| Abstract |
Understanding the demographic responses of wild animal populations to different factors is fundamental to make reliable prediction of population dynamics. Both bottom–up processes and top–down regulation operate in terrestrial and marine ecosystems, but their relative contribution remains insufficiently known. In addition, direct weather effects on demographic rates have been overlooked in marine ecosystems and inferences on the demographic effects of environmental drivers were overwhelmingly made from single study sites. Here, we evaluate the relative effects of bottom–up, top–down and weather processes on four vital rates and on population growth rates of a long-lived seabird, the snow petrel Pagodroma nivea, within three different breeding colonies. We used multistate capture–recapture modelling and perturbation analyses from a matrix population model based on a 36-year-long (1981–2017) individual monitoring dataset to quantify the different drivers (predation, climatic and weather covariates) of probabilities of survival, breeding, hatching and fledging according to colony, sex and breeding status of individuals. Results show that bottom–up forces and local weather affected breeding parameters, and that survival was driven by top–down regulation pressure and bottom–up processes. Breeding parameters differed between colonies and survival was sex-specific. Sensitivity analysis revealed that population regulation was mainly driven by bottom–up processes and that top–down processes played a minor role. However, there were major differences between colonies about the importance of how local weather processes affected population growth rate. Our study brings new insights into the drivers of demographic processes in a marine meso-predator, and how these drivers vary according to colonies and individual characteristics. We emphasize the importance of considering multiple study sites to make robust inferences on the effects of environmental drivers on wildlife demography. More generally, robust conclusions about the importance of environmental drivers on demography rely on considering multiple causal effects at multiple sites, while accounting for individual characteristics. |
