Inbreeding, Heterosis and Information Theory

Abstract

Measurements of the ratio of hatched eggs to total in randomly breeding and inbred populations of Drosophila melanogaster show that the two differ strikingly. In the former the ratio is high and its variance low; in the latter the situation is reversed. The inbred embryo has a low and variable probability of developmental success even in a very constant environment. Recent attempts to apply information theory and cybernetics to biological phenomena suggest a new way of looking at these facts employing the concepts of redundancy, noise and equivocation with respect of developmental efficiency. Biological information is passed from one generation to another coded in the genetic material. It is generally agreed that DNA plays a major role in this coding and there is very good evidence that very slight changes in nucleotide sequences can produce phenotypic change. But there is also excellent evidence that the number of possible permutations of the DNA is far, far greater than the minimum necessary for coding the genetic information and that actually the information must be coded in numerous different ways with different degrees of redundancy. It is suggested that the low and variable hatching ratio of the inbred population is the result of a lowering of redundancy in a genome where more loci are occupied by identical alleles. The data concerning hatching ratio exactly parallel those for heterotic characters generally and it appears likely that redundancy in the genetic code may be a general explanation of heterosis more satisfactory than either overdominance or the covering of deleterious recessives.