Abstract: To identify the spatial variations in the relationship between dissolved organic carbon (DOC) and dissolved inorganic carbon (DIC) in global lakes and the major driving factors, we constructed a “hydrological-nutrient” scenario analysis framework using methods including MissForest imputation, classification and regression trees (CART), and generalized additive models (GAMs). Based on global lake DOC and DIC observational data, we compared the correlations between DOC and DIC under different environmental conditions and analyzed their driving factors. The results showed that the CART model identified two primary environmental segmentation points: an annual rainfall of approximately 457 mm and a total nitrogen concentration of about 1.7 mg·L^-1. Global lakes could thus be classified into four scenarios: humid-high nitrogen, humid-low nitrogen, arid-high nitrogen, and arid-low nitrogen. There were significant differences in the relationship between DOC and DIC across these scenarios. In humid-high nitrogen lakes, a weak positive correlation was found, with water temperature near 17.5 ℃ exhibiting nonlinear response characteristics, suggesting that biological processes and water stratification may jointly influence the DOC-DIC relationship under eutrophic conditions. In humid-low nitrogen lakes, the correlation was near zero, likely due to DOC being primarily influenced by soil organic carbon inputs from the watershed, while DIC was more controlled by groundwater recharge, rock weathering, and carbonate equilibrium processes. In arid-high nitrogen lakes, DOC and DIC showed a significant positive correlation, indicating that evaporation concentration and synchronous solute accumulation may be key drivers of their co-varying trends. In arid-low nitrogen lakes, the correlation was weaker, with a response peak in chlorophyll-a concentration near 15 μg·L^-1, suggesting that lower primary productivity may exert phased effects on the dissolved carbon component relationship. Overall, conductivity, pH, water temperature, soil organic carbon, rainfall, and chlorophyll-a were key factors driving the DOC-DIC relationship across different scenarios. The DOC-DIC relationship in global lakes did not exhibit stable synchronous changes but instead demonstrated clear scenario dependence and nonlinear characteristics, necessitating comprehensive analysis in conjunction with hydrological conditions, nutrient status, and lake physicochemical environments.

Key words: lake, dissolved inorganic carbon, dissolved organic carbon, environmental split point, synchronous variation