SUMMARY

The phytopathogenic fungus Chondrostereum purpureum is being developed to control competing forest vegetation in areas used by utilities and forest management agencies. Traditional methods of vegetation management have relied on the application of chemical herbicides to prevent resprouting from cut stems. Growing restrictions on the use of chemicals in forests, particularly near streams and other environmentally sensitive areas, are driving the search for viable alternatives for vegetation management. The overall objective of our research program is to develop safe and effective methods for the control of fast-growing hardwood trees which are considered to be "weeds" in Canadian forest renewal sites and utility rights-of-way. The biocontrol agent, living mycelium of the fungus C. purpureum, is applied to the cut surface of stumps following operational clearing and eliminates resprouting from the main stem. Several criteria, including environmental fate and risk assessment, must be met before C. purpureum can be registered for commercial use in Canada and the US. We are developing molecular genetic markers to monitor environmental fate and persistence of our lead isolate following operational use. Chondrostereum purpureum represents the first biocontrol agent for forestry applications.

INTRODUCTION

Mycoherbicides are formulations of plant pathogenic fungi that are applied directly to target weeds to repress or eliminate growth. Plant pathogens which have been developed for commercial use include the pathogenic fungus Phytophthora palmivora ( DeVine) for the control of strangler vine in citrus groves and Colletotrichum gloeosporiodes f. sp. aeschynomene (Collego) for the control of northern joint-vetch in rice and soybean fields (Templeton et al., 1984). The development of a candidate fungus usually proceeds from the discovery of diseased target weed in the field. The causal agent of the disease is cultured on artificial media, identified and then re-inoculated on the weed host to establish a causal relationship between presence of pathogen and disease symptoms. Successful reproduction of the disease and re-isolation of the fungus from the weed host establishes the fungus as a potential candidate for development as a biocontrol.

Mycoherbicides have classically been considered as selective controls for a single dominant weed species but pathogens having a wider host range should also be considered for the management of a number of weed species. This is especially important in North American reforestation sites where competition from a number of fast growing hardwood species, such as red alder, black cottonwood, aspen poplar, birch, bigleaf maple and Douglas maple, has reduced productivity of reforestation sites by causing reduced growth of the planted conifers. Availability of light, nutrients, water and space to conifer seedlings are all decreased by the more vigorously growing hardwood species hence these species are considered to be "weeds" in this context (Wall et al., 1992). These same species also pose a hazard to power lines hence hydro-electric power companies must ensure that rights-of-way (ROWs) are kept clear of trees in order to maintain uninterrupted power service. Since these same species may have commercial importance in other settings, such as hardwood plantations for pulp and paper products, control must be achieved in a highly controlled and site-specific manner. Manual brushing has been used to control these species but often results in an even more vigorous regrowth (Wall et al., 1992). Most of these species propagate by prolific resprouting of numerous shoots from cut stumps or rhizomes hence continued slashing is required following the initial cutting. To prevent resprouting from the cut stumps chemical herbicides, such as glyphosate (Round-Up^®, Vision^®, Carbopaste^®) trichlopyr (Garlon^®), and imazapyr (Arsenal^®), are routinely applied to the stump following slashing. Chemical herbicides have proven to be effective in preventing reprouting however their use is increasingly encountering public opposition due to deleterious effects on non-target vegetation and riparian zones (de Jong et al., 1990; Jobidon, 1991). We have therefore initiated a research program to develop the phytopathogenic fungus Chondrostereum purpureum (Pers: Fr.) Pouzar as as the first biocontrol agent for use in Canadian forests as an alternative to the use of chemical herbicides in reforestation sites and hydroelectric power utility rights of way (ROWs).

Because most forest "weeds" in Canada are native species and serve some useful role in forest ecology, one cannot rely on exotic pathogens or insects to effect biological control. It is preferable to consider indigenous fungi as candidate species for mycoherbicide development. The advantage to this approach is that native pathogens are subject to natural controls and normally do not persist in the environment at greater than endemic levels (Wall et al., 1992). For sites having a diverse weed population it is preferable to develop a mycoherbicide with a wider host range. The phytopathogenic fungus C. purpureum is well suited to forest vegetation management because it has a broad-spectrum pathogenicity towards many hardwood species including those considered problem species in ROWs and forest renewal sites. This fungus is relatively safe to use as it can only invade through fresh wounds or cut stumps. Chondrostereum purpureum is a primary invader of wounds and once established it can cause cambial necrosis, sapwood stain and decay, foliar discoloration and eventually host death (Wall, 1986). Upon weakening the host it is quickly replaced by other more competitive saprobic fungi such as Coriolus versicolor and Schizophyllum commune. This rapid succession of fungi ensures that C. purpureum will not persist at high levels following the local inundation of stumps within a field site (Wall et al., 1992). The natural occurrence of C. purpureum in Canadian forests allows application of this fungus without the risk of introduction of a foreign species.

RESULTS AND DISCUSSION

Efficacy trials. Forest vegetation is managed on variety of industrial sites and the type of vegetation management used is often site-specific and integrates several approaches. For silvacultural purposes, competing vegetation is controlled to improve tree seedling survival and growth. Methods used to manage vegetation are chosen based on plant type, terrain and intended use of the area and can include manual, mechanical, and chemical treatments. On industrial sites, such as hydro-electric ROWs, combinations of methods are often recommended to remove high-growing vegetation and encourage the growth of compatible low-growing vegetation (CVMA Handbook, 1996). Newly developed biocontrol options must be sufficiently flexible to fit into current vegetation management practices. From a biogeoclimatic perspective British Columbia is a large and diverse province and generally is considered to be a cool, moist, mountainous, forested region however the province also has areas with Mediterranean-type, semi-arid, subarctic and alpine climates. While most areas are forested there are also extensive areas of grassland, wetlands, scrub and tundra (Meidinger and Pojan, 1991). Valentine et al. (1978) have divided British Columbia into five primary physiographic regions including to the south the Coastal Mountains and Islands, the Interior Plateau, the Columbia mountains and Southern Rockies, and to the north the Northern and Central Mountains/Plateaus, and the Great Plains. These regions have been further divided into 14 biogeoclimatic zones according to location, climate, dominant plant species, general ecological conditions and general features of zonal ecosystems (Meidinger and Pojan, 1991). Because of the diversity of target hardwood species in ROWs and reforestation sites across province additional research on efficacy the C. purpureum on several target species in the diverse biogeoclimatic ecozones must be undertaken.

The first full-scale experimental field trial in British Columbia was established with the cooperation of B.C. Hydro in September of 1994 in the Cowichan lake district of Vancouver Island. The objective was to determine the efficacy of C. purpureum for control of red alder compared to the commercially used herbicide glyphosate at a site on Vancouver Island. An even-aged stand of red alder was selected having a tree height of about 10 m and a stem diameter of between 5 and 7 cm. For this comparative trials the trees were cut by chainsaw at a height of 15 cm and all stumps of a clump (up to 30 stems) were similarly treated with one of: a mycelial preparation of C. purpureum formulated in a nutrient paste, a 12 % glyphosate spray (Vision^® or RoundUp^®), or a paste formulation of glyphosate (Carbopaste^®). Control stems were cut and either left untreated (slash alone) or treated with the nutrient paste formulation lacking mycelium. Five replicates of each treatment were setup in a randomized block design. Ten months after treatment (July 1995) the average number of shoots per stump, the average dieback of the stump, and the maximum shoot height were assessed for each treatment. The data were subjected to ANOVA and multiple range tests (DMRT) analysis. Analysis of the results indicated that control of red alder with C. purpureum was as effective as that obtained with both paste and spray applications of chemical herbicide glyphosate. The numbers of live shoots and maximum shoot height of treated stumps was not significantly different between the herbicide-treated and C. purpureum-treated stumps and both were significantly reduced from the controls (slash alone and formulation alone). The measured parameters were not significantly different between the slash alone and the formulation indicating that components of the nutrient formulation were indeed themselves inert and did not interfere with reprouting. The number of live resprouts and the maximum height that the resprouts attained were the most important variables studied in this trial with regard to vegetation management. The results indicated that there was no statistical difference between treatment of red alder with C. purpureum and glyphosate hence both could be considered to equally effective. Parallel trials are being established to test the performance of C. purpureum on several other target species including birch, aspen, and Sitka alder in the drier regions of the interior plateau and the colder regions of the Northern and Central Mountains.

Genetic characterization. While C. purpureum is an indigenous microorganism occurring naturally in all of the ecozones of Canada, little is known about the population structure of this fungus across the country. Prior to the production and release of a biocontrol agent it is necessary to determine whether a single isolate can be applied across a large area or if treatment must be regionalized using local isolates collected from each region. There is a possibility that a rare virulence allele could be introduced into a local population of the pathogen from a released isolate hence the genetic variation of the population must be determined in order to assess the impact of the release of a single isolate on the natural population.(Templeton and TeBeest, 1979). Ideally strains chosen for development should be genetically similar to local populations to reduce the risk of the introducing novel alleles into a local ecosystem. To address these issues we initiated an examination of the population structure of C. purpureum by first measuring variation in the ribosomal DNA (rDNA) repeat of 107 isolates collected from Europe, New Zealand and North America (Ramsfield et al., 1996). Three distinct nuclear type patterns were identified in the non-transcribed spacer (NTS) of the rDNA using the restriction endonuclease Hae III. The North American populations could be easily distinguished from the European and New Zealand populations according to this marker but there were only minor rDNA differences observed within each continent. The two rDNA patterns found in North America demonstrated an unequal east-west distribution with a higher frequency of each type on the coasts. In central North America the two types were found in roughly equal frequency and there were no coastal sites where the rarer type was entirely excluded. If gene flow across the country is frequent and equal there would be a reduced risk of releasing a single isolate across large geographic distances. These initial results suggested that there were no major barriers to gene flow for C. purpureum across North America (Ramsfield et al., 1996) however continued research is required to support this hypothesis. We are in the process of assessing genetic variation in North American populations of C. purpureum using techniques such as of randomly amplified polymorphic DNA (RAPD) and sequence characterized amplified regions (SCAR) markers (McDermott et al., 1994; Paran and Michelmore, 1993; Milgroom and Lipari, 1995) which are more sensitive to low genetic variation and will allow us to better measure the relative rate of gene flow across the country. These genetic markers will also allow us to "DNA fingerprint" our lead isolate and allow us to follow the environmental fate of released isolate.

Environmental fate and risk assessment. The application of C. purpureum to the cut stumps in the Cowichan valley field site increased the local concentration of the lead isolate relative to the natural background of C. purpureum found as part of the local ecosystem. Approximately 18 months after inoculating the stumps we observed that most were abundantly covered in shallow bracket fruitbodies. Chondrostereum purpureum, being an obligate pathogen, forms fruit bodies once the host stump has been killed in order to release spores and continue its life cycle. The fruitbodies are quite ephemeral and the incidence of C. purpureum, as measured by the presence or absence of these fruiting structure, drops back to endemic levels within two to three years (Wall et al., 1992). Chondrostereum purpureum is normally disseminated by short-lived sexually-derived wind-borne basidiospores which are released during periods of prolonged rainfall and high relative humidity (Spiers, 1985; 1988; Wall , 1995). If two spores of compatible mating type come in contact on a fresh wound of a susceptible host, infection and mating can occur and the cycle is then repeated. Commercial fruit orchards could be at high risk of infection during the early spring due to pruning practices which can provide entry points for the fungus. The risk of infection to non-target trees was assessed in the Netherlands by series of field experiments, surveys, simulations, and models of basidiospore dispersal using a Gaussian plume model (deJong et al., 1991). The added infection caused by inundative application of C. purpureum was concluded to be of the same order of magnitude as that from sporulation of naturally occurring basidiocarps and therefore did not significantly add to current risks of infection from natural sources of inoculum. Based on these results, the Plant Protection Service of the Netherlands concluded that the risk of use of C. purpureum as a biological control is acceptable when applied at least 500m away from fruit orchards (deJong et al., 1991).

To further assess the risk of infection to non-target plantations by basidiospores from a release site we have developed a spore-trapping technique and have initiated spore-trapping experiments at the field release sites. With the development of the RAPD and SCAR strain-specific markers it will be possible to measure the frequency of encountering the released isolate at various distances from the test site and to measure the extent of genetic introgression of the released isolate into the local population. Recovered C. purpureum will be genetically characterized and allele frequency data, collected from the field sites over a period of two years, will be used to determine guidelines for safe use of this fungus as a biocontrol for hardwood species.

ACKNOWLEDGMENTS

The research support of B.C. Hydro, the National Research Council IRAP program and the Science Council of British Columbia Technology B.C. program are gratefully acknowledged.

REFERENCES

Jobidon, R. (1991) Some future directions for biologically based vegetation control in forestry research. For. Chron. 67: 514-519.

Jong, M.D. de, Scheepens, P.C., and Zadoks, J.C. (1990) Risk analysis for biological control: A Dutch case study in biocontrol of Prunus serotina by the fungus Chondrostereum purpureum. Plant Disease, 74: 189-194.

Jong, M.D. de, Wagenmakers, P.S., and Goudriaan, J. (1991) Modelling the escape of Chondrostereum purpureum spores from a larch forest with biological control of Prunus serotina. Neth. J. Pl. Path., 97: 55-61.

McDermott, J.M., Brändle, U., Dutly, F., Haemmerli, U.A., Keller, S., Müller, K.E., and Wolfe, M.S. 1994. Genetic variation in powdery mildew of barley: Development of RAPD, SCAR and VNTR markers. Phytopathology, 84: 1316-1321.

Meidinger, D., and Pojar, J. (editors) 1991 Special Report Series 6: Ecosystems of British Columbia, B.C. Ministry of Forests.

Milgroom, M.G. and Lipari, S.E. (1995) Population differentiation in the chestnut blight fungus, Cryphonectria parasitica, in eastern North America. Phytopathol. 85: 155-160.

Paran, I. and Michelmore, R.W. (1993) Development of reliable PCR-based markers linked to downy mildew resistance genes in lettuce. Theor. Appl. Genet. 85: 985-993.

Ramsfield, T., Becker, E., Rathlef, S., Tang, Y., Vrain, T., Shamoun, S. and Hintz, W. (1996) Assessment of the Genetic Variability of Chondrostereum purpureum with restriction fragment length polymorphism in the ribosomal DNA. Can. J. Bot (in press).

Spiers, A.G. (1985) Factors affecting basidiospore release by Chondrostereum purpureum in New Zealand. Eur. J. For. Pathol. 15: 111-126.

Spiers, A.G. and Hopcroft D. H. (1988) Ultrastructural studies of basial and basidiospore development and basidiospore release in Chondrostereum purpureum. Eur. J. For. Pathol. 18: 367-381.

Templeton, G.E. and Greaves, M.P., (1984) Biological control of weeds with fungal pathogens. Tropical Pest Management 30(4): 333-338.

Templeton, G.E., and TeBeest, D.O. 1979. Biological weed control with mycoherbicides. Ann. Rev. Phytopathol. 17: 301-310.

Valentine, K.W.G., Sprout, P.N., Baker, T.W., and Lavkulich, L. M. (editors). 1978. The soild landscapes of British Columbia. B.C. Min. Environ., Victoria, B.C.

Wall, R.E. (1986) Pathogenicity of Chondrostereum purpureum to yellow birch. Plant Disease 70: 158-160.

Wall, R.E., Prasad, R., and Shamoun, S.F. 1992. The development and potential role of mycoherbicides for forestry. For. Chron. 68: 736-741.

Wall, R.E., Shamoun, S., and Prasad, R. 1995. Supporting data for registration of the fungus Chondrostereum purpureum as a biological control for hardwoods in forests and rights-of-way. Pacific Forestry Centre, Canadian Forest Service, Victoria, B.C. (unpublished).