Diversity and Composition of Rumen Bacteria, Fungi, and Protozoa in Goats and Sheep Living in the Same High-Altitude Pasture

Simple Summary Tibetan goats and sheep graze together but have different growth performances, immune responses, and feeding preferences in the Tibetan pasture. Rumen microbiota composed of bacteria, fungi, and protozoa are necessary for a healthy ruminant. Therefore, in this study, we comprehensively describe composition and diversity of bacteria, fungi, and protozoa in the high- altitude rumen. Compared with sheep, the bacteria that degrade crude protein and produce volatile fatty acids (VFA) were increased in the rumen of goats (Saccharofermentans and Lachnospiraceae_XPB1014) (p < 0.05). In addition, when compared with goats, the fungi and protozoa that degrade fiber were increased in rumen of sheep (Neocallimastigaceae and Metadinium) (p < 0.05). Furthermore, VFA were significantly increased in the rumen of goats compared with sheep (p < 0.05). The VFA level was consistent with differences in the microbiota composition in the rumen between goats and sheep. Under mixed grazing conditions, goats tend to select a high-crude protein diet that is good for growth, whereas sheep tend to select a high-lignin diet that is difficult to digest. Therefore, the different microbiota in the rumen of goats and sheep may be explained by dietary preference. Abstract Environmental adaptation of ruminants was highly related to microbiota in the rumen. To investigate the diversity and composition of bacteria, fungi, and protozoa in the rumen of high-altitude animals, amplicon gene sequencing was performed using rumen fluid samples derived from both Tibetan goats and sheep at the same pasture in a highland (altitude > 4800 m). Between these two species, the ruminal bacteria and fungi were significantly different at multiple taxonomic levels. The alpha diversity of bacteria was significantly high in goats (p < 0.05). One hundred and sixty-four and 29 Operational Taxonomy Units (OTUs) with significant differences were detected in bacteria and fungi, respectively. The abundance of bacteria, fungi, and protozoa in the rumen was characterized at multiple taxonomic levels, and we determined that Firmicutes, Bacteroidetes, Neocallimastigomycota, and Ciliophora were the most abundant bacteria, fungi, and protozoa. The family Neocallimastigaceae and the genus Metadinium had cellulose degradation capacity in the rumen with high abundance, thereby, suggesting that fungi and protozoa played an essential role in rumen fermentation. In addition, by comparing microbiota in the rumen of goats and sheep it was found, that the fiber-degrading fungi genus (Cyllamyces) was increased in the rumen of sheep (p < 0.05) whereas VFA-producing bacteria (Saccharofermentans and Lachnospiraceae_XPB1014) were increased in the rumen of goats (p < 0.05). Interestingly, in the rumen, no differences in protozoa were observed between goats and sheep (p > 0.05). Furthermore, when compared to sheep, level of acetic acid, propionic acid, and total volatile fatty acid (TVFA) were significantly increased in the rumen of goats (p < 0.05). Taken together, these results suggested microbiota in the rumen drive goats to better adapt to high-altitude grazing conditions.


Animal Handling, Diet, and Sampling
Grazing goats and sheep were raised in the Qiangtang Plateau (Nagchu, Tibet, China; altitude more than 4800 m). A total of forty goats and sheep (n = 20 each) were subjected to simultaneous estrus and artificial insemination. After the goat kids and lambs were born, they followed the dam for grazing together, and no artificial feeding was used. One-year-old female kids and lambs (n = 5 each) were randomly selected to collect rumen fluid samples using oral stomach probing. To reduce invasion of the body, hot water-heated Poly Vinyl Chloride pipes were inserted to a depth of approximately 120-150 cm through the esophagus at 3 h after feeding, then the rumen contents were immediately collected using a vacuum pump, to avoid possible contamination of the samples with saliva. The samples (approximately 50 mL) were strained by a nylon membrane, then, samples were immediately frozen in liquid nitrogen for further study [15].
All experimental procedures were approved by the Institutional Animal Care and Use Committee of the Northwest A&F University under permit number 2014ZX08008002.

Amplicon Gene Sequence Processing
Singletons and low-quality reads were removed from raw reads using Cutadapt (v1.9.1) (http: //code.google.com/p/cutadapt/) [19]. According to 97% similarity, clean sequences were clustered into operational taxonomic units (OTUs) using the UPARSE method [20]. Using the Mothur method [21], annotation analysis of OTUs was performed to obtain representative species and count the numbers of OTUs in the taxonomic levels (kingdom, phylum, class, order, family, genus, and species) per sample. Then, the OTUs of bacteria and protozoa were identified and assigned to the bacteria and protozoa SILVA database (version 11.9) (http://www.arb-silva.de) [22]; the OTUs of fungi were assigned in the database of UNITE Release 7.0 database (http://unite.ut.ee) [23].

Volatile Fatty Acids
After thawing at 4 • C for 2 h, rumen fluid samples were centrifuged for 10 min at 4 • C with 12,000 rpm. Then, 1.5 mL of supernatant was transferred to 2 mL centrifuge tubes containing 200 µL 25% metaphosphoric acid. After incubating overnight at 4 • C, samples were centrifuged for 15 min at 4 • C at 13,000 rpm. A total of 1 mL supernatant was absorbed into the 1.5 mL centrifuge tube containing 200 µL 0.2458% crotonic acid. After incubation at 4 • C for 4 h, 800 µL of the mixture was transferred into vials for further analysis by gas chromatography. The VFA concentrations were measured using an Agilent Technologies 7890A GC system (Agilent Technologies, Santa Clara, CA, USA) as follows: the injector and FID detector temperatures were set at 200 • C, the column temperature was maintained at 100 • C for 1 min, then the temperature increased to 145 • C at 3 • C/min, followed by a temperature increase from 145 to 200 • C at 20 • C/min, the temperature was maintained at 200 • C for 6 min [24].

Data Analysis
The differences in Alpha diversity and VFA were calculated by One-way ANOVA using SPSS v18. The Unifrac distance was computed by QIIME v1.9.1 (http://qiime.sourceforge.net/) [25]. Bar and box graphs were plotted by Graph Prism v6 (https://www.graphpad.com/). The pie chart was plotted by Microsoft Office Excel 2010 and was based on the relative abundance of microbiota. The network, principle coordinate analysis (PCoA) and LDA effect size (LEfSe) graphs were plotted using the Major bio I-Sanger Cloud Platform (https://www.i-sanger.com/), in which the Linear discriminant analysis (LDA) threshold was higher than 2.5.n

An Overview of Ruminal Bacteria, Fungi, and Protozoa Composition
The ruminal bacteria contained 2283 OTUs by 16S rRNA gene sequencing. In addition, 24 bacteria phyla were detected, and the prevalence and abundance of each phylum were different among individuals ( Figure 1D, Supplementary Material Table S1). Moreover, 15 phyla appeared in all individuals, in which Firmicutes (50.6% in goats, 50.3% in sheep) and Bacteroidetes (41.2% in goats, 40.6% in sheep) were the most abundant phyla. The relative abundance of Actinobacteria ranged from 0.48% (goat_5) to 6.05% (sheep_5). Epsilonbacteraeota was less prevalent and not detected in 11 samples (Table S1). Of the OTUs, about 90.5% in goats, 88.9% in sheep and 99.8% in goats, 100% in sheep obtained from the core microbiota (Firmicutes and Bacteroidetes), were enriched in Clostridiales and Bacteroidales orders (Supplementary Material Table S2). The major orders for other phyla with the relative abundance ≥1% were Coriobacteriales (Phylum: Actinobacteria, 90.0% in goats, 89.7% in sheep), Spirochaetales (Phylum: Spirochaetes, 100% in goats, 100% in sheep) (Table S2). At the genus level (the relative abundance ≥1%), the abundance of both Rikenellaceae_RC9_gut and Prevotella_1 belonged to Bacteroidetes (24.5% in goats, 22.6% in sheep) ( Figure 1D, Supplementary Material Table  S3). By ITS rRNA gene sequencing, 1398 OTUs were obtained, including, Neocallimastigomycota (51.6% in goats, 77.6% in sheep) and Ascomycota (36.0% in goats, 18.0% in sheep) were the core fungi in 13 phyla ( Figure 2D, Table S1). Blastocladiomycota, Cercozoa and Kickxellomycota were detected in one sample. The relative abundance of Neocallimastigomycota was highly different between samples (sheep_5: 89.65% to goats_3: 32.13%) (Table S1). Neocallimastigaceae (Phylum: Neocallimastigomycota, 100% in goats and sheep), Pleosporales (Phylum: Ascomycota, 60.8% in goats, 55.6% in sheep) and Filobasidiales (Phylum: Basidiomycota, 38.0% in goats, 45.5% in sheep) were the dominant orders (Table S2). At the genus level, the abundance of unclassified_f__Neocallimastigaceae, Orpinomyces, Caecomyces and Piromyces, which all belonged to Neocallimastigomycota, was the highest ( Figure 2D, Table S1). For protozoa, 270 OTUs were obtained from the rumen of all goats and sheep. Ciliophora, the core microbiota (85.9% in goats, 97.8% in sheep), was unique in 6 phyla, which presented in all individuals ( Figure 3D, Table S1). The OTUs of Ciliophora were enriched in the order Trichostomatia (Table S2). Ciliophora contained only 14 OTUs. Genus with the relative abundance ≥1% belonged to Ciliophora, in which Metadinium had the highest abundance (58.6% in goats, 75.0% in sheep) ( Figure 3D, Table S3). shown. The red color represented that the abundance of taxa was higher in goats and a blue color represented that the abundance of taxa was higher in goats. The higher the Linear Discriminant Analysis (LDA) score, the greater the effect of the taxa abundance on the difference between goats and sheep. , family (f), genus (g) levels are shown. The red color represented that the abundance of taxa was higher in goats and a blue color represented that the abundance of taxa was higher in goats. The higher the Linear Discriminant Analysis (LDA) score, the greater the effect of the taxa abundance on the difference between goats and sheep. shown. The red color represented that the abundance of taxa was higher in goats and a blue color represented that the abundance of taxa was higher in goats. The higher the Linear Discriminant Analysis (LDA) score, the greater the effect of the taxa abundance on the difference between goats and sheep.

Different Composition of Ruminal Bacteria, Fungi, and Protozoa in Goats and Sheep
Goats and sheep shared 1762 OTUs for bacteria. However, 398 and 123 OTUs were not shared in the rumen of goats and sheep, respectively ( Figure 1A). For fungi, goats possessed 706 unique OTUs, while sheep possessed 236 unique OTUs, and shared 456 OTUs (Figure 2A). For protozoa, goats contained 224 unique OTUs, while sheep contained 15 unique OTUs, and shared 31 OTUs ( Figure 3A). The number of OTUs for bacteria, fungi and protozoa was higher in the rumen of goats. Bacterial alpha diversity was higher in goats than sheep based on the Shannon index (p = 0.01) ( Figure  1B). However, the alpha diversity of fungi and protozoa was not significantly different between goats and sheep ( Figures 2B and 3B). Subsequently, the Bray-Curtis distance was used for whole microbiota abundance. The PCoA plot showed a distinct clustering of bacteria (ANOSIM: R = 0.46; p = 0.01) and fungi (ANOSIM: R = 0.37; p = 0.04) composition between goats and sheep ( Figures 1C and 2C). However, such a pattern was not observed in the PCoA of protozoa (ANOSIM: R = 0.03; p = 0.32) ( Figure 3C).
Based on LEfSe analysis, comparing the abundance of bacteria between goats and sheep, the phylum Actinobacteria was higher in sheep, and the family F082 was higher in goats ( Figure 1E). At the genus level (the relative abundance ≥ 1%) (Table S3), the relative abundance of norank_f_F082, Saccharofermentans, Lachnospiraceae_XPB1014, and unclassified_f_Rikenellaceae were higher in the

Different Composition of Ruminal Bacteria, Fungi, and Protozoa in Goats and Sheep
Goats and sheep shared 1762 OTUs for bacteria. However, 398 and 123 OTUs were not shared in the rumen of goats and sheep, respectively ( Figure 1A). For fungi, goats possessed 706 unique OTUs, while sheep possessed 236 unique OTUs, and shared 456 OTUs (Figure 2A). For protozoa, goats contained 224 unique OTUs, while sheep contained 15 unique OTUs, and shared 31 OTUs ( Figure 3A). The number of OTUs for bacteria, fungi and protozoa was higher in the rumen of goats. Bacterial alpha diversity was higher in goats than sheep based on the Shannon index (p = 0.01) ( Figure 1B). However, the alpha diversity of fungi and protozoa was not significantly different between goats and sheep ( Figures 2B and 3B). Subsequently, the Bray-Curtis distance was used for whole microbiota abundance. The PCoA plot showed a distinct clustering of bacteria (ANOSIM: R = 0.46; p = 0.01) and fungi (ANOSIM: R = 0.37; p = 0.04) composition between goats and sheep ( Figures 1C and 2C). However, such a pattern was not observed in the PCoA of protozoa (ANOSIM: R = 0.03; p = 0.32) ( Figure 3C).
Based on LEfSe analysis, comparing the abundance of bacteria between goats and sheep, the phylum Actinobacteria was higher in sheep, and the family F082 was higher in goats ( Figure 1E). At the genus level (the relative abundance ≥ 1%) (Table S3), the relative abundance of norank_f_F082, Saccharofermentans, Lachnospiraceae_XPB1014, and unclassified_f_Rikenellaceae were higher in the goat group, whereas the relative abundance of Olsenella was increased in sheep ( Figure 1E). At the OTU level, 164 OTUs of ruminal bacteria were significantly different between the two groups, in which the abundance of 132 OTUs was increased in the goat group, and 32 OTUs were increased in sheep group. The core phylum encompassed 129 OTUs (Bacteroidetes: 51 OTUs; Firmicutes: 78 OTUs) (Supplementary Material Table S4). For fungi, the relative abundance of the phylum Neocallimastigomycota was higher in the sheep group ( Figure 2E). Taxa with a high abundance was not significantly different at the family level ( Figure 2E, Table S2). Moreover, at the genus level with high abundance, the relative abundance of Cyllamyces was higher in the sheep group ( Figure 2E, Table S3). A total of 29 OTUs was significantly different, in which only 3 OTUs were increased in the sheep group. In all OTUs with significant differences, OTU145, OTU99, and OTU1108 represented the highest abundance, and all belonged to the phylum Neocallimastigomycota (Table S4). In the protozoa, OTU22 was only observed in goats and belonged to the phylum Streptophyta (Table S4). No significant differences were observed at other taxonomy levels.
In this study, the results of amplicon sequencing showed that most of the remarkably different bacteria were related to the production of VFA. Therefore, we measured the concentrations of VFA in the rumen to compare the rumen fermentation capacity between goats and sheep. Notably, when compared to sheep, the concentration of acetic acid, propionic acid, and TVFA was significantly increased in the goat rumen (p < 0.01). Other VFAs did not show a significant difference ( Figure 5).

Discussion
In this study, rumen microbes were characterized in goats and sheep that lived in the same pasture. Both diversity and composition of rumen microbiota were significantly different in goats

Discussion
In this study, rumen microbes were characterized in goats and sheep that lived in the same pasture. Both diversity and composition of rumen microbiota were significantly different in goats

Discussion
In this study, rumen microbes were characterized in goats and sheep that lived in the same pasture. Both diversity and composition of rumen microbiota were significantly different in goats and sheep, and the differences in rumen microbe diversity and composition may be driven by the food preferences of goats and sheep.

Feature of Rumen Microbiota in the High-Altitude Pasture
In all goats and sheep (n = 5 each), a diversity of 24 bacteria phyla (2283 OTUs), 13 fungi phyla (1398 OTUs), and 6 protozoa phyla (270 OTUs) was detected. In addition, microbial diversity was highly observed in high-altitude ruminants [26,27]. The rarefaction curves of bacteria, fungi, and protozoa (Supplementary Material Figure S1) suggested that the number of OTUs was sufficiently captured and that the sequencing depth of the samples was sufficient. When compared to other studies, this study firstly expounded the prevalence and abundance of bacteria, fungi, and protozoa in the rumen of high-altitude ruminants at various taxonomic levels. One of the main functions of the rumen was cellulose degradation. Especially, Fibrobacter (a fiber-degrading bacteria) was not observed in this study. The Neocallimastigomycota of fungi (producing high fiber degrading enzyme) [28], and the genus Metadinium of protozoa [29] (Figures 2D and 3D, Tables S1 and S3) appeared in the rumen with high abundance. Fungi and protozoa might play an important role in the capability of degrading cellulose in the rumen.
The Tibetan Plateau is known for its harsh environment including cold and hypoxic conditions, as well as a low biomass. In order to adapt to the harsh environment, ruminants that live at a high-altitude pasture had evolved low-methane and high-VFA phenotypes [10]. Methane emission is an important energy-consuming pathway, resulting in energy loss of the diet [30]. The increase in VFA could greatly inhibit the production of methane by competing for hydrogen in the methane-producing pathway [31]. In this study, Treponema and Neocallimastigomycota ( Figures 1D and 2D, Tables S1 and S3), having a H2-production capacity, were present in the rumen at high abundance [32]. Markedly, Prevotella_1, is involved in the metabolization of proteins, peptides, starch, hemicellulose, and pectin, and VFA production (Table S3 and Figure 1D) [33]. Treponema produces acetate from hydrogen as a substrate that appeared in the rumen of goats and sheep with high abundance [34]. Therefore, in the high-altitude rumen, more hydrogen might be involved in the synthesis of VFA than methane emission. Previous studies have indicated that rumen microbiota have played a key role in the adaption of the ruminant to high-altitude pasture.

Rumen Microbiota May Reflect Host's Adaptive Capacity of Environment
Alpha diversity was analyzed using One-way ANOVA. The Shannon index was high in goats. PCoA analysis showed that 10 individuals clustered in accordance with the host. A few studies on livestock had shown similar host-related stratification in microbiota. The microbiota in dairy cattle and yellow cattle was significantly different from that in two yak herds that lived at different altitudes [35]. In a study on Yak and Tibetan sheep, it was shown that the prokaryotic community structure between yak and Tibetan sheep was significantly different (p < 0.01) [36]. In another study on yak rumen, a high bacterial diversity was observed when compared to Tibetan sheep [37]. These studies considered the host as the major influencing factor but did not consider the differences in feeding habits. In this study, Cyllamyces (the family Neocallimastigaceae) was related to fiber degradation [38], which was increased in the rumen of sheep ( Figure 2E). Saccharofermentans fermented several carbohydrates, and mainly produced propionic acid, which was increased in the rumen of goats ( Figure 1E) [39]. Moreover, the concentration of propionic acid was higher in goat rumen ( Figure 5). The high abundance of Saccharofermentans indicated the high level of crude protein in the diet [40]. Also, a high crude protein diet significantly increased the growth performance of ruminants under grazing conditions [41]. Lachnospiraceae_XPB1014 produced an array of bacteriocins and butyrate [42], which was increased in the rumen of goats ( Figure 1E). In this study, no significant differences were observed in the concentration of butyrate acid between goats and sheep ( Figure 5). The reason might be that Lachnospiraceae_XPB1014 was not just involved in butyrate production. The increase of fiber-degrading microbiota caused more hydrogen production in the rumen of sheep compared to goats. However, the decrease in VFA-producing microbiota indicated higher methane emissions and more energy consumption in the rumen of sheep. In a previous study, it was suggested that methanogens attach to protozoa and fungi (which degrade fibers to produce hydrogen) to obtain hydrogen [43].
VFA played an important role in ruminant growth and immunity. VFA could make intestinal epithelial cells highly keratinized, which provided a physical barrier for the rumen environment [44]. Acetic and propionate acid could activate the GPR43 and GPR41 receptor to product PYY and GLP-1 hormones which increased glucose utilization for body energy [45]. Notably, acetic acid, the major VFA in the rumen, could decrease the abundance of Escherichia coli to maintain rumen health [46]. In this study, TVFA, acetic acid, and propionate acid were significantly increased in goat rumen, suggesting that goats were more suitable to the harsh conditions of Tibetan pastures compared to sheep.

The Role of the Diet in Rumen Microbiota of Goats and Sheep
In this study, the diversity and composition of rumen microbiota were different between goats and sheep that lived in the same pasture. A study on the gut microbiome of pigs during nursing and weaning suggested that the diet played a key role in shaping the gut microbiome, and that a microbiome with different functions reflected the dietary composition [47]. Similarly, in another study, it was found that the gut microbiome was highly dynamic, exhibiting daily cyclical fluctuations in composition influenced by diet, and it was suggested that diet was the dominant influencing factor for gut microbiota [48]. In our study, we could not eliminate the impact of the host gene, although goats and sheep had the same father by artificial insemination and the same age and sex by estrus synchronization. In a study on primates, it was a discovered that Varecia variegata and Lemur catta had a diverse gut morphology. After being fed a diet of a rotating fruit and vegetable mix, V. variegata and L. catta had similar microbial structures [49]. Several studies on the effect of altitude on rumen microbiota showed that Christensenellaceae R7, Succiniclasticum and Ruminococcaceae highly corrected with altitude [35,50]. These bacteria could secrete various enzymes and be associated with food efficiency [51][52][53]. Notably, the numbers of these bacteria in the rumen depended on the diet [54,55]. The bacteria related to altitude appeared in this study with high abundance ( Figure 1D). Interestingly, Alloprevotella, that appeared in the rumen in this study (Table S3), decreased the risk of cardiovascular disease [56], and the absence of oxygen could lead to cardiovascular disease [57]. Therefore, the role of diets might be more important for the colonization of rumen microbiota. In the case that goats and sheep grazed with a high-intensity in mixed herds and goats reduced dry matter intake and increased feed intake of crude protein. However, the intake of food containing mature fiber forage increased in sheep [58]. We inferred that the difference in rumen microbial diversity and composition between goats and sheep might be driven by feeding preferences.

Conclusions
In our study, we characterized ruminal bacteria, fungi, and protozoa at multiple taxonomic levels in the highland pasture, and found differences in rumen microbial diversity and composition between goats and sheep that lived in the same pasture. The family Neocallimastigaceae and the genus Metadinium were enriched in sheep, whereas Saccharofermentans and Lachnospiraceae_XPB1014 were enriched in goats. These differences caused the production of different metabolites between goats and sheep. Furthermore, we found that compared to sheep, the rumen of goats had a higher VFA level. The above-mentioned results suggested that rumen microbiota facilitated goats to better adapt to high-altitude pastures. Therefore, in future studies, we aim to investigate the differences in feeding preference for types of plants between goats and sheep by measuring the concentration of alkanes. Ultimately, we hope to improve ruminant tolerance for the plateau area by diet.
Supplementary Materials: The following are available online at http://www.mdpi.com/2076-2615/10/2/186/s1, Table S1: Relative abundance of the phylum level in goats and sheep. Table S2: Relative abundance of the order level in goats and sheep. Table S3: Relative abundance of the genus level in goats and sheep. Table S4: Operational taxonomy units (OTUs) showing significant differences in goats and sheep. Figure S1: Rarefaction curves.

Funding:
The present study was supported by the Tibet Autonomous Region Natural Science Fund Project (XZ2018ZR G-51Z), Local Grants (2017NY-072 and 2018KJXX-009), Tibet Science and Technology Department "13th Five-Year Plan" Major Agriculture Special Project (XZ201901NA02), and the Special Fund for Modern Agricultural Industry Technology System Construction (CARS- . None of the funding agencies had any role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.