Bacterial Community Research of Migratory Bird in Chifeng, Neimeng via High-Throughput Sequencing


 Background: In August 2018, a large number of migratory birds died in Chifeng, Neimeng. We were entrusted by local animal disease prevention and control center to collect the migratory bird epidemic materials and their living environmental water, and in 2019, we collect the local migratory bird stool and their living environmental water again. The bacterial communities in migratory bird epidemic materials, water and aquatic plants are profifiled by high-throughput sequencing of the V3–V4 hypervariable region of 16S rDNA gene.Results: We found that the dominant phylum between migratory bird epidemic materials, water and aquatic plant were Proteobacteia, Bacteroidetes, Firmicutes, Fusobacteria and Verrucomicrobia in 2018. One year later, we found that the dominant phylum between migratory bird stool, water and aquatic plant were Proteobacteia, Firmicutes, Cyanobacteria, Bacteroidetes. The relative abundance among bacterial phylum notably differed between two years. The relative abundance of Fusobacteria and Verrucomicrobia were higher in samples in 2018, while that of Cyanobacteria was higher in water, 2019. The relative abundance of Fusobacteria in migratory bird epidemic materials began to decline in the later period over time. At genus level, the relative abundance of Vibrio, Clostridium and other patnogenic bacteria decreased markedly and disappeared in 2019. The salt content and pH show a downward trend.Conclusions: Differences in diet and geographical location can lead to diversification in migratory bird intestinal flora. It is necessary to pay attention to diversification in intestinal flora of migratory birds, especially the abundance of vibrio in intestine. The overall structure of intestinal flora and relative abundance changes of various species are showed intuitively by 16S rDNA amplicon. But comparing to PCR which using specific primers, the accuracy and sensitivity are relatively poor. In the case of purposeful detection, it can be used in combination with both 16S rDNA amplicon and PCR which using specific primers. They are more accurately, because of supplement each other.

Migratory birds were malnutrition, wasting, diarrhea, and a large number of them were death in August 2018, Chifeng, Neimeng. Among dead birds were red duck, red-billed gull, black-winged sandpiper, silver gull, spot-billed Ducks and other precious birds. We found some birds with heart and intestine atulence, bleeding in the intestinal mucosa, enlarged liver, but no abnormalities in the stomach and spleen by anatomized the migratory birds epidemic materials. We analyzed migratory bird epidemic materials and plankton in local environment by 16S rDNA amplicon, and we analyzed local healthy migratory bird stool and plankton from environment in 2109, once again. V. cholerae and V. metschnikovii were found in migratory bird epidemic material. V. parahaemolyticus was not found in Chifeng, Neimeng. V. cholerae was not found in healthy migratory bird stool. Results pH, salt content, total vibrio count and total bacteria count of migratory birds habitat Table 1 showed the variations of pH and salt content during the sampling periods from 2018 to 2019 in Chifeng, Neimeng. Several environmental factors were measured during two years : salinity (0.87-0.008%) and pH (9.159-8.40), decreased slowly over time in Beihekou; salinity (0.076-0.017%) and pH (9.12-8.88), decreased slowly over time in Ganggenghu. The pH of water samples from 2019, except for Beian and Nanyang, which showed a general decrease. The salt content of water samples from 2018 was higher than the second year. The count of total vibrio and bacteria was strong reduction from 2018 to 2019. The pH of Nanyang aquatic plants was slightly lower than Nanyang and the salt content was almost same to Nanyang, but the count of total of vibrio and bacteria were higher than Nanyang.  Figure S1, Table 3 ). According to the weighted unifrac PCoA ( Fig. 3), A2F, A11G, A18C and H2C were clustered together, which implied a high similarity among bacterial communities from different samples. Compared to 2018, stool samples and environmental samples were scattered distribution, didn't clustered together. Only RS2.4 and RS3.1 were clustered together ( Supplement Table S1 ).
The samples OTUs could be divided into three types in 2018 by UPGMA ( Fig. 4). We found that some migratory bird epidemic materials were closely related to plankton from environment. The samples OTUs could also be divided into two categories in 2019, but we found that genetic relationship in gure that feces had close relationship with other planktonic bacteria, except water planktonic bacteria RS2.1 and RS2.3.
According to the heatmap ( Fig. 5 ), the abundance of Plesiomonas in A13Y was high in 2018, which made cluster distance between A13Y and others was far. The samples showed a diverse structure, and each sample with its own unique higher abundance genus in 2019. The LEfSe showed resulted in signi cant differences in the samples from 2018 to 2019, See Supplement Figure S2 and Supplement Table S2 for details.

Detection of pathogenic Vibrio
Vibrio was identi ed by PCR as V. cholerae and V. metschnikovii. V. cholerae identi ed by PCR did not contain CTX, rtxA, but contained hlyA and was identi ed as non-O1 and O139 V. cholerae by serum in 2018. However, Vibrio was identi ed as V. metschnikovii by PCR, and V. cholerae was not detected in 2019 ( Table 4 ).

Discussion
In recent years, the research on the intestinal ora had gradually increased, but the research on the structure of migratory animals intestinal ora, except swans and ducks., was still relatively poor. Dietary habits of animals had a great relationship with the habitat, plankton from habitat could be used as the indicator of migratory birds health status [11,12]. We found that more than 20 kinds of known phylum and more than 1300 kinds of known genus could be carried in migratory birds by 16S rDNA analysis of migratory birds epidemic material and stool ( Supplement Table S3 ). Our results were consistent with studies on the intestinal bacteria of swan and mallard, the dominant phylum of them were Proteobacteia, followed by Bacteroidetes and Firmicutes [13,14]. At the same time, the water environment as the main source of migratory bird diet, it played an important role in the composition of migratory bird gut microbiota [15]. The results implied that the microbial species richness in environment was higher than that in migratory birds, there were differences between the composition structure of the microbiota, migratory bird epidemic materials and environmental samples in 2018, were no signi cant. The differences in the composition of the microbiota between various samples were not signi cant in 2019, while the differences might be ascribed to changes in relative abundance.
Except common phylum in intestinal tract, the relative abundance of Fusobacteria and Verrucomicrobia in epidemic materials was relatively higher than other samples, while the relative abundance of them in stool was poor. Among them, the relative abundance of Fusobacteria in migratory birds intestine was relatively higher than that in other samples in 2018, and the relative abundance in the habitat was extremely poor, which was the same as habitat in 2019. Fusobacteria was not the main phylum in environment, but they were carried by migratory birds from other places to there. It was well colonized in migratory birds intestinal tract, and little planktonic bacteria have formed. This might be one of the reasons why the relative abundance of Fusobacteria was greatly reduced. Gut microbiota was certain differences between feces and intestines of migratory birds. However, Fusobacteria should not be the main phylum in the intestine. When Fusobacteria was signi cantly elevated in the intestine, it would cause intestinal ora imbalance, which may cause diarrhea or other chronic diseases [16].
We also found that Fusobacteria caused signi cant difference between intestinal samples and other samples by LEfse, 2018. Fusobacterium and Cetobacterium were Fusobacteria, and the relative abundance was relatively higher than others at genus level of samples in 2018. Heatmap showed that abundance of Fusobacterium and Cetobacterium in Bird NO.3 was higher than other birds. Fusobacterium and Cetobacterium were detected in bird NO.3 intestine at same time, and abundance of them was relatively high, and Cetobacterium was also detected in bird NO.3 lungs, which also had a high abundance.
However, the relative abundance of Fusobacteria in epidemic materials began to decline in the later period over time. Abundance of Fusobacterium and Cetobacterium in birds NO.18 and NO.19 in intestine was little, and it was not the dominant genus of birds NO.18 and NO. 19. We believed that other phylum have begun to breed, especially Proteobacteia, resulting in the relative abundance of Fusobacteria reduced. We found that abundance of Vibrio, Aeromonas and Plesiomonas in bird NO.18 intestine was relatively higher than others, and abundance of Bacteroides, Megamonas, Lactococcus and Parabacteroides in bird NO.19 intestine was relatively higher than others by heatmap. Among them, the relative abundance of Vibrio, Aeromonas, Plesiomonas and Bacteroides was also higher, which was the dominant genus, and they were Proteobacteia at phylum level. There were signi cant differences from the organ samples and other samples, because of Plesiomonas. There was a relatively high abundance in bird NO.13 pancreas, but the relative abundance in environment was also extremely low ( Fig. 5 and Table 3 ). Vibrio and Aeromonas were detected in migratory bird habitats and epidemic materials in 2018. Except for bird NO.2 lung and bird NO.18 intestine, the abundance of Vibrio in the remaining samples was slightly higher than that of Aeromonas ( Fig. 5 ). In 2019, the abundance of Vibrio and Aeromonas decreased signi cantly. Abundance of Aeromonas in water was higher than that of other samples, while Vibrio only with a lower abundance in stool. And in 2018, we isolated a large number of non-O1/O139 V. cholerae and V. metschnikovii in epidemic materials. It's interesting that we only found Vibrio with extremely low abundance in migratory feces by 16S rDNA amplicon, but we found a large number of V. metschnikovii in migratory bird feces and water in 2019 by PCR, which with speci c primer. There were differences with the results of 16S rDNA amplicon. We thought that the bacteria were enriched after feces were streaked onto the solid medium. And, during the process of DNA binding with universal primers, defective combintion would also lead to deviations in the results [17]. The PCR detection of speci c primers used speci c primers of V. metschnikovii for identi cation, and there would be no defective combintion of bacterial DNA and primers [18].
The abundance of Vibrio on surface of aquatic plants was higher than water in 2018. It might be a large number of plankton in water. And Vibrio without advantage in abundance. Abundance of Vibrio in lungs was higher than intestines and other organs. Vibrio infection was usually caused by eating food contaminated with Vibrio, which changed the permeability of the small intestinal epithelial cells, resulting in abnormal Na / K + pumps and diarrhea [19,20].
If migratory birds ate foods infected with V. cholera or V. metschnikovii, which caused V. cholerae to be infected in intestine, and then carried to Chifeng, the intestine was the initial invasion site of bacteria. The abundance of bacteria should be the highest, but the abundance of V. cholera in lungs was high. We speculated that V. cholera were not invaded by the intestine, they might be invaded by the lungs, and then spread throughout body. Due to the relative abundance, it was not ruled out that relative abundance of Fusobacteria in the intestine was higher, and relative abundance of Vibrio in Proteobacteia was lower. We could also nd the relative abundance of Fusobacteria in bird NO.18 intestine, and abundance of Vibrio increased by heatmap. We speculated that Vibrio was also present in other birds intestines, but it was less than Fusobacteria abundance. And we also veri ed our conjecture by separated and identi ed Vibrio in samples. The abundance of Vibrio in water was relatively poor, and bio lm formation capacity of the bacteria may be strong, resulting in more Vibrio on surface of aquatic plants and relatively few plankton in water. By analyzing the dominant genus in epidemic materials in 2018, it was closer to the dominant bacterial structure of sh [21]. It might be caused by migratory birds ate contaminated sh, opportunistic bacteria invasion, resulting in poor absorption of nutrients, weight loss, intestinal mucosal bleeding and other conditions. Salinity and pH were indicators to assess whether Vibrio could grow well in one place. For V. cholerae disappeared in 2019, we compared the indicators of local water in these two years [22].
It was found that both pH and salinity had decreased, but pH and salinity were also within the range of growth for V. cholerae in 2019. And there are a large number of V. metschnikovii from samples in 2019. It indicated that Chifeng, which suitable for the growth for Vibrio. The decrease of pH and salinity was not the reason for the disappearance of V. cholerae, it might be carried here by migratory birds. Poultry was susceptible for V. metschnikovii, rst discovered in a bird epidemic materials [23]. But, it has not aroused social concern. There was no international report on the pathogenic ability of V. metschnikovii. It with higher adaptability than other kinds of Vibrio.
V. cholerae without any direct lethal toxin, only contained hlyA. It might be opportunistic pathogen in migratory birds and didn't cause migratory birds death rapidly in a short time. It could cause intestinal mucosal bleeding, which was consistent with the pathological characteristics of migratory birds.
Cyanobacteria was the dominant phylum in water, 2109 and the relative abundance of the Cyanobacteria had dropped signi cantly in water, 2108. Cyanobacteria was a common phylum in water, and it was an indicator to assess whether water was polluted. It indicated that there was cyanobacteria pollution in the water, Chifeng [24,25]. We also found that Cyanobacteria which was to be signi cantly different between water and stool at phylum level, and that was unidenti ed_Cyanobacteria at genus level. However, Enterobacter was to be signi cantly different between feaces and water at genus level. Although Cyanobacteria was polluted in water, it was not impact on migratory birds.

Conclusions
This study provided basic knowledge of the relationship between environmental microbiota and the migratory birds in aquaculture using high-throughput sequencing. In our study, Proteobacteia, Bacteroidetes, Firmicutes were the dominant phylum of gut microbiota in migratory birds from 2018 to 2019. Migratory migratory birds with long distance, it caused gut microbiota to complicate. And it could carry pathogenic bacteria to transfer.
Neimeng, as a natural bacterial storage, was not only suitable for various rare animals to survive here, but also suitable for various microorganisms to survive here. Differences in diet and geographical location could lead to change migratory bird gut microbiota. It was necessary to pay attention to change in migratory birds intestinal ora, especially the abundance of vibrio in intestine.
It could intuitively re ect the relative abundance changes of entire gut microbiota and various species in it by 16S rDNA amplicon. But its accuracy and sensitivity were poor compared to PCR, which using speci c primers. If the relative abundance of the species was very poor, it was eliminated. Abundance of some pathogenic bacteria was very low in organism before it broke out. Once it was eliminated, it would cause a lot of unnecessary trouble and might cause a pandemic. Therefore, 16S rDNA amplicon and PCR, which using speci c primers could be used simultaneously as complementary and more accurate in case of purposeful detection.

Sample Collection
Consequent two years of Chifeng with migratory bird migration backgrounds were selected for sampling.  Sequences were clustered followed by chimera ltering, OTUs for species classi cation, each OTU is considered to represent a species. We picked a representative sequences for each OTU and used the RDP (RDP, http://rdp.cme.msu.edu)classi er to annotate taxonomic information for each representative sequence [27,28]. Based on OTU abundances and taxonomic annotation of OTUs, we obtained relative abundance pro les at the phylum, class, order, family, genus, and species. This makes it easy to understand the overall situation of each classi cation level annotated.
The abundance of different species, species clustering, and sample clustering information for sample are re ected by color gradients of heatmap.It drawed by heatmap.2, gplots package, R.
In the molecular evolution research, studying the evolutionary relationship of the system. Through phylogenetic relationship to reveal the difference between a certain level of OTUs sequence, combined with the species annotation information represented by each OTUs sequence, then built a evolution tree.
According to the results of species classi cation, the dominant species were selected.
The abundances of main 10 genus were used for constructing species classi cation tree, abundance differences and evolution relationships of dominant species in single or multiple samples were understood from the entire classi cation system. We calculated the value of sample Alpha Diversity Index by QIIME software made the corresponding dilution curve.
Dilution curve was a curve, using the relative proportions of various known OTUs in measured 16S rDNA sequence to calculate the expected value of each Alpha diversity index when extracting n (n is less than the total number of measured Reads sequences), and then based on a set of n values ( it's generally a set of equivariance series less than the total sequence number) and the expected value of the corresponding Alpha diversity index [29]. And we made Alpha Diversity Index statistical table. In addition, the 16S rDNA gene amplicon sequence data were also analyzed based on weighted UniFrac distance, alpha diversity (PD index, Chao 1 index, Shannon index, Simpson index) and principal coordinates analysis (PCoA).
According to OUTs evolutionary relationship in the sample, UPMGA was used to draw an evolutionary tree to analyze the relationship between the samples.
LEfSe analysis: LEfSe uses linear discriminant analysis (LDA) to estimate the magnitude of the effect of each component (species) abundance on the difference effect, and nd out the communities or species that have a signi cant difference in sample division. we uses LEfSe Tools in this study [30].

Declarations
Ethics approval and consent to participate