Animal and Dairy Science Department, University of Georgia, Athens Key Words: Beef Carcass, Composition, Fat, Muscling, Ultrasound

Transcription

1 Gluteus medius and rump fat depths as additional live animal ultrasound measurements for predicting retail product and trimmable fat in beef carcasses 1,2 C. E. Realini, R. E. Williams 3, T. D. Pringle 4, and J. K. Bertrand Animal and Dairy Science Department, University of Georgia, Athens ABSTRACT: This study was conducted to determine the ability of additional ultrasound measures to enhance the prediction accuracy of retail product and trimmable fat yields based on weight and percentage. Thirty-two Hereford-sired steers were ultrasonically measured for 12th-rib fat thickness, longissimus muscle area, rump fat thickness, and gluteus medius depth immediately before slaughter. Chilled carcasses were evaluated for USDA yield grade factors and then fabricated into closely trimmed, boneless subprimals with 0.32 cm s.c. fat. The kilogram weight of end-point product included the weight of trimmed, boneless subprimals plus lean trim weights, chemically adjusted to 20% fat, whereas the fat included the weight of trimmed fat plus the weight of fat in the lean trim. Prediction equations for carcass yield end points were developed using live animal or carcass measurements, and live animal equations were developed including ultrasound ribeye area or using only linear measurements. Multiple regression equations, with and without ultrasound rump fat thickness and gluteus medius depth, had similar R 2 values when predicting kilograms of product and percentages of product, suggesting that these alterna- tive variables explained little additional variation. Final unshrunk weight and ultrasound 12th-rib fat thickness explained most of the variation when predicting kilograms of fat. Rump fat and gluteus medius depth accounted for an additional 10% of the variation in kilograms of fat, compared with the equation containing final weight, ultrasound ribeye area, and ultrasound 12th-rib fat thickness; however, the two equations were not significantly different. Prediction equations for the cutability end points had similar R 2 values whether live animal ultrasound measurements or actual carcass measurements were used. However, when ultrasound ribeye area was excluded from live animal predictions, lower R 2 values were obtained for kilograms of product (0.81 vs 0.67) and percentages of product (0.41 vs 0.17). Conversely, the exclusion of ultrasound ribeye area had little effect on the prediction accuracy for kilograms of fat (0.75 vs 0.74) and percentage fat (0.50 vs 0.40). These data substantiate the ability of live animal ultrasound measures to accurately assess beef carcass composition and suggest that the alternative ultrasound measures, rump fat and gluteus medius depth, improve the accuracy of predicting fat-based carcass yields. Key Words: Beef Carcass, Composition, Fat, Muscling, Ultrasound 2001 American Society of Animal Science. All rights reserved. J. Anim. Sci : Introduction The need for a functioning value-based marketing system has led to increased interest in developing objective methods for accurately assessing carcass traits in live animals. Ultrasound technology is important in the de- 1 This research was partially funded by the National Cattlemen s Beef Association and was part of HATCH project # The authors would like to thank the personnel at the Central Branch Exp. Sta., at Eatonton, GA, and the Whitehall Beef Cattle Unit at Athens, GA, for their management of the steers in this project. 3 Present address: American-International Charolais Association, Kansas City, MO Correspondence: 212 Animal Science Complex (phone: ; fax: ; dpringle@arches.uga.edu). Received August 7, Accepted January 19, velopment of such a system, as it is a fast, repeatable, relatively inexpensive, and nondestructive measure of carcass components before slaughter (Faulkner et al., 1990). Previous studies have shown that ultrasound provides accurate measures of live animal fat thickness and longissimus muscle area (Robinson et al. 1992; Herring et al. 1994a), provided that the technicians collecting and interpreting images are trained and experienced. Furthermore, these live animal measures have been shown to accurately predict beef carcass composition when used in conjunction with live weight (Herring et al., 1994b; Williams et al., 1997). Williams et al. (1997) also investigated alternative measurements as predictors of beef carcass composition and found that biceps femoris depth was a significant variable in models predicting trimmable fat. These authors concluded that 1378

2 Use of alternative ultrasound measures 1379 other easily obtainable live animal measurements need to be investigated as a further aid in the prediction of carcass retail yield. Gluteus medius muscle depth may also have merit as an alternative live animal measure because Johns et al. (1993) suggested that it improves the prediction accuracy of retail product yield, and it can be measured in the same image as rump fat, minimizing the need for additional image collection. The inclusion of additional ultrasound measures, such as rump fat depth and gluteus medius depth, may be beneficial because these measurements are derived from an easily obtained image with potential for automated data collection. Thus, the objective of this research was to determine the value of live animal ultrasound scanning of gluteus medius depth and rump fat depth as additional predictors of retail product and trimmable fat in beef carcasses. Materials and Methods Experimental Design. Thirty-two Hereford-sired steers, approximately 16 mo old, were used to collect live animal ultrasound and carcass composition data for this study. After weaning, all steers were backgrounded for approximately 6 mo before being placed in a feedlot. Initially, steers were fed an adjustment diet, consisting of 78% rolled corn and 15% cottonseed hulls, for approximately 40 d. During the final finishing phase, which was approximately 90 d, the diet consisted of 73% rolled corn, 10% cottonseed hulls, and 10% soybean meal. The finishing phase was terminated when the entire group averaged approximately 1 cm of 12th-rib fat thickness, as determined by ultrasound. Within 3 d of slaughter, steers were measured ultrasonically on the right side by a Beef Improvement Federation-certified technician for 12th-rib fat thickness (UFAT), longissimus muscle area (UREA), rump fat thickness (URUMP), and gluteus medius depth (UGM). Gluteus medius depth was measured immediately below the juncture of the gluteus medius and the biceps femoris muscles between the hook and pin bones with the transducer placed approximately 2.5 cm dorsal to the hook bone and parallel to the backbone (Figure 1). Ultrasound images were collected using an Aloka 500-V ultrasound unit (Corometrics Medical Systems, Wallingford, CT) with a 17-cm, 3.5-MHz linear probe and were interpreted, by the above mentioned technician, using Beef Information Manager software, Version 3.0 (Critical Vision, Inc., Atlanta, GA). Final, unshrunk weight (FINALWT) was also recorded. Steers were then transported to a commercial packing facility where they were held overnight and humanely slaughtered the next day, following normal industry procedures. Carcasses were weighed immediately after slaughter and chilled for 48 h at 4 C. Carcasses were then evaluated for USDA yield grade factors (USDA, 1997) by trained university personnel. Longissimus muscle areas at the 12th rib were traced onto acetate paper and later measured with a planimeter. The right side of each carcass was transported to the University of Georgia for fabrication into boneless subprimals, according to the procedures outlined by Herring et al. (1994b). Kilograms of retail product (KGPROD) was defined as the weight of closely trimmed (0.32 cm s.c. fat), boneless retail cuts from the round, loin, rib, chuck, plate, foreshank, brisket, and flank, plus the lean trim weight from these cuts, which was mathematically adjusted to a 20% fat basis after chemical analysis. Lean trim samples were homogenized and analyzed for chemical fat using the chloroform:methanol procedure of Folch et al. (1957). Kilograms of trimmable fat (KGFAT) was defined as the weight of fat removed when the carcass was trimmed to 0.32 cm of s.c. fat plus the weight of fat from the lean trim samples (based on chemical analysis). Percentage of retail product (PERPROD) and percentage of fat (PERFAT) were obtained by expressing KGPROD and KGFAT as a percentage of cold carcass weight, respectively. Statistical Analysis. All statistical analyses were conducted using SAS (SAS Inst. Inc., Cary, NC). Pearson product-moment correlations were estimated between live animal and carcass traits, and KGPROD, PER- PROD, KGFAT, and PERFAT. All prediction equations were developed by regression procedures using either live animal or carcass measurements as the independent variables. Of primary interest in this study was the inclusion of two new variables, URUMP and UGM, to the current live animal measurements, UFAT and UREA, used to predict carcass yields. Therefore, the initial regression equations reported in Table 4 include FI- NALWT, UREA, and UFAT representing the variables that are most often used by the beef industry to predict live animal carcass yields (Equation 1). Equation 2 includes the variables from Equation 1 along with UR- UMP, whereas Equation 3 includes the variables from Equation 1 along with UGM, and Equation 4 includes all of the above-mentioned variables. In Table 5, only linear measurements are considered. Equation 1 includes FINALWT and UFAT; Equation 2 includes FI- NALWT, UFAT and URUMP; Equation 3 includes FI- NALWT, UFAT and UGM; and Equation 4 includes all of the linear variables. The F-test statistic used to determine whether the addition of a new variable in the model significantly reduced the sums of squares for error compared with ignoring it was (SSE RM SSE FM )/MSE FM, where SSE FM and SSE RM are the sums of squares for error of the full model and reduced model, respectively, and MSE FM is the mean squared error for the full model. The degrees of freedom for the numerator in the above expression were derived from the difference in the number of variables between the two models. The denominator s degrees of freedom were the degrees of freedom associated with MSE FM (McClave and Dietrich, 1982). Carcass trait prediction equations were developed as a comparison to equations developed using live animal measurements. Carcass measurements used in the prediction equation were longissimus muscle area (CREA); adjusted 12th-rib backfat fat (ADJCFAT); estimated percentage of kidney, pelvic, and heart fat (CKPH); and

3 1380 Realini et al. Figure 1. Ultrasound rump image collected approximately 2.5 cm dorsal to the hook bone (ilium) and parallel to the vertebral column. Reference points A and B show the apex of the biceps femoris muscle and the ilium, respectively; and measurement sites C and D show the measurement for rump fat and gluteus medius depth, respectively. hot carcass weight (HCW), which are the same measurements used in the USDA yield grade equation (USDA, 1997). All equations were reported on a whole-carcass basis, and genetic and environmental effects were not considered in the modeling process. Results and Discussion Descriptions of the acronyms assigned to variables are given in Table 1. Means, standard deviations, and minimum and maximum values for live animal and carcass traits are presented in Table 2. Note that the cattle used in this study produced considerably fatter and lower-yielding carcasses than the average of the industry reported in the most recent National Beef Quality Audit (Smith et al., 1995). Means for UFAT were lower than for ADJCFAT (1.25 vs 1.62 cm); however, the simple correlation coefficients showed a moderately high positive relationship between UFAT and ADJCFAT (r = 0.79; P < 0.01), URUMP and ADJCFAT Table 1. Description of acronyms Acronym Definition FINALWT Final unshrunk weight, kg UFAT Ultrasound rib fat thickness, cm UREA Ultrasound longissimus muscle area, cm 2 URUMP Ultrasound rump fat thickness, cm UGM Ultrasound gluteus medius thickness, cm ADJCFAT Adjusted carcass fat thickness, cm CKPH Estimated kidney, pelvic, and heart fat, % CREA Carcass longissimus muscle area, cm 2 HCW Hot carcass weight, kg KGPROD Weight of boneless retail cuts from the round, loin, rib, chuck, plate, foreshank, brisket, and flank trimmed to 0.32 cm s.c. fat thickness and weight of adjusted lean trim (20%) from those cuts PERPROD Boneless retail cuts from the round, loin, rib, chuck, plate, foreshank, brisket, and flank trimmed to 0.32 cm s.c fat thickness and adjusted lean trim from those cuts, expressed as a percentage of cold carcass weight KGFAT Weight of trimmed fat from retail cuts, trimmed to 0.32 cm of s.c. fat, plus the weight of fat from the lean trim PERFAT Trimmed fat from retail cuts, trimmed to 0.32 cm of s.c. fat, plus the fat from the lean trim, expressed as a percentage of cold carcass weight

4 Use of alternative ultrasound measures 1381 Table 2. Means, standard deviations, and minimums and maximums for live animal ultrasound measurements, carcass traits, and cutability end points (n = 32) a Trait Mean SD Minimum Maximum Live FINALWT, kg UFAT, cm UREA, cm URUMP, cm UGM, cm Carcass HCW, kg COLDWT, kg ADJCFAT, cm CREA, cm CKPH, % End point USDA Yield Grade KGPROD, kg PERPROD, % KGFAT, kg PERFAT, % (r = 0.69; P < 0.01), as well as between URUMP and UFAT (r = 0.54; P < 0.05). Ultrasound estimates of UREA were lower than CREA values (73.0 vs 75.9 cm 2 ), showing a moderately high correlation (r = 0.69, P < 0.01). This corresponds with work by Henderson- Perry et al. (1989), which indicated that ultrasonic measures underestimate longissimus muscle area. Ultrasound estimates of UGM tended to correlate with CREA (r = 0.30; P < 0.10); however, the association was much lower than between UREA and CREA. Previous research using ultrasound has shown similar correlations between ultrasound and carcass measurements for fat thickness (r = 0.81 to 0.86) and longissimus muscle area (r = 0.61 to 0.76) (Henderson-Perry et al., 1989; Stouffer et al., 1989; May et al., 2000). All ultrasound and carcass measurements were significantly correlated with KGPROD, ranging from 0.30 for URUMP to 0.90 for HCW (Table 3). Hot carcass weight and FINALWT (r = 0.77) were the variables most closely correlated with KGPROD, followed by UREA and CREA. Fat measurements showed significant but lower correlations with KGPROD than muscle measurements. Correlations between live animal and carcass traits with PERPROD were lower and less significant than with KGPROD, ranging from 0.03 for UGM to 0.45 for CREA. Although the correlations between UGM and PERPROD were low in this study, Johns et al. (1993) reported a correlation of 0.42 between a similar ultrasonic measure of the gluteus medius muscle depth and percentage of separable lean. In addition, other researchers have reported that ultrasound and carcass measurements are generally more highly correlated to weight-based carcass yield characteristics than to those based on percentage (Johns et al., 1993; Herring et al., 1994b; Williams et al., 1997). Fat measurements (UFAT, URUMP, AD- Table 3. Pearson product-moment correlations between carcass cutability end points and live animal and carcass traits a Trait KGPROD PERPROD KGFAT PERFAT FINALWT 0.77** ** 0.34* UFAT 0.46** ** 0.48** UREA 0.67** 0.40* URUMP * 0.64** 0.58** UGM 0.53** ** 0.21 CREA 0.64** 0.45** HCW 0.90** ** 0.44** ADJCFAT 0.47** ** 0.54** CKPH 0.39* ** 0.52** USDA Yield Grade 0.38* 0.48** 0.76** 0.68** P < *P < **P < 0.01.

5 1382 Realini et al. Table 4. Regression equations for predicting kilograms and percentage of retail product and trimmable fat using live animal measurements a Partial regression coefficients Dependent variable and equation SE b R 2 Intercept FINALWT, kg UREA, cm 2 UFAT, cm URUMP, cm UGM, cm KGPROD * 1.09* * 1.06* * 1.02* * 1.00* PERPROD * * * * KGFAT * * * * * * * * 9.72* 5.41* PERFAT * * 3.19* * * 1.13 b Root mean square error for the model. P < *P < JCFAT, CKPH) showed a negative relationship with PERPROD and a positive relationship with KGFAT and PERFAT (Table 3). Ultrasound rump fat thickness and UFAT were more closely correlated with PER- PROD than ADJCFAT. As expected, fat measurements showed stronger correlations than muscle measurements with KGFAT and PERFAT, with ADJCFAT being most closely correlated with KGFAT, and UR- UMP with PERFAT. Multiple regression equations were developed to examine the ability of live animal measurements to predict the weight in kilograms and the percentage of retail product and trimmable fat (Table 4). In agreement with our correlation analysis and prior studies (Herring et al., 1994b; Williams et al., 1997), the accuracy of predicting KGPROD and KGFAT was greater than for PERPROD and PERFAT. Also, the inclusion of URUMP and(or) UGM explained little additional variation in KGPROD and PERPROD, suggesting limited application of these alternative variables as live animal predictors of carcass composition. These results are somewhat inconsistent with Tait et al. (2000), who suggested that the inclusion of ultrasound gluteus medius depth could increase the prediction accuracy for the weight of retail product from the beef round. In our study, the variables having the strongest influence on the prediction of KGPROD were FINALWT and UREA, followed by UFAT, whereas UREA had the strongest influence on the prediction of PERPROD. The importance of UREA in predicting these variables is in contrast to Hamlin et al. (1995), who indicated that longissimus muscle area was not an accurate predictor of retail product percentage, accounting for less than 15% of the variation. Researchers have reported that 12th-rib fat thickness over the longissimus muscle is the most accurate measurement for determining beef carcass composition (Murphey et al. 1960; Crouse et al., 1975); however, other researchers have reported that measurements of fat at different locations are useful for predicting lean and fat in beef cattle (Wallace et al., 1977; Johns et al., 1993). Williams et al. (1997) reported that URUMP was consistently a significant variable in equations predicting retail product, expressed either on a weight or percentage basis. This is in contrast to the results of our study, in which URUMP was not a significant variable in equations predicting kilograms or percentage of retail product (Table 4). The reason for the discrepancy between this study and other literature may be related to differences in population variation between the studies. The variables UFAT and FINALWT explained the majority of the variation in KGFAT. Ultrasound rump fat depth and UGM explained an additional 3 to 4% of the variation in KGFAT, individually, and an additional 10% collectively, compared with Equation 1 containing FINALWT, UREA, and UFAT. In agreement with Williams et al. (1997), the prediction of PERFAT using ultrasound measures was less accurate than KGFAT. Although the addition of URUMP and UGM explained

6 Use of alternative ultrasound measures 1383 Table 5. Regression equations for predicting weight in kilograms and percentage of retail product and trimmable fat using linear live animal measurements a Partial regression coefficients Dependent variable and equation SE b R 2 Intercept FINALWT, kg UFAT, cm URUMP, cm UGM, cm KGPROD ** ** ** ** PERPROD KGFAT ** 16.77** ** 12.61* ** 16.41** * 10.63* 10.03* 4.96* PERFAT * * * * 0.73 b Root mean square error for the model. P < *P < approximately 15% more of the variation in PERFAT, the best model still accounted for less than half of the variation. The addition of URUMP and UGM to Equation 1 did not significantly reduce the standard error for predicting KGFAT; however, for PERFAT, models that included URUMP and UGM were different (P < 0.05) from the model containing the variables FI- NALWT, UREA, and UFAT (Table 4). These findings suggest that both URUMP and UGM can enhance the prediction accuracy for fat-based yield traits, particularly when yield is measured as a percentage. Generally, ultrasound is more accurate at predicting fat thickness, a one-dimensional measurement, than longissimus muscle area. Research has shown that accuracy is highly dependent on the technician and the level of experience (Perkins et al., 1992). Herring et al. (1994a) identified two main sources of error during the ultrasound data collection process: image acquisition and image interpretation. Variation between technicians in image interpretation according to their experience was found to be a significant source of variation. Hassen et al. (1998) studied the accuracy and repeatability of ultrasound measurements of fat thickness and longissimus muscle area and indicated that there was a clear technician difference in the magnitude of the variance components for measurements of longissimus muscle area. Similarly, Houghton and Turlington (1992) and Perkins et al. (1992) reported that the accuracy of ultrasound longissimus muscle area is inconsistent among technicians and ultrasonic instrumentation. Thus, gluteus medius muscle depth was selected for investigation because it is a linear measurement of muscle that can be obtained from the same image as URUMP (Figure 1). In addition, these linear measure- Table 6. Regression equations for predicting kilograms and percentage of retail product and trimmable fat using carcass measurements a Partial regression coefficients Dependent variable R 2 Intercept HCW, kg CREA, cm 2 ADJCFAT, cm CKPH, % KGPROD * 0.67* PERPROD * KGFAT * 0.34* * PERFAT * P < *P < 0.05.

7 1384 Realini et al. ments may allow automation of image collection and analysis. In order to evaluate this theory, regression analysis was conducted using only single-dimension measurements (Table 5). Prediction equations of kilograms and percentage of retail product using linear measurements had lower R 2 values than equations including UREA. Model 4 from Table 4 explained about 13% more of the variation in KGPROD, and 24% more of the variation in PERPROD, than Model 4 from Table 5. The full model that contained UREA was different (P < 0.05) from the reduced model that dropped UREA. When UREA was excluded from the models estimating KGPROD, UGM became slightly significant. Prediction equations for KGFAT and PERFAT, using linear measurements (Table 5) had R 2 values similar to those of equations including longissimus muscle area (Table 4). These results suggest that the use of one-dimensional ultrasound measurements to predict fat-based yield traits, in particular KGFAT, is similar in accuracy to the use of single- and multidimensional measures combined. Additional data needs to be collected on a population of greater size and variation in order to substantiate these findings. As a comparison between ultrasound-derived predictions and actual carcass measurement predictions, regression equations for KGPROD, PERPROD, KGFAT and PERFAT using the same variables currently found in the USDA yield grade equation (USDA, 1997) are reported in Table 6. Hamlin et al. (1992) indicated that live animal ultrasonic prediction of beef carcass yield is slightly less predictive than carcass measurements. However, Bullock et al. (1991), Herring et al. (1994b), and Williams et al. (1997) have reported that live animal measurements are as predictive as carcass measurements for retail yield or percentage lean of the beef carcass. In the present study, the best models using live animal measurements had R 2 values similar to models using actual carcass measurements when predicting retail product weight (0.81 vs 0.87, respectively) and percentage (0.41 vs 0.40), and trimmable fat weight (0.75 vs 0.82) and percentage (0.50 vs 0.50) (Tables 4 and 6). The best models using only linear live animal measurements accounted for approximately 20% and 23% less of the variation in KGPROD and PERPROD, respectively; however, linear live animal equations (Table 5) were similar to carcass-based equations for trimmable fat weight and percentage (Table 6). Implications This research indicates that ultrasound is an effective method of measuring 12th-rib fat thickness and longissimus muscle area and that these measurements can be combined with other live measurements to estimate retail yield and trimmable fat with a degree of accuracy that is similar to predictions based on carcass measurements. Furthermore, it seems that inclusion of the alternative ultrasound measures, rump fat thickness and gluteus medius depth, may be beneficial for ultrasound data collection. These measures account for additional variation in trimmable carcass fat and are derived from an easily obtained image with potential for automated data collection. Further research is warranted to validate the ability of these additional ultrasound measurements to improve the prediction of carcass yields and should focus on cattle that are representative of the current industry. Literature Cited Bullock, K. D., J. K. Bertrand, L. L. Benyshek, S. E. Williams, and D. G. Lust Comparison of real-time ultrasound and other live measures to carcass measures as predictors of beef cow energy stores. J. Anim. Sci. 69: Crouse, J. D., M. E. Dikeman, R. M. Koch, and C. E. Murphey Evaluation of traits in the U.S.D.A. yield grade equation for predicting beef carcass cutability in breed groups differing in growth and fattening characteristics. J. Anim. Sci. 41: Faulkner, D. B., D. F. Parrett, F. K. McKeith, and L. L. Berger Prediction of fat cover and carcass composition from live and carcass measurements. J. Anim. Sci. 68: Folch, J., M. Lees, and G. S. H. Stanley A simple method for the isolation and purification of lipids from animal tissues. J. Biol. Chem. 226: Hamlin, K. E., R. D. Green, L. V. Cundiff, M. E. Dikeman, and T. L. Perkins Relationship between real-time ultrasonic estimates and estimates of carcass merit and measures of retail yield in diverse biological types of beef steers. Texas Tech. Univ. Agric. Sci. Tech. Rep. No. T-5-317, Lubbock. Hamlin, K. E., R. D. Green, L. V. Cundiff, T. L. Wheeler, and M. E. Dikeman Real-time ultrasonic measurement of fat thickness and longissimus muscle area: II. Relationship between realtime ultrasound measures and carcass retail yield. J. Anim. Sci. 73: Hassen, A., D. E. Wilson, R. L. Willham, G. H. Rouse, and A. H. Trenkle Evaluation of ultrasound measurements of fat thickness and longissimus muscle area in feedlot cattle: Assessment of accuracy and repeatability. Can. J. Anim. Sci. 78: Henderson-Perry, S. C., L. R. Corah, and R. C. Perry The use of ultrasound in cattle to estimate subcutaneous fat thickness and ribeye area. J. Anim. Sci. 67(Suppl. 1):433 (Abstr.). Herring, W. O., D. C. Miller, J. K. Bertrand, and L. L. Benyshek. 1994a. Evaluation of machine, technician, and interpreter effects on ultrasonic measures of backfat and longissimus muscle area in beef cattle. J. Anim. Sci. 72: Herring, W. O., S. E. Williams, J. K. Bertrand, L. L. Benyshek, and D. C. Miller. 1994b. Comparison of live and carcass equations predicting percentage of cutability, retail product weight, and trimmable fat in beef cattle. J. Anim. Sci. 72: Houghton, P. L., and L. M. Turlington Application of ultrasound for feeding and finishing animals: A review. J. Anim. Sci. 70: Johns, J. V., P. O. Brackelsberg, and M. J. Marchello Use of real-time ultrasound to determine carcass lean and fat in beef steers from various live and carcass measurements. Iowa State Univ. Beef and Sheep Res. Rep. A.S. Leaflet R1020, Ames. May, S. G., W. L. Mies, J. W. Edwards, J. J. Harris, J. B. Morgan, R. P. Garrett, F. L. Williams, J. W. Wise, H. R. Cross, and J. W. Savell Using live estimates and ultrasound measurements to predict beef carcass cutability. J. Anim. Sci. 78: McClave, J. T., and F. H. Dietrich II Statistics. 2nd ed. Dellen Publishing Co., San Francisco, CA. Murphey, C. E., D. K. Hallett, W. E. Tyler, and J. C. Pierce, Jr Estimating yields of retail cuts from beef carcasses. J. Anim. Sci. 19(Suppl. 1):1240 (Abstr.).

8 Use of alternative ultrasound measures 1385 Perkins, T. L., R. D. Green, and K. E. Hamlin Evaluation of ultrasonic estimates of carcass fat thickness and longissimus muscle area in beef cattle. J. Anim. Sci. 70: Robinson, D. L., C. A. McDonald, K. Hammond, and J. W. Turner Live animal measurement of carcass traits by ultrasound: Assessment and accuracy of sonographers. J. Anim. Sci. 70: Smith, G. C., H. G. Dolezal, and J. W. Savell The Final Report of the National Beef Quality Audit National Cattlemens Beef Assoc., Inglewood, CO. Stouffer, J. R., T. C. Perry, and D. G. Fox New techniques for real-time ultrasonic evaluation of beef cattle. J. Anim. Sc. 67(Suppl. 1):120 (Abstr.). Tait, J. R., G. H. Rouse, D. E. Wilson, C. L. Hays Prediction of lean in the round using ultrasound measurements Iowa State Univ. Beef and Sheep Res. Rep. A.S. Leaflet R1733, Ames. USDA Official United States Standards for Grades of Carcass Beef. Agriculture Marketing Service, USDA, Washington, DC. Wallace, M. A., J. R. Stouffer, and R. G. Weservelt Relationships of ultrasonic and carcass measurements with retail yield in beef cattle. Livest. Prod. Sci. 4: Williams, R. E., J. K. Bertrand, S. E. Williams, and L. L. Benyshek Biceps femoris and rump fat as additional ultrasound measurements for predicting retail product and trimmable fat in beef carcasses. J. Anim. Sci. 75:7 13.