Time-structured models of population growth in fluctuating environments
Abstract
1. Although environmental variability is expected to play a more prominent role under climate change, current demographic models that ignore the differential environmental histories of cohorts across generations are unlikely to accurately predict population dynamics and growth. The use of these approaches, which we collectively refer to as non time-structured models or nTSMs, will instead yield error-prone estimates by giving rise to a form of ecological memory loss due to their inability to account for the historical effects of past environmental exposure on subsequent growth rates. 2. To address this important issue, we introduce a new class of time-structured models or TSMs that accurately depict growth under variable environments by splitting seemingly homogeneous populations into distinct demographic cohorts based on their past exposure to environmental fluctuations. By accounting for this cryptic population structure, TSMs accurately simulate the historical effects of environmental variability, even when individuals exhibit different degrees of phenotypic plasticity. 3. Here, we provide a conceptual framework, the mathematical tools needed to simulate any TSM, and a closed form solution for simple exponential growth. We then show that traditional nTSMs yield large errors compared to TSMs when estimating population dynamics under fluctuating temperatures. Overall, TSMs represent a critical tool for predicting population growth in a variable world.