View Resource

Return to search results   |   Browse resources
The "Niche-Variation" Hypothesis: A Test and Alternatives

abstract1 (full description below): The American Naturalist, Vol. 104, No. 935. (Jan. - Feb., 1970), pp. 85-97.
The "Niche-Variation" Hypothesis. PDF/Acrobat file     1970. Michael Soule; Barbara Rindge Stewart

INTRODUCTION Often a theory gains acceptance almost for no other reason than its inherent plausibility, the case in point being what has been called the "niche variation model" (Van Valen 1965). This is the hypothesis that certain inds of genetic variation and variation of morphological characters, such as bill size in birds, are correlated with and to some extent determined by the variety of foods and habitats used by a population. In its broadest form the hypothesis seeks to explain why some bird species on some islands have greater bill-size variation than on continents (Van Valen 1965; Grant 1967), and why some "central" populations of Drosophila are more polymorphic for chromosomal inversions than some peripheral populations (da Cunha and Dobzhansky 1954; da Cunha et al. 1959). In such discussions a term like "niche width" is usually employed, and it is argued that populations in wider niches are more variable than populations in narrower niches, and that the greater morphological variation of the former type of population is the evolutionary expression of the increased fitness of the extreme individuals. For example, the smallest and largest individuals in a population of birds colonizing a species-poor island might choose food items which are relatively small and relatively large, respectively. In the absence of competing species and in the presence of a uniform distribution of food item size, these extreme individuals would prosper compared to the more abundant typical individuals. The resulting reproductive advantage of the extreme individuals will lead to an evolutionary increase in the variance of bill size. The argument so stated seenis reasonable, other things being equal. Essentially the same argument is often couched in less specific terms, namely that there will be "less selection" (a relaxation of stabilizing selection) in a relatively commodious environment. When phrased in this more neutral fashion, the argument is, if anything, less controversial, but it still assumes a relative increase in the reproductive success of extreme individuals. On the other hand, if diet-item diversity and variation are uncorrelated, a model invoking intrapopulation feeding specialization in "euryphagous" species is unsupported. A good theory should be testable and should generate verifiable predictions. For instance, one would expect to find a correlation of bill size and food-item size within a population under natural and experimental conditions. Besides such behavioral predictions there are also evolutionary tests. This is an attempt at the latter type of verification. I've reasoned as follows. If resource subdivision can be the basis for reduction of ultraspecific or intrapopulation competition, and if diversity in type or size of food item is such a resource variable, then such a subdivision should be most clearly expressed in animals with especially varied diets. More particularly, if bill-size differences within a species is a potential means of food apportionment, then putatively "euryphagous" species, like crows, should have more variable bills than putatively "stenophagic" birds like flycatchers.