- Email: [email protected]
Use of linear measurements of m. longissimus∗ to predict the muscle content of beef carcasses
PII:
SO309-1740(96)00117-9
Meat Scieme, Vol. 45, No. 3, 321-327, 1997 0 1997 Elsevier Science Ltd A11 rights reserved. Printed in Great Britain 0309-1740/97 S17.00+0.00
ELSEVIER
Use of Linear Measurements of m. fongissimus* to Predict the Muscle Content of Beef Carcasses E. R. Johnson & D. A. Baker Department
of Farm Animal Medicine and Production, The University of Queensland, P.O. Box 125, Kenmore, Queensland 4069, Australia
(Received 11 May 1996; revised version received 20 September 1996; accepted 22 September 1996)
ABSTRACT Five linear measurements associated with the eye muscle (m. longissimus), together with hot carcass weight, 10th rib fat thickness, eye muscle area and an estimate of eye muscle volume (eye muscle area x a carcass length measurement) were made on 53 chilled beef carcasses (hot weight 143-384 kg). The right side of each carcass was anatomically dissected into muscle, bone, fat and connective tissue. Correlation and regression analyses were used to identtfy the most accurate predictors of weight and percentage of side muscle. In simple regression, hot carcass weight and the estimate of eye muscle volume were the most accurate predictors of side muscle weight; 10th rib fat thickness and MN, a depth measurement of muscle and fat over the loin, were the most accurate predictors of percentage side muscle. In multiple regression, the addition of either eye muscle volume or eye muscle area to hot carcass weight and 10th rib fat thickness gave the most accurate predictions of side muscle weight and percentage side muscle, but in the case of each dependent variable, the improvement in accuracy was slight compared with that of the two most accurate regressors, hot carcass weight and 10th rib fat thickness. Although eye muscle volume was a more accurate predictor of side muscle weight than eye muscle area in simple regression, their contributions in multiple regression with hot carcass weight and 10th rib fat thickness were similar. None of the five linear measurements associated with m. longissimus contributed significantly to improving the prediction of weight or percentage of side muscle. 0 1996 Elsevier Science Ltd. All rights reserved
INTRODUCTION Eye muscle area is commonly used as a regressor with carcass weight and subcutaneous fat thickness to improve the accuracy of prediction of weight or percentage of commercial yield (carcass muscle together with a specified account of fat) in beef carcasses (Murphey et al., 1960; Brungardt & Bray, 1963; Abraham et al., 1968). Although eye muscle area alone is a poor predictor of total dissectible carcass muscle (Cole et al., 1960; Go11 et al., 1961; Johnson et al., 1992), its use in multiple regression generally leads to a modest improvement in the accuracy of prediction (Crouse et al., 1975; Alliston, 1982; Porter et al., 1990; Johnson et al., 1992, 1995). The measurement is widely used in Australia in the *In this paper, m. longissimus signifies m. longissimus thoracis et lumborum. 321
322
E. R. Johnson, D. A. Baker
Chiller Assessment Scheme (Anonymous, 1991) and as an optional measurement in BREEDPLAN (McDonald, 1991). In cattle, the object of measuring eye muscle area is to give an indication of the muscling status of the animals concerned for marketing or genetic selection purposes. Although scanning ultrasound is highly accurate in determining back fat thickness in live animals (Brethour, 1992; Robinson et al., 1992; Smith et al., 1992), many problems still exist in determining eye muscle area accurately, especially in heavier animals (Smith et al., 1992; Herring et al., 1994). Because of the relatively large prediction errors associated with estimates of eye muscle area (Cross, 1989; Smith et al., 1992; Waldner et al., 1992), ultrasound technology is probably not yet at the stage where live animal measurements can replace costly and laborious serial slaughter programmes in growth studies. Cross ef al. (1989) reported errors of up to 56% in eye muscle area measurement, whereas Smith et al. (1992) found that about 50% of feed-lot steers could not be measured to within 6.4 cm* of actual eye muscle area. Robinson et al. (1992) described a residual standard deviation (RSD) of 5.1 cm2 when estimating areas of about 60 cm*. Despite the rapid development, real-time ultrasound cannot be relied upon to measure eye muscle area accurately in all cattle. In particular, the deep and medial boundaries of m. lungissimus have often been difficult to discern (Cross, 1989; Smith et al., 1992). This may have accounted for inaccuracies of real-time ultrasound in measuring eye muscle area. Wilson (1992), who noted the success of ultrasound in improving the accuracy of prediction of body composition in swine, suggested that considerable research and development were necessary before the technology could be used successfully in beef cattle. However, the development of new transducers (Herring et al., 1994) and their application to carcass morphometric problems (Swatland et al., 1994) may eliminate major inaccuracies in the measurement of eye muscle area. If linear measurements of muscles, particularly the eye muscle, could be used to improve the quantification of carcass muscle as much as eye muscle area, then they could supplant this measurement, thus obviating any problems associated with area measurement. Some workers have investigated linear measurements of the eye muscle in predicting body composition in cattle. Porter et al. (1990) found eye muscle depth useful, and Gillis et al. (1973) noted that two depth measurements of the eye muscle at the 1I-12th rib were highly correlated with area. These and similar comparative or applied studies have not addressed the fundamental question of whether highly accurate linear measurements of muscle, made on the carcass, are satisfactory predictors of carcass muscle, determined accurately from total dissection. In the following study, conducted on chilled carcasses, the contribution of a number of linear measurements made on m. longissimus to the prediction of weight and percentage muscle, determined by total anatomical dissection was examined.
MATERIALS
AND METHODS
Fifty-three steer carcasses (11 Friesian, 17 Angus, 12 Hereford, 10 Charolais cross-bred and 3 Murray Grey) weighing from 143 to 384 kg (hot) were chilled at 3°C for 24 h before measurements were made. The measurements were subcutaneous fat thickness at the 10th rib, eye muscle (m. longissimus thorucis et lumborum) area and breadth (B) at the 10th rib, eye muscle volume and four measurements, CT, R, W and MN, described by Yeates (1952). Eye muscle area was measured with a transparent cellulose acetate grid divided into l-cm squares. CT is a carcass length measurement, from the anterior edge of the pecrenossis pubis to the junction of the seventh cervical and first thoracic vertebrae, which closely approximates the length of m. longissimus. R and W are muscle depth measurements at the 10th rib, the former being the greatest depth of m. longissimus, and the latter the
Linear measurements to predict the muscle content of beef
323
greatest depth of a muscle mass which includes m. spinalis dorsi, m. longissimus and m. multifdis dorsi (Fig. 1). MN is a depth measurement involving subcutaneous fat and m. longissimus at a defined point at the 4th lumbar vertebra. This point is located by measuring a distance equivalent to that from the surface of the subcutaneous fat to the bottom of the neural canal, laterally and at right angles to the chine (Fig. 2). Eye muscle volume was estimated by multiplying CT by eye muscle area at the 10th rib. The right side of each carcass was dissected into its component tissues, muscle, bone, fat and connective tissue within 96 h of slaughter. Each side took from 50 to 70 h to dissect and the dissection room was maintained at high humidity to minimise evaporative losses. The study extended over almost 2 years. Correlation and regression were used to determine the usefulness of measurements as predictors of the weight and percentage of side muscle. Regressors were added in a stepwise fashion and evaluated on the basis of their residual standard deviation and coefficient of determination.
RESULTS
AND DISCUSSION
The variables that contributed significantly to the prediction of weight or percentage of side muscle in simple regression are shown (Table 1). For weight predictions, hot carcass weight, eye muscle volume and eye muscle area were the most accurate predictors. Eye muscle area was clearly superior to B x R, indicating that a product of width and depth of m. longissimus at the 10th rib was not a good substitute for eye muscle area. For the prediction of side muscle weight, hot carcass weight was the major contributor. For the prediction of percentage side muscle, 10th rib fat thickness and tissue depth (MN) at the 4th lumbar vertebra were the only measurements to give a reasonably accurate prediction. Regression analyses showed that the width (B) and depth (R) of m. longissimus, eye muscle area and eye muscle volume changed in a linear fashion with hot carcass weight, but carcass length (CT), loin tissue depth (MN) and depth of muscle mass at the 10th rib (W) were better described by quadratic regression. These regressions were negative for CT and W and positive for MN. The addition of a second regressor to the best single predictor, hot carcass weight, showed that fat thickness at the 10th rib was easily the next most accurate contributor to
Fig. 1. Cross-section of beef side at the 10th rib showing measurements. R and W: muscle depth measurements; B: width of m. longissimus.After Yeates (1952).
324
E. R. Johnson, D. A. Baker
the prediction of side muscle weight (Table 2). Other regressors which contributed significantly to prediction with hot carcass weight were tissue depth (MN), eye muscle volume, carcass length (CT) and eye muscle area. The first three improved prediction accuracy to a similar degree and were superior to eye muscle area. For the prediction of percentage side muscle, the same five second regressors contributed significantly to regression, but only fat thickness at the 10th rib and hot carcass weight improved the accuracy of percentage prediction over fat thickness at the 10th rib alone. Much less variance was accounted for in predicting the percentage of side muscle. m. loaissimus jumborum
..r-...:..t-~:;.;..r..
~y*:.y.~.~~._y.$~ ,, .. MUSCLE
...I ‘BONE
Fig. 2. Cross-section of beef side at the 4th lumbar vertebra showing measurement MN. PQ = OM and MN is parallel to PQ with N located on the transverse. process of the vertebra. After Yea& (1952).
Carcass Measurements
that Contributed
Carcass measurement Total muscle weight (mean = 66778 g) Hot carcass weight” Eye muscle volume” Eye muscle area** CT** B x R” R” * B MN” W” Total muscle (%) (mean = 56.31) Fat thickness at 10th rib** MN” Hot carcass weight” R” Eve muscle area* Eye muscle volume’ W’ B x R’ l
TABLE 1 to the Prediction of Weight or Percentage of Side Muscle in Simple Regression RSD (g or %)
5630 6107 7585 8996 9580 10 936 11830 12 155 12 521
3.40 3.63 4.29 5.20 5.26 5.27 5.32 5.39
Coeficient of determination (9)
0.84 0.81 0.71 0.59 0.54 0.39 0.29 0.25 0.21
0.62 0.57 0.40 0.12 0.10 0.10 0.08 0.06
RSD: residual standard deviation; CT: a carcass length measurement; B, R, MN and W: linear measurements of m. longissimus. Defined in Materials and Methods.*p < 0.05; “p < 0.01.
325
Linear measurements to predict the muscle content of beef
Because hot carcass weight and fat thickness at the 10th rib was the most accurate combination of regressors for predicting the weight and percentage of side muscle, a number of third regressors was added to this combination in an effort to improve prediction accuracy (Table 3). Only eye muscle volume and eye muscle area proved useful, each improving the prediction of weight and percentage of side muscle slightly but significantly. Of these two measurements, eye muscle volume, either alone or together with hot carcass weight, was generally superior to eye muscle area in predicting the weight and percentage of side muscle. However, when used as a third regressor, there was little between them in improving the accuracy of prediction except that eye muscle area did improve, modestly, the accuracy of percentage prediction (from RSD, 3.11 to 2.90). Johnson (1993) showed that eye muscle area was most valuable as a predictor when the carcass weight range was wide. It is likely, therefore, that over the extensive carcass weight range in the current study (143-384 kg), eye muscle area was exerting a relatively large effect. If so, the influence was still minor. A number of workers (Kempster et al., 1981; Alliston, 1982; Anderson et al., 1983) suggested that the contribution of eye muscle area to the estimation of carcass muscle is small or of no real value. If the most beneficial use of eye muscle area is to differentiate among cattle or carcasses of similar weight (young bulls or carcasses directed to the one market), it would seem to have a restricted use in quantification. This study indicates that none of the linear measurements made on m. longissimus at the 10th rib was of value in improving the prediction of weight or percentage of side muscle. Carcass length (CT) was of value, but only for estimating eye muscle volume and this product gave little or no improvement over the contribution of eye muscle area itself. Furthermore, it should be noted that a reliable carcass length measurement, such as CT, is very difficult to make in the live animal. Modern ultrasound is capable of measuring fat TABLE 2 Carcass Measurements Carcass measurementf
Total muscle weight (mean = 66778.5) HCW” + F-f,0 HCW” + MN” HCW” + EMV” HCW” + CT” HCW” + EMA” Total muscle (%) (mean = 56.31) ** HCW” + FT,,, HCWNS + MN” HCW” + EMV” HCW” + EMA” HCW** + CT*
that Contributed to the Prediction of Weight or Percentage of Side Muscle in Multiple Regression Using Two Variables RSD (g or %)
3437 4569 4648 4818 5253
3.11 3.60 3.68 3.90 4.06
Coeficient of determination (ti)
0.94 0.89 0.89 0.88 0.86
0.69 0.58 0.56 0.51 0.46
RSD: residual standard deviation; HCW: hot carcass weight; FTra: fat thickness at 10th rib; EMA: eye muscle area at 10th rib; EMV: eye muscle volume; MN and CT: tissue depth and carcass length measurements, respectively, defined in Materials and Methods. “No linear measurements of m. longissimus contributed significantly to regression. l p < 0.05; “p < 0.01. Significance levels are for each term in the prediction after the inclusion of the others.
326
Carcass Measurements
E. R. Johnson, D. A. Baker TABLE 3 Used With Carcass Weight and Fat Thickness in an Attempt to Improve the Prediction of Weight or Percentage of Side Muscle
Carcass measurements added to hot carcass weight andfat thickness at f&h rib Total muscle weight (mean = 66 778 g) EMV” EM A’ RNs BNs MNNS B x RNS CTNS WNs Total muscle (%) (mean = 56.31) EM A’ EMV’ RNs WNs MNNS B x RNS CTNS BNs
RSD (g or %)
3213 3287 3327 3409 3443 3445 3450 3462
2.90 2.98 3.10 3.10 3.10 3.12 3.13 3.14
Coeficient of determination (rZ)
0.95 0.95 0.94 0.94 0.94 0.94 0.94 0.94
0.73 0.71 0.69 0.69 0.69 0.68 0.68 0.68
RSD: residual standard deviation; EMA, eye muscle area; EMV: eye muscle volume; R, B, MN, CT and W: linear measurements of tissue or carcass, defined in Materials and Methods. l p < 0.05; “p < 0.01; NS: not significant. Significance levels apply to the third regressor with hot carcass weight and fat thickness both contributing significantly to regression.
depth and fat area with a high degree of accuracy (Brethour, 1992; Smith et al., 1992) and a large volume of evidence shows that, with scanners, fat depth and fat area are the best quantifiers of carcass lean (Kempster et al., 1981; Alliston, 1982; Anderson et al., 1983; Porter er al., 1990). Smith et al. (1992) and Wilson (1992) urged caution in making breed or management decisions based on estimates of eye muscle area using current technology and expertise. Results of this study, based on 10th rib measurements, support the conclusion that scanning technology should be directed at obtaining an accurate measurement of eye muscle area rather than investigating linear measurements of the eye muscle. When this is achieved, the true value of eye muscle area, currently unclear from the literature, may be determined.
ACKNOWLEDGEMENTS The authors are grateful to Dr David Chant, Graduate School of Education, The University of Queensland for statistical advice and to MS Lyn Knott, Department of Farm Animal Medicine and Production, The University of Queensland for technical assistance.
Linear measurements to predict the muscle content qf beef
321
REFERENCES Abraham, H. C., Carpenter, Z. L., King, G. T. & Butler, 0. D. (1968). J. Animal Sci., 27, 604. Alliston, J. C. (1982). Animal Prod., 35, 361. Anderson, B. B., Busk, H., Chadwick, J. P., Cuthbertson, A., Fursey, G. A. J., Jones, D. W., Lewin, P., Miles, C. A. & Owen, M. G. (1983). Live. Prod. Sci., 10, 133. Anonymous (1991). Chiller Assessment. A Pictorial Guide. AUS-MEAT, Australian Meat and Livestock Corporation, Sydney, Australia. Brethour, J. R. (1992). J. Animal Sci., 70, 1039. Brungardt, V. H. & Bray, R. W. (1963). J. Animal Sci., 22, 177. Cole, J. W., Orme, L. E. & Kincaid, C. M. (1960). J. Animal Sci., 19, 89. Cross, H. R. (1989). Proc. Beef Improvement Federation, pp. l-6. Cross, H. R., Whittaker, D. & Savell, J. W. (1989). The Automated Measurement ofBeef, eds L. E. Brownlie, W. J. A. Hall & S. U. Fabiansson. Australian Meat and Live-stock Corporation, Sydney, p. 1. Crouse, J. D., Dikeman, M. E., Koch, R. M. & Murphey, C. E. (1975). J. Animal Sci., 41, 548. Gillis, W. A., Burgess, T. D., Usborne, W. R., Greiger, H. & Talbot, S. (1973). Can. J. Animal Sci., 53, 13. Goll, D. E., Hazel, L. N. & Kline, E. A. (1961). J. Animal Sci., 20, 264. Herring, W. O., Miller, D. C., Bertrand, J. K. & Benyshek, L. L. (1994). J. Animal Sci., 72, 2216. Johnson, E. R. (1993). The Evaluation of Bulls Using a Beef Progeny Testing System Based on Carcass Merit. Australian Association of Cattle Veterinarians, Gold Coast Proceedings, Upjohn Pty Ltd, Sydney, p. 85. Johnson, E. R., Taylor, D. G. & Priyanto, R. (1992). Proc. 38th Int. Cong. Meat Sci. Technol. Clermont Ferrand, France, p. 911. Johnson, E. R., Taylor, D. G. & Priyanto, R. (1995). Meat Sci., 40, 13. Kempster, A. J., Cuthbertson, A., Jones, D. W. & Owen, M. G. (1981). .I. Agric. Sci. Camb., 96, 301. McDonald, A. (1991). Specialist School on BREEDPLAN, Handbook, ed. R. Fenwick. University of New England, Armidale, Australia. Murphey, C. E., Hallett, D. K., Tyler, W. E. & Pierce, J. C., Jr. (1960). J. Animal Sci., 19 (Abst), 1240. Porter, S. J., Owen, M. G., Page, S. J. & Fisher, A. V. (1990). Animal Prod., 51,489. Robinson, D. L., McDonald, C. A., Hammond, K. & Turner, J. W. (1992). J. Animal Sci., 70, 1667. Smith, M. T., Oltjen, J. W., Dolezal, H. G., Gill, D. R. & Behrens, B. D. (1992). J. Animal Sci., 70, 29. Swatland, H. J., Ananthanarayanan, S. P. & Goldenberg, A. A. (1994). J. Animal Sci., 72, 1475. Waldner, D. N., Dikeman, M. E., Schalles, R. R., Olson, W. G., Houghton, P. L., Unruh, J. A. & Corah, L. R. (1992). J. Animal Sci., 70, 3044. Wilson, D. E. (1992). J. Animal Sci., 70,973. Yeates, N. T. M. (1952). Aust. J. Agric. Res., 3, 68.
SO309-1740(96)00117-9
Meat Scieme, Vol. 45, No. 3, 321-327, 1997 0 1997 Elsevier Science Ltd A11 rights reserved. Printed in Great Britain 0309-1740/97 S17.00+0.00
ELSEVIER
Use of Linear Measurements of m. fongissimus* to Predict the Muscle Content of Beef Carcasses E. R. Johnson & D. A. Baker Department
of Farm Animal Medicine and Production, The University of Queensland, P.O. Box 125, Kenmore, Queensland 4069, Australia
(Received 11 May 1996; revised version received 20 September 1996; accepted 22 September 1996)
ABSTRACT Five linear measurements associated with the eye muscle (m. longissimus), together with hot carcass weight, 10th rib fat thickness, eye muscle area and an estimate of eye muscle volume (eye muscle area x a carcass length measurement) were made on 53 chilled beef carcasses (hot weight 143-384 kg). The right side of each carcass was anatomically dissected into muscle, bone, fat and connective tissue. Correlation and regression analyses were used to identtfy the most accurate predictors of weight and percentage of side muscle. In simple regression, hot carcass weight and the estimate of eye muscle volume were the most accurate predictors of side muscle weight; 10th rib fat thickness and MN, a depth measurement of muscle and fat over the loin, were the most accurate predictors of percentage side muscle. In multiple regression, the addition of either eye muscle volume or eye muscle area to hot carcass weight and 10th rib fat thickness gave the most accurate predictions of side muscle weight and percentage side muscle, but in the case of each dependent variable, the improvement in accuracy was slight compared with that of the two most accurate regressors, hot carcass weight and 10th rib fat thickness. Although eye muscle volume was a more accurate predictor of side muscle weight than eye muscle area in simple regression, their contributions in multiple regression with hot carcass weight and 10th rib fat thickness were similar. None of the five linear measurements associated with m. longissimus contributed significantly to improving the prediction of weight or percentage of side muscle. 0 1996 Elsevier Science Ltd. All rights reserved
INTRODUCTION Eye muscle area is commonly used as a regressor with carcass weight and subcutaneous fat thickness to improve the accuracy of prediction of weight or percentage of commercial yield (carcass muscle together with a specified account of fat) in beef carcasses (Murphey et al., 1960; Brungardt & Bray, 1963; Abraham et al., 1968). Although eye muscle area alone is a poor predictor of total dissectible carcass muscle (Cole et al., 1960; Go11 et al., 1961; Johnson et al., 1992), its use in multiple regression generally leads to a modest improvement in the accuracy of prediction (Crouse et al., 1975; Alliston, 1982; Porter et al., 1990; Johnson et al., 1992, 1995). The measurement is widely used in Australia in the *In this paper, m. longissimus signifies m. longissimus thoracis et lumborum. 321
322
E. R. Johnson, D. A. Baker
Chiller Assessment Scheme (Anonymous, 1991) and as an optional measurement in BREEDPLAN (McDonald, 1991). In cattle, the object of measuring eye muscle area is to give an indication of the muscling status of the animals concerned for marketing or genetic selection purposes. Although scanning ultrasound is highly accurate in determining back fat thickness in live animals (Brethour, 1992; Robinson et al., 1992; Smith et al., 1992), many problems still exist in determining eye muscle area accurately, especially in heavier animals (Smith et al., 1992; Herring et al., 1994). Because of the relatively large prediction errors associated with estimates of eye muscle area (Cross, 1989; Smith et al., 1992; Waldner et al., 1992), ultrasound technology is probably not yet at the stage where live animal measurements can replace costly and laborious serial slaughter programmes in growth studies. Cross ef al. (1989) reported errors of up to 56% in eye muscle area measurement, whereas Smith et al. (1992) found that about 50% of feed-lot steers could not be measured to within 6.4 cm* of actual eye muscle area. Robinson et al. (1992) described a residual standard deviation (RSD) of 5.1 cm2 when estimating areas of about 60 cm*. Despite the rapid development, real-time ultrasound cannot be relied upon to measure eye muscle area accurately in all cattle. In particular, the deep and medial boundaries of m. lungissimus have often been difficult to discern (Cross, 1989; Smith et al., 1992). This may have accounted for inaccuracies of real-time ultrasound in measuring eye muscle area. Wilson (1992), who noted the success of ultrasound in improving the accuracy of prediction of body composition in swine, suggested that considerable research and development were necessary before the technology could be used successfully in beef cattle. However, the development of new transducers (Herring et al., 1994) and their application to carcass morphometric problems (Swatland et al., 1994) may eliminate major inaccuracies in the measurement of eye muscle area. If linear measurements of muscles, particularly the eye muscle, could be used to improve the quantification of carcass muscle as much as eye muscle area, then they could supplant this measurement, thus obviating any problems associated with area measurement. Some workers have investigated linear measurements of the eye muscle in predicting body composition in cattle. Porter et al. (1990) found eye muscle depth useful, and Gillis et al. (1973) noted that two depth measurements of the eye muscle at the 1I-12th rib were highly correlated with area. These and similar comparative or applied studies have not addressed the fundamental question of whether highly accurate linear measurements of muscle, made on the carcass, are satisfactory predictors of carcass muscle, determined accurately from total dissection. In the following study, conducted on chilled carcasses, the contribution of a number of linear measurements made on m. longissimus to the prediction of weight and percentage muscle, determined by total anatomical dissection was examined.
MATERIALS
AND METHODS
Fifty-three steer carcasses (11 Friesian, 17 Angus, 12 Hereford, 10 Charolais cross-bred and 3 Murray Grey) weighing from 143 to 384 kg (hot) were chilled at 3°C for 24 h before measurements were made. The measurements were subcutaneous fat thickness at the 10th rib, eye muscle (m. longissimus thorucis et lumborum) area and breadth (B) at the 10th rib, eye muscle volume and four measurements, CT, R, W and MN, described by Yeates (1952). Eye muscle area was measured with a transparent cellulose acetate grid divided into l-cm squares. CT is a carcass length measurement, from the anterior edge of the pecrenossis pubis to the junction of the seventh cervical and first thoracic vertebrae, which closely approximates the length of m. longissimus. R and W are muscle depth measurements at the 10th rib, the former being the greatest depth of m. longissimus, and the latter the
Linear measurements to predict the muscle content of beef
323
greatest depth of a muscle mass which includes m. spinalis dorsi, m. longissimus and m. multifdis dorsi (Fig. 1). MN is a depth measurement involving subcutaneous fat and m. longissimus at a defined point at the 4th lumbar vertebra. This point is located by measuring a distance equivalent to that from the surface of the subcutaneous fat to the bottom of the neural canal, laterally and at right angles to the chine (Fig. 2). Eye muscle volume was estimated by multiplying CT by eye muscle area at the 10th rib. The right side of each carcass was dissected into its component tissues, muscle, bone, fat and connective tissue within 96 h of slaughter. Each side took from 50 to 70 h to dissect and the dissection room was maintained at high humidity to minimise evaporative losses. The study extended over almost 2 years. Correlation and regression were used to determine the usefulness of measurements as predictors of the weight and percentage of side muscle. Regressors were added in a stepwise fashion and evaluated on the basis of their residual standard deviation and coefficient of determination.
RESULTS
AND DISCUSSION
The variables that contributed significantly to the prediction of weight or percentage of side muscle in simple regression are shown (Table 1). For weight predictions, hot carcass weight, eye muscle volume and eye muscle area were the most accurate predictors. Eye muscle area was clearly superior to B x R, indicating that a product of width and depth of m. longissimus at the 10th rib was not a good substitute for eye muscle area. For the prediction of side muscle weight, hot carcass weight was the major contributor. For the prediction of percentage side muscle, 10th rib fat thickness and tissue depth (MN) at the 4th lumbar vertebra were the only measurements to give a reasonably accurate prediction. Regression analyses showed that the width (B) and depth (R) of m. longissimus, eye muscle area and eye muscle volume changed in a linear fashion with hot carcass weight, but carcass length (CT), loin tissue depth (MN) and depth of muscle mass at the 10th rib (W) were better described by quadratic regression. These regressions were negative for CT and W and positive for MN. The addition of a second regressor to the best single predictor, hot carcass weight, showed that fat thickness at the 10th rib was easily the next most accurate contributor to
Fig. 1. Cross-section of beef side at the 10th rib showing measurements. R and W: muscle depth measurements; B: width of m. longissimus.After Yeates (1952).
324
E. R. Johnson, D. A. Baker
the prediction of side muscle weight (Table 2). Other regressors which contributed significantly to prediction with hot carcass weight were tissue depth (MN), eye muscle volume, carcass length (CT) and eye muscle area. The first three improved prediction accuracy to a similar degree and were superior to eye muscle area. For the prediction of percentage side muscle, the same five second regressors contributed significantly to regression, but only fat thickness at the 10th rib and hot carcass weight improved the accuracy of percentage prediction over fat thickness at the 10th rib alone. Much less variance was accounted for in predicting the percentage of side muscle. m. loaissimus jumborum
..r-...:..t-~:;.;..r..
~y*:.y.~.~~._y.$~ ,, .. MUSCLE
...I ‘BONE
Fig. 2. Cross-section of beef side at the 4th lumbar vertebra showing measurement MN. PQ = OM and MN is parallel to PQ with N located on the transverse. process of the vertebra. After Yea& (1952).
Carcass Measurements
that Contributed
Carcass measurement Total muscle weight (mean = 66778 g) Hot carcass weight” Eye muscle volume” Eye muscle area** CT** B x R” R” * B MN” W” Total muscle (%) (mean = 56.31) Fat thickness at 10th rib** MN” Hot carcass weight” R” Eve muscle area* Eye muscle volume’ W’ B x R’ l
TABLE 1 to the Prediction of Weight or Percentage of Side Muscle in Simple Regression RSD (g or %)
5630 6107 7585 8996 9580 10 936 11830 12 155 12 521
3.40 3.63 4.29 5.20 5.26 5.27 5.32 5.39
Coeficient of determination (9)
0.84 0.81 0.71 0.59 0.54 0.39 0.29 0.25 0.21
0.62 0.57 0.40 0.12 0.10 0.10 0.08 0.06
RSD: residual standard deviation; CT: a carcass length measurement; B, R, MN and W: linear measurements of m. longissimus. Defined in Materials and Methods.*p < 0.05; “p < 0.01.
325
Linear measurements to predict the muscle content of beef
Because hot carcass weight and fat thickness at the 10th rib was the most accurate combination of regressors for predicting the weight and percentage of side muscle, a number of third regressors was added to this combination in an effort to improve prediction accuracy (Table 3). Only eye muscle volume and eye muscle area proved useful, each improving the prediction of weight and percentage of side muscle slightly but significantly. Of these two measurements, eye muscle volume, either alone or together with hot carcass weight, was generally superior to eye muscle area in predicting the weight and percentage of side muscle. However, when used as a third regressor, there was little between them in improving the accuracy of prediction except that eye muscle area did improve, modestly, the accuracy of percentage prediction (from RSD, 3.11 to 2.90). Johnson (1993) showed that eye muscle area was most valuable as a predictor when the carcass weight range was wide. It is likely, therefore, that over the extensive carcass weight range in the current study (143-384 kg), eye muscle area was exerting a relatively large effect. If so, the influence was still minor. A number of workers (Kempster et al., 1981; Alliston, 1982; Anderson et al., 1983) suggested that the contribution of eye muscle area to the estimation of carcass muscle is small or of no real value. If the most beneficial use of eye muscle area is to differentiate among cattle or carcasses of similar weight (young bulls or carcasses directed to the one market), it would seem to have a restricted use in quantification. This study indicates that none of the linear measurements made on m. longissimus at the 10th rib was of value in improving the prediction of weight or percentage of side muscle. Carcass length (CT) was of value, but only for estimating eye muscle volume and this product gave little or no improvement over the contribution of eye muscle area itself. Furthermore, it should be noted that a reliable carcass length measurement, such as CT, is very difficult to make in the live animal. Modern ultrasound is capable of measuring fat TABLE 2 Carcass Measurements Carcass measurementf
Total muscle weight (mean = 66778.5) HCW” + F-f,0 HCW” + MN” HCW” + EMV” HCW” + CT” HCW” + EMA” Total muscle (%) (mean = 56.31) ** HCW” + FT,,, HCWNS + MN” HCW” + EMV” HCW” + EMA” HCW** + CT*
that Contributed to the Prediction of Weight or Percentage of Side Muscle in Multiple Regression Using Two Variables RSD (g or %)
3437 4569 4648 4818 5253
3.11 3.60 3.68 3.90 4.06
Coeficient of determination (ti)
0.94 0.89 0.89 0.88 0.86
0.69 0.58 0.56 0.51 0.46
RSD: residual standard deviation; HCW: hot carcass weight; FTra: fat thickness at 10th rib; EMA: eye muscle area at 10th rib; EMV: eye muscle volume; MN and CT: tissue depth and carcass length measurements, respectively, defined in Materials and Methods. “No linear measurements of m. longissimus contributed significantly to regression. l p < 0.05; “p < 0.01. Significance levels are for each term in the prediction after the inclusion of the others.
326
Carcass Measurements
E. R. Johnson, D. A. Baker TABLE 3 Used With Carcass Weight and Fat Thickness in an Attempt to Improve the Prediction of Weight or Percentage of Side Muscle
Carcass measurements added to hot carcass weight andfat thickness at f&h rib Total muscle weight (mean = 66 778 g) EMV” EM A’ RNs BNs MNNS B x RNS CTNS WNs Total muscle (%) (mean = 56.31) EM A’ EMV’ RNs WNs MNNS B x RNS CTNS BNs
RSD (g or %)
3213 3287 3327 3409 3443 3445 3450 3462
2.90 2.98 3.10 3.10 3.10 3.12 3.13 3.14
Coeficient of determination (rZ)
0.95 0.95 0.94 0.94 0.94 0.94 0.94 0.94
0.73 0.71 0.69 0.69 0.69 0.68 0.68 0.68
RSD: residual standard deviation; EMA, eye muscle area; EMV: eye muscle volume; R, B, MN, CT and W: linear measurements of tissue or carcass, defined in Materials and Methods. l p < 0.05; “p < 0.01; NS: not significant. Significance levels apply to the third regressor with hot carcass weight and fat thickness both contributing significantly to regression.
depth and fat area with a high degree of accuracy (Brethour, 1992; Smith et al., 1992) and a large volume of evidence shows that, with scanners, fat depth and fat area are the best quantifiers of carcass lean (Kempster et al., 1981; Alliston, 1982; Anderson et al., 1983; Porter er al., 1990). Smith et al. (1992) and Wilson (1992) urged caution in making breed or management decisions based on estimates of eye muscle area using current technology and expertise. Results of this study, based on 10th rib measurements, support the conclusion that scanning technology should be directed at obtaining an accurate measurement of eye muscle area rather than investigating linear measurements of the eye muscle. When this is achieved, the true value of eye muscle area, currently unclear from the literature, may be determined.
ACKNOWLEDGEMENTS The authors are grateful to Dr David Chant, Graduate School of Education, The University of Queensland for statistical advice and to MS Lyn Knott, Department of Farm Animal Medicine and Production, The University of Queensland for technical assistance.
Linear measurements to predict the muscle content qf beef
321
REFERENCES Abraham, H. C., Carpenter, Z. L., King, G. T. & Butler, 0. D. (1968). J. Animal Sci., 27, 604. Alliston, J. C. (1982). Animal Prod., 35, 361. Anderson, B. B., Busk, H., Chadwick, J. P., Cuthbertson, A., Fursey, G. A. J., Jones, D. W., Lewin, P., Miles, C. A. & Owen, M. G. (1983). Live. Prod. Sci., 10, 133. Anonymous (1991). Chiller Assessment. A Pictorial Guide. AUS-MEAT, Australian Meat and Livestock Corporation, Sydney, Australia. Brethour, J. R. (1992). J. Animal Sci., 70, 1039. Brungardt, V. H. & Bray, R. W. (1963). J. Animal Sci., 22, 177. Cole, J. W., Orme, L. E. & Kincaid, C. M. (1960). J. Animal Sci., 19, 89. Cross, H. R. (1989). Proc. Beef Improvement Federation, pp. l-6. Cross, H. R., Whittaker, D. & Savell, J. W. (1989). The Automated Measurement ofBeef, eds L. E. Brownlie, W. J. A. Hall & S. U. Fabiansson. Australian Meat and Live-stock Corporation, Sydney, p. 1. Crouse, J. D., Dikeman, M. E., Koch, R. M. & Murphey, C. E. (1975). J. Animal Sci., 41, 548. Gillis, W. A., Burgess, T. D., Usborne, W. R., Greiger, H. & Talbot, S. (1973). Can. J. Animal Sci., 53, 13. Goll, D. E., Hazel, L. N. & Kline, E. A. (1961). J. Animal Sci., 20, 264. Herring, W. O., Miller, D. C., Bertrand, J. K. & Benyshek, L. L. (1994). J. Animal Sci., 72, 2216. Johnson, E. R. (1993). The Evaluation of Bulls Using a Beef Progeny Testing System Based on Carcass Merit. Australian Association of Cattle Veterinarians, Gold Coast Proceedings, Upjohn Pty Ltd, Sydney, p. 85. Johnson, E. R., Taylor, D. G. & Priyanto, R. (1992). Proc. 38th Int. Cong. Meat Sci. Technol. Clermont Ferrand, France, p. 911. Johnson, E. R., Taylor, D. G. & Priyanto, R. (1995). Meat Sci., 40, 13. Kempster, A. J., Cuthbertson, A., Jones, D. W. & Owen, M. G. (1981). .I. Agric. Sci. Camb., 96, 301. McDonald, A. (1991). Specialist School on BREEDPLAN, Handbook, ed. R. Fenwick. University of New England, Armidale, Australia. Murphey, C. E., Hallett, D. K., Tyler, W. E. & Pierce, J. C., Jr. (1960). J. Animal Sci., 19 (Abst), 1240. Porter, S. J., Owen, M. G., Page, S. J. & Fisher, A. V. (1990). Animal Prod., 51,489. Robinson, D. L., McDonald, C. A., Hammond, K. & Turner, J. W. (1992). J. Animal Sci., 70, 1667. Smith, M. T., Oltjen, J. W., Dolezal, H. G., Gill, D. R. & Behrens, B. D. (1992). J. Animal Sci., 70, 29. Swatland, H. J., Ananthanarayanan, S. P. & Goldenberg, A. A. (1994). J. Animal Sci., 72, 1475. Waldner, D. N., Dikeman, M. E., Schalles, R. R., Olson, W. G., Houghton, P. L., Unruh, J. A. & Corah, L. R. (1992). J. Animal Sci., 70, 3044. Wilson, D. E. (1992). J. Animal Sci., 70,973. Yeates, N. T. M. (1952). Aust. J. Agric. Res., 3, 68.