Implications
Breeding tropical sheep or goats for enhanced resistance to nematode parasites should lead to sustained improvements in animal health and performance. It is recommended that selection is based on phenotypic measurements such as performance under conditions of nematode challenge, faecal egg count and measures of anaemia such as packed cell volume or Famacha© score. With currently available technologies, it is unlikely that genetic markers will make a large contribution to such breeding programmes.
Introduction: context for breeding for nematode resistance
Tropical livestock production systems, particularly those for small ruminants, are diverse and influenced substantially by both genetic and environmental factors, and their interaction. Within such production systems, animal adaptation to the environmental circumstance is one of the key factors affecting the success of the production system. Identifiable factors impacting on animal adaptation include heat tolerance (Silanikove, Reference Silanikove2000; Thornton et al., Reference Thornton, van de Steeg, Notenbaert and Herrero2009), ability to cope with often poor-quality forages and ability to cope with numerous endemic disease challenges. This paper addresses the ability of sheep and goats to cope with disease, specifically disease caused by gastrointestinal nematode parasites, the most important of which is the strongyle parasite Haemonchus contortus.
Disease has major impacts on tropical livestock production systems, including production losses, uncertain family food security and loss of income, and some diseases and the strategies used to manage them directly impact human health. In market-oriented terms, the costs of disease are estimated as 35% to 50% of turnover within the livestock sector in the developing world. All animal production systems are subjected to infectious disease and the consequent production losses. Disease also poses threats across species barriers. Firstly, several animal infections pose zoonotic threats to human health, and diseases in one species may act as reservoirs for infections in other species. Furthermore, uncertainties in terms of threats to food security and instability of livestock production systems are a further impact of animal disease.
Options for the management of infectious disease include vaccination, chemotherapy, improved management including better nutrition, diagnosis and removal of infected animals, and genetic change of the host. Each of these options can contribute to solving the problems caused by disease; however, often none on their own is sufficient to control the disease problem. Furthermore, disease management options may interact; the use of one option may enhance or diminish the effectiveness of another. A detailed assessment of the interaction of various disease control measures for the case of nematode infections in sheep is given by Jackson et al. (Reference Jackson, Bartley, Bartley and Kenyon2009).
Disease management may also break down, for example, the infectious pathogen or parasite may evolve to avoid the control measure. Strategies based on a single approach are the most vulnerable, whereas those based on multiple approaches are inherently more resilient and stable. For example, integrated pest management, a strategy that aims to combine several approaches, is well established in the management of plant diseases with the concepts easily transferable to livestock (Campbell, Reference Campbell1994). In the context of nematode infections, anthelmintic resistance, the evolution of parasites to become resistant to the chemicals used for their control, is a major concern worldwide (Waller, Reference Waller2003).
This paper focuses on gastrointestinal nematode parasitism, because of its ubiquitous importance to grazing ruminants. In a study considering diseases impacting upon livestock producers in developing countries, in which diseases were ranked on their economic impact, human health impact, animal welfare and implications on international trade, Perry et al. (Reference Perry, McDermott, Randolph, Sones and Thornton2002) concluded that helminthosis (i.e. diseases including those caused by nematode parasites) was the most important disease in sheep and goat production systems. In fact, from Perry's ranking system, helminthosis was, overall, the most important livestock disease in developing countries.
This paper discusses options for breeding small ruminants, sheep and goats, for resistance to gastrointestinal nematode parasites, specifically strongyle parasites such as H. contortus. A precise definition of resistance is given below. Evidence for genetic variation in resistance to H. contortus and other nematode parasites will be briefly summarised, options for genetic improvement of resistance will be considered and issues such as potential impacts on animal performance and the likelihood of parasite evolution will be discussed. Typically, these small ruminants will be kept in small-holder production systems, in which security of production is critical to sustainability. It is also likely that many of these communities will have weak infrastructure, limiting the sophistication of the breeding programmes that can be implemented.
Evidence for genetic variation in nematode resistance
The term disease resistance is often used loosely and generically to cover both resistance to infection as well as resistance to the disease consequences of infection, that is, disease tolerance. Following the terminology of Bishop and Stear (Reference Bishop and Stear2003), resistance to infection essentially describes the host's ability to interact with and control the lifecycle of the parasite. In the context of nematode infections, this may include the establishment probabilities of ingested larvae, rate and extent of parasite development within the host, parasite mortality and parasite fecundity – and hence faecal egg count (FEC). It is important to note that resistance is a relative term, and complete resistance to nematode infection is unlikely to be ever seen. Disease tolerance is often used to describe a host's ability to withstand pathogenic effects of infection. Resilience is related to tolerance, and describes an animal's ability to maintain performance in the face of a disease challenge.
The ability to breed animals for improved nematode resistance is dependent upon the existence of genetic variation between animals in their resistance to (or tolerance of) such infections and the ability of the breeder to identify and breed from the more resistant animals. Genetic variation in nematode resistance at both the between-breed and within-breed levels has been summarised by Bishop and Morris (Reference Bishop and Morris2007). In their review, these authors concluded that breed differences in resistance to nematode infections have been well documented, particularly for tropical or sub-tropical sheep facing H. contortus challenge. Although many of the published breed comparison studies lack power, a consensus between studies has emerged. For example, in sheep there is ample evidence for the Barbados Blackbelly, St Croix, Florida Native and Gulf Coast Native breeds from the Caribbean and southern United States being relatively resistant, when compared with non-adapted breeds (Baker and Gray, Reference Baker and Gray2004). In India, the Garole breed appears to be relatively resistant compared with other breeds (Nimbkar et al., Reference Nimbkar, Ghalasi, Swan, Walkden-Brown and Kahn2003). In Africa, considerable evidence has demonstrated the favourable resistance and tolerance characteristics of the Red Maasai breed (Baker and Gray, Reference Baker and Gray2004), and this resistance translates into improved performance under many environmental conditions (Baker et al., Reference Baker, Mugambi, Audho, Carles and Thorpe2004). Similarly, in goats the West African Dwarf goat shows both trypanotolerance and resistance to nematode infections (Chiejina and Behnke, Reference Chiejina and Behnke2011).
Most within-breed studies of genetic resistance use FEC as the indicator trait for resistance, and significant heritabilities are invariably found, coupled with extensive between-animal variation in FEC (Bishop and Morris, Reference Bishop and Morris2007). Most heritabilities for FEC in sheep are generally in the range from 0.2 to 0.4, although values in goats tend to be slightly lower, for example, from 0.1 to 0.35. Importantly, resistance to different strongyle parasites appears to be strongly genetically correlated, perhaps approaching unity (Gruner et al., Reference Gruner, Bouix and Brunel2004), and even between Strongyle and Nematodirus FEC, genetic correlations are at least 0.5 (Bishop et al., Reference Bishop, Jackson, Coop and Stear2004).
Quantitative trait loci (QTL) studies for nematode resistance, using microsatellite markers, have now been undertaken in many countries (Paterson et al., Reference Paterson, McEwan, Dodds, Morris and Crawford1999; Beh et al., Reference Beh, Hulme, Callaghan, Leish, Lenane, Windon and Maddox2002; Diez-Tascon et al., Reference Diez-Tascon, Macdonald, Dodds, McEwan and Crawford2002; Cockett et al., Reference Cockett, Bishop, Davies, Hadfield, Eng and Miller2005; Davies et al., Reference Davies, Stear, Benothman, Abuagob, Kerr, Mitchell and Bishop2006; Moreno et al., Reference Moreno, Gruner, Scala, Mura, Schibler, Amigues, Sechi, Jacquiet, François, Sechi, Roig, Casu, Barillet, Brunel, Bouix, Carta and Rupp2006; Gutiérrez-Gil et al., Reference Gutiérrez-Gil, Pérez, Alvarez, Martinez-Valladares, de la Fuente, Bayon, Meana, San Primitivo, Rojo-Vazquez and Arranz2009; Marshall et al., Reference Marshall, Maddox, Lee, Zhang, Kahn, Graser, Gondro, Walkden-Brown and van der Werf2009). A conclusion from these published results is that most significant QTL effects tend to be scattered throughout the genome. Only regions on chromosomes 3 (near to the interferon gamma locus) and 20 (within or adjacent to the MHC region) consistently appear significant across studies. Although some QTL appear to have large sizes of effect, caution must be used in interpreting these results, as it is the effects with the largest overestimation bias that will be the most significant.
Attention is now turning to genome-wide association studies using dense single nucleotide polymorphism (SNP) arrays. These studies attempt to either (i) find SNPs in population-wide linkage disequilibrium with mutations with large effects on the phenotype and/or (ii) develop genomic predictors of performance using the entire SNP chip. Initial results from studies of nematode resistance in sheep (e.g. Kemper et al., Reference Kemper, Emery, Bishop, Oddy, Hayes, Dominik, Henshall and Goddard2011), as well as other disease conditions in livestock and humans, suggest that resistance tends to be very polygenic, with few SNPs having large effects. This contrasts with QTL studies, where reported family-specific effect sizes are generally larger.
Selection for nematode resistance
Considerable evidence has now accumulated in developed countries that selection in sheep for resistance to nematode parasites, with selection based solely on FEC, is feasible and likely to be successful (Woolaston and Piper, Reference Woolaston and Piper1996; Morris et al., Reference Morris, Vlassoff, Bisset, Baker, West and Hurford1997; Woolaston and Windon, Reference Woolaston and Windon2001). One of the most spectacular demonstrations of such selection was documented by Kemper et al. (Reference Kemper, Palmer, Liu, Greeff, Bishop and Karlsson2010a). Following selection for reduced FEC in merino sheep after a natural challenge, FEC and worm burdens were ca. 20% of those in an unselected control line, after only 15 years of selection. Further, as indicated by the results of Kemper et al. (Reference Kemper, Palmer, Liu, Greeff, Bishop and Karlsson2010a) and the strong genetic correlation between resistance to difference nematode species (Gruner et al., Reference Gruner, Bouix and Brunel2004), selection for resistance to one species will tend to increase resistance to other species. Successful short-term selection for nematode resistance in goats has also been reported (Vagenas et al., Reference Vagenas, Jackson, Russel, Merchant, Wright and Bishop2002). Responses to artificial selection for nematode resistance have yet to be demonstrated in tropical sheep or goat production systems; however, long-term natural selection has produced many breeds that are resistant or tolerant of infection, as described above.
One of the main benefits of genetically improving resistance to nematodes is the epidemiological effect arising from reduced FEC (Bishop and Stear, Reference Bishop and Stear1997). Resistant sheep excrete fewer eggs and this ultimately leads to a reduced pasture larval contamination, as more eggs are removed from the pasture, through mortality or ingestion, than are returned. This reduced pasture larval contamination leads to reduced challenge and lower FEC, and hence a greater apparent response to selection, as well as benefits in performance (Bishop and Stear, Reference Bishop and Stear1999). Ultimately, this will benefit all animals grazing the same pasture, not only the genetically more resistant animals. Experimental evidence supporting this concept has been presented by Leathwick et al. (Reference Leathwick, Atkinson, Miller, Brown and Sutherland2002), Gruner et al. (Reference Gruner, Cortet, Sauve, Limouzin and Brunel2002) and Williams et al. (Reference Williams, Greeff, Vercoe, Dobson and Karlsson2010).
SWOT analysis of breeding programmes
In their report to the FAO (Opportunities for Incorporating Genetic Elements into the Management of Farm Animal Diseases: Policy Issues), Bishop et al. (Reference Bishop, de Jong and Gray2003) identified generic strengths, weaknesses, opportunities and threats for breeding programmes applied to disease resistance under tropical conditions. These may be adapted to the situation of nematode infections as follows:
Strengths
1. Genetic change is permanent
2. Consistency of effect: ‘just keeps working’
3. No continued input required once established
4. Prolong/protect the effectiveness of other methods (lifetime)
5. Broad-spectrum effects, that is, increased resistance to one disease can increase resistance to others
6. Adds to diversity of management strategies
Weaknesses
1. Goals of production system may change more quickly than genetic change can be implemented
2. Uncertainty of genetic outcomes in different environments and production systems
3. Need for some level of controlled breeding
4. Cost of measurements and analysis
5. Adds to technologies to be understood and implemented
Opportunities
1. Marketing of disease-resistant stock
2. Infrastructure that can be used for the other purposes, for example, performance recording
3. Mobilise communities for related activities including training and acquiring skills for farm and community management, marketing, politics
Threats
1. Inappropriate stock may become cheap and/or widely available
2. Genetic material may not be owned by local stakeholders
3. Opposition/competition from existing investors in other control options, for example, chemical suppliers
Clearly, a main benefit of breeding for resistance is its sustainability, that is, the fact that it is permanent and continues with little or no intervention once achieved. However, this has to be offset against the complication of applying what might be quite complex technology in the first instance. However, the infrastructure required to achieve genetic gain can be seen as an opportunity to obtain added value, through gaining of skills and other technical opportunities such as performance recording. An identifiable threat, which has occurred on occasions, is the use of inappropriate livestock, for example, imported animals from Europe or North America that lack adaptation characteristics to local conditions.
Choice of appropriate phenotype
So far, only FEC has been considered as the indicator of relative resistance to nematodes. However, there are several indicator traits that could be considered. These may be classified as follows:
- Measures of resistance: FEC, worm burden, worm size and fecundity.
- Immune response: Eosinophilia, antibodies such as IgA, IgG and IgM.
- Measures of impact of infection: anaemia, pepsinogen or fructosamine concentrations.
- Resilience: growth rate and required treatment frequency.
Of these traits, FEC, anaemia and growth rate are obvious goal traits, that is, traits that should be improved. In animals infected with H. contortus, anaemia can be easily measured using either packed cell volume (PCV) or Famacha© score, an indicator of anaemia in the eyelid. PCV is heritable in sheep and goats (e.g. Baker et al., Reference Baker, Audho, Adudo and Thorpe2001; Baker et al., Reference Baker, Nagda, Rodriguez-Zas, Southey, Audho, Aduda and Thorpe2003; Mandonnet et al., Reference Mandonnet, Menendez-Buxadera, Arquet, Mahieu, Bachand and Aumont2006) and emerging data also demonstrate that Famacha© scores are heritable in sheep (Riley and Van Wyk, Reference Riley and Van Wyk2009). Furthermore, anaemia scores are consistently negatively genetically correlated with FEC and positively correlated with live weight, making definition of selection goals straightforward, that is, decrease FEC and increase PCV and live weight gain.
Considerable work has been carried out looking at the genetic properties of the other mentioned traits, that is, concentrations of various antibodies (Strain et al., Reference Strain, Bishop, Henderson, Holmes, McKellar, Mitchell and Stear2002; Gutiérrez-Gil et al., Reference Gutiérrez-Gil, Pérez, de la Fuente, Meana, Martinez-Valladares, San Primitivo, Rojo-Vazquez and Arranz2010; Murphy et al., Reference Murphy, Eckersall, Bishop, Pettit, Huntley, Burchmore and Stear2010), eosinophils (Stear et al., Reference Stear, McKellar, Mitchell, Seeley and Bishop2002; Mandonnet et al., Reference Mandonnet, Menendez-Buxadera, Arquet, Mahieu, Bachand and Aumont2006), pepsinogen (Davies et al., Reference Davies, Stear and Bishop2005; Gutiérrez-Gil et al., Reference Gutiérrez-Gil, Pérez, de la Fuente, Meana, Martinez-Valladares, San Primitivo, Rojo-Vazquez and Arranz2010) and fructosamine (Stear et al., Reference Stear, Eckersall, Graham, McKellar, Mitchell and Bishop2001). All these measurements show desirable properties, such as being moderately to highly heritable and often strongly correlated with goal traits such as FEC or worm burdens (Davies et al., Reference Davies, Stear and Bishop2005). In fact, the results of Davies et al. (Reference Davies, Stear and Bishop2005) suggested that selection indices may rather easily be constructed with a number of additional indicator traits.
However, interpretation of indicator traits that describe animals’ immune responses or the impact of disease is complicated by the fact that these traits are a part of a dynamically changing response of the host to the parasite. Thus, they change over time and they are also affected by the previous challenge levels that animals have faced. In a study of longitudinal changes in genetic relationships, Davies (Reference Davies2006) showed that not only did genetic correlations of traits such as IgA or eosinophil concentrations with FEC or worm fecundity change over time, but that often they changed sign between times when lambs presumably had immature immune responses (e.g. 3 months of age) to when lambs had more mature immune responses (e.g. 6 months). This cautions against their use in breeding programmes, unless their genetic properties are well characterised. Consequently, breeding programmes based on phenotypic measurements in the lamb would more robustly focus on FEC, live weight gain and a measure of anaemia, such as PCV or Famacha© score, as described above. Care must be always taken that phenotypic measurements are truly comparable between animals. For example, the effectiveness of the Famacha© score as an indicator of anaemia will vary between breeds, and its utility may be limited in sheep that are only lightly infected (Moors and Gauly, Reference Moors and Gauly2009).
Resistance or resilience?
An issue that is often debated by parasitologists and breeders is whether it is more appropriate to select for resistance or resilience, the latter clearly being a desirable attribute in ruminants. Although resilience is usually thought of as the ability of an animal to maintain performance in the face of parasitic challenge, it has also been defined in terms of anthelmintic treatment requirements (specifically “the age at which a first post-weaning anthelmintic treatment is required to maintain acceptable growth in lambs grazing nematode-contaminated pasture” Morris et al., Reference Morris, Bisset, Vlassoff, Wheeler, West, Devantier and Mackay2010) and anaemia following H. contortus infection (Baker et al., Reference Baker, Nagda, Rodriguez-Zas, Southey, Audho, Aduda and Thorpe2003). Long-term selection for decreased treatment requirements has been shown to be successful and was accompanied by an increase in growth rate and a decrease in breech soiling (Morris et al., Reference Morris, Bisset, Vlassoff, Wheeler, West, Devantier and Mackay2010); however, it was complex to implement under practical conditions and did not improve resistance. Conversely, selection for a combination of resistance and performance should encompass the concept of resilience, and also allow some of the epidemiological benefits of selection described above to be captured. The choice of the optimal trait to select on will often depend on the feasibility of trait recording under practical conditions.
Use of genetic markers
Incorporation of genotype information, using genetic markers, is often argued as a means of making selection quicker and more efficient, avoiding the requirement for animals to be challenged with nematodes. However, this may prove problematic in practice, particularly for sheep or goats raised under tropical conditions. Reasons for this conclusion are briefly explored.
Most reported genetic marker studies for nematode resistance have detected and explored microsatellite-based QTL (Paterson et al., Reference Paterson, McEwan, Dodds, Morris and Crawford1999; Beh et al., Reference Beh, Hulme, Callaghan, Leish, Lenane, Windon and Maddox2002; Diez-Tascon et al., Reference Diez-Tascon, Macdonald, Dodds, McEwan and Crawford2002; Cockett et al., Reference Cockett, Bishop, Davies, Hadfield, Eng and Miller2005; Davies et al., Reference Davies, Stear, Benothman, Abuagob, Kerr, Mitchell and Bishop2006; Moreno et al., Reference Moreno, Gruner, Scala, Mura, Schibler, Amigues, Sechi, Jacquiet, François, Sechi, Roig, Casu, Barillet, Brunel, Bouix, Carta and Rupp2006; Gutiérrez-Gil et al., Reference Gutiérrez-Gil, Pérez, Alvarez, Martinez-Valladares, de la Fuente, Bayon, Meana, San Primitivo, Rojo-Vazquez and Arranz2009; Marshall et al., Reference Marshall, Maddox, Lee, Zhang, Kahn, Graser, Gondro, Walkden-Brown and van der Werf2009). However, results from such studies are often difficult to utilise in breeding programmes for a number of reasons. The primary difficulty is that the QTL are generally detected within families, and the linkage phase of markers with causative mutation(s) is family specific. This places an onerous task on re-establishing linkage phase within each family, and consequently leads to a continued large-scale requirement for phenotyping. Further, QTL will not be detected in homozygous families. Added to these difficulties are the observations that QTL for nematode resistance generally have small effects, that is, many are required to explain a substantial proportion of the genetic variation, and the technical challenges and costs associated with performing large numbers of microsatellite genotypes. Considerable infrastructure will be required to collect and extract DNA, prepare it for genotyping and perform the actual genotyping, Together, these factors mean that such QTL are unlikely to be sustainably incorporated into breeding programmes.
An alternative to microsatellite-based QTL is the exploitation of SNP associations, that is, SNPs that are associated with favourable phenotypes across an entire population. Because this technique uses SNPs that show population-wide disequilibrium with the causative mutation, or may even be the causative mutation in rare cases, the issue of family-specific linkage phase is avoided. To date, very few SNP association studies have been reported for nematode resistance in sheep or goats, and available evidence suggests that individual SNPs are likely to be associated with very small effects (Kemper et al., Reference Kemper, Emery, Bishop, Oddy, Hayes, Dominik, Henshall and Goddard2011). In other words, resistance is somewhat polygenic. Consequently, many SNPs would need to be included in a breeding programme to achieve reasonable genetic progress.
An alternative use of SNPs is in genomic selection. In this technique, all SNPs on an SNP chip, for example, >50 000 SNPs, are used simultaneously to predict the genetic merit of an individual, irrespective of the significance of individual SNPs (Meuwissen et al., Reference Meuwissen, Hayes and Goddard2001). This approach is now widespread in the dairy cattle breeding industry, which is characterised by a large census size but small effective population size, and a cumbersome and time-consuming progeny testing strategy. In such cases, selection accuracies of up to 80% can be achieved in the absence of phenotype information (Daetwyler et al., Reference Daetwyler, Pong-Wong, Villanueva and Woolliams2010), once robust genomic prediction equations have been developed. The situation for small ruminants is more challenging. The requirement to genotype and phenotype many thousands of animals in order to develop prediction equations will be difficult to achieve. Further, in many populations, for example, the Merino, effective population sizes are large suggesting that chips of 50 000 SNPs will be too small to achieve adequate accuracy of genotype prediction (Kemper et al., Reference Kemper, Emery, Bishop, Oddy, Hayes, Dominik, Henshall and Goddard2011). Lastly, extrapolation of results across distantly related populations is poor (Hayes et al., Reference Hayes, Bowman, Chamberlain and Goddard2009), with the genetically fragmented nature of sheep and goat breeds making it difficult to use the results on anything other than the population in which they were derived. Therefore, genomic selection, at least with current technologies, is likely to be expensive and logistically difficult to implement in tropical sheep and goats.
In summary, sustainable breeding programmes in sheep and goats for nematode resistance are likely to be based on phenotypic measurements, at least in the short term. Sustainable breeding will require the correct choice of genotype, that is, breed, accompanied by efficient phenotype measurement and parentage recording. At least with current technologies, genetic markers are unlikely to make a substantial contribution to such breeding programmes under tropical conditions.
Issues associated with nematode resistance
Genetic correlations with performance and genotype by environment interactions
The genetic relationship between performance and resistance to nematodes is often misunderstood (Bishop and Stear, Reference Bishop and Stear2003). Incorrect inferences are often drawn from observations such as the fact that locally adapted breeds tend to be small with poor production characteristics, whereas high-performing exotic breeds often have poor disease-resistance characteristics. These breed differences are likely to simply reflect their selection history; it cannot be concluded that they reflect an antagonism between resistance and performance. To deduce such relationships, it would be necessary to perform a within-breed genetic study.
As summarised by Bishop and Stear (Reference Bishop and Stear2003), genetic relationships between performance and resistance are the outcome of a complex set of factors. Firstly, although infection inevitably compromises performance, the ability of an animal to withstand infection or disease depends on the environment, particularly the protein nutrition available to the animal (Coop and Kyriazakis, Reference Coop and Kyriazakis2001). It may also be asserted that the genes controlling resistance per se and those influencing performance are generally not associated, that is, they are different genes with different gene variants. Consequently, genetic relationships between resistance and performance may be thought of as the outcome of a balance between two opposing factors: the resources used by the host to fight or protect against infection v. the damage caused by infection. If the resources used to protect the host outweigh the benefits of being more resistant, then the relationship will be unfavourable. If the benefits of being resistant, that is, the damage that is avoided, outweigh the costs of achieving resistance, then the relationship will be favourable.
This framework predicts that the genetic relationships of resistance and performance will vary with both the disease and the environment. As resources are improved, for example, better nutrition, then there will be less trade-off and a weaker relationship. Conversely, if the challenge level increases, then the benefits of being resistant increase and a stronger relationship is predicted. This framework also predicts genotype by environment interactions, for example, the ranking of animals on their performance may differ in environments with different challenge levels. This last prediction has been demonstrated in the comparison of Red Maasai (resistant and tolerant of nematode infections) v. Dorper sheep (susceptible; Baker et al., Reference Baker, Mugambi, Audho, Carles and Thorpe2004). In this comparison, breed differences in performance that were present in the face of strong nematode challenge (favouring the resistant Red Maasai) disappeared in an environment with a low-level challenge.
Sustainability of resistance and worm evolution
A feature of widespread and indiscriminate use of anthelmintics is the evolution of resistance by the nematode population to the specific chemical used for their control. Such anthelmintic resistance is now widespread (Waller, Reference Waller1997; Waller, Reference Waller2003; Jackson et al., Reference Jackson, Bartley, Bartley and Kenyon2009). Therefore, an equivalent concern is whether host genetic resistance would similarly breakdown over a period of time, with the nematodes evolving to adapt to the resistant hosts. The apparently polygenic nature of parasite resistance in sheep and goats would suggest that worm evolution is likely to be slower than the evolution of anthelmintic resistance, as worms would have to evolve against many more targets. Further, there is no published evidence for apparently resistant breeds losing their relative advantage in resistance compared with more susceptible breeds. Nevertheless, it is useful to consider the available evidence.
Kemper et al. (Reference Kemper, Elwin, Bishop, Goddard and Woolaston2009) reported serial passage of two strongyle species, H. contortus and Trichostrongylus colubriformis, through genetically susceptible and resistant hosts, for many parasite generations. The parasite strains were then tested not only in the host populations within which they were passaged but also in unselected outbred hosts. The experiment did produce some evidence of evolution in the worm populations; essentially, the passaging technique was equivalent to performing a weak selection experiment for fecundity within the parasite population. However, there was no evidence for differential adaptation between the resistant and susceptible hosts.
To make more general predictions on the likelihood of parasite evolution, and to develop a likely time horizon over which such evolution may occur, a predictive framework for parasite evolution was then developed, consisting of interacting models at three levels: disease epidemiology, host genetics and parasite genetics (Kemper, Reference Kemper2010; Kemper et al., Reference Kemper, Goddard and Bishop2010b). These models successfully predicted rapid rates of development of anthelmintic resistance. However, they predicted only slow rates of evolution in the parasite in response to resistant hosts. This, coupled with likely fitness costs to the parasite (Jorgensen et al., Reference Jorgensen, Leathwick, Charleston, Godfrey, Vlassoff and Sutherland1998), leads to the conclusion that even with some, albeit slow, change in parasites subjected to resistant hosts, the advantage in terms of reduced FEC in these hosts is likely to be maintained, at least for many parasite generations. Therefore, selection for resistant hosts can be considered a sustainable control strategy, although it will be most effective when used to complement other control strategies.
Conclusions
Breeding for enhanced resistance to nematode parasite may greatly facilitate and enhance parasite control. In tropical production systems where flock sizes are likely to be small, with diverse production systems and host genotypes, selection on phenotype is likely to be the most feasible option. The first step will be the choice of resistant or tolerant breeds, followed by measurement of traits such as growth rate in the face of challenge, FEC and anaemia. With current technologies, the use of genetic markers is likely to be too expensive and logistically difficult. With a correctly derived selection index and adequate management, selection for enhance resistance to nematodes should improve rather than compromise performance. Further, it is unlikely that the parasites will evolve to overcome the enhanced host resistance, at least in the foreseeable future.
Acknowledgements
Funding for research described in this paper has come from diverse sources including the BBSRC, EU, Defra, MLC, ODA and RERAD. I would like to thank many colleagues, particularly Kathryn Kemper, Gail Davies and Mike Stear, all of whom have contributed many ideas to this manuscript.
