The mammals of the Nansei Shoto may be inconspicuous and difficult to observe, but their distribution, and the relationships between the different species and populations in these islands, provides insight into the past history of the archipelago. This interesting mixture of animals has links in the north to Honshu and in the south to Taiwan and the Asian continent. Unique forms have evolved locally, especially in the central islands of the archipelago.

It takes no real stretch of the imagination to picture how rising sea levels could have flooded lowlands, separating larger islands into smaller ones and marooning creatures on the separate parts of what was once a larger homeland. Small mammals, rodents, rabbits and the like would be unable to swim the deeper channels that were formed in this way between the new islands. In isolation they might evolve slowly in tune with the local conditions, and so end up distinctly different from what were once close relatives.

Where flooded channels between new islands were shallower and narrower, however, it would have remained possible for terrestrial animals to swim across. The regular input of genetic material from one island to the next might help keep them a single species.

Species dependent on living in trees might be reluctant to cross even shallow water, so while some species might cross a channel of a given size, others might not.

We might therefore expect that the closer islands are together, the more likely they are to share the same fauna and flora, and conversely, the further apart they are, the more likely it would be that their fauna and flora would become different over time. These very patterns are observable among the mammals of the Nansei Shoto, as I have described in the last two articles.

It is easy to imagine animals that live on the ground being divided by water, but it is much harder to think of birds being affected by the same barriers -- yet that is exactly what we find. Given that typical birds can fly, why then would they be so affected by the distribution of islands and the distance between them?

By no means are all birds affected. Many species are migratory, and fly from one island to the next with little effort. I have stood in the hills of northern Okinawa at dawn and watched in awe as dozens of Chinese sparrowhawks flew overhead, moving steadily southward.

It was particularly exciting knowing that other observers on islands further north and south were tracking the same birds. We knew that the sparrowhawks were moving out of the continent, migrating down the Korean Peninsula, passing over Kyushu and island hopping south via the Nansei Shoto to Taiwan and beyond.

Many smaller species of birds pass through on migration too, and they are quite obviously able to cross water, so how is it that some are unable?

Recent research has shown that even within the same species, some populations migrate longer distances than others, while some populations may not migrate at all. Migratory populations and migratory species, it turns out, have longer wings than nonmigratory ones. It seems then, that although nonmigrants have wings for flight, they don't have wings for migration.

Birds living in a rich forest environment may have small territories and need to fly only short distances. In general, fewer predators live on islands, especially small islands, so the birds living there even have less reason to fly; they no longer have predators to flee from, so again their ability to fly long distances may be reduced. In fact, among island bird species quite a high proportion are entirely flightless, having given up the ability.

Typical are the rails. These sturdy, omnivorous, ground-dwelling birds do well on islands, or at least they did until humans introduced rats, cats and the like.

Flightless island rails are especially vulnerable to alien predators. The native Guam rail was completely wiped out in the wild by 1987 as a consequence of brown tree snakes being introduced to its home island. The snakes, which were alien to the island, were intended for rodent control, but finding the native birds easy to catch, they tucked into them quickly. Captive breeding has saved the Guam rail and made it possible to introduce them to a nearby snake-free island, but many other flightless island species have disappeared before such conservation efforts could be put into action.

The virtually flightless Okinawa rail is confined to the forests of the northern part of Okinawa, which it shares with the equally restricted and even rarer Pryer's woodpecker. To the north, on Amami-Oshima, lives the very attractive forest-dwelling Lidth's jay and the secretive Amami thrush.

Sharing that island with them is the Amami woodcock. The Amami woodcock seems more reluctant to fly than its woodcock relatives in other parts of the world, but it can fly and it is also found on both Tokunoshima and Okinawa in addition to Amami-Oshima. Did the Amami woodcock evolve on one of these islands and then somehow manage to cross later to the others, or did it previously have a wider range on what was once a larger island that has now been divided into several by changes in sea level?

Either way, here in these central Nansei Shoto islands there are at least five distinct bird species that occur nowhere else in the world, not even on the northern or southern islands of the same archipelago. An extraordinary coincidence, for they live in the very same central region, where the Nansei Shoto's unique mammals are also found.

If you recall, the northern and southern islands have few isolated species of mammals. That pattern is also repeated among the birds. Although the ranges of a number of birds, such as the watercock and the white-breasted waterhen, creep north just far enough to include the southernmost Nansei Shoto islands and no other part of Japan, none of the species there are unique to the locality. Similarly, some species more typical of Kyushu and further north just reach into the northern islands and no further south, but there again, no birds are found only on those northern islands.

It seems then that the basic pattern of distribution of isolated species in these islands is pretty much the same for both mammals and the birds, even though the latter can fly.

There are mammals, such as the Ryukyu fruit bat, that occur on a number of islands in the Nansei Shoto archipelago, but not beyond it. Once again, there are parallels among the birds, with species such as the Ryukyu robin and the Ryukyu minivet having archipelago-wide ranges, without being restricted to any particular islands.

Further investigation may show that species with more general, broader habitat requirements have the wider ranges within the islands, while other species are more restricted because of their narrower habitat requirements.

There is still much to learn about these islands, but every piece of the puzzle that is found and fitted into place makes them seem even more the Galapagos of Asia.