Predicting Species Relative Abundance in Ecological Communities

Ecologists are increasingly appreciating using statistical models to predict aspects of species ecology including their abundance and distribution due to their importance in biological conservation and management practices. The aim of this study is to propose a statistical model that allows predicting previously unknown plant species relative abundance (SRA) in an unsurveyed region based on small sub-samples of the whole community. We apply the model to a biodiversity data set which includes plant relative abundances collected from sub-samples of varied communities in central Europe. The results show that the predicted plant relative abundances in unsurveyed sites are close in value to those in the known sites, reflecting the accuracy and the predictive power of the model in estimating species relative abundance in previously unsurveyd ecological sites. The importance of our model is discussed in relation to conservation biology and management.