Photo L/R : Gary Trethewey (BKB), Richard Morgan (back) and Murray Rogerson in discussion at the Sale of Robbie Blaine’s Wauldby Dohnes, Amabele
Murray Rogerson is a prominent Dohne Merino Breeder in the state of Victoria, Australia. Murray’s Stirling Dohne Merino stud is situated in the South Western part of Victoria at Glenthompson. Murray describes the area as a “soft damp area” with an average yearly rainfall of 540mm. Sheep have to handle “long wet grass” for approximately seven months of the year and therefore “strong feet” are not negotiable. Murray also mentioned that this particular area, similar to the Riverina, is well-known for producing some of the best wools in Australia. The Stirling stud was established circa 2004 by means of embryo transfers as well as 490 stud ewes purchased from the Uardry Dohne Merino stud. Stud Sires were obtained from almost exclusively the Uardry stud. The Stirling stud already in 2010 sold more than 100 rams and its 7th annual on-property ram sale, 50 rams, will take place on November 10th.
- The first important aspect influencing accurate evaluations by means of EBV’s, that cannot be over emphasized, is accurate data recording at farm level. The basic principle that inaccuracies affect performance testing figures, also applies to EBV calculation. Although the system of EBV calculation is more robust, Murray, in his presentation pointed out, cases of “unacceptable” changes in EBV’s. He also indicated that the adjustment of the base of comparison every year may lead to breeders confusing this base year change with an interpretation that the system does not account accurately for enviromental effects.
- The practice of recording birth dates inaccurately by using “average dates” for ewes lambing in some extreme cases even longer than a week apart, seriously affect the accuracy of weaning indexes and also weaning weight direct and maternal breeding values. The over and under estimation of growth rate may be as high as 30 percent according to Murray Rogerson. Murray therefore suggested a system of measuring total weight of lamb produced per unit of body weight for ewes as an alternative. It must however be remembered that the inherent inaccuracy of improper birth recording is not removed and the ewe productivity figures will remain subject to the same inaccuracy.
Photo : Dohne Merino Mothers together with their newly born lambs
- The practice of performing final performance testing at the age of 250 days with a minimum wool growth of five months, also seriously compromises accuracy of evaluation. According to Murray, he would rather suggest to performance test only ewes at an age of two years. In this instance, the worst case scenario is debatable. Having only half of a dataset that is accurate compared to complete but inaccurate data may have equally inadequate breeding value figures as a result. It is a well known scientific fact that the repeatability of measurements at an earlier age than two tooth stage, is questionable. The practice of measuring wool traits on a short period of wool growth is also questionable and leads to excessively large coefficients of variation in national datasets compared to experimental data.
Photo : Excellent Australian Dohne Merino wool!
- In the Australian case, EBV’s are calculated for number of lambs weaned. In the conversation with Murray it became clear that the result of this calculation was implemented prematurely. Sufficient “depth” of pedigree data is necessary to produce usable and more reliable EBV’s for this trait with an extremely low heritability. The problem in Australia seems to be that breeders are losing their confidence in these figures that were not reliable at the time of implementation. Another more positive approach of breeders would be to strive for an improvement in the reliability of these most necessary figures that are in any case of more value than phenotypic measurements only.
- Murray as an individual farmer, did not have access to the entire Australian data base and therefore could not comment on the amount of linage in the Australian case. However it may be sufficient to conclude that imperfections in across year and across stud genetic links may also contribute to unsatisfactory end results. In South Africa, breeders are in some cases, not following the linkage guidelines properly and statistically totally unbalanced data causes breeding values to be not comparable.
The bottom line is that the system usually gets blamed instead of the data! Resorting back to a less sophisticated methods is definitely not the solution to this international problem. Performance ratios are less robust and are therefore more affected by the discrepancies mentioned by Murray Rogerson. The only difference is that performance ratios “hide” the reality of erroneous data from breeders. Breeders resorting back to these older inadequate techniques of animal breeding can, tongue in the cheek, be compared to an ostrich hiding its head in the sand. Animal scientists and progressive breeders should actually encourage sound recording practices to facilitate more accurate and more complete comparisons between animals. The trick for animal breeders nowadays, seems to be in the long run, to follow breeding objectives producing the highest possible commercial profit by using as few as possible, but accurate measurements.
Photo L/R : Murray Rogerson and Quentin Rogers (back in South Africa from New Zeeland after judging the world shearing championships) relaxing after the Wauldby Dohne Merino sale