COS 136-2 - Examining effects of forest fire on woody carbon pools in coastal forests of British Columbia, Canada: Boulder Creek Fire case study

Thursday, August 10, 2017: 8:20 AM
B116, Oregon Convention Center
Kate F. Peterson1, Bianca N.I. Eskelson1, Vicente J. Monleon2 and Lori D. Daniels3, (1)Forest Resource Management, University of British Columbia, Vancouver, BC, Canada, (2)PNW Research Station, USDA Forest Service, Corvallis, OR, (3)Forest and Conservation Sciences, University of British Columbia, Vancouver, BC, Canada
Background/Question/Methods

Forest fires are uncommon in the coastal region of British Columbia (BC), Canada. Minimal information is available on the effects of wildfires on stand carbon dynamics in this area. However, ecologically similar areas in the northwestern United States store substantial amounts of carbon. In 2015, four notably large and severe fires burned in the coastal forest zone, which may be a sign of an increase of fire frequency and size in the region due to climate change. We conducted a case study of the Boulder Creek fire, which burned 7,000ha of submaritime coastal western hemlock forest. Our goals were to examine how woody carbon pools vary between fire severities one year after the fire, and to establish baseline measurements that can be used to examine post-fire dynamics under a changing fire regime in the region. Using Landsat imagery, the burn area was partitioned into unburned control and three fire severity classes. A random sample of ground plots was selected from each class. Carbon pools – stemwood of living trees, snags, and downed coarse woody debris (CWD) – were quantified in each plot. Carbon mass by pool as a function of fire severity was estimated using Poisson pseudo maximum likelihood.

Results/Conclusions

The snag carbon pool was significantly lower on unburned plots (1.48Mg/ha) than on plots that burned at any severity (p<0.0028), with high fire severity areas having 23 times more snag carbon than the unburned areas. There were no significant differences in snag carbon among fire severity classes (p≥0.2467). For live trees, high fire severity areas showed significantly less carbon (0Mg/ha) than unburned plots (30.87Mg/ha) (p<0.0001) as well as plots that burned at low or moderate severity (p<0.0001). CWD carbon was 2.19 (p=0.02) and 2.54 (p=0.0503, suggestive but inconclusive) times higher in low and high severity areas respectively than moderate severity areas (8.35Mg/ha). No difference in CWD carbon was found between burned and unburned plots (17.5Mg/ha; p≥0.1551). Clearly, wildfire has a significant, direct impact on standing carbon pools. Fire effects on CWD are inconclusive, possibly due to CWD being consumed by fire irrespective of severity, but accumulating post-fire due to falling snags. To date, this study is the first fire impact study in coastal BC. It shows that severe wildfires cause significant transitions from carbon in living to dead and downed woody pools, which can have an effect on future carbon storage.