Cirsium arvense (L.) Scop., Creeping Thistle

Account Summary

Growth form and preferred habitats

C. arvense is by far the most persistent thistle in B & I, being a true polycarpic perennial. The overwintering organ is an extremely vigorous, rapid and far-spreading root system that can expand up to 6 m per year and penetrate to a soil depth of 200 cm on occasions. The plant does not possess a rhizome, nor any stolons, both of which are horizontally spreading shoot tissues found in many other species capable of spread or clump formation (Donald 1994). Instead, a thickened, horizontally spreading, propagative root system produces numerous erect, aerial, adventitious shoots at intervals from subterranean adventitious root buds, enabling the formation of extensive clonal clumps of the thistle that can become locally dominant and which are extremely difficult, or almost impossible, to eradicate. There is evidence from observations of clonal patches, that Creeping Thistle displays thinning shoot density and lessening height at the centre of older stands, strongly suggesting that local senescence takes place (Donald 1994).

The root connections linking new aerial shoots to the parent stock degenerate after about two years, resulting in the creation of separate, clonal, daughter plants. Thus shoot height and density are reckoned to be greatest about 3 m back from the advancing front of a thistle patch, and stress conditions such as drought can give rise to the creation of a hollow centre and the formation of a ring-like colony or clone (Tiley 2010). Even very small portions of the perennating root system can regenerate the whole plant and fragments can readily be transported in root-contaminated soil (Moore 1975).

The erect, slender, grooved flowering stems, 30-150 cm tall, are almost hairless when young but become increasingly hairy with age (Holm et al. 1977).

Like the even more common Cirsium palustre (Marsh Thistle), Creeping Thistle has small flower-heads, 15-25 mm across, which are clustered, but it differs from the latter in that the stem is not winged. The upper leaves, however, clasp the stem and the leaf-base may run down the stem a short way forming a spiny wing, though this is never continuous as is the case in C. palustre (Donald 1994).

The characteristic deep root system allows the plant to survive in moderately dry or periodically droughted situations, but it means the species cannot cope with sites where the soil regularly becomes waterlogged or there is a high water-table. It also avoids shallow, stony soils (Tiley 2010).

C. arvense is primarily a lowland species of temperate agricultural land and it can readily become locally dominant in disturbed or tilled ground, and particularly so in poorly managed, dry, fertile meadows and pastures. This agricultural menace is one of five injurious weeds listed in the 1959 UK Weeds Act, and is regarded as one of the world's worst weeds (Holm et al. 1977). A survey of English and Welsh farms carried out between 1974 and 1977 found C. arvense was a serious weed in 8% of pastures and, "it was more prevalent on beef than on dairy farms" (Peel & Hopkins 1980).

The extensive geographical distribution of C. arvense suggests it is adapted to grow on a wide variety of soil types. Although it favours moist, fertile substrates, Creeping Thistle avoids wet, poorly aerated and strongly acidic peaty soils and it tends to be absent from both woodland shade and the most exposed sites. In Germany, Korsmo (1930) noted that C. arvense was most persistent on agricultural land that contained chalk. On the other hand alkaline or high calcium soil horizons are believed to limit root development (Donald 1994).

The leaves and stems of C. arvense are heavily armed with abundant sharp spines that deter most browsing animals except horses, donkeys, goats and pigs. However, the spines become less of a deterrent on old plants and on those whose leaves have wilted for any reason, including those after cutting, trampling or other forms of physical damage, when they then may be readily be eaten (Mitchell & Abernethy 1995; Van Toor 1995). Cattle and sheep also graze Creeping Thistle in the autumn when the species is senescing and availability of alternative grazing is limited (Tiley 2010). At other times of year, heavy grazing of pastures containing C. arvense tends to favour the thistle since its spiny armour protects it and competition from surrounding grass and other weeds becomes reduced, since they, rather than the thistle, become the browser's target species and are eaten down and/or removed (Tiley 2010).

Fermanagh occurrence

Creeping Thistle is the second most frequent thistle in Fermanagh, being commonly recorded across 427 tetrads, 80.9% of those in the VC. It follows the most common and widespread thistle in the county, C. palustre (Marsh Thistle) which, for comparison, has been recorded in 494 tetrads, 93.6% of the VC tetrad total.

Probable geographical origin and fossil history

Based on the evidence of its widespread Eurasian temperate phytogeographical distribution (Preston & Hill 1997), C. arvense is believed to have originated in temperate parts of the Middle East and the Mediterranean, which also is the centre of origin of arable agriculture, and to have spread from there with early human migration and the expansion of farming cultures in Europe from east to west and south to north, and then across the Atlantic and into the Southern hemisphere (see below and Hultén & Fries 1986, Map 1862; Tiley 2010). The species was probably transported accidently as a weed contaminant of cereal and other crop seeds.

The earliest records of C. arvense in Britain are of fossil fruits (ie achenes or 'seeds') from phases of open vegetation at the beginning and end of glacial stages in the Pleistocene. These date from the Ipswichian and the Late Weichselian period zones I/II transition and zone III (Godwin 1975). Later, more recent fossil records are from late in the Flandrian, the current interglacial we are living in at present and come from archaeological diggings and date from the Late Bronze Age, through the Roman occupation to Mediaeval sites (Godwin 1975).

Vegetative reproduction

In C. arvense, seed is probably of lesser local dispersal significance in comparison with the amazing vegetative reproductive capability and local diffusion displayed by the species, ie it achieves a high degree of local dispersal, accomplished purely by lateral vegetative growth assisted by root fragmentation, enabling frequent accidental or incidental transport in soil through farming, soil and mud movement and building activities, and rapid re-establishment in fresh sites. It has been reported that pieces of either vertical or horizontal roots as small as 3-6 mm in length are capable of producing adventitious buds and can rapidly regenerate whole plants (Hayden 1934; Moore 1975).

Flowering reproduction

The lilac or pale whitish-purple flower-heads are generally considered dioecious, although the sexual separation is incomplete: the male flowers have an abortive ovary and the female ones abortive stamens. It is therefore not unknown for some clones to be bisexual (hermaphrodite) and self-fertile (Kay 1985). Male plants are the most common and some of them occasionally form achenes, meaning they are not totally male but, rather, they are semi- or partially-hermaphrodite (Hodgson 1968). The species might then be better described as being 'near-dioecious' or 'subdioecious'. Donald (1994) summarised the unusual sexual behaviour as follows, "Most flowers on male plants are not fertilized because pollen fails to adhere to the poorly developed papillae on stigmas of male flowers, even though male flowers produce well-formed egg cells. In female plants, anthers prematurely abort due to degeneration of the tapetum before the onset of meiosis.". The tapetum is the tissue layer in the wall of the anther sac that generates pollen grains.

Flowering takes place mainly from July to September and a large variety of insects, mainly honey bees but also Diptera (flies) and Sphecidae (wasps), make visits to the honey-scented flower heads, making this an essentially cross-bred, outbreeding species (Proctor et al. 1996). As a result of this breeding system there is a considerable range of genetic variation within the species and many ecotypes and five varieties are sometimes distinguished (Moore 1975; Grime et al. 1988, 2007; Sell & Murrell 2006). When both male and female plants are present and not more than about 30 m apart, viable fruits are produced freely, usually numbering from 20 to 200 per capitula (Salisbury 1964). Seed predation by insects before dispersal from the fruiting flower-head can seriously reduce the number of viable seeds capable of spreading and increasing the species. In one case, in Canada, 20 to 85% of seed-heads in a population sample were attacked in this way (Forsyth & Watson 1985).

The fruiting calyx of each floret produces a pappus of hairs but, unlike C. vulgare, it is very easily detached from the achene, so it may or may not assist the observed colonisation of new sites by seedlings (Grime et al. 1988, 2007). An experimental study showed that only about 1% of freshly collected seed germinated, and single figure percentages were recorded over the following twelve months for seeds in dry storage (Grime et al. 1981). Other experimental studies have shown good seed viability and seed ready to germinate immediately after release, up to 95% of it successfully emerging (eg Hayden 1934; Hodgson 1964). In a pot experiment, seed survival was especially good for deeply buried achenes (at depths of 55 and 105 cm). It was found that secondary dormancy developed, permitting survival for up to 21 years (Toole & Brown 1946). Thus some seed scientists think C. arvense is capable of forming a long-term persistent seed bank in the soil (Moore 1975), while others believe it germinates almost immediately upon release so that its seed may be absent from the soil seed bank altogether. Alternatively, seeds may only be found in small numbers (Roberts 1981). Other experiments found that seed could germinate when exposed on the soil surface (Wilson 1979) and fresh seed germinated best when sown at a depth of only 4-5 mm, appearing on average after eight to nine days (Lund & Rostrup 1901). In comparison with this, Kolk (1947) concluded the optimum depth for good levels of seedling emergence was 1 cm, although seedlings would emerge from depths up to 6 cm, the maximum depth that he tested.

Reflecting this divergence in experience, the authors of The soil seed banks of NW Europe survey found a variety of buried seed survival estimates for C. arvense, with numerous researchers suggesting the majority of seed germinated either immediately after release, or within twelve months of their production, while other findings indicated the long-term persistence of seed buried in soil for periods of up to five years, with seed densities of up to 1,200/m² recorded in some sites (Thompson et al. 1997).

Optimum seed germination took place at pH levels between 5.8 and 7.0, rates dropping off rapidly beyond these limits in both directions (Wilson 1979). Donald (1994) concluded that, "after seed maturation, primary dormancy is short lived, but it is followed by longer term environmentally enforced secondary dormancy.". Secondary dormancy develops if seed experience environmental conditions that prevent germination (Bakker 1960). In his review, Moore (1975) remarked that, "seedlings become established in recently ploughed or disturbed soils" and, "they survived only if competition is limited and light intensity is high".

Established strategy and competitive ability

Like other common perennial weeds such as Agrostis stolonifera (Creeping Bent), Elytrigia repens (= Elymus repens) (Common Couch) and Ranunculus repens (Creeping Buttercup), well known for their ability to multiply asexually, the established strategy of C. arvense was categorised by Grime (1977) as a perennial competitive ruderal (his C-R strategist). In plants of this type, selection operates to maximize the capture of resources in conditions where disturbance, even if it occurs fairly regularly, does not take place during the main growing season, or at least before the production of vegetative propagules. Their powers of lateral spread allow C-R plants to compete even more effectively for light, water, mineral nutrients and physical space than other plants and, indeed, the high density of often interconnected plants of the same species, minimises any risk of their elimination (Leakey 1981). Interestingly, Grime et al. (1988, 2007) later revised the established strategy to 'C', a straightforward Competitor, clearly regarding the competitive ability of the species as more significant in its establishment than the undoubted ruderal capabilities it possesses.

Solé et al. (2004) used molecular techniques to examine genotypic and genetic diversity in C. arvense populations and found they did not decline with time but, rather, they were maintained at a high level both within and between populations. As Tiley (2010) commented, "They attributed this finding to a combination of successful recruitment of sexually out-crossed seedlings in the early stages of succession, together with subsequent efficient vegetative reproduction, producing established clonal stands. The two means of reproduction, seed production and vegetative spread, resulted in efficient colonization of sites, followed by long-term local persistence, and thus working in tandem both contributed to the success of the species worldwide.".

British and Irish occurrence

As the species' hectad map in the BSBI Atlas 2020 very clearly demonstrates, C. arvense is almost ubiquitous in habitats disturbed or modified by man everywhere throughout B & I, being absent only from high ground over about 700 m in mountain regions. The mapped distribution remains stable at least at the hectad level, despite the long-term attempts of farmers and land managers across both islands working since the 1960s to limit and control what has been recognised as an 'injurious weed' under the UK government Weed Act, using herbicides and biological control efforts (F.H. Perring & S.M. Smart, in: Stroh et al. 2023).

European and world occurrence

C. arvense is regarded as native from N Africa northwards across the Mediterranean to N Scandinavia, although absent from Svalbard, Crete and the Macronesian region (ie the Azores, Madeira, the Savage islands, the Canaries and the Cape Verdes) (Press & Short 1994), and recognised as an introduced rare inventive in SW Greenland, and likewise in Iceland and the Faeroes (Ostenfeld & Gröntved 1934; Böcher et al. 1968; Löve 1983). It is also considered native in Asia Minor and W Asia eastwards through Afghanistan and Siberia to China and Japan. It has been widely introduced with agriculture across the world from N & S America (eg Chile), S Australia and New Zealand (Hultén & Fries 1986, Map 1862; Tiley 2010).

Weedy behaviour

Cirsium arvense was a troublesome weed throughout S Europe by the beginning of the 16th century and by the mid-18th century was very common throughout Europe (Dewey 1901). The weed was introduced to N America (where it is referred to as 'Canada Weed') very early in the colonial period, probably in the 17th century. It most likely arrived as a crop or pasture grass and clover seed contaminant, or with transported hay or straw, and it became so common and widespread that by 1795 control legislation was enacted in Vermont and in 1831 by New York (Moore 1975; Donald 1994).

C. arvense is no longer a really important weed in B & I, perhaps partially because it supports up to 100 species of herbivorous insects, as well as the rather moderate effect of attempted biological control measures. However, it remains a very successful and economically significant weed of temperate regions throughout both hemispheres in areas where it has been introduced during the last 400–500 years (Moore 1975).

Weed control measures

Attempts to eliminate or control C. arvense in N American cereal crops have shown that some ecotypes are resistant to herbicides such as 2,4-D and glyphosate. It can survive even after four annual pre-harvest applications of these chemicals combined with spring applications of MCPA (Darwent et al. 1994). Nevertheless, spring application of a broadleaf herbicide was effective in preventing seed production and this, combined with one or more pre-harvest applications of glyphosate, produced low densities of 'Canada Thistle'.

Other control experiments using disking tillage on established thistle populations with and without herbicide, found that tillage alone, while destroying or damaging emerged shoots, released other buds from dormancy and could actually increase the number of C. arvense shoots in dry land sites (Zimdahl & Foster 1993). This is a very active field of research and it may be that some form of biological control will eventually emerge.

Threats

An injurious and difficult or impossible weed to eradicate mechanically, it can be controlled by a combination of herbicide, ploughing and crop rotation or reseeding.

References

• Find out more about the format of the species accounts
• View abbreviations
• View references for all species accounts
• View a gazetteer of all the sites