On the Relative Abundance of Species: Postfire Changes in a Coastal Sage Scrub Rodent Community

Abstract

Mac Arthur's Q—Minimization theory of competition—mediated community structure suggests a protocol for predicting species abundances from knowledge of the availabilities and species utilizations of critical resources. Using this protocol, we compared predicted changes in rodent abundances with those observed over the two and one—half years following a fire in a topographically uniform area of California coastal sage scrub vegetation. Approximately half of our permanent census grid covered an area burned in 1979, and half covered an adjacent unburned area. A separate trapping survey at a nearby unburned site showed that the five rodent species common to the permanent grid had distinct capture frequencies in four structural microhabitats. We used these microhabitat—specific capture frequencies as resource—utilization functions for each species, because structural microhabitats appear to be related to underlying resources that limit rodent populations. Resource—availability functions were derived from censuses of vegetation and other structural features on burned and unburned sides of the permanent grid. Predictions from Mac Arthur's theory were upheld in that (1) abundances of each species on one side of the grid relative to the other correlated significantly with relative K values on the two sides of the grid (K values were calculated as the overlap between microhabitat utilization and availability functions); (2) relative abundances of the five rodent species on the burned side of the grid were correlated with those predicted from relative K values; and (3) the total microhabitat utilization of the rodent community on each side of the grid matched overall microhabitat availability on that side well and matched availability on the other side poorly. These results provide support for a potentially powerful protocol for predicting identities and relative abundances of coexisting species. Although this protocol should provide results similar to those from a more inductive one relying on empirical foreknowledge of spatial correlations between resource and consumer abundances, it differs in important respects.