Spatial and temporal variability of chlorophyll-a in U.S. rivers and streams: insights into water quality and nutrient dynamics

Lada Kuzmina

Bakalářská práce

Abstract:

This thesis examines spatiotemporal variability in algal biomass in flowing freshwater systems across the United States in 2018–2019, with an emphasis on understanding environmental factors that contribute to eutrophication. To achieve this goal, a comprehensive dataset of 2,107 water quality measurements provided by the U.S. Environmental Protection Agency was carefully processed. Temporal trends were assessed by aggregating daily measurements, interpolating missing values using linear fitting, and applying moving average smoothing to distinguish short-term fluctuations from long-term trends. Spatial mapping in ArcGIS Pro identified regions with elevated algal biomass, particularly highlighting 91 hotspots associated with agricultural runoff and urban influences. Correlation analyses performed using Pearson and Spearman methods revealed significant monotonic relationships between algal concentrations and nutrient parameters, despite modest linear associations. Linear regression models quantified these relationships, showing that elevated nutrient loads, particularly total nitrogen (β = 1.582, p-value = 2.49e-13*) and total phosphorus (β = 7.712, p-value = 9.55e-14*), were strongly associated with higher algal biomass, while temperature and precipitation showed weaker direct effects. Furthermore, application of generalized additive models (GAMs) confirmed the nonlinear nature of the nutrient-chlorophyll relationship, explaining 56% of chlorophyll-a deviance. Multivariate analysis using principal component analysis (PCA) was used to reduce dimensionality and identify the most influential environmental gradients. The first five principal components, which explained approximately 78% of the total variance, were strongly associated with nutrient dynamics and provided a concise insight into the complex interactions between water quality parameters. Clustering methods further classified sampling locations into 3 distinct eutrophication risk categories – high risk (10 sites), moderate risk (128 sites) and low risk (1,969 sites), showing that hotspots are predominantly located in regions with intensive agricultural practices. This thesis contributes to an integrative analytical framework that combines advanced statistical, temporal and spatial methods to elucidate the factors influencing eutrophication in freshwater systems. The results provide an insight on how nutrients and water quality affect algal biomass.

Keywords

Environmental Monitoring, Eutrophication, Multivariate Analysis, Spatiotemporal Dynamics, Ecological Modeling

Jazyk práce: angličtina

Datum vytvoření / odevzdání či podání práce: 3. 4. 2025

Obhajoba závěrečné práce

• Obhajoba proběhla 27. 5. 2025
• Vedoucí: doc. Zhongbing Chen
• Oponent: Ye Su, externi

Česká zemědělská univerzita v Praze

Fakulta životního prostředí

Bakalářský studijní program:
Územní technická a správní služba v životním prostředí