Cascading effects of agricultural practices on weed floral resources, pollinators and pollination: a trait-based approach

Agricultural management affects pollination through multiple, interacting pathways, yet the role of weed-mediated floral resources has received comparatively little attention. Here, we propose a trait-based cascade linking agricultural practices to weed community assembly and floral traits, plant–pollinator trait matching, pollinator communities, and ultimately pollination functions and services. Management shapes this cascade by filtering weed species and ecological strategies and modifying the quantity, quality, accessibility and temporal continuity of floral resources. However, changes in weed communities do not necessarily cascade to pollinators. Building on response–effect trait theory, we identify four mechanisms that may strengthen or buffer cascading effects: response–effect trait coupling, spatial resource substitution temporal floral complementarity and pollinator functional redundancy. Direct effects of farming practices on pollinators may coexist with weed-mediated pathways, while landscape context modulates resource availability across spatial and temporal scales. Importantly, changes in pollinator communities and interactions need not translate into proportional changes in pollination services. Evidence remains uneven: weed research has largely focused on richness, abundance and cover, whereas landscape studies often describe land-use patterns without characterising the floral resources to which pollinators respond. Quantitative floral traits are particularly poorly documented in annual cropping systems. We therefore advocate integrated approaches combining management timing and intensity, repeated seasonal measurements of floral resources, weed and pollinator functional traits, interaction networks and direct measures of pollination outcomes. Beyond agroecosystems, this framework helps predict when environmental changes cascade across interacting communities and when their functional consequences are buffered or amplified.