Influence of climate and land use on watershed anthropogenic phosphorus inputs and riverine phosphorus export dynamics: A global analysis

Abstract

Many studies have found predictive relationships between riverine phosphorus (P) export and net anthropogenic P inputs (NAPI) at the watershed scale, but the global or regional extent of these relationships has not been empirically quantified. Herein, we present a data-driven global assessment of the response of riverine total P (TP) fluxes to NAPI based on 358 watersheds. NAPI exhibited high spatial heterogeneity (2–12,085 kg P km−2 yr−1) and was well correlated with riverine TP fluxes. Riverine TP export fractions of NAPI were primarily regulated by NAPI components, hydroclimate factors, and land-use as determined through a random-forest meta-analysis. In watersheds dominated by disturbed land-use (e.g., agricultural and developed lands), runoff emerged as pivotal climate-related factors influencing riverine export fractions of NAPI. In watersheds dominated by natural land-use, runoff, precipitation and temperature were identified as the most critical factors. We developed a mixed-effects meta-regression model (R2 = 0.63–0.70, RMSE = 19–78 %, n = 87–202) to examine the quantitative relationship between riverine TP fluxes and NAPI, which avoids subjectivity in selecting influencing factors and regression forms. The model estimated that legacy P contributed 14–17 % of annual riverine TP fluxes in Chinese watersheds, 25 % in North American watersheds and 11–27 % in European watersheds. Annual NAPI contributions to annual riverine TP flux were 83–86 % in China, 75 % in North America and 73–89 % in Europe. The model forecasted 52–67 %, 69–71 % and 74–77 % reductions in riverine TP fluxes across Chinese, North American, and European watersheds by 2050 under five shared socio-economic pathway scenarios compared to 2010 baseline conditions, respectively. This study provides a straightforward and reliable method for quantifying anthropogenic P input and riverine P export dynamics within an acceptable error range. It provides guidance for developing phosphorus pollution control strategies to counter potential increases in phosphorus inputs due to expected changes in climate and land use.