Variation of cotton yields from year to year has been a major problem in the Sudan Gezira for a long time. Previous studies indicated involvement of pest attack, soil nitrogen levels and lengths of growing season in this and suggested that adequate control of these factorswould lead to consistently high yields. Factorial experiments to test this hypothesis, comparing plots sown in early or late August, with or without frequent insecticide spraying and with orwithout nitrogenous fertilizer were therefore carried out at each of five sites in the Sudan Gezira and the first five years’ results are presented.
There were large, positive responses to nitrogen, spraying and early sowing at all sites. Nitrogen and spraying interacted positively, i.e. the response to nitrogen was much greater on sprayed plots and the response to spraying much greater on fertilized plots. Where plots were sprayed, early sowing also increased the response to nitrogen.
Treatment responses varied from year to year in such a way that the optimum combination resulted not only in much higher average yields but also in greatly improved stability of yield from year to year. The average yields, and the response to nitrogen on sprayed plots, were up to the levels which previous regression studies had estimated would be obtained in the absence of serious pest attack and were similar to peak yields in earlier years. During the course of theseexperiments commercial crop practice has changed towards earlier sowing, heavier nitrogen dressing and more frequent spraying, with apparently beneficial effects on yield and consistency of cropping.