Introduction

Ruminants, including sheep, have a rumen, a primary fermentation chamber that supports a complex ecosystem of microorganisms such as bacteria, archaea, fungi, protozoa, and viruses, as well as substantial populations in the hindgut1. This microbiome works in the rumen and hindgut, converting complex carbs into volatile fatty acids (VFAs), synthesizing vitamins, and detoxifying toxic metabolites2. A balanced diet has been shown to improve the resilience and stability of the gut environment3, however, sudden dietary changes can upset this balance, resulting in dysbiosis (microbiota imbalance)4. The ruminants’ diet significantly affects the composition and function of their hindgut microbiota. Ruminants that graze on high-fiber diets, such as grasses, often their hindgut have a higher population of fibrolytic bacteria, such as Fibrobacter succinogenes and Ruminococcus spp.5,6. However, those cattle reared inside for fattening and fed high-concentrate diets for extended periods may encounter changes in their hindgut microbiota, with a greater abundance of amylolytic bacteria such as Streptococcus bovis and Prevotella spp.7,8. In sheep and other ruminant species, this microbiota can play a role in feed digestion, absorption, immunological function, and general performance. Thus, understanding the link between diet and hindgut microbiota composition is crucial for improving sheep nutrition and health.

Calcium is one of the major minerals that is abundant in pastures and forages, meeting the requirements of grazing sheep9. Dietary calcium is mainly absorbed in the small intestine, specifically the duodenum and jejunum, with absorption impacted by dietary calcium levels, vitamin D status, and the animal’s physiological state (e.g., pregnancy or lactation)9,10. Adequate dietary calcium absorption is required for bone and tooth growth (where 99% of the body’s calcium is stored), as well as for a variety of physiological processes such as successful blood clotting, muscular contraction, cell signaling, and enzyme activity10. However, factors such as long-term indoor rearing and high-grain feeding can disturb calcium homeostasis, potentially causing morphological abnormalities (e.g., rickets) and metabolic diseases (e.g., hypocalcemia)11. Furthermore, the regulation of calcium homeostasis in ruminants is distinct, separating them from monogastric animals while also demonstrating variability within ruminant species12. The hindgut of ruminants harbors diverse microbiota, primarily Firmicutes and Bacteroidetes, which ferment undigested fiber and produce short-chain fatty acids, aiding mineral solubilization and absorption5. This microbiota activity plays a crucial role in mineral utilization, particularly calcium, influencing bone health and metabolic functions. For example, Böswald, Dobenecker12 found that hindgut fermenting animals like Diceros bicornis (Rhinoceros) and Elephas maximus (Asian elephant) absorb more calcium than domestic ruminants like Bos taurus (Cow) and Ovis aries (Sheep).

Recent research has shown that calcium supplementation can affect the makeup and function of the hindgut microbiota of ruminant species such as goat and cattle13,14,15. Calcium can influence the growth and activity of many bacterial populations by changing their pH levels16. High calcium levels may encourage the growth of fibrolytic bacteria that aid in fiber breakdown, resulting in enhanced digestion and nutritional absorption17. Furthermore, calcium’s capacity to bind to bile acids in the intestine reduces their reabsorption, affecting the availability of bile acids18, which some gut bacteria might be able to use as a nutritional source. These variations in calcium digestion imply that dietary calcium levels can impact hindgut microbiota composition. Calcium absorption and interactions with other minerals are well studied in domestic ruminants, including cattle and goats; however, information on ruminal transit and impact on hindgut microbiota composition is limited, particularly in sheep. Moreover, sheep exhibit unique feeding behavior than cattle, characterized by a higher rate of mastication and rumination, resulting in finer feed particle size and a faster digesta passage rate19. These factors may influence calcium solubility and availability along the gastrointestinal tract20, thereby altering microbiota dynamics and nutrient utilization in the hindgut. Keeping all in view, we hypothesized that dietary calcium levels would exert dose-dependent effects on the hindgut microbiota of sheep by shifting hindgut microbiota composition and enhancing microbial energy-harvesting pathways. Moreover, to our knowledge, this is the first study to include data on changes in sheep’s hindgut microbiota population caused by dietary calcium level modulation. It seeks to gain a better knowledge of mineral homeostasis disturbances and the ability of such animals to adjust to limited or abundant mineral supply. The current study aims to evaluate the relationship between dietary calcium levels and hindgut microbiota composition and functional gene prediction in sheep.

Materials and methods

Experimental station and statement of ethical approval

This experiment was conducted at the Yunnan Animal Science and Veterinary Institute (YASVI), Kunming City, Yunnan Province, China (26° 22′ N; 103° 40′ E) after approval by the ethical committee of the Yunnan Animal Science and Veterinary Institute (202009002). Moreover, all procedures were carried out following the approved protocols and guidelines by the State Science and Technology Commission of the People’s Republic of China, 1988, and the Standing Committee of Yunnan Provincial People’s Congress 2007.10). In addition, the current study experimental procedures were in accordance with ARRIVE guidelines detailed by Percie du Sert et al.21.

Animals’ selection, research design, diets, and husbandry practices

For this experiment, thirty male rams of the Yunnan Semi Fine Wool Sheep by considering their age (approximately 10 months each ram) and live body weight (40.37 ± 0.49 kg/ram) were selected from the research farm of the YASVI, randomized into three groups (n = 10 rams/group), and each group was allotted to one of the dietary treatments varying in calcium contents by following the rules of the completely randomized design. The dietary treatments contained 0.53% (Q_1), 0.74% (Q_3), and 0.98% (Q_5) calcium contents on a dry matter basis. The dietary calcium levels were adjusted to lower (0.53%) and higher (0.98%) by considering the calcium requirement (0.74%) of growing sheep according to the Chinese National Feeding Standards. All the dietary treatments were isonitrogenous (Crude protein contents = 10.4%), isocaloric (Metabolizable energy = 9.3 MJ/kg), and contained homogenous fiber (Neutral detergent fiber = 34%) contents. The feed formulation and chemical composition are given in Table 1. The detailed methods followed in carrying out their chemical analysis, and feed-offering protocols are detailed in our recent publication by Ni et al.9. In brief, rams were housed individually in sand-bedded pens, and feed was supplied twice a day (at 08:00 and 17:00 h of the day) with a 10% refusal modification daily for ad libitum consumption, as well as fresh water, which was available around the clock. A 14-day dietary adaptation phase was given, followed by a 30-day feeding trial.

Table 1 Experimental diet formulation and chemical composition on a dry basis.
Full size table

Fecal sampling and preservation

On the 21th day of the feeding trial, fecal samples were collected for nine rams from each treatment by following the sterile protocols detailed in our previous publication Khan et al.10. In brief, 10 g of feces were harvested from the terminal part of the rectum by ensuring a sterile collection, transferred into the prelabeled tube with the tag number of respective ram, and then immediately stored at − 196 °C in a liquid nitrogen-filled tank. The samples were transferred to the department laboratory and stored at the − 80 °C facility until microbial analysis.

DNA extraction, microbial DNA amplification, and sequencing

The preserved feces samples were transferred to Hangzhou Lianchuan Biotechnology Co., Ltd, Hangzhou, China, for microbial analysis. The microbial DNA was extracted from the fecal samples using the E.Z.N.A.®Stool DNA Kit (D4015, Omega, Inc., USA). The protocol provided by the manufacturer was followed to homogenize fecal samples, lyse bacterial cells, and purify the extracted DNA. The quality and concentration of the extracted DNA were measured using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and verified by agarose gel electrophoresis. The V3–V4 region of the 16S rRNA genes was amplified to assess the taxonomic composition of the fecal microbiota, using the primers 341F (5′-CCTACGGGNGGCWGCAG-3′) and 805R (5′-GACTACHVGGGTATCTAATCC-3′), both of which had unique barcodes for multiplexing. The 5′ ends of the primers tagged with specific barcodes were sequenced with universal primers. PCR amplification was performed in a total volume of 25 μL reaction mixture containing 25 ng of template DNA, 12.5 μL PCR Premix, 2.5 μL of each primer, and PCR-grade water to adjust the volume. The PCR conditions to amplify the prokaryotic 16S fragments consisted of an initial denaturation at 98 °C for 30 s; 32 cycles of denaturation at 98 °C for 10 s, annealing at 54 °C for 30 s, and extension at 72 °C for 45 s; and then final extension at 72 °C for 10 min. The PCR products were confirmed with 2% agarose gel electrophoresis. Throughout the DNA extraction process, ultrapure water was used to exclude the possibility of false-positive PCR results as a negative control. The PCR products were purified by AMPure XT beads (Beckman Coulter Genomics, Danvers, MA, USA) and quantified by Qubit (Invitrogen, USA). The amplicon pools were prepared for sequencing, and the size and quantity of the amplicon library were assessed on the Agilent 2100 Bioanalyzer (Agilent, USA) and with the Library Quantification Kit for Illumina (Kapa Biosciences, Woburn, MA, USA), respectively. The libraries were sequenced on the NovaSeq PE250 platform.

Bioinformatics and statistical analysis

Raw Illumina NovaSeq paired-end sequences (LC-Bio) were demultiplexed, merged (FLASH), and quality-filtered (fqtrim v0.94) before chimera removal (Vsearch v2.3.4). High-resolution Amplicon Sequence Variants (ASVs) were generated using DADA2. After rarefaction, alpha diversity (observed ASVs, Chao1, Shannon/Simpson, Good’s coverage) and beta diversity (PCoA/NMDS with Bray–Curtis/UniFrac) were analyzed in QIIME2, with statistical validation (PERMANOVA, Wilcoxon tests). Relative abundance analysis (QIIME2) identified dominant bacterial taxa by calculating mean proportional abundances at phylum/genus levels, with significance testing via ANCOM for differentially abundant features by adjusting the p-value at p < 0.05. Bray–Curtis dissimilarity matrices (QIIME2) visualized through PCoA plots revealed sample clustering patterns, validated by PERMANOVA (999 permutations) to test group separation significance. Venn diagrams (R ‘VennDiagram’ package) quantified shared/unique ASVs among groups. Taxonomic classification employed SILVA 132 (BLAST-aligned), while functional potential was predicted via PICRUSt2, with all visualizations generated in R (v3.5.2).

Results

16S rRNA-based sequencing

A total of 1,624,912 high-quality ASVs of the 16S rRNA gene sequences were generated from 27 samples, with each group (9 samples/group) contributing 524,677 (Q_1), 542,196 (Q_3), and 558,039 (Q_5) sequences, respectively. To minimize depth screening biases, 1,909,270 sequences were randomly subsampled across all samples. Amplicon Sequence Variants were rarified at a sequencing depth of 5755 sequences per sample, using a 100% sequence similarity threshold. Representative sequences from these ASVs were taxonomically assigned to 23 bacterial phyla, 47 classes, 98 orders, 181 families, and 490 genera. The Venn diagram (Fig. 1) shows a total of 31,854 ASVs, with 9840 (30.89%), 10,035 (31.50%), and 11,979 (37.60%) ASVs belonging to the Q_1, Q_3, and Q_5 groups, respectively. The Q_1 group shared 2161 (13.19%) and 2060 (11.24%) ASVs with the Q_3 and Q_5 groups, out of their combined gene pools of 16,373 and 18,321 ASVs, respectively. Similarly, the Q_3 and Q_5 groups shared 2383 (12.86%) ASVs from their combined total pool of 18,516 ASVs.

Fig. 1
figure 1

Venn diagram illustrating the shared and unique fecal microbial amplicon sequence variants (ASVs) among Yunnan Semi-Fine Wool rams fed diets with different calcium levels: Q_1 (0.53% Ca, light green), Q_3 (0.74% Ca, red), and Q_5 (0.98% Ca, yellow). Each circle represents the ASVs detected within a specific dietary group, while overlapping regions indicate ASVs common between two or all three groups.

Full size image

Gut microbiota richness and composition

The rarefaction curve analysis was carried out to estimate the diversity indices (Fig. 2). Rarefaction curves for the indices Chao1 (Fig. 2A), Goods_coverage (Fig. 2B), Observed ASVs (Fig. 2C), and Simpson (Fig. 2D) reveal the microbiological diversity across feces samples. The alpha diversity analysis revealed distinct patterns across calcium level groups (Table 2), with Q_5 demonstrating both the highest species richness (observed ASVs: 1479–2371; Chao1: 1498–2435) and most consistent evenness (Shannon: 8.74–9.63; Simpson: 0.99–1.00). While Q_1 showed moderate richness (1146–1954 ASVs) and stable diversity indices (Shannon: 7.90–9.00), Q_3 exhibited greater variability in both richness (1161–2283 ASVs) and diversity (Shannon: 7.22–9.55), particularly with one sample (Q_3_8) showing markedly lower diversity (Shannon: 7.22). All groups achieved complete coverage (Goods_coverage = 1.00), confirming sufficient sequencing depth. These results suggest a potential positive relationship between calcium levels and microbiota diversity, with Q_5 supporting the most robust and stable communities, while Q_3’s intermediate position indicates possible transitional dynamics in community structure. The number of observed ASVs did not differ significantly among the groups (p > 0.05) (Fig. 3A). However, comparisons of the Shannon diversity index (Fig. 3B) and Good’s coverage (Fig. 3C) revealed significant differences among the three groups (p < 0.052, Kruskal–Wallis test). Pairwise comparisons using the Wilcoxon test showed significant differences in alpha diversity metrics between calcium level groups. The Shannon and Simpson diversity indices demonstrated significant contrasts between Q_5 and both Q_1 (Shannon: p = 0.01; Simpson: p = 0.02) and Q_3 (Shannon: p < 0.001; Simpson: p = 0.01). The most pronounced differences appeared between Q_3 and Q_5, which were significant for both diversity metrics (p < 0.01). Goods coverage also differed significantly between Q_1 and Q_5 (p = 0.02), though other comparisons for this metric were non-significant. These results suggest that while species richness (observed ASVs and Chao1) remained similar across different calcium groups, diversity measures incorporating species evenness (Shannon and Simpson) showed meaningful variation, particularly between the highest (Q_5) and other calcium level groups.

Fig. 2
figure 2

Refraction curve analysis of fecal microbiota in Yunnan Semi-Fine Wool rams fed different dietary calcium levels: Q_1 (0.53%), Q_3 (0.74%), and Q_5 (0.98%) for the indices Chao1 (A), Goods_coverage (B), Observed ASVs (C), and Simpson (D).

Full size image
Table 2 Tabulated summary (average per group).
Full size table
Fig. 3
figure 3

Alpha diversity analysis of fecal microbiota in Yunnan Semi-Fine Wool rams fed different dietary calcium levels: Q_1 (0.53%), Q_3 (0.74%), and Q_5 (0.98%). Violin plots represent (A) Observed AVSs, (B) Shannon Diversity Index, and (C) Good’s Coverage, highlighting the distribution and variation of microbial diversity within each dietary group. Statistical comparisons between groups were conducted using Kruskal–Wallis test, with significance set at p ≤ 0.05.

Full size image

Beta diversity analysis was performed for the pairwise comparison to examine the differential fecal microbiota of different sheep groups fed varying dietary calcium levels. Principal Coordinate Analysis (PCoA) and Non-Metric Multidimensional Scaling (NMDS) using both unweighted and weighted metrics indicated clear distinctions in the fecal microbiota of rams across different dietary calcium levels (Fig. 4A,B). Beta diversity analysis revealed significant compositional differences, as evidenced by the PCoA plot (PERMANOVA, p = 0.001), indicating strong separation between groups along the MDS1 and MDS2 axes. The first PCoA axis (PCoA1) explained 16.98% and 21.28% of the variance in the two subplots, with coordinate values ranging from − 0.1 to − 0.4, suggesting distinct clustering patterns among groups. NMDS further supported these findings, yielding a stress value of 0.17, which reflects a reliable representation of the data’s underlying structure, while MDS1 values (− 0.1 to − 0.5) indicated consistent gradients or group separations. Together, these results demonstrate that calcium level groups exhibit significant heterogeneity in their compositional profiles, highlighting measurable differences in beta diversity. Moreover, the fecal microbiota clustered more tightly in the Q_5 group compared to the Q_1 and Q_3 groups. Relative abundance analysis (Bray–Curtis distance plot) of the top 30 phyla revealed that Firmicutes (60.30 ± 9.05%) was the most dominant phylum, followed by Bacteroidetes (23.45 ± 5.54%), Proteobacteria (4.84 ± 6.03%), Fibrobacteres (4.61 ± 6.15%), and Spirochaetes (3.29 ± 2.06%) (Fig. 5). A total of approximately 490 genera were detected, of which 147 exhibited significant differences in relative abundance among the different groups (p < 0.05). Specifically, 40, 66, and 41 genera showed significant differences between the Q_1, Q_3, and Q_5 groups (Fig. 6). Genus-level analysis identified Ruminococcaceae_UCG-005 (15–25%), Lachnospiraceae_unclassified (10–20%), and Fibrobacter (5–15%) as the most abundant taxa, while species-level resolution highlighted uncultured Ruminococcaceae and Fibrobacter spp. as key contributors (Fig. 7). Notably, Treponema_2 and Prevotella genera exhibited group-specific variations, suggesting potential functional adaptations. Unclassified taxa comprised < 10% of communities, indicating robust classification accuracy.

Fig. 4
figure 4

Beta diversity analysis of fecal microbiota (ASV level) in Yunnan Semi-Fine Wool rams fed diets with varying calcium levels: Q_1 (0.53%), Q_3 (0.74%), and Q_5 (0.98%). (A) Principal Coordinate Analysis and (B) Non-metric Multidimensional Scaling plots visualize the microbial community structure among dietary groups. Clustering patterns reflect compositional differences in microbial communities across treatments.

Full size image
Fig. 5
figure 5

Relative abundance analysis of bacterial phyla in fecal samples from Yunnan Semi-Fine Wool rams fed different dietary calcium levels: Q_1 (0.53%), Q_3 (0.74%), and Q_5 (0.98%). The stacked bar plot is based on Bray–Curtis distance at the phylum level. Each color represents a distinct bacterial phylum. Relative abundance is expressed as the percentage of total sequences within each group, reflecting compositional shifts in the gut microbiota across dietary treatments.

Full size image
Fig. 6
figure 6

Relative abundance analysis of bacterial genera in fecal samples from Yunnan Semi-Fine Wool rams fed diets containing different calcium levels: Q_1 (0.53%), Q_3 (0.74%), and Q_5 (0.98%). The stacked bar chart illustrates the composition of the gut microbiota at the genus level. Each color represents a different bacterial genus. Relative abundance is expressed as a percentage of total sequences within each group, highlighting genus-level shifts in microbial populations in response to dietary calcium variations.

Full size image
Fig. 7
figure 7

Relative abundance analysis of bacterial species in fecal samples from Yunnan Semi-Fine Wool rams fed diets with varying calcium levels: Q1 (0.53%), Q3 (0.74%), and Q5 (0.98%). The stacked bar chart displays the composition of the gut microbiota at the species level. Each color corresponds to a distinct bacterial species, as indicated in the legend. Relative abundance is expressed as the percentage of total sequences within each group, illustrating species-level shifts in microbial populations across dietary treatments.

Full size image

Differential abundance of microbiota

At genera level (Fig. 8), comparative analysis revealed significant microbiota shifts across groups (Q_1, Q_3, Q_5), with Q_5 showing marked enrichment of SCFA-producing genera (Ruminococcaceae_UCG-010: log2FC + 2.3 in Q_5 vs. Q_1; Faecalibacterium: + 1.8 in Q_5 vs. Q_3) and mucin-degrading Akkermansia (+ 1.5 in Q_5 vs. Q_3). Q_3 exhibited higher abundances of Bacteroides (log2FC + 1.6 vs. Q_1) and Prevotella_9 (+ 1.2), while Q_1 retained lipid-metabolizing Eubacterium_coprostanoligenes_group (+ 1.9 vs. Q_5). Notably, Christensenellaceae_R-7_group (associated with leanness) was depleted in Q_5 (− 1.8 vs. Q_1). These trends aligned with phylum-level Firmicutes/Bacteroidetes ratio shifts and were statistically validated (Kruskal–Wallis, FDR < 0.05). At the species level (Fig. 9), Q_5 showed marked enrichment of beneficial SCFA-producers, including Faecalibacterium prausnitzii (log2FC + 2.1 vs. Q_1, p < 0.01) and Roseburia spp., while Q_3 exhibited higher abundance of mucin-degrading Bacteroides uniformis (+ 1.8 vs. Q_1). The Q_1 group retained lipid-metabolizing species like Eubacterium coprostanoligenes (+ 1.9 vs. Q_5). Notably, Akkermansia glycaniphila was uniquely abundant in Q_5 (+ 2.3 vs. Q_3), suggesting enhanced gut barrier function. Rare but significant species (< 1% abundance), including Christensenellaceae R-7 group, showed calcium-dependent patterns (Q_5 depletion: − 1.8 log2FC). The log2 (abundance + 1) transformation captured both dominant (> 5% relative abundance) and rare (< 0.1%) but biologically significant species, highlighting calcium-dependent microbiota selection.

Fig. 8
figure 8

Differential abundance analysis of bacterial genera in fecal samples of Yunnan Semi-Fine Wool rams fed diets with varying calcium levels: Q_1 (0.53%), Q_3 (0.74%), and Q_5 (0.98%). Comparisons were made between (A) Q_1 versus Q_3, (B) Q_1 versus Q_5, and (C) Q_3 versus Q_5. Boxplots display log2 transformed relative abundances of genera that differed significantly between groups based on the Kruskal–Wallis test with false discovery rate (FDR) correction (p < 0.05). These results highlight specific genera that responded to dietary calcium variation.

Full size image
Fig. 9
figure 9

Differential abundance of bacterial species in fecal samples of Yunnan Semi-Fine Wool rams fed diets with different calcium levels: Q_1 (0.53%), Q_3 (0.74%), and Q_5 (0.98%). Boxplots display log2 transformed relative abundances of species showing significant differences among groups (Kruskal–Wallis test, FDR-corrected p < 0.05). Notable taxa include SCFA-producing Faecalibacterium prausnitzii and Roseburia spp. (enriched in Q_5), mucin-degrading Bacteroides uniformis (enriched in Q_3), and lipid-metabolizing Eubacterium coprostanoligenes (enriched in Q_1), indicating functional microbial shifts in response to dietary calcium levels.

Full size image

Relationship between bacterial community and predicted function

According to the PICRUSt2 analysis (Fig. 10), there were significant differences in microbial metabolic pathways across calcium-level groups (Q_1, Q_3, Q_5). Q_5 exhibited marked enrichment of butyrate production pathways, including pyruvate fermentation to butanoate (log2FC + 1.8 vs. Q_1, p = 0.001) and l-lysine fermentation to acetate and butanoate (+ 1.5 vs. Q_1), aligning with its higher abundance of SCFA-producing taxa (Faecalibacterium, Roseburia). Q_3 showed elevated l-lysine biosynthesis I (+ 1.2 vs. Q_5) and chorismate biosynthesis (+ 0.9), consistent with its mucin-degrading Bacteroides dominance. Q_5 showed enhanced energy metabolism (↑aerobic respiration I, p = 0.0004; ↑reductive TCA cycle, + 1.6 vs. Q_1) and lipid catabolism (↑fatty acid β-oxidation, + 1.3), while Q_3 specialized in nucleotide biosynthesis (↑purine/pyrimidine pathways, p < 0.01). In the Q_5 group, the functional genes with significant differences were primarily involved in energy metabolism, including pathways such as 3-phenylpropionate degradation, aspartate super pathway, chondroitin sulfate degradation I (bacterial), fatty acid β-oxidation I, fatty acid salvage, glucose and glucose-1-phosphate degradation, glutaryl-CoA degradation, glycolysis V (Pyrococcus), inosine-5′-phosphate biosynthesis III, l-1,2-propanediol degradation, pyruvate fermentation to butanoate, queuosine biosynthesis, starch degradation V, succinate fermentation to butanoate, and sucrose degradation III. Additionally, the second most abundant functional genes in the Q_5 group were associated with protein metabolism. These included pathways such as acetylene degradation, adenosine nucleotide degradation II, aerobic respiration I (cytochrome c), l-leucine degradation I, l-lysine fermentation to acetate and butanoate, NAD salvage pathway I, the super pathway of heme biosynthesis from glutamate, the super pathway of (R,R)-butanediol biosynthesis, the super pathway of arginine and polyamine biosynthesis, and the super pathway of histidine, purine, and pyrimidine biosynthesis.

Fig. 10
figure 10

PICRUSt2-based functional pathway analysis of fecal microbiota in Yunnan Semi-Fine Wool rams fed diets containing different calcium levels: Q_1 (0.53%), Q_3 (0.74%), and Q_5 (0.98%). The figure compares the relative abundance of metabolism-associated functional genes across groups: (A) Q_1 versus Q_3, (B) Q_1 versus Q_5, and (C) Q_3 versus Q_5. Only pathways showing statistically significant differences (p < 0.05) are indicated with an asterisk (*). Error bars represent the 95% confidence interval, highlighting functional shifts in microbial metabolic potential in response to dietary calcium variation.

Full size image

Discussion

It is well known that feed-associated changes, including variations in macro and micronutrient levels or their density, directly influence gastrointestinal microbiota profiling of small ruminants13. Previous research stated that dietary mineral deficiencies or improper proportions often impair gastrointestinal microbiota, ultimately suppressing feed intake22. Earlier studies have highlighted the importance of major and trace elements in microbiota metabolism in ruminant species23,24,25. Typically, the calcium-phosphorus ratio in dietary dry matter is adjusted to 1:2, respectively, considering the animal’s requirements until now. However, no studies have specifically examined the impact of calcium on hindgut microbiota, particularly in sheep. Therefore, it remains unknown whether different calcium levels influence hindgut microbiota profiling. In the current experiment, an omics approach was employed to assess the association of dietary calcium contents with fecal microbiota profiles and their functions in growing sheep. The alpha diversity analysis revealed distinct patterns across calcium level groups. Moreover, Beta diversity analysis revealed significant compositional differences, suggesting distinct clustering patterns among groups. The fecal microbiota had clear difference between the Q_5, Q_1 and the Q_3 groups. The increasing diversity and richness of fecal microbiota with higher calcium contents are noteworthy, indicating that it affects the normal microbiota composition of the hindgut. These results are consistent with Meng et al.26 as who reported that dietary copper contents changed the gastrointestinal microbiota composition of grazing Mongolian sheep. Similarly, Lee et al.27 documented changes in the gastrointestinal microbiota profile of cattle supplemented with trace minerals containing bolus. A study by Scholz-Ahrens et al.28 found that dietary supplementation with Lactobacillus acidophilus in ovariectomized rats increased intestinal calcium absorption, reduced calcium turnover in bones, and increased bone density by boosting calcium content in bone. Similarly, dietary supplementation with Lactobacillus plantarum has been shown to stimulate osteoblastic activity, leading to greater calcium and phosphorus deposition29. Another study reported that dietary calcium supplementation enhanced the small intestine length and enriched populations of Bacteroides, Prevotella, and Bifidobacterium in the ileum of obese mice30. Chickens fed a diet enriched with calcium, lactose, and probiotics had significantly higher populations of lactic acid bacteria and coliforms in the hindgut31.

Interestingly, ruminants fed precise dietary regimens, which promote higher diversity and richness of specific gastrointestinal microbiota associated with streamlined metabolic pathways, can ensure greater energy efficiency for the host animal32. Conversely, animals fed unbalanced diets may encounter nutrient deficiencies and exhibit a more diverse gastrointestinal microbiota that utilizes a variety of substrates, resulting in the production of metabolites that the host may not efficiently use32. Among the top 30 bacterial phyla, this study found that Firmicutes (60.30 ± 9.049%) was the most abundant, followed by Bacteroidetes (23.45 ± 5.543%), Proteobacteria (4.84 ± 6.025%), Fibrobacteres (4.61 ± 6.147%), and Spirochaetes (3.29 ± 2.058%). These results align with the findings of Tanca et al.33, who also reported an abundance of Firmicutes and Bacteroidetes in sheep, suggesting that these phyla are key players in sheep hindgut ecology. Several studies reported a higher abundance (> 50%) of this phylum, particularly in sheep34,35 and cattle36,37 that were fed a higher concentrate diet. The approximately 60% abundance of Firmicutes in this study is consistent with Tanca et al.33, who reported similar levels in sheep feces. It is well established that Firmicutes utilize fiber-associated carbohydrates such as cellulose, hemicellulose, galactomannan, and xylan as energy sources38. Similarly, members of the Bacteroidetes phylum possess a higher number of genes per genome for polysaccharide lyases and glycoside hydrolases, making them efficient utilizers of dietary starch, pectin, xylan, galactomannan, and arabinogalactan39. The Bacteroidetes phylum is pivotal in plant cell wall degradation and ruminal bioenergetics, as they produce butyrate through the biodegradation of complex polysaccharides40. The relative abundances of these phyla are interconnected and highly susceptible to dietary manipulations. For instance, a higher abundance of Bacteroidetes is observed in the hindgut of the cows fed high-fiber diets41, whereas higher concentrate feeding results in Firmicutes abundance42. The higher abundance of Firmicutes in this study suggests a more concentrated and less fibrous diet, possibly due to the greater availability of fermentable starch. High-starch diets can alter intestinal pH, shifting microbiota from fiber-digesting (members of Bacteroidetes phylum) to starch-utilizing bacteria (opportunistic members of Firmicutes and Proteobacteria phyla)43. In ruminants, members of the Proteobacteria phylum are involved in fermenting soluble carbohydrate fractions44.

In this experiment, approximately 490 genera were detected, with 31 showing significant differences in the relative abundance of fecal microbiota among the three groups of rams. Genus-level analysis identified Ruminococcaceae_UCG-005, Lachnospiraceae_unclassified, and Fibrobacter as the most abundant taxa, while species-level resolution highlighted uncultured Ruminococcaceae and Fibrobacter spp. as key contributors. Notably, Treponema_2 and Prevotella genera exhibited group-specific variations, suggesting potential functional adaptations. The differential analysis of this study revealed distinct calcium-dependent microbial patterns, with Q_5 showing enrichment of beneficial SCFA-producers (Faecalibacterium, Roseburia), and mucin-degrading Akkermansia, with deleted Christensenellaceae. The Q_3 favored Bacteroides and Prevotella. While Q_1 retained lipid-metabolizing Eubacterium. Members of the Ruminococcaceae family, commonly found in the gastrointestinal tract, play a significant role in breaking down the cell wall, simplifying it for host utilization45. The Bacteroides and Prevotella species (as dominated in Q_3 group), which are known to specialize in complex carbohydrate degradation. This finding may reflect an adaptive response to calcium-induced changes in digesta viscosity or bile acid composition, as these factors are known to influence the growth of mucin-degrading bacteria. Bacteroides and Prevotella species are extensive arrays of CAZymes (carbohydrate-active enzymes), particularly xylanases and arabinofuranosidases, enabling breakdown of complex plant polysaccharides such as fiber and thus in this way contribute into short-chain fatty acid production andmaintains gastrointestinal health by regulating microbiota composition, immune modulation, and nutrient metabolism41. Similarly, the Lachnospiraceae family is primarily involved in fermenting fibers, particularly highly lignified and undigested complex carbohydrates, breaking them down into short-chain fatty acids like butyrate, which are associated with various health benefits, including promoting intestinal barrier integrity and reducing inflammation45. The Rikenellaceae family, another group within the Bacteroidetes phylum, is known for producing short-chain fatty acids by fermenting various dietary carbohydrates46. The persistence of lipid-metabolizing Eubacterium species in the low-calcium group (Q_1) suggests that calcium deficiency may favor microbiota communities adapted to alternative energy substrates, possibly through increased epithelial mucus production or altered digesta retention times.

Bacterial richness and diversity are well-known indicators of microbiota functionality47. According to the functional gene analysis, Q_5 exhibited marked enrichment of butyrate production pathways, including pyruvate fermentation to butanoate. This observation aligns with previous in vitro studies demonstrating calcium’s ability to stimulate butyrate production through modulation of microbiota cross-feeding networks. The concurrent increase in Akkermansia abundance suggests that high calcium diets may promote both saccharolytic and mucolytic metabolic pathways in the sheep gut, potentially enhancing energy harvest while maintaining epithelial barrier function. Notably, we observed significant depletion of Christensenellaceae in high-calcium groups, a family increasingly recognized for its role in host metabolic regulation. This finding raises important questions about the potential systemic effects of calcium-mediated microbiota changes in ruminants, particularly regarding energy metabolism and feed efficiency. Most of the functional genes in the Q_5 group with significant differences were primarily involved in energy metabolism pathways. For instance, the 3-phenylpropionate degradation pathway involves a series of enzymatic reactions leading to the conversion of metabolites used by certain microbiota populations as an energy source48. The aspartate supermetabolic pathway is a network of biochemical reactions where the amino acid aspartate serves as a central hub for microbial protein synthesis, nucleotide biosynthesis, and energy metabolism48. The chondroitin sulfate pathway is a complex polysaccharide pathway where certain microbes break down glycosidic bonds, producing oligosaccharides and monosaccharides for bioenergetic reactions48. Similarly, fatty acid β-oxidation I is a catabolic pathway where fatty acids are oxidized into acetyl-CoA, a key component of the citric acid cycle for energy production48. Fatty acid salvage is another pathway that allows cells to reutilize or recycle fatty acids by incorporating them into the main pool of cellular lipids, conserving energy and maintaining fat homeostasis49.

The glucose and glucose-1-phosphate degradation pathways involve enzymatic reactions converting these complex molecules into simpler forms essential for bioenergetic processes such as glycolysis, gluconeogenesis, and related metabolic mechanisms that regulate blood glucose levels20. The chondroitin sulfate degradation I pathway is integral to breaking down sulfated glycosaminoglycans, which are important for maintaining connective tissue health and energy metabolism20. In contrast, gluconeogenesis II is a metabolic pathway in ruminants where glucose is synthesized from non-carbohydrate substrates like lactate, glycerol, and glucogenic amino acids, ensuring a steady energy supply during fasting or low-carbohydrate intake20. In addition to microbiota functionality, the enzymes involved in these pathways are pivotal in maintaining gastrointestinal homeostasis. For instance, the enzymes responsible for energy and carbohydrate metabolism, such as pyruvate kinase, hexokinase, and enolase, facilitate glycolysis and ATP production, providing energy for microbiota and host physiological processes50. Similarly, enzymes like chondroitinase ABC and chondroitin AC lyase, involved in chondroitin sulfate degradation, help break down glycosaminoglycans into smaller oligosaccharides, ensuring effective nutrient utilization and promoting overall gastrointestinal health51. This study demonstrates that changes in the dietary calcium content can significantly alter hindgut microbiota composition, metabolic functions, and, ultimately, animal performance. The findings underscore the importance of optimizing mineral ratios in ruminant diets to maintain microbiota diversity and functionality, leading to improved nutrient utilization and overall health. However, a key limitation of using PICRUSt2 for functional prediction in ruminants lies in its reliance on reference genomes, many of which are derived primarily from human and model organism-associated microbiomes. Ruminant gut environments harbor a large proportion of uncultured and poorly characterized microbial taxa, which may not be adequately represented in these reference databases. Consequently, the accuracy of predicted metagenomic functions may be limited, particularly for taxa unique to or abundant in the rumen. These constraints should be considered when interpreting the functional inference results, and future studies employing shotgun metagenomics or metatranscriptomics could provide more precise insights into microbial functional dynamics in ruminant systems.

Conclusion

This study demonstrates that dietary calcium levels significantly influence hindgut microbiota composition and function. Feeding higher levels of calcium led to the development of a distinct microbiota profile characterized by SCFA-producing bacteria (Faecaliboccaceae, Roseburia) and enhanced butyrate synthesis pathways in sheep, suggesting potential gut health benefits. Importantly, the robust correlation between taxonomic shifts and predicted metabolic functions highlights calcium’s role as a key modulator of gut ecosystem functionality of the sheep. These findings provide compelling evidence that calcium supplementation may be a viable strategy for targeted microbiome modulation in ruminant species, particularly for enhancing butyrogenic capacity.