Active restoration of degraded alpine grassland weakens mineral-associated soil organic carbon retention
Abstract
Active restoration is predominant to reverse grassland degradation and soil organic carbon (SOC) loss. However, early studies focused on the bulk soil carbon (C) dynamics upon grassland degradation and restoration, with limited evidence involving SOC fractions. Here, we explored how the particulate and mineral-associated OC (POC and MAOC), metal-bound OC, soil aggregate-associated OC, SOC molecular structure, and microbial-derived OC respond to grassland degradation and restoration on the Tibetan Plateau, and also studied the effects of active restoration on SOC fractions across grassland ecosystems through a global-scale meta-analysis. We found that POC and soil aggregate-associated OC contents substantially decreased, while MAOC content remained stable after grassland degradation. In contrast, short-term active restoration induced significant decrease in MAOC and metal-bound OC due to lower soil pH, but POC content remained stable, indicating that grassland restoration resulted in a decrease in the SOC stability. Moreover, soil aggregate structure and its stability recovered rapidly during short-term active restoration, but aggregate-associated OC did not recover or even declined, suggesting asynchrony between them. Nuclear magnetic resonance spectroscopy revealed no significant changes in the SOC molecular composition between intact, degraded, and restored grasslands, suggesting that molecular structure alone does not mirror SOC stability. Furthermore, the meta-analysis revealed that the response of bulk SOC, MAOC and POC in degraded grassland to active restoration occurred sequentially in three phases characterized by vulnerability, retention, and re-loss of SOC, respectively.
Synthesis and applications
Combining a field experiment with a global synthesis, we found that grassland restoration induces divergent shifts in SOC fractions, with threshold effects during grassland restoration chronosequence. Relying solely on overall changes in bulk SOC may misrepresent the true impact of grassland restoration on SOC retention. Therefore, it is essential to pay more attention to the response of the functionally distinct SOC fractions to grassland restoration, which could help us more accurately assess and predict the grassland soil C-climate feedback.
Data files include:
Supplementary_data_1.xlsx: Experimental design parameters, edaphic properties, physico-chemical protection, and soil organic carbon (SOC) content (total and fractions) for paired grassland treatments (intact, degraded, restored) across three geographically distinct sites. All data were collected by the authors. Blank cells represent missing data.
| variables | units | description |
|---|---|---|
| Land type | categorical | the paired treatments (i.e., intact, degraded and restored grassland) |
| Site | categorical | the paired treatments by three geographically spaced demonstration sites |
| Sampling | categorical | five replicates measured 30 m × 30 m at different land types |
| SOC | mg / g dry soil | soil organic carbon |
| POC | mg / g dry soil | particulate organic carbon |
| MAOC | mg / g dry soil | mineral-associated organic carbon |
| Metal-OC | mg / g dry soil | metal-bound organic carbon |
| Fe-OC | mg / g dry soil | iron- bound organic carbon |
| Ca-OC | mg / g dry soil | calcium-bound organic carbon |
| GRSP | mg / g dry soil | glomalin-related soil protein. |
| Macroaggregate | % | macroaggregate(2000-250 um) |
| Microaggregate | % | microaggregate(250-53um) |
| Clay+Silt | % | clay+silt fraction(<53 um) |
| MWD | mean weight diameter | |
| Macro-OC | mg / g dry soil | macroaggregate-associated organic carbon |
| Micro-OC | mg / g dry soil | microaggregate-associated organic carbo |
| Clay+Silt-OC | mg / g dry soil | clay+silt-associated organic carbon |
| Fed | mg / g dry soil | dithionite-extractable Fe oxides |
| Caexe | mg / g dry soil | exchangeable calcium |
| pH | soil pH level | |
| SWC | % | soil water content |
| TN | mg / g dry soil | total nitrogen |
| C:N | soil organic carbon/total nitrogen | |
| Clay | % | volume fraction of clay(0-2 μm) |
| Silt | % | volume fraction of silt(2-50 μm) |
| Sand | % | volume fraction of sand(50-2000 um) |
| CS | % | the sum of clay and silt volume fractions |
Supplementary_data_2.xlsx: Molecular composition of SOC derived from solid-state 13C NMR spectroscopy for paired grassland treatments (intact, degraded, restored). All data were collected by the authors. Blank cells represent missing data.
| variables | units | description |
|---|---|---|
| Land type | categorical | the paired treatments (i.e., intact, degraded and restored grassland) |
| Alkyl C | % | Alkyl carbon (0–45 ppm) |
| N-alkyl and methoxyl C | % | N-alkyl and methoxyl carbon (45–60 ppm) |
| O-alkyl C | % | O-alkyl carbon (60–95 ppm) |
| Di-O-alkyl C | % | Di-O-alkyl carbon (95–110 ppm) |
| Aromatic C | % | Aromatic carbon (110–145 ppm) |
| Phenolic C | % | Phenolic carbon (145–165 ppm) |
| Amide/carboxyl C | % | Amide/carboxyl carbon (165–210 ppm) |
| A/A-O | Alkyl C:O-alkyl C ratio (0–45/45–110 ppm) | |
| Aromaticity | aromatic C peak area (110–160 ppm)/total peak area (0–160 ppm) |
Supplementary_data_meta_analysis.xlsx: A global meta-analysis dataset evaluating the effects of grassland restoration on bulk SOC and its fractions (POC, MAOC). Includes study-level metadata, effect sizes, and statistical parameters. All data were collected by the authors. Blank cells represent missing data.
| variables | units | description |
|---|---|---|
| Study | number | study number |
| Citations | data source | |
| Longitude | degrees | longitude coordinate |
| latitude | degrees | latitude coordinate |
| MAT | ℃ | mean annual air temperature |
| MAP | mm | mean annual precipitation |
| Climatic zones | categorical | climate classification |
| Restoration year | number | year of active restoration |
| Cn | number | sample size of control groups |
| Tn | number | sample size of treatment groups |
| C_M_SOC | mg / g dry soil | mean soil organic carbon (SOC) content in control groups |
| C_SD_SOC | mg / g dry soil | standard deviation of SOC in control groups |
| T_M_SOC | mg / g dry soil | mean SOC in treatment groups |
| T_SD_SOC | mg / g dry soil | standard deviation of SOC in treatment groups |
| Vi_SOC | variance of the natural logarithm of response ratios for SOC | |
| RR_SOC | natural logarithm of response ratios for SOC | |
| C_M_POC | mg / g dry soil | mean particulate organic carbon (POC) content in control groups |
| C_SD_POC | mg / g dry soil | standard deviation of POC in control groups |
| T_M_POC | mg / g dry soil | mean POC in treatment groups |
| T_SD_POC | mg / g dry soil | standard deviation of POC in treatment groups |
| Vi_POC | variance of the natural logarithm of response ratios for POC | |
| RR_POC | natural logarithm of response ratios for POC | |
| C_M_MAOC | mg / g dry soil | mean mineral-associated organic carbon (MAOC) content in control groups |
| C_SD_MAOC | mg / g dry soil | standard deviation of MAOC in control groups |
| T_M_MAOC | mg / g dry soil | mean MAOC in treatment groups |
| T_SD_MAOC | mg / g dry soil | standard deviation of MAOC in treatment groups |
| Vi_MAOC | variance of the natural logarithm of response ratios for MAOC | |
| RR_MAOC | natural logarithm of response ratios for MAOC |