Publications

Soil respiration (Rs) is projected to be substantially affected by climate change, impacting the storage, equilibrium, and movement of terrestrial carbon (C). However, uncertainties surrounding the responses of Rs to climate change and soil nitrogen (N) enrichment are linked to mechanisms specific to diverse climate zones. A comprehensive meta-analysis was conducted to address this, evaluating the global effects of warming, increased precipitation, and N enrichment on Rs across various climate zones and ecosystems. Data from 123 studies, encompassing a total of 10,377 worldwide observations, were synthesized for this purpose. Annual Rs were modeled and their uncertainties were associated with a 1-km2 resolution global Rs database spanning from 1961 to 2022. Calibrating Rs using ensemble machine learning (EML) and employing 10-fold cross-validation, 13 environmental covariates were utilized. The meta-analysis findings revealed an upsurge in Rs rates in response to warming, with tropical, arid, and temperate climate zones exhibiting increases of 12 %, 13 %, and 16 %, respectively. Furthermore, increased precipitation led to stimulated Rs rates of 11 % and 9 % in tropical and temperate zones, respectively, while N deposition affected Rs in cold (+6 %) and tropical (+5 %) climate zones. The machine learning technique estimated the global soil respiration to range from 91 to 171 Pg C yr−1, with an average Rs of 700 ± 300 g C m−2 yr−1. The values ranged between 314 and 2500 g C m−2 yr−1, with the lowest and highest values observed in cold and tropical zones, respectively. Spatial variation in Rs was most pronounced in low-latitude areas, particularly in tropical rainforests and monsoon zones. Temperature, precipitation, and N deposition were identified as crucial environmental factors exerting significant influences on Rs rates worldwide. These factors underscore the interconnectedness between climate and ecosystem processes, therefore requiring explicit considerations of different climate zones when assessing responses of Rs to global change.

Paddy fields, as the largest anthropogenic wetlands on Earth, face a high risk of micronutrient loss through surface runoff and leaching due to their frequent irrigation-drainage cycles, as well as removal with crop harvest. While micronutrient's losses largely impede biological nitrogen fixation (BNF) in soils, agricultural practices and mechanisms that retain micronutrients and thus increase BNF in paddy soils remain underexplored. Here we showed that the long-term (40 years) application of fertilizers increased the contents of microbial necromass in paddy soils by 20%-43% compared to the unfertilized control. In particular, long-term organic fertilizations increased poorly crystalline minerals through intensifying mineral weathering, which further contributed to the increased stable carbon burials in paddy soils. Synchrotron radiation based spectromicroscopy analysis provided direct evidence demonstrating a differential control of mineral hydroxyl on C functional groups at the submicron scale in paddy soils. Notably, microbial necromass and short-range ordered minerals had a strong correlation with the content and bioavailability of micronutrients in paddy soils, indicating their key roles in the retention of micronutrients. Metagenomic sequencing analysis further indicated that the content and bioavailability of micronutrients were strongly correlated with the abundance of the key N-fixing genera (i.e., Azospirillum and Bradyrhizobium). Unexpectedly, structural equation modeling (SEM) identified that microbial necromass exerted the strongest control on N-fixing genera, highlighting an underappreciated role of microbial necromass as a reservoir of micronutrients. Based on micronutrient's bioavailability and metagenomic sequencing, we conclude that micronutrients are the key factor for BNF in paddy soils, offering significant implications for managing BNF in paddy soils.

We propose and successfully applied a new approach for 3-source-partitioning based on a combination of 14C labeling with 13C natural abundance. By adding 14C-labeled glucose to soil after C3 – C4 vegetation change, we partitioned three C sources in three compartments, namely CO2, microbial biomass and dissolved organic C (DOC). This enabled us to estimate mechanisms and sources of priming effects (PE). Glucose application at low and high rate (GL: 100 and GH: 1000 μg C g−1, respectively) caused positive PE both short-term (during 1–3 days) and long-term (3–55 days). Despite a 10-fold difference in the amount of substrate added, the PE observed was larger by a factor of only 1.6 at the high versus low rate of glucose. The real and apparent priming effects were distinguished by partitioning of microbial C for glucose-C and SOM-derived C. As the amount of primed CO2 respired during short-term PE was 40% lower than microbial C, and the contribution of soil C in microbial biomass did not increase, we concluded that such short-term PE was apparent and was mainly caused by accelerated microbial turnover (at GL) and by pool substitution (at GH). Both the amount of primed CO2–C, which was 1.3–2.1 times larger than microbial C, and the increased contribution of soil C in microbial biomass allowed us to consider the long-term PE as being real. The sole source of real PE (GL treatment) was the “recent” soil organic matter, which is younger than 12-year-old C. The real PE-induced by a glucose amount exceeding microbial biomass (GH) was due to the almost equal contribution of ‘recent’ (<12 years) and ‘old’ (>12 years) C. Thus, the decomposition of old recalcitrant SOM was induced only by an amount of primer exceeding microbial C. We conclude that combining 14C labeling with 13C natural abundance helped disentangle three C sources in CO2, microbial biomass and DOC and evaluate mechanisms and sources of PE.

No abstract is provided for this article.

Summary Microbial elevational diversity patterns have been extensively studied, but their shaping mechanisms remain to be explored. Here, we examined soil bacterial and fungal diversity and community compositions across a 3.4 km elevational gradient (consists of five elevations) on Mt. Kilimanjaro located in East Africa. Bacteria and fungi had different diversity patterns across this extensive mountain gradient—bacterial diversity had a U shaped pattern while fungal diversity monotonically decreased. Random forest analysis revealed that pH (12.61% importance) was the most important factor affecting bacterial diversity, whereas mean annual temperature (9.84% importance) had the largest impact on fungal diversity, which was consistent with results obtained from mixed‐effects model. Meanwhile, the diversity patterns and drivers of those diversity patterns differ among taxonomic groups (phyla/classes) within bacterial or fungal communities. Taken together, our study demonstrated that bacterial and fungal diversity and community composition responded differently to climate and edaphic properties along an extensive mountain gradient, and suggests that the elevational diversity patterns across microbial groups are determined by distinct environmental variables. These findings enhanced our understanding of the formation and maintenance of microbial diversity along elevation, as well as microbial responses to climate change in montane ecosystems.

A quantitative assessment of the carbon balance was performed in gray forest soils of the former agricultural lands abandoned in different time periods in

Despite broad information on greenhouse gas (GHG) emissions from the soil surface, only few studies have examined the depth related GHG concentration and production within the soil. We coupled chamber surface GHG flux measurements with analysis of subsurface GHG concentrations at five depths (0, 10, 30, 50 and 70 cm) using silicon tubes in a grassland and cropland soils in a semi-arid agro-pastoral ecotone over a full year. Grassland and cropland soils behaved as net CO2 and N2O sources, but sinks of CH4. CO2 and N2O concentrations increased but CH4 decreased with soil depth. Subsurface GHG fluxes calculated using Fick’s law decreased with depth. The modeled GHG and measured surface fluxes agreed much better for CO2 than N2O or CH4. Spring-thaw and rainfall events triggered N2O pulses, CH4 uptake peaks and CH4 concentration drops. Annual emissions or uptake based on chamber method were 4820–7580 kg C ha−1 yr−1, 2.4–1.5 kg C ha−1 yr−1, and 0.13–0.14 kg N ha−1 yr−1 for CO2, CH4, and N2O across the grassland and cropland, respectively. Modeled annual accumulation of GHG fluxes dropped with depth, whereas the topsoil (0–20 cm) contributed about 65% for CO2, 68% for N2O and 72% for CH4 of the total profile. The vertical distribution of GHG fluxes reflected those of soil organic carbon and root biomass. The grassland had higher annual CO2 flux but lower CH4 uptake than the cropland. Annual N2O fluxes were similar between the two ecosystems.

Fractionation of soil organic carbon (SOC) is crucial for mechanistic understanding and modeling of soil organic matter decomposition and stabilization processes. It is often aimed at separating the bulk SOC into fractions with varying turnover rates, but a comprehensive comparison of methods to achieve this is lacking. In this study, a total of 20 different SOC fractionation methods were tested by participating laboratories for their suitability to isolate fractions with varying turnover rates, using agricultural soils from three experimental sites with vegetation change from C3 to C4 22–36 years ago. Enrichment of C4-derived carbon was traced and used as a proxy for turnover rates in the fractions. Methods that apply a combination of physical (density, size) and chemical (oxidation, extraction) fractionation were identified as most effective in separating SOC into fractions with distinct turnover rates. Coarse light SOC separated by density fractionation was the most C4-carbon enriched fraction, while oxidation-resistant SOC left after extraction with NaOCl was the least C4-carbon enriched fraction. Surprisingly, even after 36 years of C4 crop cultivation in a temperate climate, no method was able to isolate a fraction with more than 76% turnover, which challenges the link to the most active plant-derived carbon pools in models. Particles with density >2.8 g cm−3 showed similar C4-carbon enrichment as oxidation-resistant SOC, highlighting the importance of sesquioxides for SOC stabilization. The importance of clay and silt-sized particles (<50 μm) for SOC stabilization was also confirmed. Particle size fractionation significantly outperformed aggregate size fractionation, due to the fact that larger aggregates contain smaller aggregates and organic matter particles of various sizes with different turnover rates. An evaluation scheme comprising different criteria was used to identify the most suitable methods for isolating fractions with distinct turnover rates, and potential benefits and trade-offs associated with a specific choice. Our findings can be of great help to select the appropriate method(s) for fractionation of agricultural soils.

Carbon (C) derived from roots, rhizodeposition of living roots and dead root decomposition, plays a critical role in soil organic C (SOC) sequestration. Recent studies suggest that root inputs exert a disproportionate influence on SOC formation, and therefore, it is necessary to test separately the effects during root growth (i.e. rhizodeposition) and dead root decomposition (i.e. belowground litter) undernaturalconditions. A field experiment was carried out in grasslands with three typical plants: Stipa bungeana (St.B), Artemisia sacrorum (Ar.S), and Thymus mongolicus (Th.M) to differentiate the effects of root growth vs decomposition on C inputs by using in-growth soil cores and litter bag methods. For root growth experiment, the SOC content increased (2.5–3.8 g·kg−1) with root biomass, especially in soil under Ar.S. C input increased mineral-associated organic C (MAOC%) from 57% to 65%. The SOC δ13C was consistent with roots δ13C, indicating the roots were the primary source of SOC. The root decomposition experiment showed that the increase in SOC (0.74–2.6 g·kg−1) was highest at 90 days. Root decomposition rate was fast before 45 days, and SOC increased with the MAOC% during this period. After 45 days, particulate organic C % (POC%) was raised and was higher than in the control soil. The increase in SOC under root growth (2.5–3.8 g·kg−1) after one year was greater as compared to the rate under root decomposition (-0.4–0 g·kg−1). It suggested that C released during root growth was more effectively retained in the soil than that caused by dead root decomposition. The rise in MAOC during root growth and decomposition mainly explains the increase in SOC. Our results provided direct field based evidence illustrating the specific contribution of root growth and decomposition to SOC.