Breeding abundance of threatened raptors as estimated from occurrence data

A model derived from the negative binomial distribution (NBD) has been proposed to solve the problem of predicting abundance of species from occurrence maps. The viability of NBD was explored for predicting the breeding abundance of five threatened species of raptor: Bonelli's Eagle Hieraaetus fasciatus, Golden Eagle Aquila chrysaetos, Peregrine Falco peregrinus, Lanner Falco biarmicus and Lesser Kestrel Falco naumanni. First, the accuracy of the NBD was tested in a reference area where the species abundance and occurrence were known through intensive field surveys. Next, an estimation of regional abundance derived from NBD was made for each species. These estimates were then compared to the existing regional data for the five raptors. The spatial distributions of the species were strongly aggregated, with F. peregrinus correctly showing the most widespread area of occupancy. The NBD gave a good approximation of the breeding abundance of the raptors, but tended to overestimate real data, particularly the regional data for falcons. Difficulties in species detection, insufficient sampling effort (F. biarmicus) or data collected over long time spans when population size increased (F. naumanni) may have reduced the NBD's resolution power. The ability of the method to predict local abundances over large areas from readily available presence-absence data, with relatively low fieldwork effort, could have considerable applications in conservation biology. © 2008 The Author.