Incorporating variability in simulations of seasonally forced phenology using integral projection models. Issue 1 (26th November 2017)

• Record Type: Journal Article
• Authors: Goodsman, Devin W.; Aukema, Brian H.; McDowell, Nate G.; Middleton, Richard S.; Xu, Chonggang
• Is Part Of: Ecology and evolution. Volume 8:Issue 1(2018)
• Page Start: 162
• Page End: 175
• Publication Date: 2017-11-26
• Subjects: bark beetle -- individual‐based model -- insect -- phenology model; Ecology -- Periodicals; Evolution -- Periodicals; 577.05
• DOI: 10.1002/ece3.3590
• Languages: English
• ISSNs: 2045-7758

Abstract

Abstract: Phenology models are becoming increasingly important tools to accurately predict how climate change will impact the life histories of organisms. We propose a class of integral projection phenology models derived from stochastic individual‐based models of insect development and demography. Our derivation, which is based on the rate summation concept, produces integral projection models that capture the effect of phenotypic rate variability on insect phenology, but which are typically more computationally frugal than equivalent individual‐based phenology models. We demonstrate our approach using a temperature‐dependent model of the demography of the mountain pine beetle ( Dendroctonus ponderosae Hopkins), an insect that kills mature pine trees. This work illustrates how a wide range of stochastic phenology models can be reformulated as integral projection models. Due to their computational efficiency, these integral projection models are suitable for deployment in large‐scale simulations, such as studies of altered pest distributions under climate change. Abstract : Phenology models are important tools for forecasting the effects of climate change on ecosystems. We derive integral projection models of phenology that are deterministic but which retain the effects of stochastic rate variation and seasonal forcing. The resultant integral projection models are useful for integration in large scale earth system models due to their computational efficiency.