The Knapweed War: Light at the End of the Tunnel?

 Tim Seastedt, October, 2000

 Diffuse knapweed, a plant in the Aster family that has come to dominate many pastures on Boulder Open Space lands, continues to be a major problem in Boulder County. This species is classified as a noxious weed by the State of Colorado, which mandates that this plant be controlled. Control of this weed, however, remains a contentious issue. Expensive control measures don't appear to provide any long-term relief. Regardless of how the plant is killed, there are substantial data to show that the weed will simply come back, either from seed that is in the soil or from outside sources. This plant is a 'tumbleweed', and can disperse seed long distances by wind dispersal.

 In 1997, the Boulder County Commissioners requested that I attempt a field study to demonstrate that this weed could be controlled without herbicides. That same year the County spent several thousand dollars on biocontrols - insects that eat knapweed - in hopes of finding a biological solution. Most of the insects purchased simply vanished without having any effect on the weed. However, among those insects released were 200 weevils that fed on knapweed seed heads.

Colorado Department of Agriculture personnel had also released this same species of insect at the Walker Ranch in Boulder County in 1996. In 1999 the County Weed Coordinator suggested that I check on that population. To my surprise, not only did I find an abundance of weevils, but I also found that the knapweed housing this weevil had virtually no seed production. That same year, these weevils were observed in large numbers at the release site near Superior.

This summer, I estimate that there are 20 million weevils on the Superior knapweed site, and that seed production by this weed diminished up to 75% from the previous year in areas where the weevil was particularly abundant. By the end of the growing season, about fifty percent of the seed heads in the release areas produced weevils rather than knapweed seed. The weevil numbers diminish quickly away from the release sites, but these insects have been found on adjacent properties. Further inspection of knapweed up at the Walker Ranch release site confirms the continued presence of large numbers of the weevil and little knapweed seed production. The weevil is having a major impact on the reproductive success of this weed.

In addition to the weevil, four other species of knapweed-eating insects are present at the site near Superior. These insects further reduce seed production and survivorship of knapweed. This year, a root-feeding beetle that lives in the first-year knapweed plants (the rosettes) appears to have killed many of these non-flowering plants. While we won't know for sure until next year, it's possible that the number of flowering plants will be greatly reduced. That should make the seed-feeding weevils even more effective. Two other insects, flies that live in the knapweed seedheads, also were more abundant in 2000 than in previous years. Overall, the average seed production of a knapweed plant in the area where these biocontrols were released was less than 40 seeds per plant. Obviously, it only takes one seed to replace the plant, but there are numerous sources of seed, seedling, and rosette mortality. Few if any of those 40 seeds will survive to reproduction; this number may be too low to maintain the knapweed populations.

The key question is, will this be the solution to our knapweed problem, and a potentially important secondary question is, will the weevils or other insects become a problem by attacking desirable plant species? The results suggest that control by the insects may be possible, but that the time from release of these insects to the time that knapweed is no longer a threat to neighboring areas is perhaps five years. Many knapweed-infested sites in Boulder County and elsewhere in the Front Range have not been treated as yet with these insects. Also, we do not as yet know if these insects will be able to contain the knapweed growing along roadsides and disturbed areas. Finally, some knapweed appears to flower after the weevils and other insects have finished their destructive activities. If enough of these plants are able to produce sufficient seed, then the problem remains.

I'm not overly concerned that knapweed-feeding insects will feed on other plant species. While anything is possible, the insects' feeding and reproductive activities are closely tied to the very unique chemistry created by the knapweed. No native plants have this type of chemistry. I therefore believe that the insects are much less likely to harm non-target species than is repeated use of chemicals. Further, not only are these species low risk, they're also extremely low budget items. The insects appear to be a major contribution to finding a sustainable solution to our knapweed problem.

Integrated pest management (IPM) uses management activities directed to control pest species that minimize economic, human health, and environmental risks. Weed managers may need to develop a new perspective about the presence of knapweed. The weed may be effectively controlled by insects (that is, produce no viable seed), but the flowering plants may continue to appear because of the seed bank in the soil, or because the seed continues to be introduced from other sites. Such areas should not be treated with chemicals - or even hand-pulled - because the killing of adult knapweed plants will harm the long-term ability of the insects to control the weed. Insect surveys may have to become a necessary part of weed mapping activities to develop a successful integrated pest management program.

 While I would like to end on an optimistic note, the scientist in me requires that I add the following caveat. Until we know the exact cause for the presence of this weed, its demise simply opens up the opportunity for the next "noxious weed" to appear. A critical research need is to identify those management activities that maximize the abundance of the species we deem most desirable for the specific site. To this end I strongly urge the continuation and expansion of partnerships between the land manager agencies attempting to implement stakeholder goals and research institutions whose mission is to provide the scientific knowledge required for such management activities.