http://hdl.handle.net/10012/11866

Increased loading of nitrogen and phosphorus from agricultural and urban intensification has led to severe degradation of inland and coastal waters. Lakes, reservoirs, wetlands, and streams retain and transform these nutrients, thus regulating their delivery to downstream waters. While the processes controlling nitrogen and phosphorus removal from the water column are relatively well-known, there is a lack of quantitative understanding of how these processes manifest across spatial scales.

This thesis explores the relationship between hydrologic and biogeochemical controls on nutrient processing in a lentic water body (lakes, reservoirs, and wetlands). Here, our work revolves around three research questions: 1) What are the emergent patterns between nutrient processing rates and residence times in lentic systems? 2) What are the underlying mechanisms contributing to the observed patterns? 3) What is the relative magnitude of nutrient retention as a function of wetland size? These questions are addressed through a meta-analysis of existing literature, the development of a modelling framework, and an analysis through upscaling of the results.

Within the meta-analysis, we synthesized data from 600 sites across the world and various lentic systems (wetlands, lakes, reservoirs) to gain insight into the relationship between hydrologic and biogeochemical controls on nutrient retention. Our results indicate that the firstorder reaction rate constant, k [T-1], is inversely proportional to the hydraulic residence time, τ [T], across six orders of magnitude in residence time for total nitrogen, total phosphorus, nitrate, and phosphate. This behavior prompted the hypothesis that the consistency of the relationship points to a strong hydrologic control on biogeochemical processing. Specifically, we hypothesized that small systems have a higher sediment surface area to water volume ratio that would facilitate the biogeochemical processes of the system.
To validate the hypothesis, we developed a two-compartment model that links the major nutrient processes with system size: the water column and the reactive sediment zone are coupled through a mass exchange process, with nitrogen being removed through denitrification in the sediments and phosphorus transferring to long term storage via particle settling. The model analyses validated our hypothesis by replicating the empirical inverse k-τ relationship through deterministic modelling. Additionally, we demonstrated the inverse relationship between the sediment surface area to water volume ratio and size through an analysis of the bathymetric relationships.

Finally, we focused on wetland systems that have been relatively less studied, and upscaled the k-τ relationships to the landscape scale using a wetland size-frequency distribution. Results highlight the disproportionately large role of small wetlands in landscape scale nutrient processing, such that for the same wetland area removed, the nutrient removal potential lost is larger when smaller wetlands are lost. The disproportionately larger role of small wetlands in landscape scale nutrient processing is important given previous research on the preferential loss of smaller wetlands from the landscape.

Through the use of a cross-system meta-analysis that spanned multiple orders of magnitude of system size, we were able to quantify multi-scale behavior that is less apparent when studying individual systems. Our study highlights the need for a stronger focus on small lentic systems as potential nutrient sinks in the landscape due to their high reactivity rates in comparison to larger water bodies. With a growing recognition that wetlands play a critical role in landscape nutrient cycling, our work will help policy makers and water managers to better understand the suite of functions that is associated with the different size classes and types of wetlands.

http://hdl.handle.net/10012/11866