An integrated analysis of micro- and macro-habitat features as a tool to detect weather-driven constraints: A case study with cavity nesters.

The effects of climate change on animal populations may be shaped by habitat characteristics at both micro- and macro-habitat level, however, empirical studies integrating these two scales of observation are lacking. As analyses of the effects of climate change commonly rely on data from a much larger scale than the microhabitat level organisms are affected at, this mismatch risks hampering progress in developing understanding of the details of the ecological and evolutionary responses of organisms and, ultimately, effective actions to preserve their populations. Cavity nesters, often with a conservation status of concern, are an ideal model because the cavity is a microenvironment potentially different from the macroenvironment but nonetheless inevitably interacting with it. The lesser kestrel (Falco naumanni) is a cavity nester which was until recently classified by as Vulnerable species. Since 2004, for nine years, we collected detailed biotic and abiotic data at both micro- and macro-scales of observation in a kestrel population breeding in the Gela Plain (Italy), a Mediterranean area where high temperatures may reach lethal values for the nest content. We show that macroclimatic features needed to be integrated with both abiotic and biotic factors recorded at a microscale before reliably predicting nest temperatures. Among the nest types used by lesser kestrels, we detected a preferential occupation of the cooler nest types, roof tiles, by early breeders whereas, paradoxically, late breeders nesting with hotter temperatures occupied the overheated nest holes. Not consistent with such a suggested nest selection, the coolest nest type did not host a higher reproductive success than the overheated nests. We discussed our findings in the light of cavity temperatures and nest types deployed within conservation actions assessed by integrating selected factors at different observation scales.