Using Crop Management Scenario Simulations to Evaluate the Sensitivity of the Ohio Phosphorus Risk Index

Abstract

Phosphorus (P) risk indices are commonly used in the USA to estimate the field-scale risk of agricultural P runoff. Because the Ohio P Risk Index is increasingly being used to judge farmer performance, it is important to evaluate weighting/scoring of all P Index parameters to ensure Ohio farmers are credited for practices that reduce P runoff risk and not unduly penalized for things not demonstrably related to runoff risk. A sensitivity analysis provides information as to how sensitive the P Index score is to changes in inputs. The objectives were to determine 1) which inputs are most highly associated with P Index scores and 2) the relative impact of each input variable on resultant P Index scores. The current approach uses simulations across 6134 Ohio point locations and five crop management scenarios (CMSs), representing increasing soil disturbance. The CMSs range from all no-till, which is being promoted in Ohio, rotational tillage, which is a common practice in Ohio to full tillage to represent an extreme practice. Results showed that P Index scores were best explained by soil test P (31.9%) followed by connectivity to water (29.7%), soil erosion (13.4%), fertilizer application amount (11.3%), runoff class (9.5%), fertilizer application method (2.2%), and finally filter strip (2.0%). Ohio P Index simulations across CMSs one through five showed that >40% scored <15 points (low) while <1.5% scored >45 points (very high). Given Ohio water quality problems, the Ohio P Index needs to be stricter. The current approach is useful for Ohio P Index evaluations and revision decisions by spatially illustrating the impact of potential changes regionally and statewide.

Using artificial neural networks to improve phosphorus indices

“…Subsequently, an N-based rate of 192 kg P 2 O 5 ha -1 y -1 for swine effluent was determined for fields in the data set with STP levels ≤50 mg kg -1 following the steps outlined in Joern and Brichford (2003). Assuming that a corn-soybean rotation (with an average yield goal of 10,080 kg ha -1 for corn and 3,350 kg ha -1 for soybean similar to Dayton et al [2017]) has an average crop removal rate of 56 kg P 2 O 5 ha -1 y -1 , fields with STP levels between 51 to 100 mg kg -1 and 101 to 200 mg kg -1 were assigned organic P rates of 84 kg P 2 O 5 ha -1 y -1 and 56 kg P 2 O 5 ha -1 y -1 , respectively. Values for P application method and timing (FPA and OPA) were randomly assigned using modified uniform probability distributions.…”