Quantifying the Environmental Determinants of Plant Demography

The environment dictates population growth rates (lambda). Rapid global change makes quantifying the roles of environmental stressors on populations a priority. We can scale from environmental effects on individuals to the consequences for lambda by using structured population models. However, collecting the data population modelling is resource intensive. I illustrate how data-driven approaches and experiments can be used to understand the consequences of environmental variation for individual performance and lambda, reducing fieldwork demands. First, I show that crowding effects and habitat quality can be approximated through model-selection and spatial autocorrelations of vital rates respectively. My crowding analysis shows that good habitat quality can mask strong intraspecific competition for the critically endangered carnivorous plant, Drosophyllum lusitanicum. I study the negative responses of British Drosera rotundifolia populations to experimental nitrogen addition. This peatland indicator species varied in responses to treatments and vital rates across sites, highlighting the need to spatially replicate demographic studies. I go on to apply retrospective decompositions to a range of ecological systems, comparing the functional decomposition approach with more common decomposition analyses of life table response experiments. I demonstrate that the functional decomposition approach is a simple, precise way to quantify the contribution of environmental variation and treatments on observed differences in lambda. I build site-specific integral projection models of D. rotundifolia and show that treatment-induced changes in vital rates can have strong interactive effects on lambda. Moreover, treatments that affect single vital rates can have non-additive effects. I use functional decompositions to understand multiple treatment effects on vital rates as an aggregate contribution to a change in lambda. I explain how retrospective decompositions have a useful role in informing population management strategies. This thesis illustrates how we can quantify and disentangle various environmental determinants of vital rates and their contributions to lambda across a range of ecosystems.