70 Plant Protecti on Quarterly VoI.7(2) 1992 distribution. In these cases it was usually Southern African plants naturalized in Australia: a not possible to decide whether the plants in Australia originated from southern Af review of weed status and biological control potential rica or elsewhere. The species were separated into those John K. Scott, CSIRO Division of Entomology, Private Bag P.O., Wembley with distributions centred in sub-Saharan Africa, but which include southern Africa W.A. 6014, Australia. (T), those species widely distributed in E.S. Delfosse, CSIRO Division of Entomology, GPO Box 1700, Canberra ACT southern Africa but which also occur in 2601 , Australia. Present address: National Bio logical Co ntrol Institute, USDA, sub-Saharan Africa (A); and those species APHIS, Office o f the Administrator, Federal Building, Room 538, which ex tend into Madagascar (M). 6505 Belcrest Roadd, Hyattsville, MD 20782 USA. Within the southern African region we have also indica ted those plants restricted (endemic) to so uthern Africa (S) and Summary those plants restricted (endemic) to the Forty-three families, 141 genera and 251 plants have become weeds of importance. Cape Floristi c Region (C). This area most spedes of southern African plants have The need for the bi ological control of closely resembles the Mediterranean cli been recorded as established in Aus weeds of southern Africa n origin has been matic region of southern Australia and is tralia, mostly in southern regions. The recognized for some time (Stride 1960), defined by Bond and Goldblatt (1984) as families with the most species are the but only recently have detailed surveys the south western Cape Province of South Iridaceae (20% of species), Asteraceae started. As part of the planning of re Africa, from the Olifants River Mouth to (15% of species) and Poaceae (13"Yo of sea rch on the biological control of weeds Port Elizabeth, incorporating most of the species). Sixty-eight species are re it is necessary to know which species fynbos (heath) vegetation type. Species garded as weeds and 33 species have come from southern Africa, and which distributions within Africa were taken been serious enough to warrant research species are a serious problem for the agri from Bond and Goldblatt (1984), and expenditure as measured by publica culture and environment of Australia. taxonomi c reviews listed in Gibbs RusseU tions. Significant weeds include Given the number of introduced weeds in ef al. (1985, 1987). Aretothull calendula, C/"ysaut/remoides Australia and limited research funds it is The presence of southern African spe moftili/ern, Cucumis spp., Emex also necessary to set priorities as to which cies in Australia was noted from state and australis, Eragrostis curou/n, Homerin weeds should be subject to a biological regional fl oras, Flora of Australia (1981- spp., Lycium /erocissimllm, Oxalis pes control program. 1990) and taxonomic revisions. The fol caprae, Peuuisetum mncrourllm, In this review we list southern African lowing were the principal references con Romillea rosea, Senecio madagas plants in Australia and assess their status sulted: Australian Capital Territory carieusis, Urocl,lon mnximum and as weeds. We then examine the priorities (Burbidge and Gray 1976); central Aus Zm'tedesc/Jin aetlriopicn. Most of these for the biological control of the more im tralia (Jessop 1981); New South Wales plants are agricultural problems. Sig portant species. (Ja co bs and Pickard 1981 ; Jacobs and nificant weeds of conservation areas are Lapinpuro 1986); Northern Territory inadequately documented at present, Methods (Hnatiuk 1990); Queensland (Hnatiuk but include De/airen odoratn, E1fr/Jnrta 1990); South Australia (Jessop 1984; Kloot erecta, Meliuis repens, Myrsip/Jyllum Cel/SIIS of sOlltherll Africal/ plants 1986); Tasmania (Buchanan ef al. 1989); nsparagoides, Polygnla spp., The selection of species was limited. by the Victoria (Forbes and Ross 1988); and Protasparagus a/ricaulls, Seuecio following definitions. A naturalized plant WestemAustralia (Green 1985, 1987). The pteropltonls and Watsouia spp. is a species alien to Australia that has Flora of Australia trea tment was accepted The southern African flora estab spread beyond the possibility of eradica in most cases where the distributions in lished in Australia was assessed for bio ti on and has existed as a reproducing the various references were not in agree logical control priority. Of the weeds not population for at least 25 years (see Kloot ment. Species names and authorities fol currently the subject of biological con 1986 for more on this aspect). We have low the usage in Gibbs Russell ef al. (1985, trol research, Oxnlis pes-capra#! was taken this definiti on in a wide sense and 1987) with a few exceptions based on identified as having the highest priority. have included plants that may be taxonomic treatments since 1985. Recent The assessment of the biological and adventive or casual (K loot 1986) if it ap synonyms in the Australian literature economic importance of Arctotlreca pears from the literature that they might were also noted because not all taxonomic calendula and weeds of conservation re be naturalized in at least part of their re changes are universa lly accepted and be serves were identified as research areas corded distribution. Species were also in cause of differences between the interpre requiring attention. cluded in the list where it was not possible tation of species names between southern to determine the degree of establishment. Africa and Australia. Common names are Introduction Southern Africa n plants were defined based on Hartley (1 979). An important part of the introduced nora as those species indigenous to the region in Australia is of southern Africa n origin. south of the Kunene and Limpopo Rivers Weed status The accidental transfer of plants to Aus (Botswana, Lesotho, Namibia, South Af Plants were recorded as weeds if they tralia has been favoured by climatic simi rica and Swaziland). The species found in have been described as such in the litera larities and southern Africa being on the this region are listed by Gibbs Russell et ture. The principal sources used were immigration route of European colonists. al. (1985, 1987). We have limited the spe Auld and Medd (1987), Hyde-Wyatt and In addition, the horticultural attributes of cies under consideration to those that Morris (1980), Kleinschmidt and Johnson a number of southern African plant fami have distributions restricted to sub-Saha (1977), Kl oot (1986), Parsons (1973) and lies, especially the Asteraceae and ran Africa. This was done to exclude the Tothill ef al. (1982). Species were given lridaceae, has ensured the deliberate in large number of species indigenous to important weed status only if they were troduction of a number of species which southern Africa, but which are cosmo the topic of research publica tions specifi have subsequently escaped cultivation politan or include Europe and/or Asia or ca ll y concerned with their detrimental at and become naturalized. So me of these elsewhere as part of their natural tributes in A ustralia. In this case checklists Plant Protection Quarterly VoI.7(2) 1992 71 and weed floras were excluded. An esti versible control should the agents be suc- Results mate of the level of importance of a weed. cessful. For this reason it was necessary to was obtained by counting the number of determine which southern African spe SOllthern African plants in Australia references to the weed in the abstracting cies were economica lly useful. We have attempted to list all southern Af journal, Weed Abstracts, from 1976 to Host-specificity and hence safety of in rican vascular plant species naturalized in 1990 (C.A.B. International). Only refer troduction isof primary importance when Australia, including 43 plant families, 141 ences to weed science, including biologi selecting biological control agents (mainly genera and 251 species (Appendix I). The ca l control, in Australia were counted. insects and fungi) . In practice, potential most important families are the Irida ceae biological control agents are often found (20% of species), Asteraceae (15% of spe Biological control potentwl to feed on or infest species closely related cies) and Poaceae (13% of species). Two The existence of potential biological con to the target. This implies that a search for thirds of the families (30) are represented trol agents can only be established follow biological control agents will be less likely by three or less species. Most species are ing thorough surveys in the weed's region to find suitably host-specific organisms found in Western Australia (143) fol of origin. Consequently no weed should on a weed in its natural range if it has lowed by South Australia (139), New be excluded as a potential target for bio many related crop and/or native species South Wales (135), Victoria (1l3), Queens logical control. However, priOrities can be in Australia . In contrast, weeds with no or land (66), Tasmania (51), Northern Terri establi shed based on the likelihood of few related species in Australia w ill have tory (14) and the Australian Capital Terri (inding host-specific agents, and the more potential biological control agents tory (14) (Appendix I). The distribution knowledge of insects and fungi in the re available since the level of host-specificity of most species is restricted to southern gion of origin that might be suitable for may be less critica l. The number of native Africa, whereas 37 extend into the rest of use as biological control agents. It is also species and genera in Australia was de Africa, including six also found in Mada important to know if there might be any termmed from the references used to es gascar, while 10 have their distributions conflicts of interest that would militate tablish the census. centred in sub-Saharan Africa . Approxi against the use of permanent and non-re- mately a third (32%) of the species come
Table 1. Plants of southern African origin that have become weeds of importance in Australia.
Species States Main industries where Reference gi ving weed is important importance as a weed
WEEDS OF AGRICULTURE AND FORESTRY Arctotheca calendula NSW, V[CT, WA Pasture, cereal crops Scott & Way 1990a Cucilmis myrioenrpus QLD Pasture (toxic) McKenzie et al. 1988 Emex australis WA, SA, NSW Cereal crops, sheep pasture Gilbey & Weiss 1980 Gorteria personata W A Fleece contammant Dodd 1986, 1987 Hameria jlaccida SA, VICT, WA Pasture (toxic) Pearce 1979, Parsons 1973 Hamerin. miniafa SA, VICT, WA Pasture (toxic) Pearce 1979, Parsons 1973 MelimltJws cosmosils VICT Pasture and watercourses Parsons 1973 Mesembrymlthemum crystallimwI SA Pasture Kloot 1983 Mesembryn."tiremum 'ladjjloru", WA Pasture (toxic) Jacob & Peet1989 Oncosipiron pilliliferum WA Pasture Dodd 1990 01lcosiphon sliffruticOSll11l SA, WA Pasture Stephenson 1987, Dodd 1990 Oxalis glabra WA Pasture Peirce 1990a,b Oxalis pes-caprae SA, VrcT Cereal crops, pastu re (toxic) Michael 1964, Ca tt 1970 Oxnlis pllrpuren WA Pasture Peirce 1990a Romulen. rosen. VICT Pasture Eddy 1973 Seuecio mndagnscariellsis NSW Pasture (toxic) SindeI1986, Sindel & Michael 1988 Setaria sphncefntn QLD Pine plantations Costantini & Podberscek 1987 Solnnum lillltaem1ll1tl NSW, VICT, WA Pasture Parsons 1973, Rutherford 1979 Sporobolus africnnlls NSW Pasture Matthews & Dyason 1987 Trachynndra divn ricnta W A Pasture (toxic) Huxtable et al. 1987 Uroclllon mnximll1n QLD Sugar-cane and maize crops Crosthwaite 1983, Swain & Johnston 1984 Zmlfedeschia nethiopica W A Cattle pasture (toxic) Panetta 1988
WEEDS OF NATURE CONSERVATION AND AMENITY AREAS Chrysn.nthemoides m01li[ifera NSW, SA, VICT Nature conservation Weiss 1986 Delniren odoratn. VICT Nature conservation Fagg 1989 Ehrhnrta erecta VICT Nature conservation Mcintyre & Ladiges 1985 Erngrostis cllrvula NSW Road verges, pasture Campbell 1983 Lycium /erocissimll11l NSW, QLD, VICT Road verges, pasture Parsons 1973, Lee 1978 Melinis repens NSW, QLD Road verges Swain & Johnston 1984 Myrsipilyllu11I asparagojdes SA, VICT,WA Nature conservation, road verges Scott & Kleinjan 1991 Pennisetllm mncrollrll1n VICT, WA Road verges, pasture Parsons 1973, Dean 1987 Polygala myrtifolia SA, V[CT Nature conservation Carter et al . 1990 Polygala virgata NSW,SA,WA Na ture conservation Carter eI al. 1990 Protasparagus afrienlllls QLD Na ture conservation Conran & Forster 1986 Senecio pterophorlls SA Nature conservation, forests Moore et al. 1975 Watsmlia mer;nna VICT, WA Road verges Pa rsons 1973 72 Plant Protection Quarterly VoI.7(2) 1992 from the small area known as the Cape Table 2. Research activity in Australia as measured by the number of Floristic R.gion (Bond and Goldblatt abstracts in Weed Abstracts 1976 -1990 (C.A.B. International) for each 1984). weed (including synonyms) (species with weed status of lor 2 in Appen dix 1). No references were found for the species not listed. Weeds of SOli them Africa" origill Sixty-eight species (27%) have been listed Number of abstracts 1976-80 1981-85 1986-90 Total as weeds in general weed texts (Appen WEEDS OF AGRICULTURE dix I). Of these 33 have been regarded as Arctotheca calendula 10 3 11 24 seri ous enough to warrant expenditure, Cllcllmis spp. 0 3 1 4 as indicated by resea rch that resulted in Emex australis 17 17 7 41 publications relating to their biology and Hom eria spp. 2 0 2 importance as weeds in Australia (Table a Mesembryanthem lllfl crysialiilillm 0 1 0 1 1). The degree of research activity on these Mesembryallthemum llodij1orum 0 I I weeds can be seen in Table 2. Two thirds a Gxaiis pes-caprae 4 5 2 11 of the species in Table I appear in Weed Romulea rosea 2 1 0 3 Abstracts. We have divided the important Senecio madagascarieusis 1 3 4 weeds into two categories: those prima a Setaria sphacela ta 0 1 1 rily affecting agriculture and those prima a SporobolllS africmllls 0 3 3 rily affecting nature conservation. a TrachylZlldra divarica.ta 0 0 1 1 Importa"t weeds of agriculture Urochlon maximum 1 3 1 5 Zalltedeschia at!lhiopica. 0 ] 1 2 Twenty-two species have been recorded as weeds of agriculture (Table I). Most WEEDS OF CONSERVATION AN D AMENITY AREAS research activity since 1976 has concen Cltrysautl,emoides mOllilifera 1 5 4 10 trated on Emex allstralis and Arctotheca DeTairea odorata 0 0 I 1 calendll/a (Table 2) . These weeds are by fa r Ehrllllrta erecta 0] 0 1 the most important to agriculture of Eragrostis cUTVu la 1 2 4 7 plants of southern African o rigin. Both Lycium jerocissimum 3 1 0 4 cause measured yield losses in wheat Me/illis repens 0 1 0 1 crops (Poole and Gill 1987). Both are also Penlliset1l11l macrOllrllm 5 4 0 9 rated as weeds in sheep pastures, with A. ca /wdllia estimated to cost the industry though both are notable in W.A. and Vic weeds. Concern for weeds of conserva $32.8 million per year, and E. nlls/ralis $5.7 toria for their toxic properties and persist tion areas is very recent; consequently this million per year (Anon. 1988). Gilbey and ence in pasture due to abundant produc category of weeds is under-represented in Weiss (1980) estimated that E. allslralis ti on of corms (Parsons 1973, Pearce 1979). the weed literature, despite being recog causes losses to cereal crops and pastures The remaining agricultural weeds, listed. nized as one of Australia's greatest con in W .A. of more than $20 million per year. in Tables 1 and 2 are of localized signifi servation problems (Carr el nl. 1986). Another weed of southern African ori · ca nce. Thus, it is premature to establish from the gin with published estimates of costs is literature which species are important, Senecio madagascaritllsis. This plant was Importallt weeds of Ilature conservation apart from C. mouilifera . estimated to cost the dairy industry in areas N.S.W. $250,000 in 1985 (Sindel and Seven species have been reco rded. as Discussion Michael 1988). The comparatively recent weeds of nature conservation areas (Table identification and spread of this weed ex 1). The plant that has attracted most atten Approved targets for biological control plains the low number of publications in tion because of its invasiveness in nature Four weeds, E. australis, C. monill/era, M. Table 2. Fi eld (1984) lists the labour cost in conservation areas is Chrysanthemoides asparagoides and S. II1f1dngnscariensis, are Victoria for the control of declared nox lIlolliilfem (Weiss 1986) (Table 2) . Other approved target organisms for biological ious weeds of southern African origin im portant weeds, as judged by research control. during 1982 /83 as follows: Homerin activity, are Pell/lise!"", macrourUnl and flaccida and H. miniala, $44,945; Oxalis pes Emgrost is clITVllla (Table 2). These grasses Successful biological control in areas cnprne, $30,714, E. nllslralis $2,187 and spread primarily in road verges (impor outside Australia MeiianOws comoslIs $0, tant for conserving remnant vegetation Many biological control programs (e.g., Objective evaluation of the importance (Saunders and Hobbs 1991)) from where Oplllltia species) have been developed fol of weeds are difficult to make and have they are also a threat to pasture because lowing successful control reported else rarely been published . Oxalis pes-capme is they are unpalatable to stock (Campbell where. There has only been one non-Aus a case where a plant ha s been the subject 1983, Dean 1987). Lycillm ferocissimum is a tralian control program directed. against a of a moderate amount of research activi ty weed. of road verges in Victoria and like weed of southern African origin. This was (Table 2) and is judged to be an important wise is regarded as a threa t because it the successful contro l o f E. (lustralis in weed (Auld and Medd 1987). The plant is spreads into pasture (Lane 1979, Parson Hawaii by the southern African weevil, a weed of cereals of southemAustralia, in 1973). Field (1984) lists the labour cost in Perapioll allliqllll!/l (Gylienhal) (Julien particular South Australia, where it Victoria for the control of declared nox 1987) . causes reductions in crop yield and oxalic ious weeds of southern African origin acid poisoning in sheep (Michael 1964). during 1982/83 as follows: Senecio Weeds with known potential biological Catt (1970) estimated that O. pes-capme pleropilorlls, $0; C. !/Iollilifera, $55,545; P. co ntrol agents could reduce cereal yields by 20 to 50%, IIlflCTOllTlml, $27,475; E. curvula, $0; and L. Surveys for biological control agents have but quantified estimates are lacking ferocissil1WlIl, $274,062. Recently been completed for E. allstralis (Scott and (Poole and Gi ll 1987). Humphries et nl. (1992) included C. Way 1990b, Scott and Shivas 1990). Pa rtial There are few recent publications on moniTlfern and MyrsiphyllllTn aspnrngoides surveys show that potential biological Homeria and Romulea species (Table 2), al- among Australia's "top" environmental control agents are available for Plant Protection Quarterly VoI.7(2) 1992 73 Table 3. Ranking of weeds of sou thern African origin found in Australia Another group of possibly difficult tar· for priority for biological control research. get weeds are those plants that belong to genera containing many native species, Species lmpor- Conflict Closely Number of Biological again the reason being the unlikelihood of tancel of interest? related closely related control finding agents that are sufficiently host crops? native spp. priority specific. Examples are Solmwm WEEDS OF AGRICULTURE lhmaeanum and Senecio spp. (AppendiX 2). Arctotheca calendula 1 yes' no 0 I ' Cllcumis spp. 2 no yes 1 3 Priority weeds for biological control Emex australis 1 no no 8 1 Given the importance of va rious weeds, Garleria persona fa 3 no no 0 3 the possible conflicts of interest and de Hamerin spp. 2 no no 0 2 gree of taxonomic isolation, it is possible Melialllhlls C0511105 U5 3 no no 0 3 to rank the species in order of suitability MesembryanthemunI spp. 2 no no 0 2 for biological control (Table 3). Some of OncosiplJon spp. 2 no no 0 2 the species identified are already under Oxalis spp. I no no 9 I investigation. These include E. australis, RomuJea rosen 2 no no 0 2 C. monilifera, M. asparagoides and Senecio mndagascnriellsis I no no 46 2 Protasparaglls afriC/mus . Of highest prior· Setaria sphncelata 2 yes yes 5 3 ity among weeds not being investigated is Solanum !hmaeanum 3 no yes 94 3 O. pes·caprae (Table 3) . Sporoboills africal1l1s 2 no no 13 2 Trachyandra divaricata 3 no no 0 2 Con cl u sion s: priorities for research Urochloa maximum I yes yes 13 3 Of first priority is the completion of Zantedesc1tia aethiopicn 2 no no 0 2 projects already underway on Asparagaceae species, C. monillfera, E. WEEDS OF CONSERVATION australis and S. madagascariensis . The CllTystZuthemoides monilifera I no no 0 I weed with the highest priority for a new Delairea odoratn 2 no no 0 2 biological control project is O. pes-cnprae. Ehrharta spp. 2 yes yes 2 3 Biological control of this species could be Eragrostis cltrvula I yes yes 43 3 undertaken, pOSSibly without conflicts of Lyciuftl jerocissil1lum 2 no no I 2 interest, and at least one potential agent Melinis repens 3 no yes 0 3 has already been identified (Kluge and MyrsiphyllllTn tZsparagoides 2 no no 0 2 Claassens 1990). Pennisetum ml1crourum I no yes 3 2 Future developments towards the bio Polygala spp. 2 no no 19 2 logical control of weeds of southern Afri· Protnsparagus africallus 2 no no I 2 can origin will depend on increasing our Senecio pterophorus 3 no no 46 3 understanding of the importance of these Wntsonin spp. 2 no no 0 2 weeds in Australia. In conclusion, the fol
I Explanation of scores: lowing set of research objectives will help 1 = important weeds, based on the number of publications; towards such an understanding. . 2 = important weeds, which have been studied to a lesser extent; 1. Examination of confli cts of interest is 3 = weeds which have had little research activity or where studies have shown the sues. The most important weed. to be plant not to be a serious problem. examined is A. cnlendula. Here it is im 2 Depending on clarification of weed status. portant to establish the overalJ benefit of the plant in pasture and its impor A. calendula (Scott and Way 1989a), C. evident is the case of A. calendula. This tance to the honey industry. mOllilifera (Adair and Scott 1991 , Scott and plant is clea rly an important weed of 2. Assessment of priorities for weeds of Adair 1990), Myrsiphyllum and crops and pastures throughout southern conservation areas. Examples where Protasparagus spp. (Kleinjan and Scott Australia, but is also regarded as a useful further work is needed is on the 1990, Scott and Kleinjan 1991). Recently pasture plant and a pollen source for bees Asparagaceae species, Watsonia spp. Kluge and Claassens (1990) reported that (Clemson 1985, Scott and Way 1990a). and Polygala spp. the larvae of a noctuiid moth was host Further assessment of the weed status of 3. Predicti on of weed threat. The list of specific to a restricted range of Oxa/is spe this species is needed. Other weeds which southern African plants in Appendix 1 cies in laboratory tests. The moth has been may be of value to apiarists are L. is the first comprehensive listing of an found in the field in areas of similar cli ferocissi11lu11I and O. pes-caprne (Clemson exotic fl ora in Australia. Attributes of mate to parts of Australia with infesta 1985). Lycillm ferocissimllm is also used as these species could be examined to de tions of O. pes-capme. Surveys have also a hedge plant (Parsons 1973). velop a prediction of weediness. This been made of the insect fauna associated could lead to eradication of weeds with Berkheya rigida (Clarke 1988), S. Weeds where host-specificity may where this is feasible, e.g., Acacia karroo madagascariensis (Marohasy 1989), and preset!t a problem was identified during the preparation Solanum limmemwm (Okkers and Hulley It is likely that only a restricted set of bio· of Appendix I as a threat to the Austral· 1989, 1991). logical control agents (such as pathogens, ian environment and is now undergo eri ophyid mites) would be found that are ing eradication in the few places where Conflicts of interest sufficiently host-specific for weeds that it is established in Western Australia EJ"harta species, E. curvllla, Setaria belong to genera including cultivated spe (Scott 1991). Alternatively, biological splracelata and Urochloa maximum are ex cies (Appendix 2). Examples are Cucumis control can be attempted once a weed cluded as targets for classical biological myriocarplls, Ehrharta spp., E. cllrvlI la , P. has been identified as a threat, but is control since they are also economical1 y macrollrum, SOimlllTfl lilmaemtu11l and beyond eradication. useful plants in parts of their range. Less Sporobolus afric!IIlIls . 4. Surveys for biological control agents in 74 Plant Protection Quarterly VoI.7(2} 1992 southern Africa . Weeds whose poten 1990, pp. 116-20. ria, Melbourne.) tial biological control agents have been Catt, M.j. (1970) . The control of soursob Gibbs Russell, G.E., Reid, e., Van Rooy, j. identified become attractive proposi (Oxalis pes-caprae) by the use of herbi and Smook, L. (1985). List of species of tions for further research. Since south cides in South Australia . Australian southern African plants, 2nd ed., part 1. ern Africa is relatively unexplored (in Weeds Conference, November 1970, Memoirs of the Botanical Survey of South biologkal terms) it is not possible to ex Hobart,8(a}17-8(a}19. Africa No. 51, 1-152. tract this information from the IiterahIre Clarke, M.M. (1988). Insect herbivore Gibbs Russell, G.E., Weiman, W.G., (in contrast to Europe) and it is neces communities colonizing the fl ower Retief, E., lmmelman, K.L., sary to carry out such surveys in the heads of Berkheya in South Africa and Germishuizen, G., Pienaar, B.J., Van field. High priority for surveys should Carduoideae in Europe and California. Wyk, M. and Nicholas, A. (1987). List of be given to Homeria spp ., L./erocissj"lIlm Unpubli shed PhD Thesis, University of southern African plants, 2nd ed., part 2. and Roml/lea rosen since there is little Cape Town, 232 pp. Memoirs of the Botanical Survey of South knowledge of the associated fauna and Clemson, A. (1985). 'Honey and pollen Africa No. 56, 1-270. pathogens. fl ora'. 263 pp. (Inkata Press, Mel Gilbey, D.j . and Weiss, P.W. (1980). The bourne.) biology of Australian weeds 4. Em ex Acknowledgements Conran, j.G. and Forster, P.1. (1986). australis Steinh. Joumal of the Allstra1ian We thank the librarians of CSIRO for their Protasparaglls africmws (Asparagaceae) Institute ofAgn 'cu/turnl Science 46,221-8. help with the literature and j. Dodd, P.B. a serious weed for south-eastern Green, j.W. (1985) . 'Census of the vascu Edwards, RH. Groves, j.N. Matthiessen Queensland. Allstrobaiteya 2, 300-4. lar plants of Western Australia'. 312 pp. and F.D. Panetta for comments on earlier Costantini, A. and Podberscek, M. (1987). (Western Australian Herbarium, drafts. Control of Setaria spltncelata var. sericea Perth.) in afforestation o f pasture country with Green, j.W. (1987) . 'Census of the vascu References Pinus carilmea var. hOlldurellsis. Plaut lar plants o f Western Australia, cumu Adair, Rj. and Scott, j.K. (1991). Distribu Protection Quarterly 2,112-4. lative supplement No.5'. 23 pp. (West tion, life history, host specificity and Crosthwaite, I.e. (1983). Maize growing ern Australian Herbarium, Perth.) suitability of an undescribed Chrysolilla on the Atherton Tableland. Queellslmld Hartley, W. (1979) . 'A checklist of eco species (Coleoptera: Chrysomelidae) AgriCllItural/oIlTllaI1983,40-6. flomic plants in Australia'. 214 pp. for the biological control of Cullen, j.M. and Delfosse, E.S. (1985). (CS1RO, Melbourne.) Ch,ysanthemoides moni!ifera (Com Ecll;u", plantagineum: catalyst for con Henderson, M. and Anderson, j.G. (1966). positae). Bulletin of EntomologiCli I Re flict and change in Australia . Proceed Common weeds in South Africa. Botani sea rch 81, 235-42. ings of the Vl International Symposium cal 5l1rv