Site icon Conservation news

Pre-Columbian Amazon settlement primarily ate fish — more sustainable?

  • A study of the Central Amazon’s Hatahara settlement found that, circa 750-1230 AD, 76 percent of the animals people ate were fish and just 4 percent were mammal — very different from American / European prehistoric groups who ate more meat than fish.
  • Thirty-seven different fish taxa were identified in the Hatahara samples, indicating that the people of that time were exploiting a much more diverse spectrum of food species than today, perhaps making their fishing and dietary habits more sustainable.
  • One mystery: just one river turtle genus (Podocnemis) dominated the reptile diet, even though a diversity of turtle taxa can be found in the region.
  • While these results are intriguing, more study is needed at more locations (inland, interfluvial and wetland settlements) to arrive at a regional understanding of available animal resources and the diets of Pre-Columbian Amazon settlements over time.
Excavations at the Hatahara prehistoric site revealed a large diversity of animal bones. The vast majority were fish, followed by reptiles, particularly river turtles. Photo courtesy of Val Moraes

Uncovering the history of human habitation in the Amazon rainforest is challenging. The humid, tropical climate and dynamic forest landscape, where organic materials decay rapidly, can be an enemy of archaeological remains. But researchers have persevered and over the last two decades have gathered evidence that overturns the once commonly-held theory that the ancient Amazon was a sparsely populated, largely “pristine” wilderness.

Modern archaeology instead paints a picture of a far more densely populated and heavily manipulated landscape, occupied by flourishing Pre-Columbian communities.

A recent study has unearthed the first “zooarchaeological” evidence — animal bones that indicate dietary habits — for one such Central Amazonian community, located on a bluff above the confluence of the Amazon and Negro rivers between 750 and 1230 AD.

Excavations at an earth mound at Hatahara, a Central Amazonian settlement on a bluff near the confluence of the Amazon and Negro rivers. Photo courtesy of Val Moraes

The Hatahara site dates back even further, but during this time period the settlement had expanded to cover about 20 hectares (49.4 acres), and its residents produced ceramics, cultivated crops, and modified vegetation cover so extensively that their work indicates “intense landscape transformation related to population growth,” the researchers wrote.

But what did they eat? Sieving 300 liters (79.3 gallons) of excavated sediment, from 4.5 to 9 feet beneath an artificial mound, in a funerary part of the settlement, researchers discovered a surprising answer: fish.

“When I first started this study at Hatahara, I was expecting to find [evidence of the consumption of the] mammals currently hunted in the Amazon, such as monkeys, peccary, spotted paca,” Gabriela Prestes-Carneiro, who led the study, told Mongabay, explaining that mammals were also usually an important part of the subsistence diets of prehistoric American and European peoples. “[To] my surprise, mammals corresponded only to 4 percent of the remains, while fishes and reptiles corresponded to the other 96 percent.”

But it wasn’t just the quantity of fish — totalling 76 percent of the remains — that was a revelation; 37 different fish taxa were identified in the samples, indicating that the people of that time were “exploiting a much more diverse spectrum [of species] than we do now,” said Prestes-Carneiro, of the Paris Natural History Museum. The most common species found was paiche, or arapaima; others included catfish, tambaqui, piranha and swamp eels.

Harvesting so many species on a regular basis would have clearly required a far-ranging, in-depth knowledge of fish habitat, behavior, predation, and migration specific for each species. Today, for example, paiche are caught mostly when they are restricted to lakes during the dry season, when they surface to breathe air. This had led Prestes-Carneiro to suspect that the “Hatahara people knew more about fishes than we, as scientists, know now.”

The Hatahara study sieved 300 liters of Amazon sediment, and recovered the remains of 37 different fish taxa. This photo was taken at Monte Castelo, a different Amazon site. Photo courtesy of Eduardo Góes Neves

Podocnemis, a genus of river turtle, dominated the reptile samples found; animals of this genus can weigh up to 90 kilograms (198.4 pounds). A small mystery: despite a diversity of other turtle taxa in the region, no other turtles appear to have been hunted. What’s more, the remains of all individuals ranged between 30-70 centimeters (11.8-27.6 inches), suggesting that not only did the Hatahara prefer this large genus over others, but that juvenile and very large individuals were preferentially avoided. A finding at a settlement 15 kilometers (9.3 miles) from Hatahara indicates that turtles may have even been kept in corrals.

The contrast between present day fishing patterns in the Amazon, which tend to favor fewer species, and those of hundreds of years ago, which tend to be more diverse, is illuminating, Prestes-Carneiro says. One of the motivations for conducting the study concerned the current debate over sustainable resource use in the past as compared to the present.

Was fishing as undertaken by the Hatahara people sustainable? “Probably yes,” Prestes-Carneiro said. “Hatahara fishermen were able to adapt fishing strategies to different [locales] and climate constraints. Furthermore, we cannot underestimate the protein intake of plants in the [diet] such as palms [and] grains that were exploited by Hatahara inhabitants.”

Some primary Hatahara fish, such as paiche and catfish, are still favored by present-day river communities. However there is a potentially important difference between then and now: the paiche bones recovered by the study suggest that fish weighing from 5-100 kilograms (11-220 pounds) were regularly being caught by the Hatahara people. Such large individuals are rare today; the paiche species is listed as Data Deficient by the IUCN, and Prestes-Carneiro thinks the species may even be at risk of global extinction. The decrease in size and abundance of the species indicates that “at some point they have started to be overhunted,” she said. “It would be interesting to investigate when and how it occurred.

Study leader Gabriela Prestes-Carneiro surrounded by the clutter of her fieldwork at the Laboratory of Zooarchaeology of the Natural History Museum, Paris. Prestes-Carneiro hopes that studying past settlements, and what appear to be their sustainable fishing practices, could be illuminating for present-day sustainable resource use in the Amazon. Photo courtesy of Gabriela Prestes-Carneiro

A dietary preference for fish over mammals could be construed in a number of ways: had mammals been overhunted in this part of the Amazon by 750 AD-1230 AD? Or was there a cultural reason for avoiding mammal species? “We still don’t know,” admitted Prestes-Carneiro. But it does appear certain that protein was plentiful for the Hatahara. “We have to [abandon] the idea that people were fighting against starving in Amazonia. They probably had largely the choice to select what they were going to eat.”

Understanding the extent of fishing across ancient Amazonian communities, and its potential for maintaining long-term settlements, will require a lot more research, Prestes-Carneiro said, cautioning against extrapolating too much from the new study. “It would be very important to expand zooarchaeological studies to different environments, such as inland settlements, interfluvial areas, [and] wetlands in order to construct a regional long-term understanding of animal resources and their relation with human settlements.”

Prestes-Carneiro also thinks that combining archaeological research with contemporary studies could be fruitful: “I do think that, as archaeologists, we should work more on comparing archaeological evidence with traditional / indigenous knowledge in order to understand how human management shaped environments through time,” she said.



Prestes-Carneiro, G., Béarez, P., Bailon, S., Py-Daniel A.R., and Neves E.G. (2015) Subsistence fishery at Hatahara (750–1230 CE), a pre-Columbian central Amazonian village. Journal of Archaeological Science: Reports. In press: doi:10.1016/j.jasrep.2015.10.033