The in vitro data, such as gas production, in vitro degradation, kinetics of gas production, total NH₃, volatile fatty acids (VFA), fatty acids composition, were analyzed with RStudio 2023.06.0 Build 421 (Posit Software, PBC, Boston, MA, USA), using analysis of variance (ANOVA) based on the completely randomized design (CRD) (Akhmad, 2019). Normality of data distribution was assessed before analysis. The mathematical model used is a linear model for CRD: Yij = μ + αi + εij, where Yij is the observation value, μ is the overall mean, αi is the treatment effect, and εij is the experimental error. The Duncan multiple-range test was used to determine the significance of differences among the groups. The outcomes were presented as the mean ± standard error of the mean (SEM). Statistical differences were considered significant at p<0.05.

A total of 140 isolates were obtained from selective media (MRS Agar) enriched with LA derived from goat rumen fluid. The pure isolates exhibited rod-shaped and spherical colonies, and microscopic examination confirmed that all isolates were Gram-positive. Additionally, they tested negative for catalase and oxidase, which are characteristic of lactic acid bacteria (Table 3). A phylogenetic tree was generated to visualize the genetic similarities among sequencing results (Figure 1). All 140 isolates were evaluated for CLA synthesis in MRS broth enriched with LA using an active culture technique based on UV spectrophotometry. Following 24 h of incubation, 23 isolates were identified as producing CLA concentrations ranging from 9.0 to 12.3 μg/mL from LA (Figure 2). The highest CLA production was observed in isolate RI-11, whereas the lowest production was recorded for RI-8. Only two isolates (RI-11 and RI-41) produced CLA concentrations exceeding 12.0 μg/mL, while the majority of isolates exhibited CLA production of approximately 9.0 μg/mL in MRS-supplemented broth. Therefore, RI-11 was selected as the potential CLA-producing bacterium to be incorporated into the in vitro gas production technique for evaluating its fermentation characteristics. The positive isolates underwent molecular confirmation, yielding five strains of lactic acid bacteria, as detailed in (Table 3): Lactobacillus fermentum (15), Lactobacillus plantarum (4), Pediococcus acidilactici (2), Lactobacillus argentoratensis (1), and Lactobacillus pentosus (1).

Supplementation with sunflower oil (SFO) and L. plantarum-R11 (LP) affected in vitro degradability and gas production (Table 4). The inclusion of 2% SFO significantly suppressed early gas production at 2 h and 6 h of incubation; the LP and LPSFO groups produced significantly more gas than the control group. Additionally, the LPSFO combination significantly yielded the highest gas production at 72 and 96 h. Simultaneously, the asymptotic gas potential (b) of the LPSFO group was higher compared to other groups. However, the LPSFO group exhibited the lowest IVDMD and IVOMD at 24 h and maintained lower dry matter degradability at 48 h, whereas the SFO group restored degradability to control levels at this time point.

Fermentation parameters are affected by the combination of L. plantarum-R11 and sunflower oil (Table 5). The LPSFO produced the highest ammonia (NH₃) levels, and the pH remained stable. The VFA and ruminal fatty acids profile (Table 6) were not significantly affected by treatments.

Heatmap analysis illustrated differences in relative abundance patterns of selected bacterial taxa at the species level among treatments (Figure 3). The microbial profiles were similar across the group, with the baseline community structure characterized by a higher relative abundance of Ruminococcus. Acinetobacter indicus and butyrate-producing bacteria show higher relative abundance in the LP group. In contrast, a distinct abundance pattern, with Blautia spp. and Selenomonas ruminantium appearing more, was observed in the LPSFO group compared with the other treatment groups. Additionally, a prominent display in Butyrivibrio fibrisolvens in the SFO and LPSFO groups, followed closely by the CON group. B. fibrisolvens is a key microbe involved in ruminal biohydrogenation and CLA synthesis. All observations are descriptive and based on relative abundance patterns derived from heatmap visualization.

A clear difference in community structure between the sample groups in the taxonomic analysis at the species level (Figure 4). Variation in the relative abundances of key taxa, including Streptococcus lutetiensis, Streptococcus macedonicus, Ruminococcus flavefaciens, and unclassified members of Ruminococcus and Prevotella. A substantial unassigned fraction and other categories indicated a considerable degree of microbial diversity that requires further investigation. The compositional changes suggest that experimental treatments exerted significant pressure on the microbial community.

Our present study screened rumen-derived lactic acid bacteria (LAB) from goats for their potential to produce CLA under laboratory conditions. L. plantarum-R11 was evaluated for its effects on in vitro ruminal fermentation and microbial community structure. A total of 140 rumen-derived isolates, 23 LAB strains, were identified to produce CLA concentrations ranging from 9.0 to 12.3 μg/mL after 24 h of incubation. According to Sosa-Castañeda et al. (2015) (Sosa-Castañeda et al., 2015), a higher CLA value was observed when LAB were cultured in skimmed milk supplemented with 2% LA at 36 °C for 48 h. Numerous factors may have contributed to the difference in CLA content produced, including incubation time, LA availability, and culture medium. Skimmed milk provides a more favorable substrate and reduces microbial competition compared with MRS-based systems, while longer incubation allows greater enzymatic conversion of LA to CLA. We identified five distinct LAB species: Lactobacillus fermentum (15 isolates), Lactobacillus plantarum (4 isolates), Pediococcus acidilactici (2 isolates), Lactobacillus argentoratensis (1 isolate), and Lactobacillus pentosus (1 isolate), demonstrating the diversity of CLA-producing bacteria within the LAB of the rumen microbiome. The predominance of L. fermentum among CLA producers aligns with the findings of Tyagi et al. (2020) (Tyagi et al., 2020), who reported that this species possesses robust linoleic acid isomerase activity. LAB strains isolated from the rumen exhibit variation in CLA production efficiency, even within the same species, underscoring the specific metabolic capabilities of each strain (Chen et al., 2024).

These findings align with the need for systematic screening to identify efficient CLA-producing strains rather than relying solely on species-level identification. Based on our preliminary CLA-producing ability in MRS medium, L. plantarum-R11(12.3 μg/mL) was selected using a UV-based spectrophotometric method. However, the introduction of L. plantarum-R11 into the in vitro system did not enhance ruminal CLA accumulation during in vitro fermentation. Still, it did exert measurable effects on gas kinetics and microbial community composition. The UV-based method used during initial isolate screening is a rapid, cost-effective preliminary screening tool. However, it has some limitations in resolving individual CLA isomers compared with gas chromatography. The disparity observed between the MRS medium and the rumen conditions for CLA production might be attributed to the media used in both experiments. The MRS condition provides abundant substrates and minimal microbial competition, thereby favoring LA conversion. In contrast, the rumen environment is surrounded by intense microbial competition and extensive biohydrogenation, which likely explains the lack of increased CLA under rumen-simulated conditions. Previous studies observed discrepancies between pure-culture screening and rumen systems, restating that CLA production observed in laboratory media does not necessarily translate to ruminal conditions (Özer & Kılıç, 2021).

Inoculation with LP and SFO, individually and in combination, influenced in vitro gas production parameters. We observed suppressed gas production at 2 h in the SFO and LPSFO in the earlier phase. Due to the antimicrobial effects of the medium-chain fatty acids present in SFO, which temporarily slow microbial activity during the early fermentation phase, as suggested by Jenkin (1993) (Jenkins, 1993). However, by 6 h, the LPSFO and LP groups marked enhanced gas production compared to the control, suggesting the L. plantarum-R11 promotes microbial adaptation and fermentation despite the presence of dietary oil. Elevated asymptotic gas potential (b) and the highest cumulative gas production at 72 and 96 h were observed in the LPSFO group, indicating improved and sustained fermentation of the insoluble fiber fraction during prolonged incubation. An increase in the asymptotic gas parameter (b) reflects substrate degradation, particularly slowly fermentable components such as cell wall polysaccharides, and its increase indicates improved utilization of residual fibrous substrates. Previous studies reported that gas production increases in the late phase when dietary lipids are combined with microbial additives, reducing lipid biohydrogenation and inhibiting microbial activity in the early phase, thereby promoting fermentation more efficiently at later incubation times (Fitriyah et al., 2024);(Kim et al., 2020). The study suggests a synergistic interaction between L. plantarum-R11 and SFO that enhances long-term fermentative capacity, possibly through selective modulation of the microbial community.

Despite improved gas production, the LPSFO group exhibited lower IVDMD and IVOMD at 24 h than the control. This contradiction between gas production and digestibility reflected a shift in fermentation rather than a reduction in microbial activity. The PUFA’s inhibitory effects on fiber-degrading rumen bacteria might contribute to the reduced degradability in the LPSFO group (Gao et al., 2019);(Jenkins, 1993). Sunflower oil is rich in LA; thus, its presence could disrupt microbial cell membranes and interfere with cellulolytic activity, particularly among gram-positive fiber-degrading species. Interestingly, by 48 h, the SFO group had restored degradability, suggesting that the microbial community had adapted to the oil supplement over time. Moreover, the LPSFO group recorded a high ammonia-nitrogen concentration (13.15 nM), suggesting enhanced protein degradation. The NH₃ level does not exceed the optimal range for microbial growth (5-20 mM), which may support microbial protein synthesis (McDonald et al., 2010). Supplementation with L. plantarum-R11 did not disrupt rumen buffering capacity, which is critical for maintaining optimal microbial function; thus, this is reflected in the stable pH across all treatments (6.67-7.09). The VFA and fatty acid profiles remained stable across the groups.

LAB is not the most dominant member of the rumen microbiota, but its presence might influence microbial interactions under in vitro conditions. Rumen microbial community analysis provides insight into the potential of L. plantarum-R11 to shift microbial communities. The relative abundances of the treatments suggest selective microbial modulation of the rumen ecosystem; however, these differences are descriptive and based on relative abundance patterns rather than formal differential abundance or beta-diversity analyses. Ruminococcus spp. and Prevotella spp are the dominant in the control group, representing fibrolytic and amylolytic bacteria essential for plant polysaccharide degradation (Flint et al., 2008). The LP treatment showed dominance of Ruminococcus, with Acinetobacter indicus and butyrate-producing bacteria also present. The LPSFO group showed a distinct abundance pattern, with S. ruminantium and Blautia spp. appeared more prominent (Figure 4). We observed a stable pH across the treatment, even though L. plantarum-R11 is associated with lactic acid production (Liu et al., 2021);(Lu et al., 2023). Both taxa are associated with lactate utilization, which may partly explain the stable ruminal pH. However, as lactate concentration was not measured, the interpretation is based on a potential mechanism rather than a confirmed effect. The established relationship between lactate-producing and lactate-utilizing bacteria likely contributed to the extended gas production and improved fermentative stability in the LPSFO group. Butyrivibrio fibrisolvens, a key bacterium responsible for CLA production through LA isomerization in the rumen (Hussain et al., 2016), was enriched in both the SFO and LPSFO. The degree of B. fibrisolvens on the relative abundance in oil-supplemented groups confirms that sunflower oil provided sufficient LA substrate to stimulate CLA-producing bacterial populations. However, stable CLA content in the ruminal fatty acid profile at 24 h suggests that either the sampling time did not capture peak CLA levels or that rapid conversion to stearic acid via complete biohydrogenation occurred. The combined supplementation of L. plantarum-R11 and sunflower oil appears to produce a synergistic effect on long-term gas production and microbial community structure.

Overall, the potential mechanism is that L. plantarum-R11 produces antimicrobial compounds, such as bacteriocins, which may selectively inhibit competing bacteria and modulate rumen microbial activities (Urban et al., 2024). This modulation could create a more favorable fermentative ecology characterized by stable pH, reduced microbial competition, and enhanced cross-feeding. It could promote the activities of lactate-utilizing bacteria such as S. ruminantium to limit lactate accumulation and support the activities of fibrolytic and biohydrogenating bacteria, including Butyrivibrio spp. (Guo et al., 2020). However, these mechanisms remain hypothetical because bacteriocin production and interspecies interactions were not confirmed in the study. Alternatively, the presence of L. plantarum-R11 in rumen ecology could enhance the bioavailability of LA from sunflower oil through lipase activity, providing a substrate for biohydrogenation bacteria (Tyagi et al., 2020). The observed changes in microbial community composition validate the hypothesis that L. plantarum-R11 acts as a modulator of rumen ecology rather than only a CLA producer in the complex rumen environment.