Diparopsis watersi (Roths.) is the principal pest of cotton in the Abyan Delta, West Aden Protectorate, an almost rainless area where the crop is sown in August–September on floodland watered by seasonal streams rising on the Yemen plateau. From December onwards, a large proportion of the pupae formed exhibit a facultative diapause, and adults emerging from these are responsible for infesting the following season's crop, the size of the initial population in which depends upon the extent to which this moth flight is intercepted by cotton in a suitable stage for breeding. When initial populations are low, most of the crop is set before the population has greatly increased by short-cycle breeding, shedding of flower buds and young bolls following attack by D. watersi is slight in relation to natural shedding, and little crop loss is attributable to this boll-worm. When a large part of the moth flight is intercepted, initial populations are high, heavy shedding is caused, and varying degrees of loss ranging up to virtual crop failure can result.

Investigations show that, when land is watered, emergence from diapause pupae is almost completed within six weeks; it is thought that watering reduces soil temperature below the high level that appears to inhibit diapause development throughout the summer. The resulting moth flight is largely spent before flower buds become available and the initial population in the crop is low. When land that was under cotton the previous year is not watered, soil temperatures remain high, and this is thought to account for a delay in emergence until temperatures fall, giving rise to a moth flight extending from November to January. Much of this is intercepted by cotton in a condition suitable for breeding by D. watersi, the initial population is high and heavy losses can result. The annual variation in the volume of the floods results in corresponding fluctuations in cotton acreage, a varying proportion of the previous year's cotton land remaining unwatered. The histories of the four cotton crops 1958–59 to 1960–61 indicate the extent to which infestation of the crop depends on this factor. In a year of poor floods following a large acreage in the previous year, infestations tend to be severe, and, conversely, a large crop following a smaller one the year before is normally lightly attacked. Failure to observe a close season may also promote heavy infestation, since it results in the provision of a food supply for the offspring of the main emergence from diapause pupae, which otherwise would not survive.

Endrin has proved very effective in the control of D. watersi, and Sevin also shows promise, but several applications over the two months following the first flower buds are necessary and it is doubtful whether these could be economic. A close season is essential, and the effects of its violation have been seen in recent years. The available flood water should be used more economically, using a system of area rotation of crops whereby cotton would be followed by another crop and that in turn by a dry fallow. The spring floods, which normally run to waste, could thus be utilised, and this would reduce the incidence of unwatered cotton land.