KARST TOPOGRAPHY AS AN INFLUENCE ON
LAND USE IN WEST CENTRAL FLORIDA
ELIZABETH MORGAN FURR ABBOTT
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
Grateful acknowledgment is extended to the members of my com-
mittee for their patience and encouragement. They are: Drs. James
R. Anderson, Clark I. Cross, Shannon McCune, Richard A. Edwards, and
Anthony F. Randazzo. I especially thank Committee Chairman Dr. James
R. Anderson for his help with the field research. The contribution
of Dr. H. K.Brooks on Florida topography was much appreciated. Drs.
R. L. Ott and James McClave are given credit for the statistical
appraisal. Professor Gilbert Tanner directed final preparation of
TABLE OF CONTENTS
List of Tables ................................
List of Figures ............................................
List of Maps ...... ..................................
The Study and Its Objectives ..............
Overview of the Study Area ................
Relevant Studies of Karst Areas ...........
Notes to Chapter I ....................
Physiographic Background for the Study
Karst Areas of the World .........
Karst Areas in the United States
Karst of Florida ................
The Eocene Limestone Plains ......
Notes to Chapter II .........
Chapter III Methods, Techniques, and Topics
Employed in Karst Area Studies ......
Topical Approach ..................
Notes to Chapter III ..........
The Effect of Karst Topography on
Agricultural Land Use .......................
Historical Perspective ....................
Pine Land ................................
Hammock Land ..............................
Prairie Land ..............................
Sinkhole Density and Land Use ..............
Land Use and Soil Type ....................
Soil Productivity .........................
Case Studies ...............................
Notes to Chapter IV ...................
TABLE OF CONTENTS (continued)
Hazards and Problems Associated
with Karst .........................
Dimensions of the Problem ........
Hazardousness of Place ...........
Hazards to Structures ............
Hazards to Equipment and Livestock
Hazards to Humans ................
Hazards and Problems to Highways
Notes to Chapter V ...........
a Resource ....
to Chapter VI..
to Chapter VI .
Chapter VII Land Values and Karst ..........
Notes to Chapter VII ...
Appendix I .................................
Appendix II .................................
Biographical Sketch ..........................
LIST OF TABLES
Population Statistics in Eight-County Study Area .
Kentucky Sinkhole Enumeration . . . . . .
Karst Areas of the World (outside of the United
States) . . . . . . . . . .
Karst Areas of the United States . . . . .
Relative Proportion of Karst Count to Total Land
and Per Cent of Land in Each Use, by Soil
Associations, 1966 . . . . . . . .
6. Normalized Average Yields for 35-Year Period:
Coastal Plain Physiographic Area . . .
7. Selected Crop Yields Computed for Gilchrist
County . . . . . . . . .
8. Selected Crop Yields Computed for Levy County
9. Average Yields Per Acre and Management Level,
Coastal Plains Physiographic Area, by
Soil Types . . . . . . . .
10. List of Natural Hazards with Sinkholes Added
11. Procedures and Comments on Specific Sinkhole
Problems Encountered by the Department of
Transportation . . . . . . .
12. Functions and Uses of Limestone . . . .
13. Value of Mineral Production in the Study
Area Counties, 1969-70 . . . .
14. Springs in the Study Area . . . . .
. . 83
. . 84
. . 87
. . 127
. . 142
. . 153
- . 158
LIST OF TABLES (continued)
15. Some Characteristics of the Shorelines of
Florida Lakes . . . . . . . ... .164
16. Water-Related Names for Streets . . . ... .170
17. Market Data on Rainbow Lakes Acreage ...... .175
18. Summary of Costs for Lake-oriented Lots in
Melrose, Florida . . . . . . ... 177
LIST OF FIGURES
Frontispiece: Scenic Pools, University of Florida,
1 View of Typical Landscape of Limestone Plains of
Willis ton . . . . . . . . .. . . 14
2 Typical Doline-Filled Karst Plain Northeast of
Chiefland . . . . . . . . . . 30
3 Cenote About Twenty Feet in Diameter (Jerome Sink
North of Newberry) . . . . . . ... 34
4 Ponor or Natural Well About Two Feet in Diameter
(Three Miles North of Newberry) . . . ... 35
5 Trapped Brahma Calf inWell Sinkhole on Highway
320 East of Chiefland . . . . . . . . 36
6 Borrow Pit Wall, Good Vertical Cross Section
of Karstification on Giglia Farm Near Chiefland 38
7 Outcrop of Surface Limestone in Plowed. Field
.Three Miles West of Gainesville . . ... 39
8 The Iron Forty, Acreage Northeast of Chiefland
Showing Iron Ore Outcrops . . . . . . .. 41
9 Planted Pine Replacing Original Longleaf Pines
(East of Newberry) . . . . . . . .. 62
10 The Writer Observing a Stand of Oak Trees
in Western Marion County . . . . . .. 64
11 Hammock Land With Sinkhole Lakes on Bellamy Road
in Northern Alachua County . . . . .. . 65
12 Erosion Around Sinkholes, Hazard to Fences on
Highway 320 East of Chiefland . . . . .. 71
13 Adverse Drainage Conditions in Cornfield North
of Alachua . . . . . . . .. .. . 93
14 High Degree of Mechanization Possible (Valley
Sprinkler) When No Sinkholes are Present (Near
High Springs) . . . . . . . . . 94
LIST OF 1IG:1UES (continued)
15 Effective Efforts to Retard Excessive Drainage
Around Sinkhole, Nort'Tcrn .'i.chua County . .
16 Residential S.inkhole Damage: Courtesy Bartow
Area Civil Defense . . . . . . .
17 Specialized Insurance Programs for Sinkhole
Coverage, Advertisement . . . . . .
18 Accepted Risk With Livestock, Hazard to Animals
on Highway 320 East of Chiefland . . .
19 Sinkhole Damage to Highways: Courtesy DOT
(Taylor County) . . . . . . . .
20 Limerock Boulders Used to Fill Small Roadside
Sinkhole near Chiefland . . . . . .
21 Cleaned Limerock Surface Showing Numerous Solut
Holes, Buda Pit Near Newberry: Courtesy DOT
22 Florida Phosphate Production, 1890-1960 . .
23 Signpost Indicates Water-Oriented Names for
Streets in Rainbow Lakes Estates . . . .
. . 147
. . 152
. . 172
24 Rainbow Lakes Estates Beach and Recreation Area
. . 95
. . 106
. . 107
. . 116
. . 125
1. Surface and Near-Surface Eocene Limestone
(Study Area) . . . . . . .
2. Regional Landforms . . . . . . .
3. Population Centers and Major Highways . .
4. Principal Soils of Northwest Central Florida .
5. The Ocala Arch . . . . . . . .
6. Bell-Trenton-Chiefland Limestone Plain, Karst
Features . . . . . . . .
7. Natural Vegetation . . . . . . .
8. Marschner's Land Use . . . . . . .
9. Selected Agricultural Areas with Insets of
Sample Areas . . . . . . .
.0. Vulnerability of Areas to Sinkhole Damage .
1. Piezometric Surface . . . . . . .
2. Mineral Resources . . . . . . .
.3. Major Springs . . . . . . . .
.4. Rainbow Lake Estates Lake Area Fluctuations
between 1940 and 1964 and the Develop-
ing Plat of the Rainbow Lakes Community
LIST OF MAPS
. . 5
. . 8
. . 10
. . 28
. . 51
. . 160
. . 169
. . 121
. . 145
. . 160
Abstract of Dissertation Presented to the
Graduate Council of the University of Florida in Partial
Fulfillment of the Requirements for the Degree of Doctor of Philosophy
KARST TOPOGRAPHY AS AN INFLUENCE ON
LAND USE IN WEST CENTRAL FLORIDA
Elizabeth Morgan Furr Abbott
Chairman: Dr. James R. Anderson
Major Department: Geography
A limestone terrane is a distinctive landscape with a particu-
lar set of contingencies to which man must adapt. The most important
adaptation is in man's use of the land. Limestone outcrops cover
several million square miles of the earth's surface, and they can vary
extraordinarily from place to place. Rock solubility is the dominant
factor in the formation of karst landscapes.
Karst is found extensively in North America, and most of Florida
is underlain by limestone sediments deposited during the Cenozoic
Era. Eight counties of west central Florida have limestone outcrops
of Eocene age, which constitute one of the largest.contiguous exposures
known anywhere. These lands were selected for study in a treatment
which is original for the area and for its scope, depth, and broad use
of today's geographical methods, to evaluate the influence of karst on
the use of the land.
Sinkholes and sporadic subsidence of land is commonplace through-
out. Other karst features include springs of varying size and water-
filled depressions. Land use has always shown some shifting, although
the shifts tend to remain within the boundaries of the original
natural vegetation, so that the overall percentages of land in crops,
pastures, and forests remain fairly stable. Yields on the limestone
plains, where soils are thinnest, are substantially lower than else-
where in the Gulf Coastal Plain Province.
Sinkhole counts were made to establish density ratios, with
counts up to 24 large sinkholes per square mile on the Bell-Trenton-
Chiefland Plain. These and other subsidence features form a hazard
which the inhabitants perceive at only a low level. The hazards
accumulate however, and when their effects are summed up, they cause
the death of livestock, imperil humans, and add to farming and produc-
tion costs. Sinkholes degrade highways and are an added expense in
road maintenance. Karst features also add to building costs, requir-
ing special exploration. The accessibility of surface pollutants to
deep aquifers through solution ducts is a threat to remote farms, as
well as to major centers such as Gainesville and the University of
Population levels have tended to be static in the study area,
owing in large part to the limited subsistence farming which is the
base of the region's economy. Karst as a resource has provided sup-
plemental wage employment from time to time, in the form of rock and
mineral extractive operations. Springs assume some economic impor-
tance as resort sites, while still others have seen intensified use
as state-sponsored recreational parks. A new factor is the develop-
ment of residential communities for sale to out-of-state persons,
using the presence of lakes as a chief sales attraction. Such land
use is new to the area, and can have a major impact on the demography
and economics of the region.
In overview, the chief influence of karst is seen to be a limit-
ing one. The areas were settled sparsely, as second-choices after more
desirable lands were taken up, remained as small economic units and
limited the production and growth of the region. Limited crop yields
persist to the present, and farming is either marginal or declining.
There is much natural beauty in the region, however, and new communi-
ties using karst features may completely shift the basis of land use
in the future.
The Study and Its Objectives
Physical or natural factors of the environment include climate, topo-
graphy, soils and water cycles. These factors are interrelated in their
effect on man and in his response to the environment. One of the more
significant is topography, to be discussed here as it affects man's use
of the land for his home and work. The physical milieu sets limits on
human activity and provides the framework within which man must make his
way. He, in turn, modifies or disrupts the natural operation of the phys-
ical factors. A limestone terrane is a distinctive landscape offering a
particular set of contingencies to which man must adapt, his response
being most measurable in the observation of his use of the land.
The major objective of this study is to collect and organize sel-
ected information end data concerning the karst country of west central
Florida for the purpose of gaining an understanding of the effect of sol-
ution landforms on land use in the region. This region is a limestone
terrane, chiefly the Eocene outcrop on the axis of the Ocala Uplift. The
counties referred to in this study lie wholly or largely within the most
extensive contiguous exposure of the Ocala Group in Florida.
Secondary objectives are as follows: to examine the geomorphology
of the region in order to contribute to existing works concerning the
karst features present; to assay the peculiar impact of karst landscape
upon patterns of agriculture; to analyze economic circumstances as in-
fluenced by limestone topography; and to relate the study to certain re-
cent investigations concerning environmental hazards.
Overview of the Study Area
An introduction to the eight-county region shows Dixie, Levy, and
Citrus counties as coastal lands almost wholly representative of the
Eocene outcrop. Lafayette, Gilchrist, Alachua, Marion, and Sumter
counties are more or less representative. Outliers of Miocene deposits
remain to a significant extent as the strip of Alachua Clay or Hawthorne
deposits running through the area.
As much as possible, any discussions or descriptions included in
the study will be relative to karst. Difficulty arises with appraisal
of certain economic factors such as forest products or industrial out-
put, but emphasis will be on the more direct influences such as lime-
stone evaluated as a resource.
This introduction to the study area will briefly describe location,
population, climate, soils and vegetation, with the view that an understand-
ing of the regional geography is basic to understanding the relationships
between land use and karst. A close examination of the geomorphology of
the area is included as a major component of this work.
The study area comprises over six thousand square miles in central
peninsular Florida, approximately between 29* and 30* north latitude
(see lap 1). The region borders on the low-energy Gulf Coast of the
upper peninsula. The counties lie in the physiographic province of the
Coastal Plain, as defined by Fenneman,1 but extend into the higher areas
of "upland" Forida known as the Central Florida Ridge (see Map 2). The
Eocene-age outcrop is rimmed on the east by Miocene deposits and on the
north and south by Suwannee limestone of Oligocene age.
The study area includes the Steinhatchee River, some of the Suwannee
River, and parts of the Santa Fe River, the Oklawaha River, and the With-
lacoochee River. The Waccasassa River is entirely within the study area.
The research is focused on the "heartland" of this physical region, the
limestone plains stretching north-to-south through Gilchrist, Levy, Alachua,
and Marion counties. Chiefly agricultural areas, these plains offer indi-
vidual examples for a closer view. In addition to the selected case studies,
use was made of the Conservation Needs Inventories made in 1958 and 1966
by the United States Soil Conservation Service. The focus was again on
the limestone plains as the central and most influential sub-region.
The area has never supported a large population and has grown only
very slowly since the first American settlement. The earliest inhabitants
were Indians of the Archaic Phase who subsisted by hunting, fishing, and
foraging. Traces of this early and intermittent land use are found along
the coast and on the banks of streams and springs. Mounds of shells remain
today, and there is evidence of former coastal occupation by Indians in the
"drowned" shallows of the Gulf shore. The Apalachee and Timucua of North
Florida reached a near-civilization grouped in permanent towns with pali-
saded walls and large houses. They grew corn in extensive fields surround-
ing the villages.3
The earliest white men in the area were those of the party of Pan-
filo Narvaez, a Spanish explorer journeying northward from Tampa Bay to
w W w
---- -- - OUVAL
SNNE, CM' / Io ,- ge
S0 oO I ST JOHNS
COLUMBIA C v \
/E 1 OBRADFORD|I / C--
Gap o a
,; Ridge-. ^A ^ r^^Palat ka ,1
I San M I teo
S. E 1 Hills
l ig L^\ \ o ,, i;^ o) f
S- a inata r..
/' Sumter Ao |
Map 2 EFA 72
-ource: Florida Geological Survey
Source: Florida Geological Survey
10 Up larfd-
the vicinity of Tallahassee. Narvaez, as did de Soto later, likely tra-
versed ancient Indian trails. Some of these same trails, along favorable
topography, later became the early roads through the region. Spanish
missions were emplaced as far west as present-day Alachua County by 1606,
but it is thought that they survived only about a hundred years. There
are practically no remains of these settlements today, but evidence indi-
cates that the early Indians and the Spanish colonists as well subsisted
on diets of corn, pumpkins, beans, game, and fish.5
With the decay of the missions and the mounting trouble with the
English, the sparsely scattered Spanish occupants vanished almost without
trace. Activity by white men was at that time located chiefly along the
coast. Though back in Spanish hands yet another time, west central pen-
insular Florida was little influenced by Spanish culture. Even with
increasing economic and military movement in Florida, the limestone plains
of the central area remained almost unaffected, probably because of the
lack of surface streams for inland navigation. With the gradual settle-
ment by farmers, the area became part of the Middle Florida agriculture
belt.6 The Lafayette land grant, a gift to the Marquis in 1824, did much
to advertise the region to settlers, though Lafayette himself never saw
Transportation into the interior was aided as roads were built to
connect forts during the Seminole Wars. Congress authorized a road from
Pensacola to St. Augustine which later became known as the Bellamy Road.
Portions of this roadway remain today, with one such marking the use of
the "natural bridge" route over the Santa Fe River. The town now named
Lake City was once called Alligator, leading to the term Alligator Road
for part of the route.
Cotton was grown throughout the region but subsistence farming was
the most prevalent land use. The soil was quickly depleted, requiring
more and more clearing of "new" land. The coming of the railroads led
to exploitation of the forests forlumber and turpentine. The region re-
mained in general farming with little urbanization through the period
of the War Between the States and even into the twentieth century. Urban
places are few even today with Gainesville and Ocala (neither actually
wholly in the study area) being the largest (see map 3). Population
statistics illustrate the economic situation in these counties (see Table
1), not especially prosperous when considered in relation to the rest of
Climate, Soils and Vegetation
In the study area the average January temperature is close to 580 F.
and the average July temperature is 800 F. There is frost in winter, so
that tropical conditions do not exist; but cold spells rarely last more
than three days. Except in coldest periods, the temperature is above freezing
in the daytime. Maximum rainfall occurs usually in June and July, and the
minimum is in October, November, and December; the average annual rainfall
amounts to approximately 50 to 55 inches. Relative humidity is high and
heavy morning fogs occur occasionally. There is a growing season of
approximately 290 to 310 days per year.
The soil map of Florida indicates 34 soil associations and three
miscellaneous land types in the state as a whole. The associations are
composed of soils with similar origins and characteristics. Many such
associations are found in the study area, but only the principal groupings
POPULATION CENTERS and MAJOR HIGHWAYS
U4' *.r. u I
s I .-.J . '
\\ ,,.'r c ,' S/ F - "
S.. / I .. .. 3 4 L a k e .
',- ,, ,, C'.,,:1
F F 7Bel ie
1 -' '- -
1 0 I l il
Trent N-- ,,,
C EFA 72
Source:, Florida Geological Map 3
t2i Tr, .ei ': ," i
Rr J .+ ..... .4 + --,
are shown on the soil map (see Map 4). The soils indicated are pre-
dominantly sandy and developed on materials little altered by climate
and vegetation. Variations appear to reflect parent material, because
climatic differences are not pronounced enough to cause contrasts within
the small region. The parent material is principally limestone and sand
with some clay and alluvial deposits present also.
Since the study area encompasses parts of the Central Florida Ridge
and the Coastal Plain Province, the soils have somewhat greater differ-
ences than might be expected in a small area. However, the underlying
limestone with its high porosity predisposes an overall excessively
drained aspect except in areas with clay or silt. The groupings in the
area range from solid dominantly thick to moderately thick acid sands
(formed on Pleistocene sands, dunes and terrances) to soils dominantly
thick to thin sands influenced by alkaline materials.
The surface of the region is nearly flat to gently sloping and
the native vegetation varies with soil types and drainage. High hammock
land occurs in the eastern parts characterized by a mixture of hardwoods,
pines, shrubs, and grasses. Some low hammock land with heavier plant
growth is found near lakes in the southern part of the region, but dom-
inant vegetation throughout the entire area is pine woods. Stands of
turkey oak and longleaf pine are typical of the sand hills.
There are few remains of native vegetation or virgin forests.
Practically all of the area has been farmed or logged off. In the lime-
stone plains of the Bell-Trenton-Chiefland and High Springs-Newberry-
Archer-Williston-Reddick areas, the presence of clumps of trees, espec-
ially live oaks, usually indicates sinkholes or outcrops of limestone.
NORTHWEST CENTRAL FLORIDA
0 10to 20
Source: General Soil Map of Florida
-rl (n r4l CM r-
O 1 C1 0 1 r-
4-1 0 c
0 04r 0
-It en CM r-l
-~ ') '.J 1-
rl rll 00 C' 0 rc
41 %0 -4 Lin
0 : a ao a
E-4i 'M' 0 Lon ,,
00 0 r-
-3" v CMI
0 13- Cc
I 01 r-4 0
I a as a
(' -4 CM
tn tn cn
4 -2 -4
0 0 0
0 IP I t I 0 cO
1-1 0 p
O C' uO O C', CM CO N -
-l 0 -T o r cu r 0 Lf -
~3 r- c cn M M
10 4. pJ
ul <4 p 1-4 _- Z on
M M W
0 a C
0 0 0
fr< < E-
Soils in general show the effects of erosion. In certain local
areas, wind is a chief agent in the removal of the topsoil. The land-
scape is often beautiful, with a "rolling" aspect such as that found in
the Bluegrass region of Kentucky. Green pastures with small and large
clumps of trees and vegetation are typical (see Figure 1). The last
forests of pines were cut chiefly in the 1920's and 1930's, During
the 1930's wind erosion became such a severe problem on land stripped
.of vegetation that the area was sometimes compared to the dust bowl of
the west. The problem was greatly reduced by the introduction of
SPensacola Bahia grass as cover for the soil.*
The well-known World War II writer, Ernie Pyle, travelling through
this region in 1935 on Highway 41, noted the wind erosion and the mar-
ginal character of the subsistence homesteads. Pyle's observations
are quoted verbatim in deference to his unique powers of description.
We drove up through north-central Florida toward
Georgia. That part of Florida is no more like the
winter-resort ads than an electric razor is like the
Brooklyn Bridge. It is plain old deep South, except
that it isn't so good-looking; it lacks the luxuriant
vegetation, the greenness and freshness. It is a land
of gray sandy soil, brown scrub pines, little farms,
and many shacks. It is country that makes you mel-
The wind was blowing a steady, insistent gale.
It was hot with an uncanny, foreboding heat. You
had a feeling that this wind would slowly, relent-
lessly blow away north Florida. Queer clouds were
in the sky. And then we began to run through dust
storms. From every open field the dust streamed
across the road; from big fields it rose high into
gray clouds. We had to keep running up the windows,
and we were soon dusty all over and grit was in our
teeth. Although we never had to turn on our lights,
as we did in west Kansas, we said to ourselves,
"This is an incipient dust bowl. It has every ear-
mark of the desolated areas in Kansas and Oklahoma.
State Office Agricultural Stabilization Conservation Service.
The soil is light, and lifts more easily than the
western loam. The timber is scrubby, and they're
clearing too many fields. If the government doesn't
start doing something, they'll be making moving
pictures of the great Florida dust bowl in a few
We stopped in a town for lunch. I asked the
woman at the counter if it blew like this every
spring. She said, "Oh, no, this is the first wind
we've had." I went up to two men on a street cor-
ner. They were farmers-typical overalled cracker
farmers. I talked to them about the possibility
of this section's blowing away.
"No, no," they said, "this don't amount to
nothing Next month will be worse than this, but
this country ain't blowin' away. We get lots of
rain here, better than fifty inches a year. And
there's lots of springs hereabouts. No, there's
no danger of a dust bowl here."
This was corn, tobacco, and peanut country.
1 talked with the farmers a long time. I couldn't
get the dust-bowl idea out of my head-for I had
seen the real dust bowl, and they hadn't. I put
one last question: "Have you ever had any bad
droughts around here?"
"No, sir," one of them said, "we've never had
a single one, except once. That was pretty bad.
The water got up over the highway, and they had
to rebuild the road. It drowned out all the
crops around here. The water was even right up
in the streets here in town."
That seemed to settle the question. I had
intended asking if they ever had any floods,
but after that I didn't see any sense in it.
Double rows of planted pines outline many fields, evidence of
other efforts to slow the relentless winds. They add to the beauty of
the landscape as they bound the secondary roads and fencerows. Geometric
rows of huge acreages of planted pine add green to the drab fields be-
Farms today are dispersed rather evenly throughout the limestone
plains and livestock graze.in large fields. Corn is a chief crop because
it is required for feed for hogs, an important animal in the economy.
Karst landforms are widespread except in the flatwoods areas un-
derlain by plastic sediments in abandoned river valleys or in 3raben-
like structures. Sinkholes and outcrops are the chief karst features
and are numerous in the plains area where there is no surface drainage
Relevant Studies of Karst Areas
The history of this study can be traced through the published
literature and unpublished research accounts. The materials generally
fall into one of two categories--subsurface geology and geomorphology.
Only a few studies have been made which attempt to correlate the physical
milieu with cultural, or human, aspects. Most of these refer to climate
as a major influence on man. Rarely does an author pursue his study
into the increasingly complex relationship between physical factors of
soil, vegetation, slope and drainage, and man in his work on the earth.
That a relationship exists is of course obvious, but there are many
unanswered questions concerning the degree of significance.
The earliest recorded scientific analysis of the characteristic
limestone landscape is probably that of Lamarck. Well-known in other
fields, he made this single contribution to geology in Hydrogeology,
published at his own expense in 1802. He states his conclusions con-
cerning the origin of limestone rocks, drawn from careful observations
of fossils and stratigraphic relationships. His statements about
geologic time are surprising, considering the religious climate of his
day. His understanding of the various "displacements" of the ocean
basins along with an awareness of subaerial erosion processes offer
a beginning for analysis of limestone terranes.
A century of increasing study of the earth's surface finally pro-
duced a classic in landform analysis. William Morris Davis, well known
as "the great definer," coined a wealth of descriptive terms for the
study of structure, process, and stage by geomorphologists. In this
vein Jovan Cvijic in 1918 adapted the cycle of erosion to a limestone
landscape of Yugoslavia. His conception of the cycle as applied to a
karst region is most important today for the substantive vocabulary,
imparting an invaluable unity to the students of solution geomorphology.
For instance, this is the first use of such nonfamiliar terms as "ponor"
or "polje." Many of the words are of Serbo-Croat derivation but French
and German nouns are also employed.
E. M. Sanders (1921) condensed and summarized Cvijic on karst in
the Geographical Review, thus establishing a core vocabulary directed
toward physical geographers.11 Much more recently, H. Lehmann (1960)
described attempts at compilation of a uniform and descriptive karst
vocabulary for the Commission on Karst Phenomena on the occasion of the
XIX International Geographical Congress.12 The recommendation that an
International Atlas of Karst be prepared did not come to fruition, but
-the United States Geological Survey did publish a Glossary of Karst
Terms (1970) by Watson Monroe.13 This work is a synthesis of definitions
of karst words as they are generally used, without significant departure
from the vocabulary of textbooks by Lobeck or Thornbury.
The understanding of patterns of world karst depends heavily on
an agreement as to the meaning of terms used in the literature. With
the latest work by Monroe there has evolved at least an English language
karst vocabulary. The matter of definition gains importance when attempts
are made to enumerate karst features or to tabulate sinkhole densities.
Counts of Indiana and Kentucky sinkholes have been occasionally attempted.
Some examples are "Land of Ten Thousand Sinks" in Transactions of Kentucky
Academy of Science (1924),14 or "Sink Holes" in This Sculptured Earth
(1959).15 (See Table 2.)
As stated before, publications are numerous concerning the geology
of the solution landscape. Notable among these are Cotton (1941) on
"Limestone Landscapes,"16 Swinnerton (1942) on "Hydrology of Limestone
Terranes,1 and Stringfield and Le Grand (1971) on Artesian Water in
-Tertiary Limestone in the Southeastern United States (1966).18 These geo-
logical publications offer much in the way of background information for a
study of a limestone terrane but do not include the cultural aspects so
vital to a determination and analysis of use of the land. A study by
Wellington Jones, "A Method of Determining the Degree of Coincidence in Dis-
tribution of Agricultural Uses of Land With Slope-Soil-Drainage Complexes"
(1930) compares features on topographic maps with distribution of crops and
other agricultural land uses.19 More recently Woodford Garrigus describes
relief of land as a determinant of use in "Roughness of Terrain as a
Factor in the Areal Variation of Agricultural Productivity in Ohio" (1965).20
These publications do not refer specifically to limestone terranes, but
they serve as background material for the type of study attempted here.
Dr. Garrigus offers an interesting justification for his study, as
---it may be said that the results of this
investigation should prove instructive in
the .study of the association between land
use and land surface conditions as that
association might be taken up in connection
with physical geography, economic geography,
Table 2 Kentucky Sinkhole Enumeration
Number of Sinkholes
Largest in area is on Brownsville Sheet, 3,114
acres. Author estimates that the mapped areas
of the Mississippian Plateau in Kentucky con-
tains between 60,000 and 70,000 sinkholes of
varying size and description.
Source: "Land of Ten Thousand Sinks" W. R. Jillson
Kentucky Academy of Science (1924).
agronomy, or any number of other subjects.
The results may be of general interest even
though not used for any specific purpose.
More practical gains might be expected to come from a much more
Intensive study of land use in relation to land forms, including not
only a detailed examination of the areal association between the two,
but including also an attempt to discover so far as possible the causal
relationships: Just how, precisely, does the form of the land surface
act to influence the various kinds of land use?
Most helpful is an assemblage of research directly based on the
limestone landscape. A pioneer in this respect is Carl Ortwin Sauer
with his Geography of the Pennyroyal (1927). This book is a study of
the influence of geology and physiography upon the industry, commerce,
and life of the people of a Kentucky area. He uses the limestone land-
scape as a basis for regional unity and develops this idea through
description of historical land uses.
Actually more geological than geographical in scope is a recent-
work by Placido La Valle (1967).22 His purpose was to analyze spatial
patterns of karst features, but his emphasis was on elongation and
orientation of sinkholes rather than on uses of the land. A more
classic geographical approach is found in "The Central Kentucky Karst"
by White, Watson, Pohl, and Brucker (1970).23 This accurate and des-
criptive work features the influence of climate on karst processes and
the rates of denudation of karst plains in Kentucky.
Doubtless the most significant and most pertinent recent research
in a similar vein, yet wider scope, is a dissertation by James M.
Goodman (1962) describing yet again the Pennyroyal area and analyzing
the influence of its landscape on agricultural land use and productivity.
This is a case study approach dealing with an area of outcropping lime-
stone in the Kentucky karst. Goodman has defined ten different terrane
types from various combinations of soil, drainage, slope and lithology.
These types form a framework for examination of land use and agricultural
productivity. Patterns and variations for each terrane type were
determined and compared with crop yields and land values.
It is evident that a study of this type has not yet been made of
Florida karst. The range of physiographic terrane types is not as large
as that of other karst areas because of the relatively low relief, and
also the soils are more homogeneous as well. On the other hand, the
variety of expression is surprising and the sinkholes vary widely in
size, shape, and depth.
NOTES TO CHAPTER I
IN. M. Fenneman, Physiography of the Eastern United States
(New York: McGraw-Hill Book Co., 1938), p. 46.
Gordon E. Bigelow, "History of Man in the Oklawaha Valley,"
Environmental Impact of the Cross-Florida Barge Canal (Gainesville:
Florida Defenders of the Environment, 1971), p. 47.
3Florida Handbook (Tallahassee: The Peninsular Publishing Com-
pany, 1971), p. 293.
Charlton W. Tebeau, A History of Florida (Coral Gables: Uni-
versity of Miami Press, 1971), p. 53.
51bid., p. 52.
Ibid., p. 125.
Florida Agricultural Experiment Station, Principal Soil Areas
of Florida, a Supplement to the General Soil Map (University of Florida,
1967), p. 15.
Ibid., p. 4.
E. T. Pyle, Home Country (New York: William Sloan Associates,
Inc., 1947), pp. 340-341.
J. B. Lamarck, Hydrogeology (Urbana: University of Illinois
Press, 1964), first published at the author's expense in 1802.
E. M. Sanders, "The Cycle of Erosion in a Karst Region (After
Cvijic)," The Geographical Review (1921).
12H Lehmann, "Introduction," Second Rapport of the Commission on
Karst Phenomena, XIX International Geographical Congress, Norden (1960),
3Watson Monroe, A Glossary of Karst Terminology, United States
Geological Survey Water Supply Paper 1899-K (1970).
1L. R. Jillson, "Land of Ten Thousand Sinks," Sketches in
Geology (Louisville: C. T. Dearing Printing Co., 1928, Reprinted from
Transactions, Kentucky Academy of Science, Vol. I, 1924),pp. 143-144.
15John A. Shimer, "Sink-holes," This Sculptured Earth: The
Landscape of America (New York: Columbia University Press, 1959),
C. A. Cotton, "Limestone Landscapes," Landscape (Cambridge:
University Press, 1941).
17Allyn Swinnerton, Hydrology of Limestone Terranes, Physics
of the Earth IX, Meinzer, ed. (New York: McGraw-Hill Book Co., 1942).
V. T. Stringfield and H. E. Le Grand, Artesian Water in Tertiary
Limestone in the Southeastern United States, United States Geological
Survey Professional Paper 517 (1966).
W. D. Jones, "A Method of Determining the Degree of Coinci-
dence in Distribution of Agricultural Uses of Land with Slope-Soil-
Drainage Complexes," Transactions, Illinois Academy of Science, Vol. 22
W. M. Garrigus, "Roughness of Terrain as a Factor in the Areal
Variation of Agricultural Productivity in Ohio," paper presented at the
Annual Meeting of the Association of American Geographers, Columbus,
Carl Ortwin Sauer, Geography of the Pennyroyal (Frankfort:
The Kentucky Geological Survey, 1927).
22Placido La Valle, "Some Aspects of Linear Karst Depression
Development in South Central Kentucky," Annals of the Association of
American Geographers, Vol. 57, No. 1 (1967).
William B. White, Richard A. Watson, E. R. Pohl, and Roger
Brucker, "The Central Kentucky Karst," The Geographical Review, Vol.
LX, No. 1 (1970).
24James A. Goodman, Physiography and Agricultural Land Use in
a Portion of the Pennyroyal Region in Warren and Edmonson Counties,
Kentucky (Northwestern University Dissertation, 1963).
PHYSIOGRAPHIC BACKGROUND FOR THE STUDY
Karst Areas of the World
Carbonate rocks constitute about five to ten percent of world-
wide sedimentary rocks; and outcrops of limestone measure several
million square miles of the earth's surface. The distinctive lime-
stone terrane can be found on every continent, but quite often it varies
extraordinarily from place to place. Rock solubility is the primary
element in the formation of karst landscapes. Certainly the carbonate
rock,or other soluble material, must be present for karstification to
occur. Other important considerations include rock structure and
World karst areas have been studied by workers and comparisons
and contrasts recorded. Some of the areas most often noted for re-
search projects outside the United States are listed in Table 3.
Karst is-considered to be the surface expression of solution
phenomena and is a descriptive geomorphological term. Generally it is
"sinkhole topography" named after the Yugoslavian word "kras" meaning
"stone."2 The Italian place-name "Carso' and the French word "Causses"
are sometimes regarded in the generic sense as equivalent to karst.
The classic region for the observation of karst is the plateau of the
western Balkans, the study area for Jovan Cvijic who established a
substantive vocabulary for his description of a karstt cycle."3
Table 3 Karst Areas of the World (outside of the United States)
Middle America and the Caribbean
Northern Yucatan and Tabasco (Mexico)
The West Indies
The Pennines of Yorkshire and Derbyshire (England)
The Mendyes and South Wales
County Clare in Western Ireland
The Grand and Petits Causses (Southern France)
The Dinaric Alps (Yugoslavia)
Southeast Asia and Oceania
The South Coast of Central and East Java
Central New Guinea
North Canterbury, New Zealand
Nearly every country has its own individualized set of karst terms
coined to fit its particular landscape features. In the presence of
carbonate rocks, a variety of geologic and hydrologic features account
for the almost infinite variation of karst expression. It should be
remembered that there are also innumerable areas where karst features
are present as surface landforms, but do not dominate the landscape.
These areas are not included in this review.
It has been noted that world karst areas vary with climate and
rock composition and structure. There may be well-drained uplands
completely bare of soil, or jungle-covered lowlands mantled with thick
organic muck. The rocks may be flat-lying to steeply dipping. The
presence of caves may or may not constitute a karst terrane, for it
is the surface expression that is being described.
Karst Areas in the United States
Karst terranes develop most efficiently when certain factors in
addition to the presence of carbonate rocks prevail. The soluble
rocks should be dense, highly jointed, and preferably thin-bedded.
Permeability in the sense of mass permeability may be unfavorable be-
cause that condition could allow water to pass through the entire rock
mass rather than to be channeled along restricted pathways. This con-
centration of flow is necessary for the karstification to take place.
Carbonate terranes occur in each of the four major groundwater
regions of the United States as described by Meinzer (1949).7 Some
of the more prominent of each region are described briefly by String-
field and Le Grand in Hydrology of Carbonate Rock Terranes A Review.
The Karstlands Map in the National Atlas of the United States further
delimits actual karst terranes. The distribution is generally sum-
marized as in Table 4. Alaska and Hawaii have only pseudo-karst areas.
Puerto Rico is karstland along the northern coast, west of San Juan.
Karst of Florida
This brief review of world and United States karst serves as back-
ground for this study of Florida karst. The physical region of central
Florida takes on more importance when its size is compared with that
of other karst areas, for it has the largest area of contiguous exposed
karst surface in the United States.
Florida is composed of sedimentary layers resting on the contin-
ental platform far below the surface. Deposition was carried on through
the Cenozoic under shallow seas and then continued intermittently through
the Pleistocene Epoch. The peninsula underwent some modification in atti-
tude because of disturbances in the basement rocks during a Post-Oligocene
Orogeny. The deformation resulted in uplift of an area of west central
Florida and is termed the "Ocala Arch." (See Map 5.) Once-horizontal
beds of limestone, clay and sand were domed upward from below as erosion
took place to lower the surface from above. The resulting geomorphology
of the study area is that of an eroded dome or anticline, with very
gentle dips, exposing older rocks at the center and younger strata
along the periphery. Thus, rocks of Eocene age are widespread on
the surface at the center of the Ocala Arch, whereas the Oligocene is
not represented until Live Oak, Florida,is reached to the north, and
Brooksville in the south. During the several transgressions and regressions
Table 4 Karst Areas of the United States
Atlantic and Gulf Coast Plain
West-central and central Florida (Ocala Uplift)
South-central and east Georgia
East-central South Carolina
Southeastern North Carolina
Appalachian Mountains from New England to northern
Alabama (The Great Valley)
Central Tennessee-Kentucky-southern Indiana
Southern Missouri-northern Arkansas (Ozarks)
Southeastern New Mexico
Northern Arizona (Colorado Plateau)
THE OCALA ARCH
PLEISTOCENE 7 ]
Source: Florida Geological Survey
of the seas, varying portions of the topographic "highs" were left sub-
aerial. This phenomenon complicates the geology, causing unconformi-
ties as well as disconformities in the structure of the rocks. Re-
gional terrace deposits or ancient shorelines further influence the
The study area is centered on the Ocala Uplift; specifically it is
the area of outcropping or near-surface, Eocene rocks. The most typical
manifestation of this landscape is the limestone plain where the Eocene
rocks are only lightly covered with permeable sands. (See Figure 2.)
Ranging north-to-south through the limestone plain is the area called
the Brooksville Ridge. This ridge is composed of younger deposits of
sand and clay overlying the Eocene rocks. Some areas are hilly, probably
a result of differential subsidence of underlying materials which had
become voided by solution. Other parts of the ridge seem to be sands
and dunes remaining from Pleistocene terrace deposits.
Tn Gilchrist and Levy counties and practically "within" the Brooks-
ville Ridge is the area termed the Waccasassa Flats, partly underlain
by a graben developed during the movement of the rocks that formed the
Ocala Uplift at the end of Oligocene time. The land features are not
flat. but are sandy hills underlain by Miocene and Pleistocene clays
that retard percolation of water and result in perched lakes and a
pocosin type of swamp drainage.
These "higher" areas just described are essentially nonkarst, their
surface expression not particularly resulting from solution. Their
appearance is distinctly different from that of the karstified lime-
k .. .
~~p~q~lT L ~r-4
stone plains areas to the east and west of the ridge-hill-dune complex.
The transition zones are sometimes gradual and both topographic types
are seen on a given traverse. More often, however, the borders are
abrupt, showing as distinct boundaries, especially on aerial photographs.
On the Gulf coast and in almost all of Dixie County the water
table is quite high, presenting a "drowned karst" aspect. This coastal
area is not presently considered karst because of the masking of the
features by water and shallow surface deposits.
The areas remaining for study are the limestone plains' This area
is, in the opinion of the writer, the most distinctive karst in Florida.
There is another Eocene outcrop in the Panhandle (or west Florida) near
Marianna, but this area is not as extensive as the peninsular outcrop.
The Eocene Limestone Plains
The limestone plains have been traditionally considered as agri-
cultural areas. The outcrop is divided into two parts by the overlying
Waccasassa Flats and the Brooksville Ridge, called Bell Ridge in its
most northern extent. Vernon has at different times called portions of
the plains by place-names (Chiefland and Williston), but for this research
they will be termed The Bell-Trenton-Chiefland Limestone Plain and The
Newberry-Williston Limestone Plain. The two areas are quite similar,
especially in their origin. There are, however, some differences in the
character of the sinkhole topography which are often difficult to inter-
The most characteristic karst feature here is the doline, a sink /
or sinkhole with sloping sides found almost continuously in the lime-
stone plains. The landscape is characteristically rolling, with the
shallow basins of interior drainage separating low rocky interfluves.
The area is not unlike parts of the Pennyroyal of Kentucky in the "lay
of the land." Low areas of subdued relief could sometimes fit the
definition of polje, or karst plain. Their origin could possibly be
tectonic, downfaulted or downfolded. In any case, these expressions
are further modified by solution, with lowering by sub-surface solution
a factor of considerable importance. Florida poljes lack the steep
walls found in some European karst. Generally, the term which might
best be applied to Florida karst landscape is "solution-subsidence,"
descriptive of the formation of the karst plains. The shallow basins
are often coalesced, forming compound sinks.
Water percolates downward through the sides and bottoms of the
sinks, sometimes flowing rapidly into one of several smaller sinkholes
in the floor of the doline. Hardly any water stands or flows on the
surface except in scattered areas of clay deposits which are probably
residual or reworked Miocene formations. Therefore, there are few
sinkhole ponds or karst lakes perched on the limestone plains.
The land has been used mostly for agriculture and some subse-
quently planted pines, but clumps of hardwood trees remain that are
more often than not in deep-sinkhole locations. Farmers have placed
boulders from the fields into the sinkholes for so many years that the
topographic expression is now virtually an outcrop of limerock rather
than a void. These same cherty liierock boulders are sometimes found
placed by the farmer along the fence rows.
Scattered throughout the karst plain are vertical-walled sinkholes
called cenotes. (See Figure 3.) Origin of the cenote is from the
collapse of overlying rocks into a cavity rather than from a lowering
of the surface slowly downward by "sagging" or by solution beneath a
soil cover. Whereas dolines may be formed either of two ways, the
steep-walled cenote is evidently catastrophic in its formation.
Cenotes of large size are readily found in the eastern range of
the limestone plain, especially near Newberry. They vary from five or
six feet across to as much as fifty feet. They may be connected to
one another by underground passages which may be caves or contain water.
An interesting karst feature most prevalent in the western plains,
near Chiefland, is the vertical-walled ponor or natural well. (See
Figuye*4.) These are often perfectly round and are of varying depths,
containing water in some cases. They are sometimes closed from the top,
appearing to have formed from deeper zones upward. This mysterious
closed-top well can be seen when the observer is below ground level,
as in an excavation such as an old quarry or highway borrow-pit. The
smoothed and rounded wells are seen also as cave chimneys leading up-
ward from cavities, sometimes opening at the surface, sometimes seeming
to taper inward to small size, then closing. Most ponors observed are
obviously open at the top or they would not be noted in a surface re-
connaissance. It is this type of sinkhole, as well as the cenote, that
often must be fenced to lessen its hazard to men, animals, and equipment.
(See Figure 5.)
Origin and location of the tubes and pipes are most often joint-
controlled; and they may run as galleries along bedding planes. The
t. -- .-.( -.
*'~- *b- ?-. (2' *.~ -*
Figure 3 -- Cenote About Twenty Feet in Diameter (Jerome Sink North of
-. --C. 4
*'L -I.- .. I
,,i ADW r
r r %*.
"~..*~5 I ~ ..*
4. r' -
r -)r r .' -
circular regularity of some wells may be a result of swirling of vadose
water downward, or phreatic water upward, expanding under artesian con-
editions which have existed in the past. The roundness is unusual in
that there is a smoothed inner wall, with truncated fossil shells,
evidencing a process of corrosion, rather than solution alone.
These vertical wells and also numerous "inclined" or horizontal
pipes or tunnels give a honey-combed aspect to a vertical cross-section.
(See Figure 6.) It is easily observed that the cavities fill with non-
indurated surface materials, sand and clay, which tend to move downward
through gravity and movement of water. When this displacement occurs,
often after a heavy rain, new holes of all sizes open up. The consensus
is that the most "favorable" time is during or immediately following
a prolonged heavy rain breaking a long drought.
A karst feature of considerable importance is the ubiquitous out- "
crop of limerock. (See Figure 7.) The karstification of the limestone
plains has left uncovered areas of rock. Often there are boulders pre-
sent which have been hardened by concretions of silica precipitated
into cavities. The unusually hard durable rocks are sometimes piled
up to remove them from cultivated fields. They fill sink hole sites
or line fences where they are available for use by a crusher operator
for aggregate. Limestone only sparsely covered with soil and grass can
be seen in the ditches beside the highways, especially Highway 41 be-
tween High Springs and Williston.
The rather poor, excessively drained soils of the limestone plain
are often several feet thick on the irregular surface of the underlying
.^*p3-. A-:- .l.-.f .l
-,'.I 1^^*^ 7'
L. a if
; s ~~~
rocks, yet outcrops are frequent. An unusual occurrence in the study
area is a deposit of "bog iron" ore outcropping in northern Levy County.
The ore is a high-grade limonite possibly deposited in the limerock by
rising artesian waters carrying iron dissolved from Lower Eocene strata.3
The ore is reddish-brown in color and stains the fields and sandy roads
for several miles. One quarter of a quarter-section is so rocky that
nearby residents refer to it as "The Iron Forty." (See Figure 8.)
The karst of Florida as a whole is not typified by the Eocene out-
crop area. There is in fact an almost infinite variety of expression.
-Areas beyond the Eocene outcrop are influenced geomorphologically by
the Eocene levels. These underlie younger formations flanking the
Ocala Uplift. The beds serve as the principal aquifer of Florida.
Solution cavities in this rock sometimes result in the differential
subsidence of overlying formations to create hills or lake basins.
There is said to be a rough ratio between the thickness of the over-
burden and the size of a sinkhole collapsed or sagged to fill a void
below. The sinkholes at Gainesville or to the east in the Hawthorne
deposits would therefore be large and spaced well apart. Those in the
limestone plains with almost no overburden are often quite small and
closely spaced. Soil or cover must be considered too because case
studies show that even though holes may occur under sandy soil, they
fill rapidly again and are usually not reported as significant.
Most of the Florida lakes and prairies could be called karst
features because they owe their origin to a lowering of the surface
by the solution of underlying rocks. The shallow prairie lakes sur-
Y KL~~.-L ~Cr~l? ~ rS1
1..1.4 -S~C ( \,
rounding the study area are lined with fine grained sediments making
the limestone impermeable. When hydrostatic pressure builds up,these
prairie lakes "break through" and drain into an adjoining sinkhole.
The preceding sections have involved a formal description of karst
as a complex of landforms. Next followed the identification of the
more important karst areas of the world and of the United States,
ending with an expanded outline of karst in all of Florida. With the
foregoing to serve as a frame of reference, the focus of the present
research is reached. Its raison d'etre is the description and explan-
ation of the Eocene karst features as they affect land use in this
selected region of Florida.
NOTES TO CHAPTER II
IAllyn C. Swinnerton, Hydrology of Limestone Terranes, Physics
of the Earth IX, Meinzer, ed. (New York: McGraw-Hill Book Co., Inc.,
1942), p. 656.
C. A. Cotton, "Limestone Landscapes," Landscape (Cambridge:
University Press, 1941), p. 281.
V. T. Stringfield and H. E. Le Grand, Hydrogeology of Limestone
Terranes in the Coastal Plain of the Southeastern United States,
Geological Society of America Special Paper No. 93 (1966), p. 351.
William D. Thornbury, Principles of Geomorphology (New York:
Wiley and Sons, Inc., 1969), p. 304.
0. E. Meinzer, Ground Water, Physics of the Earth IX, Meinzer,
ed. (New York: McGraw-Hill Book Co., Inc., 1949), p. 162.
V. T. Stringfield and H. E. Le Grand, Hydrology of Carbonate
Rock Terranes--A Review, Journal of Hydrology, Vol. 8, No. 3 (1969),
c/ United States Department of the Interior, Geological Survey,
National Atlas of the United States of America (Washington, D. C.,
1970), p. 77.
William J. Yon, Jr., and Harbans S. Puri, "Geology of the
Waccasassa Flats, Gilchrist County, Florida," Bulletin of the Ameri-
can Association of Petroleum Geologists, Vol. 46, No. 5 (1962), p. 674.
Robert O. Vernon, Geology of Citrus and Levy Counties, Florida,
Florida State Geological Survey Bulletin No. 33 (1951), p. 41.
Harbans S. Puri, J. William Yon, Jr., and R. Oglesby Woodson,
Geology of Dixie and Gilchrist Counties, Florida, Florida State Geo-
logical Survey Bulletin No. 49 (1967), p. 33.
Vernon, Geology of Citrus and Levy Counties, p. 286.
METHODS, TECHNIQUES, AND TOPICS EMPLOYED
IN KARST AREA STUDIES
The focus of the study, as indicated before, is upon the limestone
plains, termed for geographic reference, Bell-Trenton-Chiefland and
In the literature reviewed in the background chapter, the scientists
have used several methods and numerous techniques for research. The
formal statistical method is best demonstrated by Placido La Valle in
his study of Kentucky sinkholes.l In an area of 800 square miles in
South Central Kentucky, using USGS quadrangle maps, La Valle carried
out detailed studies on a random sample of 25 percent chosen from an
arbitrarily placed grid. His measurements constituted the basis for
a morphometric analysis of linear karst development. This examination
revealed two facets reflecting linear trends: the depression elongation
ratio and the relative orientation of the depression axes. His results
show that major axes parallel joint strikes and that spatial distribu-
tions correlate with structural weaknesses such as numerous joints and
thin bedding planes. Studies have been made which confirm that this
condition also exists in the Florida karst, but the results are not
based on formal statistical methodology.
The experimental method is one in which data are secured under con-
ditions in which some forces are held fairly constant while other forces
are measured. In the literature surveyed, the best example of this re-
search method is described by Clifford A. Kaye in "The Effect of Solvent
Motion on Limestone Solution" (1957). He performed actual experiments
with blocks of limestone, water, and acid solutions. His demonstration
proved that agitation and motion of solvent flow are positive factors
of importance in limestone solution patterns. From the principle that
the greater the flow velocity, the greater the rate of solution, he
suggests that differential solution in proportion to velocity takes
place and that this process causes preferential conduit enlargements of
cavities by phreatic waters.
Few actual experiments exist as to degrees and patterns of solu-
bility of Florida limestone. Engineering studies have been made con-
cerning relative strengths of wet and dry limerock, but controlled
experiments are only infrequently described in the literature. C. V.
Dolliver conducted several physical projects in the laboratory for her
thesis "The Geomorphology of the West-Central Florida Peninsula."3
These are not exclusively performed on limestone rocks, however, but
involve deposition and arrangement of materials by moving water in the
Gulf Limestone Plain area. H. K. Brooks has performed a large number
of laboratory experiments to analyze the chemical composition of sur-
face and ground waters throughout the karst area of northern peninsular
Florida. His quantitative results are interesting in that they show
the overall rate of solution in the karst terrane as about 1.5 inches
per 1,000 years. These figures were based upon field samples of water
collected and analyzed in 1967 and show that the lowering of the re-
gional landscape takes place at a rate only a little less rapid than
that calculated by Sellards in 1908.5
The general research method of analogy has been used rather broadly
in research on limestone terranes. The efforts of scholars to make
comparisons and to attribute certain characteristics of one karst area
to those of another are frequent. They are not applied without reser-
vation, however, because the karst landscape varies so extremely with
physical exigencies. The effort to synthesize a karst vocabulary is
one way to emphasize the analogies that do exist. Florida karst is
rarely compared to that of other areas. The gentleness of dip, the
high water table, and the subtropical climate combine with the structural
characteristics, dense and highly jointed, to present a unique landscape.
Tropical karst has been researched by Sweeting,6 Meyerhoff,7 Monroe,8
and others, but only Corbel has actually made comparisons of another
landscape and the Florida karst. His description of features in Mexico
mentions similarities and contrasts between Yucatecan and Floridian
The present study will employ two methods other than the historical
method already being utilized. These are the informal statistical
method and the case study method. The informal statistical method is
offered as an alternative to the formal when the available data are
discontinuous or otherwise unsuitable for formal quantitative handling.
The writer considers that the counting of sinkholes using various tech-
niques is a problem-oriented approach, but the statements obtained from
analysis of the Conservation Needs Inventory data are somewhat more
formalistic in presentation.
The case study method has been used most often in the literature
specifically about karst terranes. As cited before, Sauer and Goodwin
both used the Pennyroyal of Kentucky as their case in point. In addition,
Dicken and Brown have been especially erudite in their presentation of
a case study describing erosion problems of Kentucky "limestone soils."10
Central Tennessee is another case area covered in Limestone Hydrology in
the Upper Stones River Basin by Moore, Burchett, and Bingham.11
Florida karst has been specifically investigated by several students of
the area in case study approaches. Abbott (1971) in "Twenty Springs of
the Oklawaha" describes springs as karst features of a particular area.2
Lipchinsky in "A Study of the Origin of Limestone Caverns in Florida"
offers detailed descriptions and explanations of features found in fifteen
Florida caves.13 0. Girard, Jr.,14 and J. D. Vormelker15 describe the
geology of the High Springs Quadrangle and that of the Gainesville West
Quadrangle, respectively. These two studies describe lands which are fairly
representative of local karst terranes, as is the underground Santa Fe
River chosen for research by R. T. Skirven.16 All these projects skirt
the present study area of the limestone plains, except one conducted by
H. K. Brooks, before cited. His topic circumscribes the entire area
in a systematic approach which is geological in content. The geographical
purview, thus, has been left for the present study.
The methods reviewed include the five classic ways of presenting
research: the formal statistical method, the informal statistical
method, the experimental method, the analogy and the case study. Many
techniques or skills have been employed in the use of the research
methods reviewed here. The field techniques are generally reconnais-
sance, measurement, gathering of samples, photography, and map and photo
interpretation. The interview has been little used or is unreported,
very likely because it is often considered vague or unscientific in
approach. All the above mentioned techniques are used in the present
research, except for the collection and analysis of samples in the
laboratory. Especial attention is called to photography, a tool which
can be used to help greatly in the description of the limestone terrane.
(See Figure 8.) Almost all available research into Florida topics in-
cludes air and ground photographs.
The technique of remote sensing of the environment is comparatively
new for limestone research, dating from 1940, or about thirty years.
Earliest photos used for this paper are the Agricultural Stabilization
and Conservation Service contact prints in stereo coverage of the two
counties of the western portion of the limestone plains, Gilchrist and
Levy Counties. The sample area, only part of the plain, was chosen for
study as an example of an agricultural area, defined as such on
Marschner's Land Use Map of Florida. Stereo pairs were examined care-
fully for evidence of existence of sinkholes, quarries, and springs.
Outcrop could not be distinguished. The forested and nonforested areas
are fairly easily discerned on the photos. The result of painstaking
marking of sinkholes large enough to show up on the pictures is that
the pattern on the land as well as the density of such sinkholes can
be calculated. (See Map 6.)
Extremely sophisticated techniques have been reported recently in
determining areas which may be "sinkhole-prone." Thermal infra-red
imagery and interpretation have led to possibly extravagant claims that
collapse-prone areas can be predicted or detected by analysis of data.
Coker, Marshall and Thompson worked with multispectral imagery to show
that some oval patterns on the recognition maps delineated subsidence
areas near Bartow, Florida.17 J. W. Stewart came somewhat closer geo-
graphically to the Eocene limestone plains in his remote sensing report
on lakes in west central Florida. He used imagery-producing remote
sensors (1) to determine interrelationships between lakes of dif-
ferent characteristics; (2) to develop a classification of the lake
shore features; and (3) to establish a classification of lakes in a
karst topography. Most of these goals are also sought after here in the
present research effort, but through use of conventional photography and
techniques of interpretation.
Other reports on remote sensing experiments have not proved that
use of thermal infra-red imagery is much more effective than use of
stereo pairs. Some features, especially rock outcrops and wet areas,
may be detected more accurately by the use of the expensive and sophis-
ticated techniques, but sinkholes in general are discernible to about
the same degree with either type of imagery. The hydrobiological in-
BELL TRENTON CHIEFLAND
* ; .* *.
.. :. .*: ..
. ** *! ** . *
.. . *,
: . *
55 ***:V **
S* 55 S
.* S * *
550 0 g O
-. 0 *. 5* ***5
0* *es. I
S ~ ~ g g
vestigations of lakes, the Everglades, selected coastal areas, the
Oklawaha Valley, and muck deposits now being conducted will prove the
value of modern techniques for analyses of karst-influenced topography.
The rapid drainage prevailing in the limestone plains themselves is
probably the factor rendering aerial photographs almost as effective
as multispectral reflections.
The methods and techniques used for research serve to facilitate
the accomplishment of certain tasks and the attainment of selected goals.
Another matter of choice is the form of presentation of the results of
the work. The topical approach is used here because of the geographical
nature of the research. The emphasis, after the background setting, is
upon man and his use of the land.
The limestone plains are traditionally agricultural in use. Crops
have been grown and livestock tended on this land through the years.
"The Effect of Karst Topography on Agricultural Land Use" will be the
topic reviewed initially. Also included is the contrasting position-that
of examining the effect that man's use has had on the condition of the
land. This first topic is treated in Chapter IV, the theme of the present
research project, because of the fact that agriculture is the dominant
land use in the area.
The second topic is "Hazards Associated with Karst." Three
divisions of Chapter V seem logical in the examination of assembled
data: the place of sinkhole hazard in contemporary geographical re-
search regarding hazardousness of location or place; the history and
status of sinkhole insurance; and the Department of Transportation
problems with highway and right-of-way subsidence.
The third subject to be presented is a view of "Limestone as a
Resource," Chapter Six. Historically, the presence of minerals has
helped in some ways to settle the area and to establish the transpor-
tation routes, chiefly roads and railroads. Limerock, dolomite and
phosphate are the three outstanding minerals in the case of the Eocene
The last topic to be discussed, as Chapter Seven, concerns "Land
Values and Karst." Booming prosperity and increasing demands for land
in Florida put new values upon some almost abandoned tracts of rural
land. The advantage gained by the presence of karst features, lakes
in this instance, is explored in an analysis of the enhancement of land
values when these features are present and able to be exploited.
These four topics offer a geographic presentation in the study of
karst topography not before attempted. The topical approach emphasizes
the use of the land by man and his works.
NOTES TO CHAPTER III
Placido La Valle, "Some Aspects of Linear Karst Depression
Development in South Central Kentucky, Annals of the Association of
American Geographers, Vol. 57, No. 1 (1967), p. 59.
Clifford A. Kaye, "The Effect of Solvent Motion on Limestone
Solution," Journal of Geology, Vol. 65, No. 1 (January, 1957), p. 35.
Claire Vincent Dolliver, "The Geomorphology of the West-
Central Florida Peninsula" (University of Florida Thesis, 1965),
H. K. Brooks, Rate of Solution of Limestone in the Karst
Terranes of Florida (Florida Water Resources Research Center, Publi-
cation No. 6, 1967), p. 1.
E. H. Sellards, Underground Water Supply of Central Florida,
Florida State Geological Survey (1908), p. 48.
M. M. Sweeting, Hydrogeological Observations in Parts of the
White Limestone Areas in Jamaica (Kingston: Geological Survey, 1955).
Howard A. Meyerhoff, "The Texture of Karst Topography in
Cuba and Puerto Rico," Journal of Geomorphology 1-2 (1938-9).
8Watson H. Monroe, "The Karst Features of Northern Puerto
Rico," National Speleological Society Bulletin, Vol. 30, No. 3
J. Corbel, Notes sur les Karsts Tropicaux. Revue de Geographie
de Lyon XXX, 1 Lyon (1955). Cited by H. Lehmann in Report of the
International Geographical Union Commission on Karst Phenomena (1956),
S. N. Dicken and H. B. Broun, Jr., "Soil Erosion in the Karst
Lands of Kentucky," United States Department of Agriculture Circular
No. 490 (1938).
Gerald K. Moore, Charles R. Burchett, Roy H. Bingham,
Limestone Hydrology in the Upper Stones River Basin, Central Tennessee,
State of Tennessee Department of Conservation, Division of Water Re-
sources (1969),pp. 8-9.
Elizabeth F.Abbott, "Twenty Springs of the Oklawaha"
(Gainesville: Florida Defenders ofthe Environment, 1971).
Zelek Lawrence Lipchinsky, "A Study of the Origin of Lime-
stone Caverns in Florida" (University of Florida Thesis, 1963).
1Oswald W. Girard, Jr., "The Geology of the Gainesville West
Quadrangle, Alachua County, Florida" (University of Florida Thesis,
Joel David Vormelker, "Geology of the High Springs Quadrangle,
Florida" (University of Florida Thesis, 1966).
Raymond Taylor Skirven, "The Underground Course of the Santa
Fe River Near High Springs, Florida" (University of Florida Thesis,
A. E. Coker, R. Marshall, and N. S. Thonpson,"Application of
Computer Processed Multispectral Data to the Discrimination of Land
Collapse (Sinkhole) Prone Areas in Florida'" (1969).
Joseph W. Stewart, "Synoptic Remote-Sensing Survey of Lakes
in West-Central Florida" (1969).
THE EFFECT OF KARST TOPOGRAPHY
ON AGRICULTURAL LAND USE
Patterns of agriculture are necessarily determined by both physical
and cultural influences. One or the other may be dominant, but both must
be considered in a geographic appraisal. They physical landscape provides
the setting for the complex interactions of man's influences through time.
Therefore, changes through historical time are important in explaining
present patterns of land use.
The study area is fairly homogeneous climatically because of the
rather small size of the region. There are differences in types of land-
forms, however, as described in Chapter II. These geomorphological vari-
ations have pre-determined the soil types which, therefore, are mostly
azonal (influenced principally by factors other than climate). Parent
material of limestone, sand, silt, and clay underlie the soils and, along
with drainage conditions, are responsible for their characteristics.
The distribution of natural vegetation as it existed almost con-
tinuously over the land also was not strongly influenced by climate
(see Map 7). Patterns of vegetation responded to soils, which are
principally related to geology. Vegetative cover or physical appearance
of the landscape helped early settlers make decisions about how to use
the land initially.
Settlement of the study area took place slowly and the people lived
on the land, farming only the better soil areas. Early farming was mainly
of a subsistence type, for markets were remote. Corn, rice, sugar cane,
and sweet potatoes were grown, and stock raising was also important.
The growing of sea-island cotton became popular, especially following The
War years (1861-1865) and much hammock land was cleared for cotton.
Alachua and Marion Counties were settled first, and substantial
planters were included among the first Americans in that area. Alachua
County was settled in the 1820's and 1830's (in the eastern and southern
sections of the present county); Marion County was settled in the 1830's
and early 1840's, despite the fighting during the Second Seminole War
which took place near Ocala. In both these counties, agricultural
staples were produced early, along with the plantation economy based on
slave labor, but this settlement mainly involved richer hammock lands
east of the Eocene outcrop area.
In the sparsely settled county to the west in central Florida,
lonely farmsteads of the successful land claimants grew patches of corn
and beans, cane and tobacco among stumps of cleared and burned land.
The first town mentioned in the area by Cash is Archer, settled in 1859,
coincident with the building of the Florida railroad through the section.2
Practically all the countryside was open range for livestock.
Cattle and hogs were fairly well fed on pasture grasses in the prairie
lands and oak mast in the hammocks. Stock raising was the total occu-
pation for some of the population because prairie land unsuited for crop-
land often made good pasturage.
land often made good pasturage.
In the southern parts of the study area, citrus became important,
beginning in 1859 with a single orange grove at Micanopy. Soil and
climate were suitable for oranges and the industry expanded, increasing
land values appreciably. With the discovery of phosphate, the area
boomed even more, and land speculation followed. Towns were laid out
and many lots surveyed, but not all were settled, as is typical of
American land booms. There was a general depression following the
collapse of the boom. Prosperity had not yet completely returned when
the great freeze of 1894-1895 killed all the orange trees in the area.
Prior to the freeze, almost every farm in the southern area had its
own orange grove. With citrus and stock raising as primary ventures,
not much land was farmed except for subsistence crops. Turpentining
became profitable and the industry grew rather rapidly, sometimes com-
peting with logging operations for the pine trees.
The start of commercial forestry first provided a source of supple-
mental income for subsistence farmers but was conducted on rather a
small scale in the karst areas because of the lack of waterways. In
northern Florida only the timber tracts convenient to the ports of
Pensacola, St. Marks, or Jacksonville were extensively logged off before
1900.6 So the years of early settlement did not drastically change the
appearance of the forests of the study area.
Truck farming was greatly extended after the freeze and was
fostered by the extension of the railroads into the region. Cantaloupes
and watermelons, as well as peppers and some eggplants, were the prin-
cipal vegetable crops then as now.7 In the areas of hard rock phosphate
mining, on the other hand, hardly any attention was given to agriculture,
although much of the forest was being cut over to provide fuel for drying
operations. By the second decade of this century, the growing of tuck
crops and corn was following some clearly recognized practices. Vegetables,
such as tomatoes, snapbeans, and cabbages)were planted very early in the
growing season and fertilized heavily. Then a corn crop was put in the
same field to take advantage of the remaining fertilizer in the soil.
This provided a rotation program of sorts with good yields.
Although a little tobacco had been grown and sometimes sold by the
pioneer settlers, six acres of bright-leaf tobacco in the Suwannee River
Valley,in 1921, marked the first commercial venture in the area. In
1923 some 200 acres were harvested, giving tobacco growing importance as
a money maker.10
The most active lumbering took place in the late 1920's and early
1930's. Whereas before only small local sawmills had used logs for
local construction, now large companies ravaged the land, removing chiefly
the longleaf pine. Small gauge railroads and primitive logging roads
criss-crossed many areas, along with a few farm roads utilized by farmers
who came in and began clearing the brushy cut-over lands for cropland.
The choice of specific uses for land in farms was made by the grower,
mostly in response to soil type, indicated to him by the native vegetation
(see Map 7). It is apparent that there is a definite relationship between
the patterns of soils and the details of the geology in the area. The
main features of the geology have been summarized in Chapter II, and it
is shown that limestone now exposed at the surface once was covered with
Sand Pine'' "
SPine F.at o -'
[ F reshwater Marsh
I 1 000000
' i 1 ___
SOUnCE: Darvs. General map of Naaural Vegeteton ol Florida. 1967
M a p 7
younger strata of clay and sand of which remnants, remain on the uplands.
Over all is the mantle of residual material which is the matrix for the
formation of the soils. The native vegetation type is frequently utilized
to distinguish soil areas, even after the removal of the virgin growth.
Residents of the study area commonly use the terms "pine land," "hammock
land," or "prairie land," as names for subregions. This nomenclature is
employed here to emphasize man's position and view of the physical matrix
of his environment. For instance, in the vernacular, a farmer comments
"I have most of my 'prairie land' in pasture." Thus the geographic sub-
regions are discussed here as land types rather than by geology, soils,
vegetation, and land use separately. As a geographer, this writer emphasizes
man's perception of and interaction with his environment; therefore the
vegetative terms for land types are used here, in preference to the com-
plex spectrum of soil variations.
The chief natural growth in the study area once was the longleaf
pine; these lands were referred to as either "rolling pine lands," or
"flatwoods." The rolling pine lands are slightly undulating, well-drained
lands and can be subdivided into: "open pine," "pine with blackjack
undergrowth," and "pine with turkey oak undergrowth." The open pine areas
with tall longleaf pines are almost completely gone today, and much of
this land is now cropland, pastures, or planted pines (see Figure 9).
Blackjack oak (Quercus marilandicus Muench.) still abounds in the driest
or most barren areas and the turkey oak (Quercus laevis Walt.) predominates
on soils a little finer (see Figure 10). The soil in the rolling
pine lands is mostly thick sand with minor components of silts and clays.
The hills overlie an uneven limestone surface with scattered large sink-
holes. Small outcrops of flinty limestone are present in the soils and
humus is sparse or entirely lacking because of repeated fires. The
gopher mouse or "salamander," as it is called by the native Floridians,
burrows into the sandy dry soils, leaving low mounds a foot or more in
diameter. Not only does this help to mix the soil, but also encourages
young pine seedlings to get a better start in the fresh mound. The bare
sand also offers the very young tree some protection from fire. When
the tree is older it withstands ground fires well, and most mature trees
are blackened at some time or other during their growth.
"Hammock" as a vegetative term is applied to areas where the timber
growth is dense enough to result in the accumulation of leaf mold on the
surface. There is a wide variation of vegetation types and plant groups
in the makeup of different hammocks, owing to the variables of soil and
drainage conditions. Those having the poorer, sandy soils will have a
higher percentage of pines than those with richer soils. The richest
soils of all will have large stands of fine hardwoods and associated
flora, which give the hammocks their wide appeal to farmer and naturalist
alike (see Figure 11).
Within the hammocks there is competition for dominance among the
hardwoods, and between the hardwoods and pines. Pines do not tolerate
the crowding which results from good growth on nutritious soils, and
~8Z~Cin-~ru CP -Y
I -~~5i~Ij;li '
9, n- s~l ~
-- '51 1 In rrr I r Ear
I~i~1~C~3t3 ~Q~ilZC1-~J-a~CS~?IC, ~ ~i ~L~5d~jl~Y~ blA;rl. E-Z:
~c, ~Lr~ ;,ic
~;c. ~l~j~t~z~ r .,
-_ r~1~4~;LcC*~ ?
.$ ;i III rf~
C;~ '' 1 i :1 uil~r~
I tif +j
r - ..r
Ir '~5 _~ Pb~l:
r r r
r 'r bi j r
r. 5 t; iii ~e4jsj Y
.rb- It i
j B ;r 'i
'' C Y;
4. ,, 'L~~.' :il
'S~t; ~i ..
hardwoods do not tolerate the fires to which pines are resistant. Fires
across the land have always controlled the landscape in Florida, and
pines quickly replace hardwoods which have been burned out. Rocky out-
crops or wet sink areas offer some fire protection to forest cover.
Hardwood hammocks will slowly accumulate leaf mold around their boundaries
by the steady dropping of leaves in these sub-tropical conditions and
thus, by soil enrichment, slowly invade the surrounding pinelands.
The sandy soils of most hammocks are internally well-drained, with
frequent limestone sinks and sinkholes. Rainfall finds its way to the
sink areas, where almost no surface streams exist. Where hammocks have
been cleared off for cultivation, the surface soil varies from light
brownish-gray, to pale brown, to yellowish gray.3 The depth to lime-
stone may vary greatly from a few inches to many feet. Most of these
soils are planted to corn and peanuts, with considerable tobacco and
watermelon acreage as well. Even though the area soils are not especially
suitable for corn, it is grown extensively because of hog-raising needs.
This designation in Florida is a specialized use of the term "prairie,"
being applied by the early settlers to lands affected by freshwater in-
undation. The numerous lakes and ponds of the study are are all shallow
with nearly flat shores. The shorelines are subject to wide fluctuation
with the rainy cycles, sometimes being inundated, sometimes exposed.
Such areas, as well as the silted-in beds of former lakes, are termed
prairie land.14 The beds of former lakes represent either advanced stages
of eutrophication, or lakes whose sealed-in bottoms have become punctured
by natural causes, allowing the waters to drain away into the underlying
limestone. The huge Payne's Prairie in Alachua County is a well-known
example of such a karst lake which entered a downdrawn cycle in recent
historical times, to become grazing land.
The soil of these lake margins and other prairies in the study area
varies from rather sterile sand to peaty sand mixtures and occasional
silicified limestone. The vegetation in the deeper and more permanent
pools is entirely aquatic, whereas the outer pond margins can support a
few small trees and shrubs, and while these two vegetation types are very
different, there are all gradations in between. All treeless areas, per-
manently or periodically inundated, are therefore included under the de-
signation of prairie lands.15 To this writer, they represent one of the
more distinctive features separating Florida's from other karst terranes.
General Land Use Patterns
Land use in the study region has always undergone some changes, but
agricultural areas have remained fairly definite in location. The first
land use map of the entire United States was made by F. J. Marschner, an
agricultural economist with the Bureau of Land Economics,16 during the
period from 1945 to 1950. The compilation scale was 1:1,000,000 with a
map prepared for most states at that scale. The final more generalized
map was published in 1950 at a scale of 1:5,000,000. For the National
Atlas of the United States published in 1971, J. R. Anderson revised the
original Marschner map for publication at a scale of 1:7,500,000. He
mainly used aerial photography (mosaics of contact prints), as well as
some statistical and other surveys to make decisions about land uses
for each state* (see Map 8). The study area for west central Florida
is delineated as combinations of Cropland, Forest Land, or Grazing
Land, by letters C, F, and G-the order and case signifying the relative
importance of dominant uses. The study area is chiefly designated as
Cgf on the limestone plains, which means that cropland is the dominant
land use in the Bell-Trenton-Chiefland area with grazing and forest
land as secondary uses. "F" appears most frequently around the periphery,
sometimes in combination with "s" meaning swamp. The two lobes of the
limestone plain can be seen distinctly; especially well defined is the
Bell-Trenton-Chiefland extent. This localization of land uses corresponds
to areas similarly delimited on both vegetation and soil maps. Soil
associations are dependent upon geology; natural vegetation is governed
by soils; land use patterns seem to overlie vegetation tracts.
When agricultural areas were chosen for more intensive study of
the impact of karst features, Marschner's areas marked as cropland (in
combination with other uses) were used as a base for sampling and case
studies. From the natural growth, the hammock lands most often yield
to crops; wet prairie land, even within agricultural regions, is little
changed; but pine land, depending on economic conditions, has often
The effect of the limestone topography on agriculture can also be
seen, though not so dramatically, in examination of individual farmscapes.
On the case study farms, size and arrangement of fields sometimes indicate
*James R. Anderson, Chief Geographer, United States Geological Survey.
1MASCHNER'S LAND USE
Wet forest, swamp
and forest land
Mostly forest with some
crop and grazing land
1 1 000 000
0 10 20
presence of karst features. Dividing lines or field boundaries may
follow lines of sinkholes, adding an unwelcome problem of fencing to
that of sinks and erosion (see Figure 12).
The effect of the karst terrane on agriculture is that it is a
limiting factor on man's use of the land. Less freedom of choice at
everyturn is offered the farmer. The original choice for specific use
for subsistence, and later the response to economic factors, are always
influenced by the "lay of the land." In the karst terrane of the study
area, agriculture is practiced on the limestone plains despite the
limits and hazards of the landscape.
The pale sandy soils are of variable thickness and have been sub-
ject locally to severe wind erosion. They are excessively well-drained
xericc) because of the permeability of the underlying limestone, and
when left without cover, they can become severely eroded by wind. After
the land was cut over for lumber and subsequently cropped, the problem
of blowing in the 1930's was so drastic in local areas that the area
between High Springs and Newberry was compared to the dust bowl of the
Great Plains.7 The problem was somewhat alleviated by the introduction
of Pensacola Bahia grass in the late 1940's.18 Other solutions were tried
also, such as windbreaks of planted pines along highways and fence rows.
The strips of gred are an attractive addition to the landscape today.
In the whole of the study area the proportion of land in cropland,
forests, and pasture is approximately 3:3:2. Soil Conservation Service
samples chosen in agricultural lands in 1966 showed 37,36, and 23 percent,
respectively.19 Forests predominate in Dixie County and coastal areas as
respectively. Forests predominate in Dixie County and coastal areas as
well as in the river valleys and pocosin swamps. Cropland predominates
in the limestone plains, but both planted pines and improved pasture are
becoming increasingly important. Chief crops vary through the area gen-
erally,with tobacco bringing in more money and using less land than other
crops. The allotment system keeps the acreage small, but the yields are
high because of the intensive methods employed. Soil areas unmarked by
sinkholes or outcrops are chosen for tobacco, and irrigation is extensively
employed. Peanuts are important especially in Levy County and are grown
scientifically in the sandy soils of former pine lands. Corn is ubiquitous
and is mostly used for feed. Sweet corn, melons, and vegetables are grown
anywhere in the area. Citrus is important in the southernmost counties of
the study area-Citrus, Sumter, and Marion. Each year more land is used
for pasture as cattle raising becomes more desirable.
Sinkholes are scattered through the cropland and pastures on almost
every farm. Usually those in pastures that have precipitous slopes are
fenced or filled. Deep sandy soils mask small subsurface pipes and wells
by sliding and filling the hole by gravity and washing. Shallow soils
expose rocky outcrops and well-defined sinks.
Sinkhole Density and Land Use
Very few attempts have been made to count sinkholes to establish
density ratios. The introductory chapter describes the extent of previous
investigations. At the present time no counts have been made of sinkholes
of Florida until the present research was undertaken in land use for
agriculture. Several techniques were employed by the writer to reach
certain conclusions. A large mosaic was assembled from alternate air
photographs, sinkholes were located with stereoscope and magnifying glass,
and locations plotted on an overlay map (see Map 6 in Chapter III). The
area selected was chosen because it is a major agricultural area located
on the Bell-Trenton-Chiefland Limestone Plain. Sinkhole density over
the whole plain is estimated to be about 4 per square mile, or about one ,-
sinkhole for every 160 acres. This estimate is probably low rather than
high because in forested areas many sinkholes were obscured as well as was
any existing orientation. Also it must be remembered that most of the
sinkholes smaller than five feet in diameter could not be counted. There
is a definite clustering pattern in some areas of open land. No clustering
could be seen in forest land. The highest density in any square mile is
25, and there are a number of square-mile areas that do not have any sink-
holes. One advantage to this type of count is that patterns and longi-
tudinal trends in placement could be discerned. There is orientation in
a northerwest-southeast direction as well as north-south and east-west
strikes. This orientation as well as the clustering reflect the pattern
of joints in the limestone directly beneath the surface. Differences in
thickness of overburden doubtless influenced the surficial expression.
There were fewer holes along the periphery of the plain where alluvial
deposits and continuous forests occur.
Another tool for enumeration of sinkholes in agricultural lands is
the topographic map. Selected areas were counted, calling a closed con-
tour with hachures as a sink. Averaging counts of the several topographic
quadrangle maps of the Bell-Trenton-Chiefland Limestone Plain showed about
24 sinks per square mile. This figure of one sinkhole for each 27 acres
is a much higher figure than computed from the aerial mosaic. Another
map used for a partial enumeration was the Alachua County Soil Map with
sinkholes identified. The highest density, 22 per square mile, occurred
in sandy soils near High Springs. The sinkholes are scattered in a
northwest-southeast orientation apparently irrespective of differences in
soil type in the limestone plain of western Alachua County.
Field reconnaissance by the writer proved both counts to be less
than actual numbers by a factor of up to 50 percent. Actual field ob-
servation was the only method feasible for counting outcrops except in
the sample areas covered by the Conservation Needs Inventory, which yielded
fairly accurate data on karst features. These counts were actually made
in the field, however, and recorded on photos by soil scientists.
The Soil and Water Conservation Needs Inventory for Florida was
developed as part of the National Inventory for the 50 states. The pur-
pose was to provide information for federal departments and other agencies
to use in programs for resource conservation. The first National Inventory
was carried out in 1958 and updated in 1966. Its purposes include esti-
mates of acreage in each land use and treatment needs for that use. Land
uses are defined as follows:
Cropland-Land being used for field crops, tillage rotation, or-
chards and goves, temporarily idle land, and open land formerly cropped.
Pasture-Lands producing forage plants, principally introduced
species, for animal consumption.
Forest-Land at least 10 percent stocked by forest trees of any
size and not currently developed for a nonforest use.
Other land-The concept was changed between inventories. The 1958
inventory included most idle land in this category. The updating shows
most of this in other uses. This is land which is not cropland, pasture,
or forest land and includes strip mines, borrow and gravel pits.
In the process of inventory, the surveyors were directed to make
note of sinkholes, outcrops, wet spots, and miscellaneous karst features,
to serve as additional information to the primary purposes of the inventory.
The updating in 1966 re-examined the same sample areas as were studied in
1958. Randomized samples were selected by the statistical laboratory at
Iowa State University, and the sample units were located on county maps.
The standard size of the sample unit was 160 acres. The basic sampling
rate was 2 percent in a county or other area of 250,000 to 500,000 acres.
The study area in the limestone plains sums to more than 400,000 acres,
providing data of an acceptable degree of reliability. The intent was to
county sinkholes and other karst features in land use types (see Map 9).
An additional experiment was intended to provide data on karst features
for lands of various soil types.
In updating the Inventory, a different system was used for ob-
taining more data from the sample areas. Each sample area was marked with
random points (36 points per 160 acre sample area) at which time a record
was made of conditions existing at the site of each point. A template
was spun to give random orientation to the points and varied the number
fallingin the sample from 32 to 38. Land use determinations at each
GILCHRI5T ) /
(70 !High Springs
SETS OF SAMPLE AREAS
. ~ ALACrtUA
\.I -./ .
l Location of Sample Plots
SCS Conservation Needs Inventory
3 10 20 3
designated point were in terms of the field in which the point fell.
Therefore, one point could be interpreted as representative of land
use of 4.4 acres.
Fifty-five samples fell in the selected agricultural areas of the
limestone plains which encompass Bell-Trenton-Chiefland and High Springs-
Williston-Reddick. Every statistical treatment considered to prove valid
was executed with the sample plots. The results have provided sinkhole
counts generally, as well as counts categroized by land use.
These samples total 8,800 acres, or more than the 2 percent required
for validity. Results are shown in the paragraphs which follow as per-
centages of land estimated to be in a particular land use and soil
Cropland: The percent of land in the selected agricultural areas
that was used for cropland in 1958 was 37 percent and it remained at
this proportion in 1966. Of the total in cropland, more than four-fifths
was on Jonesville-Chiefland-Hernando soils (numbers 4 and 9 on the soil
map) despite the fact that the incidence of sinkholes was extremely high.
Cropland counts yielded: one sinkhole per 35 acres, one outcrop per 54
acres, and one "other" karst feature (wet spot, karst pond, quarry, etc.)
per 295 acres. The average is one karst feature for each 20 acres in
cropland, which fact indicates that farmers do not avoid cropping lands
with sinkholes if the lands are other wise suitable.
Forest: Thirty-three percent of the total land area was used for
forest in 1958 and 36 percent in 1966. The largest amount of this forest
was on Lakeland-Eustis-High Blanton soils (number 3) with few sinkholes
counted. Next highest count is on Jonesville-Chiefland-Hernando soils which
have many karst features. Forest land counts show: one sinkhole per
56 acres, one outcrop per 22 acres, and one "other" per 50 acres. The
average is one karst feature for every 12 acres of forest land. This
high density was not perceptible from study of the stereo-pairs. It
may indicate some land use changes from other uses.
Pasture: The proportions of land in pasture rose from 16 percent
in 1958 to 23 percent in 1966. Largest amounts of pasture are found on
Jonesville-Chiefland-Hernando and on Arredondo-Zuber-Fellowship soils
(numbers 5 and 8). Highest incidence of sinkholes in pasture land is
on soil grouping 5 8. In land used for pasture, there is one sink-
hole per 51 acres, one outcrop per 9 acres, and one other karstt" feature
per 63 acres. When the total number of these features is divided into
the total acreage involved, an average of one feature per 7 acres results.
This is the highest incidence recorded in the survey. Information derived
from the case study interviews suggests that cropland sinkholes are
treated by pushing in, whereas no such effort is exerted in pasture areas,
since mechanical equipment is used much less frequently on pastureland.
The CNI count included a category labelled "idle and other lands."
The chief use of this category is to base all acreage figures on 100
percent. The small amounts of land involved preclude drawing any re-
levant conclusions from this "filler" element.
Land Use and Soil TvDe
An attempt was made to compare land use and soil types to determine
whether there was any preference or lack of preference for using soils
with karst influences. The first step was to calculate what percent of
total land was in each soil grouping. Soils with similar properties were
grouped to simplify results. (See first part of Table 5.) This portion
of the table relates the density of counted karst features to soil asso-
ciation. The second part of the table gives the percent of land in each
land use, also by the soil associations numbered and named on the table.
Association 3 shows a disproportionately low density of karst
features, which the writer believes may be attributed to the deep sands
characteristic of this association, which fill in subsidence events and
mask their surface expressions. There are more actual counted features
on the Jonesville-Chiefland-Hernando soils (numbers 4 and 9) because
more of the land is in this soil association than any other. The
largest percent of the land classified as cropland is also in this
association. Soils 4-9 directly overlie limestone, and subsidence is
common. Soils 5-8 account for about one-sixth of the land and more than
one-fourth of the counted karst features. These are mostly hammock lands,
fertile, and influenced by the subsurface paleo-karst conditions respon-
sible for the origin of the calcareous hammock soils in the first
instance. The last two categories in Table 5 do not require comment.
Land Use-Soil Type Correlation
In addition to the tabular matrix of data recorded in Table 5, a
statistical problem was proposed to examine less obvious idiosyncracies.
Using four land uses and five soil associations, an analysis of karst
features in each of twenty categories was attempted. (See Appendix I.)
4 N r, -.
0 1 -N en
- U N CO Cr 0 O
0 0 -
0 C 0 C
on C co to vr 0 o
H E > ". C 3
O 0 [-
Z 0 J0 Ci 0 QH 0 0
c (. C03 C C C.3C
*1 0 .- 0 O --4 CA- 0 O
S- 3- 0.' ) 0 0 0 0 O-' .- 0
z- .-iz -L) < N L.. < Od H
Fifty-five photos were used as the basis for a sampling count, each
covering 160 acres on the Bell-Trenton-Cheifland limestone plain. It
was decided first to compare all land use categories disregarding soil
types and then to compare all soil types disregarding land use. The
ultimate purpose was to ascertain numbers of karst features per land
use-soil type category. (See Appendix I.)
The results suggest to this student that true numbers of features
in cropland and idle land were less related to soil type than those in
forest land or pasture land. Or, put another way, if land is perceived
as cropland, it will be used as cropland, despite its karst features.
There was also clear evidence that soils of group 3 were least affected
by land use-soil type correlation.
The present condition of the soil is the result of continuous action
of cultural influences and physical factors. No doubt soil structure and
composition reflect the presence or absence of limestone near the sur-
face. Drainage factors and subsequent erosion damage also indicate sub-
The actual productivity of the soil is best judged by examination of
crop yields under certain conditions. Production data for the Georgia-
Florida physiographic area were selected as a broad basis for comparison.
These data were published by the U. S. Study Commission for the Southeast
River Basins in 1961, and indicate average yields computed annually over
a 35-year period for crops grown in the Coastal Plain Physiographic
Province. The figures were adjusted to account for extremes influenced
by climate, etc., hence are "normalized." As such, they may be used for
comparison against figures for any given individual year, such as census
years, for example. Table 6 lists crops selected as common to both the
west central Florida counties and the Coastal Plain.
For the study area, average yields computed for Gilchrist and Levy
Counties, as compiled by the Census of Agriculture, 1964, indicate the
comparative productivity of the limestone plain. (See Tables 7 and 8.)
These counties were chosen as representative because most of their farming
areas are in the limestone plain chosen for sampling.
Additional information given in Table 6 shows possible yields
under various management practices. Group A includes farmers who use
average to below average management practices for seedbed preparation,
fertilization, seed selection, cultivation, and use of pesticides.
Most of the farmers in Gilchrist and Levy Counties are considered to
conform to Group A criteria. Group B makes full use of all applicable
recommended management practices except irrigation. Only a few farmers
practice irrigation in the study area, tobacco and truck crops being
most affected. Management Group C encompasses all criteria for Groups
A and B plus irrigation. This factor gives Group C very high yield
estimates which affect the summarized figure for "All Levels." This
figure is a weighted average, computed by use of percentages of land
in each management level group. Thus, the tables show the limestone
plain to be producing close to average yields for corn, somewhat lower
than average for tobacco production, and moderately high yields for
peanut crop production.
Table 6 Normalized Average Yields for 35-Year Period:
Plain Physiographic Area
A B C
20 40 58
Average Management Practices
Improved Management Practices
Improved Management Practices
Source: U. S. Study Commission, Southeast River Basins, 1961.
Table 7 Selected Crop Yields Computed for Gilchrist County
Source: Census of Agriculture, 1954, 1959 and 1964
Table 8 Selected Crop Yields Computed for Levy County
Enterprise Per Acre 1954 1959 1964
Corn bushels 15 21 26
flue-cured pounds 991 1031 1295
Peanuts pounds 636 912 1502
Bermuda tons NA NA 2.31
All other tons .91 1.32 .76
Source: Census of Agriculture, 1954, 1959 and 1964
The soils of the agricultural area may be considered also and
have been grouped for this purpose. Information available indicates
that no significant differences prevail in average yields for a given
soil group by river basins within a physiographic area.20 Therefore,
data for the Suwannee River Basin soils are considered to be fairly
accurate measures of area soil productivity (see Table 9). Highest
yields in the basin areas for corn are in soil grouping 5 8 (which
once was mostly high hammock land). These are the calcareous hammock
soils typically of the Zuber-Fellowship Association. Lowest yields
are in the areas of Hernando-Chiefland soils which are almost directly
overlying limestone and are locally eroded and sometimes wind blown.
Tobacco yields appear highest in the sandy soils of the former
rolling pine lands. These are Lakeland, Eustis, and High Blanton
Associations often with a deep sandy profile. Limestone is several
feet below the surface, but its irregular contours influence the
topography greatly. Lowest yields are again in the lernando-Chief-
Peanuts do well on the same soils that are good for corn. They
also yield profitably on the sandy soils of Association 3. They are
not suited especially to the Hernando-Chiefland or to the Prairie Land
soils. Patterns for hay follow those of corn and peanuts, with the
alternately wet and dry prairies being best suitable.
The influence of the limestone on the soils and thence on the
suitability of the land for crops is obvious here. Culturally in-
stilled farming methods have had some effect upon decisions made by
Table 9 Average Yields Per Acre and Management Level,* Coastal
Plains Physiographic Area, by Soil Types
and Land Type
3 Pine Land
4-9 Hammock &
% of land in
A B C All
15 37 48 20
7 18 44 10
36 44 67 30
15 26 33 17
80 17 3 100
A B C All
855 1412 1710 1306
446 743 1635 907
855 1412 1710 1306
743 1487 1784 1316
35 35 30 100
* Level A:
Average Management Practices.
Level B: Improved Management Practices
Level C: Improved Management Practices
Source: U. S. Study Commission, Southeast River Basins, 1961.