Predicting the Impact of H2Ohio Practices on Edge-of-Field Water Quality

Using Edge-of-Field Data and Modeling to Inform H2Ohio Eutrophication driven by excess nutrient loading to receiving waters may be the most rampant global water quality issue facing society, leading to dire environmental, social and economic consequences. Recently in the United States, a significant shift in freshwater streams and lakes from an oligotrophic to mesotrophic, and mesotrophic to eutrophic states has been documented. In Lake Erie, greater than 70% of phosphorus (P) delivery has been linked to nonpoint sources and has led to the recurrence of harmful algal blooms (HABs). Significant voluntary, incentive, and regulatory efforts have been focused on addressing agricultural phosphorus (P) loss within the Western Lake Erie Basin (WLEB) watershed. The H2Ohio initiative was developed to promote conservation practice adoption to ensure clean waters for Ohio. However, much of the data used to identify and promote the conservation practices included in the 4R nutrient stewardship and H2Ohio frameworks were based on studies from locations that did not have the unique heavy clay soils and intensive tile drainage found in the Ohio WLEB watershed. Direct quantitative measurements of agricultural management practice impacts on tile-drain water quality at the field-scale is impractical across a broad array of landscapes due to economic and logistical constraints. Process-based numerical models at the field-scale can aid in filling the gaps, but often have uncertain parameters that require calibration and validation using long-term monitoring records at specific sites. For this project, we will focus on numerical modeling of the USDA-ARS SDRU edge-of-field (EOF) monitoring network in order to quantify 1) typical water quality loadings in tile- drained landscapes as it pertains to landscape variability, and 2) nutrient load reduction potential for prevailing water quality management practices such as drainage water management, subsurface fertilizer placement, and cover crops. Objective 1: We will select an appropriate field-scale model (e.g., APEX, DRAINMOD, RZWQM2, or field- scale SWAT) to simulate hydrology and water quality dynamics across study sites in the USDA-ARS EOF monitoring network that span typical soil texture gradients, soil nutrient levels, and prevailing cropping practices of Ohio to obtain a baseline scenario. Uncertain hydrologic and water quality modeling parameters will be calibrated for each site. Next, we will run a Monte Carlo (randomization) analysis to simulate how various combinations of landscape variables (e.g., soil texture, soil nutrient levels, and climate variables) impact annual nutrient loadings. We aim to develop nomographs to describe the baseline relationships between area-normalized annual nutrient loadings and landscape variables. We will consider uncertainty by considering ranges of calibrated model parameters from our initial model evaluation analysis. Objective 2: We will select four prevailing water quality management practices (initially nutrient management planning/rate, subsurface placement, drainage water management, and conservation crop rotation) that have been implemented in USDA-ARS SDRU before-after-control-impact (BACI) studies. As additional practice data (i.e., VRT and manure incorporation) becomes available we have the potential to project the impact of those practices as well since the modeling framework will be established. We will initially evaluate the utility of our field-scale model to simulate practice impacts (e.g., drainage water management). Based on findings, we may modify model formulations (as needed) to improve representation of physical processes. Similar to Objective 1, uncertain model parameters will be calibrated, in order to consider a realistic range for our uncertainty analysis. Agricultural management practices will be incorporated into the Monte Carlo simulations and we will generate nomographs that include anticipated nutrient loadings for sites when implementing prevailing agricultural water management practices. Further, we will also present nomographs showing percent reductions from baseline levels simulated.