Document: DAV-3-82-15

A simple model of species viability in stochastic landscapes.

BOUGHTON, D.A.*

USDA Forest Service, Corvallis, OR 97331 USA ¹

Abstract:
Stochasticity is an important agent for extinction in populations and metapopulations. Such models usually assume the habitat itself is static. Yet in nature, habitat patches are not static; they turn over because of disturbance and succession. Here, I examine species viability in model landscapes where habitat patches are dynamic. To simplify the problem, I assume a perfect species with no demographic or dispersal constraints other than carrying capacity. The stochastic patch model that I developed has the following assumptions: 1) Each landscape has three stationary traits: a rate of patch births, a mean patch size (or carrying capacity per patch) and a mean patch lifetime; 2) Patch dynamics have two kinds of stochasticity: Patch stochasticity is due to the discrete nature of patches and organisms. Landscape stochasticity is due to variability in the traits mentioned above; 3) Patch lifetimes are either fixed (in the succession version of the model), or geometrically distributed (in the disturbance version). Numerical simulation showed that extinction risk had a threshold when either patch formation or patch size were reduced decrementally. The position of the threshold (in terms of mean capacity of the landscape) varied over at least three orders of magnitude, as an effect of: 1) whether patch birth rate or patch size was changed; 2) landscape stochasticity in each variable; and 3) whether succession or disturbance was assumed. The analysis suggests some conceptual links between viability concepts and hierarchy theory. The results predict a variety of ways in which disturbance regimes might be designed so as to incorporate species into managed landscapes. In particular, if disturbances are more evenly distributed in time and space, extinction risk may be reduced. However, realistic models are needed to examine the situation more precisely.

Keywords: biodiversity, landscape, patch dynamics, viability, disturbance regimes, succession

This abstract is being presented at: 10:30 AM in session:
Poster Session #5: Landscape Ecology.