Where a licence is displayed above, please note the terms and conditions of the licence govern your use of this document. When citing, please reference the published version. Take down policy While the University of Birmingham exercises care and attention in making items available there are rare occasions when an item has been uploaded in error or has been deemed to be commercially or otherwise sensitive.

Riverine fluxes of carbon and inorganic nutrients are increasing in virtually all large permafrost-affected rivers, indicating major shifts in Arctic landscapes. However, it is currently difficult to identify what is causing these changes in nutrient processing and flux because most long-term records of Arctic river chemistry are from small, headwater catchments draining <200 km2 or from large rivers draining >100,000 km2. The interactions of nutrient sources and sinks across these scales are what ultimately control solute flux to the Arctic Ocean. In this context, we performed spatially-distributed sampling of 120 subcatchments nested within three Arctic watersheds spanning alpine, tundra, and glacial-lake landscapes in Alaska. We found that the dominant spatial scales controlling organic carbon and major nutrient concentrations was 3–30 km2, indicating a continuum of diffuse and discrete sourcing and processing dynamics. These patterns were consistent seasonally, suggesting that relatively fine-scale landscape patches drive solute generation in this region of the Arctic. These network-scale empirical frameworks could guide and benchmark future Earth system models seeking to represent lateral and longitudinal solute transport in rapidly changing Arctic landscapes.

Quantifying nutrient attenuation at watershed scales requires long-term water chemistry data, water discharge, and detailed nutrient input chronicles. Consequently, nutrient attenuation estimates are largely limited to long-term research areas or modeling studies, constraining understanding of the ecological characteristics controlling nutrient attenuation and complicating efforts to protect or restore water quality in developed and developing regions. Here, we combined long-term data and a broad suite of biogeochemical parameters from 49 watersheds in northwestern France to test how well instantaneous measurements can predict nitrogen (N) and phosphorus (P) attenuation at watershed scales. We evaluated 13 biogeochemical and 12 hydrological proxies of hydrological flowpaths, residence time, and biogeochemical transformation. Across the 49 watersheds, nutrient attenuation ranged from 88 to −2% for N and 99-96% for P. The strongest biogeochemical proxies of N attenuation were NO − 3 isotopes, rare earth elements (REEs), radon, and turbidity, together explaining 75% of observed variation. For P attenuation, REEs, NO − 3 isotopes, molecular weight of dissolved organic matter, and radon were the strongest proxies, but only explained 27% of observed variation. However, a single hydrological parameter-annual runoff-explained 91% of N attenuation and the relative abundance of schist bedrock explained 56% of P attenuation. We discuss how runoff both controls and reflects watershed hydrology, biogeochemistry, and nutrient attenuation. For example, runoff was correlated with long-term decreases in nutrient concentration, demonstrating how leakier watersheds recover more quickly from nutrient saturation. Given the immense fertilization capacity of modern society, we propose that eutrophication can only be solved by reducing nutrient inputs, though hydrochemical proxies can provide valuable information on where to carry out essential food production activities.