contingent upon both
patterns of thought and feeling and environmental circumstanc-
es" (Boyd and Richerson 1985:36).
The importance of excluding behavior and the products of
behavior, i.e., traditional definitions of culture in materi-
alist paradigms, from their definition of culture is to
provide a focus on social learning as a mechanism that
evolutionary forces can act upon. In their words, "Only by
distinguishing culture from behavior can we see clearly how
social learning interacts with environmental contingencies to
produce behavior" (Boyd and Richerson 1985:36). This then
provides the basis for their dual inheritance model. The
cultural repertoire is analogous to the genotype and social
learning provides the means to produce phenotypes. It is
argued that social learning has the properties of an inheri-
tance system and that random variation, drift, and natural
selection provide the mechanisms of cultural variability.
Cultural materialists likely feel comfortable with the
cultural selectionist school of thought. It, unlike sociobi-
ology, does not invoke genetic control over the development of
human cultural systems but does provide a strong metaphor from
the natural sciences for explaining social science phenomena.
However, unlike the ecology school of thought from which
cultural materialism grew, cultural selectionism argues that
cultural evolution is the result of a natural evolutionary
process (Rindos 1984:38). Where cultural ecology views
culture change as a result of non-equilibrium conditions,
cultural selectionism argues that all human behavior is
determined by the interaction of two inheritance systems--one
genetic and the other cultural.
Rindos (1984:23) points out that there are weaknesses
inherent in the cultural ecology school of thought, especially
in the branch known as systems theory. He finds that even
though cultural ecologists take a basic materialist view of
the development of agriculture the weaknesses in the paradigm
force proponents to argue in terms of intent and teleology.
Intent emerges as the ultimate cause in explaining
origins or evolution of systems because systems theory forces
explanation into a course of action requiring choices. Once
choice is invoked, intent becomes an inherent part of the
explanation. Teleology results when explanations within
systems theory allow us to place emphasis on equilibri-
um/stability, resulting in difficulty explaining any change,
or on change, providing us with a situation in which the
mechanism of change is not inherently specifiable; thus we can
choose one that best fits our prejudices (Rindos 1984:24).
Cultural selectionism, on the other hand, seeks to
identify functional or proximate modes of explanation and
separate them from evolutionary or ultimate modes of explana-
tion. The emphasis is placed on variability, heritability,
and differential fitness in explaining changes in learned
Evolutionary models such as cultural selectionism may
prove beneficial for cultural evolutionists. However, their
application at this time is very limited. Among researchers
of the Belle Glade culture, materialist strategies, especially
cultural ecology, dominate our current understanding.
However, historical particularism is also strong.
Theoretical Perspectives on Belle Glade Culture
Historical particularism and cultural ecology are the
dominant theoretical approaches to the study of the precolum-
bian history of the Lake Okeechobee Basin. These views lead
to markedly different conclusions about cultural developments.
In this section, the various researchers and their approaches
are contrasted to provide a focus for the approach taken in
Sears (1977:6-7, 1982:191) views the introduction of
maize agriculture and a settled village economy as having come
to South Florida's interior from Venezuela via the Antilles.
This introduction, according to Sears, comes in the form of
migration, the earliest of which was prior to 500 B.C. He
recognizes similar designs and similar manufacturing tech-
niques on the pottery of this time, as well as the terraform-
ing techniques that enabled savannah agriculture. As time
progressed, Sears believes that "the society here [at the Fort
Center site] developed a complex ceremonial center within a
century or two of the beginning of the Christian era"
(1977:7). By A.D. 1000, the ceremonial center was abandoned
and dwellings moved to small, single house mounds, each with
an adjacent linear earthwork up to 1200 ft long and 100 ft
wide. This, he believes, was due to continued contact with
the South American population (Sears 1977:7-9, 1982:200). To
further substantiate his claim of a South American origin for
cultural developments in South Florida, he cites Granberry
(1971; see also Granberry 1989 for a recent update) for the
origin of the Timucuan language (a language group that
dominated north-central Florida prior to European contact and
that is presumed to be similar to the language of South
Florida) from Venezuela.
Thus, for Sears, the migration of a South American
population via the Antillean route explains cultural develop-
ment in South Florida's interior. Such an explanation is
typical of a historical particularist strategy. It provides
a comprehensive explanation for the presence of complexity in
a given area without having to explain the reasons for the
development of that complexity in its original setting.
Bridging the historical particularist approach and the
cultural ecology school of thought is Robert Carr's (1975,
1985) research in the area. In his study of the area's
circular earthworks, he invokes a diffusionist argument for
their origin in Southeast Florida (Carr 1985:300). On the
other hand, he notes that all of the circular earthworks occur
in inundated environments that are located adjacent to
elevated hammocks or other upland environments. Thus, he
reasons that Sears' assertion that the earthworks functioned
for draining garden plots has to be rethought in terms of
other environmental contingencies (Carr 1985:299).
Gordon Willey's ecological approach focuses on a diet
based on wild foods (Willey 1949:71, 129). Moreover, he
believes that the environmental conditions of the area are not
suited for agriculture, especially in aboriginal times (Willey
1949:17). Indeed, he concludes "The area is not now, nor was
it in the past, well-suited for maize" (1949:18).
Stephen Hale (1984:183), who also takes an ecological
approach to the understanding of Belle Glade cultural adapta-
tions, has found that portions of the Belle Glade earthworks
are aligned with the direction of surface water flow and may
represent a method of drainage for agricultural production.
Though his work on the zooarchaeological materials from Fort
Center and other sites in the Basin provides abundant data on
the use of wild foods, he advocates an agricultural subsis-
tence for the Belle Glade peoples prior to A.D. 1545 (Hale
1984:183). Subsequently, Hale (1989:187-191) has retreated
from this position.
The question of maize among the South Florida natives has
been discussed at length. Most of these arguments center
around ethnohistoric documentation and/or archaeological
research. In all cases, the specific arguments can be said to
take a materialist approach [though Sears' (1971, 1982) work
on the subject is certainly couched in a historical particu-
The conclusions about cultural development are typical of
the approaches represented. Historical particularists will
invariably see diffusion as the source for culture change and
will often invoke migration theories to explain why and how
such change took place. Ecologically-minded researchers will
cite the high productivity of the natural environment and
assume the development of complex social phenomena as a
natural outcome. Such ecological approaches fall neatly in
the cultural materialist school of thought, which is basic to
the research approach reported here. However, the present
study departs from the strictly ecological school in its
approach to adaptation.
Adaptation appears in the title of this dissertation for
two reasons. First, it emphasizes the relationship between
culture and nature, and second, it draws attention to a
materialist approach. In the former, adaptation does not
refer to an equilibrium between cultural and natural systems
as it so often does in ecological models. Rather, it refers
to the juxtaposition of a cultural system in concert with a
natural system. Adaptation in this view becomes a methodolog-
ical tool for a materialist paradigm. The following chapters
are consistent with this position and provide results that
will change interpretations of the Belle Glade culture.
THE BELLE GLADE CULTURE AREA
The Okeechobee Basin geographical area is bounded by the
Western Flatlands (Davis 1943:43-48), the Allapattah Flats and
the Big Cypress Swamp (DeLorme 1987:102,113-114). According
to Brooks (1974:256), the area of the lake is dependent upon
water level and varies from 560 to 730 square miles (1450 -
1890 km2). The total area of the watershed (including the
lake surface) is over 4500 square miles (11,655 km2). Of
this, the Kissimmee River accounts for about 3000 square
miles. The basin itself, excluding the lake, can be said to
encompass between 3940 square miles (10,200 km2) and 3770
square miles (9760 km2) depending on water level of the lake.
If one excludes the Kissimmee River, then the area of the
basin drops to between 940 square miles (2430 km2) and 770
square miles (1990 km2).
However, there is good evidence to include Lake Kissimmee
and the Kissimmee River drainage in this cultural area.
Precolumbian cultural affinities represented in burial mounds,
earthworks, and ceramics are found as far north as Lake
Tohopekaliga (Milanich and Fairbanks 1980:26). Similarly,
both the Kissimmee River drainage and the Basin are character-
ized by relatively poor, sandy soils (composed of quartzitic
sands developed over calcareous marine deposits), relatively
low elevations with little relief, and numerous ponds and
sloughs. Outcrops of silicified limestone, or chert, which
were often exploited by aboriginal peoples as raw material
sources for the manufacture of stone tools, do not occur in
the area (cf. Lane 1980; Lane et al. 1980). The closest known
outcrops lie to the west along the Peace River (Upchurch et
al. 1982). The vegetation cover consists primarily of pine
and palmetto flatlands, wet prairies and occasional hammocks
of live oak usually mixed with cabbage palm. Cypress swamps
are found bordering larger lakes and ponds and a wide assort-
ment of wildlife exists in the area.
The region was first inhabited during at least the
Transitional period, c. 1000 B.C. 500 B.C. (Milanich and
Fairbanks 1980:23). Evidence for an earlier Archaic period
population is found nearby in the upper St. Johns "from north
of Palatka to well down opposite the Indian River region"
(Sears 1977:3). However, no sites in the Okeechobee Basin
have been assigned to this period. The earliest occupations
came from the Fort Center site (8G113), where semi-fiber
tempered ceramics have been found in limited numbers (Goggin
1952:58; Griffin 1952:329; Sears and Sears 1976:53; Sears
1982:26). According to Sears (1977:7; 1982:192-193), this
initial occupation is coeval with the construction of the
earliest of the Okeechobee Basin's spectacular earthworks--
i.e., the circular ditches.
Following the Transitional period is the Belle Glade
period, c. 500 B.C. A.D. 1700 (Milanich and Fairbanks
1980:23). During this period, cultural activities in the
Okeechobee Basin increase dramatically suggesting the greatest
cultural complexity in South Florida centered around this
inland territory (Milanich and Fairbanks 1980:181). Several
researchers have attempted to build a chronology for this area
but the lack of comprehensive surveys coupled with very
limited excavations leaves this region of Florida one of the
most poorly known regions of the state.
To date, the most comprehensive chronology for the Belle
Glade culture area was developed by Sears (1971, 1977, 1982)
based on his work at the Fort Center site. A combination of
stratigraphy, seriation, and radiocarbon dates allowed him to
develop a four-period chronology beginning with Period I and
ending with Period IV. These periods are roughly coeval with
the Transitional period as defined by Bullen (1959, 1970), and
the Belle Glade I and II periods as defined by Willey
Sears' (1982:26,185) Period I from c. 800 1000 B.C. to
A.D. 200 most closely fits the Transitional period definition.
At the early end of Period I, Sears (1982:192-193) notes small
populations of perhaps one or two families living along the
riverbank on small house mounds with fish and turtle being
particularly important to the diet. Only semi-fiber tempered
pottery was in use during this time but by the end of this
period most of the pottery was the sand-tempered plain ware.
Distinguishing this period from the Transitional period is the
presence of the circular ditches. Sears (1982:193-194)
believes the ditches functioned as drained fields for maize
Sears' (1982:26-31,186-189) Period II, from A.D. 200 to
about A.D. 600-800, is coeval with the Belle Glade I period,
except that there are no Glades decorated wares at Fort
Center. Griffin (1988:126) notes that sand-tempered plain
pottery is dominant at this time but is accompanied by the
appearance of Belle Glade Plain ware. The latter increases in
its popularity throughout Period II. Furthermore, trade wares
are present representing influence from Deptford, Crystal
River, Cartersville, Pasco and St. Johns areas to the north.
According to Willey (1949:125), evidence for the Belle
Glade I period consists of Belle Glade Plain and Glades Plain
(sand-tempered plain) ceramics with small percentages of
Glades decorated wares (see also Widmer 1988:87-88). There
are no defining dates associated with the Belle Glade I period
but, by inference, Widmer (1988:87) provides a terminal date
at A.D. 900 1000.
The Belle Glade II period is defined by the presence of
Belle Glade Plain and Glades Plain ceramics coupled with a
lack of Glades decorated wares and the appearance of St. Johns
(Biscayne) Check Stamped (Willey 1949:125). Sears'
(1982:27,31,189-190) Periods III and IV most closely resemble
this period. Period III lasts from A.D. 600 800 to A.D.
1200 1400 and is characterized as a period of time which
experienced minimal cultural change (Sears 1982:199). Griffin
(1988:126) indicates that sand-tempered plain pottery contin-
ued to decrease while Belle Glade Plain increased.
Period IV, on the other hand, is described as a time of
excelled craftsmanship, revived focus for authority, and
participation in a larger social system than the one existing
on the site. The introduction of linear embankments occurs in
this period as well as the first appearance of European
objects and artifacts made from European-derived materials.
In addition, Griffin (1988:126) notes that only a small amount
of sand-tempered plain ceramics accompany the large amount of
Belle Glade Plain ware. Furthermore, new rim forms, particu-
larly expanded flat and comma shaped varieties appear (Griffin
1988:126). Sears (1982:200) believes that during this period
Fort Center was part of the sixteenth and seventeenth-century
Little is known about the terminus of Belle Glade
culture. Few ethnohistoric accounts exist and only one is
widely publicized. Fontaneda's (1945:13) account of the
sixteenth century South Florida native groups refers to a
people who lived around a large lake known as the Mayaimi.
His narrative provides scanty information on population
density, subsistence, environment, and political affiliations
with the coastal Calusa. Some Belle Glade sites, including
Ortona (8G15), Belle Glade (8PB41), and Fort Center (8G113),
contain European artifacts and European-derived raw materials
such as silver, iron, and gold. Thus, there exists ample
evidence to suggest Belle Glade participation in a larger
economy than existed solely in the Okeechobee Basin.
The demise of the Belle Glade peoples was surely linked
to their participation in this larger economy. The introduc-
tion of European diseases is documented throughout the
southeastern U.S. and is considered a major cause in the death
of the South Florida native groups. It is thought by many
researchers that by the time the Seminoles entered into this
part of Florida, all of the South Florida aborigines were
Summary of Cultural Adaptations
Archaeological sites in the Okeechobee Basin include
earthworks (Allen 1948; Carr 1975:14-36; McGoun 1987, 1988;
Sears 1982:130-183; Willey 1949:73-77), burial mounds (Carr
1973:14; Sears 1982:130-183; Willey 1949:20-23; Williams
1975:15-34), and habitation mounds or middens (Goggin 1952;
Sears 1982:130-183; Willey 1949:19-20). Among the more
spectacular sites are the large earthworks found in the open
savannah, often bordering creeks or major environmental zones
(see Hale 1984:181). These include circular ditches, linear
embankments, and combined mound and embankment complexes.
Sears (1982:185) provides evidence indicating that the
circular ditches are the earliest earthworks at Fort Center,
dating earlier than 450 B.C. His research demonstrates that
some of the habitation middens are coeval with the ditches.
Habitation mounds are believed to have been built during his
Periods I and II (Sears 1982:186,195-199) and again during his
Period IV, but the latter include linear embankments (Sears
1982:199-201). Burials in mounds date to his Period I,
possibly II, and IV (Sears 1982:140-141,200-201). Willey
(1949:125) was unable to ascertain the date of the burial
mound at the Belle Glade site, but believed it was coeval with
the habitation mound. Williams (1975:33-34) obtained a
radiocarbon date of 200 B.C. under one of the burials he
excavated, but he believes the date is erroneous.
The combined mound and embankment complex known as Big
Mound City is thought to have been laid out according to a
master plan, although no conclusions have been made as to
whether it was built all at once (Willey 1949:73-77).
However, a collection from one of the mounds at the site
indicates it is coeval with the late period of occupation at
the Belle Glade site. At least 21 mound and embankment
complexes have been identified and include Barley Barber I
(8Mt19, Carr 1973); the Big Circle Mounds, also known as
Tony's Mound (8Hn3, Allen 1948); the Boynton Mound Complex
(8PB100 Jaffe 1976; McGoun 1987, 1988); Cowbone Prairie,
also known as Hendry Earthworks (8Hn25), L-8 Earthworks;
Lakeport Earthworks (8G126); Little Cypress Slough; Maple
Mound (8Hn5); Nicodemous Earthworks (8G19); Ortona (8G15);
South Mound City, also known as Big Gopher (Hale 1984:181);
Benchmark 24 Earthworks; Pepper Earthworks (Hale 1989); Fort
Center (8G113, Sears 1982); Kissimmee Circle Earthworks
(8G139); South Lake Mounds (8Hn33), and Summer Earthworks
(8Hn26). In addition, Johnson (1990) lists the Kissimmee
Circle (80b31), Lonesome Island (8Hg634), and Palmdale
Earthworks (8G176) as mound and embankment complexes.
Some of these sites have yielded European artifacts, such
as the Ortona Mound in Glades County (Goggin n.d.a:679),
indicating a colonial period occupation. Most, however, have
not been dated.
The defining ceramic assemblage for Belle Glade culture
is the presence of Belle Glade Plain and Glades Plain wares.
Lithic tools such as stemmed projectile points are found at
some of the sites, e.g. Fort Center (Steinen 1982:75-76) and
Belle Glade (Willey 1949:34, Plate 6), but not at others, e.g.
Barley Barber sites (8Mtl9, 28, 29, Carr 1973; Williams 1975).
Wooden artifacts have been found at two Belle Glade sites,
Fort Center (Sears 1982:38-58) and Belle Glade (Willey
1949:53-59), and with further research in the area may be
found at others. The currently limited distribution of wooden
artifacts precludes them from being considered part of the
defining archaeological assemblage. Similarly, other artifact
categories have been found to have a limited distribution.
At the Fort Center site, Steinen (1982:68-102) has
identified the nonceramic artifacts. These include eight
morphological categories of sharks' teeth, six types of bone
artifacts, chipped stone points and blades, ground stone
celts, hones, and grinding stones, shell and stone plummets,
shell adzes, celts, axes, and gouges, and chert and quartz
The morphological categories of sharks' teeth represent
two classes of tools. The first is triangular-shaped,
asymmetrical, serrated, and has bilateral wear patterns, and
the second is tapering, usually symmetrical, and lacks cusps.
Similarly, the six categories of bone artifacts represent only
five types of tools. Of the five, only one has been classi-
fied into a functional type: fids.
Comparable nonceramic assemblages are known from only two
other sites in the Okeechobee Basin: the Belle Glade type site
(Willey 1949:34-53) and the Platt site (Goggin 1952:55).
Willey (1949:69) reports socketed and bi-pointed bone projec-
tile points, bone pins, shell picks, adzes, and hammers, shell
celts, conch shell drinking cups, shell and stone plummets,
stone pipes and tubes, and chipped stone projectile points
were recovered from Belle Glade. In addition, fossil shark
teeth, porpoise teeth, and possible bear teeth are in the
Belle Glade site assemblage (Willey 1949:45). At the Platt
site, Goggin (1952:55) found chipped stone projectile points,
shell picks, gouges, and celts, shell and stone plummets, bone
points and pins, and an antler adze socket. However, none
were found in sufficient quantity to indicate significant
temporal differences (Goggin 1952:55).
A recent paper on some of these artifact types, specifi-
cally the bone points and shell and stone plummets, indicate
fishing technology (Walker 1989). Long considered representa-
tive of terrestrial technology and/or ceremonial regalia,
Walker (1989) presents a convincing argument that these
artifact types represent two parts of a composite fish hook.
The importance of her argument for South Florida artifact
assemblages is to focus attention on fishing technology.
Chronological distinctions in Belle Glade ceramics have
been hampered by a lack of decorative motifs on surfaces.
Virtually all Belle Glade wares are plain. However, Sears
(1982:112) provides a seriation based on relative amounts of
plain ceramics. Semi-fiber tempered wares are earliest and
are gradually replaced by sand tempered wares. St. Johns
Plain pottery appears abruptly and relatively abundant at an
early date, but dissipates rapidly with the introduction of
Belle Glade Plain. Minor amounts of St. Johns Plain continue
to be used throughout the occupation at Fort Center. Belle
Glade Plain ceramics are initially a minor component of the
total ceramic assemblage, but become the dominant type at the
expense of sand tempered wares.
In addition to Sears' (1982:112) seriation of ceramic
types, he indicates that some technological distinctions are
related to chronological differences. Specifically, he
provides a seriation based on rim forms that shows flat lip
rims are the dominant form throughout much of the occupation
at Fort Center (Sears 1982:113). These are superseded by
expanded flat rims late in the occupation. This work appar-
ently builds on Porter's (1952) analysis of rim forms from the
The presence of monitor or platform type ceramic pipes at
Fort Center are thought to provide a good chronological
indicator because the pipes are said to be very similar to
those from Marksville sites (Sears 1982:117), known to be a
Hopewellian culture dating from Middle Woodland times. Sears'
(1982:117) associated radiocarbon dates place the pipes in his
One other major category of Belle Glade artifacts that
serve as good chronological indicators are the artifacts
derived from European contact. These have been discussed by
Willey (1949), Sears (1982), and Leader (1985). Their
usefulness as indicators of sixteenth and seventeenth century
period activities has not gone unnoticed, but their origin and
form have been the focus of most analyses.
At the Belle Glade site, Willey (1949:59-61) identified
gold, silver, and copper beads, an iron spike and iron
fragments, a lead plummet, four types of glass beads, and
European-made ceramics. Sears' (1982:59-67) work with the
collection from Fort Center categorized the metal artifacts as
symbol badges, disc-shaped ornaments, a cast gold jaguar,
silver and copper plates, and miscellaneous gold and silver
pieces. Sears (1982:59) states, "It is apparent that the
metal specimens were made from metal of Spanish origin,
reworked in most cases into Indian forms." Leader's (1985)
work on the Fort Center collection focused on this statement.
In his quest to discern the origins of these materials he
grouped the artifacts by morphological characteristics into
the following categories: disks, domes, beads, hemisphere and
cones, plaques, and miscellaneous fragments. His conclusions
confirm Sears' (1982:59) statement (but noted that some of the
artifacts were actually made by South and, perhaps, Middle
American natives; the Spaniards were evidently shipping the
artifacts with other booty to Spain when the materials were
salvaged from shipwrecks by the Florida natives.)
Of all the artifacts made from European-derived materials
the symbol badges have drawn the most attention. The Fort
Center symbol badges are made of heavy slabs or ingots of
silver, 1/16 inch thick, cut into tablet forms and etched with
native designs. Sears (1982:60) states they are larger and
heavier than others found in South Florida.
Allerton et al. (1984) have summarized the distribution
of these badges and have found they are concentrated around
Lake Okeechobee. Leader (1985:81) states, "the ethnic origin
of the metal artifacts from the Fort Center metal collection
is American aborigine." Unfortunately, activity areas of
manufacture have not been identified. As such, no one has
been able to ascertain the culture group. However, their
known distribution clustering in the Lake Okeechobee Basin and
the largest and heaviest coming from Fort Center strongly
suggests they are Belle Glade artifacts and should be consid-
ered part of the defining archaeological assemblage.
Major sites in the Belle Glade culture area have been
identified by Carr (1975). His survey is one of the most
comprehensive listings of archaeological and historic sites in
the Lake Okeechobee Basin, but it was limited to areas that
were to be inundated by the Army Corps of Engineers proposal
to raise the regulation range of Lake Okeechobee from 13.5 -
15.5 feet above mean sea level to 15.5 17.5 feet. His
survey lists 16 precolumbian sites including mounds, middens,
An even more comprehensive listing of precolumbian sites
in the Okeechobee Basin can be found on a map provided by Hale
(1984:181). However, even his figure does not include
earthworks such as Barley Barber I (8Mt19, Carr 1973:9-11),
West Okeechobee Circle (8G157, Carr 1985:295-297), and others.
As such, there is not a complete list of all major sites in
the Belle Glade culture area. To date, eight circular
earthworks, 17 circular-linear earthworks, nine linear
embankment sites, and six miscellaneous earthwork sites are
known to be in the Okeechobee Basin. Of these, major excava-
tions have been limited to the Belle Glade and Big Mound City
sites (both in the 1930s, see Willey 1949) and the Fort Center
site (in the 1960s, see Sears 1971, 1977, 1982).
As for settlement practices, Sears (1982:184-190)
indicates a Belle Glade cultural adaptation to the open
savannah with artificially raised areas occupied during
flooding. The focus of occupation takes place at large sites
(such as Fort Center), though Sears (1982:175) acknowledges
occupation of "the small midden sites on higher land through-
out the sand country." On the other hand, Willey finds Belle
Glade affinities extending from the Okeechobee Basin to all of
South Florida, from the Florida Keys (Willey 1949:26 cites
Goggin n.d.a) to Melbourne (Ferguson 1951:36-39) and the
southwest coast (Willey 1949:124-125). However, he believes
the Belle Glade peoples differ significantly from the east
coast groups and states, "the most likely guess would be that
Belle Glade was a village of the Calusa domain, presumably
occupied by a people related to them" (Willey 1949:127).
Thus, his scenario depicts a wider range of adaptation to a
number of environments than does Sears.
A discussion of social organization provided by Sears
(1982:1919-201) indicates that an egalitarian population
inhabited the Fort Center site during his Period I. This was
followed by a marked degree of social stratification in
subsequent periods. Trade with other Florida groups is
apparent by the number of exotic artifacts he found. Further-
more, the presence of maize pollen and the earthworks suggests
to him that the Fort Center inhabitants had periodic contact
with a South American population (and, indeed, trace their
ancestry to this population). Inclusion of the Fort Center
peoples within the sphere of colonial period Calusa hegemony
is believed to have occurred in the final aboriginal occupa-
tion of the Fort Center site, i.e. his Period IV.
Goggin and Sturtevant (1964:180-182, 195-197) suggest
Calusa hegemony occurs early in the cultural development of
the South Florida native groups. Indeed, they hint at the
possibility of the Belle Glade peoples representing an inland
population of the Calusa culture. However, theirs appears to
be a minority view.
Only a minimal discussion of beliefs has been presented.
Sears (1982:42,58) suggests that the Fort Center inhabitants
had strong totemic beliefs (as represented by carved wooden
figures of animals and birds). He also argues for a belief in
an afterlife (as represented by burial practices). Such
beliefs would not be surprising given the material remains
found in the Belle Glade culture area.
Based on what is known from the above, the diet of the
Belle Glade peoples has been interpreted to be either one
based on wild foods (Willey 1949:71,129) or one based on maize
agriculture supplemented with an abundance of wild foods,
especially turtle (Sears 1982:184-201). However, Walker's
(1989) recent analyses of fishing technology has alerted the
archaeological community to the potential contribution from
aquatic resources; an area of subsistence activities that
needs further exploration in the Okeechobee Basin. Hale
(1984:183) concurred with Sears (1982) noting that the
planting surfaces he identified are aligned with the surface
water flow facilitating drainage for agricultural production.
However, Hale (1989) has subsequently retreated from this
position. Analysis of the soils of one of the alleged
planting surfaces (reported hence) indicates it could not
support annual harvests of maize. In the following chapter I
explore the question of maize in detail.
MAIZE AND SOILS IN SOUTH FLORIDA
Maize was obviously an important cultigen in the develop-
ment of certain New World cultures. Richard MacNeish (1970)
studied the evolution of this cultigen in the Tehuacdn Valley
of Mexico and concluded that settlement changes coincided with
increased reliance on maize. According to MacNeish
(1972:500), the use of more productive hybrids of maize
coupled with increased interactions between developing areas
caused changes in settlements rather than the reverse.
Similarly, Sears (1982) believes maize was an important basis
for cultural change within the Lake Okeechobee Basin of South
Florida. However, the presence of maize in South Florida, has
led to much debate over its use, importance, origins, and
Precolumbian Maize in South Florida
Evidence for precolumbian maize use in South Florida is
limited to maize pollen documented at the Fort Center site by
Sears and Sears (1976; see also E. Sears 1982). At the site,
Sears found numerous pollen grains in a variety of contexts
including at least one excellent precolumbian context dating
between A.D. 100 and 500. Strips of white pigment adhering to
a wooden carving that was found at the bottom of a pond
provided this context (Sears 1982:122; Milanich 1987:177).
Other contexts from which maize pollen was found are soil
samples from earthworks and coprolites. On the basis of size
and exine pattern, Sears and Sears (1976:54) distinguish the
maize pollen grains from other grasses. In addition, they
believe the size range of the maize pollen grains from Fort
Center eliminates the possibility that they could be modern
In addition to the maize pollen, Sears (1982:171-173)
identified lime deposits at the Fort Center site that were
created by burning shell. He hypothesizes that the creation
of the lime was directly related to the processing of dried
maize into a mush (Sears 1982:173).
Sears (1982) believes the earthworks at Fort Center
served as planting surfaces for maize cultivation. The
origins of the maize and the earthworks are believed to be the
result of a migration of peoples from South America. Sears
(1977:6-7, 1982:191) suggests four possible routes for such
contact, but feels the most likely one was from Venezuela
through the Antilles (Figure 3). He notes similar ceramic
manufacturing techniques and stylistic elements on pottery
between South Florida and Venezuela, as well as similarity
between terraforming techniques (presumably related to
agriculture). His excavations at the Fort Center site
indicate the development of complex mortuary activities within
Hypothesized routes for maize origins in
South Florida (after Sears 1971).
a century or two of the beginning of the Christian era
(1977:7). By A.D. 1000, the mortuary center was abandoned and
individual houses were placed on small, single house mounds,
each with an adjacent linear earthwork up to 1200 feet long
and 100 feet wide. This innovation, Sears believes, was due
to continued contact with the South American population
Thus, for Sears, the migration of a South American
population with the knowledge of maize cultivation in wet
savannah environments explains cultural development in South
Florida's interior. This interpretation provides a comprehen-
sive explanation for the origin of the maize and the earth-
works but has left a myriad of questions in its wake.
The Antillean Route: A Not-so-Likely Course
Perhaps the most difficult problem with Sears' chronology
and origins of maize in South Florida is its early arrival, as
much as 1000 years prior to its presence in the Eastern
Woodlands. William Keegan (1987) addresses this issue and
concludes that present evidence indicates the introduction of
maize into South Florida via the Antilles could not have
occurred until maize was already well established in the
Eastern Woodlands (circa A.D. 1000).
According to Keegan (1987:331), human groups arrived in
the Greater Antilles by 4000 B.C. These peoples were aceram-
ic, non-agricultural hunter-gatherers who exploited terrestri-
al and marine resources. Their way of life changed with the
arrival of the agricultural Arawakan-speaking population. The
Arawakan-speakers, also known as the Tainos, began their
expansion into the Antilles from eastern Venezuela around 200-
300 B.C. (Haviser 1989). Their arrival into the northernmost
parts of the Antilles, i.e. the Bahamas and Cuba, was not
until A.D. 600 700. As with other parts of the Antilles,
the initial colonization period diet included cultivated roots
and tubers with terrestrial animals and a relatively minor
contribution from marine resources. Shortly after coloniza-
tion, a rapid shift to increased reliance on marine resources
occurred but root crops remained the dietary staple and maize
did not appear until late in precolumbian times (Keegan
Keegan (1987:336) concludes that the available evidence
indicates that the Antilles was not the source of this
cultigen in South Florida. However, he points out that this
conclusion concerns only one of four possible contact avenues
hypothesized by Sears (1977) for the arrival of maize in South
Florida and that little is known of the precolumbian history
of western Cuba. That region may yet be implicated in "a more
direct trans-Caribbean jump from Central or South America to
south Florida" (Keegan 1987:339).
The Soils at Fort Center: Not Productive for Maize
Regardless of Keegan's (1987:339) assertion that alterna-
tive routes of entry remain unexplored (allowing for the
possibility of a direct entrance of maize into South Florida),
a problem closer to the location of the pollen has been
discovered. Analysis of the soils from one of the circular
earthworks at Fort Center has yielded information on their
potential to support annual harvests of maize.
The Fort Center archaeological site is located in Glades
County, Florida (Figure 4). It was first reported by a
professional archaeologist when Goggin (1952) investigated it
and the nearby Platt site. At Fort Center, Goggin recorded
four middens, three small mounds, a large sand mound and its
associated platform, a large circular canal or ditch, a ridge
with refuse on its surface, and a restricted concentration of
nineteenth century artifacts (Goggin 1952:50). The latter
presumably mark the location of Fort Center, a small block
house established at the site during the Seminole wars. Also
recorded were several pairs of earth ridges, two with terminal
mounds, another small mound, and a long narrow ridge in the
prairie to the east. Goggin and Sturtevant (1964:197) suggest
that these earthworks, and similar ones elsewhere in the Lake
Okeechobee Basin, were built by the coastal-dwelling Calusa as
Location of Fort
Highway Map 1984).
Center (Source: General
Sears (1971, 1974, 1977, 1982), on the other hand,
suggests that the original inhabitants of the Lake Okeechobee
Basin (the Belle Glade culture) migrated from the northern
part of South America sometime before 500 B.C. According to
Sears (1982:192), continued contact with the mother population
resulted in the adoption of a variety of economic behaviors,
one of which was the practice of draining fields for producing
maize. He states "the economic system supporting the center
. acquired stability through dependable maize crops"
(1982:197). Sears believes this economic system was accom-
plished in the wet savannah environment by the creation of
planting surfaces, formed by the excavation of ditches, for
growing a stable, dependable maize crop. Initially the
ditches served to drain fields but later they were designed to
create raised fields. Hale (1984:183) suggests that this was
a response to rising water levels in the Basin. According to
Sears (1982:178,191-201), such a system sustained the economy
of these people for over a two thousand year period.
His evidence for this belief is twofold. First is the
presence of the circular and linear embankments at the Fort
Center site (Figure 5) and second, the identification of maize
pollen grains from a number of contexts at the site (E. Sears
1982:118-129), including at least one grain from a soil sample
from the Great Circle ditch. The former suggests to Sears the
ability of the inhabitants to have practiced drained and
raised field agriculture, while the latter provides evidence
Schematic drawing of the Fort Center site
(after Carr 1985:290).
of the presence of maize. Combined with other evidence
regarding the social complexity of the precolumbian inhabit-
ants, these data form the basis for Sears' (1982:178, 191-201)
conclusions that the earthworks were used as agricultural
The Circular Ditches
Among the various earthworks at the Fort Center site are
two remnant circular ditches each of which have a diameter
approximately 140 m across. These are enclosed by a much
larger circular ditch referred to as the Great Circle (Sears
1982:175). Excavations revealed that the remnant ditches were
built prior to the Great Circle (Milanich 1968), but it is not
known how much earlier, nor which of them was dug first.
Radiocarbon dating of midden materials found in the base of
the Great Circle ditch date to 450 B.C. + 105 years (Sears
1982:116). Thus, Sears (1982:178) states all three circular
ditches were completed by then.
Sears (1982:6) believes these ditches functioned as
drainage ditches to enable food production. The basis for
this belief comes from his observation that "ditch bases were
dug below the hardpan so that water would drain across the
hardpan and into the ditch, drying the surrounding area"
(1982:6). He finds additional support for this belief in "the
need for drained garden plots if agriculture was to be
practiced in this environment and by the presence of maize
pollen in soil samples from points inside the ditches and the
[nearby] contemporaneous midden" (Sears 1982:186).
The Maize Pollen
Evidence for maize in the diet consists of maize pollen
grains found in coprolites, white pigment samples applied to
wooden carvings, and soil samples from several of the earth-
works at Fort Center (E. Sears 1982:120). The maize pollen
grains, specifically those from coprolites and the pigment,
lead Milanich (1987:177) to conclude "that maize was being
cultivated at Fort Center sometime during the period ca. A.D.
100-500." However, Milanich and Ruhl (1986:2) state that the
importance of maize within the Belle Glade people's diet is
uncertain. They conjecture that it might have been grown and
used for special purposes rather than as a dietary staple.
Evidence of Complexity at Fort Center
Sears' excavations documented numerous artifacts, long
distance trade, increased population through time, extensive
earthworks, and complex ceremonial activity areas. These
features suggest a fairly high degree of complexity when
compared to the contemporary cultures of the rest of South
Florida (Milanich and Fairbanks 1980:181).
Among the artifacts found at the Fort Center site are a
number of intricately carved wooden objects, including a
series of animal figures which formed an elaborate burial
platform. Similar wooden objects have been documented by
Willey (1949) at the Belle Glade site, also located in the
Lake Okeechobee Basin. Additionally, numerous wooden carvings
were uncovered by Cushing (1895, 1897) on the southwest coast
of Florida, but Sears (1982:58) believes they do not have any
real resemblance or relationship to the wooden objects from
Fort Center [but see Schwehm (1983) for an alternative
explanation Nevertheless, the craftsmanship of all the
carvings appears to reflect a degree of craft specialization.
Evidence for long distance trade includes the presence of
non-local artifacts including chipped stone tools, Deptford,
Crystal River, Cartersville, Pasco and St. Johns ceramic
types, marine shell tools, sharks' teeth and European-derived
items. All indicate participation in a wider economy than
just the Lake Okeechobee Basin. Most of these materials
indicate trade with coastal dwelling populations. Sears
(1982:32) suggests "at least fringe participation in the
Hopewell phenomenon" because of the presence of ceramic types
which co-occur with the strictly ceremonial Yent complex.
Sears (1982:32) also lists Deptford series sherds, Deptford
copies on chalky paste, and Crystal River series specimens as
evidence of the Yent complex at Fort Center. Milanich and
Fairbanks (1980:84) agree that similarities with materials of
the Yent complex suggests trade between the Woodland peoples
to the north (e.g., Adena, Hopewell, Cartersville, Copena,
early Weeden Island) and the South Florida peoples. However,
one must keep in mind that Sears (1962) defined the Yent
complex and may be biased in his interpretations.
Increases in population through time are also apparent at
the site, according to Sears (1982:193). He believes "that
there were not more than one or two families on the site at
any one time" during his Period I (approximately 1000 B.C. to
A.D. 200) but that number increases up to three-fold during
his Period II (Sears 1982:193-195). Moreover, Sears
(1982:197) suggests that this increase coincides with the
participation of additional peoples in the economic activities
at Fort Center. Numerous small sites throughout the savannah
region were occupied by peoples who, according to Sears
(1982:197), were supported by the economic system at the Fort
Center site. Finally, population density at the site drops
back to one or two families during the last period of occupa-
tion, Period IV (Sears 1982:200). However, Sears
(1982:200-201) believes that the inhabitants at that time were
part of the sixteenth- and seventeenth-century Calusa domain.
Thus, a considerable increase in the population which partici-
pated in the Fort Center economy can be discerned through
While most population estimates were derived by midden
deposits on house mounds, the burial population of the Mounds
A and B and charnel pond complex provided reinforcement for
Sears' estimates (Sears 1982:175). This mortuary area was
treated as a single unit by Sears because "one could not be
investigated without disturbing at least one other" (Sears
1982:147). Furthermore, an earthen embankment surrounding
Mound B "continues on at both ends to Mound A, thereby tying
both mounds and the pond into a unit" (Sears 1982:148). Thus,
Sears (1982:147-148) believes it is important to treat the
area as a single unit.
Mound A, the shorter of the two, yielded midden materi-
als, indications of two to four dwelling units, ceremonial
artifacts, and lime deposits. The midden materials contained
the usual assortment of faunal remains [although Hale
(1984:177) notes a larger amount of deer bone than found at a
similar deposit elsewhere at the site] and ceramic sherds but
also included shell tools, pipes, some stone grinding tools,
and a few other items with no "meaningful clusters or associa-
tions" (Sears 1982:171). The lime deposits, created from
burned shells, and the pipes, limited to this one mound (save
for the few fragments found at the UF Mound and Mound 3, Sears
1982:34), led Sears (1982:175) to believe that the people
living on Mound A were ceremonial specialists. Their services
"providing lime for treatment of their corn and probably pipes
and tobacco," are thought to have been performed for the
people who "lived elsewhere, in the small middens sites on
higher land throughout the sand country" (Sears 1982:175).
Mound B and the charnel pond, on the other hand, served
principally for mortuary activities. Approximately 150
burials were recovered from the pond along with numerous
carved wood figures, midden materials, and coprolites. The
wood figures are thought to be the remains of a charnel
platform that collapsed after a fire destroyed its structural
integrity (Sears 1982:167). Another 150 burials, many with
grave goods, and mud-streaked sand were documented in Mound B.
Sears (1982:157) believes the mud-streaked sand occurred when
skeletal remains were retrieved from the pond, after the
charnel platform collapsed, and placed in Mound B.
The extensive earthworks and complex ceremonial activity
areas documented at the site attest to a highly organized
population. The circular and linear ditches accompanied by
mounds, middens, and a charnel pond provide ample evidence of
large-scale work projects that could only be accomplished by
an organized labor force. Thus, it is not surprising that
Sears' discovery of maize pollen at the site led him to
believe that the inhabitants could only have developed such a
highly complex society on the stability of maize cultivation.
However, Feldman (1980:4) cautions that "once a trait has been
found in the archaeological record, any number of additional
features are often assumed to have been present from that date
forward. For example, the presence of maize at a site--
however small the amount recovered--is often assumed to
indicate that the subsistence of the site was based on
efficient maize farming."
Testing Sears' Model: A Soil Science Application to Archaeo-
Soil science applications to archaeological interpreta-
tion have concentrated on comparisons between natural soils
and human influenced soils (e.g., Collins and Shapiro 1987).
Such applications have proven useful in avoiding misidentifi-
cation of natural formations as cultural. In this test of
Sears' model of maize-based complexity, soil science applica-
tions are utilized to identify the limitations of the archae-
Soil samples were taken from one of the remnant circular
ditch formations in order to examine the potential of the
soils associated with the earthwork to support long-term maize
agriculture (Figure 5). Three column soil samples were
removed with a bucket auger. These bisected the ditch with
column #1 removed from the berm of the earthwork, #2 taken
from the bottom of the ditch, and #3 from outside of the
ditch. Each was excavated to 188 cm below the surface, the
maximum depth that the bucket auger could reach. These
resulted in a profile view of the ditch and its associated
berm (Figure 6).
Morphological descriptions of each field specimen were
conducted prior to preparing the samples for laboratory
analyses. These included horizon designation, moist Munsell
Schematic profile view of Columns 1-3 (not
color, structure (Soil Survey Staff 1981), and any other
observable characteristics (Table 1). Unfortunately, removal
of the samples with a bucket auger and immediate storage in
specimen bags resulted in some disturbance to them. Thus,
structure was reported only as to type. However, careful
storage and handling of the specimens prevented mixing and,
therefore, structure disturbance should not be considered a
threat to the reliability of the data.
Four soil analyses were performed on the samples: organic
carbon content; pH in water, and in potassium chloride and in
calcium chloride solutions; particle-size distribution (Soil
Survey Staff 1984); and aluminum content (Rhue and Kidder
1984). These data, along with data obtained from the morpho-
logical descriptions, provide a general basis for classifying
the soils and making some interpretations about their forma-
Based on the available data and recent mapping of the
county (Soil Survey Staff 1989), the soils associated with the
ditch belong to the order Spodosols. Spodosols are character-
ized as naturally infertile (Soil Survey Staff 1975:333).
This is due, at least in part, to their high acidity. Figure
7 provides evidence of the highly acid nature of the soils
Table 1. Morphological characteristics of soils.
Depth Horizon Moist Colors Structure Roots
18 Apl 10YR 3/1 subangular blocky abundant
31 Ap2 moderate
46 Ap3 granular; few
61 Ap4 ; *
71 Ap5 ..
81 Ap6 7.5YR 3/0 single-grain none
112 Ab3 10YR 3/1 some
122 Ebl 10YR 5/1
130 Eb2 10YR 4/1 none
147 Bhb 10YR 2/1 subangular blocky; **
155 Bwbl 10YR 3/2 single-grain
170 Bwb2 10YR 4/4 some
183 Bwb3 none
18 Ap 10YR 3/1 subangular blocky abundant
31 El 10YR 4/1 some
46 E2 none
61 E3 10YR 3/4
74 E4 "single-grain
86 E5 10YR 4/4
102 Bhl 10YR 5/4 subangular blocky; **
119 Bh2 10YR 4/3 **
135 Bwl 10YR 6/3 ; **
152 Bw2 **
163 Cl 10YR 6/2 single-grain
178 C2 10YR 7/2
188 C3 10YR 7/1
18 Ap 10YR 3/1 subangular blocky abundant
31 El 10YR 4/1 none
48 E2 10YR 5/1
61 Bhl 10YR 5/1 ; **
w/ 10YR 2/1
74 Bh2 10YR 2/1 "" **
86 Bh/E 10YR 2/1 ; **
w/ 10YR 7/2
104 B'hl 10YR 2/2 ; **
119 B'h2 10YR 3/2 ; **
137 Bh3 10YR 3/2 ; **
150 Bw 10YR 4/3 ; **
163 Cl 10YR 6/3 single-grain
188 C3 10YR 7/2
fragmented cemented spodic horizon found in matrix.
** occurs in single-grain matrix.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
As measured in
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
As measured in
2 3 4 5 6 7 8 9 10 11 12 13 14
As measured in
pH distribution of soils in water, potassium
chloride, and calcium chloride solutions.
I I I I I I I I
tested. Nowhere in the soil columns is the pH value higher
than 4.9. Darrel Metcalfe and Donald Elkins (1980:340) state
that maize will not grow adequately below pH 5.5. Further-
more, Clyde Evans and E.J. Kamprath (1970) indicate that
acidic mineral soils, such as these, combined with the
presence of exchangeable aluminum, yield poor maize growth.
Figure 8 provides evidence of the very high concentrations of
aluminum in these soils. While not providing the amount of
exchangeable aluminum, this figure does suggest that aluminum
toxicity could be a problem for maize. As such, the likeli-
hood that the earthworks could support maize cultivation for
the long time span suggested by Sears (1982:191-201) is not
very high without considerable amendments.
Evans and Kamprath (1970:894) indicate that maize growth
can be improved on acidic mineral soils, such as these, by
liming the soil. This practice, along with fertilizing, would
have resulted in good maize growth. Sears believes the
residue in the ditches could have served as a source of
fertilizer. He states, "The muck that accumulates in the
ditches, particularly after a flood, is practically protoplas-
mic, wiggling with great quantities of fish and amphibian
small-fry and eggs. The nitrogen content at least should be
most useful in maize growing, and the tiny bones might provide
some phosphorus. Certainly the ditches had to be cleaned
occasionally if they were to function" (1982:189).
0 100 200 300 400 500 600 700 800 900 1000
0 I I I Horizon
S -25-- Ap2
D -75-- Ap5
0 100 200 300 400 500 600 700 800 900 1000
0 I Horizon
S -2 El
r -5 / E3
o -125- Bh2
U *- 9wl
C -150- Bw2
-175 /.- C1
0 100 200 300 400 500 600 700 800 900 1000
0| I 1 Horizon
- -50- E2
i In Bh3
- -150- R Bw
Figure 8. Al distribution in columns 1-3.
But, if the ditches were cleaned occasionally and if this
cleaning resulted in a preference towards dumping the debris
from the ditch into the interior of the earthwork (as would be
the case if the residue caught in the ditches was being used
as fertilizer), then a pattern in the particle-size distribu-
tion of the upper levels of the column located on the berm
(i.e. column #1) would reflect the particle-size distribution
of the lower levels of the column located inside the ditch
(i.e. column #2). That is to say, if the ditch were cleaned
periodically and excavated below the Bh (spodic horizon) as
suggested by Sears (1982:6), then the particle-size distribu-
tion found in the lower levels of column #2 should be the same
as that found in the upper levels of column #1. However,
instead of periodic cleaning, the particle-size distribution
of column #1 reflects a single episode of activity which built
the ditch. This can be viewed graphically in Figure 9.
Figure 9C, column #3, the column taken outside the ditch,
provides a model for what the others looked like prior to
excavation of the ditch. Its particle-size distribution shows
a trend towards an increase in the percentage of the larger
particles (i.e. vcs + cs) and a trend toward a decrease in the
smallest particles (i.e. fs + vfs and silt + clay) with
increasing depth. This pattern is even more pronounced in the
lower levels of column #2 (Figure 9B). Thus, if periodic
cleaning of the ditch resulted in excavation below the Bh,
then a substantial increase in the percentage of the larger
A. COLUMN 1
) 10 20 30 40 50 60 70
I I :
B. COLUMN 2
0 10 20 30 40 50 60 70
C* COLUMN 3
0 10 20 30 40 50 60 70
I / /
\ / /
SA A *
Particle size distribution of columns 1-3
(where vcs = very coarse sand, cs = coarse
sand, ms = medium sand, fs = fine sand, vfs
= very fine sand).
I i : :
particles should be evident in the top layers of column #1
(Figure 9A). However, the particle-size distribution for
column #1 reveals an almost mirror image of particle-size
distribution between the Ab and the Ap horizons. That is to
say, the Abl and the Ap6 are almost identical. Likewise, the
Ab2 and the Ap5 are very similar. This "mirror image"
correlation continues with the Ab3 and Ap4, the El and Ap3,
the E2 and Ap2, and the Bh and Apl layers. Nowhere in this
profile are there substantial increases in the larger parti-
cles as would be expected if periodic cleaning of the ditches
Supporting evidence for this model is provided by the
graphic illustration of the distribution of organic carbon
content for each column. Figure 10 provides a comparison of
the distribution of organic carbon content among the three
columns and indicates the highest percentage is located
between 75 and 100 cm below the surface in column 1. Although
it is evident that column 3 also shows a high percentage of
organic carbon deep in the soil profile, note that it coin-
cides with the Bh horizon. This is because the Bh horizon is
defined as a horizon with an accumulation of organic matter
(Soil Survey Staff 1975). However, the high percentage of
organic carbon in column 1 coincides with the Ap6. (Note that
this horizon has the lowest concentration of aluminum content
within the Ap and Ab horizons of column 1 [Figure 8] and,
therefore removes the possibility that a Bh horizon was
Organic Carbon (%)
0.25 0.50 0.75 1.00
Organic Carbon (%)
0.25 0.50 0.75 1.00 1.25
Organic Carbon (%)
0.50 0.75 1.00
Organic carbon distribution of columns 1-3.
mistakenly identified as an Ap horizon.) Thus, I argue that
the ditch is the result of an excavation that began with
removal of the A horizon and subsequent discard on top of the
old ground surface (effectively beginning the berm). This was
followed by excavation of the E and Bh horizons, which were
heaped on top of the A horizon. Subsequent pedogenetic
processes resulted in the accumulation of organic carbon,
creating an artificially thickened Ap horizon as seen in
column 1. Further proof of this scenario is provided by the
traces of the Bh (as seen in Table 10) in the Ap3 and Ap4
horizons. This evidence, in conjunction with the particle-
size distribution data, supports the model of a single episode
of activity rather than periodic cleaning of the ditch. An
important archaeological implication of this is the amount of
labor involved in construction of the ditch. While it is not
possible to determine the exact number of individuals nor the
amount of time it took to construct, the model suggests an
organized labor force over a relatively short period.
As for the liming of the fields, Sears (1982:171-174,178)
identified lime deposits on the site that are contemporaneous
with the use of the Great Circle. These deposits were found
associated with the Mound A midden deposits. Conceivably,
they could represent lime processing activity areas for
artificially raising the pH of the soils within the Great
Circle, but liming without fertilizing would still result in
poor maize growth. On the other hand, liming with fertilizing
should have left some evidence of the fertilizer--e.g., the
large accumulations of midden materials at the site. Since
these could easily have served as a source of fertilizer, one
could expect that some would have been found within the
alleged planting surfaces. However, Sears (1982:176-177)
indicates that repeated attempts to find materials within the
Great Circle were made by surface inspection and excavation
but nothing was ever found. Thus, the evidence indicates that
neither lime nor fertilizer was used within the Great Circle
to prepare the soils for maize cultivation.
Finally, Sears (1982:186) states that the ditches were
dug through an impermeable hardpan (i.e. the term that Sears
uses for the spodic horizon) in order to drain the planting
surface. However, this analysis of the soils reveals that
doing so results in additional aluminum in the upper horizons
of the soil (Figure 8) well within the rooting zone. As such,
Sears' explanation for the ditching would have a net result of
adding toxins to the surface prior to planting maize. Thus,
soil science data indicate the circular earthworks were not
planting surfaces for "a stable, dependable maize crop," but
must have served some other purposess.
Ethnohistoric Evidence: Conflicting Accounts
Perhaps the strongest evidence challenging Sears' (1982)
belief that maize served as the dietary mainstay comes from an
eyewitness account of life in sixteenth century South Florida.
Do. d'Escalante Fontaneda, a shipwreck survivor, lived among
South Florida's natives for 17 years from 1548 to 1565. His
memoir, focusing on the Calusa of southwest Florida, provides
some details on subsistence activities within the Okeechobee
Basin. He refers to a bread made from roots as the common
food most of the time except when the lake rises in some
seasons so high that the roots cannot be reached and they are
"for some time without eating this bread" (Fontaneda 1945:13).
He does not mention maize at all. Sears (1982:201) acknowl-
edges this problem, but questions whether Fontaneda would have
recognized maize if he had seen it.
Of course, there were probably many things Fontaneda did
not recognize and could not describe, but one must wonder why
it is that both the Narvaez and de Soto entradas described
maize in Florida years before Fontaneda lived among the
Calusa. Milanich (1987, 1989) believes this is because both
of these Spanish explorations landed in present-day Tampa Bay,
well north of the South Florida aboriginal groups that
Fontaneda described. On the other hand, Dobyns (1983)
believes that Narvaez and de Soto landed in Charlotte Harbor
and described maize use among the Calusa. In either case,
Fontaneda surely would have known about maize by the time he
had written his memoir. If maize had been as important in the
Calusa diet as Dobyns (1983) contends, how could Fontaneda
have omitted it?
Such an argument might go on indefinitely except for the
fact that of all the archaeological research conducted in
South Florida only the Fort Center site has produced any
evidence for maize (save that which has been found in Seminole
Period contexts). On the southwest coast, the few systematic
attempts to recover plant remains have produced no evidence of
maize (Marquardt et al. 1985). Similarly, Margaret Scarry
(1982) found none on the southeast coast. Indeed, John
Griffin (1988:298) remarks that there is no evidence for maize
cultivation in the Caloosahatchee or Everglades archaeological
areas. Based on the maize pollen at Fort Center, Milanich
(1987:177) accepts that it was present at the site sometime
during the period ca. A.D. 100-500. But, whether it was
present before or after that he feels is speculative.
Milanich and Ruhl (1986:2) point out that the importance of
maize within the Belle Glade people's diet is uncertain.
Johnson and Collins (1989) concur with Milanich (1987)
that the maize pollen found at Fort Center indicates maize was
grown at the site but believe that the circular earthworks
were not dug to create planting surfaces for a stable,
dependable maize crop. Following Milanich and Ruhl (1986:2),
they believe the maize grown at Fort Center must have served
some other purposes) other than a dietary staple.
In addition, Carr (1985:299) notes that all of the
circular earthworks are located near elevated hammocks or
other upland environments that could have naturally offered
the same drainage characteristics that Sears (1982) believes
the ditches offered. Indeed, the only advantage that the
ditches offered was the accumulation of organic sediments that
could serve as a source of nutrients for Sears' hypothesized
planting surfaces (Carr 1985:299) and that advantage apparent-
ly was not realized.
On the other hand, Donald Lathrap's (1987) study of the
introduction of maize into precolumbian eastern North America
questions current wisdom and warns us "not to be overly
skeptical of the presence of maize in situations where its
preservation is a matter of chance, such as in 'Green Corn'
utilization." Such skepticism may well be pertinent since
some ethnohistoric evidence for "Green Corn" use exists. For
instance, Oviedo (1959:2) is apparently speaking of green corn
use among the Taino of the Greater Antilles when he says "when
the ears are tender they are eaten almost like milk."
Furthermore, Lathrap (1987) outlines a plausible early
introduction of maize which conforms to Sears' (1982) hypothe-
sis of its introduction into Florida and takes exception to
Keegan's (1987) view that maize could not have passed through
the Antilles prior to the arrival of the Taino. Lathrap
(1987:347) notes that navigation in the Circum-Caribbean is
well documented as early as 3100 B.C. and that the early maize
in the Southeast discussed by Sears (1982) had a totally
different origin from the maize that was utilized by Missis-
sippian societies. Furthermore, he believes that a
two-pronged entry of maize agriculture into the eastern United
States may well explain the heterosis and success of Midwest
Hybrid Maize (1987:347).
Mainstream opinion on this topic, however, has been at
odds with Lathrap. Walton Galinat (1985) has provided a
recent summary of the domestication and diffusion of maize
into North America (north of Mexico) and while he readily
admits that isolation of races of maize and subsequent
heterosis from reconvergence of divergent populations accounts
for the Corn Belt Dent variety (what Lathrap is calling the
Midwest Hybrid Maize), the origins of the heterosis are to be
found in the introduction by the Spanish explorers of white
dent corn from Mexico blending with the Northeastern Flint
variety (originating as Maize de Ocho in the Southwest around
A.D. 200 700).
The soil data presented here strongly support a view that
is contradictory to Sears' model. Periodic cleaning of the
ditches, fertilizing the planting surface, and the effects of
digging below the hardpan are the main elements of Sears'
model that have been addressed. Soil science applications to
each element indicate annual harvests of maize were not
possible from the planting surface identified by Sears (1982).
These, however, do not refute the presence of maize pollen at
the site nor do they suggest that maize was never grown at the
site. Rather, the soil science data modify Sears' model in
two fundamental ways. First, they indicate the role of maize
in the Belle Glade peoples diet was probably more akin to
Milanich and Ruhl's (1986) view that maize was not a dietary
staple and second, they suggest that Sears' purely functional
interpretation of the earthworks is incorrect.
In the case of the latter, Goggin and Sturtevant's
(1964:194-197) view that the earthworks were ceremonial is
certainly the most reasonable interpretation at this time. A
cursory inspection of the literature on drained and raised
fields in South America (Denevan 1970; Denevan et al. 1987;
Farrington 1985; Smith et al. 1968) reveals morphological
similarity throughout the areas studied. In no case do these
agricultural fields resemble the Belle Glade earthworks.
Indeed, the only earthworks that are morphologically similar
are those in the midwestern United States. Current research
on them supports a ceremonial interpretation (e.g., Brose and
As to the former, Sears' model lacks corroborative
evidence to support his arguments. If Lathrap (1987) is
correct in his assertion that a preservation bias exists
because of "Green Corn" utilization, then we must look for
corroborative evidence in the pollen record. If possible,
phytolith analysis should be undertaken as well. As long as
the Fort Center maize pollen stands alone as evidence, the
ability to interpret the role of maize in the Belle Glade
culture is compromised.
REMOTE SENSING, AND ARCHAEOLOGICAL FIELD AND LAB METHODS
The methods used for obtaining data on the Belle Glade
culture area involved a background and literature review, a
review of aerial photographs, a pedestrian survey for ground-
truthing sites found on the aerial photographs and informant
interviews. The background and literature review amassed data
from all over the Basin, but the review of aerial photographs,
pedestrian survey, and informant interviews were limited to an
area on the west side of the lake as part of the West Okeecho-
bee Basin Project (Figure 11). These methods, as well as the
methods used for obtaining data from the field and for
laboratory analysis of the collected data, are explicated
Background and Literature Review
The background and literature review provided a frame of
reference for evaluating archaeological sites. Pertinent
references on the precolumbian archaeology of the area were
consulted. Furthermore, a review of the Florida Master Site
File provided information on all known archaeological sites in
the project area.
Figure 11. West Okeechobee Basin Project Area.
Unfortunately, many of the known sites in the project
area were recorded on the basis of scanty information and
recorded locations, at times, are vague. Thus, a conscious
effort was made to obtain exacting information on recorded
sites that were lacking locational data.
Early maps of the project area were checked for informa-
tion on aboriginal site locations, early homesteads, and
military roads. The original survey maps (Florida State
Archives 1837) were considered the best early maps for this
information. One aboriginal site, the Old Venus Mound (8Hgl),
two historic homesteads, and numerous military roads were
discovered by this method. However, only the Old Venus Mound
was surveyed while in the field and, as such, the homesteads
and military roads were not given site numbers and do not
appear in the results section.
The review of aerial photographs proved most enlightening
as not only were 10 previously unrecorded sites found but
significant additional information on the architecture of six
previously recorded sites was obtained. As stated in the
introduction, early aerial photographs are available in
stereopairs at the Map Library on the University of Florida
campus. Reviewing these involved locating the project area on
an index, retrieving the pertinent photographs, and viewing
Earthwork sites were easily discerned and plotted on USGS
quadrangle maps. Stereoscopes were employed at this stage to
provide detail on the sites. Whenever a possible mound was
found the stereoscopes were used to aid in deciding whether or
not it should be plotted.
Once a site was identified and plotted, later aerial
photographs of the site were viewed for additional detail.
These provided remarkably more information than could any one
photograph because portions of a site not visible in one were
often visible in another. Many factors are responsible for
this variation and include, but are not limited to, the time
of day the photograph was taken, season the photograph was
taken, angle of the camera, and moisture conditions at the
site. Thus, the best image of any one site is a composite
from many photographs.
Overall, this remote sensing application is most useful
in relatively flat, savannah areas of the Basin. Forest
cover, regardless of habitat, obscures manifestations of
cultural activity. Thus, in areas of forest cover, this
method is useful only in cases of known site locations.
Pedestrian Survey and Informant Interviews
The fieldwork was conducted via a pedestrian survey
coupled with informant interviews. The pedestrian survey
included the employment of judgmentally placed shovel tests.
Informant interviews included contacting landowners and other
knowledgeable persons during the fieldwork portion of the
Subsurface tests included shovel tests and bucket auger
tests. The shovel tests were approximately 50 cm in diameter
and were dug in 20 cm levels. The matrix from each was
screened through 1/4-inch hardware mesh and, for some levels,
through 1/16-inch window screen. (The addition of the
1/16-inch fraction was requested by the zooarchaeology
department personnel at the Florida Museum of Natural History,
the repository for the faunal remains.) However, artifact
analysis is based solely on the 1/4-inch fraction.
Soil from each shovel test was described in the field
notebook. The color of contrasting zones and their thickness-
es were noted. All were dug to culturally sterile soil unless
groundwater seepage or soil conditions prevented it. The
tests were placed judgmentally unless more than one was dug at
a site. In those cases, additional shovel tests were placed
on a grid pattern with spacing of the subsurface tests
depending on the local conditions.
In the event that cultural materials were recovered in a
shovel test no attempt was made to delineate the subsurface
boundaries of the site. Rather, the tests served as a sample
of the materials from the site for comparative purposes as
well as document information on the site's integrity and
The bucket auger tests, on the other hand, were generally
used to obtain soil samples from sites. However, there were
cases in which the bucket auger was used to test whether or
not cultural materials were present at a particular locale.
The bucket auger is an 8 cm in diameter coring device that
allows quick removal of a soil core. The soil was not
screened when using this method. Rather, the soil is poured
from the coring device onto a tarp and visually examined for
the presence or absence of cultural materials. If any
cultural materials were found, they were noted and a site
number was assigned. On the other hand, if no cultural
materials were found, the location was not considered a site.
Thus, the accuracy of this method is limited by the variabili-
ty in cultural material content at a site. A site that is
characterized by a relatively high density of cultural
materials is easily discernible with the use of the bucket
auger. However, sites that are characterized by a low density
of cultural materials (i.e., no cultural materials found in
the bucket auger sample) were not identified as sites.
Therefore, caution must be observed when determining whether
or not a locale is a site with this method.
When the auger tests were used in this manner they were
placed judgmentally on the site. Once cultural materials were
encountered (usually within the first 5 cm), the presence of
materials was noted, the site named, and the matrix poured
back into the auger test hole. If no cultural materials were
encountered, the test was dug below the dark colored soils to
what appeared to be culturally sterile soils and then back-
filled. Generally, only one such test was excavated at each
Laboratory analysis of the materials collected from this
survey involved separating the specimens from each provenience
into four gross categories. These categories are lithics,
ceramics, animal bone, and soil samples.
The lithics were measured across their longest axis and
grouped into the following categories: very small (<1 cm),
small (1-2 cm), medium (2-4 cm), large (4-6 cm), very large
(6-8 cm), and extra large (8-10 cm). They were then counted
and, based on the presence or absence of cortex, were classi-
fied into primary flakes (those flakes with cortex completely
covering the dorsal surface), secondary flakes (those flakes
with little cortex on the dorsal surface), and tertiary flakes
(those flakes with no cortex on the dorsal surface).
The ceramics were compared with typology collections at
the Florida Museum of Natural History and subjected to
microscopic analysis (when warranted) in order to determine
their typological classification. Each sherd was measured
across its longest axis and grouped into the following
categories: very small (<1 cm), small (1-2 cm), medium (2-4
cm), large (4-6 cm), very large (6-8 cm), and extra large
(8-10 cm). They were then counted and are reported in the
results section as total number of sherds/number of rim sherds
in that total.
The animal bones were examined for obvious diagnostic
elements (e.g., Lepisosteus spp. vertebra and scales, Odocoil-
eus virginianus astragali, etc.) and species observed in the
assemblages are noted. However, no attempt was made to
provide a full analysis of the materials. Instead, they were
weighed by provenience (note that some weights are skewed by
the presence of impurities, usually concretions) for compara-
tive information on relative density of deposits.
The soil samples were subjected to either morphological
analysis or pH analysis. Morphological analysis was reserved
for auger samples from two of the earthworks. It is used to
describe the soil horizons, thereby providing insight on their
cultural manipulation. Analysis of pH, on the other hand, was
utilized at a number of the sites to provide a means of
checking for possible preservation bias. The readings are
based on an E.M. System Soil Tester.
A total of 38 previously unrecorded archaeological sites
were discovered in this survey. In addition, the review of
aerial photographs revealed significant new data on nine
previously recorded sites. These are described in this
A discussion of the previously unrecorded sites is
included in the section New Sites and the previously recorded
sites are listed under the heading Updated Sites. These are
followed by the section Negative Data, which provides informa-
tion on the field inspection of two localities that were
thought to have mound features based on inspection of aerial
photographs, but the pedestrian survey revealed they did not.
80b27 Wedqworth Mound. The Wedgworth Mound is located in
Township 33S, Range 35E, Section 25 (Figure 12). The site was
located via informant interview and consists of a large sand
mound measuring 48 m north-south by 50 m east-west and is at
least 2 m high. It was reported to the author by the owner,
Mr. George Wedgworth, via Dr. William H. Marquardt of the
(b" + '- . 2j:C SA'; 3;E LOCATION
2 214 J. II
.4%.. ) I '*",M.C
S-, / 1
*! a 1. 214" -
A;u s0 .
\ -6 ..... -_ : .- -_ -,- -. .- - - -----
.3"-, .:0b2 .7. :.
..5 0 1 MILE -
S1 .5 0 1 KILOMETER
Figure 12. Wedgworth (80b27) and Tennis Court (80b34)
.5 0 1- KILOMETE
Figur 2. edwrh(02)adTnis or 8b4
Florida Museum of Natural History. Inspection of aerial
photographs before and after fieldwork revealed no earthworks
in the vicinity.
One of two shovel tests yielded a few highly deteriorated
bone fragments and a single sherd. The sherd is the type
Dunn's Creek Red, suggesting St. Johns period cultural
affiliations. The bone fragments are not identifiable and
their condition precludes curation. Since it has been
documented that soil acidity can affect bone preservation (see
Gordon and Buikstra 1981), the soil pH was obtained. It
revealed that soil acidity is probably not responsible for the
poor condition of the bone as the pH registered neutral (at
The mound probably did not serve as a habitation site.
Its large size coupled with its lack of preserved faunal
remains suggests alternative functions.
It is thought that additional testing may reveal other
cultural affiliations however, based on the limited data
available, the site's contents and its presence near the Fort
Drum Creek drainage suggests a cultural boundary between the
Belle Glade and St. Johns regions.
80b34 Tennis Court Mound. The Tennis Court Mound is
located in Township 33S, Range 35E, Section 24 (Figure 12).
The site was located via informant interview and consists of
a small oval mound measuring approximately 15 m in diameter
and is at most 0.5 m high. It, like the Wedgworth Mound, was
reported to the author by Mr. George Wedgworth. Inspection of
aerial photographs was done after fieldwork and revealed no
earthworks in the vicinity.
One of two shovel tests produced one flake at a depth of
60 cm below the surface (cmbs). Furthermore, these tests
revealed what appeared to be a buried midden stain in the soil
profile between 45 and 90 cmbs. No animal bone was discovered
nor does the soil pH (6.9) suggest that soil acidity has
created a preservation bias. As such, it is difficult to
determine the functions) of this mound.
At this time, no specific cultural affinities can be
discerned but the presence of the flake and the buried midden
stain strongly suggest that the mound is precolumbian.
80b28 Fort Kissimmee Earthworks. The Fort Kissimmee
Earthworks are located in Township 34S, Range 31E, Section 11
(Figure 13). The site was located via remote sensing (aerial
CYW-4H-178) and consisted of a large rectangular shaped ditch
with rounded corners measuring approximately 500 m north-south
by 265 m east-west. The northwest corner of the rectangle
opened to a bend in the Kissimmee River such that water could
flow from the River into the ditch. Two embankments measuring
approximately 110 m each were located within the interior of
the rectangular earthwork and may have terminated at mounds.
Unfortunately, this site was not visited during the field
survey. Access is restricted by the Kissimmee River on one
side and the Kissimmee Canal on the other. Aerial photographs
S- - FLORIDA,
'- QUADRANGLE LOCATION
_.._ -... . .. . ---.- /.
\N ._/'- ._ : ., -1; -- ,, / *"^,* ..--
*-- ,:': N' .
,/i_-. -- .j.,. ,
^wj(r ~^ H "-_' ,. -
/ / V-u~v/ -) *-
it 49 X I t
1 .5 0 1 MILE
S.5 0 1 KILOMETER
Fort Kissimmee Earthworks (80b28).
reveal that construction of the Kissimmee Canal resulted in
destruction of about 2/3 of the site leaving only the north-
west corner of the earthwork intact (aerial 12055-174-126).
(Note that figure 13 suggests the northwest corner of the site
was destroyed by the canal; this is not the case. Instead, it
is the only portion of the site that is intact. The problem
with the figure is due to an error in placement of the
Kissimmee Canal by the U.S.G.S.) The embankments and most of
the ditch were dug through during excavation of the canal (see
The limited data on this site do not allow for definitive
statements of its origin, much less its function. However,
its overall morphology, especially in light of the data on the
next two sites, suggests it originates with Belle Glade
80b29 Fulford Earthworks. The Fulford Earthworks are
located in Township 34S, Range 32E, Section 31 (Figure 15).
The site was located via remote sensing (aerial CYW-4H-117)
and consists of a large rectangular shaped ditch with rounded
corners measuring approximately 230 m north-south by 170 m
east-west. In addition, a linear ditch extends from the
southwest corner of the earthwork approximately 200 m west
connecting the earthwork to the Kissimmee River. Similar to
the Fort Kissimmee Earthworks, an embankment measuring
approximately 100 m is located within the interior of the
earthwork and may terminate at a mound.