The world’s most studied rainforest is still yielding new insights

Conservationists generally disapprove of flooding species-rich habitats. But the law of unintended consequences works in mysterious ways. For it was just such a flood, in 1913, that created Barro Colorado Island in central Panama. Gatun, the lake surrounding the island, was, at the time of its birth, the largest artificial body of water in the world. It formed the middle passage of the Panama Canal. Barro Colorado, meanwhile, has become the most intensively scrutinised scrap of tropical rainforest on Earth.

The field station from which this scrutiny is conducted opened in 1924, and in 1946 it was taken over by the Smithsonian Institution, which dubbed the operation the Smithsonian Tropical Research Institute (STRI). Since then, the island, and some neighbouring areas of forest on the mainland that have fallen under the STRI’s auspices, have turned into biology’s equivalent of a big physics facility like CERN, the home of the Large Hadron Collider. Though the STRI itself employs only 35 scientists, a further 1,200 visit each year, many of whom are regulars.

To mark the field station’s centenary, a horde of old hands and enthusiastic youngsters gathered on June 18th in Gamboa, the island’s nearest town, for three days of festivities to celebrate the past and plot the future. They discussed everything from how new species arise and the effect of lightning on insect ecology, to tropical forests’ role in curbing climate change.

Biology’s hundred-year scrutiny of Barro Colorado means its wildlife has been recorded in exquisite detail. The current roll lists 110 species of mammal (70% of which are bats), 384 birds, 33 frogs and toads, two salamanders, 23 lizards, 41 snakes, more than 400 ants, some 600 butterflies, 10,000 beetles and even 100 cockroaches. There are also 1,400 species of plants. Given this abundance, it is hardly surprising that the place has been the source of several fundamental ecological findings, and many new techniques of observation.

Early on, for example, it provided the first good test of the idea that the number of species a patch of habitat can support depends on its area. The so-called species-area relationship predicted that Barro Colorado’s sudden isolation would cause local extinctions. Which it did. Ornithologists saw bird species dwindle in number by more than a quarter, while remaining unchanged in similar habitats on the mainland. Among other things, that result demonstrates the value of scale, if nature reserves are to be effective.

Barro Colorado also hosted the first serious study of wild primates (howler monkeys, by Clarence Ray Carpenter in the 1930s), the first big field investigation of echolocation in bats (by Donald Griffin in the 1950s), and the first attempt to attach numbers to the obvious truth that the tropics support more species of creepy-crawly than temperate climes do. This was launched in 1971, by Terry Erwin, an entomologist who “fogged” tree canopies on the island with insecticide, collected what fell out, and recorded the arthropods he found in each tree type. Combining this information with estimates of the number of tree species around the world, he concluded that Earth supported about 30m sorts of insects and other terrestrial arthropods. The current calculation is about a quarter of this, so it was not a bad estimate.

The island’s plants have not been neglected. Between 1981 and 1983 a block of 50 of its 1,560 hectares was surveyed in detail by Robin Foster and Stephen Hubbell, to record the location and species of every tree and shrub with a stem-diameter of more than 1cm (initially, these numbered 242,000). The Forest Dynamics Plot, as it is known, has since been resurveyed every five years, permitting the life cycles of many woody plants to be understood in far better detail than was previously possible. It has also inspired the establishment of 77 similar plots in forests around the world, both tropical and temperate, containing 7m specimens of 13,000 species.

On top of all this, the STRI has itself provided a salutary example of the social evolution of American science over the past century. The field station’s founders were men carved in the image of Theodore Roosevelt, the American president whose methods for getting the Panama Canal built included involving his country in a minor war. (The old-boy culture endured, as sexual-misconduct allegations revealed in 2021 made plain.) Today, though, the institute has become—fittingly—more representative of human diversity.

It is one thing to measure and describe the luxuriant biodiversity of the tropics, though, and quite another to explain it. Speciation’s details can be hard to demonstrate. But in the case of trees of the genus Inga, which boasts 300 species, Phyllis Coley of the University of Utah, one of the station’s old hands (she began her work there in 1995), thinks she has done so. The driving force behind Inga’s diversity, according to Dr Coley and her team, is herbivory—specifically by the moth and butterfly caterpillars that eat its members’ leaves.

Over the years, they have studied 174 of these lepidopteran persecutors. They have shown that most are picky eaters, choosing their meals based on which toxic chemicals (there are about 200 varieties) particular trees pack into their leaves. Such specialisation in turn encourages diversification, as new species with new, temporarily insect-proof toxin profiles emerge.

Predation by insects is not, however, the only thing that encourages arboreal speciation. Habitat differences can help. Camila Pizano, of Lake Forest College in Illinois, studies the island’s population of Trema micrantha. This comes in two types, distinguishable by the sizes of their seeds, which prefer subtly different habitats.

As she explained, both require the forest’s otherwise-uninterrupted canopy to have been opened to the sky by some accident to provide light for their growth. One variety, though, does best if the opening was caused by a landslip, which clears the soil of nitrogen-rich leaf litter and exposes the phosphorus-rich subsurface. The other thrives in the aftermath of a tree fall, when the soil remains nitrogen-rich and phosphorus-poor. Whatever the details, T. micrantha looks like a species dividing in two.

How insects speciate is also under investigation at the STRI. Owen McMillan and his group study a genus of toxic butterflies called Heliconius, which warn predators off by evolving distinctive colour patterns mimicked by other, less toxic, species. As part of their work this group’s researchers have used CRISPR genetic editing to create a novel butterfly wing-colour pattern that looks, superficially, at least, like a new species. Whether the new morph, with its ivory-coloured wings, would prosper in the wild is unlikely ever to be tested, for releasing gene-edited organisms is frowned on.But the ability to create new heliconid designs is intriguing.

There was more. Much more. Rachel Page, heir to Griffin as head of the institute’s bat research, reviewed classic studies of how the island’s 74 bat species are defined not only by what they eat (fruit, frogs, fish, insects and, for vampires, blood) but also by when and where, thus minimising competition and increasing the effective number of ecological niches.

Kane Lawhorn entertained the troops with an explanation of how the frequency of lightning strikes regulates beetle populations (by providing food in the form of dead and injured trees), while Andrew Seiler talked of a different sort of “weather”: the constant rain of arthropods pitter-pattering from the canopy to the forest floor.

Roland Kays described his adventures tracking kinkajous (arboreal relatives of raccoons) with GPS-enabled radio tags and told of one particularly large and dominant male being posthumously crapped on in apparent triumph by rivals. His collaborator, Meg Crofoot, explained that kinkajous, which have smaller brains than the local monkeys, are just as successful at foraging for food, suggesting better foraging is not the reason big brains have evolved.

Such findings are striking. But Barro Colorado’s most important contribution to ecology may be in serving as inspiration to try similar things elsewhere. Those 77 simulacra of the Forest Dynamics Plot, for example, constitute, together with the Panamanian original, a network called the Forest Global Earth Observatory (ForestGEO), which is an attempt to compare and contrast forest ecology across the planet. Stuart Davies, ForestGEO’s director, told the meeting that ForestGEO itself is being recruited, along with similar networks covering tropical, temperate and boreal forests, into a megaproject called GEO-trees. The organisers of this hope to use local estimates of biomass to calibrate satellite measurements of the same areas. That would make extrapolations from space-based measurements more reliable.

This is important. Tropical rainforests are huge stores of carbon that might otherwise add to atmospheric carbon dioxide. Understanding whether they are, in any given place, increasing their carbon stocks or releasing them into the air, is crucial to monitoring global warming. More accurate measurements would also enable more regular check-ups to police untoward changes. For, much as urgent change is needed to combat global warming, when it comes to forests, the urgent need is for nothing to change at all.

© 2024, The Economist Newspaper Limited. All rights reserved. From The Economist, published under licence. The original content can be found on www.economist.com