Search

The abandonment of insecticide treatments to achieve sustainable crop production calls for a combination of methods to obtain satisfactory pest control. To this end, net houses and biological control are two promising methods, though we suspected that most natural enemies are blocked outside the net houses. In Kenya, tomato crops are particularly threatened by Trialeurodes vaporariorum (Westwood 1856) and Tuta absoluta (Meyrick, 1917) that are targets of most insecticide treatments in this country. We investigated in an on-farm experiment (i) the abundances of T. vaporariorum and T. absoluta in open fields and in net houses, (ii) the diversity of arthropod natural enemies of these pests, and (iii) the effect of net houses on the natural regulation. To complete, in the laboratory, we checked the capacity of natural enemies to pass through different net types. We identified several natural enemies, mainly the mirid bug Nesidiocoris tenuis (Reuter 1895) that was the most abundant predator for both pests. One parasitoid species was also identified for each pest, i.e. Encarsia formosa Gahan (1924) and Copidosoma sp. for T. vaporariorum and T. absoluta respectively. Net houses reduced drastically the pest populations. Predators were less abundant under net houses, while parasitoids did not seem to be affected by the nets. Encarsia formosa was shown to be able to pass through the different net types, while N. tenuis adults were not. These results raise the potential for augmentative biological control under net houses. Using kairomones to attract natural enemies, or introducing natural enemies under the net houses through inoculative releases, are two approaches to increase the abundance of natural enemies under net houses.

The analysis of net (bottom to surface vertical hauls) and trap samples taken at night shows that net hauls underestimate abundance of all planktonic organisms from nauplii and mollusc larvae to adult calanids. There is no obvious explanation for this. The underestimation probably results from low net filtering efficiency. The mean ratio between trap and net observed densities is 1.46 (1.2 to 1.9) and does not change significantly according to the organisms.