La Mojana Canals Built by Communities, Not Rulers

The network of canals and raised fields in Colombia’s La Mojana region has long stood as a textbook case of engineering that only a powerful ruler could have ordered. A study published August 11, 2026, in the journal Communications Sustainability, led by an archaeological team from the University of Cambridge, calculated for the first time how much labor actually went into the work, and came away with a different picture. The project, the data show, was locally driven, collective, and took far less time than assumed.

La Mojana lies in the Caribbean lowlands of Colombia. The ridges and canals cover more than 500,000 hectares, an area comparable to Grand Canyon National Park or three Londons, and are visible from high altitude. The system was in use from the first centuries BCE to the 11th century CE, and fell out of use before the Spanish arrived in the 16th century. The communities that built it were contemporaries of the Maya, yet the region shows no trace of a state-level society. Finds of worked gold and ceramics had previously led researchers to conclude the builders were hierarchical chiefdoms on their way to statehood.

These field systems are so impressive that people assumed they must have been the work of hierarchical chiefs. Our data tell a very different story.

That is how Jasmine Viera, a researcher at Cambridge’s McDonald Institute for Archaeological Research and the paper’s lead author, summed up the finding in a statement from the university. The paper carries the DOI 10.1038/s44458-026-00128-5 and is co-authored by Marcos Martinón-Torres of Cambridge, Fernando Montejo Gaitán of the Colombian Institute of Anthropology and History, and Miguel Carrero-Pazos.

What the Team Actually Measured

Using satellite imagery, the researchers analyzed a 500-hectare sample in the municipality of San Marcos, in the department of Sucre, in detail. Within that sample they catalogued the dimensions and volume of 1,600 ridges and 63 residential platforms. When new, the ridges stood roughly 1.4 meters high, measured from the top of the ridge to the lowest point of the adjacent canal.

On top of that data, the team applied a new method for estimating the labor invested. The calculation factored in the volume of moved earth, local geology, Indigenous work practices, population density, and the technology available at the time. The soil is heavy clay, and metal shovels did not exist in the region back then. Workers may have used wooden tools, though that remains unproven.

How Many Workers Did It Take?

The starting estimate is that around 1,600 people lived on the studied land. The calculation was then run with a workforce of 800 people, a figure the authors stress is a conservative assumption, since ethnographic data show that subsistence farming has traditionally involved men and women alike, and likely children too.

We are confident that one of these canal and ridge systems could have been built in under two months, roughly half a dry season. That is far less labor than has been assumed until now, and it is achievable through cooperation.

That figure applies to one canal-and-ridge system, meaning a single cluster of the kind studied in the sample, not the full 500,000 hectares. The whole network was built up and maintained over centuries by the work of many local communities, and the study offers no timeline for that larger process. The point of the measurement lies elsewhere: a single formation required no coercion and no central command, only agreement among a few family groups within one dry season.

Our findings show that local communities were capable of building these systems on very short timelines and of organizing themselves, rather than relying on a powerful outside force.

A Levee That Breaks and a Ridge That Lets Water Through

During heavy rains, the system held back large volumes of water, then released it gradually toward the fields in drier months. The canals channeled water from permanent bodies of water around the raised areas where houses stood. The result was protection from both flood and drought, more fertile land to farm, and, likely, places to fish.

Modern Colombia has taken the opposite approach. Instead of redirecting water, the state has tried to block it with levees and adapt the land to models built for areas that do not flood. The Cara de Gato levee has breached repeatedly since 2021, damaging surrounding communities and the ecosystem. At other times, the same region suffers severe droughts.

Policymakers have shown interest in these ancient systems, but they lacked the data. Our study is the first to show that these sustainable solutions can be organized locally and collectively.

Montejo Gaitán of the Colombian Institute of Anthropology and History said this. The paper grew out of the REVERSEACTION project, which studies how technologies are maintained in societies without a state. Project lead Martinón-Torres added that arguments for top-down centralized planning, which the data do not support, risk obscuring the agency of pre-Columbian Indigenous populations, and that these societies worked with nature rather than trying to fight it. The calculation code has been released openly, so the same method can be applied to data from other regions.

The limits of the findings are clear. The 500-hectare sample measured in detail amounts to a tenth of a percent of the total area, so the result for the rest of La Mojana is carried over as an estimate, not a measurement. The two-month figure is a model built on an assumed workforce and volume of moved earth, not a construction record. The study does not prove that no central authority existed in the region, only that one was not necessary for this kind of work. It also remains an open question why the system was abandoned, with the working assumption that the lowland population moved toward the mountains. The fact that something worked a thousand years ago is linked to a possible solution for today’s floods, but it does not guarantee one on its own.