2020 ESA Annual Meeting (August 3 - 6)

COS 171 Abstract - Do two rights make a wrong? Individual and combined mosaics of burn severity in short-interval re-burned forests

Brian Harvey, School of Environmental and Forest Sciences, University of Washington, Seattle, WA and Monica Turner, Department of Integrative Biology, University of Wisconsin, Madison, Madison, WI
Background/Question/Methods

In fire-prone forests, a key mechanism of resilience to fire is spatial heterogeneity in burn severity, as it governs important factors including the spatial distribution of post-fire biotic legacies (e.g., seed sources on mature, unburned trees). Recent studies have documented spatial patterns of burn severity over fire-prone forests in different regions of the world. However, a period of increasing fire activity has led to many forests re-burning in short succession, and the landscape patterns of burn severity in overlapping fires, as well as their resultant effect on twice-burned forest landscapes, have been less explored. In this study, we examined the individual and cumulative burn severity patterns in areas of overlapping fires (two fires occurring between the period of 1984 and 2017) in the US Northern Rocky Mountains. We used field-validated satellite burn severity maps to characterize several landscape metrics of burn severity (e.g., high severity amount, patch size, and configuration) in each fire, as well as the cumulative patterns of fire and remaining unburned forest that resulted from overlaying and summing effects of both fires.

Results/Conclusions

Within the spatial overlap of forests that were burned twice, the amount of stand-replacing fire (areas with ~100% tree mortality) was less in the 2nd fire than in the first. Further, landscape patterns of stand-replacing fire in the 2nd fires were characterized by greater heterogeneity than the first fire; that is, smaller patch size, and greater patch complexity (edge:area). However, the cumulative landscape patterns of stand-replacing fire from both fires, when combined, were more homogeneous than either fire individually. The greater area, larger patch size, and less patch complexity of stand-replacing fire from these overlapping burn severity patterns resulted in substantially more area that was far from seed sources (> 150m from a patch edge where mature unburned trees persisted) than did the landscape patterns of either fire individually. Our findings indicate that two overlapping fires in a landscape can produce individual burn severity patterns that are heterogeneous, and progressively more so with subsequent fires. However, when the individual burn severity patterns of overlapping fires are considered together, they can produce homogeneous twice-burned landscapes that substantially reduce the extent of old, unburned forest. Such landscape mosaics have important implications for forest resilience to changing disturbance regimes as the climate warms.