Plant Protection Quarterly Vol.23(1) 2008 29 Chrysanthemoides seed fl ies Three Mesoclanis spp. have been consid- Progress on boneseed (Chrysanthemoides monilifera ered for biological control of boneseed subsp. monilifera (L.) Norlindh) biological control: and bitou bush in Australia (Edwards and Brown 1997; Neser and Morris 1985). These the boneseed leaf buckle mite Aceria (Keifer) sp., the fl ies lay their eggs into Chrysanthemoides fl owerheads and the larvae can destroy lacy-winged seed fl y Mesoclanis magnipalpis Bezzi substantial proportions of developing and the boneseed rust Endophyllum osteospermi ovules, thus suppressing seed production. When introduced to Australia in 1996, the (Doidge) A.R.Wood bitou seed fl y (Mesoclanis polana Munro) (BSF) rapidly colonized almost the entire A B T.B. Morley and L. Morin range of bitou bush (Edwards et al. 1999). A Department of Primary Industries, PO Box 48, Frankston, Victoria 3199, However, in South Africa the BSF does not Australia. utilize boneseed and prevails at latitudes B CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia. much closer to the equator than the more southerly bitou bush infestations in Aus- tralia. The BSF was therefore considered unlikely to effectively suppress seed pro- duction of Australian boneseed or the more Introduction stimulates dispersal. Dispersive mites can southerly bitou bush infestations, and this So far six exotic organisms have been re- walk to adjacent uncolonized shoot tips or appears to be true (Robin Adair personal leased in Australia as potential biological can be wind-dispersed to other boneseed communication and Morley unpublished control agents for the environmental weed plants. observations). Conversely, in South Africa boneseed (Chrysanthemoides monilifera sub- the lacy-winged seed fl y (LWSF) (Figure sp. monilifera (L.) Norlindh (Asteraceae)). Introduction to Australia 2) utilizes a range of C. monilifera subspe- None of these has established on boneseed Following host specifi city testing in South cies (Munro 1950, Edwards and Brown although one, the bitou tip moth (Comos- Africa in 2002 that showed the BLBM 1997) including boneseed, bitou bush tolopsis germana Prout (Lepidoptera: Ge- would be safe to introduce to Australia (most prosperously at latitudes similar to ometridae)) has colonized all of the major (Morley 2004), the mite was imported into southerly Australian bitou bush infesta- infestations of the closely-related invasive quarantine at the Department of Primary tions (Edwards and Brown 1997)), subsp. species bitou bush (C. monilifera subsp. ro- Industries – Frankston, Victoria in 2006 tundata (DC.) Norlindh). Currently three and was due for fi eld release in Spring organisms are under investigation for 2007. In collaboration with local land man- possible use as biological control agents agers, community groups and researchers, for boneseed. They are the boneseed leaf releases are planned for core boneseed in- buckle mite (Aceria (Keifer) sp. (Acari: Eri- festations in Victoria, South Australia and ophyidae)) (BLBM), the lacy-winged seed Tasmania and may be extended to New fl y (Mesoclanis magnipalpis Bezzi (Diptera: South Wales in the future. There are no Tephritidae)) (LWSF) and the rust fun- plans to release in Western Australia as gus (Endophyllum osteospermi (Doidge) very few boneseed infestations are present A.R.Wood) that induces witches’ brooms. and State law requires eradication of plants Figure 2. Lacy-winged seed fl y They are all endemic natural enemies of when found. Mesoclanis magnipalpis. boneseed in South Africa and each has the potential to suppress boneseed vigour and/or seed production.

Boneseed Leaf Buckle mite Characteristics The BLBM (Figure 1) is vermiform (worm- like), about 0.15 mm long and its preferred host is boneseed (Morley 2004). The BLBM induces formation of leaf galls known as erinea. Erinea are composed of abnormally dense patches of hair-like structures that resemble felt. Initially erinea are white but turn brown within a few weeks. Erinea may develop anywhere on a leaf, have ir- regular shapes and size and may range from one to several hundred square mil- limetres. Erineum formation is induced by BLBM feeding at shoot meristems and commences before a leaf becomes visible. Erineum growth interrupts the normal ex- pansion of affected leaves, resulting in leaf disfi gurements ranging from small dim- ples to gross distortion of mature leaves. Figure 1. Boneseed leaf buckle mite Aceria sp. (Courtesy Charnie Craemer, The BLBM feeds and breeds in erinea un- Plant Protection Research Institute, Pretoria, and Alan Hall, University of til overcrowding or erineum deterioration Pretoria, South Africa). 30 Plant Protection Quarterly Vol.23(1) 2008 pisifera (L.) Norlindh and Chrysanthemoides sites has yielded nearly 4000 BSF (Morley should it be released in Australia, it is not incana (Burm.f.) Norlindh. Thus the LWSF unpublished data)). The reason for this expected to have a signifi cant impact on seemed a good candidate biological con- outcome is not known but the following most Australian bitou bush populations. trol agent. speculations seem plausible: Climate modelling has shown that the rust 1) LWSF is present but in such low num- would likely have its greatest impact in Failure of LWSF to establish bers that it has not yet been detected. If Tasmania and southern Victoria (Wood et Boneseed this is the case then further sampling al. 2004). Attempts to establish the LWSF in Aus- should eventually detect it. tralia have been in progress for nearly a 2) There are distinct biotypes of LWSF in Infection decade. It was fi rst released on boneseed South Africa that have particular Chry- E. osteospermi systemically infects its hosts in 1998 and also in 1999, 2000 and 2005. santhemoides preferences and LWSF ex via immature foliage and stems. After These efforts involved LWSF reared from C.m. pisifera is a biotype not suited to 1–2 years, infected plants develop witch- imported C. monilifera subsp. pisifera fruit the bitou bush populations found in es’ broom branches with multiple swol- using a special AQIS-approved direct re- Australia. Intraspecific variation of len stems, short internodes and smaller lease protocol to ensure the released in- Mesoclanis has not been studied but and slightly chlorotic leaves (Wood 2002, dividuals were free of pests and diseases. based on the experience described Wood and Crous 2005a) (Figure 3). The Releases were made in the You Yangs, on above and an observation by Munro rust produces fruiting bodies on leaves of the Mornington Peninsula and at Frank- (1950), biotypism seems a real pos- witches’ brooms when conditions are con- ston, Victoria. The LWSF did not establish sibility. Munro (1950) observed that ducive, generally in winter to late spring but on three occasions offspring were re- fl ies of a collection of M. magnipalpis (Wood 2002). covered in the same season as the release ex C. incana were ‘larger than usual’. (Aline Bruzzese and Robin Adair personal Importation of LWSF ex C.m. rotundata Impacts communications, Morley unpublished at a latitude comparable to Australia’s The rust has been shown to have a sig- data). This confi rmed the apparent suit- southerly infestations (e.g. St Francis nifi cant impact on boneseed populations ability of boneseed as a host for LWSF ex Bay, South Africa, where large popula- in South Africa, reducing growth and C.m. pisifera. The reason for establishment tions of LWSF can be found (Edwards reproduction and in some instances kill- failure is not known but it appears that and Brown 1997)) would address the ing plants (Wood 2002, Wood and Crous LWSF ex C.m. pisifera is not able to survive biotype theory. This could perhaps im- 2005a). Deformed infected branches of between the end of boneseed’s fruiting prove the chance of establishment on diseased plants produced far fewer buds, season in early summer and the beginning bitou bush and is the next most obvious fl owers and fruits than branches of healthy of the next fl owering season in late win- strategy to try. plants and usually die within 1–4 years. ter. In South Africa, LWSF ex C.m. pisifera 3) Competition with the BSF has prevent- The rust also has an indirect effect on the might not be constrained in this way either ed or impeded LWSF estab- because the interval between C.m. pisifera lishment. If this is the rea- fruiting and fl owering is suffi ciently short son for failure of the LWSF or there are alternative Chrysanthemoides to establish on bitou bush, (a) hosts whose fl owerheads are suitable and then it might be overcome available at critical times. Importation and by releasing larger num- release of the LWSF ex C.m. monilifera has bers and/or using other been previously considered but not at- techniques to promote es- tempted largely because boneseed’s rela- tablishment (e.g. exclusion tively short fl owering season combined of BSF from bitou bush at with the logistic and mandatory delays release points). of passage through Australian quarantine make this option diffi cult. Populations of Boneseed rust LWSF on C.m. monilifera are also often dif- The South African rust fungus fi cult to fi nd in South Africa (Robin Adair E. osteospermi is a microcyclic personal communication). Notwithstand- species only recorded on a ing these diffi culties, in light of the failure small group of related plants of LWSF ex C.m. subsp. pisifera to establish of the genera Chrysanthemoides on boneseed in Australia it might now be and Osteospermum (Calend- (b) worthwhile investigating the introduction uleae: Asteraceae) in South of LWSF ex. C.m. monilifera in case it exists Africa (Doidge 1926, Morris as a biotype specifi cally adapted to toler- 1982, Wood 1998, Wood and ate boneseed’s fl owerhead phenology. Crous 2005b). It is thought to have considerable potential Bitou bush for the biological control of In 2005, the LWSF ex C.m. pisifera was also boneseed, and to a lesser ex- released on bitou bush on the New South tent bitou bush, in Australia Wales south coast. This was done with the (Morin 1997). Although the view that LWSF could be redistributed disease has been recorded on from these release sites to boneseed in- a few occasions on bitou bush festations in Victoria, South Australia and in South Africa, this was only Tasmania, in much larger numbers than at the southernmost part of it is practical to import, thereby perhaps the plant’s range (Wood and enhancing the possibility of establish- Crous 2005b). Consequently, Figure 3. Boneseed rust, Endophyllum ment on boneseed. However, the LWSF while the rust may attack bi- osteospermi; (a) fruiting bodies on leaves has not been recovered from bitou bush tou bush in the southern part and (b) witches’ brooms caused by the rust post-release (although sampling of release of its New South Wales range, infection. Plant Protection Quarterly Vol.23(1) 2008 31 growth and reproduction of non-infected Summary International Symposium on Biological branches of diseased plants. Field meas- BLBM was due for release in Australia in Control of Weeds, eds J.M. Cullen, D.T. urements showed that the production of Spring 2007. Further attempts to establish Briese, D.J. Kriticos, W.M. Lonsdale, fl owers and fruits on non-infected branch- LWSF on: 1) bitou bush in Australia with L. Morin and J.K. Scott, pp. 297-300. es decreased as the severity of the disease collections ex C.m. rotundata from St Fran- (CSIRO Entomology, Canberra, Aus- on plants increased (Wood 2002, Wood cis Bay, South Africa and 2) boneseed in tralia). and Crous 2005a). The systemic nature of Australia with collections ex C.m. monilif- Morris, M.J. (1982). A systemic rust fungus E. osteospermi is a desirable characteristic era in South Africa would be worthwhile. infecting Chrysanthemoides monilifera for biological control purposes, as once The boneseed rust has a signifi cant impact subsp. monilifera in South Africa. Phyto- the fungus is established within the host, on C.m. monilifera populations in South phylactica 14, 31-4. the infection is retained until the death of Africa. Substantial progress on risk as- Munro, H.K. (1950). Trypetid fl ies (Dip- infected branches (Morin 1997). sessment has been made and if approval tera) associated with the Calendulae, for release is granted then an importation plants of the family Compositae in Risk assessment strategy that accommodates the rust’s South Africa. I. A bio-taxonomic study Because of the nature of the rust, which unusually long life cycle will need to be of the genus Mesoclanis. Journal of the develops visible symptoms only 1–2 years negotiated with AQIS and Biosecurity Entomological Society of Southern Africa after infection of its host, an initial series Australia. 13, 37-52. of host-specifi city tests were performed on Neser, S. and Morris, M.J. (1985). Prelimi- detached leaves of test plant species. This Acknowledgments nary observations on natural enemies was done to determine, using microscopy Work on BLBM and LWSF is funded by of Chrysanthemoides monilifera in South techniques, whether the rust was capable the Australian, Victorian and Tasma- Africa. Bitou bush and boneseed: pro- of penetrating epidermal cells of non- nian Governments and was conducted ceedings of a Conference on Chrysan- target plant species (as it successfully does at the Department of Primary Industries, themoides monilifera, eds A. Love and on its host C. monilifera) (Wood 2006). Suc- Frankston, Victoria. Current work on the R. Dyson, pp. 105-109. Port Macquarie, cessful penetration was observed on bone- boneseed rust is funded by the Australian New South Wales. seed and its close relative species tested Government and CSIRO and is conducted Wood, A.R. (1998). Endophyllum osteosper- within the Calenduleae tribe, but also on at CSIRO Entomology, Canberra, ACT. mi, a new combination for Aecidium os- four other species outside the Calendule- teospermi (Basidiomycetes: Uredinales: ae. Additional tests were carried out on References Pucciniaceae). South African Journal of leaves still attached to plants of some of the Doidge, E.M. (1926). A preliminary study Botany 64, 146. non-target plant species as well as the of the South African rust fungi. Bothalia Wood, A.R. (2002). Infection of Chrysanthe- target weed species boneseed and bitou 2, 1-228. moides monilifera by the rust fungus En- bush, to confi rm accuracy of results ob- Edwards, P.B., and Brown, E.M. (1997). dophyllum osteospermi is associated with tained with detached leaves (Wood 2006). Mesoclanis seed fl ies (Diptera: Tephriti- a reduction in vegetative growth and Since penetration of epidermal cells does dae) on Chrysanthemoides (Asteraceae) reproduction. Australasian Plant Pathol- not necessarily imply that the infection in South Africa: distribution, attack ogy 31, 409-15. process will continue and be successful, strategy and impact on seed produc- Wood, A.R. (2006). Preliminary host specif- more tests on whole plants of the species tion. Bulletin of Entomological Research icity testing of Endophyllum osteospermi where penetration occurred in initial tests 87, 127-35. (Uredinales: Pucciniaceae), a biological are currently underway to determine if Edwards, P.B., Holtkamp, R.H. and Adair, control agent against Chrysanthemoides the fungus is capable of colonizing tis- R.J. (1999). Establishment and rapid monilifera ssp. monilifera. Biocontrol Sci- sue of these species. Results from these spread of the bitou seed fl y, Mesoclanis ence and Technology 16, 495-507. additional tests will provide the neces- polana Munro (Diptera: Tephritidae), in Wood, A.R. and Crous, P.W. (2005a). Epi- sary information to fully assess the risk eastern Australia. Australian Journal of demic increase of Endophyllum oste- of signifi cant impact on these non-target Entomology 38, 148-50. ospermi (Uredinales: Pucciniaceae) on species should the rust be released in Morin, L. (1997). The rust fungus Endo- Chrysanthemoides monilifera. Biocontrol Australia. phyllum osteospermi, a potential biologi- Science and Technology 15, 117-25. If permission is granted to release E. cal control agent for Chrysanthemoides Wood, A.R. and Crous, P.W. (2005b). Mor- osteospermi in Australia, an acceptable ap- monilifera in Australia. Proceedings phological and molecular characteriza- proach for its release from quarantine will of the Bitou Bush Workshop, eds R. tion of Endophyllum species on peren- have to be negotiated with AQIS and Bi- Holtkamp, R. Groves and S. Corey, pp. nial asteraceous plants in South Africa. osecurity Australia. Because of the rust’s 22-3, Adelaide. (NSW National Parks Mycological Research 109, 387-400. long generation time, it will not be possible and Wildlife Service and the Coopera- Wood, A.R., Crous, P.W. and Lennox, C.L. to maintain infected plants in quarantine tive Research Centre for Weed Manage- (2004). Predicting the distribution of until spores are produced and therefore an ment Systems). Endophyllum osteospermi (Uredinales: alternative strategy will be required (e.g. Morley, T.B. (2004). Host-specifi city testing Pucciniaceae) in Australia based on its surface decontamination of infected plants of the boneseed (Chrysanthemoides mon- climatic requirements and distribution before transfer from quarantine facility to ilifera ssp. monilifera) leaf buckle mite in South Africa. Australasian Plant Pa- ordinary glasshouse to provide optimal (Aceria neseri). Proceedings of the XI thology 33, 549-58. sporulation conditions; or direct release of spores collected in South Africa).