Document: JAN-3-92-12

Inferring process from pattern: Negative frequency-dependence and the importance of spatial scale.

MOLOFSKY, J.* ¹, J.D.BEVER ² and J.ANTONOVICS ³

University of Vermont, Burlington, VT 05405 USA ¹
University of California, Irvine, Irvine, CA ²
University of Virginia,Charlottesville, VA 22903 USA ³

Abstract:
The ability to infer process from pattern has a long history in ecology. Nevertheless, the accuracy with which patterns reflect the processes that give rise to them is hotly debated. This study uses a general spatially explicit model to explore how the scale of dispersal interacts with the scale and strength of negative frequency-dependence to determine patterns of species distribution. Counter to expectation, strong local frequency-dependent interactions result in random spatial patterns. When dispersal scale and interaction scale are decoupled, the resulting patterns are not necessarily random. For strong negative frequency dependence, stable bands result when the scale of interaction exceeds the scale of dispersal, and travelling waves result when the scale of dispersal exceeds the scale of interaction. However, for weaker interactions occuring over intermediate scales, only random patterns result. Thus, our results call into question the utility of inferring any ecological interaction from only the spatial distributions of the putatively interacting species. Correspondingly, for populations that are spatially sub-structured, it is also not clear what the correct null model is for a spatial distribution when there are no species interactions (the neutral case). In this case, the pattern also depends on the scale of dispersal. We suggest that simple spatial models of species interactions can be used to provide the appropriate null model.

Keywords: cellular automata, competition scale, dispersal scale, stochastic spatial model, spatial patterns

This abstract is being presented at: 10:15 AM in session:
Oral Session #39: Theoretical Ecology.