Background & Summary

Winter cover crops provide many ecosystem services such as weed suppression, improved soil structural and hydraulic properties, increased soil organic carbon (C) stocks, reduced erosion, reduced winter nitrogen (N) leaching, and N provision to cash crops1,2,3,4. Cover crop biomass production is often positively correlated to ecosystem service provisions1. For example, previous research has shown that weed biomass often decreases with greater cereal rye (Secale cereale) residue5,6, and soil organic C often increases with cover crop biomass7,8. Similarly, there were greater reductions in nitrate leaching with increases in non-leguminous cover crop shoot biomass in a global meta-analysis3. Elucidating the agronomic, soils, and climate controls on winter cover crop performance can help farmers determine the optimal time to terminate cover crops for maximum agronomic benefits, and support broader adoption of winter cover crops for increased climate resilience, reduced soil erosion, and lower nutrient pollution2.

We focused on performance data of cereal rye, the most commonly used cover crops in the United States9,10. To acquire data for this study, we reached out to potential data contributors through regional cover crop groups and the Precision Sustainable Agriculture and Getting Rid of Weeds networks, two national research consortia with a major focus on cover crop research. Recruiting data contributors through this network enabled us to assimilate many plot-level observations from both on-station and on-farm studies (Fig. 1). Our goal was to create a dataset on cereal rye cover crop biomass quantity and quality across heterogeneous agronomic, soils, and climate conditions with broad coverage across the eastern half of United States (Fig. 1).

Fig. 1
figure 1

Map of research locations with the point color scaled by the number of available observations.

Full size image

Methods

We collected data on cereal rye cover crop performance metrics (biomass, N content, and C:N ratio), additional agronomic and soil data, and metadata such as any associated publications (Table 1). The minimum data required from a location for inclusion in our dataset was aboveground (shoot) cereal rye biomass, the respective harvest or sampling date, the experimental site name, year, latitude and longitude, cereal rye planting date, cereal rye planting method (drilled vs. broadcast), and whether N fertilizer was applied during cover crop growth (Table 1).

Table 1 Structure of the data records with file name, purpose, and variables included for each tabular dataset included.
Full size table

Additional data gathered included cereal rye cultivar; seeding rate; row spacing; plant population; tiller density; growth stage at sampling; cumulative growing degree days; shoot N (concentration or content); shoot C:N ratio; N fertilizer rate, type (form), and application date; presence of fall tillage; and previous and subsequent cash crop. Root C and N data were requested but not available for any study. Additional soil data gathered included texture class and/or clay, silt, and sand percentages; bulk density; soil pH; soil ammonium, nitrate, and/or total inorganic N; soil organic matter or C; as well as soil sampling depth and timing. The overwhelming majority of observations were recorded as plot level data. In the few cases (17 observations) when plot level data were not available, we collected treatment means with standard deviations (Table 1). At least 14 of the 28 studies correspond to existing publications, which are detailed in the study metadata CSV file11,12,13,14,15,16,17,18,19,20,21,22,23,24,25. Plot sizes varied from as small as 6.1 × 6.1 m to as large as 30.5 × 42.7 m depending on the study. These published datasets described methods for cereal rye biomass collection and some soil analyses in each publication. The DOI from any publications associated with data provided is listed in the “publication_DOI” column in the study metadata file, and “NA” is listed for studies with unpublished data. The methods for cereal rye biomass collection and some soil analyses and unpublished are detailed in the “methods_unpublished_data” column of the study metadata file.

To focus on the vegetative biomass of cereal rye that is planted in the fall and followed by cash crops in the spring, we only included biomass sampling that occurred in February, March, April, or May. We excluded observations from sites that terminated winter cover crops after May—such as forage experiments where cover crop vegetative biomass and grain were sampled in June or later. We also omitted biomass data outliers that fell above the 99.9th percentile (n = 6) because the upper end of the distribution appeared to have unrealistic values. All figures and summary statistics were created using the statistical software R (v4.1.3)26 and the following packages: ggplot2 v3.4.127, ggmap v3.0.128, and ggsn 0.5.029.

Data Records

The metadata and primary data collected can be accessed through the following repository: https://doi.org/10.5061/dryad.tx95x6b3h30. The data are organized in tabular CSV files including a data dictionary, study metadata, rye data, agronomic data, and soil data. (Table 1). In 26 out of 245 site years, location data from private farms were obscured to ensure privacy and have lower location accuracy than the rest of the dataset; as such, those locations are indicated as “TRUE” in the “location_obscured” variable. We reported location data with as much precision as possible for each datapoint, as a result, there are varying levels of precision in reported latitude and longitude coordinates.

Technical Validation

To check the validity of the data we collected, we examined the spread of the data and found some unreasonable values. Cereal rye biomass data outliers that fell above the 99.9th percentile were excluded. The final dataset had a mean of 3,428 kg ha−1, a median of 2,458 kg ha−1 and a standard deviation of 3,163 kg ha−1 (Fig. 2). We also checked overall data validity by assessing whether cereal rye biomass generally increased with time elapsed from planting to termination date. This increase was observed; however, there was a large amount of variability (Fig. 3). This variability in cereal rye biomass production may be explained by differences in agronomic, soil, and climate factors. A subset of commonly reported ancillary data is summarized in Table 2 and Fig. 4.

Fig. 2
figure 2

Histogram of cereal rye shoot dry biomass data.

Full size image
Fig. 3
figure 3

Cereal rye shoot biomass versus time elapsed between cereal rye planting to termination date.

Full size image
Table 2 Summary of commonly-reported ancillary data from the rye growth dataset.
Full size table
Fig. 4
figure 4

Summary of planting method data.

Full size image