Chemical components of ground beef produced on grass-or grain diets and their relation to beef flavor

Forty-eight yearling Angus, Angus-Hereford, and Angus-Charolais steers were randomly placed on one of two pasture systems containing 24 animals per system. One pasture (A) contained ryegrass (Rye) and the other (B) contained fescue, orchardgrass and clover (FOG). All animals were placed on pasture in the fall, followed by hay and legume feeding during winter and then were returned to pasture during the spring and summer. After pastures died in the summer, each group of animals were subdivided into three equal sub-groups, two of which were put in the feedlot and fed a supplemented whole shelled com diet. The third subgroup from each pasture was slaughtered directly off pasture (0 days on feed). The others were serially slaughtered at 56 and 112 days on feed (DOF). Ground beef was formulated to contain approximately 20% fat using the biceps femoris muscle and external fat from the round of each steer's carcass. The raw ground beef from each steer was analyzed for fatty acid (FA) composition of total lipids and lecithin.content. Each variable was statistically analyzed as a function of pasture, DOF and their interaction. Fatty acids from raw ground beef samples were correlated with levels of volatiles and intensity of flavor descriptors from quantitative descriptive analysis for cooked ground beef samples as reported previously. Regression analysis of beef fat flavor score (BFF) as a function of FA composition of the raw ground beef lipids and the concentration of 11 selected volatiles isolated from cooked ground beef samples was also determined.

In general, the percentage of monounsaturated FA's (MONO) increased and the percentages of saturated Fa's (SAT), polyunsaturated v omega-3 FA's (PUFA-3) and unknown FA's decreased in beef lipids with increasing DOF from 0 to 112. The percentage of polyunsaturated omega- 6 FA's (PUFA-6) in lipids from steers grazed on FOC pasture followed by a com diet increased significantly from 0 to 112 DOF, but did not change significantly in lipids from steers grazed on Rye pasture and then fed a com diet. Lecithin content of ground beef (g / 100 g beef) was significantly affected by DOF. The FA types: SAT, PUFA-3 and unknown were negatively correlated with BFF and MONO types were positively correlated (P<.05). The relationship of these FA types to milky-oily flavor was reversed from that of BFF.

Finally, the concentration of 4 variables: 18:0, 18:2ω6, pentanal and unknown 7 volatile explained 72.95% (R²) of the variance in the BFF score of cooked ground beef from 22 steers grazed on pasture A or B and then fed a com diet 0, 56 or 112 days. BFF for ground beef samples from 11 steers treated in a similar manner as the 22 steers used to develop this model was also calculated and compared with their actual BFF. BFF predicted from the 4-term model had a correlation coefficient of .83 (P<.05) with the actual BFF of the 11 samples.