Eric Klopfer

Advisor: Tony Ives

PhD Abstract: Ecological and evolutionary consequences of explicit spatial structure in exploiter-victim systems

A large and diverse literature on spatial models in ecology exists. Yet, in many ways the science of spatial ecology remains "the final frontier for ecological theory," with many of the fundamental principles of the science still developing.

One class of spatial model which has been widely used in ecology has been termed "pseudo-spatial models" and classically employs various types of aggregation in studying the coexistence of competing parasitoids. Yet, little is known about the relative effects of each of these aggregation behaviors. Thus, in Chapter 1 I choose to examine three types of aggregation and explore their relative strengths in promoting coexistence of two competing parasitoids. While it is difficult, or impossible, to state a priori which of the three mechanisms has the strongest tendency to promote coexistence, I provided various tools for examining the common statistical properties of these mechanisms that promote coexistence. Providing these tools to examine different mechanisms in a common light serves to unify this class of spatial models.

A striking shortcoming of spatial models in ecology to date is that there is a relative lack of use of spatial models to investigate problems on the evolutionary as opposed to ecological time scale. Consequently, in Chapter 2 I chose to start with a classic problem of evolutionary time scale - the evolution of virulence and predation rates. Debate about this problem has continued through several decades, yet many instances are not adequately explained by current models. In this study I explored the effect of explicit spatial structure on exploitation rates by comparing a cellular automatia (CA) exploiter-victim model which incorporates local dynamics to a metapopulation model which does not include such dynamics. These models showed that in the absence of the model which does not include such dynamics. These models showed that in the absence of the spatial localization, selection always favors higher rates of exploitation, but in the presence of spatial localization, intermediate rates of exploitation may be the evolutionary stable strategy.

One advantage of CA models is that they are defined by simple rules rather than the often complex equations of other types of spatial models. This is an extremely useful attribute when one wants to convey results of models to an audience with an applied bent that is often uncomfortable with hard-to-understand equations. Thus in Chapter 3 I chose to use CA models to investigate some potential problems with biocontrol. Through the use of CA models, I showed that there are spatial phenomena which alter the impact of introduced predators and that these phenomena are potentially important in the implementation of biocontrol programs.

The relatively recent incorporation of spatial models into the ecological literature has left most ecologists and evolutionary biologists without the ability to understand, let alone employ, spatial models in evolutionary problems. In order to give the next generation of potential ecologists a better understanding of these modesl, in Chapter 4 I developed an interactive tutorial in which students are able to explore the most well studied of these models (the evolution of cooperation in a spatial environment).