Using wildfires as a natural experiment to evaluate the effect of fire on southern California vernal pool plant communities

Fires in Mediterranean-type ecosystems (MTEs) have been studied widely with emphasis on shrub and grassland vegetation types. Although vernal pools comprise a very small fraction of MTEs, they are important to regional biodiversity due to high local endemism. Fire frequency has been increasing in MTEs and while altered fire regimes have been shown to threaten native shrub communities, their effect on vernal pools is uncertain. Due to the number of at-risk species in this habitat, experiments with potentially harmful effects are problematic. Therefore, we initiated this study to take advantage of two anthropogenic but unplanned fire events. The analysis uses data collected from 2001 to 2009 on a site burned in 2000 and 2003. We analyzed the data in an exploratory framework and applied unadjusted and adjusted models using different parameterizations of the exposure variables. The results did not provide evidence that fire reduced the abundance of native vernal pool species in southern California. There is provisional evidence of a positive but temporary effect of fire on native vernal pool species. Our analysis demonstrates an exploratory analytical approach for use with problematic data sets that can arise when conservation objectives constrain opportunities for experimental studies.