FICKLE FLUXES: CARBON BIOGEOCHEMISTRY ACROSS SPATIO-TEMPORAL SCALES

In lakes and estuaries, a multitude of physical, chemical and biological factors interact to drive large spatial and temporal variability in carbon distributions. These factors also preserve large concentration gradients between the water and the atmosphere, supporting globally-significant carbon exchange between these systems and the atmospheric carbon dioxide reservoir. However, these systems are functionally diverse, and this between-system variability causes considerable uncertainty when estimating their impact on the global carbon cycle. The goal of this dissertation research was to better constrain the factors leading to this variability in a set of tractable and representative systems. First, a well-resolved two-year dataset is used investigate the watershed-scale drivers of air-water carbon dioxide exchange in adjacent North Carolina estuaries. Variable inputs of fresh water are linked with annual-scale variability, and we demonstrate that freshwater residence time exerts a strong control on carbon biogeochemistry that is generalizable across estuarine morphologies. Next, we compare the response of the same two estuaries to historic flooding in 2016, and show that massive freshwater inputs drive large carbon dioxide emissions, despite elevated primary production. The annual carbon budget of both estuaries was significantly impacted by flooding, and we argue that these events should be accounted for in regionally and globally-scaled carbon cycles. The physics driving gas exchange in estuaries is complex, and introduces a large uncertainty into modeled carbon dioxide emissions from these systems. We combine a direct, eddy covariance determination of carbon dioxide efflux with in-situ measurements of gas concentration in an effort to better refine gas transfer parameterizations in estuaries. Due in part to large background variability in these systems, our parameterization does not differ significantly from literature ones when assessed over annual scales. Over shorter time scales, though, we demonstrate large differences in gas transfer between day and night, presumably due to water-side thermal convection. This is a previously unreported finding for estuaries, and one that will greatly improve global assessments of estuarine carbon dioxide emissions. Finally, an empirical stable carbon isotope approach is combined with a modeling exercise to show that conditions of carbon dioxide limitation may promote the dominance of potentially-toxic cyanobacteria. Over all, this dissertation research made progress towards the stated goal by showing that the delivery of fresh water, both during normal conditions and during storm-driven flooding events, significantly affects estuarine carbon budgets. Furthermore, it appears that previously unaccounted for physical factors may drive variations in gas transfer rates between day and night.Doctor of Philosoph