PS 81-140 - Wildfire caused changes in forest structure increase reburn severity

Friday, August 16, 2019
Exhibit Hall, Kentucky International Convention Center
Yi-Chin (Sunny) Tseng, Department of Forest Resources Management, University of British Columbia, Vancouver, BC, Canada and Bianca N.I. Eskelson, Forest Resource Management, University of British Columbia, Vancouver, BC, Canada
Background/Question/Methods

Given regional increases in wildfire frequency in western North America, understanding wildfire behaviour has become a crucial topic for assisting post-fire management. The shift in the vegetation structure after fires results in unique post-fire fuel conditions. How these post-fire fuel conditions influence the reburn severity is of particular interest as it may guide post-fire activities to prevent subsequent severe fires. A wildfire in 2017, the Hanceville-Riske Creek wildfire in British Columbia, Canada, provided a unique opportunity to study reburn severity. A total of 70 permanent plots were burned in 2009/2010 fires, and all plots were burned again in the 2017 Hanceville-Riske Creek wildfire. All plots were surveyed before and after the fires. Specifically, the shift of vegetation structure caused by previous fires was quantified by six plot level vegetation variables. Furthermore, the severity of the Hanceville-Riske Creek wildfire was categorized by field measured soil fire severities (i.e., five categories from 0 to 4). We used ordinal logistic regression to investigate the relationship between the change in vegetation structure caused by previous fires and the soil fire severity in the Hanceville-Riske Creek wildfire.

Results/Conclusions

The change in dead tree density, live aspen basal area, and live conifer basal area were identified as significant predictor variables of reburn severity using ordinal logistic regression. The final ordinal model met the proportional odds assumption (Brant Wald test p-value > 0.05) and was significantly different from the null model (likelihood ratio test p-value < 0.05). The change in live conifer basal area was the most important predictor variable in the model (i.e., had lowest p-value) given the other two predictor variables in the model. This implies that the amount of dead conifers left after fires could potentially be a key indicator of reburn severity. Furthermore, all three significant variables were positively related with the reburn severity. Specifically, the higher the change in vegetation structure caused by previous fires, the higher the reburn severity.