| Abstract: | Nitrogen (N) fertilizer use in cotton (Gossypium hirsutum L.) production is a potential source of nitrate (NO3 ?) contamination of soils, groundwater, and streams. The McConnell–Mitchell plots, a long-term study of cotton responses to N-fertilization and irrigation methods, were utilized to determine the NO3 ?-N in soil cropped to continuous cotton. The McConnell–Mitchell plots had a split-block experiential design. The main blocks of this test were irrigation methods. Each block of plots was irrigated using a single irrigation method for the entirety of the testing. Nitrogen fertilization rates were tested within each irrigation block. The soil NO3 ?-N content of two irrigation blocks, furrow flow (FI) and center pivot (CP), were compared to the dryland (DL) control block. Nitrogen treatments tested within each irrigation block ranged from 0 to 168.0 kg N ha?1 in 33.6-kg N ha?1 increments. Nitrogen treatments were tested for 18 years (1982 through 1999), discontinued for 4 years (2000 through 2003), and resumed in 2004. Soil samples were taken in the early spring (2000 and 2004) to a depth of 1.50 m in 0.15 m increments and analyzed for NO3 ?-N. Soil samples taken in 2004 were prior to any fertilization treatment. Irrigation method was found to influence the distribution of soil NO3 ?-N. Little accumulation of soil NO3 ?-N was observed in either irrigation block or under dryland production when N rates were less than 67.2 kg N ha?1. Distribution of soil NO3 ?-N in the FI block was significantly different with sample depth and N treatment but not the interaction of depth and treatment in both 2000 and 2004. Presumably, the small and close values of the means and the greater variability of interactions compared to main effects precluded significant interactions. Differences in soil NO3 ?-N in the FI block after suspending N treatments for 4 years were similar to those found in 2000, although the soil NO3 ?-N was generally depleted in 2004 compared to 2000. The distribution of soil NO3 ?-N in the CP-irrigated block was dependent on the interaction of sample depth with N treatment in both 2000 and 2004. Soil NO3 ?-N values and differences tended to be too small to be of discernable or practical importance under CP irrigation. The distribution of soil NO3 ?-N in the DL block was dependent on the interaction of sample depth with N treatment in 2000 and 2004. Soil NO3 ?-N was minimal in the three lowest N treatments (0, 33.6, and 67.2 kg N ha?1) in 2000. Greatest amounts of soil NO3 ?-N were found in conjunction with the 134.4 and 168.0 kg N ha?1 treatments both years. Depletion of soil NO3 ?-N was evident in the surface 0.45 m of the 100.8, 134.4, and 168.0 kg N ha?1 treatments under DL conditions in 2004. |
