Frangula alnus (alder buckthorn)

F. alnus was introduced to North America from Europe more than 100 years ago. Once established it maintains itself due to prolific seed production, vigorous growth over an extended growing season and its ability to regenerate following burning and cutting. These characteristics make it difficult to eradicate. Repeated cutting and application of herbicides required to eliminate F. alnus is laborious and expensive. Consequently, most restoration work has been conducted in natural ecosystems of special interest. Its adverse effect on native species arises because F. alnus shades out understorey plants. Its aggressive character, especially in wetlands, is widely noted (Catling and Porebski, 1994).

In the past, Frangula was treated as a subgenus of Rhamnus, distinguished from Rhamnus sensu stricto by the presence of bisexual instead of unisexual flowers and 5-merous instead of 4-merous flower parts, and on the basis of seed characteristics. Most authorities now recognize this section as a separate genus, using Frangula alnus Mill.

The genus name is possibly derived from the Latin 'frangere' meaning to break, perhaps a reference to the brittle nature of its branches. The specific name 'alnus' and the common name 'alder buckthorn' may be related to the frequent occurrence of F. alnus in wet areas, the preferred habitat for alders (Alnus spp.; Haber, 1997). The wood has been used to make butchers' skewers which are known in some areas as 'dogs', hence the alternative common name of black dogwood (Fraser, 2003).

F. alnus is a deciduous shrub or small tree usually 4-5 m in height (Tutin et al., 1968), but may grow to 7 m (Gleason, 1963). It develops an erect, slender habit with branches somewhat irregular in alternate pairs, ascending at an acute angle to the main stem (Godwin, 1943). Young twigs are green but turn grey-brown with age and develop red-brown to dark violet tips. Lenticels may be evident as white dots and stripes. Lemon-yellow inner-bark tissues are exposed when the outer-bark is damaged and the young wood is dark brown. Old bark is smooth, except in very old specimens, and readily peels off dead wood. Spines are absent from F. alnus. Leaves are petiolate, obovate in shape, 2-7 cm in length and usually little more than half as wide. They are cuspidate to acuminate in shape, typically ending with a short pointed tip. Leaf margins are entire but wavy, although in seedlings leaves may be serrated. The lower surface of young leaves is pubescent, being covered with dense brownish hairs which are later shed so that older leaves are glabrous and shiny green in colour. Sun leaves are relatively broader and more shiny than shade leaves. The leaves turn yellow, then red in the autumn. Lateral veins are conspicuous on the upper surface of the leaves with 6-12 (commonly 7) pairs running more or less parallel to each other.

F. alnus develops sessile umbels in the leaf axils on young wood with 2-8 flowers borne on stout, unequal, glabrous pedicels 3-10 mm long; occasionally single flowers develop. Individual flowers are greenish-white, about 3 mm in diameter and bisexual. The flowers are 5-merous with broadly obovate petals 1-1.4 mm long and cleft at the tip (Gleason, 1963). Fruits are 6-10 mm in diameter, and change from green to red, then to violet-black on ripening; flowering and fruit development are rather asynchronous, hence all stages of ripening may be present. Each drupe usually contains 2, but occasionally 3, pyrenes or stones which are broadly obovoid in shape, about 5 mm long and 2 mm thick; they have a faint ridge running down the inner face and a deep furrow at the base. Young and Young (1992) report 52 seeds/g for F. alnus. Germination is hypogeal (Godwin, 1943).

F. alnus is found throughout Europe, except Iceland and the extreme north, with a latitudinal limit for the species ranging from 64.5°N in Norway to about 67°N in Russia. Although it is found throughout Portugal and Spain it is rare in the Mediterranean region and absent from Sicily, Corsica and the Balearics; it is present in Algeria and Morocco (Godwin, 1943). It is one of Northern Ireland's rarest trees, and remains mostly in the region of Lough Neagh where it is protected under the 1985 Wildlife Order. This Wildlife Order prohibits intentional picking, uprooting or destruction of F. alnus, and even collection of flowers and seeds except with permission. It is apparently declining in Northern Ireland due to bog exploitation (Ulster Museum, 1999).

In North America, F. alnus is found most extensively in the northeastern part of the continent. It is found throughout the New England states from Maine to New Jersey, as far south as Tennessee and west to Wyoming and Colorado in the USA. In Canada it is reported mainly from southern Ontario where it occurs in the southeast and southwest of the province; it is absent from most of the Canadian Shield region (Catling and Porebski, 1994). F. alnus is also common in Quebec and occurs eastwards to Nova Scotia and Prince Edward Island; some individuals are recorded from Manitoba.

F. alnus was probably introduced to North America before 1800, but did not become widespread and naturalized until the early 1900s. In Canada it is thought to have spread from the three urban centres where it was first introduced: London (1898), Ottawa (1899) and Guelph (1906). F. alnus was subsequently recorded in Toronto by 1953, and in the Georgian Bay area of Ontario by 1968. In southeast Ontario, it was recorded in Kingston in 1953, and east of Ottawa in 1975 (White et al., 1993; Catling and Porebski, 1994; Haber, 1997). The presence of F. alnus in Prince Edward Island is attributed to its introduction as a hedge plant at the Experimental Farm in Charlottetown; it has now become naturalized in the province along hedgerows and wood edges (Catling et al., 1985). Haber (1997) reported that the earliest and westernmost occurrence of F. alnus in southern Ontario was documented by a collection on Walpole Island in Lake St Clair in 1985, but noted that it probably arrived in that region much earlier, and perhaps had spread from Michigan where it was recorded in 1949 in Oakland County on the west side of Lake St Clair.

Cultivated stocks of F. alnus are still available from nurseries, even though it is recognized as an invasive exotic in North America. In several US states and Canadian provinces, F. alnus is listed as a restricted noxious weed and it is prohibited to import, sell and transport it or its propagating parts. Since January 2002, the USA requires seeds of all species to have phytosanitary certificates, so this has limited importation from overseas suppliers.

Natural Dispersal (Non-Biotic)

Natural dispersal is minimal. Many fruits fall directly beneath the parent shrubs resulting in high seed and seedling densities (in the order of 100 seedlings/m²) under mature plants (Godwin, 1943). The fresh fruits of F. alnus are reported to float for about 3 weeks and dry seeds for about 1 week (Ridley, 1930). However, Catling and Porebski (1994) noted that ripe fruits sank immediately. Unlike Rhamnus cathartica, the fruits do not dry out on the shrub over winter, so it is unlikely that dispersal occurs by water.

Vector Transmission (Biotic)

Dispersal is principally by birds. In Europe, field fares (Turdus pilaris), missel thrushes (T. viscivorus) and pheasants (Phasianus colchicus) have been observed feeding on the fruit. Field mice (Apodemus sylvaticus) also collect and eat the stones (Godwin, 1943). Kollmann and Pirl (1995) concluded that F. alnus was dispersed primarily by birds in the Lake Constance area of Germany. Although fruit availability was highest at the time of autumn migration, fewer fruits were consumed at this time than at other times of the year. They also noted that seed predation by rodents was highest under mature shrubs. In North America, waxwings (Bombycilla sp.), starlings (Sturnus vulgaris), American robins (Turdus migratorius) and rose-breasted grosbeaks (Pheucticus ludovicianus) are important (Catling and Porebski, 1994). The spread of F. alnus in Ohio is attributed mainly to starlings (Sturnus vulgaris; Howell and Blackwell, 1997).

F. alnus is also used by moose (Alces alces) for browse (Borkowska and Konopko, 1994) and this accounts for about 8% of their early winter food intake (Morow, 1976). The winter diet of male moose in Belarus is mainly conifers and F. alnus, whereas females mostly feed on willows; calves are less selective. Daily food intake was 12.5 kg for males, 7.8 kg for females and 5.7 kg for calves. For the period January-March it was estimated that moose consumed 2,670 kg of F. alnus, equivalent to about 1% of the total food eaten (Dunin, 1989).

Agricultural Practices

The use of F. alnus as a landscape plant would have assisted dispersion in the past, but further movement cannot be attributed to current agricultural practices. However, the role of F. alnus as a potential windbreak for centre-point irrigation has been investigated (Scholten et al., 1993).

Accidental Introduction

No documented instances.

Intentional Introduction

F. alnus has been used extensively as an ornamental shrub. Some demand for F. alnus may also be expected because of its use as a herbal remedy. Some countries, including the Netherlands, former Czechoslovakia, Poland and Russia have commercial plantations as a source for high-grade charcoal.

F. alnus is a problem species in native communities because it establishes in dense stands which shade out other understorey species. Possessky et al. (2000) reported a reduction in composition and abundance of the herbaceous cover in riparian habitats in the northern Allegheny Plateau (of Pennsylvania, New York and Ohio, USA) following invasion by F. alnus . Similarly, Reinartz (1997) described how an undisturbed bog community in Wisconsin was invaded by F. alnus in 1955 with a dense tall shrub canopy dominating the site within 12 years. The species is listed as an invasive weed in Tennessee and Wisconsin, USA (Southeast Exotic Pest Plant Council, 1996; Hoffman and Kearns, 1997). F. alnus was recently rated as one of the six principal invasive aliens of wetlands in Canada, and one of four principal invasive aliens in Canadian uplands. In a national survey it was rated second to purple loosestrife (Lythrum salicaria) with respect to the extent to which it is spreading in natural habitats and its severity of impact in Canada (White et al., 1993).

F. alnus is associated with crown rust (Puccinia coronata) which infects several cool season turfgrasses, native grasses and cereals. The uredia, telia and basidiospores are produced on the graminoid hosts, the aecia and pycnia are produced on F. alnus (and Rhamnus cathartica; Partridge, 1998). Alfalfa mosaic virus, which infects a wide variety of plants, including crops, and is vectored by aphids, has also been isolated from young leaves and root cuttings of F. alnus in Italy (Marani and Giunchedi, 2002).

F. alnus has a superficial resemblance to F. alnus, although the former is readily distinguished by its smaller leaves and absence of spines. The alderleaved buckthorn (Rhamnus alnifolia) which is native to North America can be distinguished from F. alnus by the presence of small teeth on its leaves.

F. alnus is native to Eurasia and North Africa where it is generally associated with moist scrubland and open stands of deciduous trees and conifers. Although it shows a preference for seepage areas and gley soils, it will also survive in moderately dry sites. However, F. alnus cannot tolerate dry soils, such as those associated with chalk and limestone, nor will it survive permanent waterlogging. It increased in abundance in raised bogs in the Great Hungarian Plain as a result of human induced drainage (Simon, 1992). F. alnus is generally intolerant of shade and will survive only in the understorey of open woodland (Godwin, 1943). In the UK it is normally considered a lowland species being found generally below 305 m elevation, but commonly may grow up to 700 m on a variety of soils in Europe with some specimens at 1500 m (Godwin, 1943).

In North America, F. alnus has become established in bogs, marshes and fens, along riverbanks, in mesic upland sites, as well as drier prairies, abandoned farmland and roadsides. It is especially aggressive in alkaline bogs and swamps. It grows in a wide range of soil textures, and is prevalent in alder thickets and in wetlands associated with calcareous substrates and limestones. F. alnus has become widespread in North America due to various disturbances, such as drainage, lack of fire, woodland grazing and cutting, which created ideal habitats for seedling establishment and maintenance of mature individuals.

Genetics

The chromosome number is 2n = 20 (Tutin et al., 1968).

Physiology and Phenology

In Europe, F. alnus grows actively in spring and summer, with shoot growth typically greatest in the early part of the season (Raulo and Leikola, 1975). Leaves appear in early April to late May, turn yellow in late September and October, and are shed in October (Godwin, 1943). In North America, Converse (1984) noted that F. alnus leafs out in mid- to late May prior to most woody deciduous plants, and sometimes can retain leaves into November. Flowering begins in late May and June. Flowering will normally continue until August, with some blossoms developing as late as September, but ends as the green fruits begin to turn black. Flowering may be terminated by low temperatures in the autumn. In England, fruits begin to redden in July, some may turn black in August, with the majority reaching maturity in September (Godwin, 1943). In North America, fruit ripening occurs from July to October (Young and Young, 1992). Fruits are dispersed from September through November, although some persist until December. Germination mainly occurs in the spring (Godwin, 1943).

Reproductive Biology

Natural reproduction in F. alnus is through seed, with good seed crops produced each year. Vegetative reproduction has not been observed. The majority of flowers are insect pollinated, although self pollination may occasionally occur (Godwin, 1943). Outcrossing is promoted because the anthers normally mature before the pistil is receptive. Numerous insects are attracted to the flowers through odour, pollen and nectar secretions. In Europe the principal pollinators include bees (Apis mellifica, Bombus agrorum, B. jonellus and B. proteus), wasps (Vespa sylvestris [Dolichovespula sylvestris], Eumenes pomiformis), flies (Lophosia fasciata) and beetles (Corymbites sjaelandicus). Medan (1994) considered flies to be the most important pollinators of F. alnus in Spain, although 21 species of insects visited the flowers. However, reproductive success was limited with only 2.8% of open-pollinated flowers producing fruits, and in these only about 50% of the ovules developed into seeds. Although overall reproductive success in terms of ripe ovules averaged only 1.4%, this was equivalent to 430-1560 potential offspring per plant (Medan, 1994). This compares to 1804 fruits produced by the end of September by a single plant in England; of these 1268 had fallen to the ground by the end of December with the remainder presumably dispersed by birds. In drier sites the flowers drop before maturity and seed crops are small. Seed viability of at least 3.5 years is reported by Godwin (1943). Seedling emergence in F. alnus is usually high near the seed source, and a density of 540/m² has been reported (Converse, 1984).

Environmental Requirements

Seed germination typically occurs in the spring following dispersal, although germination occurs within a few days in pyrenes extracted from ripe fruits (Godwin, 1943). After drying, the pyrenes exhibit dormancy. Young and Young (1992) noted that seeds of F. alnus require 20 minutes acid scarification prior to prechilling at 1-5°C for 8 weeks; germination occurs at temperatures of 20-30°C under an 8-hour photoperiod.

Seedling success is greatest in areas where the previous vegetation has been removed and the soil cultivated compared with areas that have been burned, lightly raked or left untreated (Converse, 1984). Experiments conducted in coniferous woodland in Sweden showed that recruitment was limited by the absence of safe sites on the forest floor, rather than a shortage of seed (Eriksson and Ehrlen, 1992). Field-mice (Apodemus sylvaticus) collect the pyrenes and store them in caches of 30-100, and some may be eaten (Godwin, 1943).

Seedlings on peat soil can suffer from frost heaving. The seedlings are tolerant of long periods of submersion when they are not actively growing, but are killed if submerged during periods of active leaf expansion in the spring. In wet sites, the entire root system is above the water table. Raising water levels has been suggested as a means of eliminating shrubs and seedlings (Haber, 1997).

F. alnus is a typical pioneer species in young carr communities in eastern England (Friday, 1997). Godwin (1943) describes the species as typically marginal and seral in woodland because it is somewhat shade intolerant. In dense stands, the lower branches lose their leaves and the shrub develops a more columnar growth habit (Converse, 1984). The status of F. alnus as a seral species is supported by observations that it appears after forest clearance or similar types of land use change.

Associations

The entire root system is mycorrhizal and is associated with vesicular-arbuscular fungi. F. alnus serves as a host for crown rust (Puccinia coronata) which infects grasses and cereals. Alfalfa mosaic virus has also been isolated from young leaves and root cuttings. The main concern of F. alnus is its ability to invade native habitats, especially moist mesic soils where it can become the dominant species.

Godwin (1943) noted that F. alnus bushes of all ages suffer from terminal die-back which in some years may be so widespread that the dominance of F. alnus may be challenged by Rhamnus cathartica. The pathogen (Nectria cinnabarina) enters the plant through the base of dead lateral branches. Several species of arthropods and fungi are associated with the genera Rhamnus and Frangula, and about a dozen of these are currently being reviewed as potential biological control agents. At one site in the Czech Republic, 13 species of wood-boring and subcortical beetles and 3 predaceous beetles were found on F. alnus (Simandl, 1993). The most abundant species exhibited preferences for wood of different diameters; differences were also noted in terms of the condition of the wood. The earliest successional colonizer (Tetrops praeusta) was found on partly fresh branches that were just beginning to wither, although most species were associated with dead, dry wood. Unlike leaf-eating insects, the xylophagous species do not attack live plants. Potential biocontrol agents are discussed in more detail in the section on Biological Control.

F. alnus is also associated with crown rust (Puccinia coronata) which infects several cool season turfgrasses, native grasses and cereals. The uredia, telia and basidiospores are produced on the graminoid hosts, the aecia and pycnia are produced on F. alnus (and Rhamnus cathartica; Partridge, 1998). Alfalfa mosaic virus, which infects a wide variety of plants, including crops, and is vectored by aphids, has also been isolated from young leaves and root cuttings of F. alnus in Italy (Marani and Giunchedi, 2002).

Godwin (1943) noted that bushes of all ages suffer from terminal die-back which in some years may be so widespread that the dominance of F. alnus may be challenged by Rhamnus cathartica. The pathogen (Nectria cinnabarina) enters the plant through the base of dead lateral branches. Several species of arthropods and fungi are associated with the genera Rhamnus and Frangula, and about a dozen of these are currently being reviewed as potential biological control agents. At one site in the Czech Republic, 13 species of wood-boring and subcortical beetles and 3 predaceous beetles were found on F. alnus (Simandl, 1993). The most abundant species exhibited preferences for wood of different diameters; differences were also noted in terms of the condition of the wood. The earliest successional colonizer (Tetrops praeusta) was found on partly fresh branches that were just beginning to wither, although most species were associated with dead, dry wood. Unlike leaf-eating insects, the xylophagous species do not attack live plants. Potential biocontrol agents are discussed in more detail in the section on Biological Control.

At present the economic impact of F. alnus is comparatively small, despite it being an alternate host for crown rust (Puccinia coronata) and Alfalfa mosaic virus. However, considerable time and expense have been invested to remove the shrub from native habitats.

F. alnus forms dense, even-aged stands and causes habitat degradation by shading out rare species and causing a decline in native species diversity (Taft and Solecki, 1990). Lovely (1981) observed a decrease in the small white lady's-slipper orchid (Cyripedium candidum) under F. alnus in Wisconsin, USA. In North America it is especially prevalent in wetland environments where it is characterized as an aggressive alien.

No direct social impacts have been reported, except that the presence of F. alnus can visually detract from local vistas. Local access to sites may be impeded where dense stands have developed.

F. alnus has been used for a variety of purposes. The wood was traditionally used in the form of young twigs and branches, and the supply of small wood was ensured by coppicing. The wood sharpens well and has been used to make arrows, nails and skewers; it has also been used for walking sticks and, when split, for cane chair seats and basket work. Because it is hard and durable, larger pieces have been used to make shoes (Anon., 2003). However, the principal use since Mediaeval times has been as a source of charcoal for gunpowder, and it was in demand for fuses in explosives up to the end of World War II. The hard straight twigs also make excellent charcoal for artists. Some countries, including the Netherlands, the former Czechoslovakia, Poland and Russia have commercial plantations (Fraser, 2003).

The leaves and bark produce a yellow dye, and unripe berries provide a green dye that was used in calico printing and the woollen industry. The ripe berries give various shades of blue and grey (Fraser, 2003).

F. alnus also has many medicinal uses. Traditional herbal remedies are derived from the bark of 1- to 2-year old branches which are harvested in the early spring. The bark contains the glucoside frangulin, which has value as a purgative (Lewis and Elvin-Lewis, 1977). Because fresh bark is poisonous, it is allowed to cure for at least a year before use. It has long been prescribed in various concoctions to cure ailments, such as constipation, gout, jaundice, gum disease, sore throats, lice infestations, dry skin and to heal wounds. In charcoal preparations, F. alnus has been used to treat flatulence and poisoning, and also as a deodorizing substance (Fraser, 2003). As an antifungal agent it has shown activity against Aspergillus, Fusarium and Trichophytum (Anon., 2003).

There has been some interest in F. alnus as a honey plant (Risnes, 1980).

Introduction

F. alnus is difficult to eradicate because of its heavy fruit production, high germination rate and long-term seed viability. Mature plants that are cut near the base sprout vigorously and a combination of different control methods is normally needed, together with follow-up treatments over a period of several years.

Cultural Control

Cultural methods include cutting, mowing girdling and burning (Converse, 1984). Cutting in early June and late August for two or three consecutive years reduced the number and size of the stems that developed and increased the vigour of the herbaceous groundcover. Cutting is very labour intensive and is normally suitable only for small or highly significant sites. Mowing at a height of 2-13 cm from the ground once or twice in the summer can arrest development but may stimulate vigorous resprouting; it does have the beneficial effect of limiting flowering, and prevents seedling establishment. Cutting alone without subsequent herbicide application is usually not effective (Reinartz, 1997). Sinclair and Catling (1999) noted an increase in the cover and number of native species on plots where F. alnus had been cut and, although it grew back quickly, they suggested that even a single cutting may give native species an opportunity to replenish the seed bank.

Girdling plants at the base with a 2- to 3-cm-wide saw cut into the phloem limits resprouting. Phloem should be removed without damaging the xylem, and the girdles should be checked after a few weeks to ensure that new bark is not developing (Heidorn, 1991). Likewise, heating the stem with a flame torch will kill the cambium; this is recommended for smaller stems up to 4.5 cm in diameter. Follow-up herbicide treatment is recommended following girdling (Reinartz, 1997)

F. alnus can potentially be topkilled by fire in late spring just after it has leafed out, because food reserves are low at this time and resprouting vigour is reduced (Converse, 1984). However, burning in autumn has little effect, and studies in Michigan, USA, indicate that stem density was twice as great the following summer as before the burn; the resprouts were one-third the height of the pre-burn stems. Cutting F. alnus in the spring at the leaf expansion stage, and again in the fall, followed by spring burning in the next two years may also be effective in fenland. Successive burns are generally required because F. alnus typically grows in moist areas; this limits fire severity and roots and seeds are often unaffected. Burning kills most seedlings, but new plants arise from buried seed reserves and quickly recolonize the area (Post and Klick, 1988; Post et al., 1989; Taft and Solecki, 1990). A combination of burning followed by planting with groundcover species reportedly limits the re-establishment of F. alnus (and Rhamnus cathartica). Only six Rhamnus/Frangula seedlings (species not differentiated) were recorded in a 144-m² test plot treated in this way compared with 99 seedlings in a burned plot, and 191 seedlings in the control plot (Scriver and Leach, 1998).

The raising of water levels in wetlands where the water table has been lowered artificially (Simon 1992), is also a means of controlling F. alnus (Heidorn, 1991).

Mechanical Control

Seedlings or small plants can be removed by weeding. Larger plants may be pulled out with a "weed wrench" or heavy equipment. However, excavation often disturbs roots of adjacent plants, or creates open areas that are readily colonized by new Rhamnus/Frangula seedlings. This technique may be most useful to control invasion at low densities, or along trails, roads and woodland edges (Anon., 2002).

Chemical Control

Chemical control is the most common method of managing F. alnus . Because F. alnus grows in wetland environments, it is preferable to carry out treatment in the winter months when the soil is frozen and there is less risk of damage to native species from trampling and herbicide overspray (Reinartz, 1997). The preferred method is to cut and apply herbicide to stumps greater than 10 cm in diameter. Basal bark treatments can be applied on smaller shrubs or in areas where openings in the understorey are not necessary for restoration. Repeated treatment over several seasons is usually required to eliminate recruitment from the seed bank. Converse (1984) lists the following herbicides and methods that have achieved good control against F. alnus in the USA: wick-applied glyphosate in May-June; misting of cut stumps <5 cm in diameter) with glyphosate in August; misting with fomasine (ammonium salt) in September. Foliar application of 2,4-D in March-August was not effective because new leaves developed later in the year. Similarly, resprouting occurred following misting of larger stumps (>12 cm in diameter) with glyphosate in August.

Biological Control

Godwin (1943) lists several saprophytes associated with F. alnus, and noted that the fungus Nectria cinnabarina was identified as causing die-back in the 1930s. Experiments with this fungus have been conducted in North America in an attempt to eradicate F. alnus from Wisconsin, USA (Reinartz and Parker, 1989). Current research has focused on the role of insects as potential biocontrol agents (Gassmann et al., 2002); the principal species under evaluation are the following:

Zygana suavis (Cicadellidae) appears to have a strong preference for F. alnus and is common in northern Germany. Its potential impact is to impair photosynthesis and to act as a vector for pathogens. It is one of the very few insects strongly associated with the species.

Sorhagenia janiszewskae (Cosmopterigidae) attacks both F. alnus and Rhamnus cathartica. It is relatively common and easy to collect, but it is difficult to overwinter the adults. S. janiszewskae bores into the above-ground parts of the target shrubs. Its potential impact is to impair growth of the current year's shoots.

Ancylis apicella (Tortricidae) appears to have a preference for F. alnus . It is a bivoltine species with a prolonged impact on the target host.

Oberea pedemontana (Cerambycidae) bores into the branches and trunks of F. alnus and Rhamnus cathartica and causes structural damage and occasionally the death of the host tree. Work with this species is handicapped by a high rate of parasitism and by the time needed to complete its life cycle of 3 years.

Bucculatrix frangulella (Bucculatrigidae) is the only leaf miner recorded on F. alnus, but because it is active late in the season it is likely to have little impact on photosynthesis and water loss.

The univoltine defoliating butterfly Gonepteryx rhamni (Pieridae) shows a clear preference for F. alnus in Europe.

Synanthedon stomoxiformis (Sesiidae) is the sole root-boring species on F. alnus and Rhamnus cathartica. However, there is evidence that it may also attack Sorbus aria (Rosaceae) and more rarely Corylus avelana (Betulaceae) as alternate hosts.

The mites Aceria rhamni and Tetra rhamni have some potential as biological control agents, and may facilitate disease transmission.

Preliminary observations suggest that more information is needed on the habitat of F. alnus in North America and in Europe to determine which biological control agent is most suitable for a particular type of habitat and also which native buckthorn species might be at risk if they occur within the physiological range of a selected biological control agent. Sorhagenia janiszewski, Ancylis apicella, A. derasana, Oberea pedemontana and Zygana suavis merit further study as potential biocontrol agents for North America according to Gassmann et al. (2002).

Integrated Control

Cultural, mechanical and chemical control methods have been used in varying combinations, but no "classic" integrated control methods have been devised for F. alnus.

In Europe, F. alnus grows actively in spring and summer, with shoot growth typically greatest in the early part of the season (Raulo and Leikola, 1975). Leaves appear in early April to late May, turn yellow in late September and October, and are shed in October (Godwin, 1943). In North America, Converse (1984) noted that F. alnus leafs out in mid- to late May prior to most woody deciduous plants, and sometimes can retain leaves into November. Flowering begins in late May and June. Flowering will normally continue until August, with some blossoms developing as late as September, but ends as the green fruits begin to turn black. Flowering may be terminated by low temperatures in the autumn. In England, fruits begin to redden in July, some may turn black in August, with the majority reaching maturity in September (Godwin, 1943). In North America, fruit ripening occurs from July to October (Young and Young, 1992). Fruits are dispersed from September through November, although some persist until December. Germination mainly occurs in the spring (Godwin, 1943).

Natural reproduction in F. alnus is through seed, with good seed crops produced each year. Vegetative reproduction has not been observed. The majority of flowers are insect pollinated, although self pollination may occasionally occur (Godwin, 1943). Outcrossing is promoted because the anthers normally mature before the pistil is receptive. Numerous insects are attracted to the flowers through odour, pollen and nectar secretions. In Europe the principal pollinators include bees (Apis mellifica, Bombus agrorum, B. jonellus, and B. proteus), wasps (Vespa sylvestris [Dolichovespula sylvestris], Eumenes pomiformis), flies (Lophosia fasciata) and beetles (Corymbites sjaelandicus). Medan (1994) considered flies to be the most important pollinators of F. alnus in Spain, although 21 species of insects visited the flowers. However, reproductive success was limited with only 2.8% of open-pollinated flowers producing fruits, and in these only about 50% of the ovules developed into seeds. Although overall reproductive success in terms of ripe ovules averaged only 1.4%, this was equivalent to 430-1560 potential offspring per plant (Medan, 1994). This compares to 1804 fruits produced by the end of September by a single plant in England; of these 1268 had fallen to the ground by the end of December with the remainder presumably dispersed by birds. In drier sites the flowers drop before maturity and seed crops are small. Seed viability of at least 3.5 years is reported by Godwin (1943). Seedling emergence in F. alnus is usually high near the seed source, and a density of 540/m² has been reported (Converse, 1984).

The entire root system is mycorrhizal and is associated with vesicular-arbuscular fungi. F. alnus serves as a host for crown rust (Puccinia coronata) which infects grasses and cereals. Alfalfa mosaic virus has also been isolated from young leaves and root cuttings. The main concern of F. alnus is its ability to invade native habitats, especially moist mesic soils where it can become the dominant species.

Natural dispersal is minimal. Many fruits fall directly beneath the parent shrubs resulting in high seed and seedling densities (in the order of 100 seedlings/m²) under mature plants (Godwin, 1943). The fresh fruits of F. alnus are reported to float for about 3 weeks and dry seeds for about 1 week (Ridley, 1930). However, Catling and Porebski (1994) noted that ripe fruits sank immediately. Unlike Rhamnus cathartica, the fruits do not dry out on the shrub over winter, so it is unlikely that dispersal occurs by water.

Dispersal is principally by birds. In Europe, field fares (Turdus pilaris), missel thrushes (T. viscivorus) and pheasants (Phasianus colchicus) have been observed feeding on the fruit. Field mice (Apodemus sylvaticus) also collect and eat the stones (Godwin, 1943). Kollmann and Pirl (1995) concluded that F. alnus was dispersed primarily by birds in the Lake Constance area of Germany. Although fruit availability was highest at the time of autumn migration, fewer fruits were consumed at this time than at other times of the year. They also noted that seed predation by rodents was highest under mature shrubs. In North America, waxwings (Bombycilla sp.), starlings (Sturnus vulgaris), American robins (Turdus migratorius) and rose-breasted grosbeaks (Pheucticus ludovicianus) are important (Catling and Porebski, 1994). The spread of F. alnus in Ohio is attributed mainly to starlings (Sturnus vulgaris; Howell and Blackwell, 1997).

F. alnus is also used by moose (Alces alces) for browse (Borkowska and Konopko, 1994) and this accounts for about 8% of their early winter food intake (Morow, 1976). The winter diet of male moose in Belarus is mainly conifers and F. alnus , whereas females mostly feed on willows; calves are less selective. Daily food intake was 12.5 kg for males, 7.8 kg for females and 5.7 kg for calves. For the period January-March it was estimated that moose consumed 2,670 kg of F. alnus, equivalent to about 1% of the total food eaten (Dunin, 1989).

Seed germination typically occurs in the spring following dispersal, although germination occurs within a few days in pyrenes extracted from ripe fruits (Godwin, 1943). After drying, the pyrenes exhibit dormancy. Young and Young (1992) noted that seeds of F. alnus require 20 minutes acid scarification prior to prechilling at 1-5°C for 8 weeks; germination occurs at temperatures of 20-30°C under an 8-hour photoperiod.

Seedling success is greatest in areas where the previous vegetation has been removed and the soil cultivated compared with areas that have been burned, lightly raked or left untreated (Converse, 1984). Experiments conducted in coniferous woodland in Sweden showed that recruitment was limited by the absence of safe sites on the forest floor, rather than a shortage of seed (Eriksson and Ehrlen, 1992). Field-mice (Apodemus sylvaticus) collect the pyrenes and store them in caches of 30-100, and some may be eaten (Godwin, 1943).

Seedlings on peat soil can suffer from frost heaving. The seedlings are tolerant of long periods of submersion when they are not actively growing, but are killed if submerged during periods of active leaf expansion in the spring. In wet sites, the entire root system is above the water table. Raising water levels has been suggested as a means of eliminating shrubs and seedlings (Haber, 1997).

F. alnus is a typical pioneer species in young carr communities in eastern England (Friday, 1997). Godwin (1943) describes the species as typically marginal and seral in woodland because it is somewhat shade intolerant. In dense stands, the lower branches lose their leaves and the shrub develops a more columnar growth habit (Converse, 1984). The status of F. alnus as a seral species is supported by observations that it appears after forest clearance or similar types of land use change.

Control of the plant as a weed

F. alnus is difficult to eradicate because of its heavy fruit production, high germination rate and long-term seed viability. Mature plants that are cut near the base sprout vigorously and a combination of different control methods is normally needed, together with follow-up treatments over a period of several years.

Cultural methods include cutting, mowing girdling and burning (Converse, 1984). Cutting in early June and late August for two or three consecutive years reduced the number and size of the stems that developed and increased the vigour of the herbaceous groundcover. Cutting is very labour intensive and is normally suitable only for small or highly significant sites. Mowing at a height of 2-13 cm from the ground once or twice in the summer can arrest development but may stimulate vigorous resprouting; it does have the beneficial effect of limiting flowering, and prevents seedling establishment. Cutting alone without subsequent herbicide application is usually not effective (Reinartz, 1997). Sinclair and Catling (1999) noted an increase in the cover and number of native species on plots where F. alnus had been cut and, although it grew back quickly, they suggested that even a single cutting may give native species an opportunity to replenish the seed bank.

Girdling plants at the base with a 2- to 3-cm-wide saw cut into the phloem limits resprouting. Phloem should be removed without damaging the xylem, and the girdles should be checked after a few weeks to ensure that new bark is not developing (Heidorn, 1991). Likewise, heating the stem with a flame torch will kill the cambium; this is recommended for smaller stems up to 4.5 cm in diameter. Follow-up herbicide treatment is recommended following girdling (Reinartz, 1997)

F. alnus can potentially be topkilled by fire in late spring just after it has leafed out, because food reserves are low at this time and resprouting vigour is reduced (Converse, 1984). However, burning in autumn has little effect, and studies in Michigan, USA, indicate that stem density was twice as great the following summer as before the burn; the resprouts were one-third the height of the pre-burn stems. Cutting F. alnus in the spring at the leaf expansion stage, and again in the fall, followed by spring burning in the next two years may also be effective in fenland. Successive burns are generally required because F. alnus typically grows in moist areas; this limits fire severity and roots and seeds are often unaffected. Burning kills most seedlings, but new plants arise from buried seed reserves and quickly recolonize the area (Post and Klick, 1988; Post et al., 1989; Taft and Solecki, 1990). A combination of burning followed by planting with groundcover species reportedly limits the re-establishment of F. alnus (and Rhamnus cathartica). Only six Rhamnus/Frangula seedlings (species not differentiated) were recorded in a 144-m² test plot treated in this way compared with 99 seedlings in a burned plot, and 191 seedlings in the control plot (Scriver and Leach, 1998).

The raising of water levels in wetlands where the water table has been lowered artificially (Simon 1992), is also a means of controlling F. alnus (Heidorn, 1991).

Seedlings or small plants can be removed by weeding. Larger plants may be pulled out with a "weed wrench" or heavy equipment. However, excavation often disturbs roots of adjacent plants, or creates open areas that are readily colonized by new Rhamnus/Frangula seedlings. This technique may be most useful to control invasion at low densities, or along trails, roads and woodland edges (Anon., 2002).

Chemical control is the most common method of managing F. alnus. Because F. alnus grows in wetland environments, it is preferable to carry out treatment in the winter months when the soil is frozen and there is less risk of damage to native species from trampling and herbicide overspray (Reinartz, 1997). The preferred method is to cut and apply herbicide to stumps greater than 10 cm in diameter. Basal bark treatments can be applied on smaller shrubs or in areas where openings in the understorey are not necessary for restoration. Repeated treatment over several seasons is usually required to eliminate recruitment from the seed bank. Converse (1984) lists the following herbicides and methods that have achieved good control against F. alnus in the USA: wick-applied glyphosate in May-June; misting of cut stumps <5 cm in diameter) with glyphosate in August; misting with fomasine (ammonium salt) in September. Foliar application of 2,4-D in March-August was not effective because new leaves developed later in the year. Similarly, resprouting occurred following misting of larger stumps (>12 cm in diameter) with glyphosate in August.

Godwin (1943) lists several saprophytes associated with F. alnus, and noted that the fungus Nectria cinnabarina was identified as causing die-back in the 1930s. Experiments with this fungus have been conducted in North America in an attempt to eradicate F. alnus from Wisconsin, USA (Reinartz and Parker, 1989). Current research has focused on the role of insects as potential biocontrol agents (Gassmann et al., 2002); the principal species under evaluation are the following:

Zygana suavis (Cicadellidae) appears to have a strong preference for F. alnus and is common in northern Germany. Its potential impact is to impair photosynthesis and to act as a vector for pathogens. It is one of the very few insects strongly associated with the species.

Sorhagenia janiszewskae (Cosmopterigidae) attacks both F. alnus and Rhamnus cathartica. It is relatively common and easy to collect, but it is difficult to overwinter the adults. S. janiszewskae bores into the above-ground parts of the target shrubs. Its potential impact is to impair growth of the current year's shoots.

Ancylis apicella (Tortricidae) appears to have a preference for F. alnus . It is a bivoltine species with a prolonged impact on the target host.

Oberea pedemontana (Cerambycidae) bores into the branches and trunks of F. alnus and Rhamnus cathartica and causes structural damage and occasionally the death of the host tree. Work with this species is handicapped by a high rate of parasitism and by the time needed to complete its life cycle of 3 years.

Bucculatrix frangulella (Bucculatrigidae) is the only leaf miner recorded on F. alnus, but because it is active late in the season it is likely to have little impact on photosynthesis and water loss.

The univoltine defoliating butterfly Gonepteryx rhamni (Pieridae) shows a clear preference for F. alnus in Europe.

Synanthedon stomoxiformis (Sesiidae) is the sole root-boring species on F. alnus and Rhamnus cathartica. However, there is evidence that it may also attack Sorbus aria (Rosaceae) and more rarely Corylus avelana (Betulaceae) as alternate hosts.

The mites Aceria rhamni and Tetra rhamni have some potential as biological control agents, and may facilitate disease transmission.

Preliminary observations suggest that more information is needed on the habitat of F. alnus in North America and in Europe to determine which biological control agent is most suitable for a particular type of habitat and also which native buckthorn species might be at risk if they occur within the physiological range of a selected biological control agent. Sorhagenia janiszewski, Ancylis apicella, A. derasana, Oberea pedemontana and Zygana suavis merit further study as potential biocontrol agents for North America according to Gassmann et al. (2002).

Horticultural cultivars include F. alnus 'Asplenifolia' (Fernleaf Buckthorn), a narrow-leaved form with fine-textured foliage that gives it a graceful, ferny appearance, and 'Columnaris' (Tallhedge Buckthorn) a tall, narrow form that creates a dense canopy supported by numerous twisting stems which makes it ideally suited for row plantings and hedges. Both cultivars are marketed throughout most of eastern and central USA (Anon., 2001). The chromosome number is 2n=20 (Tutin et al., 1968).

F. alnus was introduced to North America from Europe more than 100 years ago. Once established it maintains itself due to prolific seed production, vigorous growth over an extended growing season and its ability to regenerate following burning and cutting. These characteristics make it difficult to eradicate. Repeated cutting and application of herbicides required to eliminate F. alnus is laborious and expensive. Consequently, most restoration work has been conducted in natural ecosystems of special interest. Its adverse effect on native species arises because F. alnus shades out understorey plants. Its aggressive character, especially in wetlands, is widely noted (Catling and Porebski, 1994).

Cultivated stocks of F. alnus are still available from nurseries, even though it is recognized as an invasive exotic in North America. In several US states and Canadian provinces, F. alnus is listed as a restricted noxious weed and it is prohibited to import, sell and transport it or its propagating parts. Since January 2002, the USA requires seeds of all species to have phytosanitary certificates, so this has limited importation from overseas suppliers. At present the economic impact of F. alnus is comparatively small, despite it being an alternate host for crown rust (Puccinia coronata) and Alfalfa mosaic virus. However, considerable time and expense have been invested to remove the shrub from native habitats. F. alnus forms dense, even-aged stands and causes habitat degradation by shading out rare species and causing a decline in native species diversity (Taft and Solecki, 1990). Lovely (1981) observed a decrease in the small white lady's-slipper orchid (Cyripedium candidum) under F. alnus in Wisconsin, USA. In North America it is especially prevalent in wetland environments where it is characterized as an aggressive alien.