2

Sustainable approaches for managing haemonchosis in sheep and goats

Final Report of

FAO Technical Co-operation Project

in South Africa

2001

Sustainable approaches for managing haemonchosis in sheep and goats

Final Report of Food and Agriculture Organization (FAO) Technical Co-operation Project No. TCP/SAF/8821(A)

Prepared by : G F Bath

Department of Production Animal Studies, Faculty of Veterinary Science, University of Pretoria, Private Bag X04, Onderstepoort, 0110, South Africa

J W Hansen Animal Production and Health Division, Food and Agriculture Organization of the United Nations,

Viale delle Terme di Caracalla, Rome 00100, Italy Present address : The Royal Danish Embassy, GPO Box 2056, Dhaka-1212, Bangladesh

R C Krecek Department of Veterinary Tropical Diseases, Faculty of Veterinary Science, University of Pretoria,

Private Bag X04, Onderstepoort, 0110, South Africa

J A van Wyk Department of Veterinary Tropical Diseases, Faculty of Veterinary Science. University of Pretoria,

Private Bag X04, Onderstepoort, 0110, South Africa

A F Vatta Division of Parasitology, Agricultural Research Council-Onderstepoort Veterinary Institute, Private

Bag X05, Onderstepoort, 0110, South Africa

With the assistance of : L M Michael, M O Stenson, E F van Wijk : Division of Parasitology, Onderstepoort Veterinary Institute, Private Bag X05, Onderstepoort, 0110, South Africa H M Boshoff, D T Durand : Department of Veterinary Tropical Diseases, Faculty of Veterinary Science, University of Pretoria, Private Bag X04, Onderstepoort, 0110, South Africa A Sutherland, J T K van der Toorn : Department of Production Animal Studies, Faculty of Veterinary Science, University of Pretoria, Private Bag X04, Onderstepoort, 0110, South Africa B A Letty, J F de Villiers : Farming Systems Research Section, Cedara Agricultural Experimental Station, KwaZulu-Natal Department of Agriculture, Private Bag X9059, Pietermaritzburg, 3200, South Africa H T Groeneveld, R J Grimbeek, M J van der Linde : Department of Statistics, Faculty of Natural and Agricultural Sciences, University of Pretoria, Pretoria, 0002, South Africa G S Molebiemang, W Motswatswe : Department of Agriculture, Conservation and Environment, North-West Province, Private Bag X2039, Mmabatho, 2735, South Africa S H Barnard : P O Box 1672, Witbank, 1035, South Africa S A Bisset : AgResearch, Wallaceville Animal Research Centre, P O Box 40063, Upper Hutt, New Zealand F S Malan : Intervet Malelane Research Unit, P O Box 124, Malelane, 1320, South Africa

C D Wessels : Vygeboom, P O Box 136, Badplaas, 1190, South Africa

The designations employed and the presentation of material in this publication do not imply the expression of any opinion whatsoever on the part of the Food and Agriculture Organization of the United Nations concerning the legal status of any country, territory, city or area of its authorities, or concerning the delimitation of its frontiers or boundaries. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying or otherwise, without the prior permission of the copyright owner. Applications for such permission, with a statement of the purpose and extent of the reproduction, should be addressed to the Director, Information Division, Food and Agriculture Organization of the United Nations, Viale delle Terme di Caracalla, 00100 Rome, Italy. Further copies of this report are also available at this address.

© FAO 2001

Contents Introduction ......................................................................................................... 1 1. Managing haemonchosis in sheep using treatment according to the degree of

anaemia, and a method for reversion of anthelmintic resistance Section 1 : General materials and methods ................................................... 2

2. Section 2 : Anthelmintic resistance in South Africa ........................................ 4

3. Section 3 : Clinical evaluation of anaemia in sheep : Early (Badplaas) trials, before development of the FAMACHA© system ............................. 6

4. Section 4 : Accuracy of anaemia estimation using the FAMACHA© system ....... 10

5. Section 5 : Validation of the FAMACHA© system on commercial sheep farms ... 21

6. Section 6 : Analysis of phenotypic and genotypic relationships of FAMACHA©

data compared to other production data ...................................... 39

7. Section 7 : Dilution of anthelmintic resistant with susceptible worm strains : Preliminary results in the field ..................................................... 57

8. Survey of nematophagous fungi from agricultural environments ........................ 60 9. Haemonchus in goats farmed under resource-poor conditions in South Africa :

Incidence, effect on haematocrit and treatment according to ocular mucous membrane colour .......................................................................................... 63

10. Conclusions and recommendations : Tools for integrated worm management

for sheep and goats ...................................................................................... 77 References ........................................................................................................... 86 Acknowledgements ............................................................................................... 89 Appendices Appendix 1 : FAMACHA© chart ......................................................................... i Appendix 2 : Practical implementation of the FAMACHA© system ........................ iii Appendix 3 : Information pamphlet .................................................................. viii Appendix 4 : �Histogram� ................................................................................ xiii Appendix 5 : Fb5 � Marking ............................................................................. xv Appendix 6 : Perceptions questionnaire for farmers ........................................... xvii Appendix 7 : Perceptions questionnaire for veterinarians .................................... xx Appendix 8 : Instructions for bleeding of sheep for demonstrations ..................... xxii Appendix 9 : Perceptions questionnaire for farm workers ................................... xxv Appendix 10 : Proficiency certificate ...................................................................xxvii

1

Introduction Africa is faced with the challenges of growing population numbers and shortages of resources. Some facts illustrate this. The world population in 1992 was 5.5 billion and is estimated to be 11 billion by 2050. Africa mirrors this scenario. Africa also has the lowest annual per capita income of US $ 650 compared with Asia (US $ 2,490) and the USA (US $ 27,000). In addition, 6 of the 18 water-scarce nations in the world are in Africa and, during the next 30 years, 6 more in Africa are projected to join these figures. Only 15 % of South Africa is arable and most of this land has only limited opportunities for irrigation and subsequent improvement for crop production. The situation is becoming even more difficult due to the rapid increase in the human population (now at more than 42 million) and the growing urbanization rate (more than 55 % living in cities). As importation of large amounts of food will not be feasible because of a shortage of foreign exchange, there will be an urgent need to increase animal production and productivity in the very near future. Animal production should be enhanced in all suitable marginal lands in an attempt to cover the envisaged food deficiency of the country. At present, the South African animal sub-sector plays an important role in enhancing food security. In South Africa, sheep and goat populations are approximately 32 million, with 27 million sheep and 5 million goats. The total global value of the meat production from sheep and goats is in the range of US $ 210 � 250 billion. The South African Government is greatly concerned about the status of health of livestock, particularly of those belonging to resource-poor livestock producers. Both the commercial sheep producers and the resource-poor farmers experience serious animal disease-related problems. Gastrointestinal nematode infections, particularly with Haemonchus contortus, constitute one of the most important constraints to small ruminant production. The situation is even more critical in areas in which levels of animal production are poor and farmers have limited access to relevant information and stock remedies. This report is a culmination of the results of two years of testing, modification and validation of new technologies of worm management in sheep and goats within the context of both commercial and resource-poor (small scale or subsistence) farming systems. The aim of this project, in which regional, national and international researchers were involved, was to test and modify, over a two-year period, new technologies of worm control in sheep and goats within the context of both commercial and resource-poor farmers. This report presents validated, sustainable and integrated parasite control strategies based on the project findings. It is considered that these will provide additional instruments for the �toolbox� needed for the �decision support systems� approach which will aid the farmer to make a well-informed decision for managing internal parasites in sheep and goats.

2

Managing haemonchosis in sheep using treatment according to the degree of anaemia, and a method for

reversion of anthelmintic resistance

Section 1 General materials and methods

Introduction The FAMACHA© System was developed before the commencement of the FAO Technical Cooperation Project (Malan and Van Wyk, 1992; Bath et al., 1996; Van Wyk et al., 1997b) to make it practical for the farmer to identify clinically and drench only those sheep in a flock that are unable to withstand heavy challenge with Haemonchus spp., a blood-sucking worm instead of (as is usually the case) treating the entire flock when there are indications of overwhelming infection in some of the individuals. In addition, in this way selection of the worms of species for anthelmintic resistance can be reduced considerably (Besier, 1997). The name of the system was coined at the suggestion of one of us (GFB) to honour its originator, Francois (Faffa) Malan : FAffa MAlan CHArt (Van Wyk et al., 1997b). A reproduction of the FAMACHA© chart is shown in Appendix 1. In the early 1990s a trial was conducted at Badplaas (Mpumalanga) to test whether it was possible to evaluate the degree of clinical anaemia caused by infection with the worm clinically by classifying the colour of the ocular mucous membranes (Malan and Van Wyk, 1992; Malan et al., 2001). In this trial the colour variations were judged subjectively, without colour standards. Thereafter, when encouraging results were obtained, a colour chart (called the FAMACHA© chart) was developed, with which the colour of the socular mucous membranes of the animal could be compared (Bath et al., 1996). After the initial trials in the early 1990s, we commenced on-farm training of veterinarians, technicians, farmers and stockmen in 1997. Subsequently we were able to continue validation trials, using funds obtained, inter alia from the FAO, and eventually the project involved, at different levels of intensity, 13 farms in Gauteng, Mpumalanga, Free State and KwaZulu-Natal Provinces. On all of these farms we had the wonderful privilege to have farmers who were prepared: to learn techniques such as haematocrit determinations; to allow investigations that were experimental, with no guarantee that there would not be deaths in their animals; to risk production losses (growth, wool and reproduction); to allow frequent handling of their sheep; to permit severe stressing of the animals after separating resistant/resilient animals from susceptible individuals; and to permit a seemingly endless process of sampling. The intensity of the investigations conducted varied from farm to farm according to the frequency with which the farms could be visited, whether individual sheep record-keeping was or was not done on the farm, and the extent to which the farmers were prepared to invest time and effort in the project.

Aims The overall aim was to test to what extent estimates of the FAMACHA© categories could be used with confidence under a wide variety of conditions by the farmer and the stockman to ensure that individual animals that required drenching would be identified for timely treatment, without the animals being subjected to undue stress during the process. The aims of the trials also varied according to the level of intensity at which we were able to work on each of the farms concerned.

3

The level of intensity at which the aims of the trials were applied did however, vary to some extent on each of the farms concerned. The major aims were : 1. To test the reliability of clinical anaemia evaluation in sheep, as determined by veterinarians,

technicians, farmers and stockmen against measured haematocrit values in sheep 2. To develop methods and materials for training and evaluating persons who are to apply the

system and/or train others in its use 3. To test the FAMACHA© for its ability to identify individual animals unable to withstand worm

challenge and to compare the results obtained with existing indices of worm infections 4. To test whether the FAMACHA© worm control management system could be applied at a

level where phenotypic differences between animals are magnified by heavy worm challenge, without causing the death of any animals

5. To test applicability of the FAMACHA© system for :

a) Identifying individual animals for treatment, instead of drenching the whole flock when only some of the animals require treatment

b) Deciding if and when entire flocks of sheep need to be drenched, if at all c) Reducing selection pressure for anthelmintic resistance d) Selecting the most resistant/resilient animals and/or culling of the least resistant/resilient

sheep 6. To evaluate the opinion of farmers, farm workers and veterinarians concerning application of

the system after they had committed themselves to testing it for at least 1 �worm season�. 7. To test whether anthelmintic resistance can be lowered in practice, under field conditions, by

dilution of a resistant worm strain with a susceptible one.

Materials and methods Faecal worm egg counts (FECs), larval cultures, post mortem worm recovery and worm counting and identification of worms were performed according to Reinecke (1973). Infective larvae (L3) were identified according to Van Wyk et al. (1997a). Faecal egg count reduction tests (FECRT) were conducted according to Presidente (1985). Faecal specimens were cooled with ice after collection. In the laboratory they were maintained in refrigerators at ± 4°C until processed. The FAMACHA© card was used clinically to evaluate the degree of anaemia of the sheep in the various trials. All sheep were subjected to normal farming practices and routine measures then in use on each farm, with the addition of our experimental procedures. Standard electronic apparatus and techniques were used for microhaematocrit determination and weighing the animals.

4

Section 2 Anthelmintic resistance in South Africa

Introduction Anthelmintic resistance has reached the stage where we are experiencing progressive problems in South Africa in managing especially Haemonchus contortus infection in small ruminants with the available anthelmintics. The following comprises a summary of results of a series of investigations (Van Wyk et al., 1999). In a survey of 52 farms in Mpumalanga and KwaZulu-Natal Provinces, rafoxanide, levamisole, albendazole and ivermectin (representing the 4 anthelmintic activity groups available in South Africa) were used in Faecal Egg Count Reduction surveys to gauge the prevalence of anthelmintic resistance (Van Wyk et al., 1999). Every one of the farms surveyed had experienced problems with resistance, which was very serious on a large proportion of the farms (Table 2.1 and 2.2). Table 2.1 Strains of Haemonchus spp. resistant to 3 (or all 4) anthelmintic groups

Efficacy %a

<95 % <60 % <40 %

Locality Farms(n)

% of worm strains

Mpumalanga 26 69 (19) 16 (8) 8 (8)

Ermelo (1996)

KwaZulu-Natal

Mount Currie (1996)

Estcourt (1996)

Underberg (1996)

12

5

9

75 (8)

60 (60)

78 (0)

8 (8)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

Various (1992)b 28 60 (14) 11 (7) 7 (4)

Total/Mean 80 68.4 (20.2) 7.0 (6.2) 3.0 (2.4) aIn brackets : Resistance to compounds from all 4 groups. bGauteng, Mpumalanga, KwaZulu-Natal, Free State, Northern and Eastern Cape Provinces. Table 2.2 Rate of failure of the individual anthelmintics tested, against the 80 strains of Haemonchus spp. in 5 surveys in South Africa

% of worm strainsa Mean efficacy on 80 Farms

Alben Iver Lev Rfx

<95 % 79 73 23 89

<60 % 35 24 8 54

<40 % 22 16 5 40 aPercentage of the worm strains with a geometric mean susceptibility of <95 %, <60 % or <40 %, to albendazole (Alben), ivermectin (Iver), levamisole (Lev) and rafoxanide (Rfx).

5

On 68.4 % of the farms included in the survey, 3 of the 4 anthelmintics tested were <95 % effective, and on 20 %, all 4 were affected. Even more serious is that on 6.2 % and 2.4 % of the farms all 4 anthelmintics were <60 % and <40 % effective, respectively.

6

The results in Tables 2.1 and 2.2 do not imply that there are no effective anthelmintics remaining on the affected farms but rather that, because every nemicidal anthelmintic on the market in South Africa is related to one or more of the 4 tested, the continued use of any one of more than 60 registered anthelmintic trade names on the worst affected farms will not be sustainable even in the short term, unless steps are taken drastically to reduce further selection for anthelmintic resistance. Table 2.3 Resistance of Haemonchus spp. of sheep to various anthelmintics : Comparison of South Africa with 4 South American countries

Anthelmintic : % resistant worm strains (farms) Country Survey farms (n)

Alben Iver Lev Rfx Clos B+L Overall

South Africaa 80 79 73 23 89 - - 98f

Paraguayb 37 70 67 47 - - - ?

Uruguayc 242 61 1 29 - - - 93

Brazild 182 68 7 19 - 20 15 97

Argentinae 65 37 2 8 - - 5 46 aVan Wyk et al. (1999); bMaciel et al. (1996); cNari et al. (1996); dEchevarria et al. (1996); eEddi et al. (1996); fWhen analysed by the RESO method all 3 South African surveys show resistance on 100 % of the commercial forum investigated. When the results of surveys in 4 South American countries were reported anthelmintic resistance in Paraguay was described as the worst in the world (Maciel et al., 1996). However, despite the fact that, to give the anthelmintics the best chance to show efficacy, a more lenient method of analysis was used by Van Wyk et al. (1999) than the RESO method used in South America, the resistance in South Africa proved to be the worst for 3 of the 4 drugs tested (Table 2.3). This observation is supported by the fact that resistance to the following compounds was recorded first in South Africa : rafoxanide, closantel, disophenol, nitroxynil and ivermectin, as well as the first case of resistance of a worm strain simultaneously to compounds from 5 anthelmintic activity groups, namely benzimidazoles, imidasothiazoles, macrolactones, organophosphorus compounds and halogenated salicylanilides (Van Wyk et al., 1997). The fact that anthelmintic resistance has escalated to the present levels globally, indicates that the anthelmintics have been used incorrectly. It appears that this is because the worms in refugia (free-living on pasture) are not considered sufficiently in the drenching policy. While refugia is commonly mentioned in passing in most papers on anthelmintic resistance as a factor that may be of importance, almost without exception it is not incorporated into the final control/management proposals, which usually centre around much less important phenomena such as alternation of anthelmintics and ensuring that the animals are not given less than the recommended therapeutic dose rates (Van Wyk, 2001). Formerly we always thought in terms of worm �control�, but this is too finite and should be replaced in our thinking by worm �management�. For almost a century we have attempted to eradicate worms (Theiler, 1912), or at least to reduce them to levels where they can have no effect at all on animal production. But this has failed. So, we should rather find and develop ways to �live with� or be able to �farm with� worms, in a manner of speaking.

7

Section 3 Clinical evaluation of anaemia in sheep : early (Badplaas) trials, before development of the

FAMACHA© system

Introduction Waller (1997) described anthelmintic resistance as the biggest threat to grazing livestock production globally. The costs involved in developing new anthelmintics and researching them are so high (McKellar, 1994) that it is completely beyond the budgets of research institutions in South Africa. We therefore decided to concentrate on developing biological methods of worm management for delaying further escalation of resistance to existing anthelmintics. It is common knowledge that, while the ocular mucous membranes of healthy animals are red in colour, those of animals that are terminally ill from Haemonchus contortus infection are practically white as a result of the anaemia produces by this blood-sucking worm. As this pernicious worm species is the most important helminth species in small ruminants in most of South Africa, we investigated whether progressive paling of the conjunctivae could give a reliable indication of the magnitude of the worm infection in individual animals. The following trial (preliminary results of which have been published previously - Malan and Van Wyk, 1992; Van Wyk et al., 1997b; Malan et al., 2001) was the first of a series to test the hypothesis that various levels of anaemia could be graded clinically. The aim was to evaluate whether a system for identifying and drenching only individual animals that would have otherwise succumbed to haemonchosis, could be used to replace routine strategic and tactical drenching of the entire flock under conditions of heavy infection with H. contortus.

Methods A commercial farm near Badplaas in Mpumalanga Province was selected for the trial because of severe problems with haemonchosis on irrigated pastures. The climate is mild in winter and hot and humid in summer. A flock of 388 sheep ewes on irrigated kikuyu (Pennisetum clandestinum) pasture was drenched with levamisole at a dosage of 7.5 mg kg-1 at the commencement of the trial which was run for 125 days, from March to July 1991. During this time routine drenching of the whole flock was terminated and replaced by salvage treatment of severely affected animals only. Hence conditions were ideal for the proliferation of H. contortus. The sheep were classified into one of the following 5 categories according to the colour of the conjunctival mucosae : red, red-pink, pink, pink-white or white. They were bled for haematocrit determination, and only those with a packed cell volume of 15 % or lower were dewormed with levamisole (�Tramisol�, Hoechst Roussel Vet � now Intervet SA) at the dosage level mentioned above. At every weekly examination the sheep were also inspected for the presence of submandibular oedema (�bottle-jaw�). Each weekly evaluation of the sheep was usually carried out by only one person but during the course of the trial several individuals were concerned. A total of 166 ewes (45 % of the flock) lambed : 112 shortly before commencement of the trial (and were thus lactating during the trial), 51 soon afterwards (hence pregnant during part of the trial), and 3 both before and after the trial. While not part of the trial, the ± 1.500 other sheep on the farm were similarly examined but were not bled for haematocrit evaluations. Those judged to require treatment were drenched.

8

Results The results are summarised in Tables 3.1-3.6. None of the animals in the trial, or among the ± 1,500 other sheep mentioned, died from haemonchosis during the course of the trial. On about 10 occasions (not recorded separately at the time) sheep developed bottle-jaw, and were drenched on the strength of this observation, independent of changes in the colour of the conjunctivae. Table 3.1 Correlation between haematocrit and clinical anaemia

category, ignoring reproductive classes and drenching Day of trial Pearson correlation coefficient between

haematocrit and clinical anaemia estimatea

0 -0.289 (8.4 %)

41 -0.624 (38.9 %)

51 -0.535 (28.6 %)

63 -0.456 (20.8 %)

92 -0.506 (25.6 %)

136 -0.507 (25.7 %)

Total : Day 41-136 -0.580b

aIn brackets : r2x100 (the percentage to which the variation is declared by the values correlated); All correlations are highly significant (P<0.0001). bExcluding Day 0 value. Table 3.2 Success rates of estimates of anaemia

Estimate error Percentage of 2367 observations

1. Correct category estimated 54.6 %

2. Estimate too high (more serious)

a) Category 1 vs 2

b) Category 1 vs 3

c) Category 1 vs 4

d) Category 1 vs 5

25.0 %

8.0 %

1.9 %

0.2 %

Total % (above, plus unlisted) 40.3 %

3. Estimate too low (less serious)

a) 1 category

b) 2 or more categories

3.3 %

1.8 %

Total % 5.1 %

9

Table 3.3 Drenching during the trial Ewes No of drenches per ewe (% of class)

Class Number (n) 0 1 2 3 4

Dry 206 83.0 % 13.6 % 2.4 % 0.5 % 0.5 %

Lactating 112 44.6 % 32.1 % 16.1 % 5.4 % 1.8 %

Pregnant 51 70.6 % 21.6 % 5.9 % 2.0 % 0 %

Lactating + Pregnant

3 100.0 % 0 % 0 % 0 % 0 %

Total number(% of total number)

372 260 (70 %)

75 (20 %)

26 (7 %)

8 (2 %)

3 (1 %)

Table 3.4 Summary of Table 3.3

Class of ewe Not drenched Drenched

Dry 171 (83.0 %) 35 (17.0 %)

Lactating 50 (44.6 %) 62 (55.4 %)

Pregnant 36 (70.6 %) 15 (29.4 %)

Lactating + Pregnant 3 (100 %) 0 (0 %)

Total 260 (69.9 %) 112 (30.1 %)

Table 3.5 Anaemia : Drenched/Not drenched

Group Number (n) Percentage Mean haematocrit ±

SD

Mean clinical anaemia

category ± SD

Not drenched 260 69.9 % 28.4 ± 3 1.13 ± 0.4

Drenched 112 30.1 % 23.8 ± 3 1.46 ± 0.5

Table 3.6 Mean haematocrits and clinical evaluation category values

for the different lambing classes, irrespective of the number of treatments received

Effect Number (%) of ewes Mean clinical category (± S.D.)

Mean haematocrit (± S.D.)

Dry 206 (55.4 %) 1.18 (± 0.4) 28.17 (± 3.3)

Lactating 112 (30.0 %) 1.57 (± 0.6) 24.46 (± 3.5)

Pregnant 51 (13.7 %) 1.23 (± 0.3) 27.82 (± 3.6)

Lactating + Pregnant 3 (0.8 %) 1.10 (± 0.2) 29.67 (± 3.2)

While the various clinical anaemia categories correlate highly significantly with the haematocrit values of the various sheep (Table 3.1), it is important to note that a total of 45.4 % of the estimates fell outside the corresponding haematocrit categories. However, 38.1 % of the �mistakes� were not serious, as the haematocrits of the animals concerned were not low enough to merit drenching. On the other hand, in 2.1 % of cases the errors could have had serious consequences, had haematocrit determinations not been done and the animals drenched, since

10

the sheep could otherwise perhaps have died. The presence of bottle-jaw served as a �safety factor� to identify animals requiring treatment that were not detected by examining the conjunctivae. Furthermore, Table 3.2 shows that in only 5.1 % of cases was the anaemia less severe than estimated, compared to 40.3 % in which the anaemia was more severe. Despite ideal conditions for H. contortus, 83 % of the dry ewes did not require to be drenched during the period of heavy worm challenge (Table 3.3). Even in the case of the pregnant animals, less than 30 % could have succumbed had they not been drenched. On the other hand, the lactating sheep were by far the most susceptible of the reproductive classes, since only 44.6 % of the ewes were able to withstand the worm challenge unaided. It should be noted that the discrepancy between the total number of sheep in the flock (388) and the totals listed in the tables (372) is caused by some missing values that disqualified some of the animals from inclusion in the analyses. Three of the ewes lambed twice during the course of the trial but despite this, their mean haematocrit values were even higher than those of the dry ewes.

Discussion When judging the results of the Badplaas trial, it should be kept in mind that at the time it was conducted, there was neither a colour standard against which the observed colours of the conjunctivae could be compared [for example, the FAMACHA© chart that was developed later (Bath et al., 1996; Van Wyk et al. 1997b], nor was the range of possible haematocrit values divided into categories. In other words, there was no standard for the clinical anaemia categories when the trial was conducted; the colours were merely classed as red (category 1), red-pink (category 2), pink (category 3), pink-white (category 4) or white (category 5). Despite this fact, it is encouraging that the correlations between the clinical classification (developed later for the FAMACHA© system) of the state of anaemia of the sheep in the trial and the haematocrits of the animals were highly significant (Table 3.1). On the other hand, it is important to note that the r2 values were not high enough to give reasonable surety that the haematocrit values of individual animals could be predicted with a high degree of confidence from their clinical classifications. In general, the clinical classifications of the sheep in the trial tended to underestimate the real degree of anaemia from which individual animals were suffering. This indicates that at the critical level of 15 % haematocrit, the colour of the conjunctiva appeared redder to the evaluator than reflected by the haematocrit � in other words that treatment should have been undertaken before the mucosae appeared extremely pale. A total of 2.1 % of the sheep in categories 4 and 5 (prominent anaemia) were seriously at risk of succumbing to haemonchosis, as they were classified in the �healthiest� categories (Table 3.2). Despite this fact, it is encouraging that none of the sheep in the trial died from haemonchosis. The reason for this appears to be that animals that were missed in one week, were correctly identified at an examination thereafter, and were duly treated. The fact that the persons who did the clinical evaluations varied from time to time, may also have played a role in making the classifications less accurate. Nevertheless, it sounded a warning that this required attention. Another matter that was shown to require attention, was that high burdens of Trichostrongylus spp. and Ostertagia spp. were recorded by Malan et al. (2001) from 7 of 14 sheep that were slaughtered at the conclusion of the Badplaas trial. In the light of the trial plan it was to be expected that the sheep that required salvage drenching would have lower mean haematocrits than the others, as shown in Table 3.5, but the high mean haematocrit values of 25-31 % in the undrenched animals (28 % ± 3) were unexpected.

11

Section 4 Accuracy of anaemia estimation using the FAMACHA©

system

Introduction One of the most important prerequisites for field application of the FAMACHA© concept is that this method of clinical evaluation can give a relatively reliable approximation of the haematological status of animals suffering from haematophagous worm infection, especially at lower haematocrit values. Therefore, after they had been trained, persons with or without further experience in applying the system were tested by comparing their FAMACHA© classifications of sheep with the corresponding haematocrit values in different flocks. The haematocrit was regarded in these investigations as the �gold standard� for evaluating the haematological status of the animals involved.

Methods As reported in Section 3 above, when the preliminary (Badplaas) trial described above was conducted, the FAMACHA© method of clinical evaluation of anaemia had not been formulated. Furthermore, the 5 clinical categories then used had not been related to specific haematocrit values. Nevertheless, for the purpose of comparison with the later current, the Badplaas results above were classified to correspond with the FAMACHA© categories formulated after conclusion of the first (Badplaas) trial (Table 4.1) : Category 1 (haematocrit ≥ 28 %), 2 (23-27 %), 3 (18-22 %), 4 (13-17 %) or 5 (haematocrit ≤ 12 %). In an attempt to determine whether the haematocrit ranges were correctly chosen for the various FAMACHA© categories, the FAMACHA© evaluations (for approximately 170 sheep) of 2 experienced persons were plotted on a graph, against the corresponding haematocrit values. From the graph the ranges of haematocrit percentages were read from the dividing lines between the FAMACHA© categories obtained. Haemonchus spp. largely dominated the worm infections in the sheep in the trials, both preliminary and subsequent, although Trichostrongylus colubriformis was also important in some individual animals on two of the farms.

Training In all but the preliminary trial the persons tested for their ability to estimate the state of anaemia from the colour of the ocular mucous membranes were trained as follows : As described in more detail in Appendix 2, training took the form of an initial information session (varying in content according to the level of education and literacy of the trainees), followed by a description of the FAMACHA© system, and a demonstration using cut-outs from the FAMACHA© chart and sheep with varying haematocrits and degrees of clinical anaemia. Sheep with as large a range of haematocrits as possible were selected from available flocks at each locality where the training was done. If necessary, sheep with low haematocrit values were obtained by exsanguination of predetermined volumes of blood, followed by fluid replacement when necessary (see Appendix 8 for details). In each test that followed a training session (Tests 5-9, Tables 4.3, 4.4 and 4.5) the animals were presented at random to the participant trainers. With the exception of Test 9 (Tables 4.3, 4.4 and 4.5), one sheep at a time was held on its side on a table in strong, natural light, and as each trainee passed by, one eye was opened to expose mainly the conjunctiva on the inside of the lower eyelid, while the upper lid was depressed, as

12

described in the FAMACHA© instructions (Appendix 3). At the same time the FAMACHA© colour chart was held next to the eye, while continuously being moved to-and-fro (with the colour bars of the chart being held close to the exposed conjunctiva), so that each trainee had the chance to see each colour bar in juxtaposition with each sheep�s conjunctiva. The trainees were repeatedly requested to write down their evaluation without making any comments, to prevent the others from being influenced. In Test 9 the 20 sheep used were held, 5 at a time, each by a different helper with the necessary experience. The eyes of the sheep were opened as described above by the different helpers, but the sheep were restrained in a standing position. Another difference from the first 8 tests was that 2 of the sheep used for prior demonstration (in Test 9) were also included in this test, and the persons who may have noted the numbers of these 2 sheep in spite of requests not to do so, would have been influenced. The experienced persons (Tables 4.3, 4.4 and 4.5) evaluated each sheep clinically in turn (with the eye being opened either by an assistant, or by the evaluator himself), whereafter the animals were bled for haematocrit determination on the same day.

Accuracy evaluation Two methods of scoring (see Table 4.2 and below) were used to evaluate the extent to which success was achieved with the clinical classification of the colours of the conjunctival mucous membranes of sheep into the 5 haematocrit categories set for the test (Table 4.1). In Table 4.3 the results of tests with considerably smaller groups, by experienced and inexperienced persons are compared by both methods. While the former persons classified relatively large groups of sheep after some months of experience, the trainees were tested after the system had been introduced to them after only one training session (per group) of about an hour, more or less half of which comprised specific training in the application of the FAMACHA© technique. Table 4.1 Various sets of haematocrit ranges used for evaluating the

success rate of clinical classifications, using the FAMACHA© system

Haematocrit ranges in subsequent trials FAMACHA© category Clinical evaluation in preliminary Badplaas

triala Initial ranges set Ranges obtained from line graph

1 Red ≥ 28 % ≥ 31 %

2 Red-Pink 23-27 % 21-30 %

3 Pink 18-22 % 16-20 %

4 Pink-White 13-17 % 12-15 %

5 White ≤ 12 % ≤ 11 % aBefore the FAMACHA© concept had been initiated and refined. A correct classification is awarded a score of 0 in Method 1, rising according to the extent to which the answer deviates from reality. In contrast, a correct answer is allotted a score of 10 in Method 2, diminishing thereafter in relation to the degree of divergence from the correct haematocrit category, with negative values being given for extreme deviations. For both methods a variation was introduced to increasingly penalise those errors that would tend to endanger the life of the sheep more seriously, because the estimated FAMACHA© category indicated a �safer� haematocrit value (thus a higher FAMACHA© category), than the observed percentage (Table 4.2).

13

The score allocation and penalisation is explained further by way of the following example, that is illustrated in Table 4.2. Given a sheep with a haematocrit of 15 %, then, if the FAMACHA© clinical classification was correct (Category 4), in Method 1 a value of 0 would be allocated for both unpenalised and penalised evaluation, while the corresponding value for Method 2 was 10 for both. However, if the clinical evaluation was 3 for the same sheep, then the unpenalised value for Method 1 was 3, and penalised 4.5, with the corresponding values of 7 (unpenalised) and 1 (penalised), for Method 2, etc. Hence the perfect score for Method 1 will be 0 for both the unpenalised and penalised evaluation, irrespective of the number of sheep examined, while the perfect score in Method 2 equals the number of sheep, n, times 10.

Results Table 4.2 Example of scores allocated for evaluating the rate of

success in applying the FAMACHA© system HAEMATOCRIT (%)a FAMACHA© category

Method 1 Method 2

27 % 27 %

26 % 26 %

25 % 25 %

24 % 24 %

2

23 % [8] {16} 23 % [-2] {-6}

22 % 22 %

21 % 21 %

20 % 20 %

19 % 19 %

3

18 % [3] {4.5} 18 % [7] {1}

17 % 17 %

16 % 16 %

15 % [0] {0} 15 % [10] {10}

14 % 14 %

4

13 % 13 % aIn square brackets : Unpenalised, and, in {brackets}, penalised weights for evaluating the success rate of clinical evaluation of anaemia.

14

Table 4.3 Accuracy of FAMACHA© clinical scoring by experienced and inexperienced evaluators in relation to the predetermined

haematocrit categories Method 1a Method 2a

Data set No. Animals Persons Unpenalised :

Mean (range) Penalised :

Mean (range) Unpenalised : Mean (range)

Penalised : Mean (range)

Experienced evaluators

1 184 1 1.8 (n/a) 2.3 (n/a) 8.6 (n/a) 7.4 (n/a)

2 181 1 1.4 (n/a) 1.8 (n/a) 9.0 (n/a) 8.2 (n/a)

3 172 1 1.2 (n/a) 1.4 (n/a) 9.2 (n/a) 8.4 (n/a)

4 336 1 1.8 (n/a) 2.2 (n/a) 8.7 (n/a) 8.2 (n/a)

Trainees (inexperienced evaluators)

5 20 18 3.9 (2.1-6.3) 4.0 (2.2-6.3) 8.1 (9.1-6.7) 8.0 (9.0-6.6)

6 20 36 2.2 (1.3-4.3) 2.5 (1.4-5.0) 8.6 (9.4-6.0) 7.8 (9.3-4.9)

7 19 14 1.9 (1.4-2.6) 2.6 (1.7-4.3) 8.3 (9.1-6.9) 6.9 (7.9-5.0)

8 31 23 4.8 (3.6-6.0) 4.8 (3.6-6.0) 7.7 (8.5-6.3) 7.6 (8.5-6.2)

9 20 32 2.4 (0.8-4.8) 2.8 (0.9-4.9) 8.5 (9.6-7.0) 7.9 (9.4-6.3) an/a : not applicable.

15

Table 4.4 Clinical evaluation accuracy of FAMACHA© scoring versus haematocrit categories :

Comparing initial categories with others obtained from a graph fitted to the results of an experienced evaluator

Mean (range) clinical evaluation success

Unpenalisedb Penalisedb

Data set No. Animals Persons Method 1 :

Initial scores

Method 1 : Scores

from graph

Difference (Initial minus graph)

Method 1 : Initial scores

Method 1 : Scores

from graph

Difference (Initial minus graph)

Experienced evaluator

1 184 1 1.8 (n/a)

1.3 0.5 2.3 (n/a)

1.8 0.5

2 181 1 1.4 (n/a)

1.2 0.2 1.8 (n/a)

1.6 0.2

3 172 1 1.2 (n/a)

0.9 0.3 1.4 (n/a)

1.3 0.1

4 336 1 1.8 (n/a)

- - 2.2 (n/a)

- -

Meana - - 1.47 1.13 0.34 1.83 1.57 0.26

Trainees (inexperienced evaluators)

5 20 18 3.9 (2.1-6.3)

3.6 (1.4-5.1)

0.3 4.0 (2.2-6.3)

4.3 (1.8-6.0)

-0.3

6 20 36 2.2 (1.3-4.3)

1.8 (1.4-3.3)

0.4 2.5 (1.4-5.0)

2.4 (1.6-4.2)

0.1

7 19 14 1.9 (1.4-2.6)

1.3 (0.7-2.2)

0.6 2.6 (1.7-4.3)

2.0 (1.0-3.6)

0.6

8 31 23 4.8 (3.6-6.0)

4.0 (3.1-5.0)

0.8 4.8 (3.6-6.0)

4.8 (3.8-5.6)

0.0

9 20 32 2.4 (0.8-4.8)

1.8 (0.4-3.6)

0.6 2.8 (0.9-4.9)

2.3 (0.5-4.2)

0.5

Mean 22 - 3.04 2.50 0.54 3.34 3.16 0.18 aExcluding Data Set 4. bn/a : not applicable.

16

Table 4.5 Analysis of haematocrit values of the sheep in the various tests of the FAMACHA© system

Haematocrit analysis FAMACHA© : Mean success

scorea Percentages of sheep in the various FAMACHA© categories

Data set No.

Animals (n)

Persons (n)

Method 1

Method 2

Mean (Range)

1 2 3 4 5

Experienced evaluators

1 184 1 1.8 8.6 22.1 % (11-35 %)

13.6 34.7 29.3 22.3 1.1

2 181 1 1.4 9.0 22.4 % (12-36 %)

16.1 30.6 36.7 16.1 0.5

3 172 1 1.2 9.2 22.4 % (12-36 %)

16.1 30.6 36.7 16.1 0.5

4 336 1 1.8 8.7 26.2 % (9-40 %)

42.7 33.5 18.7 4.5 0.6

Trainees (inexperienced) evaluators

5 20 18 3.9 8.1 27.1 % (16-40 %)

45.0 25.0 20.0 10.0 0.0

6 20 36 2.2 8.6 21.5 % (10-34 %)

20.0 25.0 25.0 25.0 5.0

7 19 14 1.9 8.3 21.6 % (10-34 %)

21.1 26.3 21.1 26.3 5.3

8 31 23 4.8 7.7 29.8 % (20-40 %)

67.7 22.6 9.7 0.0 0.0

9 20 32 2.4 8.5 24.8 % (14-34 %)

40.0 25.0 15.0 20.0 0.0

10 19 28 Nd Nd 25.7 % (9-36 %)

36.8 36.8 21.1 0.0 5.3

11 19 58 (51)b Nd Nd 22.5 % (15-35 %)

15.8 26.3 42.1 15.8 0.0

aUnpenalised only (see text and Table 4.3 for the corresponding penalised scores). b51 persons, but 58 sets of results (see text).

17

Table 4.6 Regression analysis of FAMACHA© (F©) vs faecal egg count (FEC) and the haematocrit

(Ht) and of the latter versus the FEC Regression analysis (r2 value x 100)a

Trial/Date F© vs Ht Ht vs FEC F© vs FEC

Intensive trial (n=170)

Data Set 1b 43.2 67.9 38.9

Data Set 2b 44.3 45.0 25.5

Data Set 3 58.5 - 34.1

Mean 48.7 % 56.5 % 32.8 %

Early trial (n=370)

14/3 8.4 - -

22/4 38.9 - -

02/5 28.6 - -

14/5 20.8 - -

11/6 25.6 - -

24/7 25.7 - -

Mean 24.7 (28.0) %c - - aThe r2 value x 100 gives an indication of the extent to which the variability of one of a given pair of observations is explained by the value of the other; with a perfect match the r2 x 100 = 100 %. All of the values are highly significant (P<0.0001). bData Sets 1 and 2 were from the same group of sheep, evaluated on the same day by 2 different persons with experience. cIn brackets : Mean, excluding the results of 14/3. Table 4.7 Accuracy of basing decisions on drenching on FAMACHA©

estimates, versus on haematocrit determinations Data set : Accuracy (%)a

Category Data set 1 Data set 2 Data set 3

No drench required

Correct (None given) 89.8 85.3 86.5

Wrong (Drenched) 2.8 9.8 8.4

Drench required

Correct (Drenched) 5.1 3.3 3.9

Wrong (Animals at risk) 2.3 1.6 1.1

Total 100 % 100 % 100 % aSee Tables 4.3-4.6 for more particulars on Data Sets 1-3.

18

Discussion The primary aim of these investigations into the FAMACHA© system was to find a method by which those individual sheep in a flock that are unable to withstand Haemonchus spp. challenge could be identified clinically, without the need for laboratory tests. Another objective was to obtain an indication of the extent to which clinical evaluation of anaemia according to the colour of the conjunctivae can approximate evaluation by the haematocrit, used here as the �gold standard� for indicating the haematological status of a sheep. As can be deduced from the example of the scoring in Table 4.2, the great majority of the evaluations by the 3 experienced persons (Tables 4.3, 4.4 and 4.5) were either correct (being in the correct haematocrit category), or in an adjacent category, while the mean results of the inexperienced persons were more inaccurate. Nevertheless, it will be noted in Table 4.3 that some of the inexperienced trainees obtained better results than the �old hands� (e.g. Test 9, where one person had mean scores of 0.8 and 9.6 by the 2 methods of scoring, compared to the best corresponding scores of 1.2 and 9.2 of the experienced persons). Among the trainees the large differences between the persons who fared best and worst (e.g. means of 0.8-4.8 in Test 9 - unpenalised, Method 1 - in Table 4.3) may indicate corresponding differences in the basic ability of persons to apply the FAMACHA© system. It would be interesting to be able to follow up groups of these persons as they gain experience, to determine the rate of improvement (if any), and to compare the abilities of those who fared best initially, with the others who were not as successful. There were only small differences between the unpenalised and penalised weighted values, indicating that there were few cases where the sheep that were incorrectly classified may have been endangered by the mistakes. In addition, while the large differences in the numbers of animals involved may have invalidated statistical evaluation of the differences between the experienced and the inexperienced evaluators, it is striking that there was a large difference in the mean rate of success of the evaluations, in the favour of the experienced persons.

Haematocrit categories selected for use in the FAMACHA© system As described above, the haematocrit categories were selected by allocating a width of 5 percentage points to each of the 3 intermediate haematocrit categories, 2, 3 and 4 (previously B, C and D - Van Wyk et al., 1997b). Thereafter the colours on the FAMACHA© chart used in all of these investigations were selected by examining and photographing the conjunctivae of a large number of sheep, of which the haematocrit values were known. Relatively narrow haematocrit ranges were chosen per category to enable division of particularly the lower haematocrit ranges (that is, for the most anaemic sheep) into a number of FAMACHA© categories, in an attempt to discriminate effectively between sheep still able to manage without treatment, and those at risk of being overwhelmed. However, when the chart was produced, we did not know to what extent the colours chosen would, on clinical comparison with the conjunctivae, correspond with the matching haematocrit categories. It is probable that a perfect match between clinical FAMACHA© classification and haematocrit categories would be theoretically possible but only if the relationship (unknown when the chart was produced) between the colour of the conjunctivae of a sheep and its haematocrit is linear in the critical intermediate haematocrit ranges. However, there was no �gold standard� for evaluating the �correct� selection of the different colours for the various haematocrit categories. Eventually the method described above was used to investigate the relationship : A graph was plotted for the results of Data Set 2 in Table 4.3, whereafter the ranges of haematocrit values for the various FAMACHA© categories were obtained from this line graph. Then the results of all the persons and groups listed in Table 4.3 were re-evaluated by Method 1, using the modified haematocrit ranges obtained from the graph.

19

In Table 4.1 the haematocrit ranges read from the graph are compared with the original chosen for the chart, and the comparative results are listed in Table 4.4. It will be noted that the relationship between the FAMACHA© categories and the various original haematocrit categories was not linear, but curved. When, as appears necessary from the graph, FAMACHA© Category 2 is widened in relation to the higher (thus more �anaemic�) categories, the mean success rate of the majority of the various persons tested is considerably higher than previously (Table 4.4), except for those tests where there were no, or very few highly anaemic animals. This was most prominent in the case of the penalised scores. One is not able straightforwardly to interpret the differences in the success rates between the clinical classifications according to the initial haematocrit ranges used in these investigations and those after the haematocrit ranges had been adapted according to the graph (Table 4.4). While the superior results with the latter ranges may indicate an inexact match of the colours chosen to represent the various categories, it should be kept in mind that, as shown in Table 4.5, few of the sheep were patently anaemic, most of their haematocrit values being in the lower (less anaemic) FAMACHA© categories (Categories 1-3, Table 4.1). Therefore, the striking widening of the haematocrit range of FAMACHA© Category 2 may have made it easier to obtain better evaluation scores than with the initial set of haematocrit ranges, because there were so many sheep in which haematocrit values caused them to fall into that category. Since inception of the FAMACHA© concept 3 years ago, we progressively came to the conclusion that the colours needed to be adapted slightly, as they did not match the observed colours of the different categories closely enough. However, this decision and the subsequent adaptations were done before the above evaluation using the haematocrit ranges from the graph. Hence this seems to indicate that the superior test results with the values from the graph may result from both theories propounded above (that the improvement could be due to a larger proportion of animals in the lower FAMACHA© categories, and that the colours are not completely representative). What should now be done in an attempt to resolve the issue, is to test the success rates of groups of both experienced and inexperienced persons in classifying relatively large flocks of sheep with equal numbers of animals in each FAMACHA© category. To a certain extent the suggestion of using equal numbers of sheep per category is supported by the data shown in Table 4.5 (discussed below). On the other hand, as shown below (Section 6), the heritability of the FAMACHA© values appears to be, from the single set of suitable data obtained to date, practically on a par with that of the haematocrit values of Merino sheep tested, indicating that the initial haematocrit ranges did prove to be very useful. In Table 4.5 the distribution of the haematocrit values of the different sheep to the various FAMACHA© categories is examined in relation to the success rates of the persons tested, while keeping in mind that relatively few test groups have been evaluated. When the mean success rates of the persons tested; of the mean haematocrit values; and of the total percentage of animals in FAMACHA© categories 1-3 (Table 4.5) are ranked separately for the experienced persons and trainees, the ranks correspond, with the single exception of Tests 6 and 7 (Table 4.5) that differed little from one another. However, the distributions in haematocrit values do not appear to have had such a big effect on the success rates of the experienced persons, compared with those of the trainees. It seems possible that in the trainees differences may have resulted because those who had to classify groups of sheep with a narrow range of colour variation among the different sheep seemed to �magnify� the relatively small differences observed between sheep, which would have caused them to record a wider range of variation than existed in reality. On the other hand, note, from the data given in Table 4.3, that in all but one of the tests some of the trainees were at least as successful with the clinical classification of the sheep as were those persons with experience, irrespective of the distribution of sheep according to their haematocrit categories.

20

While the persons who did the FAMACHA© evaluations in the early (Badplaas) trial varied from time to time and their individual evaluations were not recorded separately, the data from the intensive trials (Table 4.6) were from only 2 persons, of whom one did the evaluations of both Data Set 2 and Data Set 3 and the other only Data Set 1 (Table 4.6). This complicates comparison of the different sets of data from the early trial with the later ones. It should be noted that throughout this report the evaluation of the Data Sets 1, 2 and 3 were as described above, with one person being responsible for Sets 2 and 3, and the other only for Set 1 (Tables 4.3-4.5). In the regression analysis the level of significance of correlations alone is not sufficient for evaluating the usefulness of the FAMACHA© system for detecting anaemic sheep. The symbol r is used to represent the extent to which any 2 sets of data correlate. When 2 identical sets of data are compared, r=1. The larger the numbers of observations in the pairs of data that are compared, the lower is the r value that is still statistically significant at a given level of confidence. This is best illustrated by the fact that with the relatively large numbers of animals in some of the later trials, the value of 8.4 (early trial, F© vs Ht, 14/3 in Table 4.6) is also highly significant. For example, with the total of 370 sheep used in the early (Badplaas) trial, the mean extent to which the haematocrit of an individual sheep (in the data set of Day 0 in Table 3.1) will declare variation in the FAMACHA© of the same animal is only 8.4 %. While this correlation is still highly significant, it is of little use in the field for estimating the haematocrit from the clinical evaluation of the colour of the conjunctivae. The fact that all of the values in Table 4.6 are highly significant (for all the columns), is encouraging, but unfortunately this signifies only that one can be certain that there is indeed a correlation between the variables that are compared, and not that one of them can predict the other with a high degree of accuracy for a given sheep. In Table 4.6 the accuracy of the clinical evaluation according to the FAMACHA© system is indicated by the comparison between the FAMACHA© value (the dependent variable) and the haematocrit (the �gold standard�, independent variable). From this it would appear that the person who was responsible for the evaluations of Data Set 1 fared better than the other person responsible for Set 2, but the difference was not significant (P>0.05). Furthermore, there was a marked contrast between the mean success rate in the current, intensive trial and the earlier Badplaas trial, which was conducted when the FAMACHA© chart had not yet been produced, nor had the FAMACHA© categories been established in relation to the haematocrit values of sheep. While keeping in mind that the differences in trial design limit the deductions that can be made concerning the significance of the differences in results, the introduction of the FAMACHA© chart and accessory system of classification does seem to have brought a considerable improvement in the success rate for evaluating the haematological status of the animals by the FAMACHA© system. Another method of evaluation of the data was used for judging the applicability of the FAMACHA© system under field conditions. This was by estimating the extent to which decisions (by 2 experienced persons) based on the FAMACHA© evaluation for drenching, would be �right� or �wrong� (Table 4.7). This indicates that, according to the analyses of 3 different data sets, only very small percentages of the animals under heavy Haemonchus challenge (the figures in bold italics in the table) were at risk of succumbing to haemonchosis. This small percentage of sheep that may have been at risk is even more insignificant if it is considered that sheep seldom die from haemonchosis without first developing bottle-jaw. Furthermore, when sheep are driven, the animals that are being overwhelmed by the worm infection tend to lag behind, and can also be identified in this way. In other words, by applying the FAMACHA© system and also closely scrutinising sheep under stress from haemonchosis, it is unlikely that deaths will occur. The conclusion we draw from the analyses of the results of the trials conducted to date is that it is essential to have at least some sheep in each of the FAMACHA© categories 1-4, preferably also including Category 5 for trainees to have the best chance of being able to apply the system

21

effectively after the training session. In addition, at least one strikingly anaemic and another very healthy animal should be used in the practical demonstration, before the candidates are tested. However, since farmers conventionally drench sheep flocks as a whole when any of them are unable to withstand the worm challenge, it is usually difficult to obtain severely anaemic sheep in FAMACHA© category 5, and even 4, for training in the field. Hence it will often be necessary beforehand to bleed sheep to create anaemia artificially, or else the persons may not be properly trained when they start to apply the FAMACHA© system.

Education and literacy While this has also not been analysed statistically, the level of education does not appear to affect the ability of a person to learn to apply the FAMACHA© system. Poorly literate or illiterate people apparently learn as well and as quickly as those with a tertiary education. Various persons in Tests 7, 8 and 11 were poorly literate, having had minimal schooling, and yet obtained good scores in the tests. Notwithstanding this observation, the only person who entirely misunderstood how the system is applied (a person in Test 9, whose results were excluded from those recorded in the various tables) is poorly literate; he classified 9 consecutive animals into corresponding consecutive FAMACHA© categories, �1-9�.

22

Section 5 Validation of the FAMACHA© system on commercial

sheep farms

Aim The aim of this part of the validation procedure was to determine if the FAMACHA© system would be implementable with minimal supervision by veterinarians and the research team on well-managed, commercial farms in several summer rainfall regions without changing management systems then employed.

Materials and methods The trials varied somewhat in format according to the level of intensity at which the research team was able to visit the farms and be involved in sampling and evaluation, as well as to the levels of commitment and animal record-keeping of the farmers. The basic approach was for farmers and their workers periodically to examine all the trial animals, using the FAMACHA© technique, and to pay closer attention to those sheep that had been judged to be markedly anaemic. In most instances the inspections were commenced at the start of the Haemonchus season (± September/October), and were repeated at intervals of 3 weeks until the worm challenge intensified, whereafter the intervals between inspections were reduced. Eventually, towards the end of the worm season, the sheep were examined weekly on most of the farms. Each farmer was advised at the beginning of the worm season to drench only sheep judged to be in FAMACHA© categories 4 and 5, but subsequently (as larger proportions of the flock became anaemic), also to drench those in category 3, or even, if deemed essential, the whole flock. The more intensive the trial, the more expansive the trial plans were. For instance, in the most intensive trials individual animals were drenched during the Haemonchus season (from September to May) only after faecal and blood sampling had indicated that their lives (of such animals) would be endangered if they were not treated. Nevertheless, as a result of the relatively long distances between the farms and the fact that all but one of the trials were conducted under practical farming conditions on private property, in almost all of the cases we were unable entirely to realise our ideals and had to settle for less. On the farms on which we worked most intensively one or more times, both the haematocrit percentage and the FAMACHA© category estimate were recorded for each sheep in each trial flock. At the other extreme there were a few farms to which the trial team paid few visits, with the result that the farmers had to make their own decisions as to whether or not to apply individual and/or flock drenching. They based these on periodic discussions with the research team members, on their own clinical observations with the aid of the FAMACHA© chart and on other clinical signs of worm infection, such as the presence of bottle-jaw or diarrhoea.

Parameters for identifying suitable farms to test the FAMACHA© system While not all of the parameters mentioned below were met on each farm, they served as guidelines for what was expected from the farmers. The parameters (the most important of which are indicated with an asterisk) were :

23

* Liver fluke not a problem; * Good level of management; * Accurate record keeping; * Proven wireworm problem; * Determined to carry the investigation through to its conclusion; * Sheep numbers preferably >200; Good handling facilities and sufficient staff; Willingness to co-operate/follow instructions; Motivation to succeed; Proven track record; Knowledge of disease control; Reasonably near veterinary assistance; Good relationship with their veterinarian; Willing to consult with veterinarian on problems; Willing to bring in dead sheep for examination; Willing to stress animals; Willing to allow animals to be frequently handled; Ability to use Faecal Egg Count/Faecal Egg Count Reduction Test (FEC/FECRT); Overall suitability. A number of farmers were preliminarily selected (short-listed) to participate in the trials, after which each was assessed separately by 3 evaluators for his/her ability to apply the FAMACHA© system and was then given a combined rating. Only the candidates with the highest scores were selected from Gauteng, Mpumalanga, Free State and KwaZulu-Natal Provinces.

Farm sites selected Two farms were selected from Gauteng, 5 from Mpumalanga, 5 from Free State and 1 from KwaZulu-Natal. Four of these were intensively grazed, while, on the others, more extensive sheep farming systems were practised.

Requirements from participating farmers Each participating farmer was expected to : 1. Supervise staff who were using the FAMACHA© system. 2. Allow the staff to practise the use of FAMACHA© in trial runs before clinical haemonchosis

was expected to occur. 3. Select a trial group of preferably at least 100 sheep of uniform type, or failing this, 2 or more

distinct types (examples: unweaned lambs, weaned lambs, dry ewes, pregnant ewes, lactating ewes, wethers or rams, weaned replacement ewes being preferred by us).

4. Examine the trial group every 2-3 weeks until the first cases of clinical anaemia were

encountered. 5. Examine sheep (particularly lambs) weekly during the peak worm season. 6. To score, in flocks in which the sheep were not individually identified, all the animals at

every examination into categories 1 to 5 according to the FAMACHA© card, to record (on form FB3 � Appendix 4*) the numbers of animals classified into each of the categories 1, 2, 3, 4 or 5, and to take the following action : 6.1 Sheep classified as 1 or 2 to be released without further action. 6.2 Dose sheep classified as 4 or 5 with an appropriate anthelmintic (levamisole preferred,

but not one with residual action), and to mark them permanently (as outlined in FB5, Appendix 5).

6.3 Dose also all sheep judged to be in Category 3, but perhaps a doubtful Category 4. 6.4 Sheep classified as 3, which were dosed, were also to be marked (see 7 below). 6.5 If there were too many sheep in a group, or if the farmer did not want to examine all

sheep weekly, he could examine 20 % of the flock. If more than 10 % were very anaemic (categories 4 or 5) then the whole flock was to be examined.

6.6 Record if dung was taken for FECs, as well as the results thereof. 6.7 Record every instance where treatment was given. 6.8 Record if camps were changed.

24

7. To resist from removing sheep from the trial group until the end of the trial. If any were removed, then numbers, dates and reasons were to be supplied.

25

8. To take all trial group sheep which died to the supervising veterinarian for a post mortem examination. When this was not possible, the farmer was to examine the dead sheep and record his/her findings and send these to the veterinarian.

The information required in the record or action to be taken when the necropsy was done by the farmer was (if possible) : 8.1 Whether the sheep was anaemic or not (pale mucous membranes, watery blood, pale

carcass). 8.2 The presence or absence of bottle-jaw. 8.3 The presence and severity of wireworm infection in the abomasum. 8.4 Whether a faecal sample was taken for FEC. 8.5 If samples of worms were taken. Large canned-fruit bottles or similar containers and

10 % formalin were left with the farmer, with the request that the entire content of the abomasum be preserved in it, or failing this to tie the abomasum at both ends and then to place it in a plastic bag containing 10 % formalin.

8.6 When the sheep was last dosed and with which anthelmintic. 8.7 The date of death of the sheep. 8.8 Any other abnormalities observed.

9. To dose the whole flock with an effective anthelmintic (preferably levamisole, but not a very

long-acting remedy) if more than 10 % of the total flock was judged �anaemic� (categories 4, 5 and doubtful cases of 3). Thereafter the grazing camp could be changed after 2-3 weeks if necessary, in consultation with the veterinarian and researchers.

10. To make groups of sheep in the trial flocks available for the taking of samples by the

research team with prior arrangement. 11. To vaccinate all sheep in the trial flocks at the commencement of the trial at least against

enterotoxaemia. 12. To treat nasal worm conservatively (but only if absolutely necessary) with rafoxanide (trade

name : �Nasalcur�, Hoechst Roussel Vet - now Intervet SA). Then the entire flock was to be treated, not merely those showing clinical signs.

13. To liaise closely with the supervising veterinarian. 14. To not make changes to the agreed procedure without first consulting the supervising

veterinarian. 15. To submit trial records to the supervising veterinarian during, and at the end of the trial. 16. To compile assessment forms at the end of the trial. 17. To sign undertaking and indemnity agreements (below).

26

The following agreement was signed by each farmer before the commencement of the trial : Undertaking By Farmer I, ����������������������������������������� hereby undertake to conduct a trial on the FAMACHA© anaemia guide and system according to the attached schedule and not to deviate from it before I have consulted with the supervising veterinarian. I undertake to inform him/her as soon as possible of any problems which may be emerging connected with the use of the system so that, if necessary, an investigation can be carried out and recommendations made. Indemnity Form I, ����������������������������������������� hereby declare that I am taking part in the trial of the experimental FAMACHA© concept of my own free will and in the full knowledge that results cannot be predicted with any certainty. I further declare that I am competent to make this indemnity and have carefully studied and understood the information supplied on the implementation of the FAMACHA© anaemia guide and undertake to abide by its instructions and provisions. I therefore without qualification indemnify all persons and organisations involved in the trial, including all veterinarians, technicians, the Worm Workshop, the Livestock Health and Production Group of the South African Veterinary Association, Hoechst Roussel Vet Animal Health and any other advisors, functionaries and organisations whatsoever, from any responsibility for losses from any cause arising directly or indirectly from the use of the FAMACHA© card and techniques. Signed ������������������������������������������������� Date, place, witnesses �����������������������������������������

Requirements of local veterinarians Where possible, we identified and requested local veterinarians to monitor the situation on the selected farms and to investigate and deal with any crises that were reported to them. This meant that we had local experts to report to us and to solve some of the problems that arose. Undertaking by Supervising Veterinarian I, ����������������������������������������� hereby undertake to supervise the FAMACHA© trial/s as set out in the attached schedules (FV1 and FV2) free of charge (except where a written agreement determines otherwise) and to supply the required information to the research team. I further undertake to inform them in good time of any emerging problems which may arise in the course of the trials.

Recording of results All data in this section could be collected on a single, flock histogram form (FB3, Appendix 4) and for the classification of the sheep into the 5 FAMACHA© categories, use of the chart was designed to be easy even for illiterate persons. The farmer was to fill in the remaining items.

Marking of sheep Treated sheep were to be marked by one of the methods illustrated in FB5 in Appendix 5.

Support visits Support visits were planned for at least every 6 weeks for every farm during the Haemonchus season to monitor progress, but most farmers were visited every 3 weeks and a few less frequently. Some had unscheduled visits in addition. At each visit progress was discussed and

27

problems (if any) identified and solved. Data were also collected and subsequently collated for input into a database. FECRTs were done on a bulk (flock) basis according to the method described in �General Methods�. Faecal egg count reduction tests were performed on a representative sample of sheep on some of the farms.

Training of participants The farmers and their staff were all trained as described in Appendix 2.

Perception questionnaire interviews at the conclusion of the trials Two sets of questionnaires (one for farmers and the other for stockmen and veterinarians) were compiled for gauging their experience and perceptions and evaluation of the FAMACHA© system by participating farmers and staff, as well as veterinarians who had been involved in the trial, (Appendices 6 and 7). Each participating farmer and the principal worker involved in the trials were interviewed independently at the conclusion of the trial. Care was taken to put them at their ease and not to direct their responses. The questions for farmers (Appendix 6) were designed to elicit their perceptions on their understanding of the system, its practicality, its effects on animal production, its financial impact, its effect on their awareness of helminth control, and its overall usefulness. The questions were asked in random order, in the form of both positive and negative statements concerning the FAMACHA© system, to obtain responses to each of the 6 aspects and the clusters of questions were pooled for recording. The 34 questions in Appendix 6 were designed to cross-check the 6 main categories of evaluation, as shown in Table 5.4 and Appendix 7. The duration of the trial was divided into 2 periods. In each of these periods the costs of worm remedies was estimated by the farmers, the first being before the implementation of the FAMACHA© system and the second after its implementation. Details hereof are given in Tables 5.2 and 5.3.

Results Farms selected and data used Thirteen trial farms were selected, 1 in KwaZulu-Natal (Farmer 9 in Table 5.3), 2 in Gauteng (1 and 2), 5 in Mpumalanga (5 and 10 and 3 unlisted farmers) and 5 in Free State (3, 4, 6, 7 and 8). Some farmers (1, 2 and 10) had recorded data which were patchy and not easily interpreted or analysable and on these grounds were thus also largely excluded. Bar graphs are presented of a representative selection of the flocks examined, where the data form a meaningful series or illustrate important findings. Data are also given of estimated or calculated financial effects. One farmer (8) recorded the final weights of each sheep, as well as how many times each sheep had been treated during the season. This data is presented in Table 5.1.

28

Table 5.1 Numbers and average weights of sheep (ewe replacements) treated 0, 1, 2 or 3 times in

a summer (October 1998 to February 1999, inclusive) on one of the farms (Farmer 8)

Number of treatments Number of sheep Percentage of flock Average weight (kg)

0 81 32.4 33.75

1 139 55.6 32.60

2 28 11.2 32.70

3 2 0.8 29.75

Table 5.2 Effect of the FAMACHA© system on the costs of treating

sheep for worms on one farm (5, Table 5.3) Period Remedy Costs (R) Cost/sheep

Sept. 1996 to April 1997a, inclusive Cydectin 11,308.99 R 7.53

Sept. 1998 to April 1999b, inclusive Cydectin 5,988.46 R 3.99 aBefore and bafter implementation of FAMACHA© system. Table 5.3 Estimated or calculated reductions in costsa of worm

remedies used on some of the farms Reductions in costsb Farmer

Before FAMACHA© After FAMACHA©

% Reduction

1 3520 (doses) 200 96 %

2 5000 (rands) 1000 80 %

3 Estimate Estimate 66 %

4 Estimate Estimate 50 %

5 10669 (rands) 5988 44 %

6 Estimate Estimate 50 %

7 6021 (rands) 3588 40 %

8 Estimate Estimate 60 %

9 Estimate Estimate 60 %

10 16000 (rands) 10000 38 %

Average of 10 farms : 58.4 % aFigures supplied by the farmers involved. bComparing the average annual costs in the period during which the FAMACHA© system was tested (1998-1999) to the immediate preceding period : 1995-1998, 1996-1998, or (mostly) 1997-1998.

29

Table 5.4 Farmer/Worker/Veterinarian opinions and perceptions Response rating to questions (% of responses)a

Rating category Category cluster :

Questions asked of farmers

Bad 1

Poor 2

Fair 3

Good 4

Excellent 5

Positive (4+5)

Understanding : Pamphlets, training, explanations, illustrations, back-up help

Farmer 7.7 10.3 5.1 11.6 64.9 (77 %) Worker 4.7 4.7 0 23.8 66.6 (90 %) Veterinarian 0 0 0 66.6 33.3 (100 %)

Average 4.1 5.0 1.7 34.0 54.9 89 % Practicality: Time taken, difficulty, reliability, fitting in programme

Farmer 0 6.4 3.8 23.3 66.2 (89 %) Worker 4.5 4.5 0 13.6 77.2 (91 %) Veterinarian 0 0 0 33.3 66.6 (100 %)

Average 1.5 3.6 1.2 23.4 70.0 93 % Animal production: Stress, untoward effects, deaths and problems

Farmer 1.2 1.2 6.2 27.5 63.7 (91 %) Worker 0 13.6 4.5 54.5 27.2 (82 %) Veterinarian 0 0 0 66.6 33.3 (100 %)

Average 0.4 4.9 3.5 49.5 41.4 91 % Financial aspects : Expenses, savings, cost/benefit analysis, less treatment

Farmer 1.9 0 0 9.8 88.2 (98 %) Worker 0 9.0 0 45.4 45.4 (91 %) Veterinarian 0 0 0 0 100 (100 %)

Average 0.6 3.0 0 18.4 77.8 96 % Awareness : Worms, resistance, control measures, worker awareness, veterinary involvement

Farmer 2.5 6.4 7.7 12.9 70.1 (83 %) Worker 0 9.0 0 18.1 68.1 (86 %) Veterinarian 0 0 0 0 100 (100 %)

Average 0.8 5.1 2.5 10.3 79.4 90 % Usefulness : Will use again, will recommend, want similar systems

Farmer 1.2 1.2 3.7 15.1 78.4 (94 %) Worker 4.7 0 0 19.0 76.1 (95 %) Veterinarian 0 0 0 0 100 (100 %)

Average 1.9 0.4 1.2 11.3 84.9 96 % Overall averages :

Farmer 2.4 4.2 4.4 16.7 71.9 (89 %) Worker 2.3 6.8 0.7 29.0 60.1 (89 %) Veterinarian 0 0 0 27.7 72.2 (100 %)

Average 1.5 3.6 1.7 24.4 68.0 93 % aResponses from the 13 farmer participants, 22 farm workers and 3 supervising veterinarians (all percentages rounded off).

30

Problems identified during the trials ! Wrong diagnosis

Many farmers were inclined to attribute any loss or problem encountered during the trial to the FAMACHA© system rather than to anything else. Some examples of problems encountered are : # Fasciolosis � as it also causes anaemia and bottle-jaw, it is easy to confuse it clinically

with haemonchosis. This problem, encountered on 2 farms, should be investigated further, as the FAMACHA© should be equally effective in identifying animals unable to cope with liver fluke infection; the important difference being to ensure that, on those farms where liver fluke is a problem, an anthelmintic is used that is effective against both this parasite and Haemonchus spp.

# Trichostrongylosis � poor production and growth on 1 farm was associated with this

problem together with haemonchosis. # Bluetongue – what appeared to be sudden deaths on one farm (Farm 10, Table 5.3)

was attributed by the farmer to acute haemonchosis, but investigation revealed that bluetongue was undoubtedly the cause of the deaths.

# Pasteurellosis � encountered as the cause of deaths on Farm 4; haemonchosis was

excluded as playing a role. ! Young lambs and weaners � it is well known that young sheep are unable to mount as

effective a response to worm infections as are older sheep. This is borne out by our findings - both the graphs and farmers� opinions support this observation.

! Periparturient ewes � in common with the results in the preliminary trial at Badplaas,

ewes in late pregnancy or early lactation on several farms were clearly more susceptible than other classes to the effects of parasitism. They therefore need closer monitoring in the system.

! Low condition scores � where sheep were in poor condition, whether due to inadequate

nutrition or other causes, they were unable to respond adequately to challenge. ! Bottle-jaw � in several cases (especially Farm 4) we encountered sheep with obvious

submandibular oedema before anaemia had became severe. This emphasized that sheep must be examined for bottle-jaw as well as for anaemia.

! Wrong interpretation � some farmers became complacent and too confident of their

ability to score sheep without reference to the FAMACHA© card for calibration. This resulted in problems and even losses.

! Irregular examination � if management on a farm did not ensure that sheep were

examined at the prescribed intervals, there could be a sudden worsening of the situation with many anaemic sheep to be treated.

! Anthelmintic resistance � on one farm where the status of worm resistance to

anthelmintics was unknown at the commencement of the trial (Farm 2), an inappropriate remedy was used, leading to apparent failure of the FAMACHA© system.

31

Conclusions ! It is essential that the principal worm species, as well as others that could escalate in

numbers and importance owing to reduced drenching without their being detected by the FAMACHA© system, be known and monitored where the system is to be applied. Fasciola spp., Oesophagostomum columbianum and Trichostrongylus spp. are particularly important in this regard.

! The efficacy of the FAMACHA© clinical anaemia evaluation was adequate to detect animals unable to withstand haematophagous worm challenge.

! It is clear from the bar graphs (Figures 5.1-5.18 and Table 1) that most sheep in the ewe replacement category needed no, or else only one treatment under conditions of strong challenge by H. contortus. Many lambs and weanlings, as well as pregnant or lactating sheep were also able to manage without treatment.

! It is apparent from the financial analyses that considerable savings in anthelmintic treatment were obtained or estimated (Tables 5.2 and 5.3).

! A number of problems were identified, none of which are insurmountable. However, they do emphasize the necessity of implementing the system with caution by suitably trained persons and under proper supervision.

! In the final analysis, the feasibility and acceptability of such a system has to be judged subjectively by farmers and workers who have to implement it. Their perceptions are presented together (Table 5.4). As, in all groups, the overall positive rating approached 90 %, often with two-thirds or more rating the system excellent, we need have no hesitation in concluding that in good hands, and subject to the limitations listed in the instructions pamphlet, the FAMACHA© system works well and, when implemented correctly, will be a decided success.

! Our overall conclusion is that the FAMACHA© system has now been properly tested in field trials and can be released for use by farmers. The training and distribution system needs some organisation and planning, but can proceed in the summer of 2000/2001. On the other hand, vigilance for possible problems that may arise, as well as the use of other systems (such as condition scoring) to test the principle of drenching only animals that are unable to cope with infections of worm species other than H. contortus, must be emphasised.

32

Figure 5.1

Figure 5.2

������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������

������������������������������������������������������������������������

���������������������������������������������������������������������������������

������������������������������������������������������������������������

������������������������������������

���������������������������������������������������������������

����������������������������������������

���������������������������������������������

���������������������������������������������

������������������

��������������������������������

������������������������������������

���������������������������������������������

���������������������������������������������������������������������������������

������������������������������������������������������

��������������������������������������������������������

���������������������������������������������������������������

��������� ��������� �������� ��������� ��������� ��������� ��������� ����������������

������������������

���������

Plot of Fc-SCORES (%) for Season 1998-1999Farm - BESTER

Sheep category - EW EREP

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

27-11-98 05-12-98 15-12-98 26-01-99 04-02-99 16-02-99 23-02-99 02-03-99 21-05-99

DATE of Fc - EVALUATIONApproximate flock size: 290 - 210

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������

��������������������������Fc - 5��������������������������Fc - 4��������������������������Fc - 3��������������������������Fc - 2��������������������������Fc - 1

Dosed

��������������������

������������������������������������������������������������������������������������������

������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������

����������������������������������������������������������������������������������������������������

��������������������������������������������������������������������������������

��������������������������������������������������������������������������������

��������������������������������������������������������������������������������

��������������������������������������������������������������������������������������������������������������������������������������������

����������������������������������������������������������������������

������������������������������������������������������

���������������������������������������������������������������

������������������������������������������������������������

����������������������������������������������������������������������

����������������������������������������

����������������������������������������

��������������������

��������������������

������������������

������������������

�������������������� ����������

������������������������������������������������������������

��������������������������������������������������

��������� �������������������� ����������

��������������������

��������� ��������������������

����������

Plot of Fc-SCORES (%) for Season 1998-1999Farm - KÖHLER

Sheep category - EWEREP

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

15-10-98 25-11-98 15-12-98 06-01-99 28-01-99 17-02-99 03-03-99 11-03-99

DATE of Fc-EVALUATIONApproximate flock size: 325

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

�������������������������� Fc - 5

�������������Fc - 4

�������������������������� Fc - 3

�������������Fc - 2

�������������������������� Fc - 1

Dosed

33

Figure 5.3

Figure 5.4

������������������������������������

���������������������������������������������������������������

���������������������������������������������������������������

���������������������������������������������������������������������������������

������������������������������������������������������

������������������������������������

������������������������������������

������������������������������������

���������������������������������������������������������������������������������

������������������������������������������������������������������������������������������������������������

������������������������������������������������������������������������������������������

������������������������������������������������������������������������������������������

������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������

������������������������������������������������������������������������������������������

������������������������������������������������������������������������

������������������

��������� ��������� ��������� ��������� ���������������������������

������������������������������������

������������������������������������

������������������ ��������� ��������� ������������������

������������������

Plot of Fc-SCORES (%) for Season 1998-1999Farm - CILLIERS

Sheep category - EWEREP

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

26-11-98 14-12-98 06-01-99 09-02-99 17-02-99 24-02-99 05-03-99 10-03-99 01-04-99

DATE of Fc - EVALUATIONApproximate flock size: 260

PER

CEN

TAG

E of

SH

EEP

in F

c - S

CO

RE

GR

OU

PS

�������������Fc - 5

��������������������������Fc - 4

�������������Fc - 3

��������������������������Fc - 2

�������������Fc - 1

Dosed

��������������������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������

������������������������������������������������������������������������

������������������������������������������������������

������������������������������������������������������������������������

������������������������������������

���������������������������������������������������������������

��������������������������������������������������������

���������������������������������������������������������������������������������������������������

������������������������������������������������

������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������

������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������

���������������������������������������������������������������������������������

���������������������������������������������������������������������������������

��������������������������������������������������������������������������������

���������������������������������������������������������������

���������������������������

������������������

������������������������������������������������������

���������������������������

����������������

������������������

������������������

������������������ ���������

������������������

����������������

������������������

Plot of Fc-SCORES (%) for Season 1998-1999Farm - OBERHOLZER

Sheep category - EWEREP

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

01-12-98 11-12-98 28-12-98 07-01-99 18-01-99 02-02-99 15-02-99 01-03-99 23-03-99

DATE of Fc-EVALUATIONApproximate flock size: 140

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

�������������Fc - 5

��������������������������Fc - 4

�������������Fc - 3

�������������Fc - 2

��������������������������Fc - 1

Dosed

34

Figure 5.5

Figure 5.6

��������������������������������������������������������������������������������

�������������������������������������������������

�����������������������������������������������������������������������������

���������������������������������������������������������������

�����������������������������������������������������������������������������

��������������������������������������������������������

����������������������������������������������������������������������

������������������������������������������������������������������������

��������������������������������

�������������������������������������������������

���������������������

�����������������������������������

���������������������

�����������������������������������

����������������������������

��������������������������������

�������� ������� ������� ������� ������� ��������������

������� ����������������

������� ������� ���������������

Plot of Fc-SCORES (%) for Season 1998-1999Farm - NICOLAU

Sheep categories - EWEREP, EWEPRG, EWEDRY

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

EWEREP

25-11

-98

EWEPRG

02-11

-98

15-12

-98

06-01

-99

28-01

-99

17-02

-99

11-03

-99

EWEDRY

24-02

-99

DATE of Fc-EVALUATIONApproximate flock size: 350

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

��������������������������Fc - 5��������������������������Fc - 4��������������������������Fc - 3��������������������������Fc - 2��������������������������Fc - 1

Dosed

���������������������

��������������������������������

������������������

��������������������

���������������������

����������������

��������������������

����������������

���������������������

����������������������������

������������

���������������������

��������������������������������

��������������������������������������������������������

���������������������������������

������������������������������������

���������������������

���������������������

����������������������������

����������������������������������������

���������������

��������������������������������������������

������������������

���������������������

����������������������������

������������������������

������������������������������������

������������������������

����������������������������������������

������������������������������������

��������������������������������

���������������������

��������������������������������

������������������������������

���������������������

������������������������

���������

������������������������

������������������������

������������������������

��������������������������������

��������������������

������������������������������

����������������

���������������������������

������

���� ������

��������

��������

��������

����������������

������

���� ������

������

��������

������

��� ���

Plot of Fc-SC O R ES (% ) for Season 1998-1999Farm - D A SER R A

Sheep categories - EW EPR G , EW ELAC , LAM B S, W EAN ER

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

EWEP

RG

1/4/

99

1/11

/99

2/1/

99

2/22

/99

3/1/

99

3/9/

99

3/15

/99

EWEL

AC

10/1

3/98

12/7

/98

1/4/

99

5/3/

99

LAM

BS

10/1

3/98

10/2

6/98

11/9

/98

12/7

/98

1/11

/99

1/25

/99

WEA

NER

12/2

3/98

1/4/

99

1/25

/99

2/8/

99

2/22

/99

3/1/

99

DATE of Fc-E VALUATIO N

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

�������������Fc - 5

�������������Fc - 4

�������������Fc - 3

�������������������������� Fc - 2�������������������������� Fc - 1

Dosed

35

Figure 5.7

Figure 5.8

���������������������������������������������������������������

�����������������������������������������������������������������������������

����������������������������������������������������������������������

������������������������������������������������������������������������������������

���������������������������������������������������������������

��������������������������������������������������������

������������������������������������������������������

���������������������������������������������������������������

��������������������������������������������������������

�������������������������������������������������

��������������������������������������������������������

������������������������������������������

�������������������������������������������������

����������������������������������� �������

���������������������

��������������������������������������������������������

������������������������

���������������������

��������������������������������������������������������

���������������������������������������������������������������

������� ������� ������� �����������������������������������

�������������� ������

������������������

������������������������������������������

��������������

��������������

������� ������������

�������������

Plot of Fc-SCORES (%) for Season 1998-1999Farm - KÖHLER

Sheep categories - LAMBS, WEANER

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

LAM

BS

02-1

1-98

15-1

2-98

WEA

NER

25-1

1-98

06-0

1-99

28-0

1-99

WEA

NER

03-0

3-99

11-0

3-99

25-0

3-99

31-0

3-99

DATE of Fc-EVALUATIONApproximate flock size: 250

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

��������������������������Fc - 5��������������������������Fc - 4

�������������Fc - 3

��������������������������Fc - 2��������������������������Fc - 1

Dosed

������������������������

��������������������

��������������������������������

������������������������

��������������������������������������������������������

������������������������������������

����������������������������

��������������������������������

��������������������

������������������������

������������

����������������

��������������������������������

������������������������������������������������

����������������������������

��������������������������������

����������������������������������������

��������������������

����������������������������

������������������������������������������������������������

��������������������������������������������

����������������������������������������

������������������������������������

������������������������������������

������������

��������������������������������

����������������������������������������

������������������������������������

������������������������������������������������

��������������������������������������������

��������������������������������������������������������

����������������������������������������������������

������������������������������������

����������������

������������������������������������

������������������������������������

����������������������������

������������������������������������������������

����������������������������������������

��������

���� ������������

������������

���� ���� ��������

��������

���� ��������

Plot of Fc-SCO RES (% ) for Season 1998-1999Farm - DA SERRA

Sheep categories - EW EDRY, EW EREP, O THER

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

EWED

RY

10/1

3/98

12/7

/98

1/25

/99

5/3/

99

EWER

EP

12/2

3/98

1/11

/99

1/11

/99

1/25

/99

2/22

/99

3/1/

99

3/9/

99

3/15

/99

OTH

ER

11/9

/98

11/2

3/98

1/18

/99

1/25

/99

2/1/

99

2/8/

99

2/15

/99

11/3

0/99

DATE of Fc-EVALUATION

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

�������������Fc - 5

������������� Fc - 4

�������������Fc - 3

�������������������������� Fc - 2

�������������Fc - 1

Dosed

36

Figure 5.9

Figure 5.10

������������������������������������

����������������������������������������

���������������������������������������������

���������������������������������������������

��������������������������������������������������

��������������������������������������������

��������������������

�������������������������

��������������������

����������������������������

��������������������������������

������������������������������

���������������������������������������������

������������������������������������������������

��������������������

�������������������������

������������������������������������

��������������������������������������������������

����������������������������������������

�����������������������������������

����������������������������������������

����������������������������

����������������������������������������

�������������������������������������������������������

��������������������������������������������������

����������������������������������������

��������������������������������

���������������������������������������������

���������������������������������������������

������������������������ ����

����������������������������

�������������������������������������������������������

���� ����� ����������

���������������

����� ���� ����������������

���������������

��������������������

��������

������������

��������������������

����� ������������������������

���������������

����� ����� ���� ����� ����������

����� �����

Plot of Fc-SCORES (%) for Season 1998-1999Farm - KLEU

Sheep categories - EWEPRG, LAMBS, WEANER

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

EWEPRG36

136

3617

136

213

3622

036

228

3623

436

269

3628

3

LAMBS

3619

236

206

3621

3

WEANER36

220

3622

836

234

19-04

-9936

283

DATE of Fc-EVALUATION

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

�������������Fc - 5

������������� Fc - 4

�������������������������� Fc - 3

�������������Fc - 2

�������������������������� Fc - 1

Dosed

��������������������������������������������������������������������������������

����������������������������������������������������������������������

�������������������������������������������������������

������������������������������������������������������������������������

������������������������������������������

������������������������������������������������������

�����������������������������������

��������������������

����������������������������������������

�����������������������������������

��������������������������������������������������

���������������

�������������������������

�����������������������������������

������������������������������

������������������������������������������������������������������

������������������������������������������������

���������������������������������������������

�������������������������������������������������������

��������������������������������������������������

�������������������������������������������������������

���������������������������������������������

����������

������������������

������������

������������������

��������������������

�������������������������

����������

����������

����������� ����� �����

Plot of Fc-SCORES (%) for Season 1998-1999Farm - KLEU

Sheep categories - HAMELS, EWEREP, RAMMAT, EWEDRY, OTHER

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

HAM

ELS

12/7

/98

EWER

EP

12/7

/98

1/4/

99

RAM

MAT

3/1/

99

EWED

RY

4/19

/99

OTH

ER

1/4/

99

2/1/

99

2/8/

99

2/15

/99

29-0

3-99

5/3/

99

DATE of Fc-EVALUATION

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UP

�������������Fc - 5

�������������Fc - 4

��������������������������Fc - 3

��������������������������Fc - 2

�������������Fc - 1

Dosed

37

Figure 5.11

Figure 5.12

���������������������������������������������

������������������������������������������������������������������������������

����������������������������������������������������������������������

������������������������������������������������������������������������������������

����������������������������������������

��������������������������������������������������

���������������������������������������������������������������������������

����������������������������������������

���������������������������������������������������������������������������

����������������������������������������������������������������������

������������������������������������������������������������

����������������������������������������

������������������������������������������������������������������������������

�����������������������������������

������������������������

���������������

������������������

���������������������������������������������

�����������������������������������

���������� �����

�����������������������������������

����������

���������������

��������������������

����������������������������������������

������������������������

����������

����� ����� ����������

����������

����������

Plot of Fc-SCORES (%) for Season 1998-1999Farm - BÜHRMANN

Sheep category - EWEPRG, EWELAC, EWEREP

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

EWEPRG

22-09

-98

28-10

-98

30-11

-98

05-01

-99

08-02

-99

EWELAC

24-09

-98

08-12

-98

EWEREP

21-09

-98

30-10

-98

11-12

-98

07-01

-99

17-03

-99

29-09

-99

DATE of Fc-EVALUATION

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

������������������������ Fc - 5

������������Fc - 4

������������Fc - 3

������������������������ Fc - 2

������������Fc - 1

Dosed

��������������������������������������������������������������������������������

��������������������������������������������������������������������������������

��������������������������������������������������������

���������������������������������������������������������������������������

�����������������������������������������������������������������

���������������������������������������������

�����������������������������������������������������������������

����������������������������������������������������������������������

��������������������������������������������������

�������������������������������������������������������

��������������������������������������������������������������������������������

�������������������������������������������������������

����������

����������

����������������

���������������

��������������������

����������������������������������������

��������������������

���������������

������������������������������

������������������������������

����������

�������������������������

����� ���������� �����

����������

����������

���������������

���������� �����

�����

����������

����� ����������

Plot of Fc-SCORES (%) for Season 1998-1999Farm - BÜHRMANN

Sheep category - RAMMAT, HAMELS, RAMREP, OTHER

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

RAM

MAT

08-1

2-98

HAM

ELS

11-1

2-98

07-0

1-99

RAM

REP

21-0

9-98

29-1

0-98

05-1

1-98

08-1

1-98

30-1

1-98

10-1

2-98

21-0

2-99

03-0

3-99

29-0

9-99

OTH

ER

06-0

2-99

DATE of Fc-EVALUATION

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

��������������������������Fc - 5

�������������Fc - 4

��������������������������Fc - 3��������������������������Fc - 2�������������

Fc - 1

Dosed

38

Figure 5.13

Figure 5.14

������������������������������������������

��������������������������������������������������������

������������������������������

��������������������������������������������������������

������������������������������������������

�����������������������������������

������������������������������

������������������������������������������

������������������������������������������

������������������������������������������������������

���������������������������������������������������������������

������������������������������������������������������������������������

���������������������������������������������������������������

������������������������������������������������������������

������������������������������������������������������������������������������������

������������������������������������������������������������������������

����������������������������������������������������������������������

������������������������������������������������������������

������������

������� ������� ������ ���������������������

Plot of Fc-SCORES (%) for SEASON 1999Farm - SMIT

Sheep categories - WEANER, RAMMAT, EWEDRY, OTHER

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

WEANER

17-02

-99

25-03

-99

RAMMAT

20-01

-99

27-01

-99

17-02

-99

EWEDRY

16-04

-99

22-04

-99

OTHER

20-01

-99

27-01

-99

DATE of Fc-EVALUATION

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

�������������������������� Fc - 5�������������������������� Fc - 4

�������������Fc - 3

�������������������������� Fc - 2�������������������������� Fc - 1

Dosed

���������������������������������������������������������������

���������������������������������������������������������������

������������������������������������

������������������������������������

������������������������������������������������������

������������������������������������������������������

���������������������������������������������������������������������������������������������������������������������������������������

������������������������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������������������������������������������

������������������������������������������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������

���������������������������������������������

������������������

��������� ������������������

���������������������������

������������������

������������������������������������

��������� ��������� ���������

Plot of Fc-SCORES (%) for SEASON 1999 & 2000Farm - SMIT

Sheep category - LAMBS

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1999 17-02-99 23-02-99 01-03-99 04-03-99 10-03-99 17-03-99 2000 31-01-00

DATE of Fc-EVALUATION

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

�������������������������� Fc - 5

�������������������������� Fc - 4

�������������������������� Fc - 3

�������������������������� Fc - 2

�������������������������� Fc - 1

Dosed

39

Figure 5.15

Figure 5.16

����������

�������������������������

��������������������

������������������������������

������������������������

����������������������������������������

�������������������������

�������������������������

�������������������������

��������������������

��������������������������������������������������

������������������������������

�������������������������

��������������������

��������������������

�������������������������

�����������������������������������������������������������������

�������������������������������������������������������

�������������������������������������������������������

���������������������������������������������

����������������������������������������

����������������������������������������

�������������������������������������������������������

��������������������������������������������������

��������������������������������������������������

����������������������������������������

������������������������������

���������������������������������������������

����������������������������������������

��������������������������������

����������������������������������������

���������������������������������������������

����������

����� ����������

����������

���� ����� ����� ����������

����������

��������

����� ���������� �����

����������

������������

�������������������������

���������������

����� ����� ���� ����� ����� ����� ���� ����������

��������

��������� �����

Plot of Fc-SCORES (%) for SEASONS 1998-1999 & 1999-2000Farm - SMIT

Sheep category - EWELAC

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1998

199

908

-1228

-1205

-0127

-0117

-0223

-0201

-0304

-0310

-0317

-03

1999

200

023

-1212

-0131

-0114

-0221

-0220

-03

DATE of Fc-EVALUATION

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

�������������Fc - 5

�������������Fc - 4

�������������������������� Fc - 3

�������������Fc - 2

�������������Fc - 1

Dosed

������������������������������������������������������������������������

������������������������������������������������������������������������

������������������������������������

������������������������������������������������������

������������������������������������������������

���������������������������������������������

���������������������������������������������

������������������������������������������������������

��������������������������������

��������������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������������������������

����������������������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������������������������

������������������������������������������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������������������������

������������������������������������������������������������������������������������������������������������������������

��������� ��������� ������������������������

������������������

��������� ��������� �����������������

Plot of Fc-SCORES (%) for Season 1999Farm - SMIT

Sheep category - EWEPRG

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

17-02-99 23-02-99 01-03-99 04-03-99 10-03-99 17-03-99 25-03-99 16-04-99 22-04-99

DATE of Fc-EVALUATION

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

�������������Fc - 5

�������������Fc - 4

�������������Fc - 3

�������������Fc - 2

�������������Fc - 1

Dosed

40

Figure 5.17

Figure 5.18

��������������������

������������������������������������

������������������������������������������������������������������������������������������

��������������������������������������������������������������������������������

������������������������������������������������������������

������������������������������������������������������

��������������������������������������������������������������������������������

����������������������������������������

����������������������������������������������������������������������������������������������������������������������������������������������������������������

���������������������������������������������������������������������������������������������������������������������������������������

��������������������������������������������������������������������������������������������������������������

��������������������������������������������������������������������������������������������������������������

����������������������������������������������������������������������������������������������������������������������������������

������������������������������������������������������������������������������������������������������������������������������

��������������������������������������������������������������������������������������������������������������

����������������������������������������������������������������������������������������������������������������������������������������������������������������

������������������������������

������������������

���������� ��������������������

��������������������

��������������������

Plot of Fc-SCORES (%) for Season 1998-1999Farm - SMIT

Sheep category - EWEREP

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

26-11-98 07-12-98 14-12-98 28-12-98 05-01-99 20-01-99 25-03-99 08-04-99

DATE of Fc - EVALUATION

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

�������������������������� Fc - 5

�������������������������� Fc - 4

�������������������������� Fc - 3

�������������������������� Fc - 2

�������������������������� Fc - 1

Dosed

������������������������������������������������������������������������������������������������������������������������

������������������������������������������������������������������������������������������������������������������������������������������������������������

������������������������������������������������������������������

��������������������������������������������������������������������������������������������������������������

����������������������������������������������������������������������������������������

������������������������������������������������������������������������

������������������������������������������������������������������������������������������������������������

������������������������������������������������������������������������������������

������������������������������������������������������������

��������������������������������������������������������������������������������������������������������������

�������������������������������������������������������

����������������������������������������������������������������������������������������

������������������������������������������������������������������������������������������������������������������������������������

������������������������������������������������������������������������������������

������������������������

������������ ���������������������������������

���������������������������������

���������������������������������

������������������������

������������������������������������

������������ ����������� ����������������������

����������� ����������������������� ����������� ����������� ������������

Plot of AVERAGE Fc-SCORES (%) for Season 1998-1999 Farms- BESTER, BÜHRMANN, CILLIERS, KÖHLER, OBERHOLZER, SMIT

Sheep category - EWEREP

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

BESTER BÜHRMANN CILLIERS KÖHLER OBERHOLZER SMIT ALL-6

FARMS

PER

CEN

TAG

E of

SH

EEP

in F

c-SC

OR

E G

RO

UPS

����Fc - 5

����Fc - 4

����Fc - 3

�������� Fc - 2�������� Fc - 1

Dosed

Died

41

Section 6 Analysis of phenotypic and genotypic relationships of FAMACHA© data compared to other production data

Introduction Since the advent of modern anthelmintics, starting 4 decades ago, there has been a common tendency for farmers to attempt to exclude worm infections as a factor in animal production by frequent and/or strategic drenching. The effect was that the natural selection that occurred prior to this in nature, to the advantage of the more resistant and resilient animals, was practically neutralised. Nevertheless, as seen in the initial Badplaas trial and as is known from the literature (Barger, 1985), worm burdens are overdispersed among individuals in a host population, with similarly large differences in the susceptibility of animals to the effects of the worms. This is also dramatically illustrated by the use of the FAMACHA© system in the field, where numerous animals with normal, bright red clinically visible mucous membranes often predominate in a flock of sheep, while individuals are dying of haemonchosis. We hypothesised that the FAMACHA© system would make it possible to identify and cull the overly susceptible animals on the strength of paling of their conjunctivae, compared to the flock mean. This could have the effect of eliminating the �worm taxis� that previously practically determined the rate of drenching of the entire flock. Furthermore, granted that the phenomenon of increased susceptibility is hereditary, the culling of such animals will result in a slow rate of improvement of the ability of the flock to withstand worm infection. However, at the same time we hoped that the FAMACHA© system would be able to contribute to genetic selection of the most resilient and possibly also resistant animals, which would enable selection of the best rams and even ewes for future breeding, in which case faster genetic progress is theoretically possible. This question of the potential value of the FAMACHA© system for phenotypic and/or genotypic selection of animals was investigated in the field over 2 breeding systems. The use of selective drenching requires a reliable means to differentiate between those animals which, if left untreated, would be at risk of developing severe helminthosis and possibly dying, and those which would be in no immediate danger. Bisset et al. (1994 and 1996) in studies involving Romney sheep subjected to an extended period of natural challenge predominated by Ostertagia spp. and Trichostrongylus spp., tested selective drenching strategies based on animal condition or relative live-weight gain. However, in the summer rainfall area of South Africa, where the Merino is the predominant sheep breed and H. contortus - a blood feeder - is the predominant nematode parasite, an assessment of host anaemia may be a more appropriate means to determine when treatment is required. Haematocrit measurements are generally accepted as providing the most accurate indication of the severity of anaemia, but the process itself is time consuming and requires specialised equipment which few if any farmers possess. We have developed the FAMACHA© system which is a quick, simple, and cheap alternative to haematocrit measurements. At the same time as the implications of using the FAMACHA© scores as a guide on which to base selective drenching of sheep and goats are currently being investigated in the present trials, the feasibility of using a selective drenching strategy to identify sheep which are genetically better able to withstand challenge by H. contortus (and hence have lower drench treatment requirements) is also being investigated, with the aim of establishing a breeding programme to select for this trait in South Africa.

42

Principles of breeding sheep and goats for resistance/resilience to gastrointestinal helminth infection 1. Types of host response involved in limiting helminthosis Two different types of host response can potentially be targeted when seeking to breed

sheep or goats which require minimal anthelmintic treatment in the face of nematode parasite challenge. These are �resistance� (the ability of the host to suppress the establishment and/or subsequent development of a parasite infection) and �resilience� (the ability to maintain acceptable production while subjected to parasite challenge/infection).

Clunies-Ross (1932) was among the first to recognise the distinction between �resistance to

infection� and �resistance to the effects of infection�. The term �resilience� was subsequently adopted by Riffkin and Dobson (1979) to describe the latter trait.

Most studies of the genetics of resistance have focused on faecal FECs in animals under

limited worm challenge as the indicator trait/selection criterion. Assessing resilience in a practical situation is somewhat less straightforward. Riffkin and Dobson (1979) defined resilience as the ability to withstand the pathogenic effects of nematode infection. Pathogenic effects of infection (and/or challenge) can range from depression in growth rate, diarrhoea and/or anaemia, to death depending on the severity of challenge and the predominant species of parasite involved. In theory the ability of a host to withstand the effects of nematode infection can be defined and measured in terms of any of these parameters. In the absence of DNA markers, testing for resilience generally requires the animals being tested to be exposed to parasite challenge for extended periods, so that those least affected by challenge can be identified.

In reality the ability of an animal to limit helminthosis probably reflects a combination of both

resistance and resilience factors. Indeed Albers et al. (1987), in a study of Merino sheep/H. contortus interactions in Australia, found a moderate positive genetic correlation (r = 0.56) between resistance (FEC) and resilience (measured in terms of depression of growth rate while infected). This suggested that the traits were in some way linked and that therefore selection for either trait would ultimately achieve a similar endpoint. However, it is possible that the relationship between these traits differs for different sheep breed/parasite species interactions, as there is evidence of unfavourable correlations between FEC and various production traits in the case of Romney sheep/Ostertagia spp./Trichostrongylus spp. interactions (Bisset and Morris, 1996).

Anaemia is probably the main cause of production loss and death in sheep and goats in

areas of South Africa where H. contortus is the predominant parasite species. Selective breeding for the ability of sheep and/or goats to limit anaemia while exposed to challenge might therefore be expected to lead to sheep which require less drenching. Logically the severity of anaemia in an infected host can be expected to be related to both the ability of a host animal to resist infection and its ability to restrict blood loss caused by those worms which do establish themselves. However, results of work in Australia by Le Jambre (1995) have suggested that the most important factor in sheep is the ability to limit the size of infection, although there is some evidence that haemoglobin type may also be a factor in limiting anaemia (Allonby and Urquhart, 1976; Altaif and Dargie, 1978).

2. Heritabilities The heritability of a trait indicates what proportion of the phenotypic variation observed is

due to genetic factors as opposed to environmental factors, and therefore it gives an indication of the genetic progress that can be expected with selective breeding for a particular characteristic. A low heritability is indicative of a low correlation between

43

genotype and phenotype. It also often indicates that variation in a trait is due to additive gene action rather than a major gene. As an example, most estimates of the heritability of FEC in sheep are around 0.23 (i.e. about 23 % of the phenotypic variation seen is due to genetic factors). This is not substantially different from heritability estimates for many important productivity traits. Notwithstanding the fact that environmental factors obviously have a large influence on FEC, substantial progress can still be made in selective breeding, as indicated by the fact that the mean FECs of nematode resistant and susceptible sheep flocks developed at Wallaceville, New Zealand, currently differ by a factor of 21.

Heritability estimates are based on how much more relatives resemble each other for certain

traits than non-relatives. Estimates can be derived in several ways but all require knowledge of the animals� pedigree. The most commonly used method of calculating heritabilities for particular traits in livestock is probably the half-sib analysis. This generally requires the generation of relatively large data sets and the use of specialised software to adjust the data appropriately to account for �environmental� factors such as the source, birth date, birth rank and dam age, which may have influenced the animals� performance, and to calculate heritability estimates and other genetic parameters.

3. Genetic correlations/Correlated responses Genetic correlations indicate whether selection for a particular trait will have an effect (either

desirable or otherwise) on other traits. They are estimated either statistically by assessing the relationships between traits in different sire progeny groups (as opposed to individual animals) or by selecting for one trait over a period of time and noting whether or not there is a change (correlated response) in other traits not directly selected for. As noted above, in New Zealand studies on Romney sheep under challenge mainly by Ostertagia spp., and Trichostrongylus spp., there tends to be a slightly unfavourable genetic correlation between FEC and some production traits (Bisset and Morris, 1996). However, Australian evidence suggests that this may not be the case for Merino sheep under challenge with H. contortus.

4. Effect of environment on genetic rating A common concern is that animals ranked as more resistant or resilient on 1 farm or in 1

year may not maintain their ranking on another farm or in another year. This concern has been the subject of several studies in New Zealand, all of which have indicated that such factors have little or no effect on genetic merit under practical farming conditions (McEwan et al., 1997).

5. Calculation of breeding values As indicated above, the performance of an individual (i.e. its phenotype) is influenced by

both �environmental� and genetic factors. The lower the heritability of a trait, the lower the correlation between its genotype and its phenotype. For this reason in order to make rapid progress in selective breeding programmes it is usually preferable to design a breeding strategy that will allow the calculation of breeding values (estimates of genetic merit) rather than to rely on phenotypic ratings as the basis of identifying desirable breeding animals. Breeding values take into account the performance of an individual�s relatives as well as its own performance, and thus an accurate pedigree for all animals being tested is essential for the analysis. For some traits (such as FEC) transformation is needed prior to analysis. Calculation of breeding values is normally done using specialised genetics software. Various scales are used to express breeding values, but commonly they are expressed in terms of the number of standard deviations above or below the flock mean value for that trait or alternatively as a percentage above or below the average flock value.

44

6. Indices Indices are a way of combining breeding values from different traits so that the best overall

progress is made for each. The influence of each trait on the final index is determined by how much weight that trait is given in relation to other traits. The amount of weight given to each trait depends on its relative economic value (since all traits are not equal in this respect), its heritability, and its genetic correlations with other traits.

Veld ram system for comparing rams from different breeders A veld ram club (literally translated as �field ram club�), of which there are more than a dozen in South Africa, comprises a field comparison of young (more or less 6 months old at the start) rams of club members on common pasture, over a period of 10-11 months. At the end of the period the rams are evaluated for growth and wool quality and production, in addition to certain conformational features, and the best performers are sold on a public auction. While it was common previously to attempt to exclude worms as a factor in comparing the rams by regular drenching as frequently as every 4 weeks, the members of the Oos Vrystaat Veldramklub agreed to include the dominant worms of their region (overwhelmingly Haemonchus spp.) as a factor in the comparisons during the 1998/1999 and the 1999/2000 tests. Merino stud During the 1999/2000 Haemonchus spp. season the investigations involving the veld ram club were extended to a Merino stud, with genealogy records since 1989. The farm is situated east of Ermelo in Mpumalanga Province. The farmer concerned follows a similar approach to the veld ram club, in that the rams he prepares for auction are also compared with one another on veld grazing, instead of being fed in stalls, as is commonly practised by stud breeders. Thus this can also be regarded as a veld ram production system, with all the lambs (rams and also ewes) belonging to the same owner. But the principal difference in the Merino stud, compared to the veld ram club, is that pedigree data are available for genetic analyses.

Methods Veld Ram Club Management The between-ram comparisons described below for the veld ram club were continued over only about 7 months per annual ram intake, whereafter all of the animals were drenched with an effective anthelmintic and finished with supplementary feed for the rest of the period. At intervals of 3 weeks during the tests the 6 month-old rams were weighed and evaluated by the FAMACHA© system. A haematocrit determination was done for every animal with a FAMACHA© score of 4 or 5, and only those animals with haematocrit scores of 15 % or below were drenched. As the Haemonchus season advanced and the worm challenge became intense from January 1999, onwards, the FAMACHA© evaluation and sampling of individual animals for haematocrit analyses and drenching was done more frequently, until the animals were examined weekly during the peak worm season, in February. This increased frequency of inspection was not necessary during the 1999/2000 worm season, as lower levels of infection were experienced, and only one ram required salvage drenching. Data analysis Phenotypic relationships between several traits were assessed by correlation analyses, using ranks rather than raw data to avoid problems associated with dealing with a mixture of continuous and discrete variables, as well as differences in distribution of the data.

45

Calculations of individual (phenotypic) merit for weight gain under challenge, FEC and haematocrit percentage were undertaken at Ruakura Agricultural Research Centre using specialised genetics software. As is standard procedure in such analyses, the data sets for each of the above traits were checked to see if adjustments for fixed (non-genetic) effects were required. Only �source of ram� was considered as a fixed effect as other data normally considered, such as date of birth, birth rank and dam age were not available from the veld ram club flock. Effects related to source of rams proved not to be significant for any of the traits. Individual performance ratings for all animals with respect to each trait were expressed in terms of the number of standard deviations above or below the flock mean value for that trait (residuals). These were subsequently converted to a scale whereby animals of average merit in a given trait received a value of 100, with those of above or below average receiving merits respectively more or less than 100, with 1 standard deviation equivalent to 20 rating points. This was done in order to allow presentation of performance ratings for the resistance and resilience traits on a scale familiar to the participating farmers. In calculating individual performance ratings, adjustment was made in 1998/1999 only in the case of those animals that were given an individual drench during the test period; in 1999/2000 the levels of infection were relatively low, to the extent that only one animal received an individual treatment; 5 weeks later a FEC again indicated a relatively heavy infection compared to most of the others in the flock. In the case of FEC data, a transformation to loge (FEC+100) was used to normalise distribution of data before analysis. For both ram intakes exploratory analyses were undertaken to examine phenotypic relationships between each of the traits considered to be of potential importance in assessing an animal�s ability to withstand haemonchosis (viz. relative growth rate under worm challenge, FEC, haematocrit level and FAMACHA© score). On introduction to the common pasture in each year, FECs and haematocrit determinations were done on every ram, after which all were drenched with a combination of anthelmintics and monitored 10-14 days later to confirm whether or not the treatment had been effective. On both occasions it was highly successful. Thereafter only salvage drenching was done as described below. Only at the termination of the test was the whole flock drenched again. 1998/1999 ram intake Individual ram anthelmintic requirements were not analysed as such, as the onset of haemonchosis in the flock was rapid and hence the time frame of the selective treatment programme was very short resulting in minimal between-animal variation in drench timing. The short time frame also meant that no animal received more than one additional anthelmintic treatment relative to its flock-mates before a decision was taken to administer treatment to all members of the flock. One way analysis of variance tests and/or Kruskal-Wallis non-parametric tests were used to compare the performance of contrasting sub-groups within the flock. Correlation analyses were used to assess phenotypic relationships between selected traits. Where appropriate these were undertaken using ranks rather than raw data to avoid problems associated with dealing with a mixture of continuous and discrete variables and also differences in distribution of the data. Effects related to source of rams proved not to be significant for any of the traits tested and so no adjustments were required. Individual performance ratings for all animals with respect to each trait were expressed in terms of the number of standard deviations above or below the flock mean value for that trait (residuals).

46

Individual performance ratings for weight gain under challenge (period 15/12/98 to 16/2/99) and FEC (16/2/99) were calculated only for those animals not receiving drench treatment before that date (i.e. those of higher than average phenotypic merit in terms of low drench requirements). This avoided the potentially confounding influence of selective drench treatment on these parameters. In the case of FEC, a transformation to loge (FEC +100) was used to normalise distribution of data before analysis.

47

Performance ratings for haematocrit values were calculated using an average of measurements taken on 16/2/99 and 26/2/99 (after appropriate adjustments had been applied in the case of those animals which had earlier received selective drench treatment). For animals drenched on 10/2/99, haematocrit values taken on 16/2/99 were adjusted to account for drench treatment by deducting 1.3 percentage points from each (i.e. the average amount lost over the intervening period by animals sampled on 10/2/99 but not drenched). A similar approach was used to adjust the 25-26/2/99 values of those animals which required drenching on 16/2/99 (2.8 percentage points deducted in this case). While it is recognised that this adjustment may to some extent have been prone to bias, any effect is expected to have been relatively minor and indeed to have had no influence on recognition of the most resilient or resistant animals. In order to present performance ratings for the resistance and resilience traits on a scale familiar to participating farmers, ratings were converted to a scale whereby animals of above average merit for a particular trait received a score of >100 and those below average merit received a score of <100, with 1 standard deviation being equivalent to 20 rating points. 1999/2000 ram intake A similar management system was used for testing the stud animals as in the case of the veld ram club, excepting that sheep classified as FAMACHA© 3, 4 and 5 were drenched. The methods were similar to those used in the 1998/1999 ram intake. However, the level of worm infection was much lower, with the result that the ram-ram evaluation on the veld grazing had to be terminated before the rams had been stressed nearly as much with heavy Haemonchus spp. challenge as in the previous ram intake. Merino stud The availability of both performance records and pedigree data for 1999-born ewe and ram hoggets from the Merino Stud flock allowed the calculation of heritability estimates for various resilience, resistance and productivity traits of potential interest, as well as phenotypic and genetic correlations between them. In addition, the pedigree records made it possible to calculate breeding values (estimates of the genetic merit) for each of the animals included in the analysis. Performance records, including liveweights, haematocrit levels, FAMACHA© scores, faecal worm egg counts, greasy fleece weights and fibre diameters from a total of 523 1999-born ewe and ram hoggets, were examined in the analysis. The hoggets involved were the progeny of a total of 21 different sires. Data from those animals whose sire was not known were excluded from the analysis. Faecal egg count data were analysed following transformation to natural logarithms. Because there were some zero values in the FEC data, it was necessary to add a constant (100) to all values before transformation. Preliminary fixed effects models were analysed for each trait of interest using GENSTAT (1994). On the basis of these preliminary analyses, fixed effects which were taken into account in the subsequent analyses included sex, age of dam (2, 3, 4, 5+), birth rank (1, 2+), and date of birth. Multi-trait restricted maximum likelihood (REML) procedures were used to calculate heritabilities of the traits examined as well as phenotypic and genetic correlations among them, and individual breeding values for all animals involved (and their sires). In addition, single-trait REML estimates of heritability were calculated for comparative purposes. It was not practical to analyse drench treatment data as such as very few animals had been deemed to require treatment on each of the occasions that they were brought in for assessment. In the ram flock, a total of only 9 animals received individual drench treatments on the basis of their FAMACHA© scores during the test period (2 on 12/1/00, 2 on 2/2/00, 1 on 18/2/00, and 4 on 21/2/00). In the ewe flock a total of 18 animals were drenched over the same period on a similar

48

basis. In order to achieve a useful level of genetic variation in relation to individual drenching requirements, approximately 50 % of flock members would need to have received at least 1 treatment during the test period.

Results 1998/1999 ram intake Under the selective drenching strategy used in the veld ram test, 32 of the 188 yearling rams being tested were treated with anthelmintic on 10/2/99 on the basis of their low haematocrits. A further 23 were treated on the same basis on 16/2/99, 14 more were treated on 23/2/99, and another 21 on 25/2/99. All members of the flock, regardless of previous drench treatments, were subsequently drenched on 26/2/99. In order to further investigate the implications of using haematocrit levels as the basis for selective drenching, the performance of animals deemed to require drenching on 16/2/99 was compared to that of those not requiring drenching. Results are summarised in Table 6.1. Table 6.1 Comparison of mean live-weight (mass), live-weight gain

(15/12/98 – 16/2/99), condition score, haematocrit level, FAMACHA© score and FEC in rams deemed on the basis of haematocrit level to require drench treatment on 16/2/99 and those not requiring treatmenta Drench required No drench required Significance

Live-weight (kg) 43.35 45.95 **

Wt gain (15/12 � 16/2) (kg) 2.48 3.92 **

Condition score 2.33 2.47 *

Haematocrit % 15.7 24.0 **

FAMACHA© score 3.48 2.20 **

FEC 22400 11559 **

aNote : data from animals drenched on 10/2/99 were excluded from this analysis. *P<0.05; **P<0.01. As can be seen, every aspect of animal performance examined differed significantly between these groups. The results strongly suggested that haematocrit measurements had effectively identified those animals which exhibited both resistance to Haemonchus contortus infection (as indicated by FEC) and resilience to infection (as indicated by growth and condition parameters). The concept that FAMACHA© scores can provide a useful indicator of anaemia was also supported by these results. Results of rank correlation analyses between the above traits in animals sampled on 16/2/99 but which had not received drench treatment on 10/2/99 are presented in Table 6.2. As can be seen there were significant correlations, in the directions expected, between all the indicators of resistance/resilience examined with the exception of FEC and liveweight. They thus provide further support for the trends identified in Table 6.1 and, in addition, give an indication of the strengths of the relationships between each of the traits at the phenotypic level.

49

Table 6.2 Correlations between live-weight (mass), live-weight gain ( 15/12/98 – 16/2/99), condition score, haematocrit level,

FAMACHA© score and FEC in rams on sampling date 16/2/991 [Cell contents : Correlation; P-value]a

Live-weight L-wt gain Condition Haematocrit FAMACHA©

L-wt gain 0.217 0.007

Condition 0.385 0.000

0.274 0.001

Haematocrit 0.202 0.012

0.343 0.000

0.288 0.000

FAMACHA© -0.300 0.000

-0.338 0.000

-0.274 0.001

-0.774 0.000

FEC -0.113 0.198

-0.323 0.000

-0.232 0.008

-0.665 0.000

0.596 0.000

aNote : data from animals drenched 10/2/99 were excluded from this analysis. The strong relationship between FAMACHA© scores and haematocrit levels proved to be maintained regardless of drenching history, as indicated by the results of a rank correlation analysis of mean haematocrit levels and FAMACHA© scores of all animals sampled on 16/2/99, including those which had been drenched on 10/2/99 (r = �0.765; P<0.0001). Thus, 61 of the 98 animals in the flock which received a mean FAMACHA© score of 2.5 or more on 16/2/99 had a mean haematocrit value of 20 % or less, while the remaining 37 had values >20 %. Importantly however, there were only 3 animals which received a mean FAMACHA© score of <2.5 which had haematocrit values of <20 % (i.e. there was a very low number of �false positive� cases, that would have been drenched unnecessarily). Significantly there was also a relatively strong relationship between haematocrit levels and FECs in the animals tested. When only data from those animals left untreated on 10/2/99 were considered, the rank correlation between haematocrit levels and FECs in the animals on 16/2/99 was highly significant (r = �0.665; P<0.0001). In the case of animals remaining untreated on both 10/2/99 and 16/2/99, this relationship proved to be even stronger on 26/2/99 (r = �0.735; P<0.0001). 1999/2000 ram intake Results of rank correlation analyses between selected traits in animals sampled on 8/3/00 are presented in Table 6.3. There were significant correlations, in the directions expected, between haematocrit readings and the other resistance, resilience and productivity traits examined. Notably however, with the lower worm challenge levels experienced in this season, the correlations between FAMACHA© scores and the other traits examined were considerably lower than the estimates of the 1998/1999 season, and statistically significant only in the case of condition score and haematocrit. This indicates the limitation of the FAMACHA© scoring system in discriminating between animals when under minimal parasite stress. Under such circumstances haematocrit readings might be expected to be rather more sensitive. Veld ram auction (1999/2000) Price paid per ram compared to worm infection indices At the conclusion of the veld ram test of 1999/2000, 50 of the rams were auctioned on a public sale in July 2000.

50

The price paid per ram was correlated by multiple regression analysis (with price as dependent variable), including the various production and �worm� indices, as well as the breeder of each ram. In the former a stepwise deletion process was used, in which the values correlating least significantly with the price per ram were excluded in turn from the multiple regression analysis. The results indicated that, according to the multiple regression method, the prices were most highly significantly partially correlated with the fibre diameter (P<0.0001) and live mass indices (P<0.0019) of the rams, and also significantly partially correlated with the worm resilience indices (P<0.02). There was, however, no significant partial correlation with the worm resistance indices (P>0.176). We fitted a model where the variables (each expressed as an index per sheep) �breeder�, �live mass�, �fibre diameter�, �resistance� (of the sheep to worm infection) and �resilience� (ditto) were used as predictors for price. The interactions among the predictors were also tested, but were found to be totally insignificant. Hence we concentrated on the model for price with the above-mentioned 5 variables as predictors. The model was fitted to the data and an r2 value of 64 % was found. We can therefore conclude that 64 % of the variation in price can be ascribed to the 5 predictors. If the predictors were examined with respect to the partial effect of each on price, the following was found : Only the indices �live mass� and �fibre diameter� were found to be significant. The model estimated that the price paid at the sale increased by R 74 for every kg increase in �live mass�, and, secondly, that for every unit decrease in �fibre diameter� the increase in price was estimated to be R 173. Next we fitted the second model where we excluded �breeder�. The reason for this was the suspicion of a possible correlation between �breeder� and �resistance� and �resilience�. Now the r2 decreased to 46.3 %, and �live mass� was a significant predictor, with a P-value of P<0.0001. Also, �fibre diameter� was a highly significant predictor of price, and, lastly, �resilience� now also became a significant predictor (P=0.0192). This strengthened our hypothesis that �breeder� and �fibre diameter� were probably correlated. If �resilience� was also deleted from the model, and only �live mass�, �fibre diameter� and �resistance� were included, then �resistance� also became significant (P=0.0184). This indicated to us that there was also some degree of correlation between �resilience� and �resistance�. Finally, another model was run, with �resilience� as dependent variable (instead of �price�, as above), and with �breeder�, �live mass� and �fibre diameter� as predictor variables. In this model �breeder� did not come up as a significant predictor (P=0.1547). However, if we observe the individual �resilience� means per �breeder�, fairly large differences were observed, with a range of �resilience� index values of 69.0-145.1. Hence this analysis indicated a correlation between �breeder� and �resilience�, which supported our earlier contention.

51

Table 6.3 Correlations between live-weight (mass), live-weight gain (29/9/99 – 8/3/00), condition score, haematocrit level, FAMACHA© score and FEC in rams on Van Zyl property on sampling date 8/3/00 [Cell contents : Correlation; P-value]

Live-weight L-wt gain Condition Haematocrit FAMACHA©

Live-wt gain 0.284 0.001

Condition 0.461 0.000

0.355 0.000

Haematocrit 0.311 0.000

0.257 0.004

0.313 0.000

FAMACHA© -0.100 0.265

-0.016 0.858

-0.241 0.007

-0.266 0.003

FEC 0.028 0.762

0.145 0.108

-0.022 0.813

-0.263 0.003

0.044 0.622

Merino stud Multi-variate estimates of heritability for weight gain (ram hoggets : 11/11/1999 � 21/12/1999; ewe hoggets : 12/11/1999-12/1/2000); log(FEC+12000), FAMACHA© score and haematocrit reading (ram hoggets : 23/2/2000; ewe hoggets : 17/2/2000); and greasy fleece weight and fibre diameter (ram and ewe hoggets : both 13/4/2000), along with phenotypic and genetic correlations among them, are presented in Table 6.4. By way of explanation, multi-variate estimates of heritability (which are shown on the diagonal in Table 6.4) take into account not only the genetic variability of a particular trait (effectively the variation between sire progeny groups) but also the strength of the genetic correlations with the other the traits measured. Thus, if weight gain for example has a low estimate of heritability when considered by itself (i.e. a uni-variate estimate) but is strongly correlated with another trait which has a much higher heritability, the model will indicate that the uni-variate analysis has under-estimated the heritability of weight gain and adjust it upwards accordingly. The phenotypic correlations among the various traits are shown above the diagonal in Table 6.4. These are effectively just straight forward correlations among the traits in individual sheep but with corrections applied to adjust for fixed (non-genetic) effects such as sex, birth-date, age of mother, birth rank etc. The genetic correlations are shown below the diagonal. The easiest way to think about these is that they are correlations among the average values for each of the sire progeny groups included in the analysis. Thus, in the case of genetic correlations, n = the number of sires used, not the number of individual animals tested.

52

Table 6.4 Heritabilities, phenotypic and genetic correlations (± S.E.) among resistance, resilience and productivity traits estimated by multi-variate ASREML analysis (Merino Stud flock) (heritabilities on diagonal [bold] phenotypic correlations above and genetic correlations below the diagonal)

WtGain LogFEC F'A Haem FlcWt FDiam

WtGain 0.22±0.10 -0.08±0.05 -0.12±0.05 0.13±0.05 0.16±0.05 0.21±0.05

LogFEC -0.91±0.19 0.47±0.16 0.59±0.04 -0.72±0.03 -0.18±0.05 -0.15±0.05

F'A -0.63±0.24 1.00±0.05 0.55±0.17 -0.73±0.03 -0.17±0.05 -0.12±0.05

Haem 0.73±0.22 -0.99±0.04 -0.99±0.03 0.55±0.17 0.16±0.05 0.19±0.05

FlcWt 0.17±0.39 -0.64±0.25 -0.75±0.21 0.63±0.25 0.21±0.11 0.24±0.05

Fdiam 0.76±0.18 -0.66±0.19 -0.65±0.19 0.57±0.20 -0.30±0.35 0.36±0.12

Multi-variate estimates of heritability for FEC, FAMACHA© and haematocrit readings were all reasonably high. Although the limited amount of data used in this analysis meant that the standard errors were also relatively high, the heritability estimates clearly indicate promising levels of between-sire variability for these traits, suggesting that good genetic progress should be possible in selective breeding programmes. Phenotypic and genetic correlations among these traits were also very high, particularly in the case of genetic correlations, which were essentially all unity. Thus, the sire progeny groups which on average maintained the highest haematocrit levels (and lowest FAMACHA© scores) under challenge, were consistently those which also had the lowest faecal worm egg counts and vice versa. While these values would undoubtedly have been somewhat lower had the worm challenge experienced by the sheep been less intense (especially perhaps in the case of FAMACHA© score), the results nevertheless indicate the value of using genetic analyses in the case of parameters which appear to have limited sensitivity in a phenotypic sense. Heritability estimates for the productivity traits (weight gain and greasy fleece weight) and fibre diameter were in a similar range to those seen for these traits in sheep in most other data sets. Importantly however, there were strong favourable correlations between the productivity traits and both haematocrit levels and FAMACHA© scores. Thus, sire progeny groups which, on average, were able to maintain high haematocrit levels (and low FAMACHA© scores) despite exposure to natural challenge by H. contortus, also tended to show the highest growth rates and greasy fleece weights. Not unexpectedly, correlations (both genetic and phenotypic) between fibre diameter and anaemia levels were unfavourable, there being a tendency for sire progeny groups which suffered higher than average levels of anaemia to have the finest wool. In contrast to results of similar analyses carried out on data derived from Romney sheep exposed to natural challenge with Trichostrongylus spp. and Ostertagia spp. in New Zealand, genetic correlations between FEC (i.e. resistance to infection) and productivity traits were both significant and favourable. This is very important because it indicates that selection undertaken using haematocrit, FAMACHA© or FEC in Merinos challenged with H. contortus would each achieve a similar end point (i.e. both resistance and resilience to Haemonchus), granted that the present results are repeatable and that other Merino flocks do not differ much in this respect from the one in which we did the investigation.

53

Uni-variate estimates of heritability for each of the above traits are presented in Table 6.5. These take into account the genetic variability of just one trait at a time without considering the strength of the genetic correlations. In general these estimates are somewhat lower than estimates calculated using multi-variate analyses, but nevertheless they still support the multi-variate estimates in suggesting useful levels of genetic variability. Table 6.5 Heritabilities of traits of interest estimated by uni-variate

ASREML analysis h2

WtGain 0.09±0.09

LogFEC 0.31±0.14

F'A 0.33±0.14

Xhaem 0.41±0.16

FlcWt 0.31±0.15

Fdiam 0.33±0.16 Because we know how well their progeny performed in relation to the various traits measured we can also calculate a breeding value for each of the sires used on the Merino Stud property. These are presented in Table 6.6 below. Table 6.6 Breeding values for selected traits for each of the sires used

in the 1999 breeding season on the Merino Stud property Sire ID LWGbv FECbv F'Abv Haebv GFWbv FDbv HaeInd

922643 -0.0326 -0.7344 -0.4693 3.8150 0.4355 0.4164 0.7111

939539 -0.5363 -0.0450 0.0142 -0.1048 0.0636 -0.8076 -0.0202

940173 0.0545 0.2483 0.2362 -1.2800 -0.1766 0.2511 -0.2797

950113 -0.7575 0.9338 1.4290 -5.5510 -0.4048 0.1706 -1.4244

960229 0.0594 -0.5283 -0.7109 3.9270 0.2611 -0.5173 0.8510

960419 0.0768 0.0852 -0.2033 -0.1867 -0.2170 0.2583 0.0834

970062 -0.6418 0.6847 0.8302 -3.3950 0.2141 -1.6940 -0.8483

970152 0.2565 -0.3343 0.2475 0.9667 -0.1871 0.9488 -0.0112

970178 1.0390 -0.1828 -0.5859 1.3700 -0.1387 1.0520 0.4734

970226 0.3962 0.0879 0.6130 -2.3970 -0.0468 -0.0781 -0.6130

970261 -1.3480 0.5093 0.6084 -2.6410 0.0460 -0.2527 -0.6403

970268 0.7501 -0.0482 -0.1619 1.0610 0.0178 -0.0232 0.2141

970279 0.4303 0.0181 -0.2735 0.4618 -0.0755 -0.2634 0.1993

970305 -0.1591 -0.0643 -0.2329 0.5274 0.0470 -0.1645 0.1861

970385 -0.4907 0.1411 0.7722 -2.0800 0.0054 0.1862 -0.6574

970406 -0.5422 0.8302 0.9691 -4.2590 -0.1716 -0.3259 -1.0264

970442 -0.2706 -0.4680 -1.1590 3.0280 0.2518 -0.0854 0.9752

970521 0.7450 -0.4890 -0.5818 3.0260 0.1578 0.2959 0.6735

970791 0.2032 0.3820 0.3504 -2.5000 -0.3879 -0.1738 -0.4884

970807 0.5902 -0.3819 -0.5679 2.2090 0.0940 0.4499 0.5664

971010 0.1774 -0.6444 -1.1240 4.0020 0.2119 0.3571 1.0759

54

Discussion and conclusions Veld Ram Club The Veld Ram Club results over the 2 seasons were subjected only to phenotypic analyses as pedigree information was not available for the rams being evaluated by the club and even if the information had been available, sire progeny groups represented in the test flock would almost certainly have been too small to undertake meaningful genetic analyses. Although it was possible to assess the individual (phenotypic) merit of particular rams in relation to resistance and/or resilience from an analysis of the present data, it is important to remember that an individual�s phenotype is influenced by both �environmental� and genetic factors. Thus a number of factors such as birth rank (i.e. whether a single, twin or triplet), birth date, nutritional level and experience of infection before entering the test flock, as well as the age and physical condition of the mother, may have had a significant impact on the performance of the rams tested. The individual performance ratings obtained should therefore be treated with caution, as they provide only a broad guide to the potential genetic merit of particular rams in relation to the traits measured. Similarly, phenotypic correlations between different traits do not always provide a good indication of the potential genetic relationships between them. Thus, in some cases they give a poor indication of the strength or even the direction of correlated responses that are likely to occur over time in flocks selected for a particular characteristic. Performance ratings were calculated for 3 traits � FEC (as an indicator of resistance to infection); live weight gain (gain in mass) while under challenge (as an indicator of resilience); and haematocrit (which the correlation analyses above suggest may provide an indication of a combination of both resistance and resilience factors). As it was not possible to adequately assess genetic correlations between the traits (or estimate heritabilities) from the veld ram club data set, there was no adequate basis on which to attempt to calculate indices which combined more than one of the traits. Based on the analyses above, it was therefore concluded that performance ratings derived from haematocrit values would provide potential ram buyers with the most useful and reliable measure of an animal�s ability to withstand challenge by H. contortus. Heritability estimates are essential to predicting the progress that might be made in selecting for particular traits in a breeding programme. Using some of the rams tested in the Veld Ram flock to generate progeny for testing in a subsequent season could provide an opportunity to generate the data required to calculate heritability estimates for haematocrits and FAMACHA© scores. Either of two approaches could potentially be employed. Both would require a knowledge of at least the sire of each of the ram lambs to be tested, and preferably also its date of birth, and whether it was raised as a single, twin or triplet.

Recommendations relating to practical breeding programmes and generation of genetic parameters Breeding strategies The main components required to establish a successful programme to breed sheep for the ability to withstand challenge by H. contortus include (1) a flock (or flocks) of adequate size in which the ewes are single-sire mated and the lambs are tagged and recorded at birth (to ensure that all pedigrees are accurately known), and (2) owners or managers who have, or are prepared to put in place, a carefully maintained performance recording system, and who are prepared to expose young stock (particularly ram lambs) to a test period of untreated worm challenge/infection (a situation requiring the ability to closely monitor the animals to ensure none develop severe helminthosis).

55

From the results of this veld ram club project, haematocrit readings of animals under natural challenge would appear to provide the most direct and reliable indication of their ability to cope with H. contortus infection. Another factor in favour of haematocrits is that they are also positively correlated (at least at the phenotypic level) with most other indicators of resistance/resilience. While FAMACHA© scores could be used instead, if necessary, they might be expected to be somewhat less reliable. FECs could also be used if preferred although the evidence suggests that they would probably be a less direct and reliable indicator of an animal�s ability to cope with nematode challenge. On the other hand FECs might have the advantage that they would provide a better indication of the size of burdens of parasite worms other than Haemonchus should these become a problem. Challenge protocol and data collection In order to assess an animal�s ability to withstand parasite challenge in the absence of suitable DNA markers, it is obviously necessary to subject it to challenge for an appropriate period. As pointed out by Campbell (1986) in relation to selective breeding for disease resistance traits under field conditions, more genetic progress is likely to be made in years (or in flocks) where challenge is higher, due to clearer discrimination among progeny groups revealing greater genetic variation. Of course, in a commercial farm situation this must be carefully balanced against the greater risk to productivity and survival of the animals involved. Calculation of genetic parameters As emphasised above, in order to calculate genetic parameters it is essential to have a good pedigree and performance recording system in place, to have adequate numbers of progeny from each of the sires used, and to have all animals tested in the same environment or, if not, to have adequate genetic linkages (e.g. use of reference sires) between the different test environments involved. Accurate estimation of genetic parameters requires data transformation in some cases, adjustments to be made for fixed (�environmental�) effects (such as birth date, birth rank, dam age where available, pre-test data, etc.) and animal relationships to be taken into account. These usually involve the use of specialist genetics software. Veld ram auction (1999/2000) While our investigations into differences in susceptibility of the veld rams were publicised in a few articles in lay periodicals and in meetings with members of the veld ram club, the results of the analyses were available to the persons attending the auctions only shortly beforehand. Furthermore, very little information concerning the full implications of the differences in indices between individual animals and of those between the resistance and resilience indices was given before the auction of 1999/2000 to the buyers. It can also be expected that it will take time before any possible potential value of such results can gain acceptance by the farmers. This should also be viewed in light of the fact that before the present trials, no animals that had been investigated as to susceptibility to worms or their effects have been offered on sale to farmers in South Africa. It is probably to be expected that differences between breeders should exist in the mean resistance and resilience of the sheep in the different flocks to worm challenge, owing to the degree to which each breeder has through drenching protected the animals from production losses from worm infection. Our data support such a possibility, but this should be repeated with more data from different regions, before it can be accepted with confidence. Merino stud From analyses of the replacement lambs in the stud it can be seen that the sires with the highest genetic merit for ability to withstand the effects of Haemonchus challenge (i.e. those with minimal tendency to develop anaemia under challenge according to the haematocrit/FAMACHA© index) are 971010, 970442, 960229 and 922643 (Table 6.6). Reflecting the strong genetic correlations

56

between haematocrit, FAMACHA© and FEC in the flock, the progeny of these rams also tended to have, on average, the best genetic merit for resistance to infection (i.e. negative breeding values for FEC). Although these progeny groups were not amongst those with the lowest fibre diameter, they were amongst the best for fleece weight. While it would be up to the owner to decide what weighting he wanted to give to each of the traits listed in making his ram selections, appropriate use of the information to determine which rams to use in the upcoming season could enable him to make significant progress in the ability of his lambs to withstand Haemonchus challenge. Importantly, by selecting those rams with a good haematocrit index to breed from, he should also be able to significantly reduce the level of pasture contamination with Haemonchus eggs deposited by his lambs and thus gain an additional epidemiological benefit in terms of productivity and drenching requirements.

Recommendations for practical breeding programmes General considerations The results above indicate that with appropriate breeding strategies in place it should be possible to make good genetic progress in selecting for Merino sheep which can withstand challenge by H. contortus with minimal recourse to anthelmintics. The high (and favourable) genetic correlations between indicators of anaemia (haematocrit levels and FAMACHA© scores) and faecal egg count potentially give an additional bonus, as selecting animals on the basis of high breeding values for anaemia would result in a simultaneous improvement in resistance (i.e. lower faecal worm egg counts), leading to lower pasture contamination rates and associated epidemiological benefits. For most commercial sheep farmers the simplest and most economical route to incorporate genes for resistance and resilience to nematode infection in their flocks would generally be to buy their replacement rams from an established ram breeder or stud owner who has a performance recording system in place and is testing for the desired traits. As indicated in the previous report, the main components required to establish a successful programme to breed sheep for the ability to withstand challenge by H. contortus include (1) a flock (or flocks) of adequate size in which the ewes are single-sire mated and the lambs are tagged and recorded at birth (to ensure that all pedigrees are accurately known), and (2) owners or managers who have, or are prepared to put in place, a carefully maintained performance recording system, and who are prepared to expose young stock (particularly ram lambs) to a test period of untreated worm challenge/infection (a situation requiring the ability to closely monitor the animals to ensure none develop severe helminthosis). These requirements are unlikely to be practical for most commercial sheep farmers other than ram breeders or stud owners. In addition, in order to capture as much genetic variability as possible (and thus maximise genetic progress), it would be desirable for groups of breeders to establish genetic links between their flocks for the purpose of enabling them to screen much larger pools of animals for desirable characteristics. Flocks do not need to be run together for this purpose but provision does need to be made to use selected �reference sires� across more than one flock to form links for genetic analyses. This is easily accomplished where artificial insemination is practised simply by using semen from selected sires to generate progeny (15-20 male lambs) in each of the linked flocks. With appropriate planning the same sort of result can be readily achieved using natural mating. In relation to this it should be pointed out that the strategy currently used by the veld ram club involved in the project is of a poor design if optimal genetic progress is to be made in breeding for the ability of sheep to cope with Haemonchus challenge. In New Zealand a service has been established to help sheep breeders select for animals which are resistant to roundworm infection as well as being highly productive. This service is known as

57

�WormFECTM�, and it provides advice to participating breeders, offers a faecal worm egg counting service and calculates breeding values for several resistance traits (in New Zealand these are mainly based on faecal worm egg counts but an alternative immunological test is also available) along with an overall index of resistance. Breeders are able to select animals based on the information provided by the WormFEC service, along with data collected for other traits. In addition, they can supply breeding values for these traits with all sheep sold to their clients. Selection traits Haematocrit readings from sheep under natural challenge undoubtedly provide the most direct and reliable indication of their anaemia status and hence their ability to cope with H. contortus infection. At a phenotypic level FAMACHA© scores would be expected to be less sensitive than haematocrits, and FECs may provide a less direct and reliable indicator of an animal�s ability to cope with the effects of nematode challenge. Nevertheless, the high genetic correlations between these traits indicate that the use of any of them in an appropriate selective breeding strategy would lead to worthwhile genetic improvements in the others. The use of FEC, either by themselves or in parallel with one of the indicators of anaemia, would have the advantage that they would provide a better indication of the presence of potentially pathogenic burdens of parasite worm genera other than H. contortus should these become a problem. In addition, it is probable that FECs would have an advantage over indicators of anaemia (particularly FAMACHA© scores) when seasonal conditions or farmer decisions lead to the sheep being exposed to lower than optimal levels of nematode challenge. While the high heritability estimates for indicators of anaemia in the Merino Stud flock clearly indicate the presence of good levels of genetic variability under high levels of nematode challenge, the ability to discriminate among progeny groups using this trait could be expected to decline at a faster rate under lower levels of nematode challenge than would be the case for FEC, because FECs are likely to be more sensitive to the presence of relatively low numbers of H. contortus than would the haematocrit of their host. Nevertheless, in South Africa the expense of FEC and haematocrit evaluations, compared to the profitability of stud breeding and the dearth of institutions that can perform the tests make it unlikely that these would be practical. On the other hand, the FAMACHA© system can be easily learnt and applied, and may hold promise for genetic selection if the levels of worm challenge required do not have too severe an effect on the production of the sheep. In commercial farming situations, where speed of genetic progress must be carefully balanced against risk to productivity and survival of the animals involved, such considerations will be of primary importance in decisions relating to the preferred selection traits. Suggested test procedures to select for low anthelmintic requirements in Merino sheep in South Africa On the basis of the results above, the following test procedures are suggested as being the most appropriate, from the point of view of both practicality and maximising genetic progress, for use in flocks belonging to Merino ram breeders and stud owners in South Africa. For useful progress to be made at least 80 ram lambs should be performance tested - preferably all the ram lambs to be retained for breeding purposes or for sale. All should have been grazed together from as soon after birth as practical to ensure a similar experience of nematode challenge. Sire, dam, date of birth and birth rank should be recorded. This requires ewes to have been single-sire mated and lambs to have been tagged at birth. To get reliable genetic estimates, at least 6 ram lambs per sire should be included in the test flock � preferably more. The test period should be at a time of year when adequate natural challenge is likely to be encountered by the animals to be tested. In order to achieve a reasonable level of standardisation, it would be best to treat all animals at the start of the test period with an effective anthelmintic to eliminate any existing nematode infection. In New Zealand the start of the test

58

period is generally weaning (when lambs are around 3 months old), but this may not be appropriate in the South African context.

59

Following the start of the test period all animals should be left untreated until such time as the condition, haematocrit levels or FAMACHA© scores of some animals indicate that treatment is essential. Alternatively a �trigger level� faecal worm egg count could be agreed upon with the flock owner and 10-15 randomly selected animals sampled regularly to determine when the test should be carried out. WormFECTM in New Zealand uses an 800 eggs/g trigger level but a higher count would be necessary where H. contortus is the predominant worm species present. At this stage all animals should have haematocrits or FAMACHA© scores read and faecal samples taken to assess FECs and generic composition. [Note : Haematocrit readings would probably be more reliable than FAMACHA© scores particularly at lower challenge levels, but provided worm challenge is sufficient to get a reasonable spread of FAMACHA© scores, they would be faster, cheaper and (the results suggest) almost as effective as haematocrit readings provided subsequent animal selection is based on properly calculated breeding values.] We recommend that following sampling ALL animals are treated with an effective drench and that they are then left for a second similar test period without treatment prior to being re-sampled as before. For those farmers prepared to invest more effort in order to make greater genetic gains, a similar procedure could be followed with the ewe lambs although it should be noted that considerably greater genetic progress can be made by applying selection pressure to rams than to ewes because of their greater reproductive potential. The pedigree and performance data collected should be used to calculate breeding values for the desired traits which, estimate the �genetic merit� of the animals tested and their parents. Our preference would be to select on the basis of a �breeding index� which combines both an indicator of anaemia (resilience) (i.e. either haematocrit or FAMACHA©) and of resistance (FEC). Taking into account the high genetic correlations between these traits, such an index could be expected to provide a very reliable indicator of genetic merit for the ability to cope with worm challenge as well as for the ability to limit pasture contamination. In the light of the results, which indicate a strong favourable correlation between indicators of resilience and resistance in Merinos infected with H. contortus, we would advise against the general use of a selective drenching approach as a breeding strategy for commercial ram breeders in South Africa. While a selective drenching procedure might be expected to allow somewhat better differentiation between the more resistant/resilient animals at a phenotypic level (because of the greater period of challenge without drench intervention for these animals), at the same time the procedure has the disadvantages of stressing a bigger proportion of the flock (thus risking greater economic losses) and complicating the genetic assessment of other important traits. It should be noted that, although they are using a selective drenching approach to develop an experimental line of sheep in New Zealand, this is mainly because there is an unfavourable genetic association between resistance and resilience in dual purpose sheep infected with Trichostrongylus colubriformis and Trichostrongylus vitrinus under these conditions. There is currently no general recommendation in place for commercial sheep farmers to use the procedure in New Zealand because of the disadvantages discussed above. For many smaller scale farmers in South Africa, who must rely on breeding their own replacements rather than buying them in, the above genetic approaches may not be a practical option at this time. Under these circumstances a phenotypic selection strategy involving the use of FAMACHA© scoring and individual drenching of the most anaemic (least resilient) animals may be a more appropriate option. Although genetic progress could be expected to be comparatively slow, the relatively high heritability estimates for the resilience/resistance traits in sheep exposed to sufficient challenge by Haemonchus suggest that phenotypic selection on the basis of productivity in those animals not requiring drench treatment may be of some value. In most instances, the

60

approach would of necessity have to consist largely of culling (or at least avoiding breeding from) those animals least able to cope with Haemonchus challenge. The availability of simple blood tests to identify the presence of genetic markers for traits of interest will invaluable to such farmers in the future to assist them in identifying the most desirable rams to breed from.

61

Section 7 Dilution of anthelmintic resistant with susceptible

worm strains : Preliminary results in the field

Introduction Once anthelmintic resistance has developed in a given worm strain, withdrawal of the anthelmintic concerned is of little use for obtaining reversion to susceptibility. Even after periods of as long as 10 years, any reversion that may have developed is undone with a few drenches following reintroduction of the compound (Borgsteede and Duyn, 1989; Scott et al., 1995). One option that may seem feasible for obtaining reversion would appear to be to withdraw all susceptible animals from the pasture for up to 2 years, whereafter practically all of the dominant worm species in the Sub-Tropics or Tropics should have died. However, this option is unlikely to give good results in a country such as South Africa, where worm strains on few, if any, of the farms in the main sheep-producing regions do not show some degree of anthelmintic resistance (Van Wyk et al., 1999 and Section 2 above). Hence, even in the unlikely event that this option may be financially feasible and the resistant worm strain is eradicated, resistance is very likely to be re-introduced with the animals that are brought in when the farm is restocked. Thus far only one method seems to have shown promise for obtaining reversion in the field : Van Wyk and Van Schalkwyk (1990) succeeded under semi-laboratory conditions (using small paddocks) to obtain reversion by thoroughly deworming animals and then artificially infecting them with a known susceptible strain of H. contortus. This has been confirmed by Bird et al. (2001) Table 7.1 Reversion under semi-laboratory conditions, by diluting a resistant

worm strain of H. contortus on pasture with a susceptible strain (Van Wyk and Van Schalkwyk, 1990)

Worm strain Susceptibility to albendazolea

Susceptible strain 99.9 %

Strain on pasture :b

Before 29.4 %

After : Spring (September) 96.0 %

After : Spring (October) 99.6 % aEstimated by worm recovery from the trial animals. bBefore and after the attempt at obtaining reversion by dilution. In Table 7.1 it can be seen that when the replacement of the resistant strain was attempted on the Highveld of Gauteng Province in September or October (when there are few infective larvae of H. contortus on pasture), a strain that was previously 29.4 % susceptible to albendazole reverted to susceptibilities of 96.0 % and 99.6 % in 2 different camps (= paddocks). The attempts in summer (on heavily infected pastures) were very erratic, varying from ± 50 % reversion in 2 camps, to 96.1 % in a third (Van Wyk and Van Schalkwyk, 1990). Aim of preliminary trials in the field We are attempting to obtain reversion under practical farming conditions, and have obtained some preliminary results.

62

Materials and methods After having gauged the susceptibility of the strains of H. contortus by means of Faecal Egg Count Reduction (FECR) tests on the 2 farms involved, all the sheep were drenched with combinations of anthelmintics that were most likely to be highly effective. Subsequently, after suitable periods to allow for the residual effects of the anthelmintics that were used, the sheep were all infected before commencement of the Haemonchus season, with infective larvae of the Moredun strain of susceptible H. contortus. When lambs that were born on the 2 farms became infected with the worms that had developed on the pasture after the artificial infections, FECR tests were done on them. In addition, at the end of the worm season a further FECR test was conducted on adult sheep on each farm. From 10-15 sheep were included per trial group in the FECR tests, according to the numbers of animals available. Results Table 7.2 Reversion on farm 1

Worm strain susceptibility (FECRT)a Anthelmintic

Before After (lambs) After (ewes)

Albendazole 0 % 75.1 % 91.6 %

Ivermectin 30.7 % 96.5 % 88.0 %

Rafoxanide 76.7 % 95.8 % 96.9 %

Levamisole 73.8 % 97.1 % 97.9 % aArithmetic means, before and after the attempt at obtaining reversion to susceptibility. Table 7.3 Reversion on farm 2

Worm strain susceptibility (FECRT)a Anthelmintic

Before After (lambs) After (ewes)

Levamisole 36.4 % 98 % 99.7 %

Albendazole 24.9 % 7 %/84 %b 98.2 %

Ivermectin 23.8 % -c 71.0 %

Rafoxanide 58.1 % - 94.3 % aSee Table 2. bArithmetic/Geometric mean. c- Not tested (too few lambs). Discussion It can be seen from Tables 7.2 and 7.3 that both farmers were in trouble when our trials commenced. Their prospects for sustainable worm management were bleak, with the result that the viability of the sheep enterprises were under threat. The results in Tables 7.2 and 7.3 are encouraging, since a considerable improvement was observed in the susceptibility of the worm strains on the 2 farms. However, much more work is required to streamline the application of the method and to establish why some of the results appeared to be erratic, particularly in the case of ivermectin and albendazole. On Farm 2, 2 of the lambs in the albendazole group of 10 had exceptionally high faecal egg counts after drenching, while the rest showed considerable reductions, as can be deduced from the difference between the arithmetic and geometric mean efficacies. A surmise that the 2 lambs concerned may have been drenched incorrectly, is supported by the high percentage of efficacy of albendazole in adult sheep in the second test, at the end of the worm season.

63

A fortunate coincidence that probably greatly improved the chances of success with the present technique for obtaining reversion, is a grazing system that was developed by Kirkman and Moore (1995) at the Nooitgedacht experimental farm in Mpumalanga Province. These workers showed that if all the animals on a farm are concentrated on half the available pasture while the remainder is rested for a year, the quality and amount of herbage improves dramatically. We are using this principle to some extent on the farms where we are testing reversion, with the result that pastures with very low numbers of Haemonchus spp. are available annually - almost ideal for the method for reversion. However, unfortunately this advantage is somewhat watered down in practice, as it is almost impossible after the switch not to have contact with pastures that were previously infected with resistant worms, for example, around the handling pens. In the case of Haemonchus spp. and other haematophagous worm species, the reversion is also greatly assisted by the FAMACHA© system for clinical evaluation of the extent of infection of individual animals (Bath et al., 1996; Malan and Van Wyk, 1992). This system makes it possible to drench only severely affected animals, instead of the entire flock or herd, as is the usual practice. In this way the selection for resistance is reduced to such an extent, that even limited success with reversion should make it possible for worm management to be sustainable for considerable periods of time, but this should be investigated further in the field. One of the most difficult problems we encountered to date was that the strains of H. contortus on the two farms were so resistant to most anthelmintics, that we had to use moxidectin in combination with a variety of the older compounds to reduce the resident worm burdens to minimal levels at the time of artificial infection with the susceptible H. contortus. As a result, the long residual effect of the moxidectin complicated the attempt at replacement. The strains on both farms were highly resistant to ivermectin and, the long �tail� of dwindling residual efficacy of moxidectin is probably compromised. If this is correct, it will lead to a lengthy period during which no susceptible, and only resistant larvae, can develop in the treated sheep. In the case of the strain, of which the results are presented in Table 7.3 the results may have been affected by the following : the farmer unexpectedly (and without prior notice) moved his animals away from the pasture on which the susceptible strain had become established, to pasture that was shared with extraneous sheep, infected with large numbers of a strain of H. contortus of unknown susceptibility. The farmer did not realise the implications of his action, despite our previous efforts at explaining the importance of principles such as maintaining a closed flock, and placing animals that are brought in, under quarantine. It is encouraging that, while we expected much reluctance from the farmers to the principle of infecting their animals artificially with a susceptible worm strain, this has not been the case. As predicted by Van Wyk and Van Schalkwyk (1990), every farmer we approached thus far with the proposal to attempt reversion when sustainable worm management appeared to be on the brink of collapse in the short to medium term, was very keen for us to attempt the reversion, despite the fact that the procedure is still being developed experimentally. The potential of the present dilution technique for obtaining reversion has been criticised because it will probably not be practical to apply to all of the thousands of farms affected by resistance in most of the important sheep producing countries. While the results are not yet such that large-scale application can be contemplated, this criticism needs to be seriously considered before too much time is spent at attempting to improve the technique. The redeeming feature is most probably that farmers are unlikely to consider reversion by such a relatively difficult method unless in dire straits. And such cases will hopefully emerge at a rate that is manageable with the technique if and when consistent results become practical in the field. If such severe anthelmintic resistance becomes more widespread, it may be possible to find ways of making the technique more practical on a large scale.

64

Section 8 Survey of nematophagous fungi from agricultural

environments

Introduction The nematophagous fungus, Duddingtonia flagrans, has been shown to be effective in reducing pasture contamination with cyathostome larvae from horses (Larsen et al., 1995; Fernandez et al., 1997) and Ostertagia ostertagi from calves (Grønvold et al., 1993). In Mexico promising results have been obtained in a trial to assess the effect of orally administered chlamydospores of a local isolate of D. flagrans on Haemonchus contortus larvae from sheep (Mendoza et al., 1998). The resistance of H. contortus to anthelmintics in South Africa necessitates the formulation of alternative strategies for control. Among these is the possibility of isolating a local strain of D. flagrans and testing it for efficacy in reducing pasture contamination with the infective larvae of H. contortus. The aim of this study was to isolate species of nematophagous fungi in order, it was hoped, to obtain a South African strain of D. flagrans, adapted to local environmental conditions, for use in an integrated worm control programme.

Method The modified sprinkling technique of Larsen et al. (1991) was used for the isolation of nematophagous fungi. Soil, faeces from livestock, leaf litter or compost samples were sprinkled in a cross configuration on 2 % water agar containing 0.02 % tetracycline hydrochloride to suppress bacterial growth. The sample was baited with infective nematode larvae. The plates were incubated at 26°C, baited at least twice a week for 3 weeks and examined for signs of nematophagous activity every 2 to 3 days. Conidia, associated with trapped larvae were sketched, measured and in some cases photographed for identification purposes. Conidia growing from trapped larvae on water agar plates were transferred by means of sterile glass inoculation needles to corn meal agar plates to purify the isolate. Conidia and chlamydospores of D. flagrans were prepared for scanning electron microscopy (SEM). One-centimetre blocks of corn meal agar cultures were vapour fixed with osmium tetroxide for twelve hours and air dried for 2 days. The air-dried blocks were mounted on aluminium stubs, gold coated and viewed in a Philips XL 20 Scanning Electron Microscope.

Results Duddingtonia flagrans was isolated from 2 of the samples. Sixty-eight samples out of a total of 384 were positive for other species of nematophagous fungi. The samples were collected from 5 provinces of South Africa and included leaf litter, soil, faeces from livestock and compost. Certain isolates were photographed using light microscopy and scanning electron microscopy.

65

Table 8.1 Nematophagous fungi isolates from agricultural environments Source Localities Number of

samples Nematophagous

fungi Trapping

mechanism Identification

3-D Nets Monacrosporium sp. Arthrobotrys oligospora

Soil Onderstepoort, University of Pretoria (OP/UP), Onderstepoort Veterinary Institute (OVI), Mooikloof, Rust de Winter, Rhenosterpoort, Botanical Gardens

76 10 Positive

Constricting rings Arthrobotrys dactyloides

Constricting rings Arthrobotrys dactyloides Faeces OP/UP, OVI, Roodeplaat, Thabazimbi, Rust de Winter, Mooikloof, Rhenosterpoort, Dullstroom, Hluhluwe, Vaal Island, Bloemfontein, Borakalalo

117 4 Positive 3-D Nets Arthrobotrys conoides ?

Arthrobotrys oligospora 3-D Nets Arthrobotrys oligospora Constricting rings Arthrobotrys dactyloides

Faeces + Soil

OP/UP, Mooikloof, Rust de Winter, OVI, Rhenosterpoort, Dullstroom, Thabazimbi

29 9 Positive

2-D Nets Monacrosporium gephyropagum

Constricting rings Arthrobotrys dactyloides 2-D Nets Monacrosporium

gephyropagum

Leaf litter

Rhenosterpoort, Borakalalo, Botanical Gardens, Thabazimbi, Mountain View

81 34 Positive

3-D Nets Arthrobotrys oligospora Arthrobotrys oviformis

3-D Nets Duddingtonia flagrans Arthrobotrys oligospora

Hyphal loops Arthrobotrys amerospora 2-D Nets Monacrosporium

gephyropagum

Compost Just Nature Organics, Mountain View 71 11 Positive

Constricting rings Arthrobotrys dactyloides Arthrobotrys anchonia

Larval cultures

OP/UP 10 2 Positive 3-D Nets Arthrobotrys oligospora

Totals 384 70 aOP/UP : Faculty of Veterinary Science, University of Pretoria; OVI : Onderstepoort Veterinary Institute.

66

Discussion Eighteen percent of the samples cultured were positive for nematophagous fungi. D. flagrans has for the first time been isolated in Africa. The isolation of a South African stain of D. flagrans provides the opportunity to compare it with strains from elsewhere in the world, for efficacy in the reduction of infective nematode larvae in the environment as well as molecular characterisation. This fungus is to be used as part of an integrated worm control programme in grazing animals.

67

Section 9 Haemonchus in goats farmed under resource-poor conditions in South Africa : Incidence, effect on

haematocrit and treatment according to ocular mucous membrane colour

Introduction Small ruminants play an important socio-economic role within traditional farming systems in many developing countries, including South Africa. However, little is known about the epidemiology of internal parasites and their effects on the production of small ruminants raised under resource-poor conditions. In contrast, Boomker, Horak and others have studied worms in goats raised under commercial farming conditions (Boomker et al., 1989; Horak et al., 1991) and Boomker in indigenous goats on an experimental farm (Boomker et al., 1994). The epidemiology of helminths in sheep raised within the summer-rainfall area of South Africa, where the present study was conducted, has also been well described (Horak, 1978; Horak and Louw, 1977). Indiscriminate and overuse of anthelmintics has led to the development of severe anthelmintic resistance on commercial sheep farms and resistance has also been reported in helminths in sheep raised under resource-poor conditions in South Africa (Van Wyk et al., 1999), but has not yet been reported from goats in South Africa. An emphasis is currently being placed on assisting resource-poor farmers to improve the production of their livestock and in this respect the management of worms is seen to be a researchable constraint. Worm control strategies in small ruminants, appropriate to the resource-poor farmer and sustainable in terms of maintaining anthelmintic efficacy, need to be tested. The present study had a number of objectives, which are outlined as follows : to conduct a longitudinal study of the differential nematode faecal egg counts (FECs) and haematocrits in herds of goats owned by resource-poor farmers; to determine the incidence of Haemonchus spp.; to evaluate the effect of the level of worm infection on the haematocrit; to use the FAMACHA© system to determine which animals require dosing with anthelmintics to prevent mortalities; and to use the FAMACHA© system to ensure more effective dosing to slow down the development of anthelmintic resistance.

Materials and methods 1. Study sites, sampling and diagnostic techniques

Three study sites were selected within the summer-rainfall area of South Africa : Rust de Winter, Gauteng; Impendle (2 different flocks), KwaZulu-Natal; and Kraaipan, North-West Province (Fig. 9.1).

68

Figure 9.1 Study sites for sampling

Table 9.1 contains a summary of the trial dates and frequencies of visits; of sample sizes; breeds of goats; grazing practices, vegetation types and rainfall; and of anthelmintics used for each study site. At Rust de Winter all the weaner and adult goats present at each visit were sampled. For each of the 2 farmers at Impendle (Impendle I and II), a representative sample of the weaner and adult animals was selected (to serve as monitor groups), based on the first animals brought into the crush at the first visit, and the same goats were sampled/scored throughout the trial period. Similarly for Kraaipan, a monitor group was selected from the herd, but the animal numbers started to dwindle and for this reason every 10th goat brought into the crush in May 1999 was added to the group. Faecal samples were collected at each visit from the goats and were processed for nematode FECs, following the method of Van Schalkwyk et al. (1995) and at a sensitivity of 100 eggs per gram of faeces (epg). Strongyloides spp., Nematodirus spp. and Trichuris spp. eggs were counted separately from the other nematode eggs, which are herein referred to as �strongyle� eggs (Order Strongylida - Molin, 1861). Samples were screened for Fasciola eggs by means of the sedimentation method (Van Wyk et al., 1987) performed on samples pooled from 10 goats selected at random from each site at each visit. Faeces remaining after the FECs had been processed were cultured for third-stage nematode larvae (L3) at a temperature of approximately 25°C, until November 1999 when a new room was used at a temperature of approximately 26°C. The L3 were identified using the keys of Van Wyk et al. (1997a) and Dunn (1978). Where possible, at least 50 L3 were identified per culture. The proportions of the strongyle L3 were used to estimate the proportional FECs of the various strongyle worm genera. The goats were bled and their haematocrits determined by the microhaematocrit method. The animals were scored for body condition (Vatta et al., 2000). The efficacies of the anthelmintics used in the trial were assessed towards the end of the trial by means of the faecal egg count reduction (FECR) test, according to the methods of the World Association for the Advancement of Veterinary Parasitology (WAAVP) (Coles et al., 1992) and of Presidente (1985).

2. Scoring for level of anaemia and evaluation of the FAMACHA© clinical assay At the scheduled visits, one of the authors (A F Vatta) or one of the assistants on the project scored each animal for level of anaemia using the FAMACHA© card. It was ensured that each assistant for whom scores were recorded had been adequately trained in the method. Excepting for the few visits that Dr Vatta could not undertake, the scoring was always performed under his direct supervision. Occasionally, monitoring was done in-between

69

scheduled visits by the farmer at Impendle I, and by the animal health technicians assisting (AHTs) with the project at Kraaipan, but these scores were not included in any of the analyses discussed below. Only the animals that were considered to be pale, i.e. FAMACHA© categories 4 and 5, were treated with an anthelmintic. At times, animals scored as category 3 were erroneously treated by the AHTs at Kraaipan and the farmer at Impendle II initially misunderstood the aim of the trial and treated all his goats sometime between 24 November and 22 December 1998. Between 23 November and 21 December 1999, 10 goats were apparently treated by the shepherd of the farmer at Impendle I, but the animal identifications were not recorded. On 7 January 1999 one of the goats at Kraaipan showed signs of Oestrus ovis infection, indicated by a mucoid nasal discharge. The animal showed difficulty in breathing through the nose and was treated with rafoxanide [Nasalcur�, Hoechst Roussel Vet (now Intervet), 7.5 mg/kg]. In January 1999, 22 of the goats at Impendle I and 25 of the goats at Impendle II were treated with niclosamide [Ex-a-lint�, Hoechst Roussel Vet (now Intervet), 50 mg/kg] for cestodes. Two data sets for 1998/1999 and 1999/2000 were created from the FAMACHA© scores and haematocrit values obtained from all the goats at Rust de Winter and the representative sample sets at Impendle and Kraaipan. Two-way frequency tables of haematocrit by FAMACHA© were drawn up, with FAMACHA© values 4 and 5 (or 3, 4 and 5) considered positive for anaemic animals and FAMACHA© values 1, 2 and 3 (or 1 and 2) considered negative test results, respectively (Table 9.2). Haematocrit was used as the �gold standard� by which anaemia was measured and two cut-off values for anaemia were assigned, namely less than 18 % and less than 19 %. In establishing the properties of a test, cut-off values are assigned to define the level of a test result that is needed to confirm or reject a diagnosis, in this case a diagnosis of anaemia (Smith, 1995, p.42). Sensitivity, specificity, predictive value of a negative, and predictive value of a positive were calculated for the data. Smith (1995, p. 51) defines sensitivity as the proportion of infected or diseased individuals with a positive test, or in the case of the FAMACHA© clinical assay, the proportion of anaemic animals correctly identified as anaemic. Test specificity, on the other hand, is defined as the proportion of disease-free individuals that test negative, or the proportion of non-anaemic animals that are correctly categorised. In the case of the FAMACHA© method, predictive value of a negative is the probability that an animal is not anaemic when the test result is negative for anaemia and vice versa for the predictive value of a positive. The sensitivity and specificity were tested statistically by means of Fisher�s Exact Test for a two-by-two contingency table. The authors have chosen to maximise the sensitivity and specificity of the FAMACHA© method in goats when the average of the sensitivity and specificity attains its highest value. This was calculated as follows : (sensitivity + specificity)/2 (1) In order to determine the percentage of animals treated in each case (FAMACHA© cut-off of 4 as opposed to FAMACHA© cut-off of 3), the following calculation was also applied to the data : (true positive + false positive)/total number of animals x 100 (2)

70

Table 9.1 Summary for study sites including trial dates and frequencies of visits; sample sizes; breeds of goats; grazing practices, vegetation types and rainfall; and anthelmintics used

Study site Trial dates Frequency of visits

Approximate numbers of

animals at each visit

Mean number of animals sampled/

scored (range)

Breed Grazing practice

Vegetation (Acocks,

1975)

Rainfallb Anthelmintic drenched per os (dosage)

Rust de Winter

Sep 1998 - Apr 2000

Fortnightly ±30 31

(18-47) Boer goat crossbreed

Private Mixed bushveld

610 (Rust de Winter, 10 km)

Levamisole (7.5 mg/kg)c

Kraaipan Oct 1998 - Apr 2000 Monthlya ±60 16

(10-23) Indigenous crossbreed

Communal Sourish mixed bushveld

539 (Mmabatho, 60 km)

Levamisole (7.5 mg/kg)c

Impendle I ±60 18 (9-22)

Impendle II

Nov 1998 - Apr 2000

Monthly ±35

19 (14-24)

Indigenous Zulu crossbreed

Communal

Highland sourveld and Döhne sourveld

993 (Donnybrook, 35 km)

Ivermectin (0.2 mg/kg)d

aExcepting for two visits over three months at start of trial. bLong-term average annual rainfall in mm (weather station, approximate kilometres in a direct line from study site). Source: South African Weather Bureau. cPredominantly Tramisol� (Hoechst Roussel Vet, now Intervet); on a few occasions initially, Levisol� (Bayer). dIvomec� tablets for sheep (Logos Agvet) used extra-labelly in goats: ½ tablet of 10mg for goats less than 25 kg and 1 tablet for goats more than 25 kg. Table 9.2 Two-by-two contingency table of haematocrit-by-FAMACHA© score with haematocrit (Ht) cut-off of 18 % (or 19 %) and

FAMACHA© scores 4 and 5 (or 3, 4 and 5) considered positive test results ANAEMIA

Present Ht <18 % (Ht <19 %)

Absent Ht ≥18 % (Ht ≥19 %)

Negative 1,2,3 (1,2)

False negative (FN) True negative (TN)

FAM

AC

HA

©

scor

e

Positive 4,5 (3,4,5)

True positive (TP) False positive (FP)

Sensitivity = TP/(TP + FN) x 100 Predictive value of a negative = TN/(FN + TN) x 100 Specificity = TN/(FP + TN) x 100 Predictive value of a positive = TP/(TP + FP) x 100

71

Results 1. Results of faecal nematode egg counts and haematocrits

Figs. 9.2-9.5 illustrate the mean FECs and haematocrits for the study sites. Complete results for the L3 cultures were not obtained during October 1998, the beginning of November 1998 and late July 1999 owing to problems in the laboratory. In these cases, the averages of the proportions for Haemonchus spp. and for the other nematode genera for the visit dates immediately prior to and following the dates of missing data were applied to the FECs. These results are indicated by dotted lines in the figures. Since Haemonchus predominated in many of the cultures, the graphs were drawn to reflect the mean Haemonchus FECs and the mean total FECs for the other genera. Mean Strongyloides FECs were less than 100 epg for all sites and all visits except for 31 December 1998 and 10 February 2000 when mean counts of 103 epg and 116 epg, respectively, were recorded in the goats of Rust de Winter. Trichuris and Nematodirus FECs were negligible, maximum individual counts never exceeding 200 epg. In summary, L3 of Haemonchus spp., Teladorsagia/Trichostrongylus spp., Oesophagos-tomum spp. and Strongyloides spp. were identified in the faecal cultures from each of the trial sites. Periods of heavier worm infection during which time Haemonchus was the predominant species occurred from December/January to March at Rust de Winter; from December to March/April at Impendle; and from November/December to February or April at Kraaipan. During these periods the mean haematocrit values decreased. At Rust de Winter, FECs started to rise during late August 1999 (Fig. 9.2). Mean FECs for the worm genera other than Haemonchus reached a level of over 1000 epg in early April 2000. In addition to the above-mentioned genera, L3 resembling those of Gaigeria or Bunostomum spp. were also found in the faecal cultures from Impendle II. These L3 occurred with a low frequency in November and December 1998 and in January, May, August and September 1999. The mean FECs at Kraaipan remained below 1200 epg throughout the period of the investigation. The drops in haematocrit during the periods of heavier worm infection are less noticeable than those recorded for the sites discussed above. Tick burdens were not deemed heavy enough to be important contributors to anaemia in any of the study areas. Fasciola eggs were found in faecal samples of goats at Rust de Winter and Impendle II in low numbers (range : 2-4 epg, when positive) and incidence (Vatta et al., unpublished data). Periods of poorer body condition, as an indication of nutritional status, occurred during the dry season, from August to early December 1999 in Rust de Winter and from June to September 1999 in Impendle. These periods did not clearly correspond to any periods of lower haematocrit. The goats at Kraaipan did not show any clear seasonal trend in body condition. The results of the FECR tests indicated that anthelmintic resistance was not found in any of the groups tested, except for the goats at Rust de Winter where resistance to levamisole was detected by the WAAVP method of calculation of FECR (Vatta et al., 2001).

2. Evaluation of the FAMACHA© clinical assay (Tables 9.3, 9.4.1 and 9.4.2)

Treatment of goats from November to April (period of heavier Haemonchus infection) is summarised in Table 9.3.

72

Figure 9.2 Strongyle faecal egg counts (FECs) and haematocrits of goats at Rust de Winter

0

500

1000

1500

2000

2500

3000

3500

4000

45002

4/0

9/9

8

22

/10

/98

19

/11

/98

17

/12

/98

14

/01

/99

11

/02

/99

11

/03

/99

08

/04

/99

06

/05

/99

03

/06

/99

01

/07

/99

29

/07

/99

26

/08

/99

23

/09

/99

21

/10

/99

18

/11

/99

15

/12

/99

13

/01

/00

10

/02

/00

08

/03

/00

06

/04

/00

--- data not available

FEC

(eg

gs p

er g

ram

)

0

5

10

15

20

25

30

35

40

45

50

Hae

mat

ocri

t (%

)

Mean Haemonchus egg counts Mean egg counts for other genera Mean haematocrits

Midsummer

Midsummer

73

Figure 9.3 Strongyle faecal egg counts (FECs) and haematocrits of goats at Impendle I

0

500

1000

1500

2000

2500

3000

3500

4000

4500

24

/11

/98

22

/12

/98

19

/01

/99

16

/02

/99

16

/03

/99

13

/04

/99

11

/05

/99

08

/06

/99

06

/07

/99

03

/08

/99

31

/08

/99

28

/09

/99

26

/10

/99

23

/11

/99

21

/12

/99

18

/01

/00

15

/02

/00

14

/03

/00

11

/04

/00

FEC

(eg

gs p

er g

ram

)

0

5

10

15

20

25

30

35

40

45

50

Hae

mat

ocri

t (%

)

Mean Haemonchus egg counts Mean egg counts for other genera Mean haematocrits

Midsummer

Midsummer

74

Figure 9.4 Strongyle faecal egg counts (FECs) and haematocrits of goats at Impendle II

0

500

1000

1500

2000

2500

3000

3500

4000

4500

24

/11

/98

22

/12

/98

19

/01

/99

16

/02

/99

16

/03

/99

13

/04

/99

11

/05

/99

08

/06

/99

06

/07

/99

03

/08

/99

31

/08

/99

28

/09

/99

26

/10

/99

23

/11

/99

21

/12

/99

18

/01

/00

15

/02

/00

14

/03

/00

11

/04

/00

FEC

(eg

gs p

er g

ram

)

0

5

10

15

20

25

30

35

40

45

50

Hae

mat

ocri

t (%

)

Mean Haemonchus egg counts Mean egg counts for other genera Mean haematocrits

Allgoats treated

Midsummer

Midsummer

75

Figure 9.5 Strongyle faecal egg counts (FECs) and haematocrits of goats at Kraaipan

0

500

1000

1500

2000

2500

3000

3500

4000

4500

15

/10

/98

26

/11

/98

07

/01

/99

02

/02

/99

02

/03

/99

30

/03

/99

29

/04

/99

25

/05

/99

22

/06

/99

20

/07

/99

17

/08

/99

14

/09

/99

12

/10

/99

09

/11

/99

07

/12

/99

04

/01

/00

01

/02

/00

29

/02

/00

28

/03

/00

26

/04

/00

--- data not available

FEC

(eg

gs p

er g

ram

)

0

5

10

15

20

25

30

35

40

45

50

Hae

mat

ocri

t (%

)

Mean Haemonchus egg counts Mean egg counts for other genera Mean haematocrits

Midsummer

Midsummer

76

Table 9.3 Number of goats treated from November to April (Haemonchus season) and May to October

Location Total examined F© 4 & 5 treated (percentage)a

F© 3, 4 & 5 treated (percentage)b

Rust de Winter Nov 98 Apr 99 434 9.4 % 47.2 % Nov 99 Apr 00 372 8.1 % 46.8 % May Oct 99 398 6.3 % 34.4 %

Impendle I Nov 98 Apr 99 129 12.4 % 52.7 % Nov 99 Apr 00 83 19.3 % 55.4 % May Oct 99 141 14.9 % 58.2 %

Impendle II Nov 98 Apr 99 134 8.2 % 48.5 % Nov 99 Apr 00 104 16.3 % 62.5 % May Oct 99 134 2.2 % 35.1 %

Kraaipan Nov 98 Apr 99 121 11.6 % 43.0 % Nov 99 Apr 00 91 12.1 % 57.1 % May Oct 99 105 2.9 % 35.2 %

aFAMACHA© values 4 and 5 treated. bFAMACHA© values 3, 4 and 5 treated. Theoretical, since only animals scored as 4 or 5 were treated (see text).

Both the percentage of animals treated (FAMACHA© values 4 and 5) and the theoretical percentage of animals that would have been treated, had goats scored as FAMACHA© values 3, 4 and 5 been treated, are given in Table 9.3. Except for Impendle I, relatively more goats were treated during the Haemonchus seasons (November to April) than in-between (May to October).

The data used to draw up the two-way frequency tables to calculate the sensitivities and specificities included paired values of haematocrit and FAMACHA© score for the periods November to April of 1998/1999 and 1999/2000. A total of 787 pairs of haematocrit and FAMACHA© values were included for the former and 648 for the latter. The sensitivities, specificities and predictive values for the 2 different levels of positive FAMACHA© scores (4 and 5, and 3, 4 and 5, respectively) and for the 2 different haematocrit cut-off values for anaemia (less than 18 % and less than 19 %, respectively) are given in Table 9.4.1. The results are for the 2 summer seasons (1998/1999 and 1999/2000).

77

Table 9.4.1 Comparison of results for application of the FAMACHA© system in goats during 1998/1999 and 1999/2000

Predictive value (%) Sensitivity (%) Specificity (%) Negative Positive

98/99 99/00 98/99 99/00 98/99 99/00 98/99 99/00 FAMACHA© values 4 and 5 considered positive test results

Ht cut-off <18%a 31.1 35.6 91.2 90.2 95.6 94.9 17.7 21.3 Ht cut-off <19% 23.0 28.4 91.3 90.4 91.9 91.6 21.5 25.3

FAMACHA© values 3, 4 and 5 considered positive test results Ht cut-off <18% 80.0 86.7 54.3 50.7 97.8 98.1 9.6 11.6 Ht cut-off <19% 75.7 85.1 55.3 52.0 95.6 96.8 14.9 17.0

aHaematocrit cut-off value used. The results for the application of equations (1) and (2) and for Fisher�s Exact Test are recorded in Table 9.4.2. Table 9.4.2 Comparison of results for application of the FAMACHA© system in goats

during 1998/1999 and 1999/2000 (continued)

(Sensitivity + specificity)/2a Goats treated (%)b P values for Fisher’s

Exact Test

98/99 99/00 98/99 99/00 98/99 99/00 FAMACHA© values 4 and 5 considered positive test results

Ht cut-off <18%a 61.2 62.9 10.0 11.6 4.5x10-5 9.6x10-6 Ht cut-off <19% 57.2 59.4 10.0 11.6 7.1x10-10 5.3x10-5

FAMACHA© values 3, 4 and 5 considered positive test results Ht cut-off <18% 67.2 68.7 47.6 51.9 8.6x10-10 6.1x10-7 Ht cut-off <19% 65.5 68.6 47.6 51.9 3.2x10-10 3.8x10-9

aEquation 1 (see text). bEquation 2 (see text).

Discussion 1. General

In common with the results of Boomker et al. (1994) in goats at Roedtan (in the Northern Province, approximately 80 km in a direct line from Rust de Winter), this study has demonstrated a seasonal distribution of Haemonchus spp. with the highest level of infection in the summer months. Similarly, Horak (1978) and Horak and Louw (1977) reported greatest numbers of Haemonchus contortus in sheep from January to May or June on dryland and irrigated pastures, respectively. Tick burdens, worm infections and inadequate nutrition were considered as differential diagnoses of conditions causing anaemia. Of these, the most probable and important of the possible causes of anaemia in the animals studied was Haemonchus spp.

An increase in worm infection in spring is known as a �spring rise� (Gordon, 1973) and is thought to arise from the resumption of development of larvae retarded in the fourth-stage during the cooler months of the year (Michel, 1974). Hence, the rise in FECs during late August 1999 at Rust de Winter is probably to be attributed to this phenomenon, since no rain fell during June to August 1999 (South African Weather Bureau). While Horak and Louw (1977) and Horak (1978) reported retarded larval development in sheep, Boomker et al. (1994) suggest that arrested development is not significant in indigenous goats at Roedtan. On the other hand, lactating animals show a periparturient relaxation of resistance (PPRR) which also manifests as a rise in worm infection (Gordon, 1973). Since many of the female goats at Rust de Winter kidded in August and September 1999, the rise in FECs is perhaps also partially attributable to this phenomenon in the female animals. The mean haematocrits

78

are seen to decrease concurrently with the increase in FECs, which would seem to agree with the findings by Dorny et al. (1995). These workers reported a significant drop in haematocrit and a significant rise in FECs (predominantly Haemonchus contortus infection) during the periparturient period in 2 sheep flocks and 1 of 2 goat herds kept under the traditional husbandry systems of peninsular Malaysia. In the second goat herd the changes were similar, though not significant. In order to confirm the identity of the L3 resembling Gaigeria or Bunostomum, animals from the site would need to be necropsied for worm recovery or sufficient L3 would need to be obtained to infect an experimental animal on-station for subsequent necropsy and worm recovery.

At Kraaipan, nematodes assumed greater importance only during the periods of higher FECs from November/December to February or April, when Haemonchus predominated.

2. Evaluation of the FAMACHA© clinical assay

The FAMACHA© assay is a clinical estimate of anaemia and results for the accuracy of estimation of anaemia using the system in sheep, as described in Section 4, indicates that there are large differences in the basic ability of different persons to apply the FAMACHA© method. This should be borne in mind, since the vast majority of the results with goats of resource-limited farmers reflects the scores of only Dr Vatta, with some imput by others trained by him. However, there is some agreement between periods of heavier worm infection, lower haematocrits and paler mucous membranes scored according to the FAMACHA© method (Table 9.3) with more animals being treated during the periods of heavier worm infection (November to April) than during the winter period (May to October). Fewer of the goats at Impendle I were treated during the periods of heavier Haemonchus infection than during the other months. It is possible that the number of treatments given at scheduled visits was reduced because of drenching by the farmer or his goatherd in-between the scheduled visits. Treatment by animal health technicians of animals at Kraaipan in-between scheduled visits may have similarly reduced the number of treatments given at the scheduled visits. A cut-off haematocrit value for anaemia of less than 18 % was assigned when the two-way frequency table was drawn up initially in order to evaluate the FAMACHA© method, 17 % being the highest value subjectively chosen for FAMACHA© category 4 in sheep (Van Wyk, 2000). For the period 1998/1999, the sensitivity of the FAMACHA© system to identify goats that are anaemic (i.e. with a haematocrit less than 18 %) when only those animals falling into FAMACHA© categories 4 and 5 are treated, was poor at 31.1 %. The specificity of the method was, however, good at 91.2 %. A greater sensitivity of the FAMACHA© method is preferred since the consequences of not treating an anaemic animal (possible mortality) are more severe than treating an animal that did not require treatment. It was then hypothesised that considering FAMACHA© categories 3, 4 and 5 as anaemic may render a better sensitivity. The sensitivity at a haematocrit cut-off of less than 18 % increased from 31.1 % to 80.0 %. However, the specificity decreased from 91.2 % to 54.3 %. Schalm�s Veterinary Hematology lists a normal haematocrit range for goats as 19-38 % (Jain, 1986). Using a cut-off value of 19 % and where FAMACHA© categories 4 and 5 are considered anaemic, the sensitivity and specificity for the 1998/1999 data are 23.0 % and 91.3 %, respectively (Table 9.4.1). These values change to 75.7 % and 55.3 %, respectively, when categories 3, 4 and 5 are considered anaemic. Hence, as is to be expected for any diagnostic test (Smith, 1995, pp. 39-41), the sensitivity increases while the specificity decreases when the FAMACHA© cut-off changes from 4 to 3. Where FAMACHA©

79

categories 4 and 5 are considered anaemic, Fisher�s Exact Test shows that the sensitivity of the FAMACHA© system to detect animals with a haematocrit less than 19 % is highly significant but not as significant as with a haematocrit less than 18 %. This may indicate that the subjectively assigned range for categories 4 and 5 (less than 18 %, Van Wyk, 2000) may be accurate. However, the sensitivities for a FAMACHA© cut-off of 4 do not compare favourably with those for a cut-off of 3. Moreover, for the latter sensitivities, Fisher�s Test also gives more highly significant P values. Classifying animals as anaemic in categories 3, 4 and 5 and as non-anaemic in categories 1 and 2, the sensitivity and specificity (equation 1, Table 9.4.2) are maximized for the data of both 1998/1999 and 1999/2000. Unless later results differ, the cut-off for anaemia in the goat should be taken as a �3�. Recommendations that resource-poor farmers are receiving reflect the convention of commercial agriculture, i.e. to treat all small ruminants for worms every time that this is deemed necessary for any one of them (personal observation, 1998/1999). Horak et al. (1976) recommended 4 to 5 strategic anthelmintic treatments for roundworms per year. Moreover, to achieve maximum production in lambs during their first year, these authors recommended that regular, short-interval treatments be applied, presumably even at 4-weekly intervals, which was the regimen tested in the study. However, the emergence of anthelmintic resistance and investigations into its subsequent causes have shown that high frequency of anthelmintic treatment of all the animals in a herd within or close to the prepatent period of the worm in question increases the selection pressure for the development of anthelmintic resistance (Jackson, 1993). Moreover, Van Wyk (2001) has suggested that it is the proportion of the worms in refugia compared with that within the animals at the time of drenching that determines the degree of selection for resistance. It is critical that resource-poor farmers do not follow the same path of those commercial farmers who have selected for severe anthelmintic resistance (Van Wyk et al., 1999). The question that arises, then, is whether the application of the FAMACHA© system represents a potential improvement on the current recommendation to treat the whole herd at any one time, irrespective of the number of animals that require treatment. Based on the fact that only between 47.6 % and 51.9 % of individual treatments are carried out if only goats categorised as 3, 4 or 5 are dewormed (equation 2, Table 9.4.2), the use of the FAMACHA© method would reduce the selection pressure for anthelmintic resistance considerably because a large proportion of the animals would be left untreated. In addition, the untreated animals would be able to deposit the eggs of anthelmintic-susceptible worms on the pasture, which would maintain a reservoir of susceptible larvae in refugia, and should slow down the development of anthelmintic resistance (Besier, 1997). Resistance was detected at Rust de Winter, but an FECR test was not carried out at the start of the trial. Hence it is not possible to determine whether the resistance status of the flock changed over the trial period and it is unknown whether the treatment of animals according to the FAMACHA© system may or may not have affected the development of the resistance. Although the origin of the goats at Rust de Winter was not confirmed, it seems that the animals were brought on to the farm in 1989 when the farmer settled there. It is possible that the animals were infected with resistant worm strains on commercial farms before being purchased at local auctions by the resource-poor farmer. During the trial period the farmer also moved some of the goats between the farm and the farmer�s usual place of residence, a township where communal grazing of animals is practised. Although unlikely, the animals may have been infected with resistant worm strains on the communal grazing. On the other hand, the farm was purchased from a commercial farmer (for resettlement of developing farmers) and the resistant worms may have originated from the commercial farming enterprise.

80

It cannot be recommended that the FAMACHA© clinical assay be used as the only indicator of worm prevalence in a worm control strategy in any of the resource-poor areas studied. Rather, it should be recommended as part of an integrated approach to worm control. Peaks in FECs were seen in the goats at Rust de Winter in February 1999 and late January/early February 2000, in February 1999 and 2000 at Impendle, and in March 1999 and February 2000 at Kraaipan. A tactical anthelmintic treatment of all the animals during late January may be warranted in addition to the use of the FAMACHA© system. A study by Connor et al. (1990) in southern Tanzania showed that weaner goats had significant liveweight gains following a single anthelmintic treatment administered after the rainy season. However, a concern with this approach is that it may select strongly for resistant parasites since at the end of the dry period most of the worms are to be found within the animals and not in refugia (Van Wyk, 2001). A tactical treatment before the peak in worm infection as suggested here may be a more sustainable approach. Supplementation of the animals with a protein and energy source particularly during the times of lower body condition found in this investigation is recommended, since poor nutrition is one of the most important problems in animals within resource-poor areas. Such supplementation would probably also assist the animals partially to overcome the detrimental effects of worms during the winter period. Supplementation of indigenous Tuli cattle in Zimbabwe with cottonseed meal between July and October (i.e. towards the end of the dry period) improved the liveweight gains of the animals above anthelmintic-treated and untreated controls in the presence of subclinical gastrointestinal nematode parasitism (Magaya et al., 2000). Other worm control strategies that may form part of an integrated approach are considered in Section 10.

Conclusion The FAMACHA© assay represents a potential revolution in internal parasite management (J W Hansen, personal communication, 1998) and its validation for goats for use by resource-poor farmers, is of particular relevance for South Africa and indeed for the tropics and sub-tropics of sub-Saharan Africa and elsewhere. In many of these regions haemonchosis represents a major disease constraint on increasing production in small ruminants. Where appropriate, the method should be taught as part of an integrated approach to worm control within participatory rural extension programmes and further testing of the FAMACHA© clinical assay should also be carried out in other goat farming systems.

81

10. Conclusions and recommendations : Tools for

integrated worm management for sheep and goats This chapter has been compiled from the global literature on this subject, as well as the results of the present project. Current useful worm management principles The prerequisite for the development and implementation of successful sustainable parasite management programmes is to have an epidemiological knowledge of the parasites present on a specific farm and even of those in specific camps. Knowledge can be obtained from : # Publications on trials performed in the area or specific climatic zone. # The local veterinarians who have obtained knowledge through experience. # Consultation with veterinary helminthologists and other experts. # The farmer�s own experience � it is important to document all deaths related to parasitism

and the circumstances that may have played a role (such as date, rainfall, temperature and nutrition).

It should be emphasized that each farm has a unique identity regarding its parasite composition. To name but a few, factors that can play a role are : introduction of animals with resistant worm strains, use of anthelmintics, pasture management, irrigated pastures, wet areas, high rainfall and type of animal farmed. We are currently faced with the major predicament that the cornerstone of parasite control during the past 30-40 years (i.e. chemical control with anthelmintics) appears to be at the point of collapse, thus endangering the viability of livestock farming. Resistance of wireworm strains to all groups of drugs has been well documented, and selection for these strains is taking place at an alarming rate, and has already led to some commercial sheep farmers abandoning farming with sheep altogether. Before embarking on or before continuing with a strategy which only uses anthelmintics in controlling parasites the following issues should be taken into account : nutritional status of the animal, basic management requirements, breeding for suitable animals, monitoring systems, optimal anthelmintic use, proper planning, improving the efficacy of the remedy, promising systems of management being developed, and integration. Nutritional status of the animal Parasitism is aggravated by undernutrition or the lack of specific essential dietary components. Malnutrition in livestock most commonly comprises undernutrition, or lack of specific dietary components (such as copper, cobalt and phosphorus). Optimal nutrition reduces production losses and mortality rates from worm parasites in livestock. Basic management requirements $ Identify the groups most at risk

The most susceptible animals are lambs/weanlings and pregnant/lactating ewes. While the former are susceptible because they cannot mount an effective immune response to infection, in the latter there is a temporary suppression of immunity. These groups must get

82

special attention, as well as pregnant ewes at the stage when implantation of embryos takes place.

$ Separation of groups Since different classes of animals vary in their susceptibility to worm infection and its effects, they should be separated into groups which are grazed, treated and managed as distinct entities. If these distinctions are not made, one may be forced to treat the flock or herd according to the most susceptible group. The solution is to manage and treat the most susceptible groups more intensively within a mixed flock. This applies especially in resource-poor farming where not only animals of different ages but also flocks belonging to different owners are grazed together.

$ Separation of pastures Unless pastures can be divided by fencing or herding, all sheep and goats will be exposed to a similar challenge, regardless of whether they are susceptible or resistant to infection and its effects. This will complicate any differentiation in management and treatment. Electric fences can be used as temporary pasturage dividers, particularly on intensive farming systems.

$ Resting of pastures and strategic movement of flocks If pastures can be separated, it is then possible to rest them, which has decided advantages to pasture management and improvement. It also allows the pastures to be rested long enough to have a significant effect on the survival of worm larvae (which die after some time if not ingested by the host) and therefore the infection rate of the flock. Although the time needed for effective resting of pastures will vary with the climate, weather and worm species, a useful rule of thumb for effective resting is at least 3 months. The longer the rest, the better it is for worm management. The aim should be to create "safe" (not "worm-free") pastures. By planning changes in camps or paddocks, stock will be subject to lower challenges and need fewer chemical treatments. The "50:50" system, where half the farm is rested for an entire growing season, is particularly effective in this respect. When safe pastures have been created, it is crucial for reducing development of anthelmintic resistance not to drench all of the animals when they are moved to the rested pasture. As many animals as possible (with a minimum of approximately 15 % of the flock/herd), that are not suffering from heavy worm burdens should be left undrenched at the time. The FAMACHA© system is ideal for identifying the animals that are coping well with infection with haematophagous worm species, while other clinical evaluation system such as condition scoring may prove to be useful for non-bloodsucking worm species.

$ Movement of sheep When sheep are moved from one farm to another they should be treated with an effective remedy and kraaled overnight in a suitable area. Thereafter they should be placed on infected pasture, so that any resistant worms that may have survived the treatment are genetically swamped by the worms� free-living stages in refugia.

$ Alternation with other species Sheep and goats share the same worm species and cannot be used in a system of effective alternation. However, other species such as cattle, horses and ostriches are generally not susceptible to the worms of sheep and goats. If they are used to graze pastures before or after sheep or goats, they act as "vacuum cleaners" to the pasture, as they ingest many larvae and prevent later infection of the sheep and goats. The other advantage is that the pasture can still be utilised in its growing season, which prevents the grass from becoming senescent and optimises its usefulness. This aids in maintaining the profitability of the farm.

83

$ Avoiding concentrations of infective larvae Where sheep or goats have to be penned for lengthy periods (usually at night, to combat theft or predation) there can be a fatal build-up of larvae on the grass growing in the pens. Sheep become hungry overnight and will eat any grass present. In consequence they will ingest massive numbers of larvae. It is therefore necessary to remove all grass from such pens. Water points (e.g. troughs, windmills) should not be allowed to leak, as this encourages the growth of grass and the moisture and shade prolongs the life expectancy of the worm larvae. Since this is where sheep concentrate, the area can become lethally contaminated by nematode larvae. Leaking water troughs also enhance the survival of other parasites such as coccidia and the snails that transmit immature liver and conical flukes.

$ Disturbing the survival of larvae on pasture • Use of annual pastures (e.g. wheat) and making use of stubble lands, can temporarily

stop transmission of parasites between livestock. • Making hay. This minimises helminth infections in livestock. • Tilling, ploughing, replanting. Disturbing the soil reduces larval survival rates. Tilling of fields will reduce the numbers

of worm larvae on pasture. This may have application where camps are alternatively used for growing of crops and grazing of animals on fallow lands.

• Setting fire to grazing. Fire in dry pastures can reduce larval and egg survival, but its effects on the pasture both

in terms of fodder production and its long term viability have to take precedence. Burning as part of veld management will help parasite control as an added bonus. However, larvae harboured underground (i.e. in soil or deep in the mat of the grass) will not be affected by burning. Consult an expert for burning at the correct time.

• Through zero grazing, where feed is brought to the animals in feedlots, excessive contamination can be prevented. This will help control both coccidiosis and Strongyloides, given that the pens are not muddy and the animals are prevented from defaecating in the feeding troughs.

$ Avoiding muddy pens

This applies particularly in resource-poor agricultural areas where the transmission of infective larvae of Gaigeria, Bunostomum and Strongyloides through the skin may be enhanced.

$ Fence off moist areas Areas particularly prone to high moisture and therefore the survival of worm larvae, such as streams and marshes, should be fenced off to reduce the challenge to the flock. Freshwater snails are the intermediate hosts for liver and conical fluke, and fencing off these areas will prevent animals from becoming infected with these parasites.

$ Buying in rams from veld ram clubs If rams are bought from veld ram clubs or group breeding schemes, they may carry worms originating from multiple farms and having varying degrees of resistance. Treat these rams on arrival with a highly effective remedy (e.g. moxidectin), isolate them, and do faecal egg counts after 10-14 days to ensure that they are practically clean of infection. Thereafter place them on an infected pasture, as suggested above.

84

$ Quarantine Animals brought into the flock or herd should be treated with a highly effective remedy (e.g. moxidectin), then do faecal egg counts and ensure that they are grazed on infected pasture.

$ Making use of hired grazing Hired grazing, provided it has not been grazed by sheep or goats for an entire year or at least a worm season, is a practical way of resting the farm�s pastures for a similar period. This will reduce the need for treatment and can assist with the reintroduction of susceptible helminth strains, on condition that the animals are not drenched when they are moved to the rested pasture.

$ Manure management Where this is possible, periodic removal of animals from pasture (for example, by placing them in pens at night) will cause a considerable proportion of the worm eggs to be deposited away from the pasture. In this way the numbers of free-living worm eggs and larvae on pasture will be considerably reduced, in relation to the time spent in the pens. Nevertheless, be on the lookout for worm species such as hookworm (Gaigeria pachyscelis or Bunostomum trigonocephalum) and Strongyloides papillosus that can penetrate the skin and lead to losses when the manure in the pens is thoroughly wet by rain.

Breeding for suitable animals $ Selection for resistance

Resistance (the acquired or innate ability to prevent or minimise infection by parasites) is heritable and can be selected for, by measuring the faecal egg counts (FECs) and using only those sheep with the lowest FECs for breeding. For practical and economical reasons, this is usually only done for rams. Some successful breeding programmes have been undertaken particularly in Australia and New Zealand, but they require good organisation and meticulous record keeping.

$ Selection for resilience (FAMACHA© system)

Resilience (the ability to withstand the effects of infection and produce satisfactorily in spite of it) is also heritable. At present, only one proven approach exists, use of haematocrit determination and the FAMACHA© system which can be used only where wireworm is the major parasite. By treating according to clinical anaemia (an indicator of poor resilience), only those sheep unable to cope with wireworm are treated - this reduces selection pressure for anthelmintic resistance and at the same time allows the farmer to cull the non-copers (animals not able to withstand worm infection). In the long term the farmer is thus able to breed an animal better adapted to the environment. It is also possible to select rams by a system of allocation of indices (Section 6). This is currently under further investigation and will probably require measuring individual ram haematocrits to make the measurement more accurate, but results of a single investigation using the FAMACHA© system were most encouraging.

$ Weighing of ewes This may be another way of identifying the least worm-competent animals for culling, and the most resilient for selection. As stated above, condition scoring is being investigated as an alternative to weighing.

$ Dag scoring Dag score in hyperallergic animals is an indicator of excessive reaction to helminths, which is not desirable because of its association with blowfly attack.

85

Monitoring systems $ Faecal egg counts (FECs)

Regular (monthly or every 2 months) monitoring of FECs from 10 animals per group or flock will help to indicate when dosing is really needed, and equally important, when it can be delayed or even omitted. A bulk (composite) FEC is certainly cheaper than dosing the whole flock unnecessarily.

$ Faecal egg count reduction (FECR) tests Farmers should have their veterinarian check for drug resistance in the worm populations on every farm, at regular intervals of not less than two years. Only by knowing exactly what the state of anthelmintic resistance on a farm is, can appropriate action be taken. Generalisations such as "benzimidazole resistance is found on most farms" are not much use since they cannot tell us what the situation is on a particular farm. What is just as important, is that we must know not only that resistance is present, but how severe it is (i.e. can we still use a particular drug group at all ?).

$ FAMACHA© evaluation Apart from selection and culling, this system also allows frequent, cheap and easy monitoring of the current situation as regards worm infection, but it applies only to haemonchosis.

$ Submandibular oedema (bottle-jaw) Bottle-jaw detection should accompany the examination of ocular mucous membranes, as submandibular oedema can precede or accompany clinical anaemia. Detecting the oedema introduces an added safety factor for ensuring that individual sheep are not overwhelmed by worm infection.

$ Autopsy for helminth detection Doing postmortem examinations (or autopsies) on animals that have died is an important adjunct to monitoring the extent of worm infections and gives vital information.

$ Abattoir reports Regular reports from abattoirs on the condition of livers for liver fluke and other parasites such as Stilesia hepatica help to form a picture of the helminth situation on a given farm.

Optimise anthelmintic use $ Establish the importance of parasites

Unless the incidence and importance of worm species is known, worm control becomes dangerous and unpredictable guesswork. The basis for any intervention is a knowledge of the predominant worm species and their epidemiology.

$ Use the most suitable drug If the parasites are ranked in order of economic importance, it is possible to decide which drug(s) will be the most suitable in each situation. This forms part of a cost/benefit analysis.

$ Avoid too frequent and too much treatment The old approach of "clean dosing" must be completely abandoned. The aim is to treat only to maintain the equilibrium between parasite, host and environment (that is, worm management). Overtreatment ensures that only resistant parasites can survive. Minimal effective treatment programmes must be the new watchwords.

86

$ Treat all and stay This is a major departure from the recommendations made for close to a century. If all sheep are to be treated, they should remain in the camp (paddock) where they were grazing. This will prevent sheep from contaminating a new pasture with only resistant parasites which survived treatment, thus unwittingly aiding selection of parasites for resistance. In most cases sheep or goats should remain in the paddock for 2-3 weeks after drenching to pick up susceptible larvae before being moved to the safe pasture. However, should a long-acting anthelmintic be used, this period will have to be longer (i.e. 2 to 3 weeks after the effective residual action ends).

$ Treat selectively It is preferable to treat only those sheep or goats unable to cope with the current infection challenge, provided the percentage of non-copers remains below 20 %. This can be done with the FAMACHA© system for haemonchosis, or possibly with condition scoring for other parasites. An immediate move to clean pasture in this case would not be detrimental.

Proper planning $ Use the expert

Only knowledgeable veterinarians who are familiar with the area, farming systems and risks can construct a simple, practical, economic and effective total strategy. In resource-poor agricultural systems, the farming community is often able to plan to achieve its own objectives. Here the objective is not so much to be an expert, but a facilitator and a source of technical knowledge.

$ Use a programme Unless a basic planned system is in place and is used, actions will inevitably be largely reactive and based on ad hoc or panic decisions. It is probably true on most farms that animals are either dosed too often, or with inappropriate drugs, or at the wrong times, or with no coherent plan. By setting up a well thought out dosing plan, the number of ineffective doses can be reduced which only add to the selection pressure for parasite resistance. This is one of the areas in which the knowledge and skills of the local veterinarian are vital for success. The local veterinarian must be up-to-date with the latest thinking on worm management.

$ Flexibility The programme must be flexible to allow for changes in weather, management and farming systems, drug costs or other factors.

Improving the efficacy of the remedy $ Dose over the tongue

By placing the tip of the gun towards the back of the mouth, over the tongue, closure of the oesophageal groove does not occur and thus the full dose lands in the rumen where it is absorbed more slowly - this is particularly important for anthelmintic groups which rely on prolonged blood levels for their effect, such as the benzimidazoles and macrocyclic lactones. This prolonged level of activity (a long so-called "killing zone�) means that the drug against which worms have developed a moderate degree of resistance can be made more effective, although, of course, the resistance of the worms is not reduced, but rather partially

87

overcome. However, dosing over the tongue, if done carelessly, can result in 2 very severe consequences : 1. the dose can land up in the lungs, and cause pneumonia; and/or 2. the nozzle of the dosing gun can penetrate the wall of the pharynx and cause severe,

fatal infection. If the sheep jumps forward, the operator must let the gun �ride� with the sheep, and not oppose it, and the dose must be delivered by a measured, steady pressure rather than a fast, single squeeze.

$ Reduce feed intake It has been shown, in the case of benzimidazoles and closantel, that reducing feed intake (i.e. starvation) for 24 hours prior to treatment will improve the absorption of the remedy because of the lower rate of flow of ingesta. As in the case of dosing over the tongue, this results in a more effective exposure of the parasite to the drug. In turn, this means that the drug is clinically more effective and can partially overcome drug resistance.

$ Repeat the dose This only applies to benzimidazoles and macrocyclic lactones. Two doses given 12 hours apart will increase the �killing zone� of these drugs, allowing more time for a cumulative killing effect. Thus resistant worms can still be killed, although this is achieved at a cost since 2 normal doses rather than one are needed. A double dose, given at one time, will have practically no beneficial effect with these 2 groups of anthelmintics.

$ Increase the dose (beware of toxicity, however) This only applies to drugs which rely mainly on peak concentrations for their effect. In this case, a double amount of drug given at one time can sometimes be effective for killing resistant worms. This is useful for the imidazoles (levamisole only; morantel is not currently registered in South Africa).

$ Correct dosage It may seem too obvious, but a lot of problems are caused by not weighing sheep, not calibrating the dosing gun for accuracy, not checking for repeatability and not reconciling the amount of drug used with the number of sheep treated. Underdosing may be a factor leading to anthelmintic resistance.

$ Use combinations It appears that combining drugs from different groups in 1 dose will not only improve the effective clinical action of these drugs in most cases, but also, mathematical models indicate that it may also have an effect on development of anthelmintic resistance. If drugs are mixed, this can only be effective if the formulation has been fully tested and the mixing carried out by experts, in registered products. Use of home-made combinations is dangerous and against the law.

$ Sustained delivery Medicated blocks or controlled release preparations will increase the clinical efficacy of drugs, which rely on prolonged action for their effectiveness. However, we have to bear in mind that prolonged exposure to a drug at low levels will increase selection for resistance. This approach will therefore only bring temporary relief.

$ Goats are different Because of differences in the rate of metabolising drugs, goats often resemble cattle more closely than sheep in this respect. In many cases this means that goats must be given a higher dosage rate than sheep. Note that many anthelmintics may not be registered for use

88

in goats, or that the recommended dose given is the same as for sheep. If the product is not registered for use in goats, or is drenched at a dosage that is higher than the recommended dosage, the user has no redress if the product is used and fails, or causes losses.

Promising future systems $ Predacious fungi

Predacious (nematophagous) fungi in the soil can constrain larval survival and reduce numbers of larvae available to infect grazing animals.

$ Dilution of resistance By the re-introduction of susceptible strains to a farm where a parasite strain has become resistant to anthelmintics, it is possible to reduce significantly the degree of resistance by means of a dilution effect.

$ Protein supplementation Since resistance and resilience are dependent on adequate nutrition, and the most important factor identified is protein, it is possible to ameliorate the effects of parasites by feeding animals well. However, we need to know when and how much of what supplement must be supplied to which class of animal, and what the cost/benefit ratio would be. However, since protein is expensive and as circumstances dictate the detailed requirements from time to time and from place to place, experts in this field should be consulted before embarking on protein supplementation.

$ Vaccination Although this �holy grail� of worm control still remains a mirage as a practical, economical solution, it is theoretically attainable and can be a potent factor when the problems are overcome.

$ Condition scoring There are indications that this may be feasible for assessing brown stomach worm, bankrupt worm and other non-haematophagous worm infections, but at present it is only a theoretical proposal and requires research and validation.

Integration of worm management principles If only 1 or 2 of the foregoing principles are used exclusively, failure will be certain; it is only by using a prudent mix of strategies that sustainable, cost-effective measures can be established. As above, the decision on which measures are to be used in a given situation can only be made by an expert who is conversant with local conditions. This programme will, of course, have to be drawn up in close consultation with the livestock owner(s). In the resource-poor set-up it will have to be the farming community that makes the decisions in consultation with the expert if the programmes are to be sustainable. Whether the farming system is based on communal ownership, subsistence farming, small-scale farming, commercial farming or stud farming, or is a commercial resourced or a resource-limited system, the principles remain the same. Only the mixture and weighting of worm management measures will vary according to circumstances. Conclusion By every means of evaluation we used in the preliminary evaluation of the FAMACHA© system for drenching only animals unable to cope with heavy worm challenge, it is apparent that it is certainly practical for field use, granted that the precautions and limitations of the system are heeded. It will not only reduce drenching costs, but will also drastically reduce selection for anthelmintic

89

resistance. Farmers who participated in the trials also reported 2 most important by-products of applying the system, in that the appropriate training not only vastly improved the self-confidence and interest of the stockmen involved, but also helped the more timely detection of detect general, non-helminthological problems. The above conclusions are borne out by the fact that few sheep (probably less than 1 %, according to our estimates) died from haemonchosis in the trials reported here, despite the fact that the sheep were drenched only at a stage when they were severely challenged by the worm infection. There is a potential danger that persons using the FAMACHA© system will become complacent concerning worm infection. It should be kept in mind that the system can identify only infection by haematophagous worm species. Unless recognised in time, the presence of a species such as Oesophagostomum columbianum can wreak havoc, as it has a most intensive effect on the production of the sheep, compared with Haemonchus which is deadly but not nearly so erosive concerning the growth of the animals. Trichostrongylus spp. can also be a serious complication, and can easily be overlooked. On one of the farms where we are working, T. colubriformis has caused problems, and we are at present attempting to find a practical solution. It is essential, when applying the FAMACHA© system, that the anthelmintic susceptibility of the worm strains concerned, be known. On one of the farms where we conducted intensive investigations, it appeared (when the farmer was applying the system with little supervision by us) that the system was not effective, as numerous sheep developed severe bottle-jaw, and were in dire straits. However, our subsequent investigations indicated that the compound the farmer was using to drench the affected animals was well-nigh completely ineffective. When he switched to effective anthelmintics, excellent results were obtained. Another obvious advantage of the FAMACHA© system is that animals that are overly susceptible to infection (present with pale conjunctivae) can be identified at little cost. These animals can then be prepared for market, and in the process the average ability of the remaining flock to withstand the Haemonchus challenge is improved. Furthermore, there appears to be the exciting prospect that, in the case of the stud farmer who is prepared to stress the animals with Haemonchus challenge before treating, selection of animals on the strength of FAMACHA© evaluation and progeny testing will result in considerable genetic improvement of the flocks, as regards resilience and possibly also resistance against this worm species.

90

References Acocks, J.P.H., 1975. Veld types of South Africa with accompanying veld type map. Memoirs of the

Botanical Survey of South Africa No. 40, 2nd ed. Botanical Research Institute, Department of Agricultural Technical Services, South Africa: 128 pp.

Albers G.A.A., Gray G.D., Piper L.R., Barker J.S.F., Le Jambre L.F. & Barger I.A., 1987. The genetics of resistance and resilience to Haemonchus contortus infection in young Merino sheep. International Journal for Parasitology 17: 1355-1363.

Allonby E.W. & Urquhart G.M., 1976. A possible relationship between haemonchosis and haemoglobin polymorphism in Merino sheep in Kenya. Research in Veterinary Science 20: 212-214.

Altaif K.I. & Dargie J.D. 1978. Genetic resistance to helminths. The influence of breed and haemoglobin type on the response of sheep to primary infections with Haemonchus contortus. Parasitology 77: 161-175.

Barger I.A., 1985. The statistical distribution of thichostrongylid nematodes in grazing lambs. International Journal for Parasitology 15: 645-649.

Bath, Gareth, Malan, Francois & Van Wyk, Jan, 1996. The "FAMACHA©" Ovine Anaemia Guide to assist with the control of haemonchosis. In: Proceedings of the 7th Annual Congress of the Livestock Health and Production Group of the South African Veterinary Association, Port Elizabeth, 5-7 June 1996: 5 pp.

Besier, R.B., 1997. Ecological selection for anthelmintic resistance: re-evaluation of sheep worm control programs. In: Van Wyk, J.A., Van Schalkwyk, P.C. (Eds.), Managing anthelmintic resistance in endoparasites. Workshop held at the 16th International Conference of the World Association for the Advancement of Veterinary Parasitology, August 1997, Sun City, South Africa: 30-38.

Bird, J., Schulaw, W.P., Pope, W.F. & Bremer, C.A., 2001. Control of anthelmintic resistant endoparasites in a commercial sheep flock through parasite community replacement. Veterinary Parasitology 97: 219-226.

Bisset S.A. & Morris C.A., 1996. Feasibility and implications of breeding sheep for resilience to nematode challenge. International Journal for Parasitology 26: 857-868.

Bisset S.A., Morris C.A., Squire D.R. & Hickey S.M., 1996. Genetics of resilience to nematode parasites in young Romney sheep - selective drenching using weight gain under challenge to assess drench requirements. New Zealand Journal of Agricultural Research 39: 314-323.

Bisset S.A., Morris C.A., Squire D.R., Hickey S.M. & Wheeler, M., 1994. Genetics of resilience to nematode parasites in Romney sheep. New Zealand Journal of Agricultural Research 37: 521-534.

Boomker, J., Horak, I.G. & MacIvor, K.M. de F., 1989. Helminth parasites of grysbok, common duikers and Angora and Boer goats in the Valley Bushveld in the eastern Cape Province. Onderstepoort Journal of Veterinary Research 56: 165-172.

Boomker, J., Horak, I.G. & Ramsay, K.A., 1994. Helminth and arthropod parasites of indigenous goats in the northern Transvaal. Onderstepoort Journal of Veterinary Research 61: 13-20.

Borgsteede, F.H.M. & Duyn, S.P.J., 1989. Lack of reversion of a benzimidazole resistant strain of Haemonchus contortus after 6 years of levamisole usage. Research in Veterinary Science 47: 270-272.

Campbell A.G., 1986. Selection strategies for animal disease resistance. New Zealand Agricultural Science Journal 20: 169-171.

Clunies-Ross I., 1932. Observations on the resistance of sheep to infestation by the stomach worm Haemonchus contortus. Journal of the Council for Scientific and Industrial Research 5: 73-80.

Coles, G.C., Bauer, C., Borgsteede, F.H.M., Geerts, S., Klei, T.R., Taylor, M.A. & Waller, P.J., 1992. World Association for the Advancement of Veterinary Parasitology (WAAVP) methods for the detection of anthelmintic resistance in nematodes of veterinary importance. Veterinary Parasitology 44: 35-44.

Connor, R.J., Munyuku, A.P., Mackyao, E. & Halliwell, R.W., 1990. Helminthosis in goats in southern Tanzania: investigations on epidemiology and control. Tropical Animal Health and Production 22: 1-6.

Dorny, P., Symoens, C., Jalila, A., Vercruysse, J. & Sani, R., 1995. Strongyle infections in sheep and goats under the traditional husbandry system in peninsular Malaysia. Veterinary Parasitology 56: 121-136.

Dunn, A.M., 1978. Veterinary Helminthology, 2nd ed. William Heinemann Medical Books, London, 323 pp. Echevarria, F., Borba, M.F.S., Pinheiro, A.C., Waller, P.J. & Hansen, J.W., 1996. The prevalence of

anthelmintic resistance of sheep in Southern Latin America: Brazil. Veterinary Parasitology 62: 199-206. Eddi, C., Caracostantogolo, J., Pena, M., Schapiro, L. Marangunich, L., Waller, P.J. & Hansen, J.W., 1996.

The prevalence of anthelmintic resistance of sheep in Southern Latin America: Argentina. Veterinary Parasitology 62: 189-197.

Fivaz, B.H., Horak, I.G. & Williams, E.J., 1990. Helminth and arthropod parasites of Angora goats on irrigated Kikuyu grass pastures in the eastern Cape Province. Journal of the South African Veterinary Association 61: 112-116.

91

Gordon, H. McL., 1973. Epidemiology and control of gastrointestinal nematodoses of ruminants. Advances in Veterinary Science 17: 395-437.

Horak, I.G., 1978. Parasites of domestic and wild animals in South Africa. V. Helminths in sheep on dryland pasture on the Transvaal Highveld. Onderstepoort Journal of Veterinary Research 45: 1-6.

Horak, I.G. & Louw, J.P., 1977. Parasites of domestic and wild animals in South Africa. IV. Helminths in sheep on irrigated pasture on the Transvaal Highveld. Onderstepoort Journal of Veterinary Research 44: 261-270.

Horak, I.G., Honer, M.R. & Schröder, J., 1976. Live mass gains and wool production of Merino sheep: three treatment programmes for parasite control. Journal of the South African Veterinary Association 47: 247-251.

Horak, I.G., Knight, M.M. & Williams, E.J., 1991. Parasites of domestic and wild animals in South Africa. XXVIII. Helminth and arthropod parasites of Angora goats and kids in Valley Bushveld. Onderstepoort Journal of Veterinary Research 58: 253-260.

Jackson, F., 1993. Anthelmintic resistance - the state of play. British Veterinary Journal 149: 123-138. Jain, N.C., 1986. Schalm�s Veterinary Hematology, 4th ed. Lea and Febiger, Philadelphia: 1221 pp. Kirkman, K.P. & Moore, A., 1995. Veld management strategies for livestock farmers in the sourveld regions

of South Africa. Bulletin of the Grassland Society of South Africa 6: 25-28. Le Jambre L.F., 1995. Relationship of blood loss to worm numbers, biomass and egg production in

Haemonchus infected sheep. International Journal for Parasitology 25: 269-273. Magaya, A., Mukaratirwa, S., Willingham, A.L., Kyvsgaard, N. & Thamsborg, S., 2000. Effects of

anthelmintic treatment and feed supplementation on grazing Tuli weaner steers naturally infected with gastrointestinal nematodes. Journal of the South African Veterinary Association 71: 31-37.

Maciel, S., Giménez, A.M., Gaona, C., Waller P.J. & Hansen, J.W., 1996. The prevalence of anthelmintic resistance of sheep in Southern Latin America: Paraguay. Veterinary Parasitology 62: 207-212.

Malan, F.S. & Van Wyk, J.A., 1992. The packed cell volume and colour of the conjunctivae as aids for monitoring Haemonchus contortus infestations in sheep. In: Anonymous, 1992. Proceedings of the South Africa Veterinary Association Biennial National Veterinary Congress, 7-10 Sept., Grahamstown: P139.

Malan, F.S., Van Wyk, J.A. & Wessels, C.D., 2001. Clinical evaluation of anaemia in sheep : early trials. Onderstepoort Journal of Veterinary Research 68: 165-174.

McKellar, Q.A., 1994. Chemotherapy and delivery systems - helminths. Veterinary Parasitology 54: 249-258.

McEwan J.C., Bisset S.A. & Morris C.A., 1997. The selection of sheep for natural resistance to internal parasites. In: Sustainable Control of Internal Parasites in Ruminants; Animal Industries; Lincoln University.

Michel, J.F., 1974. Arrested development of nematodes and some related phenomena. Advances of Parasitology 12: 279-366.

Nari, A., Salles, J.l, Gil, A., Waller, P.J. & Hansen, J.W., 1996. The prevalence of anthelmintic resistance of sheep in Southern Latin America: Uruguay. Veterinary Parasitology 62: 213-222.

Presidente, P.J.A., 1985. Methods for detection of resistance to anthelmintics, in Resistance in nematodes to anthelmintic drugs edited by N. Anderson & P.J. Waller, CSIRO Division of Animal Health, Australian Wool Corporation Technical Publication: 13-27.

Reinecke, R.K. 1973. The larval anthelmintic test in ruminants. Technical Communication No 106, Department of Agricultural Technical Services, Republic of South Africa: 20 pp.

Riffkin G.C. & Dobson C., 1979. Predicting resistance of sheep to Haemonchus contortus infections. Veterinary Parasitology 5: 365-378.

Scott, E.W., Robbins, H., Jackson, F., Jackson, E. & Clarkson, M., 1995. Limiting the spread of benzimidazole resistant nematodes: a farm study. Proceedings of the Sheep Veterinary Society, 1993-1994 18: 181-182.

Smith, R.D., 1995. Veterinary Clinical Epidemiology: A Problem-oriented Approach, 2nd ed. CRC Press, Boca Raton: 279 pp.

Soll, M.D., 1997. The future of anthelmintic therapy from an industry perspective, in Managing anthelmintic resistance in endoparasites, edited by J.A. van Wyk & P.C. van Schalkwyk. Workshop held at the 16th International Conference of the World Association for the Advancement of Veterinary Parasitology, 10-15 August 1997, Sun City, South Africa: 1-4.

Theiler, A., 1912. Wire-worms in sheep and their treatment. South African Agricultural Journal. Bulletin No. 63: 17 pp.

Van Schalkwyk, P.C., Schröder, J., Malan, F.S. & Van Wyk, J.A., 1995. Worm Workshop: Recommendations on Worm Control, 1st rev. Division of Helminthology, Onderstepoort Veterinary Institute, Pretoria, 32 pp.

92

Van Wyk, J.A., 2000. Managing gastrointestinal helminths biologically: some aspects of intensive investigations in the field in South Africa. In: Anonymous, FAO TCP Workshop on Sustainable Worm Control Programmes for Sheep and Goats, June 2000, Faculty of Veterinary Science, University of Pretoria, Pretoria, South Africa: 64-73.

Van Wyk, J.A., 2001. Refugia - overlooked as perhaps the most potent factor concerning the development of anthelmintic resistance. Onderstepoort Journal of Veterinary Research 68: 55-67.

Van Wyk, J.A. & Van Schalkwyk, P.C., 1990. A novel approach to the control of anthelmintic resistant Haemonchus contortus in sheep. Veterinary Parasitology 35: 61-69.

Van Wyk, J.A., Alves, R.M.R. & Michael, L.M. (Abstract), 1997a. A novel key for identifying nematode infective larvae (L3) from domesticated ruminants. In: Anonymous, Proceedings of the 16th International Conference of the World Association for the Advancement of Veterinary Parasitology, August 1997, Sun City, South Africa: p84.

Van Wyk, J.A., Malan, F.S. & Bath, G.F., 1997b. Rampant anthelmintic resistance in sheep in South Africa � what are the options? In: Van Wyk, J.A. & Van Schalkwyk, P.C., 1997. Managing Anthelmintic Resistance in Endoparasites. Workshop held at the 16th International Conference of the World Association for the Advancement of Veterinary Parasitology, 10-15 August 1997, Sun City, South Africa: 51-63.

Van Wyk, J.A., Malan, F.S. & Randles, J.L., 1997c. How long before resistance makes it impossible to control some field strains of Haemonchus contortus in South Africa with any of the anthelmintics. Veterinary Parasitology 70: 111-122.

Van Wyk, J.A., Schröder, J., Van Schalkwyk, P.C., Horak, I.G., 1987. Tegnieke : Helmintologie. In: Schröder, J. (Ed.), Proceedings of the Worm Resistance Workshop, August 1987, Pretoria, South Africa.

Van Wyk, J.A., Stenson, M.O., Van der Merwe, J.S., Vorster, R.J. & Viljoen, P.G., 1999. Anthelmintic resistance in South Africa: surveys indicate an extremely serious situation in sheep and goat farming. Onderstepoort Journal of Veterinary Research 66: 273-284.

Vatta, A.F., Letty, B.A., van der Linde, M. J. & Krecek, R.C., 2000. Preliminary report: testing of a chart for the diagnosis of ovine clinical anaemia caused by haemonchosis for use in goats farmed under resource-poor conditions in South Africa. In: Anonymous, FAO TCP Workshop on Sustainable Worm Control Programmes for Sheep and Goats, June 2000, Faculty of Veterinary Science, University of Pretoria, Pretoria, South Africa: 44-55.

Vatta, A.F., Letty, B.A., Van der Linde, M.J., Van Wijk, E.F., Hansen, J.W. & Krecek, R.C., 2001. Testing for clinical anaemia caused by Haemonchus spp. in goats farmed under resource-poor conditions in South Africa using an eye colour chart developed for sheep. Veterinary Parasitology 99: 1-14.

Waller, P.J., 1997. Anthelmintic resistance. Veterinary Parasitology 72: 391-412.

93

Acknowledgments Organisations

The contributions to the funding and the support provided by the Faculty of Veterinary Science at the University of Pretoria, ARC-Onderstepoort Veterinary Institute, National and Provincial Departments of Agriculture (including Gauteng, North-West, Mpumalanga, KwaZulu-Natal, Free State), Food and Agriculture Organization, Intervet SA, and National Wool Growers� Association were absolutely essential. This work was also supported by a grant from the Technology and Human Resources for Industry Programme (THRIP); a partnership programme funded by the Department of Trade and Industry (DTI) and managed by the National Research Foundation (NRF).

People From the University of Pretoria we were given invaluable support by Morkel Boshoff, Ali Chuene, Rachel Coetzee, Frans Magwai, Sarie Jacobs, Christine Jansen van Vuuren, Estelle Mayhew, Amanda Olivier, Rina Serfontein, Anatulie Sutherland, Lawrence Tshikhudo, Heleen Smit, Ryno Watermeyer, Dirk Booyse and Susan Milne. From the FAO we received the support of Seatla Nkosi, Florence Chenoweth and M M Sefeane and N M Ncokazi. Staff of the Onderstepoort Veterinary Institute were extremely supportive and played a vital role. In particular we must mention Theo de Waal, Daniel Chipana, William Shima, Frans Masubelle, Christo Troskie, Roy Williams, Ofhani Nefolovhodwe, Fanie Buys, Lettie Beyer, Laura Lopez-Rebollar, Marie Watson, Pienie Vorster, Marina Lubbe, Solet Prinsloo and Fréda van Wyk. Staff of National and Provincial Departments of Agriculture were crucial to the project. In particular we mention Emily Mogajane, Tertius Bester, Jacoba Wessels, Org Fourie, Ronél Smit, Venty Mahlangu, Lee Pachonic, Louis van Rooyen, Johan Scholtz, Reinet van Niekerk, Sibusiso Madiba, Gininda Msiza and Wynton Rabolao. The National Wool Growers� Association�s Arno Moore and Danie van Niekerk were particularly helpful. Staff at Hoechst Roussel Vet (now Intervet) were consistently supportive. Peter Oberem, Hawie Havemann, Ernst Arndt, Alan Kloeck, Gaynor Russell and Gideon Shabangu deserve special mention. Logos AgVet is thanked for donation of anthelmintics. Numerous farmers were an integral part of the various aspects of the project and were not only prepared to do a great deal of work to collect data, but also bore the brunt of the risk of losses by partaking in the project : Fanie Barnard, Marais and Andries Bester, Johan Bührmann, Carel and Willem Cilliers, Frank da Serra, Maqiti Duma, Petros Khumalo, Peter Kleu, Jean Köhler, B K Mthombeni, Mike Nicolau, Obie Oberholzer, Ishmael Segwe, Gert and Ronél Smit, Dewald Steyn and Pieter van Zyl. Others who were vital to the success of the project were Hannes Dreyer, Alta Stenson and Eddy Krecek. Chris Morris and Sharon Hickey of AgResearch, Ruakura Agricultural Research Centre, New Zealand, undertook the analyses required to calculate performance ratings of animals in the veld ram club test flock, breeding values of animals in the Merino stud flock and genetic parameters (heritability and correlation estimates) for the resistance, resilience and productivity traits measured. Their contributions are much appreciated.

94

David Smith (Moredun Veterinary Institute, UK) supplied a susceptible strain of H. contortus. We are grateful to Helena Theron for analysing most of the Badplaas data. The South African Weather Bureau is thanked for supplying the climatic data. Roy Tustin is thanked for his assistance with editing, and Koos van der Lende for the cover photograph. The families of the authors are thanked for their ongoing support.

i

Appendix 1

FAMACHA© chart

ii

Colours are not a completely accurate reflection of the original card

iii

Appendix 2

Practical implementation of the FAMACHA© system

iv

Training of farmers and workers This will be done mainly by veterinarians already trained by the research group. However, it is possible that in specific circumstances this training may be supervised by other competent, trained and knowledgeable people.

Phases 1. Explain the problem of wireworm (haemonchosis) 2. Explain the FAMACHA© guide 3. Demonstrate the technique using live sheep 4. Allow participants to make their own evaluations under supervision 1. Explain the problem Veterinarians will use background information as supplied by the research and development

group to supplement and reinforce information given on the FAMACHA© guide and accompanying information pamphlet. While the exact format cannot be laid down, it is suggested that the trainer (veterinarian or other) covers at least the following : # Worms in general and their effects # Wireworm as the biggest problem # Explain its bloodsucking nature # Explain how this leads to anaemia # What is anaemia? # How is it measured? # How it can be seen in mucous membranes # Other causes of anaemia # Causes of inflammation or congestion which can mask clinical anaemia # The importance of faecal egg counts # The importance of an integrated worm management programme # The problem of resistance to anthelmintics

2. Explain the FAMACHA© card

# Do not expect that everyone can read # Explain all the symbols and be sure that everyone understands # Be prepared to repeat something not properly understood # Ask people what they understand by the symbols # Be sure that participants all know what they should do, depending on the examination of

sheep # Emphasise advantages and problems

3. Demonstration of use

# Take participants to a number of sheep and show them what has been discussed # Allow time for questions # Don�t be afraid of repetition

4. Supervised evaluation

# Allow participants to evaluate sheep on their own # Correct where necessary # Issue certificate to successful participants

Training – Example of a session for stockmen with little education

[Encourage participants and trainees, lead them, and praise them every time they get something right !]

v

The problem ! Good afternoon ! Thank you for giving us the chance to tell you something about worms,

which make sheep and goats on the farm ill, and cause you to lose money. ! I will speak about worms of sheep, goats and cattle � have you ever seen such worms,

which occur in the stomach of sheep ? [Demonstrate a sample of worms]. ! Worms are one of the most important disease problems of sheep. [Agreement?] ! In those parts of the country where the rain falls mostly in summer, the wireworm is the

most important of the worms making the sheep sick, or killing them. ! Why ? These worms suck blood and in this way cause the sheep to lose blood : No blood �

no sheep ! We say the sheep is anaemic if it loses too much blood. ! Have you seen what the eye of a healthy sheep looks like � the inside of the eyelid is red, as

I�ll show you shortly. But when worms suck out the blood of the sheep, the eye becomes pink in colour, and later, shortly before the animal dies, the eye is white (like milk). Who of you has seen this already ? [Wait for an answer]

! Even when there are very many worms, only some of the animals will become very pale, with white eyes and eyelids. Most of them will still look healthy, with eyes that are red, but the eyes of just some of the sheep become white, and it is only these animals that die from the worms.

! Remember, there are also other things that can cause the eyes of sheep to become �white�, but most of the time it is wireworms that suck the blood of the sheep, and cause their eyes to become white when they have almost been killed by the worms.

Demonstration of symbols and pictures (obtained by cutting up one of the cards)

This section was compiled primarily for persons who are poorly literate and therefore aims at ensuring that the course participant understands the principal of the system. The testing (below) with symbols tends to give these persons confidence. However, also keep in mind that some educated people are colour blind, and may perhaps also not be able to rank the colour categories correctly. ! A series of questions is posed to a volunteer person [Ask : What does the following

mean to you ? and demonstrate the following small pictures (symbols) of drenching guns, etc.] # Little sketch on the flipside of the FAMACHA© card, showing how to open the eye. # Drenching gun � [show picture]. Another way we can stop the worms from becoming a

problem, is to dose the sheep when necessary, and to move the animals to another camp if the worms are too bad.

# Danger sign [skull and crossbones] � what does this mean ? # A correct tick [show picture] � what does this mean ? # A question mark [show picture] � what does this mean ? # The colours of the FAMACHA© card : [Demonstrate them separately, and then

show on the chart where to look in the eye of the sheep, and where the blocks of colour are located, to compare. Let each of a few volunteers sort the coloured blocks in colour sequence : Firstly according to the colour sketches of the sheep’s eyes, and then the coloured blocks separately]

# The numbers 1 to 5 [what they represent]

Warnings ! But we must be careful not to look only at the eye, because other diseases than worms can

make the eye red. For example, fever and other diseases and dust can cause the eye to turn red, so that the white eye caused by wireworms, cannot be seen. Also, when the sheep have walked for a long distance, the eyes will be redder.

vi

! The biggest problem is that if you treat all the sheep too often, then you make the worms

able to fight against the medicine, and thereafter it is very difficult to kill the worms. So, we use this FAMACHA© card to make it easier to see which sheep are becoming ill. Then it is only very seldom necessary to dose all of the sheep. Most people still treat all the sheep every time, when only a few of them have a worm problem.

! Now, if you put a little of the sheep�s blood in a machine that works like a �milk separator�,

then we can measure how much blood a sheep has. [Demonstrate microhaematocrit machine and microhaematocrit tubes]. But this is too difficult on a farm, and that is the reason why we rather use the colour of the eye to decide which animals have to be treated.

! To repeat, it is the blood that makes the eye of the sheep red. So, if there is too little blood,

then the eye of the sheep also becomes less red, later pink, and then whiter and whiter, until it is as white as milk in sheep that die from too many wireworms.

! When the sheep becomes very sick from wireworms, then it also gets a swelling under its

bottom jaw [show where]. [Who of you have seen this already ?] ! Remember, if the farmer loses money because the worms kill the sheep, he will

no longer farm, and all of us here will also no longer be certain of having work.

Worm egg counts ! We do not look only at the eye of the sheep to see whether it has too many worms � we

also look at the manure. Why do we look at the manure ? [Wait for answer] The worms live in the stomachs and intestines of the sheep, and when they lay eggs, these eggs come out of the sheep in the manure. The eggs of the worms are too small to see with the eye, but we can see them with a microscope, and count them. So, by counting the eggs, we can make sure that the sheep will not be killed by the worms, because we can treat the animals before the worms become too bad. We then also look at the eye of the sheep to see which sheep are weak against the worms. This is what we want to help you learn today � to see when the eye of the sheep is white.

Demonstration ! Show trainees how the eye is opened. Using the FAMACHA© card, show how the colours are

compared and matched. Let the trainees give their opinions, then give your rating. If haematocrits have been done, this will help confirm the rating. Show the examination procedure several times.

! Next, allow trainees to evaluate other sheep themselves, with trainers looking on, assisting and correcting. Build their confidence and reinforce correct evaluations.

Evaluation ! Finally, let trainees look at a number of different sheep and give their rating individually,

without any help. Every participant must do it without voicing an opinion, so that each person must do the evaluation him or herself, and the trainer, in turn, can evaluate each person.

vii

! After about 20-30 sheep it becomes clear which trainees have the knack, confidence and natural ability, and which do not. Some people can also be colourblind. These people are then not given this task on the farm.

Suitable sheep ! If at all possible, there should be a range of haematocrit values in the group of sheep to be

used for the training; usually if Haemonchus is a problem and it is late summer, there should be both normal and anaemic sheep. It is also possible to prepare sheep to show anaemia by controlled exsanguination (protocol attached).

Certification ! Those trainees who successfully complete the course and show a good, consistent aptitude

to evaluate the colour of mucous membranes, can be issued with a certificate by the trainer (who must also have passed the course). A copy of this certificate is included in the supporting documents.

Monitoring and backup ! Phone the farmers regularly (every 2 weeks), to ensure they are examining the sheep. Ask

to see the flock histograms (see �Instructions� insert) to get the numbers of sheep or goats in the flock assigned to each category. Check to see that there are no untoward things happening! (This is often blamed on the system, whatever the cause). Investigate and give advice. Make sure that faecal egg counts are done regularly and the results are given to you for evaluation and advice.

viii

Appendix 3

Information pamphlet

ix

x

xi

xii

xiii

Appendix 4

“Histogram”

xiv

Vormnommer : 9 November 1998

FAMACHA© Proef / Trial FAMACHA© Anemie-Opgawe / Anaemia Record

vir kudde skape / for flock sheep

Boer/Farmer : Datum/Date : Middel/Remedy : Plaas/Farm : Kudde/Flock : Gedoen deur/Done by :

A B C D E

50 100 150 200

$ Getel / Counted [A,B,C] Υ Getel en behandel / Counted and dosed [D,E, (C)] Υ Kwakkeel / Bottle jaw Getal diere in totale trop/Total no. animals in flock Kamp verwissel?/Camp changed? MET/FEC Getal diere in trop ingebring/No. animals brought in to flock Datum van kampverwisseling/Date of camp change Datum/Date Getal geboortes/No. of births Nuwe kamp ID/New camp ID Kultuur/Culture Getal diere uit trop verwyder/No. animals removed from flock Maande sonder kleinvee/Months small-stock-free Getal vrektes/No. of deaths

xv

Appendix 5

Fb5 - Marking

xvi

xvii

xviii

Appendix 6

Perceptions questionnaire for farmers

xix

FAMACHA© EVALUERING DEUR BOER FAMACHA© EVALUATION BY FARMER

Boer Farmer

Plaas Farm

Distrik District

Datum Date

Plaaswerker Farm worker :

Merk asseblief die kolom (%) wat die meeste ooreenstem met jou waarneming. Daar is geen ‘regte’ of ‘verkeerde’ antwoorde nie. Please tick the column (%) which most closely corresponds with your experience. There are no ‘right’ or ‘wrong’ answers.

Stem h

ewig saam

/Strongly agree

Stem saam

/Agree

Neu

traal/Neutral

Verskil/D

isagree

Verskil h

ewig/Strongly disagree

Die sisteem het vir my geld bespaar 1. The system saved me money Die sisteem was moeilik om te gebruik 2. The system was difficult to use Ek het te veel vrektes gehad 3. I had too many deaths Dit het my meer bewus van wurms gemaak 4. It made me more aware of worms Die toepassing van die sisteem is deeglik verduidelik 5. Its application was well explained Dit is �n nuttige sisteem 6. It is a useful system Die sisteem was onprakties 7. The system was impractical Dit het my kudde se produksie hewig benadeel 8. It badly affected my flock�s production Die illustrasies was swak 9. The illustrations were poor Ek sal graag ander soortgelyke programme wil laat ontwikkel 10. I would like to have other, similar programmes developed Ek het meer besorgd geraak oor wurmmiddelweerstand 11. I have become more concerned about drug resistance Die sisteem was duurder as wat ek verwag het 12. The system was more expensive than I had anticipated

xx

Die gebruik van wurmmiddels het merkbaar afgeneem 13. The usage of worm remedies decreased significantly Dit het te lank geduur om die skape te ondersoek 14. It took too long to examine the sheep Deur nie te doseer nie, is die skape te veel benadeel 15. Not dosing affected the sheep to severely Persone wat met die skape werk, het meer bewus van wurms geraak

16.

People who work with the sheep became more aware of worms Ek sal die sisteem weer in die toekoms gebruik 17. I will use the system again in future Die inligtingspamflette en instruksies was duidelik 18. The information pamphlets and instructions were clear Die sisteem was die uitgawes werd 19. The system was worth the costs Ek het die sisteem betroubaar gevind 20. I found the system reliable Daar was te veel onverwagse probleme 21. There were too many unexpected problems Ek het meer kontak met my veearts gehad 22. I had more contact with my veterinarian Daar was goeie opvolging van probleme wat ondervind is 23. There was good backup for problems encountered Ek sal die sisteem by ander boere aanbeveel 24. I will recommend the system to other farmers Ek het net die kaart en instruksies nodig gehad 25. I only needed the card and instructions Hierdie behoort nog verder ontwikkel te word 26. This is something that should be developed still further Weerstand teen wurmmiddels is onbelangrik in skaapboerdery 27. Anthelmintic resistance is not an important factor in sheep farming Die sisteem het nie my boerderybestuur onvergegooi nie 28. The system did not upset my farming management Dit was net �n groot foefie wat nie gewerk het nie 29. It was just a big sham which did not work Ek moes te veel diere uitskot 30. I had to cull too many animals Haarwurm het die oorhand in my trop gekry 31. Wireworm gained the upper hand in my flock Die sertifikaat was a goeie idée 32. The certificate was a good idea Geen opleiding was nodig nie 33. No practical training was necessary Ek kon dit nie in my skedule inpas nie 34. I could not fit this into my schedule

Kommentaar/Comments

xxi

Appendix 7

Perceptions questionnaire for veterinarians

xxii

FAMACHA© - PROEF / TRIAL

ALGEMENE INDRUKKE DEUR DEELNEMENDE VEEARTSE

OVERALL IMPRESSIONS BY PARTICIPATING VETERINARIANS

Merk asseblief die kolom (%) wat die meeste ooreenstem met jou waarneming. Daar is geen ‘regte’ of ‘verkeerde’ antwoorde nie. Please tick the column (%) which most closely corresponds with your experience. There are no ‘right’ or ‘wrong’ answers.

Sleg / Bad

Swak / Poor

Matic / Fair

Goed / G

ood

Uitsteken

d / Excellent

BEGRIP illustrasies, pamflette, verduidelikings, inligtingstukke

UNDERSTANDING illustrations, pamphlets, explanations, backup

TOEPASLIKHEID hoe moeilik, tyd geneem, betroubaarheid PRACTICALITY difficulty, time taken, reliability

DIEREPRODUKSIE effek, stress, vrektes, probleme ANIMAL PRODUCTION effects, stress, deaths, problems

FINANSIëLE ASPEKTE uitgawes, besparings, koste/voordele FINANCIAL ASPECTS expenses, savings, cost/benefit

BEWUSTHEID van wurms, van weerstand, van beheermaatreëls, van arbeid, oor die veearts

AWARENESS of worms, of resistance, of control measures, of labour, about the vet

ALGEMEEN soortgelyke sisteme benodig, sal weer gebruik, sal by ander aanbeveel

OVERALL similar systems needed, will use again, will recommend to others

SPESIFIEKE KOMMENTAAR SPECIFIC COMMENTS

xxiii

Appendix 8

Instructions for bleeding of sheep for demonstrations

xxiv

Considerations ! The sheep has a contractile spleen which can raise the haematocrit (Ht) 10-13 % (not

percentage points). ! The normal ovine haematocrit should be 25-35 % (seldom above 40 %). ! Blood volume is 17 % in ovine neonates, 10 % at 2 months and 6½ % in animals 27-48 kg

(% expressed as volume per body mass). ! When the Ht drops below 15 %, blood transfusions are considered in dogs. ! Only 1 % of body weight in blood is currently removed from blood donor animals at Veterinary

Academic Hospital (VAH), Onderstepoort. ! At Onderstepoort Veterinary Institute (OVI), sheep blood donors infected with heartwater for

vaccine production are bled at 2 % of bodyweight with very few deaths (in which case the effect of Heartwater infection must be considered).

! The Ht will remain much the same immediately after exanguination (decrease in blood volume only) and will only stabilise at an adjusted lower level 12 hours later.

! Starting with an Ht of around 30-35 %, bleeding 3 % of BW can be calculated to drop the Ht to not less than 15-18 %, taking splenic contraction into account. This Ht is not life-threatening for a sheep.

! Haemopoeitic effects (new erythrocyte production) will only be seen after 3 days ! Hypovolaemic shock is the main danger in this procedure Materials ! Sheep : 2-6 healthy, normal sheep of the same sex, age and condition (preferably ram lambs

of 3-12 months old) should be used. ! Bleeding system : Jugular bleeding using bleeding packs or a sterile, heparinised jar. Method ! Initial bleeding: the starting Ht and body weight (BW) is taken on day -2. The amount of

blood to be removed at the first stage is calculated as follows :

WV = BWBV x x Ht

Ht-Hts

ts

where Hts = starting haematocrit Htt = target haematocrit BV = Blood volume/body weight in ml/kg (= 65 in adults or 80 in lambs) BW = Live body weight in kg WV = Withdrawn blood volume (ml)

The initial bleeding is aimed at reducing the Ht to between 15-18 %, depending on the final target, but not withdrawing more than 3.3 % body weight. # Subsequent bleeding : the Ht is measured again on day -1, and a further calculation made

as above to reduce Ht further to final target figures. # Final target figures should be 10-12 % (2 sheep) which will correspond to category 5 (Ht

≤12 %), 13-17 % (2 sheep) which will correspond to category 4 (median Ht 15 %) and 18-22 % (2 sheep) which will approximately correspond to category 3 (median Ht 20 %).

# Minor adjustments may be made in the afternoon of day -1. Monitoring ! Timing : sheep must be monitored closely for the first 2 hours after bleeding, then hourly

for another 4 hours, then twice daily. ! Symptoms : hyperventilation, hypoxia, collapse, pallor, shock. ! Stress : animals should be handled as little as possible after bleeding and not be made

to run. Driving should be done at a slow walking pace.

xxv

Untoward effects ! If sheep show severe signs of hypovolaemic shock, they should either be retransfused with

their own blood or (if less severe) with I.V. fluid. ! If it becomes clear the animal is not responding or terminal (we have not experienced any

such cases), it must be euthanased. ! If sheep show transient, mild signs they should be closely observed only. ! If mild signs occur during bleeding, this may be temporarily discontinued. ! It may be preferable, in order to minimise the possibility of hypovolaemic shock, to

immediately transfuse an isovolaemic amount of suitable fluid (Ringer�s lactate solution or even Normal Saline) when exsanguination is complete.

Termination of the procedure The sheep should be observed twice daily for 1 week after the demonstration. If deemed necessary auto-retransfusion can be done if Ht is below 15 % after 1 week.

xxvi

Appendix 9

Perceptions questionnaire for farm workers

xxvii

FAMACHA© PROEF / TRIAL

Indrukke deur plaaswerkers Impressions of farm workers

Boer

Farmer Plaas Farm

Distrik District

Datum Date

Plaaswerker Farm worker :

Merk asseblief die kolom (%) wat die meeste ooreenstem met jou waarneming. Daar is geen ‘regte’ of ‘verkeerde’ antwoorde nie. Please tick the column (%) which most closely corresponds with your experience. There are no ‘right’ or ‘wrong’ answers.

Stem h

ewig saam

/Strongly agree

Stem saam

/Agree

Neu

traal/Neutral

Verskil/D

isagree

Verskil h

ewig/Strongly disagree

Ek het die illustrasies, verduidelikings en hoe die FAMACHA© gebruik word, verstaan / I understood the illustrations, explanation how the and system is used

Die FAMACHA© werk maklik, vinnig en is betroubaar / The system was workable, easy, quick and reliable

Die skape het nie van die wurms siek geword of gevrek nie / The sheep did not get sick or die of worms

Ons het die skape minder as tevore (in vorige jare) gedoseer / We dosed the sheep less than before (in previous years)

Ons het meer bewus geword van wurmbeheer en wat om daaroor te doen / We became more aware of worm control and what to do

Die FAMACHA© is nuttig en het goed gewerk / It was a useful system and worked well

Kommentaar/Comments

xxviii

Appendix 10

Proficiency certificate

xxix

SERTIFIKAAT VAN BEVOEGDHEID CERTIFICATE OF COMPETENCE

FAMACHA© ANAEMIE GIDS/ FAMACHA© ANAEMIA GUIDE

Wie dit mag aangaan / To whom it may concern :

Dit word hiermee gesertifiseer dat / It is hereby certified that

________________________________________

___

van / of

___________________________________________

___________________________________________

met welslae die kursus in die gebruik van die FAMACHA© gids, om bystand met haarwurmbeheer in skape te verleen,

voltooi het / has successfully completed a course in the use of the FAMACHA© guide

xxx

to assist with the control of wireworms in sheep

___________________ _____________ Veearts / Veterinarian Datum / Date