Dataset DOI: 10.5061/dryad.p2ngf1w77

Description of the data and file structure

This repository contains processed and derived data, model code, and analysis scripts associated with the study “Elevated CO₂ mitigated warming-enhanced carbon loss from northern peatland.” The data were collected as part of the Spruce and Peatland Responses Under Changing Environments (SPRUCE) experiment, which investigates the responses of a northern peatland ecosystem to whole-ecosystem warming and elevated atmospheric CO₂. The experiment includes five warming levels (+0, +2.25, +4.5, +6.75, and +9 °C) under ambient and elevated CO₂ conditions. The repository is organized into three main components: (1) processed and derived data used in the analyses, (2) Python analysis and plotting code, and (3) the TECO-SPRUCE model code and associated model inputs and outputs.

Files and variables

File: data_sources.zip

Description: Contains the processed and derived data used directly for the analyses and figures presented in the associated article. All tabular data files are provided in Microsoft Excel (.xlsx) format. 
The archive contains the following major files and directories:

• 1_sourcedata.xlsx: Processed data used for the primary analyses of ecosystem carbon dynamics and their responses to warming and elevated CO₂.
• 1_sourcedata_minus_2013.xlsx: Data corresponding to the primary source data after accounting for pretreatment/baseline differences among experimental enclosures.
• 2_sourcedata_noacc.xlsx: Processed data from the analysis excluding photosynthetic acclimation.
• 2_sourcedata_noacc_minus_2013.xlsx: Corresponding data from the no-acclimation analysis after accounting for pretreatment/baseline differences among experimental enclosures.
• 3_simu_vs_obs/: Data used to compare TECO-SPRUCE model estimates with independent or corresponding observations.
• 4_water_table_fluxes_minus_2013.xlsx: Processed water-table and carbon flux data used to examine relationships between hydrological conditions and ecosystem carbon dynamics.
• 5_community_structure_data.xlsx: Data describing vegetation community structure and plant functional type responses.
• 6_parameters_for_plot.xlsx: TECO-SPRUCE model parameter information used for the experimental plots.
• 7_NCE_baseline_sensitivity_plotdata.xlsx: Data used for the sensitivity analysis of net carbon exchange (NCE) to baseline treatment differences.

The experimental plots are identified by plot/enclosure IDs. The warming treatments are expressed in °C above ambient conditions, and the CO₂ treatments distinguish ambient and elevated atmospheric CO₂ conditions. Major carbon-cycle variables include gross primary productivity (GPP), ecosystem respiration (ER), autotrophic respiration (Ra), heterotrophic respiration (Rh), net primary productivity (NPP), aboveground NPP (ANPP), belowground NPP (BNPP), net ecosystem exchange (NEE), net CO₂ exchange (NCE), and methane (CH₄) flux. Carbon pool variables include leaf carbon (cLeaf), stem carbon (cStem), total plant carbon (cPlant), soil carbon (cSoil), and carbon pool size (C). Other variables include tree leaf photosynthetic rate (photo), leaf area index (LAI or lai), water table depth (zwt), maintenance respiration (rMaint), and growth respiration (rGrowth). Carbon fluxes and productivity are generally reported in g C m⁻² yr⁻¹ for annual quantities, while carbon pools are reported in g C m⁻². Leaf area index is expressed as m² leaf area m⁻² ground area.

File: TECO-SPRUCE_code_v20251209.zip

Description: Contains the version of the TECO-SPRUCE model used for the model simulations and analyses reported in the associated article. TECO-SPRUCE is written primarily in Fortran and requires a compatible Fortran compiler to compile and run. The archive includes:

• TECO_inputs/: model input data;

• TECO_outputs/: model simulation outputs;

• TECO_outputs_noacc/: outputs from simulations excluding photosynthetic

  acclimation;

• TECO_outputs_opt/: model outputs associated with optimized parameter sets;

• TECO-SPRUCE_v3.1/configs/: model configuration files;

• TECO-SPRUCE_v3.1/scripts/: supporting model scripts;

• TECO-SPRUCE_v3.1/src/: Fortran source code;

• README files containing additional model information.

• TECO-SPRUCE_v3.1/README.md: Introduction to how to run TECO-SPRUCE model.

The model source directory includes the Fortran modules required for the TECO-SPRUCE simulations. Model outputs were subsequently processed and analyzed using the Python scripts provided in code_sources.zip.

The same P## identifiers are used throughout the TECO_inputs, TECO_outputs, TECO_outputs_noacc, and TECO_outputs_opt directories to identify plot-specific model inputs and outputs. Files within each plot folder therefore correspond to the treatment combination listed above.

File: code_sources.zip

Description: Contains the analysis and plotting code used to process the data, perform the statistical analyses, and generate the results and figures presented in the paper. The scripts use the processed datasets provided in data_sources.zip. They include analyses of ecosystem carbon fluxes, warming and CO₂ treatment responses, plant functional type responses, sensitivity analyses, and figure generation.

Code/software

The data analyses and figure generation were primarily conducted using Python 3. The main Python packages used include pandas, NumPy, SciPy, Matplotlib, statsmodels, scikit-learn, and pyGAM. The analysis and plotting scripts are provided in the code_source/ directory and use the datasets provided in the data_source/ directory.

The Jupyter notebooks were run using Python 3.12.0. The main Python packages
used for data analysis, statistical analyses, and figure generation were:

• pandas 2.2.3
• NumPy 1.26.4
• Matplotlib 3.9.2
• seaborn 0.13.2
• statsmodels 0.14.0
• SciPy 1.15.2

The notebooks were executed in a Jupyter environment. Required input datasets
are provided in data_sources.zip. File paths used in the notebooks are
relative to the original project directory structure and may need to be
adjusted if the files are placed in a different directory.

The TECO-SPRUCE model was compiled and run using a Fortran compiler. The complete model code used in this study is provided in the TECO-SPRUCE_code_v20251209/ directory. Model outputs were subsequently processed and analyzed using the scripts provided in code_source/.

All software required to access the data files and run the analysis code is freely available or open source.

Access information

Other publicly accessible locations of the data:

• The data used and processed in this study were derived in part from publicly available datasets from the SPRUCE (Spruce and Peatland Responses Under Changing Environments) experiment. The original datasets are publicly accessible through their respective SPRUCE/ESS-DIVE repository records and are subject to the data-use and licensing terms specified by those repositories.

Data was derived from the following sources:

• SPRUCE shrub-layer growth assessments: https://doi.org/10.25581/spruce.052/1433837
• SPRUCE tree growth assessments of Picea and Larix: https://doi.org/10.25581/spruce.051/1433836
• SPRUCE Sphagnum productivity and community composition: https://doi.org/10.25581/spruce.049/1426474
• SPRUCE peat physical and chemical characteristics: https://doi.org/10.25581/spruce.102/1878603
• SPRUCE root production resolved to plant functional type: https://doi.org/10.25581/spruce.127/2570059
• SPRUCE large-collar in situ CO₂ and CH₄ fluxes: https://doi.org/10.3334/cdiac/spruce.034
• SPRUCE leaf area index data: https://doi.org/10.25581/spruce.058/1491566
• SPRUCE photosynthesis and respiration of Picea mariana and Larix laricina, 2016–2017: https://doi.org/10.25581/spruce.056/1455138
• SPRUCE photosynthesis and respiration of Picea mariana and Larix laricina, 2019: https://doi.org/10.25581/spruce.121/2503934