Harold Van Es1, Jeff Melkonian1, Jean M. Sogbedji2, Ivy Tan1, Beverly Kay3, R.S. Dharmakeerthi4, and H. Dadfar4. (1) Cornell Univ, 1005 Bradfield Hall, Ithaca, NY 14853, (2) Univ de Lome, B.P. 1515, Lome, Togo, (3) Dept. of Land Res. Science, Univ of Guelph, Guelph, ON N1G 2W1, Canada, (4) Univ of Guelph, Land Resource Science, Guelph, ON N1G 2W1, Canada
In humid regions, N transformations are strongly influenced by dynamic processes that may result in rapid losses, mostly through leaching and denitrification. These losses of N result from complex interactions among weather, soil hydrology, crop water and N uptake, and management practices. Most current recommendations for N management do not directly account for the dynamic behavior of soil N, limiting our ability to more precisely and efficiently manage N. Increasing N fertilizer cost, elevated concerns about water contamination, and increased evidence of greenhouse gas impacts are prompting a new look at N management. We summarized results of five maize N response studies conducted in New York, USA and Ontario, Canada for multiple years at multiple locations, which demonstrate the need for a dynamic approach to N management. All studies involve a spatial and a temporal component in that different soil types and landscape positions were involved and the studies were conducted for multiple years. Results show that high precipitation during the critical late spring period, when maize ET is still low and soil nitrate levels and temperatures have increased, is a strong determinant for N losses through leaching and denitrification, and thereby greatly affects crop N availability. This process may in some cases interact with soil type and landscape position, where finer-textured, poorly-drained soils, or lower-landscape-position soils may experience greater losses. Accurately predicting fertilizer-N requirements therefore requires a dynamic approach that incorporates this complexity. Soil N tests are expensive, primarily due to high labor costs, and high spatial and temporal variability makes their prediction accuracy often unsatisfactory. Model-based approaches are showing considerable promise for accurately predicting this seasonal variation in N dynamics. Due to the critical nature of weather and soil processes, such models need to accurately represent the important processes and be dynamically linked to real-time weather information.