PREDATOR POLYMORPHISM AND APOSTATIC SELECTION.

Abstract

The phenomenon of polymorphism is widespread in living organisms, but its adaptive significance remains obscure in most cases. Both genetic and environmental factors involved in the evolution of polymorphism have been studied, and in these studies it has been shown that a variety of factors influence both the extent and the type of polymorphism. Plumage polymorphism, independent of sex and age, in birds has been the subject of numerous studies, most extensively reviewed by Mayr (1942) and Huxley (1955). Huxley recorded polymorphism in 21-25 of 134 bird families; in only seven was it considered frequent. I have not compiled an exhaustive list, but it is clear that the phenomenon is really frequent in only a few families: Procellariidae, Ardeidae, Accipitridae, Falconidae, Stercorariidae and Strigidae. When present in other families it is sporadic, often occurring in only a single or a few species of a large group. Examples are the ruff (Philomachus pugnax), a strikingly polymorphic species in the otherwise monomorphic Scolopacidae; the bush-shrikes (Malaconotus) of the Laniidae; and one goose (Anser caerulescens) of the Anatidae. Polymorphisms in birds vary in extent from those that can be observed only at close range (as the ring on Uria aalge) through those that are minor but conspicuous (as the head color in Poephila gouldiae) through those that involve the entire ventral coloration (as in many birds of prey) to those that encompass the greater part or all of the plumage (as in some herons). Genetic polymorphism in any form has been considered adaptive in providing an increased diversity of genotypes, leading to reduced intraspecific competition and increased buffering against environmental changes (Mayr, 1963: 252). Thus there should be selection for variants, especially in situations in which the pressure from interspecific competition may be reduced, as on islands. Yet there is no reason for variation in external appearance to increase, and indeed there should be selection against such variation in species that recognize each other visually, as do birds. An alternate possibility is that variation in appearance may be used for individual recognition and facilititates social interactions, in which case it would be adaptive. But here a gradual range of variation would seem to be more advantageous than two or several sharply distinct phenotypes. One hypothesis for the adaptiveness of external polymorphism that has received attention recently is that of apostatic selection, or selection for variation for its own sake (Clarke, 1962a). In apostatic selection, a given phenotype is favored in direct proportion to its rarity through frequency-dependent predator pressure. Rand (1967) has suggested that this mechanism can be invoked to explain the great, and to our eyes often bizarre, diversity of tropical insects, i.e., that there has been selection for phenotypes that look as different as possible from the other phenotypes present in an area. Tinbergen (1960) and Mook, Mook and Heikens (1960) presented evidence that species of insects were taken by titmice (Parus spp.) less often than expected (by their frequency in the environment) when at low density and more often than expected when at higher densities. They interpreted this to mean that when a given prey species was more common, the birds encountered it with sufficient frequency to be able to form a search image of the prey, thus increasing their proficiency at finding it.