David Kottelenberg, Jochem B. Evers, Niels P.R. Anten, Magdalena Rangs and Lammert Bastiaans
Cereal-legume intercrops can suppress weeds more effectively than would be expected from the average of their component sole crops. This enhanced weed suppression has been attributed to selection, where the more competitive species disproportionately determines the weed suppression level of the intercrop. Yet, the conditions under which enhanced weed suppression occurs require further study to inform the design of weed-suppressive intercrops. This study tested whether selection drives weed suppression in cereal-legume intercrops and how competitive asymmetry and spatial design modulate this effect. We conducted four field experiments over three years (2022–2024), varying crop species composition (barley, rye, triticale, wheat combined with faba bean, lupine, pea), cereal:legume mixing ratio (1:1, 3:1, 1:3), and spatial design (within-row mixing vs alternate rows, and 12.5 or 37.5 cm row distance). Crop species with faster canopy development (generally cereals, especially rye) showed stronger weed suppression in sole crops, and intercrop canopy cover resembled the cereal sole crop more than the legume. The harmonic mean, which serves as a prediction model accounting for the selection effect, accurately predicted intercrop weed biomass based on sole-crop weed biomasses. Even at a 1:3 triticale:faba bean ratio, weed suppression was comparable to sole-crop triticale. Close proximity of component species was a prerequisite of selection, as weed suppression benefits were larger in systems with narrower row distances or within-row mixing designs. Our findings establish selection as the dominant weed-suppression mechanism in cereal-legume intercrops and demonstrate how both crop species composition and intercrop configuration influence this effect.