A predictive model of avian natal dispersal distance provides prior information for investigating response to landscape change

摘要: 1. Informative Bayesian priors can improve the precision of estimates in ecological studies or estimate parameters for which little no information is available. While analyses are becoming more popular ecology, use strongly informative remains rare, perhaps because examples not readily available published literature. 2. Dispersal distance an important parameter, but difficult to measure and scarce. General models that provide prior dispersal distances will therefore be valuable. 3. Using a world-wide data set on birds, we develop predictive model median natal includes body mass, wingspan, sex feeding guild. This predicts well when using fitted independent test set, explaining up 53% variation. 4. this model, predict priori 57 woodland-dependent bird species northern Victoria, Australia. These then used investigate relationship between ability vulnerability landscape-scale changes habitat cover fragmentation. 5. We find evidence woodland with poor predicted vulnerable fragmentation than those longer distances, thus improving understanding phenomenon. 6. The value constructing from existing also demonstrated. When as four example species, reduced 95% credible intervals posterior by 8-19%. Further, should have wished collect avian relate it species' responses loss fragmentation, 221 individuals across would been required obtain same provided general model.