This field study was conducted over a 3-year time period in Louisiana to determine which soybean ( Glycine max L.) tillage practice discharged the least amount of nutrients and sediment from experimental plots after rainfall/runoff events. In addition, tillage effect on soybean yield was investigated. Experimental design consisted of three Louisiana soybean tillage treatments [conventional (CT), stale seedbed (SS), and no-till (NT)] with three replications per treatment. A randomized complete-block design was used for statistical analysis. Each of the nine treatment plots measured 27.1 m by 106.4 m and was equipped with an automatic runoff sampler integrated with a continuously recording flow meter and H-flume. Composite runoff samples were analyzed for ammonium N (NH 4+-N), nitrate N (NO 3--N), total Kjeldahl nitrogen (TKN), total phosphorus (TP), orthophosphorus (ortho-P), total organic carbon (TOC), and total solids (TS). Analyte discharge (kg ha -1) per rainfall/runoff event was calculated using runoff concentrations and total runoff flows (L). Statistical analysis showed that discharge treatment means were highly variable and that tillage practice had little or no effect on total runoff and on the amount of N and P discharged from treatment plots. Treatment differences over the study were nonsignificant for all N and P forms 93% and 61% of the time, respectively. Only 21% of the time was mean treatment total runoff significant ( P?0.05). Stale seedbed and NT practices reduced sediment discharges over segments of the soybean growing seasons. Total organic carbon discharge from the NT plots was significantly greater 42% of the time. Soybean yields were highly variable within and between treatments and strongly influenced by rainfall, disease, and insects.
