The study was conducted in Tianzhu, Gansu, China (37° 09′ N, 102° 99′ E), in May 2019 and 2020. The altitude of this region is approximately 2,630 m, the annual average temperature is −2°C, and the average annual precipitation is 400–450 mm. At the study location, the soil type was loam, the terrain was flat, and the soil fertility was uniformly moderate. Lettuce seeds (cv. “Yong’an lettuce—-Zilong lettuce”) were sown at Fuzhou Kexiang Seed Industry Co., Ltd., and thinned thereafter. The conventional fertilizers used in this study were urea (N ≥ 46%), calcium superphosphate (P₂O₅ ≥ 16%), and potassium sulfate (K₂O ≥ 52%). The bio-organic fertilizer used in the experiment (organic matter content ≥ 40%, Bacillus subtilis, and Pseudomonas stutzeri with effective viable count ≥ 20 million CFU/g) was purchased from Gansu Lvneng Agricultural Technology Co., Ltd. The calcium superphosphate and bio-organic fertilizers are both applied as basal fertilizers. In this study, 30% of the urea and potassium sulfate were used as basal fertilizers. Moreover, 70% of the urea and potassium sulfate were top-dressed twice; of this, 30% was applied in the seedling stage and 40% in the rhizome expansion stage.

Five treatments were used in this experiment: (i) chemical fertilizer with conventional fertilization, as used by local farmers (CK; NPK dosage: 965.2 kg ha^–1: 314.3 kg N ha^–1, 478.5 kg P ha^–1, and 172.4 kg K ha^–1); (ii) chemical fertilizer reduction by 30% + 6,000 kg ha^–1 bio-organic fertilizer (T1; NPK dosage: 675.6 kg ha^–1: 293.9 kg N ha^–1, 117.2 kg P ha^–1, and 264.5 kg K ha^–1); (iii) chemical fertilizer reduction by 30% + 9,000 kg ha^–1 bio-organic fertilizer (T2; NPK dosage: 675.6 kg ha^–1: 293.9 kg N ha^–1, 117.2 kg P ha^–1, and 264.5 kg K ha^–1); (iv) chemical fertilizer reduction by 40% + 6,000 kg ha^–1 bio-organic fertilizer (T3; NPK dosage: 579.1 kg ha^–1: 251.9 kg N ha^–1, 100.5 kg P ha^–1, and 226.7 kg K ha^–1); and (v) chemical fertilizer reduction by 40% + 9,000 kg ha^–1 bio-organic fertilizer (T4; 579.1 kg ha^–1: 251.9 kg N ha^–1, 100.5 kg P ha^–1, and 226.7 kg K ha^–1). Each treatment included three replicates, and the area of each experimental plot was 80 m². We used the ridge, double row, and semi-film cultivation modes for the experiments. The cultivation density was 57,750 plants ha^–1, the ridge width was 50 cm, the furrow width was 40 cm, and the plant spacing was 30 cm. We ensured that all field management practices (other than fertilization) were consistent across treatments.

After lettuce harvesting in August 2020, soil samples were collected from the plow layer (0–20 cm) of each replicate plot for each treatment using a stainless-steel auger. Each soil sample was a composite of five randomly collected soil cores that were pooled in a sterile plastic bag and transported to the laboratory on ice. The soil samples were sieved using a 2-mm mesh sieve and divided into two subsamples before thorough homogenization. One part was air-dried to determine the physical and chemical properties of the soil, and the other part was stored at −80°C for DNA extraction.

The soil pH and electrical conductivity (EC) were measured after shaking the soil-water suspension (1:5 wt/vol) for 30 min. The pH of the filtrate was determined using a glass electrode (PHS-3E, Shanghai Jingke, China), and the EC was measured by inserting a conductivity meter (DSJ-308A, Shanghai Jingke) into the filtrate. The SOM content was measured using a titration method after oxidation with K₂Cr₂O₇. The soil’s total nitrogen (TN), total phosphorus (TP), and total potassium (TK) were determined using the H₂SO₄-H₂O₂ wet digestion method. The TN was measured using the Kjeldahl method using a fully automatic Kjeldahl K1100F apparatus (Jinan Hanon Instruments Company, Jinan, China). The TP was measured using the Mo-Sb colorimetric method and analyzed using a UV-1780 spectrophotometer [Shimadzu Instruments (Suzhou) Co., Ltd., Suzhou, China]. The TK was analyzed using a ZEEnit 700P atomic absorption spectrometer (Analytik Jena, Germany). All samples were tested in triplicates.

The genomic DNA of the soil microbial community was extracted from 0.5 g soil samples using the E.Z.N.A.^® soil DNA Kit (Omega Bio-tek, GA, United States) according to the manufacturer’s instructions. The DNA extract was analyzed on a 1% agarose gel, and DNA concentration and purity were determined using a NanoDrop ND-2000 UV-VIS spectrophotometer (Thermo Scientific, Wilmington, DE, United States). The hypervariable V3–V4 region of the bacterial 16S rRNA gene was amplified with the primer pairs 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R (5′-GGACTACHVGGGTWTCTA AT-3′) (Chen et al., 2016) using an ABI GeneAmp 9700 PCR thermocycler (Applied Biosystems, CA, United States). The internal transcribed spacer (ITS) region of fungi was amplified using the primers ITS1F (5′-CTTGGTCATTTAGAGGAAGTAA-3′) and ITS2R (5′-GCTGCGTTCTTCATCGATGC-3′) (Blaalid et al., 2013) on the same thermal cycler. The PCR was conducted using 20 μl reaction mixtures containing the template DNA (10 ng), 4 μL of 5 × TransStart FastPfu buffer (TransGen Biotech, Beijing, China), 2 μL of 2.5 mM dNTPs, 0.8 μL each of the forward and reverse primers (5 μm), 0.4 μl of TransStart FastPfu DNA Polymerase, and ddH₂O. The PCR conditions were as follows: initial denaturation at 95°C for 3 min, 27 cycles of denaturation at 95°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 45 s, a single extension at 72°C for 10 min, and storage at 4°C. The PCRs were performed in triplicate. The PCR product was extracted from a 2% agarose gel, purified using the AxyPrep DNA gel extraction kit (Axygen Biosciences, CA, United States) according to the manufacturer’s instructions, and quantified using a Quantus Fluorometer (Promega, WA, United States).

Purified amplicons were pooled in equimolar amounts and paired-end sequenced (2 × 300 bp) on an Illumina MiSeq PE300 platform (Illumina, CA, United States) by Shanghai Majorbio Bio-pharm Technology Co., Ltd. (Shanghai, China). The raw sequences were analyzed using the Fastp software (version 0.19.6)¹ (Chen et al., 2018) for quality control. The paired-end reads were merged using the FLASH software (version 1.2.11)² (Magoč and Salzberg, 2011). The reads were dereplicated, sorted, and clustered into operational taxonomic units (OTUs) at the default 97% similarity using UPARSE (version 7.0.1090)³ (Edgar, 2013), and dechimerized against the UCHIME reference dataset (Edgar et al., 2011). Taxonomic labels were assigned to the OTUs using the Silva 16S bacterial database and the UNITE fungal database (version 7.2) (Nilsson et al., 2019) using the RDP classifier (version 2.11)⁴ with a minimum confidence threshold of 0.7.

In August 2019 and 2020, the lettuce crops were harvested after they had reached commercial standards. The yields of all plots were determined. The nitrate content was determined using the salicylic acid-sulfuric acid method (Cataldo et al., 1975). The 2,6-dichloroindophenol staining method was used to determine the vitamin C content (Arya et al., 2000), and the Coomassie Brilliant Blue method was used to estimate the levels of soluble protein (Sedmak and Grossberg, 1977). The soluble sugar content was measured using the anthrone-sulfuric acid assay (Wang et al., 2020).

Microsoft Excel 2013 was used to calculate the mean and standard error (SE) of the crop yield and quality of lettuce and the chemical properties of soil in different treatment groups. One-way analysis of variance (ANOVA) was applied to evaluate the effects of different treatments on the lettuce yield and quality and soil chemical properties. Duncan’s multiple range test (significance level: p < 0.05) was used to compare the differences in mean values among different fertilization treatments. The SPSS software (version 23.0, IBM Corp., NY, United States) was used for one-way ANOVA.

Alpha diversity indices (coverage, ACE, Chao1, Shannon index, and Simpson index) at the OTU level were calculated using the Mothur software (version 1.30.2). For the analysis of beta diversity, we used unweighted UniFrac distances to generate principal coordinate analysis (PCoA) maps to assess the differences between the bacterial community members of soil samples with different fertilization treatments (Han et al., 2021). In addition, we used PCoA based on Bray–Curtis distances to evaluate differences between the fungal community members of soil samples with different fertilization treatments. Redundancy analysis (RDA) was applied to analyze the correlation between microbial community composition and soil chemical properties (Yang et al., 2019c). The Mantel test based on Bray–Curtis distances was used to determine the relationship between microbial community composition and soil properties at the OTU level. The PCoA, heatmap, and RDA were computed using the vegan package in R (Oksanen et al., 2015).

Table 1 lists the chemical properties of soils with different fertilization treatments. Compared with the CK treatment, the four treatments (T1, T2, T3, and T4) with reduced chemical fertilizer combined with bio-organic fertilizer significantly increased the soil pH (Table 1). The SOM values of the T1, T2, T3, and T4 treatment groups were significantly higher compared to those in the CK treatment group by 61.94, 59.18, 65.60, and 71.12%, respectively. The EC in the T1, T2, T3, and T4 treatment groups was significantly lower than that in the CK group; however, the EC of T1 and T2 was significantly higher than that of T3 and T4. The T2 treatment group had the highest TN and TK levels, which were significantly higher (by 35.71 and 16.69%, respectively) than that in the CK group. There was no significant difference in TP among the T2, T3, T4, and CK treatments.

Coverage refers to the coverage of each sample library. The higher the value, the higher the probability that the sequence in the sample will be detected. This index reflects whether the sequencing results represent the actual microorganisms in the sample. The coverage indices for both bacteria and fungi exceeded 0.96 (Table 2), indicating that the sequencing results reflected the actual microorganisms in the soil samples. The higher the values of community richness indices (Chao1 and ACE), the higher the community richness. In terms of the bacterial community, the Chao1 and ACE indices in the T1–T4 treatment groups were higher (2,950.93–3,058.16 and 2,942.68–3,060.67, respectively) than those in the CK treatment group (2,859.13 and 2,864.02, respectively). In particular, the Chao1 and ACE indices of the T3 treatment group were significantly higher than those of the CK group, although there was no significant difference between the other three treatment groups and the CK treatment group (Table 3, p < 0.05). For the fungal community, the Chao1 and ACE indices of the T1–T4 treatment groups were lower (582.66–690.62 and 585.28–695.51, respectively) than those in the CK treatment group (723.01 and 708.37, respectively). In particular, the Chao1 and ACE indices of the T2 treatment group were significantly lower than those of the CK treatment group, and there was no significant difference between the other three treatment groups and the CK treatment group (Table 3, p < 0.05). We also used the Shannon and Simpson diversity indices to assess the diversity of each sample. The higher the Shannon index, the higher the microbial diversity of the sample, and the higher the Simpson index, the lower the microbial diversity of the sample. The Shannon indices of bacteria were higher in the T2 and T3 treatment groups (6.393 and 6.408, respectively) than in the CK treatment group (6.388), but the difference was not significant (Table 2). The Simpson index of bacteria was the lowest in the T2 treatment group. The Shannon indices of fungi were significantly lower in the T1–T4 treatment groups than in the CK treatment group (p < 0.05), whereas the Simpson index showed the opposite trend.

The PCoA analysis showed that the community composition of soil bacteria and fungi was altered by the reduction of chemical fertilizer combined with bio-organic fertilizer (Figure 1). For the bacterial community structure, the first and second principal coordinates explained 9.97 and 8.65% of the differences among the five treatments, respectively (Figure 1A). Moreover, the bacterial communities of the T1–T4 treatment groups were considerably separated from those of the CK treatment group (Figure 1A). The bacterial community structures of the T1 and T3 treatment groups appeared to be similar. For the fungal community structure, the first and second principal coordinates explained 36 and 17.21% of the variance, respectively (Figure 1B). The fungal communities of the T1–T4 treatment groups were clearly distinguished from those of the CK treatment group along the first principal component axis. Moreover, the fungal community structure was similar between the T3 and T4 treatment groups.

The phylum composition of the bacteria in the five treatment groups is shown in Figure 2A. The top five bacterial phyla were Actinobacteria (31.25–32.68%), Proteobacteria (17.22–19.88%), Chloroflexi (16.86–19.54%), Acidobacteria (12.65–16.01%), and Gemmatimonadota (4.31–4.61%). These five bacterial phyla accounted for more than 85% of the total sequence reads. Compared to the CK group, the relative abundance of Actinobacteria was higher in the T1, T2, T3, and T4 treatment groups. The relative abundance of Gemmatimonadota in the CK treatment was lower than those in the T1–T4 treatment groups. There was no significant difference in the relative abundance of other phyla among all treatments, except for Bacteroidetes (p < 0.05). The relative abundance of Bacteroidetes was significantly lower in the T1, T2, and T4 groups than in the CK treatment group (Figure 2C, p < 0.05).

The dominant fungal phyla in all treatment groups were Ascomycota, Mortierellomycota, and Basidiomycota, which accounted for 65.10–72.87%, 18.39–27.28%, and 3.37–4.70% of all sequence reads, respectively (Figure 2B). Overall, these three fungal phyla accounted for more than 90% of the high-quality sequences. Compared with the CK treatment group, the T2 treatment decreased the relative abundance of Ascomycota and increased that of Basidiomycota. Furthermore, the Kruskal–Wallis H-test showed that the relative abundances of Mortierellomycota, Olpidiomycota, Glomeromycota, Blastocladiomycota, and Zoopagomycota were significantly different among the five treatment groups (Figure 2D, p < 0.05).

The RDA analysis in Figure 3 shows the relationship between soil chemical properties and microbial community composition. The first two axes of the RDA analysis explained 40.46% of the total variance in the soil bacterial community (axis 1: 31.12%; axis 2: 9.34%) (Figure 3A). The TN showed the highest correlation with the soil bacterial community structure, followed by TP, SOM, TK, pH, and EC. The first two axes of the RDA analysis explained 42.50% of the total variance in the soil fungal community (axis 1: 38.95%; axis 2: 3.55%) (Figure 3B). Soil chemical properties influenced the soil fungal community composition in the following order: EC > TP > TN > SOM > pH > TK. The results of a Mantel test (Table 3) revealed that TN, SOM, pH, and TK were significantly (p < 0.05) or extremely significantly (p < 0.01) positively correlated with bacterial community composition, and that TN was significantly positively correlated with fungal community composition (p < 0.01).

Among the five fertilization treatments, the T2 treatment group had the highest yield (Figures 4A,B). The 2-year yields of the T1 and T2 treatment groups were significantly higher than that of the CK group (p < 0.05). Compared with the 2-year yield of the CK treatment group, the yield of the T2 treatment group increased by 19.43 and 16.63%, respectively. The nitrate content of lettuce was significantly lower in the T1–T4 treatment groups (by 25.25, 25.46, 32.88, and 32.44%, respectively) than in the CK group (Figure 4B, p < 0.05). However, there was no significant difference in vitamin C content among the five fertilization treatments (Figure 4C, p < 0.05). The soluble protein content of the T2 and T4 treatment groups were significantly higher (by 22.49 and 24.30%, respectively) compared to that in the CK group (Figure 4D, p < 0.05). The soluble sugar content of lettuce was higher in the T1–T4 treatment groups than in the CK group (Figure 4E). Compared to the soluble sugar content of the CK group, that of the T2 treatment increased significantly by 32.19%, and those of T1, T2, and T3 groups increased by 19.62, 22.52, and 12.29%, respectively (Figure 4F, p < 0.05).

The results of the present study demonstrated that the soil chemical properties changed significantly after the application of chemical fertilizer combined with bio-organic fertilizer. Specifically, the pH and SOM of the four treatment groups (reduced chemical fertilizer combined with bio-organic fertilizer) were significantly improved compared with the values in the control (CK) group (Table 1, p < 0.05). This was consistent with the results of previous studies showing that the application of organic fertilizer improved the soil chemical properties in rice (Wang et al., 2014). The significant increase in soil pH may be ascribed to the mitigation of soil acidification through the decomposition of organic matter (Xiaojing et al., 2018). Organic fertilizers are an important source of SOM, and SOM with a higher C content is important for maintaining soil fertility and improving soil properties (Yang et al., 2019b; Zhu et al., 2019). Pinggera et al. (2015) found that adding an appropriate amount of nitrogen facilitates the microbial decomposition of organic matter and adjusts the C/N ratio to levels more suitable for the growth of soil microorganisms. This accelerates decomposition and the release of organic nutrients by microorganisms and increases the content of SOM, which may explain the increase in SOM in the T1–T4 treatment groups in this study. In addition, the TN content was increased by a 30% reduction in chemical fertilizers combined with 6,000 or 9,000 kg ha^–1 bio-organic fertilizer (Table 1), which is consistent with the results of Yang et al. (2019b) and Wu et al. (2021b).

Previous studies on arable soil have shown that conventional fertilization with chemical fertilizers reduced bacterial richness and increased fungal richness (as measured using the Chao1 and ACE indices) (Shen et al., 2010; Zhong et al., 2010). In this study, the group with conventional chemical fertilization had the lowest Chao1 and ACE indices for bacteria and the highest indices for fungi (Table 2); thus, these results were consistent with those of the aforementioned studies. Ai et al. (2012) also reported that the application of mineral nitrogen fertilizers generally promotes the growth of fungi. In contrast, we found that the Chao1 and ACE indices were significantly lower in the T2 treatment group (30% fertilizer reduction + 9,000 kg ha^–1 bio-organic fertilizer) than in the T3 and T4 treatment groups (40% fertilizer reduction + 6,000 or 9,000 kg ha^–1 bio-organic fertilizer, p < 0.05). This may be because among the T1–T4 treatment groups, the T2 group had the highest pH, and studies have shown that within a pH range of 4–8.3, lower pH promotes the growth of fungi (Rousk et al., 2009). There was no significant change in the Shannon diversity index of bacteria in the treatment groups (p < 0.05), indicating that the bacteria showed adaptability to the soil under different fertilization systems. Luo et al. (2015) discovered that compared to inorganic fertilizers, organic fertilizers are more conducive to increasing fungal diversity. However, in this study, the Shannon diversity index of fungi was reduced by a treatment with reduced chemical fertilizers combined with bio-organic fertilizer, and the Simpson index showed the opposite trend. Studies using culture-dependent methods have shown that long-term organic fertilization—especially with manure—can reduce fungal diversity (He et al., 2008). These contrasting results are likely due to differences in experimental methods and soil types. Moreover, the PCoA analysis revealed that the bacterial and fungal community compositions were clearly separated between the control and treatment groups. Similar results have also been reported by Yu et al. (2013) and Ding et al. (2017).

Inorganic or organic fertilization leads to the enrichment of specific bacteria or fungi that can effectively utilize these nutrients, thereby altering the composition of the microbial community. Elucidating soil microbial taxa at the phylum level can reveal the ecological consistency of a microbial community (Guo et al., 2018). In this study, high-throughput sequencing showed that the dominant bacterial phyla were Actinobacteria, Proteobacteria, Chloroflexi, Acidobacteria, and Gemmatimonadota (Figure 2A), which is consistent with previous studies on agricultural soils (Chen et al., 2016; Guo et al., 2018; Wu et al., 2021b). Actinobacteria are eutrophic bacteria that are considered an indicator taxon and play a key role in the soil carbon and nitrogen cycle (Zhu et al., 2019). The present study showed that the relative abundance of Actinobacteria was increased in the four treatment groups (T1–4). Wang et al. (2017a) reported that the relative abundance of Bacteroidetes increases after 10 years of inorganic fertilization. Similarly, in this study, the relative abundance of Bacteroidetes was significantly higher in the control group (conventional chemical fertilization) than in the combined fertilization groups (reduced chemical fertilizer and bio-organic fertilizer). Consistent with previous studies, Ascomycota was the most abundant fungal phylum in all fertilization treatments (Chen et al., 2016; Bei et al., 2018). Ascomycota and Basidiomycota are typical saprophytes that are environmentally friendly, decompose organic matter (cellulose, lignin, and pectin), and play an important role in degrading litter with high lignin content in the soil (Yelle et al., 2008; Guo et al., 2018; Liu et al., 2019). Compared with the CK treatment group, the T2 group (30% reduction of chemical fertilizer combined with 6,000 kg ha^–1 bio-organic fertilizer) had a lower relative abundance of Ascomycota and increased relative abundance of Basidiomycota (Figure 2B). This may be due to a competitive relationship between the Ascomycota and Basidiomycota microbes that results in a negative correlation between the relative abundances of the two phyla (Ye et al., 2020a). Several studies have shown that compared with bacteria, fungi are more sensitive to invasive plants (Xiao et al., 2014), salt stress (Chowdhury et al., 2011), soil moisture (Kaisermann et al., 2015), fertilizer compost (Shen et al., 2019) and other factors. Yang et al. (2020b) demonstrated that compared to soil bacteria, soil fungi are more sensitive to the application of organic fertilizer (pig manure). Furthermore, the Kruskal–Wallis H- test also showed that the only bacterial phylum with significant differences among the five treatments was Bacteroidetes, whereas the other fungal phyla—including Mortierellomycota, Olpidiomycota, Glomeromycota, Blastocladiomycota, and Zoopagomycota—were significantly different between groups (Figure 2, p < 0.05).

TN is significantly associated with ammonia-oxidizing bacterial communities and is one of the main factors promoting changes in the soil microbial community (Qu et al., 2016; Yang et al., 2017). Actinobacteria is a dominant taxon in the bacterial community and is positively correlated with the soil nitrogen pool; therefore, TN has been proven to drive changes in the soil bacterial community (Nemergut et al., 2010). In this study, TN was found to be the main factor driving changes in the bacterial community (Figure 3A). pH is also considered the most significant factor determining the composition of the soil bacterial community (Zhao et al., 2015; Kim et al., 2016), and the results of the Mantel test showed that TN and pH were significantly correlated with soil bacterial community composition (Table 3, p < 0.05). Fungal community composition is affected by soil nutrient content (i.e., TN and TP) (Paungfoo-Lonhienne et al., 2015; He et al., 2016). Accordingly, the main driving factors of the fungal community were EC, TP, and TN in this study (Figure 3B). The results of the Mantel test showed that TN was significantly positively correlated with the fungal community composition (Table 3). Soil nutrients may indirectly affect fungal community composition by regulating changes in the aboveground portions of plants (Haiyan et al., 2016). Previous studies have demonstrated that SOM is the primary driver factor regulating soil fungal community composition (Liu et al., 2015). This is inconsistent with the results of this experiment and suggests that the mineralization of SOM by organic fertilizers is a long-term process. This makes it difficult to assess the full effect of their nutritional potential on the soil microbial community in a short-term study period (Zhang et al., 2021). Thus, the short-term effects of organic fertilizers on soil microbial communities appear to be more modest than those of chemical nitrogen fertilizers.

Excessive chemical fertilization is related to several problems, including high levels of N–P–K in vegetable production and the safety of vegetable products in the market. Bio-organic fertilizers are a potential alternative to mineral fertilizers in crop production that can address these urgent problems (Naeem et al., 2006; Flores-Félix et al., 2013). The application of bio-organic fertilizers to improve plant nutrition can help reduce mineral fertilization and is becoming a popular strategy for sustainable agriculture. Bio-organic fertilizers are a new type of organic fertilizer that contains various beneficial microbial communities. These unique microbial communities can activate the soil, improve its physical and chemical properties, and increase soil microbial diversity (Bandyopadhyay et al., 2010; Xue et al., 2010). However, reducing the use of chemical fertilizers in agriculture without causing loss of productivity is a feasible—but difficult—challenge. Bio-organic fertilizers can replace 23–52% of nitrogen fertilizers without causing a loss of yield (Rose et al., 2014). In this study, we found that a 30% or 40% reduction in chemical fertilizers combined with 6,000 or 9,000 kg ha^–1 bio-organic fertilizer did not cause a significant loss of yield, and the yields of the T1 and T2 treatment groups were higher than that of the control groups with conventional chemical fertilization (Figures 4A,B). Thus, our results reveal the potential role of bio-organic fertilizers in reducing the chemical fertilizer load in soil. Similar results have also been reported in cabbages (Sarkar and Rakshit, 2021), cucumbers (Zhao et al., 2018), tomatoes (Ye et al., 2020b), peppers (Wu et al., 2015), and other crops, and the use of bio-organic fertilizers has also been shown to increase crop yield.

Human diets containing vegetables with a high nitrate content can cause serious health hazards. Excessive application of mineral fertilizers—especially N fertilizers—is the main reason for high levels of nitrate accumulation in plant tissues (Wang et al., 2008). In the present study, the nitrate content of the T1–4 treatment groups was significantly lower (by 25.25–32.88%) than that of the CK group. This may be because the improvement in soil microorganisms reduces the nitrate content by increasing the activity of nitrate reductase, thus improving the safety and quality of vegetables (Chatterjee et al., 2012). Qi et al. (2021) found that the levels of soluble sugars and vitamin C increased in “yellow rose” cabbage treated with bio-organic fertilizer. In this study, we found that in lettuce, the levels of soluble sugars, soluble protein, and vitamin C increased to varying degrees in the T1–4 treatment groups (Figure 4). Similar results have also been reported in tomatoes (Brunetti et al., 2019) and cauliflowers (Yang et al., 2020a). Based on these findings, we speculate that the application of reduced chemical fertilizer combined with bio-organic fertilizer can promote the yield and quality of lettuce by improving soil chemical properties and microbial communities (Figure 5).