Weather extremes in the Mediterranean winter are associated with reduced apparent survival and delayed initiation of egg-laying in a migratory raptor

Climate change increases the frequency of extreme weather events, to which migratory raptors are particularly vulnerable. We monitored a population of Eurasian Kestrels (Falco tinnunculus) breeding in nest boxes in western Finland (approx. 63°N, 23°E) and applied a capture–mark–recapture scheme to adults encountered at nests from 1985 to 2013 to examine how environmental conditions affect their apparent survival. We included density data of their primary prey (voles), collected during the early breeding season within the study area, as a predictor of annual survival. We also recorded the initiation dates of egg laying from 1985 to 2018 to test whether variables representing winter weather conditions predict laying phenology, with prey density included as a covariate. We found that both frequent and infrequent winter rain were associated with reduced apparent survival, and that vole abundance in the breeding area was negatively correlated with apparent survival. Additionally, we found that frequent precipitation in the wintering grounds led to delayed laying the following spring in years of low vole abundance.

Methods
Breeding adult kestrels were captured at their nests using swing-door traps. Each individual, identified by a unique leg-ring number, was assigned a capture history by arranging successful capture events (denoted as “1”) and capture occasions during which the individual was not detected (denoted as “0”) into a timeline of 29 digits, with each digit representing a capture occasion (year). Captured individuals were sexed and aged into two classes: second-year (“2”) and after–second-year (“>2”).

Timing of egg laying was estimated either by revisiting nests with incomplete clutches or by back-dating from the hatching date, assuming a two-day laying interval and a 30-day incubation period.

Data on Eurasian Kestrels were accompanied by annual abundance indices of their primary prey (voles of the genera Microtus and Myodes), collected in mid-May within the study area. The number of voles trapped with snap traps during one sampling period was totaled and divided by the number of trap-nights to derive an index of prey abundance.