Understanding future big sagebrush regeneration potential

Contact: Daniel R Schlaepfer ORCiD

This repository contains the code developed for the manuscript

Understanding the future of big sagebrush regeneration: challenges of projecting complex ecological processes

by Daniel R Schlaepfer, John B Bradford, William K Lauenroth, and Robert K Shriver

Table of contents

1. File structure
2. Data
3. Overview
4. Dependencies

File structure

The code assumes the following organization of directories and files:

• 1_Simulation_Setup/ holds netCDF files that describe and/or define the underlying simulation experiment; see section Data

• 1_Additional_Setup/ holds additional files that describe and/or define the underlying simulation experiment, e.g., "SFSW2_project_variables_metadata.csv" describing the variables and their organization in netCDF files

• 2_Simulation_Output/ holds netCDF files that contain the simulation experiment output used by the analysis; see section Data

• 3_Prepared_Data/ holds intermediate files generated by the analysis code

• 4_Analysis_Results/ will hold the analysis results including figures and tables produced by executing the R scripts

• Code/ contains the R scripts to run the analysis

Data

Data are available as netCDFs from the USGS ScienceBase-Catalog at https://doi.org/10.5066/P9MB2QB8

The content of the group "Simulation_Setup" (i.e., all files and sub-folders) should be copied to folder 1_Simulation_Setup/. The remaining groups should be copied (without renaming) to folder 2_Simulation_Output/.

There is a separate netCDF file for each variable simulated with SOILWAT2 under separate soil types (where appropriate) and forced by a separate climate forcing data set (i.e., historical observed meteorological data or future projections derived from one of 37 global circulation models under one of two representative concentration pathways).

Overview

The files SagebrushRegeneration_1_Functions.R and SagebrushRegeneration_2_DescribeExperimentAnalysis.R

contain specific functions used for the analysis and prepare data structures (read in from 1_Simulation_Setup/) that define and organize the simulations.

The analysis code is organized in several independent scripts that each produce a part of the overall results:

• Objective 1: Geographic patterns of big sagebrush regeneration

  • SagebrushRegeneration_3_Part1a_HistoricalPatterns.R for Figure 1 "4_Analysis_Results/Output1_GeographicPatterns/HistoricalPatterns/Fig_Maps_GISSM_vs_Shriver2018_Current-SoilSiteSpecific_v4.pdf", Table 1, Appendix S1 Table S3, Appendix S2 Figure S1

  • SagebrushRegeneration_3_Part1b_FuturePatterns.R for Figure 3 "4_Analysis_Results/Output1_GeographicPatterns/FuturePatterns/Fig_Maps_GISSM_vs_Shriver2018_SoilSiteSpecific_median_hybrid-delta-3mod.d90yrs.RCP85_v4.pdf", Table 1, Appendix S1 Table S3, Appendix S2 Figure S2

• Objective 2: Robustness of projections

  • SagebrushRegeneration_3_Part2a_Uncertainty-Agreement.R for Figure 4 "4_Analysis_Results/Output2_Uncertainty/Fig_Agreement_GISSM_vs_Shriver2018_SoilSiteSpecific_median_hybrid-delta-3mod.d90yrs.RCP85_90perc-agreement_v4.pdf", Table 1, Appendix S2 Figure S3

  • SagebrushRegeneration_3_Part2b_Uncertainty-Sensitivity.R for Figure 5 "4_Analysis_Results/Output2_Uncertainty/Fig_Sensitivity_ExperimentalFactors_resps_v4.pdf", Appendix S1 Table S4, Appendix S2 Figure S4

  • SagebrushRegeneration_3_Part2d_Uncertainty-Extrapolations.R for Appendix S1 Table S5, Appendix S2 Figure S5-S8,

• Objective 3: Regeneration patterns in relation to climate and soils

  • SagebrushRegeneration_3_Part3a_EnvRelationships_Shriver2018.R for Appendix S2 Figure S9

  • SagebrushRegeneration_3_Part3b_GeneralEnvRelationships.R for Figure 6 "4_Analysis_Results/Output3_EnvRelationships/Current_GeneralEnvRelationships/Fig_VisReg_GeneralEnvRelationships_Current_nclusters2.pdf", Table 2, Appendix S1 Table S2, Appendix S1 Table S6

  • SagebrushRegeneration_3_Part3c_SoilDifferenceRelationships.R for Table 2, Appendix S2 Figure S10

  • SagebrushRegeneration_3_Part3d_GeneralFutureDiffEnvRelationships.R for Figure 7 "4_Analysis_Results/Output3_EnvRelationships/FutureDelta_GeneralEnvRelationships/Fig_VisReg_FutureDiffEnvRelationships_merged-ensemble_hybrid-delta-3mod.d90yrs.RCP85_v4.pdf", Table 2, Appendix S1 Table S7, Appendix S2 Figure S11-S13

Dependencies

The R code utilizes several R packages available from CRAN or github including

• raster, sp, sf, and ncdf4 packages

• rSW2st and rSW2analysis for manipulation of SOILWAT2/rSOILWAT2 simulations; they can be installed with

  remotes::install_github("DrylandEcology/rSW2st")
  remotes::install_github("DrylandEcology/rSW2analysis")