Forestry and Forest Products, CSIRO, PO Wembley 6014, Western Australia

Background and objectives
Phytophthora is the most devastating forest pathogen worldwide. P. cinnamomi has been introduced into European and North American hardwood and softwood forests and southern Australia. There it decimates species from a wide range of plant families in native forests, woodlands and heaths, changing community structure and function. The disease threatens rare flora and fauna and world heritage areas. As an introduced organism P. cinnamomi has the opportunity to interact with and evolve in association with a wide range of new hosts in a wide range of ecosystems and microsites, e.g. SW Australia has about 2000 indigenous susceptible species, mostly long-lived woody ones. Distribution of the pathogen at local, national and international levels is of concern to land managers. New introductions have the potential to contribute even more unwanted variation to the pathogen's genetic pool and exacerbate the challenges of ecosystem-scale control.

Results and conclusions
P. ;cinnamomi is a highly variable pathogen worldwide. Genetic variation within mating and isoenzyme types is as substantial as phenotype variation in pathogenicity, physiology and morphology. Both mating types occupy some sites, giving the potential for gene flow through the population sexually and asexually. Crucial genetic characters for this soilborne pathogen are a capacity to change unpredictably in any character, including those affecting disease and adaptation to edaphic or other environmental factors. Environmental variables including a range of site, soil and seasonal factors substantially change disease expression to P. cinnamomi. Such factors may directly affect both host resistance and capacity of the pathogen to invade host tissue. For example, variation in temperature and tissue moisture can cause highly resistant or highly susceptible responses in clonal eucalypt lines inoculated with highly pathogenic isolates. Understanding these and other variables is crucial to programs of selection for resistance in native forest, woodland and heath species. The understanding also improves prediction of differences in disease vulnerability between sites, thus affecting the choice between land management options and mitigating disease impacts.

Large differences exist in the susceptibility amongst species, both between and within some species of native plant genera. Conservation for biodiversity of the genetic structure of native host species is an important consideration in managing this disease. In infected sites, individuals of plant species at the very susceptible end of the scale do not survive to set seed; others species regenerate predominantly by vegetative growth so that in some communities all the woody species are replaced by resistant sedges and annuals. In species with some field resistance, population resistance may be enhanced by natural selection or promoted by management introductions of more resistant forms from selection and breeding programs. Resistance in native vegetation involves non-co-evolved host-pathogen associations and it is most probably due to general and not host-specific genetic resistance factors. Disease control in long-lived woody plants is a long-term venture. Programs selecting and breeding for disease resistance in native forest, woodland and heath species are challenged by the need to develop long-term resistance.

Resistance to P. cinnamomi in a wide range of woody species can be enhanced by foliar spraying with low concentrations of phosphite. Phosphite remains effective for more than 4 years in species having slow growth rates and high biochemical recycling of nutrients. It provides an option for disease management in native vegetation, especially for rare and endangered communities. It does not kill the pathogen in soil, but appears to reduce propagule production of treated plants (Shearer & Komorek, personal communication).

Phosphite application, deploying resistant plants, quarantine measures and other types of disease management are a considerable cost to the community. This is particularly so when control must be achieved in large tracts, often on public lands serving many, sometimes conflicting, interests including timber production, recreation, biodiversity and conservation.