In a previous post I touched briefly on two species of New Zealand rail in the genus Porphyrio. As mentioned, the pūkeko is one subspecies (Porphyrio porphyrio melanotus) that belongs to a cosmopolitan species complex. The takahē (Porphyrio hochsetterii), on the other hand, is endemic to New Zealand. In addition to these two extant species there is an extinct member of the genus, the mōho (Porphyrio mantelli), also known as the North Island takahē. The pūkeko and takahē are morphologically similar as would the mōho have been, which is what you expect given that they are congeneric (i.e. in the same genus). From fossil evidence we know that the mōho and the takehe were more similar to each other than to the pūkeko: the main difference being that the mōho had longer legs, was larger and was more slender than the takahē. The strong morphological similarity between the mōho and takahē led early taxonomists to classify them as two subspecies, Porphyrio mantelli mantelli and Porphyrio mantelli hochsetterii respectively. Because all three species are in the same genus, we know that they share a common ancestor. Therefore, pragmatically we might expect that they all evolved from this common ancestor in New Zealand.
However, this expectation is not supported by both fossil and molecular (DNA) evidence. In the mid-nineties, Steven Trewick (note the link to his personal page is broken but this takes you to his lab website) undertook a study of New Zealand rails to determine their origins, evolutionary history and the relationships among them. What he found was somewhat surprising in regard to the mōho and takahē. Rather than sharing a common ancestor with an in-situ divergence event, the mōho and takahē were the result of two separate colonisations by similar volant ancestors—probably a species similar to the pūkeko. This evidence resulted in a revision of the two subspecies into two separate species as indicated above. It also raised some interesting questions regarding the evolution of flightlessness among New Zealand rails.
New Zealand has (including recently extinct species) a relatively high diversity of rails (18), about half of which are flightless. An interesting question that Steve Trewick and his research group are asking is why seemingly mobile, volant species evolve into flightless range-restricted species. This comes partially from the fact that the likely ancestors of the mōho and the takahē came from the same lineage as the pūkeko. Because birds are highly mobile - the pūkeko being a perfect example in this case the - gene flow tends to be high reducing the probability of speciation. In fact, there is little evidence for speciation among birds on islands. In other words, it takes a substantial barrier to stop gene flow in birds which reduces the probability of speciation. Yet there have been several independent rail colonisations in NZ with many resulting in a flightless species. To me this indicates that there is strong selection for rails to fly when necessary (i.e. in the presence of predators) but it is easily relaxed without predators. But the question of even rates of gene flow remains.