USDA Agricultural Research Service --Lincoln, Nebraska


Date of this Version



Published in Biological Control 35 (2005) 319–329.


All introduced natural enemies present a degree of risk to nontarget species. Since most biological control programs use relatively host-specific natural enemies, the risk to nontarget species is generally very low, particularly from biological control of weeds, which uses extensively tested and validated host-specificity testing procedures to predict risk. However, many of the published comments about risks of biological control are superficial or misleading, often inappropriately lumping risk from all taxa of agents as “the risk of biological control,” and ignore the potential benefits, rather than dealing with species-by-species risk and benefits. Particularly confounding accurate predictions is the common mixing of parameters of hazard and exposure in discussions of risk. In this paper, traditional risk analysis techniques are discussed and adapted for biological control. How people perceive risk is the key to understanding their attitude to risk. Some of the criticisms of biological control relating to inadequate post-release monitoring are valid and the ethical responsibilities of biological control scientists in this area are also discussed. Biological control scientists should address objectively the criticisms of biological control, continue to review and adjust current host-specificity testing procedures and make appropriate changes. This process will result in better science, ultimately delivering more focused programs, and altering the perception of risk from biological control agents by objective observers.