Quantitative Modeling Across Scales: Behavior, Demography, and Trophic Interactions in Population and Ecosystem Dynamics

Abstract

Ecological dynamics emerge from processes operating across vastly different scales, yet theory and empirical work have often treated these scales in isolation. This dissertation applies quantitative modeling from the scale of the individual social bond up to population dynamics and even to ecosystem stoichiometry. In Chapter 2, I developed a synthetic demographic model of barren-ground caribou (Rangifer tarandus groenlandicus) to investigate the drivers of long-period population cycles. Motivated by Indigenous knowledge of decadal fluctuations in caribou abundance, the model links cohort effects with density dependence and environmental stochasticity. I used wavelet analyses and sensitivity analyses to reveal drivers of sustained oscillatory dynamics, broadening the theoretical scope of population cycling beyond the short-lived species where it is most studied. In Chapter 3, I explored how producer quality, nutrient enrichment, and space limitation interact to shape community dynamics using a two-producer, one-consumer ecological stoichiometry model. Under a Holling Type II functional response, the system undergoes a cascade of bifurcations not observed in classical two-species stoichiometric models, including an abrupt, irreversible transition in which both the preferred producer and the consumer are lost; these results have potential implications for ecosystems undergoing climate-change driven change in vegetation composition. In Chapter 4, I analyzed six years of social association data from a reintroduced population of scimitar-horned oryx (Oryx dammah) in Chad to characterize de novo formation of non-random social structure. Using MRQAP regression and Generalized Affiliation Indices across 18 seasonal networks, I found that gregariousness and experience differences explain a large portion of variation in dyadic associations, yet explanatory power declines over time as residual social modularity strengthens. The emerging modules show no alignment with release group, sex, age class, or range residency, suggesting that innate social preference combines with structural factors and individual life history to shape social landscapes. Taken together, these chapters argue that the processes governing ecological dynamics, including demography, stoichiometric constraint and social preference, are deeply entangled across scales, and that quantitative models parameterized for real systems are essential tools for exploring these phenomena.