MORPHOLOGICAL AND PHYLOGENETIC RELATIONS AMONG THE DARWIN'S FINCHES.

Abstract

Despite the ecological and evolutionary significance of the Darwin's finches (Lack, 1947; Bowman, 1961; Abbott et al., 1977; Boag and Grant, 1981; Schluter and Grant, 1984), their phylogenetic relationships are incompletely understood. Yang and Patton's (1981) electrophoretic study on 11 of the 13 Galapagos species supported Lack's (1947) division of the finches into three major groups (Fig. 1): ground finches (Geospiza), tree finches (Camarhynchus and Platyspiza), and the warbler finch (Certhidea olivacea). Biochemical differences within these groups were slight, however, leaving details of species relationships largely unresolved. In this report I use a new method to estimate evolutionary relationships among the finches based on morphological data. The analysis includes all 13 Galapagos species and Pinaroloxias inornata from Cocos Island, putatively a warbler finch (Lack, 1947). Ecological studies have indicated that morphology in the Darwin's finches has been strongly influenced by natural selection (Boag and Grant, 1981; Price and Grant, 1984; Schluter and Grant, 1984). Under these conditions morphological differences between species might be expected to reflect differences in environmental history rather than evolutionary relationships (Barrowclough, 1983). Indeed, poor concordance between estimates of phylogeny based on morphological data and estimates based on biochemical data (ostensibly neutral) has been noted in other groups (e.g., Zink, 1982) as well as in the Darwin's finches (Barrowclough, 1983). However, most standard measures of morphological distance (e.g., Cherry et al., 1982) lack explicit evolutionary justification, and this may partly explain poor concordance when it is observed. Here I employ a measure of morphological distance based on a theoretical study of multivariate evolution (Lande, 1979). The resulting evolutionary relationships inferred agree reasonably well with both the biochemical (Yang and Patton, 1981) and the traditional estimates (Lack, 1947), where comparison is possible.