The Effect of Homozygosity on Developmental Stability.

LTHOUGH close inbreeding does divide a population up into lines, members A of which are in general very similar-wRIGHT claims there was never any difficulty in telling to which of his 23 lines of inbred guinea pigs any individual guinea pig belonged-it has often been observed that some characters in some lines become more variable with inbreeding. The characters for which this has been found to be true are those usually called quantitative. It is common to find that the phenotypic variance of a quantitative character increases with extreme inbreeding (EMMENS 1939; REEVE and ROBERTSON 1953; MCLAREN and MICHIE 1954; BIGGERS and CLARINGBOLD 1954). Since close inbreeding reduces genotypic variance, except as described by ROBERTSON (1952) in the early generations which are not under consideration here, one must suppose that homozygosis has actually increased phenotypic variance by increasing nongenetic variance. This seems to be confirmed by the finding of BIGGERS and CLARINGBOLD (1954) that response to estrogen treatment was most variable in inbred lines, less so in random-bred lines but least variable in F, crosses between inbred lines. Members of such F, crosses would be genotypically identical but less homozygous even than the random-bred group. The experimental findings leave no doubt that inbreeding, even to the point of homozygosis at virtually all loci, can result in an increase in total variance which must result from an increase in nongenetic variance. The reason why a highly inbred individual should be more susceptible to nongenetic changes than an outbred one is not immediately obvious. It is well known that mutants are more variable than their corresponding wild-type genotypes and it may be that inbreeding fixes recessive genes at some loci in every line, and that these loci are responsible for increases in variability in the characters they affect. It must not be assumed that such an increase in variability is a characteristic of the homozygous recessive locus itself. WRIGHT (1934a, b) found abnormalities such as extra toes and otocephaly appearing in some of his lines of inbred guinea pigs but not in others. The frequency of occurrence of these abnormalities was different in the lines in which it occurred; extra toes for example occurred in lines at frequencies of 0.3, 1.7, 2.0, 7.7, 13.5, and 19.0 percent. The average seventy of the abnormality was directly related to its frequency in a line. In this example we have two types of variation exhibited, that which distinguishes normal from abnormal individuals and that which renders some abnormals more extreme than others. Detailed work of GOLDSCHMIDT (1937) , GOLDSCHMIDT and HONER (1937) and others, leaves little doubt that in cases of this sort it is not any par-