1. Small shallow ponds are widespread but understudied water bodies in agricultural landscapes. Agricultural run-off (ARO) transports pesticides and nutrients into adjacent aquatic ecosystems where they occur dissolved in the water column or are bound to sediments. Consequently, aquatic communities are affected by ARO via different exposure pathways. We hypothesize that sediment-bound ARO mainly affects submerged rooted macrophytes, while phytoplankton and periphyton are more prone to ARO in water. These primary producers compete for resources resulting in a regime shift between alternative stable states of macrophyte or phytoplankton dominance. We hypothesize that warming increases nutrient release from sediments and thereby facilitates the occurrence of phytoplankton dominance. 2. Using a full-factorial microcosm design, we exposed aquatic primary producers to either sediment or water application of a mixture of common pesticides (terbuthylazine, pirimicarb, tebuconazole and copper) and nitrate at two concentrations and two temperatures (22°C and 26°C) for 4 weeks. Initial and final concentrations of pesticides and nitrate, final biomass of macrophytes, periphyton and phytoplankton, pesticide accumulation in macrophytes and changes in carbon, nitrogen and phosphorus content and selected exoenzyme activities in the sediment were measured. 3. We found lower final macrophyte biomass for both ARO treatments compared to controls, indicating a prevalence of negative effects by herbicides and competition for light with other phototrophs. In contrast, phytoplankton and periphyton biomass increased, but only when exposed to ARO via the water column, indicating a prevalence of positive effects by nutrient supply. Microbial carbon and nutrient cycling in sediments was not affected by ARO. Higher temperature mitigated ARO-related effects on macrophytes under sediment exposure. 4. Synthesis and application. ARO poses a strong risk of submerged macrophyte loss and establishment of turbid conditions with phytoplankton dominance in aquatic ecosystems. In conclusion, exposure pathways as well as indirect and interacting effects of multiple stressors need to be considered when designing appropriate mitigation measures. Under climate change, we suggest to prioritize local measures as buffer strips a reduced use of pesticides and fertilizers, and sediment removal as appropriate measures to protect these vulnerable but widespread aquatic systems, which are highly relevant for biodiversity in agricultural landscapes.
