The Extended Phenotype

The Extended Phenotype is a 1982 book by the evolutionary biologist Richard Dawkins, in which the author introduced a biological concept of the same name. The book's main idea is that phenotype should not be limited to biological processes such as protein biosynthesis or tissue growth, but extended to include all effects that a gene has on its environment, inside or outside the body of the individual organism.

Dawkins considers The Extended Phenotype to be a sequel to The Selfish Gene (1976) aimed at professional biologists, and as his principal contribution to evolutionary theory. In the 1999 reissue and subsequent reprintings an afterword by Daniel Dennett is included.

The central thesis of The Extended Phenotype, and of its predecessor by the same author, The Selfish Gene, is that individual organisms are not the true units of natural selection. Instead, the gene — or the 'active, germ-line replicator' — is the unit upon which the forces of evolutionary selection and adaptation act. It is genes that succeed or fail in evolution, meaning that they either succeed or fail in replicating themselves across multiple generations.

These replicators are not subject to natural selection directly, but indirectly through their "phenotypical effects". These effects are all the effects that the gene (or replicator) has on the world at large, not just in the body of the organism in which it is contained. In taking as its starting point the gene as the unit of selection, The Extended Phenotype is a direct extension of Dawkins' first book, The Selfish Gene.

Dawkins argues that the only thing that genes control directly is the synthesis of proteins; restricting the idea of the phenotype to apply only to the phenotypic expression of an organism's genes in its own body is an arbitrary limitation that ignores the effect a gene may have on an organism's environment through that organism's behaviour.

Dawkins proposes there are three forms of extended phenotype. The first is the capacity of animals to modify their environment using architectural constructions, for which Dawkins provides as examples caddis houses and beaver dams.

The second form is manipulation of other organisms: The morphology of a living organism, and possibly of that organism's behaviour, may influence not just the fitness of the organism itself, but that of other living organisms as well. One example of this is parasite manipulation. This refers to the capacity, found in some parasite-host interactions, for the parasite to modify the behaviour of the host in a way that enhances the parasite's own fitness. One well-known example of this second type of extended phenotype is the suicidal drowning of crickets infected by hairworm, a behaviour that is essential to the parasite's reproductive cycle. Another example is seen in female mosquitoes carrying malaria parasites. The mosquitoes infected with the parasites whose preferred hosts are humans have been shown in a field experiment to be significantly more attracted to human breath and odours than uninfected mosquitoes when the parasites are at a point in their life cycle where they can infect a human target.

The third form of extended phenotype is action at a distance of the parasite on its host. A common example is the manipulation of host behaviour by cuckoo chicks, which elicit intensive feeding by the host birds. Here the cuckoo does not interact directly with the host (which could be meadow pipits, dunnocks or reed warblers). The relevant adaptation lies in the cuckoo producing eggs and chicks that resemble sufficiently those of the host species so that they are not immediately ejected from the nest. These behavioural modifications are not physically associated with individuals of the host species but influence the expression of its behavioural phenotype.