The Aggregate-Mediated Restoration of Degraded Black Soil via Biochar and Straw Additions: Emphasizing Microbial Community Interactions and Functions

Abstract

The synergistic application of biochar and straw could improve soil properties and influence soil microbial community. However, its impacts on microbial community interactions and functions within various aggregate fractions remain unclear. We conducted a three-year field trial in black soil in northeastern China, under the restoration measures of biochar application (BR, 30 t ha⁻¹ once), straw return (SR, 5 t ha⁻¹ year⁻¹), and the combination of BR and SR (BS, BR at 30 t ha⁻¹ once and SR at 5 t ha⁻¹ year⁻¹). Utilizing high-throughput sequencing, we assessed the influence of different straw-returning methods on the structure and function of microbial communities in the mega-aggregates (ME, >2 mm), macroaggregates (MA, 0.25–2 mm), and microaggregates (MI, <0.25 mm). Relative to the control (CK), the BR, SR and BS treatments significantly decreased the bacterial Shannon index, mainly dependent on ME (p < 0.05). Conversely, compared with the CK and SR treatments, both BR and BS treatments notably reduced the fungal Shannon index, largely influenced by MI (p < 0.05). Moreover, the BS treatment significantly increased the relative abundance (RA) of Mortierellomycota (p < 0.05) compared to the CK, BR and SR treatments. Meanwhile, the SR and BS treatments substantially reduced the RA of Nitrospirae (p < 0.05) in comparison to the CK and BR treatments. Furthermore, compared with the CK, the BR and SR treatments enhanced microbial network connectivity, while the BS treatment diminished it, especially in ME and MI. Concurrently, the keystone of co-occurrence networks shifted from Phycisphaeraceae, Blastocatellaceae, and Glomeraceae in the CK treatment to uncultured_bacterium_c_JG37-AG-4 and DA111 in the BS treatment. Additionally, BR and SR exhibited synergistic effects on most microbial community functions (e.g., enhanced chitinolysis and carbon fixation but reduced nitrogen-cycling functions), but they also possessed distinct differential functions. In short, the combined application of biochar and straw adversely impacted soil microbial community diversity and stability, especially in ME and MI.

1. Introduction

The degradation of agricultural land poses a serious threat to both agricultural production and food security. This issue is particularly pertinent to the black soil region of Northeast China. Following approximately seven decades of widespread overuse, the degradation of black soil in Northeast China has become increasingly severe [1]. This degradation transpires as a consequence of multiple factors including the excessive use of agricultural machinery, over-application of fertilizers, monoculture practices, and the removal of crop residues [2]. It has manifested as soil compaction, acidification, salinization, and a diminished level of organic matter content as well as microbial community diversity [3], which seriously affects regional ecological security and sustainable agricultural development.

Soil productivity can be categorized into restrictive, available, inherent, and exhausted types [2]. The term “inherent soil productivity” invokes the synergistic impact of soil organic matter (SOM), nutrients, aggregates, and microorganisms [2]. Collectively, these factors foster an optimal environment for crop growth, characterized by robust buffering capacity and stability. The fundamental approach to enhancing inherent soil fertility lies in augmenting the level of soil organic carbon (SOC). Organic material additions led to an increase in SOM content, which subsequently resulted in elevated nutrient contents and microbial activities [2,4]. This process also facilitated the formation of macroaggregates [2,5]. China, recognized as one of the world’s leading producers of crop straw, generates approximately 700 million tons annually [6]. Crop straw is a rich source of carbon (C) and nutrient elements such as nitrogen (N), phosphorus (P), and potassium (K) [7]. Its return to the soil can not only effectively mitigate soil degradation, but also reduce the air pollution caused by straw burning [8]. Moreover, biochar is a solid byproduct of high-temperature pyrolysis of natural organic materials in the absence of oxygen (O₂) [9]. As one of the crucial methods of indirect straw return (SR), biochar application (BR) increases the proportion of aromatic C [10]. In contrast, SR introduces a greater quantity of active organic C than BR [11]. Previous studies have shown that both BR and SR can enhance soil aggregate stability, improve soil pH, and increase aggregate-associated organic C and nutrient contents [8,12,13,14,15]. Despite the carbon-rich nature of both crop straw and its biochar, they may have opposite effects on soil properties [16]. This variation can be attributed to the distinct characteristics of their C sources: crop residues are abundant in active C, while biochar is predominantly composed of recalcitrant C [17].

Soil microorganisms contribute significantly to nutrient cycling and the maintenance of soil structure, and their diversity serves as a sensitive indicator reflecting subtle alterations in soil quality [18]. Alterations in soil physicochemical conditions can either beneficially or detrimentally influence the structure and functions of soil microbial communities by changing their habitats [17]. The preponderance of scientific studies concur that SR and BR enhance the efficacy of soil microbiota in terms of their activity and diversity [13,19,20,21]. However, some studies have also shown that BR reduces the richness and diversity of bacterial and fungal communities [17,22]. The influence of biochar on soil physicochemical conditions, including pH, water retention, and nutrient provision, is regarded as paramount in affecting microbial communities [17]. Given that fungi typically thrive in lower pH environments, a pH increase induced by biochar might prioritize bacterial over fungal growth, particularly in the immediate vicinity of the biochar particle following its application [11]. However, long-term or excessive application of biochar makes the soil bacterial community structure develop in a specific direction, and finally leads to the decline of its diversity and abundance [23,24]. In addition, previous studies have shown that SR can enhance the stability of soil bacterial communities [24,25]. Similarly, BR has been shown to exert a positive influence on the stability of soil bacterial and fungal communities [24,26]. Currently, limited studies revealed that the combined straw and biochar application (BS) could promote SOC sequestration and mitigate C loss, yet there are marked differences in the C sequestration between SR and BR [27,28,29]. However, there are limited studies on the effects of BS on soil microbial community structure and function.

Soil aggregates, as fundamental components of soil structure and functional units of soil ecosystems, mediate many chemical and biological processes [30]. Each aggregate represents a distinct ecological niche suitable for microbial colonization, and the distribution of aggregate sizes impacts the availability of C [31]. Organic compounds serve as the principal cementing agents, critically influencing both the formation process and stability of aggregates [32]. Generally, organic material additions alter soil aggregation, which may subsequently influence the habitats where microbes are heterogeneously distributed [33]. Besides, the formation of aggregates is influenced by the diversity of bacterial and fungal communities, and the resulting changes in soil structure in turn affect future microbial life [34,35]. The majority of soil microorganisms inhabit periodically interconnected communities that are intimately associated with soil aggregates (<2 mm) [34]. At the spatial scales pertinent to microbial biogeochemistry, soils are predominantly constituted of macroaggregates (MA, 0.25–2 mm), which limit O₂ diffusion and regulate water flow, and of microaggregates (MI, <0.25 mm), which bind SOC and protect it from erosion-induced removal [34]. Moreover, the impact of management practices on soil C is modulated by soil aggregate sizes and their associated microbial communities [36]. This influence is more pronounced in MA compared to MI [18]. In addition, BR enhanced microbial interactions, particularly in mega-aggregates (ME, >2 mm) or MA fractions [1,37]. However, the effect of BS on microbial community interactions within various aggregate fractions remains largely unclear. Concurrently, the existence of synergistic effects (e.g., similar effect or further excitation) between BR and SR on microbial community functions within soil aggregates is yet to be fully understood.

Therefore, the objective of this research was to investigate the effects of three straw-returning methods (SR, BR, and BS) on the structure and function of soil microbial communities in different aggregate fractions (ME, MA, and MI). We hypothesized that BS could promote microbial community interactions, especially in large aggregates, while BR and SR have synergistic effects on microbial community functions. This work expands our understanding of the interactions and functions of microbial communities in response to BS within soil aggregates, enhancing our knowledge of the impact of different straw-return methods on the improvement of inherent soil productivity.

2. Materials and Methods

2.1. Field Experiment Description

This field experiment was initiated in 2016 and located in the middle of Jilin Province, China (43°59′ N, 125°41′ E). The study site has a temperate continental monsoon climate. According to the USDA texture classification system, the study soil is a Mollisol. Prior to the start of this experiment, the soil properties of 0–20 cm layer were as follows: pH of 6.92; electrical conductivity (EC) of 23.6 μs cm⁻¹; soil organic C (SOC) of 7.36 g kg⁻¹; total N (TN) of 1.07 g kg⁻¹; and total P (TP) of 0.50 g kg⁻¹ [37]. The four treatments included the following: (1) CK, without biochar or straw addition, as a control; (2) BR, biochar application (30 t ha⁻¹ once); (3) SR, maize straw return (5 t ha⁻¹ year⁻¹); (4) BS, combined application of biochar (30 t ha⁻¹ once) and maize straw (5 t ha⁻¹ year⁻¹). The treatments were applied to 25 m² (5 m × 5 m) plots and arranged in a complete randomized block design with three replicates. The biochar, produced through the pyrolysis of maize straw at 450 °C in the absence of O₂, was procured from Liaoning Jinhefu Agricultural Technology Co., Ltd. (Anshan, China). The total contents of C, N, P, and K in the biochar were 72.21%, 1.08%, 0.72%, and 1.64%, respectively. Biochar was manually spread over the soil and immediately mixed into the 0–20 cm layer before maize sowing in April 2016. The total contents of C, N, P, and K in the maize straw were 60.63%, 0.92%, 0.37%, and 1.38%, respectively. Maize straw was manually spread over the soil at a rate of 5 t ha⁻¹ yearly before maize sowing, and then the maize straw was mixed into the 0–20 cm soil layer. Fertilizers were applied to the soil every year at rates of 90 kg N ha⁻¹, 100 kg P₂O₅ ha⁻¹ and 100 kg K₂O ha⁻¹ before maize sowing, after which the fertilizers were incorporated into the 0–20 cm soil layer using rotary tillage machine. Maize was sown at a density of 65,000 plants ha⁻¹ in early May each year, using a handheld hole-sowing machine. During the maize silking stage, 135 kg N ha⁻¹ was applied to all plots with the same sowing device.

2.2. Soil Sampling and Analysis

Soil samples were collected from the 0–20 cm layer at harvest in October 2018. Five sampling points were randomly selected in a S-shape distribution within each plot to form a single composite sample. The soil columns, collected parallel to the soil section, were excavated vertically between two maize plants at each sampling point [37]. These undisturbed profiles were stored in sterilized plastic containers for aggregate fractionation and then were immediately transported to the laboratory. Additionally, the study by Luo et al. [36] was referenced, in which undisturbed soil samples were divided into three aggregate fractions: ME, MA, and MI. The fractional soil samples were further divided for microbial community analysis (stored at −80 °C), ecological enzyme activity examination (stored at −20 °C), and evaluation of SOC, TN, and TP contents (air dried and sieved through 0.15 mm). SOC and TN were analyzed by the dichromate oxidation method and the Kjeldahl method, respectively [38,39]. TP was determined by digestion with HClO₄-H₂SO₄ and TK was assessed via digestion with HClO₄-HF [38,39]. The total contents of C, N, P, and K in the maize straw and biochar were determined according to the methods for measuring SOC, TN, TP, and TK. The activities of β-1,4-glucosidase (βG), leucine aminopeptidase (LAP), N-1,4-acetyl-glucosaminidase (NAG) and alkaline phosphatase (ALP) were determined using respective enzyme-linked immunosorbent assay (ELISA) kits from Jiangsu Meimian Industry Co., Ltd. (Yancheng, China), following the manufacturer’s instructions [37].

2.3. Illumina Sequencing and Bioinformatic Analysis

The bacterial 16S rRNA and fungal ITS genes were chosen for microbial community analysis via Illumina sequencing, utilizing the Illumina HiSeq 2500 Platform (Illumina Inc., San Diego, CA, USA), which was performed by Biomarker Technologies Corporation (Beijing, China). DNA was extracted from 0.5 g soil samples utilizing a NucleoSpin^® Soil DNA extraction kit (Macherey-Nagel, Duren, Germany). The concentration and quality of the extracted DNA were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA) [37]. The universal primer pairs 338F/806R [40] and ITS1/ITS2 [41] were selected for PCR amplifications of 16S rRNA (V3 + V4 region) and ITS (ITS1 region) genes, respectively. The Solexa PCR mixture (20 μL) contained 2.5 μL of 2 μMMPPI-a, 2.5 μL of 2 μMMPPI-b, 10 μL of 2 × Q5 HF MM, and 5 μL of the targeted PCR product [37]. The PCR protocol involved an initial denaturation at 98 °C for 30 s followed by 10 cycles of 10 s at 98 °C, 30 s at 65 °C, and 30 s at 72 °C, with a final extension at 72 °C for 5 min [37]. The raw sequences were merged using FLASH (version 1.2.11) and filtered with Trimmomatic (version 0.33), and chimeras were removed using UCHIME (version 8.1) [37] The clustering of operational taxonomic units (OTUs) with high-quality reads at 97% sequence similarity, and the taxonomic annotation of OTUs using taxonomy databases of Silva (Release128, http://www.arb-silva.de, accessed on 21 January 2019) and UNITE (Release 7.2, http://unite.ut.ee/index.php, accessed on 21 January 2019) for bacteria and fungi, respectively. The raw sequences were available at the NCBI Sequence Read Archive database under the BioProject accession number of PRJNA656076.

2.4. Statistical Analysis

The MicrobiomeAnalyst platform (https://www.microbiomeanalyst.ca, accessed on 26 May 2020) was employed to calculate and visualize the relative abundances (RAs) at the phylum level as well as the Shannon diversity index of α-diversity. An analysis of variance (ANOVA) was conducted to evaluate the impact of straw-returning methods and aggregate fractions on soil properties and the phyla RAs and α-diversity of bacterial and fungal communities. When a significant F-value was observed, the least significant difference (LSD) test was applied for mean comparisons using SPSS (version 20.0). In all instances, differences were considered significant if p < 0.05. Refer to Zhang et al. [37] for redundancy analysis (RDA), between-class analysis (BCA), co-inertia analysis (COIA), and co-occurrence network analysis. The co-occurrence networks were individually visualized utilizing Gephi v0.9.2, using Spearman’s rank correlation coefficients for taxonomic families with a correlation magnitude of ∣r∣> 0.9 and a significance level of p < 0.01. Topological features including nodes, links, modularity, average degree, and betweenness centrality were estimated using Gephi v0.9.2 to infer complex associations and keystone taxa. In the networks, nodes were utilized to represent individual taxa, which were colored according to their respective microbial phyla. The links within the networks served to indicate the pairwise correlations, both positive and negative, implying the presence of either cooperative or competitive interactions between nodes. The degree of modularity increased proportionally to the tightness of the connections among nodes of co-occurring or co-evolving microbiota. Average degree reflected the overall connectivity density of the network. Betweenness centrality reflected the role of nodes as hubs in the network. Network hubs in the top 5% for both hubness and betweenness centrality were identified as keystone. Line discriminant analysis effect size (LEfSe), FAPROTAX function prediction, FUNGuild function prediction, and Tax4Fun function prediction were performed in BMKCloud (https://www.biocloud.net, accessed on 19 August 2024). In this study, the effect size of LEfSe was set to LDA value ≥ 4. The results of function prediction were visualized by Statistical Analysis of Metagenomic Profiles (STAMP, version 2.1.3) based on a t-test. The error bars reflected the proportion of variation in community function within the 95% confidence interval.

3. Results

3.1. Microbial Community Diversity

In ME, relative to the CK, the BR, SR, and BS treatments significantly decreased the bacterial Shannon index (p < 0.05, Figure 1A). In MA, only the BR and BS treatments caused a significant decrease in the bacterial Shannon index compared to both the CK and SR treatments (p < 0.05, Figure 1A), while the BS treatment also notably reduced the fungal Shannon index in comparison to the CK, BR, and SR treatments (p < 0.05, Figure 1B). In MI, the bacterial Shannon index was significantly lower in the BR treatment compared to the CK, SR, and BS treatments (p < 0.05, Figure 1A), whereas the fungal Shannon index decreased significantly in both the BR and BS treatments in comparison to the CK and SR treatments (p < 0.05, Figure 1B). Overall, compared with the CK, the BR, SR and BS treatments led to a substantial reduction in the bacterial Shannon index (p < 0.05, Figure 1C), while only the BR and BS treatments caused a significant reduction in the fungal Shannon index (p < 0.05, Figure 1D).

3.2. Bacterial Community Composition

In bacterial populations, relative abundance (RA) shifts were observed in response to various treatments compared to the CK treatment (Figure 2A). Specifically, the BR, SR, and BS treatments led to a significant increase in the RA of Proteobacteria (p < 0.05). In contrast, these treatments also significantly reduced the RAs of Actinobacteria, Chloroflexi, Planctomycetes, and Verrucomicrobia (p < 0.05). Furthermore, relative to the CK, both the SR and BS treatments markedly reduced the RA of Nitrospirae (p < 0.05), while the BS treatment significantly decreased the RA of Saccharibacteria (p < 0.05). At the same time, aggregate fraction trends demonstrated distinct patterns for different bacterial groups (Figure 2A). For instance, the RAs of Nitrospirae and Verrucomicrobia exhibited the ME ≥ MA ≥ MI pattern, while Proteobacteria had the MI ≥ MA > ME pattern (p < 0.05). The RAs of Acidobacteria and Gemmatimonadetes in the MI fraction were lower than those in the ME and MA fractions (p < 0.05). Additionally, the distribution of Actinobacteria followed the MI ≥ ME ≥ MA pattern, and Chloroflexi displayed the ME > MI ≥ MA pattern (p < 0.05).

The LEfSe analysis of bacterial community (Figure A1) revealed that in ME, the BR treatment significantly enriched o_Nitrospirales, s_uncultured_bacterium_f_Nitrosomonadaceae, and s_uncultured_bacterium_f_Gemmatimonadaceae (p < 0.05). In MA, the CK treatment led to an increase in p_Acidobacteria and c_Betaproteobacteria (p < 0.05); o_Rhizobiales was notedly increased in the BR treatment (p < 0.05); while the BS treatment significantly enriched g_Gemmatimonas, c_Holophagae, and o_Myxococcales (p < 0.05). In MI, the CK treatment caused an enrichment of s_uncultured_bacterium_o_Gaiellales, g_Pseudarthrobacter, and c_KD4_96 (p < 0.05); the BR treatment led to an increase in f_Xanthomonadaceae, g_Sphingomonas, and g_Rhodanobacter (p < 0.05); the SR treatment resulted in an enrichment of o_Xanthomonadales and o_Sphingomonadales (p < 0.05); lastly, an enrichment of f_Comamonadaceae was observed in the BS treatment (p < 0.05).

3.3. Fungal Community Composition

The relative abundance (RA) of fungal populations also exhibited various changes under the different treatments (Figure 2B). In comparison to CK treatment, the SR treatment notably increased the RAs of Olpidiomycota and Rozellomycota (p < 0.05), while the BS treatment significantly raised the RA of Mortierellomycota (p < 0.05). Furthermore, the RAs of Basidiomycota and Glomeromycota in ME were greater than that in MA and MI (p < 0.05). Conversely, the RA of Ascomycota in MI surpassed that in ME and MA (p < 0.05). Notably, the RAs of Chytridiomycota and Mortierellomycota were higher in MA than in ME and MI (p < 0.05).

The LEfSe results of fungal community (Figure A2) shown that in ME, g_Podospora and s_Glomus_indicum were notably enriched in the CK treatment (p < 0.05); the BR treatment led to an increase in g_Rhodotorula, g_Naganishia, and f_Glomeraceae (p < 0.05); the SR treatment resulted in the enrichment of s_Olpidium_brassicae (p < 0.05); and f_Lasiosphaeriaceae was notedly increased in the BS treatment (p < 0.05). In MA, the lineages enriched in the BR treatment were derived from s_Mortierella_alpina, g_Penicillium, f_Pseudeurotiaceae, and o_Thelebolales (p < 0.05); the SR treatment led to an increase in f_Herpotrichiellaceae and g_Microdochium (p < 0.05); and the BS treatment caused an enrichment of g_Sphaerosporella and g_Mortierella (p < 0.05). In MI, the lineages enriched in the CK treatment were derived from s_Cladosporium_delicatulum, s_Chaetomidium_gallecicum, s_Guehomyces_pullulans, f_Pleosporaceae, and f_Xylariaceae (p < 0.05); f_Nectriaceae was notably enriched in the BR treatment (p < 0.05); while the SR treatment significantly enriched g_Talaromyces and g_Chaetomium (p < 0.05).

3.4. Redundancy Analyses

The RDA results of the soil bacterial community demonstrated that the first and second ordination axes accounted for 35.16% and 16.10% of the total variance, respectively (Figure 3A). Notably, the bacterial community exhibited an extremely significant positive correlation with the TN/TP ratio (p < 0.001; Figure 3A), a very significant positive correlation with both TN and TP contents (p < 0.01; Figure 3A), and a significant negative correlation with the SOC/TN ratio (p < 0.05; Figure 3A). On the other hand, the RDA results of the soil fungal community revealed that the first and second ordination axes explained 29.53% and 17.78% of the total variance, respectively (Figure 3B). Moreover, compared with the CK treatment, the BR, SR, and BS treatments significantly increased the TN and TP contents in MI, and the BR treatment significantly elevated the TN and TP contents in MA (p < 0.05,

Table A1. Differences in soil chemical and biochemical properties among different straw-returning methods in three aggregates.

Aggregates	Treatments	SOC (g kg−1)	TN (g kg−1)	TP (g kg−1)	SOC/TN	SOC/TP	TN/TP	βG (IU g−1)	LAP (IU g−1)	NAG (IU g−1)	ALP (IU g−1)	C/N	C/P	N/P
ME	CK	7.31 c	0.95 b	0.49 a	7.71 c	14.95 c	1.94 b	328.00 a	1728.00 a	2848.67 a	414.67 a	0.07 a	0.80 a	11.17 a
	BR	10.31 a	0.54 c	0.52 a	18.97 a	19.98 b	1.05 c	362.67 a	1859.00 a	2591.33 a	427.33 a	0.08 a	0.85 a	10.57 a
	SR	10.47 a	1.03 a	0.44 b	10.23 b	24.05 a	2.35 a	301.33 a	1843.33 a	2485.00 a	411.00 a	0.07 a	0.73 a	10.57 a
	BS	9.34 b	0.97 b	0.52 a	9.64 b	18.10 b	1.88 b	307.67 a	1792.67 a	2485.00 a	372.67 a	0.07 a	0.82 a	11.57 a
MA	CK	7.73 c	1.05 b	0.45 bc	7.36 b	17.75 c	2.41 a	360.00 a	1555.00 a	2412.00 a	383.67 b	0.09 a	0.95 a	10.53 a
	BR	11.61 a	1.42 a	0.58 a	8.18 b	20.29 bc	2.48 a	326.33 a	1453.67 a	2516.00 a	402.67 ab	0.08 a	0.82 ab	10.03 a
	SR	11.12 ab	1.10 b	0.50 ab	10.14 a	22.24 b	2.20 a	340.00 a	1651.33 a	2589.00 a	511.00 a	0.08 a	0.67 b	8.33 a
	BS	10.47 b	1.05 b	0.40 c	9.94 a	26.51 a	2.67 a	396.67 a	1605.33 a	2775.33 a	440.00 ab	0.09 a	0.91 a	10.00 a
MI	CK	8.72 b	1.02 c	0.47 c	8.51 a	18.61 a	2.18 b	346.67 a	1349.67 b	2659.67 a	435.33 a	0.09 a	0.80 a	9.30 a
	BR	8.60 b	1.25 b	0.55 a	6.95 b	15.52 b	2.25 b	302.00 a	1710.00 ab	2579.67 a	375.67 a	0.07 ab	0.81 a	11.60 a
	SR	7.33 c	1.23 b	0.51 b	5.96 b	14.45 b	2.43 b	355.67 a	1938.33 a	2808.33 a	396.00 a	0.08 ab	0.90 a	12.13 a
	BS	10.94 a	1.54 a	0.55 a	7.12 b	20.17 a	2.85 a	299.00 a	1843.33 a	2914.67 a	373.00 a	0.06 b	0.80 a	12.83 a

SOC: soil organic carbon; TN: total nitrogen; TP: total phosphorus; βG: β-1,4-glucosidase; NAG: N-1,4-acetylglucosaminidase; LAP: leucine amino peptidase; ALP: alkaline phosphatase; C/N: βG/(NAG + LAP); C/P: βG/ALP; N/P: (NAG + LAP)/ALP; ME: mega-aggregates (>2 mm); MA: macroaggregates (0.25–2 mm); MI: microaggregates (<0.25 mm); CK: without biochar application or straw return; BR: biochar application of 30 t ha⁻¹; SR: straw return of 5 t ha⁻¹ year⁻¹; BS: combined application of biochar (30 t ha⁻¹) and straw (5 t ha⁻¹ year⁻¹). Least Significant Difference (LSD) was used for the post hoc test in analysis of variance (ANOVA) with three replicates. Different lowercase letters following the mean values indicate significant differences among the treatments (p < 0.05).

). Additionally, the fungal community displayed a highly significant positive correlation with the ALP activity (p < 0.01; Figure 3B) and a highly significant negative correlation with the βG/ALP ratio and the (NAG + LAP)/ALP ratio (p < 0.01; Figure 3B). Furthermore, relative to the CK treatment, the SR treatment markedly enhanced the ALP activity in MA and notably reduced the βG/ALP ratio in MA (p < 0.05, Table A1). Overall, under different straw-returning methods, the TN and ALP had the greatest impacts on the changes in bacterial and fungal communities, mainly in MA and MI.

3.5. Between-Class and Co-Inertia Analyses

The BCA was performed to check for dissimilarities in the soil microbial community. The composition of the bacterial and fungal communities varied significantly and explained 95.1% (p  <  0.001) and 86.5% (p  <  0.001) of the total variation, respectively (Figure 4A,B). The ellipses depicting the bacterial community composition of the SR and BS treatments were distinctly separated from those representing the CK and BR treatments (p  <  0.001, Figure 4A). The ellipses denoting the bacterial community composition of the BR treatment were distinctly separated from those of the CK treatment across three aggregate fractions (p  <  0.001, Figure 4A). The ellipses illustrating the fungal community composition of the BR and BS treatments were notably separated from those of the CK and SR treatments in MA and MI (p  <  0.001, Figure 4B). The ellipses representing the fungal community composition of the SR treatment was distinctly separated from that of the CK treatment in MA (p  <  0.001, Figure 4B). In addition, the ellipses depicting the microbial community composition of MI were distinctly separated from those of ME and MA (p  <  0.001, Figure 4A,B).

We employed the COIA to ascertain the co-variance between bacterial and fungal community structures, which revealed that 19.6% of the variation in microbial communities could be explained by this method (Figure 4C). The length of the arrow indicated the strength of the interaction, with a shorter arrow denoting a stronger interaction. The closer the direction of the arrow, the more similar the co-variation in microbial communities. In ME, the arrow length trend across different treatments was SR > BS > CK > BR; in MA, it was BR > SR > CK > BS; while in MI, the trend was BR > SR > BS > CK. Furthermore, in ME, the BR treatment’s arrow direction was approximately parallel to that of the CK treatment, whereas the SR and BS treatments were in contrasting directions to the CK treatment. In MA, the BR treatment shared the same arrow direction as the CK treatment, the SR treatment was in an opposing direction, and the BS treatment was perpendicular. In MI, the arrow direction for the CK treatment diverged from that of the other treatments.

3.6. Microbial Community Co-Occurrence Networks

To identify the co-occurrence patterns of soil bacterial and fungal communities, as well as niche partitioning across various treatments, we proceeded to construct microbial networks (Figure A3). The topological features distinguishing the bacterial and fungal co-occurrence patterns varied in relation to the number of nodes and links, average degree, and modularity. The links and average degree of both BR and SR treatments surpassed those of the CK treatment, whereas the BS treatment exhibited lower links and average degree in comparison to the CK treatment (

Table 1. Topological features, module hubs and keystone taxa of the microbial co-occurrence networks under three straw-returning methods.

		CK	BR	SR	BS
Topological features	Nodes	221	207	226	218
	Links	634	692	873	456
	Positive	60.60%	59.10%	60.60%	62.70%
	Negative	39.40%	40.90%	39.40%	37.30%
	Average degree	5.738	6.686	7.726	4.183
	Modularity	1.954	1.558	1.344	2.155
Module hubs	Bacterial phyla	Planctomycetes	Chloroflexi	Chloroflexi	Chloroflexi
		Acidobacteria	Acidobacteria	GAL15	Proteobacteria
		Nitrospirae	Proteobacteria	Verrucomicrobia	Acidobacteria
				Proteobacteria
				Gemmatimonadetes
	Fungal phyla	Glomeromycota	Ascomycota	Ascomycota	Basidiomycota
		Basidiomycota
Keystone taxa	Family	Phycisphaeraceae	uncultured_bacterium_c_P2-11E
		Blastocatellaceae_[Subgroup_4]	uncultured_bacterium_c_JG30-KF-CM66		uncultured_bacterium_c_JG37-AG-4
		Glomeraceae	Acidobacteriaceae_[Subgroup_1]		DA111

CK: without biochar application or straw return; BR: biochar application of 30 t ha⁻¹; SR: straw return of 5 t ha⁻¹ year⁻¹; BS: combined application of biochar (30 t ha⁻¹) and straw (5 t ha⁻¹ year⁻¹).

). Conversely, the modularity of the BR and SR treatments was lower than that of the CK treatment, and the modularity of the BS treatment was higher than that of the CK treatment (Table 1). In summary, the BR and SR treatments had higher complexity and connectivity of co-occurrence networks, while the BS treatment had higher modularity of co-occurrence networks.

Furthermore, within the CK network, the Planctomycetes, Acidobacteria, Nitrospirae, Glomeromycota, and Basidiomycota were identified as module hubs. Notably, the bacterial families Phycisphaeraceae (Planctomycetes) and Blastocatellaceae_[Subgroup_4] (Acidobacteria), along with the fungal family Glomeraceae (Glomeromycota), were categorized as keystone taxa (Table 1). In the BR network, module hubs included Chloroflexi, Acidobacteria, Proteobacteria, and Ascomycota. The bacterial families uncultured_bacterium_c_P2-11E (Chloroflexi) and uncultured_bacterium_c_JG30-KF-CM66 (Chloroflexi), as well as Acidobacteriaceae_[Subgroup_1] (Acidobacteria), were classified as keystone taxa (Table 1). The SR network had module hubs from Chloroflexi, GAL15, Verrucomicrobia, Proteobacteria, Gemmatimonadetes, and Ascomycota (Table 1). For the network of BS treatment, module hubs comprised Chloroflexi, Proteobacteria, Acidobacteria, and Basidiomycota. Here, the bacterial families uncultured_bacterium_c_JG37-AG-4 (Chloroflexi) and DA111 (Proteobacteria) were designated as keystone taxa (Table 1).

3.7. Function Prediction

The FAPROTAX functional prediction results showed that, relative to the CK treatment, the BR and SR treatments significantly enhanced bacterial chemoautotrophy in ME (Figure 5A,B), while the BS treatment significantly reduced bacterial respiration of sulfur and its compounds (Figure 5C). In MA, the BR treatment significantly promoted bacterial chemoautotrophy, whereas the BS treatment significantly reduced bacterial aerobic nitrite oxidation (Figure 5D,E). In MI, the BR, SR, and BS treatments significantly enhanced chitinolysis and significantly reduced the degradation of aromatic compounds; the SR and BS treatments significantly decreased aerobic nitrite oxidation (Figure 5F–H). Moreover, the BR treatment led to a significant decrease in photoautotrophy and nitrate reduction, whereas the SR treatment resulted in a significant reduction in nitrification and N fixation (Figure 5F,G). The FUNGuild function prediction results demonstrated that, in comparison to the CK treatment, the SR and BS treatments significantly enhanced ectomycorrhizal in MI (Figure 5I,J).

The results from the Tax4Fun function prediction revealed that, relative to the CK treatment in MI, the BR, SR, and BS treatments significantly enhanced the biofilm formation of Pseudomonas aeruginosa, Vibrio cholerae, and Escherichia coli (Figure 6). They also significantly facilitated bacterial secretion system, bacterial chemotaxis, flagellar assembly, C fixation in photosynthetic organisms, phosphonate and phosphinate metabolism, and plant-pathogen interaction (Figure 6). Conversely, these treatments led to a significant decrease in ABC transporters, as well as the degradation of aromatic compounds, atrazine, and polycyclic aromatic hydrocarbons (PAHs) (Figure 6). Furthermore, the BR and BS treatments significantly promoted two-component system and one C pool by folate (Figure 6A,C). Conversely, the SR and BS treatments significantly reduced N-Glycan biosynthesis, quorum sensing, and streptomycin biosynthesis (Figure 6B,C). Meanwhile, the BR and SR treatments significantly enhanced citrate cycle (Figure 6A,B). Additionally, the BR treatment led to a significant decrease in various forms of N-glycan biosynthesis, while the SR and BS treatments significantly promoted it (Figure 6).

4. Discussion

4.1. Straw-Returning Method and Aggregate Fraction Change Soil Microbial Community Structure

Overall, different straw-returning methods and aggregate fractions significantly affected the diversity of soil bacterial and fungal communities. In this study, the BR treatment significantly reduced the Shannon index of bacterial community across all aggregate fractions, while the SR treatment only markedly diminished the Shannon index of bacterial community in >2 mm aggregates, and the BS treatment notably decreased the Shannon index of bacterial community in >0.25 mm aggregates (p < 0.05). The above results indicated that the different straw-returning methods could reduce the bacterial community diversity, only that the different straw-returning methods affected different aggregate fractions. Similarly to the results of our study, Chen et al. [42] demonstrated that BR notably reduced the Shannon index of bacterial community in an incubation experiment under alternating wetting and drying. Conversely, there are also inconsistencies with our findings [33,43,44]. Bai et al. [33] noted that BR significantly increased the Shannon index of bacterial community of silt + clay in a wheat–rice rotation planting system. Additionally, Zhao et al. [43] found that SR notably enhanced the Shannon index of bacterial community in an incubation experiment. We speculated that the reasons for the different results might be (1) different types and application amounts of straw and biochar, and (2) different environments or soil types at the test sites. In addition, the trend of bacterial Shannon index among different aggregate fractions was ME > MA > MI, indicating that the micro-environment differences in aggregate fractions produce distinct microbial communities [34].

In the present study, the BR treatment significantly reduced the Shannon index of fungal community in <0.25 mm aggregates (p < 0.05), the BS treatment notably decreased the Shannon index of fungal community in <2 mm aggregates (p < 0.05), while the SR treatment had no significant effect, indicating that biochar had a significantly negative effect on fungal community diversity. Similarly to our findings, Chen et al. [42] demonstrated that biochar significantly reduced the diversity of fungal community, which may be due to the fact that BR increased soil pH, leading to a decrease in the diversity of fungal community preferring low pH environments [11]. However, previous studies have shown inconsistent effects of SR on fungal community diversity [22,42,45], which may be due to the different straw types and return conditions. Additionally, the trend of the fungal Shannon index among different aggregate fractions was ME ≥ MA > MI, indicating that the microenvironment differs significantly between different aggregates, consequently leading to distinct impacts on fungal community.

4.2. Straw-Returning Method and Aggregate Fraction Change Soil Bacterial Community Composition

Briefly, the composition of soil bacterial communities was significantly influenced by various straw-returning methods and aggregate fractions. Proteobacteria are eutrophic bacteria that prefer to colonize nutrient-rich environments, and the decomposition of straw and biochar provides a suitable environment for their rapid growth in soil [46]. Representative species of the phylum Proteobacteria, such as Rhizobium, Nitrosomonas, Desulfovibrio, Pseudomonas, and Acetobacter, have functions such as soil N fixation, nitrification, sulfate reduction, inorganic P solubilization, and acetic acid production [1]. In the present study, the BR, SR, and BS treatments significantly increased the RA of Proteobacteria (p < 0.05), which may be due to the increase in soil nutrients after application of biochar and straw. Consistent with our findings, previous studies have demonstrated that BR significantly enhanced the RA of Proteobacteria [1,42,47]. Moreover, in this study, the BR, SR, and BS treatments significantly decreased the RAs of Actinobacteria, Chloroflexi, Planctomycetes, and Verrucomicrobia (p < 0.05). Consistent with our results, previous studies have shown that BR significantly reduced the RAs of Actinobacteria and Chloroflexi [33,42,48]. However, previous findings were also inconsistent with our observations that BR and/or SR significantly enhanced the RAs of Actinobacteria, Chloroflexi, Verrucomicrobia, and Planctomycetes [44,49,50]. Actinobacteria are oligotrophic bacteria that prefer soil environments with low resource availability, and they play a crucial role in the degradation of aromatic compounds and environmental pollutants (e.g., hydrocarbons and pesticides), and antibiotic biosynthesis (e.g., streptomycin, tetracycline, and erythromycin) [27,44]. The increase in available C under different straw-returning methods may be responsible for the decrease in Actinobacteria [33]. Additionally, in this study, the distribution of Acidobacteria, Gemmatimonadetes, Chloroflexi, Nitrospirae, and Verrucomicrobia was significantly higher in ME compared to MI (p < 0.05). Conversely, Proteobacteria and Actinobacteria exhibited a greater distribution in MI as opposed to ME (p < 0.05). The above findings suggested that there were ecological niche differences in microbial taxa in different microenvironmental spaces.

In this study, the biomarkers associated with the CK treatment played a pivotal role in biological N fixation (e.g., c_Betaproteobacteria in MA and g_Pseudarthrobacter in MI) [51,52]. Meanwhile, the main functions of the biomarkers in the BR treatment were related to soil N cycle, including nitrification (e.g., o_Nitrospirales and s_uncultured_bacterium_f_Nitrosomonadaceae in ME) [53], N fixation (e.g., o_Rhizobiales in MA) [54], and denitrification (e.g., g_Rhodanobacter in MI) [55], P dissolution (e.g., s_uncultured_bacterium_f_Gemmatimonadaceae in ME) [56], and organics degradation (e.g., f_Xanthomonadaceae and g_Sphingomonas in MI) [44,57]. Moreover, the biomarkers in the SR treatment were primarily known for their organics degradation and bioremediation potential (e.g., o_Xanthomonadales and o_Sphingomonadales in MI) [44,57]. In addition, the biomarkers in the BS treatment were mainly known for their P solubilization capacity (e.g., g_Gemmatimonas in MA) [56], organics degradation (e.g., c_Holophagae in MA and f_Comamonadaceae in MI), and microbial community regulation (e.g., o_Myxococcales in MA) [58]. The above findings indicated that SR and BR alone or in combination have important roles in promoting the degradation and conversion of aromatic compounds into humus in the microaggregates, with the BR and SR treatments having the same functional taxa. Furthermore, the application of biochar, either singularly or in combination with SR, may enhance P solubilization, subsequently augmenting soil P utilization. Additionally, based on the results of our study, it can be inferred that SR alone or combined with BR may reduce biological N fixation.

4.3. Straw-Returning Method and Aggregate Fraction Change Soil Fungal Community Composition

Generally, the composition of soil fungal communities was significantly influenced by various straw-returning methods and aggregate fractions. In the present study, it was observed that the SR treatment notably enhanced the RAs of Olpidiomycota and Rozellomycota (p < 0.05). Olpidiomycota and Rozellomycota, both known for their ability to parasitize plants, play important roles in ecosystems by regulating host population dynamics and contributing to C cycling [59]. We speculated that the increased RAs of Olpidiomycota and Rozellomycota in the SR treatment resulted from direct return of plant residues. Moreover, in this study, the BS treatment significantly increased the RA of Mortierellomycota (p < 0.05). Mortierellomycota are predominantly recognized for their capacity to degrade organic matter and maintain soil health, and some Mortierellomycota species also have plant-growth promoting properties that help plants absorb nutrients and resist pathogenic bacteria [60,61,62]. Furthermore, the RA of Mortierella was positively correlated with the SOC content, and Mortierella species can compose recalcitrant substances and contribute to SOM storage [63,64]. Thus, we believed that the mechanism of C sequestration in the BS treatment was not only due to the large amount of C return, but also related to the biological C sequestration of Mortierellomycota.

Furthermore, aggregate fractions had significant impacts on the RAs of Ascomycota (higher in MI), Basidiomycota (higher in ME), Glomeromycota (higher in ME), Mortierellomycota (higher in MA), and Chytridiomycota (higher in MA) (p < 0.05). Previous studies have demonstrated that Ascomycota exhibits a preference for high-nitrogen soils, while Basidiomycota and Mortierellomycota are more adapt to low-nitrogen soils [61,65]. We inferred that the increased N limitation due to large amounts of C return may occur in ME and MA compared to MI. The Basidiomycota species exhibit a higher efficacy in degrading lignocellulosic organic matter compared to other fungal groups, while Ascomycota species demonstrate a restricted capacity to degrade litter that contains recalcitrant lignin [61]. Meanwhile, Chytridiomycota is capable of breaking down complex organic matter such as cellulose and chitin [66]. Therefore, our findings suggested that the decomposition of recalcitrant lignin is more evident significant in ME compared to MI, while the decomposition of cellulose and chitin is more pronounced in MA. Additionally, in this study, we observed that the BS had a negative impact on the RAs of Ascomycota, Basidiomycota, and Glomeromycota, compared to their individual applications, whereas it had a positive effect on the RA of Mortierellomycota (p < 0.05). As we known, Glomeromycota plays an important role in promoting plant nutrient uptake, enhancing stress tolerance, and improving soil aggregate stability [67]. This suggested that the BS may have a negative effect on soil aggregate stability, relative to the application of biochar or straw alone.

In addition, the fungal lineages enriched in the CK treatment were g_Podospora and s_Glomus_indicum of ME, as well as s_Cladosporium_delicatulum, s_Chaetomidium_gallecicum, s_Guehomyces_pullulans, f_Pleosporaceae, and f_Xylariaceae of MI. Cladosporium is usually found on the leaves and stems of wilted or dead herbaceous plants, and some Cladosporium species exhibit phytopathogenic traits [68]. Moreover, the fungal lineages that were enriched in the BR treatment included g_Rhodotorula, g_Naganishia, and f_Glomeraceae of ME; s_Mortierella_alpina, g_Penicillium, f_Pseudeurotiaceae, and o_Thelebolales of MA; as well as f_Nectriaceae of MI. Mortierella_alpina can compose recalcitrant substances and contribute to SOM storage. Mortierella_alpina and Penicillium have been demonstrated to be effective in the biological control of plant diseases [69,70], and some species of Mortierella and Penicillium have exhibited the capability to solubilize phosphate, thereby promoting plant development [71,72]. Meanwhile, the fungal lineages enriched in the SR treatment were s_Olpidium_brassicae of ME; f_Herpotrichiellaceae and g_Microdochium of MA; as well as g_Talaromyces and g_Chaetomium of MI. Olpidium_brassicae can directly infect plant roots and cause plant diseases, and it is also a vector for many plant viruses, posing a threat to agricultural production [73,74]. Microdochium have been demonstrated to be effective in the biological control of plant diseases [75], and Chaetomium can promote plant growth, improve soil health, and mitigate the continuous cropping obstacles [76]. Furthermore, the fungal lineages that were enriched in the BS treatment included f_Lasiosphaeriaceae of ME, as well as g_Sphaerosporella and g_Mortierella of MA. Sphaerosporella can promote plant growth and reduce heavy metal concentrations [77]. The above results demonstrated that the BR and BS treatments in MA had the same functional taxa of Mortierellomycota, which indicated that BR brought more organic matter degradation and plant promotion of macroaggregate.

4.4. Straw-Returning Method and Aggregate Fraction Affect the Interaction of Microbial Community

Microbial co-occurrence networks with a greater number of links and a higher average degree exhibit more complex coupling and strength between microorganisms [78], and the modularity of the network reflects the integrity of microbial community interactions [79]. Previous studies have shown that the application of either straw or biochar can increase the complexity of microbial co-occurrence networks [42], which is consistent with our results. Meanwhile, it has been reported that more complex networks may be formed in nutrient-rich environments [80]. In this study, the application of biochar and straw increased the complexity, strength and integrity of microbial community interactions, while the BS reduced them. This indicates that the BS had a negative effect on the complexity, strength and integrity of microbial interactions, possibly due to excessive C return. Moreover, previous studies have shown that BR enhanced microbial community interactions in terms of the number of links and modularity, especially in MA [1]. The results of co-inertia analysis can reveal the strength and similarity of microbial community interactions [81]. Combined with the results of our co-inertia analysis, it can be inferred that BR promoted microbial community interactions mainly in ME, while the BS reduced microbial community interactions predominantly in ME and MI. Furthermore, in this study, microbial community interactions in the BR treatment were similar to the CK treatment and diametrically opposed to the SR treatment, especially in ME and MA. Meanwhile, the opposing effects of BR and SR on microbial interactions appeared to be superior for BR in ME, while in MA the two appeared to be equivalent. Variations in microbial community interactions can be attributed to the diverse structures within different aggregate fractions of the microbial community.

Additionally, the module hubs of the microbial co-occurrence network shifted significantly across various treatments (e.g., Chloroflexi, Proteobacteria, and Ascomycota), characterizing an altered core function of the microbiota. In particular, the members of bacterial family uncultured_bacterium_c_P2-11E (Chloroflexi), uncultured_bacterium_c_JG30-KF-CM66 (Chloroflexi), and Acidobacteriaceae were categorized as keystone taxa in the BR treatment; meanwhile, the members of bacterial family uncultured_bacterium_c_JG37-AG-4 (Chloroflexi) and DA111 (Proteobacteria) were identified as keystone taxa in the BS treatment. Chloroflexi is capable of non-oxygenic photosynthesis and specialized C fixation pathways [82]. The highly connected keystone taxa in the networks can explain compositional changes in the microbiome more independently than the combination of all groups, highlighting the important role of maintaining the integrity and function of the microbiome [83]. The above results demonstrated that Chloroflexi, a key member of the microbial co-occurrence networks, could synergize with other microbial groups (e.g., Proteobacteria and Acidobacteria) to maintain network stability.

4.5. Straw-Returning Method and Aggregate Fraction Affect the Function of Microbial Community

The sequencing of 16S rRNA suggested an increased expression of genes associated with C (rbcL, acsA, gltS, aclB, and mcrA) and N (nifH and amoA) transformation following the addition of biochar [1]. In this study, the FAPROTAX functional prediction showed that BR treatment had the potential to decrease the function of nitrate reduction in MI, which may lead to a reduction in its N₂O emissions [84,85]. Furthermore, both the SR and BS treatments significantly reduced the RA of Nitrospirae, which are mainly involved in nitrification. Meanwhile, the SR treatment could potentially reduce N fixation and nitrification (e.g., aerobic nitrite oxidation) in MI, as well as the BS treatment presented the potential to diminish aerobic nitrite oxidation in MA and MI, which may result in a reduced source of N available for plant uptake [84,86].

Moreover, some species of Chloroflexi, Verrucomicrobia, and Planctomycetes possess the capability to degrade complex organics (e.g., petroleum hydrocarbons and PAHs), oxidize sulfides, or reduce sulfates in environmental contexts. Our findings suggested that the decreased RAs of Chloroflexi, Verrucomicrobia, and Planctomycetes observed in the BS treatment could be responsible for the diminished respiration of sulfur and its compounds in ME. Further combined with Tax4Fun function prediction, the BR, SR, and BS treatments demonstrated the potential to enhance chitinolysis and C fixation in photosynthetic organisms, while potentially mitigating the degradation of aromatic compounds in MI, thereby these potential functional changes, to some extent, could promote C sequestration in microaggregates. In addition, the BR, SR, and BS treatments presented the capability to enhance the related functions of biofilm formation and bacterial chemotaxis in MI, despite reducing various types of N-Glycan biosynthesis. Meanwhile, the FUNGucarbonild function prediction results demonstrated that the SR and BS treatments possessed the potential to augment ectomycorrhizal in MI. The above results indicated that microbial communities under BR and SR alone or in combination could improve C fixation, biofilm cementation, and mycelial entanglement, especially in MI, which further causes improvement in soil aggregate structure [87].

In this study, the BR, SR and BS treatments also significantly reduced atrazine degradation in MI. On the one hand, the pores within the biochar act as microbial habitats, which may minimize the probability of interactions between microorganisms and pesticides. On the other hand, the addition of biochar to the soil may increase or decrease the quantity and population of microorganisms, which may alter the adsorption, degradation, or mineralization of pesticides. Consequently, these two factors collectively contribute to a decrease in the decomposition of pesticides in the soil following the application of biochar [88]. Moreover, the BR and BS treatments notably enhanced two-component system and one C pool via folate. Conversely, the SR and BS treatments markedly diminished quorum sensing and streptomycin biosynthesis. In summary, BR and SR had synergistic effects on most microbial community functions, but they still had their own dominant differential functions, which should be determined by C source characteristics rather than quantity.

Due to the inherent limitations of this study, such as the lack of multi-point long-term field comparative research and that the analysis of microbial functional differences is based on predicted results rather than actual direct measurements, the generalizability and applicability of the conclusions of this study may be limited.

5. Conclusions

In this study, relative to the CK, the different straw-returning methods reduced the bacterial community diversity, primarily due to the influence of ME, while the BR and BS treatments decreased the fungal community diversity, largely attributed to the impact of MI. In addition, compared with the CK, the BR and SR treatments had higher complexity and connectivity of co-occurrence networks, while the BS treatment had higher modularity of co-occurrence networks. Further combined with co-inertia analysis, the BR treatment promoted the microbial community interaction mainly in ME, while the BS treatment reduced the microbial community interaction predominantly in ME and MI. Functional predictions showed that BR and SR had synergistic effects on most microbial community functions in MI, e.g., enhancement of chitinolysis, biofilm formation, bacterial chemotaxis, and C fixation in photosynthetic organisms and reduction of aromatic compounds, PAHs, and atrazine degradation. Nevertheless, different straw-returning methods exhibited distinct functions. For instance, the BR treatment could potentially reduce nitrate reduction in MI, while the SR treatment possessed the potential to decrease N fixation and nitrification in MI. Additionally, the BS treatment had the potential to diminish the respiration of sulfur and its compounds in ME, as well as aerobic nitrite oxidation in MA and MI. In summary, BR and SR alone or in combination improved C fixation and biofilm formation in MI; however, they had different effects on N-cycling functions. Meanwhile, the BS had a negative effect on soil microbial community diversity and interaction, especially in ME and MI. These findings provide useful insights for understanding microbial community interactions and functions in soil aggregates mediated by biochar and straw additions. Certainly, the current study has limitations. In the future, further research should be conducted to provide measurable direct validation of microbial functions.