Lepidium draba
(hoary cress)

Lepidium draba is found in a wide range of habitats ranging from roadsides (highly disturbed sites) to rangelands, meadows and pastures, cultivated fields, home gardens, national parks and wasteland. It exists at high and low latitudes, above and below sea level. It is found under irrigation and in soils of high moisture content and also exists in arid regions. It invades all soil types, although it prefers alkaline conditions and can germinate in saline soils. All of these characteristics and adaptations provide exceptional advantages for this noxious weed and reflect its high potential to invade different agricultural systems. The absence of L. draba plants from plantations or sites where it has not been reported may not be due to the species' ability to invade, survive, grow or develop in these sites but rather to the absence of initial infestation. The weed can invade both agricultural and natural ecosystems (Larson et al., 2000). In summary, L. draba is a serious threat to agriculture and the environment. It is of a high ecological tolerance and is potentially harmful to human and animal health. Its ability to invade and reproduce by different means and to host plant pathogenic agents make this weed invasive with significant problems to man and natural resources.

L. draba has been reported in many Asian and European countries of moderate climate, and has also spread as a noxious weed in relatively cold regions of Canada and USA. In addition, it has been reported in several African countries, such as South Africa.

Although native to Mediterranean and western and central Asia, L. draba has become widely naturalized in many parts of the world, probably as a contaminant of seed supplies. It is one of the worst weeds in all European countries (Clay, 1987) and is also widespread in arid and semi-arid regions of certain parts of the world, such as Africa (Al-Ahmad, 1982; Qasem, 1995). This wide distribution reflects the phenotypic plasticity of the weed, its adaptability and ability to grow and proliferate under different environmental conditions.

L. draba is a ruderal perennial weed (Simova-Tosic et al., 1996), widely reported from coastal, maritime and montane habitats. It is widespread on arable land, waste ground and roadsides (Morris, 1982; Qasem, 1993). Cropping systems such as limited tillage systems may favour its spread (Fernandez-Quintanilla et al., 1984) and it is especially vigorous under irrigation. It is found on a variety of loamy soils, from light to heavy, and prefers neutral to alkaline soils (Olah, 1979). It favours disturbed soils (turned and/or cultivated) with moderate moisture, especially roadsides, ditch banks, sub-irrigated pastures and rangeland (Elpel, 2000), or overgrazed and other areas where native plants face growth problems (http://mtwow.org/whitetop.html). It grows under open conditions, without shade, in different economically important crops and is found along roadsides. It is also reported in 6- to 9.6-year-old communities of saltcedar (Tamarix ramosissima) with 25% coverage (Brothrson et al., 1984) and occurs in Purshia tridentate/Festuca scabrella and Pseudoroegneria spicata habitats (Guenther et al., 1993). In other parts of the USA, it is associated with Artemesia tridentata subsp. tridentata and A. tridentata subsp. wyomingensis communities (McInnis et al., 1993). In Jordan, it is widely spread from the north of the country (1600 m above sea level; annual rainfall of 500-600 ml) to the drier regions of the centre and north-east, which receive less than 150-200 ml annual rainfall. Thick populations of low, vigorous plants, receiving the lowest annual rainfall, have been found close to the desert and sporadic plants or small patches have been recorded in lucerne in the Jordan Valley at 355 m below sea level (Qasem, unpublished data). L. draba has been observed at altitudes as high as 2100 m in the Alps and up to 1600 m in Iran (Holm et al., 1997); at 1700-2400 m in Baha plateau, Saudi Arabia; and at 200 m above sea level in Argentina (Conticello and Gandullo, 1991).

Although L. draba has been reported in more than 38 countries, only a limited number of host crops have been mentioned. As it is a perennial weed, it prefers non- or less-disturbed habitats and thus is commonly found as a weed of fruit trees and rangelands. However, it also has a strong ability to invade field crops and to cause great yield losses through competition or allelopathy, or both.

Genetics

L. draba is self-incompatible and is pollinated by insects (Lacey and Lacey, 1985).

Physiology and Phenology

L. draba produces large amounts of seed, some 2300 seeds per stem or 1200-4800 seeds per plant have been recorded (Elpel, 2000). In temperate zones such as Canada, the flowering period is from May to July, with mature seed produced 1 month later. The minimum temperature for germination reported by Brown and Porter (1942) was 0.5°C, the maximum was 40°C, and the optimum was between 20 and 30°C. Light was not required for germination but enhanced germination rates, whereas oxygen levels below 10% and above 55% were unfavourable to germination. Seeds buried at a depth of 10-15 cm for 3 years rapidly decreased in viability; however, they were reported to retain viability in the soil for about 2 years (Elpel, 2000). Kiemnec and Larson (1991) observed that germination of L. draba decreased with lower osmotic potential. Increases in salinity up to an electrical conductivity (EC) of 12 ds/m had no effect on germination, but root growth was reduced by decreases in osmotic potential.

L. draba perennates by rootstocks that have abundant resources, and can overwinter and produce vigorous new shoots in the spring. Miller et al. (1994) reported similar phenology of the weed in different locations, despite differences in drought occurrence between the study sites. They reported 13C translocation to below-ground tissues occurring within 1 hour of labelling, which reached a peak within 24 hours. However, the greatest enrichment of roots occurred during the flowering stage. The peak in below-ground carbon allocation occurred at a stage when leaf conductance was declining rapidly. The same authors concluded that the short period of maximum carbon allocation below ground, the large proportion of subterranean tissue, and the wide variation of phenology among plants at a given time may account for the difficulty in chemically controlling L. draba.

Neururer (1986) concluded that undiversified rotations, minimal tillage, chemical selection and inappropriate control timing within a rotation are factors that have contributed to an increasing proliferation of perennial weeds in sugarbeet, including L. draba. The weed competes highly for soil moisture in arid regions, and its growth increases with increasing water consumption (Al-Ahmad, 1982).

Reproductive Biology

L. draba reproduces by seeds, root stock, and creeping roots or rhizomes (http:mtwow.org/whitetop.html). It can spread vegetatively at a rate of 2 m/year (Prach, 1988); a single plant can spread over an area of 3.6 m in diameter in 1 year (Miller, 1986) and fragments can give rise to new shoots throughout the season (Miller et al., 1994). The majority (approximately 76%) of the weed biomass is located below the ground and the deeply penetrating, creeping roots make it difficult to eradicate. If the weed is left unchecked, it soon colonizes large areas, choking the other plants present.

Branching patterns generally start with development of the first roots from the radicle in the initial 2-3 weeks of growth; first-order lateral roots grow outwards and then downwards to become vertical roots. Secondary lateral roots and shoot buds usually develop just below where the lateral roots turn downwards. Buds at or below the soil surface may become subterranean roots that are able to produce shoot buds later, whereas buds higher up the plant become rosettes. New crowns form rapidly from adventitious buds on upper roots if the crown is damaged. In this way, dense colonies of the weed can easily exclude other vegetation (Holm et al., 1997).

L. draba is implicated as a host of several  economically important pests and diseases which include, Beet western yellows virus, Tobacco mosaic virus and Pieris brassicae (Mushtaque and Mohyuddin, 1984) to name a few. L. draba, is also attacked by several insect and mite species and Hinz et al., (2003) list 211 different organisms that attack L. draba to varying degrees of specificity. Only three of the herbivores, Aceria draba, Ceutorhynchus cardariae and Ceutorhynchus turbatus, are considered host specific enough to have potential as biological control agents of this noxious weed. Some others such as Pieris brassicae (Mushtaque and Mohyuddin, 1984), can attack many other plant species including some important crops. L. draba can therefore serve as an alternative host for these agricultural pests.

Human food and beverage

• Spices and culinary herbs
• Vegetable

L. draba subsp. draba has been previously confused with the other hoary cresses, L. draba subsp. chalepense  (formerly Cardaria chalepensis and C. draba var. repens) and L. appelianum (formerly Cardaria pubescens) in North America. These are primarily weeds of arable land.

Distinctions between the hoary cresses are based on a few fruit characters: L. draba subsp. draba has a fruit shaped like an inverted heart, with a persistent style, hairless and deflating at maturity, usually containing two seeds; in L. draba subsp. chalepense, the fruit is oval- to lens-shaped, has a slightly longer style than in subsp. draba, and it stays inflated at maturity, usually containing four seeds; the fruit of L. appelianum is globose, hairy and remains inflated when mature, usually containing four seeds.

Lepidium sativum, L. latifolium, L. spinescens, L. spinosum and L. aucheri have a racemose inflorescence, simple or paniculate, and ebracteate. Their petals are equal, longer or sometimes shorter than sepals, or extremely reduced. L. draba and L. appelianum have corymbose, flowering racemes, whereas L. draba subsp. draba has white petals twice as long as the sepals and its leaves are not clasping.

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

Soil cultivation can prevent reproduction of L. draba by creeping roots (Lipa, 1983), whereas direct drilling of cereals encourages growth (Fernandez-Quintanilla et al., 1984). Rouquand and Videla (1996) found that a combination of hoeing and glyphosate or weed removal when above-ground biomass was at a maximum in August and September, followed by glyphosate application in February, was the most effective strategy for control of L. draba. In Canada, L. draba was eliminated in 3 years following intensive tillage (24 operations at 2-week intervals) with disc or sub-surface cultivators (Holm et al., 1997). The seeds were killed after being buried in moist compacted manure for 1 month (http://www.nwcb.wa.gov/weed_info/hoarycress.html). Early spring blowing and planting to a cereal crops can control L. draba (Mulligan and Findlay, 1974) and out-compete the weed through dense stands of perennial grasses and legumes such as lucerne (Elpel, 2000). Grazing by sheep and goats is effective when the weed is the most abundant food source, or at the bud stage. Cattle eat L. draba more in the spring and grazing may be effective, but only at early growth or seed pod stages. Although the weed is nutritious, it is generally unpalatable and not attractive when other sources of food are available. In addition, it can cause a lot of health complications to grazing animals due to its content of glucosinolate glycosides. In general, the best management of L. draba is probably through a combination of competitive crops, tillage and herbicides.

Mechanical Control

Mechanical control seems an ineffective method of control because hand-weeding is impractical for perennial weed control, where weeds regenerate from below-ground vegetative parts. The weed must be tall enough to pull out from the soil, but hidden rhizomes or creeping roots below the soil surface can re-establish the weed. Mowing with a lawn mower or weed whacker can reduce the competitive ability of the weed and its ability to produce seeds if practised on time, but it is confined to small area and is ineffective against buried seeds and creeping roots or rhizomes. Cultivation may be a time- and energy-consuming method of control for perennial weed infestation unless it is repeated many times at the proper time and under the proper conditions, short intervals and during re-budding stage of L. draba, but may also enhance rhizome dissemination and increase infestation in wet soil.

Chemical Control

The herbicide 2,4-D has been widely used for control, but must be applied for 3 years or more at the early bud stage before flowering or in the late autumn rosette stage (Holm et al., 1997). Herbicides that have been found to be useful include chlorsulfuron (Swinnerton et al., 1984; Chirita and Henegar, 1985), 2,4-D + dicamba, bromoxynil + MCPA or 2,4-D at emergence in winter wheat. Effective control of L. draba in oats was obtained with chlorsulfuron, trisulfuron and fluroxypyr (Heap and Mitchell, 1992) and with glyphosate, atrazine + nopon 11E oil or a mixture of atrazine + sun 11E oil at the two-leaf stage in maize (Dobrovodski, 1975). In maize, a combination of eradicance, atrazine and 2,4-D was most effective and gave good control (Naederi and Khajehpour, 1997). In lucerne, Hernado et al. (1987) and Kontsiotou (1982) found that secbumeton gave good results, whereas in lentil, prometryn provided the best control of different broadleaved weeds including L. draba, and gave the highest crop yields.

In fruit tree orchards, norflurazon (Monserrat-Delgado, 1995), simazine at pre-emergence followed by glyphosate (Sarpe et al., 1989), simazine + paraquat, aminotriazole + diuron + bromacil + glyphosate, or terbumeton have all been used (Budoi et al., 1981). Glyphosate (Kafadaroff, 1977; Brattain and Fay, 1980) has been used successfully in vineyards and in oak plantations (Wallis, 1978; Waterhouse and Mahoney, 1983). Combination treatment of 24% urea and 6% yolk gave excellent control in pistachio fields (Davarynejad and Ak, 2001). Other effective herbicides reported were chlorsulfuron or metsulfuron applied during the budding or early bloom stages; picloram, but with little effect (Elpel, 2000); and imazethapyr, which provided >90% control in lucerne/Dactylis glomerata pastures only during the year of application (Stougaard et al., 1999).

Problems in the chemical control of L. draba have been reported by Julliard and Ancel (1973) in France, by MAFF (1977) in the UK and by the Tasmanian Department of Agriculture (1978) regarding aminotriazole, glyphosate and metribuzin. However, the short period of maximum C allocation to the below-ground tissues, the large proportion of below-ground tissues, and the wide variation of phenology among plants at a given time may account for the difficulty in chemical control of L. draba (Miller et al., 1994).

Biological Control

Lipa (1981) studied the monophagous silver mite, Aceria drabae, as a suitable species for biological control of L. draba. It did not attack any other crucifer growing in the vicinity of the weed, and was able to colonize it successfully. The mite first attacks the flowers, penetrating the seeds; flower heads become completely damaged and infested plants do not set seed (Lipa, 1976; Lipa et al., 1977). The seed feeding  beetle, Ceuthorhynchus turbatus, has also been found to attack L. draba (Lipa, 1974). Sobhian (1976) observed the chrysomelid Colaphellus hoeftii feeding on L. draba in Iran. The adults and larvae were capable of denuding young plants of vegetation and preventing them from flowering. Larvae fed on groups of immature plants. At the initiation of the biological control programme by CABI in 2001, literature and fieldwork in Europe identified at least 211 species associated with L. draba, of which five weevils, one flea beetle, two gall midges and one gall mite had potential as biological control agents. Six of these species have been under investigation: Ceutorhynchus turbatus (Coleoptera: Curculionidae) attacks the seeds; Psylliodes wrasei (Coleoptera: Chrysomelidae) feeds on developing shoots and vegetative points, and C. merkli (Coleoptera: Curculionidae) mines the stems of hoary cress. The remaining species form galls in the inflorescences (Aceria draba (Acari: Eriophyidae), stems/petioles (Ceutorhynchus cardariae) (Coleoptera: Curculionidae); and roots (Ceutorhynchus assimilis (Coleoptera: Curculionidae). By 2018, only three species, A. draba, C. cardariae and C. turbatus, have been have been prioritized as promising biological control agents.

Volatile materials from leaves of Artemisia absinthium have been reported to inhibit growth of L. draba seedlings and aqueous extracts of its leaves can prevent germination (Chirca and Fabian, 1973).

Integrated Control

Integrated control of L. draba includes prevention, education, early detection and eradication. Prevention is integrated with cultural, physical, mechanical, biological, chemical and monitoring methods of control and includes banning hay, grinding seeds with lucerne hay or barley grain, thus reducing weed emergence by 98-100%. Rigorous processing during manufacturing of hay/grain pellets can reduce the risk of disseminating weed seeds from pelleted feed (Cash et al., 1998). Only certified weed-seed-free hay should be used, equipment should be pressure-cleaned before entering national parks and common gardens (Olliff et al., 2001), new infestations should be surveyed, and vehicles (especially the undercarriage) should be careful cleaned by washing after driving out of an infested area (http://mtwow.org/whitetop.html).

22/01/18 Updated by:

Philip Weyl, CABI, Delémont, Switzerland