https://orcid.org/0000-0002-6462-9313
Figures
Abstract
Adaptive radiations offer an excellent opportunity to understand the eco-evolutionary dynamics of gut microbiota and host niche specialization. In a laboratory common garden, we compared the gut microbiota of two novel derived trophic specialist pupfishes, a scale-eater and a molluscivore, to closely related and distant outgroup generalist populations, spanning both rapid trophic evolution within 10 kya and stable generalist diets persisting over 11 Mya. We predicted an adaptive and highly divergent microbiome composition in the trophic specialists reflecting their rapid rates of craniofacial and behavioral diversification. We sequenced 16S rRNA amplicons of gut microbiomes from lab-reared adult pupfishes raised under identical conditions and fed the same high protein diet. In contrast to our predictions, gut microbiota largely reflected phylogenetic distance among species, rather than generalist or specialist life history, in support of phylosymbiosis. However, we did find significant enrichment of Burkholderiaceae bacteria in replicated lab-reared scale-eater populations. These bacteria sometimes digest collagen, the major component of fish scales, supporting an adaptive shift. We also found some enrichment of Rhodobacteraceae and Planctomycetia in lab-reared molluscivore populations, but these bacteria target cellulose. Overall phylogenetic conservation of microbiome composition contrasts with predictions of adaptive radiation theory and observations of rapid diversification in all other trophic traits in these hosts, including craniofacial morphology, foraging behavior, aggression, and gene expression, suggesting that the functional role of these minor shifts in microbiota will be important for understanding the role of the microbiome in trophic diversification.
Citation: Heras J, Martin CH (2022) Minimal overall divergence of the gut microbiome in an adaptive radiation of Cyprinodon pupfishes despite potential adaptive enrichment for scale-eating. PLoS ONE 17(9): e0273177. https://doi.org/10.1371/journal.pone.0273177
Editor: Axel Meyer, University of Konstanz, GERMANY
Received: January 20, 2022; Accepted: August 3, 2022; Published: September 16, 2022
Copyright: © 2022 Heras, Martin. 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: Sequence data is available through the NCBI Sequence Read Archive database under the BioProject accession PRJNA867380, which contains BioSamples: SAMN30196614 –SAMN30196619. Raw sequences and dataset files used for our analyses are available in the Dryad repository (doi:10.5061/dryad.5hqbzkh8n).
Funding: This research was funded by the National Science Foundation DEB CAREER grant #1749764, National Institutes of Health grant 5R01DE027052‐02, and the University of California, Berkeley to CHM. Lastly, additional support came from California State University, San Bernardino (startup funds to JH). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Rapid evolutionary change can alter ecological processes which in turn change the course of evolutionary processes [1, 2]. This process is described as eco-evolutionary dynamics, which provides a framework for understanding the interplay between evolution and ecological interactions [3, 4]. The emergence of studies that focus on eco-evolutionary dynamics has provided insight for community assembly, ecological speciation, and adaptive radiations [3]. A better understanding of these eco-evolutionary dynamics can be applied to host-microbiota interactions, in which microbes and host relationships impact host performance and fitness [5–7]. The microbial community may also play a large role in ontogeny, immunity, physiology, and the ecology and evolution of the host [8–11].
Several studies have now examined gut microbiome diversification in an adaptive radiation of hosts, including fishes [6, 10, 12–14]. Similarity among host species microbiomes that recapitulates the evolutionary history of the host species is known as phylosymbiosis and is frequently the primary hypothesis in these studies [15, 16]. However, these studies rarely examine outgroups to the focal radiation in order to compare rates of microbiome divergence. Furthermore, phylosymbiosis [comparable to phylogenetic conservatism; 17] contrasts with the expectations of a rapid burst of phenotypic diversification during adaptive radiation, which would suggest that the microbiome within an adaptive radiation should diverge far more quickly than outgroup taxa due to rapid ecological divergence and specialization [18–22]. Thus, we predicted greater microbiome divergence within a recent adaptive radiation of trophic specialists than among outgroup generalist taxa.
An adaptive radiation of Cyprinodon pupfishes provides an excellent opportunity to test the relative roles of rapid trophic divergence and phylosymbiosis in shaping the gut microbiome. Pupfishes are found in hypersaline lakes and coastal areas throughout the Caribbean and Atlantic (most are allopatric) and within isolated desert pools and streams [23–25]. However, there are only two sympatric adaptive radiations of trophic specialists across this range [26]. One radiation is endemic to San Salvador Island, Bahamas, containing a generalist algivorous and detritivorous species, Cyprinodon variegatus, and two trophic specialist species, a molluscivore (durophage) C. brontotheroides and a scale-eater (lepidophage) C. desquamator [26–28; there is also a fourth intermediate scale-eating species not including in this study: 28]. Scale-eating and molluscivore niches are uniquely derived within this sympatric radiation on San Salvador Island relative to a generalist or omnivore diet of macroalgae and micro-invertebrates in all other Cyprinodon species spread across the Caribbean and desert interior of North America [23, 29], including the most closely related extant genus Cualac [26]. The adaptive radiation of Cyprinodon pupfishes on San Salvador Island is estimated to be around 10,000 years old based on the age of the hypersaline lakes on the island which filled with rising sea levels following the last glacial maximum [30–32]. In contrast, the most divergent generalist population in our study, the checkered pupfish Cualac tessellatus, occurs only in the El Potosí desert spring system in Mexico and last shared a common ancestor with Cyprinodon 11.2 Mya [33]. The two trophic specialist species on San Salvador Island are derived from a generalist ancestor and each shows signatures of adaptive introgression and the reassembly of standing genetic variation in generalist populations from across the Caribbean [24, 34]. Thus, this radiation provides an excellent opportunity to compare microbiome divergence within a sympatric adaptive radiation of trophic specialists nested within a large clade of generalist/omnivorous which have not substantially shifted their dietary niches over millions of years.
Despite extensive craniofacial and behavioral divergence between trophic specialists endemic to San Salvador Island there are very few fixed genetic differences between these species: there are only 157 fixed SNPs between molluscivores and scale-eaters out of 10 million segregating SNPs and only 87 deletions fixed in scale-eaters relative to molluscivores [34]. These are likely overestimates of fixed differences due to our smaller sample sizes of each species (approximately 30 per species). However, these fixed genetic differences may be driving differences in gut microbiome composition. Intriguingly, the only fixed coding indel detected so far in this system is a fixed deletion in all scale-eater populations of the fifth exon of the gene gpa33 [34]. This is an oncogene expressed in the intestinal epithelium and mice knockouts display a range of inflammatory intestinal pathologies [35], suggesting it may play a role in the gut microbiota composition of scale-eaters. Overall genetic differentiation among species is minimal, even within the same lake, ranging from Fst = 0.1–0.3, suggesting that soft sweeps and allele frequency changes among ecotypes may also play a larger role in their adaptation to different diets.
We raised all species in our study in a common laboratory environment for at least one generation and fed them an identical commercial pellet diet for one month before sampling gut microbiomes. All pupfishes were fed pellet food throughout their lives after feeding exclusively on newly hatched brine shrimp for approximately the first month after hatching. Importantly, all species and outgroups were treated in exactly the same way during lab-rearing. We addressed the following questions: 1) Do gut microbial communities primarily reflect dietary specialization or phylogenetic distance among species? 2) Is there a microbiome signal associated with lepidophagy (scale-eating) or molluscivory? We found enrichment only for the microbial family Burkholderiaceae in our two independent scale-eating pupfish colonies. This is significant because members of this microbial family digest collagen, the major component of fish scales. Overall, we infer a minor but potentially adaptive shift in the scale-eater microbiome, even when rearing hosts in identical environments on identical non-scale diets.
Materials and methods
Sampling and preparation of gut microbiome samples
Colonies of Cyprinodon pupfishes were collected from two hypersaline lakes on San Salvador Island, Bahamas (Crescent Pond and Osprey Lake) and Lake Cunningham, Bahamas in March, 2018 and were reared for two generations in aquaria at the University of North Carolina at Chapel Hill and the University of California, Berkeley. One generalist population was collected in May, 2018 from Fort Fisher Estuary in North Carolina and wild fish raised in the lab for one year were used for this study because no lab-reared fish were yet available at sufficient size. All fish were collected and exported with research permits from the Bahamas Environmental Science and Technology (BEST) commission or the U.S. Fish and Wildlife Service. Cualac tessellatus eggs were provided by the Zoological Society of London and reared in the lab for two generations before used for the four samples in this study. Exact generation times are unknown for this captive colony, but likely exceeded ten generations in captivity. All samples, except for the recently collected NC population, came from first or second-generation captive-bred individuals reared in aquaria (40–80 L) at 5–10 ppt salinity (Instant Ocean synthetic sea salt) and between 23 to 30°C. Colonies were always kept isolated by species and location. Individuals used for this study were fed once daily ad libitum with a single commercial pellet food (New Life Spectrum Cichlid Formula, New Life International, Inc., Homestead, FL), containing 34% crude protein, 5% crude fat, and 5% crude fiber, for one month without exposure to any other food or tankmates. Before this period, individuals were reared on pellet food throughout their lives following approximately one month feeding exclusively on newly hatched baby brine shrimp (Artemia spp.) after hatching. All animal care and experiments were conducted under approved protocols and guidelines of the University of California, Berkeley Institutional Animal Care and Use Committee (AUP-2018-08-11373).
In total, forty fishes were euthanized in an overdose of MS-222 and the entire intestinal tract was immediately excised (Cyprinodontidae do not possess stomachs; [36]) for DNA extraction. Standard length (the distance from the tip of the snout on the fish to the base of the caudal fin), and gut length were measured for all samples (S1 Table), and analyzed with an ANCOVA. Five individuals (F2 generation) from each of three species (C. variegatus, C. brontotheroides, and C. desquamator) in two replicate lake populations from San Salvador Island were sampled (n = 30 total), Crescent Pond and Osprey Lake. Conspecific populations show minimal genetic differentiation (Fst = 0.1–0.3) between these lakes [29, 37], providing independent replicates of the ecotypes across two hypersaline lake environments. In addition, we included the following generalist pupfish species as outgroups to our study: C. laciniatus (F1 generation; Lake Cunningham, New Providence Island, Bahamas; n = 4), C. variegatus (F0 generation raised in the lab for one year; Fort Fisher, North Carolina, United States; n = 2) plus liver tissue as a tissue control to compare with the gut microbiome, and Cualac tessellatus (long-term captive colony; San Luis Potosí, Mexico, n = 4).
Each gut was divided into proximal and distal regions for all San Salvador Island samples to compare microbial composition between these regions. We subsampled the gut only for the San Salvador Island (our focal group) samples to determine if the microbiome composition differs throughout the intestine. The outgroup guts were not the primary focus of this study, therefore we sampled the entire gut for all outgroups. In addition, the microbial community was haphazardly sampled from aquaria water in two tanks which contained F2 individuals of Osprey Lake C. variegatus and Crescent Pond C. variegatus, for a control sample of the existing aquatic microbial community in the common garden lab environment (n = 2). These two samples were taken concurrently with the end of our sampling for this study. The Vincent J. Coates Genomics Sequencing Laboratory at the University of California, Berkeley also generated three controls, including a positive control and two no template controls (NTC). Microbial DNA extractions were performed in batches (stored on ice) immediately after intestinal dissections with the Zymobiomics DNA Miniprep Kit (Zymo Research, Irvine, CA).
16S amplicon sequencing of gut microbiomes
All extracted microbiome DNA samples were quantified with a Nanodrop ND-1000 spectrophotometer (range 4.2–474.9 ng/μl). All samples were then sent to the QB3 Vincent J. Coates Genomics Sequencing Laboratory at the University of California, Berkeley for automated library preparation and sequencing of 16S rRNA amplicons using an Illumina Mi-Seq v3 (600 cycle). As part of the QB3 library preparation, the Forward ITS1 (ITS1f)–CTTGGTCATTTAGAGGAAGTAA and Reverse ITS1 (ITS2)–GCTGGGTTCTTCATCGATGC primers [38] were used for DNA metabarcoding markers for fungi [38]. We removed all eukaryotic sequence reads from our analyses which included fungi reads. QB3 also used the following 16S rRNA primers for amplification of prokaryotes (archaea and bacteria): Forward 16S v4 (515Fb)– GTGYCAGCMGCCGCGGTAA, and Reverse 16S v4 (806Rb)– GGACTACNVGGGTWTCTAAT [39, 40].
Bioinformatic analysis/quantification and microbial ecology assessment of samples
All 16S rRNA amplicon sequences were processed through qiime 2.0 [41] to identify microbe species and estimate abundances. Sequences from all 78 microbiome preps were imported into qiime 2 v. 2019.10.0. We used dada2 [q2-dada2 version 2019.10.0; 42] for modeling and correcting Illumina-sequenced amplicon errors, removing chimeras, trimming low quality bases, and merging of forward and reverse reads using the following parameters:–p-trunc-len-f 270 –p-trunc-len-r 210. The end product of dada2 is an amplicon sequence variant (ASV) table. We used the qiime alignment mafft software to align sequences alignment mask to filter non-conserved and highly gapped columns from the aligned 16S sequences [43]. Next, we used qiime phylogeny midpoint-root to root the phylogeny of our 16S amplicon sequences. Finally, we used qiime diversity alpha-rarefaction on all samples and we set the—p-max-depth to 10,000. We removed samples with 5,000 (only one sample fell below this threshold, Cualac tessellatus; S1 Fig) or less from our analyses as suggested [44, 45].
We compared the beta diversity (qiime emperor plot) of proximal and distal gut microbiomes of the San Salvador samples with a two-tailed paired t-test and found no significant differences between proximal and distal regions of the gut microbiome (P = 0.29). Therefore, we merged the proximal and distal samples for each individual from San Salvador Island, resulting in 48 samples, which included experimental controls and quality controls from the QB3 facility (S2 Table). There was no difference between the means of amplicon sequence reads in the foregut and the hindgut (paired t-test, P = 0.29). We also removed one Cualac tessellatus sample because of low read count (129 reads; S1 Fig).
We used the classifier Silva 132 99% 515F/806R (silva-132-99-515-806-nb-classifier) for training in identification of taxa from our samples. All qiime commands described above were completed in qiime 2 v. 2019.10.0. Afterwards the following files generated in qiime were used in r (v. 4.0.0) for further statistical analyses: table.qza, rooted-tree.qza, taxonomy.qza, and sample-metadata.tsv. We used the following R packages for further analyses: phyloseq v.1.32.0 [46] with the following functions: distance, plot_bar, plot_ordination, and plot_richness. We used ggplot2 v.3.3.2 [47] for creating plots that we generated from phyloseq. Before conducting any analyses, we removed the following taxa from our analyses, uncharacterized and Opisthokonta (eukaryotic sequences mainly due to fish 16S amplicons). We plotted alpha diversity by using the plot_richness function as part of phyloseq, and we plotted both Chao1 and Shannon’s diversities (Fig 1). For beta diversity, we used the plot_ordination function and non-metric multidimensional scaling (NMDS) based on Bray-Curtis distances among samples. We conducted a Permutational multivariate analysis of variance (PERMANOVA) based on Bray-Curtis similiarty matrix to test whether diet or species impact the fish gut microbiomes. This was done with adonis2 which is part of the vegan package (https://CRAN.R-project.org/package=vegan) with 9,999 permutations. We also tested for significance at group-level differences using multivariate homogeneity of groups dispersions with betadisper, which is also part of the vegan package. Hierarchical clustering was generated with the distance function along with hclust as part of fastcluster [48] using the average linkage clustering method. The plot_bar function in the phyloseq package was used to visualize relative abundance of taxa. In only our taxa plots, we removed abundance counts of less than 400 from our analyses to provide the prevalent taxa with higher abundances across samples [49]. We used ggplot2 to generate all figures [47]. We used the linear discriminant analysis effect size [lefse version 1.0; 50] algorithm to identify microbial taxa that were significantly enriched in each of our specialists (molluscivore and scale-eater) in comparison to all other samples. This analysis was used to determine the features (i.e. organisms, clades, operational taxonomic units) to explain differences in assigned metadata categories. We used the nonparametric factorial Kruskal-Wallis rank-sum test to detect taxa with significant differential abundances between specialist samples and all generalist samples (scale-eater versus generalist + molluscivore, molluscivore versus generalist + scale-eater). We then used a Wilcoxon test for all pairwise comparisons between taxa within each significantly enriched class to compare to the class level.
Lake 1 indicates Crescent Pond and Lake 2 represents Osprey Lake, both located on San Salvador Island in the Bahamas. Alpha diversity is represented by (A) Chao1 and (B) Shannon diversity for the estimate of species richness from gut microbiomes from all fishes in this study.
Lastly, we used generalized linear models (GLMs) in r to test the effects of diet (generalist, scale-eater, molluscivore), the fixed effect of location (Osprey Lake, San Salvador Island; Crescent Pond, San Salvador Island; Lake Cunningham, New Providence Island; Fort Fisher, North Carolina; and San Luis Potosí, Mexico), and their interaction on the response variables of principal coordinates axes 1 and 2. We used Principal Coordinates Analysis (PCoA) for only the GLMs to conducts these tests.
Results
Gut microbiome diversity and divergence among taxa and intestinal lengths
We sequenced a total of 11,152,147 reads across all samples (S2 Table). We identified 5,174 bacterial taxa in 48 samples. Similar to other ray-finned fishes [51], proteobacteria is the predominant microbial taxon (S3 Fig). We did not find any significant differences among species in Chao1 or Shannon diversity indices (Kruskal-Wallace [pairwise], P > 0.05; Fig 1). Twenty-one San Salvador Island pupfishes clustered together relative to the three outgroup generalist species, indicating strong host phylogenetic signal associated with overall microbiome diversity (S4 Fig). We also noticed that eight San Salvador Island pupfishes were more dissimilar than the outgroup generalists; however, this may have been due to limited microbial material sampled from these individuals. Throughout the dendrogram, we noticed that within the San Salvador pupfishes, the two specialists (molluscivore and scale-eater) clustered with San Salvador generalists (Figs 2 and S4 and S5). Water and tissue controls were scattered throughout the NMDS plots but were clearly distinct from Cyprinodon microbiome samples with the exception of one tissue control that clustered near the outgroup species, possibly due to contamination during dissections (Fig 2). The ordination stress values for our NMDS plots were 0.191, 0.207, 0.200 for all samples in the study including controls (Fig 2A), San Salvador Island and outgroup species gut microbiomes (Fig 2B), and San Salvador Island species gut microbiome (Fig 2C), respectively. Using PERMANOVA, we found significant differences among gut microbiome communities by diet (df = 1, F = 2.210, P = 1e-04, and betadisper P = 0.512); and by species (df = 5, F = 2.196, P = 1e-04, and betadisper P = 0.000166).
A) NMDS plot based on all Cyprinodon pupfish gut samples labeled according to species and diet including controls (n = 44). Closed circles represent the two specialists (scale-eater and molluscivore) and open circles represent generalists. Open squares and triangles represent controls used in this study. B) Shows a NMDS plot of the three Cyprinodon pupfish species (F2 generation) from San Salvador Island and outgroup members gut microbiomes (n = 39). According to Echelle et al. [33], there is ~11.2 Mya phylogenetic divergence of Cualac tessellatus and Cyprinodon pupfish species, which you can see in the gray arrows and text. The black arrow indicates the 10 kya phylogenetic divergence of C. desquamator and the other Cyprinodon pupfish species from San Salvador Island. C) Lastly, a NMDS plot of only the San Salvador Island species gut microbiomes (n = 30).
Multiple regression analyses of the effects of dietary specialization (generalist, scale-eater, or molluscivore) and the fixed effect of population origin (two different lakes on San Salvador Island, Lake Cunningham, North Carolina, and El Potosí) on PCoA axes 1 and 2 confirmed that population origin and scale-eating had a significant effect on microbiome divergence along both axes (Axis.1: scale-eater P = 0.001 with 52.8% proportion of variance; Axis.2: scale-eater P = 0.018 with 10.3% proportion of variance). However, when evaluating residual deviance compared to the null deviance for both generalized linear models, both showed no differences when using a chi-squared test (glm, P > 0.05). In our comparison of the pupfish gut lengths in our study, we found no significant difference in gut lengths after controlling for specimen size (S2 Fig; ANCOVA with covariate of log-transformed SL; F5,33 = 0.916, P = 0.483).
Linear discriminate analyses of trophic specialist microbiota
We found an excess of reads belonging to taxa within the family Burkholderiaceae in all lab-reared scale-eater individuals from two different lake populations relative to all other gut microbiome samples (Figs 3 and 4; linear discriminant analysis log score = 4.85). We also identified 108 taxa in the family Burkholderiaceae across all scale-eater gut microbiomes (S6 Fig). In addition, we found a deficiency of Vibrionales, Vibrionaceae, and Vibrio in these scale-eater individuals relative to all other gut samples (LDA log scores = -5.22, -5.22, and -5.08, respectively; Fig 4). Similarly, we found an excess of reads belonging to taxa in the family Rhodobacteraceae and class Planctomycetacia in the molluscivores relative to all other gut samples (Fig 5; LDA log scores of 4.39 and 4.37, respectively).
Bars show proportions (relative abundance) of taxa at the family level per individual gut microbiome. Lake 1 indicates Crescent Pond and Lake 2 represents Osprey Lake, both located on San Salvador Island in the Bahamas. Taxa which contained uncharacterized and Opisthokonta (eukaryotic sequences) were removed and taxa with a count of 400 or greater were represented. Taxa were grouped according to species and location (controls included).
A) Log scores of the top four dominant loadings on lefse discriminate axis separating scale-eaters from all other pupfish samples. B) Relative abundance of the family Burkholderiaceae and the order C) Vibrionales among all pupfish gut microbiomes.
A) Log scores of the top two dominant loadings on the lefse discriminate axis separating molluscivores from all other pupfish samples. B) Relative abundance of the family Rhodobacteraceae and the class C) Planctomycetacia from all Cyprinodon gut microbiomes.
Discussion
Using a common garden experiment, we show that differences in gut microbial diversity across Cyprinodon pupfish species largely reflect phylogenetic distance in support of phylosymbiosis [52], rather than the recent rapid evolution of novel trophic specialization as predicted by adaptive radiation theory. Our study is consistent with Ren et al. [53] who also found limited microbiome divergence and minimal associations with ecomorph in an adaptive radiation of Puerto Rican Anolis lizards, even within wild lizards. From our NMDS plot, we see clear differences between the generalists from San Salvador Island and the outgroup generalists, included the most closely extant genus Cualac spanning 11 Mya of evolutionary history (Fig 2). Similar studied found gut microbiome diversity to associate more strongly with geography than phylogeny [54] or a combination of geography, diet, and host phylogeny [55]. These emerging studies of microbiome divergence within adaptive radiations of hosts provide an important counterpoint to the classic expectation of rapid phenotypic diversification and speciation during adaptive radiation [18–21]. For example, a previous study found that craniofacial traits were diversifying up to 1,000 times faster on San Salvador Island than neighboring generalist pupfish populations, including many of the same populations analyzed for gut microbiota in this study [23].
One major caveat is that we did not examine the microbiota of wild-collected animals feeding on their diverse natural resources of macroalgae, scales, and snails. Scales form up to 50% of the diet in scale-eaters [27] and wild gut microbiome samples surely would have revealed more substantial differences in microbiome diversity and composition among generalist and specialist species on San Salvador Island. However, our goal with this common garden study using lab-reared animals spanning multiple generations and fed an identical generalist-type diet for one month was to uncover any genetically based microbiome differences in these taxa by eliminating environmental effects as much as possible. Pupfishes exhibit no parental care and deposit external eggs on the substrate so vertical transmission also appears highly unlikely (but see Satoh et al. [56] for a potential example of vertical transmission in a scale-eating cichlid). It is also possible that even second-generation lab-reared scale-eater colonies consumed more scales from their conspecific tankmates due to elevated behavioral aggression; however, aggression levels are comparable in both trophic specialists [57]. Furthermore, by including two lab-reared colonies of each generalist and specialist species on San Salvador from genetically differentiated and ecologically divergent environments of Crescent Pond and Osprey Lake [23, 28], we aimed to connect significant differences in microbiome composition observed in our specialist species to their specialized diets, rather than their lake environment or genetic background. These results from our ordination plots (Fig 2) are even more surprising because trophic specialists show very little genetic differentiation from San Salvador Island generalists (Fst = 0.1–0.3; 24, 29), yet we see differences in microbiome taxa and abundance across all three species from San Salvador Island and we identified community level differences based on diet from our PERMANOVA. In addition, there is a clear difference between the San Salvador Island pupfishes microbiomes and the generalist outgroup microbiomes, which can be best explained by phylogenetic distance and habitat differences of the host (Fig 2). This has been demonstrated by across multiple fish taxa where host habitat is the key determinant of gut microbiome composition [58]. Even though there were varying conditions throughout our experiment including captive generation (F0, F1, and F2), salinities ranging from 5–10 ppt, and temperatures ranging from 23–30°C, these environmental ranges were distributed haphazardly across our species and populations over the course of treatment period and likely had minimal impact on our conclusions; furthermore, all San Salvador Island populations were lab-reared at least through the F1 generation. Potentially, if we collected wild pupfish gut microbiomes, we may find further resolution from each location, especially between the two locations of San Salvador Island pupfishes (Crescent Pond and Osprey Lake). In addition, we did not detect a gut length difference among the generalist, molluscivore, and scale-eaters F2 pupfishes from San Salvador after controlling for size. We had expected to see shorter gut lengths in the two trophic specialists because of their shift in diet. In marine prickleback fishes (stichaeids), there are differences in gut length due to dietary diversity [59]. Overall, metabolic processes were the single most enriched category among all differentially expressed genes between these trophic specialists at the 8 dpf larval stage, accounting for 20% of differential expression [60], suggesting that gene expression differences may explain differences in microbial communities rather than gross anatomical differences.
Enriched microbiota in scale-eating pupfish
Fish scales are composed of a deep layer that is mostly collagen type I [61]; therefore, we predicted that any adaptive microbes within the scale-eater gut would have collagen degrading properties. We found significant enrichment of one family, the Burkholderiaceae in both scale-eater populations (Figs 3, 4, and S5). Burkholderiaceae is a family of proteobacteria which contains many human and animal pathogens [62], plant and insect symbionts [63, 64], and can be found in soil, water, and polluted environments [65, 66]. Importantly, they also include some collagenase-producing bacteria, such as Burkholderia pseudomallei [UniProtKB–A3P3M6; 67], which is the causative agent of melioidosis in humans [68]. The significant shift in a major collagenase-producing group suggests the potential for an adaptive scale-eater microbiome, even in the absence of dietary scales (except perhaps for incidental aggression and ingestion of scales among tankmates).
In contrast to a microbiome study of the adaptive radiation of Tanganyikan cichlids [69], we found no evidence of Clostridia enrichment in scale-eaters nor a reduction of microbial diversity in this carnivorous species. This may be due to the very young 10 kya age of the scale-eating pupfish relative to the comparatively ancient 12 Mya Tanganyikan radiation and Perissodus scale-eating clade [27, 70].
Enriched microbiota in molluscivore pupfish
We found enrichment of the families Rhodobacteraceae and Planctomycetacia within the molluscivore gut from both lake populations (Fig 5). Both of these taxa are present in the water controls, but not at the level present in the molluscivores (S7 Fig). However, these families have no clear role in anything related to mollusc digestion or even increased levels of protein, lipids, or chitin in the diet (due to some molluscivores specializing on ostracods during periods of abundance). Taxa from these taxonomic group are known to be found within aquatic environments [71, 72]. As noted by Simon et al. [71], marine microbes from the family Rhodobacteraceae play a crucial role in biogeochemical cycling, make up about 30% of bacterial communities within pelagic environments, and generally have a mutualistic relationship with eukaryotes providing vitamins to these groups. Both families are known for aquatic cellulose-decomposing taxa [73, 74], which suggests this microbiome shift may help more with macroalgae digestion rather than molluscs, despite previous observations that macroalgae forms the largest component of the generalist pupfish diet in the hypersaline lakes of San Salvador Island, Bahamas [27].
Conclusion
Many studies have focused on understanding digestion and assimilation within a variety of vertebrates and invertebrates, but there is limited information about the cooperative process between the host intestine cells and gut microbiota, and their role in eco-evolutionary dynamics during rapid species diversification [10, 59, 75]. We found evidence for enrichment of taxa in the Burkholderiaceae family within the scale-eater microbiome (Fig 2), even when hosts were reared in identical environments on identical non-scale diets. However, it is still unknown to what extent this microbiome shift will improve digestion of the collagen found in scales, for example, as demonstrated for the gut fauna in the scale-eating khavalchor catfish [76]. Despite unique and highly specialized pupfish dietary adaptations within shared hypersaline lake habitats, overall gut microbial diversity did not follow the expected pattern of rapid diversification and divergence as observed in their hosts, calling into question how eco-evolutionary dynamics between host and symbiont proceed during adaptive radiation.
Supporting information
S1 Fig. Rarefaction for all 48 (16S) microbiome samples used in this study.
Rarefaction curve constructed based on Amplicon Sequence Variant (ASVs), and samples with less than 6,000 reads (sequence depth) are shown with labels.
https://doi.org/10.1371/journal.pone.0273177.s001
(TIF)
S2 Fig. Scatter plot of the covariate (log standard length) and the outcome variable (log gut length) for all Cyprinodon pupfish species in our study.
Closed circles represent the two specialists (scale-eater and molluscivore) and open circles represent generalists.
https://doi.org/10.1371/journal.pone.0273177.s002
(TIF)
S3 Fig. Total abundance of gut microbes across all Cyprinodon pupfish species used in this study.
Thirty-two phyla of microbes represented across all gut microbiomes, not including controls.
https://doi.org/10.1371/journal.pone.0273177.s003
(TIF)
S4 Fig. A cluster dendrogram based on pupfish gut microbiome taxa using a Bray-Curtis distance (averaged).
For the San Salvador Island samples only, individuals numbered as 1–5 represent Crescent Pond and 6–10 represent Osprey Lake. Scale = scale-eater, Moll = molluscivore, and Gen = generalist.
https://doi.org/10.1371/journal.pone.0273177.s004
(TIF)
S5 Fig. A cluster dendrogram based on pupfish gut microbiome taxa using a Bray-Curtis distance (averaged).
For the San Salvador Island samples only, individuals numbered as 1–5 represent Crescent Pond and 6–10 represent Osprey Lake. Outgroup species to our study are in different shades of blue. Samples which did not cluster with the majority of the San Salvador Island samples as depicted in S4 Fig were removed from the analysis to determine if the same clustering pattern appeared.
https://doi.org/10.1371/journal.pone.0273177.s005
(TIF)
S6 Fig. Abundance counts of genera within the family Burkholderiaceae found in the Scale-eater pupfish gut microbiomes.
Individuals numbered as 1–5 and 6–10 had parental colonies from Crescent Pond and Osprey Lake, respectively.
https://doi.org/10.1371/journal.pone.0273177.s006
(TIF)
S7 Fig. Abundance counts of Rhodobacteraceae and Planctomycetacia of Cyprinodon pupfishes gut microbiomes based on parental location and diet type along with controls.
Lake 1 indicates Crescent Pond and Lake 2 represents Osprey Lake, both located on San Salvador Island in the Bahamas.
https://doi.org/10.1371/journal.pone.0273177.s007
(TIF)
S1 Table. Sample size, location, standard length, gut length, and relative gut length of Cyprinodon pupfish guts.
https://doi.org/10.1371/journal.pone.0273177.s008
(DOCX)
Acknowledgments
We thank L. Smith in the Evolutionary Genetics Lab at the University of California, Berkeley, for generous logistical assistance in preparing microbiome samples; S. McDevitt, C. Miller, and D. Pappas at the Vincent J. Coates Genomics Sequencing Laboratory California Institute for Quantitative Biosciences (QB3) for processing our microbiome samples for 16S amplicon sequencing; R. Berlemont at California State University, Long Beach and C. Weihe from the Microbiome Consortium at the University of California, Irvine for suggestions on microbiome extraction protocols and bioinformatics workflow; and the Gerace Research Centre and Troy Day for logistical support in the Bahamas. We thank the Zoological Society of London for providing C. tessellatus eggs and the governments of the Bahamas and United States for permission to collect and export Cyprinodon samples.
References
- 1. Turcotte MM, Reznick DN, Daniel Hare J. Experimental test of an eco-evolutionary dynamic feedback loop between evolution and population density in the green peach aphid. The American Naturalist. 2013 May 1;181(S1):S46–57. pmid:23598359
- 2. Matthews B, Aebischer T, Sullam KE, Lundsgaard-Hansen B, Seehausen O. Experimental evidence of an eco-evolutionary feedback during adaptive divergence. Current Biology. 2016 Feb 22;26(4):483–9. pmid:26804555
- 3. Rudman SM, Barbour MA, Csilléry K, Gienapp P, Guillaume F, Hairston NG Jr, et al. What genomic data can reveal about eco-evolutionary dynamics. Nature ecology & evolution. 2018 Jan;2(1):9–15. pmid:29158555
- 4. Post DM, Palkovacs EP. Eco-evolutionary feedbacks in community and ecosystem ecology: interactions between the ecological theatre and the evolutionary play. Philosophical Transactions of the Royal Society B: Biological Sciences. 2009 Jun 12;364(1523):1629–40. pmid:19414476
- 5. Gould AL, Zhang V, Lamberti L, Jones EW, Obadia B, Korasidis N, et al. Microbiome interactions shape host fitness. Proceedings of the National Academy of Sciences. 2018 Dec 18;115(51):E11951–60. pmid:30510004
- 6. Macke E., Tasiemski A., Massol F., Callens M., & Decaestecker E. (2017). Life history and eco‐evolutionary dynamics in light of the gut microbiota. Oikos, 126(4), 508–531.
- 7. Walters AW, Hughes RC, Call TB, Walker CJ, Wilcox H, Petersen SC, et al. The microbiota influences the Drosophila melanogaster life history strategy. Molecular ecology. 2020 Feb;29(3):639–53.
- 8. Videvall E, Song SJ, Bensch HM, Strandh M, Engelbrecht A, Serfontein N, et al. Major shifts in gut microbiota during development and its relationship to growth in ostriches. Molecular Ecology. 2019 May;28(10):2653–67. pmid:30916826
- 9. Videvall E, Song SJ, Bensch HM, Strandh M, Engelbrecht A, Serfontein N, et al. Early-life gut dysbiosis linked to juvenile mortality in ostriches. Microbiome. 2020 Dec;8(1):1–3.
- 10. Baldo L, Pretus JL, Riera JL, Musilova Z, Bitja Nyom AR, Salzburger W. Convergence of gut microbiotas in the adaptive radiations of African cichlid fishes. The ISME journal. 2017 Sep;11(9):1975–87. pmid:28509910
- 11. Trevelline BK, Kohl KD. The gut microbiome influences host diet selection behavior. Proceedings of the National Academy of Sciences. 2022 Apr 26;119(17):e2117537119. pmid:35439064
- 12. Baldo L, Riera JL, Salzburger W, Barluenga M. Phylogeography and ecological niche shape the cichlid fish gut microbiota in Central American and African Lakes. Frontiers in microbiology. 2019:2372. pmid:31681230
- 13. Loo WT, García-Loor J, Dudaniec RY, Kleindorfer S, Cavanaugh CM. Host phylogeny, diet, and habitat differentiate the gut microbiomes of Darwin’s finches on Santa Cruz Island. Scientific Reports. 2019 Dec 11;9(1):1–2.
- 14. Rennison DJ, Rudman SM, Schluter D. Parallel changes in gut microbiome composition and function during colonization, local adaptation and ecological speciation. Proceedings of the Royal Society B. 2019 Dec 4;286(1916):20191911. pmid:31795865
- 15. Brooks AW, Kohl KD, Brucker RM, van Opstal EJ, Bordenstein SR. Phylosymbiosis: relationships and functional effects of microbial communities across host evolutionary history. PLoS biology. 2016 Nov 18;14(11):e2000225. pmid:27861590
- 16. Lim SJ, Bordenstein SR. An introduction to phylosymbiosis. Proceedings of the Royal Society B. 2020 Mar 11;287(1922):20192900. pmid:32126958
- 17. Losos JB. Phylogenetic niche conservatism, phylogenetic signal and the relationship between phylogenetic relatedness and ecological similarity among species. Ecology letters. 2008 Oct;11(10):995–1003. pmid:18673385
- 18. Stroud JT, Losos JB. Ecological opportunity and adaptive radiation. Annu. Rev. Ecol. Evol. Syst. 2016 Jan 1;47(47):507–32.
- 19.
Schluter D. The ecology of adaptive radiation. OUP Oxford; 2000 Aug 31.
- 20. Martin CH, Richards EJ. The paradox behind the pattern of rapid adaptive radiation: how can the speciation process sustain itself through an early burst?. Annual Review of Ecology, Evolution, and Systematics. 2019;50(1):569–93.
- 21. Gillespie RG, Bennett GM, De Meester L, Feder JL, Fleischer RC, Harmon LJ, et al. Comparing adaptive radiations across space, time, and taxa. Journal of Heredity. 2020 Jan;111(1):1–20. pmid:31958131
- 22. Rundell RJ, Price TD. Adaptive radiation, nonadaptive radiation, ecological speciation and nonecological speciation. Trends in Ecology & Evolution. 2009 Jul 1;24(7):394–9.
- 23. Martin CH, Crawford JE, Turner BJ, Simons LH. Diabolical survival in Death Valley: recent pupfish colonization, gene flow and genetic assimilation in the smallest species range on earth. Proceedings of the Royal Society B: Biological Sciences. 2016 Jan 27;283(1823):20152334. pmid:26817777
- 24. Richards EJ, McGirr JA, Wang JR, John ME, Poelstra JW, Solano MJ, et al. Major stages of vertebrate adaptive radiation are assembled from a disparate spatiotemporal landscape. bioRxiv. 2020 Jan 1.
- 25.
Echelle AA, Echelle AF. Mode and pattern of speciation in the evolution of inland pupfishes of the Cyprinodon variegatus complex (Teleostei: Cyprinodontidae): an ancestor-descendat hypothesis. Systematics, historical ecology, and North American freshwater fishes (ed. by Mayden R.L.). 1992 pp. 691–709. Stanford University Press, Stanford, CA.
- 26. Martin CH, Wainwright PC. Trophic novelty is linked to exceptional rates of morphological diversification in two adaptive radiations of Cyprinodon pupfish. Evolution: International Journal of Organic Evolution. 2011 Aug;65(8):2197–212.
- 27. Martin CH, Wainwright PC. On the measurement of ecological novelty: scale-eating pupfish are separated by 168 my from other scale-eating fishes. PLoS One. 2013 Aug 19;8(8):e71164. pmid:23976994
- 28. Richards EJ, Martin CH. Adaptive introgression from distant Caribbean islands contributed to the diversification of a microendemic adaptive radiation of trophic specialist pupfishes. PLoS genetics. 2017 Aug 10;13(8):e1006919. pmid:28796803
- 29. Martin CH, Feinstein LC. Novel trophic niches drive variable progress towards ecological speciation within an adaptive radiation of pupfishes. Molecular Ecology. 2014 Apr;23(7):1846–62. pmid:24393262
- 30.
Holtmeier CL. Morphological and trophic diversification among pupfish (Cyprinodontidae): Dietary, genetic and ontogenetic effects. Cornell University; 2000.
- 31. Holtmeier CL. Heterochrony, maternal effects, and phenotypic variation among sympatric pupfishes. Evolution. 2001 Feb;55(2):330–8. pmid:11308091
- 32. Turner BJ, Duvernell DD, Bunt TM, Barton MG. Reproductive isolation among endemic pupfishes (Cyprinodon) on San Salvador Island, Bahamas: microsatellite evidence. Biological Journal of the Linnean Society. 2008 Nov 1;95(3):566–82.
- 33. Echelle AA, Carson EW, Echelle AF, Van Den Bussche RA, Dowling TE, Meyer A. Historical biogeography of the new-world pupfish genus Cyprinodon (Teleostei: Cyprinodontidae). Copeia. 2005 May;2005(2):320–39.
- 34. McGirr JA, Martin CH. Few fixed variants between trophic specialist pupfish species reveal candidate cis-regulatory alleles underlying rapid craniofacial divergence. Molecular biology and evolution. 2021 Feb;38(2):405–23. pmid:32877534
- 35. Williams BB, Tebbutt NC, Buchert M, Putoczki TL, Doggett K, Bao S, et al. Glycoprotein A33 deficiency: a new mouse model of impaired intestinal epithelial barrier function and inflammatory disease. Disease models & mechanisms. 2015 Aug 1;8(8):805–15. pmid:26035389
- 36.
Wilson JM, Castro LF. Morphological diversity of the gastrointestinal tract in fishes. In Fish physiology 2010 Jan 1 (Vol. 30, pp. 1–55). Academic Press.
- 37. Patton AH, Richards EJ, Gould KJ, Buie LK, Martin CH. Hybridization alters the shape of the genotypic fitness landscape, increasing access to novel fitness peaks during adaptive radiation. eLife. 2022 May 26;11:e72905. pmid:35616528
- 38. Smith DP, Peay KG. Sequence depth, not PCR replication, improves ecological inference from next generation DNA sequencing. PloS one. 2014 Feb 28;9(2):e90234. pmid:24587293
- 39. Caporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Lozupone CA, Turnbaugh PJ, et al. Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample. Proceedings of the national academy of sciences. 2011 Mar 15;108(Supplement 1):4516–22.
- 40. Apprill A, McNally S, Parsons R, Weber L. Minor revision to V4 region SSU rRNA 806R gene primer greatly increases detection of SAR11 bacterioplankton. Aquatic Microbial Ecology. 2015 Jun 4;75(2):129–37.
- 41. Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nature biotechnology. 2019 Aug;37(8):852–7. pmid:31341288
- 42. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJ, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. Nature methods. 2016 Jul;13(7):581–3. pmid:27214047
- 43.
Stackebrandt E, Goodfellow M. Nucleic acid techniques in bacterial systematics. Wiley; 1991.
- 44. Hughes JB, Hellmann JJ, Ricketts TH, Bohannan BJ. Counting the uncountable: statistical approaches to estimating microbial diversity. Applied and environmental microbiology. 2001 Oct 1;67(10):4399–406. pmid:11571135
- 45. Cameron ES, Schmidt PJ, Tremblay BJ, Emelko MB, Müller KM. Enhancing diversity analysis by repeatedly rarefying next generation sequencing data describing microbial communities. Scientific reports. 2021 Nov 16;11(1):1–3.
- 46. McMurdie PJ, Holmes S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PloS one. 2013 Apr 22;8(4):e61217. pmid:23630581
- 47.
Wickham H. ggplot2: elegant graphics for data analysis. springer; 2016 Jun 8.
- 48. Müllner D. fastcluster: Fast hierarchical, agglomerative clustering routines for R and Python. Journal of Statistical Software. 2013 May 29;53:1–8.
- 49. Simons AL, Churches N, Nuzhdin S. High turnover of faecal microbiome from algal feedstock experimental manipulations in the Pacific oyster (Crassostrea gigas). Microbial biotechnology. 2018 Sep;11(5):848–58. pmid:29749083
- 50. Segata N, Izard J, Waldron L, Gevers D, Miropolsky L, Garrett WS, et al. Metagenomic biomarker discovery and explanation. Genome biology. 2011 Jun;12(6):1–8. pmid:21702898
- 51. Youngblut ND, Reischer GH, Walters W, Schuster N, Walzer C, Stalder G, et al. Host diet and evolutionary history explain different aspects of gut microbiome diversity among vertebrate clades. Nature communications. 2019 May 16;10(1):1–5.
- 52. Bordenstein SR, Theis KR. Host biology in light of the microbiome: ten principles of holobionts and hologenomes. PLoS biology. 2015 Aug 18;13(8):e1002226. pmid:26284777
- 53. Ren T, Kahrl AF, Wu M, Cox RM. Does adaptive radiation of a host lineage promote ecological diversity of its bacterial communities? A test using gut microbiota of Anolis lizards. Molecular ecology. 2016 Oct;25(19):4793–804.
- 54. Godoy-Vitorino F, Leal SJ, Díaz WA, Rosales J, Goldfarb KC, García-Amado MA, et al. Differences in crop bacterial community structure between hoatzins from different geographical locations. Research in Microbiology. 2012 Apr 1;163(3):211–20. pmid:22313738
- 55. Antonopoulou E, Nikouli E, Piccolo G, Gasco L, Gai F, Chatzifotis S, et al. Reshaping gut bacterial communities after dietary Tenebrio molitor larvae meal supplementation in three fish species. Aquaculture. 2019 Mar 30;503:628–35.
- 56. Satoh S, Awata S, Tanaka H, Jordan LA, Kakuda U, Hori M, et al. Bi-parental mucus provisioning in the scale-eating cichlid Perissodus microlepis (Cichlidae). Biological Journal of the Linnean Society. 2019 Dec 6;128(4):926–35.
- 57. St. John ME, McGirr JA, Martin CH. The behavioral origins of novelty: did increased aggression lead to scale-eating in pupfishes? Behavioral Ecology. 2019 Mar;30(2):557–69. pmid:30971862
- 58. Kim PS, Shin NR, Lee JB, Kim MS, Whon TW, Hyun DW, et al. Host habitat is the major determinant of the gut microbiome of fish. Microbiome. 2021 Dec;9(1):1–6.
- 59. German DP, Sung A, Jhaveri P, Agnihotri R. More than one way to be an herbivore: convergent evolution of herbivory using different digestive strategies in prickleback fishes (Stichaeidae). Zoology. 2015 Jun 1;118(3):161–70. pmid:25769813
- 60. McGirr JA, Martin CH. Parallel evolution of gene expression between trophic specialists despite divergent genotypes and morphologies. Evolution letters. 2018 Apr;2(2):62–75. pmid:30283665
- 61. Harikrishna N, Mahalakshmi S, Kiran Kumar K, Reddy G. Fish scales as potential substrate for production of alkaline protease and amino acid rich aqua hydrolyzate by Bacillus altitudinis GVC11. Indian journal of microbiology. 2017 Sep;57(3):339–43.
- 62. diCenzo GC, Mengoni A, Perrin E. Chromids aid genome expansion and functional diversification in the family Burkholderiaceae. Molecular biology and evolution. 2019 Mar 1;36(3):562–74. pmid:30608550
- 63. Gyaneshwar P, Hirsch AM, Moulin L, Chen WM, Elliott GN, Bontemps C, et al. Legume-nodulating betaproteobacteria: diversity, host range, and future prospects. Molecular plant-microbe interactions. 2011 Nov;24(11):1276–88. pmid:21830951
- 64. Takeshita K, Kikuchi Y. Riptortus pedestris and Burkholderia symbiont: an ideal model system for insect–microbe symbiotic associations. Research in Microbiology. 2017 Apr 1;168(3):175–87.
- 65. Coenye T, Vandamme P. Diversity and significance of Burkholderia species occupying diverse ecological niches. Environmental microbiology. 2003 Sep;5(9):719–29. pmid:12919407
- 66. Estrada-De Los Santos P, Rojas-Rojas FU, Tapia-García EY, Vásquez-Murrieta MS, Hirsch AM. To split or not to split: an opinion on dividing the genus Burkholderia. Annals of Microbiology. 2016 Sep;66(3):1303–14.
- 67. Rainbow L, Wilkinson MC, Sargent PJ, Hart CA, Winstanley C. Identification and expression of a Burkholderia pseudomallei collagenase in Escherichia coli. Current microbiology. 2004 Apr;48(4):300–4.
- 68. Holden MT, Titball RW, Peacock SJ, Cerdeño-Tárraga AM, Atkins T, Crossman LC, et al. Genomic plasticity of the causative agent of melioidosis, Burkholderia pseudomallei. Proceedings of the National Academy of Sciences. 2004 Sep 28;101(39):14240–5.
- 69. Baldo L, Riera JL, Tooming-Klunderud A, Albà MM, Salzburger W. Gut microbiota dynamics during dietary shift in eastern African cichlid fishes. PloS one. 2015 May 15;10(5):e0127462. pmid:25978452
- 70. Koblmüller S, Egger B, Sturmbauer C, Sefc KM. Evolutionary history of Lake Tanganyika’s scale-eating cichlid fishes. Molecular phylogenetics and evolution. 2007 Sep 1;44(3):1295–305. pmid:17383901
- 71. Simon M, Scheuner C, Meier-Kolthoff JP, Brinkhoff T, Wagner-Döbler I, Ulbrich M, et al. Phylogenomics of Rhodobacteraceae reveals evolutionary adaptation to marine and non-marine habitats. The ISME journal. 2017 Jun;11(6):1483–99. pmid:28106881
- 72. Yilmaz P, Yarza P, Rapp JZ, Glöckner FO. Expanding the world of marine bacterial and archaeal clades. Frontiers in microbiology. 2016 Jan 8;6:1524. pmid:26779174
- 73. Ringø EZ, Zhou Z, Vecino JG, Wadsworth S, Romero J, Krogdahl Å, et al. Effect of dietary components on the gut microbiota of aquatic animals. A never‐ending story? Aquaculture nutrition. 2016 Apr;22(2):219–82.
- 74. Kim JW, Brawley SH, Prochnik S, Chovatia M, Grimwood J, Jenkins J, et al. Genome analysis of Planctomycetes inhabiting blades of the red alga Porphyra umbilicalis. PLoS One. 2016 Mar 25;11(3):e0151883.
- 75.
Terra WR, Barroso IG, Dias RO, Ferreira C. Molecular physiology of insect midgut. In Advances in insect physiology 2019 Jan 1 (Vol. 56, pp. 117–163). Academic Press.
- 76. Gosavi SM, Kharat SS, Kumkar P, Navarange SS. Interplay between behavior, morphology and physiology supports lepidophagy in the catfish Pachypterus khavalchor (Siluriformes: Horabagridae). Zoology. 2018 Feb 1;126:185–91. pmid:29191622