ISOLATION, ENDEMISM, AND MULTIPLICATION OF SPECIES IN THE DARWIN FINCHES

Abstract

In this paper we use multiple-regression analysis to consider the general evolutionary problem of the relative importance of the various factors governing the proliferation of related subspecies or species on member islands of an archipelago. From this quantitative analysis of the relation of isolation to endemism and sympatry' in the Darwin finches, deductions are made concerning (i) the location within the Galapagos Archipelago of insular sites of origin of the species and (ii) whether or. not these sites are also the ones where the remarkable, external morphological divergences characteristic of the group occurred. Together these considerations constitute a reinvestigation of factors underlying the processes of species multiplication and differentiation within an archipelago, and it is hoped that this will encourage students of other groups of organisms to make comparable comparisons and determine if our generalizations are universally applicable. For the Darwin finches (Geospizinae), which have undergone a moderate, adaptive radiation2 within the strongly isolated complex of the Galapagos Islands, we have found it useful to ask the question: Why, in this bird group of restricted distribution, is there a tendency for many species and few endemic subspecies to occur on some islands, and for few species but more endemics to occur on other islands (see fig. 1)? Explanation of the negative relation (correlation coefficient or r =-0.60) between insular numbers of species and insular numbers of endemic subspecies relates to the evolutionary problem cited. Under various guises, the problem has been taken up by others, with Ridgeway (1896), Swarth (1931, 1934), Lack (1947), and Bowman (1961) being major references to the topic. We seek in the present study of the Darwin finches an evaluation that is more statistically oriented than is usual for such problems (cf., however, Kramer and Mertens, 1938). Our determination of partialregression coefficients by the method of least squares involves estimation of the partial influences on total insular numbers of either species or endemic subspecies of the following four factors: insular area; insular numbers of land plant species; insular isolation as measured by distance from nearest island; and insular isolation as measured by distance from Indefatigable Island located in the central archipelago area.