Streambank Erosion from Grazed Pastures, Grass Filters and Forest Buffers Over a Six-Year Period.

In agricultural landscapes, streambank erosion, as a source of non-point water pollution, is one of the major contributors to stream habitat degradation. Streambank erosion rates from riparian forest buffers, grass filters and grazed pastures (stocking rates ranged from 0.23 to 1.15 cow-days ha^-1 m^-1 yr), in three landform regions of Iowa, were measured.  Erosion pins were measured seasonally, except in winter, over a six year period. Severely eroded streambank areas, soil bulk densities and total streambank soil-P concentrations were also measured to calculate total soil and P losses from stream banks. Buffers and grass filters had significantly lower soil loss (respectively, ranged from 4 to 20 tons km^-1 yr^-1, p= 0.002 and 0 to 73 tons km^-1 yr^-1, p= 0.001) from stream bank erosion than did grazing pasture practices (92 to 376 tons km^-1 yr^-1). Similar significant trends among these practices were also found for soil-P loss, severely eroded bank area and erosion rate as well. From the three treatments only the pasture practices had a correlation between seasonal erosion rate and amount of precipitation as it directly contributes to discharge (p= 0.0001; R² = 0.42). This suggests that precipitation recorded from the closest weather station in these pasture sites can be used to account for a significant portion of the streambank failure that took place under different frequencies and intensities that were not directly observed in this study. It can also be suggested that significant correlation between precipitation and erosion rate from grazed pastures could be an indication of low streambank soil resistance to stream discharge which, in turn, increased soil loss.