The study was established in a greenhouse vegetable field at the Dahe experimental station, in Hebei Province, China (38°08′N, 114°23′E) in 2009, with winter-spring cucumber (Cucumis sativus Lcv. Bomei No. 11) and autumn-winter tomato (Lycopersicum esculentum Mill. cv. Jinpeng No. 11) rotation system. The experiment included four treatments: (i) chemical fertilizer application alone (CF), (ii) 2/4 fertilizer N+2/4 organic fertilizer N (CM), (iii) 2/4 fertilizer N+1/4 organic fertilizer N+1/4 straw N (CMS), and (iv) 2/4 fertilizer N+2/4 straw N (CS). The total amounts of nutrient (N, P₂O5, and K₂O) applied to each treatment was equal. For more information on field location experiments, see Rong et al. (2018). The specific N and C inputs in each fertilization treatment are shown in Supplementary Table S1 .

Soil samples (0–20 cm) were randomly collected from ten positions in each plot in June 2021 (the 24^th cultivation season). These samples were placed on ice and transferred to the laboratory. Soil samples were sieved through a 2 mm mesh after the removal of stones and plant residues. A part of soil samples was used for incubation experiments. The other parts were used to determine the bioinformatic analyses and basic physicochemical parameters.

Two factors were designed for the incubation experiments, i.e. incubation temperature (15, 25, and 35°C) and fertilization treatment (CF, CM, CMS, and CS). The soil samples (equivalent to 20 g dry soil) and 0.1 g ¹³C-labeled maize straw (94.9 atom% ¹³C) were thoroughly mixed and transferred to 100 mL unsealed glass bottles. Simultaneously, a soil sample (CK) with ¹²C-labeled maize straw was also prepared. The treatments were set up in 36 replicates (3 incubation temperatures × 3 plot samples × 4 incubation periods). Dark aerobic incubation experiments were carried out in artificial climate chambers at 15, 25, and 35°C with soil moisture content maintained at 75% of field capacity for 60 days. Our previous research found that the 7 days of straw decomposition is a period of rapid decomposition, and straw decomposition is basically over after 30 days (Ma et al., 2024). Therefore, we analyzed soil samples on the 7^th and 30^th day of the incubation period.

Soil DNA was isolated from 0.50 g of fresh soil using FastDNA SPIN Kit DNA on a Fast Prep-24 Homogenisation System (MP Biomedicals, Irvine, CA, United States) as described by the manufacturer. DNA concentration was determined by Nanodrop spectrophotometer (Nanodrop, Peq Lab, Germany).

Soil DNA extracted from the addition of ³C-straw and ^{1 12}C-straw treatments on the 7^th and 30^th day of incubation was stratified by ultra-highspeed centrifugation (Guo et al., 2022). For each sample, we mixed soil DNA (3 μg), CsCl (1.85 g ml^-1), and a gradient buffer (1 mM EDTA, 0.1 M KCl, 0.1 M Tris-HCl, pH = 8.0) to achieve a final mixture density of 1.725 g ml^-1. The solution was centrifuged for 44 h (190,000 × g) at 20°C by a Vti65.2 vertical rotor (Beckman Coulter Inc., Palo Alto, CA, United States) and Quick-Seal polyallomer ultracentrifugation tubes (5.1 ml). After centrifugation, the solution in the centrifuge tube was divided into 15 equal portions (380 μl each) using a syringe pump. The density of each buoyant was determined by a digital hand-held refractometer. We purified the fractionated DNA samples and then re-eluted them with 30 μl of sterilized ultrapure water for subsequent analyses.

The 16S rRNA genes of each soil microcosm system were analyzed by qPCR in triplicate using the iCycler system (Bio-Rad). The reaction system (20 μl) included 2.0 μl of DNA, 4 μM of each primer, 10 μl qPCR Master Mix (Vazyme Biotech Co., Ltd.). Primers 515F (5’-GTGCCAGCMGCCGCGGTAA-3’) and 806R (5’-GGACTACHVGGGTWTCTAAT-3’) (Zhang et al., 2022).

To reveal taxonomic characteristics (ecological attributes), we identified bacterial communities involved in straw decomposition on Illumina 16S rRNA gene sequencing by Novogene Co., Tianjin, China. A total of 228 (72 from the ¹³C treatments after gradient fractionation, 72 from the ¹²C treatment after gradient fractionation, 72 from the ¹³C treatments before gradient fractionation, and 12 from before the start of the incubation experiment) DNA composite samples were selected for amplicon sequencing by targeting the V4 region of the 16S rRNA gene with the primer sets 515F (5’-GTGCCAGCMGCCGCGGTAA-3’) and 806R (5’-GGACTACHVGGGTWTCTAAT-3’). Sequences analysis was conducted using Uparse software (Edgar, 2013). Sequences with ≥97% similarity were grouped into the same OTU and representative sequences from each OTU were screened for further annotation. For each representative sequence, the Silva Database (https://www.arb-silva.de/)was used based on Mothur algorithm to annotate taxonomic information (Quast et al., 2012). All raw reads were archived in the NCBI Sequence Read Archive database (accession number: PRJNA1089403).

To reveal bacterial functional profiling (physiological attributes), DNA-SIP-based shotgun metagenomic sequencing was conducted on Illumina NovaSeq/Hiseq Xten at Majorbio Bio-Pharm Technology Co., Ltd. (Shanghai, China). A total of 36 samples were selected from the ¹³C treatments DNA composite samples after gradient fractionation for library construction and shotgun metagenomic sequencing. Extracted DNA was fragmented at an average of 400 bp (Genetics Co. Ltd., China) to construct paired-end libraries using Covaris M220. Extracted DNA was fragmented at an averagely of about 400 bp through Covaris M220 (Gene Co., Ltd., China) to construct paired-end libraries. Data were analyzed on the Majorbio Cloud Platform (www.majorbio.com) free online platform. Representative sequences from the non-redundant gene catalogue were aligned to the NR database using Diamond for taxonomic annotation (Buchfink et al., 2015). Carbohydrate-active enzymes (CAZy) annotation was conducted using hmmscan (http://hmmer.janelia.org/search/hmmscan) against the CAZy database. About 10 Gbp of Illumina data were obtained for each sample. All raw reads were archived in the NCBI Sequence Read Archive database (accession number: PRJNA1089381).

Soil nitrate-N (NO₃ ⁻-N) was extracted by 2 M KCl and measured using a flow injection autoanalyzer (Smartchem 200, Alliance, Paris, France) (Norman et al., 1985). Soil total nitrogen (TN) and organic carbon (SOC) were determined through an elemental analyzer (Elementar Analysensy steme GmbH, Hanau, Germany). Soil pH was determined using a pH meter (Mettler Toledo, Switzerland) with a water/soil ratio of 2.5:1. Soil available phosphorus (P) and available potassium (K) were measured by the Olsen method and the flame photometry method, respectively (Olsen et al., 1954; Helmke and Sparks, 1996).

The Kruskal–Wallis H test was performed with STAMP statistics (https://beikolab.cs.dal.ca/software/STAMP) to compare the abundance of OTU -indicators in the ¹³C-treatment samples with those in the ¹²C-treatment samples.

One-way analysis of variance (ANOVA) and tests of multiple comparisons across treatments (Duncan’s post hoc test, P <0.05) were performed on the measurements under different treatments using IBM SPSS statistical software (SPSS, Inc., Chicago, IL, USA). Principal coordinate analysis (PCoA) was performed based on the Bray–Curtis dissimilarity of bacterial communities using the “vegan” packages of R. Linear discriminant analysis (LDA) effect size analyses were performed using the LEfSe tool (https://bioincloud.tech/standalone-task-ui/lefse). Correlation analyses between species abundance and functional abundance based on the relative species and functional abundance of samples to identify the functional contribution of specific species were analyzed on the platform (www.i-sanger.com) provided by Majorbio Co., Ltd. (Shanghai, China).

As shown in Figure 1A , the dominant bacterial phylum under different fertilization treatments was Proteobacteria, which comprised 25.7% of the total sequences on average, followed by Firmicutes, Actinobacteria, and Acidobacteria, which represented 14.8%, 9.7%, and 5.4% of the sequences, respectively. Compared to the chemical fertilizer application alone (CF), organic-amended treatments (CM, CMS, and CS) increased the relative abundance of Firmicutes, Bacteroidetes, Acidobacteria, Gemmatimonadetes, and Crenarchaeota by an average of 4.1%, 43.9%, 23.2%, 9.1%, and 48.6%, respectively. In contrast, the relative abundance of Proteobacteria and Actinobacteria were lower in organic-amended treatments than in CF treatment.

Principal coordinates analysis (PCoA; Figure 1B ) of soil bacteria at the OTU level for the 12 samples shows that the PCo1 and PCo2 explained 30.3% and 14.9% of the total variance in bacterial community profiles, respectively. Meanwhile, the soil samples were distinctly separated into three groups, i) CF, ii) CM, and iii) CMS and CS. Compared to CF treatment, organic-amended treatments, especially straw-amended treatments (CMS and CS), exhibited higher bacterial Alpha diversity (as indicated by the higher values of Shannon, Chao1, and Ace; Table 1 ).

As shown in Figure 2A , there are five dominant bacterial phyla (relative abundance > 1.0%) under different incubation temperatures and fertilization treatments. Proteobacteria had the highest relative abundance (22.6%-44.8%), followed by Actinobacteria (13.0%-25.2%), Firmicutes (13.1%-18.3%), Bacteroidetes (1.9%-6.87%), and Acidobacteria (1.3%-4.3%).

Principal coordinates analysis (PCoA) as conducted on soil bacteria from 36 samples on the 7^th and 30^th days of straw decomposition at the OTU level ( Figures 2B, C ). The results showed that PCo1 and PCo2 explained 22.6% and 14.9% of the total variance in bacterial community profiles on the 7^th day, respectively, and 35.2% and 16.6% of the total variance on the 30^th day, respectively. On the 7^th day of incubation, the bacterial community structure at different temperatures can be divided into three groups: i) 15°C, ii) 25°C, and iii) 35°C; meanwhile, the bacterial community structure under different fertilization treatments can also be divided into three groups: i) CF, ii) CM, and iii) CMS and CS. On the 30^th day of incubation, the bacterial community structure at different temperatures can be divided into two groups: i) 15°C and ii) 25°C and 35°C; additionally, the fertilization treatments have no significant impact on the composition of bacterial community structure.

The Chao1 and Ace in the soil bacterial Alpha diversity were significantly higher at 15°C than at 25°C and 35°C on the 7^th day, and incubation temperature did not significantly affect soil bacterial Alpha diversity on the 30^th day. As compared to CF treatment, organic-amended treatments, especially straw-amended treatments, exhibited higher bacterial Alpha diversity (as indicated by the higher values of Shannon, Chao1, and Ace; Table 2 ).

On the 7^th day of incubation, 67-538 bacterial OTUs exist in the ¹³C-heavy DNA under different incubation temperatures and fertilization treatments. These bacterial OTUs were dominated by Proteobacteria (30.1%-64.4%) and Actinobacteria (16.2%-67.7%), followed by Bacteroidetes (0.1%-10.2%), Firmicutes (0.0%~14.3%), Chlorobacteria (0.1%~2.3%), Verrucomycota (0.0%~1.5%), and Acidobacteria (0.0%~0.9%). On the 30^th day of incubation, there are 106-470 bacterial OTUs in the ¹³C-heavy DNA under different incubation temperatures and fertilization treatments. These bacterial OTUs were dominated by Proteobacteria (25.4%-63.7%) and Actinobacteria (2.5%-64.9%), followed by Firmicutes (0.1%-20.6%), Bacteroidetes (0.1%-12.9%), Chloroflexi (0.0%-2.3%), Verrucomycota (0.0%-0.7%), and Acidobacteria (0.0%-1.4%).

As shown in Figure 3A , the relative abundance of soil Verrucomicrobia, Chloroflexi, and Acidobacteria involved in straw decomposition tended to increase with increasing temperature on the 7^th day, and the relative abundance of Actinomycetes showed a decreasing trend. The organic-amended treatments, especially straw-amended treatments, increased the relative abundance of soil Proteobacteria, Firmicutes, and Acidobacteria, and decreased the relative abundance of Actinobacteria compared to CF treatment. The relative abundance of soil Chlorobacteria and Acidobacteria involved in straw decomposition tended to increase with increasing temperature on the 30^th day, whereas the relative abundance of Verrucomicrobia and Firmicutes tended to decrease. Compared to the CF treatment, organic-amended treatments, especially straw-amended treatments, increased the relative abundance of soil Actinobacteria, and decreased the relative abundance of Chloroflexi.

We identified specific phylotypes responding to different incubation temperatures and fertilization treatments at different sampling times by using LEfSe ( Figures 3B–E ). On the 7^th day of incubation, Alphaproteobacteria and Actinomycetia were the common and most predominant class involved in straw decomposition under different incubation temperatures and fertilization treatments. Gammaproteobacteria (Pseudomonas, Steroidobacter, Acidibacter, and Arenimonas) were the unique and predominant class involved in straw decomposition at medium and high temperatures (25°C and 35°C) as well as in the straw-amended treatments (CMS and CS). On the 30^th day of incubation, Actinomycetia was the common and most predominant class involved in straw decomposition under different incubation temperatures, and Alphaproteobacteria was the common and most predominant class involved in straw decomposition under different fertilization treatments.

According to the redundancy analysis (RDA) of the bacterial community structure constrained by soil physico-chemical properties, OC (P=0.001), TN (P=0.001), and AP (P=0.001) are the main environmental factors affecting the bacterial community structure on the 7^th day of incubation ( Figure 4A ). Meanwhile, OC (P=0.001), TN (P=0.001), and C/N (P=0.001) are the main environmental factors affecting the bacterial community structure on the 30^th day of incubation ( Figure 4B ). In sum, the bacterial community structure involved in straw decomposition has a stronger response to OC and TN during each incubation period.

As shown in Figure 5A , the relative abundance of glycosyl transferases (GT), carbohydrate-binding modules (CBM), and polysaccharide lyases (PL) enzyme genes involved in straw decomposition tended to increase with increasing temperature on the 7^th day of incubation, whereas the relative abundance of glycoside hydrolases (GH) and carbohydrate esterases (CE) enzyme genes tended to decrease. The organic-amended treatments, especially straw-amended treatments, increased the relative abundance of GT and auxiliary activities (AA) enzyme genes and decreased the relative abundance of GH, CE, and PL enzyme genes. Significant markers of CAZymes involved in straw decomposition were GH and PL in CF treatment, and GT and AA in CMS treatment ( Figure 5B ). Significant markers of CAZymes involved in straw decomposition were CE at 15°C, and CBM at 35°C ( Figure 5C ).

As shown in Figure 6 , the dominant bacteria with a higher proportion have a higher contribution to CAZymes under different incubation temperatures and fertilization treatments. The sum of the relative contributions of dominant bacteria (Actinobacteria, Alphaproteobacteria, and Gammaproteobacteria) to GH, GT, CE, AA, PL, and CBM is averagely 75.9%, 72.6%, 78.2%, 79.1%, 70.7%, and 67.1%, respectively.

Fertilization can affect soil microbial proliferation and functional metabolism by altering soil nutrient conditions and physical properties (Mi et al., 2016; Wei et al., 2017). Researches-based on grain crops (corn, wheat, and rice) and vegetables (open fields and facilities) have shown that long-term organic fertilization (i.e., crop straw and manure) were beneficial for increasing microbial diversity (Ai et al., 2018; Luan et al., 2020; Jin et al., 2022; Wang et al., 2023). Our results also indicated that organic-amended treatments increase bacterial diversity as indicated by Shannon, Chao1, and Ace. Moreover, we found that the CMS treatment (i.e., high C diversity), rather than CS treatment (i.e., high C quantity), were more conducive to enriching bacterial diversity. The likely reason for these findings was that organic fertilizer or straw is rich in carbon and nitrogen resources, which not only provides “food” for microorganisms after being put into the soil, but also improves the soil structure, thus providing a more suitable “micro-environment” for microbial growth and reproduction (Mi et al., 2016; Wei et al., 2017). Organic fertilizer or straw application shifts the structural composition of soil microbial communities towards fast-growing communities (symbiotic nutrient groups), increases the diversity and sustainability of soil microbial communities, and facilitates the formation of soil microbial taxa associated with healthy crop growth (Reilly et al., 2013; Gonthier et al., 2014; Lupatini et al., 2017). There is an equilibrium point between soil organic carbon input and mineralization, when the equilibrium point is reached, the soil organic carbon will not increase and reach the threshold, the soil microbial growth and reproduction also exists in this law (Zhang et al., 2020). Therefore, in our study, the diversity of the soil microbial communities did not increase with an increase in carbon inputs. Combining manure with straw makes it easier to form a nutrient-rich environment, which can provide more effective nutrients and energy substances for the more active symbiotic flora. In our study, the soil microbial Alpha diversity index of the CMS treatment was the highest and had the strongest potential to withstand external environmental changes (Delgado-Baquerizo et al., 2016; Bei et al., 2018).

Recently, several studies revealed that adding straw reduced the microbial diversity (Sanaullah et al., 2016; Zhao et al., 2019; Ye et al., 2022). This finding was confirmed by our study ( Tables 1 , 2 ) and indicated that symbiotic populations stimulate a subset of the entire community, whereas other subsets may not respond to straw, leading to a decrease in microbial community diversity (Tardy et al., 2015; Sanaullah et al., 2016). Meanwhile, we found that the response of soil microbial diversity to different fertilization treatments was consistent with that before straw addition, i.e., soil microbial diversity was higher in the organic-amended treatments than in the chemical fertilizer application alone. Previous research on the microbial diversity during straw decomposition at different temperatures has shown that the diversities of the bacterial community have no significant response to temperature during straw decomposition (Wahdan et al., 2023). Another study reveals that the microbial diversity index is higher under low- and medium-temperature treatments than under high-temperature treatments in the early stage of straw decomposition; meanwhile, the bacterial community diversity index was the highest under medium-temperature treatment in the later stage of straw decomposition (Zhou et al., 2016). Our research results indicate that in the early stage of straw decomposition, the diversity of soil bacterial communities decreases with increasing temperature. In the later stage of straw decomposition, there was no significant difference in the diversity index of soil bacterial communities at different incubation temperatures, with the highest diversity index at 25°C. In the early stage of straw decomposition, the soil bacterial community has a high reactivity to climate change (temperature), which explains our results (Glassman et al., 2018).

Long-term organic application can improve soil microbial community structure, enhance microbial activity, and thereby affect straw decomposition (Fang et al., 2018). The results of previous studies showed that the soil microorganisms of straw decomposition were dominated by soil Proteobacteria, Actinobacteria, Firmicutes, and Bacteroidetes in the soil of grain field crops (corn, wheat, and rice) through DNA-SIP (Fan et al., 2019; Kong et al., 2020; Guo et al., 2022; Zhang et al., 2022). Meanwhile, there are variable effects of organic-amended treatments on different dominant phyla in straw decomposition. Guo et al. (2022) indicated that soil Bacteroidetes (Flavobacterium) was an indicator bacterial group for the straw-amended treatments, whereas Chloroflexi (Herpetosiphon) and Gammaproteobacterial (Psudoxanthomonas) were indicator bacterial groups for the chemical fertilizer application alone. Partial results of our study support the above research that microorganisms dominating straw decomposition mainly contain Proteobacteria, Actinobacteria, Firmicutes, and Bacteroidetes. Meanwhile, we found that Alphaproteobacteria and Actinomycetia were the common and most predominant class involved in straw decomposition under different fertilization treatments during the period of rapid straw decomposition, and Gammaproteobacteria (Pseudomonas, Steroidobacter, Acidibacter, and Arenimonas) were the unique and predominant class involved in straw decomposition in the straw-amended treatments (CMS and CS). When resource conditions are favorable, eutrophic groups (Proteobacteria, Actinobacteria, Firmicutes, and Bacteroidetes) have higher nutrient requirements, preferentially consume unstable soil organic carbon pools, show higher growth rates, and have an increased proportion in the soil (Fazi et al., 2005). Our result confirmed the first hypothesis. Some soil bacterial phyla show a tendency to be symbiotic and have high abundance in soils with high organic carbon, Proteobacteria and Bacteroidetes usually show co-nutritional properties in combination with organic carbon-rich substrates; Actinobacteria possess several genes related to carbon degradation in plant residues, namely endoglucanases, β-glucosidases, α-glucosidases, and α-mannosidases, as well as binding proteins involved in sugar translocation (Fierer et al., 2007; Yu et al., 2020);. Therefore, soil Proteobacteria, Actinobacteria, Firmicutes, and Bacteroidetes dominated straw decomposition in our study. Crop straw is rich in cellulose, which is the most abundant renewable natural organic matter on earth, and Pseudomonas play a key role in cellulose decomposition (Kumar et al., 2015). The abundance of Pseudomonas in straw-amended treatments was significantly higher than that in the other treatments, which made the treatments accelerate straw decomposition.

Temperature accelerates straw decomposition by changing the composition of the soil microbial community, which in turn affects extracellular enzyme activities related to straw decomposition (Feng and Simpson, 2009; Yergeau et al., 2012; Stark et al., 2015). Wahdan et al. (2023) indicated that the main dominant phylum of soil bacterial community in the straw decomposition were Proteobacteria, Actinobacteria, and Bacteroidetes, and the abundance of Enterobacteriaceae and Pseudomonas showed an increasing trend with the increase of temperature. Previous studies on lignin-degrading microorganisms in tropical forest (high-temperature) soils have indicated that soil Alphaproteobacteria and Gammaproteobacteria are the main active flora (Olson et al., 2007; Jiang et al., 2015). In this study, we used DNA-SIP technology to investigate the changes in soil microbial communities during straw decomposition at different temperatures, which can accurately identify the microbial groups utilizing straw-C. Our results showed that the dominant phyla utilizing straw-C at different temperatures were Proteobacteria and Actinobacteria in the early stage of straw decomposition, and Proteobacteria, Actinobacteria, and Firmicutes in the later stage. Meanwhile, the Gammaproteobacteria (Acidibacter and Lysobacter) were the unique and predominant class involved in straw decomposition at medium and high temperatures (25°C and 35°C). Bao et al. (2021) found that Actinobacteria play important ecophysiological roles at various stages of straw decomposition and maintain functional composition during the decomposition process. Previous studies pointed out that Gammaproteobacteria is the relevant group of bacteria involved in cellulose degradation (assimilation of straw-C) and the main carbon-utilizing bacteria (Bernard et al., 2007; Eichorst and Kuske, 2012; Rime et al., 2016);. This shows that Gammaproteobacteria is one of the reasons for the fast rate of straw decomposition under medium- and high-temperature conditions. In addition, we found that the incubation temperature had a higher effect on soil microbial community composition than the fertilization treatments in our study ( Figures 2B, C ).

Plant residue decomposition processes and microbial communities are closely influenced by the functional characteristics of microorganisms (Wegner and Liesack, 2016; Guo et al., 2022; Zhang et al., 2022; Bao et al., 2023). Previous studies of DNA-SIP experiments have confirmed that soil bacteria contain large amounts of ¹³C labels not only as a result of cross-feeding in the experiments, but also due to their secretion of hydrolytic enzymes for actual degradation of plant residues (López-Mondéjar et al., 2018; Wilhelm et al., 2019; López-Mondéjar et al., 2020). Bao et al. (2023) found that the carbohydrate metabolism of bacteria decreased significantly with increasing soil fertility. López-Mondéjar et al. (2020) indicated that soil bacteria contain an abundance of genes encoding catabolic cellulases and hemicellulases, which play a key role in plant residue degradation, and that most of the CAZymes families have broad substrate specialization through DNA-SIP with shotgun metagenomic sequencing analysis. Our results partially support the above conclusion that the organic-amended treatments increased the relative abundance of GT and AA enzyme genes during the pre-straw decomposition period, and the higher contributions to CAZymes genes were from Actinomycetes, Alphaproteobacteria, and Gammaproteobacteria. Our result confirmed the second hypothesis. High-fertility soils were selected for fast-growing but inefficient bacteria, whereas low-fertility soils were selected for slow but efficient bacteria to degrade plant residues, and nutrient status influences the ecological strategies and metabolic trade-offs of bacteria due to the cytological economics of energy partitioning between growth rate and substrate utilization efficiency (Roller et al., 2016; Niederdorfer et al., 2017). Actinomycetes are highly competitive for carbon sources and adaptable to their environment, so they have the highest contribution to CAZymes genes (Swarnalakshmi et al., 2016; van Bergeijk et al., 2020). The presence of CBM enzyme genes in bacterial laccases permits direct binding of bacterial cells to the target polysaccharide, increasing the potency of the laccases and decreasing competitive rejection for cellulose degradation (Donohoe and Resch, 2015). In our study, the unique and predominant class of CAZymes at high temperature was the CBM, which may be one of the reasons for the fast rate of straw decomposition under high temperatures.