Global acceleration of sediment fluxes under post-1950 agricultural intensification

Since the mid-20th century, agricultural intensification and expansion have profoundly altered sediment fluxes from cultivated landscapes to freshwater systems. However, the long-term (>70 years) regional and global imprint of these changes remains poorly quantified. Here, we present a global synthesis of sediment accumulation records from 812 lakes and reservoirs draining agricultural catchments affected by land use worldwide.By compiling sediment accumulation rates (SAR), mass accumulation rates (MAR) and associated geochemical proxies constrained by robust age–depth models, we reconstruct multi-decadal sediment flux trajectories from 1900 to 2010 at global and regional scales, and compared them with global land use statistics. These trajectories provide an integrated proxy for long-term land degradation. Our results reveal a pronounced and sustained post-1950 increase in sediment fluxes, with global MAR and SAR rising by approximately 500% and 350%, respectively. This acceleration is observed across all regions of the world, although its timing and relationship with land-use change differ markedly. In Europe and North America, sediment fluxes increased earlier (1950s–1960s; ≈140%) despite declining agricultural land area, suggesting an anticorrelation with land extent but a strong link to the intensification of agricultural practices. In contrast, Africa, Asia and Latin America exhibited later accelerations (1980s–2000s) that are positively correlated with the agricultural expansion.Together, these findings demonstrate that lakes and reservoirs in agricultural regions worldwide record a coherent sedimentary response to post-1950 agricultural changes, while highlighting regional contrasts in the mechanisms linking land use, land management and sediment delivery. This synthesis provides a long-term reference for evaluating the impact of agricultural intensification on soil degradation and freshwater systems during the Anthropocene.