Membership Information
 Table of Contents
 Editor's Page
 Late Quaternary Environments in...
 A Distribution Analysis of the...
 Florida Anthropologist Interview...
 Paleo-Indian Manifestations in...
 A Preliminary Report on the Excavations...
 An Extinct Bison Kill Site, Jefferson...
 A Mammoth Kill Site in the Silver...
 Explorations of a Paleo-Indian...
 A Late Paleo-Indian Site, Cape...
 Unusual Marks Found on Giant Land...
 An Engraved Mammoth Tusk from Vero...
 Amino Acid Racemization Date from...
 Back Cover

Group Title: Florida anthropologist
Title: The Florida anthropologist
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00027829/00011
 Material Information
Title: The Florida anthropologist
Abbreviated Title: Fla. anthropol.
Physical Description: v. : ill. ; 24 cm.
Language: English
Creator: Florida Anthropological Society
Conference on Historic Site Archaeology
Publisher: Florida Anthropological Society.
Place of Publication: Gainesville
Frequency: quarterly[]
two no. a year[ former 1948-]
Subject: Indians of North America -- Antiquities -- Periodicals -- Florida   ( lcsh )
Antiquities -- Periodicals -- Florida   ( lcsh )
Genre: periodical   ( marcgt )
Summary: Contains papers of the Annual Conference on Historic Site Archeology.
Dates or Sequential Designation: v. 1- May 1948-
 Record Information
Bibliographic ID: UF00027829
Volume ID: VID00011
Source Institution: University of Florida
Holding Location: Department of Special Collections and Area Studies, George A. Smathers Libraries, University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 01569447
lccn - 56028409
issn - 0015-3893

Table of Contents
    Membership Information
        Unnumbered ( 3 )
    Table of Contents
        Table of Contents
    Editor's Page
        Page 2
    Late Quaternary Environments in Florida and the Southeast
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
    A Distribution Analysis of the Clovis/Suwannee Paleo-Indian Sites of Florida -- A Geographic Approach
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
    Florida Anthropologist Interview with Ben Waller
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
    Paleo-Indian Manifestations in the Tampa Bay Region, Florida
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
    A Preliminary Report on the Excavations at Harney Flats, Hillsborough County
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
    An Extinct Bison Kill Site, Jefferson County Florida
        Page 81
        Page 82
    A Mammoth Kill Site in the Silver Springs Run
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
    Explorations of a Paleo-Indian Site on the Aucilla River
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
    A Late Paleo-Indian Site, Cape Haze Peninsula, Charlotte County, Florida
        Page 98
        Page 99
        Page 100
    Unusual Marks Found on Giant Land Tortoise Remains in Hillsborough River
        Page 101
        Page 102
        Page 103
        Page 104
    An Engraved Mammoth Tusk from Vero Beach
        Page 105
    Amino Acid Racemization Date from Paleo-Indian Sites in Florida
        Page 106
        Page 107
        Page 108
    Back Cover
        Page 109
        Page 110
Full Text


2000 Florida Anthropological Society Inc.

The Florida Anthropological Society Inc. holds
source text of the Florida Anthropologist
considered the copyright holder for the text
these publications.

all rights to the
and shall be
and images of

The Florida Anthropological Society has made this publication
available to the University of Florida, for purposes of
digitization and Internet distribution.

The Florida Anthropological Society reserves all rights to this
publication. All uses, excluding those made under "fair use"
provisions of U.S. Code, Title 17, Section 107 are restricted.

Contact the Florida Anthropological Society for additional
information and permissions.



i.' 111 rl liVOLU E i3 I
'' VOLUME 36




`'' '

THE FLORIDA ANTHROPOLOGIST is published quarterly in March, June, September, and
December by the Florida Anthropological Society, Inc., Geoarcheological Research
Center, Department of Geology, University of Miami, Coral Cables', Fl. 33124. "
Subscription is by membership in the Society for individuals and institutions
interested in the aims of the Society. Annual dues are $10.00 ($12.00 after
March 1). Requests for membership dues and changes of address should be addressed ::
to the Membership Secretary; manuscripts for publication and orders for back
issues to the Editor; and newsletter items to the President. Address changes
should be made at least 30 days prior to the mailing of the next issue. Second
class postage paid at Miami, Florida 33124.


President: John Beriault Treasurer and Resident Agent:
3550 Bolero Way Ruth Thomas
Naples, FL 33942 545 Bayberry Drive
Lake Park, FL 34403
1st Vice President: Claudine Payne
Tallahassee, FL 32301 Directors-at-Large

'2nd Vice President: Karen Malesky Bender Three years: Mitchell Hope
Palmetto, FL 33561 1113 Sunset Drive
Sebring, FL 33870
SSecretary: M. Katherine Jones
Tallahassee, FL 32301 Two years: Mary Lou Watson
229 Woodlawn Drive
.Membership Secretary: Joan Deming Panama City, FL 32407
1839 Pine Cone Circle, #28
Clearwater, FL.33520 One year: Jerry Hyde
4233 Oristano Road
Jacksonville, FL 32210


Editor: Robert S. Carr
Geoarcheological Research Center
Department of Geology
University of Miami
Coral Cables, FL 33124

Editorial Board:
': Morgan Crook
VWest. Georgia College
;,,.Carrollton, GA 30118

-.?etltieen A. Deagan
Hi.strie St. Augustine Preservation
SBoard .
St. 'Augustine, FL 32084


Assistant Editor: George M. L
Sarasota, FL

Guest Editor: Jim Dunbar
Tallahassee, FL

; :' .


.:. ...-. I

.* .. *: :
...* ::
... ,.. -
: :: ie

: "!: :s .

"- :.

Yscar Iscan
Department of Anthropology
Florida Atlantic University
Boca Raton, FL 33431

George Percy
Division of Archives, History and
Records Management
Tallahassee, FL 32304

"The Hunt". Courtesy of Frank H. Gilson.

-. ,.:..i :: ::

:: .-;"R
..:. ** "

.. ..I. ?.,





Editor's Page . . .. . . . 2

Late Quaternary Environments in Florida and the Southeast
by Victor A. Carbone . . . . 3

A Distribution Analysis of the Clovis/Suwannee Paleo-Indian
Sites of Florida--A Geographic Approach
by James S. Dunbar and Ben I. Waller. . ... 18

Florida Anthropologist Interview with Ben Waller. ... 31

Paleo-Indian Manifestations in the Tampa Bay Region, Florida
by Albert C. Goodyear, Sam B. Upchurch, Mark J. Brooks,
and Nancy N. Goodyear. . . . ... 40

A Preliminary Report on the Excavations at Harney Flats,
Hillsborough County
by Randy Daniel and Michael Wisenbaker. ... 67

An Extinct Bison Kill Site, Jefferson County, Florida
by S. David Webb, Jerald T. Milanich, Roger Alexon,
and James Dunbar . . . . ... 81

A Mammoth Kill Site in the Silver Springs Run
by Charles A. Hoffman. . . ... ... 83

Explorations of a Paleo-Indian Site on the Aucilla River
by Don Serbousek . . . . ... 88

A Late Paleo-Indian Site, Cape Haze Peninsula, Charlotte
County, Florida
by Dan Hazeltine. . . . ... 98

Unusual Marks Found on Giant Land Tortoise Remains in
Hillsborough River
by Danny H. Clayton . . . .. 101

An Engraved Mammoth Tusk from Vero Beach
by E. H. Sellards. . . . ... .105

Amino Acid Racemization Dates from Paleo-Indian Sites
in Florida
by Paul M. Lien . . . .. 106


This special double-issue on the Paleo-Indian is a testimony
to the hard work of Jim Dunbar, co-editor for this issue, who devoted
a tremendous amount of time soliciting the articles and the out-
standing cover art. The front cover is the work of Frank H. Gilson,
an illustrator and lab technician for Florida's Division of Archives,
History and Records Management. Mr. Gilson's work has appeared
previously in The Florida Anthropologist and in the 1981 Florida
Atlas. His cover illustration, "The Hunt", depicts two mammoths
approaching a small oasis pond in an otherwise dry river channel.
Although fanciful, his illustration dramatically renders a Paleo-
Indian kill site, perhaps not unlike those that existed along the
rivers of North Florida.

Not included in this issue is an article by Carl Clausen on the
Paleo-environment as reconstructed from data from Little Salt Springs.
His article arrived too late for publication here, but hopefully will
be included in a future issue of The Florida Anthropologist.

A number of books have been printed within the past year that are
of interest to our readers. The long awaited excavation report on Ft.
Center by Dr. William Sears was published by University Presses of Flor-
ida and is available for $21.50. Students of Seminole studies will be
interested in James Covington's Billy Bowlegs War, the only book I
am aware of published exclusively on the Third Seminole War. It is
available for $9.95 and is published by Mickler House Publishers.
An Abandoned Black Settlement on Cumberland Island, Georgia by Mary
R. Bullard is available for $11.95. This book is about the conditions
and adaptation of a small black settlement in post-Civil War coastal
Georgia. Orders may be sent to the Georgia Historical Society, 501
Whitaker Street, Savannah, Georgia, 31499.

Robert S. Carr


Victor A. Carbone

The purpose of this paper is to present an overview of the paleoenviron-
mental evidence which is available from Florida and the Southeast in order to
arrive at a reconstruction of the landscape at the end of the Ice Age which
served as a backdrop to Paleo-Indian and Early Archaic peoples. In order to
understand the cultural patterns of this time period which are evident in the
archeological record we need to have an appreciation of the environmental
setting in which these early peoples played out their daily routines.

That these late Pleistocene and early Holocene environments of unglaciated
Eastern North America were drastically different from what they are today is
now a matter of accepted fact. In the late forties Deevey (1949) presented
evidence which challenged the long-held view of long term stability and con-
tinuity of the deciduous forest formations as espoused by Braun (1955).
Throughout the fifties and early sixties paleoecological investigations in
the unglaciated regions resulted in a small but definitive body of data which
conclusively proved that these contentions regarding the continuity of the
eastern forests were not true. However, these findings took some time in
filtering out to a broader public. The situation in the early sixties could
best be exemplified by reference to two articles which appeared in the same
volume, the 1965 compendium entitled The Quaternary of the United States edited
by Wright and Frey (1965). In it, Williams and Stoltman (1965:669) in their
review of Southeastern prehistory drew certain inferences about early Paleo-
Indian population distributions using Braun's (1955) conclusions on the envi-
ronment to back them up while two hundred pages before, in the same volume,
Whitehead (1965:417) presented data which in essence demolished Braun's con-

With the explosion of studies of both floral and faunal sites which have
been carried out in the last 20 years we need no longer be concerned with the
broader reconstructions of zonal biotic and climatic gradients. The data now
is at such a level that we can begin to deal with it at a finer level of reso-
lution and begin to consider in detail the structure of environmental mosaic \
which we now know was the hallmark of this transitional period. Reevaluation
of both the late glacial floral and faunal evidence indicates that the different
environmental zones were not as distinct as they are today. Rather a new con-
sensus has evolved which sees these late glacial and early postglacial environ-
ments as a hodgepodge or mosaic of microhabitats which allowed for the existence
of a wide array of presently unrelated species.

The evidence for this includes:

(1) The persistence of mixed boreal and temperate, coniferous and deciduous
pollen types within a broader matrix of grasses, sedges, helophytes, hydro-
phytes and xerophytes which suggest open and closed, wet and dry environments
coexisting in close proximity;



(2) The presence in a number of deposits and differing contexts of a
varied megafauna such as the woodland musk ox, the grazing mammoth, the
browsing mastodon, giant moose of swampy forests, grassland species of bison
and horse, woodland peccaries along with the smaller whitetailed deer, caribou,
elk, and giant species of beaver and ground sloth, armadillo and tortoise and
the arctic walrus;

(3) The presence of seemingly anomalous assemblages of small ecologically
sensitive animals which today occupy extremely diverse environments, such as
the yellow cheeked vole from the taiga, the collared lemming from the tundra,
the white footed mouse and pine vole of temperate forests, two prairie species--
the thirteen lined ground squirrel and the sharp tailed grouse, the Hudson Bay
toad and the common frog.

This type of mixture is more characteristic of the Middle Atlantic pro-
vince but a number of sites through Florida and the Southeast support the pic-
ture of an environmentally diverse mosaic confronting these early populations.
Some of the sites are shown in Figure 1 and the list includes faunal sites
such as Edisto Island in the Charleston area (Roth and Laerm 1980), the Georgia
Piedmont (Voorhies 1974), numerous Florida sites, especially those in the Vero
area, the Tombigbee River in Alabama and Mississippi (Frazier 1982) and floral
sites from Columbia, South Carolina (Watts 1980), northern (Watts 1975), central
Piedmont (Sheehan, Whitehead and Jackson 1982), and south Georgia (Watts 1971);
northern and central Florida (Watts 1971,1975,1980) and Alabama (Delcourt
and Delcourt 1981; Whitehead and Sheehan 1982). All support the picture of an
environmentally diverse mosaic confronting these early populations.

The remainder of this paper will be devoted to a discussion of the paleo-
climatic parameters which were responsible for this environmental structure
for which there are no modern analogs. A graphic summary of the available
floral and faunal evidence for the Southeast in general and Florida in par-
ticular also will be presented.


The logical scenario that would result in such a mixture of species is
,one that demands cold summers and warm winters. And indeed, Bryson and Wend-
land (1967) proposed such a scenario fifteen years ago. According to their
model,the late glacial climatic pattern was controlled primarily by the pre-
sence of the Laurentide ice sheet. In winter, because the continental ice
sheet acted as a barrier to the very cold, low level arctic air, there were
fewer incursions of this cold air into the United States thus accounting for
the milder winters. On the other hand in summer air masses would be cooled
by the ice sheet causing a cooler and drier summer regime since evaporation
would be reduced. Wright (1982) and Solomon (1982) have recently reviewed the
situation in papers dealing with the sensitivity of natural systems to climatic
change. According to Wright:

A case has been made by Bryson that the massive Laurentide ice sheet
confluent on the west with the high Cordilleran ice sheet, entrapped
in winter the frigid arctic air in the polar regions. The winter
outbreaks of frigid air that today bring blizzards to the northern
plains and cold weather to Texas thus may not have occurred during
the glacial periods, and in summer the ice sheet provided a supply



BBP-Bob Black Pond (Watts, 1970)
BLB-B. L. Bigbee Oxbow (Whitehead and Sheehan, 1981)
DD-Devil's Den (Martin and Webb, 1974)
EI-Edisto Island (Roth and Laerm, 1980)
GS-Goshen Springs (Delcourt, 1980)
LA-Lake Annie (Watts, 1975)
LKC-Little Kettle Creek (Voorhies, 1974)
LQ-Ladds Quarry (Ray, 1967)
LSS-Little Salt Spring (Clausen, et al. 1979)
N-Nodoroc Bog (Sheehan, Whitehead and Jackson, 1982)
P-Pennington (Watts, 1980)
SL-Sheelar Lake (Watts and Stuiver, 1980)
TT-Tombigbee (Frazier, 1981)
VM-Vero-Melbourne Area (SAI-Webb, 1981)
WP-White Pond (Watts, 1980)
LL-Lake Louise (Watts, 1971)



of cool air that decreased evaporation. The admixture of
species (in pollen records), inexplicable in terms of modern North
American vegetation, may reflect the relatively mild winter con-
ditions and cool summers mentioned above. So also can be explained
the superposed ranges of certain fossil vertebrates that now are
separated in their distributions .

In summary, what we are dealing with is decreased seasonality, reduced
differences between winter and summer temperatures and increased dryness. In
this regard it should be noted that Coleman's (1982) recent analysis of rain-
fall patterns over the Florida peninsula suggests that lower temperatures
generally have resulted in a rainfall pattern that favors an increase of winter
precipitation to the north while the suppression of cyclonic activity which
results from the lowered temperatures results in extended drought conditions
for the southern part of the peninsula. As will be seen below in the vegeta-
tional reconstructions, as well as in Dunbar's (1981) studies of hydrologic
fluctuations, Coleman's conclusions seem to be supported by the data available
from these other fields.

Vegetation History

Figure 2 presents a summary of the vegetational history of the Southeast
for the period between the full glacial maximum and the mid-Holocene hysither-
mal interval. This figure has been adapted from the work of Delcourt and
Delcourt (1981:124) which presents vegetation maps for Eastern North America
from 40,000 B.P. to the present. It should be stressed that even though the
maps show broad vegetational zones, this was probably not the case. What the
figures attempt to convey is the general character of the dominant vegetation
type. Again, in thinking about these early floral and faunal communities for
which there are no modern analogs the major emphasis should be on the mosaic
aspect of the environment. Also, as the Delcourts (1981:124) point out, another
major factor to be considered is the transitional aspect of the vegetational
history. They state that "The paleovegetation maps emphasize that more than 60
percent of the last 40,000 years has been characterized by environmental con-
ditions transitional between extreme glacial and nonglacial regimes." In other
words, environmental change has been a constant of human experience throughout
Paleo-Indian and Archaic times.

Full Glacial Vegetation (18,000 yrs. B.P.)

Delcourt and Delcourt (1981) hypothesize a warm temperate forest of oak,
hickory and southern pine across the southeastern coastal plain from the Gulf
to the Atlantic, including northern Florida. A hardwood forest is projected
along the major river valleys, while along the northern margins of this zone
a band of mixed northern hardwood with boreal conifer forest is assumed as
the transition to the fully boreal jackpine spruce forests of the Middle
Atlantic region. The central and southern parts of Florida including the
Gulf continental shelf were covered with scrub vegetation consisting of xeric
herbs and shrubs such as rosemary and sandhill polygonella, etc. on active
sand dunes with scrub oak patchily intermingled.

More recent evidence from both the Piedmont and south-central section
of Georgia (Sheehan, Whitehead and Jackson 1982; Watts 1971 and north Florida
(Sheelar Lake) (Watts and Stuiver 1980) support the interpretation of a more
patchy mosaic which included local stands of boreal conifers, as well as areas
mantled with more mesic hardwood forests. The Georgia data which is based on


















(from Delcourt and Delcourt, 1981)



both pollen and plant macrofossil evidence clearly demonstrates the persistence
of spruce in the lower Southeast with the southern limit somewhere in south
Georgia. Sheehan, Whitehead and Jackson (1982) describe the full glacial
vegetation of the Georgia Piedmont as follows:

Spruce and fir were not common in the vegetation but were present
on the Piedmont, probably in special microhabitats. Oak was un-
common as were most deciduous trees. Herbs were much more important
in the vegetation than in modern times. Their abundance and
diversity, and the presence of several taxa with "boreal affinities,
give the impression of a park like vegetation, interrupted fre-
quently by patches of trees and shrubs.

In northern Florida, Watts and Stuiver (1980) infer the following vegeta-
tion around Sheelar Lake:

S. an open pine forest with some populations of broad leaved trees.
'Prairie and sandhill herbs may have been dispersed on local unforested

The central and lower peninsula had open xeric scrub vegetation on active
sand dunes, while a coastal strip of savannah is projected along the Gulf.

Late Glacial Vegetation (14,000 yrs. B.P.)

By 14,000 B.P., as conditions began to ameliorate there was a northward
expansion of the mixed conifer-hardwood forests,while the greater southeastern
oak-hickory-pine forest remained relatively stable. A major change involved
the contraction of the land mass of peninsular Florida due to rising sea levels,
and thus a considerable reduction in open xeric habitats (Delcourt and Delcourt

This interpretation again has to be qualified by more recent data. In
the'Georgia Piedmont

there was a decrease in pine dominance, accompanied by a decline in
the fir and spruce populations. Oak and hickory replaced these species
to some extent, reflecting increased warmth. The continuation of high
percentages of herbs and shrubs suggest that the increased warmth was
not accompanied by increased precipitation. Shrubs became some-
what less important, but herbs remained a prominent feature of the
generally open vegetation .(Sheehan, Whitehead and Jackson 1982).

In northern Florida around 14,000 B.P., oak, hickory and hackberry pollen
dominated the spectrum and herbs amounted to up to 40% of the pollen sum.
"The vegetation is interpreted as dry oak hickory stands with local prairie
in a warm, dry climate" (Watts and Stuiver 1980). Around 14,000 B.P. there
was an abrupt expansion of mesic trees in northern Florida suggesting increased
precipitation. Beech became a major component of forest stands.

Through the time period between 14,000 and 10,000 B.P. the same trends
are evident from northern Georgia to northern Florida. Forests became grad-
ually denser with pine and oak gradually colonizing previously unforested
areas, suggesting a return to drier conditions.


Early Postqlacial Vegetation (10,000 yrs. B.P.)

By 10,000 B.P. the mixed hardwood forests to the north had expanded con-
siderably. The oak-hickory-southern pine areas became more densely forested.
Hemlocks and chestnuts were present along the floodplains in northern Georgia.
Oaks and pines were the major components in northern Florida vegetation. In
the peninsula the scrub vegetation was replaced by an oak savannah, with a
strip of prairie persisting along the Gulf coast.

Mid-Postqlacial Vegetation (5,000 yrs. B.P.)

By the mid-Holocene,the hypsithermal interval once again induced a change
towards more open vegetation. According to the Delcourts (1981)

oak and hickory were replaced by southern pine on the sandy
uplands of the Gulf and Atlantic Coastal Plains during the mid-
Holocene interval. Oak-hickory-southern pine forest was restricted
to the Piedmont. In peninsular Florida, southern pine forest
replaced oak savannah. Extensive swamps and marshes developed along
the present coasts, and the subtropical hardwood forest became
established along the southernmost tip of Florida ..

At Sheelar Lake in northern Florida the period between 9500 and 7200 B.P.
saw an increase in oak and herbs (e.g., ragweed, grasses, etc.) with a decrease
in pine suggesting a mixture of oak stands with intervening prairie. The late
Holocene saw a reversal to pine dominance and the establishment of modern
conditions, including the development of cypress swamps.

Faunal History

The faunal record is not as easily resolved as the vegetational record
because of the problems associated with the mixed character of the various
deposits. There has also been less systematic work done on faunal sites
which generally do not lend themselves to the kind of stratigraphic resolution
that is found in pollen sites. Nevertheless, thanks to the work of people like
S. David Webb and his associates at the Florida State Museum, Mike Frazier in
Alabama and Mississippi, as well as a number of others throughout the South-
east, a clearer understanding of faunal distributions during the late glacial
and early postglacial times is beginning to emerge. Two major studies which
have received limited distribution will serve as a basis for this discussion.
The first, and what is perhaps the major synthesis of the late Quaternary
faunal history of the Southeast, is Webb's contribution to the study sponsored
by the Bureau of Land Management, entitled A Cultural Resources Survey of
the Continental Shelf from Cape Hatteras to Key West (Science Applications,
Inc. 1981). In Volume I, Introduction and Physical Environment, Webb presents
the most up-to-date and comprehensive review of late Pleistocene paleontology
available for the Southeast. Other studies which will be cited are Martin
and Webb's (1974) analysis of the Devil's Den fauna in Levy County, Florida
which presumably dates to 8000 B.P. and Frazier's (1981) work along the Tombig-
bee River and its tributaries in Mississippi and Alabama.

Webb's study is based on his analysis of the evidence from paleontological
sites in the Charleston area (Edisto Island), sites in the Vero-Melbourne area
and his extensive knowledge of other localities throughout Florida and the


Southeast. Table I presents an incomplete list of primarily extinct species
which are found in these deposits. Looking at these data one can readily
appreciate the kaleidoscopic array of environments which these remains repre-
sent. Webb discusses the ecological implications of the Edisto Island verte-
brate assemblages as follows:

The predominant vertebrate fossils are large grazers, most of which
were herd ungulates. These include horses, camels, mammoths and
bison. Giant tortoises, glyptodonts and most of the ground sloths
also fall into this broad category.

Browsing vertebrates were also present, notably mastodons, tapirs
and peccaries. Large fresh water mammals, notably giant beavers,
giant capybaras and abundant muskrats, not to mention fishes,
turtles and alligators indicate the proximity of a major river
system. The aquatic and terrestrial vertebrate fauna suggest a
mosaic of deciduous woodland and grassland savannah, crossed by
'major meandering streams (Science Applications, Webb 1981:

Webb goes on to emphasize the high productivity of such a varied ecosystem.


Extinct Species from Southeastern Faunal Localities

- Extinct Land Tortoise
Giant Armadillo
Jefferson's Ground Sloth
Giant Ground Sloth
Harlan's Ground Sloth
Giant Beaver
Giant Capybara
Extinct Wolf
Extinct Jaguar
Extinct Spectacled Bear
Extinct Tapir
Extinct Horse
Extinct Peccary
Extinct Camel
Extinct Bison
Extinct Bog Lemming
Sabertooth Cat
Extinct Peccary
Extinct Capybara

Geochelone crassicutatta
Dasypus bellus
Glyptotherium floridanus
Megalonyx jeffersoni
Eremotherium mirabile
Glossotherium harlani
Castoroides ohioensis
Neochoerus pinckneyi
Canis dirus
Panthera onca augusta
Tremarctos floridana
Mammut americanum
Mammuthus columbi
Tapirus haysii
Equus fraternus
Mylohyus fossilis
Palaeolama mirifica
Bison antiquus
Synaptomis australis
Smilodon sp.
Platygonus compressus
Hydrochoerus holmesi


Writing of the Vero area sites, Webb contrasts the situation with the
Charleston area as follows:

The extensive samples of grazing ungulates and sloths indicate
more grassland than now appears in coastal Florida. Freshwater
taxa are well represented, though not as abundantly as at Edisto.
Overall the Florida setting was somewhat warmer and drier
than the South Carolina setting but the major features of the
predominent megafauna were comparable.

Martin and Webb (1974:135) describe the surroundings of the Devil's Den
deposits at the time of accumulation as follows:

Probably the sinkhole was surrounded by some mesic forest that
gave way to a predominantly xeric savannah .

The results of the investigations along the Tombigbee river in north-
eastern Mississippi suggest that this kaleidoscopic pattern was prevalent
throughout most of the deep south. The evidence from Mississippi also strongly
supports the cool summer/mild winter climatic scenario. The presence of the
Blanding's turtle, the more northerly distributed porcupine, the southern bog
lemming, and the meadow vole suggests cooler conditions. This contrasts with
the giant tortoise, the extinct armadillo and other larger mammals such as the
capybara and the tapir which favor warm, wet environments. Thus the data support
the mild winter/cool summer climatic scenario as well as the extremely patchy
environments which these conditions engendered.

In summary, it is useful to consider these distributions in terms of the
latitudinal gradients which may have existed. In his work, Webb has defined
three broad bioclimatic regions from the late glacial southeastern coastal
plain. These are shown in Figure 3.

The Boreal Faunal Region

This area extends from the north as far south as coastal North Carolina.
This region is characterized by the occurrence of species which currently are
primarily known to have inhabited both tundra and boreal woodland habitats.
The larger vertebrates include thewoolly mammoth, horses, bison and caribou.
Along the coast the range of the walrus extended as far south as North Carolina.
This is a considerable extension of their modern range and the southern limit
today lies at the Gulf of St. Lawrence.

In this boreal faunal region there was probably a vegetational gradient
ranging from tundra at the ice margin and higher elevations and grading down
to an open spruce-fir-pine woodland.

The Temperate Faunal Region

This was an area where the major vegetation consisted of communities of
temperate hardwoods with some boreal conifers interspersed with grasslands.
The zone is estimated to have extended from Cape Fear, North Carolina to the
Charleston, South Carolina area. According to Webb (SAI-Webb 1981:1-73-82):



- - --am, l m m lm ~ r -l gll



(from Webb, 1981)



The vertebrate samples from this region are diverse and include
complex megafaunal samples of both browsers and grazers. The most
striking distinction from the northern fauna is the regular pres-
ence of sizeable percentages of browsers, especially the mastodon.
Other browsers such as the white-tailed deer, elk, and Jeffer-
son's ground sloth also occur in a vegetation (which) was probably
mixed mesophytic woodland.

The grazing elements imply continued large expanses of grasslands as
evident by the presence of bison and horses along with the more southerly
distributed camelids and amphibious rodents such as the capybaras. Webb sees
the overall environmental situation as follows:

Within this compressed temperate zone, there was surely a number
of different habitats in a complex mosaic. Woodland habitats in-
cluded northern conifer forests, northern hardwood forests, oak-
hickory and oak pine forests, among others that have been relatively
well-differentiated in some sites on the basis of local pollen
samples. Open country habitats included extensive coastal grass-
lands and marshlands, as well as more local, often edaphically con-
trolled, scrub habitats.

The Subtropical Faunal Region

This region extended from Charleston, South Carolina down through penin-
sular Florida and westward along the Gulf coastal plain. The vertebrate assem-
blage of this region consisted of the giant ground sloth and the giant tortoises,
both of which are herbivores. The giant tortoise gives conclusive evidence
that winters must have been considerably milder probably without freezing
temperatures as far north as Charleston. Rounding out the faunal list were
the long nosed peccary, the spectacled bear, the southern llamas, the extinct
armadillo, and the giant armadillo. The variety of species represented suggests
that this subtropical region was one of the richest and most diverse environ-
ments in the East.

In summary, it should once more be emphasized that Webb's plot of these
zones tends to obscure the true richness, variety, and diversity of these late
Quaternary environments and should simply serve as a heuristic device to inte-
grate the range of data available.

Cultural Implications of Late Quaternary Environments

The major emphasis in the above descriptions of late Quaternary environ-
ments has been on the fact that there are no modern analogs for the kinds of
environmental situations confronting early man. If we are ever successfully to
take the "emic leap" in attempting to understand these early lifeways we must
develop the capacity to conceptualize daily rounds in ways which are totally
foreign to the modern archeologists' experience, and this is perhaps the greatest
challenge facing us in understanding these early peoples. A number of general-
izations can be made which will make it easier. First, we must consider the
notion of environmental grain which theoretical biologists have used in the
strategic analysis of evolutionary processes and adaptation (Levins 1968).
Environmental grain relates to the size of patches in the environment. A homo-
geneous environment would be seen as coarse grained whereas a heterogeneous


environment would be fine grained. In terms of adaptation, coarse grained
environments lead to greater specialization whereas fine grained environments
lead to generalization. The late glacial and postglacial environments in
Eastern North America, and especially in the Southeast, were highly fine grain-
ed, probably in contrast to the western environments. Thus, the fine grained
nature of the environment demanded a rather generalized adaptive strategy for
these eastern Paleo-Indian populations, again, in contrast to the seemingly more
specialized western tradition.

These Late Quaternary environments were also optimal in terms of Butzer's
(1971) classification system which is partly based on carrying capacity. Be-
cause of this optimality, as well as the fine grained character of the environ-
ment, one can assume that subsistence strategies were secondary determinants
\in settlement patterns and site distributions. As Gardner (1977) has shown
the availability of high quality cryptocrystalline raw material was probably
the primary determinant in these early lifeways.

'The richness of these eastern environments probably in large part accounts
for the huge number of Paleo-Indian finds in Eastern North America, as the recent
compilation of Paleo-Indian points in the journal Archeology of Eastern North
America (1982) has amply demonstrated.

Finally, a rather generalized hunting and foraging strategy with sites
probably tied to mesic habitats can be inferred. Pollen analyses of Clovis
floors at the Thunderbird site (Carbone 1976) suggest that the starchy tubers
Sof pondweed and other aquatic plants were probably a major staple of the Paleo-
Indian diet, and thus when looked at in toto, these mesic habitats satisfy the
largest number of preconditions for a generalized way of life.

References Cited

Archeology of Eastern North America
1982 A Compilation of Fluted Points of Eastern North America by
Count and Distribution: An AENA Project. AENA 10:27-45.

Braun, E. L.
1955 The Phytogeography of Unglaciated Eastern United States and
Its Interpretation. Botanical Review 21:297-375.

Bryson, R. A. and W. M. Wendland
1967 Tantative Climatic Patterns for Some Late Glacial and Post-
glacial Episodes in Central North America. In Life, Land,
Water. Edited by W. Mayer-Oakes, University of Manitoba
Press, Winnipeg, pp. 271-298.

Butzer, Karl
1971 Environment and Archaeology. Aldine, Chicago.

Carbone, Victor A.
1976 Environment and Prehistory in the Shenandoah Valley. Ph.D.
dissertation, Department of Anthropology, Catholic University
of America, Washington, D.C.


Clausen, C. J., A. D. Cohen, C. Emiliani, J. A. Holman and J. J. Stipp
1979 Little Salt Spring: A Unique Underwater Site. Science

Coleman, James M.
1982 Recent Seasonal Rainfall and Temperature Relationships in
Peninsular Florida. Quaternary Research 18:144-151.

Deevey, E. S.
1949 Biogeography of the Pleistocene. Geological Society of
America Bulletin 60:1315-1416.

Delcourt, P. A.
1980 Goshen Springs: Late Quaternary Vegetation Record for
Southern Alabama. Ecology 61:371-386.

Delcourt, P. A. and H. R. Delcourt
1981 Vegetation Maps for Eastern North America: 40,000 yr. B.P.
to the Present. In Geobotany II. Edited by R. C. Romans.
Plenum Publishing Corp.

Dunbar, James
1981 The Effect of Geohydrology and Natural Resource Availability
on Site Utilization at the Fowler Bridge Mastodon Site
(8Hi393c/uw) in Hillsborough County, Florida. Interstate 75
Highway Phase II Archeological Reports.

Frazier, Michael K.
1981 Paleontology. In A Study of Late Quaternary Environments
and Early Man Along the Tombiqbee River, Alabama and
Mississippi. Edited by Guy Muto and Joen Gunn. Report
submitted to the National Park Service under Contract
C-55027(79). IAS, Atlanta.

Gardner, William M.
1977 Flint Run Paleoindian Complex and Its Implications for Eastern
North American Prehistory. In Amerinds and Their Paleoenviron-
ments in Northeastern North America. Edited by W. S. Newman
and Bert Salwen. Annals of the New York Academy of Sciences

Levins, R.

Martin, R. A.

Ray, C. E.

Evolution in Changing Environments: Some Theoretical Explora-
tions. Princeton University Press, Princeton.

and S. D. Webb
Late Pleistocene Mammals from the Devil's Den Fauna, Levy
County. In Pleistocene Mammals of Florida. Edited by S. David
Webb, University Presses of Florida. Gainesville, pp. 114-145.

Pleistocene Mammals from Ladds, Bartow County, Georgia.
Georgia Academy of Science Bulletin 25:120-150.


Roth, J. A.

and J. Laerm
A Late Pleistocene Vertebrate Assemblage from Edisto Island,
South Carolina. Brimleyana (Journal of the North Carolina
State Museum) 3:1-29.

Science Applications, Inc.
1981 A Cultural Resources Survey of the Continental Shelf From
Cape Hatteras to Key West. Volume I: Introduction and
Physical Environment. Report submitted to the Bureau of Land
Management under Contract AA551-CT8-40.

Sheehan, M., D. Whitehead and S. Jackson
1982 Late Quaternary Environmental History of the Richard B.
Russell Multiple Resource Area. Report submitted by the
Thunderbird Research Corporation under contract C-54056
with the National Park Service. IAS, Atlanta.

Solomon, A. M.

Voorhies, M.

Watts, W. A.




Plant Community to Decreased Seasonality During Full Glacial
Time. In Program and Abstracts of the Seventh Biennial Con-
ference of the American Quaternary Association, Seattle,

Pleistocene Vertebrates with Boreal Affinities in the Georgia
Piedmont. Quaternary Research 4:85-93.

The Full Glacial Vegetation of Northwestern Georgia. Ecology

Postglacial and Interglacial Vegetation History of Southern
Georgia and Central Florida. Ecology 52:676-690

A Late Quaternary Record of Vegetation from Lake Annie,
South-Central Florida. Geology 3:344-346.

Late Quaternary Vegetation History at White Pond on the Inner
Coastal Plain of South Carolina. Quaternary Research 13:187-

Watts, W. A. and M. Stuiver
1980 Late Wisconsin Climate of Northern Florida and the Origin of
the Species Rich Deciduous Forest. Science 210:325-327.

Whitehead, D. R.
1965 Palynology and Pleistocene Phytogeography of Englaciated
Eastern North America. In The Quaternary of the United
States. Edited by H. E. Wright, Jr. and D. G. Frey,
pp. 417-432.

Whitehead, D. R. and M. Sheehan
1981 Palynology. In A Study of Late Quaternary Environments and
Early Man Along the Tombigbee River, Alabama and Mississippi.
Edited by G. Muto and J. Gunn.


Williams, S. and J. Stoltman
1965 An Outline of Southeastern United States Prehistory with
Particular Emphasis on the Paleo-Indian Era. In The Quaternary
of the United States. Edited by H. E. Wright, Jr. and D. G.
Frey, pp. 669-684.

Wright, H. E.

Sensitivity of Natural Systems to Climatic Change. In Program
and Abstracts of the Seventh Biennial Conference of the Americai
Quaternary Association, Seattle, Washington.

Wright, H. E. Jr. and D. G. Frey (Editors)
1965 The Quaternary of the United States. Princeton University
Press, Princeton.

Victor A. Carbone
Archeological Services Branch
National Park Service
Atlanta, Georgia 30303


James S. Dunbar and Ben I. Waller


The intent of this paper is to update our previous effort (Waller and
Dunbar 1977) to plot the distribution of Clovis/Suwannee horizon Paleo-Indian
sites in Florida and further define the regions of major settlement and/or
site concentrations. An emerging pattern of site clusters within these regions
is discussed, at least two "types" of site clusters are proposed, and a re-
stricted and somewhat predictable pattern of site concentrations is examined,
occurring either within or directly adjacent to the two Tertiary age karst
regions of Florida.

Since presenting the last distribution of Clovis/Suwannee horizon Paleo-
Indian sites, the catalogued sample of sites has grown and our thanks must be
extended to many professional and avocational archaeologists, as well as to
the private land owners who have freely shared information. The earlier
attempt at plotting site distributions had been simply to plot the number of
sites per county, and even though this scale was highly generalized, a rather
distinct pattern of site occurrences was revealed along the central Florida
limerock ridge extending westward toward the Florida Gulf coast. In the pre-
sent distribution analysis, our controls for plotting have become more refined
in a further attempt to define site patterning.

The overall distribution of sites is based on the distribution of the
diagnostic Clovis, Suwannee, and Simpson projectile points and carved ivory
foreshafts or pins. Originally, butcher-marked bone of Pleistocene megafauna,
such as mammoth, horse and giant land tortoise, had also been included in the
sample; however, evidence from sites such as Bon Tera Farm (Connert 1932:516
and Neill 1953:170-171), Guest Mammoth Kill Site (Rayl 1974:43-45), and Warm
Mineral Springs (Cockrell and Murphy 1978:1-12) suggest that these now extinct
forms survived to be hunted as late as the notched point subphase (Greenbriar-
Bolen) of the Late Paleo-Indian ("Early Archaic"), thus they are considered
nondiagnostic and were deleted from the sample.

Another consideration, which may eventually eliminate part of the sample
presented here, is the possible inclusion of Mid Paleo-Indian diagnostic forms.
Gardner (Gardner and Verrey 1979:13-45), from work conducted at the Thunderbird
Site, Virginia, has shown evidence that suggests that the smaller and thinner
Clovis forms were being intentionally manufactured during a subsequent evolu-
tionary stage of the Eastern American Paleo-Indian Period, or the Mid Paleo-
Indian subphase. This sequence would essentially mirror the established Clovis
to Folsom sequence of the southwestern United States. This evidence, however,
seems too premature and the intentionally manufactured smaller Suwannee and Clovis
forms of Florida are still included in the distribution sample. It should be
mentioned that the smaller forms have a differing pattern of concentration



compared to the distribution of the large Clovis/Suwannee forms. This differ-
ence may represent a contemporaneous method of lithic resource conservation
since the small forms are concentrated in areas where no chert outcrops occur.
This observation would tend to support Goodyear,'s hypothesis (Goodyear 1979:12)
that "among mobile (Paleo-Indian) hunter-gatherers, the use of cryptocrystallin
raw materials (chert, jasper, etc.) is a strategy for creating portable and
flexible technologies to offset geographic incongruences between resources and

To date we have partially catalogued an overall sample of 1296 diagnostic
Clovis/Suwannee artifacts; however, we have not had time to obtain full data
on all of the sample. We have, therefore, broken the sample into two parts.
First is the control sample which represents the fully catalogued sites along
with the number of diagnostics recovered. The control sample will be the main
focus of this distribution analysis and is represented by 126 sites containing
401 diagnostic artifacts.

The second sample represents the control sample plus all the partially
catalogued collections throughout the state. An example of a partially cata-
logued site area would be the lower Santa Fe River basin where a total sample
approaching 500 diagnostics have been reported, yet only 138 diagnostics from
29 sites represents the control sample. Like our other partially catalogued
sample we have not had the time to interview all the informants at length so
as to identify the individual sites and separate the number of diagnostics
recovered from each. For the purposes of this paper it should be remembered
that the term site is used rather loosely, since one site may represent the
location of the recovery of a single diagnostic artifact while another site
may represent a location where over 100 diagnostics have been recovered.

Site Distribution

Since the initial observation that the Clovis/Suwannee phase Paleo-Indian
sites were concentrated along the Ocala Lime Rock Ridge (Waller and Dunbar
1977), research has lead us to expand the regions of site concentration to in-
clude the two near-surface Tertiary age karst regions of Florida. By plotting
the distribution density of diagnostic artifacts on a map also showing the
Northern Panhandle and Central Gulf Coastal karst regions (Fig. 1,- the control
sample; Fig. 2, the total sample), the major concentration of Paleo sites is
obviously concentrated within the regions of Tertiary age limestone outcrops.
Further, many of the sites not located within either karst region are located
near or even in sinkholes, spring caverns or other karst features which pene-
trate through considerable overburden to expose the limestone. Examples of
sites outside of the karst region which occur in or near deep karst features

1. Several sites surrounding Choctawhatchee Bay are in the vicinity
of Choctawhatchee Springs submerged in the bay.
2. A Leon County site discovered when Lake Jackson was dry in the
1930s is adjacent to a sink in the lake's bottom.
3. A site on the University of Florida Agricultural Experimentation
Station (IFAS) property just north of Gainesville is adjacent to
Blue Creek Sink.
4. A site adjacent to Salt Springs in Marion County (St. Johns
River Basin).


5. A site in and adjacent to Silver Glen Springs, Marion County
(St. Johns River Basin).
6. Sites submerged in Tampa Bay where karst is exposed on the
bay bottom and at least one submerged spring is located in
the bay.

Finally, of the sites that occur in outlying areas away from known ex-
posures of Tertiary age limestone, four are located in or adjacent to large
flat-bottomed lakes, two are modern coastal beach line sites and the remaining
four are located in the St. Johns River drainage basin.

Looking at the control sample, 100 sites or 79.4% of the sample occur
within the regions of Tertiary age karst. An additional 16 sites or 12.7%
of the sample occur outside the karst regions but are located near deep karst
features which expose the limestone leaving only 10 sites or 7.9% of the sample
occurring in outlying areas away from any known Tertiary age karst exposure.
It should be mentioned that five sites in the outlying areas may eventually
prove to be Mid Paleo in age and may thus be eliminated from the control

In order to see just how closely this distribution conforms to areas of
karst exposures, the portion of the Central Gulf Coast Karst Region having
the largest site concentration was selected, and more detailed information was
obtained from the Florida Bureau of Geology, Environmental Geology Map Series
(Valdosta, Gainesville, Tarpon Springs and Orlando sheets). This series of
maps accurately depicts where limestone outcrops frequently occur within
the karst regions as opposed to areas where "islands" of plastic sediments
totally bury the limestone. Figure 3 depicts the distribution of sites from
the mid-Suwannee River area southward to the top of Tampa Bay and from the
Gulf coast eastward to just east of the Oklawaha River. Of the 80 sites de-
picted on this map, 78 were correctly plotted on the statewide map (Fig. 1)
as being within the karst region. With the increase of detail, however,
J(Fig. 3) we can see that 67 of the 78 karst region sites occur in subregions
or areas of frequent karst exposures while the other 11 sites are marginal to
the areas of frequent karst exposures. Of the 11 sites marginal to the areas
of surface karst all but one are located either in or adjacent to erosional
features where the limestone is exposed in river valleys, springs, caverns,
or karst sinks. Thus, the close association of Clovis/Suwannee horizon Paleo-
vIndian sites with the Tertiary age karst features of Florida suggests similar
physiographic settings may have been one of the most important selective factors
of the settlement pattern. By looking at groups of sites or site clusters in
and marginal to the Central Gulf Coastal Karst Region (Fig. 4) a pattern of
similar localized physiographic provinces is revealed.

Eleven site clusters in and adjacent to the Central Gulf Coastal Karst
Region are:

1. Lake Jackson/Lake Lafayette Cluster (Leon County) represents four
sites located adjacent to or in the lake bottoms. The lake bottoms
are karst connected via sinkholes and are subject to become dry.
2. Aucilla/Wacissa Cluster (Jefferson and Taylor Counties) is repre-
sented by five sites with others partially catalogued. Sites are
located in or overlooking river channel segments.


* ,





Karst Regions of Florida

one dot equals one diagnostic


Figure 1.

Map showing the distribution density
of the control sample of Clovis/Suwannee
Paleo-Indian diagnostics.

* ",0,




3. Upper Suwannee/Withlacoochee North Upland Valley Cluster (Columbia,
Hamilton, Suwannee and Madison Counties) is represented by four
sites located in these river channels. These sites, even though
located in the river channel, are atypical of most river basin
clusters. There are several large cave systems functioning as
active springs.
4. Waccasassa Cluster (Levy County) is represented by three sites.
Two sites are located in the river channel and one is on land
overlooking a submerged rock shelter-like cave.
5. Lower Santa Fe/Ichetucknee Cluster (Alachua, Columbia, Gilchrist
and Suwannee Counties) represents the largest cluster with 29 sites
and other sites still to be completely catalogued. These sites are
located overlooking the river above the flood plain, in the river's
channel and within the flood plain elevated above normal river
6. Paynes Prairie/Orange Lake Cluster (Alachua and Marion Counties)
is represented by seven sites seemingly widely dispersed but all
located in, or on the margins of, the many flat-bottomed karst
peneplain lakes of the area. These sites are located in, or on,
Paynes Prairie, Levy Lake, Lochloosa Lake, Orange Lake and Johnsons
Lake. All but one of these lakes is known to have been karst con-
nected via sinks which have caused the lakes to be occasionally
drained. Paynes Prairie and Levy Lake are examples of extant,
drained lake bottoms which are now savannas.
7. Silver Springs/Oklawaha Cluster (Marion County) is represented by
seven sites. These sites are located in and overlooking the springs
and river channel and in Silver Springs cave.
8. Chain of Lakes Cluster (Lake County) is represented by three sites
located on the margins of the lakes. A line of several springs runs
from the southwest corner of Lake Apopka then along a slough con-
necting Lake Apopka to Lake Harris and continues along the southern
shore of Lake Harris. A muck filled slough runs westerly to about
a near-surface karst area and Lake Deaton. The sites are located
on Lake Deaton, Harris and Griffin.
9. Tsala Apopka/Panasoffkee/Withlacoochee South Cluster (Marion,
Citrus, and Sumter Counties) is represented by four sites with
several other sites to be catalogued. The sites in this cluster
re either located in the channel of the Withlacoochee River and
its tributary, Rainbow Springs, or occur marginally to Tsala Apopka
and Panasoffkee Lakes. It should be mentioned that Lake Tsala Apopka
is not in an area where the limestone is exposed and the lake bottom
has a relatively thick sequence of sediment covering the limestone.
The Withlacoochee River borders the east side of Tsala Apopka and
represents the line east of which the near-surface karst area re-
sumes. The area encompassed by this cluster is admittedly rather
large, however, since the river basin, lakes and the large swamp
forest (Jumper Creek Swamp) all interconnect forming one broad
lowland depression, the feature is considered as one cluster area.
10. Brooksville/Lecanto Upland Cluster (Hernando and Citrus Counties)
is represented by four sites. These sites are located in the vicinity
of upland dry caves and along the margins of deep karst depressions.
The cluster area is typified by extreme topographic relief for the
central peninsula of Florida, with hills well above 60 meters above
sea level and some karst depressions being only 3 meters above sea


~)&* S


| Tertiary Karst Regions '

one dot equals one diagnostic \

Figure 2. Map showing the distribution density
of the total sample of Clovis/Suwannee
Paleo-Indian diagnostics.




r. I ~~ )


11. The Hillsborough/Harney Cluster (Hillsborough County) is repre-
sented by ten sites. These sites are located in and adjacent to
the channel of the Hillsborough River or marginal to broad wetland
areas such as Harney Flats and the wide swamp forest associated
with the river. The river and associated swamp are in karst areas
whereas the adjacent Harney Flats is not. However, Harney Flats
is karst connected via several springs caverns.

Another site cluster is submerged in Tampa Bay. This cluster had not
been defined adequately until recently (see Goodyear et al., this issue).
This cluster is indicated by artifacts that had been dredged from the Bay.
Future research may be able to specifically document the specific locations
of these dredge sites.

The eleven site clusters described above can be divided into two separate
types. The first and most common type is that associated with large lowland
karst depressions which are currently at or below present ground water tables
(i.e., lakes, rivers, swamps, savannas). The second type of cluster is the
upland karst cluster which is associated with relatively minor hydraulic
features. The upland karst clusters are located in the Upper Suwannee/
Withlacoochee North and Brooksville/Lecanto Upland areas. The other nine
clusters represent examples located in lowland karst depressions.

The two upland karst clusters and the one other isolated upland karst
site west of Gainesville are atypical of the remaining distribution in that
they are located adjacent to minor water sources that are subject to becom-
ing dry during extended drought. In this sense, the upper Suwannee River from
White Springs west to the mouth of the Withlacoochee River and the Withlacoo-
chee and Alapaha Rivers from the Suwannee northward, cannot be considered major
lowland water features since their narrow channels are located in deep valleys
without flood plains. These areas of the rivers are subject to cease flowing
during extended droughts and many sections of the channel become dry (e.g.,
droughts of 1932, 1977 and 1981; see Mansfield 1937:Fig.2). Thus, the major
aspect which differentiates the upland sites from the remaining major distri-
bution of Clovis/Suwannee sites is the lack of expansive lowland surface water
features. There are also two other aspects that all of the upland karst sites
share in common. First they are located in areas where large open cavern
systems are relatively common, and second these areas have very abundant chert

The other nine site clusters in lowland karst depressions are associated
with rivers with wide, low-lying flood plains, large lakes, large wetland
savannas, swamp forest, or any combination of the above. In every instance,
a Paleo cluster surrounding and within a particular lowland karst area is sit-
uated in the same physiographic setting as the others. That is to say, if the
present vegetation and surface water features were removed, the results would
leave a large trough or confined basin-like depression, perforated by sinkholes
and caves. At least some of the higher ground surrounding these lowland clus-
ters serves as the catchment area to funnel rainfall runoff toward the lowland

The lowland karst clusters associated with confined basin-like depressions
are located along the interior border of the Central Gulf Coastal Karst Region.
These clusters include the Lake Jackson, Paynes Prairie, Chain of Lakes, and
Harney Flats areas. In each instance these basins are the least developed in
terms of being karst connected, with anywhere from one to a half dozen or so


Site location

Outlying area
boardcring the
karst region

Areas of Irequent
outcrops in the
karst region

SAreas where "islands"
of plastic sediments cover
the limestone in the karst

map compilation from the Florida
Bureau of Geology environmental
series & the geologic map of


Figure 3. Distribution of Clovis/
Suwannee sites in the f
middle and southern
Central Gulf Coastal
Karst Region.


sinkholes penetrating through clayey bottom sediments to the limestone. When
the sinks in these basins are below the water table (piezometric surface) of
the limestone aquifer (Floridan Aquifer) they hold sufficient water to form
lakes. However, when the piezometric surface of the Floridan Aquifer falls
significantly below these lake basins, the weight of the water in these lakes
causes it to be funneled through the sinks until the lake's water reaches
equilibrium with the piezometric surface of the Floridan Aquifer. The effects
of this phenomenon inmodern times has effectively drained Lake Lafayette, Paynes
Prairie, and Levy Lake, leaving wetland savannas where the lakes once stood
(Williams et al.1977:21-23). Lake Jackson, Lake Lochloosa and Orange Lake have
also been drained by sinks, only to become eventually refilled.

Harney Flats, near Tampa, is the southernmost of the lowland karst basins
and is the only basin adjacent to the coast. As such, it has not drained in
modern times. On the contrary, spring flow from the Floridan Aquifer helps
sustain a water level sufficient to maintain a wetland savanna. Likewise, the
water levels in the Chain of Lakes are augmented by spring discharge (Rosenau
et al.-1977). Within the set of site clusters associated with lowland karst
depressions, those that are centered around discrete basins occur along a zone
directly adjacent to either side of the Central Gulf Coast Karst Region. The
reason that lakes occupy, or once occupied, these basins is tied directly to
the level or piezometric surface of the Floridan Aquifer. These basins also
have a relatively impervious clayeyy) bottom substrate except where sinks
penetrate to the limestone. Although great amounts of water are no longer
held by the drained basins, the nature of the bottom sediments tends to perch
sufficient water to maintain a wetland savanna environment (Williams et al.1977).
These savanna environments are climate dependent and sustained by modern rain-
fall amounts (Brooks 1967).

A second subtype of site clusters located in lowland karst areas is the
trough-like depressions which are associated with karst river basin segments
having relatively wide, low flood plains and other associated lowland features.
They are the most developed in terms of irregular karst topography with numer-
ous large spring caverns, irregular crevasses and abrupt cliffs. Generally,
the river channels are entrenched well into the limestone, giving the impres-
sion of a sheer rock-walled arroyo. In some parts of the typical river channel
a rocky bottom juts upward, forming a rapid or ripple area shallow enough to
afford an easy river crossing area for wild game. These game trail crossing
areas have often produced major Paleo-Indian sites (Waller 1969:37-39). The
flood plains of these cluster areas are elevated above the limestone entrenched
channels and can be considered nonkarst in that these broad low-lying areas
are generally blanketed by clay sediments. Clayey river deposits tend to perch
pockets of water above the limestone and retain soil moisture for extended
periods of time. Karst features are also found in the flood plain and, in
extreme examples such as the Santa Fe River, rock cliffs 3 to 5 meters tall
skirt the flood plain as vivid evidence of its Early Pleistocene origins
(Edwards 1948:125-132). Site clusters in this type of lowland area frequently
extend into the heart of the Gulf Coastal Karst Region. This distribution
accounts for sites located in the Lower Aucilla, Waccasassa, Lower Santa Fe,
Silver Springs and Hillsborough River Clusters and sites occurring in and along
the Pithlachascotee, Suwannee, Steinhatchee, St. Marks, Wakulla and Crystal

Unlike the textbook model of most rivers, those that flow through the
karst regions owe their sustained flow to the Floridan Aquifer. During the



r, 1

:::: "
i:: \ .~'~,::'.';
i; '~: -~
E' :" ~
^ -~

1.Lake Jackson/
Lafayette area
2.Aucilla/Wacissa area
3.Upper Suwannee/
Withlacoochee North area
4.Waccasassa area
5.Lower Santa Fe/
Ichetucknee area
6.Paynes Prairie/Orange
Lake area
7.Silver Springs/Oklawaha
8.Chain of Lakes area
9.Tsala Apopka/Withlacoochee
10.Brooksville/Lecanto area
11.Hillsborough/Harney area

O. Central Gulf Coastal
Karst Region

: '

Figure 4.

The location of and
approximate area encom-
passed by the Paleo-Indian
site clusters in and adja-
cent to the Central Gulf
Coastal Karst Region.

-., ..


period of recorded data, the piezometric surface of the Floridan Aquifer has
remained relatively stable in its relationship to the surface of lowland karst
rivers. During the high water stages, the flood water rises above the pie-
zometric surface so that the many sinks, and in some cases, springs, through-
out the channel and flood plain, take in water in an attempt to reach equi-
librium with the aquifer's surface. During extended droughts the surface run-
off from rain no longer supplements the river's flow and the aquifer maintains
the river's base flow via springs discharging into the channel (Puri et al.
1967:17). In other words, the piezometric surface of the Floridan Aquifer
is elevated above the bottom of the river channel. If the piezometric surface
of the aquifer dropped below the level of the river's channel as drought ensued,
the channel would become partially or totally dry as do the Alapaha and
northern Withlacoochee Rivers.

Paleo-Indian site clusters centered around and in the karst river channels
are similar to those centered around and in the karst lake basin depressions
in that the position of the aquifer's surface determines whether or not they
are capable to holding water for extended periods of time. They are also
similar in that a major portion of these low-lying areas are blanketed by
relatively impervious sediments capable of perching small, climate dependent
ponds and can retain soil moisture over extended periods. The major difference
between the karst river and lake lowland clusters is that the rivers have fre-
quently occurring karst features and the lake basins do not. This fact alone
implies the river clusters should have the largest selection of chert resources
because the limestone is more frequently exposed.

Summary and Conclusions

An analysis of the distribution of Clovis/Suwannee horizon Paleo-Indian
sites has shown that, with the sample thus far plotted, the major concentration
of sites is located within the two Tertiary age karst outcrop regions of
Florida. It has also shown that a major portion of sites located outside
these karst regions is situated in the vicinity of deep sinks and spring
caverns which penetrate through other sediments to expose the Tertiary age
limestones below. If viewed in total, the number of sites associated with
and/or having close access to open karst features is 116 or 92% of the control
sample. One aspect of the data we have chosen not to emphasize in the main
body of the paper is the generalized distribution of the partially catalogued
sample included with the control sample (Fig. 2). This overall sample compared
with the generalized border of the Panhandle and Gulf Coastal Karst Regions
shows a major increase occurring within the karst regions and minor additions
outside the border.

By examining the Central Gulf Coastal Karst Region in a closer manner,
the control sample shows that the Clovis/Suwannee sites most frequently occur
in areas of frequent karst outcrop and that two major types of site clusters
seem to occur. First and most frequent are the clusters centered in and around
lowland karst depressions. Second and least prevalent are the upland site
clusters located in areas of abundant chert outcrops where large caves and
minor climate dependent water sources occur.

Sites in the lowland karst areas can be further divided into two subtypes,
those associated with extant and extinct karst lake basins and those asso-
ciated with karst river basin having wide low-lying flood plains. The karst


lake areas tend to be located marginal to the inland border of the Gulf Coastal
Karst Region while the river basin areas tend to extend well inside the karst
regions. Both have the similar modern geohydrologic characteristics of first
being dependent upon the limestone aquifer's piezometric surface and secondly
on climate. Perhaps the greatest difference between the two lowland clusters
is that the river basins are more likely to have abundant chert resources due
to the more frequent outcrops of limestone.

This paper has attempted to describe objectively the geographic distribu-
tion of Clovis/Suwannee age Paleo-Indian sites from a noncultural standpoint.
We have also tried to ignore nonconstant natural phenomena such as vegetation
patterns in order that the site concentrations might be compared to the more
time stable constants such as surface geology. We hope to continue adding
additional data to this ongoing distribution and encourage anyone having in-
formation regarding uncatalogued sites to share their information with us.

References Cited

Brooks, H. K.
1967 Rate of Solution of Limestone in the Karst Terrain of Florida.
Florida Water Resources Research Center, Publication No. 6.

Cockrell, W. A. and L. Murphy
1978 Pleistocene Man in Florida. Archaeology of Eastern North
America Vol. 6 Summer.

Connert, Jack H.
1932 Recent Find of Mammoth Remains in the Quaternary of Florida,
Together with Arrow-Head. Science 75:516.

Edwards, Richard A.
1948 An Abandoned Valley Near High Springs, Florida. The Quarterly
Journal of the Florida Academy of Sciences 11(4):125-132.

Gardner, William M. and R. A. Verrey
1979 Typology and Chronology of Fluted Points from the Flint Run
Area. Pennsylvania Archaeologist 49(1):13-45.

Goodyear, Albert C.
1979 A Hypothesis for the Use of Cryptocrystalline Raw Materials
Among Paleo-Indian Groups of North America. Research Manuscript
Series No. 156. Institute of Archaeology and Anthropology,
University of South Carolina.

Mansfield, W. C.
1937 Mollusks of the Tampa and Suwannee Limestones of Florida.
Florida Geological Survey. Bulletin No. 15. Tallahassee.

Neill, W. T.
1953 Notes on the Supposed Association of Artifacts and Extinct
Vertebrates in Flagler County, Florida. American Antiquity

Puri, H. S., J. W. Yong and W. R. Oglesby
1967 Geology of Dixie and Gilchrist Counties, Florida. Florida
Geological Survey, Geological Bulletin No. 49, Tallahassee,


Rayl, S. L.
1974 A Paleo-Indian Mammoth Kill Site Near Silver Springs, Florida.
Unpublished Masters Thesis. Northern Arizona University.

Rosenau, J. D., G. L. Faulkner, C. W. Hendry, Jr., and R. W. Hull
1977 Springs of Florida. Florida Department of Natural Resources,
Bureau of Geology Bulletin #31 (Revised), Tallahassee,

Waller, B. I.
1969 Paleo-Indian and Other Artifacts from a Florida Stream Bed. The
Florida Anthropologist 22(1):37-39.

Waller, Ben I. and James Dunbar
1977 Distribution of Paleo-Indian Projectiles in Florida. The
Florida Anthropologist 30(2):79-80.

Williams, K. E., D. Nicol and A. F. Randazzo
1977 The Geology of the Western Part of Alachua County, Florida.
Bureau of Geology. Report of Investigations No. 85.

James S. Dunbar
Bureau of Historic Sites and Properties
R. A. Gray Building
Tallahassee, Florida

Ben I. Waller
Ocala, Florida


On January 15, 1983, James Dunbar
conducted an interview with Ben Waller,
an avocational underwater archaeologist
in Ocala, Florida. Mr. Waller has more
practical experience investigating and
observing innundated Clovis/Suwannee age
Paleo-Indian sites than any other under-
water archaeologist in the state (pro-
fessional or avocational). He has spent
thousands of hours of dive time pursuing
this endeavor. Mr. Waller's experience
as a former animal trainer with Ross
Allen's Reptile Institute, his personal
study and his varied experience as a
diver have given him insight and a field
savey which have led to numerous important

* I

Figure 1. Ben Waller. Photograph
courtesy of Don Serbousek.

FA: When did you first begin to Scuba dive?

BW: I started diving in the late 1940s as a spear fisherman. Then I stopped
spearing and began photographing fish. Eventually, I became interested
in clear water photography which led me to cave diving. We started find-
ing things that were unique in the caves--stalactites that shouldn't have
been there, animal remains--the kinds of things that sparked my curiosity.

FA: How did the early Scuba equipment compare to today's standards?

BW: My first regulator I made from instructions in a Popular Science magazine
which showed you how to make it out of a Maxwell House coffee can. The
one I made wasn't real successful.

That must have been in 1948 or 1949. I bought my first tank and regulator
in 1950. It was the very last part of 1950 and finally several of them
came in at the same time. I wanted to buy the first one I ever saw and
as I remember, that first regulator cost $25.

FA: When did you first begin to notice and investigate underwater sites in







VOL. 36 NOS. 1-2


BW: We were interested in and trying to explore caves, when a man with the
State geological survey asked me to let him know where the caves were
and what was in them. I didn't collect at the time.

One of my early dives was at Little Salt Springs. I can remember we
were decompressing from that dive when I began to hear a metal-like
sound--dink, dink, dink. It was one of the other divers clanking two
fossil bones together. Gosh, that's embarrassing to think of now with
all the very good work that is being done on the site recently.

Soon after, we began finding more and more things underwater and we
contacted Dr. Clayton Ray, then with the Florida State Museum in Gaines-
ville. Dr. Ray convinced us that we should be working with the Museu.
recording the things that we were recovering underwater.

I eventually worked on diving expeditions with Bob Allen and Dr. Goggin
of the University of Florida. Since then I have worked with Dr. Webb,
Dr. Hoffman, Ripley Bullen, and a little with Dr. Purdy, even though
the latter two were not divers.

FA: In the early days of diving, how many others were involved in looking
at underwater archaeological sites?

BW: I dove the Santa Fe River for more than a year before I ever saw another
diver in the water. We saw other Scuba divers in the river for the first
time in 1958. We'd found a fossil site that was pretty nice and we went
out there one day and there were some divers over the site. We thought
that the word was out. There were some girls up in the boat, two guys
in the water, and one of them started going down and we were sitting
there just crestfallen because we thought we had lost our site. However,
after a few minutes one of the divers came up to the surface and said,
"Oh, my gosh, don't try to go down there; I went down 110 feet and never
saw the bottom." I knew then we were all right because the actual bottom
is only 22 feet deep. It was just dark water.

FA: Did you, and do you, ever share information with other people interested
in diving on underwater sites to get a better understanding of the sites?
Did you make agreements with other divers and the like?

BW: Yes, Jarl Malwin and I have been best friends and diving buddies for
years and we've spent hundreds and hundreds of hours working together.
Jarl is one of the most honorable people I've ever dove with and we had
a kind of unwritten agreement. There was never any thought of going to
another person's site. Jarl and I knew that Tom Gouchnour was diving a
certain area of the Santa Fe River, and we never once dove that area;
and likewise he did not dive in our areas. There was enough river that
you didn't have to go claim jumping; so, we talked freely about sites
and the things that we were finding. It's not that way now because
everybody is so darn secretive about it. However, we have freely given
information and I prefer to do it that way; like we are doing here.
Because with this kind of exchange there is information that may assist
all of us.

FA: What are some of the examples of the types of projects you've participated


W: I did work on the Guest Mammoth Kill Site at Silver Springs. Bill Guest
was a fellow here in Ocala that wanted some antique bottles that I had
and he knew I was interested in fossils. One day he came to me with a
lot of elephant bones he wanted to trade for the bottles. They were just
broken remains, but I decided I would trade him the bottles if he would
show me where the site was. We went out to the site and found a set of
beautifully preserved mammoth jaws and two bone pins. We then contacted
Bates Littlehales with National Geographic magazine, who had asked us to
find something that he could photograph in clear water. The day we first
showed him the site we found another bone pin and a broken banner stone.
We stopped handfanning and tried to get an archaeologist to come and take
over the excavation. However, no one had time. Eventually, Charlie
Hoffman from Northern Arizona University did the excavation; almost three
years later. You can imagine how difficult it was to leave it for almost
three years with jungle cruise boats going by there about eight times a
day in some cases.

FA: What are your major interests in the field of archaeology?

BW: I'm interested in the animals that were hunted by the first people that
came here--the Paleo-Indians. I'm torn between paleontology and archeo-
logy, but I don't think I have to be because what it boils down to is
that I am most interested in the time when now-extinct animals and
man coexisted. There is more and more evidence being discovered to
indicate that Paleo man was close to the extinct animals that lived
here. That is most pleasing to me because it creates a situation where
paleontology and archaeology become a bit of the same.

FA: What kinds of experiences have you had which have helped you to interpret
underwater sites?

BW: I was part of a body recovery team for a long time and there was a rule
of thumb that if you knew where a victim had entered the water, the body
coul8 be found within one and one-half times the depth of the water. If
a victim dropped in 30 feet of water you could expect to find the body
within 45 feet from where he originally fell in.

FA: Current or no current?

BW: It didn't make any difference. That was the rule of thumb; even in the
river it turned out to be that way. It didn't exactly hold true, but the
whole point is that the bodies didn't wash very far even though a body
is not hydrodynamically smooth and considering the grasping and clutching
that must have gone on. It used to please us a lot when everyone in boats
would say, "Hey, I'll go downstream to the bends and catch him." We would
have to wait for the boats to clear out so we did not have to worry about
the outboard propellers. Then we would enter the water and in several
instances find the victim within just a few minutes.

FA: Did you begin to notice the same kind of things in underwater sites?

BW: Yes, I kept hearing a theory that artifacts dropped in at point A and
washed down to point Z; therefore, they had meaningless context. I didn't
feel that was true because normally when we found sites we would find
concentrations of points and artifacts within a confined area. If arti-
facts had washed randomly down river they would have been scattered up


and down the river and not concentrated. I worked on a project with
the Department of Paleontology of the Florida State Museum where we
uncovered two million-year-old antelopes still partially articulated.
Now that's a pretty good argument against the idea of downstream washing
of the much later period artifacts. In another instance we found the
carcass of a cow that had fallen in the river. Nobody wanted to be
around that site because who would want all that rotted flesh. However,
I felt like it was a good opportunity to occasionally observe the remains
through time. There were a number of people who knew where the remains
were and each week or so the remains were observed. Finally, over a
year had passed and the bones were still there on the ledge just where
they were originally discovered. I even remember when the jaws disappear-
ed. One was discovered up on the bank, and we knew that somebody had
taken them and discarded one of them.

We did another thing that was interesting and it's something that needs
to be followed up. We took 500 flint spalls, painted them and then
placed them in a very fast moving area of the river to check if hori-
zontal movement occurred. The degree of movement over two years was
almost nothing. Finally, we spent one day recovering them and removed
about 375 spalls. We would have probably gotten more if we had had
more time to remove the sand cover. We noticed that the fast-moving
water had removed much of the paint. In another experiment we did some-
thing else that I thought was on the right track. We recovered all the
bones from two different kill sites in the river. Now that's a job!
You don't do that in one day. The sites had just thousands and thousands
of bone fragments, but we picked them all up and took them to the lab,
then in the Segal Building. We then separated the bones according to
the various types; such as deer, horse,alligator, elephants, turtles,
etc., and then tried to identify the minimum number of individuals repre-

We found that manatee was the most common animal at the early sites.
'The second most common bone was elephant and then there was horse in that
order. Deer came in fourth or fifth. At Archaic sites deer invariably
was first and turtle was second.

FA: Do you know if the manatee bone was of late Pleistocene age or could it
have been earlier fossilized material?

BW: I don't really know although I couldn't believe it when a paper came
out stating that manatee bones had not been used by Indians. I think it
would have been a lot easier to hunt manatees than elephants. When you
find manatee ribs and flint spalls together, it is an almost sure key
that you have located a Paleo-Indian site. Also, I have found tools
carved from manatee bone such as a beautiful banner stone out of a manatee

FA: Do you have any carved manatee bone that you think is Paleo-Indian?

BW: Yes, I have a "clubhead" which has been described as being Paleo-Indian

FA: Let's move to questions that deal with Paleo-Indian river sites. How
would you characterize the typical Paleo-Indian river channel site or
are there different types?


BW: As I have previously described, there are the types of sites containing
much animal bone. I believe some of these are the kill sites but you
don't always find points at sites that produce fossil bone. In other
places you just have large out-croppings of flint and almost no fossil
there at all, so obviously, I don't think those are kill sites. Basically,
the places that I was describing as kill sites are places that you hate
to see if you drive an outboard. In other words, in areas of swift
water rapids, areas that are shallow that could have served as a river
ford for game animals. I think animals were killed and butchered on
these sites although I don't think we have yet made an adequate defini-
tion for these sites. Most of the time you will find areas of the kill
sites where rock juts out into the river and for some reason you find
more tools than points. I believe that as the animals waded across,
their legs were impeded by the water and they became easy hunting prey.
As a result, the river rapid areas are good places to look. We spent
hours and hours finding points in places where our tanks weren't even
under water.

FA: In other words, these sites occur in areas of irregular topography?

BW: Yes.

FA: Which river has produced the largest concentration of Paleo-Indian sites
in the state?

BW: Well, I don't think there is any doubt that the Santa Fe River has
produced more than all the other Florida rivers put together. For
example, I always wondered why large multicomponent sites in the Suwannee
River only produced low percentages of Paleo-Indian material. In the
Santa Fe River, a tributary of the Suwannee, similar sites have high
percentages of Paleo material.

FA: Where is the single largest Paleo-Indian river site you have investigated
in terms of the number of diagnostic artifacts recovered?

BW: Without any question, the biggest site is the Santa Fe 1 site. It-is the
most significant Paleo site I've ever seen. It is a classic kill site
where a shallow rapids area gives way and the channel bottom drops off
to deep water. There is a tremendous amount of debitage and points.
At one time I kept track of the number of points from the site, and there
would be over one hundred whole and broken Paleo points which include
Clovis/Suwannee and Simpson types in various collections. I have per-
sonally found 27 complete Paleo-Indian points from this site which is
only about 200 feet long by 100 feet wide. About 50 complete points
have been found altogether. Another interesting aspect of the site is
that relatively few Bolen points have been recovered.

FA: It is my understanding that sites have been discovered that have produced
a few diagnostic points along with the remains of what appeared to be a
single late pleistocene species (i.e., horse, mammoth, etc.). These
sites may be surface finds in river channels; however, late cultural
material or the presence of other fossil bone is absent suggesting the
possibility of an isolated kill site. Can you describe as many of these
sites as possible?


BW: Yes, of course, there is the Guest Mammoth Kill Site but the one I
enjoyed was the horse site at the junction of the Suwannee and Santa
Fe Rivers--that was my favorite site.

FA: Can you describe that site a little?

BW: Yes, in 1963 we had been finding a number of points there--over one
hundred that day--when we came across the front part of a Pleistocene
horse snout eroding out a submerged portion of the north bank of the
river. The maxilla is not a part of the skull that you often find
preserved, so we collected it, and then uncovered the left mandible.
Then we started finding articulated vertebra and realized that it might
be a significant fossil site. I tried laying the skull pieces back down
in the exact same positions we had seen them. After a little more work
we suddenly uncovered a Suwannee point in the neck region pointing down
towards the legs. Later we found another one not quite a foot away from
the first and still another about 22 inches from the first two. All
three were in the neck region. Finally, we found a fourth around the
pelvic girdle area. We then covered the site back up and tried to get a
professional to come to the site, but nobody could; that was on a
Saturday. The next week it was the same story. Nobody came. On the
following week it started raining and we were afraid that if the water
got dark,by the time it cleared back up, the whole bank in the site area
might have collapsed, so we removed the horse. Bear in mind that the
horse site was but part of a much larger site (8Su2/8Gil), but in that
particular area the only things found were the partially articulated
horse remains with the four Suwannee points.

In another case, we found the isolated remains of an elephant in the
Waccasassa River with nothing else around. That is, nothing until we
noticed a rib with cut marks and upon turning the rib over we found
four inches of a Clovis-like base.

.Other faunal sites with associated points is another in the Santa Fe
River where elephant remains were found eroding out of the bank. We
found one whole and one broken Paleo point and those were the only two
pieces of flint in the area. We also found a number of bone pins at
this site. Bob Alan also found a mammoth site in the Santa Fe River
with one point and the one animal. Then we found a tapir site in the
Oklawaha River, but it was not a Paleo site, or I should say, that the
point found with the remains is known to us as the Wacissa type. That's
the only time I've ever seen any point with a tapir. There was a site
in the Steinhatchee which had one elephant with one point. But perhaps
more significant were the other sites with a tremendous quantity of bone
pins associated with elephant remains. These sites were in the Stein-
hatchee River and Aucilla River. I have seen two of the elephant sites
and have heard of a third one. The people who found the sites recover-
ed tremendous quantities of these pins.

Then, of course, there is the Alexon Bison Site in the Wacissa River.
This site produced the bones of an extinct bison with a point fragment
imbedded in the skull. That was the first time I had ever seen anything
like that. That's really mind-boggling.


FA: You mentioned bone pins in conjunction with Paleo sites. Do you think
bone pins are an important Paleo tool?

BW: Yes. Of course, I've been pushing that idea for years. I totally
believe that they are the most common Paleo tool found in Florida.
When I say pins, I am generally speaking of the double-pointed variety.

FA: Do you think they may have been used as projectiles?

BW: I am uncertain of their use, however, I would like to say that the
people who have seen the Steinhatchee elephant sites agree that bone
pins were common in early tool kits. However, there are a lot of people
that just won't buy the idea. I guess years ago the earliest date for
their manufacture was believed to be five to six thousand years ago;
therefore, even the obvious can be questioned. But I believe that the
bone points are in situ on Paleo sites and are part of the Paleo tool

FA: Do the double-pointed bone pins have any distinctive features like basal
roughening that might separate them from those of a later age?

BW: I don't know, but the ones I'm speaking of are made of the cannon bone
of the deer. I think four pins can be made from each cannon bone. I
think they cut off the epiphysis of the bone, split the remaining long
shaft into four sections, and sharpened both ends of each of the sections.

FA: You also mentioned butcher-marked Pleistocene bone from one of the river
sites; how many have you seen from Florida?

BW: Well, I think it's safe to say that there are at least eleven Pleistocene
animal bones displaying butchering marks. I was fortunate to have dis-
covered the first one to be reported (Editor: see Bullen, Webb and
Waller, American Antiquity 35(2):203-205). Tne specimen was a vertebral
spine of a large mammoth recovered from the Santa Fe River. What was
most significant, as Dr. Webb was quick to realize, is that the cut marks
were in strategic places; for example, ligament attachments-were cut
away to separate the meat from the bone. Since then, there have been
other specimens found with even more distinctive cut marks.

FA: What are some of the traits that you would look for that would charac-
terize a river bottom Paleo site?

BW: Well, as I mentioned earlier, the occurrence of manatee rib and flint
spalls is diagnostic. In the Santa Fe River you quite often notice and
come across a site that has chert which is not extremely translucent,
nor of the highest grade. I don't know my stone that well, so just
bear with me in this poor definition of it, but sometimes the chert has
multicolors that go through it, like white, yellow, etc. Anyway, I
think that type of stone is frequently found at the early sites.

FA: How frequently do you find good bone preservation at the river Paleo

BW: Very often. Many paleontologists feel that some of the best preserved
bone in the United States has come out of the Florida rivers. I think
the rivers' tanic acid is one of the big reasons for this excellent


FA: What about preservation of the fragile organic materials, such as wood,
etc., that date to the Paleo-Indian period?

BW: The objects that I really consider to be wooden artifacts are rare and
are generally more recent. I do think it is possible to find early wood
artifacts because of the wood preservation at the 100,000 year old sloth
site in Daytona Beach. At that site, logs were strewn throughout the
area and had to be removed to get at the fossil remains.

FA: Have you ever seen a component of a Paleo river channel site eroding
out of fresh-water marl other than the Guest site?

BW: In the Santa Fe there is a real fine-grained white sand, usually up next
to a bank, and I have seen Paleo material come from that type of sediment.
There is usually very little bone at these sites and I would not call
them kill sites. Near Wilson Springs is a good area to find this
setting. You can go up near the bank adjacent to a channel site and
.hand fan and find points in this extremely fine-grained sand.

FA: Do the river channel sites occur at great depth or are they shallow or
do they vary from site to site?

BW: I've never seen any at great depths in the river; although, we've seen
Paleo points in deep holes in the rivers.

FA: Deep holes?

BW: Yes, 60 feet or more. One of these had several points. Of course, not
many rivers have sixty-foot deep holes. Most of the holes in rivers
that we think of as so deep are only 25 feet deep when checked with a
depth gauge.

FA: What is the average depth of the Paleo-Indian river channel sites?

BW: Most of the Paleo sites are in 12 to 20 feet of water, but probably
averaging closer to 12 feet.

FA: Let's move to a few questions regarding Paleo sites submerged in Florida
lakes. Have you ever worked on a Paleo lake site?

BW: I dove in a number of lake sites but didn't find any good Paleo material.
There was one site--the Nalcrest Site--that Rip Bullen and I visited.
Two elderly gentlemen had pretty much worked out the site by the time I
looked at it and I was really amazed to find out that they had recovered
the artifacts by using rakes. They had done such a good job of raking
the site--I was just astounded. They didn't have the opportunity to use
Scuba gear so they waded out there with a snorkel and a rake. The water
was often over their heads so they would rake and then hold themselves
on the surface with the rake handle. They did some hairy diving not to
be using Scuba gear. When the rake touched rock they would hear it and
reach down to pick it up. After that, it was a dash back over to shal-
lower water where they could get a breath of air. They used a criss-cross
pattern and I don't think the tiniest flake survived. They did a tremen-
dous job of surface collecting and I always had to admire those men.


FA: Have you ever investigated an offshore Atlantic or Gulf Paleo site or
fossil locality?

BW: Yes, I have visited fossil localities. We found elephant remains con-
siderable distances out in the Gulf of Mexico--six miles out. Usually
you get into an area. where you just see fragments of bones that are
highly covered with barnacles. There is a lot of terrestrial material,
for example, off Venice.

FA: Late Pleistocene age fossil bone?

BW: Yes, one nice Paleo point came from the beach as well.

FA: What avenues of research would you like to see take place in the future?

BW: First of all, I think that many people need to get over the idea that
Florida rivers wash artifacts great distances downstream. The classic
argument that modern refuse--like Coke bottles--can be found together
with Paleo points does not negate the fact that artifacts from a river
site were originally deposited within that site's boundaries and not
ten miles upstream. I think the river banks adjacent to the extra good
sites should be tested. Sometime in the future the site we have all been
waiting for will be found. Unfortunately, no systematic search effort
has ever been attempted.

Personally, I would like to continue plotting the distribution of these
early sites. I think there will be a staggering number of sites even-
tually plotted once more people start sharing that kind of information.
People need to share this information and professionals and avocational
archaeologists need to do more active field work because I think we're
going to flip when we see the amount of material and good information
that is retrieved.


Albert C. Goodyear, Sam B. Upchurch, Mark J. Brooks, and Nancy N. Goodyear

In Florida, the Suwannee point and other early lanceolate points are the
most common and reliable archeological indicators of Early Man. Although there
may have been earlier human populations in what is now the state of Florida,
the Paleo-Indian groups that made Suwannee points constitute the first unam-
biguous and prominent cultural horizon in the same way Clovis and other fluted
point complexes do elsewhere in North America. Compared to the rest of the
country, the study of Paleo-Indian in Florida is somewhat unique in that the
typically unfluted Suwannee is the most common diagnostic hafted biface form.
In addition, the paleoenvironmental and hydrological conditions to Florida at
the end of the Pleistocene and early Holocene resulting from a greatly lowered
sea level, provide for significant human ecological relationships due to
geographic restrictions in surface water (e.g., Clausen et al. 1979). One area
that has long been known to produce Paleo-Indian and Archaic artifacts from
sites now drowned by the Gulf of Mexico is that of Tampa Bay (Warren 1962, 1964;
Goodyear and Warren 1972; Milanich and Fairbanks 1980). The strong association
between Suwannee points and aquatic environments, such as rivers, springs, lakes
and coastal bays, has been repeatedly noted by Florida researchers (Warren 1962;
Neill 1964: Waller 1970; Waller and Dunbar 1977).

The present paper analyzes a sample of Suwannee and technologically re-
lated lanceolate points from the Tampa Bay region as an initial phase of a
longer range study aimed at understanding late Pleistocene-early Holocene
aboriginal adaptations along the west coast of Florida. This broader research
plan seeks to investigate environmental dynamics related to climatic changes
and sea level rise, particularly the development of aquatic environments, and
associated changes in aboriginal lithic technologies.

From an archeological standpoint, we are interested in understanding
variability and change in chipped stone technologies spanning from Paleo-Indian
through Archaic times and the raw material characteristics of these technol-
ogies. The recent development by Upchurch (Upchurch, Strom and Nuckels 1981)
of a geological method of chert source identification has made lithic raw
material studies feasible in Florida. Specifically, data are presented and
analyzed for Paleo-Indian lanceolates regarding: 1) techno-functional patterns,
i.e., how they were manufactured and may have been used, and 2) the raw material
type of each specimen based on the Upchurch classification (Upchurch et al.
1981), and trends in raw material selection as indicated by the various lance-
olate types.

The Nature of the Sample

A total of 27 specimens was obtained for measurement, photography, raw
material analysis and other observations. As can be seen in Figure 1, the 27
examples are drawn from several locations dispersed throughout the area which
should enhance their value as a regional sample. Because all of these are



borrowed from private collections or otherwise not accessioned to scientific
institutions, for purposes of this study, each has been given a number which
corresponds to a photograph (Figs. 2,3,4). These specimen numbers are also
used to show the location where each specimen was found in the Tampa Bay region
(Fig. 1). The 27 specimens and their proveniences of recovery have all been
verified to the satisfaction of the authors. Although it was our goal to
physically study as large a sample as possible, the present group does not
exhaust the known cases of Suwannee point finds in the area.

In order to document to the fullest extent the known occurrences of early
lanceolates in this region, the following cases should be added to the regional
inventory. Most of these are indicated in Figure 1 by an asterisk.

Additional finds include one found by Francis Bushnell from a fill (Pi98)
pumped up from Tampa Bay on the extreme southern end of Pinellas County imme-
diately east of the Sunshine Skyway bridge (U.S. 19) (Fig. 1); one specimen
excavated from the Safety Harbor phase village of Maximo Point (Pil9) by Francis
Bushnell (Bushnell 1962: Plate IV, top row, eighth specimen from right) (Fig. 1)
two Suwannees found by Phil Jordan at an unnumbered terrestrial site known
locally as Japanese Gardens Mobile Home Park near where U.S. 19 intersects
highway 60 (Fig. 1, between 18 and 26); one Suwannee found by Albert Goodyear
from a dredged canal and fill on the western shore of Old Tampa Bay (Fig. 1,
near 26); one Suwannee found by Lyman O. Warren which came from fill material
pumped out of St. Josephs Sound to make Caladesi Causeway which connects Dunedin
Beach on Honeymoon Island with the City of Dunedin (Fig. 1), a photograph of
which was published by Warren (1968:Fig. 2, center); two Suwannees from the
site of Fish Creek (Hil05), which was a fill on the eastern shore of Old Tampa
Bay north of the interstate 75 bridge (Fig. 1, 16), photographs of which were
published by Karklins (1970:Fig. 3, a,e); and three Suwannees found in the
oyster shell deposits dredged commercially from Tampa Bay. The latter three
specimens and the circumstances of their finds are discussed by Goodyear and
Warren (1972:58). One of the three, the one found by Warren, is illustrated
by Bullen (1968:Fig. 1, f). The exact original locations of these oyster shell
specimens, shown by asterisks in Figure 1, as well as specimens 4 and 15
(Fig. 1) of this study are unknown; but it is likely that they all came from
one or more oyster shell rich areas of the drowned channels of the Alafia and
Little Manatee Rivers (see Goodyear and Warren 1972:Fig. 1).

Gordon Willey (1949), in his Archeology of the Florida Gulf Coast, illus-
trates two .(Plate 54B, h,j) Suwannees from Parrish Mound 3, and a possible
example from the Safety Harbor site (Pi2) (Plate 55, i). All of these speci-
mens, as they were found culturally redeposited in late prehistoric Safety
Harbor phase sites, should be added the one just mentioned from the Maximo
Point site excavated by Bushnell, plus specimen 3 of this study, all indicate
that the collecting of artifacts in the Tampa Bay area goes back several cen-
turies. Re-use of older hafted bifaces by Safety Harbor phase peoples was

Finally, the recent excavations at Harney Flats (Fig. 1) (see Daniel and
Wisenbaker, this issue) yielded 14 Suwannee points and approximately that many
more reforms and blanks. Counting the above listed cases and the 28 from
Harney Flats, a total of 42 specimens beyond the 27 presented here can be
noted for a grand total of 69 points and reforms for the Tampa Bay region.


Figure 1.

Map showing the study area and locations of recovery
of Paleo-Indian lanceolates.


Figure 2. Large Suwannee-type lanceolates. a-#18; b-#17; c-#5;
d-#15; e-#13; f-#16; g-#27; h-#9; i-#21; j-#14; k-#23.



e f

0 1 2 in.
0 5 cm.

Figure 3. The smaller Suwannee-type and Simpson-type lanceo-
lates. a-#10; b-#11; c-#7; d-#20; e-#22; f-#6;
g-#8; h-#2; i-#4; j-#1; k-#12.


0 I2 in.
0 5 cm.

Figure 4. Slender, lozenge-shaped lanceolates. a-#24; b-#3; c-#25;
d-#19; e-#26.

Typological and Technological Observations

Although the type name Suwannee is the most commonly used term for the
majority of Paleo-Indian lanceolates in Florida, other types have been devel-
oped to accommodate other modalities seen within the lanceolates (Bullen 1975:

The Suwannee type, thought early on to be related to the North American
Folsom complex (Simpson 1948), has been defined as "A usually large and fairly
heavy, lanceolate.shaped, slightly waisted point with concave base, basal ears,
and basal grinding of bottom and waisted parts of sides. Basal thinning and
suggestions of fluting are but rarely present" (Bullen 1975:55).

The Simpson point is distinguished from the Suwannee in that it is a
smaller point, has a relatively wide blade, a noticeably indented or waisted
haft area compared to the blade, and is slightly thinner. The basal ears are
not as prominent as those of the Suwannee.. Simpsons and Suwannees are alike
in that both have basal and lateral grinding, ears, and neither is generally
basally thinned or fluted (Bullen 1975:56). It should be noted that Bullen's


(1975:56) subtypes 1, 2, and 3 could represent a resharpening or retipping
sequence resulting in loss of length (subtype 1) and loss of blade width
(subtype 3).

The Clovis type has also been described as occurring in Florida (Bullen
1975:57; Bullen 1962: Fig. 1, upper right; Bullen 1969: Fig. 1; Neill 1964).
However, it is apparent that instances of Florida specimens which match the
Clovis type found elsewhere in the United States are noteworthy for their
rareness (Bullen 1962:87; 1969:37). Neill (1964:18) points out that when
Suwannee points are basally thinned enough to be described as fluted, that
often it is only on one face. On the other hand, lanceolates in other states
which are not strongly fluted are called "Clovis", but resemble Suwannees in
Florida (Neill 1971:63). In a sample of 51 Suwannees from the collection of
Ben Waller, chosen because they were whole, the senior author noted 5 bi-
facially fluted points and a sixth that was unifacially fluted. (This inci-
dence may be inflated because whole specimens were specifically chosen from
a much larger number in different degrees of breakage.) It is usually the
case with Suwannees, as is illustrated by most of the specimens in the present
study, that the bases are laterally thinned rather than thinned by flakes
originating from the basal concavity (see Fig. 5). This distinctive techno-
logical feature does not seem to have been emphasized in previous descriptions
in the literature.

-, t r/

,- ,

Figure 5. Line drawing illustrating lateral thinning
scars on the base. Note how scars approach
the top of the basal concavity.

Other early lanceolate types thought to be in the Paleo-Indian period
are Beaver Lake, Sante Fe and Tallahassee (Bullen 1975:45-47). These appear,
based on size and resharpening, to be in the late Paleo-Indian or Early Archaic
period, perhaps on the same horizon as Dalton in the rest of the Southeast,
i.e., 10,500 to 9900 B.P. (Goodyear 1982). With the exception of the serrated
Tallahassee, the typically resharpened Dalton form does not seem to be very
common in Florida.


The difficulties in applying the above type categories to our sample of
anceolates is typical of most attempts in lithic analysis to accommodate all
normal variability within discrete, nominal categories. That is, while some
pecimens may fit the ideal type descriptions, many will lack all of the de-
ining criteria. The result is that not all morphological variability is
ccommodated, much less explained. The source of the difficulty lies in the
lassification of formerly dynamic tools into categories which are a reaction
7 the classifier to modal forms in the specimens, rather than an understanding
f the processes which lead to formal patterns ("types") and gradations between.
nce the sources of variation are understood, then the selection of attributes
nd construction of types becomes more meaningful behaviorally, Ideally,
Classification becomes more successful in that variation is accounted for
father than ignored or qualified by generating subtypes.

To describe and evaluate some of the processes suspected to lie behind
ae formal variations in our sample of 27, the metric attributes were chosen
Lth the goal of monitoring areas of change and stability expected for a hafted
Facial tool. Generally speaking, the area of a biface covered by the haft
aght to be more stable, i.e., less variable morphologically because of the
iape requirements of the haft mechanism, which also exempts that part of the
)ol from alterations due to damage and maintenance. The blade and tip areas,
Alternatively, are exposed to breakage, wear from use, and are more variable
:cause of repairs and resharpening or retipping (cf. Judge 1973; Goodyear
)74; Gardner and Verrey 1979).

Maximum tool length (Fig. 6, a) is a measure of overall tool size. Length
Dependent on the size and type of core or flake serving as a blank for the
production of a biface, and it is also a measure sensitive to tip and blade
iduction related to repair, retipping and resharpening.

Maximum blade width (Fig. 6, b) helps measure blade configuration and
Also sensitive to reworking of the tip and blade margins. Patterning in
ie blade width-to-minimum width across the stem ratio is also a means of
asuring the "waistedness" typified by the Simpson type.







Figure 6. Diagram indicating where attribute
measurements were made on lanceolates.


Maximum width at the ears (Fig. 6; c) is a measure of the haft area. It
is measured at the maximum width across the ears. The statistical behavior
of this measure, like those of D and X (Fig. 6), should indicate something
about the haft tolerances of these tools. Measure D is probably the more
reliable of these in that the statistical value of C can be affected by ears
which break or are crumbled in the haft and which are subsequently repaired.
Fractures across the area of D presumably would cause a complete failure in
the tool. Basic metric and other data pertinent to each of the 27 specimens
are presented in Table 1.

For comparative purposes, the sample can first be treated in terms of
conventional types (Bullen 1975). Viewing the photographs, Figure 2 (a-k)
depicts the typical Florida Suwannee. They are relatively large and shaped
primarily by a well executed percussion method of bifacial thinning. It is
worthwhile to note here, along with the lateral thinning of the base (Fig. 5),
that the broad, expanding flake scars usually surviving on the central portion
of a Suwannee can be described as a technological hallmark of this point.
This method of manufacture not only helped thin and shape the preform, but
created a remarkably flat bifacial tool when viewed in cross section. Judging
from the size of these scars, they were struck from much larger reforms.
Their frequent presence on smaller finished points indicates such scars were
left there intentionally. Subsequent pressure flaking is restricted to the
biface margin for purposes of final shaping and for sharpening the tip and
blade margins.

Specimens 18, 17, 5 and 15 (Fig. 2, a-d) appear to represent a manufac-
turing sequence from a thick crude percussion preform (18 and 17) through
finished and possibly used and completed early stage Suwannees (specimens 5
and 15, Fig. 2, c-d). None of these specimens is ground or dulled in the
basal area except 15 which has heavy grinding restricted to the ears. The
rest of the specimens (Fig. 2, e-k) appear to have functioned as hafted bifaces.
Specimen 16 (Fig. 2, f) exhibits an unusual mode of resharpening, almost
Dalton-like (Morse 1973), but care was taken not to create a shoulder effect,
or a break in the line where the blade joins the haft. Also, two of the
Suwannees were reworked into endscrapers (Fig. 2, e,i). These specimens have
been described before in the Florida literature (Goodyear, Thompson and Warren
1968). All of these specimens (Fig. 2) exhibit the lateral thinning along
the base, except specimen 21 (Fig. 2, i) which is basally thinned on both

The specimens shown in Figure 3 (a-g) also have the outline of the
Suwannee type, except perhaps for specimen 8 (g) which could represent a
preform stage for the specimens to its left (e and f). These specimens appear
to be smaller than the examples discussed above, smaller in some cases, than
the Simpsons to be discussed next. The thinner of these points (Fig. 3, c,f),
specimens 7 and 6 (Table 1), do not exhibit the remnants of broad, percussion
scars as do the other specimens and the larger Suwannees (Fig. 2), suggesting
they were manufactured from smaller flakes. Specimen 6 (Fig. 2, f) is inter-
esting for it seems to be the only reasonably good example of heat treating
among the 27 specimens. It is pink and has a gloss, criteria not met by any
of the others.

Four specimens seem to fit Bullen's (1975) concept of the Simpson point
(Fig. 3, h-k). All four manifest the flat expanding scar remnants of the con-
trolled bifacial thinning. Specimen 2 (Fig. 3, h) is an unusually well pre-
served early stage Simpson, providing a glimpse as to what this type may have

Table 1. Attribute Data For Suwannee Points From Tampa Bay, Florida


cimen Provenience
No. (Fig. 1)

1 Kellog Fill
(Warren 1970)

2 Golden Gate
Speedway, Hi 486

3 Maximo Pt.
Midden, Pi 19

4 Oyster Shell
(Goodyear and
Warren 1972)

5 Fill

6 Golden Gate
Speedway, Hi 486

7 Fill,
Indian Rocks

8 Maximo Park
Beach, Pi 31

9 Fill, Pi 98

10 Bootranch
(Warren 1966)

11 Golden Gate
Speedway, Hi 486

12 Bootranch
(Warren 1966)

13 Curlew Road
(Goodyear et al.

14 Downtown
St. Petersburg

15 Oystershell
(Goodyear and
Warren 1972)








9.3 23.9








Max. Min.
Stem Stem
Thick Width



































Heavily Resharpened

Early Stage

Slender Lozenge-



Heat Treated?
(Glossy and Pink)

Organic Manganese
Oxide Coating


Large Suwannee
















Reworked to

Preform But
Appears Used

Table 1 (Cont.)

Specimen Provenience
No. (Fig. 1)

16 Fish Creek Fill
Hi 105 (Karklins 1970)

17 Lake Seminole

18 Bayview Gardens
Fill, Tampa Bay

19 Fill, 131st St.

20 Maximo Park Beach
Pi 31

21 Tarpon Springs

22 Ben Davis Beach
(Fill) Hi 456

23 Fill, 131st St.

24 Pinellas Pt.
Beach, Pi 61
(Goodyear 1968)

25 Pinellas Pt.
Beach, Pi 61
(Goodyear 1968)

26 Shorelanes Fill

27 Fill. 131st St.






















14. 1




























7.2 18.4












Heavily Resharpened


Early Stage Preform

Small Lozenge Form

Heavily Resharpened

Reworked to

Basal Thinning

Sante Fe?
Extremely Weathered







3 or 5?




H Small Lozenge Form

I Probable Broken



looked like in regions where raw material was abundant and available in large
packages. The tip is sharp and serrations have been set on the blade margins.
It is not ground anywhere on the stem (Table 1). Although there is obvious
variation in the lengths and widths of their blades, the metric uniformity
in the haft areas of these Simpsons is remarkable (Table 1).

Finally, Figure 4 illustrates an array of small, delicately made lance-
olates which should fit into the Paleo-Indian period based on their techno-
logical attributes. Specimens 24 and 3 (Fig. 4, a-b) appear to be the Sante
Fe type (Bullen 1975:46). Neither is basally thinned but both are thinned
laterally. Specimen 25 (Fig. 4, c) seems to be a small Suwannee, based on
its outline. The remaining points (Fig. 4, d,e) are narrow lozenge-shaped
forms. Specimen 19 (d) reminds one of the Beaver Lake type illustrated by
Bullen (1975:47), except that it is thinned basally by lateral thinning scars.
The last specimen is similar in that it is also laterally thinned. Basal
grinding is only present on specimen 26 (Fig. 4, e).

Using the metric attributes discussed above and their intended role in
analyzing the dynamic aspects of the points, we can evaluate certain expected
technological and functional patterns.

Besides configurational attributes, such as expanding blades, waisted
haft areas and prominent ears, size seems to be one important factor used to
classify Suwannee, Simpson and the other lanceolates (Bullen 1975). The
measurements most reliable for size should be those taken on the base of stem.
The basal area would have been covered by the haft mechanism and thus the
original basal dimensions would be preserved in spite of modifications of the
blade area. Of the three basal measures (Fig. 6, x,d,c), maximum thickness
of the stem was disgarded because the differences in thickness among the
various types are not great (Table 1).

A cross plot of maximum width at the ears against minimum width across
the stem (Fig. 7) indicates some separation by size of the types described by
Bullen (1975). Suwannee reforms plot at the extreme upper right followed
next by large Suwannees. Proceeding down the axis, there is a very tight
clustering of Simpson points. The small lozenge-shaped points and the other
slender points (Fig. 4) fall at the very bottom of the axis of size.

Just as interesting as the size patterning indicated in Figure 7, is
the highly predictable relationship between these two areas of the haft. The
correlation coefficient for these variables is exceedingly high (rs=.99).

All of the metric data can be similarly examined by calculating all pair-
wise Spearman rank correlation coefficients between attributes (Table 2). All
of the haft area measurements indicate a high degree of proportional regularity.
As noted, for any given specimen, the relationship between the maximum width
at the ears and the minimum width across the stem is extremely constant
(rs=.99). Maximum stem thickness is less highly correlated with other stem
dimensions, with rs=.75 with maximum width across the ears and rs=.77 with
minimum width across the stem. These results suggest that beyond the desired
physical size of a stem, i.e., the appropriate width and thickness, that pro-
portion or shape was also critical to the haft mechanism.

As expected, maximum length is poorly correlated to the basal area meas-
ures (Table 2). There is a moderately high relationship with blade width


(rs=.66), indicating that longer points have some tendency to have wider blades
One result which was not anticipated is the very high relationship of maximum
blade width with maximum width at the ears (rs=.89) and minimum width across
the stem (rs=.91) (Table 2). This indicates that the width of the blade did
not change much during blade reduction since blade width matches the stability
of the haft area, the latter having been argued to be unchanged since the

Table 2. Spearman Rank Correlation Coefficients Between Attributes

Blade Length

= 11
= 67.95
= 17.99

Blade Width

n = 19
x = 30.6
s = 12.49

r = .66
n = 11
p < .02

Blade Width

r = .42
n = 11

r = .89
n = 18
(p < .0001)

r = .39
n = 11

r = .71
n = 18
(p < .001)

r = .75
n = 26
(p < .0001)


Functional Imolications

The current primary method of determining the uses of stone tools is
through microscopic analysis of utilized edges (e.g., Keeley 1980). In the
present study, a microscopic analysis was not performed due to time limita-
tions. Furthermore, it was suspected that owing to the nature and condition
of the lithic raw material, such an approach would not be very rewarding.
This is not to say that lithic materials from Florida are not amenable to


n = 26
x = 30.6
s = 11.62


n = 26
x = 7.03
s = 2.27


@ stem

n = 26
x = 28.65
s = 11.60

r = .45
n = 11






microwear analysis. However, sophisticated studies 'that control for the re-
action of these cherts to use and the chemical and physical alterations these
immature cherts sustained subsequent to their deposition in the archeological
record are only in their infancy (Purdy 1981).

Macroscopically, there appears to be a general light dulling on the blade
margins of many of the specimens. Crisp, sharp serrated edges such as that
possessed by two of the Simpsons (Fig. 3, h,k), were not common in our sample.
It may be significant that these two specimens were from upland terrestrial
sites and 18 of the 27 cases are from dredges, fills and beaches. Although
none of the specimens is water worn or tumbled, the fact that they came from
inundated environments may affect the preservation of fine micro edges. Althouc
only a few have whole undamaged tips, in those cases where the tip is intact,
care seems to have been taken to make it sharp. In a few cases, however, tips
seem somewhat rounded presumably from use. Extreme weathering on a few of the
specimens prevents any sort of wear analysis (Table 1).

Some observations relative to use can be preferred, however, that rest
on morphological and statistical patterns. It is significant that in every
case where a tip survives, care was taken in both the manufacture and re-
working to maintain a symmetrical blade. The skewing of blade margins as
they form a tip, a pattern common in later Florida stemmed bifaces, is non-
existent. This symmetry is in keeping with Paleo-Indian point outlines else-
where in the East. Although reworking and resharpening of the blade is obvious
from many fluted point sites, care was taken to maintain a sharp tip and avoid
a break in the blade outline as it joins the stem or haft area. This stands
in contrast with later Dalton and notched point forms of the Early Archaic,
the hallmark of which is intensive resharpening of blade edges, often leaving
a shoulder (Goodyear 1982).

In this study, overall length is poorly correlated to any other attribute
indicating that reduction in the length of the blade was relatively common.
Thus, the blade, particularly the tip, would appear to be the area of the tool
receiving the greatest maintenance, hence, the area receiving the greatest use
In this way, these lanceolates are like other eastern U.S. Paleo-Indian points
Among the latter, variability in length is usually the greatest compared to
other attributes (e.g., Gardner and Verrey 1979).

A powerful clue as to projectile use is the presence or absence of distal
impact fractures. On the High Plains, many of the hafted bifaces used for
projectiles or thrusting implements posses flute-like scars on the blade face
and lateral burin scars on blade margins which originate from the distal end
(e.g., Frison 1978). None of the specimens in this study exhibited such impact
scars. Such scars have been noticed on two Suwannee points in the collection
of Ben Waller. Obviously, re-tipping of broken blades would tend to erase any
evidence of impact fractures assuming that a point could be salvaged. An
example of a probable re-tipped Simpson is pictured in Figure 3 (k).

The final data on the functions) of the lanceolates pertains to their
overall size. The Suwannee, especially the larger ones (Fig. 2), give the
impression of being used as heavy cutting tools. The relatively large width
of their blades and stems would not seem conducive to penetration required for
spearing and thrusting. Smaller Suwannees with identical forms (Fig. 3), occu
at the lower and middle ranges of the size spectrum (Fig. 7), and would not be
eliminated as projectily points based on size. The smaller sizes of the Simp-
son and more slender lanceolate forms such as Beaver Lake and Sante Fe and the






' 40






40 50

Figure 7.

Scattergram showing distribution of various
lanceolate types along a size axis.


0 -0

o oso
0 00

* *


lozenge-shaped specimens (Figs. 3 and 4), suggest that projectile functions
would be possible. The variability in the size of the stem width of the
Suwannees versus the Simpson and smaller lanceolates (Fig. 7) could perhaps
be a function of the hafting mechanisms of each. The high variation in Suwan-
nees might reflect a wooden haft since wood is easy to shape and can be worked
to any size. The small widths of the other forms as well as their tighter
statistical range might reflect hafts based on bone, antler or ivory where it
is easier to chip stone to match natural diameters of more intractable mate-
rials (cf. Judge 1973).

A larger sample of all of these forms will be necessary in order to document
impact fractures, use-wear from cutting and scraping, patterns in basal grind-
ing, and tip condition by type. For example, among the 51 whole specimens
observed in Ben Waller's collection, the smaller more delicate examples only
had light grinding or rubbing on their stems while the large Suwannees were
thoroughly ground. Heavier grinding might be necessary where a biface is used
in a cutting mode (cf. Goodyear 1974:32). Private collections with large
sample sizes provide data adequate to the investigation of these patterns and
suspected associations of attributes. A detailed microwear study which cross
cuts all size and shape factors would be a next step since multiple or secondary
functions seem likely for these curated implements. Multiple use and recycling
is a pattern typical of eastern United States Paleo-Indian lithic technologies
(Goodyear 1979).

Raw Material Analysis

The study of types and sources of lithic raw material has proven to be a
highly productive area of research in archeology. Paleo-Indian lithic raw
material patterns are distinctive in that high quality, easily flaked crypto-
crystalline silicates were the focus of their technology (Wilmsen 1970; Gardner
1974). Gardner (1974) has noted that, geographically, the density of fluted
points is dependent on the availability of cryptocrystalline sources. His
point is well born out in Florida as the distribution of Suwannee points tracks
very neatly the chert exposures of the state (Waller and Dunbar 1977:Fig. 1;
Purdy 1981:Maps 1 and 6).

Another fascinating characteristic of North American Paleo-Indian chipped
stone technologies is the long distance transport of lithic artifacts, re-
sulting in the widespread distribution of exotics. Exotic in this sense means
items which are foreign to a locality. Elsewhere, arguments have been pre-
sented that the dispersion of exotics reflects the movement patterns of highly
mobile hunter-gatherers relying on a curated technology (Binford 1979; Goodyear

One of the goals of the present study is to document Paleo-Indian lithic
raw material selection in the Tampa Bay region. These raw material descrip-
tions provide a baseline from which to compare the technologies of subsequent
time periods, and thus a means for determining continuity and change in raw
material utilization. Because a catalog now exists for many of the Florida
chert quarries (Upchurch et al. 1981), it should be possible to recognize the
presence of chert tools which are foreign to a region.

Within the Tampa Bay area, a number of different chert sources are present.
Analysis of the chert types represented by the 27 specimens is also a means of
determining if certain cherts were favored or ignored. For example, it has


been observed by collectors that a distinctive faded, mottled chert was often
used for Suwannee and Bolen points and the accompanying well-made plano-convex
unifaces. These artifacts usually come from drowned sites lying in bay bottom
sediments, recovered as a result of dredges and filling operations. Also, the
use of certified corals, identified informally on the basis of fossil "stars",
has been thought to have been minimal during the earliest periods, but enjoyed
a great popularity during the middle and late Holocene. Given the amount of
environmental change in coastal Florida, such as aeolian movement of sand and
rising sea levels, these patterns could refer to changes in the availability
of chert sources as well as changes in cultural systems.

Finally, analysis of the cherts represented by the 27 specimens is also
a means of checking the comprehensiveness of the Upchurch classification
system. If, for example, specimens are present for which no sources are known,
it might indicate that more variability exists in the chert bodies than was
suspected or that certain sources are now inaccessible to modern survey. One
such source, Hawthorn chert from the Turtlecrawl Point Quarry Cluster (Upchurch
et al. 1981), is unknown from present day terrestrial sources but was revealed
by offshore filling operations from an innundated site and quarry in Boca Clega
Bay (Goodyear, Upchurch and Brooks 1980). Given the extensive rise in Holocene
sea level, the probability of drowned quarries seems great.

Through identification of chert fabrics, fossils and inclusions, it was
determined that all of the specimens were obtained within the Hillsborough
River and Caladesi Quarry Clusters (Fig. 8; Upchurch, Strom and Nuckels 1981).
These quarry clusters are characterized by having distinctive chert types ana
known Indian quarry sites. In addition, the Miocene Tampa Formation, which is
the host horizon for the cherts of the Hillsborough River Quarry Cluster, is
variable in lithology (King 1979), and cherts can be attributed to local out-
zrop areas within the quarry cluster. Figure 8 shows the location of the
relevant quarry clusters.

The artifacts were subdivided into six groups, including a category for
unknown provenience, according to protolith fabric (texture of material re-
placed by chert). Each fabric type is characterized by a limited set of known
exposures and can be used to determine probable local sources of lithic raw

Type 1--Replaced Coral
Coral (Siderastraea, and other species) is a common fossil from the
Miocene of central Florida. Coral replacements by quartz range from intricate
reservations to virtually obliterated fabric. Since corals are widespread
in Florida (Upchurch et al. 1981), detailed sources cannot be assigned to
specific artifacts. Figure 9 shows areas in the Tampa Bay region where sili-
cified corals are known to occur. Because the source areas are so widespread,
there is no basis for assuming that coral artifacts were derived from greater
distances elsewhere.

Silicified coral is highly anisotropic and generally requires heat treat-
ment in order to be worked. It weathers easily and older artifacts are usually

Type 2--Speckled, Black Chert
This chert type is a micromosaic quartz replacement (Upchurch et al. 1981;
Upchurch, Strom and Nuckels 1982) of opaline sediments (Upchurch et al. 1982)
and is incompletely silicified. The chert is translucent, black to dark gray


Figure 8.

Map of study area showing location
of Upchurch (Upchurch et al. 1981)
Quarry Clusters.

Figure 9. Map showing locations of known
exposures of type 1 chert.


with light-colored spots and fossil remnants. It is found most commonly in
the Caladesi Island Quarry Cluster (Fig. 8). A similar, unspeckled chert
occurs in the Turtlecrawl Quarry Cluster and the material is also associated
with an opaline deposit in Hillsborough Bay (Strom, Upchurch, and Rosenzweig
1981). Only two artifacts were composed of this material and each was found
near a known source (Fig. 10).

The material is relatively homogenous and is easily worked. It is anal-
ogous to flint in fabric and the only problem in working it lies in the
impurities present. It patinates to a blue-gray rind and corticates easily.

Type 3--Chalcedonic Chert
This material consists of microspherules of chalcedone (fibrous quartz).
The chert is pale brown to tan, is translucent, and contains faint peloids.

It is homogenous and should work relatively easy. The chert is fairly
resistant to weathering and is infrequently heat treated. Only two pieces
were made of this material, which is surprising considering how common it is
(Figure. 11). Both specimens were found near known sources.

Type 4--Silicified Mudstones and Wackestones
This material is composed of microgranular quartz and/or microspherulitic
chalcedony (Upchurch et al. 1981; Upchurch et al. 1982). It is opaque, light
colored, and fossil content is highly variable. Fossils are primarily fora-
minifera, although molluscs are present. Quartz sand is an important inclusion
for this material. The amount of quartz sand present in the Tampa Formation
increases to the northeast and north, which coincides with the artifacts that
contain quartz sand. Figure 12 shows the geological exposures of this type
and the distribution of artifacts. Most of the specimens were transported to
the west away from sources.

This material varies in homogeneity and may or may not need thermal alter-
ation. It corticates readily, but the fabric remains easily identified.

Type 5--Peloidal Grainstones and Breccias
The opaline sediments of the Miocene of Tampa Bay are characterized by
brecciation and subsequent infilling by opal clasts and chalcedony void fill-
ing. Quartzose cherts also show these fabrics and are through to have a
common origin with the opaline material (Upchurch et al. 1981; Upchurch et al.
1982). These cherts are light gray to tan and the peloids and breccia frag-
ments look like clotted pudding. Foraminifera are common fossils.

One source at the head of Old Tampa Bay is known. Two others are sus-
pected to exist beneath Tampa Bay (Fig. 13). This material was so frequently
used for the Suwannee and Bolen points found on fills derived from inundated
portions of the bay, that amateur archeologists in the area refer to it as
"old flint" or "bay bottom chert".

The artifacts of this material are associated with these known or sus-
pected sources and show a clear distributional pattern (Fig. 13). The fills
and beaches of the extreme southern end of the Pinellas peninsular have pro-
duced a large number of tools made of this chert, suggesting that an adjacent
source might be drowned in Tampa Bay.


Figure 10.

Map showing locations of known and
suspected exposures of type 2 chert.

Figure 11. Map showing locations of known
exposures of type 3 chert.


Figure 12.

Map showing locations of known
exposures of type 4 chert.

Figure 13. Map showing known and suspected
exposures of type 5 chert.


The material is relatively homogenous and heat treatment is not neces-
sarily required before tool manufacture. It corticates easily and the fabric
may be obscured if weathering is thorough.

Type 6--Unknown
Only one specimen (7) (Fig. 3, c) could not be attributed to a source.
It is coated with an opaque, black material that obscures the fabric of the
rock (Table 1).

Although limited by a sample of 27, a couple of patterns in raw material
selection are worth nothing. It is significant that 6 of the 11 larger Suwan-
nees and 4 of the 7 smaller Suwannees are made of type 5 material, the so-
called "bay bottom chert" (Table 1). Experimental work needs to be done to
evaluate the knapping qualities of this material, but it seems highly likely
that it flakes easily and comes in larger packages sufficient for manufac-
turing the bigger Suwannees. It is known that this material was also favored
in the succeeding Early Archaic period. Accordingly, artifacts of the type 5
chert could be diagnostic of Early Man in this part of Florida.

Three of the 5 slender, lozenge shaped points are made of the replaced
coral (type 1). This association complements the observations of Neill and
Ferguson (1977). In the New Port Richey area, they reported finding highly
weathered Sante Fe points made from certified corals. Coral quarries can
be observed on the shoreline there today.

The fact that only two of the Suwannee type were made from coral is in
accord with previous raw material observations for this region. Suwannees
exhibiting the "star" fossils are unusual. The early stage Simpson (Fig. 3, h)
is a rare exception. The minimal use of replaced coral cherts for Suwannees
forms a strong contrast to middle and late Archaic period hafted bifaces where
this material, usually in a heat treated condition, was very popular.

Finally, no exotic cherts are represented in our sample of 27. Given the
pronounced pattern of exotic dispersal elsewhere in the eastern U.S., our data
do not seem to conform. We were hoping to see specimens made of material
originating from the north of Tampa Bay, such as the Brooksville and Ocala
areas. With a larger sample, such exotics might be detected. Because of the
fairly continuous availability of chert down the central spine of Florida, it
may be that the degree of curation was less such that tools were not used for
long enough periods of time to travel as far as the person. It should also
be considered that the geographic orientation of Paleo-Indian settlement
systems during the early Holocene may not have been north and south, following
the configuration of Florida today, but may have been more east and west with
the mouths of rivers existing several kilometers further west in what is now
the Gulf of Mexico. Tampa Bay, for example, no doubt represents the drowned
upland or source area of the early Holocene Hillsborough River.


This paper has presented data relating to Paleo-Indian technological
systems focusing on how hafted bifaces were made and used, and the raw material
basis of these tools. The Suwannee point and other lanceolates, it was noted,
were manufactured by a distinctive well controlled percussion method. The
basal or stem area, rather than basally thinned or fluted, was laterally


thinned with well placed scars that run across the mid line of the base,
often extending down to the basal concavity. The statistical analysis of
attributes indicated that blade reduction probably related to tip use and
repair explains most of the variability in tool length. This suggests they
may have been broken in a projectile or thrusting mode, although other uses
are not precluded. Plotting the specimens by size revealed that the different
mophological types can be segregated according to basic modalities in haft or
stem widths. The modalities in stem proportions may be related to differences
in haft mechanisms, which in itself may be a clue to tool functions.

The raw material identification indicated a minimal use of the certified
corals during Paleo-Indian times. This contrasts greatly with the use of such
materials in later periods. The few specimens that were made of coral serve
to indicate that such materials were available to Early Man but, in a manner
parallel to Paleo-Indian cryptocrystalline utilization elsewhere, were
essentially ignored in favor of the limestone replaced cherts, which are more
homogenous and easier to flake. Whether or not thermal alteration was employed
to improve these latter types is a question yet to be dealt with in a rigorous
experimental way. However, based on the criteria of glossiness and pink color,
these cherts seemed to have been used as they were in nature. No exotics,
i.e., chert types found outside the Tampa Bay region, were seen among the 27
points. The settlement and geographic implications of this finding requires
further evaluation using a larger sample. Also, a raw material study such as
this one needs to be undertaken in regions to the north to see if tools of
Tampa Bay chert are present.

The high probability of drowned late Pleistocene and early Holocene sites
on the continental shelf of North America is a subject of increasing scientific
interest among archeologists (Edwards and Merrill 1977: Newman and Salwen 1977).
The existence of such sites off the west coast of Florida is amply illustrated
in this paper. The bays along the Gulf Coast were, in former times, the head-
waters of Pleistocene drainage systems, as well as the locations of prehistoric
aboriginal settlement systems. The fact that 18 of the 27 specimens and 13
of the 19 contexts of recovery were from underwater, illustrates in two ways
that the archeological record of Early Man in this part of Florida is removed
from normal methods of terrestrial site observation.

In the past, there has been a tendency to overlook surface data in favor
of sites with "secure" contexts. The use of artifacts from dredged sites,
such as these from Tampa Bay, have enjoyed even less research status. Hope-
fully, this paper demonstrates that the artifacts themselves, especially studies
which reveal how they were made and used and their raw material characteristics,
can tell us much about the broader aspects of past cultural systems to which
they were no doubt linked.


This research was partially supported by the Institute of Archeology and
Anthropology, University of South Carolina, Dr. Robert L. Stephenson, Director.
We would like to thank the following institute staff for their contributions:
Darby Erd for his fine artistic illustrations, Gordon Brown for his superb
photography, and Mary Joyce Burns and Azalee Swindle for their conscientious
typing. Those people who so graciously loaned us their artifacts are Lyman
O. Warren, Richard McDonnell, Butch Evans and Ben Waller. The previous


research of Lyman O. Warren, Dr. Ripley P. Bullen, and Dr. Wilfred T. Neill
on the Paleo-Indian period and technology is also acknowledged. Warren's
observations on the Tampa Bay area during the 1960's has laid a foundation
for our present work. Finally, this paper is dedicated to the memory of Ralph
"Bud" Nelson, first president of the Central Gulf Coast Archaeological Society.

References Cited

Binford, Lewis R.
1979 Organization and Formation Processes, Looking at Curated
Technologies. Journal of Anthropological Research 35:255-273.

Bullen, Ripley P.
1962 Suwannee Points in the Simpson Collection. The Florida
Anthropologist 15:83-88.

1968 Beveled Stemmed Points from Tampa Bay. The Florida
Anthropologist 21:89-90.

1969 A Clovis Fluted Point from the Sante Fe River, Florida.
The Florida Anthropologist 22:36-37.

1975 A Guide to the Identification of Florida Projectile Points.
Gainesville: Kendall Books.

Bushnell, Francis
1962 The Maximo Point Site 1962. The Florida Anthropologist

Clausen, C. J., A. D. Cohen, Cesare Emiliani, J. A. Holman and J. J. Stipp
1979 Little Salt Spring, Florida: A Unique Underwater Site.
Science 203:609-614.

Daniel, Randy and Michael Wisenbaker
1983 A Preliminary Report on the Excavations at Harney Flats,
Hillsborough County. The Florida Anthropologist 36(1-2):67-80.

Edwards, R. L. and A. S. Merrill
1977 A Reconstruction of the Continental Shelf Areas of Eastern
North America for the Times 9,500 B.P. and 12,500 B.P.
Archaeology of Eastern North America 5:1-43.

Frison, George C.
1978 Prehistoric Hunters of the High Plains. New York: Academic

Gardner, William M.
1974 The Flint Run Complex: Pattern and Process During the
Paleo-Indian to Early Archaic. In The Flint Run Paleo-Indian
Complex: A Preliminary Report, 1971-73 Seasons. Edited by
W. M. Gardner, pp. 5-47. Occasional Publication No. 1,
Archaeology Laboratory, Department of Anthropology, The
Catholic University of America, Washington, D.C.


Gardner, William M. and Robert A. Verrey
1979 Typology and Chronology of Fluted Points from the Flint Run
Area. Pennsylvania Archeologist 48:13-46.








Albert C.
Pinellas Point: A Possible Site of Continuous Indian
Habitation. The Florida Anthropologist 21:74-82.

The Brand Site: A Techno-functional Study of a Dalton Site
in Northeast Arkansas. Arkansas Archeological Survey,
Research Series 7.

A Hypothesis for the Use of Cryptocrystalline Raw Miaterials
Among Paleo-Indian Groups of North America. Research
Manuscript Series 156. University of South Carolina,
Institute of Archeology and Anthropology, Columbia.

The Chronological Position of the Dalton Horizon in the
Southeastern United States. American Antiquity 47:382-395.

Albert C., William Thompson and Lyman O. Warren
Suwannee Style End Scrapers From Pinellas County. The
Florida Anthropologist 21:91.

Albert C., Sam Upchurch and Mark J. Brooks
Turtlecrawl Point: An Inundated Early Holocene Archeo-
logical Site on the West Coast of Florida. In Holocene
Geology and Man In Pinellas and Hillsborough Counties,
Florida. Southeastern Geological Society Guidebook 22,
Sam B. Upchurch, assembler, pp. 24-33. Tallahassee.

Albert C. and Lyman O. Warren
Further Observations On The Submarine Oyster Shell Deposits
of Tampa Bay. The Florida Anthropologist 25:52-66.

Judge, W. James
1973 Paleo-Indian Occupation of the Central Rio Grande Valley
in New Mexico. Albuquerque: University of New Mexico


The Fish Creek Site, Hillsborough County, Florida. The
Florida Anthropologist 23:62-80.

King, K. C.

Tampa Formation of Peninsular Florida--A Formal Definition.
Masters Thesis, Florida State University.

Milanich, Jerald T. and Charles H. Fairbanks
1980 Florida Archaeoloov. New York: Academic Press.

Morse, Dan F.

Dalton Culture in Northeast Arkansas. The Florida
Anthropologist 26:23-38.


Neill, Wilfrec


The Association of Suwannee Points and Extinct Animals.
The Florida Anthropologist 17:17-32.

A Florida Paleo-Indian Implement of Ground Stone. The
Florida Anthronoloaist 24:61-70.

Neill, Wilfred T. and George R. Ferguson
1977 The Age of the Sante Fe Projectile Point Type. The Florida
Anthropologist 30:18-21.

Newman, W.

S. and B. Salwen, eds.
Amerinds and Their Paleoenvironments in Northeastern
North America. Annals of the New York Academy of
Sciences 288.

Purdy, Barbara A.
1981 Florida's Prehistoric Stone Technology. Gainesville:
University Presses of Florida.

Simpson, J. Clarence
1948 Folsom-Like Points from Florida. The Florida Anthropologist

Strom, R. N., S. B. Upshurch and A. Rosenzweig
1981 Paragenesis of "Box-Work Geodes", Tampa Bay, Florida.
Sedimentary Geology 30:275-289.

Upchurch, S. B., R. N. Strom and M. G. Nuckels
1981 Methods of Provenance Determination of Florida Cherts.
Unpublished contract report submitted to Florida Bureau
of Historic Sites and Properties, Tallahassee.

1982 Silicification of Miocene Rocks from Central Florida. In
Miocene of the Southeastern United States, Special
Publication No. 25, Florida Bureau of Geology, T. M. Scott
and S. B. Upchurch, editors, pp. 251-84. Tallahassee.

Waller, Ben I.
1970 Some Occurrences of Paleo-Indian Projectile Points in
Florida Waters. The Florida Anthropologist 23:129-134.

Waller, Ben I. and James Dunbar
1977 Distribution of Paleo-Indian Projectiles in Florida. The
Florida Anthropologist 30:79-80.

Warren, Lyman O,
1962 Early Pottery in the Tampa Bay Area. The Florida
Anthropologist 15:71-72.



Possibly Submerged Oyster Shell Middens of Upper Tampa
Bay. The Florida Anthropologist 17:227-230.

A Possible Paleo-Indian Site in Pinellas County. The
Florida Anthropologisr 24:39-41.


1968 Caladesi Causeway: A Possible Inundated Paleo-Indian
Workshop. The Florida Anthropologist 21:92-94.

1970 The Kellogg Fill from Boca Ciega Bay, Pinellas County,
Florida. The Florida Anthropologist 23:163-167.

Willey, Gordon R.
1949 Archeology of the Florida Gulf Coast. Smithsonian
Miscellaneous Collections 113.

Albert C. Goodyear
Institute of Archeology and Anthropology
University of South Carolina
Columbia, South Carolina 29208

Sam B. Upchurch
Department of Geology
University of South Florida
Tampa, Florida 33620

Mark J. Brooks
Piper Archaeological Research, Inc.
2112 4th Street North
St. Petersburg, Florida 33731

Nancy N. Goodyear
Department of Public Health
University of South Carolina
Columbia, South Carolina 29208


Randy Daniel and Michael Wisenbaker


The purpose of this report is to provide preliminary results of the exca-
vations recently conducted at Harney Flats--one of the archaeological sites
of the 1-75 Highway Salvage Program conducted by the Bureau of Historic Sites
and Properties. The Florida Department of State, Division of Archives, History
and Records Management, Bureau of Historic Sites and Properties, has maintained
a cooperative agreement with the Florida Department of Transportation since
1966. The primary focus of this agreement has been for the Bureau to assess
the impacts of highway construction on cultural resources within the state of

As a requirement of this program, a Phase I survey was conducted in 1978
by B. Calvin Jones of 41.2 miles of the proposed Interstate 75 corridor through
Hillsborough County. It revealed a rich inventory of preceramic archaeological
sites. Selective testing of specific physiographic features along the right-
of-way such as ridge tops, slopes, plateaus, river terraces, and levees re-
vealed 31 sites. Additional sites were noted also in a larger survey corridor,
subjected to only surface inspection. Thirteen of the 31 sites discovered by
the survey were chosen for Phase II projects. After testing was underway, four
of the Phase II sites were then considered significant enough to warrant Phase
III mitigative salvage excavations. When analyses are completed on all of
these Hillsborough sites, a great deal of information will be generated re-
garding Paleo-Indian and Archaic cultures in Florida (Jones and Tesar 1982).
Harney Flats (8Hi507) was one of two sites that Jones believed had the poten-
tial for the recovery of Paleo-Indian materials. Although no such artifacts
were recovered in Jones' survey, his examination of amateurs' collections re-
vealed the presence of Suwannee points allegedly recovered from the Tampa By-
Pass canal in the vicinity of Harney Flats (Jones, personal communication)

Deep sand lithic sites such as Harney Flats are abundant in central
Florida and often contain a greater than 2 meter deposit of cultural lithic
material intermixed within a homogeneous deposit of sand. Due to the acidic
nature of the soil, organic material such as bone or wood is seldom preserved.
Similarly, midden stains or feature outlines are rarely found.

The Harney Flats Site

The Harney Flats site is in north-central Hillsborough County (which is
located in west-central peninsular Florida) just over 2 miles east of the city
of Temple Terrace (Fig. 1). Harney Flats, a low inland basin covering several
square miles, is located near the interface of the Gulf Coastal Lowlands and
the Polk Upland physiographic regions. More specifically, the Flats are near



A 6




1 0 2 3Mile
R R R R = = = = = =r
75 i SCALE0

Map of site location.

Figure 1.


the north end of the western boundary of the Polk Upland, where the Pamlico
scarp (25 ft. elevation) overlaps the higher Wicomico scarp (100 ft. elevation)
that delimits the extent of the Upland (White 1970). The Flats are part of a
plain that slopes gently upward from Hillsborough Bay. The plain is a former
bay bottom that was in existence sometime during the Pleistocene and was part
of an estuary larger than the present Hillsborough Bay (Motz 1975). The Flats
are now (or were prior to the construction of the Tampa By-Pass Canal) a low
swampy plain rimmed by a scarp. The site itself is not in Harney Flats but
rather on the scarp overlooking them (Fig. 2).

In this region of Florida, especially in the vicinity of Tampa Bay and the
Hillsborough River Basin, outcrops of St. Marks Formation (or Tampa Stage) lime-
stones are common. The outcrops frequently contain silicified corals and cherts
(Carr and Alverson 1959:60). These silicates provided aboriginal inhabitants
with a plentiful source of raw materials from which tools could be made. In
addition, such karst features as springs, sinks, and cavities are prevalent
in these Tertiary limestone formations (Wright 1973), thus explaining their
presence in or near Harney Flats.

Site Setting
Although certain areas of the Harney Flats site are characterized by well-
drained deep sandy deposits, those sediments found south of an abandoned house
and east of the right-of-way center line--the main locus of the Paleo-Indian
component--are classified as Leon Fine Sand, the most common soil type in
Hillsborough County. This type of soil is extremely acidic and has acted as
a catalyst in the decomposition of wood and bone that otherwise might have been
preserved at the site. Leon Fine Sand normally has a pan layer (ca. 14 to 30
inches below the surface) composed of organic matter and minerals. This hardpan
perches water during and after periods of heavy-to-moderate rainfall (Leighty
1958), but was not formed until after the deposition of preceramic cultural
remains (Hunt and Hunt 1957).

Although it should be mentioned that much of the site surface had been
disturbed prior to testing, there were several noticeable zones of vegetation
found within the Phase II test areas. The northern third of the site, a well-
drained hillside, was covered with xerophytic flora. To the south of this
xeric zone, young pines (Pinus sp.) were planted in rows running-diagonally
from northwest to southeast. Mesic vegetation including cypress (Taxodium sp.),
cabbage palm (Sabal palmetto), bay (Magnolia virginiana), and magnolia (Magnolia
grandiflora) were observed down slope (south and east) of the planted pines on
the southeastern portion of the right-of-way. Lastly, the extreme southern
portions of the corridor were covered with mucky grasslands on the west and a
hydric swamp to the east. Much of this southern portion was disturbed as a
result of the construction of the Tampa By-Pass Canal and levee (Fig. 3).

In terms of available water, a segment of the Hillsborough River is less
than 1.4 miles northwest of the site. A USGS topographic map indicates that
a sinkhole is located about one-third of a mile north of the site but the sink-
hole is now inundated by the Tampa By-Pass Canal. The site is also only 0.4
mile northeast of the Six Mile Creek, the primary tributary of the Palm River.
In addition, several small artesian springs are located within Harney Flats
(Rosenau et al. 1977).

Sea Level and Climate
It is important to keep in mind that 10,000 to 14,000 years ago sea level
on the Gulf of Mexico may have been 200 feet lower than at present (Ballard

--L-,, -~ ,, "- L -:-:./ --..'' -,
70 r
_-- -

S- -- &,- I-


7,. 0 5Ud
.0 ,

RHi507 '
:- o, >L--


Y_1 _JO

:_ _: -.---- -F :

:,t.-t.Y__. HARNEY FL-ATS = '7

-2 -.,-_= =_- ._ _- -- m-ie

Figure 2. Map of Harney Flats.


and Uchupi 1970). Harney Flats would have been more than 100 miles further
inland from the Gulf than at present and Tampa Bay would have been nothing
more than a catch basin and channel for drainages such as the Hillsborough
River. The flow of rivers, streams, and springs in the region would have been
greatly reduced (if flowing at all) compared with that of recent times, due to
the lower levels of the Floridan Aquifer and lesser amounts of precipitation.
As a consequence of these conditions, many of Florida's shallow lakes such as
Lake Thonotosassa) and water courses were probalby dry prior to 8,500 years ago
(Watts 1980).

Not only was the peninsula of Florida drier when Paleo-Indians occupied
it, but the climate was more temperate (as opposed to the humid subtropical
climate of today) with the landscape characterized by sclerophyllous oak (a
desert-like adaptation where exceptional development of sclerenchyma in leaves
results in a thickened, hardened foliage resistant to loss of water) with open
patches of prairie and savanna (Watts 1980).

Despite a drier climate, a greater diversity of fauna was present during
Paleo-Indian times, especially with regard to mammals. The remains of sloth
(Megalonyx sp.), dire wolf (Canis dirus), North American spectacled bear
(Tremarctos floridanus), sabercat (Smilodon floridanus), mastodon (Mammut
americanum), horse (Equus sp.), and peccary (Platygonius ccmpressus) have been
recovered from the Devil's Den site in north-central Florida (Webb and Martin
1974). At the Little Salt Spring site in Sarasota County, just south of Tampa
Bay, a giant land tortoise (Geochelone crassicutata), believed to have been
killed by man, was recovered (Clausen et al. 1979).

Phase II Testing

Late in the summer of 1981, a small crew spent ten weeks conducting pre-
liminary test excavations at Harney Flats. This testing was designed to locate
artifact concentrations and to determine the site boundaries within the 1-75
right-of-way. The segment of the right-of-way to be tested was approximately
350 meters long and just over 100 meters at its widest point. The testing
consisted initially of fifty centimeter square test pits located approximately
every 30 meters across the site. Since only ten weeks were allocated for
testing an area of almost ten acres, and since deep deposits (ca. two meters)
were anticipated, a backhoe was used to help reduce the work load in excavating
the test pits. We used the backhoe to dig a short trench that allowed a crew
person to enter and then excavate a 50 centimeter wide column in the wall
relatively rapidly to a depth of two meters. Excavation was done in 20 centi-
meter arbitrary levels with the fill being screened through .25 inch mesh
(Fig. 3).

Based on artifact counts from 58 of these test pits, a series of 30 meter
square blocks were selected on the eastern half of the right-of-way for further
sampling. This was done by manually excavating randomly selected 1 and 1 x 2
meter test units in areas of higher artifact concentration. However, due to
high water from summer rains, test pits could not be finished in the southern
and central portion of the right-of-way, and excavations were redirected to
areas at higher elevations. An additional 72 square meters were excavated in
this manner. Also, series of exploratory backhoe trenches were placed along
the eastern half of the right-of-way.

72 SOi 401 1O ZO 0 O o 0t X0 OI Sot o T O

Ook f
."O '
-. -- "

1O S A 1 le r
.OS- / C

\ \ I Mixed Hordl i I l [ I


.01.. '. ., ... .. .

10 0 1 ,

'a' TI~h S-- -

gose- sie .
o- //j-ioo r

leoo a can J I -

so,_ .. Pond
a o

7oo Rudwsl vfg*loe
00os- / *

amas- *. ,
avs -" --
sees- .. s .

-- W Mo *
31o ...

Figure 3. Site map.

Trench 7

Area I

Area 3

. -A-. -.46-


0 5 10 20 mets

Figure 4. Phase III excavation areas.

O I.Om Square Excavation Unit
a Previously Excavated
I Exploratory Trench
--. Right-of-Way Boundory
A Bench Mark
0 Permanent Transit Station
-M Uf Drainage Ditch





Cultural material was found to extend from the surface to 160 to 180
centimeters below the surface. The Paleo-Indian component was found to lie
in a 60 centimeter deep deposit that was located from about 120 to 180 cm
below the surface, although apparently concentrating from between 120 and 140
cm below surface. A Middle Archaic component was also present, indicating the
stratified nature of the site. This component was manifested by the recovery
of stemmed projectile points and debitage which was confined to the upper 80
to 100 centimeters of the site, but appeared to concentrate in the 60-80 cm

Stratigraphically, the uppermost soil zone consists of a thin layer of
modern duff and humic sand. Beneath this is a thick gray sand zone extending
in depth from 40 centimeters to as much as 80 centimeters below the surface.
The zone is underlain by a chocolate brown hardpan which varies somewhat in
thickness and consistency but is about 10 centimeters thick. Beneath the hard-
pan is a tan sandy zone, the uppermost portion of which is somewhat darker as
a result of leaching from the hardpan. This sandy zone continues to a depth
of at least 2 meters, a depth that water was usually encountered during the
excavations. However, on the east edge of the right-of-way, a blue-gray clayey
sand was encountered at about 140 centimeters below the surface. This stratum
was found to be culturally sterile. The clay was laid down during the middle
to late Miocene time, when beds of clay were deposited over the pre-existing
calcium carbonates and which now cover much of the limestone in the area (Davies
and LeGrand 1972).

It is worth noting that the exact geological and chemical nature of the
hardpan is not clear at this time and is under study as part of the analysis
of site formation processes. Apparently, the hardpan "sealed" and protected
the early cultural component and prevented the mixing of cultural materials
that is common on many sites in the area. It was found that the hardpan dis-
appeared gradually at about the centerline of the right-of-way. At this point,
the stratification basically becomes one of undifferentiated deposits of sand
similar to the dune formations common to the area.

Although no features were identified, test pits often revealed clusters
of artifacts believed to be in primary archaeological context. These included
a pair of sandstone abraders, a hammerstone and core fragments, a bifacial core
and a large unifacially retouched flake and a "turtleback" scraper associated
with another unifacially modified flake. All of these tools came from the
lower levels, below the hardpan. The identification of the component as Paleo-
Indian was made on the basis of the recovery of point reforms and lanceolate
points (Fig. 5) that are generally classified as Suwannee or Simpson types in
Florida (Bullen 1975:55-56).

Numerous points from the Paleo-Indian period have been found at a number
of underwater locations including sinkholes and streambeds in Florida. These
sites often are interpreted as being animal kill sites or special use loca-
tions. Only one possible dry-land habitational site from this period has been
previously excavated in the state. This site, located near Silver Springs,
yielded a few Suwannee points in a deep sand context, although the significance
of that site has not been determined and much of it has been hauled away for
fill (Neill 1958).

In summary, the Phase II testing did accomplish most of the goals for
which it was designed. It demonstrated the existence of deeply stratified

Figure 5. Paleo-Indian projectile points from the Harney Flats site.

Unifacial tools from the Harney Flats site.

Figure 6.


in situ deposits of cultural material dating to Paleo-Indian and transitional
Paleo-Indian periods. Moreover, it isolated significant areas of occupational
concentrations within the total site area. The Paleo-Indian component was
isolated in an area that covered approximately four acres. It should be noted
that an, as yet, unknown portion of the site extends beyond the eastern limits
of the right-of-way.

Phase III Salvage

Since virtually no terrestrial Paleo-Indian site had been excavated in
the state of Florida, more extensive salvage work was recommended. A proposal
that included a budget for one field supervisor, two assistants, and a crew of
twenty-three for four months of field work was submitted to and accepted by
the Florida Department of Transportation and the Federal Highway Administra-
-tion. The focus of the Phase III was to obtain data on intra-site structural
patterning. Therefore the strategy of the fieldwork was to excavate large
contiguous areas in order to reveal really extensive levels of the site that
when combined with the piece-plotting of artifacts would provide invaluable
data on the behavioral patterns of early man.

As a result of the Phase II testing, three areas totaling some 10,000
square meters were chosen for sampling within the right-of-way. However, it
was felt that because of the large size of the right-of-way and since the
initial testing was restricted because of the high water table, additional
test pits would be necessary to confirm or modify the sampling areas. There-
fore, additional test pits were placed approximately every ten meters along
the eastern half of the right-of-way. Based on these results the three areas
were slightly modified. The three'areas consisted of 283 square meters exca-
vated in Area 1.; 377 square meters excavated in Area 2; and 196 square meters
excavated in Area 3 (Fig. 4).

Prior to the excavation approximately one meter of topsoil was removed
by heavy machinery from each area. Before this was done, however, over 1,000
surface elevations were taken in order to reconstruct the original ground

Briefly, the principal excavation technique employed was that of flat
shovel-skimming. This technique, along with hand troweling when any suspected
tool or large artifact was encountered, served as an efficient means for soil
removal while providing a precise method for uncovering in situ artifactual
materials. Lithic tool specimens were plotted immediately after they were
uncovered by the excavations. This artifact category included any tool forms
that were morphologically recognizable and identifiable under conventionally
accepted terminology (e.g., projectile points, biface, uniface, cores, modified
flake tools, hammerstones and various tool fragments). These tools were
accurately plotted with regard to both horizontal and vertical provenience.

Excavators worked in single 1 x 1 meter squares which were contained
within a larger 4 meter square block, each such excavation block consisting
of sixteen 1 x 1 meter units. Excavation proceeded in 20 centimeter arbitrary
levels as previously described. All soil from each level was screened through
.25 inch mesh. All debitage and any tool fragments that escaped detection
were thereby provenienced in a 1 square meter by 20 centimeter level. Soil
samples both for phosphate analysis and for flotation (to recover microflake
debitage and botanical remains) were also collected. A more detailed descrip-
tion of the excavation technique will be given in the final report.


Field Impressions

Based on observations concerning artifact density and distribution within
the three areas, some general impressions concerning intra-site differences
were formed and appeared tentatively to reinforce the possibilities of defining
activity areas within the site. The exact nature of these areas, however,
awaits further analysis.

A few side-notched Bolen points were recovered in each of the excavated
areas at Harney Flats. Although these points can be roughly placed in time by
comparisons with similar side-notched and corner-notched styles in other areas
of the Southeast (e.g., Palmer, Big Sandy I), the exact cultural assignment
in Florida is still not fully understood. Some archaeologists (Milanich and
Fairbanks 1980:45-48; Purdy 1981:24-26) assign them to a late Paleo-Indian
time period while others indicate early Archaic placement (Goodyear 1982a:390;
1982b:166-167). The Bolens from Harney Flats appear to come from just above
as well as within the Suwannee levels, although final interpretation awaits
the completion of the analysis.

If anything of certainty can be said at this point, it is that the uni-
facial tool is the hallmark of the stone tool kit recovered from Harney Flats.
It was readily apparent during the excavation that several variable forms
probably existed. Although unifaces were the predominent tool type recovered,
a number of bifaces (point reformss, cores, and hammerstones also were re-

Analysis Goals

Several problem areas concerning early man in Florida will be addressed
in the analysis of Harney Flats. A few of the primary analytical objectives
will be outlined below.

As noted earlier, the goal of the Phase III excavation was to obtain data
concerning the intra-site structure at Harney Flats. In essence, this dictated
the way the site was excavated; i.e., opening large contiguous areas with
emphasis on three-dimensional plotting of individual tools. This analytical
objective will be achieved by analyzing the spatial dimensions of tool and
debitage clusters in attempting to isolate activity areas as well as task
specific functional tool assemblages and tool classes. To this end, it is
anticipated that computer-generated statistical tests of association and com-
puter maps (SYMAP) will be utilized.

Another major goal of the analysis will be to formulate a tool typology,
particularly from the large number of recovered unifaces. A preliminary sorting
of the unifaces suggests a variety of tool shapes and sizes. The tools from
this site have both large size as well as variable forms (Fig. 6). They appear
to be more difficult to classify in relation to traditional typologies than it
has been customary to North American Paleo-Indian studies. Albert C. Goodyear
(personal communication) believes this is related to the fact that most of our
"classic" Paleo-Indian assemblage studies are from kill or habitation sites
and not quarry-related'sites such as Harney Flats.

A functional interpretation of the stone tools will be based on overall
tool morphology, use wear analysis, edge angle information, and an assessment
of other indicators of use such as repair and breakage patterns. All tools


will be microscopically examined for traces of use-wear in the form of edge
rounding, polish, striations, and patterned microflaking. Placement and type
of wear will be evaluated to establish mode of use; i.e., cutting, scraping,
drilling, perforating and to provide an indication of the kinds of materials
worked. This analysis will be conducted in association with George R. Ballo
at the University of South Florida.

Another interesting problem that will be addressed is that of the strati-
graphic relationship of the Suwannee and the Bolen points at Harney Flats.
Can a meaningful vertical separation be made at this site? This, of course,
relates to any temporal depositional events that may exist in the 60 centimeter
zone that constitutes the earliest cultural component. Although Bolens are
later in time, do they reflect early Archaic or late Paleo-Indian events?

Finally, geological investigations also will play an important role in
the overall understanding of Harney Flats. These investigations are being
conducted by Sam Upchurch of the University of South Florida. This will involve
reconstructing the depositional history of the site as well as attempting to
understand the Paleo environmental factors that were of importance in why
an upland area such as the Pamlico scarp overlooking Harney Flats was utilized
by early man. Additionally, analysis of the utilized stone in relation to
recent work concerning the defining of source areas or quarry clusters (Upchurch
et al. 1981), also will be an important facet of the analysis. The heavy reli-
ance upon quality cryptocrystalline stone is now understood to be a critical
component of the Paleo-Indian stone tool kit, based on the special technolog-
ical needs of highly mobile hunter-gatherers (Goodyear 1979).

Summary Summary

In summary,almost seven months of fieldwork were spent excavating over
900 square meters at Harney Flats. The recovered data, when analyzed, will
offer for the first time a rare opportunity not only to enhance our under
standing of Paleo-Indian lithic technology in Florida, but also to provide
insight into other important questions of site organization and function.


The authors would like to acknowledge the assistance of the following
people in the completion of this paper. Special thanks are extended to Charlie
Poe who assisted in the field supervision and also drew the figures. Thanks
also are due to Albert C. Goodyear who contributed his observations concerning
the tool assemblage and analysis. Ron Jones prepared the figures for publica-
tion. Finally, Claudine Payne edited the manuscript and Carolyn Moore typed
the final copy.

References Cited

Ballard, Robert D. and Elazar Uchupi
1970 Morphology and Quaternary History of the Continental Shelf of
the Gulf Coast of the United States. Bulletin of Marine Sciences


Bullen, Ripley P.
1975 A Guide to the Identification of Florida Projectile Points.
Revised edition. Kendall Books, Gainesville.

Carr, Wilfred J. and Douglas C. Alverson
1959 Stratigraphy of Middle Tertiary Rocks in Part of West-Central
Florida. U.S. Geological Bulletin 1092. Washington.

Clausen, Carl J., A. D. Cohen, C. Emiliani, J. A. Holman and J. J. Stipp
1979 Little Salt Spring: A Unique Underwater Site. Science

Davies, W. E. and H. E. LeGrand
1972 Karst of the United States. IN Karst, edited by M. Herak and
V. T. Stringfield. Elsevier Publishing Co., Amsterdam,
pp. 467-505
Goodyear, Albert C.
1979 A Hypothesis for the Use of Cryptocrystalline Raw Materials
Among Paleo-Indian Groups of North America. Research Manuscript
Series 156, Institute of Archaeology and Anthropology, University
of South Carolina.

1982a The Chronological Position of the Dalton Horizon in the South-
eastern United States. American Antiquity 47(2):382-395.

1982b Review of Florida Archaeology, 1980, Jerald T. Milanich and
Charles H. Fairbanks, Academic Press. American Anthropologist

Hunt, C. B. and A. P. Hunt
1957 Stratigraphy and Archaeology of Some Florida Soils. Bulletin
of the Geological Society 68:797-806.

Jones, B. Calvin and Louis D. Tesar
1982 An Update on the Highway Salvage Program in Florida. The
Florida Anthropologist 35(1):59-62.

Leighty, Ralph G.
1958 Soil Survey of Hillsborough County, Florida. U. S. Department
of Agriculture, Washington.

Milanich, J. T. and Charles H. Fairbanks
1980 Florida Archaeology. Academic Press, New York.

Motz, Louis H.
1975 Hydrologic Effects of the Tampa Bypass Canal System, Florida.
Bureau of Geology. Report of Investigations #82, Tallahassee.

Neill, Wilfred T.
1958 A Stratified Early Site at Silver Springs, Florida. The
Florida Anthropologist 11:33-48.

Purdy, Barbara A.
1981 Florida's Prehistoric Stone Technology. University Presses
of Florida, Gainesville.


Rosenau, J. C., G. L. Faulkner, C. W. Hendry, Jr., and R. W. Hull
1977 Springs of Florida, Florida Bureau of Geology. Bulletin
#31, Tallahassee.

Upchurch, Sam B., Richard N. Strom and Mark G. Nuckels
1981 Methods of Provenience Determination of Florida Cherts.
Department of Geology, University of South Florida. Report
submitted to Florida Bureau of Historic Sites and Properties
in compliance with State Grant No. 80-072.

Watts, W. A.
1980 The Quaternary Vegetation History of the Southeastern United
States. Annual Review of Ecology and Systematics 11:387-409.

Webb, S. David and Robert A. Martin
1974 Late Pleistocene Mammals from the Devil's Den Fauna, Levy
County. IN Pleistocene Mammals of Florida, edited by S. D.
Webb, pp. 114-145. University Presses of Florida, Gainesville.

White, W. A.
1970 The Geomorphology of the Florida Peninsula, Florida Bureau of
Geology. Geological Bulletin #51, Tallahassee.

Wright, Alexandra P., ed.
1973 Environmental Geology and Hydrology, Tampa Area, Florida.
Florida Bureau of Geology. Special Publication #19,

Randy Daniel
Michael Wisenbaker
Florida Division of Archives, History
and Records Management
Tallahassee, Florida


S. David Webb, Jerald T. Milanich, Roger Alexon,
and James Dunbar

In October, 1981, Roger Alexon, accompanied by Bill Mathen and Bob
Gingery, recovered a portion of the skull roof of a Bison from the Wacissa
River in Jefferson County, Florida. A right horn core was also collected
at the same time. Subsequent examination showed that the core connected with
the skull table. Protruding from the right fronto-parietal area of the
skull was a fragment of a chert projectile point. The skull, originally
recovered in six pieces and now partially reconstructed (see Figure 1) has
been dated by biostratigraphy and radiocarbon analysis of bone collagen
to 11,000 years ago. The find demonstrates unquestionable association of
the Bison with humans during the Paleo-Indian period in Florida. This
brief report summarizes information produced by an examination of the skull
and the artifact, both of which were presented to the Florida State Museum
(catalog number UF43201). A more detailed description and interpretation
of the skull and fragment of point will appear at a later date.

Figure 1. Reconstructed Bison skull and horn cores. Arrow
points to chert point fragment.


VOL. 36 NOS. 1-2



The Bison skull came from the river bottom at a depth of approximately
2.4 m below the present surface of the water. It was immediately behind
(downstream) several large boulders which lie roughly east-west across
the north-south river bottom. This location is beside a deeper depression
in the river bottom which contains deposits of organic mud. At the time
the Bison bones were laid down it is possible that the depression was
a shallow, limestone-bottomed pool. Such a water hole might have attracted
both animals and human hunters at a time when water levels and the avail-
ability of surface water in Florida were much less than they are at present.

Metric examination of the excellently preserved Bison specimen, which
includes the skull and both horn cores, positively identifies the species
of the animal. Using systems of cranial and horn measurements and indices,
it is certain that the Wacissa Bison is a female Bison antiquus six to
ten years of age. Furthermore, the species is an evolutionarily late
representative of that now extinct species, suggesting a late Pleistocene

Absolute dating is provided by radiocarbon analysis of bone samples.
Three pieces of skull collected from the same locale, but not having recog-
nizable contacts with the rest of the skull, were submitted to Beta Analytic,
Inc., in Coral Gables. A date of 9990200 radiocarbon years: 8040 B.C.
(Beta-5941) was obtained. A second sample was also submitted, consisting
of the distal end of a Bison humerus found within one meter of the skull.
Its excellent preservation and position in somewhat deeper bottom sediments
than the three skull fragments make us suspect that it was less subject
to foreign carbon contamination. A date of 11,170130 radiocarbon years:
9220 B.C. (Beta-5942) came from the analysis.

The chert fragment sticking through the skull is a section of a
tapering and lenticular in cross section blade of an unidentified point.
Broken off at both ends, the fragment is 1 cm long. Approximately one
half of it protrudes out from the skull. Apparently the point shattered
when it struck the skull, breaking both the tip and the blade. Measurements
of the fragment and comparisons with other points (from all time periods)
in the collections of the Florida State Museum, suggest that it could have
come from almost any type of lanceolate point and that it stuck into the
animal's cranium 1-2 cm. It was probably not the cause of death and could
have even been thrust into the skull just after the animal's death.

In summary, the Wacissa Bison antiquus clearly indicates an association
between such species and humans in Florida during the Paleo-Indian period
at the end of the Pleistocene about 11,000 years ago. Most likely the
Bison was hunted and killed by humans; perhaps it was ambushed and butchered
at a now inundated locale adjacent to a former water hole.

S. David Webb Jerald T. Milanich
Florida State Museum Florida State Museum
Gainesville, FL. Gainesville, FL.

Roger Alexon James Dunbar
Daytona Beach, FL. Florida Division of Archives
and History
Tallahassee, FL.


Charles A. Hoffman

In January 1973, Northern Arizona University began excavations of the
site of reported findings of bones of extinct megafauna associated with Indian
artifacts near the Silver Springs attraction in north-central Florida. The
first efforts were carried out jointly by archaeologists from Northern Arizona
University, geologists from the University of Florida, and archaeologists from
the State of Florida's Division of Archives, History, and Records Management.

Our efforts were supported by a grant-in-aid from the American Broadcast-
ing Company, owners of Silver Springs. ABC Scenic & Wildlife Division not only
provided a cash grant for operating expenses, but the loan of equipment needed
to excavate aqualungss, barges, glass-bottom boats, compressed air), logistic
support, and even the use of Silver Springs personnel. The cooperation be-
tween a state university and private enterprise has been an exciting aspect
of this nine-year project and could be the subject of another paper in itself.

Excavations began at a location where Mr. George William Guest, a resident
of the Silver Springs area, discovered bones exposed in the bank of the Silver
Springs Run, several kilometers east and downstream from the main boil of the
Springs. Mr. Guest informed Benjamin Waller, one of the top avocational archae-
ologists in Florida and an expert on the Paleo-Indian about his discovery.
Waller then notified professional archaeologists and scientists about the dis-

Working with Waller and geologist H. K. Brooks of the University of
Florida, I submitted a proposal to Silver Springs, Inc., to excavate t! site.
They accepted and we worked out an arrangement to purchase part of our equip-
ment from Don Serbosek in Daytona. Because the University of Florida had an
earlier holiday schedule than that of Northern Arizona University, Brooks was
to begin removing the overburden in December, and then to draw profiles of the
stratigraphy. Northern Arizona University arrived at the site in January 1973,
to excavate the bone bed. However, when we arrived in January, the University
of Florida operation was just getting started.

The bones had been discovered eroding from the south bank of the Silver
Springs Run, about three meters beneath the surface of the water and lying
beneath one to two meters of marl sediments. The matrix of the layer in
which the bones were lying consisted primarily of finely crushed shell. This
stratum varied in thickness at the edge of the bank from 10 to 20 cm being
thickest towards the south into the bank. Lying on top of this layer of
crushed shell was a meter and a half of mucky marl. The marl lying above the
bone bed was a cross-bedded formation consisting primarily of light brown muck
with a lot of vegetal matter. On top of it were thin layers of marls of vary-
ing colors and consistencies. We assumed that all the layers above the bone
bed were deposited long after the bone bed, but we wanted to understand better
the sequence of events before and subsequent to the deposition of the bones,
and we later took core samples of these deposits.



By the end of January we had removed the overburden from much of the site,
exposing a considerable quantity of bone. We took great pains not to disturb
anything until we could get some feeling for the overall context. I wanted to
excavate the site just as we would any land site--and not remove the bones as
they were found, but leave them in for recording by drawing and photography.

Part of our control was through a series of iron pipes (wooden stakes
later were used) driven into the ground at two-meter intervals in north-south
and east-west directions, forming a grid over the site. We then used a 2 x 1
meter metal grid for mapping and measuring. This unit was sub-divided every
ten centimeters with fine wire. Vertical control was determined from a common
datum line permanently established by a reference point in a tree above the
site. All vertical measurements were referred to that line.

Silver Springs Run is the outflow from freshwater springs. This run and
others in Florida have for years yielded faunal remains and artifacts from the
Paleo-Indian period. Williams and Stoltman (1965:676) have calculated that
more fluted Paleo-Indian projectile points have been tabulated for Florida
than for any other state in the southeastern United States. Over 80 percent
of those found were in areas near the Suwannee and Santa Fe Rivers, not far
north of Silver Springs Run. This observation led Williams and Stoltman (1965)
to suggest a correspondence between the distribution of fluted projectile
points and the remains of recent megafauna, both of which occur most frequently
near north-central Florida rivers, including Silver Springs Run. However, the
well-documented association of artifacts with the remains of these megafauna
has been lacking, at least during the time we began our work, anywhere east
of the Mississippi River.

The run is bordered by zones of marsh vegetation that interface the run
and sandy ridges and knolls on both banks. Vegetation along the banks ranges
from swamp forest communities to an abbreviated savannah of palms and tall
grasses. The water in the river maintains a steady 720 F throughout the year
as compared to other rivers in the area that are from 150 to 200 F warmer during
the summer. The river's cooler temperature causes the summer air temperature
along the river to be lower.

At the end of the first field season in January 1973, we had uncovered
the remains of at least three mammoths, one of which was fairly complete. We
also found one or two bone fragments of bison, large cat, deer, turtle, and
alligator. We photographed bones in place (Fig. 1), drew them on graph paper,
and then reluctantly removed some of them hoping to find evidence of man be-
neath the bones. The only evidence recovered was a large quantity of tiny
thinning chert flakes found in the rib area of the mammoth. We then backfilled
the site and left.

In the summer of 1973, a small crew from Northern Arizona University
returned and worked with personnel from the Florida's Division of Archives,
History, and Records Management. We removed the debris we had placed on top
of the site and began to remove additional bone fragments, going deeper into
the crushed shell stratum. We encountered more mammoth bone, apparently
getting closer to the original ground surface on which the animal had laid
when it died or was killed. We then quickly began to encounter artifactual
material. A small stemless point (Fig. 2) was found by Sandra Rayl, an NAU
graduate student from Jacksonville, in the vicinity of the proximal end of
the right femur close to the ilium. Chert flakes were found in the area of
the ribs and vertebra. All this material was made from blue or white chert.


We mapped our finds, cataloging many of the bones in the field, and took
two core samples of the sediments adjacent to the main portion of the site.
With the help of E. Thomas Hemmings, we began to map the site and to prepare
a profile of the river at the site location.

The third field season produced very little. We concentrated on recording
the site stratigraphy and took three additional core samples from the site area,
and another from a promising site upstream. We did discover a "pile" of bones,
including alligator and turtle, just centimeters south of where most of the
bones had been previously uncovered.

During the fourth season we excavated a north-south profile at right angles
to the bank, and then extracted eight cores along this profile. This additional
data caused us to modify considerably our knowledge of the stratigraphic se-
quence. What we had thought of as a series of essentially horizontally laid
beds of marl was generally accurate, but these strata had been interrupted by
a buried stream channel.

Beginning with the 1973 summer season, Professor Brooks no longer worked
with us on the site, since I did believe his use of a water-cannon allowed much
control. As a result we progressed quite slowly, but we seemed to have greater
control. The fourth season we used a special pumping system, and were able to
move sediment quickly and efficiently. Because of this technique we were able
to recover most of a St. Johns Incised bowl from well above the bone bed. This
bowl is roughly square in shape and is divided down the center of the inside
by a ceramic rib or buttress. We also obtained samples of peaty muck from
inside the bowl and from beneath the bowl for future radiocarbon dating.

I returned to the site in the summers of 1976 and 1977, mostly to survey
other parts of the Silver Springs Run including a site across the river that
B. Waller and Bill Franklin had excavated (Martin 1969) in 1966. I was fortu-
nate in being able to obtain sufficient data as to revise our previous under-
standing of the stratigraphy of the later site and the relationship of its
stratigraphy to that of the Guest site. It was determined that an estimated
17,000 B.P. date assigned to it was probably incorrect. I believe that the
bone beds of both sites were probably contemporaneous and dated from around
10,000 years B.P.

Much of the work since finding the projectile point associated with the
mammoth bones has been anticlimactic. Dr. Richard H. Hevly, botanist at
Northern Arizona University, has almost completed the study of the first cores
that we took to Flagstaff. This includes the analyses of diatoms, pollen,
seeds, shells, and ostracods. All of the remaining cores have been opened
and the lithology illustrated. Processing of these is now underway and we
should have the results soon.

In the summer of 1978, I took Dr. Hevly to the Guest site. I re-exposed
the profiles for him,showed him the sequences which we had previously uncovered
and took more cores of the sediments. We then toured much of Florida, trying
to find aquatic habitats similar to those he was reconstructing for the bone
bed at Silver Springs. Again, the American Broadcasting Company came to our
aid. We received additional assistance from the Florida State Fresh Water
Fish and Game Commission, the State Division of Geology, and the University
of Florida. With their aid, and that of others, we visited such exciting
places as the Fisheating Creek near Lake Okeechobee, Lake Helen Blazes at the
head of the St. Johns River, Paines Prairie, and many fast- and slow-moving


A i4

1* 4

Figure 1. The bones of mammoths as they were exposed at the kill
site near Silver Springs. Iron pipe is marked in 10 cm
intervals; a small projectile point was found just to
right of the datum pin in the upper right hand corner.


~*":~i' ="~
i ~s~LV'
s' L- "$

Sp ^iL

^rc S8




Since Professor Brooks would not re-
spond to requests for information concern- F.
ing any radiocarbon dates, we do not have '
any for the sediments, or his interpreta- -
tion on stratigraphy of the sediments, above
the bone bed. However, by writing to all
the universities in Japan which do radio-
carbon dating, we were able to find out
that Professor Brooks sent several samples
to Gakushuin University in Tokyo. They
informed us that laboratory analysis re-
sulted in a 9840 190 (GaK-4512) B.P. date
on collagen of the mammoth bone.

A detailed study as the result of
Northern Arizona University field work
of the site was made by Sandra Rayl for l,
her Master's Thesis (Rayl 1974), and a "
detailed discussion of the site's archae-
ology and ecology is being prepared for
publication by myself and Professor Hevly, '" "
which should be available in the fall of

Figure 2. A stemless projectile S .
point found between the ."
upper right femur and the i
pelvis of a mammoth at the '
Guest site near Silver
Springs. Point measures
3 cm in length.

References Cited

Martin, Richard A.
1969 Eternal Spring. Great Outdoors Publishing Co., St. Petersburg,

Rayl, Sandra L.
1974 A Paleo-Indian Mammoth Kill Site Near Silver Springs, Florida.
M.A. Thesis on file, Department of Anthropology, Northern
Arizona University.

Williams, Stephen and James B. Stoltman
1965 An Outline of Southeastern United States Prehistory With a
Particular Emphasis on the Paleo-Indian Era. IN The Quaternary
of the United States, edited by H. E. Wright, and David G. Frey
Princeton University Press, pp. 669-683.

Charles A. Hoffman
Northern Arizona University
Flagstaff, Arizona 86011


Don Serbousek

The following is a preliminary report of
discoveries made while exploring a Paleo-Indian
site on the Aucilla River in north Florida. -V-
This site was investigated on extended
weekends every winter since 1968 by myself
with the assistance of other scuba divers, -
particularly John Cotrill and Jeff Rupple.
As an amateur archeologist, I recognize
my limitations regarding professional
excavation skills and capability, and
appeal to the professional community to
consider seriously this site. Many
years and much hard work have been put
into the recovery of artifacts from this
site but few significant conclusions can
be made without more information which
can only be obtained with the help of
trained professionals.

I asked Ben I. Waller, an amateur
archeologist from Ocala, Florida, to
view the artifacts from this site.
He identified the projectile points
from the site and provided valuable
information about the rest of the

!I V

Figure 1. Don Serbousek with special
lighting equipment.



The following are his comments:

In this remarkable assemblage of tools, spalls, projectile
points, and faunal remains, we have the unique opportunity to look
at the debris and artifacts found in a small area from the bottom
of a hole in the river.

Although each item carries its own characteristics, and
although we have no stratigraphic data, it is most interesting
to see this dramatic variety of artifactual evidence that seems
to indicate continual habitation for a period that may exceed
six thousand years.

The side-notched projectiles might be good indicators of
pre-Archaic times, during which times the remaining Pleistocene
megafauna probably became extinct.

Very likely, many of the unifacial tools from the site
might belong to pre-Archaic times, and the frequency with which
these tools occur at other pre-Archaic sites leads us to believe
that they might well be a primary tool in use during the times of
the Bolens, Greenbriers, Daltons, etc.

Following the Paleo-Indian period, we find projectile points
used during the Archaic period. These points include stemmed
varieties and other types indigenous to Archaic sites.

It is rare for a private collector to gather all this evidence
from a given location. This collection and information could give
us the opportunity to comprehend better submerged sites and to
speculate about how the artifacts were used as well as to make an
attempt to designate specific tools to specific time sequences
(Ben I. Waller, personal communication).

Equipment Description

Different pieces of equipment are required for underwater exploration,
as well as for different methods of data collecting. I have been diving in
the rivers and underground streams of central and northern Florida for twenty
years, and have had either to modify existing underwater equipment or to
invent specific equipment for specific conditions. This particular site
required a special light because of the high concentration of tannic acid;
therefore, I used a quartz halogen light in a separate unit with a reflector
behind the bulb and a lens in front of the bulb (Fig. 1). The distance from
the bulb to the lens can be varied, which varies the circumference of the beam
of light. This light uses the river water for cooling and is connected to
nickel cadmium batteries. Without this type of light, the exploration in the
Aucilla River could not have been as successful.

An underwater dredge was used to clear away the deposits of silt and
sand. Much of the time spent at this site was used in operating the dredge
to transport sand from one side of the river to the other, and in insuring
that the dredged silt and sand would not interfere with the course of the


Since the road to the site
was in such poor condition, I
equipped my van with a ten-ton
electric winch (to pull it out of
deep mud holes), a one-quarter
inch steel plate under the gas tank
for protection, and chains on the rear
wheels. The van also has positraction,
which enables it to go almost any place
a four-wheel drive vehicle can go.

Since this exploration has been conducted during the
winter months, the 560 temperature (Fahrenheit) of the water
has necessitated wearing double wet suits. We have poured
warm water between the two layers of suits for further insu-
lation, but even with this added protection, we have not
stayed in the cold water for more than two hours at a time.

Figure 2. Map of Florida
with site location.

Site Description

The Aucilla River flows out of southern Geogria across the panhandle of
Florida to the Gulf of Mexico. The southern section of the river forms the
border between Taylor and Jefferson counties (Fig. 2). This particular site
is located in the Aucilla River 35 miles west of Perry between U.S. 98 and
1-10. The river flows under 1-10 and, in some places, is a very narrow,
picturesque stream lined with oaks, cypress, a few large cedar and magnolia
trees. It threads through the Ocala Limestone Ridge, going underground, then
reappearing as small sink holes or streams up to a mile long. The site re-
ferred to in this paper is located in the Half Mile Rise, although there the
river rises and flows for more than a mile before going underground again.
The Aucilla continues as sink holes until it reaches the Gulf of Mexico where
the ancient river bed extends further into the Gulf. This association with
the Gulf causes 12 to 14 inch tides in the river's sink holes and streams.

At the upper end of the Half Mile Rise, and at other places in the Aucilla,
there is a lack of faunal material and artifacts. The bottom of the river in
these areas contains rock formations which cover the true bottom. This in-
dicates that these sections were covered by land bridges which collapsed.
The absence of faunal material and artifacts points to the collapse of the
land bridges in a post-Pleistocene period. The limestone formations are
eroded with sharp, jagged edges and some form caves or tunnels through which
a diver can swim.

The Wacissa River flows into the Aucilla River in this Half Mile Rise
portion, and creates a small waterfall of about 3 to 4 feet, depending upon
the rainfall. The Wacissa is spring-fed and swift-flowing most of the year,
which results in very clear water. However, the Aucilla flows very slowly,


going underground frequently and collecting much debris along the way. This
makes the bottom sediment of silt vary in depth from a few inches to more
than 5 feet, and the water of the river dark brown because of the tannic acid
from leaves and branches. This slow movement of the water precludes the
possibility of objects being carried downstream by the river (Waller 1970).
Further, as this section of the river appears from underground and disappears
approximately one mile downstream, the material collected probably was not
carried to the site by water movement. There has been recent discussion,
however, of the possibility of bio-turbidity with underwater objects. This
theory requires more thorough examination, but it is logical to assume that
some very small objects could be moved about by marine creatures.

Sea level may have been as much as 400 feet lower 20,000 B.P. during the
last glacial period (Leet and Judson 1961:37-38). When the glaciers began to
melt, their melt waters returned to the sea, thereby raising sea level. The
rate of rise was not steady, and different authorities list different sea
levels at different times (Brooks 1973:7-14; Gifford 1976; Gross 1972:12-15).
Land subsidence as well as sea level changes must be considered when deter-
mining water tables in Florida (Lazarus 1965:49-58). About 10,000 to 12,000
B.P., there may have been no water or very little water in this part of the
Aucilla River except during heavy rainy seasons and at the end of the Half
Mile Rise where the water depth is now 130 feet. The deeper end would have
been a natural watering hole.

The river is 75 feet wide at the point where it drops down approximately
130 feet before going underground. Upstream about 50 feet is an underwater
ledge that stretches the width of the river. This ledge is 5 feet underwater
in the center of the river, 7 feet underwater at the east bank and 10 feet
underwater at the west bank. Drifting sand and silt have formed a wedge-
shaped bank against the south side of the ledge which is 12 to 15 feet wide
at the bottom of the ledge and extends the width of the river. On the east
side of the river approximately 20 feet downstream from the ledge is another
ledge which extends about one-third the width of the river. On the north
bank of this second ledge and opposite the first one, there is a cave 4 feet
in height, 6 feet deep and 6 feet wide. This cave is 4 feet from the bottom
of the river and faces the first ledge. Two logs have been lying on the
bottom of the river between the two ledges since we first began diving at
this site in 1968. In front of the second ledge and probably 20 feet south
of the first ledge, the river bottom starts its slope down to the 130-foot
deep hole.


When we first began exploring this site, we found a bison leg bone
partially covered by sand between the two ledges in 25 feet of water, and
excavated it by fanning the sand away from the bone with our hands. The
fanning process is very slow and requires much energy underwater, but we
continued in this manner and found other faunal material as well as points
and chippings in the same location. Then we concentrated our efforts around
the logs in the area below the cave. We went through two feet of sand before
we realized that it would be very difficult to attempt to uncover the material
buried under this depth of sand. It was, therefore, decided to use a water
dredge to facilitate removal of sand. We would vacuum away the sand until
we began to see the rubble, then we would fan the rest of the sand to reach
the material in and under the rubble. We could feel projectile points,


needles, chippings and bone under this sand and rubble, but 'even with the
water dredge, this became a major operation as the sand would continually
avalanche down to the bottom of the river where we were working, and the silt
was constantly being stirred up. We then tried to remove the sand from the
top of the ledge in order to prevent it from avalanching down to the bottom.
We also wanted to view the material in situ, and knew we had to remove all
of the sand to accomplish this. After working in this manner for 11 hours
one weekend, we had cleared a portion 6 feet down from the top of the ledge,
4 feet wide and 25 feet long The work was very difficult and produced no
material along the upper part of the ledge.

Discouraged with the slow progress of this method, we went back to stair-
stepping the face of the sand slope which uncovered a foot or two of the rubble
at a time at the bottom. This method produced bone and lithic material as
well as projectile points. In one day, we picked up 14 points (including
Clovis and Suwannee), 10 pounds of stone chippings and 20 bone
tools. These were all found in the same area as the
first bison leg bone, between the two ledges.
During excavation by the stair-stepping
method, the lower jaw of a juvenile
mammoth was discovered well back
in the sand ledge (Fig. 3).

Figure 3. A Mammoth mandible.



The complete mammoth mandible weighs about 60 lbs. and contains two
unerupted molars which can be seen pushing out the first set of molars.
Further professional identification of the material collected may reveal
other juvenile mammoth bones or bone fragments; however, to date, the rest of
the skull has not been discovered. Other Pleistocene elephant evidence in-
cludes an upper vertebra, tooth plates, skull fragments, and vertebra frag-

One bison leg bone fragment shows small (possibly rodent) teeth marks
on the end of the bone. A deer leg bone fragment as well as an unidentified
bone fragment also show similar markings. Of the 385 pieces of bone material
(including teeth and jaws), only the three described have indications of teeth
marks, although closer examination may reveal more.

Other animals represented at this site include bear, beaver, camel, horse,
rabbit, skunk and sloth. Fish, alligator, turtle, turkey and other bird bones
were also discovered. Other species have not been positively identified yet.
Eight coprolites were discovered, possibly from alligator.


Projectile points found indicate periods of use at this site from 12,000
B.P. to 4,000 B.P. Twenty-seven of the 60 points recovered were identified
as Bolen, but the three Clovis points indicate the earliest time period
(Fig. 4). The points were found in association with various fauna--some being
directly adjacent to the bones while others were a few feet away. Table 1
lists the projectile points.

Over 600 pieces of stone were found which include drills, unifacial and
bifacial scrapers, hammers and flakes (Fig. 5). Sixty-three pieces of stone
have definite side notches; some with single side notches and others with
both sides notched. Some of these pieces show beginning stages of notching
with simple curved parts having been flaked from the side. Other pieces have
definite V shapes cut from the sides.

Paleo tools have been found in apparent association with Pleistocene
animals in Florida springs and rivers by many other people. Ripley Bullen
(1969) reported a discovery by Jack Dowdy of a Clovis point in the Santa Fe
River. Ben Waller (1969) described Paleo points and Pleistocene animals also
found in the Santa Fe River. Wilfred T. Neill (1964) lists 51 sites which
have produced Suwannee and Clovis-like points in apparent association with
early animals, especially mastodon. All reports so far state that no con-
clusions can be reached; however, these reports are certainly suggestive of

Bone Tools

Most of the bone tools discovered are awl-like or needle-like, and
three harpoons were identified, one with incised marks on it. One unidenti-
fied bone was cut on the side and also cut to a point. Another was cut
straight across the top. One deer antler was notched with a cut at the top
which encircled the bone, and two fish hooks were identified.


.' 1
r- -'.e A:*

* I.

Figure 4.

A small sample of the
projectile points from
the Aucilla site.
(Photograph by Don

Figure 5. Lithic tools. (Photograph
by Don Serbousek.)



~:- :~-~~":'Zi


Table 1. Projectile Points Dimension Range (in mm)

Typology Length Width Thickness Quantity

Clovis 78-82 27-32 4-8 3
Suwannee 109 30 6 1
Greenbriar 48-54 28-33 5-7 5
Bolen 40-83.5 27-42 4-9 27
Gilchrist 76 31 10 1
Wacissa 51-70.5 42-60 10 2
Florida Spike 72 32 15 1
Hamilton 55 37 11 1
Kirk Serrated 50-69 34-37 8-13 5
Savannah River 54.5-66 32-50 9-11 3
Fla. Archaic Stemmed 54-103 30.5-53 8-12 3
Levy 60-110 35-52 9-11 4
Lafayette 68 43 10 1

Deposits on Material

All material collected which was not covered with sand had a deposit of
an iron-like substance which varied from one-quarter inch to one-half inch
thick. As the faunal material or artifacts were dried, this deposit could be
scraped off, but showed a strong adherence to bone material. This deposit
has not been identified, but appeared on only those objects exposed in the
water and not those imbedded in the sand.

Visitors to Site

Ripley Bullen visited the site with Ben Waller and myself in May, 1969.
Bullen dug a number of test holes in the land around the river, but found no
evidence of habitation. He expressed great interest in the material collected
St that time, especially the harpoons. Dr. C. Vance Haynes, an authority on
he Paleo-Indian from the University of Arizona, looked at the collected
materiall in September, 1979. Dr. Haynes (1979) agreed with the identifica-
tion of the Clovis points and that the area was used by early man for a long
Period of time.

Future Research

We feel that continued exploration of this site at deeper water levels
should produce even more Paleo-Indian and faunal material than so far re-
covered. Better equipment is needed for this type of exploration and better
scientific methods should be used in order to gain more information. Also,
xplorations of the surrounding land should be conducted.


Ceramics were not found at the site which may indicate a pre-ceramic
site or a non-habitation site. However, further excavation on the banks of
the river as well as in the cave and further upstream is necessary before a
pre-ceramic designation can be made.

Without professional exploration and excavation, few conclusions can be
made regarding the assemblege of Paleo-Indian tools and points in occurrence
with late Pleistocene animal remains.

Considering that the water table in Florida was much lower at the times
indicated by the projectile points, the river bed might have resembled a deep
ravine or gorge. The cave between the two logs might have been used as a
hiding place and the area in front of the cave might have been a kill or
butcher site. As animals came into the ravine to use the watering hole, early
man might have killed and butchered them, which could account for the large
number of broken knives and tools.

The bottom of the river at the southern end of the site is more than 130
feet under water and contains sediment so deep that we have not been able to
explore or to excavate there. We feel this area will produce much more fauna
and Paleo-Indian material.

The ancient river bed, which extends into the Gulf of Mexico, probably
contains much more Paleo material and excavations there may provide more
information about the association of man and Pleistocene animals (Bullen,
personal communication, 1969).

We would be glad to provide assistance to the professional community at
this particular site or in this area.

References Cited

Brooks, H. K.
1973 Geological Oceanography. A Summary of Knowledge of the Eastern
Gulf of Mexico, State University System of Florida Institute of
Oceanography, St. Petersburg, p. 1-21.

Bullen, Ripley P.
1969 Personal communication.

1969 A Clovis Fluted Point from the Santa Fe River, Florida. The
Florida Anthropologist 22(1):36.

Gifford, John A.
1976 Sea Levels and Ancient Seafaring. American Institute of
Nautical Archaeology Newsletter Vol. 3, No. 2.

Gross, M. Grant
1972 Oceanqraphy, A View of the Earth. Englewood Cliffs, New Jersey:
Prentice Hall, Inc.

Haynes, C. Vance
1979 Personal communication. University of Arizona, Departments
of Anthropplogy and Geosciences, Tucson.


,azarus, William C.
1965 Effects of Land Subsidence and Sea Level Changes on Elevation
of Archaeological Sites on the Florida Gulf Coast. The Florida
Anthropologist 18(1):49-58.

,eet, L. Don and Sheldon Judson
1961 Physical Geology. 2nd Edition. Englewood Cliffs, New Jersey:
Prentice Hall, Inc.

eill, Wilfred T.
1964 The Association of Suwannee Points and Extinct Animals in
Florida. The Florida Anthropologist 17(1):17-32.

aller, Ben I.
1969 Paleo-Indian and Other Artifacts from a Florida Stream Bed.
The Florida Anthropologist 22(1):37-39.


Some Occurrences of Paleo-Indian Projectile Points in Florida
Waters. The Florida Anthropologist 23(4):129-134.

Don Serbousek
333 South Yonge Street
Ormond Beach, FL. 32074

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs