Laboratory and field host utilization by established biological control agents of Lantana camara L. in South Africa

摘要

Varieties of Lantana camara (lantana) have been introduced into many countries of the world as ornamental plants and have become invasive weeds in many countries including South Africa. In South Africa, it mostly invades the sub-tropical eastern and northern range. Mechanical and chemical control options are expensive and ineffective. A biocontrol programme was initiated in South Africa in 1961. To date, 22 insect species, and a fungus have been introduced, of these 10, and the fungus have established. Three indigenous lepidopteran species and an exotic generalist pest mealybug are also associated with the weed. The variable success of some of the agents released on L. camara worldwide has been ascribed to a few factors. One important aspect is the large range of varieties encountered in the field. It is therefore essential to be able to predict the possible establishment and impact of agents on many varieties. Laboratory trials on five of the established agents showed clear varietal preferences. In the field, most of the biocontrol agents had limited geographic ranges, linked to altitudinal conditions, as higher populations were recorded at low lying warm summer rainfall areas. A pink and orange flower corolla lobe and throat colour combination and plants with few to medium leaf hairs were most abundant in South Africa. Most of the agent species had individual preferences towards different flower colour combinations, as the agents built up different population levels on varieties in the field, within the suitable geographic region for the insect species. Eight agents preferred smooth leaved varieties, while three preferred hairy leaves, and three had no specific preference to leaf hairiness. Varietal preferences thus did play a significant role in agent populations and accompanied impact achieved in the field. New candidate agents need to be proven specific under quarantine conditions and the results extrapolated to predict specificity in the field, while avoiding potential non-target effects. Many authors have questioned the validity of laboratory host specificity trials. The conventional wisdom is that insects portray a far wider host range in the laboratory than what they would do in the field. In other words, laboratory studies measure the physiological host range of an agent and are conservative and usually don’t reflect the ecological host range of agents in the field. To avoid unnecessary rejections of biocontrol agents, this study has made a retrospective study of the host specificity of agents established in the field. Their laboratory and field host ranges were compared and it was found that virtually all the agents reflect similar or less non-target effects in the field than predicted during multiple choice trials. Of the 14 agents, only one introduced species, Teleonemia scrupulosa, and the indigenous species, Hypena laceratalis and Aristea onychote were able to sustain populations on non-target species in the field in the absence of L. camara. Insect populations on non-target species were much reduced compared to that on L. camara. Furthermore non-target effects were only recorded on plant species closely related to the target weed. The multiple choice trials therefore predict field non-target effects accurately. Predictions of non-target effects of candidate agents can therefore be accurately predicted by laboratory studies, in terms of species likely to be affected and to what extent. One field that need further study though is the impact of non-target effects, especially on Lippia species by L. camara biocontrol agents.