The Kingdom of Rarities (11 page)

When we sat down to breakfast the next morning, we disturbed a clump of fruit flies camped in the bowl of bananas. These shortlived insects feasted on the ripe flesh and skin of fruit. They might
spend a fruit fly eternity, measured in days, in a home range no larger than a serving bowl. Across the river, scarlet macaws preened their lustrous feathers in the bright sunshine. Fruit flies and scarlet macaws lead lives that are polar opposites in range, age span, and relative rarity. The noisy birds live for decades and feed on a variety of nuts and seeds. Before George's large-predator project took off, macaws had been one of his main study subjects, along with quetzals and bellbirds, enigmatic bird species of Central America that present urgent conservation challenges. These birds all range widely, and so—in a biological sense—they are surprisingly similar to the big cats. They are “area sensitive,” meaning that they need huge expanses of habitat to survive. What can we learn about rarity from their study?

“We've found that macaws can fly hundreds, maybe thousands, of kilometers in search of an available nest cavity,” George said, gesturing toward the flock. They seem to know when and where to move to find food, no matter how far away. It is a triumph of avian memory in this vast forested region and possibly a trait under strong selection pressure. Individual birds that could time their food-searching forays accurately enough to find bonanzas of ripening nuts most likely survived to produce more offspring than those that could not.

On the habitat-use continuum from fruit flies to macaws, individuals of most species lean toward the fruit fly end and live out their lives in relatively small areas. Individuals that range over large distances often belong to the very species that networks of protected areas strive but often fail to conserve. So George had begun to focus on another important question: How were the rovings of macaws, jaguars, and peccaries related to specific features of their habitats? He was particularly interested in the distribution of forest tree species that might affect all three species and, of particular interest for the macaws, the spacing of the bonanza nut trees in their range and critical places to nest.

Like the sakis, macaws love to eat the unripe seeds of large forest
trees. In fact, they are seen as pests by Brazil nut harvesters, though research has shown they have little impact on yield. With their preferred food supplies dispersed throughout the region and seasonal in abundance, there is no other way for the macaws to live.

Macaws at a mineral lick, with a jaguar (
Panthera onca
) below

Macaws search not just for their bonanza nut trees but also for safe tree cavities in which to raise their young. The blue-and-gold macaws prefer the palm swamps that spring up in old oxbows of the
main rivers and nest in the tops of the
Mauritia
palms that grow there. The red-and-green macaws favor as one of their primary nesting trees a canopy genus called
Dipteryx
, which as a fruiting adult, as opposed to a sapling, occurs only about once every five hectares in this region. So, unlike the jaguars and pumas, which prowl at low densities because of energetic laws, or the sakis and other mature forest dwellers that are limited by dense-canopy sanctuaries, the macaws' spaced-out lifestyle at low density is defined by the availability of tree cavities in which to nest, and their preferred nesting-tree genus,
Dipteryx
, is rare.

So we have finally descended the tropical food chain, from the top carnivores to peccaries, monkeys, macaws, seeds, and the species that produce those seeds and give structure to the rain forest. Is the relative scarcity of adult
Dipteryx
the exception or the rule among tropical trees? Are tropical trees any different in the cause of their rarity from the animals that live in or below them? Of the 60,000 to 70,000 species of trees and woody climbers that fill the canopy across the tropical belt, how many are considered rare, and why? To answer these questions, ecologists have to understand the spacing of forest trees and its causes. In Madre de Dios, the problem was that botanists had recorded more than 1,000 tree species in the Los Amigos area alone (to put this in perspective, only 15 tree species are native to the United Kingdom).

The distribution of tropical trees presents the central puzzle of rain forest ecology and, by extension, of one major form of rarity—life at low density over a large range. Explain this ecological brainteaser and many things fall into place. To answer how such rarity is generated requires an understanding of why rain forest trees are so diverse in the first place. From New Guinea to the Amazon to the Congo or Indonesia—wherever you go, you typically find an extraordinary number of tree species in a small plot of land, say one hectare, often with each species represented by a single individual. Why is this pattern so prevalent in the tropics but less so in the temperate zone and boreal forests?

All tropical trees use the same photosynthesis system, utilizing light to convert carbon dioxide and water into sugar and oxygen. Different tree species may “specialize,” growing better in different light levels, but the tree species that occupy the canopy (as compared with the understory) have roughly equal access to light and to the available nutrients in the soil. If nature is said to abhor a vacuum, it also seems here to abhor a monoculture. Many tropical scientists believe that specialized insects and pathogens that primarily attack a single tree species check that species from dominating the forest. Foresters in both the temperate and boreal forests know that on plantations of trees or in vast stands of few species, such as the white or black spruce trees that blanket the upper latitudes, pest outbreaks are common. Such eruptions are even more common in tropical tree farms where trees of the same species are planted in large numbers—essentially a monoculture. Besides, in tropical regions, plant herbivory can be more intense than in forests closer to the poles; in the latter, the long winters suppress the activity of pest species. In contrast, tropical plant eaters are typically active every day of the year, especially in the wettest rain forests. If a given tree became common as a seedling, sapling, or small tree, the plant predators to which that species was susceptible would proliferate in the presence of more abundant food and would soon devour the increased supply. The combined effect of these leaf chewers, suckers, and defoliators is that individual trees that are spaced far from other members of their species—often a distance of a hectare or more—have higher survival rates. So the very nature of the hot, humid, and stable climate, especially warm winters, that leads to considerable diversity of trees also fosters a rich coterie of the creatures that inhibit their spread.

An allied theory accepted by many tropical biologists has held that patterns of seed predation and dispersal best explain the low density of canopy tree species. According to this hypothesis, if most of the seeds of an adult tree fell beneath its crown, waves of seed predators, from ants to agoutis, would finish them all off. Only those
seeds carried a safe distance from the parent by some animal, be it a fruit-eating bird, mammal, or even fish, would find a better and safer germination site. Biologists note that as much as 90 percent of all plants in tropical forests produce fruits that are dispersed by animals, whereas in the temperate zone the percentage is much lower, often as low as 10 percent, and most tree seeds are dispersed by wind, water, or gravity. Left relatively undisturbed and undiscovered by the seed predators congregating at the base of the parent, a distant seed would germinate and prosper. This phenomenon also contributes to tropical canopy trees growing far apart from one another.

For George and Sue, knowing the distribution of the trees that were important resources for the macaws and monkeys was a vital research question. So, like many before them, they worked with teams of botanists to map the distributions. There was one simple but pervasive problem, though: how to identify all those trees in the first place. Until the late 1980s, a lack of reliable identification guides to nonflowering and nonfruiting plants had held back progress in field tropical biology. Botanists relied on flowers and fruits to identify species, but few plants were ever in a reproductive phase at the same time, and the leaves of many species all seemed to converge on the same basic shape—narrow, without teeth on the margins, and with an elongated tip. Not much to go on.

Then came Alwyn Gentry, a curator at large with the Missouri Botanical Garden. While in the field he noticed subtle features that escaped the attention of most collectors and curators, many of whom had prematurely decided that leaves alone, or other vegetative features, would be of little value in on-the-spot species identification. Gentry worked with another fine tropical botanist, Robin Foster, to identify tropical trees and shrubs, including many rare ones, relying only on parts on display every day—leaves, stipules, glands, twigs, bark, thorns, trunk, or exposed roots. The clues to identifying species from these features never appeared in a botanical text. That comprehensive field guide lived in Gentry's head.

“Fortunately, he started to write it all down for the rest of us,”
said Adrian Forsyth, who helped finance the publication of Gentry's magnum opus,
A Field Guide to the Families and Genera of Woody Plants of Northwest South America (Colombia, Ecuador, Peru)
. While the book was in its final editing stage in 1993, Gentry's small plane crashed into a mountainside in western Ecuador. Gentry died at the age of forty-eight, approaching the apex of his career.

Raul Tupayachi is a top Peruvian botanist and the head of George's botany field team. Raul's uncle had worked for Gentry and handed down his inside knowledge. On the second day of field-work, George, Raul, and I hiked on a trail through the Tambopata forest to join the plant collectors. En route, we came upon a carpet of beautiful white blossoms resembling pincushions. I stopped to pick up some of the fragrant
Inga
flowers that had fallen from a tall tree. “You know, Eric,” Raul said, “we've already recorded twenty-two species of
Inga
, and we may have twenty-five. It's the most diverse group of trees in this part of the Amazon.” This was a striking instance of the interplay of rarity and abundance in the tropics.

What's more, many of those
Inga
species could be found on the same hectare. According to standard textbook niche theory, the reason so many tree species, and so many within the same genus, can share the same hectare of forest is that the trees divide up the available niches in microvariations of their environment. These variations occur among many different dimensions, such as light levels; concentrations of nitrogen, iron, or potassium in the soil; and relatively wet or dry spots on the forest floor. Niche theory holds that the ecological separation of many species is a reflection of their differing abilities to use such limiting resources; scientists merely lack the tools to detect all those dimensions just yet. However, a visit I had made to a tropical forest research site in the Lambir Hills of Sarawak, Malaysia, a few years earlier made me wonder if at least one part of niche theory deserved a recall. There in the northern Borneo rain forests, as many as twenty-five species in the genera
Shorea
,
Parashorea
, and
Eugenia
all occurred on the same hectare, like the many species of
Inga
in Amazonian Peru! On
what axis of limiting resources required by plants—nutrients, light, moisture—could trees micromanage their needs in order to live so packed together?

The Lambir Hills site was part of a remarkable experiment to study and map rain forests at an unprecedented scale—fifty hectares, about the size of a small vineyard. The first to do this were ecologist Steve Hubbell and botanist savant Robin Foster. Their landmark study on Barro Colorado Island, Panama, in the early 1980s illustrated the patterns seen everywhere in the tropics: an incredible diversity of tree species—cresting at 125 to 150 species per hectare—with virtually all appearing on their plot at very low densities. Among some species, an adult tree was found only every few hectares. How different this is from a temperate zone forest, where on a single hectare one might find the same number of individual stems but fewer than two dozen tree species. Since then, twenty-three plots of similar size have been intensively studied across equatorial forests from Peru to the Philippines. Each has yielded census results similar to those from Hubbell and Foster's fifty-hectare samples.

Reanalysis of some of these large data plots yielded, in the first decade of this century, yet another paradigm of tree diversity and local rarity. Steve Hubbell proposed a theory of “functional equivalence” in which he suggested that many tropical tree species function as ecological duplicates of one another and the species that grew in a given location in the forest was determined purely by chance. Thus, many plant species exist in tropical forests because they all have an equal chance of occupying spaces that open up when trees fall or die. Rarity in tropical trees and vines is thus in part a function of limited habitat required for successful propagation of a new generation, in particular a shortage of space for seedling survival, such as the infrequent light gaps, or openings, in the forest for species whose seedlings and saplings require intense sunlight. A tree that is the only one of its kind in a forest stand could also be a recent immigrant whose seed somehow managed
to disperse and survive far from its clan. The take-home message of Hubbell's work was that the forest was not in a permanent state of balance. The commoners might fluctuate in number, but species that were extremely rare—say, one individual per fifty hectares—could also just as easily disappear.