In his book, The Extended Phenotype, Dawkins discusses his take on three distinct types of EP. The first is what we all think of when we hear EP: the ability of an animal to construct architectural forms from its environment to aid in its overall fitness. An example might be the caddis house, which is constructed by a silk produced in the caddis' salivary gland. The silk has nothing to do with the phenotype of the caddis that might cause a mate to choose it, but the way the silk is used will ultimately play a deciding role.The ability of the caddis to create a tighter wrapped silk house or a more hidden placement of the silk house are all advantages that can cause increased fitness against predators and increase the likelihood of finding a mate.
The second type is what he called "parasite manipulation," meaning that parasites can manipulate the behaviors of their hosts to create a distinct advantage in their overall fitness. A good example is the zombie like behavior of ants caused by the infection of a type of fungus known as ophiocordyceps unilateralis. Once the fungus infects the ant it causes a chemical imbalance in the brain assuming complete control. The ant will first be forced to drop from the canopy and find specific conditions where it will clamp down on the underside of a leaf vain, known as the death grip, in conditions that are moist, humid, and relatively warm (perfect conditions for fungal growth). The mechanism by which the fungus is able to control the ants' behavior and to the extent that it does is unknown. Once the fungus has absorbed all the nutrients within the ant it will sprout a stalk from the ants head and disperse its spores. Still other types of fungus cause some insects to climb to higher parts of leaves and assume a sexual position so that when a mate arrives the mate will be infected as well, even though its sexual partner is dead.
Here is a link to a video about this parasitic fungus!
The third is a little more complicated to understand. It encompasses the idea that the behavioral phenotype of the host can be manipulated by the parasite even if it is not physically associated with the host. The example given by Dawkins are the chicks of cuckoo birds. Some species within this family of birds are considered brood parasites, meaning they lay their young in the nests of other birds; whether it is the same species or not. Some cuckoo birds have even developed a selective advantage of closely replicating the appearance of the host bird's eggs that they are "trying to fool." The cuckoo chicks are thought to have genetically acquired the instinct of throwing out the chick eggs of the host bird, as they will hatch much more quickly than the host's own chicks. The chicks will then begin a rapid begging call or continually have their mouth opened, both signal stimuli that they need feeding. This is the part that I think Dawkin is referring to when he says the parasite is altering the behavior of the host. Or it might even include the whole situation with the parasite being the mother cuckoo which lays the eggs, then becomes completely dissociated with the host. The fitness is increased for the cuckoo because the "new mother" is now forced to take care of the chicks because she thinks they are her own especially since her own were thrown out of the nest never to be born. Therefore, the host's behavioral phenotype is changed by the continual searching of food to feed "her" young.
Articles, Diagrams, and Videos