The influence of climate variability on demographic rates of avian Afro-palearctic migrants

Abstract

Climate is an important driver of changes in animal population size, but its effect on the underlying demographic rates remains insufficiently understood. This is particularly true for avian long-distance migrants which are exposed to different climatic factors at different phases of their annual cycle. To fill this knowledge gap, we used data collected by a national-wide bird ringing scheme for eight migratory species wintering in sub-Saharan Africa and investigated the impact of climate variability on their breeding productivity and adult survival. While temperature at the breeding grounds could relate to the breeding productivity either positively (higher food availability in warmer springs) or negatively (food scarcity in warmer springs due to trophic mismatch), water availability at the non-breeding should limit the adult survival and the breeding productivity. Consistent with the prediction of the trophic mismatch hypothesis, we found that warmer springs at the breeding grounds were linked with lower breeding productivity, explaining 29% of temporal variance across all species. Higher water availability at the sub-Saharan non-breeding grounds was related to higher adult survival (18% temporal variance explained) but did not carry-over to breeding productivity. Our results show that climate variability at both breeding and non-breeding grounds shapes different demographic rates of long-distance migrants.

Figure 1

Responses (mean regression slopes ± 95% confidence intervals) of breeding productivity of long-distance migratory birds to (a) spring advancement at the breeding grounds and (b) water availability at the non-breeding grounds (carry-over effect). Results are shown for models with a single variable (single covariate model—filled circles) and for the best model combining one variable of spring advancement and one variable of carry-over effect (best model—empty triangles), if this latter model performed better than the single covariate model (see Table 1 and Supplementary Table S3 for full results). GDD5—growing degree days (accumulated temperature above 5°C); T34—mean temperature in March & April, T56—mean temperature in May & June; Salix caprea, Tilia cordata, Sambucus nigra—date anomaly of 10% leaf unfolding for given plant species (number of days before the long-term mean); AET/PET—ratio of actual to potential evapotranspiration in species’ whole sub-Saharan non-breeding range, its Sahelian part and the part south of the Sahel, respectively (see Supplementary Fig. S1b).

Figure 2

Responses (mean regression slopes ± 95% confidence intervals) of adult survival of long-distance migratory birds to water availability at their non-breeding grounds. Water availability is expressed as a ratio of actual to potential evapotranspiration (AET/PET) in (i) each species’ whole sub-Saharan non-breeding range, (ii) its Sahelian part and (iii) the part south of the Sahel (see Fig S1b). Each variable was tested in a single model.