Wildfire burn severity differs between riparian buffers and whole stream catchments
On this page:
Wildfire is a characteristic disturbance process affecting Pacific Northwest (PNW) ecosystems, but fire regimes are variable across the region. Increasing wildfire frequency and severity can interact with other anthropogenic disturbances in the PNW to influence aquatic ecosystems already susceptible to climate changes. Quantifying aquatic ecosystem responses to fire requires assessment of burn severity for spatial extents at which wildfire-aquatic linkages are most directly connected. For example, wildfires in riparian buffers are expected to have less severe impacts on soil and vegetation compared to those in upland areas, but proximity to streams likely results in stronger wildfire-aquatic linkages. Therefore, the relationship of burn severity with in-stream ecological responses and recovery trajectories likely varies with the spatial extent at which burn severity is quantified (i.e. riparian buffer vs. whole stream catchment). In this poster, we assess paired differences in burn severity between 100m riparian buffers and whole stream catchments for 85 long-term stream monitoring sites.
Long-term ecological monitoring programs in fire prone regions provide valuable datasets with which to assess stream ecological responses to fire. The Aquatic and Riparian Effectiveness Monitoring Program (AREMP) was initiated in 2002 to evaluate Northwest Forest Plan management practices by collection of long-term, in-stream data at 1,411 study sites. Of these study sites, 376 (27%) have experienced fire within their 100m upstream riparian buffer since 1984 according to raster layers derived from the Monitoring Trends in Burn Severity (MTBS) database. Because many of these sites are nested within drainage networks, we selected the most downstream sites (n = 85) and assessed paired median differences in burn severity between 100m riparian buffers and whole stream catchments in terms of (1) percent of total area burned; (2) percent area of moderate to high burn severity; and (3) percent area of low burn severity. Paired differences in burn severity extents showed total area burned was 2.4% lower and moderate to high severity burns were 5.0% lower in riparian buffers compared to whole stream catchments (both p < 0.001). Low severity burns appeared to be slightly higher in riparian buffers, but this result was less significant (p = 0.11). Ongoing research will use mixed effects models, including nested sites within drainage networks, to account for the influence of watershed size, elevation, slope, aspect, and precipitation on riparian vs. whole stream catchment burn severity differences. Our results support existing literature suggesting that riparian buffers are typically less severely burned than upslope areas. These differences highlight a need to consider wildfire impacts on streams at multiple spatial extents as burn severity and proximity likely influence in-stream biomonitoring responses and decisions around ecosystem management.