Effect of multiple births on reproduction rate in a control flock.

Abstract

Proc. Aust. Soc. Anim. Prod. (1974) 10; 254 EFFECT OF MULTIPLE BIRTHS ON REPRODUCTION RATE IN A CONTROL FLOCK J.R.W. WALKLEY* and C.H.S. DOLLING* Summary A positive selection differential, in number of lambs born per ewe joined, has been demonstrated in a hypothetical control flock in which male and female replacements are chosen at random. Annual genetic gains of 0.0129 and 0.0103 have been calculated for control flocks with 30% and 157o multiple births in which 16.7% of the ewes and half of the rams are replaced each year. The increase in repr oduction rate observed in a c ontrol flock of medium Peppin Me rinos is di scuss ed. I. INTRODUCTION The possible sources of error in the use of a control group to estimate the genetic change in a selected population have been reviewed by Hill (1972). One source is a genetic trend in the control due to natural selection, Natural selection, opera.ting through differences in reproduction rate among the parent individuals or in viability among their offspring, may causg changes in gene frequency between the parent individuals and the offspring generation (Falconer 1961). If natural selection should operate in a control population by virtue of differences in number of lambs born among parent individuals, this could be a problem in the interpretation of a selection experiment in which number of la,mbs born is either under selection itself or is under observation for a correlated response with another character which is under selection. here. A method of compensating for this effect of natural selection is described II. MATERIALS AND METHODS (a) Estimation of Natural Selection Young and Turner (1965) gave theoretical consideration to the improvements in both number of lambs born and clean wool weight per head which might be expected from a number of selection schemes in Australian Merino sheep. In each of the schemes, both rams and ewes were selected on various criteria relating to reproduction or wool weight or both. We have made the same assumptions as Young and Turner (1965), but all replacement rams and ewes have been selected at random. The assumptions made are set out below. Ewes were 2-7 years of age at lambing, and 12% bore no lamb. At the two levels of twinning (307o and Is%), the numbers of lambs born per ewe joined (%J ) were 1.18 and 1.03 respectively. Rams were j oined at 3% of the number of ewe s and were used in two consecutive year s. The rams were 2 and 3 year s old at the b irth of their offspring. * Department of Agriculture, Adelaide, South Australia. Of the single lambs 80% were assumed to survive until selection; twins, 60%. of the ,No deaths in adult she ep were allowed for; one sixth (16.7%) of the breeding ewe flock and 50% of the wo rking ram flock was replaced each year. (b) Avoiding Natural Selection The aim in avoiding natural selection is to establish a zero selection differential, in LBJ , for the rams and ewes chosen as replacements. Details of the calcula.tions are shown in the results. III. a RESULTS (a) Estimation of Natural Selection Let P be the proportion of the flock replaced each year, P the prop orti on of twins among the animals a. vzlilable at selection, LBJ the original lambs b orn per ewe joined. Then the selection differential in number of lambs born, that is, based on the dam's reproductive performance, is (i) 30% Multiple births . The percentages of ewes bearing 0, 1 a,nd 2 lambs are 12, 58 and 30 respectively. If all replacements are then selected at random, the expected percentages of twins and singles chosen would be 43.7 and 56.3. The percentages of ewes bearing 0, 1 and 2 lambs are 12, 73 and 15 respectively. We would then expect 23.6% of twins and 76.4% of singles to be selected at random. The selection differential, in LBJ , for the rams and ewes becomes (iii) Genetic gains The rate of genetic gain per generation in number of lambs born if ewes are chosen as twin-born, that is, if selection is based on the dam's reproduction rate, is 255 (b) Avoiding Natural Selection Given the flock replacement rate, the mean reproduction rate LBJ, rnd the number of sheep of each sex and birth type survivin>; to selection, the percentage of .. twin and single replacements required to achieve a zero selection differential can be calculated for a control flock. We have shown earlier that the selection differential IQ is 1 + p - Lg-7 where p is the proportion of twins a,mong the animals available at selection, The selection Thus with 30% multiple births, differential will be zero when p = 1. P = 1.18 - 1 = 0,18 and so 18% of the 16.7% ewe replacements required in the theoretical flock are twins and 82% are singles. Similarly for 15% multiple births, P = 1.03 - 1 = 0.03 and 3% of the 16.7% ewe replacements req-uired are twins and 97% are singles. The percentages of twin and single ram replacements needed are the same as the ewe percenta.ges required at each level of multiple births. %J - IV. DISCUSSION T-he reproduction rate in a control flock of medium Peppin Merinos (Turner, McKay end. @inane 1972) is characterised by an increase in the number of lambs born per ewe lambing (L& f or ewes from 3 to 8 years of age at lambing. The lambings covered were those from 1950 to 1966 inclusive; the increase in LOP was evident in 1959 and later years. The percenta ge chance of choosing a, twin fr om the weaners available each year has been calculat ed (Table 1); the 7 lambs born as triplets have been bulked with the twins in the 2,105 weaned lambs considered. It is seen from Table 1 thet there is z general increase in the probability that twin-born sheep would be chosen as replacementf in this control flock. Should, in fact, twins have been selected, the increase in BP described by Turner, McKay . and Guinane (1972) could have resulted from ns.tural selection for reproduction rate as described in this paper. 256 Differences in favour of singles have been observed by Dun and Grewal (1963) for weaning weight, 18 month body weight, and greasy fleece weight. Turner (1961) has shown a difference of 0.154 kg clean wool between handicapped (twins and the progeny of young ewes) and unhandicapped (singles, the progeny of adult ewes) sheep. Inclusion of more twins could, therefore, result in an over-estimation of the response to selection involving these characters, when measured as percentage deviations from the control, unless the numbers of twins and singles in the control flock and the flock under selection were the same or type of birth adjustments were made. The increase in described for a random control flock, would be expected to occur in commercial flocks. If this has not occurred, it could be because handicapped animals are generally not identified at selection, as a result of which they are selected against with a resultant lowering of the twinning rat& (Turner 1961). TABLE 1 Percentage chance of choosing a twin from weaned lambs available each year in the control flock of Turner, McKay and Guinane -(1972) V. ACKNOWLEDGEMENTS The advice and criticism of Ellen Bennett, Dr L. Bailey and Mr R. Polkinghorne, Department of Agriculture, South Atistralia, is gratefully acknowWe wish to thank Dr Helen Newton Turner of C.S.I.R.O. for making ledged. available data from the control flock. This study was undertaken as part of a project which is financed by the Wool Research Trust Fund. VI. REFERENCES DUN, R-B.' and GREWAL, R-S. (1963). Aust. J. exp. Agric. Anim. Husb. 2: 235. FALCONER, D-S- (1961). 'Introduction to Quantitative Genetics'. (Oliver and Boyd: Edinburgh and London). Anim. Breed. Abstr. 40: 193.% HILL, W-G. (1972). Aust. J. azic. Res. 12: 974. TURNER, Helen Newton (1961). Aust. J. TURNER, Helen Newton, MCKAY, Elaine, and GUINANE, Fay (1972). agric. Res. 23: 131. 'Quantitative Genetics in TURNER, Helen NewEn, and YOUNG, S-S-Y. (1969)(Macmillan of Australia: Melbourne). Sheep Breeding'. Aust. J. agric. Res. 16: 863. YOUNG, S.S.Y., and TURNER, Helen Newton (1965). YOUNG, S.S.Y., TURNER, Helen Newton, and DOLLING, C.H.S. (1963). Aust. J. agric. Res. 11: 604. 257