Plant Protection Quarterly Vol.8(4) 1993 159 disturbance. Weed taxa that are closely al- lied to economically important plants or Overview and use of biological control in pasture indigenous plants e.g., Poaceae should not be excluded from consideration due situations to the high level of host specificity dem- onstrated by some organisms. Australian D.A. McLaren, Department of Conservation and Natural Resources, Keith native plants that are weeds outside their Turnbull Research Institute, PO Box 48, Frankston, Victoria 3199, Australia. natural range may also make suitable tar- gets for biological control, particularly where substantial geographical barriers Summary It has been estimated that ragwort exist between their introduced and native A number of successful biological con- populations have been reduced by range. The protocols for conducting such trol programs against pasture weeds are 60–70% through the combined actions of programs are not developed, but require discussed and related to the state of the the cinnabar moth, Tyria jacobaeae (L.), the attention before they should be under- programs in Australia. Factors that af- ragwort flea beetle, Longitarsus jacobaeae taken. fect the establishment of biological con- (Waterhouse) and the ragwort seed fly, The ultimate criteria for the successful trol agents and how biological control Botanophila seneciella (Meade) (Brown control of environmental weeds, regard- can be integrated into pasture manage- 1990). less of the techniques used, should be ment programs are discussed. Biological control of blackberry, Rubus measured as either the level of replace- constrictus Lef & M., has been successful ment of the target infestations with other How successful have biological in Chile where the rust fungus, Phrag- vegetation, or the level of protection pro- control programs been against midium violaceum (Shulz), has signifi- vided to uninfested vegetation by a reduc- pasture weeds? cantly reduced infestations (Oehrens and tion in the rate of spread. Control pro- There have been a number of examples of Gonzalez 1977). The rust fungus builds grams that fail to reduce the fitness of successful biological control of pasture up its innoculum over summer and heavy weed infestations to levels below the weed species around the world. St. John’s rust infestations defoliates blackberry ‘critical ecological threshold’ that allows wort, Hypericum perforatum L., occupied plants. This allows light to penetrate natural regeneration, or significantly re- more than 1 424 300 hectares in North through the blackberry thicket, enabling duce dispersal rates cannot be regarded as America (Tisdale 1976). Biological control seed from competing plant species to ger- successful. Where the target weed is re- was started in 1946 with the introduction minate and start growing up through the placed by undesirable vegetation with an of two chrysomelid beetles, Chrysolina blackberry. This gradually takes nutrients equal or greater weed status, limited suc- quadrigemina (Suffrian) and Chrysolina and light away from the blackberry until cess may be claimed but the contribution hyperici (Forster). Significant reductions it is displaced by other plant species. of the control program to the protection were observed by 1954 and St. John’s Musk thistle, Carduus nutans L. has or enhancement of biological conserva- wort now occupies less than 3% of its been successfully controlled by the wee- tion must be rated as negligible. In this re- original density over much of its former vil Rhinocyllus conicus Froel. in Virginia. spect, biological control programs for distribution in California. The beetles The weevil has reduced musk thistle den- environmental weeds should be incorpo- only feed on St. John’s wort in sunny situ- sity by 95% and six years after release the rated into integrated management plans ations, so its distribution has now been weevil had spread 32 km (Kok and Surles that aim to manipulate post-control suc- largely restricted to shaded locations. Bio- 1975). cession from weedy to native vegetation. logical control programs against St. Evaluation processes that record changes John’s wort have been most successful in Biological control of pasture weeds in vegetation composition and structure areas with summer droughts, where de- in Victoria and the reproductive fitness of the target foliated plants succumb to water stress Table 1 provides a summary of the cur- weed as well as the more traditional and die. It is still a problem in localities rent status of biological control projects measures of plant density, biomass and with summer rainfall such as British Co- against pasture weeds in Victoria. area of distribution are required to deter- lumbia in Canada, as the plant will regen- St. John’s wort has been partially con- mine success of environmental weed con- erate and recover after defoliation and its trolled by biological control in Victoria, trol programs. seed will germinate more readily especially in agricultural areas experienc- (Williams 1985). ing Mediterranean conditions. The best An excellent example of biological con- results have occurred when farmers have trol in Australia is the spectacular success improved their pasture management of the moth, Cactoblastis cactorum (Berg), practices by sowing strong competing in controlling prickly pear cactus, Opuntia pasture and by keeping it well fertilized stricta Haw. Prickly pear once occupied (Parsons 1957). It appears that the combi- more than 20 million hectares of pastoral nation of stress imposed by the biological and agricultural land in south west to control agents and the competition from central Queensland and was invading the improved pasture, displaces the new country at the rate of two hectares weed. St. John’s wort is still a serious every minute (Waterhouse 1978). Fifty weed in many areas of Victoria and re- one biological control agents were intro- search is currently being carried out on duced over a 22 year period (1913–1935) the host specificity of a fungus, Colleto- of which seventeen were released and 12 trichum gloeosporioides, and introduction established. C. cactorum was introduced of a mite, Aculus hyperici, for its control. in 1925 and had controlled most of the There are three insect species estab- prickly pear infestations by 1932 (Wilson lished on ragwort in Victoria. These are 1960). the ragwort leaf and crown boring moth, Another pasture weed, ragwort, Senecio Cochylis atricapitana (Stephens), and the jacobaea L., has been biologically control- ragwort flea beetles, Longitarsus flavi- led in Western Oregon in North America. cornis Ste. and L. jacobaeae. Thus far, the 160 Plant Protection Quarterly Vol.8(4) 1993 flea beetles have only established in high What are our chances of success The ragwort flea beetle, L. flavicornis is altitude, high rainfall locations. Where and what factors affect the likeli- another insect species that appears to be the flea beetles have established, they hood of successfully controlling a limited in its distribution by climate and have reduced ragwort infestations but pasture weed? insect strain. So far, L. flavicornis has only have spread slowly and taken a long time The overall figure on worldwide success established in high altitude, high rainfall to control the weed. The ragwort leaf and in biological control programs on weeds sites. Attempts were made to bring in sev- crown boring moth is building up at es- is 24% (Julien 1989). In Australia, it has eral other strains of this flea beetle from a tablished sites and is expected to start been estimated that biological control has range of climatic localities in Europe. having an impact on ragwort infestations saved in excess of $1 billion dollars However, this species has still only estab- (McLaren 1992). (Harley and Forno 1992). Factors that will lished in higher altitude/rainfall locali- An illegal introduction of a strain of the determine the likelihood of success are: ties. Another flea beetle species L. blackberry rust fungus, P. violaceum, was jacobaeae was introduced in 1988 and has found in Victoria in 1984 (Marks et al. Climate, host range, agent strain established at some lower altitude sites. 1984). This rust only attacks four of the Biological control agents have evolved Another way climate and insect strain nine blackberry species in Victoria but has with their host in their country of origin may impact on effectiveness of biological been shown to have a significant impact for many thousands of years. As a conse- control is shown by impact of the ragwort on these species (E. Bruzzese personal quence, many insects have become spe- seed fly, Botanophila jacobaeae, on ragwort communication). A strain of P. violaceum cifically adapted to particular sets of cli- infestations in New Zealand. There is that attacks the most widespread black- matic conditions and in many cases poor synchronization of emergence of the berry species (Rubus procerus) was re- formed strains that will only survive in seed fly and the flowering period of rag- leased in 1990 and observations suggest particular environmental locations. Out- wort, as it emerges approximately six that this could have a considerable impact side this climatic range, the species strain weeks before the peak flowering period, on blackberry populations in the next may not reproduce or complete their life resulting in 80–90% of ragwort seeds es- decade. cycle. It is therefore important that caping predation (Dymock 1987). The Biological control of Paterson’s curse searches are made for biological control milder climate in New Zealand enables re-commenced in 1988 with the release of agents in areas that have similar climatic ragwort to flower over a prolonged pe- the Paterson’s curse leaf miner, Dialectica conditions to those that the insect species riod. It may be possible to overcome this scalariella (Zeller). This insect is now es- will encounter when released on plant in- problem by importing seed flies from lo- tablished in the warmer parts of the state festations in Australia. cations in Europe that have similar cli- but has yet to have an impact on Climate may impact on establishment mates to those in New Zealand. Paterson’s curse infestations. Several of biological control agents in a number other biological control agents are cur- of ways. It has been shown for St. John’s The genetic diversity of the weed rently being reared for release or are wort that locations that have summer The way a weed reproduces may be an being tested for host specificity. These “droughts” are more likely to enable suc- important factor in determining the likeli- include a root boring weevil, Mogulones cessful biological control than locations hood of controlling the weed. It would be geographicus (Goeze), a crown boring with cooler, moister conditions that allow expected that a genetically uniform weed weevil, M. larvatus Shultze and two sap plants to recover and for seed to germi- species would be easier to control than a sucking bugs, Dictyla echii Schrank and D. nate. In areas of the world where St. species that exhibits wider genetic diver- nassata Puton. John’s wort has become a serious weed, it sity (Burdon et al. 1980). For example, Strains of the thistle receptacle weevil has only been well controlled in locations prickly pear cactus, Opuntia spp., repro- Rhinocyllus conicus are being mass reared where there is a true Mediterranean cli- duces asexually and clonally and when and released to control spear, slender and mate (i.e., California in North America introduced into Australia, it essentially variegated thistle. This insect has been and parts of north east Victoria or south- formed a huge monoculture of genetically very difficult to rear and has yet to estab- ern NSW in Australia.) uniform structure. Once an agent was lish in the field. A fly, Urophora stylata (F.) Another good example of how climate found that would feed on and destroy that galls the heads of spear thistle is and host range can affect a biological con- prickly pear (the moth, C. cactorum), the currently being reared, but has yet to be trol agent is the Paterson’s curse leaf plant was controlled in a relatively short released. miner, D. scalariella. In its countries of ori- time. The limiting factor of control being A biological control project on hore- gin, D. scalariella is oligophagous and will the temperature requirements of the in- hound began in 1991 with the introduc- survive on other Boraginaceae species sect (Wilson 1960). tion of a plume moth, Pterophorus when Paterson’s curse is not available Weed species that reproduce sexually spilodactylus (Curtis) whose larvae defoli- (Wapshere 1985). As these host plants do (i.e., ragwort, Paterson’s curse, thistles) ate horehound. It has undergone host not occur in Australia, D. scalariella is may have a greater genetic diversity specificity testing and it is anticipated that restricted to climatic zones or locations which gives these plant species a greater the first release of this moth will occur late where some Paterson’s curse plants are chance of survival from insect attack than in 1993. actively growing all year round. species that primarily reproduce vege- tatively (prickly pear cactus, St. John’s Table 1. Biological control of pasture weeds in Victoria wort). They have a greater probability of Common name Scientific name Date of Success adapting through natural selection and 1st release producing strains that may be more re- sistant to insect attack in the future. St. John’s wort Hypericum perforatum 1930 Partial Another example of how genetic diver- Ragwort Senecio jacobaea 1930 Partial sity of a weed can affect the success of a Blackberry Rubus fruiticosus 1984 Partial biological control program, is the biologi- Paterson’s curse Echium plantagineum 1988 Slight cal control of skeleton weed Chondrilla Spear thistle Cirsium vulgare 1991 Not established juncea L., using the rust fungus Puccinia Slender thistle Carduus pycnocephalus 1993 Not established chondrillina Bubak & Syd. Skeleton weed Variegated thistle Silybum marianum 1988-89 Not established is a widespread weed in wheat/fallow Horehound Murrubium vulgare Not released cultivations of south-eastern Australia. Plant Protection Quarterly Vol.8(4) 1993 161 In Australia, there are three main forms of vegetative plant stages in autumn, winter London 244, 462-4. skeleton weed which have been termed and spring. Attack by the flea beetle re- Dymock, J.J. (1987). Population changes narrow leaf, intermediate leaf and broad duced the capacity of ragwort to recover of the seedfly, Pegohylemyia jacobaeae leaf (Hull and Groves 1973). A rust fun- from attack by the cinnabar moth. (Diptera: Anthomyiidae) introduced gus P. chondrillina, that was virulent for biological control of ragwort. New against the most common “narrow leaf” Pasture management and integrated Zealand J. Zool. 14, 337-42. form of skeleton weed was introduced control Harley, K.L.S. and Forno, I.W. (1992). and released in Australia in 1971 (Hasan Farmers should not rely on biological ‘Biological Control of Weeds, a hand- 1972). It was extremely effective and sig- control as a total cure for weed problems, book for practitioners and students’. nificantly reduced skeleton weed infesta- as biological control cannot achieve weed (Inkata Press, Melbourne). tions (Cullen et al. 1973). However, the eradication. Approximately 75% of bio- Hasan, S. (1972). Specificity and host spe- narrow leaf form of skeleton weed was logical control agents released, fail to con- cialization of Puccinia chondrillina. Ann. partially replaced by the intermediate and trol their target. If biological control does Appl. Biol. 72, 257-63. wide leaf forms which has necessitated work and the farmer does not improve Hasan, S. (1981). A new strain of rust fun- the search for rusts that attack these skel- his pasture, then the weed controlled gus Puccinia chondrillina for biological eton weed forms (Hasan 1981). through biological control may well be control of skeleton weed in Australia. replaced by another weed species. Ann. Appl. Biol. 99, 119-24. Predators, parasites and disease Biological control should be viewed as Hull, V.J. and Groves, R.H. (1972). Varia- Biological control agents are cultured and part of an integrated approach to control- tion in Chondrilla juncea L. in south-east- screened for at least one generation in ling weeds in pasture situations. To help ern Australia. Aust. J. Bot. 21, 113-35. quarantine before they are released, to aid competition from desirable plant spe- Julien, M.H. (1989). Biological control of ensure that parasites and/or diseases are cies, the pasture should be improved weeds worldwide: trends, rates of suc- not introduced with them. However, through application of fertilizer, sowing cess and the future. Bio. News. Info. predators, parasites and/or diseases have new grasses or adjusting stocking rates. 10(4), 299-306. still played significant roles in preventing Farmers must rehabilitate their pasture Kok, L.T. and Surles, W.W. (1975). Suc- agent establishment. Repeated attempts while biological control is taking place or cessful Biocontrol of Musk Thistle by an at establishing the cinnabar moth for bio- they may only be removing one weed introduced weevil, Rhinocyllus conicus. logical control of ragwort from 1930 to species and replacing it with another. Envir. Ent. 4(6), 1025-7. 1981 were prevented initially by disease Biological control should not be used as Marks, G.C., Pascoe, I.G and Bruzzese, E. and then by the action of a number of na- an excuse by farmers not to do weed con- (1984). First record of Phragmidium tive insects, the most damaging being the trol on the rest of their properties. If bio- violaceum on blackberry in Victoria. scorpion fly, Harpobittacus nigriceps (Selis) logical control agents are released at a Aust. Plant. Path. 13(1), 12-13. (Bornemissza 1966). Attempts were made particular location on a property, then McEvoy, P.B., Cox, C.S., James, R.R., and at protecting field cultures in cages. This they should allow a 50 metre buffer area Rudd, N.T. (1989). Ecological mecha- was successful but once cages were re- around the release site and control weeds nisms underlying successful biological moved, the cinnabar moth larvae were on the rest of the property. weed control: Field experiments with heavily predated upon and failed to es- Improving the competitiveness of de- ragwort Senecio jacobaea. Proceedings tablish (Bornemissza 1966). sirable pasture species will take nutrients V11 International Symposium Biologi- Other important predators include and resources away from weeds being cal Control of Weeds, 6–11 March birds, spiders, ants, and centipedes. To try damaged and weakened by biological 1988, Rome, Italy. Ed. E.S. Delfosse. pp. and overcome some of these problems, control agents. This “pasture improve- 55-66. increasing effort is being placed into do- ment” will aid in the displacement of the McLaren, D.A. (1992). Observations on ing caged releases. This helps protect the weed and will also help prevent re-infes- the life cycle and establishment of insects from predators while also keeping tation by preventing germination of weed Cochylis atricapitana (Lep:Cochylidae), a the insects together so they can mate, lay seed or from the weed being replaced by moth used for biological control of eggs and establish. other undesirable weed species. Senecio jacobaea in Australia. Ento- mophaga 37(4), 641-8. Increasing stress on the weed by intro- References Oehrens, E.B. and Gonzalez, S.M. (1977). ducing a variety of biological control Bornemissza, G.F. (1966). An attempt to Dispersion, ciclo biologico y danos agents that attack different parts of the control ragwort in Australia with the causados par Phragmidium violaceum weed at different times of the year Cinnabar moth, Callimorpha jacobaeae (Shultz) Winter en zarzamora (Rubus There is a deliberate strategy in most (L.) (Arctiidae: Lepidoptera). Aust. J. constrictus ef. et y R. ulmifolius Schott) weed biological control projects to intro- Zool. 14, 201-43. en la zonas centro-sur y sur de Chile. duce a range of biological control agents Brown, R.E. (1989). Biological control of Agro. Sur 5, 73-85. that will attack different stages of the life Tansy ragwort (Senecio jacobaea) in Parsons, W.T. (1957). St. John’s Wort in cycle of the target weed at different times Western Oregon, USA, 1975–87. Pro- Victoria. History, distribution, control. of the year. The aim is to accumulate ceedings V11 International Symposium J. Agric. Vic. Dec. pp. 781-8. stress on a weed species so it is unable to Biological Control of Weeds. 6–11 Tahvanainen, J.O. and Root, R.B. (1972). recover from agent attack and enable March 1988, Rome, Italy. Ed. E.S. The influence of vegetational diversity more desirable plant species to out com- Delfosse. pp. 299-305. on the population ecology of a special- pete the weed and replace it. McEvoy et Burdon, J.J., Marshall, D.R. and Groves, ized herbivore. Oecologia (Berl) 10, 321- al. (1989) have shown that the combined R.H. (1980). Aspects of weed biology 46. action of the cinnabar moth, T. jacobaeae, important to biological control. Pro- Tisdale, E.W. (1976). Vegetational re- and the ragwort flea beetle, L. jacobaeae, ceedings V International Symposium sponses following biological control of led to faster and more effective ragwort Biological Control of Weeds. Brisbane, Hypericum perforatum in Idaho. North- control than when either agent attacked Australia, Ed. E.S. Delfosse. pp. 21-9. west Science 50(2), 61-75. ragwort in isolation. The cinnabar moth Cullen, J.M., Kable, P.F. and Catt, M. Wapshere, A.J. (1985). The phyto- attacks the reproductive plant stages in (1973). Epidemic spread of a rust im- phagous guild on Echium in the western summer while the flea beetle attacks the ported for biological control. Nature, Mediterranean and the biological 162 Plant Protection Quarterly Vol.8(4) 1993 control of the weed, Echium plantagineum. Acta. Ecol. Appl. 6(4), 331- 8. Summary and conclusion Waterhouse, D.F. (1978). Biological con- trol. CSIRO Information Service. Sheet Robert V. Edgar, Department of Conservation and Natural Resources, Keith No.1-14, 1-6. Turnbull Research Institute, PO Box 48, Frankston, Victoria 3199, Australia. Williams, K.S. (1985). Climatic influences on weeds and their herbivores: Biologi- cal control of St. John’s Wort in British Columbia. Proceedings V1 Interna- The speakers at this workshop gave an ex- This workshop provides policy makers tional Symposium Biological Control of cellent overview of biological control pro- and managers at both the State and Na- Weeds. 19–25 August 1984, Vancouver, grams in Victoria and the relationship to tional level with information to review ex- Canada. Ed. E.S. Delfosse. pp. 127-32. the National scene. isting and proposed biological programs. Wilson, F. (1960). A review of the biologi- The following major points were made: Programs must have a stronger focus on cal control of insects and weeds in Aus- • Biological control programs may re- producing effective outcomes with tangi- tralia and New Guinea. Technical Com- quire 10–20 years to achieve an effec- ble results in the field. munication No. 1. Commonwealth In- tive outcome. The proceedings will also serve as a stitute of Biological Control, Ottawa, • Strict protocols are in place to ensure valuable tool both for promoting biologi- Canada. pp. 51-6. the scientific integrity of the programs, cal control and in increasing the level of to protect natural and modified ecosys- understanding in the government, indus- tems and to provide a mechanism for try and the community. This will help the the resolution of conflict of interest development of stronger links between where it arises in the community. workers in research and in the field, and • The major benefit of successful biologi- provide for the integration of biological cal control will be at the community control into regional weed control strate- level although some benefits will flow to gies and programs. individuals. With the current level of resourcing • Both government and industry or- there is a need for a reduction, or at least ganizations have provided significant a consolidation, in the number of target funding over many years, but there is species and agents under study. It would currently serious questioning of both the be preferable to adopt a “best bet” ap- time span and level at which this fund- proach with a view to achieving a limited ing will continue to be provided. number of recognized successes rather • A large number of control agents are than maintaining a larger number of pos- being tested against a range of pest sibilities of success without adequate re- species. Are resources spread too sources to complete the required work. thinly? Professional publicity should be given • The level of resources that are used to to programs where success can be dem- study plant and agent interactions onstrated. This publicity should not raise should be reviewed with a view to in- unrealistic expectations, but rather clearly creasing the availability of resources for spell out the level of success and the limi- mass rearing insects for field releases. tations. Publicity of this nature will help to • Community expectations on the poten- sustain community, industry and political tial value of biological control vary, and support to help the continued flow of the “success” to the scientist may not be necessary long term funding. Effective in- seen in the same light by farmers or tegration of the community into rearing others in the community. and release programs will also build and • In many bushland situations biological maintain support for programs. control appears to offer the only chance Although there is excellent communica- of returning the vegetation to near natu- tion between research workers in State ral condition. and Commonwealth agencies, the impact • There is an exciting opportunity to link of programs could be enhanced through mass rearing and release programs to the development of a stronger National schools and community groups. This focus. The Australian Weeds Committee would give the benefits of increased and the co-ordinator, recently appointed availability of insects coupled with a jointly by the Wool and Meat Research better community understanding of the Industry Councils, are two avenues which place of biological control in weed con- could help in the development of this fo- trol. cus. • To be effective biological control pro- Finally, congratulations are due to both grams must be linked to regional weed the members of the Weed Science Society control strategies and not be seen as an of Victoria and the staff of the Depart- add-on that may have some benefit. ment of Conservation and Natural Re- • Although there is good communication sources for their work in organizing and and co-ordination between State and presenting this important workshop. Commonwealth research workers, the Their efforts are well appreciated and I impact of biological control would be look forward to the continued co-opera- enhanced through the development of tion between the Department and the So- a better National focus. ciety in the coming years.