CANCELLED - The effects of local weather patterns on Soay sheep dynamics

Background/Question/Methods

Variation in the weather influences herbivores directly, via imposing energetic costs, and indirectly via plant growth and senescence. Variation in plant productivity between seasons and years is rarely quantified in natural ecosystems. We use data from the individual-based long-term study of Soay sheep on St. Kilda, Scotland, to explore Climate-Plant-Herbivore interactions in detail over multiple decades using novel statistical approaches including random forest models. We investigate the influence of climatic drivers, both on forage productivity and on sheep demography to allow parameterization of models of climate change. To reconstruct past local weather, we use statistical approaches to interpolate climatic data using weather station daily observations from nearby Stornoway.

Results/Conclusions

The use of integrating detailed, local climate data revealed important climatic processes that were obscured when large-scale drivers like the North Atlantic Oscillation (NAO) were used. Our results show the winter NAO is not an accurate predictor of the local climate on St. Kilda, and that the effects of the climatic variation on demography is often subtle and complex. We show that the dynamics of Soay sheep are, in fact, influenced by the interaction between strongly non-linear density dependence, food and climate, and our analysis of weather variables marks an important new step towards understanding this system.