Geography 26 Project Papers: Spring 1999
The Use of Aerial Photography and GIS: A Part of a Reintroduction Viability Study of Ocelots in Texas
Robert Thomas

The Use of Aerial Photography and GIS: A Part of a Reintroduction Viability Study of Ocelots in Texas

By

Robert E. Thomas

For

Geography 26- Data Acquisition, Spring 1999                                   American River College, Sacramento, CA

And

v:shapes="_x0000_s1029">

Science Division                                                                               Defenders of Wildlife, Washington, DC

 

ABSTRACT

 

The use of GIS and aerial photo to determine habitat suitability and site selection has been common practice in wildlife conservation and land ecology in recent times. At present, there is an International multi-agency project that is comprised of individuals from private, public, and non-profit institutions who seek recovery for endangered felids; ocelots, jaguars, and jaguarundis throughout their historic range in the American southwest (Arizona, New Mexico & Texas), and adjacent Mexican states of Sonora, Chihuahua, and Tamaulipas.  The primary focus of this paper will be on the use of aerial photography to examine areas that have been indicated using GIS to be suitable for these cats. The Lower Rio Grande Valley region of south Texas will be the area of attention for this study. As a pilot area, the area of concentration will be in six counties: Maverick, Zapata, Cameron, Hildalgo, Starr, Kleberg, and Wallacy  (29.0625 to 26.4375 N; -97.6250 to -103.4375W).  ArcInfo’s grid analysis module was used to assist in determining percent coverage of vegetation after examining and evaluating the photos for vegetation. The use of this technique was reasonably successful in determining whither vegetation was present in each view.  In addition, the use of ArcInfo’s grid analysis was only moderately successful, due to the subjectivity of vegetation isolation in the analysis and quality of the image being analyzed. Overall, this process could be beneficial as a preliminary tool in habitat selection for these species.

INTRODUCTION

 

The use of Geographical Information Systems (GIS) to analyze and help make decision about species habitat problems is extensive (Chou, 1997; DeMers, 1997; ESRI, 1997). In particular, the use of GIS in the study of small bordercats (i.e. Ocelots, Jaguars & Jaguarundis, (Felis pardalis, Panthera onca and Felis yagouaroundi respectively)) in recent times has also occurred (Anderson, et. al., 1997; Haveson, Tewes & Anderson, 1997).

At present, there is an International multi-agency project that is comprised of individuals from private, public, and non-profit institutions who seek recovery for ocelots, jaguars, and jaguarundis throughout their historic range in the American southwest (Arizona, New Mexico & Texas), and adjacent Mexican states of Sonora, Chihuahua, and Tamaulipas. One of the working goals of this project is to focus on the decline of suitable habitat and begin preserving suitable areas. To address this goal, studies will build off the work of Anderson, et. al., did using GIS to study ocelot habitat and the classification structure of the Asian Tiger study done by the World Wildlife Fund and Wildlife Conservation Society (1997). The primary focus of this paper will be on the use of aerial photography to examine areas that have been indicated using GIS to be suitable for these cats. The Lower Rio Grande Valley region of south Texas will be the area of attention for this study. As a pilot area, the area of concentration will be in six counties: Maverick, Zapata, Cameron, Hildalgo, Starr, Kleberg, and Wallacy  (29.0625 to 26.4375 N; -97.6250 to -103.4375W). It is hoped that if this technique is successful, the results of this study will be able to be generalized to the lager domain of southern Texas (panhandle downward) as a whole when at a later time, it can be used to determine the extent of decline in suitable habitat since the early 1900's to the present.

Background:

The Lower Rio Grande Valley (LRGV) region of south Texas is composed of the counties of Starr, Hidalgo, Cameron and Willacy. In addition, the counties of Maverick, Webb and Zapata have been included in this study area because of their inclusion within the boundaries of the Tamaulipan Biotic Province, a region known for containing Tamaulipan brushland ecosystem. The importance of this area is that it is the only area of it's kind anywhere in the United States that has the combination of climate, vegetation and associated wildlife. The dense brush in this unique ecosystem provides food, nest sites and cover for many wildlife species (Jahrsdoerfer & Leslie, 1988). Many neotropical genera of mammals, reptiles and birds reach the northern limits of their distribution in the LGRV (Blair, 1950). In particular, the felid of this study, ocelots, use tracts of dense brush for cover and travel lanes in this area (Tewes & Everett, 1982).

Throughout this past century, human impacts on Tamaulipan brushland have been tremendous. Since the 1920's, more than 95% of the native brushland in the LRGV has been converted to agricultural or urban use (USFWS, 1980; Parvin, 1988a,b). More than 90% of the riparian habitat on the US side of the Rio Grande has been cleared (Collins, 1984). Of the lush subtropical region of the delta area of the LRGV, it has been estimated that 98% in the US (USFWS, 1980), and a large percentage of similar habitat in Mexico has been cleared (Collins, 1984).



The ocelot is a spotted medium-size neotropical cat with a moderately long tail that ranges from southern Texas and Arizona south to Peru and northern Argentina (Figure 1) (Murray & Gardner, 1997; Hall, 1981). It has been reported, as late as the 1800's, to be found in east and central Texas, western Louisiana and southern Arkansas (Laack, 1991; Navarro-Lopez, 1985; Woodward, 1980). Historically, in the United States, two subspecies of ocelots have been determined. The Texas ocelot, Felis pardalis albescens, and the Sonora ocelot, Felis pardalis sonoriensis (Hall, 1981; Harwell & Siminsk, 1990). It has been documented that there are eight other subspecies of F. pardalis or Leopardus pardalis[1] (Murray & Gardner, 1997).

The ocelot, now listed as of 1982, as an Endangered Species in Arizona and Texas (Federal Register, 1982) also is afforded international protection since 1992 under CITES Appendix 1 (Conservation on International Trade in Endangered Species of Wild Fauna and Flora, 1983).  In spite of these listings, the ocelots is still being trapped for the fur industry in central and south America and the habitat crucial for their survival being destroyed worldwide.  Finding and protecting large areas of suitable habitat, will be a key to this species long-term survival.

 

MATERIALS AND METHODS

 

Electronic Data Sources:

            All of the data sets, maps and images are listed below with sources.

Base Map

Texas Parks & Wildlife Department (TPWD) GIS Lab Homepage (http://www.tpwd.state.tx.us/admin/gis/download.htm).

Texas Vegetation

Natural Regions of Texas

                        Major Natural Regions

                        Subregions

Texas Counties

ESRI Data & Maps CD#3 (Western States of the US)

Texas Roads Data Set

Texas Rivers Data Set

Aerial Photography:

Microsoft TerraServer (http://terraserver.microsoft.com/advfind.asp)

14 images of 6 counties

 

Data Preparations:

The creation of the base map done by extracting coverage's from the downloaded .e00 files from TPWD and converting them to shapefiles.

To determine possible habitat locations based on vegetation data, literature reviews were conducted.  Based on associated plants listed in the various physiognomic regions, 13 vegetation classes chosen and mapped.  Roads and River data were brought in as separate themes and overlaid onto the possible habitat map, to determine possible fragmentation due to these to barriers.  Next using a historical sightings maps by county (Davis & Schmidly, 1994), points were on-screen digitized.  Doing a theme-by-theme selection, all vegetation classes that fell within the sightings area were clipped and mapped as possible habitat locations. 

Using these county locations, aerial photos of various locations within these areas were downloaded from the Microsoft/USGS TerraServer.  Each corner point of these photos was brought into ArcView and overlaid on the base map of possible habitat. 

Using ArcInfo, photos were imagegrid processed. After determining the lightest dark spot that indicated vegetation, the photos were reclassified to 1's & 0's (W or B) (Namecover = con (photoname < ##, 1, 0)) based on the cellvalue. They were then brought ArcView as grid coverages.  Using Spatial Analyst, coverages were evaluated.  Looking at the cell counts (the 0's & 1's), in the legend editor, the counts were brought into Excel and percent cover computed (1's/total).

 

RESULTS

The evaluation of the vegetation classes (149 classes within 8 physiognomic regions) yield 13 classes:

Oak-Mesquite-Juniper Parks/Woods

Mesquite_Lotebrush_Brush

 Mesquite_Juniper_Live_Oak_Brush

Mesquite_Juniper_Brush

Mesquite_Juniper_Shrub

Mesquite_Granjeno_Woods

Mesquite_Granjeno_Park

Mesquite_Saltcedar_Brush/Woods

Creosotebush_Lechuguilla_Shrub

Mesquite_Hackbarry_Brush/Woods

Juniper_Mixed_Brush

Mesquite_Blackbrush_Brush

Yucca_Ocotilo_Shrub.

 

These classes encompassed 254/1533 polygons, which calculated out to be ~39779142 acres or ~twelve % of the state.  After conducting the overlay process of sightings, four classes were left; Creosotebush_Lechuguilla_Shrub, Mesquite_Blackbrush_Brush, Mesquite_ Granjeno_Woods and Mesquite_Granjeno_Park, for a total of ~10925482 acres or ~6.3 % of the state.

 

Of the fourteen photos that were processed, only one could be analyzed for percent-coverage due to problems with software (segmentation error-ArcView) (Table 1).

Images

No Vegetation Cells

Vegetation Cells

Total Cell Count

% of Coverage

Match Rating Lw/Med/Hi (1-3)

Cam1Cover

 

 

 

 

 

Cam2Cover

 

 

 

 

 

Hid1Cover

 

 

 

 

 

Hid2Cover

 

 

 

 

 

Hid3Cover

 

 

 

 

 

Klb1Cover

 

 

 

 

 

Klb3Cover

 

 

 

 

 

MavCover

71,477

61,131

132,608

46.1%

3

Mav1Cover

 

 

 

 

 

Mav2Cover

 

 

 

 

 

Str1Cover

 

 

 

 

 

Str2Cover

 

 

 

 

 

Wal1Cover

 

 

 

 

 

Wal2Cover

 

 

 

 

 

Zap1Cover

 

 

 

 

 

Summary (Averages)

X

X

X

 

 

 

The Maverick county photo-1 (Mav1Cover) produced a total count of 132,608 cells of which 61131 cells that met the color value of the photo (has vegetation) for a percent coverage of 46.1% and a matchablility index of high (3).

 The use of this technique was reasonably successful in determining whither vegetation was present in each view.  In addition, the use of ArcInfo’s grid analysis was only moderately successful, due to the subjectivity of vegetation isolation in the analysis and quality of the image being analyzed.

CONCLUSION

 

This technique’s success and failure are due to the limitations of the process.  For example the variations in photo quality ( 2nd generation photos, darkness, & tiling done by USGS/MS Terraserver) and the subjectivity of color determination, and the imagegrid, all had limitations, could have synergistic negative effect on the process. As an example, the USGS logo was picked up as vegetation due to the black lettering fit the selection criteria. Though this is a negligible amount of cells, it nevertheless shows the lack of precise control necessary to increase to accuracy.

Overall, this process could be beneficial as a preliminary tool in habitat selection of this species.

 

WORKS CITED

 

Anderson, G. L., M. Bray, D. Griffin, P. K. McDowell, W. Swanson, and M. E. Tewes 1997. Using remote sensing and GIS to map vegetation and determine Ocelot habitat. Proceedings of the 16th Biannual Remote Sensing Workshop, Weslaco.

 Blair, W. F. 1950. The biotic provinces of Texas. Texas Journal of Science, 2(1):93-117.

Chou, Y, 199x. Exploring spatial analysis in geographical information systems. OnWord Press:Santa Fe:29.

Collins, K. 1984. Status and management of native south Texas brushland. U.S Fish and Wildlife Service, Ecological Service, Corpus Christi, TX.

Conservation on International Trade in Endangered Species of Wild Fauna and Flora (CITES), 1983. CITES appendices-Mammals. U.S. Department of the Interior. Washington, DC.


Davis, W. B. and D. J. Schmidly, 1994.  Distribution of the Ocelot in Texas based on known county records.  The mammals of Texas- online edition (http://www.nsrl.ttu.edu/tmot1/images/dmap258.jpg).

DeMers, M. M. 1997. Fundamentals of geographic information systems. John Wiley & Sons:New York.

ESRI, 1997. Getting to know ArcView GIS. Environmental Research Institute:Redlands, CA,:xx.

Hall, E. R. 1981. The mammals of North America, Vol. II. Ronald Press Co., New York.

Harwell, G and P. D. Siminski 1990. Listed cats of Texas and Arizona: recovery plan (with emphasis on the Ocelot). For U. S. Fish and Wildlife Service, Region 2, Albuquerque, New Mexico. Ed. A. M. Shull, S. Van Riper, S. P. Thompson & S. E. Jahrsdoerfer.

Haveson, P. M., M. E. Tewes, and G. L. Anderson 1997. Using GIS to analyze habitat selection by Ocelots and Bobcats. Proceedings of the Caesar Kleberg Wildlife Research Institute: Current Research 1996-97, Kingsville, TX,:33.

Jahrsdoerfer, S. E. 1988. Tamaulipan brushland of the lower Rio Grande valley region of south Texas: description, human impacts and management options. Biological Report, 88(36). For U. S. Fish and Wildlife Service, Region 2, Albuquerque, New Mexico.

Laack, L. L. 1991. Ecology of the Ocelot (Felis pardalis) in south Texas. MS Thesis, Texas A & I University, Kingsville.

Murray, J. L. and G. L. Gardner, 1997. Leopardus pardalis. Mammalian Species, 548:1-10.

Navarro-Lopez, D. 1985. Status and distribution of the Ocelot in south Texas. MS Thesis, Texas A & I University, Kingsville.

Parvin, B. 1988a. Valley under siege. Defenders, 63:18-29.

               . 1988b. The disappearing wild lands of the Rio Grande Valley. Texas Parks & Wildlife, 46:2-15.

Tewes, M. E., and D. D. Everett 1982. Study of endangered ocelot occurring in Texas. Year-end report. U.S. Fish and Wildlife Service, Region 2, Albuquerque, NM.

Shindle, D. B. and M. E. Tewes 1998. Woody species composition of habitats used by ocelots (Leopardus pardalis) in the Tamaulipan Biotic Province. Southwestern Naturalist, 43(2):273-279.

U.S. Fish and Wildlife Service, 1980. Department of the Interior Habitat Preservation Plan- preservation of areas of important fish and wildlife habitat: Cameron, Hidalgo, Starr and Willacy counties. U. S. Fish and Wildlife Service, Region 2, Albuquerque, New Mexico.

Woodward, D. W. 1980. The Ocelot. Selected vertebrate endangered species of the seacoast of the United States. FWS/OBS-80/01.9.

World Wildlife Fund-US 1997. A framework for identifying high priority areas and actions for the conservation of Tigers in the wild. World Wildlife Fund, Washington, DC.



[1] These two naming schemes are used interchangeably throughout the literature.  For this paper, Felis pardalis will be used and be the main species of focus.