My latest Mind and Matter column in the Wall Street Journal:
A recent paper in the journal Nature concluded that species extinction caused by habitat loss is happening less than half as fast as usually estimated. The normal method for calculating rates of extinction assumes that doomed species merely cling temporarily to a shrunken patch of habitat, on their way to disappearing (an idea called "extinction debt"). Apparently, this isn't the case: Although a larger patch of habitat has more species in it, shrinking a patch does not lead to a proportional rate of species loss.
According to the authors of the study, the biologists Stephen Hubbell and Fangliang He, estimates of extinction rates based on the usual method are "almost always much higher than those actually observed." Though you need a big patch of forest to attract a rare species, you do not need such a big patch to retain it once it is there. Mr. Hubbell added: "The method has got to be revised. It is not right."
This is good news for conservationists and should cause them to redouble their efforts to rebuild and reconnect the scraps of wilderness that are left-as is happening in some parts of the world, even as habitats are lost elsewhere. The new study tells us that no patch of wild habitat, however small, is a hopeless cause, and that the local disappearance of a species can often be reversed. Consider, for example, the Brazilian coastal rain forest. Ninety percent of it has been destroyed, but not a single bird species has gone extinct as a result (though one survives only in captivity).
It also suggests that in focusing on habitat loss we have been neglecting the chief causes of species extinction. The species-extinction crisis is largely-and always has been-about the disappearance of unique and isolated creatures from islands, not the vanishing of species on continents: for example, the dodo from Mauritius, a host of birds from Hawaii and scores of unique fish from Lake Victoria (lakes are water islands).
In nearly all such cases, the damage was done not by habitat loss but by the introduction of predators, competitors or parasites: monkeys and pigs in Mauritius, rats and other birds in Hawaii, Nile perch in Lake Victoria. The species-extinction crisis on islands peaked around 1900, but it continues today. In the Galapagos and other places, newly arrived animals are driving endemic species to the brink.
By contrast-and so long as you count Australia as an island, because its rash of extinctions was caused mostly by introduced aliens-the rate at which continents are losing species is remarkably slow, despite huge changes in habitat wrought by human beings. According to the Red List of the International Union for the Conservation of Nature, 122 bird species and 58 mammals have gone extinct in the last 500 years. But of these, the independent scholar Willis Eschenbach has concluded, only six birds and three mammals were on continents-out of 8,971 and 4,428 continental species, respectively. None was exclusively a forest dweller, and none was extinguished exclusively by habitat loss. (Here's Eschenbach's graph; his essay is well worth reading:)
Europe got through the 20th century without losing a single species of bird. (The Faroese pied raven was at most a subspecies.) The last European breeding bird to die out altogether was the great auk-an island species-in the 1840s. In a drastic and unusual case of habitat destruction, an underwater volcano off Iceland finally did in the flightless bird, after centuries of human persecution. The eruption sank the great auk's last breeding colony, an island called Geirfuglasker. A forlorn few pairs subsequently tried breeding on the much less suitable island of Eldey, but they were killed by a collector of rare birds.
This month there was a swarm of earthquakes at the location of Geirfuglasker, so perhaps the island will be reborn. Can genetic engineers do the same for the great auk?