Temporal Changes in Ruminal Microbiota Composition and Diversity in Dairy Cows Supplemented with a Lactobacilli-Based DFM

1. Introduction

The rumen of dairy cows is a complex and vital ecosystem responsible for breaking down plant material and extracting essential nutrients for maintenance and production. This process is driven by a diverse microbial community consisting of bacteria, protozoa, phages, anaerobic fungi and archaea. These microbes ferment fibre, produce essential nutrients, and detoxify harmful substances [1,2]. Bacteria, particularly from Bacillota and Bacteroidota, dominate, while archaea and eukaryotes (protozoa and anaerobic fungi) form smaller portions of the community [3]. Most of these microbes reside in the solid phase of the digesta, with genera like Succiniclasticum and Prevotella playing key roles in fermentation [4]. The ruminal microbiota offer valuable insights into digestive health and the microbial dynamics that influence digestion, complementing the understanding provided by the rumen itself [4]. A common nutritional problem in ruminants is reduced feed efficiency due to imbalanced gut microbiota, leading to poor digestion and nutrient absorption. The experimental direct-fed microbial (DFM) formulation addresses this issue by introducing specific strains of probiotics that restore microbial balance, enhance enzymatic activity, and improve fibre breakdown. This targeted intervention promotes better nutrient utilisation, improving feed efficiency and overall animal health [5,6].

Recent research has highlighted the critical role of microbiota in livestock health, significantly influencing digestion, nutrient absorption, and overall production efficiency [3,7]. Zeineldin et al., [7] reported that cows with a wide variety of microbial species in the rumen experienced better overall health and productivity. Furthermore, maintaining microbial diversity through appropriate dietary management and DFM supplementation can enhance immune function and ultimately contribute to the sustainability of dairy farming practices. Ogunade, et al. [8] examined the effects of different dietary interventions on ruminal microbiota composition but not on fermentation end-products or energy status of the experimental steers. The fiber-enriched diet serves as a substrate for microbial growth, and the DFM provides targeted microbial supplementation to maximise the fermentation and digestion process. Interactions between diet and age play a critical role in shaping the composition of the gut microbiota, driving changes in microbial populations [9,10]. Evidence suggests that dietary interventions can effectively modulate microbial communities, promoting improved health outcomes in cattle [11]. Optimising the diet can boost beneficial microbial populations, thereby enhancing digestive efficiency. Feed additives like DFMs are added to modulate the gut microbiota, aiming to enhance the health and productivity of dairy cows [12].

Lactobacilli-based DFM supplementation influenced microbial populations in the digestive tract to potentially enhance fermentation efficiency and nutrient absorption [13]. Further investigation is warranted to better understand the functional implications of these changes and their long-term effects on the health and productivity of dairy cows. Despite its potential advantages in improving ruminant production and health, the impact of DFM supplementation on the rumen microbiota composition and diversity in dairy cows remains largely unexplored. Further research is needed to provide insights on ways to optimise DFM interventions and improve the health and productivity of dairy cows. This study aimed to analyse the microbial communities in the ruminal fluid samples (with digesta) of DFM-supplemented and un-supplemented (control) cows. Through a comparative analysis of these microbial communities over time, the research sought to elucidate the temporal dynamics and identify specific taxa influenced by the DFM supplement. The hypothesis was that supplementation with the DFM would lead to beneficial shifts in the ruminal microbiota by changing the diversity and composition. This study enhances understanding of how dairy cows' ruminal microbiota composition and diversity are altered over time when supplemented with a lactobacilli-based DFM.

2. Materials and Methods

Study Location, Study Herd, and Study Animals

The complete details of the study location, herd, production performances, and study design have been published elsewhere [14]. This longitudinal, negatively controlled study, with a randomised design blocked on treatment groups, was conducted at a commercial dairy farm in Harrisville, Queensland, Australia, from September 2021 to January 2023. The milking herd, comprising approximately 350 Holstein cows, included the randomly selected study cows. The cow herd was managed in two groups for housing, feeding, and milking. Both groups followed a partial mixed ration (PMR) feeding system, where cows received a mixed ration on a covered feed pad during the day and grazed pasture at night. The composition of the mixed ration provided once daily consisted mainly of maize or barley silage, lucerne hay, soybean silage, canola meal, and barley or wheat grain. An additional 1.5 kg of grain was fed twice daily during milking. All cows in the study had free access to both water and their allocated pasture. Typical of Southeastern Queensland dairy production systems, the study cows were managed as two separate groups such that the pasture allowance (up to 6 kg of DM/cow per day) and the PMR feed were sufficient for the maintenance and production requirements of a 600-kg cow producing 35 L of milk/d. The target dry matter intake (DMI) was 21 to 22 kg/cow/day. The pasture consisted of an 80:20 mix of ryegrass (Lolium perenne) and white clover (Trifolium repens). The chemical analysis of the total ration is reported elsewhere [14]. The study farm had 11 well-defined grazing paddocks that were similarly managed concerning grazing time, rotation frequency, and the irrigation program. Each of these 11 paddocks was subdivided along its length to create 22 paired grazing subpaddocks, each approximately 1.5 ha in size. The paired subpaddocks were grazed for approximately 2 d, and then the cows moved to the next pair of subpaddocks according to the grazing rotation program. If required, the grazing period on any pair of subpaddocks was adjusted based on the consumption pattern of the cows. Both groups were housed in a single dry lot partitioned to provide separate feeding and loafing areas and free access to water. The DFM group received an additional 10 mL/cow/day of a DFM supplement (manufacturer recommended dose), top-dressed onto their mixed ration using a 2 L manual pressure sprayer (245 kPa maximum pressure, Aqua Systems Australia). The DFM contained approximately 3.5 × 10⁹ CFU/mL each of three live bacterial strains: Lentilactobacillus buchneri Lb23, Lactocaseibacillus casei Lz26, and Lactocaseibacillus paracasei T9. The control group was not supplemented and received a PMR diet only. For this study, assuming a difference of 50% in microbiome structure, an alpha 2.5%, power 80%, no change in rumen microbiome structure the control group, a change in the rumen microbiome structure in the DFM group of 50%, and a difference of 15% is negligible, the minimum number of animals in each group was 20. We further inflated the sample size by 25% to account for any loss of follow-up. A total of 50 Holstein cows (average body weight 590±67 kg), including both primiparous and multiparous cows, were randomly selected based on parity and days in milk (DIM), and assigned into two groups: control (n = 25) and DFM (n = 25).

Sample Collection

Cows were securely restrained in a chute with their heads manually restrained. An oral speculum, an oro-ruminal sampling tube, and a manually operated pump (Double action hand pump, Wanderer) were used to collect approximately 200 mL of ruminal fluid from each cow approximately 3 hours post-diurnal feeding (9:00 a.m.). Samples were collected at approximately 2-month intervals at 8 points of time throughout the study, covering all stages of lactations. The fluid was collected into an Erlenmeyer vacuum filter flask. Ruminal fluid samples (with digesta) were placed into sterile 5 mL polypropylene flat-bottom tubes (Interpath, Melbourne, Australia) and stored at -20°C before being transported to the laboratory on dry ice for analysis.

Quality Control and Sequence Read Counts

To provide a sufficient representation of the taxa present, a minimum read count of 30,000 for all bacteria and 5,000 for archaea was used. Those samples that fell below the required sequencing depth (n =12 Bac) were re-sequenced. FastQC (Version0.12.0) was used to assess read quality and determine whether trimming was needed (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/).

DNA Extraction and PCR

DNA was extracted from ruminal samples using the Maxwell® RSC Fecal Microbiome DNA Kit (Promega) following the manufacturer’s standard protocols, including a bead beating step on a FastPrep machine at 4m/s for 1 min, twice, with a 5 min break between cycles. The V3-V4 region of the 16S rRNA gene was amplified for all bacteria and archaea (sequences provided in Table 1). Thermocycling conditions were: 95oC/5 min: 30 cycles of 98oC/20 s; 55oC/15 s; 72oC/1 min: hold at 4oC (3-step PCR) for bacteria; 95oC/5 min: 30 cycles of 98oC/20 s; 72oC/1 min: hold at 4oC (2-step PCR) for Archaea. In a second PCR, sequencing indexes were added to the amplicons using 96 forward indexes from the Nextera XT Index 1 plate (Illumina, New York, USA) and three reverse indexes (R97, R98, R99), creating 288 unique index combinations. Negative controls were included in both the target amplification and in the indexing PCRs. Library purification, mixing and sequencing followed the Illumina 16S Metagenomic Sequencing Library preparation document (#15044223). Paired-end sequencing (2 × 300 bp) was performed using the MiSeq Reagent Kit v3 (600 cycle, Illumina).

Statistical and Bioinformatic Analysis

Denoising and trimming the raw data was done using the DADA2 plugin [18] within the QIIME2 platform [19]. DADA2 was then used to produce representative sequences and amplicon sequence variants (ASV), filtering by sample and feature. ASVs were then merged with metadata for further analysis (feature-table). Final representative sequences were compared against the 16S database, GreenGenes2 [20], using the feature-classifier. Multiple sequence alignment was used to group the sequences with the highest homologies, masking was used to remove errors and ambiguous sequences, and a phylogenetic tree was produced using Fasttree (Version 2.1). MicrobiomeAnalyst v2.0 [21] was used for further analysis, with the feature table, taxonomy, metadata and phylogenetic tree files from QIIME2. Alpha and beta-diversity metrics were calculated to assess the microbial diversity within and between samples over time and with or without the DFM supplement. Alpha-diversity analysis measured the Chao1 [22], Observed [23], and Shannon [24] indices. Beta-diversity analysis, via principal component analysis (PCoA; [25]) and non-metric multidimensional scaling (NMDS; [26]), compared the effect of the DFM supplement at the different experimental time points. The linear discriminant analysis effect size (LEfSe) was used to identify which taxa most likely drove the differences between groups. Both alpha- and beta-diversity and LEfSe were performed at the genus level with statistical significance set at a p-value ≤ 0.05.

To test for associations between longitudinal changes in alpha diversity over time and for the different treatment groups, we performed linear mixed-effects (LME) regression analysis for each diversity index. This accounted for subject-specific variation by using cow ID as a random effect whilst allowing identification of longitudinal differences in alpha/beta diversity due to treatment group by using that category as a fixed effect. The LME models were fitted with a first-order autoregressive correlation. Fitted residuals were assumed to follow a normal distribution with a mean of zero and a variance of σ2. Overall model fit was based on The Akaike information criterion (AIC), Bayesian information criterion (BIC) and visual assessment of Pearson’s residuals against fitted values, Q-Q standardised residuals against standardised normal quantiles violated the normality assumption [27]. All analyses used nlme and lme4 [27,28] statistical packages in R (R team) [29].

3. Results

Bacterial Microbiota

The sequencing of the bacterial V3-V4 region of the 16S rRNA gene produced 18,443,231 raw reads, with reads per sample ranging from 32,339 to 233,255 (mean 59,302). For archaea ruminal samples, sequencing generated 3,513,714 raw reads, with reads per sample ranging from 5,011 to 129,376 (mean 8,378). After filtering the mean number of reads per sample was 52, 004 and 7,948 for total ruminal and archaeal samples, respectively. The number of amplified sequence variants (ASVs) identified was 38,140 for ruminal and 1186 for archaeal samples. The proportion of Bacillota over time ranged from 33 to 54%, with Bacteroidota ranging from 42 to 50%. Other phyla that comprised a smaller proportion of the ruminal microbiota included Actinomycetota (2.2%), Fibrobacterota (2.2%), Patescibacteria 3.8%), Pseudomonadota (3.6%), Spirochaetota (2.3%) and Verrucomicrobiota (1%). The relative abundance of ruminal microbial communities in CON and DFM samples across different months from April 2021 to June 2023 was graphed for visual display (Supplementary Figure S1). Overall, there are visible differences in the microbial community structure between CON and DFM groups. Over time, some taxa were consistently more abundant in samples from the DFM group than in samples from the CON group.

Bacterial alpha-diversity (genus level) differed within the CON and DFM samples for Shannon diversity indices at various time points, including September 2021 (p = 0.01), September 2022 (p = 0.04) and January 2023 (p = 0.01). The other indexes (Observed; p > 0.05 and Chao1; p > 0.05) did not differ at any time points from September 2021 to January 2023 (Figure 1; Supplementary Table S1). Ruminal bacterial diversity tested at the genus level using beta-diversity analysis, differed significantly over the study period across six-time points from September 2021 to January 2023 except for April 2022 (p = 0.11) and November 2022 (p = 0.21) (Figure 2; Supplementary Figure S2: Supplementary Table S2).

Figure 3 presents the significant differences in the abundance of various genera between the CON and DFM groups. Genera such as Eubacterium_Q, UBA7741, and Sharpea are significantly more abundant in the DFM group. At the same time, Bacillus_P_294101 and SFMI01 show higher abundance in the CON group with positive LDAscores. The results highlight significant shifts in microbial community structure between the CON and DFM groups, with several genera showing marked differences in abundance (p ≤ 0.05).

Archaeal Microbiota

The abundance data for pooled CON and DFM samples over time identified three families, Methanobacteriaceae (99.3%), Methanomethylophilaceae (0.69%) and Methanosarcinaceae (0.001%) within the Methanobacteriota.

The archaeal alpha-diversity analysis at the genus level from the CON and DFM cows over various time points demonstrated significant differences in the Observed and Chao1 indices in the September 2021 (p ≤ 0.01), April 2022 (p ≤ 0.01) and January 2023 (p ≤ 0.01) samples. Signifcant differences in September 2021 (p ≤ 0.01), December 2021 (p ≤ 0.01), April to September 2022 (p ≤ 0.01) were observed for the Shannon index (Figure 4; Supplementary Table S3).

The beta-diversity analysis at the genus level of archaea identified no significant differences (p > 0.05) at any time point between the CON and DFM cows (Figure 4; Supplementary Table S4, Supplementary Figure S3). The beta-diversity did significantly vary over time.

Figure 4. Alpha diversity analysis (genus level) within the archaea of ruminal fluid of CON (blue) and DFM (orange) cows. Observed (A) and Chao1 (B) and Shannon index (C). For individual time point statistics, refer to Supplementary Table S3. The bars across each box represent the median whilst the top and bottom whiskers represent the upper and lower quartiles respectively.

Figure 4. Alpha diversity analysis (genus level) within the archaea of ruminal fluid of CON (blue) and DFM (orange) cows. Observed (A) and Chao1 (B) and Shannon index (C). For individual time point statistics, refer to Supplementary Table S3. The bars across each box represent the median whilst the top and bottom whiskers represent the upper and lower quartiles respectively.

Figure 5. Beta-diversity analysis (genus level) of the archaea of ruminal fluid over the time points investigated. F-value: 17.7; R²: 0.3; p-value: 0.001; Stress: 0.2. The microbial diversity of the archaea changes significantly over time. F-value: 17.73, R²: 0.27, p-value: 0.001, Stress, 0.16. Panels A and B, respectively, display the principal component analysis and non-metric multidimensional scaling plots of the data.

Figure 5. Beta-diversity analysis (genus level) of the archaea of ruminal fluid over the time points investigated. F-value: 17.7; R²: 0.3; p-value: 0.001; Stress: 0.2. The microbial diversity of the archaea changes significantly over time. F-value: 17.73, R²: 0.27, p-value: 0.001, Stress, 0.16. Panels A and B, respectively, display the principal component analysis and non-metric multidimensional scaling plots of the data.

4. Discussion

In the present study, Firmicutes and Bacteroidetes were the predominant phyla in the rumen fluid samples of dairy cows, with other less prevalent phyla, including Actinomycetota , Fibrobacterota, Patescibacteria, Pseudomonadota, Spirochaetota, and Verrucomicrobiota. These findings are consistent with previous studies [3,30]. Prevotella was the predominant genus (29%) within the phylum Bacteroidota in the rumen, which is crucial for protein degradation and starch utilisation [31]. Henderson et al. [32] noted that while the microbiota of ruminants from different geographic regions varied, a core microbiota could be identified. This core included the seven most abundant bacterial groups, including Prevotella, Butyrivibrio, Ruminococcus, and unclassified members of Lachnospiraceae, Ruminococcaceae, Bacteroidales, and Clostridiales. These groups were present in all samples regardless of species, diet, or geographical location but did vary in abundance. Their study encompassed a variety of ruminant species, including cattle, buffalo, bison, sheep, goats, deer, giraffes, and camelids.

The dominance of Bacillota and Bacteroidota among rumen microbiota sampled in this study is consistent with previous studies [33]. Temporal variations observed in the current study indicated that external factors, such as diet, environment, and management practices, significantly influence microbial communities [34]. Additionally, the lack of significant differences in alpha diversity between DFM-supplemented and control cows indicates that DFM supplementation does not drastically alter overall microbial diversity [35]. This observation is consistent with other studies examining the effect of feed additives like DFMs on ruminal microbiota in which subtle changes are typical rather than any dramatic shifts in microbial diversity [36]. This indicates that the supplement impact is nuanced and more likely to affect specific taxa rather than the entire microbial community [37,38]. Beta-diversity analysis and LEfSe results suggest that the DFM supplement influenced specific microbial taxa [39,40], supporting the conclusion that while overall diversity remained stable, the supplement altered the abundance of certain bacteria. Such changes in microbial composition could impact fermentation processes and nutrient absorption in the cows. Consequently, optimising the digestive processes has the potential to improve cattle health and productivity. Finally, the beta-diversity analysis and LEfSe results indicate that the DFM supplement influences the composition of specific microbial taxa within the rumen, as demonstrated by genera like Prevotella, which had significant changes in abundance. The result aligns with prior research suggesting that DFM selectively promote or inhibits specific microbial populations, thereby impacting the overall fermentation process [41]

Recent studies have supported similar effects in response to various dietary interventions and feed additives [42,43]. These outcomes underline the dynamic responses of microbial communities to external interventions and bring to focus the risk of significant implications for animal health (enhanced nutrient absorption and immune responses). This emphasises the importance of comprehending and optimising the rumen microbial balance in ruminant nutrition [7]. The observed changes in specific microbial taxa in this study, such as the increase in Prevotella abundance among DFM-supplemented cows, may carry significant functional implications with the potential to influence enhancing complex carbohydrate degradation and improve nutrient absorption [31,44]. Similarly, alterations in the abundance of other key genera (such as Ruminococcus and Succiniclasticum) could have implications for the breakdown of fibre and the production of short-chain fatty acids crucial to the energy metabolism of dairy cows [45,46]. The findings underline the intricate complexity of the interaction between microbial composition and metabolic processes in the rumen and serve further to highlight the importance of understanding and optimising microbial communities to achieve enhanced animal performance [47,48].

Archaea constituted approximately 2.5% of the ruminal microbiota [49], higher than that observed in the present study. Previous research has indicated that the bacterial community in the rumen is taxonomically richer than the ruminal archaea, reflecting a pattern of similar and limited diversity of the archaeal populations when compared to bacteria, also found in the current study [32]. The archaeal microbiota, particularly members of the Methanobacteriaceae family, play a pivotal role in ruminal fermentation by driving methanogenesis, which balances hydrogen levels but contributes to energy losses and greenhouse gas emissions [50]. Exploring the impact of DFM supplementation on archaeal populations is essential, as DFMs may alter substrate availability or interspecies interactions, potentially mitigating methane production and improving rumen efficiency [51].

The temporal changes observed are crucial to understanding the adaptability and resilience of the microbial community in response to external interventions, like the DFM supplement. This study unveiled significant temporal variations in microbial diversity, consistent with previous research. This finding highlights the dynamic nature of the ruminal microbiota, and the potential influence of diet, environment, and management practices. These fluctuations suggest that the relative abundances of the core rumen and ruminal microbial groups fluctuate over time, in turn, potentially affecting the fermentation efficiency and overall health of cattle [52,53]. Therefore, continuous monitoring and effective management practices are required to optimise rumen microbial balance. Further research should explore the functional implications of these microbial shifts to elucidate the long-term effects of DFM supplements on dairy cows. Comparative studies across various breeds, diets, and geographic regions would provide a more comprehensive understanding of the core microbiota and the drivers of the fluctuations. Valuable insights into the factors influencing microbial dynamics in dairy cows may be discovered, allowing improved tailoring of management practices that optimise animal health and productivity compatible with local breed(s), geography, ration ingredients, and other inputs. The findings of this study offer valuable insights into how the DFM supplement often affects the microbiota composition and diversity in dairy cows. There are significant potential implications for animal health and productivity due to improved knowledge of the complexity of the ruminal microbiota. All this serves to emphasise the potential role of DFMs in the planned and controlled modulation of microbial communities in livestock.

To advance this study's findings, future research should prioritise longitudinal studies investigating the persistence of any effects of the DFM supplement on microbial community structure and function. Employing metagenomic and metabolomic approaches for functional analysis would offer deeper insights into the functional roles of the affected microbial taxa. Comparative studies across diverse breeds, diets, and geographic locations would facilitate a comprehensive understanding of the core ruminal microbiota and its variations, elucidating broader patterns and influences. Furthermore, the study focused on temporal variations in microbial diversity, but the study design did not account for potential seasonal effects or diet changes during the 18-month period that could influence microbial composition. While 16S rRNA sequencing provides insights into microbial taxonomy, it does not capture functional activities or metabolic pathways, limiting understanding of how microbial shifts impact cow health and production. Furthermore, not all microbial communities were identified using current databases. Expanding microbial databases and improving taxonomic classification methods are essential for accurately characterising microbial communities and their functional roles. In particular, upgrading bacterial and archaeal databases is crucial for better understanding fungal microbiota. Collectively, these efforts would enhance our understanding of how DFM supplementation shapes microbial communities, influences functional dynamics, and informs targeted strategies to optimise animal health, productivity, and welfare in cattle production systems. Finally, the results of this study can only be extrapolated to dairy herds similar to the study animals with similar settings. Although the study has strong internal validity, the external validity is limited. Therefore, the results cannot be generalised to the larger dairy cow populations.

5. Conclusions

This study demonstrated that a lactobacilli-based DFM supplement over an extended period alters the composition and diversity of the ruminal microbiota in dairy cows. The temporal change of ruminal microbiota was mainly explained by calendar month, which highlights the dynamic nature and potentially other factors that influenced temporality that were not accounted for in this study. Future research using metagenomic and metabolomic approaches and comparative studies across different breeds, diets, and geographic locations is recommended for a better understanding of the supplement's impact.