Maintaining Genetic Variation in Fitness

Abstract

Although natural selection is expected to remove additive genetic variation in fitness, the recent evidence is that in natural populations considerable amounts of such variation remain in fitness and related traits, just as for other quantitative traits. In assessing this it is important to distinguish between heritability which may be low and the genetic coefficient of variation which may be high. Here we concentrate mainly in the artificial selection context, where selection criteria are known and experiments have been conducted, and review rather than present new results. In an artificial selection programme, the selection index can be regarded as the measure of fitness. In this case, analyses of rates of loss of variation can be used that have been derived for quantitative traits in general. These indicate that effective population size is a critical factor, in particular when mutation to useful variation is taken into account. In many long term selection experiments and breeding programmes variation has been maintained for very many generations, without a plateau in response. If fitness is regarded as the set of health, viability and reproductive traits not directly related to the selection objective, for example in a breeding programme for meat animals, then maintenance of variation in fitness is more highly dependent on effective population size than is variation in the directly selected traits. Examples of improvement programmes show that, while such measures of fitness have declined in the past, there has remained genetic variation such that subsequent selection to reduce or reverse the trends in some fitness associated traits has been effective.