(Opposite) Fruits of the ñejilla palm (Bactris brongniartii) for sale in the Belen market of Iquitos, Peru.

FOUR

Fruits from the Amazon Floodplain

Iquitos, Peru (S3°44´56´´, W73°15´13´´) and environs, 1984–1987

During my three years in Peruvian Amazonia, I spent a lot of time in the Belen market of Iquitos—a twenty-block, byzantine sprawl of stores, stalls, houses, and canoes, reached by steep concrete stairs. During certain times of the year, much of the market is underwater, but residents of this part of Iquitos appeared to have developed two strategies for dealing with the annual flooding of the Ucayali River. They either built their houses on pontoons that would gradu­ally start to float as the floodwaters rose, or they lived in two-story houses and moved upstairs for the duration. The latter required more work because they had to dig the mud out of the first floor every year after the floodwaters receded. Yet inhabitants of both types of houses, surprisingly, would keep their electrical lines connected, live, and dangling during the flooding. They might be floating on the tide or have a basement full of water and mud—but they still had electricity to run their stereos.

I saw a number of unexpected sights in this market. I saw a man carrying a catfish that was bigger than me in a basket on his back. I saw several species of monkeys being grilled over a brazier; bundles of hand-rolled cigarettes, known as mapachos, used for smoking, getting a botfly out of an arm, or casting a spell; grimy glass bottles filled with cane alcohol infused with strips of bark for improving one’s sex life or, conversely, for use as a contraceptive; and stacks of T-shirts made in China printed with cryptic slogans in questionable English. But the main item of interest in the market was the dazzling diversity of native fruits. There were dozens of different types of fruits, and the variety changed from month to month. My colleague in Iquitos, Christine Padoch, who was investigating the production and marketing of fruits while I focused on wild fruit trees, counted over a hundred different species of native fruits being offered for sale in the market. The majority of these fruits were wild harvested.

The fruits of sacha mangua, or “false mango” (Grias peruviana)¹ have a large seed, carrot-colored flesh, and a fuzzy brown covering; they are the size and shape of a large potato. They are eaten peeled, with slices of the stiff, orange flesh separated from the seed, either raw with fariña—granules of toasted manioc (Manihot esculenta)—or grilled. An oil can also be extracted by boiling the fruit; the seed, which contains saponins (foamy glycosides), is used as soap in many rural areas. Nutritional analyses have shown that the fruits are rich in vitamin A.

Sacha mangua trees grow in the understory of seasonally flooded forests in Peruvian Amazonia, where they can form dense aggregations containing two hundred adult trees per hectare. The tree has a distinctive growth form, and it is easy to recognize in the ­forest. Adult trees are rarely taller than 20 meters, and the slender, unbranched stems terminate in a cluster of one-meter-long spatula-shaped leaves. The flowers and fruits grow directly out of the trunk—a bit like a tree designed by Dr. Seuss. Given the ease with which the tree, its seedlings, and its saplings can be recognized, the use and value of its fruits, and the propensity of the species to form dense stands, I selected sacha mangua as one of my study species. I laid out eight plots in a tract of flooded forest located about a kilometer up a blackwater tributary of the Ucayali River and started counting and labeling trees.² Quantifying the production of fruits was relatively straightforward. I selected fifteen adult trees, visited them every week, and counted the number of new fruits along the stem of each tree. I marked the new fruits with paint each week to avoid duplication in subsequent counts.

I noticed during my weekly visits to the sample trees that several of the immature sacha mangua fruits had scratches on them. Each week there would be one more scratch—I counted them—on the fruits. This continued for several weeks until the fruits were mature. At that time, when I visited the tree all that was left was the stalk of the fruit, most of the brown covering, and a little ­orange pulp. The entire seed was gone. My field assistant Umberto finally explained what was going on. Squirrels were also visiting my sample trees each week and making small scratches to see if the fruits were ripe yet, like savvy shoppers testing cantaloupes in the produce aisle. Once they saw that the fruits were mature or, more important, that the seeds were fully formed, they gnawed into them and made off with the seeds.

I continued monitoring my plots for about a year and a half and compiled a complete set of demographic data for the species, consisting of the number of individuals in different size classes, how many died and how fast the rest grew into the next size class, and the size-specific fruit production. I put all these data into a computer model to assess the stability of the population and to perform a series of simulations to estimate how many fruits could be harvested without affecting the long-term stability of the population.³

The sacha mangua population contained over five hundred individuals with a plethora of seedlings and saplings; the species was actively regenerating itself on my site. The adult trees in the population flowered and fruited almost constantly throughout the year, except during the flood peak and for a couple of months thereafter. All together, they produced more than eight thousand fruits per hectare—minus the ones the squirrels made off with. The total annual production of fruits was in excess of 2.5 metric tons per year.

My computer simulations revealed that the sacha mangua population was close to stability, in fact was gradually increasing in size each year. The simulations also showed that collectors could harvest about 80 percent of the fruit produced each year, roughly ­sixty-five hundred fruits, with little impact on the long-term stability of the population. Based on this finding, I became curious to see what level of harvest would overexploit the population. How many fruits would I have to collect to drive the population to extinction? A morbid question, I know, but I was only doing a computer simulation. In my model I simulated a harvest of 95 percent of the fruits each year, ran the program to calculate the resultant size of the population, and continued doing this for eighty years. The results from this rather tedious exercise were not what I had expected.

The population was able to maintain its stability, with birthrates and death rates in the population balanced, on the site for almost thirty years. Levels of seedling establishment were gradually dropping during this period, but dead trees in the larger size classes were continually being replaced by smaller trees. This changed as I continued the simulation. Adult trees started to die, the rate of seedling establishment plummeted, and the seedling and sapling classes began to empty out completely. There were still large adult trees that were producing harvestable quantities of fruit, but the population was recruiting no new individuals each year. This meant that a fruit collector who was only looking at the adult trees might conclude that everything was fine. But a plant ecologist, looking around in vain for seedlings, saplings, and young sacha mangua poles, would immediately appreciate that a crisis was at hand. After about fifty years, even a fruit collector would notice that something was wrong because there would not be much left to harvest. After eighty years, the last sacha mangua tree disappeared from the site—in my computer simulation.

Harvesting too much from a plant population can drive it to extinction. But unless the population is monitored carefully, harvesters might not notice the impact their fruit collecting was having for forty or fifty years. I have a feeling that we are at this stage with a number of the more valuable wild harvested fruits in Amazonia, such as the aguaje palm.

The fruits of the aguaje palm (Mauritia flexuosa)⁴ are highly esteemed in Iquitos. They are eaten raw or processed into a sweet paste for beverages and ice cream, and the pulp and seed kernel yield an edible oil of good quality. Aguaje fruits have a large seed, a thin layer of tasty orange pulp, and a tight, protective covering of small red scales. To get to the sweet part of the fruit, you have to nibble off the scales one by one, spit them out, and then scrape the fleshy orange pulp off the seed with your teeth. It takes some practice to learn how to do this with grace, and eating an aguaje fruit is not something one does in a hurry. They say in Iquitos that the easiest way to get a girl to chat with you for a few minutes is to buy her an aguaje fruit.

Aguaje is a massive palm with large fan-shaped leaves. It is easy to identify in the forest, especially since it frequently occurs in locations that are sparsely populated with other trees. Aguaje is one of the few species that can tolerate the waterlogged conditions in the swamps and permanently flooded backwater forests of Amazonia, and over a million hectares of aguajales (dense aggregations of aguaje palms) are found in Peru alone. These aggregations can contain 150 adults and 400–500 aguaje seedlings, saplings, and pre-reproductive palms per hectare. Not only are wild populations of aguaje exceedingly dense, the large number of individuals in the smaller size classes suggests that the species is regenerating itself in the swamps at a rate sufficient to maintain the population at its current density. The adults in these populations produce a staggering amount of fruit—over six metric tons per hectare in some of the populations that have been studied in Peru.⁵ The aguaje, then, is an extremely well adapted palm that produces large quantities of valuable market fruit and forms high-density aggregations in habitats that are not suited to other forms of land use.

Fruit collectors in Peru intensively exploit the local aguaje forests. The average demand for aguaje fruit in Iquitos has been estimated at about 15 metric tons per day, and the majority of the fruit is wild harvested.⁶ While the intensity of exploitation is not in itself a problem, a key aspect of the biology of the palm coupled with the destructive, short-sighted method collectors currently use to harvest the fruits has created a relationship between people and plant that is decidedly dysfunctional.

Aguaje palms have separate male and female trees. Only about half the trees in the population will produce fruit; the remaining trees produce the pollen required to complete the reproductive process.⁷ Fruit collectors have developed several methods of harvesting the fruits from tropical trees. They pick the fruit up after it falls from the tree, knock it out of the tree with a long pole, climb the tree to collect it (native Amazonians have devised a number of ingenious ways to scale fruit trees using rope belts and slings for their feet), or, as seems to happen with increasing frequency these days, cut the entire tree down. The latter method, unfortunately, is especially common with palms that have large, heavy bunches of fruit and thick trunks, and are extremely tall—like the aguaje palm. Most of the aguaje fruits offered for sale in the market and on the streets of Iquitos were harvested by cutting the palm down. But only the female palms are felled. The aguajales and palm swamps that were once extensive in the vicinity of Iquitos—the swaths of forest that look so photogenic from an airplane—are actually barren stands of male aguaje trees. They contain no fruits, no seeds, and no seedlings, and, before long, they won’t contain any aguaje palms either.

The growth form of a palm can have a major influence on the pattern, intensity, and ultimate sustainability of its use by local communities. A comparison of two closely related palm species, açaí and huasai, offers a useful example. Açaí (Euterpe oleracea)⁸ is a slender, multistemmed palm that forms dense aggregations in the Amazon estuary and along whitewater rivers in eastern Brazil. The species is an important source of palm hearts, and a beverage made from the fruit pulp is a stable component of the regional diet. The huasai palm (Euterpe precatoria)⁹ grows in similar environments in western Amazonia, also occurs in dense, natural population, and also produces palm hearts and edible fruits. The salient difference between the two species is that huasai is a single-stem palm.

Palms, unlike broadleaf trees, have only one primary meristem, and this important collection of cells is located at the apex of the plant where it elongates the stem and makes new leaves.¹⁰ In addition, multistemmed palms have buds at the base of the stem that can be activated to produce additional shoots—which have only a single primary meristem. Single-stemmed palms do not have these buds. When we order palm hearts in our salad we are getting the primary meristem of a palm, and the extraction of palm hearts from a single-stemmed palm kills the plant. Extracting palm hearts from a multistemmed palm kills the shoot on which the heart grows, but triggers the basal buds that will eventually produce new shoots.

Palm hearts can be extracted from several different species, but rarely will palm hearts from huasai appear on the menu. Not enough of these palms grow at a reasonable distance from Iquitos to supply the market. They did at one time, but then a palm heart canning factory was opened in Iquitos. The factory was eventually forced to close for lack of raw material after the huasai palms were completely harvested from local forests. Commercial exploitation of wild populations of a single-stemmed palm for palm hearts is risky business.

Forest farmers in Brazil, on the other hand, have developed sophisticated systems for managing wild populations of açaí palms for both palm hearts and fruits. Closely controlled palm heart extraction is used to stimulate the production of new fruit-bearing shoots that are harvested every year. There are extensive stands of açaí palm throughout the Amazon estuary, and more are being created through enrichment planting and forest refinement to supply the growing market for the fruit. Forest habitats are being sustainably managed and conserved, and local communities are generating additional income for themselves. The multistemmed growth habit of the açaí palm—its ability to re-sprout after cutting—is largely responsible for this.

As noted earlier, the best-known, most widely cited characteristic of Amazonian forests is the large number of different tree species that grow in them. Small tracts of forest may contain several hundred tree species per hectare, and new taxa are being discovered every year. An unavoidable correlate with high species diversity, however, is that the trees of a given species are usually scattered throughout the forest at relatively low densities. Many common trees occur at densities of only one or two individuals per hectare, and some of the rarer species occur at even lower densities.

The scattered distribution and low abundance of many Amazonian tree species represent a dilemma for local collectors. Harvesting takes considerably more time and effort when the trees are spaced far apart, and yields per unit area are likely to be low. Tree populations with only a few adult trees also have a limited capacity for regeneration, are extremely sensitive to the effects of destructive harvesting, and can easily be overexploited. Several commercial forest resources have already been depleted through over­exploitation, and many of the more valuable tree species are currently grown in plantations or small-scale agroforestry systems, not wild harvested, testimony to the difficulties of extracting commercial quantities of useful products from species-rich forests.

All the Amazonian fruit species mentioned so far—camu-camu, sacha mangua, aguaje, huasai, and açaí—form dense, natural populations in habitats where seasonal flooding or permanent swamp conditions preclude the formation of more species-rich mixtures. The oligarchic forests—forests dominated by a small number of species—formed by these species are extremely dense, highly productive, and contain more juveniles than adult trees, suggesting that the populations are actively regenerating themselves.¹¹ In terms of both density and yield, the oligarchic forests of native fruits in Amazonia rival many of the commercial fruit orchards that have been established in the tropics.

It is puzzling, therefore, that the extractive reserves and community forestry initiatives in the region have invariably been based in species-rich forests. It would seem more logical to focus attention on forest ecosystems that have the greatest potential to sustain a program of commercial resource use. In terms of density of useful products, productivity, familiarity to local communities, and relative ease of management, oligarchic forests are much better suited to this form of land use than other types of plant community in Amazonia.

A number of promising native fruits in flooded forests do not form dense populations, however. Uvos (Spondias mombin),¹² for example, usually occurs at densities of five to ten trees per hectare, and finding them takes some work. But it is worth the effort. Uvos trees produce large numbers of bright orange, grape-sized fruits that have an astringent and pleasingly tart flavor. The seeds are fibrous and the fruits float. The trees flower when water levels are low and they drop their fruit into the water during the flooding season. Paddling through a flooded forest and coming upon thousands of florescent orange uvos fruits bobbing up and down in the water, then reaching out and scooping up handfuls to eat is one of the joys of doing ecological research in lowland Amazonia.

Local people use almost every part of the uvos tree. The fresh fruits are eaten raw or made into juice, ice cream, and jellies; the wood is used in light construction; and a tea made from the leaves and bark is considered medicinal. An alcoholic infusion of the bark is said to be an effective contraceptive. I remember that one of the materos (woodsmen) who worked with me would strip off a piece of bark every time we walked by an uvos tree and extol the virtues of the species as a means of birth control.¹³ I would always nod my head, but I maintained a certain degree of skepticism, because I frequently saw his wife and ten children in the village.

The huito tree (Genipa americana)¹⁴ also produces a fruit that is collected and sold in local markets. While the fruits are edible and are used to make juice and ice cream, a somewhat more unusual use is as a body paint. When the juicy liquid from the fruit oxidizes, it stains the skin dark purple or black. Indigenous Amazonian men use the juice to paint their faces before hunting, going into battle, or visiting a girl friend. The dye obtained from huito is essentially permanent, which explains the large number of tourists with black lines on their faces getting on the airplane in Iquitos to go home after a visit to a “traditional” Amazonian village.

In spite of the relatively low densities of uvos and huito trees in flooded forests, both species appear to be actively regenerating and maintaining their populations. The high demand for uvos ice cream and juice in Iquitos drives collectors to harvest large quantities of the fruit, but apparently a sufficient number of seeds escape collection and germinate to produce a new crop of seedlings each year when the river level drops. The demand for huito fruits is much lower. Only a few vendors in the market sell huito, and the juicy exudate from a single fruit is sufficient to paint several faces. Many of the fruits produced by this tree species remain in the forest.

I became friends with a well-known botanist from Missouri, Al Gentry, who would occasionally visit Iquitos, and during one of our dinners together we discussed the idea of trying to put an economic value on all the non-timber resources obtainable from a tropical forest. He had just finished a detailed inventory of a tract of forest near Iquitos, and I had been collecting data on fruit production by forest trees and had been following the market price of different forest products for several years. I still remembered a little from the resource economics class that I took in forestry school, and I knew that if we could quantify the number of fruit-producing trees in the forest and estimate how many fruits each one produced in a year, then put these two numbers together and multiply the result by the market price for each resource, we could come up with an estimate of the value of one year of fruit production in the forest. If we did this for a number of years and brought the combined value of all future harvests back to the present using a discount rate, we could also calculate the net present value, or NPV, of the forest. If the same measurements were made for alternative land use scenarios such as logging or clearing the forest to make a cattle pasture, a comparison of the NPVs should show which form of land use represented the best investment.

We had both seen the bustling commerce in the local market and knew that the collection of forest fruits was an important part of the livelihoods of local communities. We also knew that a forest exploited for non-timber resources seemed to be in considerably better shape than one that had been logged or cleared to make a plantation or a pasture. We enlisted the help of an economist to help us with the calculations, wrote a small opinion piece, and submitted the manuscript to a prestigious scientific journal. It was accepted and published in June 1989.¹⁵

What we found out in our analyses was that the net present value of the tropical forest outside Iquitos, harvested annually for fruits and latex, was a little over $6,000. Selectively logging the timber on a twenty-year rotation yielded a net present value of $490; clearing the forest to make a plantation of fast-growing timber trees or create a pasture both generated net present values of about $3,000—half of the NPV calculated for the sustainable harvest of non-timber forest resources from the forest. We pointed out that harvesting for fruit and latex appeared to be the best way to use a tropical forest. The harvest of fruit and latex yielded higher net revenues than timber, and the two resources could be harvested with considerably less damage to the forest. We closed the piece by suggesting that comparative economics might provide the most convincing justification for the conservation and sustainable use of tropical forests.

Soon after our commentary was published, prominent pieces about our findings appeared in the New York Times and the Washington Post.¹⁶ I was interviewed on National Public Radio, and copies of the article, I was told, were distributed to all the delegates at the United Nations. I got a lot of telephone calls. And, of course, we also received the requisite criticisms of the piece because our study forest was located so close to Iquitos, and the discount rate could have been different: if the collectors had harvested all the fruit from all the forests the market value of these products would drop to zero, and different forests with different floristic compositions would surely exhibit a different net present value. I do not dispute any of these criticisms, and acknowledge that we could have done things differently. That said, the collection of forest fruit does generate a significant amount of income to local communities in Peruvian Amazonia. Non-timber forest products, like fruits, can be harvested from tropical forests with less ecological impact than harvesting timber, and comparative economics always provides the most convincing explanation for the land-use decisions that human beings make. That a tropical ecologist and a botanist were the first people to bring up these issues paints a pretty clear picture of the state of non-timber resource and sustainable forest use thirty years ago. A lot of people still think I’m an economist.