Biological Control of Grass Weeds in Australia

62 Plant Protection Quarterly VOI.5(2) 1990

son 1972, Savile 1979, 1987). For these rea Biological control of grass weeds in Australia: sons the tribe Stipeae is placed as belonging an appraisal to both Pooideae and Arundinoideae in Table 1 and this combined relation would have to be taken into account when consider A.J. Wapshere, CSIRO Division of Entomology, GPO Box 1700, Canberra, ACT ing biological control of grasses in this tribe. 2601, Australia. The weeds Hordeum spp. and Agropyron spp. are the only ones related tribally or more Summary closely to the important crop grasses, Hor Two biological control methods are consid iii) if not and potential agents occur in the deum spp. (barley), Triticum spp. (wheat) ered for the more than 250 grasses regarded grass weed's native range, whether they and Secale cereale L (rye) (Table 1). A as weeds in Australia, 170 of which, includ have sufficient host restriction to be in group of andropogonid weeds are related ing the most important, are of exotic origin. troduced into Australia. tribally to Zea mays L (maize), Sorghum bi 1. Inundative or bioherbicidal control color (L) Moench (sorghum) and Sac· where agents already present in Australia, Introduction charum officinarum L (sugar cane). Weeds probably fungi, would be used like herbi The possibility of biologically controlling in the genera Aira. Aven~ An-henalherum. cides. Mostly it is grass weeds important in grasses which have become weeds in Austra Holcus, Lophochloa, Molineriella are sub crops which would repay such an approach lia has been discounted by commentators on tribally or more closely related to the impor but only a few of these have 8 range orfuogi the subject for some considerable time. This tant crop grass Avena sativa L (oats) and already infesting them in Australia. has been due mainly to the close relation of many other weedy grass genera are tribally 2. Classical or inoculative control, where these weeds to cereal crops and pasture related to that crop (Table 1). Weeds in the agents from the home range of the grass grasses. This paper discusses the biological genera Avena, Echinochloa, Hordeum, would be introduced into Australia. Only a control of grass weeds in Australia in the light Oryza, Sorghum, Pennisetum, Panicum and few types of agents, notably gall makers and of recent developments in the SUbject and Setaria are closely related generically to the smut fungi, have sufficient specificity to be indicates which grass weeds would be most important crop grasses Avena saliva (oats), considered for introduction and then only if susceptible to the different methods of bio Echinochloa spp. (Siberian and Japanese the grass weeds are not related generically logical control. millets), Hordewn vulgare L (barley), Oryza to crop or pasture/lawn grasses. sativa L (rice), Sorghum bicolor (sorghum), The taxonomic relations of the grass The Australian Grass Weeds Pennisetum glaucum (L) R. Br. (pearl weeds to crop and pasture/lawn grasses are More than 250 members of the grass family millet), Panicum miliaceum L. (proso) and detailed. Conflicts of interest that arise be Poaceae (; Gramineae) are regarded as Setaria italica (L) Beauv. (foxtail millet) reo cause many grasses either are both crop weeds either in standard Australian weed spectively. weeds and valuable pasture or lawn grasses texts and lists (Auld and Medd 1987, Klein Table 2 lists the weedy grasses which are or are generically related to them, are schmidt and Jobnson 1977, Lamp and Collet closely related (uG" ; same genus) to crop listed. 1979, Parsons 1973, Swarbrick 1983, Whittet or pasture/lawn grasses and those cases Examples are given of the possible use of 1968, Wilding et aL 1986) or are aquatic where the same grass ("S" ; same species) is the two methods as follows: weeds (Mitchell 1978, Sainty and Jacobs both a weed in some situations (i.e. crops) Avena spp., Hordeum spp and Echinochloa 1981) or have herbicide recommendations and an important component of native or spp. because of their close relation to crop for their control (Swarbrick 1984). Of these, improved pastures are also indicated based grasses, Stipa spp. because they are native 170 species, including most of the important on comments on the pasture importance of pasture grasses and Cynodon dacJylon be weeds, are of exotic origin. grasses in Whittet (1969), Reid (1981), Bur· cause both a lawn grass but also a crop The taxonomic relations of Australian bidge (1966, 1968, 1970, 1984), Lazarides weed, could only be controlled using agents grass weeds and crop grasses are based on (1970), Wheeler et al (1982), and Tothill and already present in Australia inundatively. Watson and Dallwitz (1985) and Clayton and Hacker (1983). The biological control of Bromus spp. which could also be the subject Renvoize (1986) (Table 1). There are only these grasses could be compromised bY con of bioherbicidal control, Holcus s pp. and, in two major differences between their classifi flicts of interest between land users wishing particular, Phragmites spp. which has a cations. One, the separation of the centoth to maintain these grasses in pasture or lawns large number of apparently specific agents ecoid group as a separate sub-family in Clay or wishing to grow related crops and others in its Old World home range, could possibly ton and Renvoize (1986) and its inclusion as seeking to control the same or closely related be controlled by the classical introduction a tribe of the Oryzaneae within the sub·fam· grasses biologically. Indeed, many grasses of exotic agents. Sorghum hakpense could ily Bambusoideae bY Watson and Dallwitz would be considered valuable fodder for have rhizome-feeding agents introduced to (1985), is not relevant here as there are no stock but weeds when that same land was control it but not agents attacking aerial Australian weeds, crops or pasture grasses in ploughed for crops bY the same farmer. parts of the plant which would infest crop that centothecoid group. The other is the Table 3 lists the grass weeds for which there sorghum. Eleusine indica could possibly be placement of the tribe Stipeae in the sub would not be economic conflicts of interest in controlled using both methods. Nassella family Arundinoideae by Watson and Australia although some are regarded as trichowma is probably too closely related to Dallwitz (1985) instead of in the sub·family minor lawn or decorative garden plants and native Stipa spp. to allow the introduction of Pooideae bY Clayton and Renvoize (1986). others such asAmmophila arenaria (L.) Link agents. Recent discussions on the taxonomic posi are used for sand dune stabilization. It is concluded that each genus of weedy tion of the Stipeae either concur with Watson grasses and in some cases each weedy grass and Dallwitz (1985) (Watson et aL 1985, Biological Control Methods species has to be considered individually Barkworth and Everett 1986) or suggest that Two types of biological control will be consid and the type of biological control selected this tribe is basal to the Pooideae rather than ered here. according to the following features:- belonging to it (Kellogg and campbell 1986). 1. Inundative or bioherbicidal control, where i) whether the weedy grass is related to crop The evolutionary position of both the rust an agent is artificially increased, bulked up, and/or pasture grasses, fungi and sm ut fungi infesting Stipeae sug and applied bY the land user in the same ii) wheth-;r a pool of potential agents occurs gests that this tribe is intermediate between manner as a chemical herbicide. Disease in Australia already, the Arundinoideae and the Pooideae (Wat- organisms such as fungi and nematodes Plant Protection Quarterly Vot.5(2) 1990 63 Table 1. Taxonomic position of Australian grass weed genera TRIBE TRITICEAE Hordeum, Agropyron, CROPS; Triticum, Secale Position of grass crop genera indicated where different from weed (2 tribes in Triticanae (W. & D.)) genera. Based on Clayton and Renvoize (1986) (c. & R.), relevant SUB-FAMILY CHLORlDOlDEAE differences between them and Watson and Dallwitz (1985) (W. & D.) TRIBE ERAGROSTIDEAE as indicated. SUB-TRIBE ELEUSININAE Leplochloa, Dinebra, Eragrostis, FAMILY POACEAE (= GRAMINEAE) Triraphis, Eleusine, SUB-FAMILY BAMBUSOlDEAE Dactyloctenium, Dip/aehne TRIBE BAMBUSEAE SUB-TRIBE SPOROBOLINAE SUB-TRIBE ARUNDINARIlNAE Sparobolus Anmdinaria TRIBE CYNODONTEAE SUB-TRIBE BAMBUSINAE SUB-TRIBE CHLORIDINAE Bamb"sa, Phyllostachys Chloris, Brachyachne, Sparlina, Cynodon TRIBE ORYZEAE SUB-TRIBE ZOYSIlNAE Oryza, Leersia Tragus, Perotis TRIBE EHRHARTEAE (all above chloridoid sUb-tribes Ehrharta combined (W. & D.)) SUB-FAMILY ARUNDINOlDEAE SUB-FAMILY PANICOIDEAE TRIBE ARUNDINEAE TRIBEPANICEAE Danthonia (= Rylidosperma). Cortaderia, SUB-TRIBE SETARIINAE (both in Dantbonieae, (W. & D. )),Arnndo, Pan;cum , Echinochloa , Brachiaria, Urochloa , Phragmites Paspalum , Axonopus I Setaria , Paspalidium , TRIBE ARISTIDEAE Eriochloa, Stenotaphrum Aristida SUB-TRIBE MELINIDINAE SUB-FAMILIES ARUNDINOJDEAE/POOlDEAE Rhynchelytrnm, Melinis TRIBE STIPEAE SUB-TRIBE DIGiTARIINAE Stipa, Nassella, Oryzopsis (= Piptathenlm) Digilaria (all 3 in Stipeae, Arundinoideae (W. & D. ) SUB-TRIBE CENCHRINAE but all 3 in Stipeae, Pooideae (c. & R. )) Cenchrns, Pennisetum SUB-FAMILY POOlDEAE TRIBE ANDROPOGONEAE TRIBEPOEAE SUB-TRIBE SACCHARINAE Fesluca, Lolium, Vulpia, Psi/urns, Cynosurns. Imperata, CROP; Saccharum Lamarelda, Poa, Desmazeria, (= Catapadiwn 1 SUB-TRIBE SORGHINAE Dactylis, Brim Sorghum, Dichanlhillm , Chrysopagon, TRIBE HAINARDIEAE BOIhriochloa Pholiunls, Parapholis, Hainardia (= Monenna) SUB-TRIBE ANDROPOGONINAE (all 3 in Poeae (W. & D.)) Andropagon TRIBE MELICEAE SUB-TRIBE ISCHAEMINAE Glyceria Ischaemum TRIBE A VENEAE (a ll above sub-tribes in uawned" SUB- TRIBE A VENINAE Andropogoneae (W. & D.)) A vena, A"henatherum, Ho/cus, Periballia SUB-TRIBE ANTHISTIRIINAE (= Molineriella1 Aira, Rostraria Hyparrhenia , Themeda , [seilema, ( =Laphoch/oa) Heleropogon SUB-TRIBE PHAIARlDINAE SUB-TRIBE ROTTBOELLIINAE Anlhoxanlhum, Phalaris Hemarlhrio, Rottboellia SUB-TRffiE ALOPECURINAE (above 2 sub-tribes in "awnless" Agrostis, Ammophila, i.agurus, Andropogoneae (W. & D.)) Polypogon, Alopecurns, SUB-TRIBE CHIONACHININAE Gastridium, Echinopogon Chionachne (all 4 tribes in Poanae (W & D)) SUB-TRIBE TRIPSACINAE TRIBE BROMEAE CROP:Zea Bromus (above 2 sub-tribes in Maydeae (W. & D. ))

are particularly able to be developed as specificity either to tbe grass weed alone or agents present in Australia would normally bioherbicides (Wapshere 1982). Bioherbi to it and a few unimportant close relatives, be considered and they would have to be de cides cou Id be developed from diseases almost all types of phytophagous organ veloped as bioherbicides for these grass already present in Australia on tbe weed isms cou ld be introduced as classical weeds. grass concerned or, if sufficiently specific, agents. They would then pose no risk to As it is costly both to develop and to apply disease organisms on the grass could be any crop or pasture/lawn grass whether a bioherbicide this method of biological con imported and then developed as bioherbi native or imported. trol could be used mainly for grass weeds cides (Wapshere 1987). which are major crop weeds, where the cost 2. Classical or inoculative control, where the Bioherbicidal Control of Grass Weeds of development and of application would be agent is simply introduced, released, then For all those grass weeds listed in Table 2 repayable. Table 4 lists the grass weeds which disperses and self-propagates achieving with importance as crop, pasture and lawn are important in crops (Auld and Medd control without further human interven grasses it would not be possible to introduce 1987, Wilding el al. 1986) andlor have the tion (Wapshere 1982). Given adequate biological control agents from overseas. Only most herbicide recommendations (Wapsh- 64 Plan I Protection Quarterly VOI.5(2) 1990 Table 2 . Australian grass weeds closely related to crop, pasture and lawn grasses Table 3. Grass weeds with no or little connict of interest for control

Weed Crop or pasture or lawn grass Relation Aira spp. S = Same species Alopecllrus spp. G = Same genus Ammophila arenaria AndrojXJgon virginicus Agropyron repens Agropyron spp. (wheat grasses) G Anthoxanthum odoratum Agrostis spp. A. avenacea (blown) and S&G Arislida spp. Agrostis spp. (bents) A"henalherom elatius Aristida spp. Aristida spp. (three-awn) S&G Arondinaria spp. Avena spp. A. sativa (oats) G Arondo donax Bothriocltloa macro B. erianthoides (satin top) and G Axonopl/S spp. Bothriochloa spp. (blues) BambllSa spp. Brachiaria spp. B. mUlfea (para) and BrachiaTia spp. S&G Brachyachne spp. Bromlls spp. E. catharticllS (prairie) S&G Briza spp. Cenchms spp. C. ciliaris (buffel) and S&G Chionachne hubbardiana C .seliger (birdwood) Cortaderia spp. Chloris spp. C. gayana (rhodes) and Chloris spp. S&G Cynosllnls spp. Chrysopogon aciculatus Chrysopogon spp. (golden-beards) G Desmazen·a rigida Cynodon spp. C. dacty/on (couch) S&G Dinebra retrofle.xa Dactylis glomerata D.glomerata (cocksfoot) S Diplachne spp. Dactylocteniwn spp. D. radulans (button) S&G Ehrharta spp. Dantltonia spp. Danthonia spp. (wallaby) S&G Eleusine spp. Dichanthium spp. D. sericeum (Queensland blue) and S&G Gastridiwn ph/eoides Dichanthiwn spp. (blues) Hainardia cylindrica Digitaria spp. Digitaria spp. (summer grass) G Hemarthria UI1cinata Echinochloa spp. E./nunenlacea and E. utilis S&G Heteropogon conlortllS (Siberian and Japanese millets) HolcllS spp. Echinopogon spp. Echinopogon spp. (hedgehogs) S&G Hypa"henia hirta Eragrostis spp. E. curvl/la (African love) and S&G Imperata cylindrica Eragrostis spp. (loves) Ishaemum rugosum Eriochloa spp. Eriochloa spp. (early spring) S&G Lagurns ovatus Festuca spp. Festuca spp. (feseues) S&G Lamarckia aurea Glyceria maxima G. maxima (water meadow) S Leersia spp. Hordeum spp. H. vulgare (barley) G Leplochloa spp. Iseilema spp. Iseilema spp. (flinders) S&G Laphochloa cristata Loliumspp. L. perenne (perennial rye grass), S&G Molineriella minula L. multiflorwn (Italian rye) and Nassella trichotoma L. rigidum (Wimmera rye) Parapholis incurva Melinis minutiflora M. minutiflora (molasses) S PerOlis rara Oryza spp. O. sativa (rice) G Pholiuros pannonicus Panicum spp. P. miliaceum (proso) and S&G Phragmites spp. Panicum spp. (panics) Phyllostachys spp. Paspolidium spp. P. globoideum (shot), S&G Piptathen/m (= Oryzopsis) miliacellm P. jubiflorum (Warrego summer) and Polypogon spp. Paspalidillln spp. (panics) Psi/emu incllfllUS Paspalum .pp. P. dilatatum (paspalum), S&G RhyncheIyIrll/n repens P. scrobiculatum (serobic) and Rottboellia emitata PaspaIum spp. Spartina Townsendi; Pennisetum spp. P.gIaucwn (pearl millet), S&G Sporobolus spp. P. clandestinllm (kikyu) and Triraphis mallis Pennisetum spp. Vulpia spp. Phalaris spp. P. aquatica (phalaris) and S&G Phalaris spp. (canaries) ere 1987). Of these, those with a conflict of Poa .pp. P.pratensis (Kentucky blue) S&G interest and for which agents that could pos and Poa spp. (tussocks) sibly be developed as bioherbicides are al ready present in Australia, areAgrostis lenuis Setaria spp. S. sphacelata (setaria) and S&G S. italica (foxtail millet) Sibth., Avena spp., Bromus spp., Cenchrus spp., Chloris spp., Cynodon dactylon (L) Sorghum spp. S. bicolor (sorghum), S. sudanese G (Sudan) and S. x a/mllm (Columbus) Pers., Digitaria spp., Echinochloa spp., Era gros/is spp., Hordeum spp., Lalium spp., Slen olaphn~m secundatum S. secundatum (buffalo) S Stipa spp. Stipa spp. (spears) S Panicum spp., Paspa/wn spp., Pennisetum Themeda .pp. T. australis (kangaroo) and S&G clandeslinum Chiov., Phalaris spp., Poa an· T. avenacea (native DaiS) nua L, Setaria spp., Sorghum halepense (L.) Pers. and Urochloa panicoides Beauv_ The Tragus australianus T. allstralianus (small burr) S Urochloa spp. U. panicoides (liverseed) S&G few remaining grass weeds in Table 4 which involve less conflict of interest and for which Plant Protection Quarterly VoI. 5(2) 1990 65 Table 4. Principal grass weeds suitable Table 5. Relation of weedy grass genera in Australia to native and overseas for bioherbicidal control grasses Based on those listed as crop weeds in Auld and Medd ( 1987) and Wilding et al. (1986) Based on figures for world and Australian species in each grass genus in Baines (1981). and on those with most herbicide recommen dations (Wapshere 1987) A) GRASS GENERA IN WHICH WEEDS ARE ALL AUSTRALIAN NATIVES. I) GENERA WHICH ARE SOLELY AUSTRALASIAN. Agrostis lenllis Danthonia (=Rytidosperma), Echinopogon. A"henatherom ela/ius 2) GENERA WITH EXOn C REPRESENTATIVES BUT WITH LARGE Avena spp. NUMBERS OF SPECIES NAnVE TO AUSTRALIA. Briza minor Aristida, Dichanthium, Iseilema, Paspalidiwn, Stipa. Bromus spp. 3) GENERA WITH MOSTREPRESENTA'nVES EXOnC BUT WITH A CenchnLY spp. FEW AUSTRALIAN SPECIES. Chloris spp. Bothriochloa, Brachyachne, Chionachne, Chrysopogon, Diplachne, Cynodon dactylon Enochloa, Hemarthnn, Heteropogon, Imperato, Leptoclrloa, Peralis, Digitaria spp. *Phragmites, Themeda. Tragus. Triraphis. Echinochloa spp. B) GRASS GENERA IN WHICH ONE OR MORE WEEDS ARE EXOn C AND E/eusine indica OTHERS AUSTRALIAN NATIVES. Eragroslis spp. I) GENERA WITH EXOnC REPRESENTATIVES BUT WITH LARGE H o/ells lana/us NUMBERS OF SPECIES NATIVE TO AUSTRALIA. Hordewn spp. Agrostis, Brachiaria, CII/oris, Digitaria, Eragrostis. Panicum, Pon, Lolium spp. Sorghum, Sporobolus. Panicwn spp. 2) GENERA WITH MOST REPRESENTATIVES EXOTIC BUT WITH A Paspalwn spp. FEW AUSTRALIAN SPECIES. Pennisetwn clandestinum Agropyron, Cenchms, Cynodon, Dactylocteniwn, Echinochloa, Festllca, Phalaris spp. Glyceria, Leersia, Oryza, Paspalum, Pennisetwn, Setaria. Poaannua C) GRASS GENERA IN WHICH WEEDS ARE EXOnC TO AUSTRALIA. Selaria spp. 1) GENERA WITH ONE OR VERY FEW AUSTRALIAN SPECIES Sorghum ha/epense AlopecunlS, Anmdinaria, Bambusa, Bromlls, Hyparrhenia, Ischaemwn, Urochloo panicoides Rottboellia. Vulpia spp. 2) GENERA EXOTIC TO AUSTRALIA Aira, Ammophila. Andropogon, AnthoXllnlhwn, ATThenathenun, Anmdo, Avena, Axonopus, Briza, Corladeria, CynosllnlS, Dactylis, it might be possible to consider introducing Desmazeria, Dinebra, Ehrharta, *Eleusine, Gastridiuln, Hainardia, agents from overseas for SUbsequent devel· HO/CllS, Hordeum, Lagums, Lamarckia, Loliwn, Lophochloa, Melinis, opment as biohcrbicides, are A"henalhernm Molineriel/a, Nassella, Parapilolis, Pilalaris, Pholill",s, Phyllostachys. elo/ius (L.) Presl, Briza spp., Eleusine indica Piptathenlln, Polypogon,Psilttnls. Rhynchelymlln, Spar/ina, (L.) Gaertn., HolClls lanalus L. and Vulpia Stenolaphnun, Urochlon, Vulpia. spp. As none of these are native to Australia, additional agents could be found for intro • See later discussion for these 2 genera. duction from their borne ranges. However, of these grasses only E. indica does not have of A vena, Hordeum and Lolium have a large ticularly [hose related to the weedy grasses. tribal relations with an important crop or group of inadvertently introduced agents on As already noted, it has been established pasture grass. them here (Simmonds 1966, Sampson and that the greatest number of agents specific to Any attempt to control a particular grass Walker 1982, Woodcock and Clarke 1983, a weed or its close relatives occur at its centre weed within a cereal crop or amongst other Shivas 1989, Cook and Dulle 1989, Queens of origin and/or where the groups of related • grasses in a pasture sward would require a land Department of Primary Industry, Un species occur (Wapshere 1974a). The classi certain level of speCificity. It has been shown pub!,) but most of the grasses in the exotic cal method depends on finding suitable that a larger guild of specific agents occurs genera listed in C2 have very few fun gi re agents for the weed concerned elsewhere in where a given weed occurs together with a corded on them in Australia. the world. Thus weeds of exotic origin and large group of species in the same genus, as those native to Australia but with greater at evolutionary centres of genera or subgen Classical Control of Grass Weeds species representation in the same genera era (Wapshere 1974a). Table 5 shows the This method does not have the economic elsewhere (in A3, B2, CI and C2, Table 5) relation of Australian grass weed genera in constraints of bioherbicidal contrOl, but it is should initially be considered. Weeds of gen terms of their species distribution between severely constrained as far as grasses are con era limited to Australia or belonging to gen Australia and elsewhere. Large groups of cerned by the impossibility of introducing any era of wider distribution but with large spe specific or near specific agents on Australian agent that would infest any crop or pasturel cies groups here (in AI, A2 and BI, Table 5) grasses would be expected in those genera lawn grass once released. Thus none of the are less likely to have suitably specific agents which occur only in Australia and close re weeds listed in Table 2 as the same species as available outside Australia. Except for Spo gions or have large groups of species in Aus crop or pastureftawn grasses could be con roboulS spp. all genera in Table 3 belong to tralia (in AI, A2 and BI, Table 5). Fewer trolled in this way, and the others could only groups A3, B2 and C in Table 5, indicating specific or near specific agents would be be considered if any agents overseas were that suitable agents could be found to cont rol likely to be present here On grasses in those restricted to the weedy species of that grass many of them in their regions of origin and genera with few or no Australian representa genus. The discu ssion below indicates that perhaps elsewhere overseas for the more tives (in A3, B2, C I and C2 in Table 5), and species specificity is rare amongst grass or widespread grasses. nearly specific agents would probably only be ganisms compared with generiC specificity. As specificity or near specificity is of prime found if they had been inadvertently intro The risk posed to the large number of native importance for any agent introduced to con duced. A few exotic grass weeds, e.g. species grass species has also to be conSidered, par- trol a grass weed, the distribution of recorded 66 Plant Prolection Quarterly VOI.5(2) 1990 Table 6 Recorded host range or various types or organisms (a) Attacking Australian grass weeds or European or igin in Europe

No and % of Species in Each Host Range Level

Type of Recorded One One Total Reference Organism on: Species Genus 2-3 4+ Species Only Only' Genera Genera ofType

INSECTS NOCfUIDAE No 10 10 14 90 114 Forster & (Cutworms) % 8.8 8.8 12.3 78.9 Wohlfahrt (1971) MICROLEPIDOPTERA No 7 16 28 12 56 Schutze (1931) (Small MothS) % 12.5 28.6 50.0 21.4 DlPTERA, No 15 15 13 22 50 Seguy (1934) BRACHYCERA (Flies) % 30.0 30.0 26.0 44.0 CECIDOMYIlDAE No 17 39 7 5 51 Barnes (1946) (Gall-midges) % 33.3 76.5 13.7 9.8 APHIDAE No N/A 15 10 20 45 Borner (1952) (Non-hOst-alternating Aphids) % 33.3 22.2 44.5

FUNGI USTILAGINALES No 16 37 16 6 59 Zundel (1953) (Smuts) % 27.1 62.7 27.1 10.2 UREDlNALES No 4 7 7 10 24 Cummins (1971) (Rusts asexual phase only) % 16.7 29.1 29.1 41.7

HABIT LEAF-MINERS No 10 38 28 55 121 Hering (1957) TYPES (Insects) % 8.3 31.4 23.1 45.5 GALL-MAKERS No 29 42 9 16 67 Buhr (1964, 1%5) (Arthropods, Nematodes & Fungi) % 43.0 62.6 13.5 24.9 b) Attacking grasses world wide

FUNGI ASCOMYCETES PHYLLACHORA spp. No N/A 85 25 \3 123 Parberry (1966,1971) DEUTEROMYCETES % 69.1 20.3 10.6 CERCOSPORA spp. No N/A 22 4 3 29 Chupp (1953) % 75.9 13.8 10.3 STAGONOSPORA spp. No N/A 54 11 9 74 Costellani % 73.0 14.9 12.1 & Germano (1975)

• Note that figure for one species only is included in figure for one genus only so that total species of given type of organism is sum of last 3 columns. host range of organisms on a group of grasses from only 9% ofnoctuids limited to both one (Table 6). requires examination. species and one genus of grass, to as high as In the case of habit types, 31 % of leaf-min The only group of grasses for which ade 33% and 77 % of cecidomyiids limited to one ers from all insect groups arc restricted to quate data are readily available are those species and one genus respectively. Of the one grass genus. However, gall-makers are from Europe. A group of Australian grass fungi the smuts which have 27% and 63% of fa r more specific, 43% of gall-makers of all weeds were selected because of their occur their species limited to one species and one types (arthropods, nematodes and fungi), rence in Europe (Tutin el al. 1980). For all genus respectively are the most specific to being restricted to one grass species and 63% organisms listed in Table 6 European or their grass hosts. to one grass genus (fable 6). Thus, of those world wide host lists were used. Some other fungal genera appear to have organisms conSidered, except for the noc Cummin's (1971) world list of rust fu ngi restricted host ranges. Based on world lists, tuids, there are several types with different on grasses was used rather than Gaumann's about 69% of graminicolous species in the habits which have sufficient specificity to (1959) European list of rust hosts because ascomycetous genus Phyllachora are only serve as classical biological control agents, the latter has narrower specific distinctions recorded from one grass genus (Table 6b) particu larly if the grass weed belongs to a between rust species. Cum min's (1971) list and the deuteromycetous genera CercosfX1ra genus distinct from those of crop and pas therefore gives a conservative result for the (Hyphomycetes) and Stagonospora (Coelo ture!lawn grasses. host range of grass rusts. mycetes), have more than 70% of their In Europe, the organisms which are par As well as considering grass organisms in members recorded only from a single grass ticularly restricted in recorded host range to taxonomic groupings it is also possible to use genus. However, the host restrictions have one grass species or genus are:- amongst the the European data to investigate their speci not been confirmed by cross-inoculation and insects; elachistid moths (Lepidoptera); chlo ficity when in a particular habit. Leaf miners the taxonomy of CercosjXJra and Stagono rapid flies (Diptera, Brachycera); (Hering 1957) and gall makers (Buhr 1%4- sporn are in a state of flux at the moment cecidomyiid gall midges; and chalcid ga ll 65) were used for this part of the analysis. (Walker, New South Wales Department of wasps in the genus Telramesa (= Hannolila): The distribution of recorded host ranges Agriculture, pers. comm.). Despite this, the amongst the fungi; the smuts (Ustilaginales). (fable 6a) of the principal organisms attack data for these fungi arc comparable with Furthermore, in the USA each Tetramesa ing grasses in Europe vary, for the insects, those for the arthropods and habit types species tested has been shown to be re- Plant Protection Quarterly VOI.5(2) 1990 67 stricted to a single grass genus (Phillips Australia. The genera which are little differ ered suitable for further study by Thurston 1920). entiated from their respective agriculturally and Cussans (1976). The other five damag The rust fungi (Uredinales) on grasses are important relatives (Clayton and Renvoize ing fungi were all known pathogens of culti not particularly specific according to Cum 1986) are as follows :Ammophila, Gastridium vated cereals. P. hibernica is now included in min's (1971) data but would be considered to and Polypogon from Agrostis;Arrhenalherum Phoma herbantm Westend, a saprophytic be so if Gaumann's (1959) specific distinc from Avena; Brachyachne from Cynodon; fungus with a broad host range (Boerema tions had been followed. However I nearly all Briza from Poa; Imperata from Sacchanun; 1964) which occurs in Australia (Woodcock grass rusts have alternate hosts in the other uerzia from Oryza; Nassella and Piplath and Clarke 1983). In England, the seed plant families. Although persistence in Aus erom (= Orywpsis) from Slipo ; and Vulpia borne fungus Pyrenophora avenae Ito and tralia may not be a problem, as many exotic from Festuca. There still remains a large Kuribay has been investigated as a potential rusts of grasses which have found their way number of grass species that could be consid control agent (Wilson and Hall 1987). This here can persist without their alternate hosts ered for classical biological control (Table 3). fungus infests both A vena sterilis Land A. being present (McAlpine 1906), introduction However, there is no certainty that the sativa but could not maintain itself on wbeat could only occur if possible alternate hosts of agents specific at the generic level would be (Trilicum aeslivwn L.) or barley (Hordeum agricultural or conservational importance the ones able to control a given weed. If it vulgare), producing only hypersensitive were not infested. The same problem of in was necessary to consider agents with a host necrotic spots on these two cereals. Thus it festing alternate hosts amongst the dicotyle range extending to tribal level this would de could be used to control weedy Avena spp. in dons applies also to many grass aphids, 37% lete from consideration the weeds in the wheat or barley crops (Wilson and Hall of European grass aphids having alternate Panicoid family and the weeds in the Pooid 1987). P. avenae has been recorded in Aus hosts (Borner 1952). For this reason, only tribes, Poeae, Aveneae and Triticeae (Table tralia in its conidial stage Drechslera avenae the host range of non-alternating grass 1). This would seriously reduce the number (Eidam) Scherif (= Hebninthosporium ave aphids is given in Table 6a and these aphids of grass weeds that cou Id be considered for nae) (Simmonds 1966). are specific at the genus rather than species classical biological control to species of In Canada, both Hebninlhosporium spp. level. Arondinaria, Bambusa, Dactylocleniwn, (Watson and Harris 1975) and Collelol Some species of chloropid flies, Dinebra, Diplochne, Ehrhar/a, Eleusine, richum graminicola (Ces.) Wilson cecidomyiid gall midges, chalcid wasps and Hainardia, uplochloa, Parapholis, Pho (Mortensen 1983) have been recommended aphids infesting cereals are all major pests of liunls, Phylloslachys and Triraphis. Consid for detailed study as agents to control Avena these crop grasses (Balachowsky and Mesnil eration could also be given to species of spp. More recently, Mortensen and Hsaio 1935). Similarly, some of the rust fungi, Arundo, Cortaderia and Phragmites, which (1987) studied the fungi infesting seeds of smuts, SeplOria, CercosfXJra, Drechslera and are related only to a few native grass genera Avena spp. there and found, of tbe five com Bipolaris (both previously Helminlho including Danthonia. Species of Hainardia, monest, only Drechslera avenacea (Curtis ex sporium) spp. (Fischer and Holton 1957, Parapholis and Pholiurus could also not be Cooke) Shoem. was restricted to Avena spp., Sprague 1950) cause major diseases of their considered if Macfarlane and Watson's the others were equally or more pathogeniC cereal hosts. Species from these groups of (1982) and Macfarlane's (1987) classifica to wheat, barley and rye (Mortensen and insects and fungi could have the same dam tions were used. Hsaio 1987). This fungus occurs in Australia aging effect on grass weeds to which they (Woodcock and Clarke 1983). were specific or near specific. Indeed the Comments on Particular Grass Weeds 10 Au stralia, a simulation model of the rusts and smuts of cereals have caused major The grass weeds selected for comment are population dynamics of Avena spp. suggests disease outbreaks when they have inadver important in Australia and illustrate the that controlling seed production or survival tently been introduced into Australia problems of biological control of these would be the most efficient way of controlling (McAlpine 1906, 1910). weeds. these annual grass weeds (Medd and Ridings All genera of grass weeds are not infested 1989). Strains of an aerially transmitted fun equally by organisms restricted to one species Avena spp. gus disease which has a broad host range or genus. Avena spp. and Hordeum spp. have Wild oats, A vena fatua L., and other A vena amongst pasture grasses and cereals are only a few restricted organisms infesting spp. are major annual weeds of cereal crop being investigated as regards pathogenicity them whereas Phragmiles australis (Cav.) ping and other cultivations in Australia and specificity to weedy Avena spp. The Steud. has many specific or near specific or (Wilding el aL 1986, Auld and Medd 1987). eventual aim is to develop suitable strains as ganisms infesting it (see belOW). The host A. latua is probably of Central Asian origin mycoherbicides (Medd and Ridings 1989). range distribution of organisms on the group and the other weedy A vena spp. are of Medi A number of fungi have been recorded on of European grasses can be considered rep terranean, Middle Eastern and European the weedy species ofA vena in Australia and a resentative of the situation likely to exist for origin (Holm el aL 1977). A. latua and the much larger number have been recorded on grass weeds from other parts ofthe world but other Avena spp. are all closely related to the cropA. saliva (Simmonds 1966, Sampson for which there are no adequate data on host each other and to the crop grass, cultivated and Walker 1982, Woodcock and Clarke range distributions. oats, Avena sativa, and hybridization occurs 1983, Walkden Brown 1987, Shivas 1989, It is accepted by workers in classical bio between them. There are noAvena spp. na Cook and Dube 1989, Walker NSW Depart logical control that plants related to the weed tive to Australia. Studies on the possibilities ment of Agriculture pers comm.). are more at risk from biological agents infest of the biological control of A vena spp. have ing that weed than unrelated plants. Thus, been carried out in Europe, within the weeds Hordeum spp. close taxonomic relation is a major criterion native range, and in Canada where they are The barley grasses are serious pasture weeds for selecting plants to demonstrate the safety also unwelcome introductions, as well as in causing stock damage because of their sharp of agents for introduction (Wapshere Australia. Thurston and Cussans (1976) sur awns and some are weeds of winter -grown 1974b). Table 3 lists the weeds that are gen veyed the organisms infesting Avena spp. in crops (AUld and Medd 1987, Wilding el aL erically separate from important crop and Europe and concluded that only seed infest 1986). However, two speeies, Hordeum le pasture~awn grasses. However to be on the ing fungi were worth further study. Kiewnick porinum Link and H. marinum Huds. are conservative side it would be appropriate to (1%3, 1964) had previously studied the fungi self-seeding pasture grasses producing useful consider, initially, only those genera of grass infesting seeds ofA.larua in Germany and of forage before heading (Reid 1981). They are weeds which are well separated from the the six most damaging species only one closely related to cultivated barley Hordeum grass genera of agricultural importance in Phoma hibernica Grimes et aL was consid- vulgare L. and are tribally related to both 68 Plant Protection Quarterly YOI.5(2) 1990 wheat, Triticum spp., and rye, Secale cereale Renvoize 1986, Watson and Dallwitz 1985) Europe, which is now an important weed of (Table 1~ The weedy species of Hordeum in and this separation is reinforced by the latest western USA. This work has not, however, Australia are natives of Eurasia. classification of the Pooideae (Macfarlane progressed beyond a list of fungi and nema Hordeum spp. have not been considered for 1987). todes occurring on the weed in North Amer biologica l control in other paris of the world There is one native species of the genus, ica (Peeper 1984). and there 3rc no biological control studies of Bromlls arenarills LabilL (Burbidge 1984) A survey of the arthropodS and fungi of agents in their home ranges but they have and there are no other native genera in the Bromus spp. in Europe indicates a feworgan been recommended as one of the annual tribe Bromeae (Watson and Dallwitz 1985). isms sufficiently specific to them to be consid grass weeds for mycoherbicidal control in Bramlls calharticus Yah!. (= B. unioloides ered as biological control agents in Australia Australia (Medd and Ridings 1989). In Aus Kunth), prairie grass, regarded as a usefut (Table 7). At first, the most interesting in tralia, there is a large number of fungi re pasture grass, is a native of South America. sects would be the lepidopterous leaf miners corded on the crop barley, some of which However, the major weedy species originate Elachista spp. and the gall-forming insects, have also been recorded on the weedy barley in Europe, the Mediterranean region and the aphid Diuraphis (Holcaphis) bromicola grasses (Simmonds 1966, Sampson and Middle East. One of these Bromlls mollis L. (H.R.L.) and the eneyrtid wasps Tetramesa Walker 1982, Woodcock and Clarke 1983, is also of some value as a pasture grass (Reid (= Harmolita) spp. However, the records Sparrow and Doolette 1987, Shivas 1989, 1981). come mainly from northern Europe and Cook and Dulle 1989, Qld DPI unpub!.). The only weedy brome that has received most Australian weedy bromes are of Medi Many of them also infest wheat and rye as attention as far as biological control is con terranean origin. Very little is known con well as barley (Ioc. cit. and Lovett 1987). cerned is Bromus /ee/orum L., a native of cerning the insects and fungi of bromes in

Stipa spp. Table 7 (a) Arthropods specific or near specific to Bromus spp. in Europe Most spear grasses 3rc natives to Australia and arc important clements of native pas Order/Family Species Comments/habitat References tures (Burbidge 1984). However, the seeds with their spiral awns contaminate wool and Lepidoptera penetrate the skins, mouths and eyes of Pyralidae Agriplrila latistria (Haw.) L'Homme (1923-49) sheep (Auld and Medd 1987). Although spe Elachistidae Elachista lastrella Chret. Leaf miner Emmett (1979) cies occur in Eurasia and the Americas, the E. bromella Chret. Leaf miner Hering (1957) Australian species form a distinct, probably Diptera basal, group of the genus and of the tribe Sti Cecidomyiidae Con/arinia sp. Flowers Barnes (1946) peae (Barkworth and Everett 1986). It can Mayetiola sp. Stem-gall Buhr (1964) be expected that organisms adapted to the Agromyzidae Agromyza brami Spencer Leaf-miner Spencer (1972) Australian spear grasses would already occur in Australia and that there would be little Hemiptera Aisoon point in searching other parts of the world for Miridae Amblytylus albidus (Hahn) Corynophorus Wagner & Weber suitable agents. canescens (1964) Approximately 17 fungi are recorded on Acetropis gimmenhali (Flor.) Stichel (1955-1962) Stipa spp. in Australia, including 6 smuts Aphidae Diuraphis (Holcaphis) Shoot-gall Borner (1952) (McAlpine 1910, Simmonds 1966, Sampson bromicola (H.R.L.) Buhr (1964) and Walker 1982, Woodcock and Clarke Cicadellidae Mogangina bromi EM. Bey-Bienko (1967) 1983, Cook and Dulle 1989). Although little Coccidae Lecanopsis tOllrica Borchs Bey-Bienko (1967) is known about the insects which infest them, it can be assumed that many of the native Hymenoptera insects in Australia that have switched to ex Eurytomidae Tetramesa (= Harmolita) Stem-gall Claridge (1%1) otic crop and pasture grasses (Whittet 1969, maculata (Howard) Hassan 1977, Lazenby and Matheson 1987) Tetramesa (= Harmolita) sp. Stem-gall Buhr (1964) are derived from the fauna on native grasses. Since the agents adapted to the native spear Table 7(b) Fungi specific or near specific to Bromus spp. in Europe grasses are already in Australia and some of these grasses are important in native pas Order/Family Species Comments!habitat References tures, the spear grass problem could only be tackled by inundative or bioherbicidal meth Basidiomycetes ods. However, little is known about the or Ustilaginales Ustilago bromivora (Tulasne) Possibly part of ganisms infesting them and more detailed Fisch U. bullata Berk. Zundel (1953) studies of the fungal flora and nematode Tille/ia bromina Maire F10wer head smuts fauna of spear grasses in Australia would be T. guyotiana Hariot required to determine whether a sufficient pool of suitable agents on these grasses is Ascomycelae Pyrenophora bromi (Died) Oudemans (1919) available for development as bioherbicides. Drechsler Deuteromycetae Drechs/era Bromus spp. (=Helminthosparium) bromi Leaves. Teleomorph Oudemans (1919) Brome grasses are major pasture weeds in (Died) Shoem. Pyrenophora bromi temperate Australia and can be important in Sep/oria bromi Sacco Leaf spot Sprague (1950) crops (Wilding et al. 1986, Auld and Medd Stagonospora bromi AL. LeafblOlch 1987). Despite a close relation to members Smith & Ramsb. of the tribe Triticeae, the large genusBromus Pseudosep/oria and twO other small grass genera are placed (= Selenophoma) bromigena Leaf spot in the separate tribe Bromeae (Clayton and Sprague & Johnson Plant Protection Quarterly VOLS(2) 1990 69 that region of Europe. None of the insect Table 8(a) Arthropods specific or near specific to Holcus spp. in Europe species listed in Table 7a is k.nown to occur in Australia. However, of the specific fungi Order!Family Species Comments!Habitat References listed in Table 7b on Bromus spp. in Europe, USlilago bromivora (TuL) Walsh. (as U. bul Lepidoptera lata Berk.), SepLOria brain; Sacco and Psychidae Reisseronia tarnierella Traugott-Olsen & Drechslera bromi (Died.) Shoem. are all re (Bruand) Schmidt-Nielsen (1977) corded in Australia. Forms of the rust Puc L'Homme (1923-49) cinia recondita ROb. ex Desm. infesting Elachistidae Elachista rufocinerea A1s0A"henathenun Emmet (1979) Bromus spp. also occur here. Only two of the (Haw.) elatius? Leaf miners Hering (1957) Bromus specific fungi of certain identity, E. pulehella (Haw.) Slagonospora brain; Smith and Ramsb. and Pseudoseptoria bromigena B. Sulton are not Diptera already found in Australia (Sampson and Cecidomyiidae Mayeliola holci Kieff. Stem-gall Barnes (1946) Walker 1982, Woodcock and Clarke 1983, Contarinia sp. Flower Buhr (1964) Shivas 1989, Cook and Dube 1989, Walker, Dasyneura sp. Flower NSW Dept of Agriculture, pers. comm.). Silodiplosis sp. Aower Lestodiplosis sp. Flower Holcus spp. Agromyzidae Cerodonlha Leaf miner Hering (1957) The two fog grasses H. lanatus Land H. flavocingulala (StrobL) mollis L., are pasture weeds in cool higher Metopomyza Leaf miner rainfall regions of New South Wales and Vic flavonolala (HaL) toria (Auld and Medd 1987). HolClls lanaals Opcmyzidae Geomym balachowskyi Stern miner Balachowsky & Mesnil has been selected as a pasture grass but the Mesnil ( 1935) improved varieties are not used in Australia (Reid 1981, Wheeler el aL 1982). Both spe Hemiptera cies are of European origin and there are no Aphidae Diuraphis (Holcaphis) Shoot-gall Borner (1952) native Australian species of the genus. The holci (H.R.L.) genus is however related to 12 native grass Schimphis !wlci (H.R.L.) Buhr (1964) genera (Watson and Dallwitz 1985). Delphacidae Muellerianella Ossianilsson (1978) There have been no reported studies of fainnairei (Perris) the biological control of Holeus species. The placing of the genus Holeus in the sub-tribe Hymenoptera Aveninae (Clayton and Renvoize 1986) indi Eurytomidae Tetramesa Stem-gall Buhr (1964) cates a sub-tribal relation with oats, Avena (= Harmolila)sp. sativa (Table 1). However, a search of the lit erature in Europe reveals a range of organ Table 8(b) Fungi specific or near specific to Holeus spp. in Europe isms that have only been recorded from Holeus spp. (Table 8). Order!Family Species Comments(Habitat References Of these, the gall making insects, the cecidomyiids, the encyrtid wasp Tetramesa Basidiomycetes sp. and tbe aphid Diuraplris (Holcaplris) Ustilaginales Entyloma holci (Liro) Leaf smut Zundel (1953) holei (H.R.L.) are the most likely to be suffi Fisch. ciently specific. Preliminary studies on D. Tillelia holei Flower head smut holei indicate that this apbid is dependent on (Westerdorp) De Toni stimulating gall-like stunting of the grass host to build up populations and this only occurs Deuteromycetae Septoria tritid f. sp. Leaves, f. sp. hold Sprague (1950) on Holeus spp. (Packham 1982). None of the holci Sprague specific to Holcus insects listed in Table Sa appear to occur in Australia. larly important as a weed in subtropical crop Programs to find and develop bioherbicides Although a few fungi are recorded on ping regions of the country (Auld and Medd against S. halepense are underway in several Holeus spp. here, none of the apparently spe 1987, Burbidge 1984, KleinSChmidt and States of the USA. In North Carolina the cific ones listed in Table 8b occur in Australia Johnson 1977, Tothill and Hacker 1983). fungus Bipolaris sorghieola (Lefebvre and (Sampson and Walker 1982, Woodcock and The plant is a native of Mediterranean re Sherwin) Alcorn (= Helminlhosporiwn sor Clarke 1983, Cook and Dube 1989). How gions of Eurasia. The related S. venicillijlo ghicola) causes lesions on the leaves of S. ever, of the two smutsEntyloma holci (Lim) rum (Steud.) Stapf., an African species, is a halepense. After confirming efficacy in the Fisch. is now regarded as a synonym of E. widespread weed in tropical Australia (KJein greenhouse, field testing with mass produced dactylidis (Pass.) Ciferri, which does occur in schmidt and Johnson 1977, Tothill and spores was undertaken. Results varied from Australia. Tilletia holei (Westerdorp) De Hacker 1983). Both are closely related to 73 % kill of seedlings of the grass to as low as Toni is not certainly a synonym of another grain sorghum, Sorghum bicolor with which 1% between 2 years in North Carolina and grass smut. Thus all insects and one of the S. halepense hybridizes to form Columbus 44% and zero kill with heavy leaf damage in smut fungi listed could be considered as po grass Sorghum x abnum Parodi and also to Mississippi (Van Dyke and Winder 1985, tential classical agents. the pasture sorghums, Sorghum sudanese Winder 1989). B. sorghieola occurs on tbe (Piper) Stapf. and hybrids (AUld and Medd crop S. bieolor in Australia (Alcorn and May Sorghum halepense (L.) Pers. 1987, Burbidge 1984). At the tribal level, ers 1975). Jobnson grass is a major weed of crops and both are related to sugar cane Sacchamm Phytotoxins have been isolated from the roadsides in all mainland States of Australia offieinanlln being inCluded in the group of leaf spot fungi of S. halepense, Dreehslera and is also poisonous to stock. It is particu- "awned" Andropogoneae (Table 1). sorghieola (Lefebvre and Sherwin) (= B. sor- 70 Planl Protection Quarterly VO!.5(2) 1990 ghicola) and Bipolaris sp. (Sugawara el aL Metacrambus caraetellus Zell. whose larvae necrosis. The strain of C. [unatus from 1987, Pena-Rodriguez el aL 1988) as a prel feed on and damage the rhizome of the plant E. crus-galli did not cause necrosis symptoms ude to their consideration as natural herbi (Gerling and Kugler 1973). However, there on the dicotyledonous crops tomato and cides. Except for maize which was affected, was no examination of the specifiCity of this bean, Lycopersicon esculentum Miller and these toxins have yet to be tested against a moth. Phaseolus vulgaris, L. and only produced series of crop grasses. In Pakistan, an elachistid CosmiOles sp. Dr. minor necrosis on older leaves of the grass The smut Space/Olheca hold Jack. which is illectella was found mining the leaves and crops, barley, maize, rye and wheat even considered by some to be a form of Space- four borers were found in the stems of S. when combined with atrazine in tbe case of 10lheca cmenta (Kuhn.) Potter did not infest halepense, but only two of these, the chloro maize (Schccpens 1987). cultivars of S. bieolor or S. sudanense but pid fli es ScoLiopthalmlis micans Lamb and C. lunatus as its imperfect stage Cwvularia readily infested S. halepense. Despite its spe Polyodaspis sp., were not known as crop lunala occurs on Echinochloa spp. in Austra cific name it does not infest Holcus spp .. The pests. Preliminary examination of the speci lia but it also occurs here on a large number smut is systemic and plants remain infected ficity of first instar larvae of C. sp. nr. illectella of other grasses including the crop grasses, from year to year and as the growth of S. showed they would not feed on wheat, rice, 0. sativa, grain sorghum, S. bic%r, halepense is adversely affected it loses its T. aeslivum and sugar cane, S. officinarum wheat, T aestivwn and maize, Z. mays (Old. competitive advantage in crops. Smut in but developed to adults on S. halepense. lar DPI unpub!'). fested plants do not set seed. Smut spores vae of the phycitine moth, Palna rhizolineata from California and Louisiana readily in Bradley were found boring in the rhizome Eleusine indica (L.) Gaertn. fested plants under glasshouse and field con and its specificity to S. halepense also re Crowsfoot grass is a weed of crops and pas ditions (Milhollon 1985, Massien and Lindow ceived a preliminary examination. Larvae of tures in tropical and subtropical Australia 1986). S. holci does not appear to occur in the moth did not survive on maize, Zea mays, and is also a st.ock poisoner (AUld and Medd Australia (Simmonds 1966, Shivas 1989, Qld grain sorghum S. bicolor or oats A. Saliva. 1987, Burbidge 1984, Kleinschmidt and DPI unpub!') and if its speCificity is con However, survival on S. halepense itself was Johnson 1977, Tothill and Hacker 1983). It is firmed it could be considered for introduc only 10% (Baloch el al. 1978, Khan el aL probably of Asian origin (Holm el al. 1977). tion here to control S. halepense. 1978,1980, 1981). AsS. halepense is not na A closely related species Eleusine trislachya, As well as studying a form of S. cruenta, tive to Pak.istan, the low survival of larvae on (Lam.) Lam., a minor weed in Australia, is which has the same specificity and host dam this grass weed suggests that it is not its nor probably also a naturalized exotic (Auld and aging effects as S. holci, from S. halepense in mal host but that the moth normally main Medd 1987, Burbidge 1984). If this is so, LouiSiana, EI-Wakil el aL (1985), also col tains itself on a native Pakistan grass. A rhi there is no Eleusine species native to lected and demonstrated the pathogenicity of zome-infesting insect specific to Sorghum Australia. E. indica is closely related to finger a Colletotrichum and a Phyllosticta species of spp. would not pose a threat in Australia to millet Eleusine comcana (L.) Gaertn., a fungi from the same grass weed. Strains of the annual grain and fodder sorghums S. staple food in Africa and India (Purseglove tbe anthracnose fungus Collelotrichum bicolor and S. sudanese. However, some of 1972), but this crop is not grown in Australia. graminico/a (Ces.) Wilson and the zonate the native Sorghum spp. are perennial, form £. indica is only distantly related to the prin leaf spot Gloeoeercospora sorghi Bain and ing tussocks (Tothill and Hacker 1983, Bur Cipal crop and pasture grasses here being Edg. have been isolated from S. halepense in bidge 1968, Lazarides 1970). placed in the subfamily Chloridoideae (Table Arkansas and tested for virulence on the 1). However, it is tribally related to the native weed and against Olher Sorghum spp. for Echinochloa spp. pasture grasses Eragroslis spp. (love grasses) specificity. G. sorghi was more virulent but Of the barnyard grasses, Eehinochloa ems and Dactyloelenium radulans (R. Br.) Slightly less specific tban the selected strain of galli (L.) Beauv. is a major weed of rice and Beauv. (button grass) and to 5 otber native C. graminicola which did not attack all culti other irrigated crops in Australia and is a na grass genera. It is also related to pasture and vars of the crops, S. bieolor. A mathematical tive of Europe and India. Echinochloa native Chloris spp. and lawn Cynodon spp. model of the biomass loss caused by these colonum (L.) Link is also a crop weed par and several other native grass genera in Wat two fungi when infesting S. halepense can be ticularly in subtropical and tropical Australia. son and DaJlwitz's (1985) broader classifica used to forecast the effect of inundative in It is of Asian origin. Five other introduced tion of the subfamily. oculation of these 2 fungi onto weed infesta Echinochloa spp. are regarded as occasional Despite the importance of E. indica as a tions (Mitcbell 1989). C. graminico/a occurs weeds as are a few of the several native Ech; weed in the tropics (Holm el aL 1977) at in Australia on S. halepense and otber Sor noehloa spp. These weeds are closely related tempts at biological control are very recent. ghum spp. and G. sorghi occurs on the crop S. to the fodder crops Echinochloa frumenta Figliola el al. (1988) investigated two fungi bicolor (Simmonds 1966, Shivas 1989). cea Link (Siberian millet) and E. uti/is Ohwi Bipolaris selariae (Saw.) Shoem. and Pyricu In Hungary, a strain of the bacterium and Yabuno (Japanese millet), both of which /aria grisea (Cke.) Saccoas possible bioherbi Pseudomonas syringae van Hall causes leaf are grown in Australia as fodder crops and cides in South Carolina. Both fungi have spot disease of S. halepense. Despite the some of the native species of the genus have broad host ranges in tbe Poaceae and broad bost plant spectrum of tbis bacterium, agronomic value (Burbidge 1984, Tothill and P. grisea occurs in Australia on E. indica (Qld this strain was considered to be specific to the Hacker 1983). Echinoehloa spp. are tribally DPI unpub!'). The only knowledge of other grass weed althougb only maize was tested related to the important pasture genera Digi possible agents are those recorded on the and remained uninfested by it (Mikulas and taria, Panicwn, Paspalum, Pennisetum, Se weed as an alternate host for grass crop dis Sule 1979). P. syringae is recorded from laria but the genus is only distantly related to eases and as pests or diseases of E. coracana other Sorghum spp. in Australia (SimmondS rice, Oryza Saliva (Table 1), the crop witbin (Ramalcrishnan 1%3, Holm el aL 1977, 1966, Shivas 1989). which they are such important weeds. Purseglove 1972). All bave broad bost ranges S. halepense bas been surveyed for poten In Europe, Coehliobolus lunatus (imper amongst the Poaceae and several of the dis tial anhroJXXI agents in Israel and Pakistan. fect stage Curvularia lunala (Wakker) ease fungi are present on the weed in Austra Unlike annual grain sorghum, S. bicolor and Boedijn which causes shoot and leaf necrosis lia (Qld DPI, unpub!'). Melanopsichium other Sorghum species occurring in Israel, S. of E. clUS-galli was found to be only weakly eleusinis (Kulk.) Mundk. and Thirum. (= ha/epense is a rhizomatous perennial. For pathogenic. However, treating the weed us USlilago eleusinis KUlk.) is a smut fungus this reason biological studies in Israel, within ing a suspension of spores of the fungus to which is only recorded from Eleusine and the original home range of S. halepense, have gether with the herbicide atrazine synergisti Dactylocleniwn spp. (Zundel 1953). This been concentrated on the crambid moth, cally increased the frequency and level of smut is not recorded from E. indica in Aus- Plant Protection Quarterly VOI.5(2) 1990 71 tralia (Simmonds 1966, Qld DP! unpubl.) Auld and Coote 1981 ,Auld elaL 1982, Vere all , except one species, with known alterna te despite oomment to the oontrary by and Campbell 1978a,b, 1979, Vere el a/. dicotyledonous aecidia l hosts and a stem Ramakrishnan (1963). However, it has ten 1981). smut Usrilago spp. tatively been recorded on the native grass The genus Nassella is closely related (0 the The above large number of apparently D. radulans in Queensland (Simmonds genus Slipo (Clayton and Renvoize 1986), specific fungi is based mainly on Oudemans 1966). numerous native Australian species of which (1919) and although synonymy and more are important components of unimproved recent knowledge of other hosts of many of Cynodon dactylon (L.) Pers pastures (Burbidge 1984). The rusts Puc the fungi would reduce this number consid Couch grass is a major weed in gardens, vine cinia and Uromyces spp. and the smut erably there would still be a larger number of yards and cultivations and an occasional Til/elio hypsophi/a Speg. infests grasses of specific fungal species in the Northern Hemi stock poisoner (Auld and Medd 1987). How both genera in their South American home sphere on this grass than in Australia. ever, it is also a major lawn grass in Australia range (Cummins 1971 , Zundel 1953). There None of the specific insects have been re (Burbidge 1984). Amongsl the crop and pas is no record of any rust or smut on oorded from this grass in Australia and only 4 ture grasses, it is tribally related only to N. trichotoma itself, nor is there any readily of the many specific or near specific fungi on Chloris (Table 1). available knowledge concerning the anhro this grass weed in the Northern Hemisphere In Florida, where C. dactylon is regarded pcxis infesting N. trichotoma in its home Old World are recorded in Australia. These as a major weed of lawns, the eriophyid gall range. No fungi or arthropcxis are recorded are:- Deightonie/la anmdinacea (Corda) mite Aeeria cynodoniensis Sayed has been from serrated tussock in Australia. Hughes, Helerosporium phragmitis Sacc., suggested as an agent because it causes Hadrarrichum phragmitis Fuckel and the stunting of the grass (Cromroy 1983). This Phragmiles australis (Cav.) Steud. rust, Puccinia magnusiana Korn (Simmonds mite occurs in Australia (Gibson 1967). Al Common reed is a widespread major weed of 1966 Sampson and Walker 1982, Woodcock though c. dacty/on is regarded by many au irrigation channels, drainage ditches and and Clarke 1983, Cook and Dub