PREDICTING LOCAL FISH SPECIES RICHNESS IN THE GARONNE RIVER BASIN

The aim of this work was to predict local fish species richness in the Garonne river basin using three environmental variables (distance fro m the source, elevation and catchment area). Commonly, patterns of fis h species richness have been investigated using simple or multi-linear statistical models. Here, we used backpropagation of artificial neura l networks (ANNs) to develop stochastic models of local fish diversity . Two independent data collections were used, the first one to build a nd test the model; the second one to validate the model. Correlation c oefficients between observed values and predicted values both in the t esting and the validation procedures were highly significant (r = 0.90 4, P < 0.001 and r = 0.822, P < 0.001, respectively). The ANN model ob tained using only three environmental variables succeeded in explainin g ca 70 % of the total variation in local fish species richness. Throu gh these findings, ANNs can be seen as a powerful predictive tool comp ared to traditional modelling approaches. ((C) Academie des sciences/E lsevier, Paris.)