Taxon Cycles in the West Indian Avifauna

Abstract

Analysis of distribution patterns of land birds in the West Indies supports the empirical theory that once species invade islands they progress through a series of evolutionary changes that eventually greatly increase the probability of extinction. Species are arbitrarily assigned to four stages of the "taxon cycle" based on their distribution and differentiation within islands: (I) expanding or widespread, with little differentiation into morphologically distinct populations; (II) widespread but highly differentiated; (III) restricted, usually with a fragmented range; and (IV) endemic to one island. These stages are shown, beyond reasonable doubt, to represent a temporal sequence. Immigration has been observed to occur very rapidly, and expanding species usually reach most islands in a stepping-stone fashion. Recent colonists usually belong to species that are abundant and distributed in open habitats on the mainland. Once colonists reach islands, differentiation is usually rapid, although hampering effects of gene flow can be recognized. Changes in morphology and ecology also occur during the taxon cycle. In particular, habitat distributions are frequently restricted, often to mature forest. Extinction within recorded history has occurred with much higher frequency among Stage IV species than among species assigned to earlier stages. Slopes of species-area curves increase with stage of taxon cycle from 0.075 for Stage I species to 0.42 for Stage IV species, demonstrating that extinction rate is more sensitive to island size at progressively later stages of the cycle. Moreover, slopes vary with feeding behavior, from 0.04 for hummingbirds to 0.21 for flycatchers, swallows, and swifts. On the basis of these and other observations, we postulate that for a particular species the cycle results from progressively reduced competitive ability caused by counterevolution of the island biota toward that species. The cycle is driven by competition from newly arrived colonists that are initially free of a counteradaptive load. The presence of old populations on small remote islands rarely receiving new colonists attests to the importance of competition in the cycle. We expect groups with differing rates of immigration, differentiation, and extinction to exhibit different patterns of distribution among islands, but that suitable island groups, whose size and spatial patterns normalize these differences, can be found for any group of organisms. Moreover, the concept underlying the taxon cycle undoubtedly applies to mainland faunas and may help explain such patterns as variation in dominance and relative abundance of species.