ManureDB - National database of manure nutrient content and other characteristics: 1998 - 2023

The vision for a manure nutrient database, ManureDB (Bohl Bormann et al., 2024), was to collect and evaluate historical manure data from multiple agricultural commercial and university laboratories in the United States (U.S.) and then create a database to aggregate dynamic data in space and time while meeting FAIR data principles (Findable, Accessible, Interoperable, and Reusable). A research team of manure management and testing experts sought to investigate, collect, and aggregate manure sample data from across the U.S. through a collaborative effort between universities, laboratories, and related interest groups and transform that information into a dynamic database. A development team at the Minnesota Supercomputing Institute helped scaffold the database architecture. Named ManureDB, the manure database project can be found at http://manuredb.umn.edu/ where options to view summary statistics and download data are available to the public. Working with input from a stakeholder group, the project team developed a database schema, sample template, data-sharing agreement, data upload process, and website to support the database. Laboratories shared manure data without customer names and limited geographical data to protect customer privacy. Data that met minimum requirements was uploaded into the database. The prototype database was developed as a Python WebApp using a sqlite3 database, and a Linux VM deployed the web application. The database schema outlined the organization, labeling, and layout of the database.

A standardized survey data template with versioning was created and iteratively refined to capture the information coming from laboratories while placing that data into a standard format prior to uploading it into the database. The template contents included lab name, year analyzed, date sampled, date received, date analyzed, date reported, sample ID, account number, ZIP Code, state, sample notes, manure type, animal or other amendment type, manure treatment, agitation status, bedding type, storage type, length of storage, and application method options, if known. The variety of date options came from interviews with laboratory managers and their date reporting procedures; only “year analyzed” was a required value. The sample notes column included any notes the customer wrote on the sample submittal form and were included with the data submitted to the ManureDB project. During the data cleaning process, prior to database upload, a ManureDB team member searched the sample notes column with keywords to see if other labels for a sample could be added. The manure types were based on commonly used manure consistency descriptions by laboratories: liquid, slurry, semi-solid, separated solid, separated liquid, digested solid, digestate, runoff, sludge, as-excreted (urine), and as-excreted (feces), litter, or compost. The manure treatment category was added because of some inclusion of treatment labels by some laboratories. They included litter amendments (alum, KLASP, PLT, Poultry Guard), a feed additive (phytase), a process (dried, aerobic), other treatment, or no treatment. An agitation category had a yes, no, or unknown label for whether the manure was agitated or not when sampled if known. The template offered a wide selection of analyte options from macronutrients, micronutrients, metals, characteristics, and calculations. For each analyte reported, the analytical method, units reported, and wet or dry basis options were selected from controlled options on the template.

The animal or other amendment type originally was populated with common livestock species, however, once the data started to roll in, more specific sex and life stages were added along with other less common species. Laboratories also analyzed organic amendments for nutrient analysis, so the team also decided to include types of organic amendments used for agricultural purposes that had multiple samples analyzed to the list of options. For bedding type, besides unspecified bedding or no bedding, specific bedding material options were also available since some laboratories captured that information. The specific bedding type options included: hardwood sawdust, hardwood shavings, paper, peanut hulls, pine sawdust, pine shavings, rice hulls, sand, and straw.

The storage type column had options based on MidWest Plan Services Manure Storages Publication MWPS-18 Section 2, which included: lagoon, uncovered pit or tank, covered pit or tank, earthen basin, runoff holding pond, solid, underfloor solid storage, stockpile under cover, stockpiled outdoor, and cage-free poultry (Fulhage et al., 2001). The length of storage category was added when a laboratory had length of storage on their submittal forms. The manure length of storage category included daily haul, 0-3 months, 3-6 months, 6-12 months, and 12 plus months. Application method information was also captured on some submittal forms and the application methods options listed were irrigation, broadcast, broadcast – incorporated within 24 hours, broadcast – incorporated after 24 hours, and injection. For inclusion in the database, the only requirements were the year the sample was analyzed, the laboratory where analyzed, and the animal type or other organic amendment category for inclusion in the database.

This project carefully protected data privacy with established rules. A participating laboratory or data partner and the University of Minnesota both signed a data use agreement prior to data transfer. Laboratories shared past manure data with no customer names or addresses shared to avoid privacy concerns. Only the state and/or first three digits of a ZIP Code were uploaded into the database. The public-facing database website and this dataset did not include ZIP Codes, laboratory identities, account information, or samples notes. To show up in public summaries such as this dataset, at least five samples per year from a state, region, or U.S. were required. The public-facing website and this dataset showed aggregate summary data for a region, animal type, or time span. As of the February 19, 2024 dataset, ManureDB data spanned from 1998-2023 and included >490,000 samples from 49 states, 14 laboratories, over 65 animal types, and 18 organic amendments.

References

• Bohl Bormann, N. L., Wilson, M. L., Cortus, E. L., Silverstein, K. A. T., Janni, K. A., & Gunderson, L. M. (2024). ManureDB. http://manuredb.umn.edu/
• Fulhage, C., Hoehne, J., Jones, D., & Koelsch, R. (2001). Manures Storages—Manure Management System Series. MidWest Plan Service, Iowa State University, Ames, IA. https://store.extension.iastate.edu/product/Manures-Storages-Manure-Management-System-Series