Technical Service Center
Environmental Applications and Research Group — Publications
Yellow Starthistle Biocontrol at the Bureau of Reclamation, Red Bluff, CA, Facility
Technical Memorandum No. 8220-99-02
Yellow Starthistle Background and Biology
Yellow Starthistle Control Options
Hairy Weevil Background and Biology
Release Site Description
Obtaining and Releasing the Insects
Yellow starthistle is one of California's worst noxious weeds, infesting 8 million acres of varied habitats. It is also the worst weed problem at the Bureau of Reclamation's Red Bluff, CA Facility. A program to control this yellow starthistle using the biocontrol insect Eustenopus villosus began June 1998. Over a period of three to five years this insect should begin to affect the density of the weed by decreasing the seed output. Revegetation of the area with desired plant species will be an important facet in reclaiming this habitat and reducing further weed problems.
Yellow starthistle (Centaurea solstitalis L. (Asteraceae: Cardueae)) is a winter annual from the Mediterranean region that arrived in California about the mid-1800's. It is estimated to infest about 8 million acres in California alone (Maddox and Mayfield, 1985) and is also found in Idaho, Oregon and Washington.
Yellow starthistle is a gray-green to blue-green plant with a vigorous taproot, and height ranging from 6 inches to over 3 feet. Stems and leaves are covered with small hairs that give them a whitish appearance. Stems may appear flattened and "winged". The most easily recognized feature is the yellow dandelion-like flower with sharp spines at the base.
Yellow starthistle reproduces only by seed. The number of seeds produced can vary greatly, generally from 700 to 10,000 seeds per plant. Higher and lower numbers have been obtained from crowded plant populations showing about 100 seeds per plant, and plants in a herbicide study that produced 170,000 seeds per plant when allowed to regrow the following year (Thomsen, 1991). Plant density and precipitation during the growing season likely have the greatest effect on seed output.
Most of the seed produced is viable and will germinate during the autumn rains, but a very small percentage of the seed can remain dormant in the soil for up to ten years (Callihan et al, 1993). Most seed falls within 2 feet of the parent plant, but wind, birds, and vehicles can play a minor role in dispersal. Predation of seed by birds is common; species include ring-necked pheasants, California quail, house finches, and American goldfinches (Roche, 1992).
This weed can form a monoculture, replacing desirable vegetation and habitat for wildlife and rangeland. The sharp spikes on the flowerheads make passage through heavily infested areas extremely unpleasant. In addition, when it is eaten in sufficient quantities by horses, it can lead to a fatal nervous disorder called "chewing disease".
A number of options have been tested for controlling yellow starthistle (Thomsen, 1991 & 1994). They include, briefly:
- Mowing must be done just before seed formation (early bud stage), and repeated at least once. Mowing too early or too late will either encourage growth or aid in seed dispersal.
- Cultivation just after seedlings emerge in the fall can be effective. However, it will bring more seeds to the surface to germinate. Cultivation in the spring must be much deeper to have any effect on the deep taproot.
- Controlled grazing can manage large stands of yellow starthistle, but animals must be actively managed. Choice of ruminant, timing of grazing, and grazing intervals are important considerations.
- Burning is most effective when done just before flowering. However, starthistle is still green at this time, and there must be enough dry biomass from other plants to carry a fire. It is important to obtain a burn permit and have all safety measures in place before conducting a prescribed burn.
- Plant competition is used in combination with other control methods. Annual or perennial grasses are generally chosen, and irrigation can be important for success.
- Chemical control can be accomplished with both pre- and post-emergent herbicides. Non-target effects must be seriously considered, and chemical costs make application to large areas impractical.
- Biological control with insects is a fairly new method of control. Many specifics are still unknown, such as best choices of insects, insect habitat preferences, and degree of control expected. However, basic data has been collected since 1990, and control can occur if selections are made from this initial information. We have chosen to use this method at Reclamation's Red Bluff facility because of the acreage of infestation, potential non-target effects of herbicides, and the staff time requirements of other methods (grazing management and mowing).
The hairy weevil (Eustenopus villosus (Boheman) [EUVI] (Coleoptera: Curculionidae)), native to Greece, was first introduced into California in 1990 for the control of yellow starthistle. Extensive testing has shown this insect to be highly host specific to yellow starthistle and it will not attack commercially valuable crops or native plants.
EUVI has one generation per year, with adults living for about one year. The overwintering adults become active in May and feed on young, closed flowerbuds, causing the buds to die. Eggs are laid in older flowerbuds (stage Bu-4, Appendix IV), and the larvae eat most of the seeds. In one study, the rate of seed consumption per one larva was 100% in small seedheads, and 70 to 99% in larger seedheads (Fornasari and Sobhian, 1993). The new generation of adults emerges from the buds during July through September.
The weevil establishes readily in most areas of California, but prefers the hotter interior parts of the state over cooler climates (coastal, higher elevations and latitudes) (Woods, 1998). High moisture and humidity may also be unfavorable.
After a number of years (three to five) in areas where the insect has suitable habitat, an established population of EUVI may begin to affect the density of yellow starthistle by substantially decreasing seed production. Unfortunately, yellow starthistle frequently responds to a decrease in plant density by producing more and larger seedheads per plant, so the expected time to control may be increased.
Release site selection at Red Bluff was based on criteria from the CDFA (Villegas, 1998): this site is above the minimum acreage needed, has mixed infested habitats nearby, and will remain undisturbed for at least five years. Insect release points were recorded with a PLGR-GPS unit, flagged, and photographed (Appendix I, Appendix II).
Insects were collected on June 25, 1998 by Debra Eberts (Denver, Colorado) at a field day sponsored by the CDFA, hosted by Baldo Villegas. After training, participants used sweep-nets to collect insects at a site in Redding, CA. The insects were sorted, counted, and placed into cardboard containers which were kept cool in a picnic cooler until the insects were released the following day (June 26, 1998). Three release points were chosen at the Red Bluff facility with the assistance of Wayne Stokes, and about 330 insects were released at each of these points. The insects were released without cages directly at each of the three points, rather than being scattered about.
Future plans will follow the recommendations given by the CDFA, Biological Control Program (Villegas, 1998) as summarized below:
- Population monitoring will be done at least once, but if possible two, time(s) per year. Each release site should be monitored starting from the time floral buds B3-B4 (Appendix IV) start to show up at the release site. (In the Sacramento Valley this takes place from mid May through July.) EUVI feed on all closed flower buds (B1-B4), mate on B3-B4 flower buds, and the females oviposit on B4 buds. Initial surveys for colonization and establishment should be done either visually by looking for the weevils on the yellow starthistle floral buds or by sweeping the top 12 inches of bolting plants that have the appropriate floral buds preferred by EUVI.
- Another visit to the nursery site should be made during September-November for a visual count of larval damaged seedheads, as emergence or exit holes made by F1 weevils upon exiting their pupal chambers in the seedheads are difficult to see. During this visit only seedheads exhibiting signs of larval damage should be counted during a 15-minute time period within the release site. An effort should be made to see how far from the release site the larval damaged seedheads can be found.
- The above monitoring activities should be done each year to determine EUVI establishment and population buildup. After three years the EUVI population at each site may be high enough for redistribution. To determine this, take visual counts of all EUVI seen during three separate five-minute transects of the site. The five minute transects can be supplemented by sweeping the YST seedheads with an insect net. If fifteen or more weevils are seen/swept per five-minute transect, the EUVI population is considered well established, and redistribution collections are recommended.
Revegetation of the area with desired plant species will be an important facet in reclaiming this habitat and reducing further weed problems. Revegetation plans will be developed during FY 1999 & 2000. Recommendations will be based on input from the local managers as to their desired future uses of the area, and on guidance from local experts and studies of plants suitable for the area. These plants will need to compete well with weeds, and establish without irrigation (if possible). Because of the time lag of three to five years until decreases in plant density are seen, actual implementation of plans would probably not be necessary until 2001. However, progress of the insects and effects on plant density will be monitored each year to determine any need to change the implementation date.
Callihan, Robert H., Timothy S. Prather, and Francis E. Northam. 1993. Longevity of Yellow Starthistle (Centaurea solstitialis) Achenes in Soil. Weed Technology 7:33-35.
Fornasari, L. and R. Sobhian. 1993. Life History of Eustenopus villosus (Coleoptera: Curculionidae), a Promising Biological Control Agent of Yellow Starthistle. Environmental Entomology 22:684-692.
Maddox, D.M., and A. Mayfield. 1985. Yellow starthistle infestations are on the increase. Calif. Agric. 39:10-12.
Roche, Ben F. Jr. 1992. Achene Dispersal in Yellow Starthistle (Centaurea solstitialis L.). Northwest Science 66:62-65.
Thomsen, Craig D., Marya E. Robbins, and Stephanie Larson. 1991. Yellow Starthistle Control. University of California, Department of Agronomy and Range Science, Range Science Report No. 30.
Thomsen, C.D., W.A. Williams, M. Vayssieres, and F.L. Bell. 1994. Yellow Starthistle Control. University of California, Department of Agronomy and Range Science, Range Science Report No. 33.
Villegas, Baldomero. 1998. Work Plan for Colonization Releases of the Hairy Weevil Eustenopus villosus (Boheman) for the Biological Control of Yellow Starthistle in California. personal communication.
Woods, Dale M., Editor. 1998. Biological Control Program Annual Summary, 1997. California Department of Food and Agriculture, Plant Health and Pest Prevention Services, Sacramento, California. 85pp.
We would like to express our appreciation for the contributions and support from the following:
Bureau of Reclamation Research Program WATER Project EE007, "Development of Improved Aquatic Site Pest Management Methods" Bureau of Reclamation Program Analysis Office.
Bureau of Reclamation, Red Bluff Field Office, Red Bluff, CA Baldomero Villegas, California Department of Food and Agriculture, Biological Control Program, Sacramento, CA.