2002 Denver Annual Meeting (October 27-30, 2002)

Paper No. 9
Presentation Time: 3:55 PM

FIRST YEAR POSTFIRE EROSION IN BITTERROOT NATIONAL FOREST, MONTANA


SPIGEL, Kevin Malone, Geography, Univ of Wisconsin-Madison, 550 North Park Street, Madison, WI 53715 and ROBICHAUD, Peter R., USDA Forest Service, Forestry Sciences Laboratory, 1221 South Main Street, Moscow, ID 83843, kmspigel@students.wisc.edu

Accelerated runoff and erosion are common occurrences following forest fires due to the loss of protective forest floor material. The combustion of organic forest floor material may lead to high percentages of bare soil being exposed to overland flow, raindrop impact, and water repellent soil formation. This study focused on the variability of erosion rates on steep slopes representing two categories of stand density. Burning indices were not different between stands. Twenty-four silt fences were installed to study postfire erosion on a storm-by-storm basis in the Bitterroot National Forest (BNF) of West-Central Montana after the 2000 fire season. Erosion rates were estimated from the volume of sediment trapped behind the silt fence downslope of a contributing area of 100 m2. Collected sediment was weighed onsite and a sub-sample taken for dry weight correction, particle size distribution, organic matter content, and nutrient content. Rainfall intensity was the most significant factor in explaining postfire erosion rate variability. Short duration, high intensity thunderstorms (maximum 10-minute rainfall intensity 76 mm/hr) caused the highest erosion rates (86 t/ha). Long duration, low intensity rain events produced little erosion (0.01 t/ha). Other environmental variables were not statistically significant in individual rainstorm events. Losses of total C and N were correlated with erosion rates, while minimal losses of Mg, Ca, and K occurred. Specific site characteristics were entered into the WEPP model to compare observed field data against annual estimates of erosion predicted by the computer model. Compared to observed field erosion rates, results of this comparison show the model was well within the ±50% accuracy level.