The RStudio project directory EAB_soilflux.Rproj contains all of the code to reproduce the analysis (from the raw data) for the manuscript:
Matthes, J.H., A.K. Lang, F.V. Jevon, S.J. Russell. Tree stress and mortality from emerald ash borer does not systematically alter short-term soil carbon flux in a mixed northeastern U.S. forest. Submitted to Forests, 21 Dec 2017.
The file EAB_soilflux_workflow.R steps through the analysis presented within the manuscript using functions with the R/ directory and data from the input/ directory. The workflow produces output files and figures that match those presented within the manuscript to the output/ directory.
clean_experiment_data directory: Contains the calculated soil CO2 and CH4 flux measurements and associated covariates that will be submitted to the Dryad data repository upon publication of the manuscript.
input/ directory:
- climate_NOAA/: weather data from the Concord Airport (GHCND:USW00014745) station from 1981-2010 (climatology data) and 2015-2017 (experiment data)
- raw_CO2CH4_data/: raw data files from the Los Gatos Research Ultraportable Greenhouse Gas Analyzer and the Picarro GasScouter analyzer used for soil-atmosphere trace gas measurements
- TF_insttempmoisture.csv: instantaneous soil moisture and soil temperature data collected at the time of flux measurement
- TF_SoilCovariates.csv: soil covariates (pH, texture, root biomass) measured at the conclusion of the experiment
- TF_TreeHealth.csv: EAB impact status of study trees within the experiment
- times_key.csv: flux measurement start times that correspond to the soil-atmosphere analyzer measurements
output/ directory:
- data/: output tables produced by the workflow code that corresponds to the Bayesian mixed models within the manuscript
- figures/: figures produced from the analysis that are described within the manuscript
R/ directory:
- functions/: set of R functions loaded and called within the workflow.R script to format raw data and to create figures with data and model output
- stan/: the Stan Bayesian mixed models code used to fit the models described within the manuscript