Location: Invasive Plant Research LaboratoryTitle: An abundant biological control agent does not provide a significant predator subsidy ) Author
Submitted to: Biological Control
Publication Type: Peer reviewed journal
Publication Acceptance Date: 7/23/2013
Publication Date: 7/31/2013
Citation: Tipping, P.W., Martin, M.R., Pratt, P.D., Rayamajhi, M.B., Center, T.D. An abundant biological control agent does not provide a significant predator subsidy. Biological Control. 67:212-219. 2013. Interpretive Summary: Melaleuca quinquenervia is a serious wetland weed from Australia that threatens the Everglades ecosystem. Classical biological control programs have resulted in the release of specialist insect herbivores that only feed on the weed. However, predators can attack these biological control agents which may potentially affect food webs by providing prey subsidies as well as reducing the beneficial effects of the agents on the weed. The melaleuca snout weevil, Oxyops vitiosa, uses chemical defenses derived from the plant to deter predators like Podisus mucronatus, a native predacious stinkbug. As a result, predation had no effect on the degree of suppression exerted on the weed by the weevil. During the course of these multi-year experiments, there was no evidence of a significant prey subsidy provided to this generalist predator that might alter food webs.
Technical Abstract: Classical weed biological control agents, regardless of their effectiveness, may provide subsidies to predators and parasites. The chemically defended weevil Oxyops vitiosa Pascoe is a successful agent that was introduced to control the invasive tree Melaleuca quinquenervia. Two consecutive small plot experiments that lasted two and three years, respectively, examined the population dynamics of O. vitiosa while subjected to predation by Podisus mucronatus, a native generalist predator. During this time the estimated mean (+ SE) percent predation of O. vitiosa larvae was 7.2 + 1.7% per sample date in the two year study and 8.4 + 0.8% in the three year study. There was no relationship between the number of larvae per tree and the number that were predated in either experiment. Consumer losses from predation did not cascade down to the producer level and influence any plant variable in either experiment. Time series analysis found no autoregressive processes for predation in either experiment while there were strong first through fourth-order auto-correlations for live larvae in both experiments, indicating the presence of strong trends in prey density. If longevity was a gauge of the relative importance of a predator subsidy, then any provided by O.vitiosa was negligible because predation was unlikely to increase over two consecutive sample periods despite increasing prey populations. The benign presence of sustained populations of a biological control agent provides a counter argument to studies that imply inevitable and perilous linkages between introduced agents and community food webs.