A meta-analysis highlights globally widespread potassium limitation in terrestrial ecosystems.

Potassium (K+ ) is the most abundant inorganic cation in plant cells, playing a critical role in various plant functions. However, the impacts of K on natural terrestrial ecosystems have been less studied compared with nitrogen (N) and phosphorus (P). Here, we present a global meta-analysis aimed at quantifying the response of aboveground production to K addition. This analysis is based on 144 field K fertilization experiments. We also investigate the influences of climate, soil properties, ecosystem types, and fertilizer regimes on the responses of aboveground production. We find that: K addition significantly increases aboveground production by 12.3% (95% CI: 7.4-17.5%), suggesting a widespread occurrence of K limitation across terrestrial ecosystems; K limitation is more prominent in regions with humid climates, acidic soils, or weathered soils; the effect size of K addition varies among climate zones/regions, and is influenced by multiple factors; and previous N : K and K : P thresholds utilized to detect K limitation in wetlands cannot be applied to other biomes. Our findings emphasize the role of K in limiting terrestrial productivity, which should be integrated into future terrestrial ecosystems models.

Disentangling the variability of symbiotic nitrogen fixation rate and the controlling factors.

Symbiotic nitrogen (N) fixation (SNF), replenishing bioavailable N for terrestrial ecosystems, exerts decisive roles in N cycling and gross primary production. Nevertheless, it remains unclear what determines the variability of SNF rate, which retards the accurate prediction for global N fixation in earth system models. This study synthesized 1230 isotopic observations to elucidate the governing factors underlying the variability of SNF rate. The SNF rates varied significantly from 3.69 to 12.54 g N m-2 year-1 across host plant taxa. The traits of host plant (e.g. biomass characteristics and taxa) far outweighed soil properties and climatic factors in explaining the variations of SNF rate, accounting for 79.0% of total relative importance. Furthermore, annual SNF yield contributed to more than half of N uptake for host plants, which was consistent across different ecosystem types. This study highlights that the biotic factors, especially host plant traits (e.g. biomass characteristics and taxa), play overriding roles in determining SNF rate compared with soil properties. The suite of parameters for SNF lends support to improve N fixation module in earth system models that can provide more confidence in predicting bioavailable N changes in terrestrial ecosystems.

Aridity threshold for alpine soil nitrogen isotope signature and ecosystem nitrogen cycling.

Determination of tipping points in nitrogen (N) isotope (δ15N) natural abundance, especially soil δ15N, with increasing aridity, is critical for estimating N-cycling dynamics and N limitation in terrestrial ecosystems. However, whether there are linear or nonlinear responses of soil δ15N to increases in aridity and if these responses correspond well with soil N cycling remains largely unknown. In this study, we investigated soil δ15N and soil N-cycling characteristics in both topsoil and subsoil layers along a drought gradient across a 3000-km transect of drylands on the Qinghai-Tibetan Plateau. We found that the effect of increasing aridity on soil δ15N values shifted from negative to positive with thresholds at aridity index (AI) = 0.27 and 0.29 for the topsoil and subsoil, respectively, although soil N pools and N transformation rates linearly decreased with increasing aridity in both soil layers. Furthermore, we identified markedly different correlations between soil δ15N and soil N-cycling traits above and below the AI thresholds (0.27 and 0.29 for topsoil and subsoil, respectively). Specifically, in wetter regions, soil δ15N positively correlated with most soil N-cycling traits, suggesting that high soil δ15N may result from the "openness" of soil N cycling. Conversely, in drier regions, soil δ15N showed insignificant relationships with soil N-cycling traits and correlated well with factors, such as soil-available phosphorus and foliage δ15N, demonstrating that pathways other than typical soil N cycling may dominate soil δ15N under drier conditions. Overall, these results highlight that different ecosystem N-cycling processes may drive soil δ15N along the aridity gradient, broadening our understanding of N cycling as indicated by soil δ15N under changing drought regimes. The aridity threshold of soil δ15N should be considered in terrestrial N-cycling models when incorporating 15N isotope signals to predict N cycling and availability under climatic dryness.

Relative importance between nitrification and denitrification to N2 O from a global perspective.

Nitrous oxide (N2 O) is a potent greenhouse gas, and its mitigation is a pressing task in the coming decade. However, it remains unclear which specific process between concurrent nitrification and denitrification dominates worldwide N2 O emission. We snagged an opportunity to ascertain whence the N2 O came and which were the controlling factors on the basis of 1315 soil N2 O observations from 74 peer-reviewed articles. The average N2 O emission derived from nitrification (N2 On ) was higher than that from denitrification (N2 Od ) worldwide. The ratios of nitrification-derived N2 O to denitrification-derived N2 O, hereof N2 On :N2 Od , exhibited large variations across terrestrial ecosystems. Although soil carbon and nitrogen content, pH, moisture, and clay content accounted for a part of the geographical variations in the N2 On :N2 Od ratio, ammonia-oxidizing microorganisms (AOM):denitrifier ratio was the pivotal driver for the N2 On :N2 Od ratios, since the AOM:denitrfier ratio accounted for 53.7% of geographical variations in N2 On :N2 Od ratios. Compared with natural ecosystems, soil pH exerted a more remarkable role to dictate the N2 On :N2 Od ratio in croplands. This study emphasizes the vital role of functional soil microorganisms in geographical variations of N2 On :N2 Od ratio and lays the foundation for the incorporation of soil AOM:denitrfier ratio into models to better predict N2 On :N2 Od ratio. Identifying soil N2 O derivation will provide a global potential benchmark for N2 O mitigation by manipulating the nitrification or denitrification.

Dominant role of the non-forest woody vegetation in the post 2015/16 El Niño tropical carbon recovery.

The extreme dry and hot 2015/16 El Niño episode caused large losses in tropical live aboveground carbon (AGC) stocks. Followed by climatic conditions conducive to high vegetation productivity since 2016, tropical AGC are expected to recover from large losses during the El Niño episode; however, the recovery rate and its spatial distribution remain unknown. Here, we used low-frequency microwave satellite data to track AGC changes, and showed that tropical AGC stocks returned to pre-El Niño levels by the end of 2020, resulting in an AGC sink of 0.18 0.14 0.26 $$ {0.18}_{0.14}^{0.26} $$ Pg C year-1 during 2014-2020. This sink was dominated by strong AGC increases ( 0.61 0.49 0.84 $$ {0.61}_{0.49}^{0.84} $$ Pg C year-1) in non-forest woody vegetation during 2016-2020, compensating the forest AGC losses attributed to the El Niño event, forest loss, and degradation. Our findings highlight that non-forest woody vegetation is an increasingly important contributor to interannual to decadal variability in the global carbon cycle.

Enhanced CO2 uptake is marginally offset by altered fluxes of non-CO2 greenhouse gases in global forests and grasslands under N deposition.

Despite the increasing impact of atmospheric nitrogen (N) deposition on terrestrial greenhouse gas (GHG) budget, through driving both the net atmospheric CO2 exchange and the emission or uptake of non-CO2 GHGs (CH4 and N2 O), few studies have assessed the climatic impact of forests and grasslands under N deposition globally based on different bottom-up approaches. Here, we quantify the effects of N deposition on biomass C increment, soil organic C (SOC), CH4 and N2 O fluxes and, ultimately, the net ecosystem GHG balance of forests and grasslands using a global comprehensive dataset. We showed that N addition significantly increased plant C uptake (net primary production) in forests and grasslands, to a larger extent for the aboveground C (aboveground net primary production), whereas it only caused a small or insignificant enhancement of SOC pool in both upland systems. Nitrogen addition had no significant effect on soil heterotrophic respiration (RH ) in both forests and grasslands, while a significant N-induced increase in soil CO2 fluxes (RS , soil respiration) was observed in grasslands. Nitrogen addition significantly stimulated soil N2 O fluxes in forests (76%), to a larger extent in grasslands (87%), but showed a consistent trend to decrease soil uptake of CH4 , suggesting a declined sink capacity of forests and grasslands for atmospheric CH4 under N enrichment. Overall, the net GHG balance estimated by the net ecosystem production-based method (forest, 1.28 Pg CO2 -eq year-1 vs. grassland, 0.58 Pg CO2 -eq year-1 ) was greater than those estimated using the SOC-based method (forest, 0.32 Pg CO2 -eq year-1 vs. grassland, 0.18 Pg CO2 -eq year-1 ) caused by N addition. Our findings revealed that the enhanced soil C sequestration by N addition in global forests and grasslands could be only marginally offset (1.5%-4.8%) by the combined effects of its stimulation of N2 O emissions together with the reduced soil uptake of CH4 .

Drivers of microbially and plant-derived carbon in topsoil and subsoil.

Plant- and microbially derived carbon (C) are the two major sources of soil organic matter (SOM), and their ratio impacts SOM composition, accumulation, stability, and turnover. The contributions of and the key factors defining the plant and microbial C in SOM along the soil profile are not well known. By leveraging nuclear magnetic resonance spectroscopy and biomarker analysis, we analyzed the plant and microbial C in three soil types using regional-scale sampling and combined these results with a meta-analysis. Topsoil (0-40 cm) was rich in carbohydrates and lignin (38%-50%), whereas subsoil (40-100 cm) contained more proteins and lipids (26%-60%). The proportion of plant C increases, while microbial C decreases with SOM content. The decrease rate of the ratio of the microbially derived C to plant-derived C (CM:P ) with SOM content was 23%-30% faster in the topsoil than in the subsoil in the regional study and meta-analysis. The topsoil had high potential to stabilize plant-derived C through intensive microbial transformations and microbial necromass formation. Plant C input and mean annual soil temperature were the main factors defining CM:P in topsoil, whereas the fungi-to-bacteria ratio and clay content were the main factors influencing subsoil CM:P . Combining a regional study and meta-analysis, we highlighted the contribution of plant litter to microbial necromass to organic matter up to 1-m soil depth and elucidated the main factors regulating their long-term preservation.

Threshold responses of soil gross nitrogen transformation rates to aridity gradient.

The responses of soil nitrogen (N) transformations to climate change are crucial for biome productivity prediction under global change. However, little is known about the responses of soil gross N transformation rates to drought gradient. Along an aridity gradient across the 2700 km transect of drylands on the Qinghai-Tibetan Plateau, this study measured three main soil gross N transformation rates in both topsoil (0-10 cm) and subsoil (20-30 cm) using the laboratorial 15 N labeling. The relevant soil abiotic and biotic variables were also determined. The results showed that gross N mineralization and nitrification rates steeply decreased with increasing aridity when aridity was less than 0.5 but just slightly decreased with increasing aridity when aridity was larger than 0.5 at both soil layers. In topsoil, the decreases of the two gross rates were accompanied by the similar decreased patterns of soil total N content and microbial biomass carbon with increasing aridity (p < .05). In subsoil, although the decreased pattern of soil total N with increasing aridity was still similar to the decreases of the two gross rates (p < .05), microbial biomass carbon did not change (p > .05). Instead, bacteria and ammonia oxidizing archaea abundances decreased with increasing aridity when aridity was larger than 0.5 (p < .05). With an aridity threshold of 0.6, gross N immobilization rate increased with increasing aridity in wetter region (aridity < 0.6) accompanied with an increased bacteria/fungi ratio, but decreased with increasing aridity in drier region (aridity > 0.6) where mineral N and microbial biomass N also decreased at both soil layers (p < .05). This study provided new insight to understand the differential responses of soil N transformation to drought gradient. The threshold responses of the gross N transformation rates to aridity gradient should be noted in biogeochemical models to better predict N cycling and manage land in the context of global change.

Global variations and controlling factors of anammox rates.

Soil anammox is an environmentally friendly way to eliminate reactive nitrogen (N) without generating nitrous oxide. Nevertheless, the current earth system models have not incorporated the anammox due to the lack of parameters in anammox rates on a global scale, limiting the accurate projection for N cycling. A global synthesis with 1212 observations from 89 peer-reviewed papers showed that the average anammox rate was 1.60 ± 0.17 nmol N g-1 h-1 in terrestrial ecosystems, with significant variations across different ecosystems. Wetlands exhibited the highest rate (2.17 ± 0.31 nmol N g-1 h-1 ), followed by croplands at 1.02 ± 0.09 nmol N g-1 h-1 . The lowest anammox rates were observed in forests and grasslands. The anammox rates were positively correlated with the mean annual temperature, mean annual precipitation, soil moisture, organic carbon (C), total N, as well as nitrite and ammonium concentrations, but negatively with the soil C:N ratio. Structural equation models revealed that the geographical variations in anammox rates were primarily influenced by the N contents (such as nitrite and ammonium) and abundance of anammox bacteria, which collectively accounted for 42% of the observed variance. Furthermore, the abundance of anammox bacteria was well simulated by the mean annual precipitation, soil moisture, and ammonium concentrations, and 51% variance of the anammox bacteria was accounted for. The key controlling factors for soil anammox rates differed from ecosystem type, for example, organic C, total N, and ammonium contents in croplands, versus soil C:N ratio and nitrite concentrations in wetlands. The controlling factors in soil anammox rate identified by this study are useful to construct an accurate anammox module for N cycling in earth system models.

Dryness limits vegetation pace to cope with temperature change in warm regions.

Climate change leads to increasing temperature and more extreme hot and drought events. Ecosystem capability to cope with climate warming depends on vegetation's adjusting pace with temperature change. How environmental stresses impair such a vegetation pace has not been carefully investigated. Here we show that dryness substantially dampens vegetation pace in warm regions to adjust the optimal temperature of gross primary production (GPP) ( T opt GPP ) in response to change in temperature over space and time. T opt GPP spatially converges to an increase of 1.01°C (95% CI: 0.97, 1.05) per 1°C increase in the yearly maximum temperature (Tmax ) across humid or cold sites worldwide (37o S-79o N) but only 0.59°C (95% CI: 0.46, 0.74) per 1°C increase in Tmax across dry and warm sites. T opt GPP temporally changes by 0.81°C (95% CI: 0.75, 0.87) per 1°C interannual variation in Tmax at humid or cold sites and 0.42°C (95% CI: 0.17, 0.66) at dry and warm sites. Regardless of the water limitation, the maximum GPP (GPPmax ) similarly increases by 0.23 g C m-2 day-1 per 1°C increase in T opt GPP in either humid or dry areas. Our results indicate that the future climate warming likely stimulates vegetation productivity more substantially in humid than water-limited regions.

Increased annual methane uptake driven by warmer winters in an alpine meadow.

Pronounced nongrowing season warming and changes in soil freeze-thaw (F-T) cycles can dramatically alter net methane (CH4 ) exchange rates between soils and the atmosphere. However, the magnitudes and drivers of warming impacts on CH4 uptake in different stages of the F-T cycle are poorly understood in cold alpine ecosystems, which have been found to be a net sink of atmospheric CH4 . Here, we reported a year-round ecosystem daily CH4 uptake in an alpine meadow on the Qinghai-Tibetan Plateau after a 5-year warming experiment that included a control, a low-level warming treatment (+2.4℃ at 5 cm soil depth), and a high-level warming treatment (+4.5℃ at 5 cm soil depth). We found that warming shortened the F-T cycle under the low-level warming and soils did not freeze under the high-level warming. Although both warming treatments increased the mean CH4 uptake rate, only the high-level warming significantly increased annual CH4 uptake compared to the control. The warming-induced stimulation of CH4 uptake mainly occurred in the cold season, which was mostly during spring thaw under low-level warming and during the frozen winter under high-level warming due to a longer period with thawed soil. We also found that warming significantly stimulated daily CH4 uptake mainly by reducing near-surface soil water content in the warm season, whereas both soil water content and temperature controlled daily CH4 uptake in different ways during the autumn freeze, frozen winter, and spring thaw periods of the control. Our study revealed a strong warming effect on CH4 uptake during the entire F-T cycle in the alpine meadow, especially the unfrozen winter. Our results also suggested the important roles of soil pH, available phosphorus, and methanotroph abundance in regulating annual CH4 uptake in response to warming, which should be incorporated into biogeochemical models for accurately forecasting CH4  fluxes under future climate scenarios.

Dryness weakens the positive effects of plant and fungal ß diversities on above- and belowground biomass.

Plant and microbial diversity are key to determine ecosystem functioning. Despite the well-known role of local-scale α diversity in affecting vegetation biomass, the effects of community heterogeneity (ß diversity) of plants and soil microbes on above- and belowground biomass (AGB and BGB) across contrasting environments still remain unclear. Here, we conducted a dryness-gradient transect survey over 3000 km across grasslands on the Tibetan Plateau. We found that plant ß diversity was more dominant than α diversity in maintaining higher levels of AGB, while soil fungal ß diversity was the key driver in enhancing BGB. However, these positive effects of plant and microbial ß diversity on AGB and BGB were strongly weakened by increasing climatic dryness, mainly because higher soil available phosphorus caused by increasing dryness reduced both plant and soil fungal ß diversities. Overall, these new findings highlight the critical role of above- and belowground ß diversity in sustaining grassland biomass, raising our awareness to the ecological risks of large-scale biotic homogenization under future climate change.

An integrated belowground trait-based understanding of nitrogen-driven plant diversity loss.

Belowground plant traits play important roles in plant diversity loss driven by atmospheric nitrogen (N) deposition. However, the way N enrichment shapes plant microhabitats by patterning belowground traits and finally determines aboveground responses is poorly understood. Here, we investigated the rhizosheath trait of 74 plant species in seven N-addition simulation experiments across multiple grassland ecosystems in China. We found that rhizosheath formation differed among plant functional groups and contributed to changes in plant community composition induced by N enrichment. Compared with forb species, grass and sedge species exhibited distinct rhizosheaths; moreover, grasses and sedges expanded their rhizosheaths with increasing N-addition rate which allowed them to colonize belowground habitats. Grasses also shaped a different microenvironment around their roots compared with forbs by affecting the physicochemical, biological, and stress-avoiding properties of their rhizosphere soil. Rhizosheaths act as a "biofilm-like shield" by the accumulation of protective compounds, carboxylic anions and polysaccharides, determined by both plants and microorganisms. This enhanced the tolerance of grasses and sedges to stresses induced by N enrichment. Conversely, forbs lacked the protective rhizosheaths which renders their roots sensitive to stresses induced by N enrichment, thus contributing to their disappearance under N-enriched conditions. This study uncovers the processes by which belowground facilitation and trait matching affect aboveground responses under conditions of N enrichment, which advances our mechanistic understanding of the contribution of competitive exclusion and environmental tolerance to plant diversity loss caused by N deposition.

Variations and controlling factors of soil denitrification rate.

The denitrification process profoundly affects soil nitrogen (N) availability and generates its byproduct, nitrous oxide, as a potent greenhouse gas. There are large uncertainties in predicting global denitrification because its controlling factors remain elusive. In this study, we compiled 4301 observations of denitrification rates across a variety of terrestrial ecosystems from 214 papers published in the literature. The averaged denitrification rate was 3516.3 ± 91.1 µg N kg-1  soil day-1 . The highest denitrification rate was 4242.3 ± 152.3 µg N kg-1  soil day-1 under humid subtropical climates, and the lowest was 965.8 ± 150.4 µg N kg-1 under dry climates. The denitrification rate increased with temperature, precipitation, soil carbon and N contents, as well as microbial biomass carbon and N, but decreased with soil clay contents. The variables related to soil N contents (e.g., nitrate, ammonium, and total N) explained the variation of denitrification more than climatic and edaphic variables (e.g., mean annual temperature (MAT), soil moisture, soil pH, and clay content) according to structural equation models. Soil microbial biomass carbon, which was influenced by soil nitrate, ammonium, and total N, also strongly influenced denitrification at a global scale. Collectively, soil N contents, microbial biomass, pH, texture, moisture, and MAT accounted for 60% of the variation in global denitrification rates. The findings suggest that soil N contents and microbial biomass are strong predictors of denitrification at the global scale.

Ecosystem restoration and belowground multifunctionality: A network view.

Ecological restoration is essential to reverse land degradation worldwide. Most studies have assessed the restoration of ecosystem functions individually, as opposed to a holistic view. Here we developed a network-based ecosystem multifunctionality (EMF) framework to identify key functions in evaluating EMF restoration. Through synthesizing 293 restoration studies (2900 observations) following cropland abandonment, we found that individual soil functions played different roles in determining the restoration of belowground EMF. Soil carbon, total nitrogen, and phosphatase were key functions to predict the recovery of belowground EMF. On average, abandoned cropland recovered ~19% of EMF during 18 years. The restoration of EMF became larger with longer recovery time and higher humidity index, but lower with increasing soil depth and initial soil carbon. Overall, this study presents a network-based EMF framework, effectively helping to evaluate the success of ecosystem restoration and identify the key functions.

Fine-root functional trait responses to experimental warming: a global meta-analysis.

Whether and how warming alters functional traits of absorptive plant roots remains to be answered across the globe. Tackling this question is crucial to better understanding terrestrial responses to climate change as fine-root traits drive many ecosystem processes. We carried out a detailed synthesis of fine-root trait responses to experimental warming by performing a meta-analysis of 964 paired observations from 177 publications. Warming increased fine-root biomass, production, respiration and nitrogen concentration as well as decreased root carbon : nitrogen ratio and nonstructural carbohydrates. Warming effects on fine-root biomass decreased with greater warming magnitude, especially in short-term experiments. Furthermore, the positive effect of warming on fine-root biomass was strongest in deeper soil horizons and in colder and drier regions. Total fine-root length, morphology, mortality, life span and turnover were unresponsive to warming. Our results highlight the significant changes in fine-root traits in response to warming as well as the importance of warming magnitude and duration in understanding fine-root responses. These changes have strong implications for global soil carbon stocks in a warmer world associated with increased root-derived carbon inputs into deeper soil horizons and increases in fine-root respiration.

Divergent responses of primary production to increasing precipitation variability in global drylands.

Interannual variability in precipitation has increased globally as climate warming intensifies. The increased variability impacts both terrestrial plant production and carbon (C) sequestration. However, mechanisms driving these changes are largely unknown. Here, we examined mechanisms underlying the response of aboveground net primary production (ANPP) to interannual precipitation variability in global drylands with mean annual precipitation (MAP) <500 mm year-1 , using a combined approach of data synthesis and process-based modeling. We found a hump-shaped response of ANPP to precipitation variability along the MAP gradient. The response was positive when MAP < ~300 mm year-1 and negative when MAP was higher than this threshold, with a positive peak at 140 mm year-1 . Transpiration and subsoil water content mirrored the response of ANPP to precipitation variability; evaporation responded negatively and water loss through runoff and drainage responded positively to precipitation variability. Mean annual temperature, soil type, and plant physiological traits all altered the magnitude but not the pattern of the response of ANPP to precipitation variability along the MAP gradient. By extrapolating to global drylands (<500 mm year-1  MAP), we estimated that ANPP would increase by 15.2 ± 6.0 Tg C year-1 in arid and hyper-arid lands and decrease by 2.1 ± 0.5 Tg C year-1 in dry sub-humid lands under future changes in interannual precipitation variability. Thus, increases in precipitation variability will enhance primary production in many drylands in the future.

Global soil-derived ammonia emissions from agricultural nitrogen fertilizer application: A refinement based on regional and crop-specific emission factors.

Ammonia (NH3 ) emissions from fertilized soils to the atmosphere and the subsequent deposition to land surface exert adverse effects on biogeochemical nitrogen (N) cycling. The region- and crop-specific emission factors (EFs) of N fertilizer for NH3 are poorly developed and therefore the global estimate of soil NH3 emissions from agricultural N fertilizer application is constrained. Here we quantified the region- and crop-specific NH3 EFs of N fertilizer by compiling data from 324 worldwide manipulative studies and focused to map the global soil NH3 emissions from agricultural N fertilizer application. Globally, the NH3 EFs averaged 12.56% and 14.12% for synthetic N fertilizer and manure, respectively. Regionally, south-eastern Asia had the highest NH3 EFs of synthetic N fertilizer (19.48%) and Europe had the lowest (6%), which might have been associated with the regional discrepancy in the form and rate of N fertilizer use and management practices in agricultural production. Global agricultural NH3 emissions from the use of synthetic N fertilizer and manure in 2014 were estimated to be 12.32 and 3.79 Tg N/year, respectively. China (4.20 Tg N/year) followed by India (2.37 Tg N/year) and America (1.05 Tg N/year) together contributed to over 60% of the total global agricultural NH3 emissions from the use of synthetic N fertilizer. For crop-specific emissions, the NH3 EFs averaged 11.13%-13.95% for the three main staple crops (i.e., maize, wheat, and rice), together accounting for 72% of synthetic N fertilizer-induced NH3 emissions from croplands in the world and 70% in China. The region- and crop-specific NH3 EFs of N fertilizer established in this study offer references to update the default EF in the IPCC Tier 1 guideline. This work also provides an insight into the spatial variation of soil-derived NH3 emissions from the use of synthetic N fertilizer in agriculture at the global and regional scales.

Vital roles of soil microbes in driving terrestrial nitrogen immobilization.

Nitrogen immobilization usually leads to nitrogen retention in soil and, thus, influences soil nitrogen supply for plant growth. Understanding soil nitrogen immobilization is important for predicting soil nitrogen cycling under anthropogenic activities and climate changes. However, the global patterns and drivers of soil nitrogen immobilization remain unclear. We synthesized 1350 observations of gross soil nitrogen immobilization rate (NIR) from 97 articles to identify patterns and drivers of NIR. The global mean NIR was 8.77 ± 1.01 mg N kg-1  soil day-1 . It was 5.55 ± 0.41 mg N kg-1  soil day-1 in croplands, 15.74 ± 3.02 mg N kg-1  soil day-1 in wetlands, and 15.26 ± 2.98 mg N kg-1  soil day-1 in forests. The NIR increased with mean annual temperature, precipitation, soil moisture, soil organic carbon, total nitrogen, dissolved organic nitrogen, ammonium, nitrate, phosphorus, and microbial biomass carbon. But it decreased with soil pH. The results of structural equation models showed that soil microbial biomass carbon was a pivotal driver of NIR, because temperature, total soil nitrogen, and soil pH mostly indirectly influenced NIR via changing soil microbial biomass. Moreover, microbial biomass carbon accounted for most of the variations in NIR among all direct relationships. Furthermore, the efficiency of transforming the immobilized nitrogen to microbial biomass nitrogen was lower in croplands than in natural ecosystems (i.e., forests, grasslands, and wetlands). These findings suggested that soil nitrogen retention may decrease under the land use change from forests or wetlands to croplands, but NIR was expected to increase due to increased microbial biomass under global warming. The identified patterns and drivers of soil nitrogen immobilization in this study are crucial to project the changes in soil nitrogen retention.