A new statistical tool to predict phenology under climate change scenarios

Abstract

Climate change will likely affect the phenology of trophic levels differently and thereby disrupt the phenological synchrony between predators and prey. To predict this disruption of the synchrony under different climate change scenarios, good descriptive models for the phenology of the different species are necessary. Many phenological models are based on regressing the observed phenological event against temperatures measured over a fixed period. This is problematic, especially when used for future predictions because the paradoxical situation could arise that the phenological event occurs before the period over which temperature is measured. Such models are unable to predict population variation in phenology. Here, we propose a 'proportional hazards model' to describe phenology and illustrate it with an example from breeding time in birds. This type of model circumvents the above-mentioned problems and is generally applicable for describing phenology, which is essential when assessing the ecological impact of climate change. [KEYWORDS: climate change ; descriptive models ; egg laying dates ; Parus major ; phenology ; proportional hazards model]