Demographic and phenotypic signals of impending population change

A major goal in population ecology is to predict how populations will respond to environmental change. To achieve this goal we must identify the early warning signals of demographic regime shifts. Most studies have focused on phenomenological identification of such signals based on basic population statistics. Recent studies have highlighted rapid phenotypic changes accompanying and often preceding demographic regime shifts, indicating that a trait-based demographic approach can improve our ability to predict a population's response to environmental change and help design management action plans well in advance. Furthermore, these studies show that some species can adapt to environmental change more easily than others through a range of evolutionary and ecological mechanisms; understanding these differences is crucial for identifying species that are more susceptible. In this study, I will investigate the links between the abiotic environment, population density, individual traits (such as body size) and demography, and address the following questions: - How do populations respond demographically, ecologically and evolutionarily to environmental perturbations? - What are the trait and demography-based statistics that can be used as early warning signals of demographic regime shifts? To address these questions, I will use a three-tier approach. I will first investigate the trait-demography relationships and how these change over time using long-term data from three contrasting mammalian systems: Soay sheep, yellow-bellied marmots and prairie voles. Recent studies on these systems have reported phenotypic change preceding, or occurring simultaneously with, a population dynamic shift. Furthermore, there are important contrasts in phenotype-demography relationships. Next, using a perturbation analysis on these three systems, I will simulate population and phenotypic dynamics under multiple scenarios. These scenarios will include alternative demographic, plastic and evolutionary responses to further perturbation. The simulations will yield time-series data from which I will estimate several trait and demography-based statistics. These statistics will be evaluated prospectively using the simulations and retrospectively using the long-term datasets. Finally, using a laboratory microcosm, I will experimentally test the ability of these statistics to predict demographic responses to environmental perturbation. Showcasing these well-studied natural systems, identifying trait and demography-based early warning signals of population change and testing these on an experimental system will provide much-needed predictive insight into how species respond demographically, ecologically and evolutionarily to rapid environmental change.