Lakes store, transport and transform a large share of the dissolved organic matter moving through the global carbon cycle, and they are also a source of greenhouse gases to the atmosphere. Yet the large-scale drivers of dissolved organic matter composition in lakes have remained poorly understood, as has the question of how climate change will modify them. A new synthesis published in Carbon Research addresses that gap by compiling dissolved organic matter data and ambient physico-chemical conditions from 429 lakes spread across five contrasting limnetic regions of China, drawing on more than 2,000 observations extracted from 91 refereed publications, 58 in English and 33 in Chinese, and subsequently scrutinized for consistency.
The study, led by Siyue Li with colleagues including Liuqing Zhang, Isaac R. Santos, Lars J. Tranvik, Liza K. McDonough, Y. Jun Xu, Kaishan Song and Chen Ye, reports that dissolved organic matter and its light-absorbing fraction, chromophoric dissolved organic matter, are more abundant in arid and semi-arid environments than in humid and rainy areas. In lakes of the temperate continental and subtropical monsoonal climate biomes, the chromophoric fraction originates from both the surrounding terrestrial watersheds and indigenous primary production within the lakes. In the plateau and temperate monsoonal climate biomes, by contrast, terrestrial sources dominate. The authors argue that this synthesis covers a wider range of environmental conditions than previous comparable studies, revealing patterns not discernible in work confined to narrower environmental gradients.
The database behind the analysis was built from field sampling conducted between 2000 and 2020, with each studied lake covering an area greater than 0.01 square kilometers. The lakes span China’s five limnetic regions: the East, Northeast, Inner Mongolia-Xinjiang, Yungui and Tibet-Qinghai limnetic regions. The Tibet-Qinghai region alone represents around 50 percent of China’s lake area at 82,232 square kilometers, while the East and Inner Mongolia-Xinjiang regions together contribute 40 percent of total lake area. Water quality parameters included pH, electrical conductivity, dissolved organic carbon, total nitrogen, total phosphorus, chlorophyll a, total suspended matter and Secchi depth. Optical indicators of the chromophoric fraction included the absorption coefficient at 254 nanometers, the specific ultraviolet absorption ratio SUVA254, the absorption ratio at 250 to 365 nanometers, the humification index and the fluorescence index.
To place each lake in its climatic context, the researchers registered elevation, temperature and precipitation at each lake location, drawing altitude data from Google Earth and extracting 11 bioclimatic variables from the WorldClim database. These variables included annual mean temperature, annual temperature range, maximum temperatures in the warmest and coldest months and quarters, annual precipitation, and precipitation in the wettest and driest months and in the warmest and coldest quarters. The literature search was limited to studies published up to 2020, a cutoff the authors set to ensure a consistent, pre-pandemic dataset. They acknowledge that methodological heterogeneity is inherent in compiling independent studies, but note that widely reported optical indices were extracted and standardized, and that large-scale directional patterns are likely detectable despite such variation.
The statistical approach combined a linear mixed effects model, with dissolved organic carbon as the dependent variable and region as a random effect, with three fixed effects models using the absorption coefficient at 254 nanometers, SUVA254 and the fluorescence index as response variables. Fixed effects were selected through backward selection based on the Akaike Information Criterion, and variables with variance inflation factors above 4 were removed to control collinearity. After excluding samples with missing values, 294 complete sample points entered the models for dissolved organic carbon, the 254-nanometer absorption coefficient and SUVA254, while 120 complete points entered the fluorescence index model. Results were reported as percent changes for dissolved organic carbon and the absorption coefficient, and as unit changes for the fluorescence index and SUVA254.
The link between eutrophication and organic matter emerged clearly. Dissolved organic carbon concentration, the 254-nanometer absorption coefficient and the humification index all increased with the trophic state index, while the 250-to-365-nanometer absorption ratio, a proxy for molecular weight, decreased as the trophic state index rose, all at p below 0.01. Dissolved organic carbon, the absorption coefficient and the humification index showed positive linear correlations with total nitrogen and total phosphorus, with r-squared values between 0.12 and 0.32, whereas the molecular weight proxy correlated negatively with both nutrients. Both the humification index and the fluorescence index were positively related to chlorophyll a. These correlations held consistently across all five lake regions, and lake eutrophication coincided with accumulation of dissolved organic matter, its chromophoric fraction and humic substances, suggesting that external loadings of nutrients and organic matter were coupled.
Climate variables showed equally strong statistical associations. The mixed effects model with dissolved organic carbon as the dependent variable explained 59.0 percent of its variation through fixed climatic factors alone, with a further 9.0 percent explained by regional differences. Dissolved organic carbon decreased by 15.5 percent per 1 degree Celsius increase in annual mean temperature, but increased by 29.8 percent per 1 degree Celsius increase in maximum temperature. Every 1 degree Celsius increase in annual temperature range was associated with a 12.2 percent increase in the 254-nanometer absorption coefficient, whose fixed-factor model explained 27.5 percent of variation. SUVA254 decreased by 0.6 units per 1 degree Celsius increase in maximum temperature, in a model where fixed factors explained 43.6 percent of variation. Precipitation effects depended on the timescale: dissolved organic carbon rose by 40 percent for every 100 millimeter increase in precipitation in the wettest month, and the absorption coefficient rose 7.7 percent per 100 millimeters in the warmest quarter, yet both declined slightly as annual precipitation increased.
The authors interpret these opposing precipitation signals through dilution and flushing. Precipitation can flush large amounts of terrestrial organic material into lakes, and extreme weather events such as summer heat waves and anomalous precipitation may stimulate accumulation of dissolved organic matter and its chromophoric fraction. At monthly and quarterly scales, precipitation was positively related to terrestrial organic matter inputs. At the annual scale, however, rainfall appears to exert a dilutional effect, lowering the abundance of lake organic matter in monsoonal climate regions, a pattern consistent with prior work in the Yangtze River showing that dissolved organic matter decreases with increasing water discharge under monsoonal precipitation. The authors caution that their climate relationships derive from long-term WorldClim normals for 1970 to 2000 paired with year-specific lake sampling, so the findings reflect associations with the overarching spatial climate template rather than contemporaneous weather, and the models are not designed to predict responses to annual climate variability.
Spatially, the dataset showed average dissolved organic carbon of 14.8 plus or minus 21.2 milligrams per liter, with a range from 0.2 to 250.2 milligrams per liter, similar on average to a previous study of 560 Swedish lakes at 13.2 plus or minus 6.08 milligrams per liter but spanning a far wider concentration range. The Chinese dataset also ranged widely in total nitrogen from 0.04 to 57.0 milligrams per liter, total phosphorus from 0.001 to 8.4 milligrams per liter, annual precipitation from 23 to 1,999 millimeters, and annual mean temperature from minus 7.5 to 21.6 degrees Celsius, a 29-degree span. The highest dissolved organic carbon and absorption values occurred in the temperate continental and temperate monsoonal climate biomes, SUVA254 was significantly higher in the subtropical and temperate monsoonal biomes, and the molecular weight proxy peaked in the plateau climate biome.
Source fingerprints varied by region. Humification index values above 10, indicating strongly humified, exogenous organic material, characterized the temperate continental and temperate monsoonal biomes, while fluorescence index values below 1.4 point to terrestrial sources and values above 1.9 to in-situ production. The lowest and highest mean fluorescence index values occurred in the plateau and temperate continental biomes respectively. In high-elevation plateau lakes, catchment runoff and atmospheric deposition appear to be the main exogenous sources, and very saline conditions may inhibit microbial activity and suppress autochthonous signatures. Some brackish plateau lakes showed fluorescence index values as low as 0.2 to 0.8, while Himalayan lakes influenced by glacial meltwater exhibited high autochthonous signatures, consistent with earlier findings that glacial-meltwater-dominated lakes can contain more than 80 percent autochthonous chromophoric material. Overall, the synthesis concludes that both eutrophication and climate are important drivers of lake dissolved organic matter in China: concentration and terrestrial character increase toward more eutrophic systems and decrease toward warmer and wetter climates, patterns that differ from previous cross-system comparisons spanning narrower environmental gradients.
Subject of Research: Large-scale drivers of dissolved organic matter composition in lakes across China's climate and nutrient gradients
Article Title: Large-scale patterns in lake dissolved organic matter driven by nutrients and climate
Article References: Li, S., Zhang, L., Santos, I. R., Tranvik, L. J., McDonough, L. K., Xu, Y. J., Song, K., & Ye, C. (2026). Large-scale patterns in lake dissolved organic matter driven by nutrients and climate. Carbon Research, 5(1), Article 47. https://doi.org/10.1007/s44246-026-00279-z
Image Credits: AI Generated
DOI: 10.1007/s44246-026-00279-z
Keywords: dissolved organic matter, chromophoric dissolved organic matter, lakes, eutrophication, climate, carbon cycle, China, CDOM, trophy state index, biogeochemistry, precipitation, water quality
News Source: Bethany Barker. (October 29, 2026). Nutrients and Climate Shape Dissolved Organic Matter in 429 Chinese Lakes. Scienmag.
