Bacteria and bacteriophage consortia modulate cecal SCFA production and host metabolism to enhance feed efficiency in ducks

Bacteria and bacteriophage consortia modulate cecal SCFA production and host metabolism to enhance feed efficiency in ducks | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Bacteria and bacteriophage consortia modulate cecal SCFA production and host metabolism to enhance feed efficiency in ducks Xia Wang, Shujie Tian, Yunsheng Zhang, Li Yang, Di Hu, Zezhong Wang, and 10 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7357899/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 04 Mar, 2026 Read the published version in Microbiome → Version 1 posted 4 You are reading this latest preprint version Abstract Background: The gut microbiota critically influences poultry health, nutrition, feed efficiency (FE), and overall productivity. However, the relationship between gut microbes, including bacteria and bacteriophages, and FE in ducks remains underexplored. To address this, we integrated cecal 16S amplicon, metagenome, microbiota derived short-chain fatty acids (SCFAs) profiling, liver transcriptome, and serum metabolome data to illustrate the contribution of gut microbiome (bacteria and viruses) to duck FE. Results: We reconstructed viral genomes and prokaryotic metagenome-assembled genomes (MAGs), annotated their genes using comprehensive databases. Prokaryotic hosts of viruses were also predicted to understand virus-host dynamics within the gut ecosystem. Our results revealed that high-FE ducks have higher concentration of propionate and butyrate in cecum compared with low-FE ducks. The metagenome sequencing revealed distinct cecal microbiota profiles between two groups, with increased relative abundance of representative SCFA producers, especially Paraprevotella sp905215575 and Bacteroides sp944322345 , and enhanced SCFA-biosynthesis pathways in high-FE ducks. Virome genome assembly identified two bacteriophages encoding auxiliary metabolic genes (AMGs) involved in pyruvate metabolism, enhancing nutrient availability for host bacteria to produce SCFAs (e.g., temperate phage-encoded pyruvate phosphate dikinase) or exploiting host central metabolic pathways for viral replication (e.g., lytic phage-encoded formate C-acetyltransferase). Furthermore, these representative SCFA-producing bacteria and bacteriophage consortia were associated with serum metabolites (including L-histidine and 4-hydroxydecanedioylcarnitine) linked to duck FE. Conclusion: Collectively, these findings provide novel insights into the gut microbial factors regulating FE in ducks, offering potential strategies to optimize poultry nutrition and productivity. Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction Efficient feed utilization is critical for sustainable and profitable poultry production. Feed costs account for 60–70% of total production expenses in commercial poultry operations, making feed efficiency (FE) a primary determinant of economic viability [ 1 ]. In the context of a growing global demand for animal protein, enhancing FE in poultry species like chickens and ducks holds significant economic and environmental importance through lower feed waste and nitrogen excretion. In non-ruminant animals like poultry and pigs, short-chain fatty acids (SCFAs), primarily produced through microbial fermentation of fiber in the cecum or colon, can supply 60–70% of the energy requirements of intestinal epithelial cells [ 2 ]. This process not only sustains normal intestinal epithelial function but also plays a vital role in overall gut health and metabolic regulation. Beyond energy provision, SCFAs (such as acetate, propionate, and butyrate) function as signaling molecules by activating G protein-coupled receptors (GPCRs), including FFAR2 (GPR43), FFAR3 (GPR41), and HCAR2 (GPR109A). Activation of these receptors modulates energy expenditure, preadipocyte differentiation, and appetite control [ 3 , 4 ]. Given these functions, modulating the gut microbiota to enhance SCFA production represents a promising strategy for improving feed utilization. For examples, probiotic supplementation accelerates gut microbiota maturation, increasing average daily gain (ADG) by 6.2% while reducing FCR by 4.5% compared to antibiotic-free controls [ 5 ]. Similarly, microbiota-accessible fibers boost populations of SCFA-producing bacteria (e.g., Bacteroidetes ) by up to 40%, enhancing energy harvest and intestinal barrier function in laying hens [ 6 ]. While bacterial contributions to FE through SCFA production are well-documented [ 7 – 9 ], the specific roles of bacteriophages in modulating these bacterial functions and their combined impact on duck feed efficiency remain uncharacterized, preventing the development of comprehensive microbiome-based strategies to optimize feed utilization. The gut virome, dominated by bacteriophages (phages infecting bacteria), represents one of the densest viral ecosystems in nature, with estimates of more than 10 8 to 10 10 virus-like particles per gram of intestinal content [ 10 , 11 ]. These phages exert profound influence over microbial community dynamics and metabolic processes through predator-prey interactions, horizontal gene transfer, and the regulation of bacterial lysogenic-to-lytic life cycles. For instance, phages can reshape bacterial diversity by selectively lysing dominant taxa like Bacteroides or Bacillota , thereby altering niche competition and nutrient flow [ 12 ]. Phage-encoded auxiliary metabolic genes (AMGs) further modulate bacterial metabolism, such as enhancing carbohydrate utilization in Bacteroides [ 13 ] and Oscillospiraceae [ 14 ]. Such phage activities can cascade to reshape microbial functional outputs, such as bile acid metabolism and SCFA production, with indirect implications for host physiology. Despite their ubiquity and functional significance, the role of gut viruses in shaping host phenotypes remains understudied outside human disease contexts (e.g., inflammatory bowel disease, where phage dysbiosis alters bacterial consortia and immunity [ 15 ]). In livestock, emerging evidence suggests agricultural relevance: rumen phages in cattle may regulate host nutrient utilization via AMGs and virus-host linkages, potentially influencing FE [ 12 , 16 ]. These findings position the virome as a hidden effector bridging microbial ecology and host biology. However, data on viral-bacterial interactions and their association with FE in ducks are limited, and specific microbial taxa or functional genes linked to duck FE remain unidentified. Elucidating these interactions is essential for developing microbiome-based strategies to enhance FE in ducks. This study characterized gut microbial communities, including bacteria and bacteriophages, in ducks stratified by FE. We hypothesized that high-FE ducks harbor distinct gut microbiota compositions and functional capabilities compared with low-FE ducks. We further aimed to resolve virus-host dynamics and assess their impact on bacterial community structure and function. By integrating 16S amplicon, metagenome, SCFAs, liver transcriptome, and serum metabolome data, we provide a comprehensive assessment of the duck gut microbiome’s association with FE. Our findings may inform microbiome-targeted strategies (e.g., probiotics, dietary interventions) to optimize feed utilization in duck production. Results 1. FE traits and SCFAs in CF and ZF ducks The statistics of average daily feed intake (ADFI), average daily gain (ADG), feed conversion ratio (FCR), and fatness traits between two duck populations are shown in Fig. 1 A&B and Table S1 . Compared to ZF ducks, CF ducks exhibited 5.45% higher ADFI ( p < 0.0001), 4.59% lower ADG ( p < 0.001), 9.45% higher FCR ( p < 0.0001), 55.83% greater SFW ( p < 0.001), and 26.64% greater AFW ( p < 0.001). These results demonstrate significantly superior FE and reduced adiposity in the ZF population relative to CF ducks. SCFAs derived from gut microbiota have been proven to be associated with animal FE previously [ 8 , 9 ]. To compare cecal SCFAs in two groups, we analyzed 10 SCFAs from a randomly selected subset of both groups (20 CF, 16 ZF), which also showed significant difference in FCR and ADG (Fig. S1 A&B), using targeted metabolomics. We found that ZF ducks exhibited higher propionate and butyrate concentrations in cecal content compared with CF ducks ( p < 0.05), while the others have no significant difference (Fig. 1 C&D and Fig. S1 C-J). These results combined suggest that higher propionate and butyrate may improve FE in ducks. 2. Divergence of cecal bacterial diversity, composition, and function in High- and Low-FE Ducks Because SCFAs in duck is primarily derived from microbial fermentation of fiber in the cecum, we first investigated cecal bacterial composition and interactions in CF and ZF groups using 16S rRNA gene amplicon sequencing, which revealed a conserved phylum-level structure across both groups, dominated by Bacteroidetes followed by Bacillota_A and Bacillota_C (Fig. 2 A and Fig. S2A-F). While observed ASV richness was comparable between groups, ZF ducks exhibited significantly elevated α-diversity versus CF ducks as measured by Chao1, Pielou, and Shannon indices (Fig. 2 B). The co-occurrence networks of two groups have low adjust rand index (ARI) (0.136) [ 50 ], indicating that the network structural and topological characteristics varied for each group. Detailed information of network topology is shown in Fig.S2G&H and Tables S2-3. To further identify the specific intestinal microbial species that contribute to FE, we analyzed cecal microbiome from the subset of both groups (20 CF, 16 ZF) using metagenomics. Prokaryotic genome reconstruction from metagenomic data yielded 1,820 MAGs through single-sample assembly (mean contigs per MAG: 482). Quality filtering (completeness ≥ 50%, contamination < 10%) retained 943 medium-to-high-quality MAGs. Taxonomic classification assigned these MAGs to 16 phyla, with 890 MAGs (94.4%) resolving to 121 genera and 620 MAGs to 144 known species. Notably, 327 MAGs (34.7%) lacked matches to reference genomes, representing potentially novel species. We performed the analysis of composition of microbiomes with bias correction (ANCOM-BC) and identified 21 species significant different in terms of abundance between two groups ( p < 0.05) (Fig. 2 C). Specifically, gut bacterial communities in ZF ducks were mainly enriched in these species, such as Phocaeicola barnesiae , Merdibacter sp900759455 , Paraprevotella stercoravium , Paraprevotella stercorigallinarum , and Avelusimicrobium sp944324735 , while CF animals were more abundant in Alistipes finegoldii , Desulfovibrio sp944324525 , gemmiger formicilis_B , and UMGS1491 sp905211185 . We then performed spearman correlations between these differentially abundant bacterial species and FCR/ADG among individuals. Multiple species, including Phocaeicola barnesiae , UMGS263 sp949289045 , Paraprevotella sp905215575 , and Bacteroides sp944322345 , were negatively correlated with FCR ( r = -0.43 ~ -0.37, p < 0.05) but positively correlated with ADG. UMGS1491 sp905211185 exhibited a positive relationship with FCR ( r = 0.40, p < 0.05) but negative correlation with ADG (Fig. 2 D and Fig.S2I). We further analyzed the correlation of these differential bacterial species with propionate or butyrate, the results revealed most of differential bacteria, such as UMGS1449 sp900551925 , Paraprevotella sp905215575 , and Mediterraneibacter quadrami , were positively correlated with propionate or butyrate ( r = 0.33 ~ 0.39, p < 0.05), while a few of them, including Desulfovibrio sp944324525 and Gemmiger formicilis_B , displayed negative association with propionate ( r = -0.46 ~ -0.34, p < 0.05) (Fig. 2 E and Fig.S2I). To further compare the functional features of microbiota in two groups, we annotated the microbial functions of each metagenome via the KEGG database using a gene-centric approach [ 51 ] and quantified the abundance of each KEGG orthologous (KOs) in different samples. Non-metric multidimensional scaling (NMDS) plot showed an obvious difference in the gut microbial KOs between the CF and ZF groups (PERMANOVA, R 2 = 0.167, p < 0.001; Fig. 2 F). We identified 545 differentially abundant KOs ( p < 0.05), with 339 (62.2%) enriched in ZF ducks. These differential KOs were significantly enriched in multiple KEGG pathways including “Carbon metabolism”, “Sulfur metabolism”, “Phosphonate and phosphinate metabolism”, “Methane metabolism”, “Citrate cycle (TCA cycle)”, “Nucleotide metabolism”, and “Pyruvate metabolism” (Fig.S2J). Because “Pyruvate metabolism” is a central hub in the anaerobic glycolysis pathway of gut bacteria, driving the synthesis of short-chain fatty acids (SCFAs) like acetate, propionate, and butyrate, ZF microbiomes showed significant enrichment of aceE (K00163; pyruvate dehydrogenase) within pyruvate oxidation modules (M00307), converting pyruvate to acetyl-CoA, the central precursor for SCFA biosynthesis (Fig. 2 G). This predicted enhanced SCFA production was consistent with our experimental measurement that ZF ducks exhibited higher propionate and butyrate concentrations in cecal content compared with CF ducks ( p < 0.05). Regarding carbohydrate metabolism and energy metabolism, it is worth mentioning that the KOs associated with “Methane metabolism” (9/12 KOs) was predominantly enriched in the CF microbiome. While the KOs involved in “Citrate cycle (TCA cycle)” (5/6 KOs), “Pentose and glucuronate interconversions” (6/6 KOs), and “Phosphonate and phosphinate metabolism” (5/6 KOs) were more abundant in the gut microbiome of ZF group (Fig. 2 H&I). These enrichment patterns align with known energy-partitioning mechanisms: ZF’s enhanced pentose/glucuronate interconversions promote SCFA yield [ 52 , 53 ], while CF’s methane metabolism redirects carbon from SCFA production [ 54 ]. 3. Gut virome characterization and host interactions Metagenomic assembly of 36 cecal samples yielded 4,334 quality-filtered viral contigs. Among these, 835 high-confidence contigs (completeness ≥ 50%) had a mean length of 41,983 bp (N50 = 50,390 bp). Clustering at 95% average nucleotide identity (ANI) identified 1,986 viral operational taxonomic units (vOTUs), including 428 with ≥ 50% completeness (Fig. 3 A). All vOTUs represented bacteriophages, predominantly within Caudoviricetes (tailed phages), Malgrandaviricetes (microviruses), Megaviricetes , and Arfiviricetes (Fig. 3 B). Protein-sharing clustering of high-quality vOTUs generated 392 viral clusters (VCs). Host prediction revealed a total of 1,984 virus-host linkages, comprising 808 lytic VC-hMAG (host metagenome assembled genome) and 1,176 temperate VC-hMAG linkages. The hosts of temperate viruses spanned across 11 phyla (Fig. 3 C), whereas those of lytic viruses were associated with 12 phyla (Fig. 3 D). As expected, Bacteroidota (comprising 58.17% of lytic and 69.56% temperate viral hosts) and Bacillota_A (comprising 15.97% of lytic and 15.73% temperate viral hosts), highly predominant gut bacterial taxa, were the most common hosts of gut viruses at phylum level. Bacteroides and Phocaeicola were the most prevalent hosts at genus level (Fig. S3A&B). Among these VC-hMAG linkages, VC2778, VC2779, and VC2563 exhibited the highest host diversity, each associated with 62, 54, and 49 distinct host species (Fig. S3C). 4. Virome divergence and metabolic modulation in high- and low-FE Ducks To compare gut viral communities between two groups, we examined viral α- and β-diversity of two groups. The result showed that two groups exhibited similar α-diversity profiles except for the Shannon index at the VC level (Fig. 4 A), while β-diversity analysis (NMDS, Bray-Curtis) revealed distinct clustering of individuals from two groups (PERMANOVA, R 2 = 0.096, p < 0.001; Fig. 4 B). To further investigate gut viral signatures associated with FE, we performed ANCOM-BC analysis and identified that 47 VCs were significantly decreased, while 32 VCs were increased in CF group compared with ZF group (Fig S3D). Then we did the correlation analysis between these differential VCs and FE-related phenotypes. 12 viruses (5 positive and 7 negative) and 10 viruses (6 positive and 4 negative) displayed significant correlations with FCR and ADG, respectively (Spearman, | r | ≥ 0.3, p < 0.05) (Fig. 4 C). For example, VC2859, VC2785, and VC0894 showed consistent positive ADG or negative FCR correlations. As phages may link to phenotype by modulating the composition and dynamics of bacterial communities, we predicted interactions of these FE-associated VCs (either co-occurrence or mutual exclusion) with differentially abundant bacteria (Fig. 4 D). Spearman correlation analysis revealed distinct phage-bacteria association patterns. One group of VCs, including VC2785, VC2396, VC2667, VC0894, and VC2518) displayed significant positive correlations (co-occurrence) with high-FE favored bacteria (e.g., Paraprevotella sp905215575 , Phocaeicola barnesiae , Bacteroides sp944322345 , Paraprevotella stercorigallinarum ) ( r = 0.34 ~ 0.75, p < 0.05), while exhibiting significant negative correlations (mutual exclusion) with low-FE favored bacteria (e.g., Alistipes finegoldii , UMGS1491 sp905211185 , Desulfovibrio sp944324525 , Gemmiger formicilis_B ) ( r = -0.70 ~ -0.33, p < 0.05). Conversely, a second group of VCs (e.g., VC4010, VC2559, VC2862, VC2755, VC2722, VC1483) displayed the opposite association pattern, positively correlating with low-FE favored bacteria and negatively correlating with high-FE favored bacteria. Then we searched AMGs from viral contigs and found 109 AMGs that were involved in a wide range of metabolic processes, including the metabolism of nucleotides, carbohydrates, amino acids, and sulfur (Fig. 5 A). The most prevalent AMG was DNMT1, a DNA (cytosine-5)-methyltransferase that prevents viruses from being recognized and cleaved by the restriction-modification systems of their hosts [ 55 , 56 ], which was detected in more than 40 high quality viral contigs. CysH was the second highly abundant AMG which encodes a phosphoadenosine phosphosulfate reductase involved in the synthesis of sulfite from sulfate. Additionally, another abundant AMG, metK encoding S-adenosylmethionine (SAM) synthetase, was important for genomic DNA methylation and cell division. Two AMGs, NAMPT and pncA, were important for maintaining duck’s NAD + level which is critical for maintaining energy homeostasis. CobS and cobT were two key AMGs involved in synthesis of vitamin B12. We subsequently conducted a differential analysis of AMGs between two groups and found 16 AMGs showing significant differential abundance. Notably, two differential AMGs, pyruvate phosphate dikinase (PPDK) and formate C-acetyltransferase (pflD), involved in pyruvate metabolism were identified in viral contigs (Fig. 5 B). This PPDK shared 83% amino acid identity with PPDK in Prevotella sp. E2-28 (GenBank: CP091788) (Fig. 5 C), while the pflD was 52% identical to a pflD in Catenibacterium mitsuokai strain DSM 15897 (GenBank: CP102271) (Fig. 5 D). Specifically, in the anaerobic cecal environment, temperate virus (VC2785)-encoded PPDK, catalyzing the reversible conversion of phosphoenolpyruvate (PEP) to pyruvate and ATP [ 57 ], was higher abundant in ZF group compared with CF group, which may augment the fermentative capacity of host bacteria (e.g., Bacteroides , Paraprevotella , and Phocaeicola ) in ZF group (Fig. 5 E). This is in line with abovementioned observation that VC2785 displayed positive correlation with both ADG and propionate (Fig. 4 D). While lytic virus (VC0717)-encoded pflD, converting pyruvate to acetyl-CoA and formate, was higher abundant in CF group compared with ZF group, which may accelerate viral replication by hijacking the host’s metabolic machinery in CF group (Fig. 5 F). Moreover, its host, Prevotella sp015074785 , belonging to the Prevotellaceae family, significantly contributes to the production of SCFAs which may induce lytic viral replication and lysis of cells [ 58 , 59 ]. These results together suggest that gut viruses could potentially optimize nutrient availability for their hosts or exploit host central metabolic pathways for their own replication, underscoring their influence on host physiology including FE. 5. Serum metabolome profiling reveals systemic metabolic divergence between two groups Serum metabolomics identified 221 differentially abundant metabolites between CF and ZF ducks (|log 2 FC| ≥ 1, p < 0.05), with 184 upregulated and 37 downregulated in CF compared with ZF (Fig. S4). According to the annotation of Human Metabolome Database (HMDB), these differential metabolites were mainly categorized into lipids and lipid-like molecules (38.01%), organic acids and derivatives (10.41%), organoheterocyclic compounds (9.95%), and organic oxygen compounds (5.43%) (Fig. 6 A). Metabolite pathway enrichment analysis indicated that these differential metabolites were mainly enriched in 13 pathways (Fig. 6 B). Among them, 3 pathways including “pyrimidine metabolism”, “butanoate metabolism”, and “pantothenate and CoA biosynthesis” showed concordant enrichment in liver transcriptomes (Fig. 6 C). For example, in butanoate metabolism, we observed higher abundance of succinic semialdehyde in CF group compared with ZF group (log 2 FC = 1.37, p < 0.05), this may stem from downregulation of the succinic semialdehyde dehydrogenase (ALDH5A1) gene (log 2 FC = -0.75, p < 0.01), a mitochondrial homotetramer protein that catalyzes the NAD+ -dependent conversion of succinic semialdehyde into succinate which can be further fermented to butanoate (Fig. 6 D&E). In functional enrichment analysis of DEGs, we also observed multiple pathways related to lipid metabolism, such as “insulin resistance”, “PPAR signaling pathway”, and “glycerophospholipid metabolism”, explaining the predominant lipid metabolite differences observed between two groups (Fig. 6 C). 6. Gut microbiome-serum metabolome associations with feed efficiency Among differential serum metabolites, 37 metabolites were significantly correlated with FE traits (| r | ≥ 0.3, p < 0.05), with most of them positively correlated with FCR (21/37) (Fig. 7 A). For example, L-histidine and 4-hydroxydecanedioylcarnitine were negatively correlated with ADG, while positively associated with FCR. To detect the potential effect of differentially abundant bacteria on host FE-associated metabolites, spearman’s rank correlations revealed a total of 188 strong correlations (| r | ≥ 0.5, p < 0.05; Fig. 7 B). Among these correlations, most of them were negative between differential bacteria and FE-associated metabolites (e.g., Paraprevotella sp905215575 vs. L-histidine: r = -0.67), while UMGS1491 sp905211185 positively correlated with multiple FE-linked metabolites (e.g., 4-hydroxydecanedioylcarnitine: r = 0.48). Spearman’s rank correlations between differential viruses and FE-associated metabolites were also performed, most FE-associated metabolites showed positive correlations with 26 VCs (e.g., VC2540 vs. 4-hydroxydecanedioylcarnitine: r = 0.64) and negative correlations with 27 VCs (e.g., VC2785 vs. L-histidine: r = -0.56), respectively, while the correlation of CDP-DG(a-25:0/20:4(6E,8Z,11Z,14Z) + = O(5)) and (2R)-O-Phospho-3-sulfolactate with VCs showed the reversed pattern (Fig. 7 C). Discussion In the past decade, gut microbiome and its associated roles have attracted increasing attention. Although several studies have investigated the bacterial contributions to FE of poultry using 16S rRNA sequencing [ 8 , 61 ], the role of gut virome remains poorly characterized. Our multi-omics integration (metagenomics, transcriptomics, metabolomics) reveals how both bacterial and viral components of the duck gut microbiome collectively influence FE-associated phenotypes. We identified that ZF ducks have higher FE traits and cecal SCFAs (e.g., propionate and butyrate) concentration compared with CF group. Due to pivotal roles of SCFAs in maintaining energy homeostasis and promoting animal health through multifaceted mechanisms, including maintenance of intestinal barrier integrity and improvement of nutrient absorption [ 62 ], thus animals producing more SCFAs usually may have high FE. Previous studies showed that butyrate supplementation improved FE in chicken and duck [ 63 , 64 ]. In a sheep study, higher concentration of propionate was also detected in rumen of high FE group [ 9 ]. In duck, SCFAs are chiefly produced in the cecum, thus the difference of SCFAs between two groups may arise from variation of cecal bacterial diversity, composition, and function. We found that cecal bacteria in high-FE (ZF) and low-FE (CF) ducks shared core phyla ( Bacteroidetes and Bacillota ) but exhibited divergent diversity and network architectures. ZF microbiomes displayed higher modularity (Fig. S2G&H), indicative of more persistent and stable community-level and within-community interactions [ 65 , 66 ]. Besides network topology, the relative abundances of bacteria at various taxonomic levels in two groups also exhibit differences, most of the differential species are enriched in ZF group. Crucially, these ZF-enriched bacteria (e.g., Phocaeicola barnesiae , Paraprevotella sp905215575 , and Bacteroides sp944322345 ) potentially drove efficient pyruvate flux toward SCFA production rather than methanogenesis (Fig. 4 E-F). This is consistent with the observation that elevated cecal propionate/butyrate in ZF ducks. It has reported that Paraprevotella was more enriched in the leaner individuals [ 67 ] which usually have high FE traits [ 8 , 68 ]. Conversely, CF-enriched Desulfovibrio sp944324525 and Gemmiger formicilis_B correlated negatively with SCFAs (Fig.S2I), aligning with inefficient energy harvesting, this was also observed in chicken gut [ 69 ]. As the “dark matter” of the gut microbiome, bacteriophages (predominantly Caudoviricetes ) can alter the bacterial community structure by lysing certain bacteria or through lysogeny, thereby affecting animal phenotypes. There was significant difference in β diversity of viruses, suggesting distinct viral community structure in two groups. In addition, multiple differentially abundant viruses showed significant correlation with FE traits in ducks. From the predicted phage-host linkages, we found that some FE-correlated phages could target SCFA producers, such as Paraprevotella sp905215575 and Bacteroides sp944322345 , which are favorable for high FE. These results position phages as key phenotype mediators, which is supported by our human gut findings [ 14 ]. During the phage infection of bacteria, AMGs carried by bacteriophages play crucial roles in optimizing the host’s metabolism and nutrient acquisition, which affect several important ecological processes, including methane metabolism in different habitats [ 70 ], polysaccharide digestion in the ruminants [ 12 ], and SCFAs metabolism in the human gut [ 14 ]. However, the contribution of viral AMG in poultry gut has not been reported yet. Our results identified a total of 109 AMGs in duck cecum, which participated multiple processes, such as antiviral defense (e.g., DNMT1), sulfur metabolism (e.g., cysH and metK), energy homeostasis (e.g., NAMPT and pncA), and vitamin biosynthesis (cobS and cobT), suggesting that the gut virome can also affect feed digestibility in duck. We identified a total of 16 differentially abundant AMGs between two groups, with two AMGs, PPDK and pflD, involved in pyruvate metabolism process. In anaerobic environments like the cecum, microbes rely on fermentation pathways. PPDK enables the regeneration of PEP, which serves as a key intermediate for substrate-level phosphorylation, a primary ATP-generating mechanism under anaerobic conditions [ 57 ]. Given the requisite integration of temperate viruses into the host genome, the temperate phages-encoded PPDK enriched in ZF potentially boosted pyruvate metabolism and SCFAs production of host bacteria, thereby conferring communal benefits and, in turn, reciprocal advantages to the viruses themselves [ 71 ]. While the lytic phage-encoded pflD enriched in CF could hijack host resources to produce more acetyl-CoA (SCFAs), which may promote viral infection and host lysis [ 72 ]. However, the underlying mechanisms remain to be explored and need further experimental validation. Because gut microbiotas produce a variety of molecules, some of which are taken up into the bloodstream and affect health. Therefore, characterization of the interactions between gut microbiota and host plasma metabolites is also important for understanding the effects of the gut microbiota on animal phenotypes. Our results revealed that L-histidine and succinic semialdehyde were more abundant in serum of CF group compared with ZF group. Consistently, recent papers have reported that plasma L-histidine was downregulated in lambs with high ADG [ 73 , 74 ]. Besides, our RNA-seq data revealed reduced hepatic ALDH5A1 expression in CF ducks, which could impair succinic semialdehyde→butanoate conversion and lead to accumulation of succinic semialdehyde (Fig. 7 D). Intriguingly, succinic semialdehyde level were found to be significantly positively correlated with insulin resistance and liver fat content [ 75 ], concordant with dysregulated lipid pathways, such as “insulin resistance”, “PPAR signaling pathway”, and “glycerophospholipid metabolism”, in CF group. We reported that these serum FE related metabolites were associated with duck cecal microorganisms, where most of them displayed strong negative associations with multiple differential bacterial and viral species between groups. The negative correlations of plasma L-histidine level with Bacteroides and Paraprevotella are in harmony with the results observed in piglets, rats, and fish [ 76 – 78 ]. This is probably due to the contribution of Bacteroides and Paraprevotella genera to L-histidine degradation [ 79 , 80 ]. Due to “virus-host” interactions, where viruses lysis host bacteria or boost their metabolism, gut viruses also display correlations with plasma metabolites. However, these links still require experimental validation in further study. Such information will provide evidence highlighting the possibility of animal blood metabolites through manipulating gut microbial functions. Conclusion Taken together, to elucidate the mechanisms by which gut microbiota affect FE of ducks, we proposed two potential convergent pathways: (1) SCFAs are the major end products from the fermentation of gut bacteria. Viruses directly target SCFA-producing bacteria, altering community function and output, thereby jointly impacting FE through collaborative interactions with bacteria; (2) Gut bacteria and viruses interact to modulate host animal metabolism via blood circulation and thus influencing FE (Fig. 8 ). However, further studies are required to validate the function of bacteriophages-encoded AMGs and causal relationships between representative SCFA-producing bacteria and bacteriophage consortia and host physiological outcomes. Materials and Methods Animals management Two Pekin duck populations with divergent FE were used: a high-FE group (ZF) and a low-FE group (CF). A total of 300 one-day-old male ducklings (150 per group) with similar initial body weights were housed individually in wire-floor cages (200 × 100 × 40 cm) equipped with nipple drinkers and tubular feeders at the National Changping Comprehensive Agricultural Engineering and Technology Research Center (Livestock and Poultry Sub-Center). Ducks were reared in two phases (days 1–14 and days 15–42) according to standardized practices. Both groups received identical commercial diets (Hope Feed Co., Ltd., Fangshan District, Beijing, China); diet composition and nutritional components matched those described in our previous study [ 17 ]. Feed Efficiency Measurement and Sample Collection Body weight (BW) was recorded at 22 and 42 days of age. Total feed intake (FI) was measured from 22 to 42 days. Average daily feed intake (ADFI), average daily gain (ADG), and feed conversion ratio (FCR) were calculated as follows: ADFI (g/day) = Total FI / 21 days (1) ADG (g/day) = (BW 42 – BW 22 ) / 21 days (2) FCR = Total FI / (BW 42 – BW 22 ) (3) Valid phenotypic records were obtained for 283 ducks. After a 12-hour fast, ducks were humanely euthanized according to standard commercial protocols. Fatness traits were recorded, including skin fat weight (SFW), skin fat percentage (SFP), abdominal fat weight (AFW), and abdominal fat percentage (AFP), calculated as: SFP (%) = (SFW / carcass weight) × 100 (4) AFP (%) = (AFW / carcass weight) × 100 (5) DNA Extraction, 16S amplicon and Metagenomic Sequencing Ducks with valid FCR records are euthanized by cervical dislocation and dissected. Cecal contents (including chyme and mucosa) were immediately collected, snap-frozen in liquid nitrogen, and stored at − 80°C until further processing. Microbial DNA was extracted from cecal samples of 283 ducks using the QIAamp DNA Stool Mini Kit (QIAGEN, Hilden, Germany). The V3-V4 hypervariable region of the bacterial 16S rRNA gene was amplified using primers 338F (5′-ACTCCTACGGGAGGCAGCA-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′). Amplified libraries were sequenced (2 × 250 bp) on an Illumina NovaSeq 6000 platform. A subset of 40 ducks (20 CF, 20 ZF) was randomly selected. Microbial DNA was extracted from cecal content using a CTAB-based method. DNA quality and quantity were assessed using a Qubit 3.0 Fluorometer (Invitrogen). Four ZF samples were excluded due to insufficient DNA yield, resulting in 36 samples (20 CF, 16 ZF) for library preparation. Libraries were constructed using the VAHTS Universal Plus DNA Library Prep Kit (Illumina, San Diego, CA, USA) and sequenced (2 × 150 bp) on an Illumina NovaSeq X platform following standard protocols. SCFA concentration measurement Cecal samples from the 36 ducks used for metagenome sequencing were also analyzed for the concentration of ten SCFAs: acetic acid, propionic acid, isobutyric acid, butyric acid, isovaleric acid, valeric acid, hexanoic acid, heptanoic acid, nonanoic acid, and decanoic acid. Briefly, samples were thawed on ice. Approximately 0.5 g of cecal content was homogenized in 1 mL of deionized water for 4 minutes at 40 Hz, followed by ultrasonic oscillation for 15 minutes on ice. The homogenate was centrifuged at 5,000 rpm for 20 minutes at 4°C. A 0.8 mL aliquot of the supernatant was mixed with 0.1 mL of 50% (v/v) sulfuric acid (H 2 SO 4 ) and 0.8 mL of extraction solution containing 25 mg/L internal standard (methyl tert-butyl ether, MTBE). The mixture was vortexed vigorously for 10 minutes and centrifuged at 10,000 rpm for 15 minutes at 4°C. The resulting supernatant was analyzed using a SHIMADZU GC2030-QP2020 NX gas chromatography-mass spectrometer (GC-MS). Absolute SCFA concentrations were determined using external calibration curves generated from standard solutions of each SCFA at varying concentrations. Determination and analysis of plasma metabolome Plasma metabolites were analyzed using ultra-high-performance liquid chromatography (UHPLC) coupled with tandem mass spectrometry (MS/MS). Chromatographic separation was performed on an ACQUITY UPLC ® HSS T3 column (100 mm × 2.1 mm, 1.8 µm average particle size, Waters Corporation) using 0.1% (v/v) formic acid aqueous solution and acetonitrile as mobile phases. To mitigate analytical bias from instrumental variation, a pooled quality control (QC) sample was created by combining equal aliquots of all sample supernatants. This QC sample was injected at regular intervals throughout the analytical sequence. Raw MS data were acquired using MassLynx ™ V4.2 software (Waters Corporation). Data processing—including peak detection, alignment, and integration—was performed in Progenesis QI (Nonlinear Dynamics). Metabolite annotation was conducted against the METLIN database ( https://metlin.scripps.edu ) within the Progenesis QI platform, using a mass accuracy threshold of ≤ 10 ppm and MS/MS spectral matching. Preparation of libraries for RNA-sequencing and data analysis Liver tissues were collected from three CF and three ZF ducks immediately post-euthanasia, snap-frozen in liquid nitrogen, and stored at -80°C. Total RNA was extracted using TRIzol ™ Reagent (Thermo Fisher Scientific, Waltham, MA) following manufacturer protocols. RNA integrity was assessed by UV spectrophotometry (NanoDrop ™ 2000; Thermo Fisher Scientific) and Electropherogram analysis (Agilent 2100 Bioanalyzer; Agilent Technologies, Santa Clara, CA). Ribosomal RNA was depleted using the Illumina TruSeq Stranded Total RNA Library Prep Kit with Ribo-Zero Gold. Resulting mRNA was fragmented and reverse-transcribed into cDNA libraries using the same kit. Libraries were sequenced (2 × 150 bp) on an Illumina HiSeq X platform. Raw reads were quality-trimmed and adapter-filtered using fastp (v0.20.0) [ 18 ]. Processed reads were mapped to the Pekin duck reference genome (ZJU1.0) using STAR (version 2.7.9a) [ 19 ] with default parameters. Gene-level counts were generated with STAR’s --quantMode GeneCounts option and analyzed for differential expression using DESeq2 (v1.26.0) [ 20 ] in R/Bioconductor, with normalization via the median-of-ratios method. 16S amplicon sequencing data processing Raw sequencing reads underwent quality assessment using FastQC. Adapters and low-quality bases were trimmed using fastp (v0.20.0). Processed reads were denoised, error-corrected, and filtered for chimeras using the DADA2 plugin (v1.18) [ 21 ] in QIIME2 (v2020.6) [ 22 ], generating amplicon sequence variants (ASVs). ASVs were filtered to retain those with relative abundance > 0.0001% (10⁻⁶) and presence in ≥ 2 samples. Taxonomy was assigned against the Genome Taxonomy Database (GTDB r207) using DADA2’s native Bayesian classifier with default parameters. Alpha diversity (Shannon, Chao1, Pielou, and Richness) metrics were calculated in QIIME2. Phyloseq (v1.34.0) [ 23 ] was used for evaluation of sequencing depth and preparation of final ASV tables for downstream statistical analysis. Metagenome-assembled genomes (MAGs) Recovery and Annotation Raw metagenomic reads were preprocessed using fastp (v0.20.0) for adapter trimming and quality filtering. Host-derived reads were removed by alignment to the duck reference genome (ZJU1.0) using Bowtie2 (v2.4.2) [ 24 ]. De novo assembly of filtered reads was performed per sample using metaSPAdes (v3.15.0) [ 25 ] with default parameters. Assembled contigs were subsequently binned into MAGs using an integrative approach with MetaBAT2 (v2.17) [ 26 ], MaxBin2 (v2.2.7) [ 27 ], and CONCOCT (v1.1.0) [ 28 ]. Consensus bins were generated and refined using DAS Tool (v1.1.3) [ 29 ]. The quality of the refined MAGs was assessed using CheckM2 (v1.0.1) [ 30 ] with default parameters. Taxonomic classification was performed against GTDB r207 using GTDB-Tk (v2.1.0) [ 31 ]. Community-level functional annotations were performed using HUMAnN 3.0 (v3.6) pipeline with the clean reads as input [ 32 ]. Species-level differential abundance between FE groups was determined using ANCOM-BC [ 33 ]. Open reading frames (ORFs) were predicted from all contigs using Prodigal (v2.6.3) [ 34 ] in metagenomic mode. Proteins were annotated using eggNOG database with eggNOG-mapper v2 [ 35 ]. A non-redundant gene catalog was constructed with CD-HIT (v4.8.1) [ 36 ] at 95% nucleotide identity and 90% coverage. Gene abundances were quantified by aligning reads to the non-redundant gene catalog using Salmon (v1.10.2) [ 37 ] in alignment-based mode. Differential abundance of KEGG Orthologs (KOs) was tested using MaAsLin2 (v1.12.0) [ 38 ] with default parameters. Viral contig identification and annotation Putative viral contigs were first filtered from assembled metagenomic contigs using geNomad (v1.8.0) [ 39 ] with a relaxed threshold of –min-score = 0.75 to increase detection sensitivity. The putative viral contigs were annotated using our previously developed ViroProfiler pipeline [ 40 ] for precise viral contig identification and annotation. Briefly, The ViroProfiler pipeline utilizes CheckV (v0.9.0) [ 41 ] to remove host region in provirus contigs, and then use VirSorter2 (v2.2.4) [ 42 ] and VIBRANT (v1.2.1) [ 43 ] to identify viral contigs. Auxiliary metabolic genes (AMGs) were annotated using VIBRANT and DRAM-v (v1.4.4) [ 44 ]. Viral hosts were predicted using iPHoP (v1.2.0) with confidence score ≥ 90 [ 45 ]. Taxonomic classification was performed using a combination of geNomad and VITAP (v1.7.1) [ 46 ]. VITAP taxonomy was preferred due to its more specific annotations, often to genus or species level with high precision, while geNomad taxonomy was used when a contig was not annotated by VITAP. Viruses were classified as temperate if they are identified as provirus by geNomad, VIBRANT, or CheckV (v0.9.0), or classified as temperate viruses by BACPHLIP (v0.9.6) [ 47 ]. Viral contigs were then dereplicated to generate a non-redundant catalog of viral sequences. Specifically, pairwise average nucleotide identity (ANI) was calculated among these contigs using the clustering algorithm within CheckV, applying thresholds of ≥ 95% ANI and ≥ 85% alignment fraction (AF) to define species-level viral operational taxonomic units (vOTUs). The longest contig within each vOTU cluster was selected as its representative sequence. To further investigate relationships among the identified vOTUs, compare them with known viral genomes, and identify potential novel viral clusters (VCs), gene-sharing network analysis was conducted using vConTACT2 (v.0.11.3) [ 48 ]. VCs identified by vConTACT2 represent groups of related viral genomes, offer insights into potential genus-level relationships, and highlight vOTUs that may constitute novel taxa. Abundance of viruses were estimated by mapping clean reads to vOTUs using Bowtie2, and then quantified using CoverM (v0.7.0) ( https://github.com/wwood/CoverM ). Statistical Analyses All statistical analyses were performed in R (v4.1.0). Data manipulation and visualization utilized the tidyverse ecosystem (v1.3.0). Student’s t-test and Wilcoxon rank sum test were employed for group comparisons. For transcriptome data, differentially expressed genes (DEGs) were identified using DESeq2 with thresholds of adjusted p -value (Benjamini-Hochberg) < 0.05 and |log 2 FC| ≥ 0.585. KEGG pathway enrichment of DEGs was performed using ClusterProfiler (v4.0) [ 49 ]. For metabolomics data, peak areas were normalized via total sum scaling, differential metabolites between two groups were identified with threshold of |log 2 FC| ≥ 1 and p < 0.05. Metabolic pathway enrichment was analyzed in MetaboAnalyst 5.0 ( https://www.metaboanalyst.ca ). Declarations Ethics approval All animal procedures involving Pekin ducks (Anas platyrhynchos domesticus) were approved by the Animal Welfare and Ethics Committee of the Institute of Animal Sciences (IAS), the Chinese Academy of Agricultural Sciences (IAS20160401, CAAS, Beijing, China). Data availability The raw data for multi-omics cohort are deposited in the National Genomics Data Center website (https://ngdc.cncb.ac.cn/). All the sequencing data are accessible with the identifier CRA026543 (amplicon), CRA028005 (metagenome), and CRA027933 (transcriptome). The metabolomics data are accessible with the identifier OMIX011117. The code used for data analysis is available at the GitHub repository https://github.com/rujinlong/duckbiome. Competing interests The authors declare that they have no competing interests. Funding This work was supported by National Natural Science Foundation of China (#32341055) awarded to YSZ, China Agriculture Research System of MOF and MARA (CARS-42-2) awarded to XW, the Fundamental Research Funds for the Central Universities (grant no. 226-2025-00030) awarded to FYL, the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation - Emmy Noether program, Project No. 273124240; and SPP2330, Project No. 464797012), and the European Research Council Starting grant (ERC StG 803077) awarded to LD. Authors' contributions All authors contributed intellectually to and agreed to this submission. XW, YSZ, and JLR designed the experiments. SJT, ZZW, XQY, SFL, JW, WZ, SQW, and YSZ conducted the experiments and collected the data. SJT, LY, and JLR analyzed and interpreted the data. SJT, PYL, DH, and YSZ prepared figures. XW interpreted the data and wrote the manuscript, while JLR, FYL, LD and SSH provided substantial feedback. JLR, XW, and PYL contributed to the conceptual design of the study. XW, YSZ, FYL, LD and SSH provided funding support. All authors read and approved the final manuscript. Acknowledgements The authors sincerely thank Prof. Guang Yao from the University of Arizona and Prof. Xujun Liang from Northwest A&F University for their constructive discussions and valuable suggestions. Additionally, we acknowledge the Technical University of Munich (TUM), the Helmholtz Zentrum München (HMGU), and Northwest A&F University (NWAFU) for providing high-performance computing resources. References Zampiga, M., et al., Application of omics technologies for a deeper insight into quali-quantitative production traits in broiler chickens: A review . J Anim Sci Biotechnol, 2018. 9: p. 61. Wrzosek, L., et al., Bacteroides thetaiotaomicron and Faecalibacterium prausnitzii influence the production of mucus glycans and the development of goblet cells in the colonic epithelium of a gnotobiotic model rodent . BMC Biol, 2013. 11: p. 61. Han, H., et al., From gut microbiota to host appetite: gut microbiota-derived metabolites as key regulators . Microbiome, 2021. 9(1): p. 162. Ikeda, T., et al., Short-chain fatty acid receptors and gut microbiota as therapeutic targets in metabolic, immune, and neurological diseases . Pharmacol Ther, 2022. 239: p. 108273. Gao, P., et al., Feed-additive probiotics accelerate yet antibiotics delay intestinal microbiota maturation in broiler chicken . Microbiome, 2017. 5(1): p. 91. Yang, J., et al., Microbiota-accessible fiber activates short-chain fatty acid and bile acid metabolism to improve intestinal mucus barrier in broiler chickens . Microbiol Spectr, 2024. 12(1): p. e0206523. Yang, C., et al., Metagenomic insights into the relationship between intestinal flora and residual feed intake of meat ducks . Poult Sci, 2024. 103(7): p. 103836. He, Z., et al., Combined effect of microbially derived cecal SCFA and host genetics on feed efficiency in broiler chickens . Microbiome, 2023. 11(1): p. 198. Zhou, G.C., et al., Multi-omics revealed the mechanism of feed efficiency in sheep by the combined action of the host and rumen microbiota . Animal Nutrition, 2024. 18: p. 367–379. Hoyles, L., et al., Characterization of virus-like particles associated with the human faecal and caecal microbiota . Res Microbiol, 2014. 165(10): p. 803–812. Lepage, P., et al., Dysbiosis in inflammatory bowel disease: a role for bacteriophages? Gut, 2008. 57(3): p. 424–425. Yan, M., et al., Interrogating the viral dark matter of the rumen ecosystem with a global virome database . Nat Commun, 2023. 14(1): p. 5254. Campbell, D.E., et al., Infection with Bacteroides Phage BV01 Alters the Host Transcriptome and Bile Acid Metabolism in a Common Human Gut Microbe . Cell Rep, 2020. 32(11): p. 108142. Thiele Orberg, E., et al., Bacteria and bacteriophage consortia are associated with protective intestinal metabolites in patients receiving stem cell transplantation . Nat Cancer, 2024. 5(1): p. 187–208. Mangalea, M.R., et al., Individuals at risk for rheumatoid arthritis harbor differential intestinal bacteriophage communities with distinct metabolic potential . Cell Host Microbe, 2021. 29(5): p. 726–739.e5. Liu, X., et al., Rumen DNA virome and its relationship with feed efficiency in dairy cows . Microbiome, 2025. 13(1): p. 14. Zhang, Y., et al., Genetic parameters for residual feed intake in a random population of Pekin duck . Asian-Australas J Anim Sci, 2017. 30(2): p. 167–170. Chen, S., et al., fastp: an ultra-fast all-in-one FASTQ preprocessor . Bioinformatics, 2018. 34(17): p. i884-i890. Dobin, A., et al., STAR: ultrafast universal RNA-seq aligner . Bioinformatics, 2013. 29(1): p. 15–21. Love, M.I., W. Huber, and S. Anders, Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2 . Genome Biol, 2014. 15(12): p. 550. Callahan, B.J., et al., DADA2: High-resolution sample inference from Illumina amplicon data . Nat Methods, 2016. 13(7): p. 581–583. Bolyen, E., et al., Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2 . Nat Biotechnol, 2019. 37(8): p. 852–857. McMurdie, P.J. and S. Holmes, phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data . PLoS One, 2013. 8(4): p. e61217. Langmead, B. and S.L. Salzberg, Fast gapped-read alignment with Bowtie 2 . Nat Methods, 2012. 9(4): p. 357–359. Nurk, S., et al., metaSPAdes: a new versatile metagenomic assembler . Genome Res, 2017. 27(5): p. 824–834. Kang, D.D., et al., MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies . PeerJ, 2019. 7: p. e7359. Wu, Y.W., B.A. Simmons, and S.W. Singer, MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets . Bioinformatics, 2016. 32(4): p. 605–607. Alneberg, J., et al., Binning metagenomic contigs by coverage and composition . Nat Methods, 2014. 11(11): p. 1144–1146. Sieber, C.M.K., et al., Recovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy . Nat Microbiol, 2018. 3(7): p. 836–843. Chklovski, A., et al., CheckM2: a rapid, scalable and accurate tool for assessing microbial genome quality using machine learning . Nat Methods, 2023. 20(8): p. 1203–1212. Chaumeil, P.A., et al., GTDB-Tk: a toolkit to classify genomes with the Genome Taxonomy Database . Bioinformatics, 2019. 36(6): p. 1925–1927. Beghini, F., et al., Integrating taxonomic, functional, and strain-level profiling of diverse microbial communities with bioBakery 3 . Elife, 2021. 10: p. e65088. Lin, H. and S.D. Peddada, Analysis of compositions of microbiomes with bias correction . Nat Commun, 2020. 11(1): p. 3514. Hyatt, D., et al., Prodigal: prokaryotic gene recognition and translation initiation site identification . BMC Bioinformatics, 2010. 11: p. 119. Cantalapiedra, C.P., et al., eggNOG-mapper v2: Functional Annotation, Orthology Assignments, and Domain Prediction at the Metagenomic Scale . Mol Biol Evol, 2021. 38(12): p. 5825–5829. Li, W. and A. Godzik, Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences . Bioinformatics, 2006. 22(13): p. 1658–9. Patro, R., et al., Salmon provides fast and bias-aware quantification of transcript expression . Nat Methods, 2017. 14(4): p. 417–419. Mallick, H., et al., Multivariable association discovery in population-scale meta-omics studies . PLoS Comput Biol, 2021. 17(11): p. e1009442. Camargo, A.P., et al., Identification of mobile genetic elements with geNomad . Nat Biotechnol, 2024. 42(8): p. 1303–1312. Ru, J., et al., ViroProfiler: a containerized bioinformatics pipeline for viral metagenomic data analysis . Gut Microbes, 2023. 15(1): p. 2192522. Nayfach, S., et al., CheckV assesses the quality and completeness of metagenome-assembled viral genomes . Nat Biotechnol, 2021. 39(5): p. 578–585. Guo, J., et al., VirSorter2: a multi-classifier, expert-guided approach to detect diverse DNA and RNA viruses . Microbiome, 2021. 9(1): p. 37. Kieft, K., Z. Zhou, and K. Anantharaman, VIBRANT: automated recovery, annotation and curation of microbial viruses, and evaluation of viral community function from genomic sequences . Microbiome, 2020. 8(1): p. 90. Shaffer, M., et al., DRAM for distilling microbial metabolism to automate the curation of microbiome function . Nucleic Acids Res, 2020. 48(16): p. 8883–8900. Roux, S., et al., iPHoP: An integrated machine learning framework to maximize host prediction for metagenome-derived viruses of archaea and bacteria . PLoS Biol, 2023. 21(4): p. e3002083. Zheng, K., et al., VITAP: a high precision tool for DNA and RNA viral classification based on meta-omic data . Nat Commun, 2025. 16(1): p. 2226. Hockenberry, A.J. and C.O. Wilke, BACPHLIP: predicting bacteriophage lifestyle from conserved protein domains . PeerJ, 2021. 9: p. e11396. Bin Jang, H., et al., Taxonomic assignment of uncultivated prokaryotic virus genomes is enabled by gene-sharing networks . Nat Biotechnol, 2019. 37(6): p. 632–639. Wu, T., et al., clusterProfiler 4.0: A universal enrichment tool for interpreting omics data . Innovation (Camb), 2021. 2(3): p. 100141. Peschel, S., et al., NetCoMi: network construction and comparison for microbiome data in R . Brief Bioinform, 2021. 22(4): p. bbaa290. Qin, J., et al., A metagenome-wide association study of gut microbiota in type 2 diabetes . Nature, 2012. 490(7418): p. 55–60. Levi Mortera, S., et al., Functional and Taxonomic Traits of the Gut Microbiota in Type 1 Diabetes Children at the Onset: A Metaproteomic Study . Int J Mol Sci, 2022. 23(24): p. 15982. Wang, S., et al., Dose-related effects of early-life intake of sn-2 palmitate, a specific positionally distributed human milk fatty acid, on the composition and metabolism of the intestinal microbiota . J Dairy Sci, 2023. 106(12): p. 8272–8286. Hao, L., et al., Novel syntrophic bacteria in full-scale anaerobic digesters revealed by genome-centric metatranscriptomics . ISME J, 2020. 14(4): p. 906–918. Hampton, H.G., B.N.J. Watson, and P.C. Fineran, The arms race between bacteria and their phage foes . Nature, 2020. 577(7790): p. 327–336. Murphy, J., et al., Bacteriophage orphan DNA methyltransferases: insights from their bacterial origin, function, and occurrence . Appl Environ Microbiol, 2013. 79(24): p. 7547–7555. Deramchia, K., et al., Contribution of pyruvate phosphate dikinase in the maintenance of the glycosomal ATP/ADP balance in the Trypanosoma brucei procyclic form . J Biol Chem, 2014. 289(25): p. 17365–17378. Feng, W.H., et al., Lytic induction therapy for Epstein-Barr virus-positive B-cell lymphomas . J Virol, 2004. 78(4): p. 1893–1902. Westphal, E.M., et al., Activation of lytic Epstein-Barr virus (EBV) infection by radiation and sodium butyrate in vitro and in vivo: a potential method for treating EBV-positive malignancies . Cancer Res, 2000. 60(20): p. 5781–5788. van Kempen, M., et al., Fast and accurate protein structure search with Foldseek . Nat Biotechnol, 2024. 42(2): p. 243–246. Wen, C., et al., Joint contributions of the gut microbiota and host genetics to feed efficiency in chickens . Microbiome, 2021. 9(1): p. 126. Liu, Y., Fatty acids, inflammation and intestinal health in pigs . J Anim Sci Biotechnol, 2015. 6(1): p. 41. Ji, Z.Q., et al., Effects of coated sodium butyrate on growth performance, intestinal morphology, and caecal short-chain fatty acids of growing Pekin ducks . Br Poult Sci, 2023. 64(4): p. 529–533. Yang, Q., et al., Butyrate in combination with forskolin alleviates necrotic enteritis, increases feed efficiency, and improves carcass composition of broilers . J Anim Sci Biotechnol, 2022. 13(1): p. 3. Liu, S., et al., Ecological stability of microbial communities in Lake Donghu regulated by keystone taxa . Ecological Indicators, 2022. 136: p. 108695. Maslov, S. and K. Sneppen, Specificity and stability in topology of protein networks . Science, 2002. 296(5569): p. 910–913. Wang, Y., et al., Effect of host genetics and gut microbiome on fat deposition traits in pigs . Front Microbiol, 2022. 13: p. 925200. Bergamaschi, M., et al., Gut microbiome composition differences among breeds impact feed efficiency in swine . Microbiome, 2020. 8(1): p. 110. Yoon, J.H., et al., Inclusion of Lacticaseibacillus paracasei NSMJ15 in broiler diets induces changes in jejunal immune cell population and cecal microbiota . Anim Biosci, 2024. 37(12): p. 2155–2166. Zhong, Z.P., et al., Viral potential to modulate microbial methane metabolism varies by habitat . Nat Commun, 2024. 15(1): p. 1857. Obeng, N., A.A. Pratama, and J.D.V. Elsas, The Significance of Mutualistic Phages for Bacterial Ecology and Evolution . Trends Microbiol, 2016. 24(6): p. 440–449. Luo, X.Q., et al., Viral community-wide auxiliary metabolic genes differ by lifestyles, habitats, and hosts . Microbiome, 2022. 10(1): p. 190. Feng, T., et al., Metabolite Profile of Sheep Serum With High or Low Average Daily Gain . Front Vet Sci, 2021. 8: p. 662536. Wang, B., et al., Impact of feeding regimens on the composition of gut microbiota and metabolite profiles of plasma and feces from Mongolian sheep . J Microbiol, 2020. 58(6): p. 472–482. Li, S., et al., Longitudinal changes of serum metabolomic profile after laparoscopic sleeve gastrectomy in obesity . Endocr Connect, 2024. 13(11): p. e240292. Han, Z., et al., Fecal microbiota transplantation accelerates restoration of florfenicol-disturbed intestinal microbiota in a fish model . Commun Biol, 2024. 7(1): p. 1006. Li, Z., et al., Traditional Tibetan Medicine Twenty-Five Wei'er Tea Pills Ameliorate Rheumatoid Arthritis Based on Chemical Crosstalk Between Gut Microbiota and the Host . Front Pharmacol, 2022. 13: p. 828920. Zheng, X., et al., Integrated Metagenomic and Metabolomics Profiling Reveals Key Gut Microbiota and Metabolites Associated with Weaning Stress in Piglets . Genes (Basel), 2024. 15(8). Huang, Z., et al., Gut microbiota and blood metabolites: unveiling their roles in hippocampal volume changes through Mendelian randomization and mediation analysis . Metab Brain Dis, 2025. 40(4): p. 178. Yan, J., et al., Fecal microbiota transplantation significantly improved respiratory failure of amyotrophic lateral sclerosis . Gut Microbes, 2024. 16(1): p. 2353396. Additional Declarations No competing interests reported. Supplementary Files suppduckBiome0812.docx Supplementary Fig.S1: FE traits and SCFAs concentrations between two groups. Supplementary Fig.S2: Bacterial composition, network analysis, and functional correlations between two groups. Supplementary Fig.S3: Host association and differential abundance of viral clusters (VCs). Supplementary Fig.S4: Volcano plot of differentially abundant metabolites in CF vs. ZF. Supplementary Table S1: Growth, feed efficiency and fatness traits in two groups. Supplementary Table S2: Comparative topological properties of the CF and ZF networks. Supplementary Table S3: Degree and eigenvector centrality of the hub nodes in CF and ZF networks. Cite Share Download PDF Status: Published Journal Publication published 04 Mar, 2026 Read the published version in Microbiome → Version 1 posted Editorial decision: Revision requested 10 Sep, 2025 Editor assigned by journal 10 Sep, 2025 Submission checks completed at journal 13 Aug, 2025 First submitted to journal 12 Aug, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7357899","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":499956919,"identity":"ba9a778d-4817-47e9-9c20-1081be0cffa5","order_by":0,"name":"Xia Wang","email":"","orcid":"","institution":"Northwest A\u0026F University","correspondingAuthor":false,"prefix":"","firstName":"Xia","middleName":"","lastName":"Wang","suffix":""},{"id":499956920,"identity":"71991618-c93d-4f21-b94c-822bec1223b3","order_by":1,"name":"Shujie Tian","email":"","orcid":"","institution":"Northwest A\u0026F University","correspondingAuthor":false,"prefix":"","firstName":"Shujie","middleName":"","lastName":"Tian","suffix":""},{"id":499956921,"identity":"b04e83c1-a82d-40c6-b96b-cf3d09c0b84e","order_by":2,"name":"Yunsheng Zhang","email":"","orcid":"","institution":"Chinese Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Yunsheng","middleName":"","lastName":"Zhang","suffix":""},{"id":499956922,"identity":"6f08c516-0cde-47e4-9ecf-6e181431edf1","order_by":3,"name":"Li Yang","email":"","orcid":"","institution":"Northwest A\u0026F University","correspondingAuthor":false,"prefix":"","firstName":"Li","middleName":"","lastName":"Yang","suffix":""},{"id":499956923,"identity":"271b6179-0618-4799-b082-e6f9e52b51df","order_by":4,"name":"Di Hu","email":"","orcid":"","institution":"Northwest A\u0026F University","correspondingAuthor":false,"prefix":"","firstName":"Di","middleName":"","lastName":"Hu","suffix":""},{"id":499956924,"identity":"f107d962-a819-4f58-8d6d-5bbac8398cd9","order_by":5,"name":"Zezhong Wang","email":"","orcid":"","institution":"Northwest A\u0026F University","correspondingAuthor":false,"prefix":"","firstName":"Zezhong","middleName":"","lastName":"Wang","suffix":""},{"id":499956925,"identity":"e1db22c1-bcfa-41f7-b6e1-448b2e42e479","order_by":6,"name":"Xueqin Yang","email":"","orcid":"","institution":"Northwest A\u0026F University","correspondingAuthor":false,"prefix":"","firstName":"Xueqin","middleName":"","lastName":"Yang","suffix":""},{"id":499956926,"identity":"f8b42032-f518-476f-a796-202659aa3e97","order_by":7,"name":"Shaofei Li","email":"","orcid":"","institution":"Northwest A\u0026F University","correspondingAuthor":false,"prefix":"","firstName":"Shaofei","middleName":"","lastName":"Li","suffix":""},{"id":499956927,"identity":"23511896-8ef4-4ee5-9183-437bf43ac470","order_by":8,"name":"Jie Wei","email":"","orcid":"","institution":"Chinese Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Jie","middleName":"","lastName":"Wei","suffix":""},{"id":499956928,"identity":"38235c26-959c-4217-9d40-3f951a2806a7","order_by":9,"name":"Wei Zhou","email":"","orcid":"","institution":"Chinese Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Zhou","suffix":""},{"id":499956929,"identity":"02d8dc77-2e3e-4b49-92f5-360522c895d0","order_by":10,"name":"Shuaiqin Wang","email":"","orcid":"","institution":"Chinese Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Shuaiqin","middleName":"","lastName":"Wang","suffix":""},{"id":499956932,"identity":"cdafae21-5ce6-4d1d-bb1a-a09220aec49b","order_by":11,"name":"Li Deng","email":"","orcid":"","institution":"Technical University of Munich","correspondingAuthor":false,"prefix":"","firstName":"Li","middleName":"","lastName":"Deng","suffix":""},{"id":499956933,"identity":"a6dd46b7-a2d1-41a4-ab5d-89dafb2997a6","order_by":12,"name":"Fuyong Li","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"prefix":"","firstName":"Fuyong","middleName":"","lastName":"Li","suffix":""},{"id":499956934,"identity":"03c3da07-c669-488d-892c-99a690ae94b9","order_by":13,"name":"Shuisheng Hou","email":"","orcid":"","institution":"Chinese Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Shuisheng","middleName":"","lastName":"Hou","suffix":""},{"id":499956935,"identity":"62e30025-1c9d-4db3-b8b5-d4663b3e7765","order_by":14,"name":"Pengying Li","email":"","orcid":"","institution":"Northwest A\u0026F University","correspondingAuthor":false,"prefix":"","firstName":"Pengying","middleName":"","lastName":"Li","suffix":""},{"id":499956936,"identity":"3367f767-90c2-42a3-8235-a19d41ee0907","order_by":15,"name":"Jinlong Ru","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwUlEQVRIiWNgGAWjYBACxmYGxgMQJjOYZmwgQgsDVAtbYgNRWkAAqoXHkDgtzO28Bw583FHLYD4j5/tj3hwG2X7CDuNLODjzzHEGmRu5G5t5tzEYzyRkDWMzj8Fh3rZjDBISEC2JGw4QryXnIVjLfiK11IC0MEJsIewXHoODM9sO8EjwPDOcOXebhPEMQrYY9p8xfPCxrU5Ogj35wYe322xk+xsIaYEoOMwD5UsQchYDgzyEqiOschSMglEwCkYuAAD/JUDsG8fFYgAAAABJRU5ErkJggg==","orcid":"","institution":"Technical University of Munich","correspondingAuthor":true,"prefix":"","firstName":"Jinlong","middleName":"","lastName":"Ru","suffix":""}],"badges":[],"createdAt":"2025-08-12 16:38:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7357899/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7357899/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s40168-026-02368-y","type":"published","date":"2026-03-04T15:59:53+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":90152846,"identity":"411b9c74-eed4-4f2a-8303-8d7b689ae96b","added_by":"auto","created_at":"2025-08-29 07:35:21","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":145858,"visible":true,"origin":"","legend":"\u003cp\u003eFE traits and SCFAs in two groups. \u003cstrong\u003eA \u003c/strong\u003eand \u003cstrong\u003eB.\u003c/strong\u003e Average daily gain (ADG) and feed conversion ratio (FCR) of CF and ZF ducks. \u003cstrong\u003eC\u003c/strong\u003e and \u003cstrong\u003eD.\u003c/strong\u003e Absolute concentrations of propionate and butyrate in two groups. *\u003cem\u003ep\u003c/em\u003e\u0026lt;0.05, ***\u003cem\u003ep\u003c/em\u003e\u0026lt;0.001, ****\u003cem\u003ep\u003c/em\u003e\u0026lt;0.0001.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-7357899/v1/346c9b40735df5e400d47786.png"},{"id":90152850,"identity":"d623ced5-0f85-47bb-a05c-e8fc0caa3887","added_by":"auto","created_at":"2025-08-29 07:35:21","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":661648,"visible":true,"origin":"","legend":"\u003cp\u003eTaxonomic compositions, diversity, and co-occurrence networks of gut bacterial communities revealed by 16S rRNA gene amplicon sequencing. \u003cstrong\u003eA. \u003c/strong\u003eThe average composition of top 15 bacterial communities at phylum level in CF and ZF duck populations. *\u003cem\u003ep\u003c/em\u003e\u0026lt;0.05, **\u003cem\u003ep\u003c/em\u003e\u0026lt;0.01. \u003cstrong\u003eB.\u003c/strong\u003e α-diversity (Chao1, Pielou, Richness, and Shannon indices) in two groups. \u003cstrong\u003eC.\u003c/strong\u003e Differential abundance of bacterial species between two groups (\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05). Values represent natural log fold changes from ANCOM-BC analysis. The color gradient indicates the direction and magnitude of change: blue to white represent decreases, while white to red represent increases in ZF compared with CF. \u003cstrong\u003eD.\u003c/strong\u003e Spearman correlation and linear regression analyses between bacterial species and feed conversion ratio (FCR). \u003cstrong\u003eE.\u003c/strong\u003e Spearman correlation and linear regression analyses between bacterial species and SCFAs (butyrate and propionate). The correlation coefficient r and the corresponding \u003cem\u003ep\u003c/em\u003evalue are displayed in each plot. The regression line (gray) is shown with its 95% confidence interval (shaded area). \u003cstrong\u003eF.\u003c/strong\u003eNon-metric multidimensional scaling (NMDS) analysis of the Bray–Curtis distances based on gut microbial species. Ellipsoids represent a 95% confidence interval surrounding each group. \u003cstrong\u003eG.\u003c/strong\u003e Relative abundance (total sum scaling) of aceE (K00163; pyruvate dehydrogenase) in two groups. \u003cstrong\u003eH.\u003c/strong\u003e Absolute concentrations of propionate and butyrate in two groups. \u003cstrong\u003eI.\u003c/strong\u003eBar plot showing the \u003cem\u003ep\u003c/em\u003e-values (-log10 transformed) of differentially abundant KOs in modules. Red, ZF-enriched KOs; blue, CF-enriched KOs. \u003cstrong\u003eJ.\u003c/strong\u003e Metabolic pathways involved in SCFAs biosynthesis and methanogenesis.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-7357899/v1/8aeebc1db3176f7bcf6320cf.png"},{"id":90152848,"identity":"71e20d89-6238-4e01-80e5-1895108c2607","added_by":"auto","created_at":"2025-08-29 07:35:21","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":232358,"visible":true,"origin":"","legend":"\u003cp\u003eThe viral contig characteristics and virus-host associations from 36 cecal metagenomes. \u003cstrong\u003eA.\u003c/strong\u003e The distribution of the estimated completeness of 1,986 viral operational taxonomic units (vOTUs). \u003cstrong\u003eB.\u003c/strong\u003e The classification information of the class level of 1,986 vOTUs. \u003cstrong\u003eC\u0026amp;D.\u003c/strong\u003e Number of temperate and virulent viral clusters (VCs) with \u0026gt;50% completeness for each predicted host at the phylum level.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-7357899/v1/8c2456bf530c857f4f175f4c.png"},{"id":90152851,"identity":"46e1f6f4-c274-48cf-bcef-f60f375dad27","added_by":"auto","created_at":"2025-08-29 07:35:21","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":380710,"visible":true,"origin":"","legend":"\u003cp\u003eViral diversity and their relationship with FE related traits. \u003cstrong\u003eA.\u003c/strong\u003e Viral diversity (Chao1, Pielou, Richness, Shannon index) between two groups. \u003cstrong\u003eB.\u003c/strong\u003e The NMDS plot of cecal viral communities between two groups based on Bray-Curtis dissimilarity. \u003cstrong\u003eC.\u003c/strong\u003e Spearman's rank correlation between FE traits and differential VCs, with color gradients representing positive (red) and negative (blue) correlation coefficients (\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05). \u003cstrong\u003eD.\u003c/strong\u003e Spearman's rank correlation between differential VCs in \u003cstrong\u003eC\u003c/strong\u003e and differential bacteria (CF-enriched bacteria are colored red). ***\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001, **\u003cem\u003ep\u003c/em\u003e\u0026lt; 0.01, *\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-7357899/v1/129d498e00487efd06354880.png"},{"id":90153917,"identity":"974f3dd0-8a8c-4503-89bf-6947aaf12b62","added_by":"auto","created_at":"2025-08-29 07:43:21","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":346892,"visible":true,"origin":"","legend":"\u003cp\u003eAuxiliary metabolic genes (AMGs) carried by duck cecal viruses. \u003cstrong\u003eA.\u003c/strong\u003e A bar plot showing the occurrence (log10) of AMGs identified in the temperate (green) and virulent (yellow) viruses. \u003cstrong\u003eB.\u003c/strong\u003e The relative abundance of two AMGs (\u003cem\u003ePPDK\u003c/em\u003e and \u003cem\u003epflD\u003c/em\u003e) in two groups. ***\u003cem\u003ep\u003c/em\u003e\u0026lt; 0.001, *\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05. \u003cstrong\u003eC.\u003c/strong\u003e Genome synteny maps of two viruses (ctg4378 is identified in this study) and one bacterium containing the \u003cem\u003ePPDK\u003c/em\u003e gene. \u003cstrong\u003eD.\u003c/strong\u003e Genome synteny maps of two viruses (ctg2056 is identified in this study) and one bacterium containing the \u003cem\u003epflD\u003c/em\u003e gene. Homologous genes from different organisms are filled in the same color and connected using the shading links. \u003cstrong\u003eE.\u003c/strong\u003e Predicted tertiary structure and function of the phage encoded PPDK protein. The phage-encoded PPDK structure is superimposed onto the reference structure of PPDK from \u003cem\u003eClostridium symbiosum\u003c/em\u003e (UniProt: P22983; colored silver) based on Foldseek search [60]. PPDK (EC 2.7.9.1) catalyzes the conversion of AMP, phosphoenolpyruvate, and diphosphate to ATP, pyruvate, and inorganic phosphate. \u003cstrong\u003eF.\u003c/strong\u003ePredicted tertiary structure and function of the phage encoded pflD protein. The phage-encoded pflD structure is superimposed onto the reference structure of pflD from \u003cem\u003eEnterococcus faecium\u003c/em\u003e(UniProt: A0A132P2N7; colored silver) based on Foldseek search. pflD (EC 2.3.1.54) catalyzes the conversion of pyruvate and CoA to formate and acetyl-CoA.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-7357899/v1/e785f75772c170f38707ef6e.png"},{"id":90153918,"identity":"903ba947-b7da-40d0-a54b-86279769e851","added_by":"auto","created_at":"2025-08-29 07:43:21","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":350209,"visible":true,"origin":"","legend":"\u003cp\u003eSerum metabolome and liver transcriptome profiles reveal systemic metabolic divergence between the two groups. \u003cstrong\u003eA.\u003c/strong\u003e Classification and composition of differentially abundant serum metabolites between two groups. \u003cstrong\u003eB.\u003c/strong\u003e Functional enrichment analysis of differentially abundant metabolites. \u003cstrong\u003eC.\u003c/strong\u003eKEGG pathway enrichment analysis of differentially expressed genes (DEGs) in the liver between two groups. \u003cstrong\u003eD.\u003c/strong\u003eRelative abundance of succinic semialdehyde in two groups. \u003cstrong\u003eE.\u003c/strong\u003e Expression level of succinic semialdehyde dehydrogenase (ALDH5A1) in two groups. **\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01, *\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-7357899/v1/13b832564e42b0af000a03b7.png"},{"id":90152858,"identity":"b7ef59fc-e1cf-47a2-8d34-bbc7304c3602","added_by":"auto","created_at":"2025-08-29 07:35:21","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":690290,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelations between gut microbiome and serum metabolome. \u003cstrong\u003eA.\u003c/strong\u003e Spearman’s rank correlations between differential serum metabolites and FE traits. \u003cstrong\u003eB.\u003c/strong\u003e Spearman’s rank correlations between FE-associated metabolites and differentially abundant bacteria. \u003cstrong\u003eC.\u003c/strong\u003e Spearman’s rank correlations between FE-associated metabolites and differential viruses. Metabolites are clustered by their biochemical classifications. ***\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001, **\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01, *\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-7357899/v1/b706dcd2db638c92bdb4034d.png"},{"id":90152864,"identity":"d2fb692d-4360-44c1-bae0-8ea0bfc8e37e","added_by":"auto","created_at":"2025-08-29 07:35:22","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":186251,"visible":true,"origin":"","legend":"\u003cp\u003eThe potential mechanism by which bacteria and bacteriophage regulate cecal SCFA production and host metabolism to improve FE in ducks. Bacteriophages modulate bacteria to produce SCFAs through encoded AMGs, PPDK and pflD.\u003c/p\u003e","description":"","filename":"image8.png","url":"https://assets-eu.researchsquare.com/files/rs-7357899/v1/3aa995ad994e5130416632ed.png"},{"id":104251593,"identity":"c688c725-156d-4c90-888f-d97d8b38efe4","added_by":"auto","created_at":"2026-03-09 16:14:15","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4143108,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7357899/v1/934d3d19-1f39-46a7-816b-3535433be252.pdf"},{"id":90152854,"identity":"94c8022a-ffe1-4e74-bd60-2045cf2c22f7","added_by":"auto","created_at":"2025-08-29 07:35:21","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1727537,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Fig.S1: FE traits and SCFAs concentrations between two groups.\u003c/p\u003e\n\u003cp\u003eSupplementary Fig.S2: Bacterial composition, network analysis, and functional correlations between two groups.\u003c/p\u003e\n\u003cp\u003eSupplementary Fig.S3: Host association and differential abundance of viral clusters (VCs).\u003c/p\u003e\n\u003cp\u003eSupplementary Fig.S4: Volcano plot of differentially abundant metabolites in CF vs. ZF.\u003c/p\u003e\n\u003cp\u003eSupplementary Table S1: Growth, feed efficiency and fatness traits in two groups.\u003c/p\u003e\n\u003cp\u003eSupplementary Table S2: Comparative topological properties of the CF and ZF networks.\u003c/p\u003e\n\u003cp\u003eSupplementary Table S3: Degree and eigenvector centrality of the hub nodes in CF and ZF networks.\u003c/p\u003e","description":"","filename":"suppduckBiome0812.docx","url":"https://assets-eu.researchsquare.com/files/rs-7357899/v1/7f4dadab1615a8f46681c80f.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Bacteria and bacteriophage consortia modulate cecal SCFA production and host metabolism to enhance feed efficiency in ducks","fulltext":[{"header":"Introduction","content":"\u003cp\u003eEfficient feed utilization is critical for sustainable and profitable poultry production. Feed costs account for 60\u0026ndash;70% of total production expenses in commercial poultry operations, making feed efficiency (FE) a primary determinant of economic viability [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. In the context of a growing global demand for animal protein, enhancing FE in poultry species like chickens and ducks holds significant economic and environmental importance through lower feed waste and nitrogen excretion.\u003c/p\u003e\u003cp\u003eIn non-ruminant animals like poultry and pigs, short-chain fatty acids (SCFAs), primarily produced through microbial fermentation of fiber in the cecum or colon, can supply 60\u0026ndash;70% of the energy requirements of intestinal epithelial cells [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. This process not only sustains normal intestinal epithelial function but also plays a vital role in overall gut health and metabolic regulation. Beyond energy provision, SCFAs (such as acetate, propionate, and butyrate) function as signaling molecules by activating G protein-coupled receptors (GPCRs), including FFAR2 (GPR43), FFAR3 (GPR41), and HCAR2 (GPR109A). Activation of these receptors modulates energy expenditure, preadipocyte differentiation, and appetite control [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Given these functions, modulating the gut microbiota to enhance SCFA production represents a promising strategy for improving feed utilization. For examples, probiotic supplementation accelerates gut microbiota maturation, increasing average daily gain (ADG) by 6.2% while reducing FCR by 4.5% compared to antibiotic-free controls [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Similarly, microbiota-accessible fibers boost populations of SCFA-producing bacteria (e.g., \u003cem\u003eBacteroidetes\u003c/em\u003e) by up to 40%, enhancing energy harvest and intestinal barrier function in laying hens [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. While bacterial contributions to FE through SCFA production are well-documented [\u003cspan additionalcitationids=\"CR8\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e], the specific roles of bacteriophages in modulating these bacterial functions and their combined impact on duck feed efficiency remain uncharacterized, preventing the development of comprehensive microbiome-based strategies to optimize feed utilization.\u003c/p\u003e\u003cp\u003eThe gut virome, dominated by bacteriophages (phages infecting bacteria), represents one of the densest viral ecosystems in nature, with estimates of more than 10\u003csup\u003e8\u003c/sup\u003e to 10\u003csup\u003e10\u003c/sup\u003e virus-like particles per gram of intestinal content [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. These phages exert profound influence over microbial community dynamics and metabolic processes through predator-prey interactions, horizontal gene transfer, and the regulation of bacterial lysogenic-to-lytic life cycles. For instance, phages can reshape bacterial diversity by selectively lysing dominant taxa like \u003cem\u003eBacteroides\u003c/em\u003e or \u003cem\u003eBacillota\u003c/em\u003e, thereby altering niche competition and nutrient flow [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Phage-encoded auxiliary metabolic genes (AMGs) further modulate bacterial metabolism, such as enhancing carbohydrate utilization in \u003cem\u003eBacteroides\u003c/em\u003e [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e] and \u003cem\u003eOscillospiraceae\u003c/em\u003e [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Such phage activities can cascade to reshape microbial functional outputs, such as bile acid metabolism and SCFA production, with indirect implications for host physiology.\u003c/p\u003e\u003cp\u003eDespite their ubiquity and functional significance, the role of gut viruses in shaping host phenotypes remains understudied outside human disease contexts (e.g., inflammatory bowel disease, where phage dysbiosis alters bacterial consortia and immunity [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]). In livestock, emerging evidence suggests agricultural relevance: rumen phages in cattle may regulate host nutrient utilization via AMGs and virus-host linkages, potentially influencing FE [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. These findings position the virome as a hidden effector bridging microbial ecology and host biology. However, data on viral-bacterial interactions and their association with FE in ducks are limited, and specific microbial taxa or functional genes linked to duck FE remain unidentified. Elucidating these interactions is essential for developing microbiome-based strategies to enhance FE in ducks.\u003c/p\u003e\u003cp\u003eThis study characterized gut microbial communities, including bacteria and bacteriophages, in ducks stratified by FE. We hypothesized that high-FE ducks harbor distinct gut microbiota compositions and functional capabilities compared with low-FE ducks. We further aimed to resolve virus-host dynamics and assess their impact on bacterial community structure and function. By integrating 16S amplicon, metagenome, SCFAs, liver transcriptome, and serum metabolome data, we provide a comprehensive assessment of the duck gut microbiome\u0026rsquo;s association with FE. Our findings may inform microbiome-targeted strategies (e.g., probiotics, dietary interventions) to optimize feed utilization in duck production.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cb\u003e1. FE traits and SCFAs in CF and ZF ducks\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe statistics of average daily feed intake (ADFI), average daily gain (ADG), feed conversion ratio (FCR), and fatness traits between two duck populations are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA\u0026amp;B and Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e. Compared to ZF ducks, CF ducks exhibited 5.45% higher ADFI (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0001), 4.59% lower ADG (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), 9.45% higher FCR (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0001), 55.83% greater SFW (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and 26.64% greater AFW (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001). These results demonstrate significantly superior FE and reduced adiposity in the ZF population relative to CF ducks.\u003c/p\u003e\u003cp\u003eSCFAs derived from gut microbiota have been proven to be associated with animal FE previously [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. To compare cecal SCFAs in two groups, we analyzed 10 SCFAs from a randomly selected subset of both groups (20 CF, 16 ZF), which also showed significant difference in FCR and ADG (Fig.\u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eA\u0026amp;B), using targeted metabolomics. We found that ZF ducks exhibited higher propionate and butyrate concentrations in cecal content compared with CF ducks (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), while the others have no significant difference (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC\u0026amp;D and Fig.\u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eC-J). These results combined suggest that higher propionate and butyrate may improve FE in ducks.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003e2. Divergence of cecal bacterial diversity, composition, and function in High- and Low-FE Ducks\u003c/b\u003e\u003c/p\u003e\u003cp\u003eBecause SCFAs in duck is primarily derived from microbial fermentation of fiber in the cecum, we first investigated cecal bacterial composition and interactions in CF and ZF groups using 16S rRNA gene amplicon sequencing, which revealed a conserved phylum-level structure across both groups, dominated by \u003cem\u003eBacteroidetes\u003c/em\u003e followed by \u003cem\u003eBacillota_A\u003c/em\u003e and \u003cem\u003eBacillota_C\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA and Fig. S2A-F). While observed ASV richness was comparable between groups, ZF ducks exhibited significantly elevated α-diversity versus CF ducks as measured by Chao1, Pielou, and Shannon indices (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). The co-occurrence networks of two groups have low adjust rand index (ARI) (0.136) [\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e], indicating that the network structural and topological characteristics varied for each group. Detailed information of network topology is shown in Fig.S2G\u0026amp;H and Tables S2-3.\u003c/p\u003e\u003cp\u003eTo further identify the specific intestinal microbial species that contribute to FE, we analyzed cecal microbiome from the subset of both groups (20 CF, 16 ZF) using metagenomics. Prokaryotic genome reconstruction from metagenomic data yielded 1,820 MAGs through single-sample assembly (mean contigs per MAG: 482). Quality filtering (completeness\u0026thinsp;\u0026ge;\u0026thinsp;50%, contamination\u0026thinsp;\u0026lt;\u0026thinsp;10%) retained 943 medium-to-high-quality MAGs. Taxonomic classification assigned these MAGs to 16 phyla, with 890 MAGs (94.4%) resolving to 121 genera and 620 MAGs to 144 known species. Notably, 327 MAGs (34.7%) lacked matches to reference genomes, representing potentially novel species.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eWe performed the analysis of composition of microbiomes with bias correction (ANCOM-BC) and identified 21 species significant different in terms of abundance between two groups (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC). Specifically, gut bacterial communities in ZF ducks were mainly enriched in these species, such as \u003cem\u003ePhocaeicola barnesiae\u003c/em\u003e, \u003cem\u003eMerdibacter sp900759455\u003c/em\u003e, \u003cem\u003eParaprevotella stercoravium\u003c/em\u003e, \u003cem\u003eParaprevotella stercorigallinarum\u003c/em\u003e, and \u003cem\u003eAvelusimicrobium sp944324735\u003c/em\u003e, while CF animals were more abundant in \u003cem\u003eAlistipes finegoldii\u003c/em\u003e, \u003cem\u003eDesulfovibrio sp944324525\u003c/em\u003e, \u003cem\u003egemmiger formicilis_B\u003c/em\u003e, and \u003cem\u003eUMGS1491 sp905211185\u003c/em\u003e. We then performed spearman correlations between these differentially abundant bacterial species and FCR/ADG among individuals. Multiple species, including \u003cem\u003ePhocaeicola barnesiae\u003c/em\u003e, \u003cem\u003eUMGS263 sp949289045\u003c/em\u003e, \u003cem\u003eParaprevotella sp905215575\u003c/em\u003e, and \u003cem\u003eBacteroides sp944322345\u003c/em\u003e, were negatively correlated with FCR (\u003cem\u003er\u003c/em\u003e = -0.43 ~ -0.37, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) but positively correlated with ADG. \u003cem\u003eUMGS1491 sp905211185\u003c/em\u003e exhibited a positive relationship with FCR (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.40, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) but negative correlation with ADG (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD and Fig.S2I). We further analyzed the correlation of these differential bacterial species with propionate or butyrate, the results revealed most of differential bacteria, such as \u003cem\u003eUMGS1449 sp900551925\u003c/em\u003e, \u003cem\u003eParaprevotella sp905215575\u003c/em\u003e, and \u003cem\u003eMediterraneibacter quadrami\u003c/em\u003e, were positively correlated with propionate or butyrate (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.33\u0026thinsp;~\u0026thinsp;0.39, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), while a few of them, including \u003cem\u003eDesulfovibrio sp944324525\u003c/em\u003e and \u003cem\u003eGemmiger formicilis_B\u003c/em\u003e, displayed negative association with propionate (\u003cem\u003er\u003c/em\u003e = -0.46 ~ -0.34, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE and Fig.S2I).\u003c/p\u003e\u003cp\u003eTo further compare the functional features of microbiota in two groups, we annotated the microbial functions of each metagenome via the KEGG database using a gene-centric approach [\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e] and quantified the abundance of each KEGG orthologous (KOs) in different samples. Non-metric multidimensional scaling (NMDS) plot showed an obvious difference in the gut microbial KOs between the CF and ZF groups (PERMANOVA, R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.167, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eF). We identified 545 differentially abundant KOs (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), with 339 (62.2%) enriched in ZF ducks. These differential KOs were significantly enriched in multiple KEGG pathways including \u0026ldquo;Carbon metabolism\u0026rdquo;, \u0026ldquo;Sulfur metabolism\u0026rdquo;, \u0026ldquo;Phosphonate and phosphinate metabolism\u0026rdquo;, \u0026ldquo;Methane metabolism\u0026rdquo;, \u0026ldquo;Citrate cycle (TCA cycle)\u0026rdquo;, \u0026ldquo;Nucleotide metabolism\u0026rdquo;, and \u0026ldquo;Pyruvate metabolism\u0026rdquo; (Fig.S2J). Because \u0026ldquo;Pyruvate metabolism\u0026rdquo; is a central hub in the anaerobic glycolysis pathway of gut bacteria, driving the synthesis of short-chain fatty acids (SCFAs) like acetate, propionate, and butyrate, ZF microbiomes showed significant enrichment of aceE (K00163; pyruvate dehydrogenase) within pyruvate oxidation modules (M00307), converting pyruvate to acetyl-CoA, the central precursor for SCFA biosynthesis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eG). This predicted enhanced SCFA production was consistent with our experimental measurement that ZF ducks exhibited higher propionate and butyrate concentrations in cecal content compared with CF ducks (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e\u003cp\u003eRegarding carbohydrate metabolism and energy metabolism, it is worth mentioning that the KOs associated with \u0026ldquo;Methane metabolism\u0026rdquo; (9/12 KOs) was predominantly enriched in the CF microbiome. While the KOs involved in \u0026ldquo;Citrate cycle (TCA cycle)\u0026rdquo; (5/6 KOs), \u0026ldquo;Pentose and glucuronate interconversions\u0026rdquo; (6/6 KOs), and \u0026ldquo;Phosphonate and phosphinate metabolism\u0026rdquo; (5/6 KOs) were more abundant in the gut microbiome of ZF group (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eH\u0026amp;I). These enrichment patterns align with known energy-partitioning mechanisms: ZF\u0026rsquo;s enhanced pentose/glucuronate interconversions promote SCFA yield [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e], while CF\u0026rsquo;s methane metabolism redirects carbon from SCFA production [\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e].\u003c/p\u003e\u003cp\u003e\u003cb\u003e3. Gut virome characterization and host interactions\u003c/b\u003e\u003c/p\u003e\u003cp\u003eMetagenomic assembly of 36 cecal samples yielded 4,334 quality-filtered viral contigs. Among these, 835 high-confidence contigs (completeness\u0026thinsp;\u0026ge;\u0026thinsp;50%) had a mean length of 41,983 bp (N50\u0026thinsp;=\u0026thinsp;50,390 bp). Clustering at 95% average nucleotide identity (ANI) identified 1,986 viral operational taxonomic units (vOTUs), including 428 with \u0026ge;\u0026thinsp;50% completeness (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). All vOTUs represented bacteriophages, predominantly within \u003cem\u003eCaudoviricetes\u003c/em\u003e (tailed phages), \u003cem\u003eMalgrandaviricetes\u003c/em\u003e (microviruses), \u003cem\u003eMegaviricetes\u003c/em\u003e, and \u003cem\u003eArfiviricetes\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). Protein-sharing clustering of high-quality vOTUs generated 392 viral clusters (VCs). Host prediction revealed a total of 1,984 virus-host linkages, comprising 808 lytic VC-hMAG (host metagenome assembled genome) and 1,176 temperate VC-hMAG linkages. The hosts of temperate viruses spanned across 11 phyla (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC), whereas those of lytic viruses were associated with 12 phyla (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD). As expected, \u003cem\u003eBacteroidota\u003c/em\u003e (comprising 58.17% of lytic and 69.56% temperate viral hosts) and \u003cem\u003eBacillota_A\u003c/em\u003e (comprising 15.97% of lytic and 15.73% temperate viral hosts), highly predominant gut bacterial taxa, were the most common hosts of gut viruses at phylum level. \u003cem\u003eBacteroides\u003c/em\u003e and \u003cem\u003ePhocaeicola\u003c/em\u003e were the most prevalent hosts at genus level (Fig. S3A\u0026amp;B). Among these VC-hMAG linkages, VC2778, VC2779, and VC2563 exhibited the highest host diversity, each associated with 62, 54, and 49 distinct host species (Fig. S3C).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003e4. Virome divergence and metabolic modulation in high- and low-FE Ducks\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo compare gut viral communities between two groups, we examined viral α- and β-diversity of two groups. The result showed that two groups exhibited similar α-diversity profiles except for the Shannon index at the VC level (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA), while β-diversity analysis (NMDS, Bray-Curtis) revealed distinct clustering of individuals from two groups (PERMANOVA, R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.096, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB).\u003c/p\u003e\u003cp\u003eTo further investigate gut viral signatures associated with FE, we performed ANCOM-BC analysis and identified that 47 VCs were significantly decreased, while 32 VCs were increased in CF group compared with ZF group (Fig S3D). Then we did the correlation analysis between these differential VCs and FE-related phenotypes. 12 viruses (5 positive and 7 negative) and 10 viruses (6 positive and 4 negative) displayed significant correlations with FCR and ADG, respectively (Spearman, |\u003cem\u003er\u003c/em\u003e| \u0026ge; 0.3, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC). For example, VC2859, VC2785, and VC0894 showed consistent positive ADG or negative FCR correlations. As phages may link to phenotype by modulating the composition and dynamics of bacterial communities, we predicted interactions of these FE-associated VCs (either co-occurrence or mutual exclusion) with differentially abundant bacteria (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD). Spearman correlation analysis revealed distinct phage-bacteria association patterns. One group of VCs, including VC2785, VC2396, VC2667, VC0894, and VC2518) displayed significant positive correlations (co-occurrence) with high-FE favored bacteria (e.g., \u003cem\u003eParaprevotella sp905215575\u003c/em\u003e, \u003cem\u003ePhocaeicola barnesiae\u003c/em\u003e, \u003cem\u003eBacteroides sp944322345\u003c/em\u003e, \u003cem\u003eParaprevotella stercorigallinarum\u003c/em\u003e) (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.34\u0026thinsp;~\u0026thinsp;0.75, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), while exhibiting significant negative correlations (mutual exclusion) with low-FE favored bacteria (e.g., \u003cem\u003eAlistipes finegoldii\u003c/em\u003e, \u003cem\u003eUMGS1491 sp905211185\u003c/em\u003e, \u003cem\u003eDesulfovibrio sp944324525\u003c/em\u003e, \u003cem\u003eGemmiger formicilis_B\u003c/em\u003e) (\u003cem\u003er\u003c/em\u003e = -0.70 ~ -0.33, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Conversely, a second group of VCs (e.g., VC4010, VC2559, VC2862, VC2755, VC2722, VC1483) displayed the opposite association pattern, positively correlating with low-FE favored bacteria and negatively correlating with high-FE favored bacteria.\u003c/p\u003e\u003cp\u003eThen we searched AMGs from viral contigs and found 109 AMGs that were involved in a wide range of metabolic processes, including the metabolism of nucleotides, carbohydrates, amino acids, and sulfur (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). The most prevalent AMG was DNMT1, a DNA (cytosine-5)-methyltransferase that prevents viruses from being recognized and cleaved by the restriction-modification systems of their hosts [\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e], which was detected in more than 40 high quality viral contigs. CysH was the second highly abundant AMG which encodes a phosphoadenosine phosphosulfate reductase involved in the synthesis of sulfite from sulfate. Additionally, another abundant AMG, metK encoding S-adenosylmethionine (SAM) synthetase, was important for genomic DNA methylation and cell division. Two AMGs, NAMPT and pncA, were important for maintaining duck\u0026rsquo;s NAD\u0026thinsp;+\u0026thinsp;level which is critical for maintaining energy homeostasis. CobS and cobT were two key AMGs involved in synthesis of vitamin B12.\u003c/p\u003e\u003cp\u003eWe subsequently conducted a differential analysis of AMGs between two groups and found 16 AMGs showing significant differential abundance. Notably, two differential AMGs, pyruvate phosphate dikinase (PPDK) and formate C-acetyltransferase (pflD), involved in pyruvate metabolism were identified in viral contigs (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). This PPDK shared 83% amino acid identity with PPDK in \u003cem\u003ePrevotella sp. E2-28\u003c/em\u003e (GenBank: CP091788) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC), while the pflD was 52% identical to a pflD in \u003cem\u003eCatenibacterium mitsuokai strain DSM 15897\u003c/em\u003e (GenBank: CP102271) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD). Specifically, in the anaerobic cecal environment, temperate virus (VC2785)-encoded PPDK, catalyzing the reversible conversion of phosphoenolpyruvate (PEP) to pyruvate and ATP [\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e], was higher abundant in ZF group compared with CF group, which may augment the fermentative capacity of host bacteria (e.g., \u003cem\u003eBacteroides\u003c/em\u003e, \u003cem\u003eParaprevotella\u003c/em\u003e, and \u003cem\u003ePhocaeicola\u003c/em\u003e) in ZF group (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eE). This is in line with abovementioned observation that VC2785 displayed positive correlation with both ADG and propionate (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD). While lytic virus (VC0717)-encoded pflD, converting pyruvate to acetyl-CoA and formate, was higher abundant in CF group compared with ZF group, which may accelerate viral replication by hijacking the host\u0026rsquo;s metabolic machinery in CF group (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eF). Moreover, its host, \u003cem\u003ePrevotella sp015074785\u003c/em\u003e, belonging to the \u003cem\u003ePrevotellaceae\u003c/em\u003e family, significantly contributes to the production of SCFAs which may induce lytic viral replication and lysis of cells [\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e]. These results together suggest that gut viruses could potentially optimize nutrient availability for their hosts or exploit host central metabolic pathways for their own replication, underscoring their influence on host physiology including FE.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003e5. Serum metabolome profiling reveals systemic metabolic divergence between two groups\u003c/b\u003e\u003c/p\u003e\u003cp\u003eSerum metabolomics identified 221 differentially abundant metabolites between CF and ZF ducks (|log\u003csub\u003e2\u003c/sub\u003eFC| \u0026ge; 1, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), with 184 upregulated and 37 downregulated in CF compared with ZF (Fig. S4). According to the annotation of Human Metabolome Database (HMDB), these differential metabolites were mainly categorized into lipids and lipid-like molecules (38.01%), organic acids and derivatives (10.41%), organoheterocyclic compounds (9.95%), and organic oxygen compounds (5.43%) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA). Metabolite pathway enrichment analysis indicated that these differential metabolites were mainly enriched in 13 pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB). Among them, 3 pathways including \u0026ldquo;pyrimidine metabolism\u0026rdquo;, \u0026ldquo;butanoate metabolism\u0026rdquo;, and \u0026ldquo;pantothenate and CoA biosynthesis\u0026rdquo; showed concordant enrichment in liver transcriptomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC). For example, in butanoate metabolism, we observed higher abundance of succinic semialdehyde in CF group compared with ZF group (log\u003csub\u003e2\u003c/sub\u003eFC\u0026thinsp;=\u0026thinsp;1.37, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), this may stem from downregulation of the succinic semialdehyde dehydrogenase (ALDH5A1) gene (log\u003csub\u003e2\u003c/sub\u003eFC = -0.75, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01), a mitochondrial homotetramer protein that catalyzes the NAD+ -dependent conversion of succinic semialdehyde into succinate which can be further fermented to butanoate (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eD\u0026amp;E). In functional enrichment analysis of DEGs, we also observed multiple pathways related to lipid metabolism, such as \u0026ldquo;insulin resistance\u0026rdquo;, \u0026ldquo;PPAR signaling pathway\u0026rdquo;, and \u0026ldquo;glycerophospholipid metabolism\u0026rdquo;, explaining the predominant lipid metabolite differences observed between two groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003e6. Gut microbiome-serum metabolome associations with feed efficiency\u003c/b\u003e\u003c/p\u003e\u003cp\u003eAmong differential serum metabolites, 37 metabolites were significantly correlated with FE traits (|\u003cem\u003er\u003c/em\u003e| \u0026ge; 0.3, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), with most of them positively correlated with FCR (21/37) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eA). For example, L-histidine and 4-hydroxydecanedioylcarnitine were negatively correlated with ADG, while positively associated with FCR. To detect the potential effect of differentially abundant bacteria on host FE-associated metabolites, spearman\u0026rsquo;s rank correlations revealed a total of 188 strong correlations (|\u003cem\u003er\u003c/em\u003e| \u0026ge; 0.5, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05; Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eB). Among these correlations, most of them were negative between differential bacteria and FE-associated metabolites (e.g., \u003cem\u003eParaprevotella sp905215575\u003c/em\u003e vs. L-histidine: \u003cem\u003er\u003c/em\u003e = -0.67), while \u003cem\u003eUMGS1491 sp905211185\u003c/em\u003e positively correlated with multiple FE-linked metabolites (e.g., 4-hydroxydecanedioylcarnitine: \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.48). Spearman\u0026rsquo;s rank correlations between differential viruses and FE-associated metabolites were also performed, most FE-associated metabolites showed positive correlations with 26 VCs (e.g., VC2540 vs. 4-hydroxydecanedioylcarnitine: \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.64) and negative correlations with 27 VCs (e.g., VC2785 vs. L-histidine: \u003cem\u003er\u003c/em\u003e = -0.56), respectively, while the correlation of CDP-DG(a-25:0/20:4(6E,8Z,11Z,14Z)\u0026thinsp;+\u0026thinsp;=\u0026thinsp;O(5)) and (2R)-O-Phospho-3-sulfolactate with VCs showed the reversed pattern (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eC).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn the past decade, gut microbiome and its associated roles have attracted increasing attention. Although several studies have investigated the bacterial contributions to FE of poultry using 16S rRNA sequencing [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e], the role of gut virome remains poorly characterized. Our multi-omics integration (metagenomics, transcriptomics, metabolomics) reveals how both bacterial and viral components of the duck gut microbiome collectively influence FE-associated phenotypes.\u003c/p\u003e\u003cp\u003eWe identified that ZF ducks have higher FE traits and cecal SCFAs (e.g., propionate and butyrate) concentration compared with CF group. Due to pivotal roles of SCFAs in maintaining energy homeostasis and promoting animal health through multifaceted mechanisms, including maintenance of intestinal barrier integrity and improvement of nutrient absorption [\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e], thus animals producing more SCFAs usually may have high FE. Previous studies showed that butyrate supplementation improved FE in chicken and duck [\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e]. In a sheep study, higher concentration of propionate was also detected in rumen of high FE group [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eIn duck, SCFAs are chiefly produced in the cecum, thus the difference of SCFAs between two groups may arise from variation of cecal bacterial diversity, composition, and function. We found that cecal bacteria in high-FE (ZF) and low-FE (CF) ducks shared core phyla (\u003cem\u003eBacteroidetes\u003c/em\u003e and \u003cem\u003eBacillota\u003c/em\u003e) but exhibited divergent diversity and network architectures. ZF microbiomes displayed higher modularity (Fig. S2G\u0026amp;H), indicative of more persistent and stable community-level and within-community interactions [\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e]. Besides network topology, the relative abundances of bacteria at various taxonomic levels in two groups also exhibit differences, most of the differential species are enriched in ZF group. Crucially, these ZF-enriched bacteria (e.g., \u003cem\u003ePhocaeicola barnesiae\u003c/em\u003e, \u003cem\u003eParaprevotella sp905215575\u003c/em\u003e, and \u003cem\u003eBacteroides sp944322345\u003c/em\u003e) potentially drove efficient pyruvate flux toward SCFA production rather than methanogenesis (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE-F). This is consistent with the observation that elevated cecal propionate/butyrate in ZF ducks. It has reported that \u003cem\u003eParaprevotella\u003c/em\u003e was more enriched in the leaner individuals [\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e] which usually have high FE traits [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e]. Conversely, CF-enriched \u003cem\u003eDesulfovibrio sp944324525\u003c/em\u003e and \u003cem\u003eGemmiger formicilis_B\u003c/em\u003e correlated negatively with SCFAs (Fig.S2I), aligning with inefficient energy harvesting, this was also observed in chicken gut [\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eAs the \u0026ldquo;dark matter\u0026rdquo; of the gut microbiome, bacteriophages (predominantly \u003cem\u003eCaudoviricetes\u003c/em\u003e) can alter the bacterial community structure by lysing certain bacteria or through lysogeny, thereby affecting animal phenotypes. There was significant difference in β diversity of viruses, suggesting distinct viral community structure in two groups. In addition, multiple differentially abundant viruses showed significant correlation with FE traits in ducks. From the predicted phage-host linkages, we found that some FE-correlated phages could target SCFA producers, such as \u003cem\u003eParaprevotella sp905215575\u003c/em\u003e and \u003cem\u003eBacteroides sp944322345\u003c/em\u003e, which are favorable for high FE. These results position phages as key phenotype mediators, which is supported by our human gut findings [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eDuring the phage infection of bacteria, AMGs carried by bacteriophages play crucial roles in optimizing the host\u0026rsquo;s metabolism and nutrient acquisition, which affect several important ecological processes, including methane metabolism in different habitats [\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e], polysaccharide digestion in the ruminants [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], and SCFAs metabolism in the human gut [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. However, the contribution of viral AMG in poultry gut has not been reported yet. Our results identified a total of 109 AMGs in duck cecum, which participated multiple processes, such as antiviral defense (e.g., DNMT1), sulfur metabolism (e.g., cysH and metK), energy homeostasis (e.g., NAMPT and pncA), and vitamin biosynthesis (cobS and cobT), suggesting that the gut virome can also affect feed digestibility in duck. We identified a total of 16 differentially abundant AMGs between two groups, with two AMGs, PPDK and pflD, involved in pyruvate metabolism process. In anaerobic environments like the cecum, microbes rely on fermentation pathways. PPDK enables the regeneration of PEP, which serves as a key intermediate for substrate-level phosphorylation, a primary ATP-generating mechanism under anaerobic conditions [\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e]. Given the requisite integration of temperate viruses into the host genome, the temperate phages-encoded PPDK enriched in ZF potentially boosted pyruvate metabolism and SCFAs production of host bacteria, thereby conferring communal benefits and, in turn, reciprocal advantages to the viruses themselves [\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e]. While the lytic phage-encoded pflD enriched in CF could hijack host resources to produce more acetyl-CoA (SCFAs), which may promote viral infection and host lysis [\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e]. However, the underlying mechanisms remain to be explored and need further experimental validation.\u003c/p\u003e\u003cp\u003eBecause gut microbiotas produce a variety of molecules, some of which are taken up into the bloodstream and affect health. Therefore, characterization of the interactions between gut microbiota and host plasma metabolites is also important for understanding the effects of the gut microbiota on animal phenotypes. Our results revealed that L-histidine and succinic semialdehyde were more abundant in serum of CF group compared with ZF group. Consistently, recent papers have reported that plasma L-histidine was downregulated in lambs with high ADG [\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e]. Besides, our RNA-seq data revealed reduced hepatic ALDH5A1 expression in CF ducks, which could impair succinic semialdehyde\u0026rarr;butanoate conversion and lead to accumulation of succinic semialdehyde (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eD). Intriguingly, succinic semialdehyde level were found to be significantly positively correlated with insulin resistance and liver fat content [\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e], concordant with dysregulated lipid pathways, such as \u0026ldquo;insulin resistance\u0026rdquo;, \u0026ldquo;PPAR signaling pathway\u0026rdquo;, and \u0026ldquo;glycerophospholipid metabolism\u0026rdquo;, in CF group.\u003c/p\u003e\u003cp\u003eWe reported that these serum FE related metabolites were associated with duck cecal microorganisms, where most of them displayed strong negative associations with multiple differential bacterial and viral species between groups. The negative correlations of plasma L-histidine level with \u003cem\u003eBacteroides\u003c/em\u003e and \u003cem\u003eParaprevotella\u003c/em\u003e are in harmony with the results observed in piglets, rats, and fish [\u003cspan additionalcitationids=\"CR77\" citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e]. This is probably due to the contribution of \u003cem\u003eBacteroides\u003c/em\u003e and \u003cem\u003eParaprevotella\u003c/em\u003e genera to L-histidine degradation [\u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e, \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e]. Due to \u0026ldquo;virus-host\u0026rdquo; interactions, where viruses lysis host bacteria or boost their metabolism, gut viruses also display correlations with plasma metabolites. However, these links still require experimental validation in further study. Such information will provide evidence highlighting the possibility of animal blood metabolites through manipulating gut microbial functions.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eTaken together, to elucidate the mechanisms by which gut microbiota affect FE of ducks, we proposed two potential convergent pathways: (1) SCFAs are the major end products from the fermentation of gut bacteria. Viruses directly target SCFA-producing bacteria, altering community function and output, thereby jointly impacting FE through collaborative interactions with bacteria; (2) Gut bacteria and viruses interact to modulate host animal metabolism via blood circulation and thus influencing FE (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). However, further studies are required to validate the function of bacteriophages-encoded AMGs and causal relationships between representative SCFA-producing bacteria and bacteriophage consortia and host physiological outcomes.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003eAnimals management\u003c/h2\u003e\u003cp\u003eTwo Pekin duck populations with divergent FE were used: a high-FE group (ZF) and a low-FE group (CF). A total of 300 one-day-old male ducklings (150 per group) with similar initial body weights were housed individually in wire-floor cages (200 \u0026times; 100 \u0026times; 40 cm) equipped with nipple drinkers and tubular feeders at the National Changping Comprehensive Agricultural Engineering and Technology Research Center (Livestock and Poultry Sub-Center). Ducks were reared in two phases (days 1\u0026ndash;14 and days 15\u0026ndash;42) according to standardized practices. Both groups received identical commercial diets (Hope Feed Co., Ltd., Fangshan District, Beijing, China); diet composition and nutritional components matched those described in our previous study [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eFeed Efficiency Measurement and Sample Collection\u003c/h3\u003e\n\u003cp\u003eBody weight (BW) was recorded at 22 and 42 days of age. Total feed intake (FI) was measured from 22 to 42 days. Average daily feed intake (ADFI), average daily gain (ADG), and feed conversion ratio (FCR) were calculated as follows:\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eADFI (g/day)\u0026thinsp;=\u0026thinsp;Total FI / 21 days (1)\u003c/p\u003e\u003cp\u003eADG (g/day) = (BW\u003csub\u003e42\u003c/sub\u003e \u0026ndash; BW\u003csub\u003e22\u003c/sub\u003e) / 21 days (2)\u003c/p\u003e\u003cp\u003eFCR\u0026thinsp;=\u0026thinsp;Total FI / (BW\u003csub\u003e42\u003c/sub\u003e \u0026ndash; BW\u003csub\u003e22\u003c/sub\u003e) (3)\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eValid phenotypic records were obtained for 283 ducks. After a 12-hour fast, ducks were humanely euthanized according to standard commercial protocols. Fatness traits were recorded, including skin fat weight (SFW), skin fat percentage (SFP), abdominal fat weight (AFW), and abdominal fat percentage (AFP), calculated as:\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eSFP (%) = (SFW / carcass weight) \u0026times; 100 (4)\u003c/p\u003e\u003cp\u003eAFP (%) = (AFW / carcass weight) \u0026times; 100 (5)\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eDNA Extraction, 16S amplicon and Metagenomic Sequencing\u003c/h2\u003e\u003cp\u003eDucks with valid FCR records are euthanized by cervical dislocation and dissected. Cecal contents (including chyme and mucosa) were immediately collected, snap-frozen in liquid nitrogen, and stored at \u0026minus;\u0026thinsp;80\u0026deg;C until further processing.\u003c/p\u003e\u003cp\u003eMicrobial DNA was extracted from cecal samples of 283 ducks using the QIAamp DNA Stool Mini Kit (QIAGEN, Hilden, Germany). The V3-V4 hypervariable region of the bacterial 16S rRNA gene was amplified using primers 338F (5\u0026prime;-ACTCCTACGGGAGGCAGCA-3\u0026prime;) and 806R (5\u0026prime;-GGACTACHVGGGTWTCTAAT-3\u0026prime;). Amplified libraries were sequenced (2 \u0026times; 250 bp) on an Illumina NovaSeq 6000 platform.\u003c/p\u003e\u003cp\u003eA subset of 40 ducks (20 CF, 20 ZF) was randomly selected. Microbial DNA was extracted from cecal content using a CTAB-based method. DNA quality and quantity were assessed using a Qubit 3.0 Fluorometer (Invitrogen). Four ZF samples were excluded due to insufficient DNA yield, resulting in 36 samples (20 CF, 16 ZF) for library preparation. Libraries were constructed using the VAHTS Universal Plus DNA Library Prep Kit (Illumina, San Diego, CA, USA) and sequenced (2 \u0026times; 150 bp) on an Illumina NovaSeq X platform following standard protocols.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eSCFA concentration measurement\u003c/h3\u003e\n\u003cp\u003eCecal samples from the 36 ducks used for metagenome sequencing were also analyzed for the concentration of ten SCFAs: acetic acid, propionic acid, isobutyric acid, butyric acid, isovaleric acid, valeric acid, hexanoic acid, heptanoic acid, nonanoic acid, and decanoic acid. Briefly, samples were thawed on ice. Approximately 0.5 g of cecal content was homogenized in 1 mL of deionized water for 4 minutes at 40 Hz, followed by ultrasonic oscillation for 15 minutes on ice. The homogenate was centrifuged at 5,000 rpm for 20 minutes at 4\u0026deg;C. A 0.8 mL aliquot of the supernatant was mixed with 0.1 mL of 50% (v/v) sulfuric acid (H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e) and 0.8 mL of extraction solution containing 25 mg/L internal standard (methyl tert-butyl ether, MTBE). The mixture was vortexed vigorously for 10 minutes and centrifuged at 10,000 rpm for 15 minutes at 4\u0026deg;C. The resulting supernatant was analyzed using a SHIMADZU GC2030-QP2020 NX gas chromatography-mass spectrometer (GC-MS). Absolute SCFA concentrations were determined using external calibration curves generated from standard solutions of each SCFA at varying concentrations.\u003c/p\u003e\n\u003ch3\u003eDetermination and analysis of plasma metabolome\u003c/h3\u003e\n\u003cp\u003ePlasma metabolites were analyzed using ultra-high-performance liquid chromatography (UHPLC) coupled with tandem mass spectrometry (MS/MS). Chromatographic separation was performed on an ACQUITY UPLC\u003csup\u003e\u0026reg;\u003c/sup\u003e HSS T3 column (100 mm \u0026times; 2.1 mm, 1.8 \u0026micro;m average particle size, Waters Corporation) using 0.1% (v/v) formic acid aqueous solution and acetonitrile as mobile phases. To mitigate analytical bias from instrumental variation, a pooled quality control (QC) sample was created by combining equal aliquots of all sample supernatants. This QC sample was injected at regular intervals throughout the analytical sequence. Raw MS data were acquired using MassLynx\u003csup\u003e\u0026trade;\u003c/sup\u003e V4.2 software (Waters Corporation). Data processing\u0026mdash;including peak detection, alignment, and integration\u0026mdash;was performed in Progenesis QI (Nonlinear Dynamics). Metabolite annotation was conducted against the METLIN database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://metlin.scripps.edu\u003c/span\u003e\u003cspan address=\"https://metlin.scripps.edu\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) within the Progenesis QI platform, using a mass accuracy threshold of \u0026le;\u0026thinsp;10 ppm and MS/MS spectral matching.\u003c/p\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003ePreparation of libraries for RNA-sequencing and data analysis\u003c/h2\u003e\u003cp\u003eLiver tissues were collected from three CF and three ZF ducks immediately post-euthanasia, snap-frozen in liquid nitrogen, and stored at -80\u0026deg;C. Total RNA was extracted using TRIzol\u003csup\u003e\u0026trade;\u003c/sup\u003e Reagent (Thermo Fisher Scientific, Waltham, MA) following manufacturer protocols. RNA integrity was assessed by UV spectrophotometry (NanoDrop\u003csup\u003e\u0026trade;\u003c/sup\u003e 2000; Thermo Fisher Scientific) and Electropherogram analysis (Agilent 2100 Bioanalyzer; Agilent Technologies, Santa Clara, CA). Ribosomal RNA was depleted using the Illumina TruSeq Stranded Total RNA Library Prep Kit with Ribo-Zero Gold. Resulting mRNA was fragmented and reverse-transcribed into cDNA libraries using the same kit. Libraries were sequenced (2 \u0026times; 150 bp) on an Illumina HiSeq X platform.\u003c/p\u003e\u003cp\u003eRaw reads were quality-trimmed and adapter-filtered using fastp (v0.20.0) [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Processed reads were mapped to the Pekin duck reference genome (ZJU1.0) using STAR (version 2.7.9a) [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e] with default parameters. Gene-level counts were generated with STAR\u0026rsquo;s --quantMode GeneCounts option and analyzed for differential expression using DESeq2 (v1.26.0) [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e] in R/Bioconductor, with normalization via the median-of-ratios method.\u003c/p\u003e\u003cp\u003e\u003cb\u003e16S amplicon sequencing data processing\u003c/b\u003e\u003c/p\u003e\u003cp\u003eRaw sequencing reads underwent quality assessment using FastQC. Adapters and low-quality bases were trimmed using fastp (v0.20.0). Processed reads were denoised, error-corrected, and filtered for chimeras using the DADA2 plugin (v1.18) [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e] in QIIME2 (v2020.6) [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], generating amplicon sequence variants (ASVs). ASVs were filtered to retain those with relative abundance\u0026thinsp;\u0026gt;\u0026thinsp;0.0001% (10⁻⁶) and presence in \u0026ge;\u0026thinsp;2 samples. Taxonomy was assigned against the Genome Taxonomy Database (GTDB r207) using DADA2\u0026rsquo;s native Bayesian classifier with default parameters. Alpha diversity (Shannon, Chao1, Pielou, and Richness) metrics were calculated in QIIME2. Phyloseq (v1.34.0) [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] was used for evaluation of sequencing depth and preparation of final ASV tables for downstream statistical analysis.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eMetagenome-assembled genomes (MAGs) Recovery and Annotation\u003c/h2\u003e\u003cp\u003eRaw metagenomic reads were preprocessed using fastp (v0.20.0) for adapter trimming and quality filtering. Host-derived reads were removed by alignment to the duck reference genome (ZJU1.0) using Bowtie2 (v2.4.2) [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. \u003cem\u003eDe novo\u003c/em\u003e assembly of filtered reads was performed per sample using metaSPAdes (v3.15.0) [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] with default parameters. Assembled contigs were subsequently binned into MAGs using an integrative approach with MetaBAT2 (v2.17) [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e], MaxBin2 (v2.2.7) [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e], and CONCOCT (v1.1.0) [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Consensus bins were generated and refined using DAS Tool (v1.1.3) [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. The quality of the refined MAGs was assessed using CheckM2 (v1.0.1) [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e] with default parameters. Taxonomic classification was performed against GTDB r207 using GTDB-Tk (v2.1.0) [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Community-level functional annotations were performed using HUMAnN 3.0 (v3.6) pipeline with the clean reads as input [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Species-level differential abundance between FE groups was determined using ANCOM-BC [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Open reading frames (ORFs) were predicted from all contigs using Prodigal (v2.6.3) [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e] in metagenomic mode. Proteins were annotated using eggNOG database with eggNOG-mapper v2 [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. A non-redundant gene catalog was constructed with CD-HIT (v4.8.1) [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e] at 95% nucleotide identity and 90% coverage. Gene abundances were quantified by aligning reads to the non-redundant gene catalog using Salmon (v1.10.2) [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e] in alignment-based mode. Differential abundance of KEGG Orthologs (KOs) was tested using MaAsLin2 (v1.12.0) [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e] with default parameters.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003eViral contig identification and annotation\u003c/h2\u003e\u003cp\u003ePutative viral contigs were first filtered from assembled metagenomic contigs using geNomad (v1.8.0) [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e] with a relaxed threshold of \u0026ndash;min-score\u0026thinsp;=\u0026thinsp;0.75 to increase detection sensitivity. The putative viral contigs were annotated using our previously developed ViroProfiler pipeline [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e] for precise viral contig identification and annotation. Briefly, The ViroProfiler pipeline utilizes CheckV (v0.9.0) [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e] to remove host region in provirus contigs, and then use VirSorter2 (v2.2.4) [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e] and VIBRANT (v1.2.1) [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e] to identify viral contigs. Auxiliary metabolic genes (AMGs) were annotated using VIBRANT and DRAM-v (v1.4.4) [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. Viral hosts were predicted using iPHoP (v1.2.0) with confidence score\u0026thinsp;\u0026ge;\u0026thinsp;90 [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. Taxonomic classification was performed using a combination of geNomad and VITAP (v1.7.1) [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. VITAP taxonomy was preferred due to its more specific annotations, often to genus or species level with high precision, while geNomad taxonomy was used when a contig was not annotated by VITAP. Viruses were classified as temperate if they are identified as provirus by geNomad, VIBRANT, or CheckV (v0.9.0), or classified as temperate viruses by BACPHLIP (v0.9.6) [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. Viral contigs were then dereplicated to generate a non-redundant catalog of viral sequences. Specifically, pairwise average nucleotide identity (ANI) was calculated among these contigs using the clustering algorithm within CheckV, applying thresholds of \u0026ge;\u0026thinsp;95% ANI and \u0026ge;\u0026thinsp;85% alignment fraction (AF) to define species-level viral operational taxonomic units (vOTUs). The longest contig within each vOTU cluster was selected as its representative sequence. To further investigate relationships among the identified vOTUs, compare them with known viral genomes, and identify potential novel viral clusters (VCs), gene-sharing network analysis was conducted using vConTACT2 (v.0.11.3) [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. VCs identified by vConTACT2 represent groups of related viral genomes, offer insights into potential genus-level relationships, and highlight vOTUs that may constitute novel taxa. Abundance of viruses were estimated by mapping clean reads to vOTUs using Bowtie2, and then quantified using CoverM (v0.7.0) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/wwood/CoverM\u003c/span\u003e\u003cspan address=\"https://github.com/wwood/CoverM\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003eStatistical Analyses\u003c/h2\u003e\u003cp\u003eAll statistical analyses were performed in R (v4.1.0). Data manipulation and visualization utilized the tidyverse ecosystem (v1.3.0). Student\u0026rsquo;s t-test and Wilcoxon rank sum test were employed for group comparisons. For transcriptome data, differentially expressed genes (DEGs) were identified using DESeq2 with thresholds of adjusted \u003cem\u003ep\u003c/em\u003e-value (Benjamini-Hochberg)\u0026thinsp;\u0026lt;\u0026thinsp;0.05 and |log\u003csub\u003e2\u003c/sub\u003eFC| \u0026ge; 0.585. KEGG pathway enrichment of DEGs was performed using ClusterProfiler (v4.0) [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. For metabolomics data, peak areas were normalized via total sum scaling, differential metabolites between two groups were identified with threshold of |log\u003csub\u003e2\u003c/sub\u003eFC| \u0026ge; 1 and \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05. Metabolic pathway enrichment was analyzed in MetaboAnalyst 5.0 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.metaboanalyst.ca\u003c/span\u003e\u003cspan address=\"https://www.metaboanalyst.ca\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eEthics approval\u003c/h2\u003e\n\u003cp\u003eAll animal procedures involving Pekin ducks (Anas platyrhynchos domesticus) were approved by the Animal Welfare and Ethics Committee of the Institute of Animal Sciences (IAS), the Chinese Academy of Agricultural Sciences (IAS20160401, CAAS, Beijing, China).\u003c/p\u003e\n\u003ch2\u003eData availability\u003c/h2\u003e\n\u003cp\u003eThe raw data for multi-omics cohort are deposited in the National Genomics Data Center website (https://ngdc.cncb.ac.cn/). All the sequencing data are accessible with the identifier CRA026543 (amplicon), CRA028005 (metagenome), and CRA027933 (transcriptome). The metabolomics data are accessible with the identifier OMIX011117. The code used for data analysis is available at the GitHub repository https://github.com/rujinlong/duckbiome.\u003c/p\u003e\n\u003ch2\u003eCompeting interests\u003c/h2\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003ch2\u003eFunding\u003c/h2\u003e\n\u003cp\u003eThis work was supported by National Natural Science Foundation of China (#32341055) awarded to YSZ, China Agriculture Research System of MOF and MARA (CARS-42-2) awarded to XW, the Fundamental Research Funds for the Central Universities (grant no. 226-2025-00030) awarded to FYL, the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation - Emmy Noether program, Project No. 273124240; and SPP2330, Project No. 464797012), and the European Research Council Starting grant (ERC StG 803077) awarded to LD.\u003c/p\u003e\n\u003ch2\u003eAuthors\u0026apos; contributions\u003c/h2\u003e\n\u003cp\u003eAll authors contributed intellectually to and agreed to this submission. XW, YSZ, and JLR designed the experiments. SJT, ZZW, XQY, SFL, JW, WZ, SQW, and YSZ conducted the experiments and collected the data. SJT, LY, and JLR analyzed and interpreted the data. SJT, PYL, DH, and YSZ prepared figures. XW interpreted the data and wrote the manuscript, while JLR, FYL, LD and SSH provided substantial feedback. JLR, XW, and PYL contributed to the conceptual design of the study. XW, YSZ, FYL, LD and SSH provided funding support. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003ch2\u003eAcknowledgements\u003c/h2\u003e\n\u003cp\u003eThe authors sincerely thank Prof. Guang Yao from the University of Arizona and Prof. Xujun Liang from Northwest A\u0026amp;F University for their constructive discussions and valuable suggestions. Additionally, we acknowledge the Technical University of Munich (TUM), the Helmholtz Zentrum M\u0026uuml;nchen (HMGU), and Northwest A\u0026amp;F University (NWAFU) for providing high-performance computing resources.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eZampiga, M., et al., \u003cem\u003eApplication of omics technologies for a deeper insight into quali-quantitative production traits in broiler chickens: A review\u003c/em\u003e. J Anim Sci Biotechnol, 2018. 9: p. 61.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWrzosek, L., et al., \u003cem\u003eBacteroides thetaiotaomicron and Faecalibacterium prausnitzii influence the production of mucus glycans and the development of goblet cells in the colonic epithelium of a gnotobiotic model rodent\u003c/em\u003e. BMC Biol, 2013. 11: p. 61.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHan, H., et al., \u003cem\u003eFrom gut microbiota to host appetite: gut microbiota-derived metabolites as key regulators\u003c/em\u003e. Microbiome, 2021. 9(1): p. 162.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIkeda, T., et al., \u003cem\u003eShort-chain fatty acid receptors and gut microbiota as therapeutic targets in metabolic, immune, and neurological diseases\u003c/em\u003e. Pharmacol Ther, 2022. 239: p. 108273.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGao, P., et al., \u003cem\u003eFeed-additive probiotics accelerate yet antibiotics delay intestinal microbiota maturation in broiler chicken\u003c/em\u003e. Microbiome, 2017. 5(1): p. 91.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYang, J., et al., \u003cem\u003eMicrobiota-accessible fiber activates short-chain fatty acid and bile acid metabolism to improve intestinal mucus barrier in broiler chickens\u003c/em\u003e. Microbiol Spectr, 2024. 12(1): p. e0206523.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYang, C., et al., \u003cem\u003eMetagenomic insights into the relationship between intestinal flora and residual feed intake of meat ducks\u003c/em\u003e. Poult Sci, 2024. 103(7): p. 103836.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHe, Z., et al., \u003cem\u003eCombined effect of microbially derived cecal SCFA and host genetics on feed efficiency in broiler chickens\u003c/em\u003e. Microbiome, 2023. 11(1): p. 198.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhou, G.C., et al., \u003cem\u003eMulti-omics revealed the mechanism of feed efficiency in sheep by the combined action of the host and rumen microbiota\u003c/em\u003e. Animal Nutrition, 2024. 18: p. 367\u0026ndash;379.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHoyles, L., et al., \u003cem\u003eCharacterization of virus-like particles associated with the human faecal and caecal microbiota\u003c/em\u003e. Res Microbiol, 2014. 165(10): p. 803\u0026ndash;812.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLepage, P., et al., \u003cem\u003eDysbiosis in inflammatory bowel disease: a role for bacteriophages?\u003c/em\u003e Gut, 2008. 57(3): p. 424\u0026ndash;425.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYan, M., et al., \u003cem\u003eInterrogating the viral dark matter of the rumen ecosystem with a global virome database\u003c/em\u003e. Nat Commun, 2023. 14(1): p. 5254.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCampbell, D.E., et al., \u003cem\u003eInfection with Bacteroides Phage BV01 Alters the Host Transcriptome and Bile Acid Metabolism in a Common Human Gut Microbe\u003c/em\u003e. Cell Rep, 2020. 32(11): p. 108142.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eThiele Orberg, E., et al., \u003cem\u003eBacteria and bacteriophage consortia are associated with protective intestinal metabolites in patients receiving stem cell transplantation\u003c/em\u003e. Nat Cancer, 2024. 5(1): p. 187\u0026ndash;208.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMangalea, M.R., et al., \u003cem\u003eIndividuals at risk for rheumatoid arthritis harbor differential intestinal bacteriophage communities with distinct metabolic potential\u003c/em\u003e. Cell Host Microbe, 2021. 29(5): p. 726\u0026ndash;739.e5.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu, X., et al., \u003cem\u003eRumen DNA virome and its relationship with feed efficiency in dairy cows\u003c/em\u003e. Microbiome, 2025. 13(1): p. 14.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang, Y., et al., \u003cem\u003eGenetic parameters for residual feed intake in a random population of Pekin duck\u003c/em\u003e. Asian-Australas J Anim Sci, 2017. 30(2): p. 167\u0026ndash;170.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen, S., et al., \u003cem\u003efastp: an ultra-fast all-in-one FASTQ preprocessor\u003c/em\u003e. Bioinformatics, 2018. 34(17): p. i884-i890.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDobin, A., et al., \u003cem\u003eSTAR: ultrafast universal RNA-seq aligner\u003c/em\u003e. Bioinformatics, 2013. 29(1): p. 15\u0026ndash;21.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLove, M.I., W. Huber, and S. Anders, \u003cem\u003eModerated estimation of fold change and dispersion for RNA-seq data with DESeq2\u003c/em\u003e. Genome Biol, 2014. 15(12): p. 550.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCallahan, B.J., et al., \u003cem\u003eDADA2: High-resolution sample inference from Illumina amplicon data\u003c/em\u003e. Nat Methods, 2016. 13(7): p. 581\u0026ndash;583.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBolyen, E., et al., \u003cem\u003eReproducible, interactive, scalable and extensible microbiome data science using QIIME 2\u003c/em\u003e. Nat Biotechnol, 2019. 37(8): p. 852\u0026ndash;857.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcMurdie, P.J. and S. Holmes, \u003cem\u003ephyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data\u003c/em\u003e. PLoS One, 2013. 8(4): p. e61217.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLangmead, B. and S.L. Salzberg, \u003cem\u003eFast gapped-read alignment with Bowtie 2\u003c/em\u003e. Nat Methods, 2012. 9(4): p. 357\u0026ndash;359.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNurk, S., et al., \u003cem\u003emetaSPAdes: a new versatile metagenomic assembler\u003c/em\u003e. Genome Res, 2017. 27(5): p. 824\u0026ndash;834.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKang, D.D., et al., \u003cem\u003eMetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies\u003c/em\u003e. PeerJ, 2019. 7: p. e7359.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWu, Y.W., B.A. Simmons, and S.W. Singer, \u003cem\u003eMaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets\u003c/em\u003e. Bioinformatics, 2016. 32(4): p. 605\u0026ndash;607.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAlneberg, J., et al., \u003cem\u003eBinning metagenomic contigs by coverage and composition\u003c/em\u003e. Nat Methods, 2014. 11(11): p. 1144\u0026ndash;1146.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSieber, C.M.K., et al., \u003cem\u003eRecovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy\u003c/em\u003e. Nat Microbiol, 2018. 3(7): p. 836\u0026ndash;843.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChklovski, A., et al., \u003cem\u003eCheckM2: a rapid, scalable and accurate tool for assessing microbial genome quality using machine learning\u003c/em\u003e. Nat Methods, 2023. 20(8): p. 1203\u0026ndash;1212.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChaumeil, P.A., et al., \u003cem\u003eGTDB-Tk: a toolkit to classify genomes with the Genome Taxonomy Database\u003c/em\u003e. Bioinformatics, 2019. 36(6): p. 1925\u0026ndash;1927.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBeghini, F., et al., \u003cem\u003eIntegrating taxonomic, functional, and strain-level profiling of diverse microbial communities with bioBakery 3\u003c/em\u003e. Elife, 2021. 10: p. e65088.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLin, H. and S.D. Peddada, \u003cem\u003eAnalysis of compositions of microbiomes with bias correction\u003c/em\u003e. Nat Commun, 2020. 11(1): p. 3514.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHyatt, D., et al., \u003cem\u003eProdigal: prokaryotic gene recognition and translation initiation site identification\u003c/em\u003e. BMC Bioinformatics, 2010. 11: p. 119.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCantalapiedra, C.P., et al., \u003cem\u003eeggNOG-mapper v2: Functional Annotation, Orthology Assignments, and Domain Prediction at the Metagenomic Scale\u003c/em\u003e. Mol Biol Evol, 2021. 38(12): p. 5825\u0026ndash;5829.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, W. and A. Godzik, \u003cem\u003eCd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences\u003c/em\u003e. Bioinformatics, 2006. 22(13): p. 1658\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePatro, R., et al., \u003cem\u003eSalmon provides fast and bias-aware quantification of transcript expression\u003c/em\u003e. Nat Methods, 2017. 14(4): p. 417\u0026ndash;419.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMallick, H., et al., \u003cem\u003eMultivariable association discovery in population-scale meta-omics studies\u003c/em\u003e. PLoS Comput Biol, 2021. 17(11): p. e1009442.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCamargo, A.P., et al., \u003cem\u003eIdentification of mobile genetic elements with geNomad\u003c/em\u003e. Nat Biotechnol, 2024. 42(8): p. 1303\u0026ndash;1312.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRu, J., et al., \u003cem\u003eViroProfiler: a containerized bioinformatics pipeline for viral metagenomic data analysis\u003c/em\u003e. Gut Microbes, 2023. 15(1): p. 2192522.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNayfach, S., et al., \u003cem\u003eCheckV assesses the quality and completeness of metagenome-assembled viral genomes\u003c/em\u003e. Nat Biotechnol, 2021. 39(5): p. 578\u0026ndash;585.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGuo, J., et al., \u003cem\u003eVirSorter2: a multi-classifier, expert-guided approach to detect diverse DNA and RNA viruses\u003c/em\u003e. Microbiome, 2021. 9(1): p. 37.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKieft, K., Z. Zhou, and K. Anantharaman, \u003cem\u003eVIBRANT: automated recovery, annotation and curation of microbial viruses, and evaluation of viral community function from genomic sequences\u003c/em\u003e. Microbiome, 2020. 8(1): p. 90.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShaffer, M., et al., \u003cem\u003eDRAM for distilling microbial metabolism to automate the curation of microbiome function\u003c/em\u003e. Nucleic Acids Res, 2020. 48(16): p. 8883\u0026ndash;8900.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRoux, S., et al., \u003cem\u003eiPHoP: An integrated machine learning framework to maximize host prediction for metagenome-derived viruses of archaea and bacteria\u003c/em\u003e. PLoS Biol, 2023. 21(4): p. e3002083.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZheng, K., et al., \u003cem\u003eVITAP: a high precision tool for DNA and RNA viral classification based on meta-omic data\u003c/em\u003e. Nat Commun, 2025. 16(1): p. 2226.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHockenberry, A.J. and C.O. Wilke, \u003cem\u003eBACPHLIP: predicting bacteriophage lifestyle from conserved protein domains\u003c/em\u003e. PeerJ, 2021. 9: p. e11396.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBin Jang, H., et al., \u003cem\u003eTaxonomic assignment of uncultivated prokaryotic virus genomes is enabled by gene-sharing networks\u003c/em\u003e. Nat Biotechnol, 2019. 37(6): p. 632\u0026ndash;639.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWu, T., et al., \u003cem\u003eclusterProfiler 4.0: A universal enrichment tool for interpreting omics data\u003c/em\u003e. Innovation (Camb), 2021. 2(3): p. 100141.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePeschel, S., et al., \u003cem\u003eNetCoMi: network construction and comparison for microbiome data in R\u003c/em\u003e. Brief Bioinform, 2021. 22(4): p. bbaa290.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQin, J., et al., \u003cem\u003eA metagenome-wide association study of gut microbiota in type 2 diabetes\u003c/em\u003e. Nature, 2012. 490(7418): p. 55\u0026ndash;60.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLevi Mortera, S., et al., \u003cem\u003eFunctional and Taxonomic Traits of the Gut Microbiota in Type 1 Diabetes Children at the Onset: A Metaproteomic Study\u003c/em\u003e. Int J Mol Sci, 2022. 23(24): p. 15982.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang, S., et al., \u003cem\u003eDose-related effects of early-life intake of sn-2 palmitate, a specific positionally distributed human milk fatty acid, on the composition and metabolism of the intestinal microbiota\u003c/em\u003e. J Dairy Sci, 2023. 106(12): p. 8272\u0026ndash;8286.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHao, L., et al., \u003cem\u003eNovel syntrophic bacteria in full-scale anaerobic digesters revealed by genome-centric metatranscriptomics\u003c/em\u003e. ISME J, 2020. 14(4): p. 906\u0026ndash;918.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHampton, H.G., B.N.J. Watson, and P.C. Fineran, \u003cem\u003eThe arms race between bacteria and their phage foes\u003c/em\u003e. Nature, 2020. 577(7790): p. 327\u0026ndash;336.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMurphy, J., et al., \u003cem\u003eBacteriophage orphan DNA methyltransferases: insights from their bacterial origin, function, and occurrence\u003c/em\u003e. Appl Environ Microbiol, 2013. 79(24): p. 7547\u0026ndash;7555.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDeramchia, K., et al., \u003cem\u003eContribution of pyruvate phosphate dikinase in the maintenance of the glycosomal ATP/ADP balance in the Trypanosoma brucei procyclic form\u003c/em\u003e. J Biol Chem, 2014. 289(25): p. 17365\u0026ndash;17378.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFeng, W.H., et al., \u003cem\u003eLytic induction therapy for Epstein-Barr virus-positive B-cell lymphomas\u003c/em\u003e. J Virol, 2004. 78(4): p. 1893\u0026ndash;1902.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWestphal, E.M., et al., \u003cem\u003eActivation of lytic Epstein-Barr virus (EBV) infection by radiation and sodium butyrate in vitro and in vivo: a potential method for treating EBV-positive malignancies\u003c/em\u003e. Cancer Res, 2000. 60(20): p. 5781\u0026ndash;5788.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003evan Kempen, M., et al., \u003cem\u003eFast and accurate protein structure search with Foldseek\u003c/em\u003e. Nat Biotechnol, 2024. 42(2): p. 243\u0026ndash;246.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWen, C., et al., \u003cem\u003eJoint contributions of the gut microbiota and host genetics to feed efficiency in chickens\u003c/em\u003e. Microbiome, 2021. 9(1): p. 126.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu, Y., \u003cem\u003eFatty acids, inflammation and intestinal health in pigs\u003c/em\u003e. J Anim Sci Biotechnol, 2015. 6(1): p. 41.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJi, Z.Q., et al., \u003cem\u003eEffects of coated sodium butyrate on growth performance, intestinal morphology, and caecal short-chain fatty acids of growing Pekin ducks\u003c/em\u003e. Br Poult Sci, 2023. 64(4): p. 529\u0026ndash;533.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYang, Q., et al., \u003cem\u003eButyrate in combination with forskolin alleviates necrotic enteritis, increases feed efficiency, and improves carcass composition of broilers\u003c/em\u003e. J Anim Sci Biotechnol, 2022. 13(1): p. 3.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu, S., et al., \u003cem\u003eEcological stability of microbial communities in Lake Donghu regulated by keystone taxa\u003c/em\u003e. Ecological Indicators, 2022. 136: p. 108695.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMaslov, S. and K. Sneppen, \u003cem\u003eSpecificity and stability in topology of protein networks\u003c/em\u003e. Science, 2002. 296(5569): p. 910\u0026ndash;913.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang, Y., et al., \u003cem\u003eEffect of host genetics and gut microbiome on fat deposition traits in pigs\u003c/em\u003e. Front Microbiol, 2022. 13: p. 925200.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBergamaschi, M., et al., \u003cem\u003eGut microbiome composition differences among breeds impact feed efficiency in swine\u003c/em\u003e. Microbiome, 2020. 8(1): p. 110.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYoon, J.H., et al., \u003cem\u003eInclusion of Lacticaseibacillus paracasei NSMJ15 in broiler diets induces changes in jejunal immune cell population and cecal microbiota\u003c/em\u003e. Anim Biosci, 2024. 37(12): p. 2155\u0026ndash;2166.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhong, Z.P., et al., \u003cem\u003eViral potential to modulate microbial methane metabolism varies by habitat\u003c/em\u003e. Nat Commun, 2024. 15(1): p. 1857.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eObeng, N., A.A. Pratama, and J.D.V. Elsas, \u003cem\u003eThe Significance of Mutualistic Phages for Bacterial Ecology and Evolution\u003c/em\u003e. Trends Microbiol, 2016. 24(6): p. 440\u0026ndash;449.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLuo, X.Q., et al., \u003cem\u003eViral community-wide auxiliary metabolic genes differ by lifestyles, habitats, and hosts\u003c/em\u003e. Microbiome, 2022. 10(1): p. 190.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFeng, T., et al., \u003cem\u003eMetabolite Profile of Sheep Serum With High or Low Average Daily Gain\u003c/em\u003e. Front Vet Sci, 2021. 8: p. 662536.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang, B., et al., \u003cem\u003eImpact of feeding regimens on the composition of gut microbiota and metabolite profiles of plasma and feces from Mongolian sheep\u003c/em\u003e. J Microbiol, 2020. 58(6): p. 472\u0026ndash;482.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, S., et al., \u003cem\u003eLongitudinal changes of serum metabolomic profile after laparoscopic sleeve gastrectomy in obesity\u003c/em\u003e. Endocr Connect, 2024. 13(11): p. e240292.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHan, Z., et al., \u003cem\u003eFecal microbiota transplantation accelerates restoration of florfenicol-disturbed intestinal microbiota in a fish model\u003c/em\u003e. Commun Biol, 2024. 7(1): p. 1006.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, Z., et al., \u003cem\u003eTraditional Tibetan Medicine Twenty-Five Wei'er Tea Pills Ameliorate Rheumatoid Arthritis Based on Chemical Crosstalk Between Gut Microbiota and the Host\u003c/em\u003e. Front Pharmacol, 2022. 13: p. 828920.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZheng, X., et al., \u003cem\u003eIntegrated Metagenomic and Metabolomics Profiling Reveals Key Gut Microbiota and Metabolites Associated with Weaning Stress in Piglets\u003c/em\u003e. Genes (Basel), 2024. 15(8).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHuang, Z., et al., \u003cem\u003eGut microbiota and blood metabolites: unveiling their roles in hippocampal volume changes through Mendelian randomization and mediation analysis\u003c/em\u003e. Metab Brain Dis, 2025. 40(4): p. 178.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYan, J., et al., \u003cem\u003eFecal microbiota transplantation significantly improved respiratory failure of amyotrophic lateral sclerosis\u003c/em\u003e. Gut Microbes, 2024. 16(1): p. 2353396.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"microbiome","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mbio","sideBox":"Learn more about [Microbiome](http://microbiomejournal.biomedcentral.com/)","snPcode":"40168","submissionUrl":"https://submission.nature.com/new-submission/40168/3","title":"Microbiome","twitterHandle":"@MicrobiomeJ","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7357899/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7357899/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground:\u003c/h2\u003e\u003cp\u003eThe gut microbiota critically influences poultry health, nutrition, feed efficiency (FE), and overall productivity. However, the relationship between gut microbes, including bacteria and bacteriophages, and FE in ducks remains underexplored. To address this, we integrated cecal 16S amplicon, metagenome, microbiota derived short-chain fatty acids (SCFAs) profiling, liver transcriptome, and serum metabolome data to illustrate the contribution of gut microbiome (bacteria and viruses) to duck FE.\u003c/p\u003e\u003ch2\u003eResults:\u003c/h2\u003e\u003cp\u003eWe reconstructed viral genomes and prokaryotic metagenome-assembled genomes (MAGs), annotated their genes using comprehensive databases. Prokaryotic hosts of viruses were also predicted to understand virus-host dynamics within the gut ecosystem. Our results revealed that high-FE ducks have higher concentration of propionate and butyrate in cecum compared with low-FE ducks. The metagenome sequencing revealed distinct cecal microbiota profiles between two groups, with increased relative abundance of representative SCFA producers, especially \u003cem\u003eParaprevotella sp905215575\u003c/em\u003e and \u003cem\u003eBacteroides sp944322345\u003c/em\u003e, and enhanced SCFA-biosynthesis pathways in high-FE ducks. Virome genome assembly identified two bacteriophages encoding auxiliary metabolic genes (AMGs) involved in pyruvate metabolism, enhancing nutrient availability for host bacteria to produce SCFAs (e.g., temperate phage-encoded pyruvate phosphate dikinase) or exploiting host central metabolic pathways for viral replication (e.g., lytic phage-encoded formate C-acetyltransferase). Furthermore, these representative SCFA-producing bacteria and bacteriophage consortia were associated with serum metabolites (including L-histidine and 4-hydroxydecanedioylcarnitine) linked to duck FE.\u003c/p\u003e\u003ch2\u003eConclusion:\u003c/h2\u003e\u003cp\u003eCollectively, these findings provide novel insights into the gut microbial factors regulating FE in ducks, offering potential strategies to optimize poultry nutrition and productivity.\u003c/p\u003e","manuscriptTitle":"Bacteria and bacteriophage consortia modulate cecal SCFA production and host metabolism to enhance feed efficiency in ducks","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-29 07:35:16","doi":"10.21203/rs.3.rs-7357899/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-09-10T13:34:30+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-09-10T11:51:12+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-08-13T12:08:10+00:00","index":"","fulltext":""},{"type":"submitted","content":"Microbiome","date":"2025-08-12T16:28:07+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"microbiome","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mbio","sideBox":"Learn more about [Microbiome](http://microbiomejournal.biomedcentral.com/)","snPcode":"40168","submissionUrl":"https://submission.nature.com/new-submission/40168/3","title":"Microbiome","twitterHandle":"@MicrobiomeJ","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"9c53e60f-1e08-4b4d-a109-47115ee39663","owner":[],"postedDate":"August 29th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-03-09T16:09:51+00:00","versionOfRecord":{"articleIdentity":"rs-7357899","link":"https://doi.org/10.1186/s40168-026-02368-y","journal":{"identity":"microbiome","isVorOnly":false,"title":"Microbiome"},"publishedOn":"2026-03-04 15:59:53","publishedOnDateReadable":"March 4th, 2026"},"versionCreatedAt":"2025-08-29 07:35:16","video":"","vorDoi":"10.1186/s40168-026-02368-y","vorDoiUrl":"https://doi.org/10.1186/s40168-026-02368-y","workflowStages":[]},"version":"v1","identity":"rs-7357899","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7357899","identity":"rs-7357899","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}