Microplastics reshape rhizosphere microbial communities and nutrient cycling in tomato agroecosystems Authors: Yongshi Liang, Manying Zhang, Mingzhu Guan, Sijia Liu, Jiugeng Chen, Rongchao Yang, Yueqin Zhang Published: 2026 ABSTRACT AbstractBackgroundMicroplastics (MPs) are emerging soil contaminants increasingly recognized for their capacity to alter agroecosystem functioning; however, their impacts on horticultural crops and rhizosphere microbial processes remain insufficiently understood.ResultsWe examined the impacts of MPs, polyethylene (PE), polypropylene (PP), and polyvinyl chloride (PVC), on rhizosphere soil physicochemical properties, bacterial community structure, and functional genes associated with carbon (C), nitrogen (N), and phosphorus (P) cycling in tomato (Solanum lycopersicum L.), a major horticultural crop cultivated worldwide. Both PVC-MPs and PE-MPs significantly reduced soil ammonium nitrogen and available phosphorus contents, modifying rhizosphere elemental composition. In contrast, PP-MPs had no significant effects on these nutrients but significantly increased soil pH by 0.76% compared with the control (p < 0.05). PE-MPs exposure markedly reduced bacterial α-diversity and selectively enriching bacterial taxa involved in recalcitrant organic matter degradation and nitrogen cycling. High-throughput quantitative PCR (qPCR) revealed that MPs suppressed the degradation of labile carbon substrates, including starch, cellulose, and hemicellulose (as evidenced by downregulation of apu, manB, and xylA), as well as methane metabolism, while promoting carbon fixation and phosphorus cycling.ConclusionsOur results indicate selective reprogramming of bacterial carbon metabolism, enhancing recalcitrant carbon turnover while constraining labile carbon mineralization, thereby disrupting C-P metabolic coupling and promoting soil organic carbon accumulation. Additionally, MPs stimulated nitrogen fixation and nitrification while inhibiting denitrification, particularly under PE- and PVC-MPs treatments. These findings provide bacterial ecological insights into MPs-induced reshape rhizosphere processes and nutrient cycling dynamics in horticultural systems.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12915-026-02633-4.Keywords: Microplastics pollution, Tomato, Rhizosphere microbiome, Soil ecosystem, Nutrient cycling Background Plastics are extensively used worldwide due to their exceptional versatility, chemical stability, and durability [3, 5, 16, 31]. However, the rapid increase in plastic consumption has led to growing concerns about environmental pollution from plastic waste. Over time, plastic waste degrades into particles smaller than 5 mm, referred to as microplastics (MPs), through prolonged exposure to environmental factors such as ultraviolet (UV) radiation, hydrolysis, mechanical abrasion, and microbial activity [6, 15, 24]. As emerging environmental pollutants, MPs have been detected across diverse ecosystems, including agroecosystems [8, 18]. Huang et al. (2020) estimated that agricultural plastic films in China could release approximately 100 trillion plastic particles into the soil, with 7000–43,000 particles per kilogram of soil [14]. These findings highlight the widespread presence and potentially profound impacts of MPs on soil ecosystems, raising significant global environmental concerns. Previous studies have demonstrated that MPs can significantly alter soil physicochemical properties and nutrient availability [9, 40, 41]. For example, MPs have been reported to modify soil bulk density, porosity, and field water-holding capacity [4, 12, 28]. Moreover, even low concentrations of polyvinyl chloride MPs (PVC-MPs) and polypropylene MPs (PP-MPs) have been found to significantly reduce available phosphorus levels in soil [20, 38]. However, the effects of MPs on other essential macronutrients, such as potassium (K), calcium (Ca), and magnesium (Mg), as well as micronutrient availability in the rhizosphere, remain largely unknown. Soil microbes play a vital role in soil structure formation, nitrogen fixation, and the decomposition of organic matter and minerals [7, 17, 26]. Exposure to MPs has been reported to significantly influence the diversity, composition, and abundance of soil microbial communities [2, 21]. For instance, Fei et al. (2020) found that the addition of 5% polyvinyl chloride (PVC) and 1% or 5% polyethylene (PE) MPs to acidic loam soil led to a reduction in microbial α-diversity [10]. Changes in soil microbial community composition can further affect the cycling of key elements such as carbon (C), nitrogen (N), and phosphorus (P), thereby disrupting soil ecosystem stability [13, 34]. Moreover, exposure to 0.3% and 1% (w/w) concentrations of PET-MPs, PVC-MPs, and PLA-MPs have been shown to significantly reduce the abundance of nirS and nirK genes involved in denitrification [36]. Therefore, investigating the impacts of MPs on rhizosphere microbial communities and soil nutrient cycling is essential for understanding their broader ecological effects on soil ecosystems and plant–microbe interactions. Tomato (Solanum lycopersicum L.) is one of the most widely cultivated horticultural crops worldwide, with a total cultivation area of approximately 4.9 million hectares [9]. Tomatoes are rich in essential nutrients, including vitamins, minerals, dietary fiber, carotenoids, and phytosterols, and they contain high levels of natural antioxidants such as β-carotene, lycopene, and vitamins C and E, along with diverse bioactive phenolic compounds (e.g., quercetin, kaempferol, naringenin, lutein, and hydroxycinnamic acids) [1]. Despite its substantial nutritional value and widespread cultivation, the effects of MPs pollution on tomato growth and the underlying physiological and ecological mechanisms remain poorly understood. Although increasing evidence suggests that MPs can alter soil physicochemical properties and microbial communities, their polymer-specific effects on rhizosphere biogeochemical processes remain poorly resolved. Existing studies have largely focused on single MPs and have often examined soil chemical properties, microbial community structure, or nutrient cycling functions separately, limiting mechanistic understanding of how MPs reshape rhizosphere nutrient dynamics, particularly in horticultural cropping systems. To address these gaps, we conducted a systematic and comparative assessment of three widely used MPs, PE, PP, PVC, in the tomato (Solanum lycopersicum L.) rhizosphere. By integrating analyses of soil physicochemical properties, bioavailable elemental composition, microbial community structure, and the abundance of functional genes involved in carbon (C), nitrogen (N), and phosphorus (P) cycling, we provide a unified framework for evaluating MPs-induced alterations in rhizosphere nutrient cycling. We hypothesized that different MPs exert polymer-specific effects, leading to distinct modifications of C, N, and P cycling processes and their coupling. We further hypothesized that MPs-induced changes in functional gene abundance are closely associated with shifts in soil physicochemical characteristics and nutrient availability. Testing these hypotheses enables mechanistic insight into how distinct MPs reorganize rhizosphere microbial processes and nutrient cycling dynamics, thereby strengthening assessment of MPs-induced ecological risks in agricultural soils and informing sustainable management of horticultural production systems under increasing MPs contamination. Results Effects of different types of MPs on soil physicochemical properties To comprehensively assess the effects of different types of conventional microplastics (MPs) on soil physicochemical properties, this study selected three commonly detected MPs in polluted soils: polyethylene (PE), polypropylene (PP), and polyvinyl chloride (PVC)—and applied them to tomato cultivation. After 3 months of growth, rhizosphere soil samples were collected for physicochemical analysis. The results demonstrated that all three MPs treatments significantly increased soil pH, with PP-MPs showing the most pronounced increase (0.76% higher than the control, p < 0.05). In addition, all MPs treatments significantly enhanced soil organic matter (SOM) and nitrate nitrogen (NO3−-N) contents. PE-MPs and PP-MPs treatments markedly increased SOM, with PE inducing the greatest increase (5.17% above the control), whereas PVC-MPs significantly reduced SOM by 2.22%. All MPs treatments also significantly elevated NO3−-N content, with PVC-MPs showing the most pronounced effect (a 104.9% increase). Conversely, regarding ammonium nitrogen (NH4+-N), MPs treatments generally reduced its levels, with PVC-MPs causing the most substantial decline, decreasing by 36.74%. The effects on NO3−-N content varied by MPs type: PE-MPs increased NO3−-N by 13.11%, PP-MPs decreased it by 16.44%, and PVC-MPs showed no significant effect. Moreover, all MPs treatments reduced available phosphorus (AP) content, with PE-MPs inducing the most notable reduction of 36.88% compared to the control, while PP-MPs and PVC-MPs showed no significant changes (Table 1). These findings indicate that the impacts of MPs treatments on soil physicochemical properties are highly dependent on MPs type, with PE-MPs and PVC-MPs exerting more pronounced effects. Table 1Effects of three microplastics (polyethylene (PE), polypropylene (PP), polyvinyl chloride (PVC)) exposure on soil physicochemical propertiesSoil propertiesCKPEPPPVCpH4.42 ± 0.01 b4.44 ± 0.01 ab4.46 ± 0.02 a4.43 ± 0.01 bOrganic matter (g kg−1)146.59 ± 1.08 c154.17 ± 0.36 a149.37 ± 0.69 b143.33 ± 0.97 dNH4+-N (mg kg−1)2.93 ± 0.35 a1.96 ± 0.37 ab2.60 ± 0.63 ab1.85 ± 0.59 bNO3−-N (mg kg−1)1.92 ± 0.13 b2.17 ± 0.15 a1.61 ± 0.03 c2.12 ± 0.08 abNO2−-N (mg kg−1)0.09 ± 0.01 c0.14 ± 0.01 b0.14 ± 0.02 b0.18 ± 0.01 aAvailable P (mg kg−1)2.18 ± 0.26 a1.38 ± 0.26 b1.56 ± 0.38 ab1.56 ± 0.50 abNote: Values are presented as mean ± standard error (n = 3). One-way analysis of variance (ANOVA) followed by LSD tests was performed; different lowercase letters indicate significant differences between treatments (p < 0.05) Effects of three microplastics (polyethylene (PE), polypropylene (PP), polyvinyl chloride (PVC)) exposure on soil physicochemical properties Note: Values are presented as mean ± standard error (n = 3). One-way analysis of variance (ANOVA) followed by LSD tests was performed; different lowercase letters indicate significant differences between treatments (p < 0.05) Effects of different types of MPs on soil available element content Soil available elements not only serve as essential mineral nutrients for plant growth but also play a pivotal role in shaping soil physicochemical properties and overall fertility. To further explore the effects of different types of MPs on soil available element content, we measured the concentrations of both macronutrients and micronutrients (Table 2). The results showed that all MPs treatments increased the concentrations of available sodium (Na), magnesium (Mg), and calcium (Ca). Among them, PE treatment resulted in the greatest increase in Na (18.06% above the control), while PVC had the most pronounced effects on Mg and Ca, with increases of 37.40% and 47.83%, respectively. Similarly, only PE and PVC treatments significantly altered potassium (K) content, with PVC inducing the most substantial increase of 75.81%. Additionally, PVC treatment also led to the most notable rise in the heavy metal chromium (Cr), increasing by 19.66%. Both PP and PVC treatments significantly elevated lead (Pb) content, with PP causing the highest increase (19.08%). Conversely, all MPs treatments reduced the availability of iron (Fe), cobalt (Co), and nickel (Ni) concentrations, with the PP treatment showing the most substantial decreases. No significant changes were observed in the manganese (Mn), copper (Cu), or zinc (Zn) contents across the three MPs treatments. Altogether, these findings indicate that MPs exposure can enhance the availability of certain macronutrients and heavy metals, such as Cr and Pb, while reducing the bioavailability of specific micronutrients in tomato rhizosphere soil. Notably, PE-MPs and PVC-MPs exhibited more pronounced effects, with PVC-MPs showing the strongest overall impact. Table 2Effects of three MPs types (PE, PP, PVC) on the bioavailability of soil elementsSoil elementsCKPEPPPVCNa (mg/kg)87.01 ± 4.06 b102.72 ± 2.98 a88.37 ± 3.31 b95.89 ± 4.06 aMg (mg/kg)69.93 ± 7.32 c87.27 ± 2.19 b73.68 ± 3.24 c96.08 ± 2.29 aK (mg/kg)26.15 ± 1.88 c41.32 ± 0.70 b23.68 ± 0.84 c45.98 ± 1.63 aCa (mg/kg)318.06 ± 34.95 c426.66 ± 8.21 b350.82 ± 16.37 c470.19 ± 14.05 aFe (mg/kg)297.93 ± 6.68 a287.57 ± 7.58 a289.05 ± 1.04 a294.56 ± 6.67 aMn (mg/kg)37.09 ± 1.03 ab34.92 ± 1.55 b34.82 ± 1.05 b37.83 ± 1.07 aCo (mg/kg)1.13 ± 0.03 a1.04 ± 0.06 b1.03 ± 0.02 b1.04 ± 0.03 bNi (mg/kg)1.81 ± 0.03 a1.74 ± 0.06 ab1.67 ± 0.02 b1.75 ± 0.05 abCu (mg/kg)4.86 ± 0.01 ab4.76 ± 0.16 ab4.64 ± 0.06 b4.94 ± 0.16 aZn (mg/kg)22.85 ± 0.34 ab22.01 ± 0.84 ab21.76 ± 0.29 b23.11 ± 0.78 aCd (mg/kg)0.07 ± 0.002 c0.07 ± 0.001 BC0.07 ± 0.001 b0.08 ± 0.002 aPb (mg/kg)1.17 ± 0.01 b1.24 ± 0.04 b1.39 ± 0.03 a1.36 ± 0.06 aNote: Values are presented as means ± standard errors (n = 3). One-way ANOVA followed by LSD tests was conducted, with different lowercase letters indicating significant differences among treatments (p < 0.05) Effects of three MPs types (PE, PP, PVC) on the bioavailability of soil elements Note: Values are presented as means ± standard errors (n = 3). One-way ANOVA followed by LSD tests was conducted, with different lowercase letters indicating significant differences among treatments (p < 0.05) Effects of different types of MPs treatments on soil bacterial community diversity To evaluate the effects of different types of MPs on the diversity of the tomato rhizosphere bacterial community, we analyzed both α-diversity and β-diversity of the soil bacterial community under MPs treatments. The results showed that all three types of MPs significantly reduced the Shannon diversity index, with PE-MPs treatment showing the most pronounced effect, a 6.89% reduction compared to the control (Fig. 1a). Although the Chao1 and ACE indices did not show significant change, PE-MPs treatment notably reduced the phylogenetic diversity index by 21.24% compared to the control (Fig. 1b–d). Effects of different types of MPs treatments on bacterial α- and β-diversity. a Shannon index; b Chao1 index; c ACE index; d Phylogenetic diversity index; e Bacterial β-diversity analysis based on Bray–Curtis distance. Values are presented as means ± standard errors (n = 3). *p < 0.05; **p < 0.01 Principal coordinate analysis (PCoA) revealed a clear separation between the treatment groups and the control group, with replicates from each treatment clustering closely together. This indicates that MPs exposure significantly altered the structure of the soil bacterial community. The first principal component (PC1) accounted for 56.14% of the total variation, while the second principal component (PC2) explained 12.52%, demonstrating high experimental reproducibility and reliable data quality (Fig. 1e). In conclusion, MPs treatments significantly decreased the α-diversity of the tomato rhizosphere bacterial community and markedly reshaped the bacterial community structure, with PE-MPs exerting the most pronounced impact. Effects of different types of MPs treatments on soil bacterial community structure To further understand the impact of different types of MPs on the structure of tomato rhizosphere bacterial communities, we analyzed the taxonomic composition at both the phylum and genus levels under MPs treatments (Fig. 2). The results revealed that the top ten dominant bacterial phyla were Proteobacteria (22.47–46.48%), Actinobacteriota (8.41–11.85%), Bacteroidota (4.55–15.13%), Acidobacteriota (4.46–10.02%), Firmicutes (0.85–23.18%), Chloroflexi (2.76–7.18%), Patescibacteria (1.00–5.88%), unclassified_Bacteria (2.57–6.04%), Gemmatimonadota (1.75–4.20%), and Verrucomicrobiota (0.67–2.98%). Under MPs treatments, the relative abundances of six dominant phyla: Proteobacteria, Actinobacteriota, Acidobacteriota, Chloroflexi, Patescibacteria, and Gemmatimonadota, significantly increased compared to the control. Among them, five phyla (excluding Patescibacteria) exhibited the most pronounced increases under PE-MPs treatment. Conversely, four other dominant phyla showed significant reductions in relative abundance, with PE-MPs treatment again showing the most notable effect. Effects of different types of MPs treatments on bacterial community structure. a Relative abundance of bacterial community structure at the phylum level; b Relative abundance of bacterial community structure at the genus level At the genus level, the top ten dominant genera were unclassified_Micropepsaceae (0.06–5.75%), Rhodanobacter (0.04–5.79%), unclassified_Bacteria (2.57–6.04%), unclassified_Xanthobacteraceae (0.23–4.28%), unclassified_LWQ8 (0.06–3.55%), Sphingomonas (1.16–2.88%), Chujaibacter (0.05–3.39%), unclassified_Chitinophagaceae (0.46–3.10%), Pseudolabrys (0.04–2.88%), and unclassified_Gemmatimonadaceae (1.09–2.37%). With the exception of unclassified_Bacteria, the relative abundances of the remaining nine genera were significantly elevated under MPs treatments. To identify statistically significant bacterial biomarkers among different treatment groups, LEfSe analysis was performed across taxonomic levels (LDA threshold = 4.0) (Fig. 3). A total of 21 bacterial taxa were enriched in the control, 26 in PE-MPs treatment, 11 in PP-MPs treatment, and 4 in PVC-MPs treatment, indicating that PE-MPs had the most substantial impact on bacterial community structure. At the phylum level, Firmicutes, Bacteroidota, unclassified_Bacteria, and Verrucomicrobiota were significantly enriched in the control group. In contrast, Proteobacteria, Acidobacteriota, Actinobacteriota, and Gemmatimonadota were significantly enriched under PE-MPs treatment, while Patescibacteria was specifically enriched under PP-MPs treatment. At the genus level, Ligilactobacillus, unclassified_Bacteria, and Pseudomonas were significantly enriched in the control group. In the PE-MPs treatment group, Rhodanobacter, Chujaibacter, Pseudolabrys, unclassified_Frankiales, Granulicella, and unclassified_Chloroflexi were significantly enriched. Additionally, unclassified_LWQ8 and unclassified_Chitinophagaceae were enriched under PP-MPs treatment, while unclassified_Micropepsaceae was specifically enriched under PVC-MPs treatment. These results support and extend previous findings, highlighting the dominant role of PE-MPs in reshaping the structure of rhizosphere bacterial communities. Identification of enriched bacterial taxa under different microplastic treatments using linear discriminant analysis effect size (LEfSe). a Cladogram from LEfSe analysis (from phylum to genus). Each circle represents a taxon, and the circle’s diameter is proportional to the taxon’s relative abundance; b histogram of linear discriminant analysis (LDA) scores showing the distribution of significantly enriched taxa Effects of different types of MPs on the abundance of carbon, phosphorus, and nitrogen functional genes in tomato rhizosphere soil Microbially driven biogeochemical cycles are essential for maintaining soil ecosystem functions, and shifts in the abundance of key functional genes serve as sensitive indicators of changes in elemental cycling processes [43]. To further explore the effects of different types of MPs on these processes, we examined the abundance of functional genes associated with carbon (C), phosphorus (P), and nitrogen (N) cycling in tomato rhizosphere soil. Effects of different MPs on carbon cycling functional genes MPs treatments significantly influenced the abundance of several key genes involved in carbon fixation pathways (Fig. 4a). Notably, PE-MPs treatment markedly increased the relative abundance of aclB and korA genes associated with the reductive tricarboxylic acid (rTCA) cycle, with increases of 368.56% and 46.85%, respectively, compared to the control. Conversely, PP-MPs treatment significantly reduced korA abundance by 60.60%. In the reductive acetyl-CoA pathway, both PE-MPs and PVC-MPs treatments significantly enhanced the abundance of acsE, with PE-MPs showing the greatest increase (298.68%), while acsA remained unaffected. Additionally, pccA, a key gene in the 3-hydroxypropionate bi-cycle, exhibited the highest increase under PE-MPs treatment (173.37%). All three MPs treatments significantly decreased smtA abundance, with the most pronounced reduction observed under PVC-MPs treatment (82.49%). Effects of different MPs treatments on carbon and phosphorus cycling genes abundance. a Relative abundance of carbon fixation functional genes under different MPs treatments; b relative abundance of carbon degradation functional genes under different MPs treatments; c relative abundance of methane metabolism functional genes under different MPs treatments; d relative abundance of phosphorus cycling functional genes under different MPs treatments. Values are presented as mean ± standard error (n = 3). *p < 0.05; **p < 0.01 For carbon degradation pathways, genes associated with starch degradation (amyA, apu, and sga) exhibited widespread reductions under MPs exposure, with PP-MPs (65.27%) and PVC-MPs (79.80% and 61.99%) showing the most significant declines. Similarly, genes involved in hemicellulose degradation (manB and xylA) were notably reduced, particularly under PE-MPs (68.28%) and PVC-MPs (89.16%) treatments. MPs exposure also led to significant decreases in the cellulose degradation gene cex and the chitin degradation gene chiA, with the most substantial reductions observed under PP-MPs treatment (57.45% and 49.71%, respectively). In contrast, genes linked to lignin degradation (glx and mnp) were significantly upregulated under PE-MPs and PVC-MPs treatments, with PE-MPs inducing the most substantial increases (1401.49% and 430.84%, respectively) (Fig. 4b). Regarding methane metabolism, MPs treatments generally suppressed the abundance of mxaF and pqq-mdh, with the most pronounced reductions observed under PVC-MPs (89.93%) and PE-MPs (43.08%) treatments, respectively (Fig. 4c). Interestingly, only PE-MPs treatment significantly enhanced the abundance of pmoA, increasing by 157.23% compared to the control. In summary, MPs exposure promoted carbon fixation while inhibiting carbon degradation and methane metabolism processes. Among the treatments, PE-MPs exhibited the most pronounced effects on carbon fixation, while the impacts on carbon degradation varied by MPs type. Moreover, methane metabolism was most disrupted under PE-MPs and PVC-MPs treatments. Effects of different MPs on phosphorus cycling functional genes MPs treatments also affected the abundance of key functional genes involved in phosphorus (P) cycling (Fig. 4d). In the process of inorganic phosphate solubilization, both PE-MPs and PVC-MPs treatments significantly increased the relative abundance of the gcd gene, with PE-MPs inducing an 85.36% increase compared to the control. Additionally, only PVC-MPs treatment significantly enhanced the abundance of the pqqC gene, resulting in a 91.51% increase. For organic phosphorus mineralization, PE-MPs and PVC-MPs treatments significantly elevated the relative abundance of phnK and phoX genes, with PVC-MPs showing the most pronounced increases, 121.95% for phnK and an impressive 549.00% for phoX. These findings suggest that PVC-MPs treatment exerts a particularly strong influence on phosphorus cycling functional genes, potentially altering P availability and turnover in the rhizosphere more substantially than other MP types. Effects of different MPs on abundance of nitrogen cycling functional genes in tomato rhizosphere soil In the nitrogen (N) cycle, MPs treatments significantly affected the abundance of genes associated with key nitrogen transformation processes, including ammonification, nitrification, denitrification, and nitrogen reduction (Fig. 5). Effects of different MPs treatments on nitrogen cycling genes. a Effects of PE treatment on nitrogen cycling genes; b effects of PP treatment on nitrogen cycling genes; c effects of PVC treatment on nitrogen cycling genes. Percentage indicates the relative change compared to the control. Red represents an increase in gene abundance, while green indicates a decrease under MPs treatments. *p < 0.05; **p < 0.01 For ammonification-related genes (ureC, gdhA), most exhibited a decreasing trend under MPs exposure, with the exception of ureC. Nitrification-associated genes showed contrasting responses. The genes amoA2 and nxrA exhibited significant upregulation under PE-MPs and PVC-MPs treatments, with amoA2 under PVC-MPs exhibiting the most pronounced increase (382.24% compared to the control). In contrast, the amoB gene was consistently downregulated across all three MPs treatments, with the most substantial reduction (42.42%) observed under PVC-MPs treatment. Genes involved in denitrification (narG, nirS3, nirK3, and nirK1) showed a consistent decline in abundance across all MPs treatments. Similarly, genes participating in both dissimilatory (napA) and assimilatory (nasA) nitrogen reduction processes were significantly downregulated, with PE-MPs and PVC-MPs treatments showing the most notable reductions (73.99% and 74.08%, respectively). Interestingly, the hzsB gene, involved in anaerobic ammonium oxidation (anammox), was upregulated exclusively under PE-MPs treatment, showing a 117.18% increase compared to the control. Collectively, these findings demonstrate that PE-MPs and PVC-MPs treatments exert substantial impacts on nitrogen cycling in the rhizosphere by suppressing denitrification, as well as both dissimilatory and assimilatory nitrogen reduction processes. Relationships between soil physicochemical properties, elemental content, and the abundance of C, N, and P cycling functional genes under different MPs treatments To evaluate the relationships among soil physicochemical properties, C, N, and P cycling functional genes, and the different MPs treatments in tomato rhizosphere soil, a Mantel test analysis was conducted (Fig. 6). The results revealed that Na, Mg, K, Ca, and nitrate nitrogen (NO3−-N) significantly influenced C, N, and P cycling processes (p < 0.05) (Fig. 6). Correlation analysis between carbon, nitrogen, and phosphorus cycling genes and soil physicochemical properties and elemental content under different MPs treatments using the Mantel test. Edge width corresponds to the Mantel r value, while edge color indicates statistical significance. Pairwise correlations between variables are represented by a color gradient reflecting Pearson correlation coefficients, with purple and green indicating positive and negative relationships, respectively Moreover, the abundance of carbon degradation-related genes showed a strong correlation with ammonium nitrogen (NH4+-N) and nitrite nitrogen (NO2−-N) (p < 0.05), whereas organic matter content significantly impacted methane metabolism (Fig. 6). However, the effects of these environmental factors varied across specific functional genes involved in C, N, and P cycling pathways. Heatmap analysis further demonstrated that Na, Mg, K, Ca, and NO3−-N were positively correlated with genes involved in carbon fixation and phosphorus cycling processes, particularly korA, acsA, acsE, pccA, gcd, pqqC, phnK, and phoX. In contrast, these soil properties showed significant negative correlations with genes involved in carbon degradation (apu, sga, manB, and xylA) and methane metabolism (pqq-mdh). For nitrogen cycling, Na, Mg, K, Ca, and NO3−-N exhibited significant positive correlations with the nitrification genes amoA2 and nxrA, while showing notable negative correlations with gdhA, napA, narG, nirS1, and nirS (Fig. S2). Overall, these findings suggest the complex interactions between soil physicochemical properties, nutrient element dynamics, and bacterial functional gene abundance associated with C, N, and P cycling in response to different MPs treatments. Discussion Since the recognition of microplastics (MPs) as emerging soil contaminants [32], increasing attention has been paid to their impacts on soil physicochemical properties and microbial communities. However, mechanistic understanding of how MPs regulate rhizosphere nutrient cycling through microbe-mediated processes remains limited, particularly in horticultural systems. By integrating soil chemistry, microbial community composition, and functional gene profiles, this study provides mechanistic insight into how chemically distinct MPs reshape carbon (C) and nitrogen (N) cycling in the tomato rhizosphere. Soil microorganisms play a critical role in nutrient cycling, plant growth, and ecosystem stability [11, 27, 37]. In this study, both the Shannon index and PD-tree index significantly declined, indicating reduced microbial diversity in the tomato rhizosphere under MPs exposure. Moreover, the enrichment of Proteobacteria and Actinobacteriota under PE-MPs exposure suggests a shift toward copiotrophic taxa capable of utilizing labile carbon substrates and adapting to heterogeneous microhabitats created by MPs, aligning with previous studies [19]. MPs have been reported to influence carbon distribution within the plant-soil system by modulating the abundance of carbon cycling-related genes [39, 42]. In line with these findings, our results illustrate functional genes associated with autotrophic and anaplerotic carbon fixation pathways were enhanced, while genes encoding enzymes for polysaccharide and cellulose degradation were suppressed. This pattern indicates a functional decoupling between carbon input and decomposition processes, whereby MPs promote microbial carbon assimilation while constraining heterotrophic carbon mineralization. Such imbalance likely contributes to the accumulation of soil organic carbon observed under MPs exposure and suggests that MPs may alter the efficiency and directionality of rhizosphere carbon fluxes rather than simply increasing or decreasing total microbial activity. MPs also influenced nitrogen cycling in the rhizosphere by reshaping the abundance of key functional genes [23, 33, 35]. In our study, nitrogen fixation and nitrification-related genes (nifH, amoA2, nxrA) were upregulated under PE-MPs and PVC-MPs treatments, which suggest that MPs favor bacterial groups capable of exploiting low-competition nitrogen niches or responding to altered oxygen and substrate availability, creating microsites with enhanced aeration or localized nutrient gradients. In contrast, the suppression of denitrification-related genes (nirS1/2/3, narG) implies a shift away from anaerobic or facultatively anaerobic nitrogen loss pathways. This suppression may be linked to both physical changes in soil structure induced by MPs and shifts in bacterial community composition away from denitrifying taxa. Collectively, these changes indicate that MPs not only alter individual nitrogen transformation steps but also reconfigure the overall balance between nitrogen retention and loss in the rhizosphere, with potential consequences for plant nitrogen availability and ecosystem nitrogen use efficiency. Notably, the contrasting responses observed among PE-, PP-, and PVC-MPs highlight the type-dependent manner of MPs effects. The comparatively weaker impacts of PP-MPs suggest that not all MPs exert equivalent ecological pressure, emphasizing the need to move beyond generic “MPs effects” toward polymer-specific risk assessment. Importantly, the coordinated shifts in bacterial community structure, functional gene profiles, and soil nutrient pools observed in this study indicate that MPs disrupt the coupling between C and N cycling processes in the rhizosphere. Such decoupling may have cascading effects on nutrient availability, bacterial competition, and plant–microbe interactions, particularly in intensively managed horticultural systems. Conclusions This study systematically compared the effects of PE, PP, and PVC microplastics on soil physicochemical properties, elemental bioavailability, bacterial communities, and functional genes involved in C, N, and P cycling in tomato rhizosphere. MPs exposure significantly increased macronutrients (Na, Mg, Ca) and heavy metals (Cr, Pb) concentrations while reducing NH4+-N and available P. PE-MPs exerted the strongest effects, decreasing bacterial diversity and enriching taxa associated with carbon and nitrogen turnover. Across all MPs types, carbon fixation and nitrogen retention pathways were promoted, whereas carbon degradation and denitrification were inhibited, indicating a functional reorganization of rhizosphere nutrient cycling driven primarily by bacterially mediated mechanisms. Environmentally, such MPs-specific shifts may alter soil fertility, nutrient use efficiency, and heavy metal mobility, with potential consequences for crop productivity and agroecosystem sustainability. These findings highlight the need for integrating microbial functional traits into MPs risk assessment and management strategies in agricultural soils. Methods Experimental materials and soil sampling In this study, three types of MPs: PE, PP, and PVC were all synthesized and supplied by Fengtai Polymer Materials Co., Ltd. (Guangzhou, China). According to the manufacturer, all MPs had a purity > 99%. The polymer types of the MPs were further confirmed using Raman spectroscopy, and the corresponding spectra are shown in Additional file 1: Fig. S1. The MPs consisted of irregularly shaped particles with an average size of approximately 10 μm. Prior to use, all MPs were rinsed thoroughly with ultrapure water and air-dried at room temperature to remove potential surface contaminants. Based on field surveys of MPs contamination in agricultural soils (Huang et al., 2020), a concentration of 100 mg·kg−1 dry soil was selected to simulate realistic MPs contamination in farmland soils. The experiment consisted of four treatments: control, PE-MPs treatment, PP-MPs treatment, and PVC-MPs treatment, with each treatment including three biological replicates (each replicate consisted of six individual plant samples). The soil used in this study was collected from the agricultural farmland in Zhanjiang, Guangdong Province, China (longitude 110°41′ E, latitude 21°20′ N). Surface soil (0–20 cm) was collected, air-dried, homogenized, and passed through a 2-mm stainless steel sieve to remove stones and plant debris. The key physicochemical properties of the soil, including pH, soil organic carbon, inorganic nitrogen, and available nutrient contents, are presented in Additional file 2: Table S1. After sieving, the soil was thoroughly mixed with MPs according to the respective treatments to ensure homogeneous distribution. The tomato (Solanum lycopersicum) cultivar M82 was used in this study. Seeds were obtained from the Tomato Genetics Resource Center (TGRC), University of California, Davis. M82 is a widely used modern large-fruited tomato cultivar and is commonly employed in physiological, genetic, and soil–plant interaction studies, making it representative for assessing MPs effects on horticultural crops. Germinated seedlings were transplanted into plastic pots containing 1.5 kg of dry soil per pot. Plants were grown in a controlled climate chamber with a day/night temperature of 26/22 °C, a light intensity of 350 μmol·m−2·s−1, a 12-h/12-h light/dark cycle, and relative humidity of 45–60%. After 90 days of cultivation, rhizosphere and bulk soil samples were collected. Plants were carefully removed from the pots, and loosely adhering soil was gently shaken off. Rhizosphere soil was operationally defined as the soil tightly adhering to the root surface (< 2 mm) and was collected by gently brushing the roots with sterile brushes into sterile tubes. For each treatment, rhizosphere soil from six randomly selected plants was pooled to form one biological replicate, resulting in three biological replicates per treatment. Bulk soil samples were collected simultaneously, sieved through a sterile 2-mm mesh, and stored together with rhizosphere samples at − 80 °C for further analysis. All sampling procedures were conducted under sterile conditions to avoid contamination. Determination of soil physicochemical properties Soil organic matter and available phosphorus were quantified using the dichromate oxidation method and the molybdenum blue colorimetric method, respectively [29, 44]. For SOM analysis, 5.0 g of air-dried soil (passed through a 2-mm sieve) was digested with 50 mL of 0.8 mol·L−1 K2Cr2O7–H2SO4 at 180 °C for 5 min. After cooling, distilled water and phenanthroline indicator were added, and the solution was titrated with 0.2 mol·L−1 FeSO4 until it turned brick-red. SOM content was calculated based on the titration volume. Available phosphorus was extracted with Bray-1 solution (NH4F–HCl; soil-to-solution ratio of 1:10, w/v) by shaking at 25 °C and 200 rpm for 30 min and filtered. The filtrate was reacted with ammonium molybdate–antimony potassium tartrate reagent, and after 15–30 min of color development, absorbance was measured at 660 nm. AP concentration was calculated using a phosphorus standard curve. The concentrations of ammonium nitrogen (NH4+-N) and nitrate nitrogen (NO3−-N) were determined following the method described by Mulvaney (1996) [25], while nitrite nitrogen (NO2−-N) was analyzed according to Rayment and Lyons (2011) [30]. Bioavailable macronutrients (e.g., Na, Mg, K, Ca) were extracted using 1 M BaCl2, whereas micronutrients (e.g., Fe, Mn, Cu, Zn) were extracted following the DTPA/TEA method developed by Lindsay and Norvell (1978) [22]. Specifically, 10 g of soil was mixed with 20 mL of an extractant solution containing 0.005 M diethylenetriaminepentaacetic acid (DTPA), 0.1 M triethanolamine (TEA), and 0.01 M calcium chloride (CaCl2) in a 50-mL centrifuge tube. The mixture was shaken at 220 rpm for 2 h and filtered using quantitative filter paper with a medium filtration rate. Elemental concentrations were measured using inductively coupled plasma mass spectrometry (ICP-MS). 16S rRNA-based bacterial diversity analysis Bacterial diversity was assessed using the Illumina NovaSeq sequencing platform. The 16S rRNA gene was amplified using region-specific primers targeting the V3–V4 hypervariable regions (forward: ACTCCTACGGGAGGCAGCA; reverse: GGACTACHVGGGTWTCTAAT). Bioinformatics processing of raw sequencing data was performed on the Biomarker Cloud platform. Bacterial α-diversity was evaluated using Shannon, Chao1, and ACE indices, as well as phylogenetic diversity measures. Principal component analysis (PCA) based on Bray–Curtis distances was conducted to compare community differences across treatments. The bacterial community composition at the phylum and genus levels was visualized using stacked bar charts. Linear discriminant analysis effect size (LEfSe) was used to identify bacterial biomarkers in rhizosphere soil under different MPs treatments. Quantification of functional genes involved in carbon, nitrogen, phosphorus, and sulfur cycling A high-throughput quantitative PCR (qPCR) array was used to quantify the abundance of 71 functional genes associated with carbon, nitrogen, phosphorus, and sulfur cycling. The qPCR analysis was performed on the SmartChip Real-Time PCR System. The cycle threshold (Ct) values of individual genes in each sample were converted to relative gene copy numbers using the formula: Relative copy number = 10.(31−Ct)/(10/3) To normalize the gene abundances, the relative copy number of each functional gene was divided by the relative copy number of the 16S rRNA gene. The absolute quantification of 16S rRNA was determined using a Roche instrument, and the absolute copy numbers of other genes were derived using the equation: Relative abundance of 16S rRNA/Absolute abundance of 16S rRNA = Relative abundance of target gene/Absolute abundance of target gene. Data analysis One-way analysis of variance (ANOVA) and least significant difference (LSD) multiple comparisons were used to assess differences in soil physicochemical properties, bacterial diversity, and functional gene abundance among treatments. Statistical significance was set at p < 0.05. Mantel tests were performed to evaluate the correlation between soil physicochemical properties and the abundance of C, N, and P cycling-related functional genes. All experiments were conducted with at least three replicates. Statistical analyses were performed using SPSS 22.0 (IBM SPSS Statistics 22). Supplementary Information Additional file 1: Fig. S1 Characterization of the microplastics used in this study. Fig. S2 Heat map showing the potential relationship of functional gene abundance with soil physicochemical factors and available elements. Additional file 2: Table S1 The soil physicochemical properties. Acknowledgements We thank Changbin Wei from Public Technology Center, South Subtropical Crops Research Institute, Chinese Academy of Tropical Agricultural Sciences for technical assistance. We also acknowledge the support of the Public Technology Center, South Subtropical Crops Research Institute, Chinese Academy of Tropical Agricultural Sciences for providing access to facilities and for technical support. Abbreviations Microplastics Polyvinyl chloride Polyethylene Polypropylene Soil organic matter Available phosphorus Authors’ contributions YS Liang: Investigation, Data curation, Methodology. MY Zhang: Investigation, Data curation, Software. MZ Guan: Investigation, Validation. SJ Liu: Investigation. JG Chen: Investigation, Formal analysis. RC Yang: Validation, Funding acquisition. YQ Zhang: Project administration, Supervision, Methodology, Writing-Reviewing and Editing. All authors read and approved the final manuscript. Funding This work was funded by the National Natural Science Foundation of China (No. 32370359). Data availability Raw data are available from NCBI with BioProject accession PRJNA1455477. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Yongshi Liang and Manying Zhang contributed equally to this work. References Associated Data Supplementary Materials Additional file 1: Fig. S1 Characterization of the microplastics used in this study. Fig. S2 Heat map showing the potential relationship of functional gene abundance with soil physicochemical factors and available elements. Additional file 2: Table S1 The soil physicochemical properties. Data Availability Statement Raw data are available from NCBI with BioProject accession PRJNA1455477.