The Changbai Mountain is located in Jilin Province, northeastern China (Zhao et al., 2022). The peatlands in the Changbai Mountain are dominated by sedges (Bao et al., 2010; Wang et al., 2018). The climate in this region is a typical continental monsoon climate; the mean annual temperature (MAT) in the study area ranges from −0.2°C to 3.9°C, and the mean annual precipitation (MAP) ranges from 580 to 770 mm in the study area (Zhao et al., 2022).

Samples were collected in July 2020. Soil cores at 0–30 cm depth with an interval of 5 cm were collected in four altitude gradients (200–500, 500–800, 800–1,200 and 1,200–1,500 m; Figure 1). Two peat samples were collected in each altitudinal gradient. Peat samples were kept at −20°C immediately after collection. Each sample was separated into two subsamples. One was used for the measurement of soil water content (SWC), and the other was freeze-dried and sieved for the measurement of extracted microbial lipids and soil physicochemical properties.

The analysis of microbial abundance was used by phospholipid fatty acids (PLFAs) technology. The extraction and separation of PLFAs were performed followed the Bligh Dyer method (Bligh and Dyer, 1959). The sample test method was described by Zhao M. L. et al. (2021). The detected compounds were identified in the MIDI library (MIDI, Inc., Newark, United States) (Zhang et al., 2019). The PLFAs were used as the microbial biomarkers according to Joergensen (2021). The gram-positive bacteria (G+) were the sum of the abundances of Firmicutes and Actinobacteria, the G+ and gram-negative bacteria (G-) belong to bacteria (B). The total microbial abundance was the sum of the abundances of bacteria, fungi (F) and unspecific microbial biomarkers.

Soil water content (SWC) was measured by the gravimetric method. Soil pH was measured by a potentiometric test with a soil to water ratio of 1:10. The total organic carbon (TOC) and dissolved organic carbon (DOC) were detected on a TOC analyzer (Shimazu, Japan). The total phosphorus (TP) and total nitrogen (TN) were determined by an automated analyzer (Smartchem140, AMS-Alliance, and French; Lu, 2000).

Co-occurrence network was used to show microbial biomarker interactions, the analysis was visualized using Gephi v.0.9.1. Each node means one microbial lipid and each edge means a strong relationship between two nodes. The topology of the co-occurrence networks was assessed referred to previous studies (Tu et al., 2020; Li J. et al., 2022). Briefly, average degree indicates the complexity of the co-occurrence network. Average path length indicates the distance between any two members of the co-occurrence network. A higher average degree indicates a higher complexity of the network, a shorter average path length suggests a stronger correlation between members.

Two-way ANOVAs were used to examine the main effect of altitude and soil depth and their interaction on peat physicochemical properties and soil microbial abundance. Redundancy analysis (RDA) was conducted to analyze the relationship of microbial abundance to peat physicochemical properties. Variation decomposition analysis (VDA) was used to analyze the effects of peat physicochemical properties and microbial co-occurrence networks on soil total microbial abundance. The physical (pH and SWC) and nutrient (DOC, TN, TOC, and TP) variables were used as abiotic factors and the microbial ecological interaction network was used as the biotic factor in the prediction model. Based on previous studies, the microbial co-occurrence networks are represented by the first two axes’ scores of the principal component analysis for microbial community composition (Purahong et al., 2016). A positive coefficient of VDA indicates a positive effect on the prediction of the total microbial abundance, and a negative coefficient suggests the opposite (Gross et al., 2017). The data were log₁₀-transformed to conform to normality and homogeneity of variance. The analyses were performed using SPSS 21.0, Canoco 5.0, Origin 29.0, and R 4.1.1 with the packages vegan (Oksanen et al., 2020), MuMIn (Bartoń, 2022), performance (Lüdecke et al., 2021), ggplot2 (Wickham, 2016), and ggh4x (van den Brand, 2021).

Soil depth and altitude significantly affected the total microbial abundance (Table 1). The effect of soil depth on the abundance of the microbial group was stronger than the altitude (Table 1). The soil layer of 5–10 cm had the highest total microbial concentration, which then decreased with soil depth (Figure 2A). The concentration of total microbial PLFAs was higher at 500–800 m than at other altitudes (Figure 2B). The concentrations of G+, G−, and F showed a similar trend with the total microbial concentration (Figure 2).

Soil depth significantly affected DOC, TN, and pH (Table 1). As soil depth increased, DOC decreased and TN generally increased. No significant difference was found in SWC, TOC, and TP between different depths. Altitude significantly affected SWC (Table 1). SWC generally increased with the altitude. No significant difference was found in soil pH, TN, TP, and TOC along the altitudinal gradient.

Microbial co-occurrence network did not show significant differences along the altitudinal gradient (Figure 3A), but it had significant differences between soil depths (Figure 3B). As soil depth increased, the negative interaction ratio and average path length of co-occurrence networks decreased, and the positive interaction ratio, average degree, average clustering coefficients, and graph density of co-occurrence networks gradually increased (Table 2; Figure 3B).

The RDA analysis found that the first and second axes together explained over 60.0% of the total variation in microbial groups (Figure 4). Soil pH, TN, DOC, and depth were the main factors that significantly affected the microbial abundance. Samples from the top soil layers were located in the lower quadrant of the RDA graph while samples from deeper soil layers were located in the upper quadrant (Figure 4).

The VDA analysis showed that the total microbial abundance was significantly affected by both abiotic and biotic factors. Soil nutrients including DOC and TN had a positive correlation with the total microbial abundance. Soil physical properties (pH and SWC) and the microbial interaction network had a negative relationship with the total microbial abundance (Figure 5).

Soil depth has a great impact on soil microbes because of the unequal distribution of plant roots and soil nutrients across soil profiles (Rousk et al., 2010). Soil nutrients and physical environments were the main factors affecting the total microbial abundance in peatlands. Soil nutrients provided energy for the growth, metabolism, and reproduction of microorganisms (Zhang et al., 2022). In our study, soil depth significantly affected soil physicochemical properties and microbial abundance in peatlands. Soil nutrients had a positive influence on the soil microorganisms and soil physical factors had a negative impact on the abundance of soil microbes (Figure 5).

Soil nutrient is a key factor affecting microbial abundance in peatlands. DOC can be used as a carbon substrate for soil microbes, and TN can alleviate carbon limitation on soil microorganisms (Guo et al., 2011; Zhou et al., 2017). In our study, DOC and TN differed significantly between soil depths, and they significantly affected the total microbial abundance (Table 1; Figure 5). This result was consistent with former research (Bradley et al., 2006; Jia et al., 2020; Zhao M. L. et al., 2021). DOC is the most active intermediate in carbon cycle process because of its high mobility and bioavailability (Marschner and Kalbitz, 2003). DOC is utilized as a substrate, leading to microbial mineralization and CO₂ emissions in peatlands (Battin et al., 2008; Zhang et al., 2022). Nitrogen accumulation increases microbial abundance because it increases the utilization rate of nitrogen resources, which can alleviate carbon limitation on soil microorganisms or inhibit the limitation of carbon caused by soil acidification (Guo et al., 2011; Zhou et al., 2017). This finding indicated that nitrogen accumulation had a positive effect on microbial abundance. The terrestrial surface temperature is projected to exceed 2°C by the end of this century, and the atmospheric nitrogen deposition and the intensity of extreme precipitation events will increase (IPCC, 2022). These changes may lead to more DOC exports from the peatlands (Cole et al., 2002; Clark et al., 2010) and an increasing nitrogen accumulation in peatlands (Zhang et al., 2018, 2022). Our findings suggest that DOC and TN positively affect soil microbial abundance and microbial activities, which may lead to higher CO₂ emissions in peatlands.

Soil physical properties also affect soil microbial abundance. Soil pH changed cell membrane charge and thus influenced the nutrient absorption by soil microorganisms and the enzyme activity in metabolic processes (Rousk et al., 2010). The acidic environment in peatlands is conducive to the growth of microorganism, and the microbial abundance generally decreased as soil pH increased in our study, which is consistent with former research (Anderson et al., 2010). Water regime could also affect soil microbial abundance. The effects of water drainage on soil microbial community and enzyme activity were dependent on soil depth in peatlands (Xu et al., 2021). When SWC was high, soil microbial abundance decreased because microbial heterotrophic respiration of microorganisms was inhibited (Wagner et al., 2015).

The complexity of the co-occurrence network indicated microbial interactions along the environmental gradient (Ma et al., 2020). Environmental variables are essential for microbial niche differentiation, which enables distinct microbial groups to obtain adequate substrate and survive under various environmental conditions (Wiens et al., 2010; Li et al., 2020). The wide range of edaphic environments has driven the assembly process of soil microorganism (Dini-Andreote et al., 2015; Tripathi et al., 2018). In our study, the complexity of microbial co-occurrence networks increased and the microbial abundance decreased with soil depth (Table 2), and the VDA analysis indicated that microbial community assembly was one major factor affecting microbial abundance (Figure 5). The availability of carbon, energy and oxygen decreased with soil depth in peatlands, the competition of microorganisms for the resource increased, and their interactions increased. Resource limitation leads to the reduction of microbial abundance (Lu et al., 2020). Our results are consistent with a recent study which found the depth effect on bacterial community assembly processes in paddy soils (Li W. T. et al., 2022).