Assessing Conservation Practice Effectiveness with Lorenz Inequality Results
Open Access
- Author:
- Biertempfel, Julia
- Graduate Program:
- Biorenewable Systems
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- June 21, 2022
- Committee Members:
- Suat Irmak, Program Head/Chair
Heather Elise Preisendanz, Thesis Advisor/Co-Advisor
Cibin Raj, Committee Member
Tamie Veith, Committee Member - Keywords:
- water quality
Lorenz inequality
Gini coefficient
TMDL
Chesapeake Bay
nutrient loading
conservation practices
best management practices
BMPs
nutrients
phosphorus
nitrogen
agriculture
concentration-discharge
legacy nutrients - Abstract:
- The Chesapeake Bay watershed is home to over 18 million people and contains over 87,000 working farms, which are two aspects that have led to the Bay’s long-standing impaired status. In response to the federal mandate to reduce pollutant loadings to the Bay and to meet its portion of the watershed-wide load reduction goals, the Commonwealth of Pennsylvania has divided its counties into four prioritization tiers based on potential pollutant reduction. Twenty watersheds within the Susquehanna and Potomac portions of the Chesapeake Bay were selected from among the tiers to compare watersheds with varying levels of documented agriculturally-focused best management practice (BMP) implementation to those dominated by forested land cover. Although spatial targeting of BMPs has been extensively studied for reducing nutrient and sediment loads from agriculturally-dominated landscapes of the Bay watershed, effectiveness of BMP implementation to restore natural biogeochemical variability to the nitrogen and phosphorus cycles remains unknown. Furthermore, the time frame between widescale BMP implementation and watershed recovery at the county level is a complex subject which requires more assessment. Both these research inquiries could affect policy decisions, as available funding constraints often force BMPs to be implemented where they will be most effective. Research results could help land managers and policymakers assess the effectiveness of BMP implementation for achieving load reduction goals, as well as provide a better understanding of their effectiveness not only in reducing loads, but in restoring variability to nutrient cycling. This research investigated the hydrologic, biogeochemical, anthropogenic, and physiographic factors that influence the degree of temporal inequality exhibited by nutrient time series data in the Chesapeake Bay watershed. Temporal inequality was assessed via Lorenz inequality curves and their corresponding Gini equations, as well as testing and comparing the concentration-discharge relationships of TN and TP, within twenty targeted watersheds within the Susquehanna and Potomac portions of the Chesapeake Bay. Spatial analysis was additionally conducted to analyze the relationship between Gini values, nutrient reduction progress, forested and agricultural land use percentages, coefficients of variation, and b values. Specifically, this thesis examined how the implementation of BMPs affects the degree of temporal inequality exhibited by nutrient transport at individual gauging stations over time and how quickly improvements are reflected within flow and load data. The watersheds of interest were selected using tier ranking from Phase III of the Pennsylvania Watershed Implementation Plan (WIP). The goal of this thesis was to compare watersheds with varying levels of documented BMP implementation to those with largely forested land cover in order to assess comparisons in degrees of temporal inequality of flow and nutrient load, as well as whether BMPs restore variability to nutrient dynamics within the studied timeframe. Hydroclimic variables were also scrutinized based on how they affect Gini Coefficients for nutrient loads. The hypothesis that drove this research was that agriculturally-dominated watersheds would show less variability of nutrient concentrations observed in streams due to nutrient legacy sources and current-day TN and TP sources, while less impacted watersheds would exhibit higher variability. I anticipated that if BMPs have helped to restore the natural variability of nutrient concentrations, then the degree of temporal inequality exhibited by a given gauging station location will increase over time after the BMPs have been implemented. However, due to the different pathways TN and TP pollution take in the environment, I anticipated TP to respond faster to BMP implementation than TN. Results drawn after data analysis indicate that there were few significant data markers of the improvement in TN variability, even after looking at both time series data and spatial analysis. Though this was the expected result for watershed sub-catchments with a higher percentage of agricultural land-use, this persisted even in more forested sub-catchments. No sub-catchment investigated in this thesis had a CVTN:CVQ value above 0.3 by the end of the data series, which highlights how difficult it is to restore TN variability in a landscape. Ultimately, I concluded that noticeable trend changes in TN variability may not be detectible only eight years after the TMDL model was implemented. Conversely, I concluded that there were visible signs that indicated several sub-watersheds within the scope of this research that displayed improving variability patterns for TP. The Gini coefficients of TP were overall more responsive to change than TN, particularly when looking at tier 1 and 2 sub-catchments. This is likely due to a combination of more progress completed towards TP reductions via infrastructure implementation (BMPs) and the different pathways TN and TP pollution take in the environment. A discussion of the implications and limitations of using the Weighted Regression on Time, Discharge, and Season (WRTDS) method to analyze long-term surface water-quality data is included at the end of this thesis.