Masters Theses

Date of Award

8-1977

Degree Type

Thesis

Degree Name

Master of Science

Major

Animal Science

Major Professor

E.R. Lidvall

Committee Members

W.T. Butts Jr, J.W. Holloway

Abstract

Data from 180 Hereford feeder heifers were used to study the relationship between subjective scores and objective measures. Analyses were run to determine which variables within the subjective scores and objective measures accounted for the greatest amount of variance in the description of feedlot cattle. In addition, simultaneous correlations were run to determine maximum agreement among the subjective scores and objective measures. A definition of the subjective scores was obtained by regression analysis on the objective measures. Then, 180 steer calves of the breeds Angus, Hereford, Angus X Hereford, and Charolais crosses were used to determine grader effects on subjective evaluation of feeder cattle.

Prime, Choice and Good feeder heifers and steers were purchased through East Tennessee Feeder Calf Sales. The heifers had a mean weight of 495 lbs., while the steers averaged 630 lbs. The heifers were allotted to pens of 6 heifers each. The heifers were fed a standard growing and finishing ration consisting of corn silage ad libitum plus a concentrate equivalent of 1.40 percent of body weight (pen average). Amount of concentrate was adjusted each 2 weeks for the first 18 weeks of the trial and weekly thereafter until slaughter. Pens of heifers were slaughtered when the mean fat thickness of the 6 heifers reached 12 millimeters. No pens were sold prior to 18 weeks on feed.

The heifers and steers were photographed, weighed, measured for shoulder width, and sonorayed for subcutaneous fat thickness upon arrival at the farm and at 2 week intervals until slaughter. Heifers were independently scored by a single grader at initial, mid-way through the feeding period, and at slaughter on a scale of 1 to 15 for frame size, height, length, general trimness, head shape, and muscle expression. Estimated age was recorded in months, while fat at the 12th rib and predicted finish weight were estimated in millimeters and pounds, respectively.

The steers were independently graded at the beginning of test as feeder steers by a committee of 10 graders varying widely in live animal evaluation experience. These scores were then used to study agreement among graders and with initial objective measures.

It was found by factor analysis that major differences in Hereford heifers was accounted for by size rankings. Subjectively height, length, frame size, and predicted slaughter weight, along with objective weight and depth showed the highest factor loadings. When factor analysis was not describing a size dimension, it was scaling cattle on muscle expression.

Utilizing canonical correlation analysis, the first canonical variate correlating subjective evaluation with objective measures in Hereford heifers described size and fat differences in both data sets. The second canonical variate was similar to the first explaining more size and fat differences while the third and smallest canonical variate subjectively correlated muscle expression to objective width and depth.

To further define the subjective scores, they were regressed on the objective measures using a stepwise regression analysis. R-square values for stepwise regression analysis indicated for the initial time period that models describing height, length, frame size, fat and predicted slaughter weight of the Hereford heifers explained .40 or more of the variation, while those explaining other subjective terms was .11 or lower. For models relating to the structural dimensions of the heifers, objective height was the principal source of variation with partial F values in excess of 75. Measured height consistently entered more models than any other objective measure.

Subjective evaluation of the feeder steers by a committee of 10 graders revealed r-square values of .60 or more for height, length, frame size, and predicted slaughter weight, .45 for fat and .25 for muscle expression. Coefficients of correlation among graders ranged from .61 to .71 for subjective height versus measured height; .43 to .60 for subjective fat versus measured fat; and .30 to .41 for subjective length versus measured length. Muscle expression had a low correlation with all subjective and objective variables in the data sets. Coefficients of correlation among graders ranged from .69 to .92 for frame size and predicted slaughter weight, .45 to .79 for subjective fat estimate, and .15 to .60 for muscle expression. Graders were more consistent scoring traits associated with structural dimensions, somewhat less so in estimating fat and quite inconsistent in scoring muscle expression.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS