'Ring species' occur when one species grades into two at the overlap of a circular population distribution. Good examples are rare, but one case has now passed some rigorous tests.
In principle, ring species should constitute “the perfect demonstration of speciation”1. According to the classic model, such rings are composed of populations of an organism that started from a single ancestral population and are now distributed geographically in a circular manner. Neighbouring populations in the ring can interbreed, except for the two terminal, overlapping populations which cannot and behave as if they are distinct species. The process is a culmination of a gradual adaptation of the organism to a gradient of habitats around the ring such that, when populations meet at the terminus, species-level divergence has occurred. An especially attractive feature of ring species is that they represent a piece of history, in that various stages in the formation of species are preserved and can be studied.
For various reasons there may be no such thing as the perfect ring species. But on page 333 of this issue, Irwin et al.2 describe how they have revisited one of the earliest putative examples — that of the greenish warbler (Phylloscopus trochiloides; Fig. 1), which is distributed around the margins of the Tibetan plateau in Asia and was first described over 60 years ago3. They conclude that, although this warbler ring is not perfect, it meets most of the requirements of the classic model.
Identifying ring species is complicated. First, the geographical conditions are critical: there needs to have been divergence from a single ancestral source population around some kind of barrier. An appropriate balance must be reached between the opposing forces of selection (which causes geographically neighbouring populations to diverge in their adaptations) and continuing gene flow through interbreeding (which maintains genetic continuity). Moreover, the ring of populations should be continuous, although it is almost impossible to satisfy this criterion. Ranges of species are subject to constant change — they expand and contract over time, with parts becoming separated then reconnecting to form secondary contacts. In other instances, populations become extinct, producing gaps. In the case of the greenish warbler, there is a gap in China, which seems to have been caused by human destruction of the bird's forest habitat.
A further complication is that the two terminal populations are usually identified first, and are often named as different species. Given this initial taxonomic bias, subsequent studies may overlook the possibility that a ring species exists. In other instances, possible examples have fallen prey to changing concepts of what a species is; several former ring species are no longer treated as a single species taxonomically, but rather as groups of closely related species (circumpolar gulls are a classic example).
One of the best-studied cases is the salamander Ensatina eschscholtzii , which has populations around the Central Valley of California with sharply divergent forms coexisting in southern California4. But whether or not this is a genuine ring species is a matter of controversy. There is one break in the distribution and another area where there is a 'weak link' in the chain of differentiated but intergrading forms. Also, there is debate over the most appropriate taxonomy because of different interpretations of genetic data5,6,7,8,9.
From Irwin and colleagues' work2, however, it seems that greenish warblers satisfy most of the stringent requirements for a ring species. One form ranges from the Baltic region to western Siberia, where it spreads south and intergrades with another form along the western and southern slopes of the Himalayas (see Fig. 1 of the paper on page 334). Intergradation with other forms occurs to the east, then northward, until an eastern Siberian form meets the western Siberian form in the vicinity of the Yenisei River. There the two behave as distinct biological species, and do not interbreed. For instance, as Irwin et al. show, they have more complex songs than southern populations, and no longer recognize each other as potential mates — this is sexual selection as a driving force in speciation. The birds are thought to have originated in the south and moved around the Himalayas, the eastern and western populations coming into contact in the north, but remaining distinct, following the end of the last glacial epoch. This fulfils the geographical and biological requirements of a ring species.
Irwin et al. also collected mitochondrial DNA evidence from the various populations. The DNA sequences form two well-supported clades, one eastern and the other western. In the region of the putative ancestral populations, the clades are intermixed. Typical phylogeographic analysis10,11 would interpret these clades as historical entities: that is, that they arose from eastern and western isolates that diverged when they were geographically isolated, and subsequently expanded their ranges both north and south to reconnect. In the north, songs diverged such that coexisting populations behave as different species, whereas in the south birds remained sufficiently similar in song characteristics to merge and form an interbreeding unit.
This interpretation is considered, but rejected, by Irwin et al.2. Instead, they have conducted simulations (not published) which show that antecedents of the two major clades of DNA sequences might have arisen within the same area. The sequences are not very divergent, and Irwin et al. argue that sorting of the two primordial lineages of sequences took place in the region in which they originated. One lineage came to dominate in the eastern region and the other in the western, as the populations expanded around the mountain mass. This interpretation runs counter to a central tenet of phylogeography, which sees history as having been recorded in the phylogeny of sequence lineages, and so is likely to be controversial.
Finally, the paper is notable for incorporating two different explanations for species formation. First, the authors take sexual selection to be responsible for the elaboration and divergence of songs in the greenish warbler. By contrast, they believe that the similarity in body size of the northern forms stems from natural selection acting on size. Both natural selection (especially in so-called ecological speciation)12,13 and sexual selection14,15 have been the subject of close recent interest. So the work of Irwin et al.2 makes a multifaceted contribution to the current debate on species formation.