Spatial Heterogeneity of Vibrio spp. in Sediments of Chinese Marginal Seas

Vibrio is an important component of natural marine microbial populations in terms of pathogenicity and roles in carbon cycling. Compared to the marine pelagic environment, our knowledge of the diversity and distribution pattern of Vibrio spp. in sediment is limited. Here, we show higher Vibrio abundance in Chinese marginal seas than in other studied sediments. There was a clear spatial differentiation of Vibrio abundance and community composition in different sea areas. The benthic Vibrio community displayed a strong distance-decay pattern across a wide spatial range, which was formed under the combined effects of spatial and environmental factors. These results provide deep insights into the ecological dynamics of Vibrio and its environmental controls, facilitating a more comprehensive understanding of the marine Vibrio ecology.

ABSTRACT

Vibrio spp. are ubiquitous marine bacteria with high metabolism flexibility and genome plasticity. Previous studies have revealed the dynamics of planktonic vibrios in relation to environmental forces, such as temperature and salinity. However, little is known about Vibrio ecology in benthic environments. Here, we elucidate the abundance, diversity, and spatial distribution of Vibrio spp. in sediments of the Chinese marginal seas, with a wide spatial range from north to south covering the Yellow Sea (YS), East China Sea (ECS), and South China Sea (SCS). Quantitative analysis showed that Vibrio spp. were most abundant in the SCS (∼9.04 × 10⁵ copies/g) compared to the YS (∼1.00 × 10⁵ copies/g) and ECS (∼8.86 × 10⁵ copies/g). Vibrio community compositions derived from Illumina sequencing of Vibrio-specific 16S rRNA genes varied significantly between sampling areas, which was reflected by a strong distance-decay pattern. The spatial distribution of Vibrio was governed by a joint effect of spatial and environmental factors (especially temperature, salinity, and SiO₃²⁻), and their respective pure effects explained only a small fraction of the community variation. Moreover, we identified the most prominent operational taxonomic units (OTUs) that were partitioned in these sea areas. Whereas Vibrionaceae OTU20 and Photobacterium lipolyticum were prevalent in the YS, Vibrio gigantis and Photobacterium piscicola, and P. piscicola, Photobacterium lutimaris, and Photobacterium alginatilyticum were prevalent in the ECS and SCS, respectively. Our study demonstrated clear spatial heterogeneity of Vibrio spp. in sediments of the Chinese marginal seas, laying a foundation for fully understanding the marine Vibrio ecology and the ecological roles of the species.

IMPORTANCE Vibrio is an important component of natural marine microbial populations in terms of pathogenicity and roles in carbon cycling. Compared to the marine pelagic environment, our knowledge of the diversity and distribution pattern of Vibrio spp. in sediment is limited. Here, we show higher Vibrio abundance in Chinese marginal seas than in other studied sediments. There was a clear spatial differentiation of Vibrio abundance and community composition in different sea areas. The benthic Vibrio community displayed a strong distance-decay pattern across a wide spatial range, which was formed under the combined effects of spatial and environmental factors. These results provide deep insights into the ecological dynamics of Vibrio and its environmental controls, facilitating a more comprehensive understanding of the marine Vibrio ecology.

INTRODUCTION

The genus Vibrio, belonging to the Gammaproteobacteria of the phylum Proteobacteria, comprises more than 120 valid species (1, 2). Vibrio spp. have several common characteristics, including a halophilic nature, a short generation time, motility by means of polar flagella, and broad metabolic capacities (3, 4). The study of Vibrio ecology has a long history, largely because several Vibrio spp. act as potential pathogens (5, 6). Many well-known Vibrio spp., such as Vibrio cholerae, V. parahaemolyticus, V. vulnificus, V. anguillarum, and V. harveyi, cause disease in humans and/or marine animals (1, 7).

Although pathogenic Vibrio strains are hazards for coastal systems (5, 8–10), most Vibrio strains are harmless (5). Vibrios are present in a great variety of aquatic environments from coastal areas to the deep sea (2, 5), indicating that they occupy a wide range of environmental niches. Importantly, vibrios represent some of the most abundant cultivable bacteria in the ocean. The ecological success of vibrios is reflected by the short multiplication time and association with plants (microalgae, macroalgae, and filamentous cyanobacteria) and animals (e.g., loligonoid squids, flashlight fish, and anglerfish) (5, 11, 12). As a consequence, Vibrio spp. can rapidly respond to nutrient pulses and become the dominant members of a total bacterial assemblage, resulting in bloom events (3, 13–15). Thus, Vibrio spp. are copiotrophic and opportunistic strategists. They have been found to utilize a wide range of organic carbon compounds, such as chitin, alginate, and agar (16). These findings suggest that vibrios may exert large impacts on biogeochemical processes in the marine ecosystem (3, 17).

Many studies have been conducted to investigate the ecology of Vibrio populations in different marine environments worldwide (3, 7, 18–20), including a coral reef ecosystem in Ishigaki, Japan (18); Venice Lagoon (Italy) (20); and subtropical coastal water of Hong Kong (21). Most earlier studies were carried out by culture-based techniques (18, 22, 23), which are problematic due to the high frequency of Vibrio bacteria entering a viable but nonculturable (VBNC) state (24, 25). More recently, the distribution pattern of Vibrio populations has been investigated by quantitative PCR (qPCR) and high-throughput sequencing on the basis of Vibrio-specific 16S rRNA gene primers (1, 7, 26). Typically, the abundance of Vibrio spp. is 10³ to 10⁶ CFU/liter and 10⁴ to 10⁸ 16S rRNA gene copies/liter in estuarine and coastal waters (3, 27). As for community composition, different Vibrio spp. may exist in various marine environments, likely driven by diverse biotic and abiotic factors (3, 5). For example, Vibrio sp. operational taxonomic unit 13800 (OTU13800) and V. mimicus were the dominant groups in the Sydney Harbor estuary (7), whereas the most prominent species in the surrounding seawater of coral reefs in Ishigaki, Japan, were V. hyugaensis, V. owensii, and V. harveyi (18).

Current knowledge on the ecology of Vibrio is mostly derived from studies of pelagic environments. Compared to the water column, sediment contains a significantly (10⁴- to 10⁵-fold) higher concentration of organic matter (28–30), which results in bacterial densities in marine sediment that are generally 3 orders of magnitude higher, especially in the upper surface layer, than in the water column (28, 31, 32). Hence, sediment may be a favorable ecological niche or act as a reservoir for the copiotrophic Vibrio spp. to thrive, and the benthic Vibrio populations may evolve ecological features different from those in the pelagic environment (26). However, few studies have investigated the Vibrio community in sediments. Vezzulli et al. (26) found that the abundances of both culturable and total Vibrio spp. were at least 1 order of magnitude higher in Mediterranean Sea sediments than in seawater (26). A study of the Vibrio community in the tidal flat sediment of the Oregon estuary focused on the oyster-pathogenic Vibrio spp., represented mainly by V. coralliilyticus (33). Moreover, diverse culturable Vibrio spp. occurred in different beach sediments (34–36), characterized by species-specific responses to environments, e.g., V. alginolyticus diminished substantially at >10°C, whereas V. vulnificus increased rapidly at >20°C (34). Thus, we hypothesize that Vibrio spp. exhibit substantial spatial heterogeneity across different benthic environments. However, no data are available on the geographic pattern of Vibrio populations in sediments.

Chinese marginal seas include the Bohai Sea (BS), Yellow Sea (YS), East China Sea (ECS), and South China Sea (SCS), which consist of 1.8 × 10⁴ km of continental coastline and have significant differences in environmental conditions (37). For example, in spring, the bottom seawater temperature ranged from 6 to 24°C, whereas the mean salinity varied from 31‰ to ≥35‰ among the YS, ECS, and SCS (37–39). Two major waterways, namely, the Changjiang River and the Pearl River, which are two of the largest rivers in the world, flow into the ECS and SCS, respectively (40). These rivers discharge a significant amount of terrigenous material into the marginal seas, contributing up to 10% of the world’s river sediment load (41). The combined effects of complex water masses, as well as ocean currents, result in distinct distribution patterns of sediments in the Chinese marginal seas (37, 39, 42, 43). Due to their complex nature, the Chinese marginal sea sediments are excellent candidates for evaluating the spatial distribution of benthic Vibrio spp. We hypothesize that sediments are favorable reservoirs for Vibrio spp. and that Vibrio spp. exhibit spatial heterogeneity across various sedimentary environments. To test this and to better understand the distribution of Vibrio spp. in the context of environmental heterogeneity, we investigated the sedimentary vibrios in the YS, ECS, and SCS off the China coast. The Vibrio abundance and community components were studied using qPCR and high-throughput sequencing and were related to a range of environmental and spatial variables.

RESULTS

Environmental parameters.

All the samples collected from the Chinese marginal seas (Fig. 1) were divided into three clusters based on geographic location and environmental parameters, which are summarized in Table 1. The bottom seawater temperature of the sites varied significantly among the three groups, i.e., YS, ECS, and SCS (P < 0.01) (Fig. 2), with the mean value increasing by 14.64°C from north to south. In addition, the salinity (32.71 ± 0.52 practical salinity units [PSU]) and SiO₃²⁻ (137.98 ± 32.39 μmol/liter) and NO₃⁻ (0.58 ± 0.67 μmol/liter) concentrations in the YS group were significantly lower than in the other two groups (Fig. 2). However, the concentrations of dissolved oxygen (DO) (5.80 ± 1.43 to 6.63 ± 0.44 mg/liter), NH₄⁺ (56.17 ± 27.22 to 96.87 ± 69.00 μmol/liter), NO₂⁻ (0.12 ± 0.20 to 0.43 ± 0.41 μmol/liter), and PO₄³⁻ (1.90 ± 1.23 to 3.69 ± 2.09 μmol/liter) showed no significant differences among the three groups (P > 0.05). Other parameters showed different patterns. For chlorophyll a (Chl a), there were differences between the YS and SCS (P < 0.05), and the lowest value was 0.25 ± 0.24 μg/liter in the YS, whereas the highest was 0.85 ± 0.42 μg/liter in the SCS. Detailed physicochemical attributes of each site in different regions are shown in Table S1 in the supplemental material. The concentrations of salinity, DO, NO₃⁻, NO₂⁻, and NH₄⁺ changed more broadly in the ECS than in the other sea areas, and several factors (e.g., PO₄³⁻) generally decreased with the decrement of latitude (see Table S1).

FIG 1.

Study area and sampling stations. Samples were collected from the YS, the ECS, and the SCS. (The map was created using Ocean Data View [version 5.1.2; R. Schlitzer, Ocean Data View, https://odv.awi.de, 2018].)

TABLE 1.

Environmental parameters in different areas

Physicochemical parameter	Valuea
	YS	ECS	SCS
Longitude (°E)	123.43 ± 0.99	122.11 ± 0.83	115.10 ± 2.94
Latitude (°N)	35.07 ± 1.01	28.18 ± 1.01	22.46 ± 1.52
Temp (°C)	8.11 ± 1.89	18.88 ± 1.02	22.75 ± 0.67
Salinity (PSU)	32.71 ± 0.52	33.83 ± 0.98	34.09 ± 0.52
Chl a (μg/liter)	-	0.43 ± 0.31	0.85 ± 0.42
DO (mg/liter)	-	5.80 ± 1.43	6.63 ± 0.44
NO3− (μmol/liter)	0.58 ± 0.67	3.12 ± 2.03	2.74 ± 1.21
NO2− (μmol/liter)	0.12 ± 0.20	0.43 ± 0.41	0.35 ± 0.11
NH4+ (μmol/liter)	65.87 ± 64.12	96.87 ± 69.00	56.17 ± 27.22
PO43− (μmol/liter)	3.69 ± 2.09	1.90 ± 1.23	2.49 ± 2.04
SiO32− (μmol/liter)	137.98 ± 32.39	230.46 ± 61.63	264.30 ± 76.72

^aStatistical differences after chi-square test. The hyphens indicate that there were too few data (<3) to calculate.

FIG 2.

Spatial distribution of significantly varied environmental parameters in the YS, ECS, and SCS. Tem, temperature; Sal, salinity.

Abundances of total Vibrio spp. in different sea areas.

Quantitative PCR was used to detect the abundance of the vibrionic 16S rRNA gene, which showed a range from 2.40 × 10⁴ (site YS05) to 1.10 × 10⁷ copies/g (site ECS08) (see Table S1). Samples from sites located in the south tended to have higher copy numbers of the vibrionic 16S rRNA gene, and the mean value increased from 1.00 × 10⁵ to 9.04 × 10⁵ copies/g (except ECS08) with the decrement of latitude (Fig. 3A). After chi-square test analysis, the Vibrio abundance (log value) in the YS was significantly lower than in the other two areas (P < 0.01), while no obvious differences were recorded between the ECS and the SCS (P > 0.05) (Fig. 3A). Unlike the result for Vibrio abundance, the abundance of total bacteria had little relationship to latitude, and the highest value (9.70 ± 0.21 log copies/g) appeared in the ECS (Fig. 3B). Also, there were significant differences (P < 0.01) (Fig. 3B) among the three areas in the abundance of total bacteria (log value).

FIG 3.

Total Vibrio and bacterial abundance (log copies per gram) determined by qPCR in the YS, ECS, and SCS. The boxes show the distribution of values. The lower and upper hinges correspond to the first and third quartiles, respectively. The middle hinges correspond to the medians. The upper and lower whiskers extend from the hinge to the largest value no further than 1.5 times the interquartile range from the upper and lower hinges, respectively. The cycle points in boxes represent mean values. The outlier points are the data beyond the end of the whiskers. The asterisks denote significant differences between areas. **, P < 0.01.

The correlations between abundance (Vibrio spp. and total bacteria) and environmental factors were calculated using Spearman’s rank correlation coefficients. Across the entire data set, Vibrio abundance was positively correlated with temperature (r = 0.610; P = 0.0001), salinity (r = 0.532; P = 0.0010), NO₃⁻ (r = 0.590; P = 0.0005), NH₄⁺ (r = 0.364; P = 0.0370), and SiO₃²⁻ (r = 0.627; P = 0.0001) and negatively correlated with longitude (r = -0.574; P = 0.0004) and latitude (r = -0.562; P = 0.0010). As for the total bacteria, they were positively correlated with NO₃⁻ (r = 0.411; P = 0.0220), NH₄⁺ (r = 0.556; P = 0.0010), and SiO₃²⁻ (r = 0.499; P = 0.0030).

Richness and diversity estimators.

In total, 1,247,075 overlapped reads, ranging from 30,818 to 43,849 in all the samples, were generated through Illumina sequencing. Of these, 945,078 reads were left after quality control (Table 2), and the read number in each sample was limited to 18,622 after rarefaction. The total sequences yielded 2,131 OTUs at a 97% sequence similarity level (Table 2). The Good’s coverage values, indicating whether the current sequences represented the majority of the bacterial community, ranged from 99.94 to 99.99% across entire samples. In addition, the Chao 1 index values (a measure of richness) were comparable to the number of OTUs in each sample, indicating that the retrieved sequences could represent most of the Vibrio community at the studied sites. The Chao 1, evenness, and Shannon index values (including both richness and evenness) were shown to be variable across sea areas (Fig. 4); they differed significantly among the YS, ECS, and SCS (Wilcoxon’s test; P < 0.01) (see Table S2 in the supplemental material) and increased gradually from the YS to the SCS (Fig. 4). In contrast, the phylogenetic diversity in the YS was comparable to that in the ECS. There were positive correlations between the Chao 1 and Shannon indices and temperature, salinity, NO₃⁻, and SiO₃²⁻ (P < 0.01) and negative correlations with spatial distance (see Table S2). The Shannon index also showed positive correlation with Chl a and DO. Both the evenness and the phylogenetic distance showed no correlation with environmental parameters (P > 0.05) (see Table S2).

TABLE 2.

Observed richness and diversity estimates of Vibrio spp. based on 97% OTU clusters

Sample identifier	No. of overlapped reads	No. of filtered reads	No. of rarefraction reads	No. of OTUs	Chao 1	Equitability	Shannon	Phylogenetic distance	Coverage (%)
YS01SS	37,011	31,971	18,622	48	48.33	0.63	2.28	7.42	99.99
YS02SS	37,054	33,453	18,622	31	32.00	0.45	2.00	3.84	99.99
YS03SS	31,141	30,146	18,622	25	26.50	0.61	1.11	3.36	99.98
YS04SS	36,394	34,930	18,622	17	17.50	0.59	1.56	7.13	99.99
YS05SS	31,268	29,118	18,622	13	13.00	0.67	1.23	6.73	99.99
YS06SS	37,105	35,034	18,622	30	30.50	0.67	1.94	1.86	99.99
YS07SS	39,652	31,301	18,622	44	45.00	0.57	2.07	4.78	99.99
ECS01SS	38,889	30,113	18,622	68	73.25	0.55	2.57	8.04	99.96
ECS02SS	36,928	26,953	18,622	61	67.43	0.65	1.82	5.57	99.95
ECS03SS	41,876	27,130	18,622	74	83.00	0.63	2.53	6.21	99.95
ECS04SS	33,354	23,907	18,622	67	71.00	0.55	2.29	8.04	99.96
ECS05SS	37,827	28,330	18,622	82	84.14	0.48	2.76	3.93	99.97
ECS06SS	35,345	30,426	18,622	61	66.00	0.50	2.41	4.83	99.97
ECS07SS	31,612	23,731	18,622	67	67.00	0.54	2.40	5.62	99.99
ECS08SS	43,030	28,567	18,622	75	81.00	0.65	2.45	5.53	99.95
ECS09SS	40,664	26,574	18,622	65	67.14	0.56	2.62	3.86	99.97
ECS10SS	37,166	25,494	18,622	62	75.75	0.55	2.23	3.85	99.94
ECS11SS	32,528	24,668	18,622	66	68.00	0.57	2.33	8.89	99.97
ECS12SS	42,990	30,232	18,622	61	64.00	0.58	2.51	4.25	99.97
ECS13SS	32,835	25,421	18,622	60	69.00	0.63	2.72	8.69	99.95
ECS14SS	31,675	25,447	18,622	44	44.17	0.59	1.71	2.44	99.99
ECS15SS	32,748	23,923	18,622	73	80.86	0.62	2.13	6.80	99.94
SCS01SS	41,443	33,490	18,622	79	82.00	0.55	3.06	8.52	99.98
SCS02SS	43,767	31,471	18,622	82	82.86	0.54	2.76	2.50	99.98
SCS03SS	43,849	33,422	18,622	83	86.75	0.57	2.79	5.50	99.97
SCS04SS	32,016	22,856	18,622	70	70.60	0.70	2.78	6.20	99.98
SCS05SS	34,123	23,716	18,622	92	95.50	0.59	3.05	7.83	99.96
SCS06SS	42,250	27,312	18,622	73	76.50	0.63	2.52	3.13	99.96
SCS07SS	30,818	20,669	18,622	85	92.50	0.59	2.76	6.01	99.95
SCS08SS	41,727	30,126	18,622	76	81.14	0.57	2.40	7.21	99.95
SCS09SS	31,356	22,258	18,622	74	74.38	0.35	2.47	6.05	99.98
SCS10SS	40,239	27,722	18,622	66	70.67	0.61	2.74	3.81	99.96
SCS11SS	35,015	26,545	18,622	81	83.50	0.44	2.42	7.89	99.97
SCS12SS	31,380	18,622	18,622	76	76.43	0.65	2.80	2.75	99.98
Total	1,247,075	945,078	633,148	2,131

FIG 4.

Richness, evenness, and diversity estimators of Vibrio spp. in the YS, ECS, and SCS. The asterisks denote significant differences between areas. *, 0.01 < P ≤ 0.05; **, 0.001 < P ≤ 0.01; ***, P ≤ 0.001. The richness and evenness of Vibrio spp. in different groups were calculated with Chao 1 (A) and Shannon even (B) indices, while the Shannon index and phylogenetic distance were used to estimate the diversity of Vibrio spp. (C and D).

Spatial distribution and community classification of Vibrio spp.

An unweighted UniFrac analysis, an effective distance metric for microbial community comparison (39), was conducted at the OTU level to illustrate the diversity pattern of Vibrio spp. across all samples. The results of the fast UniFrac all-environment significance test (P < 0.05) and P test (P < 0.01) indicated significant community differentiation in different samples. Basically, the samples clustered according to spatial location, as shown in principal-component analysis (PCA) (Fig. 5A), and were separated into three clusters, i.e., YS, ECS, and SCS. The first two principal components explained 46.43% of the total variation. Samples in the YS were separated from ECS and SCS samples along the first principal coordinate (analysis of similarity [ANOSIM]; r = 0.7019; P = 0.001), which was mainly due to the predominance of OTU20 and OTU28 (P < 0.01) in the YS. Meanwhile, samples from the ECS, with a dominance of OTU58, were separated from the SCS samples along the second principal coordinate (P < 0.05) (Fig. 5A). ECS14 and ECS15 were located south of the ECS and were positioned close to the SCS cluster, indicating spatial continuity of the benthic Vibrio community.

FIG 5.

Community analysis of Vibrio at the OTU level. (A) Unweighted PCA with PC1 and PC2. The dominant OTUs of different sampling groups are indicated by red arrows. (B) RDA analysis illustrating the relationship between the Vibrio community at the OTU level and its top environmental variables.

Taxonomic assignment based on the EzBioCloud database showed 29 abundant species (>99.5%) across all sites (Fig. 6). The most dominant species was Vibrio sp. OTU132, occupying 21.22% of all sequences, followed by Photobacterium lipolyticum OTU28, V. gigantis OTU58, Photobacterium piscicola OTU69, Photobacterium lutimaris OTU135, and V. harveyi OTU7, which jointly accounted for 38.66% of all sequences. OTU132 was not resolved at the species level and showed the highest 16S rRNA gene similarity to Vibrio sp. strain 3459 (NCBI Taxonomy ID no. txid1676692) (99.6%) and V. gallicus (95.5%) (see Fig. S1 in the supplemental material). Thus, OTU132 may represent a novel Vibrio species. The remnant 23 relatively abundant species, from V. galatheae OTU173 to Photobacterium alginatilyticum OTU150, comprised 33.63% of all sequences in total (0.53% to 4.63%).

FIG 6.

Vibrio community compositions at the species level across all samples.

Although the relative abundance of Vibrio sp. 3459 (OTU132) was the highest across all sediments, it showed relatively low abundance in the YS compared to the ECS and SCS. In addition, there were high relative abundances of P. lipolyticum OTU28 and Vibrionaceae species OTU20 in the YS (P < 0.01); V. gigantis OTU58 and P. piscicola OTU69 in the ECS (P < 0.05); and P. piscicola OTU69, P. lutimaris OTU135, and P. alginatilyticum OTU150 in the SCS (P < 0.01). The components of Vibrio communities were more similar between the YS and ECS than the YS-SCS and ECS-SCS (one-way analysis of variance [ANOVA]; P < 0.05). The proportions of OTU135, OTU150, P. kishitanii OTU134, Vibrio sp. OTU96, V. sinaloensis OTU137, V. hangzhouensis OTU205, V. alfacsensis OTU81, and V. galatheae OTU173 were similar in the YS and ECS but differed significantly from the SCS (one-way ANOVA, Duncan test; P < 0.05). In contrast, other OTUs displayed different patterns. Vibrio sp. 3459 (OTU132), V. harveyi OTU7, Vibrio sp. OTU91, and Photobacterium sp. OTU208 showed no difference in abundance in the YS and SCS but differed from the ECS. OTU28, OTU69, OTU20, and Ferrimonas sediminum OTU67 had comparable abundances in the ECS and SCS but differed from the YS (P < 0.05). Similar to the OTU level PCA, the sample clustering based on the most abundant species also revealed three groups basically according with the spatial distribution (YS, ECS, and SCS) (see Fig. S2 in the supplemental material), with the exception that ECS14 and ECS15 were grouped in the SCS cluster instead of the ECS cluster (see Fig. S2). These observations revealed significant spatial heterogeneity in the Vibrio community in the coastal areas. Indeed, we found that there was a significant distance-decay pattern for the Vibrio community (Fig. 7), with a Spearman correlation coefficient of 0.512 (P < 0.01) between the Bray-Curtis community dissimilarity and geographic distance.

FIG 7.

Distance-decay relationships between the Vibrio community and geographic distance. Pairwise dissimilarities (the Bray-Curtis index) are plotted as a function of the distance among sites. The data are pairwise dissimilarities between the communities at 34 sites. The black line represents the best linear regression result.

Effects of environmental and spatial variables on the Vibrio community.

The relationship between environmental and spatial factors and Vibrio community components at the OTU level was shown by redundancy analysis (RDA) (Fig. 5B). Chl a and DO data were available for only two samples and were missing from most samples from the YS (see Table S1). Thus, the two variables were removed from the RDA. The first axis explained 40.66% of the total variance, whereas the second axis explained 8.68%. Monte Carlo permutation tests identified five factors, i.e., latitude (F = 8.3; P = 0.001), temperature (F = 8.3; P = 0.001), salinity (F = 5.7; P = 0.001), longitude (F = 5.6; P = 0.001), and SiO₃²⁻ (F = 2.6; P = 0.015), that significantly affected the Vibrio community structure. No significant correlation was observed between NO₃⁻, NO₂⁻, NH₄⁺, or PO₄³⁻ and the communities (P ≥ 0.05).

The Spearman correlations between the top 29 most abundant species and the environmental factors were calculated (Table 3). The most abundant species, i.e., Vibrio sp. 3459 (OTU132), had significant correlations with DO (r = −0.503; P = 0.006) and NO₂⁻ (r = 0.433; P = 0.015). V. gallaecicus OTU211, the unique dominant group in ECS15, was positively related to temperature, salinity, Chl a, DO, and SiO₃²⁻ and negatively to latitude and longitude (P < 0.05). As for the most abundant species in the YS, P. lipolyticum OTU28 and Vibrionaceae species OTU20 had the same relationship with factors in that they were both positively correlated with latitude and longitude and negatively correlated with temperature, salinity, Chl a, SiO₃²⁻, and NO₂⁻ (P < 0.05). Similar to the correlation mentioned above, P. lutimaris OTU135, as well as P. alginatilyticum OTU150 (dominant in the SCS), showed positive correlation with temperature and Chl a and negative correlation with longitude and latitude. In addition, P. alginatilyticum OTU150 was also correlated with several nutrients, including SiO₃²⁻, NO₃⁻, and PO₄³⁻ (P < 0.05). V. gigantis OTU58 in the ECS was related only to NH₄⁺, while P. piscicola OTU69 showed close relationships with temperature, longitude, and latitude and all the nutrients except NH₄⁺ (P < 0.05). Detailed correlation coefficients between all taxa and environmental factors are summarized in Table 3.

TABLE 3.

Correlations between percentage composition of taxa and environmental factors

OTU	Correlation coefficienta
	Longitude	Latitude	Temp	Salinity	Chl a	DO	NH4+	SiO32−	NO3−	NO2−	PO43−
Vibrio sp. 3459 (OTU132)						−0.503				0.433
P. lipolyticum OTU28	0.881	0.887	−0.876	−0.349	−0.575	−0.419		−0.520	−0.458
V. gigantis OTU58							0.387
P. piscicola OTU69	−0.662	−0.663	0.670					0.352	0.545	0.476	−0.458
P. lutimaris OTU135	−0.642	−0.721	0.657	0.396	0.445
V. harveyi OTU7					0.479	0.748
P. alginatilyticum OTU150	−0.910	−0.900	0.889		0.589			0.431	0.541		−0.449
Vibrionaceae species OTU20	0.850	0.877	−0.858	−0.379	−0.454			−0.504	−0.392
P. aquae OTU5							0.379
Vibrio sp. OTU91				0.491		−0.429		0.446	0.597	0.416	−0.435
P. aquimaris OTU52		−0.442	0.375	0.436		−0.526
Photobacterium sp. OTU208							0.479		0.420
V. gallaecicus OTU211	−0.540	−0.658	0.650	0.437	0.538	0.397		0.415
V. atypicus OTU87							0.491		0.425
Vibrio sp. OTU96	−0.717	−0.772	0.747	0.554				0.481	0.597	0.440	−0.459
P. kishitanii OTU134	−0.666	−0.634	0.702		0.654	0.454			0.508
Uncultured bacterium OTU138	−0.568	−0.596	0.604	0.448				0.415	0.725	0.554	−0.531
V. sinaloensis OTU137	−0.752	−0.731	0.736		0.656	0.544		0.389	0.445
V. orientalis OTU142	−0.598	−0.643	0.646		0.456	0.415		0.484
V. hangzhouensis OTU205	−0.720	−0.683	0.710		0.605	0.419		0.460	0.718	0.392	−0.389
Vibrio sp. OTU95				0.339	−0.515	−0.498
Uncultured bacterium OTU48
V. atypicus OTU210
P. swingsii OTU204							0.390	0.357
F. sediminum OTU67	−0.442	−0.485	0.455	0.362				0.409	0.650	0.447
Vibrio sp. OTU88				0.472	−0.493
V. alfacsensis OTU81	−0.591	−0.511	0.541		0.630	0.548
Chloroflexi sp. OTU37		0.421	−0.415	−0.521
V. galatheae OTU173	−0.668	−0.742	0.710	0.345	0.704	0.464		0.385	0.406
Other

^aOnly significant correlations (P < 0.05) are shown. −, negative; boldface, P < 0.01; lightface, P < 0.05.

To discern the relative importance of spatial and environmental factors, variation partition analysis (VPA) was conducted. The two types of factors jointly explained 46.7% of the total community variation (see Fig. S3 in the supplemental material). The pure effect of spatial variables (9.2%) was greater than that of environmental factors (5.9%) (see Fig. S3). Remarkably, the shared effects of environmental and spatial factors explained 31.6% of the variation.

DISCUSSION

The abundances and community compositions of Vibrio populations in different marine environments worldwide have been reported previously. Most of these works have focused on pelagic environments (17–19); however, little is known about their ecology in sediments. Here, as a complement to previous studies, our study assessed the spatial distribution of Vibrio communities, as well as individual taxa, in the Chinese marginal seas. To our knowledge, this is the first investigation regarding the geographic distribution pattern of vibrios in sediments. Our results showed that Vibrio spp. exhibited substantial spatial heterogeneity with a joint effect of spatial and environmental factors.

Spatial distribution pattern of Vibrio spp. in sediments of Chinese marginal seas.

Previous studies have demonstrated distinct Vibrio abundances and community compositions in separate marine environments worldwide (3, 26, 33). In this study, we showed that Vibrio communities differed significantly in sediments along the coast of China, magnifying different spatial distributions between certain OTUs. To resolve this, we found a significant distance-decay relationship (Fig. 7), which has been the most frequently observed pattern underpinning microbial biogeography (44, 45). Many previous studies have reported microbial biogeography targeting the total bacterial population, and thus, specific lineages have been neglected. Our results suggest that a specific phylogenetic group, i.e., Vibrio, can also exhibit a clear spatial distribution pattern. It is likely that a specific phylogenetic group represents rare taxa. This may apply to Vibrio, as the relative abundance of Vibrio in the total bacteria was less than 0.1% (Fig. 3). Thus, the discovery of spatial heterogeneity in Vibrio spp. was consistent with that observed in rare bacteria (46).

Relating the Vibrio distribution pattern to abiotic factors has revealed the effects of various factors, i.e., hydrological gradients (e.g., temperature and salinity), inorganic and organic nutrients, the abundance of host organisms, and other factors (e.g., pigments) (3, 5, 7, 22). Similarly, the present study showed that the benthic Vibrio community was governed by temperature, salinity, several nutrients (NO₃⁻, NH₄⁺, and SiO₃²⁻), and latitude and longitude. Temperature and salinity are the most frequently observed factors that are influential in the Vibrio community structure (26, 33, 34). It was noticeable that latitude and longitude also contributed to the community variation. This may be explained by the differential discharge from the Changjiang River and Pearl River into the YS, ECS, and SCS (40, 41), resulting in space-related distinctions in sediment sources. To tease apart the effects of spatial and environmental factors, VPA was performed, which revealed that the pure effect of spatial variables was slightly stronger than that of environmental factors. However, a higher proportion of community variation was explained by the joint effect of the two types of factors. Thus, our study expanded upon previous studies, where only the influence of environmental factors was investigated (1, 4, 7, 18), by showing an interaction of spatial and environmental variables in determining the distribution of Vibrio communities. These interactions include the latitude-dependent change in temperature.

Interestingly, two samples from near the Taiwan Strait, ECS14 and ECS15, were separated from other ECS samples but closer to samples from the SCS. The reason may be that the environmental conditions of ECS14 and ECS15 were affected by the Taiwan Warm Current (47). In addition, the Taiwan Strait, which is known to have the highest erosion rate in the world, receives sediments from the Chinese mainland and the island of Taiwan, and thus, nutrients such as NH₄⁺, SiO₃²⁻ and PO₄³⁻ are concentrated (48). The complex sedimentation processes and currents may facilitate the growth of a community different from those in adjacent areas (48). For example, OTU211 of V. gallaecicus, which shows a wide range of growth temperatures (4 to 35°C) and a capacity to reduce nitrate to nitrite (49), was abundant in and specific to ECS15.

High abundance of Vibrio spp. in the benthic environment and potential adaptive strategy.

Consistent with previous reports of coastal sediments as reservoirs for Vibrio spp. (26, 33–35), the median density of Vibrio spp. in sediments reported here was 3 orders of magnitude higher than values (10⁵ to 10⁶ copies/liter) reported in the water column from the same area (50), highlighting sediments as a favorable ecological niche for Vibrio spp., which may evolve ecological features different from those of species that reside the water columns (26). The reason might be that sediment provides biotic and abiotic surfaces for bacterial attachment, and its organic matter concentration is much higher than that of the overlying water column (26). Additionally, the numbers of culturable Vibrio spp. were 10⁰ to 10⁶ CFU/liter at three beaches of the northern Adriatic Sea (Italy) and 10² to 10³ CFU/g (wet sediment) in sediment from the estuarine zone of the Bengal delta (24, 34, 36). In contrast, the total Vibrio abundances derived from qPCR were 1 × 10⁴ to 3.2 × 10⁵ copies/g in sediment from the Oregon estuary (United States) (33) and 1 × 10⁵ to 1 × 10¹⁰ cells/liter in the Mediterranean Sea (Italy) (26). This discrepancy indicates that many marine Vibrio spp. are as yet unculturable. Compared to previous qPCR results, a higher Vibrio abundance (2.4 × 10⁴ to 1.1 × 10⁷ copies/g) occurred in sediments of Chinese marginal seas (see Table S1). This may suggest unique environmental conditions in the Chinese marginal seas, which should be considered one of the most important coastal ecosystems for further Vibrio studies. Specifically, the abundance of Vibrio spp. showed obvious differences between the YS and ECS-SCS (P < 0.01), with an increasing trend toward the south. The observed differences in Vibrio abundance measured by qPCR may reflect special conditions of separate benthic environments and may also be contributed to by the varied copy numbers of 16S rRNA genes (from 6 to 14 operons) between Vibrio spp. (16). Taken together, the ecological importance of benthic Vibrio populations merits further investigation.

The abundant Vibrio populations were diverse and contained several species that were dominant and evenly distributed across the sampling areas, such as Vibrio sp. 3459 (OTU132), V. gigantis OTU58, V. gallaecicus OTU211, V. atypicus OTU87, and V. orientalis OTU142. The most common species, i.e., Vibrio sp. 3459 (OTU132) and V. gigantis OTU58, comprised the highest proportion in the YS, ECS, and SCS (Fig. 6), indicating they can adapt to a wide range of environments, probably through their capacity to utilizing a broad range of substrates. Indeed, it has been reported that a large number of compounds (e.g., glucose, melibiose, amygdalin, and glycerol) can be utilized by V. gigantis as sole carbon sources (51). In contrast, other common species comprised a relatively low community proportion, and instead, specific local species occurred in different areas with higher relative abundance than the common species, except Vibrio sp. 3459 (OTU132) and V. gigantis OTU58. The reason might be that environmental forces selected for specific species that adapt and grow faster than common species. Despite the widespread distribution of several abundant species, the distance-decay pattern seen in the total Vibrio community (Fig. 7) thus can be attributed to these site-specific species.

Because of their genetically and metabolically diverse nature, vibrios can survive in all ocean environments, ranging from coastal waters to open ocean and from surface to deep water (5). It was found that the most predominant Vibrio sp. in sediments in the Chinese marginal seas (Vibrio sp. 3459) differed from V. lentus in sediments from the Oregon estuary (33). Vibrio spp. may evolve several adaptive strategies to survive in benthic environments, e.g., utilize a wide range of organic carbon sources, enter the VBNC state, or use sediments as reservoirs (3, 16, 52, 53). Most Vibrio spp. can utilize a wide range of organic carbon compounds as carbon and energy sources, such as polysaccharides derived from macroalgal cell walls or zooplankton exoskeletons (3, 16, 17). When environmental conditions become unfavorable, many Vibrio spp. (e.g., V. fluvialis) can enter into the VBNC state and recover after conditions once again become favorable (3, 52). Furthermore, Vibrio spp. may use sediment as a reservoir rather than entering the VBNC state (52, 53). Indeed, compared to the Vibrio abundance in the water column of the northern Chinese marginal seas (∼1.4 × 10⁶ copies/liter in summer and ∼1.9 × 10⁵ copies/liter in winter) (50), the median densities of Vibrio spp. in sediments reported here was much higher. Furthermore, the overall Vibrio community structure showed consistency in sediment and seawater at the Oregon estuary, suggesting a closed interaction between the two reservoirs (33). To reveal the survival strategy of Vibrio spp., comparison between seawater and sediment in the Chinese marginal seas should be further analyzed.

The most abundant taxa are thought to be the most important in fluxes of dissolved organic matter and carbon cycling, whereas recent evidence has indicated that rare taxa are also key for primary ecosystem functions (5, 54, 55). A widely held belief about vibrios is that they play a relatively minor role in chemical transformations in the ocean, which is largely based on the low relative abundance of Vibrio (5, 56). However, Takemura et al. suggested that the role of vibrios has been underestimated due to their high biomass and rapid bloom and disproportionate predation by protozoa and viruses (5). Indeed, it has been reported that many vibrios isolated from seawater in Chinese marginal seas have the ability to hydrolyze and metabolize chitin (330 out of 581 strains), alginic acid, and agarose (3, 57, 58). Therefore, the putative function of Vibrio spp. in marine biogeochemical processes should be reevaluated (3, 5, 13, 14), and more research should be conducted in different sedimentary environments.

High proportion of Photobacterium spp. in the community composition.

Many species belonging to the genus Photobacterium were recovered in our community composition analysis. Photobacterium bacteria are phylogenetically close to Vibrio bacteria, and their species have been observed to show good congruence in terms of morphological and rRNA characteristics (59). For example, Photobacterium spp. are able to grow on regular media (e.g., marine 2216E agar) and selective media, such as thiosulfate-citrate-bile salts-sucrose (TCBS) agar, which is widely used in Vibrio isolation (60). Thus, the evolutionary relationships between Vibrio and Photobacterium are difficult to define. In fact, due to their common evolutionary origin, there is a high frequency of phylogenetic placement of Photobacterium species within other genera (e.g., Vibrio, Aliivibrio, Grimontia, and Enterovibrio) (59, 61, 62) in the family Vibrionaceae. Photobacterium spp. have been observed to be globally distributed in the marine environment and to be more abundant in bottom seawater and sediment; this may be attributed to their ability to produce polyunsaturated fatty acid, cold-adapted lipase, esterase, and antimicrobial compounds (62, 63). Most Photobacterium spp. currently known (∼23 out of 31) were isolated from the deep sea or sediment, existing either as free-living organisms or associated with marine organisms (e.g., fish and squid). Environments may have determined the distribution pattern of Photobacterium spp., as several Photobacterium OTUs were shown to be related to environmental factors. For example, P. piscicola OTU69 is correlated with various parameters, including temperature and inorganic nutrients; this might be due to its physiological characteristics, such as the optimal temperature for growth and the substrate utilization spectrum (64, 65).

The taxonomic status and adaptive capability of Photobacterium spp. may affect the analysis of the Vibrio community, and previous studies even grouped Photobacterium and Vibrio together for diversity analysis (66). Most early studies performed taxonomic assignment based on the SILVA and Greengenes databases, which are not resolved to the species level (7, 33). In this study, to obtain a more accurate result, after annotation in SILVA, taxonomy was reassigned against the EzBioCloud database, which provides the latest taxonomy of microbial isolates and has a complete taxonomic hierarchy from phylum to species (67, 68). Several OTUs annotated as Vibrio (OTU58) and Photobacterium (OTU69, OTU135, and OTU52) in SILVA were assigned to V. gigantis, P. piscicola (OTU69), P. lutimaris (OTU135), and P. aquimaris (OTU52) in the EzBioCloud database.

Finally, 16S rRNA gene analysis may have some limitations, such as the fact that it does not contain highly variable domains that are probably required to analyze relationships of closely related organisms (69). It was reported that next-generation sequencing (NGS) of protein-coding marker genes, including those encoding heat shock protein 60 (hsp60) (1), ferric uptake regulator (fur) (70), and uridylate kinase (pyrH) (18), may replace the 16S rRNA gene as excellent markers for differentiating closely related Vibrio spp. (1). However, it is difficult to amplify environmental Vibrio spp. using these marker genes, and the Illumina MiSeq technology can accurately determine only genes less than 600 bp in length. NGS of protein-coding marker genes paired with information generated from sequencing the more widely used 16S rRNA gene can dramatically improve the study of Vibrio spp. in future.

Conclusions.

The spatial distribution of total Vibrio spp. in sediments of Chinese marginal seas was revealed using qPCR and high-throughput sequencing. Overall, according to qPCR, Vibrio spp. showed higher abundance in sediments of the Chinese marginal sea than in other areas. Both the abundance and community composition varied between different sea areas, with the Vibrio assemblages displaying a clear distance-decay pattern. This suggests marine surface sediment is a favorable ecological niche for Vibrio spp., which, however, may differ in composition in distinct environmental settings. Both spatial and environmental (especially temperature, salinity, and SiO₃²⁻) factors played important roles in regulating the assembly of the Vibrio community, and the influence of pure spatial variables appeared to be stronger than that of pure environmental components. Several Vibrio spp. showed no difference among the YS, ECS, and SCS, demonstrating their broad niche breadth, likely facilitated by strong adaptive capabilities. Further attention should be paid to exploring the linkages between Vibrio dynamics and roles in biogeochemical cycling.

MATERIALS AND METHODS

Site description and sample collection.

To compare the sedimentary Vibrio communities in different niches of the Chinese marginal seas, a total of 34 sites, with water depths ranging from 15.1 to 83.5 m, were sampled, representing three sea areas, i.e., the YS, ECS, and SCS. Samples were collected aboard the R/V Dong Fang Hong 2 and an offshore vessel, R/V Haili, during two cruises from 27 March to 17 May 2017 (i.e, spring) (Fig. 1), using stainless steel box samplers. Based on spatial locations, the samples were divided into three groups (see Table S1), i.e., 7 sites in the YS (YS01 to YS07), 15 sites in the ECS (ESC01 to ECS15), and 12 sites in the SCS (SCS01 to SCS12).

Surface sediments of ∼200 g from 0- to 2-cm depth were collected with small sterilized shovels and stored in sterilized polyethylene bags (71); a total of ∼3 g sediments were pooled into three 5-ml sterile tubes. All the sediments were stored at −20°C on board and transferred to −80°C in the laboratory until DNA extraction. Pore waters were collected with Rhizon samplers (Rhizosphere Research Products; Netherlands), inactivated with HgCl₂, and stored at 4°C before measurement of dissolved inorganic nutrients. The pore waters were not used for community analysis. Dissolved inorganic nutrients (NO₃⁻, NO₂⁻, NH₄⁺, PO₄³⁻, and SiO₃²⁻) in sediment pore waters and temperature, salinity, Chl a, and DO in bottom seawater were determined as previously described (72).

DNA extraction and quantification analysis.

Genomic DNA was extracted from sediment (0.25 g [wet weight]) using a Power Soil DNA isolation kit (Mo Bio Laboratories, Inc., Carlsbad, CA, USA) according to the manufacturer’s instructions, with the modification that a FastPrep-24 cell disrupter (MP Biomedicals, Irvine, CA, USA) was used. The concentration and quality of the extracted DNA were measured using a Nanodrop spectrophotometer (ND-2000; Thermo Fisher Scientific), and the remaining DNA was stored at −80°C.

Quantitative PCR with SYBR green detection, using the StepOnePlus real-time PCR system (Applied Biosystems) and StepOne software version 2.2, was used to quantify the abundance of total Vibrio bacteria and the bacterial community. The 16S rRNA oligonucleotide primers used in the qPCR were V567F and V680R (73, 74), specific for the genus Vibrio, whereas B967F and B1046R (74, 75) were used to quantify the domain Bacteria. Based on previous studies (7, 73–75), the primer sequences, reaction mixture, and cycling conditions were modified and are provided in Table 4. The standards were prepared from 16S rRNA gene nucleic acid templates of V. rotiferianus WXL191 (our laboratory) and referred to the methods of Zheng et al. (76). All assays were conducted in triplicate with negative controls, and all amplification efficiencies were between 95% and 105% with R² values of >0.99.

TABLE 4.

Primers and conditions for PCR amplification and sequencing

Primers	Sequences (5′–3′)	Length of amplicon (bp)	Reaction mixture	Thermal profile	Organism	Reference(s)
V567F/V680R	GGCGTAAAGCGCATGCAGGT/GAAATTCTACCCCCCTCTACAG	113	20 μl; 5.0 mM MgCl2, 0.2 μM each primer, 2 μl template DNA	95°C for 5 min, followed by 35 cycles of 30 s at 95°C and 60 s at 64°C	Genus Vibrio; 16S rRNA	73, 74
B967F/B1046R	CAACGCGAAGAACCTTACC/CGACAGCCATGCANCACCT	79	20 μl; 5.0 mM MgCl2, 0.4 μM each primer, 2 μl template DNA	94°C for 3 min, followed by 35 cycles of 30 s at 94°C, 45 s at 57°C, and 30 s at 72°C	Total bacteria; 16S rRNA	74, 75
V169F/V680R	GGATAACCTATTGGAAACGATG/GAAATTCTACCCCCCTCTACAG	511	20 μl; 1× Fast Pfu buffer, 0.25 mM deoxynucleoside triphosphates, 0.2 μM each primer, 0.4 μl Fast Pfu polymerase, 0.2 μl bovine serum albumin, 10 ng template DNA	95°C for 3 min;, 35 cycles of 30 s at 95°C, 30 s at 55°C, and 45 s at 72°C; and a final extension at 72°C for 10 min	Genus Vibrio for Illumina MiSeq plaform; 16S rRNA	7

High-throughput sequencing and read processing.

To reveal the overall composition of the Vibrio community, we combined the Vibrio-specific 16S rRNA primers V169F and V680R to target the variable regions V2 to V4 (7). The PCR system and cycling conditions were modified and are shown in Table 4. The raw data were filtered according to the pipeline of Quantitative Insights into Microbial Ecology (QIIME) (77). Reads were assigned to samples according to their barcodes without mismatch. The raw reads that had a quality score higher than 20 over a 5-bp window size and a minimum length of 100 bp were retained (78). Paired-end DNA sequences were joined with at least a 50-bp overlap and less than 5% mismatches using FLASH (79). A Perl script, i.e., daisychopper.pl, was used to randomly subsample sequences from each sample according to the lowest read numbers to equalize sampling efforts (80). OTU clustering and taxonomic assignment were also performed in QIIME. Specifically, OTUs were defined at a 97% sequence similarity level, and then chimera sequences were detected and removed with UCHIME (81), as recommended by QIIME tutorials. Taxonomy was assigned using the RDP Classifier v2.2 (82) against the SILVA v128 16S rRNA gene reference database (http://www.arb-silva.de) with a minimum support threshold of 70%. To remove the effect of sampling effort upon analysis, sequences were then rarefied to the lowest read number for all samples with a “single rarefaction” QIIME script (71, 77).

Statistical analysis.

Spatial differences in environmental parameters were tested by one-way ANOVA. To compare total Vibrio spp. and bacterial abundances, data were log |x + 1| transformed. The chi-square and Kruskal-Wallis tests were also performed. Correlations between the abundance of Vibrio spp. and environmental parameters were assessed with Spearman’s rank correlation analysis. All the above-mentioned analyses were performed using Statistica version 22.0 (StatSoft, Tulsa, OK, USA).

The diversity indices for alpha diversity analysis, including Good’s coverage, phylogenetic diversity, Chao 1, equitability (a Shannon index-based measure of evenness), and Shannon, were calculated using MOTHUR software packages (83). Wilcoxon’s test was used to compare the alpha diversity indices between different groups of samples. For beta diversity, PCA was performed with Canoco 5 software (Microcomputer Power) at the OTU level. One-way ANOSIM was conducted in PRIMER 6 (Plymouth Routines in Multivariate Ecological Research) using the Bray-Curtis resemblance matrix calculated from square-root-transformed OTU tables to determine the significant differences among different samples. The distance-decay pattern of the Vibrio community was examined, and the significances of correlations were tested by Spearman’s rank correlation tests.

The relative contributions of geographic distance and environmental factors were estimated by VPA, with adjusted R² coefficients based on RDA. The relative contributions of both components were explained by pure spatial variables (S), pure environmental variables (E), and the combined effects of both space and environment (S and E), respectively. The relationships between phylotypes and environmental factors were evaluated by RDA in Canoco 5 with 9,999 Monte Carlo permutation tests using square-root-transformed data. The Spearman rank correlation coefficients were calculated to determine the relationship between environmental factors and diversity indices of the Vibrio community at species level using STATISTICA v 22.0.

Data availability.

The Illumina sequences were deposited in the NCBI SRA database under accession number SRP159585 (BioProject accession number PRJNA488965).