Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/springer-journals/geography-and-elevation-as-drivers-of-cloacal-microbiome-assemblages-yaj8hfomaU
- Publisher
- Springer Journals
- Copyright
- Copyright © The Author(s) 2023
- eISSN
- 2524-4671
- DOI
- 10.1186/s42523-022-00219-3
- Publisher site
- See Article on Publisher Site
Abstract
Background Empirical field studies allow us to view how ecological and environmental processes shape the biodi- versity of our planet, but collecting samples in situ creates inherent challenges. The majority of empirical vertebrate gut microbiome research compares multiple host species against abiotic and biotic factors, increasing the potential for confounding environmental variables. To minimize these confounding factors, we focus on a single species of pas- serine bird found throughout the geologically complex island of Sulawesi, Indonesia. We assessed the effects of two environmental factors, geographic Areas of Endemism (AOEs) and elevation, as well as host sex on the gut microbiota assemblages of the Sulawesi Babbler, Pellorneum celebense, from three different mountains across the island. Using cloacal swabs, high-throughput-amplicon sequencing, and multiple statistical models, we identified the core microbi- ome and determined the signal of these three factors on microbial composition. Results The five most prevalent bacterial phyla within the gut microbiome of P. celebense were Proteobacteria (32.6%), Actinobacteria (25.2%), Firmicutes (22.1%), Bacteroidetes (8.7%), and Plantomycetes (2.6%). These results are similar to those identified in prior studies of passeriform microbiomes. Overall, microbiota diversity decreased as elevation increased, irrespective of sex or AOE. A single ASV of Clostridium was enriched in higher elevation samples, while lower elevation samples were enriched with the genera Perlucidibaca (Family Moraxellaceae), Lachnoclostridium (Fam- ily Lachnospiraceae), and an unidentified species in the Family Pseudonocardiaceae. Conclusions While the core microbiota families recovered here are consistent with other passerine studies, the decreases in diversity as elevation increases has only been seen in non-avian hosts. Additionally, the increased abundance of Clostridium at high elevations suggests a potential microbial response to lower oxygen levels. This study emphasizes the importance of incorporating multiple statistical models and abiotic factors such as elevation in empirical microbiome research, and is the first to describe an avian gut microbiome from the island of Sulawesi. Keywords Microbiota, Community structure, Elevational shifts, Avian gut microbiome, Sulawesi Babbler, Pellorneum celebense, Elevational gradient, Alpha diversity, Beta diversity *Correspondence: Rachael L. Joakim rjoakim@amnh.org Rauri C. K. Bowie bowie@berkeley.edu Full list of author information is available at the end of the article © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. Joakim et al. Animal Microbiome (2023) 5:4 Page 2 of 12 30], compositional microbiome differences were found Background between syntopic resident and migratory populations The complex relationship between microbial symbionts within a species (the barn swallow, Hirundo rustica [12]). and their hosts is a functionally important, medically rel- Whether these differences were due to migration stress evant, and often understudied component of global bio- or region-specific microbe uptake is difficult to deter - diversity. In every habitable natural system, there exists mine. An alternative approach would be to compare the communities of microscopic organisms, including those microbial composition within allopatric populations of residing in and on other organisms. Humans and other a single, widespread avian host species, occurring across vertebrates harbor communities of microbes in the gas- a repeated environmental gradient (e.g., elevation). To trointestinal (GI) tract known as the gut microbiota that resolve if region-specificity affects the microbial com - directly contribute to nutrient uptake, immune function, position of avian hosts in Southeast Asia, we sampled and fitness plasticity in stochastic environments [1–5]. one passerine species in three areas of endemism across Studies of vertebrate hosts report various ecological and a 1000 m elevational range on the island of Sulawesi, evolutionary factors driving gut microbiota commu- Indonesia. nity assemblage, including diet [6–8], sex [9], reproduc- Sulawesi is the eleventh largest island in the world tive behavior [10], habitat type [11], movement [12] and (Fig. 1) and has a complex geological history, with differ - host phylogeny [13–16]; although many of these factors ent landmasses accreting and breaking apart over the past are interrelated in terms of microbe-specific selection 30 million years [31, 32]. Recent studies have revealed pressures. However, as the majority of the vertebrate Sulawesi remained partially submerged until less than gut microbiota literature focuses heavily on mammalian 1 MYA [32]. As a result of these processes, seven differ - systems [17, 18], there is a significant knowledge gap in ent geographically-isolated Areas of Endemism (AOEs) terms of how these processes affect other major taxa. have been identified, supported by clear species bounda - Additional studies within non-mammalian hosts, par- ries within terrestrial vertebrates [33–35]. Sulawesi is also ticularly in birds, will allow for generalizable insights unusual for an island in that it harbors 25 high-elevation regarding host-gut micriobiota dynamics across verte- mountains (> 2000 m), which have likely contributed to its brates. Avian hosts provide an ideal study system as they unusually high percentage of endemic flora and fauna [31, are known vectors of human diseases (e.g. West Nile 33, 36, 37]; endemic species make up approximately 48% of Virus, High-pathogenic Bird Flu), demonstrate diverse all birds and 36% of all mammals described on the island ecologies and mating systems, and serve as model sys- [38]. The mountains on Sulawesi naturally encompass tems for comparative analyses in ecology and evolution- steep gradients in abiotic variables over short distances, ary biology [19]. making it possible to rapidly sample multiple individu- Songbirds (Order: Passeriformes) make up more als of a single species spanning a broad environmental than 50% of global bird diversity [20]. In general, the cline. The replicate sampling of high-elevation mountains gut microbiota of Passeriformes are dominated by the on Sulawesi provides an opportunity to investigate how bacterial phyla Firmicutes, Actinobacteria, Tenericutes, avian gut microbiome diversity is shaped by both geogra- Bacteroidetes, and Proteobacteria, with a particularly phy (allopatric isolation on different mountains) and steep high abundance of Firmicutes in bird species that feed clines in abiotic variables (elevational gradients). on insects [21–26]. Comparative field studies have sug - Evidence for a relationship between elevation and gut gested that host phylogeny is the most significant, albeit microbial diversity and abundance in mammals, squa- weakly associated [19], driver influencing the gut micro - mates, and birds has been mixed [21, 30, 39–42]. How- biome structure of birds. This relationship is independent ever, we predict a decrease in both diversity and relative of other factors such as diet or habitat [6, 19, 21, 26–28]. abundances of microbial taxa in individuals with increas- While placental mammals inoculate their offspring with ing elevation, corresponding to the lower diversity and a portion of their microbiota through live birth, avian abundance of prey, parasites, and environmental micro- young are exposed to microbes in nesting material and biota found at higher elevations [30, 40, 43, 44]. Further, receive parental microbiota via incubation and paren- although AOEs on Sulawesi represent distinct population tal saliva [12, 29], suggesting birds are more susceptible barriers for terrestrial vertebrates, this may not extend to to environmental variation in microbial source pools volant organisms and their gut microbiota. We therefore (e.g., due to geographic or ecological gradient effects) predicted that higher elevation would be negatively cor- than viviparous organisms [15, 19, 26, 28]. While evi- related with microbial diversity and relative abundances dence suggests only small differences in avian microbial of dominant taxa, and that elevation would be a more communities for intraspecific bird populations from significant factor in microbial composition than area of both temperate and tropical locations [15, 20, 21, 26, Joak im et al. Animal Microbiome (2023) 5:4 Page 3 of 12 Fig. 1 Map of collecting sites on the island of Sulawesi, Indonesia. Grey bars indicate the seven Areas of Endemism (AOEs) described in [32]. The axis values are in degrees endemism within intraspecific populations of endemic mountains, each within a different AOE, to determine Sulawesi passerines. if the diversity and relative abundances of gut micro- To determine the effect of AOE and elevation on avian bial assemblages correlate with both elevation and area gut microbiota assemblages within these montane sys- of endemism on the island of Sulawesi (Fig. 1). Host sex tems, we chose the Sulawesi babbler (Pellorneum cel- was also included in analyses to account for potential sex- ebense) as a focal taxon. This species is a common, dependent variation. We tested the hypotheses that H : endemic insectivorous passerine widespread throughout diversity and abundance of P. celebense gut microbiota is the island, where it occupies dense forest undergrowth negatively correlated with elevation; and H : areas of end- from sea-level to 1500 m in elevation, and was the most emism do not have significant effect on host intraspecific abundant species found [45–47]. We sampled P. celebense microbial variation. individuals along 1000 m elevational transects on three Joakim et al. Animal Microbiome (2023) 5:4 Page 4 of 12 Results (t = 0.81, p = 0.42), so only the results from unrarefied A total of 4339 ASVs from 40 Pellorneum celebense indi- data are reported. viduals were recovered for downstream analyses. The R package “decontam” did not find any reads from posi - P. celebense microbiome composition tive or negative controls that were found in all corre- Only one ASV, an unidentified member of the Entero - sponding samples, so no matching reads were removed bacteriaceae family (Phylum Proteobacteria), was found from the dataset. After transforming raw read counts in 90% of individuals. An additional ASV in the Entero- and removing low abundance ASVs, 3445 were retained coccaceae family (Phylum Firmicutes) was shared among in the transformed dataset, with 3028 ASVs recovered 90% of the Torompupu individuals. The five most prev - after removing individuals with < 3000 reads (n = 2) and alent bacterial phyla within the gut microbiome of P. rarefying the remaining 38 individuals to the lowest read celebense were Proteobacteria (32.6%), Actinobacteria count (n = 3389 reads, Table 1). A t-test comparing ASV (25.2%), Firmicutes (22.1%), Bacteroidetes (8.7%), and richness by sample between the transformed and rarefied Plantomycetes (2.6%). The relative abundances of these datasets did not reveal a difference in microbial richness phyla by individual are visualized in Fig. 2. Trends in P. celebense microbe diversity By Sex and Mountain To determine if host sex or AOE is directly correlated with changes in alpha diver- Table 1 Comparison between amplicon sequence variant (ASV ) sity, ANOVAs of individual Shannon indices using preprocessing of the original 4339 mountains and sex as explanatory variables were run. ASVs Samples Mean Results did not reveal differences in mean group diver - ASV per sample sity (F = 1.594, p = 0.217; F = 1.420, Mountain Mountain Sex p = 0.255, Fig. 3). These results did not change when Sex Transformed 3445 40 95.3 these ANOVAs were replicated using Chao1 as the Rarified 3028 38 86.1 diversity index (Fig. 3). However, PERMANOVAs com- Transformed reads were log transformed with all samples below 0.00001 paring Unifrac beta diversity indices indicated a slight removed, while rarified reads were randomly subset to 3028 reads per sample Fig. 2 Absolute frequencies in all samples (A) and the relative abundances of each individual (B) of the 5 most abundant phyla of ASVs from the P. celebense cloacal microbiome compositional dataset, identified using the 16S SILVA ribosomal RNA gene database Joak im et al. Animal Microbiome (2023) 5:4 Page 5 of 12 difference in microbial membership (unweighted Uni - of elevation as an environmental gradient also revealed a frac) among mountains (F = 1.216, r = 0.062, p = 0.062, clear negative trend between elevation and microbial rel- Fig. 4a), though this difference was no longer significant ative abundance (F = 1.230, p = 0.004, Fig. 5a). When this when relative abundance of ASVs (weighted Unifrac) analysis was repeated adding mountain as a conditional was included in the analysis (F = 1.283, r = 0.065, variable, these changes in abundance became less sig- w w p = 0.254, Fig. 4b). A PCoA plot of weighted Unifrac nificant, though plotting regression lines suggests a con - distances revealed a higher clustering of samples from tinued negative correlation with elevation (F = 1.202, Dako, suggesting that microbial communities of this p = 0.077, Fig. 5b). sample population may have more phylogenetic similar- ity than populations on the other mountains, as a non- Modeling interacting factors dimensional ordination did not show the same clustering To determine if mountain or host sex is producing a of Dako samples (Fig. 4c). random effect potentially confounding the influence of By Elevation While an ANOVA of individual Shan- elevation on microbial assemblages, three robust mixed- non diversity by elevational category did not show a sig- effect models were run with elevation as the predictor nificant decrease in mean diversity as elevation increases effect and mountain and sex as separate and interacting (F = 5.174, p = 0.072, Fig. 3c), a linear regression using random effects. Because robust models do not gener - numerical elevation data revealed a significant decrease ate p values, the significance of models are confirmed if −05 as elevation increases (t = 4.967, p = 1.56e , Fig. 3d). the slope and 95% CI intervals do not intercept 0 [48]. PERMANOVAs of beta diversity indices revealed a sig- All three models were significant under these conditions nificant difference by elevational category only in micro - and resulting predictor plots revealed identical negative bial composition (weighted Unifrac), not membership correlations between microbial diversity and elevation, alone (unweighted Unifrac) (F = 2.521, r = 0.120, regardless of which categorical variable was used as the w w p = 0.044; F = 1.180, r = 0.060, p = 0.071). A PER- random effect (Additional file 1: Fig. S2 and Table S2). MANOVA of CCA ordination residuals testing the effect Fig. 3 Microbial alpha diversity plots and associated p values using number of ASVs and Shannon Index compared by the variables: A sex (adults) and juveniles, B mountain, C elevational as a factor (low = < 700 m, mid = 700–1200 m, high = > 1200 m), and D continuous elevation Joakim et al. Animal Microbiome (2023) 5:4 Page 6 of 12 Fig. 4 Beta-diversity PCoA plots using unweighted Unifrac (A), weighted Unifrac (B), and NDMS (C) ordinations. Stress value for the NDMS was 0.190. Individuals are represented by a point shaped by sex and colored by elevation. Variations by mountain are represented by ordination ellipses Elevational effects on microbe taxonomic abundance Pellorneum celebense. In general, cloacal microbiota To determine which microbial families changed in abun- appeared to decrease in abundance and diversity as dance at each end of the elevational gradient, logfold val- elevation increases, regardless of mountain or host sex. ues were only compared between the “low” and “high” Surprisingly, there seems to be no influence of host elevation categories. The families Pseudonocardiaceae, sex on microbiome composition in this system. While Lachnospiraceae, and Moraxellaceae were significantly a PERMANOVA comparing microbial membership more abundant at lower elevations, whereas Clostridi- (unweighted Unifrac distances) did reveal a significant aceae was enriched at higher elevations (Fig. 6). influence of mountain, the most likely explanation is that this species was only found above 900 m on Mt. Dako. Discussion Therefore, the differences are most likely attributed to In this study, we determined the effects of eleva - a lack of sampling at lower elevations at this locality, as tion, mountain (representing independent AOEs), and this significance was lost when abundance-weighted sex on microbial assemblages in a Sulawesi songbird, Joak im et al. Animal Microbiome (2023) 5:4 Page 7 of 12 Fig. 5 Canonical Correspondence Analyses (CCAs) using only elevation as an environmental gradient (A) and with both elevation as a gradient and mountain as a conditional variable (B). Mountains are represented as separate regression lines, with shaded areas represent 95% confidence intervals. P values of each model indicates that the correlation between microbial diversity and elevation is less significant when sampling location is considered microbial composition (weighted Unifrac distances) was the decrease in diversity at higher elevations is associ- compared (Additional file 1: Fig. S1). ated with lower diversity of invertebrate prey species, a While previous avian gut microbiota studies did not lack of dietary analyses prevents a definitive conclusion. reveal correlations with elevation, decreases in micro- Additionally, selection pressures, especially those related bial alpha and beta diversities at higher elevations are to oxygen levels and air pressure, are likely increasing as also seen across populations of the toad-headed lizard, elevation increases. Expanding elevational analyses to Phrynocephalus vlangalii [39]. This decrease is thought include avian host species with a wide elevational range, to be influenced by hypoxic conditions resulting from especially those differing in feeding guilds, would further decreased oxygen partial pressure, though it is worth validate this negative correlation with microbial diversity noting the elevational range of P. vlangalii (2900–4250 m) at the community scale. is much higher than that of P. celebense. Conversely, a Observed patterns from specific microbial taxa found study of high-altitude pikas (Ochotona curzoniae) found in P. celebense cloacal samples also offer insights into the higher diversity and functional enrichment as eleva- underlying factors influencing the microbiome commu - tion increased [40], while studies of Tibetan ruminants nity. Clostridium sensu stricto 1 is an obligate anaerobic (Bos spp. and Ovis spp.) and mesquite lizards (Scelopo- fermenter metabolizing a range of compounds such as rus grammicus) did not find any relationship between carbohydrates, amino acids, alcohols, and purines [51]. diversity and elevation that was not consistent with The increased presence of Clostridium in higher eleva - subsequent dietary shifts [49, 50]. While we can assume tion hosts may suggest an increased dependency on Joakim et al. Animal Microbiome (2023) 5:4 Page 8 of 12 Fig. 6 Differential abundances between low and high elevational categories. ASV’s enriched in higher elevations fall to the left side of the x-axis, while those enriched in lower elevations fall to the right microbial symbionts for metabolism due to lower oxygen incorporating measured immune activity could discern levels, as microbes related to host metabolic pathways if these microbes are associated with avian immune were also seen in higher proportions in high-elevation function. S. grammicus lizard populations [50]. The genus Perlu - cidibaca has only been isolated from aquatic samples, Conclusions which does not provide a clear biological explanation for The results of this study illuminate the importance of its increased abundance in hosts at low elevations [52]. assessing abiotic and biotic factors in empirical gut A possible explanation is that these microbes are present microbiota research by documenting intraspecific vari - in sampling sites near bodies of standing water (which ation seen in wild host populations. Studies focusing on are typically seen only in lower elevations), though no host phylogeny and diet in wild systems may miss the environmental samples were collected in this study so potential confounding effects of environmental factors this could not be confirmed. Members of the family if they are not included in these analyses. In the absence Pseudonocardiaceae are known to produce antibacterial of elevational data, this dataset would reveal limited metabolites, especially in high nitrogen environments spatial or sex-dependent variation in microbial commu- [53, 54]. The genus Lachnoclostridium is associated with nities. Sulawesi montane ecosystems provide an ideal metabolism of similar metabolites as Clostridium, but has study system of an isolated, endemic avian community. been identified as an indicator of early stages of colorec - By incorporating an elevational gradient and testing for tal cancer and therefore may indicate localized immune interacting factors, we show that the cloacal microbiota activity in the cloaca [55, 56]. Both families were also membership, structure, and overall abundance in P. cele- significantly more abundant in hosts at lower elevation, bense populations significantly decreases in higher eleva - which may suggest increased immune activity. Because tions in all three different areas of endemism that were the majority of functional analyses on gut microbes are sampled. Additionally, the higher abundance of meta- human-based, however, these potential associations bolic microbe ASVs at high elevations suggests altitude- are merely speculative. Future gut microbiota studies related shifts in community structure. Future community Joak im et al. Animal Microbiome (2023) 5:4 Page 9 of 12 studies can confirm if these elevational shifts are consist - Table 2 Sampling site information for Pellornuem celebense cloacal swabs used in this study ent across host feeding guilds and phylogeny. This study lays the foundation for future work on montane host Mountain Dako Katopasa Torompupu communities on Sulawesi, and contributes to a growing Total n 9 15 16 global microbiome dataset by providing the first report of Summit Elevation (m) 2260 2835 2495 avian gut microbiota from this remarkably unique island. Area of Endemism (AOE) N Peninsula E Peninsula NW Central Core Sampling range (m) 924–1406 364–1340 660–1446 Methods Sampling date Jul-18 Aug-17 Nov-17 Study region Each of the three mountains surveyed shared general habitat gradients relative to elevation (Fig. 1). We cat- egorized elevations under 1000 m as lowland forest, This reagent was chosen over RNAlater as it is less likely dominated by large canopy trees with thick undergrowth. to degrade DNA when stored at warmer temperatures, Above 1000 m we observed mossy transitional forests, which is unavoidable during remote tropical fieldwork with understory dominated by rattan vines (subfamily [57, 58]. A tube of ethanol was used as a negative for Calamoideae). Summit ecosystems (> 1600 m) are cate- every new site to account for potential reagent contami- gorized as mossy forests with a sparse understory of Rho- nation. Individuals were sexed by dissection and visual dodendron spp., though no P. celebense individuals were inspection of gonads. If there were no developed ova found in ecosystems above 1450 m. or testes and the skull was not fully ossified, birds were defined as juvenile. Upon returning from the field, cloacal Gunung Torompupu NW central Core AOE. Summit swabs were stored in a – 80° freezer until processed. All is 2495 m. Located west of the Palu-Koro fault, which sampling protocols were approved by institutional ani- has been shown to be a species boundary for terres- mal care and use committees at the University of Califor- trial vertebrates [34]. The understory at all elevations nia, Berkeley (AUP-2016-04-8665-1) and the American was dominated by dense rattan. Museum of Natural History (AMNHIACUC-20171020). Gunung Katopasa Eastern peninsula AOE. Summit Research permits (Surat Izin Penelitian) for three expedi- is 2835 m. Sampling sites between 700 and 1300 m tions undertaken in 2017 and 2018 were obtained from included fire damage and clearings for small planta - Ministry of Research, Technology and Higher Education tions. (KEMENRISTEKDIKTI) (no. 213/SIP/FRP/E5/Dit.KI/ Gunung Dako Northern Peninsula AOE. Summit VIII/2017), with sample export documents for each expe- is 2260 m. Because the mountain is more accessible dition provided by the Research Center for Biology, Indo- than most, lower elevation habitat was fragmented nesian Institute of Sciences (LIPI). by small plantations with cleared understories. Given that this species specializes in dense forest under- Microbiome sequencing growth, the level of cultivation likely explains why DNA extractions were conducted using the MoBio we did not observe P. celebense below 900 m on this Power Soil kit (MoBio Laboratories Inc., Carlsbad, CA) mountain. with an extra wash step to maximize DNA recovery. DNA was PCR-amplified in triplicate using the Earth Microbi - ome Project’s 16S rRNA PCR protocol primers 515RB (5′-GTG YCA GCMGCC GCG GTAA-3′) and 806RB (5′- Sample collection GGA CTA CNVGGG TWT CTAAT-3′) and sequenced on Birds were sampled during three, month-long expedi- the Illumina MiSeq platform, generating 250 base pair tions during the dry seasons between 2017 and 2018 paired-end amplicon reads [59, 60]. PCRs and sequenc- (Fig. 1, Table 2) using mist-net transects spanning ele- ing, including of negative controls, were performed at the vational gradients at each mountain. Once captured in Argonne National Laboratory Sample Processing Facility mist-nets, individuals were promptly removed and placed (Lemont, IL, USA). in cloth holding bags while transported to camp for processing. To profile the avian gut microbiota, cloacal Data preprocessing swabs (Copan Diagnostics, Murrieta, CA) were collected Preprocessing of raw reads was conducted using pack- from live animals once at camp. First, 3% H O and a 2 2 ages implemented in QIIME2 v.2018.11 [60]. With the fresh KimWipe was used to sterilize the skin surround- DADA2 plugin, single-end reads were trimmed and chi- ing the cloaca. A sterile flocked nylon swab was gently meric sequences were removed before being paired and inserted, turned ½ rotation clockwise and ½ rotation classified as Amplicon Sequence Variants (ASVs) [61]. counterclockwise, and placed into a tube of 96% ethanol. Joakim et al. Animal Microbiome (2023) 5:4 Page 10 of 12 A multiple sequence alignment of paired-end reads richness, using elevation as the continuous fixed effect was generated using the ‘mafft’ program [62], and Fast - and combinations of sex and mountain as random Tree was used to create a midpoint-rooted phylogenetic effects. DESeq2, a logistical regression of dispersions tree [63]. Taxonomic identification of each ASV was weighted by normalized mean counts of unfiltered tax- generated using the Naïve Bayesian q-2 feature classi- onomic data [70], was used to assess which microbial fier trained on the 16S SILVA 132 ribosomal RNA gene taxa changed in abundance among mountain, host sex, database [64]. Downstream analyses were conducted in and elevational category. RStudio v.1.2.1335. The R package ‘decontam’ was used to assess potential contamination using field and laboratory Supplementary Information controls [65]. Reads that matched chloroplast and host The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s42523- 022- 00219-3. mitochondrial sequences were then removed. Due to the compositional nature of Illumina MiSeq data, [66] reads Additional file 1: Table S1. Specimen data for samples used in this study. were transformed to relative abundances and ASVs with Fig. S1. Histogram of total elevation values for each specimen, and total a relative abundance of < 0.00001 were removed for cer- values by mountain. The inconsistency in elevational sampling gradient at each mountain explains variation seen in mountain-based compari- tain downstream analyses using the R package ‘phyloseq’ sons. Fig. S2. PCoAs of Unweighted and Weighted unifrac distances by [67]. However, a rarefied dataset using the ‘phyloseq’ mountain. Fig. S3. Robust linear model residual effect plot with elevation command rarefy_even_depth was analyzed in parallel to as the predictor effect. Tic marks along the x-axis represent elevations of individual data points. All plots were identical, regardless of whether rule out potential effects of these preprocessing steps. mountain or host sex were set as conditional variables. Table S2. Slope and confidence intervals for robust mixed-effect models using elevation as an environmental gradient and mountain and sex as random effects. Statistical analyses All three models were not significant as the confidence intervals do not All statistical analyses and visualizations were per- intersect zero [48]. formed in R 4.1.2 (R Core Team 2020). The top 20 microbial phyla were identified, and the shared micro- Acknowledgements bial taxa of all samples were determined using the The authors would like to thank L. Bloch, Daniel, J. Childers, D. Wait, Pungki L, A. Riyanto, Apandi, Suparno, the Sulawesi Mountaineering club, and Mardin core_members function of ‘phyloseq’. Mountain and Sarkam, Safir, Mursalim for contributions to fieldwork, and Safir for keep - host sex were used as categorical variables, while ing us alive and fed on the summits; L. Smith and C. Claypool-Wang for lab elevation was analyzed as either a continuous or cat- assistance; S. Holmes and the instructors of the WHOI STAMPS workshop for assistance with statistical analyses; Kementerian Riset, Teknologi dan egorical variable: “low” for samples from elevations Pendidikan Tinggi (KEMENRISTEKDIKTI RI), Balai Konservasi Sumber Daya Alam under 700 m (n = 16), “mid” for 700–1200 m (n = 9), (BKSDA) Sulawesi Tengah, and Dinas Kehutanan Provinsi Sulawesi Tengah for and “high” for elevations over 1200 m (n = 15). Alpha providing research permits. diversity means were calculated using Shannon diver- Author contributions sity and Chao1 diversity indices using the estimate_ RLJ collected samples, performed laboratory assays, analyzed the data, and richness function of ‘phyloseq’ and compared using wrote the manuscript. MI, TH, YD, and SA collected samples and edited the manuscript. ASA and JAM led the field expeditions and edited the manuscript. ANOVAs (for categorical factors) with post-hoc Tukey SLP conceived the study, edited the manuscript, and contributed resources. tests and a general linear model (for elevation). Beta RCKB conceived the study, collected samples, edited the manuscript, and diversity indices were calculated as the sum of phylo- contributed resources. All authors read and approved the final manuscript. genetic branch lengths (Unifrac distances) to compare Funding variances in microbial membership (unweighted Uni- This research was funded by the National Science Foundation (DEB 1457845 frac) and composition (abundance-weighted Unifrac) awarded to JAM and RCKB and DEB 1457654/2039257 awarded to SP). among categorical factors using PERMANOVAS with Availability of data and materials the adonis function [68]. We also included an NDMS All data files and reproduceable code can be found at https:// github. com/ ordination based on Bray–Curtis distances. To evalu- rjoak im/ babbl er_ micro biome ate elevation as a continuous environmental gradient, Canonical Correspondence Analyses (CCAs) were Declarations run [69], using categorical factors as conditional vari- Ethics approval and consent to participate ables. To determine directional correlations between Capture, handling, and sampling protocols were approved under permit num- microbial diversity and elevation, robust mixed-effect bers AMNHIACUC-20171020 and AUP-2016–04-8665–1, in accordance with Institutional Animal Care and Use Committees at the University of California, models were created and evaluated using the R pack- Berkeley and the American Museum of Natural History, respectively. age “robustlmm,” as the robust model is more appro- priate than other mixed models for datasets with small Consent for publication All authors listed on the manuscript have consented to the submission of this sample sizes and high degrees of freedom [48]. We manuscript for publication in Animal Microbiome, and each participated in its used Shannon diversity index as the response variable development as stated in Author Contributions. in this model to detect shifts in ASV abundance and Joak im et al. Animal Microbiome (2023) 5:4 Page 11 of 12 Competing interests in passerines is not explained by ecological divergence. Mol Ecol. The authors declare that they have no competing interests. 2017;26:5292–304. https:// doi. org/ 10. 1111/ mec. 14144. 16. Ingala MR, Becker DJ, Bak Holm J, Kristiansen K, Simmons NB. Habitat Author details fragmentation is associated with dietary shifts and microbiota vari- Department of Biology, The City College of New York, 160 Convent Avenue, ability in common vampire bats. Ecol Evol. 2019;9:6508–23. New York, NY 10031, USA. The Graduate Center of The City University of New 17. Amato KR. Co-evolution in context: the importance of studying gut York, Biology Program, 365 5Th Ave, New York, NY 10016, USA. Sack ler microbiomes in wild animals. Microbiome Sci Med. 2013;1:10–29. Institute for Comparative Genomics, American Museum of Natural History, 18. Zmora N, Suez J, Elinav E. You are what you eat: diet, health and the New York, NY 10024, USA. The Richard Gilder Graduate School, American gut microbiota. Nat Rev Gastroenterol Hepatol. 2019;16:35–56. https:// Museum of Natural History, New York, NY 10024, USA. Museum Zoologicum doi. org/ 10. 1038/ s41575- 018- 0061-2. Bogoriense, Research Centre for Biology, National Research and Innovation 19. Bodawatta KH, Hird SM, Grond K, Poulsen M, Jønsson KA. Avian gut Agency, Jl. Raya Jakarta - Bogor Km 46, Cibinong 16911, Indonesia. Sciences microbiomes taking flight. Trends Microbiol. 2021. https:// doi. org/ 10. Department, Museums Victoria, Carlton, VIC, Australia. BioSciences Depar t-1016/j. tim. 2021. 07. 003. ment, University of Melbourne, Parkville, VIC, Australia. Museum of Ver tebrate 20. Bodawatta KH, Sam K, Jønsson KA, Poulsen M. Comparative analyses of Zoology and Department of Integrative Biology, University of California, the digestive tract microbiota of New Guinean Passerine birds. Front Berkeley, CA 94720, USA. Microbiol. 2018;9:1–13. 21. Hird SM, Sánchez C, Carstens BC, Brumfield RT. Comparative gut micro - Received: 4 February 2022 Accepted: 13 December 2022 biota of 59 neotropical bird species. Front Microbiol. 2015. https:// doi. org/ 10. 3389/ fmicb. 2015. 01403. 22. Kreisinger J, Čížková D, Kropáčková L, Albrecht T. Cloacal microbiome structure in a long-distance migratory bird assessed using deep 16sRNA pyrosequencing. 2015. https:// doi. org/ 10. 1371/ journ al. pone. 01374 01. References 23. Lewis WB, Moore FR, Wang S. Characterization of the gut microbiota of 1. Spor A, Koren O, Ley R. Unravelling the effects of the environment and migratory passerines during stopover along the northern coast of the host genotype on the gut microbiome. Nat Rev Microbiol. 2011;9:279–90. Gulf of Mexico. J Avian Biol. 2016;47:659–68. https:// doi. org/ 10. 1038/ nrmic ro2540. 24. Kropáčková L, Pechmanová H, Vinkler M, Svobodová J, Velová H, 2. Russell JA, Dubilier N, Rudgers JA. Nature’s microbiome: introduction. Mol Těšičký M, et al. Variation between the oral and faecal microbiota in a Ecol. 2014;23:1225–37. https:// doi. org/ 10. 1111/ mec. 12676. free-living passerine bird, the great tit (Parus major). PLoS ONE. 2017. 3. Carding S, Verbeke K, Vipond DT, Corfe BM, Owen LJ. Dysbiosis of the gut https:// doi. org/ 10. 1371/ journ al. pone. 01799 45. microbiota in disease. Microb Ecol Health Dis. 2015. https:// doi. org/ 10. 25. Grond K, Sandercock BK, Jumpponen A, Zeglin LH. The avian gut 3402/ mehd. v26. 26191. microbiota: community, physiology and function in wild birds. J Avian 4. Koskella B, Hall LJ, Metcalf CJE. The microbiome beyond the horizon Biol. 2018. https:// doi. org/ 10. 1111/ jav. 01788. of ecological and evolutionary theory. Nat Ecol Evol. 2017;1:1606–15. 26. Capunitan DC, Johnson O, Terrill RS, Hird SM. Evolutionary signal in the https:// doi. org/ 10. 1038/ s41559- 017- 0340-2. gut microbiomes of 74 bird species from Equatorial Guinea. Mol Ecol. 5. Suzuki TA, Ley RE. The role of the microbiota in human genetic adapta- 2020;29:829–47. https:// doi. org/ 10. 1111/ mec. 15354. tion. Science. 2020. https:// doi. org/ 10. 1126/ scien ce. aaz68 27. 27. García-Amado MA, Shin H, Sanz V, Lentino M, Martínez LM, Contreras 6. Phillips CD, Phelan G, Dowd SE, McDonough MM, Ferguson AW, Delton M, et al. Comparison of gizzard and intestinal microbiota of wild neo- Hanson J, et al. Microbiome analysis among bats describes influences tropical birds. PLoS ONE. 2018;13:1–16. of host phylogeny, life history, physiology and geography. Mol Ecol. 28. Videvall E, Strandh M, Engelbrecht A, Cloete S, Cornwallis CK. Measur- 2012;21:2617–27. ing the gut microbiome in birds: comparison of faecal and cloacal 7. Delsuc F, Metcalf JL, Wegener Parfrey L, Song SJ, González A, Knight R. sampling. Mol Ecol Resour. 2018;18:424–34. https:// doi. org/ 10. 1111/ Convergence of gut microbiomes in myrmecophagous mammals. Mol 1755- 0998. 12744. Ecol. 2014;23:1301–17. 29. van Veelen HPJ, Falcao Salles J, Tieleman BI. Multi-level comparisons of 8. Hale VL, Tan CL, Niu K, Yang Y, Knight R, Zhang Q, et al. Diet versus phylog- cloacal, skin, feather and nest-associated microbiota suggest consider- eny: a comparison of gut microbiota in captive colobine monkey species. able influence of horizontal acquisition on the microbiota assembly of 2018;75:515–27. https:// doi. org/ 10. 1007/ s00248- 017- 1041-8 sympatric woodlarks and skylarks. Microbiome. 2017;5:156. 9. Roggenbuck M, Bærholm Schnell I, Blom N, Bælum J, Bertelsen MF, Pon- 30. Woodhams DC, Bletz MC, Becker CG, Bender HA, Buitrago-Rosas tén TS, et al. The microbiome of new world vultures. Nat Commun. 2014. D, Diebboll H, et al. Host-associated microbiomes are predicted by https:// doi. org/ 10. 1038/ ncomm s6498. immune system complexity and climate. Genome Biol. 2020;21:1–20. 10. Jacob S, Parthuisot N, Vallat A, Ramon-Portugal F, Helfenstein F, Heeb 31. Hall R. Southeast Asia’s changing palaeogeography. Blumea - Biodi- P. Microbiome affects egg carotenoid investment, nestling develop - versity, Evol Biogeogr Plants. 2009;54:148–61. https:// doi. org/ 10. 3767/ ment and adult oxidative costs of reproduction in Great tits. Funct Ecol. 00065 1909X 475941. 2015;29:1048–58. https:// doi. org/ 10. 1111/ 1365- 2435. 12404. 32. Nugraha AMS, Hall R. Late Cenozoic palaeogeography of Sulawesi, 11. Sullam KE, Rubin BER, Dalton CM, Kilham SS, Flecker AS, Russell JA. Indonesia. Palaeogeogr Palaeoclimatol Palaeoecol. 2018;490:191–209. Divergence across diet, time and populations rules out parallel evolution 33. Evans BJ, Andayani N, Melnick DJ, Supriatna J, Setiadi MI, Cannatella in the gut microbiomes of Trinidadian guppies. ISME J. 2015;9:1508–22. DC. Monkeys and toads define areas of endemism on Sulawesi. Evolu- https:// doi. org/ 10. 1038/ ismej. 2014. 231. tion. 2006;57:1436. 12. Turjeman S, Corl A, Wolfenden A, Tsalyuk M, Lublin A, Choi O, et al. 34. Shekelle M, Meier R, Wahyu I, Ting N, Meier R, Wahyu I, et al. Molecu- Migration, pathogens and the avian microbiome: a comparative study in lar phylogenetics and chronometrics of tarsiidae based on 12S sympatric migrants and residents. Mol Ecol. 2020;29:4706–20. https:// doi. mtDNA haplotypes: evidence for miocene origins of crown tarsiers org/ 10. 1111/ mec. 15660. and numerous species within the Sulawesian Clade. Int J Primatol. 13. Sanders JG, Powell S, Kronauer DJC, Vasconcelos HL, Frederickson ME, 2010;31:1083–106. Pierce NE. Stability and phylogenetic correlation in gut microbiota: les- 35. Allen R, Damayanti CS, Frantz LAF, Leus K, Hulme-Beaman A, Gillemot S, sons from ants and apes. Mol Ecol. 2014;23:1268–83. et al. Synchronous diversification of Sulawesi’s iconic artiodactyls driven 14. Carrillo-Araujo M, Tas N, Alcántara-Hernández RJ, Gaona O, Schondube by recent geological events. Proc R Soc B Biol Sci. 2018;285:20172566. JE, Medellín RA, et al. Phyllostomid bat microbiome composition is 36. Moss SJ, Wilson MEJ. Biogeographic implications of the Tertiary palaeo- associated to host phylogeny and feeding strategies. Front Microbiol. geographic evolution of Sulawesi and Borneo. Biogeogr Geol Evol SE 2015;6:1–9. Asia. 1987;133:63. 15. Kropáčková L, Těšický M, Albrecht T, Kubovčiak J, Čížková D, Tomášek O, et al. Codiversification of gastrointestinal microbiota and phylogeny Joakim et al. Animal Microbiome (2023) 5:4 Page 12 of 12 37. Esselstyn JA, Achmadi AS, Handika H, Giarla TC, Rowe KC. A new climbing 57. Dominianni C, Wu J, Hayes RB, Ahn J. Comparison of methods for fecal shrew from Sulawesi highlights the tangled taxonomy of an endemic microbiome biospecimen collection. 2014.https:// doi. org/ 10. 1186/ radiation. J Mammal. 2019;100:1713–25. https:// doi. org/ 10. 1093/ jmamm 1471- 2180- 14- 103 al/ gyz077. 58. Bodawatta KH, Puzejova K, Sam K, Poulsen M, Jønsson KA. Cloacal swabs 38. Wikramanayake E, Dinerstein E, Loucks CJ, Olson DM, Morrison J, Lam- and alcohol bird specimens are good proxies for compositional analyses oreaux J, et al. Terrestrial ecoregions of the Indo-Pacific: a conservation of gut microbial communities of Great tits (Parus major). Anim Microbi- assessment. Electr Green J. 2002. https:// doi. org/ 10. 5860/ choice. 40- 0287. ome. 2020. https:// doi. org/ 10. 1186/ s42523- 020- 00026-8. 39. Zhang W, Li N, Tang X, Liu N, Zhao W. Changes in intestinal microbiota 59. Caporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Huntley J, Fierer N, across an altitudinal gradient in the lizard Phrynocephalus vlangalii. Ecol et al. Ultra-high-throughput microbial community analysis on the Illu- Evol. 2018;8:4695–703. https:// doi. org/ 10. 1002/ ece3. 4029. mina HiSeq and MiSeq platforms. ISME J. 2012;6:1621–4. https:// doi. org/ 40. Li H, Zhou R, Zhu J, Huang X, Qu J. Environmental filtering increases 10. 1038/ ismej. 2012.8. with elevation for the assembly of gut microbiota in wild pikas. Microb 60. Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, Biotechnol. 2019;12:976–92. et al. Author correction: reproducible, interactive, scalable and extensible 41. Bodawatta KH, Koane B, Maiah G, Sam K, Poulsen M, Jønsson KA. Species- microbiome data science using QIIME 2 (Nat Biotechnol 2019;37(8):(852– specific but not phylosymbiotic gut microbiomes of New Guinean pas- 857). https:// doi. org/ 10. 1038/ s41587- 019- 0209-9). Nat Biotechnol. serine birds are shaped by diet and flight-associated gut modifications. 2019;37:1091. https:// doi. org/ 10. 1038/ s41587- 019- 0252-6. Proc R Soc B Biol Sci. 2021. https:// doi. org/ 10. 1098/ rspb. 2021. 0446. 61. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. 42. Herder EA, Spence AR, Tingley MW, Hird SM. Elevation correlates with sig- DADA2: High-resolution sample inference from Illumina amplicon data. nificant changes in relative abundance in hummingbird fecal microbiota, Nat Methods. 2016;13:581–3. but composition changes little. Front Ecol Evol. 2021;8:534. 62. Katoh K. MAFFT: a novel method for rapid multiple sequence alignment 43. Foster JT, Woodworth BL, Eggert LE, Hart PJ, Palmer D, Duffy DC, et al. based on fast Fourier transform. Nucleic Acids Res. 2002;30:3059–66. Genetic structure and evolved malaria resistance in Hawaiian honey- 63. Price MN, Dehal PS, Arkin AP. FastTree 2 – approximately maximum- creepers. Mol Ecol. 2007;16:4738–46. https:// doi. org/ 10. 1111/j. 1365- 294X. likelihood trees for large alignments. PLoS ONE. 2010;5:e9490. https:// doi. 2007. 03550.x.org/ 10. 1371/ journ al. pone. 00094 90. 44. Galen SC, Witt CC. Diverse avian malaria and other haemosporidian 64. Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA parasites in Andean house wrens: evidence for regional co-diversification ribosomal RNA gene database project: improved data processing and by host-switching. J Avian Biol. 2014;45:374–86. web-based tools. Nucleic Acids Res. 2013. https:// doi. org/ 10. 1093/ nar/ 45. Moyle RG, Oliveros CH, Andersen MJ, Hosner PA, Benz BW, Manthey gks12 19. JD, et al. ARTICLE tectonic collision and uplift of Wallacea triggered the 65. Davis NM, Proctor DM, Holmes SP, Relman DA, Callahan BJ. Simple statisti- global songbird radiation. 2016. https:// doi. org/ 10. 1038/ ncomm s12709 cal identification and removal of contaminant sequences in marker-gene 46. Gelang M, Cibois A, Pasquet E, Olsson U, Alström P, Ericson PGP. Phylog- and metagenomics data. Microbiome. 2018. https:// doi. org/ 10. 1186/ eny of babblers (Aves, Passeriformes): major lineages, family limits and s40168- 018- 0605-2. classification. Zool Scr. 2009;38:225–36. https:// doi. org/ 10. 1111/j. 1463- 66. McMurdie PJ, Holmes S. Waste Not, Want not: why rarefying microbiome 6409. 2008. 00374.x. data is inadmissible. PLoS Comput Biol. 2014;10. 47. Cai T, Cibois A, Alström P, Moyle RG, Kennedy JD, Shao S, et al. Near-com- 67. McMurdie PJ, Holmes S. Phyloseq: an R package for reproducible interac- plete phylogeny and taxonomic revision of the world’s babblers (Aves: tive analysis and graphics of microbiome census data. PLoS ONE. 2013. Passeriformes). Mol Phylogenet Evol. 2019;2019(130):346–56. https:// doi. https:// doi. org/ 10. 1371/ journ al. pone. 00612 17. org/ 10. 1016/j. ympev. 2018. 10. 010. 68. Lozupone C, Knight R. UniFrac: a new phylogenetic method for compar- 48. Koller M. Robustlmm: an R package for Robust estimation of linear Mixed- ing microbial communities. Appl Environ Microbiol. 2005;71:8228–35. Eec ff ts models. J Stat Softw. 2016. https:// doi. org/ 10. 18637/ jss. v075. i06.https:// doi. org/ 10. 1128/ AEM. 71. 12. 8228- 8235. 2005. 49. Zhang Z, Xu D, Wang L, Hao J, Wang J, Zhou X, et al. Convergent 69. Paliy O, Shankar V. Application of multivariate statistical techniques in evolution of rumen microbiomes in high-altitude mammals. Curr Biol. microbial ecology. Mol Ecol. 2016;25:1032–57. 2016;26:1873–9. 70. Love MI, Huber W, Anders S. Moderated estimation of fold change and 50. Montoya-Ciriaco N, Gómez-Acata S, Muñoz-Arenas LC, Dendooven L, dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:550. Estrada-Torres A, Aníbal H, de la Vega-Pérez D, Navarro-Noya YE. Dietary https:// doi. org/ 10. 1186/ s13059- 014- 0550-8. effects on gut microbiota of the mesquite lizard Sceloporus grammicus ( Wiegmann, 1828) across different altitudes. Microbiome. 2020. https:// Publisher’s Note doi. org/ 10. 1186/ s40168- 020- 0783-6. Springer Nature remains neutral with regard to jurisdictional claims in pub- 51. Alou M, Ndongo S, Frégère L, Labas N, Andrieu C, Richez M, et al. lished maps and institutional affiliations. Taxonogenomic description of four new clostridium species isolated from human gut: `Clostridium amazonitimonense’, `Clostridium merdae’, `Clostridium massilidielmoense’ and `Clostridium nigeriense’. N Microbes N Infect. 2018;21:128–39. https:// doi. org/ 10. 1016/j. nmni. 2017. 11. 003. 52. Song J, Choo YJ, Cho JC. Perlucidibaca piscinae gen. nov., sp. Nov., a freshwater bacterium belonging to the family Moraxellaceae. Int J Syst Evol Microbiol. 2008;58:97–102. https:// doi. org/ 10. 1099/ ijs.0. 65039-0. 53. Platas G, Morón R, González I, Collado J, Genilloud O, Peláez F, et al. Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : Production of antibacterial activities by members of the family pseu- donocardiaceae: influence of nutrients. World J Microbiol Biotechnol. fast, convenient online submission 1998;14:521–7. https:// doi. org/ 10. 1023/A: 10088 74203 344. thorough peer review by experienced researchers in your field 54. Thoemmes MS, Cove MV. Comparing the microbial communities of natural and supplemental nests of an endangered ecosystem engineer. rapid publication on acceptance bioRxiv. 2019. https:// doi. org/ 10. 1101/ 727966. support for research data, including large and complex data types 55. Uchimura Y, Wyss M, Brugiroux S, Limenitakis JP, Stecher B, McCoy KD, • gold Open Access which fosters wider collaboration and increased citations et al. Complete genome sequences of 12 species of stable defined mod- erately diverse mouse microbiota 2. Genome Announc. 2016;4:951–67. maximum visibility for your research: over 100M website views per year https:// doi. org/ 10. 1128/ genom eA. 00951- 16. 56. Liang JQ, Li T, Nakatsu G, Chen YX, Yau TO, Chu E, et al. A novel faecal At BMC, research is always in progress. Lachnoclostridium marker for the non-invasive diagnosis of colorectal Learn more biomedcentral.com/submissions adenoma and cancer. Gut. 2019;69:1248–57. https:// doi. org/ 10. 1136/ gutjnl- 2019- 318532.
Journal
Animal Microbiome – Springer Journals
Published: Jan 16, 2023
Keywords: Microbiota; Community structure; Elevational shifts; Avian gut microbiome; Sulawesi Babbler; Pellorneum celebense; Elevational gradient; Alpha diversity; Beta diversity