Conifer seedling demography reveals mechanisms of initial forest resilience to wildfires in the northern Rocky Mountains

Climate warming and an increased frequency and severity of wildfires are expected to transform forest ecosystems, in part through altered post-fire vegetation trajectories. Such a loss of forest resilience to wildfires arises due to a failure to pass through one or more critical demographic stages, or "filters," including seed availability, germination, establishment, and survival. Here we quantify the relative influence of microclimate and microsite conditions on key stages of post-fire seedling demography in two large, lightning-ignited wildfires from the regionally extensive fire season of 2017 in the northern Rocky Mountains, U.S.A. We tracked conifer seedling density, survival, and growth in the first three years post-fire in 69 plots spanning gradients in fire severity, topography, and climate; all plots were limited to within 100 m of a seed source to assure seed availability. Microclimate conditions were inferred based on measurements in a subset of 46 plots. We found abundant post-fire conifer regeneration, with a median of 2,633 seedlings per hectare after three years, highlighting early resilience to wildfire. This robust regeneration was due in part to moderate post-fire climate conditions, supporting high survivorship (>50% on average) of all seedlings tracked over the study period (n = 763). A statistical model based on variables describing potential seed availability, microclimate, fire severity, understory vegetation, and soil nitrogen availability explained 75% of the variability in seedling density among plots. This analysis highlights the overarching importance of fine-scale heterogeneity in fire effects, which determine microclimate conditions and create diverse microsites for seedlings, ultimately facilitating post-fire tree regeneration. Our study elucidates mechanisms of forest resilience to wildfires and demonstrates the utility of a demographic perspective for anticipating forest responses to future wildfires under changing environmental conditions.See "README_for_ClarkWolf_et_al_2022_Data.docx" for metadata information. Funding provided by: Joint Fire Science ProgramCrossref Funder Registry ID: http://dx.doi.org/10.13039/100015231Award Number: 18-01-53These data were collected in 69 stratified-random field plots in two recent wildfires in the northern Rocky Mountains, USA over the first three growing seasons post-fire. Plots spanned gradients in fire severity (unburned, moderate, high) and biophysical setting (via elevation and aspect), and were limited to within 100 m of a live tree to ensure seed availability. Variable-width belt transects were used to sample post-fire conifer seedling regeneration, and permanent 1-m2 subplots were established to monitor the vertical growth and survival of individual seedlings. Plots were revisited annually over the first three years post-fire to measure seedling density within transects and monitor individual seedlings in subplots. Additional plot measurements were collected, including: overstory tree density and basal area, ground cover and canopy cover, and soil nitrogen content and availability. In addition, near-ground microclimate was monitored in a subset of sites using data loggers and used to develop statistical models to predict microclimate conditions across all sites and sample years. The attached files include both raw data and derived data (e.g., plot averages, modeled microclimate estimates); see README file for more information