If you need an accessible version of this item please contact JSTOR User Support

A Multivariate Statistical Analysis of Direct and Correlated Response to Selection in the Rat

William R. Atchley, J. J. Rutledge and David E. Cowley
Evolution
Vol. 36, No. 4 (Jul., 1982), pp. 677-698
DOI: 10.2307/2407882
Stable URL: http://www.jstor.org/stable/2407882
Page Count: 22
  • Download PDF
  • Cite this Item

You are not currently logged in.

Access your personal account or get JSTOR access through your library or other institution:

login

Log in to your personal account or through your institution.

If you need an accessible version of this item please contact JSTOR User Support
A Multivariate Statistical Analysis of Direct and Correlated Response to Selection in the Rat
Preview not available

Abstract

This paper demonstrates a pronounced asymmetry in the correlated response of a suite of 17 skull and jaw traits to directional selection for 3-9 week weight gain in the rat. These results parallel those for direct response to selection where the upselected replicates show about 2.6 times as much divergence as the down-selected ones in terms of average univariate divergence. When the effect of intercorrelated characters is taken into account the upselected rats have diverged about 1.5 times as much as the down-selected rats. The extent of phenotypic divergence between selection treatments, replicates of selection treatments and sexes within replicates is described. The narrow-sense heritability estimate of the canonical vectors suggest that the patterns of among-groups covariance are highly heritable in both sexes. Further, there is a general inverse trend between the heritability of a canonical vector and the magnitude of its eigenvalue. A quantitative genetic distance estimate equivalent to the Mahalanobis distance is computed to estimate the extent of genetic divergence between selection lines. The phenotypic and genetic distance estimates are highly correlated in both sexes; further, these results suggest that about two standard deviations of difference occur when there is no additive genetic divergence between the selection lines. A statistically significant Kluge-Kerfoot effect is shown to occur in females of these rats. However, no genetic association can be made with this effect in these data and it is suggested that the Kluge-Kerfoot effect in these data stems from variation in the environmentally labile portion of the phenotype rather than being genetically determined. Alternatively the Kluge-Kerfoot phenomenon may be a statistical artifact.

Page Thumbnails

  • Thumbnail: Page 
677
    677
  • Thumbnail: Page 
678
    678
  • Thumbnail: Page 
679
    679
  • Thumbnail: Page 
680
    680
  • Thumbnail: Page 
681
    681
  • Thumbnail: Page 
682
    682
  • Thumbnail: Page 
683
    683
  • Thumbnail: Page 
684
    684
  • Thumbnail: Page 
685
    685
  • Thumbnail: Page 
686
    686
  • Thumbnail: Page 
687
    687
  • Thumbnail: Page 
688
    688
  • Thumbnail: Page 
689
    689
  • Thumbnail: Page 
690
    690
  • Thumbnail: Page 
691
    691
  • Thumbnail: Page 
692
    692
  • Thumbnail: Page 
693
    693
  • Thumbnail: Page 
694
    694
  • Thumbnail: Page 
695
    695
  • Thumbnail: Page 
696
    696
  • Thumbnail: Page 
697
    697
  • Thumbnail: Page 
698
    698