Vanderbilt Graduate Student Provides Accessible Data for Inland Aquatic Ecosystems

June 23, 2026

By Joanne Liu, Torrey Pines High School Student and SDSC Communications Intern 

When dusk descends, human-built environments light up. Cities, roads and residential areas emit artificial light throughout the night, raising important questions about how light pollution affects nearby ecosystems. In freshwater environments, these effects are still not fully understood, particularly in small lakes where biological and geochemical processes can change over daily and seasonal timescales. Standing at the shores of murky waters, each ripple beckons for a dive towards deeper research.

With support from the National Science Foundation (NSF) DeCODER program, Mahir Tajwar, a doctoral student at Vanderbilt University studying aquatic biogeochemistry, developed a FAIR and reproducible framework for organizing, modeling, visualizing and sharing lake ecosystem metabolism data. His project focused on Stephens Lake, an oligotrophic headwater lake in Franklin, Tennessee, and used seasonal field measurements, open-source code and standardized metadata to make freshwater ecosystem data more accessible and reusable.

Caption: Deploying a water-quality profiler at Stephens Lake, Tennessee. High-frequency measurements of dissolved oxygen, temperature, pH and light availability were used to support lake metabolism modeling and depth-resolved visualization of ecosystem dynamics. 

“Our study site, Stephens Lake in middle Tennessee, served as a natural outdoor laboratory because of its shallow depth, carbonate-rich setting, and clear diel biogeochemical signals,” Tajwar said. “These characteristics make it possible to examine how light, temperature, dissolved oxygen and carbon cycling interact over short timescales.”

Tajwar collected high-frequency field data during seasonal 48-hour monitoring campaigns in Fall 2024, Winter 2025, Spring 2025 and Summer 2025. Measurements included dissolved oxygen, temperature, pH, photosynthetically active radiation (PAR), wind speed and related water chemistry parameters. This collected data was then used to model lake ecosystem metabolism, including gross primary production (GPP), ecosystem respiration (ER) and net ecosystem production (NEP).

Caption: Nighttime water sampling at Stephens Lake, Tennessee during a seasonal 48-hour monitoring campaign. Field measurements and water samples were collected across diel cycles to evaluate dissolved oxygen dynamics, lake metabolism, and freshwater biogeochemical processes. 

Tajwar applied schema.org metadata standards to the collected data using JSON-LD, allowing him to properly structure, document and improve the discoverability and reuse of environmental datasets. In addition, he developed a reusable Jupyter Notebook code that reads the data and generates depth-resolved three-dimensional visualizations of dissolved oxygen and photosynthetically active radiation (PAR), consistent with the FAIR data framework promoted by DeCODER.

The project also incorporated a customized ecosystem metabolism modeling workflow based on the LakeMetabolizer R package. This workflow was adapted to estimate high-resolution diel patterns in gross primary production (GPP), ecosystem respiration (ER) and net ecosystem production (NEP), providing greater insight into short-term metabolic variability across seasons. Together, the metadata framework, modeling tools and visualization products provide a reproducible approach for organizing, analyzing and sharing freshwater ecosystem data. 

“These deliverables provide a reproducible framework for future lake metabolism studies,” Tajwar said. “By combining open-source code, standardized metadata and accessible data products, this project helps make inland aquatic ecosystem data more findable, interpretable and reusable.” .

Tajwar presented related findings at the American Geophysical Union (AGU) 2025 Annual Meeting. His AGU poster and abstract, titled “Coupled Carbonate Equilibrium and Metabolic Fractionation Drive δ¹³C-DIC Variability in Freshwater Ecosystems,” were also  published in the ESS Open Archive.