0 . I . I . .. . . I . 4 I . . OI -. I s n p . I . .. O . . I I . c o o c I o u n u u c Q g .. I I . u . I . ~ . ... . . . n . . I . . . . . 1 - ‘ I I I . . . .. . w I . . . : . a . . . . . . . I OI . I. . . . . . . . r . . .. . .. O I , . v . y. . . o ..40 I .avnlvl . , . . .. ..I. 'n ‘— o .o I. ... y . . . . . . .I . . . . . '14» . . ..v I I n a u y o . . ... I I .. . . I u . , .. I . .... . . . I I V V‘ I O Q ~ I .I. . n c u . II . vI _ p c o. I o . . . A u I . . a n . . I . . ‘. . . . I I . . I; .. I n I o o ...v..1..d a . . . . . .. .I'q .0 1 u . I . . . O n u o a ll . I O I t‘ . .. u. . I . . I n _ . o . - ~ 0 . . . .. . _ . . . - o . . - - . . . I .. . . n .. I. .. ... 1| AI-I v u C _ h I . . I. . . . . . n . a. . . .. .. . r I I. I ...-la. I4 o p . . I . s . . .. I ... .IJ...I¢.MJ - fi’ .. .o .B. I . n . I ’.I. I IDJOI' I o I u. u by a o. I. . . . t.. 4 F . . O - I4. l..l . IA . . .- .. . . . . . .- . ...? I . III a .s . . .5 . .0 .0 .Ea fl; . . v » : I . ,_ I u' 9 ... .a .. .4 d. . 0 .- . . .. I .. ~ 4 u I . m. III ~ ¢ . _ . f. u r u A . I . .... u ‘i a a s ... < u . n . . .- ~ .0. a L .. . -- I. .i . D . . . . . _ 0 A. .. I T. . ‘ I u . E . {D A .. l . I . .5 ‘ .v a . I u I I I . . . r. . o . A . . o o . u 1 . _ . . . o . . . o to . I I I I o . I I I A I II 1‘ I l-l.v . o _- . . . J . . 1.. . .0 c . T .. . . . . v or-.. . . I ' . . I . , . . . . ) x -.. . .. . . . a .A 11-. . I .5 . . . . . . . . . - - . .‘l' . . .-. . ‘ . o .I ... I 1.. I I O I . I u I n I. a u l . n s, I . I . d . I.a It .I .. ll . I \ I I . I‘I . T o m a . . . . . . . . .I .. . . A .-. v .u n r .o . . . . 4. . . . . . on . . .. ‘ . A .4. u..r . ...| .. r . . n . .- I. . . o o n n . . I u A: .0 . .711. l I III I - ‘(ll . . I . .a . a. . . :. ¢. I) ...! ‘ .n B I . . I. . .. . .. . v . ‘ I . . I . I . . I ' I '1 ...! u . . . . . . . . o u .. ' n o . t . Q . .l n u I. . . o I I I I -' . u _, . n. ._ . .« t. . I . O I. ' .I ‘ I . .. ... AI... ..oo.... ....4... . ‘ \v. . .I . . ... .o . ....Io-c..|.9I..‘.O Ia;uoI....I.....IOPJabI..I. ......IIOO..sI.I..On.oOc...Jo.3.o.ll24.oooooo..o ...I.....Jo.4.n.\ a.4.o....l\0‘r.o.nfi.rhtn|.HIII.‘-.§VI....0.\.' I. o. I . . . .- -......... ....Z..Tncr.........,....- ...A.n..a.;....- 1....“ 31... ....J.-..-..... 3:32:11. {ff/...)» 3.52...- ..Arrs: . .u.. .. _ .... .. . - J.-- fl o . I I. . L0. 4. a . 3 ‘ll \H}; Ii} ~' II ’il’lll'l WEE? —4-..‘- “-u .... LIB RA R Y Miching State University I he ABSTRACT \ ACCURACY AND REPEATABlLlTY OF SEVERAL SUBJECTIVE LIVE ANIMAL ESTlMATES 0F BEEF CATTLE AND SWiNE CARCASS DESIRABlLITY By Patrick D. Vitlo This study was made to estimate the accuracy and repeatabilities of judges in estimating carcass traits. The judges estimated loin- eye area to the nearest O.l inch and percent preferred cuts ( percent trimmed round and loin ) to the nearest O.l%. Bulls were evaluated twice by each judge, the evaluations being four days apart. All of the judges worked independentiy. There were i6 bulis in l963 and and 27 in l964. The only prior imformation known to the judges were the bulls live weights and loin-eye area and percent preferred aver- ages from bulls slaughtered the previous years. Analysis of variance ‘showed judges did not differ significantly in their eifiretes of loinm eye area or percent preferred c;ts for either year, exeept for percent preferred in 1963. Simpie correiatioas between each judgeb estimates and actual values of loinweye area for individual bulls were highly significant each year. The simple correlations for the judges between actual and estimated percent preferred cuts were quite variable. in l96e all but one judge showed a negative, non=significant correlation, one judge showing a positive highly significant correlation. In l963 all judges showed nonmsignificant correlations, some positive and some negative. The repeatabiiity of judges in their estimates of loinueye Patrick D. Vitlo area w s highly significant {0.70 m 0.85) eacn year. The repeatability l-l‘ .- of judges in .stimating percent preferred cuts was lower. Two jrdge. in f In if) l963 and three in l964 tad highly significant repeatability estimate ranging from 0.6l t3 0.85. This study Indicated that judges were more accurate in estimating laihmeye area than percent preferred cuts. The A similar study was carried oat with swine. Here four judges estimated backfat thicku \ LI‘ uss, lshgthg and perceht ham to the rearest sl) 0.l unit. There mere two repl.eatighs on the live timate on the carcass. All judges wsrked iid:p61defitly and khea 33th“ ing abau (1‘ (:1. 1... «I: U. :13 (t T LL) 0 ! “ad of the sw;he involved. Ahalysis of variance at a: in backfat thickhess estimates to be due to the ihd;vfdxal pigs, the judges, and the interaction be- tween the two. The same was trte for estimates of perceht ham, whereas variation in estimates of length could be accounted for by the pigs and the jsdgas but there were no interaction effects. This indicates that m the judge. essehtially rahked the pigs similarly. fl Judges did pearest ;h estimating backfat thickness in the live animal but were more consistent in these estimates as they moved from the live animal to the carcass. Esckfat thickness was more accurately evaluated in the carcass, however the reverse was true in estimat percent ham. Estimates of length were equally accurate in the carcass —l and live animal. On the Eve animal, judges were most accurate in pre- dicting length and were least accurate in predicting backfat thickness. h .‘g d.) ri- (r U‘ 5: '1 Ln 0 \_ A. y—r J) Repeatability of live estima‘i' gh for all judges, save judge D, and ranged fram 0.10 ta 0.96 with a mean value of 0.62. Patrick D. Vitlo Rank correlation was calculated between live placings and carcass placings. It was found that judges were better able to place light" weight swine than heavyuweight swine but in neither case were the cor» relations significant. in fact the correlations («.02 and 0.M8) indim cated that judges were not able to rank the live pigs on carcass merit. Correlations between carcas index ard the factors constituting the index were generally found to be moderately righ (0.80) with meat quality playing essentially no role. Percent tam was by far the most important single factor deterrining carcass merit, followed by loin- ’3 r) eye area. Standard partial regress o. .oefficients of index on index factors suggested that each factor did not contribute its allotted share of the variation to the overall index. 3 “+1 :1 ‘3 “J (T) '.. U} 0 ’T‘ r 2' ‘1- ’D ..- O :) ll Simple correlation coefficients betwee eye and loinueye area, baekfat thickness, and length ind loinneye area was the most highly reiated of the three to the firm” ness, the correlation bein~ m.h4. r“ ..7 ACCURACY AND REPEATABILITY OF SEVERAL SUBJECTIVE LIVE ANIMAL ESTIMATES OF BEEF CATTLE AND SWINE CARCASS DESIRABILITY By Patrick D. Vitlo A THESIS Submitted to Michigan State University in partial fquIIlment of the requirements for the degree of MASTER OF SCIENCE Department of Animal Husbandry I965 ACKNOWLEDGEMENTS The author wishes to eXpress his appreciation to Dr. W. T. Magee, Professor of Animal Breeding, for his suggestions, assist- ance, and guidance involving the analyses and preparation of this thesis. The writer is also indebted to the various members of the faculty and graduate staffs of the Animal Husbandry and Food Science Departments, without whose participation and cooperation in the re- search, this manuscript would not have been possible. Deepest thanks are also extended to my wife, Fran, for her continual thoughtfulness and confidence throughout the authors col- lege career. TABLE OF CONTENTS Page Introduction . . . . . . . . . . . . . . . 2 Literature Review . . . . . . . . . . . . . 5 Data . . . . . . . . . . . . . . . . . . 14 Methods of Analysis . . . . . . . . . . . . . 18 Results and Discussion . . . . . . . . . . . . 22 Conclusion . . . . . . . . . . . . . . . . 52 Bibliography . . . . . . . . . . . . . . . 55 Appendix D O O O O O O O O O O O O O O O 58 Table ll 12 l3 l4 LIST OF TABLES Means and Standard Deviations for Actual and Estimated Values of Loin-eye Area and Percent Preferred Cuts for BUIIS O O O O O O C O O O O O O O O . C Simple Correlation Coefficients Between Actual and Estimated Carcass Traits. . . . . . . . . . . Mean Correlation Coefficients of Live Bulls Estimates with Actual Carcass Values . . . . . . . . . . Simple Correlations Between Estimated and Actual Loin- eye Area and Standard Partial Regression Coefficients Between Estimated and Actual Loin-eye Area, Carcass Wt. held Constant . . . . . . . . . . . . . Simple Correlations Between Estimated and Actual Per- cent Preferred and Standard Partial Regression Coef- ficients Between Estimated and Actual Percent Prefer- red, Carcass Wt. held Constant. . . . . . . . . Repeatability Estimates Between Judges Replications for 1963 and 196# O O O O O O O O O O O O 0 Analysis of Variance for Loin-eye Area (l963). . . . Analysis of Variance for Loin-eye Area (l964). . . . Analysis of Variance for Percent Preferred Cuts (l963). Analysis of Variance for Percent Preferred Cuts (I964). Means and Standard Deviations for Actual, Live Esti- mates and Carcass Estimates of Backfat, Length and Percent Ham for Four Judges. . . . . . . . . . Simple Correlation Coefficients Between Live Estimates and Carcass Values in Swine by Judge and Replication . Repeatability of Live Estimates in Swine for Four Judges Analysis of Variance for 196A Farmers Week Live Esti- mates of Backfat Thickness . . . . . . . . . . Page 23 23 25 25 26 26 29 29 3] 31 33 34 35 35 IS 20 2l 22 23 24 25 26 27 28 29 30 3] Analysis of Variance for l964 Farmers Week Live Estiu mates of Percent Ham . . . . . . . . . . . . Analysis of Variance for I964 Farmers Week Live Esti- mates of Length. . . . . . . . . . . . . . Simple Correlations Between Estimated Live Values and Estimated Carcass Values. . . . . . . . . . . Simple Correlation Coefficients Between Carcass Traits and Between Estimates of Carcass Traits. . . . . . Simple Correlation Coefficients Between Carcass Esti- mates and Actual Carcass Values in Swine by Judges . . Simple Correlation Coefficients Between Carcass Plac- ing and Index Factors and Between Index and Index Fac- tors, l965 Farmers Week Swine . . . . . . . . . Partial and Multiple Correlation Coefficients for l965 Farmers Week Heavy-Weight Swine . . . . . . . . Partial and Multiple Correlation Coefficients for I965 Farmers Week Light-Weight Swine . . . . . . . . Partial and Multiple Correlation Coefficients for I965 Farmers Week Light and Heavy-Weight Swine Combined . . Standard Partial Regression Coefficients of Index on Actual Carcass Values. I965 Farmers Week Swine . . . Standard Partial Regression Coefficients of Index on Index Point Factors. I965 Farmers Week Swine. . . . Standard Partial Regression Coefficients of Carcass Placing on Carcass Values. I965 Farmers Week Swine. . Standard Partial Regression Coefficients of Carcass Placing on Index Point Factors. I965 Farmers Week Swine Standard Partial Regression Coefficients of Index on Actual Carcass Values. I965 Spring Barrow Show . . . Standard Partial Regression Coefficients of Index on Index Point Factors. I965 Spring Barrow Show Swine . . Simple Correlations Between Loin-eye Firmness and Backfat Thickness, Length, and Loin-eye Area . . . . Relationship Between Firmness of Loin-eye and Back- fat thickness, Length, and Loin-eye Area . . . . . 36 36 38 38 39 hi #3 Ah 45 47 li7 A8 A8 49 49 49 5l Figure LIST OF FIGURES Page Path Coefficient Diagram Showing the Relationships Between Carcass Wt., Actual Loin-eye Area and Est. Loin-eye Area for Judge A (I964) . . . . . . . . 26 Relationships Between Firmness of the Loin-eye and Backfat Thickness, Length, and Loin-eye Area . . . . - 5I Table 2A 28 LIST OF APPENDIX TABLES Judges Estimates of Loin—eye Area and Percent Preferred Cuts, Two Replications, I963 . . . . . . . . . Judges Live Estimates of L.E.A. and Percent Preferred (1961+) . JUdges I and 2 O O O O O O I O O O O Judges Live Estimates of L.E.A. and Percent Preferred (I964). Judges 3 and 4 . . . . . . . . . . . Actual L.E.A. and Percent Preferred Cuts (I963 and I964) I964 Farmers Week Live Estimates of Backfat, Length and (j/DHam O O O O O O O O O O I O O O O I O Judges Estimates of Carcass Traits for I964 Farmers Week I964 Farmers Week Pigs, Carcass Values . . . . . . Live Placing, Carcass Placing, Actual Values and Point Values, I965 Farmers Week Swine . . . . . . . . Michigan Quality Pork Score Card . . . . . . . . Carcass Placing, Actual Carcass Values and Point Values for the I965 Spring Barrow Show Entries Slaughtered at Michigan State University . . . . . . . . . . Actual Values for Backfat, Length, Loin-eye Area and Firmness. I965 Spring Barrow Show Pigs Slaughtered at Farmer FEELS. o o o o o o o o o o o o o 0 vii 59 6O 62 63 64 65 66 67 68 INTRODUCTION For many years the livestock producer has been selecting his breeding stock on a system that requires a subjective live animal score. Although this score is relatively easily obtained, once the stockman has had ample eXperience in the field of evaluation, there is still some doubt as to the degree of the correlation of this score with ac- tual carcass values on an individual animal basis. In the past it has seemed that at best these predictions have fallen short of having a high correlation with the carcass values. There is much research in this field concerning the-use of live animal measures in predicting carcass value, but these measurements too have had highly questionable accuracy. Measures which have been used are length and width of round, loin,rib, and/or rump, creatine determinations, radioisotOpe readings and length of sound waves. How- ever it is interesting to note that none of these measures thus far has shown to be any more reliable than has visual appraisal. For the producer of superior livestock to improve his animals by selection it is necessary that he base his selection on those traits that seem to be at least moderately heritable (0.30 or more). Only through the selection for these traits can the stockman make major improvements in the genetic composition of his livestock. Arthaud _£fl§l, (I964) reported several heritability estimates for beef carcass traits as follows; plumpness of round and fullness of loin, 0.42; loinueye area, 0.52; and live grade, 0.43. There have also been many heritability estimates of carcass traits in swine. Carroll _t__i. (I962) have presented average estimates of several workers for heritu abilities of carcass traits in swine. These estimates include the fol- lowing: carcass length 0.59, loin-eye area 0.48, backfat thickness 0.49, and percent ham 0.58. The past research indicates that most of the carcass character- istics on which breeders have based their swine selection seem to be quite highly heritable. The primary objective of this researc is to estimate the accu- racy and repeatability with which carcass traits can be evaluated in the live animal. In conjunction with this, a study was initiated to evaluate the accuracy of predicting percent ham, average backfat thickness, and length from swine carcass evaluation. Since the overall value of the carcass is the most important factor in determining an animals merit, a study was carried out to determine the correlation of live placing with carcass placing. An overall carcass score or index value, based on several factors, was determined for each swine carcass. The factors included were carcass length, average backfat thickne"s, percent ham, loin-eye area at the tenth rib and quality. The quality grade consisted of a composite of three factors, color, firmness, and marbling, all of which recieved equal emphasis. To determine whether the official carcass measures actually ac- counted for the amount of emphasis each was supposed to contribute to the overall index value, a standard partial regression analysis was utilized. Carcass score was based on l00 points. Five variables, each accounting for a certain percentage of the overall value, were used in this study. Percent ham along with average backfat thickness and loin- eye area were each allotted a maximum value of 25 points. Length con- tributed I0 points and overall quality I5 points. Data from 87 barrows entered in the I965 Spring Barrow Show were analyzed for evaluating the relationships between firmness of loin-eye and length, average backfat thickness and loin-eye area. Subjective evaluation of livestack has been used for many years in the selection sf livestock and mast likely will be used for many stock judge is effective in detecting real differences in carcass cs from live animal appraisal. Var.ati3n in the eval- .IJ uation by judges is quite common since cu :titative trait predictions are based on some qualitative measure such as score. Gregory et al. (I962) cancluded that experienced judges were rea- sonably accurate 2n predicting the means of groups of steers for sever~ I al traits including, carcass we ght (ggven live weight), fat thickness, loin-eye area, percent kidney fat, cutability, and grade, providing the judges had some previous knowledge of the feeding program. Simple carrelations between est;mated and actual carcass weights ~0.97 t3 0.98) while the correlation (J .1 0) Q. I 3 D... LO (D U; fa were quite high f percent kidney knob were less for 0 indicate that the 3 D. < D. I: (1) Ill ‘1’- (I) (D ”i U) f“ i O :1 O I 0‘ V. -4 Z i“ U‘ (I) (D (I) 3 U) ('1- th, basis of carcass weight 77 (l (D 0 LI) fi- fI' (D 0 .I graders could accurately ran live weight, but were unable to rank the h providing -hey were given th steers on the basis of percent kidney knob. Simple correlati ns between the estimated and the actual fat thickness at the twelth rib were in the range of 0.3 to 0.5. Similar results were obtained for live cutability estimates and estimated cut- ability from carcass measures. ‘fl’ ‘ n ‘ 1 I ' 7‘ . ' ‘ , :3 rtmn_:"'~ .___ _‘_ , v 1 ~-:',:: :1. r. ..h ..., .3 ' f. 2; '.!-._ “j 1.“ .... ‘- :1 .5.” "l ‘ ,5 F9. :1 l. U 41" b'd I'Jg'Ofl-r“; . no. I ‘4’ "o“~u '1. a I."-". “a L“..'» ~c|H-. .— ‘Ih L...:-. L-guu -Ii‘a' J 4 5 quite accurate in grading the groups of cattle they were rather poor I, in their estimates of care 35 grade on ini;vidual stce.< with corral» 1 rtions ringing tram 0.l: t: 0.l8. U Correlations between estimated cutability and actual cutability using the average estimates of the three graders, involving two or more traits that are used as indicators of cutability, are of the same magnitude as those involved in estimating the cutability by individual graders (0.37 to 0.5!). Simpie correlations between cutability and the individual estimates of those traits involved in cutability were in the range of -.l7 to -.L2. The correlations were negative (~.25) between estimates of loinueye area and cutability because of the relatively high positive correiation that exists between carcass weight and loinm eye area. When carcass weight was added to loinmeye area in the mul- tiple correlation equation a positive multiple correlation was obtained (0.h9) between loinueye area + carcass weight and cutability. Thus, some additional precision was gained by increasing the number of inde- pendent variables evaluated, however the contribution of percent kidney knob was negligible. It was further found that conformation grades based on cutability estimates were more accurately predicted than were carcass quality grades, the correlation between estimated cutability live and actual cutabllity was 0.52 as Opposed to 0.29 between estimated carcass grade and actual carcass quality grade. Wilson _£._l. (i96k) found the correlation between live estimates of fat thickness and a single fat thickness measurement to be 0.5l. This suggests that overall fatness of the beef carcass may be predicted with reasonable accuracy. The correlation between the actual cutabil- ity and the live estimates of cutability averaged 0.h4. Based on a prediction equation (live weight and an estimate of loin-eye area, fat thickness, and percent kidney knob) a multiple correlation coefficient of 0.51 was obtained between carcass cutability and the predicted cutw ability on a live animal basis. Estimated fat thickness alone was found to be able to account for 2] percent of the variation in carcass cutability which suggests that fat thickness plays a relatively impor- tant role in the determination of cutability. Hence, it can be con- cluded from this study that a single estimate of fat thickness is prac- tically as good an indicator of cutability as the equation containing the four independent variables. In a study involving the evaluation of yearling steers by three judges Gregory at 31. (l964) found that judges were able to estimate group means for cutability and carcass grade quite accurately. How- ever visual appraisal was not nearly an accurate enough indicator of the actual carcass traits of individual steers to be of significant value. These results would be expected since some of the errors made in the estimates of any individual steers would tend to cancel each other, making the estimated group means coincide more closely with the actual group means. They also found that the live estimates were more accurate than were several carcass measurements in estimating cutability. Group means from live estimates of cutability were closer to the actual group values than were the means that were obtained from the regression equa- ’4 tion according to Murphey 23 El, (3960)! using several cooler measure- ments and estimates. Kidwell (l955) found a highly significant correlation between live grade and carcass grade (0.60). Cook (l95l) and Yao (I953) rem ported values of 0.59 an? 0.7! respectivelg=whioh were quite class to those reported by Kidmali. Wilson _£_§l. (l96h) found a simple correlation coefficient of 0.25 between the estimated grade and the actual quality grade. it was felt that the prediction of percent yield was more accurate than was the prediction of overall quality grade (0.4% versus 0.25). Qual- ity grade, being composed of both a conformation grade and a meat qual- ity grade, would seem to be more difficult to estimate since quality has been found to be quite difficult to estimate on the live animal. Wheat and Holland (l960) in a study involving the estimation of carcass grades, under differing conditions, obtained correlations of 0.56, 0.38, and 0.22 between the estimated carcass grade and conformat- ion, quality grade before ribbing, and quality grade after ribbing, reSpectively. There were no differences in the ability of the 12 l. (l96l) found that meat quality grade is most reSpon- sible for the error between live animal estimates and carcass estimates of overall grade. Correlations between live muscling scores and actual muscling (based on loin-eye area alone) were significant at the 0.05 level. I Cutability = 52.66 - 5.33(estimated fat thickness, in) + 0.665 (estimated L.E.A., sq. in.) ~ 0.0065(estimated carcass weight, lb.) Davis gt al. (l964) in a study comparing ultrasonic measurements J) and visual appraisals of total muscling in beef steers found that graders could be ranked according to their ability to assign individual steers to three different muscling groups. The muscling groups being designated as light, medium, and heavy. The three muscling groups were based solely on the mean loin—eye area1 as the indicator of total mus- cling. Analysis indicated that there was a 0.99 square inch increase in loinmeye area for every lOO pounds increase in live weight and a de» crease of l.0h square inches for each 0.l inch increase in fat thickm ness. Thus if the graders were able to accurately appraise the steer‘s live weig.t and fat thickness in relation to the other steers they could accurately rank the cattle into one of three muscling groups using loin- 1;) eye area as the indicator of total muscling. However current re earch indicates that loinneye area is not a good indicator of total muscling in the beef steer. Davis gt al. (l964l using Ultrasonics and other live animal esti- mates for lambs, showed that there was a correlation of 0.59 between live subjective estimates and actual loinneye area. A correlation of 0.25 was found between live estimates of fat thickness and actual fat thickness. Ultrasonic estimates of fat thickness were more accurate than were subjective estimates, but the reverse was true in estimating loin-eye area. Thus it seems that most of the research in this field indicates that carcass conformation can be more accurately evaluated on the live lAdjusted Mean L.E.A. = Actual L.E.A. adjusted for live weight and fat thickness. animal by competent j; dges than can the overall carcass quality grade and that fat thick“ is the most important single criterion in dew termining carcass cutability. Orme (1958) cor ducted a study to es imate the merst of several live as t °mates ar .d measurem nts of steers for various traits. Cora relations were obtained between two 5 bje c tive measures of evaluation. One measure was an unadjusted live animal score which was determined by subjective evaluation of a"chemical steers. The second measure was called an adjusted score. it was made by evaluations of each 5 ea while his legs were obscured from vi The author was working on the premise that if the legs could not be seen the various estimates of the individual traits, as well as the overall score, would not be highly correlated between the adju U3 te ed and unadjusted evaluations. however this was not the case. Correlation coefficients betwe een adjusted and unadjusted scores for live animal traits were as follows; type score (0.8h), estimated carcass grade (0.87), dressing percent (0.90) fat covering (0.79) and loinmeye area (0.89}. For both adjusted and ur adj;:_ ted evaluations, correlations be« tween the actual carcass grade and live animal scores were determined. The highest relationships were obtained be tween actual carcass grade and estimated grade, dressing percent, fat thickness, and loinmeye area. in order to determine the utility of visual appraisal in estimat- ing various carcass measures cor relation coefficients were calculated betwt * carcass measures and unadjusted live animal scores. Subiective live animal scores were significantly correlated with their corresponde no I U) a {h b n .1 II I 3 I I I I I ‘ f¢ n r I :i l 1:. h o -r ’T‘ d a (‘1 _I‘ It E"‘ r" Y ,1. k‘!‘ S' m n J (K {1. L!‘ LI) ... 0 CL. H _ K 3 m .1 \l l h I! ’ 0 E 0‘ S l) 3 O 1 I 11' Q a l 3 L gl ’ [1 <2 m l ‘- O- :3 :1 r 1‘ =!I LL "h J 3 (-0- T D ‘r D U) i ‘Y‘ [I .i‘ i (... 1" 21‘ . ) E J! '3 i Y) Also a large share of the estimated live animal traits were fotnd to be highly related to the acttal carcass grade. Cregory 2E.E£s (l962l obtained correlation coefficients ranging from 0.0% to 0.L3 between lire muscling scores and specific variables in the carcass. Overall res;lts indicated that the graders were able HI .1. O 3 J O I” ' i 0 {D II: U! ('1' ’i 11) (—f U) to account for about 20% to 25% (‘1 based on live animal scores. Holland and hazel {l958fi presents: correlation coefficients bee tween live animal condition a: f\ U 'i (h 1! ‘3 U a.” ll ‘1 LL t a 1 fl LL- :-r J \h If) U) 0) O m '0 m ) C). ”n ‘7‘ ' I O (I: J I? J! lean cats and percent fat cut, in swine (“.25 and 0.22 respectively). :1’ The asthors felt that live evalzation Stores were no highly correiam ted with percent 1&1: or percent tat nut: Eratzlsr and Margergm {1953) studiei the relationships between live swine scores and carcass measures. They found that judges were " - as \--.":»-~ ”an '-I --°. -- A: --¢-"~. " |.\ - I . ‘ , ‘.I . on‘ least accsrate in the.r evaloation or pezcent preferred cuts on a live i ’h \D ‘ r- O” J! U3 J 3 In D. r; r..- 6 L) .3 Q T‘ n 4. mates made on heavymweight swine were less accurate than those made on lighter swine. The highest correlam tion coefficients were those for estimated body length and backfat. These estimates were Highly significant for the light and mediummweight D hogs. They concltaed that visual scares for swgne were not highly rem ll Zoellner_§£_gl. (l963) found that judges varied in their sub- jective scores for swine. The correlation between judges for total desirability scores (0.60) was higher than the correlations between judges for individual items of meatiness (0.32) and finish (0.52). Repeatability of live animal scores is one important phase in» volved in the selection of livestock. if the judge is highly repeat- able in his estimates he most probably will be better able to select his livestock, as the sources of error in subjective measures will tend to decrease. Wheat and Holland (l960) found the correlation between slaughter grades per animal by different judges to be 0.50. Ternan st 21. (l959) found the repeatability of conformation scores to fall between 0.50 and 0.76. Zoellner'gt'gl. (1963), in a study involving two scoring systems found, by comparing the total correlations with the intra-season cor- relations, that judges were consistent in their scoring of swine from one season to the next. This indicates, according to the author, that the judges' picture of the ideal did not change and that one judge scored the swine the same way as did the other judges from season to season. Pooling of the scores for each pig resulted in a correlation of 0.76 between the two scoring sy if: tems used, which indicated that the pooled scores would be highly repeatable. From this review of literature it seems, that cutability of beef cattle is more accurately estimated in the live steer, providing the live weight is known _nd the judges are aware of some of the managing procedures, than is carcass grade. it also seems that fat thickness is l2 the most important single factor in determining cutability and that it is as good a predictor of cutability as any other single criterion. In fact, from work by Wilson gt a1. (l96h) it appears that fat thick- ness alone is essentially as good an 'ndicator of cutability as is any equation that was developed utilizing several factors. Repeatability of live estimates of carcass traits also seems to be moderately high. The work of Gregory 23 a1. (l96h) points out that groups of cat- tle can be more accurately evaluated for cutability, loin-eye area and grade than can individual steers. Subjective methods of live evaluation of swine do not seem to be reliable estimators of percent lean or percent fat cuts. Backfat thickness seems to be the easiest live trait to estimate in swine. DATA Cattle Evaluation Cattle used in this study were yearling Hereford bulls from the research herd located at the Lake City Experiment Station. The data wee obtained from l6 bulls in 1963 and 27 bulls in l96h. These bulls were part of a breeding project currently being carried out at Michi- gan State University and were progeny from bulls selected either on tenderness or leanness. The I963 bulls represented four sires, two selected on tenderness and two selected on leanness. Bulls evaluat- ed in l96h represented five sires. Twelve were from the foundation sires and fifteen from bulls either in the tender line or leanness line. Evaluation of the bulls was carried out two weeks prior to slaughter and each bull was appraised twice by each of four judges, the appraisals being four days apart. Bulls from both years were evaluated individually by each judge and all judges worked inde- pendently. Prior knowledge concerning the bulls was restricted to live weight, percent preferred averages and loin-eye averages of bulls from prior years. Live estimates included percent preferred cuts, to the nearest 0.l%, and loin-eye area estimates to the nearest 0.l square inch. Actual values for these measures were obtained on the bulls follow- ing a 48 hour chilling period. Loin-eye areaém the l2th rib, for each side of the carcass, was traced on acetate paper. Two l3 34 planimeter readings of loin-eye area were obtained for each side of the carcass. The average of the four readings was then used as the actual loin-eye area. Percent preferred cuts were calculated by sepa- rating the round and loin from each half of the carcass and trimming them to no more than 3/8 of an inch of fat. The percentage was then calculated as the ratio of the preferred cuts to the total carcass weight. Swine Evaluation Two groups of swine were used in this phase of the experiment. The swine used represented several breeds and on occasion crossbreds. Thirteen head were used initially and were obtained from swine entered in the 196A Farmers Week Contest held at Michigan State University. The second group of swine came from entries in the 1965 Farmers Week Contest. The data from these swine were subdivided into two cate- gories, heavy-weight barrows (2l5 - 235), and light-weight barrows (190 - Zlh). There were lh heavy-weights and l6 light-weights. Swine from the l96h contest were evaluated by four experienced livestock judges, two of the judges are well known carcass judges and the remaining two are well known live animal judges. All of the judges have had prior experience in evaluating both carcasses and live animals, thus all four participated in both evaluation phases. The swine were appraised twice by each judge prior to slaughter and the carcasses were evaluated once by each judge after a #8 hour chill. Judges did not know any of the live weights nor did they have any previous knowledge of the animals to be judged. Each judge worked lf independently and estimated the following characters in both live and carcass evaluation; percent ham to the nearest 0.l%, average backfat thickness to the nearest 0.] inch, and length to the nearest 0.l inch as measured from the first rib to the aitch bone. The actual carcass measurements of backfat were the average of three measurements from opposite (l) the first rib, (2) the last rib, (3) the last lumbar vertabra. The length of the carcass was measured from the leading edge of the first rib to the leading edge of the aitch bone. The ham was removed from the carcass and was trimmed to 3/8 of an inch of fat. The percentage of ham was calculated as the ratio of the ham weight to the carcass weight. Data from the I965 Farmers Week Contest swine were analyzed to estimate the correlation between live placing in the ring and carcass placing. Two experienced swine judges ranked the live animals basing their judgements on live estimates of carcass desirability. Two main factors were considered by the judges in arriving at the final placing, overall fatness and muscling. Estimates of leanness were based on turn of the top (curvature over the loin edge), trimness of jowl, underline and shoulder, firmness at the base of the ham and overall firmness of the animal. The basis for the muscling estimates depended on width through the center of the ham, depth of ham, length of rump and ham, turn of the top, and muscle movement in the stifle region when walking. Length did not become a factor unless the estimated length fell below 29.0 inches. Two groups of swine were placed sepa- rately. The heavy-weight hogs were ranked from I to lh and the light- weight hogs from i to l6. # U3 Carcass placing was determined by the use of an index value based on overall carcass desirability, the maximum value being lOO. The index value was determined by several factors including, percent ham, length, backfat thickness, loin-eye area and meat quality. These factors and the points assigned to them are presented inappendix table 8. Evaluation of the Index The third phase of this study was concerned with the evaluation of the carcass score that is currently being used at Michigan State University for placing swine carcasses entered in competition. Use of an index is valuable in determining the overall merit of swine carcasses. Hewever it is important that the factors that contribute to the overall variation of the index are actually accounting for the amount of the variance that they were originally designed to contri- bute. For this reason data were collected from several sources over a period of time to try to establish what variance the individual com- ponents were actually reSponsible for and if this portion was the actu- al amount that was assigned to them in setting up the index. Thirty pigs entered in the l965 Farmers Week Contest and thirty-three bar- rows entered in the Spring Barrow Show were used in this study. Eighty-seven additional barrows entered in the Spring Barrow Show and slaughtered at Farmer Peets were used to estimate the degree of relationship between firmness of the loin-eye and length, backfat thickness, and loin-eye area. METHODS OF ANALYSIS All analyses, save the varianCe ratios, computed on the Control Data Corporation 3600 computer at Michigan State University. Correla- tion coefficients were used to estimate the judges' ability or accuracy. Cattle Simple correlation coefficients were obtained from both groups of cattle and all judges. Correlations calculated were as follows: each judge's first estimate with his second estimate (repeatability), each judge's first and second estimates with the actual carcass values, the average of each judge's first and second estimates, the average of all judges' first and second estimates with the actual carcass values, and the average of all estimates with the actual carcass val- ues. Simple correlations were also calculated for actual loin-eye area with actual percent preferred cuts as well as for estimated loin- eye area with estimated percent preferred cuts. Means and standard deviations for each of the judge's estimates of the cattle were calculated and compared with the actual group means and standard deviations for both years. This included each judge's mean estimates and the mean of all the judges' estimates. Significance of simple, partial and multiple correlations as well as standard partial regression coefficients were obtained accord- ing to Snedecor (l956). l7 l8 Partial correlation coefficients were computed according to the following formula supplied by Snedecor (1956). r12-3 =l/(b'12.3)(b'21.3) A three by three completely randomized randon-effects model was used in the analysis of variance. This was performed on all of the bull data each year to test the significance of the different sources of variation. The three way interaction term was used as the denomi- nator in the F test for testing interaction effects. The expected mean square for bulls (A) times judges (B) interaction is UZABC + CUZAB and the expected mean square for bulls (A) times Judges (8) times replications (C) is UZABC hence by using ABC interaction as the denominator in the variance ratio the only variance remaining in the AB interaction tern is that which is due to cazAB , which is the var- iance component we are interested in.1 To obtain a variance ratio for the effects due to bulls, judges, or replications the following formula was used. MSA F' (3-] ox) = This formula was used because no exact F test exists for testing the direct effects for a three factor model such as was employed for this 2 problem. Thus an approximate procedure due to Satterthwaite was used. l Guenther, Analysis of Variance. p. l30. 2 lbid. p. 131. l9 Since MSAB + ”SAC - MSABC and MSA both have the same expected 2 2 .,. .. 4‘3: ...” ' i"- k . . .. r- AB + b? AC , under tau “all hypotheSis, if the va. iance ratio exceeds l.0 the excess will be that variance due to factor 2 values, 0 ABC + Co A. The degrees of freedom for the denominator used were approximated by using the following formula also due to Satterthwaite. ( “5A3 + MSAC ' “SABt )2 2 2 2 _ (MSAB) + (MSAC) (MSABC) (ea-I) (b-i) Ta-lNc-Il * (a-n (b-mc-I) Swine Simple correlations were also calculated for the judge's first and second replications of live swine estimates with the actual car- cass values. Correlations between estimates made by judges from car- cass observations and the actual carcass values were computed. Simple correlations were calculated between the carcass valces and the averages of each judge's two live estimates and the average of eight (four judges X two replications) l.ve estim C", U U? u‘i" Analysis of variance was performei on the live estimates of care cass traits. The method employed for the analysis of the bull data was also used for the analysis of the swine data. Rank correlations between carcass placing and live placing for light and heavynweight swine was performed according to the formula as applied by Spearman], where d is the difference between the two ranks assigned to an individual. atdz' 5 — niniml) Snedecor, a. w., Statistical Methods. p. The value of the sw.ne carcass andex was evaluated on data col- lected from three different sources. To evaluate whether each component xpected portion, based on the amount (T: of the index was contributing its of points assigned to it, of the total index, standard partial regres= sion coefficients were calculated. Using some of the data from the above experiment the individ- ual relationships between firmness of the loin-eye with carcass length, backfat thickness, and loin-eye area were estimated by the use of standard partial regression coefficients. Simple correlations were also calculated. Multiple correlations were also employed for this phase of the analysis. The standard partial regression coefficients are presented as path coefficients. RESULTS AND DISCUSSION The results of the various phases of this experiment may most advantageously be presented in tabular form accompanied with appro- priate comments and discussion where nedded. Table I shows the mean live estimates of loin-eye area and per- cent preferred cuts to be quite close to the actual mean values of the bulls reported in the carcass data. Mean loin-eye area was more accu- rately predicted in both years than was the mean percent preferred cuts. The reason for these results may be seen in the following table. Tabie twg shows that in both 1963 and l96h the simple correla- tion between actual loin-eye area and actual percent preferred cuts was essentially negative or zero, whereas in both years the estimated loin-eye area and estimated percent preferred cuts were moderately cor- related, with loin-eye area accounting for about 30% of the variation in percent preferred cuts. Thus it appears that the judges estimates of percent preferred cuts was based to some extent on their estimates of loin-eye area. In general those bulls with the highest estimated loin-eye area received the higher percent lean cut values when in fact this should not have been the case since actual values showed no cor- relation of any significance. Table 3 indicates that all judges had some competence in pre- dicting the loin-eye area of bulls in both years. However, in pre- dicting the percent preferred cuts all judges, save judge h in l96h, could not accurately predict percent preferred cuts in either year. 2i 22 Table l. Means and Standard Deviations for Actual and Estimated Val- ues of Loin-eye Area and Percent Preferred Cuts for Bulls. 1963 l96h Trait or Judge _flggfl_ St. Dev. _£E§fl_l St. Dev. Loin-eye Area Actual 10.01 0.78 l0.06 0.58 l l0.59 0.63 lO.lh 0.76 2 9.46 0.98 l0.l3 0.94 3 l0.l0 0.84 9.97 l.2h 4 9.98 l.00 l0.l7 0.80 Ave. l0.08 0.86 l0.l0 0.99 Percent Pref. Actual 38.87 l.27 39.78 l.00 l h0.l8 0.57 38.58 0.923 2 38.19 l.3l 38.8A l.03 3 39.05 0.79 38.83 l.Ol 4 39.36 0.87 38.63 0.98 Ave. 39.20 0.88 38.72 0.99 Table 2. Simple Correlation Coefficients Between Actual and Estimat- ed Carcass Traits. l263 l264 Actual L.E.A. with Actual Percent Preferred 0.l3 -.2l Estim. L.E.A. with Estim. Percent Preferred. 0.52* 0.67** 23 Table 2 may help to explain why estimates of loinmeye area were more ac- curate than were estimates of percent preferred cuts. Since percent preferred estimates were based to some extent on estimates of loin-eye area and since the judges overestimated this relationship, correlations between estimated and actual percent preferred cuts were lower than the correlations between actual and estimated loin-eye area. Table 3 also shows that judge 4 apparently readjusted his sights in 1964 as only he was able to register a highly significant correla- tion between estimated and actual percent preferred cuts. Table 4 shows the standard partial regressions of actual loin- eye area on estimated loin-eye area with carcass weight held constant. Also included in this table are the simple correlations between esti- s. The correlations indicate that the m mated and actual loin=eye are judges were able to predict, with reasonable accuracy, the loinneye areas of the bulls. however it seems that the judge‘s estimates were based to a large extent on weight as the regressions of actual on estimated loinmeye areas, with carcass weight held constant, were of small magnitude. Thus, with the exception of judge 3 in l963, judges were unable to predict the loineeye area with any significant degree of accuracy when the carcass weight was held constant. Figure 1 shows the path coefficient diagram of the relationship of actual loin-eye area with carcass weight and estimated loin-eye area. The numerical values are for judge 4 in l964. The reason for the -.88 may best be explained by remembering that the numerator of the standard partial regression (path coefficient) is equal to the correlation between the estimated and actual loinmeye area minus the Table 3. Mean Cerrsletien Coefficients cf Live Bulls Estimates with Actual Carcass Values. l963 l964 Trait or Judge Meana Meana Loin-eye Area l 0.l-+5 0, 58th? 2 0.70** 0,5oee 3 0.75** 0,53fifi 4 0 . 64th": 0 . 4597‘}: Ave.b 0.67** 0.53** Percent Pref. l -.09 -.25 2 -.l2 -.23 3 «.05 -.33 4 0.09 0,52ee Ave.b -.09 -.09 2 Mean = Average of each judge's two replications. Ave. = Arithmetic average of the four judges. Table 4. Simple Correlations Between Estimated and Actual Loin-eye Area and Standard Partial Regression Coefficients Between Estimated and Actual _£cin-e”e Area Carcass Wt. held Const. 1963 1964 Rep. 1 b' r bu r Judge 1 0.03 0.40 0.05 0.52ne 2 0.38 0.52% 0.03 0.55** 3 0.65% 0.71** 0.10 0.57** 4 0.63 0.60% -.03 0.55** Rep. 2 Judge l 0.l0 0.44 0.34 0.60** 2 0.42 0,51% -,43 0,39en 3 0.68% 0.61% -.30 0.43** 4 0.62 0.59 -.88** 0.34 £5 Table 5. Simple Correlations Between Estimated and Actual Percent Preferred and Standard Partial Regression Coefficients Be- tween Estimated and Actual Percent Preferred, Carcass Wt. held Const. 1963 1964 Rep. 1 b' r b' r Judge l -.06 -.14 -.OS -.19 2 0.21 -.04 0.13 -.29 3 0.42 0.05 -.08 -.38 4 0.10 0.07 0.46 0.47* Rep. 2 JUdge I 0.17 0.12 '017 -022 2 -.08 -02," 0035 -013 3 0.32 -.14 0.32 -.26 4 0.22 0.11 0.40 0.47* Figure 1. Path Coefficient Diagram Showing the Relationships Between Carcass Wt., Actual Loin-eye Area and Est. Loin-eye Area for Judge 44(j964l, . Estimated L.E.A. 0.34 0.87 Actual L.E.A. Ca cass Weight Table 6. Repeatability Estimates Between Judges Replications for 1963 and 1964. Est. Loin-eye Area (1) with Est. Loin-eye Area (2) 12.5% 1264 Judge 1 0-7 ** 0. 6*w 2 0.60* 0.77** 3 0.55% 0.90** h 0.75** 0.83** Ave. 0.66** O.83** Est. Percent Pref. (1) with Est. Percent Pref. (2) Judge 1 -.36 0.33 2 0.66** 0.71** 3 0.68** 0.80** 4 0.55* 0.6l** Ave. 0.38 0.6l** product of the correlations of estimated loin-eye area with carcass weight and actual loin-eye area with carcass weight. Thus in this case, the correlation between the estimated and the actual loin-eye area would have had to exceed 0.56 (product of 0.64 X 0.87) in order for the standard partial regression of actual on estimated loin-eye area to be positive. Since the correlation was only 0.34 the standard partial regression coefficient was negative. The relationships between estimated and actual percent preferred cuts are presented in table 5. It seems quite obvious that percent preferred cuts were more difficult to estimate on live bulls than was loin-eye area. Except for judge 4 in 1964, all simple correlations and standard partial regressions were non-significant, some were even negative. Comparing the correlations to the regressions however in- dicates that some of the judges were more accurate in their estimates when carcass weight was held constant.. Considering both tables 4 and 5 it can be postulated that esti- mates of carcass'weight played a double role in affecting those esti- mates of loin-eye area and percent preferred cuts. From past experi- ence the judges realized that carcass weight was positively correlated with loin-eye area. Therefore when the affects of carcass weight were removed statistically, a major portion of the correlation between it and loin-eye area was also removed. Hence the correlations between estimated and actual loin-eye area were substantially reduced. Since the judges initially overestimated the relationship between carcass weight and loin-eye area some of the correlations dropped to the negative Side, 27 The fact that the judges expected a moderately high positive cor- relation between loin-eye area and percent preferred cuts, in addition to the positive correlation between loin-eye area and carcass weight, caused the estimates of the percent preferred cuts to be positively correlated with the carcass weight. The estimates of percent preferred cuts was then interdependent on carcass weight through its relationship with the estimated loin-eye areas of the bulls. In reality the relat- ionship between the actual percent preferred cuts and the actual car- cass weight was negative rather than positive. Thus removing the ef- fect of carcass weight increased the relationship between the estimated and the actual percent preferred cuts, thereby causing the correlation to be positive rather than negative. Table 6 indicates that the judges were highly repeatable in their estimates of loin-eye area and most judges were also highly repeatable in estimating percent preferred cuts. Analysis of variance (tables 7 and 8) showed that in 1963 and 1964 there was a highly significant difference between bulls for estimated loinmeye area. in l963, differences 'n loin-eye area estimates made by judges were also highly significant but in 1964 this was not the case. Apparently by 1964 all the judges had readjusted their sights in the loin-eye area prediction and the average estimate of all of the bulls mean loin-eye area for each judge were in closer agreement. ,Although in 1963 there was a significant difference between the average of all judges' estimates of each bull and also between the average estimate of each judge's prediction on all bulls, there were no significant differ- ences due to the interaction between bulls and judges. Thus the Table 7. Anaiviis "‘ 5 ~ r‘ i -x .- r . _. 7‘ ,3 ‘, " ‘fi . r . . - . . _ . - --m—‘m-A -..--.. .‘n‘D-a-~--D~ - I‘- AA... ..C. y f. L :H.‘ 7.: Source 2&5; Mean Elgare T Bulls is 3.234 7.57:2— Judges 2 3.64 32.50/‘ Bulls X Judges 30 .24 l.67 Replications l .0} .07 Bulls X Replications TB .36 2.5} Judges X Repiications 2 .02 .lh Bulls X Judges X Reps. 3O .l4 Total (after the mean) 95 Table 8. Analysis of Variance For Loinueye Area jj96h). Source 0.? Mean Square F Bulls 26 6.45 lh.70** Judges 3 .47 .54 Bulls X Judges 78 .27 2.:l Replications l .l3 .lh Bulls X Replications 26 .33 2.32‘ Judges X Replications 3 .7h 5.75' Bulls X Judges X RepS. 78 .13 Total (after the mean) 2l5 29 judges tended to rank the bulls in the same order but the estimates of the means for the bulls was different among judges. There was a sig- nificant interaction between bulls and judges in l964. Thus in this year the judges tended to rank the bulls in a different order on loin- eye area. In l96h the estimates of loin-eye area for each bull was different from one replication to the next, also the judges average estimates of the mean loin-eye area were not in the same order for the two replications. Differences among judges was the only significant source of vari- ance in estimated percent preferred cuts in I963. That is, there was a significant difference between the average estimate of each judges evaluation on all bulls. In l96h the sources of variance in estimates of percent preferred cuts were essentially the same as for loin-eye area except the average estimates of all four judges for each bull was not significantly different from one replication to another. As can be seen in Table 9 only three judges were included in the analysis of variance. Judge two was excluded from this phase of the study because on an examination of the data it was established that he had dropped his estimates of percent preferred cuts a full five points for all bulls involved in the analysis. By doing this the judge introduced unwarranted variation into the problem. Since the primary interest was in determining which sources were contributing to the variation, his readjustment of percent preferred ctus to sub-normal level caused an uneXplainable interaction effect which in turn affected the analysis. Hence his estimates were deleted from the analysis of variance. 30 Table 9. Analysis of Variance for Percent Preferred Cuts ()9631; Source 2‘5; Mean ngare F Bulls 15 2.h2 2.#8 Judges 2 l0.27 6.82* Bulls x Judges 30 1.02 1.28 Replications l .0h .05 Bulls X Replications IS .75 .94 Judges X Replications 2 l.28 l.62 Bulls X Judges X Reps. 30 .79 Total (after the mean) 95 Iéble l0. Analysis of Varignge for Percent Preferred Cuts (196h). Source 2‘5‘. Mean Square F Bulls 26 h.08 2.8l** Judges 3 .47 .06 Bulls X Judges 78 1.23 2,90%* Replications l l.07 .l6 Bulls X Replications 26 .64 l.50 Judges X Replications 3 6.32 l5.00** Bulls X Judges X Reps. 78 .h2 Total (after the mean) 2l5 3i Live and Carcass Evaluation in Swine The analysis listed in table ll shows the judges were able to estimate the average for length more accurately than for either back- fat thickness or percent ham. The relative standard deviations of the judges estimates as compared to the actual standard deviations of the pigs indicate that the pigs actually varied more than the judges estimates. Carcass estimates of backfat seemed to be better indica- tors of actual backfat than did live estimates. However, the live estimates of percent ham were closer to the true values than were carcass estimates. The judges' live estimates of percent ham seemed to vary more than their estimates of percent ham on a carcass basis. this is shown by the relative sizes of the standard deviations. Correlations listed in table l2 indicate that the variance of the average of all judges was associated with about 50% of the total variance in length and percent ham but only 2l% of the variance in backfat. Generally Speaking the judges second replication was closer to the actual value than was the first. Judge D seemed to have reset his sights considerably between replications. Repeatability of live estimates (table l3) was high for all. except judge D. This is due to the fact that he had readjusted his sights between replications. Backfat estimates tended to be more highly rexaable than were estimates of percent ham or length. Analysis of variance for live estimates of backfat thickness in- dicated the major sources of variance to be the estimates among pigs, judges, and the interaction between the two sources. Thus the average of all judges estimates differed among pigs. Effects due to judges 32 Table ll. Means and Standard Deviations for Actual, Live Estimates and Carcass Estimates of Backfat, Length and Percent Ham »_for Four Jud-es. Trait or Judge Live-l - 5 - Carcass Mean éefle Mean §e2e Backfat Actual l.35 0.2A l.35 0.2h A l.59 0.ll l.AO 0.17 B l.h2 0.22 l.h5 0.07 C l.56 0.l8 l.33 0.l7 D l.42 0.09 l.36 0.l6 Mean l.50 0.l5 1.38 0.lh Length Actual 29.6 0.99 29.6 0.99 A 29 8 0.32 29.5 0.52 B 29 2 0 59 29.h 0 #3 C 29 5 0.h7 29 5 0 54 D 29.9 0.43 30.0 0.5h Mean 29.6 0.#5 . 29.6 0.5l Percent Ham Actual l9.7 l.h8 l9.7 l.h8 A 20.0 l.20 l9.2 l.l5 B 19 1 0.80 19 0 0 39 C l9.9 l.26 l9.l 0.82 D 18.3 0.h7 l8.8 0.67 Mean l9.3 0 93 l9.0' 0 76 33 Table 12. Simpie Correiation Coefficierts Between Live Estimates and Carcass Values in Swine by Judge and Fepiication. Replication Trait or Judge (ll (2) (l + 2)/2 Estimated Backfat A 0.26 0.28 0.28 B 0.39 0.35 0.37 C 0.48 0.48 0.48 D 0.09 0.5M 0.39 Mean 0.30 0.4] 0.39 Average Estimate of 8 = 0.h6 Estimated Length A 0.56* 0.4A 0,67* 8 0.52 0.h3 0.50 C 0 .561: 0 Jew: o .68~.2~.2- D 0.06 0.59* 0.53 Mean 0.A2 0.55* 0.59* Average Estimate of 8 = O.70** Estimated Percent Ham A O.h9 0.50 0.52 B 0.12 0.5A 0.36 C 0.79** 0.60* 0,71nn D 0.09 0.60* 0.43 Mean 0.37 0.55* 0.5l Average Estimate of 8 = 0.7l** 7.1.} .o’ ‘I 1 “ L : 2“. . . I’ .. "F’fi‘ "\ ... 6' g ”I .- ‘ jut“ ' 'fll In Tah.e 13. Repsa-at.l.tg of Line F5.i ates in .Nitw 731 i'ur uhd335. Judges Trait A B C D Est. 3.1:. (1) with Est. B.F. (2). 0.867%- 0.96%: 0.96='=2- 0.22 Est. Lgh. (i) with Est. Lgh. (2). 0.13 0.73-2.2': 0.88-Int— 0.10 Est. warn (1) with Est. %Ham (2). 0.81='r='r 0.72-:c'.- 0.87='n': 0.23 Table lh. Analysis of Variance for 1964 Farmers Week Live Estimates of Backfat Thickness. Source of Variation D.F. Mean Square F Pigs 12 .135 5.h** Judges 3 .2l2 l0.l** Pigs X Judges 36 .023 7,6** Replications l .007 2.3 Pigs X Replications l2 .005 l.6 Judges X Replications 3 .OCl 0.3 Pigs X Judges X ReFS. 36 .003 Total (after the mean) l03 Table l5. Analysis of Variance fer l96a Farmers Week L°ve Estimates of Percent He“. Source cf Variation D.F. Mean 9}“. F Figs l2 [Ll-+7 li.l8‘v'3‘«'~‘ Table l6. Analysis of Variance f-r l964 Pa'mers ive Estinates of Length. So-.ce of Variation D.F. Main Square F Pigs X Judges X Reps. 36 Total (after the mean) l03 36 were also significant and this shows that the judge's average estimates of the group mean for backfat thickness were not the same. The signi- ficance of the interaction of pigs X judges suggests that the judges could not agree on a general ranking of the pigs based on the backfat thickness alone. The analysis in table is indicates much the same story for per- cent ham estimates as table lh did for backfat estimates, whereas table l6 indicates that the judges could agree on ranking the swine based on length alone. Though the judges ranked the pigs the same the average of all pigs from judge to judge was significantly different. Although the judges did poorest in the live estimates of back- fat thickness, they were most consistent in these estimates as they moved from the live evaluation phase of the study to the carcass evalu- ation phase. These results are illustrated in table 17. This table also shows that the pooled live estimates of all four judges more near- ly approached the true values for all three traits than did the average estimates of any one judge with the exception of judge C on backfat thickness estimates. The coefficients in table l8 show that the correlations between estimated backfat thickness and the other two traits were generally higher than the correSponding correlations among actual carcass traits. The correlations between carcass estimates and actual values are shown in table l9. These correlations suggest the variance in the judges estimates are associated with about 64% of the variance in actual backfat thickness. In comparison to the live estimates reported in table l2, backfat thickness can be more accurately evaluated in the 37 Table l7. Simple Correlat 'ons Between Estimated Live Values and Estimated Ca c. s I ‘I . Judges Trait A B C D Est. B.F. (i+2)/2 with Est. C. B.F. 0.55* 0.60% 0.67* 0.21 Est. Lgh. (l+2)/2 with Est. C. Lgh. 0.51 0.48 0.6l* 0.48 Est. %Ham (l+2)/2 with Est. C. %Ham. 0.5} 0.37 0.37 0.h7 Est. B.F. (ave. of 8) with Est. C. B.F. (ave. of A) = 0.6h* Est. Lgh. (ave. of 8) with Est. C. Lgh. (ave. of h) = 0,63% Est. %Ham (ave. of 8) with Est. C.%Ham. (ave. of 4) = 0.55% Table l8. Simple Correlation Coefficients Between Carcass Traits and Between Estimates of Carcass Values. Judges Trait A B C D Actual Backfat with Actual Length -.€9 -.09 -.09 -.09 Estim. Backfat with Estim. Length -.33 -.h0 -.57* --07 Actual Backfat with Actual % Ham -.43 -.43 -.43 -.43 Estim. Backfat with Estim. % Ham -.73** -.68** -.93** -.20 Table 19. Simple Corr e and Actual C ass Es timate.s S i. Trait or Judge Estimated Backfat A B C D Mean Estimated Length Mean Estimated % Ham Mean Simple Correlations O.8l** O.68** 0.66% 0.87** 0.75 Average Estimate of Four Judges = 0.83** o.71=z--:: 0.8l='=': 0.50 0.19 0.63=’~‘ Average Estimate of Four Judges = 0.75** Average Estimate of Four Judges = 0.65* 39 carcass than it can by subjective live animal evaluation. Subjective evaluations of beth°the live animal and the carcass were equally ef- fective in evaluatiOns of lengthwend percent ham. Rank Correlation and Index Evaluation The rank correlation was calculated between live placing and carcass placing on both heavy and light-weight swine. These correla- tions showed that there was little if any relationship between the live placing and the carcass placing of either group of swine. The correla- tion between heavy-weight swine placings was -.02 and that for the light-weight swine 0.h8. Although the correlation coefficient for light-weight swine was not significant, it does Indicate that judges tend to be more nearly correct with regard to carcass merit in their live placings of light-weight swine than In their placings of the heavier swine. The material in table 20 illustrates that index values placed on the swine carcasses were reSponsible for their placings. Since carcass placings are graduated in units of one from place to place and index values do not follow the same pattern but instead graduate quite differently and Sporadically the correlation would not be expected to be perfect. In addition if two pigs were to have the same index value the tie would be broken by awarding the higher place to that pig which cut the higher percent ham. This too would tend to decrease the corre- lation. The coefficients in table 20 also show that there exists quite a difference in the amount of emphasis each component part of the index ho Table 20. Simple Correlation Coefficients Between Carcass Placing and Index Factors and Between Index and Index Factors, l965 Farmers Week Swine. Swine Within Within Factor Lights Heavies Overallc Carcass Placing with Index -.7l** -.93** -.92** Backfat 0.63** 0.39 0.73** Length 0.05 -.#3 -.0l Percent Ham -.5#* -.85** -.79** Loin-eye Area -.76** 0.l5 -.35 Quality Points 0.08 -.16 -.l6 Backfat Points -.55* -.50 -.69** Length Points -.h8 -.IO -.0l ‘% Ham Points -.6h** -.87** -.89** Loin-eye Area Points -.75** 0.07 -.h2* Index with Backfat -.76** -.6l* -.76** Length 0.28 O.hl 0.l6 Percent Ham 0.50* 0.87** 0.70** Loin-eye Area 0.60* -.26 0.3h Quality Points 0.02 0.2l 0.l3 Backfat Points 0.93** 0.60* 0.83** Length Points 0.63** 0.l3 0.30 % Ham Points O.89** 0.88** 0.83** Loin-eye Area Points 0.85** -.09 0.52** a in D.F. b 12 D.F. c 28 D.F. Ll plays in the index values and thus in the overall carcass placing, for lightwweight pigs compared to heavymweight pigs. This indicates that variation in the factors in the two classes of swine is quite different. For example quality points played a heavier role in placing the heavy pigs than the light pigs. This suggests that quality in the heavy pigs was more variable between pigs and hence played a more important role in determining the final placing, whereas bacfat thickness seemed to assume a similar role in the placing of lightmweight pigs. When the data for light and heavy-weight swine were combined, on the-whole, percent ham and backfat thickhess recieved more emphasis than any of the other factors. Quality and length played essentialiy to role. Tables 2i, 22, and 23 show the multiple correlation coefficients between the index values and between the carcass placing and the various contributing factors. in all cases the multiple correlation coeffECe ehts of the various indepehdent factors with the depehdeht index values all have values which are highly significant. The same is true concerning the carcass placings and the independent factors. The partial correlation coefficients of index with actual care cass values are not l.0 because each unit of difference in the car- cass values is not awarded l unit in overall index value. ’hat is, for example, backfat thickness values of l.37 inches and l.h0 inches both recieve the same number of backfat thickness points. However, the multiple correlation shows that the majority of the variation in index values is accounted for by the actual values of the component parts. 152 Table 2]. Partiala and Multiple Correlation Coefficients for I965 Farmers Week Heavy Weight Swine. Variable Partial Correlation Index with Backfat Thickness Length Percent Ham Loin-eye Area Quality Backfat Points Length Points % Ham Points Loin-eye Area Points Quality Carcass Placing with Backfat Points Length Points % Ham Points Loin-eye Area Points Quality O.6l* 0.05 0.86** 0.45 0.60 R = 0.9A** 1 . 00w: 1 . men:- I . 00-m- l . 00-m- 1.00%: R = l.007'r7': R = o .9Ln'n': a Partial correlations were calculated on each variable with all other variables held constant. 45 Carcass data from both light and heavy-weight pigs wer pooled for further analysis and the partial and multiple correlations are shown in table 23. Standard partial regression coefficients were also calculated on the pooled data and are presented in tables 24 through 27. These measure the amount of emphasis that is placed on each of the factors considered in the overall evaluation of the index. Each component part should contribute its allotted share of emphasis to the overall index value. Loineeye area was assi,ned a maximum point Q value of 25 yet, in tables 25, 26, and 27, the direct effects of loin- eye area " ccounted for only about l8% cf the Index va :e. Percent ham was the only variable that could consistently account for at least its allotted share of the index value on a point basis. This is shown most clearly in table 27 where percent ham po nts for the combined data was contributing over 55% of the emphasis in the carcass placing. Additional information gatnered during tn (D Spring Barrow Shtw was analyzed in the same manner. These results are.presented in tables 28 and 29. Generally speaking these tables show much the same result as the previous tables. index values, hence carcass placings, to be more dependent on percent “am than on any other factor, with the direct effects of quality points contributing next to nothing in the overall ranking. Factors Affecting the Firmness of the Loin-Eye Simple correlations show that barkfat thickness, length, and loinmeye area are all related to the firmness of the ioinmeve muscle an pigs with loinaeye area diSplaying the most marked relationship Tabie 24. Standard Parziai Ra.. rs 3r Cce" . ~'t" ' injcw or Actgai Carts": BEQ' 5; :_6S ?‘"”5"E;W ck " r3. - ' y: Qua D- e P “ ~F f\ f i2 variadle -v_:_~—_. I _-v--. 8A1. LC?- . (4".L A Backfat Thickness -.h3 0.l9 Length 3.28 0.08 Percent Ham 0.“? 0.l8 Loin-eye Area 0.35 0.13 Quality 0.32 _Q;12_, 3.38 Tabie 25. Standard Partiai Regression Coefficients of index on Index Point Factors. l965 Farmers Week Swine. Variable Std. Part. Reg. Coef. (A) __Lfiti Backfat Points 0.36 O.i3 Length Points 0.22 0.05 Percent Ham Points 0.56 0.3] Loin-eye Area Points 0.29 0.09 Quality Points 0.22 0.05 “7 Table 26. Standard Partial Regression Coeffigients of Carcass Placing on Carcass Values. lQES Farmers Week Swine. Variable Std. Part. Reg. Coef. (A) ._LALE Backfat Thickness 0.39 ' 0.15 Length -.12 0.02 Percent Ham -.56 0.32 Loin-eye Area -.26 0.07 Quality -.38 0.15 0.7] Table 27. Standard Partial Regression Coefficients of Carcass Placing on Index Point Factors. l96§ Farmers Week Swine. Variable Std. Part. Reg. Coef. (A0 _1513 Backfat Points -.l3 0.02 Length Points -.06 0.00 Percent Ham Points -.76 0.57 Loin-eye Area Points -.27 0.07 Quality Points -.20 _Jld¥i_ 0.70 #8 Table 28. Standard Partial Regression Coefficients of index on Actual Carcass Values. l965 Spring Barrow Show. Variable Std. Part. Reg. Coef. 1A) _jfil£ Backfat Thickness 0.ll 0.0] Length -.36 0.l3 Percent Ham 0.5k 0.29 Loin-eye Area 0.27 0.07 Quality 0.06 _Q;QQ__ 0.50 Table 29. Standard Partial Regression Coefficients of index on index Point Factors. l965 Spring Barrow Show Swine. Variable Std. Part. Reg. Coef. jflQ _jfil£_ Backfat Points 0.l8 0.03 Length Points 0.42 0.l8 Percent Ham Points 0.50 0.25 Loin-eye Area Points 0.38 0.15 Marbling Points 0.l0 0.0] Color Points 0.l4 0.02 Firmness Points 0.ll 0.0l 0.65 Table 30. Simple Correlations Between Loin-eye Firmness and Backfat Thickness, Length and Loin-eye Area. Variable Firmness Backfat Thickness 0,38ne Length 0,22% Loin-eye Area _,4unn #9 which was negative (table 30). This indicates that as the area of the loin-eye increased the firmness of the muscle decreased. Correlation [coefficients were calculated between-firmness and backfat thickness, length, and loin-eye area. Loinneye area and backfat thickness were the most highly related of the three, the correlations being ~r4h and 0.38 reapectively. Length had the least effect (0.22). Figure 2 is a path coefficient diagram showing the effects of length, backfat thickness, and loin-eye area on firmness. Although these factors do contribute to the variance in firmness between pigs their average effects can only account for about 19% of the variation, thus about 8l% of the variation is dependent on other factors. The coefficients in table 3] show the relationship of the dif- ferent factors when the other two are held constant. Loin-eye area seems to be the major contributor to the overall variance in this case as well. I Figure 2. Reiati orshi, 9e: een Firmness of Loin-eye and Backfat Th: ck.ess, Length} 2-5 Loan—eye Areafi Firmness 0'0 Qackfat Thickness -043\K // fig? Loin-eye Area Partiela and Multiple Correlation Coeff:c:erts for Lergth, and Loin-eye Area with Table 3i. . Backfat Thickness, Firmness. ial Correlation Variable Part Firmness with Backfat Thickness 0.l7 Length 0.l7 0.36* Loin-eye Area R = 0.52** Partial correlations were calculated on each variable a . With all other variables held constant. CONCLUSION It is apparent that live evaluation of beef cattle is a neces- sary part of selecting beef bulls and cows for breeding purposes. This study dealt with the accuracy and repeatability of judges in pre- dicting the loin-eye area and percent preferred cuts of beef bulls, both of which are significantly heritable. Judges were able to ac- curately predict the mean loin-eye area and percent preferred cuts of grOUps of bulls, providing the live weight was known. 0n evaluating each individual bull the judges were not nearly as accurate in pre- dicting percent preferred cuts as loin-eye area. Judges assumed that as the loin-eye area increased the percent preferred cuts also increas- ed ( r = 0.50 ), whereas in reality the carcass data indicated a nega- tive relationship. Repeatability of loin-eye area estimates were con- siderably higher than were estimates of percent preferred cuts. A similar study was performed on swine since live evaluation is also an important gJide in the selection of swine breeding stock. Live evaluation of swine indicated that the four judges were least accurate in estimating the backfat thickness and most accurate in predicting length both on the individual animal and on groups as a whole. However repeatability of backfat thickness estimates was the highest ( r = 0.90 average ). Percent ham was intermediate between length and backfat thickness in accuracy of prediction. Carcass evaluation of swine showed that judges were better able to predict backfat thickness and less able to estimate percent ham in Si 52 the carcass as compared to live animal estimates. in live evaluation as compared to carcass evaluation, correlations between backfat thick- ness estimates were the highest. It seems that judges can quite ac- curately predict percent ham and len th on the live animal but can not accurately predict the backfat thickness. Rank correlation between live placing and carcass placing indi- cated that judges, on the whole, can not rank sWine carcasses by live evaluation ( r = 0.48 and -.02 ). Evaluation of an index used to place swine carcasses based on backfat thickness, length, percent ham, loinweye area, and quality of meat showed that by far the most important single factor in placing swine carcasses was percent ham ( b' = 0.36 ). Multiple correlation coefficients between the index and the index factors were all highly Relationships between firmnes‘ of the loinmeye and loinmeye area, length, and backfat thickness indicated that loinmeye area and backfat thickness were by far the most important factors concerned with the firmness. The correlation was negative and significant ( n.4h ) for loineeye area and positive and significant ( 0.38 ) for backfat thick- ness. As an overall conclusion one could say that judges were able to accurately predict group means for several carcass traits in both cat- tle and swine but were not able to estimate these parameters on an indi- vidual animal basis. Carcass evaluation of swine was about as good as was live evaluation, being more accurate in predicting backfat thickness but less accurate in predicting percent ham. The correlation between 53 live placing and carcass placing indicated that judges were not able to accurately estimate the overall carcass merit of swine by live animal evaluation. The index used for placing swine carcasses appears to be adequate for this purpose but it is apparent that the individual fac- tors did not all contribute their allotted share of emphasis to the overall index value on this rather superior sample of swine. Loin-eye area was significantly related to firmness of the loin-eye muscle in swine, as was backfat thickness, whereas length was not. The relation- ship between loinweye area and firmness was negative and could account for about 20% of the variation in the firmness of the loin-eye. Back- fat thickness could account for about 15% of the variation in the firm- ness. BiBLiOGRAPhY Arthaud, V. H., C. H. Adams, L. A. Swiger, K. E. Gregory and R. M. Koch. i96h. Beef Cattle Progress Report. Animal Husbandry Dept., Univ. of Neb. Bratzler, L. J., and E. P. Margerum Jr. l953. The Relationship between Live Hog Scores and Carcass Measurements. J. An. Sci. l2:856. Carroll, W. E., J. L. Krider and F. N. Andrews. l962. Swine Production (3rd Ed.). McGraw-Hill Book Co., inc., New York, New York. Cook, A. C., M. L. Kohli and W. M. Dawson. 1951. Relationships of five Body Measurements to Slaughter Grade, Carcass Grade, and Dres- sing Percentage in Milking Shorthorn Steers. J. An. Sci. l0:386. Craft, W. A. l958. Fifty Years of Progress in Swine Breeding. J. An. Sci. 17:980. Davis, D. L., R. W. Ockerman, V. R. Cahili, W. J. Tyznik and C. F. Parker. l964. Ultrasonic Evaluation of Lambs. J. An. Sci. 23:l202 (Abstr.). Davis, J. K., R. A. Long, R. L. Saffie, E. P. Warren and J. L. Carmon. l96h. Use of Ultrasonics and Visual Appraisal to Estimate Total Muscling in Beef Cattle. J. An. Sci. 23:638. Enfield, F. 0., and J. A. Whatley Jr. l961. Heritability of Carcass Length, Carcass Backfat Thickness, and Loin Lean Area in Swine. J. An. Sci. 20:63l. Fredeen, h. T. 1953. Genetic Aspects of Can dian Bacon Production. Canadian Dept. of Agr. Pub. 889. h! Gifford, Warren, C. J. Brown and M. L. Ray. l9Sl. A Study of Cla5a sification Scores of Hereford Cows. J. An. Sci. 10:378. Good, D. L., G. M. Dahl, S. Wearden and D. J. Weseli. l96l. Relation~ ships among Live and Carcass Characteristics of Selected Slaughter Steers. J. An. Sci. 20:698. Gregory, K. E., L. A. Swiger, V. H. Arthaud, R. B. Warren, D. K. Hallet and R. M. Koch. l962. Relationships among Certain Live and Carcass Characteristics of Beef Cattle. J. An. Sci. 2l:720. 5h 55 Gregory, K. E., L. A. Swiger, B. C. Breidenstein, V. H. Arthaud, R. B. Warren and R. M. Koch. 196A. Subjective Live Appraisal of Beef Car- cass Traits. J. An. Sci. 23:1176. Guenther, W. C. 196#. Analysis of Variance. Prentice Hall inc., Englewood Cliffs, New Jersey. Hetzer, H. 0., and J. H. Zeller. 1956. Selection for High and Low Fatness in Duroc and Yorkshire Swine. J. An. Sci. 15:l215. Holland, L. A., and L. N. Hazel. 1958. Relationship of Live Measure- ments and Carcass Characteristics of Swine. J. An. Sci. l7:825. Jonsson, Per, and J. W. B. King. 1962. Sources of Variation in Danish Landrace Pigs at Progeny Testing Stations. Acta Agr. Scand. 12:68. Kidwell, James F. 1955. A Study of the Relationship between Body Con- formation and Carcass Quality in Fat Calves. J. An. Sci. 14:233. Lush, J. L. 1936. Genetic Aspects of the Danish System of Progeny Testing Swine. Iowa Agr. Exp. Sta. Res; Bul. No. 20“. Murphey, C. E., D. K. Hallet, W. E. Tyler and J. C. Pierce Jr. 1960. Estimating Yields of Retail Cuts from Beef Carcasses. J. An. Sci. 19:1240 (Abstr.). Orme, L. E. 1958. Methods for Estimating Carcass Characteristics in Beef. Ph.D. Thesis. Michigan State Univ., E. Lansing, Michigan. Reddy, V. B., J. F. Lasley and L. F. Tribble. 1959. Heritability and Heterosis of some Economically important Traits in Swine. Mo. Agr. Exp. Sta. Res. Bul. No. 689. Robison, 0. W., J. H. Cooksey, A. B. Chapman and H. L. Self. 1960. Estimating Carcass Merit of Swine from Live Animal Measurements. J. An. Sci. 19:1013. Snedecor, G. W. 1956. Statistical Methods (5th Ed.). iowa State College Press, Ames, iowa. Ternan, P. R., J. F. Kidwell, J. B. Hunter, C. E. Shelby and R. T. Clark. 1959. Association among Conformation Scores, among Body Measurements and the Relations between Scores and Measurements in Yearling Steers. J. An. Sci. 18:880. Wheat, J. 0., and L. A. Holland. 1960. Relationships between Slaughter and Carcass Grades in Beef Cattle. J. An. Sci. 19:722. Wilson, L. L., C. A. Dinkel, H. J. Tuma and J. A. Minyard. 196A. Live Animal Prediction of Cutability and Other Beef Carcass Characteristics by Several Judges. J. An. Sci. 23:1102. 56 Willman, J. F., and J. L. Krider. 19h3. A Study of the Characteristics of Live Market Hogs as Related to the Quality and C rcass Produced. J. An. Sci. 2:231. Yao, T. S., W. M. Dawson and A. C. Cook. 1953. Relationships between Meat Production Characters and Body Measurements in Beef and Milking Shorthorn Steers. J. An. Sci. 12:7?5. Zobriskey, S. E., D. E. Brady, J. F. Lasley and L. A. Weaver. 1959. Significant Relationships in Pork Carcass Evaluation. ii. Measure- ments and Cuts of Fat as Criteria for Live Hog Value. J. An. Sci. 18:583. Zoellner, K. 0., J. F. Lasley, L. F. Tribble and B. N. Day. 1963. Selection for Thinner Backfat in Swine. Mo. Agr. Exp. Res. Bul. No. 831. n-n- -r 11 aoie 1. Judges Estimates of Loinweye Area and Ferzent Preferred Cuts, T1013 F915] 1 55121055 9 I963 ' Jufige 1 ..".-d-';e 2 L E. A._ A Prefix 1. EL;EL_ % Pref. 13:11.1 (1)3 (23:3 :1) . 12;; 0) 321 1 10.3 10.5 41.0 39.0 10.5 9.0 40.0 37.0 2 10.5 11.5 41.5 39.5 10.0 13.0 40.0 37.0 3 11.0 10.5 42.0 39.3 11.0 30.0 42.0 37.0 4 10.5 12.8 33.0 4i.0 13.0 8.3 37.0 34.0 5 11.5 13.8 41.5 40.5 10.0 9.0 39.5 36.0 6 11.5 1.3 40.0 40.0 11.0 10.5 41.0 36.0 7 10.5 13.3 41.5 39.0 9.5 8.0 38.0 33.0 8 10.0 9.5 39.5 39.5 10.0 8.0 40.0 35.0 9 11.0 11.5 41.0 39.5 11.0 9.5 41.0 36.0 10 10.5 10.5 39.0 40.0 10.5 8.5 40.5 34.0 11 11.0 10.0 40.0 40.5 10.0 8.0 39.0 34.0 12 11.3 11.0 39.0 40.0 10.5 9.5 40.0 36.0 13 9.5 9.8 39.0 39.0 9.5 8.0 38.0 35.0 14 10.0 10.3 42.0 39.8 9.0 8.0 37.0 ’ 34.0 15 11.0 11.3 40.5 40.5 10.0 9.5 38.0 35.0 16 9.0 9.0 38.0 39.8 7.0 8.0 38.0 35.0 Judge_3 Judge 4 1 10.3 10.5 39.0 40.5 11.0 10.5 41.5 40.5 2 10.0 11.2 38.8 40.0 10.0 11.5 39.5 50.0 3 11.3 11.5 40.5 40.5 10.5 10.5 40.0 40.0 4 11.0 10.0 39.8 39.0 9.8 10.0 37.5 40.0 5 11.0 10.5 40.0 39.0 10.8 10.5 42.5 42.0 6 10. 11.0 39.8 39.5 .11.3 11.0 39.5 40.0 7 10.5 8.7 39.0 38.0 9.0 9.0 36.5 39.5 8 9.0 8.5 38.0 38.0 9.? 9.0 93.5 39.5 9 11.0 11.0 40.0 39.5 11.0 11.5 41.0 39.0 10 10.0 19.0 39.0 39.0 10.0 9.5 40.0 40.0 11 9.8 10.0 38.5 38.5 9.0 9.0 41.0 41.0 12 9.0 10.7 38.3 39.5 10.4 10.0 39.5 39.0 13 8.8 9.5 37.8 38.0 8.5 9.5 38.5 39.0 14 9.8 9.5 38.5 38.5 9.5 9.0 37.0 39.0 15 9.5 10.0 38.0 39.0 9.5 10.0 38.0 39.5 16 8.5 8.5 37.5 38.0 8.3 8.5 39.5 39.0 a Rep1icaLion. 1964). L referred ( _______ _‘ Hl' ‘ . v. , 'gwrcr+ E .r‘i c'~*--- I i _-_ pan, a u~d L) %Pref. 1 .0 U) 5 r we L.E.A. (2) L1) (2) E.A, (l) (2) Buil 08050000000“000000500550000 o o o o o o o o o o o o o o o o o o o o o o o o o o o 0908899777888889197009770089 L.39333333333334.5259333333141433 001770000008520022000000700 ......... 0000000 9887981199800890871900 Ali—1.14.51.14.142524.533.53.5.531414.3314.33143.41... 05030.4550505890880500010080 O 0 O 0 1019a 1093899880011891992081 ail-null I'll-HI Inn-8110! all 9005360«o01855.107085n49802h8939 ntviulaa..vq1.nu1183.U3301.1101379111U91090 ‘H'QI‘HI 31‘1““. all] I] 0008505050055552000000550088 ........... ..10008839w171831/39/88800780889989 14.14.95.522.) 5L. 3Q53?53.5.531414 .5314. 333.533 058085505005820580005008508 o o o o o o o o o o o o o o o o o o O o o o o o O 9787,18W/887R.8rb88]789790770888 221JQJQJ2JQJ2JQJ2JquQJQJQJQJLWQJQJQJQJQJAIQJQJLWQJQJQJ 880808C/r308000258C¢222C2225280 09449Oficfiu09~5~U99009fl01|9A01109109.1 1'13] II] I] I 5580.8022—5032nun/.OOBOOOSSOSOonwO 1....909nL1|OnLOJnJ099019AU119019910911 1'11] Cell I 1'] I] I i a - ..a ' . . . - 4 .13. - v- : ...) . ‘ . u - - a ~ — - - n .- “-O “--...‘—-m “um-n: .... ‘A—Q- -_ ....- - .—_ ‘.-s.‘ -_ ‘ ...—“n. “A F 59 :.'.-+9~-+-_,.- .rw - 1.3. - .. V 5" -..--M-.‘ C u. -..-n M“.-u—u--ww.--M -..—.4.-- -‘ .1 r1~-:r‘ '. ...-.. 10142014183814.62070295003002389). \ll 0 O O O O O O O O O O O O O O O O O O O O O O O O O O 2 007,797.96971771098307ca7/77AU8877.. /.\ 14.14..533.53333.5.5L.3.514b3.533.514.3339) 61‘. 0.0.22 4.108.422.102360 130.171-495ANO...) /. O O O 0 m ..HOQJI A7887990080700310000074789809 L{ z 1.... i4. 3.53.5.5 3.533.514 3.5.5.4 .51... .53 25.5.5514 .5 ”Pr E d u. L.0Jn..0032.00711.4. 26.521.13.12 6.4.14.1...0721 n-.. o o 19.119.33.100. .4..OJOI3).QJOJOJ..11:7_1931:99239.1. ...... Q.‘ Qufl‘qlqn' 1311'] 1' . ’1‘ (2) 90408.1”..10 AL8 ..1. nd88111..)31u_-..)3)0.8AL _..,I79C)3SQU. A....... r...» 0.. «... ...o. .....OJAAJQ)..A4)3:\U3J ....JAXVGJn.OJOJ.1...A..VAJJn I‘I‘- .‘IQII . ..JIIQI 1.111.. c.) EJ30.EJ88CJ/h: EJthnan8CJ2338AJ~I A/._CJ788_L.J O C O O O C C O I O O I O O O . 33.488333415388899q194 .83A4....70300.. ...». 31» .1) ...)...H 05......(.A....9).<1..«,...)1u A)n4. .1) 7%.)?!“ .).)b 7) 7J1». cF (2‘.- .44 '1...“ r nlndau8n/h8.k.:. .8CU8fi12n475flv1/32n.8EJEJAJJC) C 1 ‘1). o o o o o o o o o o o 1 .51“)... Quolnawux a... ..I...~........i 00.47.331.109 371.13.107.41 I14 1.1.. .4; Q... a.) 7) .11. ...... 1.73....) «(... 1) 7-. .51-...3...) 7.. ...) 31.». ..(J 2.1». ...). ...) 9.). .3 . L \L nJ/p277f.r.J.Ku—I.On_5358_)3.. QJAUAJJLWCJQJCJA. 7.. 114u19.19_AJ. 88992.9.11A11a882887_A18). . (L 1‘1‘} .51....1-“ ‘1'} 11".- ‘I o n48—7/C’AJAI_VIE/AJHJA‘JBQJC/OOOZA1 AIHZOAIWCJBSZ I.“ \j o o o o o o o o o o o o o .1 1938513189. 88.419110031991181. ...I. Inc)... 3111 569189.01154 314 EJ6783nv ..II 11 n15 31 H- w_ laxiflfll‘ulql‘lll 0‘22 3:567 22222 £21.]- L E A.a PETVc: ._ PFQFE ’E‘Eda 1 33.0 38.8 2 3M 39.0 3 ll 3 38.7 L‘ 9.8 8.3.8 5 - .. 6 30.] 36.5 7 30.5 39.5 8 9-4 41.3 9 - - m ”-6 39.5 F] 9.8 38.9 32 9-7 39.5 I3 8.8 37.5 M 9°5 38.2 ‘5 9-3 37.6 ‘6 9-2 38.2 i 30.8 38.9 2 30.6 40.7 3 9.8 38.9 1‘ 9-8 39.3 5 9-‘ 39.2 6 ‘0-2 39.7 7 ‘0-0 40.1 8 9-8 39A 9 9-3 39A 10 9.3 38.8 H 10.2 39.8 ‘2 ”-4 38A ‘3 9.9 42.2 11+ 10.2 39.2 35 33.3 39.7 V" 9-0 80.9 37 33.0 40.3 18 39.9 38.5 13 9.2 A].§ 23 ”-7 39.9 2‘] 10.9 38.5 22 IO.2 40.3 23 9-8 14.3.8 2” “-3 39.8 25 30.0 39.2 25 9.9 42.0 27 11.0 39.3 a 14 bulls in 3963 and 27 buifis in 3964. 6] Judge B Judge A %Ham Length Backfat %Ham LU Backfat Length 11) Pig,(l) L2) (1) (1) (2) Q) (1L (2) (2) (2) 5555055055505 888898L099898 111111 221111.] 5555500050055 888889L099980 llllll 2211-112 5550050555500 00.00.0000... 9889980099098 2222223322322 5555500550055 0 O O O O O O O O C O O 0 9889999999098 2222222222327. 3664.4681433667 O O O O O I O O O O O I O lullsIu-nlifl-illo lllllll 26643680246E16 1992112099808 211222221112] 0005550005550 0 I ....... 1999012190898 5550550550050 0 ... O C C. I O O. 3223323223322 0550505550000 0 .... O '0. C ... 0990909990100 3223232223333 #7756656657 0 O C O. O O llllllllalll] ......... 111111111111] Judge D Judge C 5966050508000 0 O O O O O O O O O O O 0 88788898970087 1111111111111 02.05055nwoo3nuoo...) _/.om800000098 9800.9Rw lllllllllllll 11111111111]! ............. Illlllllllll] 5050050055050 0 o ccccccccccc 1970102009988 2112222221111- 5050500005050 0000...... o 1991102010988 211222222211] 693705.“.9514229 0000000000000 0899099999098 3222322222322 08h5052876539 0000000000000 0899099999098 3222322222322 277506u555789 0 0000.00... 11111111111]! llllllllll‘lll . 0 I .- - . I .— ..' ngfi... -— A ... 2'70. - ‘1' 4 h .3. o - m .” L-M_J-—- M -..-L.— _A_ _ _‘__ r. "7“ ‘A‘ _ gr. s . 1'1“: "' 5.0.?- --~‘ ~- I I P h -10 !r4.g -- \=. p- ...—n m.-- - 5 ”I-‘ CI.- . _ 1...... 0. 5'1?“ 1‘ ./Q."'am Gin Cl I,‘ lot :13 R L —i» . Judge A :1. 385 L.,J .. VJ“: J fl: SZOSSSOOSanws 8999899998nxm98 llllilflnflllllilflfil QAUSOSSOSSSAUOAUOS o o o o o o hv989939939399 322222n42nl~nrhfl uc.0¢ ucm mo:.m> .m:uu< .mc.om.m mmmuumu qmcmum.¢ m>.. .N,m.an \JI)\n/\II) 5.14. 3....- nU 5 (I\ V! VI P: r- V: 6 D r V1,... 0 a: D vi a \«L \aJ \/))))\vi)\ul ... 00 6 ”W .... C) :7... 3 1 oo oo ...... 0 n. .Y ...... .... Ac. 64 pup—p.22?! .... a. mu. afv ..-... \.. )\Il\sl\../\ll)\./)\./\5/\JI I... Il\l(\(l\.lu\ll\(( C D n... M” w _ 5. 5L. 3 2 3 73 9.4 .5 O .L .J .. ..L __ 2L 2 ...... 2 Z 2 .1- ..I S r 8 ...r .J 5 c. (k (/1\(\(((((((( .... .t F O Y” a. (a .nm 0 .«L C0 5.. mm. p9+94914990dfi nu OJGOWbOP rd 0.... .4. .... o W .K 0 4L 3 Eer 78 U ..I .... 1i ... o .._ ....“ 1J1Npfiu998763 a. 1| 0 o o o c o FMPJCu!» «.... Avr£2nz7~1fllvl1lnullunu ... 013111114. f... .... n». ra. .ni ... ....a—u-n-—_—cu —. u--—.a ... e O ... ...... OUPDMUC/nbsNUOAUOQJ o +L01l.l-1III.II.NI 2; mi 0 o o O o o o o o o 0 Pk. flul r./- all 5.26 78 \I14 \I/‘n’r\...jl\u11) N. .1. .C 0... .5- wnl 0 an. 9 QJ 8 .../:6 ”L/ o D- o o o o o o o .L/ .Dt4. 2.14 fiJ u fl 7. ML 7. 2 ad 1!. ...i 1.: «2- 7| .3. R. n. 1| II 1| .II 1 al. .I.. In; ((11.2{sl \ A“. .-u .... 5 _ g k. . . .T. M..- g 6.. 0" “U IIIIII ...... .u DI «L d e ..4 ad .1. MC cl .4 .1. 3 a...“ m.- n M... _ \1. )\../)\.I Q Pt. 5 O .... hr 5. \11 \1/\dl))\ll\al)\bl\ul\al) an... O 5 2 0 .. . .2 5b. 3 2 .-- n... ..D ....-. .2). 1| n.“ a... .3 I! .K . 3 V. ...7 7.. nl. A... 7. CL a. 61. ..I 11. I.\ (((( 3 “I .91... r .... (((l‘ In\(!l\(((( r .d U a .5 .1 [A ...” C P... T. D. Vu Q i a L. .3 3 a; .. m. 3 4.31214. 94 914. 3....4. n .H Dual H. .1. r... F. C 2 9 7.4 r . 9.11.1... 2 w 3 U o o w. o o o o o o o o o O .L 8 8 3 ., ...: dq/I4uhn43242u.4.d Ann/”2H... : K! n... he Pu .4 s a . . . _ . . . . .8 S a. .5 an. m a; ...H .u .b3SOF\/nUSfiJSnd9 .. 0.45203 ... ... 7.. O 71:) 7.. C 7C).... ...u OJ .. o o o o \))\./\u/\|/ .... ....V o o o o o o o o ... 01.887 .1.) .u QJan. w. .L 5731...... h. ...H 3 3 7; .12 F; 2 7.. 2 7. {\(f...{\{\ C: a C. R. 5 ..2. C d 3 .nl L“ .... ..- 1-- . 1.]. "Vv A”. at... ... .3 3. .1 a V 3. A). vim WI. r AC ..._ M S S T D 66 .141 mr-w-L3W\.d” w-r-N‘mwwr—mJoF-moq- mmF-r—ou—mr-r—mmm a. £2 L1. LnOC‘JONM --cxl;\l-—t\1 ow mm 3 .N 2 ON mm mu. mN JN mm mm 5 mm mm mm .\ mm mm mm d c MN "N MN ....f LfiLfld' LnMOJ LnF—F—mdd'dgfo—F-OJ mm-—.:1~ommmmmmq r0 mmr—n—mQ—moo—a—o—u——mmo——QMMo—Omomo—o—p—u—Jm o L. I: o O L.) I‘ML‘xu'; CNS] (J‘IDLukC 7‘00“} LI‘\.;r--\0;f omr-me-mxoo—ooo .‘~\O Lnnfi O .:r:r-:r\o.;r~.ommmm.:r.a‘.:r.:rmmmde-J Md“ Lnd‘ md’ MJJMJM m-j‘ 0‘erch mood? mcoxooo N m»— ox-m mmN 00 moo "'u‘ \0 ?‘~‘7‘¢"‘"‘ ”“0 <1 0 LLJ I O ..l F '- MNI\OO\MO\I‘\O\O\-d‘ Md‘ F-o—o—N :— ow mw o o F-F-O\O\'-G\O\NOOO®O\O\OWON®O\®OOK\O\F~ NNp—p-Np—NNNNI—F-I—Nr—F-p—o—F-P-P-u—u— (DMFONQU‘OJommm—JmfiNw—NNWJJNO O o o o o o o o o o omooomOOOQD—ONomoo—OowmmNooo OOOONNLnfi-NNOLnNCDQNCDCDmCOOmmM O MNMNNMNNMMMMNNMMMNNNMNM O O O C O O O I O o—_—_—__——_——I-F-—'_—I-— MQNMMOOMW\OMO(‘DT\M «nmm: ermmmq'd‘mstmNN O FNMJUN\DP\CDO\ I ...- ..l -. 'fl. '.."1'.I.-.% li.‘l "II‘I‘IIIII'I *1“! ~‘1I.‘ II .xu.mgm>_c a; wumuw .2 gm bwLuucunm_m mw_L+ £C_£U .z.ncm zoLme mc_Lam mmm_ ox“ Low mm3_m> u:_om vcm mm:_m> ammo mu ~m_.uu< .m:_um.m mmmULmu .m m_mmh 67 Table 10. Actual Values for Backfat, Length, Loin-eye Area and Firm- ness. 1965 Spring Barrow Show P195 S1aughtered at Farmer Peets. qu, B.FJ Lgh. L.E.A. Fns. P19 B.F. Lgh. L.E.A. Fns. l 1.13 31.3 6.31 3 45 1.13 29.7 4.30 3 2 1.27 29.8 4.48 3 46 1.40 30.5 4.08 3 3 1.13 30.0 5.91 2 47 1.50 31.5 4.12 4 4 1.20 29.5 5.09 3 48 1.50 31.7 4.72 3 5 0.97 29.5 4.92 3 49 1.40 30.0 4.18 3 6 1.03 30.5 6.16 3 50 1.57 33.0 4.85 3 7 1.30 29.4 6.21 2 51 1.30 30.7 3.75 3 8 1.27 29.0 5.30 3 52 1.43 31.1 4.05 2 9 1.30 29.0 5.48 4 53 1.43 29.9 4.30 3 10 1.23 29.5 5.80 4 54 1.37 29.6 3.95 4 11 1.20 31.4 5.64 2 55 1.47 30.5 4.48 4 12 1.30 30.7 4.90 3 56 1.70 30.1 4.03 4 I3 1.37 30.5 4.10 4 57 1.37 30.3 4.51 4 14 1.13 31.0 6.85 2 58‘ 1.50 30.5 4.57 4 15 1.43 32.0 4.68 4 59 1.30 29.8 5.00 3 16 1.30 32.5 5.15 4 60 1.33 29.0 5.41 3 17 1.13 30.2 6.00 3 ' 61 1.47 29.0 3.71 3 18 1.37 29.6 5.68 4 62 1.30 30.4 5.02 3 19 1.13 29.8 4.10 3 63 1.30 27.5 4.38 l 20 1.33 30.0 4.56 3 64 1.17 28.6 4.72 3 21 1.30 30.4 4.05 4 65 1.43 30.0 4.05 4 22 1.27 30.5 5.23 4 66 1.23 30.8 5.15 4 23 1.07 30.6 5.60 3 67 1.13 30.5 4.95 2 24 1.03 29.5 4.92 3 68 1.13 28.5 4.14 2 25 0.97 29.7 5.33 3 69 1.13 30.0 4.21 3 26 1.17 29.8 6.95 2 70 1.27 29.9 4.56 2 27 1.20 30.5 5.20 3 71 1.27 29.2 4.02 3 28 1.07 29.7 5.26 3 72 1.20 30.0 4.75 4 29 1.23 30.0 5.66 3 73 1.07 27.3 3.67 4 30 1.20 29.0 6.56 2 74 1.43 28.7 4.62 4 31 1.13 30.0 6.35 2 75 1.20 30.2 4.85 3 32 1.43 30.1 4.50 4 76 1.23 29.0 6.43 3 33 1.27 30.8 4.85 4 77 1.57 31.7 4.43 4 34 1.23 29.5 4.05 3 78 1.43 29.5 4.04 4 35 1.07 28.8 4.48 3 79 1.27 30.1 5.35 3 36 1.07 30.0 4.71 2 80 1.27 29.7 4.23 5 37 1.03 30.0 4.92 3 81 1.27 31.2 4.37 5 38 1.27 28.1 4.46 3 82 1.10 29.8 6.13 3 39 1.20 28.8 5.57 I 83 1.30 29.0 5.43 3 40 1.60 30.4 3.77 3 84 1.23 28.2 5.75 3 41 1.30 29.5 3.95 5 85 1.10 30.5 5.75 2 42 1.50 30.1 3.61 5 86 1.17 28.6 5.38 2 43 1.27 29.9 5.45 2 87 1.20 29.0 6.41 2 44 1.30 30.0 5.50 2 MICHIGAN STATE UNIVERSITY LIBRARIES I ll llllll III II 3 129313177 7422