Abstract
Two case studies are presented here of nitrogen (N) dynamics in potato/maize systems. Contrasting systems were investigated from (1) the highland tropics of Dedza, Malawi in southern Africa and (2) the northern temperate Great Lakes region of Michigan. Formal surveys were conducted to document grower perceptions and N management strategies. Survey data were linked with N budgets conducted by reviewing on-farm data from representative farms in the targeted agroecosystems and simulation modeling to estimate N losses. Potential N-loss junctures were identified. Interventions that farmers might accept are discussed. The Malawi system uses targeted application of very small amounts of fertilizer (average 18 kg N ha-1) to growing plants. This low rate is on the steep part of plant response to N curve and should serve to enhance efficiency; plant growth, however, is generally stunted in Malawi due to degraded soils and weed competition. Very limited crop yields reduce N efficiency from a simulated 60 kg grain per kg N to an actual of ~20 kg grain per kg N (at 40 kg N ha-1 applied). Legume-intensified systems could improve growth potential and restore N use efficiency through amelioration of soil quality and transfer functions and from biological fixation N inputs. In the Michigan system, N efficiency is enhanced currently through multiple, split applications of N fertilizer tailored to plant growth rate and demand. Fertilizer N rates used by growers, however, averaged 32% higher than recommended rates and 40% higher than N removed in crop product. Application of 50 kg N ha-1 to cover crops in the fall may contribute to the apparent high potential for N leaching losses. Careful consideration of N credits from legumes and residual soil N would improve N efficiency. Overall, N budgets indicated 0 to 20 kg N ha-1 loss potential from the Malawi systems and tenfold higher loss potential from current practice in Michigan maize/potato rotations. Best management practices, with or without integration of legumes, could potentially reduce N losses in Michigan to a more acceptable level of about 40 kg N ha-1.