Monitoring the freshness of Salmo salar within cold chain logistics is paramount for ensuring food safety. However, conventional physicochemical and microbiological assays are impeded by inherent limitations, including destructiveness and significant time latency, rendering them inadequate for the real-time, non-invasive inspection demands of modern industry. Here, we present a novel detection framework synergizing hyperspectral imaging (400–1000 nm) with the Transformer deep learning architecture. Through a rigorous comparative analysis of twelve preprocessing protocols and four feature wavelength selection algorithms (Lasso, Genetic Algorithm, Successive Projections Algorithm, and Random Frog), prediction models for Total Volatile Basic Nitrogen (TVB-N) and Total Viable Count (TVC) were established. Furthermore, the capacity of the Transformer to capture long-range spectral dependencies was systematically investigated. Experimental results demonstrate that the model integrating Savitzky-Golay (SG) smoothing with the Transformer yielded optimal performance across the full spectrum, achieving determination coefficients (R²) of 0.9716 and 0.9721 for the Prediction Sets of TVB-N and TVC, respectively. Following the extraction of 30 characteristic wavelengths via the Successive Projections Algorithm (SPA), the streamlined model retained exceptional predictive precision (R² ≥ 0.95) while enhancing computational efficiency by a factor of approximately six. This study validates the superiority of attention-mechanism-based deep learning algorithms in hyperspectral data analysis. These findings provide a theoretical foundation and technical underpinning for the development of cost-effective, high-efficiency portable multispectral sensors, thereby facilitating the intelligent transformation of the aquatic product supply chain.

Nanoplastics are ubiquitous in aquatic environments, and elucidating their bioaccumulation behavior is essential for assessing toxicity and trophic transfer risks. While most studies focus on nanoplastics properties (e.g., type, size, surface charge), the influence of organismal growth stage remains unclear. Through bioaccumulation kinetic experiments with 10 mg/L polystyrene nanoplastics (PSNPs), this study found that Tetrahymena thermophila (T. thermophila) in the lag phase (cell density 4 × 10⁴ cells/mL) exhibited the highest uptake rate of nanoplastics, 1.2–5.8 times that of the exponential phase (1–5 × 10⁵ cells/mL) and 7.7 times that of the stationary phase (>5 × 10⁵ cells/mL). Lag phase cells also had a larger specific surface area (0.319 vs. 0.271/0.269 μm⁻¹), supporting their heightened uptake capacity. Under PSNP exposure, exponential and stationary phase cells showed significantly elevated reactive oxygen species (ROS) levels, accompanied by downregulated superoxide dismutase (SOD) and stable catalase (CAT) activity, indicating impaired antioxidant defense and potential redirection of energy toward stress mitigation. Consistent with efficient internalization, confocal imaging revealed clear PSNP colocalization within food vacuoles of lag period cells. Proteomic and transcriptomic analysis further confirmed the upregulation of carrier proteins, FAD/FMN oxidoreductases, and pathways associated with cellular components (membrane and organelle membrane) and molecular functions (transporter activity and transmembrane transporter activity) in lag-phase T. thermophila. Collectively, these findings provide a molecular-level understanding of the multi-phase-dependent bioaccumulation of PSNPs, offering critical insights for assessing the environmental risks of polystyrene nanoplastics in dynamic aquatic ecosystems.

The gut microbiome of rainbow trout (Oncorhynchus mykiss) and Atlantic salmon (Salmo salar) is increasingly recognized as a functional interface linking dietary inputs, epithelial barrier integrity, mucosal immunity, environmental stress, disease susceptibility, and antimicrobial-resistance risk in intensive aquaculture. Based on available salmonid studies and relevant evidence from broader fish and aquaculture systems, this review synthesizes current knowledge on salmonid gut microbial composition, nutritional modulation, microbiome–mucosal immune interactions, aquaculture stressors, antibiotic exposure, antibiotic resistance genes (ARGs), mobile genetic elements (MGEs), metagenomics, multi-omics, and emerging microbiome-informed decision-support tools. Current evidence does not support a universally stable single-core microbiota in these species. Instead, community structure is shaped by developmental stage, freshwater–seawater transition, intestinal segment, digesta versus mucosa sampling, diet, temperature, stress, health status, and methodological workflow. Feed substitution and functional additives can remodel the gut microbiota, but these shifts should be interpreted alongside histology, barrier function, metabolic profiles, immune indicators, and disease-resistance phenotypes. Antibiotic exposure may reduce acute bacterial disease pressure while disturbing community structure and potentially enriching ARGs or ARG–MGE associations. Risk assessment should therefore move beyond ARG abundance toward host–ARG–MGE linkage using shotgun metagenomics, metagenome-assembled genomes, long-read sequencing, Hi-C, and externally validated multi-omics models. Machine learning and artificial intelligence approaches may support feature screening, risk stratification, and decision support, but their application in salmonid gut-health management remains at an early stage and requires external validation across sites, production stages, diets, and seasons.

This study evaluated the endpoint biochemical responses and associated physiological status of rainbow trout (Oncorhynchus mykiss) of different body sizes during salinity acclimation under varying temperature and salinity conditions. An L₉ (3⁴) orthogonal design was applied using three nominal body-size levels (100, 250, and 500 g), three temperatures (10, 15, and 20 °C), and three salinities (10, 20, and 30‰) over 60 days. Hepatic superoxide dismutase (SOD), alkaline phosphatase (ALP), and acid phosphatase (ACP) activities and plasma malondialdehyde (MDA), urea nitrogen (BUN), ammonia, alanine aminotransferase (ALT), and aspartate aminotransferase (AST) were measured. For inferential analysis, the three fish sampled from each tank were averaged to obtain one tank-level value, yielding 18 tank-level observations. Tank-level main-effects analysis showed that body size was significantly associated with ALP, ACP, MDA, BUN, ammonia, ALT, and AST. Temperature was significantly associated with SOD, ACP, MDA, BUN, and AST, whereas salinity was significantly associated with MDA, ALT, and AST. MDA and AST were the only endpoints associated with all three factors. The biochemical responses were pathway-specific rather than uniformly directional across treatment combinations. Because interactions could not be independently estimated in the L₉ design, the identified main effects should be interpreted as orthogonal screening associations. Overall, body-size class was associated with the broadest range of endpoints and should be incorporated into future experimental assessments of rainbow trout within the tested salinity range of 10–30‰.

Drought stress severely restricts early growth in Brassica napus, while selenium (Se) can enhance stress tolerance. This study examined how foliar sodium selenite pretreatment affects polyethylene glycol (PEG)-induced drought stress in B. napus seedlings. Seedlings were sprayed with 10, 20, or 40 mg L⁻¹ Na₂SeO₃ for 7 days and then exposed to 10% (w/v) PEG 6000 for 2 days. Drought reduced fresh weight and increased malondialdehyde (MDA), O₂·⁻ production, H₂O₂ accumulation, and antioxidant enzyme activities. The response was nonlinear, and 20 mg L⁻¹ Na₂SeO₃ was most effective. Relative to drought alone, this treatment increased fresh weight from 2.6 to 5.1 g plant⁻¹ and reduced MDA, O₂·⁻, and H₂O₂ by 50.5%, 50.9%, and 44.3%, respectively. Notably, catalase (CAT) and superoxide dismutase (SOD) activities declined markedly from the drought-induced levels, indicating lower oxidative pressure rather than enhanced enzymatic scavenging. Integrated transcriptomic and metabolomic analyses showed associations between selenium pretreatment and pathways related to photosynthesis, carbon fixation, porphyrin metabolism, phenylpropanoid biosynthesis, and selenocompound metabolism. Together, these findings suggest that foliar application of 20 mg L⁻¹ Na₂SeO₃ improves drought tolerance in B. napus seedlings by alleviating oxidative pressure rather than simply increasing antioxidant enzyme activity.