The Potential of Fungal Pathogens to Control Hypericum Species In
Total Page:16
File Type:pdf, Size:1020Kb
Plant Protection Quarterly Vol.12(2) 1997 81 necrotic areas, producing acervuli and of- ten perithecia – the sexual stage. The potential of fungal pathogens to control The C. gloeosporioides hyperici strain has Hypericum species in Australia been described as an ‘orphan’ myco- herbicide (Templeton 1992). It has been A A B demonstrated to be effective in the field D.A. McLaren , E. Bruzzese and I.G. Pascoe for specific weed control but has not been A Department of Natural Resources and Environment, Keith Turnbull Research developed for commercial use. A low level Institute and Co-operative Research Centre or Weed Management Systems, PO market of potential in comparison to that Box 48, Frankston, Victoria 3199, Australia. of broad-spectrum chemical herbicides is a B Institute for Horticultural Development, 621 Burwood Highway, Knoxfield, major reason for lack of commercial inter- Victoria 3180, Australia. est in this potential mycoherbicide (Templeton 1992). Shepherd (1995) tested two Canadian Abstract isolates of C. gloeosporioides against three Two fungal pathogens that attack either The Colletotrichum genus has been de- recognised Australian strains of H. St. John’s wort (Hypericum perforatum) scribed (Barnett and Hunter 1972). The perforatum (narrow-leaved, intermediate- or tutsan (Hypericum androsaemum) are genus is part of the Imperfect fungi and leaved and broad-leaved; see Campbell et discussed. The fungus, Colletotrichum has the following distinguishing features: al. 1997) and untyped plants collected from gloeosporioides, is host-specific and • Acervuli disc-shaped or cushion- various areas of Victoria, New South causes significant damage to H. shaped, waxy, subepidermal, typically Wales and Canada. Both C. gloeosporioides perforatum populations in Nova Scotia in with dark spines or setae at the edge or isolates killed all strains of Australian St. Canada. Its potential as a biological con- among the conidiospores, John’s wort but isolate S19 was more viru- trol agent in Australia is reviewed. • Conidiospores simple, elongate, lent than isolate DAOM. Both isolates were In October 1991, infestations of tutsan • Conidia hyaline, 1-celled, ovoid or ob- effective at killing field-collected St. John’s in the Otway Ranges of Victoria were long, wort (Figure 1). Virulence was influenced found to be significantly attacked by the • Parasitic; imperfect states of Glomerella. by dew period and inoculation concentra- fungus Melampsora hypericorum. The This genus differs from Glomeosporium in tion. It was concluded that if C. gloeo- rust had a devastating impact on the weed having spines, which may be absent under sporioides were introduced into Australia, population, killing entire hillsides. A certain cultural conditions. it could probably be used as a classical bio- subsequent inspection in 1994 found only Colletotrichum pathogens have been logical control agent. one live tutsan seedling. The tutsan rust successfully developed as mycoherbicides Preliminary host specificity-testing has fungus is possibly the most successful ex- (Templeton 1987, Wymore et al. 1988, shown three out of twenty six test species ample of weed biological control ever Makowski and Mortensen 1989, were susceptible to attack by C. gloeo- witnessed in Victoria. Templeton et al. 1989); some species have sporioides. Hypericum canadense (in the excellent qualities (host specific, consistent field), H. virginicum (in laboratory trials) Introduction weed control under field conditions, low- and tomato (Lycopersicon esculentum) St. John’s wort (Hypericum perforatum) cost of production, stable formulation (wound-inoculated in laboratory trials) (Clusiaceae) is a perennial herb that origi- with minimum shelf life of six months) for were attacked (Hildebrand and Jensen nated in Europe, western Asia and North weed control (Charudattan 1989, Jackson 1991). Shepherd (1995) tested a further ten Africa; in Australia it now occupies more et al. 1995). species and found no attack by C. gloeo- than 188 000 ha in New South Wales and K.I.N. Jensen (personal communica- sporioides. However, even with these 175 000 ha in Victoria (Parsons and tion) described how C. gloeosporioides at- promising results, more testing is required Cuthbertson 1992). It has been found in all tacks St. John’s wort. The pathogen gener- before this fungus could be considered as States of Australia except the Northern ally attacks young stems but also infects a likely biological control agent for ap- Territory where it invades poorly man- leaves and flowers. It causes stem girdling proval for entry into Australia. aged grazing land, sparse bushland, lesions that are characteristic of Colleto- roadsides and neglected areas such as trichum attack. Sporulating ascervuli com- Biological control of tutsan, abandoned mine sites (Parsons and monly cause secondary infection. Given Hypericum androsaemum, by the rust Cuthbertson 1992). St. John’s wort con- favourable conditions for sporulation, a fungus, Melampsora hypericorum tains hypericin, a compound that causes single infection is potentially lethal to St. Tutsan (Hypericum androsaemum) is an photosensitization in stock that ingest it. John’s wort. Infected leaves turn a reddish erect perennial shrub that originated in Symptoms develop on areas of skin di- colour, making infected plants easily iden- Europe, Asia Minor and North Africa (Par- rectly exposed to light, such as ears and tified. Seed from diseased plants produces sons and Cuthbertson 1992). Tutsan was noses. Affected areas then become itchy diseased seedlings, though there is evi- estimated to be infesting more than of and painful (Everist 1974). Animals be- dence that seed viability may outlast the 200 000 ha of Victoria in 1980 (Lane et al. come irritable, lose condition and in severe seedborne pathogen. There is also circum- 1980) and grew densely in the Otway, cases may develop convulsions and die stantial evidence that the disease can be Strzelecki and Dandenong Ranges. (Kingsbury 1964). transmitted by Chrysolina beetles feeding In October 1991, significant infestations on infected plants and carrying the muci- of tutsan in the Otway Ranges of Victoria Biological control of St. John’s wort, laginous spores to new St. John’s wort were found to be attacked by the rust fun- Hypericum perforatum, by the fungus, plants. In addition there are signs that the gus, Melampsora hypericorum (de Candolle) Colletotrichum gloeosporioides pathogen can survive as a weak parasite (Bruzzese and Pascoe 1992). The rust has In 1984, a fungal pathogen, Colletotrichum or saprophyte on plant species other than since been observed attacking tutsan at gloeosporioides (Penz.) Penz & Sacc., was St. John’s wort. Thus, symptomless carri- Myrtleford (E. Bruzzese personal observa- identified as causing substantial mortality ers such as dandelion and clover may har- tion) and near Morwell in the Strzelecki to St. John’s wort infestations in Nova bour dormant appressoria which become Ranges (D.A. McLaren personal observa- Scotia, Canada (Hildebrand and Jensen active only when leaves senesce after tion). It is not known how M. hypericorum 1991). which the fungus becomes active in the entered Australia. 82 Plant Protection Quarterly Vol.12(2) 1997 ab Figure 1. a) Control plant of St. John’s wort plant and b) St. John’s wort plant attacked by the fungus Colletotrichum gloeosporioides. At Apollo Bay in the Otway Ranges of Samples of M. hypericorum collected plant may be defoliated and killed. The Victoria, tutsan was a dominant weed on from tutsan have been placed on St. John’s impact of this rust was so great in Victoria, hillsides, outcompeting native vegetation wort but no sign of infection has been ob- it was as though a fire had swept through and invading pastures (M. Doueal per- served (E. Bruzzese personal communica- the undergrowth and selectively sonal communication). Since the rust was tion). Similarly, a plant nursery at Apollo browned, shrivelled, defoliated and killed first found, it has been having a devastat- Bay containing numerous native and ex- the tutsan plants (D.A. McLaren personal ing impact on tutsan populations, with otic species, including Hypericum observation). whole hillsides being killed. An inspection gramineum, reported no sign of attack by The Melampsora family forms a charac- conducted in the Apollo Bay area in 1993 M. hypericorum. This suggests that the teristic sub-epidermal crust of sessile, lat- by two Government officers with twelve tutsan rust fungus may be extremely spe- erally adherent, single celled teliospores local landholders located only a single cific (E. Bruzzese personal communica- near the surface of the infected host tutsan seedling after a full day’s search (M. tion). (Littlefield 1981). Melampsora hypericorum Doueal personal communication). The The symptoms of Melampsora was described by Plowright (1889): present situation in the Apollo Bay area is hypericorum and the fungus’s cycle of at- Uredospores – Sori orange, small, pul- that the tutsan populations continue to be tack on tutsan have been described by verulent, scattered, mostly hypophyllous. decimated with only the occasional seed- Baker (1955). The rust symptoms first ap- Spores globose or elliptical, finely echinu- ling being found. These tutsan seedlings pear in late spring or early summer. Yel- late, orange-yellow, 14–21 × 12–17 µ. Para- grow to only 10–15 cm before M. low to red irregular blotches appear on the physes absent. hypericorum infects and kills them (J. upper surface of tutsan leaves. Golden Teleutospores