Friday, March 8, 2002
11:30 am, Science Building, Small Auditorium

David Hiebeler

Cornell University

Evolution of Dispersal in Spatially Structured Heterogeneous Habitats

Abstract: Dispersal is a key aspect of an individual's life history, since it determines the context within which the rest of that life history is played out. Also, habitat loss and fragmentation play an increasingly important role in the fate of populations. How should an organism allocate reproductive effort to near versus far dispersal? It depends on the distribution of suitable habitat and on intraspecific competition due to spatial aggregation, among many other things. I have been investigating the evolutionary stability of short-distance versus long-distance dispersal strategies (e.g. wind-dispersed seeds versus clonal growth, or seeds with varying dispersal distances) on heterogeneous landscapes with varying amounts of suitable habitat and varying levels of local spatial correlations in habitat type. The model can also be applied to dispersal in artificial organisms; the Code Red II internet worm which caused so much damage as it spread across the internet late last summer used a mixture of short- and long-distance dispersal when looking for suitable hosts to infect. Finally, I will also mention some work trying to estimate dispersal parameters in Tree Swallows, via statistical models which correct for inherent spatial biases caused by the geometry of finite study areas.

The contact process involves a random population whose members interact, move and reproduce in space. While Ted Harris introduced the isotropic model in the 1970ies to analyze infectious diseases, a theme that perhaps for obvious reasons promises to be of continued importance well into the present century, it also serves as a model for tumor growth, competition, and the propagation of any computer virus and provides a fruitful labyrinth to study phase transitions.

In recent work, I introduced the anisotropic contact process to account for topographic heterogeneities and investigated in which sense the isotropic model's behaviour is typical in the richer landscape of the anisotropic model, especially, near the discontinuous phase transition between global and local survival of the population. We will describe the contact process along with its background, discuss selected results, and mention a number of open questions on this spatial stochastic process.