Skip to main content
ARS Home » Southeast Area » Fort Lauderdale, Florida » Invasive Plant Research Laboratory » Research » Publications at this Location » Publication #229967
Title: Can efficacy of new agents be predicted before their release?

Author
item VAN DRIESCHE, R.G. - UNIV OF MASSACHUSETTS
item Center, Ted
item HODDLE, M - University Of California
item MILLS, N - UNIV OF CA, BERKELEY

Submitted to: International Symposium on Biological Control of Invertebrates
Publication Type: Proceedings
Publication Acceptance Date: 12/1/2008
Publication Date: 12/31/2008
Citation: Van Driesche, R., Center, T.D., Hoddle, M., Mills, N. 2008. Can efficacy of new agents be predicted before their release?. International Symposium on Biological Control of Invertebrates.

Interpretive Summary: Biological control involves finding natural enemies of an exotic pest by searching in the native home of the pest then prioritizing the candidates to determine which is most suitable for release where the pest causes problems. It has been suggested that we could save time and money when developing unproven biological control agents if we could first predict whether or not they will become effective. While this is something that all biological control practitioners agree upon, it is not really possible. Simply measuring the damage that an individual insect does will not suffice inasmuch as control of a pest species depends on the individual effect multiplied by the number of individuals present. This depends on the agent establishing viable populations after being released and then increasing in number until populations are large enough to suppress the pest. Population growth of the newly released agent is affected by a number of factors, such as climate, parasites, diseases, predators, host strains, etc., many of which are novel in the new area and cannot be duplicated prior to release. Some biocontrol project case histories have been analyzed that can provide valuable insights into how best to focus attention on potentially effective species.

Technical Abstract: Prediction of the efficacy of new biocontrol agents before their release, while desirable and of scientific interest, is not possible at this time. Several approaches that attempt to do so are discussed. Measurement of per capita impact is recommended by some weed biocontrol scientists but does not allow comparison if agents affect plants in different ways (e.g., defoliation vs seed reduction, etc) and is useless if agents do not develop large populations. Studies of efficacy in the native range have potential value but are rarely done. Climate matching programs that provide match indices comparing collection and release areas have potential but their effective use requires understanding of which particular aspects of climatic really affect the agent. Use of isolated agent attributes (fecundity, longevity, etc) is flawed because actual species often reflect evolutionary selection for several attribute combinations, which does not allow all good attributes to be simultaneously maximized. Also, some critical attributes like host finding cannot be measured in quarantine. Analysis of natural enemy guild structures and pest life tables to identify unfilled niches for enhanced natural enemy attack is a potentially effective means of determining the kind of agent most likely to be effective, but pest models have so far been of little value in predicting agent efficacy. Analyses of biocontrol project case histories can provide valuable insights into how best to reduce initial agent pools to focus attention on species with higher probabilities of efficacy.