Mathematical Biology Seminar Heather Lynch, Stony Brook University Wednesday, March 27, 2019 3:05 pm LCB 225 Adventures in mathematical biology inspired by a bird's eye view of penguin colonies in Antarctica Abstract: Satellites and drones are radically expanding our ability to study wildlife in the world?s most remote places, and the data acquired is generating some new opportunities for using mathematical models to understand biological processes. In this talk, I will introduce some of the work that my lab is doing to understand the fine-scale spatiotemporal dynamics of penguin colonies. In particular, I will discuss a long running and ongoing effort to understand how biogeomorphological dynamics constrain penguin population growth and generate complex spatial patterning of nests within the breeding colony. Using individual-based models that track the fate of hundreds of thousands of individuals evolving on an empirically-derived three-dimensional landscape created by drone imagery, we can understand the links between spatial patterning and biological processes. We find that declining abundance leads to fragmentation even in a homogeneous environment, which has population-level consequences for reproductive success because predation is biased towards colony edges. Strong edge effects from heterogeneous predation coupled with fragmentation in response to population declines creates a positive feedback cycle that can accelerate population decline. I will explain how nest site fidelity among penguins creates a ``frozen herd" whereby the fluid re-arrangement expected according to Hamilton's selfish herd hypothesis is thwarted. The result is a nest arrangement that is sub-optimal from the perspective of fitness and leads to a non-linear tipping point beyond which declining populations cannot easily recover. This work provides a mechanistic understanding of complex spatial structuring in penguin colonies, provides a link between current spatial patterning and past dynamics, and suggests the possibility of critical collapse in seabird populations. Finally, I will also discuss a number of additional opportunities for mathematical modelling in the area of spatial ecology that arise from the proliferation of high resolution spatial datasets.