Mixed cropping of narrow-leaved lupin and oat enhances plant health and yield buffering while maintaining protein yield

Context and objectives: Diversifying arable systems with pulse-cereal mixtures can stabilise yield and reduce external inputs, yet performance data for narrow‑leaved lupin (NLL, Lupinus angustifolius L.)–oat combinations remain scarce, especially for grain quality and protein yield. We investigated whether NLL–oat mixed cropping influences plant health and development, weed suppression, grain and protein yield, and yield quality relative to pure stands. Methods: Within a two‑year randomised complete‑block field experiment at two Swiss sites, the performance of three spring‑NLL varieties (Lunabor, Probor, Jowisz) and three spring‑oat varieties (Bison, Lion, Troll) grown in pure and pairwise mixed stands was compared. Agronomic, health and quality traits were recorded. Linear mixed‑effects models with contrast tests estimated treatment and varietal effects. Results and conclusions: Mixed cropping significantly reduced cereal leaf beetle incidence in oats by 41 % with Lion and by 46 % with Troll. Weed volumes in mixed NLL treatments decreased by up to 87 %. Oat yield in mixtures exceeded the expected performance based on their sowing ratio, while their quality increased with regards to elevated protein content and hectolitre weight. NLL yield remained stable or was slightly reduced, while maintaining their crop quality, including their high protein contents. Lunabor mixed with Troll showed a significantly elevated land use efficiency, while all other combinations remained stable across pure and mixed stands. This buffering also applied to their protein yield. Further, mixed cropping had no effect on attributes such as plant developmental stages, height in late growth stages, and oat tiller numbers. The absence of a mixed cropping response underscores the potential for forecasting mixture performance based on monoculture data. Significance: NLL–oat mixtures improved pest control, weed suppression and yield buffering without compromising total protein yield or land use efficiency. The investigated system offers a low‑input approach to sustainable intensification. Because most structural traits were mixture‑neutral, monoculture data can reliably be used to predict mixture performance.