UC Santa Barbara

Wildfire return intervals are expected to decrease in shrubland communities across southern California due to increasing anthropogenic fire ignitions and climate change. These shortened fire intervals may initiate a positive feedback, placing native chaparral species at risk of replacement by alien annual grasses. This shift is predicted because many chaparral species require multiple years to recover from a wildfire disturbance (i.e., to replenish the seedbank or to replenish underground carbohydrate reserves) and if a second wildfire occurs in quick succession, these species may experience a severe population decline or even extirpation as new seedlings or young resprouts are killed.

After generating a wildfire occurrence map, I selected twelve polygons that experienced two wildfires within five years to evaluate vegetation change following a short-interval fire. All polygons were located in Ventura or Los Angeles County, California and spanned a temporal range from 1956 to 2003. These polygons were then compared to adjacent polygons that experienced only one wildfire within the same five-year period. In order to capture prefire vegetation conditions, historical aerial photographs (HAPs) were selected as-close-to before the first wildfire as possible. In order to capture the maximum postfire growth, HAPs were chosen to be no less than six years following the second wildfire. Prefire and postfire images were georectified and used to calculate ground cover using five community types: chaparral, alien annual grass, sage scrub, tree or bare ground/exposed rock.

To determine the role of regional moisture gradients and environmental conditions in predicting vegetation change I investigated aspect, location in relation to the Santa Ana wind corridors, distance from the coast, time since fire, and prefire cover values as potential predictors of the strength and direction of vegetation responses. Results showed no significant differences in vegetation cover from short interval wildfires compared to adjacent single wildfires. Prefire vegetation cover was highly correlated with postfire cover. Chaparral and sage scrub cover showed strong trends in relation to the time since fire, especially following a single wildfire: positive vegetation change for chaparral and negative vegetation change for sage scrub. Location in relation to the Santa Ana wind corridors and distance from the coast were not found to be significant factors of vegetation change. By contrast, ‘aspect’ correlated with significant differences in vegetation cover, regardless of wildfire history, for chaparral and sage scrub communities: chaparral cover declined and sage scrub cover increased after wildfire on north, but not south, aspects. This study did not find evidence of chaparral loss strictly as a result of a single short interval wildfire. I propose that additional factors, such as aspect and initial community cover (extent), may be equally or more important than a single short interval fire when predicting vegetation changes following wildfire in southern California.