The experiment was conducted from July 18 to September 25, 2024, in a greenhouse located in Lanling County, Shandong Province, China (34 °52′N, 117 °56′E; elevation 44 m). The experimental timeline included three key stages: before greenhouse solarization (July 20), after greenhouse solarization (August 20), and post-transplantation (September 20). The greenhouse solarization period lasted from July 20 to August 10. After August 10, the plastic film and greenhouse covering used for greenhouse solarization were removed to allow ventilation. Organic fertilizer was applied on August 20, followed by cucumber transplantation on August 25. September 24 marked 30 days after transplantation. Grafted cucumber was used as the experimental crop, with rootstock variety Zhenxinxin 431 (RSM NO.431) and scion variety Boxin 10-2. The planting density was 3,700 plants per mu (approximately 55,500 plants per hectare), using a wide-narrow row planting method with row spacing of 80 cm and 40 cm, and a plant spacing of 30 cm. Drip irrigation was used, and fruit harvesting was carried out during the cucumber harvesting stage.

The experiment was conducted using a randomized complete block design with four treatments based on whether greenhouse solarization was applied and the type of organic amendment used after disinfection, with four replicates per treatment. Nutrient input levels for each treatment are detailed in Supplementary Table S1. The treatments included: (1) Only greenhouse solarization (CK), where no calcium cyanamide disinfection was applied and 33.3 t·ha⁻¹ of chicken manure was used as basal fertilizer; (2) Famers’ Practice (FP), with calcium cyanamide applied via irrigation before greenhouse solarization and organic fertilizer input identical to CK; (3) bio-organic fertilizer treatment (BF), which involved greenhouse solarization as in FP but replaced chicken manure with 6 t·ha⁻¹ of bio-organic fertilizer and 750 kg·ha⁻¹ of kaolin clay; and (4) organic material treatment (OM), where greenhouse solarization was also conducted as in FP, with organic materials including 6.3 t·ha⁻¹ biochar, 2.4 t·ha⁻¹ peat, and 1.5 t·ha⁻¹ fungal residue, and soil conditioner applied as in BF. The calcium cyanamide, sourced from Shandong Jiuzhouyuan Biotechnology Co., Ltd., was applied at 1,410 kg·ha⁻¹in FP, BF, and OM. Before greenhouse solarization, diluted calcium cyanamide was applied through drip irrigation at 120 mm to the FP, BF, and OM plots. Following irrigation, the soil was covered with plastic film and the greenhouse was sealed to maintain high temperature conditions for 20 days.

Statistical analyses were performed using IBM SPSS Statistics 24.0. One-way analysis of variance (ANOVA) was conducted to assess the effects of different treatments, with significance set at p < 0.05. Heatmaps were generated using Chiplot, while bar charts were created with GraphPad Prism (v8.0.2). Network diagrams were constructed using Cytoscape (v3.8.0) (Smoot et al., 2011), and other visualizations were produced using R software (v4.4.1) and its packages.

To investigate the effects of FP, BF, and OM treatments on greenhouse cucumber, yield, nitrogen uptake, dry weight, and nutrient uptake of potassium, calcium, and magnesium were measured (Figure 1). The results showed that compared to the control, yields in FP, BF, and OM treatments increased by 1.20, 1.53, and 1.33 times, respectively. The yield under BF treatment was significantly higher than that of CK, FP, and OM treatments, while no significant difference was observed between FP and OM (Figure 1A). Regarding fruit dry weight, FP, BF, and OM increased it by 22.7, 48.7, and 25.3%, respectively, compared to CK (Figure 1C), which showed a significant increase under BF treatment. In terms of nutrient quality, nitrogen uptake was highest in the BF treatment at 184.3 kg·ha⁻¹, whereas no significant differences were found between FP (151.7) and OM (160.2) (Figure 1B). Furthermore, nutrient yield, which integrates crop yield, nutrient concentration in fruits, and dietary recommendations, is an important indicator for evaluating crop quality. Compared to the CK group, FP, BF, and OM significantly increased the nutritional yield of potassium, calcium, and magnesium in cucumbers, with BF showing the optimal nutritional yield. There were significant differences between BF and both FP and OM, while no significant difference was observed between FP and OM (Figures 1D–F). In summary, different fertilizer treatments effectively improved yield, fruit dry weight, and nutrient yield of greenhouse cucumber, with BF showing the best performance.

Soil DIN, including ammonium nitrogen (NH₄⁺) and nitrate nitrogen (NO₃⁻), not only influences nutrient supply for cucumber but also affects the soil environment. The content of DIN in the soil profile from 0 to 200 cm under different treatments showed an overall decreasing trend across all four treatment groups. Notably, the CK and FP treatments exhibited significantly higher DIN contents in the 0–60 cm soil layer compared to BF and OM, with this difference being especially pronounced in the 0–20 cm layer (90.8, 100.4, 60.2, and 40.1 mg N kg⁻¹, respectively) (Figure 2). As soil depth increased, DIN content in BF and OM treatments remained lower than that in CK and FP. The lowest value was observed in BF at 180–200 cm depth, with 17.6 mg N kg⁻¹. These results indicate that CK and FP treatments led to nitrogen accumulation in the topsoil, promoting varying degrees of DIN migration to deeper soil layers and resulting in nitrogen leaching losses. Overall, BF and OM treatments effectively mitigated nitrogen loss while maintaining cucumber yield and quality, with BF showing relatively better performance.

Given the significant changes in DIN observed within the 0–60 cm soil layer, we further examined the variations in basic soil properties (Figure 3). Except for available calcium, all measured parameters-including pH, soil organic carbon (SOC), total nitrogen (TN), C/N ratio, Olsen-P, available potassium (Ac-K), and available magnesium (Ac-Mg), showed significant differences in the 0–20 cm soil layer. BF and OM significantly increased soil pH at 0–20 cm (Figure 3A), with no notable differences at deeper layers. For SOC, the FP, BF and OM treatment significantly decreased its content in the 0–20 cm layer compared with CK (Figure 3B). Although TN content in FP was higher than that in BF and OM at 0–20 cm (Figure 3C), the C/N ratio was significantly higher in BF and OM compared to FP and CK (Figure 3D). For DIN in the 0–60 cm layer, the highest values were found in FP, followed by CK, BF, and OM (Figure 3E). Similar to DIN, available phosphorus (Olsen-P) was significantly lower in BF and OM compared to CK and FP at 0–20 cm (Figure 3F). Ac-K content was highest in FP across the 0–40 cm layer, with significantly lower values in BF and OM (Figure 3G). Available calcium (Ac-Ca) showed significant differences only at 0–40 cm, where FP had the highest levels, followed by CK (Figure 3H). For Ac-Mg, FP consistently exhibited the highest content across the 0–60 cm layer, while OM showed the lowest (Figure 3I). In summary, the FP treatment led to the accumulation of most soil nutrients in the topsoil (0–20 cm), whereas BF and OM effectively reduced their excessive accumulation and potential waste.

Microbial activity is closely linked to crop performance and environmental conditions. To evaluate the impact of fertilization treatments on microbial community dynamics, 16S rRNA sequencing was performed on soil bacterial communities from the 0–20 cm layer (Figure 4). Based on a 97% OTU sequence similarity threshold, a total of 983,771 high-quality sequences were obtained. The bacterial communities included 47 phyla, 123 classes, 357 orders, 617 families, and 1,227 genera. The bacterial Shannon rarefaction curves plateaued toward the end, indicating sufficient sequencing depth (Supplementary Figures S1A,B), and the coverage of all samples was close to 1. In terms of α-diversity, both Chao and Shannon indices revealed that bacterial diversity was highest under the BF treatment, followed by OM (Figures 4A,B), indicating that optimized fertilizer applications significantly enhanced soil bacterial diversity. Principal component analysis (PCA) further revealed clear separations among CK, FP, and BF/OM treatments (Figure 4C), with BF and OM clustering closely together, suggesting significant differences in bacterial communities between optimized and conventional treatments. PLS-DA analysis supported this finding, demonstrating that fertilization management effectively altered microbial community structures (Figure 4D).

At the phylum level, Proteobacteria, Actinobacteria, and Chloroflexi were dominant taxa, accounting for more than 50% of the total relative abundance (Figure 4E). At the genus level, the abundance of Bacillus was significantly reduced in both BF and OM treatments, with the greatest decrease observed in OM (Figure 4F). In contrast, the relative abundance of norank_c_Subgroup_6, Candidatus_Nitrososphaera, Meiothermus, and norank_o_Gaiellales markedly increased in BF, by 47.8, 62.8, 477.1, and 173%, respectively, compared to FP, becoming dominant genera. Cluster analysis of the top 50 most abundant genera revealed two distinct groups: one dominated by BF and OM, and the other by CK and FP, further confirming that nutrient management practices significantly altered the composition of the soil microbial community.

To further investigate microbial community differences among treatments, LEfSe was employed to identify characteristic bacterial taxa associated with each fertilization regime (Figure 5A). In the CK group, Actinomadura, Lysobacter, Microvirga, and Streptomyces were identified as characteristic genera (Figure 5A). At the family level, Nocardioidaceae and the order Streptomycetales had the greatest impact on group differentiation, with LDA scores of 4.14 and 3.76, respectively (Figure 5B). In the FP group, Bacillus, Chitinophaga, and Devosia were the dominant genera, with enrichment further supported by the class Bacilli and the order Bacillales (Figure 5B).

In the BF treatment, norank_c__Subgroup_6 and norank_o__Gaiellales were identified as key indicator taxa, with notable enrichment of the corresponding families (f_norank_c__Gaiellales, LDA = 4.03; f_norank_c__Subgroup_6, LDA = 3.96). For OM, Meiothermus was identified as the dominant indicator microorganism, with the highest LDA score (4.26), suggesting a strong influence on community structure. To visually assess the relative abundance of indicator taxa across treatments, Kruskal-Wallis rank sum tests were performed at the phylum and genus levels (Figure 5C). At the phylum level, Firmicutes were significantly enriched in the FP group (p = 0.008), while Acidobacteria and Chloroflexi were dominant in BF (p = 0.040) and OM (p = 0.008), respectively. At the genus level, Bacillus (p = 0.007) and Sporosarcina (p = 0.008) were significantly more abundant in FP, whereas norank_c__Subgroup_6 (p = 0.049), Gaiella (p = 0.008), and Meiothermus (p = 0.008) were dominant in BF and OM (Figure 5D).

Redundancy analysis (RDA), canonical correspondence analysis (CCA), and correlation analysis were performed to evaluate the relationships between soil properties and microbial communities in the 0–20 cm layer under each treatment (Figure 6). In the CCA plot (Figure 6A), available Mg (Ac-Mg), available K (Ac-K), available Ca (Ac-Ca), mineral nitrogen (Nmin), Olsen-P, and total nitrogen (TN) exerted strong positive influences on the microbial communities of the CK and FP treatments, with most of these soil properties showing the highest correlations with FP. Soil organic carbon (SOC) was not significantly correlated with any treatment, whereas pH and the C/N ratio were more closely associated with OM. RDA yielded a pattern consistent with CCA (Figure 6B), confirming that differences in soil properties induced by the various treatments shaped microbial community composition.

Correlation analysis (Figure 6C) revealed that Sphingomonas was strongly negatively correlated with pH (r = −0.85), while norank_c__Subgroup_6 was positively correlated with pH (r = 0.64). With respect to TN, significant correlations were observed for Bacillus (r = 0.72), Gaiella (r = −0.81), Candidatus Nitrososphaera (r = −0.74), and Pseudomonas (r = −0.89). The C/N ratio showed significant negative correlations with Bacillus (r = −0.71) and Sphingomonas (r = −0.78), whereas significant positive correlations were found with norank_c__Subgroup_6 (r = 0.64), Gaiella (r = 0.72), and Candidatus Nitrososphaera (r = 0.66). Similar trends were detected for mineral nitrogen and Olsen-P, both were positively correlated with Bacillus and Pseudomonas (r = 0.74–0.86) and negatively correlated with Gaiella, norank_c__Subgroup_6, and Candidatus Nitrososphaera. Collectively, the RDA/CCA and correlation analyses indicate that the BF treatment effectively mitigated soil nutrient losses and identified soil properties that drive microbial community composition.

Functional prediction of the bacterial communities (Figure 6D) showed that BF was enriched in pathways related to Metabolic pathways, Microbial metabolism in diverse environments, Phenylalanine metabolism, Nitrogen metabolism, Citrate cycle, and Carbon metabolism. These functions may enhance the transformation and utilization of carbon and nitrogen, thereby facilitating cucumber nutrient uptake and reducing residual soil nutrients.

A typical characteristic of protected vegetable production systems in China is their high cropping index, excessive fertilizer input, and severe leaching losses, all of which place significant pressure on environmental protection. In this study, optimizing fertilizer input ratios effectively improved cucumber yield. Compared to the CK treatment, the BF treatment increased yield by 53% and improved dry weight by 48.7%. This may be related to soil-borne diseases caused by excessive fertilizer application. Excessive fertilization not only exacerbates soil acidification but also creates favorable conditions for the occurrence of diseases (Wei et al., 2024; Zhang et al., 2022). In the BF and OM treatments, the incorporation of kaolinite and organic materials helped restore soil quality by reducing the input of excessive manure-based nitrogen, thereby mitigating the risk of soil-borne diseases. In addition, the study found that the optimized integrated management strategy of organic and inorganic fertilizers has a positive impact on cucumber growth. The use of slow-release organic fertilizers has been shown to enhance soil DIN concentrations and improve both grain yield and nitrogen use efficiency in rice production (Wang Y. et al., 2025). Mercer also demonstrated that under moderate irrigation and nitrogen inputs, cotton yield as well as nitrogen and water use efficiency were maximized without the need for excessive resource input (Baird et al., 2024). Furthermore, in the BF treatment, the application of commercial animal-derived organic fertilizer may have promoted nitrogen uptake in cucumber, thereby increasing yield. This effect could be attributed to enhanced microbial transformation processes and the presence of bioactive compounds (Quan et al., 2015; Chaichi et al., 2017). This study also introduced the concept of nutritional yield to evaluate the effects of different nutrient management strategies on the uptake of potassium, calcium, and magnesium in cucumber. The BF treatment significantly outperformed FP in terms of nutritional yield. By modifying the type and dosage of organic fertilizers and adding high-pH soil amendments rich in calcium, magnesium, and silicon, cucumber growth was promoted and nutrient uptake of Ca and Mg was enhanced.

In the present study, we found that the application of organic fertilizers significantly increased the concentrations of soil DIN, available phosphorus, and exchangeable potassium in surface soil. Moreover, compared to the FP treatment, the addition of bio-organic fertilizers and organic materials significantly enhanced the carbon-to-nitrogen (C/N) ratio in the topsoil, which is likely attributable to the type and amount of organic inputs. Previous studies have reported that under conventional fertilization regimes, large applications of organic manure lead to nutrient accumulation in greenhouse vegetable soils, consistent with our findings (Bai et al., 2020).

On one hand, the application of organic fertilizers increases total nutrient input, which may promote the downward movement of nitrogen. On the other hand, some studies suggest that the characteristics of organic fertilizers can also reduce nitrogen leaching losses (Tei et al., 2020; Xu et al., 2020). Our results indicate that the FP treatment led to significantly higher DIN concentrations in the 0–60 cm soil profile than treatments with bio-organic fertilizers or organic materials. During the early growth stages of cucumber, root systems are not yet fully developed, making it difficult for roots to access nitrogen below 20 cm.

Soil microorganisms interact closely with the rhizosphere environment of crops. Based on RDA/CCA analysis of bacterial community composition at the genus level and soil environmental factors, we observed strong associations between bacterial communities and soil physicochemical properties across different treatments. Previous studies have confirmed significant correlations between variations in soil nutrient content and changes in bacterial community structure, particularly with available potassium, total nitrogen, and total carbon (Han et al., 2025; Lin et al., 2025; Peng et al., 2025). In our study, post-transplantation soil nutrient status exerted a major influence on the structure of the bacterial community.

Specifically, Actinomadura, Lysobacter, Microvirga, and Streptomyces were identified as characteristic genera in the CK group, Bacillus was dominant in FP, norank_c__Subgroup_6 and Gaiella were enriched in BF, while Meiothermus was predominant in OM. Microvirga and Streptomyces have been shown to be positively correlated with consecutive years of tobacco monocropping (Li et al., 2024), and are also considered deleterious microbes contributing to yield decline in soybean (Neupane et al., 2021). The presence of these genera in non-greenhouse solarization cucumber soils may further affect soil conditions and crop growth. Norank_c__Subgroup_6 has been identified as a beneficial rhizosphere bacterium recruited by the halophyte Suaeda salsa to promote root development (Diao et al., 2024). Additionally, it has been implicated in antibiotic resistance regulation, serving as the only potential host of the sul2 gene in wheat rhizosphere soils (Guo et al., 2020). In the intercropping system of Pteris vittata L. and Sedum alfredii Hance, the application of mixed fertilizers increased the abundance of norank_c__Subgroup_6, contributing to the remediation of heavy metal-contaminated soils (Yang et al., 2025).

Regarding Gaiella, its relative abundance was increased by the addition of maize root exudates (from 2.7 to 4.7%), and was positively correlated with shoot phosphorus concentration and soil phosphatase activity, suggesting its potential role in phosphorus mobilization (Wang G. et al., 2025). Meiothermus was found to increase by 8.75% in relative abundance during summer fallow when soil solarization was combined with manure amendment, indicating its contribution to improved soil quality (Wang et al., 2023). The abundance changes of these microorganisms under different treatments mainly depend on the type of fertilizer, application ratio, and management practices. In summary, the BF treatment significantly enhanced soil bacterial diversity, increased the abundance of beneficial bacteria, and altered the microbial community structure.