Predicting mammal species richness and distributions: testing the effectiveness of satellite-derived land cover data

摘要: Mapping species richness and distributions is an important aspect of conservation land use planning, but the time cost producing maps from field surveys prohibitive. It useful, therefore, if mappable environmental variables, a readily accessible source, can be used as surrogates for attributes. We evaluated power satellite-derived cover information, Land Cover Map Great Britain, to predict occurrences terrestrial mammals in one hundred 10×10 km quadrats, four regions Britain. The predictive data was relatively poor – with few exceptions, explained less than half variation mammal occurrence regression models. Predictive considerably stronger when were analyzed separately together, best fitting models varied between taxa. also affected (positively or negatively depending on taxon) PCA-ordinated variables predictors. strength probably limited mostly by high proportion British geographic changing rapidly independently (and hence non-saturation preferred habitats), lesser extent shortcomings data, influence landscape factors other distributions. results suggest that regional stratification essential attempting distributions, even across areas such conclude caution necessary using information systems this basis planning decisions.