Insecticides alter wild and managed pollination dynamics on oil crop yield

Global production of Ethiopian rape ( Brassica carinata ) is increasing for biofuel and feed purposes, yet its pollination ecology remains poorly understood despite its similarity to other mass‐flowering Brassica crops. We experimentally investigated how floral resource availability, insecticide use, and managed honey bee hives interact to shape pollination and yield at different scales. We conducted a 2‐year field experiment across 35 sites in the Northern Great Plains, USA, monitoring over 800 focal plants. We quantified pollinator visitation, diversity, movement, flower traits, and seed set across plant, field, and site scales, and assessed how insecticide application altered pollinator behaviour and crop performance. Pollinators increased seed set by 26.8%, with 28 visiting species recorded. Wild pollinators were more efficient than honey bees, contributing 1.01 seeds per visit compared with 0.79. In pesticide‐free fields, interactions between wild and managed pollinators produced non‐additive effects, suggesting context‐dependent competition and facilitation. In contrast, insecticide application disproportionately reduced wild pollinators, while honey bees partially compensated, buffering yield losses. However, wild pollinators consistently delivered higher per‐visit efficiency across spatial scales, although this advantage declined under chemical exposure. Synthesis and applications . Our results show that combining honey bee hives with insecticide use does not automatically enhance crop yield. Pollination outcomes are strongly scale‐ and context‐dependent, with wild pollinators providing disproportionately important services. Agricultural management should therefore prioritize integrated strategies that reduce pesticide pressure, enhance habitat, and support wild pollinator diversity to maintain resilient pollination services in intensive agroecosystems.