Spatial variability in dissolved organic matter quantity and composition in Midwest reservoirs, USA

Reservoirs are globally ubiquitous and critical for inland carbon (C) cycling, but the coupled influence of land use and hydro-climatic fluctuations on reservoir C source and composition is relatively unexplored. In this study, we used chromophoric dissolved organic matter (CDOM) absorbance and fluorescence excitation and emission matrices modeled by parallel factor analysis to characterize the source and composition of reservoir CDOM. Water samples were collected from reservoirs along a land use gradient in Missouri, USA during a wet and dry summer, respectively. Dissolved organic C (DOC) and CDOM quantity assessed by absorbance at 350 nm (a 350) were significantly (r 2 = 0.5; p < 0.001) correlated, with CDOM quantity per unit DOC increasing with percentage agriculture. Stronger correlations were observed in the wet summer compared to the dry summer. The influence of land use and hydro-climatic variability was further confirmed by redundancy analysis between CDOM compounds, reservoir limnological parameters, watershed characteristics, and precipitation anomaly (Palmer Drought Severity Index). Agricultural reservoirs were shallow and abundant in a 350, aromatic, humic acid- and fulvic acid-like components (C2), chlorophyll-a, and particulate organic matter. These OM signatures in agricultural reservoirs were more pronounced during wet summer conditions, indicating the greater watershed influence. In comparison, CDOM composition in deeper reservoirs was less influenced by watershed influxes. Further, reservoirs with longer residence times experienced internal CDOM processing and an abundance of low-molecular-weight, microbial (C3), and protein-like (C4) CDOM, particularly during drier conditions. Understanding the sensitivity of reservoir CDOM dynamics to watershed land use and variable hydro-climatic conditions will be useful for effective monitoring and resource management.