Centromadia pungens

From Bugwoodwiki
Jump to: navigation, search

Authors: John Randall, Global Invasive Species Team, The Nature Conservancy


Kingdom: Plantae
Phylum: Magnoliophyta
Class: Magnoliopsida
Order: Asterales
Family: Asteraceae
Genus: Centromadia
Species: pungens
Subspecies: ssp. pungens
Scientific Name
Centromadia pungens ssp. pungens
(Hook. & Arn.) Greene
Scientific Name Synonym
Hemizonia pungens
(Hook. & Arn.) Torr. & Gray
Common Names

common tarweed, common spikeweed, spikeweed


Latin Names: Centromadia pungens Greene
Syn. Hemizonia pungens Torr. & Gray
Syn. Hartmannia pungens H. & A.

Common Name: Common Spikeweed

General Description:

Some taxonomists place H. pungens in section Centromadia which is regarded as a natural group within the genus Hemizonia based on analyses of morphology, distribution, genetics and cytology (Venkatesh 1958). H. fitchii and H. parryi whose ranges overlap that of H. pungens are also included in section Centromadia.

The genus Hemizonia contains about 31 species native to California and Baja California (Mabberley 1987). Species in this genus and several closely related genera in the subtribe Madiinae, including Madia and Holocarpha, are commonly referred to as tarweeds (Munz and Keck 1959). The genus name is Greek for 'half girdle' in reference to the phyllaries which half enclose the ray flowers (Hickman 1993). The specific epithet apparently refers to the 'pungent', or sharp-pointed, upper leaves and/or the 'pungent' scales on the receptacles of the flowerheads.

Hemizonia pungens is an annual composite with rigid, bristly, freely branching stems reaching 10-120 cm in height. Following germination (in autumn - late spring depending on rains) the plants grow into small rosettes with distinctive pinattifid (twice divided) leaves. Rosettes bolt in late spring and early summer and plants may flower from late June through the summer depending on location and weather patterns. Large plants may have well over 100 flowerheads, while the smallest may have just one or two. It is one of the few species in its habitat that flowers so late in the season. Others include certain Hemizonia spp., Grindelia spp. and the non-native invasive yellow starthistle (Centaurea solstitialis).

The leaves and stems are rough to the touch with both long and short spreading hairs which may have tiny glands which exude a strong-scented resin. On bolted plants leaves are spine tipped and generally stiff; lower leaves are 5-15 cm long, linear-lanceolate and deeply pinnattifid; upper leaves are 1 to 2 cm long, linear to awl-like and spine tipped, glabrous to scabrous with stiff margins and generally with axillary leaf clusters.

The flowering heads are terminal on short lateral stems which are axillary along the main branches and have condensed internodes and crowded leaves. They measure 0.5-1.0 cm across and contain both disk and ray flowers. The leaves appear to be whorled or involucrate beneath some flowerheads, each of which has a single series of 25-40 true phyllaries. Each phyllary partly encloses a single pale yellow disk flower at its base. The ray flowers are 3-5 mm long and two lobed. They produce fertile seeds (achenes) which are about 2 mm long and beaked with a smooth or minutely roughened surface and without a pappus. The disk flowers are darker yellow but are usually staminate (male only) and do not produce fertile seeds. The receptacle is flat with chaffy, spine- tipped scales.

The Jepson Manual of Higher Plants of California (Hickman 1993) recognizes four subspecies of Hemizonia pungens: ssp. laevis; ssp. maritima; ssp. pungens; ssp. septentrionalis. These differ from one another in the shape of their receptacle bracts, roughness of leaves and bracts and size of flowerheads. Subspecies pungens was introduced to southwestern California and to Oregon and Washington (Hickman 1993)and is an agricultural pest in the Columbia Basin. Subspecies laevis is found in grasslands in the southern coast ranges and the Peninsular ranges of California and is regarded as rare.

An analysis of the flavonoids in exudates produced by the leaves of Hemizonia species indicated that the chemistry of all 18 species examined was distinct except H. australis and H. pungens ssp pungens(Tanowitz et al.). All 18 species contained patuletin, quercitin, 3-methylquercitin and luteolin). H. pungens also contained naringenin, 3-methylpatuletin and 5,3,4-trihydroxy-3,6,7-trimethoxyflavone.

A good illustration of the species may be found in Volume IV of Illustrated Flora of the Pacific States (Abrams and Ferris 1960). Another good, but small, illustration may be found in The Jepson Manual: Higher Plants of California (Hickman 1993).

Information for the description was taken from Abrams and Ferris (1960), Hickman (1993), Mason (1957) and Hitchcock et al. (1955).


Hemizonia pungens (spikeweed) is an annual composite with small (0.5 - 1.0 cm across) flowerheads and rigid, bristly, freely branching stems reaching 10-120 cm in height. It is native to California where it is commonly found in dry grasslands below 500m and around alkaline wetlands and other marshy areas in the state's southern Central Valley. The species is common in certain seasonally flooded areas of the Cosumnes River preserve south of Sacramento. It was introduced to the Columbia River basin of northeastern Oregon and southeastern Washington within the last century and behaves as an invasive weed and agricultural pest there. Several counties in northeastern Oregon and southeastern Washington have designated it a noxious weed and as such landowners are legally required to prevent it from spreading. In 1988 it was discovered on the 386 acre Lindsay Prairie Preserve in Oregon. At that time it dominated a total of four acres but the infestation spread rapidly. By 1992 fifteen acres of the preserve were infested with spikeweed densities averaging 750 individuals/m2. Morrow County, where Lindsay Prairie is located, listed the species as a noxious weed and in 1992 served the Oregon field office legal warning that it must control the infestation on the preserve.

By this time the population was too large to control manually and experimental burns did not reduce its numbers. Experiments indicated that chlorsulfuron (Telar®) and picloram (Tordon®) reduced spikeweed densities by 80-90% when applied in early spring at 100% of the recommended rate. Use of Tordon® was discontinued on the preserve due to concerns that it might leach down to lower layers in the soil and poison the roots of native shrubs. An additional study confirmed that spring applications of chlorsulfuron at 100% of the recommended rate will control spikeweed. There are plans to spray infested areas one last time in April 1995 and then to remove surviving individuals later in the summer. Manual control efforts perhaps involving several volunteers will be needed to control the lower number of plants expected to sprout on the preserve in 1996 and thereafter. If control is neglected in the future, spikeweed will almost surely re-infest the preserve.

Hemizonia pungens was introduced to the Columbia River basin of northeastern Oregon and southeastern Washington within the last century where it behaves as an invasive weed. In 1988 it was discovered on the 386 acre Lindsay Prairie Preserve in Oregon's Columbia River Basin. At that time it dominated a total of 4 acres but the infestation spread rapidly. By 1992 Fifteen acres of the preserve were infested and spikeweed densities averaged 750 individuals m2. Morrow County, where the preserve is located, has placed spikeweed on its Noxious Weed List as has neighboring Umatilla County and several counties across the river in Washington. In November 1992, Morrow county served the Oregon Field Office legal warning that it must control the infestation on Lindsay Prairie.

The species poses no known threats within its native range



Hemizonia pungens is native to dry grasslands below about 500m in California's Central Valley (Sacramento county to Kern county), the Sierra Nevada and Cascade foothills, and the coast ranges. It is also native to depressions, alkaline seasonal wetlands and other marshy areas in the southern two thirds of the state's Central Valley.

Subspecies pungens was introduced to southwestern California (Los Angeles, San Bernardino and San Diego counties) and to the Columbia Basin of Oregon and Washington. It has become an agricultural pest in portions of the Columbia Basin, especially around Walla Walla (WA), and in Umatilla and Morrow Counties (OR) which border the Columbia River.

Subspecies septentrionalis was also apparently introduced to grasslands north to Washington state (Hickman 1993).

Hemizonia pungens grows in dry grasslands and seasonal wetlands and thrives on a wide variety of soils and will move into dryland cereal and irrigated crops. At the Cosumnes River Preserve, where it is native, it colonizes low areas that may be flooded following winter storms and is most conspicuous where the soil has been disturbed (by plowing, construction, etc.) and/or flooded for several weeks during the winter months. These areas support herbaceous vegetation dominated by introduced mediterranean grasses and forbs (Avena fatua, Hordeum leporinum, Bromus spp., Lolium multiflorum, Erodium cicutarium, and Centaurea solstitalis) and a few natives (Eremocarpus setigerus, Madia spp., Hemizonia spp., Amsinckia intermedia). It is now difficult to determine the nature of the pre-settlement herbaceous vegetation but it is thought that native bunchgrasses, particularly Stipa pulchra, Elymus triticoides and Poa scabrella dominated with forbs such as Madia spp., Orthocarpus spp., Amsinckia intermedia, and Eschscholtzia californica occupying areas between bunches.

Spikeweed was introduced to northeastern Oregon and is regarded as a pest at The Nature Conservancy's Lindsay Prairie Preserve where it infests the big sagebrush bottomland community which has a heritage rank of G3-S2. Unfortunately, the community is in degraded condition. Erosion of the ephemeral creek bed that runs through the area and the consequent lowering of the water table are believed to have allowed several exotic species, including spikeweed, to invade. As a result, it is difficult to determine the area's original vegetation. Heritage program ecologists believe the site was dominated by Big sagebrush (Artemisia tridentata) and Great Basin wildrye (Elymus cinereus). Saltgrass (Distichlis stricta), Sandberg's bluegrass (Poa sandbergii), and squirrel tail (Sitanion histrix) are also thought to have been present. Now, scattered, small patches containing only saltgrass are present and intervening areas have been taken over by spikeweed and other weedy species. Where patches of un-eroded soil remain, small clusters of natives like Sandberg's bluegrass, squirrel tail, and on rare occasions, Great Basin wildrye, hang on. The native shrubs Big sagebrush, grey rabbitbrush (Chrysothamnus nauseosus), and matchweed (Xenotriche sarathrae) have become more abundant than they are believed to have been originally but the only species in the community which occur with greater than 80% frequency (1m2 frame) are Sandberg's bluegrass and the non-native weeds cheatgrass (Bromus tectorum) and spikeweed.

Soil grab samples from areas supporting conspicuous populations of the species on the Cosumnes River (CA) and Lindsay Prairie (OR) Preserves were analysed for general fertility, salinity, sodicity and pH (Soil study 1995). The portion of the Cosumnes River Preserve from which samples were taken may be inundated for several weeks at a time during rainy winters but was formerly farmed and supported a corn-wheat-tomato rotation. These soils were slightly acidic (pH 6.7-6.9) while those from Lindsay Prairie were nearly neutral to basic (6.9-10.0). Cosumnes soils were less sodic (0.1 - 0.85 meq/100g) and less saline (0.34- 0.8 mmhos/cm) than those from Lindsay Prairie (0.29-11.5 meq/100g and 0.4-2.4 mmhos/cm respectively). Cosumnes soils were richer in Phosphorus (13-37 ppm) Potassium (13-163 ppm) and magnesium (4.1-6.7 meq/100g) than Lindsay soils (6-18 and 6-18 ppm and 2.1-4.1 meq/100g respectively) but had lower levels of calcium (6.5-10.7 meq/100g at Cosumnes vs 4.7-34.6 at Lindsay).

Further analyses indicated that at Lindsay Prairie the distribution of Hemizonia pungens was correlated with the distribution of the more eroded, sodic soils (1994). This may be because these soils are too harsh for all but a few species, reducing the competition that keeps H. pungens out of other areas of the preserve. If this is true, the species may be limited to the Lindsay series and similar alkaline flood silt soils in this region.

We found no literature on the ecology of Hemizonia pungens but some work has been done on other members of the genus. One study by Morse (1988) found that seed set in another summer-flowering species, H. luzulifolia DC ssp. rudis, was insensitive to both atmospheric and soil drought. The plants coped with the drought by continual root growth into deep soil-water reserves and by storing water in extracellular polysaccharides. The polysaccharides apparently help maintain tissue water balance and buffer cells from rapid changes by releasing water as plant water status declines. Morse (1988) also found that tissue water storage (capacitance) of above-ground shoots increased from low values in the relatively wet spring to very high values in the hot, dry summer. Greenhouse studies indicated that the increase was genetically determined, not environmentally induced. During drought periods, plants that received moisture as dew were able to continue photosynthesizing longer and to keep their flowerheads open for pollinator visitation longer. The latter factor was important because Morse (1988) found that seed set could be limited by reductions in pollinator visitation resulting from early flowerhead closure. However, the primary factor responsible for limiting seed production was the destruction of flowerheads and seeds by herbivores.

We were not able to obtain information on the insects that feed on Hemizonia pungens. English-Leob (personal communication) made observations of the closely related species H. luzulaefolia in California's Central Valley and found that it was fed on by larvae of the noctuid moths Heliothodes diminutivus and Heliothis phloxiphaga, and the plume moth Trichoptilus californicus, a tree cricket (Gryllidae) in the genus Oencanthus, flea beetles (Chrysomelidae) and the spittle bugs Philaenus spumarius.


Management Requirements:

Populations in any area to which Hemizonia pungens has been introduced should be monitored to determine if densities or the area infested are increasing. In areas where the species is being controlled it should be monitored carefully in order to determine if the methods in use are effective.

For details on the spikeweed control program at Lindsay Prairie preserve in northeastern Oregon contact Berta Youtie, P.O. Box 1188, LaGrande, OR 97850.


None known.


In 1988 a prescribed burn was conducted on a 50m2 portion of the infested area of Lindsay Prairie in order to determine whether fire might be used to control spikeweed. That season the spikeweed population in the treated area was reduced relative to that of the control area but the following year there was no significant difference. Unfortunately, the burns also appeared to reduce numbers of native shrubs and grasses and increase numbers of introduced grasses.


Studies conducted at The Nature Conservancy's Lindsay Prairie Preserve indicated that chlorsulfuron (trade name Telar® or Glean®) and picloram (trade name Tordon®) significantly reduced densities of spikeweed for at least one year (Youtie 1995). The first study was begun in March 1991, when small plots were established in the infested areas and randomly assigned treatments of 100, 75, 50, or 25% of the recommended dosages of Tordon®, Telar® or Roundup® (active ingredient glyphosate). Two months later spikeweed densities in areas treated with the highest rates of Tordon® and Telar® were reduced by 80-90% relative to pre-treatment counts and control plots. Forb densities were so low, both before and after treatment, that changes due to the treatment could not be determined. Densities in plots treated with Roundup and with lower herbicide dosages had higher spikeweed densities.

Chlorsulfuron is relatively insoluble in water but picloram is relatively soluble so it was feared that picloram might leach into the soil and be taken up by native shrubs. Therefore, a follow-up study using just chlorsulfuron (Telar®) was initiated in 1993. Five blocks were established in the infested area and each block divided into three units for this study. One unit in each block was randomly assigned as a control (no herbicide treatment) and the other two were treated with 100% of the recommended rate of Telar® in March 1993 and again in April 1994. Spikeweed densities in each unit were assessed prior to treatment (March 1993) by counting stems in fifteen 10 x 25 cm plots randomly located in each unit. Densities were assessed again following the first treatments (June 1993) and again following the second treatment (June 1994). Densities in control and treatment units were not significantly different in March 1993 (means 199m2 in controls and 110m2 in treatment units). Following the initial Telar® applications in June, however, densities were significantly lower in treated units (mean 24m2) than in control units (mean 587m2). Densities remained significantly different following the second Tarlon® applications; means were 10m2 in treated units and 289m2 in controls.

Plant height and numbers of mature and immature flowerheads on surviving spikeweed plants in the treated and control units were assessed in September 1993 (Youtie 1995). Differences were not statistically different but there was a tendency for plants in the treated areas to be taller (mean 15 cm) and to have more mature (6.9) and immature flowerheads (6.0) than plants in the control units (9.0 cm, and 2.7 mature & 0.5 immature flowerheads per plant). Plants in the treated areas may have benefited from the reduced competition resulting from the elimination of nearby individuals. It has not yet been determined whether total seed production is lower in Tarlon® treated areas or not. Handpulling all surviving plants from the treated areas before they set seed is far easier, however, than clearing untreated areas.

Whitson and Costa (1986) also studied the efficacy several herbicides for spikeweed control on alkaline silt loam soils in Oregon. They found that several provided excellent control in pastureland for at least one year. They treated 10 x 40 foot plots each containing 104 spikeweed plants per square foot with one of the following herbicides: chlorsulfuron, clopyralid, dicamba, metsulfuron-methyl, picloram, triclopyr, 2,4-D ester and 2,4-D amine. They also tested the following herbicide combinations: dicamba + 2,4-D amine and triclopyr + 2,4-D ester. Each herbicide and herbicide combination was tested at three concentrations. They found that the highest rates of chlorsulfuron, clopyralid, dicamba, metsulfuron-methyl, picloram and dicamba + 2,4-D amine treatments controlled more than 99% of the spikeweed three months after application. triclopyr and both 2,4-D formulations were not as effective. One year later, spikeweed densities had returned to pre-treatment levels in plots treated with dicamba, triclopyr, 2,4-D amine, 2,4-D ester, dicamba + 2,4-D and triclopyr + 2,4-D. Excellent control was maintained, however, in plots that had been treated with chlorsulfuron, clopyralid, metsulfuron methyl and picloram.

Based on this work Costa (n.d.) made the following recommendations:

1. for grass pasture applying picloram at a rate of 0.25 to 0.5 pounds of active ingredient per acre in early spring before seedlings germinate or at 0.5 pounds per acre after they emerge but before the rosettes exceed two inches in diameter or dicamba (trade name Banvel®) at 0.75 to 1.0 pounds active ingredient per acre after germination but before rosettes reach two inches in diameter.

2. for non-cropland follow recommendations for grass pasture or apply chlorsulfuron at 0.5 to 1.0 ounces of product per acre in early spring(before germination) or mid-spring (before rosettes exceed two inches) or fall.

3. for dryland cereals apply chlorsulfuron in the fall or spring at a rate of 0.33 to 0.5 ounces of product per acre in the fall or spring.


Jepson (1911) noted that sheep exterminated dense stands of H. pungens from fields they grazed during the summer months in Solano county, CA. It is not clear if he meant that the spikeweed colonies were truly eliminated from these field or if they were simply no longer a problem for the rest of the season.


Not tested


Hand pulling can be effective in small areas. It may also be useful as a follow-up method in areas that were treated with herbicides earlier in the growing season. Plants should be pulled while they are still green and relatively 'soft', and even then gloves must be worn. Berta Youtie notes that at Lindsay Prairie spikeweed should be pulled by May.

Mowing and discing were not tested in natural areas.

Monitoring may include mapping the extent of the infested area and determining densities of spikeweed in plots located randomly in the infested area/area of concern. Densities should be determined in early summer. If there are not enough resources to make density counts, it may be possible to take frequeny data in randomly placed nested quadrats annually. Analysis of frequency data for trends (population increasing, decreasing or stable over time) is straightforward and relatively easy.

Hemizonia pungens populations are being monitored at the Lindsay Prairie preserve to determine whether control efforts are having the desired effect. See CHEMICAL control methods section above for details or contact Berta Youtie for more information

Research on Hemizonia pungens seed dormancy and on how to restore infested areas is needed. Additional work on control methods aimed at reducing or eliminating the need to use herbicides would be useful. Management Research Needs:


Management Research Programs:

Reduction of spikeweed densities is part of a larger effort to restore bottomland habitat at the Lindsay Prairie preserve.

Hemizonia pungens poses no known threats in its native habitat but where it has been introduced it behaves as an invasive weed in natural areas and is an agricultural pest. In some counties it is listed as a NOXIOUS species and its control is legally mandated. For small, recently established infestations hand-pulling may suffice but for larger infestations the only successful control methods known involve the use of herbicides (see CHEMICAL control section, below).

Hemizonia pungens is managed as an invasive non-native pest at the Lindsay Prairie preserve in northeastern Oregon. The infestation is being controlled with spring applications of the herbicide chlorsulfuron (Tarlon®). It will be treated again in April 1995 and, if the number of surviving spikeweed plants is as low as is anticipated, they will be removed by hand in late spring or early summer. It is believed that this will eliminate most of the spikeweed seed input to the preserve, sharply reducing the number of seedlings emerging the following spring. If so seedlings that do sprout can be controlled by hand. Further herbicide applications should not be necessary unless control is neglected for a period and populations are allowed to build up again.



Abrams, L. and R.S. Ferris. 1960. Illustrated Flora of the Pacific States, Volume IV: Boraginaceae to Compositae. Stanford University Press, Stanford, California. 732 pp.

Costa, R. n.d. Control of common spikeweed.

Jepson, W.L. 1911. A Flora of Western Middle California, second edition. Cunningham, Curtiss and Welch, San Francisco. 515 pp.

English-Loeb, G. personal communication. Research Entomologist. Cornell Experiment Station, Geneva, NY.

Hickman, J.C. (ed.) 1993. The Jepson Manual: Higher Plants of California. University of California Press, Berkeley. 1400 pp.

Hitchcock, C.L., A. Cronquist, M. Ownbey and J.W. Thompson. 1955. Vascular Plants of the Pacific Northwest, Part 5: Compositae. University of Washington Press, Seattle. 343 pp.

Mabberley, D.J. 1987. The Plant-Book: a portable dictionary of the higher plants. Cambridge University Press, Cambridge. 707 pp.

Mason, H.L. 1957. A Flora of the Marshes of California. University of California Press, Berkeley. 878 pp.

Morse, S.R. 1988. Drought tolerance and reproduction of annual tarweeds (Madiinae, Compositae). Ph.D. Dissertation, University of California, Berkeley. 159 pp.

Munz, P.A. and D.D. Keck. 1959. A California Flora. University of California Press, Berkeley. 1681 pp.

Soil Study. 1995. The Nature Conservancy Lindsay Prairie special soils investigation.

Tanowitz, B.D., G. Leeder, P.N. Ross and P. Proksch. 1987. Foliar flavonoid exudates and sectional taxonomy of Hemizonia. Biochemical Systematics and Ecology 15:535-540.

Venkatesh, C.S. 1958. A cytogenetic and evolutionary study of Hemizonia, section Centromadia. American Journal of Botany 45:77-84.

Whitson, T.D. and R. Costa. 1986. Spikeweed control in pastureland. Western Society of Weed Science 1986 Research Progress Report. San Diego, CA 1986.

Original Document

Element Stewardship Abstract; John M. Randall, 1995.

Personal tools

Other Bugwood Resources
Export Current Page