--- slug: intercropping-polyculture type: pattern summary: "Growing two or more crops in the same field at the same time so the species partition resources rather than competing, buying land-use efficiency without a yield collapse when the pairing and ratio are chosen for the field." created: 2026-06-20 updated: 2026-06-20 related: crop-rotation: relation: complements note: "Crop Rotation diversifies across time on one field; Intercropping diversifies across space at the same time, and the two are often run together." alley-cropping: relation: complements note: "Alley Cropping is the woody-perennial form of the same spatial-partitioning idea; Intercropping is its herbaceous-annual companion." cover-cropping: relation: complements note: "A relay or living-mulch intercrop is the boundary case where the second species is not harvested, which is what Cover Cropping does on purpose." integrated-pest-management: relation: uses note: "Push-pull is an intercropping-based IPM system: a repellent intercrop and a trap border manage stemborers and Striga without synthetic inputs." biological-nitrogen-fixation: relation: uses note: "The legume component is why maize-legume mixes outperform a sole cereal on protein and nitrogen-fertilizer efficiency." soil-food-web: relation: supports note: "Mixed root architectures and residues feed a wider slice of the soil food web than a single crop does." mycorrhizal-networks: relation: supports note: "Belowground complementarity, including shared mycorrhizal hosting, drives part of the land-equivalent-ratio gain." nutrient-balance: relation: measured-by note: "Nutrient Balance and Nitrogen Surplus accounting is how the fertilizer-efficiency claim of a legume mix is verified rather than asserted." food-sovereignty: relation: complements note: "The milpa maize-bean-squash polyculture is a culturally-rooted intercrop with deep ties to Food Sovereignty and Mesoamerican farming communities." single-practice-claim: relation: prevents note: "Intercropping resists the trap of selling one row-ratio result as a universal yield uplift; the land-equivalent ratio is pairing- and site-specific." sustainability-linked-loan: relation: supported-by note: "Sustainability-Linked Loan can carry the equipment and learning costs of moving a sole-crop operation to a mixed-row system." --- # Intercropping and Polyculture > **Pattern** > > A named solution to a recurring problem. *Grow two or more crops in the same field at the same time, arranged so the species partition resources in time, space, or both, rather than competing head-to-head.* *Also known as: mixed cropping, companion planting, polyculture; the named systems milpa and push-pull.* A sole crop uses one root depth, one canopy height, one nutrient demand curve, and one harvest date. Intercropping breaks that uniformity on purpose: a tall cereal and a short legume, a deep root and a shallow one, an early crop and a late one, sharing the same ground at the same time. When the pairing fits, the two crops together capture more light, water, and nitrogen than either would alone on the same area. The operator-grade question is never "is more diversity better." It is narrower: which species pairing, at which row ratio and relative planting date, on which field, buys a land-equivalent ratio above 1 without a mechanization or labor penalty the operation cannot absorb. ## Understand This First - [Crop Rotation](crop-rotation.md) — the time-axis sibling; intercropping is the spatial-and-simultaneous case, and the two combine. - [Alley Cropping](alley-cropping.md) — the woody-perennial form of the same spatial-partitioning idea. - [Biological Nitrogen Fixation](biological-nitrogen-fixation.md) — why a legume in the mix changes the cereal's fertilizer budget. - [Cover Cropping](cover-cropping.md) — the boundary case where the second species is grown but not harvested. ## Context Intercropping is the spatial axis of field diversification. It runs at almost any scale, from a hand-weeded milpa plot to a mechanized strip system on a row-crop operation. The family has four recognizable forms. **Row or strip intercropping** alternates rows or narrow strips of two crops, such as maize and soybean, so each strip is wide enough for equipment but narrow enough that the crops still interact at the edges. **Relay intercropping** seeds the second crop into a standing first crop before harvest, so their growth windows overlap but their harvests do not. **Mixed intercropping** scatters both crops with no row structure, the oldest form and the hardest to mechanize. And the named **polycultures** are specific cultural systems: the Mesoamerican milpa of maize, beans, and squash, and the East African push-pull system of maize intercropped with desmodium and bordered by Napier grass. The unifying mechanism is resource partitioning. Two crops that draw from different soil depths, peak their nitrogen demand at different times, or fill different layers of the canopy compete less for the same resource and waste less of it. A cereal and a grain legume are the canonical pairing because the legume fixes its own nitrogen and the cereal does not, so they are not bidding against each other for the same fertilizer pool. > **Confidence: medium** > > The land-use advantage of well-matched intercrops is well supported by large field-experiment datasets. The size of the advantage is pairing-specific and geographically uneven: the strongest evidence comes from Chinese smallholder maize-legume systems, and the numbers do not transfer untouched to a mechanized temperate row operation. ## Problem A sole crop leaves resources on the table. A single root depth ignores nutrients above and below it. A single canopy lets light hit bare ground between rows for part of the season. A single nitrogen demand curve forces the operator to buy fertilizer the crop cannot all use at once. The field produces one thing, on one schedule, exposed to one set of pests that learn the crop and return for it. The opposite mistake is treating diversity as a free win. An operator can mix two crops on paper and hit any one of four walls in the field: the combine can't harvest them separately, one crop shades the other into a yield loss, the planting and harvest calendars collide, or the labor to manage two crops in one field exceeds the value of the land saved. The honest claim isn't that two crops always beat one. It's that a *matched* pair, at the right ratio and timing, can produce more total output per hectare than either monoculture, while a mismatched pair produces less of both and costs more to manage. ## Forces - **Land-use efficiency pulls against single-crop grain yield.** A mix can produce more total output per hectare than either monoculture while still yielding slightly less grain than the single best sole crop. - **Biological complementarity pulls against mechanization.** Hand-managed mixed cropping captures the most interaction; equipment-scale row and strip intercropping captures less of it but stays harvestable. - **The legume buys nitrogen efficiency but adds management risk.** Fixation pays only when establishment, inoculation, and the cereal-legume ratio all work. - **Narrow-geography evidence pulls against broad-claim ambition.** A pairing validated on Chinese smallholder plots is a hypothesis on a Midwestern operation, not a transferable result. - **Pest suppression can be a co-benefit or the whole point.** In push-pull the intercrop is the pest-control system; elsewhere it is a side effect of diversity. ## Solution **Choose the species pair, the row ratio, and the relative planting date for the specific field, then verify the land-equivalent ratio against the right monoculture baselines rather than a vendor's headline number.** The pattern lives in those three choices, not in the abstract idea of mixing crops. Start with complementarity, not crop names. The pairing should partition something: root depth, canopy height, nitrogen source, or growth window. A cereal and a grain legume partition the nitrogen source, which is why maize-legume is the most-studied mix. A tall crop and a shade-tolerant short crop partition light. An early-maturing crop relay-seeded into a late one partitions the calendar. If two crops compete for the same resource at the same time, the mix loses to both monocultures. Set the ratio and the geometry next. Row ratio (how many rows of crop A per row of crop B), strip width, and plant density decide whether the crops complement or one suppresses the other. Wider strips ease mechanization but reduce the edge interaction that drives the yield advantage; narrower strips capture more interaction but may need specialized or slower equipment. There's no universal ratio. The 2:1 or 3:1 maize-to-legume arrangements common in the literature are starting points to calibrate against the field, not settings to copy. Time the planting deliberately. In relay intercropping the second crop's seeding date relative to the first crop's maturity is the whole design: too early and the crops compete, too late and the second crop has no season left. In simultaneous mixes the relative emergence timing decides which crop establishes dominance. The calendar is as load-bearing as the species choice. Then measure the right thing. The standard metric is the **land equivalent ratio (LER)**: the land area a sole-crop system would need to match the intercrop's yields, summed across both crops. An LER of 1.2 means the intercrop produces on one hectare what monocultures would need 1.2 hectares to produce, a 17 percent land saving. The catch is that LER is only as honest as its denominators: it compares the intercrop's yields against chosen monoculture yields, and a generous choice of comparison yields can inflate it. Treat any single LER figure as a diligence question about which monocultures it was measured against, not as a settled fact. > **💡 Tip** > > Write the intercrop as a plan with the pairing, the row ratio, the two planting dates, the harvest method for each crop, and the monoculture yields you will measure the land-equivalent ratio against. If the harvest-method or the baseline-yield line is blank, the design is not ready for the operating plan. ## How It Plays Out **The maize-legume evidence base.** The largest synthesis to date, Li and colleagues in *PNAS* (2023), pooled 226 field experiments and 934 data records across intercropping systems. It found a mean land equivalent ratio of about 1.23, roughly a 19 percent land saving for the same output. The honest tradeoff sits alongside that headline: intercrops yielded about 4 percent *below* the single most productive monoculture for grain, while matching it for protein, and maize-legume mixes specifically delivered about 10 percent more protein and 11 to 18 percent better nitrogen-fertilizer efficiency than the best sole crop. That is a defensible claim a farmer can plan against and a lender can diligence: a gain in land use and protein output, a small loss in raw grain yield, paid for in part by lower nitrogen inputs. **The geographic-narrowness flag.** A maize and grain-legume meta-analysis in *Agronomy for Sustainable Development* (2022) reported a mean yield gain of about 1.45 tonnes per hectare from intercropping, but the average hides a wide split: China averaged about 2.3 tonnes per hectare of gain while sub-Saharan Africa averaged about 0.90. Much of the strongest intercropping evidence comes from intensively managed Chinese smallholder plots with high labor availability. A pattern validated there does not transfer untouched to a mechanized temperate row operation, where the labor and equipment economics are different. The land-use advantage is real; its magnitude isn't portable. **Push-pull in East Africa.** The push-pull system, developed by the International Centre of Insect Physiology and Ecology (ICIPE) with East African smallholder communities, intercrops maize with desmodium and borders the plot with Napier grass. Desmodium volatiles repel stemborer moths (the push), the Napier border attracts and traps them (the pull), and desmodium root exudates suppress the parasitic Striga weed's seed bank. ICIPE field data report push-pull maize yields around 4 tonnes per hectare against roughly 1.5 in monocrop controls, with the gain attributable to pest and weed suppression rather than added fertilizer. A third-generation drought-tolerant version, evaluated in *Experimental Agriculture*, extended the system to fall armyworm control. Push-pull is the case where the intercrop *is* the pest-management system, not a co-benefit of diversity. **The milpa.** The Mesoamerican milpa intercrops maize, climbing beans, and squash, often called the "three sisters." The maize provides a stalk for the beans to climb, the beans fix nitrogen the maize uses, and the squash's broad leaves shade the soil and suppress weeds. The milpa is traditional ecological knowledge developed over millennia by Indigenous Mesoamerican communities, and it is the cultural and agronomic root of much of what the formal intercropping literature later quantified. It belongs to those communities of origin, not to an anonymous "regenerative wisdom." ## Consequences **Benefits.** A matched intercrop can produce more total output per hectare than either monoculture, raise protein output and nitrogen-fertilizer efficiency through a legume component, suppress pests and weeds (decisively so in push-pull), spread market and weather risk across two crops, and keep more of the canopy and root zone working through the season. It gives a capital allocator something concrete to diligence: a measured land-equivalent ratio against named monoculture baselines says more than a "diversification" label. **Liabilities.** Intercropping adds management load and often fights mechanization. Two crops in one field can mean two planting passes, two harvest methods, separation of mixed grain, and more in-season decisions. The yield advantage is pairing- and site-specific: the wrong ratio or timing produces less of both crops and costs more to manage. The evidence base is geographically uneven, so a result from one region is a hypothesis elsewhere. And the headline land-equivalent ratio is sensitive to which monoculture yields it is measured against, which makes a vendor or program claim a place to ask questions rather than a number to take at face value. The carbon and soil claims need the same restraint applied to other diversification patterns. Mixed roots and residues plausibly feed a wider soil community, but a whole-profile soil-carbon stock increase is an audited outcome, not a default of mixing crops. Treat intercropping as a strong land-use-efficiency and, in the right design, pest-management pattern first. Treat soil carbon as something to measure. > **Disclaimer** > > Pattern descriptions are not site-specific recommendations. Local conditions, > soil type, climate, and regulatory context govern application. ## Sources - Li, Zhang, and colleagues' [2023 *PNAS* synthesis, "The productive performance of intercropping"](https://pmc.ncbi.nlm.nih.gov/articles/PMC9926256/) pooled 226 field experiments and 934 records to report a mean land equivalent ratio of about 1.23, the grain-versus-protein tradeoff, and the nitrogen-fertilizer-efficiency gain of maize-legume mixes. - The maize and grain-legume meta-analysis in [*Agronomy for Sustainable Development* (2022)](https://link.springer.com/article/10.1007/s13593-022-00816-1) reported a mean yield gain near 1.45 tonnes per hectare with a wide China-versus-sub-Saharan-Africa split, the basis for this entry's geographic-narrowness caution. - The [International Centre of Insect Physiology and Ecology's climate-smart push-pull program](https://www.icipe.org/news/icipe-climate-smart-push-pull-benefitting-farmers-east-africa) documents the desmodium and Napier-grass system controlling stemborers and Striga, with push-pull maize yields reported around 4 tonnes per hectare against roughly 1.5 in controls. - The [field evaluation of a third-generation push-pull technology in *Experimental Agriculture* (Cambridge)](https://www.cambridge.org/core/journals/experimental-agriculture/article/field-evaluation-of-a-new-third-generation-pushpull-technology-for-control-of-striga-weed-stemborers-and-fall-armyworm-in-western-kenya/C321C6FB102E7B0FFE6AA11AC70C82AC) extended the system to drought tolerance and fall-armyworm control in western Kenya. - The milpa maize-bean-squash polyculture review in [*Agriculture* (2025), 15(16):1737](https://www.mdpi.com/2077-0472/15/16/1737) surveys the three-sisters system's agronomy and its roots in Indigenous Mesoamerican farming. - The methodological caution on [land-equivalent-ratio robustness in *Experimental Results* (Cambridge)](https://www.cambridge.org/core/journals/experimental-results/article/robustness-of-land-equivalent-ratio-as-a-measure-of-yield-advantage-of-multicrop-systems-over-monocultures/8FD89D9D9AC127C7546AFA35887A4AE0) shows that LER is sensitive to the monoculture yields it is compared against, the basis for treating a single LER figure as a diligence question. - The [2025 ecological-drivers paper in *npj Sustainable Agriculture*](https://www.nature.com/articles/s44264-025-00110-z) examines the ecological mechanisms that determine when intercrop pairings deliver a land-use advantage and when they do not. --- - [Next: Alternate Wetting and Drying Rice](alternate-wetting-drying.md) - [Previous: Crop Rotation](crop-rotation.md)