Introduction

Herbicides and other pesticides are major technological tools that are used successfully throughout the world. However, an adverse consequence of persistant application has been the emergence of populations resistant to pesticides. Resistance to pesticides is a global phenomenon that has occurred with fungicides, bactericides, insecticides, rodenticides, nematicides and herbicides (6). Since the first report in 1970 (21) there have been many reports of herbicide resistance. Most have appeared in the northern hemisphere and involve resistance to the triazine herbicides in areas where these herbicides have been extensively used (11). Resistance also has developed to a diverse range of other herbicides (II). It is apparent that repeated use of the same, or chemically similar herbicides, will lead to the development of resistant weed populations.

Herbicide resistance has been confirmed in five weed species in Australia. All of these species are important and are widespread weeds of crops and pastures throughout southern Australia. Considerable research is underway on these resistant weeds. Of greatest practical and scientific concern is the appearance of biotypes of the grass weed, annual ryegrass, with cross resistance to a range of herbicides that are chemically dissimilar and have different modes of action. A cross-resistant weed biotype is defined as a biotype which has developed resistance after selection from one herbicide and which then exhibits resistance to herbicides which differ chemically and which have different modes of action. This review will discuss the research on herbicide resistant weed populations in Australia.

Cross resistance in biotypes of Loliurn rigidum (annual ryegrass) selected in crops and pastures.

Occurrence. Annual ryegrass is ubiquitous throughout the cropping zones of southern Australia and is possibly the most important weed of Australian crops. Before the advent of selective post-emergent herbicides a combination of practices was used for annual ryegrass control (20). Since registration in 1978 the aryloxphenoxypropionate herbicide, diclofopmethyl, has been widely and successfully used for post-emergent control of annual ryegrass in cereal and legume crops. Since 1982 the sulfonylurea herbicide chlorsulfuron, which selectively controls annual ryegrass in cereal crops, has become widely used. In 1989 the sulfonylurea herbicide, triasulfuron, was introduced for selective control of annual ryegrass in cereal crops. These three herbicides are applied on vast areas in southern Australian cropping regions for the control of annual ryegrass and other weeds.

In 1982 an infestation of annual ryegrass in the Bordertown region of South Australia was reported as resistant to diclofop-methyl (7). Subsequent studies with biotypes from this area (4,8) revealed cross resistance to other aryloxyphenoxypropionate herbicides (fluazifop-butyl, haloxyfop, quizalafop) to cyclohexanedione herbicides (sethoxydim, tralkoxydim) to sulfonylurca herbicides (chlorsulfuron, etsulfuron, triasulfuron) to imidazolinone herbicides and to the dinitroaniline herbicide trifluralin. Studies show that cross-resistant biotypes of annual ryegrass are present in all mainland states of southern Australia and that biotypes selected in different areas can exhibit different degrees and spectrum of cross resistance. Table 1 lists the herbicides to which annual ryegrass biotypes have been found to be resistant. Farmers report that upon failure of diclofop-methyl, a switch to one or more of the alternative herbicides listed in Table 1 may or may not control the annual ryegrass. In some instances control is effected by herbicides such as sethoxydim or chlorsulfuron while in other instances cross resistance is evident to these same herbicides. Two to three years use of any of the herbicides listed in Table 1 may result in failure due to the further extension of cross resistance.

Control of the resistant biotypes. It is a practical reality that chemical control of the cross-resistant biotypes of annual ryegrass can be difficult. The options for control are limited as the list in Table 1 represents most of the herbicides registered to control annual ryegrass in crops. Depending upon the spectrum of cross resistance, control cannot be achieved by alternative selective herbicides. Research into viable strategies for the control of cross-resistant biotypes is being undertaken at the Waite Agricultural Research Institute and Department of Agriculture laboratories in Victoria and Western Australia. Clearly, integrated strategies to control cross-resistant annual ryegrass in crops and pastures are required. These will involve alternation of crops, pastures, judicious choice of herbicides and cultural operations. Guidelines prepared by researchers and a Herbicide Resistance Action Committee of AVCA were recently published (18).

Table 1. Herbicides against which Lolium rigidum (annual ryegrass) has exhibited cross-resistance following exposure to diclofop-methyl.

Aryloxyphenoxypropionates

Diclofop-methyl
Fluazifop-butyl
Fenoxaprop-ethyl
Haloxyfop-methyl
Quizalafop-ethyl

Cyclohexanediones

Alloxydim
Sethoxydim
Tralkoxydim

Sulfonylureas

Chlorsulfuron
Metsulfuron-methyl
Triasulfuron

Dinitroanilines

Trifluralin

Cross resistance in a biotype of Lolium rigidum (annual ryegrass) selected along railway lines

Occurrence. In Australia, as in other parts of the world, railway authorities have relied on mixtures of contact and residual herbicides for broad-spectrum weed control along railway lines. In Western Australia, 5000 km of railway line have been treated once annually for more than 10 years (to 1988) with an amitrole-atrazine mixture. Until recently this has provided excellent control of annual ryegrass. However, in recent years annual ryegrass has survived this treatment along extensive segments of Western Australian railway line. Experiments show that this biotype is resistant to the triazole herbicide amitrole, to pre- or post-emergent applications of triazine herbicide atrazine and there is cross resistance to a wide range of other triazine herbicides to which the population has never been exposed (3). A surprise finding was that cross resistance also extended to a wide range of urea herbicides (3).

Control of the resistant biotype. Strong selection pressure from the repeated applications of high rates of the same herbicides to railway lines provides classical conditions in which resistance can develop. Since 1988, the use of amitrole and atrazine for weed control on railway lines in Western Australia has been discontinued because of the failure to control annual ryegrass. It is noteworthy that much of the railway line infested by the cross-resistant annual ryegrass traverses crop lands containing large areas of lupins, Lupinus spp, in which simazine is almost universally applied. No cases are known of ingression of the cross-resistant biotype from railway lines into adjacent lupin fields.

Following the appearance of cross resistance in annual ryegrass along railway lines, the railway authorities switched from amitrole and atrazine to an alternative herbicide combination of glyphosate and sulfometuron-methyl. Glyphosate and sulfometuronmethyl have different and specific sites of action and both herbicides control the amitrole and atrazine resistant biotype (3). However, as annual ryegrass has already developed resistance to a mixture of the dissimilar herbicides amitrole and atrazine, it is possible that if glyphosate and sulfometuron-methyl are used repeatedly in the same manner then resistance to these herbicides will also develop. Recognition of this maxim is not generally appreciated at this time.

Diclofop-methyl resistance in Avena fatua (wild oat)

Occurrence. The aryloxyphenoxypropionate herbicide diclofop-methyl is successfully used for Avena fatua control in cereal and other crops in many parts of the world, including Australia. In 1985, a farmer at York, Western Australia, obtained poor control with diclofop-methyl and subsequent field and pot experiments confirmed that recommended rates of 563 g a.ilha had little effect on this population. Field control of the diclofop-methyl resistant Avena fatua has been obtained with the aryloxyphenoxypropionate herbicide, fluazifop-butyl, and the cyclohexanedione herbicide, tralkoxydim.

Diclofop-methyl resistant Avena fatua populations have since been identified near to the original resistant site at York and at Spencers Brook in Western Australia, in central and southern New South Wales and at 3 sites in the north-east of Victoria and in some sites in South Australia. Given the widespread usage of diclofop-methyl in Australia it is likely that many further cases of diclofop-methyl resistance will appear.

Control of the resistant biotype. Resistance to diclofop-methyl in Avena fatua is of considerable importance as this species is a serious weed of cereal cropping in many countries, including Australia. Alternative herbicides to diclofop-methyl are registered for the post-emergent control of Avena fatua in cereals (fenoxaprop, tralkoxydim) and broadleaf crops (various aryloxyphenoxypropionate and cyclohexanedione herbicides). If there is exclusive and persistent reliance on these alternative herbicides then a very strong selection pressure will be applied to the diclofop-methyl resistant populations and the potential will exist for the development of cross resistance as has been previously observed in annual ryegrass (see below discussion of resistance in Avena sterilis).

Aryloxyphenoxypropionate resistance in A vena sterilis (wild oat)

Occurrence. In 1989 a farmer near Bordertown, South Australia, reported a failure of haloxyfop-methyl to control wild oats in the first year of treatment with haloxyfopmethyl. This population had previously been treated with diclofop-methyl and fluazifop-butyl. In pot experiments this population of Avena sterilis has shown no mortality at rates up to 416 g haloxyfop-methyl/hectare, whereas Avena sterilis collected from an adjacent field showed mortality at 26 g/hectare. This biotype is also resistant to other aryloxyphenoxypropionate herbicides such as diclofop-methyl, fluazifop-butyl, fenoxaprop-ethyl and quizalafop. This very recent development is obviously of substantial concern because of the importance of wild oat as a weed of crops and the appearance of resistance to a number of herbicides registered for wild oat control.

Paraquat and diquat resistance in Hordeum glaucum, Hordeum leporinum (both known as barley grass) and Arciotheca calendula (capeweed)

Occurrence. A biotype of the important annual weed known as barley grass (Hordium glaucum) infesting a lucerne field near Ararat in the Western District of Victoria, was found to be resistant to paraquat in 1982 (28). This resistant biotype is 250 times less sensitive to paraquat than normal Hordeum glaucum (14). The resistant biotype was dominant in a lucerne field which had been sprayed annually since 1969 but was not detected in untreated adjoining fields. A survey of lucerne fields with a history of paraquat/diquat usage was conducted throughout Victoria and South Australia in 1985 and 1986 and revealed resistant 1-lordeum glaucum was present on nine lucerne fields on four separate farms within the Ararat area (24). There was no evidence of spread of the resistant biotype from one original site; however, the survey did reveal that in some cases the resistant biotype had been introduced to uninfested areas by hay or stock from an infested site (24).

A biotype of the important and widespread annual dicot, Aretotheca calendula (capeweed), infesting a lucerne field at Elmhurst in the Ararat area is resistant to paraquat/diquat (16). In addition to capeweed this field is infested with the paraquate/diquat resistant Hordeum glaucum (24), and recent studies show that this field also contains a resistant biotype of the polyploid grass Hordeum leporinum, also termed barley grass (25). Therefore, this field hosts three weed species resistant to paraquat/diquat. This field was sown to lucerne in 1958 and had been sprayed once annually with paraquat and/or diquat from 1963 to 1986. Throughout this period the field received a total of 3.6 kg ai/ha paraquat and 3.9 kg ailha diquat. The repeated use of paraquat/diquat for 24 years without cultivation or other herbicides has provided the selection pressure allowing resistant Hordeum glaucum, capeweed, and Hordewn leporinum biotypes to dominate. Several other weed species in this field remain susceptible to paraquat and diquat.

Recently, we have documented that there are two distinct fields in Tasmania which contain paraquat and diquat resistant Hordeum leporinum. Both of these are lucerne fields and have a long history of paraquat/diquat use.

Control of the resistant biotypes. Resistant Hordeum glaucum has only appeared on a few lucerne fields with a history of paraquat/diquat application (24). The resistant capeweed has only been found on one lucerne field. Resistant Hordeurn leporinum has only been found on one field in Victoria and two fields in Tasmania. These paraquat/diquat resistant biotypes have not been found in any other crops. Farmers have controlled the resistant biotypes with alternative herbicides as the resistance is specific for paraquat and diquat. Both non-selective knockdown control and in-crop control of the resistant biotypes have been achieved, although control of the resistant biotypes has involved extra cost.

Tactics For The Prevention Or Management Of Herbicide Resistance

1. Tactics to overcome herbicide resistance when resistance has developed

Although there have been many cases of herbicide resistance throughout the world there has been little need for concerted control strategies because control of the resistant biotypes can usually be obtained by alternative herbicides. Management factors, especially the choice and combination of herbicides, have changed in response to the appearance of resistance. Triazine resistant weeds in the northern hemisphere exhibit little cross-resistance and have been controlled by alternative herbicides or a combination of triazine herbicide and a herbicide with a different mode of action. This has also been the case in Australia with the Avena

fatua resistant to diclofop-methyl and the weeds that are resistant to paraquat/diquat. Control has simply been acheived by switching to alternative herbicides. Thus the agricultural impact of the resistant weed biotypes has been limited as alternative herbicides are readily available, albeit sometimes at greater cost. It should be stressed that continued reliance on a single alternative herbicide can result in the development of resistance to the replacement herbicides.

Cross resistance to chemically dissimilar herbicides such as that evident in annual ryegrass biotypes in Australia (Table 1) means that cross-resistant biotypes cannot be controlled simply by changing herbicides. Practical control of cross-resistant biotypes in crops and pastures can be a problem. The pattern of resistance to aryloxyphenoxypropionate herbicides in a biotype of Avena sterilis may indicate that this species is exhibiting the same phenomenon. Perusal of the insecticide resistance literature reveals that many insect species exhibit cross resistance and that their control by chemicals means is difficult (6).

The management of herbicide cross-resistant annual ryegrass requires an integrated weed management strategy (cultural practices together with crop, pasture and herbicide rotation). Integrated weed management techniques have long been used to control recalcitrant weed species throughout the world (although the control measures adopted are rarely defined as integrated weed management). An example in Australia is the control of Vulpia fasciculata (silvergrass) for which there are few selective herbicide options ( Vulpia is not controlled by the aryloxyphenoxypropionate or the cyclohexanedione herbicides). Control is achieved by a number of methods including cultivation, pasture management, spray topping, crop rotations and use of trifluralin and simazine. Farmers controlling Vulpia fasciculata by these techniques are practising integrated weed management, although they may not define it in this way. The appearance of herbicide cross-resistant ryegrass (or other weed species) will require similar consideration. We are fortunate that, in addition to herbicides, there are a number of other weed management tools which can be used to control herbicide cross-resistant weeds.

2. Tactics to minimise the development of herbicide resistance

It must be recognised that the great majority of Australian farmers have always practised integrated weed management for the control of weeds. While farmers and advisers may not have used this term to describe the use of cultivation, the grazing animal, pasture-topping, burning, etc., there is no doubt that these practices, together with pasture and crop rotation, constitute an integrated weed management system. By the use of these varied weed control practices, farmers have (unconsciously) acted to avoid or greatly delay the emergence of herbicide resistant weed biotypes. Australian farmers should be encouraged to maintain this diversity of operations in the control of weeds. The converse is represented by those relatively small numbers of Australian farmers who rely heavily (or exclusively) on the use of selective and non-selective herbicides for weed control and who crop intensively or continuously on some fields. There is no doubt that it is these farmers who are at greatest risk of developing herbicide resistance. Our observations clearly identify that farmers practising continuous cropping, or intensive cropping, run a much greater risk of developing resistance. Under such conditions, a consistent selection pressure is placed on the weed population and herbicide resistance is the inevitable result. Farmers practising such cropping regimes need to be alerted to the probability of resistance appearing under these conditions and should consider modifications to their cropping practices.

3. A need for a change in emphasis by industry and advisers: Advocation of strategies which will minimise the likelihood of herbicide resistance

A great challenge to weed scientists and to public and private sector advisers in Australia and elsewhere will be the development of herbicide management strategies which minimise the rate of development of resistance. Currently, herbicides are promoted and used by all sectors of the industry with scanty recognition that their persistent use will lead to resistance. There are virtually no recommendations that herbicide types be rotated or used sparingly so as to minimise the likelihood of resistance. Recognition of the biological inevitability of resistance is, in our view, currently lacking in the marketing strategies adopted by many producers of herbicides, as it is in many other sectors of the weed control community. There is a need for a considerable change in attitude among herbicide producers, herbicide advisers and herbicide users if we are to maintain the arsenal of herbicides which currently is available. It is incumbent on weed scientists and advisers in the public and the private sector and farmers to recognise the threat of herbicide resistance and to investigate and promote strategies to minimise the development of resistance. We are convinced that with a modicum of creativity it is possible to manage herbicide usage so as to minimise herbicide resistance and thereby maintain the range of herbicides available to agriculture in this country and elsewhere.

Literature Cited

1. Arntzen, C.J., Pfister, K. and Steinback, K.D. (1982). In: Herbicide resistance in plants. Eds. H.M. LeBaron and J. Gressel. John Wiley and Sons, New York. pp.185-214.

2. Bishop, T., Powles, S.B. and Comic, G. (1987). Aust J.Plant Physiol. 14:539.

3. Burnet, M.B., Heldebrand, O.B., Holtum, J.A.M. and Powles, S.B. Weed Science (submitted).

4. Christopher, J.C., Powles, S.B., Holtum, J.A.M. and Liljegren, D. (1990) Plant Physiol. (submitted).

5. Fuerst, E.P. and Vaughn, K.C. (1990). Weed Tech. (in press).

6. Georghiou, G.P. (1986). In: Pesticide Resistance: Strategies and Tactics for Management. Nat.Acad.Press: Washington, pp. 14-43.

7. Heap, J. and Knight, R. (1982). Aust.J.Agric.Res. 48:56.

8. Heap, J. and Knight, R. (1986). Aust.J.Agric.Res. 37:149

9. Holtum, J.A.M. Plant Physiol. (submitted)

10. Islam, A.K.M.R. and Powles, S.B. (1988). Weed Res 28:393

11. LeBaron, H.M. and McFarland, J. (1990). In: A.C.S. Symposium Series 421. Managing resistance to agrochemicals: From fundamental research to practical strategies. Eds. M.B. Green, H.M. LeBaron and W.K. Moberg. Am.Chem. Soc.,Washington.

12. Matthews, J.M., Holtum, J.A.M., Liljegren, D. and Powles, S.B. (1990). Plant Physiol. (in press).

13. Moss, S.R. and Cussans, G.W. (1987). In: Combatting resistance to xenotiotics: Biological and chemical approaches. Ed. M.G. Ford. pp.200-213.

14. Powles, S.B. (1986). Weed Res. 26:167.

15. Powles, S.B. and Comic, G. (1987). Aust,J.Plant Physiol. 14:8 1.

16. Powles, S.B., Tucker, E.S. and Morgan,T.W. (1989). Weed Sci. 31:60.

17. Powles, S.B., Matthews, J.M., Holtum, J.A.M. and Liljegren, D. (1990). In: A.C.S. Symposium Series 421. Managing resistance to agrochemicals: From fundamental research to practical strategies. Eds. M.B. Green, H.M. LeBaron and W.K. Moberg. Am.Chem.Soc., Washington, pp 394-406.

18. Powles, S.B. and Howat, P. (1990). Weed Tech 4:l78.

19. Ray, T.B. (1984). Plant Physiol. 7.5:827.

20. Reeves, T. (1976) Weed Res .16:57.

21. Ryan, G.F. (1970). Weed Sci. .18:614.

22. Secor, J. and Cseke, C. (1988). Plant Physiol. 86:10.

23. Thill, D.C., Mallory, C.A., Saari, L.L. and Cotterman, J.C. (1989). WSSA Abstracts 29: 132.

24. Tucker, E.S. and Powles, S.B. (1988). Plant Prot. Quart. 3:l9.

25. Tucker, E.S. and Powles, S.B. (1990). Weed Sci. (in press).

26. Tucker, E.S. and Powles, S.B. (1990). Weed Res. (in press).

27. Tucker, E.S. and Powles, S.B. (1990). Weed Res. (in press).

28. Warner, R.B. and Mackie, W.B.C. (1983). Aust.Weed Res.Newsl. 3 1:16.

29. Yaacoby, T., Schonfeld, M. and Rubin, B. (1986). Weed Sci. 3.4:181.