Prevalent macroecological patterns have been identified across a wide range of ecosystems, and these patterns have proven effective for understanding ecosystems at scales relevant for conservation and management. However, empirical studies and macroecological theory to this point have largely focussed on static patterns for ecosystems in steady state, and there is increasing interest in understanding how these metrics change over time and in respond to disturbance.

In my dissertation, I use the Maximum Entropy Theory of Ecology (METE) as a starting point to predict how ecosystems will respond to disturbance and analyze the corresponding shifts in macroecological patterns. METE uses the principal of maximum entropy to predict various macroecological patterns and has proven effective for ecosystems at steady state, though its predictions appear to fail for disturbed ecosystems. The first chapter of my dissertation studies how deviations from METE predictions can inform us about underlying biology by studying macroecological patterns across land uses of different intensities for arthropods in the Azores. I then look at how we can modify METE to improve its predictions for ecosystems out of steady state. In my second chapter I present a new model that extends the spatial predictions of the theory to include intraspecific negative density dependence. Finally, in my third chapter I discuss my work developing DynaMETE: a new hybrid theory of macroecology that combines the maximum entropy methods of METE with explicit mechanisms to predict how patterns change in time. I present a method for iterating this theory in time, and code that implements this iteration scheme.

Ecosystems are faced with increasing levels of human disturbance from habitat fragmentation, to land management, to climate change. This makes it important to study macroecological patterns out of steady state as we work toward understanding how ecosystems will respond to disturbances at the large scales relevant for conservation and management.