Behavior of Lithium Isotopes in the Dogxung Zangbo River Basin: Constraints from Geothermal Water Inputs

Lithium (Li) isotopes have emerged as a powerful geochemical tool for tracing silicate weathering processes. However, the sources of dissolved Li and its isotopic composition in high-altitude rivers, particularly those draining the Tibetan Plateau, still remain debated. This study presents an investigation of dissolved Li concentrations and its isotopic compositions in the Dogxung Zangbo River basin of the Tibetan Plateau through systematic sampling and analysis of river water, geothermal water, lake water and rainwater samples. The results show that river waters in this high-altitude basin exhibit significantly elevated dissolved Li concentrations (1.7–141.9 μg/L, mean 51 μg/L, n = 11) and notably lower δ7Lidis values (+3.2–+11.6‰, mean +7.9‰) compared to most global river waters. Moreover, most data points predominantly cluster around the theoretical dilution line for Na and Cl of geothermal water, and the calculated contributions of geothermal waters to river water vary between 87 and 95%. Therefore, these distinctive geochemical signatures are primarily attributed to substantial inputs from local geothermal waters, which display extremely high Lidis (5840–6269 μg/L, mean 6052 μg/L, n = 3) and characteristically low δ7Lidis values (+2.8–+3.1‰). A pronounced downstream increase in riverine δ7Lidis values is observed following the confluence of geothermal inputs with the mainstream. This spatial evolution reflects the influence of secondary minerals, where the formation and incorporation of secondary minerals into river systems drives continuous Li isotope fractionation. This study shows that in high-altitude river systems, the commonly observed low δ7Lidis values may not only from dissolution of primary silicate rocks or evaporites, but also from contributions of isotopically light Li derived from geothermal sources. Geothermal water input may significantly alter riverine δ7Li values and interfere the information we derive about silicate weathering processes, thus impacting our understanding of the relationship between riverine δ7Li values and silicate weathering intensity.