Cotton crop environment, growth and yield in the Sudan Gezira: II. The causes of variation in yield from year to year

Abstract

SUMMARYAt three sites in the Sudan Gezira the variations in seed cotton yields between 1941 and 1961 were due, in the main, to variations in insect pest damage and, at one site only, to a downward trend in soil nitrogen level.At each site seed cotton yield was predetermined by total plant top dry weight before the opening of the first cotton bolls. The cotton crop was sown in August. At Turabi in the northern Gezira and at the Gezira Research Station in the south-central Gezira, variations in plant growth up to early October had little effect on subsequent yield. The yield was determined by growth during October, November and December; the total top dry weight at the end of December was much more important than the pattern of growth by which this weight was attained. At these sites yield was much more closely dependent on the amount of leaf recorded as free from pest and disease damage than on total leaf weight. Damage due to jassid (Empoasca lybica), thrips (Caliothrips spp.) and leaf curl accounted for 69% of the year to year yield variation at Turabi. Damage due to jassid and thrips accounted for 56% of the year-to-year variation at the Gezira Research Station.Severe pest damage was associated with low July plus August rains and high temperatures during the growing season but was not found to be dependent on leaf nitrogen levels.At Hag Abdulla in the southern Gezira, there was a general downward trend in yields over the years as well as variation from year to year consequent on variations in thrips damage. The downward trend inyields was associated with a trend to lower soil nitrate and plant nitrogen levels which resulted in anearlier cessation of leaf growth in recent years than hitherto.Variation in the level of pest attack accounted for 61% of the year-to-year variation in nitrogen response at Turabi, while at Hag Abdulla the response to nitrogen was determined by the deleterious effect of thrips and by a trend to increasing response as natural soil nitrate levels declined.Gezira commercial crop yields were closely correlated with those of the appropriate experimental plots for the years considered (r = 0·859 for the S-type crop, 0·736 for the L-type crop) inspite of some differences in cultural practice. This suggests that the major causes of variation were the same in both cases and that adequate protection against insect pests should do much to prevent the recurrence of seasons of extremely low yield with their serious economic consequences.