SELECTION INTENSITY IN NORTH AMERICAN HOUSE SPARROWS (PASSER DOMESTICUS).

Abstract

House sparrows (Passer domesticus) were first introduced at New York City in 1852 and have since colonized most of North America. Present North American populations are mostly descendent from birds from central England (Barrows, 1889). As house sparrows dispersed, populations responded to different selective regimes determined by local conditions. In little more than 100 generations since its introduction, Johnston and Selander ( 1971 ) showed that development of morphological variation has approached the amount seen in this species for the whole of Europe. They related geographic trends in morphological variation to environmental factors, and their interpretations of trends in body size and proportions for North American populations are generally consistent with the ecogeographic rules of Bergmann and Allen. Selection pressures on sparrow populations must vary considerably throughout the continent. A method of measuring such selection intensity is provided by O'Donald (1968a, b, 1970a, 1972, 1973) that uses the proportionate change in fitness, Ar7/rw, as a measure of selection intensity. The model assumed for fitness is a quadratic fitness function, w = 1 a K( -)2. As a character-state, x, deviates from an optimum phenotype, 0, fitness of the individual decreases. Actual fitness is given if the constant a is known, or if a = 0, then fitnesses are relative; K is a constant. From this model, O'Donald (1970b) shows that the change in mean fitness is A?v = Vwl/r, where w7v is mean fitness and Vw is variance in fitness. If the character x is normally distributed, then Afvlfv/ = (x-'C)2/ V + 1/2{ [(X-)2 + (V $-V )V]/ V,}2, where x and V$ are the initial mean and variance of the character and x' and V'$ are these values after some period of selection (O'Donald, 1973). Death can occur for any member of a population through predation, disease, environmental stresses, starvation or accidents. Certain elements of populations may be more susceptible to mortality and in such cases selective deaths will occur. Since Johnston and Selander (1971) have shown that patterns of morphological and climatic variation are highly congruent, suggesting a cause-effect relationship, I felt that any geographic variation in selection on house sparrows was caused primarily by environmental mortality factors. Such influences continuously affect species' populations, and I initially investigated selection in house sparrows with the view of determining relationships between selection intensities and climatic factors. Additionally the questions of when selection initially operated in adapting North American sparrows and to what extent selection still operates on house sparrows were examined.