Relative contributions of subcutaneous and intermuscular fat to yields and predictability of retail product, fat trim, and bone in beef carcasses.

Carcass data from one side of 610 steers born from 1988 to 1990 in Cycle IV of the Germ Plasm Evaluation research program were analyzed to develop means for carcass traits and retail product percentages at two fat trim levels (.76 and .00 cm) by yield grade categories. Weights of subcutaneous (s.c.) fat and intermuscular (i.e.m.) fat were recorded separately at each trim level. Quadratic regression curves were plotted for percentages of roast and steak meat (R&S), retail product (RP), and fat trim components relative to incremental changes in USDA yield grade. Prediction equations were developed on a randomly chosen half of the 610 carcasses to predict weights and percentages of R&S, RP, and fat trim using carcass traits obtained at the time of USDA grading and then tested on the remaining half of the carcasses. In addition, prediction equations were developed using s.c. and i.e.m. fat plus carcass traits to evaluate the contribution of each to carcass fabrication yields. Percentage of RP, trimmed to either .76 cm or .00 cm of fat, decreased by an average of 3.5% for each full yield grade increase. Trimming to .00 cm of fat resulted in about 5.3% less RP compared to trimming to .76 cm. A prediction equation for percentage of RP trimmed to .00 cm using adjusted fat thickness, carcass weight, longissimus muscle area, and percentage of kidney knob had an R2 value of .54. The variations in percentage of R&S and percentage of RP at both trim levels were reduced by removing s.c. fat trimmed to .76 cm; however, considerable variation still existed. Subcutaneous fat expressed as a percentage of the sum of i.e.m. and s.c. fat increased as yield grade increased, but the percentage of i.e.m. fat was higher than the percentage of s.c. fat for all yield grades. On the basis of partial correlation coefficients, i.e.m. fat was approximately twice as important as s.c. fat in accounting for variations in fabrication yields.