Genomic and metagenomic approaches decode host-microbiome-environment interactions by mapping microbial community dynamics under health, disease, or stress. These methods identify keystone taxa (e.g., opportunistic pathogens like Aeromonas) and functional shifts linked to dysbiosis, enabling targeted interventions. High-throughput sequencing (e.g., 16S rRNA gene profiling, shotgun metagenomics) enhances resolution of microbial profiles across host health and environmental gradients. These approaches identify key microbial players such as opportunistic pathogens (e.g., Aeromonas, Vibrio) and beneficial taxa (e.g., Cetobacterium, Weissella), with abundance shifts strongly correlating to host status. Critically, functionality is strain-specific: pathogenic potential varies within Aeromonas/Vibrio, and probiotic properties are not universal in Cetobacterium/Weissella (Wang T. et al., 2024; Li et al., 2025). Metagenomics facilitates pathogen discovery and dysbiosis-disease linkages, exemplified by metabarcoding identifying a novel Flavobacterium species causing peracute skin disease in rainbow trout, distinct from classical columnaris strains (Zamparo et al., 2025). Viral profiling through virome analyses detects pathogens like white spot syndrome virus in environmental reservoirs and novel caliciviruses linked to mass fish mortality (Mercer et al., 2024; Su et al., 2024). Advanced techniques such as Oxford Nanopore sequencing enhance resolution in complex matrices like fish mucus (Domingo-Bretón et al., 2024). However, these DNA-centred tools have well-recognized constraints: short-read assemblies can mask low-abundance taxa and hamper strain-level resolution; draft metagenomes rely on gene annotations that are predictive rather than experimental; and high host DNA backgrounds can dilute microbial signals, particularly in gut, gill, or skin biopsies. Stable isotope probing (DNA-SIP/RNA-SIP) and long-read sequencing are now being combined with metagenomics to assign functional genes to active taxa and partially alleviate these blind spots (Alcolombri et al., 2022; Han Y. et al., 2024). Spatial heterogeneity is critical, with distinct microbiomes inhabiting mucosal surfaces (gill, skin, gut, ovary). For instance, the gut of olive flounder harbors more antibiotic resistance genes and Vibrionaceae than functionally diverse gill/skin communities, underscoring the need for site-specific probiotics (Yu et al., 2025).

Metabolomics reveals how microbial metabolites (e.g., SCFAs) mediate host-microbe crosstalk, influencing immune pathways and stress resilience. This mechanistic insight identifies therapeutic targets for precision engineering. For example, butyrate and other SCFAs serve dual roles as enterocyte energy sources and immunomodulators, acting via histone deacetylase (HDAC) inhibition or G-protein-coupled receptor signaling. Dietary interventions, such as supplementation with Clostridium butyricum in shrimp, elevate beneficial metabolites and enhance mucosal immunity and pathogen resistance (Li S. et al., 2022; Liao et al., 2023). Yet metabolite profiles alone seldom reveal which organism produced a given compound. Emerging compound-specific stable isotope labeling (e.g., ¹³C or ¹⁵N SIP) tracked by high-resolution MS, as well as spatial metabolomics coupled with fluorescence in situ hybridization (FISH-SIMS), now help connect metabolite fluxes to specific microbial producers (Uengwetwanit et al., 2020; Alcolombri et al., 2022). Beyond SCFAs, metabolomics detects broader shifts in bile acid and amino acid metabolism that contribute to immune resilience. Critically, metabolomics reveals pollutant-induced dysbiosis, where microplastics, pesticides (e.g., deltamethrin), and polychlorinated biphenyls disrupt gut-liver axes. These pollutants alter lipid metabolites like lysophosphatidylcholines, trigger oxidative stress, and dysregulate signaling pathways (e.g., PPAR, apoptosis genes), detectable through integrated metabolomic analyses (Song et al., 2025; Zhong et al., 2025).

Transcriptomics and epigenetics uncover host response mechanisms to microbiome shifts, including immune gene regulation and epigenetic priming of ‘trained immunity’—even in invertebrates lacking adaptive immunity. Nevertheless, transcript counts rarely translate directly into protein function; low-expression genes may be missed, and host RNA often dominates libraries. Coupling metatranscriptomics with metaproteomics or ribosome profiling can overcome these bottlenecks by verifying actual protein synthesis (Zhao C. et al., 2023; Heyer et al., 2025). Probiotics such as Lactiplantibacillus plantarum and microbial metabolites like butyrate upregulate immune genes (e.g., proPO, lysozyme, IL-10, antimicrobial peptides) and pathways (Toll, Imd, NLRP3 inflammasome), enhancing defense mechanisms (Shan et al., 2021; Wang T. et al., 2024; Guzman et al., 2025). Under environmental stress, transcriptomics uncovers conserved response pathways: in Pacific white shrimp exposed to heat stress, it reveals energy repartitioning through glycolysis, immune modulation via C-type lectin and IL-17, and glutathione-mediated antioxidant defense (Liu et al., 2025). Similarly, ammonia stress in fish disrupts amino acid metabolism and activates apoptosis pathways (Peng et al., 2025). Epigenetic mechanisms mediate microbiome effects on immunity; butyrate, as an HDAC inhibitor, alters chromatin accessibility (e.g., enhancing IL-17D expression in tilapia for neutrophil recruitment) and influences DNA methylation. Microbiome-induced “trained immunity” occurs even in invertebrates lacking adaptive immunity through epigenetic priming (Liao et al., 2023; Wang T. et al., 2024; Guzman et al., 2025). Gnotobiotic models confirm that microbiome composition directly shapes host physiology, including immune modulation and spatial microbial distribution (Adade et al., 2023).

Integrated multi-omics bridges microbial composition, host physiology, and environmental interactions to identify biomarkers (e.g., microbial ratios signaling dysfunction) and refine precision strategies. This integration yields systems-level insights linking microbial dynamics, host physiology, and environmental interactions. It identifies robust biomarkers, such as genus-level microbial ratios (e.g., Vibrio/Photobacterium in shrimp) signaling stress-associated dysbiosis and immune disruption, and unravels complex mechanisms. For example, multi-omics links Vibrio proliferation to glutathione depletion and immune dysregulation specifically in shrimp under combined ammonia/salinity stress, and Vibrionaceae increase (with a shift in commensal genera) in seabream during parasitic infection (Li et al., 2025; Toxqui-Rodríguez et al., 2025). Because each single-omics layer is imperfect, integrated designs (e.g., genome-resolved metaproteomics, metabolite SIP linked to metagenome-assembled genomes [MAGs]) are indispensable for triangulating taxon→gene expression→metabolite relationships and thereby closing attribution gaps highlighted above (Rasmussen et al., 2022; Hansen et al., 2023). These insights directly inform precision strategies: optimizing probiotic formulations (e.g., Lactococcus lactis D1813 for specific salinity/dissolved oxygen conditions), designing targeted dietary supplements (e.g., antioxidants based on stress pathways), and developing novel therapies like antimicrobial peptides (e.g., Lvvibriocin-GK). Omics reveals that such peptides reduce pathogen loads, stimulate immunity (e.g., lysozyme activity), modulate microbiota, and suppress inflammation (Adil et al., 2025; Sun et al., 2025). Nutraceuticals like astaxanthin further mitigate tissue damage by regulating apoptosis and metabolism genes (He et al., 2025). Collectively, multi-omics underpins three synergistic pillars of microbiome engineering: identifying biomarkers to inform probiotic/synbiotic design; leveraging omics-derived targets (e.g., butyrate-producing taxa) for pathogen suppression; and using transplantation strategies (e.g., synthetic consortia) for ecological restoration. This closed-loop system—where omics guides intervention design and outcomes refine models—enables sustainable disease management in aquaculture.

A key premise in probiotic development is that strain-specific traits (e.g., spore formation, bacteriocin production) must align with host physiology and environmental conditions for efficacy. Probiotics exert their protective effects in aquaculture through multifaceted mechanisms, including the production of antimicrobial compounds (e.g., bacteriocins, organic acids, enzymes), competitive exclusion of pathogens, and immunomodulation (Figure 1; Table 1). These mechanisms are harnessed in microbiome engineering to design targeted interventions, though their efficacy depends critically on strain selection, host compatibility, and environmental stability. For instance, Bacillus subtilis subsp. inaquosorum BSXE-2102 synthesizes 12 secondary metabolites with potent antagonistic activity against aquatic pathogens (Qin et al., 2025), while Lactobacillus plantarum strains from kefir produce bacteriocins that suppress Vibrio alginolyticus (Tseng et al., 2023). Strain-specific efficacy is evident in Bacillus velezensis FiA2, which produces broad-spectrum oxydifficidin (Khan et al., 2025), and Streptomyces sp. D-6, which delivers novel bioactive metabolites (Zhao et al., 2025). Meta-analyses confirm Bacillus spp. outperform Lactobacillus in pathogen inhibition due to spore-forming resilience, whereas Streptomyces strains show unique bioremediation potential (Giri et al., 2024; Hoseinifar et al., 2024). However, limitations persist: survivability under gastrointestinal stress varies (31–75% acid tolerance at pH 2–4), adhesion capacity to the GI tract (a critical determinant of colonization success) is often strain-specific, efficacy is host-dependent, and inconsistent trial results occur across environments or host genotypes. For example, Bacillus subtilis NTU-18 improves growth in Anguilla japonica glass eels but requires precise dosing for other species (Lin et al., 2024). These constraints highlight the need for engineered solutions, such as encapsulation or host-adapted consortia, to improve reliability.

Microbiome transplantation strategies represent two distinct approaches for enhancing aquaculture sustainability. Fecal microbiota transplantation (FMT) leverages naturally evolved communities from healthy donors to restore dysbiotic hosts, functioning primarily as a restorative intervention (El-Son et al., 2025). In contrast, synthetic microbial communities (SynComs) employ rationally designed, precision-engineered consortia of defined beneficial strains to modulate host phenotypes beyond restoration (Guo et al., 2023; El-Son et al., 2025).

FMT has demonstrated efficacy in reversing antibiotic-induced dysbiosis by reintroducing keystone taxa and metabolites. Recent research confirms FMT rapidly reverses antibiotic-induced dysbiosis in fish by restoring aromatic amino acid metabolism and glutathione synthesis—critical for mucosal repair—while accelerating recovery 2.5-fold compared to natural restoration (Han Z. et al., 2024). For example, florfenicol-treated koi carp (Cyprinus carpio) exhibited reduced beneficial genera (e.g., Lactobacillus, Bifidobacterium) and mucosal damage, but FMT rapidly restored these populations and normalized critical metabolites like aromatic amino acids and glutathione, accelerating recovery compared to natural restoration (Han Z. et al., 2024). FMT also restores key metabolites, including short-chain fatty acids and lipid metabolism-related molecules, supporting a balanced metabolic profile (Xiao et al., 2020; Han Z. et al., 2024). Gut microbiota-derived metabolites like indole 3-propionic acid further demonstrate non-gut protective roles, such as radiation toxicity mitigation (Xiao et al., 2020). Similarly, FMT in large yellow croaker (Larimichthys crocea) larvae improved growth performance, digestive enzyme activity, and intestinal morphology by enhancing microbial diversity and introducing functional taxa linked to nutrient metabolism (Zhang et al., 2023). Notably, FMT can transfer phenotype-specific effects: Transplantation of microbiota from antipsychotic-exposed common carp induced behavioral abnormalities in recipient fish, confirming causal microbiota-host nervous system interactions (Chang et al., 2024). This supports evidence that FMT can transfer non-gut traits like behavior and stress tolerance in aquatic animals (Ouyang et al., 2023), and disease susceptibility/resistance between species (Ma et al., 2022). A meta-analysis revealed that higher donor strain engraftment significantly correlates with clinical success (p = 0.017) (Ianiro et al., 2022). The study found that species from the phyla Bacteroidota (e.g., genera like Bacteroides and Parabacteroides) and Actinobacteria (e.g., families like Bifidobacteriaceae and Coriobacteriaceae) demonstrated significantly higher engraftment success than most species from the Firmicutes phylum. This underscores the critical role of recruiting specific, highly engraftable taxa in achieving ecological resilience and clinical efficacy after FMT. Beyond restoration, early-life microbial interventions can induce durable and even intergenerational protection, as demonstrated in the Pacific oyster (Crassostrea gigas), where microbial exposure protected against Pacific Oyster Mortality Syndrome (POMS) across generations via epigenetic reprogramming and sustained immune gene expression (Fallet et al., 2022). Early-life FMT modulates immune programming and may enhance disease resistance, though epigenetic mechanisms and cross-generational evidence specifically from FMT remain emerging areas (Cao et al., 2023; Dong et al., 2024).

SynComs transcend restoration by enabling targeted manipulation of host functions through ecological and functional strain selection. The advanced design incorporates AI-predicted strain interactions and CRISPR-enhanced traits (e.g., bile tolerance), validated within in vitro gut-on-chip systems—microfluidic devices that culture intestinal epithelial cells under controlled flow with co-cultured microbes to emulate gut physiology—before deployment (Guo et al., 2023). Gnotobiotic models have been pivotal in elucidating strain-specific roles, such as Cetobacterium dominance in probiotic-fed grey mullet correlating with upregulated immune genes (IL-1β, TNF-α) and improved survival against Nocardia seriolae (Chan et al., 2024). Rational design is exemplified by a shrimp SynCom (Paracoccus, Ruegeria, Microbacterium, Demequina, Tenacibaculum) that suppressed V. parahaemolyticus, improved growth, and restored immune parameters through competitive exclusion and metabolic synergy (Guo et al., 2023). This SynCom reduced pathogenic Vibrio by 89% and enhanced thermal resilience via Tenacibaculum-mediated metabolic adjustments (Guo et al., 2023). SynComs can also improve environmental resilience (e.g., thermal/salinity tolerance) via optimized resource utilization, member replacement, and microbial communication (Jiang et al., 2024; Dubey et al., 2025). Case studies highlight their precision: a Bacillus spp. and Lactobacillus plantarum mix reduced Streptococcus agalactiae mortality by 40% in Nile tilapia via gut barrier enhancement (Huang X. et al., 2024), while Streptomyces sp. D6 suppressed Aeromonas veronii in crucian carp using bacteriocin-like compounds (Zhao et al., 2025). Encapsulation technologies further optimize SynCom delivery by enhancing microbial persistence in hydrogels (Fu S. et al., 2025).

Both strategies require rigorous biosafety frameworks due to inherent risks. FMT may facilitate horizontal gene transfer (HGT) of antibiotic resistance or virulence factors, particularly if donors harbor dysbiotic or pro-inflammatory communities (e.g., carrageenan-induced microbiota exacerbating colitis) (Wu et al., 2021, 2022). Engraftment failure is common when recipient conditions (e.g., diet, water chemistry) mismatch donor niches, as shown in interspecific transplants where dietary alignment was critical for stability (Li W. J. et al., 2022; Ruiz et al., 2024). Multi-omics-guided keystone taxon identification (e.g., Akkermansia for mucus integrity) minimizes ecological disruption during transplantation (Huang et al., 2025). For SynComs, an incomplete understanding of strain interactions risks community collapse or dominance of opportunistic taxa. Furthermore, environmental persistence of introduced strains could disrupt native microbiomes or nutrient cycles (Zhao Z. et al., 2023). Standardized donor screening, multi-omics-guided keystone taxon identification, and ecological risk assessments are thus essential for responsible application.

Gut-on-chip platforms—advanced microfluidic devices that emulate the structural complexity, cellular organization, and dynamic physiology of the intestinal tract—are revolutionizing the study of host–microbe interactions by simulating the intestinal microenvironment of aquatic species. These systems recreate critical features including mucus-secreting epithelia, vascular-like perfusion, mechanical peristalsis, and oxygen gradients, enabling physiologically relevant modeling of gut barrier integrity, immune responses, and probiotic colonization dynamics. By integrating living cells from target species (e.g., fish intestinal epithelium), these chips allow precise testing of probiotic efficacy, host-pathogen interactions, and metabolite exchange under controlled yet dynamic conditions. This technology bridges the gap between traditional in vitro models (oversimplified) and in vivo trials (ethically and logistically challenging), offering high-throughput screening of microbiome interventions. For example, a recent dual-sample microfluidic LAMP ‘gut-on-chip’ detects ten aquatic pathogens—including Vibrio parahaemolyticus—within 30 min (93% clinical sensitivity), providing a rapid pathogen-challenge module that can be coupled to probiotic screening workflows (Zhou et al., 2021). Integration with lateral flow assays (μLAMP-LFA) further enhances field applicability by enabling multiplexed diagnostics while preventing aerosol contamination (Zhu et al., 2025). Gnotobiotic models, such as germ-free zebrafish (Solis et al., 2020; Jia et al., 2024), have laid the groundwork for these platforms by demonstrating how specific bacterial communities influence host immunity and pathogen resistance. For instance, germ-free zebrafish colonized with synthetic microbial communities (SynComs) revealed that commensal bacteria like Cetobacterium somerae enhance glucose homeostasis via acetate production (Wang et al., 2021), insights critical for validating Gut-on-Chip responses. Recent studies using germ-free Atlantic salmon (Gómez de la Torre Canny, 2023) further highlight how mucosal barrier function and adipose tissue dynamics depend on microbiota composition, parameters that can be monitored in vitro using chip technology. By integrating multi-omics data from gnotobiotic models, Gut-on-Chip systems allow high-throughput screening of probiotics, reducing reliance on live animal tria0-9ls while accelerating the development of targeted microbial therapies. These platforms thus serve as physiologically relevant validation hubs within the engineering cycle.

Artificial intelligence (AI) and machine learning (ML) are transforming microbiome engineering by predicting host–microbe interactions and optimizing probiotic formulations through the integration of complex multi-omics datasets, environmental metadata, and phenotypic traits. For instance, support vector machines (SVM) and random ferns (RFerns) achieved 100% accuracy in detecting water quality parameters linked to aquaculture disease outbreaks (Çakir et al., 2023), while conditional forest and random forest algorithms accurately predicted Salmonella contamination in agricultural waters using microbiome signatures (Chung et al., 2023). Neural networks discriminated probiotics from non-probiotics with >90% accuracy by analyzing tRNA information content (Bergamini et al., 2022), and decision tree models predicted the in vivo immunomodulatory activity of lactic acid bacteria in snails with 88% accuracy by prioritizing phenotypic traits like hydrophobicity and autoaggregation (Charizani et al., 2024). Meta-analyses of zebrafish gut microbiota have identified stage-specific microbial taxa, data that can train ML algorithms to design age-specific probiotics (Garibay-Valdez et al., 2024). AI-driven tools like VirOncoTarget refine pathogen risk assessment by screening viral oncoproteins via adversarial networks (98% accuracy) (Beltrán et al., 2024), while ensemble spatial models predict methane emissions by correlating microbial taxa with environmental stressors (Chen C. C. et al., 2025). Bayesian networks (e.g., SAMBA tool) model how farming conditions alter gut microbiome diversity and predict responses to environmental shifts (Soriano et al., 2023). In shrimp aquaculture, SynComs enriched with Paracoccus and Ruegeria to suppress Vibrio parahaemolyticus, a process optimizable via ML-predicted synergistic combinations. Similarly, Stagaman et al. (2024) linked benzo[a]pyrene toxicity to microbiome diversity in zebrafish, demonstrating ML’s capacity to correlate pollutants with dysbiosis. ML also models climate impacts, such as warming-driven enterotype migration in Litopenaeus vannamei to forecast disease risk (Zeng et al., 2025). AI-driven resources like Microbiome Atlas (Lin et al., 2021) catalog bacterial growth patterns for colonization modeling, while foundation models with transfer learning adapt predictions to specific aquaculture contexts despite data sparsity (Han et al., 2025). These approaches are critical for aquaculture, where dynamic conditions demand adaptive strategies. Collectively, AI/ML provides the computational engine for in silico design and optimization of microbial consortia before empirical testing, enabling precision modulation of microbiomes for disease prevention, growth enhancement, and environmental resilience.

CRISPR-based genome editing is emerging as a powerful tool to engineer probiotics with enhanced pathogen-inhibiting capabilities. Recent research demonstrates how engineered riboregulators and CRISPR-based devices can enhance auxotrophic biocontainment in genetically modified probiotics. For instance, Cas9-assisted containment systems in Bacteroides thetaiotaomicron blocked transgene dissemination and prevented escape via thymidine auxotrophy, establishing a model for biosafety in aquaculture-engineered strains (Hayashi et al., 2024). Applications include CRISPR-Cas9 knockout of the hfq gene in Vibrio alginolyticus, reducing motility and mortality in scallops by disrupting biofilm formation (Ma et al., 2023), and Cas13a-mediated targeting of white spot syndrome virus (WSSV) in shrimp, extending survival by 33% post-infection (Pudgerd et al., 2024). Regulatory frameworks for CRISPR applications in aquaculture, particularly in the EU and Asia, are evolving to address biosafety and horizontal gene transfer risks, as highlighted in recent policy reviews (Gutási et al., 2023). Furthermore, phage-plasmid systems encoding toxin-antitoxin modules and CRISPR-Cas systems have been shown to limit horizontal gene transfer, enhancing host microbial immunity (Sayid et al., 2024). Direct host genome editing has also shown promise: CRISPR-modified tilapia exhibited enhanced resistance to Streptococcus agalactiae by targeting immune pathways like TLR2 (Yang et al., 2022), highlighting dual strategies for microbiome and host engineering. For instance, Aeromonas veronii secretes GlcNAc-binding protein (GbpA), which stimulates intestinal epithelial proliferation in zebrafish (Banse et al., 2023). CRISPR could augment GbpA expression or modify adhesion factors to improve probiotic persistence. Similarly, Cetobacterium somerae, which activates TLR2-mediated antiviral immunity (Liang et al., 2024), might be engineered to overexpress immunostimulatory exopolysaccharides. While CRISPR applications in aquaculture remain nascent, studies on Vibrio sp. and Aeromonas sp. (Xin et al., 2020) demonstrate how innate immune pathways can be targeted via genetically modified microbes. Challenges include ensuring horizontal gene transfer prevention and addressing regulatory concerns, but CRISPR’s precision offers unparalleled potential for tailored microbial solutions. This technology enables precise trait augmentation in probiotic chassis, feeding directly into AI-designed SynCom blueprints.

Rising water temperatures threaten aquaculture productivity, necessitating probiotics that thrive under thermal stress. A case study by Guo et al. (2023) illustrates how SynComs designed for shrimp resilience can be adapted using ML. By analyzing microbiota from biofloc systems, their team identified heat-tolerant taxa like Tenacibaculum, which were incorporated into SynComs to enhance Vibrio resistance. Biopolymer encapsulation (e.g., alginate-microencapsulated Bacillus licheniformis) ensures >90% gastric viability and sustained function across temperature fluctuations (Cota-Gastélum et al., 2025), while directed evolution yields strains like Bacillus subtilis TLDK301120C24, validated in gnotobiotic zebrafish to displace pathogens via competitive biofilm exclusion (Nayak et al., 2023). Such climate-resilient strategies align with SDG 13 (Climate Action), as highlighted in Table 3, which maps microbiome engineering approaches to their contributions toward mitigating environmental stressors. ML algorithms can further optimize these communities by predicting strain interactions under simulated warming scenarios. For example, Cetobacterium, a dominant genus in zebrafish guts (Garibay-Valdez et al., 2024), shows temperature-dependent acetate production (Wang et al., 2021), a trait ML could exploit to design climate-resilient consortia. For cold-water species like olive flounder (Paralichthys olivaceus), AI-guided SynComs enriched with Pseudomonas and Comamonas restore PUFA synthesis at suboptimal temperatures, countering metabolic dysregulation (Chen C. C. et al., 2025). Integrating environmental metagenomics with host transcriptomic data (Andersen-Civil et al., 2023) will refine these models, enabling probiotics that buffer against both pathogen outbreaks and climate variability. Deploying these advances on working farms, however, brings regulatory, ecological, and socioeconomic hurdles into sharp relief.

These technologies form a closed-loop engineering cycle: AI/ML leverages multi-omics data to design optimized SynComs; CRISPR introduces precision traits into probiotic chassis; Gut-on-Chip platforms validate host–microbe interactions under simulated environmental conditions; and encapsulation ensures field resilience. For instance, AI-designed Vibrio-targeting consortia can be genome-edited for enhanced bile tolerance, functionally validated in microfluidic intestines, and deployed via temperature-stable encapsulates (Vijayaram et al., 2024; Ding et al., 2025). This iterative workflow bridges computational prediction with empirical delivery, establishing a scalable framework for precision microbiome engineering in aquaculture.

Probiotic formulations—particularly resilient Bacillus spp.—dominate commercial adoption in major aquaculture regions like Asia-Pacific, where they competitively exclude pathogens (Vibrio, Aeromonas, Edwardsiella) via quorum-quenching mechanisms and enhance digestive physiology (Muras et al., 2021; Santos et al., 2021). EU-approved probiotics (e.g., Pediococcus acidilactici) demonstrate immune modulation in salmonids, reducing mortality from pathogens like Yersinia ruckeri by 25–50% (Villumsen et al., 2020; Yousuf et al., 2023). Synbiotics show accelerated growth, with shrimp-specific formulations improving yields by >600 kg/pond through ammonia reduction and microbiome stabilization (Shinde et al., 2023; Wang Q. et al., 2024). Postbiotics gain traction for storage stability and reduced regulatory burdens; heat-inactivated Lactiplantibacillus plantarum and engineered Lactococcus lactis (expressing host cytokines) enhance immune responses and disease resistance in tilapia and salmon without live-cell risks (Muñoz et al., 2021; Ferro et al., 2024). Plant-derived supplements (e.g., fermented herbal blends, resveratrol) and encapsulated bioactives (e.g., nano-astaxanthin) further augment growth and stress tolerance, though adoption varies regionally (Fujaya et al., 2023; Elbahnaswy and Elshopakey, 2024; Kari et al., 2024; Lee et al., 2024).

Field validations confirm significant productivity gains: Bacillus velezensis supplementation in shrimp ponds increased survival by 23% and upregulated hepatopancreatic immune genes (Abdelsamad et al., 2024), while synbiotic strategies in giant freshwater prawns induced complete resistance to Aeromonas veronii (Chin et al., 2025). Postbiotics accelerated microbiome recovery post-antibiotics in shrimp, reducing pathogenic Vibrio by 40% (Luo et al., 2024b). Species-tailored blends (e.g., Pediococcus/Lactococcus/Weissella consortia in trout) enhanced intestinal morphology and cytokine expression (Mahmoodian et al., 2025). Performance gains are most consistent when formulations align with host ecology—e.g., temperature-adapted probiotics in olive flounder improved winter survival (Lee et al., 2024), and microalgae-phytase synergies boosted seabass growth without disrupting core microbiota (Peralta-Sánchez et al., 2024).

Rational SynComs designed from native taxa show promise in pathogen suppression (e.g., V. parahaemolyticus in shrimp) and stress resilience (e.g., salinity/thermal tolerance) but remain at the pilot stage due to challenges in donor-recipient compatibility and ecological instability (Wang Z. et al., 2023). FMT effectively restores antibiotic-disrupted microbiomes in controlled settings yet faces barriers in horizontal gene transfer (HGT) risk (e.g., antimicrobial resistance (AMR) propagation) and farm-scale workflow standardization (Behling et al., 2024; Tian et al., 2024). Encapsulation (e.g., pH-responsive microcapsules) is critical for SynCom viability and intestinal delivery, though ecological matching and gnotobiotic validation are prerequisites for field success (Dremova et al., 2023; Jiang et al., 2024).

AI/ML tools enable risk prediction and in silico consortium design but require integration into telemetry-linked farm trials to address domain shifts between lab and pond conditions (Bergamini et al., 2022; Çakir et al., 2023). CRISPR applications (e.g., containment-enhanced probiotics, host-immunity tuning) are constrained by regulatory ambiguity—particularly GMO classification in the EU—and unresolved technical hurdles like off-target effects and biocontainment validation (Okoli et al., 2022; Gutási et al., 2023; Sheng et al., 2025). Field-validated environmental DNA (eDNA) surveillance (Technology Readiness Level [TRL] 6) demonstrates utility in species detection but lacks standardized ecosystem integration (Kim et al., 2024).