Tropical Scientists Find Fewer Species Than Expected

It is important to know how fast biodiversity is being lost, but this is hard to gauge without a solid baseline, Weiblen said. Scientists advising governments on policies to curtail species losses must have credible estimates of species numbers if they are to shape appropriate policies. The stakes are high because losses of too many species or certain kinds of species can cripple tropical forest ecosystems, which normally stabilize soil and climate, purify and recycle water, and produce food, medicine, building materials or other useful products, he said.

The current study took a cross-disciplinary approach; besides Weiblen, the principal scientists were insect expert and project coordinator Scott Miller of the Smithsonian Institution and insect community ecologists Yves Basset of the Smithsonian Tropical Research Institute in Panama and Voytech Novotny of the Czech Academy of Sciences.

The team compared insect communities feeding on 51 tropical plant species, most belonging to either the fig family, the mulberry family or the coffee family. While previous work had also based estimates of total species on numbers of insects--the most species-rich class of animals known--scientists had assumed that each plant-eating insect species tended to feed on one or very few plant species. If true, the estimates of the number of herbivorous insects should be tied to the number of plant species.

"Instead, most insects turn out to be specialized not to plant species, but rather to a genus [grouping of species] or family [grouping of genera] of plants," Weiblen said. "Fewer effective plant 'hosts' means fewer herbivores. Where people had assumed that different insects' food sources overlapped very little, we found that many insects share their food plants with other insect species. There are actually few extreme specialists among tropical insect herbivores."

New Guinea's species diversity made the island an ideal laboratory for the study. Its tropical forests are about the same area as Texas but contain five percent of the world's biodiversity, Weiblen said. The island has 12,000 species of plants; in comparison, Weiblen's home state of Minnesota has only about 1200 species. The work began in 1994 with studies of the insects eating 15 species of figs--Weiblen's area of specialty--that co-occur in the rainforest. The study then expanded to include the mulberry family, which is related to figs, then members of the coffee family and more distantly related plants. The team used DNA sequences to sort out the evolutionary relationships between plants and looked at how insect species were distributed on them.

They found that the tropical forest plant community was dominated by clusters of closely related plant species and that insects tend to feed on multiple close relatives in a given plant genus or family. Because the number of insect species was tied to broader--and therefore less numerous--categories of plants, estimates of their numbers had to be diminished, Weiblen said. Once the estimate of herbivorous insects had been made, the researchers plugged that number into equations to estimate the total number of species in the world. Such equations take into account the proportion of insects that are herbivorous, the proportion of all species that are insects or other members of the arthropod group, the proportion of all plant species that are found in New Guinea, and other factors. Depending on whether insects as a whole follow the pattern of beetles, whose food preferences are quite broad, or moths and butterflies, which are more particular, they estimated that the world contains a total of between about 4.8 million and 6 million species.

"Our estimates bring some reality to predictions about declining biodiversity in the sense that the consequences for insect herbivores of losing a particular host plant species may not be as dire as previously thought," Weiblen said. "But that is no reason to ignore the decreasing number of species worldwide. [Harvard biologist] E.O. Wilson predicted in his book 'The Future of Life' that half of all species will suffer extinction in 50 years if current land use patterns continue. Because the consequences are so severe, we've got to refine our predictions and conserve as many species as possible."