Performance of a variable tillage system based on interactions with landscape and soil

Understanding tillage system interaction with landscape variability is important in prescribing appropriate tillage systems that are profitable and environmentally sound. A three-year (1997-1999) study was conducted on a gently sloping, poorly drained lacustrine landscape to evaluate tillage, landscape, and soil interactions on grain yield. Tillage systems investigated were a reduced tillage (RT) system [no-tillage after soybean (Glycine max (L.) Merr), fall chisel plowing after corn (Zea mays (L.) var. mays)], and a conventional tillage (CT) system (fall chisel plowing after soybean and fall moldboard plowing after corn). Fall primary tillage was followed with a pre-plant field cultivation in the spring. Runoff and pollutant losses from the two tillage systems were also measured under a 63 mm h^-1 simulated rainfall. Runoff and pollutant (total solids, chemical oxygen demand, total P, dissolved molybdate reactive P) losses were similar, or lower (6.6, 8.0, 7.7, 5.5, and 4.1 times, respectively) in the RT than the CT system. Tillage system, landscape elevation, and soil type interactions on crop yield varied depending upon whether it was a wet or dry growing season. Using the interactions, soybean yield differences among the modeled fixed-RT, fixed-CT, and variable tillage (VT) systems in a wet year were less than 0.1 Mg ha^-1. During a dry year, corn yield was higher in the RT and the VT systems than in the CT system. When no new purchase of tillage equipment(s) is necessary to implement the RT, VT, or CT system, the modest yield benefits during relatively dry years, plus the improved runoff water quality by using reduced tillage system in all or part of the landscape, would justify the use of RT and VT systems over the CT system in the lacustrine landscape.