Conceptual framework of connectivity for a national agroecosystem model based on transport processes and management practices

Authors: Arnold, Jeffrey; White, Michael; ALLEN, P - Baylor University; GASSMAN, P - Iowa State University; BIEGER, K - Texas Agrilife Research

Submitted to: Journal of the American Water Resources Association
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/29/2020
Publication Date: 2/1/2021
Citation: Arnold, J.G., White, M.J., Allen, P.M., Gassman, P.W., Bieger, K. 2021. Conceptual framework of connectivity for a national agroecosystem model based on transport processes and management practices. Journal of the American Water Resources Association. 57(1):154-169. https://doi.org/10.1111/1752-1688.12890.
DOI: https://doi.org/10.1111/1752-1688.12890

Interpretive Summary: The USDA uses national models to: 1) determine the impact of conservation practices on the environment (streamflow, sediment and nutrients) and 2) to run future land use and management scenarios to determine conservation policy. In the previous national assessment, similar soils and land uses were lumped into single homogeneous areas and only large rivers were simulated. One of the major problems with this approach is that much of the sediment and nutrients delivered to water bodies originates in small headwater streams and we did not explicitly simulate processes in those streams. In this study, we propose a conceptual framework to divide the country into five spatial zones to more accurately simulate processes and management in smaller steams. Previous studies have delineated 4.2 million agricultural fields and the National Hydrography Dataset (NHD+) delineates all the streams in the U.S. We show how it is conceptually possible to simulate each of the 4.2 million ag fields and the connections to each stream and reservoir across the U.S. The framework is demonstrated on a large watershed in central Illinois and Indiana to show how fields and streams are connected in low relief, tile drained watersheds. The concept is also demonstrated for a field in steeper terrain in central Iowa. When implemented, the new framework will allow future national environmental assessments to simulate each ag field and each stream and reservoir in the U.S. It also provides the capability of simulating off-field conservation practices in headwater streams. All the data used to formulate the model is readily available and will enhance current conservation assessment and policy.

Technical Abstract: The model currently used for USDA national conservation assessment lumps land use and management within 8-digit USGS hydrologic unit codes and does not explicitly simulate low order streams. In addition, most individual watershed studies typically develop hydrologic connectivity based on computer storage and run times, without consideration of flow and sediment transport processes. Numerous studies have shown that a large portion of sediment and nutrients delivered to water bodies across the U.S. originate from headwater streams. In this study, we propose to conceptually divide the watershed continuum into five discrete spatial zones or domains: field (1-40 ha), transition (0.2-2.0 km2), headwater (1-15 km2), tributaries (10-150 km2), and main river (> 150 km2). The process domain concept assumes that the spatial variability in geomorphology controls processes that are distributed over the landscape. Agricultural management at each domain were also identified. Previous research by others has delineated 4.2 million agricultural fields across the U.S. Methods were developed in this study to identify the other domains using the National Hydrography Dataset (NHD+). An example 8-digit watershed in central Illinois and Indiana (07120002) was shown to demonstrate connectivity of fields, transition, and other process domains. This watershed is typical of low relief, tile drained agriculture. An example of field and transition domains in steeper terrain in central Iowa was also shown to demonstrate differing processes and management in a steeper agricultural landscape. The new connectivity will improve the USDA conservation assessment model in two important areas. First, each ag field will be simulated using dominate soil, management and slope without lumping into hydrologic response units. This allows use of a single rain gage for each field, eliminating areal weighting of precipitation. Second, each headwater stream will be explicitly simulated, eliminating lumping all sediment delivery processes into a delivery ratio. This is critically important since much of the sediment and legacy nutrients originate from headwater and tributary streams. It also provides the capability of simulating off-field management practices in headwater and tributary streams. All the data used to formulate this conceptual model is readily available across the U.S. However, limitations include: 1) management data is not available for every field, 2) transition channels from each field are not easily defined from NHD+, and 3) the resulting national model is spatially complex and computationally intensive. These limitations are not currently prohibitive and future improvements in data collection and computer speed will alleviate the limitations.