UC Santa Barbara

Species dynamics and interactions in nature often fail to conform to classical ecological models because species interactions can be driven by environmental forcing in complex ways. Appropriate management requires an understanding of how these biotic and abiotic forces jointly drive dynamics, particularly for valuable resources that face the additional pressure of human harvesting. My research uses modern quantitative tools for spatial and time series analysis to help unravel complex ecological dynamics. The first part of the dissertation uses nonlinear time delay embedding to reconstruct and analyze species interactions in a California kelp forest across decades of environmental variability. We show that environmental context greatly alters the strength and direction of species interactions, and therefore equilibrium assumptions about kelp forest community dynamics are likely inadequate for predicting how these systems will respond to environmental change. The second portion of the dissertation investigates a human-wildlife conflict between California sea otters and the southern California red sea urchin fishery at a productive fishing ground. We use Bayesian models that combined fishing and otter predation to reconstruct the past dynamics of the sea urchin fishery and predict the likely effects of future otter population growth. We find that in the past, the urchin fishery rather than otter predation contributed to urchin population decline, but in the near future, continued sea otter population growth will severely reduce sustainable fishery harvest levels. In the final part of the dissertation, I investigate broad spatial patterns in interannual fluctuations in the distributions of two important harvested species in the eastern Bering Sea: snow crab and one of its primary predators Pacific cod. I show how both species distributions respond to environmental gradients related to temperature and depth, but in distinct and often opposite directions, leading to emergent spatial patterns in expected predation risk. In combination, the projects in my dissertation reveal how long term spatial and time series data can be leveraged through modern quantitative methods to better understand and manage dynamic species distributions and interactions.