A decade‐long study of repeated prescription burning in California native grassland restoration.

Native bunchgrass communities dominated by Stipa pulchra are widely distributed in California but share dominance with non‐native annual grasses. Restoration of these grasslands focuses on altering the balance of native to non‐native grasses to favor the former. This study investigated the impact of burning on vegetation recovery. In the first postfire year burning showed a 70% reduction in cover of non‐native annual grasses (Bromus diandrus exhibited the greatest reduction) and minimal impact on S. pulchra recovery. In the following 3 years, S. pulchra recovered to levels comparable to controls, whereas the annual grasses remained below control levels until the fifth year. Also, in response to reduced annual grass cover on burned sites several species of non‐native Erodium increased from 10 to 30% relative cover, however, the low growth form of these forbs presented a less competitive threat to bunchgrasses than the non‐native annual grasses, and by the third postfire year returned to near control levels. The rare native geophyte Brodiaea kinkiensis was present throughout these grasslands and was not inhibited by burning treatments. To document the reliability of these patterns a second prescription burn was conducted on these sites 5 years after the first burn and vegetation recovery followed for the subsequent 4 years. Patterns observed after the first burn were duplicated following the second burn. The cover of S. pulchra varied in response to precipitation, with the 95% credible intervals of precipitation parameters overlapping zero, however, the cover of non‐native grasses varied greatly with precipitation and had similar trajectories in unburned and burned plots. [ABSTRACT FROM AUTHOR]