Submitted to: XI Symposium on Biological Control of Weeds
Publication Type: Proceedings
Publication Acceptance Date: 9/30/2003
Publication Date: 7/1/2004
Citation: Balciunas, J.K. 2004. Are mono-specific agents necessarily safe? the need for pre-release assessment of probable impact of candidate biocontrol agents, with some examples. In: Proceedings of the XI Symposium on Biological Control of Weeds, April 27-May 2, 2003, CSIRO Entomology, Canberra, Australia. p.252-257.
Interpretive Summary: Invasive weeds cause enormous economic and environmental losses in the United States, as well as other parts of the world. Controlling some of these weeds by using carefully selected and approved biological control agents is frequently considered as the most environmentally friendly management technique, and for some widespread weeds, is the only hope for control over extensive areas. However, there has been growing concern from ecologists and other observers, about potential non-target impacts resulting from the release of overseas biological control agents. In this chapter, I review various non-target impacts, along with some examples. I also suggest that increased emphasis on releasing agents that are effective, as well as host-specific, may help reduce the chances of non-target impacts.
Technical Abstract: Historically, weed biocontrol practitioners searched for highly effective agents that were also safe. Protecting agronomic crops was the original focus for risk evaluation, but to this has been added protecting native plants, especially those related to the target weed. Host range tests are now weed biocontrol's hallmark tool, with concern about the efficacy of the candidate agent sometimes being secondary. However, even a highly specific agent can disrupt ecosystem pathways in unpredictable ways, especially if it becomes abundant on its target, but fails to reduce the target weed's populations. I review some of the current concerns about non-target impacts, both direct and indirect, as well as criticisms about the inefficient "lottery" approach that wastes scarce resources in introducing many agents, some of which never contribute to controlling the target weed. Effective agents can help alleviate some of these concerns, and there is increasing demand that we should strive to release agents that are not only narrowly host specific, but that have also demonstrated their ability to damage the target weed. While still not yet routine, pre-release consideration of the proposed agents probable efficacy is receiving increased attention. These are usually conducted overseas, in the native range of both the target weed and candidate agent. I review some of the different approaches used in these overseas evaluations. But pre-release impact assessments can also be performed under containment conditions in a quarantine. I discuss the results of two 'dosage' trials I conducted with a gall-making fly that is being considered as a biological control agent for Cape ivy. Plants exposed to both low and high densities of gall flies, were smaller, and had fewer leaves than the ungalled controls. Pre-release evaluations of a candidate agent's potential impact should lead to fewer ineffective agents being released, thereby making weed biocontrol more efficient, and reducing [but not eliminating] the possibility of negative indirect impacts on non-targets.