Asia there are giant squirrels, and so on. As we contemplate the mass extinction of tropical animals, we can speculate that if it were not for a dramatic historical extinction we might be considering the dodo (Raphus cuculatus) as a candidate for domestication and transtropical exploitation. This giant (25-kilogram) pigeon was flightless and probably subsisted on fruits, just like the paca! Just as the dispersal strategies of some Neotropical trees seem to have coevolved with the pacas, so, it seems, have some of the Mauritian fruits coevolved with the dodo. Temple (1977) suggested that the hard, nutlike seed of Calvaria major depended for its germination on its abrasion in the gizzard of the dodos.
Janzen and Martin (1982) suggested that at least 30 species of trees in Costa Rica depended for their dispersal on the digestive processes of some now extinct, large (sometimes giant), herbivores of the Neotropics. These, like the dodo, were probably eminently edible. Similarly, the giant ground sloths of South America and the giant birds of New Zealand, known from historic times, could have been candidates for domestication. This is not fruitless speculation but a notice that our regrets about these extinctions are certainly likely to be matched soon by more massive regrets about the ongoing and imminent extinctions.
Our knowledge of plants with a high potential for future exploitation is very sketchy indeed. The Board on Science and Technology for International Development within the National Research Council has highlighted some examples (NRC, 1975, 1979, 1981, 1984, 1985). In Chapter 10 of this volume, Iltis has drawn attention to some other dramatic examples. Again on a priori grounds we can predict that there must be many plants that have or are in themselves valuable, undiscovered products.
The tropical forests of the world are the most intense battlegrounds for species competition on the face of the Earth. This intensity is a direct result of extraordinary biodiversity and the continuingly ferocious evolutionary arms race. Huge assemblages of animal species eat leaves, and it is no exaggeration to say that tropical plants must have evolved a formidable array of insecticidal compounds and insect deterrents. We already use some of these. The picture is similar in the case of fungicides. The moist tropical forest is a superb environment for fungi of all kinds. Fungi relish warmth, wetness, and an abundance of organic substrates. The leaves of tropical trees are not only assaulted by insects and other animals but also by fungi and epiphyllic plants. What a place to look for fungicides and herbicides!
An illustration of how fundamental studies lead to discoveries in this field comes from a report of the food preferences of leaf-cutter ants by Hubbell et al. (1983). They found that leaf-cutter ants (Atta sp.) rejected the leaves of the leguminous tree Hymenaea courbaril. Such leaves kill the fungus that the ants grow in their nests as a food resource. It is entirely improbable that the tree evolved the fungicide to deter leaf-cutter ants. Rather, it probably had to cope with a broad spectrum of epiphyllic fungi and evolved a broad-spectrum fungicide to (literally) keep its leaves clean. Laboratory tests on the fungicide, a terpene-caryophyllene epoxide, show antifungal activity against a wide range of pathogenic fungi. This precisely illustrates the points I make above. It surely means that we can confidently predict, because of the nature of tropical forest ecology, that its plants should be a source of a wide range of useful products. These await discovery. There is little doubt that