The composition of host-associated microbial communities may correlate with the overall status of the host, including physiology and fitness. New bi-directional hypotheses suggest that sexual behaviors can shape, and be shaped by reproductive microbiomes, which may be particularly important for species with mating systems that feature strong sexual selection. These dynamics have been particularly understudied in female animals. Using 16S rRNA sequencing, we compared the cloacal microbiome of females and males from two socially polyandrous bird species that vary in the strength of sexual selection, Jacana spinosa (Northern Jacana) and J. jacana (Wattled Jacana). We hypothesized that the strength of sexual selection would shape cloacal microbial diversity, such that the more polyandrous J. spinosa would have a more diverse microbiome, and that microbiomes would be more diverse in females than in males. If the reproductive microbiome is indicative of competitive status, we also hypothesized that cloacal microbial diversity would be associated with competitive traits, including plasma testosterone levels, body mass, or weaponry. We found no differences in microbial alpha diversity between species or sexes, but we did find that microbial beta diversity significantly differed between species. We also found a positive relationship between microbial alpha diversity and testosterone in female J. spinosa. Future experiments are needed to explore the potential drivers of correlations between the cloacal microbiome and competitive phenotypes in socially polyandrous jacanas.
La composición de las comunidades microbianas asociadas a los hospederos puede reflejar el estado general del huésped, incluyendo su fisiología y éxito reproductivo. Hipótesis recientes sugieren que los microbiomas reproductivos pueden intervenir en fenotipos seleccionados sexualmente, los cuales pueden ser particularmente importantes para especies con sistemas de apareamiento que presentan una fuerte selección sexual. Estas dinámicas han sido poco estudiadas, particularmente en hembras. Utilizando secuenciación de ARNr 16S, comparamos el microbioma cloacal de hembras y machos en dos especies de aves socialmente poliándricas, la jacana norteña (Jacana spinosa) y la jacana carunculada (J. jacana). Hipotetizamos que la fuerza de la selección sexual daría forma a la diversidad microbiana cloacal, de manera tal que la jacana norteña, al ser más poliándrica, tendría la mayor diversidad microbiana y los microbiomas serían más diversos en hembras que en machos. Si el microbioma reproductivo es indicativo del estado competitivo, hipotetizamos también que la diversidad microbiana cloacal estaría asociada con rasgos competitivos, como los niveles de testosterona en plasma sanguíneo, la masa corporal o el armamento. Encontramos una correlación positiva entre la diversidad alfa microbiana y los niveles de testosterona y armamento en las hembras de la jacana norteña. No encontramos diferencias en la diversidad alfa entre especies ni sexo, pero sí encontramos que la diversidad beta de microbios difirió significativamente entre las dos especies. Nuestros resultados sugieren que el microbioma cloacal puede ser un componente clave del fenotipo competitivo en jacanas socialmente poliándricas, respaldando la hipótesis de que la interacción entre la selección sexual y la comunidad microbiana puede moldear la fisiología y aptitud en aves salvajes huéspedes. KEYWORDS: reproductive microbiome, social polyandry, testosterone, Jacana, sexual selection
Almost all multicellular organisms host microbial communities, which consist of microorganisms (i.e., bacteria, viruses, unicellular algae, protozoans, and fungi) and their metagenomes -known as the 'microbiome' (Marchesi and Ravel 2015). The composition of host microbiomes varies with diverse internal and external factors, including host genetics (Goodrich et al. 2014), diet (Bodawatta et al. 2022), and social interactions (Tung et al. 2015). The microbiome is not only shaped by its host; it also influences processes within the host, including digestion, metabolism, and protection from infection (Grond et al. 2018). Following the theory on biodiversity-ecosystem function, greater microbial diversity may increase its functionality in the host. For instance, hosts with higher microbial diversity may have a broader uptake of nutrients (Tennant et al. 1971), lower risk of immunodeficiency (Le Chatelier et al. 2013) and disease (Minamoto et al. 2015), and more resistance to invasion by microbial pathogens (Spragge et al. 2023). Because of the myriad ways microbiomes can impact host health, microbial diversity and composition are potentially an essential component of organismal fitness (Williams et al. 2024).
A recent focus in microbial ecology and evolution is the reproductive microbiome, which consists of microorganisms within structures, organs, fluids, or tissues of the host reproductive tract (Rowe et al. 2020). Though past work has emphasized sexually transmitted pathogens (Sheldon 1993;Lombardo 1998), the reproductive microbiome can affect a variety of reproductive traits including the host's fertility (Williams et al. 2019), reproductive compatibility (Grieves et al. 2019), and hatching success (Van Dongen et al. 2019). Patterns of sexual transmission may influence the composition of the reproductive microbiome, as mating can facilitate the transfer of microbiota from one partner to another (Kulkarni and Heeb 2007;Rowe et al. 2020). One consequence of this microbial transmission is that mating partners may converge to have similar reproductive microbiomes (White et al. 2010;Pruter et al. 2023).
Another consequence is that females with multiple sexual partners may have higher reproductive microbiome diversity than monogamous females (MacManes 2011; White et al. 2011). Thus, the reproductive microbiome may be impacted by the magnitude and direction of sexual selection, including variation in mating behavior from monogamy to polygamy (Rowe et al. 2020).
Microbiomes may also correlate with external morphological traits and internal physiological states associated with reproductive fitness. Wild non-model organisms, including birds, have been valuable for examining the relationships between microbial diversity and fitness-related traits (Hird 2017;Grond et al. 2018). For instance, a study of northern cardinals (Cardinalis cardinalis) found that ornamentation and body condition were positively correlated with cloacal microbiome diversity (Slevin et al. 2024). In rufous collared sparrows (Zonotrichia capensis), cloacal microbiome diversity was positively correlated with testosterone levels (Escallón et al. 2016). In birds and most other species, studies have focused on the relationship between the microbiome and sexually selected traits in males, leaving the relationship between female microbiomes and competitive traits relatively understudied. Sexual experience throughout an individual's lifetime may also be reflected in their reproductive microbiome, as age was positively correlated with cloacal microbiome richness in female tree swallows (Tachycineta bicolor) (Hernandez et al. 2021). Though directionality can be challenging to disentangle from observational studies, these results nevertheless suggest an interesting relationship between competitive phenotypes and microbial diversity. Because there are inter-and intraspecific differences in the degree of sexual selection, it is important to study microbiome-fitness dynamics among multiple species and between both sexes.
Here, we conduct an exploratory study on the reproductive microbiomes of two socially polyandrous shorebirds, Jacana spinosa (Northern Jacana) and J. jacana (Wattled Jacana). Both J. spinosa and J. jacana are socially polyandrous, in which females mate with multiple males and are typically larger in body size and weaponry, and males mate monogamously and conduct the majority of parental care (Jenni and Collier 1972;Emlen and Wrege 2004b). However, J. spinosa has a greater degree of sexual dimorphism, higher average number of male mates, and longer sperm morphology, suggesting that sexual selection is stronger in this species (Lipshutz 2017;Lipshutz et al. 2023). Therefore, we predicted that female J. spinosa would have higher microbial diversity than female J. jacana. Given sex differences in patterns of sexual transmission, in which females mate with multiple males, but males only mate with one female, we expected higher microbial diversity among females than males. Finally, we evaluated the relationship between microbial diversity and sexually selected traits. We predicted that females with larger competitive traits would have more diverse reproductive microbiomes.
Jacana spinosa and J. jacana are tropical, freshwater shorebirds, and can be found in marshes and agricultural landscapes from Mexico to Central and South America. We studied wild, free-living populations of J. spinosa in La Barqueta,Chiriqui,Panama (8.207N,82.579W) and J. jacana between Las Guabas (8.384N, 80.447) and Chepo, Panama (9.166N, 79.122W).
Our fieldwork took place during the rainy season from May 16 to July 1, 2018, when breeding is most prevalent in jacanas, though jacanas may breed all year when freshwater is available (Emlen and Wrege 2004a). We focused on reproductively active adults, including territorial females and their male mates. Male jacanas breed asynchronously and cycle between two breeding stages: courtship, when males actively copulate with their female mate, or parenting, when male are incubating eggs, brooding, or foraging with chicks (Lipshutz and Rosvall 2020).
We confirmed these breeding stages with behavioral observations of copulation, and the presence of nests and/or brood patches.
In total, we sampled 39 individuals: 11 J. spinosa females, 12 J. spinosa males (five courting, seven parenitng), 9 J. jacana females, and 7 J. jacana males (four courting, three parenting). These individuals were collected for a study on the neurogenomic mechanisms of social polyandry, which involved terminal collection of brain and peripheral tissues. Birds were euthanized with an air rifle, followed by an overdose of isoflurane anesthetic. These sample sizes are limited, stemming in part from the logistical challenges of tropical fieldwork in wetland habitats. Nonetheless, these samples are a valuable contribution, as the microbial communities of this avian family, Jacanidae, have not yet been described.
In birds, most research focuses on the gut microbiome (Sun et al. 2022), which is typically isolated non-lethally from fecal or cloacal samples (Berlow et al. 2020). However, the cloacal cavity is an endpoint for intestinal, urinary, and reproductive functions, and also receives bacteria incorporated into ejaculates via copulation (White et al. 2010). Thus, cloacal samples may also reflect the reproductive microbiome (Rowe et al. 2020), and we use the term cloacal microbiome to represent the reproductive microbiome. To collect the cloacal sample, we cleaned the outside of the cloaca with an alcohol wipe to remove environmental and eased a sterile swab (Puritan 25-3316-U Ultra Flocked Swab, USA) into the cloaca. Once the head of the swab was fully inserted into the cloaca (~20mm), we gently turned the swab for 3-5 s. Swabs were stored in 500uL of RNAlater (Invitrogen; Carlsbad, CA USA), flash frozen on dry ice, and later stored at -80°C until DNA extractions.
We measured competitive phenotypes that have previously been associated with territoriality and reproductive success in J. jacana (Emlen and Wrege 2004a). We measured body mass to the nearest gram using a Pesola scale, and wing spurs to the nearest millimeter using calipers. An analysis of testosterone levels in circulation has been published for these same J. spinosa individuals from both sexes (Lipshutz and Rosvall 2020) and of sperm morphology for these same males from both species (Lipshutz et al. 2023).
We extracted DNA from whole swabs using the DNeasy PowerSoil DNA isolation kit (Qiagen, Inc.) with minor modifications to the manufacturer's recommended protocol (Vo and Jedlicka 2014). We pipetted off the RNAlater supernatant from the swabs prior to DNA extraction. In addition to Solution C1, we added 10μl of Proteinase K for the tissue lysing step.
The two solutions (Solutions C2 and C3) which precipitate non-DNA substances were combined (Berlow et al. 2022). Final DNA was eluted from each spin column twice with the same 100 μl of Solution C6. DNA extracts were frozen at -30°C until PCR.
PCR and sequencing follow the methods described in (Houtz et al. 2023). We amplified the V4 region of the 16S rRNA gene using the primers 515F and 806R with Illumina adapters added. We ran 10 μl PCR reactions in triplicate for each sample and included 5 μl of 2× Platinum Hot Start Master Mix (Invitrogen, Waltham, MA, USA), 0.5 μl of 10 μM primers, 3 μl of nuclease free water, and 1 μl of template DNA. Cycling conditions were 3 min at 94°C followed by 35 cycles of 94°C for 45 s, 50°C for 60 s, and 72°C for 90 s before a final extension at 72°C for 10 min. We pooled the three replicate reactions for each sample and ran a 1% agarose gel to confirm that amplification was successful for the V4 region of the 16S rRNA gene (~350 bp).
Each PCR run included negative controls (nuclease-free water in place of template DNA but were not sequenced). We also extracted, amplified, and sequenced 4 negative kit reagent controls. We submitted our final pooled PCR products to the Cornell Biotechnology Resource Center for quantification, normalization, library preparation, and sequencing on a single Illumina MiSeq run (2 x 250 bp).
We used QIIME2 (Quantitative Insights Into Microbial Ecology; v 2024.5) to process demultiplexed paired-end forward and reverse sequences as FASTQ files (Bolyen et al. 2019).
We used the DADA2 plugin to truncate the sequences to the 180 bp position to remove low quality base pairs based on interactive sequence quality plots. After trimming off primers, these trimmed sequences were used to create amplicon sequence variants (ASVs). We assigned taxonomy to the ASVs by fitting a naive-Bayes classifier trained on the Silva 138 database using the sk-learn classifier (Yilmaz et al. 2014). The phylogeny plugin was applied to construct a rooted phylogenetic tree by employing FastTree and MAFFT. We removed singletons, chloroplasts, mitochondria, eukaryotes, archaea, and any ASVs unassigned to a bacterial phylum.
We exported the filtered ASV table, taxonomy, and phylogenetic tree artifacts from QIIME2 into R v 4.3.1 for all subsequent analyses. Using the 'phyloseq' package v 1.46.0, we combined the ASV table, sample metadata, taxonomy table, and phylogenetic tree into a phyloseq object (McMurdie and Holmes 2013). We used the 'decontam' package v 1.22.0 to remove 4 contaminant ASVs and were left with 2,227 ASVs across 39 total samples after removing negative controls (n = 4) (Davis et al. 2018).
To limit bias due to differing sequencing depths, we rarefied samples to 14,000 reads, removing 2 samples (WAF3 and WAM6) with less than 14,000 reads, leaving 1,989 across 37 samples. With the 'phyloseq' package, we calculated alpha diversity metrics including Chao1 index (Chao 1984) and Shannon index (Shannon and Weaver 1949). Faith's Phylogenetic Diversity (PD) (Faith 1992) was calculated using the 'picante' package v 1.8.2 (Kembel et al. 2010). Chao1 index measures the richness of the samples, Shannon's index considers the richness and evenness of the samples, and Faith's PD assigns the diversity score based on how related the species are to each other from our phylogenetic tree.
For bacterial beta diversity, we calculated Bray-Curtis dissimilarity (richness and abundance of the ASVs in the communities; (Sørensen 1948) and Jaccard distance (ASV richness only; (Jaccard 1908). All samples were processed to remove exceptionally rare taxa that could disproportionately influence beta diversity metrics. ASVs were excluded if they had fewer than 50 reads in total across all samples. Bray-Curtis dissimilarity and Jaccard matrices were calculated using reads rarefied to 14,000 as described for alpha diversity metrics. After prevalence filtering, 352 ASVs were retained from rarefied data.
All statistical analyses were run in R v 4.3.1. We used the package 'lme4' v 1.1-35.3 (Bates et al. 2015) to run linear mixed effect models with each microbial alpha diversity. The package 'DHARMA' v 0.4.6 was used to carry out model diagnostics (Hartig 2018). To test for changes in microbial beta diversity, we conducted a separate permutational multivariate analysis of variation (PERMANOVA) with each distance matrix (Bray-Curtis or Jaccard) as the response variable to statistically partition the sources of variation in microbial community structure with the adonis2 function in the 'vegan' package v 2.6-6.1 (Oksanen et al. 2022), with the 'by' parameter set for 'margin' to account for marginal effects of the tested variables. An assumption of the adonis test is that groups have homogeneity of variance. The dispersion of the groups outlined in each section below was checked for homogeneity of variance using the Betadisper and permutest functions from the 'vegan' package. All figures were created using the 'ggplot2' package v 3.5.1 (Wickham 2016) and cartoons were added in PowerPoint.
We used Spearman correlations to compare competitive traits, log testosterone levels, and alpha diversity metrics using the cor.test function in R. We analyzed correlation matrices using the cor.mtest function and visualized them with the package corrplot (Wei and Simko 2024).
After agglomerating ASVs by genus, we conducted a similarity percentage (SIMPER) analysis to identify the average contribution of each bacterial genus to the Bray-Curtis dissimilarity (beta diversity) of the significant species comparison in PAST (v 4.15), with a 70% cutoff for low contributions to dissimilarity. Beta diversity ordinations for Bray-Curtis and Jaccard were visualized with principal component analyses (PCoA) calculated with the 'phyloseq' package.
We only tested for relationships between bacterial genera in female J. spinosa because we found relationships between testosterone and alpha diversity in female J. spinosa. We used MaAsLin2 (Microbiome Multivariable Associations with Linear Models; (Mallick et al. 2021) to identify differentially-abundant genera with log testosterone in female J. spinosa as the fixed effect with the following functions: analysis_method = "CPLM" (Compound Poisson Linear Model), transform = "AST" (Arcsine Transformation), and correction = "BH" (Benjamini-Hochberg or False Discovery Rate for multiple comparisons correction). We filtered the ASV table to only include female J. spinosa and removed sample NOF6 because it did not have testosterone data (n = 10). We used the rarefied ASV table agglomerated to the genus level as the input for the MaAsLin2 analysis, as MaAsLin2 based on rarefied data produces the most consistent results across datasets (Nearing et al. 2022).
Before filtering, there were 1,254,953 total reads and 2,445 unique ASVs across 43 samples (n = 39 cloacal samples, n = 4 negative controls). Mean reads per sample were 29,185 with samples ranging from 257 to 53,578 reads. After filtering (i.e., removal of singletons, chloroplasts, mitochondria, eukaryotes, archaea, and any ASVs unassigned to a bacterial phylum), we retained 1,225,236 reads and 2,231 unique ASVs that were used to calculate cloacal microbial alpha and beta diversity metrics. Mean reads per sample were 28,493.9 with samples ranging from 221 to 52,860 reads.
We did not find a significant relationship with alpha diversity and species, sex, nor their interaction for all three diversity metrics including Chao1 index (sex*species: β = 24.5, CI = -70.7 -119.70, P = 0.60, Supplementary Material Table S1, Supplementary Material Figure S1A), Shannon index (sex*species: β = -0.33, CI = -1.27 -0.61, P = 0.48, Supplementary Material Table S2, Supplementary Material Figure S1B), and Faith's PD (sex*species: β = 1.18, CI = -5.13 -7.49, P = 0.71, Supplementary Material Table S3, Supplementary Material Figure S1C).
For Bray-Curtis dissimilarity, we found a significant relationship with beta diversity and species (R 2 = 0.05, F = 1.72, P = 0.04), but not for sex (R 2 = 0.04, F = 1.30, P = 0.16) nor their interaction (sex*species: R 2 = 0.03, F = 1.14, P = 0.26) (Figure 1A). For Bray-Curtis beta dispersion, there was no significant effect of species (F = 0.35, P = 0.56) on, but there was a trend that the sexes had marginally different dispersions (F = 3.15, P = 0.08). For Jaccard distance, we also found a significant relationship with beta diversity and species (R 2 = 0.04, F = 1.29, P = 0.03), but not for sex (R 2 = 0.03, F = 1.06, P = 0.27) nor their interaction (sex*species: R 2 = 0.03, F = 1.08, P = 0.22) (Figure 1B). There was no significant effect of species (F = 0.06, P = 0.81) or sex (F = 2.30, P = 0.14) on Jaccard beta dispersion.
The SIMPER analysis revealed that over 80% of the dissimilarity between J. jacana and J. spinosa can be attributed to changes in the relative abundances of 21 bacterial genera. A higher relative abundance of Catellicoccus and Aeromonas in J. jacana explained 14.08% and 6.98% of the average dissimilarity between the two species, respectively. A higher relative abundance of Campylobacter in J. spinosa explained 9.52% dissimilarity between species (Supplementary Material Table S4). Other notable genera of lower relative abundance across both species and sexes include Clostridium sensu stricto 1, Escherichia-Shigella, Helicobacter, Corynebacterium, and Acinetobacter, among others (see Supplementary Material Figure S2 for relative abundance taxa barplots). distance (beta diversity) of jacana cloacal microbiomes by species and sex. Red indicates Jacana jacana (Wattled Jacanas) and yellow indicates J. spinosa (Northern Jacanas). Circles represent females and triangles represent males. courtship (dark blue circles) and parenting (light blue triangles) breeding stages.
The MaAsLin2 analysis found significant correlations between log testosterone and 12 differentially-abundant bacterial genera in female J. spinosa (Supplementary Material Figure S4 and Table S5). Testosterone levels were positively correlated with 8 genera including Alistipes (coef = 91.94, p < 0.001), Devosia (coef = 84.40, p < 0.001), Dysgonomonas (coef = 14.09, p < 0.001), an uncultured genus from the family Desulfovibrionaceae (coef = 9.77, p < 0.001), Rikenella (coef = 9.30, p < 0.001), Mucispirillum (coef = 9.03, p < 0.001), Parabacteroides (coef = 7.37, p = 0.005), and Methylobacterium-Methylorubrum (coef = 5.72, p = 0.007). Testosterone levels were negatively correlated with 4 genera including Cronobacter (coef = -123.68, p < 0.001), an alphal cluster genus from family Beijerinckiaceae (coef = -113.07, p < 0.001), pair fertilizations (Raouf et al. 1997). Jacanas defend territories containing multiple male mates, and females with the largest competitive phenotypes have higher reproductive success (Emlen and Wrege 2004a). However, we do not know whether females with higher alpha diversity had more sexual partners, as we did not measure this in the field. Testosterone levels in female J. spinosa positively correlated with 8 bacterial genera, including Rikenella and Mucispirillum. These genera, respectively, have been found in fecal samples of female rats with prenatal androgen exposure (Sherman et al. 2018) and associated with hyperandrogenism in female mice (Chen et al. 2024).
An alternative, but not mutually exclusive explanation, is that higher testosterone suppresses immune function, facilitating the colonization of additional microbes in the reproductive tract. This potential trade-off between competitive phenotypes, sexual transmission, and immune function may explain the relationship between testosterone concentrations and the relative abundance of Chlamydia, a potentially pathogenic bacteria, in rufous collared sparrows (Escallón et al. 2016). In support of potential testosterone-mediated immune suppression, testosterone was positively correlated with bacterial genera, Dysgonomonas and Parabacteroides, which are associated with disease in the cecum and colon of ostriches (Videvall et al. 2020). Dysgonomonas is also related to an increased phytohaemagglutinininduced immune response in barn swallow nestlings (Kreisinger et al. 2018). Another bacterial genus positively correlated with testosterone in jacanas, Alistipes, may increase susceptibility of mice to orchitis, inflammation of testes, during gut microbiota dysbiosis (Guo et al. 2024).
Although we are unsure why the relationship between alpha diversity and competitive traits was only significant in J. spinosa, one explanation is that sexual selection is strongest in this species. J. spinosa females have a greater degree of polyandry, and more extreme female-biased sexual dimorphism than J. jacana females (Lipshutz 2017). Future work should test whether female jacanas with more mating partners receive a higher diversity of microbes via sexual transmission, to directly test this relationship. It is important to acknowledge the null hypothesis, that the diversity of the cloacal microbiome is a byproduct of other physiological processes unrelated to sexual behavior or immune function. Future studies are needed to explore the potential drivers of these relationships.
In male J. spinosa, Shannon diversity and testosterone levels correlated with body mass.
A previous study of these same individuals found that courting males had both higher testosterone and larger body mass than parenting males (Lipshutz and Rosvall 2020). It is possible that microbial diversity is impacted by male breeding stage. We were unable to analyze male breeding stages separately due to small sample sizes, but future studies could examine these dynamics.
We predicted that J. spinosa would have higher microbial alpha diversity than J. jacana, due to species differences in the strength of sexual selection (Lipshutz 2017;Lipshutz et al. 2023). We also expected higher microbial diversity among females than males, because of their socially polyandrous mating systems. However, we did not find significant differences in microbial alpha diversity between species nor between sexes. We found a significant effect of species on cloacal microbial beta diversity, but no effect of sex. Although cloacal microbial alpha diversity did not differ between sexes or species, beta diversity differences suggest that J. spinosa and J. jacana may differ in their cloacal microbial community composition while maintaining similar alpha diversity levels. As studies of other species have found sex differences in microbial diversity e.g. (Gongora et al. 2021), future work in jacanas should re-examine these patterns with larger sample sizes.
J. spinosa and J. jacana may differ in their microbial compositions due to geographic isolation, environmental variation, and/or dietary strategy. Previous ecological niche modeling found significant differences in habitat suitability between the species, with suitable habitats for J. spinosa being wetter and warmer (Miller et al. 2014). The two species hybridize where their geographic ranges overlap (Lipshutz et al. 2019), but there are very few individuals of both species in this region, and our samples are from individuals outside the hybrid zone.
Observations from the field suggest that both species eat seeds and aquatic vegetation, as well as insects, aquatic invertebrates, and small fish and amphibians (Jenni and Kirwan 2020). The main bacterial genera that explained the highest percentage of the dissimilarity between species included Catellicoccus and Campylobacter, both of which are found at differing relative abundances in wild waterbirds (Boukerb et al. 2021). We also found the presence of Aeromonas in the animal prey of jacanas including fish and insects (Dubey et al. 2022). Future research should characterize the diets of both species via diet metabarcoding of fecal samples and investigate potential relationships with the microbiome. Future work should also sample the microbiomes of other parts of the reproductive and digestive tracts, to evaluate the extent to which cloacal swabs represent the gut versus the reproductive microbiome.
Sexual transmission of bacteria in birds may be asymmetrical, as males copulate into the cloaca of the female (Kulkarni and Heeb 2007); however, transmission could be bi-directional, and how these patterns change in polyandrous vs. polygynous systems is still an open question.
Within each species, we found no difference in microbial beta diversity between males and females. At the population level, sex differences in the diversity of reproductive microbiomes may be eroded by increasing sexual transmission (Rowe et al. 2020). Males and females may homogenize their reproductive microbial compositions through copulations. Kittiwakes share a similar cloacal microbial composition with their mates, but after inseminations are experimentally blocked, the cloacal communities of mates became increasingly dissimilar (White et al. 2010). Transmission of cloacal bacteria between males and females is likely limited to the breeding season (Escallón et al. 2019), when there is direct cloacal contact between individuals.
Thus, the year-round breeding season of jacanas (Emlen and Wrege 2004a) may allow for continuous intersexual transmission of cloacal microbiota.
Though this was an exploratory study, the patterns we uncovered between microbial diversity and competitive traits pose exciting new directions for understanding female-driven dynamics shaping reproductive microbiomes in wild birds. Promising research directions should include experimental manipulations of the reproductive microbiome via antibiotics or probiotic supplementation to investigate causal links between the avian microbiome and competitive phenotypes.