Patterns and drivers of the scale of effect of landscape structure on diurnal raptors in a fragmented tropical dry forest

Abstract

Theoretical models propose that the spatial extent at which landscape structure best predicts species responses (scale of effect, SoE) depends on habitat and dietary specialization, landscape metrics, and response variables. However, empirical support for such models is scarce, especially for apex predators. To determine SoE for diurnal raptors, and test for differences among ecological traits of habitat and dietary specialization, landscape metrics, and response variables. We conducted 1.5 km transect surveys of diurnal raptors at 26 sites in a tropical dry forest from western Mexico. We measured four landscape metrics (forest cover, matrix hardness, forest patch density, edge density) in 16 concentric landscapes (400 to 3400 ha) around each survey site. We then assessed the landscape size at which each landscape metric best predicted each response variable (abundance, species richness, temporal beta diversity). Finally, we tested for differences in SoE among ecological traits of raptors, landscape metrics, and response variables. Landscape composition metrics of forest cover and matrix hardness decreased with increasing landscape size, and were the main landscape predictors influencing diurnal raptors. SoE differed significantly among landscape metrics, being larger for forest cover (2433.3 ha) than matrix hardness (1500 ha). However, SoE did not differ significantly among ecological traits of raptors, or response variables. SoE was mainly driven by spatial metrics, with diurnal raptors being more strongly associated with forest cover measured over larger scales. This supports findings for other taxa and suggests that forest cover influences dispersal success across larger scales.