Habitat-GIS-based models for ground beetles (Coleoptera: Carabidae) distribution in agricultural landscape

Predicting the distribution of animals is not only important for conservation, but also for assessing the potential impacts of environmental changes. Ground beetles (Coleoptera: Carabidae), which is one of common environmental indicator species, were investigated by pitfall trapping in different habitats during May to September, 2006 in agroecosystems of Northern China. A logistic regression model based on the relationship between habitat pattern metrics and spatially correlated beetles presence/absence data were then introduced. We extrapolated the models in Xinbaozhuang, a traditional village in Yanqing county, Beijing, where with the support of RS and GIS tools, a number of environmental factors and patch metrics could be derived from QuickBird (TM) satellite image of this area. And the predicted carabids spatial distribution maps were generated subsequently. These simulated maps were validated by the field survey in 2007. The results showed high correlation between land use type and species occurrence, and the nature habitats (forest, woodlands) which were less disturbed harbor more species than the farmlands. It may suggest that modern farming or agricultural intensification may lead to the decline of biodiversity and environmental quality in rural landscape. The GIS-based approaches could be used more widely to estimate the distribution pattern of species richness for the purpose of environment.