Claims checked › Claim
Intercropping makes crops drought-resilient
“Growing a cereal and a legume together makes the crop resilient to drought.”
The productivity gain is real and well measured. The resilience claim rests on a misreading of it. The only multi-site European stability trial found intercrops more stable than sole crops only in productive conditions, and the buffering that exists protects against the legume failing, not against dry weather.
Where it comes from
Meta-analyses showing that mixtures yield 1.2 to 1.3 times the average of their sole crops, and European project communications that translate "higher yield stability" into "resilience to climate change". The two are not the same statement, and the second usually omits its comparator.
Who benefits
Seed merchants selling mixtures, and research projects whose funding case is climate adaptation. Neither is acting in bad faith; the phrase is simply easier to say than the finding.
What the evidence says
The productivity gain is real. A meta-analysis of 934 observations puts the land equivalent ratio at 1.23 (Li et al. 2023, PNAS), and European wheat–pea, barley–pea and wheat–faba trials agree at low input.
The stability gain appears only where conditions are good. Seven European trials in 2017 and the drought year 2018 tested barley–pea and wheat–faba bean mixtures against their sole crops and found the intercrops as stable or less stable at most sites, more stable only in more productive locations, with stability rising as mean yield rose (Weih et al. 2021, Agriculture). For a farm that is already low-yielding and water-limited, that is the wrong way round.
What buffering exists is against the legume failing. In the 2018 Swedish drought, faba bean alone yielded 1.08 t/ha; in a wheat–faba mixture the wheat compensated for most of the loss. The ReMIX project's stability claim is likewise made against the legume sole crop, not the cereal (ReMIX policy brief). So the mixture insures the field against one component collapsing. It does not insure it against drought.
No water saving has been shown. No European rainfed trial we found reports a mixture using less water per hectare than a sole crop. Efficiency per unit of yield is a different measure, and the trials reporting it were irrigated.
What to say instead
"More grain per hectare at low input, and insurance against the legume failing. Not insurance against drought." Ask any stability claim what it is being compared with. The intercropping page grades the productivity and the drought questions separately for exactly this reason.