Soil burial has a greater effect on litter decomposition rate than nitrogen enrichment in alpine grasslands

Abstract

Aims

Litter is frequently buried in the soil in alpine grasslands due to grassland degradation, serious rodent infestation and frequent strong winds. However, the effects of various litter positions on litter decomposition rates and nutrient dynamics under nitrogen (N) enrichment in such areas remain unknown.

Methods

A field experiment was performed in the alpine grasslands of northwest China to investigate the influence of litter position (surface, buried in the soil and standing) and N enrichment on litter decomposition, using data from two dominant grass species (Festuca ovina and Leymus tianschanicus) in control and N-enriched plots.

Important Findings

Litter decomposition rates were much faster in buried litter and slower in standing litter than in surface litter. N enrichment significantly affected litter quality and then influenced decomposition. But no significant differences in litter mass remaining were observed between control and N-enriched soil burial. These results indicated that N enrichment significantly affected litter decomposition by changes in litter quality. In addition, all litter exhibited net carbon (C) and phosphorus (P) release regardless of treatments. Litter exhibited net N accumulation for litter from the control plots but showed N release for litter from N enrichment plots. These suggested that litter decomposition can be limited by N and N enrichment influenced N cycling of litter. Current study presented direct evidence that soil buried litter exhibited faster mass loss and C release, and that soil burial can be a candidate explanation why litter decomposes faster than expected in dryland.

摘要

高寒草地土壤掩埋对凋落物分解速率的影响大于氮素富集

由于高寒草地退化、鼠害严重、大风频繁等原因，凋落物被频繁掩埋在土壤中。但凋落物的位置变化和氮富集对高寒草原凋落物分解速率和养分动态影响的认识尚不清楚。为了研究凋落物 位置变化(地表、掩埋10 cm和悬空60 cm)和氮富集对高寒草原优势植物凋落物分解的影响，本研究依托2009年在新疆天山巴音布鲁克高寒草原设置的长期模拟氮沉降研究平台，以对照和氮富集处理样方的优势植物羊茅(Festuca ovina)和赖草(Leymus tianschanicus)凋落物为试验材料，测定分解过程中凋落物质量损失和碳氮磷含量的变化特征。研究结果表明，掩埋凋落物分解速率显著快于地表凋落物，悬空处理凋落物分解速率慢于地表凋落物。氮富集显著影响凋落物质量，进而影响凋落物分解。而凋落物质量残留在对照与氮富集土壤掩埋之间无显著差异。这些结果表明，氮富集通过凋落物质量而不是通过土壤环境因素，影响短期凋落物分解。不同处理的所有试验凋落物均有碳和磷的释放现象。对照处理的凋落物，凋落物氮以累积为主，而氮富集处理的凋落物，凋落物氮以释放为主。这表明凋落物分解可能受到氮元素限制，氮富集改变了凋落物分解调控的氮循环过程。本研究提供了直接证据，掩埋处理的凋落物有更快的质量损失和碳元素释放，土壤掩埋是旱地凋落物分解速率比模型预测的快的一个候选解释。

INTRODUCTION

Plant litter decomposition in terrestrial ecosystems contributes substantially to global annual carbon (C) fluxes and nutrient cycling (Ciais et al. 2013). Estimates show that the annual release of CO₂ due to litter decomposition accounts for 70% of the global annual C flux (Raich and Schlesinger 1992). In most ecosystems, litter decomposition is controlled by climate, litter quality and soil organisms (Austin and Vivanco 2006; Ayres et al. 2009; Kaspari et al. 2007). Initial litter quality has been widely reported as influencing decomposition rates (Bradford et al. 2016; Cornwell et al. 2008; Kaspari et al. 2007). High-quality litter, with higher N content and lower lignin content, usually decomposes faster (Bradford et al. 2016; Mooshammer et al. 2012). Therefore, litter quality can explain a large proportion of the variation in litter decomposition at local and regional scales (Bradford et al. 2014; Cornwell et al. 2008).

Atmospheric nitrogen (N) deposition, due to agricultural fertilizer application, rapid urbanization and livestock development, has been predicted to increase at the global scale (Lamarque et al. 2005) and is considered one of the key components of global change (Ciais et al. 2013). It has been estimated that the annual rate of N deposition increased by 60% between 1980 and 2010 in China (Liu et al. 2013), but shows great spatial variability with a decreasing trend from the meadow steppes in the northeast to the alpine steppes in the northwest of China (Wen et al. 2020). N enrichment could alter the litter quality via affecting litter N content and C:N ratio (Hou et al. 2017; Su et al. 2021) and these changes in litter quality could produce large effects on subsequent litter decomposition (Hou et al. 2021; Pichon et al. 2020; See et al. 2019). Due to the vast amount of C held in plant litter (70–150 Pg, even very small changes will have tremendous impacts on global C fluxes and nutrient cycling (Schimel 1995). In addition, the relationships between climate and litter decomposition can be partly controlled by litter quality (Cornwell et al. 2008; Currie et al. 2010). Traditional decomposition models factor in climate variables when predicting litter decomposition rates through temperature dependent effects on soil microbial activity in forest and mesic ecosystems. However, these models fail to predict the observed litter decomposition rates in dry soils, which are higher than expected based on climatic variables (Austin 2011; Bradford et al. 2014). Previous studies have indicated that litter decomposition in arid regions is significantly affected by photodegradation and litter position, especially standing and burial (Austin and Vivanco 2006; Gliksman et al. 2018; Hewins et al. 2012; Liu et al. 2015; Wang et al. 2017).

High litter decomposition rates in buried litter compared with surface litter have been reported, with clear evidence from desert grasslands (Erdenebileg et al. 2020), the Chihuahuan desert (Joly et al. 2017) and the Patagonian steppes (Austin et al. 2009). This may be due to favorable soil temperature and moisture conditions supporting microbial decomposition in buried litter. In contrast, decomposition rates may be higher in surface litter than in buried litter due to lower soil microbial activity under unfavorable conditions (Fan et al. 2021; Lee et al. 2014). N enrichment significantly affects soil bacterial and fungal diversity and composition (Liu et al. 2020), changes which significantly influence surface litter decomposition (Allison et al. 2009). In addition, changes in litter quality caused by N enrichment have been widely reported (Allison et al. 2009; Hou et al. 2021; Liu et al. 2010). Even so, information on the effect of N enrichment on buried litter decomposition remains limited.

In dry soils, especially in the grassland ecosystems of Central Asia, the effects of degradation, serious rodent infestations and strong winds (Chen et al. 2020; Li et al. 2020) create various microenvironments which, in concert with soil/sand transport and altered light and moisture regimes, are widely recognized to cause substantial impacts on litter nutrient cycling (Austin et al. 2016; Erdenebileg et al. 2020). Photodegradation can break down the lignin structure in secondary cell walls, promoting subsequent biotic degradation (Austin et al. 2016). However, the combination of high temperature and low litter moisture may constrain the activity of microbial decomposers (Huang and Li 2017). Standing litter is exposed to these conditions. Standing and surface litter are exposed to strong fluctuations in precipitation, but standing litter is air-dried more quickly, so that that standing litter is less influenced by microbial decomposition variations due to lower litter moisture levels. However, previous studies have shown that standing litter of Cleistogenes squarrosa decomposed more rapidly because high humidity conditions at night supports rapid decomposition of standing litter in temperate steppes (Wang et al. 2017). Because the degree of litter decomposition is dependent on the decomposer species assemblage concerned, more experiments need to be carried out using cultivated decomposer stocks to confirm the different capacities of different species.

Soil movements will probably increase in the future due to on-going changes in land use and climate, such as extreme drought (Okin et al. 2009), and the global effects of litter burial in the soil on litter decomposition in extensive arid lands may become more important. Only a few studies have investigated the simultaneous effects of litter position (standing, surface and buried in the soil) and N enrichment on litter decomposition. This paper presents the results of a comprehensive field litter decomposition experiment, carried out in the alpine grasslands near the Bayinbuluk Grassland Ecosystem Research Station in the Tianshan Mountains of northwest China. The aim of the study was to test the effects of N enrichment and three typical litter positions on litter decomposition and nutrient release. We studied two dominant perennial grasses (Festuca ovina and Leymus tianschanicus) in both control and N-enriched plots, taking advantage of an established N enrichment field experiment that has been running since 2009 (Li et al. 2012).

MATERIALS AND METHODS

Study site

The Bayinbuluk Grassland Ecosystem Research Station is situated in the Tianshan Mountains of northwest China (42.89° N, 83.71° E). The long-term annual average temperature was −4.8 °C, with monthly average temperatures ranging from −27.4 °C in January to 11.2 °C in July. The long-term average annual precipitation was 282.3 mm from 1980 to 2012, with approximately 78% of the rainfall occurring during the growing season (Li et al. 2012). The dominant species in the ecosystem were F. ovina and L. tianschanicus, which constitute 64%–75% of the community biomass. This grassland received approximately 8.0 kg N ha⁻¹ yr⁻¹ from atmospheric deposition (Li et al. 2015).

Experimental design

The nitrogen addition experiment, using five levels of added N (0,10, 30, 90 and 150 kg N ha⁻¹ yr⁻¹, hereafter abbreviated as N0, N1, N3, N9 and N15, respectively), was established at the end of May 2009 and has been maintained from 2009 to 2020. In comparison, the N levels of previous studies were: in typical steppe habitats (Bai et al. 2010, N levels 0, 17.5, 52.5, 105, 175 and 280 kg N ha⁻¹ yr⁻¹); in meadow steppe (Ma et al. 2020, N levels 0, 20, 40, 80, 160 and 320 kg N ha⁻¹ yr⁻¹); in desert steppe (Wang et al. 2019, N levels 0, 12.5, 25, 50, 100 and 200 kg N ha⁻¹ yr⁻¹); and in alpine steppe (Peng et al. 2017, N levels 0, 10, 20, 40, 80, 160, 240 and 320 kg N ha⁻¹ yr⁻¹). The experimental plots were arranged within randomized blocks with four replicates (blocks). Each plot was 4 m × 8 m. The distance between any two adjacent blocks was 1 m and the plots within a block were separated by 1 m buffer strips. N fertilizer (NH₄NO₃) was weighed and dissolved in 8 L water, and applied uniformly to each plot in late May and June each year using a backpack sprayer. The control plots received the same amount of water but without added N (Li et al. 2015).

Our aim was to investigate the effects of the three litter positions and N enrichment on leaf litter decomposition. We placed the litter either on the soil surface, buried in the soil or suspended above the surface to simulate standing litter. Loss of litter mass was determined in each plot using a litterbag method where 4 g of oven-dried litter was placed in 15 cm × 20 cm (0.5 mm mesh) nylon litterbags. These litterbags were evenly distributed among each plot (4 treatments × 4 kinds of litter quality × 4 sampling timepoints × 5 bags per sampling time = 320 bags) on 1 September 2019. Surface litterbags were placed flat on the soil surface of the plots after manually clipping the vegetation to ground level within the center (1 m × 2 m) of each plot. The litterbags were then fixed to the ground surface using small pieces of wire and then lightly buried with soil from the control plots (N0) or from the N9 treatment plots. Buried litter was placed below the soil surface by removing soil and placing the litterbags flat at 5 cm below the soil surface and then backfilling with soil. Standing litter was simulated by suspending litter samples 60 cm above the soil surface using iron wire. Four treatments were therefore created (Supplementary Fig. S1). Photographs of the experimental plots are provided in Supplementary Fig. S1. Litterbags were harvested after 8, 9, 10 and 12 months of decomposition. The burial of litter by soil due to grassland degradation, serious rodent infestation and frequent strong winds, is a common natural occurrence in this alpine grassland. To assess the effects of N enrichment on litter decomposition, we collected leaf litter from the control (N0) and N9 treatment plots. In this study, two N levels were chosen, contrary to previous studies on the effects of N enrichment on litter decomposition in temperate grasslands (Hou et al. 2021).

Two dominant perennial grasses (F. ovina and L. tianschanicus) were chosen from the control and N9 treatment plots: L. tianschanicus from N9 treatment plots (N9-L. T litter); L. tianschanicus from N0 treatment plots (N0-L. T litter); F. ovina from N9 treatment plots (N9-F. O litter); and F. ovina from N0 treatment plots (N0-F. O litter), to make two types of litter quality from each species. The leaf litter collected was oven-dried at 65 °C to a constant weight.

Litter mass loss and chemical analysis

The litter remaining after decomposition was removed from the litterbags in the laboratory, cleaned of extraneous material, oven-dried at 65 °C and weighed. The litter mass remaining (MR) was expressed as a percentage of the initial mass. The remaining litter was ground to a fine powder using a ball mill and the total C and N concentrations analyzed using an N/C analyzer (Analytik Jena, Germany). The concentration of P was measured using an AA3 Continuous Flow Analyzer (Seal Analytical, Germany). The concentration of lignin, hemicellulose and cellulose in the litter was determined using the Van Soest method (Van Soest and Wine 1968). The MR and the remaining C, N and P were calculated as follows: $MR(%)=Mt/M0×100=MtCt/M0C0×100$⁠, where M_t and M₀ represent the remaining and the initial dried litter mass (g); C₀ represents the initial concentration of C and C_t represents the concentration of C remaining after each sampling collection (and similarly for calculations for N and P).

Statistical analysis

The single-exponential decomposition model was used: $Mt/M0=e−kt$⁠, where M_t/M₀ is the fraction of the MR at time t and k is the decomposition constant (Olson 1963). All data were checked for normal distribution and homogeneity of variances using the Kolmogorov–Smirnov test and Levene’s test, respectively. The paired t-test was used to test the initial response of litter decomposition to N enrichment. A one-way analysis of variance (ANOVA) was used to estimate the effects of litter position on litter mass loss and the levels of C, N and P remaining in the litterbag after each collection period, with the significance level set at P = 0.05. A three-way ANOVA was conducted to evaluate the impacts of litter position, decomposition time and litter quality on litter mass loss. Statistical analyses were performed using SPSS 23.0 software (SPSS Inc., IL, USA).

RESULTS

Initial chemical traits of the litter

N enrichment significantly affected the initial nutrient content of the leaf litter of both experimental species (Table 1). N enrichment significantly increased both the N concentration and N:P ratio of the litter, but tended to decrease the leaf P concentration, leaf pH and the C:N and lignin:N ratios. In addition, N enrichment enhanced litter lignin, cellulose and hemicellulose levels of F. ovina, but not to a significant degree. However, N enrichment did significantly increase the lignin and hemicellulose contents of L. tianschanicus litter.

Impact of litter position on litter mass loss

Litter quality, decay time and experimental treatments all had significant effects on litter mass loss (Table 2). Litter buried in the soil decomposed significantly faster than standing and surface litter. However, there were no significant differences between the control plots and the N9 treatment plots, indicating that N-enriched soil had no significant effect on litter decomposition. After 12 months of decomposition, the mass of surface litter remaining (63.75%) was 1.87 times higher than that of the buried litter (34.14%), but 0.86 times lower than that of the standing litter (74.02%) (Fig. 1). In addition, the changes in litter quality caused by N enrichment significantly affected litter mass loss (Figs 1 and 2). Leymus tianschanicus from the control plots decomposed faster than that from the N9 treatment plots. However, F. ovina from the N9 treatment plots showed faster decomposition than that from the control plots. The effect of N enrichment on litter decomposition was therefore species dependent.

The constant k for litter decomposition rate

Litter quality and litter position significantly influenced the constant k for litter decomposition (Fig. 2; Supplementary Table S1). k was greater for all of the buried litter than for both standing and surface litter. Results indicated that decomposition rate of L. tianschanicus induced by N enrichment was lower than that of L. tianschanicus from control treatment, but contrasting result was observed for F. ovina (Fig. 2; Table 2).

Litter nutrient release dynamics and patterns in the different treatments

Litter position and N enrichment showed no significant impacts on litter C concentration across the different periods of decomposition (Supplementary Fig. S2). Litter position had significant effects on litter C release dynamics and patterns. As decomposition progressed, the C remaining in litter consistently decreased across all litter types (Fig. 3). Buried litter had faster C release than standing and surface litters. N-enriched soil had no significant effect on litter C release, but changes in litter quality due to N enrichment did significantly affect C release (Fig. 3).

Litter position and N enrichment showed notable impacts on litter C concentration across different durations of decomposition (Supplementary Fig. S3). Litter N dynamics were significantly affected by litter quality and litter position, and the response differed in both species (Fig. 4). For L. tianschanicus litter, N9-L. T litter exhibited N release, whereas N0-L. T litter showed N accumulation over a 12-month period of decomposition. Similar phenomenon was also observed for F. ovina litter. N enrichment significantly changed litter N cycling during decomposition due to changes in litter quality caused by N enrichment. Litter P concentration was also significantly influenced by litter position and N enrichment across different periods of decomposition (Supplementary Fig. S4). Litter P was released throughout the 12-month decomposition period in both experimental specie. Surface litter released more P than buried litter over a 10-month decomposition period (Fig. 5).

DISCUSSION

The effects of changes in litter quality caused by N enrichment on litter decomposition rates

Litter quality is an important factor in controlling decomposition rate (Bradford et al. 2016; Cornwell et al. 2008). Numerous previous studies have reported that high-quality litter, with higher N content and lower lignin content, usually decomposes faster (Bradford et al. 2016; Mooshammer et al. 2012). This is partially consistent with our results. In our study, in L. tianschanicus litter from N-enriched plots, with higher N and lower lignin contents, N decomposed faster than litter decomposing in the control plots. In contrast, in F. ovina litter from N-enriched plots, with higher N and lower lignin contents, N decomposed faster than litter decomposing in the control plots. This indicates that changes in litter quality caused by N enrichment significantly influenced litter decomposition and that the effects of N enrichment were species dependent. We hypothesize the following explanations. First, N enrichment significantly increases the lignin and hemicellulose contents of L. tianschanicus litter, but has no significant effect on the lignin and hemicellulose contents of F. ovina litter. Litter with high lignin and hemicellulose contents usually decompose more slowly (Bradford et al. 2016; Mooshammer et al. 2012). Second, litter stoichiometry has been reported to affect litter decomposition, litter decomposition rate being higher in species with lower C:P ratios (Chen et al. 2016) while N enrichment significantly increases the litter leaf C:P ratio of L. tianschanicus, which may lead to slow decomposition of this litter. This effect has remained poorly understood due to limited samples and reports. More studies should be conducted on the effects of the litter C:P ratio on litter decomposition rates.

The effects of litter position and N enrichment on litter decomposition rates

To our knowledge, our study is the first to comprehensively quantify the effects of litter position and N enrichment on litter mass loss in alpine grasslands. The results indicate that litter buried in the soil decomposes at a significantly increased rate; 0.5–1 times faster than soil surface litter. This result supports many previous studies conducted in desert and grassland ecosystems (Table 3), which confirmed that burying litter under the sand significantly accelerated its loss in mass (Hewins and Throop 2016; Joly et al. 2017; Liu et al. 2015). However, these results differ from those of a recent study by Li et al. (2019), who reported that litter decomposition was significantly inhibited when buried under sand in extremely arid desert grasslands. Another recent study also found that when litter was buried under 5 cm of sand, there was no significant effect on litter decomposition, but litter decomposition was significantly decreased when buried 10 or 20 cm under the sand. This decrease in litter decomposition when buried more deeply was related to reductions in light intensity and soil temperature, which were the most important factors affecting the speed of litter decomposition (Qu et al. 2021). A study by Austin et al. (2016) reported that the decrease in buried litter decomposition rates were mainly caused by reduced photodegradation leading to reduced lignin decomposition and subsequently slower litter decomposition. Austin and Ballare (2010) showed that photodegradation played a dominant role in driving litter decomposition in the semi-arid Patagonian steppe and that soil microbial activity played only a very minor role in litter decomposition rates.

Nitrogen enrichment alters the belowground microbial community in soils and disturbs the litter-mediated nutrient cycle (Pichon et al. 2020). However, our results did not show a significant difference in the decomposition rates of litter buried in the soil between N-enriched plots and the control plots. However, litter quality did significantly influence litter decomposition, showing that N enrichment affects litter decomposition rates through changes in litter quality rather than soil factors. In fact, N enrichment changes the diversity indices and dominance patterns of bacterial and fungal communities in this alpine grassland (Hao et al. 2020). Previous studies have reported that soil microbial organisms have important impacts on litter decomposition (Dirks et al. 2010; Dong et al. 2020; Wang et al. 2017). Consequently, the limited number of replications and short time span of our work may result in some uncertainties and further studies are needed to verify our findings. Overall, the present results highlight the fact that litter buried under the soil decomposes faster and releases more C into the atmosphere, with positive feedback impacts on global warming.

The decomposition of surface litter has received most attention and very few studies have looked at the effects of standing litter height (Gliksman et al. 2018). Our results show that standing litter decomposes more slowly than surface litter. This contradicts many previous studies that have reported standing litter decomposing faster than surface litter (Table 3), with evidence from desert grasslands (Liu et al. 2015), temperate steppes (Wang et al. 2017) and Mediterranean shrub forest ecosystems (Gliksman et al. 2018). However, opposing evidence can also be found. For example, Erdenebileg et al. (2020) reported that even within the same desert grassland ecosystem, decomposition rates of standing litter were not significantly different from those of surface litter. Decomposition of standing litter can not only be strongly influenced by abiotic factors, such as photodegradation and wind abrasion, but also by enhanced microbial decomposition induced by beneficial combinations of temperature and moisture at night (Dirks et al. 2010; Wang et al. 2017). In our study area, multiyear mean temperature is −4.8 °C and the long-term average annual precipitation is 282.3 mm. In addition, frequent strong winds occur mainly from April to July each year. This causes litter to air-dry faster, which is detrimental to subsequent microbial decomposition. Our results provide direct evidence that standing litter decomposes more slowly in alpine grassland ecosystems, deepening our understanding of the impact of litter position on litter-mediated nutrient cycling, and significantly advances our ability to accurately evaluate and predict global C cycling.

The effects of litter quality and litter position on nutrient dynamics during litter decomposition

Numerous studies have reported that litter nutrient accumulation and release are determined primarily by the litter quality and the stoichiometric requirements of the microbial decomposers (Chen et al. 2016; Manzoni et al. 2010; Zhang et al. 2018). All of our experimental litter samples showed net C release, consistent with observed litter mass loss. Litter C release has been found in many studies, resulting from high initial activity of polysaccharide-degrading enzymes (Gong et al. 2020; Hobbie et al. 2012; Jing et al. 2019; Zhang et al. 2018). Litter quality significantly effected the amount of N remaining in the litter. Leymus tianschanicus and F. ovina litter from the control plots showed net N immobilization during the 12-month decomposition period, whereas litter from the N-enriched plots showed net N release. A number of studies from N-limited ecosystems have also observed N accumulation during litter decomposition (Gong et al. 2020; Prieto et al. 2019; Sollins et al. 1987). Our study suggested that initial litter N from the control plots can be insufficient to support the decomposition of litter and that the litter needed more N to decompose adequately. Therefore, microbial immobilization nutrients from the soil are needed to maintain stoichiometric homeostasis in litter (Manzoni et al. 2010). N enrichment significantly increased litter N by 100% in the experimental species in this study. Our results indicate that all of the litter from N-enriched treatment plots showed net N release. This is consistent with previous studies showing that N enrichment improved litter quality (higher litter N) and stimulated litter N release during decomposition (Chen et al. 2016; Jing et al. 2019). Microbial decomposers transfer litter N to meet their own requirements when colonizing litter, and this may explain the N release in higher N content litter (Zhang et al. 2016). Our results indicate that changes of litter quality caused by N enrichment significantly affect litter N cycling.

In our study, the litter quality and position significantly influenced the patterns of P release of the experimental species during litter decomposition. All of the experimental litter showed net P release. This result is consistent with previous studies (Jing et al. 2019; Tu et al. 2014; Yan et al. 2020). Surface litter had the fastest P release compared with buried and standing litter. Liu et al. (2015) indicated that litter position had no significant effect on litter P loss in desert grasslands. It is possible that leaching by precipitation is an important factor regulating litter P dynamics. Dou et al. (2018) also showed that leaching is an important avenue for phosphorus release, with rates six times higher than those of C and N. More studies are needed in future to verify this result, especially in grassland ecosystems.

CONCLUSIONS

Our results showed that litter buried in the soil decomposes significantly more quickly, and that standing litter decomposes significantly more slowly, than surface litter. N enrichment affects litter decomposition via litter quality rather than changes in decomposing environments; the influence of litter quality induced by N enrichment on litter decomposition depends on species identity. Meanwhile, N enrichment significantly affects litter N accumulation and release and disturbs the mediation of N cycling by litter. The relative role of precipitation in litter P dynamics requires further study. This study presents direct evidence that litter buried in the soil decomposes significantly more quickly, suggesting a potential mechanism to explain why observed litter decomposition is faster than previous model predictions in dry soils. Soil movement will likely exacerbate these results and given the current increases in soil wind erosion and grassland degradation in northwest China and Central Asia in general, the effect of decomposition of litter buried in the soil in extensive arid lands globally may be magnified under future conditions. More litter C will be released into the atmosphere as climate and land use change, suggesting future positive feedbacks to climate warming.

Supplementary Material

Supplementary material is available at Journal of Plant Ecology online.

Figure S1: Examples of the litter position and N enrichment treatments effect on litter decomposition.

Figure S2: Changes of litter C concentrations across differential period of decomposition.

Figure S3: Changes of litter N concentrations across differential period of decomposition.

Figure S4: Changes of litter P concentrations across differential period of decomposition.

Table S1: Litter decomposition rate, and predicted time to 50% and 95% mass loss calculated using the first-order exponential decay model for experimental species under differential position treatments.

Funding

This work was supported financially by the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA20050103), Natural Science Foundation of Xinjiang Uygur Autonomous Region (2019D01C066), Tianshan Cedar Project of Xinjiang Uygur Autonomous Region (2020XS26), the National Natural Science Foundation of China (41425007, 41673079) and the ‘Light of West China’ Program of the Chinese Academy of Sciences (W.X. Han).

Conflict of interest statement. The authors declare that they have no conflict of interest.

REFERENCES