Abstract:
An infiltration capacity and surface erodibility study was conducted six years after forest harvesting in the Oregon Cascades. A portable rainfall simulator was utilized to obtain field measurements on the Coyote Creek and Hi-15 Watersheds during summer and fall, 1977. Seasonal variations were found to occur in infiltration
capacities and surface erodibility. Infiltration capacities increased by 1.4 times from summer to fall, while surface erodibility characteristics, suspended sediment concentration and sediment yield, decreased from summer to fall. Surface limiting conditions during the summer and soil profile controlled conditions in the fall were hypothesized to explain this seasonal variation. Nearly all timber harvesting treatments for each study area
had statistically equal summer infiltration capacities in comparison with adjacent unlogged areas. In addition, summer surface erodibility characteristics on treated areas were typically less than those found on undisturbed areas. Only certain skid trails,
cable log paths and severely disturbed sites such as tractor windrowed and burned areas had substantially reduced infiltration
capacities and increased surface erodibility. However, all areas, including the most severely disturbed, had fall infiltration capacities that exceeded usual and maximum fall precipitation intensities.
Many skid trails and other highly disturbed and compacted areas at Coyote Creek appeared to have greatly recovered since logging six years ago. Freezing/thawing, biological activity, and shrinking and swelling of soils may account for this recovery in
infiltration capacities, surface erodibility and soil properties. Skid trails and severely disturbed areas may partially account for peak flow increases and minor sedimentation the first few years after logging. However, data from this study collected six years following timber harvesting do not support the premise that continued increases in peak flows are caused by changes in infiltration capacities, except perhaps for a tractor windrowed and burned area. Predictive models for infiltration capacity (normally distributed)
and surface erodibility characteristics (requiring normalizing transformations) were not found using regression techniques
because of large amounts of variance. Variation in estimates of infiltration capacities and surface erodibility for individual plots and between study areas was identified.
