INTRODUCTION
Although the knowledge base on the ecology of the RIFA (red imported fire ant, Solenopsis invicta Buren) is robust, causes for change in distribution and abundance in landscape mosaics are poorly understood. At a meso-scale (100 to 1,000,000 ha), the fire ant can be viewed from a meta-population perspective. Many entomologists are unfamiliar with the technologies used in a project of this scale and perspective. Populations are linked through dispersal behavior of adults, and are spatially separated by areas where the environmental conditions are less optimal. An important research question, relevant to integrated pest management of the RIFA, centers on how the spatial structure of habitat patches influences distribution and abundance of RIFA across complex landscapes.The goal of this project is to develop a risk rating system for RIFA in a post oak savanna landscape. Objectives are:
1. To develop a spatially referenced database for the post oak savanna study site,
2. To measure the distribution and abundance of RIFA at the study site,
3. To analyze the distribution and abundance of RIFA at the study site,
4. To synthesize the information to develop a risk rating system.
STUDY SITE
 |
|
| Post Oak Savanna is characterized by a mix of grasslands dominated by introduced coastal bermudagrass (Cynodan dactylon), and woodlands predominatly of native post oak (Quercus stellata), blackjack oak (Q. marilandica) and yaupon (IIex vomitoria). |  |
 |
|
| The landscape also typically contains cleared woodlands, agricultural fields, farm ponds and road corridors. |  |
 |
|
METHODS
Color infrared aerial photography was captured by the Texas Forest Service
in February, 1998 at 823.5 meters above ground level. Prints of the
large-scale positive color transparencies were scanned at 600 dots per
inch (0.23m resolution). Images were geo-referenced and rectified
to the Snook Digital Ortho Quadrangle (USGS) and mosaicked. USGS
specifications for DOQs require 10 meter spatial precision; however, RMS
error for CIR image registration by first order Affine transformation was
consistently better than 3.0 meters. Histogram matching was employed
to normalize the coloring between images.
Spatial data base development: CIR image mosaic of Sawdust
Ranch. Yellow arrows represent optimal flight path for image capture. Blue X's indicate
image principle points.
Landscape Classification
Landscape elements were classified by outlining, scanning and digitizing patches. Each polygon was assigned a unique identification code according to its corresponding landscape element type. Sampling protocol required that field records of landscape element were compared against the element types assigned by photo interpretation to verify accuracy of the classification.
| Modified versions of Anderson's (et al., 1976) land classification system and Lynch's (1996) landscape ecological mapping symbology were used. Landscape ecological symbology is based on a parent / daughter relationship for patches. Note the different patterning for each type of patch (stripes for introduced patches and checkers for disturbance patches). Color was used similarly for vegetation types. Dark green was used for all patches with post oaks, while pink was used in areas where agricultural products are grown. |
Waypoint Determination and NavigationFive percent of potential plots were randomly selected, each with a unique spatial identifier. Plot ID's and map coordinates were uplinked to a GPS receiver as waypoints. In the field, Real-time Differential GPS capabilities provided navigation to within two meters of plot centers. At the sample point, the operator collected a minimum of 20 GPS epochs and logged the collected data fields into its integrated data collector. |
 |
Sample PlotsPlots of 0.01 ha (5.65m radius measured by rope) provided a compromise between an area large enough to include numerous RIFA mounds yet small enough to minimize field time. Sampling was conducted in May-June, 1998; August-September, 1998; December, 1998; and February,1999. Numbers of active and dormant RIFA mounds by size class were logged electronically and recorded on field survey sheets. From a central location, the navigator recorded sample data for each point. |
| Field recorded land cover types were compared to the photo-interpreted base map to verify the landscape classification. |  |
Gyne Sampling
Sampling of nearly 80 ant mounds, distributed across the study area, was
conducted during June, 1998. This sampling aided in determining the
spatial distribution of single (monogyne) and multiple (polygyne) queen
forms. Determination of queen number follows Greenburg et al. (1985).
A mosaic of monogyne and polygyne colonies was found across the study site. |
 |
SAWDUST RANCH RIFA MOUNDS
PER SAMPLE
CONCLUSIONS
Several tools, not normally used in the study of insect ecology, have been employed to study RIFA distribution and abundance in a spatial context at mesoscale. Tools used include aerial photography, videography, geographic information systems, and real-time differential global positioning. Such technology enables researchers to truly randomize sample plots throughout the study area and conduct spatially referenced studies in a non-invasive manner on an actively managed resource.Use of the landscape ecological patch / corridor / matrix model facilitates interpretation of how RIFA utilizes the landscape. RIFA were observed in all landscape elements throughout the post oak savanna study site, however, considerable variation in numbers of active RIFA mounds was found. Preliminary study indicates this variation is not simply a function of patch type; other potential influences include patch size, patch interface composition, and distance to water. In further analysis, the distribution and abundance of RIFA will be examined with regard to content and context of landscape elements. It is likely that pattern in distribution and abundance will be understood by examining data at varying resolution and scale.
REFERENCES CITED
Anderson, J. R., E. E. Hardy, J. T. Roach and R. E. Witmer. 1976. A land use and land cover classification system for usewith remote sensor data, Geological Survey Professional Paper 964, United States Printing Office, Washington.
Avery, T. E. and G. L. Berlin. 1992. Fundamentals of Remote Sensing and Airphoto Interpretation, 5th Edition. Prentice
Hall, Upper Saddle River, New Jersey. Pp. 472.
Bray, W. L. 1905. Distribution and adaptation of the vegetation of Texas, Bulletin of the University of Texas, no. 82,
Scientific Series no. 10. Pp. 108.
Forman, R. T. T. and M. Godron. 1986. Landscape Ecology. John Wiley and Sons, New York, NY Pp. 620.
Forman, R. T. T. 1997. Land Mosaics: The Ecology of Landscapes and Regions. Cambridge University Press, Cambridge,
United Kingdom Pp. 632.
Frye, R. G., K. L. Brown and C. A. McMahan. The Vegetation Types of Texas, Texas Parks and Wildlife Department,
1984.
Green, L. R. 1976. Factors Affecting Detection of Red Imported Fire Ant (Solenopsis invicta Buren) Mounds With
Aerial Photography, Ph.D. Dissertation, Texas A & M University, College Station, TX. Pp. 158.
Greenburg, L., D. J. C. Fletcher and S. B. Vinson. 1985. Differences in worker size and mound distribution in monogynous
and polygynous colonies of the fire ant (Solenopsis invicta Buren), Journal of the Kansas Entomological Society,
58 (1) pp. 9 - 18.
Gould, F. W. 1975. Texas Plants, A Checklist and Ecological Summary, The Texas A & M University System, The Texas
Agricultural Experiment Station, College Station, TX. Pp. 121.
Lynch, K. M. 1996. A New Landscape Ecology Mapping Scheme for Coastal Environments: Galveston Island, Texas,
Master's Thesis, Texas A & M University, College Station, TX. pp. 113
Martin, J. B., B. M. Drees, W. E. Grant, E. K. Pederson, C. L. Barr and S. B. Vinson. 1998. Foraging range of the
polygynous form of the Red Imported Fire Ant, College Station, Texas. Southwestern Entomologist 23 (3)
pp. 221 - 228.
McMahan, C. A., R. G. Frye and K. L. Brown. The Vegetation Types of Texas, Including Cropland.
Snook Digital Orthometric Quadrangle, Texas Orthoimagery Program. 1m, 24-bit Color Infrared.
