Short-term effects of different forms of straw return on rhizosphere and bulk soil bacterial communities and microenvironmental conditions in saline-alkali maize fields

Salt stress and low nutrient availability in saline-alkali soils constrain crop productivity. However, the effects of different forms of exogenous carbon input on rhizosphere bacterial communities and their predicted functional potential remain unclear. A field experiment was conducted in a coastal saline-alkali maize field with four treatments: no organic amendment (CK), crushed straw return (CS), straw biochar return (CB), and granulated straw return (GS). The effects of different straw forms on soil physicochemical properties, bacterial community structure, predicted functional profiles, and maize yield were compared between rhizosphere and bulk soils. Compared with CK, GS significantly reduced rhizosphere electrical conductivity (EC) and salt content by 57.97% and 59.56%, respectively. Among the different straw forms, GS resulted in significantly lower rhizosphere salt content than CS and CB, while its EC was comparable to CS. Both CS and GS significantly increased rhizosphere available phosphorus content and increased maize yield by 24.87% and 23.35%, respectively. Straw-return treatments significantly altered bacterial community structure in both rhizosphere and bulk soils, with GS enriching a greater number of treatment-specific bacterial taxa in both habitats. Functional prediction based on bacterial community composition showed that, compared with CK, GS was associated with significantly lower predicted relative abundances of multiple MetaCyc pathways related to aerobic respiration, the tricarboxylic acid cycle, and branched-chain amino acid biosynthesis in both rhizosphere and bulk soils. Overall, different straw-return forms exerted distinct short-term effects on the soil microenvironment and bacterial communities. Notably, GS showed a relative advantage in reducing rhizosphere salinity, particularly salt content, while both CS and GS improved rhizosphere nutrient conditions and maize yield.

Soil salinization is a major abiotic stress that limits global agricultural productivity, threatening food security, and affecting ecosystem sustainability (Zhao et al., 2025). High concentrations of soluble salts and exchangeable sodium can induce osmotic stress, ionic imbalance, and restricted nutrient uptake in crops. They can also alter soil physical structure, carbon and nitrogen transformation processes, and microbial community composition, thereby limiting the recovery of soil ecological functions and crop yield improvement in saline-alkali land (Lei et al., 2025; Shrivastava and Kumar, 2015). Therefore, the improvement of saline-alkali land should focus not only on reducing soil salinity but also on improving the coordination among soil microenvironmental conditions, bacterial community functions, and crop growth. Compared with bulk soil, the rhizosphere is a microenvironment where root activity, nutrient fluxes, salt movement, and microbial activity are highly concentrated. Root exudate input, root uptake, and the metabolism of bacteria and other microorganisms jointly shape a highly heterogeneous rhizosphere environment, making it the most sensitive zone for crop adaptation and soil improvement responses under saline-alkali stress (Ling et al., 2022). Therefore, analyzing the effects of saline-alkali soil improvement from the perspective of the rhizosphere microenvironment is important for understanding how exogenous carbon input affects rhizosphere bacteria, their functional potential, and crop productivity.

Straw return contributes to both agricultural residue recycling and soil improvement, and it is an important measure for alleviating production constraints and enhancing soil ecological functions in saline-alkali land (Ninkuu et al., 2025). Previous studies have shown that straw return can affect the soil environment in saline-alkali land by supplying organic substrates, improving soil pore structure, regulating water and salt migration, and promoting nutrient release (Zhang et al., 2020; Li et al., 2024). However, different forms of straw may function differently after being incorporated into soil. Crushed straw is relatively easy to decompose. It can provide readily available carbon sources for soil microorganisms such as bacteria, and can affect rhizosphere nutrient supply by improving soil structure. Biochar, as a product of high-temperature pyrolysis, has high stability and a well-developed porous structure. It can influence bacterial colonization and community assembly by adsorbing salt ions, regulating active carbon fractions, and altering microhabitat conditions (Zhang et al., 2020; Akhtar et al., 2015; Li and Li, 2022). Granulated straw has a certain degree of structural stability as well as decomposability. Its regular form and relatively slow carbon release process may continuously affect bacterial carbon utilization and the pattern of carbon supply in the rhizosphere (Chen et al., 2020). Therefore, the effects of different straw forms on saline-alkali land may reflect not only differences among amendment types but also differences in the physicochemical properties and spatial structure of exogenous carbon inputs.

Although crushed, carbonized, and granulated straw have all been shown to improve the physicochemical properties of saline-alkali soils, most existing studies have considered straw return as a relatively uniform soil improvement practice. Insufficient attention has been paid to the spatially heterogeneous effects of different physical forms of straw in rhizosphere and bulk soil. As the microzone where plant roots interact most strongly with the soil environment, the rhizosphere differs markedly from bulk soil in physical structure, nutrient flux, salt migration, and microbial metabolic activity (Ling et al., 2022). Under saline-alkali stress in particular, differences in soil water, salt, and gas transport caused by different forms of straw may be further amplified in the rhizosphere microenvironment. These changes may indirectly affect how crops recruit functional bacterial taxa through root exudates, thereby helping alleviate osmotic stress and nutrient limitation. Previous studies have suggested that crops can regulate the rhizosphere microhabitat through root exudates and recruit salt-tolerant or growth-promoting bacterial taxa, thereby alleviating osmotic stress and insufficient nutrient acquisition (Li et al., 2021; Liu et al., 2022; Xun et al., 2024). Bacterial communities are important participants in straw-derived carbon decomposition, nutrient transformation, and the maintenance of rhizosphere functions. They therefore represent a key link connecting exogenous carbon input, changes in the soil microenvironmental conditions, and crop growth responses (Chai et al., 2026; Zhang et al., 2024). However, previous studies have mainly focused on bulk soil or overall changes in soil communities. A systematic understanding is still lacking of how different forms of straw regulate the rhizosphere microenvironment and further affect bacterial community restructuring and potential functional responses. Therefore, analyzing the effects of different forms of straw return on bacterial communities and functional potential in saline-alkali soils from the perspective of rhizosphere and bulk spatial differences can help clarify how exogenous carbon inputs influence soil ecological functions and crop productivity.

Based on this background, a field experiment was conducted in saline-alkali soil to compare three forms of straw return, namely crushed maize straw return, maize straw biochar return, and granulated maize straw return, with a no-organic-amendment control, on soil microenvironmental conditions, bacterial community structure, predicted functional potential, and maize yield in rhizosphere and bulk soil. This study aimed to explore the relative advantages of different forms of straw return in improving soil physicochemical properties and enhancing maize productivity in the short term. Accordingly, we hypothesized that, despite comparable C inputs, differences in material form and carbon stability may create distinct resource conditions and bacterial community responses between rhizosphere and bulk soils. In particular, granulated straw may induce rhizosphere responses that differ from those of crushed straw and straw biochar.

Study site

The field experiment was conducted in the Yellow River Delta Agricultural High-Tech Industrial Demonstration Zone, Guangrao County, Dongying City, Shandong Province, China (37°02′N, 118°25′E). The flat plain terrain was favorable for maintaining relatively uniform field conditions. The study area has a warm temperate continental monsoon climate, with distinct seasons and concurrent rainfall and heat. The mean annual temperature is approximately 12.3 °C, and the frost-free period lasts about 200 days. The mean annual precipitation is 532 mm. The soil type is mainly coastal saline-alkali soil. Before the experiment, the soil properties in the 0–20 cm layer were as follows: pH, 8.46; salt content, 1.88 g kg -1; soil organic carbon (SOC), 7.6 g kg -1; total nitrogen (TN), 1.01 g kg -1; alkali-hydrolysable nitrogen (AN), 50.41 mg kg -1; available phosphorus (AP), 34.98 mg kg -1; and available potassium (AK), 393.71 mg kg -1.

Experimental design

The study area was abandoned land that had not been cultivated or fertilized before 2023. The experiment was initiated in 2023 using a randomized block design with four treatments: no organic amendment (CK), crushed maize straw return (CS), maize straw biochar return (CB), and granulated maize straw return (GS). Each treatment was replicated three times, and each plot covered an area of 30 m 2 (3 m × 10 m). Before the experiment, crushed maize straw (15 t ha -1), maize straw biochar (8 t ha -1), and granulated maize straw (15 t ha -1) were thoroughly incorporated into the 0–10 cm soil layer by rotary tillage. The maize straw return rate was determined based on the local high straw yield level of 15 t ha -1. Based on the measured TOC contents of the amendments (Supplementary Table 1), the biochar application rate was set at 8 t ha −1 to provide a TOC input comparable to that of 15 t ha −1 straw (Wang et al., 2024). Here, the equal C input principle refers to equal amounts of total organic carbon (TOC) applied across the three straw-return treatments. The maize straw was collected from local maize fields, air-dried, crushed, and then applied to the soil. Maize straw biochar was produced from the same maize straw under anaerobic conditions at 700 °C (Cong et al., 2022). The nutrient contents of maize straw and straw biochar are shown in Supplementary Table 1. The spring maize cultivar Ludan 506 was sown on May 10, 2023. The wide row spacing was 0.7 m, the narrow row spacing was 0.5 m, and the plant spacing was 0.2 m, with a planting density of 90,000 plants ha -1. In addition, 750 kg ha -1 controlled-release fertilizer with an N:P:K ratio of 28:6:6 was applied at the bell-mouthed stage, together with 300 kg ha -1 urea. Irrigation was conducted twice, once after sowing and once at the bell-mouthed stage, with 750 m 3 ha -1 applied each time. Other field management practices followed local conventional maize cultivation standards.

Soil and crop sampling

After harvest in 2023, bulk soil was collected from the 0–20 cm soil layer within 30 cm around the maize plants. Rhizosphere soil was collected by gently brushing off the soil adhering to maize roots within approximately 2 mm using a sterile brush. Soil samples were composited in equal proportions and divided into two subsamples for subsequent analyses. One subsample was placed in a sterile plastic bag, properly labeled, quickly stored in an ice box, and transported to the laboratory. It was then stored at −80 °C for microbial analysis (Dong et al., 2021). The other subsample was air-dried, sieved, and ground for the determination of soil pH, AK, AP, and other soil properties (Bao, 2010; Lu, 2000). The detailed methods for measuring soil physicochemical properties are provided in Supplementary Table 2. Maize grain samples were collected from each plot using two rows with a length of 5 m. After air drying, the samples were threshed to determine maize yield.

Statistical analysis

Soil bacterial communities were analyzed by high-throughput sequencing of the 16S rRNA gene, as described in Supplementary Table 2. Demultiplexed paired-end reads were processed using Quantitative Insights Into Microbial Ecology 2 (QIIME 2). After primer removal, reads were quality-filtered, denoised, merged, and screened for chimeras using the Divisive Amplicon Denoising Algorithm 2 (DADA2) plugin to generate amplicon sequence variants (ASVs) and an ASV feature table. Representative ASV sequences were taxonomically assigned against the SILVA database. Before alpha- and beta-diversity analyses, the ASV feature table was rarefied to the minimum sequencing depth. Linear mixed-effects models were used to evaluate the effects of straw-return treatment, soil habitat, and their interaction on soil physicochemical properties, with treatment and soil habitat as fixed effects and PlotID as a random effect. One-way analysis of variance (ANOVA) followed by Fisher’s least significant difference (LSD) test was used to compare straw-return treatments within each soil habitat and maize yield among treatments. Figures were generated using Python 3.11 and matplotlib. Data processing was conducted using pandas and numpy, while scikit-bio, scipy.stats, matplotlib, and related libraries were used for community structure visualization and analysis.

Soil physicochemical properties and maize yield

Linear mixed-effects models revealed significant interactions between straw-return treatment and soil habitat for electrical conductivity (EC), salt content, AP, AN, SOC, particulate organic carbon (POC), and mineral-associated organic carbon (MOC) (p < 0.05; Supplementary Table 3), indicating that the responses of soil physicochemical properties to straw-return treatments differed between rhizosphere and bulk soils. In rhizosphere soil, straw-return treatments significantly altered soil salinity-related properties and nutrient availability (Figure 1). Compared with CK, CS significantly increased rhizosphere soil pH by 7.15%, whereas the numerical increases under CB and GS were not significant. CS, CB, and GS significantly reduced EC by 39.55%, 21.71%, and 57.97%, respectively. Among different straw returning forms, rhizosphere EC under GS was significantly lower than that under CB but did not differ significantly from that under CS. Rhizosphere salt content was significantly reduced under CB and GS, whereas the decrease under CS was not significant relative to CK. GS produced a 59.56% reduction relative to CK and was significantly lower than both CS and CB. In bulk soil, EC did not differ significantly among treatments. Salt content under GS was significantly higher than that under CK, significantly lower than that under CB, and comparable to that under CS. In addition, CS and GS significantly increased available phosphorus by 54.82% and 75.26%, respectively. However, AN decreased under CB and GS treatments, suggesting that different straw forms produced contrasting effects on soil nutrient availability. Straw-return treatments also modified the contents of soil carbon fractions. In rhizosphere soil, SOC was significantly lower under CS and GS than under CK. CB and GS increased POC by 48.46% and 16.52%, respectively, whereas GS decreased MOC by 6.01%. In bulk soil, GS increased POC by 15.56%, while CB increased MOC by 3.11%. Evidently, different straw return forms led to differential accumulation of organic carbon fractions in the rhizosphere and bulk soil. Compared with CK, maize yield was significantly increased by 24.87% and 23.35% under CS and GS treatments, respectively (Figure 1), whereas CB did not significantly affect yield.

Bacterial community structure

The alpha diversity results showed that, compared with CK, the Chao1 index of rhizosphere soil was significantly decreased under CB treatment, whereas no significant changes were observed in the other treatments or in the Observed ASVs and Shannon indices (Supplementary Table 4). In bulk soil, the Chao1, Observed ASVs, and Shannon indices under CB and GS treatments were all significantly lower than those under CK.

Principal coordinate analysis (PCoA) showed that the first two coordinate axes explained 38.55% of the variation in bacterial community structure, reflecting the effects of different straw return practices on rhizosphere bacterial communities (Figure 2a). Permutational multivariate analysis of variance (PERMANOVA) indicated significant differences in community structure among the four treatments (Pseudo-F = 2.761, p = 0.001). Further permutational analysis of multivariate dispersions (PERMDISP) showed no significant differences in within-group dispersion among treatments in rhizosphere soil (F = 1.670, p = 0.269), indicating that the PERMANOVA result for rhizosphere soil mainly reflected differences in community composition among treatments (Supplementary Table 5). In terms of sample distribution, the treatments showed a relatively clear separation pattern. The CB group was clearly separated from the CK and CS groups, while the GS group was mainly distributed along the positive direction of PCoA1. The CK and CS groups were close to each other, with partial overlap. The envfit results showed that, after Benjamini–Hochberg false discovery rate (FDR) correction, POC, SOC, and AN were significant environmental factors (q ≤ 0.05). Among them, POC mainly pointed toward the distribution area of the CB group, whereas SOC and AN were more closely associated with the CK and CS groups. PCoA of the bulk-soil bacterial community revealed significant differences among treatments (PERMANOVA, Pseudo- F = 3.037, p = 0.001; Figure 2b), with the first two axes explaining 42.01% of the total variation. PERMDISP also showed significant differences in within-group dispersion (F = 29.935, p = 0.001; Supplementary Table 5). Given the balanced sampling design, the PERMANOVA result supports treatment-related multivariate differentiation, while the significant PERMDISP result indicates that differences in within-group heterogeneity also contributed to the observed community pattern (Supplementary Table 5). Unlike the rhizosphere habitat, the four treatments occupied distinct regions of the bulk-soil ordination plot. Although no environmental factor was significant after multiple-testing correction (q > 0.05), the projection directions of AN, SOC, MOC, and salt content were retained for descriptive interpretation of the bulk-soil community pattern.

Variation partitioning analysis showed that salinity-related factors, carbon fractions, and nutrient variables jointly explained 18.4% and 18.3% of the bacterial community variation in rhizosphere and bulk soil, respectively. In rhizosphere soil (Figure 3a), carbon fractions had the highest independent explanatory proportion at 9.2%, followed by nutrient variables (6.3%) and salinity-related factors (2.8%), indicating that changes in rhizosphere bacterial communities were mainly associated with carbon fractions and nutrient conditions. In bulk soil (Figure 3b), carbon fractions and salinity-related factors independently explained 4.8% and 3.9% of the bacterial community variation, respectively, while the shared explanatory proportion between carbon fractions and nutrient variables was relatively high (6.6%). This suggests that bacterial community differentiation in bulk soil was more closely associated with the combined variation in carbon, nutrient, and salinity conditions. The relatively high residual proportions indicate that other unmeasured factors may also be involved in bacterial community changes.

Key differential bacterial taxa and community composition

The total number of ASVs among treatments followed the order CS (8843) > CK (7653) > GS (7555) > CB (6269) (Figure 4a). CS had the highest total number of ASVs (8843), and also showed the largest number of unique ASVs (3217). In contrast, CB had the lowest number of unique ASVs, and shared relatively fewer ASVs with the other treatments in pairwise comparisons. In bulk soil, the total number of ASVs followed the order CKB (8436) > CSB (8292) > CBB (6300) > GSB (5914) (Figure 4b). A total of 2530 ASVs were shared by all four treatments, representing the largest shared group. CK had the highest number of unique ASVs, with 3179 ASVs, whereas GS had the lowest number, with 1535 ASVs. CB still shared relatively fewer ASVs with the other treatments in pairwise comparisons. Compared with rhizosphere soil, bulk soil had higher numbers of unique ASVs under CK and CS treatments.

At the phylum level (Figure 4c), the bacterial community was mainly composed of Proteobacteria, Actinobacteriota, Gemmatimonadota, and Acidobacteriota. Proteobacteria was the most abundant phylum in rhizosphere soil across all treatments, with a relative abundance ranging from 21.73% to 25.62%. It was followed by Actinobacteriota (16.92%–22.00%), Gemmatimonadota (12.33%–19.30%), and Acidobacteriota (12.13%–12.97%). The dominant phyla in bulk soil were generally consistent with those in rhizosphere soil (Figure 4d), and were still dominated by Proteobacteria, Gemmatimonadota, Actinobacteriota, and Acidobacteriota. Their relative abundances ranged from 21.94% to 29.25%, 15.60% to 21.37%, 14.02% to 16.54%, and 10.61% to 11.13%, respectively. At the genus level (Supplementary Figure 1), the distribution of dominant bacterial genera was generally similar between rhizosphere and bulk soil, and no obvious replacement of dominant genera was observed.

Linear discriminant analysis effect size (LEfSe) analysis with a linear discriminant analysis (LDA) score > 3.5 was used to identify biomarker taxa in each group (Figure 5a), and the corresponding biomarker cladogram was also generated (Figure 5c). The results showed that different straw forms not only altered bacterial community structure but were also associated with the differential enrichment of specific bacterial lineages. In rhizosphere soil, CS was mainly associated with Gemmatimonadota-related lineages, whereas GS enriched a broader range of taxa affiliated with Actinobacteriota and Proteobacteria (Figures 5a, c). In bulk soil, the biomarkers in CK were mainly associated with Gemmatimonadota and anaerobic Chloroflexi lineages. CS was enriched with Gemmatimonadota and myxobacterial taxa, while CB was associated with the enrichment of only a small number of biomarker taxa. GS had the largest number of enriched lineages, involving taxa related to Actinobacteriota, Proteobacteria, and Bacteroidota (Figures 5b, d). To further assess the statistical robustness of the LEfSe biomarkers, Kruskal–Wallis test results with FDR-adjusted q values are provided in Supplementary Table 6 and Suppelementary Table 7. All listed biomarker taxa remained significant after correction (q < 0.05), supporting the treatment-associated differences identified by LEfSe.

Relationships among key bacterial taxa, environmental factors, and maize yield

Correlation analysis showed that, in rhizosphere soil (Figure 6a), Tistrellales and Geminicoccaceae exhibited similar correlation patterns. Both taxa were significantly positively correlated with electrical conductivity, salt content, and AN, but significantly negatively correlated with available phosphorus (AP) and particulate organic carbon (POC) (FDR-adjusted q < 0.05). In contrast, Sphingomonas was significantly negatively correlated with electrical conductivity, salt content, and AN, while showing a significant positive correlation with AP (FDR-adjusted q < 0.05). Electrical conductivity was strongly positively correlated with salt content (r = 0.99, FDR-adjusted q < 0.001), and both variables were significantly negatively correlated with maize yield. AP was significantly positively correlated with yield (r = 0.63, FDR-adjusted q < 0.05). In addition, Tistrellales and Geminicoccaceae showed negative correlations with yield, although these relationships were not significant after FDR correction. In bulk soil (Figure 6b), Pontibacter was strongly positively correlated with POC (r = 0.94, FDR-adjusted q < 0.001), indicating a close association between its relative abundance and POC. The S0134 terrestrial group was significantly negatively correlated with Rhizobiales, Micrococcales, and Pseudoxanthomonas. Rhizobiales was also significantly negatively correlated with AN (FDR-adjusted q < 0.05). Rhizobiales, Micrococcales, and Pseudoxanthomonas showed positive correlations with yield, but these relationships were not significant after FDR correction. Exploratory correlations for all LEfSe biomarker taxa and environmental variables, based on unadjusted p values, are presented in Supplementary Figure 2.

Predicted functional profiles of bacterial communities

Based on MetaCyc pathway abundances predicted by Phylogenetic Investigation of Communities by Reconstruction of Unobserved States 2 (PICRUSt2), significant treatment-related differences were mainly observed in pathways associated with energy metabolism, amino acid biosynthesis, and cellular maintenance (Figure 7; Supplementary Tables 8, 9). In both rhizosphere and bulk soils, compared with CK, GS showed significantly lower predicted abundances of the aerobic respiration pathways PWY-3781 and PWY-6969, the tricarboxylic acid (TCA) cycle, the branched-chain amino acid biosynthesis pathways PWY-5101, VALSYN-PWY, ILEUSYN-PWY, and BRANCHED-CHAIN-AA-SYN-PWY, as well as PWY-5103. In rhizosphere soil, CB also significantly reduced the predicted abundances of several branched-chain amino acid biosynthesis pathways and PWY-5103, whereas its effects in bulk soil were limited mainly to PWY-6969 and the TCA cycle. CS generally showed no reduction in these predicted pathways, although PWY-5973 was significantly higher under CS in rhizosphere soil.

Straw-return forms differentially modified the rhizosphere and bulk soil microenvironments

Changes in the soil physicochemical environment are important for bacterial activity and crop growth. GS significantly reduced rhizosphere EC and salt content relative to CK (Figure 1), with salt content significantly lower than under CS and CB and EC significantly lower than under CB but comparable to CS. The concurrent reduction in salinity indicators and increase in maize yield indicate that GS created more favorable root-zone conditions during the growing season (Akhtar et al., 2015). Previous studies have shown that granulated straw can reduce soil bulk density, increase total and capillary porosity, and improve aggregate stability (Dong et al., 2021; Fan et al., 2026). In saline soils, improved aggregation and porosity can limit surface salt accumulation and promote salt leaching (Xie et al., 2020; Li et al., 2024). Accordingly, the reported effects of granulated straw on soil aggregation, pore structure, and hydraulic properties suggest that GS may have modified local water and salt movement, thereby contributing to the reduction in rhizosphere EC and salt content. An important feature of the GS treatment was the contrasting salinity response between rhizosphere and bulk soils. Rather than a uniform decrease throughout the sampled soil layer, this pattern suggests that GS may have modified the local distribution of water and salts around maize roots. Salinity gradients can develop between the soil–root interface and bulk soil because root-associated water movement is spatially heterogeneous (Jorda et al., 2018). Field observations in saline soils have likewise shown that water movement can create localized desalination zones while concentrating salts at the margins or lower positions of the wetting front (Wang et al., 2023). In this context, changes in local water-flow pathways associated with GS, together with root-associated water movement, may have contributed to the spatial redistribution of salts between root-adhering and surrounding bulk soils. Such spatial differentiation could help maintain a lower-salinity environment around the roots and may partly explain the yield response under GS. However, because the present comparison was confined to rhizosphere and bulk soils within the 0–20 cm layer, the relative contributions of horizontal redistribution and downward transport remain to be further resolved.

All amendments induced a slight increase in rhizosphere pH, which may have been jointly affected by ion exchange, alkaline salt migration, and organic matter mineralization in the short term (Wang et al., 2021). Therefore, this pH change should not be interpreted as a long-term trend. In the present study, the slight change in pH did not weaken the positive effects of granulated straw and crushed straw on maize growth. Instead, yield improvement occurred together with reduced electrical conductivity and alleviated salt stress. This suggests that, for seasonal improvement of coastal saline-alkali soil, reducing osmotic stress and improving the rhizosphere resource environment may be more critical than regulating pH alone (Akhtar et al., 2015; Li et al., 2025). Granulated straw and biochar increased rhizosphere POC, while granulated straw also decreased MOC. This indicates that amendment input did not simply increase soil carbon pools, but was also associated with redistribution between active and relatively stable carbon fractions (Wang et al., 2024; Yuan et al., in press). In the rhizosphere, the input of root exudates, straw decomposition, and enhanced bacterial metabolism may jointly promote active carbon turnover, leading to an increase in POC and a decrease in total SOC (Wang et al., 2025; Zhou et al., 2024). Such an increase in active carbon, accompanied by short-term fluctuations in total SOC, may reflect the dynamic distribution of microbial necromass carbon and plant-derived carbon rather than soil quality decline (Kalu et al., 2024; Lin et al., 2024). Notably, although CB produced the largest increase in rhizosphere POC, it did not significantly increase maize yield. This increase reflects greater carbon accumulation in the particulate fraction but does not necessarily indicate an immediate improvement in plant-available nutrient supply. Biochar-derived carbon is relatively stable and may contribute more to carbon accumulation than to short-term nutrient supply. Consistently, CB reduced rhizosphere AN and did not significantly increase AP, suggesting that temporary microbial nutrient immobilization and biochar-mediated nutrient adsorption or retention may have limited its short-term yield benefit. A previous field study similarly found that biochar enhanced soil carbon sequestration without significantly increasing maize yield during the first three years, whereas yield was positively related to labile organic carbon (Liang et al., 2021). Under CS and GS, enhanced microbial activity may have further contributed to the increase in rhizosphere AP. This was likely related to the chemical chelation of insoluble phosphorus by straw decomposition products and the biological activation of phosphorus-solubilizing bacteria (Li and Li, 2022; Sulastri et al., 2025). Although AN decreased significantly under granulated straw treatment, this does not necessarily indicate nutrient deficiency. It may be related to microbial immobilization, plant uptake, or nitrogen transformation processes. Selective microbial immobilization may have caused a short-term decrease in rhizosphere AN. However, during the key growth stage within a single season, this process may have helped reduce nitrogen leaching under high-intensity irrigation and enhance nutrient retention (Xu et al., 2021; Wang et al., 2026). Under the conditions of this experiment, these improvements in the soil physicochemical environment were associated, to some extent, with the yield response (Yu et al., 2024).

Responses of soil bacterial community diversity and composition to different forms of straw return

Species richness and community structure are important indicators for evaluating the response of soil bacterial communities to abiotic stress. Alpha diversity analysis showed that, in the short term, rhizosphere bacterial communities maintained a certain buffering capacity under root-mediated effects (Supplementary Table 4). Except for the decrease in Chao1 under CB, the rhizosphere alpha-diversity indices did not differ significantly from those under CK. In contrast, biochar and granulated straw significantly decreased bacterial diversity in bulk soil. This indicates that, in the absence of root involvement, straw amendments may have imposed a more direct environmental filtering effect on bacterial communities (Cong et al., 2025; Yuan et al., in press), thereby reducing some taxa that were sensitive to changes in salinity or substrate availability. Such a decline in diversity does not necessarily indicate weakened ecological function. Under strong stress conditions in saline-alkali soil, the adjustment of bacterial communities toward more adaptive taxa is also a common short-term ecological response (Delgado-Baquerizo et al., 2016; Shrivastava and Kumar, 2015). Combined with soil properties and yield results, crushed straw improved nutrient supply and increased yield while maintaining rhizosphere bacterial diversity, indicating a mild and beneficial regulatory effect. Although granulated straw decreased bacterial diversity in bulk soil, it did not significantly reduce diversity in the rhizosphere. At the same time, it alleviated salt stress and increased maize yield. This suggests that its effect was more likely associated with the selective enrichment of key responsive taxa rather than an adverse disturbance alone (Chen et al., 2020; Cong et al., 2019).

Beta diversity analysis showed significant differences in rhizosphere bacterial community structure among treatments (Figure 2a). The partial overlap between CK and CS suggests that the rhizosphere bacterial community under CS remained comparatively similar to that under CK, despite the observed improvement in nutrient conditions (Chai et al., 2026; Xie et al., 2020). In contrast, biochar was clearly separated from CK and crushed straw, and POC mainly pointed toward the biochar treatment, suggesting that biochar had a stronger effect on rhizosphere active carbon fractions and microhabitat conditions (Song et al., 2018). Granulated straw was separated from the other treatments along the positive direction of PCoA1, which may be associated with the reductions in salinity and changes in nutrient conditions observed under GS (Chen et al., 2020; Gai et al., 2025). Combined with the variation partitioning analysis (VPA) results, changes in rhizosphere bacterial communities were not driven by a single salinity or nutrient factor (Figure 3a), but were more closely associated with carbon input, nutrient transformation, and salinity alleviation. This was consistent with the envfit results showing the important roles of POC, SOC, and AN, indicating that rhizosphere bacterial communities were more sensitive to carbon source supply and resource regulation than to salinity stress alone. In bulk soil, the clear separation among treatments in the PCoA supported distinct treatment-associated changes in bacterial community composition (Figure 2b; Li et al., 2025; Yuan et al., in press). The significant PERMDISP result further indicated that these compositional changes were accompanied by differences in within-treatment heterogeneity, possibly reflecting spatially variable bacterial responses to amendment-derived carbon, nutrients, and salts. Although no significant environmental factor was identified after multiple-testing correction, the potential gradients of AN, SOC, MOC, and salt content, together with the VPA results (Figure 3b), suggest that bulk bacterial community differentiation was associated with the combined changes in carbon pools, nutrient conditions, and salinity. Overall, the PCoA, envfit, and VPA results indicate that different forms of straw return shaped bacterial communities through habitat-specific changes in soil microenvironmental conditions. Carbon and nutrient-related factors were more closely associated with rhizosphere bacterial communities, whereas combined environmental factors were more evident in bulk soil.

From the perspective of ASV richness and sharing patterns, different treatments affected bacterial communities through distinct pathways. In rhizosphere soil (Figure 4a), crushed straw return increased the supply of available substrates and microhabitat heterogeneity on the basis of root activity (Ninkuu et al., 2025; Chai et al., 2026). This may have provided more opportunities for bacterial colonization and coexistence and expanded rhizosphere niches (Chai et al., 2026; Wang et al., 2025). Together with its positive effects on available potassium, available phosphorus, and maize yield, these results indicate that crushed straw improved resource supply and was associated with a comparatively large rhizosphere ASV pool. In contrast, biochar showed relatively low total ASV numbers and unique ASV numbers in both habitats, indicating that it imposed stronger selective pressure on bacterial communities (Wang et al., 2021; Yuan et al., in press). Under the short-term experimental conditions, this filtering effect was more likely to reflect rapid changes in the soil microenvironmental conditions, carbon source availability, and ionic conditions after biochar input (Li and Li, 2022; Zhang et al., 2020). These changes may have inhibited some sensitive taxa, resulting in a contraction of the bacterial community member pool (Akhtar et al., 2015; Kalu et al., 2024). In rhizosphere soil, granulated straw did not show the same obvious community contraction as biochar, while it was also associated with reduced salinity and increased maize yield (Cong et al., 2019). This pattern may reflect the restructuring of bacterial communities under alleviated saline-alkali stress rather than a simple loss of taxa (Gai et al., 2025; Wang et al., 2026). In bulk soil (Figure 4b), granulated straw had the lowest number of unique ASVs, indicating that its return strengthened the selection of salt-tolerant or highly efficient resource-utilizing taxa in soil away from the root system (Chen et al., 2020). This pattern represents lower taxon richness but a more specific community shift, and it should be interpreted together with its positive rhizosphere effects and further long-term experimental evidence.

In terms of taxonomic composition, both rhizosphere and bulk soil were dominated by Proteobacteria, Actinobacteriota, Gemmatimonadota, and Acidobacteriota (Figures 4c, d). At the genus level, no obvious replacement of dominant genera was observed (Supplementary Figure 1). These results indicate that short-term return of different straw forms changed the number of taxa and their relative structure, but did not disrupt the basic framework of bacterial communities in saline-alkali soil (Vendruscolo et al., 2022; Wang et al., 2021). In other words, the amendments mainly regulated internal community proportions and niche distribution, rather than reconstructing an entirely new community (Delgado-Baquerizo et al., 2016; Tong et al., 2025). Overall, although crushed straw and granulated straw acted through different pathways, both showed more favorable effects on bacterial community regulation than biochar in the short term (Cong et al., 2025). Crushed straw was associated with a comparatively large rhizosphere ASV pool, whereas granulated straw was associated with a distinct bacterial community structure, which may reflect environmental filtering under altered salinity and nutrient conditions.

Key bacterial biomarker taxa and their associations with environmental factors and maize yield

Identifying bacterial taxa associated with soil environmental variation and crop performance is important for understanding bacterial responses to different straw-return practices. LEfSe analysis showed that the effects of different straw return forms on bacterial communities were reflected not only in overall community structure, but also in the significant enrichment of treatment-specific taxa. In rhizosphere soil (Figures 5a, c), crushed straw mainly enriched lineages related to Gemmatimonadota, whereas granulated straw enriched more taxa associated with Actinobacteriota and Proteobacteria and showed the widest lineage coverage. This indicates that both treatments were associated with selective enrichment of bacterial communities in saline-alkali soil during the growing season (Gai et al., 2025; Cong et al., 2019), with GS being associated with a broader range of biomarker taxa. Combined with the soil physicochemical and yield results, crushed straw and granulated straw improved the rhizosphere environment and were accompanied by the enrichment of key bacterial taxa and increased maize yield. In bulk soil, biomarker lineages were more widely distributed and showed stronger treatment specificity (Figures 5b, d), indicating that amendments exerted a more direct filtering effect on bacterial communities in the absence of root regulation (Wang et al., 2021). Crushed straw enriched Gemmatimonadota- and Myxococcota-related taxa, suggesting that it favored the formation of a bulk community structure characterized by enhanced organic substrate utilization and bacterial interactions (Xiao et al., 2022). Biochar enriched only a small number of taxa related to chitin degradation, further indicating that the dominant response it induced in the bulk environment was relatively limited. Granulated straw enriched the broadest range of biomarker taxa, suggesting that it also had a strong selective effect in bulk soil (Zhou et al., 2026).

The correlation results indicate that Tistrellales and Geminicoccaceae tended to occur in environments characterized by stronger salt stress and relatively limited phosphorus and particulate organic carbon availability (Figure 6a). These taxa may therefore serve as potential indicators of short-term saline-alkali stress (Fu et al., 2025; Liu et al., 2022). Polyhydroxyalkanoate production has been reported within Tistrellales, suggesting that this group may also respond to changes in carbon availability under different resource conditions (Shi et al., 2002). However, the ecological responses of bacterial taxa may vary across soil types, resource conditions, and experimental durations (Delgado-Baquerizo et al., 2016; Tong et al., 2025). Therefore, the enrichment of these taxa in the present study more likely reflects their adaptation to salt and nutrient gradients, rather than directly indicating a role in promoting crop growth or alleviating salt stress. In contrast, Sphingomonas appeared to be more closely associated with rhizosphere environments with lower salinity and higher nutrient availability, suggesting that it may be sensitive to improvements in soil conditions (Li and Li, 2022; Li et al., 2025). Some Sphingomonas strains possess plant-growth-promoting traits, including auxin and siderophore production and 1-aminocyclopropane-1-carboxylate deaminase activity, which may promote root development and enhance plant responses to environmental stress (Mazoyon et al., 2023). These reported traits provide a possible ecological context for the association of Sphingomonas with lower salinity and higher AP in the rhizosphere. Considering the relationships among salinity, available phosphorus, and maize yield, the alleviation of salt stress and improvement in nutrient availability may be important factors influencing rhizosphere conditions and crop performance. Changes in the bacterial community were not random but appeared to occur as the rhizosphere environment shifted from salinity limitation toward improved resource availability (Li et al., 2021; Xun et al., 2024). The reductions in salinity and improvements in nutrient availability under crushed straw and granulated straw treatments may have created more favorable conditions for the enrichment of certain bacterial taxa and were associated with better crop growth performance (Cong et al., 2025; Yu et al., 2024). In bulk soil (Figure 6b), the strong positive correlation between Pontibacter and POC suggests that its distribution may be closely related to particulate carbon accumulation. Some Pontibacter strains are adapted to saline environments, and the application of a Pontibacter actiniarum-based seed coating has been associated with lower soil EC, higher AP, and improved maize yield in saline soil (Huo et al., 2024). Together with its positive correlation with POC, these reported characteristics suggest that Pontibacter may respond to carbon inputs and saline soil conditions. In the absence of continuous root inputs and regulation, nitrogen status and organic carbon availability may exert more direct effects on bacterial community composition (Xu et al., 2021; Hansen et al., 2026). Although several bacterial taxa showed trends consistent with maize yield, these relationships were not significant after FDR correction and should not be regarded as stable indicators of crop productivity. More reasonably, these taxa may reflect the combined effects of soil nutrient transformation, substrate utilization, and changes in the root–soil environment. Their relationship with crop productivity still requires further verification using larger sample sizes or longer-term experiments (Banerjee et al., 2018; Zhang et al., 2024).

The differences in PICRUSt2-inferred functional potential suggest that bacterial responses to straw-return materials may depend not only on carbon inputs, but also on material form, local water and salt conditions, and the rhizosphere environment (Figures 7a, b). Crushed straw is relatively dispersed in soil and may be less likely to form stable localized regulatory effects in the short term. Its influence on water and salt movement and resource distribution may therefore be limited (Zhang et al., 2020). Persistent saline-alkali stress may favor bacterial taxa with stress-adaptive traits (Shrivastava and Kumar, 2015). The functional changes associated with biochar appeared to be mainly influenced by the rhizosphere environment, suggesting that root activity may enhance its regulatory effects on bacterial communities. Root exudate inputs, nutrient transformation, and microbial activity together create a more active rhizosphere environment, in which the porous structure and adsorption properties of biochar, as well as its effects on local physicochemical conditions, may be more readily expressed. In soils lacking continuous root inputs, however, these effects may be relatively limited (Zhang et al., 2020). The effect of granulated straw may be related to its pelletized structure. Compared with crushed straw, granulated materials differ in their spatial distribution in soil and in their contact with soil particles. These characteristics may be more favorable for maintaining relatively stable local conditions and reducing the influence of salt fluctuations on bacterial communities (Cong et al., 2019; Wang et al., 2026). This also suggests that material form itself may be an important factor affecting short-term soil regulation, rather than the input of straw as organic matter alone (Chen et al., 2020; Su et al., 2026).

The affected predicted functional pathways were mainly related to energy supply, central carbon metabolism, and branched-chain amino acid biosynthesis, which are closely associated with cell maintenance, osmotic regulation, and stress adaptation (Fu et al., 2025). The consistently lower predicted relative abundances of these pathways under GS indicate that granulated straw was associated with a distinct predicted functional profile in both rhizosphere and bulk soils. In the rhizosphere, the marked reductions in electrical conductivity and salt content suggest that alleviated osmotic stress may have modified environmental filtering and the relative abundances of salt-adapted bacterial taxa. This interpretation is consistent with previous evidence that soil salinity influences the recruitment of root-associated bacteria and that amendment-induced reductions in salinity can be accompanied by shifts in microbial community composition (Li et al., 2021, 2025). However, similar predicted pathway responses were also observed in bulk soil, where salinity was not reduced relative to CK, suggesting that salt alleviation alone may not explain the pattern shared by the two habitats. This contrast also points to the possible contribution of bacterial community reorganization. Because PICRUSt2 infers functional potential from ASV composition and phylogenetically inferred gene content, the predicted pathway changes may reflect both taxonomic turnover and changes in alpha diversity (Douglas et al., 2020; Yan et al., 2017; Xun et al., 2019). Under GS, alpha diversity did not change significantly in rhizosphere soil, suggesting that shifts in the relative abundances of taxa with different inferred functional repertoires may have contributed to the observed predicted pathway pattern. In bulk soil, the concurrent decline in alpha diversity may have had an additional influence by reducing some pathway-associated taxa. These community responses may also be related to the spatial distribution and gradual decomposition of granulated straw, which could modify local substrate supply and microhabitat conditions (Chen et al., 2020; Wang et al., 2026). The treatment-specific bacterial lineages identified by LEfSe are also consistent with taxonomic differentiation under GS. Overall, the predicted functional response may reflect the combined influence of changes in soil conditions and bacterial community organization, with diversity loss potentially contributing more strongly in bulk soil. The lower predicted pathway abundances should therefore be interpreted cautiously as a composition-associated redistribution of predicted functional potential rather than metabolic downregulation within individual bacterial taxa or an overall decline in bacterial functioning.

Limitations and future perspectives

This study provides field evidence for the distinct short-term effects of different straw-return forms on the soil microenvironment, bacterial communities, and maize yield in saline-alkali soil. However, the experiment was conducted during a single growing season, and the persistence of the observed responses remains to be evaluated. In addition, the PICRUSt2-based results represent functional potential inferred from bacterial community composition rather than direct measurements of metabolic activity. Future studies should extend the present field experiment across multiple growing seasons and integrate metagenomic, metatranscriptomic, or metabolomic analyses to evaluate the persistence of these responses and verify the predicted functional changes.

Different straw forms exerted distinct short-term effects on the soil microenvironment and bacterial communities in rhizosphere and bulk soils of maize grown in saline-alkali soil. Granulated straw significantly alleviated rhizosphere salt stress and increased rhizosphere available phosphorus and particulate organic carbon. Crushed straw improved rhizosphere nutrient conditions, and both crushed and granulated straw produced clear short-term yield benefits. Different forms of exogenous carbon input further induced divergence in bacterial community structure, enrichment of biomarker taxa, and shifts in predicted functional profiles in rhizosphere and bulk soils. Overall, both granulated and crushed straw improved the maize rhizosphere microenvironment and increased crop yield in the short term. Under the conditions of this short-term field experiment, granulated straw showed a relative advantage in reducing rhizosphere salinity, particularly salt content.