Attempts to achieve biological weed control with insects are met with stringent risk assessment in the United States. Before insects are released, their potential to attack economically important or threatened plants is closely evaluated. This assessment focuses on risk and does not adequately address host-range data, especially results from multiple-choice and open-field tests; therefore, it may result in missed opportunities for safe, effective, and natural weed control.

An article in the journal Invasive Plant Science and Management examines five successful cases of insects released in the United States for weed control. Prerelease and postrelease data collected for these insects (known as agents) are compared to evaluate the safety of biological weed control. In general, experimental host range data accurately predicted or overestimated the risks to nontarget plants.

Before releasing an insect to control weeds, the benefits and risks are weighed. An herbivore is tested, one plant species at a time, under confined conditions to determine its fundamental host range. But to establish its realized host range, under natural field conditions, an insect is allowed to choose from among the target weed and other potential host plants.

The U.S. Department of Agriculture Animal Plant Health Inspection Service is the agency responsible for biological control introductions. Recent decisions by the agency favor a conservative approach that focuses solely on the fundamental host range. While this is a safe methodology, the authors of this analysis argue that this criterion significantly overestimates the risks posed by an agent, thus limiting biological weed control options.

The authors contend that five agents, historically proven as successful, would not have been released under today’s assessment standards. These include a leaf beetle, mite, and weevil that reduced populations of St. Johnswort, bindweed, and toadflax, respectively.

A reassessment of current policy is proposed to enable the consideration of both benefits and risks of all management options, such as biological control, but also herbicides or non-action. Focus should be at the habitat level, rather than for individual species of concern.  In the long term, the authors believe that biocontrol legislation should be amended to include this risk–benefit analysis, ideally, early on in the control program, noting that this has proven effective in New Zealand and Australia.

Full text of the article “Successes We May Not Have Had: A Retrospective Analysis of Selected Weed Biological Control Agents in the United States,” Invasive Plant Science and Management, Vol. 7, No. 4, October–December 2014, is now available.