From my hotel room in the Solomons, there were two kinds of animals that stubbornly made their presence known: birds and ants. The tiny ants can almost always be seen scuttling along the baseboards in orderly lines and the birds keep up the chatter, a steady and persistent racket that is impossible to ignore (I think several live just outside my window). Scientists must find them impossible to ignore as well. The birds and ants of the Solomons, and the Pacific Islands in general, have long helped scientists try to understand how species develop.

Usually studying one of these two groups, birds or ants, researchers such as Robert MacArthur, Edward O. Wilson, Ernst Mayr and Jared Diamond used the Pacific Islands as ocean-spanning natural laboratories to create theories about why certain species are where they are, how they got there and how they evolved to their present forms.

One of the long-held assumptions was that islands are evolutionary dead-ends; species go to islands, evolve to suit the environment and then eventually go extinct with the arrival of other species or some change in the habitat. The trip from continent to island is a one-way affair, it was thought, and all island communities, no matter how far flung, are descended directly from continental sources. For example, the flycatchers in Marquesas, on the far eastern side of Polynesia, are descended directly from birds that arrived there from Australasia. Same for the birds in the much closer Solomons and Fiji in between. This led Mayr to conclude that there are really no Polynesian birds, only birds that arrived there and changed.

This work was based on taxonomic data--descriptions of the birds and their behavior--which has obvious limits in reconstructing the natural history of a species. Chris Filardi and his colleagues at the American Museum of Natural History recently applied DNA analysis to the same question (and the same birds) and have now overturned many of the long-held assumptions that came from Mayr’s work. Essentially, Filardi found that islands are hardly evolutionary dead-ends, but rather engines that may help drive diversification all over the world. His results were published Nature in November 2005 (if this sounds familiar, the work generated some media attention when it came out, including a piece by Carl Zimmer in The New York Times).

Filardi found that the 50 or more flycathers found throughout the Pacific were the result of a single radiation. Flycatchers therefore do not go to an island from the continent, diversify then die. Rather, they hop between islands, diversify and hop some more. And some have even made it back to the continent (namely, Australia). This might be the big finding of the study, that islands are not dead-ends but rather can generate biological diversity and then send it back to a continent. And all of this seems to have happened in a relatively short time, perhaps 1.5 million years.

The sad part of the story, of course, is that many of the birds from this family are seriously endangered or have recently gone extinct. In that sense, islands are to some extent evolutionary dead-ends because species there are always at high risk of extinction.

I’ve left the Solomons for Vanuatu, which, refreshingly, lacks the history of ethnic tension that has affected other islands. While Vanuatu only has 83 islands to the Solomons 900, it is just as culturally diverse. Like the Solomons, many of these cultures and languages are still alive today and most people still live in small traditional villages. I’ll be spending my time here on the largest island, Espiritu Santo, where I’ve started diving the enormous SS Calvin Coolidge, a luxury liner that was converted to a military transport and sunk by American mines in 1942. It’s enormous, quite deep and bloody terrifying to be inside. I’ve got two weeks to explore the whole thing.

(I've been in it four times by the time I finally got around to posting this entry. The Holy Shit factor is just off the charts.)