The Evolutionary Psychology FAQ

Edward H. Hagen, Institute for Theoretical Biology, Berlin

Why is the heritability of adaptations generally zero?

One often reads in the paper that researchers have discovered the gene for homosexuality or shyness, or, more generally, that some trait is 'heritable'. Heritability means that variation in some trait (like the presence or absence of homosexuality) is correlated with genetic variation (e.g., the presence or absence of some gene). It is a common mistake to assume that if a trait is heritable, it is adaptive. The opposite is generally the case!

Adaptations evolve by the differential reproduction of alternative alleles. Novel alleles arise by mutation. If the phenotypes associated with these novel alleles interact with aspects of the environment in such a way as to gain a reproductive advantage over phenotypes associated with other alleles at the same locus, the novel allele will (with some exceptions discussed below) go to fixation--that is, its frequency in the population will be 100%. Heritability is the proportion of variance in a phenotypic trait that is accounted for by genetic variance. As we have just seen, however, genes that confer a reproductive advantage generally go to fixation. Because their frequency in the population is 100%, the genetic variance at the loci of these genes is zero, so any variance in the corresponding phenotypic traits cannot be attributed to the non-existent genetic variance. Even though such traits are genetically specified, their heritability is zero! Everyone has the same genes.

An obvious exception would occur if the population were sampled before an allele went to fixation. If half the population had an allele, and having the allele meant having the trait, then the heritability of the trait at that moment in time would be unity; most genes under positive selection, however, go to fixation relatively quickly. Another exception occurs when there is strong selection for a trait in one geographical area, but no selection, or even counter-selection, in other regions--the sickle-cell variant of the beta globin gene is a classic example. Because heterozygotes have a reproductive advantage in malaria-infested regions, the sickle-cell variant is maintained at high frequencies in those regions; because homozygotes are reproductively disadvantaged, the trait, prior to recent migrations, was virtually absent elsewhere. These dynamics prevent the sickle-cell allele from going to fixation, so the heritability of sickle-cell is high.

The sickle-cell allele is an example of an adaptive genetic polymorphism: an allele that is maintained at a much higher frequency in the population than can be accounted for by mutation and drift, yet has not gone to fixation (and therefore coexists with other alleles at the same locus--thus the term 'genetic polymorphism'). Population or geographically specific selection pressures are one avenue for the maintenance of adaptive genetic polymorphisms; frequency dependent selection is another. An allele might be able to increase in frequency when it is rare simple because individuals possessing the gene are 'different'--this might happen if, say, members of the opposite sex preferred 'difference'. As its frequency increases, however, individuals possessing the allele are no longer 'different', so they are no longer as preferred as mates, and the allele will not go to fixation. Instead, it will exist in equilibrium with other alleles at the same locus. This type of frequency dependent selection (so called because the selection pressure on an allele depends on the frequency of the allele in the population) can result in adaptive genetic polymorphisms.

Most adaptations of interest, however, are not like sickle-cell--they are not coded for by single genes, they are not population or geographically specific, nor are they frequency dependent. Adaptations like hearts and lungs are coded for by an enormous number of genes, and are universal in the human species. We will refer to adaptations coded for by single (or relatively few) genes as simple, and those coded for by many genes as complex. Complex adaptations must be universal, and their heritability (construed as the presence or absence of the adaptation) must be essentially zero. Complex adaptations evolve 'piece-by-piece'. The alleles coding for fundamental elements of what could eventually become a complex adaptation must arise before there will be a selection pressure favoring alleles that code for additional 'components' of the adaptation. An organism must first evolve light-sensitive tissue before there will be a selection pressure to evolve a lens, for example. Because complex adaptations are coded for by many genes, because the genes that code for the rudimentary functionality must evolve before there was selection pressure for the evolution of advanced functionality, and because genes under positive selection go to fixation relatively rapidly, almost all complex adaptations (and the genes that code for them) must be universal. Further, the probability that any individual would lack all (or even many) of the genes needed to code for a complex adaptation is essentially zero, so the heritability of the presence or absence of a complex adaptation is also zero. There is overwhelming physiological evidence for these claims. Although there are some simple adaptive physiological differences between populations (skin color, for example), the vast majority of tissues and organs are functionally identical in all humans. (If a complex organ was evolving in one population but not others, that population would almost certainly become reproductively isolated, branching into a new species).

Aside from the simple, geographically specific adaptations noted above, or genetic diseases (which also consist of one or a few genes), what kinds of traits are heritable? In general, traits that are not under selection can have substantial heritable components, because if they were under positive or negative selection, the genes underlying those traits would go to fixation or be eliminated (respectively), eliminating the heritability; it is this fact that implies that if the trait in question is highly heritable, then it is probably not an adaptation. This is not to say, however, that there no aspect of a complex adaptation can be heritable. Although vision is not heritable, if a particular visual performance parameter were measured precisely enough, a small degree of heritability in that parameter could no doubt be detected. These minor differences would usually be genetic noise, however, not adaptive differences.

Copyright 1999-2002 Edward H. Hagen