Models explaining the fix success rate (FSR) of lightweight-GPS collars tested in stationary sites (N = 59) under different habitats, vegetation configuration and sky availability.

All candidate models represented alternative hypotheses expressed as logistic models. The response variable registered the percentage of successful fixes. Models are ranked from the most explanatory model after Akaike Information Criterion (AIC) diagnosis for small and dispersed sample size (QAICc); K indicates the number of parameters; ΔQAICc the change in QAICc and ω values of weighted analysis.