Use of climatic data for estimating nitrogen fertilizer requirements of dryland wheat

The effect of moisture stress on N fertilizer needs of spring wheat (Triticum aestivum L.) was assessed usingyield and meteorological data from N x P and N x P x K factorial trials, conducted on farmers’ fields in a summer and a winter rainfall area in South Africa in 1979–84. Three to five rates of N fertilizer between 15 and 135 kg N/ha (plus control) and two to four replicates were used. The data were statistically analysed by a multiple regression yield model using applied N rate, N, and seasonal moisture stress S, as the explanatory variables. A good positive correlation (r = 0.83 S.E. = 0.341 t/ha, d.f. = 48) was obtained between the estimated and the observed yields. The effects of S on Nmaxand Nopt (i.e. the N rates corresponding to maximum yield and an economically optimal rate, respectively), were evaluated. For 0.05 < S < 0.3, the range of S values representativeof dryland cropping conditions, N max was largely independent of S, whereas N opt increased sharply with decreasing S. The results indicated that N-response models which include moisture stress parameters may be used to determine the cropping areas climatically most suitable for wheat cultivation.