During the nineteenth century a “war of extermination” was being raged against turtles in the Amazon basin, a slaughter so intense that contemporary observers saw the threat of extinction as a real and immediate danger. The reason? To provide fuel and feed the region’s ever-increasing influx of settlers, who ignored the sustainable turtle-harvesting practices of the native Indians and relentlessly pursued the arrau turtle (podocnemis expansa), the most prized of freshwater turtles. These reptiles are typically called the tartaruga grande, when mature reaching as much as to 90 kilos in weight and over one metre in length.

Often referred to as “the beef of the Amazon”, turtle meat leant itself, in the words of the nineteenth-century Brazilian natural historian and explorer Lopes Gonçalves, “to the most exquisite condiments”. Roasting was the most straightforward cooking method – and by some accounts the most delicious – although it was also shockingly cruel, even according to standards of the time: the process involved roasting a turtle alive and serving its in its own shell and juices, nothing whatever being added. Extremely versatile, turtle meat could also be stewed with vegetables or fried in turtle fat. No part of the animal was wasted. Flesh from the sides of the carapace and the breast was used as prime filet steak, while less tender cuts might be partially minced and mixed with farinha – manioc flour, a Brazilian staple; moistened with tucupí  (a vinegar-like liquid made from juice squeezed from manioc), pieces of lean meat were simply grilled on a spit; turtle blood was used to produce a kind of a blood pudding, similar to that made with pig’s blood; entrails were chopped up and made into a soup called sarapatel, a dish that was often boiled in a turtle’s shell used as a kettle; the flippers were cooked in a kettle with tucupí; and large sausages were made of the thick-coated stomach, which was filled with minced-meat and boiled. Prime cuts were also preserved by being fried in turtle fat and then canned, a process that for the meat of one animal required the fat from two animals.

*                     *                     *

The Amazon basin is an immense hydrographic catchment area of over six million square kilometres, some two-thirds of which are in Brazil with the rest located in Bolivia, Colombia, Ecuador, Guyana, Peru, Suriname and Venezuela. The world’s second longest river, the Amazon stretches 6,115 kilometres from west to east, a body of water that in parts is several kilometres wide. In the river’s vast hinterland are tributaries and lakes that during the intense rains of December through May, link together to form vast inland seas. But the “River Sea”, as the Amazon is sometimes called, does not simply bear a visual resemblance to an ocean – there is also a more real, if ancient, relationship to the saltwater bodies of water. Around eighty or ninety million years ago, just before the dinosaurs became extinct, the continent of Gondwanaland began to rupture, eventually leading to the emergence of South America and Africa, separated by a new ocean, the Atlantic. Until the Andes mountain chain emerged around fifteen million years ago, the Amazon actually flowed westwards into the Pacific Ocean. There followed a period when the river had no outlet at all to an ocean, in effect becoming a gigantic lake and swamp ecosystem. Finally, perhaps ten million years ago, the Amazonian lake cut itself eastwards, again transforming itself into a river, this time flowing into the Atlantic Ocean.

A consequence of these geological changes was that marine creatures were left stranded, cut off from their origins in the Pacific Ocean. While most were unable to evolve to survive in the emerging freshwater ecosystems, some made remarkable adaptations. Today, for example, there are over twenty species of stingrays in Amazonian waters; their nearest marine relatives are in the Pacific, far off across the Andes. More remarkable still are manatees – peixe boi (ox fish) in Portuguese – a huge, ungainly, herbaceous sea mammal that colonised both the Amazon and Orinoco river systems. Just how they entered these rivers remains a mystery. One possibility is that they entered from the Atlantic, gradually making their way inland. It is also conceivable that, like stingrays, they colonized from the Pacific, although they are no longer found grazing on that western coast of South America. Dolphins and porpoises also adapted to the emerging freshwater environment, and schools of these friendly creatures still accompany boats as they ply up and down the region’s rivers. And then there are the turtles, some species of which retain clear physical and behavioural characteristics of those of the world’s great oceans.

*                *                *

In the mid-sixteenth century, Europeans began to explore the Amazon basin, initially entering the uncharted region from the west, with Spanish armies crossing the Andes from Quito or Lima. Their aim was to forge routes from the silver mines of Peru that could serve as direct outlets to the Atlantic, providing an alternative to the dangerous trek across the Isthmus of Panama and through the pirate-infested Caribbean Sea. They failed to establish such a route, but by the seventeenth century, European nations and commercial interests had launched a mass of competing claims to the Amazon area, drawn by dreams of untapped riches in this largely unexplored region. According to the Treaty of Tordesillas – the Papal directive signed in 1494 that divided South America between Spain and Portugal – the entire Amazon was Spanish territory. Nevertheless, Portuguese forces were persistent in asserting a presence, and there were also Dutch, English and even Irish trading posts and forts.

Exploration of the Amazon westwards, from the Atlantic, had the advantage of allowing parties to move steadily upstream without first having to cross the Andes. But navigating against the current was much harder work than floating downriver, a feat made all the more time consuming because of forks in the river network, forcing explorers to choose which was the main river and which might simply be an equally impressive-looking tributary that would ultimately lead nowhere. Passing through a completely alien environment, navigating rivers that were previously unknown to all but the native Indians, explorers found themselves delayed by backtracking great distances, and being so far from outposts of European settlement meant that finding food was a constant problem. Some intrepid explorers travelled alone or in small parties, but in the sixteenth to eighteenth centuries expeditions of a thousand or more men frequently set off together to protect themselves from hostile Indians and to allow for at least a reasonable number of survivors to report back on the newly “discovered” territory. Early explorers, such as the Spaniards Francisco Orellana in the 1540s and Lope de Aguirre in the 1560s, maintained their parties by raiding the stores of the Indians whom they encountered, rather than adopting a more far-sighted approach to survival of seeking to win friendships that might also sustain fellow countrymen following in their footsteps. Sustaining large parties, however, was beyond the capability of the indigenous inhabitants, regardless of their treatment, because their economies were wholly subsistence in nature. To survive the months – even years – separated from civilization, the explorers were forced to rely largely on their own foraging skills, collecting fruit, shooting game and catching fish and turtles.

To the Spanish explorer Pedro de Acuña, the lives of the native Indians and the river and forests that he found in the 1630s presented a veritable Eden. Exclaiming that “these barbarians never know what hunger is”, Acuña described a subsistence agricultural system that supplemented fruit and game from the forests. From the rich waters of the rivers and ponds, fish were caught, with enough dried or preserved with salt (which was collected from burned palm leaves) to carry the inhabitants over periods of scarcity and shortage. But what particularly struck Acuña was the capture and husbandry of turtles, providing the Indian communities with year-round availability of fresh meat. For travelling Amazon explorers, however, turtle was not a food source that could be relied on. Unlike on ocean-going vessels, where live turtles were kept on their backs or in tanks to be butchered as needed, the river boats, usually mere canoes, were small and could carry few turtles, so it was necessary to rely on taking their chances in hunting and fishing along the way.

*                     *                     *

The region seemed to promise great abundance and potential wealth to European explorers. From the seventeenth century on, rumours abounded of remarkable natural resources that were believed to lie hidden amidst the Amazon’s dense tropical rainforest. There were healing drugs that were believed to be just awaiting discovery within the forest and huge trees that would enable ships to be built more cheaply than elsewhere. Tobacco, cotton and cacao were all said to be of excellent quality, and it was claimed that sugar, maize, manioc, plantains, pineapples, guavas and coconuts could all be grown with ease. Then there were the creatures living in the vast expanse of the Amazon basin: the exotic game was assumed to be there for the taking in forests, the fish that could be caught so effortlessly in the lakes and rivers, manatees whose skin was strong enough for shields and whose flesh made divine eating, and turtles, the meat of which was considered every bit as good as that of the green turtle, so favoured on Europe’s fashionable dinner tables.

Trading vessels that plied the Madeira and Amazon.

But for Indians as well as for explorers and settlers, these resources were not easy to exploit. The thin soils were unsuited to commercial agriculture, as heavy tropical rains quickly washed away the nutrients, while the semi-nomadic slash-and-burn farming methods of the Indians only provided for subsistence levels of production, not for commercial exploitation. These methods were not suitable for explorers on the move, nor were they adequate for the increasing numbers of settlers. Despite the Amazon’s rich diversity of flora and fauna, the logistical problems of tapping these resources meant that food supply was a constant trial.

As the region’s population increased in the eighteenth and nineteenth centuries, more food was required, heavily dependent on supplies brought in from great distances and at considerable cost. Bread, produced using North American flour, was a great luxury; instead, manioc meal was the main carbohydrate.  Most food – beans, rice as well as manioc – was transported upstream from Belém do Pará, the important river port near the mouth of the Amazon. Milk was rarely available, at any price, and meat was also extremely limited. Even in the few large settlements fresh beef was rarely obtainable, due to lack of pasture.

Intrepid Englishman Henry Walter Bates hunting.

Game was difficult to track and kill even by experienced hunters, and it was hardly seen during the long rainy season, when villages and towns were cut off from their surroundings by the rivers’ and lakes’ floodwaters. Nevertheless, to travellers and residents alike, game was regarded as something that, with luck, could supplement an otherwise tedious diet. Henry Walter Bates, the famous English naturalist and explorer, spent eleven years in the 1840s and 1850s collecting flora and fauna in Amazonia. He found that tapir – a pig-like mammal with a large, elongated snout – was a particular treat when it appeared, albeit only occasionally, on the dinner table. Archie Macintyre, a Scottish missionary who travelled extensively in eastern and central Amazonia in the first decade of the twentieth century, was introduced to boiled jaguar, although he failed to record whether or not he enjoyed this dish.

At least bird life might be expected to be a source of animal protein. In fact relatively few species were sufficiently large, tasty and nourishing to warrant the efforts required of a skilled hunter. The long-legged, ground-nesting curassow, plentiful on the banks of rivers of western Amazonia, was sought after, but to claim just a brace or two could involve an expedition lasting several days. The months of June and July were always greatly anticipated, when large flocks of Cuvier’s toucan – a large bird instantly recognised by its long, black and yellow beak and shiny black and white plumage – made their appearance in the central Amazon. Arriving well fed, these plump birds were shot in such quantities that families could enjoy feasting on stewed and roasted toucan daily for many weeks.

The waters of Amazonia were seemingly a more reliable source of sustenance. “At the Falls of the Madeira”, observed with wonder American missionary James Fletcher in the 1850s, “the traveller will halt and gaze with wonder at the vast multitude of fish of all kinds and sizes – from the huge sea-cow [manatee] to the little sardine.” Fletcher claimed that all you had to do if you wanted a fish was to take a canoe paddle and strike it into the river: it would be impossible not to miss.

Hauling a manatee ashore.
Hauling ashore a manatee.

Despite this apparent natural abundance, commercial fisheries were virtually non-existent, with fish almost always in short supply in the markets of riverside settlements. The limited labour force, most of which was employed extracting rubber and other forest products, hampered the development of fisheries. Fishing activity was largely restricted to the dry season, and what little fish was marketed took the form of pirarucú, dried fish, an article of commerce suited to the region’s sweltering temperatures. Instead of jerked or sundried beef, versatile staples of much of the rest of Brazil, pirarucú was the most readily available source of animal protein in the Amazon’s towns and villages. That said, pirarucú could not be stored indefinitely as, after just a few weeks, it was completely inedible in the prevailing humid conditions.

Visitors frequently identified other river creatures for their potential as food. Four porpoise species were potentially edible, but their slaughter was taboo amongst the indigenous people, its flesh only rarely eaten and its fat never used in lamps as it was believed that doing so would lead to blindness.  But in the wet season peixe boi, the manatee – described by Bates as possessing a bovine-like upper lip, a head and skin of a seal, a body similar to a hippopotamus without legs and a semi-circular flat tail – was keenly sought by Indians and was caught either with strong nets that were placed across inlets, or struck by harpoon, something requiring tremendous patience and skill due to the creature’s acute sense of hearing. While the peixe boi was a constant source of fascination to naturalists, for most visitors it was the manatee’s qualities on the dinner table that merited comment, the flesh said to resemble, in both taste and texture, the finest cuts of veal.

*                    *                    *

The one species that was not only central to the diet of the Amazon’s indigenous peoples but was also to become highly significant to the region’s commerce was the turtle. Even when the rubber boom of the second half of the nineteenth century transformed the Amazon, the exploitation of turtles remained a major economic activity. Well into the twentieth century, thousands of people worked in turtle-related activities, both locals supplying food for themselves and migrant workers employed in turtling gangs for large-scale business enterprise.

Numerous species of turtle were endemic to Amazonia, with undoubtedly the most eye-catching being the matamata, identifiable by its flat and almost triangular head, connected to a long neck that acts like a snorkel, and a carapace (up to 45cm long) and neck (protruding upwards by a further 20cm) with appendages of skin, warts and rough ridges. Looking like a rock encrusted with algae, the matamata favoured calm, slow-moving water, spending most of the time half submerged in the muddy or sandy river bottom, occasionally stretching out its neck to expose its long nostrils to the air. As they were so well hidden, matamata were difficult to catch. They were also much valued, not simply for taste but for the supposed healing properties of their flesh, in particular as a remedy for rheumatism.

The flesh of the tracajá, a small turtle that inhabits the Amazon basin’s large rivers, was also greatly appreciated, but it was difficult to keep in captivity for more than a few days. Some land tortoises, such as the jaboti, were also enjoyed. A particularly fat jaboti would develop an enormous liver, which was regarded as a delicacy: to tenderise and further enlarge the organ, a live animal would be ritually hurled three times high into the air, each time landing on the ground.

But it was the tartaruga grande – the large turtle or, more accurately, the podocnemis expansa – that was the most sought after of these freshwater turtles of the Amazon. The tartaruga grande’s smooth, dark-coloured shell is flat and oval, growing up to a metre long and two-thirds of a metre wide, with females weighing up to fifty kilos. They inhabit large rivers and deep waters, from Venezuela’s Orinoco and Essequibo rivers to the north of the Amazon and, especially, in the Madeira, Purus, Napo, Ucayalli, and Huallaga rivers, the Amazon’s western tributaries. The tartaruga grande bear a striking physical resemblance and taste to sea turtles, in particular to the green turtle. Their breeding and laying habits are also strikingly similar. Whereas most species of river turtles nest in isolation, laying a small clutch of eggs just once a year, the tartaruga grande is much more visible in its nesting habits and produces many more eggs. James Fletcher observed in the 1850s that “the streams [are] fairly speckled with [turtles], paddling their clumsy carcasses up to their native sand-bar”.

The visibility and vulnerability of turtles on beaches made them the favoured target of predators, with eagles, herons and alligators all feasting on hatchlings, jaguars pouncing on the full-grown specimens, joined by vultures and other scavengers, a natural scene that was described as “carnage” by a shocked British witness. Predators also raided nests for eggs, and those hatchings that succeeded against the odds in making it to the water found ravenous fish awaiting them. “The destruction of turtle life is incredible”, wrote James Orton in the 1860s, predicting that the tartaruga grande would be extinct within a matter of few years. But the odds against survival and the devastation that the American naturalist was referring to was not that caused by other animals but was instead inflicted by the most audacious of predators, humans, as they developed a systematic industry and commerce around the exploitation of the tartaruga grande.

*                    *                    *

The Amazon basin’s network of rivers ebbs like a great annual tide. During the course of seven or eight months of constant rain, the crisscrossing rivers and creeks flood vast swathes of land, enabling fish and turtles to scatter over an area far too spread out for fishermen to exploit. Towards the end of June, however, a few clear days herald the start of the short dry season, when the clouds above the green wilderness no longer distil heavy rains on a daily basis and, more important, when the amount of water flowing into the tropical lowlands from the river’s mountain sources is reduced. Men would take out and mend their fishing nets, caulk and repair their canoes, restring their powerful bows and stock up on arrows and harpoons. For villagers, half-starved by the long scarcity of fish and game and the lack of crops from their flooded gardens, the waters’ retreat was a time of hope and excitement.

Tartaruga grande were caught using various methods. In the dry season they would leave the lakes, where they had been during the months of floods, and ascend the river, pursued by Indians and other river-dwellers using either a lance or bow and arrow, methods similar to those used by turtle hunters in many other parts of the world. The harpoon heads generally featured a socket into which fitted the sharpened end of a lance (a shaft made of palm-tree wood or similar) some three to four metres long. Secured to the harpoon head was a long, thin cord of twisted carauá (a plant related to a pineapple) fibre that passed through a loop about the centre of the lance. Upon striking a turtle, the harpoon head would detach and the wooden part of the lance would rise to the surface, acting as a buoy or float.

Bows and arrows, used in a similar fashion to the lance and harpoon, were a more common means of catching turtles. Some two metres long, bows were traditionally fashioned from hardwood, while arrows, onto which feather wings were secured by thread, were typically one and a half metres long and made of arrowgrass, a plant that was ideal for its strength and lightness and which grew on swampy shores. At the end of the arrow was a loose spear. Arrow heads were three or four centimetres long and had small barbs cut into their edges; they were secured by means of wax and thread to a piece of hard wood attached to a socket, into which the shaft of the arrow was loosely fitted. As an arrow struck the shell of its target, the cord attached to the head would uncoil as the wounded turtle dived, carrying with it the inserted weapon, the shaft floating on the water’s surface indicating the exact position where the creature had been struck. Feeling wounded, the turtle would dive rapidly, and the line would unroll itself from the arrow and spiral on the water to enable the turtler in his boat to locate it. When the arrow ceased moving it was understood that the turtle was exhausted, and the line could then be gently drawn in to bring the ailing animal to the surface. As soon as the turtle appeared above water, it would either be harpooned or killed by a blow to its head with a heavy stick or, if wanted alive, secured by being dragged ashore and rolled onto its back.

Turtle-hunting spears.

Striking a turtle with an arrow required even more skill than shooting a bird in flight. Turtles very rarely show their shell above the water; only their nostrils protrude above the surface when they need to breathe. So slight would be the rippling of the water that none but the most experienced eyes could detect movement and the position of a turtle. If an arrow was shot at too narrow an angle it would glance off the turtle’s smooth carapace, and so the missile had to be fired into the air in such a way and with such accuracy to enable it to descend nearly vertically into the turtle’s shell. Often attached to the arrow were small stones pierced with holes, and as the arrow cleaved its way through the air, the stones would make a noise similar to a bird’s cry, so as not to frighten a turtle into taking a sudden dive.

Taking aim at turtles with bows and arrows.

As the dry season progressed, long and wide sand, silt and mud beaches emerged at the bends of the river. During the day alligators would enjoy basking in the early-morning and late-afternoon sunshine, while on moonlit nights turtles would arrive to lay their eggs.  To these beaches were drawn local Indians pursuing the simplest means of catching turtles. As the turtles hauled themselves back into the water after depositing their eggs, the Indians would intercept and immobilise them by turning them over onto their backs, a common method used by sea turtlers. The turtler would then slaughter the disabled creature on the spot or, more usually, drill holes into its carapace, through which ropes would be tied so the turtles could be attached to canoes, and tow them to the home village.

Regardless of the hunting method, the harvesting of turtles was only sustainable when the Indians for their own needs sought animals. The breeding grounds were soon decimated as the taste for turtle grew among the expanding settler population, leading to large-scale commercial exploitation. Edward Mathew, a railway engineer working in the Madeira River area during the 1870s, described a ten-kilometre long beach with rows of turtles eight or ten deep where thousands of turtles were being caught. Not satisfied with capturing mature female turtles, the turtlers often collected hatchlings that they found just emerging from their eggs or scurrying towards the river. They would place the baby turtles in pails of clean water for a day or two before stewing them; the remains of yolk in the entrails apparently contributed flavour to this much-appreciated delicacy.

If well cared for, the tartaruga grande could be kept alive in a pond or wooded palisade for over a year, providing fresh meat during the long wet season. Each year, a single worker would remove hundreds of mature female turtles from the nesting beaches both for himself and to be sold in settlements across the region.  As staple fare in every private house, restaurant and hotel in Manaus, Santarém and even Belém, live turtles met with insatiable demand, with virtually every steamer, schooner and small craft travelling downriver being sure to be laden with tartaruga grande. Even at the rate of capture, demand for these turtles exceeded supply, and prices rose: Henry Bates recorded that in 1850 a medium-sized turtle could easily be bought for nine pence, but by the time he left Brazil in 1859 it was difficult to purchase one for eight or nine shillings. In 1874, the cost in Manaus of a good-sized turtle, suitable for a dinner for fifteen people, was over double of just a few years earlier.

Writing from Barra do Rio Negro in 1851, Richard Spruce noted that turtle was central to every meal. The English naturalist was fortunate in being able to dine each day at the home of an Italian merchant who served excellent food, not least turtle. “I know not how many forms it is cooked in”, wrote Spruce, “but we have never fewer than five dishes of turtle at table, viz. 1. Tartaruga guisada (cooked or stewed); 2. Tartaruga assada a casca (i.e. roasted in the shell); 3. Tartaruga picada (minced); 4. Tartaruga a la rosbif;  5. Sopa de Tartaruga. Of these the picada is the most recherché, but I prefer the guisada.”

Both foreign and Brazilian visitors to Amazonia generally recorded enjoyment of turtle, perhaps feeling a sense of privilege at having as a staple an ingredient that in Europe and the United States was such a luxury, with “nourishing”, “delicious”, “palatable”, “succulent”, “tender”, “wholesome” and “toothsome” some of the words that were used to describe the tartaruga grande. Even so, not everyone shared such views: after just two years in the Amazon, Henry Bates reached a point that he could no longer even bear to smell it cooking. As a consequence of turtle’s “cloying” effect on him and the lack of alternatives, Bates claimed that he often suffered actual hunger, no longer able to digest or even physically swallow the meat.

*                    *                    *

If hunting was not enough to threaten the survival of the tartaruga grande, the species was brought to the very brink of extinction by the indiscriminate harvesting of eggs. With nests that were easily located by humans, theirs were the most commonly consumed of turtle eggs. Indeed the Indians of the Amazon were somehow aware that the tartaruga grande returned to the beach where they were born to lay their eggs, a migration pattern similar to that of the sea turtle and which has only comparatively recently been confirmed by scientists. While certainly magical, this nesting habit was to contribute to the rapid devastation of tartaruga grande. As with harvesting of grown turtles, the Indians traditionally only took eggs for their own, limited, consumption. This situation changed, however, with the arrival of outsiders intent on the exploitation of the Amazon’s natural resources, men who soon learnt the secrets of the nesting beaches.

The tartaruga grande lays her eggs at night on the central and highest part of the beaches, as far as thirty metres from the river’s shore, beginning at dusk and continuing until dawn. The nesting was an awe-inspiring sight, with the concentration of tartaruga grande at least as great as with olive ridley turtles during their so-called “arribada” on Mexican Pacific beaches. In western Amazonia, tartaruga grande in their thousands arrived together on the banks of the nesting beaches, the noise of their shells striking against each other said to be so loud that it could be heard from a great distance. The tartaruga grande’s nesting technique was essentially the same as that of sea turtles, the main difference being the fact that the river turtle had developed sharply clawed flippers that enabled them to excavate a nest in the soft sand at a faster pace. Like those of its distant cousin the green turtles, the nests of tartaruga grande were typically between half-a-metre and a metre deep, with each female laying between 80 and 120 eggs – or occasionally up to 160 eggs – in a single clutch.

In 1872, German engineer Franz Keller-Leuzinger employed by the Brazilian government witnessed nesting on the long Tamanduá beach, finding the sight of so many creatures quite incredible, although he could not help but feel “a sensation of horror and disgust” by the activity, somehow appalled by the turtle’s efforts at proclivity. Keller-Leuzinger pointed out that when swimming, turtles were timid creatures, liable to dive out of sight at the slightest noise, whereas when nesting they were deaf and blind to any noise or danger, and therefore were easily laid on their backs by the fishermen and local rubber tappers, hundreds of whom assembled on these occasions, like birds of prey round dead game, as much to harvest eggs as to capture turtles.

Beaches could literally be covered with nesting turtles, but they were not disturbed until they deposited their eggs. Locals would guard the beaches from intrusion, but they would then lose no time in flocking to the scene as soon as they judged nesting to be complete.

Nesting beach at night: turning turtles.

Like turtle meat, turtle eggs were a source of food to Indians and settlers both, although an inefficient source in terms of the numbers of eggs needed. Turtle eggs constituted a favourite meal and they could be stored for two or three weeks – longer if smoked – with around twenty eggs normally able to satisfy a hungry adult. Although they were sometimes served boiled, the white coagulates only slowly, taking about an hour to begin to solidify, much the same time needed to boil eggs of sea turtles. More commonly, eggs were consumed raw, the yolk blended with farinha (manioc flour) and seasoned. Turtle eggs were also appreciated as a dessert, for which they were usually boiled and served in their skins with sugar. At the beginning of the twentieth century, missionary Archie Macintyre tried to introduce more complicated recipes to a bewildered local Indian population. Producing puddings that reminded him of those that were served back home in Scotland, he blended turtle eggs with corn flour, tinned condensed milk, and sugar. Although Macintyre himself judged the results to be “truly wonderful”, he was unable to persuade either the suspicious Indians or his black Brazilian attendant to sample his creations.

*                    *                    *

Turtle eggs were infinitely more valuable as an article of commerce than as a food source, in particular in the production of manteiga, or “butter”. In Around the World in Eighty Days, an adventure novel first published in 1873, Jules Verne likened turtle-egg manteiga to “the finest Normandy or Brittany butter”, a product that he believed was good enough to spread on bread. In reality though, this so-called “butter” was a foul-tasting concoction with almost no gastronomic qualities. Instead, this form of manteiga was used across the Amazon region as lamp fuel, in the same way that whale oil or leatherback turtle oil was used at the time in other parts of the world.

Lardner Gibbon, a United States naval officer surveying the western Amazon in the early 1850s, reported seeing hundreds of men, and sometimes women, both locals and from as far south as Cuiabá, engaged in gathering turtle eggs. Although most of the gang members were of Indian origin, they were usually led by Brazilians of Portuguese descent. These so-called “oil-men” arrived on the beaches with a fleet of canoes, manned with workers and loaded with provisions of food and equipment such as copper boilers, spades and baskets. From years of experience they knew the precise time that turtles would lay their last eggs and on what beaches. The life was a hard one: the voyage from their home towns or villages was long and arduous, and conditions in the encampments were primitive in the extreme, with only meagre provisions and constant outbreaks of life-threatening fever. Work was carried on at a rapid rate as eggs had to be harvested almost immediately, before young turtles would start to form in the egg, which would impair the quality of the oil.

Nests were invisible to all but the most experienced of eyes, with the river turtles showing considerable cunning in burying their eggs. Whether deliberately or not, they would double back on their tracks and, as if to disguise the place where the eggs are hidden, would make a general upheaval all round before returning to the river. That said, turtles’ skills were no match for an experienced harvester of eggs. By walking barefoot along the known turtle-laying beaches, an experienced collector could easily locate nests, as the foot sinks slightly into the sand. The less skilled wandered up and down the beaches sounding for covered eggs with long, slender sticks, but often destroyed many, despite the tough shell, in their eagerness to outstrip rivals. After locating a nest, the collector would toss the eggs “like potatoes” to other gang members, who placed them in baskets. The eggs were then piled in heaps that resembled “the stacks on cannon-balls seen at a navy-yard”, each as large as six metres in diameter, and of a corresponding height, and numbering several thousand eggs.

Raiding turtle nests on sandbanks.

While still fresh, the eggs were thrown, shell and all, into wooden canoes or other large vessels. Next, they were broken-up with forked sticks and trodden with bare feet, rather as grapes were traditionally crushed for wine, until they formed a gelatinous yellow paste. The thin, soft, eggshells were then pitched out and water was poured into the canoe, with the remaining substance exposed to the sun for several days. The oily matter of the eggs would rise to the surface, to be skimmed off and purified over a fire in copper cauldron. The resulting clarified substance, with an appearance of melted butter, was placed in twelve-litre earthenware jars and dispatched across to illuminate in towns across the vast Amazon region. At best the manteiga retained merely a rancid, fish-like flavour that only the most hardened palates could just about tolerate as cooking oil. But the manteiga was usually impossible to eat, not least because many of the eggs used in the production process had been partially hatched and the baby turtles had decomposed, lending impurities to the substance and giving off a wretched odour that a German traveller likened to “Russian-leather and tanneries [that] renders it thoroughly disgusting to a civilised Christian’s palate at least”.

Conibo Indians producing oil from turtles’ eggs.

The destruction wrought on the tartaruga grande by manteiga production was on a phenomenal scale. One witness calculated that based on a turtle laying a minimum of eighty eggs, to make one twelve-litre jar of manteiga required the eggs of forty turtles, a total of some 3,200 eggs. Another horrified observer referred to a stretch of beach that annually yielded two thousand jars of oil, each jar requiring about twenty-five hundred eggs, in that one locality alone causing the destruction of five million eggs. It was estimated that by the late nineteenth century a total of around 250 million turtle eggs were being destroyed each year simply for the production of manteiga. Oil was used locally, but most of the production was loaded on boats and sent to expanding Amazon towns where there was a seemingly insatiable demand for cooking and lighting.

*                    *                    *

By the early twentieth century the number of eggs harvested was being greatly reduced, partly a consequence of the introduction of gas lighting to towns across Amazonia, but largely due to a marked decrease in the turtle population. As the century progressed, the turtle oil trade remained of some local importance, though much reduced. Even in the previous century as the species was being hunted to extinction, there was a sense among both native Indians and settlers that the taste for turtle was not sustainable. That Brazilian and foreign travellers – not only naturalists but also explorers, engineers and missionaries – commented throughout the nineteenth century on this destruction is evidence of an early awareness that the tartaruga grande was already an endangered species. “It is clear that, with the present procedure, they must rapidly decrease, and that, at no distant date, they will be counted amongst the things of the past”, wrote Franz Keller-Leuzinger, noting as early as 1874 a staggering decrease in the number of turtles nesting within just a few years.

In the 1850s the first limited measures were put in place on a local level to protect turtle stocks by restricting the harvest of eggs. It was observed that turtles were not returning to once-important nesting beaches, either out of a sense of fear or, more likely, due to the lack of mature females. An 1882 provincial law was passed to prohibit what an American visitor considered the “barbarous custom” of taking hatchings: during the four-month dry season, nesting beaches were to be guarded by police, and transgressors were liable to severe financial penalties. This law, however, was completely ignored. Other early attempts at conservation were just as fruitless. By the early twentieth century, attempts were made to introduce quotas for commercial turtle hunting and egg harvesting, with nesting beaches rented from the local government, which held the tenant responsible for the protection of turtles on his beach and which could fine tenants for exceeding their quota. This too was unsuccessful. In the twenty-first century the tartaruga grande maintains a precarious existence, with a place on the IUCN Red List of Endangered Species for being “critically endangered”. No longer actively slaughtered as in previous centuries, the main danger now is the continued destruction of the turtles’ natural habitat. The future of the tartaruga grande is not encouraging.