| dc.description.abstract | Drinking water supplies for much of society originate in forests. To preserve the capability of these forests to produce clean and easily treatable water, source water supply and protection strategies focus in particular on potential disturbances to the landscape, which include prescribed forest harvesting and wildfires of varying intensity. While decades of work have revealed relationships between forest harvesting and stream flow response, there is a considerable lack of synthesis disentangling the interactions of climate, wildfires, stream flow, and water quality. Revealing the mechanisms for impacts on downstream waters after disturbances of harvesting and wildfire will greatly improve land and water management. In this dissertation, I combined synthesis of previously published or available data, novel mathematical analyses, and deterministic modeling to disentangle various disturbance effects and further our understanding of processes in forested watersheds. I broadly sought to explore how streamflow and water quality change after forest disturbances, and how new methods and analyses can provide insight into the biogeochemical and ecohydrologic processes changing during disturbances.
First, I examined the effect of wildfire on hydrology, and developed a novel Budyko decomposition method to separate climatic and disturbance effects on streamflow. Using a set of 17 watersheds in southern California, I showed that while traditional metrics like changes in flow or runoff ratio might not detect a disturbance effect from wildfire due to confounding climate signals, the Budyko framework can be used successfully for statistical change detection. The method was used to estimate hydrologic recovery timescales that varied between 5 and 45 years, with an increase of about 4 years of recovery time per 10% of the watershed burned.
Next, in Chapter 3 I used a meta-analysis approach to examine the effect of wildfire on water quality, using data from 121 catchments around the world. Analyzing the changes in concentrations of stream water nutrients, including carbon, nitrogen, and phosphorus, I showed that concentrations generally increased after fire. While a large amount of variability existed in the data, we found concurrent increases in the constituents after fire highlighting tight coupling of the biogeochemical cycles. Most interestingly, we found fire to increase the concentrations of biologically active nutrients like nitrate and phosphate at a greater rate than total nitrogen and phosphorus, with median increases of 40-60% in the nitrate to TN, and SRP to TP ratios.
Next, I conducted an analysis of both water quality and hydrology together after fire in Chapter 4, using a set of 29 wildfire-impacted watersheds in the United States. Concentration-discharge relationships can be used to reveal pathways and sources of elements exported from watersheds, and my overall hypothesis was that these relationships change in post-fire landscapes. I developed a new methodology, using k-means clustering, to classify watersheds as chemostatic, dilution, mobilization and chemodynamic, and explored how their position within the cluster changed in post-fire landscapes. I found that the behavior of nitrate and ammonium was increasingly chemostatic after fire, while behavior of total nitrogen, phosphorus, and organic phosphorus was increasingly mobilizing after fire.
Finally, I developed a coupled hydrology-vegetation-biogeochemistry model to simulate and elucidate processes controlling the impact of harvesting on downstream waters. I focused on the Turkey Lakes watershed where a significant amount of data has been collected on vegetation and soil nutrient dynamics, in addition to traditional streamflow and water quality metrics, and developed a novel multi-part calibration process that used measured data on stream, forest, and soil characteristics and dynamics. Future work would involve using the model to explore the data driven relationships that have been developed in the earlier chapters of the paper.
The work presented in this dissertation highlights new small and large-scale relationships between disturbances in forested watersheds and effects on downstream waters. With more threats predicted to escalate and overlap in the coming years, the novel results and methodologies that I have presented here should contribute to improving land and water management. | en |
