Two belugas, three Pacific white-sided dolphins, one common bottlenose dolphin, and nine Cape fur seals kept in the Marine and Waterpark in Atlantis hotel, Sanya, China, were surveyed in this study (Supplementary Table S1). All animals were maintained for public display in indoor pools with natural seawater filtered by the same filtration system based on mechanical filtration with percolator filters, pressurized sand filter, and protein skimmers. All seawater was treated with chlorination and ozone sterilization. The diet of all surveyed animals consisted of whole frozen fish including herring (Clupea harengus) and capelin (Mallotus villosus).

For each individual animal, at least one fecal sample was collected during June to July, 2020. All 35 fecal samples were collected from animals under operant conditioning (behavioral training) during routine medical examinations. Fecal samples were harvested by veterinarians between 8:00 and 9:00 in the morning. Whales and dolphins were asked for a “belly-up” position and feces were collected directly from the rectum with a Levin’s tube (Vygon, France), with a diameter of 4mm, inserted 10–20cm into the rectum. Fecal samples of fur seals were collected using anal swabs, with a diameter of 12mm, and inserted 10–15cm into the rectum. All fecal samples were frozen at −20°C until analysis.

DNA was extracted using MoBio PowerSoil extraction kits (Mo Bio Laboratories, Carlsbad, CA, United States), according to the manufacturer’s instructions. Three extraction blank control samples were also included. The extracted DNA was quantified with a Qubit fluorometer (Invitrogen Inc. Manufacturer: Life Technologies Holdings Pte Ltd., Singapore) and used for amplification of the V4 region of the 16S rRNA gene with the primer pair 515f Modified and 806r Modified (Walters et al., 2015). The PCR cycling conditions were as follows: denaturation at 95°C for 3min, followed by 27cycles at 95°C for 30s, 55°C for 30s, and 72°C for 45s, and a final extension at 72°C for 10min. Triplicate PCR amplifications were combined after purification with a TaKaRa purification kit (TaKaRa, Japan). The PCR products were prepared for library construction with the TruSeq DNA sample preparation kit (Illumina, San Diego, CA, United States), according to the manufacturer’s instructions. The libraries were sequenced at MajorBio Co. Ltd. (Shanghai, China) using the HiSeq platform (Illumina) with a paired-end 250bp sequence read run.

After sequencing, the raw reads were assigned to their respective samples according to their barcodes and forward and reverse primers (one mismatch of each was allowed). Paired-end reads of sufficient length with at least a 30-bp overlap were combined into full-length sequences by using FLASH program version 1.2.8 (Magoč and Salzberg, 2011). The average fragment length was 253bp. Btrim program (version 0.2.0) was used to filter out low-quality sequences. The quality score was set to >20 with a 5-base window size as the standard and any sequences containing Ns or<200bp were discarded. Sequences with lengths of 245–260bp were retained (Kong, 2011). We used UNOISE3 to correct sequencing errors to determine the real biological sequences at single-nucleotide resolution by generating amplicon sequence variants (ASVs) with default settings (Edgar, 2016). A representative sequence from each ASV was selected for taxonomic annotation by comparison with the SILVA 132 database (Quast et al., 2013), which includes bacterial, archaeal, and eukaryotic sequences. To account for varying sequencing depths, the ASVs were randomly resampled to normalize the reads of each sample. The raw sequencing reads of all samples were deposited to the NCBI database¹ under BioProject accession number: PRJNA743584.

A total of 1,927,942 sequences were obtained from 35 fecal samples of different marine mammals after quality assessment, the marine mammals included two belugas, three Pacific white-sided dolphins, one common bottlenose dolphin, and nine Cape fur seals. An average of 55,084±10,935 sequences was obtained from each sample. In order to obtain a more accurate result of α-diversity, we rarefied to 33,334 sequences per sample, and then used this set for analysis of microbial diversity, composition, and structure. The α-diversities of microbial communities from the gut of different marine mammals were calculated. The Shannon, Inverse Simpson, and Chao1 indices and observed richness indicated that the α-diversity of the gut microbiomes from belugas and Cape fur seals was higher than those from Pacific white-sided dolphins and common bottlenose dolphin (Figure 1).

NMDS analysis of microbial communities clearly showed three principal groups: belugas, two dolphin species, and Cape fur seals (Figure 2). These results suggest that whales, dolphins, and fur seals possess different gut microbial communities, even in the context of similar food and environment. Furthermore, the MRPP, ANOSIM, and PERMANOVA showed significant differences in the gut microbial communities of whales, dolphins, and fur seals (p<0.01). By contrast, no significant difference was detected between the Pacific white-sided dolphin and the common bottlenose dolphin (p>0.05; Table 1).

The relative abundance of microorganisms was evident at the phylum, family, and genus levels with a similarity of 97% for ASV classification and provided detailed information on the composition of the microbial communities (Figures 3–5). Firmicutes were the dominant bacterial lineage in the fecal samples of belugas, accounting for 94%~96%. The dominant microbial phyla for two of the three Pacific white-sided dolphins were also Firmicutes, the remaining animal (Duomi) was dominated by Proteobacteria (Gamma) (76%~83%). The majority of the fecal samples from the lone common bottlenose dolphin in this study were dominated by Proteobacteria (Gamma), with one exception that was dominated by Firmicutes. The gut microbial communities of Cape fur seals were dominated by Firmicutes, Bacteroidetes, Fusobacteria, Epsilonbacteraeota, and Proteobacteria (Gamma). However, the respective compositions of different seal fecal samples were slightly different. At the family taxonomic level, Lachnospiraceae, Ruminococcaceae, and Peptostreptococcaceae, which are affiliated with the Clostridiales, were the dominant bacterial lineages in the beluga fecal samples. Enterobacteriaceae, Enterococcaceae, Clostridiaceae 1, and Peptostreptococcaceae were dominant in the fecal samples of Pacific white-sided dolphins. The gut microbial communities of the common bottlenose dolphin was also dominated by Enterobacteriaceae, Clostridiaceae 1, and Peptostreptococcaceae, but Vibrionaceae was also one of the dominant bacterial lineage in some of the fecal samples. Campylobacteraceae, Fusobacteriaceae, Family XI (affiliated with the Clostridiales), Moraxellaceae, and Bacteroidaceae were the dominant bacterial lineages in the fecal samples of Cape fur seals. Furthermore, at the genus taxonomic level, the gut microbial communities of belugas were mainly composed of Lachnoclostridium, Romboutsia, Fournierella, and Eubacterium fissicatena group. Meanwhile, Escherichia-Shigella, Clostridium sensu stricto 1, and Enterococcus were the dominant bacterial genera in the fecal samples of Pacific white-sided dolphins, while Vibrio, Morganella, and Clostridium sensu stricto 1 were dominant in samples from the common bottlenose dolphin. As for Cape fur seals, Fusobacterium, Campylobacter, Psychrobacter, Bacteroides, Marinifilum, and Ezakiella were the main bacterial genera.

To better understand the core microbial communities of different marine mammals, we explored the detailed differences at the ASV level between different marine mammal fecal samples. Quality control and random resampling of the 35 samples were conducted, and the sequence reads were clustered into 572 ASVs at the 97% similarity level. As shown in Figure 6, in the fecal samples of belugas, Pacific white-sided dolphins, common bottlenose dolphin, and Cape fur seals, contained 13, 44, 7, and 206 unique ASVs, respectively. Cape fur seals shared 132, 20, and 15 ASVs with Pacific white-sided dolphins, belugas, and common bottlenose dolphin, respectively, while Pacific white-sided dolphins shared 16 and 13 ASVs with common bottlenose dolphin and belugas, respectively. By contrast, only two ASVs were shared between belugas and common bottlenose dolphin. Furthermore, 44 ASVs were detected in all four different marine mammal species.

Indicator organisms analysis showed that Lachnoclostridium, Proteus, Eubacterium nodatum group, Tyzzerella, Ruminiclostridium, Oscillospira, Eubacterium fissicatena group, Romboutsia, Hydrogenoanaerobacterium, GCA-900066225 (Ruminococcaceae), Ruminococcus gauvreauii group, Negativibacillus, Mycobacterium, Fournierella, Faecalibacterium, Butyricicoccus, and some other unclassified genera were highly associated with the gut microbial communities of belugas (Figure 7). Enterococcus was the only indicator genus in the gut microbial communities of Pacific white-sided dolphins. In the gut of the common bottlenose dolphin, Ureaplasma and one unclassified genus were the indicator organisms. Three indicator genera, Fusobacterium, Bacteroides, and Peptoclostridium, were observed in the fecal samples of Cape fur seals. The heatmap based on these indicator ASVs reveals that belugas and Cape fur seals harbor distinct gut microbial communities (Figure 8). Meanwhile, the gut microbial communities of Pacific white-sided dolphins and common bottlenose dolphins could not be separated (Figure 8), consistent with the NMDS results.

In this study, we reported the gut microbial communities in two belugas, three Pacific white-sided dolphins, one common bottlenose dolphin, and nine Cape fur seals. These marine mammals resided at the Marine and Waterpark in Atlantis hotel, Sanya, China. Moreover, these creatures also shared the same water filtration system and diet including herring (Clupea harengus) and capelin (Mallotus villosus). The α-diversity indices showed that the gut microbial diversity of belugas and Cape fur seals were higher than that of Pacific white-sided dolphins and common bottlenose dolphin, with the gut microbial diversity of the two dolphin species being similar. Given the same food sources and water quality, we speculate that the low gut microbial diversity of Pacific white-sided dolphins and common bottlenose dolphin may attribute to the poor health status of the animals themselves.

In their pioneering work, Ley et al. (2008) explored the fecal microbiota of humans and 59 other mammalian species living in captivity and in the wild. Results indicated that host diet and phylogeny both influence microbial diversity, and the influence increases from carnivory to omnivory, and to herbivory (Ley et al., 2008). The gut microbiota makes an important contribution to host metabolism by contributing enzymes that are not encoded by the host genome such as the decomposition of polysaccharides and polyphenols and the synthesis of vitamins (Rowland et al., 2018). Different diets can shape gut microbial communities in both the short and long term (Flint et al., 2012). For instance, low intake of accessible carbohydrates and dietary fibers may reduce the microbial diversity and change the microbial communities, contributing to the depletion of specific microbial taxa (Sonnenburg and Sonnenburg, 2014; Sonnenburg et al., 2016). In our study, four different marine mammals are carnivores, living in the same environment, and eating the same food. Therefore, if diet and environment affects gut microbial community composition and structure more than the phylogeny, the gut microbial β-diversity of belugas, Pacific white-sided dolphins, common bottlenose dolphin, and Cape fur seals should tend to be similar. However, according to the results of NMDS and different statistical analyses, belugas, dolphins, and Cape fur seals possess significantly different gut microbial communities. Whereas, the gut microbial communities of Pacific white-sided dolphins and common bottlenose dolphin were statistically undifferentiated, we propose the reason for the similarities in β-diversity may attributed to the animals’ health status and medical treatments prior to sampling; however, further study is required.

Members of Firmicutes, Bacteroidetes, Fusobacteria, and Epsilonbacteraeota constituted the vast majority of the microorganisms in the gut of Cape fur seals. The most common bacterial genera were Fusobacterium, Campylobacter, Psychrobacter, Bacteroides, Marinifilum, and Ezakiella. However, to date, no Cape fur seal gut microbial communities have been reported. The skin microbial communities of Antarctic fur seal (Arctocephalus gazella) showed that Proteobacteria, Bacteroidetes, Firmicutes, and Actinobacteria were the dominant bacterial lineages (Grosser et al., 2019). Although there are definite similarities, the Fusobacteria, rather than the Actinobacteria were the dominant bacterial lineage in the gut of Cape fur seals. Nevertheless, at the genus level, only Psychrobacter was detected as a dominant bacterial lineage both in the gut and the skin microbial communities of fur seals, and Chryseobacterium and Jeotgalibaca, which were dominant in the skin microbial communities, were either not detected (Chryseobacterium) in the gut microbial dataset or identified as a single ASV (Jeotgalibaca) within in the gut of AX (0.003–0.05%, relative abundance, Pacific white-sided dolphin), and Kun, Flaca (0.003–0.02%, relative abundance, Cape fur seals). The Fusobacterium, as an animal-associated genus affiliated with the Fusobacteriaceae, are widely found in the gut of humans and other animals (Brennan and Garrett, 2019) including rats (Zhang et al., 2020), macaques (Nugeyre et al., 2019), pigs (Wylensek et al., 2020), penguins (Tian et al., 2021), striped dolphin (Godoy-Vitorino et al., 2017), and whales (Li et al., 2019), and some birds and fishes (Michl et al., 2017; Rasheeda et al., 2017; Chen et al., 2018; Michel et al., 2018). These organisms may play a symbiotic or pathogenic role (Brennan and Garrett, 2019).

The genus Campylobacter are very common bacteria in animal digestive tracts, whose powerful flagellar motility is very important in nutrient acquisition, whereby bacterial cells attach to surfaces and rotate their flagella to increase nutrient flux for metabolism (Beeby, 2015). Different species of Campylobacter can have variable characteristics of either pathogen or commensal, depending on their host immune status and the pathotype of the bacteria themselves (Wigley, 2015). For example members of the Campylobacter, such as Campylobacter jejuni and Campylobacter coli, are among some of the most common human gastrointestinal pathogens, causing bacterial diarrheal illness (Fitzgerald, 2015). In addition, the ASVs classified into Campylobacter in this study were further identified by phylogenetic analysis. The results showed that these ASVs clustered together far from the common species of Campylobacter (Supplementary Figure S2). Therefore, the role of Campylobacter in the gut of Cape fur seals remains unclear. The members of the Psychrobacter, another common genus, can be found in different marine environments and marine creatures, such as Arctic fjord (Zeng et al., 2015), Arctic glacier (Zeng et al., 2016), deep sea (Evans-Yamamoto et al., 2019), penguins (Kämpfer et al., 2020), corals (Zachariah et al., 2016), and marine crustaceans (Romanenko et al., 2009). In this and previous studies, the genus Psychrobacter was dominant in the gut of Cape fur seals and the skin of Antarctic fur seal (Arctocephalus gazella). At the same time, we barely detected this bacterium in the gut of belugas, Pacific white-sided dolphins, and common bottlenose dolphin. The members of Psychrobacter, as one of the commensal bacteria, could improve the microbial diversity of gastrointestinal tract, and upregulate the expression of immune-related genes and is therefore considered to be probiotic (Sun et al., 2011; Yang et al., 2011; Makled et al., 2017). According to previous study, it is interesting that the decrease in the relative abundance of Psychrobacter may associate with depressive conditions of hosts. Meanwhile, the results of metabolomics showed that the lower levels of amino acids, and fatty acids, and higher amounts of bile acids, hypoxanthine, and stercobilins were detected in the depression groups. However, the roles of these bacteria in the different stages of depression are not clear (Yu et al., 2017). Moreover, the Psychrobacter, one of the key genera of young-like gut microbial community of African turquoise killifish (Nothobranchius furzeri), was considered to play an important role in extending host life span (Smith et al., 2017). Additionally, after infection by porcine epidemic diarrhea virus, the relative abundance of Psychrobacter, which is a symbiotic group of bacteria in the gut of pigs, was decreased.

Members of the Bacteroidetes, one of the major lineages of friendly bacterial commensuals, possess large numbers of genes encoding carbohydrate active enzymes, which allows them to switch readily between different energy sources in the gut, depending on availability, using sophisticated regulatory mechanisms to control gene expression (Thomas et al., 2011). Polysaccharides comprise the most abundant type of biological polymer and, as such, also the most abundant biological food source. Carbohydrate fermentation by Bacteroides and other intestinal bacteria results in the production of a pool of volatile fatty acids that are absorbed through the large intestine and utilized by the host as an energy source, providing a significant proportion of the host’s daily energy requirement (Hooper et al., 2002). Although most of the Bacteroides are commensals in the gut, some species can also be responsible for infection, including Bacteroides fragilis, Bacteroides distasonis, Bacteroides ovatus, Bacteroides thetaiotaomicron, Bacteroides vulgatus, and Bacteroides uniformis, with significant morbidity and mortality (Wexler, 2007). However, in this study, the genus Bacteroides was only found to be dominant in the fecal samples of Cape fur seals and was rarely detected in the other three marine mammals. Moreover, ASV 12, ASV 24, and ASV 35 were the most predominant ASVs and were classified to Bacteroides at the genus level. The phylogenetic analysis, based on the neighbor-joining method, showed that the strain Bacteroides plebeius DSM 17135 was most closely related to ASV 12, ASV 24, and ASV 35 (Supplementary Figure S3). Bacteroides plebeius is a common bacterium in human faces, with numerous strains (Kitahara et al., 2005). Furthermore, compared to healthy cats, the relative abundance of B. plebeius was significantly decreased in the fecal microbiome of animals with Feline chronic enteropathy (CE; Marsilio et al., 2019). The genus of Marinifilum is often isolated or detected in coastal sea water and sediments (Na et al., 2009; Ruvira et al., 2013; Xu et al., 2016) and is most likely involved in the hydrolysis and fermentation of proteins, carbohydrates, and lipids (Müller et al., 2018). The members of Ezakiella, a genus of Gram-stain positive, coccus-shaped, anaerobic bacteria are often found in the intestinal and vaginal tracts of healthy people (Patel et al., 2015; Diop et al., 2018). However, based on the very limited reports currently available, it is not clear what role(s) this genus may play in the gut and vagina.

The bacterial genus of Lachnoclostridium is commonly found in the healthy human gut (Tidjani Alou et al., 2016a,b; Traore et al., 2017). But, Lachnoclostridium was also isolated from the gut microbiota of an obese patient (Amadou et al., 2016). A previous study reported that the genus of Romboutsia was predominant in the samples of mammalian gut microbiota (Gerritsen et al., 2018). However, this genus can be treated as one of the gut microbiotal markers for obesity in patients (Zeng et al., 2019). The Fournierella can be found in the fecal samples of humans, and it has also been reported to be significantly associated with intrahepatic fat accumulation (Yaskolka Meir et al., 2021). The Eubacterium fissicatena group is closely associated with obesity and obesity-related metabolic disorders of the host. Previous studies found that medium-, long-, and medium-chain (MLM) structured lipids have anti-obesity effects (Song et al., 2021). After supplementation with MLM structured lipids, the relative abundance of E. fissicatena group were decreased (Yue et al., 2020). The dominant gut bacterial lineages of belugas in this study indicated that they were obese or overweight, which we propose to be an essential environmental adaption strategy for the belugas to defense extremely cold in the arctic waters.

Escherichia-Shigella, Clostridium sensu stricto 1, and Enterococcus were the dominant bacterial lineages in the fecal samples of Pacific white-sided dolphins. However, only one ASV (ASV3) was assigned to Escherichia-Shigella. The results of phylogenetic analysis indicate that ASV3 is closely related to either Cronobacter turicensis or Erwinia iniecta (Supplementary Figure S4). Cronobacter turicensis is an opportunistic foodborne pathogen, which can lead to severe disease manifestations, such as brain abscesses, meningitis, necrotizing enterocolitis, and systemic sepsis (Lehner and Stephan, 2004; Stephan et al., 2011). Only one paper has reported that E. iniecta possess virulence to the Russian wheat aphid (Campillo et al., 2015). The most abundant ASV, ASV2, was assigned to Clostridium sensu stricto 1 at the genus level and was closely related to Clostridium perfringens based on the phylogenetic analysis (Supplementary Figure S5). ASV11 was present in the fecal samples of Pacific white-sided dolphins and was closely related to Enterococcus lactis (Supplementary Figure S6), which can produce an antimicrobial substance that is stable over a wide range of pH (2–10) and after heating at 100°C for 15min (Ben Braïek et al., 2018a). Therefore, E. lactis is classified as a probiotic (Sharma et al., 2012; Ben Braïek et al., 2018b). Clostridium perfringens is a typical anaerobic, Gram-positive, spore-forming bacillus that can secrete more than 20 virulence toxins and has been associated with intestinal diseases ranging in severity from diarrhea to necrotizing enterocolitis and myonecrosis in both animals and humans (Ohtani and Shimizu, 2016; Kiu and Hall, 2018). Clostridium perfringens was found to be dominant in the fecal samples of the common bottlenose dolphin of this study. The other two dominant bacterial lineages were Vibrio and Morganella. Phylogenetic analysis showed that ASV40 was closely related to Vibrio fluvialis (Supplementary Figure S7), and ASV 5 was closely related to Morganella morganii (Supplementary Figure S8). Vibrio fluvialis and M. morganii are both pathogenic bacteria. Vibrio fluvialis is a halophilic, Gram-negative pathogen commonly found in coastal environments including sea water and seafood. Several outbreaks and sporadic cases of acute diarrhea caused by V. fluvialis have been reported extensively (Igbinosa and Okoh, 2010; Chowdhury et al., 2012, 2013; Liang et al., 2013). Morganella morganii is a gram-negative bacterium found in the gut of humans and is associated with many diseases including cellulitis, abscessation, sepsis, bacteremia, and diarrhea (Falagas et al., 2006; Kim et al., 2007; Liu et al., 2016). These results indicated that the Pacific white-sided dolphins and common bottlenose dolphin are in poor health condition.

Indicator species refers to a species constrained to one or a few habitat types, and as “specialists” have a greater susceptibility to local or regional extinction. These organisms can also potentially represent a better ecological indicator of environmental change than habitat generalists (Carignan and Villard, 2002). For example, in the vegetation studies, the species of plant that are absolutely inclined to present in a single or, at the most, a few vegetation types are identified as indicator species, and this method is beneficial to the identification of vegetation types in research surveys (Chytrý et al., 2002). To find out the association between single species and one or several groups of sites representing habitat types, community types, or other categories, the indicator value index (IndVal) is used to assess these relationships (De Cáceres and Legendre, 2009). Because there were an uneven number of fecal samples from the different groups of marine mammals in this study, we used the index of IndVal.g to eliminate the effect of different numbers in each group. According to the set threshold (IndVal.g>0.95, p<0.001), our results show that the gut microbial communities of belugas harbor 28 indicator species representing by ASVs. While five and two ASVs were identified in the gut microbial communities of Cape fur seals and common bottlenose dolphin, respectively. However, only a single ASV was detected in the gut microbiota of Pacific white-sided dolphins. Heatmap results based on these indicator species show that different species of marine mammals can harbor distinct microbial communities with the exception of Pacific white-sided dolphins and common bottlenose dolphin, which cannot be separated. Interestingly, in our study, the microbial phyla with the highest relative abundance in the fecal samples of common bottlenose dolphin were Proteobacteria (Gamma) and Firmicutes at the phylum level, and Vibrionaceae, Enterobacteriaceae, Clostridiaceae 1, and Peptostreptococcaceae at the family level. These findings were consistent with previous studies (Soverini et al., 2016;Suzuki et al., 2019 ; Robles-Malagamba et al., 2020). The Fusobacteriaceae of Fusobacteria was also documented in the previous studies as a dominant bacterial lineage (Suzuki et al., 2019; Robles-Malagamba et al., 2020), however, we barely detected Fusobacteria in the gut microbial communities of common bottlenose dolphin but found that it dominated the fecal samples of Cape fur seals. Moreover, at the genus level, previous work has shown that the genera Cetobacterium (70.9%), Clostridium XI (15.3%), and Clostridium sensu stricto (4.2%) had the highest relative abundance in the fecal samples of common bottlenose dolphins, while in our microbial datasets, Vibrio (Vibrio fluvialis), Morganella (Morganella morganii), and Clostridium sensu stricto 1 (Clostridium perfringens) were the dominant bacterial lineages, all of which are pathogenic bacteria, especially the C. perfringens, which have been reported to be the causative agent of common bottlenose dolphins (Buck et al., 1987; Jaing et al., 2015). Furthermore, we also detected lager number of C. perfringens in the fecal samples of Pacific white-sided dolphins, because the C. perfringens can enter other dolphins via skin wounds, which serve as the focus for these bacteria to penetrate and spread (Buck et al., 1987). Meanwhile, the spread of same bacteria between Pacific white-sided dolphins and common bottlenose dolphin may contribute the similarity of microbial community structure between them. Unfortunately, we did not find any published reports on the gut microbiota of Pacific white-sided dolphins. For a comparative and comprehensive understanding, further studies of the microbial communities of Pacific white-sided dolphins from different places should be conducted in the future.