Characterization of Intestinal Bacterial Communities of Western Lowland Gorillas (Gorilla Gorilla Gorilla), Central Chimpanzees (Pan Troglodytes Troglodytes), and a Forest Elephant (Loxodonta

TROPICS Vol. 23 (4) 175-183 Issued March 1, 2015

ORIGINAL ARTICLE Characterization of intestinal bacterial communities of western lowland gorillas (Gorilla gorilla gorilla), central chimpanzees (Pan troglodytes troglodytes), and a forest elephant (Loxodonta africana cyclotis) living in Moukalaba-Doudou National Park in Gabon

Sayaka Tsuchida1 and Kazunari Ushida1*

1 Graduate School of Life and Environmental Sciences, Kyoto Prefectural University, Shimogamo, Kyoto, 606-8522, Japan * Corresponding author: [email protected]

ABSTRACT Intestinal microbiota play an important role in digestion and host’s health. Furthermore, their composition is complex, with large differences between animal species. Intestinal microbiota have been intensively studied in humans, whereas those in animals, especially in the wild, have not been thoroughly studied. In this study, we focused on the intestinal microbiota of wild western lowland gorillas, chimpanzees, and an elephant in Moukalaba-Doudou National Park in Gabon by using pyrosequencing analysis for understanding their characteristics. Pyrosequencing of their fecal samples yielded 16,898 reads with a read length of 390 nucleotides (10,860,585 nucleotides in total) on average, and taxonomic analysis of metagenomic reads was performed by BLAST search. In almost all animal feces, Prevotellaceae, Clostridiaceae, and Lachnospiraceae were detected as major bacterial families. At the genus level, no-rank Operational Taxonomic Units (OTUs), 80%-90% identities with known sequences, covered a major of fecal microbiota which seemingly determine the enterotype of the host. However, in principal coordinate analysis using weighted UniFrac, their fecal bacteria were clustered by species of host. The result in the present study suggests that it is necessary that no-rank OTUs and minor populations of the fecal bacteria should be analyzed in detail to understand the true characteristics such as functionality of intestinal microbiota.

Key words: Intestinal microbiota, Metagenomes, Pyrosequencing, Wild animals

INTRODUCTION stock, including some pet animals, due to their economic importance (Zhou et al. 2007, Dowd et al. 2008, Turnbaugh Intestinal microbiota in the mammal develop a com- et al. 2008, Hill et al. 2010, Handl et al. 2011). Indeed, plex ecosystem of vast diversity after birth. For example, many studies are concerned with gut microbiota in order to humans and mice, as experimental animals, possess at least characterize its particular relationship to metabolic disor- 1,000 and 400 phylogenetically different bacteria, respec- ders and chronic diseases in view of the loss of health- tively (Hooper 2004, Qin et al. 2010). It has been speculated promoting indigenous bacteria (Andoh et al. 2007, Wen et that such an ecosystem may develop in a host-speciﬁc man- al. 2008, Kellermayer 2013). Such protected microbiota ner (Ley et al. 2008). This was evidenced by the earlier should have a particular relationship with their host, and studies done by Mitsuoka and Kaneuchi (1977) in which such a relationship can be explained by the concept of co- each animal species possesses a particular composition of evolution between the host and its intestinal microbiota microbiota. The interaction between bacteria and the intes- (Amato 2014). Comprehensive analyses on the microbiota tinal mucosa of the host explains the selection of bacteria to of wild animals have been out of focus for such nutritional some extent (Yamamoto et al. 1996, Kelly et al. 2005, and pathological studies. Accordingly, there are only a few Uchida et al. 2006). The feeding behavior of the hosts fur- studies elucidating on the microbiota of wild animals except ther selects the bacteria that can reside in their intestine. In for those of the great apes under captivity (Kisidayová et al. this context, there may be clear differences in the composi- 2009, Vlčková et al. 2012). We believe that the surveys on tion of microbiota between herbivores and carnivores the intestinal microbiota of wild animals have tremendous (Mitsuoka and Kaneuchi 1977, Ley et al. 2008, Endo et al. importance for the understanding of the co-evolution 2010). Intestinal microbiota have been intensively studied between the host and its intestinal microbiota. Our previous in humans, model animals such as rats or mice, and live- study reveals that the fecal microbiota of wild chimpanzees 176 TROPICS Vol. 23 (4) Sayaka Tsuchida and Kazunari Ushida

(Pan troglodytes verus) were clearly different from those ically experiences two seasons: the rainy season from mid- under captivity with some particular influences from hu- October to May and the dry season from June to September. man-associated bacteria (Uenishi et al. 2007, Ushida 2010). Mean annual rainfall (2002-2006) was 1,777 mm (range: Studies on captive animals may have limitations in reveal- 1,583-2,163 mm). The mean monthly minimum and maxi- ing the original composition of intestinal microbiota of the mum temperatures varied from 21.3℃ to 24.1℃ and target animals. 29.3℃ to 33.7℃, respectively (Takenoshita et al. 2008). Our previous study adopted 16S rDNA-based tempera- ture gradient gel electrophoresis (TGGE), which enables us to analyze bacteria of top 20-level abundance (Uenishi et al. Sampling of feces 2007). Recent developments in sequencing technology can characterize the individual differences of human microbiota The feces were collected in the forests of Boutiana and by deep sequencing (Wu et al. 2010). Therefore, we decided Douguetsi in Moukalaba-Doudou National Park in Gabon. to study the intestinal microbiota of gorillas, chimpanzees, In this national park, anthropological and ecological studies and an elephant in the wild by using pyrosequencing analy- have been carried out since 2003, and a group of gorillas is sis for a more precise understanding of their intestinal mi- now habituated (Ando et al. 2008). Chimpanzees are not crobiota. We have selected western lowland gorillas (Gorilla yet well habituated, but they sometimes allow the approach- gorilla gorilla) and central chimpanzees (P. troglodytes ing researchers to collect fresh feces. Elephants are one of troglodytes) in Moukalaba-Doudou National Park in Gabon most dangerous animals in this study area, but their numer- as targets because they are sympatric in depending on nearly ous fresh feces are relatively easily collected. The fresh feces the same food variety, with the exception of temporary in- of western lowland gorillas were collected on November 23 gestion of insects, which is one of preferred food for chim- and December 16, 2011, in Boutiana. From the volume and panzees in general (Tutin and Fernandez 1993, Yamagiwa size of the feces, one fecal sample was judged to be from a and Basabose 2006). As mentioned above, the food habits male silverback gorilla (SBG) of this group, and the other of the hosts select the intestinal bacteria. It is important to the feces of an infant gorilla (IG). The feces of chimpanzees compare the intestinal bacteria between sympatric apes, go- (CH1 and CH2) were collected on December 29, 2011, in rillas, and chimpanzees, which reflect their adaptation to a the forest of Douguetsi adjacent to that of Boutiana. Fresh particular food habit. In this context, we are also interested feces from an elephant (EP) was collected in Boutiana on in intestinal microbiota of forest elephants (Loxodonta December 20, 2011. A portion of feces free from contami- africana cyclotis), which are the major herbivore animals in nation such as soil or dead leaves was sampled in an RNA the study area that forage grass, leaves, and fruits. In gener- later solution. Samples were stored in a dark place in the al terms, the chemical components of their food seem to be campsite of Boutiana until the end of the field research similar to those of the food of gorillas and chimpanzees. (January 8, 2012). Samples were then transported to the This is the first report on the comparison of the intesti- laboratories of the Research Institute of Tropical Ecology nal microbiota of sympatric wild gorillas and wild chim- (IRET) at Libreville, where samples were placed in a refrig- panzees. This is also the first bacteriological study on the erator and later transported to Kyoto Prefectural University; wild forest elephant. there, samples were stored at －20℃ until DNA extraction.

MATERIALS AND METHODS Culture-independent method

Study site Bacteria were recovered from the fecal samples by centrifugation with exhaustive washing with phosphate- The study was conducted in Moukalaba-Doudou Na- buffered saline to remove residual RNA later solution. The tional Park, Gabon (Yamagiwa 2015). The park covers an resultant bacterial pellets were subjected to DNA extraction area of 5,028 km2. The study area covers about 120 km2 in with cell disruption by zirconia beads beating (Microsmash, the southeastern part of the park at an altitude of 50-800 m. TOMY, Tokyo) and a DNA stool mini kit (QIAGEN, To- The research station was located at 2° 20′ and 10° 34′ E. kyo). After quantiﬁcation by spectrophotometry, a portion The vegetation is a complex mosaic of semi-primary forest, (100 ng) of each DNA was subjected to PCR ampliﬁcation secondary forest, Musanga cecropioides-dominated forest, of a partial 16S rRNA gene (V1-V2 region) using ExTaq savanna, and swamp (Iwata and Ando 2007). This area typ- polymerase (Takara, Kyoto). PCR primers, 27F (5’-AGAGT Intestinal microbiota of gorilla, chimpanzee and elephant 177

TTGATCCTGGCTCAG-3’) and 520R (5’-ACCGCGGCTG fashion by means of the UniFrac distance metric (Lozupone CGGC-3’) (Lane 1991), were used to gain nearly 500bp and Knight 2005). All steps were carried out in an automat- PCR amplicon in size. Both primers were attached with ed fashion within QIIME (Caporaso et al. 2010). UniFrac barcode sequences. The primers were supplied from Hok- analysis was carried out in a weighted fashion, which takes kaido System Science Co., Ltd. (Sapporo, Japan). PCR was into account the relative proportions of each individual (Wu performed under the following conditions: 3 min of initial et al. 2010). Clustering was visualized for weighted UniFrac denaturation at 95℃ followed by 30 cycles (95℃ for 30 s, data using principal coordinate analysis (Gower 1966). 55℃ for 40 s, and 72℃ for 90 s) and ﬁnal extension at 72℃ for 4 min. The amplicons were puriﬁed by the PCR Clean- Up System (Promega, USA). Accession numbers

Nucleotide sequence date reported are available in the Pyrosequence analysis DDBJ Sequenced Read Archive under the accession numbers DRA002738 (IG), DRA002742 (SBG), DRA002743 Pyrosequencing and sequence analysis was performed (CH1), DRA002744 (CH2), and DRA002745 (EP). at Hokkaido System Science. Pyrosequencing was performed by the 454 Genome Sequencer FLX (Roche, USA) according to the manufacturer’s instructions. RESULTS The sequences were cleaned by custom script on LINUX to remove sequences comprising a base-call other Pyrosequencing than A, T, C, or G. Then the sequences shorter than 250 bp were removed from whole sequence reads by the same ap- The total nucleotides of 16S rRNA genes detected from plication. The resultant cleaned sequence reads were further the feces, number of sequence reads, and average read length subjected to BLAST search by the stand-alone program are shown in Table 1. On average, pyrosequencing on fecal NCBI BLAST (http://www.ncbi.nlm.nih.gov/books/ samples of IG, SBG, CH1, CH2, and EP yielded 16,898 NBK52640/) to remove sequences showing less than 50 % reads with a read length of 390 nucleotides (10,860,585 nu- alignment similarity with known sequences, because those cleotides in total). Rarefaction curves calculated by Chao1 sequences possibly contained chimeric sequences. richness estimator in QIIME indicated that microbial diversity reached saturation at 14,000 sequence reads. Taxonomic classification analysis on pyrosequencing data: The cleaned sequences from the previous step were assigned their phylotypes using DDBJ database according Phylogenetic profile of the fecal bacteria at phylum to QIIME-1.5.0 454 Tutorials (http://qiime.org/tutorials/ and class levels tutorial.html) with 80 % conﬁdence threshold. In this taxonomic analysis, we defined sequences as identified OTUs Firmicutes and Bacteroidetes were detected as major when their similarities to known sequences were lager than bacterial phyla in the feces of all animal studied. Then, 90 %. The sequences were deﬁned as no-rank OTUs if their Clostridia and Bacteroidia were detected as the major bac- similarities to known sequences were smaller than 90 %. terial classes in the feces of all animal studied. Erysipelotri- chi was detected as the major class of the fecal Firmicutes in chimpanzees (Fig. 1). UniFrac cluster analysis

The cleaned sequences were compared in a pair-wise

Table 1. Pyrosequencing results of 16S rRNA genes in feces of gorillas, chimpanzees and an elephant

IG PG CH 1 CH 2 EP Total 16S rRNA gene (nucleotides) 10,065,107 10,121,642 9,684,893 12,706,879 11,724,402 Number of sequence reads 17,437 16,300 15,736 19,503 15,514 Average read length (nucleotides) 414.22 384.81 386.16 386.03 378.87 178 TROPICS Vol. 23 (4) Sayaka Tsuchida and Kazunari Ushida

Fig. 1. Phylogenetic proﬁle of the fecal bacteria at phylum and class levels of (A) IG, (B) SBG, (C) CH1, (D) CH2, and (E) EP. Others1, 3, 5, 7, and 9: Other minors belong to Firmicutes. Others 10: Other minors belong to Bacteroidetes. Others were the phylum and class whose percentages accounted for <5 %. No rank indicates the sequences unidentiﬁed by BLAST search.

Phylogenetic proﬁle of the fecal bacteria in gorillas at bacteria in IG. At the genus level, the fecal bacteria of IG family and genus levels was dominated by no-rank OTUs which relates to unknown Firmicutes (GU428814_1) followed by the OTUs assigned The major bacterial families (>5 % in total population) as Prevotella. Within no-rank OTUs at genus level, which detected in the feces of IG were Prevotellaceae, Clostridia- covered 40 % of the total sequence reads retrieved from the ceae, Ruminococcaceae, Eubacteriaceae, and Lachnospira- fecal bacteria of IG, the most prevalent no-rank OTU was ceae, each having 6,378 (37 %), 1,764 (10 %), 1,609 (9 %), suggested to relate with the sequence of unknown Firmic- 818 (5 %), and 786 (5 %) sequences, respectively (Table 2). utes (GU428814_1), the second to unknown Firmicutes These five families accounted for 66 % of the total fecal (AB262677_1), and the third to unknown Bacteroidetes Intestinal microbiota of gorilla, chimpanzee and elephant 179

Table 2. Phylogenetic proﬁle of the fecal bacteria at family level of gorillas (SBG, IG), chimpanzees (CH1, CH2) and an elephat (EP)

Relative abundance (%) Bacterial family SBG IG CH1 CH2 EP No rank 18.0 25.3 27.5 25.2 40.3 Prevotellaceae 32.6 36.6 11.8 13.8 8.7 Lachnospiraceae 18.8 4.5 15.4 18.0 8.8 Clostridiaceae 6.6 10.1 8.4 9.7 7.6 Erysipelotrichaceae 1.8 5.9 13.0 3.8 Ruminococcaceae 1.8 9.2 5.5 2.9 3.0 Eubacteriaceae 3.7 4.7 4.7 2.4 4.9 Veillonellaceae 4.9 5.2 4.1 1.1 Oscillospiraceae 3.7 6.0 2.5 2.4 Coriobacteriaceae 3.7 3.6 3.6 1.1 Acidaminococcaceae 2.4 1.6 1.8 Victivallaceae 4.5 Mycoplasmataceae 1.7 Spirochaetaceae 1.7 Sutterellaceae 1.2 Bacteroidaceae 1.1 Oxalobacteraceae 1.1 Porphyromonadaceae 1.1 Peptostreptococcaceae 1.0 Others 4.6 3.1 3.9 4.8 7.5 Others were the families whose percentages accounted for ＜1 %. No rank indicates the sequences unidentiﬁed by BLAST search.

(AB547676_1). OTU was similar to Oribacterium sinus, the second to In identified OTUs at genus level, which covered 60 % Prevotella bryantii, and the third to Prevotella paludivivens. of the total sequence read, the most prevalent OTU was sim- Their similarities with known sequences ranged from 92 % ilar to Prevotella copri, the second to Prevotella oulorum, to 98 %. and the third to Clostridium indolis. Their similarities with known sequences were from 92 % to 96 %. For the fecal bacteria of SBG, Prevotellaceae, Lachno- Phylogenetic profile of the fecal bacteria in chimpan- spiraceae, Clostridiaceae, and Veillonellaceae were the major zees at family and genus levels families, each having 5,306 (33 %), 3,062 (19 %), 1,080 (7 %), and 795 (5 %) sequences, respectively (Table 2). The major families (>5 % in total population) in the These four families accounted for 64 % of the total fecal feces of CH1 were Lachnospiraceae, Prevotellaceae, Clos- bacteria in SBG. tridiaceae, Oscillospiraceae, Erysipelotrichaceae, Rumino- At the genus level, the fecal bacteria of SBG was dom- coccaceae, Veillonellaceae, and Eubacteriaceae, each hav- inated by no-rank OTUs which relates to unknown Bacte- ing 2,420 (15 %), 1,862 (11 %), 1,317 (8 %), 944 (6 %), 935 roidetes (EU728760_1) followed by the OTUs assigned as (6 %), 868 (6 %), 815 (5 %), and 734 (5 %) sequences, re- Prevotella. Within no-rank OTUs at genus level, which spectively (Table 2). These eight families accounted for covered 34 % of the total sequence reads retrieved from the 63 % of the total fecal bacteria in CH1. fecal bacteria of SBG, the most prevalent no-rank OTU was At the genus level, the fecal bacteria of CH1 was dom- suggested to relate with the sequence of unknown Bacteroi- inated by no-rank OTUs which relates to unknown Bacte- detes (EU728760_1), the second to unknown Firmicutes roidetes (AB239491_1) followed by the OTUs assigned as (GU429031_1), and the third to unknown Bacteroidetes Prevotella. Within no-rank OTUs at genus level, which (AJ009933_1). In identified OTUs at genus level, which covered 49 % of the total sequence reads retrieved from the covered 66 % of the total sequence read, the most prevalent fecal bacteria of CH1, the most prevalent no-rank OTU was 180 TROPICS Vol. 23 (4) Sayaka Tsuchida and Kazunari Ushida suggested to relate with the sequence of unknown Bacteroi- detes (AB239491_1), the second to unknown Bacteroidetes (GQ131410_1), and the third to unknown Firmicutes (GU470893_1). In identified OTUs at genus level, which covered 51 % of total sequence read, the most prevalent OTU was similar to Oscillibacter valericigenes, the second to Prevotella oulorum, and the third to Prevotella oris. Their similarities with known sequences ranged from 92 % to 96 %. In the feces of CH2, the major bacterial families were Lachnospiraceae, Prevotellaceae, Erysipelotrichaceae, and Clostridiaceae, each having 3,505 (18 %), 2,695 (14 %), 2,533 (13 %), and 1,886 (10 %) sequences, respectively Fig. 2. Comparison of the relative abundance of the fecal (Table 2). These four families accounted for 55 % of the microbiota in each individual using weighted UniFrac. total fecal bacteria in CH2. At the genus level, the fecal bacteria of CH2 was dominated by no-rank OTUs which relates to unknown covered 48 % of the total sequence read, the most prevalent Firmicutes (GU470893_1) followed by the OTUs assigned OTU was similar to Prevotella ruminicola, the second to as Prevotella. Within no-rank OTUs at genus level, which Prevotella copri, and the third to Prevotella oulorum. Their covered 44 % of the total sequence reads retrieved from the similarities with known sequences ranged from 92 % to fecal bacteria of CH2, the most prevalent no-rank OUT was 94 %. suggested to relate with the sequence of unknown Firmic- utes (GU470893_1), the second to unknown Bacteroidetes (AB239491_1), and the third to unknown Bacteroidetes Common bacterial families and genus in the feces of (GQ422745_1). In identified OTUs at genus level, which gorillas, chimpanzees, and an elephant covered 56 % of the total sequence read, the most prevalent OTU was similar to Oribacterium sinus, the second to At the family level, Prevotellaceae and Clostridiaceae Prevotella oulorum, and the third to Prevotella oris. Their were detected as major bacterial families in the feces of all similarities with known sequences were from 92 % to 97 %. animals. Excluding IG, Lachnospiraceae was detected as a major bacterial family in the feces of adult animals. Un- identified families accounted for >20 % of the population in Phylogenetic profile of the fecal bacteria in an all animals (Table 2). elephant at family and genus levels At the genus level, Prevotella was detected as the most dominant bacterial genus in the feces of all animals. The major families (>5 % in total population) in the feces of EP were Lachnospiraceae, Prevotellaceae, Clostri- diaceae, and Eubacteriaceae, each having 1,357 (9 %), UniFrac cluster analysis 1,347 (9 %), 1,185 (8 %), and 752 (5 %) sequences, respectively (Table 2). These four families accounted for 26 % of Weighted UniFrac analysis, which takes into account the total fecal bacteria in EP. the information of abundance, shows the relative difference At the genus level, the fecal bacteria of EP was domi- in composition of fecal microbiota of gorillas, chimpanzees, nated by no-rank OTUs which relates to unknown Firmic- and an elephant (Fig. 2). utes (EU281854_1) followed by the OTUs assigned as Prevotella. Within no-rank OTUs at genus level, which covered 52 % of the total sequence reads retrieved from the DISCUSSION fecal bacteria of EP, the most prevalent no-rank OTU was suggested to relate with the sequence of unknown Firmic- Intestinal microbiota play an important role in utes (EU281854_1), the second to unknown Bacteridetes digestion, absorption of nutrients, and the host’s health (AB501166_1), and the third to unknown Firmicutes (Sekirov et al. 2010, Kau et al. 2011). This theory is sub- (AB596885_1). In identified OTUs at genus level, which stantiated by many works in humans and their model exper- Intestinal microbiota of gorilla, chimpanzee and elephant 181 imental animals, as well as in livestock. Such a relationship unknown bacteria that have never been isolated or is now understood as a context of co-evolution (Amato characterized. This seems to be in contrast to those analyzed 2014). Studies on the intestinal microbiota of wild animals for human fecal microbiota. In the latter case, no-rank are important for recognizing the co-evolution of the host OTUs constituted less than 20 % of the total OTUs detected and its intestinal microbiota. However, the intestinal micro- when analyzed by the same method (E. Inoue, personal biota in wild animals have not been well studied due to the communication). Human-associated bacteria have been technical difficulties of cultivation in field conditions. studied more intensively than those of wild animals for a In developing culture-independent analyses, some of long time. The database for human-associated bacteria is the studies succeeded in showing the characteristics of the obviously better documented than that for wild animals. To intestinal microbiota of wild animals (Uenishi et al. 2007, some extent, the rumen bacteria of ruminant livestock have Glad et al. 2010). However, the techniques allowed analysis been focused so far due to their economical importance only of the top 20 to top 200 levels of bacteria. The recent (Flint et al. 2008, Duan et al. 2009, Jami and Mizrahi 2012). development of the so-called Omics approach allows for This may limit the application of metagenomic analyses on analysis of some 1,000 bacterial OTUs. This development fecal microbiota in wild animals particularly mono-gastric is helpful for better understanding the characteristics of animals. Because of this limitation, only the phylum-level intestinal microbiota of particular wild animals. comparison has been reported so far, which only allows for In the present study, we believe that the host-specific the rough comparison between host animal species and may characteristics of the fecal microbiota of wild western not allow for characterization in detail. As shown in the lowland gorillas, central chimpanzees, and a forest elephant recent reports based on metagenomic analyses (Ley et al. that are living in a relatively narrow area and foraging quite 2008, Bhatt et al. 2013), phylum-level analyses can reveal similar food resources are shown. In previous studies, by that the mammals possess nearly the same composition of using metagenomic analyses, Bacteroidetes and Firmicutes microbiota. For example, in our study, an elephant has a were shown to be the major intestinal bacterial phyla of similar composition to those of gorillas and chimpanzees at various mammals, including gorillas, chimpanzees, and the phylum level (Fig. 1). However, its composition was elephants in the wild and in captivity (Ley et al. 2008, clearly different from those of gorillas and chimpanzees Moller et al. 2013). Our results also demonstrate that the when a comparison was made with family-level analyses Firmicutes and Bacteroidetes were the common intestinal (Table 2). The difference was, in fact, attributable mostly to bacteria for gorillas, chimpanzees, and forest elephants at no-rank OTUs. We believe that no-rank OTUs should be the phylum level. However, the relative proportion of these analyzed in detail for the better understanding and compari- phyla were not the same for host animal species tested (Fig. son of intestinal microbiota of animals except for humans. 1). Therefore, as shown by UniFrac analysis, the fecal To understand the co-evolution between the host and its in- microbiota in gorillas and in chimpanzees were different, testinal microbiota, further efforts to isolate and identify the and that of an elephant was further distant from those of unknown bacteria corresponding to no-rank OTUs are of gorillas and chimpanzees. importance. Among the identified OTUs at genus levels, excluding SBG, Prevotella oulorum was detected as the dominant ACKNOWLEDGEMENTS This study was conduct- OTU of all animals, and Prevotella oris was the dominant ed in cooperation with the Centre National de la Recherche OTU of chimpanzees. In addition to these, Prevotella copri Scientifique et Technologique (CENAREST, Gabon) and was detected as the dominant OTU of IG and EP. However, Institute des Recherches en Ecologie Tropicale (IRET, only a few identified OTUs were commonly shared by three Gabon). We thank the Ministère des Eau et Fôret and different animal species. Agence National des Parcs Nationaux of the Gabonese The most important finding of this study is the government for permission and support for our research quantitative importance of no-rank OTUs in the fecal project in Gabon. We also thank the members of our re- microbiota of all animals tested. In our results at genus search group for their cooperation. We are indebted to all level, no-rank OTUs covered from 35 % (SBG) to 52 % field assistants of Moukalaba-Doudou National Park and (EP) of the total sequences in each animal subject. the people in Doussala, Konzi, and Mboungou for their Moreover, they constituted the richest OTUs in all animals. kind support and hospitality. We are also greatly indebted to As indicated above, the similarity of sequences of no-rank Dr. Takahiro Segawa, National Institute of Polar Research, OTUs to those registered in the data bank was lower than and Mr. Dylan Bodington, Tokyo Institute of Technology of 90 % in this study. These no-rank OTUs are, accordingly, Bioscience and Biotechnology, for their help in statistical 182 TROPICS Vol. 23 (4) Sayaka Tsuchida and Kazunari Ushida analysis of the data. This study was financially supported 325-338. by the program Science And Technology Research Partner- Handl S, Dowd SE, Garcia-Mazcorro JF, Steiner JM, Suchodolski ship for Sustainable Development (SATREPS) of the Japan JS. 2011. Massive parallel 16S rRNA gene pyrosequencing reveals highly diverse fecal bacterial and fungal communities Science and Technology Agency (JST) and the Japan Inter- in healthy dogs and cats. FEMS Microbiology Ecology 76: national Cooperation Agency (JICA). 301-310. Hill DA, Hoffmann C, Abt MC, Du Y, Kobuley D, Kirn TJ, Bushman FD, Artis D. 2010. Metagenomic analyses reveal REFERENCES antibiotic-induced temporal and spatial changes in intestinal microbiota with associated alterations in immune cell homeo- Amato KR. 2014. Co-evolution in context: The importance of stasis. Mocosal Immunology 3: 148-158. studying gut microbiomes in wild animals. Microbiome Sci- Hooper LV. 2004. Bacterial contributions to mammalian gut devel- ence and Medicine 1: 10-29. opment. Trends in Microbiology 12: 129-134. Ando C, Iwata Y, Yamagiwa J. 2008. Progress of habituation of Iwata Y, Ando C. 2007. Bed and bed-site reuse by western lowland western lowland gorillas and their reaction to observers in gorillas (Gorilla g. gorilla) in Moukalaba-Doudou National Moukalaba-Doudou National Park, Gabon. African Study Park, Gabon. Primates 48: 77-80. Monographs Supplementary Issue 39: 55-69. Jami E, Mizrahi I. 2012. Composition and similarity of bovine Andoh A, Sakata S, Koizumi Y, Mitsuyama K, Fujiyama Y, Benno rumen microbiota across individual animals. PloS One 7 (3): Y. 2007. Terminal restriction fragment length polymorphism DOI: 10.1371/journal.pone.0033306. analysis of the diversity of fecal microbiota in patients with Kau AL, Ahern PP, Griffin NW, Goodman AL, Gordon JI. 2011. ulcerative colitis. Inflammatory Bowel Diseases 13: 955-962. Human nutrition, the gut microbiome and the immune system. Bhatt VD, Dande SS, Patil NV, Joshi CG. 2013. Molecular analysis Nature 474: 327-336. of the bacterial microbiome in the forestomach fluid from the Kellermayer R. 2013. Prospects and challenges for intestinal mi- dromedary camel (Camelus dromedarius). Molecular Biology crobiome therapy in pediatric gastrointestinal disorders. World Reports 40: 3363-3371. Journal of Gastrointestinal Pathophysiology 4: 91-93. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman Kelly D, Conway S, Aminov R. 2005. Commensal gut bacteria: FD, Costello EK, Fierer N, Peña AG, Goodrich JK, Gordon Mechanisms of immune modulation. Trends in Immunology JI, Huttley GA, Kelley ST, Knights D, Koenig JE, Ley RE, 26: 326-333. Lozupone CA, McDonald D, Muegge BD, Pirrung M, Reeder Kisidayová S, Váradyová Z, Pristas P, Piknová M, Nigutová K, J, Sevinsky JR, Turnbaugh PJ, Walters WA, Widmann J, Petrzelková KJ, Profousová I, Schovancová K, Kamler J, Yatsunenko T, Zaneveld J, Knight R. 2010. QIIME allows Modrý D. 2009. Effects of high- and low-fiber diets on fecal analysis of high-throughput community sequencing data. fermentation and fecal microbial populations of captive chim- Nature Methods 7: 335-336. panzees. American Journal of Primatology 71: 1-10. Dowd SE, Callaway TR, Wolcott RD, Sun Y, McKeehan T, Lane DJ. 1991. 16S/ 23S rRNA sequencing. In: Stackebrandt E, Hagevoort RG, Edrington TS. 2008. Evaluation of the bacte- Goodfellow M (eds) Nucleic acid techniques in bacterial sys- rial diversity in the feces of cattle using 16S rDNA bacterial tematics. John Wiley & Sons, Chichester. 115-175. tag-encoded FLX amplicon pyrosequencing (bTEFAP). BMC Ley RE, Hamady M, Lozupone C, Turnbaugh PJ, Ramey RR, Microbiology 8: DOI: 10.1186/1471-2180-8-125. Bircher JS, Schlegel ML, Tucker TA, Schrenzel MD, Knight Duan CJ, Xian L, Zhao GC, Feng Y, Pang H, Bai XL, Tang JL, Ma R, Gordon JI. 2008. Evolution of mammals and their gut QS, Feng JX. 2009. Isolation and partial characterization of microbes. Science 320: 1647-1651. novel genes encoding acidic cellulases from metagenomes of Lozupone C, Knight R. 2005. UniFrac: A new phylogenetic meth- buffalo rumens. Journal of Applied Microbiology 107: 245- od for comparing microbial communities. Applied and Envi- 256. ronmental Microbiology 71: 8228-8235. Endo A, Futagawa-Endo Y, Dicks LM. 2010. Diversity of Lactoba- Mitsuoka T, Kaneuchi C. 1977. Ecology of bifidobacteria. The cillus and Bifidobacterium in feces of herbivores, omnivores American Journal of Clinical Nutrition 30: 1799-1810. and carnivores. Anaerobe 16: 590-596. Moeller AH, Peeters M, Ndjango JB, Li Y, Hahn BH, Ochman H. Flint HJ, Bayer EA, Rincon MT, Lamed R, White BA. 2008. 2013. Sympatric chimpanzees and gorillas harbor convergent Polysaccharide utilization by gut bacteria: potential for new gut microbial communities. Genome Research 23: 1715- insights from genomic analysis. Nature Reviews Microbiology 1720. 6: 121-131. Qin J, Li R, Raes J, Arumugam M, Burgdorf KS, Manichanh C, Glad T, Bernhardsen P, Nielsen KM, Brusetti L, Andersen M, Aars Nielsen T, Pons N, Levenez F, Yamada T, Mende DR, Li J, J, Sundset MA. 2010. Bacterial diversity in faeces from polar Xu J, Li S, Li D, Cao J, Wang B, Liang H, Zheng H, Xie Y, bear (Ursus maritimus) in Arctic Svalbard. BMC Microbiolo- Tap J, Lepage P, Bertalan M, Batto JM, Hansen T, Le Paslier gy 10:10. DOI: 10.1186/1471-2180-10-10. D, Linneberg A, Nielsen HB, Pelletier E, Renault P, Sicheritz- Gower JC. 1966. Some distance properties of latent root and Ponten T, Turner K, Zhu H, Yu C, Li S, Jian M, Zhou Y, Li vector method used in multivariate analysis. Biometrika 53: Y, Zhang X, Li S, Qin N, Yang H, Wang J, Brunak S, Doré J, Intestinal microbiota of gorilla, chimpanzee and elephant 183

Guarner F, Kristiansen K, Pedersen O, Parkhill J, Weissenbach gorilla gorilla). Journal of Microbiological Methods 91: 45- J, MetaHIT Consortium, Bork P, Ehrlich SD, Wang J. 2010. A 51. human gut microbial gene catalogue established by metage- Wen L, Ley RE, Volchkov PY, Stranges PB, Avanesyan L, nomic sequencing. Nature 464: 59-65. Stonebraker AC, Hu C, Wong FS, Szot GL, Bluestone JA, Sekirov I, Russell SL, Antunes LC, Finlay BB. 2010. Gut microbi- Gordon JI, Chervonsky AV. 2008. Innate immunity and intes- ota in health and disease. Physiological Reviews 90: 859-904. tinal microbiota in the development of type 1 diabetes. Nature Takenoshita Y, Ando C, Yamagiwa J. 2008. Fruit phenology of the 455: 1109-1113. great ape habitat in the Moukalaba-Doudou National Park, Wu GD, Lewis JD, Hoffmann C, Chen YY, Knight R, Bittinger Gabon. African Study Monographs Supplementary Issue 39: K, Hwang J, Chen J, Berkowsky R, Nessel L, Li H, Bushman 23-39. FD. 2010. Sampling and pyrosequencing methods for char- Turnbaugh PJ, Bäckhed F, Fulton L, Gordon JI. 2008. Diet- acterizing bacterial communities in the human gut using induced obesity is linked to marked but reversible alterations 16S sequence tags. BMC Microbiology 10: 206. DOI: 10.1186/ in the mouse distal gut microbiome. Cell Host & Microbe 3: 1471-2180-10-206. 213-223. Yamagiwa J, Basabose AK. 2006. Diet and seasonal changes in Tutin CEG, Fernandez M. 1993. Composition of the diet of chim- sympatric gorillas and chimpanzees at Kahuzi-Biega National panzees and comparisons with that of sympatric lowland Park. Primates 47: 74-90. gorillas in the Lopé Reserve, Gabon. American Journal of Yamagiwa J. 2015. Preface for ＂Conservation of Biodiversity in Primatology 30: 195-211. Tropical Forest through Sustainable Coexistence between Hu- Uchida H, Kinoshita H, Kawai Y, Kitazawa H, Miura K, Shiiba man and Wild Animals＂, a special issue of Tropics. TROPICS K, Horii A, Kimura K, Taketomo N, Oda M, Yajima T, Saito 23 (4) : 135. T. 2006. Lactobacilli binding human A-antigen expressed in Yamamoto K, Miwa T, Taniguchi H, Nagano T, Shimamura K, intestinal mucosa. Research in Microbiology 157: 659-665. Tanaka T, Kumagai H. 1996. Binding speciﬁcity of Lactoba- Uenishi G, Fujita S, Ohashi G, Kato A, Yamauchi S, Matsuzawa T, cillus to glycolipids. Biochemical and Biophysical Research Ushida K. 2007. Molecular analyses of the intestinal micro- Communications 228: 148-152. biota of chimpanzees in the wild and in captivity. American Zhou H, Gong J, Brisbin JT, Yu H, Sanei B, Sabour P, Sharif S. Journal of Primatology 69: 367-376. 2007. Appropriate chicken sample size for identifying the com- Ushida K, Uwatoko Y, Adachi Y, Soumah AG, Matsuzawa T. 2010. position of broiler intestinal microbiota affected by dietary Isolation of Biﬁdobacteria from feces of chimpanzees in the antibiotics, using the polymerase chain reaction-denaturing wild. The Journal of General and Applied Microbiology 56: gradient gel electrophoresis technique. Poultry Science 86: 57-60. 2541-2549. Vlčková K, Mrázek J, Kopečný J, Petrželková KJ. 2012. Evalua- Received: December 2, 2013 tion of different storage methods to characterize the fecal bac- Accepted: June 16, 2014 terial communities of captive western lowland gorillas (Gorilla