Ophiostoma ulmi
(Dutch elm disease)

The fungal pathogens causing Dutch elm disease are cautionary examples of the dramatic effects that the introduction of exotic fungal pathogens can have. It was the international trade of timber and other products that made their intercontinental spread possible. O. ulmi caused the first pandemic of Dutch elm disease and is therefore among the most invasive tree pathogens known to man. Its remarkable spread throughout Europe, western and central Asia, and especially North America is often used as an example of the dangers of accidental introductions. After its introduction from an unknown origin, O. ulmi met several widely distributed, highly susceptible host species in three continents. Its close association with bark beetles from the genus Scolytus and Hylurgopinus made spread over large areas and integration into a fully functioning host-vector-pathogen system possible. Man has further helped to quicken the spread by transport of infested timber and firewood over large distances and/or removing obstacles such as areas free of elm trees. During its most active time (ca 1910-1940), O. ulmi killed at least 10% of the European elm population and caused even higher losses in the more susceptible American elm species (Brasier, 2001). However, O. ulmi has today lost most of its importance due to competition with O. novo-ulmi, which shows higher virulence, higher bark colonization ability and a higher growth rate at lower temperatures.

O. ulmi was the causal organism of the first Dutch elm disease pandemic starting around 1910. It was first recognized as a new disease in the Netherlands in 1919 and subsequently spread throughout most of Europe, western Asia and North America. In the 1930s/1940s a decline in the incidence of the disease was noted in Europe (but not in North America). Since the introduction of O. novo-ulmi (ca 1940; Brasier, 1990) into Europe and North America, O. ulmi has been replaced in most parts of its distribution range by this much more aggressive pathogen (Brasier, 2001).

Most of the distribution range of elm has been or was previously colonized by O. ulmi in Europe, western and central Asia and North America. Dutch elm disease is only absent in some counties of California and a few other states of the USA. These areas have quarantine regulations to prevent introduction of the pathogen (Anon., 1995, 2002b). Although O. ulmi is at present being replaced in most of its distribution range by O. novo-ulmi, it still poses a phytosanitary hazard. Possible introduction of the pathogen to other countries and continents, such as Australia, should not be disregarded.

O. ulmi is listed as a quarantine pest in several countries. It is still listed as a quarantine pest in several European countries (Switzerland, Hungary, Poland, Norway, Iceland) although the pathogen is (or was) well established there (OEPP/EPPO, 1999). Other countries which list O. ulmi as quarantine pest are Algeria, Morocco (OEPP/EPPO, 1999), Canada (Anon., 2003b), China (Anon., 1991), Australia (Anon., 2003a) and New Zealand (Anon., 2002a). However, it is not listed as a quarantine pest in any country of the European Union (OEPP/EPPO, 1999) and is not mentioned on the APHIS list of regulated pests for the USA (Anon., 2000). This may pose a significant risk, because the fact that a particular fungal species has been recorded in a given region does not mean that there is nothing to fear from further introductions of what is believed to be the same species elsewhere (Gibbs, 2001). This is highlighted by the fact that O. novo-ulmi was introduced into the UK unrecognized, when O. ulmi was already there. The discovery of the endemic, highly aggressive O. himal-ulmi in the Himalayas shows that a number of DED fungi or other pathogens of elm may still be undiscovered in some parts of Asia, awaiting introduction.

Most danger of introduction comes from logs and branches with the bark attached (including firewood), loose bark used for mulching and especially elm wood (with the bark attached) used for crates. The latter go mostly unrecognized and have been the source of introduction in many cases (Sinclair and Campana, 1978). Although the fungus may also be present in all parts of an infected elm (except for the seeds), the highest risk for introduction of the disease comes from the bark because it also harbours the vectors of O. ulmi. To date, there have been no reported cases of deliberate introduction of O. ulmi. However, it is theoretically possible that, for example, international mail may be used for such an intention.

O. ulmi was the causal organism of the first Dutch elm disease pandemic (ca 1910-1940) (Brasier, 1991). The disease is assumed to have had greatest impact on trees already weakened by some other factor such as insect damage or drought. O. ulmi mostly affected elms in the urban environment, where only single clones had been planted (e.g. in the Netherlands), whereas a large proportion of the elm population remained healthy (Gibbs, 1978).

Natural infections by DED fungi have been found only in elms (Ulmus spp.) and Zelkova carpinifolia. Both hosts belong to the family Ulmaceae (Sinclair and Campana, 1978). All three European elm species are susceptible, but U. minor shows at least some resistance against O. ulmi. Of the six American elm species, Ulmus americana, U. thomasii, U. alata, U. serotina and U. rubra are reported to be highly susceptible, whereas U. crassifolia shows some resistance. No data are available for U. mexicana (Smalley and Guries, 2000). The Asian elm species U. parvifolia and U. pumila show high levels of resistance and have been used in breeding programmes. U. pumila has been planted extensively to replace the endemic elm species in North America and Europe (Stipes and Campana, 1981). Other elm species from Asia that show high levels of resistance include U. davidiana, U. szechuanica, U. gaussenii, U. bergmanniana and U. castaneifolia (Smalley and Guries, 2000). Although Zelkova carpinifolia has been reported to be susceptible, Z. serrata is resistant (Sinclair and Campana, 1978). Inoculation studies were carried out with other members of the Ulmaceae and Celtidaceae. Celtis spp. always proved to be immune, whereas Planera aquatica proved susceptible. Trema spp., Holoptelea integrifolia and Hemiptelea davidii were resistant (Sinclair and Campana, 1978; Smalley and Guries, 2000).

Genetics

O. ulmi and its sister species O. novo-ulmi have been the subject of extensive genetic investigation because of the dramatic impact of Dutch elm disease. Most research has been aimed at differentiating between O. ulmi and O. novo-ulmi; however, relatively little is known about the basic genomic organization of O. ulmi. Five chromosomes have been identified in its haploid genome (Dewar and Bernier, 1993). The mitochondrial DNA of O. ulmi is 74-88 kb, compared to 48-71 kb in O. novo-ulmi.

Several methods have been used for differentiation between the two pathogens. Protein and isoenzyme polymorphisms are detectable which clearly separate O. ulmi from O. novo-ulmi (Bernier et al., 1983; Jeng and Hubbes, 1983; Jeng et al., 1988). Differences between the two species have been found in RFLPs of mitochondrial (Bates et al., 1993b), nuclear (Bates et al., 1993a) and ribosomal DNA (Hintz et al., 1993). RAPDs were used to differentiate the two species and to identify variation within the species (Pipe et al., 1995; Hoegger et al., 1996). The two species can be separated most effectively by PCR-RFLP of ribosomal DNA (Hintz et al., 1993; Harrington et al., 2001).

A strong unidirectional mating barrier exists between O. ulmi and O. novo-ulmi, supporting the species rank of both. In pairings, O. novo-ulmi strongly rejects O. ulmi as male, while O. ulmi accepts O. novo-ulmi fully as the male mating partner. Hybridization between the two species can give rise to hybrid isolates, which show altered morphology and a considerable loss of fitness. Such hybrids are considered to be transient in the pathogen population (Kile and Brasier, 1990), but may act as genetic bridges between the species (Brasier et al., 1998). In fact, recent investigations indicate that O. novo-ulmi has acquired important parts of its genome from O. ulmi (Brasier, 2000a, 2001).

Physiology and Phenology

O. ulmi is considered to be the least virulent of the three DED fungi. It causes a low percentage of defoliation on elm trees (e.g. 10-35% on the moderately resistant Ulmus x Commelin) (Brasier and Mehrotra, 1995) and is only a weak colonizer of elm bark (Kile and Brasier, 1990). However, O. ulmi appeared to be more virulent than O. novo-ulmi on Ulmus americana under controlled conditions at temperatures above 24°C (Smalley et al., 1993b). This may be due to the fact that the optimum growth temperature for O. ulmi is between 27.5 and 30°C, whereas that for O. novo-ulmi is around 22°C (Brasier, 1981). O. ulmi is thus considered to be naturally adapted to a subtropical environment (Brasier and Mehrotra, 1995).

Phytotoxic compounds are thought to be involved in the pathogenesis of DED fungi. In particular, the role of the hydrophobin cerato-ulmin (CU) in the development of DED has been the subject of extensive investigation. This compound had previously been reported as a wilt toxin and a correlation between high levels of CU production and high virulence has been demonstrated (Takai, 1974, 1980). CU production is another feature which differentiates O. ulmi from novo-ulmi: O. ulmi (the less aggressive species) produces no, or very low amounts of, CU in liquid culture (Scheffer et al., 1987). The low virulence of O. ulmi compared to that of O. novo-ulmi has been attributed to a weaker expression of CU caused by differences in the promoter sequence of the CU gene and the derived amino-acid sequence (Jeng et al., 1996). However, further investigations have shown that CU is not the main source of virulence for O. ulmi and O. novo-ulmi. Among other evidence, the over-expression of CU in a transformed isolate of O. ulmi did not lead to higher virulence. CU is therefore currently considered to be a parasitic fitness factor rather than a virulence gene (Temple et al., 1997; Temple and Horgen, 2000). This emphasises the fact that the physiological mechanisms involved in the pathogenicity of O. ulmi and O. novo-ulmi are still not fully understood and need further investigation.

The study of the vegetative incompatibility (vic) system of O. ulmi has proved to be most useful for the investigation of population structure (Brasier, 1984; Mitchell and Brasier, 1994). The vic system of O. ulmi strongly restricts the movement of nuclei and cytoplasmic elements between the mycelia of different isolates. This includes the spread of deleterious cytoplasmic virus particles (d-factors), which are present in many populations of O. ulmi (Brasier, 1984). High diversity in vic-types helps the pathogen population to maintain its fitness. Many inferences on population structure have been made by analysing the ratios of the two mating types in the populations of this heterothallic fungus. A striking difference was revealed between the structure in O. ulmi populations in Europe and North America. Populations from Europe showed high levels of heterogeneity for vic-types and ratios of mating types were almost equal. In contrast, a single vic-type predominated in North American populations and mating type A was dominant. This result suggested that mycoviruses were not as common in the American populations of O. ulmi as they were in Europe, or that the more susceptible American elms allowed the O. ulmi population in North America to tolerate a continuous high level of virus infection (Mitchell and Brasier, 1994).

Reproductive Biology

The life cycle of the DED fungi is one of the best studied and is often used as a textbook example of insect-fungus interactions. Two stages are distinct in the life cycle of O. ulmi: the parasitic phase involving host colonization, overcoming host resistance, and growth inside the host tissue; and the saprophytic phase inside the bark involving sporulation inside the breeding galleries of the bark beetle (Scolytus and Hylurgopinus) vectors. Because O. ulmi is reliant on dissemination by bark beetles, its lifestyle is perfectly adapted to its vectors.

When the young bark beetles emerge from their pupation chambers in spring they already carry the fungus. During the winter and spring the fungus has spread and sporulated extensively inside the breeding system of the beetles. The young beetle is often surrounded in its pupal chamber by synnemata of the fungus and has ample opportunity to catch a large number of spores (Webber, 1990). For maturation feeding, the young beetles fly to the twigs of healthy elms where they feed on the phloem and on the xylem tissue in twig crotches (Scolytus spp.) or branches (Hylurgopinus rufipes). The fungus is inoculated into a new host during this essential phase. After the establishment of infection, the fungus spreads throughout the xylem vessels of the elm tree in its yeast-like stage and causes a vascular wilt (Sinclair and Campana, 1978). To complete the disease cycle the pathogen must again come into contact with its vectors. Scolytid beetles must breed in the bark of the infected elm for this to occur. Because the galleries of the bark beetles superficially penetrate the xylem, the fungus is 'released' from the vascular system and starts growing into the phloem (bark) of the elm. Other O. ulmi genotypes are introduced into the bark with the beetles (Webber et al., 1987). The combined process of beetle and pathogen colonization of elm bark is known as the saprophytic phase (Webber et al., 1987). During this phase, O. ulmi grows in the breeding galleries of the bark beetles and a marked sequence in sporulation can be seen. The production of fruiting structures appears to follow a predetermined, ontogenetic sequence which starts with the mycelial sporothrix stage, is followed by the formation of synnemata, and ends with the production of perithecia (Webber et al., 1987). Sexual propagation is important for the fungus, to maintain the fitness of the population by creating new genotypes (Brasier, 1984). O. ulmi is a heterothallic fungus, therefore, both mating types (termed A and B; Brasier, 1991) have to be present for successful sexual sporulation (i.e. formation of perithecia). When the young beetles, which by now carry the fungus, emerge they fly to the branches of healthy elm trees for feeding and after successful inoculation of the fungus into the new elm host the disease cycle is complete.

Environmental Requirements

O. ulmi inhabits the xylem and bark of elms, particularly in and around the breeding galleries of scolytid beetles. Climatic conditions in the breeding galleries are important for the growth of the fungus inside the bark and its subsequent sporulation (saprophytic phase). Microclimatic conditions are optimal for the fungus in relatively thick parts of the bark, where the moisture content is highest. Prolonged exposure to high summer temperatures and the combined action of lower initial moisture content and lack of nutrients in the outer bark usually inhibits sporulation of O. ulmi in stems and branches of smaller diameter (Webber, 1990). The optimum temperature for growth of O. ulmi is between 27.5 and 30°C, which may be the reason why O. ulmi has never spread to the upper limit of elm distribution in the mountains or further to the north. For example, O. ulmi has never established in Norway, Finland, Denmark and northern Scotland (Gibbs, 1978). O. ulmi is considered to be adapted to a subtropical environment because of its high optimum growth temperature (Brasier and Mehrotra, 1995).

It is believed that the outbreak of O. ulmi in the first half of the twentieth century was triggered, to some extent, by human influence on the ecosystem such as the lowering of the groundwater table, pollution by aerosols, and intensified root and crown damage resulting from urbanisation (Brasier, 1990).

Associations

The DED fungi are most closely associated with their vectors, bark beetles from the genera Scolytus (North America, Europe and western Asia) and Hylurgopinus (in North America only). The bark beetles act not only as vectors but also provide sites for sporulation of O. ulmi in their breeding galleries. Scolytus scolytus is the most effective vector in Europe, carrying up to 350,000 spores (Webber, 1990). Two smaller species, S. multistriatus and S. pygmaeus, are also effective vectors (Faccoli and Battisti, 1997). Spore load varies between spring-emerging beetles (e.g. 58% beetles infested with spores) and summer-emerging beetles (e.g. only 8% beetles infested with spores) (Faccoli and Battisti, 1997). This is most likely to be related to climatic conditions in the bark, which are probably too dry in summer for extensive sporulation. The relationship between DED fungi and their vectors can be described as mutualistic, because both profit from their combined attack on elms. However, Webber and Gibbs (1989) found that the fungus has some negative effects on the bark beetle brood as the larvae have to feed on phloem tissue that is free of fungal growth for successful development from the first to the third instar.

The influence of mites, present in the breeding galleries of the bark beetles, on O. ulmi must not be disregarded. Although the mites eat a significant portion of the fungal biomass they also play an important role in disseminating fungal spores inside the breeding galleries of the bark beetles (Brasier, 1978). Brasier (1984) showed experimentally that the activity of mites could drastically enhance sexual sporulation of the fungus. Fransen (1939) reported that mites drag spores of the pathogen through frass in the larval galleries to the pupal chambers, thus forming a carpet of synnemata inside the pupal chamber, which enhances the opportunities for distribution of the spores.

Mites (Acarina spp.) that live in the galleries of the bark beetle vectors of O. ulmi could be considered as natural enemies of O. ulmi because they eat a large amount of its mycelium and spores. However, these mites also play an important role in the dissemination of different O. ulmi genotypes within the breeding galleries of bark beetles and thus facilitate the sexual sporulation of O. ulmi. Their effect on O. ulmi therefore appears to be more beneficial than detrimental (Brasier, 1978).

Fungi that compete with O. ulmi (except O. novo-ulmi) in the colonization of diseased bark are Phomopsis oblonga [Diaporthe eres], Botryosphaeria stevensii and Nectria coccinea. D. eres is a very effective antagonist of O. ulmi in vivo as it is already present in healthy elms and can invade the diseased phloem before the bark beetles start to breed (Webber, 1981; Brasier, 1996).

However, the most important natural antagonists for O. ulmi are fungal viruses and virus-like RNAs, which are of common occurrence in fungi. An extranuclear virus-like factor (the d-factor) which causes degenerative disease in DED fungi was first identified by Brasier (1983). D-factors are associated with multiple virus-like RNA segments (Rogers et al., 1986). Infected isolates show an unstable amoeboid colony morphology and infection results in a severe reduction in growth rate and aggressiveness (Brasier, 1983, 1986a). D-factors are spread within the fungal population by hyphal anastomosis and transmission is most effective in isolates of the same vegetative incompatibility group (Brasier, 1983). The impact of d-factors on populations of O. ulmi is believed to be high, because the spread of d-factor within the European population of O. ulmi was probably involved in the recession of the first Dutch elm disease pandemic in Europe (Brasier, 2000b).

Malt extract agar (MEA) supplemented with streptomycin and cycloheximide is the best medium for isolation of O. ulmi from field samples (Brasier, 1981). Chips of symptomatic (discoloured) wood, which has been surface sterilized, are placed on the selective medium and the fungus will grow on it within a few days. Alternatively, bark beetle frass or spore masses taken from the galleries are placed onto the agar. Asexual sporulation may be observed after a week. Perithecia are only formed on elm wood or media containing elm wood (e.g. elm sapwood agar; Brasier, 1981) when both mating types are present. Differentiation of O. ulmi from the similar O. novo-ulmi relies on laboratory testing; colony morphology and growth rate at 33°C are used as a means of evaluation. Petri dishes containing Oxoid MEA are inoculated with an isolate of the fungus and incubated in darkness at 20°C for 48 h. The diameters of two colonies are then measured at right angles from the reverse of each plate. The plates are incubated for a further 5 days and measured again. The mean growth per day is calculated. O. ulmi will grow at 2.0-3.1 mm/day, whereas O. novo-ulmi grows at 3.1-4.8 mm/day. The plates are kept for a further 10 days in diffuse light at room temperature to bring out the characteristic colony morphology. It is important that Oxoid MEA is used for morphology examination as this medium brings out very clearly the distinct characteristics of both species. A further growth test at 33°C must be conducted. O. ulmi will grow at 33°C with a growth rate of 1.1-2.8 mm/day, whereas O. novo-ulmi will only grow between 0 and 0.5 mm/day (Brasier, 1981, 1991).

The detection of Dutch elm disease on elm is usually not difficult as the wilting of branches is very conspicuous. However, early symptoms are often difficult to detect in large trees. Examination of the xylem of an affected branch will often reveal a brown discoloration of the outer year ring on a transverse cut. Brown streaking will be seen in the outer xylem of a twig or stem stripped of its bark. Laboratory isolation of the pathogen is needed to rule out infection by other pathogens.

Bark samples of logs or firewood of elm should be inspected for signs of bark beetle breeding. If any bark beetle galleries are found they should be inspected carefully for the presence of mycelium or fruiting structures of O. ulmi. No sporulation will be seen under conditions that are adverse to the fungus even if it is present in the bark. Bark samples should be put into a moist chamber and inspected after a few days for fungal growth. For assurance, isolations are needed.

The symptoms caused by O. ulmi infection can easily be mistaken for those caused by O. novo-ulmi; however, in the case of O. ulmi the disease progresses much more slowly. For identification, it is essential that the fungus is isolated from its host and further laboratory examinations performed (Brasier, 1981).

Another vascular wilt of elm is Verticillium wilt, caused by Verticillium dahliae and V. albo-atrum. These soilborne pathogens have been reported from Europe, Asia and North America (Stipes and Campana, 1981). Elm infection with Verticillium fungi usually starts with wilting of the leaves on one or several twigs on a branch. External symptoms of Verticillium wilt resemble those of Dutch elm disease, but crown involvement is less prominent. Systemically infected elms show foliage discoloration, sharply reduced twig growth, and partial leaf abscission on severely infected branches. Branches often die during the dormant season. Before abscission leaf blades may develop necrotic patches on their surface. A characteristic symptom of Verticillium wilt is sapwood discoloration in the roots, stems and branches. Extensive invasion of the elm xylem with the fungus may ultimately result in the death of younger trees and decline in older trees. Positive diagnosis of Verticillium wilt relies on isolation of the pathogen on suitable media in the laboratory (Stipes and Campana, 1981). Verticillium wilt is more prevalent in nurseries and on ornamental trees than in forests (Butin, 1995).

Dothiorella wilt is another wilt disease that affects elm trees. The causal organism, Dothiorella ulmi, is only known to occur in North America but is not believed to be native there (Stipes and Campana, 1981). Symptoms of this disease include wilting, yellowing of leaves and dieback of the affected branches. Flat cankers develop on newly killed parts of the bark and pycnidia develop on the bark. Brownish discoloration can be observed in the outer annual rings of the xylem. The disease progresses slowly and may take several years to kill a mature elm tree (Stipes and Campana, 1981). This disease is distinguished from Dutch elm disease by the slower progression of dieback and the presence of cankers and fruiting bodies on affected branches. Laboratory isolation of the fungus may be necessary in some cases to distinguish it from the DED fungi (Boyce, 1938; Stipes and Campana, 1981).

Elm yellows (also known as elm phloem necrosis) is a debilitating or lethal disease on elms caused by elm yellows phytoplasma. Although the disease was first noticed in the USA (Stipes and Campana, 1981), it is now believed to be widespread in Italy and some parts of France (Mittempergher, 2000). The disease has been largely overlooked in these areas because it causes only minor disease effects on European elms and the number of symptomatic plants is low. Symptoms on European elms include the formation of witches' brooms and stunted growth. However, the situation is much more dramatic in North American elm species. Elm yellows is widespread in the eastern USA, where sporadic epidemics have killed endemic elms, but not those of Eurasian origin. Elm yellows symptoms on American elm species usually involve rootlet necrosis, degeneration of conductive phloem and cambium in the roots and the lower trunk, foliar epinasty, yellowing and leaf casting. Foliage changes colour in only a few weeks from green to yellowish-green and reddish-gold, before the leaves are cast or suddenly wilt, shrivel, turn brown but remain attached to the twigs. At first only individual branches may be affected but eventually the whole tree will be affected and die. The degenerating phloem is a uniform yellow colour and may have small, necrotic lesions (elongated brown spots on the inner bark surface). Freshly diseased phloem in most North American elms species has a characteristic winter-green odour (Sinclair, 2000). Elm yellows can be distinguished from Dutch elm disease by the different combination of symptoms (in Elm yellows all branches are affected at once) and the discoloration of the phloem which is not usually seen in a fresh infection with O. ulmi. The causal organism cannot be isolated onto artificial media (Tainter and Baker, 1996).

Low levels of nutrients, summer drought, re-plantation of elm trees, and insect and mite injury can result in leaf discoloration, wilting and leaf fall (Stipes and Campana, 1981). These conditions can be confused with symptoms of Dutch elm disease; however, the whole tree is usually affected, not only single branches, and recovery is common in the following year. Isolations are needed to distinguish these causes from infection by O. ulmi or O. novo-ulmi.

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.

Phytosanitary Measures

The import of elm wood with the bark attached is not allowed into countries with a quarantine for O. ulmi (e.g. Australia; Anon., 2003a). Debarked elm timber has to be properly treated for import. Treatments can consist of kiln drying, ethyl bromide fumigation or other methods able to disinfect the logs (e.g. Anon., 2003b). All shipments of elm wood have to carry a phytosanitary certificate (FAO, 1997) and are subject to inspection. Shipment of live plants of Ulmus are prohibited to some countries such as China (Anon., 1991). The same quarantine regulations apply to Zelkova spp. in some countries (OEPP/EPPO, 1999), and also to Planera sp. (Anon., 2002b). Although seeds have not been proven to carry the disease, the import of elm seeds is prohibited into India with reference to the DED fungi (Anon., 1989).

Cultural Control and Sanitary Methods

Sanitation is the most effective way of controlling DED fungi and may account for 80% of the total effort in urban control programmes (Stipes, 2000). It is directed against both the pathogen and its vectors. Sanitation can significantly reduce the probability of infection, but cannot eradicate the pathogen or vector. It involves the destruction (burning) of all elm wood attractive to beetle infestation, such as any elm stems and branches that are weakened, dying, or dead from the disease or any other cause.

Removal of the entire vascular lesion by pruning is often sufficient for an elm infected by O. ulmi to recover from Dutch elm disease. For this approach to be effective, two prerequisites should be taken into account: the elm may not wilt before mid-summer, and early disease detection is necessary, i.e. the elm may not have more than 5% crown damage (Himelick and Ceplecha, 1976). Wilt symptoms observed early in the season are indicative of a DED infection that occurred in a previous year. In this situation, the disease will be far more established than is apparent from foliar symptoms, and therapeutic pruning will probably fail. All branches showing the characteristic streaking in the outer sapwood have to be removed. However, if the streaking has reached the stem the infection will persist in the tree. Cutting tools should be sterilized between different trees. The success of eradicative pruning is limited when bark beetle populations are large, and is most effective in areas where multiple or successive infections are rare (Sinclair and Campaign, 1978). Eradicative pruning is mostly used in urban areas, often together with the application of other (chemical) disease treatments. However, it is not practical in the forest (Haugen, 1998; Stennes, 2000).

Root grafts have been shown to spread O. ulmi infection between trees, especially in hedgerows (Haugen, 1998). Root severance must therefore be carried out between infected and healthy trees. A spade or trench-digging machine is used for this purpose depending on the size of the tree (Burdekin and Gibbs, 1974).

Host-Plant Resistance

Extensive work has been done in Europe and North America to exploit resistance in the host. Early elm breeding was carried out, mostly in the Netherlands, and several clones with good resistance against O. ulmi were released. The clone 'Commelin' was extensively planted in the late 1960s for its good shape and high resistance against O. ulmi (Heybroek, 1993a). A range of elm cultivars with good resistance against O. ulmi were also released in North America (Townsend and Santamour, 1993; Smalley et al., 1993a). However, with the appearance of O. novo-ulmi, many of these cultivars proved susceptible to this more aggressive species and breeding programmes were set back to the start.

Biological Control

Injection of conidial spores of a hypovirulent strain of Verticillium dahliae into the xylem of a healthy elm tree has proven to be effective in inducing resistance to DED fungi (Scheffer, 1990; Elgersma et al., 1993). This biological control system has been tested extensively in the Netherlands and USA and is now commercially available in both countries (Voeten, 2003).

Chemical Control

Chemical control against DED fungi has been investigated since the mid-1930s and over 600 compounds have been tested for their ability to manage DED (Stipes, 2000). Chemicals can be applied by soil application or by injection into the vascular system of the tree. The former method has certain environmental drawbacks and therefore the injection of systemic fungicides is now preferred.

Six chemicals are currently available in the USA for injection: three benzimidazole compounds (carbendazim phosphate, thiabendazole hypophosphite, benzimidizole carbamate), two triazole compounds (propiconazole and tebuconazole) and a patented formulation of copper sulphate pentahydrate (Haugen and Stennes, 1999). Thiabendazole hypophosphite and propiconazole are the compounds most often used as they show good systemic qualities and work selectively against certain ascomycetes and fungi imperfecti (Klopping, 1960). Injection or infusion (without applying pressure) of systemic fungicides into the xylem is either applied as a prophylactic or therapeutic measure (Stennes, 2000).

Chemicals are applied by exposed root flare injection. Only elm trees in good condition (despite O. ulmi infection) and with minor symptoms (5-15% symptoms in the crown) should be treated with fungicides (Scheffer et al., 1988; Stipes, 2000). Systemic fungicides move with the transpiration stream through infected sapwood that is still functional, stop the pathogenic action of the fungus (but do not kill it), and allow the tree to wall off the infection with a layer of new sapwood. It is most important that the chemical is distributed completely throughout the crown of the affected tree during the injection procedure (Haugen and Stennes, 1999). Symptomatic branches should be removed after successful application. Success rates of between 55 and 79% were achieved for thiabendazole and propiconazole (Stennes, 2000). If the infection has spread into the roots, no chemical treatment will be effective.

The application of chemicals has certain disadvantages. Generally it does not kill the fungus but only inhibits it, thus the remission of symptoms can be expected. Furthermore, severe wounding at the injection site in the elm tissue and phytotoxic effects on the leaves are commonly observed. Additionally, the treatment has to be repeated at maximum intervals of 3 years. For economic reasons, chemical control is usually only applied to single trees of high value (Haugen, 1998).

IPM Programmes

Many cities in North America, for example, Minneapolis, USA (Stennes, 2000) and Winnipeg, Canada (Allan, 2000) have made DED control a priority and have invested in effective disease management programmes. A 15-year study by the United States Department of Agriculture concluded that control programmes cost 37-76% less than doing nothing and having to remove killed trees and replant new plants (Anon., 1977). An annual loss of 1-2% of the trees in a community is seen as successful management (Cannon et al., 1977). Communities that do nothing may lose 90% of their elms within a decade. A successful example of disease management is the City of Fredericton (New Brunswick, Canada) which was able to save 70% of its elms after 30 years of Dutch elm disease in the area. In 1990 the annual loss was only 0.5% (Magasi et al., 1993).

Sanitation combined with the use of insecticides and root graft severance is the most effective approach to control (Anon., 1977). Sanitation is the most important component because it is aimed at both the pathogen and the vector. A primary inventory of elm stands present should be conducted for the timely detection of disease symptoms. Subsequent systematic surveys should be carried out at least twice each year during the growing season (Stipes and Campana, 1981). Elm wood infested with bark beetles or trees that could provide sites for breeding have to be removed promptly. All diseased trees have to be felled and destroyed as soon as possible after infection is detected, within 2-3 weeks during the growing season or before April in the dormant season (Haugen, 1998). All elm firewood has to be destroyed by early spring. Harvested elm wood has to be debarked and dried. Wood can be immersed in water for long-term storage (Stipes and Campana, 1981). Declined branches that may attract beetles must be pruned off. If the disease is localised in a single branch, eradicative pruning may be sufficient to halt disease development. The severance of root grafts is another effective measure used to contain the disease in urban plantations. Single trees of high value can be further treated with fungicides.

Pheromone trapping of vectors can help to estimate the population size and identify high-risk locations (Haugen, 1998; Gadgil et al., 2000). The use of trap logs has been successful in reducing the number of bark beetle vectors. Trap logs should be treated with an insecticide or debarked and burned (or buried) before the beetles emerge (Stipes and Campana, 1981). The application of insecticides to the crown of the tree in order to kill the vectors is often not effective and not environmentally justifiable (Haugen, 1998). The treatment of the lower bole with an insecticide in late summer or early autumn is an effective means of killing or eliminating overwintering adults of Hylurgopinus rufipes (Haugen, 1998).

The replacement of susceptible elms with resistant cultivars is necessary to reduce the impact of the disease. Plantation of other species is a common practice and should be done because it was monocultures of elm that made the immense impact of Dutch elm disease possible in the first place. Elms are remarkably protected when grown in a diverse or heterogeneous community of other tree species (Stipes and Campana, 1981). The eradication of all elms from an area has been suggested, and should also be considered under certain conditions (Haugen, 1998).