- Email: [email protected]
Weight-based targeted selective treatment of gastrointestinal nematodes in a commercial sheep flock
Veterinary Parasitology 164 (2009) 59–65
Contents lists available at ScienceDirect
Veterinary Parasitology journal homepage: www.elsevier.com/locate/vetpar
Weight-based targeted selective treatment of gastrointestinal nematodes in a commercial sheep flock K.A. Stafford *, E.R. Morgan, G.C. Coles Department of Clinical Veterinary Science, University of Bristol, Langford House, Langford, Somerset BS40 5DU, UK
A R T I C L E I N F O
A B S T R A C T
Keywords: TST Weight gain Anthelmintic resistance Epidemiology Reduced chemical use Productivity Parasite control Low input Easy care Sheep breed
Targeted selective treatment (TST) strategies, in which a proportion of the flock or herd is left untreated so that anthelmintic-susceptible genotypes are preserved, are increasingly advocated as a means of prolonging the effective life of current anthelmintic drugs. The major limitation to this approach is a lack of efficient indicators for selection, which can be applied effectively on commercial farms to identify individuals that can be left untreated without fear of disease or production loss. With the advent of electronic identification and automated weighing technology, monitoring of short-term changes in weight gain shows promise as such an indicator, but its operation in the field as part of TST has yet to be evaluated. Widespread deployment of weight-based TST will be highly dependent on the likely production penalty from leaving the fastest growing animals untreated. On a commercial flock in south-west UK, the weight gain of 508 lambs of various breeds was tracked using an automated identification and weighing system, every one to ten weeks from June to December (one to four weeks in summer), and a variable proportion of the fastest growing individuals that also appeared to be in good condition with little breech soiling was left untreated during whole-flock dosing in June, July and August. In total, 51 lambs were selected for non-treatment on at least one occasion, while the other lambs were treated two or three times during the summer. Subsequent weight gain of untreated animals was not reduced relative to their peers in either the short-term or over the whole grazing season. Faecal egg counts from untreated individuals did not differ significantly from those of the rest of the flock, suggesting that animals left untreated on the basis of weight gain can contribute effectively to refugia. The application of TST in this case is cautious in its extent, but this is appropriate on a commercial farm with associated aversion to production loss. Results suggest that such losses can be avoided while leaving part of the flock untreated, and should encourage wider application of this approach to slow the development of anthelmintic resistance. Since the cost of investment in weighing and recording systems is likely to prove prohibitive to many farmers, other selective indicators should also be investigated. The co-ordination of TST with pasture use to maximise the benefit in terms of environmental refugia, and its integration with other control strategies, also requires further attention. ß 2009 Elsevier B.V. All rights reserved.
1. Introduction Modern farming methods have seen an increase in the stocking density at which sheep are kept. This has
* Corresponding author. Tel.: +44 117 9289495. E-mail address: [email protected] (K.A. Stafford). 0304-4017/$ – see front matter ß 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.vetpar.2009.04.009
primarily been made possible by the application of fertilisers to increase grass growth and regular wholeflock treatment with anthelmintics to control nematode infections. However, routine anthelmintic treatment inevitably selects for drug resistance in parasite populations. The common practices of regular whole-flock treatment, and dose and move to clean or rested pasture, have undoubtedly accelerated this process. Anthelmintic
60
K.A. Stafford et al. / Veterinary Parasitology 164 (2009) 59–65
resistance is now widespread in gastrointestinal nematodes of sheep worldwide. In the UK as in many other countries the situation is serious, with the first farms ceasing production due to the development of multiple resistant Teladorsagia circumcincta (Sargison et al., 2005; Blake and Coles, 2007). It has long been appreciated by parasitologists that helminth infections are typically aggregated (or, equivalently, overdispersed), with the majority of worms in the minority of animals (Barger, 1985; Gaba et al., 2005). As well as influencing host–parasite population interactions (Anderson and May, 1978; Quinnell et al., 1990), aggregation has practical implications for parasite control. By its nature, aggregation influences the effect of chemotherapy on morbidity in a group of hosts (Medley et al., 1993), and also its effects on parasite transmission (Churcher et al., 2005) and selection for drug resistance (Churcher and Basanez, 2008). Diagnosis of herd or flock infection levels as a trigger for treatment also become complicated when parasites are aggregated (Morgan et al., 2005). Importantly, overdispersion provides an opportunity for targeted control of parasites by treating relatively few hosts, since selective elimination of the burdens of the most heavily infected individuals will have a disproportionate effect on the parasite population as a whole. More than 20 years ago, Michel (1985) suggested that in the control of nematodes of grazing livestock, leaving some animals untreated could prove beneficial in terms of preserving the efficacy of anthelmintics. This should delay the development of resistance by leaving a pool of susceptible alleles of genes conferring resistance, thus diluting the frequency of resistant alleles in any given population. However, the importance of this approach was not widely recognised until it was reintroduced by Van Wyk (2001) and considered as part of general nematode control strategies (Coles, 2002; Sissay et al., 2006). Although theoretically attractive, the concept of targeted selective treatment (TST) runs into practical problems when applied in a commercial setting. Its major drawback is the risk that some animals will be left with parasite burdens sufficient to cause sub-clinical or even clinical disease and hence production loss, as well as compromised welfare. Even if relatively direct indices of parasitism such as faecal egg counts are used to select which animals to treat, by the time elevations are noted production loss has already occurred, since many species cause most damage at immature stages. Practical methods that allow rapid identification of individuals that are least in need of treatment, in the field and in time to inform the decision on whether or not to treat, are therefore urgently needed (Jackson and Miller, 2006). Such a system has been developed for sheep in warmer climates, where Haemonchus contortus dominates. The FAMACHAß method uses the colour of the mucous membrane of the lower eyelid to estimate the degree of anaemia, leading to substantial decreases in the number of sheep treated (Van Wyk and Bath, 2002), although greater care may be needed in lambs, in which fatal anaemia can develop quickly (Kaplan et al., 2004). A different approach to TST, which targets treatment of highly productive individuals and therefore those most susceptible to the effects of parasit-
ism, has also been successful in dairy goats (Hoste et al., 2002). Where H. contortus is not the dominant species, as in most temperate regions, anaemia is not a suitable indicator for selective treatment. There are a number of possible alternatives, including faecal egg counts, faecal moisture content, breech soiling (dag score), body condition scoring and weight gain. Faecal egg counts or faecal moisture content are not practical in large commercial flocks since the time needed to collect individual samples, count the eggs and then re-muster the lambs for dosing is prohibitive even without considering laboratory fees. Faecal egg counts are therefore more relevant to grazing animals of higher value kept in smaller numbers, for example horses (Eysker et al., 2006). Body condition scoring is useful in breeding ewes or fat lambs (Russel, 1984), but inconsistent in growing lambs. Breech soiling is related to the number of eggs in the faeces (Broughan and Wall, 2007) and could be useful for nematode species other than H. contortus. However, factors other than nematode infection also cause breech soiling. Weight gain has the advantage that it can be measured quickly and non-invasively, and is of sufficient interest to farmers for reasons other than parasite control that routine monitoring through the grazing season is a realistic proposition. With the impending introduction of compulsory electronic identification of sheep in the European Union, as a result of increasing emphasis on food traceability, and the availability of automated weighing systems, monitoring the weight of individual lambs is less labour intensive than ever before. Commercially available automated weighing systems generally link to computerised record keeping and flock management software, such that weight gain and parasite treatment data can be easily integrated into flock health and production planning. As an index for selective anthelmintic treatment, weight gain has the added advantage that alterations are detectable early in the course of sub-clinical parasitic gastroenteritis (PGE), and can be reversed with prompt treatment (Kyriazakis et al., 1996; Louie et al., 2007). Moreover, the major consequence of sub-clinical PGE to the farmer is reduced weight gain, so this index is more closely linked to the farmer’s commercial interests than even direct measures of parasite burden. The production consequence of error is low: thus, if individuals with high burdens are missed because their weight gain is rapid, they are by definition showing resilience to the effects of infection and their non-treatment has few penalties beyond increased pasture contamination. In the context of TST to build refugia, such mistakes could actually be beneficial. The willingness of commercial farmers to deploy TST in the fight against resistance will depend on the perceived short-term costs of lost production, which are balanced against the long-term costs of anthelmintic failure. Given the lack of available data on the costs of TST in the field, most farmers are reluctant to consider this approach and continue to routinely treat on a whole-flock basis. A second potential drawback of weight-based TST is that if animals that are growing well have very low parasite burdens, their contribution to environmental refugia of
K.A. Stafford et al. / Veterinary Parasitology 164 (2009) 59–65
anthelmintic-susceptible genotypes will be minimal. Although leading to reduced chemical use, this would defeat the objective of TST to slow the spread of resistance. Over-efficient exploitation of parasite overdispersion in targeting treatment is therefore epidemiologically advantageous, but counterproductive in terms of managing anthelmintic resistance. The aim of the present study was to determine whether TST based on weight gain, i.e. leaving the fastest growing animals untreated, can be practically incorporated into the normal working of a commercial farm. Further, we set out to determine the cost of TST in terms of reduced weight gain in untreated animals, and the contribution of these animals to environmental refugia by output of eggs from nematodes unexposed to anthelmintics. 2. Materials and methods The work was conducted on a commercial farm in an intensive sheep producing area of the south west of England, which routinely used an automated weighing system (Shearwell Data Ltd., Wheddon Cross, Somerset, UK) for monitoring weight gain in growing lambs at pasture. The system registers the identity of individual lambs as they pass through or are held in a weighing crate, by means of unique electronic ear tags. Weights are recorded and compiled for export to a computer spreadsheet. The system also has the capacity to draft individual lambs from the weigh crate in different directions according to pre-determined criteria, e.g. weight. Four breeds of sheep were present on the farm: Texel and Suffolk (meat breeds), Friesland (dairy breed, with most lambs reared for breeding and milk production), and Easycare (Table 1). The Easycare breed is derived from the Wiltshire Horn and Welsh Mountain breeds, and has been selected for easy maintenance, with rapid lamb growth at grass, wool shedding, and disease resistance prioritised in selection (http:// www.easycaresheep.com/). It is therefore a logical
61
candidate for production systems that seek to minimise intensive inputs including chemoprophylactic treatment of gastrointestinal nematodes. Previous work has established that T. circumcincta and Trichostrongylus spp. are the dominant causes of PGE in the study area, with H. contortus important on some farms but not that used in the present study. 2.1. Monitoring and selective treatment Grazing management and anthelmintic treatment during the study period proceeded according to the farmer’s judgement with no interference from the researchers, but with the stated aim of leaving a proportion of the lambs untreated. This proportion was not fixed, and a cautious approach was taken in order to avoid commercial loss resulting from sub-clinical and clinical effects of untreated nematode infections. Each time the flock was brought in for anthelmintic treatment, individual lambs were identified that were least likely to be experiencing ill effects from nematode infection, and that would therefore be expected to continue to perform well even if not treated. These lambs had to be (i) in good general condition, (ii) free of faecal soiling in the breech (perineal) area, and (iii) in the top 25% of their peer group in terms of weight gain (initially weight, then weight gain since last weighing). Approximate reference ranges for weight and weight gain were estimated by weighing a sub-sample of 20–25 lambs before beginning treatment. These lambs were selected by moving the group around the holding pen in a circular motion and then opening the gate at random to allow access to the race and weigh crate. 2.2. Effect of selection on weight gain Weight gains in non-treated individuals were compared with those of their treated peers. The average weight gain in each breed and sex category was calculated during
Table 1 Number of lambs of each breed and sex sampled on each weighing occasion, and selected for non-treatment on at least one occasion. Texel Female 18 July 27 July 3 August 11 August 28 August 4 September 17 September 21 September 29 September 8 October 13 October 20 October 4 November 27 December Not treated % Not treated
Suffolk Male
22
Suffolk Texel
Friesland
Female
Male
Female
Female
Male
Female
Easycare Male
Castrated
56
42
17
41
32
120
35
75
56
42
17
41
32
120
35 35 31
75
56 56
42 41
17 16
41
32 32
120 115
17 22 17 17 22 22 17
75 72 26
41 22 22
43 17
40 33
36
17
30 27
8
71
43
12 19
6
14
22
24
3
0
1
1
1
5
5
15
7
13
12
0
2
2
6
11
14
11
17
15
Numbers declined from August onwards as lambs were sold or sent to separate holdings.
K.A. Stafford et al. / Veterinary Parasitology 164 (2009) 59–65
62
Table 2 Average weight gain between weighing occasions (kg). The first weighing was on 26th June. Sample sizes are given in Table 1. Texel Female 18 July 27 July 3 August 11 August 28 August 4 September 17 September 21 September 29 September 8 October 13 October 20 October 4 November 27 December
Suffolk Male
3.09
Suffolk Texel
Friesland
Easycare
Female
Male
Female
Female
Male
Female
Male
Castrated
2.24
1.81
4.03
2.87
3.39
3.32
3.80
1.36
1.83
2.36
1.18
1.15
1.36
1.73
1.57 3.57 4.90
3.05
2.97 3.43
2.00 4.13
4.75 5.00
2.22
2.05 3.25
3.37 2.75
3.88 1.05 0.74 3.38 2.11 2.20 4.28
3.60 2.03 2.09
5.09 0.23 3.00
1.24 3.76
2.62 2.75
1.14
2.44
3.04 2.93
2.08
1.19
2.00
2.34 4.64
3.83
0.02
4.18
each time interval, and subtracted from individual weight gains of untreated individuals, to give a figure for relative weight gain. This was compared with zero using a onesample t-test. The effect of the decision to leave lambs untreated on at least one occasion during the summer on total summer weight gain was also examined. The total weight gain of each lamb from the end of June until mid September was calculated, and the interacting effects of breed, sex and decision to treat examined in fully factorial analysis of variance. SPSS v12.0 (SPSS Inc., Chicago, USA) was used for statistical analysis.
1.61
3.2. Effect of selection on weight gain The weight gain of lambs that were routinely treated is given in Table 2. Weight gain relative to the average for routinely dosed lambs of the same breed and sex was largely unaffected by the decision not to treat (Fig. 1).
2.3. Faecal egg output Faecal samples were collected opportunistically from lambs routinely treated and those selected for nontreatment at each weighing from June to September. Density of trichostrongyle-type eggs was estimated using the modified McMaster method (MAFF, 1986). 3. Results 3.1. Weighing and anthelmintic treatment A total of 508 lambs were monitored by automated weighing, which was conducted on 14 occasions (Table 1). Not all lambs were weighed on each occasion: groups were selected for weighing by the farmer based on perceived need and convenience, and each lamb was weighed between 5 and 7 times during the study period. Lambs were grazed in two groups on separate parts of the farm. Each group consisted of mixed breeds and towards the end of the grazing season there was some movement between groups. The proportion of lambs selected for non-treatment was similar between groups. Each group was treated with anthelmintic two (group 1, n = 276) or three (group 2, n = 232) times during the grazing season. Levamisole was used in late June (group 2), mid July (group 1) and late July (group 2), and ivermectin in mid August (both groups). Treatment was purposely withheld from 51 lambs on a total of 59 occasions, some lambs being selected for non-treatment on two occasions.
Fig. 1. Weight gain in lambs selected for non-treatment relative to that of treated lambs of the same age, breed and sex. Bars above the horizontal axis indicate higher average weight gain than that of the treated animals, and bars below the horizontal axis average weight gain below that of the treated animals. (a) Lambs not treated at the first dosing (26 June); (b) Lambs not treated at the second dosing (18 July). Months are July (7) to December (12). Each data point represents a weighing date: 18 July, 3 August, 4 and 17 September, 8 and 13 October, 27 December. Weight gain in the first interval is taken from 26 June. Error bars represent 95% confidence intervals. Arrows indicate the time of treatment.
K.A. Stafford et al. / Veterinary Parasitology 164 (2009) 59–65
63
Table 3 Analysis of variance in live weight gain from the end of June to mid September. Source
Type III sum of squares
Corrected Model Intercept Dose Breed Sex Dose breed Dose sex Breed sex Dose breed sex
728.702 16581.026 5.487 78.885 156.655 3.664 40.908 134.453 0.241
Error Total Corrected total
5079.018 53994.200 5807.720
df
Mean square
F
p
9 1 1 1 2 1 2 1 1
80.967 16581.026 5.487 78.885 78.327 3.664 20.454 134.453 0.241
5.356 1096.910 0.363 5.219 5.182 0.242 1.353 8.895 0.016
<0.001 <0.001 0.547 0.023 0.006 0.623 0.260 0.003 0.900
336 346 345
15.116
Factors are dose (number treated = 301, not treated = 45), breed (Friesland = 83, Easycare = 263) and sex (female = 181, male = 79, castrated male = 86). Adjusted R squared = 0.102. df = degrees of freedom.
Fig. 2. Faecal egg output of treated (shaded bars) and untreated (open bars) lambs in early (period 1, weighing date 26 June), mid (period 2, weighing date 18 July) and late (period 3, weighing dates 25 July and 3 August) summer. Error bars represent bootstrapped 95% confidence intervals (1000 repetitions with replacement). Numbers below the bars are the sample size, and numbers above the bars are the maximum faecal egg count (eggs per gram) in each group.
Lambs not treated in mid July showed a tendency to gain less weight than their routinely treated peers in the following 16-day period, but the difference was not significant (mean difference 0.38 kg, t45 = 1.45, p = 0.15). Since few Texel or Suffolk lambs were selected for non-treatment (n = 6 in 5 categories), these breeds were excluded from analysis of factors affecting seasonlong weight gain. In Friesland and Easycare breeds, weight gain from the end of June to mid September (mean 11.8 kg, standard deviation 4.10, n = 303) was significantly affected by breed, sex, and the interaction between breed and sex (Table 3). However, selection for non-treatment did not affect summer weight gain. 3.3. Faecal egg output The faecal output of lambs selected for non-treatment was not significantly different from that of lambs routinely treated (Fig. 2). 4. Discussion Based on weight gain relative to their peers, and visual assessment of body condition and breech soiling as a failsafe, 51 lambs were selected for non-treatment on a total
of 59 occasions, while the remaining 457 lambs were treated two or three times as part of routine whole-flock treatments. TST in this case therefore saved 59 out of 1248 treatments (=4.7%). The untreated lambs made a contribution to environmental refugia of anthelmintic-susceptible alleles, since their egg output per capita was similar to that of treated peers at the time of treatment, and presumably much greater after treatment. The proportion of animals left untreated was fairly low compared with that in other TST systems such as FAMACHAß (Van Wyk and Bath, 2002) and could be lower than required to optimize the trade-off between nematode control and selection for resistance. However, the proportion of the parasite population that must be left untreated in order to slow anthelmintic resistance has not been determined. The practical uptake of TST is highly dependent on farmer perception of production loss, with most reluctant to engage with this strategy when risks to production are unknown. Once conservative application of TST has been demonstrated on a farm with good results in terms of production, there is scope for refining the system and increasing the proportion left untreated in future years. The expectation prior to this study was that animals showing higher than average weight gain would have lower than average levels of nematode infection, and that weight gain as an indicator would enable parasite overdispersion to be exploited in selective treatment. However, comparison of faecal egg counts in treated and untreated lambs suggests that weight gain was largely independent of nematode burden in the animals sampled. This is in spite of moderately high average faecal egg counts that would normally be consistent with a risk of sub-clinical disease. It is likely that regular treatment of the flock kept levels of infection below the level at which it would have detectable effects on weight gain, and that untreated lambs benefited from this reduced level of challenge. The continued egg output from untreated lambs is of course consistent with the aims of TST, and the fact that faecal egg counts were not lower in animals selected for non-treatment underlines their benefit in terms of generating refugia. However, more vigorous selection such that a larger proportion of lambs is left untreated could result in epidemiologically important
64
K.A. Stafford et al. / Veterinary Parasitology 164 (2009) 59–65
levels of pasture contamination from untreated animals, and the risk of production loss across the flock. This tradeoff is inherent to the TST approach. Given the unexpected lack of bias towards lower egg counts in individuals selected for non-treatment, incorporation of routine faecal egg count monitoring alongside weighing, e.g. through comparison of pooled egg counts from treated and untreated groups (Morgan et al., 2005) might be advisable. The rate of weight gain in lambs selected for nontreatment was not significantly depressed either in the short-term or over the whole grazing season. This is encouraging, since it suggests that production loss can be completely avoided on commercial farms while using TST. However, weight gain of untreated animals was compared with that of peers matched by breed and sex. It is therefore possible that growth of untreated animals was impaired when compared with their potential growth, and actual performance was reduced to the flock average through non-treatment. Since these could be the best genetic stock and therefore the most desirable replacement animals, or the first to be ready for market, the commercial penalty of reduced growth in these animals could be disproportionate. This requires further study. In a study of TST in New Zealand, untreated lambs grew more slowly than treated lambs over some periods, but overall penalties in average weight gain were not discernible, as individuals whose growth fell behind that of the group were unlikely to be selected twice for non-treatment (Leathwick et al., 2006a). In the same study, groups of lambs subjected to TST by non-treatment of the 10% heaviest individuals on five occasions matched the average growth of those given five whole-group treatments. The implications of reduced rate of growth differ between production systems. Thus, for meat lamb production rapid growth on grass and early dispatch to market is of key importance, whereas for breeding animals slower growth in the face of moderate nematode challenge can be made up in the future, and the benefits of immunity are carried forward in terms of reduced impact of parasites on ewes and reduced pasture contamination for lambs. Further field studies of TST should include a range of breeds and production systems, including those geared to rapid finishing of meat lambs. Since multiple management practices can affect the rate of selection for resistance, TST should moreover be integrated with other control strategies, especially grazing management and treatment regimes for older age classes, to enhance its role in generating pasture refugia for anthelmintic susceptibility (Coles, 2002; Lawrence et al., 2006a; Leathwick et al., 2006b). The uptake of TST will depend strongly on the ease, rapidity and cost of the relevant indicator for treatment. While automated weighing makes short-term changes in weight gain a practical proposition as a selective indicator, the high cost of investment in equipment is likely to exclude many producers. Moreover, farmers would be reluctant to base treatment purely on weight gain if there are other signs that relatively fast-growing animals would benefit from treatment. In the present study, treatment based on farmer judgement alone, regardless of weight gain, was permitted, and the sign most commonly used to
justify such treatment was breech soiling. Faecal egg counts on animals identified in this way suggested that their nematode burdens were elevated (unpublished data). Visual assessment of breech soiling could therefore be usefully incorporated into weight-based TST, or even used as an alternative where automated weighing facilities are not available. This should form the basis for future work. Automated weighing systems could also be improved to ease their use in TST, in particular through software development to automatically compute incremental weight gain, and enable simple recording and recall of information on individual anthelmintic treatment. The present study demonstrates that weight-based TST can be conservatively applied to commercial sheep farms in temperate areas with no detectable loss in productivity, and with reasonably high levels of egg production from untreated individuals. Further work is needed to determine the proportion of the flock that can be left untreated without production loss, although this is likely to be highly variable between production systems and with epidemiological conditions. Importantly, the effect of TST on the rate of development of anthelmintic resistance needs to be determined if farmers are to be persuaded to adopt this approach. Development of a theoretical framework for predicting such effects is under way (Gaba et al., 2006), but field validation will of course be necessary, and will have to include commercial holdings if farmers are to be persuaded of its applicability. Although the present study reports relatively small reductions in the number of treatments given, this is an improvement on the present situation in which whole-flock treatments are very much the norm. It is not surprising that farmers prefer to take the short-term benefits of intensive anthelmintic treatment when the production implications of TST are so poorly characterized, even though the longer range costs of this strategy in terms of anthelmintic resistance are increasingly apparent (Bath, 2006; Waghorn et al., 2006; Lawrence et al., 2006b). Perhaps this situation will change when overt failure of anthelmintic therapy becomes more common, but waiting until this occurs will result in high levels of suffering and economic loss. Demonstration of TST using weight gain and also other more practical and accessible indicators for treatment, with minimal production loss, should be prioritized if this approach is to progress beyond the research stage. Conflict of interest There are no conflicts of interest. Acknowledgements This work was funded by the European Union FP6 under the PARASOL project. We gratefully acknowledge the help and co-operation of Peter Baber with data collection, and useful discussions with Richard Webber and staff at Shearwell data systems. Research within this paper was either partially or wholly supported by the EU Parasol project (FOOD-CT2005-022851) and by other governmental/institutional/ university funding.
K.A. Stafford et al. / Veterinary Parasitology 164 (2009) 59–65
References Anderson, R.M., May, R.M., 1978. Regulation and stability of host–parasite population interactions. 1. Regulatory processes. J. Anim. Ecol. 47, 219–247. Barger, I.A., 1985. The statistical distribution of trichostrongylid nematodes in grazing lambs. Int. J. Parasitol. 15, 645–649. Bath, G.F., 2006. Practical implementation of holistic internal parasite management in sheep. Small Rumin. Res. 62, 13–18. Broughan, J.M., Wall, R., 2007. Faecal soiling and gastrointestinal helminths infection in lambs. Int. J. Parasitol. 37, 1255–1268. Blake, N., Coles, G.C., 2007. Flock cull due to anthelmintic-resistant nematodes. Vet. Rec. 161, 36. Churcher, T.S., Ferguson, N.M., Basanez, M.G., 2005. Density dependence and overdispersion in the transmission of helminth parasites. Parasitology 131, 121–132. Churcher, T.S., Basanez, M.G., 2008. Density dependence and the spread of anthelmintic resistance. Evolution 62, 528–537. Coles, G.C., 2002. Sustainable use of anthelmintics in grazing animals. Vet. Rec. 151, 165–169. Eysker, M., van Doorn, D.C.K., Lems, S.N., Weteling, A., Ploeger, H.W., 2006. Frequent de-worming in horses; the advantage often more damage than good. Tijdschrift voor dier Geneeskunde 131, 524–530. Gaba, S., Ginot, V., Cabaret, J., 2005. Modelling macroparasite aggregation using a nematode-sheep system: the Weibull distribution as an alternative to the Negative Binomial distribution? Parasitology 131, 393–401. Gaba, S., Cabaret, J., Ginot, V., Silvestre, A., 2006. The early drug selection of nematodes to anthelmintics: stochastic transmission and population in refuge. Parasitology 133, 345–356. Hoste, H., Le Frileux, Y., Pommaret, A., 2002. Comparison of selective and systematic treatments to control nematode infection of the digestive tract in dairy goats. Vet. Parasitol. 106, 345–355. Jackson, F., Miller, J., 2006. Alternative approaches to control—Quo vadit? Vet. Parasitol. 139, 371–384. Kaplan, R.M., Burke, J.M., Terrill, T.H., Miller, J.E., Getz, W.R., Mobini, S., Valencia, E., Williams, M.J., Williamson, L.H., Larsen, M., Vatta, A.F., 2004. Validation of the FAMACHAß eye colour chart for detecting clinical anemia in sheep and goats on farms in the southern United States. Vet. Parasitol. 123, 105–120. Kyriazakis, I., Anderson, D.H., Oldham, J.D., Coop, R.L., Jackson, F., 1996. Long-term subclinical infection with Trichostrongylus colubriformis: effects on food intake, diet selection and performance of growing lambs. Vet. Parasitol. 61, 297–313. Lawrence, K.E., Leathwick, D.M., Rhodes, A.P., Jackson, R., Heuer, C., Pomroy, W.E., West, D.M., Waghorn, T.S., Moffat, J.R., 2006a. Management of gastrointestinal nematode parasites on sheep farms in New Zealand. N. Z. Vet. J. 55, 228–234.
65
Lawrence, K.E., Rhodes, A.P., Jackson, R., Leathwick, D.M., Heuer, C., Pomroy, W.E., West, D.M., Waghorn, T.S., Moffat, J.R., 2006b. Farm management practices associated with macrocyclic lactone resistance on sheep farms in New Zealand. N. Z. Vet. J. 54, 283–288. Leathwick, D.M., Waghorn, T.S., Miller, C.M., Atkinson, D.S., Haack, N.A., Oliver, A.M., 2006a. Selective and on-demand drenching of lambs: impact on parasite populations and performance of lambs. N. Z. Vet. J. 54, 305–312. Leathwick, D.M., Miller, C.M., Atkinson, D.S., Haack, N.A., Alexander, R.A., Oliver, A.M., Waghorn, T.S., Potter, J.F., Sutherland, I.A., 2006b. Drenching adult ewes: implications of anthelmintic treatments pre- and post-lambing on the development of anthelmintic resistance. N. Z. Vet. J. 54, 297–304. Louie, K., Vlassoff, A., Mackay, A.D., 2007. Gastrointestinal nematode parasites of sheep: a dynamic model for their effect on liveweight gain. Int. J. Parasitol. 37, 233–241. MAFF, 1986. Manual of Veterinary Parasitological Laboratory Techniques. Reference Book 418, 3rd ed. HMSO, London. Medley, G.F., Guyatt, H.L., Bundy, D.A.P., 1993. A quantitative framework for evaluating the effect of community treatment on the morbidity due to ascariasis. Parasitology 106, 211–221. Michel, J.E., 1985. Strategies for the use of anthelmintics in livestock and their implications for the development of drug resistance. Parasitology 90, 621–628. Morgan, E.R., Cavill, L., Curry, G.E., Wood, R.M., Mitchell, E.S.E., 2005. Effects of aggregation and sample size on composite faecal egg counts in sheep. Vet. Parasitol. 131, 79–87. Quinnell, R.J., Medley, G.F., Keymer, A.E., 1990. The regulation of gastrointestinal helminth populations. Phil. Trans. Royal Soc. Lond. B 330, 191–201. Russel, A., 1984. Body condition scoring of sheep. In Practice 6, 91–93. Sargison, N.D., Jackson, F., Bartley, N.D., Moir, A.C.P., 2005. Failure of moxidectin to control benzimidazole-, levamisole- and ivermectinresistant Teladorsagia circumcincta in a sheep flock. Vet. Rec. 156, 105– 109. Sissay, M.M., Asefa, A., Uggla, A., Waller, P.J., 2006. Anthelmintic resistance of nematode parasites of small ruminants in eastern Ethiopia: exploitation of refugia to restore anthelmintic efficacy. Vet. Parasitol. 135, 337–346. Van Wyk, J.A., 2001. Refugia—overlooked as perhaps the most potent factor concerning the development of anthelmintic resistance. Onderstepoort J. Vet. Res. 68, 55–67. Van Wyk, J.A., Bath, G.F., 2002. The FAMACHAß system for managing haemonchosis in sheep and goats by clinically identifying individual animals for treatment. Vet. Res. 33, 509–529. Waghorn, T.S., Leathwick, D.M., Rhodes, A.P., Lawrence, K.E., Jackson, R., Pomroy, W.E., West, D.M., Moffat, J.R., 2006. Prevalence of anthelmintic resistance on sheep farms in New Zealand. N. Z. Vet. J. 54, 271–277.
Contents lists available at ScienceDirect
Veterinary Parasitology journal homepage: www.elsevier.com/locate/vetpar
Weight-based targeted selective treatment of gastrointestinal nematodes in a commercial sheep flock K.A. Stafford *, E.R. Morgan, G.C. Coles Department of Clinical Veterinary Science, University of Bristol, Langford House, Langford, Somerset BS40 5DU, UK
A R T I C L E I N F O
A B S T R A C T
Keywords: TST Weight gain Anthelmintic resistance Epidemiology Reduced chemical use Productivity Parasite control Low input Easy care Sheep breed
Targeted selective treatment (TST) strategies, in which a proportion of the flock or herd is left untreated so that anthelmintic-susceptible genotypes are preserved, are increasingly advocated as a means of prolonging the effective life of current anthelmintic drugs. The major limitation to this approach is a lack of efficient indicators for selection, which can be applied effectively on commercial farms to identify individuals that can be left untreated without fear of disease or production loss. With the advent of electronic identification and automated weighing technology, monitoring of short-term changes in weight gain shows promise as such an indicator, but its operation in the field as part of TST has yet to be evaluated. Widespread deployment of weight-based TST will be highly dependent on the likely production penalty from leaving the fastest growing animals untreated. On a commercial flock in south-west UK, the weight gain of 508 lambs of various breeds was tracked using an automated identification and weighing system, every one to ten weeks from June to December (one to four weeks in summer), and a variable proportion of the fastest growing individuals that also appeared to be in good condition with little breech soiling was left untreated during whole-flock dosing in June, July and August. In total, 51 lambs were selected for non-treatment on at least one occasion, while the other lambs were treated two or three times during the summer. Subsequent weight gain of untreated animals was not reduced relative to their peers in either the short-term or over the whole grazing season. Faecal egg counts from untreated individuals did not differ significantly from those of the rest of the flock, suggesting that animals left untreated on the basis of weight gain can contribute effectively to refugia. The application of TST in this case is cautious in its extent, but this is appropriate on a commercial farm with associated aversion to production loss. Results suggest that such losses can be avoided while leaving part of the flock untreated, and should encourage wider application of this approach to slow the development of anthelmintic resistance. Since the cost of investment in weighing and recording systems is likely to prove prohibitive to many farmers, other selective indicators should also be investigated. The co-ordination of TST with pasture use to maximise the benefit in terms of environmental refugia, and its integration with other control strategies, also requires further attention. ß 2009 Elsevier B.V. All rights reserved.
1. Introduction Modern farming methods have seen an increase in the stocking density at which sheep are kept. This has
* Corresponding author. Tel.: +44 117 9289495. E-mail address: [email protected] (K.A. Stafford). 0304-4017/$ – see front matter ß 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.vetpar.2009.04.009
primarily been made possible by the application of fertilisers to increase grass growth and regular wholeflock treatment with anthelmintics to control nematode infections. However, routine anthelmintic treatment inevitably selects for drug resistance in parasite populations. The common practices of regular whole-flock treatment, and dose and move to clean or rested pasture, have undoubtedly accelerated this process. Anthelmintic
60
K.A. Stafford et al. / Veterinary Parasitology 164 (2009) 59–65
resistance is now widespread in gastrointestinal nematodes of sheep worldwide. In the UK as in many other countries the situation is serious, with the first farms ceasing production due to the development of multiple resistant Teladorsagia circumcincta (Sargison et al., 2005; Blake and Coles, 2007). It has long been appreciated by parasitologists that helminth infections are typically aggregated (or, equivalently, overdispersed), with the majority of worms in the minority of animals (Barger, 1985; Gaba et al., 2005). As well as influencing host–parasite population interactions (Anderson and May, 1978; Quinnell et al., 1990), aggregation has practical implications for parasite control. By its nature, aggregation influences the effect of chemotherapy on morbidity in a group of hosts (Medley et al., 1993), and also its effects on parasite transmission (Churcher et al., 2005) and selection for drug resistance (Churcher and Basanez, 2008). Diagnosis of herd or flock infection levels as a trigger for treatment also become complicated when parasites are aggregated (Morgan et al., 2005). Importantly, overdispersion provides an opportunity for targeted control of parasites by treating relatively few hosts, since selective elimination of the burdens of the most heavily infected individuals will have a disproportionate effect on the parasite population as a whole. More than 20 years ago, Michel (1985) suggested that in the control of nematodes of grazing livestock, leaving some animals untreated could prove beneficial in terms of preserving the efficacy of anthelmintics. This should delay the development of resistance by leaving a pool of susceptible alleles of genes conferring resistance, thus diluting the frequency of resistant alleles in any given population. However, the importance of this approach was not widely recognised until it was reintroduced by Van Wyk (2001) and considered as part of general nematode control strategies (Coles, 2002; Sissay et al., 2006). Although theoretically attractive, the concept of targeted selective treatment (TST) runs into practical problems when applied in a commercial setting. Its major drawback is the risk that some animals will be left with parasite burdens sufficient to cause sub-clinical or even clinical disease and hence production loss, as well as compromised welfare. Even if relatively direct indices of parasitism such as faecal egg counts are used to select which animals to treat, by the time elevations are noted production loss has already occurred, since many species cause most damage at immature stages. Practical methods that allow rapid identification of individuals that are least in need of treatment, in the field and in time to inform the decision on whether or not to treat, are therefore urgently needed (Jackson and Miller, 2006). Such a system has been developed for sheep in warmer climates, where Haemonchus contortus dominates. The FAMACHAß method uses the colour of the mucous membrane of the lower eyelid to estimate the degree of anaemia, leading to substantial decreases in the number of sheep treated (Van Wyk and Bath, 2002), although greater care may be needed in lambs, in which fatal anaemia can develop quickly (Kaplan et al., 2004). A different approach to TST, which targets treatment of highly productive individuals and therefore those most susceptible to the effects of parasit-
ism, has also been successful in dairy goats (Hoste et al., 2002). Where H. contortus is not the dominant species, as in most temperate regions, anaemia is not a suitable indicator for selective treatment. There are a number of possible alternatives, including faecal egg counts, faecal moisture content, breech soiling (dag score), body condition scoring and weight gain. Faecal egg counts or faecal moisture content are not practical in large commercial flocks since the time needed to collect individual samples, count the eggs and then re-muster the lambs for dosing is prohibitive even without considering laboratory fees. Faecal egg counts are therefore more relevant to grazing animals of higher value kept in smaller numbers, for example horses (Eysker et al., 2006). Body condition scoring is useful in breeding ewes or fat lambs (Russel, 1984), but inconsistent in growing lambs. Breech soiling is related to the number of eggs in the faeces (Broughan and Wall, 2007) and could be useful for nematode species other than H. contortus. However, factors other than nematode infection also cause breech soiling. Weight gain has the advantage that it can be measured quickly and non-invasively, and is of sufficient interest to farmers for reasons other than parasite control that routine monitoring through the grazing season is a realistic proposition. With the impending introduction of compulsory electronic identification of sheep in the European Union, as a result of increasing emphasis on food traceability, and the availability of automated weighing systems, monitoring the weight of individual lambs is less labour intensive than ever before. Commercially available automated weighing systems generally link to computerised record keeping and flock management software, such that weight gain and parasite treatment data can be easily integrated into flock health and production planning. As an index for selective anthelmintic treatment, weight gain has the added advantage that alterations are detectable early in the course of sub-clinical parasitic gastroenteritis (PGE), and can be reversed with prompt treatment (Kyriazakis et al., 1996; Louie et al., 2007). Moreover, the major consequence of sub-clinical PGE to the farmer is reduced weight gain, so this index is more closely linked to the farmer’s commercial interests than even direct measures of parasite burden. The production consequence of error is low: thus, if individuals with high burdens are missed because their weight gain is rapid, they are by definition showing resilience to the effects of infection and their non-treatment has few penalties beyond increased pasture contamination. In the context of TST to build refugia, such mistakes could actually be beneficial. The willingness of commercial farmers to deploy TST in the fight against resistance will depend on the perceived short-term costs of lost production, which are balanced against the long-term costs of anthelmintic failure. Given the lack of available data on the costs of TST in the field, most farmers are reluctant to consider this approach and continue to routinely treat on a whole-flock basis. A second potential drawback of weight-based TST is that if animals that are growing well have very low parasite burdens, their contribution to environmental refugia of
K.A. Stafford et al. / Veterinary Parasitology 164 (2009) 59–65
anthelmintic-susceptible genotypes will be minimal. Although leading to reduced chemical use, this would defeat the objective of TST to slow the spread of resistance. Over-efficient exploitation of parasite overdispersion in targeting treatment is therefore epidemiologically advantageous, but counterproductive in terms of managing anthelmintic resistance. The aim of the present study was to determine whether TST based on weight gain, i.e. leaving the fastest growing animals untreated, can be practically incorporated into the normal working of a commercial farm. Further, we set out to determine the cost of TST in terms of reduced weight gain in untreated animals, and the contribution of these animals to environmental refugia by output of eggs from nematodes unexposed to anthelmintics. 2. Materials and methods The work was conducted on a commercial farm in an intensive sheep producing area of the south west of England, which routinely used an automated weighing system (Shearwell Data Ltd., Wheddon Cross, Somerset, UK) for monitoring weight gain in growing lambs at pasture. The system registers the identity of individual lambs as they pass through or are held in a weighing crate, by means of unique electronic ear tags. Weights are recorded and compiled for export to a computer spreadsheet. The system also has the capacity to draft individual lambs from the weigh crate in different directions according to pre-determined criteria, e.g. weight. Four breeds of sheep were present on the farm: Texel and Suffolk (meat breeds), Friesland (dairy breed, with most lambs reared for breeding and milk production), and Easycare (Table 1). The Easycare breed is derived from the Wiltshire Horn and Welsh Mountain breeds, and has been selected for easy maintenance, with rapid lamb growth at grass, wool shedding, and disease resistance prioritised in selection (http:// www.easycaresheep.com/). It is therefore a logical
61
candidate for production systems that seek to minimise intensive inputs including chemoprophylactic treatment of gastrointestinal nematodes. Previous work has established that T. circumcincta and Trichostrongylus spp. are the dominant causes of PGE in the study area, with H. contortus important on some farms but not that used in the present study. 2.1. Monitoring and selective treatment Grazing management and anthelmintic treatment during the study period proceeded according to the farmer’s judgement with no interference from the researchers, but with the stated aim of leaving a proportion of the lambs untreated. This proportion was not fixed, and a cautious approach was taken in order to avoid commercial loss resulting from sub-clinical and clinical effects of untreated nematode infections. Each time the flock was brought in for anthelmintic treatment, individual lambs were identified that were least likely to be experiencing ill effects from nematode infection, and that would therefore be expected to continue to perform well even if not treated. These lambs had to be (i) in good general condition, (ii) free of faecal soiling in the breech (perineal) area, and (iii) in the top 25% of their peer group in terms of weight gain (initially weight, then weight gain since last weighing). Approximate reference ranges for weight and weight gain were estimated by weighing a sub-sample of 20–25 lambs before beginning treatment. These lambs were selected by moving the group around the holding pen in a circular motion and then opening the gate at random to allow access to the race and weigh crate. 2.2. Effect of selection on weight gain Weight gains in non-treated individuals were compared with those of their treated peers. The average weight gain in each breed and sex category was calculated during
Table 1 Number of lambs of each breed and sex sampled on each weighing occasion, and selected for non-treatment on at least one occasion. Texel Female 18 July 27 July 3 August 11 August 28 August 4 September 17 September 21 September 29 September 8 October 13 October 20 October 4 November 27 December Not treated % Not treated
Suffolk Male
22
Suffolk Texel
Friesland
Female
Male
Female
Female
Male
Female
Easycare Male
Castrated
56
42
17
41
32
120
35
75
56
42
17
41
32
120
35 35 31
75
56 56
42 41
17 16
41
32 32
120 115
17 22 17 17 22 22 17
75 72 26
41 22 22
43 17
40 33
36
17
30 27
8
71
43
12 19
6
14
22
24
3
0
1
1
1
5
5
15
7
13
12
0
2
2
6
11
14
11
17
15
Numbers declined from August onwards as lambs were sold or sent to separate holdings.
K.A. Stafford et al. / Veterinary Parasitology 164 (2009) 59–65
62
Table 2 Average weight gain between weighing occasions (kg). The first weighing was on 26th June. Sample sizes are given in Table 1. Texel Female 18 July 27 July 3 August 11 August 28 August 4 September 17 September 21 September 29 September 8 October 13 October 20 October 4 November 27 December
Suffolk Male
3.09
Suffolk Texel
Friesland
Easycare
Female
Male
Female
Female
Male
Female
Male
Castrated
2.24
1.81
4.03
2.87
3.39
3.32
3.80
1.36
1.83
2.36
1.18
1.15
1.36
1.73
1.57 3.57 4.90
3.05
2.97 3.43
2.00 4.13
4.75 5.00
2.22
2.05 3.25
3.37 2.75
3.88 1.05 0.74 3.38 2.11 2.20 4.28
3.60 2.03 2.09
5.09 0.23 3.00
1.24 3.76
2.62 2.75
1.14
2.44
3.04 2.93
2.08
1.19
2.00
2.34 4.64
3.83
0.02
4.18
each time interval, and subtracted from individual weight gains of untreated individuals, to give a figure for relative weight gain. This was compared with zero using a onesample t-test. The effect of the decision to leave lambs untreated on at least one occasion during the summer on total summer weight gain was also examined. The total weight gain of each lamb from the end of June until mid September was calculated, and the interacting effects of breed, sex and decision to treat examined in fully factorial analysis of variance. SPSS v12.0 (SPSS Inc., Chicago, USA) was used for statistical analysis.
1.61
3.2. Effect of selection on weight gain The weight gain of lambs that were routinely treated is given in Table 2. Weight gain relative to the average for routinely dosed lambs of the same breed and sex was largely unaffected by the decision not to treat (Fig. 1).
2.3. Faecal egg output Faecal samples were collected opportunistically from lambs routinely treated and those selected for nontreatment at each weighing from June to September. Density of trichostrongyle-type eggs was estimated using the modified McMaster method (MAFF, 1986). 3. Results 3.1. Weighing and anthelmintic treatment A total of 508 lambs were monitored by automated weighing, which was conducted on 14 occasions (Table 1). Not all lambs were weighed on each occasion: groups were selected for weighing by the farmer based on perceived need and convenience, and each lamb was weighed between 5 and 7 times during the study period. Lambs were grazed in two groups on separate parts of the farm. Each group consisted of mixed breeds and towards the end of the grazing season there was some movement between groups. The proportion of lambs selected for non-treatment was similar between groups. Each group was treated with anthelmintic two (group 1, n = 276) or three (group 2, n = 232) times during the grazing season. Levamisole was used in late June (group 2), mid July (group 1) and late July (group 2), and ivermectin in mid August (both groups). Treatment was purposely withheld from 51 lambs on a total of 59 occasions, some lambs being selected for non-treatment on two occasions.
Fig. 1. Weight gain in lambs selected for non-treatment relative to that of treated lambs of the same age, breed and sex. Bars above the horizontal axis indicate higher average weight gain than that of the treated animals, and bars below the horizontal axis average weight gain below that of the treated animals. (a) Lambs not treated at the first dosing (26 June); (b) Lambs not treated at the second dosing (18 July). Months are July (7) to December (12). Each data point represents a weighing date: 18 July, 3 August, 4 and 17 September, 8 and 13 October, 27 December. Weight gain in the first interval is taken from 26 June. Error bars represent 95% confidence intervals. Arrows indicate the time of treatment.
K.A. Stafford et al. / Veterinary Parasitology 164 (2009) 59–65
63
Table 3 Analysis of variance in live weight gain from the end of June to mid September. Source
Type III sum of squares
Corrected Model Intercept Dose Breed Sex Dose breed Dose sex Breed sex Dose breed sex
728.702 16581.026 5.487 78.885 156.655 3.664 40.908 134.453 0.241
Error Total Corrected total
5079.018 53994.200 5807.720
df
Mean square
F
p
9 1 1 1 2 1 2 1 1
80.967 16581.026 5.487 78.885 78.327 3.664 20.454 134.453 0.241
5.356 1096.910 0.363 5.219 5.182 0.242 1.353 8.895 0.016
<0.001 <0.001 0.547 0.023 0.006 0.623 0.260 0.003 0.900
336 346 345
15.116
Factors are dose (number treated = 301, not treated = 45), breed (Friesland = 83, Easycare = 263) and sex (female = 181, male = 79, castrated male = 86). Adjusted R squared = 0.102. df = degrees of freedom.
Fig. 2. Faecal egg output of treated (shaded bars) and untreated (open bars) lambs in early (period 1, weighing date 26 June), mid (period 2, weighing date 18 July) and late (period 3, weighing dates 25 July and 3 August) summer. Error bars represent bootstrapped 95% confidence intervals (1000 repetitions with replacement). Numbers below the bars are the sample size, and numbers above the bars are the maximum faecal egg count (eggs per gram) in each group.
Lambs not treated in mid July showed a tendency to gain less weight than their routinely treated peers in the following 16-day period, but the difference was not significant (mean difference 0.38 kg, t45 = 1.45, p = 0.15). Since few Texel or Suffolk lambs were selected for non-treatment (n = 6 in 5 categories), these breeds were excluded from analysis of factors affecting seasonlong weight gain. In Friesland and Easycare breeds, weight gain from the end of June to mid September (mean 11.8 kg, standard deviation 4.10, n = 303) was significantly affected by breed, sex, and the interaction between breed and sex (Table 3). However, selection for non-treatment did not affect summer weight gain. 3.3. Faecal egg output The faecal output of lambs selected for non-treatment was not significantly different from that of lambs routinely treated (Fig. 2). 4. Discussion Based on weight gain relative to their peers, and visual assessment of body condition and breech soiling as a failsafe, 51 lambs were selected for non-treatment on a total
of 59 occasions, while the remaining 457 lambs were treated two or three times as part of routine whole-flock treatments. TST in this case therefore saved 59 out of 1248 treatments (=4.7%). The untreated lambs made a contribution to environmental refugia of anthelmintic-susceptible alleles, since their egg output per capita was similar to that of treated peers at the time of treatment, and presumably much greater after treatment. The proportion of animals left untreated was fairly low compared with that in other TST systems such as FAMACHAß (Van Wyk and Bath, 2002) and could be lower than required to optimize the trade-off between nematode control and selection for resistance. However, the proportion of the parasite population that must be left untreated in order to slow anthelmintic resistance has not been determined. The practical uptake of TST is highly dependent on farmer perception of production loss, with most reluctant to engage with this strategy when risks to production are unknown. Once conservative application of TST has been demonstrated on a farm with good results in terms of production, there is scope for refining the system and increasing the proportion left untreated in future years. The expectation prior to this study was that animals showing higher than average weight gain would have lower than average levels of nematode infection, and that weight gain as an indicator would enable parasite overdispersion to be exploited in selective treatment. However, comparison of faecal egg counts in treated and untreated lambs suggests that weight gain was largely independent of nematode burden in the animals sampled. This is in spite of moderately high average faecal egg counts that would normally be consistent with a risk of sub-clinical disease. It is likely that regular treatment of the flock kept levels of infection below the level at which it would have detectable effects on weight gain, and that untreated lambs benefited from this reduced level of challenge. The continued egg output from untreated lambs is of course consistent with the aims of TST, and the fact that faecal egg counts were not lower in animals selected for non-treatment underlines their benefit in terms of generating refugia. However, more vigorous selection such that a larger proportion of lambs is left untreated could result in epidemiologically important
64
K.A. Stafford et al. / Veterinary Parasitology 164 (2009) 59–65
levels of pasture contamination from untreated animals, and the risk of production loss across the flock. This tradeoff is inherent to the TST approach. Given the unexpected lack of bias towards lower egg counts in individuals selected for non-treatment, incorporation of routine faecal egg count monitoring alongside weighing, e.g. through comparison of pooled egg counts from treated and untreated groups (Morgan et al., 2005) might be advisable. The rate of weight gain in lambs selected for nontreatment was not significantly depressed either in the short-term or over the whole grazing season. This is encouraging, since it suggests that production loss can be completely avoided on commercial farms while using TST. However, weight gain of untreated animals was compared with that of peers matched by breed and sex. It is therefore possible that growth of untreated animals was impaired when compared with their potential growth, and actual performance was reduced to the flock average through non-treatment. Since these could be the best genetic stock and therefore the most desirable replacement animals, or the first to be ready for market, the commercial penalty of reduced growth in these animals could be disproportionate. This requires further study. In a study of TST in New Zealand, untreated lambs grew more slowly than treated lambs over some periods, but overall penalties in average weight gain were not discernible, as individuals whose growth fell behind that of the group were unlikely to be selected twice for non-treatment (Leathwick et al., 2006a). In the same study, groups of lambs subjected to TST by non-treatment of the 10% heaviest individuals on five occasions matched the average growth of those given five whole-group treatments. The implications of reduced rate of growth differ between production systems. Thus, for meat lamb production rapid growth on grass and early dispatch to market is of key importance, whereas for breeding animals slower growth in the face of moderate nematode challenge can be made up in the future, and the benefits of immunity are carried forward in terms of reduced impact of parasites on ewes and reduced pasture contamination for lambs. Further field studies of TST should include a range of breeds and production systems, including those geared to rapid finishing of meat lambs. Since multiple management practices can affect the rate of selection for resistance, TST should moreover be integrated with other control strategies, especially grazing management and treatment regimes for older age classes, to enhance its role in generating pasture refugia for anthelmintic susceptibility (Coles, 2002; Lawrence et al., 2006a; Leathwick et al., 2006b). The uptake of TST will depend strongly on the ease, rapidity and cost of the relevant indicator for treatment. While automated weighing makes short-term changes in weight gain a practical proposition as a selective indicator, the high cost of investment in equipment is likely to exclude many producers. Moreover, farmers would be reluctant to base treatment purely on weight gain if there are other signs that relatively fast-growing animals would benefit from treatment. In the present study, treatment based on farmer judgement alone, regardless of weight gain, was permitted, and the sign most commonly used to
justify such treatment was breech soiling. Faecal egg counts on animals identified in this way suggested that their nematode burdens were elevated (unpublished data). Visual assessment of breech soiling could therefore be usefully incorporated into weight-based TST, or even used as an alternative where automated weighing facilities are not available. This should form the basis for future work. Automated weighing systems could also be improved to ease their use in TST, in particular through software development to automatically compute incremental weight gain, and enable simple recording and recall of information on individual anthelmintic treatment. The present study demonstrates that weight-based TST can be conservatively applied to commercial sheep farms in temperate areas with no detectable loss in productivity, and with reasonably high levels of egg production from untreated individuals. Further work is needed to determine the proportion of the flock that can be left untreated without production loss, although this is likely to be highly variable between production systems and with epidemiological conditions. Importantly, the effect of TST on the rate of development of anthelmintic resistance needs to be determined if farmers are to be persuaded to adopt this approach. Development of a theoretical framework for predicting such effects is under way (Gaba et al., 2006), but field validation will of course be necessary, and will have to include commercial holdings if farmers are to be persuaded of its applicability. Although the present study reports relatively small reductions in the number of treatments given, this is an improvement on the present situation in which whole-flock treatments are very much the norm. It is not surprising that farmers prefer to take the short-term benefits of intensive anthelmintic treatment when the production implications of TST are so poorly characterized, even though the longer range costs of this strategy in terms of anthelmintic resistance are increasingly apparent (Bath, 2006; Waghorn et al., 2006; Lawrence et al., 2006b). Perhaps this situation will change when overt failure of anthelmintic therapy becomes more common, but waiting until this occurs will result in high levels of suffering and economic loss. Demonstration of TST using weight gain and also other more practical and accessible indicators for treatment, with minimal production loss, should be prioritized if this approach is to progress beyond the research stage. Conflict of interest There are no conflicts of interest. Acknowledgements This work was funded by the European Union FP6 under the PARASOL project. We gratefully acknowledge the help and co-operation of Peter Baber with data collection, and useful discussions with Richard Webber and staff at Shearwell data systems. Research within this paper was either partially or wholly supported by the EU Parasol project (FOOD-CT2005-022851) and by other governmental/institutional/ university funding.
K.A. Stafford et al. / Veterinary Parasitology 164 (2009) 59–65
References Anderson, R.M., May, R.M., 1978. Regulation and stability of host–parasite population interactions. 1. Regulatory processes. J. Anim. Ecol. 47, 219–247. Barger, I.A., 1985. The statistical distribution of trichostrongylid nematodes in grazing lambs. Int. J. Parasitol. 15, 645–649. Bath, G.F., 2006. Practical implementation of holistic internal parasite management in sheep. Small Rumin. Res. 62, 13–18. Broughan, J.M., Wall, R., 2007. Faecal soiling and gastrointestinal helminths infection in lambs. Int. J. Parasitol. 37, 1255–1268. Blake, N., Coles, G.C., 2007. Flock cull due to anthelmintic-resistant nematodes. Vet. Rec. 161, 36. Churcher, T.S., Ferguson, N.M., Basanez, M.G., 2005. Density dependence and overdispersion in the transmission of helminth parasites. Parasitology 131, 121–132. Churcher, T.S., Basanez, M.G., 2008. Density dependence and the spread of anthelmintic resistance. Evolution 62, 528–537. Coles, G.C., 2002. Sustainable use of anthelmintics in grazing animals. Vet. Rec. 151, 165–169. Eysker, M., van Doorn, D.C.K., Lems, S.N., Weteling, A., Ploeger, H.W., 2006. Frequent de-worming in horses; the advantage often more damage than good. Tijdschrift voor dier Geneeskunde 131, 524–530. Gaba, S., Ginot, V., Cabaret, J., 2005. Modelling macroparasite aggregation using a nematode-sheep system: the Weibull distribution as an alternative to the Negative Binomial distribution? Parasitology 131, 393–401. Gaba, S., Cabaret, J., Ginot, V., Silvestre, A., 2006. The early drug selection of nematodes to anthelmintics: stochastic transmission and population in refuge. Parasitology 133, 345–356. Hoste, H., Le Frileux, Y., Pommaret, A., 2002. Comparison of selective and systematic treatments to control nematode infection of the digestive tract in dairy goats. Vet. Parasitol. 106, 345–355. Jackson, F., Miller, J., 2006. Alternative approaches to control—Quo vadit? Vet. Parasitol. 139, 371–384. Kaplan, R.M., Burke, J.M., Terrill, T.H., Miller, J.E., Getz, W.R., Mobini, S., Valencia, E., Williams, M.J., Williamson, L.H., Larsen, M., Vatta, A.F., 2004. Validation of the FAMACHAß eye colour chart for detecting clinical anemia in sheep and goats on farms in the southern United States. Vet. Parasitol. 123, 105–120. Kyriazakis, I., Anderson, D.H., Oldham, J.D., Coop, R.L., Jackson, F., 1996. Long-term subclinical infection with Trichostrongylus colubriformis: effects on food intake, diet selection and performance of growing lambs. Vet. Parasitol. 61, 297–313. Lawrence, K.E., Leathwick, D.M., Rhodes, A.P., Jackson, R., Heuer, C., Pomroy, W.E., West, D.M., Waghorn, T.S., Moffat, J.R., 2006a. Management of gastrointestinal nematode parasites on sheep farms in New Zealand. N. Z. Vet. J. 55, 228–234.
65
Lawrence, K.E., Rhodes, A.P., Jackson, R., Leathwick, D.M., Heuer, C., Pomroy, W.E., West, D.M., Waghorn, T.S., Moffat, J.R., 2006b. Farm management practices associated with macrocyclic lactone resistance on sheep farms in New Zealand. N. Z. Vet. J. 54, 283–288. Leathwick, D.M., Waghorn, T.S., Miller, C.M., Atkinson, D.S., Haack, N.A., Oliver, A.M., 2006a. Selective and on-demand drenching of lambs: impact on parasite populations and performance of lambs. N. Z. Vet. J. 54, 305–312. Leathwick, D.M., Miller, C.M., Atkinson, D.S., Haack, N.A., Alexander, R.A., Oliver, A.M., Waghorn, T.S., Potter, J.F., Sutherland, I.A., 2006b. Drenching adult ewes: implications of anthelmintic treatments pre- and post-lambing on the development of anthelmintic resistance. N. Z. Vet. J. 54, 297–304. Louie, K., Vlassoff, A., Mackay, A.D., 2007. Gastrointestinal nematode parasites of sheep: a dynamic model for their effect on liveweight gain. Int. J. Parasitol. 37, 233–241. MAFF, 1986. Manual of Veterinary Parasitological Laboratory Techniques. Reference Book 418, 3rd ed. HMSO, London. Medley, G.F., Guyatt, H.L., Bundy, D.A.P., 1993. A quantitative framework for evaluating the effect of community treatment on the morbidity due to ascariasis. Parasitology 106, 211–221. Michel, J.E., 1985. Strategies for the use of anthelmintics in livestock and their implications for the development of drug resistance. Parasitology 90, 621–628. Morgan, E.R., Cavill, L., Curry, G.E., Wood, R.M., Mitchell, E.S.E., 2005. Effects of aggregation and sample size on composite faecal egg counts in sheep. Vet. Parasitol. 131, 79–87. Quinnell, R.J., Medley, G.F., Keymer, A.E., 1990. The regulation of gastrointestinal helminth populations. Phil. Trans. Royal Soc. Lond. B 330, 191–201. Russel, A., 1984. Body condition scoring of sheep. In Practice 6, 91–93. Sargison, N.D., Jackson, F., Bartley, N.D., Moir, A.C.P., 2005. Failure of moxidectin to control benzimidazole-, levamisole- and ivermectinresistant Teladorsagia circumcincta in a sheep flock. Vet. Rec. 156, 105– 109. Sissay, M.M., Asefa, A., Uggla, A., Waller, P.J., 2006. Anthelmintic resistance of nematode parasites of small ruminants in eastern Ethiopia: exploitation of refugia to restore anthelmintic efficacy. Vet. Parasitol. 135, 337–346. Van Wyk, J.A., 2001. Refugia—overlooked as perhaps the most potent factor concerning the development of anthelmintic resistance. Onderstepoort J. Vet. Res. 68, 55–67. Van Wyk, J.A., Bath, G.F., 2002. The FAMACHAß system for managing haemonchosis in sheep and goats by clinically identifying individual animals for treatment. Vet. Res. 33, 509–529. Waghorn, T.S., Leathwick, D.M., Rhodes, A.P., Lawrence, K.E., Jackson, R., Pomroy, W.E., West, D.M., Moffat, J.R., 2006. Prevalence of anthelmintic resistance on sheep farms in New Zealand. N. Z. Vet. J. 54, 271–277.