Following increased yields of cotton in the Sudan Gezira by DDT spraying, during the 1952/53 season 22,300 and 6,700 feddans of X1730A cotton were sprayed once and twice, respectively, in the Gash Delta of eastern Sudan, where nearly 60,000 feddans were grown under controlled flush irrigation. Each spray consisted of 1 lb. technical DDT per feddan and was applied by aircraft in 2 gals, of spray per feddan.

Systematic observations were made on the incidence of cotton pests in 30 observation stations scattered throughout the Gash Delta, such stations being selected more or less at random from cotton sprayed once and twice and unsprayed, ten amongst each of these three treatments.

A single DDT spray applied 50–70 days after sowing gave entirely satisfactory control of the Jassid, Empoasca lybica (de Berg.), and the thrips, Hercothrips fumipennis Bagn. & Cam. and H. sudanensis Bagn. & Cam., throughout the growth of the crop. There was little lasting control of the flea-beetle, Podagrica puncticollis Weise. The second spray, applied 70–90 days later, had little effect on any of these pests which were then present in low numbers.

The incidence of bollworms was observed between December and March, covering the important fruiting period. During these months, over 70 per cent, more larvae of Diparopsis watersi (Roths.) and 40 per cent, more larvae of Earias insulana (Boisd.) were observed on sprayed than on unsprayed cotton. Significantly more larvae of D. watersi were recorded on cotton sprayed twice than sprayed once; conversely, significantly fewer larvae of E. insulana were recorded on twice- than on once-sprayed cotton. Moreover the estimated total number of fruits damaged by or shed in association with bollworm attack between January and March was over 30 per cent, greater, and nearly 30 per cent, more bollworm damage to nearly mature bolls was recorded between December and March, on sprayed than on unsprayed cotton.

Yields of seed cotton were significantly less from twice-sprayed than from the other treatments. Once-sprayed cotton yielded less but not significantly so, than unsprayed cotton. Yield was negatively correlated with the number of sprays but the correlation coefficient just failed to reach significance.

Further analysis of the data from stations where yield differences were most marked indicated that D. watersi lowered the yield potential of the crop as expressed by fruit production and retention, as a result of continuous shedding of the damaged fruit. Attack, however, was concentrated on cotton with the best yield potential. The effect of spraying was to increase the numbers of D. watersi and thus to give rise to excessive shedding, increased production of fruit primordia, an increased number of damaged bolls, and finally reduced yield.

It is concluded that any benefits which the crop enjoyed, as a result of elimination of leaf-feeding insects by DDT spray, were completely lost through increased bollworm attack, which moreover reduced the yield below that of unsprayed cotton.