Field experiments were conducted in the Coastal Plain and Ridge and Valley regions of Virginia during the 1981-82 through 1985-86 winter wheat growing seasons. The treatments in all experiments consisted of varying amounts of N fertilizer rates applied at Zadoks' growth stages 25 (GS 25) and 30 (GS 30). The research was divided into three studies. The first study was conducted to assess the effect of N fertilizer rates and climatic conditions on the amounts and patterns of the crop N uptake. Dry matter production and total N concentration were measured in total above ground plant material at different growth stages, as well as in leaves, stems, and spikes. Plant N uptake was affected by the climatic conditions in the different growing seasons. Temperature and precipitation variations in early spring determined the differences in amounts and patterns of the N uptake by whole plants and by the various plant portions. Maximum N uptake daily rates were obtained in the period immediately after GS 30 suggesting that this is the wheat growth stage in which the highest efficiency of N fertilizer use could be expected. Crop N uptake at GS 30 also appeared to be a potentially good indicator of the plant N requirements.

The second study was designed to develop models for determining critical N levels and optimum N fertilizer rates for winter wheat. Two nonlinear models were successfully developed to determine critical N levels at GS 30 utilizing plant N concentration at GS 30 (N%30) and crop N uptake at GS 30 (NUP30). The R² values for the models utilizing N%30 and NUP30 as independent variables were 0.87, and 0.82, respectively. Simple regression models were successfully developed to predict N rates required at GS 30 to obtain either maximum or economically optimum grain yields. The models utilized NUP30 as the independent variable and had high correlation coefficients and good predicting ability.

The objective of the third study was to determine the recommended amount of N fertilizer to be applied at GS 25 that would optimize the use of the simple linear regression models developed in the second study. Quadratic programming models were developed with the objective of maximizing marginal profit with N fertilizer application. The models were then solved to determine the amounts of N at GS 25 and at GS 30 that would produce the maximum attainable profit. Four recommended nitrogen fertilizer rates at GS 25 (N25) were evaluated: 0, 30, 60, and 100 kg N ha⁻¹. The difference (D) between the yields with maximum attainable profit (Y) and the yields when N25 was forced to be 0, 30, 60, and 100 kg N ha⁻¹ (Υ^R) was then calculated (D = Y - Y^R). The best N25 recommendation was the one that minimized the mean value, standard deviation, and coefficient of variation of D. This methodology was used for 3 nitrogen fertilizer : wheat price ratios (2.0, 4.0, and 8.0). The results indicated that the best recommendations for N25 were 50 - 60 kg N ha⁻¹ for N fertilizer : wheat price ratios of 2.0 - 4.0, and 40 - 50 kg N haha⁻¹ for a price ratio of 8.0. Sensitivity analysis was then performed to study the effect of variations in the N fertilizer : wheat price ratio on the recommended N rates. The results indicated that the recommended N rates were essentially insensitive to the variations in the price ratio of N fertilizer : wheat