Data from: Necromass carbon but not microbes constrain soil carbon release in restoration of degraded alpine grassland
Data files
Feb 06, 2026 version files 35.73 KB
-
AlpineGrassland_NecromassC_Data.xlsx
26.74 KB
-
README.md
8.99 KB
Abstract
Alpine grassland restoration, a critical strategy for enhancing soil organic carbon (SOC) sequestration in high-altitude ecosystems, profoundly influences plant‒soil‒microbe interactions that govern the magnitude of carbon (C)-climate feedback. However, mechanisms driving plant and microbial regulation of SOC mineralization (i.e., soil CO2-C release) during degraded alpine grassland restoration remain unresolved, limiting predictions of SOC cycling in these vulnerable ecosystems. Here, by integrating passive and active restoration experiments with aerobic incubation, high-throughput sequencing, and biomarker analyses, we disentangled how restoration-induced shifts in SOC composition (plant- and microbial-derived C) and microbial activity and diversity regulate soil CO2-C release in degraded alpine grassland on the Qinghai-Tibetan Plateau. Our results showed that soil CO2-C release increased significantly with restoration progression under both passive and active approaches. Alpine grassland restoration markedly enhanced plant-derived C accumulation and its SOC contribution, while microbial-derived C remained unchanged due to reduced necromass accumulation coefficients. Notably, although active restoration accelerated plant-derived C accumulation, its oxidation decomposition degree was lower compared to passive restoration and even to unrestored heavily degraded grasslands, increasing the SOC pool lability. Fungal community restructuring, particularly in the saprophytic fungal community, emerged as a hallmark of restoration. More importantly, we found that elevated soil CO2-C release during degraded alpine grassland restoration was not primarily mediated by microbial activity and diversity shifts but strongly linked to divergent plant- and microbial-derived C accumulation patterns, especially the dynamics of plant-derived C. These insights underscore the critical roles of plant- and microbial-derived C redistribution in grassland restoration and suggest new mechanisms for restoration-induced soil C dynamics.
Dataset DOI: 10.5061/dryad.j9kd51ct1
Description of the data and file structure
The data collected in this study aims to investigate the limiting factors of soil carbon release during the restoration of degraded alpine grasslands, with a particular focus on the relative roles of plant-derived carbon and microbial-derived carbon (necromass carbon). The experiment encompasses grassland sites with varying degrees of degradation (from non-degraded to heavily degraded) and restoration approaches (active and passive restoration). Multiple sets of indicators were measured, including soil physicochemical properties, plant- and microbial-derived carbon compounds, microbial community biomass and diversity, and soil carbon mineralization rates, in order to decipher the key biogeochemical factors driving soil carbon dynamics along the restoration gradient.
Files and variables
File: AlpineGrassland_NecromassC_Data.xlsx
Description: This file contains the metadata and data dictionary for the entire dataset. The table provides a comprehensive list of all variables (attributes), their descriptions, and their units of measurement.
Variables
| Attribute (Abbreviation) | Description | Category/Units |
|---|---|---|
| PRC | Passive restoration | – |
| ARC | Active restoration | – |
| HD | Heavy degradation | – |
| MD | Moderate degradation | – |
| LD | Light degradation | – |
| ND | Non-degradation | – |
| A3 | Active restoration (3 years) | – |
| A7 | Active restoration (7 years) | – |
| SOC | Soil organic carbon content | mg/g soil |
| TN | Total nitrogen content | mg/g soil |
| C:N | Soil organic carbon-to-total nitrogen ratio | – |
| Clay | Clay content | % |
| Silt | Silt content | % |
| Sand | Sand content | % |
| pH | Soil pH | – |
| SWC | Soil water content | % |
| Tlp | Total lignin phenols | mg/kg soil |
| S | Syringyl monomer | mg/kg soil |
| V | Vanillyl monomer | mg/kg soil |
| C | Cinnamyl monomer | mg/kg soil |
| plant-derived C | Plant-derived carbon | % |
| (Ad/Al)v | Acid-to-aldehyde ratio (vanillyl) | – |
| (Ad/Al)s | Acid-to-aldehyde ratio (syringyl) | – |
| GlcN | Glucosamine | mg/g soil |
| GalN | Galactosamine | mg/g soil |
| ManN | Mannosamine | mg/g soil |
| MurA | Muramic acid | mg/g soil |
| TAS | Total amino sugars | mg/g soil |
| FNC | Fungal necromass carbon | mg/g soil |
| BNC | Bacterial necromass carbon | mg/g soil |
| MNC | Microbial necromass carbon | mg/g soil |
| MNC/SOC | Contribution of microbial necromass carbon to SOC | % |
| NAC | Necromass accumulation coefficient (MNC/MBC) | – |
| MBC | Microbial biomass carbon | mg/g soil |
| MBN | Microbial biomass nitrogen | mg/g soil |
| FS | Fungal Shannon index | – |
| FA | Fungal ACE index | – |
| BS | Bacterial Shannon index | – |
| BA | Bacterial ACE index | – |
| CO2-C | Soil CO₂–C release | mg CO₂–C g⁻¹ soil |
Code/software
All statistical analyses were performed using R version 4.2.2 (R Core Team, 2023). The data files (e.g., .xlsx，.csv) can be viewed with any standard spreadsheet software or text editor.
Key R Packages for Analysis
The following R packages, along with their primary use in this study, were essential for the statistical workflow:
- piecewiseSEM: Used to construct and evaluate the piecewise structural equation model (SEM) to assess direct and indirect effects of biotic and abiotic factors on soil CO2-C release.
- randomForest: Employed to run the random forest model for evaluating and ranking the relative importance of various soil variables on CO2-C release.
- rfPermute: Used in conjunction with randomForest to determine the statistical significance of variable importance via permutation tests.
- vegan: Utilized for conducting Principal Component Analysis (PCA) to create comprehensive metrics (PC1) from correlated variable groups for input into the SEM.
- stats (base R): Used for checking assumptions of homogeneity of variance and normality, and for performing one-way analysis of variance (ANOVA) to test for differences among treatment groups.
Analytical Workflow
The statistical workflow proceeded in the following logical sequence, with each step dependent on the output of the previous one:
- Data Preparation & Assumption Checking: Data were imported and cleaned. Assumptions of homogeneity of variance and normality were checked prior to parametric tests.
- Group Comparisons (ANOVA): One-way ANOVA was used to test for significant differences in soil biotic and abiotic factors across restoration treatments.
- Dimensionality Reduction (PCA): PCA was performed separately on highly correlated variable groups (plant-derived C, microbial-derived C, microbial biomass/diversity, edaphic factors) to extract the first principal component (PC1) for each group, which was used as a composite metric in subsequent modeling.
- Path Analysis (SEM): A piecewise SEM was built using the piecewiseSEM package to evaluate the network of direct and indirect influences on soil CO2-C release. Model fit was assessed using Fisher’s C statistic, Akaike information criterion (AIC), and associated p-values.
- Variable Importance (Random Forest): The randomForest and rfPermute packages were used to train a model predicting CO2-C release from all measured soil variables. The analysis ranked predictor importance based on the percentage increase in mean squared error (%IncMSE) and tested the significance of each variable set.
Access information
Other publicly accessible locations of the data:
- None. This dataset is submitted here for the first time and is not currently stored in any other public repository or platform.
Data was derived from the following sources:
- This dataset consists of original data specifically collected and generated by the authors for this study. It is not derived from any pre-existing datasets. All data were obtained through the field sampling and laboratory analyses conducted for this research.
- Associated Publication: This dataset supports the accepted manuscript titled "Necromass carbon but not microbes constrain soil carbon release in restoration of degraded alpine grassland" (Manuscript ID: EAP25-0349) in the journal Ecological Applications.