• TABLE OF CONTENTS
HIDE
 Abstract
 Methods
 Geographical distribution...
 Ecological distribution in...
 Status of tortoise populations...
 Louisiana
 Mississippi
 Alabama
 Georgia
 South Carolina
 Status of tortoise populations...
 Recommendations
 Acknowledgments
 References






Title: Status and distribution of the gopher tortoise (gopherus polyphemus)
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00000974/00001
 Material Information
Title: Status and distribution of the gopher tortoise (gopherus polyphemus)
Physical Description: Book
Creator: Auffenberg, Walter
Franz, Richard
Affiliation: University of Florida -- Florida State Museum
University of Florida -- Florida State Museum
 Record Information
Bibliographic ID: UF00000974
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: ltqf - AAA0755
 Related Items
Other version: Alternate version (PALMM)
PALMM Version

Table of Contents
    Abstract
        Page 95
    Methods
        Page 95
        Page 96
    Geographical distribution in Florida
        Page 97
        Page 98
    Ecological distribution in Florida
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
    Status of tortoise populations in Florida
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
    Louisiana
        Page 113
    Mississippi
        Page 114
        Page 115
    Alabama
        Page 116
    Georgia
        Page 117
        Page 118
    South Carolina
        Page 119
    Status of tortoise populations outside of Florida
        Page 120
        Page 121
    Recommendations
        Page 122
        Page 123
    Acknowledgments
        Page 124
    References
        Page 124
        Page 125
        Page 126
Full Text
The Status and Distribution of the Gopher Tortoise (Gopherus polyphemus)
By
Walter Auffenberg
Florida State Museum University of Florida Gainesville, Florida 32611
and
Richard Franz
Florida State Museum University of Florida Gainesville, Florida 32611
Abstract
The gopher tortoise (Gopherus polyphemus) is the only extant species of the genus found east of the Mississippi River. Here it is distributed in small, disjunct populations along the southeastern Coastal Plain between Jasper and Hampton counties, South Carolina, and Tangipahoa Parish, Louisiana, primarily on well-drained soils associated with late Cretaceous, late Tertiary, and Quaternary deposits and with certain xerophytic plant associations (e.g., longleaf pine-oak uplands, xeric hammocks, sand pine-oak ridges, and ruderal successional types). Precise distributional data and accompanying ecological interpretations are presented in this report by state, though we realize that tortoises disregard political boundaries. Throughout the entire range, populations are declining, and action is needed to prevent its eventual demise. Though some human activities (e.g., certain types of forest clearing) encourage tortoise population increase or dispersal into previously unfavorable areas, most are deleterious (e.g., improved pasture practices, clearing for agricultural crops, roads and real estate development, extensive mining operations, planting large tracts in slash pine, and human predation).
Concern is increasing that populations of the gopher tortoise (Gopherus polyphemus) are being reduced at an accelerating rate by human exploitation and habitat modification. Several protective programs for the species are under consideration by appropriate conservation agencies, but basic information on the status of the species is unavailable. This paper summarizes the status and distribution of G. polyphemus in the southeastern United States.
The main goals of the present study were to: (1)establish range limits for the species; (2) investigate the major factors limiting distribution;
(3) estimate man's effect on gopher tortoises;
(4) assess the prospects for future conservation and management of gopher tortoises; and (5) make appropriate recommendations.
Detailed food studies have been initiated in Florida. Later, these data will be combined with those from studies of food caloric value and yield of food resource plants. Data pertaining to activity range and movement are being obtained by several independent studies in different parts of the range. Until these data become available, our knowledge of factors affecting the distribution of the gopher tortoise will remain incomplete.
Methods
Populations of gopher tortoises were censused by counting the number of tortoise burrows located along transects that measured ISO m long by 7 m wide. In 1973, a total of 1,602 transects
95


The Status and Distribution of the Gopher Tortoise (Gopherus polyphemus)
By
Walter Auffenberg
Florida State Museum University of Florida Gainesville, Florida 32611
and
Richard Franz
Florida State Museum University of Florida Gainesville, Florida 32611
Abstract
The gopher tortoise (Gopherus polyphemus) is the only extant species of the genus found east of the Mississippi River. Here it is distributed in small, disjunct populations along the southeastern Coastal Plain between Jasper and Hampton counties, South Carolina, and Tangipahoa Parish, Louisiana, primarily on well-drained soils associated with late Cretaceous, late Tertiary, and Quaternary deposits and with certain xerophytic plant associations (e.g., longleaf pine-oak uplands, xeric hammocks, sand pine-oak ridges, and ruderal successional types). Precise distributional data and accompanying ecological interpretations are presented in this report by state, though we realize that tortoises disregard political boundaries. Throughout the entire range, populations are declining, and action is needed to prevent its eventual demise. Though some human activities (e.g., certain types of forest clearing) encourage tortoise population increase or dispersal into previously unfavorable areas, most are deleterious (e.g., improved pasture practices, clearing for agricultural crops, roads and real estate development, extensive mining operations, planting large tracts in slash pine, and human predation).
Concern is increasing that populations of the gopher tortoise (Gopherus polyphemus) are being reduced at an accelerating rate by human exploitation and habitat modification. Several protective programs for the species are under consideration by appropriate conservation agencies, but basic information on the status of the species is unavailable. This paper summarizes the status and distribution of G. polyphemus in the southeastern United States.
The main goals of the present study were to: (1)establish range limits for the species; (2) investigate the major factors limiting distribution;
(3) estimate man's effect on gopher tortoises;
(4) assess the prospects for future conservation and management of gopher tortoises; and (5) make appropriate recommendations.
Detailed food studies have been initiated in Florida. Later, these data will be combined with those from studies of food caloric value and yield of food resource plants. Data pertaining to activity range and movement are being obtained by several independent studies in different parts of the range. Until these data become available, our knowledge of factors affecting the distribution of the gopher tortoise will remain incomplete.
Methods
Populations of gopher tortoises were censused by counting the number of tortoise burrows located along transects that measured ISO m long by 7 m wide. In 1973, a total of 1,602 transects
95


were surveyed in Florida in four vegetation types by W. Auffenberg (Table 1). Because not all tortoise burrows were occupied, a correction factor must be applied to burrow numbers to derive a population estimate. Long-term data (up to 15 years) that were obtained on 122 burrows suggested that an average of 38.6% were unoccupied; therefore, this percentage is taken as a correction factor for all density estimates in the present study.
A system of 100% forage clipping (Campbell and Cassidy 1951) was used to compare gopher tortoise food resources in different habitats. This method yields a quantitative measurement of available forage expressed in weight (kg) per hectare. All grasses and forbs (Table 2) were weighed. Vegetation studies were conducted during June-July in most areas. A total of 178 quadrats (each 1 m2) were studied to determine forage yield in each of four major plant associations (Table 1). Most samples were taken within 3 m of an active burrow so that the plants obtained would reflect the types available to the tortoises. Because food species diversity is apparently not great in gopher tortoises (Auffenberg, unpublished data), a minimum of five quadrats per sample area was taken as a standard.
Several workers have previously attempted to estimate the total area originally represented by each of the major plant associations in Florida (see Harlow 1959). However, our fieldwork and newer data (Davis 1967) suggested that the estimates for tortoise habitats should be revised. Our recalculation was based on projection of the Davis (1967) vegetation map onto a Florida county map, with all known tortoise habitats superimposed. Each habitat area in each county was delineated and its extent calculated by using a Hayashi Denko area meter.
The data and conclusions for the other states in which this species is found are based on 197 tortoise colonies in five states, investigated between January 1974 and April 1975 by R. Franz (Table 3). The locations of 35 additional colonies were provided by other workers. Ecologic and geologic data were collected at most localities14 common plant species were listed as present or absent; soil profiles were obtained where possible; and soil samples from the mounds at burrow mouths were examined. Points on the range maps sometimes represent more than one locality if the colonies were near each other.
Belt transects (150 x 7 m) were also used to estimate tortoise population densities within colonies. At least two transects were obtained from each site; in general, more transects were established for larger colonies.
Burrows were judged active if the soil at the mouth had recently been disturbed by a tortoise, inactive if the soil were undisturbed but the burrow appeared to be maintained, and old if the mouth had been washed in or covered with debris.
Although we realize that species disregard political boundaries, the following information is presented by states for the convenience of users.
Table 1. Distribution of study quadrats and density transects according to vegetation and regions of Florida.
Vegetative association Number of Number of and region quadrats transects
Longleaf pine-oak
Florida panhandle 10 230
North peninsula 32 363
Central peninsula 10 158
South peninsula 0 0
Subtotal 52 751
Xeric hammock
Florida panhandle 3 32
North peninsula 15 60
Central peninsula 20 27
South peninsula 4 I
Subtotal 42 120
Sand pine-oak
Florida panhandle 0 0
North peninsula 21 61
Central peninsula 10 175
South peninsula 5 13
Subtotal 41 249
Ruderal communities
Florida panhandle 2 17
North peninsula 26 426
Central peninsula 10 19
South peninsula 43 482
Subtotal 81 944
Totals 216 2064


Table 2. Glossary of plant names used in the present study.
Scientific name Common name
Trees and Shrubs
Acer rubrum Red maple
Carya glaba Pignut hickory
Cornus florida Dogwood
Crataegus ssp. Hawthorn
Diospyros virgin Persimmon
Ilex opaca American holly
Juniperus silieicola Red cedar
Kalmia lalifolia Mountain laurel
Magnolia grandiflora Magnolia
M. virginiana Sweet bay
Myrica cerifera Wax myrtle
Persea borbonia Redbay
Pinus clausa Sand pine
P. echinata Shortleaf pine
P. elliotti Slash pine
P. paluslris Longleaf pine
P. taeda Loblolly pine
Prunus augustifolia Chickasaw plum
P. serotina Black cherry
Quercus chapmanii Chapman oak
Q. falcata Southern red oak
Q. geminata Twin oak
Q. incana Bluejack oak
Q. laevis Turkey oak
Q. laurifolia Laurel oak
Q. margaretta Post oak
Q. marilandica Blackjack oak
Q. myrtifolia Myrtle oak
Q. virginiana Live oak
Sabal etonia Scrub palmetto
S. palmetto Sabal palm
Serenoa repens Saw palmetto
Vaccinium arhoreum Sparkleberry
V. floridanum Lowbush blueberry
Yucca filamemosa Slender yucca
Grasses and Forbs
Andropogon virginicus Broom sedge
A. elliottii Broom sedge
Aristida stricia Wire grass
Eupalorium capillifolium Dog fennel
E. compositifolium Dog fennel
Opumia lata Prickly pear cactus
Panicum ssp. Panic grass
Rubus ssp. Blackberry
Smilax glauca Green briar
Sporobolus poiretii Smut grass
Cladonia sp. Reindeer moss
Geographical Distribution in Florida
General
Gopherus polyphemus is widely distributed in Florida (Fig. 1), and the stated range usually covers all of the mainland (Carr and Goin 1959; Conant 1958). Within this area, however, gopher tortoises are restricted to particular soil and vegetative types (Wright et al. 1915; Hallihan 1923; DeSolaand Abrams 1933; Carr 1940, 1952; Duellman and Schwartz 1958; Hanson 1963).
Reptiles with similar distributions in Florida include several species of considerable zoogeo-graphic and evolutionary significance (Highton 1956; Mount 1965; Telford 1965, 1966; Jackson 1973). Zoogeographic and ecologic correlates of G. polyphemus distribution have been fully discussed by Neill (1957). Dispersal routes for the xeric reptile fauna of Florida (of which G. polyphemus is a part) have been discussed by Auffenberg (1963) and Auffenberg and Milstead (1965).
Geological papers important in the distributional analysis of G. polyphemus in Florida and its ecological and zoogeographical equivalents include Cooke and Mossom (1929), Vernon (1942, 1951), Parker and Cooke (1944), and
Table 3. Summary of fieldwork outside Florida.
Level
Number Number Areas needing of
of of additional confi-
State trips colonies' work dence''
Louisiana 1 6 Tangipahoa, St. Helena Parishes **
Mississippi 2 31 Longleaf Pine Hills District west of Pearl River **
Alabama 6 40 Southeastern Alabama ***
Georgia 8 120 Coastal islands *
South Carolina 3 0 Coastal region east of Hwy U.S. 301
"Includes only the number of colonies located during fieldwork for this study.
'Some areas were not worked as extensively as other because of limited funds and long travel distances. *=greatest confidence; ***=least confidence.


Moore (1955). Important soil publications are Harper (1925), Anon. (1962), and Smith et al. (1967). General ecological references, including detailed descriptions of Florida habitats in which gopher tortoises occur are: Harper (1914, 1921), Mulvania (1931), Wright and Wright (1932), Cooke (1939), Carr (1940), Kurtz (1942), Laessle (1942, 1958), Davis (1943, 1967), Duellman and Schwartz (1958), Harlow (1959), and Snedaker and Lugo (1972). Mitchell and Ensign (1928) present the most comprehensive treatment of Florida climate. Gunter (1948) gave useful eleva-tional data. Two recent Florida atlases contain much general information of importance to the interpretation of gopher tortoise distribution (Raisz and Dunkle 1964; Wood and Fernald 1974). Additional statistical data are available (Anon. 1960, 1971, 1974). Plant identifications largely follow Small (1933), West and Arnold (1952), and Kurtz et al. (1962). A glossary of plant names used in the present study is provided in Table 2.
Gopher tortoise density in Florida depends on several factors, of which the most important is light intensity at ground level. Generally, large areas of closed canopy habitat support small tortoise populations. Highest densities are usually found where light levels are high and where there is a diversity of grasses and forbs, some of which serve as food resources.
Soil type is also an important factor in gopher tortoise population distribution and density; the greatest concentrations are usually found on well-drained, sandy soils.
Man-induced factors are usually deleterious. Sometimes, however, forest clearing may encourage an increase in tortoise populations or a dispersal into previously unfavorable areas. Adverse effects of human intervention include most practices to improve pastures, almost all clearing for citrus groves, cultivation of agricultural crops, and the presence of roads, real estate developments, and large tracts of slash pine. Mining operations in central Florida have destroyed thousands of acres of suitable habitat. Although direct predation by man is low at the present time, it persists in most rural areas where the species is commonly collected for food. Cattle and horses apparently compete with gopher tortoises only on a low, inconsequential level, for large colonies are sometimes found in heavily grazed pastures and, under some circumstances, grazing may in fact benefit tortoise populations. However, where grasses and forbs have been completely removed
Fig. 1. Expected distributional limits of G. polyphemus in Florida. Localities at which they are knowntooccur are indicated by dots. Colonies outside the major range are represented by numbers: (1) in cutover mesic hammock, (2) on small isolated ridges of xeric hammock, and (3) small, isolated patches of sand pine-scrub oak. The contour-like lines show the expected limits for the tortoise.
over sizeable areas, by whatever means, food is no longer available and tortoise populations are generally extirpated. Land-use factors that may favorably influence tortoise habitats are pulp-wood and timber cutting and controlled burning.
Species Distribution
The distribution of G. polyphemus in Florida is based on data from 291 major localities, combining colonies within a few kilometers of one another. The tortoise occurs from the eastern to western (80 to 87.31 W longitude) and from the northern to near the southern state boundaries (25.1 to 31 N latitude). In Florida the species is found from levels just above high tide to about 100 m elevation. Populations have been located in all counties except Okeechobee and Hendry, which are low and marshy. Only small and isolated colonies exist in the counties of Baker, Bay, Brevard, Broward, Charlotte, Collier, Dade, Dixie, Escambia, Flagler, Franklin, Glades, Gulf, Holmes, Indian River, Jackson, Lafayette, Martin, Monroe, Nassau, Orange, Osceola, Palm Beach, St. Johns, St. Lucie, and Wakulla, This distributional pattern is in part natural and in part due to the manipulation of Gopherus habitats (see


below). Most of the present range is in central and northern Florida especially near and to the west of the longitudinal axis of the peninsula. The tortoises are generally absent or scarce in the low zones near the coast, particularly in the poorly drained southeastern half of the peninsula. They were historically abundant over large areas in most counties along the northern boundary and thence south along the central axis to Highlands County. Populations were historically common along most beach areas from near Pensacola to at least Panama City, spotty along the upper part of the west coast of the peninsula to Hernando County, and more or less continuous to Marco Island. An isolated population is found behind the beach on Cape Sable, at the southwestern tip of the mainland. Isolated populations were formerly located from Cutler Ridge near Perrine, Florida, to Pompano Beach. From this point northward, colonies occur more or less continuously to the mouth of the St. Mary's River near Fernandina Beach.
Landform regions (Raisz and Dunkle 1964) from which Gopherus are generally absent are the Coastal Terraced Lowlands of the panhandle, the Terraced Low Marine Zone in general around the entire periphery of the state, and the Intermediate Terraced Zone of the eastern portion of the central and northern parts of the peninsula. They are also generally absent from the High Flatlands of the Central Highlands and parts of the Marianna Lowlands on the Northwestern Plateau. Gopherus are particularly widespread in the following divisions of the Central Highlands: Trail Ridge, Sink Lake Region, Central Lake Region, Lake Wales Ridge, Dissected Highlands, Archer Ridge, and Brooksville Ridge. They are also locally abundant and widespread in the Limestone Plain of the Coastal Lowlands in the western half of the central and northern parts of the peninsula.
Individual Distributions
In all non-ruderal habitats where tortoises are not disturbed, individuals exhibit a strong tendency for colony formation. This pattern is quickly disrupted with either major habitat modification or minor predation. Such colonies may include as many as 57 individuals (~ = 11.3). Adjacent colonies may be separated by several hundred meters to several kilometers, even in a uniform habitat. Interburrow distances may vary
from several centimeters to more than 150 m (~ = 86.9 47.3 m, N = 127 for all habitats combined) and are correlated with habitat type and overall density (see below). Two typical colonies are illustrated in Fig. 2.
Ecological Distribution in Florida
Only four vegetative associations in Florida are important in influencing gopher tortoise distribution and density: longleaf pine-oak uplands, xeric hammock, sand pine-oak ridges (including beach scrub), and ruderal successional types. The last are entirely man-produced and the others are largely man-disturbed, for there is little virgin land left in Florida. Before man entered Florida, habitats somewhat similar to ruderal were produced by extensive natural fires. At present, ruderal vegetational types are represented by successional stages in cutover areas, grove and orchard edges, fence rows, roadways, and similar areas. The importance to the distributions of reptiles and amphibians of ruderal and other Florida habitats is discussed by Carr (1940).
To a large extent, Florida vegetation is correlated with soil types (Table 4) and water levels. A difference in elevation as slight as 20 cm often results in marked vegetational changes, and significantly different soils and plant associations are often closely juxtaposed. Tortoise colony distribution is closely correlated with such differences (Fig. 3).
Longleaf Pine-Oak Uplands
General Description
Some of the highest land in Florida (slightly over 100 m) is found within the longleaf pine-oak uplands vegetational association (Fig. 4). The topography is generally rolling, and the mature stands are usually open, park-like savannas. The dominant tree, the longleaf pine, is usually associated with one or another species of xerophi-lous oak, which forms the lower tree canopy. Most commonly this oak is turkey oak, although bluejack oak, twin oak, and myrtle oak are locally common. Wire grass and smut grass are usually the dominant lower-level plants. Under some conditions (mixed pine-oak upland), this association may include some mesic species. In Florida the longleaf pine-oak association supports only low tortoise population densities (unless modified


Fig. 2. G. polyphemus colonies (inhabited burrows=solid circles; abandoned=hollow circles). Right side of the figure shows a colony near Astor Park, Lake County, in old field habitat surrounded by sand pine-scrub oak. Left side shows nonrandom distribution in pasture surrounded by longleaf pine-southern red oak near Alachua, Alachua County, Florida.
by man), but it is an important habitat in other southeastern states.
The soils are well drained, with the water table usually more than 1 m below the surface. Soil fertility is always low. The subsoils in which the tortoises excavate their burrows are friable and well drained. The predominant soils, mainly noncalcareous sands and sandy clays, are classified by the series names of Norfolk, Orlando, Eustis, Blanton, Lakeland, and Lakewood.
Land Use Patterns
Practices such as burning, cattle grazing, and timber cutting have been much reduced during the last two decades, resulting in a dense overstory and sparse surface flora. Fire and moderate timber cutting often result in higher yields of tortoise food species.
Gopher Tortoise Population Density
In general, the longleaf pine-oak association provides the most extensive gopher tortoise habitat in the state, originally comprising 29.3% of the total area and 15% (2,110,256 ha) of the land surface. Although much of the original habitat remains in the Florida panhandle, particularly in state and federal forests and parks, most of the central peninsula has been converted to orange groves, improved pastures, and watermelon fields.
"Timber stand improvement," the practice of removing oaks and other noncommercial trees from the pinelands, has become popular during the last decade. This modification does not seem to significantly alter tortoise population densities. Data for three different sites selectively cut in Alachua County show that tortoise density per hectare before cutting (1963) was 2.5, 3.5, and 2.9;


8135'
Fig. 3. Area near Astor Park, Lake County, Florida, showing correlation of soil type and distribution of G. polyphemus colonies (circles).


W. AUFFENBERG AND R. FRANZ Table 4. Soil characteristics within Gopherus polyphemus habitats.
Soil groups Occupy
Topography
Principal vegetative associations
Major soil associations
Red and yellow soils
Dry soils
Well-drained uplands, Developed on limestone or marine sands.
Excessively drained scrub ridges of the peninsula and dunes along the beach
Well-drained areas of several types, including all of above
Undulating to rolling
Rolling
Rolling to flat
Longleaf pineoak uplands (mixed pine oak in NW)
Xeric hammock
Sand pinescrub oak ridges
Ruderal and similar successional stages on all of the above, as well as in cutover areas of mesic hammock, tropical hammock and, rarely, flatwoods habitats
Norfolk-Ruston-Orangeburg;
Magnolia-Faceville-Tifton;
Lakeland-Eustis-Blanton
Jonesville-Chiefland-Hernando; Hague-Zuber-Fellowship; Blanton-Klej (part)
St. Lucie-Lakeland-Pomello; Palm Beach-Cocoa (beach)
All of above, plus:
Arrendondo-Gainesville-
Ft. Meade; Leon-Immokalee-
Pompano; Leon-Blanton-
Plummer
"Fide Henderson (1939).
"Smith et al. (1967).
5 years later (1968) these same sites provided estimates of 3.3, 6.1, and 1.3, respectively, showing an average density increase of 0.6 tortoise per ha. Similar data are available for several other sites in Florida (Table 5). Thus, selective logging of pines or other trees appears to have little effect on resident colonies and may sometimes even be beneficial to tortoise populations. The more common, current practice of clear-cutting and replanting in dense stands of slash or sand pine, however, has a definitely detrimental effect on tortoise populations. Even where original tortoise colonies were not eliminated during clearing operations, few grasses grow in the dense shade, and the habitat is generally poor for gopher tortoises (Table 5).
At present, about 1,211,000 ha (10.2%) of the original Florida tortoise habitat is in planted slash pine, of which about 23% (283,286 ha) was formerly longleaf pine-turkey oak.
The longleaf pine-oak association (Fig. 4) occurs in two phases. One occurs on less sandy, more moist soils and supports a diverse flora with mesic elements. This mixed oak-pine association constitutes 5.6% (792,994 ha) of the land area.
Grass and forb yields are generally high in ruderal areas (Table 6) due to moderate soil fertility. However, undisturbed areas often have a closed canopy with heavy shade, and tortoise population densities are low.
Fig. 4. Distribution of longleaf pine-oak uplands in Florida (dark areas).


Table 5. Tortoise population density per hectare and habitat modification in Florida.
Habitat Years
and between data Change in
county How modified collections density/ha
Longleaf pineoak
Alachua Clear cut 8 (1961-1970) -3.63
Selective cut 5 (1963-1968) +0.79
Selective cut 5 (1963-1968) +2.57
Selective cut 5 (1963-1968) -1.61
Columbia Slash pine planting 3.5 (1972-1975) -4.92
Selective cut 7 (1968-1975) +0.84
Levy Selective cut 6 (1969-1975) -0.27
Improved pasture 3.3 (1967-1971) -1.75
Sand pine-scrub oak
Lake Clear cut 5 (1969-1974) -0.77
Selective cut 5 (1970-1975) -1.06
Xeric hammock
Levy Improved pasture 7.5 (1967-1975) +0.22
Improved pasture 6 (1966-1972) -0.07
aData for more than 8 years were excluded because of successional effects.
The more important phase is the longleaf pine-turkey oak association, comprising 9.4% (1,313,621 ha) of the total land area. The mean corrected number of tortoises per transect is 0.35, or 0.86/ha (Table 7). Mean density of tortoises within colonies is 10.23/ha. There is significant geographic variation in tortoise population density estimates (Table 7). In general, northern peninsular populations in this vegetative type show a higher overall density (~ = 4.15 2.00) than do those in the central peninsula and panhandle portions of the state (x= 1.68 0.07 and 0.49 1.56, respectively). The reasons for this variation are not entirely clear, although the level of predation on gopher tortoises in the panhandle area is high (see below). It is doubtful that heavy predation occurs in the populations of the central peninsula.
Xeric Hammock
General Description
Hammocks of several types, from very dry to seasonally flooded, are found interspersed with other vegetation types in which gopher tortoises are commonly found. Gopher tortoises occur only in the well-drained xeric type. In many parts of the state, xeric hammocks occur within more mesic habitats or between even drier longleaf pine-oak uplands and sand pine-scrub oak. Many of the upper canopy trees are live oak, laurel oak, myrtle oak, and American holly. Grasses and forbs are often sparse (Table 6). In most areas, the xeric hammock association is distributed as small, scattered stands several hectares in extent, and these are most often
Table 6. Forbs and grasses (kilograms per hectare, with standard deviations in parentheses) in Florida tortoise habitats.
labitats Forbs Grasses Total
Longleaf pine-oak upland
60.1 (8.1)
337.4 (45.5)
397.5 (53.6)
Mixed pine-oak
Xeric hammock
Sand pine-scrub oak ridges
Ruderal
15.2 (9.6) 2.0 ( 0.2) 9.4 ( 1.5) 225.4 (31.8)
122.0 (31.1) 2.9 (0.5) 3.1 (0.5)
519.0 (73.4)
137.2 ( 29.5) 4.9 ( 0.7) 14.6 ( 2.4) 744.4 (105.3)-


Table 7. Mean number* (x) of tortoises per transect in Florida longleaf pineturkey oak uplands.
x Holes x Tortoises
County per transect per transect
Alachua 0.95 0.58
Bay 0.23 0.14
Citrus 1.17 0.84
Clay 1.14 0.70
Columbia 1.41 1.22
Gadsen 0.08 0.05
Hamilton 0.38 0.23
Hardee 1.00 0.61
Hernando 2.29 1.41
Highlands 2.00 1.23
Lake 0.21 0.13
Leon 0.07 0.04
Levy 1.46 0.90
Madison 0.71 0.44
Marion 0.68 0.42
Santa Rosa 0.07 0.04
Suwannee 0.55 0.34
Volusia 0.63 0.39
Walton 0.15 0.09
Washington 0.08 0.05
"Overall x (corrected) = 0.35 0.03; or 0.56/ha0.25.
surrounded by extensive areas of longleaf pine-turkey oak. The largest stands of the xeric hammock association occur in the central and northern parts of the peninsula; this association is largely absent in the extreme southern part (Fig. 5). The extent of the xeric hammock association combined with scrub pine types is estimated as 1,084,055 ha (7.7% of the total).
The well-drained soils are generally found over noncalcareous sands and sandy clays. The major soil series are Hernando, Orlando, Hoffman, Greenville, Tifton, Gadsden, and Scanton. Although some hammock soils are among the most productive in Florida, those of xeric hammocks are generally not very fertile. However, when the canopy is opened by cutting, tortoise food plants usually increase markedly.
Land Use Patterns
Pasturing livestock and cutting timber are the two most common uses made of xeric hammock. The former often leads to the destruction of tortoise colonies, particularly if pastures are harrowed and planted in fodder grasses. Selective timber cutting may lead to larger tortoise populations by increasing light levels and thereby en-
Fig. 5. Distribution of xeric hammock, sand pine-scrub oak, and beach scrub in Florida (dark areas).
couraging grass and forb production. Much of the original xeric hammock lands has been destroyed by housing developments because the high, dry areas with their massive live oaks were favored by early settlers for village sites.
Gopher Tortoise Population Densities
The xeric hammock association in Florida generally supports low tortoise population densities (Table 8), apparently because quantities of food plants are small (Table 6). Overall mean densities of tortoises within the habitat are 0.07/ha and within colonies are 3.09/ha. However, habitat modification causing higher light levels at the surface often results in higher population densities.
Table 8. Mean number (x) of tortoises per transect in Florida xeric hammocks.
County x Holes per transect x Tortoises per transect
Alachua 0.33 0.20
Clay 0.01 0.01
Hardee 2.00 1.23
Lake 0.20 0.12
Leon 0.11 0.07
Madison 0.33 0.20
St. Johns 0.25 0.15
Volusia 0.34 0.21
Walton 0.07 0.04
"Overall ~ (corrected) = 0.01 0.01; or 0.07/ha0.02.


Sand Pine-Scrub Oak Ridge
General Description
The sand pine-scrub oak vegetational association is usually found on rolling topography, with some dune-like areas. Indeed, all sand pine scrub oak areas are believed to be relictual shoreline features (Laessle 1958). Predominant vegetation includes Chapman oak, twin oak, myrtle oak, scrub palmetto, and sand pine. Usually few forbs and grasses are present because of the dense growth of middle-story plants (Table 6).
The predominant soil series are St. Lucie and Lakewood. Both are almost pure quartz and contain little organic matter or plant nutrients. These soils are well drained, and their agricultural value is small.
The distribution of this association, combined with xeric hammock, is shown in Fig. 5; of the total land area of Florida, only 2.4% (325,455 ha) is represented by this association. Because sand pine-scrub oak and beach scrub communities differ primarily only in the proportionate density of some tree and shrub species (Kurtz 1942), they are here combined.
Land Use Patterns
Formerly, cattle and pigs ranged over much of the scrub lands of Florida. The association gained new importance when sand pine was recognized as a pulpwood source, and much of it was cut. Within the last decade, thousands of additional hectares, particularly along the coasts, have been modified for housing and mobile home developments. Small, relictual stands are sometimes included in extensive orange groves.
Although wildfire is very destructive to mature sand pine habitat, the cones of this pine open only when exposed to high heat. As a result, controlled burning is practiced in most sand pine plantations. Brush cutting and lumbering operations are sometimes deleterious to tortoise colonies, mainly because of mechanical damage to the burrows by large machinery. The tortoises are not always able to dig their way out of the collapsed burrows.
Gopher Tortoise Population Density
In general, population densities are very low in mature phases of the association because of the low light levels and the absence of food forbs and grasses (Table 6). Tortoise populations average 1.43/ha (Table 9), but density may be much
higher in ecotonal zones or altered areas. In general, dune scrub areas contain higher densities (~= 1.19/ha) than inland ones (x- 0.37/ha); this difference is probably related to the generally higher light intensities in the generally more open seaside communities, where densities may be as high as 13.76/ha. Fires are particularly important in modifying the deep-shaded sand pine-scrub oak phase (see Succession, below).
Table 9. Mean number* (x) of tortoises per transect in Florida sand pinescrub oak (including beach scrub) habitats.
x Holes x Tortoises
County_per transect per transect
Highlands 0.37 0.23
Lake 0.01 0.01
Marion 0.00 0.00
St. Johns (open dunes) 0.00 0.00
St. Johns (open beach
scrub) 2.81 1.73
Volusia (dense beach
scrub) 0.02 0.01
"Overall x~ (corrected) = O.I20.06; or 1.43/ha 0.32.
Ruderal Communities
General Description
Ruderal communities are areas whose natural flora has been considerably altered by man (including deliberately-set fires and timbering practices). They may occur in any of the associations inhabited by gopher tortoises. Ruderal habitats may contain higher tortoise population densities than the original ones if the grasses and forbs are abundant.
Except under unusual circumstances, annually tilled fields are not inhabited by tortoises. However, after several years of abandonment, old fields begin to accumulate resident adults. These young habitats are often dominated by perennial herbaceous species, of which dog fennels (two species), broom sedges, and panic grass are the most important. Grasses and forbs are generally very dense (Table 6) because of the open canopy during early to middle successional stages. The larger plants associated with the habitat in later successional stages are often young trees of surrounding native habitats, as well as important food plants (chickasaw plum, black cherry, persimmon, prickly pear cactus, and hawthorn).


Land Use Patterns
Ruderal communities are mostly abandoned agricultural fields and overgrown pastures. However, general land clearing, road construction, and some timbering practices produce similar conditions. Although temporarily abandoned, these areas are likely to be highly modified again in the future (i.e., returned to pastures or destroyed by housing developments).
Gopher Tortoise Population Density
Of all Florida habitats, tortoise populations become most dense in ruderal types. However, density depends on several factors, the most important of which are the age of the successional stage and ground cover characteristics. High densities do not occur on land that has recently been extensively modified. Colonization depends on the availability of surrounding tortoise populations, their density, and the factors that favor dispersal. Densities varied from 0 to 4.00 tortoises per transect (Table 10); the mean densities were 26.61/ha within colonies and 8.00/ha overall.
The importance of the vegetation type surrounding areas of colonization is shown by observations made during a 20-year period near the confluence of the Santa Fe and Itchetucknee rivers, Columbia County. The immediate area is hydric to mesic hammock. No tortoises occurred there before residential development of this area in the 1950's, although colonies were found in suitable habitat (mainly longleaf pine-oak uplands) east of the hammock. Roads were bull-
dozed through the hammock area as part of the construction effort and, as a result, open grassy corridors criss-crossed unsuitable habitat. Individual tortoises from the longleaf pine-turkey oak area penetrated the mesic hammock along the roadways, but they did not move into the adjacent shaded forest. This type of penetration apparently occurs more commonly in hammocks than in any other environment.
Succession and Gopher Tortoises
The principal stages of plant succession in habitats occupied by gopher tortoises are illustrated in Fig. 6. The mesic hammock-regional climax association not inhabited by the tortoises is a deeply-shaded stage theoretically reached by either a hydric sere (not important from the standpoint of G. polyphemus) or xeric hammock (Laessle 1942). Gopher tortoises occur in communities along the psammoseric successional routes leading to xeric hammock, sand pine-scrub oak, and longleaf pineoak associations. Though tortoise population densities are low in the xeric hammock, they increase when the canopy is opened by fire or lumbering. Densities are highest in the fire-adapted associations (sand pine-scrub oak and longleaf pine-oak) or early successional stages (beach scrub and old field). Of these, longleaf pine-turkey oak usually contain the highest densities of tortoises. In the absence of fire, each of these favored xeroseres would eventually be replaced by predominantly evergreen hardwood communities (Laessle 1942, 1958),
Table 10. Mean number (x) of tortoises per transect in Florida ruderal communities.
Original x Holes x Tortoises
County_association_ per transect_per transect
Alachua Xeric hammock 2.21 1.37
Bay Longleaf pine-turkey oak 0.75 0.46
Broward Xeric hammock 2.75 1.70
Highlands Sand pine-scrub oak 2.28 1.41
Leon Longleaf pine-turkey oak 0.50 0.31
Leon Xeric hammock 0.13 0.08
Madison Xeric hammock 1.08 0.67
Madison Longleaf pineturkey oak 0.40 0.25
M artin Xeric hammock 0.40 0.25
Pasco Longleaf pine -turkey oak 4.00 2.47
Putnam Longleaf pine-turkey oak 3.33 2.06
St. Johns Land fill (spoil island) 1.09 0.67
Suwannee Xeric hammock 1.62 1.00
"Overall jf (corrected) = 0.840.09; or 8.00/ha0.86


THE STATUS AND DISTRIBUTION OF THE GOPHER TORTOISE
107
Strongly washed and torted Lett washed and sorted
Fig. 6. Schematic diagram of primary succession in Florida G. polyphemus habitats (largely after Laessle 1942). Values shown are mean number of tortoises per hectare in each successional sere.
in which tortoises are generally less abundant. Repeatedly burned xeric hammock often reverts to sand pine-scrub oak or longleaf pineturkey oak, usually resulting in an increase in gopher tortoise density.
A correlation exists between the successional stages of abandoned agriculturaMand as outlined by Laessle (1942) and the density of tortoise populations. Thus, during the first year the flora of the habitat is predominantly composed of annuals; the specific composition varies considerably due to the importance of chance colonization. Though gopher tortoises occur in these pioneer communities, colonies are never formed, and population densities are generally very low. During the second and third years, dog fennels gradually increase and tortoise population densities remain low; by the third year, broom sedges (a common food of G polyphemus) become a conspicuous element. Population density increases and tortoise colonies become established from the fourth to the tenth years, and young trees of species commonly associated with disturbed situations (such as persimmon, hawthorn, and Chickasaw plum) are abundant. Tortoise nests often become common for the first time, and populations usually prosper for several years. Populations will remain at high levels if the area is subjected to recurrent fire, pasturing, or any other factor that tends to reduce the growth and density of taller tree species in the succeeding years. Without such modification, the period after 11 years is one during which larger native trees begin to shade out and otherwise compete with ruderai species, and tortoise population densities are
generally reduced in accordance with the rate and patterns of shade production.
Tortoise populations are less affected by the temporary effects of fire and lumbering than by long-term agricultural changes. However, whereas the burning of late successional stages tend to maintain or even increase tortoise population densities, burning of the early stages may reduce densities.
Long-term changes in tortoise population densities through several successional communities are shown in Fig. 7 for a site in Lake County (several kilometers southeast of Astor Park). The site was formerly an extensive area of sand pine-scrub oak. From 1920 until 1926, when the farm was abandoned, 21.8 ha were cleared and annually planted in sweet potatoes. The successional stages in the fields began in 1926, but the entire area was periodically burned in intervening years.
In 1959 a 4.05-ha plot (selected for the long-term study) near the middle of the old field was covered with an open woodland comprised of several species of oak and sand pine. Forbs and grasses, with broom sedge (Andropogon elliotii)
733-
YEARS SINCE CLEARED
Fig. 7. Changes in tortoise population density through several successional communities at a site southeast of Astor Park, Lake County, Florida.


dominant, were abundant. Tortoise population density was calculated as 9.14 individuals per hectare, a density which was slowly reduced in the following years as the canopy closed over the area and the grasses and forbs died out. By May 1975, 16 years after burning stopped and 55 years after the land was originally cleared, all the tortoises had died or left the area. With the exception of a short period from 1965 to 1971, the rate of tortoise loss was about 1.48/ha every 5 years. This rate of loss is similar to that from several other successional sites observed over several years (Table 11). These data make it clear that tortoise population densities are highest in grassy, open-canopied associations, whether intermediate successional stages or fire-adapted, savanna-like habitats. Such habitats are most similar to those Tertiary steppe and thorn scrub environments of western North America in which the genus Gopherus is thought to have evolved (Auffenberg 1974).
Status of Tortoise Populations in Florida
Statewide Tortoise Population Estimates
One of the most difficult questions is how many gopher tortoises exist in Florida. An accurate census of tortoises in an area this size is nearly impossible, but a general estimate of the total numbers is needed so the trends in populations can be monitored.
From the estimated areas of original undisturbed conditions and of available remaining habitat (Table 12), it is evident that tens of thousands of hectares of original tortoise habitat have been altered or completely destroyed. Estimates of habitat modifications per county are
based on data provided in Citrus Industry of Florida (Anon. 1960), Raisz and Dunkle (1964), Citrus Summary (Anon. 1971), Wood and Fernald (1974), and Census of Agriculture (Anon. 1974). Agriculture, roadways, and urbanization were the only factors considered important in large-scale modification of tortoise habitat. Of the original 3,430,716 ha of tortoise habitat in Florida, about 1,500,000 ha (43.8%) are still intact. The greatest habitat loss has been in sand pine-scrub oak (70.6%), longleaf pine-oak (57.0%), and xeric hammock (37.7%) communities. Geographically, the greatest total loss has been in the central peninsula, and the greatest proportional loss in the extreme southeastern counties.
Extensive land modification has adversely at-fected the number, density, and, to some extent, the distribution of tortoise populations. The total population of gopher tortoises for each of the three major tortoise habitats in Florida (Table 13) is now estimated to be about 1,300,000; this represents a loss of about 33% of the estimated total original carrying capacity of 2,000,000. Total population estimates by county (Table 14) reflect the extent of localized habitat destruction and predation. Several Florida counties seem to lack gopher tortoises naturally; populations in others have been drastically reduced in the last decade. Certain of the northern peninsular counties nevertheless have reasonably high populations, particularly near the Suwannee River. Relatively large areas with moderate to high tortoise densities (Fig. 8), where the remaining tortoise populations of Florida are chiefly concentrated, are thus the most important sites from a conservation standpoint.
Despite the present apparently extensive range of the gopher tortoise in Florida (Fig. 1), many colonies consist of very few individuals. It is highly probable that these hundreds of small, isolated colonies will not long persist. Further-
Table 11. Tortoise population density changes and associated floral succession.
County
Prior association
Early use
Years of observation
5-year changes in no. of tor-toises per hectare
Alachua Alachua Alachua Gilchrist Lake
Mesic hammock Xeric hammock Longleaf pine-oak Longleaf pine-oak Sand pine-scrub
Cultivated field Building site Improved pasture Unimproved pasture Cultivated field
-2.72 -1.98 -0.49 -1.24 -1.48


more, of all the land environments in Florida, those inhabited by gopher tortoises are the ones that can be expected to suffer the most extensive alteration in the near future.
Impact of Man on Tortoise Populations in Florida
Although a few of man's activities have been shown to benefit tortoises (such as controlled burning), others are clearly detrimental. The most adverse of these are agricultural clearing, urban expansion, and certain forestry practices.
Agricultural Clearing
Historically, the clearing of land for cultivation has been the major means of landscape alteration. However, Florida is following a national trend toward reduction of land dedicated to agriculture, largely because of the increased efficiency of farming practices. From 1900 to 1940 in Florida, the percentage of land in farms gradually increased from about 7 to 25%, and in the next decade there was a spectacular rise to 60%. Since 1950, the amount in cultivation has slowly decreased to the present total of about 40% (Wood and Fernald 1974), and future reduction is probable.
Fig. 8. Distribution of remaining major areas containing high densities of G. polyphemus in Florida.


Table 12. Estimated total areas of original habitat (EOH) and remaining habitat (ERH) in the areas of major tortoise habitats in Florida, by counties.
Plant associations
Land Xeric Sand pine- Longleaf pine-area hammock_scrub oak_oak uplands
County (ha x 1000) EOH ERH EOH ERH EOH ERH
Alachua 238 83 44 0 0 II 3
Baker 152 0 0 0 0 0 0
Bay 200 0 0 < 1 < 1 36 29
Bradford 77 3 14 0 0 1 1
Brevard 270 38 20 10 5 0 0
Broward 316 27 2 2 0 0 0
Calhoun 146 0 0 0 0 44 33
Charlotte 191 0 0 < 1 0 0 0
Citrus 166 12 7 0 0 40 23
Clay 159 0 0 0 0 49 16.9
Collier 535 0 0 6 3 0 0
Columbia 208 15 9 0 0 37 24.3
Dade 513 7 < 1 0 0 0 0
Desoto 164 15 3 0 0 0 0
Dixie 185 18 11 0 0 19 12
Duval 206 0 0 5 2 16 2
Escambia 169 0 0 1 < 1 31 19
Flagler 128 5 3 1 < 1 0 0
Franklin 141 2 2 2 2 7 7
Gadsen 133 0 0 0 0 17 12
Gilchrist 92 25 12 0 0 40 21
Glades 206 < 1 < 1 Gulf 147 0 0 3 3 0 0
Hamilton 133 56 39 0 0 3 2
Hardee 166 22 2 0 0 0 0
Hernando 127 12 8 0 0 56 38
Highlands 276 4 Hillsborough 277 9 < 1 0 0 20 1
Holmes 125 0 0 0 0 16 8
Indian River 130 15 4 2 1 0 0
Jackson 243 3 1 0 0 39 15
Jefferson 155 0 0 0 0 0 0
Lafayette 143 34 23 0 0 0 0
Lake 273 15 6 36 18 90 51
Lee 215 0 0 5 < 1 0 0
Leon 174 20 12 0 0 44 20
Liberty 219 0 0 0 0 20 20
Madison 185 29 12 0 0 16 6
Manatee 195 32 7 Marion 420 110 32 129 127 52 15
Martin 195 0 0 10 10 0 0
Monroe 257 0 0 < 1 < 1 0 0
Nassau 169 0 0 5 4 24 21
Okaloosa 238 0 0 1 Okeechobee 201 0 0 0 0 0 0
Orange 245 27 10 0 0 76 31
Osceola 352 16 1 0 0 1 < 1
Palm Beach 519 34 < 1 2 0 0 0
Pasco 198 7 1 0 0 54 15
Pinellas 76 0 0 5 1 13 0
Polk 497 98 5 0 0 78 6


Plant associations
Land Xeric Sand pi ne Longleaf pine
area hammock scru ) oak oak uplands
County (ha x 1000) EOH ERH EOH ERH EOH ERH
Putnam 192 34 16 19 9 38 18
St. Johns 162 4 3 4 3 0 0
St. Lucie 151 0 0 9 5 0 0
Santa Rosa 263 0 0 3 3 128 126
Sarasota 152 33 12 1 < 1 0 0
Seminole 78 15 0? 0 0 3 2
Sumter 150 34 13 0 0 62 25
Suwannee 178 5 <1 0 0 77 15
Taylor 270 0 0 0 0 24 12
Union 64 2 < 1 0 0 0 0
Volusia 293 74 43 23 17 29 10
Wakulla 150 0 0 0 0 38 34
Walton 278 0 0 1 0? 58 28
Washington 157 0 0 0 0 44 26
Clear-Cutting and Monoculture
About 50% of the land in Florida remains forested, but much of it is periodically harvested for lumber and pulpwood. The fully protected public land in Florida totals 599,924 ha, of which only about 13,760 ha provide suitable habitat for tortoises. Assuming a high overall density of 2.47 tortoises/ha, we estimate that the total carrying capacity on the protected lands would be less than 34,000 tortoises. More than 90% of these would be located on State park lands, because most Federal parks and refuges in Florida are situated on low or submerged lands that are not suitable tortoise habitat.
The remainder of the forested lands on which tortoises are found is subject to continuing intensive modification by logging. As noted above, however, selective cutting is not necessarily detri-
Table 13. Estimated total remaining hectares (ERH) in favorable habitats and estimated remaining tortoises (ERT) in Florida.
Plant association ERH ERT
Longleaf pine-oak uplands 864,144 1,246,444
Xeric hammock 408,499 18,714
Sand pine-scrub oak
(including beach scrub) 223,624 29,454
Totals 1,496,267 1,294,612
mental to gopher tortoises. Furthermore, lumber production in Florida has steadily increased since 1936.
Unfortunately, selective lumber cutting has largely been replaced by pulpwood production; this usage demands dense, deeply-shaded stands of slash pine, and tortoises are not found in the older plots. Since 1928, when slash pine planting was initiated in Florida, about 1,500,000 ha have been altered for pulpwood production, and this practice will presumably continue into the foreseeable future. There is no way of estimating how many hectares of tortoise habitat were destroyed in the process, but the amount must be large. In addition, results from research on other pulp fiber may lead to further acceleration in the loss of natural habitat. Pulpwood production is highest in the northern half of the state, which is the most extensive habitat of the gopher tortoise.
Efficient pulpwood production depends on clear-cutting large areas, generally followed by deep harrowing or bulldozing to heap up stumps and unwanted material for burning. This process normally kills all the resident tortoises. The area is then densely planted in rows of slash pine. The resulting open, grassy habitat may encourage colonization for several years. Such colonies are short-lived, however, for within 10 to 15 years the fast-growing pines shade out the grasses, and the tortoises die or scatter.


Unfortunately, pulpwood production shows signs of increasing on both public and private lands and is thus a significant factor affecting the distribution and density of the tortoise population now and in the future.
Urban Expansion
Urbanization is more threatening to tortoise habitats in the future than is the expansion of agricultural clearing orforestation. Florida is now one of the fastest-growing areas in the United States, and this trend is expected to continue. The continuing population increase coupled with the trend for spacious suburban development will make significant inroads on gopher tortoise environments. Although the phenomenal human population increase in the region (78% between 1950 and 1960) abated between 1960 and 1970 (to 37%), by the year 2000 some areas in the state are expected to increase by as much as 335% (Table 14). Urbanization is expected to be greatest along the east coast and in the central area from Titusville west to Clearwater and thence southwest along the coast to Fort Myers. Projections indicate that this region will be one of the major megalopoli in the United States. Much of this area is within the range of the gopher tortoise, and the trend can be expected to result in the destruction of thousands of hectares of remaining habitat. In areas immediately adjacent to the densely urbanized areas, roadways, recreational facilities, agriculture, and related enterprises will be developed so that very little of peninsular Florida will remain unaffected. Fortunately, some parts of northern Florida can be expected to remain somewhat as they are today.
The effects of urbanization, agriculture, and roadways on tortoises over the next 25 years (Table 14) have been estimated largely on the assumptions that continued habitat destruction will remain proportional to population, and that there will be major changes neither in human attitudes nor in enforcement of regulations regarding tortoises.
The estimated tortoise population (389,200) for Florida in 2000 AD represents a loss of 68% in only 25 years and will be due almost entirely to human population pressures. Furthermore, the rate of decline, if continued, will certainly result in the elimination of all Florida tortoise colonies (except those on fully protected lands) by about 2025 AD. We estimate that the tortoise population at that time will be no more than about
21,000 individuals, scattered over a wide area in several state parks, national forests, and a few large blocks of privately owned lands.
Table 14. Human and gopher tortoise population projections for 2000 A.D.
Est. increase Tortoise population human pop., (thousands)
1972-2000 (%) Projected
County_In 1973 for 2000
Alachua 94.7 8.4 0.3
Baker 63.6 o.oa 0.0
Bay 42.8 8.8 7.8
Bradford 39.8 0.6 0.1
Brevard 77.2 5.0 0.9
Broward 117.2 0.1 0.0
Calhoun 12.8 9.9 9.5
Charlotte 326.6 o.oa 0.0
Citrus 281.1 49.1 0.0?
Clay 106.7 39.7 0.0
Collier 304.3 2.4 0.0
Columbia 55.7 56.7 34.0
Dade 64.5 0.7 0.0
DeSoto 61.6 0.1 0.0
Dixie 48.0 25.0 17.0
Duval 60.6 5.6 0.0
Escambia 42.1 5.9 4.3
Flagler 76.2 20.7 0.3
Franklin 23.8 3.8 3.0
Gadsden 0.0 5.8 1.6
Gilchrist 57.4 73.4 39.7
Glades 79.2 1.0 0.2
Gulf 13.5 73.4 39.7
Hamilton 5.1 10.4 6.4
Hardee 72.0 0.4 0.2
Hendry 145.2 o.oa 0.0
Hernando 162.7 81.5 21.6
Highlands 95.1 14.9 10.1
Hillsborough 71.4 1.7 0.0
Holmes 24.1 2.5 1.9
Indian River 124.8 1.2 0.2
Jackson 13.2 4.8 1.2
Jefferson 15.9 o.oa 0.0
Lafayette 9.4 1.1 1.0
Lake 96.0 69.0 13.3
Lee 180.1 0.5 0.0
Leon 96.6 6.1 0.2
Levy 54.0 57.5 2.7
Liberty 20.2 5.6 5.4
Madison 18.5 14.0 10.0
Manatee 118.0 0.5 0.0
Marion 63.2 145.8 19.3
Martin 177.6 8.4 6.7
Monroe 26.9 0.1 0.1
Nassau 57.7 47.8 41.0
Okaloosa 75.8 26.9 21.1


Table 14. Continued.
Est. increase Tortoise population
human pop.. (thousands)
197 2-2000 (%) Projected
County In 1973 for 2000
Orange 114.5 32.4 0.0
Osceola 235.9 0.2 0.0
Palm Beach 158.5 0.1 0.0
Pasco 335.0 15.1 0.0
Pinellas 82.4 1.1 0.0
Polk 81.2 5.9 0.0
Putnam 63.3 47.6 16.3
St. Johns 104.5 3.9 0.4
St. Lucie 101.6 4.1 1.1
Santa Rosa 74.3 40.6 10.1
Sarasota 155.2 0.5 0.0
Seminole 153.5 0.2 0.0
Sumter 107.8 26.1 0.0
Suwannee 12.0 32.3 17.0
Taylor 13.7 24.8 21.3
Union 65.5 0.1 0.0
Volusia 70.2 29.2 20.6
Wakulla 45.2 10.4 9.9
Walton 10.7 8.7 7.8
Washington 89.8 7.9 5.3
Total 1294.6 389.2
aNo gopher tortoises were found.
Louisiana Geographical Distribution
Little information is available on Gopherus polyphemus in Louisiana. According to Harold Dundee (personal communication), Gopherus was previously known only from the area north of Covington, St. Tammany Parish.
Six populations were located during this study, all in the upland area in northern St. Tammany, southern Washington, and eastern Tangipahoa parishes (Fig. 9). A seventh locality (near Tali-sheek, St. Tammany) is reported to have tortoises on the basis of two specimens seen just above the Pearl River Swamp within the past 10 years (H. Dundee, personal communication).
Woodward and Gueno (1941) divided the eastern panhandle of Louisiana into two general topographic unitsthe northern dissected upland and the southern low-lying alluvial terraces and flood and deltaic plains. The dissected upland
closely corresponds to the Longleaf Pine Hills district described by Penfound and Hathaway (1938) which corresponds to the longleaf pine-oak uplands of this paper. These authors also listed a second elevated area, the shortleaf pine-hardwood upland that occurs as a narrow band along the southern edge of the Longleaf Pine Hills. This area apparently represents a lower terrace adjacent to the upland. All Louisiana tortoise populations except one occur in the Longleaf Pine Hills (Fig. 9). The colony in Tangipahoa Parish appears to be located in the shortleaf pine-hardwood uplands.
The upland, veneered with terrace soils and composed of the Citronelle Formation, emerges from the low pine flatwoods in the south and west, gradually increasing its elevation northeastward. Its greatest elevations occur near the Pearl River in Washington Parish. The upland is apparently the result of a continuous coastward tilting along the Mobile-Tunica flexure, causing high stream gradients (Woodward and Gueno 1941) and much dissection.
The Bogue Chitto, a rather large river that flows eastward into the Pearl River, bissects the upland. Tortoises occur on both sides of the Bogue Chitto. North of the river they occur on or near the crests of low hills at elevations of 65-75 m. South of the river they are found near the crest of the bluff that parallels the Bogue Chitto at elevations between 60 and 65 m. The southern
Fig. 9. Distribution of G. polyphemus in Louisiana and Mississippi. Stippled zone in Louisiana represents the distribution of the Longleaf Pine Hills; cross-hatched zone in Mississippi represents the Oligocene limestone (Vicksburg Formation).


portions of the upland begin as a rather precipitous bluff along the river and continue as a rolling plain that slopes gently toward the southwest. Away from the crest of the bluff, the plain becomes more mesic and apparently lacks tortoises.
All the tortoise populations except those of Tangipahoa Parish and Talisheek (which occur at 36 and 31 m, respectively) are located at elevations between 60 and 75 m. Burrows are all excavated in loamy soils with a rather high clay content, and the soil mounds at the burrow mouths are usually compacted into a hard cement. In Florida, such mounds are often used for egg laying; it is not known whether females in Louisiana excavate their nests in the compacted soil. Burrows in all these colonies are usually shallow, running only a few centimeters under the surface; because of this, many collapse under the weight of vehicles or even when humans walk over them.
Ecological Distribution
Longleaf pine and blackjack oak are the dominant trees at all but one site, where a mature stand of planted loblolly pine with a closed canopy has eliminated most understory plants. Here tortoises exist only along the highway right-of-way where grasses and forbs are common. Because this area occurs within the Longleaf Pine Hills district (Penfound and Hathaway 1938), we suspect the original vegetation may have been longleaf pine. Broom sedge is the dominant herbaceous plant at all localities. Slash pine, American holly, dogwood, bluejack oak, and greenbriar are common to the colony areas; no turkey oak, wire grass, saw palmetto, slender yucca, or prickly pear cactus occur at any of the sites.
In most areas a heavy layer of leaf litter, indicating infrequent burning, tends to conceal burrow mouths. In such areas, tortoise feeding trails are useful in locating burrows.
Populations are small (Table 15), usually clumped, and occur on hillsides just below the crests. Intervening valleys apparently do not support tortoises. Most burrows are inactive, and each locality contained several old, unused burrows.
Status
Gopherus polyphemus is peripheral in Louisiana, confined to longleaf pine-mesic and sub-
mesic forests in the eastern panhandle. Tortoise populations are probably naturally small because of the mesic forest conditions, where the heavy understory tends to shade out the herbage used as food by tortoises.
"Improved" forestry practices have converted much of the state's tortoise habitat into pine plantations. Pine plantations support tortoises until the canopy begins to close; when the forage becomes less available, tortoises begin to disappear. Large-scale-deforestation, in which thousands of acres are clear-cut and site-prepared at one time, prevents recolonization from adjacent pine stands. We believe that these forestry practices will eliminate tortoises from much of their former range in Louisiana within the foreseeable future; therefore, we consider G. polyphemus to be endangered in this state.
Mississippi
Geographical Distribution
Gopherus polyphemus is associated with an upland area in Mississippi that is a continuation of the Longleaf Pine Hills of southwestern Alabama and eastern Louisiana. The upland extends across southern Mississippi, south of the Jackson Prairie physiographic unit (Lowe 1925). Tortoises are known only from the upland east of the Pearl River and south of the Oligocene limestones and marls of the Vicksburg group (Fig. 9).
The Longleaf Pine Hills slope from elevations of more than 125 m in the north to less than 30 m in the south before dipping beneath younger sediments near the coast. The coastal lowlands, including offshore islands, lack tortoises. Most of the soils in the southern part of the upland (Pearl River, Stone, George, Lamar, southern Forrest, Perry, and Greene counties) are red and yellow sandy loams probably derived from underlying Citronelle (Pliocene) Formation (Lowe 1925). In the northeastern part of the upland (Jones, Wayne, northern Perry, Forrest, and Greene counties), tortoises are apparently associated with sands of the Catahoula (Oligocene) Formation and with terrace deposits and residual sands formed from the weathering of the Hattiesburg (Oligocene) and Pascagoula (Miocene) formations. The extensive upland west of the Pearl River appears to lack tortoises. As a result of brown loams that overlie the Pliocene deposits, hardwoods are mixed with longleaf pine in this area.


Table 15. Variation in tortoise colony density and geography in four states.
No. of No. of Mean densities
Geographic areas sites transects (holes/ha)
Louisiana 6 11 1.10
Mississippi 16 42 1.16
Red clay area 3 8 0.80
Catahoula
sandstone 3 8 1.36
Citronelle,
Pascagoula,
Hattiesburg
formations 10 27 1.38
Alabama 10 48 1.58
Mio-Pliocene
formations
with terrace
deposits 3 16 2.13
Residuum 7 32 1.13
Georgia 27 58 3.13
Tuscaloosa
formation 6 12 2.87
Plio-Pleistocene
terraces 14 32 3.86
River deposits 7 14 1.62
"Densities calculated on the basis of the numberof active and inactive burrows. Old burrows, obviously not being used, were not included.
Soils associated with the Citronelle Formation, particularly those in Hancock and Pearl River counties, are primarily heavy red clays containing small amounts of sand and gravel. Soil mounds at burrow mouths pack as hard as concrete. Burrows are shallow, probably due to a slightly higher sand-gravel content in the surface layer. In other areas, residual Citronelle soils are usually more sandy, particularly in southern Forrest, Perry, and Greene counties.
Soils derived from the Hattiesburg and Pascagoula Clays are light-colored sandy clays that support a mesophyllic flora. Tortoises occur on the low hilltops, where leaching and other weathering processes have been accelerated. Populations associated with the Catahoula Formation occur in rugged, highly-eroded hilltop localities where thin beds of sandstone, sands, and clays are exposed. Several centimeters of dark sandstone cover the ground; coarse sands and thin beds of clay occur under the sandstone cap.
Ecological Association
Longleaf pine is the dominant tree throughout the range of the gopher tortoise in Mississippi, although loblolly, shortleaf, and slash pines are planted on tree farms.
Sandy soils derived from the Citronelle Formation usually have a substantial clay component, allowing for the growth of blackjack oak. Where sandy terrace materials cover the Citronelle and other formations, turkey oak and wire grass may be common. Tortoises occurring in Hancock and Pearl River counties are associated with heavy clays where the vegetation is extremely scrubby, with longleaf pine, blackjack oak, and broom sedge as the dominant plants. Conditions become more mesic eastward, and American holly, dogwood, post oak, and wax myrtle become common.
In Marion and Lamar counties, two east-west, eroded parallel ridges of the Citronelle Formation cross a flat, low-lying agricultural area. Tortoises are found on top of steep-sided hills in gravelly sands, where turkey oak and longleaf pine occur. Tortoises, shortleaf pine, and blackjack oak occur together in sandy clays on the sides of the hills. Broom sedge and wire grass occur together, though the former is more common and often invades areas where wire grass has been eliminated.
On the Pascagoula and Hattiesburg formations, tortoises are found only on residual surface sands where finer materials have been leached out. However, some clay is usually retained, allowing, for the growth of more mesophytic species such as post oak, bluejack oak, and wax myrtle, though longleaf pine remains dominant. In some areas soils are sufficiently sandy to support turkey oak and slender yucca.
A few populations of tortoises are associated with the Catahoula sandstone. These areas contain minimal herbaceous ground cover, usually wire grass or broom sedge. Burrows exist where the sandstone veneer is broken and scattered and the subsurface sands are exposed. Longleaf pine, loblolly pine, mountain laurel, and sparkleberry are the dominant plants at these sites.
Population densities of Gopherus in Mississippi are low (Table 15). The smallest populations occur in the red clays of Hancock and Pearl River counties. Populations located on surface sands associated with the Citronelle, Pascagoula, and Hattiesburg formations and in the Catahoula


sandstone are larger and have a less random distribution.
Status
Tortoises are uncommon and limited to a small area in the southeastern part of Mississippi. Rapid habitat destruction is due particularly to current tree harvesting and reforestation methods.
Alabama
Geographical Distribution
Gopherus polyphemus is limited to the southern part of Alabama (Fig. 10). Chermock (1952) reported that the tortoises were confined to Merriman's Saballian Life Zone in the state. Scattered localities are now known from outside of this area (Mount 1975; this study). Most populations occur south of the Black Belt, particularly in the Lower Coastal Plain and Red Hills regions (herpetofaunal units used in Mount 1975). The Russell and Lee County records of this study are north and east of this area and apparently represent a relict Fall Line Hills population. Mount (1975) reported additional Fall Line Hills colonies from southern Lee and Dallas counties.
Alabama tortoises occur in sands and sandy loams associated with Pleistocene terraces, certain Mio-Pliocene and Eocene deposits, and Upper Cretaceous Fall Line Hills deposits. In southern Clark, Washington, and Mobile counties in southwestern Alabama, tortoises occur on an upland (apparently part of the Lower Coastal Plain) composed of undifferentiated sediments and the Citronelle Formationa continuation of the Long-leaf Pine Hills of Mississippi and Louisiana. This upland reaches a maximum elevation of more than 175 m in southern Clark County and decreases southward. In southern Washington County and northern Mobile County, the upland almost disappears, becoming a series of low, rolling hills that eventually merge with recent coastal deposits. Tortoise populations in this area are associated with sand deposits occurring as veneers over the summits of hills rather than as integrated components of the underlying formations. These deposits represent residual Pleistocene terrace materials. In southern Clark County, the upland becomes dissected and produces a hill topography, particularly along the Tombigbee River, and the populations are found on hillsides.
Fig. 10. Distribution of G. polyphemus in Alabama. Stippled area is the Black Belt; cross-hatched area represents the Longleaf Pine Hills associated with the Citronelle Formation.Records from Mount (1975) are included on map.
Colonies known from the area north of the upland in southwestern Alabama usually occur at hilltop sites in soils that seem to be weathered Eocene sediments. The most northern of these are near Mclntyre in Clark County and south of Butler in Choctaw County (R. H. Mount, personal communication), but these were not visited during the present study.
East of the Alabama River the topography becomes flat, apparently representing an unerod-ed portion of the Citronelle Formation. The area is widely used for agriculture, and no tortoises are found here. The plain south of Monroeville is dissected by the Little River, and tortoise populations are located in coarse sand and gravel on both sides of the river. The area along Big Escambia Creek seems to be similar but was not checked.
The geology east of this area changes abruptly. Lower Coastal Plain tortoise populations in southern Conecuh, Escambia, Geneva, Houston, and southern Covington counties occur either on sand and sandy loams along rivers or on loamy soils that form part of a residuum. This residuum apparently formed when younger sediments slumped into cavities created by the weathering of underlying Eocene and Oligocene limestones. The distribution of tortoise colonies in these areas is apparently scattered, perhaps due to the relative position of the water table.
Populations in northern Conecuh, northern Covington, Crenshaw, Barbour, Hendry, and Pike counties are probably associated with the Red Hills.


Mount (1975) stated that in eastern Alabama the Red Hills region loses its integrity east of the Conecuh River. The Russell, Lee, and possibly Dallas county records occur in the Tuscaloosa Formation, which is apparently in the Fall Line Hills.
Ecological Distribution
Tortoises occur in the more scrubby portions of the Longleaf Pine Hills district that extend from Mississippi into Mobile, Washington, and Clark counties in southwestern Alabama. In Mobile and southern Washington counties, where the upland becomes a series of low hills capped with sandy terrace deposits, longleaf pine, turkey oak, saw palmetto, reindeer moss, prickly pear cactus, slender yucca, and wire grass are common. Dogwood, bluejack oak, and other mesophytic species are scattered throughout the habitat but are not common. Tortoises are most abundant in the deeper sands near the crests of hills but occur throughout much of the habitat, except in the low areas with poor drainage.
In the northern portions of the upland and in the Red Hills region, tortoises are associated with more loamy soils that promote the growth of blackjack oak and various mesophytic plants. Occasional patches of coarse sand support tortoises in a longleaf pine-turkey oak plant association.
East of the upland at Little River, tortoises are apparently confined to a small exposure of sandy gravel that supports longleaf pine, turkey oak, sparkleberry, wire grass, reindeer moss, and mountain laurel. The populations appear clumped in a narrow zone that receives maximum sun. Most of the area surrounding the site is overgrown and apparently lacks tortoises. Along the Conecuh River, small colonies exist on broad, flat levees composed of alluvial sands. Post oaks dominate these levees; longleaf pine, sparkleberry, bluejack oak, red cedar, prickly pear cactus, live oak, and broom sedge are common but scattered. Away from the rivers, tortoises occur in pinelands that resemble the longleaf pine-wire grass flat-woods described by Carr (1940). Large pines, bluejack oak, lowbush blueberry, wire grass, and blackberry form the dominant vegetation. The habitat is more or less open, and tortoise populations are small and scattered. Soils are loamy sands and not particularly well drained.
Tortoise population densities are higher in Alabama than in Louisiana and Mississippi (Table 15). In the Longleaf Pine Hills of southwestern Alabama, tortoise colonies are large, particularly on hilltops where deep surface sands are available. Here densities are nearly twice those found in other areas of the state. In colonies on the residuum, individuals are widely spaced. Populations along rivers may have a clumped or scattered distribution, depending on the extent of the available habitat.
Status
Tortoises are uncommon and limited to the southern part of Alabama where they are jeopardized by forestry practices, particularly by extensive site preparation and the development of pine plantations. Populations in southwestern Alabama could become threatened if more mechanized forestry techniques are applied.
Georgia Geographical Distribution
The range of G. polyphemus in Georgia consists of a series of small disjunct colonies occurring south and east of the Fall, Line (Fig. 11). Gopher tortoises occur in the Fall Line Hills district west of Macon and near Augusta, in weathered sandy Miocene deposits and Plio-Pleistocene marine terrace deposits of the middle and lower Coastal Plain, in recent alluvial deposits along major rivers, and on barrier islands.
Sands belonging to the Tuscaloosa Formation extend across Georgia in the area adjacent to the Piedmont and form a distinct topographic division of the Georgia Coastal Plain, called the Fall Line Hills (Cooke 1943). Large, almost continuous populations of G. polyphemus occur on these deposits in Marion, Talbot, and Taylor counties (Fig. 12). Eastward, the sand is no longer uniformly distributed but occurs in patches, and tortoises are scarce, with single colonies reported from Crawford and Bibb counties. The Richmond County record of Conant (1958) is apparently from Augusta and is most likely associated with this same topographic district. Although the habitat seemed suitable, we could not locate tortoises in the Augusta area.


Fig. 11. Distribution of G. polyphemus in Georgia. Stippled area represents the Fall Line Hills; broken line is the western limit of Hawthorne Formation. Tortoise populations associated with (a) Fall Line Hills and surficial sand capping low hills on hillsides, solid circles; (b) river dune deposits, hollow circles; and (c) beach dunes, triangles.
On the middle and lower Coastal Plain, populations of G. polyphemus are correlated with surface sands usually found as thin beds capping the low hills or on hillsides. Most often, these deposits either rest on the Hawthorne (Miocene) Formation or are residual deposits resulting from alteration of the formation. Other than populations along the Flint River and on sands in the Fall Line Hills, tortoises seem to be restricted to an area east of the western boundary of the Hawthorne Formation (Fig. 11).
Tortoises are absent from a 15-county area in southwestern Georgia known as the Dougherty Plain. Most of this low, flat area occupies the drainage basin of the Flint River and owes its topography to the highly soluble Ocala Limestone and the limy portions of the Flint River Formation (Cooke 1943). Tortoises also appear to be absent from an area south and west of Tifton; in Mitchell, Colquitt, Cook, and most of Berrien counties; and from the Okefenokee Basin.
Gopher tortoises occur on recent alluvial deposits along major creeks and rivers (Flint, Withla-coochee, Alapaha, Satilla, Altamaha, and Ocmul-
gee river systems) and on barrier beaches. Because of the shallow water table, not all river deposits support tortoises. Barrier island populations are found on Sea, St. Simons, and Cumberland islands in the southern portion of Georgia's coastal strand. Conant (1975) indicates that the population on Cumberland Island was introduced. However, Milanich's (1973) report of tortoise remains in association with Indian refuse pits on Cumberland Island (dated at about 100 A.D.) indicates that tortoises have had a long association with the islands.
Ecological Distribution
Gopherus polyphemus is intimately associated with the longleaf pine-turkey oak stands in Georgia. This association exists on sand deposits in the Fall Line Hills and on the lower and middle Coastal Plain (Wharton 1977). Populations on barrier islands are associated with xeric hammock-beach and pineland communities. In the Fall Line Hills, populations are extensive, and the habitat is almost continuous in some areas. Longleaf pine, turkey oak, wire grass, prickly pear cactus, slender yucca, lowbush blueberry, and hawthorn are the dominant plant species in these areas. Frequently bluejack and blackjack oaks occur on the more loamy soils in the ecotone between the upland habitat and the bayheads.
On the lower and middle Coastal Plain, tortoise populations are confined to small, mostly isolated patches of longleaf pine and turkey oak that occur on the sides and tops of low hills or on old river dunes. Surrounding the tortoise areas, but usually lacking tortoises, may be communities of black-
t
v i ) ...... >...
V-, Js*
>

y ~v Y
1 'v"j->; <;
> >'
a/ ~ """^ ) -------1;
Fig. 12. Distribution of G. polyphemus in the western section of the Fall Line Hills in Georgia. Extensive sandy areas are stippled.


jack oak and pine (various species), mesic hammock, or xeric hammock. Sites containing tortoises are small, frequently less than 0.2 ha. Tortoise populations in these areas have a clumped distribution and, if sufficient wire grass is available, can exist in high densities. Slender yucca, blackberry, and saw palmetto are also common at these sites. River deposits with longleaf pine and turkey oak usually have tortoise colonies, but scrubby xeric or mesic hammocks lack tortoises.
On Cumberland, St. Simons, and Sea islands tortoises are associated with extensive shell midden deposits in live oak hammocks, post-dune areas, and occasionally the leeward side of stabilized secondary dunes near the beach.Cumberland Island was cleared during the Revolutionary War period, and much of the live oak hammocks present on the island today represent second growth, dating from the Civil War period (J. Milanich, personal communication). The longleaf and slash pine areas on this island are apparently more recent, probably no more than 20 to 30 years old, originating with the Carnegie estate. Wild horses, goats, pigs, and deer prevent the shrubs from becoming dense and keep the grasses and other herbaceous plants trimmed (J. Milanich, personal communication). Sabal palm, saw palmetto, and slender yucca are common in the post-dune area. Magnolia, pignut hickory, and saw palmetto are associated with the pineland.
Population densities are highest on Plio-Pleisto-cene terrace deposits of the middle and lower Coastal Plain (Table 15). These colonies characteristically have a clumped distribution; most of the burrows are usually grouped within a limited area (less than 1 ha), particularly along an abrupt rise (such as a natural ridge, shoulder of sand road, or fire land). Densities are lowest on river deposits, and colonies are usually clumped. The Fall Line Hills populations are scattered over extensive areas, though populations maintain high densities. No estimates are available for barrier island populations.
Status
Gopherus polyphemus is uncommon but widely distributed on the Georgia Coastal Plain. The species is jeopardized in the State by habitat modification, elimination of fire from longleaf pine-turkey oak communities, repeated use of off-road vehicles in certain dune areas, and human predation.
South Carolina
Distribution
At present, G. polyphemus colonies occur in only a small part of southwestern South Carolina, specifically Jasper and Hampton counties. Colonies are restricted to the higher ridges on the east bank of the Savannah River and the west bank of the Coosahatchee River. Specific localities established during this study are Hampton County: 4.5 km northeast of Luray, 1.5 km north of Route 363, 18 km due east of Luray, 0.7 km northwest of Hwy. 601, 3 km northwest of the junction of Roads 119 and 336; and Jasper County: 6,12,16, and 18 km east of the junction of Roads 119 and 336, and 1 km south of Road 336 on the B and O Landing Road.
The major part of the known distribution of tortoises in South Carolina falls within the Savannah River Terraces of the Coastal Plain Province (Collings and Montgomery 1957). Paralleling the river, this subprovince extends well beyond the recorded colonies into western Hampton County (Fig. 13), and Gopherus probably occurs there in appropriate localities. Even so, the range of the species is very restricted, for it clearly does not occur as far north as Aiken County (Gibbons 1977). The colonies along the Coosahatchee River in central Hampton County are restricted to several small high ridges in the Blanton Subprovince (Collings and Montgomery 1957). Unlike the colonies in the Savannah River Terrace, these are small and isolated from one another. The vegetation is dominated by blackjack oak and loblolly pine. Surrounding the large colonies of the Savannah River Terrace, the vegetation is chiefly turkey oak-longleaf pine, with an understory of wire grass.
Previously the range of the gopher tortoise must have been much more extensive within the state, for during the Pleistocene it extended along the Coastal Sand Area at least as far north as Myrtle Beach, Horry County (Auffenberg 1974, unpublished data).
Investigations in the Fall Line Hills area between the Savannah River and Columbia and in the area between Orangeburg and Allendale failed to reveal gopher tortoises. The sand hills paralleling the Fall Line contain large stands of longleaf pine and turkey oak but no wire grass. Lowbush blueberry forms the dominant ground cover. None of the local people questioned knew of the gopher tortoise.


Fig. 13. Distribution of G. polyphemus in South Carolina. Cross-hatched areas. Savannah River Terraces; stippled areas, Blanton Subprovince.
Status
Throughout its South Carolina range, the gopher tortoise is rapidly being extirpated. The colonies near the Coosahatchee River are all but destroyed at present, and the more extensive ones near the Savannah River are heavily exploited for their flesh. Population density even in the best areas is very low, though apparently it was once relatively high (on the basis of destroyed burrows). Small holes are nonexistent near the major roads, suggesting that the colonies are heavily skewed toward older specimens. Much damage has been done by slash pine monoculture, though active colonies in many areas could be saved by even minimal protection from meat hunters. The state is planning to purchase some of the longleaf pine-turkey oak forests in this area with the intent of protecting the remaining colonies of gopher tortoises; if carried forward, the plan is likely to succeed. Formerly, the average colony size was apparently about eight, but human predation and disturbance have reduced the number of active holes per colony to about three. The situation is thus probably critical, and a monitored restocking program within the proposed protected area may be warranted, with the incoming animals to be drawn from surrounding heavily disturbed areas.
Status of Tortoise Populations Outside of Florida
The distribution of gopher tortoises in areas outside of Florida can be correlated with sandy
soils associated with late Cretacous, late Tertiary, and Quaternary deposits. Populations optimally require loose soils for burrow construction and low-growing grasses and forbs for food; either of these factors can become limiting. Tortoises along the Pearl River in Mississippi construct shallow burrows in dense clay soils that contain little sand except near the surface, and burrows remain near the more sandy surface. Tortoises are lacking where sand is absent.
Densely-shaded habitats, even where soil conditions seem appropriate for tortoises, usually lack tortoises. When open habitats become more shaded, tortoises become less common, eventually disappearing when low plants are eliminated.
In Georgia and southwestern Alabama the washed, coarse, marine-terrace sands support the most savanna-like sandhill associationlongleaf pine-turkey oak (Fig. 14). These are, in turn, the areas with highest tortoise densities (Table 15). Areas with more loamy soils in Louisiana, Mississippi, and Alabama support population densities only half those found in the longleaf pine-turkey oak community (Table 15). Lowest densities occur in the red clay area of Mississippi and in tree farms employing pine monoculture.
Clewell (1971) reported that the sandhill plant community is composed of more than 200 species of trees, legumes, composites, and other herbaceous and low-growing plants. However, the plant components of the sandhill association vary considerably throughout its range. It is not known whether these differences are artifacts of the extensive "cut and get out" lumbering practices of the early 1900's or reflect the actual plant composition of the original forest.
We have attempted to show these differences in a transect from west to east, from Louisiana to Georgia (Table 16). Longleaf pine is the most characteristic plant of the sandhill district, occurring at 90% of the tortoise sites. Sites without longleaf pine were obviously planted and have a loblolly and slash pine forest of uniform age. In the west, the most important species next to longleaf pine are blackjack oak and broom sedge; but these become less important in tortoise habitats eastward, where they are replaced by turkey oak and wire grass. Reindeer moss, prickly pear cactus, saw palmetto, and slender yucca are also most common in the eastern sandhills. The flora in the western portion of the tortoise's range is more diverse than in the east. Dogwood, wax myrtle, live oak, greenbriar, shortleaf pine, mountain laurel, sweet


Kig. 14. I ongleal pine-turkey oak association, a habitat type frequented by dense populations ol the gopher tortoise
bay, sweetgum, red bay, American holly, and red maple are common species in this area. These plants, combined with longleaf pine, oaks, and broom sedge, form either mesic or submesic pine-mixed hardwood forests.
In the eastern sandhills, wiregrass becomes one of the most important components of the longleaf pine-turkey oak community because of its exceptional competitive abilities which prevent the establishment of many hardwood species. Fire is important in the maintenance of wiregrass. In the western sandhills, broom sedge replaces wiregrass as the most common grass cover. In some areas, broom sedge invades areas from which wiregrass has been removed. However, it is not known whether the bluestem grasses (broom sedges) were the natural ground cover in this area historically or if they replaced wiregrass after the destruction of the longleaf pine community by the lumbering
practices of the 1900's.
Mean tortoise population densities in the different plant communities supporting tortoises are compared in Table 17. The longleaf pine-turkey oak association supports the highest densities nearly twice those of the next most productive association. The exclusion of fire in the longleaf pine-turkey oak community results in the maturing of turkey oaks, which causes rapid accumulation of leaf litter and retards the growth of wire-grass and other herbaceous foods. In such areas, gopher tortoises become scarce.
Population densities decrease westward with the increase of finer soil particles and more mesic conditions. The lowest densities occur in the longleaf pine-mesic forest and in pine monocultures because of low light intensities and lowered grass and forb production.


Table 16. Plants common to tortoise colonies in four states (number of transects given in parentheses).
Plant Louisiana (N=6) Mississippi (N=33) Alabama (N=25) Georgia (N=36)
Percent No. Percent No. Percent No. Percent No.
Dominant plants
Longleaf pine 83 5 93 31 88 22 89 32
Turkey oak 39 13 56 14 97 35
Blackjack oak 83 5 51 17 28 7 5 2
Bluejack oak 33 2 21 7 65 16 19 7
Wire grass 39 13 32 8 89 32
Broom sedge 100 6 39 13 56 14 3 1
Secondary plantsb 22
Reindeer moss 28 7 8
Prickley pear cactus 3 1 8 2 25 9
Saw palmetto 3 1 4 1 36 13
Slender yucca 9 3 4 1 33 12
Sparkleberry 6 2 28 7 11 4
Flowering dogwood 33 2 21 7 40 10 1 3
Post oak 21 7 32 8 25 9
Loblolly pine 16 1 6 2 28 7 14 5
aFound in 50% or more of the sites. bFound in 25-49% of the sites.
Recommendations
The results of this study make it clear that within certain parts of its range G. polyphemus is being extirpated at an alarming rate. About 80% of all remaining colonies are found in southern Georgia and northern Florida. The remaining 20% are scattered throughout the remainder of the species range. The situation is particularly critical in Louisiana and South Carolina, where only a few colonies remain which have been greatly reduced and are certain to suffer additional future degradation. The situation in Mississippi and Alabama is also bad, for the remaining colonies are widely scattered and have become seriously depleted in the past decade. In the Florida panhandle, many colonies have already disappeared, largely due to exploitation (for food) by local rural inhabitants. Populations are equally depleted in the extreme southeastern part of the Florida peninsula, where urbanization has been the major factor. Throughout much of the range, population decline is due largely to agriculture, including horse and cattle ranches where tortoises are often considered pests that interfere with maximum production. Additionally, dry ridges on which this species was once common have been modified by the timber industry in ways that are not conducive to colony maintenance. Combined, these factors have, in our
Table 17. Tortoise colony densities in different plant communities.
No. of No. of Mean densities Plant communities sites transects (holes/ha)
Longleaf pine-mesic (mesic
species dominant) 4 12 0.91
Longleaf pine-submesic forest (bluejack oak
prominent) 14 42 1.39
Longleaf pine-blackjack oak
forest 13 39 1.04
Longleaf pine-turkey oak 21 51 2.76
Longleaf pine parkland 4 II 1.32
Pine monoculture (early stage,
open canopy) 4 8 1.02
Pine monoculture (later stage,
closed canopy)_5_10_0.49
"Densities calculated on the basis of the number of active and inactive burrows. Old burrows, obviously not being used, were not included.


opinion, reduced the original standing crop of gopher tortoises by about 80% in the last 100 years. Furthermore, estimates for the future, at least in Florida, suggest that 70% of the remaining gopher tortoise habitat will be lost to similar activities by the year 2000, and that by 202S virtually all the existing tortoise habitat will have been destroyed. The prospects for the other states in which this species occurs are slightly brighter due to lower estimates for human population increase.
These estimates may seem pessimistic when the number of tortoise colonies in an area such as Florida (Fig. 1) is considered. Such locality maps give the impression that the gopher tortoise is a highly viable species, for it seems to be widely distributed. However, closer examination of the range maps on which colonies are indicated shows a very disturbing situationeach point designating a locality simply represents a place in which one or more gopher tortoises are known to occur. Many of these designated colonies represent only a very few individuals, and of these only a small proportion may be sexually mature. Furthermore, no data on sex ratios can be derived from these maps. The figure does not, then, reflect the distribution of known reproducing colonies. Unfortunately, we have no data as to how many adults are needed to maintain a population. The average number of adults per colony is about 11, and it is possible that this number represents a kind of critical mass needed for the colony to maintain itself in the face of heavy predation on nests and a social system that is extremely important in reproductive biology (Douglass 1976). Studies on several colonies near Gainesville, Florida, show that recruitment is generally very low. The average number of eggs per female per year is 4, or an average of 20 per colony per year. Thus, gopher tortoises have reproductive strategies completely unlike most other Florida turtles (Iverson 1977), sea turtles, or soft shells (Carr 19S2), which produce dozens or even hundreds of eggs per year. These studies also show that in many Gainesville colonies the successful hatching of at least one nest occurs only once every 5 years (W. Auffenberg, unpublished data). Additionally, mortality of the young is probably very high. As a result of these factors, the management of gopher tortoises is quite unlike that for other species that are regularly managed in the southeastern United States. The major differences are that these tortoises are rather sedentary, usually occur in colonics, have a social system that greatly affects breeding success, have a
low reproductive potential, and occur in rather restricted habitat types.
Concerned biologists and naturalists throughout the southeastern United States have recently formed a Gopher Tortoise Council, fashioned after the council formed several years ago for the desert tortoise. The prospects are that this organization will contribute much to the conservation of this species, and several worthwhile projects are already under way. However, we are of the opinion that several steps (listed below) need to be taken to insure the survival of at least representative colonies of this species before it is too late.
1. Habitat management plans on all appropriate public lands must take into consideration the uniqueness of the biota of the sandhill community, of which the gopher tortoise is often an integral part. The conservation of longleaf pine-oak habitats is most important in this regard. Its continued destruction, for whatever reason, throughout the southeastern United States is deplorable, and steps should be taken to ensure that representative areas throughout this region are preserved. Similar recommendations have been made by the Florida Committee on Rare and Endangered Species (Auffenberg 1976), but as yet there is little evidence that any significant action has been taken in any southeastern state to specifically protect this habitat.
2. Representative colonies of tortoises in several parts of the range should be conserved as soon as possible. These include the peripheral populations in Louisiana and South Carolina, the distinctive populations of the southeastern part of peninsular Florida, the populations along the Fall Line Hills of Georgia, and populations in low-lying areas in central Florida adjacent to the Gulf of Mexico, some of which are peculiar in habitat (Auffenberg and Weaver 1969) and perhaps morphology.
3. Attempts should be made to reduce the drain on surviving populations. Georgia and South Carolina protect gopher tortoises. Georgia classifies it as a nongame species and requires a scientific collecting permit for possession; South Carolina considers it an endangered species. Florida has a bag limit of five tortoises per person per day; however, there is no accompanying clarification as to sizes, sexes, or seasons in which they can be taken legally. In any event, the bag limit should, in our opinion, be eliminated, for the species is as often obtained for presumed curative properties (Auffenberg in Thomas 1978) as for actual food value. Florida Wildlife regulations forbid the sale


of gopher tortoises or their products, which is certainly a step in the right direction. The gassing of gopher tortoise burrows to obtain rattlesnakes is now also illegal in Florida. This action is particularly important in view of the fact that other protected species, such as the indigo snake (Dry-marchon corais), are common commensals in gopher tortoise burrows (see Speake and Mount 1973). Furthermore, the gopher tortoise should be placed in the threatened species category in the states of Georgia, Alabama, and Mississippi, with the possibility of being offered full protection as an endangered wildlife species should this be indicated by future studies. Georgia has a category of Species of Special Concern. This category was recommended by the Florida Committee on Rare and Endangered Species but was not accepted. Although Mississippi currently protects gopher tortoises and requires scientific collecting permits, enforcement could be improved.
4. The use of gopher tortoises in promotional stunts, such as gopher races, should be prohibited. Gopher races held in Florida require special permits from the appropriate wildlife office; however, N. Dietlein (personal communication) has reviewed the results of a large gopher race held annually in one Florida city and has been able to document that in at least some instances tortoises are injured. The practice is often cruel, for the tortoises are frequently kept in captivity for months without proper food and are often emaciated when finally released. In many areas, the annual accumulation of gopher tortoises from nearby areas has so reduced local populations that they must be obtained from areas as much as several hundred miles distant. Their capture, by "pulling" them with hooks attached to long wires or cables, often causes serious and sometimes irreparable injury. The contestant tortoises are often otherwise mistreated, and when they are finally released, it is without respect to proper habitat or colony structure.
To repeat, we believe the gopher tortoise should be accorded the protection inherent in designation as an endangered species.
Acknowledgments
The study, conducted from 1959 to 1975, was sponsored in part by the National Science Foundation (NSF Grants GB 1362 and 2725). The recent completion of the part of this study dealing
with distribution and density was made possible by financial assistance from the World Wildlife Fund (Project No. U.S. 19), provided through R. B. Bury, U.S. Fish and Wildlife Service. Facilities were provided by the Florida State Museum, University of Florida. Special mention should be made of the assistance provided by the following individuals: K. Ainslie, T. Allen, G. Einem, P. Meylan, W. T. Neill, L. D. Ober, F. S. Rose, S. Scudder, N. Tessman, W. G. Weaver, and J. Wooten; all were associated with the University of Florida at one time or another. H. W. Campbell and K. Keenlyne, U.S. Fish and Wildlife Service, provided data from the route of the proposed Florida Cross-State Barge Canal. Locality information was also provided by W. Cliburn (for Mississippi), H. A. Dundee (for Louisiana), J. W. Gibbons (for South Carolina), E. Keiser (for Louisiana, and Mississippi), R. Mount (for Alabama), and C. Wharton (for Georgia). Photographs were provided by C. Puckett; most of the illustrations and maps were prepared by N. Halliday.
References
Anonymous. 1960. Citrus industry of Florida. Fl. Dept.
Agric, Tallahassee. 134 pp. Anonymous. 1962. Generalized soil map of Florida. Fl.
Agric. Exp. Stn., Gainesville. Anonymous. 1971. Citrus summary. Fl. Agric. Stat. Fl.
Dept. Agric, Tallahassee. 134 pp. Anonymous. 1974. Census of agriculture. Bur. Census.
Part 29. Fl. Sec. 2, County Data. U.S. Printing Office,
Washington.
Auffenberg, W. 1963. The fossil snakes of Florida. Tu-
lane Stud. Zool. 10(3): 131-216. Auffenberg, W. 1974. Checklist of fossil land tortoises
(Testudinidae). Bull. Fl. State Mus., Biol. Sci. 18(3):
121-251.
Auffenberg, W. 1976. The gopher tortoise. Pp. 467-470 in
J.N. Layne, ed. Inventory of rare and endangered
biota of Florida. Fl. Audubon Soc. and Fl. Defenders
of the Environment. Microfilm. Auffenberg, W., and W.W. Milstead. 1965. Reptiles in
the Quaternary of North America. Pp. 557-568 in H.E.
Wright, Jr., and D.G. Frey, eds. The Quaternary of the
United States. Princeton Univ. Press, Princeton. Auffenberg, W., and W.G. Weaver. 1969 Gopherus ber-
landieri in southeastern Texas. Bull. Fl. State Mus.,
Biol. Sci. 13(3): 141-203. Campbell, R.S., and J.T. Cassidy. 1951. Grazing values
for cattle on 'pine forest ranges in Louisiana. La. State
Univ. Bull. 452:1-46. Carr, A.F., Jr. 1940. A contribution to the herpetology of
Florida. Univ. Fl. Publ., Biol. Sci. Ser. 3(1):118. Carr, A. 1952. Handbook of turtles. Cornell Univ. Press,
Ithaca, New York. 542 pp.


of gopher tortoises or their products, which is certainly a step in the right direction. The gassing of gopher tortoise burrows to obtain rattlesnakes is now also illegal in Florida. This action is particularly important in view of the fact that other protected species, such as the indigo snake (Dry-marchon corais), are common commensals in gopher tortoise burrows (see Speake and Mount 1973). Furthermore, the gopher tortoise should be placed in the threatened species category in the states of Georgia, Alabama, and Mississippi, with the possibility of being offered full protection as an endangered wildlife species should this be indicated by future studies. Georgia has a category of Species of Special Concern. This category was recommended by the Florida Committee on Rare and Endangered Species but was not accepted. Although Mississippi currently protects gopher tortoises and requires scientific collecting permits, enforcement could be improved.
4. The use of gopher tortoises in promotional stunts, such as gopher races, should be prohibited. Gopher races held in Florida require special permits from the appropriate wildlife office; however, N. Dietlein (personal communication) has reviewed the results of a large gopher race held annually in one Florida city and has been able to document that in at least some instances tortoises are injured. The practice is often cruel, for the tortoises are frequently kept in captivity for months without proper food and are often emaciated when finally released. In many areas, the annual accumulation of gopher tortoises from nearby areas has so reduced local populations that they must be obtained from areas as much as several hundred miles distant. Their capture, by "pulling" them with hooks attached to long wires or cables, often causes serious and sometimes irreparable injury. The contestant tortoises are often otherwise mistreated, and when they are finally released, it is without respect to proper habitat or colony structure.
To repeat, we believe the gopher tortoise should be accorded the protection inherent in designation as an endangered species.
Acknowledgments
The study, conducted from 1959 to 1975, was sponsored in part by the National Science Foundation (NSF Grants GB 1362 and 2725). The recent completion of the part of this study dealing
with distribution and density was made possible by financial assistance from the World Wildlife Fund (Project No. U.S. 19), provided through R. B. Bury, U.S. Fish and Wildlife Service. Facilities were provided by the Florida State Museum, University of Florida. Special mention should be made of the assistance provided by the following individuals: K. Ainslie, T. Allen, G. Einem, P. Meylan, W. T. Neill, L. D. Ober, F. S. Rose, S. Scudder, N. Tessman, W. G. Weaver, and J. Wooten; all were associated with the University of Florida at one time or another. H. W. Campbell and K. Keenlyne, U.S. Fish and Wildlife Service, provided data from the route of the proposed Florida Cross-State Barge Canal. Locality information was also provided by W. Cliburn (for Mississippi), H. A. Dundee (for Louisiana), J. W. Gibbons (for South Carolina), E. Keiser (for Louisiana, and Mississippi), R. Mount (for Alabama), and C. Wharton (for Georgia). Photographs were provided by C. Puckett; most of the illustrations and maps were prepared by N. Halliday.
References
Anonymous. 1960. Citrus industry of Florida. Fl. Dept.
Agric, Tallahassee. 134 pp. Anonymous. 1962. Generalized soil map of Florida. Fl.
Agric. Exp. Stn., Gainesville. Anonymous. 1971. Citrus summary. Fl. Agric. Stat. Fl.
Dept. Agric, Tallahassee. 134 pp. Anonymous. 1974. Census of agriculture. Bur. Census.
Part 29. Fl. Sec. 2, County Data. U.S. Printing Office,
Washington.
Auffenberg, W. 1963. The fossil snakes of Florida. Tu-
lane Stud. Zool. 10(3): 131-216. Auffenberg, W. 1974. Checklist of fossil land tortoises
(Testudinidae). Bull. Fl. State Mus., Biol. Sci. 18(3):
121-251.
Auffenberg, W. 1976. The gopher tortoise. Pp. 467-470 in
J.N. Layne, ed. Inventory of rare and endangered
biota of Florida. Fl. Audubon Soc. and Fl. Defenders
of the Environment. Microfilm. Auffenberg, W., and W.W. Milstead. 1965. Reptiles in
the Quaternary of North America. Pp. 557-568 in H.E.
Wright, Jr., and D.G. Frey, eds. The Quaternary of the
United States. Princeton Univ. Press, Princeton. Auffenberg, W., and W.G. Weaver. 1969 Gopherus ber-
landieri in southeastern Texas. Bull. Fl. State Mus.,
Biol. Sci. 13(3): 141-203. Campbell, R.S., and J.T. Cassidy. 1951. Grazing values
for cattle on 'pine forest ranges in Louisiana. La. State
Univ. Bull. 452:1-46. Carr, A.F., Jr. 1940. A contribution to the herpetology of
Florida. Univ. Fl. Publ., Biol. Sci. Ser. 3(1):118. Carr, A. 1952. Handbook of turtles. Cornell Univ. Press,
Ithaca, New York. 542 pp.


Carr, A.F., Jr. and C.J. Goin. 1959. Reptiles, amphibians and freshwater fish of Florida. Univ. Fl. Press, Gainesville. 341 pp.
Chermock, R.L. 1952. A key to the amphibians and reptiles of Alabama. Ala. Geol. Surv. Mus. Pap. 33:1-88.
Clewell, A.F. 1971. The vegetation of the Apalachicola National Forest: an ecological perspective. Report under contract, USDA For. Serv. Unpubl.
Collings, G.H., and P.H. Montgomery. 1957. Land resource areas in South Carolina. South Carolina Agric. Exp. Stn., Clemson Univ. 8 pp.
Conant, R. 1958. A field guide to reptiles and amphibians. Houghton Mifflin Co., Boston. 366 pp.
Conant, R. 1975. Afield guideto reptiles and amphibians of eastern and central North America. Houghton Mifflin Co., Boston. 2nd Ed. 429 pp.
Cooke, C.W. 1939. Scenery of Florida interpreted by a geologist. Fl. Geol. Surv. Bull. 17:1-164.
Cooke, C.W. 1943. Geology of the coastal plain of Georgia. U.S. Geol. Surv. Bull. 941:1-121.
Cooke, C.W., and S. Mossom. 1929. Geology of Florida. 20th Ann. Rep. Fl. Geol. Surv.: 5-294.
Davis, J.H. 1943. The natural features of southern Florida. Fl. Geol. Surv. Geol. Bull. 25:1-311.
Davis, J.H. 1967. Generalized map of natural vegetation of Florida. Agric. Exp. Stn. IFAS, Univ. Fla., Gainesville.
DeSola, C.R., and F. Abrams. 1933. Testudinata from
southeastern Georgia, including the Okefenokee Swamp,
Copeia 1933:10-12. Douglass, J.F. 1976. The mating system of the gopher
tortoise, Gopherus polyphemus, in southern Florida.
Masters Thesis, Univ. South Florida, Tampa. 79 pp. Duellman, W.E., and A. Schwartz. 1958. Amphibians
and reptiles of southern Florida. Bull. Fl. State Mus.,
Biol. Sci. 3(5): 181-324. Gibbons, W. 1977. Snakes of the Savannah River plant.
ERDA, Savannah Riv. Nat. Environ. Res. Park.
SRO-NERP 1. 26 pp. Gunter, H. 1948. Elevations of Florida. Fla. Geol. Bull.
32:1-1158.
Hallihan, T. 1923. Observations made in Duval County, northern Florida, on the gopher tortoise (Gopherus polyphemus). Copeia 1923:11-20.
Hanson. K.L. 1963. The burrow of the gopher tortoise. Quart, J. Fl. Acad. Sci. 4: 354-360.
Harlow, R.F. 1959. An evaluation of white-tailed deer habitat in Florida. Fl. Game and Freshwater Fish Comm., Tech. Bull. 5:1-64.
Harper, R.M. 1914. Geography and vegetation of northern Florida. Fl. Geol. Surv. 6th Ann. Rep.: 163-437.
Harper, R.M. 1921. Geography and vegetation of central Florida. Fl. Geol. Surv. 13th Ann. Rep.:71-307.
Harper, R.M. 1925. Generalized soil map of Florida. Fl. Geol. Survey.
Henderson, J.R. 1939. Generalized vegetation and soil maps of Florida. Fl. Agric. Exp. Stn., Gainesville, Bull. 334.
Highton, R. 1956. Systematics and variation of the endemic Florida snake genus Stilosoma. Bull. Fl. State Mus., Biol. Sci. 1(2): 73-96.
Iverson, J.B. 1977. Reproduction in freshwater and terrestrial turtles of north Florida. Herpetologica 33:205-211.
Jackson, J.F. 1973. Distribution and population phene-tics of the Florida scrub lizard, Sceloporus woodi. Copeia 1973:746-761.
Kurtz, H. 1942. Florida dunes and scrub, vegetation and geology. Geol. Bull. Fl. Dept. Conserv. (23): 1-145.
Kurtz, H., R.K. Godfrey, and K.E. Wagner. 1962. Trees of Northern Florida. Univ. of Florida Press, Gainesville. 311 pp.
Laessle, A. M. 1942. The plant communities of the Wela-ka area. Univ. Fl. Publ. Biol. Sci. Ser. 4(1):143.
Laessle, A.M. 1958. The origin and successional relationship of sandhill vegetation and sand pine scrub. Ecol. Monogr. 28:361-387.
Lowe, E.N. 1925. Geology and mineral resources of Mississippi. Miss. State Geol. Surv. Bull. 20: 5-151.
Milanich, J.T. 1973. A Deptford Phase house structure, Cumberland Island, Georgia. Fl. Anthropol. 26(3): 105-118.
Mitchell, A.J., and M.R. Ensign. 1928. The climate of Florida. Bull. Univ. Fl. Agric. Exp. Stn. 200: 91-300.
Moore, W.E. 1955. Geology of Jackson County, Florida. Fl. Geol. Surv. Bull. 37:1-295.
Mount, R. H. 1965. Variation and systematics of the scincoid lizard Eumeces egregius (Baird). Bull. Fl. State Mus., Biol. Sci. 9(5): 183-213.
Mount, R. H. 1975. The reptiles and amphibians of Alabama. Auburn Univ. Press. 347 pp.
Mulvania, M. 1931. Ecological survey ofa Florida scrub. Ecology 13:528-540.
Neill, W.T. 1957. Historical biogeography of present day Florida. Bull. Fl. State Mus., Biol. Sci. 2(7): 176-220.
Parker, G.G., and C.W. Cooke. 1944. Late Cenozoic geology of southern Florida with a discussion of the ground water. Fl. Geol. Surv. Bull. 27:1-287.
Penfound, W., and E.S. Hathaway. 1938. Plant communities in the marshlands of southeastern Louisiana. Ecol. Monogr. 8(7): 1-56.
Raisz, E., and J.R. Dunkle. 1964. Atlas of Florida. Univ. of Florida Press, Gainesville. 52 pp.
Small, J.K. 1933. Manual of the southeastern flora. Published by author. New York. 1554 pp.
Smith, F.B., R.G. Leighty, R.E. Caldwell, et al. 1967. Principal soil areas in Florida. Bull. No. 717. Univ. Fl. Agric. Exp. Stn., Gainesville. 64 pp.
Snedaker, S.C., and A.E. Lugo. 1972. Ecology of the Ocala National Forest. U.S. For. Serv., Tallahassee. 214 pp. (mimeo).
Speake, D.W., and R.H. Mount. 1973. An investigation of the effects of gassing burrows ofthe gopher tortoise, Gopherus polyphemus, on the gopher and some other inhabitants. Quart. Prog. Rep., Alabama Coop. Wildl. Res. Unit 38(4): 19-20.
Telford, S.R. 1965. Some biogeographical aspects ofthe Floridian herpetofauna. Acta Herpetol. Jpn. 2(2): 16-21.
Telford, S.R. 1966. Variation among the southeastern
crowned snakes genus Tanlilla. Bull. Fl. State Mus.,
Biol. Sci. 10(7):261-304. Thomas, M. 1978. Gopher tortoise. Fl. Nat. 51(3):2-4. Vernon, R.O. 1942. Geology of Holmes and Washington
counties, Florida. Fl. Geol. Surv. Bull. 21:1-202. Vernon, R.O. 1951. Geology of Citrus and Levy counties,
Florida. Fl. Geol. Surv. Bull. 33:1-305. West, E., and L.E. Arnold. 1952. The native trees of
Florida. Univ. Fl. Agric. Exp. Stn., Univ. Fl. Press,
Gainesville. 212 pp.


Wharton, C.H. 1977. The natural environments of
Georgia. Georgia Dept. Nat. Res., Atlanta. 227 pp. Wood, R., and E. A. Fernald. 1974. The new Florida:
atlas. Rose Printing Co., Tallahassee. 119 pp. Woodward, T.P., and A. J. Gueno, Jr. 194L The sand and
gravel deposits of Louisiana. La. Geol. Surv. Bull:l-
365.
Wright, A.H., and A.A. Wright. 1932. The habitats and composition of the vegetation of Okefenokee Swamp, Georgia. Ecol. Monogr. 2:109-232.
Wright, A.H., S.C. Bishop, and W.D. Funkhouser. 1915. A biological reconnaissance of the Okefenokee Swamp in Georgia. Proc. Acad. Nat. Sci. Phila.:l-497.
NOTE: This manuscript has been in press for over 4 years. However, even in manuscript form it has served as a stimulus and basis for several subsequent studies of this species. It is published with the full realization that some of the information has become dated in the interim. In spite of this, the authors hope that its availability in published form will continue to serve as a stimulant to additional studies of this species by others.
GPO 633


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