Olfactory conditioning in a parasitic insect and its relation to the problem of host selection
It has often been stated that insect species which breed on two or more hosts tend to become split into biological races, each of which is attached to a particular food-plant or host as the case may be. But although the fact of the existence of biological races is well known, their mode of origin is, in most cases, in dispute. It is not known how far the biological characters of such races are germinally fixed, and it has been suggested that in many cases a group of individuals may become restricted to certain host species not by any germinal change, but by a kind of conditioning, as a result of which the adult female is attracted for oviposition to the particular species on which she had fed as a larva. This is the idea behind the so-called "Host Selection Principle" of Walsh, and it is a possibility which must be carefully considered when assessing the value of experiments attempting to establish the inheritance of acquired characters. It is obvious that if a conditioning of this sort was actually occurring, a polyphagous insect species might rapidly become split into two or more populations each attached to a given species of host (plant or animal as the case may be). The barrier thus created might persist for a very long time without any hereditary specialization taking place. It is known, however, that the characters which distinguish biological races are in fact often germinally fixed, and in some instances it has been shown that even though all structural differences are absent there may be repugnance for crossbreeding between two biological races of the same species (Thorpe 1930 b ). It is nevertheless often difficult to visualize the means by which such germinal differences could become established without the aid of the ecological separation which such host conditioning might provide.
