Biology in the News Explained

We need biological control agents that work

Biological control of pest species has come under a lot of criticism in recent years. Employed by governments and private businesses since the late 1800′s, the practice was largely unregulated until the 1980′s, and regulation remains varied and complicated among different countries, and different states within the U.S.

Classical biological control specifically refers to the importation of an alien species to control pests that are usually aliens themselves. Traditionally, it has been employed in agricultural systems, in which the alien pest arrived in its new location accidentally. More recently, classical biocontrol has been undertaken to control pests in natural systems, which in the case of plants, were usually intentionally introduced through horticulture trade of botanical gardens. It is currently viewed by many ecologists as a critical tool in the battle against invasive alien species, which are growing into an ever larger economic and ecological problem.

The recent criticism of biocontrol has focused on the likelihood of “non-target” attacks, which occur when the introduced agent feeds on unintended prey or plant species, including natives. This is an especially important problem in areas with many endemic species – in the U.S., the states most affected are Hawai`i, California, and Florida because of their high native species diversity. The concern about non-target attacks is excessive, in the opinion of most biocontrol practitioners, including Messing and Wright (2006), who describe a scenario in which the peril to Hawai`i’s agriculture and ecosystems is being increased by a bureaucracy that will not allow alien biocontrol agents to be imported, even to combat serious economic and ecological pests in Hawai`i that are contributing to the destruction of native ecosystems.

Certainly their concern for the native species of Hawaii is not misplaced. Invasive alien plants such as thorny blackberry crowd out natives and are less likely to be eaten by foraging alien herbivores such as pigs and goats. Generalist insects such as the two-spotted leafhopper consume hundreds of native plant species which have no natural defenses against the aliens feeding on them. So it does seem unreasonable that Hawai`i is so stingy with its species importation permits, when researchers do their best to show that their biological control agents will not feed on any native species. (I know both of the authors personally, and can verify that they are stringent in their criteria for potential non-target interactions.)

The problem with Messing and Wright’s paper is the same problem with most discussions debating the use of biological control in the scientific literature, at conferences, and in online discussions. While we are focusing much energy and attention on predicting non-target interactions, which in many biocontrol programs has thus been adequately addressed, there is almost no discussion about doing a better job of predicting whether or not the introduced agent will actually be effective once introduced.

Messing and Wright themselves toss out that only 16% of biological control programs have been effective at controlling the target, and yet later in the paper complain about their inability to introduce agents, which is based on the assumption that the introduced agents will actually work. The high probability that (based on current practices) they will not work is never addressed in their paper. Twice the number of effective biocontrol agents, or around 33% actually establish — probably an underestimate, since follow-up generally does not extend for years, and those that are not actively controlling the target could be missed early in monitoring — and rarely has any follow-up been done to understand their role in the native ecosystems. If they are not attacking non-targets, and they are not controlling the pest, how are they interacting with native species?

It is true, as Messing and Wright point out, that in the current modern age of regulation, non-target effects have greatly decreased. But it is also true that even in the case where a known specialist was introduced, there can be indirect food web effects discovered when people have taken the trouble to look, which they rarely do. In one case (Willis and Memmott 2005), a native parasitoid attacking an specialist insect herbivore introduced to control an alien weed increased greatly in numbers and thus made native herbivores much more susceptible to attack, upsetting the food web in the system.

Effective biological control programs are indeed cost-effective, but a lot of money is spent for many years on the assessment of a single agent, and are we getting our money’s worth if only 1 in 6 actually work? A whole new science may need to be developed to further our ability to predict the effectiveness of biological control agents. Those who are generally against biocontrol find the balance of effective agents vs. the risk of non-target effects, which while getting lower, will never be zero, to be unacceptable. What if biological programs could predict effectiveness 50% of the time? That would alter the equation, and make these programs more palatable to many ecologists such as myself.

How can something as notoriously unpredictable as species interactions be better assessed before release? One method which would give researchers much better data about the role of species in both native and alien habitats is the construction of quantitative food webs. Normally when biocontrol workers go the country of the target pest’s origin, they observe only the two-species interactions between the target species an its natural predators or herbivores. This provides no predictive value because plants and animals do not interact in 2- or 3-species bubbles, they interact complexly both directly and indirectly with many species. Why don’t we construct food webs of biocontrol agents in their native habitat and figure out why they seem to be effective there? We can also do retrospective studies on established biocontrol agents that are ineffective and use food webs in both the new and native habitats to try and understand why.

Of course, ecological research is conducted at the whim of funding agencies, largely NSF in the U.S. Politics and inertia often determine what is funded more than science does. Perhaps someday, however, people will realize that for both economic and political reasons, research into the prediction of the effectiveness of biocontrol agents makes clear economic, ecological, and political sense. We do need biocontrol as an option for saving some ecosystems. But we particularly need biocontrol that works.


Messing, R.H. and Wright, M.G., 2006. Biological control of invasive species: solution or pollution? Frontiers in Ecology and Evolution 4:132-140.

Willis, A.J. and Memmott, J. 2005. The potential for indirect effects between a weed, one of its biocontrol agents and native herbivores: A food web approach. Biological Control 35:299-306.


No Responses to “We need biological control agents that work”


  1. Horticulture: For Mother Nature Fans - [...] We need biological control agents that work | Bioblog – Biological control of pest species has come under a ...

Leave a Reply

Your email address will not be published. Required fields are marked *

− one = 0

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>