Beschreibung: Agricultural land management requires strategies to reduce impacts on soil and water resources while maintaining food production. Models that capture the effects of agricultural and conservation practices on soil erosion and sediment delivery can help to address this challenge. Historic records of climatic variability and agricultural change over the last century also offer valuable information for establishing extended baselines against which to evaluate management scenarios. Here, we present an approach that combines centennial-scale reconstructions of climate and agricultural land cover with modelling across four lake catchments in the UK where radiometric dating provides a record of lake sedimentation. We compare simulations using MMF-TWI, a catchment-scale model developed for humid agricultural landscapes that incorporates representation of seasonal variability in vegetation cover, soil water balance, runoff and sediment contributing areas. MMF-TWI produced mean annual sediment exports within 9-20% of sediment core-based records without calibration and using guide parameter values to represent vegetation cover. Simulations of land management scenarios compare upland afforestation and lowland field-scale conservation measures to reconstructed historic baselines. Oak woodland versus conifer afforestation showed similar reductions in mean annual surface runoff (8-16%) compared to current moorland vegetation but a larger reduction in sediment exports (26-46 vs. 4-30%). Riparian woodland buffers reduced upland sediment yields by 15-41%, depending on understorey cover levels, but had only minor effect on surface runoff. Planting of winter cover crops in the lowland arable catchment halved historic sediment exports. Permanent grass margins applied to sets of arable fields across 15% or more of the catchment led to further significant reduction in exports. Our findings show the potential for reducing sediment delivery at the catchment scale with land management interventions. We also demonstrate how MMF-TWI can support hydrologically-informed decision making to better target conservation measures in humid agricultural environments.