- Merino breeding research in South Africa
|Last update: August 18, 2011 10:02:05 AM|
MERINO BREEDING RESEARCH IN SOUTH AFRICA
J J Olivier
Grootfontein College of Agriculture
The Merino remains South Africa's most important wool producing sheep breed. Over the years many different methods were used to genetically improve the Merino. To this regard some research was done over the years to aid the breeder. In this article a brief account of some breeding research on Merino’s in South Africa is given. It is, however, not the intention to review and summarize results from research on breeding as such. The following topics will be discussed, viz. research flocks, phenotypic and genetic parameters, applied research in the industry and some of the crossbreeding experiments with the Merino.
1. Research Flocks
Most of the basic research on breeding has been done in flocks which had a specific experimental design and hypothesis to be tested. However, especially prior to 1962, the Department of Agriculture mainly at the Grootfontein Agricultural College, maintained a number of Merino studs on which some research was done. The Merino studs of the Department of Agriculture will therefore also be mentioned.
1.1 Studs and flocks at Grootfontein prior to 1956
The first Merino stud was established as early as 1912 at the then Grootfontein School of Agriculture. Very little data concerning to this stud is available. However, Warren (1933) gave a description of this stud and how it had been converted during 1929 into a "Wanganella" stud. At the same time a second stud, namely the "Tasmanian flock" was established, as well as the "Double flock," which was a cross between the former two (Warren, 1933). Although the purpose with these studs is not entirely clear, it is assumed that adaptability of these two strains to South African conditions was evaluated.
Of the outcome of selection in these three studs little is known. However, some of these animals could have, been used in the work of De Vries (1943) on the causes of blowfly strikes in Merino sheep. Observations were made over a five year period it was found that the main causes of blowfly strikes were the following (De Vries, 1943):
- Pleats near the vulva
- Conformation faults for example "cow hocks", droopy rump and narrow hindquarters.
- Abnormal vulva.
- The length of the tail shorter than 4 inches.
The work leads to suggestions for the selection of blowfly resistant animals (Bonsma & De Vries, 1943 and De Vries & De Klerk, 1944) and the establishment of the A- and B-studs at the Grootfontein Agricultural College during 1942 (Nel, 1967).
The A-stud comprised sheep with good frontal development and medium amounts of body pleats. The B-stud consisted of sheep without any frontal body pleats (Nel, 1967). It was argued that the B-stud would be less susceptible to blowfly strike than the A- stud. This experiment ran from 1942 to 1954 and various smaller experiments sprang from it. During 1946 two separate flocks originated from these studs namely the AF-flock from stud A and the BF-flock from stud B (Nel, 1967). These flocks were kept on the veld with the purpose of measuring production in these types under natural veld conditions.
During 1941 a flock (O-flock) was established by selecting the ideal type for resistance against blowfly strike as described by Bonsma and De Vries (1943). A control group, the V-flock, was run together with the a-flock. The a-flock was maintained until 1954 and the V -flock until 1947 (Nel, 1967). The results of these experiments were summarized by Nel (1967) as lows:
- A-stud ewes produced 0,83 kg (1,85 lb) more wool than B-stud ewes and AF-flock ewes 0,45 kg (1,00 lb) more wool than BF-flock ewes.
- Both B-stud ewes and BF-flock ewes had a 10% higher lambing rate then their A-stud and AF-flock ewe counterparts.
- Mortality between 4 months and 18 months as well as in adult ewes was higher in A-stud ewes and AF-flock ewes when compared to that found in B-stud ewes and BF-flock ewes. The difference in mortality rates between AF-flock ewes and BF-flock ewes was higher than between the A- and B-stud ewes, which illustrates that the detrimental effect of excessive pleats is exaggerated under less favourable conditions.
- O-flock ewes of all ages were less susceptible, blowfly strikes than ewes of the V-flock. Nel (1967) also used data from these flocks to investigate the influence of flock structure on production and reproduction characteristics in the Merino. In addition, data from these flocks were also used by Nel (1967) to arrive at parameters for the different traits which will be discussed at a later stage (see 2.1 to 2.3). Data from the A-stud and B-stud were analyzed by Bosman (1958) to determine parameters (also see 2.1 to 2.3) for different -production traits. These parameters were the first available for South African Merino's and are included in sections 2.1 to 2.3.
1.2 The Grootfontein Merino stud
The present Grootfontein Merino stud was established in 1956 with 290 ewes purchased from various South African studs and four rams imported from Australia. Some of the A- and B-stud ewes were also incorporated in this stud. This stud was managed and selection was done as in a typical Merino stud from 1956 to 1985. However, from 1966 a half classing method (50% of animals culled on visual appraisal at a young age and replacement animals selected from the -rest at two-tooth age according to performance) was suggested but was never fully applied. As a result of this, performance records were kept from 1966 (greasy fleece mass records were kept since 1956). The annual response in production traits as predicted by mixed model methodology was as follows (Olivier, 1989):
Greasy fleece mass (1956 - 1985) = 0,014 kg (0,21%)
Clean fleece mass (1966 - 1985) = 0,024 kg (0,53%)
Body mass (1966 - 1985) = 0,166 kg (0,38%)
Fibre diameter (1966 - 1985) = 0,021 micron (0,09%)
This low genetic response in these four production traits was ascribed to the minor importance of these traits in the selection of replacement animals compared to other non-production traits (subjective traits)(Olivier, 1989). However, this stud received wide recognition throughout the Merino stud industry and was classified as a parent stud (Erasmus, 1977).
A score card describing the various subjective traits of the Merino was compiled by personnel of the Grootfontein Agricultural College during 1957 (Anon, 1957) This score card consists of eight body conformation traits and seven wool traits and was used to describe animals in the Grootfontein Merino stud. Data from the stud were used by Nel (1970) to calculate phenotypic correlations between the various subjective traits described according to the scorecard. He found low correlations (0,30) between the various subjective wool traits and greasy fleece mass. High correlations (>0,70) were found between forequarter/neck and back/barrel, general appearance and overall degree of excellence, quality and overall degree of excellence and density with staple, tip and substance. These correlations give some indication of the subjective traits which were considered to be of importance during that time.
With the recording of data on computers and the need for accurate progeny testing of AI sires, a more accurate method to describe these non-production traits was needed. Olivier et al (1987 and 1989) developed a linear score card for Merino sheep and tested its application in the Grootfontein Merino stud. The first score card had 32 traits and was later adjusted to include 21 traits in total. This scorecard has been used since 1987 in the National progeny-testing scheme of the Merino Stud Breeders' Society and was again adjusted during 1989 to eliminate some of the shortcomings. The present score card has only 11 traits which are scored from one to 50. Further research is needed on the heritability of and correlations among these non-production traits and other production traits.
Data from the Grootfontein stud were also used to demonstrate the higher accuracy of BLUP breeding values in comparison to within-group performance tests as well as the influence of known environmental effects on production traits (Olivier, 1989). Badenhorst (1987) used data recorded from 118 Grootfontein stud rams to demonstrate the increase of fibre diameter owing to increased age (2,5 micron increase form 12- to 18-months of age).
1.3 Carnarvon selection flocks
During 1962 Dr J. E. Nel started the first selection experiments on Merino sheep in South Africa. Three flocks of 200 ewes each were established at the Grootfontein Agricultural College. The methods of selection in each of the three flocks were as follows:
S-Flock - Selected for clean fleece mass with restrictions on crimp number and frontal pleat development.
B-Flock - Selected subjectively (hand-and-eye method)
K-Flock - Control flock no selection
All the animals were transferred to the Carnarvon Experimental Farm during 1964 and the experiment was suspended during 1984 owing to severe drought conditions.
The response in production traits of the two selection flocks as measured against the control flock was reported by Olivier (1980 and 1983). Owing to natural selection in the control flock, a mixed model analysis of all three flocks was done by Erasmus (198&). The annual genetic response in the two selection flocks for the different production traits was as follows:
|Clean fleece mass||0,014 kg||0,008 kg|
|Fibre diameter||0,016 micron||0,042 micron|
|Body mass||0,098 kg||0,189 kg|
As expected, the response in clean fleece mass from the S-flock was higher than the response in the B- flock. The low genetic response in the S-flock was ascribed to a low heritability (0,23) and low selection intensity owing to low reproduction rates (0,70 lambs weaned per ewe mated (Erasmus, 1988). On the other hand both fibre diameter and body mass of the B-flock showed higher genetic responses than that found in the S-flock. Erasmus (1988) found a significant response of 0,168kg per annum in body mass in the control flock. He ascribed this partly to natural selection in the harsh environment owing to the fact that some sheep were at a sub-optimum body mass at mating.
As far as the two selection methods are concerned, Olivier (1980) reported that the selection pressure in the S-flock for fleece mass was 50% higher thffi1 that in the B-flock. In this regard, Erasmus (1988) stated that: "the present study has illustrated how Merino judges place as much or more emphasis on components of wool production (body mass and fibre diameter) than the trait they are actually attempting to improve (clean fleece mass)". The same result was also found in the Grootfontein Merino stud (Olivier, 989) where selection differential for body mass was near the theoretically expected level, but the selection differentials for clean fleece mass and against thicker fibres were highly significantly lower than the expected values.
Erasmus (1988) and Delport (1989) used data from the Carnarvon experimental flocks to estimate heritabilities and genetic parameters. These will be discussed later (see 2.1 and 2.3).
The bulk of the Merino clip comes from the ewe flock. The selection of high producing ewes and the method of identification become a problem in group breeding schemes (Erasmus et al 1984). Selecting ewes which are able to not only lamb twins, but can also rear them while maintaining a high level of wool production, was preferable. In a preliminary investigation Erasmus et al (1984) found that a usable amount of variation in both mass of lambs weaned and amount of wool produced after weaning a lamb were available. A pilot scheme for ewe productivity (mass of lamb weaned + 3 x amount of greasy fleece mass), as it is called, was implemented by the SA Fleece Testing Centre and the method is fully described by Delport (1989). To test this pilot scheme, data from the Carnarvon control flock were used.
Delport (1989) investigated the whole concept of ewe productivity in the Carnarvon selection flocks and made the following important conclusions:
- Ewe productivity may lead to the exploitation of exciting new possibilities as an aid to identify "elite" ewes for the application of MOET (multiple ovulation and embryo transplant) technology.
- One record of total mass of lamb weaned is not a reliable basis for accurate predictions of ewe productivity.
- Selection of dams on the basis of breeding value predictions of ewe productivity may be fairly accurate, but the possibility of accurate selection of sires seems not so promising.
The ability of a ewe to lamb and raise a lamb in any given environment is proof of her adaptation to the specific environment and is a sound principle to keep in mind in any selection programme.
1.4 Tarka two-way selection flocks
Also in 1962 Dr J.E. Nel started a two-way selection experiment for quality (evenness of crimp, softness of handling) and the amount of wool yolk (traits were scored subjectively) at the Tarka Conservation area near Hofmeyr in the Cape Province. Each flock consisted of 100 ewes (400 ewes in total) and the experiment was suspended during 1983. Wilke (1974) did some preliminary analysis on these experiments and the results were as follow:
After nine years of selection, significant differences were found in the yolk scores between the yolk-plus and yolk-minus flocks and some indications that a se- lection limit was obtained in the yolk-minus flock. The percentage of wool yolk (objectively measured) in 1971 was 65,6 in the yolk-plus flock and 23,2 in the yolk-minus flock. Furthermore Wilke processed wool taken during the 1971 clip from each selection line. The processing characteristics of the wool of these two flocks did, however, not differ significantly. The processing ability of the wool was influenced by mainly fibre diameter, whilst the amount of wool yolk played a minor role (Wilke, 1974).
The response in the quality-plus flock was much slower than in the quality-minus flock and was as: ascribed (Wilke, 1974) to selection for quality in the base population: The scope for selection in the quality-minus flock was therefore bigger than in the quality-plus flock (Wilke, 1974). The average score after nine years of selection for the quality-minus flock was 2,7 and 4,20 for the quality-plus flock. As in the case with wool yolk, the processing abilities of wool from the two quality flocks were overshadowed by fibre diameter (Wilke, 1974).
This result suggested that quality as well as wool yolk had heritable factors and can be changed by selection. The correlation of these traits with the production traits will be discussed later.
1.5 Tygerhoek selection flocks
One of the most important and well designed selection experiments was initiated by Dr H.J. Heydenrych in 1969 at the Tygerhoek Experimental Farm of the Dept. of Agriculture near Riviersonderend in the South Western Districts of the Cape Province. The aim of this experiment was to investigate, firstly, selection for a wider secondary/primary follicle ratio (SIP ratio) at three months of age as an indirect selection method for clean fleece mass and, secondly, to arrived at genetic and phenotypic parameters and the influence of non-genetic factors on production traits (Heydenrych, 1975)
As far as the first objective of this experiment is concerned, Heydenrych et al (1984) summarized the results as follows: "The clean fleece weight of Merino sheep can be effectively increased by direct selection for this trait or by indirect selection for SIP ratio at three months of age. Although selection for SIP ratio has the advantage that it can be completed before weaning age, it is expensive and laborious. Furthermore it is not as effective as direct selection and apparently losses its efficiency with a decreasing nutritional plane".
Results on the second objective of this experiment were discussed fully by Heydenrych (1975) and is given at a later stage (see 2.1 to 2.3). Later Cloete & Heydenrych (1986) investigated the phenotypic relationship between 18-month mating mass and average lifetime reproductive performance and found a positive correlation between mating mass and reproduction rate. Heritabilities, genetic correlations and repeatabilities of reproduction rate in the Tygerhoek flocks were reported by Cloete & Heydenrych (1987a, 1987b and 1987c). From these results they suggested that selection for number of lambs weaned/ewe mated appears to be feasible in practice (Cloete & Heydenrych, 1987a). Genetic correlations of -0,25 to 0,28 were found between clean fleece mass and lambs weaned/ewe mated and low positive correlations between fibre diameter and lambs weaned/ewe mated (Cloete & Heydenrych, 1987b). They also found positive genetic correlations ranging from 0,08 to 0,67 between 18-month live mass and lambs weaned/ewe mated.
The repeatability of reproduction after first and second lambing opportunities in the Tygerhoek flocks is fully reported by Cloete & Heydenrych (1987c) and the utilization of modern techniques to accelerate reproduction rate in the current flock and in future generations by MOET techniques are discussed.
The control flock at Tygerhoek is also used to determine the genetic differences between various flocks and studs in the Merino industry. The method used is -discussed by Erasmus (1976) and results of the implementation of these methods in different flocks is given by Poggenpoel & Van der Merwe. (1977). Poggenpoel & Van der Merwe (1987) described how the control flock was used to estimate genetic response in three commercial Merino flocks.
The control flock and the flock selected for clean fleece mass are still maintained at Tygerhoek. However, the flock selected for Sip ratio was divided into two flocks during 1986; the one selected for fertility and the other "against" fertility (SWP. Cloete, 1989- pers. comm.). Fertility in this experiment is defined as number of lambs weaned per lambing opportunity. No results on these selection lines are presently available. 1.6 Other The University of the Orange Free State also run a fertility Merino flock since 1984. All base animals were collected out of the industry as twins or from highly fertile ewes (G J Erasmus, pers, comm, 1989). No results of this flock are presently available. Seven Baroola Merino rams were imported from Australia during 1984 (J.C. Greeff, 1989 pers. comm) and are used to establish a nucleus homosigotic Baroola flock from Merino base ewes. This work is being done at the experimental farm of the Animal and Dairy Science Research Institute near Ermelo and results will soon be available.
Parameters presented include heritabilities, genetic and phenotypic correlations between the various traits.
The heritability can be defined as the proportion of the phenotypic variance which is carried over to the progeny. From this it is clear that, if a trait is highly heritable, the phenotypic value gives a good indication of an animal's breeding value. On the other hand, f if a trait has a low heritability the phenotypic value is a poor indicator of an animal's breeding value and additional information (parents, co-lateral relatives, progeny results) is needed for more accurate breeding value prediction. Heritability is expressed as a proportion and the following rules can be used in interpreting heritability estimates:
Low heritability 0 - 0,20
Medium heritability 0,21 - 0,40
High heritability 0,41 - 1,00
The heritability of a trait determines the method of selection and the rate of improvement in the trait possible by selective breeding. Heritability estimates for a large number of traits has been done and those on the Merino in South Africa are summarized in Table 1 (Page 19).
The heritabilities of body mass traits increased with age. All traits younger than weaning age have low to moderate heritability. Two-tooth mass, on the other hand, is moderately too highly heritable.
Most of the economically important wool traits had moderate heritabilities. Except for greasy' fleece mass and crimp frequency, all the estimates were in a relatively narrow range.
For the subjectively scored traits, relatively few estimates were available and were mostly from the A-and B-studs and the two way selection experiments for quality and wool yolk at Tarka. The heritabilities for these traits varied from low (density score) to high (yolk score).
The estimates for reproduction traits were low. Estimates for the Tygerhoek flocks and Carnarvon flocks were available. The variation of the range was due to estimations from different sets of data of the same flock (Cloete et aI, 1987 and Delport, 1989)
2.2. Phenotypic correlations
A phenotypic correlation gives an indication of the relationship (positive or negative) between two traits of the same individual. Phenotypic correlation can be subdivided into genetic correlations (genes that affect two traits simultaneously) and the environmental correlation (traits influenced by the same differences in environmental conditions). Correlations are expressed as a fraction and normally the following rules hold:
Low correlation 0,20 - 0,40
Medium correlation 0,40 - 0,60
High correlation 0,60 - 1,00
Owing to the lack of a substantial number of estimated environmental correlations, the phenotypic correlations are presented in this article. To avoid a massive number of correlations, only the correlations of economically important traits with various other traits are presented in Table 2 (Page 20).
Correlations for most of the traits are low to medium and no big deviations from similar estimations on the Merino in other parts of the world are observed.
2.3. Genetic correlations
As previously mentioned, the genetic correlations are brought about mainly by pleiotropy (the property of a gene (genes) to affect two or more traits). Genetic correlations are also caused by linkage of genes that influence two traits. The correlations between traits caused by linkage can appear or disappear after some generations of selection.
High genetic correlations between two traits mean that selection for one trait will result in change in the correlated trait. It is therefore important to know the magnitude of genetic correlations between traits directly selected for and other economically important traits.
As for phenotypic correlations, only the correlations of the economically important traits with various other traits are listed in Table 3 (Page 21).
3. Applied research in the industry
As in the case with other farm animals, more than one trait in the Merino is considered for selection. Van der Merwe & Poggenpoel (1975) presented a selection index which combined four traits, namely body mass, clean fleece mass, fibre diameter and pleat score in a selection index. Currently selection indices are still used to various extents in ranking animals according to economic merit.
Central testing of different flocks is a method used world wide to determine the breeding value of different flocks and animals. The show for measured performance (Grove, 1987) was the first show of Merino's based on the principle of rearing all animals under standardized environmental conditions up to the same age. Owing to costs, only a limited number of studs (flocks) and animals can, however, be tested. From the show for measured performance, the so called veld ram clubs were established (Du Toit, 1988) in which rams from different breeders are reared from approximately weaning age to 14-months in standardized environmental conditions. The number of animals that can be tested has greatly increased but pre-test environment and the length of the testing period are problems that need to be eliminated.
Accurate identification of animals and parents in any breeding program is important. During 1941, Bosman, (1941) reported on the nose prints of Merino sheep as a method of identification. More than 2 000 sheep were involved and not two sheep with the same nose print had been found. Currently sheep are identified mainly by ear tags, tattoos and in the future by micro chips.
Botha (1947) investigated the fleece characteristics of wool on the pleats and between the pleats. Wool on the pleats had a 10% lower spinning count, had 57% less crimps per 25 mm and did not differ significantly in length and fibre diameter. The relationship between frontal development and various traits was investigated by Louw, Wessels & Labuscagne (195O). They found that frontal development had a phenotypic correlation of 0,34 with wool production, -0,30 with wool length, 0,07 with fleece wool, 0,31 with backs, 0,04 with short bellies and points, 0,19 with lox, -0,19 with body condition score and 0,07 with body mass.
The relatively high prices currently being paid for finer wool catch the attention. Delport (1989) found that fibre diameter of rams from the Karoo Veld Merino Club increased from 23,15 to 24,26 micron from 11 months to 18 months of age. The repeatability of fibre diameter was 0,65. The Department of Agriculture also recently established a genetic fine wool flock at the Cradock Experimental Farm (Olivier et al 1989). Material from this flock will be used to determine the production potential of genetic fine woolled animals under various environments.
The breed structure of the South African Merino was investigated during 1977 by Erasmus (1977) and it tended towards a two-tier structure rather than the classical three-tier pyramid. Most studs are more dependent on within-flock selection of rams for breeding improvement than on purchased rams. The average number of stud ewes per breeder stabilized at approximately 230 since 1966 (Erasmus, 1977).
The factors determining the price of wool have a important role to play in establishing selection objectives. Erasmus & Delport (1987) analyzed a sample of the 1984/85 wool clip and found that clean yield percentage and fibre diameter are by far the most important physical properties affecting the price of wool.
Some work is currently being done by the Animal and Dairy Research Institute on the genetic and physiological basis of early maturing characteristics of the Merino and other Merino strains. (J.C. Greeff, 1989- pers. comm). No published results are as yet available.
During the thirties and forties some experiments were conducted to improve the mutton production capabilities of the Merino. A need for this originated from the over-production of mutton in South Africa and the favourable prices for mutton during that time in England (Schuurman, 1932). Crossbreed experiments with Merino ewes and mainly British mutton breed sires, were undertaken in the Winter Rainfall area (Elsenburg -Johnston, G.W. & Bartel, L.R., 1932), the Karoo (Grootfontein- Mare, G.S., 1934) and Eastern Transvaal (Ermelo - Roux, L.L., 1935). During the sixties Erasmus (1969) also undertook crossbred experiments with Merino ewes and White Woolled mutton breeds under veld conditions in the Karoo.
Hofmeyr (1982) described the design and results of an crossbreed experiment in a high rainfall grassland region to investigate intensive sheep-production systems involving the production of wool and meat, using the Merino as foundation stock. For the first phase more than 2000 Merino ewes were mated to rams of various breeds. Hofmeyr (1982) summarized the crossbreeding work done over 60 years in South Africa and made some recommendations for future research to this regard.
Some extensive studies have been undertaken on various aspects of Merino breeding and the fact that only a small part of these studies is mentioned is by no means intended as a disparagement to the important work done. The South African Merino breeder and flock farmer have an enormous amount of available information to increase his own efficiency, but more important, to increase the overall efficiency of wool production in South Africa.
ANON., 1957. Score card for Merino Sheep. Dept. Agric. Pamphlet nr 356, Pretoria
ANON., 1961. Judging Merino Sheep. The Wool Grower, March: 10
ANON., 1986. Abstract of Agriculture Statistics. Directorate of Agricultural Economic Trends, Pretoria: 70
BONSMA, H.C. & DE VRIES, A.H., 1943. Sheep blowfly combating - Selecting and breeding for less susceptible Merino types. Farming in South Africa., 18: 581 (Afrikaans)
BOTHA, M.L., 1947. The Merino with body pleats and the evenness of its fleece. Farming in South Africa., 22: 467(Afrikaans)
BOSMAN, V., 1941. The recording of Merino sheep by using nose prints. Farming in South Africa., 16: 45 (Afrikaans)
CLOETE, S.W.P., HEYDENRYCH, H.J. & DU PLESSIS, J.J., 1984. The repeatability of reproduction characteristics in the Tygerhoek Merino flock - preliminary observations. S. Afr. J. Anim. Sci., 14: 29
CLOETE, S.W.P., HEYDENRYCH, H.J., 1986. A note on the phenotypic relationship between 18- month mating mass and average lifetime reproductive performance of Merino ewes in the Tygerhoek flock. S. Afr. J. Anim. Sci., 16: 97
CLOETE, S.W.P. & HEYDENRYCH, H.J., 1987a. Genetic parameters for reproduction rate in the Tygerhoek Merino flock. 1. Heritability. S. Afr. J. Anim. Sci., 17: 1
CLOETE, S.W.P. & HEYDENRYCH, H.J., 1987b. Genetic parameters for reproduction rate in the Tygerhoek Merino flock. 2. Genetic correlations with wool and live mass traits. S. Afr. J. Anim. Sci., 17: 8
DELPORT, G.J., 1989. A mixed model approach for selecting Merino ewes. PhD-dissertation. Univ. OFS, Bloemfontein
DEL PORT, G.J., 1989. The reliability of performance measuring. Spring Journal 1989, Karoo Veld Merino Club: 7 (Afrikaans)
DE KLERK, C.H., DOVEL, G.H. & TERBLANCHE, E. Ie F., 1983. An investigation to sheep farming in the Republic of South Africa. S.A. Institute for Extension, Univ. Pretoria, Pretoria (Afrikaans)
DE VRIES, A.H, 1943. Susceptibility of Merino sheep for blowfly contamination. Farming in South Africa., 18: 493 (Afrikaans)
DU TOIT, J., 1988. Veld ram clubs become popular. Golden Fleece, 17 (12):10 (Afrikaans)
ERASMUS, G.J., 1976. Sheep breeding programmes in South Africa. In: Sheep Breeding. Ed. by G.J. Tomes, D.E. Robertson and R.J. Lightfoot. Western Australian Institute of Technology, Perth: 40
ERASMUS, G.J., 1977. The breed structure of the South African Merino. MSc Agric-treatise. Univ. OFS., Bloemfontein
ERASMUS, G.J., 1981. Development of performance testing of woolled sheep in South Africa. Proc. Memorial Symp. in honour of the late Prof. F.X. Laubscher. 14 November 1981, Stellenbosch: 184 (Afrikaans)
ERASMUS, G.J., 1988. A mixed model analysis of a selection experiment with Merino sheep in an arid environment. PhD-dissertation. Univ. OFS, Bloemfontein
ERASMUS, G.J., DE LANGE, A.O. & DELPORT, G.J., 1984. Selecting Merino ewe replacements – The relationship between fleece weight and total weight of lamb weaned. In: Proc. World Congress Sheep and Beef Cattle Breeding. Ed. J .h. Hofmeyr and E.H.H. Meyer, Pretoria:-
ERASMUS, G.J. & DELPORT, G.J., 1985. A critical evaluation of selection criteria in woolled sheep. Karoo Agric, 3(5) :4 (Afrikaans)
ERASMUS, G.J. & DELPORT, GJ., 1987. Factors influencing the price of greasy fleece wool in South Africa. S. Afr. J. Anim. Sci., 17:111
ERASMUS, L.S., 1969. Slaughter lamb production with the Merino as mother breed. Farming in South Africa., 45(7): 17
GROVé, A.C., 1987. The evaluation of Merinos from different studs in a standardized environment. MSc Agric-treatise, Univ OFS, Bloemfontein (Afrikaans)
HEYDENRYCH, H.J., 1975. A study of flock statistics, non-genetic factors, genetic parameters and selection response in the Tygerhoek Merino flock. PhD- dissertation. Univ. Stellenbosch, Stellenbosch (Afrikaans)
HEYDENRYCH, H.J., DU PLESSIS, J.J. & CLOETE, S.W.P., 1984. Increasing the wool production of Merino sheep in the South Western Districts of South Africa by direct and indirect selection. In: Proc. World Congress Sheep and Beef Cattle Breeding. Ed. J.H. Hofmeyr and E.H.H. Meyer, Pretoria
HOFMEYR, J. H., 1982. Implications of experimental results of crossbreeding sheep in the Republic of South Africa. In: Proc. World Congress on Sheep and Beef cattle breeding. Ed. by R.A. Barton and W.C. Smith, Dunmore Press Limited, Palmers ton North, New Zealand
HUNLUN, 1988. The use of mixed model methodology in stock improvement. Technical communication No 207, Dept. Agriculture and Water Supply, Pretoria (Afrikaans)
SCHUURMAN, G.J., 1932. Crossbreeding and fat-lamb production. Farming in South Africa., 7: 165
LOUW,D.J., WESSELS, N.G., LABUSCAGNE, F. J. & BADENHORST, D.F., 1950. The relationship between Merino traits. Farming in South Africa., 25:363 (Afrikaans)
MARE, G.S., 1934. Crossbred lambs on Karooveld. Farming in South Africa., 9: 64
NEL, J.E., 1967. The influence of flock composition on production and reproduction characteristics of Merino Sheep. PhD-dissertation. Univ. Stellenbosch, Stellenbosch (Afrikaans)
NEL, J.E., 1970. Subjective judging and the prediction of breeding value in Merino sheep. Agroanimalia, 2: 145 (Afrikaans)
OLIVIER, J.J., 1980. The influence of objective and subjective selection methods and environmental factors on production and reproduction characteristics in Merino sheep at the Carnarvon experimental farm. MSc Agric-treatise, Univ. Stellenbosch, Stellenbosch. (Afrikaans)
OLIVIER, J.J., 1983. Research at Carnarvon experimental station. Mimeograph. Dept. Agriculture and Water Supply, Karoo Region, Middelburg, C.P. (Afrikaans)
OLIVIER, J.J., 1989. A linear scoring card for quantifying non-measurable traits of Merino sheep. Merino Journal, Vol. LII no 1: 51 (Afrikaans)
OLIVIER, J.1., DELPORT, G.J., ERASMUS, G.J. & EKSTEEN, T.J., 1987. Linear type scoring in Merino sheep. Karoo Agric, 3(9):1.
OLIVIER, J.J., STEYNBERG, H. & WENTZEL, D., 1989. The establishment of a genetic fine wool flock. Merino Journal, Vol. LII no 1: 49 (Afrikaans)
POGGENPOEL, D.G., HART, M. & LUND, A., 1984. The repeatability of the reproduction rate of commercial Merino ewes. S. Afr. J. Anim. Sci., 14:40 (Afrikaans)
POGGENPOEL, D.G. & V AN DER MERWE, C.A., 1975. The (use of selection indices for Merino sheep. S. Afr. J. Anim. Sci., 5: 249 (Afrikaans)
POGGENPOEL, D.G. & VAN DER MER WE, C.A., 1987. Selection response with index selection in three commercial Merino flocks. S. Afr. J. Anim. Sci., 17:70
POGGENPOEL, D.G., 1989. The relationship between performance measurements and auction prices of Veld Merino rams. S. Afr. J. Anim. Sci., 19 : 50
ROUX, C.Z., 1961. Considerations for planning and execution of breeding plans suitable for woolled sheep. MSc Agric- treatise, Univ. Stellenbosch, Stellenbosch. (Afrikaans)
ROUX, L.L., 1935. Crossbred lambs under Eastern Transvaal conditions. Farming in South Africa., 10:147
V AN WYK, J.B., NEL, J.A., V AN DER SCHYFF, W. & ERASMUS, G.J., 1985. Phenotypic and genetic parameters in a commercial Merino flock. S. Afr. J. Anim. Sci., 15: 171
WARREN, E.N.S., 1933. The sheep of the Grootfontein Agricultural school. Farming in South Africa, 8:141 (Afrikaans)
WILKE, D.P., Simultaneous two way selection for wool yolk and wool quality in Merino sheep. MSc Agric-treatise, Univ. Orange Free State, Bloemfontein (Afrikaans)
Merino journal 53 (1)