RUNOFF AND EROSION CONTROL WITH CONSERVATION TILLAGE AND REDUCED-INPUT PRACTICES ON CROPPED WATERSHEDS

The North Appalachian Experimental Watershed near Coshocton, OH was es tablished in 1935 to develop, evaluate, and refine conservation practi ces that reduce runoff and erosion under the hilly, humid conditions o f the northeastern United States. Small (0.5 to 1 ha), single-practice , gaged watersheds comprised of sandstone- and shale-derived residual soils are used to evaluate the interaction of management, climate, and soils. In a 28-year, nine-watershed study, 92% of the erosion occurre d during the corn (Zea mays L.) years of a 4-year corn/wheat (Triticum aestivum L.)/meadow/meadow rotation. These watersheds were moldboard plowed prior to planting corn and cultivation was used for weed contro l. By tilling and planting on the contour and increasing fertility lev els, soil loss was reduced more than 3-fold, but still averaged 4.7 Mg ha(-1) during corn years. Thus, annual production of row crops on a s ustainable basis was not without risk. A 6-year, six-watershed study i ndicated that by using reduced tillage (no-till, chisel, or paraplow) and herbicides, corn and soybean [Glycine max (L.) Merr.] can be grown in rotation with an average soil loss of 0.5 Mg ha(-1) yr(-1), well b elow the stipulated soil loss tolerance of 7.8 Mg ha(-1) yr(-1), if a winter cover crop of rye (Secale cereale L.) followed soybean. Under t hese conditions, however, concentrations of surface-applied herbicides and nitrate in runoff frequently exceeded drinking water standards, p articularly in the first few runoff events after application, and may be a concern. A reduced-input management practice for corn and soybean production with light disking and cultivation for weed control and ma nure and a legume (red clover, Trifolium pratense L.) to supply some o f the nitrogen was implemented to determine if a balance between losse s of soil and purchased chemical inputs could be obtained. In a B-year comparison, soil losses were similar to those under conservation till age, but the risk of yield loss increased due to inability to cultivat e in a timely manner due to weather conditions. Regardless of tillage practice, infrequent, severe storms during years when row crops were g rown caused most of the soil loss from the watersheds. Erosion predict ion models must account for the contribution of such events and manage ment practices must limit erosion caused by these storms if long-term sustainability is to be maintained. (C) 1998 Published by Elsevier Sci ence B.V.