The Discrimination of Interactions and Linkage in Continuous Variation

Abstract

Formulations including additive, dominance, linkage and interaction effects have been obtained for first and second degree statistics deriving from a backcross of an F2 family to its F1 and parents. Effects of genic interaction appear in the mean values of these backcrosses both in the presence and absence of linkage except in the case of F2XF1 which is unaffected by such effects where linkage is absent. A scaling test based on the means is capable of detecting pure i (=fixable)-interaction.[long dash]In the second degree statistics obtained from F2XF1 the effects of interaction appear in separate quadratic terms. In the case of the F2xP''s these effects are partly confounded with the additive and dominance components.[long dash]Compared with other generations the variances and covariances derived from these backcrosses require only one further definition of D and H.[long dash]A sequential analysis enables the discrimination of the effects of linkage and interaction under conditions where a correlation has been demonstrated between these effects.[long dash]An experiment in Nicotiana rustica has been analysed. In the character "height" there is no suggestion of heritable effects other than additivity and dominance. In "flowering time" the presence of non-allelic interactions is strongly suggested.