Impact of insect growth regulators on non-target species, with an emphasis on Coccinellids on citrus, in Swaziland.
This study investigated effects of insect growth regulators (IGRs), recommended for use on citrus in southern Africa, on non-target organisms, particularly species of Coccinellidae in and around three citrus estates in the lowveld region of Swaziland, i.e. Tambuti, Tambankulu and Tunzini estates. Some of the species are important predators of citrus pest insects. The distribution of coccinellids within an agricultural land mosaic and factors affecting this distribution pattern were also investigated. Results showed that coccinellid densities and species diversity were lowest in the natural veld surrounding the orchards. In contrast, citrus orchards had the highest coccinellid densities and diversity. River borders, along the Great Usuthu river, had intermediate densities and diversity. While temperature, relative humidity and ground insolation had no significant effect on coccinellid population densities, the presence / absence of prey (host plant), on the other hand, was an important factor. This was the case both for phytophagous and predatory coccinellid species. These showed close patterns of distribution with their respective host plants or prey species. Economically important species, such as Rodolia spp., were restricted to the orchards, while other species, especially those whose economic role is uncertain, such as Cheilomenes lunata and Scymnus spp., were found in most habitat types. Coccinellid population densities were assessed during four treatment programmes: 1) an integrated pest management (IPM) programme where no IGRs were used, 2) a programme where any of the recommended IGRs were to be applied, 3) conventional pesticide programme, and 4) an orchard which did not receive any chemically treatment over the last four years (at Tambankulu estate only). Significant differences between treatment programmes were observed at all three estates, when only economically important coccinellid species were assessed. Orchards under an IPM programme (with no IGRs) were found to have significantly higher population densities compared to those in IGR-treated or under a conventional pesticide regime. The untreated orchard at Tambankulu, however, had significantly higher coccinellid densities of economically important species and higher overall population densities than any of the managed orchards. Of the three estates assessed, Tunzini had significantly higher population densities of all beneficial coccinellids, (excluding the untreated orchard at Tambankulu). Natural vegetation around Tunzini and Tambuti contributed to the higher coccinellid densities compared to Tambankulu, which was surrounded by other agricultural land. Laboratory and field experiments on two non-target species, the ladybird Chilocorus nigritus (Fabricius) (all stages) and the moth Bombyx mori (Linnaeus) (larval stage) assessed specific IGR effects, if any. Three IGRs, buprofezin, teflubenzuron (both chitin synthesis inhibitors) and pyriproxyfen (a juvenile hormone analogue) were used. All three pesticides are currently recommended for use on citrus in southern Africa and were tested at the recommended dosages. Laboratory experiments showed that all three IGRs have a negative impact on both non-target species. When B. mori received direct applications, there was larval mortality as a result of the failing to complete moulting or dying immediately after moulting. In addition, no larvae were able to pupate after having fed on contaminated leaves. In the case of C. nigritus, larvae fed IGR-treated scale or sprayed with buprofezin suffered significantly higher mortality than controls, while IGR effects on those sprayed with, or fed scale-treated with, pyriproxyfen or teflubenzuron were not significant. No adults however, emerged from any pupae in any of the treatment groups. All three IGRs had ovicidal activity on C.nigritus eggs. Adult fecundity in both field and laboratory experiments was not affected significantly after exposure to any of the three IGRs. In field experiments, the proportion of larvae of the moth and ladybird that developed up to the reproductive adult stage, after exposure to buprofezin, was not significantly different from the control. This was not the case for pyriproxyfen and teflubenzuron. Buprofezin was therefore found to be the least detrimental of the three IGRs tested. Minimal drift was observed when a knapsack sprayer was used. However, there was spray drift up to 32m (the furthest distance assessed here) where commercial sprayers were used. This suggests that pesticide drift from orchards to adjacent areas would have serious implications for biological diversity in the river borders and rivers adjacent to the estates. The sensitivity of the non-target species to the IGRs tested needs serious consideration, as it indicates a broader spectrum of activity for the compounds than what is promoted. Additionally, natural control may be affected. This is because the timing of IGR applications and increasing coccinellid populations coincide, resulting in a reduced pool of coccinellid predators. This study emphasised the importance of an appropriately heterogenous landscape to maximise habitat availability for the coccinellids. Although the economic role of the multihabitat coccinellid species recorded here is unknown, they nevertheless clearly contribute to citrus pest control. Such a mosaic landscape, in conjunction with IPM, with no IGR use, promotes ecological diversity and controls pests with minimum disruption to biodiversity. The use of IGRs in citrus thus needs carefiil reconsideration in light of the non-target effects observed on the two species, especially the valuable predatory species, C. nigritus.