In this study, the experiment was conducted in a temperate meadow steppe dominated by Stipa baicalensis at the Inner Mongolia Grassland Ecosystem National Field Scientific Observation Station of the Chinese Academy of Agricultural Sciences (49°19′-49°20′N, 119°55′-120°02′E, elevation 628–649 m), situated within Hulunber Grassland in China. The Hulunber Grassland enclosure was established in 1999. The region experiences a continental climate, with a mean average temperature of approximately −1.6°C. The mean average yearly precipitation is about 400 mm, mainly concentrated from June to August. The soil type is classified as chestnut according to Chinese classification or Haplic Calcisol according to the FAO (Food and Agriculture Organization) classification. Perennial grasses, perennial forbs, and shrubs contribute 27, 58, and 5%, respectively, to the aboveground biomass.

In May 2015, a completely randomized block design was used for the experiment (Supplementary Figure S1). The design included four treatments: control, litter input, increased precipitation, and combined litter input and increased precipitation. Each treatment was replicated six times, resulting in a total of 54 plots. Due to the fact that the aboveground biomass in this region was estimated to be around approximately 200 g m⁻² yr.⁻¹ (Evans and Wallenstein, 2012; Ma et al., 2013), the litter addition rates were set at 0 (control), 60 g m⁻² (+30%), and 120 g m⁻² (+60%). Each plot had dimensions of 2 × 2 m² and was spaced 2 m apart.

During each spring season, fresh litter was collected from a 100 × 100 m² area adjacent to the experimental plots. The collected litter was carefully cleared of any non-plant debris and air-dried for subsequent use as litter addition material. This approach ensured the consistency of the added material and the plant species present in the area, providing an accurate simulation of litter input scenarios. The organic C, total N, and total P levels of the litter material were 41.26, 0.30, and 0.027%, respectively. To simulate increased precipitation, iron boxes with a base area equivalent to 15% of the plot area (85 cm in length, 71.5 cm in width, and 15 cm in height) were fixed outside the increased precipitation plots to collect natural precipitation. A 1.5 cm hole was carefully created on the side of the iron box, precisely positioned to face the plot. A 2 m tube was securely attached to the hole. To ensure efficient and uniform precipitation distribution within the plots during rainfall events, the tube was strategically arranged in a zigzag pattern across the plot surface. To further enhance the flow of precipitation, small holes were evenly spaced 3 cm apart along the length of the tube (Supplementary Figure S1).

Soil samples were collected from each plot by randomly obtaining three soil cores (5 cm inner diameter, 10 cm length) from the topsoil (0–10 cm). This sampling procedure was conducted three times throughout the year: in late May, mid-July, and mid-September of both 2016 and 2017. To ensure the representativeness of the samples, the three replicates were mixed together to create a composite sample. The composite sample was then sieved using a 2 mm sieve. From each composite sample, two subsamples of the sieved soil were obtained. One subsample was stored in a refrigerator at 4°C for routine analyses, while the other subsample was kept at −80°C for phospholipid fatty acids (PLFAs) and soil DNA analyses.

Soil temperature and soil water content measurements were conducted within three days following rainfall events. Soil water content was determined by oven-drying samples at 105°C for 24 h. Soil temperature at a depth of 15 cm was measured by an ECH₂O sensor (Em50, Decagon) between 8:00 and 9:00 in both 2016 and 2017. Soil NH₄⁺-N and NO₃-N concentrations were determined using a flow injection autoanalyzer (FIAstar 5,000 Analyzer, Foss Tecator, Denmark) as described by Ma et al. (2015). Soil microbial biomass C was measured using the fumigation-extraction method and calculated according to the formula described by Wu et al. (1994) and Rotbart et al. (2017).

where EC is the difference between C fumigated and unfumigated soil samples, Kec is the conversion factor.

To extract the phospholipid fatty acids (PLFAs) from the soil samples, we followed the procedure described by Bossio and Scow (1998). The separation and identification of the PLFAs were performed using the standard protocol of the Sherlock Microbial Identification System V4.5 (MIDI) in combination with a Gas Chromatograph (Agilent 6850, United States). During the analysis, methyl nonadecanoate fatty acid (19:0) was used as the internal standard. For the characterization of the soil microbial community, specific fatty acids were selected to represent different microbial groups. The fatty acids a13:0, i14:0, i15:0, i16:0, i17:0, and a17:0 were chosen to represent gram-positive bacteria. The fatty acids 16:1ω7c, 17:1ω8c, 18:1ω5c, 18:1ω9t, 17:0cy, and 19:0cy were selected to represent gram-negative bacteria. Additionally, three fatty acids, namely 16:1ω5c, 18:2ω6,9c, and 18:1ω9c, were chosen to represent the fungal group (Olsson et al., 1998).

Genomic DNA was extracted from soil samples using the PowerSoil DNA Isolation Kit from Mo Bio Laboratories. The extracted DNA underwent purity and quality by running it on 0.8% agarose gels. For analysis, the Illumina MiSeq PE300 platform was used. To assess the diversity and composition of soil bacteria and fungi, the V3-4 hypervariable regions of the bacterial 16S rRNA gene were amplified using the primers 806R and 338F. The region of fungal ITS was amplified using the primers ITS1F and ITS2, as described by Caporaso et al. (2012). To ensure unique identification of each soil sample, a 10-digit barcode sequence was appended to the 5′ end of both the forward and reverse primers. PCR amplification was carried out using a Mastercycler Gradient from Eppendorf. For further details on the PCR protocol, refer to the study conducted by Ma et al. (2022).

The raw data underwent quality control procedures, including the removal of sequences with low-quality scores, ambiguous bases, short lengths, and mismatches with primers and barcodes. The processed dataset was analyzed using QIIME, and OTUs were generated at a similarity threshold of 97% to investigate microbial diversity indices. Taxonomic groups were assigned using the Ribosomal Database Project Classifier tool. The OTU datasets were subsampled to compare bacterial and fungal diversity at the same sequencing depth. The number of OTUs served as an indicator of soil microbial taxonomic diversity (Caporaso et al., 2012; Ma et al., 2022).

To evaluate the effects of temporal variations (inter-or intra-annual) in litter input and increased precipitation on soil microclimate, inorganic N, soil microbial biomass C, and microbial community, multivariate analysis of variance (MANOVAs) was employed. To further analyze the differences between the means of the effects, multiple comparisons were conducted using the least significant difference test with a significance level of p < 0.05. All data management and statistical analyses were performed using SPSS 21.0 software (SPSS, Chicago, United States). Principal Component Analysis (PCA) was employed to analyze the changes in soil microbial community structure, estimated by the concentrations of phospholipid fatty acids (PLFAs), in response to litter input and increased precipitation. The PCA analysis was conducted using R statistical software (v3.3.1, R Core Team, 2016).

Throughout the 2016 and 2017 growing seasons, both soil temperature and soil water content exhibited significant seasonal variations. However, low and high litter inputs did not significantly alter soil temperature and water content (Figure 1). The influence of low and high precipitation additions on soil temperature was also not statistically significant. The interaction of litter input and increased precipitation had no significant impact on soil temperature. Furthermore, no significant interactive effects were observed among sampling time, year, and litter addition (Supplementary Table S1).

Significant seasonal variations were observed in soil ammonium content, whereas soil nitrate content remained constant. Both litter input and increased precipitation had no significant impact on the levels of soil ammonium and nitrate. Furthermore, the litter input did not exert any significant effect on soil ammonium and nitrate contents. Furthermore, no interactive effects were observed among sampling time, litter addition, and precipitation addition (Figure 2, Supplementary Table S2).

Soil microbial biomass C exhibited a notable increase with increasing litter input. In 2016, both low and high litter inputs resulted in 12% (p < 0.1) and 22.8% (p < 0.05) increases in microbial biomass C (Figure 3). Similarly, low and high litter inputs led to 12.4 and 28.1% (p < 0.05) increases in microbial biomass C in 2017. However, neither low nor high increased precipitation had a significant impact on soil microbial biomass C. Furthermore, the interactive effects of litter input and increased precipitation did not significantly affect soil microbial biomass C. Neither sampling time nor year displayed significant interactive effects with litter input and increased precipitation treatments on soil microbial biomass C (Supplementary Table S2).

Both low and high litter inputs were found to significantly increase soil bacterial PLFAs and fungal PLFAs, indicating an increase in microbial biomass. However, low and high increased precipitation did not have an effect on bacterial PLFAs and fungi PLFAs (Figure 4). Notably, there were significant interaction effects between litter input and increased precipitation on soil bacterial PLFAs and fungal PLFAs (p < 0.05, Supplementary Table S3). Interestingly, when litter input and increased precipitation interacted, the ratio of soil fungal to bacterial PLFAs remained unchanged.

Litter input had no effect on the soil fungal-to-bacterial PLFAs ratio (i.e., fungal-to-bacterial biomass ratio), whereas increased precipitation significantly increased the soil fungal-to-bacterial PLFAs ratio (Figure 5). Notably, an interaction effect was observed between litter input and increased precipitation on soil fungal-to-bacterial PLFAs (p < 0.05, Supplementary Table S3). Furthermore, sampling time and year displayed significant interactive effects with litter input and increased precipitation treatments on soil bacterial PLFAs and fungal PLFAs (Supplementary Table S3).

Low and high litter inputs significantly increased gram-negative bacterial PLFAs, while they did not affect gram-positive bacterial PLFAs (Figure 6). In contrast, increased precipitation had a significant effect on elevating gram-positive bacterial PLFAs, but it significantly reduced gram-negative bacterial PLFAs. Importantly, there was a significant interaction effect between litter and precipitation additions on both gram-positive bacterial and gram-negative bacterial PLFAs. Sampling time and year exhibited significant interactive effects linked to litter and precipitation addition treatments on soil gram-positive bacterial and gram-negative bacterial PLFAs (Supplementary Table S4).

After the three years, principal component analysis (PCA) showed that litter input and increased precipitation altered the soil microbial community structure. The first principal component (PC1) explained 47.1% of the variation in microbial community structure, and the second principal component (PC2) explained 23.8% of the variation (Figure 7). Specifically, PC1 primarily reflected the gradient of litter input, and PC2 represented the gradient of precipitation. Soil bacterial PLFAs, mainly attributed to gram-negative bacteria, and fungal PLFAs showed an increase in response to litter additions. Although the increased precipitation did not have a significant effect on the fungal-to-bacterial ratio, it did lead to an increase in gram-positive bacteria and a decrease in gram-negative bacteria. Notably, when both litter addition and precipitation addition interacted, there was an overall increase in both soil bacterial and fungal PLFAs.

Neither litter input nor increased precipitation had any effect on the soil bacterial and fungal diversity (i.e., richness) during the third growing season (Figure 8). Notably, an interaction effect was observed between litter input and increased precipitation on soil bacterial diversity (p < 0.05, Supplementary Table S5). Furthermore, sampling time and year displayed significant interactive effects with litter input and increased precipitation treatments on soil bacterial diversity and fungal diversity (Supplementary Table S5).

The findings of this study show that litter addition increased both soil fungal and bacterial biomass in the temperate grassland. This supports previous research that has also found increases in soil microbial biomass following litter addition (Bailey et al., 2002; Creamer et al., 2015). The increase in microbial biomass is likely due to the higher availability of organic C from decomposing litter, which promotes fungal growth. However, contrary to expectations based on previous research, the fungal-to-bacterial biomass ratio did not change significantly (Figure 5). Previous studies have suggested that higher litter accumulation increases this ratio (Bailey et al., 2002; Drenovsky et al., 2004), but the results of this study did not support that hypothesis. However, in the DIRT (Detritus Input Removal and Transfer) experiment, the removal or addition of litter altered the composition of soil microbial communities, leading to shifts in the proportion of soil fungi to bacteria (Siira-Pietikäinen et al., 2003; Nadelhoffer et al., 2004; Brant et al., 2006). In contrast, our results align with the findings of Kirchmann et al. (2004), Morris and Boerner (1999), and Busse et al. (2009), which also reported no significant alteration in microbial community structure despite changes in soil bacterial and fungal communities. The lack of change in the fungal-to-bacterial biomass ratio could be attributed to the complex interactions between litter quality, nutrient availability, and microbial community composition (Fanin et al., 2014). Litter inputs, when accompanied by nutrient enrichment, may alleviate nutrient limitations among microbial groups and help maintain a stable fungal-to-bacterial biomass ratio (Morris and Boerner, 1999). Understanding these dynamics is crucial for predicting the consequences of litter accumulation and decomposition on ecosystem processes and functions.

While the proportion of soil bacterial biomass within the microbial community remained non-significant, gram-negative bacteria in the bacterial community increased with extra litter addition, while the response of gram-positive bacteria biomass to litter addition was not significant (Figure 7). This aligns with the findings of Bossio and Scow (1998) and Zelles et al. (1992). Bossio and Scow (1998) studied the response of soil microbial community structure to substrate additions in agroecosystems, indicating a strong correlation between gram-negative bacterial PLFAs and the availability of soil organic matter. This finding suggests that higher organic matter availability leads to higher gram-negative biomass. Therefore, gram-negative biomass may be a predictor of soil quality in temperate grasslands.

Under the increased precipitation treatments, there was a significant increase in the fungal-to-bacterial ratio in the soil (Figure 6). This shift in the soil microbial community structure can be explained by several factors. Firstly, increased precipitation promotes plant growth, which in turn accelerates the uptake of inorganic N by plants and the leaching of nitrate. As a result, the concentration of inorganic N in the topsoil decreases significantly. This decrease in inorganic N content creates a more nutrient-restricted environment in the temperate grassland, leading to intensified competition between plants and soil microbes for limited N resources (Kuzyakov and Xu, 2013; Liu et al., 2016). Secondly, it is important to note that under nutrient-restricted conditions, soil fungi have a competitive advantage over bacteria (Shao et al., 2022). This advantage stems from the ability of fungal hyphae to extend into different soil layers, extracting the necessary nutrients and accessing a wider range of nutrient sources (Cairney, 2005). Furthermore, fungi possess filamentous characteristics that enable cytoplasmic streaming and nutrient redistribution within their cells, facilitating nutrient recycling (Bonfante and Genre, 2010; Guhr et al., 2015). Consequently, the reduction in inorganic N content resulting from increased precipitation enhances the proportion of fungi in the soil microbial community. This increase in the fungal-to-bacterial biomass ratio highlights the importance of fungal contributions to nutrient cycling and ecosystem functioning in N-limited grassland ecosystems.

Under the increased precipitation treatments, there was a notable shift in the composition of bacterial communities in the soil. Specifically, the biomass of gram-positive bacteria exhibited an increase, while the biomass of gram-negative bacteria presented a significant decrease. This finding is consistent with a previous study at the agricultural field, which revealed that arid soils were characterized by a higher abundance of gram-negative bacteria, whereas relatively moist soils were dominated by gram-positive bacteria (Drenovsky et al., 2004). The observed increase in the biomass of gram-positive bacteria under higher precipitation levels can be attributed to several factors. Firstly, the additional precipitation promotes plant growth, leading to an increased input of organic matter into the soil. Gram-positive bacteria are known to thrive in nutrient-rich environments and are efficient decomposers of organic material. Therefore, the availability of organic matter in the soil, due to increased plant growth, favors the proliferation of gram-positive bacteria (Fanin et al., 2019). Furthermore, the decrease in the biomass of gram-negative bacteria can be explained by changes in nutrient availability and competition for resources. Under nutrient-restricted conditions caused by the decrease in inorganic N content, competition for limited resources intensifies between plants and soil microbes. Gram-negative bacteria, which are more specialized in utilizing inorganic N sources, may face greater competition from plants for these resources. In contrast, gram-positive bacteria have a wider range of nutrient acquisition strategies and can effectively utilize organic matter, giving them a competitive advantage under nutrient-restricted conditions (Francis et al., 2010).

Currently, there is limited research on the effects of litter and precipitation additions on soil microbial community structure in arid and semi-arid regions (Creamer et al., 2015). The findings of this study indicated that increased precipitation had a significant effect on the ratio of soil fungal to bacterial biomass. However, intriguingly, when increased precipitation was combined with litter addition, there was no notable alteration in the fungal-to-bacterial biomass ratio. This finding suggests that the effect of increased precipitation on microbial community structure is altered when combined with litter input, indicating that the litter effect modulates the precipitation effect in the temperate grassland ecosystem. One possible explanation for this observation is that litter addition, which includes nutrient release, ameliorates nutrient limitations for microbes, thereby attenuating the impact of increased precipitation on microbial community structure.

The combined effect of litter input and increased precipitation may have a greater impact on bacterial diversity compared to their individual effects (Figure 8). There are several potential reasons for this interaction effect. Firstly, both litter input and increased precipitation can influence soil moisture, a crucial factor for microbial diversity. Litter input acts as a mulch layer, reducing evaporation and promoting water availability for microbes (Wang et al., 2016). Increased precipitation directly adds moisture to the soil, potentially creating more favorable conditions for microbial growth. Therefore, the combination of litter input and increased precipitation can result in higher soil moisture levels, enhancing microbial diversity. Secondly, litter input and increased precipitation can provide additional organic matter and nutrients to the soil, which can fuel microbial growth and activity (Ma et al., 2013; Wu et al., 2017). Litter input serves as a source of carbon and other nutrients utilized by bacteria and fungi. Increased precipitation can leach nutrients from litter and other organic materials, making them more accessible to microbial communities. This increased nutrient availability can support the growth of a wider range of bacterial species, resulting in higher diversity. In order to fully understand the underlying mechanisms and long-term effects of these interactions on microbial communities and ecosystem functioning, further research is required.

Based on the findings mentioned, future research can focus on several aspects to enhance our understanding of the interactions between organic matter, water, and soil microbial communities. Firstly, it is important to explore how organic matter and water influence the functional traits of soil microbial communities. In addition to their impact on microbial community structure, organic matter and water may also alter the functional characteristics of microbial communities, such as their ability to cycle C and N. By investigating these functional changes, we can gain valuable insights into how organic matter and water affect the overall functioning of soil ecosystems. Secondly, it would be beneficial to investigate the effects of different types of organic matter on the structure and function of soil microbial communities. Leaf litter and woody debris, for example, may have distinct impacts on microbial communities due to their unique chemical composition and rates of decomposition. By understanding how different types of organic matter influence microbial communities, we can further elucidate the underlying mechanisms driving these interactions. Lastly, integrating experimental research with field observations can provide insights into the long-term effects and ecological significance of organic matter and water on soil microbial communities. Long-term monitoring and experiments can better simulate and predict the impacts of future climate change on soil microbial communities, and provide scientific basis for ecosystem management and conservation.