GENETIC AND ENVIRONMENTAL CORRELATIONS OF MORPHOMETRIC TRAITS IN RANDOMBRED HOUSE MICE.

Abstract

The phenotypic correlation coefficient (rp) is regularly employed as a measure of association between metric traits in a variety of morphogenic studies. Multivariate techniques such as component analysis and factor analysis (Harman, 1967), extensively used with mammalian dental and/or skeletal systems (Jolicoeur and Mosimann, 1960; Wallace and Bader, 1967; Gould and Garwood, 1969; Riddle, 1971; Leamy, 1975), typically start with a matrix of such correlations. However, the phenotypic correlation between two traits is influenced by both genetic and environmental sources of variation in much the same way as is the total phenotypic variation in a single trait. Therefore, most such studies usually attempt to interpret factors, components, or forces which necessarily confound these sources. An alternative approach, only rarely used (Bailey, 1956; Yap Potter et al., 1968), involves performing the appropriate multivariate technique on separate genetic and environmental correlations rather than on phenotypic correlations only. The genetic correlation (rA) between two traits may be obtained in a conventional quantitative genetic analysis from the association of the two traits in related individuals. It is generated by the covariation of additive genetic values for the two traits, and if significantly different from 0, is an indicator of pleiotropy and/or linkage (Falconer, 1960). Genetic correlations are also of interest because they determine, in part, the amount of change which can be brought about in one character by artificial or natural selection for the other (correlated) character. Environmental sources of covariation are both many and varied, and may result ih an environmental correlation (rE) quite different from the genetic correlation in magnitude, and sometimes even in sign (see Falconer, 1960 for examples). A recently completed experiment provided data suitable for a quantitative genetic analysis of 15 osteometric and 3 external metric traits in randombred house mice (Leamy, 1974), and for calculation of genetic and environmental correlations. The purpose of this paper is the presentation and analysis (by multivariate clustering and ordination techniques) of these correlations. The patterns of association are also compared to those previously found from a principal component analysis of the phenotypic correlations among these same traits (Leamy, 1975).