- Quantifying type traits in Merinos
|Last update: August 17, 2011 03:40:45 PM|
Quantifying type traits in Merinos
J. J. Olivier
Grootfontein College of Agriculture, Middelburg CP 5900
In most Merino studs, type (subjective) traits play an important role in selection and buying of rams. This is supported by an analysis of data from three Merino veld ram clubs by Poggenpoel & Van der Merwe (1988). They found that a selection index, which is a combination of the economically important traits, accounted for only 23 % of the total variation in auction price. Even in the dairy industry of the United States of America, type traits are still of importance to the commercial producers (Wilder & Van Vleck, 1988).
Considering too many traits during selection can seriously hamper the progress in economically important traits. It is therefore of some importance to quantify these type traits and to determine their hereditary base and their correlations with economically important traits. Owing to the popularity of AI and the increase in the use of frozen semen in recent years, it becomes even more important to quantify these non-production traits in an attempt to detect faults a ram may breed and to prevent their spreading through the industry.
The first score cards for type traits were drawn up as early as 1957 (Anon, 1957) but had major shortcomings (Roux, 1961 and Nel, 1970). Scorecards should be designed in such a way as to measure traits as accurately as possible. Here are some requirements that should be met:
- To increase accuracy, a specific trait, rather than a combination of traits, should be scored.
- A trait must be scored over a wide point scale (1 to 50) to ensure a normal distribution and must be scored from one extreme to the other.
- The variation in a trait must be clearly recognizable.
Since 1985 different score cards for the Merino, which met these requirements, have been proposed and used on a lin1ited scale. The first score card dealt with (Olivier et al, 1987) had as many as 32 traits, which were lately reduced to 21 traits (Olivier, 1987). A major shortcoming of these two score cards was the time needed to score all the traits. In addition, some conformation traits could only be scored after the sheep had been shorn, which means that wool and conformation traits have to be scored at different times. To overcome these problems and to maintain a reasonable level of accuracy, some traits have since been combined on account of the medium to high correlations between them, whereas some of the less important traits have been dropped. This resulted in a final score card consisting of 11 traits (Table 1). The 11 traits are scored over a range of 1 to 50 with the following intervals:
1 - 10 Poor
11 - 20 Below average
21 - 30 Average
31 - 40 Above average
41 - 50 Excellent
Wool yolk and hocks are, however, exceptions with the ideal score being 25 with deviations from the ideal to either below or above 25.
Ideally the scoring should be done with a wool growth of between 5 and 8 months. Before commencement of scoring, a standard of all traits within the specific group to be scored should be set. A few animals over the entire scoring scale should be scored and kept as a reference throughout the scoring session.
The scorecard should be applied mainly for progeny testing, but can also be used for the description of individual animals or flocks, corrective mating, assessing trends in these traits as well as determining genetic and phenotypic correlations between these traits and other production traits.
This scorecard has been tested extensively for 2 years in the Grootfontein Merino stud, as well as in the Grootfontein Fine Wool stud. Approximately 400 animals can be scored within an 8-hour working day. Most of the traits were scored over the full range, except for traits like colour (owing to the absence of colour), variation in crimp number over the fleece (owing to little variation) and wool yolk.
The scorecard has many applications. It can be used in a progeny-testing programme as a means to identify the possible weaknesses a ram may breed. The fact that a specific trait of an animal can be quantified makes it possible to compare the animal with its contemporaries. Further, the use of pedigree information will improve the accuracy of breeding value prediction for type traits. Pedigrees will also be of greater value, because each ancestor will have a numerical value for a specific trait (Fig 1). The heritability and genetic correlations of type traits and between type and production traits can also be estimated. This will allow the development of efficient breeding plans.
The contributions of Dr G. J. Delport, T. Eksteen, D. Grove, S. L. Vorster and L. Zandberg in the finalising and testing of this scorecard are greatly appreciated.
ANON, 1957. Score card for Merino Sheep. Dept. Agric. Pamphlet Nr 365. Pretoria.
NEL, J.E., 1970. Subjective judging and the prediction of breeding value in Merino sheep. Agroanimalia, 2:145.
OLIVIER, JJ., 1989. Lineêre punteskaal vir kwantifisering van nie-meetbare eienskappe by Merinoskape. Merino Journal, Vol. LII (nr 1):51.
OLIVIER, JJ., DELPORT, GJ., ERASMUS, GJ. & ERASMUS, TJ., 1987. Linear type scoring in Merino sheep. Karoo Agric, 3(9):1.
POGGENPOEL, D.G. & VAN DER MERWE, C.A., 1988. The correlation between measurement of production characteristics and sale prices of veld Merino rams. Merino Breeders' Journal, Vol LI (no 1):44.
ROUX, C.Z, 1961. Considerations for planning and execution of breeding plans suitable for woolled sheep. MSc. Agric treatise, Univ. Stellenbosch, Stellenbosch.
WILDER, J.S. & VAN VLECK, L.D., 1988. Relative economic values assigned to milk, fat test and type in pricing of bull semen. J. Dairy Sci., 71:492.
Karoo Agric, Vol 4, No 3, 1991 (3-5)