甘肃农业大学：食品科学与工程学院（文献讲义）Comparison of bacterial diversity in traditionally homemade paocai and Chinese spicy cabbage.pdf

Contents lists available at ScienceDirect Food Microbiology journal homepage: www.elsevier.com/locate/fm Comparison of bacterial diversity in traditionally homemade paocai and Chinese spicy cabbage Zhanggen Liua,b , Zhen Penga,b , Tao Huanga,b , Yangsheng Xiaoa,b , Junyi Lia,b , Mingyong Xiea,b , Tao Xionga,b,∗ a State Key Laboratory of Food Science & Technology, No. 235 Nanjing East Road, Nanchang, Jiangxi, 330047, PR China b School of Food Science & Technology, Nanchang University, No. 235 Nanjing East Road, Nanchang, Jiangxi, 330047, PR China ARTICLE INFO Keywords: Paocai Chinese spicy cabbage High-throughput sequencing Bacterial diversity ABSTRACT This study aimed to investigate bacterial diversity in paocai and Chinese spicy cabbage and compare the microbial communities using high-throughput sequencing. Bacteria representing 26 phyla, 480 genera and 338 species were observed in these Chinese fermented vegetables. Firmicutes and Proteobacteria were the main phyla observed in both paocai and Chinese spicy cabbage. Additionally, Lactobacillus, Pediococcus, Serratia, Stenotrophomonas and Weissella were the major genera observed in both paocai and Chinese spicy cabbage. Overall, the relative abundances of Lactobacillus, Pediococcus and Weissella in Chinese spicy cabbage were much higher than those in paocai, but the proportions of Stenotrophomonas and Serratia in Chinese spicy cabbage were less than those in paocai. The results showed that the composition of the microbial community in Chinese spicy cabbage was positively correlated with total titratable acidity (TA), lactic acid and acetic acid contents but was negatively correlated with salinity. In contrast, the composition of the microbial community in paocai was negatively correlated with TA, lactic acid and acetic acid contents but was positively correlated with salinity. This study provides insights into the relationship between bacterial profiles and environmental factors in Chinese spicy cabbage and paocai, and its findings will aid in guiding future research on fermented vegetables. 1. Introduction Chinese fermented vegetables are popular foods that date back to the Zhou Dynasty (Xiong et al., 2016). Similarly to other traditional fermented foods, such as sourdough, ripened cheeses and wine (BryschHerzberg and Seidel, 2015; Garofalo et al., 2015; Lhomme et al., 2016; Nalepa and Markiewicz, 2017), fermented vegetables are usually prepared under non-sterile fermentation conditions, and a number of microorganisms participate in the fermentation process. Meanwhile, the microbial composition and population will be altered by various factors, such as geographical and climatic conditions, the parameters of fermentation process, and the kinds of vegetables used for fermentation (Cao et al., 2017; Li et al., 2017a,b; Liu and Tong, 2017; Park et al., 2012). Previous studies on Chinese sauerkraut and spicy cabbage focused mainly on lactic acid bacteria. For example, bacteria from the genera Lactobacillus and Leuconostoc (Leu) were frequently detected in paocai and included Lactobacillus fermentum, L. delbrueckii, L. oris, L. paracasei, L. plantarum, L. casei, L. zeae and Leu. mesenteroides subsp. mesenteroides (Liu et al., 2017; Xiong et al., 2012). In contrast, Chinese spicy cabbage (also known as “Chinese kimchi”), which is similar to Korean kimchi, is a fermented vegetable that is made from Chinese cabbage fermented in the presence of various species (Jeong et al., 2013); this kind of fermented vegetable was predominated by bacteria from the genera Lactobacillus, Leuconostoc, and Weissella, including L. sakei, L. curvatus, Leu. citreum and Weissella confuse (Jeong et al., 2013; Lee et al., 2005). However, the results of these previous studies do not reflect the true microbial composition of paocai and Chinese spicy cabbage. Moreover, a large-scale investigation of the bacterial communities present in paocai and Chinese spicy cabbage in China has not been performed, and few studies have reported the differences in the microbial communities in these two kinds of food. Molecular tools based on 16S rRNA (16S rDNA) sequencing have been used to analyse bacterial diversity (Cho et al., 2009; Doulgeraki et al., 2012) and gather comprehensive and detailed knowledge of the microbial flora present in fermented vegetables. For example, polymerase chain reaction-denaturing gradient gel electrophoresis (PCRDGGE) has been used to study the microbial composition of suan-cai and kimchi (Lee et al., 2005). Meanwhile, random amplified polymorphic DNA-polymerase chain reaction (RAPD-PCR), repetitive https://doi.org/10.1016/j.fm.2019.02.012 Received 1 July 2018; Received in revised form 22 January 2019; Accepted 22 February 2019 ∗ Corresponding author. School of Food Science & Technology, Nanchang University, No. 235 Nanjing East Road, Nanchang, Jiangxi, 330047, PR China. E-mail address: xiongtao0907@163.com (T. Xiong). Food Microbiology 83 (2019) 141–149 Available online 22 February 2019 0740-0020/ © 2019 Elsevier Ltd. All rights reserved. T

element PCR and species-specific PCR techniques have provided extensive and detailed descriptions of the microbial structures present in fermented vegetables from the Eastern Himalayas (Nguyen et al., 2013). However, these methods cannot completely reveal the bacterial community structure and diversity in fermented vegetables due to their limitations. High-throughput sequencing technology (HTS), also known as “next generation” sequencing technology, is characterized by the ability to sequence hundreds of thousands to millions of DNA molecules in parallel at one time and with short reading lengths (Xu et al., 2017). Compared with other methods, HTS provides a comprehensive and accurate description of bacterial communities and diversity in food, and has been applied to investigations of fermented food due to its speed, accuracy, high effectiveness during analysis, lower cost, and greater comprehensiveness (Li et al., 2017a,b; Tian et al., 2017). HTS techniques have recently been used to investigate bacterial communities in 12 types of French cheese (Dugat-Bony et al., 2016) and bacterial diversity in Spain wines (Portillo and Mas, 2016), to monitor bacterial changes in traditional Jiaozi and sourdough (Li et al., 2017a,b), and to analyse the bacterial communities present in Chinese fermented vegetables (Liu and Tong, 2017). It is known that fermented vegetables are predominated by bacteria from the genus Lactobacillus. However, information regarding the bacterial communities present in Chinese paocai and spicy cabbage and the differences between them are lacking; this information is a basis for better understanding of Chinese fermented vegetables. The present investigation thus aimed to investigate and compare the bacterial communities present in paocai and Chinese spicy cabbage from different provinces in China. We collected paocai and Chinese spicy cabbage in Southwest and Northeast China, respectively, and analysed their bacterial diversity using HTS. 2. Materials and methods 2.1. Materials Forty-five different traditionally fermented vegetables, including 27 mature paocai samples from Southwest China and 18 mature Chinese spicy cabbage samples from Northeast China, were collected from households and stored at 4 °C in the laboratory. The steps in the Chinese spicy cabbage manufacturing process are as follows: first, the outer leaves and wilted leaves are removed and the remaining leaves are washed, cut into 1.5 cm-wide strips, and put into a pot and salted for 8–15 h with 2–3% salt; next, the salted cabbage is squeezed to remove excess moisture and then put into the pot again and mixed with red pepper powder (1–2%), scallion (2–4%), ginger (0.5–1%), garlic (1–2%), and fermented anchovy sauce (0.5–1.5%); finally, the prepared Chinese spicy cabbage is incubated at 10 °C or below for 10–20 or 20–30 days, respectively (Pu, 1984; Zeng, 2002). Chinese spicy cabbages are usually fermented in winter or spring in Northeast China. The paocai manufacturing process is as follows: fresh cabbages were cut into small pieces (3–8 cm), washed, drained and put into 2.5–15 L jars with seasoning mixtures that included crystal sugar (0–2%), hot red pepper (2–4%), garlic (1–3%), ginger (1–2%), Chinese prickly ash (0–1.5%), anise (0–1%), and salt (2–10%); subsequently, the jars were usually water-sealed and stored at room temperature (20–25 °C) for 4–7 days (Xiong et al., 2012). Paocai is usually fermented year-round in Southwest China. Generally, the fermentation temperature of paocai is higher than that of Chinese spicy cabbages, while the fermentation time of paocai is shorter than that of Chinese spicy cabbages. 2.2. Physicochemical parameters The pH level of the brine samples was measured with a pH meter (PHS-25; Shanghai Precision Scientific Instruments Company, China). Using titration up to pH 8.2 ± 0.2 with 0.1 N NaOH, the TA values were calculated (Chuah et al., 2016) and interpreted as grams of lactic acid per 100 mL of brine. The lactic acid and acetic acid concentrations were monitored using the method by (Xiong et al., 2014a,b). The salt concentration was measured with a salinity meter (SA287; Qingdao Tlead International Co., Ltd., China). The nitrite contents in paocai and Chinese spicy cabbage were determined using the N-(1-naphthyl)- ethylenediamine dihydrochloride spectrophotometric method (Ding et al., 2018). 2.3. DNA extraction and PCR amplification for 16S rRNA sequence Genomic DNA was extracted using the rapid bacterial genomic DNA isolation kit (Sangon Biotech, Shanghai, China). The following universal primers were used: forward primer 338F (5′-ACTCCTACGGGA GGCAGCA-3′) and reverse primer 806R (5′-GGACTACHVGGGTWTCTAAT-3′) (Li et al., 2017a,b). The following PCR conditions were used: an initial denaturation at 94 °C for 5 min, 30 cycles of denaturation at 95 °C for 30 s, annealing at 55 °C for 50 s and extension at 72 °C for 30 s, and a final extension at 72 °C for 5 min. The PCR products were visualized using 2% agarose gel electrophoresis and recovered using a gel extraction kit (Qiagen, Germany). The amplicon library was prepared using a sample preparation kit (Illumina, USA), and HTS was performed using an Illumina HiSeq platform (Beijing Novogene Bioinformation Science and Technology Co., Ltd., China). 2.4. Sequence data analysis After sequencing, the sequence reads were filtered using the Illumina software to remove low quality sequences. The sequence data was merged using FLASH software (Version 1.2.7; Magoc and Salzberg, 2011) and filtered using QIIME software (Version 1.9.1; Caporaso et al., 2010). Sequences were removed if the sequence length was shorter than 200 bp or when the average quality score was less than 25. Chimeric sequences were detected and removed using UCHIME (Edgar et al., 2011). Afterward, OTUs defined by a 97% similarity were selected using UPARSE (version 7.0.1001, http://drive5.com/uparse/; Edgar, 2013). The α- and β-diversity indices of the 16S rRNA gene sequences were analysed using Qiime software (version 1.9.1) and R software (version 2.15.3). Rarefaction curves, the Shannon diversity index, Ace indices, Chao 1 richness, and Simpson and Good's coverage index were calculated to evaluate the alpha diversity (Cao et al., 2017). Hierarchical clustering was performed to assess the similarity of the bacterial communities present among the 45 samples (Tian et al., 2017), and canonical correlation analysis (CCA) was used to analyse the relationships between the fermented vegetable communities and environmental factors (Xiao et al., 2017). Spearman's correlation analysis was used to estimate the correlations among species in terms of richness (Cao et al., 2017). The linear discriminant analysis (LDA) effect size (LEfSe) algorithm was used to analyse the different taxa found in the paocai and Chinese spicy cabbage groups (Tian et al., 2017). LDA values > 4 were considered to indicate statistical significance. Non-metric multidimensional scaling (NMDS) was used to reveal differences between groups and within groups of samples (Portillo Mdel et al., 2016). 2.5. Statistical analysis Physiochemical properties and α-diversity indices for each group were analysed and expressed as the mean ± standard deviation (implemented in SPSS Statistics 20). Significant differences in physicochemical characteristics and α-diversity indices between the Paocai and Chinese spicy cabbage groups were analysed using t-test with PRISM 7 (Graphpad Software). The parameters of CCA and Spearman's correlation were determined by R software (version 3.4.4). The results with pvalues of less than 0.05 were considered to be statistically significant. Z. Liu, et al. Food Microbiology 83 (2019) 141–149 142

Table 1 Physicochemical properties of the Paocai samples and Chinese spicy cabbage samples. Group Sample name Sampling area Sampling location Salinity (g/100 ml) pH Titratable acidity (g/100 ml) Lactic acid content (mM/L) Acetic acid content (mM/L) Nitrite Content (mg/kg) Paocai PC1 Yunnan Province Naxi 3.18 3.36 1.69 145.56 93.10 5.64 PC2 Yunnan Province Lijiang 2.28 3.32 1.70 338.21 65.55 2.05 PC3 Yunnan Province Dali 2.04 3.94 1.10 170.47 57.18 2.56 PC4 Yunnan Province Mangshi 3.21 3.85 1.70 416.89 42.54 5.12 PC5 Yunnan Province Chuxiong 3.33 4.92 0.87 130.80 142.21 6.66 PC6 Yunnan Province Xishuangbanna 1.62 3.67 1.69 160.02 85.94 4.61 PC7 Yunnan Province Simao 2.1 3.59 1.32 188.77 40.75 6.66 PC8 Yunnan Province Kunming 0.67 4.55 0.85 104.77 64.11 5.12 PC9 Yunnan Province Qujing 4.2 3.75 0.70 105.93 35.01 3.07 PC10 Yunnan Province Zhaotong 0.43 3.84 0.96 138.50 37.69 5.12 PC11 Guizhou Province Anshun 2.25 3.7 1.25 320.56 101.82 4.10 PC12 Guizhou Province Guiyang 0.32 3.64 0.85 159.60 94.82 2.56 PC13 Guizhou Province Qiannan 4.18 4 0.56 58.74 20.96 7.17 PC14 Guizhou Province Kaili 1.91 4 0.44 61.13 44.85 0.00 PC15 Guizhou Province Qiandongnan 0.24 3.87 0.47 68.36 16.44 6.15 PC16 Guizhou Province Liupanshui 2.96 4 1.22 182.60 47.41 4.10 PC17 Guizhou Province Bijie 5.62 4 0.44 149.74 28.59 5.64 PC18 Guizhou Province Zunyi 4.12 3 1.32 187.97 36.29 4.61 PC19 Sichuan Province Xichang 3.38 3.59 0.56 93.99 13.59 5.64 PC20 Sichuan Province Leshan 3.78 3.84 0.57 83.56 15.82 29.71 PC21 Sichuan Province Zigong 8.91 3.99 0.50 65.56 18.82 2.05 PC22 Sichuan Province Yibing 5.82 3.19 0.90 158.82 12.17 3.07 PC23 Sichuan Province Neijiang 7.39 3.79 0.75 130.55 35.61 7.68 PC24 Sichuan Province Ziyang 8.62 4.45 0.38 46.28 25.99 6.15 PC25 Sichuan Province Deyang 4.87 3.72 1.10 253.56 81.60 13.83 PC26 Sichuan Province Mianyang 2.85 4.26 1.05 103.52 97.66 3.57 PC27 Sichuan Province Chongqing 8.79 3.99 0.58 123.47 102.62 5.64 Mean + SDa 3.67 ± 2.5a 3.85 ± 0.40a 0.95 ± 0.43a 153.63 ± 89.20a 54.04 ± 34.69a 5.86 ± 5.39a Chinese spicy cabbage CK1 Jilin Province Hunchun 2.36 4.02 1.56 182.69 116.58 7.97 CK2 Jilin Province Hunchun 2.59 3.64 1.95 220.45 193.64 1.99 CK3 Jilin Province Hunchun 2.81 3.98 2.29 392.96 130.22 10.46 CK4 Jilin Province Hunchun 2.13 3.76 1.63 193.95 137.99 4.48 CK5 Jilin Province Hunchun 1.93 3.97 1.53 291.71 217.45 8.96 CK6 Jilin Province Tumen 2.13 3.78 1.78 241.80 177.52 6.47 CK7 Jilin Province Tumen 2.34 3.6 2.82 372.45 171.98 7.97 CK8 Jilin Province Yanji 2.48 3.86 1.29 170.14 123.21 1.49 CK9 Jilin Province Yanji 2.04 3.8 2.16 271.72 140.63 6.97 CK10 Jilin Province Yanji 2.53 3.86 2.47 357.28 172.37 3.98 CK11 Jilin Province Longjing 2.58 4.01 1.83 299.46 136.17 5.98 CK12 Jilin Province Longjing 2.25 4 1.83 298.82 147.62 3.49 CK13 Liaoning Province Dandong 1.75 3.63 2.25 329.38 194.54 0.00 CK14 Jilin Province Helong 1.87 4.28 2.42 201.32 176.16 4.98 CK15 Jilin Province Helong 2.6 3.75 2.25 313.28 152.66 1.49 CK16 Jilin Province Tumen 2.12 4.08 2.90 313.82 151.96 1.49 CK17 Jiin Province Antu 2.25 3.87 2.13 264.65 241.11 0.00 CK18 Jilin Province Antu 1.88 3.91 2.25 289.79 201.23 4.48 Mean + SDa 2.26 ± 0.30a 3.88 ± 0.17a 2.07 ± 0.44a 278.09 ± 65.67a 165.72 ± 34.30a 4.59 ± 3.13a Yunnan Province, Guizhou Province, and Sichuan Province belong to the southwest of China; Jilin Province and Liaoning Province belong to the northeast of China. PC means Paocai; CK means Chinese spicy cabbage. a Mean ± standard deviation. Z. Liu, et al. Food Microbiology 83 (2019) 141–149 143

3. Results 3.1. Physicochemical features of the brines As shown in Table 1, the salt concentration ranged from 0.24 g/ 100 mL to 8.91 g/100 mL, with an average of 3.10 g/100 mL. The pH value ranged between 3 and 4.92, with an average of 3.86. The TA ranged from 0.38 g/100 mL to 2.90 g/100 mL, with an average of 1.39 g/100 mL. The TA of paocai and Chinese spicy cabbage ranged from 0.38 g/100 mL to 1.70 g/100 mL and 1.53 g/100 mL to 2.90 g/ 100 mL, respectively. The concentrations of lactic acid and acetic acid ranged from 46.28 mM/L to 416.89 mM/L and 12.17 mM/L to 142.20 mM/L, respectively, in paocai and from 170.14 mM/L to 392.96 mM/L and 98.71 mM/L to 241.11 mM/L, respectively, in Chinese spicy cabbage. Lactic acid concentration in the samples was positively correlated (p < 0.05) with TA and acetic acid concentration and negatively correlated with salt concentration (p < 0.05; Fig. 7). The mean TA, lactic acid, and acetic acid concentrations in Chinese spicy cabbage were 2.07 ± 0.44 g/100 mL, 278.09 ± 65.67 mM/L, and 165.72 ± 34.30 mM/L, respectively, while in paocai they were lower (0.95 ± 0.43 g/100 mL, 153.63 ± 89.20 mM/L, and 54.04 ± 34.69 mM/L, respectively). As shown in Fig. 1A, the salt content was significantly higher in the paocai groups than in the Chinese spicy cabbage groups. In contrast, the titratable acidity, lactic acid level, and acetic acid level in the paocai group were significantly lower than in the Chinese spicy cabbage group (p < 0.01; Fig. 1C, D and 1E), while no significant differences in pH were observed between the paocai group and the Chinese spicy cabbage group (Fig. 1B). The nitrite concentrations in paocai and Chinese spicy cabbage were 5.86 ± 5.39 and 4.59 ± 3 .13 mg/kg, respectively (Table 1). 3.2. Comparison of diversity indices among samples A total of 3,958,347 reads (an average of 86,529 reads for the paocai samples and 90,114 reads for the Chinese spicy cabbage samples) were generated from 45 Chinese fermented vegetable samples (Tables 2 and 3). The bacterial α-diversity indices (Shannon, Chao 1, observed species, Good's coverage index, Simpson, and ACE indices) for each sample and group were calculated (Tables 2 and 3). Based on 97% similarity, the Good's coverage index for each sample was above 0.99 (Table 2), which implied that the sequencing provided adequate coverage in terms of classification diversity. As shown in Table 2, the microbial α-diversity indices showed differences among the samples. The Shannon diversity for samples from different sites are shown in Tables 2 and 3; the diversity index obtained for paocai (3.38 ± 0.71) were lower than that obtained for Chinese spicy cabbage (3.60 ± 0.49). As shown in Fig. 1F, G and 1H, there were no significant differences in the Shannon, Chao1, and observed species indices between the paocai and Chinese spicy cabbage groups, while the Simpson index was significantly lower in paocai than in Chinese spicy cabbage (Fig. 1I). 3.3. Microbial structure The sequences comprising the total reads corresponded to 26 phyla, 186 families, and 480 genera (data not shown). As shown in Fig. 2A and B, the major phyla were Firmicutes and Proteobacteria, followed by Deinococcus-Thermus, Cyanobacteria, Actinobacteria, Acidobacteria, Bacteroidetes, Fusobacteria, Spirochaetes, and Saccharibacteria. The phyla distribution percentages varied among the samples (Fig. 2A). Among the phyla, Firmicutes was the most abundant, ranging from 12.21% to 95.29% of flora in the paocai samples and from 72.04% to 95.61% in the Chinese spicy cabbage samples. In addition to Firmicutes (average abundance of 53.95% in the paocai group and 85.83% in the Chinese spicy cabbage group), Proteobacteria was another major phylum that was observed in both groups and accounted for 39.83% and 12.47% of flora in paocai and spicy cabbage, respectively (Fig. 2B). Among the samples, Lactobacillaceae was the major family observed, followed by Enterobacteriaceae and Xanthomonadaceae. Meanwhile, the phylogenetic classification of different samples showed differences between the paocai and Chinese spicy cabbage samples in terms of 10 main bacterial families, including Lactobacillaceae, Enterobacteriaceae, Xanthomonadaceae, Leuconostocaceae, Moraxellaceae, Thermaceae, Bacillaceae, Microbacteriaceae, Methylobacteriaceae, and Rubrobacteriaceae (Fig. 2B and D). The relative richness of Enterobacteriaceae and Xanthomonadaceae was lower in Chinese spicy cabbage than in paocai, while that of Lactobacillaceae was higher in Chinese spicy cabbage than in paocai. The most representative genera were Lactobacillus, Serratia, Stenotrophomonas, Pediococcus, Weissella, Acinetobacter, Thermus, Psychrobacter, Escherichia, and Methlobacterium (Fig. 2E, F, and Fig 3A). The relative abundances of Lactobacillus (67.44%), Pediococcus (13.69%) and Weissella (3.34%) were higher in Chinese spicy cabbage than in paocai (47.22%, 3.16% and 1.4%, respectively). Meanwhile, the relative abundances of Serratia (3.37%) and Stenotrophomonas (4.67%) were lower in Chinese spicy cabbage than in paocai (17.15% and 15.32%, respectively). As shown in Fig. 3B, the species L. coryniformis, P. parvulus, L. parabrevis, and L. pentosus were more abundant in the Chinese spicy cabbage group than in the paocai group. In contrast, the relative abundances of L. acetotolerans and S. marcescens were lower in the Chinese spicy cabbage group than in the paocai group. 3.4. Community comparison NMDS analysis results revealed the separation of the bacterial communities in the two groups of samples. The bacterial communities in the Chinese spicy cabbage samples were more similar to each other than to those in the paocai samples (Fig. 4), indicating that the differences in bacterial communities were less significant (p < 0.05) in the Chinese spicy cabbage group than in the paocai group. 3.5. Bacterial cluster analysis UPGMA cluster analysis of the 45 fermented vegetable samples (27 mature paocai and 18 mature Chinese spicy cabbage samples) was performed based on the identified OTUs (Fig. 5). Two different clusters at the OTU level were observed in the UPGMA tree: Group 1, PC27, and Group 2, CK (Chinese spicy cabbage) and PC (paocai) except for PC27. This finding was consistent with the results showing that the samples in PC27 were significantly different from the Chinese spicy cabbage samples and other paocai samples (Fig. 2). Group 2 showed relatively tight clustering, which suggested the presence of a high similarity among paocai samples observed at relatively low salinity (PC8, PC3, Fig. 1. Comparison of physicochemical characteristics and α diversity indices between Paocai and Chinese spicy cabbage samples. *P < 0.05, **P < 0.01. T-test P values refer to tests carried out between the Paocai and Chinese spicy cabbage groups. Z. Liu, et al. Food Microbiology 83 (2019) 141–149 144

PC5, PC10, PC6, PC13, PC4, PC18, PC26, PC7, PC14 and PC21). Meanwhile, a significant difference in microbial composition was observed in the Chinese spicy cabbage samples and the high-salinity paocai samples. UPGMA cluster analysis found a high similarity between the Chinese spicy cabbage samples and the paocai samples with the highest salinity (PC19, PC20, PC22, PC23, PC24 and PC25). These results indicated that the microbiota in the Chinese spicy cabbage samples were more similar to each other than to those present in the paocai samples, which was consistent with the results of the NMDS analysis. 3.6. Linear discriminate analysis (LDA) effect size (LEfSe) analysis To determine the differences in relative microbial abundance between the paocai and Chinese spicy cabbage groups, the LEfSe algorithm (LDA log score threshold ≥ 4) was used at the OTU level (Fig. 6). Table 2 Alpha diversity of Paocai and Chinese spicy cabbage samples in China. Sample name Sampling area Sampling location Total reads Shannon index Chao1 index Observed species Good's coverage Simpson Ace PC1 Yunnan Province Naxi 84,019 3.09 395 227 0.998 0.725 322 PC2 Yunnan Province Lijiang 77,085 2.66 224 161 0.999 0.674 190 PC3 Yunnan Province Dali 85,719 3.66 364 211 0.998 0.844 298 PC4 Yunnan Province Mangshi 85,645 2.92 355 221 0.998 0.726 273 PC5 Yunnan Province Chuxiong 79,357 3.35 321 214 0.999 0.817 275 PC6 Yunnan Province Xishuangbanna 82,361 2.93 317 228 0.998 0.722 290 PC7 Yunnan Province Simao 88,513 3.22 411 257 0.998 0.769 354 PC8 Yunnan Province Kunming 91,068 5.30 488 370 0.999 0.933 410 PC9 Yunnan Province Qujing 85,325 3.53 247 197 0.999 0.821 231 PC10 Yunnan Province Zhaotong 96,657 3.13 467 244 0.998 0.749 349 PC11 Guizhou Province Anshun 97,837 3.55 426 190 0.998 0.852 255 PC12 Guizhou Province Guiyang 90,135 4.39 673 375 0.998 0.897 469 PC13 Guizhou Province Qiannan 82,662 2.98 428 243 0.998 0.726 320 PC14 Guizhou Province Kaili 87,254 3.05 410 256 0.998 0.764 338 PC15 Guizhou Province Qiandongnan 85,457 4.26 365 239 0.998 0.904 301 PC16 Guizhou Province Liupanshui 89,840 3.20 522 324 0.998 0.664 395 PC17 Guizhou Province Bijie 91,882 3.21 632 354 0.997 0.706 511 PC18 Guizhou Province Zunyi 81,561 2.65 377 223 0.998 0.687 318 PC19 Sichuan Province Xichang 78,526 2.33 281 220 0.999 0.551 258 PC20 Sichuan Province Leshan 88,103 3.28 361 245 0.998 0.813 319 PC21 Sichuan Province Zigong 91,650 2.66 352 230 0.998 0.747 272 PC22 Sichuan Province Yibing 80,037 3.89 308 222 0.999 0.889 271 PC23 Sichuan Province Neijiang 94,431 3.01 477 288 0.998 0.756 354 PC24 Sichuan Province Ziyang 81,461 3.31 693 300 0.997 0.769 453 PC25 Sichuan Province Deyang 86,656 3.40 459 366 0.998 0.761 447 PC26 Sichuan Province Mianyang 88,960 3.06 310 189 0.999 0.795 247 PC27 Sichuan Province Chongqing 84,083 5.25 370 252 0.999 0.938 312 CK1 Jilin Province Hunchun 88,688 4.32 382 281 0.998 0.874 339 CK2 Jilin Province Hunchun 95,811 4.02 350 212 0.999 0.902 268 CK3 Jilin Province Hunchun 80,076 3.25 744 154 0.998 0.827 230 CK4 Jilin Province Hunchun 95,251 2.67 139 95 0.999 0.730 117 CK5 Jilin Province Hunchun 83,076 3.62 355 219 0.998 0.854 272 CK6 Jilin Province Tumen 94,009 3.25 425 267 0.998 0.778 343 CK7 Jilin Province Tumen 87,536 3.60 414 216 0.998 0.870 311 CK8 Jilin Province Yanji 86,364 3.72 405 221 0.999 0.878 254 CK9 Jilin Province Yanji 91,891 3.18 413 218 0.998 0.787 274 CK10 Jilin Province Yanji 96,291 3.18 207 153 0.999 0.766 191 CK11 Jilin Province Longjing 91,122 3.88 506 209 0.999 0.896 257 CK12 Jilin Province Longjing 89,421 3.06 335 255 0.999 0.764 298 CK13 Liaoning Province Dandong 94,950 3.52 395 231 0.998 0.836 298 CK14 Jilin Province Helong 91,199 4.07 477 292 0.998 0.850 385 CK15 Jilin Province Helong 86,265 3.98 427 285 0.998 0.848 346 CK16 Jilin Province Tumen 86,454 3.93 349 199 0.999 0.900 259 CK17 Jilin Province Antu 97,327 3.09 525 250 0.998 0.771 326 CK18 Jilin Province Antu 86,332 4.51 488 293 0.998 0.914 358 Yunnan Province, Guizhou Province and Sichuan Province belong to the southwest of China; Jilin Province and Liaoning Province belong to the northeast of China. PC means Paocai; CK means Chinese spicy cabbage. Table 3 Number of total reads, observed diversity richness (OTUs), estimated OTU richness (Chao 1), Shannon, observed species, and Ace of Paocai and Chinese spicy cabbage group. Group Total reads Shannon index Chao1 index Observed species Good's coverage Simpson Ace Paocai 2336284 91.27 11,033 6846 26.952 20.999 8832 Mean ± SDa (paocai) 86,529 ± 5375a 3.38 ± 0.71a 408 ± 117a 254 ± 59a 0.998 ± 0.0006a 0.778 ± 0.090a 327 ± 78a Chinese spicy cabbage 1622063 64.85 7336 4050 17.971 15.045 5126 Mean ± SDa (Chinese spicy cabbage) 90,114 ± 4905a 3.60 ± 0.49a 408 ± 127a 225 ± 53a 0.998 ± 0.0005a 0.836 ± 0.057a 284 ± 65a a SD, standard deviation. Z. Liu, et al. Food Microbiology 83 (2019) 141–149 145