Change in the intestinal bacterial community structure associated with environmental microorganisms during the growth of Eriocheir sinensis

Abstract As an important organ to maintain the host's homeostasis, intestinal microbes play an important role in development of the organism. In contrast to those of terrestrial animals, the intestinal microbes of aquatic organisms are affected by environmental microorganisms (including water microorganisms and sediment microorganisms). In the present study, the compositional differences of intestinal microbes in three representative developmental stages of the Chinese mitten crab (Eriocheir sinensis) were studied. Meanwhile, network association analysis, and visualization of the water microorganisms of the crabs’ habitat, the environment microorganisms in the pond, and the intestinal microbes, was carried out. The results showed that the gut microbiota diversity index decreased continuously with age, and the four bacteria of Aeromonas (Proteobacteria), Defluviitaleaceae (Firmicutes), Candidatus Bacilloplasma (Tenericutes), and Dysgonomonas (Bacteroidetes) were the “indigenous” flora of the crab. In the network‐related analysis with the environment, we found that as the culture time increased, the effect of environmental microorganisms on the intestinal microbes of crabs gradually decreased, and the four “indigenous” bacteria were always unaffected by the environmental microorganisms. The results of this study identified the core bacteria of the crab and, for the first time, studied the relationship between intestinal environmental microorganisms, which will aid the practical production of crabs and will promote research into the relationship between specific bacteria and the physiological metabolism of crabs.

Only one study has examined the relationship between gut microbes and the water environment in different organs of crab (Zhang, Sun, Chen, & Cai, 2016). How intestinal microorganisms in crabs are regulated during the breeding cycle is an unanswered question.
The impact of aquatic ecosystems on aquatic animals is enormous. Chinese mitten crabs mainly live in the bottom water in the breeding pond; therefore, the influence of the sediment cannot be neglected. Previous studies focused on the composition of microbes in the intestine of the crabs and its relationship with the aquatic environment Zhang et al., 2016). However, the associations between sediment microorganisms, water environment microorganisms, and intestinal microbes in crabs are unknown.
Based on previous studies, we summarized the existing research on gut microbes of the crab and drew lessons from the study of intestinal microorganisms in fish and other crustaceans. The gut microbes of crabs were determined at three time points during the breeding cycle using 16s rRNA high-throughput sequencing, together with correlation analysis of environmental microorganisms in the breeding ponds. The results were used to study the process of gut microbe assembly and to explore the shared and differential bacteria between gut microbes and environmental microorganisms during the breeding cycle. Our results provide reliable data support for subsequent studies on microbial community functions and ecological aquaculture techniques.

| Sample collection and illumina highthroughput sequencing of bacterial 16S rRNA genes
From May to September 2017, crab samples were collected from ponds near Yangcheng Lake at three times points (Table 1). The groups of crabs had no significant differences in their specifications.
The collected Chinese mitten crabs (E. sinensis) were transported to the laboratory as quickly as possible. After measuring the crab index, the surface of the crab was washed with sterilized water and 75% alcohol and then left for 3-5 min at room temperature. The entire intestine of the crabs was dissected out and placed in sterile tubes.
The DNA was isolated from the intestines using an EZNA Tissue Water samples were collected at three sampling sites using a sterile container to take a total of 1 L from the pond for the extraction of water environment microorganisms. Crabs live in Yahu Lake, Qinghai province, before entering pond farming; therefore, water was collected from Yahu Lake as the water environment for the crab. Large suspended solids were removed using 5μm qualitative filter paper, and then the filtrate was vacuum-filtered through a 0.22μm polycarbonate membrane (Millipore, Billerica, MA, USA).

| Bioinformatic analysis
To obtain more accurate and reliable results in the subsequent bioinformatic analysis (Fadrosh, Bing, Gajer, & Sengamalay, 2014), the raw data were preprocessed to get clean data using an in-house procedure as follows: The sequences that contained more than two mismatches to the primers or more than one mismatch to the barcode were discarded, and reads of <50 bp were removed. If the window average quality value was <20, the end of the read sequence was truncated from the window, and the reads with a final read length <75% of the original read length were removed. The clean data were clustered with 97% similarity using USEARCH (v7.0.1090) (Edgar, 2013), and the operational taxonomic unit (OTU) sequences were searched against the RDP (ribosomal database) using RDP classifier (v2.2) (Cole, Wang, Fish, & Chai, 2014) with a confidence threshold of 60%. The alpha diversity index for each sample was determined using Chao1 (total species richness), ACE (abundance-based coverage estimator) and the Shannon index in Mothur (v1.31.2) (Schloss, Westcott, Ryabin, & Hall, 2009). Based on the weighted UniFrac distance, to display the differences of OTU composition between the gut, sediment, and water, principal component analysis was used to construct a two-dimensional (2-D) graph to summarize the factors mainly responsible for this difference; similarity was high if two samples are closely located by package "ade4" of software R (v3.1.1).
A heatmap was drawn based on the relative abundance of each species in each sample from the sediment, water, and gut samples. The Kruskal-Wallis test was used to analyze the bacterial composition and abundance (average abundance > 0.01%) in samples. The biological replicates were merged, and finally, the differential flora was visualized using the igraph package in R.

| General analyses of high-throughput sequencing
After removing the low-quality reads, a total of 2,916,923 valid reads and 9,814 OTUs were obtained from 38 samples. Among them, 57,425-87,343 were collected for intestinal bacterial community analysis, and 61,245-87,602 were collected for environmentassociated bacterial community analysis. The rarefaction curves showed that sufficient sampling depth was achieved for each sample ( Figure 1). The alpha diversity score was applied to analyze the complexity of species diversity for the water, sediment, and guts of crabs ( Figure 2). OTU numbers, ACE, Chao, and Shannon index can reflect the species richness of a community, and the rarefaction curve based on the three values could also be used to evaluate whether the produced data are sufficient to cover all species in the community (Schloss et al., 2009). In addition, the Shannon index can reflect the species diversity of the community. As shown in Figure 2, the bacterial community diversity in the environmental samples was higher than that in the gut samples. Moreover, the bacterial community diversity was higher in female crabs than in the male crabs.

F I G U R E 1 Rarefaction analysis of the samples
During the breeding cycle, the bacterial diversity showed a decreasing trend, with the bacterial community becoming more stable as the crab aged.

| The gut microbial composition in crab
To determine the changes in the intestinal microbiota in response to the culture environment during the breeding period, we performed hierarchical clustering using the abundance of each species at the family level. The results showed that the gut microbiota, sediment, showed low abundance in female crabs.
The four dominant bacteria showed significant changes in abundance during the life cycle of the crabs, with a similar abundance in juvenile and adult stage, but a higher abundance in the youth stage.
Although male and female crabs showed consistent trends, the four main bacteria were consistently more abundant in the female crabs than in the male crabs.

| Unique and shared bacteria groups between gut microbiota and the environment
The Kruskal-Wallis differential test was performed on the data, and  still have no effect on the environmental microorganisms.
F I G U R E 3 Frequency distribution of bacterial phyla in water, sediment, and intestines. QS represents the original water samples. YF1, YF2, and YF3 represent juvenile, young, and adult, respectively, in the female group. YM1, YM2, and YM3 represent juvenile, young, and adult, respectively, in the male group. YS1, YS2, and YS3 represent May, July, and September samples, respectively, in the water group. YT1, YT2, and YT3 represent May, July, and September samples, respectively, in sediment group

| D ISCUSS I ON
Aquatic animals have greater differences in their life history and living environment than terrestrial animals (Wong & Rawls, 2012).
Their intestinal microflora structure is more diverse and complex than terrestrial animals. For aquatic organisms, the factors that can affect the structure of the intestinal flora are the genetic background, physiological state, and biological habits (Ni, Yu, Zhang, & Gao, 2012;Xiong, Dai, Zhu, & Liu, 2016;Xiong, Wang, Wu, & Qiuqian, 2015). The physiological state can include the health status of the aquatic organisms, their feeding status, and age. Studies on the health status of aquatic organisms have commonly detected differences in sex and body weight (Li, Yan, Einar, & Wu, 2016) and in feeds (   the Defluviitaleaceae, which belongs to the Firmicutes, has not been found in the intestinal tract of aquatic organisms and was found in the human gastrointestinal tract disease (Dong, Wang, Liu, & Yang, 2017). In this study, the population abundance of Defluviitaleaceae in young crabs was higher than that in juvenile and adult crabs, while the growth curve of these bacteria in the young population showed an exponential increase. Therefore, it remains to be further studied whether this bacteria group plays a role in the intestinal tract of the crabs to maintain physiological functions for gastrointestinal stability and nutrient absorption.

Two species, Candidatus Bacilloplasma (Tenericutes) and
Dysgonomonas (Bacteroidetes), were the dominant of intestinal bacteria in this study and previous study, at the same time, it is worth mentioning that Candidatus Bacilloplasma is the "indigenous" populations of the intestinal, which is also consistent with previous studies on the intestinal bacteria of crustaceans (Kostanjsek, Strus, & Avgustin, 2007). In the study of the gut microbiota of shrimp containing pathogenic bacteria, researchers found that the intestinal abundance of Candidatus Bacilloplasma increased significantly in pathogenic shrimp. In the present study, Candidatus Bacilloplasma had the lowest abundance at the youth stage in E. sinensis, but it maintained a high abundance in the juvenile and adult stage. The youth stage is the fastest growing stage and the most advanced stage in terms of the immune system in the breeding cycle, so whether Candidatus Bacilloplasma is a potential pathogen in the intestine of the crabs, and its status and function among the intestinal bacteria remain to be further explored. Bacteroidetes has a corresponding abundance in the intestines of fish and shrimp (Baldo, Riera, Toomingklunderud, & Albà, 2015;Rungrassamee et al., 2014;Zhang, Sun, Chen, & Yu, 2014); however, its function has not been determined. The genus Dysgonomonas has a certain abundance in the gut of cockroaches and termites (Mikaelyan, Thompson, Hofer, & Brune, 2015). It is a common genus of termites intestines and plays an important role in assisting termites to digest lignocellulose, in immunity, and in reproduction (Brune, 2014;Fraune & Bosch, 2010;Scharf, 2015;Su, Yang, Huang, & Su, 2016;Warnecke, Luginbühl, Ivanova, & Ghassemian, 2007;Werren, Baldo, & Clark, 2008). E.
sinensis belongs to the crustacean class and the arthropod phylum, as do termites; and in crabs, the Dysgonomonas genus reached its highest abundance in the youth stage and was low in the juvenile and adult stages. During the sampling, the samples of the juvenile and young populations showed significant differences in their specifications; therefore, whether the function of the bacteria is related to the intestinal nutrient absorption and the level of immunity, and whether they are beneficial bacteria, requires further exploration.
Compared with previous studies, a more special kind of Marinifilum bacterium appeared in the gut microbiota of E. sinensis, with high abundance in juvenile stage, but lower abundance in the young and adult stages. The samples of this study originated from Qinghai Province, which has inland lakes with high salinity and pH, and the low temperatures. However, the intestinal microflora of wild crabs collected at the same sampling time did not contain this bacterium (data not published). Previous studies have found that crabs sampled on Chongming Island have this bacterium, but they were not found in samples from Taihu Lake Zhang et al., 2016). Comparing the changes in the living environment and the fac- Identifying the core group of crab bacteria provides the basis to study the relationship between the specific bacteria and the physiological metabolism of E. sinensis. Furthermore, the impact of food sources on the structure of crab microbial populations requires further research.

ACK N OWLED G M ENTS
This study was funded by the China Agriculture Research System (CARS-46).

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interests.

E TH I C A L S TATEM ENT
Animal use was approved by, and was under the supervision of, the Freshwater Fisheries Research Center, Chinese Academy of Fishery Sciences (FFRC, CAFS).

DATA ACCE SS I B I LIT Y
In the article, HiSeq Illumina sequencing raw sequence reads data: