Biodiversity and Remotely Sensed Habitat Types: A Comparison of Montane Meadows in the Gallatin Range, MT and in Grand Teton National Park, WY

Abstract

During the last decade, many approaches to biodiversity analysis have relied on the use of GIS (Geographic Information Systems) and remotely sensed data to categorize habitats, and then predict species assemblages expected to be found in those habitats. For example, Gap analysis uses predictions based on knowledge of the geographical limits of a species' distribution, ecological limiting factors, and habitat preferences (Scott et al. 1993). The goal of Gap analysis is to compare locations of plant and animal habitats to those of existing preserves, thereby identifying geographical gaps in habitat and/or species protection. One problem with this approach to conservation planning is that Gap analysis has not been extensively tested to determine the accuracy of its predictions (Flather et al. 1995). Thus, an important parallel approach to Gap analysis, which we describe here, involves assessing statistical relationships between species distribution patterns and remotely sensed habitat types. For the past several years, we have used plants and butterflies as taxonomic test groups to examine these relationships. Because the plant species with dominant cover play a major role in determining the spectral reflectance patterns recorded by multispectral scanners, we felt that it was imperative to test the relationship between the remotely sensed habitat types and the plant community. Butterfly species were chosen because they are moderately host-specific insects, and their diversity may be correlated with underlying plant diversity. We have found that many of the butterfly and plant species of montane meadow communities show significant differences in distribution among remotely sensed habitat types (Debinski 1996, Jakubauskas et al. 1996). Here, we pose the question whether species-habitat relationships (based on remotely sensed habitat categorization of montane meadows) in one part of the Greater Yellowstone Ecosystem will hold in another area of the ecosystem. The long-term goal of our research is to use known species-habitat relationships to predict species distribution patterns in unsurveyed sites. The test of geographic limits of species-habitat relationships is the first step in our analysis. More extensive data will be collected during 1997-1998, allowing for a more rigorous comparisons.