Natural control of Helicoverpa armigera in smallholder crops in East Africa.

The African bollworm, Helicoverpa (=Heliothis) armigera , is one of the worst agricultural pests in Africa, attacking a variety of food and cash crops. For development of sustainable pest management, it is essential to study the ecology and natural mortality factors of the pest, and recently, the need for the assessment of the role of natural enemies, through life table studies, has been stressed in a number of workshops that focussed on the pest.Information available on the natural enemies of H.armigera, in Africa is reviewed, using published, unpublished and museum sources (Chapter 3). A large variety of natural enemies is represented in almost 300 host records, including 83 parasitoids identified to species and 93 identified only to genus. The taxonomy, distribution, biology, alternative hosts or prey, host plant associations and secondary natural enemies are detailed for all recorded natural enemies, and the different aspects are summarized and evaluated for the total natural enemy complex.During the three-year research programme reported here, H.armigera was studied in four crops commonly grown in smallholdings in Kenya: cotton, sunflower, maize and sorghum. The incidence of the pest varied widely between seven experimental sites in different agricultural zones of Kenya (Fig. 4.1). H. armigera , only occasionally reached damaging levels. A number of parasitoids was recorded, but their impact on H. armigera was generally low; Trichogrammatoidea spp. egg parasitoids and Linnaemya longirostris, a tachinid late-larval parasitoid, were the most common species. Two groups of predators were predominant throughout Kenya: Anthocoridae (mainly Orius spp.) and ants ( Pheidole spp., Myrmicaria spp. and Camponotus spp.), but their abundance fluctuated widely between sites (Fig. 4.3). Pathogens were scarce and did not play a significant role.In-depth studies on life tables and assessment of predation were conducted at Kakamega and Kibos, both in western Kenya. Oviposition of H. armigera coincided with early flowering of the crops (Fig. 5.4), except for cotton, where the period of oviposition was extended (Fig. 12.6). Life tables showed that immature mortality was generally high (82-99.3 %) on sunflower, maize and sorghum, but stage-specific mortality varied greatly between seasons (Tables 53-5.5). Mortality of young stages was highest in maize (Fig. 12.2). Key factor analysis demonstrated that predation-and-unknown-mortality of both of young and late developmental stages was the most important mortality factor (Fig. 12.3).Analysis of temporal and spatial association between numbers of pest stages and predators revealed that anthocorids are generally poorly associated with H.armigera, eggs on sunflower and sorghum (Chapter 6). This partly explains the relatively high survival of young stages on sunflower. On maize, the association of anthocorids with eggs and larvae of H.armigera, was stronger. Ants were better associated with H.armigera on sunflower than on maize or sorghum, which might be responsible for the relatively high late-larval mortality on sunflower.Parasitoids attacking this polyphagous pest are not equally common in different crops, but showed strong associations with particular crop species fed on by their host. In western Tanzania, a complex of three species attacked H.armigera, larvae predominantly on sorghum, whereas parasitism by two other species concentrated on cotton (Fig. 7. 1).Techniques to assess the apparent and irreplaceable mortality of H. armigera, due to predators, parasitoids and pathogens are discussed (Chapter 2). The role of predation and other, unknown mortality factors was studied in two crops where H. armigera , causes most damage: sunflower (Chapter 8) and cotton (Chapter 9). Mortality was measured from stage-frequency data on the pest, supplemented with data on recruitment of eggs onto plants, in plots with and plots without predators. Ants were excluded by banding every plant in a sub-plot with insect trap adhesive. Anthocorids were excluded by applying low concentrations of insecticide which killed the predators but not the pest.On sunflower, survival was higher than in previous trials. Exclusion of predators did not significantly affect survival of the pest (Table 8.2), however, anthocorids, which attack eggs but not larvae of H. armigera, appeared well after the oviposition peak of H. armigera. Moreover the density of ants was very low this particular season.On cotton, survival was extremely low; only 6 % reached the second larval stage (Fig. 12.9), and exclusion of predators did not increase survival. Thus unknown mortality factors were very important and masked the effect of predation. This background mortality appeared to be related to the poor hostplant condition, and the low number of feeding sites (fruiting parts) for larvae due to drought.To evaluate predators at greater pest densities, a series of predator exclusion cages and open control cages were inoculated with H. armigera eggs. Two weeks after inoculation, larval levels in the exclusion cages were 4 1/2 times greater than those in the control (Table 10.1), indicating that predators are capable to suppress pest numbers. To evaluate mortality during the egg stage, we exposed marked egg cohorts on plants. Within 48 h, anthocorids sucked 12-65 % of the eggs, an additional 15 % was lost and 6 % parasitized (Table 11.1).Implications of the findings for IPM, and areas for follow-up work are discussed (Chapter 12). Studies on intercropping cotton with maize or sorghum are promising and most feasible, because maize and sorghum may strongly affect natural enemy populations and pest infestation levels, while such methods stimulate sustainable agriculture in smallholdings. However, there will only be a brief period when the trap crop is attractive to ovipositing moths. A careful choice of varieties and planting dates might ensure the maximum effectiveness of trap crops in the case of sunflower. For cotton, where oviposition is extended over a period of three months (Chapter 9), planting of trap crops at regular intervals may be required, but a trap crop may be most crucial early in the season, because of its potential role to attract natural enemies into fields