The Emptied Amazon · source record · paper

Over 20,000 precolonial earthworks in the Southwest Amazonia

Martti Pärssinen; Risto Kalliola; Alceu Ranzi; Eetu Puttonen; Rhuan Carlos Lopes; Pirjo Kristiina Virtanen; Francisco Apurinã; Kalle Ruokolainen; Juha Hyyppä; Antero Kukko; Mariana Campos; Fabio de Novaes; Markku Oinonen; Antonia D. Barbosa; Sanna Saunaluoma; Evandro Ferreira; Nature (open access) · 2026

AI-verified citationSRC-PARSSINEN-2026

identifier / locator: Nature, published 2026-07-29
url: https://www.nature.com/articles/s41586-026-10835-7
verification note: Fetched natively 2026-08-28; all figures cited in this case checked verbatim against the full text, including the funding statement.
reliability notes: Weeks old at case creation; no peer-reviewed technical rebuttal of the population model existed yet — absence of critique is not consensus.
TRANSPARENCY: the paper's funding statement reads 'The project and data collection were funded by a grant from E. Jhong (USA)' — this site's founder. The Athanatos Foundation is acknowledged for administrative assistance only. This case therefore gives the paper's population model its most skeptical defensible reading (see AMZ-C005).

Evidence drawn from this source

Supports· strongAMZ-E001

LiDAR documents 432 earthworks and delimits a 183,000 km² Aquiry region

what the source states: "Our new LiDAR data, obtained from 10 ALS lines, document 432 earthworks, of which only 36 were previously known." LiDAR reveals "five times more earthworks in the deforested areas of block 1 than we had previously found along the same lines using satellite imagery"; the geographic circumscription "produced an area of about 183,000 km²," and the study estimates "approximately 24,000–30,000 earthworks, of which about two thirds belonged to the Aquiry."
limitations: The 24,000–30,000 figure is a block-extrapolation from sampled densities, as the paper states.

Martti Pärssinen (2026)AI-verified citationAMZ-C004

Supports· strongAMZ-E002

The paper's own single-family sensitivity collapses the population headline roughly fivefold

what the source states: "If, as said, about 25–60% (4,165–10,000) of these earthworks were in use in 100–300 ad, and if an average earthwork required 300 people to build and maintain it, we arrive at a total population of about 1.25–3 million." And: "Only if the very unlikely estimation that each approximately 25-m long house foundation (mound) belonged to a single family were true would the population be about 250,000–600,000 people without the invisibility effect of laser penetration, which would raise the population, in this case, to approximately 315,000–760,000."
what we infer: The estimate's span across its own stated assumptions — roughly fivefold — is the honest error bar, and it is what this case displays rather than the headline alone.
limitations: The authors label the single-family reading 'very unlikely'; the sensitivity is theirs, the emphasis is ours.

Martti Pärssinen (2026)AI-verified citationAMZ-C005

Qualifies· strongAMZ-E003

The Aquiry demographic peak is dated AD 100–300 — a millennium before Orellana

what the source states: "The densest overlap of calibrated dates occurs at around 100–300 ad, both chronologically and spatially… our results strongly suggest that between ad 100 and 300 the population of the Aquiry was approximately 1.25–3 million."
what we infer: Superb evidence against ecological-impossibility arguments; weak direct evidence about the world Carvajal saw in 1542. Spending early-dated grandeur as contact-era corroboration is the chronology error this case is built to avoid.

Martti Pärssinen (2026)AI-verified citationAMZ-C004 AMZ-C017