For a specialized root-infecting fungus such as Heterobasidion annosum, infection by means of the minute inoculum represented by spores can only be expected to occur under severely circumscribed conditions (Rayner and Boddy, 1988). Infection of intact, healthy trees is prevented by active and passive host resistance, whereas at the other extreme infection of severely suppressed trees is constrained by competition from saprotrophs. By contrast, exposed wood presents a means by which a strategy of stress tolerance can be exploited to reduce competition from other organisms. However, the opportunities presented by stumps and wounded trees represent a balance between host resistance and declining selectivity against saprotrophs. Newly created conifer stumps provide the optimum avenue for infection but the period of susceptibility is short. Infection may also be limited by the availability of inoculum, by moisture and temperature relations within stumps and by competition from other organisms. Host species appears to be relatively unimportant, though infection seems to be particularly successful in pines, as noted by Cobb and Barber (1968). There is little evidence that suppressed trees are susceptible to spore infection, either with or without wounding, but vigorous trees can be infected through stem or root wounds. Stem infection is particularly common in Tsuga heterophylla and, where damage is frequent and severe, could rival that arising from stumps as a means of infection. In Europe this type of infection seems to be most common in Picea abies, but the majority of studies suggest that it occurs at relatively low levels. There is also evidence that the fungus may not persist in infected trees (Bazzigher, 1986; Dimitri and Schumann, 1989; Hallaksela, 1993; J.N. Gibbs, 1996, UK, personal communication), but in this work trees were inoculated via bore-hole wounds, which tend to heal rapidly. Larger surface wounds remain open for much longer, and the fungus may not behave in the same way. Nevertheless, wounds are potentially an important means by which H. annosum could become established in crops planted on former farmland despite protection by stump treatment. Those caused by mammals would similarly enable the disease to become established in crops on poorly drained soils which might not be thinned because of windthrow hazard. There is no evidence that dead roots can be infected by soilborne spores or that infection of minor root wounds, such as those created by stone abrasion, rodents or insects, is common. Significant root infection appears to require major wounds, such as those created by deer or harvesting machines. Most ot the available information on this method of infection concerns Picea abies and more information is required for other species, particularly those susceptible to infection through stem wounds. It follows from this review that although H. annosum has a wide host range among conifers and is present in unmanaged forest, serious disease is confined to managed forests in which regular cutting takes place. Infection of stumps and trees by spores clearly has greatest epidemiological significance in plantations newly established on land where woody inoculum is lacking. In areas remote from spore sources the risk is initially low; but once basidiocarps are formed locally, following the first infections, the risk that infection will occur in subsequent thinnings is greatly increased. The possibility that a small proportion of stumps could be particularly susceptible would favour the process in some species. Where old-growth forests are cut and regenerated, the situation isessentially similar, although the low level of infection that is likely to have been present in the original unmanaged forest can be expected to provide local spore sources. A more rapid build-up of woody inoculum might therefore be anticipated. Edmonds et al. (1989) expressed concern that early (pre-commercial) thinning of dense Tsuga heterophylla regrowth which develops following cutting of old-growth forest in the Pacific north-west of North America may lead to increased spore loads. In successive rotations of plantations established on land without a previous forest history, an increasing proportion of stumps from felled trees will be occupied by H. annosum as a result of infection of trees prior to felling through root contact with infected stumps, or by secondary infection from other trees (Stenlid, 1987; Redfern et al., 1994a) tather than by spore infection. In susceptible species such as Picea sitchensis this could be as high as 70% (Pratt, 1979a). At this stage, too, spore infection of stumps might be increasingly restricted by competing fungi as a mature forest ecosystem develops (Stenlid, 1993). Nevertheless, models (Pratt et al., 1989) and stump sampling (Piri et al., 1990; Piri, 1996) suggest a balance will be reached and some stumps will remain available for spore infection. Thus, although the importance of spore-infected stumps as a direct cause of damage can be expected to decline with time, it will continue to have significance for genetic diversity.
