Testing the ability of topoclimatic grids of extreme temperatures to explain the distribution of the endangered brush-tailed rock-wallaby (Petrogale penicillata)

摘要

Aim Many species are susceptible to climatic extremes, yet few fine-scale studies consider the factors that determine the distribution of extreme temperatures at landscape and regional scales. These factors include cold air drainage, canopy cover and topographical exposure to winds and radiation. We used the brush-tailed rock-wallaby (Petrogale penicillata) to test whether innovative topoclimatic grids of extreme temperatures are an important predictor of regional-scale species distributions. Location Hunter Valley region, New South Wales, Australia (31.2-33.4° S, 148.6-153.0° E). Methods We modelled the regional distributions of rock-wallaby observations and colonies using topoclimatic, macroclimatic, topographical and habitat factors. We employed a randomization procedure to reduce spatial clustering of records and divide the data into 10 different training and testing datasets. Models were assessed using the Akaike information criterion (AIC) on the training datasets, the area under the receiver operating characteristic curve (AUC) on the test datasets, and the consistency of the response curves. We compared multiple univariate and multivariate models, rather than producing one \u27true\u27 model, to examine the evidence that different environmental factors consistently influenced the distribution of rock-wallabies. Results The environmental factors that were consistently strongest at explaining the distribution of rock-wallabies were the topoclimatic estimate of extreme cold, the macroclimatic estimate of annual precipitation, and the amount of cleared land within 1600 m. Rock-wallaby colonies were in areas where minimum temperatures were high, rainfall was low, and there was a low proportion of cleared land. Topographical surrogates performed well in univariate models but poorly in multivariate models. Main conclusions We have shown that topoclimatic maps of extreme conditions have the potential to model the regional distributions of at least some species better than indirect surrogates based on topography or macroclimate. Topoclimatic grids are an important tool for regional conservation planning, but practitioners need to place more emphasis on the derivation and accuracy of the climate grids, not just the spatial resolution.