Sexual dimorphism shapes the gut microbiome of northern elephant seal pups across environments

Sexual dimorphism shapes the gut microbiome of northern elephant seal pups across environments | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 13 January 2025 V1 Latest version Share on Sexual dimorphism shapes the gut microbiome of northern elephant seal pups across environments Authors : Emily Yu 0009-0007-5604-2188 [email protected] , Alexandra DeCandia 0000-0001-8485-5556 , Andrea Graham , Emily Whitmer 0000-0002-5956-6944 , Cara Field , Bridgett vonHoldt 0000-0001-6908-1687 , and Stephen Gaughran 0000-0002-9413-5074 Authors Info & Affiliations https://doi.org/10.22541/au.173675473.39676661/v1 448 views 218 downloads Contents Abstract Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Northern elephant seals (Mirounga angustirostris) exhibit some of the strongest anatomical and behavioral sexual dimorphism of any mammalian species. The degree of dimorphism at the microbial level, especially in young individuals, is still relatively unknown. Here, we investigated the interplay between sex, county of stranding, rehabilitation environment, and host genetics on the gut microbiomes of 44 northern elephant seal pups that were stranded along the California coastline and brought to a rehabilitation facility. Using a metabarcoding approach, we characterized microbial communities shortly after admission to the facility and found that both sex and county of stranding contributed to variation in microbial composition. Through population genetic analyses, we showed that the effect of county of stranding on microbial composition was not driven by underlying genetic structure. More broadly, we did not find any correlation between host genetics and microbiome dissimilarity, perhaps related to the extremely low genetic diversity of this bottlenecked species. Finally, we analyzed paired samples from a subset of 24 seals at two time points: shortly after admission to the rehabilitation facility and a month post-acclimation in the facility. Although microbiome compositions became more similar over time, sex continued to contribute to variation. Sex had a weaker effect on microbiome variation at the second time point in comparison to the first, potentially due to the homogenizing effects of rehabilitation. Our findings ultimately help shape our understanding of how environment and sex shape the gut microbiomes of young NES during an understudied period of development. Introduction The mammalian gut microbiome is composed of more than a trillion diverse microorganisms (Thursby & Juge 2017) that aid in metabolic functioning, pathogen defense, and immune signaling and regulation (Belkaid & Hand 2014; Jandhyala et al . 2015; Kinross et.al 2011; Shreiner et. al 2015). Studies in humans and model organisms have shown that microbial composition and diversity can vary with age, sex, genetics, diet, environment, stress, and other factors (Hasan & Yang 2019; Kurilshikov et al. 2017; Tasnim et al. 2017). While there has been tremendous progress in characterizing the gut microbiome of humans and model organisms, there is a substantial gap in our understanding of the gut microbiomes of non-model organisms, especially in pinnipeds and other marine mammals. The extreme sexual dimorphism of northern elephant seals (NES; Beltran et al. 2022) renders them particularly suitable for considering the role of sex in shaping microbial communities. At birth, female and male NES are equivalent in size, but extreme sexual size dimorphism develops around 4-5 years of age during puberty, due to exponential growth rates in males (Le Boeuf et. al 1994). As adults, males weigh up to ten times more than females (Deutsch et al. 1994; Stewart 1997) and have sex-specific behavioral, dietary, ecological, and physiological traits (Kienle et al. 2022; Le Boeuf et. al 2000; Reiter et. al 1981; Stewart 1997). These dramatic sex-specific differences are likely the result of divergent sex-specific social and energetic needs for reproductive success, as predicted by evolutionary theory (Slatkin 1984; Williams & Carroll 2009). Despite strong sexual dimorphism in NES, little is known about the molecular mechanisms underlying this dimorphism. Sex-linked genes are predominantly responsible for major sex-specific differences, mediated through the expression or repression of genes on the X- or Y-chromosomes (Deegan & Engel 2019; Sekido & Lovell-Badge 2009), although autosomal genes, epigenetics, and gene expression patterns also likely play a role. For example, pre-pubescent physiological dimorphism has been measured in yearling NES (Jelincic et al. 2017; Kelso et al. 2012), which suggests that molecular dimorphism connected to differences in hormone expression or resource allocation may begin at early ages. In contrast to research on other wild vertebrate species with low (Bennett et al. 2016; Bolnick et al. 2014; Maurice et al. 2015; Park & Im 2020) or moderate (e.g., gorillas, Pafčo et al. 2019; baboons, Tung et al. 2015) sexual dimorphism, a few studies have detected sex as a significant factor in microbiome composition in NES pups (Stoffel et. al 2020) and in a closely related species, the southern elephant seal (Kim et al. 2020). Previous NES research was conducted during the immediate post-weaning period in which seal pups fasted and stayed in their natal rookery, thus minimizing effects of diet and habitat (Stoffel et. al 2020). It is unknown how environmental factors overshadow intrinsic factors, like sex, beyond the weaning period. Furthermore, as NES mature following weaning, they may exhibit new physiological, molecular adaptations to overcome life challenges (Pacheco-Sandoval et. al 2019). In addition to the influence of sex on the gut microbiome, another largely open question is the extent to which environmental factors associated with habitat contribute to gut microbiome composition. When comparing animals from aquatic and terrestrial environments, habitat has been suggested to be an important determinant of microbiome composition (Bik et. al 2016). It is less clear whether differences in habitat within a single marine mammal species shape microbiome composition. Studies of wild pinniped gut microbiomes have primarily included animals from a single colony or study area (Nelson et. al 2013; Pacheco‑Sandoval et. al 2022; Kim et. al 2020) due in part to logistical barriers in collecting samples from varying geographic locations ( e.g. IRB approval, transportation costs). To date, there has only been one study that has focused on this topic, and it found that natal terrestrial sites contributed to variation in wild gray seal gut microbiome composition (Watkins et. al 2022). The link between pinniped natal habitats and microbiome composition remains poorly understood. One particular challenge in investigating the influence of environmental factors, like habitat, is disentangling their effects from host genetics. Host genotype has been shown to strongly influence microbiome composition in model organisms and other wild populations (DeCandia et al. 2021; de Jonge et. al 2022; Rojas et al . 2020; Turnbaugh et al. 2009; Zhu et al . 2021). Traditionally, model organisms, such as inbred mouse lines, have been used to study the influence of environmental factors in the absence of genetic variation (Spor et. al 2011). The disentangling of these two variables is generally not possible with wild animal populations due to their much greater levels of genetic variation. However, a potential exception may be the NES. Due to their history of a severe population bottleneck, NES have very low genetic variation and high levels of inbreeding (Hoelzel et al. 2002; Weber et al. 2000), which should result in lower levels of genetically controlled phenotypic variation compared to outbred populations (Fowler & Whitlock 1999). NES may represent one of the few natural systems in which non-genetic factors, such as environment, can be assessed on a background of minimal genetic variability. Here we investigated the interplay between sex, county of stranding (county in which animal was rescued or “stranded”), and host genetics on NES gut microbiomes during an understudied age (~3–4 months), when pups conclude their monthslong fast and begin independently foraging beyond their natal rookeries. We used a metabarcoding approach to characterize the gut microbiome of NES pups shortly after admission to a rehabilitation facility in California, USA, and explored how intrinsic and extrinsic factors correlated with microbiome diversity and composition. We then paired a subset of the samples collected shortly after admission with a sample collected 1-month post-admission to track how a change in environment, the transition from the wild to a rehabilitation facility, may have impacted the gut microbiome. Given the extreme sexual dimorphism of NES, we hypothesized that sex would be a major factor in shaping microbiome composition, even at this early life stage and despite any effects of geographic origin or rehabilitation facility. Rehabilitation environment and sample collection In March 2021, NES pups were reported as stranded (washed ashore) across six counties along the coast of California (Fig. 1) and trained marine mammal rescue teams from The Marine Mammal Center (TMMC) were dispatched to recover the seals. Most seals were immediately transported to the main outdoor TMMC rehabilitation hospital in Sausalito, California (USA). Some seals that stranded in San Luis Obispo County or Monterey County were kept in a local TMMC triage facility for up to 48 hours before transfer to the main hospital. Pups were transported in individual carriers within enclosed vans. Pups were housed in groups of 3-8 conspecifics in 15x20 foot concrete pens with a closed non-heated system pool (8x8 feet, 4 feet deep), maintained at salinity of 24-30 parts per thousand with sand filtration and a continuous turnover rate of approximately 30 minutes and disinfected by ozone filtration. Pups were initially fed a slurry of herring ( Clupea spp.) with salmon oil and water by orogastric tube three times daily, and were concurrently introduced to thawed whole herring. Tube feeding was discontinued when pups were reliably eating whole fish. Each pup received vitamin B complex (VetOne, MWI Animal Health, Boise, ID, USA) via intramuscular injection once daily for three doses upon admission to care and an oral multi-vitamin supplement (Pinnivite, Mazuri, Richmond, IN, USA) daily while in care. We reviewed animal history and health data including stranding date and location, reason for stranding and admission to rehabilitation, sex, medical diagnoses and treatments in rehabilitation, and health outcomes ( i.e. , released, died, or euthanized). We collected rectal swabs from each animal during admission examination (typically within 3 days of entering care) and opportunistically during follow-up exams. Each swab was placed in a dry cryovial and stored at -80 O C until processing. Microbial DNA extraction and sequencing We randomly sorted rectal samples across two 96-well plates to minimize batch effects. To extract DNA we used a modified Qiagen DNeasy PowerSoil Kit protocol. Briefly, we first transferred swab tips to their predetermined location in a 96-well PowerBead Plate. Across batches, we reserved seven wells for negative controls (distilled water) and six wells for positive controls (ZymoBIOMICS Microbial Community Standard D6300) to assess potential sources of contamination and ensure successful amplification of bacteria. We added 750µL of PowerBead Solution to each sample and control, and then placed the plate on a Qiagen TissueLyser II for 12 minutes at 20 Hz/seconds, followed by the addition of 60µL of Solution C1. We incubated each plate for 10 minutes at 65°C. We repeated the TissueLyser step for 12 minutes for 20 Hz/sec. We then followed the standard manufacturer protocol, with the additional step of heating elution buffer C6 to 70°C before use. We used the Quant-It kit (Qiagen) to determine DNA concentrations and standardized all samples to 2.5ng/µL. Following the protocol of DeCandia et al . (2019, 2020, 2021) and Lu et. al (2023), we amplified the 16S rRNA V4 region using polymerase chain reaction (PCR) in a 13.2µL total reaction volume composed of: 5µL of 2x MyTaq HS Red Mix, 3.2µL of the forward and reverse primer mix (1.25µM; Caporaso et al. 2011), and 1.8µL of template DNA (4.5ng of DNA). We used distinct combinations of uniquely barcoded forward (n=8) and reverse (n=12) primers (Caporaso et al. 2011). An additional positive control, ZymoBIOMICS Microbial Community DNA Standard D6305, was added for PCR. The PCR cycling conditions were: initial denaturation of 94°C for 3 min; 30 touchdown cycles of 94°C for 45 s, 80°C–50°C for 60 s, 72°C for 90 s with 1°C decrease each cycle; 12 cycles of 94°C for 45 s, 50°C for 60 s, 72°C for 90 s; and a final extension of 72°C for 10 minutes. We checked a random subset of eight samples on a 2% agarose gel via electrophoresis to confirm amplification (300-400 nucleotides) of the target 16S rRNA V4 region plus primers. We pooled equal nanograms of DNA from both plates and selected for fragments 300-400 nucleotides in size using Agencourt AMPure XP magnetic beads. We submitted final libraries to Princeton University’s Lewis-Sigler Institute for Integrated Genomics (LSI) for Illumina paired-end multiplexed sequencing on the Illumina MiSeq (2x150nt; configuration Miseq v2; Micro 300nt kit). 16S rRNA demultiplexing and bioinformatic processing We demultiplexed raw sequence data and allowed for one nucleotide mismatch using a paired-end, dual barcode splitter in Galaxy (Afghan et al. 2018). We imported demultiplexed reads into QIIME2 v2022.4 for downstream processing and analysis (Afghan et al. 2018; Bolyen et al. 2019). We used the dada2 denoise-paired function to filter and trim low-quality sequences (Callahan et. al 2016). We trimmed 13 bases at the beginning of each read and then truncated sequences at the 150th base. The paired reads were then merged. For the positive controls, we looked for the presence of bacterial species expected in our mock community standards, including three gram-negative and five gram-positive bacterial species. Next, we explored the number of sequences in our negative controls, which was expected to be relatively low compared to our samples, around the few hundreds. We removed all positive and negative controls from further analyses. For all further microbiome bioinformatic pipelines, we used QIIME2 v2024.5. We removed amplicon sequence variants (ASVs) that were classified as mitochondria or chloroplasts with the filter-table function, followed by the feature-table filter-features function to remove ASVs that were found as a singleton in only one sample. Microbiome inclusion criteria We wanted to characterize the gut microbiome of pups as shortly after admission to TTMC as possible. Thus, we created an admission dataset with a cohort of seals that passed the following inclusion criteria (Fig. 2). To control for age, pups must have been admitted for rehabilitation in the month of March. As the peak number of NES births occurs in January each year (Condit et. al 2022), we assumed all pups in the study were approximately 3 months of age at admission. The first rectal sample must have been collected within 10 days of admission to TMMC in order to minimize the effects of change in environment on microbial composition. Finally, we only included individuals that were not euthanized, died in treatment, or those who did not receive any medication beyond vitamin supplementation ( i.e. anthelmintics, opioids, or antibiotics). When a seal was represented by duplicate samples, we retained the sample with the greater number of sequences. Admissions dataset: alpha and beta diversity analyses To conduct phylogenetic diversity metrics on our full sample dataset, we utilized the QIIME2 align-to-tree-mafft-fasttree pipeline. We conducted multiple sequence alignment using the mafft program followed by filtering highly variable nucleotide positions (Katoh & Standley 2013). Using this filtered sequence alignment, we implemented the FastTree program to generate a rooted phylogenetic tree with a midpoint from the longest tip-to-tip distance (Price et al . 2009). To visualize alpha diversity as a function of sampling depth, we created alpha rarefaction curves at a maximum depth of 10,000 (Fig. S1). We used the core-metrics-phylogenetic function to generate alpha and beta diversity metrics at a rarefaction depth of 4835, a threshold that we selected from the alpha rarefaction curve, which retained the maximum diversity without excluding any samples. We calculated three different alpha, or within sample, diversity metrics: Observed features (OF), a measure of community richness or number of different ASVs in the community; Pielou’s evenness (PE), a measure of community equitability (relative abundances) of species in a community; and 3) Shannon’s diversity index (SDI), a measure that considers both species richness and evenness. We implemented the Kruskal-Wallis test (Kruskal & Wallis 1952) through the alpha-group-significance QIIME2 plugin to determine differences between groups in three variables of interest: county of stranding, biological sex, and sequencing batch. Sequencing batch was added as a covariate throughout all our alpha and beta diversity analyses to minimize batch effects. We used a statistical significance threshold of p <0.05. We then calculated three different beta diversity, or between sample, metrics: Bray-Curtis dissimilarity (BC), a quantitative measure for abundance; Unweighted Unifrac (UU), a phylogenetic, qualitative measure based on ASV presence and absence; and Weighted Unifrac (WU), a phylogenetic, quantitative measure that considers abundance of ASVs. We used the tidyverse (Wickham et. al 2019), devtools (Wickham et. al 2022), and qiime2R (Bisanz 2018) packages in RStudio to visualize principal coordinate analysis (PCoA). To help interpret our PCoA plots, we performed permutational multivariate analysis of variance (PERMANOVA; Anderson 2001) using the adonis function. PERMANOVA determines if the centroid of samples differs between groups. Similar to ANOVA, PERMANOVA partitions distance matrices into sources of variation and produces a pseudo-F value. p -values are determined using a permutation test. We conducted a univariate and multivariate analysis for the following variables: county of stranding, biological sex, and sequencing batch. Using the beta-group-significance function, we conducted pairwise PERMANOVA comparisons of counties to identify individual differences that could be driving significant univariate results. The order of the variables in our PERMANOVA multivariate models was critical as adonis performs a sequential test of terms, meaning that the variance that is not explained by the first term is passed on to the second term, and so on. We designed a unique model for each of our three selected variables, where the variable of interest was last and the preceding order of covariates was controlled for and consistent across all beta diversity models (Table S1). PERMANOVA results can be driven by within-in group variation, or dispersion, rather than differences in centroid values (Anderson 2001). Thus, we tested for homogeneity of group variances for all variables of interest using permadisp within the beta-group-significance function. We set permutations at 999 and a statistical significance threshold of p <0.05 for all beta diversity analyses. Admissions dataset: differential abundance analysis A common drawback of taxonomic composition analyses is that taxa are measured as relative abundance estimates (Mandal et. al 2015). Consequently, a change in the absolute abundance of a single taxon alters its relative abundance as well as the relative abundance of all other taxa (Lin & Peddada 2020). It becomes statistically challenging to identify which taxa drive significant differences in abundance between samples or conditions. To overcome this challenge, we conducted a differential abundance analysis. We first trained a Naïve Bayes classifier using reference sequences from Greengenes 13_8 99% OTU database and then used the classify-sklearn function to assign each sequence to known taxonomy (Bokulich et al. 2018; DeSantis et al . 2006). We then used the filter-table function to remove ASVs that were not classified beyond the phylum level. We used analysis of compositions of microbes with bias correction (ANCOM-BC; Lin & Peddada 2020) to detect differential abundances between males and females on a genus and class level. We added batch and county as covariates and used Bonferroni correction to adjust p -values ( q -values). We set a significance threshold of q <0.05 and visualized our results using the da-barplot function and VegaEditor (Satyanarayan et. al 2016). Longitudinal dataset analyses We created a longitudinal dataset to understand how pup gut microbiomes changed during their stay in the rehabilitation facility. From our admission dataset, we subsetted pups with samples collected at two time points: within 10 days of admission (T1) and approximately one month after admission (T2). If multiple samples were available for T2, we chose the one with the greater number of sequences. We first combined samples from T1 and T2. Following the methods above, we created an alpha-rarefaction curve and calculated our alpha and beta diversity metrics at a rarefaction depth of 4835 (Fig. S2). We were interested in whether there was significant difference in overall alpha diversity between T1 and T2 samples, male only T1 and T2 samples, and female only T1 and T2 samples. In GraphPad Prism v10.2.3, we conducted the Shapiro-Wilk test for normality and then both applied and visualized either a two-tailed paired t-test or Wilcoxon test depending on the normality results. We ran tests for homogeneity of group variances and univariate and multivariate PERMANOVA analyses on batch, sex, county of stranding, and a new covariate called time point. To test for differential abundances between time points, we applied ANCOM-BC on a class and genus level. We then analyzed T1 and T2 samples separately. For each time point we created an alpha-rarefaction curve and calculated alpha and beta diversity metrics at a rarefaction depth of 4678 (T1) or 5808 (T2) to maximize diversity (Fig. S3). We were interested in whether there was significant difference in alpha diversity between sexes within each time point. In GraphPad Prism v10.2.3, we conducted the Shapiro-Wilk test for normality and conducted Welch’s t-test. We then tested for homogeneity of group variances, and univariate and multivariate PERMANOVA analyses for batch, sex, and county of stranding. Restriction-site associated DNA sequencing and bioinformatic processing We used blood samples stored in EDTA from the seals included in our admissions and longitudinal dataset to understand the role of host genetics in shaping the gut microbiome. We extracted genomic DNA for restriction-site associated DNA sequencing (RADseq-capture; Ali et al . 2016), which were first digested with the SbfI restriction enzyme. We then ligated a unique 8-bp barcoded biotinylated adapter to the fragmented DNA that allowed us to pool equal amounts of up to 48 samples. We sheared the pools in a Covaris LE220 to 400bp fragments, which we then enriched for fragments that contained the adapter using a Dynabeads M-280 streptavidin binding assay. Once enriched, we prepared the pools for Illumina NovaSeq paired-end (2x150nt) sequencing at Princeton University’s LSI core facility using the NEBnext Ultra II DNA Library Prep Kit and used Agencourt AMPure XP magnetic beads for any library purification or size selection step. We retained raw sequences where the read (and its pair) contained the expected unique barcode and the remnant SbfI recognition motif using the process_radtags module in STACKS v2 (Catchen et al . 2013; Rochette et al . 2019) and allowed up to a 2-bp mismatch and had a quality score ≥10. We next used the clone_filter module to remove PCR duplicates prior to mapping to the NES reference genome (NCBI assembly: ASM2128878v3). We excluded mapped reads with MAPQ<20 from all further processing, and we converted the SAM files to BAM format in Samtools v0.1.18 (Li et al. 2009). Population genetic structure analysis We implemented the gstacks and populations modules in STACKS v2 following the recommended pipeline for data mapped to a reference genome and constructing a catalog with all polymorphic sites. We further increased the stringency of SNP annotation by using the marukilow model flags –vt-alpha and –gt-alpha with p =0.01. We retained all SNPs discovered per locus and used VCFtools v0.1.17 (Danecek et al. 2011) for filtering out singleton and private doubleton alleles, and to remove individuals with more than 30% missing data. We further filtered to exclude loci with a minor allele frequency (MAF<0.03) while allowing up to 20% missingness rate per locus (–geno 0.2) in PLINK v1.90b3i (Chang et al. 2015). We further filtered for linkage disequilibrium (LD) and Hardy-Weinberg Equilibrium (HWE) to obtain a set of SNPs that were considered statistically unlinked and neutral. We used PLINK’s genotype correlation function to remove sites within a 50-SNP window whose genotypes were highly correlated (r 2 >0.5; –indep-pairwise 50 5 0.5) and excluded sites that significantly deviated from HWE (–hwe 0.001). This was the pruned SNP set used for all downstream genetic analyses. We conducted an unsupervised, non-model based principal component analysis (PCA) with the program FlashPCA v2.1 (Abraham et al. 2017) and PHATE (Moon et al. 2019) to assess the impact of geography and life history on genetic similarity. We then completed an unsupervised, maximum likelihood cluster analysis in the program ADMIXTURE (Alexander et al. 2009) to assess the likelihood at each genetic partition from K =2-10. Host genetic distance and microbiome dissimilarity analysis To test for associations between host genetic distance and microbiome dissimilarity, we used a Mantel test following the pipeline of DeCandia et al . (2021). We calculated euclidean distance between each pair of samples using the pruned SNP set with the dist function in the R package adegenet (Jombart 2008; Jombart & Ahmed 2011). We then used the R package vegan (Oksanen et al. 2019) to implement a Mantel test on the matrices of genetic distance and Bray-Curtis dissimilarity of gut microbiomes. We assessed the correlation with Spearman’s rank correlation coefficient ( ρ ) and a statistical significance threshold of p <0.05. We performed the test for genetic distance on admission samples as well as longitudinal samples (T1 and T2). Results We sequenced 179 rectal samples from 89 NES pups. We found no evidence of plate-wide contamination as per assessment of the positive and negative controls. After applying the inclusion criteria, we retained 44 unique NES individuals in the admission dataset, originating from six California counties, with samples that contained 1,235,596 sequences and 352 ASVs (Tables S2, S3). We further subsetted 19 NES individuals as part of the longitudinal dataset, where each individual had a sample collected at admission and approximately 1-month post-admission (Table S2). The admission dataset represents NES that were primarily admitted for malnourishment and did not receive additional treatments that could have altered the gut microbiome or were known to have died due to unnatural or natural causes. Effect of sex and geography on initial gut microbiome diversity While batch was our only significant factor affecting alpha diversity (Table 1), we found multiple signatures of weak correlations of Bray-Curtis (Fig. 3), Unweighted (Fig. S4A), and Weighted (Fig. S4B) Unifrac PCoA axes with batch and sex. We identified significant factors that explained variation in microbial composition using univariate and multivariate PERMANOVA analyses. Bray-Curtis dissimilarity (Table 2) and Unweighted Unifrac (Table S4) univariate and multivariate analyses showed that batch and sex were significant. With Weighted Unifrac, batch and sex were significant in the univariate analysis only (PERMANOVA; p =0.011; Table S5). We found differential abundance between sexes (Fig. 4). On a class level, males had significantly more Fusobacteriia but less Erysipelotrichi compared to females (ANCOM-BC; q =0.0177 and q =6.64E-12, respectively; Table S6). On a genus level, males had significantly more SMB53 (ANCOM-BC; q =0.0261) and significantly less of four genera (Table S6). Bray-Curtis dissimilarity also showed that county was significant for both univariate and multivariate analyses. The Bray-Curtis dissimilarity pairwise comparisons showed San Mateo was significantly different from two other counties (San Luis Obipso: q =0.045; Santa Cruz: q =0.0.045; Table S7) and nearly significantly different from a third (Monterey: q =0.060; Table S6). We did not find county of stranding to be a significant factor in the Unweighted and Weighted Unifrac PERMANOVA analyses. Homogeneity of group variances was met for all tests ( p >0.05), except for batch and county of stranding for Unweighted Unifrac (Tables 2, S4, S5). Our significant Unweighted Unifrac univariate PERMANOVA results for batch and county of stranding may be due to different dispersions between groups. Otherwise, all other PERMANOVA results were driven by differences in mean values across groups rather than differences in group variances. Population genetic analysis reveals no geographic structure To investigate if the effect from geographic origin could be driven by population genetic differentiation among northern elephant seal colonies, we constructed a SNP catalog of 164,850 loci across 17 genome scaffolds of 27 NES pups with an average of 27.4 (s.d.=14.7) depth of sequence coverage (Table S8). We retained 6,520 variants meeting all filtering thresholds. We found no genetic clustering as a function of batch, county of stranding, or by individual sex (Fig. 5). We found further support that the stranding location and sex were not crucial in driving genetic patterns with the model-based clustering (Fig. 5C). The influence of sex on gut microbiome decreases in pups rehabilitated in a “common garden” A subset of 19 NES individuals were included in the longitudinal dataset, where time between T1 and T2 sample collection varied (median: 36 days; range 25-67 days) (Fig. 2, Table S2). We first conducted analyses where T1 and T2 samples were combined. After creating alpha rarefaction curves (Figure S2), we did not find a significant increase in alpha diversity between T1 and T2 when considering all samples, males only, or females only (Table S9A, Fig. S5). Consistent across both the univariate and multivariate PERMANOVA analyses, all three beta diversity metrics yielded significant results for batch and time point (Tables 2, S4, S5) and a principal coordinate analysis (PCoA) of the Bray-Curtis dissimilarity matrix clearly showed changes over time (Fig. 6A). In addition, Bray-Curtis dissimilarity significantly decreased with time (Fig. 6B). Bray-Curtis dissimilarity and Unweighted Unifrac also yielded significant results for sex for both the univariate and multivariate analyses (Tables 2, S4). Homogeneity of group variances was met for all tests ( p >0.05), except for time point for Bray-Curtis dissimilarity and batch for Unweighted Unifrac. We recreated rarefaction curves per time point (Fig. S3) and showed that all datasets met the assumption of normality (Table S9B). At both T1 and T2, there were significant results for the batch for our Bray-Curtis dissimilarity and Unweighted Unifrac PERMANOVA univariate and multivariate analyses (Tables 2, S4). T1 showed significant results for sex using our Bray-Curtis dissimilarity univariate and multivariate analyses (Table 2). T2 showed marginally significant results for sex for our univariate ( p =0.063) and multivariate ( p =0.053) analyses (Table 2). Homogeneity of group variances was met for all tests, except for county of stranding for Bray-Curtis dissimilarity T2 and sequencing batch for Unweighted Unifrac T1 and T2 (Tables 2, S4, S5). Our differential abundance analysis on a class level revealed that compared to T1, T2 had significantly less Clostridia, Gammaproteobacteria, and Bacilli (ANCOM-BC; Fig. 4B, Table S6). On a genus level, there were 5 genera that were found in significantly greater abundance at T2 compared to T1 and 5 genera that were found in significantly less abundance at T2 compared to T1 (Fig. 4B, Table S6). Host genetic distance does not predict microbiome diversity in any environment We calculated the pairwise genetic distance for 27 pups for which we had genetic data and found no significant correlation with microbiome dissimilarity (Mantel test; ρ =-0.075, p =0.162; Fig. S6). We then calculated the genetic distance of 13 pups from the longitudinal sample set and again found that genetic distance did not correlate with microbiome dissimilarity either at admit ( ρ =-0.047, p =0.684; Fig. 6C) or after a month of rehabilitation ( ρ =0.078, p =0.496; Fig. 6D). Discussion We investigated microbiome diversity and composition at the end of weaning period (~3 months of age) for stranded northern elephant seal pups, after their monthslong fast ends and independent foraging beyond their natal rookeries begins. We surveyed the influence of sex, county of stranding, and host genetics on the gut microbiomes of NES pups at multiple time points. We discovered that shortly after admission, sex was a significant factor that shaped variation in microbial composition but not diversity. The only other study on NES gut microbiomes similarly found sex to be a determinant of composition but not diversity (Stoffel et al. 2020). Despite these similar findings, a key distinction is that previous research was conducted with younger, fasting seal pups around 1-2 months of age at the beginning to middle of the post-weaning period, while our cohort was estimated to be around 3 months of age and had initiated their life at sea. Thus, our study supports emerging evidence that sex influences the gut microbiome early in development, before extreme morphological dimorphism exists. Our result adds that this influence is evident even when young NES pups transition to the independent foraging life stage. In studies with humans and mice, microbial impacts of intrinsic factors like sex have been overshadowed by environmental factors, such as diet (Valeri & Endres 2021). Previous research was able to take advantage of the post-weaning period to study the influence of sex as NES remained in their natal rookeries while fasting completely, thereby minimizing microbial variation due to habitat and diet (Stoffel et al. 2020). The seals in our study, however, presumably originate from several different natal rookeries and may have also begun to forage in new marine locations prior to capture. We were able to minimize rehabilitation environmental variables by only including seals that were swabbed within 10 days of admission and did not receive pharmacologic treatments. Yet, even without minimizing environmental noise to the same extent as previous research, we were able to detect sex’s influence on the gut microbiome shortly after admissions. After a month of rehabilitation, sex appeared to be a less significant influence on microbiome composition than in the set of samples collected at admission to TMMC, as shown by the lower p -value at the second time point. We suspect that environmental variables within the rehabilitation facility, such as shared pools and common, single item diets, increased the similarity of microbiomes across all individuals, thereby decreasing the influence of sex. Furthermore, the rehabilitation environment likely minimized the impact of sex-specific behavioral or ecological differences on our observed gut microbiome dimorphism. A study in wild adult southern elephant seals (SES), a closely related species that also exhibits extreme sexual dimorphism, found striking sex differences in the gut microbiome that were speculated to be due to sex-specific foraging strategies and resource partitioning (Kim et. al 2020; Lewis et. al 2006). Because our seals were in an enclosed rehabilitation environment, they were unable to engage in sexually dimorphic behaviors, reducing the likelihood that sex-specific environmental and dietary factors influenced the gut microbiome. Furthermore, our rehabilitated seals were housed in shared pens and fed the same diet, thus creating a “common garden” scenario for these seals in which environmental variables are shared across all individuals. Studies on harbor seal ( Phoca vitulina ) pups in other rehabilitation facilities (Rubio-Garcia et. al 2023; Switzer et. al 2023) have found sex to shape microbiome variation, despite smaller degrees of sexual dimorphism in that species. The seals in these two previous studies were housed in similar environments and had a shared diet, with one study even being conducted in the same rehabilitation facility (TMMC) as our study (Switzer et. al 2023). On the other hand, a study of wild pup and adult harbor seals failed to find sex as a significant factor in harbor seal gut microbiome composition (Pacheco-Sandoval et. al 2022). Thus, the more controlled environment of rehabilitation facilities may allow sex-driven differences to emerge in some species. Still, each rehabilitation facility has unique exposures ( e.g. human handling, artificial and varied fish diets, disinfection protocols, etc) and are therefore less controlled than laboratory settings. In addition to understanding intrinsic sources of microbial variation, like host sex, we were also interested in understanding extrinsic sources of microbial variation. In this study, seals several hundred kilometers apart across six counties along the California coast were brought to the same rehabilitation facility, which offered us a unique opportunity to explore this understudied variable. Notably, previous research on the microbiomes of wild and weaned NES only examined animals on a single beach in Baja California, Mexico, and therefore could not explore how geography affected microbiome development (Stoffel et al. 2020). Gut microbiome studies in rehabilitated pinnipeds with a range of geographic origins also have not investigated the influence of geography (Rubio-Garcia et. al 2023; Switzer et. al 2023). By identifying individual differences between San Mateo County and both Santa Cruz and San Luis Obispo counties, we present the first findings that geographic origin is a significant determinant of microbial composition but not diversity in NES microbiomes. This result could have resulted from genetic differences between origin populations. If certain alleles or combinations of alleles influenced microbiome composition in NES, population genetic structuring along the California coast could result in distinctive microbiomes for seals of different colonies. However, our population genetic analyses revealed no significant population structure among our samples, despite animals originating from a several hundred-kilometer span of coastal California. This pattern suggests that little population genetic structure exists along the NES rookeries of California, which is consistent with both the high dispersal estimates in this species (Condit et al . 2022) and a prior study using microsatellites that found extremely low F ST estimates between rookeries in northern and southern California (Abadía-Cardoso et al. 2017). The lack of genetic structure in our dataset suggests that the influence of county of stranding is related to differential exposure originating from local environmental microbial communities (and more generally habitats) and not from the underlying genetics of the host. While previous research found that microbial communities in marine mammal hosts and their local habitat were largely distinct, there were still shared microbes, suggesting that these aquatic hosts exchanged bacteria with their environment to some extent (Bik et. al 2016; Tian et. al 2022). Further NES studies that sample microbial communities from both the surrounding terrestrial and marine environments are needed to elucidate the impacts of natural habitats upon NES microbiota. Our longitudinal studies of individuals through rehabilitation allowed us to examine several other aspects of microbiome change. After one month at TMMC, we found changes in NES gut microbiome taxonomic composition on the class and genus levels but no overall increase in diversity. Our diversity results were surprising given that dietary diversity is correlated with microbial diversity (Xiao et. al 2022; Heiman & Greenman 2017) and our seals experienced a dietary transition during their rehabilitation, from a monthslong fast with possible but minimal foraging, to mashed fish gruel with salmon oil, to whole fish. Lack of temporal variation in alpha diversity was also found in younger NES, though the seals studied were strictly fasting (Stoffel et. al 2020). The changes we observed while seals were in a rehabilitation setting may alternatively be attributed to age, as speculated in a study where only non-fasting harbor seal pups, and not weaners, saw a change in alpha diversity (Rubio-Garcia et. al 2023). Further study is required to disentangle effects of age and rehabilitation setting including facility and diet upon the gut microbiota of NES. In any case, after one month of rehabilitation, we found that the gut microbiomes of NES pups had significantly lower Bray-Curtis dissimilarity compared to when the same pups were sampled at admit, suggesting a convergence in microbiome similarity. At neither admission nor after a month of rehabilitation, however, did host genetics appear to influence microbiome composition. Studies in other species have found that genotype influences gut microbiome composition (Bonder et al. 2016; Suzuki et al. 2019), contrary to our findings in NES. The lack of host genetic signal in our analyses could be due to our relatively modest sample size or to the very low genetic variation in NES due to a history of a severe population bottleneck (Hoelzel et al. 2002). On the other hand, this low genetic diversity may in part be what allows us to identify signals of non-genetic factors influencing the NES gut microbiome. We found sequencing batch to be a consistent significant factor in our alpha and beta diversity analyses and also potentially confounded with sex. We speculate that these results are a reflection of a common issue with microbiome studies called batch effects, which are technical, non-biological factors that may affect variation in data due to how different “batches” of samples are processed (Wang & Lê Cao 2020). We took various measures to minimize batch effects and these potentially confounding variables through experimental design and computational approaches. We randomized our rectal swab samples across two 96 well plates, which were then prepared and sequenced at the same time. We nonetheless suspected that batch ( i.e ., plate) and sex may be confounded because in comparison to batch 1, batch 2 had a lower median number of sequences and was primarily composed of female samples. To account for this, we created an alpha-rarefaction curve and selected a rarefaction depth that would retain maximum diversity. This was a threshold that retained all samples, even those on batch 2, and was used when calculating alpha and beta diversity metrics. With our computational analyses, we also added batch as a covariate to be controlled for in all our beta diversity multivariate models. Overall, we found a significant signal of sex’s impact on northern elephant seal gut microbiomes in spite of batch effects and highlight the importance of accounting for batch effects throughout rRNA sequencing and subsequent bioinformatic pipelines (Regueira-Iglesias et al. 2023, Reitmeier et al. 2021, Wang & LêCao 2020). In summary, we studied the gut microbiome of pups in a species in which adults show extreme sexual dimorphism. Our study included weaned, marine-environment-staged NES pups that were found stranded in multiple northern and central California counties and brought to a common hospital environment for rehabilitation. We found that the county of stranding was a driver for differences in microbial variation that could not be attributed to host genetic influence. Although pups of this age exhibited minimal phenotypic and anatomical dimorphism, we found sex was a consistent and significant factor that contributes to microbial variation shortly after admissions and, to a lesser degree, following rehabilitation in a common environment. Our findings ultimately help shape our understanding of the extrinsic and intrinsic factors that shape the gut microbiomes of NES during an understudied period of development. Code Accessibility QIIME2 and R code used in this study is publicly available through Github under https://github.com/sjgaughran/NES_TMMC_microbiome. Data Accessibility 16S sequencing data and RADseq data analyzed are publicly available through the NCBI Sequence Read Archive under BioProject PRJNA1007377 and BioProject PRJNA1007380, respectively. Benefit-Sharing Benefits from this research accrue from the sharing of our data in the NCBI databases above. Author Contributions ALG, BVH, EY, and SJG secured research funding. CF and EW cared for the patients, collected rectal samples, and provided metadata. ALD, BVH, EY, and SJG designed the study, carried out experiments, and performed statistical analyses and interpretations. BVH, EY, SJG wrote the manuscript. ALD, ALG, BVH, CLF, ERF, EY, SJG revised the manuscript. Authorizations All response, care and sampling activities of stranded northern elephant seals were authorized under National Marine Fisheries Services Permit No. 24359. References Ali, O.A., O’Rourke, S.M., Amish, S.J., Meek, M.H., Luikart, G., Jeffres, C., & Miller, M.R. (2016). RAD capture (Rapture): flexible and efficient sequence-based genotyping. Genetics , 202(2), 389-400. https://doi.org/10.1534/genetics.115.183665 Abadía-Cardoso, A., Freimer, N. B., Deiner, K., & Garza, J. C. (2017). Molecular population genetics of the northern elephant seal Mirounga angustirostris . Journal of Heredity , 108(6), 618-627. https://doi.org/10.1093/jhered/esx053 Abraham, G., Qiu, Y., & Inouye, M. (2017). FlashPCA2: principal component analysis of biobank-scale genotype datasets. Bioinformatics, 33(17), 2776-2778. https://doi.org/10.1093/bioinformatics/btx299 Afghan, E., Baker, D., Batut, B., Van Den Beek, M., Bouvier, D., Ech, M., … Blankenberg, D. (2018). The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2018 update. Nucleic Acids Research , 46(W1), W537-W544. https://doi.org/10.1093/nar/gky379 Alexander, D., Novembre, J., Lange, K. (2009) Fast model-based estimation of ancestry in unrelated individuals. Genome Research , 19(9), 1655-1664. https://doi.org/10.1101/gr.094052.109 Anderson, M. J. (2001). A new method for non-parametric multivariate analysis of variance: NON-PARAMETRIC MANOVA FOR ECOLOGY. Austral Ecology , 26(1), 32-46. https://doi.org/10.1111/j.1442-9993.2001.01070.pp.x Belkaid, Y., & Hand, T.W. (2014). Role of the microbiota in immunity and inflammation. Cell , 157(1). https://doi.org/10.1016/j.cell.2014.03.011 Beltran, R.S., Hindell, M.A., & McMahon, C.R. (2022). The Elephant Seal: Linking Phenotypic Variation with Behavior and Fitness in a Sexually Dimorphic Phocid. In Ethology and Behavioral Ecology of Phocids (pp. 401-440). Cham: Springer International Publishing. Bennett, G., Malone, M., Sauther, M.L., Cuozzo, F.P., White, B., Nelson, K.E., … & Amato, K. R. (2016). Host age, social group, and habitat type influence the gut microbiota of wild ring‐tailed lemurs ( Lemur catta ). American Journal of Primatology , 78(8), 883-892. https://doi.org/10.1002/ajp.22555 Bik, E.M., Costello, E.K., Switzer, A D., Callahan, B.J., Holmes, S.P., Wells, R. S., …, & Relman, D. A. (2016). Marine mammals harbor unique microbiotas shaped by and yet distinct from the sea. Nature Communications , 7(1), Article 10516. https://doi.org/10.1038/ncomms10516 Bisanz, J.E. (2018). qiime2R: Importing QIIME2 artifacts and associated data into R sessions. https://github.com/jbisanz/qiime2R. Bokulich, N.A., Kaehler, B.D., Rideout, J.R., Dillon, M., Bolyen, E., Knight, R., Huttley, G. A., & Caporaso, J.G. (2018). Optimizing taxonomic classification of marker-gene amplicon sequences with QIIME 2’s q2-feature-classifier plugin. Microbiome , 6(90). https://doi.org/10.1186/s40168-018-0470-z Bolnick, D. I., Snowberg, L.K., Hirsch, P.E., Lauber, C.L., Org, E., Parks, B., … & Svanbäck, R. (2014). Individual diet has sex-dependent effects on vertebrate gut microbiota. Nature Communications , 5(1), Article 4500. https://doi.org/10.1038/ncomms5500 Bolyen, E., Rideout, J.R., Dillon, M.R., Bokulich, N.A., Abnet, C.C., Al-Ghalith, G.A., … Caporaso, J.G. (2019). Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nature Biotechnology , 37(8), 852-857. https://doi.org/10.1038/s41587-019-0209-9 Bonder, M.J., Kurilshikov, A., Tigchelaar, E.F., Mujagic, Z., Imhann, F., Vila, A.V., … Zhernakova, A. (2016). The effect of host genetics on the gut microbiome. Nature Genetics , 48(11), 1407-1412. https://doi.org/10.1038/ng.3663 Callahan, B.J., McMurdie, P.J., & Holmes, S.P. (2017). Exact sequence variants should replace operational taxonomic units in marker-gene data analysis. The ISME Journal , 11(12), 2639-2643. https://doi.org/10.1038/ismej.2017.119 Callahan, B. J., McMurdie, P. J., Rosen, M. J., Han, A. W., Johnson, A. J. A., & Holmes, S. P. (2016). DADA2: High-resolution sample inference from Illumina amplicon data. Nature Methods, 13(7), 581–583. https://doi.org/10.1038/nmeth.3869 Caporaso, J.G., Lauber, C.L., Walters, W.A., Berg-Lyons, D., Lozupone, C.A., Turnbaugh, P.J., Fierer, N., & Knight, R. (2011). Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample. Proceedings of the National Academy of Sciences , 108(supplement_1), 4516-4522. https://doi.org/10.1073/pnas.1000080107 Catchen, J., Hohenlohe, P.A., Bassham, S., Amores, A., & Cresko, W.A. (2013). Stacks: an analysis tool set for population genomics. Molecular Ecology , 22, 3124-3140. https://doi.org/10.1111/mec.12354 Chang, C.C., Chow, C.C., Tellier, L.C., Vattikuti, S., Purcell, S.M., & Lee, J.J. (2015). Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience , 4(1). https://doi.org/10.1186/s13742-015-0047-8 Condit, R., Beltran, R. S., Robinson, P. W., Crocker, D. E., & Costa, D. P. (2022). Birth timing after the long feeding migration in northern elephant seals. Marine Mammal Science , 38(3), 931-940. https://doi.org/10.1111/mms.12896 Danecek, P., Auton, A., Abecasis, G., Albers, C.A., Banks, E., DePristo, M.A., … Sherry, S.T. (2011). The variant call format and VCFtools. Bioinformatics , 27, 2156-2158. https://doi.org/10.1093/bioinformatics/btr330 DeCandia, A., Cassidy, K., Stahler, D., Stahler, E., & vonHoldt, B. (2021) Social environment and genetics underlie body site specific microbiomes of Yellowstone National Park gray wolves ( Canis lupus ). Ecology and Evolution , 11(14), 9472-9488. https://doi.org/10.1002/ece3.7767 DeCandia, A., Brenner, L., King, J., & vonHoldt, B. (2020) Ear mite infection is associated with altered microbial communities in genetically depauperate Santa Catalina Island foxes ( Urocyon littoralis catalinae ). Molecular Ecology , 29(8), 1463-1475. https://doi.org/10.1111/mec.15325 DeCandia, A.L., Leverett, K.N., & vonHoldt, B.M. (2019). Of Microbes and Mange: Consistent changes in the skin microbiome of three canid species infected with sarcoptic mange. Parasites & Vectors , 12, Article 488. https://doi.org/10.1186/s13071-019-3724-0 Deegan, D.F., & Engel, N. (2019) Sexual Dimorphism in the Age of Genomics: How, When, Where. Frontiers in Cell and Developmental Biology , 7, Article 186. https://doi.org/10.3389/fcell.2019.00186 DeSantis, T. Z., Hugenholtz, P., Larsen, N., Rojas, M., Brodie, E. L., Keller, K., … Andersen, G. L. (2006). Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Applied and Environmental Microbiology , 72(7), 5069–5072. https://doi.org/10.1128/AEM.03006-05 Deutsch, C.J., Crocker, D.E., Costa, D. P., & Le Boeuf, B. J. (1994). Sex and age-related variation in reproductive effort of northern elephant seals. In B. J. Le Boeuf, & R. M. Laws (Eds.), Elephant seals: Population ecology, behavior, and physiology (pp. 169–210). Berkeley, CA: University of California Press. Fowler, K., & Whitlock, M.C. (1999). The distribution of phenotypic variance with inbreeding. Evolution, 53(4), 1143-1156. https://doi.org/10.1111/j.1558-5646.1999.tb04528.x Hasan, N., & Yang, H. (2019). Factors affecting the composition of the gut microbiota, and its modulation. PeerJ , 7, e7502. https://doi.org/10.7717/peerj.7502 Heiman, M.L., & Greenway, F.L. (2016). A healthy gastrointestinal microbiome is dependent on dietary diversity. Molecular Metabolism , 5(5), 317–320. https://doi.org/10.1016/j.molmet.2016.02.005 Hoelzel, A.R., Fleischer, R.C., Campagna, C., Le Boeuf, B.J. & Alvord, G. (2002). Impact of a population bottleneck on symmetry and genetic diversity in the northern elephant seal: Impact of elephant seal bottleneck. Journal of Evolutionary Biology , 15, 567–575. https://doi.org/10.1046/j.1420-9101.2002.00419.x Hunnewell, Bill (n.d.). Vapor [Photography]. The Marine Mammal Center. https://www.marinemammalcenter.org/patients/es-4686 Jandhyala, S.M., Talukdar, R., Subramanyam, C., Vuyyuru, H., Sasikala, M., & Nageshwar, R.D. (2015). Role of the normal gut microbiota. World Journal of Gastroenterology , 21(29), 8787-8803. https://doi.org/10.3748/wjg.v21.i29.8787 Jelincic, J.A., Tift, M.S., Houser, D.S., & Crocker, D.E. (2017). Variation in adrenal and thyroid hormones with life-history stage in juvenile northern elephant seals ( Mirounga angustirostris ). General and Comparative Endocrinology , 252, 111-118. https://doi.org/10.1016/j.ygcen.2017.08.001 Jombart, T., & Ahmed, I. (2011). adegenet 1.3-1: New tools for the analysis of genome-wide SNP data. Bioinformatics, 27(21), 3070–3071. https://doi.org/10.1093/bioinformatics/btr521 Jombart, T. (2008). adegenet: A R package for the multivariate analysis of genetic markers. Bioinformatics , 24(11), 1403–1405. https://doi.org/10.1093/bioinformatics/btn129 de Jonge, N., Carlsen, B., Christensen, M.H., Pertoldi, C., & Nielsen, J.L. (2022). The Gut Microbiome of 54 Mammalian Species. Frontiers in Microbiology , 13, Article 886252. https://doi.org/10.3389/fmicb.2022.886252 Katoh, K., & Standley, D.M., (2013). MAFFT Multiple Sequence Alignment Software Version 7: Improvements in Performance and Usability, Molecular Biology and Evolution , 30(4), 772-780. https://doi.org/10.1093/molbev/mst010 Kelso, E.J., Champagne, C.D., Tift, M.S., Houser, D.S., & Crocker, D.E. (2012). Sex differences in fuel use and metabolism during development in fasting juvenile northern elephant seals. Journal of Experimental Biology, 215(15), 2637-2645. https://doi.org/10.1242/jeb.068833 Kienle, S.S., Friedlaender, A.S., Crocker, D.E., Mehta, R.S., & Costa, D.P. (2022). Trade-offs between foraging reward and mortality risk drive sex-specific foraging strategies in sexually dimorphic northern elephant seals. The Royal Society, 9, Article 210522. http ://doi.org/10.1098/rsos.210522Kim, M., Cho, H., & Lee, W. Y. (2020). Distinct gut microbiotas between southern elephant seals and Weddell seals of Antarctica. Journal of Microbiology , 58, 1018-1026. https://doi.org/10.1007/s12275-020-0524-3 Kinross, J., Darzi, A.W., & Nicholson. J. (2011). Gut microbiome-host interactions in health and disease. Genome Medicine, 3, Article 14. https://doi.org/10.1186/gm228 Kruskal, W.H., & Wallis, W.A. (1952). Use of Ranks in One-Criterion Variance Analysis. Journal of the American Statistical Association , 47(260), 583–621. https://doi.org/10.1080/01621459.1952.10483441 Kurilshikov, A., Wijmenga, C., Fu, J., & Zhernakova, A. (2017). Host genetics and gut microbiome: challenges and perspectives. Trends in immunology , 38(9), 633-647. https://doi.org/10.1016/j.it.2017.06.003 Le Boeuf, B., Condit, R., & Reiter, J. (2019). Lifetime reproductive success of northern elephant seals (Mirounga angustirostris). Canadian Journal of Zoology , 97(12), 1203–1217. https://doi.org/10.1139/cjz-2019-0104 Le Boeuf, B.J., Crocker, D.E., Costa, D.P., Blackwell, S. B., Webb, P. M., & Houser, D. S. (2000). Foraging Ecology of Northern Elephant Seals. Ecological Monographs, 70(3), 353–382. https://doi.org/10.1890/0012-9615(2000)070\%5B0353:FEONES\%5D2.0.CO;2 Le Boeuf, B.J., & Laws, R.M. (Eds.). (1994). Elephant seals: population ecology, behavior, and physiology. Univ of California Press. Lewis, R., O’Connell, T. C., Lewis, M., Campagna, C., & Hoelzel, A. R. (2006). Sex-specific foraging strategies and resource partitioning in the southern elephant seal (Mirounga leonina). Proceedings of the Royal Society B: Biological Sciences , 273(1603), 2901–2907. https://doi.org/10.1098/rspb.2006.3642 Li, H., Handsaker, B., Wysoker, A., Fennell, T., Ruan, J., Homer, N., … Durbin, R. (2009). The sequence alignment/map format and SAMtools. Bioinformatics , 25, 2078-2079. https://doi.org/10.1093/bioinformatics/btp352 Lin, H., & Peddada, S.D. (2020). Analysis of compositions of microbiomes with bias correction. Nature Communications , 11(1), 3514. https://doi.org/10.1038/s41467-020-17041-7 Lu, J., Hamblen, E., Brenner, L., King, J., vonHoldt, B., & DeCandia A. (2023). Ear mite infection restructures otic microbial networks in conservation-reliant Santa Catalina Island foxes ( Urocyon littoralis catalinae ). Molecular Ecology . 32(4):892-903. https://doi.org/10.1111/mec.16795 Mandal, S., Van Treuren, W., White, R.A., Eggesbø, M., Knight, R., & Peddada, S.D. (2015). Analysis of composition of microbiomes: A novel method for studying microbial composition. Microbial Ecology in Health & Disease , 26(0). https://doi.org/10.3402/mehd.v26.27663 Moon, K.R., van Dijk, D., Wang, Z., Gigante, S., Burkhardt. D.B., Chen, W.S., Yim, K., van den Elzen, A., Hirn, M.J., Coifman, R.R., Ivanova, N.B., Wolf, G., & Krishnaswamy, S. (2019). Visualizing structure and transitions in high-dimensional biological data. Nat Biotechnol 37, 1482–1492. https://doi.org/10.1038/s41587-019-0336-3 Maurice, C.F., CL Knowles, S., Ladau, J., Pollard, K.S., Fenton, A., Pedersen, A.B., & Turnbaugh, P. J. (2015). Marked seasonal variation in the wild mouse gut microbiota. The ISME journal , 9(11), 2423-2434. https://doi.org/10.1038/ismej.2015.53 Nelson, T.M., Rogers, T.L., & Brown, M.V. (2013). The Gut Bacterial Community of Mammals from Marine and Terrestrial Habitats. PLoS ONE , 8(12), e83655. https://doi.org/10.1371/journal.pone.0083655 Oksanen, J., Blanchet, F.G., Friendly, M., Kindt, R., Legendre, P., …, & Wagner, H. (2019). vegan: Community ecology package (R package version 2.5-6). https://cran.r-project.org/package=vegan Pacheco-Sandoval, A., Lago-Lestón, A., Abadía-Cardoso, A., Solana-Arellano, E., & Schramm, Y. (2022). Age as a primary driver of the gut microbial composition and function in wild harbor seals. Scientific Reports, 12(1), 14641. https://doi.org/10.1038/s41598-022-18565-2 Pacheco-Sandoval, A., Schramm, Y., Heckle, G., Brassea-Perez, G., Martinez-Porchas, M., & Lago-Leston, A. (2019). The Pacific harbor seal gut microbiota in Mexico: Its relationship with diet and functional inferences. PLOS ONE , 14, 1-21. https://doi.org/10.1371/journal.pone.0221770 Pafčo, B., Sharma, A.K., Petrželková, K.J., Vlčková, K., Todd, A., Yeoman, C.J., … & Gomez, A. (2019). Gut microbiome composition of wild western lowland gorillas is associated with individual age and sex factors. American Journal of Physical Anthropology , 169(3), 575-585. https://doi.org/10.1002/ajpa.23842 Park, J.C., & Im, S.H. (2020). Of men in mice: the development and application of a humanized gnotobiotic mouse model for microbiome therapeutics. Experimental & Molecular Medicine , 52(9), 1383–1396. https://doi.org/10.1038/s12276-020-0473-2 Price, M.N., Dehal, P.S., & Arkin, A.P. (2009). FastTree: Computing Large Minimum Evolution Trees with Profiles instead of a Distance Matrix, Molecular Biology and Evolution , 26(7), 1641–1650. https://doi.org/10.1093/molbev/msp077 Regueira-Iglesias, A., Suárez-Rodríguez, B., Blanco-Pintos, T., Relvas, M., Alonso-Sampedro, M., Balsa-Castro, C., & Tomás, I. (2024). The salivary microbiome as a diagnostic biomarker of periodontitis: a 16S multi-batch study before and after the removal of batch effects. Frontiers in Cellular and Infection Microbiology, 14, 1405699. Reiter, J., Panken, K.J., & Le Boeuf, B.J. (1981). Female competition and reproductive success in northern elephant seals. Animal Behaviour , 29(3), 670–687. https://doi.org/10.1016/S0003-3472(81)80002-4 Reitmeier, S., Hitch, T. C., Treichel, N., Fikas, N., Hausmann, B., Ramer-Tait, A. E., … & Clavel, T. (2021). Handling of spurious sequences affects the outcome of high-throughput 16S rRNA gene amplicon profiling. ISME communications, 1(1), 31. Rochette, N.C., Rivera‐Colón, A.G., & Catchen, J.M. (2019). Stacks 2: Analytical methods for paired‐end sequencing improve RADseq‐based population genomics. Molecular Ecology , 28, 4737-4754. https://doi.org/10.1111/mec.15253 Rojas, C.A., Ramírez-Barahona, S.A., Holekamp, K.E., & Theis, Kevin. R. (2020). Host phylogeny and host ecology structure the mammalian gut microbiota at different taxonomic scales. Ecology , 3, Article 33. https://doi.org/10.1101/2020.09.23.299727 Rubio-Garcia, A., Zomer, A. L., Guo, R., Rossen, J. W. A., Van Zeijl, J. H., Wagenaar, J. A., & Luiken, R. E. C. (2023). Characterising the gut microbiome of stranded harbour seals (Phoca vitulina) in rehabilitation. PLOS ONE , 18(12), e0295072. https://doi.org/10.1371/journal.pone.0295072 Satyanarayan, A., Russell, R., Hoffswell, J., & Heer, J. (2016). Reactive Vega: A Streaming Dataflow Architecture for Declarative Interactive Visualization. IEEE Transactions on Visualization and Computer Graphics , 22(1), 659–668. https://doi.org/10.1109/TVCG.2015.2467091 Sekido, R. & Lovell-Badge, R. (2009). Sex determination and SRY: down to a wink and a nudge? Trends in Genetics , 25, 19-29. https://doi.org/10.1016/j.tig.2008.10.008 Shreiner, A.B., Yao, J.Y., & Young, V.B. (2015). The gut microbiome in health and disease. Current Opinion in Gastroenterology, 31(1), 69-75. https://doi.org/10.1097/MOG.0000000000000139 Slatkin, M. (1984) Ecological causes of sexual dimorphism. Evolution, 38(3), 622-630. https://doi.org/10.2307/2408711 Spor, A., Koren, O., & Ley, R. (2011). Unravelling the effects of the environment and host genotype on the gut microbiome. Nature Reviews Microbiology , 9(4), 279-290. https://doi.org/10.1038/nrmicro2540 Stewart, B.S. (1997). Ontogeny of Differential Migration and Sexual Segregation in Northern Elephant Seals. Journal of Mammalogy, 78(4), 1101-1116. https://doi.org/10.2307/1383053 Stoffel, M.A., Acevedo, Whitehouse, K., Morales, Durán, N., Grosser, S., Chakarov, N., Krüger, O., … Hoffman, J.I. (2020). Early sexual dimorphism in the developing gut microbiome of northern elephant seals. Molecular Ecology , 29(11), 2109-2122. https://doi.org/10.1111/mec.15385 Suzuki, T.A., Phifer‐Rixey, M., Mack, K.L., Sheehan, M.J., Lin, D., Bi, K., & Nachman, M. W. (2019). Host genetic determinants of the gut microbiota of wild mice. Molecular ecology , 28(13), 3197-3207. https://doi.org/10.1111/mec.15139 Switzer, A.D., Callahan, B.J., Costello, E.K., Bik, E.M., Fontaine, C., Gulland, F.M.D. (2023). Rookery through rehabilitation: Microbial community assembly in newborn harbour seals after maternal separation. Environmental Microbiology , 1-21. https://doi.org/10.1111/1462-2920.16444 Tasnim, N., Abulizi, N., Pither, J., Hart, M. M., & Gibson, D. L. (2017). Linking the Gut Microbial Ecosystem with the Environment: Does Gut Health Depend on Where We Live? Frontiers in Microbiology , 8, Article 1935. https://doi.org/10.3389/fmicb.2017.01935 Tian, J., Sanganyado, E., Wang, Z., Kong, Z., Han, J., Lu, Z., Liu, W. (2022). Spotted seals ( Phoca largha ) harbor unique gut microbiota shaped by their host habitat. Science of the Total Environment, 832, https://doi.org/10.1016/j.scitotenv.2022.155015. Thursby, E., & Juge N. (2017) Introduction to the human gut microbiota. Biochemical Journal , 474(11), 1823-1836. https://doi.org/10.1042/BCJ20160510 Tung, J., Barreiro, L.B., Burns, M.B., Grenier, J.C., Lynch, J., Grieneisen, L.E., … & Archie, E. A. (2015). Social networks predict gut microbiome composition in wild baboons. eLife , 4, Article e05224. https://doi.org/10.7554/eLife.05224 Turnbaugh, P.J., Ridaura, V.K., Faith, J.J., Rey, F.E., Knight, R., & Gordon, J.I. (2009). The effect of diet on the human gut microbiome: a metagenomic analysis in humanized gnotobiotic mice. Science translational medicine , 1(6). https://doi.org/10.1126/scitranslmed.3000322 Valeri, F., & Endres, K. (2021). How biological sex of the host shapes its gut microbiota. Frontiers in Neuroendocrinology , 61, Article 100912. https://doi.org/10.1016/j.yfrne.2021.100912 Watkins, C.A., Gaines, T., Strathdee, F., Baily, J.L., Watson, E., Hall, A.J., Free, A., & Dagleish, M. P. (2022). A comparative study of the fecal microbiota of gray seal pups and yearlings ‐ a marine mammal sentinel species. MicrobiologyOpen , 11(3), e1281. https://doi.org/10.1002/mbo3.1281 Wickham H, Hester J, Chang W, Bryan J (2022). devtools: Tools to Make Developing R Packages Easier. https://devtools.r-lib.org/, https://github.com/r-lib/devtools. Wickham H, Averick M, Bryan J, Chang W, McGowan LD, François R, Grolemund G, Hayes A, Henry L, Hester J, Kuhn M, Pedersen TL, Miller E, Bache SM, Müller K, Ooms J, Robinson D, Seidel DP, Spinu V, Takahashi K, Vaughan D, Wilke C, Woo K, Yutani H (2019). “Welcome to the tidyverse.” Journal of Open Source Software, 4(43), 1686. doi:10.21105/joss.01686. Wang, Y., & Lê Cao, K.-A. (2020). Managing batch effects in microbiome data. Briefings in Bioinformatics, 21(6), 1954–1970. https://doi.org/10.1093/bib/bbz105 Weber, D.S., Stewart, B., Garza, J.C., & Lehman, N. (2000). An empirical genetic assessment of the severity of the northern elephant seal population bottleneck. Current Biology , 10(20), 1287-1290. https://doi.org/10.1016/S0960-9822(00)00759-4 Wickham H, Hester J, Chang W, Bryan J (2022). devtools: Tools to Make Developing R Packages Easier. https://devtools.r-lib.org/, https://github.com/r-lib/devtools. Wickham, H., Averick, M., Bryan, J., Chang, W., McGowan, L.D., François, R., …, Yutani, H. (2019). “Welcome to the tidyverse.” Journal of Open Source Software, 4(43), 1686. doi:10.21105/joss.01686. Williams, T., & Carroll, S. (2009). Genetic and molecular insights into the development and evolution of sexual dimorphism. Nat Rev Genet, 10, 797-804. https://doi.org/10.1038/nrg2687 Xiao, C., Wang, J., Su, C., Miao, Z., Tang, J., …, Zheng, J.-S. (2022). Associations of dietary diversity with the gut microbiome, fecal metabolites, and host metabolism: Results from 2 prospective Chinese cohorts. The American Journal of Clinical Nutrition , 116(4), 1049-1058. https://doi.org/10.1093/ajcn/nqac178 Zhu, L., Wang, J., & Bahrndorff, S. (2021). Editorial: The Wildlife Gut Microbiome and Its Implication for Conservation Biology. Frontiers in Microbiology , 12, Article 697499. https://doi.org/10.3389/fmicb.2021.697499 Table 1. Alpha diversity analyses for the admissions dataset only. Using the Kruskal Wallis test implemented through the alpha-group-significance QIIME2 plugin, differences within sequencing batch, sex, and county of stranding were tested for Observed features, Pielou’s evenness, and Shannon’s diversity index. Bolded values are indicative of statistical significance ( p <0.05). (Abbreviations: CA, California; H, Kruskal-Wallis H value) Diversity metric Statistical measure County of Stranding Batch Sex Observed features H 4.96 27 1.7 p 0.4212 2.07x10 -7 0.1925 Pielou’s Evenness H 6.46 5.03 2.19 p 0.2638 0.0249 0.1392 Shannon’s Diversity Index H 8.99 10.73 2.19 p 0.1093 0.0011 0.1392 Table 2. Bray-Curtis univariate, multivariate, and homogeneity of dispersion analyses. Univariate and multivariate beta diversity analyses were conducted using the adonis QIIME2 plugin, which implements PERMANOVA. Pseudo F-values (F), percentages of variation explained (R 2 ), and p -values ( p ) are given. Homogeneity of dispersion analysis was conducted using the beta-group-significance QIIME2 function. F-values ( F) and p -values are given. Significance is shown in bold ( p <0.05). Univariate Analysis Multivariate Analysis Homogeneity of dispersion Dataset F R 2 p F R 2 p F p Admissions Batch 4.72 10.1% 0.001 4.67 8.9% 0.001 0.92 0.352 Sex 3.73 8.2% 0.001 3.12 5.9% 0.001 1.03 0.278 County 1.54 16.8% 0.006 1.49 14.1% 0.010 1.93 0.462 Longitudinal Batch 4.39 10.9% 0.001 4.98 9.9% 0.001 0.58 0.426 Sex 2.91 7.5% 0.003 3.20 6.4% 0.002 0.18 0.679 County 1.06 14.2% 0.350 1.32 13.2% 0.056 1.34 0.257 Time point 4.45 11.0% 0.001 5.75 11.5% 0.001 5.04 0.021 Time point 1 Batch 2.27 11.8% 0.003 2.15 9.9% 0.002 0.05 0.822 Sex 2.33 12.1% 0.002 2.11 9.7% 0.003 0.11 0.741 County 1.06 28.9% 0.360 1.16 26.5% 0.185 0.77 0.671 Time point 2 Batch 4.37 20.4% 0.001 3.95 18.0% 0.001 3.44 0.066 Sex 1.71 9.1% 0.063 1.69 7.7% 0.053 0.55 0.469 County 0.80 23.6% 0.850 0.96 21.8% 0.557 4.27 0.037 Figure 1 . Northern elephant seals included in this study were weaned in their natal rookeries across the coast of California. (A) Map illustrating the six California counties in which northern elephant seals in our study originated from. Image was created using the tigris , ggplot2 , and ggspatial packages in R. (B) Vapor (ES4686) is a rescued, malnourished NES pup undergoing rehabilitation at The Marine Mammal Center. Photo was taken by Bill Hunnewell © The Marine Mammal Center (n.d.). A. B. Figure 2. Sample collection flow diagram. Flow diagram displays how rectal samples were omitted from or selected for our admission and longitudinal datasets. Figure 3. Bray-Curtis PCoA plots for the admissions dataset only. Using the tidyverse , devtools, and qiime2R packages in R, PCoA plots were constructed using Bray-Curtis dissimilarity distance matrices. (A ) The variables of batch, sex, and (B) county of stranding were investigated for their contributions to beta diversity variation along Axis 1 (19.99%) and Axis 2 (12.43%). A. B. Figure 4. Differential abundance analyses of microbes on a genus and class taxonomic level. (A) In the admissions data, differential abundance between sexes was tested. Graph shows relative abundance of males relative to females. (B) In the longitudinal dataset, differential abundance between time point 1 and 2 was tested. Graph shows relative abundance of time point 1 relative to time point 2. Bonferroni correction was used to adjust p -values and statistical significance was deemed as q <0.05. Log-fold change is shown. A. B. Figure 5. Population genetic cluster analysis reveals little structure. (A) Average county genetic cluster membership of individuals based on ADMIXTURE analysis, with no support for K >1. (B) Non-model clustering (left) with machine learning denoising (right). (C) The proportion of model-based probability assignments per genetic partition ( K =2) across the six California counties and sex. A. B. C. Figure 6. (A) Bray-Curtis dissimilarity PCoA when time points are analyzed together, with points from the same individual connected by a line. (B) Bray-Curtis dissimilarity is significantly higher in samples collected at intake compared to those collected after a month at TMMC. Scatter plots of pairwise genetic distance and microbial dissimilarity (Bray-Curtis) for 13 individuals with samples collected at (C) intake and after (D) about a month at TMMC. Mantel test showed no significant correlation between genetic distance and Bray-Curtis dissimilarity at either time point. A. B. C. D. Information & Authors Information Version history V1 Version 1 13 January 2025 Copyright This work is licensed under a Non Exclusive No Reuse License. Keywords gut microbiome metabarcoding seals sexual dimorphism shared environment Authors Affiliations Emily Yu 0009-0007-5604-2188 [email protected] Princeton University View all articles by this author Alexandra DeCandia 0000-0001-8485-5556 Georgetown University View all articles by this author Andrea Graham Princeton University View all articles by this author Emily Whitmer 0000-0002-5956-6944 The Marine Mammal Center View all articles by this author Cara Field The Marine Mammal Center View all articles by this author Bridgett vonHoldt 0000-0001-6908-1687 Princeton University View all articles by this author Stephen Gaughran 0000-0002-9413-5074 Princeton University View all articles by this author Metrics & Citations Metrics Article Usage 448 views 218 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Emily Yu, Alexandra DeCandia, Andrea Graham, et al. Sexual dimorphism shapes the gut microbiome of northern elephant seal pups across environments. Authorea . 13 January 2025. DOI: https://doi.org/10.22541/au.173675473.39676661/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.173675473.39676661/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9ff5ba261eb8dfa9',t:'MTc3OTM4OTAxMA=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();