Age-driven changes in the layer hen reproductive microbiome are associated with lay performance

Abstract

Eggs are a globally important food source and integral to optimal poultry production. Understanding the microbial ecology of the hen reproductive tract is essential for improving both food safety and reproductive efficiency. While the oviduct has been shown to harbor a continuous microbial community, this study is the first to demonstrate the presence of microbiota on the hen ovary surface, suggesting that the ovary is an extension of the oviductal microbial continuum. In this study, the ovarian and oviductal microbiomes of white-leghorn hens from mid-lay (high laying) and post-lay (lower laying) cohorts were analyzed. Using 16S rRNA sequencing, we identified significant shifts in reproductive tract microbiota between 9- and 18-month-old hens, coinciding with changes in lay performance. Several differentially abundant genera, including Acinetobacter, Ligilactobacillus, Bacillus, and Akkermansia, are known to modulate steroid hormone metabolism, with age-related abundance changes suggesting potential effects on hormone-driven reproductive processes. Other genera such as Ruminococcus_torques_group, Mucispirillum, and Fusobacterium—not traditionally associated with reproductive hormone pathways—may influence laying efficiency through their roles in mucin degradation, immune modulation, and inflammation. Notably, Turicibacter, newly identified on the ovary, increased with age and negatively correlated with lay performance, raising questions about its role in bile acid metabolism and stress response within the hen reproductive tract. Collectively, these findings highlight the ovary as an active microbial niche influenced by age and suggest that both hormone-associated and mucosal-interactive microbes contribute to lay dynamics. This work opens new avenues for probiotic strategies targeting key genera to support hen fertility and egg production across the productive lifespan.

Introduction

Global egg consumption has increased 20 % over the past two decades and is projected to continue growing, according to the Food and Agriculture Organization (FAO) and World Egg Organization (WEO) (Food and Agriculture Organization of the United Nations, 2025; World Egg Organization, 2022). The layer hen, or egg-laying chicken, has an agricultural productive period that is approximately 18- to 24-months in length, after which she is repurposed due to declining lay performance. In the average white leghorn, sexual maturity occurs at approximately five months of age. At that time, she has her first ovulation and begins to lay eggs which marks the start of the peaking phase. By nine months, egg production stabilizes near-peak productivity (Hy-Line International, 2020). However, by 12 months, lay performance declines by about 4 %, and by 18 months, productivity drops by 14-16 % from where it was at nine months. At this point the cost-to-income ratio becomes prohibitive, and the hens are considered spent. We propose that the hen reproductive microbiome contributes to this decline and may present opportunities for intervention.

The layer hen reproductive tract is comprised of the ovary, oviduct, vagina, and cloaca. The oviduct is further divided into specialized segments: the infundibulum, closest to the ovary, funnels a recently ovulated follicle into the oviduct; the magnum secretes egg white; the isthmus provides space for eggshell membrane formation; and the uterus forms and calcifies the eggshell. The oviductal microbiome has previously been characterized by low beta diversity and a decreasing microbial load ascending the tract toward the infundibulum (Canha-Gouveia, et al., 2023; Wen, et al., 2021). However, it remains unclear whether the ovary is colonized by this continuum.

Female reproductive microbiomes are of key interest in fertility and disease research. However, studies on the layer hen’s reproductive microbiome have primarily focused on heritability, food safety, and food security (Cheng, et al., 2024; Lee, et al., 2019; Lee, et al., 2020; Shterzer, et al., 2020; Su, et al., 2021). Few studies have investigated the microbiome’s role in hen fertility and reproductive efficiency (Dai, et al., 2023; Shterzer, et al., 2024; Wen, et al., 2021), and none have examined the ovarian microbiome.

Research on the ovarian surface is limited due to physiological barriers and the risks associated with invasive surgery, but studies of follicular fluid aspirations suggest that the presence of specific microbial genera can influence ART outcomes in humans (Babu, et al., 2017; Bernabeu, et al., 2019; Muhleisen and Herbst-Kralovetz, 2016). These findings suggest that microbes residing on the poultry ovarian surface (the cortex) and within the oviduct may also significantly impact fertility processes such as oocyte quality and folliculogenesis and thus egg production and quality. However, ovarian microbiomes and mechanisms underlying these microbial influences remain to be elucidated in the layer hen.

In this study, we examine the layer hen reproductive microbiome – including, for the first time, the ovary – throughout the reproductive window to assess microbial community fluctuations, host-microbiome interactions, and potential targets for optimizing microbiomes to enhance lay performance. Our studies indicate that the poultry ovarian surface is not sterile, suggesting that resident microbes may influence fertility. Furthermore, significant microbial shifts between 9- and 18-month-old-hens are associated with changes in steroid-related pathways, which may directly impact lay efficiency.

Materials and methods

Animal care and sample collection

All animals were acquired, housed, cared for, and euthanized as per University of Delaware Institutional Animal Care and Use Committee (IACUC)-approved protocols. All animal experiments were designed and performed in accordance with the approved regulations as per Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines. One hundred and fifty specific-pathogen-free (SPF) layer hens were acquired from Charles River Laboratories (Norwich, CT, USA) at six weeks of age and housed at the University of Delaware Poultry Research Farm as previously described (Kramer, et al., 2024; Kramer, et al., 2025).

Initially, hens were reared in temperature-controlled (21.1 °C ± 6 °C) colony housing with 12 hours of light exposure per day. To onset puberty, light exposure was gradually increased to 15 hours between 19- and 24-weeks as per commercial standards. At five months of age, animals were transferred to commercial standard cage layer housing, where they were housed individually to enable accurate daily egg production records. Lay data was analyzed in GraphPad Prism (10.4.1) by one-way Analysis of Variance (ANOVA) and Tukey’s multiple comparison test. Feed was (All Grain Layer and Breeder Crumble at 16 % crude protein (Southern States, Minneapolis, MN)) provided once a day and water ad libitum.

From the flock of 150 animals, the ten highest-producing individuals were selected and euthanized at 9, 12, and 18 months of age (Fig. 1). Animals were dissected to remove the reproductive tract. For each animal, four sites were sampled by sterile swabs (Dealmed, Brooklyn, NY, USA): the ovarian surface epithelium (cortex) and the mucosal surfaces of the infundibulum-magnum junction (Inf-Mag), isthmus-uterus junction (Ist-Ute), and cecum. Additionally, at each timepoint, environmental samples were collected from the hens’ cages, feed troughs, and waterers using sterile swabs. All swabs were flash-frozen in liquid nitrogen and stored at -80 °C until DNA isolation.

Fig. 1.

Experimental Overview. Cross-sectional sampling of animals occurred at 9, 12, and 18 months of age. Swabs from four areas of interest were collected – ovary, infundibulum-magnum (Inf-Mag), isthmus-uterus (Ist-Ute), and ceca. 16S DNA was isolated and sequenced, after which analyses of diversity, differential abundance, and metabolic prediction were performed.

16S DNA isolation and processing

DNA was isolated from swab samples using the DNeasy Blood & Tissue Kit (Qiagen, Germantown, MD, USA) (Lee, et al., 2020). Quantification was achieved using a Qubit fluorometer (Invitrogen, Waltham, MA, USA) and quality observed by nanodrop (Thermo Fisher Scientific Inc., Waltham, MA, USA). Samples were sent to the University of Delaware DNA Sequencing & Genotyping Center (University of Delaware, Newark, USA). Libraries were prepared using the Nextera XT DNA library preparation kit (Illumina, San Diego, CA, USA). The 16S rRNA microbial gene was sequenced using 357F/806R primers for the V3/V4 region on the Illumina MiSeq (Illumina, San Diego, CA, USA) to obtain paired-end 300bp amplicons. Files were received in de-multiplexed FASTQ format for downstream bioinformatic analyses.

16S rRNA data quality control and normalization

Demultiplexed paired-end sequences from 74 samples were loaded into QIIME2 (2024.5.0) for processing. The q2-cutadapt plugin was used for trimming primers and low-quality bases. The q2-DADA2 plugin was used for denoising and joining paired-end reads. This was followed by phylogenetic tree construction using the q2-phylogeny plugin. Training of taxonomic classifiers occurred using the q2-classify-sklearn Naïve Bayes on the SILVA (138.1) database for the 357F/806R primers. Files were exported from QIIME2 for further analysis in RStudio (4.4.1).

Within RStudio, the packages qiime2R (0.99.6) and phyloseq (1.48.0) were employed to load-in and organize data. Sequences were filtered after tree construction to remove amplicon sequence variants (ASVs) identified as chloroplast, mitochondria, and eukaryote as these were likely host contamination (Tan, et al., 2015). The RStudio packages data.table (1.16.0), ggplot2 (3.5.1), and dplyr (1.1.4) were used to examine sequencing depth and rarefaction curves across samples. A rarefaction curve revealed that most bacterial richness and diversity can be captured at a subsampling depth of 10,000. Samples were rarefied, resulting in the removal of five samples. The remaining samples allowed comparison group sizes greater than six samples each, which was sufficient for analysis. Rarefied-normalized data was subsequently used for all analyses and comparisons.

Comparison groups and bioinformatic analyses

Samples were organized and compared by location, age, and their combination, resulting in eight maximum comparison groups: ovary at 9- and 18-months, Inf-Mag at 9- and 18-months, Ist-Ute at 9- and 18-months, ceca at 9- and 18-months, and the environment. While environmental samples were reported, they were not used for decontamination. Current decontamination, such as decontam, assumes that true taxa are distinct from contaminants (Davis, et al., 2018). However, chickens exhibit cloacal drinking, a phenomenon where environmental microbiota can be taken up through the cloaca, leading to microbial overlap between the reproductive tract and environmental populations that cannot be considered contamination (Schaffner, et al., 1974; van der Sluis, et al., 2009). Chickens may also take up environmental microbiota through this pathway and thus the cloaca and attached reproductive tract can be expected to have some community overlap with environmental populations that cannot be considered contamination.

Alpha and beta diversity analyses were conducted in RStudio using phyloseq (1.48), vegan (2.6-8), and microbiome packages (1.26.0). Alpha diversity was assessed using observed amplicon sequence variants (ASVs) to quantify total unique taxa and the Shannon Index to evaluate taxa abundance and evenness. Normality was tested using the Shapiro-Wilk test. Groups with a normal distribution were analyzed using a Student’s t-test for pairwise comparisons and ANOVA for multiple group comparisons, while non-normally distributed groups were analyzed using the Wilcoxon test and Kruskal-Wallis test, respectively. Beta diversity was measured using Bray-Curtis dissimilarity with significance determined by Adonis2 Permutational Multivariate Analysis of Variance (PERMANOVA). Data visualization was performed using ggplot2 (3.5.1) and ggpubr (0.6.0).

Taxonomic composition was visualized, and differential abundance and correlation analyses were conducted using Analysis of Compositions of Microbiomes with Bias Correction 2 (ANCOM-BC2) (2.6.0) (Douglas, et al., 2020). Rarefied data were proportion-normalized for visualization, with samples aggregated by location and age. Genus-level taxonomic visualization focused on the 15 most abundant genera (relative abundance >5 %). For pairwise analyses within each location and across timepoints, either a Student’s t-test or Wilcoxon rank-sum was applied, depending on normality (Shapiro-Wilk test).

Differential taxonomic analysis by ANCOM-BC2 (2.6.0) was conducted for the ovary alone – a novel sampling site in the chicken model – and for the combined reproductive tract (oviduct and ovary), given the nonsignificant beta diversity differences between these sites. Differential abundance bar plots, generated from rarefied data using ANCOM-BC2, identified taxa significantly different between timepoints (P < 0.05). For the ovarian analyses, significance was set at a raw P < 0.01, with taxa also significant after FDR adjustment (P < 0.05) denoted with an asterisk. For reproductive tract analyses, all visualized taxa were significant at P < 0.05 (FDR-adjusted).

A Spearman rank correlation analysis, informed by taxa identified as significant through ANCOM-BC2, was conducted to explore relationships between ovary and reproductive tract taxa and lay performance. Taxa with a significant correlation (P < 0.05, FDR-adjusted) and a correlation coefficient (|r| > 0.5) were denoted with an asterisk.

Metabolic pathway prediction was achieved using QIIME2 hosted Phylogenetic Investigation of Communities by Reconstruction of Unobserved States 2 (q2-PICRUSt2) software resulting in 7,320 Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway hits (Douglas, et al., 2020). Results were visualized in RStudio using ggpicrust2 (1.7.3) (Yang, et al., 2023a). Pathways relating to parasitic and viral were removed, as no parasites nor viruses can be accurately quantified by 16S data, making these hits erroneous. Additionally, human disease pathways were excluded, as they are unlikely to provide relevant insights into chicken physiology in the lay context. Given non-significant beta diversity differences across the oviduct and ovary, reproductive tract samples were collapsed for analysis. LinDA statistical analysis was used to determine pathway significance at P < 0.05 (FDR-adjusted).

Results

Temporal analysis of lay performance data

Individual egg-laying frequency was recorded from start of lay until the day of euthanasia for capture of lay performance. Lay frequency from the eight to ten birds in each cross-sectional cohort showed no significant difference in the daily lay rate between 9- and 12-month-old-hens; however, a significant decline was observed at 18 months compared to both earlier timepoints (P < 0.0001) (Fig. 2). The lack of significant change between 9- and 12-month-old-hens suggests that microbiome shifts at 12 months do not meaningfully inform lay performance. Age could still be a confounding variable, however. Consequently, the 12-month timepoint was excluded from further analyses to better assess the relationship between age, lay performance, and changes in the reproductive microbiome.

Fig. 2.

Lay Performance within the Productive Window. Lay performance was sampled from 150 birds from the day of first lay until the day of euthanasia (9-, 12-, and 18-months). Lay data is presented for eight to ten birds used for microbial analysis. Data was analyzed in GraphPad Prism (10.4.1) by one-way Analysis of Variance (ANOVA) and Tukey’s multiple comparison test.

Temporal and spatial microbiota diversity analyses

In this study, a total of 1,637,928 total reads were isolated from a total of 74 samples. After filtering, 90.2 % of these remained for a total of 1,476,612 filtered reads. These were distilled to 27,029 unique ASVs and 627 identifiable genera. After rarefaction to a sampling depth of 10,000, 11,852 unique ASVs remained with 431 identifiable genera. Samples were derived from sterile swabs of the ovary, Inf-Mag, Ist-Ute, ceca, and the environment of layer hens at the time points 9- and 18-months of age.

Within-group microbial diversity was assessed using observed ASVs, which track the number of unique sequences, and the Shannon Index, which accounts for both richness and evenness. Alpha diversity analysis revealed a significant difference in Shannon diversity by sampling location (P = 0.0498, ANOVA across all groups), but not by age (P = 0.7807) (Fig. 3). Pairwise analysis indicated that the primary source of location-based differences was the ceca, which significantly differed from the reproductive tract (P < 0.05), a finding consistent with literature (Shterzer, et al., 2020). While age-related differences were not statistically significant, a visual trend suggested that the number of observed ASVs in reproductive tract samples decreased with age, while Shannon Index values increased (Fig. 3). This suggests that aging may reduce the number of unique microbial taxa in the reproductive tract while promoting a more even microbial community structure, but a greater sample size should be tested in the future to confirm.

Fig. 3.

Alpha Diversity Measures. Two methods of measure were employed to inform on alpha diversity: (A) Observed Species and (B) Shannon Index. Comparisons were across locations – ovary, infundibulum-magnum (Inf-Mag), isthmus-uterus (Ist-Ute), ceca, and environment (Env) – over two timepoints: 9- and 18-months.

Beta-diversity analysis was conducted using Bray-Curtis distances and PERMANOVA (EggsPerDay + Age + Location), revealing significant differences in microbiome composition across samples (P = 0.001), driven by differences between the oviduct and cecum (Fig. 4A). Within each timepoint, no significant beta diversity differences were observed along the reproductive continuum (P = 0.866 in 9-month-old-hens, P = 0.418 in 18-month-old hens), from uterus to ovary, supporting previous findings (Wen, et al., 2021) (Fig. 4B). For the first time, our data confirms that the ovary follows this reproductive tract continuum, suggesting that microbiota ascending the tract may establish on the ovary. Further, age-based comparisons revealed significant microbiome composition differences between 9- and 18-month-old hens in all reproductive sites (P < 0.05) but not in the ceca (P = 0.11) (Fig. 4C). These findings suggest that, during the productive window, aging exerts a greater influence on the oviduct and ovarian microbiomes than on the ceca (Awad, et al., 2016; Lu, et al., 2003; Videnska, et al., 2014).

Fig. 4.

Beta Diversity Measures. Bray-Curtis Dissimilarity visualizations were used to compare beta diversity between (A) all groups, (B) reproductive timepoints (9- and 18-months), and (C) within each reproductive sampling site (ovary, infundibulum-magnum (Inf-Mag), isthmus-uterus (Ist-Ute), ceca, and environment (Env)) across timepoints (9- and 18-months).

Analysis of taxonomic abundance

Relative abundance analysis at each location revealed common genera across sites, along with significant shifts between 9- and 18-month-old hens. The fifteen most relatively abundant genera included Alistipes, Bacteroides, Clostridia_UCG-014, Clostridia_vadinBB60_group, Faecalibacterium, Fusobacterium, Gastranaerophilales, Lactobacillus, Limosilactobacillus, Mucispirillum, Rikenellaceae_RC9_gut_group, Ruminococcus_torques_group, Staphylococcus, and uncultured taxon (Fig. 5).

Fig. 5.

Taxonomic Bar Plots of Top 15 Genera. Taxa proportions were analyzed by relative abundance for the top 15 genera. Replicates within each group were averaged and separated by both sample age (9- and 18-months) and location (ovary, infundibulum-magnum (Inf-Mag), isthmus-uterus (Ist-Ute), ceca, and environment (Env)).

Across age groups, Lactobacillus and Bacteroides consistently ranked first and second in relative abundance within the oviduct, respectively (Fig. 5). These genera have been widely reported, suggesting that they are core members of the hens’ mature reproductive microbiome (Cheng, et al., 2024; Lee, et al., 2019; Lee, et al., 2020; Shterzer, et al., 2020; Su, et al., 2021; Wen, et al., 2021). Although Clostridia and Staphylococcus showed no significant differences in our dataset, they have been reported across the reproductive tract and were linked to eggshell speckling and darker eggshell pigmentation, respectively (Cheng, et al., 2024; Shterzer, et al., 2024). Similarly, Alistipes and Faecalibacterium have been previously detected in the hen oviduct but did not exhibit significant differences in this study (Cheng, et al., 2024; Lee, et al., 2019; Shterzer, et al., 2020; Shterzer, et al., 2024).

Several identified bacterial taxa exhibited significant age-associated shifts in relative abundance across the hen reproductive tract. Fusobacterium, previously reported in the chicken reproductive tract, which is known to be differentially abundant depending on the hen breed (Su, et al., 2021; Wen, et al., 2021), exhibited significant relative decreases in our dataset between 9- and 18-month-old hens across all locations: Ovary, Inf-Mag, Ist-Ute, and Ceca (P < 0.05). Changes in other genera were more localized. Limosilactobacillus and Mucispirillum significant relative decreases in various sites while Rikenellaceae_R9_gut_group and Ruminococcus_torques_group demonstrated significant relative increases (P < 0.05) (Fig. 5).

ANCOM-BC2 differential abundance and correlation analysis

As this was the first study to analyze the layer hen ovarian microbiome, the ovary was evaluated independently using ANCOM-BC2 for differential abundance at the genus level. Due to the lack of significant beta-diversity across the reproductive tract continuum, samples from the reproductive sites (Ovary, Inf-Mag, and Ist-Ute) were collapsed and analyzed collectively for differential abundance between age groups at both the genus and family levels. To link taxa of interest back to reproductive efficiency, Spearman rank correlation analysis was conducted against lay performance.

In this first of a kind study investigating the layer hen ovarian microbiome we identified significant age-related shift between 9- and 18-month-old hens (Fig. 6A). Eubacterium_ventriosum_group was significantly less relatively abundant in the 18-month-old hens compared to the 9-month-old hens. Genera that were significantly more relatively abundant in the 18-month group included Turicibacter, Ligilactobacillus, Cerasicoccus, Candidatus_Saccharimonas, CHKCI001, and Ruminococcus_torques_group (Fig. 6A). Among these, only Turicibacter remained significant following an FDR adjustment (P < 0.05). Spearman correlation analysis identified significant (FDR, P < 0.05) moderate to strong negative associations between lay performance and the relative abundance of several ovarian genera: Ligilactobacillus (ρ = -0.88), CHKCI001 (ρ = -0.73), Ruminococcus_torques_group (ρ = -0.73), Turicibacter (ρ = -0.71), UCG-008 (ρ = -0.62), and Cerasicoccus (ρ = -0.61) (Fig. 7A).

Fig. 6.

ANCOM-BC2 Differential Abundance Bar Plots. Individual significant taxa across different ages (9- and 18-months) were determined for (A) the ovary and (B) the entire reproductive tract. All taxa are significant at an unadjusted P-value <0.01 and at an FDR adjusted P-value (P <0.05) (indicated with an asterisk).

Fig. 7.

Spearman Correlation Analyses. The strength of association between genera of interest (derived from ANCOM-BC2 analyses) and lay performance (response variable) was determined. Genera denoted with an asterisk were significant at an FDR adjusted P-value (P <0.05).

Collectively across the reproductive tract, sixteen genera were found to be significantly differentially abundant between 9- and 18-month-old hens (Fig. 6B). Genera that significantly (FDR, P < 0.05) decreased between 9- and 18-month-old hens included Akkermansia, Bacillus, Prevotellaceae_UCG-001, Megasphaera, Fusobacterium and Mucispirillum. Genera that significantly (FDR, P < 0.05) increased included Ligilactobacillus, Aeriscardovia, CHKCI001, Acinetobacter, F082, Candidatus_Saccharimonas, UCG-008, NK4A214_group, and Ruminococcus_torques_group. Spearman correlation analyses identified significant (FDR, P < 0.05) moderate to strong positive correlations between lay performance and the genera Fusobacterium (ρ = 0.58) and Megasphaera (ρ = 0.56). Significant (FDR, P < 0.05) moderate to strong negative correlations were found with Ligilactobacillus (ρ = -0.73), CHKCI001 (ρ = -0.64), NK4A214_group (ρ = -0.56), and Ruminococcus_torques_group (ρ = -0.52) (Fig. 7B).

Prediction of microbial metabolic functions using PICRUSt2

PICRUSt2 software was employed to predict potential functional impacts of relative microbial shifts observed between 9- and 18-month-old hens. Predicted functions included pathways related to cell growth and death, biosynthesis of secondary metabolites, lipid metabolism, and xenobiotic biodegradation and metabolism. In the aged 18-month-old cohort, enrichment was predicted for the p53 signaling pathway (P = 0.009), stilbenoid, diarylheptanoid, and gingerol biosynthesis (P = 0.007), ether lipid metabolism (P = 0.003), and fluorobenzoate degradation (P = 0.008). In contrast, the younger 9-month-old cohort showed predicted enrichment for steroid biosynthesis (P = 6.79 × 10⁻⁵) and steroid hormone biosynthesis (P = 0.001) (Fig. 8). Steroid biosynthesis and steroid hormone biosynthesis pathways were found to still be significant at FDR-adjusted P-values, 0.009 and 0.04 respectively.

Fig. 8.

PICRUSt2 Metabolic Prediction. Pathways predicted by PICRUSt2 are indicated. All indicated pathways are significant at an unadjusted P-value <0.05 and FDR adjusted P-value (P <0.05) (indicated with an asterisk).

Discussion

Understanding the microbial landscape of the hen oviduct is essential not only for improving food safety but also for advancing hen health and reproductive performance. Here, we show for the first time that the hen ovary also harbors microbiota on its surface, and that these communities are likely seeded by the oviductal continuum. We also observed significant shifts in both ovarian and oviductal microbial composition between 9- and 18-month-old hens. These shifts with age may reflect or contribute to changes in microbial-host interactions that ultimately influence lay efficiency and reproductive performance.

Consistent with prior literature and our own results, alpha and beta diversity were not significantly different across the oviductal continuum from uterus to infundibulum (Shterzer, et al., 2020; Su, et al., 2021; Wen, et al., 2021). Importantly, our novel characterization of the ovarian microbiome revealed a similar lack of beta diversity between the ovary and oviduct. This suggests that the ovary, like the rest of the oviduct, is seeded by the continuum (Escallon, et al., 2019; Lee, et al., 2020). However, the functional impact of microbial communities on ovarian physiology and reproductive success remains poorly understood.

Most notably, we observed significant differences in beta diversity by age across all reproductive tract sites, underscoring the influence of aging on microbial community composition and its potential link to changes in lay performance. Like the human reproductive tract, the hen oviduct was enriched in Lactobacillus (Ravel, et al., 2011), although to a lesser extent. Previous work has suggested associations between increased Lactobacillus and Bacteroides in the digestive and reproductive tracts and enhanced egg production (Su, et al., 2021); however we found no significant correlations between these genera and lay performance in our dataset.

Differential abundance analysis, paired with Spearman correlations, provided insights into how age-related microbial fluctuations are associated with lay performance. Several differentially abundant genera – CHKCI001, F082, UCG-008, Cerasicoccus, Aeriscardovia, Candidatus_Saccharimonas, and NK4A214 – have few, if any, documented associations with the female reproductive tract (Bi, et al., 2024; Yang, et al., 2023b). In contrast, among the differentially abundant genera with known relevance to reproductive biology, several have established roles in steroid hormone activity. Metagenomic functional prediction via PICRUSt2 revealed a significant decline in both steroid biosynthesis and steroid hormone biosynthesis pathways between 9- and 18-month-old hens. Four steroid-related genera – Acinetobacter, Ligilactobacillus, Bacillus, and Akkermansia – were of particular interest due to their diverse and sometimes conflicting roles in steroid hormone synthesis, degradation, and regulation.

Acinetobacter is a genus linked to various reproductive pathologies with implications for host fertility and general health (Canha-Gouveia, et al., 2023; Liang, et al., 2023; Zhang, et al., 2024a). It includes Acinetobacter baumannii, a well-known global pathogen prevalent in persistent infections (Ibrahim, et al., 2021). The genus also plays a role in steroid hormone degradation, particularly of estradiol (E2), the most potent form of estrogen (Plesiat and Nikaido, 1992; Qiu, et al., 2019; Vilela, et al., 2020; Yang, et al., 2011). Its negative correlation with lay performance suggests a shift toward a lay-unsupportive microbiome or the emergence of reproductive tract dysfunction (Cazzaniga, et al., 2022; Fredrickson, 1987; Manarolla, et al., 2011).

Ligilactobacillus is recognized for its roles in pathogen suppression, pH regulation, and anti-inflammatory responses (Yang, et al., 2023b; Yang, et al., 2024). Ligilactobacillus species, Ligilactobacillus Salivarius (formerly Lactobacillus salivarius) and Ligilactobacillus agilis (formerly Lactobacillus agilis) have previously been reported to be dominant species in Lactobacillus populations of the layer hen oviduct (Lee, et al., 2019). Ligilactobacillus is known to participate in estrogen degradation and conjugation functions (Aragon, et al., 2024), potentially reducing circulating estrogen levels. Prior reports from Mehlhorn et al. directly link increased circulating estradiol (E2) with improved lay performance (Mehlhorn, et al., 2022). A negative correlation between Ligilactobacillus and lay performance in our study suggests that while Ligilactobacillus population increases may support host homeostasis and aging-related immune responses, this may come at a cost to reproduction.

Bacillus spp. including B. subtilis, are similarly implicated in estrogen metabolism. Some species can degrade E2 into less active estrone (E1) and inactivate it entirely (Jiang, et al., 2010). However, species-specific differences exist: certain Bacillus strains produce β-glucosidases that may either inactivate or, conversely, activate estrogens (Kuo, et al., 2006; Ojanotko-Harri, et al., 1991). Notably, Bacillus supplementation has been reported to improve egg production and quality in laying hens and ducks (Cao, et al., 2022; Zou, et al., 2021). The increased abundance of Bacillus in this study highlights the need for species-level resolution to determine its functional role, particularly as some Bacillus species are closely linked to enhanced antioxidant capacity and immune performance in poultry (Qiu, et al., 2023; Zou, et al., 2022).

Among all steroid-related genera, Akkermansia is the only one directly associated with steroid biosynthesis and it is reported to be highly estrogen-responsive (Sakamuri, et al., 2023). Although its mechanism of action remains unclear, its abundance has been shown to increase with higher E2 levels. Beyond hormone interactions, Akkermansia also contributes to lipid metabolism – potentially improving egg quality (Wei, et al., 2022) – and mucin degradation in both intestinal (Derrien, et al., 2004) and reproductive tract environments (Takada, et al., 2020).

Together, these findings suggest that while some genera, such as Acinetobacter, may impair reproductive function, while others like Ligilactobacillus, Bacillus, and Akkermansia may represent adaptive shifts in the microbiome that promote health and immune-related functions over reproduction in aging hens. Importantly, not all differentially abundant genera with age contributed either directly or indirectly to steroid-hormone related processes. Ruminococcus_torques_group, Fusobacterium, and Mucispirillum have been reported to play significant roles in intestinal mucin degradation, inflammation response, and biofilm support.

Like Akkermansia, the genus Ruminococcus_torques_group is primarily recognized for its ability to degrade mucins, primarily Mucin 2, in the gastrointestinal tract. Mucin 2 (MUC2), the dominant secretory mucin of the lower intestinal lining, plays a critical role in barrier protection, lubrication, nutrient transport, and pathogen exclusion (Breugelmans, et al., 2022; Gustafsson and Johansson, 2022; Reznik, et al., 2022). Dai et al. 2023 reported increased MUC2 levels in the aged hen uterus, suggesting either age-related compromise of the digestive-reproductive tract segregation or an upregulation of uterine mucosal MUC2 despite that it is not a dominant reproductive mucin (Ariyadi, et al., 2012; Dai, et al., 2023; Jiang, et al., 2019). The increased abundance of Ruminococcus_torques_group at 18-months may be indicative of increased MUC2 availability in the aged reproductive tract (Salyers, et al., 1977; Tailford, et al., 2015). These changes could reflect a shift in mucosal homeostasis that either contributes to or coincides with a lay-unsupportive microbiome.

In contrast, Fusobacterium and Mucispirillum were found to decrease in abundance with age but were positively correlated with lay performance. Both genera have been implicated in inflammatory disorders, tumorigenesis, and metabolic dysregulation (Chen, et al., 2024; Huang, et al., 2024; Lan, et al., 2022; Mao, et al., 2023; Sheldon, et al., 2004; Yu, 2024). In poultry, dietary supplementation studies have demonstrated a positive association between Fusobacterium abundance and improvements in reproductive performance, egg quality, and antioxidant capacity (Kelly, et al., 2018; Liu, et al., 2022; Wang, et al., 2020; Tomkovich, et al., 2017). These findings suggest that Fusobacterium – or specific species within the genus – may play unrecognized beneficial roles in hen reproductive physiology and performance. Similarly, Mucispirillum has been shown to provide protective effects against Salmonella enterica serovar Typhimurium (S. Tm), a major cause of foodborne illness and reduced egg production in layer hens (Herp, et al., 2021; Okamura, et al., 2010). In our dataset, Mucispirillum abundance declined in the oviduct – an anatomical region previously identified as a reservoir for Salmonella (McWhorter and Chousalkar, 2016; Raspoet, et al., 2011). Importantly, poultry supplementation studies have linked increased Mucispirillum populations to enhanced egg quality traits (Yuan, et al., 2024), suggesting that its decline in aged hens could compromise both hen health and food safety. As with the steroid-hormone associated taxa, age-associated shifts in Ruminococcus_torques_group, Fusobacterium, and Mucispirillum may either drive or reflect the transition toward a lay-unsupportive microbiome.

Lastly, characterization of the ovarian microbiome revealed a significant increase in Turicibacter abundance, which was negatively correlated with lay performance. While Turicibacter has previously been reported in the layer hen cloaca and vagina (Wen, et al., 2021) this is the first study to identify its presence in the upper oviduct, specifically the ovary. Turicibacter has been identified as a prominent eggshell-associated microbe (Maki, et al., 2020) and is implicated in contributing to undesirable speckling phenotypes on eggshells (Cheng, et al., 2024). In other models, Turicibacter abundance has been shown to affect oocyte quality through the gut-ovary axis, particularly in light-stressed mice (Li, et al., 2023). Indeed, changes in this genus have repeatedly been associated with stress-responses (Clark and Mach, 2016; Yang, et al., 2022). Given that commercial layer hens are subjected to artificial lighting regimens designed to sustain near-daily oviposition, it is plausible that such practices may induce physiological stress (Archer, 2019; Johnson, et al., 2015; Wichman, et al., 2021), potentially influencing Turicibacter dynamics in the reproductive tract.

Most notably, Turicibacter is widely recognized for its involvement in bile acid metabolism. Interestingly, bile acids have been reported in human ovarian follicular fluid – at higher concentrations than in serum – though the mechanisms behind their presence remain debated (Nagy, et al., 2019). Some reports suggest bile acid synthesis by granulosa cells while others propose uptake from bloodstream circulation (Nagy, et al., 2019; Smith, et al., 2009). Similarly, layer hen egg yolk has been reported to contain both primary and secondary bile acids (Wang, et al., 2023; Zhang, et al., 2024b). Regardless of their origin, Turicibacter is capable of deconjugating bile acids into products that enhance vitamin absorption, particularly vitamin D, which has been linked to increased anti-Mullerian hormone and melatonin levels – both of which improve ovarian function (Li, et al., 2023; Wang, et al., 2024).

Taken together, these functions would suggest that increased Turicibacter and associated bile acid activity should support ovarian function and lay performance; however, our findings indicate a significant increase in Turicibacter abundance that was negatively correlated with lay performance. It remains unclear whether the relative increase in Turicibacter involves alternative pathways that may hinder reproductive performance or represents a compensatory response aimed at supporting ovarian function in aging hens. Our discovery of this genus on the hen ovary opens promising new avenues for understanding microbial influences on fertility and production in commercial laying hens.

Collectively, our findings establish that the oviductal continuum of the layer hen extends to and includes the ovary, and that significant microbiota shifts occur across the productive window. Differentially abundant genera such as Bacillus, Ligilactobacillus, Akkermansia, and Acinetobacter suggest intricate links between reproductive tract microbial composition and steroid hormone metabolism, with potential implications for lay performance. Meanwhile, genera not traditionally associated with steroid pathways – Ruminococcus_torques_group, Mucispirillum, and Fusobacterium – may influence reproductive outcomes through their roles in inflammation, immune modulation, and mucin degradation. Importantly, the novel identification of increased Turicibacter abundance in the ovary and its negative correlation with lay performance highlights the need for further investigation into how the ovarian microbiome may influence folliculogenesis, fertility, and the overall productivity of commercial layer hens.

Data availability

The dataset generated during the current study is available at the NCBI metagenome (Sequence Read Archive) repository, https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1260049.

CRediT authorship contribution statement

Kathryn M Ellwood: Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Validation, Writing – original draft, Writing – review & editing. Ashley E Kramer: Investigation, Writing – original draft. Aditya Dutta: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing.

Disclosures

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests

Kathryn M Ellwood reports financial support was provided by USDA. Aditya Dutta reports financial support was provided by National Institutes of Health. Aditya Dutta reports financial support was provided by University of Delaware Research Foundation Inc. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.