Intestinal Microbes of Hooded Cranes (Grus monacha) Wintering in Three Lakes of the Middle and Lower Yangtze River Floodplain

Simple Summary Intestinal microbes are critical to host health, and are affected by environmental factors. In this study, we investigated the intestinal microbes of Hooded Cranes wintering at three lakes with different environmental characteristics in the middle and lower Yangtze River floodplain in China, aiming to provide insights into the effects of habitat size and protection status of birds on their intestinal microbes. We found that the Hooded Cranes at the smaller lake had higher intestinal bacterial and fungal diversity than those at the larger lake. In addition, more diverse and abundant pathogens were found in the gut of Hooded Cranes that lived in the relatively poorly protected habitat than those that lived in well-protected habitat. This study contributes a new perspective for understanding the intestinal microbes of wintering migratory waterbirds at different habitats, and will help to understand the survival status of the vulnerable waterbirds at different habitats for their better conservation. Abstract Intestinal microbes participate in life activities of the host, and are affected by external environmental factors. Different habitat sizes and protection status provide different external environmental selection pressures for the same wintering waterbirds, which may be reflected in their intestinal microbes. Hooded Cranes are vulnerable migratory waterbirds with similar numbers wintering at three different lakes in the middle and lower Yangtze River floodplain, Poyang, Caizi, and Shengjin Lakes. Here, we analyzed the characteristics of intestinal bacterial and fungal communities of Hooded Cranes wintering at the three lakes to clarify the effect of habitat size and protection status on intestinal microbes, using high-throughput sequencing technology. Our results showed that community composition and diversity of intestinal microbes were significantly different among lakes with different habitat size and protection status. The Hooded Cranes at Shengjin Lake (small) had higher intestinal microbial alpha-diversity (for both bacteria and fungi) than those at Poyang Lake (large), which might be induced by social behavior of more waterbirds per unit area. The Hooded Cranes at Caizi Lake (relatively poorly protected habitat) had more diverse and abundant intestinal potential pathogens than Shengjin Lake (well-protected habitat). Our results indicated that the environmental pressure of a habitat might affect intestinal microorganisms and more attention might be needed for the vulnerable waterbirds at the habitat of poor protection status.


Introduction
Birds live in complex and dynamic associations with the symbiotic microbial community in their intestine [1,2]. Intestinal microbes are critical to the health of avian hosts, participating in nutrient absorption and affecting their immunity and metabolism [2][3][4]. The initial intestinal microbes of mammals are acquired during birth and the microbiota

Ethics Statement
The collection of fecal samples did not involve human interference with the birds, catching, or hunting. Approval was obtained from the local wildlife protection departments. The research process complies with Chinese current laws.

Site Selection and Fecal Sample Collection
The fecal sampling sites were composed of three different sized lakes, fostering most Hooded Crane populations wintering in China. The three lakes were Poyang Lake (28.24 • -29.46  Province. Among the three lakes, the maximum surface area of Poyang Lake is more than 4000 km 2 in wet season and less than 1000 km 2 in dry season with an average annual water level of 13.88 m [20]. The maximum surface area of Caizi Lake is 226 km 2 when water level peaks at 14.88 m and 114.97 km 2 when water level drops to 10.63 m [21]. The maximum surface area of Shengjin Lake ranges from 140 km 2 in wet season to 75 km 2 in the dry winter [18]. In this study, the difference between the maximum and minimum water surface areas (maximum surface area minus minimum) was taken as the lake habitat size. The habitat size of Poyang, Caizi, and Shengjin Lakes was more than 3000 km 2 , 111.03 km 2 and 65 km 2 , respectively. The regions where the lakes are located belong to the northern subtropical monsoon climate, with average annual temperature and precipitation of 14-16 • C and 1000-4000 mm, respectively. All the lakes are river-connected shallow with grass lands, mud flats, and paddy fields, serving as wintering and stopover sites for migratory birds on the East Asian-Australian flyway ( Figure 1) [22]. The Hooded Cranes at Poyang and Shengjin Lakes mainly feed on Polygonum criopolitanum [23,24], while those at Caizi Lake mainly feed on Oryza sativa [25].

Ethics Statement
The collection of fecal samples did not involve human interference with the birds, catching, or hunting. Approval was obtained from the local wildlife protection departments. The research process complies with Chinese current laws.

Site Selection and Fecal Sample Collection
The fecal sampling sites were composed of three different sized lakes, fostering most Hooded Crane populations wintering in China. The three lakes were Poyang Lake (28.24°-29.46° N, 115.49°-116.46° E) in Jiangxi Province, Caizi Lake (30.43°-30.58° N, 117.01°-117.09° E) and Shengjin Lake (30.15°-30.28° N, 116.58°-117.14° E) in Anhui Province. Among the three lakes, the maximum surface area of Poyang Lake is more than 4000 km 2 in wet season and less than 1000 km 2 in dry season with an average annual water level of 13.88 m [20]. The maximum surface area of Caizi Lake is 226 km 2 when water level peaks at 14.88 m and 114.97 km 2 when water level drops to 10.63 m [21]. The maximum surface area of Shengjin Lake ranges from 140 km 2 in wet season to 75 km 2 in the dry winter [18]. In this study, the difference between the maximum and minimum water surface areas (maximum surface area minus minimum) was taken as the lake habitat size. The habitat size of Poyang, Caizi, and Shengjin Lakes was more than 3000 km 2 , 111.03 km 2 and 65 km 2 , respectively. The regions where the lakes are located belong to the northern subtropical monsoon climate, with average annual temperature and precipitation of 14-16 °C and 1000-4000 mm, respectively. All the lakes are river-connected shallow with grass lands, mud flats, and paddy fields, serving as wintering and stopover sites for migratory birds on the East Asian-Australian flyway ( Figure 1) [22]. The Hooded Cranes at Poyang and Shengjin Lakes mainly feed on Polygonum criopolitanum [23,24], while those at Caizi Lake mainly feed on Oryza sativa [25].   Before sampling, we investigated the specific feeding sites of the Hooded Cranes in the three lakes. Then a telescope was used to search for separate Hooded Crane flocks to ensure that there were no other species. Fresh fecal samples were collected immediately after the cranes foraged. Fresh disposable sterile gloves were used to avoid sample contamination. Fecal samples were collected at more than 5 m intervals to avoid collecting samples from the same individual [26]. All the samples were immediately placed in dry ice in a presterilized cooler after collection, transported to the laboratory, and stored at −80 • C before analysis. When conducting formal experiments, the inner core of the feces was used to avoid possible contamination from the outside.
To control for the influence of sampling time on the experiment, sample collection took place roughly at the same time in the three lakes, during 1 March 2019 to 7 March 2019. At this point, the Hooded Cranes have almost completed the whole wintering phases and are ready to leave for the breeding site. Specifically, we collected 71 fecal samples of Hooded Cranes from the three lakes without hurting the cranes. Twenty-seven samples were collected from Poyang Lake (time: 1 March 2019; habitat: grass lands), 22 from Caizi Lake (time: 7 March 2019; habitat: paddy fields), and 22 from Shengjin Lake (time: 7 March 2019; habitat: grass lands).

Fecal DNA Extraction and Bird Species Determination
DNA extraction was performed using the QIAampR DNA Stool Mini Kit (Qiagen, Germany), and quantified by Nano Drop ND-1000 (Thermo Scientific, Waltham, MA, USA). The extracted DNA was then dissolved in elution buffer (60 µL) and stored at −20 • C.
The BIRDF1-BIRDR1 [27] primers were used to amplify the COI (Cytochrome Oxidase I) gene for identification of the bird species. PCR (Polymerase Chain Reaction) was conducted in a 50µL reaction system consisting of 100 ng fecal DNA from 1 µL DNA template (>20 ng/µL), 22 µL deionized water, 25 µL of 2× EasyTaq PCR Super Mix, 1 µL of the forward primer BIRDF1, and 1 µL of the reverse primer BIRDR1, both at a concentration of 0.2 mM. The specific operation was performed under the following conditions: 5 min of pre-degradation at 95 • C, 30 s of degeneration at 95 • C, 45 s of annealing at 55 • C, 90 s of extension at 72 • C, 35 cycles of repetition of the above process, and a final extension step at 72 • C for 10 min. The PCR results were sent to Sangon Biotech Company for sequencing and BLAST (the Basic Local Alignment Search Tool) analysis (>97% sequence similarity) in NCBI. After confirming that they were from Hooded Cranes, the DNA samples were sent for high-throughput sequencing using an Illumina MiSeq platform.

PCR and Amplicon Library Preparation
The specific process of PCR and amplicon library preparation are described in File S1 (Supplementary Materials) [28][29][30].

Sequence Analysis
The raw sequencing data were preprocessed using the Quantitative Insights into Microbial Ecology (QIIME v.2.0) [31]. Low-quality sequences were removed from our dataset using the deblur algorithm [32]. The remaining high-quality sequences were assigned to amplicon sequence variants (ASVs) based on 100 identity thresholds. The method of VSEARCH was performed to discard the chimers. Then, UNITE database was used to annotate taxonomy to each ASV. Finally, downstream analysis was used to remove the singletons from our dataset.

Statistical Analysis
The difference between intestinal bacteria and fungi among different lake groups was identified by linear discriminant analysis effect size (LEfSe), using nonparametric Kruskal-Wallis rank sum test with the default setting (alpha value: 0.05; effect size threshold: 3.5) [33]. The difference in intestinal microbial community compositions among different groups was analyzed by non-metric multidimensional scaling (NMDS) and analysis of similarity  [34]. Indicator taxa in different groups were identified by indicator analysis, using the Labdsv package in R. The normal distribution of all data was examined by the Kolmogorov-Smirnov test (p > 0.05: normal; p ≤ 0.05: non-normal) in SPSS statistics 20.0. One-way analysis of variance (ANOVA) by Tukey honestly significant difference (HSD) comparison was used for the data that followed normal distribution, and Mann-Whitney-Wilcoxon test was used for the data that followed non-normal distribution (Table S1, Supplementary Materials). The raw data were deposited in the National Center for Biotechnology Information (NCBI) Sequence Read Archive (accession number SRP290730 and SRP307396).

Intestinal Bacterial and Fungal Alpha-Diversity
All the 71 fecal samples were confirmed from the Hooded Cranes after being confirmed by bird species determination (>97% similarity). In this study, 51 samples were selected for analysis according to the ranking of similarity and sequencing quality. Twenty-two samples were from Poyang Lake, 14 from Caizi Lake, and 15 from Shengjin Lake. A total of 800,744 quality-filtered bacterial sequences and 3,164,825 quality-filtered fungal sequences were acquired from the fecal samples of Hooded Cranes at Poyang, Caizi, and Shengjin Lakes. The sequences were grouped into 1526 bacterial and 1282 fungal ASVs which were based on 100% similarity. For intestinal bacteria, the proportion of the ASVs shared by the Hooded Cranes at Poyang and Shengjin Lakes was 21.62%, while the proportion of ASVs shared by the Hooded Cranes at Shengjin and Caizi Lakes was 15.57%. ( Figure S1, Supplementary Materials). For intestinal fungi, the proportion of fungal ASVs shared by the Hooded Cranes at Poyang and Shengjin Lakes was 16.28%. The proportion of the fungal ASVs shared by the Hooded Cranes at Shengjin and Caizi Lakes was 23.57% ( Figure S1, Supplementary Materials).
The alpha-diversity of the intestinal bacteria and fungi of the Hooded Cranes at Shengjin Lake was significantly higher than that at Poyang Lake ( Figure 2). The intestinal bacterial diversity of the Hooded Cranes at Shengjin Lake was significantly higher than those of at Caizi Lake, while the intestinal fungal diversity showed no significant differences with Caizi Lake.

Intestinal Bacterial and Fungal Community Structure
In this study, the intestinal bacteria of the Hooded Cranes at the three lakes were composed of 17 phyla (ranging from 9 to 16 across the three lake samples). The intestinal bacteria phyla of the Hooded Cranes at Poyang and Shengjin Lakes were both dominated by Firmicutes, Proteobacteria, Actinobacteria, and Bacteroidetes, accounting for 98.01% and 95.05% of the total, respectively. The dominant intestinal bacterial phyla of the Hooded Cranes at Caizi Lake were Firmicutes, Proteobacteria, Actinobacteria, accounting for 99.38% of the total. The remaining intestinal bacteria phyla (i.e., Planctomycetes, Fusobacteria etc.) of Hooded Cranes at the three lakes were those with a relative abundance of less than 1%. Except for Bacteroidetes, the distribution of dominant intestinal bacterial phyla among the different lake groups was uneven, as indicated in Table 1. The intestinal bacteria of the Hooded Cranes at the three lakes were composed of 384 genera (ranging from 158 to 321 across the three lake samples). The number of the known bacterial genera (the relative abundance > 0.01%) shared by the Hooded Cranes at Poyang and Shengjin Lake were 52, and the number of the known genera (the relative abundance > 0.01%) shared by the Hooded Cranes at Shengjin and Caizi Lake was 36 (Table S2, Supplementary Materials).
proportion of the fungal ASVs shared by the Hooded Cranes at Shengjin and Caizi Lakes was 23.57% ( Figure S1, Supplementary Materials).
The alpha-diversity of the intestinal bacteria and fungi of the Hooded Cranes at Shengjin Lake was significantly higher than that at Poyang Lake ( Figure 2). The intestinal bacterial diversity of the Hooded Cranes at Shengjin Lake was significantly higher than those of at Poyang Lake, while the intestinal fungal diversity showed no significant differences with Caizi Lake.

Figure 2.
Intestinal microbial alpha-diversity of Hooded Cranes at the three lakes denoted by ASV richness and Shannon index. The panels displayed the interquartile range box plots (first and third quartiles) and whisker (dots indicate manes and the lines inside the box show the medians). Different letters above the boxes indicate highly significant differences according to Tukey HSD test (at p < 0.01 level). The same letters above the boxes represent no significant difference (at p > 0.05 level). PY: Poyang Lake; CZ: Caizi Lake; SJ: Shengjin Lake; B: bacteria; F: fungi.

(PY|SJ|CZ) (PY-SJ|SJ-CZ) (PY|SJ|CZ) (PY-SJ|SJ-CZ)
The intestinal fungi of the Hooded Cranes at the three lakes were composed of 11 phyla (ranging from 7 to 9 across the three lake samples). The intestinal fungal phyla of the Hooded Cranes at Poyang and Shengjin Lakes were dominated by Ascomycota, Basidiomycota, Mortierellomycota, and other phyla, accounting for 74.64% and 97.97% of the total, respectively. The intestinal fungal phyla of the Hooded Cranes at Caizi Lake were Ascomycota, Basidiomycota, and other phyla, accounting for 98.15% of the total. The remaining intestinal fungal phyla (i.e., Rozellomycota, Chytridiomycota etc.) of Hooded Cranes at the three lakes were those with a relative abundance of less than 1%. The three known phyla also showed obvious differentiation in different lake groups ( Table 1). The intestinal fungi of the Hooded Cranes at the three lakes were composed of 268 genera (ranging from 124 to 212 across the three lake samples). The number of the known fungal genera (relative abundance > 0.01%) shared by the Hooded Cranes at Poyang and Shengjin Lakes was 32, and the number of the known genera (relative abundance > 0.01%) shared by the Hooded Cranes at Shengjin and Caizi Lakes was 44 (Table S3, Supplementary Materials).
Intestinal bacterial and fungal community composition of Hooded Cranes among the different lake groups was identified and compared by NMDS and ANOSIM analysis. The results showed a significant separation between different groups for both bacterial and fungal communities, as indicated in Figure 3 and Table 2 (ANOSIM; p < 0.001). However, the intestinal bacterial community composition showed more obvious divergence than the fungal community composition. In addition, ANOSIM indicated that the intestinal fungal community composition of Hooded Cranes at Poyang and Shengjin Lakes (R = 0.2110) was more similar than those at Caizi and Shengjin Lakes (R = 0.7160), though only a little difference was found in the intestinal bacterial community composition between the two lake groups. fungal communities, as indicated in Figure 3 and Table 2 (ANOSIM; p < 0.001). However, the intestinal bacterial community composition showed more obvious divergence than the fungal community composition. In addition, ANOSIM indicated that the intestinal fungal community composition of Hooded Cranes at Poyang and Shengjin Lakes (R = 0.2110) was more similar than those at Caizi and Shengjin Lakes (R = 0.7160), though only a little difference was found in the intestinal bacterial community composition between the two lake groups.

Habitat-Specific Bacterial and Fungal Microbiota
LEfSe analysis revealed specific intestinal bacterial and fungal taxa that differed between the Hooded Cranes at different lake groups. For Poyang and Shengjin Lake group, there were 1 phylum, 1 class, 1 order, 1 family, and 2 genera of intestinal bacteria significantly abundant in the samples from Poyang Lake, while 3 phyla, 8 classes, 8 orders, 15 families and 11 genera of intestinal bacteria were significantly abundant in the samples from Shengjin Lake ( Figure 4A). Furthermore, the Shengjin Lake samples also harbored

Habitat-Specific Bacterial and Fungal Microbiota
LEfSe analysis revealed specific intestinal bacterial and fungal taxa that differed between the Hooded Cranes at different lake groups. For Poyang and Shengjin Lake  Figure 4A). Furthermore, the Shengjin Lake samples also harbored more differential of intestinal fungal taxa than Poyang Lake samples. There were 1 phylum, 5 classes, 7 orders, 12 families, and 9 genera of intestinal fungi significantly abundant in the samples from Shengjin Lake, while 1 phylum, 2 classes, 2 orders, 3 families and 3 genera of intestinal fungi were significantly abundant in the samples from Poyang Lake ( Figure 4B).
Animals 2021, 11, x FOR PEER REVIEW 8 of 16 more differential of intestinal fungal taxa than Poyang Lake samples. There were 1 phylum, 5 classes, 7 orders, 12 families, and 9 genera of intestinal fungi significantly abundant in the samples from Shengjin Lake, while 1 phylum, 2 classes, 2 orders, 3 families and 3 genera of intestinal fungi were significantly abundant in the samples from Poyang Lake ( Figure 4B).

Figure 4. LEfSe analysis of the intestinal bacterial biomarkers (A) and fungal biomarkers (B) of
Hood cranes at Poyang and Shengjin Lake group. Identified phylotype biomarkers were ranked by effect size and the alpha value was < 0.05. Each filled circle represents one biomarker. Cladogram represents the taxonomic hierarchical structure from phylum to genus of the biomarkers identified; Red, phylotypes statistically overrepresented in gut of Hooded Cranes at Poyang Lake; Green, phylotypes statistically overrepresented in gut of Hooded Cranes at Shengjin Lake; Yellow, phylotypes for which relative abundance is not significantly different among three lakes. PY: Poyang Lake; SJ: Shengjin Lake; B: bacteria; F: fungi.
For Caizi and Shengjin Lake group, there were 1 phylum, 1 class, 2 orders, 4 families, and 6 genera of intestinal bacteria significantly abundant in the samples from Caizi Lake, whereas for Shengjin Lake samples, 2 phyla, 6 classes, 6 orders, and 14 families and 12 genera were significantly abundant ( Figure 5A). Furthermore, there were 2 phyla, 2 classes, 5 orders, 7 families, and 8 genera of intestinal fungi significantly abundant at the Caizi Lake samples, while 2 phyla, 2 classes, 4 orders, 10 families and 11 genera of intestinal fungi were significantly abundant in the samples of Shengjin Lake ( Figure 5B). Identified phylotype biomarkers were ranked by effect size and the alpha value was < 0.05. Each filled circle represents one biomarker. Cladogram represents the taxonomic hierarchical structure from phylum to genus of the biomarkers identified; Red, phylotypes statistically overrepresented in gut of Hooded Cranes at Poyang Lake; Green, phylotypes statistically overrepresented in gut of Hooded Cranes at Shengjin Lake; Yellow, phylotypes for which relative abundance is not significantly different among three lakes. PY: Poyang Lake; SJ: Shengjin Lake; B: bacteria; F: fungi.

Figure 5. LEfSe analysis of the intestinal bacterial biomarkers (A) and fungal biomarkers (B) of Hood cranes at Caizi and
Shengjin Lake group. Identified phylotype biomarkers were ranked by effect size and the alpha value was < 0.05. Each filled circle represents one biomarker. Cladogram represents the taxonomic hierarchical structure from phylum to genus of the biomarkers identified; Red, phylotypes statistically overrepresented in gut of Hooded Cranes at Caizi Lake; Green, phylotypes statistically overrepresented in gut of Hooded Cranes at Shengjin Lake; Yellow, phylotypes for which relative abundance is not significantly different among three lakes. CZ: Caizi Lake; SJ: Shengjin Lake; B: bacteria; F: fungi.
Indicator analysis was used to identify the indicator taxa with respect to each evaluated group, which was conducted at the consensus genera and species level ASVs. If the indicator value is close to 1, which means the indicator taxonomy is a good indicator of the intestinal microbes of Hooded Cranes at the involved lake. However, if the indicator value is close to 0, it means that it is a poor indicator for the lake.
For Poyang and Shengjin Lake group, this analysis revealed 10 ASVs of Hooded Cranes at Shengjin Lake. However, only 1 ASV of Hooded Cranes at Poyang Lake was found as the rest of 9 ASVs with indicator value under 0.6. The only indicator taxa that had a high indicator value (>0.6) of Hooded Cranes at Poyang Lake was ASV372 Lactobacillus acidipiscis (51.87%). The indicator taxa of Hooded Cranes at Shengjin Lake were composed of 2 classes, 1 family, 2 genera and 4 species (Table 3).
For Shengjin and Caizi Lake group, a total of 10 ASVs of Hooded Cranes at Shengjin Lake and 10 ASVs of those at Caizi Lake were revealed. The indicator taxa of Hooded Cranes at Shengjin Lake consisted of 2 genera and 6 species. The indicator taxa of Hooded Cranes at Caizi Lake were composed of 2 families, 3 genera and 5 species (Table 4).

Figure 5. LEfSe analysis of the intestinal bacterial biomarkers (A) and fungal biomarkers (B) of Hood cranes at Caizi and
Shengjin Lake group. Identified phylotype biomarkers were ranked by effect size and the alpha value was < 0.05. Each filled circle represents one biomarker. Cladogram represents the taxonomic hierarchical structure from phylum to genus of the biomarkers identified; Red, phylotypes statistically overrepresented in gut of Hooded Cranes at Caizi Lake; Green, phylotypes statistically overrepresented in gut of Hooded Cranes at Shengjin Lake; Yellow, phylotypes for which relative abundance is not significantly different among three lakes. CZ: Caizi Lake; SJ: Shengjin Lake; B: bacteria; F: fungi.
Indicator analysis was used to identify the indicator taxa with respect to each evaluated group, which was conducted at the consensus genera and species level ASVs. If the indicator value is close to 1, which means the indicator taxonomy is a good indicator of the intestinal microbes of Hooded Cranes at the involved lake. However, if the indicator value is close to 0, it means that it is a poor indicator for the lake.
For Poyang and Shengjin Lake group, this analysis revealed 10 ASVs of Hooded Cranes at Shengjin Lake. However, only 1 ASV of Hooded Cranes at Poyang Lake was found as the rest of 9 ASVs with indicator value under 0.6. The only indicator taxa that had a high indicator value (>0.6) of Hooded Cranes at Poyang Lake was ASV372 Lactobacillus acidipiscis (51.87%). The indicator taxa of Hooded Cranes at Shengjin Lake were composed of 2 classes, 1 family, 2 genera and 4 species (Table 3). For Shengjin and Caizi Lake group, a total of 10 ASVs of Hooded Cranes at Shengjin Lake and 10 ASVs of those at Caizi Lake were revealed. The indicator taxa of Hooded Cranes at Shengjin Lake consisted of 2 genera and 6 species. The indicator taxa of Hooded Cranes at Caizi Lake were composed of 2 families, 3 genera and 5 species (Table 4).

Discussion
Intestinal microbes are thought to be affected by the habitat of host. However, due to the complexity of the habitat, there are few studies on how the habitat affects intestinal microbes. This study investigated the effects of habitat size and protection status on the intestinal microbes of Hooded Cranes.
Divergence in intestinal microbial community composition and alpha-diversity was found between the Hooded Cranes at two different size lakes with a similar habitat protection status (Figures 2 and 3). The Hooded Cranes at the small lake, Shengjin Lake, had significantly higher alpha-diversity of both intestinal bacteria and fungi than those at the large lake, Poyang Lake. Poyang Lake fosters a higher number of waterbird species and individuals (approximately 87 species and more than 400,000 individuals) than Shengjin Lake (more than 80 species and up to 70,000 individuals) [35,36]. However, we found that the number of waterbird species of per unit area of the two lakes was 0.03 (87/3000) species/km 2 and 1.23 (80/65) species/km 2 , respectively. In addition, the number of waterbirds per unit area of the two lakes was approximately 133 (400,000/3000) ind./km 2 and 1056 (70,000/65) ind./km 2 , respectively. Thus, the waterbird diversity (a comprehensive metric considering both species and relative abundance of the hosts) of Shengjin Lake per unit area was higher than that of Poyang Lake. As waterbird is a mobile entity, there could be a higher frequency of encounters and more interactions among the waterbirds at Shengjin Lake. Every year, many geese (e.g., Bean Geese, Greater White-Fronted Geese, and Swan Goose) overwinter at Shengjin and Poyang Lakes [15,37]. Previous studies have shown that there was the sharing intestinal microbes between the Hooded Cranes and sympatric other waterbirds, such as the Greater White-Fronted Goose [38], and a recent study demonstrated that the social behaviors (direct or indirect contact) between the Hooded Cranes and the sympatric Bean Goose caused the intestinal bacteria to spread and increased the diversity of intestinal bacteria for both [11]. The possible routes of crosstransmission included feather contamination, physical contact, involuntary fecal feeding, air, water, soil, or other media [39,40]. In this study, the genus of Megamonas and Turicibacter which were identified as the dominant genus of several goose were both found in the gut of the Hooded Cranes at the two lakes [7,13] (Table S2, Supplementary Materials). The relative abundance of the two genera in the gut of the Hooded Cranes at Shengjin Lake was higher than that of Poyang Lake, indicating that the interaction of intestinal microbial communities through social behaviors might be crucial for changing the intestinal microbial diversity observed in the Hooded Cranes at Shengjin Lake [41]. Furthermore, the relative abundance of the indicator species (Lactobacillus acidipiscis) in the gut of Hooded Cranes at Poyang Lake was significantly higher than that of the Hooded Cranes at Shengjin Lake, which was as high as 51.87% (Table 3). L. acidipiscis is a potential probiotic that can enhance digestion and suppress the development of certain disease [42,43]. However, the genus Streptococcus and Fusarium were more abundant in the gut of the Hooded Cranes at Shengjin Lake (Figure 4), both of which might cause disease in the host [44,45]. One effect related to the ecology of animal diseases is called amplification effect, which refers to when increase the species diversity in a community of potential animal hosts for zoonotic pathogens may increase the spread of pathogens among the hosts and increased risk of disease [46]. An example of disease amplification with increased species diversity was that when interspecific transmission is greater than intra-species transmission, the combination of species is more likely to develop pathogens than any single species combination [47]. The possible potential mechanisms include an increase in the abundance of the vectors of the pathogen (vector increase) and an increase in the frequency of encounters between the primary host individuals (encounter increase) [47]. In this study, the population size of Hooded Cranes wintering at the two lakes was small (both approximately 300 individuals). Thus, the influence of interspecific transmission was considered to be greater than intra-species transmission for the Hooded Cranes at the two lakes, which was in accord with the amplification effect example. According to the amplification effect hypothesis, the higher waterbird diversity per unit area might increase the risk of disease [46], which was consistent with our finding. Previous findings showed that dietary, and other external environment factors are important factor for shaping the gut microbial structures for the same host species [13]; however, the Hooded Cranes at Poyang and Shengjin Lakes both mainly feed on P. criopolitanum [23,24], indicating the similar diet at the two lakes. Thus, we considered the diet to be a less important factor for this lake group in our study, suggesting that other external environment factors, such as social behavior could be the main possible influencing factors.
Divergence in intestinal microbiota community composition and intestinal bacterial alpha-diversity were also found between the Hooded Cranes at the two lakes with similar size but different in protection status (Figures 2 and 3). The Hooded Cranes at Shengjin Lake had significantly higher intestinal bacterial alpha-diversity than those at Caizi Lake, which had a relatively poor protection status. In the past 30 years, the development of economic aquaculture at Caizi Lake, such as crab farming, has led to extensive nibbling on the roots and tubers of plants that could serve as the food sources of Hooded Cranes [14]. On the other hand, large-scale reclamation of farmland has degraded the wetland area of Caizi Lake and the Hooded Cranes at Caizi Lake began to mainly eat O. sativa at the farmland [25]. On the contrary, in recent years, Shengjin Lake has been implemented many effective wetland protection and management measures, including wetland vegetation restoration and banning of seine fishing [48,49]. Therefore, the wetland of Shengjin Lake has remained well protected, and the food resources of the Hooded Cranes are better guaranteed. A study showed that the diet of Hooded Cranes at Shengjin Lake was more diverse and even than that of those at Caizi Lake [25]. The habitat suitability of Shengjin Lake was better than that of Caizi Lake [50]. A previous study on wild black howler monkeys also showed that in less disturbed and relatively primitive habitats, the more diversified diet of black howler monkeys seemed to promote the acquisition and maintenance of various intestinal microbes, while in suboptimal environments, the relatively simple diet tended to reduce diversity [12]. The possible mechanism was that diverse diets provide diverse feeding niches to support diverse microbial taxa and functional taxa [51]. Therefore, a diverse diet might promote the diversity of intestinal microbes, which was consistent with our results. Moreover, the Hooded Cranes at Shengjin Lake mainly fed on high fiber P. criopolitanum in the wild. Compared with O. sativa, the nutrient content of wild food resources (such as the roots and leaves of certain plants) is relatively lower [52]. The higher diversity of intestinal bacteria might be beneficial for the Hooded Cranes at Shengjin Lake to obtain nutrients effectively from indigestible and nutrient-limited food [53]. Indeed, our study showed that the Hooded Cranes at Shengjin Lake harbored more abundant genus, such as Bacillus, Clostridium and Paenibacillus which can metabolize cellulose and improve the degradation of non-starch polysaccharides ( Figure 5) [54,55]. In addition, we found that the genera Anaerorhabdus and Ewingella were more abundant in the gut of the Hooded Cranes at Caizi Lake, which might be detrimental for the host and increase the risk of host infection, respectively ( Figure 5) [56,57]. Moreover, many indicators (Ewingella americana, Epicoccum nigrum, Enterococcus cecorum) with a high indicator value of Hooded Cranes at Caizi Lake were also potential pathogenic microbes (Table 4) [57][58][59]. For example, E. nigrum could cause respiratory fungal infections in wild birds, such as owls [58]. E. cecorum has also been reported to cause severe diseases (e.g., vertebral osteomyelitis) in birds, such as poultry [59]. Both had a relatively high abundance (3.79% and 1.6%) in the gut of Hooded Cranes at Caizi Lake, indicating the cranes there might be suffering more serious pathogen invasion. Interestingly, a previous study showed that there were more pathogenic bacteria in the intestines of Swan Geese in areas with relatively higher human disturbance [13], which was consistent with our study. Rice paddy fields are often used as poultry farms. In the process of sampling, we found that there were poultry breeding sheds near the fields. Poultry have frequent contact with humans and usually roosted in narrow sheds. They could be carriers of pathogens [60], and their feces can pollute paddy fields and water sources during foraging, which could serve as the source of contact infection with enteric pathogens. A recent study demonstrated that there was cross-transmission of intestinal potential pathogens between Hooded Cranes and domestic poultry, and the closer they are, the easier the intestinal flora spread [61], indicating that the more intestinal potentially pathogenic microbes of Hooded Cranes at Caizi Lake may be related to human activities, such as poultry farming. This implied that more conservation measures might be needed to reduce the impact of human activities on the Hooded Cranes at Caizi Lake.
In addition to the difference in intestinal bacterial and fungal community composition of Hooded Cranes at the different lake groups, clustering of intestinal microbiota depending on the habitats (lakes) was detected (Figure 3), which proved that intestinal microbiota showed habitat-preference, implying that the habitats of the host were an important factor in shaping intestinal microbial communities. Furthermore, we found the intestinal fungal community composition of Hooded Cranes at Poyang and Shengjin Lakes was more similar than that of those at Caizi and Shengjin Lakes since the birds in the former lake group have more similar diets. However, the differences in intestinal bacterial community between the two lake groups were almost the same ( Table 2), suggesting that although diet is important, environmental factors play a significant role in shaping the intestinal microbial community. Overall, the divergence degree of intestinal fungal communities of Hooded Cranes at the three lakes was less obvious than that of the bacterial communities ( Figure 3; Table 2). This is consistent with the results of recent studies in which the composition of the intestinal fungal community was shown to be more variable and less stable than that of the bacterial community [62]; however, the underlying mechanism needs to be further studied.

Conclusions
In this study, we demonstrated that intestinal microbial community composition and diversity of the Hooded Cranes from three different lakes of the middle and lower Yangtze River floodplain were significantly different. We found that the Hooded Cranes in the small lake, Shengjin Lake, had higher intestinal microbial diversity and relatively more abundant intestinal potential pathogens than those in the large lake (Poyang Lake), which might be attributed to social behavior of more waterbirds per unit area. Our study showed that the Hooded Cranes at Caizi Lake with relatively poor protection had more diverse and abundant intestinal potential pathogens than those at Shengjin Lake with relatively well protected, indicating that the cranes at the less protected habitats might be suffering more serious pathogen invasion and face greater survival pressure. In addition, intestinal bacterial community composition showed a more marked divergence than intestinal fungi. This study contributes to our understanding of the intestinal microbes of wintering migratory waterbirds in different habitats, and can provide a helpful reference for future studies. However, our study has some limitations. No soil samples were collected as controls, and although the inner core of the feces was used to avoid contamination from the outside, the possibility of environmental contamination cannot be completely removed due to the method by which the samples were collected. In addition, only 51 fecal samples were collected. These limitations need to be further clarified in future research.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/ani11051390/s1. File S1: Supporting Materials and Methods: PCR and amplicon library preparation, Figure S1: Venn diagram showing the unique and shared intestinal bacterial (A,B) and fungal (C,D) ASVs of Hooded Cranes at different lake groups. PY: Poyang lake; CZ: Caizi Lake; SJ: Shengjin Lake; B: bacteria; F: fungus, Table S1: The distribution of data was analyzed by Kolmogorov-Smirnov test. PY: Poyang Lake; SJ: Shengjin Lake; CZ: Caizi Lake. Normal distribution: p > 0.05; Non-normal distribution: p ≤ 0.05, Table S2: The relative abundance of the intestinal bacterial genera shared by the Hooded Cranes at different lake groups. The genus with a relative abundance of more than 0.01% are listed. The generic names that are the same are distinguished by family names in the bracket. PY: Poyang Lake; SJ: Shengjin Lake; CZ: Caizi Lake; B: bacteria, Table S3: The relative abundance of the intestinal fungal genera shared by the Hooded Cranes at different lake groups. The genus with a relative abundance of more than 0.01% are listed. PY: Poyang Lake; SJ: Shengjin Lake; CZ: Caizi Lake; F: fungi.