Abstract
BACKGROUND: Insecticide resistance is seriously undermining efforts to eliminate malaria. In response, research on alternatives to the use of chemical insecticides against adult mosquito vectors has been increasing. Fungal entomopathogens formulated as biopesticides have received much attention and have shown considerable potential. This research has necessarily focused on relatively few fungal isolates in order to 'prove concept'. Further, most attention has been paid to examining fungal virulence (lethality) and not the other properties of fungal infection that might also contribute to reducing transmission potential. Here, a range of fungal isolates were screened to examine variation in virulence and how this relates to additional pre-lethal reductions in feeding propensity. METHODS: The Asian malaria vector, Anopheles stephensi was exposed to 17 different isolates of entomopathogenic fungi belonging to species of Beauveria bassiana, Metarhizium anisopliae, Metarhizium acridum and Isaria farinosus. Each isolate was applied to a test substrate at a standard dose rate of 1×109 spores ml-1 and the mosquitoes exposed for six hours. Subsequently the insects were removed to mesh cages where survival was monitored over the next 14 days. During this incubation period the mosquitoes' propensity to feed was assayed for each isolate by offering a feeding stimulant at the side of the cage and recording the number probing. RESULTS AND CONCLUSIONS: Fungal isolates showed a range of virulence to A. stephensi with some causing >80% mortality within 7 days, while others caused little increase in mortality relative to controls over the study period. Similarly, some isolates had a large impact on feeding propensity, causing >50% pre-lethal reductions in feeding rate, whereas other isolates had very little impact. There was clear correlation between fungal virulence and feeding reduction with virulence explaining nearly 70% of the variation in feeding reduction. However, there were some isolates where either feeding decline was not associated with high virulence, or virulence did not automatically prompt large declines in feeding. These results are discussed in the context of choosing optimum fungal isolates for biopesticide development.