Heritabilities (ĥ2) and genetic correlations (rG) were estimated by regression of offspring on sire in two replicate, unselected lines of mice. Traits were associated with growth, feed efficiency, fat deposition and lean tissue. The ĥ2 for growth traits ranged from .34 to .42, except for 3-wk body weight, which was only .05. The ĥ2 for feed efficiency was .28. Ranges in ĥ2 were .45 to .50 for fat deposition traits and .36 to .42 for lean tissue traits. The rG involving 3-wk to 6-wk feed efficiency with hind carcass and fat measurements at 12 wk were small. Antagonisms were found between the sign of rG and the direction of usual breeding goals for pairs of traits (e.g., rG greater than 0 between fat deposition and hind carcass weight and rG less than 0 between hind carcass as a percentage of body weight and body weight). Selection indexes were developed to counteract these antagonisms. Modified selection indexes were compared where responses in individual traits rather than the aggregate breeding value were of major importance. The aggregate breeding values and selection indexes included: 1) epididymal fat pad weight and body weight, 2) hind carcass weight and body weight, or 3) all three traits. Economic weights in retrospect were calculated for the modified selection indexes. In some cases, expected correlated responses in component traits were not influenced greatly over a wide range of ratios of economic weights, but in other cases the component traits changed sharply over a narrow range of ratios.

Development of adipose tissue in five depots was investigated in mice selected for high or low 12-week epididymal fat pad weight as a percentage of body weight (HF and LF lines), or high or low 12-week hind carcass weight as a percentage of body weight (HL and LL lines). An unselected control line (RC) was maintained. Hind carcass (HC) and fat pads from subcutaneous hindlimb, subcutaneous forelimb, gonads, kidneys and mesentery were dissected and weighed at 4, 6, 9, 12 or 15 weeks of age. Generally, body weight (BW), daily gain (DG), feed intake (FI), feed efficiency (FE) and feed intake/metabolic body weight (FC) were higher (P≤0.05) in HF than in LF, and in LL than in HL. HF had more fat (as a percentage of BW) than LF in all depots (P≲-0.01), and asymmetry (P≤0.01) was detected for gonadal fat. LL consistently had a higher (P≤0.01) fat percentage than HL, and asymmetry (P≤0.01) was observed for perirenal fat. At age of selection, ranking of fat depot weights as a percentage of total fat depot weight was not changed by selection; however, gonadal fat accounted for more of the total fat in HF and LL compared with RC, while the opposite was found in LF and HL. HC percentage was higher (P≤0.01) in HL than LL, and higher (P≤0.01) in LF than HF. Growth rate of each fat depot relative to BW was not affected by selection. These results demonstrated that selection for proportion of fat in one depot or for HC percentage changed fat percentage in other depots. However, the rate of fat deposition in each depot relative to body weight gain was not altered.

The correlated response to 13 generations of selection for percent fatness and leanness is investigated in 11 mandible traits in mice. Five selection lines are examined including high fat (HF), low fat (LF), high lean (HL), low lean (LL) and a randomly selected control strain (RC. The ontogenetic patterns of growth in the RC strain serve as a model to evaluate the developmental consequences of directional selection. Selection has systematically altered the patterns of mandible growth in selection lines relative to the control strain. Further, selection has significantly altered the age-specific phenotypic covariance among these traits. In the HF strain, growth in the mandible is completed by 12 weeks of age for most traits. In other selected strains, notably LF and LL, there is a significant growth spurt that occurs between 12 and 15 weeks of age. Changes in the patterns of mandibular growth produce significant differences among strains in the final form of the mandible. Because of the changes in the patterns of growth, the differences among strains are themselves shown to vary at different postnatal ages. The phenotypically similar strains, i.e., HF and LL or LF and HL, show different but correlated patterns of divergence. Multivariate statistical analyses suggest that the temporal strain differences in these traits are multidimensional.

No abstract

No abstract