The Lost Cities of El Dorado

Abstract

For four centuries Europeans searched Amazonia for El Dorado, a city or empire of gold, and found nothing that matched the legend. The conclusion drawn from that failure, that the basin was a pristine wilderness incapable of supporting large or settled societies, has been overturned by the archaeology of the last several decades, which describes substantial populations that engineered their landscape at scale. The most durable signature of that achievement is not gold but soil: the anthropogenic dark earths, patches of fertile, carbon-rich, biologically active ground that persist in the middle of one of the most strongly weathered and nutrient-poor soil landscapes on Earth. This paper assembles the evidence record for those soils as a single body of fact. It treats the Amazonian record (composition, distribution, and formal soil classification), the question of their persistence, the living practice of dark-earth making in West Africa, the demonstration of intentional creation by contemporary Amazonian communities, and the wider global occurrence of comparable soils. It then states plainly what the record does and does not establish. The record establishes that fertile soil of this kind is real, widespread, durable, human-associated, and reproducible. It does not by itself establish the mechanism that makes such a soil self-stabilizing, which is the subject of the companion mechanistic case, and it does not require that the original soils were deliberately engineered, since the question of original intent, though increasingly resolved in favor of intent, is separable from the demonstrated fact that such soils can be and are made on purpose. The paper is the existence proof on which the framework and the mechanistic argument rest.

---

1. Introduction: the search for El Dorado and what was actually there

The legend began with a rite. On a lake in the highlands of what is now Colombia, a chief was said to have covered himself in gold dust and made offerings from a raft, and the Spanish phrase for him, el dorado, the gilded one, slid over time from a man to a city to an empire of gold somewhere in the interior of South America. The search for that empire drove a series of expeditions into Amazonia, beginning with the descent of the river by Francisco de Orellana in 1541 and 1542 and continuing, in one form or another, for centuries. No empire of gold was found. The expeditions returned with reports of large riverside populations and then, as introduced disease ran ahead of the Europeans through the basin, later travelers found a forest that seemed empty. The conclusion that hardened into orthodoxy was that the rainforest was a pristine wilderness, its soils too poor and its environment too hostile to have supported anything more than small and mobile groups.

That orthodoxy has not survived. Archaeological and ecological work over the past several decades describes a pre-Columbian Amazonia of substantial and in places dense populations that cleared, planted, mounded, and engineered their surroundings on a large scale, leaving settlements connected by roads and causeways, earthworks and geoglyphs, and a domesticated and enriched landscape rather than an untouched one (Heckenberger et al., 2003; Clement et al., 2015; Lombardo et al., 2020). The treasure those societies created and left behind was not metal. It was the ground itself. Across the basin, in and around the places where people lived, there are patches of dark, fertile, carbon-rich soil, locally called terra preta de índio, that remain markedly more productive than the surrounding land centuries after the settlements were abandoned. The real El Dorado, on this reading, was the soil, and the anthropogenic dark earths are its surviving record.

This paper assembles that record. Its purpose is narrow and evidentiary: to set out, as a single body of fact, what is known about these soils, so that the broader arguments developed elsewhere have a documented foundation to stand on. The companion framework argues that agriculture is managed extraction from a living resource web and that the sustainable ceiling of a soil can be rebuilt and raised; the companion mechanistic case argues that a durable carbon lattice can carry a soil across a threshold into a self-stabilizing, high-fertility state. Both arguments presuppose that such soils exist and are real. The record presented here is that existence proof. The paper deliberately stops at the boundary of mechanism: it documents what the dark earths are and how they came to be, and it leaves the question of why they behave as they do to the mechanistic case, to which it refers at each point where the temptation to over-read the record arises.

2. The Amazonian record: terra preta de índio

The defining contrast is between the dark earths and the soil around them. Most of the Amazon basin is covered by strongly weathered, acidic, nutrient-poor soils, principally Ferralsols and Acrisols, whose mineral reserves have been leached away over geological time and whose fertility, such as it is, cycles rapidly through the standing vegetation rather than residing in the ground. Against that background the dark earths are an anomaly. They are darker, deeper, and far more fertile, carrying several-fold to many-fold higher concentrations of phosphorus, calcium, and other nutrients, a higher cation exchange capacity, a less acidic and better-buffered pH, and a large stock of pyrogenic carbon, the residue of char, that gives them their color and much of their character (Glaser et al., 2001; Glaser, 2007; Glaser and Birk, 2012). Microbial abundance and diversity are correspondingly elevated relative to the adjacent soils (Kim et al., 2007).

The carbon is central and distinctive. Dark earths carry on the order of twice the organic carbon of neighboring soils, and a substantial part of that carbon is pyrogenic, present as condensed aromatic material that is far more resistant to decay than ordinary plant litter and that persists on a timescale of centuries (Glaser, 2007; Lehmann et al., 2008). The organic matter of dark earth is also spatially complex at very fine scales, organized in close association with mineral surfaces in a way that is characteristic of stabilized soil carbon rather than of fresh residue (Lehmann et al., 2008). The combination of properties is consistent and recognizable enough that the dark earths are now treated as a formal soil category: the international soil classification recognizes an anthropogenic, phosphorus- and carbon-enriched surface horizon, the pretic horizon, and places these soils within the Anthrosol group, the soils whose properties are dominated by human activity (IUSS Working Group WRB, 2015). A soil type defined by human making is, in itself, part of the record.

The dark earths are not a handful of curiosities. They occur at hundreds of sites across the basin, in patches that range from less than a hectare to, in some reports, areas of many hectares, concentrated along the bluffs and waterways where settlement was densest. They are extensive enough to be agriculturally significant where they occur and are still preferentially farmed by local people today for exactly that reason.

3. Persistence, and the limits of what persistence shows

The property that draws the most attention, and that is most easily over-read, is endurance. Dark earths have retained their fertility, their carbon, and their structural distinctness for centuries, and at the older sites for one to two thousand years, through a humid tropical climate that decomposes ordinary organic matter quickly and through long periods after the settlements that built them were abandoned. A soil that holds a high-function state with no one maintaining it is genuinely unusual, and it is the observation that makes the dark earths matter for sustainable agriculture rather than only for archaeology.

Two cautions must travel with this observation, and the paper states them at the outset rather than leaving them implicit. The first concerns mechanism. Endurance of this kind is over-determined: a large stock of recalcitrant pyrogenic carbon and a high, mineral-associated cation exchange capacity would persist for centuries on their own, as passive chemistry, with no need to invoke any self-sustaining biological community. The record therefore establishes that the dark earths last, and that they remain fertile while they last, but it does not by itself establish how they last, and in particular it does not show that a living, self-reinforcing community is actively holding the state in place rather than the soil simply coasting on a durable chemical legacy. The distinction between a passive durable store and a self-replenishing living state, and the behavioral test that would separate them, is the subject of the companion mechanistic case and is not adjudicated here. The record is fully consistent with that mechanistic claim; it is not, on its own, proof of it.

The second caution concerns a particular and frequently repeated observation. Practitioners and some researchers have reported that dark earth can appear to regenerate, re-darkening soil that has been removed and, in the strongest version of the claim, increasing in depth or volume over time. This is an intriguing observation and, if rigorously established, an important one, but it is at present reported rather than quantified under controlled conditions, and it should be treated as a candidate phenomenon awaiting measurement rather than as a documented property of these soils. A mechanism by which such behavior could occur, the downward and lateral transport of colonized char by soil fauna and percolating water, is developed as a testable proposition in the companion framework and mechanistic case; the present paper records the observation and its uncertain status and does not build upon it.

4. The living method: West African Dark Earths

The single most important fact for anyone asking whether these soils can be deliberately made is that they are being made now. In Liberia, Sierra Leone, Ghana, and neighboring parts of West Africa, rural communities create and maintain fertile black soils, locally named and locally valued, by the routine return of kitchen ash, charred residues, bone, and organic waste to gardens and the margins of settlements. These African Dark Earths are a living, observable practice rather than an archaeological inference, and they have been documented by researchers working directly with the communities that make them (Fairhead and Leach, 2009; Frausin et al., 2014). The soils so produced store substantially more organic carbon than the adjacent land, on the order of two to three times as much, with markedly higher pyrogenic carbon content, higher cation exchange capacity, and higher nitrogen and phosphorus, and they are more fertile and more productive than the soils around them (Solomon et al., 2016). Recent carbon-partitioning work in these soils finds that their carbon enrichment is not attributable to pyrogenic material alone, with a large and stabilized non-pyrogenic organic fraction present as well, consistent with the durable carbon protecting co-located humified organic matter (Slocum et al., 2026).

The West African case carries particular evidentiary weight for three reasons. It is contemporary, so the inputs and practices can be watched rather than reconstructed. It is unambiguously human, so it removes any doubt about whether soils of this character can have an anthropogenic origin. And the communities themselves describe the activity in purposeful terms, captured in the title of one study, that the soil was made fertile by deliberate human effort rather than found that way (Frausin et al., 2014). Whether the practice is best described as fully intentional soil engineering or as the predictable result of ordinary household and agricultural habits is itself discussed in the literature (Fraser, Leach and Fairhead, 2014), and that nuance is taken up in Section 7; either way, the soils are made by people, repeatedly and reproducibly, on weathered tropical ground.

5. Contemporary formation in Amazonia

The West African practice establishes that such soils can be made today. A study of the Kuikuro of the Upper Xingu, in the southeastern Amazon, closes the loop by showing that dark earth is made today in Amazonia itself, by descendants of the people who made the ancient soils, and that they make it on purpose. Combining direct observation of daily activity, interviews with Kuikuro residents who are co-authors of the work, and soil analysis of both modern villages and nearby archaeological sites, the study documents deliberate soil-improving practices: sweeping ash and charcoal from hearths, spreading charcoal and ash around the bases of crop trees, and accumulating fish and manioc waste in middens at the edges of the village and along its paths, which compost over years into fertile dark soil used for the most nutrient-demanding crops (Schmidt et al., 2023).

The comparison between the modern and the ancient is the heart of the result. The spatial and compositional patterns of soil enrichment around present-day Kuikuro villages resemble, on a smaller scale, those preserved at the archaeological sites, and the region shows documented cultural continuity from the ancient to the modern inhabitants. From that resemblance and that continuity the authors infer that the ancient dark earths were created by similar practices, and they state the conclusion directly, that the results demonstrate the intentional creation of dark earth (Schmidt et al., 2023). The same work quantifies the carbon involved, estimating thousands of tonnes of soil carbon stored at a single ancient site and lesser but substantial quantities in the middens of a modern village, which connects the record to the present interest in soil as a place to store carbon. For the purposes of this paper, the contemporary Amazonian evidence does what the West African evidence does, and does it in the same landscape as the original soils: it shows the making of fertile dark earth as a present, deliberate, repeatable act.

6. The global picture

The dark earths are often presented as uniquely Amazonian, and the deep, extensive, char-rich soils of the basin are indeed the canonical case, but anthropogenic enrichment of soil is not confined to the tropics. The West African dark earths already extend the phenomenon to a second continent and a different cultural context. Beyond these, soils substantially modified and enriched by long human occupation have been described in temperate Europe and at high latitudes, including sites associated with historical settlement in northern and Arctic regions, broadening the category from a single tropical anomaly to a recurring outcome of sustained human residence on the land (Asare et al., 2022). The details differ with climate, parent material, and culture, and not every enriched anthropogenic soil is a close analogue of Amazonian terra preta, but the recurrence matters: the deliberate or habitual concentration of organic and mineral residues onto inhabited ground produces durably enriched soil across very different environments. That breadth is what makes the dark earths relevant to agriculture generally rather than to one basin in particular.

7. The question of intention

Two distinct questions are often run together under the heading of intentionality, and separating them clarifies both the debate and what the present argument needs from it. The first question is whether the dark earths are anthropogenic at all. The second is whether, granting that they are human-associated, they were deliberately engineered as fertile soil or arose as the incidental byproduct of habitation.

On the first question the weight of evidence is strong and the consensus settled. The dark earths are associated with settlement, contain the artifacts and residues of human occupation, and match in composition the soils that people demonstrably make today. A recent dissent argued the contrary, proposing on the basis of neodymium and strontium isotope ratios and the radiocarbon ages of microscopic charcoal that the enrichment at a well-studied site reflected ancient riverine sediment deposited before human occupation, so that indigenous people would have used naturally fertile patches rather than created them (Silva et al., 2021). That hypothesis was contested directly, with subsequent analysis concluding that the evidence does not support a geogenic origin and confirms an anthropic one (Lombardo et al., 2022), and it sits against the large body of archaeological, compositional, and ethnographic work, including the demonstration of contemporary making, that points the other way. The honest summary is that a natural-origin hypothesis was put forward and did not withstand scrutiny, and that the anthropogenic origin of the dark earths is well supported.

On the second question the picture is less uniform but is moving in one direction. The contemporary evidence shows unambiguous intent: the Kuikuro describe and perform deliberate soil improvement, and West African communities likewise act to make their soil fertile (Frausin et al., 2014; Schmidt et al., 2023). For the ancient soils, some researchers have framed the formation as intentional cultivation and others as the predictable accumulation of refuse and hearth waste around dwellings, a distinction captured in the framing of dark-earth formation as intentional or inevitable (Fraser, Leach and Fairhead, 2014). The contemporary work tilts this debate toward intent by showing that the practices which produce dark earth today are purposeful and that the ancient patterns resemble the modern ones, but the question of how deliberate the earliest soil-making was is not fully closed.

What matters for the present argument is that it does not need to be closed. The framework and the mechanistic case require that fertile soil of this kind is real and that it can be made deliberately, not that every ancient instance was the product of conscious soil engineering. The living West African practice and the documented contemporary Amazonian practice establish the deliberate, repeatable creation of such soils beyond dispute. Whether the first dark earths were engineered or accumulated, the relevant capability, that people can build durable fertile soil on poor ground on purpose, is demonstrated. The debate over original intent is a genuine and interesting archaeological question, and it is reported here as such, but it is not load-bearing for what follows.

8. What the record establishes, and what it does not

It is worth stating the boundaries of this paper as plainly as its content, because the dark earths invite over-reading.

The record establishes the following. Soils of markedly elevated fertility, carbon, and biological activity exist in the middle of strongly weathered, nutrient-poor landscapes. They are widespread, occurring at hundreds of Amazonian sites and on at least two other continents. They are durable, holding their distinctive properties for centuries to millennia. They are human-associated, and their anthropogenic origin has survived direct challenge. And they are reproducible, made by deliberate practice today in both West Africa and Amazonia. Taken together, these are the existence proof that fertile, carbon-rich, durable soil can be built on the poorest ground, and that the capability is not lost or mysterious but living.

The record does not establish the following, and this paper does not claim it. It does not establish the mechanism by which such a soil becomes self-stabilizing, and in particular the centuries-long persistence of the dark earths is not, by itself, evidence that a living community actively maintains them, because a durable carbon and exchange legacy would persist passively for the same duration; the mechanism, and the behavioral test that would distinguish a living state from a passive store, belong to the companion mechanistic case. It does not establish that the original soils were deliberately engineered, a question that remains partly open and that the present argument does not require. And it does not establish, from the depth or the reported self-regeneration of these soils, any claim about soils mining or building themselves from below; those observations are recorded here with their uncertain status and are developed as testable propositions, not asserted as facts.

Within the larger program, the record is the foundation and not the whole structure. The framework supplies the principle, that agriculture is managed extraction and that a soil's sustainable ceiling can be rebuilt; the mechanistic case supplies the explanation, that a durable carbon lattice can carry a soil across a threshold into a self-stabilizing state; and this paper supplies the fact that such soils exist, endure, and can be made. The fact is what makes the principle and the explanation worth pursuing.

9. Conclusion

The expeditions that went looking for El Dorado were not wrong that something extraordinary had been built in Amazonia. They were wrong about what it was made of. The wealth of the pre-Columbian Amazon was not gold but fertility, deliberately concentrated and durably stored in the ground, and the anthropogenic dark earths are what remains of it. They are real, they are widespread, they have lasted for a thousand years and more, they were made by people, and they are still being made today. That is the record, and it is enough to ground everything that follows: if fertile soil of this kind has been built before, endures, and is built again now by hand, then the questions worth asking are how it works and how it could be built deliberately and at scale, which are the questions the companion papers take up. The real El Dorado was always the soil, and the soil can be made.

---

A note on claim calibration

The table states the principal claims of this paper and their evidentiary status, in the discipline used across the program: established facts are stated plainly, contested questions are marked as contested, and the boundary between what the record shows and what the mechanism would have to show is held explicitly.

Claim	Status	Support	What would strengthen or bound it
Dark earths are far more fertile, carbon-rich, and biologically active than the surrounding weathered soils	Established	Glaser et al. (2001); Glaser (2007); Glaser & Birk (2012); Kim et al. (2007)	None needed
Dark earths are a recognized anthropogenic soil class (pretic horizon, Anthrosols)	Established	IUSS Working Group WRB (2015)	None needed
Dark earths are widespread, occurring at hundreds of Amazonian sites	Established	Glaser & Birk (2012); archaeological surveys	None needed
Dark earths persist for centuries to millennia	Established	Glaser (2007); Lehmann et al. (2008)	None needed
Persistence shows a living self-stabilizing community actively maintains the soil	Not claimed here (deferred)	Endurance is over-determined by recalcitrant carbon and passive exchange	The persistence-under-export test, developed in the mechanistic case
Such soils are made deliberately and reproducibly today (West Africa)	Established	Fairhead & Leach (2009); Frausin et al. (2014); Solomon et al. (2016)	None needed
Dark earth is made intentionally today in Amazonia, by descendants of its original makers	Established	Schmidt et al. (2023)	None needed
Dark-earth carbon enrichment is not exclusively pyrogenic; durable carbon stabilizes co-located organic matter	Well supported	Solomon et al. (2016); Slocum et al. (2026)	Direct demonstration of the stabilization mechanism
Comparable anthropogenic dark earths occur beyond the tropics	Established	Asare et al. (2022)	Closer characterization of how far the analogy extends
The dark earths are anthropogenic in origin	Well established	Archaeological and ethnographic record; Lombardo et al. (2022) rebutting the geogenic hypothesis of Silva et al. (2021)	None needed
The original ancient soils were deliberately engineered (rather than incidental)	Contested, trending toward intent; not required by the program	Schmidt et al. (2023) for intent; Fraser, Leach & Fairhead (2014) for the open debate	Direct evidence of ancient intent; not load-bearing either way
Dark earth regenerates its own layer and increases in volume	Reported, not quantified	Practitioner and field observation	Controlled measurement; a candidate mechanism is offered in the companion papers

---

References

The following list is provided for development toward formal submission. Entries marked as verified were checked against primary sources or publisher records during preparation; the remainder are drawn from the established literature and are reliable as to author, year, and venue, but their fine-grained bibliographic details (volume, issue, page, and DOI) should be confirmed against the originals before submission.

Verified during preparation:

• Schmidt, M. J., Goldberg, S. L., Heckenberger, M., Fausto, C., Franchetto, B., Watling, J., Lima, H., Moraes, B., Dorshow, W. B., Toney, J., Kuikuro, Y., Waura, K., Kuikuro, H., Kuikuro, T. W., Kuikuro, T., Kuikuro, Y., Kuikuro, A., Teixeira, W., Rocha, B., … Perron, J. T. (2023). Intentional creation of carbon-rich dark earth soils in the Amazon. Science Advances, 9(38), eadh8499. https://doi.org/10.1126/sciadv.adh8499
• Silva, L. C. R., Corrêa, R. S., Wright, J. L., Bomfim, B., Hendricks, L., Gavin, D. G., Muniz, A. W., Martins, G. C., Motta, A. C. V., Barbosa, J. Z., Melo, V. F., Young, S. D., Broadley, M. R., & Santos, R. V. (2021). A new hypothesis for the origin of Amazonian Dark Earths. Nature Communications, 12, 127. https://doi.org/10.1038/s41467-020-20184-2
• Lombardo, U., Arroyo-Kalin, M., Schmidt, M., Huisman, H., Lima, H. P., et al. (2022). Evidence confirms an anthropic origin of Amazonian Dark Earths. Nature Communications, 13, 3444. https://doi.org/10.1038/s41467-022-31064-2
• Solomon, D., Lehmann, J., Fraser, J. A., Leach, M., Amanor, K., Frausin, V., Kristiansen, S. M., Millimouno, D., & Fairhead, J. (2016). Indigenous African soil enrichment as a climate-smart sustainable agriculture alternative. Frontiers in Ecology and the Environment, 14(2), 71–76. https://doi.org/10.1002/fee.1226
• Slocum, M. P., Barton, R. B., Wagner, S. J., Lehmann, J., Solomon, D., Yeboah, E., & Plante, A. F. (2026). Elevated organic carbon in African Dark Earths is not exclusively attributable to pyrogenic organic matter. Science of the Total Environment, 1016, 181481. https://doi.org/10.1016/j.scitotenv.2026.181481

Established literature (confirm fine-grained details before submission):

• Asare, M. O., Šmejda, L., Horák, J., Holodňák, P., Černý, M., Pavlů, V., & Hejcman, M. (2022). Anthropogenic dark earth: evolution, distribution, and physical and chemical properties. European Journal of Soil Science, 73, e13308.
• Clement, C. R., Denevan, W. M., Heckenberger, M. J., Junqueira, A. B., Neves, E. G., Teixeira, W. G., & Woods, W. I. (2015). The domestication of Amazonia before European conquest. Proceedings of the Royal Society B, 282, 20150813.
• Fairhead, J., & Leach, M. (2009). Amazonian dark earths in Africa? In W. I. Woods et al. (Eds.), Amazonian Dark Earths: Wim Sombroek's Vision (pp. 265–278). Springer.
• Fraser, J. A., Leach, M., & Fairhead, J. (2014). Anthropogenic dark earths in the landscapes of Upper Guinea, West Africa: intentional or inevitable? Annals of the Association of American Geographers, 104(6), 1222–1238.
• Frausin, V., Fraser, J. A., Narmah, W., Lahai, M. K., Winnebah, T. R. A., Fairhead, J., & Leach, M. (2014). "God made the soil, but we made it fertile": gender, knowledge, and practice in the formation and use of African Dark Earths in Liberia and Sierra Leone. Human Ecology, 42, 695–710.
• Glaser, B., Haumaier, L., Guggenberger, G., & Zech, W. (2001). The "Terra Preta" phenomenon: a model for sustainable agriculture in the humid tropics. Naturwissenschaften, 88, 37–41.
• Glaser, B. (2007). Prehistorically modified soils of central Amazonia: a model for sustainable agriculture in the twenty-first century. Philosophical Transactions of the Royal Society B, 362, 187–196.
• Glaser, B., & Birk, J. J. (2012). State of knowledge on properties and genesis of anthropogenic dark earths in Central Amazonia (terra preta de índio). Geochimica et Cosmochimica Acta, 82, 39–51.
• Heckenberger, M. J., Kuikuro, A., Kuikuro, U. T., Russell, J. C., Schmidt, M., Fausto, C., & Franchetto, B. (2003). Amazonia 1492: pristine forest or cultural parkland? Science, 301, 1710–1714.
• IUSS Working Group WRB. (2015). World Reference Base for Soil Resources 2014 (update 2015). FAO. (Pretic horizon; Anthrosols.)
• Kim, J.-S., Sparovek, G., Longo, R. M., De Melo, W. J., & Crowley, D. (2007). Bacterial diversity of terra preta and pristine forest soil from the Western Amazon. Soil Biology and Biochemistry, 39(2), 684–690.
• Lehmann, J., Solomon, D., Kinyangi, J., Dathe, L., Wirick, S., & Jacobsen, C. (2008). Spatial complexity of soil organic matter forms at nanometre scales. Nature Geoscience, 1, 238–242.
• Lombardo, U., Iriarte, J., Hilbert, L., Ruiz-Pérez, J., Capriles, J. M., & Veit, H. (2020). Early Holocene crop cultivation and landscape modification in Amazonia. Nature, 581, 190–193.
• Schmidt, M. J., Py-Daniel, A. R., de Paula Moraes, C., Valle, R. B. M., Caromano, C. F., Texeira, W. G., et al. (2014). Dark earths and the human-built landscape in Amazonia: a widespread pattern of anthrosol formation. Journal of Archaeological Science, 42, 152–165.