Alliaria petiolata
(garlic mustard)

A. petiolata has spread throughout much of the north-eastern and mid-western USA and Canada after its introduction from Eurasia. The species invades forested communities and edge habitats. The plant has no known natural enemies in North America, has a broad ecological amplitude with considerable plasticity, is self-fertile, maintains a seed-bank, and is quite difficult to eradicate once established. There are risks of further introduction to similar climatic zones in other continents.

Biennial
Broadleaved
Herbaceous
Seed propagated

A. petiolata is a Eurasian native, with a very wide native range extending from China, through central Asia, to western Europe and North Africa. The species has become widely naturalized throughout North America and elsewhere.

Further spread throughout eastern North America is virtually guaranteed. The ecological amplitude of this species and its phenotypic plasticity will permit it to thrive in a variety of forested habitats. High elevation, desert, and warm habitats found in much of southern and western USA, are however not as suitable for the spread of this species. Welk et al. (2002) predicts that A. petiolata will expand its range through large parts of the Great Plains, including most of South Dakota and Nebraska, from northern Utah to southern Idaho, as well as northern California and Colorado. In Canada, A. petiolata is predicted to spread to the Gaspe Peninsula, Central British Columbia, and the western slopes of the Rocky Mountains in Alberta (Welk et al. 2002). Its continued availability from seed companies and culinary uses mean that further introduction to other similar habitats in other continents cannot be discounted.

In its native environment, the plant often grows singly in hedges and fencerows, open woods, and disturbed areas (Nuzzo, 2003); rarely considered an understorey 'dominant' and is a frequent component of ruderal vegetation communities. However, its habit within invaded areas of North America is completely different. There, it usually grows in moist, rich, forest environments and in dense patches. The species has broad ecological amplitude; it thrives in areas as diverse as mesic riparian areas to relatively xeric upland hardwood forests, where it performs best in partial shade and is less successful in full shade or full sun (Dhillion and Anderson, 1999; Meekins and McCarthy, 2001). Roads and river banks have been recorded as the primary collection sites of the species, but forested riparian areas, upland forests, and urban areas are also important (Nuzzo, 1993). Shaded, moist, rich sites appear to harbour the greatest densities.

A. petiolata is not normally a weed of croplands. A. petiolata is most often found in field margins and forests, often surrounding croplands. It is known as a strong competitor and has been demonstrated in the USA to out-compete other native herbs and woody plants (Meekins and McCarthy, 1999) and ultimately alter native community composition (McCarthy, 1997).

Genetics

The chromosome number of A. petiolata is generally considered to be 2n=42 (Gleason and Cronquist, 1991) in North America and Europe. However, the sporophytic chromosome number has also been reported as 14 in India (Naqshi and Javeid, 1976) and 36 in the Netherlands (Gadella and Kliphuios, 1976). Based on this information, A. petiolata is hypothesized to be a hexaploid species that is based on a base chromosome number of n=7 (Al-Shehbaz, 1988). In addition, A. petiolata is known to be highly plastic in different habitats showing broad variation in patterns of resource allocation (Byers and Quinn, 1998; Susko and Lovett-Doust, 1998, 1999, 2000a). The results of analysis of within- and among-population genetic variability were that 61% of the total variation occurred among populations, with much less (16.3%) between North American and Eurasian populations and within populations (22.1%); however, North American and Eurasian populations were found to be significantly different (Meekins et al., 2001).

Physiology and Phenology

A. petiolata generally acts as a biennial in North America (Cavers et al., 1979), but may occasionally respond as a winter annual in parts of its native range (Grime et al., 1988). Propagation is entirely by seed. Seeds of A. petiolata are dormant at maturity (mid-summer) and require cold stratification to come out of dormancy. Viability at maturity is often 100% (Anderson et al., 1996). In northern Europe and Canada, A. petiolata seeds are commonly reported to show an 18 month dormancy period (Cavers et al., 1979), whereas in the south of its range, only 6 months may be required (Lhotska, 1975; Baskin and Baskin, 1992). Optimal seed germination temperatures were observed at 16/6°C, and declined with progressively higher temperatures (Baskin and Baskin, 1992). Germination may occur in either the light or dark. Germination is nearly complete at the end of the first germination season (>95%); however, studies in unheated greenhouses and field experiments have shown that seeds may still germinate up to 4 years later (Baskin and Baskin, 1992; Anderson et al., 1996), so the formation of a seed-bank is possible. Following germination in the spring, A. petiolata forms a slender taproot, sometimes branched.

The photosynthetic rate of first-year plants has been examined by Dhillion and Anderson (1999). They found that the species seemed more characteristic of a shade-adapted species as compared to sun-adapted species, which may explain why it performs well in partially shaded forest habitats. During the second growing season, A. petiolata matures in early spring and produces a bolting flowering stem up to 1.5 m. The success of A. petiolata in invaded areas in North America may be due to its achieving maximum photosynthetic rates before the active growth of many native ground layer species when irradiance reaching the ground layer is high when the temperature and moisture conditions are favourable for the species (Myers and Anderson, 2003).

Flower buds are usually seen in early April when the plants begin to bolt (Anderson et al., 1996). The maximum number of buds is in mid April which often coincides with the date of first flowering. By early June, flower buds are usually absent, but occasional flowers may be borne in the leaf axils, though these have not been reported to ever bear fruit in the field (Anderson et al., 1996). Fruits may mature and dehisce seeds as early as mid-July. Anderson et al. (1996) show a bimodal pattern of seed production with peaks in mid-August and mid-September. In the USA, seed rain was approximately 15,000 seeds/m² in Illinois field populations (Anderson et al., 1996), but considerably greater seed production, up to 38,000 seed/m² has been observed in Ohio (Trimbur, 1973). Seed production is often 100 times greater than observed seedling densities.

The species may have allelopathic properties and thus may interfere with woodland, plantation, or adjacent cropland species. The allelopathic potential of A. petiolata is difficult to confirm. There has been a long interest in the chemical products from this plant (Herissey and Boivin, 1927; Cole, 1975). Several phytotoxic hydrolysis products of glucosinolates have been isolated from A. petiolata (Vaughn and Berhow, 1999), especially root tissues, but their residence time in the environment is likely to be short. There is little population level variability of these products (Cipollini, 2002). McCarthy and Hanson (1998) found that water extracts from stem and root tissue had no direct effect on seed germination or seedling growth of four target assay species, while Stinson et al. (2006) reported that, under laboratory conditions, A. petiolata virtually eliminated mycorrhizal activity, limiting growth and survival of native tree species. Roberts and Anderson (2001) also proposed that allelopathic effects may be indirect and act on mycorrhizal fungi, thereby reducing the competitive ability of neighbouring plants. Alternatively, the production of these various chemicals may be solely related to plant-insect interactions. There appears to be a negative effect of these secondary compounds on egg-laying and development of butterflies in the genus Pieris (Haribal and Renwick, 1998, 2001; Renwick and Lopez, 1999; Haribal et al., 2001; Renwick et al., 2001).

Reproductive Biology

The floral characteristics of A. petiolata are consistent with a generalist pollinator syndrome. Cavers et al. (1979) observed syrphid flies, midges, and bees (Halictidae and Adrenidae) visiting flowers of A. petiolata in Ontario, Canada. Anderson et al. (1996) observed a similar suite of pollinators in Illinois, USA. Cruden et al. (1996) found that the primary pollinators were short-tongued bees and flies, and the nectar contains 51% fructose, 44% glucose, and little sucrose; a composition typical for short-tongue bee pollination. Anderson et al. (1996) report the pollen:ovule ratio to be 1455:1, which tends to be characteristic of species with facultative xenogamous (out-crossing) breeding systems, plants of late successional stage, and habitats where pollination is unpredictable. Indeed, many of these conditions are reflected by A. petiolata. However, Clapham et al. (1962) state that the flowers are visited by various small insects but are automatically self-pollinated. The species appears to be self-fertile (autogamous) and is fully capable of self-pollination and fertilization, often before anthesis. Out-crossing (xenogamy) produces similar levels of fruit and seed relative to those with autogamy (Anderson et al., 1996). The breeding system could best be described as facultative xenogamy. Propagation is entirely by seed and the seed rain can vary from 10,000 to 40,000 seeds per m², and seed are released from slender siliques over a period of time.

Environmental Requirements

The large-scale distribution of A. petiolata in its native and introduced ranges seems to be controlled by climatic requirements. The species requires a moderately high amount of moisture and cold winters of sufficient duration to ensure stratification and breaking of seed dormancy. The plant is characteristic of natural and anthropogenically disturbed areas. On a smaller scale, the relationship between the environment and invasive potential at the stand level has been examined (Meekins and McCarthy, 2001). Seeds sown into lowland habitats yielded many more and larger plants than those sown into upland habitats. Likewise, edge habitats generally produced larger more fecund individuals than forest interior plots which is consistent with allocation data of Byers and Quinn (1998). Thus, increased light and soil moisture are conditions at a local level that promote A. petiolata invasion, survival, and growth. Smith et al. (2003) found that plants growing in upland environments had relatively low reproductive outputs and that density had little influence on reproduction. Thus, invasion can occur in upland habitats, but the rate of spread and species displacement is likely to be slower. Meekins and McCarthy (2002) conducted a detailed demographic study of plants growing in low and high density populations. Survival to flowering was greatest in low density populations, but a relatively high intrinsic rate of reproduction often led to rapid expansion of the population.

There has been considerable study at the individual plant level on how biomass allocation patterns are affected by the environment. Meekins and McCarthy (2000) conducted a factorial experiment in a common garden to examine the relationship between density, nutrients, and water to growth and allocation patterns of A. petiolata, finding that both rosette and mature plant growth were increased at low densities, increased nutrients, and greater irradiance. Irradiance accounted for the greatest proportion of variation explained in the experiment. Likewise, site to site variation has been documented to affect patterns of seed maturation and abortion (Susko and Lovett-Doust, 1998) and population-level patterns of variation in seed mass were great both within and among populations (Susko and Lovett-Doust, 2000a). Susko and Lovett-Doust (1999, 2000b) also document strong patterns of within-plant resource allocation patterns and reproductive potential as it relates to source-sink issues.

In Europe, it is said to grow best on base-rich soils (Clapham et al., 1962). In Ontario it is found on a wide range of soil types from clays to sand and gravelly-loams, often associated with well fertilized sites (Cavers et al., 1979). Very dense populations have been observed in the Midwest USA on calcium-rich soils.

Associations

A. petiolata has no known strong species associations.

A. petiolata is a preferred food plant in Europe for the green-veined white butterfly (Pieris napi L.) (Lees and Archer, 1974). Cavers et al. (1979) observed European cabbage butterflies (P. rapae L.) ovipositing on plants of A. petiolata in Ontario, Canada, but little to no damage was ever observed. Likewise, the native American butterfly (P. napi oleracea) uses A. petiolata as its primary host, but the larvae rarely survive and their resultant effect is negligible (Renwick et al., 2001), thus the plant is protected from this potential herbivore. Leaf hoppers and flea beetles have been observed on A. petiolata, but again, no obvious damage has occurred.

Human food and beverage

• Chowders/soups
• salad
• Spices and culinary herbs
• Vegetable

Medicinal, pharmaceutical

• Traditional/folklore

At maturity, A. petiolata is unlikely to be confused with other species. There are only two species in the genus (Gleason and Cronquist, 1991) and the genus is distinct from most other mustards, especially in its characteristic garlic odour. Immature plants could be visually confused with other rosette-forming species such as violets (Viola spp.), avens (Geum spp.) or Cardamine spp., although again, the diagnostic garlic odour will likely prevent misidentification.

Due to the variable regulations around (de)registration of pesticides, your national list of registered pesticides or relevant authority should be consulted to determine which products are legally allowed for use in your country when considering chemical control. Pesticides should always be used in a lawful manner, consistent with the product's label.

Cultural Control

The goal of any A. petiolata management programme must be to prevent seed production (Nuzzo, 2003). Because of the presence of a seed bank, whatever control method is employed, it must be continued for a period of no less than 3 years in order to eradicate the species locally (Nuzzo, 1991). A. petiolata is not effectively controlled by grazing, and is under little to no vertebrate herbivore pressure in North America. Cavers et al. (1979) noted no grazing from white-tailed deer, and only occasional consumption by cows in Ontario, Canada, resulting in an unpleasant taste to the milk.

Mechanical Control

Hand pulling is an effective strategy for control, particularly with small or newly formed populations. When the plants have bolted in the second growing season, the stem may be easily grasped and pulled; roots will usually come out intact along with the stem. Because the fruit is photosynthetic, if fruit development has started, plants should not be left on site or hung from neighbouring vegetation as the fruits are well known to continue to develop and dehisce even while lying on the ground, thus plants should be bagged and taken off site. Given the presence of a seed bank, repeated visits to a habitat over a number of years will be required to eradicate the plant. Likewise, cutting may provide good control, but cutting at ground level is important as plants that have been mown (or 'string-trimmed') often respond by sending up new flowering shoots from the root crown. Nuzzo (1991) found that plants cut at ground level had 99% mortality and no seed production, whereas plants cut at a height of 10 cm had 71% mortality and seed production reduced by 98%. The use of cutting must be weighed against various factors, for example, certain other species that may be growing in association with A. petiolata such as native Trillium spp. are severely damaged by cutting (Nuzzo, 2003).

Chemical Control

Foliar application of herbicides can be used to control A. petiolata where mechanical methods are impractical due to population size. Glyphosate, triclopyr, and mecoprop have all been used effectively (Nuzzo, 2003) to control A. petiolata, and because these herbicides are not target specific, they should be applied to A. petiolata during the dormant season where the plant is in the rosette stage and native vegetation has not yet emerged. However, biologically active temperatures are also usually required for certain herbicides (e.g. glyphosates) to be effective. Thus, the window of time for application may be narrow. Acifluoren, bentazon and 2,4-D are not recommended for control of A. petiolata (Nuzzo, 2003).

Biological Control

There are currently no known biological control programmes in use to control A. petiolata. However, Blossey et al. (2001) indicate that they are investigating a variety of species for possible use as biological control agents. Although A. petiolata is under virtually no herbivore pressure in North American habitats, over 70 species of insect herbivores and seven fungi are associated with this plant in Europe. Many are not sufficiently host specific to use for control; five monophagous weevils and one oligophagous flea beetle are being further investigated in an effort to develop a concerted suite of attack agents for the seeds, stems, and roots (Blossey et al., 2001).

Integrated Control

Nuzzo (1991) suggests that a suite of cutting, chemical, and fire control methods can be adopted for eradication as long as they are applied sequentially for 3 years or more to exhaust the seed bank. The use of fire as a control method for A. petiolata has been well studied, but the results are somewhat conflicting. Nuzzo (1991) found that fire reduced populations of A. petiolata and that the effect was related to fire intensity; moderate intensity fires were effective whereas low intensity fires had virtually no effect. As many prescribed fires fall into this latter category, the efficacy of fire alone to control A. petiolata is questionable, but it may be used effectively in combination with other methods. Nuzzo et al. (1996) found that A. petiolata was maintained, but in a reduced condition, in forests burned repeatedly for 5 years. However, Luken and Shea (2000) found that moderate intensity dormant season fires did nothing to reduce A. petiolata abundance, and in many plots the species actually increased in abundance relative to control plots. The use of fire as a management tool should be integrated with other management objectives given the manifold effects it has on a habitat.

11/03/19 Updated by: 

Ghislaine Cortat, CABI, Delémont, Switzerland