Why Large-scale Climate Indices Seem to Predict Ecological Processes Better Than Local Weather

Large-scale climatic indices such as the North Atlantic Oscillation are associated with population dynamics, variation in demographic rates and values of phenotypic traits in many species. Paradoxically, these large-scale indices can seem to be better predictors of ecological processes than local climate. Using detailed data from a population of Soay sheep, we show that high rainfall, high winds or low temperatures at any time during a 3-month period can cause mortality either immediately or lagged by a few days. Most measures of local climate used by ecologists fail to capture such complex associations between weather and ecological process, and this may help to explain why large-scale, seasonal indices of climate spanning several months can outperform local climatic factors. Furthermore, we show why an understanding of the mechanism by which climate influences population ecology is important. Through simulation we demonstrate that the timing of bad weather within a period of mortality can have an important modifying influence on intraspecific competition for food, revealing an interaction between climate and density dependence that the use of large-scale climatic indices or inappropriate local weather variables might obscure.