Habitat dependency of phenological sensitivity to temperature. A bayesian sigmoïd modelling approach applied to constant ringing effort site

Phenological adjustment is the first line of adaptive response of vertebrates when ancestral seasonality is disrupted by climatic changes. The prevailing response is to reproduce earlier in warmer springs. But not all species and populations achieve to match their breeding phenology to changes of the environment. Phenological sensitivity to warming temperature varies between populations and is affected by environmental conditions such as species habitat preferences. We still know little about environmental factors limiting or promoting the expression of phenological sensitivity mostly because common datasets from nest boxes monitoring programs are not variable enough in term of experimented habitats. Large-scale monitoring allows investigating the determinants of between-site differences in phenological plasticity. We parametrized the evolution in proportion of juveniles with a 3-parameters function to estimate breeding phenology for each capture site. We tested the effect of forest cover and imperviousness on average phenology and its sensitivity to local temperature of two model species for research on climate-induced phenological mismatch : the Blue Tit Cyanistes caeruleus and the Great Tit Parus major. For this, we relied on standardized bird ringing scheme (189 sites monitored at least 3 years during 21 years) and high-resolution satellite images. We expected that the more the local habitat matches to ancestral species requirements (i.e. deciduous forest with limited anthropization), the more tits would rely on some preferred preys (i.e. specialization on tree caterpillars), and therefore, the more they would benefit from timing breeding according to spring temperature (i.e. the environmental cue for caterpillar peak). Since the Blue Tit is more strongly tied to forest habitat that the Great Tit, we expected that habitat-driven modulations of temperature sensitivity of phenology would be stronger is the Blue Tit. Expectations about stronger temperature-dependence of breeding phenology in more forested habitats were verified, with even a lack of evidence of temperature-dependence of breeding phenology in the habitat-generalist tit, the Great Tit. However, the influence of habitat anthropization (indexed by imperviousness) was opposite to our prediction: the more a site was surrounded by human settlements, the more blue tits were sensitive to temperature, and the earlier they bred . The study provided a framework, based on the study of juvenile ratio, which could allow a combined estimation of 3 breeding metrics (phenological timing, breeding success and synchronicity).