https://doi.org/10.5061/dryad.vdncjsz50

PI: Philip Fay, USDA-ARS Grassland, Soil, and Water Research Laboratory, Temple, TX 76502 USA.

Description of the data and file structure

Data collected: Data are from 71 grassland sites which conducted a multiple nutrient fertilization experiment for a minimum of 4 and a maximum of 14 years. All sites followed the Nutrient Network standardized experimental protocol. 5 m x 5 m experimental plots were fertilized nitrogen, phosphorus, or potassium in factorial combinations with at least three replicates per nutrient combination. Current year live aboveground biomass was sampled by harvesting, and plant community composition was measured by visual estimates of the percent cover of each plant species. Protocol details are available at www.nutnet.org and in (1).

Location of data collection: 71 sites on six continents. Site latlongs are included in the site_year_means dataset.
Date of file creation: 1st draft 14 January 2025 03 December 2020

Licenses or restrictions: none

Missing value designation: empty cells.

Files and variables

Metadata Tab:

Column A: A combined list of variable names used across the two datasets site_means and site_year_means.
Column B: Descriptions of the variables named in column A.

Site_means Tab:

Treatment means and standard errors of live_mass, effective species richness, Evenness,and plot_beta by site, averaged across years. Variable names defined in the metadata tab. Missing values (empty cells) occur in the soils variables (pct_N, ppm_P, ppm_K, PercentSand, PercentSilt, PercentClay) in two cases: 1) no data were provided for a site or 2) for the texture variables, data were collected only from control plots.

Site_year_means Tab:

Treatment means and standard error (SE) of live_mass, and number of reps by site and year. Variable names defined in the metadata tab.

Version Changes

02 April 2025: Uploaded files containing SAS code used to perform the analysis and make the graphics used in this paper. Files containing SAS code have the .sas extension, and the filename begins with a number indicating the order in which the code in the files should be run. The datafile Fay_PNAS_01142025 was not revised in this update.

 

Notes on the code files

1_libnames_options_varlists.sas: assigns library locations, creates variable lists, and sets certain SAS options. Beware that editing the variable lists will alter output from nearly every following program.

2_Read_NutNet.sas: Reads the raw dataset obtained from the Nutrient Network and performs some initial processing.

3_impute_and_clean.sas: Performs data cleaning and imputation steps as described in the published SI_Appendix.

4_build_precip.sas: Assembles precipitation data sets for joining with the raw dataset.

5_build dataset: joins the precipation dataset to the raw dataset, selects sites and observations meeting inclusion criteria, aligns precipitation years with biomass years, cross-validates the various precipitation data sources, produces the datasets used in subsequent analyses.

6_site nut effects LSM meth.sas: Fits a linear mixed model to each site, produces least square means and their p-values for use in the rubric assigning sites to the categories of nutrient limitation. The code is written to for interactions for N and P but can be modified to test other pairs of nutrients.

7_compute graph means: generates means and standard errors organized for graphing in OriginPro.

8_export graph means: Writes excel files containing the datasets produced in the preceding code file.

9_sitelist_table.sas: Produces Table S1 in the SI Appendix.

10_Mixed trts precip and temp vars: Fits various linear mixed models testing 1) which climate variables yield best fit to biomass across sites, 2) how many and which nutrients limit or co-limit biomass and influence plant species diversity variables.

11_export mixed trts tables: Writes excel files containing results from the mixed models analyses.

12_Mixed NP forms.sas: Fits linear mixed models across sites within four forms of nutrient limitation, and related sub-analyses.

13_Export mixed NPforms tables.sas: Writes an excel file used as Table S6 in SI Appendix.

14_Mgmt & Soils.sas: Fits models testing whether the classified sites differ in management, soils, and other site-level properties.

15_SEMs.sas: Fits the structural equation models and produces output tables.

Literature Cited:

1. E. T. Borer et al., Finding generality in ecology: a model for globally distributed experiments. Methods Ecol. Evol. 5, 65-73 (2014).