Rock ptarmigans (Lagopus muta) are gallinaceous birds inhabiting arctic and sub-arctic environments. Their diet varies by season, including plants or plant parts of high nutritional value, but also toxic plant secondary metabolites (PSMs). Little is known about the microbes driving organic matter decomposition in the cecum of ptarmigans, especially the last steps leading to methanogenesis. The cecum microbiome in wild rock ptarmigans from Arctic Norway was characterized to unveil their functional potential for PSM detoxification, methanogenesis and polysaccharides degradation. Cecal samples were collected from wild ptarmigans from Svalbard (L. m. hyperborea) and northern Norway (L. m. muta) during autumn/winter (Sept-Dec). Samples from captive Svalbard ptarmigans fed commercial pelleted feed were included to investigate the effect of diet on microbial composition and function. Abundances of methanogens and bacteria were determined by qRT-PCR, while microbial community composition and functional potential were studied using 16S rRNA gene sequencing and shotgun metagenomics. Abundances of bacteria and methanogenic Archaea were higher in wild ptarmigans compared to captive birds. The ceca of wild ptarmigans housed bacterial groups involved in PSM-degradation, and genes mediating the conversion of phenol compounds to pyruvate. Methanomassiliicoccaceae was the major archaeal family in wild ptarmigans, carrying the genes for methanogenesis from methanol. It might be related to increased methanol production from pectin degradation in wild birds due to a diet consisting of primarily fresh pectin-rich plants. Both wild and captive ptarmigans possessed a broad suite of genes for the depolymerization of hemicellulose and non-cellulosic polysaccharides (e.g. starch). In conclusion, there were no physiological and phenotypical dissimilarities in the microbiota found in the cecum of wild ptarmigans on mainland Norway and Svalbard. While substantial differences in the functional potential for PSM degradation and methanogenesis in wild and captive birds seem to be a direct consequence of their dissimilar diets.
Citation: Salgado-Flores A, Tveit AT, Wright A-D, Pope PB, Sundset MA (2019) Characterization of the cecum microbiome from wild and captive rock ptarmigans indigenous to Arctic Norway. PLoS ONE 14(3): e0213503. https://doi.org/10.1371/journal.pone.0213503
Editor: Alex V. Chaves, The University of Sydney, AUSTRALIA
Received: August 29, 2018; Accepted: February 24, 2019; Published: March 11, 2019
Copyright: © 2019 Salgado-Flores et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The sequence reads obtained from 16S amplicon sequencing and shotgun metagenomics are available at the Sequence Read Archive (SRA) database under the BioProject identifier PRJNA450906.
Funding: This study was funded by UiT The Arctic University of Norway, and Nansenfondet, 2014. Author P.B. Pope is supported from The Research Council of Norway (250479) and the European Research Commission (336355 - MicroDE). Author A.T. Tveit is supported by the Research Council of Norway FRIPRO Mobility grant project Time & Energy 251027/RU, co-funded by the European Research Council (ERC) under Marie Curie grant agreement no 608695. The publication charges for this article have been funded by a grant from the publication fund of UiT The Arctic University of Norway.
Competing interests: The authors have declared that no competing interests exist.
Ptarmigans (Lagopus muta) are gallinaceous birds within the subfamily Tetraoninae. With up to 30 different subspecies recognized, these birds show a circumpolar distribution in the northern hemisphere [1, 2]. Some physiological and phenotypical differences have been described between Svalbard (L. m. hyperborea) ptarmigan and rock ptarmigan from Scandinavia (L. m. muta), with Svalbard ptarmigans having a higher average body weight and size. In addition, Svalbard ptarmigans display striking variations in body weight and food intake throughout the year [3–5].
Despite being geographically isolated from one another, ptarmigans from Svalbard and ptarmigans in northern Scandinavia feed on diets with a similar polysaccharide composition, including a large fraction of plants rich in toxic plant secondary metabolites (PSMs) especially in periods from late fall to early summer [6–8]. Digestion can be compromised by the ingestion of PSMs, and in some cases being even toxic to the host . Like other herbivores, ptarmigans have high concentrations of microbes in their main anaerobic chamber, a paired ceca protruding out from the ileo-colonic junction . This cecal microbial consortium is important for the degradation of ingested plant material, yielding metabolites like Short-chain fatty acids (SCFAs) that can be absorbed by the host and used as a source of energy [10,11]. Herbivores such as the Greater Sage-Grouse (Centrocercus urophasianus) and woodrats (Neotoma lepida)) also consume diets with high levels of PSMs and possess a high proportion of PSM-degrading microbial taxa that are enriched in genes associated with the metabolism of aromatics (a key component of several PSMs) [12–14]. We therefore hypothesize that microbial groups specialized in PSM-degradation will be found in the cecum of both subspecies of ptarmigans.
In some cases a relatively high proportion of the microbiota is shared by wild and captive animals, as observed in woodrats . Furthermore, studies in captive birds indicated the presence of more homogeneous cecum microbiota, presumably due to limited dietary fiber and reduced variety of PSMs compared to their free-ranging counterparts [10, 15]. We hypothesized that captive rock ptarmigans would carry different microbiota than the two subspecies of free-ranging ptarmigans due to a strong influence of diet.
The SCFAs produced during microbial polysaccharide degradation can act as substrates for methane production by methanogenic Archaea, which catalyze the last step in the anaerobic fermentation of organic polymers . Methane production therefore represents a loss of metabolic energy for the host, in ruminants accounting for 2–12% of the gross energy intake (GEI) [17,18]. Avian-related methanogenesis has been neglected compared to other herbivores, with only a few studies assessing the methane production or characterizing the communities of methanogenic Archaea [19–23]. More research on the role of methanogenic Archaea and methane production would allow a better evaluation of the digestive physiology in birds.
A detailed description of the cecal microbiota (bacteria and archaea) from two subspecies of ptarmigans feeding on late autumn / winter natural plants on Svalbard and in northern Norway is presented. Moreover, the gut microbiota of wild and captive ptarmigans was investigated to identify how diet composition influences the microbial diversity and associated metabolic processes like PSM-degradation, hydrolysis of polysaccharides, fermentation, and methanogenesis.
Rock ptarmigans (Lagopus muta) in Norway and at Svalbard are not an endangered or a protected species. Samples for characterization of the cecum microbiome in wild rock ptarmigans from Svalbard (L. m. hyperborea) and northern Norway (L. m. muta) were collected as part of the legal hunt during fall/winter. Samples were imported from Svalbard to mainland Norway by permission from the Norwegian Food Safety Authority.
For photoperiodic experiments the birds were placed in indoor cages (L×D×H = 100×70×45 cm) in temperature and light controlled rooms specifically designed for chronobiology studies. The birds were visually, but not acoustically separated. Indoors, birds were kept under constant temperature (~5°C, well within the thermo neutral zone of ptarmigans) and defined light-dark (LD) regimes (e.g., short days LD 8:16; long days LD 16:8) for up to two months. Samples from captive Svalbard ptarmigans fed a diet of commercial pelleted feed, were collected after sacrificing the birds in our laboratory facility appropriate for that purpose, following the method for euthanasia by sedation with Ketamine (10–20 mg/kg) and xylazin (2–4 mg/kg) followed by cervical dislocation.
Housing and manipulation of captive ptarmigans as well as the method for euthanasia were approved by the National Animal Research Authority (FOTS_ID: 7971).
Free-ranging rock ptarmigans were hunted in Svalbard (SPW = Svalbard Ptarmigan in the Wild; n = 4) (78°N 16°E) and northern Norway (NPW = Norwegian Ptarmigan in the Wild; n = 4) (Troms county, 69.81°N 18.78°E) during the official hunting season (Sep—Dec). Immediately after shooting the bird cecal contents were fixed by mixing it with 70% ethanol and kept at 4°C until DNA extraction and molecular analyses. The diet composition of Svalbard ptarmigans were determined based on crop contents as previously shown . The willow species Salix polaris constituted the majority of the crop content in late fall, together with smaller quantities of Saxífraga cespitosa, Poa alpina leaves, and Drada spp. . No detailed description of the crop contents for the rock ptarmigans from mainland Norway used in this study was available. However, a description of the natural diet of rock ptarmigans during late fall/winter in northern Finland has been reported previously .
Two groups of captive Svalbard rock ptarmigans were artificially exposed to two different photoperiods mimicking light conditions during summer seasons (long photoperiod = 24 hours of light) (CP24h = Captive Ptarmigan 24 hours; n = 4) or winter seasons (short photoperiod = 6 hours of light) (CP6h = Captive Ptarmigan 6 hours; n = 4). The treatment lasted two months and was included to assess the effect of diet independent seasonal variations on the cecal microbiome, as previously observed in other animals . Captive birds were fed a commercial pelleted feed (Rypefor / Ptarmigan feed) (Norgesfor AS, Oslo, Norway) with the following composition: crude protein (10.4%), crude fat (6.2%), crude ash (8.3%), crude fiber (14.4%), and supplemented mineral and essential amino acids. Cecal contents were collected immediately after slaughter and kept frozen at -80°C, until DNA extraction and molecular analyses.
The genetic material (DNA) was extracted (for the amplicon and metagenomics sequencing studies) for each individual sample following the Repeated Beat Beating plus Column (RBB+C) method , and the resulting DNA was quantified using a NanoDrop 2000c spectrophotometer (Thermo Scientific, US).
Quantitative real-time PCR
Estimations of the cell density for methanogens and bacteria found in the different samples were carried out by quantitative real-time PCR (qRT-PCR). External standards were created for both microbial groups. Briefly, standards for methanogens were produced using purchased purified DNA previously extracted from a pure culture of Methanobrevibacter ruminantium (ATCC 35063) (DSMZ, Leipzig, Germany). The concentration of DNA was measured with NanoDrop and used to estimate the cell concentration for the axenic culture (3.97 × 107). Log serial dilutions of the DNA extract were subsequently prepared from an estimated initial cell concentration of 3.97 × 107 to 3.97 × 103 cells per mL of Mbr. ruminantium. Bacterial external standards were performed as described by Denman and McSweeney , using DNA from a pure culture of Ruminococcus flavefaciens (Sijpesteijn 1948) (DSMZ, Leipzig, Germany). Bacterial cells were anaerobically grown using a commercial culture medium (DSMZ medium 436) (DSMZ, Leipzig, Germany), counted microscopically and serially diluted from a range of 2.26 × 108 to 2.26 × 104 cells per mL. The primers for qRT-PCR were previously described in . Reactions were performed in a CFX384 Touch Real-Time PCR detection system (BioRad). A PCR mix consisting of 12.5 μL of SsoAdvanced Universal SYBR Green Supermix (BioRad, Hercules, CA, USA), 1.25 μL of each primer (400 nM), 5 μL DNA template (2 ng/μL), and distilled water to a 25 μL final volume was used. Experimental conditions were adjusted based on the DNA target (S1 Table), followed by a melting curve analysis to test for primer specificity and DNA contamination. Threshold cycles (Ct) values were automatically calculated by BioRad CFX software (v3.0), and PCR efficiency was computed based on the logarithmic fraction of the resulting sigmoid-shaped curve . Triplicates were run for each reaction and only those yielding the highest efficiency (R2≥0.996) were included in the analysis.
16S rRNA gene PCR amplification
The 16SrRNA gene for archaea and bacteria was amplified in an Eppendorf Mastercycler Gradient (Eppendorf AG, Hamburg, Germany) with a 25 μL reaction volume containing 12.5 μl of iProof High-Fidelity Master Mix (BioRad), 1 μl of each primer (400 nM), 1 μl of DNA template, and 1.25 μl of dimethyl sulfoxide (DMSO) to increase reaction efficiency. Amplification of 16S rRNA genes from archaea and bacteria was performed with the archaeal primer set 340F and 1000R , giving a 650 bp amplicon product, and the bacterial primer sets 27F and 515R [30,31], yielding a 500 bp amplicon product. Both primers sets contained the following complements in addition to their primer sequence: the Life Sciences primer A and B sequences necessary for pyrosequencing, and an 8-nt Multiplex Identifier (MID) included only in the reverse primer , which is used for downstream sample identification. Experimental conditions for both archaea and bacteria were previously described in . The resulting PCR products were checked on a 1.5% agarose gel and the DNA concentration was measured with a Qubit fluorimeter (Invitrogen). Due to the lack of amplification with primers targeting the archaeal 16S rRNA gene, some samples within the NPW group were discarded from the analyses (NPW2 and NPW5). Samples were pooled in equimolar amounts, run in an agarose gel, excised and purified using NucleoSpin Gel and PCR Clean-up kit (Macherey-Nagel). The resulting purified, pooled DNA was stored at -20°C until sequencing. Sequencing was performed with a 454/Roche GS FLX instrument, using LIB-L chemistry, at the Norwegian Sequencing Centre (NSC), in Oslo.
For captive ptarmigans, characterization of the cecal microbiota was performed using prokaryotic (16S) and eukaryotic (18S) small subunit (SSU) rRNA genes retrieved from the metagenomics dataset (see ‘Sequence analysis’ below).
For metagenomics analyses samples from wild (SPW and NPW) and captive (CP6hr and CP24hr) ptarmigans were used, giving a total of 16 samples (four per group). DNA extraction was performed as described above. Non-amplified DNA extracts were checked in a 1% agarose gel, and stored at -20°C prior to sequencing.
The sequencing service was provided by the Norwegian Sequencing Centre (www.sequencing.uio.no), a national technology platform hosted by the University of Oslo and Oslo University Hospital and supported by the "Functional Genomics" and "Infrastructure" programs of the Research Council of Norway and the Southeastern Regional Health Authorities. Libraries were prepared using TruSeq PCR-free reagents (Illumina, San Diego, CA) according to manufacturer’s instructions. Sequencing was performed on an Illumina HiSeq 3000 with 150 bp paired end reads according to manufacturer's instructions. Image analysis and base calling were performed using Illumina's RTA software version 2.7.6 and bcl2fastq v22.214.171.124. Reads were filtered to remove those with low base call quality using Illumina's default chastity criteria.
Bacterial and archaeal 16S rRNA reads from samples of wild ptarmigans were processed using the Quantitative Insights into Microbial Ecology (QIIME) software (v1.9.1) . Sequences with an average quality score above 25 were selected and trimmed to 650 bp and 500 bp length for archaea and bacteria, respectively. Reads were clustered into Operational Taxonomic Units (OTU) based on a 97% similarity criterion, discarding any putative chimeras, using the QIIME-incorporated version of USEARCH (v5.2.236). The most abundant sequence in each OTU was used as representative in downstream analyses. Representative sequences from each OTU were aligned against the Greengenes reference database (release date October 2010; http://greengenes.lbl.gov/) and taxonomically annotated as described in Salgado-Flores et al. . Weighted UniFrac distance matrices were calculated for pairwise comparisons (beta-diversity) of subsampled dataset adjusted to the one with the lowest counts to avoid any potential bias  (S2 Table). Weighted UniFrac distances were finally used to create Principal Coordinate Analysis (PCoA) plots. A network map comparing the amount of shared OTUs between pairs of samples were computed with the make_otu_network.py script in QIIME, and visualized with Cytoscape (v3.1.1) .
Shotgun metagenomic sequencing data was analyzed as previously described  (S3 Table). Briefly, paired-end sequences were assembled with Pandaseq (v2.9) , with a 10 bp minimum overlap. Assembled sequences were preprocessed with Prinseq lite (v0.20.4) ; sequences with an average quality value below 20 and more than five unambiguous bases were removed, and all but one sequence of identical sequence pools were discarded to avoid any bias due to artificial replication during sequencing. A 5,000,000 sequences subset from each sample was used for downstream analysis. DNA sequences were used as query for BLASTX analysis using DIAMOND alignment tool (‘—more_sensitive’ option; e-value 1e-4) against the RefSeq-protein database from NCBI . The same subset was screened using the ‘hmmsearch’ tool within the package HMMER  with hidden markov models (HMMs) for protein families (Pfam release 30, http://pfam.janelia.org]. All searches were carried out on the High Performance Computing (HPC) Stallo at UiT-The Arctic University of Norway (http://docs.notur.no/uit). Output files from Diamond were uploaded and visualized in MEGAN Community Edition (default parameters: minimum bit score of 50, minimum node support of 1) .
For the taxonomically comparisons between captive and wild ptarmigans, 16S and 18S rRNA gene sequences were extracted using the ‘SortMeRNA’ tool and Silva rRNA gene database 123 (July 23rd, 2015) . The extracted SSU rRNA gene reads were taxonomic annotated by comparison to the SilvaMod rRNA gene reference database using the online version of the LCA Classifier available at the Classification Resources for Environmental Sequence Tags (CREST) website (minscore: 150) (http://apps.cbu.uib.no/crest/lcaclassify) .
For qRT-PCR analysis, two-tailed Student’s t-test was applied for pair-wise comparisons between wild and captive groups’ data for methanogens and bacteria. Moreover, one-way analysis of variance (ANOVA) tests were performed for multi-group analysis for bacteria and methanogens qRT-PCR data. When statistical significance was observed, Tukey´s range test was performed to obtain further verification at a group level . All the aforementioned tests were performed with ‘R’ statistical package . A non-parametic PERMANOVA test (999 permutations), from the package ‘vegan’ in the ‘R’ tool included in QIIME, based on weighted UniFrac distances, was run to test significance for any difference between Svalbard and northern Norway ptarmigans cecum dataset (bacteria and archaea). Bar plots representing relative abundances (normalized to the total reads for each group) for the different KEGG categories were compared with the STAMP software [46, 47]. Statistical comparisons between functional categories were performed using White’s non-parametric t-test , corrected with Benjamini–Hochberg False Discovery Rate correction. Only those categories yielding significant differences were represented.
Quantitativereal time PCR analysis
We estimated the cell densities of Bacteria and methanogenic Archaea in order to assess the importance of the methane producing microbial community in comparison to the total bacterial community. Overall, bacterial mean counts were approximately 20 times higher than for methanogenic Archaea (1.40×109 vs 4.73×107–5.31×10716S rRNA gene copies/gww) (Fig 1). No significant differences were observed between the mean cell counts of bacteria (Student’s t-test: p = 0.927) or methanogens (p = 0.716) in SWP and NWP. Captive birds possessed significantly lower mean cell densities for both bacteria (CP24h: 1.02×109; CP6h: 8.35×108 16S rRNA gene copies/gww; ANOVA: p-value = 0.01) and methanogens (CP24hrs: 8.89×105; CP6hrs:4.05×105 16S rRNA gene copies/gww; ANOVA: p-value<0.01) compared to wild ptarmigans (Fig 1). However, when comparisons were performed based on individual groups (e.g. SPW versus CP24h) and not by living conditions (i.e. captive versus wild), no significant differences in bacterial mean cell densities were observed between wild and captive ptarmigans (Fig 1; ANOVA: p-value = 0.09). Methanogens mean cell densities were significantly different between all the sub-groups for captive and wild ptarmigans (Fig 1; ANOVA: p-value<0.05; Tukey´s: p<0.05). Furthermore, no significant differences in bacterial or methanogen abundances were observed between captive ptarmigans exposed to different light regimes (methanogens: p = 0.189; bacteria: p = 0.412) (Fig 1).
Bacteria (A) and methanogens (B) populations were estimated by qRT-PCR. Total counts are presented as number of cells per gram of wet weight (cells/gww). Bacterial counts were plotted using a linear scale whereas for methanogens a logarithmic scale was applied for better visualization. Wild rock ptarmigan northern Norway (NPW) (n = 4); wild Svalbard rock ptarmigan (SPW) (n = 4); captive Svalbard ptarmigan exposed to long (CP24h) (n = 4) or short (CP6h) (n = 4) photoperiods. Pair-wise statistical comparisons were calculated with a two-tailed Student’s t-test. Statistical analysis for multiple groups were calculated with ANOVA tests. Tukey´s range test was applied for further verification at group level * (p-value<0.05); ** (p-value<0.01).
Bacterial 16 rRNA gene.
A total of 100,817 and 88,259 bacterial 16S rRNA gene sequences (average length 500 bp) were obtained from the Svalbard ptarmigans and Norwegian wild ptarmigans, respectively. Clustering of OTUs resulted in a total of 1,939 chimera-free OTUs at 97% sequence similarity, from which 1,754 (90.4% total OTUs) were shared between the groups of wild ptarmigans. The bacterial communities of SWP and NWP did not separate in weighted UniFrac-based PCoA plots (Fig 2A). Further testing confirmed that there is no significant difference in the bacterial communities between wild ptarmigans from Svalbard and mainland Norway (pseudo-F = 0.823; p-value = 0.79).
Plots were generated based on weighted UniFrac distance matrices. A) PCoA plot of bacterial community structure; B) PCoA plot of archaeal community structure. Blue square: Svalbard ptarmigan. Red circle: rock ptarmigans northern Norway.
The same dominant phylotypes were observed in SWP and NWP samples, with Actinobacteria (SPW: 28.1%; NPW: 27.7% of total sequences, on average), Firmicutes (SWP: 24%; NWP: 25.1%), Bacteroidetes (SWP: 18.3%; NWP: 19%), and Synergistetes (SWP: 10.7%; NWP: 11.1%) as the major phyla (Table 1). No significant differences were found in the relative abundances of any of these phyla, between SWP and NWP (p>0.05) (Table 1). Although largely similar, some differences between shotgun metagenomics SSU and 16S rRNA gene amplicon sequencing results were observed. Firmicutes was the dominant phylotype (SWP: 33.9%; NWP: 32.6%) instead of Actinobacteria (SWP: 19.3%; NWP: 19.3%).
Analysis of SSU rRNA gene sequences extracted from shotgun metagenomics data showed that captive ptarmigans were dominated by few bacterial phyla: Firmicutes (CP24h: 81.7%; CP6h: 82.8%), Bacteroidetes (CP24h: 10.7%; CP6hrs: 10.7%) and Proteobacteria (CP24h: 6.2%; CP6h: 4.3%) (Fig 3). Similarly, only a few families dominated: Ruminococcaceae (CP24h: 41.6; CP6h: 42.8), Lachnospiraceae (CP24h: 29.6; CP6h: 31.2), and Enterobacteriaceae (CP24h: 5.5; CP6h: 3.7). No significant differences in bacterial groups between captive ptarmigans exposed to different photoperiods were observed (Fig 3).
Bar charts illustrates the taxonomical relative diversity of bacteria at phylum level from SSU sequences benchmarked against the SilvaMod rRNA reference database as described in material and methods section. Wild rock ptarmigan northern Norway (NPW) (n = 4); wild Svalbard rock ptarmigan (SPW) (n = 4); captive Svalbard ptarmigan exposed to long (CP24h) (n = 4) or short (CP6h) (n = 4) photoperiods.
Archaeal 16 rRNA gene.
In total, 109,506 16S rRNA sequences (average length 635 bp) were obtained from cecal samples of wild Svalbard ptarmigans and Norwegian ptarmigans after quality filtering. We identified a total of 67 OTUs (97% threshold), of which 54 were shared by SWP and NWP. All OTUs were assigned to one single phylum, Euryarchaeota (Table 1). Similar to bacteria, beta-diversity comparisons based on weighted UniFrac distances showed that the SWP and NWP samples grouped together (Fig 2B). In order to test if there were any differences between the archaeal community profiles a PERMANOVA test with 999 permutations was performed, and no significant differences were observed (pseudo-F = 0.23; p-value = 0.76).
The newly characterized family Methanomassiliicoccaceae (SWP: 62.8% of total sequences, on average; NWP: 71.7%) was the dominant archaeal taxon. Genera belonging to the family Methanocorpusculaceae constituted the other major archaeal group (SWP: 35.4%; NWP: 27.8%). No significant differences were found in the relative abundances of these taxa between SWP and NWP (p>0.05) (Table 1). Within captive ptarmigans, results based upon shotgun metagenomics sequencing data indicated Methanobacteriaceae as the only archaeal family identified.
PSMs degradation and methanogenesis pathways.
In order to obtain insights on the effects of diet on the genetic repertoire of the microbiome from wild and captive ptarmigans, genes encoding enzymes for PSM detoxification were mapped using Kyoto Encyclopedia of Genomes and Genes (KEGG). Xenobiotics degradation, a category comprising genes for the degradation of components found in several PSMs were more abundant in wild than captive ptarmigans (Fig 4A). Genes for pathways of chloroacrylic acid, benzoate, and nitrotoluene degradation were the most abundant sub-category of xenobiotics degradation in both wild and captive ptarmigans, and were significantly higher in wild than captive ptarmigans (Fig 4B). Furthermore, the relative abundance of genes for the degradation of aromatic compounds such as caprolactam (x2.19-fold), xylene (x1.77-fold), carbazole (x1.69-fold), were significantly more abundant in wild than captive ptarmigans (Fig 4B). Genes for the degradation of PSMs (xenobiotics KEGG category) in our wild ptarmigans were assigned to the phylum Firmicutes (families Clostridiaceae, Lachnospiraceae, and Eubacteriaceae), Bacteroidetes (genus Bacteroides), and Actinobacteria (genera Eggerthella and Slackia). The wild ptarmigan microbiota also carried genes specific for a metabolic pathway leading to the conversion of phenol and catechol to pyruvate (Fig 5). Genes for the full metabolic pathway were not present in the microbiota of captive ptarmigans.
Comparisons were carried out with calculated mean values of shotgun metanogenomic sequences from captive and wild ptarmigans. The mean proportion of total sequences is represented allocated to: A) The whole KEGG class “Xenobiotics and Biodegradation Metabolism”; B) KEGG pathways within the “Xenobiotics and Biodegradation Metabolism” class. Only those categories showing significant differences between both groups of ptarmigans are presented. Statistical tests were performed using the White’s non-parametric t-test applying the Benjamini-Hochberg False Discovery Rate correction. KO codes are presented in the beginning of each name tag (if any). Wild ptarmigans (yellow); captive Svalbard ptarmigan (blue).
Full name and KEGG entries for the enzymes involved at each step (1–5) are given in the legend box. Only a few of these genes were present (or at a minor proportion) in captive ptarmigans. A similar pathway has been previously identified in the cecum of Greater sage-grouse : 1. Phenol hydroxylase (EC.1.14.13.-); 2. Catechol-2,3-dioxygenase (EC.126.96.36.199); 2-hydroxymuconic semialdehyde hydrolase (EC.188.8.131.52); 2-keto-4-pentenoate hydratase (EC.184.108.40.206); 4-hydroxy-2-oxopentenoate (EC.220.127.116.11).
Methanogenesis plays an important role in anaerobic fermentation by catalyzing its last steps by using waste products from microbial metabolism as substrates (e.g. SCFAs) . However, methane production represents an energy loss for the host [17,18]. Genes involved in methane production were screened to obtain a better understanding on the microbial fermentation in the ceca of our wild ptarmigans. Methyl coenzyme M reductase [EC:18.104.22.168], a gene encoding the key terminal enzyme of methanogenesis, was present at an average relative abundance of 0.01% in both SWP and NWP (Table 2). Genes encoding enzymes involved in hydrogenotrophic and methylotrophic methanogenesis were also identified (Table 2; S1 Fig). Overall, higher relative abundances of genes for methanogenesis pathways were observed in wild ptarmigans (x1.7). Despite the detection of methanogenic Archaea in the SSU rRNA datasets, no genes for the methyl coenzyme M reductase were found in the captive ptarmigans (Table 2).
Identification of Pfam domains involved in polysaccharide degradation
In order to detail the carbohydrate-degrading potential in wild and captive ptarmigans, the major enzymes responsible for plant polysaccharide degradation were characterized. A screening of the major glycoside hydrolases (GHs), i.e. enzymes catalyzing the hydrolysis of the glycosidic bonds in complex sugars, and other enzymes involved in carbohydrate degradation was performed (Pfam release 25, https://pfam.janelia.org). In total, genes for 89 and 91 different GHs and carbohydrate degradation families were identified in the cecal microbiome of wild and captive ptarmigans, respectively (S4 Table; S2 Fig). No differences were found in the relative abundance of the major GH families (families accounting for ≥1% of sequences assigned to a carbohydrate degrading enzymes) between wild ptarmigans from Svalbard and northern Norway (Table 3). Genes encoding enzymes associated to GH family 77, catalyzing 4-α-glucanotransferase or amylomaltase activity (http://www.cazy.org/), were the single most abundant family in wild ptarmigans (9.5% total hits). In addition, genes involved in starch metabolism (e.g. starch phosphorylase) [EC.22.214.171.124]) were also found in wild ptarmigans (Wild: 0.19% of total KEGG-annotated sequences). GH families acting on xylo-oligosaccharides with a broad repertoire of catalytic activities like β-glucosidase, α and β-galactosidase (GH1-3, GH31, GH36, GH42) and xylosidase (GH43) were abundant in wild ptarmigans and accounted, altogether, for 49.8% total hits, on average (Table 3). Genes for rhamnose (GH78) (6% of total hits) and mannose (GH92) (4.6%) degradation were also among the major GH families in wild ptarmigans (Table 3; S4 Table). Genes for GH families involved in the degradation of pectin constituents such as rhamnose and galacturonans (e.g. GH78, GH28, GH53, GH106), in wild ptarmigans shows that the microbiota has the potential for pectin degradation (Table 3; S4 Table). Moreover, genes for pectinesterase [EC.126.96.36.199], an enzyme catalyzing the hydrolysis of pectin into pectate and methanol, were present in both groups of wild ptarmigans.
The same genes for carbohydrate degradation dominated wild ptarmigans and captive ptarmigan metagenomes (Fig A in S2 Fig), but with differences in some categories, e.g. higher relative abundances of genes for endohemicellulose and starch degradation in wild ptarmigans compared to both captive ptarmigan groups (Fig C and D in S2 Fig).
Total populations of bacteria and methanogens in cecum of wild and captive ptarmigans
The abundance of bacteria and methanogens were lower in the ptarmigans of this study than previously observed in several ruminants, and chicken [49–51] (S5 Table); however, bacterial counts were higher than described in wild seabirds . Few studies have reported estimations of methanogen populations in birds, with total numbers being higher in the crop of the wild Hoatzin (Opisthocomus hoazin) and lower in the ceca of captive chickens than in the ceca of the ptarmigans in our study [20,21].
Significant differences were observed between wild and captive ptarmigans for both bacteria and methanogens (Fig 1). Such differences may be a consequence of commercial feed (as the one given to captive ptarmigans) being ground and pelleted, which leads to faster digestion rates and, therefore, in shorter retention times and lower microbial growth . In contrast, diets possessing a higher variety of structural polysaccharides, as those presumably consumed by the wild ptarmigans, would require longer retention times for their degradation, which may lead to higher microbial growth rates for bacteria and methanogens.
Moreover, considering the important role played by anaerobic fungi in the physical disruption and degradation of fiber compounds in, for instance, the rumen [54,55], it might also be speculated that there is a higher presence of this microbial group in the cecum of wild ptarmigans. Longer feed passage rates would provide enough time to fungi to proliferate and help physically disrupt the fiber. New studies would have to integrate data from these microbes to further understand microbial fiber degradation in wild ptarmigans.
Specific bacterial taxa may mediate PSM-detoxification in wild ptarmigans
Bacterial communities in wild and captive ptarmigans were composed differently (Fig 3). Microbiome variation between wild Japanese captive Svalbard ptarmigans have been reported previously . Consistently, these wild Japanese ptarmigans and the wild ptarmigans in our study possessed similar cecal microbiota (Table 1). Genes associated with PSM-degradation in our wild ptarmigans were allocated to the bacterial phyla Firmicutes (families Clostridiaceae, Lachnospiraceae, and Eubacteriaceae), Bacteroidetes (genus Bacteroides), and Actinobacteria (genera Eggerthella and Slackia). Similarly, genes for xenobiotics biodegradation in the cecum of free-ranging Greater sage-grouse were assigned to the genera Bacteroides, Eggerthella and Clostridium . These comparisons suggest that bacterial phylotypes putatively perform the same metabolic tasks in wild ptarmigans and Greater sage-grouse, PSM-degradation being driven by several bacterial groups. On the other hand, studies from the cecum and rumen microbiota in wild Japanese ptarmigans and Hawaiian goat, respectively, both animals following a diet rich in PSM, showed most of the genes involved in PSM degradation assigned to the phylum Synergistetes [15, 56].
Genes for the conversion of phenol and catechol into pyruvate were mostly assigned to Firmicutes (class Clostridia), Actinobacteria (order Micrococcales) and Proteobacteria (class Alphaproteobacteria) (Fig 5). Enrichment of genes encoding enzymes of this metabolic pathway was also described in the cecal microbiome of the Greater sage-grouse, mostly assigned to the genus Anthrobacter (family Micrococcaceae) . Some of these genes also encode enzymes involved in xylene and benzoate metabolism, accounting for the high relative abundance of such overlapping genes in wild ptarmigans.
Chemical analysis by Hansen et al. (2011) revealed the presence of several PSMs, e.g., the two benzene ring flavonoid Catechin, in Salix polaris tissues . High contents of PSMs in Salix spp. might select for bacterial groups bearing genes mediating PSMs degradation. The presence of genes encoding enzymes involved in the degradation of phenol/catechol compounds in wild ptarmigans (Fig 5), but absence in the captive group, indicates that such compounds are utilized by the wild ptarmigan cecum microbiota. Whether the intermediates in PSM degradation (e.g., pyruvate) can be used by the host remains unknown.
Hemicellulose and starch-hydrolyzing GHs in cecum of wild ptarmigans
A broad repertoire of GHs families were found in the ceca of wild and captive ptarmigans. The total number of GHs was higher than those reported for other herbivores (e.g. bovine and reindeer rumen, hindgut fermenters, etc.) and comparable to feces from captive Asian elephants (see Table 2 in ) (S6 Table). The broad range of GHs, irrespective of the diet composition, for wild and captive ptarmigans suggests an inherent high richness in hydrolytic enzymes in the cecum microbiota of ptarmigans. Such diversity could allow this microbial consortium to readily adapt to a changing diet throughout seasons . The presence of higher relative abundances of GHs involved in starch degradation and the lower proportions of cellulases observed in wild ptarmigans compared to the captive groups was unexpected (S2 Fig). The ingestion of a natural diet would, in theory, require a higher abundance of enzymes involved in the deconstruction of complex polysaccharides (e.g. cellulose) compared to the captive birds fed commercial pellets, whose feed is compositionally consistent as well as physically ground and more easily digestible. Interestingly, crude fiber contents in the commercial pellets used in this study (about 14% of the total chemical composition) was comparable to that reported from crop contents of wild Svalbard ptarmigans during periods when S. Polaris constituted the major food item (see Materials & Methods; ). The major plant species in the crop of our wild Svalbard ptarmigans was also found to be S. polaris . This similarity in the crude fiber content for both dietary regimes may account for the comparable overall diversity in GHs present in wild and captive ptarmigans. Genes encoding endohemicellulases were found at a higher relative proportion in wild ptarmigans (Table 3; S2 Fig).
The GH families degrading hemicellulose compounds (GH1-3, GH8, GH10, GH16, GH26, GH31, GH36, and GH42) constituted the largest fraction of total GH genes in wild ptarmigans indicating the great potential to degrade this type of polysaccharides by the cecum microbiota (Table 3). These findings may suggest that, at least during late autumn / winter, the diet consumed by wild ptarmigans contains plants with high hemicellulose content. Our results may contradict previous reports on the diet of these birds indicating their capacity to select plants (or parts of plants) with high-nutritional value (low in fiber) in all periods [1,8]. However, the high presence of GH families encompassing genes for the degradation of starch (e.g.GH77, with amylomaltase activity ) (Table 3), and genes involved in starch metabolism (e.g. starch phosphorylase [EC.188.8.131.52]) also indicates that starch-containing compounds may constitute an important part of the nutrition in the wild ptarmigan diet. These findings, together with the identification of genes encoding GH families (e.g. GH78, GH28, GH53, GH106) and enzymes involved in the depolymerization of pectic compounds (e.g. pectinesterase), suggests that wild ptarmigans house a versatile microbiota that allows the digestion of plants with variable fiber contents. This versatility may be beneficial during periods when high-quality foodstuffs are scarce as in autumn / winter. Nonetheless, new studies describing the cecal metatranscriptome are necessary to find out which GH genes are expressed and the role played the different type of polysaccharides in the nutrition of wild ptarmigans.
Archaeal taxonomy and methanogenesis in wild ptarmigans
To our knowledge, this is the first study reporting the presence of Methanomassiliicoccaceae-associated members in birds (Table 1). Methanomassiliicoccaceae-related phylotypes have been described in the gut of several hervibores (e.g. rhinoceroses, muskoxen, kangaroo, elephant, etc.) [60–62], and anaerobic digesters , but at a lower relative abundance than observed in wild ptarmigans. Methanol and methylamines are the main growth-substrates for members of the Methanomassiliicocacceae family [61, 64]. Methylamines may result from the degradation of amino acids and compatible solutes, and this methanogenesis pathway was identified in peat soil and gut samples [61, 65]. Only one gene encoding one of the key enzymes mediating methanogenesis from methylamines (trymetylamine—corrinoid protein Co-methyltransfersase [EC.184.108.40.206]) was found in captive and wild ptarmigans (captive: 0.011%; wild: 0.005% of total genes) (Table 2). However, in both cases this gene was not assigned to Methanomassiliicoccaceae but to bacteria within the order Clostridiales. In wild ptarmigans, the major genus within the order Clostridiales was Acetobacterium. A species of this genus, Acetobacterium woodii, has been shown to interact with methanogens in syntrophic relationships for the transfer of H2 and acetate, but to our knowledge no indications for syntrophic methylamine metabolism have been shown .
Production of methanol for methanol-using methanogens like Methanomassiliicoccaceae may result from the microbial degradation of pectin, a common component of the middle lamella of higher plants (e.g., in fruit peels) [67, 68]. Berries of Empetrum nigrum constitute one of the major food items during fall/winter for rock ptarmigans inhabiting northern Scandinavia . Salix polaris and Saxifraga spp. are the main plant species consumed by Svalbard ptarmigans during these same seasons . Both are dicotyledonous plants, which hypothetically account for the presence of genes for hydrolytic enzymes involved in pectin degradation in another arctic herbivore, the Svalbard reindeer (Rangifer tarandus plathyrhyncus) . The presence of genes encoding for GH families and enzymes (e.g. pectinesterase) involved in pectin degradation supports the presence of pectin-degrading microbes. Taxonomic annotation of sequences for pectinesterase genes pointed to species within Bacteroidetes (mainly families Bacteroidaceae and Prevotellaceae) and Firmicutes (order Clostridiales) as the major pectin degraders in wild ptarmigans. Pectinolytic activity has been previously observed in several members of these bacterial groups [69, 70]. Genes related to methanol-based methanogenesis were identified in wild ptarmigans, almost exclusively assigned to Methanomassiliicocacceae, agreeing with the high abundances of Methanomassiliicocacceae observed (Table 1). We hypothesize that pectin metabolism in the cecum of ptarmigans leads to the formation of methanol that is utilized by Methanomassiliicocacceae-related methanogens. Methanogenesis from methanol originating from pectin degradation have been postulated from the study of frugivorous hindgut fermenters like the orangutan (Pongo pygameus), but with Methanosphaera stadtmanae being the major methanogens .
It is concluded that wild ptarmigans in these two different geographical regions (Svalbard and Norway) share highly similar microbial communities. The presence of microbial taxa with the potential to decompose PSMs reflects an ability of ptarmigans to bypass the toxicity exerted by PSMs. The broad range of GHs found in wild and captive ptarmigans suggests the existence of a versatile cecum microbiota that allows the consumption of plants with different fiber contents. Methanol-based methanogenesis seems to constitute one of the major methanogenesis pathways in wild ptarmigans based on the dominance of Methanomassiliicoccaceae methanogens and the presence of genes for this pathway. The higher abundance of methanogens in wild ptarmigans may be linked to the ingestion of a natural diet consisting of unprocessed polysaccharides that result in longer fermentation necessary to allow growth of methanogenic Archaea.
S1 Fig. Schematic representation of the main pathways of methanogenesis and related enzymes based on KEGG-mapped genes found in the ceca of wild ptarmigans.
Full name and KEGG entries (whenever possible) for the enzymes involved in each metabolic reaction (1–15) are given in the legend box. Red arrows indicate those substrates used for methane production alternative to the dominant H2-based methanogenesis. Red color numbers are used to highlight those enzymes characterized at a higher proportion in the cecal microbiome of wild ptarmigans used in this study compared to captive ptarmigans. Abbreviations: Fdred: reduced ferredoxin; Fdox: oxidized ferredoxin; F420H2 reduced coenzyme F420; F420: oxidized coenzyme F420; MFR: methanofuran; H4MPT: tetrahydromethanopterin; CoM-SH: coenzyme M; CoB-SH: coenzyme B; CoM-S-S-CoB: heterodisulfide of CoM and CoB; SH-CoA: coenzyme A; CHO-MFR: formyl-MFR; CHO-H2MPT: formyl-H4MPT; CH-H4MPT: methenyl-H4MPT; CH2-H4MPT: methylene-H4MPT; CH3-H4MPT: methyl-H4MPT; CH3-SCoM: methyl-CoM; CH3-CO-S-CoA: methyl-AcylCoA; CO-S-CoA: acyl-CoA; CoA-SH: coenzyme A. * Electrons for the reduction of methyl groups are obtained by methyl-group oxidation to CO2 reversible to methanogenesis. Modified from .
S2 Fig. Relative abundance of genes encoding hydrolytic enzymes in the cecum of wild and captive ptarmigans.
Values are given as percentages of the total sequences assigned to protein family domains involved in polysaccharide degradation. Bar charts were scaled according to the relative abundance for each functional category. Identification of each domain was performed by BLAST against the Pfam-hmm database (see Methods for more details). a) Total results for the major functional categories; b) Celullases; c) Enzymes involved in starch degradation; d) Endohemicellulases; e) Debranching enzymes; f) Olygosasaccharides hydrolases (GHs accounting for <1% of total hits each were combined in a single group named “COMBINED GROUP GH <1%”for a better visualization).
S1 Table. qRT-PCR conditions for amplification of methanogens and bacteria in cecum samples from wild and captive ptarmigans.
S2 Table. Number of sequences per sample used for 16S rRNA gene amplicon sequencing analysis.
S3 Table. Information on the shotgun metagenomics results on cecal samples from wild and captive ptarmigans.
S4 Table. Total pfam families associated to the degradation of polysaccharides in cecal samples from wild ptarmigans.
Results are represented as the percentage of sequences associated to a specific GH family in relation to the total pfam hits.
S5 Table. Bacterial and methanogens counts from different host animals.
Numbers are presented with their corresponding units.
We are grateful to the staff of animal caretakers at the Department of Arctic and Marine Biology, Hans Lian, Hans Arne Solvang and Renate Thorvaldsen, for their help with the captive birds. We thank Ruben Eidesen (The University Centre in Svalbard, UNIS) for the sampling of wild Svalbard rock ptarmigan, and Even Jørgensen (UiT) for the sampling of wild rock ptarmigans in northern Norway. We are grateful to Gabriela Wagner and Daniel Appenroth (UiT) for giving us access to the captive Svalbard rock ptarmigans, and Andreas Nord (UiT) for measuring methane emissions from captive Svalbard rock ptarmigans.
- 1. Pedersen ÅØ, Overrein Ø, Unander S, Fuglei E. Svalbard rock ptarmigan (Lagopus mutus hyperboreus)–a status report. Norsk Polarinstitutt. 2005, Tromsø, Norway. Rapportserie nr. 125.
- 2. Sahlman T, Segelbacher G, Höglund J. Islands in the ice: colonisation routes for rock ptarmigan to the Svalbard archipelago. Ecography. 2009;32:840–848.
- 3. Mortensen A, Unander S, Kolstad M, Blix AS. Seasonal changes in body composition and crop content of Spitzbergen Ptarmigan Lagopus mutus hyperboreus. Ornis Scand. 1983;14:144–148.
- 4. Steen JB, Unander S. Breeding biology of the Svalbard rock ptarmigan Lagopus mutus hyperboreus. Ornis Scand. 1985;16:191–197.
- 5. Stokkan KA, Mortensen A, Blix AS. Food intake, feeding rhythm, and body mass regulation in Svalbard rock ptarmigan. Am J Physiol. 1986;251:264–267.
- 6. Pulliainen E. Winter nutrition of the rock ptarmigan, Lagopus mutus (Montin), in northern Finland. Ann Zoo Fennici. 1970;7:295–302.
- 7. Eidesen PB, Müller E, Pedersen ÅØ. Hva spiser egentlig svalbardrypa? Norsk Polarinstitutt. 2015.
- 8. Unander S, Mortensen A, Elvebakk A. Seasonal changes in crop content of the Svalbard ptarmigan Lagopus mutus hyperboreus. Polar Research. 1985;3:239–245.
- 9. Striby KD, Wambolt CL, Kelsey RG, Havstad KM. Crude Terpenoid Influence on In Vitro Digestibility of Sagebrush. J Range Manage. 1987;40:244–248.
- 10. Hanssen I. A comparison of the microbiological conditions in the small intestine and caeca of wild and captive willow grouse (Lagopus lagopus lagopus). Acta Vet Scand. 1979;20:365–371. pmid:506874
- 11. Moss R. Gut size, body weight, and digestion of winter foods by grouse and ptarmigans. Condor. 1983;85:185–193.
- 12. Kohl KD, Dearing MD. Wild-caught rodents retain a majority of their natural gut microbiota upon entrance into captivity. Environ Microbiol Rep. 2014;6:191–195. pmid:24596293
- 13. Kohl KD, Weiss RB, Cox J, Dale C, Dearing MD. Gut microbes of mammalian herbivores facilitate intake of plant toxins. Ecol Lett. 2014;17:1238–1246. pmid:25040855
- 14. Kohl KD, Connelly JW, Dearing MD, Forbey JS. Microbial detoxification in the gut of a specialist avian herbivore, the Greater Sage-Grouse. FEMS Microbiol Lett. 2016;363: pmid:27242374
- 15. Ushida K, Segawa T, Tsuchida S, Murata K. Cecal bacterial communities in wild Japanese rock ptarmigans and captive Svalbard rock ptarmigans. J Vet Med Sci. 2015;78:251–257. pmid:26468217
- 16. Bodas R, Prieto N, García-González R, Andrés S, Giráldez FJ, López S. Manipulation of rumen fermentation and methane production with secondary metabolites. Anim Feed Sci Tech. 2012;176:78–93.
- 17. Murray RM, Bryant AM, Leng RA. Rates of production of methane in the rumen and large intestine of sheep. Br J Nutr. 1976;36:1–14. pmid:949464
- 18. Holter JB, Young AJ. Methane prediction in dry and lactating Holstein cows. J Dairy Sci. 1992;75:2165–2175. pmid:1401368
- 19. Gasaway WC. Methane production in rock ptarmigan (Lagopus mutus). Comp Biochem Physiol A Comp Physiol. 1976;54:183–185. pmid:4267
- 20. Saengkerdsub S, Anderson RC, Wilkinson HH, Kim WK, Nisbet DJ, Ricke SC. Identification and quantification of methanogenic Archaea in adult chicken ceca. Appl Environ Microbiol. 2007;73:353–356. pmid:17085694
- 21. Wright AD, Northwood KS, Obispo NE. Rumen-like methanogens identified from the crop of the folivorous South American bird, the hoatzin (Opisthocomus hoazin). ISME J. 2009;3:1120–1126. pmid:19387486
- 22. Frei S, Dittmann MT, Reutlinger C, Ortmann S, Hatt JM, Kreuzer M, et al. Methane emission by adult ostriches (Struthio camelus). Comp Biochem Physiol A Mol Integr Physiol. 2015a;180:1–5.
- 23. Frei S, Hatt JM, Ortmann S, Kreuzer M, Clauss M. Comparative methane emission by ratites: Differences in food intake and digesta retention level out methane production. Comp Biochem Physiol A Mol Integr Physiol. 2015b;188:70–75.
- 24. Bailey MT, Walton JC, Dowd SE, Weil ZM, Nelson RJ. Photoperiod modulates gut bacteria composition in male Siberian hamster (Phodopus sungorus). Brain Behav Inmun. 2010;24:577–584.
- 25. Yu Z, Morrison M. Improved extraction of PCR-quality community DNA from digesta and fecal samples. Biotechniques. 2004;36:808–812. pmid:15152600
- 26. Denman SE, McSweeney CS. Development of a real-time PCR assay for monitoring anaerobic fungal and cellulolytic bacterial populations within the rumen. FEMS Microbiol Ecol. 2006;58:572–582. pmid:17117998
- 27. Salgado-Flores A, Hagen LH, Ishaq SL, Zamanzadeh M, Wright AD, Pope PB, et al. Rumen and Cecum Microbiomes in Reindeer (Rangifer tarandus tarandus) Are Changed in Response to a Lichen Diet and May Affect Enteric Methane Emissions. PLoS One. 2016a;11:e0155213.
- 28. Liu W, Saint DA. Validation of a quantitative method for real time PCR kinetics. Biochem Biophys Res Commun. 2002;294:347–353. pmid:12051718
- 29. Gantner S, Andersson AF, Alonso-Sáez L, Bertilsson S. Novel primers for 16S rRNA-based archaeal community analyses in environmental samples. J Microbiol Methods. 2011;84:12–18. pmid:20940022
- 30. Lane D.J. 16S/23S rRNA Sequencing. In: Stackebrandt E. and Goodfellow M., Eds., Nucleic Acid Techniques in Bacterial Systematic, John Wiley and Sons, New York, 1991;115–175.
- 31. Pope PB, Mackenzie AK, Gregor I, Smith W, Sundset MA, McHardy AC, et al. Metagenomics of the Svalbard reindeer rumen microbiome reveals abundance of polysaccharide utilization loci. PLoS One. 2012;7:e38571. pmid:22701672
- 32. Hamady M, Walker JJ, Harris JK, Gold NJ, Knight R. Error-correcting barcoded primers for pyrosequencing hundreds of samples in multiplex. Nat Methods. 2008;5:235–237. pmid:18264105
- 33. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods. 2010;7:335–336. pmid:20383131
- 34. Lozupone C, Knight R. UniFrac: a new phylogenetic method for comparing microbial communities. Appl Environ Microbiol. 2005;71:8228–8235. pmid:16332807
- 35. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13:2498–2504. pmid:14597658
- 36. Tveit AT, Urich T, Svenning MM. Metatranscriptomic analysis of arctic peat soil microbiota. Appl Environ Microbiol. 2014;80:5761–5772. pmid:25015892
- 37. Masella AP, Bartram AK, Truszkowski JM, Brown DG, Neufeld JD. PANDAseq: paired-end assembler for illumina sequences. BMC Bioinformatics. 13:31. pmid:22333067
- 38. Schmieder R, Edwards R. Quality control and preprocessing of metagenomic datasets. Bioinformatics. 2011;27:863–864. pmid:21278185
- 39. Buchfink B, Xie C, Huson DH. Fast and sensitive protein alignment using DIAMOND. Nat Methods. 2015;12:59–60. pmid:25402007
- 40. Finn RD, Clements J, Arndt W, Miller BL, Wheeler TJ, Schreiber F, et al. HMMER web server: 2015 update. Nucleic Acids Res. 2015;43:W30–38. pmid:25943547
- 41. Huson DH, Beier S, Flade I, Górska A, El-Hadidi M, Mitra S, et al. MEGAN Community Edition—Interactive Exploration and Analysis of Large-Scale Microbiome Sequencing Data. PLoS Comput Biol. 2016;12:e1004957. pmid:27327495
- 42. Pruesse E, Quast C, Knittel K, Fuchs BM, Ludwig W, Peplies J, et al. SILVA: a comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB. Nucleic Acids Res. 2007;35:7188–7196. pmid:17947321
- 43. Lanzén A, Jørgensen SL, Huson DH, Gorfer M, Grindhaug SH, Jonassen I, et al. CREST—classification resources for environmental sequence tags. PLoS One. 2012;7:e49334. pmid:23145153
- 44. Tukey J. Comparing individual means in the analysis of variance. Biometrics. 1949;5:99–114. pmid:18151955
- 45. R Development Core Team (2008). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0.
- 46. Parks DH, Tyson GW, Hugenholtz P, Beiko RG. STAMP: statistical analysis of taxonomic and functional profiles. Bioinformatics. 2014;30:3123–3124. pmid:25061070
- 47. Kanehisa M, Sato Y, Kawashima M, Furumichi M, Tanabe M. KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res. 2016;44:D457–462. pmid:26476454
- 48. White JR, Nagarajan N, Pop M. Statistical methods for detecting differentially abundant features in clinical metagenomic samples. PLoS Comput Biol. 2009;5:e1000352. pmid:19360128
- 49. Denman SE, Tomkins NW, McSweeney CS. Quantitation and diversity analysis of ruminal methanogenic populations in response to the antimethanogenic compound bromochloromethane. FEMS Microbiol Ecol. 2007;62:313–322. pmid:17949432
- 50. Sundset MA, Edwards JE, Cheng YF, Senosiain RS, Fraile MN, Northwood KS, et al. Molecular diversity of the rumen microbiome of Norwegian reindeer on natural summer pasture. Microb Ecol. 2009;57:335–348. pmid:18604648
- 51. Mignon-Grasteau S, Narcy A, Rideau N, Chantry-Darmon C, Boscher MY, Sellier N, et al. Impact of Selection for Digestive Efficiency on Microbiota Composition in the Chicken. PLoS One. 2015;10:e0135488. pmid:26267269
- 52. Dewar ML, Arnould JP, Krause L, Dann P, Smith SC. Interspecific variations in the faecal microbiota of Procellariiform seabirds. FEMS Microbiol Ecol. 2014;89:47–55. pmid:24684257
- 53. Hook SE, Wright AD, McBride BW. Methanogens: methane producers of the rumen and mitigation strategies. Archaea. 2010; pmid:21253540
- 54. Akin DE, Borneman WS. Role of rumen fungi in fiber degradation. J Dairy Sci. 1990;73:3023–3032. pmid:2178175
- 55. Belanche A, Doreau M, Edwards JE, Moorby JM, Pinloche E, Newbold CJ. Shifts in the rumen microbiota due to the type of carbohydrate and level of protein ingested by dairy cattle are associated with changes in rumen fermentation. J Nutr. 142;1684–1692. pmid:22833657
- 56. Allison MJ, Mayberry WR, McSweeney CS, Stahl DA. Synergistes jonesii, gen. nov., sp.nov.: A Rumen Bacterium That Degrades Toxic Pyridinediols. Syst Appl Microbiol. 1992;15:522–529.
- 57. Hansen AH, Jonasson S, Michelsen A, Julkunen-Tiitto R. Long-term experimental warming, shading and nutrient addition affect the concentration of phenolic compounds in arctic-alpine deciduous and evergreen dwarf shrubs. Oecologia. 2006;147:1–11. pmid:16180043
- 58. Ilmberger N, Güllert S, Dannenberg J, Rabausch U, Torres J, Wemheuer B, et al. A comparative metagenome survey of the fecal microbiota of a breast- and a plant-fed Asian elephant reveals an unexpectedly high diversity of glycoside hydrolase family enzymes. PLoS One. 9:e106707. pmid:25208077
- 59. Joo S, Kim S, Seo H, Kim KJ. Crystal Structure of Amylomaltase from Corynebacterium glutamicum. J Agric Food Chem. 2016;64:5662–5670. pmid:27366969
- 60. Luo YH, Wright AD, Li YL, Li H, Yang QH, Luo LJ, et al. Diversity of methanogens in the hindgut of captive white rhinoceroses, Ceratotherium simum. BMC Microbiol. 2013;13:207. pmid:24228793
- 61. Söllinger A, Schwab C, Weinmaier T, Loy A, Tveit AT, Schleper C, et al. Phylogenetic and genomic analysis of Methanomassiliicoccales in wetlands and animal intestinal tracts reveals clade-specific habitat preferences. FEMS Microbiol Ecol. 2016;92:1.
- 62. Salgado-Flores A, Bockwoldt M, Hagen LH, Pope PB, Sundset MA. First insight into the faecal microbiota of the high Arctic muskoxen (Ovibos moschatus). Microbial Genomics. 2016b; pmid:28348861
- 63. St-Pierre B, Wright AD. Metagenomic analysis of methanogen populations in three full-scale mesophilic anaerobic manure digesters operated on dairy farms in Vermont, USA. Bioresour Technol. 2013;138:277–284. pmid:23624045
- 64. Dridi B, Fardeau ML, Ollivier B, Raoult D, Drancourt M. Methanomassiliicoccus luminyensis gen. nov., sp. nov., a methanogenic archaeon isolated from human faeces. Int J Syst Evol Microbiol. 2012;62:1902–1907. pmid:22859731
- 65. Tveit AT, Urich T, Frenzel P, Svenning MM. Metabolic and trophic interactions modulate methane production by Arctic peat microbiota in response to warming. Proc Natl Acad Sci USA. 2015;112:E2507–2516. pmid:25918393
- 66. Winter JU, Wolfe RS. Methane formation from fructose by syntrophic associations of Acetobacterium woodii and different strains of methanogens. Arch Microbiol. 1980;124:73–79. pmid:6769417
- 67. Zitko V, Bishop CT. Fractionation of pectin from sunflowers, sugar beets, apples, and citrus fruits. Can J Chemistry. 1985;43:3206–3214.
- 68. Toman R, Karacsonyi S, Kubackova M. Studies of the pectin present in the bark of white willow (Salix alba L): fractionation and acidic depolymerization of the water-soluble pectin. Carbohydrates research. 1975;43:111–116.
- 69. Gradel CM, Dehority BA. Fermentation of isolated pectin and pectin from intact forages by pure cultures of rumen bacteria. Appl Microbiol. 1972;23:332–340. pmid:4552890
- 70. Sirotek K, Marounek M, Rada V, Benda V. Isolation and characterization of rabbit caecal pectinolytic bacteria. Folia Microbiol (Praha). 2001;79–82.
- 71. Facey HV, Northwood KS, Wright AD. Molecular diversity of methanogens in fecal samples from captive Sumatran orangutans (Pongo abelii). Am J Primatol. 2012;74:408–413. pmid:22511523