| Summary: | Multitrophic and intraguild interactions influence the success of biological control. The interactions between Acyrthosiphon pisum, three natural enemies (Pandora neoaphidis, Coccinella septempunctata and Aphidius ervi) and the host plant, Vicia faba, were assessed. Volatiles released from aphid-damaged plants had a direct effect on P. neoaphidis indicating they may act as synomones. However, volatiles did not increase efficacy of the fungus suggesting it is not a bodyguard species. Transmission was greatest during plant colonisation by aphids and was not affected by plant condition. Infection by P. neoaphidis had a direct negative effect on the fitness of the aphid through reduced reproduction and early host death. Avoidance of infected colonies by predators and parasitoids could reduce the effectiveness of guilds of natural enemies for biological control. However, at the laboratory scale A. ervi and C. septempunctata did not detect infection and entered and foraged in infected aphid colonies. Aphidius ervi spent longer searching for hosts on plants that had been damaged by aphid feeding (and were emitting aphid-induced species-specific volatiles) and this may increase fungal transmission. Coccinella septempunctata and A. ervi significantly reduced populations of A. pisum when introduced as individual species whereas P. neoaphidis had no effect on aphid population size. Foraging by both C. septempunctata and A. ervi increased the abundance and distribution of P. neoaphidis which may be sufficient to initiate an epizootic. The benefits of increased transmission by C. septempunctata outweighed the fitness costs to the fungus of intraguild predation. Pandora neoaphidis was associated with a decrease in the reproductive success of A. ervi, which was further reduced as the competitive advantage of the fungus increased. This could result in competitive exclusion of the parasitoid. Poly tunnel experiments confirmed that A. ervi did not discriminate between infected and uninfected aphid colonies at this spatial scale. However, A. ervi did not incur a fitness cost from foraging in patches containing the fungus. These results indicate that C. septempunctata and P. neoaphidis may be effective as multi-species biocontrol agents. In contrast, competition between P. neoaphidis and A. ervi may reduce their overall effectiveness as control agents. Further work is required at larger spatial scales and over several generations of both the pest and natural enemy species to confirm these interactions. The implications of these results for the use of P. neoaphidis as part of a multi-species biological control program are discussed.
|
|---|
