All animal experimental procedures were conducted in accordance with the guidelines of the Laboratory Animal Ethics Committee of the Research Institute of Fisheries of the Heilongjiang River (No. 20240909-001), and the guidelines of EU Directive 2010/63/EU for animal experiments were followed throughout the study.

Cultured Amur grayling (CAG; n = 15; body length : 19.63 cm ± 0.92 cm; body weight: 112.98 g ± 24.89 g) were obtained from the Bohai Cold Water Fish Experimental Station of the Heilongjiang Institute of Fisheries Research, China. Wild Amur grayling (WAG; n = 15; body length: 18.95 cm ± 2.06 cm; body weight: 108.64 g ± 15.75 g) were captured in the Huma River, part of the upper Amur River, in October 2024. CAG came from the same cohort fed with commercial pellets, and the feed nutrient contents are presented in Supplementary Table S1. Fish were anesthetized with 100 mg/L tricaine methanesulfonate (MS-222, Sigma, Michigan, USA), after which the stomach, intestine, liver, and intestinal contents were immediately collected and stored in RNase-free tubes for subsequent analysis of feeding habits, biochemical measurements, quantitative real-time PCR (qRT-PCR), and gut microbiome analysis. All samples were immediately frozen in liquid nitrogen for 2 h and then stored at − 80°C.

The feeding habits of the Amur grayling were assessed by dissecting the fish and removing its stomach. The stomach was then immediately opened, and its contents were extracted and placed on a Petri dish containing 5 mL of sterile physiological saline solution as a diluent. Stomach contents that could be observed directly were separated by type. The frequency of occurrence for each food type was recorded, and its volume was measured. Observation of the stomach contents was performed using a digital microscope (S6D, Leica, Weztlar, Germany) at a magnification of 40 × 10 to determine the types of food consumed by the fish. Contents that could not be identified based on external characteristics due to gastric digestion were verified by DNA extraction, amplification, and sequencing. Ultimately, the contents were identified by aligning the sequencing data against reference sequences in the GenBank database (Release 262.0, 15 August 2024).

The activities of acid phosphatase (ACP, JL-T1094), alkaline phosphatase (AKP, JL-T0946), lysozyme (LYZ, JL-T1062), superoxide dismutase (SOD, JL-T0779), catalase (CAT, JL-T0900), glutathione (GSH, JL-T0906), as well as the content of malondialdehyde (MDA, JL-T0761) in the gut and liver, were measured using commercial test kits (Jianglai Co. Ltd., Shanghai, China) following the manufacturer’s instructions. Liver and gut samples were homogenized with cold extraction solution at a ratio of 1:9 (w:v, 1 g tissue per 9 mL ice-cold extraction buffer). Following centrifugation at 12,000 rpm for 10 min at 4 °C, the supernatant was collected (19). Optical density (OD) values were measured using the absorbance microplate reader (SpectraMax Plus 384, Molecular Devices, California, USA) at 405 nm (AKP), 405 nm (ACP), 530 nm (LZM), 560 nm (SOD), 510 nm (CAT), 412 nm (GSH), and 532/600 nm (MDA), respectively. Finally, the fresh weight of samples was used to calculate enzymatic activity according to the formula.

Primers were designed using the Primer Premier 6.0 software based on immune- and inflammatory-related gene sequences published in GenBank (Table 1). Primer amplification efficiency and linearity were rigorously assessed via the slope (m) and coefficient of determination (R²) of the standard curve; only primer sets exhibiting a slope between − 3.6 and − 3.1 (corresponding to 90%–110% efficiency) and an R² > 0.98 were advanced to subsequent analyses. Specificity was verified by a single, distinct melt-curve peak and the exclusive presence of a single, target-sized amplicon on agarose gel electrophoresis.

Total RNA was extracted from the liver and gut samples using the TRIzol reagent (Thermo Fisher Scientific, Wilmington, DE, USA) according to the manufacturer’s protocol. RNA quality (purity and concentration) was assessed using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific). RNA degradation and contamination were evaluated by 1% agarose gel electrophoresis. First-strand complementary DNA (cDNA) was synthesized using the Prime Script™ RT reagent kit with DNA Eraser (TaKaRa, Kyoto, Japan). The messenger RNA (mRNA) expression levels of il-lβ, il-6, il-10, Tnf-α, tlr1, tlr3, myd88, NF-κB, foxo1, igM, lyz, c3, mpo, and saa1 were detected by qRT-PCR. The reaction mixture contained 0.4 μL of each forward and reverse primer, 5 μL of 2 × TB Green Premix Ex Taq II (Tli RNaseH Plus), 3 μL of sterile distilled H₂O (dH₂O), 0.2 μL of 50 × ROX Reference Dye II (Basel, Roche, Switzerland), and 1 μL of template cDNA. The cycling parameters were established at 95°C for 180 s, followed by 40 cycles of 95°C for 5 s, 60°C for 15 s, and 72°C for 30 s. qRT-PCR was performed using an ABI 7500 real-time PCR instrument (Thermo Fisher Scientific). The level of gene expression was determined using the 2^−ΔΔCT method. β-actin (F: 5′-GGACTTTGAGCAGGAGATGG-3′; R: 5′-ATGATGGAGTTGTAGGTGGTCT-3′) was used as an internal reference to measure the relative expression of the target genes. All samples were amplified in triplicate (20).

Three gut content samples from CAG (n = 3) and gut content samples from WAG (n = 3) were collected from each experimental group. Subsequently, DNA was extracted using the E.Z.N.A. Soil DNA Kit (Omega Bio-tek Inc., Norcross, USA). The purity and integrity of nucleic acids were verified using a NanoDrop 2000 spectrophotometer (Thermo Scientific Inc., USA). The 16S rDNA was then amplified using primers 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′) on a thermocycler PCR system (GeneAmp 9700, ABI, California USA) (21). MiSeq library construction and sequencing were performed using the Illumina MiSeq PE300 platform (Illumina, San Diego, California, CA, USA) with a read length of 2 × 300 bp. Raw data underwent quality control and chimera filtering, and operational taxonomic units (OTUs) were obtained using the Uparse algorithm in Vsearch (v2.7.1) clustering. Representative OTU sequences were classified using a native Bayesian model, and species composition was analyzed at the phylum, family, and genus levels. A Venn diagram was used to identify unique and shared OTUs between the CAG and WAG groups. Alpha diversity analyses were performed using QIME software (v1.8.0) to calculate Chao1, Shannon, and Simpson indices. Based on OTUs and species abundance tables, the Bray–Curtis algorithm was used to evaluate the beta diversity of all samples, including unweighted Unifrac distance-based principal coordinate analysis (PCoA) and partial least squares discrimination analysis (PLS-DA). Functional prediction of OTUs and KEGG pathway analysis were performed using Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2) (v2.2.0). Biomarker features in each group were identified using linear discriminant analysis effect size (LefSe) software (v1.0.0). All bioinformatics data were analyzed using the Majorbio cloud platform (https://www.majorbio.com).

The biochemical data were calculated using Microsoft Excel and analyzed using the statistical software package IBM SPSS 22.0 (IBM Corporation, Armonk, NY, USA). All data are presented as the mean ± standard deviation (SD). Prior to statistical analysis, normality and homogeneity of variance were assessed using the Kolmogorov–Smirnov and Levene’s test, respectively. Independent t-tests were then utilized to determine the significance of differences, with p < 0.05 indicating a significant difference between the two groups. p-values from Spearman correlations were adjusted using the Benjamini–Hochberg false-discovery rate (FDR) method, and only correlations with FDR < 0.05 were considered significant.

The results showed that wild Amur grayling was dominated by zoobenthos (97.206%), identified based on morphological characteristics and genetic analysis. Among them, the most dominant diet was Perlidae (37.692%), followed by Ephemera (28.200%), Nothopsyche (14.002%), and Eubasilissa (5.652%). Therefore, the Amur grayling is a carnivorous fish that primarily feeds on zoobenthos. We hypothesize that this diet reflects a match between the fish’s trophic spectrum and locally available prey, thereby minimizing foraging costs. In addition, the abundance of zoobenthos precisely aligns with the realized nutritional niche of Amur grayling.

Changes in ACP, AKP, and LYZ levels in liver and gut tissue are summarized in Figure 1. The activities of ACP in the liver and gut showed no significant differences (p > 0.05) between the WAG and CAG groups. AKP activity in the gut was significantly higher in the WAG group than in the CAG group (p < 0.05), while AKP activity in the liver showed no significant difference between the two groups (p > 0.05). Similarly, LYZ activity in the liver and gut tissues of Amur grayling was significantly higher in the WAG group than in the CAG group (p < 0.05), demonstrating that the wild environment can enhance immunity in Amur grayling.

As shown in Figure 2, CAT activity in the liver was 9% higher in the captive group than in the wild group (p < 0.01). The gut antioxidant parameters of wild and cultured Amur grayling showed highly significant differences (p < 0.01). Specifically, CAT activity and GSH content were 64% and 66% higher in the wild group than in the cultured group, respectively, while MDA levels were lower in the wild group.

In the present study, the mRNA expression levels of immune/inflammatory-related genes in the liver and intestines were measured by qRT-PCR. In the liver, the mRNA expression of C3, IgM, il-6, il-10, il-lβ, lyz, mpo, myd88, NF-κB, ssa1, tlr1, tlr3, and Tnf-α was upregulated in the WAG group, while foxo1 was significantly downregulated (p < 0.05) (Figure 3). In intestinal tissue, the mRNA levels of C3, foxo1, il-6, il-10, il-lβ, tlr1, tlr3, Tnf-α, lyz, mpo, and NF-кB increased significantly in the WAG group, whereas the expression of IgM, myd88, and ssa1 decreased significantly (p < 0.05) (Figure 4).

Using high-throughput 16S rRNA gene sequencing, we analyzed the farm group (CAG; n = 3) and WAG (n = 3), yielding a total of 404,602 raw reads. The Venn diagram analysis identified 992 OTUs, with 304 OTUs unique to the CAG group and 530 OTUs unique to the WAG group (Figure 5a). Good’s coverage confirmed that the sequencing captured up to 99% of the gut microbiota in both captive and wild Amur grayling (Figure 5b). The Chao1 index indicated that wild Amur grayling had higher species richness than captive individuals (Figure 5c). Additionally, Shannon and Simpson indices showed that the gut microbiota of wild Amur grayling was more diverse, with higher Shannon and lower Simpson values compared with captive Amur grayling (Figures 5d, e). These results indicate that fish living in a natural environment exhibit higher species richness and greater species diversity. However, analyses using principal component analysis (PCA), PCoA, and nonmetric multidimensional scaling (NMDS) revealed no statistically significant differences between the gut microbiota of farmed and wild fish (Figure 6).

To assess gut microbiota differences between wild and aquaculture Amur grayling, samples from Amur Lake were compared with those from an aquaculture facility. At the phylum level, the gut microbiota of farmed Amur grayling was dominated by Pseudomonadota, followed by Fusobacteriota, Bacillota, and Actinomycetota, which together accounted for 97.3% of the community. In contrast, the wild cohort was dominated by Bacillota, followed by Pseudomonadota, Cyanobacteriota, and Actinomycetota, representing 88.7% of the community (Figure 7a). Bacillota was significantly more abundant in wild fish than in the farmed group, whereas Pseudomonadota was significantly less abundant in the wild group (p < 0.05) (Figure 7b). As shown in Figure 7c, at the family level, the farmed group was dominated by Comamonadaceae (6.10%), followed by Xanthobacteraceae (5.88%). In the wild group, Mycoplasmataceae was the most abundant, accounting for 4.2%. Compared with the farmed group, the wild group showed a significant increase in the relative abundance of Carnobacteriaceae and Mycobacteriaceae, whereas the relative abundances of Devosiaceae, Beijerinckiaceae, Xanthobacteraceae, Sphingomonadaceae, Comamonadaceae, Moraxellaceae, Burkholderiaceae, Lysobacteraceae, Pleomorphomonadaceae, Caulobacteraceae, and Rhizobiaceae were significantly decreased (p < 0.05) (Figure 7d). The most abundant genus in the gut microbiota of wild Amur grayling was Candidatus_Bacilloplasma (16.6%), followed by norank_o_Chloroplast (11.4%), Roseateles (8.2%), and Prevotella (5.9%) (Figure 7e). In captive Amur grayling, Roseateles (19.8%) was the dominant genus, followed by Bradyrhizobium (13.3%), Sphingomonas (13.1%), and Pseudomonas (11.1%). Compared with the captive group, the wild group showed a significant increase in the relative abundance of Culicoidibacter, whereas the relative abundances of Methylobacterium, Delftia, Aquabacterium, Bradyrhizobium, Sphingomonas, Brevundimonas, and Stenotrophomonas were significantly decreased (p < 0.05) (Figure 7f).

The changes in the presumptive functions of the intestinal microbiota of Amur grayling were examined by predicting the metagenomes using PICRUSt (Figure 8). Four functional pathways were more highly abundant in wild Amur grayling, including glycan biosynthesis and metabolism, immune disease, immune system, and translation. In aquaculture Amur grayling, 23 pathways were more highly abundant compared with wild fish, including those related to amino acid metabolism, biosynthesis of other secondary metabolites, carbohydrate metabolism, cell growth and death, and energy metabolism. Overall, these results clearly demonstrate that immune suppression in farmed Amur grayling corresponds to decreased abundances of Bacillota and other potentially beneficial microbes, highlighting the interconnectedness of diet, microbiota, and host immunity.

Spearman correlation coefficient was used to test the relationship between the abundance of intestinal microbiota at the phylum level and immune- and inflammatory-related indices (Figure 9). Significant correlations were observed between intestinal and liver immune-related indicators (i.e., liver ACP, liver AKP, liver LYZ, liver il-10, liver Tnf-α, liver tlr3, intestinal il-10, intestinal Tnf-α, intestinal lyz, intestinal mpo) and the gut microbiota (i.e., Pseudomonadota, Bacillota, Chloroflexota, Acidobacteriota, Thermodesulfobacteriota, Planctomycetota, Verrucomicrobiota, Fibrobacterota) in the WAG group, as shown in the heatmap. Pseudomonadota, Chloroflexota, Thermodesulfobacteriota, Planctomycetota, Verrucomicrobiota, and Fibrobacterota were positively correlated with LYZ and Tnf-α in the liver and with il-10, Tnf-α, lyz, and mpo in the intestine (R > 0.5, p < 0.01), but negatively correlated with ACP, AKP, il-10, and tlr3 in the liver (R ≤ 0.5, p < 0.01). Bacillota was strongly positively correlated with liver ACP, liver AKP, liver il-10, and liver tlr3 (R > 0.5, p < 0.01), and negatively correlated with liver lyz, liver Tnf-α, intestinal il-10, intestinal Tnf-α, intestinal lyz, and intestinal mpo (R ≤ 0.5, p < 0.01). Spearman correlation was performed between immune parameters and phylum-level intestinal microorganisms in the CAG group. The relative abundance of Actinomycetota, Bacteroidota, Chloroflexota, Acidobacteriota, Bdellovibrionota, and Nitrospirota was significantly positively correlated with liver ACP and intestinal myd88, NF-κB, and C3 mRNA levels (R > 0.5, p < 0.01). However, their relative abundance was significantly negatively correlated with intestinal ACP, il-10, Tnf-α, and IgM, as well as with liver lyz, mpo, and ssa1 (R ≤ 0.5, p < 0.01), whereas Fusobacteriota showed the opposite correlation. The abundance of Patescibacteria, Myxococcota, unclassified_k_norank_d_Bacteria, and Entotheonellaeota was positively correlated with immune indices (AKP, liver C3, intestinal lyz) and inversely correlated with genes (liver myd88, intestinal il-6, tlr1, tlr3, foxo1, mpo, ssa1) that enhance immunity.

The fish immune system is crucial for host defense and is considered an important marker for disease prevention and for maintaining fish health (24). In our trial, a series of immune-related indicators in the liver and gut tissues of wild and farmed Amur grayling were measured. Among them, significantly higher AKP and LYZ activities were found in the liver and gut tissues of wild Amur grayling. AKP and LYZ are important components of innate immunity and play a vital role in mediating host protection against bacterial invasion (25, 26). This study indicated the superior innate immune performance of wild Amur grayling.

The intestinal barrier is composed of biological, chemical, mechanical, and immune barriers, which can protect the host from pathogen invasion (27). Chemical barriers include a series of chemical substances such as complement proteins, lysozyme, and intestinal antimicrobial peptides in the outer mucus layer of intestinal epithelial cells (28, 29). In this study, the expression levels of genes that encode C3, lyz, and antibacterial enzymes such as tlr1 and tlr3 were significantly elevated in wild Amur grayling, implying superior intestinal chemical barrier function in wild Amur grayling. Inflammation is a protective biological response of the body to external injury and is mediated by several cytokines (30). In fish, the immune barrier in the intestine largely depends on their immune response (31). Lu et al. (32) have shown that a successful host immune response is generally the result of the actions of both pro- and anti-inflammatory cytokines, which clear pathogens and reduce tissue damage. The cytokines Tnf-α, il-lβ, and il-6 are considered proinflammatory cytokines (33). However, the anti-inflammatory cytokine il-10 is an immunoregulatory molecule that regulates the activity of proinflammatory cytokines (34). Currently, few studies have reported on the immunological barrier in wild and cultured Amur grayling. The present study revealed the upregulated mRNA expression trends of pro-inflammatory genes (Tnf-α, il-lβ, and il-6) and anti-inflammatory genes (il-10) in the liver and intestine of wild Amur grayling. TLRs can activate NF-κB by mediating myd88 to enhance the formation of proinflammatory factors (35). The wild individuals had higher relative mRNA expression levels of myd88 and NF-κB in the liver than the cultured group. The results of this study indicated that Amur grayling exhibit enhanced antimicrobial effects and immune function in the wild. In addition, IgM produced by B lymphocytes participates in specific immune responses and is an important indicator for measuring immune response capability (36). In this study, significantly higher mRNA expression of IgM in the liver was observed in wild Amur grayling, whereas the expression of IgM decreased significantly in the gut, implying that the farm and wild environments activate different immune-related genes to mediate various immune responses in a tissue-specific manner.

In normal metabolism, the dynamic balance of redox maintains a stable state. When fish are subjected to internal and external stress, reactive oxygen species are produced, and the balance of the antioxidant system in the body is impaired, ultimately leading to a reduction in intestinal health, immunity, and growth performance (37, 38). SOD, CAT, GSH, and MDA are important indicators of antioxidant function. SOD, CAT, and GSH can scavenge free radicals to reduce damage to the intestinal mucosa caused by oxidative stress, playing important roles in intestinal defense and repair (39, 40), while MDA content represents the degree of lipid peroxidation and indirectly reflects the extent of cellular damage (41). In the gut, the activities of antioxidant enzymes in wild Amur grayling were significantly higher, while the MDA levels in the liver and gut were lower compared with those in cultured individuals, indicating that the antioxidant capacity of wild Amur grayling is stronger than that of cultured ones.

Intestinal microbiota constitute a complex and diverse ecosystem comprising many microorganisms (42). The gut microbiome is crucial for the health and survival of fish (43). In this study, the dominant phyla in wild and captive groups were Bacillota and Pseudomonadota, respectively. This result was similar to a previous comparative study of the gut microbiota in wild and farmed Malaysian Mahseer (Tor tambroides) by Tan et al. (2019). Similarly, Pseudomonadota is also regarded as the dominant phylum in many captive fish species, such as rainbow trout (Oncorhynchus mykiss) and Atlantic salmon (Salmo salar) (44, 45). Moreover, it has been demonstrated that an increased abundance of Pseudomonadota is a potential diagnostic signature of metabolic and immune dysbiosis (46). Zhao et al. (47) reported that the abundance of Pseudomonadota increased after immune suppression. Therefore, we hypothesize that Pseudomonadota could inhibit the immune responses and decrease the expression of hepatic and intestinal immune-related genes in farmed Amur grayling.

Studies have shown that Bacillota could provide nutrients and energy substances to epithelial and gastrointestinal cells, which can improve mucus production, act as an anticancer and anti-inflammatory agent, and maintain intestinal barrier integrity (48). In cultured Amur grayling fish, the significantly increased abundance of Pseudomonadota and decreased abundance of Bacillota caused immune suppression, as indicated by the significantly downregulated relative expression of lyz, NF-κB, and il-10. Similar findings have been reported in cultured Japanese seabass (Lateolabrax japonicus) exposed to high dietary soybean meal levels (49). Dysbiosis of the intestinal microbiota could lead to disorders of the intestinal immune system and disease occurrence in farmed Amur grayling. In addition, Bacillota showed a greater abundance in the wild group than in the farm group. It was speculated that the gut health of wild Amur grayling was superior to that of captive Amur grayling. Ammonia is the inorganic nitrogen source for various Cyanobacteriota taxa (50). Compared with the farm group, Cyanobacteriota increased in the wild group, which may have been caused by the reproduction of Cyanobacteriota in the water and subsequent ingestion by the Amur grayling. Similar results have been reported for oysters (Crassostrea gigas) (51). Therefore, it is hypothesized that the ammonia content in wild habitats might be higher than that in aquaculture water, which could promote an increased stress response to ammonia exposure in wild Amur grayling.

The use of large amounts of organic and inorganic fertilizers, fecal waste generation, and low-quality feed conversion induces pollution in captive water (52). Delftia can degrade and transform organic and inorganic pollutants in the aquatic environment (53, 54). Thus, Delftia showed a higher relative abundance in the captive group. Studies have shown that frequent exposure of cultured fish to various pollutants leads to a decline in intestinal health and immunity (55). We hypothesized that this results in reduced immune function in cultured Amur grayling. High-carbohydrate and high-lipid diets have been widely used in aquaculture practices, but they can also cause excessive lipid accumulation in fish liver (56). In contrast, wild Amur grayling mainly consume small fish and zoobenthos with a high protein content. As observed in many animals, diet is a key factor influencing the gut microbiota (57). Thus, the gut microbiota in captive Amur grayling showed higher carbohydrate metabolism levels. In addition, increases in “carbohydrate metabolism” and “lipid metabolism” were observed. The pronounced disparity in dietary composition between farmed and wild Amur grayling profoundly shapes their respective gut microbial communities. We therefore encourage optimal wilderness training before artificial stock enhancement to adjust the gut microbial community of fish species and improve their adaptability to wild populations in reintroduction projects.

Interactions between the host and intestinal microbiota form the basis for the development of immunity (58). In our study, Actinomycetota, Bacteroidota, and Bdellovibrionota were positively correlated with the immune parameter (liver ACP) and intestinal immune-related genes (myd88, NF-κB, and C3) in captive Amur grayling. According to Jakubiec-Krzesniak et al. (59), Actinomycetota can secrete secondary metabolites that could inhibit pathogenic bacteria. Bdellovibrionota can disrupt the colonization of the Shigella genus, thereby enhancing the survival of zebrafish (Danio rerio) (60). Bacteroidota can improve the decomposition of nutrients and digestibility of the host, and their abundance is closely related to maintaining the intestinal environmental balance in animals (61). Therefore, the decreased abundances of Actinomycetota, Bdellovibrionota, and Bacteroidota in cultivated Amur grayling allowed undigested nutrients and harmful substances to enter into liver through the hepatic portal vein, thereby inducing a liver immune response. In addition, Actinomycetota, Bdellovibrionota, and Bacteroidota were considered important factors contributing to the increased survival rate of Amur grayling in the wild environment after artificial enhancement and release. Fusobacteriota have protective effects on fish and can produce butyrate and vitamin B12, which are involved in the immune-regulatory processes in the body (62). In our study, Fusobacteria were positively associated with intestinal immune response cytokine IgM, and both the relative abundance of Fusobacteria and the expression level of IgM increased in the captive group. It has been demonstrated that Fusobacteria are a gut microbiota biomarker closely related to the immune response in captive Amur grayling.

The gut microbiota regulates immunity via the gut axis. Planctomycetota influence complex chemical metabolism. In addition, Verrucomicrobiota benefit fish gut health by enhancing immune responses, regulating microbial communities, and promoting mucosal repair (63). Fibrobacterota are small but important phyla, mainly reported as cellulose-degrading bacteria (64). In addition, Thermodesulfobacteriota, Planctomycetota, Verrucomicrobiota, and Fibrobacterota were positively associated with inflammatory factors (hepatic LYZ and Tnf-α, intestinal il-10, Tnf-α, lyz, and mpo) but negatively correlated with liver immune- and inflammatory-related parameters (ACP, AKP, il-10, and tlr3) in the intestine tissue of wild Amur grayling. Bacillota showed the opposite correlation. The results indicated that Amur grayling immune responses can be linked to the abundance levels of Bacillota, Planctomycetota, Verrucomicrobiota, and Fibrobacterota. Based on the above studies, we postulate that intestinal microbes in Amur grayling may modulate host immunity by enhancing the expression of immune-related genes (il-10, Tnf-α, lyz, and mpo). Further research is required to understand how host–microbiome interactions coevolve and adapt, as well as the specific roles of certain microorganisms in nutrient processing and immune regulation. Studying the gut microbiomes of wild fish populations is important for understanding their natural microbial communities and ecological functions, in addition to research focused on aquaculture or laboratory fish.

Firstly, we acknowledge that our experimental groups comprised n = 15 fish per treatment, which limits the statistical power to detect subtle microbiota shifts. Secondly, all farmed samples were collected from the Bohai Cold Water Fish Experimental Station of the Heilongjiang Institute of Fisheries Research; thus, the findings may not generalize to Amur grayling in other geographic regions or rearing systems. Thirdly, all wild samples were captured in the Huma River, from the upper Amur River, in autumn. The diet of Amur grayling varies with the season; thus, the results of comparisons of gut microbiota between wild and farmed Amur grayling may differ across seasons.

Firstly, we outline plans for repeated sampling from the same individuals across an entire production cycle to capture temporal trajectories of the gut microbiota. Secondly, we propose controlled feeding experiments with candidate probiotic strains (e.g., Bacillota isolated from wild conspecifics) to test whether targeted microbial interventions can restore immune-relevant taxa in farmed fish.