Introduction

Resistance has evolved in almost every species of livestock pest that has been subjected to chemical control. It can be concluded from this that resistance is an inevitable consequence of chemotherapeutic control. Why, then, should so much reliance be placed on this form of treatment if the end result is failure to control the pest? Clearly, there are advantages. Chemical treatment has had a major impact on pests of agricultural and public health significance. Indeed, it is fair to say that the quality of life that we enjoy owes a considerable debt to the use of chemical pesticides. However, the continued success of these pesticides relies on maintaining susceptibility.

Due to the evolution of resistance (not to mention environmental consequences) there is not a direct relationship between the amount of chemical used and increased benefits. An over-reliance on chemical control hastens the rate by which resistance develops. Studying examples of resistance and the factors that influenced their rate of evolution, together with data from controlled experiments, can provide the information necessary to design management strategies that optimise the economic and social benefits from pesticide usage and limit the development of resistance.

There are a number of components that need to be considered in designing successful resistance management strategies:

1. Understanding the extent of the problem

2. The evolutionary basis of resistance

3. What strategies are available to prevent or manage resistance

4. Testing for resistance

5. How should recommendations be implemented

1. The problem

It is common to refer to a ‘resistance problem’. This is in fact incorrect. The problem is not resistance per Se, but rather, the pest we are trying to control - be they nematodes, insects or weeds. Resistance, however, becomes part of the problem because it frustrates control measures and limits the available arsenal to combat the primary problem. In economic terms, internal and external parasites have each been estimated to cost the sheep industry around $300 million annually. Similarly, estimates are no doubt available for other livestock pests and weeds. The economic significance of resistance has not been accurately assessed but undeniably becomes a component of the cost. Of equal concern, and also difficult to estimate, is the cultural cost. The emergence of multiple resistance and the lack of effective, non-chemical, alternative control measures may see landholders being forced to modify or abandon particular farming enterprises.

The extent of resistance has been detailed by previous speakers and need only be summarised here. Resistance to benzimidazole (BZ) and levamisole (LEV) anthelmintics in nematode parasites of sheep occurs on 70-90% of farms in the high rainfall areas of Australia. In these situations, ivermectin (IVM) is the only completely effective broad spectrum anthelmintic. Reports of IVM resistance occurring in other countries herald a warning to Australian farmers. The sheep blowfly has demonstrated resistance to organo-phosphorus compounds (e.g. diazinon) which is reflected as a reduction in the time sheep remain free of strike. Cyromazine (Vetrazin) however, still offers protection of up to 12 weeks. Sheep lice are showing resistance to the synthetic pyrethroids when used as a long wool, backline treatment. Cattle tick has become resistant to organo-phosphorus chemicals and resistance is emerging.

2. Development of resistance

Resistance is the result of the selection pressure applied by chemical treatment. To appreciate the selection process for resistance in nematode parasites, it must be realised that the life cycle consists of a parasitic stage in the host and a free-living stage on the pasture. Every womi must pass through the free-living stage and is acquired by the host as a separate infective larvae from the pasture. Selection for anthelmintic resistance occurs in the parasitic stages when anthelmintics are administered to the host animal. Anthelmintics are used at an efficiency of around 99% against susceptible strains. The small number of surviving worms, which are the most resistant component of the population, then contaminate the pasture with resistant offspring for subsequent generations. It has been shown that this mode of selection results in resistance being conferred by many genes (i.e. polygenic). In contrast, a persistent insecticide is usually applied at dose rates that are lethal to all genotypes. Selection occurs sonic time after treatment when concentrations decline to a level which favours resistant rather than susceptible individuals. In this instance, single gene resistance generally results.

The rate of development of resistance is influenced by many factors which can be classified as genetic, biological or operational. The most important are the operational factors because they can be manipulated by the farmer and form the bases of resistance management programs. However, it is necessary to understand the genetic and biological factors in order to arrive at the correct operational procedures.

3. Resistance management programs

Resistance management strategies can utilise chemicals in moderation, saturation or multiple attack and programs can incorporate aspects of different strategies.

(i) Control by moderation. Moderation is designed to limit selection pressure for resistance by reducing chemical treatment. Approaches include: low dose rates, increasing the proportion that escape chemical exposure, permitting a large population size before treatment and infrequent treatment. Moderation will ensure the preservation of susceptibility in the population but it can only be embraced where it provides satisfactory control.

For nematode parasites, low dose rates are inappropriate. To be sufficiently low to avoid selection, they would provide little or no control. Moderately low dose rates (70-95% kill) are also inappropriate because they are highly selective. Leaving animals untreated or permitted a large population size before treatment is also not recommended. Apart from the difficulty in identifying which animals to treat, strategic worm control utilises treatments prior to climatic conditions which favour the rapid development of infective larvae.

Infrequent treatment is, however, a highly desirable moderation approach to resistance management. Studies on the evolution of anthelmintic resistance have demonstrated a relationship between frequent treatment and the rapid development of resistance. Clearly, the more often a parasite population is exposed to a discriminatintg dose of anthelmintic, the greater the advantage to the resistant component of that population. Worm control programs such as ‘Wormkill’ or ‘Drenchplan’ rely on a few highly effective treatments.

(ii) Saturation. In essence, saturation means choosing a dose rate that is sufficiently high to kill heterozygotes - those individuals with resistance and susceptibility genes. When resistance gene frequency is low, it is extremely unlikely that an individual has only resistance genes (i.e. homozygous resistant). Therefore, removing the majority of heterozygotes will remove most of the resistance genes in that population. It must be emphasised that a saturation approach must be implemented before intensive selection results in a reasonable proportion of homozygous resistant individuals. If so. saturation will increase the level of resistance. An example of this comes from New Zealand. Underdosing was a feature of the initial treatment in a herd of milking goats which resulted in the development of resistance. When LEV failed to control Ostertagia, dose rates were increased up to 4 times that recommended. With hindsight it can be suggested that the initial low dose rate would have favoured the survival of heterozygotes and the onset of resistance. These heterozygotes, given the opportunity to interbreed, would produce homozygous resistant offspring. Once this was established, increasing the dose rates would have favoured those homozygous resistant individuals leading rapidly to a highly resistant population.

Within the constraints of the registered dose rates, worm control by saturation is achieved by knowing the weight of the heaviest few sheep in the mob and treating all animals at the appropriate dose volume for these animals.

A saturation approach will differ for persistent insecticides (e.g. organophosphorus dipping or jetting for blowfly or tick control). To maintain the chemical concentration above that lethal to heterozygote may require frequent applications. The use of synthetic pyrethroids as backline treatments against the sheep body louse (Damalinia ovis) results in a gradient of chemical concentration around the animal which provides a highly selective situation. A saturation approach cannot be achieved short of applying doses sufficiently high for the concentration on the entire surface of the animal to exceed that necessary to kill heterozygotes. The use of backline treatments, particularly on animals with long wool, has been subjected to considerable criticism because of the likelihood of resistance. It is clear from these examples that the consequences of pesticide concentration and decay characteristics are important considerations in designing resistance management programs.

(iii) Multiple attack. A program of multiple attack involves either combinations of chemicals given simultaneously or chemicals rotated sequentially over time. Both approaches should use chemicals with different modes of action (i.e. no shared cross resistance). They should also be introduced when resistance gene frequency to either chemical is low and high dose rates must be used (saturation approach). The principle underlying the success of the combination strategy is the high efficiency achieved because of the low probability that any individual has the genes for resistance to both chemicals. Rotation, on the other hand, relies on some reversion towards susceptibility of resistance to one compound while the other is being used.

Combinations. Laboratory studies have demonstrated the success of combinations of a BZ and LEV to prevent the onset of resistance. Separate lines of Trichostrongylus colubriformis (black scour worm) exposed to either a BZ or LEV anthelmintic alone developed resistance in a few generations while an equivalent line exposed to combinations of the recommended dose rate of a BZ plus LEV or a BZ plus IVM remained susceptible to all compounds (Figure 1).

While demonstrating the advantages of a combination to maintain susceptibility, this study held little practical significance because high levels of resistance to BZ and LEV occur in the field. Extending the study to the field showed that a combination of a BZ plus LEV provided high efficiency against Ostertagia (brown stomach worm) and T. colubriformis strains that were resistant to either broad spectrum when used alone. Following these and other studies, the pharmaceutical companies have marketed a number of commercial BZ plus LEV products.

Figure 1. Changes in LC₅₀ for benzimidazole resistance in Trichostrongylus colubriformis following selection with a:

BZ (— — — ),

BZ/LEV rotated each generation (———),

BVZ+LEV combination (— —— —) or

BZ+IVM (-------)

Combinations of anthelmintics are not limited to drugs effective against the same species of parasite. The Wormkill program introduced in the New England and Hunter regions of NSW involves the use of a combination of a broad spectrum anthelmintic and the narrow spectrum compound closantel. Closantel kills Haemonchus contortus (barber’s pole worm), the main parasite in this region. The broad spectrum anthelmintic is used to kill Ostertagia and Trichostrongylus which also occur in this region. The effectiveness of closantel against H. contortus has reduced the need for fequent treatment with broad spectrum anthelmintics thereby reducing selection for resistance. The particular anthelmintics used in combination in Wormkill give it some features that differ from a multiple attack program involving a combination of drugs aimed at the same species. While H.contortus may be controlled by both drugs, Trichostrongylus and Ostertagia are only controlled by the broad spectrum anthelmintic. Nevertheless, Wormkill serves as an example of combining anthelmintics for effective worm control and its benefits have been demonstrated. The long term success of Wormkill relies on continued effectiveness of closantel. Some emerging resistance is evident which threatens the program because it reduces the four-week protection period afforded by this compound.

Rotation. The concept of annual anthelmintic rotation has become pan of the parasitological folklore. Despite little experimental evidence to support the concept or indicate the best time to change anthelmintics, it has been widely accepted as a ‘best bet’ strategy. It must be recognised that alternative rotations in time and space are possible and their effects on resistance may be different.

The arguments favouring annual rotation are based on assumptions about the relative fitness of resistant and susceptible worms. In mainly susceptible populations, those few resistant worms are thought to have a lower fitness than the susceptible ones. In the presence of the anthelmintic, however, the most fit worms are those expressing resistance to that particular drug. If anthelmintics are rotated, there is an alternation of the most fit worms depending on the presence or absence of each anthelmintic. Implicit in this argument is the absence of selection of modifier genes which increases the fitness of the resistant worms. Intuitively it may seem better to change the anthelmintic each treatment to minimise selection for resistance. However, some studies have shown that this may not optimise reversion towards susceptibility which seems to occur during seasons unfavourable for larval survival. Therefore, an annual rotation is probably a better proposition.

Of particular interest is the extent to which a combination and rotation strategy can be incorporated into the one program and the sustainability of such a program. Tabulated below (Table 1) are four resistance situations that can exist within a population of worms, the effectiveness of the various anthelmintics and the possible rotation program available. There is a subtle difference between dual and multiple resistance. Dual resistance indicates a population of worms that contains individuals with resistance to either BZ or LEV. Multiple resistance indicates a population with individuals that possess the genes for resistance both BZ and LEV.

Table 1. Suitable anthelmintics and possible annual rotations over 1 to 4 years given various resistance scenarios. B = benzimidazole, L = levamisole, C = benzimidazole + levamisole combination and I = ivermectin

Resistance Status	Suitable anthelmintics				Annual Rotation (years)
	B	L	C	I	1	2	3	4
No resistance	+	+	+	+	B	L	C	I
Benzimidazole resistance	-	+	+	+	L	C	I
Levamisole resistance	+	-	+	+	B	C	I
Dual resistance	-	-	+	+	C	I
Multiple resistance	-	-	-	+	I

Potentially, a program can utilise all effective anthelmintics in an annual rotation. indeed, this is in line with the recommendation to use effective anthelmintics. However, the question is, given a situation where only BZ resistance exists, is it better to consider a two year rotation involving a combination and IVM, or a three year program including LEV? While not universally accepted by my colleagues, my bias is to the former. The logic of my reason is as follows. The frequency of multiple resistance is related to the frequency of dual resistance. Hence, susceptibility to the combination can be reduced by resistance developing to the combination or to each of the individual compounds. Population genetic theory indicates that resistance will develop more slowly to the combination than to one compound. Furthermore, if a program involving rotation of all three compounds is accepted, it is difficult not to use LEY, in either the combination or individually, over two successive years and so limit the opportunity for reversion of levamisole resistance.

During the preparation of this paper, I was encouraged by results supplied by my colleagues, Robert Dobson and Elizabeth Barnes from CSIRO, McMaster Laboratory. They demonstrated from a computer simulation model that, if three treatments are given each year, the development of multiple resistance to BZ and LEV could be delayed using a rotation program of a BZ + LEV combination one year and IVM the next. Furthermore, the development of IVM resistance was also delayed by this anthelmintic regime.

Anthelmintic efficiency. In general, the greater the anthelmintic efficiency, the slower resistance will develop. This occurs because the resistant worms leave very few offspring compared to the large population on pasture. Anthelmintics are most effective in reducing parasite numbers if given when harsh environmental conditions limit the development of eggs and larvae on pasture. For example, summer drenching is a recommended procedure for parasite control in the winter rainfall (southern) areas of Australia and forms the basis of the strategic worm control programs in this area. During the hot dry summers characteristic of this area there are few larvae available to grazing sheep. Anthelmintic treatment at this time will remove worm burdens from sheep and reduce subsequent contamination of pasture with worm eggs. Worm numbers increase to become a problem during the wet autumn to spring period. The sources of these worms are from eggs passed in the late summer and from eggs deposited during the late spring/early summer period. The relative contribution from the two sources to worms present after autumn is dependent upon climatic conditions, egg deposition rates and the efficiency of the anthelmintic treatments before and during summer. Highly efficient summer treatments will ensure that there is little contamination in the late summer. This will not only reduce the number of larvae available to grazing sheep in autumn, but also ensure that the greater proportion of them originate from eggs deposited in spring, prior to the summer anthelmintic treatments.

Strategic worm control in the summer rainfall (northern) region is also based on ecological considerations. Larval number of H. contortus are at a minimum in late winter/early spring.Treatments in August and November with a broad spectrum anthelmintic and closantel forms the basis of the Wormkill program. In order for H.contortus to persist in the presence of closantel treatment, it must survive on pasture through the adverse winter until conditions favourable for its development and survival on pasture occur in December. In some areas of the Northern Tablelands it appears that Wormkill has eradicated H. contortus. However, in the nearby areas where the winters are less severe, and as a consequence larvae survive better, H.contortus persists albeit at low levels.

Controlled release. Controlled release capsules are now available from which a BZ anthelmintic is slowly released over 100 days. Questions can be asked about the extent to which this technology will select for resistance. The device is designed to release BZ at a constant rate and, provided this gives a plasma concentration of anthelmintic above that lethal to heterozygotes, resistance should not develop rapidly. However, the device contains a BZ to which resistance in a single oral dose is widespread. Some important facts have emerged from recent research. Firstly, the capsule is ineffective against resistant worms. Secondly, it kills some, but not all, incoming larvae and, finally, it reduces contamination of pastures with worm eggs.

Field trials have shown that the capsule can provide a highly effective control of nematodes in sheep despite the fact that the worms are resistance to a BZ administered orally. Therefore, it may well be a valuable resistance management tool mediated through its effect on population size.

Controlled release devices can be incorporated into strategic worm control programs. Currently only an adult capsule is available and recommendations favour one application each year during late summer in the winter rainfall region and to pre-lambing ewes in the summer rainfall region. A weaner device may soon be available, but it is likely that recommended times for treatment will differ from those for adult animals.

4. Testing for resistance

In any management program it is necessary to detect low levels of resistance. This is necessary firstly, because many management strategies are only effective if implemented when resistance gene frequencies are low and, secondly, to monitor the effectiveness of different management programs. The faecal egg count reduction test (FECRT) is the main means of diagnosing anthelmintic resistance in the field. There are, however, limitations to the FECRT because it relies on the phenotypic expression of resistance. It is known that worms which contain both resistance and susceptible genes are predominantly susceptible. Studies have indicated that a resistance is detectable. Unfortunately, at this frequency the rate of increase in resistance with further selection is very rapid. Other tests for BZ resistance, the egg hatch and tubulin binding assays, have a similar limitation on detecting resistance. They are also more suited to laboratory studies than field diagnosis. Once resistance is obvious from any of the currently available detection tests, implementing preventative strategies may be too late for maximum effectiveness of a resistance management program. Techniques for early detection of resistance have been highlighted as a priority area for research. Perhaps the molecular biological technologies becoming more commonplace today will provide the more sensitive tests needed for field (and laboratory) investigation in the future.

5. Implementation of resistance management

Strategic worm control programs have been implemented for most sheep grazing regions of Australia. These are based on the factors known to affect parasite control and the development of resistance. Particularly, the use of a few highly effective anthelmintic treatments administered at times designed to have maximum impact on the parasite population size. Management of the programs is through the State Departments which have adopted a responsible and coordinated approach. There was wide acceptance of the first program, Wormkill, introduced to the Northern Tablelands and Hunter regions of NSW in 1984. It is generally agreed that the rapid acceptance was due to three main factors. Firstly, a desperate situation faced by graziers because of resistance in H.contortus to the broad spectrum anthelmintics. Secondly, the availability of the narrow spectrum anthelmintic, closantel, with its unique property of prolonged action against incoming larvae. Thirdly, and perhaps most significantly, was the manner in which Wormkill was ‘marketed’ by a dedicated team from CSIRO, the NSW Department of Agriculture and the NSW Pastures Protection Board headed by Keith Dash and Betty Hall. The message reached the end user through advertising, promotional literature, local meetings with farmers, Departmental and private veterinarians and field officers and an extensive backup service via personal or telephone advice.

Subsequent programs have been introduced into other areas - Drenchplan (southern NSW), Wormplan (Vic), Wormcheck (SA), Wormbuster (Qld), CRACK (WA) and Weaner watch/Drenchplan (Tas). Again their success is dependent on adoption by the fanners through service from Departmental personnel. All programs need careful monitoring and modification for particular regions and, at times, particular farms. Worm control and resistance management are moving towards a ‘one on one’ situation where experienced veterinarians or field officers can make recommendations based on their knowledge of the local situation and the general factors about worm control and the development of resistance. Similar approaches are being invoked for control of other pests species.

Licekil1. Licekill has been introduced in NSW in response to the failure of synthetic pyrethroids to control lice in long wool sheep. The program offers advice on lice management and alternative effective chemicals.

Ear1y treatment for sheep blowfly (L. cuprina). Chemicals can be used to alter the population dynamics of L. cuprina rather than just to afford protection from the pest. Current insecticide usage usually results in sheep being treated when strike is sufficiently apparent in a mob. This may not occur until late spring or early summer thus allowing L. cuprina a generation or so for population expansion, perhaps in covert strikes, on an available resource. If sheep are treated in early spring this resource is removed for L. cuprina emerging after overwintering, thus limiting population increase. While subject to some qualification the results to date are encouraging and a program based on this approach is currently being advocated in

NSW.

Note: This paper is not referenced. Readers interested in the background literature are invited to consult ‘Resistance Management in Parasites of Sheep’, edited by J.A. McKenzie, P.J. Martin and J.H. Arundel. Australian Wool Corporation Publication, 1990. This contains a number of reviews and abstracts on aspects of resistance in internal and external parasites of sheep.