Effect of a co-fermented amino acid waste liquid-rice bran feed as a soybean meal replacer on growth performance, rumen fermentation, and microbial profile in finishing cattle

1 Introduction

Driven by population growth and increasing economic urbanization, the production and consumption of beef have continuously increased over the last few decades (Chung and Hadi, 2025). As a core component of animal husbandry, the beef cattle industry serves as a critical source of animal proteins, meeting the high-protein dietary requirements of over one billion people globally (Mizoguchi, 2024). In China, the rapid growth in beef consumption is powerfully meeting the demand for high-quality protein driven by the upgrading of residents’ dietary structure (Zhu et al., 2022). Available statistics indicate that livestock and poultry products had contributed approximately 32% of the dietary protein supply for urban and rural residents in China by 2022. Among all meat categories, beef consumption recorded the fastest growth, with a cumulative increase of 63.4% from 2010 to 2024. According to USDA’s projections, China’s per capita beef and mutton consumption is expected to increase by approximately 33%, and household purchases by 163%. between 2021 and 2031 (Gale and Dong, 2023).

SBM is one of the most widely used protein sources in beef cattle diets (Nannucci et al., 2021). However, its increasing price and dependence on external supply have stimulated interest in alternative protein resources. At the same time, the efficient utilization of agro-industrial byproducts has become increasingly important for improving feed resource sustainability and reducing environmental pressure in livestock production (Shahid et al., 2025).In this context, reducing reliance on conventional feed sources and exploring cost-effective alternatives are considered critical strategies to enhance the economic efficiency of livestock production.

As a crucial industrial raw material, amino acids are extensively utilized in pharmaceuticals, human food processing, and animal feed production (Sanchez et al., 2018; Zhang et al., 2021). Supplementing animal diets with additional amino acids or free amino acids can increase feed conversion efficiency and sustainability by providing alternative nitrogen sources (Selle et al., 2020). This strategy, however, is not the only way to supply nitrogen in ruminant nutrition; ruminants are unique in that they possess a rumen that uniquely enables the efficient conversion of non-protein nitrogen (NPN) into microbial crude protein (MCP) through the ammonia nitrogen (NH₃-N) metabolic pathway. This mechanism provides a theoretical foundation for developing novel protein alternatives (Adejoro et al., 2020; Dewhurst and Newbold, 2022). Numerous studies have demonstrated that common NPN sources, such as urea and ammonium salts, are widely added to ruminant diets. However, the excessively rapid release of ammonia may lead to ammonia toxicity, and the timing of ammonia release often precedes energy substrate fermentation, resulting in an imbalance between energy and nitrogen utilization (Saro et al., 2019; Melo et al., 2022; Zheng et al., 2024). In contrast, amino acids offer notable advantages as novel NPN carriers. For instance, amino acids such as glutamine exhibit an ammonia release rate 4-fold lower than that of urea (Duan et al., 2017). Following their addition to the diet, they can reduce ruminal ammonia concentration by 37%, thereby effectively mitigating the risk of toxicity (Zhu et al., 2022). In diets based on low-quality roughage, crystalline lysine directly contributes to MCP synthesis, increasing microbial protein production by 15.6%. Moreover, its nitrogen utilization efficiency exceeds that of the urea group (Dias et al., 2023).

Amino acid solution, a byproduct of the acid hydrolysis process, is rich in free amino acids and ammonium chloride and has considerable potential for utilization as a feed resource (Karlsen et al., 2023). However, the direct inclusion of free amino acids in ruminant diets is inefficient. This is because most free amino acids are rapidly degraded by rumen microorganisms, leading to amino acid inactivation, NH₃-N production, increased hepatic burden, potential disruption of microbial composition balance, and reduced digestibility (Satter and Slyter, 1974). Additionally, the absorption rate of amino acids in the small intestine is significantly lower in free form than in peptide-bound forms. Consequently, excessive supplementation of amino acids is required to meet nutritional demands, which increases economic costs (Silk et al., 1985). To bypass these shortcomings, upgrading the feed matrix via bioprocessing prior to ingestion provides a viable alternative.

Solid-state fermentation technology, based on microbial metabolism, enables substrate transformation under low water activity conditions (Mizoguchi, 2023). This process facilitates the degradation of anti-nutritional factors, promotes the proliferation of probiotics, and enriches functional metabolites, thereby enhancing feed nutrient utilization and biosafety (Tanasković et al., 2021; De Villa et al., 2023). Yao et al. (2018) applied solid-state fermentation to improve yellow wine lees, the main byproduct in yellow wine brewing. Following fermentation, the contents of crude protein (CP), polypeptides, and amino acids increased by 14.5%, 40.9%, and 26.1%, respectively. He et al. (2024) utilized Aspergillus niger and Candida tropicalis to ferment glutamic acid waste liquid and observed a significant increase in the contents of essential amino acids and small peptides. Metabolomics analysis further confirmed a substantial enhancement in the availability of nitrogen sources for rumen microorganisms.

Amino acid waste liquid is rich in nitrogenous compounds, but its direct use in animal feeding is limited by its physicochemical characteristics and potential palatability issues (Zhu et al., 2022; He et al., 2024). Rice bran is an abundant agricultural byproduct that can serve as a carrier and fermentable substrate (Tahir et al., 2025; Artdita et al., 2026). Through microbial co-fermentation, these two resources may be converted into a more nutritionally usable feed ingredient with potential value as a partial replacement for conventional protein sources.

Previous studies have reported that fermented feeds can improve nutrient availability, modify rumen fermentation, and influence gastrointestinal microbial communities (Bana and Zsédely, 2025; Zhang et al., 2025). However, limited information is available regarding the use of co-fermented amino acid waste liquid and rice bran as a partial SBM replacer in finishing cattle diets. In particular, its effects on rumen fermentation and rumen fungal community composition remain insufficiently characterized. Rumen fungi play important roles in fiber degradation and rumen ecosystem function, yet they are less frequently investigated than bacterial communities in studies of fermented feed alternatives.

Therefore, the objective of this study was to evaluate the effects of graded replacement of SBM with a co-fermented amino acid waste liquid–rice bran feed on growth performance, rumen fermentation characteristics, and rumen fungal community composition in finishing cattle. We hypothesized that partial replacement of SBM with this co-fermented feed would maintain growth performance while modulating rumen fermentation and fungal community structure.

2 Materials and methods

2.1 Source of materials and solid-state fermentation

The amino acid solution was purchased from Wuhan Duxinglong Agricultural Science and Technology Co., Ltd. Rice bran was obtained from Wuhan Liangxin Xin Planting Professional Cooperative Company. The fermented feed ingredient used in this study was prepared by solid-state fermentation. Briefly, the fermentation substrate was formulated by mixing dried amino acid waste liquid with rice bran at 1:1 (wt/wt) ratio.The fermentation strains and process were developed based on previous optimization work conducted by our research group (Zhou et al., 2026). The selected microbial strains included Enterococcus faecium, Bacillus licheniformis, Candida utilis, Bacillus subtilis, and Lactobacillus plantarum. These strains were mixed at the optimized ratio to prepare a compound microbial inoculant, with a viable microbial content of approximately 10^6 CFU/g.The fermentation conditions were optimized using response surface methodology. Briefly, the amino acid waste liquid and rice bran mixture was adjusted to a moisture content of 37.5%, inoculated with 12.5 g of compound microbial inoculant per 100 kg of dry matter, and fermented at 34 °C for 15 days. The 15-day fermentation period was selected based on preliminary process optimization, during which the fermentation substrate reached a relatively stable state suitable for feeding. After fermentation, the product was used as the co-fermented feed for partial replacement of SBM in the experimental diets. and its nutritional composition is presented in Table 1.

Table 1. Nutritional composition of the co-fermented feed ingredient (%).

1. ND, Not detected.

2.1 Study site, animals, experimental design, and chemical analysis

The feeding trial was conducted at Hubei Jingshan Jingniu Agricultural and Animal Husbandry Cooperative Union from January 2024 to March 2024. A total of 40 healthy Simmental finishing bulls with similar initial body weight(499.2 ± 26.1kg, mean ± SD)were used in this study. A completely randomized design was used, and the bulls were randomly assigned to one of four dietary treatment groups (n = 10 per group) with each animal serving as an experimental unit. The experimental animals were subjected to a 10-day pre-experimental period followed by a 60-day experimental period. The diet was prepared in accordance with the “Beef Cattle Feeding Standards” (NY/T 815-2004). The treatments were defined by the proportion of SBM replaced by the fermented feed mixture: (1) control group (CON, 0% substitution); (2) 10% SBM substitution group (T10); (3) 20% SBM substitution group (T20); and (4) 30% SBM substitution group (T30). All feeds were prepared as total mixed ration (TMR). In addition, the bulls in all experimental groups were fed twice a day, at 08:00 and 18:00, and received water ad libitum. The details of the feed formulas and the nutritional contents of the diets are shown in Table 2. The nutritional components of the feed samples, including dry matter (DM), crude protein (CP), neutral detergent fiber (NDF), acid detergent fiber (ADF), calcium (Ca), and phosphorus (P), were determined according to the following Chinese national standard methods: GB/T 6435-2014 for DM, GB/T 6432-2018 for CP, GB/T 20806-2022 for NDF, NY/T 1459-2022 for ADF, GB/T 6436-2018 for Ca, and GB/T 6437-2018 for P (Standardization Administration of China, 2014; Standardization Administration of China, 2018a; Standardization Administration of China, 2018b; Standardization Administration of China, 2018c; Standardization Administration of China, 2022).

Table 2. Feed ingredients and nutritional composition (%) (air-dried basis).

1. The premix provided the following minerals and vitamins per kilogram of diet: VA 200000IU, VD3 25000IU, VE 4 000IU, Fe 3500mg, Mn 2000 mg, Zn 1500mg, Cu 550mg, I 30mg, Se 15 mg, Co 15mg, Ca 150g, P 60g.

2. NEm and NEg were calculated with reference to the Table of Feed Composition and Nutritional Value in China (33rd edition, 2022), while the other parameters were expressed as measured values.

2.2 Determination of growth performance

On the morning of the first day and the 60th day of the regular feeding period, the fasting weight of the experimental cattle was determined (after fasting for 8 h), and the initial body weight (IBW) and final body weight (FBW) were recorded, while the feed intake was also recorded. The average daily gain (ADG, kg/d) was determined using the formula: ADG = (FBW - IBW)/Trial Duration (days). The dry matter intake (DMI, kg/d) was calculated daily as follows: DMI = [Feed Offered (as-fed basis) - Feed Refusal (as-fed basis)] × (Dry Matter Content of the Diet/100). The feed-to-gain ratio (F/G) was calculated using the formula: F/G = Total Dry Matter Intake (kg)/Total Body Weight Gain (kg), where Total Body Weight Gain = Final Body Weight - Initial Body Weight.

2.3 Determination of rumen fermentation parameters

On the 30th day of the experiment, five beef cattle with body weight close to the group average and in good health were selected from each group for rumen fluid collection through the oral catheter method (Da Cunha et al., 2023). The catheter was inserted until it reached the rumen. The first 20 mL of turbid liquid obtained was discarded to minimize salivary contamination. Subsequently, approximately 150 mL of the ruminal content was collected in three separate aliquots with a 5-s interval between each aspiration to ensure sample homogeneity. The collected samples were pooled and filtered through four layers of sterile gauze. The filtrate was then accurately divided and added to 50 mL centrifuge tubes labeled with the corresponding experimental identifiers. A portion of the sample was immediately used for pH measurement using a portable pH meter. The remaining two portions were transferred into cryotubes, placed in precooled dry ice transport boxes, and subsequently stored in a laboratory ultra-low temperature freezer at -80 °C within 2 h of collection for the subsequent analysis of NH₃-N, volatile fatty acids (VFAs), and microbial composition. NH₃-N concentration was determined using the colorimetric method of Broderick and Kang (1980), while VFA concentrations were analyzed using an Agilent 6890N gas chromatograph (Guo et al., 2024).

2.4 Analysis of rumen microbial community structure

The preserved rumen fluid samples were analyzed by the Microbiological Testing Center of Shanghai Paisenuo Biotechnology Co., Ltd. Total DNA was extracted from the samples by using the Omega Soil DNA Kit (Omega Bio-Tek, USA) and subsequently stored following successful quality assessment by agarose gel electrophoresis and spectrometry with the NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). The ITS1 region of fungal ribosomal DNA were amplified using the ABI GeneAmp^® 9700 PCR system with primers ITS1F/ITS2R. The resulting PCR products were recovered by gel electrophoresis, purified, and quantified using the Quant-iT™ PicoGreen™ Kit. A sequencing library was then constructed in accordance with the protocol of the Illumina TruSeq Nano DNA LT Library Prep Kit. Raw sequencing data were processed using QIIME2 (through the DADA2 pipeline) to generate amplicon sequence variants (ASVs), which were taxonomically classified based on the UNITE 8.2 database (for fungi) (Kim et al., 2022). The raw sequence reads have been deposited in the Sequence Read Archive database of NCBI under accession number SUB15523124.

2.5 Statistical analysis

Data were organized in Microsoft Excel and analyzed using SPSS 26.0 (IBM Corp., USA). One-way ANOVA was performed, and means were compared by Fisher’s least significant difference (LSD) test. Because the dietary treatments formed a graded replacement series of SBM by co-fermented feed, orthogonal polynomial contrasts were used to test for linear and quadratic effects. Results are presented as means ± standard error (SE).

To control the false discovery rate (FDR) due to multiple comparisons across the different response variables (growth performance, rumen fermentation parameters, alpha-diversity indices, and differentially abundant fungal taxa identified by LEfSe), the Benjamini-Hochberg procedure was applied. For growth performance, rumen fermentation, and alpha-diversity indices, FDR correction was performed within each family of related parameters (e.g., all VFA parameters, all diversity indices) with a threshold of P<0.05q<0.05. For LEfSe output, all original P-values from the Kruskal-Wallis and Wilcoxon tests were adjusted using the same FDR method, and taxa with P<0.10q<0.10 were considered as potentially enriched biomarkers, balancing the exploratory nature of the study and the high biological variability of rumen fungal communities (Mallick et al., 2021).

3 Results

3.1 Effect of SBM replacement by a co-fermented feed on performance

As shown in Table 3, the substitution of SBM with the solid-state co-fermented feed product of amino acid waste liquid and rice bran had no effect on the ADG, DMI, and F/G ratio of finishing cattle (P > 0.05).

Table 3. Effect of SBM replacement by a co-fermented feed of amino acid waste liquid and rice bran on the growth performance of fattening cattle.

3.2 Effect of SBM replacement by a co-fermented feed on rumen fermentation parameters

The effects of replacing SBM with the co-fermented feed on rumen fermentation parameters are shown in Table 4. The overall treatment effect was significant for propionate(P < 0.05), valerate(P < 0.05), and the acetate-to-propionate ratio (A/P ratio) (P < 0.05). Compared with CON, propionate concentration increased in the replacement groups, with the highest values observed in T20 and T30. Valerate concentration was also increased by SBM replacement, particularly in T20 and T30. In contrast, the A/P ratio decreased significantly in T20 and T30 compared with CON, indicating a shift in rumen fermentation pattern toward relatively greater propionate production. No significant overall treatment effects were observed for ruminal pH, isobutyrate, butyrate, isovalerate, total VFAs, or NH₃-N(P > 0.05).

Table 4. Effect of SBM replacement by a co-fermented feed of amino acid waste liquid and rice bran on the rumen fermentation parameters of finishing cattle.

1. ^a,b,c: P ≤ 0.05 (different) and P ≥ 0.05 to ≤ 0.1 (trend) are shown by different letters (a, b, c) on the same line.

2. For variables with non-significant overall treatment effects but significant polynomial contrasts, the results indicate dose-related trends across graded replacement levels rather than significant pairwise differences among individual treatments.

3.3 Effect of SBM replacement by a co-fermented feed on rumen fungi

3.3.1 α-Diversity analysis

With an increase in sequencing depth, the dilution curves of the four rumen fungal groups almost stabilized (Figure 1A). Correspondingly, the coverage index for all samples exceeded 99% (Figure 1B). As shown in Figure 1B, SBM substitution with co-fermented amino acid waste liquid-rice bran feed in the diet did not impact the alpha diversity of colon microbiota, as indicated by the Chao1, Simpson, and Shannon (P > 0.05). In total, 1145 ASVs were identified across the CON, T10, T20, and T30 groups, with 294, 375, 367, and 302 ASVs detected in the respective groups. Among these, 19 ASVs were shared across all groups, accounting for 6.46%, 5.07%, 5.18%, and 6.29% of the total ASVs in the CON, T10, T20, and T30 groups, respectively (Figure 1C).

3.3.2 β-Diversity analysis and classification composition distribution

Principal coordinate analysis (PCoA) based on Bray-Curtis distance revealed that the rumen fungal sample points from the CON and SBM substitution groups (T10, T20, and T30) were closely clustered in the low-dimensional space (Figure 2A). Ascomycota was the dominant fungal phylum across all groups (69–77% relative abundance), while Neocallimastigomycota was more abundant in the CON group than in the T10, T20, and T30 groups (P < 0.05) (Figure 2B). At the genus level, Pichia was the most abundant among the top 10 genera (32–49%). The abundance of Candida was higher in the T20 and T30 groups than in the CON and T10 groups (P < 0.01), whereas Orpinomyces abundance was lower in all treatment groups than in the CON group (P < 0.05) (Figure 2C).

3.3.3 Linear discriminant analysis effect size differences in fungal flora

The results revealed notable differences in the structure of rumen fungal communities under different SBM substitution proportions, with each group exhibiting distinct enriched taxonomic units (Figure 3). Specifically, the CON group showed enrichment of Neocallimastigomycetes, Neocallimastigales, Neocallimastigaceae, Neocallimastix, Sebacinales, Sebacinaceae, Mucor, and Mucoraceae, which are predominantly anaerobic fungi. In the T10 group, Plectosphaerellaceae, Glomerellales, Phaeosphaeriaceae, Trimorphomycetaceae, and Saitozyma were enriched, primarily representing environmentally adaptive fungal species. The T20 group was enriched with six unclassified fungal taxa. The T30 group exhibited enrichment of Chytridiomycota, Pezizomycetes, Saccharomycetales, Spizellomycetales, Pezizales, Trichomaceae, Ascobolaceae, Candida, Myceliophthora, Ascobolus and Xeromyces, which are largely associated with aerobic fermentation processes.

4 Discussion

The lack of effect of dietary solid-state fermented mixture on ADG, DMI, and F/G demonstrates that substituting SBM with this co-fermented feed, across the inclusion levels tested (10–30%), did not compromise the overall growth performance of finishing cattle. These results demonstrate that this substitution had no adverse effects on the growth performance of finishing cattle. In this context, our findings indicate that the fermented mixture can safely replace up to 30% of dietary SBM in finishing cattle diet without adversely affecting growth performance. This presents a viable strategy for diversifying protein sources and mitigating economic constraints in beef production.

In ruminants, the functioning of the rumen fermentation process is stabilized by the equilibrium state of rumen microecology, and pH VFA and NH₃-N levels are the key parameters to evaluate rumen environmental homeostasis (Shen et al., 2023; Waters et al., 2025). The normal pH range of the rumen fluid is 5.5–7.5. The dynamic pH balance plays a core regulatory role in the stability of the rumen microbial ecosystem (Mao and Wang, 2025). In the present study, The lack of significant difference in rumen fluid pH between the co-fermented feed groups and the control group indicates that substituting soybean meal with this fermented mixture did not disrupt the ruminal fermentation environment. This finding is consistent with the result reported, who found that fermented bagasse did not affect the rumen pH of Simmental cattle (Jin et al., 2023).

As a key rumen fermentation parameter, NH3-N is the primary nitrogen source for microbial protein synthesis and is used to evaluate the dynamic balance between decomposition of rumen nitrogen-containing metabolites and microbial protein synthesis (Song et al., 2023). In the present study, no significant effect on ruminal NH₃-N concentration was observed with the inclusion of the co-fermented feed, suggesting that the rate and extent of dietary protein degradation by rumen microbes were not altered, similar to the results reported by Gunun et al. (2023). This might be because free amino acids and NPN in the amino acid waste liquid are rapidly degraded by microorganisms to yield NH₃-N, and the probiotics in the fermented feed enhance the anabolic metabolism of ammonia, thereby promoting microbial protein synthesis. Thus, SBM substitution with the co-fermented feed does not negatively impact rumen fermentation in finishing cattle and simultaneously stabilizes the rumen fermentation environment.

Volatile fatty acids (VFAs), as the primary energy source for ruminants, can provide up to 75% of the total metabolic energy (Park et al., 2024). In the present study, replacement of SBM (SBM) with the co-fermented amino acid waste liquid–rice bran feed altered the ruminal VFA profile, as shown by increased propionate and valerate concentrations and a reduced acetate-to-propionate ratio (A/P ratio) in the T20 and T30 groups. These findings are generally consistent with previous reports showing that fermented feed ingredients can modulate rumen fermentation patterns (Wang et al., 2021). Propionate is the major glucogenic precursor in ruminants and plays an important role in hepatic gluconeogenesis. Compared with acetate- or butyrate-dominated fermentation, propionate formation is generally considered to be more favorable for energy conservation in the host animal (Bergman, 1990). The decrease in the A/P ratio is physiologically relevant because it indicates a shift toward a more propionate-oriented fermentation pattern.

The polynomial contrast analysis provided further evidence for a dose-response relationship between SBM replacement level and rumen fermentation characteristics. Acetate, propionate, butyrate, valerate, total VFAs, and NH₃-N showed linear increases with increasing SBM replacement level, whereas the A/P ratio decreased linearly. Therefore, the graded substitution of SBM with the co-fermented feed appeared to induce dose-related modifications in ruminal fermentation rather than only isolated differences among individual treatments.

The increase in valerate concentration may be associated with the nitrogenous characteristics of the co-fermented feed. Compared with conventional SBM, the co-fermented amino acid waste liquid–rice bran feed may contain more soluble nitrogenous compounds, such as small peptides, free amino acids, and other fermentation-derived metabolites. Valerate can be produced during ruminal degradation and fermentation of dietary proteins, peptides, and amino acids (Allison and Bryant, 1963). Therefore, the elevated valerate concentration observed in the replacement groups may reflect altered microbial metabolism of nitrogen-containing substrates in the rumen. In addition, the nutrient-rich nature of the co-fermented feed may have influenced the growth and metabolic activity of rumen microorganisms involved in protein and amino acid fermentation.

Notably, valerate showed both significant linear and quadratic responses, increasing from CON to T20 and then showing no further increase at T30. Similar response patterns were observed for several other fermentation-related variables, particularly propionate and the A/P ratio, for which T20 and T30 showed broadly comparable values. These results suggest that the ruminal fermentation response may have approached a plateau at approximately 20% SBM replacement. Combined with the maintained growth performance and the numerically lowest F/G observed in T20, these findings suggest that 20% SBM replacement may be a promising level under the present experimental conditions.

Overall, the present results indicate that replacing SBM with the co-fermented amino acid waste liquid–rice bran feed shifted ruminal fermentation toward a more propionate-oriented pattern and altered protein-related fermentation products, particularly valerate.

Although rumen fungi have a low abundance, they play a key role in the initial colonization and physical destruction of feed pellets (Flad et al., 2020). These fungi are efficient cellulose decomposers, inhibit bacterial growth through competitive exclusion, and are displaced by fibrolytic bacteria in later colonization stages (Cox et al., 2021; Wang et al., 2025). The analysis of microbial α-diversity indicated that replacing SBM with a co-fermented amino acid waste liquid-rice bran feed had no significant effect on the Chao1, Simpson, or Shannon indices. These findings are consistent with those of Gu et al. (2021). Furthermore, PCoA based on Bray-Curtis dissimilarity indicated that the substitution did not significantly alter the overall structure of the rumen fungal microbiota in finishing cattle. This implies that the solid-state fermented mixture of amino acid waste liquid and rice bran served as a nutritionally comparable substitute to SBM, maintaining a core rumen fungal ecosystem without causing a drastic ecological disturbance. The stability at the community level is consistent with the absence of significant negative impacts on animal growth performance, indicating a successful nutritional substitution.

At the phylum level, Ascomycota is the dominant phylum, mainly involved in the decomposition of lignin, as reported by Wang et al (Wang et al., 2023). The significant reduction in the relative abundance of Neocallimastigomycota in all substitution groups (T10, T20, and T30) compared to the CON group is a critical finding. Members of Neocallimastigomycota are strictly anaerobic fungi known for their potent fibrolytic activity (Saye et al., 2021). Their decreased abundance could be directly linked to dietary change. The fermented mixture, potentially presenting a more readily fermentable substrate profile with different fiber characteristics compared to the standard diet containing SBM, may have reduced the selective advantage for these specialized fiber-degraders. A shift in dietary components can influence the eukaryotic composition of rumen microbiota, including a reduction in certain fungal groups (Dai et al., 2025). This alteration in the primary degraders likely contributed to the observed changes in the rumen fermentation end-product profile, such as the increase in propionate levels.

The genus-level composition and LEfSe results together suggest that replacement of SBM with the co-fermented amino acid waste liquid–rice bran feed was associated with restructuring of the rumen fungal community. Across all treatments, Pichia remained highly abundant, suggesting that it may represent a stable commensal yeast in the bovine rumen and potentially contribute to microbial ecosystem stability (Ji et al., 2022). However, its specific functional contribution to rumen fermentation requires further validation.

A key feature of the fungal response was the enrichment of Candida in the T20 and T30 groups, which was also supported by LEfSe analysis. This enrichment may be related to the altered nutrient profile of the co-fermented feed, which may provide more soluble nitrogenous compounds, free amino acids, small peptides, readily fermentable substrates, and fermentation-derived metabolites than conventional SBM. These substrates could favor facultative fermentative yeasts such as Candida. Previous studies have reported that Candida species can participate in carbohydrate metabolism, enzyme secretion, and microbial interactions (Garcia-Mazcorro et al., 2020; Qu et al., 2023). Therefore, the increased abundance of Candida may be associated with the observed increases in propionate and valerate and the reduced A/P ratio.

In contrast, all three replacement groups (at 10%, 20%, and 30% substitution levels) consistently showed reduced abundance of fibrolytic anaerobic fungi, including members of Neocallimastigomycota and the genus Orpinomyces, while LEfSe analysis indicated enrichment of fibrolytic taxa such as Neocallimastix in the CON group. Members of Neocallimastigomycota and Orpinomyces are important obligate anaerobic fungi involved in plant fiber degradation in the rumen (Dhakal et al., 2025; Lankiewicz et al., 2025). Their reduction may be related to changes in dietary fiber availability and substrate composition, particularly the changes in NDF and ADF with increasing replacement level. Meanwhile, the increased availability of soluble nutrients and fermentation-derived metabolites may have favored metabolically versatile yeasts over strictly fibrolytic fungi.

Nevertheless, these taxonomic changes should be interpreted cautiously. The reduction in fibrolytic fungi does not necessarily indicate reduced fiber digestion, because apparent digestibility, fibrolytic enzyme activity, and functional gene expression were not measured, and other microbial groups may compensate for changes in fungal abundance. Similarly, Candida enrichment alone does not demonstrate a direct role in VFA production or host energy metabolism. The enrichment of several unclassified fungal taxa in the T20 group suggests that this replacement level may have supported a distinct fungal consortium, but their functional roles remain unclear. Overall, the combined genus-level and LEfSe results indicate an association between fungal community restructuring and the altered ruminal fermentation profile, but further functional studies are needed to determine whether these fungal shifts directly mediate the fermentation responses to the co-fermented feed.

Considering the relatively limited sample size, the LEfSe analysis employed an FDR threshold of 0.1 to reduce the risk of false negatives. While this approach preserves statistical power to detect biologically relevant microbial shifts, it inherently carries a slightly higher false discovery rate compared to the traditional 0.05 threshold. Thus, the identified microbial biomarkers should be interpreted as strong exploratory candidates warranting further validation in larger cohorts.

5 Conclusions

Partial replacement of SBM with the co-fermented amino acid waste liquid–rice bran feed maintained growth performance of finishing cattle and imporved rumen fermentation and fungal community composition. The co-fermented feed increased propionate and valerate concentrations and reduced the acetate-to-propionate ratio. Ruminal fermentation parameters exhibited a dose-dependent linear increase to increasing SBM replacement.

We recommend 20% SBM replacement with the co-fermented amino acid liquid waste-rice bran feed in fattening beef cattle diets.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Doctor Mziwenkosi Nhlanhla Mthiyane, North-West University, South Africa

Reviewed by: Héctor A. Lee-Rangel, Universidad Autónoma de San Luis Potosí, Mexico

Pichad Khejornsart, Kasetsart University, Thailand