The research team, led by Richard Condit of the Smithsonian Tropical Research Institute's Center for Tropical Forest Science, compared data from single-hectare (2.47 acre) tropical forest plots near the Panama Canal with plots of the same size in the Yasuni National Park of Ecuador and in Peru's Manu Biosphere Reserve. After identifying, tagging and measuring more than 50,000 individual trees with stems of ten centimeters or more in diameter in all three forests, they observed that a wide swath of the western Amazon has a forest in which the species change very little over distances of more than 1000 kilometers. The tree species counts in any one locale are high, but each locale turns out to be much like the others in terms of species composition.

In contrast, forests on the Isthmus of Panama change dramatically in tree species composition from one site to the next. Forests just 50 kilometers apart in Panama are less alike than forests 1,400 kilometers apart in the western Amazon. As a result of such high landscape variation, parts of Panama have as many or even more tree species than parts of Amazonia. "Ecologists have a technical term for landscape variation in forest types: beta-diversity," Condit explained. "Beta-diversity is high when forests change a lot over short distances -as in Panama - but low when forests are similar over long distances - as in Ecuador and Peru." The unique aspect of this research by the Smithsonian team, including colleagues from France, the United States and South America, was a precise mathematical prediction of beta-diversity that helped them pinpoint its cause. A theory for beta-diversity had heretofore eluded ecologists.

"The Smithsonian theory is based on a basic ecological premise called the 'neutral theory,'" Condit said, "but adds to it the simple yet crucial observation that trees do not generally spread their seeds very far - a factor which tends to enhance beta-diversity."

The Science report provides one of the most precise tests of the neutral theory yet published.

The team concludes that the neutral theory cannot account for beta-diversity in tropical forests, and they discount the importance of random events in establishing what grows there. Instead, Panama's high beta-diversity must be due to the abrupt variation in rainfall across the Central American isthmus, from the ever-wet Caribbean shoreline to the dry Pacific slope.

Forests across western Amazonia, however, were more uniform in species composition than the theory allowed, a surprising result.

"Explanations for this uniformity will require deeper understanding of how different tropical trees are from one another," said co-author and Smithsonian scientist Egbert G. Leigh, Jr., who devised the mathematical formula that led to the undermining of the neutral theory.

"More tedious field work, it seems, is in store," Leigh concluded.

The Center for Tropical Forest Science, established within the Smithsonian Tropical Research Institute in 1990, is a consortium of forestry agencies, universities, research institutes and nongovernmental organizations around the world, each managing or involved in one or more of 17 forest dynamics plots in 14 different countries. In addition to monitoring the trees, the center sponsors training programs, scientific meetings, and communications between sites through a newsletter and Website at

The Smithsonian Tropical Research Institute, headquartered in Panama City, Republic of Panama, is one of the world's leading centers for basic research on the ecology, behavior and evolution of tropical organisms. More information is available at