Australian Journal of Experimental Agriculture, 1988, 28, 167-71 167 Carcass composition of the South Australian Merino and its crosses with the Booroola and Trangie Fertility Merino D. 0. KleemannA, R. W PonzoniB, J. E. StaffordB and R. J. GrimsonA ASouth Australian Department of Agriculture, Turretfield Research Centre, Rosedale, S.A. 5350, Australia. BSouth Australian Department of Agriculture, Box 1671 G.P.O., Adelaide, S.A. 5001, Australia. Summary. South Australian medium-wool (M), non Peppin medium-wool Booroola (B) and Peppin medium-wool Trangie Fertility (TF) Merino rams were joined to Mewes at Turretfield Research Centre, South Australia, in 2 years. Carcass composition was assessed in the ewe and wether progeny at 2 mean slaughter liveweights, viz. 24 and 38 kg. When adjusted to the same carcass weight, BxM had 13% more carcass chemical fat, 15% more subcutaneous fat, 6% less bone and the same lean tissue as M. The same result was observed for TFxM in relation to Min year 2. However, TFxM had more lean and the same amount of subcutaneous and chemical fat as Min year 1. Within the Booroola strain, there were no differences between offspring from 3 sires with genotype FF and the 1 sire with + + for any of 5 variables analysed. We conclude that crossing the Booroola with the South Australian Merino produces carcasses with the same amount of lean tissue, less bone and more fat when compared at the same carcass weight. The rank of TFxM with the other strains for the major carcass components remains obscure owing to a strainxyear interaction. Introduction The Australian Merino flock contributes significantly to the local and overseas meat trade through sales of mature and of surplus young Merino sheep, as well as forming an integral part of the prime-lamb industry. Incorporation of highly fecund Merino sheep, the medium-wool non-peppin Booroola (Turner 1978) and medium-wool Peppin Trangie Fertility (Atkins and Robards 1976), into South Australian Merino flocks to improve reproductive performance, has been evaluated by Walkley et al. (1982) and Ponzoni et al. (1984). Strong-wool and medium-wool Merino strains developed in South Australia predominate in that State, Western Australia, and in western New South Wales, and are found to a lesser extent in eastern parts of Australia (Roberts et al. 1975). These strains are larger at maturity than their Peppin and non-peppin counterparts and are popular in some areas as prime-lamb mothers when mated to terminal meat-sire breeds. Kleemann et al. (1985) have reported that the South Australian Merino has a significantly smaller fat depth at both the same age and the same carcass weight than its crosses with the Booroola and ' Trangie Fertility. However, although fat depth is related to consumer acceptability (Southam and Field 1969), it gives limited information on only 1 tissue component of the carcass. This paper reports the carcass composition of the South Australian Merino (M), BooroolaxM (BxM) and Trangie FertilityxM (TFxM) strains during the pre-fattening (24 kg slaughter liveweight) and fattening (38 kg slaughter liveweight) phases of growth described by Searle et al. (1972). Since this study was initiated, it has been established that inheritance offecundity in the Booroola is by a single gene (or a closely linked group of genes) of major effect (Piper et al. 1984). An attempt was made, therefore, to examine the effect of the fecundity gene (F) on carcass composition. Materials and methods Experimental design and mating procedure Details of the location and environment, experimental design and ewe and lamb management have been reported in Kleemann et al. (1985). A summary of the experimental design and the number of sheep involved is given in Table 1. A flock of South Australian medium-wool Merino (M) ewes of a Collinsville-type strain (Ashrose) was used as a base population. The M flock was mated to M, Booroola (B) and Trangie Fertility (TF) rams in the October November periods of 1978 and 1979 at Turretfield Research Centre, SA. After stratification on the basis of age and liveweight, base flock ewes were randomly
168 D. 0. Kleemann et al. Table 1. Experimental design M, South Australian Merino; B, Booroola; TF, Trangie Fertility. Numbers of animals are given in parentheses Sire Dam Progeny Year of birth, 1979 M (5) M (50) M (39) B (5) M (50) BxM (52) TF (5) M (50) TFxM (39) Year of birth, 1980 M (5) M (66) M (44) B (5)A M (66) BxM (35) TF (5) M (66) TFxM (48) AQne ram failed to sire progeny. allocated to the sire strains in each year. Different M, B and TF rams were used each year. Individual rams were mated to a group of base ewes. Five rams of each strain were used each year. In the second year, 1 B ram failed to sire progeny. At the end of the 6-week mating period, ewes that had been joined to individ~al sires were combined and grazed as a single flock until shortly before lambing. The ewes were then drafted into strain-of-ram groups and grazed in separate paddocks. Approximately 6 weeks after lambing began, the groups of ewes and lambs were combined and run as 1 flock. The lambs were weaned at an average age of 84 days and were grazed as 1 flock until slaughter. Each year, lambs were allocated to 2 groups to be slaughtered at 24 kg (4-5 months ofage) and 38 kg (14-15 months of age) mean liveweight. Within strains, slaughter groups were balanced for weight taken prior to the first slaughter, age, birth type (single, twin), sex (male castrate, female) and sire. Measurements The experimental progeny were fasted for approximately 18 h, and slaughtered and dressed at Turretfield Research Centre using normal commercial procedures. The carcasses were jointed between the 12th and 13th ribs. The vertical surface of the vertebral column was carefully marked with the sharp edge of a knife to assist in the accurate halving of the carcass. The left side of the carcass was stored in a plastic bag at - 15'C until required for dissection. The stored carcasses were thawed for 1-2 days at 2-4'C. The half-carcasses were weighed and dissected into subcutaneous fat, muscle including intermuscular fat (defined here as lean), bone and trimmings (including tendons in the lower, fore and hindlimbs, and ligamentum nuchae) and weighed. Subcutaneous fat was defined as fat overlying all the tissues, excluding M. cutaneous trunci (which lies in the subcutaneous fat). The bone component included cartilage and small quantities of muscle, fat and other tissues that were physically difficult to separate. The half carcass weight or side weight was defined as the sum of subcutaneous fat, lean, bone and trimming weights. Percentage weight losses (mean± s.d.) during dissection were 0 39 ± 0 26 and 0 18 ± 0 17 for the first and second slaughter groups respectively. The soft tissue of the carcass, comprising subcutaneous fat and lean (muscle plus intermuscular fat), was thoroughly mixed and then minced according to the method described by Walker et al. (1971). The mince was sampled and stored at -15'C. Percentage moisture and ether extractable fat (carcass chemical fat) were determined by the following procedures. A 100 g subsample of mince was macerated with 20 ml of water in a homogeniser. A 5 g homogenised subsample was dried at 70'C in a vacuum oven (20 mm Hg). The dried sample was extracted with diethyl ether for 5 h, dried for 1 hat IOO'C, cooled and weighed. Statistical analyses The variables listed in Table 2 were analysed by least-squares analysis of variance using the GLM procedure in SAS (Freund and Littell 1981). The basic model fitted was: J'ijklm110= µ +st;+ Jlfj+ Sijk+sg1+sxm + t,,+ (st.yr)ij + (st.sg)u + (st.sx);m + (sl.1); 11 + (yr.sg)j1+ (yr.sx)jm + (yr.f)j11 + (sg.sx)1m + (sg.t)111 + (sx.t)mn + eijklm110 where Jlijk/mno is the oth observation in the ith strain, }th year, kth sire, Ith slaughter group, mth sex, nth type of birth, µ is the common mean, and eijklmno is the random error. Strain, year, slaughter group, sex and type of birth were treated as fixed effects. The sire effect was nested within strain and year, and was treated as a random effect. The model fitted to side weight (half carcass weight) included day of birth within slaughter group as a linear covariate. All other characters included day of birth and side weight, within slaughter group, as linear covariates (Table 2). The mean square for sire within strain and year was used as the error term for testing the effects of strain and year. Since variances of the 2 slaughter groups were significantly different (P<0 05) for most characters, data were log 10 transformed prior to analysis. However, the results of this analysis and an analysis of the untransformed data were almost identical, and so for ease of presentation, the analysis of the untransformed data are presented. Results The results of the analysis of variance are presented in Table 2 and least-squares means in Table 3. With the
Carcass composition of the S. Aust. Merino 169 Table 2. Least-squares analysis of variance, with degrees of freedom and levels of statistical significance for side weight, subcutaneous fat, lean, bone, trimmings and carcass chemical fat Source of variation d.f. Side Subcutaneous LeanB BoneB TrimmingsB weighta fatb Strain (st) 2 * * ** Year of birth (y) I ** ** *** *** Sire within strain and year of birth 23 * ** ** *** Slaughter group (sg) I ** Sex (sx) I ** *** * *** *** Type of birth (I) I stxy 2 *** ** *** stxsg 2 stxsx 2 stxt 2 yxsg I *** *** *** yxsx I yxt I sgxsx I sgxt I ** * sxxt I Day of birth within sg 2 Side weight within sg 2 - *** *** *** *** Remainder 0 8146 0 0171 0 0311 0 0048 0 0001 d. f. of remainder 206 204 204 204 204 Carcass chemical fatb * * *** ** * *** 0 0443 204 *P<0 05; **P<O OI; ***P<O OOI;, not significant. ACovariate; day of birth within slaughter group. BCovariate; day of birth and side weight, within slaughter group. Table 3. Least-squares means± s.e. (g) for side weight (adjusted for day of birth), and subcutaneous fat, lean (muscle plus intermuscular fat), bone and carcass chemical fat (adjusted for day of birth and side weight) M, South Australian Merino; BxM, BooroolaxMerino; TFxM, Trangie FertilityxMerino. Within each variable and year, means not followed by a common letter differ significantly (P<0 05) Strain Side weight Subcutaneous fat Lean Bone Carcass chemical fat M 6498 ± 150 BxM 6462 ± 131 TFxM 7298 ± 154b Year of birth, 1979 654 ± 30 5039 ± 24 737 ± 30b 5017 ± 24 3 615±32 3 5105 ± 25b 1343 ± 16 1283 ± J7b 1316 ± J8 3 b 1111 ± 47 1246 ± 48b 1081 ± 50 M 7504±145 3 BxM 7006 ± 164b TFxM 7659 ± 144 3 Year of birth, 1980 593 ± 27 5185±21 694 ± 28b 5184 ± 22 720 ± 28b 5149 ± 22 1266 ± 15 3 1165±J5b 1177 ± J5b 1055 ± 42 1190 ± 44b 1213 ± 44b model fitted for the least-squares analysis of side weight, the mairi effect of slaughter group did not reach the 5% level of significance (P= 0 11). All main effects and all significant 2-way interactions, except those involving strain, were considered unimportant in relation to this study and are not commented on further. There were significant strainxyear interactions for a number of the characters measured (Table 2) and for this
170 D. 0. Kleemann et al. reason least-squares means of the strains are presented within years (Table 3). Examination of the means indicated that, for all characters, ranking of the Mand BxM strains did not change between years. There were some occasions when TFxM changed rank with the other strains. The M and Bx M strains had the same amount of lean but the BxM had less bone and more subcutaneous fat. The TFxM strain performed the same as BxM in relation to Min year 2. However, in year 1, TFxM had more lean and the same amount of fat (subcutaneous and chemical) as the M strain. The genotypes of 5 of the Booroola sires used in the present study were known (3 FF, 1 F +, 1 + + ). All of the known genotypes produced progeny in the 1 year (1979). An additional analysis of variance was conducted on all data from Booroola sires of known genotypes, including 'sire genotype' in the mathematical model. The effect of sire genotype was significant for all of the 5 variables analysed. Offspring from FF and + + sires did not differ significantly from each other for any of the variables, but were 15-20% below the corresponding value for offspring of the F + sire. However, after adjustment for side weight, differences between sire genotypes were not significant. Discussion This study has shown that important differences in carcass composition exist between the South Australian Merino and its cross with the Booroola Merino. For example, at the same carcass weight, BxM had 13% more carcass chemical fat, 15% more subcutaneous fat, 6% less bone and the same amount oflean tissue. The results for carcass fatness are consistent with observations reported earlier (Kleemann et al. 1985) on the marked effect of the Booroola Merino on carcass fat depth. The ranking of these strains with the TFxM is unclear because of a strainxyear interaction for lean and the components of carcass fat. As consumer preference shifts more towards lean meat, the amount of lean tissue at a given carcass weight becomes a prime determinant of its commercial value (Kempster et al. 1982). Our experiment indicates that the lean contents ofm and BxM carcasses were the same and therefore, on a commercial basis, could be of equivalent value. However, commercial value is currently based on carcass fatness, with no account being taken of the amount oflean tissue and bone. This highlights the importance of detailing breed and strain variation in carcass components through dissection. Two explanations are offered for the strainxyear interaction where the TFxM had more lean and less fat (subcutaneous and chemical) than the BxM in year I, but the same amount oflean and fat in year 2 (Table 3). The interaction may have been due to sampling; 5 different rams were used in each year. Alternatively, a true genotypex environment interaction may have been observed, with strains differing in their ability to produce lean tissue and fat under varying conditions (years) encountered in the experiment. In statistical terminology the environmental (year) effect may be considered as random or unpredictable (Dickerson 1962), in which case improved accuracy in measuring strain differences in average carcass composition across environments (years) would need to cover a number of environments (years). In this regard it would be inappropriate to determine strain differences using the current TFx M data if, in fact, the effects were due to a true genotypexenvironment interaction. Butterfield et al. (1983) have indicated that, at the same liveweight, a small mature size Merino strain has a greater proportion of carcass fat than a large mature size Merino strain. At the same proportion of mature liveweight, differences between the strains were small. Differences observed between the BxM and M strains for carcass fat (13-15%) and that estimated for mature size (10%; Ponzoni et al. 1984) support Butterfield's et al. ( 1983) conclusions. There is no published information on the effect of the F gene on carcass composition. Our results have to be treated with caution because of the small number of sires involved. However, they suggest that the Booroola F gene is oflittle consequence for carcass composition, which is in marked contrast to its influence on ovulation rate (Piper et al. 1984). We suggest that transfer of the gene to other background genotypes would not affect greatly carcass composition. Subjective fat scores, currently in use for commercial fat assessment, are ideally an estimation of percentage subcutaneous fat. Values for percentage subcutaneous fat in our study for M, BxM and TFxM strains at the same age were 7 8, 8 4 and 8 3 for the light slaughter group respectively; corresponding values for the heavy slaughter group were 9 9, 11 2 and 11 7. Corresponding fat scores, as estimated from the Meat and Livestock Commission (MLC) classification scheme (Kempster and Cuthbertson 1977) were l 9, 2 1 and 2 1 for the light slaughter group and 2 5, 2 8 and 3 0 for the heavier group. These latter values are in close agreement with least-squares mean values of2 5, 2 7 and 3 0 observed by D. 0. Kleemann, R. W. Ponzoni, J.E. Stafford and R. J. Grimson (unpublished data), and were O 5 of a fat score higher for the former values. Fat scores in the latter study were based on the British photographic standards from the MLC (Moxham and Brownlie 1976). Determination of the economic consequences of these strain differences for fat status will only be possible when reliable price data become available from the classification scheme introduced recently in Australia. We conclude that crossing the South Australian Merino with the Booroola produces carcasses with the same amount of lean tissue, less bone and greater amounts of
Carcass composition of the S. Aust. Merino 171 subcutaneous and carcass chemical fat, when compared at the same carcass weight. The rank of these strains with the Trangie Fertility cross and relative differences for major carcass components remain unclear; owing to a strainx year interaction. We suggest that transfer of the Booroola F gene to other background genotypes is unlikely to affect greatly carcass composition. Acknowledgments We thank: Messrs D. H. Smith and R.]. Lampe for valuable technical assistance and other staff at the Turretfield Research Centre for their contributions to this study; Mr R. Hetherington, Department of Chemistry, Adelaide, for analysis of meat samples; Dr D.R. Gifford for helpful comments on the manuscript. The Booroola and Trangie Fertility rams used in the experiment were from the CSIRO, Armidale, N.S.W., and the New South Wales Department of Agriculture, Trangie, respectively. References Atkins, K. D., and Robards, G. E. (1976). Efficiency of growth and wool production of young Merino ewes from a flock selected for fertility. Australian Jou/'llal of Experimrntal Agriculture and Animal Husband!'.)' 16, 315-20. Butterfield, R. M., Griffiths, D. A., Thompson, J.M., Zamora, J., and James, A. M. (1983). Changes in body composition relative to weight and maturity in large and small strains of Australian Merino rams. Animal Production 36, 29-37. Dickerson, G. E. (1962). Implications of genetic-environmental interaction in animal breeding. Animal Production 4, 47-63. Freund, R. J., and Littell, R. C. (1981). 'SAS for Linear Models. A Guide to the ANOVA and GLM Procedures.' (SAS Institute Incorporated: North Carolina, U.S.A.) Kempster, A. J., and Cuthbertson, A. (1977). A survey of the carcass characteristics of the main types of British lamb. Animal Production 25, 165-79. Kempster, A. J., Cuthbertson, A., and Harrington, G. (Eds) ( 1982). 'Carcass Evaluation in Livestock Breeding, Production and Marketing.' (Granada: London.) Kleemann, D. 0., Ponzoni, R. W., Stafford, J.E., Cutten, I. N., and Grim son, R. J. ( 1985). Growth and carcass characters of the South Australian Merino and its crosses with the Booroola and Trangie Fertility. Australian Jou/'llal of Experimental Agriculture 25, 750-7. Moxham, R. W., and Brownlie, L. E. (1976). Sheep carcass grading and classification in Australia. Wool Technology and Sheep Breeding 23, 17-25. Piper, L. R., Bindon, B. M., and Davis, G. H. (1984). The single gene inheritance of the prolificacy of the Booroola Merino. In 'Genetics of Reproduction in Sheep'. (Eds R. B. Land and D. W. Robinson.) pp. 127-37. (Butterworths: London.) Ponzoni, R. W., Walker, S. K., Walkley, J. R. W., and Fleet, M. R. (1984). The productivity of Bungaree, BooroolaxBungaree and Trangie FertilityxBungaree Merino ewes in South Australia. In 'Genetics of Reproduction in Sheep'. (Eds R. B. Land and D. W. Robinson.) pp. 127-38. (Butterworths: London.) Roberts, E. M., Jackson, N., and Phillips, J. M. (1975). Revised list of family groups of Australian Merino Stud Flocks. Wool Technology and Sheep Breeding 22, 6-9. Searle, T. W., Graham, N. McC., and O'Callaghan, M. (1972). Growth in sheep. I. The chemical composition of the body. Jou/'llal of Agricultural Science, Cambridge 79, 371-82. Southam, E. R., and Field, R. A. (1969). Influence of carcass weight upon carcass composition and consumer preference for lamb. Journal of Animal Science 28, 584-8. Turner, H. N. (1978). Selection for reproduction rate in Australian Merino sheep: Direct responses. Australian Joumal of Agricultural Research 29, 327-50. Walker, D. J., Potter, B. J., and Jones, G. B. (1971). Modification of carcass characteristics in sheep maintained on a saline water regime. Australian Journal of Experimental Agriculture and Animal Husband,:v 11, 14-17. Walkley, J. R. W., Walker, S. K., Ponzoni, R. W., and Smith, D. H. (1982). Plans for Booroola Merino evaluation on oestrogenic pastures. In 'The Booroola Merino'. (Eds L. R. Piper, B. M. Bindon and R. D. Nethery.) pp. 61-2. (CSIRO: Melbourne.) Received 23 August 1987, accepted 2 November 1987 /