97th ESA Annual Meeting (August 5 -- 10, 2012)

COS 124-9 - Quantifying edge and matrix permeability for a terrestrial wetland specialist

Thursday, August 9, 2012: 10:50 AM
B114, Oregon Convention Center
Scott A. Cooney, Cooperative Wildlife Research Lab, Southern Illinois University Carbondale, Carbondale, IL, Eric M. Schauber, Cooperative Wildlife Research Laboratory, Department of Zoology, Southern Illinois University, Carbondale, IL and Eric C. Hellgren, Department of Wildlife Ecology and Conservation, University of Florida, Gainesville, FL
Background/Question/Methods

The marsh rice rat (Oryzomys palustris) is specialized to wetland habitats that are often isolated by unsuitable uplands.  As a state-threatened species, southern Illinois populations of marsh rice rats may exist in a metapopulation structure that persists through inter-patch dispersal.  The likelihood of rice rats moving between wetlands through the matrix depends upon the permeability of upland cover types. The goal of this study was to quantify the edge and overall matrix permeability of 3 land cover types surrounding rice rat habitat (grassland, agriculture, and forest).  We collectively marked rice rats in wetland patches and extended traps <100 m into unsuitable cover types.  We used the slope of marked capture rates by the distance of capture as our measure of permeability such that less negative slopes indicated higher permeability.  We also estimated monthly population numbers and measured water depth and vegetative cover levels within the habitat and matrix.  We compared permeability levels between cover type, population abundance, habitat inundation and vegetative cover in a repeated measures linear regression to determine which variables predicted matrix use in our population.

Results/Conclusions

Rice rats were captured more often in agriculture (N=23) than grassland (N=11) or forest (N=9) cover, which corresponded to a higher overall permeability (regression slope of capture-rate-by-distance graphs) for agriculture (estimate=-0.02, p<0.05) than either grassland (estimate=-0.08, p<0.05) or forest (estimate=-0.05, p<0.05).  Grassland cover exhibited the highest edge captures (<15 m into matrix, N=9), while matrix captures (25-95m into matrix) could not be distinguished by land cover type.  Abundance (estimate=0.35, p<0.05) and vegetative cover <0.5 m high (estimate=0.03, p<0.05) were positively correlated with rice rat captures in the overall matrix, indicating higher matrix use during peaks in ground cover and population density.  Habitat water depth had a positive effect on edge captures (estimate=8.76, p<0.05), which may demonstrate rice rats using the matrix as a refuge during wetland inundation.  High permeability levels in agriculture can predict higher dispersal rates between habitats isolated by crop fields, while rice rats may enter grasslands for temporary foraging.  This study was the first of any kind to detect rice rats moving through agriculture, indicating that further work is needed to study these animals interacting with human-modified landscapes.