Submitted to: Proceedings of Great Plains Soil Fertility Conference
Publication Type: Proceedings
Publication Acceptance Date: 1/15/2000
Publication Date: 3/1/2000
Citation: Soltanpour, P.N., Schlegel, A., Cardon, G., Waskom, R., Halvorson, A.D. 2000. Regional n fertilizer recommendations for dryland wheat. Proceedings of Great Plains Soil Fertility Conference. 8:335-340.
Interpretive Summary: Research results from 56 dryland, winter wheat-fallow nitrogen (N) fertility studies were evaluated to develop regression algorithms to improve N fertilizer recommendations in the central Great Plains. Relative grain yield (yield/maximum site yield) as a function of N fertilizer rate plus soil NO3-N was used to normalize the data over years. Regression analysis showed that relative yield was a better parameter than absolute yield to develop functional relationships between yield and available N. Economic fertilizer N recommendation rates for different soil NO3-N levels were calculated for different yield goals and different cost/price ratios for use by farmers in the central Great Plains.
Technical Abstract: To improve our N recommendation algorithms for dryland hard red winter wheat, we regressed relative grain yield on fertilizer plus soil NO3-N using data obtained in 33 N-responsive (at the 5% level) out of 56 experiments on wheat following stubble mulch fallow in eastern Colorado and western Kansas. The quadratic response function, R2, was 0.66, and all coefficients were significant at the 5% level. We used maximum yields obtained in 52 experiments to convert our relative yield function to absolute yield functions and calculated economic N rates for each site at different cost/price ratios. The average N fertilization net return was $7.40 per acre over all sites at a cost/price ratio of 0.075 ($0.15/lb of applied N and $2.00/bu of wheat). If the cost of N was $0.30/lb, then the break even price of wheat would be $2.32/bu. Yield response to fertilizer N varied from 3 to 26 bu/ac. Only at one site did applied N reduced the yield by 2 bu/ac. Grain protein of responsive sites increased with fertilizer N plus soil NO3-N and decreased with grain yield, as shown in a multi-linear regression analysis with an R2 of 0.61 and significant coefficients. The grain protein at 100 lb N/ac and adjusted for yield, and at 11.5%, could predict N sufficiency with 73 and 67% accuracy, respectively. We conclude that: i) our relative yield function should be validated for fertilizer recommendations; and ii) grain protein is a good index for N sufficiency.