Nitrogen dynamics in conventional and no-tillage corn production following legume cover crop

In central Pennsylvania, nitrogen (N) dynamics in soil and corn (Zea mays L.) following red clover and hairy vetch cover crops were studied. The legumes were double-cropped with winter wheat in 1990 and incorporated into the soil by moldboard plow (CT) or left on the soil surface after herbicide killing (NT) in 1991. An additional NT treatment was mowing the vetch and leaving the residue on the soil surface. Fallow after the wheat harvest was the comparison. Seasonal N uptake in corn topmass and nitrate dynamics in the soil's Ap (0 to 25 cm) and upper B (25 to 45 cm) horizons were studied by periodically sampling corn and soil from selected treatments. The collected data set was used to evaluate the ability of the NCSWAP model to simulate N dynamics in soil and corn.;At corn planting in mid-May, 1991, red clover produced a total of 134 kg N ha$sp{-1}$ and hairy vetch 151 kg N ha$sp{-1}$. In the succeeding corn season, the legume 0 N CT treatments had seasonal corn N uptake similar to the fallow 200 N, seasonal soil nitrate less than the fallow 200 N but greater than the fallow 0 N throughout the season, and corn grain yields of approximately 10 Mg ha$sp{-1}$, close to the area's maximum. In contrast, the legume 0 N NT treatments had corn N uptake lower than the fallow 200 N most of the season, soil nitrate close to the fallow 0 N mid- to late season, and corn grain yields of approximately 8 Mg ha$sp{-1}$. Grain yield in the mowed vetch treatment was less than 5 Mg ha$sp{-1}$, mainly due to poor corn population and severe weed competition.;No-tillage reduced N availability in the treatments with organic N sources throughout the seasons of 1991 and 1992, and with fertilizer as the N source mid- to late season in 1991 but not in 1992. No-tillage reduced fall soil profile nitrate accumulation in all N source treatments both years. Tillage did not affect early season nitrification with fertilizer as the N source in 1991 and 1992.;By adjusting inputs controlling N levels in the active soil organic phase and a solute transport parameter, NCSWAP accurately simulated N availability in the CT with or without fertilizer addition in two years. The model accurately simulated the field-scale nitrification process when ammonium nitrate was the N source in both CT and NT treatments, but did not account for the observed nitrate losses in the fallow 200 N mid- to late season 1991. The model consistently underpredicted total available N in the legume treatments with much of the simulation error occurring early in the season. Simulation of seasonal corn growth produced generally good coincidence and degree of association compared to the observed values, but simulation of seasonal soil nitrate dynamics was much less accurate.