A Procedure to Assess and Revise The Pennsylvania Phosphorus Index
Open Access
- Author:
- Weld, Jennifer Lizabeth
- Graduate Program:
- Soil Science (PHD)
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- January 21, 2022
- Committee Members:
- Bradley Cardinale, Program Head/Chair
Tameria Veith, Outside Unit & Field Member
Charlie White, Major Field Member
Peter Kleinman, Major Field Member
Douglas Beegle, Chair & Dissertation Advisor - Keywords:
- Phosphorus
Phosphorus Index
TopoSWAT
APLE
nutrient management - Abstract:
- Phosphorus is an essential nutrient for agricultural production, but losses from agricultural fields into the environment are a major source of water pollution in many regions of the globe. While farmers have an economic incentive to manage purchased nutrients carefully, management of manure nutrients that are a waste product of on-farm animal production often require regulations to promote compliance with best management practices that can help protect water quality. Phosphorus (P) loss risk assessment tools are commonly incorporated into nutrient management regulations to evaluate and modify agricultural management practices to reduce phosphorus pollution. In many states, these tools are referred to by the name Phosphorus Index (P Index), but the structure and usage of P Indices across states is highly variable. In an attempt to unify P Index structures, scoring factors, and scoring scales across states, the US Department of Agriculture - Nature Resource Conservation Service (USDA-NRCS) revised their nutrient management standard. This revision required states to evaluate Phosphorus Index (P Index) factors and structures against standardized criteria and compare P Index values and recommendation categories to estimated or measured phosphorus (P) loss. The goal of the assessments was to minimize differences between P Indices nationally and to better connect P Index values to quantitative P loss. The objectives of this study were 1. to evaluate the Pennsylvania P Index for a mixed land use sub-watershed in Pennsylvania by comparing P Index values and rankings with P loss values estimated by water quality models, and measured P loss data and 2. To revise the Pennsylvania P Index so that it better corresponds to measured and modeled P losses. For this study, the Pennsylvania P Index version 2.0, a modification of the Soil and Water Assessment Tool (TopoSWAT), and Annual Phosphorus Loss Estimator (APLE) were used to evaluate crop, manure, and fertilizer nutrient management for the five agricultural fields in an 11-ha sub-watershed in the Ridge and Valley physiographic region located in central Pennsylvania. Pennsylvania P Index version 2.0 was evaluated against total P estimated by TopoSWAT and APLE. For comparison, the Pennsylvania P Index version 2.0 values were categorized by the associated management categories of Low, Medium, High, and Very High and estimated P losses were categorized by management categories of Low, Medium, and High defined by the USDA-NRCS national instruction. There were significant, but weak correlations between the Pennsylvania P Index version 2.0 values and total P losses estimated by TopoSWAT (r = 0.359) and APLE (r = 0.416). Using the average estimated total P values across all fields and crop years, total P loss estimated by APLE was higher (5.30 kg TP ha-1) than total P loss estimated by TopoSWAT (3.26 kg TP ha-1). TopoSWAT dissolved P to total P ratio had a mean value of 0.25 and a median value of 0.13 (range 0.003 to 0.95) indicating that particulate P loss is the predominate P loss pathway. Pennsylvania P Index version 2.0 values ranged from 30 to 114 for the five fields across five years. A low P Index rating was assigned to 64% of field scenarios, a medium rating to 20% of all fields, a high rating to 10% of fields, and a very high rating to 10% of fields. According to current state regulations, nutrient applications would be restricted on the 20% of fields ranked as high or very high. Comparison of Pennsylvania P Index version 2.0 categories and predicted P loss categories resulted in agreement for 48% of fields when using TopoSWAT and agreement for 8% of fields when using APLE. The P Index version 2.0 does not meet USDA-NRCS requirements; however, the P Index version 2.0 values did increase as predicted total P losses increased. Conducting this analysis also revealed several issues with using TopoSWAT or APLE to evaluate the P Index. Neither model directly represents the spatial distribution of P loss and there is no mechanistic handling of delivery of P to receiving waters from a field, which is the spatial unit evaluated in the P Index, TopoSWAT does not use inputs or provide outputs that are directly relatable to management inputs and outputs in the P Index, For example, there is no direct way to enter a soil test P level in TopoSWAT, but this is a very critical factor in the P Index. APLE uses a number of default inputs that may not represent the actual features of the area being evaluated. Neither model agreed very well with some limited measured P loss in the study watershed. It is unclear whether the P loss values calculated by either model correspond to the P loss criteria in the NRCS guidance, since this criterion is also based on another model. Because of these issues the modeling approach to be used in evaluating a P Index must be considered carefully to assure that valid evaluations are being made. Based on this evaluation of Pennsylvania P Index version 2.0, this study proposes a revised Pennsylvania P Index version 3.0 that includes separate evaluations for dissolved and particulate P, a separate site evaluation, and modification and prioritization of the evaluation factors, weightings, and interpretation. For example, soil loss is determined annually instead of as a rotational average. Pennsylvania P Index version 3.0 was evaluated against dissolved, particulate, and total P estimates by APLE. Runoff and rainfall inputs for APLE were calculated using regional rainfall and the SCS Curve Number (SCS-CN) method. Correlation assessment used Kendall’s Beta Tau correlation analysis. Using values for all fields across all crop year, Particulate P losses (r = 0.754) and total P losses (r = 0.140) estimated by APLE was strongly correlated to Pennsylvania particulate P Index version 3.0 values. Dissolved P losses estimated by APLE were not significantly related to Pennsylvania dissolved P Index version 3.0 values (r = 0.020). When Pennsylvania dissolved P Index version 3.0 values were determined using calculated and not categorical values for runoff, there was a stronger correlation between estimated dissolved P and Pennsylvania dissolved P Index version 3.0 values (r = 0.633). The proposed PA P Index version 3.0 does meet USDA-NRCS requirements and resulted in P Index version 3.0 values that better correlate with predicted P losses.