NITROGEN MANAGEMENT Midseason Nitrogen Fertility Management for Corn Based on Weather and Yield Prediction

approach in North Dakota as well as Minnesota and South Dakota, it may not be suitable for other environNitrogen fertilizer applications for irrigated corn (Zea mays L.) ments where high soil organic matter and high mineralbased on yield goals established before planting may result in underor overapplication of N because of weather-induced variations in yield ization reduce the need for N fertilizer to obtain YOPT. potential from year to year. This study was conducted to develop and Others have stressed the importance of accurately acevaluate a regression model to predict corn grain yield at a midpoint counting for the amount of soil N, particularly latein the growing season based on the current year’s cumulative thermal spring nitrate N, when making fertilizer recommendafactors and N fertility levels. The relationship among early-season tions (Binford et al., 1992; Blackmer et al., 1989; Vanotti growing conditions, N fertility, and corn grain yield under continuous and Bundy, 1994) and have determined relationships corn production and sprinkler irrigation was investigated from 1990– between soil nitrate N and yield. Similarly, Bundy and 1995 in southeastern North Dakota. Fertility levels on small plots Malone (1988) found no increase in yield to applied N ranged from 0 to 224 kg ha 1 applied N. There was a wide range of when soil nitrate levels were high, emphasizing the need cumulative growing degree days (GDD) and evapotranspiration (ET) for soil test data for accurate fertilizer recommendations. during the study period. Least-squares regression was used to develop equations to predict, on 10 July, grain yield based on cumulative ET Previous work indicates that yield goal selection or GDD from 1 May to 10 July and soil N plus applied N. Yields should also be based, to some degree, on soil factors predicted on 10 July corresponded well to individual observed yields (not simply the maximum desired yield) to produce a (r 2 0.80), and predicted optimum yield (YOPT) was highly correlated more site-specific recommendation (Vanotti and Bundy, to observed YOPT (r 2 0.885). The model could provide a season1994; Bundy and Andraski, 1995). Blackmer et al. (1997) specific yield potential used to modify N application during the growrecommend, on a site-specific basis, the use of a lateing season, resulting in fertilizer savings in the extreme years when spring soil test to determine the amount of N available cool early-season weather limited yield potential or fertilizer increases before rapid crop N uptake. They also recommended to take advantage of optimum growing conditions and increase yields. that yield goals no longer be used because much of Including meteorological measurements can improve fertilizer manthe N taken up by the crop is supplied by the soil. agement decisions by providing midseason adjustments to fertilizer recommendations. Additionally, they address weather’s role in modifying N recommendation where heavy spring rainfall may leach plant available N from the root zone or warm conditions may increase mineralization and available N. P N fertilizer management is crucial for both The relationship of corn maturity to air temperature economic and environmental reasons. Inadequate N and cumulative GDD in the Northern Great Plains is fertility results in low yields and lower economic returns well established. Runge (1968) found a high correlation while overfertilization can adversely affect surface and between corn yield and ET (rainfall and maximum daily ground water resources and also reduce net economic temperature). Wienhold et al. (1995) noted no differreturns. In addition, overfertilization may also result in ence in yield between two fertility levels when growing higher emissions of nitrous oxide—an important greenseason temperatures were below normal and observed house gas (Kauppi and Sedjo, 2001). Current North the lowest yields in a year with below-average cumulaDakota recommendations for fertilizing corn are based tive GDD. Several studies have indicated that a given on the yield goal concept. This is done by applying corn variety requires a minimum number of heat units 21.4 kg N ha 1 for each megagram of corn per hectare throughout the growing season to reach maturity (Anexpected (yield goal), reduced by the amount of soil drew et al., 1956; Roth and Yocum, 1997; Sutton and test nitrate N (Franzen and Cihacek, 1996). Timely soil Stucker, 1974). This relationship of yield and air tempertest data and reasonable yield goals are required to ature (in particular, cumulative GDD) has been used apply the appropriate amount of N fertilizer using this in models to predict corn grain yield (Swan et al., 1990; concept. While the yield goal concept is the accepted Bollero et al., 1996; Bauder and Randall, 1982). Taking the prediction one step further, Duchon (1986) used N.E. Derby and F.X.M. Casey, Dep. of Soil Sci., North Dakota State CERES-Maize to predict corn yield at midseason by Univ., Fargo, ND 58105-5638; R.E. Knighton, USDA-CSREES-NRE, using historical weather records to fill in the remainder Washington, DC 20250-2210; and D.D. Steele, Dep. of Agric. and of the year until crop maturity to predict yield. This Biosyst. Eng., North Dakota State Univ., Fargo, ND 58105-5626. Received 25 Mar. 2003. *Corresponding author (nathan.derby@ndsu. method assumed N was not limiting.