How Safe Is Weed Biological Control? A Global Review of Direct Nontarget Attack

Résumé

This review summarizes all known direct nontarget attack (NTA) cases of intentionally released or actively redistributed weed biological control agents, in order to allow for an objective risk-benefit analysis when choosing the most appropriate method for controlling invasive plants. Of 457 agents intentionally released until 2008, 60 (13.1%) have been recorded attacking nontarget species in the field. Of 1517 releases made using the 457 agent species, 122 (8.0%) resulted in NTA. Both proportions have declined over time. Three-quarters of all NTA cases occurred on plant species in the same family as the target weed. Approximately one-half of NTA cases were predicted/predictable. In the majority of unpredicted cases (93.5%), the respective nontarget plant species had not been tested prerelease. There were only four cases of “false negatives” (less than 1%), where the impacted plant species had been tested prerelease and deemed not at risk. The incidences of unpredicted nontarget attack of intentionally released weed biocontrol agents decreased over time and this trend is thought to continue with scientific advancements. What is most needed is more systematic post release monitoring to compare with prerelease host range testing to further advance the predictability of host use of biocontrol agents.