PARENT SESSION
Poster Session #33: Rare, Threatened, and Endangered Species.
Wednesday, August 7. Presentation from 5:00 PM to 6:30 PM. Exhibit Hall B & C, TCC
35
Habitat associations of vulnerable species: A demonstration using Columbia spotted frogs Rana luteiventris in Utah.
Welch, Nathan*,1, 1 Utah State University, Logan, UT
ABSTRACT- Conservation benefits from information about environmental factors that limit the distribution of vulnerable species, which are characterized by severely restricted distributions and sensitivity to human activities. This study uses a highly quantitative observational approach to identify associations between a vulnerable species, the Columbia spotted frog (Rana luteiventris), and features of its wetland environment. Associations are elucidated by comparing the features of occupied and apparently unoccupied habitat. Surveys during 2000 and 2001 were used to establish frog presence/absence for 64 ponds. Habitat of occupied and apparently unoccupied ponds was described between April and September 2001. General habitat descriptions included pond perimeter, connectivity, and the presence or absence of predatory fish, a woody canopy and woody debris in the pond. Vegetation, depth, water chemistry and temperature, debris and exposed substrate, and bank traits were also sampled using a systematic random design. Stepwise and backward discriminant analyses and classification trees were used to help identify potentially important habitat variables. Selected variables were used to develop explanatory models using discriminant function analyses and classification trees. Models were evaluated by considering model size, unbiased misclassification errors generated using leave-one-out cross-validation, and the biological significance of the variables. Important environmental factors will be used to instruct conservation activities, such as wetland maintenance and restoration, which may be critical to the local persistence of the species.
KEY WORDS: habitat associations, Rana luteiventris, amphibian conservation, multivariate analyses