nesses, and rare homozygotes have low fitnesses. Because The data points in these plots approximate straight lines, the allelic frequencies are close to the equilibria described by the fitness differentials. These plots are consistent with selection at these loci.
The adaptive distance plots for these polymorphisms reveal both the utility and the shortcomings of condensing data into an axis of individual heterozygosity. On the positive side, pooling homozygous genotypes yields a class of individuals with fitnesses lower than those of heterzygotes, and this has been helpful for detecting fitness differentials in multilocus data (chapter 7). These plots also reveal how much information is lost by pooling the common and rare homozygotes. Consequently, the variation in fitness explained by an adaptive distance plot can substantially exceed the variation explained when genetic data are reduced to individual heterozygosity ( Bush and Smouse 1992; Bush, Smouse, and Ledig 1987; Mitton 1989, 1993a; Smouse 1986). For example, applying the adaptive distance model to data on heterozygosity and growth in pitch pine ( Ledig, Guries, and Bonefield 1983) more than doubled the proportion of variance explained ( Bush et al. 1987).
Evolutionary biologists are still groping with the problem of fitness determination. Genetic variation is abundant in natural populations, but there is no consensus concerning whether fitness variation is determined by a few or many loci.
It was once thought that segregational load would severely limit the number of loci contributing to variation in fitness, but segregational load may have been more of a burden to evolutionary biologists than it has been in natural populations. Although some models of fitness determination generate sufficient genetic load to severely limit the number of loci contributing to variation in fitness, other models, such as truncation selection, allow all the polymorphic loci in a population to be balanced by selection.
Theoretical studies suggest that for loci whose variation is balanced by selection, fitness generally increases with heterozygosity. This expectation is consistent with many empirical studies, summarized in the next chapter. But even though fitness may generally increase with heterozygosity, a ranking of genotypes by their numbers of heterozygous loci lumps common and rare homozygous genotypes into the same class and may not adequately describe fitness differentials in natural populations.