Résumé
The distribution of rarity was predicted for Scarabaeidae and Aphodiinae in France using Generalized Linear Models and environmental variables as predictors. The entire French territory was studied, divided into 301 0.72 × 0.36 degree grid squares, and models were developed using 66 grid squares previously identified as well-sampled. The first factor of a Principal Component Analysis conducted on the matrix of squares described by three rarity indices (number of rare species, sum of rarity scores, and mean of rarity scores) was used as a synthetic rarity value. Finalmodels accounted for 67.3% and 69.2% of total deviance but with high meanjack-knife predictive errors of 77.6% and 45.8% for Scarabaeidae andAphodiinae, respectively. For Scarabaeidae, rarity hotspots corresponded todiversity hotspots. In contrast, rarity and species richness were uncorrelatedfor Aphodiinae. We argue that different types of processes have contributed tocreate and maintain the current distribution of rare species in the two dungbeetle groups. Warm-adapted Scarabaeidae and Aphodiinae rare species thathave experienced a recent range expansion from southern refuges would beexplained by current climatic factors. Contrariwise, those cold-adaptedAphodiinae rare species that recently suffered a range contraction would be less predictable by contemporary environmental variables.