Long‐term monitoring of a highly invaded annual grassland community through drought, before and after an unintentional fire.

Questions: (a) How did seedling numbers and species composition change in the first year after a wildfire during drought, relative to pre‐fire variation? (b) Has the community returned to pre‐fire composition after five years? (c) Has the degree of dominance by exotic annual grasses changed? (d) Is there any evidence that drought conditions affected community cover, before or after fire? Location: Exotic‐dominated annual grassland in southern California, USA. Methods: We monitored community cover and native annual forb densities for four years before and four (cover) to five (densities) years after an unintentional fire (fall 2013) coinciding with the spring 2012–2019 California drought. We also measured seedling emergence both before and during the first year post‐fire. We assessed post‐fire changes in cover and density relative to pre‐fire variation, and tested correlations between community cover and annual rainfall measures. Results: Seedling emergence declined strongly after fire for exotic grasses, but remained stable for exotic forbs. Seedling densities of the most common native forbs declined, but several previously‐rare natives increased. Community cover initially shifted towards the exotic forbs Erodium spp., then returned to higher exotic grass densities. Yet the previously dominant Bromus diandrus declined steeply, even as other exotic grasses and some native forbs increased. Up to five years after fire, relative cover and abundance of the most common exotic and native species still differed from pre‐fire composition. Common species were uncorrelated with annual precipitation, but several may have responded to shorter growing seasons. Conclusions: Immediate post‐fire conditions favoured exotic and native forbs over grasses, as predicted. Yet in contrast to many previous studies, the community did not return quickly to pre‐fire composition but showed persistent changes that favoured neither natives nor exotics. Our results suggest post‐fire recovery in this habitat may be contingent on abiotic conditions, with drought one potential explanation for changes. [ABSTRACT FROM AUTHOR]