The implications of seasonal climatic effects for managing disturbance dependent populations under a changing climate

Abstract

Abstract The frequency of ecological disturbances, such as fires, is changing due to changing land use and climatic conditions. Disturbance‐adapted species may thus require the manipulation of disturbance regimes to persist. However, the effects of changes in other abiotic factors, such as climatic conditions, are frequently disregarded in studies of such systems. Where climatic effects are included, relatively simple approaches that disregard seasonal variation in the effects are typically used. We compare predictions of population persistence using different fire return intervals (FRIs) under recent and predicted future climatic conditions for the rare fire‐dependent herb Eryngium cuneifolium. We used functional linear models (FLMs) to estimate the cumulative effect of climatic variables across the annual cycle, allowing the strength and direction of the climatic impacts to differ over the year. We then estimated extinction probabilities and minimum population sizes under past and forecasted future climatic conditions and a range of FRIs. Under forecasted climate change, E. cuneifolium is predicted to persist under a much broader range of FRIs, because increasing temperatures are associated with faster individual growth. Climatic impacts on fecundity do not result in a temporal trend in this vital rate due to antagonistic seasonal effects operating through winter and summer temperatures. These antagonistic seasonal climatic effects highlight the importance of capturing the seasonal dependence of climatic effects when forecasting their future fate. Synthesis. Awareness of the potential effects of climate change on disturbance‐adapted species is necessary for developing suitable management strategies for future environmental conditions. However, our results suggest that widely used simple methods for modelling climate impacts, that disregard seasonality in such effects, may produce misleading inferences.