Application of Ultrasound for Feeding and Finishing Animals: A Review' P. L. Houghton2 and L. M. Turlington3 Kansas State University, Manhattan 66506

ABSTRACT:

The ability to evaluate carcass traits in live animals is of value to research, educational, and industry personnel. Ultrasonic technology has been tested since the early 1950s and continues to be under investigation as a means of accomplishing this task. The accuracy of ultrasound in predicting carcass traits is variable and is dependent on species, ultrasonic instrumentation, and(or1the skill of the technician. Based on this review, the ranges of correlation coefficients (rl for carcass traits as predicted by ultrasound to the respective carcass measurement are as follows: swine (fat .20 to 3 4 ; longissimus muscle .27 to .931, sheep (fat .42 to .95; longissimus muscle .36 to .791 and beef @at.45 to .96; longissimus muscle .20 to .94; marbling .20 to .911. Although these correlation coefficients give an indication of the accuracy of ultrasound, it should be noted that these statistics do not reflect population variation or bias. Applications of ultrasound in swine finishing programs include the successful prediction of market weight carcass characteristics and the

prediction of percentage of lean cuts before slaughter. In contrast, the application of ultrasound in lamb finishing programs has met with limited success. Most data indicate that weight and(or1 visual estimations of fat are at least as accurate as ultrasound predictions of carcass composition. In beef finishing programs, ultrasound has, at times, been used successfully to predict fat and muscle traits before slaughter and beef carcass chemical composition. The ability to predict marbling, however, remains unclear and requires further investigation. Ultrasound has also been used in beef fdshing programs to predict days on feed to a constant body compositional end point. When summarized, these data indicate that a single ultrasonic measurement of fat can be helpful in predicting days on feed in yearling cattle. When used alone, however, a single backfat measurement does not provide adequate accuracy. Therefore, factors such as age, sex, breed type, weight, and hip height are needed to help predict days on feed more accurately.

Key Words: Ultrasound, Carcass, Cattle, Sheep, Pigs

J. Anim. Sci. 1992. 70:930-941

Introduction As the livestock industry moves closer to the concept of value-based marketing, producers are becoming more concerned about carcass traits. Although livestock producers are realizing the importance of carcass trait predictability, they are

'Contribution No. 91-129-E, Kansas State Univ. Agric. Exp. Sta., Manhattan 66506. Presented at a symposium titled

"Application of Ultrasound in Animal Science Research" at the ASAS 82nd Annu. Mtg., Ames, LA. 2Heartlaad Cattle Co., Rt. 3, Box 134, McCook, NE 69001. 3Central Soya Company, Inc., 1200 N. Second St., Decatur. IN 46733. Received November 5 , 1990. Accepted March 28, 1991.

faced with a dilemma because of the lack of accurate methods for measuring carcass value prior to slaughter. In addressing this problem, livestock producers have become interested in ultrasound as a method of determining fat thickness and muscle development in the live animal. The use of ultrasonics was first reported by Wild (19501, who stated that the ultrasonic technique is nondestructive and humane and provides a means of quantifying muscle and fatty tissues in live animals. The procedure for using ultrasound involves the application of a mineral oil to the area of the body to be measured, followed by the placement of a sensor or transducer on the chosen area. The basic principle of ultrasound is to measure an echo rebounding from soft tissues. After the transducer is placed in contact with the

930

Downloaded from https://academic.oup.com/jas/article-abstract/70/3/930/4705828 by Washington University in St. Louis user on 03 July 2018

PREDICTION OF CARCASS TRAITS BY ULTRASOUND

animal, the ultrasound equipment transfers electrical pulses to high-frequencysound waves, hence the name ultrasound. These sound waves travel into the body and are reflected from boundaries between different densities of tissues. The image that the ultrasound waves transmit back through the transducer is projected onto the screen of the ultrasound unit and the appropriate measurements are made.

Properties of Ultrasound Sound is a mechanical wave of compressions and refractions within a medium. A sound wave can be compared to a longitudinal wave having a length, frequency, and velocity. The wavelength is the distance of two similar points on a given wave. Frequency is the number of cycles or wavelengths occurring in a given time period (usually 1 SI. Velocity is derived from the computation of frequency and wavelength. Frequency is described in cycles/second or hertz (Hz). Audible sound varies from 20 to 20,000 Hz. Diagnostic ultrasound uses frequencies in the range of 2 to 10 MHz, which is well beyond the range of audible sound. If one knows the velocity and frequency, the wavelength can be calculated. Because the velocity of sound in a given tissue is constant, changing the frequency will change the wavelength. This will, in turn, affect the resolution and quality of an ultrasound image (Rantanen and Ewing, 1981; Herring and Bjornton, 1985). Diagnostic ultrasound is produced by transducers housing crystals with piezoelectric (pressureelectric) properties. When piezoelectric crystals are deformed by pressure, electricity is produced. Conversely, when an electric current is applied to them, the crystals will deform. This is the process by which ultrasound is generated and received by the transducer. When reflected, sound returns to the transducer and a slight deformation of the crystal is produced; this generates an electric current. The current is displayed on an oscilloscope as an image of the tissue interfaces (Rant% nen and Ewing, 1981).

Interaction of Sound Waves with Tissues As the sound beam passes through body tissue, a portion of the beam is reflected back to the transducer. Reflection occurs at tissue interfaces of differing acoustic impedance. The amplitude of the returning echo is determined by the absolute difference in acoustic impedance of one tissue compared to another. The closer the acoustic impedance of one tissue to a second tissue, the Downloaded from https://academic.oup.com/jas/article-abstract/70/3/930/4705828 by Washington University in St. Louis user on 03 July 2018

931

smaller the returning echo (Herring and Bjornton, 1985).

Each echo returns to the transducer, where it is changed into an electrical pulse and is displayed on a cathode-ray tube screen. The ultrasound scanner calculates the time it takes for a pulse to be emitted and the echo to be returned, which allows it to compute the exact distance of the acoustic interface from the transducer. Sound beams travel at approximately 1,540 m / s in soft tissue. Therefore, the only variable that contributes to the difference in acoustic impedance of one soft tissue to another is its density. When two tissues of different density are in contact with one another, this creates an acoustic interface or a reflecting surface (Herring and Bjornton, 1985). Sound travels through bone at approximately 3,100 m/s. The density of bone is considerable compared with that of soft tissue, through which sound travels at 1,540 m/s. Therefore, a very high impedance mismatch occurs at a soft tissue-tobone interface. (Herring and Bjornton, 1985). The absolute value of acoustic impedance of any tissue is relatively unimportant, but it is the magnitude of the difference in acoustic impedance at tissue interfaces that determines the amount of reflection of the beam (Rantanen and Ewing, 1981). Energy is removed from the sound beam as it passes through soft tissues. This energy removal is referred to as attenuation. Attenuation is caused by two predominant processes. The first process is absorption, which is the conversion of ordered motion of ultrasound into the disordered motion of heat. The amount of absorption increases with the frequency of the sound beam. The second process is scattering of the sound beam by small tissue interfaces, which results in energy loss from the sound beam. The intensity of the scattered sound increases with its increasing frequency. Because factors causing absorption and scattering of the beam are frequency-dependent, a sound of lower frequency will penetrate further into soft tissue than a higher frequency sound. (Rantanen and Ewing, 1981). For this reason, a 3-MHZ transducer is more appropriate to use for deeper locations in the body &e., muscle areal, whereas a 5-MHZ transducer is conducive for analyzing tissues close to the body surface &e., fat thickness).

Display Formats There are three basic display formats of modes. The first, called amplitude mode (A-mode),ultrasonic imaging is a one-dimensional display of returning echo amplitude and distance. This mode consists of vertical peaks along a horizontal axis. The height of the peak corresponds to the ampli-

HOUGHTON AND TURLINGTON

932

Table 1. A summary of ultrasound accuracy in swine, sheep, and beef cattle

Investigator and year

Instrumentation

Stouffer et al., 1981

Sperry reflectoscope

Measurement location*

Accuracy

kl

Swine

Gillis et al., 1972 Giles et al., 1981

A-mode ultrasonics Sonatest Scanoprobe scanogram

12th Rib fat

.92

LMA LM width LM width LMA

.72

BF BF BF

.eo

LMA Mersmann, 1982

Scanogram model 722

BF at 1/5, 1/2, and 3/4 Body length

LMA Forrest et al., 1989

Technicare 210 DX

1st Rib fat

Last rib fat Last lumbar fat lath Rib fat

LMA Lopes et al.,1987

Technicare 210 DX

10 Rib fat

Last rib fat

LMA McLaren et al., 1989

Technicare 210 DX

10th Rib fat Last rib fat Avg BF

LMA Turlington, 1990

Technicare 210 DX

Campbell et al., 1959 Moody et al., 1985 Thompson et d.,19i"

Somascope Model 5 scanogram

Gooden et al., 1980 Kempster et al., 1982

ADD prototype

Edwards et al.. l98@

scanogram

Danscanner Technicare 210 DX

1st Rib fat Last rib fat Last lumbar fat 10th Rib fat LMA

LMA LMA 12th Rib fat Tuber coxaa fat Last rib fat

LMA LMA 12th Rib fat

LMA Turlington, 1990

Technicare 210 DX

Last rib midline fat Last rib 3/4 fat Dock fat

LMA Harada and Kumazaki. 1980

scanogram

5th to 8th Rib fat

LMA Harctda et al., 1985 Stouffer and Cross, 1085

Kaijo B-mode

Marbling 12 Rib fat

Scanningacope

LMA

Technicare 210 DX

Rouse and Parrish, 1987

G. E. Datason

Smith et al.,1988

Technicare 210 DX

Turner. 1988

Technicare 210 DX

Marbling Rump fat 12th Rib fat LMA Marbling Attenuation

12th Rib fat LMA 12th Rib fat

LMA

.47 .68

.46 to 32 .75 to .80 94 .75 .20 to .e1

-

.49 .54 .85 .85

.71 .85 to .80 to .75 to .27 to .55

.88

.89 .8@ .70

.55 .82 .8 1 .74 to .83

.eo

.eo .88 to .93 .91 to .93 .44 to .79 -52to .68 .74 .84 .80 to .91 .50 .48 .82 .36 .48

.a .42 .58 .a5 to 9 7 .83 to .92 .77 to .e1

.83 .86 .78

.53 .78

.87 .21 .54 CMbI .53 a3 .8 1

.20 to .43 .81 to .94 .71 to .@4

continued

Downloaded from https://academic.oup.com/jas/article-abstract/70/3/930/4705828 by Washington University in St. Louis user on 03 July 2018

PREDICTION OF CARCASS TRAITS BY ULTRASOUND

933

Table 1. (continued) A summary of ultrasound accuracy in swine, sheep, and beef cattle Faullrner et al., 1989 Hale, 1989

Technicare 210 DX Technicare 210 DX

12th Rib fat 12th Rib fat

LMA Stouffer et al., 1989

Technicare 210 DX

12th Rib fat

LMA

Perry et al., 1989

G. E. Datason

Perry et al., 1989

G. E. Datason

Strasia et al., 1989 Brethour, 1990

Technicare 210 DX Technicare 210 DX

Duello et al., 1990

Aloka 633

Shoulder fat Foreflank fat Rump fat 12th Rib fat

Equisonia

Houghton et al., 1989

Technicare 210 DX

-

LMA

.80 .62 CMb & LI' .55 22 to .77 M b )

12th Rib fat

-

a&

& Le)

.85 .7 1

.87

12th Rib fat

Mb)

.75 .20 to .40

.78 .96 .90 .82 .63

LMA

(Mb& Lc)

.45 to .70 .87

12th Rib fat 12th Rib fat

Tecbnicare 210 DX

.56

LMA LMA Smith et al., 1990

.45 .BO 82

Attenuation If BF c 2 cm Attenuation 12th Rib fat Speckle score before slaughter Speckle score > 30 d before slaughter 12th Rib fat

G. E. Datason

.86 .76

.86 .76

LMA Perry et al., 1990

.00 to .83 .74 to .82

LMA

LMA Lvalue Henderson-Perry, 1990

.89

= longissimus muscle area BF backfat. bMarbling score as determined by visual appraid of the carcass. 'Lipid content as determined by chemical analyses.

tude of the echo (Rantanen and Ewing, 1981; Herring and Bjornton, 1985). Brightness mode (B-mode)ultrasonic imaging is another display format; it is a two dimensional display of dots. The transducer is moved across the surface of the body and a cross-sectional anatomy is depicted. The position of the dot on the screen is determined by the time it takes for an echo to return to the transducer. The brightness of the dots is proportional to the amplitude of the returning echoes. Real-time ultrasonic imaging is a form of Bmode used to record movement of structures. In real-time imaging, echoes are recorded continuously on a non-storage cathode-ray display screen. Encoders spatially orient the returning echoes on the display screen to depict tissue interfaces. With real-time units, these encoders are contained in the movable head to allow rapid transducer movement from one area to another (Rantanen and Ewing, 1981; Herring and Bjornton, 1985). The third display format is that of motion mode ultrasound (time motion; M- or TM-mode) and is a one-dimensional format that displays dots. With

Downloaded from https://academic.oup.com/jas/article-abstract/70/3/930/4705828 by Washington University in St. Louis user on 03 July 2018

TM-mode, the transducer is held in place over moving organs. The display is printed on an oscilloscope or mooing strip of white sensitive paper. This form of ultrasonics is used primarily in echocardiographic studies (Rantanen and Ewing, 1981; Herring and Bjornton, 19851.

Accuracy of Ultrasound Ultrasound technology was introduced early in the 1950s as a means for estimating compositional differences among livestock (Wild, 1950; Claus, 1957; Panier, 1957; Price et al., 1958; Hazel and Kline, 1959). Technological advances during the latter part of the 1970s and early 1980s have dramatically improved ultrasound equipment. Ultrasound units that produce real-time images are now the most commonly accepted ultrasonic instrumentation for use with livestock. Nevertheless, a number of different ultrasonic units are included in a review of ultrasound accuracy, which is summarized by species (swine, sheep, and beef cattle) in Table 1.

934

HOUGHTON AND TURLINGTON

Table 2. Frequency comparison of ultrasonically measured backfat and longissimus muscle area (LMA) in live pigs to corresponding carcass measurements* LMA

10th-Rib backfat f cm .25 .50 .75 1.00

Cumulative 0,4b 75 90 95 98

*

cm2

.63 1.25 1.68 2.50

Table 3. Frequency comparison of ultrasonically measured backfat and longissimus muscle area (LMA) in live lambs to corresponding carcass measurementsa LMA

12th-FLib 3/4 backfat

cumulative okb

f cm

Cumulative %b

43 59 76

.10

69

.20

93

.30

85

.40

99 100

sSeventy-five pigs; killed within 1 d after;-s ultra sonic instrumentation Technicare 210 DX; Turlmgton, 1990. h o t d percentage of pigs whose ultrasonic measurement was within the indicated range of the actual corresponding carcass measurement.

*

cm2

.20 .40

30 .80

h a t i v e

d

61 68 71 100

aOne hundred sixty lambs; killed within 2 d after scanninp., ultrasonic instrumentation Technicare 210 DX; Turlington, 1990.

bTotal percentage of lambs whose ultrasonic measurement was within the indicated range of the achral corresponding carcass measurement.

Almost all the accuracy data reported to date have been in the form of correlation coefficients (1-1. Although correlation coefficients may be useful in certain situations, it is also important to understand the limitations associated with this method of reporting ultrasound accuracy. These limitations include 1) the fact that population variation influences correlation coefficients Le., a larger than normal variation will produce high correlation coefficients, whereas a uniform population will result in much lower correlation coefficients); 2) correlation coefficients do not reflect bias (Le., an ultrasonic technique that consistently over or underestimates measurements); and 3) correlation coefficients are not easily understood by most producer groups. With these limitations in mind, alternate methods of reporting accuracy data should be considered. One method is to report data in the form of a frequency distribution. Tables 2,3, and 4 demonstrate this method and report recent ultra sonic accuracy information for swine (Table 21, sheep (Table 31, and beef cattle (Table 41. Regardless of the method used to report accu racy data, it is evident that there is considerable variation between species, technicians, and ultrasonic instrumentation in the ability of ultrasound to predict carcass traits. For example, much variation exists in the ability to accurately predict longissimus muscle area between the 12th and 13th ribs in beef cattle. Possible reasons for the increased variation of this measurement are explained by Stouffer (19881, who suggests that 1) dirt, hide thickness, and hair; 21 degree of fat thickness at the 12th rib; 3) the ability to match the medial and lateral halves of the longissimus muscle when using a split-screen technique; and 41 parallel interfaces to ultrasonic sound waves ke., medial and lateral boundaries of the longissimus muscle area) may contribute to the lower accuracies associated with this measurement. Downloaded from https://academic.oup.com/jas/article-abstract/70/3/930/4705828 by Washington University in St. Louis user on 03 July 2018

Because ultrasonically measured longissimus muscle area is, on the average, only moderately related to carcass longissimus muscle area, it would be a mistake to have a high level of confidence in an individual longissimus muscle measurement. Using the data in Table 4 as an example, it is possible to predict longissimus muscle area within 6.25 cm2 79% of the time; however, this means that 21% of the time longissimus muscle area was over- or underpredicted by more than 6.25 cm2 and about 12% of the time longissimus muscle area was missed by over 9.38 cm2. If these data are representative, individual longissimus muscle measurements, as estimated by ultrasound, may not be accurate enough for commercial or research purposes. In contrast, when these measurements are used to evaluate a group or treatment, the data are probably useful. For example, it is evident that error exists in this technology, but if it is assumed that the error is random across groups and enough livestock are measured, it is probably safe to say that, on the average, the data obtained from ultrasound measurements are relatively accurate. When discussing the accuracy of ultrasound, there is another point that should be considered that relates to the correlation between ultrasound measurements of backfat and longissimus muscle area to their corresponding carcass measurements. The question should be asked, How important is this relationship? The reason we have always taken these carcass measurements is because they are indicators of total carcass muscle and leanness. With this in mind, it may be more appropriate to correlate ultrasound measurements of backfat and longissimus muscle area to total carcass muscle or lean muscle mass, rather than to the carcass measurements themselves. In addition, it may be possible to identify other muscle groups that would give a better

PREDICTION OF CARCASS TRAITS BY U L M O U N D

Table 5. Carcass trait measurements in the live animal, hanging carcass, and standing carcassa

Table 4. Frequency comparison of ultrasonically measured backfat and longissimus muscle area (LMA) in live steers to corresponding carcass measurementsa Backfat

LMA

f cm

cumulative ohb

.10 .20 .30 .40 .50

58 84 90

98 100

f cm2

3.12 6.25 9.38 12.50 15.62

cumulative %b 32 79 88 97 100

%ne hundred twenty-seven steers; killed within 7 d after scanning, ultrasonic instrumentation Technicare 210 DX 11; Houghton, 1988. bTotal percentage of steers whose ultrasonic measurement was within the indicated range of the actual corresponding carcass measurement.

indication of total carcass muscle than longissimus muscle area. If this is the case, it may be a mistake to encourage widespread use of ultrasound to measure longissimus muscle area. In agreement with this concept, Dolezal et al. (19891 cautioned that fat thickness and longissimus muscle area are only predictors of total carcass composition and that, in essence, scanning for these traits merely predicts a couple of predictors.

Does Carcass Position Influence Ultrasound Accuracy? For years, ultrasonics have been used to estimate the composition of meat animals; however, several investigators have found discrepancies when they used ultrasound to predict carcass trait measurements in live animals (Giles et al., 1981; Sather et al., 19821. It has been hypothesized that some of these differences are due to changes that tissues undergo during the chilling process (Mersmann, 1982). Lauprecht et al. (19571 investigated the effects of carcass chilling position on subcutaneous fat thickness but found few differences. In contrast, Turlington (1990)concluded that carcass position does influence carcass measurements, thereby influencing the perceived accuracy of ultrasound. In this study, three groups of 25 barrows were scanned 1 d preslaughter. After slaughter, one half of each carcass was either hung on the rail in the traditional manner or was placed in a standing position by use of a specially made rack. Table 5 reports actual backfat and longissimus muscle area measurements for the standing and hanging carcass compared with ultrasonic measurements obtained from the live animal. Downloaded from https://academic.oup.com/jas/article-abstract/70/3/930/4705828 by Washington University in St. Louis user on 03 July 2018

935

Item

Live

Backfat, cm 1st Rib Lest rib Last lumbar 10th Rib Longissimus muscle area, cm2

3.9b 2.Sb 2.8b 3.0b 34Bb

Standing Hanging

3.9b 2.eb 2.7b 2.gb 34.0'

4.3c 2.8' 3.0' 3.1' 35.8d

a ultrf3sonic instnunentation: Technicare 210 DX: Turlington 1990. b~c~dMeam in the same row with different superscripts differ (P .05).

These data indicate no significant differences for backfat measurements between the live animal and standing carcass but do indicate significant differences between the live animal and hanging carcass. In all cases, backfat measurements taken from the hanging carcass exceeded those of the live animal or standing carcass. Significant differences were also found for longissimus muscle area between the live animal, standing carcass, and hanging carcass. In this case, the live animal measurement was intermediate to the standing and hanging carcass measurements. Although similar data do not exist for sheep and beef cattle, it is reasonable to assume that carcass position also influences carcass measurements in these species.

Application of Ultrasound in Swine Finishing Programs Ultrasound might be useful in swine finishing programs to predict market-weight slaughter characteristics andtor) to predict percentage of lean cuts in market weight pigs. The ability to predict market-weight slaughter characteristics was investigated by Robinson et al. (1987). In their study, 50 crossbred market swine were scanned for backfat, longissimus muscle area, and longissimus muscle depth at 14, 17 and 20 wk of age. The objective of this study was to define the relationship between measured traits at different ages and(or1 weights of pigs. These correlation coefficients are reported in Table 6. The moderate to high correlations reported in Table 6 led to the conclusion that early selection for carcass traits is possible in market pigs. Another study to predict market-weight slaughter characteristics of pigs was conducted by McClaren et al. (1989). In this study, 110 barrows

HOUGHTON AND TURLINGTON

936

Table 6. Simple correlation coefficients (r)afor carcass traits from market pigs of different agesb 17 wk

Traitc 14

LED

BF

-

-

.40

.56

-

.66

-

-

-

wk

BF LMA Lh4D 17

BF

wk

BF LMA

Lh4D

20

.53

-

LMA

-

-

-

-

.65

-

-

wk

LMA

.55

-

.62

-

LED

.55

.80

ar > .35 [P < .01). bUltrasonic instrumentation Phillips Medical Systems Model SDR 1200; Robinson et al., 1987. CBF = backfat; Lh4A = longissimus muscle area; LMD = longissimus muscle depth. and gilts were scanned every 2 wk from 42 d of age until slaughter. The pigs were scanned for backfat at the first rib, last rib, last lumbar vertebrae, and 10th rib, and for longissimus muscle area at the loth rib. Estimated lean gain/day was calculated and multiple regression was used to evaluate the predictive power of weight and ultrasound data. These researchers found that early (up to 53 kg BW) ultrasound measures had predictive power similar to preslaughter measures. For example, ultrasound data increased average adjusted R2 values for carcass traits by 22% in the growing phase, 43% in the growing and finishing phase, and 33% preslaughter. Therefore, it was concluded that ultrasound data could prove useful in early selection decisions and for selections made at market weight for carcass merit in pigs. The ability to predict percentage of lean cuts in market-weight pigs was of interest to Teny et al. (19891, who scanned 20 market pigs before and after slaughter. During the scanning process, nine fat measures were taken along the longissimus muscle. After slaughter, one side of each carcass was fabricated into the four lean cuts and prediction equations were tested for percentage of lean cuts using ultrasonic fat measurements. Of the equations tested, a two-variable equation provided the best prediction of percentage of lean cuts (R2 = .83; residual standard deviation [RSDI = 1.67). This equation is as follows: 66.55 - 4.97 (fat on the dorsal midline at the anterior tip of the gluteus medius) + 1.98 aoin eye area at the 10th rib). To verify the usefulness of this equation, it was tested on a separate sample of 20 market pigs (R2= 33; RSD = 2.04). Although some accuracy was lost upon validation, these researchers concluded that prediction equations could be useful to predict percentage of lean cuts in pigs. Furthermore, these predictions might be useful for either marketing or breeding and selection purposes. Downloaded from https://academic.oup.com/jas/article-abstract/70/3/930/4705828 by Washington University in St. Louis user on 03 July 2018

Application of Ultrasound in Lamb Finishing Programs The lamb industry must be concerned with producing lambs that yield as much lean and as little fat 8s possible to compete with other meat sources (Edwards et al., 1989). Smith and Carpenter (1973) and Johnson (1975) found that carcass fatness in lambs had the largest influence on yield of saleable product. Therefore, many researchers have investigated the use of real-time ultrasound to determine body fatness in live lambs (Gooden et al., 1980; Kempster et al., 1982). Many of these reports are contradictory as they relate to the usefulness of ultrasound to predict body composition of sheep. Therefore, Edwards et al. (1989) conducted a study in which several methods were evaluated to predict percentage of closely trimmed retail cuts in lambs. In this study, 30 lambs were evaluated before slaughter by ultrasound, linear measurements (shoulder height, heart girth, body length, and forearm circumference), and visual appraisal. Upon slaughter, carcasses were fabricated into closely trimmed primal and retail cuts and regression equations were tested to predict lamb cutability. Regression equations using ultrasound and linear measures had R2 values ranging from .29 to .41. However, the addition of visually estimated fat thickness raised R2 values from .36 to 57. It was also reported that percentage of kidney, pelvic, and heart fat was more highly related to saleable yield than was external fat cover. Additional results showed that ultrasound and linear measures accounted for more variation in yield of saleable product than did live weight. These results led to the conclusion that percentage of kidney, pelvic, and heart fat is an important predictor of saleable yield; therefore, ultrasound and linear measures were not reliable predictors in themselves because these measurements could not give a direct measure of this

PREDICTION OF CARCASS TRAlTS BY ULTRASOUND

carcass trait. It was also concluded that more accurate ultrasound measures are needed to predict compositional differences in live lambs. Until this occurs, these researchers suggested that a visual estimation of body fat by a trained livestock evaluator is still the best predictor of market lamb cutability. Prediction of carcass chemical composition is of interest in many sheep research programs. Additionally, this information could be of value to lamb consumer groups. To evaluate the usefulness of ultrasound in predicting lamb carcass composition, Turlington (19901 used 160 ram lambs. These lambs were scanned for backfat at the 12th rib and dock, and longissimus muscle area was ultrasonically measured at the 12th rib 2 d before slaughter. After slaughter, one side of each carcass was ground for compositional analyses and regression equations were tested to estimate fat-free soft tissue (lean mass). The equation that included live weight dock fat, last rib fat at the midline and 3/4 position, and longissimus muscle area as measured by ultrasound had a regression coefficient of .89. Although this reflects acceptable accuracy, it was also reported that the equation using live or carcass weight as the sole predictor provided regression coefficients from 3 6 to .91. These results led to the conclusion that weight (live or carcass) was the primary determinant of lean mass in this population of lambs. Reasons for this included the fact that these lambs had a very narrow range of fat depth; therefore, small differences in actual and estimated fat represent a large decrease in accuracy when predicting carcass composition. Although fat and muscle measures were of little value in predicting carcass composition, it was reported that fat measures at the 34 position over the loin eye were better indicators of carcass composition than measurements on the midline.

Application of Ultrasound in Beef Finishing Programs Of primary interest to feedlot managers is the ability to identify and market groups of cattle that will consistently produce carcasses of similar weight with acceptable yield and quality grades. Accurate measurements of subcutaneous fat, muscle, and marbling in the live animal would allow more effective marketing practices. Therefore, many researchers have evaluated the use of realtime ultrasound as a method to predict these carcass traits in live animals (Smith et al., 1988; Dolezal et al., 1989;Stouffer et al., 1989;Duello et al., 1990; Houghton et al., 1990;Perry et al., 19901. Ways in which this technology might be appliDownloaded from https://academic.oup.com/jas/article-abstract/70/3/930/4705828 by Washington University in St. Louis user on 03 July 2018

937

cable in beef cattle finishing programs include the prediction of lean composition at slaughter, the prediction of days on feed to a constant body compositional end point, or the prediction of carcass chemical composition for research and industry purposes. There are several times throughout the feeding period that producers may wish to use ultrasound to predict carcass traits. These include arrival at the feedyard, when cattle are reimplanted, and(or) near the end of the feeding period. Scanning cattle near the end of the feeding period should result in more accurate carcass trait predictions; however, there are disadvantages associated with scanning cattle at this time. These include additional labor, stress, and expense associated with an additional trip through the chute, increased discounts for bruising, and facility limitations as they relate to large, market-weight cattle. Another commonly accepted disadvantage is the possible reduction in gain associated with the movement of marketweight cattle. In contrast to this concept are the data generated by Houghton et al. (1989)presented in Table 7. These data indicate there were no significant differences in ADG between cattle that were worked through a chute 7 to 10 d before slaughter and cattle that were not handled before slaughter. However, cattle that were handled were only moved during cool, early morning hours, and in each case they were returned to their home pens within 2.5 h. Specific applications of ultrasound in beef cattle finishing program include the prediction of lean composition at slaughter. For example, Harada et al. (19851 scanned 64 Japanese black bulls for fat thickness, muscle, and marbling at 11 to 12 mo (Stage 11, 15 to 16 mo (Stage 21, and 21 to 22 mo (Stage 31 of age. Correlation coefficients were calculated to determine the relationship of measurements at Stage 1 to Stages 2 and 3, and are listed in Table 8. With the exception of marbling, it was concluded that ultrasonic estimates and body measurements at 11 to 12 mo of age could predict carcass traits at 15 to 16 and 21 to 22 mo of age. These researchers attributed their lack of success in predicting marbling to the fact that it is a very mobile body component that is affected by environment, and therefore its prediction early in the animal's life or feeding period is difficult. In contrast to this conclusion is a study conducted by Brethour (ISQO),in which 14 sets of cattle In = 6191 were evaluated by ultrasound speckle score at various time intervals before slaughter (up to 148 d preslaughterl. Various breeds, ages, and sexes of cattle were represented in this study. Upon slaughter, marbling scores were obtained,

938

HOUGHTON

AND

Table 7. Performance comparison of fed steers by method of handlingb Group Trial 1 Handled Not handled Trial 2 Handled Not handled Trial 3 Handled Not handled ~

No.

ADG, kgc

486

1.38

24 1

1.40

52 1 263

1.26

468

1.30

258

1.32

1.25

~~

aHandled = cattle worked through chute 7 to 10 d before slaughter; not handled = cattle were not worked through chute. bHoughton et al., 1989. CAveragedaily gain for the entire feeding period.

and those carcasses with at least Smalloomarbling were classified as USDA Choice. Results from this study indicated that speckle scores were highly repeatable (r = .go); however, the author admitted that scores might have been subconsciously affected by animal breed, condition, and conformation. Speckle score was at times moderately correlated with marbling score (r = .22 to .77), but accuracy was highly variable and showed a disturbing presence of outliers. Nevertheless, it was reported that speckle score accounted for 25 to 30% of marbling score variance. In addition, grade classification accuracy ranged from 53 to 94%; however, borderline marbling scores (Slight75 to Smal1251 were difficult to differentiate. Speckle score recorded more than 30 d before slaughter resulted in grade classification accuracy from 57 to 72%. In conclusion, speckle score showed some relationship to carcass marbling score; however, the author suggested that computer interfacing would be necessary to improve speed, accuracy, and precision of this procedure before it is taken to the field. If this is accomplished, ultrasound could prove an effective method of determining marbling in live cattle and could help cluster cattle into groups for more effective marketing. Other attempts to sort cattle into uniform

TURLINGTON

groups for marketing purposes have been made by Houghton (1988). In this study, 15- to 18-mo-old steers G.1 = 706) of similar genetic and management background were sorted into the feedlot by hip height and an ultrasonic measurement of backfat. Cattle were divided into small and large frame groups within three condition groups (light, average, and heavy). Cattle were slaughtered as a pen when a random 15% sample of that pen averaged 1.0 cm backfat or weighed 591 kg. When managed in this manner, days on feed ranged from 83 to 104 d, for a 21-d difference. Average backfat at slaughter, by pen, ranged from .92 to l.lf -27 cm. These data indicate that ultrasonic measurements were useful in marketing cattle at a constant body compositional end point of 1.0 cm of backfat. Furthermore, the SD reported here would suggest the cattle were sorted into relatively uniform body compositional groups as determined by backfat at the 12th rib. In addition, yield grade and marbling score means were very consistent, ranging from 2.7 to 3.1 and Smallo7 to Small32, respectively. These preliminary data suggested that sorting feeder cattle by frame and backfat when they enter the feedyard could result in more appropriate days on feed for cattle of different body types while achieving consistent and acceptable yield and quality grades across body type. To further evaluate this concept, Houghton et al. (1990) used 15- to 18-mO-Oldsteers In = 997) of various genetic backgrounds from two origins. These cattle were sorted into the feedlot by visual appraisal In = 448) or by an ultrasonic measurement of backfat and measured hip height In = 449). Cattle were divided into small and large frame groups within two condition groups (light and heavy). Sorted cattle were compared to unsorted controls In = 1001. As in the previous study, cattle were slaughtered as a pen when a random 15% sample of that pen averaged 1.0 cm backfat or weighted 591 kg. Results from this study indicated that ultrasonic measurements of initial backfat were more highly related to carcass backfat (r = 3 9 ) than visual estimations of initial backfat to carcass backfat (r = .16 to .33). To test

Table 8. Simple correlation coefficients (r)" among ages of beef bullsb Stage I1 (15 to 16 mol Traitc

BF

Stage I (11 to 12 mol BF LMA Marbling

.63

LMA

-

-

.86

-

-

Stage I11 (21 to 22 mol

Marbling

BF

-

.61

.78

-

LMA

Marbling

.77

-

5 c .35 (P c .OS). bUltrasonic instrumentation B-Type Ultrasonic Scanningscope; Harada et al.,1985. CBF = backfat; LMA = longissimus muscle area.

Downloaded from https://academic.oup.com/jas/article-abstract/70/3/930/4705828 by Washington University in St. Louis user on 03 July 2018

.24

PREDICTION OF CARCASS TRAITS BY ULTRASOUND

939

Table 9. Performance factors and carcass traits by method of sorting Sorting method Item

Initial backfat, cm ADG,kg Carcass backfat, cm Yield grade

Control .425 f .082 f .41

1.77 1.10

2.5

f .35 f .6

Visual

*

.42s ,088 1.73 f $32 1.00 f .25 2.4 f .8

Ultrasoundb .425 f .080 1.73 1.10

2.6

f .32 f .30 f .6

aHoughton et al., 1990. bUltrasonic instrumentation Technicare 210 DXII.

the uniformity of performance factors and carcass traits, variances of the means were statistically tested. Table 9 lists these factors by method of sorting (unsorted, visual, and ultrasound). Although visually sorted and ultrasound-sorted cattle tended to show more uniformity for certain traits, no significant differences were found when carcass uniformity and variation in ADG were tested between treatments. This might be partially explained by the fact that very little variation in body composition existed in these cattle when they were sorted into the feedlot (average initial backfat = .425f .083 cm across all treatments). It is likely that this initial uniformity made it difficult to see significant differences in feedlot performance or carcass trait uniformity. In conclusion, there was no economic benefit associated with the sorting procedure used in this study. In addition, visual sorting equaled or exceeded sorting by ultrasound and measured hip height. This might be because when cattle are sorted by ultrasound, a single measurement is used as the basis for sorting. In contrast, a visual appraiser has the opportunity to evaluate the entire animal. Therefore, an experienced visual appraiser may have an advantage when sorting cattle for uniformity. Ultrasound may be helpful, however, as a training tool for inexperienced visual evaluators. Strasia et al. (1989)also conducted a study in which feedlot cattle were sorted for marketing purposes and reported no associated economic benefits. In this study, 123 heifers were sorted for slaughter by weight and ultrasound fat thickness. Cattle were slaughtered when they reached 1.25 cm of backfat and 455 kg. This resulted in cattle being slaughtered at either 110 or 145 d on feed. Upon slaughter, carcass measurements were collected and correlation coefficients were calculated. It was found that ultrasound measurements of fat thickness were imprecise (37% within i .25 cm) in this study. Therefore, data were evaluated according to days on feed. When analyzed in this manner, it was found that the ADG of cattle fed for 110 d equaled 1.3 kg, whereas the ADG of cattle fed for 145 d represented a 14% decrease at 1.1 kg. Downloaded from https://academic.oup.com/jas/article-abstract/70/3/930/4705828 by Washington University in St. Louis user on 03 July 2018

This led to the conclusion that when weight and fat thickness were used to determine slaughter end point, it seemed that the higher performing, more efficient cattle were slaughtered first. This resulted in lower performing cattle being left on feed for additional time. An additional conclusion of this study was that ultrasonic measurements of fat thickness were not highly correlated to carcass quality grade, whereas additional time on feed was related to higher marbling scores and carcass quality grades. Additional information from steers that were slaughtered at a predetermined body compositional end point comes from Houghton and Simms (unpublished data). This information was gathered from steers In = 831) that were consigned to live animal and carcass trait evaluation programs in western Kansas. These steers entered the feedlot as weaned calves, represented a wide variation in breed type (19breeds) and origin (242consignors), and were slaughtered on an individual basis when they reached 1.0 cm backfat or weighed 591 kg. The following data represent steers from 3 y at two locations. Of these cattle, 565 (68%) were selected for slaughter according to the backfat criteria (1.0 cm). Of these 565 steers, 475 (84%) had carcass backfat measurements between .87 to 1.13 cm, whereas 548 (97%) had carcass backfat measurements from between .75 to 1.25 cm. In addition, slaughter weight, backfat, and frame score were found to be related to carcass loin eye area (r = .70)and are represented in this prediction equation: longissimus muscle area = 6.04 + slaughter weight (.0076) - backfat at the 12th rib (4.71)+ frame score C009). This equation accounted for 49% of the variation in carcass longissimus muscle area. When tested on a separate set of cattle (n = 2221,this equation accounted for 47.2% of the variation in carcass longissimus muscle area. Further information included the fact that age, initial weight, initial backfat, and frame score were related to days on feed (r = .72).Within biological type, age and weight had the most influence on days on feed (R2 = .50 to .66). This information has led to a current study in which cattle are being sorted by individual projec-

940

HOUGHTON AND TURLINGTON

tions of days on feed. Upon entering the feedyard, cattle are evaluated for age, breed type, sex, initial backfat, weight, and hip height. Backfat, weight, and hip height are also evaluated when cattle are reimplanted (when this management procedure is appropriate). Equations to predict days on feed are being developed from a combination of these factors. Subsequent sets of cattle will be used to validate these prediction equations. This sorting procedure includes the use of a chute-side computer that immediately calculates the predicted days on feed of individual steers. After this is calculated, cattle are placed into a “days on feed category” (pen) and fed accordingly. Validation of this sorting procedure is currently underway; therefore, final conclusions will be forthcoming. Interest has also been shown in the ability of real-time ultrasound to predict beef carcass chemical composition. Faulkner et al. (1989) investigated this concept by ultrasonically measuring 47 cows pre- and postslaughter. Upon slaughter, the left half of each carcass was ground and sampled for chemical analyses. Prediction equations were developed using weight, ultrasound fat, hip height, and their squared values in a stepwise regression procedure. A separate set of 36 cows was used to test the prediction equations. Equations using live variables to predict carcass composition were similar (R2= .42 to .QO) to those developed from carcass measurements EL2 = .43 to .92). Live animal variables contributing (P .05)to prediction equations were live weight, 12th rib ultrasound fat, and hip height. Quadratic components were used in this equation. Prediction equations using live measures were validated (R2 = .53 to .72), as were equations using carcass measures Et2 = .35to .84). These data led to the conclusions that ultrasound can effectively measure fat thickness, and that this measurement can be combined with other live measurements to estimate percentage of fat, weight of carcass fat and lean, and percentage of carcass bone.

useful on a commercial basis in the lamb industry. Ultrasound is currently being used on a commercial basis in the beef cattle industry to measure fat depth and 12th rib longissimus muscle area. Data presented in this review suggest that 12th rib fat measurements are accurate and could be used to enable slaughter of cattle at a predetermined body compositional end point. The accuracy of longissimus muscle area measurements, however, is inconsistent among technicians and(or) ultrasonic instrumentation. These data should cause producers to approach ultrasonic measurements of loin eye area with caution, particularly when they are presented on individual animals. Data presented in this review also suggest that the potential exists to evaluate marbling with the use of ultrasound, but the procedure is not yet ready for commercial use.

Literature Cited Brethour, J. R. 1990. Relationship of ultrasound speckle to marbling score in cattle. J. Anim. Sci. 88:2603. Campbell, D., H. H. Stonaker, and A. L. Esplin. 1959. The use of ultrasonics to estimate the size of the longissimus dorsi muscle in sheep. J. Anim. Sci. 18:1483 (Abstr.). Claus, A. 1957. The measurement of natural interfaces in the pig’s body with ul-ound. Fleischwirtschaft 9552. Dole24 H. G., M. T. Smith, and B. D. Behrens. 1989. The role of ultrasound for predicting carcass merit in livestock. Roc. Oklahoma Beef Cong. Duello, D. A. G. H. Rouse, and D. E. Wilson. 1990. Real-time ultrasound as a method to measure ribeye area, subcutaneous fat cover and marbling in beef cattle. J. Anim. Sci 68(SUppl. 1):240 (Ab&.). Edwards, J. W., R C. Cannell, R. P. Garrett, J. W. Swell, H. R Cross, and M. T. Longnecker. 1989. Using ultrasound, linear measurements and live fat thickness estimates to determine the carcass composition of market lambs. J. Anim.

sci. 67:3322. Faulkner, D. B., D. F. Parrett, F. K. McKeith, and L. L. Berger. 1989. Prediction of fat cover and carcass composition from live and carcaas measurements. pp 55-62. Beef Cattle Research Rep. Univ. of Illinois, Urbana. Forrest,J. C., C. H. Kuei, M. W. Orcutt, A. P. Schinckel, J. R. Stouffer, and M. D. Judge. 1989. A review of potential new methods of on-line pork carcass evaluation. J. Anim. Sci. 67:2164.

Implications Ultrasound is of potential use in educational and research efforts for swine, sheep, and beef cattle. From an industry standpoint, however, there seem to be uncertainties about the usefulness of ultrasound. Swine data suggest that ultrasound is useful under field conditions. In contrast, conflicting data exist for sheep. Undoubtedly, the small variation that exists in fat depth and muscle area in lambs is a contributing factor to the lack of consistent data as it relates to the usefulness of ultrasound. Therefore, more precise equipment is required before ultrasound w i l l be Downloaded from https://academic.oup.com/jas/article-abstract/70/3/930/4705828 by Washington University in St. Louis user on 03 July 2018

Giles, L. R., R. D. Murison, and B. R. Wilson. 1981. Backfat studies in growhg pigs. 2. A comparison of ultrasound and ruler probe predictors of backfat and eye-muscle measurements in the live pig. Anim. h o d . 32~47. Gillis, W. A., G. H. Bowman, H. Grieger, and G. W. Rahnefeld. 1972. A comparison of ultraaonics with the d e r probe for the prediction of carcass yield in swine. Can. J. Anim. Sci. 52:637. Gooden, J. M., A. D. Beach, and R.W. Purchas. 1980. Measurement of subcutaneous backfat depth in live lambs with an ultrasonic probe. N.Z. J. Agric. Res. 23:161. Hale, D. 1989. Beef Improvement Federation ultrasound proficiency program. Summary Rep. Harada, H., and K. Kumazaki. 1980. Relationship between f a t thickness, cross sectional area of M. Longissimus thoracis and marbling score and their ultrasonic estimates. Jpn. J. Zootech. Sci. 51(41:261.

PREDICTION OF CARCASS TRAITS BY ULTRASOUND Harada, H., K Moriya, and R.Fukuhara. 1985.Early prediction of carcass traits of beef bulk. Jpn. J. Zootech. Sci. 58(3):250. Hazel, L. N. and E. A. Kline. 1959.Ultrasonic measurements 01 fatness in swine. J. Anim. Sci. 18:815. Henderson-Perry, S. 1990. Validation of ultrasound measurements. M.S. Thesis. Kansas State Univ., Manhattan Herring, D. S.,and G.Bjornton. 1985. Physics, facts, and artjfof d i a @ ~ t i cultrasound. In:The Veterinary clinics of North America-Small Animal Practice. pp 1107-1122. W. B. Saunders Co.,Philadelphia, PA Houghton, P. L. 1988.Application of ultrasound in commercial feedlots and beef breeding programs. pp 89-99. Beef Improvement Federation Proc., Albuquerque, NM. Houghton, P. L., D. D. Simms, and J. J. Higgins. 1990.Comparison of steer feedlot performance and carcass trait uniformity by method of sorting.pp 75-77. KSU Cattlemen's Day Rep. of h o g . Houghton, P. L., D. D. Simms, and D. E. Wilson. 1989. How practical is ultrasound? pp 57-67. Proc. Range Beef Cow Symposium, Rapid City, SD. Johnson, H. K 1975. What's the opportunity with lamb? Lamb Marketing Seminars: An Update on Marketing Systems. p 85. Am. Sheep Producers Council, Denver, CO. Kempster,A. J.,D. Armll,J. C. Alliston, and J. D. Barker. 1982. A n evaluation of two ultrasonic machines (Scanogram and Danscanner) for predicting the body composition of live sheep. Anim. Prod. 34:248. Lauprecht, E., J. Sheper, and M. Schroder. 1957.Measuring the backfat thickness of live pigs with ech@ranging techniques. Milt. Dtsch. Landw. G ~ s72:881. . Lopes, D. M., S. A. Williamson.J. A. Jacobs, and M. W. Thomas. 1987.Estimation of fat depth and longissimus muscle area in swine by the use of real-time ultrasonography. Proc. Western Sect. ASAS. 38~155. McLaren, D. G., F. M. McKeith, and J. Novakofski. 1989.Prediction of carcass characteristics at market weight from serial real-time ultrasound measures of backfat and loin eye area in the growing pig. J. Anim. Sci. 87:1657. Mersmann, H. J. 19892. Ultrasonic determination of backfat depth and loin area in swine. J. Anim. Sci. 54:288. Moody, W. G., S. E. Zobrisky, C. V. Ross,and H.D. Naumann. 1965.Ultrasonic estimates of fat thickness and longissimw dorsi area in lambs. J. Anim. Sci. 24:384. Panier, C. 1957.Ultrasonics in the determination of the state of fatness of hogs. Rev. Agric. 10:482. Perry, T. C., S. J. Ainslie, M. J. Traxler, D. G. Fox, and J. R. Stouffer. 1990. Use of real-time and attenuation ultrasonic measurements to determine backfat thickness, ribeye area, carcass marbling and yield grade in live cattle. J. Anim. Sci. 88(Suppl. 1):337 [Abstr.). Perry, T.C., J. R.Stouffer, and D. G.Fox. 1989.Use of real time and attenuation ultrasound measurements to measure fat deposition, rib eye area and carcass marbling. J. Anim. Sci. 67fSUppl. 1): 120 (Abstr.). Price, J. F., H.B. Host, A. M. Pearson, and C. W. Hall. 1958. Some observations on the use of ultrasonic measurements for determining fatness and leanness in live animals. J. A n h . Sci. 171156 LAbstr.1.

Downloaded from https://academic.oup.com/jas/article-abstract/70/3/930/4705828 by Washington University in St. Louis user on 03 July 2018

941

h t a n e n , N. W., and R. L. Ewing. 1981.Principles of ultrasound application in asimals. vet. Radiol. 22:196. &binson, T. F., L. E. Orme, and R. L. Park. 1987. Growth characteristics of immature swine as determined by realtime linear array ultrasound. Proc. Western Sect. ASAS. 38:151.

Rouse, G., and F. C. Parrish. 1987. The use of real-time ultrasound to measure marbling in live beef animals and beef carcasses. pp 147-149. A. S. Leaflet R457. Sather, A. P., H. T. Fredeen, and A. H. Martin. 1982.Live animal evaluation of two ultrasonic probes as estimators of sub cutaneous backfat and carcass composition in pigs. Can J. Anim. Sci. 62:943. Smith, G. C., and Z.L. Carpenter. 1973. Estimations of lamb carcass cutability within narrow ranges of weight and fat thickness. J. Anim. Sci. 36:432. Smith, M.T.,J. W. Oltjen, H. G. Dolezal, D. R. Gill, and B. D. Behrens. 1988. Evaluation of ultrasound for prediction of carcass fat thickness and rib eye area in feedlot steers. Oklahoma Agric. Exp. Sta., Anim. Sci. Res. Fbp. MP-127: 291.

Smith, M. T.,J. W. Oltjen. H. G. Dolezal, D. R. Gill,and B. D. Behrens. 1990. Evaluation of real-time ultrasound for predicting carcass traits of feedlot steers. Oklahoma Agric. Exp. Sta., Anim. Sci. Res. Rep. MP-129:374. Stouffer, J. R. 1988.ultrasonic evaluation of beef cattle. Ad Hoc Ultrasonic Guidelines Committee. Study Guide. Cornell Univ., Ithaca, NY. Stouffer, J. R., and H. R. Cross. 1985.Evaluation of beef cattle with red-time W a r y array ultrasound. J. Anim. Sci. 61: (Suppl. 1):144 [Abstr.). Stouffer, J. R., T. C. Perry, and D. G. Fox. 1989.New Techniques for real time ultrasonic evaluations of beef cattle. J. Anim. Sci. 67(Suppl. 1):121 LAbstr.1. Stouffer, J. R., M. V. Wallentine, G. H. Wellington, and A. Diekmann. 1961. Development and application of dtr& sonic methods for measuring fat thickness and ribeye area in cattle and hogs. J. Anim. S C 20:759. ~ Strasia, C. A., H. G. Dolezal, M. T. Smith, C. P. Foutz, D. R. Gill, B. D. Behrens, R.M. Lloyd, B. J. Skaggs, R. L. Schemm, C. L Schultz, and D. L. Deen. 1989.Correlation of ultrasound and measured fat thickness in feedlot heifers. pp 151-153. Anim. Sci. Ras. Rep., Oklahoma State Univ., Stillwater. Terry, C. A., J. W. Save% H. A. Fbcio, and H. R. Cross. 1989. Using ultrasound technology to predict pork carcam composition. J. him. Sci. 87:1279. Thompson, J. M.,W. A. Pattie, and R. M. Buttefield. 1977. An evduation of the "Scanogram" as a n ultrasonic aid in aasessing carcass composition of live sheep. Aust. J. Exp. Agric. him. H u b . 17:251. Turlington, L. M. 1990. Live animal evaluation of swine and sheep using ultrasonics. M.S. Thesis. Kansas State Univ., Manhattan. Turner, J. W. 1988. Utilization of data obtained by ultrasound. Ultrasound Training School, College Station, TX. Wild, J. J. 1950. The use of ultrasonic pulses for the measurement of biological tissues and the detection of tissue density changes. Surgery 27:183.