From what little we can discern from the geologic record, a normal recovery time may require millions of years. After the dinosaur crash, for instance, between 50,000 and 100,000 years elapsed before there started to emerge a set of diversified and specialized biotas, and another 5 to 10 million years went by before there were bats in the skies and whales in the seas (Jablonski, 1986). Following the crash during the late Permian Period, when marine invertebrates lost about half their families, as many as 20 million years elapsed before the survivors could establish even half as many families as they had lost (Raup, 1986).
The evolutionary outcome this time around could prove even more drastic. The critical factor lies with the likely loss of key environments. Not only do we appear ready to lose most if not virtually all tropical forests, but there is also progressive depletion of coral reefs, wetlands, estuaries, and other biotopes with exceptional biodiversity. These environments have served in the past as preeminent power-houses of evolution, in that they have supported the emergence of more species than have other environments. Virtually every major group of vertebrates and many other large categories of animals have originated in spacious zones with warm, equable climates, notably tropical forests. In addition, the rate of evolutionary diversification—whether through proliferation of species or through the emergence of major new adaptations—has been greatest in the tropics, again most notably in tropical forests.
Of course tropical forests have been severely depleted in the past. During drier phases of the recent Ice Ages (Pleistocene Epoch), they have been repeatedly reduced to only a small fraction, occasionally as little as one-tenth, of their former expanse. Moreover, tropical biotas seem to have been unduly prone to extinction. But the remnant forest refugia usually contained sufficient stocks of surviving species to recolonize suitable territories when moister conditions returned (Prance, 1982). Within the foreseeable future, by contrast, it seems all too possible that most tropical forests will be reduced to much less than one-tenth of their former expanse, and their pockets of holdout species will be much less stocked with potential colonizers.
Furthermore, the species depletion will surely apply across most if not all major categories of species. This is almost axiomatic, if extensive environments are eliminated wholesale. The result will contrast sharply with the end of the Cretaceous Period, when not only placental mammals survived (leading to the adaptive radiation of mammals, eventually including humans), but also birds, amphibians, and crocodiles, among other nondinosaurian reptiles. In addition, the present extinction spasm looks likely to eliminate a sizeable share of terrestrial plant species, at least one-fifth within the next half century and a good many more within the following half century. By contrast, during most mass-extinction episodes of the prehistoric past, terrestrial plants have survived with relatively few losses (Knoll, 1984). They have thus supplied a resource base on which evolutionary processes could start to generate replacement animal species forthwith. If this biotic substrate is markedly depleted within the foreseeable future, the restorative capacities of evolution will be all the more reduced.
In sum, the evolutionary impoverishment of the impending extinction spasm, plus the numbers of species involved and the telescoped time scale of the phe-