Noxious Weed IVM Guide Contents, IVM for Noxious Weeds
Broom, Gorse, Knapweeds, Leafy Spurge, Purple Loosestrife,
Smooth Cordgrass, Tansy Ragwort, Yellow Starthistle
IPM Access Key Documents, Home Page
IVM Technical Bulletin
Purple Starthistle
Gather Background Information

The first step in an IVM program is to gather information on the life cycle and habits of the noxious weed.

Purple starthistle, Centaurea calcitrapa, is primarily a biennial plant (rarely an annual) that is highly branched and can grow approximately 1 to 2 feet tall in productive soils. Stems and leaves have fine, cobwebby hairs that fall off with time, giving older leaves a smooth appearance. Leaves are long and divided into narrowly linear segments, tapering at the tip. Purple starthistle’s scientific name comes from the word "caltrop," an ancient weapon with four spine-like projections that was used against mounted warriors. As its name suggests, the purple flowers are protected by sharp, straw-colored spines about one inch long. Purple starthistle can be distinguished from its close relative, the yellow starthistle, mainly by flower color. Occasionally, flowers of both plants may be pale. In this case or when the flower has not yet bloomed, the yellow starthistle can be distinguished by the wing-like leaf margins on the stem (Robbins, Bellue, and Ball 1951).

Purple starthistle is unpalatable to livestock because of its poor taste and its spine-covered flowers and seedheads. Unmanaged purple starthistle populations threaten the quality of grazing lands, dominate the most productive lands, reduce forage production, and crowd out native plant habitat. Animals avoid purple starthistle infestations, causing a loss in grazing potential and a reduction in the cattle carrying capacity of rangeland. In park lands, purple starthistle reduces recreational use (Beck 1993; Amme 1997; Abbors 1997).

Purple starthistle is native to the Mediterranean region, southern Europe, and northern Africa. It was accidentally introduced to North America early this century, and has since become a widespread problem throughout California, particularly in Solano, Napa, Sonoma, and Marin Counties. Spotty distributions range from Humboldt County in the north to San Diego County in the south. Purple starthistle is present in Idaho, New Mexico, Arizona, Washington, and Oregon where it is established but not yet a serious problem.

Look for purple starthistle in waste sites, moist, grazed areas, uncultivated fields, and along traffic corridors such as roadsides and cattle trails (Amme 1982, 1985).

Life Cycle
Purple starthistle is a biennial plant, spending roughly one year as a prostrate rosette stage before bolting and flowering. Purple starthistle can also have an annual cycle, but this form is rare. Purple starthistle is adapted to dry conditions and grows long into the summer months. It usually produces flowers from June to August after annual grasses have seeded (flowering times will vary according to climate). In moist years, purple starthistle flowers later in the summer and in dry years it flowers earlier. Seeds develop soon after bees visit the flowers. Because purple starthistle seeds have a reduced "pappus," a structure that facilitates seed dispersal, most seeds fall near the parent plant, creating a reservoir of seeds (seed bank) in the soil that can remain viable for many years. Since seedheads break off easily, animals, vehicles, or moving water can disperse them. Seedheads are particularly well suited for sticking to rubber tires, where they may be carried great distances by vehicles and harvest machinery (Amme 1997; Abbors 1997).

Special Challenges to Management
Purple starthistle seeds can remain dormant in the soil for many years; thus, constant observation and management is necessary to control present and future outbreaks. Long lasting seeds allow purple starthistle to re-invade an area where they have been absent for 2 to 3 years, particularly after heavy rains following a drought year. Viable seeds can be lodged in deep cracks in the soil. During heavy rains the buoyant seeds float to the upper layers of the soil and germinate. In addition,control measures are complicated by the lark sparrow, which selectively nests in purple starthistle infestations. Purple starthistle cannot be controlled solely by insects or pathogens; therefore, successful management requires an integrated approach that includes continual monitoring, physical removal, and herbicide spot treatments. (Amme 1997; Abbors 1997).

Site-Specific Questions
Some questions, such as those below, can only be answered on site.

Set Management Objectives

Set Realistic Goals for Your IVM Program
The answers to the following questions can help you set realistic objectives and goals.

Levels of Control
Containment - keeping an established population of the weed from spreading to non-infested areas. This strategy is especially useful when time and money are in short supply or when the infestation is very large. For example, a barrier strip between infested and non-infested areas can be maintained and monitored so that adjacent lands remain weed free. In addition, measures that stop seed production can help prevent further spread of the weed.

Reduction - reducing the area covered by purple starthistle, or reducing its dominance. This strategy can also be used against new or established weeds, but it requires more resources and more time than containment.

The "Bradley Method" (see Appendix 2), developed in Australia, is a simple yet innovative strategy for natural areas that combines containment and reduction.

Eradication - completely eliminating the weed from the management area. This strategy usually consumes the greatest amount of time and resources and is applicable mainly to newly-invading weeds that are confined to a limited number of small areas.

Establish Monitoring Programs

When planning a monitoring program, keep in mind the context of your target weed: is it invading or has it already invaded?

Locate and record purple starthistle infestations on a map. (Chapter 2 of the University of Northern Iowa IVRM Technical Manual contains a detailed discussion on how to map and inventory vegetation - see Bibliography). Note particularly sensitive areas on the map, such as critical habitat for threatened or endangered species, agricultural production areas, or areas subject to frequent disturbance and thus prone to invasion. Update maps at regular intervals.

Focus monitoring efforts on sites where purple starthistle problems are most likely to occur (see Distribution). Encourage public sighting and reporting through an education or incentive program (see Educate Vegetation Management Personnel and the Public).

Prioritize the sites you will work on. Make a realistic assessment of your weed management resources, keeping in mind the goals of your project and the cost of a follow-up program after any treatments. Without follow-up, your control efforts will be wasted. It is better to thoroughly control a weed at one or two sites than to use up resources to incompletely control the weed at many sites. If the weed is very widespread, try to determine where it poses the most serious economic, social, or environmental problem and concentrate on those areas.

Plan monitoring and treatment efforts to coincide with critical life stages of the weed. To use your resources efficiently, try to include monitoring with other planned activities in the area.

Maintain records of your monitoring activities. Creating standardized forms will make data collection easier and help remind you to gather all the information you need. Forms work best if they include labeled blanks for all pertinent information and allow the user to check or circle rather than having to write words or numbers (See Appendix 3 for some examples of forms).

Include information such as the name(s) of the person(s) collecting the data, the location, and date of monitoring; a qualitative description of the vegetation, such as the names of the plants or types of plants (native vegetation, annual/perennial weeds, trees, etc.) and stage of growth (germinating, flowering, setting seed, etc.); a quantitative description, such as percent cover, plant density, size of the patch, or if possible, the number of plants.

Note special conditions such as unusual weather events and record treatment history, including information on treatment applications (who, when, where, how, cost, difficulties, and successes). This will allow you to evaluate and fine-tune treatments.

Set Treatment Thresholds

Setting treatment thresholds includes prioritizing and balancing treatments with resources. Weeds will be treated when populations increase beyond a predetermined level. This level will largely depend on the characteristics of the site and weed. In some cases the level may be no weeds at all, and in other cases the number of weeds you can tolerate may be much greater.

Considerations for Setting Priorities
What is the size of the weed population? The opportunity for control is related to the infested area. Small patches can be more easily controlled than large infestations.

What is the level of the threat? Is the purple starthistle population changing? Is it in an area where soils are frequently disturbed? Does it threaten agriculture, pastures, or rangeland? Is it encroaching on critical habitat for a rare, threatened, or endangered species? Is it displacing the best examples of native communities?

What resources are available? Do you have the resources required for carrying out your goal?


With the advent of herbicides, prevention as a weed management technique has often been neglected; however, it is a practical, cost-effective, and extremely important part of noxious weed control.

General Weed Prevention Measures

Revegetation - Follow-up Weed Prevention
Establishing dense, competitive vegetation can help permanently replace weeds. Revegetation is critical in preventing weed infestations in areas where the soil has been disturbed or the vegetation removed. In some situations you may be able to encourage desirable vegetation that is already in place; but because of the aggressive nature of purple starthistle, it is more likely that you will need to thickly sow seeds of desirable, competitive plants.

Apply Management Methods

No individual method will control purple starthistle in a single treatment; diligence and persistence will be required over a number of years to subdue this weed. The treatment methods described in this section will help you to design an integrated program that will suit the circumstances of your particular situation.

Biological Controls
Currently, there is no biological control program for purple starthistle. Two insects that have been introduced to control other Centaurea spp. were observed feeding on purple starthistle; however, they do not specifically control purple starthistle. Although Larinus minutus, currently used to control diffuse knapweed, and Bangasterus orientalis, a predator of yellow starthistle, will feed on purple starthistle in the absence of their primary host, their effect on purple starthistle density is unknown (Jordan 1995; Sobhian 1993)

Puccinia jaceae is an airborne rust fungus that is used as a biocontrol agent for yellow starthistle and some knapweeds. The fungus also attacks purple starthistle leaves, reducing leaf longevity. Despite attack, purple starthistle naturally has a short leaf life span and can withstand rust infection with little change in root biomass. Purple starthistle plants still survive and the fungus' effect on purple starthistle is largely unknown (Shishkoff and Bruckart 1993).

Although goats and sheep have been used to control yellow starthistle and other weeds, purple starthistle is more unpalatable and there are no documented studies of purple starthistle grazing. Grazing is unlikely to be an effective control for purple starthistle; however, if goats and sheep can be coaxed to eat purple starthistle, grazers may be helpful in moderate infestations (Abbors 1997; Amme 1997).

Physical Controls

Manual Removal
When beginning a manual removal project, flag the treated areas so they can be identified for follow-up in subsequent seasons. It is easiest to work in relatively small areas of infestation. When faced with dense and/or extensive stands of purple starthistle, it is useful to divide them into grids (with flags, stakes, etc.) so that workers can thoroughly weed smaller areas before moving onto the next grid. The grid system also facilitates dividing work activities between those pulling and those removing the debris.

Manual removal is more effective on individual purple starthistle plants such as outliers and new invasions. Because manual removal is labor intensive, it is applied mostly to small populations of less than 100; however, with greater efforts, labor, and resources, larger populations can be successfully removed by hand.

Manual removal should be applied selectively to avoid excess soil disturbance, which may lead to a favorable environment for weeds to become established. Exposing and mixing soil layers can also bring buried seeds to the soil surface and stimulate seed germination. Rosettes can be found mixed in with bolting plants during the following season of treatment and both can be removed at the same time.

The tough roots of purple starthistle can be severed by a tool called a "pulaski," a long-handled hoe commonly used by firefighters. The pulaski is the only tool that can penetrate compacted soil and slice through the base of the plant. To kill purple starthistle and prevent resprouting, chop the root deep enough so that the root crown is cut out and you no longer see any purple color in the stem, at least 3 inches below the base of the plant (Amme 1985; Amme 1997; Abbors 1997).

Purple starthistle can be removed during an stage of plant’s life cycle. It is more effective to concentrate efforts on older, larger, and more conspicuous plants rather than rosettes that are difficult to find. Remove plants before seeds are formed to prevent the spread of new seeds. Continual monitoring will catch rosettes when they bolt. If plants are removed late in the season, after bees visit flowers, it is important to dispose of the plants quickly or to put them in plastic bags to prevent the spread of viable seeds (Amme 1985, Abbors 1997).

Mechanical Removal

Mowing is largely ineffective against purple starthistle. If infested areas must be mowed, such as recreational areas, lawns, and fields, the best time to mow is from mid to late summer when purple starthistle has low resprouting ability. Avoid mowing infested areas in early spring, particularly before Memorial Day in May for park goers. This will cause resprouting populations to explode and eliminate grassland competition (Amme 1997).

Integrated Example
Successful purple starthistle management involves treating current weed problems and employing preventive measures against future weed invasions. The East Bay Municipal Utility District (EBMUD) has eliminated many heavy purple starthistle infestations near San Francisco. The program involves a long-term commitment to continual monitoring, treatment according to thistle levels, and preventive measures against seed invasion. Conspicuous plants are treated before they set seed in order to exhaust the seed bank. Continual monitoring identifies young plants when they bolt. Purple starthistle may not be totally eradicated, but populations can be kept at very low manageable levels. Steve Abbors, Manager of Watershed and Recreation at EBMUD, outlined a successful long-term management procedure using an integrated approach to control purple starthistle:

1. Identify problem areas

2. Map problem areas and update maps yearly

3. Use an integration of treatment methods

4. Prevent weed growth and future seed dispersal into treatment area

5. Monitor regularly and apply treatment when necessary. Assess and re-map the area regularly. After 5 to 7 years of consistent treatment and depleting the seedbank, purple starthistle populations dramatically crashed.

Chemical Controls
In IVM programs, herbicides are considered transition tools that enable the manager to suppress weeds and replace them with desirable, competitive vegetation. Thus, it is important to select the least-toxic, low-residual herbicide that is effective against the target weed, and to apply them in a judicious manner.

The following links discuss the primary considerations when using herbicides:  Use Herbicides Properly, Herbicide Information Resources, Criteria for Selecting an Herbicide, and Application Methods.

Proper Timing
Applying herbicide to plants when purple starthistle is most susceptible (preferably before seeds are produced) is crucial to the effectiveness of the treatment. Herbicides are best applied when plant food reserves are at their lowest, just after the plant bolts but before flowers set seed. Treat purple starthistle before bees visit flowers to prevent seed production. In California, purple starthistle bolts around late May (times will vary depending on climate.) Apply herbicides directly to the tops of individual plants. Use just enough herbicide to cover the tops of the plant, but not so much that it drips onto the vegetation below. This vegetation is important in re-colonizing the site after purple starthistle is controlled. If surrounding plants are killed, purple starthistle can re-colonize the bare patches. Add a non-toxic, biodegradable dye to the herbicide to facilitate even coverage and identification of treated plants (Amme 1997; Abbors 1997).

Because purple starthistle flowers after annual grasses have seeded, a non-selective herbicide can be used with minimal impact on annual grasslands. For large populations, use a selective herbicide. To minimize negative impacts on humans, wildlife, and the environment, apply herbicide with a wick applicator or backpack sprayer. In riparian areas, remove purple starthistle manually since moist soils will facilitate pulling.


Abbors, S. 1997. Pers. Comm. Manager of Watershed and Recreation. East Bay Munincipal Utility District. 500 San Pablo Dam Rd. Orinda, CA 94563.

Amme, D. 1997. Pers Comm. Associate State Parks Resource Ecologist. 1504 Beverly Place, Albany, CA 94706.

Amme, D. 1985. Controlling purple star thistle and artichoke thistle on EBMUD Land. Design Associates Working with Nature, Inc. 24pp.

Amme, D. 1985. Controlling Purple star thistle: a case study. Fremontia 13(2): 22-23.

Amme, D. 1982. Pilot project: purple star thistle control on East Bay Regional Park District Land. Design Associates Working with Nature, Inc. 10pp.

Artz, M.C. 1989. Impacts of linear corridors on perennial vegetation in the East Mojave Desert (California, USA): implications for environmental management and planning. Nat. Areas J. 9(2):117-129.

Dewey, S.A. and J.M. Torell. 1991 What is a noxious weed? In: James et al., Noxious Range Weeds. Westview Press, Boulder, CO.

Fay, P.K., T.D. Whitson, S.A. Dewey, and R. Sheley, eds. 1995. 1995-1996 Montana-Utah-Wyoming Weed Management Handbook. Coop. Ext. Serv., Montana State University, Bozeman, MT. 245 pp.

Fuller, T.C., and G.D. Barbe. 1985. The Bradley method of eliminating exotic plants from natural reserves. Fremontia 13(2): 24-25.

Jacobs, J.S., R.L. Sheley, and B.D. Maxwell. 1997. Yellow starthistle population dynamics model. Colorado Weed Management Association 1997 Annual Conference Proceedings. Granby, CO.

Jordan, K. 1995. Host specificity of Larinus minutus Gyll. (Col., Curculionidae), an agent introduced for the biological control of diffuse and spotted knapweed in North America. J. of App. Ent. 119: 689-693.

Lacey, C.A., et al. 1988. Bounty programs—an effective weed management tool. Weed Technology 2: 196-197.

Maxwell, B. " YST software." Personal e-mail (9 Feb 1998).

Pearson, W. 1998. Pers. Comm. County Weed Extension Agent. Stillwater County, Columbus, Montana.

Randall, J.M. and J. Marinelli eds. 1996. Invasive Plants: Weeds of the Global Garden. Brooklyn Bot. Garden , Inc. Brooklyn, NY. 108pp.

Rees, N.E., P.C. Quimby, Jr., G.L. Piper, E.M. Coombs, C.E. Turner, N.R. Spencer, and L.V. Knutson, eds. 1996. Biological Control of Weeds in the West. Western Society of Weed Science, USDA/ARS, Montana Dept. of Ag., Montana State Univ., Bozeman, MT.

Robbins, W.W., M.K. Bellue, and W.S. Ball. 1951. Weeds of California. State of Calif. Documents and Pubs., Sacramento: 436-437.

Rosenthal S.S., G. Campobasso, L. Fornasari, R. Sobhian and C.E. Turner. 1991. Biological control of Centaurea spp. pp.293-302. In James et al., Noxious Range Weeds. Westview Press, Boulder, CO.

Rosenthal S.S., T. Davarci, A. Ercis, B. Platts and S. Tait. 1994. Turkish herbivores and pathogens associateed with some knapweeds (Asteraceae: Centaurea and Acroptilon) that are weeds in the United States. Proceedings of the Entomological Society of Washington 96(1): 162-165.

Sedlar L., R. Vonmoos, G. Defago, H. Kern. 1983. Fungi occurring on selected weeds: a compilation from literature data. Institut fur Phytomedizin, ETH-Zentrum, Zurich, Switzerland.

Shishkoff, N. and W. Bruckart. 1993. Evaluation of infection of target and nontaget hosts by isolates of the potential biocontrol agent Puccinia jaceae that infect Centaurea spp. Phytopathy 83(9):894-898.

Shishkoff, N. and W. Bruckart. 1996. Water stress and damage caused by Puccinia jaceae on two Centaurea species. Bio. Control 6: 57-63.

Sohbian, R. 1993. Two biotypes of Bngasternus orientalis (Coleoptera: Curculionidae) found in Greece. Proceedings of the Entomological Society of Washington 95(2): 163-164.

University of Northern Iowa. 1993. Integrated Roadside Vegetation Management Technical Manual. Produced by the Roadside Management Prog. To obtain, call Kirk Henderson at 319-273-2813.

Further Information

Last modified:  August 31, 2000
SEND MAIL - your feedback and suggestions are appreciated!
Phone: (541) 345-2272