Search for: All records

Abstract Long‐term agricultural field experiments (LTFEs) have been conducted for nearly 150 years. Yet lack of coordination means that synthesis across such experiments remains rare, constituting a missed opportunity for deriving general principles of agroecosystem structure and function. Here, we introduce the Diverse Rotations Improve Valuable Ecosystem Services (DRIVES) project, which uses legacy data from North American LTFEs to address research questions about the multifunctionality of agriculture. The DRIVES Project is a network of researchers who have compiled a database of primary (i.e., observations) and secondary (i.e., transformed observations or modeling results) data from participating sites. It comprises 21 LTFEs that evaluate how crop rotational diversity impacts cropping system performance. The Network consists of United States Department of Agriculture, university, and International Maize and Wheat Improvement Center scientists (20 people) who manage and collect primary data from LTFEs and a core team (nine people) who organize the network, curate network data, and synthesize cross‐network findings. As of 2024, the DRIVES Project database contains 495 site‐years of crop yields, daily weather, soil analysis, and management information. The DRIVES database is findable, accessible, interoperable, and reusable, which allows integration with other public datasets. Initial research has focused on how rotational diversity impacts resilience in the face of adverse weather, nutritional quality, and economic feasibility. Our collaborative approach in handling LTFE data has established a model for data organization that facilitates broader synthesis studies. We openly invite other sites to join the DRIVES network and share their data. 

This work was conducted by the Diverse Rotations Improve Valuable Ecosystem Services (DRIVES) project, based in the USDA-ARS Sustainable Agricultural Systems Lab in Beltsville, MD. The DRIVES team compiled a database of 20-plus long-term cropping systems experiments in North America in order to conduct cross-site research. This repository contains all scripts from our first research paper from the DRIVES database: "Rotational complexity increases cropping system output under poorer growing conditions," published in One Earth (in press). This analysis uses crop yield and experimental design data from the DRIVES database and public data sources for crop prices and inflation. This repository includes limited datasets derived from public sources or lacking connection to site IDs. We do not have permission to share the full primary dataset, but can provide data upon request with permission from site contacts.The scripts show all data setup, analysis, and visualization steps used to investigate how crop rotation diversity (defined by rotation length and the number of species) impacts productivity of whole rotations and component crops under varying growing conditions. We used Bayesian multilevel modeling fit to data from 20 long-term cropping systems datasets in North America (434 site-years, 36,000 observations). Rotation- and crop-level productivity were quantified as dollar output, using price coefficients derived from National Agriculture Statistics Service (NASS) price data (included in repository). Growing condtions were quantified using an Environmental Index calculated from site-year average output. Bayesian multilevel models were implemented using the 'brms' R package, which is a wrapper for Stan. Descriptions of all files are included in README.pdf.