PARENT SESSION
Oral Session # 58: Rare, Threatened, and Endangered Species III: Reptiles and Amphibians.
Presiding: K Beard
Wednesday, August 6. 1:30 PM to 5:00 PM, SITCC Meeting Room 202.
Local- and regional-scale models of habitat suitability for the Illinois chorus frog (Pseudacris streckerii illinoensis) in isolated wetlands.
Jenkins, David*,1, Kiehl, Kathryn1, Ryan-Schneider, Susan1, 1 University of Illinois at Springfield, Springfield, IL
ABSTRACT- The Illinois chorus frog (Pseudacris streckerii illinoensis) exists in Illinois only in isolated wetlands among sandy soils on the east edge of the Illinois River valley. In order to better identify critical habitat for conservation of this state-listed (threatened) amphibian, we collected a suite of abiotic and biotic data in wetlands to develop a local-scale model of habitat suitability. We also collected GIS data across three counties for regional-scale models of habitat suitability, and sampled subsets of wetlands within the three counties to test the effect of scale on models. We then developed logistic regression models to predict Illinois chorus frog presence/absence, based on the local (in-wetland) data and regional (GIS) data. Among all the abiotic and biotic variables collected in wetlands, the presence of large red water mites was most strongly and negatively associated with Illinois chorus frog presence (coefficient = -12.812, R2 = 0.84, model accuracy = 94%). Wetlands that support large red water mites appear to not support Illinois chorus frog populations. Regional-scale logistic models improved from poor (R2 = 0.2) to good (R2 = 0.8) as spatial scale was reduced from large (> 250 sites across 7000 m) to small (<20 sites across 10's of meters) scales. Landscape heterogeneity interfered with predictive models at large spatial scales and based on GIS data alone. The best predictive models for Illinois chorus frog habitat are based on neighborhood-scale sets of habitats, and will be greatly improved by the inclusion of in-wetland data, such as the presence/absence of red water mites.
Key words: regional GIS, amphibian, habitat suitability model, isolated wetlands