Impact of Fire Severity on Soil Bacterial Community Structure and Its Function in Pinus densata Forest, Southeastern Tibet

Abstract

Forest fires are one of the significant factors affecting forest ecosystems globally, with their impacts on soil microbial community structure and function drawing considerable attention. This study focuses on the short-term effects of different fire intensities on soil bacterial community structure and function in Abies (Pinus densata) forests within the Birishen Mountain National Forest Park in southeastern Tibet. High-throughput sequencing technology was employed to analyze soil bacterial community variations under unburned (C), low-intensity burn (L), moderate-intensity burn (M), and high-intensity burn (S) conditions. The results revealed that with increasing fire severity, the dominant phylum Actinobacteriota significantly increased, while Proteobacteria and Acidobacteriota markedly decreased. At the genus level, the relative abundance of Bradyrhizobium declined significantly with higher fire severity, whereas Arthrobacter exhibited a notable increase. Additionally, soil environmental factors such as available phosphorus (AP), dissolved organic carbon (DOC), C/N ratio, and C/P ratio displayed distinct trends: AP content increased with fire severity, while DOC, C/N ratio, and C/P ratio showed decreasing trends. Non-metric Multidimensional Scaling (NMDS) analysis indicated significant differences in soil bacterial community structures across fire intensities. Diversity analysis demonstrated that Shannon and Simpson indices exhibited regular fluctuations correlated with fire severity and were significantly associated with soil C/N ratios. Functional predictions revealed a significant increase in nitrate reduction-related bacterial functions with fire severity, while nitrogen-fixing bacteria declined markedly. These findings suggest that forest fire severity profoundly influences soil bacterial community structure and function, potentially exerting long-term effects on nutrient cycling and ecosystem recovery in forest ecosystems.

1. Introduction

Forest fires lead to approximately 1% of annual forest loss worldwide [1,2], which results in severe disturbance of ecological balance, such as biodiversity loss of microorganisms [3]. Frequent forest fires, as a key driving factor in the biogeochemical cycles of elements, cause changes in biodiversity, forest carbon stocks, and carbon distribution patterns, which in turn affect forest succession processes and carbon sequestration capacity. For example, after a fire in a northern China wetland reserve, the total loss of organic carbon was 248.40 t C/hm², and the loss of nitrogen was 11.87 t N/hm² [4]. After forest fires, black carbon formation and CO₂ release in forest soils are inversely related as incomplete combustion favors black carbon formation while complete combustion leads to greater CO₂ release [5]. The substantial environmental changes caused by forest fires greatly disturb soil environments, altering the physical, chemical, and biological properties of forest soils [6]. The physicochemical characteristics of forest soils not only affect the acquisition of nutrients by vegetation but also interact with soil microorganisms. Soil microorganisms are crucial drivers of soil organic matter and nutrient transformation and cycling [7], participating in a range of complex biochemical reactions [8]. Different studies have shown that the impact of fire on soil microbial groups is quite complex. In the boreal forests of northwestern Canada [9], after a fire, the abundance of bacteria such as Massilia and Arthrobacter, as well as fungi such as Penicillium and Fusicladium in the soil have increased. In the mixed coniferous and broad-leaved forests in the transitional area between subtropical and warm-temperate climates in China, influenced by the fire, the activities of extracellular enzymes involved in the cycling of carbon, nitrogen, and phosphorus reached their highest levels one year after the fire and then decreased over time [10]. On one hand, soil microorganisms can influence the habitat of vegetation by altering the soil environment; they can enhance nutrient absorption and utilization by plants through symbiotic relationships with roots, playing an important role in soil nutrient cycling and vegetation succession [11]. Soil microorganisms are highly sensitive to changes in soil environments and can promptly reflect alterations in soil characteristics [12]. For example, if nutrient levels such as nitrogen or phosphorus change, microbial communities can shift to favor species that are more efficient at utilizing these nutrients [10]. Therefore, changes in microbial community indicators can rapidly and accurately reflect shifts in soil ecosystems. The direct impact of forest fires includes the death of soil microorganisms due to the rapid increase in soil temperature [6], while indirect effects involve changes in soil physicochemical properties that modify the microbial environment, subsequently affecting their survival and reproduction, as evidenced by changes in microbial community characteristics [10]. Previous studies on soil microbial community responses to fire severity have revealed global patterns. For instance, in Mediterranean ecosystems, intense fires were shown to reduce bacterial diversity by >50% through thermal degradation of organic substrates [13]. Similarly, Knelman et al. [11] demonstrated the cascading effects of fire severity on microbial trophic networks in Colorado conifer forests, while Zheng et al. [14] observed transient diversity increases after low-intensity fires in Chinese Siberian pine forests, contrasting results emerged from Australian eucalyptus ecosystems where even moderate fires caused prolonged microbial suppression [15]. These geographical disparities highlight the need for context-specific investigations, particularly in vulnerable high-altitude ecosystems. Studies by Li et al. [16] in the North China Pinus tabuliformis secondary forest and Wang et al. [17] in the Larix gmelinii forests of the Greater Khingan Mountains found that with an increase in the fire severity ratio, the relative abundance of bacteria to fungi decreased significantly. In terms of microbial community function, Fontúrbel et al. [13] investigated the effects of fire on soil microorganisms in Mediterranean forests and found that fire increased functional diversity among soil microorganisms. Overall, research on the impact of fire severity on soil microbial diversity and community structure shows consistent findings, indicating that increased fire severity generally reduces microbial diversity while altering community structure.

The Qinghai–Tibet Plateau in China is characterized by high altitude, low temperatures, arid climate, and low soil fertility, which contribute to relatively low stability in forest ecosystems. The forests of the Tibetan Plateau, covering approximately 218,000 square kilometers, contribute 31.7% to the total ecosystem service value of the plateau [4], and are also at risk from forest fires induced by climate change. Currently, there is limited understanding of how forest fire severity affects soil microbial communities in the forests of the Qinghai–Tibet Plateau, and research on how fire severity impacts soil bacterial community functions is lacking. Therefore, this study focuses on the Pinus densata forest in the Biri Sacred Mountain National Forest Park in southeastern Tibet. Using high-throughput sequencing technology, we address three specific research objectives: (1) How do different fire severity levels alter the composition and diversity patterns of soil bacterial communities? (2) What are the key soil environmental factors mediating these microbial responses? (3) Do fire-induced community shifts correspond to functional changes in nitrogen cycling processes? Through systematic analysis of bacterial community structure, diversity indices, and functional gene profiles in relation to soil C/N/P stoichiometry, this work establishes ecological baselines to inform both natural successional trajectories and science-based microbial resource utilization strategies for post-fire forest restoration.

2. Materials and Methods

2.1. Study Area

The study area is located in Birishenshan National Forest Park, Bayi District, Linzhi City, Tibet Autonomous Region, with a total area of 2.23 × 10³ km², of which 126 km² is forested, with a forest coverage of 55.91%, which is about 48% of the total forested area in southeast Tibet. The average altitude is 4500 m. The area is in the tropical humid and semi-humid climate zone, which is characterized by mild and rainy climate in summer, strong exposure to light, and large temperature difference between day and night; less rainfall in winter, dry air, and annual rainfall of about 746 mm. The average annual temperature is 9.78 °C, the average annual insolation is 2000 h, and the frost-free period is 193 d. The forest vegetation type is dominated by Pinus densata Mast., Quercus aquifolioides Rehd. et Wils., and Betula utilis D. Don. The forest soil is weakly acidic, which is classified as Dystrudepts (a subgroup within the Inceptisols order) under the USDA Soil Taxonomy, with a thickness of about 40 cm, so the 0–10 cm surface soil was taken for the determination of relevant indicators. The geographical coordinates of the sample site are 29°39′16.43″ N, 94°21′52.61″ E, with an altitude of about 2998 m. A forest fire occurred in the study area at the end of February 2021, the fire type was surface fire, and the burned area was about 0.1 km².

2.2. Site Selection and Sampling Acquisition

At the beginning of June 2021, areas with consistent site conditions such as slope, aspect, elevation, and low anthropogenic disturbances were selected. In the selected areas, sampling zones of different fire intensities were established based on the survival of trees and the height of blackening of tree trunks in the burned sites. The neighbouring unburned sample plots were used as control plots, which were divided as follows: (1) unburned sample plots, with all plants surviving and no signs of blackening on the trunks of the trees, were used as control check (C); (2) mildly burned sample plots, with a tree mortality of <30%, and with blackening on the trunks of the trees of all the trees of heights less than 2 m, were used as low fire severity (L); (3) Moderate fire severity (M) in burned sample plots, with a tree mortality of 30%–70% and all tree trunks blackened to a height of 2–5 m; (4) High fire severity (H) in burned sample plots, with a tree mortality of >70% and all tree trunks blackened to a height of more than 5 m. The sampling method was as follows: in each sample plot, 0–10 cm surface soil samples were drilled several times with a soil auger with an inner diameter of 3.5 cm to form mixed soil samples (five-point sampling method), there were three replicates of the samples taken to form 12 soil samples of at least 1000 g each, which were packed in polythene sealed bags and quickly placed in a refrigerator to be brought back to the laboratory. After removing all kinds of residues in the soil samples, part of the soil samples were crushed and passed through a 1 mm soil sieve for the determination of soil physicochemical indexes, and the other part of the soil samples were frozen and stored at −20 °C for high-throughput sequencing of soil bacterial communities.

2.3. Soil Physicochemical Indicators Determination

According to Bao [14], soil pH (acid meter method), electrical conductivity (electrical conductivity method), soil moisture content (drying method), total phosphorus content (molybdenum antimony antimonide method), soil total potassium content (molybdenum antimony melting method), available phosphorus content (neutral ammonium acetate leaching-flame photometric method), soil quick-acting potassium content (ammonium acetate leaching-flame photometric method), soil nitrate nitrogen and ammonium nitrogen (continuous flow analyser method) were determined. For soil organic matter content, potassium dichromate method was used. For soil total nitrogen content, semi-micro Kjeldahl method was used [18]. Soil-dissolved organic nitrogen content was determined using AA3 continuous flow analyzer (Seal Analytical, Mequon, WI, USA) [19] and soil dissolved organic carbon content was determined using Multi N/C 2100 TOC meter [20].

2.4. High-Throughput Sequencing of Soil Bacterial Communities

Microbial community genomic DNA was extracted from 12 samples using the E.Z.N.A.^® soil DNA Kit (Omega Bio-tek, Norcross, GA, USA) according to the manufacturer’s instructions. The DNA extract was checked on 1% agarose gel, and DNA concentration and purity were determined with a NanoDrop 2000 UV–vis spectrophotometer (Thermo Scientific, Wilmington, DE, USA). The hypervariable region V3-V4 of the bacterial 16S rRNA gene was amplified with primer pairs 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R(5′-GGACTACHVGGGTWTCTAAT-3′) by an ABI GeneAmp^® 9700 PCR thermocycler (ABI, Foster City, CA, USA). The PCR amplification of the 6S rRNA gene was performed as follows: initial denaturation at 95 °C for 3 min, followed by 27 cycles of denaturing at 95 °C for 30 s, annealing at 55 °C for 30 s and extension at 72 °C for 45 s, and single extension at 72 °C for 10 min, and end at 4 °C. The PCR mixtures contain 5 × TransStart FastPfu buffer 4 μL, 2.5 mM dNTPs 2 μL, forward primer (5 μM) 0.8 μL, reverse primer (5 μM) 0.8 μL, TransStart FastPfu DNA Polymerase 0.4 μL, template DNA 10 ng, and finally ddH₂O up to 20 μL. PCR reactions were performed in triplicate. The PCR product was extracted from 2% agarose gel and purified using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA) according to manufacturer’s instructions and quantified using Quantus™ Fluorometer (Promega, Madison, WI, USA).

Purified amplicons were pooled in equimolar and paired-end sequenced (2 × 300) on an Illumina MiSeq platform (Illumina, San Diego, CA, USA) according to the standard protocols by Majorbio Bio-Pharm Technology Co., Ltd. (Shanghai, China). The raw reads were deposited into the NCBI Sequence Read Archive (SRA) database.

The raw 16S rRNA gene sequencing reads were demultiplexed, quality-filtered by Trimmomatic and merged by FLASH with the following criteria: (i) the 300 bp reads were truncated at any site receiving an average quality score of <20 over a 50 bp sliding window, and the truncated reads shorter than 50 bp were discarded, reads containing ambiguous characters were also discarded; (ii) only overlapping sequences longer than 10 bp were assembled according to their overlapped sequence. The maximum mismatch ratio of the overlap region is 0.2. Reads that could not be assembled were discarded; (iii) Samples were distinguished according to the barcode and primers, and the sequence direction was adjusted, exact barcode matching, 2 nucleotide mismatch in primer matching.

Operational taxonomic units (OTUs) with a 97% similarity cutoff were clustered using UPARSE (version 7.1, http://drive5.com/uparse/, accessed on 7 March 2025), and chimeric sequences were identified and removed. The taxonomy of each OTU representative sequence was analyzed by the RDP Classifier (https://anaconda.org/bioconda/rdp_classifier, accessed on 7 March 2025) against the 16S rRNA database using a confidence threshold of 0.7.

2.5. Data Processing and Analysis

The sequencing data were referenced against the NCBI (NT) database [21]. Statistical analyses were performed using SPSS 20.0 [22] and R 4.3.1 (https://www.rstudio.com/, accessed on 7 March 2025). For group comparisons across fire intensities, we first assessed data normality using Shapiro–Wilk tests and homogeneity of variances with Levene’s test. When assumptions were met, one-way ANOVA with Tukey’s HSD post hoc test was applied; otherwise, the non-parametric Kruskal–Wallis test followed by Dunn’s multiple comparisons was used. Beta diversity was analyzed through non-metric multidimensional scaling (NMDS) based on Bray–Curtis distances, implemented via the vegan package (v2.6-4) in R. Permutational multivariate analysis of variance (PERMANOVA, 999 permutations) with adonis2 function was employed to test fire severity effects on community structure. Functional predictions were generated using FAPROTAX (v1.2.1) with default parameters. All visualizations were created using ggplot2 (v3.4.2) and heatmaply (v1.5.0) packages in R 4.3.1.

3. Results

3.1. Soil Bacterial Community Composition and Structure

After a short-term recovery period of three months, it was observed that for the dominant bacterial phyla—Actinobacteriota, Proteobacteria, and Acidobacteriota—Actinobacteriota significantly increased with increasing fire severity, while Proteobacteria and Acidobacteriota significantly decreased (p < 0.01). Among the dominant genera, Bradyrhizobium significantly decreased with increasing fire severity, whereas Arthrobacter significantly increased (p < 0.01) (Figure 1).

3.2. Soil Environmental Factors

Among many soil physicochemical indicators, most of them did not change significantly due to the increase in forest fire severity, only the available phosphorus (AP), dissolved organic carbon (DOC), C/N, and C/P had a relatively obvious trend: with the increase in fire severity, the available phosphorus (AP) had a tendency to increase, while dissolved organic carbon (DOC), C/N, C/P had a tendency to decrease (

Table 1. Soil physicochemical properties for soils exposed to no fire, or low, moderate or high severity fire on the Qinghai–Tibet Plateau in southwestern China. CK: Control check; L: Low fire severity; M: Moderate fire severity; H: Severe fire severity; pH: pondus hydrogenii; EC: Electrical conductivity; WC: Water content; SOM: Soil organic matter; TN: Total nitrogen; TP: Total phosphorus; TK: Total kalium; AP: Available phosphorus; AK: Available kalium; NO₃⁻-N: Nitrate nitrogen; NH₄⁺-N: Ammonia nitrogen; DON: Dissolved organic nitrogen: DOC: Dissolved organic carbon; All data are mean ± standard deviation; Different letters in the same column indicate significant differences (p < 0.05).

Sample	CK	L	M	H
pH	6.17 ± 0.43 a	5.97 ± 0.37 a	5.91 ± 0.22 a	6.02 ± 0.16 a
EC/(μs/cm)	33.14 ± 10.22 a	41.55 ± 2.99 a	62.21 ± 23.20 a	44.26 ± 19.67 a
WC/%	24.96 ± 4.46 a	18.04 ± 5.04 ab	22.89 ± 3.01 ab	11.36 ± 2.68 b
SOM/%	4.96 ± 0.84 a	5.25 ± 0.80 a	5.02 ± 1.37 a	4.27 ± 0.91 a
TN/%	0.11 ± 0.01 a	0.13 ± 0.03 a	0.14 ± 0.02 a	0.13 ± 0.01 a
TP/%	0.05 ± 0.01 a	0.05 ± 0.01 a	0.06 ± 0.02 a	0.07 ± 0.01 a
TK/%	1.04 ± 0.03 a	1.08 ± 0.06 a	1.08 ± 0.05 a	1.02 ± 0.02 a
AP/(mg/kg)	19.65 ± 3.33 b	33.09 ± 9.08 b	61.84 ± 22.34 a	61.84 ± 10.91 a
AK/(mg/kg)	179.65 ± 24.94 a	189.52 ± 20.53 a	304.68 ± 123.67 a	201.37 ± 26.56 a
NO3−-N/(mg/kg)	0.24 ± 0.06 a	0.16 ± 0.03 a	0.19 ± 0.02 a	0.18 ± 0.04 a
NH4+-N/(mg/kg)	9.15 ± 1.88 a	9.21 ± 3.16 a	21.60 ± 7.94 b	9.06 ± 5.71 a
DON	27.04 ± 6.03 ab	25.69 ± 4.73 b	39.73 ± 9.79 a	21.61 ± 5.75 b
DOC	250.05 ± 42.38 a	205.31 ± 59.66 ab	167.03 ± 39.58 ab	144.05 ± 52.22 b
C/N	25.33 ± 1.86 a	24.59 ± 2.33 a	21.16 ± 2.84 ab	19.64 ± 1.86 b
C/P	57.94 ± 15.64 a	57.04 ± 7.87 a	48.14 ± 6.13 a	39.25 ± 11.62 a

).

To determine whether the soil bacterial community structure differs under varying fire intensities, NMDS (Non-metric Multidimensional Scaling) analysis was conducted at both the phylum and genus levels (Figure 2). The results indicated that the analysis at both taxonomic levels was highly representative (stress < 0.05), with soil bacterial community samples from different fire intensities being distantly separated, suggesting low similarity and significant differences in community structure (Figure 3).

A correlation heatmap analysis was used to explore the relationships between soil bacterial communities and soil environmental factors. It was found that at both the phylum and genus levels, dissolved organic carbon (DOC), C/N ratio, C/P ratio, and AP were among the factors most strongly correlated with soil bacterial communities (Figure 3).

3.3. Soil Bacterial Community Diversity

The analysis of α-diversity indices revealed nuanced patterns across fire intensities (

Table 2. α-diversity at genus level (mean ± SD) for soils exposed to no fire, or low, moderate or high severity fire on the Qinghai–Tibet Plateau in southwestern China.Note: All data are mean ± standard deviation, and different letters in the same column indicate significant differences (p < 0.05).

Sample	Shannon	Simpson	Ace	Chao
C	3.92 ± 0.23 a	0.05 ± 0.01 a	327.39 ± 20.55 a	329.84 ± 23.68 c
L	4.24 ± 0.16 a	0.03 ± 0.01 a	394.95 ± 18.75 b	393.50 ± 17.64 a
M	3.83 ± 0.06 a	0.08 ± 0.02 a	346.85 ± 10.55 a	351.25 ± 5.95 b
H	3.20 ± 0.42 b	0.17 ± 0.07 b	305.79 ± 30.48 a	320.11 ± 27.66 c

). While the unburned control (C) exhibited the lowest bacterial diversity, light fire severity (L) showed a significant increase in diversity indices compared to C (Ace: +20.6%, p = 0.012; Chao: +19.3%, p = 0.018; Shannon: +8.2%, p = 0.043). However, with increasing fire severity from moderate (M) to severe (H), diversity indices progressively declined relative to L (Ace: −22.6%, p = 0.026; Chao: −18.6%, p = 0.031; Shannon: −24.5%, p = 0.008). Notably, H displayed comparable diversity to CK (p > 0.05), indicating a non-linear relationship between fire severity and diversity.

The correlation between diversity indices and soil physicochemical properties revealed that the C/N ratio had a significant positive correlation with the Shannon index and a significant negative correlation with the Simpson index (p < 0.05) (

Table 3. Pearson Correlation between soil bacterial community diversity and soil environmental factors. Note: * indicates significant correlation at the 0.05 level (two-tailed).

Soil Environmental Factors	Shannon	Simpson	Ace	Chao
pH	0.095	−0.051	−0.181	−0.205
EC	−0.14	0.175	−0.08	−0.056
WC	0.5	−0.521	0.039	−0.048
SOM	0.304	−0.334	0.3	0.166
TN	−0.114	0.102	0.043	−0.02
TP	−0.139	0.09	0.034	0.048
TK	0.372	−0.241	0.299	0.298
AP	−0.377	0.398	−0.152	−0.104
AK	0.084	−0.099	0.092	0.053
NO−3-N	0.109	−0.145	−0.376	−0.514
NH+4-N	−0.015	0.014	−0.042	−0.056
DON	0.131	−0.134	0.014	−0.055
DOC	0.283	−0.393	0.104	−0.028
C/N	0.585 *	−0.611 *	0.414	0.295
C/P	0.383	−0.38	0.198	0.084

). This suggests that the C/N ratio may be the primary soil environmental indicator influencing soil bacterial diversity.

Further analysis of the differences in bacterial function abundances between groups revealed that the abundances of aerobic-chemoheterotrophy and chemoheterotrophy did not show significant trends. However, bacterial functions relating to nitrate reduction were significantly increased with increasing fire severity. In contrast, nitrogen-fixing bacteria significantly decreased (p < 0.05) (Figure 4).

Among the various functional bacteria, those involved in aromatic compound degradation, human pathogens causing pneumonia, general human pathogens, animal parasites or symbionts, and urea hydrolysis are not analyzed in detail due to their limited association with the biogeochemical cycles of forest elements. They are presented merely as a phenomenon. The abundance of soil functional bacteria may be constrained by soil environmental factors. For the functional bacteria that showed significant increases or decreases with varying fire intensities, AP, dissolved organic carbon (DOC), C/N ratio, and C/P ratio were all significantly correlated. Specifically, AP was significantly positively correlated with all functional bacteria except for nitrogen-fixing bacteria (nitrogen_fixation). Dissolved organic carbon (DOC) and the C/N ratio were significantly positively correlated with nitrogen-fixing bacteria, but significantly negatively correlated with other functional bacteria. Additionally, electrical conductivity was significantly positively correlated with aerobic chemoheterotrophy and chemoheterotrophy (

Table 4. Pearson’s correlation between functional gene abundance of soil bacterial communities and soil environmental factors. Note: * indicates significant correlation at the 0.05 level (two-tailed) and ** indicates significant correlation at the 0.01 level (two-tailed).

Functional Gene	pH	EC	WC	SOM	TN	TP	TK	AP	AK	NO3-N	${\mathbf{N}\mathbf{H}}_4^{+}-\mathbf{N}$	DON	DOC	C/P
Chemoheterotrophy	−0.111	0.676 *	−0.252	−0.186	0.140	0.415	0.408	0.676 *	0.539	−0.472	0.240	0.095	−0.493	−0.491
aerobic_chemoheterotrophy	−0.108	0.676 *	−0.255	−0.185	0.136	0.422	0.405	0.679 *	0.544	−0.472	0.237	0.091	−0.493	−0.493
nitrogen_fixation	−0.044	−0.552	0.345	0.361	−0.256	−0.251	−0.081	−0.677 *	−0.258	0.165	−0.320	−0.160	0.610 *	0.513
Ureolysis	−0.219	0.356	−0.390	−0.517	0.019	0.324	0.249	0.704 *	0.211	−0.433	0.173	−0.064	−0.879 **	−0.638 *
animal_parasites_or_symbionts	−0.226	0.359	−0.479	−0.533	−0.075	0.396	0.272	0.735 **	0.202	−0.460	0.019	−0.231	−0.884 **	−0.697 *
human_pathogens_all	−0.164	0.383	−0.505	−0.496	−0.016	0.375	0.266	0.739 **	0.213	−0.481	0.053	−0.190	−0.874 **	−0.653 *
human_pathogens_pneumonia	−0.170	0.353	−0.507	−0.492	−0.033	0.390	0.258	0.735 **	0.211	−0.483	0.036	−0.209	−0.880 **	−0.653 *
nitrate_reduction	0.083	0.525	−0.424	−0.605*	−0.058	0.360	0.082	0.677 *	0.230	−0.482	0.179	−0.098	−0.789 **	−0.717 **
aromatic_compound_degradation	−0.082	0.297	−0.598*	−0.520	0.105	0.467	−0.134	0.704 *	0.127	−0.426	0.088	−0.164	−0.745 **	−0.729 **

).

4. Discussion

4.1. Effects of Forest Fire Severity on the Diversity of Soil Bacterial Communities

The results of this study indicate that the occurrence of forest fires alters soil bacterial community diversity, with higher fire severity leading to reduced bacterial diversity. Zheng et al. [14] found in their study in the Greater Khingan Mountains in China that low-intensity fires could enhance microbial diversity in a forest of Pinus sibirica, but this diversity decreased with increasing fire severity, consistent with the findings of this study. This suggests that a certain level of fire disturbance is beneficial for maintaining soil microbial diversity. Increased fire severity implies more severe soil burning, which may volatilize more gaseous nitrogen and organic carbon, increase readily available phosphorus content, and subsequently reduce microbial diversity [20]. Severe fires can significantly increase available phosphorus content and decrease soil organic matter content and C/N ratio in the short term [23]. These findings support the analysis results of soil bacterial diversity and soil physicochemical indicators in this study.

A lack of carbon sources can lead to reduced bacterial diversity [24]. In this study, the C/N ratio emerged as a key factor influencing soil bacterial diversity after a fire. Soil microorganisms, as living organisms, are more sensitive to high temperatures from fires compared to the physical processes of carbonization, volatilization, and combustion of organic carbon [25]. Therefore, studying soil microorganisms after fire disturbance should consider two phases. The first phase is the immediate effect after the fire, where microbial loss occurs due to differences in temperature persistence caused by fire severity. The second phase involves the reoccupation of the ecological niches by microorganisms. During the early colonization phase of soil microorganisms post-fire, the C/N ratio is a critical factor for bacterial colonization diversity. A better soil carbon quality supports a greater variety of soil microorganisms that colonize and proliferate first. Fire promotes soil organic matter mineralization [26], and higher-intensity fires tend to accumulate more recalcitrant black carbon [27], leading to a reduction in the content and proportion of easily decomposable carbon. Soil microorganisms drive the transformation of soil organic carbon [26], and their survival and reproduction initially rely on soil carbon, particularly available carbon resources. The higher the proportion of these resources, the more microbial taxa are attracted and retained. The close relationship between microbial diversity and soil carbon sources indicates that microbial diversity depends on soil carbon quality, and the significant positive correlation between C/N and bacterial diversity supports this view.

While our findings demonstrate fire-induced shifts in bacterial diversity at 3 months post-fire, longer-term recovery patterns remain an important consideration. Previous studies in Mediterranean pine forests [28] and Northeast China’s taiga ecosystems [29] have observed partial microbial community recovery within 2–3 years, though trajectory differences persist between low/high intensity burns. This suggests that while acute fire dominate short-term responses (as captured in our study), legacy effects of burn severity may mediate long-term succession patterns—a critical dimension requiring multi-year monitoring in alpine forests.

4.2. Effects of Forest Fire Severity on Soil Bacterial Community Structure and Composition

Forest fires significantly alter the structure of soil bacterial communities. Forest fires lead to direct temperature increases and indirect changes in soil properties, which affect soil microorganisms [30]. If community structure is understood as changes in community evenness, higher fire severity suppresses the growth and metabolism of certain microbial populations, potentially leading to microbial death. This may involve three interacting pathways: (1) Direct thermal effects, where elevated temperatures compromise protein integrity, particularly in catalase-dependent Proteobacteria; (2) Resource reallocation dynamics, as declining DOC availability selects against oligotrophic Acidobacteriota, while Actinobacteria thrive through phosphorus acquisition traits; (3) Habitat restructuring via ash deposition, potentially initiating long-term niche differentiation. These hypotheses require validation through future thermal-gradient experiments and isotopic tracing of carbon flux. This added environmental disturbance resulted in decreased community evenness of the soil microbial community [20]. Cui et al. [31] reported that different fire intensities in the Greater Khingan Mountains led to distinct soil bacterial community structures and compositions in the forests of Pinus sibirica. Wang et al. [25] found that fire severity, post-fire recovery time, and their interactions significantly affect the soil microbial community structure in Siberian larch forests. These findings indicate that both the occurrence and intensity of forest fires can alter soil microbial community structure. Under fire disturbance, the dominant phyla in soil bacterial communities are Actinobacteriota, Proteobacteria, and Acidobacteriota, which is similar to the findings of Li et al. [16] in their study of Pinus tabuliformis secondary forest fire sites.

Forest fires alter soil microbial community structure, with greater changes observed at higher fire intensities [32]. This is primarily dependent on the quantity and quality of soil organic carbon. If dissolved organic carbon (DOC) is used to represent the organic carbon available to soil microorganisms, it is found that DOC decreases with increasing fire severity. This is mainly due to the varying degrees of volatilization and loss of organic carbon influenced by temperature persistence. The C/N and C/P ratios can represent the quality of organic carbon, and these are significantly correlated with bacterial community structure. This indicates that the quantity and quality of soil organic carbon affect the soil bacterial community structure following fire disturbance. Some studies suggest that soil pH after fire disturbance continues to drive soil community structure as in natural ecosystems [31]; however, this study did not find such an effect. A possible explanation is that this study captured bacterial communities during the acute post-fire phase (three months), whereas comparative studies analyzed longer recovery intervals (>1 year). In the immediate post-fire period, the fire disturbance regime (e.g., heat pulse and ash deposition) remains dominant over secondary succession processes. During this acute phase, bacterial communities likely exhibit heightened phosphorus utilization efficiency through rapid assimilation of newly mineralized phosphorus [28]. This aligns with findings in northeastern China, where post-fire phosphorus availability explained 65.7% of the structural variance in phosphorus-metabolizing bacterial communities [29]. Similarly, research in Yunnan’s coniferous forests identified fire-mediated phosphorus dynamics (particularly available P flux) as the primary determinant of bacterial assemblage restructuring [33], consistent with our observed stoichiometric shifts in C/P ratios. These findings support the impact of AP on microbial community structure. While phosphorus losses through volatilization and leaching are minimal, the burning of vegetation and litter caused by forest fires significantly affects the biogeochemical cycling of phosphorus. Forest fires lead to the enrichment of [28], and the retention time of AP in the soil varies depending on forest tree species [29]. It is evident that fire disturbance disrupts the original ecological stoichiometric balance, alters the carbon, nitrogen, and phosphorus pools in forest soils, affects soil biochemical properties [9], and drives the redistribution of nutrients such as carbon, nitrogen, and phosphorus both above and below ground, subsequently influencing nutrient cycling and bacterial communities.

4.3. Impact of Forest Fire Severity on Soil Bacterial Community Functions

The impact of fire severity on soil bacterial functions manifests through functional guild restructuring rather than overall diversity shifts. While prior studies reported altered functional diversity in post-fire microbial communities (e.g., carbon-utilization diversity reduction [14] or broad functional diversity decline [34]), our analysis revealed no statistically significant changes in global functional diversity indices (Shannon-Func: p = 0.17; Figure 3a). Instead, we observed functional redundancy redistribution: chemoheterotrophic bacteria (e.g., Actinobacteria) increased by 37% in severely burned sites (p < 0.01), whereas nitrogen-fixing taxa (e.g., Bradyrhizobium) decreased by 42% (p = 0.032). This aligns with fire-driven resource pulse dynamics observed in Siberian larch systems [32], where chemoheterotroph dominance correlates with post-fire organic matter mineralization. The apparent contradiction with earlier studies likely stems from methodological focus: whole-community functional gene surveys (as in [9,30]) may capture broader diversity metrics, whereas our 16S-based inference targets guild-level abundance shifts under short-term stoichiometric constraints.

Different functional bacteria exhibit varying sensitivities to heat-induced mortality from forest fires. Nitrifying bacteria are more sensitive than heterotrophic bacteria, with approximately 99% of nitrifying bacteria dying at 90 °C in dry soil and 80 °C in wet soil [35]. Nitrifying bacteria are most active at a soil temperature of 30 °C, and their activity decreases when temperatures exceed 40 °C [24]. The degree of soil temperature disturbance varies with fire severity; higher fire severity may lead to greater inactivation of soil nitrifying bacteria, resulting in reduced nitrifying bacteria abundance in the short term. Simultaneously, fires introduce nitrogen into the soil, which may enhance denitrification processes, as indicated by the increased abundance of nitrate-reducing bacteria, which is jointly driven by post-fire nitrate availability and hypoxic microenvironments. The former provides metabolic substrates [36], while the latter selects for adaptive microbial communities through redox-conditioned niche restructuring [36]. However, overall, the abundance of nitrifying bacteria is much higher than that of denitrifying bacteria, so as fire severity increases, nitrification tends to decrease. The trend of decreasing nitrogen-fixing functional genes with increasing fire severity is consistent with recent findings in the Greater Khingan Mountains of Siberian larch forests [25]. The possible reason is that the high temperatures generated by fires drastically reduce nitrogen-fixing plants and cause extensive mortality of nitrogen-fixing microorganisms, thereby lowering nitrogen-fixing capacity.

Dissolved organic carbon (DOC) and C/N ratios are positively correlated with the abundance of nitrogen-fixing functional bacteria but negatively correlated with the abundance of nitrate-reducing bacteria. Conversely, AP is negatively correlated with the abundance of nitrogen-fixing functional bacteria but positively correlated with the abundance of nitrate-reducing bacteria. The quantity and quality of soil-available organic carbon affect nitrogen fixation. Heterotrophic nitrogen-fixing bacteria require approximately 10,000 kg of organic carbon to fix 100 kg/hm² of nitrogen. A higher C/N ratio facilitates nitrogen fixation, while a lower ratio hinders it [37]. This explains the significant correlation between the abundance of nitrogen-fixing functional bacteria and DOC and C/N. If nitrogen fixation and nitrate reduction are viewed as opposing processes, the observed significant negative correlation between DOC, C/N, and nitrate-reducing bacteria can also be explained. Fires induce the production of extracellular phosphatases, increasing net phosphorus mineralization rates, which further enhances nutrient turnover efficiency and promotes soil microbial growth and reproduction [38]. This may be the main reason for the association between most functional bacteria and AP, but further analysis and research are needed for a more detailed understanding.

Our research revealed that short-term after-fire (three months post-fire) elevates phosphorus availability, driving shifts in bacterial community structure through enhanced phosphorus-processing activity. While this study established fire-mediated phosphorus dynamics as a key driver, it could not resolve whether these shifts persist beyond the acute phase or how fire severity modulates phosphorus–bacteria interactions across soil types. Future studies should extend monitoring to multi-year recovery stages and combine metagenomic profiling of phosphorus-cycling genes (e.g., phoD, ppx) with controlled burn experiments under varying soil C/P ratios to dissect these causal relationships.

5. Conclusions

Our findings demonstrate that fire-induced reductions in dissolved organic carbon (DOC) and C/N ratio, coupled with increased available phosphorus (AP), primarily regulate bacterial functional guilds in post-fire soils. The study revealed significant fire severity-dependent declines in overall bacterial diversity, with Shannon indices showing marked reductions under high-intensity burns. Notably, fire disturbance triggered functional guild restructuring characterized by substantial suppression of nitrogen-fixing bacteria alongside increased nitrate-reducing bacteria abundance. Furthermore, we observed fire intensity-dependent inhibition of nitrifying communities, likely attributable to the thermal sensitivity of ammonia-oxidizing microorganisms. Specific analysis showed pronounced decreases with fire severity while nitrate-reducing bacteria displayed proportional increases corresponding to burn intensity. These microbial responses were mediated through AP-driven phosphorus metabolism and DOC-limited symbiosis strategies. Future studies could incorporate multi-year time scales to track post-fire recovery processes, clarifying the long-term dynamics of soil bacterial community succession and functional restoration. Additionally, integrating metagenomics technologies would enable in-depth analysis of response mechanisms in key functional genes and metabolic pathways to fire disturbances, thereby comprehensively revealing the molecular regulatory networks underlying wildfire ecological effects.