Data from: Predicting global organic-matter decomposition in flowing waters

Rivers and streams contribute to the global carbon cycle by decomposing vast quantities of organic matter, but decomposition rates are highly heterogenous, and our understanding of large-scale patterns and drivers of this process remains limited. Using fine-scale climate, land-use, and water-quality data, we generated a predictive model that explains most (81%) of the variation in cellulose-decomposition rates across 514 streams spanning 135° of latitude. Our model reveals key environmental controls of decomposition, including geologic, climatic, and, importantly, anthropogenic attributes. Projections of cellulose decomposition, when combined with genus-level litter-quality attributes, predict leaf-litter-decomposition rates accurately at the global scale (70% of variance explained). High-resolution predictions of cellulose and natural-leaf-litter decomposition provide novel insight into carbon cycling in flowing waters worldwide, including vast, unstudied areas of Earth.