Development and validation of a critical source area index tool to identify areas vulnerable to phosphorus loss
Open Access
- Author:
- Lesher, Emily
- Graduate Program:
- Soil Science
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- July 20, 2021
- Committee Members:
- Bradley Cardinale, Program Head/Chair
Patrick Joseph Drohan, Thesis Advisor/Co-Advisor
Peter J A Kleinman, Committee Member
John Thomas Spargo, Committee Member
Ian Thomas, Special Signatory - Keywords:
- Phosphorus
nutrient management
critical source areas
hydrological source areas
phosphorus index - Abstract:
- The Chesapeake Bay Watershed (CBW) has experienced water quality deterioration caused by diffuse agricultural phosphorus and nitrogen runoff entering the bay area. Agricultural decision support tools (DSTs) can help to not only improve commodity production, but also reduce pest and disease damage and reduce nutrient pollution. In the U.S. a commonly used DST is the Phosphorus Index (P Index), which is a field-based tool that identifies critical source areas (CSAs) vulnerable to P loss by evaluating source and transport factors; a high P-Index field would benefit from management changes to reduce P loss risk. An alternative approach is the identification of CSAs using high resolution landscape digital elevation models to identify within field intersections between high runoff potential and high soil phosphorus concentrations. Identifying CSAs as an approach for mitigating P loss from agricultural watersheds has revealed that often small portions of a watershed are responsible for the majority of pollutant loads. An example high resolution digital elevation model tool, the CSA Index, was developed in Ireland to target the most at-risk areas in a field for producing high runoff amounts and high P soils. The CSA Index is developed using plant available P measures like Mehlich-III P. However, studies investigating if forms of soil P beyond Mehlich-III extractable P should be monitored to reduce P losses are lacking. In this thesis I apply the CSA Index to four case study areas within the WE-38 sub-watershed of the greater CBW. The CSA Index generation was assessed using both a shallow soil sampling, 0- to 5- cm, and a standard agronomic sampling depth, 0- to 15- cm approaches at different depths. The results of the CSA Index are compared to the Pennsylvania P Index results run on each of the fields located within the case study areas to determine how areas vulnerable to P loss are identified and best management practice (BMPs) recommendations are made using each DST as a guide. I then performed intense sub-field scale soil sampling and re-run the CSA Index. These results were used to evaluate CSA Index results when soil P concentrations were derived from standard agronomic soil sampling and the more intensive sub-scale field sampling. Lastly, I examined different extraction approaches for measuring forms of phosphorus in the soil. I aimed to determine if forms of phosphorus vulnerable for runoff are missed by Mehlich-III P extraction methods and whether the presence or absence of a CSA influences the P concentrations. I found that the CSA Index and P Index differ in their identification of areas vulnerable to P loss. The P Index identified 15 out of 19 fields as being a “low risk” of P loss despite the CSA Index identifying CSAs in some of these “low risk” fields. The use of the CSA Index in combination with the P Index allowed for a discussion of BMPs for these case study areas. Recommendations to reduce P loss included extension of riparian buffers, crop conversion to permanent hay, and changes in field boundaries. It was determined that the best approach for applying BMPs to fields is to use the CSA Index initially to target CSAs on a smaller sub-field scale as well as identify flow pathways that may influence management changes. The P Index should then be used to drive management changes on the remaining landscape. Three of the four sub-watersheds saw an increase in CSA area at the 0- to 15- cm which resulted in more vulnerable areas of the landscape being identified at this deeper depth. The CSA Indices generated using the sub-scale soil sampling data showed that in all case study areas CSA are decreased with the more intensive sampling approach when compared to standard agronomic sampling. Despite the decrease in CSA area the agronomic sampling models still identified the majority of the most at-risk areas. The cost of performing sub-scale sampling as well as analyzing the data is a deterrent for using this sampling approach to run the CSA Index. For this reason and the fact that the models derived using agronomic sampling still succeeded in identifying CSAs it was suggested that soil P data determined using standard agronomic soil sampling be used to run the CSA Index. The evaluation of forms of phosphorus and methods used to measure them revealed that Mehlich-III P was a better predictor of total P than OP was and that Mehlich-III P concentrations increased exponentially as total recoverable 3050B P concentrations increased. Further investigation is needed to better understand the relationship between CSA and non-CSA areas and the forms of phosphorus present in the soils of these areas. Results from this thesis show how the use of the CSA Index in the CBW can further mitigate P loss from agricultural fields by precisely identifying within field CSAs that current DSTs like the P Index were not designed to identify. The widespread availability of LiDAR data across the region has dramatically reduced the cost of CSA Index development. Available soil testing data, or existing data ranges, could be used with LiDAR data to rapidly develop the CSA Index across CBW member states. The index should be evaluated in other physiographic provinces with less topographic driven surface hydrology to assess how it will perform in such landscapes. Future studies to validate the CSA Index should also include runoff monitoring of nutrient loads to adjacent waterbodies.