Citation
Mapping Everglades alligator holes using color infrared aerial photographs

Material Information

Title:
Mapping Everglades alligator holes using color infrared aerial photographs
Series Title:
Florida Scientist, 64 (2) : Spring 2001
Creator:
Campbell, Mark R.
Mazzotti, Frank J.
Publisher:
Florida Scientist
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
University of Florida. ( LCSH )
Biotic communities -- Florida ( LCSH )
Natural history -- Florida ( LCSH )
The Everglades ( local )
Alligators ( jstor )
Vegetation ( jstor )
Everglades ( jstor )
Spatial Coverage:
North America -- United States of America -- Florida

Notes

Funding:
This collection includes items related to Florida’s environments, ecosystems, and species. It includes the subcollections of Florida Cooperative Fish and Wildlife Research Unit project documents, the Sea Grant technical series, the Florida Geological Survey series, the Coastal Engineering Department series, the Howard T. Odum Center for Wetland technical reports, and other entities devoted to the study and preservation of Florida's natural resources.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All rights reserved, Board of Trustees of the University of Florida

Downloads

This item has the following downloads:


Full Text




Made in United States ofAmerica
Reprinted from FLORIDA SCIENTIST
Vol. 64, No. 2, Spring 2001
Copyright by the Florida Academy of Sciences, Inc. 2001

MAPPING EVERGLADES ALLIGATOR HOLES USING
COLOR INFRARED AERIAL PHOTOGRAPHS


MARK R. CAMPBELL AND FRANK J. MAZZOTTI

University of Florida, Everglades Research and Education Center, PO Box 8003,
Belle Glade, FL 33430


ABSTRACT-Color infrared (CIR) aerial photographs were used to map '. .. holes in
a 2,442 km2 area (Water Conservation Area 3) of the Everglades. Open-water ponds appeared
as clearly defined black spots on the photographs. Three types ofalligator holes were identified
based on differing CIR signatures: Type I holes were encircled by a bright red ring :
a surrounding zone of woody vegetation (small trees and shrubs); Type 2 holes were not
surrounded by a change in pixel brightness or tone, -. .. a round open water pond within
the existing vegetation matrix without a surrounding ring of woody vegetation; and Type 3
holes were ..- .- adjacent to elongated red ovals (small tree islands) formed from the
construction of spoil banks and are termed "artificial holes. Eight hundred forty-five alligator
holes greater than 5 m in diameter were mapped, including 309 Type 1, 331 Type 2. and 205
Type 3 holes. These .. holes ranged from five to 15 m in diameter and 20 to 150 cm
deeper than the surrounding marsh (n = 49). Based on the ground-truthing of 89 mapped
alligator holes, 88 percent were successfully located in the field, 83 percent were correctly
characterized, and the holes were found within 60 m, on average, of their map coordinates.



AMERICAN alligators (Alligator mississippiensis) in the Florida Ever-
glades have the ability to alter the structure of the landscape (Craighead,
1968; Kushlan, 1974). Due to an alligator's large size and weight, and the
soft, organic, peat sediments of central Everglades marshes, alligators main-
tain, and sometimes create, small ponds called alligator holes by wallowing
into the peat, thereby mounding soil and vegetation around the perimeter.
This causes significant topographic and hydrologic variation in an otherwise
flat, shallow wetland landscape, resulting in increased vegetative and wildlife
diversity (Craighead, 1971).
Alligator holes are approximately 1 m deeper than the surrounding
marsh, reaching down to the limestone bedrock. They may range from 2 to
15 m in diameter and are found in a variety of major wetland habitats
including sawgrass marsh (Cladium jamaicense), wet prairie (emergent rush
marsh including Eleocharis spp. and Rhynchospora spp.), and slough [deep-
er water with floating aquatic plants such as spatterdock ".....':. luteum),
white water lily ( '. '... ... .... odorata), floating heart (Nymphoides aquatica,
and bladderwort (Utricularia spp.)]. Trees and shrubs, such as willow (Salix
caroliniana), often take root on the raised banks surrounding alligator holes.
In other cases, alligator holes may simply be small open water depressions
hidden within the surrounding vegetative matrix. Once established, alligator








No. 2 2001] CAMPBEILL AND MAZZOTTI MAPPING ALLIGATOR HOLES 149


holes are t free from encroaching vegetation by the *... .:i ...: mainte-
nance activities of the resident alligator.
Alligator holes have long been hypothesized to provide critical dry sea-
son i:. for Everglades :i ii:i (Davis, 1943; Kushlan, 197.,: As the
.............marsh dries down, fish and other aquatic organisms concentrate
in the .i: : hole, becoming an important food source for nesting wading
birds, and resident :i : .: : When water levels increase, surviving organ-
isms ...... i... and colonize surrounding wetlands. For this reason, .
holes have been recognized as important components in the process of life
and death within the Everglades ecosystem ( .: ..-.- and Brandt, 1' ::
A l:i: :: !: the importance of alligator holes in the Everglades is widely
recognized, only one alligator hole has ever been quantitatively studied
(Kushlan, 1972; 1974), and no inventory or mapping effort has been made.
As a result, the significance of alligator holes to the : .ii structure and
function of the Everglades is not known. The objective of this study was to
find a simple i" '.. to -. -- -: classify, and map :"i: ... holes using
CIR aerial photographs. i ..., : this, we asked several questions.
Could alligator holes be identified from CIR aerial .!... ... :. r What size
holes could be identified? How many types of -':. :. holes could be de-
scribed? How accurate was the identification and location of alligator holes?
Could this method be .:..... effectively over a large area?


MATERIALS AND METHODS-Stiud Area-This study focused on Water Conservation Area
3 (WCA 3), a 2,442 km2 water impoundment that extends from the L-5 levee (Palm Beach
County, Florida) in the north to U.S. 41 (Dade County. Florida) in the south (Figure 1). The
major habitats of this shallow, peat wetland arc sawgrass marsh. wet prairie, and freshwater
slough, which are dotted with occasional tree islands (clumps of small trees and shrubs) (Jordan
et al. 1997). Encompassing approximately one-third of the remaining Everglades fresh water
marsh, the major functions of WCA 3 are water supply, flood control, public recreation, and
wildlife habitat.

Data Sourc. ::. : holes were mapped using color infrared (CIR) aerial photographs
because they provided the necessary detail, were ..-' available, and covered the full extent
of the study area. In addition, CIR photographs are superior to conventional color or black and
white photographs for :- -- -. :' .. between water and vegetation, which was critical obr this
study, (Schneider, 1966; Shima and Anderson. 1976: Howland, 1980). The CIR photographs,
flown in September 1994. were obtained from the South Florida Water Management District
(SFWMD). Accompanying the photographs were a set of ground control points (GCP's) that
were collected by the SFWMD using a differentially corrected Global Positioning System
(DGPS) and helicopter. Approximately four control points per photograph were flown and
recorded as an image pinprick and as line-work on mylar overlays. Individual photographs were
9-by 9-inches at a scale of 1:24.000, resulting in a 3,025 ha photograph (5.5 km per side). A
total of 180 photos were used to map WCA 3.

Pilot study A pilot study was conducted within WCA 3 on a 250 km' area to determine
if alligator boles could be .. : .:: identified from the CIR aerial photographs. Fourteen
photos were professionally scanned at one-meter ground resolution (40 Pmi) and stored on CD-
ROM. The resulting digital color infrared (DCIR) raster images were approximately 100 Mb
each. The DCIR images were imported into EASI/PACE (PCI Geomatics. Ontario, Canada)







FLORIDA SCIENTIST


Alligator Alley


SWest
Palm
Beach




Fort
Lauderdale



Miami


S- 0 50 KM
Key West

FiG. I. Location of Water Conservation Area 3 in southern Florida.


[VOL 64










No. 2 2001] CAMPBELL AND MAZZOTTI-MAPPING ALLIGATOR HOLES 151


image processing software and georeferenced in the Universal Transverse Mercator (UTM)
ground coordinate system (Zone 17) tied to the North American Datum of 1983 (NAD 83)
using the GCP's :.- : by the SFWMD.
Potential holes were identified on the positive CIR film as small, dark, open-
water areas and subsequently located on the DCIR images on the computer. A light table and
magnifying loupe assisted in the identification of : ..... holes from the CIR photos in positive
transparency format. Color inkjet printouts of digital alligator hole enlargements were used to
record field characteristics associated with the holes such as open water, sawgrass, woody
vegetation, and shadows. The printouts contained a UTM grid for navigation to a
alligator hole with a DGPS and airboat or all-terrain vehicle. This information assisted in
classifying alligator holes into different types, and determining the minimum pond size that can
be interpreted from the photographs.

Mapping WCA .3-After the pilot study demonstrated the potential of CIR photography
for mapping alligator holes, a manual interpretation (U. S. Fish & Wildlife Service, 1994) of
all air photos covering WCA 3 was then conducted. With the aid of a light table and magnifying
loupe, potential : holes were located on each of the original 1:24,000 scale diapositives
and plotted on 9-by 9-inch clear acetate overlays with a fine point permanent marker (0.4 mm
nib). In addition, the location of the four GCP's, airboat trails, and canals were plotted on the
overlay for each photo.
To increase the amount of ground control per overlay (four GCP's are minimal), while
reducing the number of overlays for computer processing, 20 adjacent overlays were assembled
into a mosaic by lining Lup co common control points, airboat trails, canals, and alligator holes. A
second, continuous, transparent overlay then was placed over the mosaic, and all alligator holes,
airboat trails, canals, and the 20 control points were retraced onto a final single sheet. The large
overlays were reduced to 8/-by I 1-inch paper maps to accommodate the page
size of the scanner. Each reduced line drawing was scanned as a sharp black and white drawing
at 75 dpi creating a digital raster image. Once scanned, the digital line drawings were imported
into EASI/PACE image processing software and georeferenced using the ground control co-
ordinates collected by the SFWMD. Coordinates were entered from the keyboard and assigned
to their respective tick marks recorded on the digital line drawings. The image was georefer-
enced using a second order polynomial solution and a nearest-neighbor resampling algorithm.
The georeferenced images then were imported into SPANS geographic information system
(GIS) version 6.0, and compared to an existing georeferenced canal vector layer (obtained from
the SFWMD) to verify the success of georeferencing. To obtain a vector GIS layer of : ..
holes that allowed the association of attribute information to individual alligator holes, a manual
vectorization process was performed by on-screen digitizing a point layer of alligator holes (in
addition to the study area boundary and major airboat trails). Field data collected during ground-
truthing were linked to each alligator hole within the GIS database for subsequent analysis.
A random subset of a minimum of ten percent of potential .::. holes was ground
truthed to evaluate the error of commission. Aerial photographs were randomly chosen, and all
potential holes within that area were visited to maximize :* Potential holes were nav-
igated to using a DGPS and airboat. An alligator hole was determined to exist at the mapped
location if three criteria were met. First, if a depression in the nuck or limestone relative to
the surrounding marsh was found. Second, if the depression had a definite round or oval bound-
ary as indicated by the change in vegetation. Third, if the interior of the hole had no
emergent vegetation. Field information collected at the randomly selected .: .: holes in-
cluded pond length, width, and center depth, marsh waier deptl 10 meters from the edge of
pond, field UTMN coordinates, surrounding donation of vegetation, and photographs.
Due to a lack of the ability to do aerial surveys the error of omission was not evaluated.


:--Pilot study-The .: ..:' i .. of i:: ,., holes was easier
and more efficient using the positive ( ii film, i, i,: table, and .... ; ...







FLORIDA SCIENTIST


loupe than the DCIR images. The :.*. ":: film under magnification revealed
more detail than the 1-m DCIR : ..... and it was more ': .. : to ... -
ulate the photos on the light table than to pan and zoom '. ... :i images in
the computer. Within the 250-km2 pilot study area, a total of potential
holes were identified from 14 CIR :....- ::.'. Twenty-three potential al-
ligator holes were ground-truthed using the i: *::.. inkjet enlargements. The
alligator holes were located by DGPS to within fifty meters of the computer-
generated locations (varying from 0 to 26 m in the x direction and 0 to 50
m in the y ::- :: ::; The root-mean square error (RMSE) was 15.1 m.
i.. :. .:... of the image spectral characteristics showed that i .. wa-
ter *.:. ... i black, sawgrass was pink with mottled or rough texture, woody
i : was dark pink to red with a rough texture, and cattail (7ypha
domingensis) was mixed, appearing as white, i.: pink, or red, ....
on density, water depth, and other associated :*' The spectral signatures
for. :: water and woody vegetation were consistent across the 23 : I:
and could be readily identified. Five of the DCIR printouts with small islands
of woody vegetation contained neither a clear round black spot on the image
nor an open water pond in the field and were determined not to be i i: ..
holes. These small islands of woody vegetation were ; .: i investi-
gated to determine if alligator holes were present in areas where the CIR
S: ; : :i did not reveal well-defined circular black spots ... :' pond-
ed water.
Information on the 1-m resolution DCIR printouts was compared to
actual field values. There was no significant difference between the DCIR
S ... .. and actual ...! diameters (paired t-test, df = 21, P = I ,
thereby validating the accuracy of the information obtained from photo-
SThe minimum pond size : ..- -: on the :.w... ::-!- was five me-
ters in diameter. Nine ....::.. I .!i : : ponds, less than :: meters across,
were located in the field in close proximity (within 20 m) to larger ponds
.identified on the ; : '
Ai .: .. holes were classified into three i based on i::: : ..: spec-
tral characteristics on the CIR photos that resulted from varying physical
and vegetative structure. Type I alligator holes were round open water ponds
surrounded by either shrubs and trees, a ring of cattails, or a combination
of both. i i. are identified on the photos as distinct black spots surrounded
by an immediate red ring (darker in hue than the ...........:... vegetation).
Type 2 :ii, :. : holes were round open water ponds having a : ... bound-
ary clearly offset from the ... ....... vegetation matrix that are not sur-
rounded by shrubs or trees. They were identified on the photos as -.
black :...: within a red or pink matrix of marsh vegetation of varying
intensity, and are not surrounded by a bright red ring. Type 3 alligator holes
were termed ,.:.,: .,: since they were formed from the construction of
S:! banks, approximately 40 meters long by 15 meters wide, by .
up the muck with a drag line and piling it adjacent to the hole. A few
hundred of these artificial islands were created in the 1970s for the benefit


[VOL. 64







No. 2 2001] CAMPBELL. AND MAZZOTTI- MAPPING ALLIGATOR HOLES 153

of wildlife ( :. .. deer) in times of extremely high water, and are found
in close proximity (within 500 m) to each other as they string out :. --- the
levees. On the photos, Type 3 artificial holes were identified as round or
elongated black spots of open water .::::. =. 1 .:- to elongated is-
lands of bright red woody : .'. :: Only artificial islands that appeared
to have an p. *..:. pond were mapped.

:.. '.. WCA 3-A total of 845 alligator holes were identified i.....
out WCA 3 from the 180 CIR aerial ...... .:. '. This includes ':.: Type
I holes with ..... .... shrubs or trees, 331 Type 2 open-water ponds
without a .i: ::..i ring of surrounding shrubs, and 205 Type 3 .i:
holes at artificial islands (Figure 2).
A total of nine overlay-mosaics were created l overlays per mosaic)
that covered the full extent of WCA 3. After the mosaics were reduced,
scanned, and .... : .- '. the resultant RMSE averaged four pixels for a
thirty-meter resolution image (or 120 m on the ground). The canals on the
georeferenced images were approximately within two pixels (sixty meters)
of an existing : .. ... .. .: canal file (obtained from SFWMD), verifying
both the georeferencing and the accuracy of plotting information on the
acetate. This comparison of data sets indicated that the data manipulation
(plotting information on acetate, recopying, .. 1. ... and scanning) did not
compromise its spatial integrity beyond the purposes of this study. In large
part this is due to the very low : .: '.'. relief .. across WCA
3, .. ...... in m inim al : .. ... : .: : ... .: of .. .... elem ents,
but also :" ..- :...- are care and precision when manually manipulating data.
A subset of the alligator holes was located in the i: to assess the
commission error of the alligator hole map. Out of a total of 89 mapped
holes that were searched for, 11 holes could not be located, .
in an 88 percent commission rate for the map. For the individual .i
hole types, 90 percent (n = 42) of Type I was located. 74 percent (n = 27)
of Type 2 was located, and 100 percent (n = "':; of Type 3 holes searched
for was found. In addition, the commission error was not -" indepen-
dent. The majority of mis-commissioned holes (8 of I 1 or 73 percent) were
located in the southwest portion of the study area, a deep-water, open-slough
... The marsh water depth for the southwestern area (74 cm 16, n =
16) was '. ..: .. !. (t-test, P *:* I) than the central region of
WCA 3 (40 cm 1 1, n = 10). No structures indicating an alligator hole
were located for these eight potential holes, where it is : that ,'. water
within the slough habitat were misidentified. The remaining three
misidentifications, located in a northwestern sparse sawgrass marsh aver-
42 cm 12 water depth, were caused by round open water patches,
averaging nine meters in i ::.. : but without i: 1 .. in the muck.
The accuracy of the identification of ,7:. :... holes was assessed by
comparing the .i: ..... hole classifications from the photographs to the veg-
etation donation in the field (Table 1). Overall, 83 percent of the alligator







154 FLORIDA SCIENTIST [VOL. 64



j ,Hole Type


S. .\ .




-*

-* *' 0 / \ *



o. /

S,* \
\l.. ...
"; "* "' /, A/ ,,. ':.,
\ A ,\ \-








S1 10 KM,









FIG. 2. Map of Everglades alligator holes. Type I holes were surrounded by Small trees
or shrubs, type 2 holes were in the marsh and type 3 holes were associated with artificial tr
tified types. Of the 38 Type I holes located in the field, 74 percent (28











Of the 20 Type 2 holes located, 85 percent (17 holes) were found in saw-
'\. /x / N



j I


0 10KM

FIG. 2. Map of Everglades alligator holes. Type I holes were surrounded by small trees
or shrubs, type 2 holes were in the marsh, and type 3 holes were associated with artificial tree
islands.



holes investigated had physical characteristics that conformed to their iden-
tified types. Of the 38 Type 1 holes located in the field, 74 percent (28
holes) were correctly classified with surrounding trees and/or cattails, while
low shrubs and sawgrass surrounded the remaining 26 percent (10 holes).
Of the 20 Type 2 holes located, 85 percent (17 holes) were found in saw-
grass, while 15 percent (3) were surrounded by a cattail stand and therefore
misclassified. No Type 2 holes investigated were surrounded by trees. Fi-









No. 2 2001] CAMPBELL AND MAZZOTTI- MAPPING ALLIGATOR HOLES 155

TABLE I. Classification error matrix to assess the accuracy of map types.
Computer Classification (Column Total)
Field Classilication Type I Type 2 Type 3 Row Total
Type 1 20 0 0 20
Type I (cattail) 8 3 0 11
Type 2 10 17 0 27
Type 3 0 0 20 20
Column Total 38 20 20 78
Classification Accuracy 28/38 17/20 20/20 65/78
74% 85% 100% 83%



nally, all Type 3 holes investigated (20 holes) were located adjacent to ar-
tificial tree-islands, and were therefore correctly classified.
The spatial accuracy of the alligator hole locations was estimated, not
only by the RMSE generated through the georeferencing process, but more
importantly by comparing actual alligator hole ground locations to the com-
puter generated estimate. The difference between the computer coordinates
and those collected by DGPS in the field are plotted in Figure 3. The average
error between coordinate pairs was 60 meters, distributed evenly in all di-
rections. Therefore, the approximate spatial accuracy of the mapped gator
hole coordinates is +60 m.
Alligator holes ranged from five meters to over 15 m in diameter and
20 cm to 150 cm in basin depth (86 cm mean depth). A comparison of the
surface area of open water shows Type 2 holes to be smaller (43 + 33 sq
m, n = 21) than Type 1 (74 33 sq m, n = 28), (r-test, P = 0.002) (Figure
4). Large holes are generally deep, reaching the limestone substrate in ap-
proximately one meter, while small holes exhibit more variability. Most
small, deep holes were located in cattail stands.

DISCUSSION-Color infrared aerial photographs (1:24,000) were a useful
and efficient data source for mapping alligator holes over a large area. The
methods described here can be used to map other areas of the Everglades.
A total of 640 natural plus 205 artificial holes were identified using 180
photos over a 2,442 km2 area. Three types of alligator holes were identified
in this central Everglades area. In general, alligator holes with surrounding
shrubs and trees (Type 1) tend to be larger and deeper than alligator holes
without surrounding shrubs or trees (Type 2). Alligator ponds surrounded
by cattail were classified as either Type I or Type 2, and it would be useful
in the future to distinguish cattail zonation on the photographs. The third
type of alligator hole (Type 3) is identified from its association with elon-
gated, man-made, tree islands, and varies in size and shape from a few
meters in diameter to over 40 meters in length (the size of the spoil bank).
It is likely that in other areas of the Everglades with more tree islands
(Arthur R. Marshall Loxahatchee National Wildlife Refuge, Shark Slough







FLORIDA SCIENTIST


90 deg














,e* ( 100m 200m deg

\ e
**

100 m








270 deg
FIG. 3. Polar plot of the distance (meters) and bearing (degrees) errors for the computer
generated alligator hole coordinates (UTM) as compared to actual field locations found during
ground truthing. Average error is 60 meters.


in Everglades National Park) or a different substrate (rocky glades, Ever-
glades National Park) there will be other types of alligator holes.
The error of commission for alligator holes depended on size, surround-
ing vegetation, and habitat (matrix vegetation). The majority of commission
error was located in deep water, open slough habitat where breaks in the
floating surface vegetation resulted in round open-water patches. In addition,
round, vegetation-free patches were misidentified in shallow areas. These
open patches may be caused from spot fires. The overall error of omission
for the map was not determined, but alligator holes less than 5 m across
were present though they were not mapped. While it is possible that larger
alligator holes were also omitted it is likely that the majority of omission
lies below the five-meter minimum-mapping threshold. A combination of
aerial and ground surveys could be employed to investigate this error in
detail. Because of the patchy, heterogeneous makeup of the marsh vegeta-


[VOL. 64







No. 2 2001] CAMPBELL AND MAZZOTTI MAPPING ALLIGATOR HOLES


FIG. 4. Scatter plot of size vs. topographic relief (center pond water depth minus sur-
rounding marsh water depth) for both Type I and Type 2 alligator holes.



tion, identification of entities this small would rapidly increase the error of
commission. Small alligator holes with a distinct boundary were difficult to
distinguish from small, irregular open-water patches within slough habitat
that did not have a distinct vegetative boundary. Numerous potential alli-
gator holes were identified during the pilot study in two areas of the marsh
(approximately 25 km2 each) where the sawgrass becomes patchy and merg-
es into slough habitat. Investigation of these sites proved that caution should
be applied when identifying small, round, open-water ponds as alligator
holes in areas of patchy sawgrass and slough communities. After these sites
were re-evaluated, the number of potential alligator holes within the pilot
study area dropped from 548 to 173 holes.
The positional accuracy of the mapped gator hole locations was deter-
mined to be + 60 m, based on comparison with field data. This positional
accuracy could have been improved if the line work was scanned at a higher
resolution, resulting in a smaller pixel size and if the overlay did not have
to be reduced to 8.5 by 11 inches to fit the scanner. Nevertheless, the meth-
ods employed were sufficient for locating alligator holes in the field, and
acceptable for compiling a relatively small-scale map of alligator hole point
locations. The positional accuracy may not be sufficient for spatial analysis.
This mapping effort was accomplished with desktop computing systems,
using PC image processing and GIS software. Management agencies will
increasingly recognize the value of spatially analyzing environmental data.
This study demonstrates an efficient, economical method for high resolution
mapping over large spatial extent, and the value of integrating remote sens-
ing and field biology in a GIS/GPS environment.










FLORIDA SCIENTIST


ACKNOWLEDGMENTSl -This project was funded by a Challenge Grant from the Everglades
Agricultural Area Environmental Protection District and the National Fish and Wildlife Foun-
dation. We thank Dr. Peter Rosendahl for his constant support and encouragement. Color in-
frared photographs were provided by the Soumh Florida Water Management District, with a
special llanks to ILes Vilchek for his advice and generosity. Michelle Palmer assisted with
ground truthing and mapping. Logistic support, including airboals and off-road vehicles, was
provided by the Florida Game & Fresh Water Fish Commission and the Florida Cooperative
Fish & Wildlife Research Unit, University of Florida, Gainesville. Roy Welch, Laura Brandt,
and Mary IIudson Kelly reviewed the paper. .. .. .. for image processing and GIS work was
provided by PCI Geomatics, Ontario, Canada. This is a contribution of the Florida Agricultural
Experimental Station Journal Series, Journal Series Number R-07883.


LITERATURE CITED

CRAtGHEA D, E C., SR. 1968. The role of the alligator in shaping plant communities and main-
taining wildlife in the southern Everglades. The Florida Naturalist, 41: 2-7, 69-74, 94.
-- 1971. The Trees of South Florida. Vol. 1, The Natural Environments and their Succes-
sion. University of Miami Press, Coral Gables, FL.
DAVIS, .s I ., JR. 1943. The natural features of southern Florida, the vegetation and
the Everglades. Geological Bulletin No. 25, Florida Geological Survey, Tallahassee, FL.
JORDAN, E. H. l. JELKS, AND W. M. KITCHENS. 1997. Habitat structure and plant communiiity
composition in a northern Everglades welland landscape. Wetlands 17(2): 275-283.
HoW.AND, W. G. 1980. Multispectral aerial photography for wetland vegetation mapping. Pho-
togram. Engn. & Remote Sens. 46(1): 87 99.
KUSHLtAN, J. A. 1972. An ecological study of an alligator pond in the Big Cypress swamp of
Southern Florida. Masters Thesis. University of Miami, Coral Gables, FL.
-- 1974. Observations on the role of the American Alligator (Alligawor minssisippiensis)
in the southern Florida wetlands. Copiea 1974(4): 993 96.
MAZZOTTI, F J. AND L. A. BRANDT. 1994. Ecology of the American alligator in a Seasonally
Environment. Pp. 485 505. In: DAVIS. S. M. AND J. C. OGDEN (eds.). Ever-
glades: The Ecosystem and its Restoration. St. Lucie Press, Delray Beach, FL.
SCHNEIDER, W. J. 1966. Water resources in the Everglades. Photogram. Engn. 32(5): 958-965.
SHIMA, L. J. AND R. R. ANDERSON. 1976. The use of aerial color infrared photography in
mapping the vegetation of a freshwater marsh. Chesapeake Sci. 17(2): 74-85
U.S. FISH & WILDLIFE SFRVICE. 1994. Cartographic Conventions for the National Wetlands
Inventory, February,


Florida ... 64(2): 148-: 1=
Accepted: December 8, 2000.


[VOL. 64










FLORIDA SCIENTIST


ACKNOWLEDGMENTSl -This project was funded by a Challenge Grant from the Everglades
Agricultural Area Environmental Protection District and the National Fish and Wildlife Foun-
dation. We thank Dr. Peter Rosendahl for his constant support and encouragement. Color in-
frared photographs were provided by the Soumh Florida Water Management District, with a
special llanks to ILes Vilchek for his advice and generosity. Michelle Palmer assisted with
ground truthing and mapping. Logistic support, including airboals and off-road vehicles, was
provided by the Florida Game & Fresh Water Fish Commission and the Florida Cooperative
Fish & Wildlife Research Unit, University of Florida, Gainesville. Roy Welch, Laura Brandt,
and Mary IIudson Kelly reviewed the paper. .. .. .. for image processing and GIS work was
provided by PCI Geomatics, Ontario, Canada. This is a contribution of the Florida Agricultural
Experimental Station Journal Series, Journal Series Number R-07883.


LITERATURE CITED

CRAtGHEA D, E C., SR. 1968. The role of the alligator in shaping plant communities and main-
taining wildlife in the southern Everglades. The Florida Naturalist, 41: 2-7, 69-74, 94.
-- 1971. The Trees of South Florida. Vol. 1, The Natural Environments and their Succes-
sion. University of Miami Press, Coral Gables, FL.
DAVIS, .s I ., JR. 1943. The natural features of southern Florida, the vegetation and
the Everglades. Geological Bulletin No. 25, Florida Geological Survey, Tallahassee, FL.
JORDAN, E. H. l. JELKS, AND W. M. KITCHENS. 1997. Habitat structure and plant communiiity
composition in a northern Everglades welland landscape. Wetlands 17(2): 275-283.
HoW.AND, W. G. 1980. Multispectral aerial photography for wetland vegetation mapping. Pho-
togram. Engn. & Remote Sens. 46(1): 87 99.
KUSHLtAN, J. A. 1972. An ecological study of an alligator pond in the Big Cypress swamp of
Southern Florida. Masters Thesis. University of Miami, Coral Gables, FL.
-- 1974. Observations on the role of the American Alligator (Alligawor minssisippiensis)
in the southern Florida wetlands. Copiea 1974(4): 993 96.
MAZZOTTI, F J. AND L. A. BRANDT. 1994. Ecology of the American alligator in a Seasonally
Environment. Pp. 485 505. In: DAVIS. S. M. AND J. C. OGDEN (eds.). Ever-
glades: The Ecosystem and its Restoration. St. Lucie Press, Delray Beach, FL.
SCHNEIDER, W. J. 1966. Water resources in the Everglades. Photogram. Engn. 32(5): 958-965.
SHIMA, L. J. AND R. R. ANDERSON. 1976. The use of aerial color infrared photography in
mapping the vegetation of a freshwater marsh. Chesapeake Sci. 17(2): 74-85
U.S. FISH & WILDLIFE SFRVICE. 1994. Cartographic Conventions for the National Wetlands
Inventory, February,


Florida ... 64(2): 148-: 1=
Accepted: December 8, 2000.


[VOL. 64