The status of many tropical species cannot be classified either as 'definitely extinct' or 'definitely in existence', but just as 'unknown'. Instead, it becomes a matter of chance which species happens to attract the attention of some ornithologist, becomes the object of a specific search, and hence may be recognized as possibly extinct.
Here is an example. The Solomon Islands are another of my favourite bird-watching areas in the tropical Pacific Ocean, and will be recalled by older Americans and Japanese as the site of some of the fiercest fighting in the Second World War. (Remember Guadalcanal, Henderson Field, President Kennedy's PT boat, the Tokyo Express?) The ICBP lists one Solomon bird species, Meek's crowned pigeon, as extinct. Yet when I tabulated all recent observations of all 164 known Solomon bird species, I noticed that twelve of those 164 species had not been encountered since 1953. Some of those twelve species are surely extinct, because they were formerly abundant and conspicuous. Several Solomon islanders told me that those birds had been exterminated by cats. Twelve species possibly extinct out of 164 still may not sound like much to worry about. However, the Solomons are in much better shape environmentally than most of the remaining tropical world, because they have relatively few people, few bird species, little economic development, and much natural forest. More typical of the tropics is Malaysia, which is rich in species and has had most of its lowland forest cut down. Biological explorers had identified 266 fish species dependent on fresh water in Malaysia's forest rivers. A recent search that lasted four years was able to find only 122 of those 266 species—less than half. The other 144 Malaysian freshwater fish species must either be extinct, rare, or very local. They reached that status before anyone noticed it.
Malaysia is typical of the tropics in the pressure it faces from humans. Fish are typical of all species other than birds, in that they attract only patchy scientific attention. The estimate that Malaysia has already lost (or nearly lost) half of its freshwater fish is therefore a reasonable ballpark figure for the status of plants, invertebrates, and vertebrates other than birds in much of the rest of the tropics.
That is one complication in trying to pinpoint the number of extinctions since 1600: the status of many or most named species is unknown. But there is a further complication. So far, we have been trying to assess the extinction only of those species that had already been discovered and described (named). Could any species have become extinct before they were even described? Of course they could, since sampling procedures suggest that the actual number of the world's species is near thirty million, but less than two million species have been described. Two examples illustrate the certainty of other species becoming extinct before description. Botanist Alwyn Gentry surveyed the plants of an isolated ridge in Ecuador called Centinela, where he found thirty-eight new species confined to that edge. Shortly afterwards, the ridge was logged and those plants were exterminated. On Grand Cayman Island in the Caribbean, zoologist Fred Thompson discovered two new species of land snails confined to forest on a limestone ridge that was completely cleared a few years later for a housing development. The fact that Gentry and Thompson accidentally visited those ridges before rather than after they were cleared means that we have names for those extinct species. But most tropical areas that are being developed are not first surveyed by biologists. There must have been land snails on Centinela, and plants and snails on innumerable other tropical ridges, that we exterminated before we discovered them.
In short, the problem of determining the number of modern species that have become extinct seems at first to be simple and to lead to modest estimates—for example, only five or six extinct bird species in all of North America plus Europe. On reflection, though, we appreciate two reasons why published lists of species known to be extinct must be gross underestimates of the actual numbers involved. Firstly, by definition the published lists consider only named species, whereas the great majority of species (except in well-studied groups like birds) have not even been named. Secondly, outside North America and Europe and except for birds, the published lists consist only of those few named species which some biologist happened to get interested in for one or another reason and found to be extinct. Among all those remaining species of unknown status, many are likely to be extinct or nearly so—for example, about half in the case of Malaysian freshwater fishes.
Now let's move on to the second step in evaluating the mass extinction debate. Our estimates up to this point have concerned only those species exterminated since 1600 AD, when scientific classification of species was beginning. These exterminations have taken place because the world's human population has grown in numbers, reached previously uninhabited areas, and invented increasingly destructive technologies. Did these factors spring up suddenly in 1600, after several million years of human history? Were there no exterminations before 1600? Of course not. Until fifty thousand years ago, humans were confined to Africa plus the warmer areas of Europe and Asia. Between then and 1600 AD our species underwent a massive geographic expansion that took us to Australia and New Guinea around 50,000 years ago, Siberia around 20,000 years ago, most of North and South America around 11,000 years ago, and most of the world's remote oceanic islands only since 2000 BC. We also underwent a massive expansion in numbers, from perhaps a few million people 50,000 years ago to about half a billion in 1600. Our destructiveness also increased, with the development of improved hunting skills in the last 50,000 years (Chapter Two), polished stone tools and agriculture in the last 10,000 years (Chapters Ten and Fourteen), and metal tools in the last 6,000 years.
In every area of the world that paleontologists have studied and that humans first reached within the last 50,000 years, human arrival approximately coincided with massive instances of prehistoric species' extinction waves. For Madagascar, New Zealand, Polynesia, and the Americas I have described those instances in the preceding two chapters. After people reached Australia, that continent lost its giant kangaroos, its 'marsupial lion', and other giant marsupials. Around the time that Indians reached North America 11,000 years ago, it lost lions, cheetahs, native wild horses, mammoths, mastodonts, giant ground sloths, and several dozen other large mammals. Mediterranean islands like Crete and Cyprus lost dwarf elephants and pygmy hippos, while Madagascar lost giant lemurs and flightless elephant birds. New Zealand lost its giant flightless moas, and Hawaii its flightless geese and dozens of smaller birds, when the Polynesians arrived around 1000 and 500 AD, respectively.
Ever since scientists became aware of these prehistoric extinction waves associated with human arrival, they have argued over whether people were the cause or just happened to arrive while animals were succumbing to climate changes. In the case of the extinction waves on Polynesian islands, there is now no reasonable doubt that Polynesian arrival in one way or another caused them. Bird extinction waves and Polynesian arrival coincided within a few centuries at a time when no big climate change was happening, and bones of thousands of roasted moas have been found in Polynesian ovens. The coincidence of timing is equally convincing for Madagascar. But the causes of the earlier extinction waves, especially those in Australia and the Americas, are still being debated. As I explained on America's extinction waves in Chapter Eighteen, the evidence seems to me overwhelming that humans also played a role in those prehistoric cases of extinction outside Polynesia and Madagascar. In each part of the world an extinction wave occurred after the first arrival of humans, but did not occur simultaneously in other areas undergoing similar climate swings, and did not occur in the same area whenever si rnilar climate swings had occurred previously.