Part 2 · Where These Ideas Come From
The Milpa and American Polyculture
The milpa is a measured, productive maize-bean-squash system, and eastern North America had its own domesticated polyculture a thousand years before maize arrived.
The "three sisters" gets repeated in permaculture courses as a slogan, a pretty story about corn, beans and squash growing up together. It is also a real farming system with published agronomy, a measurable yield advantage on a defined basis, and known limits. This lesson gives the system as farmers and researchers describe it, then the northern polyculture that existed before maize reached the eastern woodlands.
What the milpa actually is
Milpa comes from Nahuatl, milli for a sown field and pan on top. The system is based on the Mesoamerican triad, maize (Zea mays), bean (Phaseolus species) and squash (Cucurbita species), and it carries many other crops and wild species alongside. There is no single milpa. Farmers run extensive rainfed versions with long fallow, intensive rainfed versions with short fallow, and intensively cropped or irrigated versions with no fallow at all, and the species, the land preparation, the use of fire, the water regime and the fallow length all vary with climate, soil and culture.
The agronomy rests on complementarity. The crops occupy different niches and use different resources, light, water, nutrients and space. One crop supplies something the others use, and the canopy of one changes the microclimate for the rest.
The numbers, on the right basis
The honest measure for a polyculture is the land equivalent ratio: how much monoculture land you would need to grow the same amount of each crop. A ratio of 1.0 means the polyculture is level with separate fields. A 2023 review of milpa agronomy found only eight studies whose design allowed the ratio to be calculated. In all of them the milpa was more productive, with values from 1.08 to 2.89.
A 2014 field study in Annals of Botany grew the crops in monoculture and polyculture under two levels of nitrogen and phosphorus and traced the roots by species. The maize-bean-squash and maize-bean mixtures out-yielded the monocultures on a land-equivalent basis, and the gain came mainly from complementarity between species rather than from one dominant crop. The roots divided the work: maize foraged relatively shallow, common bean used the soil profile more evenly, and squash root placement shifted with phosphorus. Earlier work put the yield advantage at up to around 30 percent over the average of the separate monocultures.
Nutrition, which is where maize alone fails
Maize on its own is not a complete diet. Two things fixed that. The first is the bean. Maize protein is short of the amino acid lysine and long on methionine; beans are the reverse, so eaten together the two make a complete protein. The second is nixtamalization, the alkaline treatment of maize in lime or ash water, which frees the bound niacin and makes the grain both more nutritious and more storable. A 2021 study in PLOS ONE examined the traditional milpa in Oaxaca and its part in household food security and nutritional self-sufficiency.
The limits, stated plainly
The polyculture advantage is a system-level result. It does not mean every crop yields more. In milpa conditions, individual maize yields are typically 1 to 3 tonnes per hectare against monoculture averages around 3.8 tonnes per hectare, and in at least one trial bean yield in the milpa fell below what the monoculture would predict. You accept a lower per-crop yield in exchange for a higher total output per unit land and far more resilience if one crop fails.
The system is also under pressure. Milpa area is falling for several reasons together: low productivity and profitability by market measures, cultural rejection and the migration of young people out of rural areas, and soil degradation from shortening the fallow. At low and moderate intensity the milpa was sustainable for millennia. Pushed to feed dense urban populations without enough fallow, it degrades. It is a good system that can be over-extended.
The northern polyculture that came first
Eastern North America was one of roughly ten independent centers of plant domestication in the world. Incipient agriculture there dates to about 5300 BCE. By about 1800 BCE, people in the woodlands were cultivating several food plants. At the Riverton site in Illinois, the first four domesticated plants were goosefoot (Chenopodium berlandieri), sunflower, marsh elder (Iva annua) and squash, and more species followed. Maize only reached the eastern woodlands after about 200 BCE, more than a thousand years into the local system.
The "three sisters" as a fixed trio is a later and largely northern expression. The eastern woodlands had their own polyculture of seed crops, and maize, beans and squash only became the familiar set after maize moved north.
Exercise
Run a land-equivalent comparison
Plant four strips of the same area: maize alone, beans alone, squash alone, and all three interplanted. Weigh every harvest separately. Compute the land equivalent ratio by dividing each polyculture yield by its monoculture yield and adding the three. A value above 1.0 means the mixture earned its ground.
Exercise
Nixtamalize a pound of maize
Simmer a pound of dried dent or flint maize in water with about a tablespoon of slaked lime or clean wood ash for 30 to 45 minutes, until the skins slip. Rinse, rub off the skins, and grind or cook. This is the step that made maize a staple rather than a hunger crop.
Fieldwork
Name the landscape that fed you this week
Walk your kitchen and write every plant food you ate this week, then mark beside each one where it grew: your ground, a neighbour's, a named region, another continent. Pick three of them and find the closest field still growing them as a mixed planting rather than a sole crop, even a garden, and photograph the layering of tall, climbing and ground crops in one frame. Bring back the list and the photograph with the place and the date noted.
Failure mode
The real failure mode is the fallow. A milpa depends on the soil recovering between cycles, and when population pressure or land shortage shortens the fallow to a year or two, yields drop and the soil degrades, which is happening across parts of Mesoamerica now. If you copy the intercropping and skip the fallow, you have copied the easy half of the system and left out the part that kept it running.
