Musk Thistle (Nodding Thistle) A. Gassmann - CABI Bioscience Centre Switzerland, Delémont, Switzerland

In: Van Driesche, R., et al., 2002, Biological Control of Invasive Plants in the Eastern United States, USDA Forest Service Publication FHTET-2002-04, 413 p.

Pest Status of Weed

Musk thistle, Carduus nutans L., is an invasive weed that has become widespread in the contiguous states of the United States. It is a highly competitive weed of Eurasian origin that has replaced much of the native vegetation in pastures and disturbed areas (Surles et al., 1974; Kok, 1978a,b).

Nature of Damage

Economic damage. Musk thistle invades pastures, rangeland, and forest lands, and areas along roadsides, railroad right-of-ways, waste areas, and stream banks. In agricultural systems, the invasive nature and prolific seed production of musk thistle result in large populations of the weed, which compete with crops for space, nutrients, and light. Thus, infestations may reduce productivity of pasture and rangeland by suppressing growth of desirable forage plants, as well as preventing livestock from eating plants growing in the vicinity of thistles due to the sharp spines on their stems, leaf margins and blooms (Trumble and Kok, 1982; Desrochers et al., 1988a). In the northeastern United States, the highest economic losses due to musk thistle infestations occur on fertile soils formed over limestone.

Ecological damage. Musk thistle generally does not pose a great threat to high-quality natural areas, although it has been known to invade native and restored grasslands despite the presence of dense, native prairie vegetation. Musk thistle may retard natural secondary succession processes. Because musk thistle is unpalatable to wildlife and livestock, selective grazing leads to severe degradation of native meadows and grasslands as grazing animals focus their foraging on other plants, giving musk thistle a competitive advantage. Successful biological control of musk thistle (Kok and Surles, 1975) is often accompanied by increased growth and coverage of pasture grasses such as fescue (Festuca arundinaria Schreb.) and orchard grass (Dactylis glomerata L.), or less desirable plants such as spotted knapweed (Kok and Mays, 1991).

Extent of losses. The rate of expansion of musk thistle populations in North America has been very rapid since the mid-1950s, when it was first recognized as a weed (Dunn, 1976). A single musk thistle per 1.49 m² can reduce pasture yields by 23%. In Canada, stands of 150,000/ha have been observed (Desrochers et al., 1988a). Direct losses are difficult to quantify due to lack of long-term monitoring programs and data.

Geographical Distribution

Musk thistle was first reported in the United States in 1953 at Harrisburg, Pennsylvania (Stuckey and Forsyth, 1971). In the 1970s, the musk thistle complex (see Taxonomy for definition) has been found in at least 3,068 counties in 42 of the mainland states, with 12% of those counties rating their infestations as economically severe (Dunn, 1976). Musk thistle is declared a noxious weed in some 20 states, including Illinois, Iowa, Kansas, Kentucky, Maryland, Minnesota, Missouri, Nebraska, North Carolina, Ohio, Oklahoma, and Pennsylvania (USDA, NRCS, 1999). Thus, musk thistle extends from the east to west coast in both the deciduous forest and prairie biomes. It grows from sea level to about 2,500 m elevation. It prefers moist alluvial soils but will grow in eroded uplands without difficulty.

Background Information On The Pest Plant

Taxonomy

The C. nutans complex in North America has been treated either as one species with four subspecies (subsp. nutans, subsp. leiophyllus [Petrovic] Stoj. and Stef., subsp. macrolepis [Peterm.] Kazmi, and subsp. macrocephalus [Desf.] Nyman), or as three species: Carduus nutans with two subspecies (subsp. nutans and subsp. macrolepis), C. thoermeri Weinm., and C. macrocephalus Desf. (McCarty, 1978; Desrocher et al., 1988b). Recent work by Desrochers et al. (1988b) has supported the existence, in Canada, of only two closely related groups of taxa referred to as subsp. nutans and subsp. leiophyllus. Carduus thoermeri Weinm. and C. nutans subsp. leiophyllus refer to the same taxon. In North America, C. nutans ssp. macrocephalus has only been collected from the United States. Carduus nutans ssp. nutans is distinguished from ssp. leiophyllus by its moderate to dense pubescence on leaves and phyllaries, by its generally smaller head diameter (1.5 to 3.5 cm in subsp. nutans and 1.8 to 4.5 cm in subsp. leiophyllus) and by the shape of its phyllary. In subsp. nutans, the lower portion of the phyllary is more or less equal to the upper portion, while in subsp. Leiophyllus, the lower portion is distinctly narrower than the upper portion. The two subspecies also can be separated by their flavonoid compounds. Carduus nutans subsp. macrocephalus differs from subsp. nutans by a wider head diameter and phyllaries. It also differs from subsp. leiophyllus by being pubescent on leaves and phyllaries, and by having phyllaries that have the lower portion more or less equal to the upper portion. Hybridization between C. nutans and Carduus acanthoides L. also has been reported (Warwick et al., 1990). Presumably, the distribution of subsp. nutans in the United States is similar to its distribution in Canada, where it is mainly distributed in the eastern part of the country, while only subsp. leiophyllus and subsp. macrocephalus are present in the Great Plains (McGregor, 1986).

Biology The biology of musk thistle has been reviewed by Desrochers et al. (1988a). Carduus nutans L. is a herbaceous biennial though occasionally it becomes a winter annual. It is 20 to 200 cm tall, with a long, fleshy taproot. The taproot is large, corky, and hollow near the surface of the ground. One or more highly branched stems grow from a common rootcrown. Musk thistle grows in all soil textures, although the soils must be well drained. Leaves are dark green with light green midribs with a white margin (Fig. 1). The plant blooms in May and June. The showy flowers (Fig. 2) are terminal, large, solitary, and nodding (slightly bent). They are deep rose to violet or purple in color. The seeds are straw colored and do not have a light requirement for germination, but are affected by temperature. Higher germination rates occur at temperatures between 20 and 28°C. Musk thistle does not appear to have any specific climatic requirements other than a cool period of vernalization, a minimum of 40 days below 10°C for flowering. It does not reproduce vegetatively and is propagated by seeds dispersed primarily by wind. Most seeds are deposited within 50 m of the release point and less than 1% are blown farther than 100 m (Smith and Kok, 1984). Up to 11,000 achenes may be produced per individual with as many as 1,500 seeds per flower head. Seed viability remains high for more than ten years.

Figure 1. Mush thistle rosette. (Photograph by L.-T. Kok.) Figure 2. Musk thistle in bloom. (Photograph by L.-T. Kok.)

Analysis of Related Native Plants in the Eastern United States

There are no native North American species in the genus Carduus. Carduus nutans belongs to the tribe Cardueae (family Asteraceae) which is largely an Old World group. The tribe is further divided into four subtribes (Echinopsidinae, Carlininae, Carduinae, and Centaureinae) including some 13 genera in North America (Bremer, 1994; USDA, NRCS, 1999). From these, only three contain native species Centaurea (two species, subtribe Centaureinae), Saussurea (seven species, subtribe Carduinae, but the position of the genus in the tribe remains uncertain), and Cirsium (subtribe Carduinae). The genus Cirsium includes about 100 native species, of which 21 species occur in the eastern United States. One of these, Cirsium pitcheri (Torr. ex Eat.) Torr. and Gray, is listed as threatened under the Endangered Species Act. This species occurs in sand dunes along the shores of the Great Lakes in Illinois, Indiana, Michigan, Wisconsin, and Ontario.

History of Biological Control Efforts in the Eastern United States

Musk thistle was among the first 19 weeds selected for biological control when the USDA overseas laboratory was established at Rome, Italy in 1959. In the early 1960s, staff of the USDA intensively surveyed Carduus spp. in Italy, whereas the Commonwealth Institute of Biological Control (now CABI Bioscience), funded by Canada Department of Agriculture, extended the survey area across Europe from western France to eastern Austria on more than 30 species in the subtribe Carduinae. The history of biological control of thistles was reviewed by Dunn (1978) and by Schroeder (1980).

Area of Origin of Weed

The genus Carduus is native to the Eastern Hemisphere, where its distribution extends over Europe, central Asia, and East Africa. Franco (1976) recognized 48 species in Flora Europaea. Several taxa have been reported in North America and separated into three groups: the slender-flowered thistles (Carduus tenuiflorus Curt. and Carduus pycnocephalus L.), the small-flowered thistles (Carduus acanthoides L. and Carduus crispus L.), and the large-flowered thistle (Carduus nutans s.l.). Southern Europe is considered to be the center of origin for Carduus because of the many endemic Carduus species found there.

Areas Surveyed for Natural Enemies

Areas surveyed included southern England, France, Austria, Germany, Italy and the northern part of the former Yugoslavia (Zwölfer, 1965; Boldt and Campobasso, 1978). Other surveys have been carried out in Pakistan, Iran, and Japan (Schroeder, 1980).

Natural Enemies Found

Some 130 insect species have been recorded on C. nutans s.l. in Europe (Zwölfer, 1965; Boldt and Campobasso, 1978). In Italy alone, 109 species from six orders and 33 families fed or reproduced on musk thistle. Some 25 species were reported to be broadly oligophagous on plants in the subtribe Carduinae (Table 1), and only very few were considered to have a host range restricted to plants in the genera Carduus, Cirsium, and Silybum, or to be monophagous. Since there was no concern about non-target impact on native thistles in the earliest phase of the program, oligophagy on several thistle species in the genera Carduus, Cirsium, and Silybum was considered as an advantage and only those species recorded as economic pests were eliminated from further consideration. After a few other candidate biological control agents had been discarded on the grounds that they did little damage to the target weeds, fewer than 10 species were considered as potential biological control agents of Carduus species and bull thistle, Cirsium vulgare (Savi) Tenore. Preference was given to seed-feeding insects for biological control of Carduus spp. and bull thistle because these weeds are short-lived species and reproduce by seeds. In contrast, defoliating beetles were selected for the perennial thistle Cirsium arvense (L.) Scop. (see Chapter on Canada thistle).

Table 1. Oligophagous Arthropods (Restricted to Carduinae) Recorded on Selected European Thistles (With Contributions from A. McClay)

In 1964, the seed-feeding weevil Rhinocyllus conicus (Frölich) was the first insect selected for biological control of Carduus species. Zwölfer (1971) believed that because of R. conicus‘ high egg potential and a tendency to disperse its eggs, this weevil should exert strong pressure on its host plant, especially after the weevil was released from limitation by its coevolved competitors and parasitoids. Shortly after biological studies had started with R. conicus, the rosette weevil Trichosirocalus horridus (Panzer) and the two rosette beetles Ceutorhynchus trimaculatus (F.) and Psylliodes chalcomera (Illiger) also were considered because they occupy different food niches and have different phenologies. Concern about non-target impact was increasing and, in the early 1980s, permission for field release of C. trimaculatus and P. chalcomera was denied. Consequently, more specific species were selected to complement the impact of R. conicus and T. horridus. The syrphid root-crown fly Cheilosia corydon (Harris) has the same feeding niche as T. horridus but it has a different phenology. The seed-feeding tephritid fly Urophora solstitialis (L.) was selected for biological control of C. acanthoides because R. conicus was not well synchronized with this thistle in many parts of North America (Surles and Kok, 1977). Later, Dunn and Campobasso (1993) showed that native North American Cirsium species were not exploited by P. chalcomera under field test conditions, and this flea beetle was finally released in the United States in 1997. Thistle insects discovered in Asia have not been exploited yet. The host specificity of Terrelia serratulae L., a trypetid fly from Pakistan, has been examined (Baloch and Khan, 1973), but it has not been considered further.

Host Range Tests and Results

Rhinocyllus conicus (Frölich). Field host records for the seed-feeding weevil R. conicus in Europe include thistles in several genera in the subtribe Carduinae (Carduus, Cirsium, Sylibum, and Onopordum)(Zwölfer and Harris, 1984). The plant species tested in the screening trials in the 1960s included primarily agricultural crops and horticultural species in the Asteraceae family, plus a few European thistles. Since the cultivated plants tested (Cynara scolymus L., Carthamus tinctorius L., Helianthus annuus L., Lactuca sativa L.) were not used by the weevil, and the potential use of native North American Cirsium species was not a concern at that time, R. conicus was approved and released in Canada (in 1968) and in the United States (in 1969). Feeding by R. conicus on native Cirsium species in North America was first reported by Laing and Heels (1978) and Rees (1978). Rhinocyllus conicus has been reported in flowerheads of nearly 20 native Cirsium spp. in the west and in the central plains and mountains (Louda, 2000). Genetic variation among populations of R. conicus does exist, but its role in host plant use is not well understood. The concept of host races associated with the main thistle species in Europe (Zwölfer and Preiss, 1983) has been challenged recently (Klein and Seitz, 1994; Briese, 1996).

Trichosirocalus horridus (Panzer). Field records of the rosette weevil T. horridus in Europe include a few genera in the subtribe Carduinae (Carduus, Cirsium, Onopordum, and Galactites). Host range studies were carried out in the late 1960s and early 1970s (Ward et al., 1974; Kok, 1975). As for R. conicus, the plant species tested included cultivated plants and a few European thistles. Some larval feeding occurred on lettuce (L. sativa), cauliflower (Brassica oleracea L.) and artichoke (C. scolymus), but none of these species supported normal larval development. Preferred hosts were species of Carduus, Cirsium, and Onopordum. Trichosirocalus horridus has only occasionally been reported to feed and develop on native North American thistles (McAvoy et al., 1987).

Cheilosia corydon (Harris). In Europe, the root-crown fly C. corydon has been reared from Carduus nutans s.l., Carduus crispus L., and Carduus pycnocephalus L., and rarely from Cirsium vulgare, Cirsium eriophorum (L.) Scop., and Cirsium palustre (L.) Scop. In laboratory tests, larvae survived on all six Carduus species tested as well as on the native North American species, Cirsium crassicaule (Greene) Jeps. None of the other nine Cirsium species (including six native North American species) were suitable for C. corydon development. In field trials in Italy, oviposition was recorded on Carduus nutans but not on the seven native Cirsium species tested (Rizza et al., 1988).

Ceutorhynchus trimaculatus (Fab.). Field records of this thistle-rosette weevil in Europe include Carduus spp., Cirsium spp., Onopordum spp., Silybum marianum (L.) Gaertn., and Galactites tomentosa Moench (Boldt et al., 1980). Ceutorhynchus trimaculatus was found to complete development on artichoke (C. scolymus) and several Cirsium species in quarantine screening tests (Kok et al., 1979, 1982; Kok and McAvoy, 1983). In field tests carried out in Italy in 1984 and 1985, larvae of C. trimaculatus were found on all three North American native Cirsium spp. exposed, but not on artichoke (Dunn and Campobasso, 1993).

Psylliodes chalcomera (Illiger). Under experimental conditions, adult feeding, oviposition, and larval development by this thistle-rosette weevil occurred on European Carduus and Cirsium species (Dunn and Rizza, 1977). Adult feeding and oviposition, but no larval development, were recorded on artichoke under no-choice conditions. In field tests carried out in Italy between 1987 and 1989, this flea beetle did not use any of the three North American Cirsium species offered (Dunn and Campobasso, 1993).

Puccinia carduorum Jacky. This rust has been accidentally introduced to North America and also was the first plant pathogen tested and released in the United States for biological control of musk thistle. In greenhouse tests, limited infection occurred on some species of Cirsium, Cynara, Saussurea, and Sylibum, but older plants were resistant. Attempts to maintain P. carduorum on 22 native North American species of Cirsium and C. scolymus failed. Musk thistle was the only host that became severely diseased (Politis et al., 1984; Bruckart et al., 1996). No rust development was observed on any of the non-target plants (10 North American Cirsium spp. and artichoke) in a field trial carried out in 1988 in Virginia (Baudoin et al., 1993). Puccinia carduorum has not been reported from native North American Cirsium species. It has spread rapidly in the eastern United States and was found in Missouri in 1994 (Baudoin and Bruckart, 1996). It can be transmitted by the thistle insects R. conicus, T. horridus, and Cassida rubiginosa Müller (Kok and Abad, 1994).

Releases Made

Information in this section is from Rees et al., 1996; Julien and Griffiths, 1999; and shipment records of L. T. Kok.

Rhinocyllus conicus. Introductions of R. conicus from eastern France via Canada began in 1969 in the United States with releases in Virginia, California, Montana, and Nebraska. Following excellent results in Virginia, weevils were collected in Virginia and released in most of the thistle-infested 48 contiguous states. These included Alabama, Illinois, Indiana, Iowa, Kansas, Kentucky, Maryland, Minnesota, Missouri, New Jersey, New York, Ohio, Oklahoma, Pennsylvania, Tennessee, Texas, West Virginia, and more recently in the southern states of Alabama, Georgia, and North Carolina.

Trichosirocalus horridus. This species was first released in Virginia in 1974 (Kok and Trumble, 1979). Weevils collected from Virginia were subsequently released in many other states, including Alabama, Colorado, Georgia, Illinois, Indiana, Kansas, Kentucky, Maryland, Montana, North Carolina, Oklahoma, Tennessee, Texas, Washington, Wyoming, several western states, and also in Argentina and Canada.

Cheilosia corydon. This fly has been released in low numbers in Maryland, New Jersey, Montana, Nevada, Oregon, and Texas.

Urophora solstitialis. This species was released in 1996, only in Montana.

Ceutorhynchus trimaculatus: This species was not released because it feeds and develops on native Cirsium species (Kok et al., 1979, 1982; Kok and McAvoy, 1983).

Psylliodes chalcomera. This species was released in 1997, in Kansas and Texas (DeQuattro, 1997).

Puccinia carduorum. This pathogen was deliberately introduced in Virginia in 1987 (Baudoin et al., 1993), but had been accidentally introduced to North America before 1987 (Julien and Griffiths, 1999).

Biology and Ecology of Key Natural Enemies

Rhinocyllus conicus (Coleoptera: Curculionidae)

The biology of this seed-feeding weevil has been described by Zwölfer and Harris (1984). Following adult emergence from overwintering sites in litter and sheltered areas, mating and oviposition occur in spring and early summer. In Virginia, overwintered adult weevils (Fig. 3) were observed to become active in mid-to-late April (Surles and Kok, 1977). Eggs are laid externally on bud bracts (Fig. 4), either individually or in small clusters of two to five eggs. Caps of masticated host plant material, which appear as “warts,“ cover and protect the eggs from predation. Larvae hatch after six to nine days and bore through the bracts into the receptacle. Larvae feed on both the developing receptacles (Fig. 5) and the florets, pushing out characteristic tufts of hair from an infested head (Fig. 6), and sometimes the supporting peduncle under the head. Four larval instars complete development in about four to six weeks (Rowe and Kok, 1985). Larval feeding induces the formation of a gall-like callus of modified parenchyma tissue that provides the larvae with additional food and shelter (Shorthouse and Lalonde, 1984). Larval survivorship is strongly density-dependent, suggesting intraspecific competition causes much of the observed larval mortality (> 80%) within heavily infested inflorescences (Zwölfer, 1979). In North America, R. conicus has acquired a large number of parasitoids, but levels of parasitism are low (Rees, 1977; Goeden and Ricker, 1977, 1978; Puttler et al., 1978; Dowd and Kok, 1981, 1982, 1983; Smith and Kok, 1983). The pupal period is seven to 10 days, and pupae (Fig. 7) usually are found from mid-June through July. A partial second generation may be found in late August and September. Adults usually remain within pupation cells (Fig. 8) for several more weeks, before emerging to disperse to overwintering sites in litter. Phenology and life-cycle details vary geographically according to local climate. Zwölfer and Harris (1984) indicated that a partial second generation could occur for individuals that complete development early, if the photoperiod exceeds 16 hours.

Figure 3. Rhinocyllus conicus adult. (Photograph by L.-T.Kok.)		Figure 4. Rhinocyllus conicus eggs on thistle head. (Photograph by L.-T. Kok.)
Figure 5. Rhinocyllus conicus larva feeding on receptacle of thistle head. (Photograph by L.-T. Kok.)		Figure 6. Tufts of hair arising in R. conicus infested thistle head. (Photograph by L.-T. Kok.)
Figure 7. Rhinocyllus conicus pupa. (Photograph by L.-T.Kok.)		Figure 8. Thistle head showing pupation chambers of R. conicus. (Photograph by L.-T. Kok.)

Trichosirocalus horridus (Coleoptera: Curculionidae)

This rosette weevil has a single generation per year. Eggs are laid on the lower side of leaves along the midrib and the primary veins and hatch in about 13 days. Larvae migrate down the petiole to rosette crowns to feed soon after hatching. Mature larvae abandon the plant and enter the soil near the roots where they create pupation cells, made from silk and soil particles (Kok et al., 1975). In Virginia, oviposition occurs from mid-December until early April, and larvae are found in rosettes from late December (first instars) through late May (third instars) (Trumble and Kok, 1979). Trichosirocalus horridus may overwinter as an adult, egg, or larva (Kok and Mays, 1989). Teneral adults appear from mid-May through June and aestivate in July through September. This life cycle is similar to that of T. horridus in southern Europe, although the climatic conditions in southwestern Virginia resemble conditions of central Europe, where the life history of T. horridus is substantially different. In central Europe, oviposition of T. horridus occurs from the middle of May through June. Pupation occurs in July and August, and adults emerge in September and overwinter.

Cheilosia corydon (Diptera: Syrphidae)

In southern Europe, adults of this root-crown fly emerge at the end of February or March, and eggs are laid from mid-March to mid-April. Larvae feed in thistle crowns and large flower-bearing stems. Eggs are laid on young leaves in the center of the thistle rosette and young shoots. Newly hatched larvae mine directly into tender, young shoots. As shoots grow, the second and third instars mine up and down the stems. There are three larval instars. In May, larvae tunnel into the shoot base and the root. Pupation occurs in November (Rizza et al., 1988). Cheilosia corydon has one generation per year.

Ceutorhynchus trimaculatus (Coleoptera: Curculionidae)

This rosette weevil has one generation per year. Adults emerge at the end of April and feed on the leaves of new rosettes or mature plants for three to four weeks. At the end of May, weevils enter the soil to aestivate. Adults gradually become active again in autumn and feed on the leaves of young rosettes. Oviposition starts in November and continues through March or April. Larvae feed gregariously, boring into leaf buds or growing tips, and moving down into the crown. Pupation occurs in the soil (Boldt and Campobasso, 1981; Kok and McAvoy, 1983).

Psylliodes chalcomera (Coleoptera: Chrysomelidae)

In southern Europe, adults of this thistle-rosette flea beetle emerge in early June, feed heavily on maturing Carduus plants, and begin aestivation during late June. Aestivation ends in early November. Oviposition takes place between January and June. Eggs are laid at the base of plants or into soil adjacent to plants. Larvae feed on leaf buds and on young rosette leaves. Larvae mature in mid-May, and pupate in the soil nearby. In the laboratory, some females are long-lived and go through two aestivation and two oviposition periods (Dunn and Rizza, 1976).

Evaluation of Project Outcomes

Establishment and Spread of Agents (from Julien and Griffiths, 1999)

Rhinocyllus conicus. Establishment of this seed-feeding weevil has been confirmed in Iowa, Illinois, Kansas, Kentucky, Maryland, Minnesota, Missouri, North Dakota, South Dakota, Pennsylvania, Tennessee, Texas, New York, and Virginia as well as in several western states. In recent years, it also has become established in the southern states of Georgia (Buntin et al., 1993) and North Carolina (McDonald and Robbins, 1993). In Virginia, dispersal was only 1.6 km three years after release, but after six years, both eggs and adults were detected 32 km from the original release site (Kok and Surles, 1975).

Trichosirocalus horridus. Establishment of the rosette weevil was confirmed within two years of its release in Virginia study sites, and weevil populations had reached high levels by the third year. The weevil was found 27 km from release sites four years after initial introduction. By 1981, T. horridus was well established in the immediate release area and covered approximately 609 km². By 1985, the weevil had extended its range to 4,345 km² despite having had to move across forested areas where no thistles occur as well as areas with low thistle populations. Dispersal by flight probably occurs after aestivation during late summer or early fall (McAvoy et al., 1987). Trichosirocalus horridus also is established in North Carolina (McDonald and Robbins, 1993), Kansas, Maryland, Missouri, and several western states.

Cheilosia corydon. Establishment has not been confirmed.

Urophora solstitialis. Establishment has not been confirmed.

Psylliodes chalcomera. Establishment has not been confirmed.

Puccinia carduorum. This species is established in Virginia and Missouri (Baudoin et al., 1993; Baudoin and Bruckart, 1996) and was recorded in Wyoming in 1996.

Suppression of Target Weed

Rhinocyllus conicus. Effects of the weevil on C. nutans in Virginia were not apparent until 1973, after a steady increase in weevil densities. By 1974, 16 out of 20 releases resulted in successful establishment, and six showed more than 75% reduction in thistle density (Surles et al., 1974; Kok 1978a, b). Establishment rates were better for spring releases of reproductive adults than summer releases (Kok, 1974). At one location, 90% of the plants were heavily infested, and in 1975 all but one of the 11 plots showed at least 90% reduction in thistle density (Kok and Surles, 1975; Kok and Pienkowski, 1985). Biological control is usually achieved in five to six years (Kok and Surles, 1975; Kok, 1986; Kok and Mays, 1991) [Figs. 9, 10]. Decrease in thistle density was slower at sites with little competing vegetation. Grass competition was found to be important in restricting thistle growth and keeping weed population levels low. Control by R. conicus is enhanced when combined with proper land management, especially prevention of overgrazing.

Figure 9. Musk thistle site before R. conicus release. (Photograph by L.-T. Kok.)

Figure 10. Musk thistle site five years after R. conicus release. (Photograph by L.-T. Kok)

Trichosirocalus horridus. Damage results from larval feeding on meristematic tissues in the rosette, resulting in crown tissue necrosis. Cartwright and Kok (1985) found that C. nutans changed its growth pattern in response to feeding by T. horridus. Infested plants produced more stems and a larger crown than uninfested plants, which did not produce multiple stems in this study. Large thistles were stimulated by weevil damage to produce larger stems and more capitula, but small and medium thistles were shorter and produced fewer seeds and capitula than uninfested thistles. Response of thistles also is influenced by larval density (Sieburth et al., 1983). In Virginia, a 96% reduction of musk thistle density occurred at two of three study sites within six years of initial releases (Kok, 1986). The collapse of thistle populations after three years of heavy weevil attack was not unusual, as pasture plants re-established and reduced thistle recruitment. The extent of thistle reduction caused by T. horridus varies. If weevil populations are large and grass competition is strong, thistle densities can be reduced dramatically. Suppression of musk thistle growth is greatest when the two weevils (R. conicus and T. horridus) act in conjunction with plant competition. Tall fescue grass (Festuca arundinacea Schreb.) together with thistle weevils suppressed musk thistle growth more quickly than the use of thistle weevils alone (Kok et al.,1986).

Recommendations for Future Work

The musk thistle program has been reassessed recently (Nechols, 2000). For a long time, the debate has focused on the effect of the biological control agents on the population level of the target thistle and the degree of their non-target feeding. Of the five insects approved for release, two have established with certainty the seed head weevil, R. conicus, and the rosette weevil, T. horridus. Long-term impact studies conducted in Virginia (Kok, 1986; Kok and Mays, 1991) suggest that the two weevils are capable of exerting some control of C. nutans, although experimental data are generally lacking from most of the other states. Thus, long-term experiments are needed in which post-dispersal seed mortality, various levels of plant competition, and the impact of both weevils (alone and combined) are considered.

Both the seed head and the rosette weevils have relatively broad host ranges. In addition to various exotic thistles, R. conicus feeds and develops in nearly 20 native North American Cirsium species, and in some cases, heavy infestations cause significant reduction in seed (Louda, 2000). There is considerable controversy over whether or not biotypes R. conicus (or other thistle head insects like U. solstitialis, reviewed by Gassmann and Louda, 2000) might exist, each with a somewhat narrower host range. The existence or absence of such biotypes has important implications in the biological control program against C. nutans and other exotic thistles in North America. This controversy might be due in part to the lack of an accepted definition of the term itself, but the existence of weevil biotypes with inherited differences in their ability to use different hosts still needs to be demonstrated. Genetic variation occurs in R. conicus reared from different thistle species (Unruh and Goeden, 1987), but the extent to which this genetic variation drives host selection and acceptance is unknown. Rather, the evidence available to date suggests that the phenology of thistle species in the subtribe Carduinae plays a major role in their exploitation by R. conicus. Therefore, the redistribution of R. conicus in areas where the weevil has not spread naturally should not be considered without an ecological assessment of the targeted area.

In contrast to R. conicus, Trichosirocalus horridus has been reported only occasionally from native North American Cirsium species (McAvoy et al., 1987). In light of available evidence to date, two questions need to be considered. (1) Is intensive exploitation of native thistles by T. horridus just a matter of time even though it has not been commonly found on non-target weeds after 25 years of release? (2) Is the exploitation of native Cirsium by R. conicus the result of the broad diet of the weevil, or the combination of phenology, host plant affinities, and other biological characteristics? The availability of reproduction sites (synchronization with flowering periods of “any” thistles) rather than preference, weevil aggregation, or altered competitive ability of R. conicus in the flower heads of thistles may play an important role in the exploitation of native Cirsium species by R. conicus (Gassmann and Louda, 2000). If this is the case, it follows that insects with biological characteristics different from those of R. conicus, such as T. horridus, will not necessarily exploit native North American Cirsium species in the same way as R. conicus.

References

Baloch, G. M. and A. G. Khan. 1973. Biological Control of Carduus sp. II. Phenology, Biology and Host
     Specificity of Terellia serratulae L. (Dipt.: Trypetidae). Technical Bulletin 16: 11-22. Commonwealth
     Institute of Biological Control, Slough, United Kingdom.

Baudoin, A. B. A. M. and W. L. Bruckart. 1996. Population dynamics and spread of Puccinia
     carduorum in the eastern United States. Plant Disease 80: 1193-1196.

Baudoin, A. B. A. M., R. G. Abad, L. T. Kok, and W. L. Bruckart. 1993. Field evaluation of Puccinia
     carduorum for biological control of musk thistle. Biological Control 3: 53-60.

Boldt, P. and G. Campobasso. 1978. Phytophagous insects on Carduus macrocephalus in Italy.
     Environmental Entomology 7: 904-909.

Boldt, P. E. and G. Campobasso. 1981. Biology of two weevils, Ceutorhynchus trimaculatus and
     Trichosirocalus horridus, on Carduus spp. in Europe. Environmental Entomology 10: 691-696.

Boldt, P. E., G. Campobasso, and E. Colonnelli. 1980. Palearctic distribution and host plants of
     Ceutorhynchus trimaculatus and Trichosirocalus horridus (Coleoptera: Curculionidae). Annals of the
     Entomological Society of America 73: 694-698.

Bremer, K. 1994. Asteraceae. Cladistics and Classification. Timber Press, Portland, Oregon, USA.

Briese, D. T. 1996. Phylogeny: can it help us to understand host choice by biological weed control
     agents? pp. 63-70. In Moran, V.C and J. H. Hoffmann (eds.). Proceedings of the 9th International
     Symposium On Biological Control of Weeds, 19-26 January 1996, Stellenbosch. University of Cape
     Town, Cape Town, South-Africa,.

Bruckart, W. L., D. J. Politis, G. Defago, S. S. Rosenthal, and D. M. Supkoff. 1996. Susceptibility of
     Carduus, Cirsium, and Cynara species artificially inoculated with Puccinia carduorum from musk
     thistle. Biological Control 6: 215-221.

Buntin, G. D., R. D. Hudson, and T. R. Murphy. 1993. Establishment of Rhinocyllus conicus
     (Coleoptera: Curculionidae) in Georgia for control of musk thistle. Journal of Entomological Science
     28: 213-217.

Cartwright, B. and L. T. Kok. 1985. Growth responses of musk and plumeless thistles (Cardus nutans
     and C. acanthoides) to damage by Trichosirocalus horridus (Coleoptera: Curculionidae). Weed
     Science 33: 57-62.

DeQuattro, J. 1997. Alticine release on Musk thistle. Chrysomela Newsletter 30: 7.

Desrochers, A. M., J. F. Bain, and S. I. Warwick. 1988a. The biology of Canadian weeds. 89. Carduus
     nutans L. and Carduus acanthoides L. Canadian Journal of Plant Science 68: 1053-1068.

Desrochers, A. M., J. F. Bain, and S. I. Warwick. 1988b. A biosystematic study of the Carduus nutans
     complex in Canada. Canadian Journal of Botany 66: 1621-1631.

Dowd, P. F. and L. T. Kok. 1981. Predators of Rhinocyllus conicus (Coleoptera: Curculionidae) in
     Virginia. Environmental Entomology 10: 136-138.

Dowd, P. F. and L. T. Kok. 1982. Parasitism of Rhinocyllus conicus in Virginia. Environmental
     Entomology 11: 71-77.

Dowd, P. F. and L. T. Kok. 1983. Microorganisms isolated from Rhinocyllus conicus, a weevil
     introduced into Virginia for thistle control. Virginia Journal of Science 34: 11-13.

Dunn, P. 1976. Distribution of Carduus nutans, C. acanthoides, C. pycnocephalus, and C. crispus in the
     United States. Weed Science 24: 518-524.

Dunn, P. H. 1978. History of the biological control of musk thistle in North America and studies with the
     flea beetle Psylliodes chalcomera, pp. 1-6. In Frick, K. E. (ed.) Biological Control of Thistles in the
     Genus Carduus in the United States. U.S. Department of Agriculture, Stoneville, Mississippi, USA.

Dunn, P. H. and G. Campobasso. 1993. Field test of the weevil Hadroplonthus trimaculatus and the
     fleabeetle Psylliodes chalcomera against musk thistle (Carduus nutans). Weed Science 41: 656-
     663.

Dunn, P. H. and A. Rizza. 1976. Bionomics of Psylliodes chalcomera, a candidate for biological control
     of musk thistle. Annals of the Entomological Society of America 69: 395-398.

Dunn, P. H. and A. Rizza. 1977. Host specificity of Psylliodes chalcomera, a candidate for biological
     control of musk thistle. Environmental Entomology 6: 449-454.

Franco, J. do Amaral. 1976. Carduus, pp. 220-232. In Tutin, T.G., V. H. Heywood, N. A. Burges, D. M.
     Moore, D. H. Valentine, S. M. Walters, and D.A. Webb (eds.). Flora Europaea Vol. 4. Cambridge
     University Press, Cambridge, United Kingdom.

Freese, G. 1993. The insect complexes associated with the stems of seven thistle species.
     Entomologia Generalis 19: 191-207.

Gassmann, A. and S. M. Louda. 2000. Rhinocyllus conicus: initial evaluation and subsequent ecological
     impacts in North America, pp. 147-183. In Wajnberg, E., J. K. Scott, and P. C. Quimby (eds.).
     Evaluating Indirect Ecological Effects of Biological Control. CABI-Bioscience, Wallingford,
     United Kingdom.

Goeden, R. D. 1974. Comparative survey of the phytophagous insect fauna of Italian thistle (Carduus
     pycnocephalus) in southern California and southern Europe relative to biological weed control.
     Environmental Entomology 3: 464-474.

Goeden, R. D. and D. W. Ricker. 1977. Establishment of Rhinocyllus conicus on milk thistle in southern
     California. Weed Science 25: 288-292.

Goeden, R. D. and D. W. Ricker 1978. Establishment of Rhinocyllus conicus on Italian thistle.
     Environmental Entomology 7: 787-789.

Julien, M. H. and M. W. Griffiths (eds.). 1999. Biological Control of Weeds. A World Catalogue of
     Agents and their Target Weeds, 4th ed. CABI-Bioscience, Wallingford, United Kingdom.

Klein, M. and Seitz, A. 1994. Geographic differentiation between populations of Rhinocyllus conicus
     Froelich (Coleoptera: Curculionidae): concordance of allozyme and morphometric analysis.
     Zoological Journal of the Linnean Society 110: 181-191.

Kok, L. T. 1974. Efficacy of spring releases in colonization of Rhinocyllus conicus for biocontrol of
     thistles. Environmental Entomology 3: 429-430.

Kok, L. T. 1975. Host specificity studies on Ceuthorhynchidius horridus (Panzer) (Coleoptera:
     Curculionidae) for the biocontrol of musk and plumeless thistles. Weed Research 15: 21-26.

Kok, L. T. 1978a. Biological control of Carduus thistles in northeastern U.S.A., pp. 101-104. In
     Freeman, T. E. (ed.). Proceedings of the 4th International Symposium on Biological Control of
     Weeds. IFAS, University of Florida, Gainesville, Florida, USA

Kok, L. T. 1978b. Status of biological control of musk thistle in Virginia, pp. 23-30. In Frick. K. E. (ed.).
     Biological Control of Thistles in the Genus Carduus in the United States. U.S. Department of
     Agriculture, Stoneville, Mississippi, USA.

Kok, L. T. 1986. Impact of Trichosirocalus horridus (Coleoptera: Curculionidae) on Carduus thistles in
     pastures. Crop Protection 5: 214-217.

Kok, L. T. and R. G. Abad. 1994. Transmission of Puccinia carduorum by the musk thistle herbivores,
     Cassida rubiginosa (Coleoptera: Chrysomelidae), Trichosirocalus horridus and Rhinocyllus conicus
     (Coleoptera: Curculionidae). Journal of Entomological Science 29: 186-191.

Kok, L. T. and W. T. Mays. 1989. Comparison of the seasonal occurrence of Trichosirocalus horridus
     (Panzer) (Coleoptera: Curculionidae) in Virginia between 1981-83 and 1979. Journal of Entomological
     Science 24: 465-471.

Kok, L. T. and W. T. Mays. 1991. Successful biological control of plumeless thistle, Carduus
     acanthoides L. [Campanulatae: Asteraceae (=Compositae)], by Trichosirocalus horridus (Panzer)
     (Coleoptera: Curculionidae) in Virginia. Biological Control 1: 197-202.

Kok, L. T. and T. J. McAvoy. 1983. Biological studies of Ceutorhynchus trimaculatus (Coleoptera:
     Curculionidae), a thistle rosette weevil. Canadian Entomologist 115: 927-932.

Kok, L. T. and R. L. Pienkowski. 1985. Biological control of musk thistle by Rhinocyllus conicus
     (Coleoptera: Curculionidae) in Virginia from 1969 to 1980, pp. 433-438. In Delfosse, E.S. (ed.).
     Proceedings of the 6th International Symposium on Biological Control of Weeds, 19-25 August
     1984, Vancouver,. British Columbia. Agriculture Canada, Ottawa.

Kok, L. T. and W. W. Surles. 1975. Successful biocontrol of musk thistle by an introduced weevil,
     Rhinocyllus conicus. Environmental Entomology 4: 1025-1027.

Kok, L. T. and J. T. Trumble. 1979. Establishment of Ceuthorhynchidius horridus (Coleoptera:
     Curculionidae), an imported thistle-feeding weevil, in Virginia. Environmental Entomology 8: 221-223.

Kok, L. T., R. H., Ward, and C. C. Grills. 1975. Biological studies on Ceuthorhynchidius horridus
     (Panzer), an introduced weevil for thistle control. Annals of the Entomological Society of America 68:
     503-505.

Kok, L. T., L. A. Andres, and P. E. Boldt. 1979. Host specificity studies on Ceutorhynchus trimaculatus
     F. (Col.: Curculionidae), a potential biological agent of musk and plumeless thistle. Environmental
     Entomology 8: 1145-1149.

Kok, L. T., T. J. McAvoy, G. R. Johnson, and P. H. Dunn. 1982. Further tests on Ceutorhynchus
     trimaculatus as a candidate for the biological control of Carduus thistles. Crop Protection 1: 67-74.

Kok, L. T., T. J. McAvoy, and W. T. Mays. 1986. Impact of tall fescue grass and Carduus thistle weevils
     on the growth and development of musk thistle (Carduus nutans). Weed Science 34: 966-971.

Laing, J. E. and P. R. Heels. 1978. Establishment of an introduced weevil Rhinocyllus conicus
     (Coleoptera : Curculionidae) for the biological control of nodding thistle (Compositae) in southern
     Ontario. Proceedings of the Entomological Society of Ontario 109: 3-8.

Louda, S. M. 2000. Negative ecological effects of the musk thistle biological control agent, Rhinocyllus
     conicus, pp. 215-243. In Follett, P.A. and J. J. Duan, (eds.). Nontarget Effects of Biological Control.
     Kluwer Academic Publishers, Boston, Massachusetts, USA.

McAvoy, T. J., L. T. Kok, and W. T. Mays. 1987. Dispersal of Trichosirocalus horridus (Panzer)
     (Coleoptera. Curculionidae) in southwest Virginia. Journal of Entomological Science 22: 324-329.

McCarty, M. K. 1978. The genus Carduus in the United States, pp. 7-10. In Frick, K. E. (ed.). Biological
     Control of Thistles in the Genus Carduus in the United States. U.S. Department of Agriculture,
     Stoneville, Mississippi, USA.

McDonald, R. C. and N. S. Robbins. 1993. Establishment and efficacy of Rhinocyllus conicus Froelich
     (Coleoptera: Curculionidae) in controlling Carduus nutans L. in North
     Carolina, pp: 30-39. In Anon. North Carolina Department of Agriculture Biological Control Laboratory
     Report of Activities, 1993. North Carolina Department of Agriculture, Raleigh, North Carolina, USA.

McDonald, R. C., K. A. Kidd, and N. S. Robbins. 1994. Establishment of the rosette weevil,
     Trichosirocalus horridus (Panzer) (Coleoptera: Curculionidae) in North Carolina. Journal of
     Entomological Science 29: 302-304.

McGregor, R. L. 1986. Great Plains Flora Association, Flora of the Great Plains. University Press of
     Kansas, Lawrence, Kansas, USA. (see 895-897).

Maw, M. G. 1976. An annotated list of insects associated with Canada thistle (Cirsium arvense) in
     Canada. Canadian Entomologist 108: 235-244.

Nechols, J. R. 2000. Biological control of musk thistle: a reassessment, pp. 245-259. In Follett, P. A.
     and J. J. Duan, (eds.). Nontarget Effects of Biological Control. Kluwer Academic Publishers,
     Boston,Massachusetts, USA.

Petanovic, R., J. Boczek, and B. Stojnic. 1997. Taxonomy and bioecology of eriophyids (Acari:
     Eriophyoidea) associated with Canada thistle, Cirsium arvense (L.) Scop. Acarologia 38: 181-191.

Politis, D. J., A. K. Watson, and W. L. Bruckart. 1984. Susceptibility of musk thistle and related
     composites to Puccinia carduorum. Phytopathology 74: 687-691.

Puttler, B., S. H. Long, and E. J. Peters. 1978. Establishment in Missouri of Rhinocyllus conicus for the
     biological control of musk thistle (Carduus nutans). Weed Science 26: 188-190.

Redfern, M. 1983. Insects and Thistles. Naturalists’ Handbooks 4. Richmond Publishing, Slough,
     United Kingdom.

Rees, N. E. 1977. Impact of Rhinocyllus conicus on thistles in southwestern Montana. Environmental
     Entomology 6: 839-842.

Rees, N. E. 1978. Interactions of Rhinocyllus conicus and thistles in the Gallatin Valley, pp. 31-38. In
     Frick, K. E. (ed.). Biological Control of Thistles in the Genus Carduus in the United States. U.S.
     Department of Agriculture, Stoneville, Mississippi, USA.

Rees, N. E., P. C. Quimby, Jr., G. L. Piper, E. M. Coombs, C. E. Turner, N. R. Spencer, and L. V.
     Knutson. 1996. Biological Control of Weeds in the West. Western Society of Weed Science,
     Montana State University, Bozeman, Montana, USA.

Rizza, A., G. Campobasso, P. H. Dunn, and M. Stazi. 1988. Cheilosia corydon (Diptera: Syrphidae), a
     candidate for the biological control of musk thistle in North America. Annals of the Entomological
     Society of America 81: 225-232.

Rowe, D. J. and L. T. Kok. 1985. Determination of larval instars, and comparison of field and artificial
     diet-reared larval stages of Rhinocyllus conicus (Col.: Curculionidae). Virginia Journal of Science 36:
     277-280.

Schroeder, D. 1980. The biological control of thistles. Biocontrol News and Information 1: 9-26.

Shorthouse, J.D. and R. G. Lalonde. 1984. Structural damage by Rhinocyllus conicus (Coleoptera:
     Curculionidae) within the flowerheads of nodding thistle. Canadian Entomologist 116: 1335-1343.

Sieburth, P. J., L. T. Kok, and M. Lentner. 1983. Factors influencing the effectiveness of Trichosirocalus
     horridus (Panzer) in the control of Carduus thistles. Crop Protection 2: 143-151.

Smith, L. M. and L. T. Kok. 1983. Hyalomydes triangulifer (Diptera: Tachinidae) parasitizes adults of
     Rhinocyllus conicus (Coleoptera: Curculionidae). Canadian Entomologist 115: 1533-1534.

Smith, L. M. and L. T. Kok. 1984. Dispersal of musk thistle (Carduus nutans) seeds. Weed Science 32:
     120-125.

Story, J. M., H. DeSmet-Moens, and W. L. Morrill. 1985. Phytophagous insects associated with
     Canada thistle, Cirsium arvense (L.) Scop., in southern Montana. Journal of the Kansas
     Entomological Society 58: 472-478.

Stuckey, R. L., and J. L. Forsyth. 1971. Distribution of naturalized Carduus nutans (Compositae)
     mapped in relation to geology in northwestern Ohio. Ohio Journal of Science 71: 1-15.

Surles, W. W. and L. T. Kok. 1977. Ovipositional preference and synchronization of Rhinocyllus conicus
     with Carduus nutans and C. acanthoides. Environmental Entomology 6: 222-224.

Surles, W. W., L. T. Kok, and R. L. Pienkowski. 1974. Rhinocyllus conicus establishment for biocontrol
     of thistles in Virginia. Weed Science 22: 1-3.

Trumble, J. T. and L. T. Kok. 1979. Ceuthorhynchidius horridus (Coleoptera: Curculionidae): life cycle
     and development on Carduus thistles in Virginia. Annals of Entomological Society of America 72:
     563- 564.

Trumble, J. T. and L. T. Kok. 1982. Integrated pest management techniques in thistle suppression in
     pastures in North America. Weed Research 22: 345-359.

Unruh, T. R. and R. D. Goeden. 1987. Electrophoresis helps to identify which race of the introduced
     weevil, Rhinocyllus conicus (Coleoptera: Curculionidae), has transferred to two native southern
     California thistles. Environmental Entomology 16: 979-983.

USDA, NRCS. 1999. United States Department of Agriculture, Natural Resources Conservation Service.
     Plants Database. http://plants.usda.gov (accessed in 2000).

Ward, R. H., R. L. Pienkowski, and L. T. Kok. 1974. Host specificity of the first-instar of
     Ceuthorhynchidius horridus, a weevil for the biological control of thistles. Journal of Economic
     Entomology 67: 735-737.

Warwick, S. L., B. K. Thompson, and L. D. Black. 1990. Comparative growth response in Carduus
     nutans, C. acanthoides, and their F1 hybrids. Canadian Journal of Botany 68: 1675-1679.

Zwölfer, H. 1965. A List of Phytophagous Insects Attacking Wild Cynareae Species in Europe.
     Progress Report 15, Commonwealth Institute of Biological Control, Slough, United Kingdom.

Zwölfer, H. 1971. Competition and coexistence in phytophagous insects attacking the heads of Carduus
     nutans L., pp. 74-81. In Dunn, P. (ed.). Proceedings of the 2nd International Symposium on
     Biological Control of Weeds, October 1971, Rome, Italy. Commonwealth Institute of Biological
     Control, Slough, United Kingdom.

Zwölfer, H. 1979. Strategies and counter strategies in insect population systems competing for space
     and food in flowerheads and plant galls. Fortschitte der Zoologie 25: 331-353.

Zwölfer, H. and P. Harris. 1984. Biology and host specificity of Rhinocyllus conicus (Froel.) (Col.,
     Curculionidae), a successful agent for biocontrol of the thistle, Carduus nutans L. Journal of Applied
     Entomology 97: 36-62.

Zwölfer, H. and M. Preiss. 1983. Host selection and oviposition behaviour in west-European ecotypes of
     Rhinocyllus conicus Fröl. (Col.: Curculionidae). Journal of Applied Entomology 95: 113-122.

[  Home  ]  [  Contents  ]   [  Previous  ]   [  Next  ]