Forest fires where an intermediate proportion of the fire area burns with stand-replacing effects have been described as “mixed-severity” fires. This term is increasingly used to include fires with a wide range of characteristic patch shapes and sizes, which obscures important variation in the potential for forest regeneration that depends on seed dispersal from live forest edges. To encourage a more biologically meaningful description of stand-replacing fire effects, we develop a simple new statistical metric to describe the rate at which the proportion of high-severity (i.e. stand-replacing) area within a fire decreases with increasing distance inward from high-severity patch edges. This metric, which we term the stand-replacing decay coefficient (SDC), is calculated as P =1/(10^(SDC *Dist)), where P is the proportion of the total high-severity area at a given distance (Dist) inwards from the live forest edge. Smaller values of SDC indicate greater area further away from a live forest edge. We tested whether this metric could be used to characterize changes in spatial patterns of stand-replacing fire in California, specifically testing for decreases in SDC over time, and in fires that are suppressed vs. managed.
The stand-replacing decay coefficient (SDC) distinguished between fires that have the same proportion and area of mapped patches of stand-replacing fire effects, but have different patch shapes and configurations. Among 463 fires examined, we found a significant trend toward smaller SDC values over time, with a mean log-SDC of -4.68 in 1984 (associated with 7.8% of a fire’s stand-replacing effects >120 m from a live forest edge), compared with a mean log-SDC of -4.99 in 2016 (associated with 15.2% of a fire’s stand-replacing effects >120 m from a live forest edge), representing a doubling of stand-replacing effects occurring beyond a realistic seed dispersal threshold. We also found significant effects of fire management on SDC. Fires managed by the National Park Service had significantly larger SDC values than fires managed by the Forest Service (P<0.001) or the California Department of Forestry (P=0.001). Furthermore, fires classified as “wildland fire use” fires had significantly larger SDC values than fires managed for suppression (P<0.001). These management effects were independent of high-severity proportion, indicating that SDC contains valuable information regarding fire effects that is not captured under the traditional metric used to describe mixed-severity fires.