A sampling programme for incidence and phenology of H. armigera and its natural enemies was set up at seven different sites throughout the country, more or less representing the major agricultural areas in different ecological zones of Kenya. They extended from the wet and high altitudes of Kakamega and Kisii through the Lake Victoria Basin site at Kibos (near Kisumu), the dry central sites of Mwea Tebere and Makueni to the coastal sites of Msabaha and Mtwapa. Each site consisted of replicated plots with three to five of the following crops: cotton, sorghum, maize and sunflower, bean, pigeon pea. The sampling sites were divided into intensive sites and minor sites. In addition, a series of field experiments were run to demonstrate the role of the different predator groups.
Together, these studies showed how the incidence of H. armigera and its natural enemies varies with season and crop in several parts of Kenya. There were no substantial indications of specificity of important natural enemy groups to particular crops. It was shown that at the population levels found at our sites, parasitoids and pathogens do not play an important role in the population dynamics of H. armigera and the damage it causes to the crops, with the likely exception of trichogrammatid egg parasitoids. Predators clearly did have an impact, but this is variable and can be masked by background mortality. The timing of movement of predators onto the crops is critical for their effectiveness; often they arrive too late to have useful impact.