A region-wide assessment of forest wildfire effects on forest attributes in the US Pacific coast states

Background/Question/Methods

Ecosystem responses after fire have been analyzed for individual fires and small groups of fires in the Pacific Northwest and Southwest. A regional analysis of ecosystem recovery after forest wildfire with regards to live and dead trees and down woody detritus is lacking. Forest Inventory and Analysis (FIA) data provide a unique opportunity to analyze the effects of wildfire on forest ecosystem components at landscape-level across a large number of fires. FIA plots measured in California, Oregon, and Washington between 2000 and 2010 that were burned within five years prior to plot measurement were included in this study. Burned plots were spatially matched with unburned plots resulting in 441 blocks of burned and unburned plots. Differences in volume of live and dead wood components between burned and unburned plots were analyzed with mixed effects models where wildfire disturbance, ownership group, elevation, and climate variables were treated as fixed effects and a random block effect was used.

Results/Conclusions

The live tree volume on unburned plots was 7.75 times as large as the live tree volume on burned plots, while the snag volume on burned plots was 5.44 times as large as the snag volume on unburned plots. The amount of down woody detritus was larger on unburned than burned plots for all size classes. Coarse woody detritus (CWD), large fine woody detritus (FWD), medium FWD, and small FWD on unburned plots were 2.27, 1.97, 1.92, and 2.35 times as large as on burned plots, respectively. Wildfire generally shifts the tree volume from live pools into dead pools and consumes most of the FWD, which is confirmed by the decrease of live tree volume, increase of snag volume, and decrease of FWD on burned plots. CWD volume has been reported to stay unchanged for a number of smaller, local fire studies. Our large-scale analysis of burned plots across three Pacific coast states shows a significant decrease of CWD volume on burned plots compared to unburned plots (p < 0.0001). No CWD was observed on 15% and 7% of the burned and unburned plots, respectively. This suggests that significant amounts of CWD are consumed during wildfires after all.