Post-fire reference densities for giant sequoia seedlings in a new era of high-severity wildfires

Many forests globally are experiencing increases in large, high-severity wildfires, often with increasingly inadequate post-fire tree regeneration. To identify areas that might need post-fire planting, forest managers have a growing need for seedling reference densities – the natural seedling densities expected to be adequate to regenerate a forest – to compare with observed post-fire seedling densities. The most useful reference densities will meet five criteria: they will (1) be specific to natural post-fire reproduction rather than planted seedlings (because planted seedlings can have substantially greater survival than natural seedlings, thus underestimating adequate natural reproduction), (2) apply to the first few years following fire (when management decisions and actions are most likely), (3) be specific to each of those post-fire years (because post-fire seedling densities can change rapidly with time since fire), (4) be associated with estimates of uncertainty, and (5) include consideration of novel environmental conditions during management applications (because most reference densities will be based on data collected under more environmentally benign conditions). The world’s most massive tree species, the giant sequoia (Sequoiadendron giganteum) of California’s Sierra Nevada, recently experienced historically unprecedented wildfires that killed an estimated 13–19% of mature sequoias across their native range. Seedlings germinating after these fires then experienced exceptional summer heat and the two most severe summer droughts of the 121-year historical record. To help inform management responses to these events, we used seedling censuses from past fires (mostly prescribed fires) to calculate sequoia seedling reference densities meeting the five criteria. The reference densities had three striking features, which are partly attributable to giant sequoia’s status as a pioneer species. First, despite being inherently conservative, the reference densities were quite high. For example, mean first-year reference density was 172,599 seedlings ha−1. Second, reference densities declined precipitously with time since fire: the mean fifth-year reference density was only 5% of the mean first-year density. Third, the reference densities were associated with relatively substantial uncertainty, a consequence of density variations among seedling plots; for example, the 95% credible interval for first-year reference density was 64,377 to 313,438 seedlings ha−1. Despite this uncertainty, a case-study sequoia grove that recently burned in a high-severity wildfire had second-year post-fire seedling densities that were significantly (and dramatically) lower than the corresponding second-year reference density, suggesting inadequate post-fire reproduction. Our results highlight the value of the five criteria for reference densities – criteria that, in current practice, are rarely all met.