Understanding post-fire vegetation recovery in southern California ecosystems with the aid of pre-fire observations from long-term monitoring

Abstract

Aims

Post-fire vegetation recovery is often determined by the similarity of post-burn with unburned sites because of a lack of in situ information on pre-fire communities. The inclusion of pre-fire data can help account for pre-existing differences and explore recovery also in terms of return to pre-fire conditions. We used long-term monitoring data in coastal sage scrub and grasslands to: (a) examine vegetation cover recovery of different functional groups; and (b) determine whether vegetation composition in burned areas has recovered in 4 years after fire with burned to unburned and pre- to post-fire comparisons.

Location

Orange County, California, USA.

Methods

We analyzed long-term vegetation monitoring (2007–2021) data from 39 grassland and 58 coastal sage scrub transects in southern California, including observations before and after the 2017 Canyon 2 fire. Linear mixed-effect models were used to determine whether forb, grass, and shrub covers differed between burned and unburned sites while considering the effects of year and repeated monitoring. We used canonical analysis of principal coordinates to analyze vegetation composition based on burn status and time of sampling.

Results

Whereas vegetation cover in grassland recovered quickly, native vegetation cover in burned coastal sage scrub remained lowered 4 years after fire, though forb and non-native grass cover were higher in some post-fire years. Community composition in burned coastal sage scrub was still in recovery 4 years after fire when compared with unburned or pre-fire composition. Although burned and unburned grassland differed after fire in dominant grass species, inclusion of pre-fire data showed that this was a pre-existing difference.

Conclusions

Coastal sage scrub had not recovered pre-fire vegetation cover and composition by 4 years after fire, whereas grassland cover rebounded quickly, albeit with shifts in composition over time; patterns that were detected only by having pre- and post-fire data from long-term monitoring efforts.