Some fish farmers have implemented intensive cleaning and disinfection procedures encompassing entire production facilities. Despite the associated costs and production downtime, these strategies have reportedly improved subsequent fish performance and system functionality.This study documents a rigorous terminal disinfection procedure conducted at an open 1700 m3 commercial freshwater rainbow trout recirculating aquaculture system (RAS). The treatment was performed during a production break after all fish had been removed and the system was bypassed circulating water without any discharge. Briefly, 2,000L of 27% sodium hydroxide (NaOH) were added first, followed by 1,000L of 35% hydrogen peroxide (H₂O₂) two days after. The system was then maintained under full recirculation for a total of three days. Water quality parameters were monitored throughout the disinfection period, and biofilter performance was assessed before, during, and after disinfection.Extreme water quality conditions were recorded, including pH values exceeding 11 for three consecutive days and circulating H₂O₂ concentrations above 100mg/L. Hydrogen peroxide concentrations declined exponentially with a half-life of approximately–8h and decreased to approximately 20mg/L within 24h after application. The disinfection procedure markedly altered microbial characteristics. The ratio of live to dead cells decreased from approximately 75:25 before disinfection to 7:93 during treatment, while total cell density increased from 3×10⁶ to nearly 2×10⁷ cells/mL. Nitrification was completely inhibited, resulting in prolonged accumulation of total ammonia nitrogen (TAN) and nitrite during system restart. Biofilter microbial activity was reduced by 91%, indicating that portions of the biofilm community survived the treatment.Because complete microbial elimination was not achieved, the findings emphasize the protective properties of biofilms and provide new insights into factors limiting terminal disinfection efficiency in RAS. The results further highlight the importance of mechanical cleaning and biofilm removal as components of effective sanitation protocols and demonstrate the need for quantitative methods to assess disinfection efficacy in aquaculture systems.
To address the challenge of efficiently harvesting live fish from large-scale deep-sea aquaculture platforms with individual yields exceeding 100 tons, a negative-pressure fish harvesting system was developed by integrating a buffer tank, fish-water separation device, vacuum priming tank, and centrifugal pump. Analysis based on the ideal gas state equation indicates that a stable operating pressure sufficient for fish harvesting can be achieved when the liquid level in the buffer tank is maintained within 5.3% to 10.2% of its total height. Model tests indicate that variation in operating pressure is kept within 4.3%, while variation in the operating liquid level does not exceed 4.2%. The survival rate of test fish reached 100%, with no obvious mechanical injury to their bodies, which verifies the feasibility of this fish harvesting principle. The engineering design of the system is carried out targeting a 100,000-ton aquaculture vessel, equipped with a variable-frequency automatic fish harvesting system taking into account the system’s operational control scheme to realize coordinated regulation of negative pressure, liquid level and flow rate. Comprehensive on-board tests demonstrated that within the variable-frequency range of 800 to1,500r/min, the delivery flow rate and negative pressure have a linear correlation with rotational speed, and the negative pressure stays within the safe range of −37.28 to −33.5 kPa under all operating conditions. The optimal operating speed range of 1,000–1,200r/min balances transport efficiency and fish protection. Under these conditions, the system achieved a stable flow rate of 180m³/h, with a sample inspection revealing a yellow croaker injury rate of only 2.5%. The observed injuries were limited to minor scale loss and fin wear, with no severe injuries such as swim bladder rupture. Throughout the entire harvesting operation, no malfunctions such as fish blockages or negative pressure instability occurred. The system’s comprehensive performance meets the requirements for large-scale, centralized, low-injury, and efficient harvesting on 100,000-ton-class aquaculture vessels.Author Profile: Lin Author Profile: LinLiqun, Associate Researcher, mainly engaged in research on fishing harvesting equipment. linliqun@fmiri.ac.cn
In grouper (Epinephelus spp.) aquaculture, the weaning period is a major developmental bottleneck and operational challenge. Existing research on automated monitoring within recirculating aquaculture systems (RAS) has primarily focused on measuring large adult fish or simply counting fry. A reliable non-contact method for real-time measurement of the highly variable body length of rapidly developing fry is lacking. Without precise data, aquaculture managers must rely on subjective experience to make decisions in terms of feeding and other aspects of fish rearing. This often leads to size heterogeneity, which, in turn, causes frequent grading-induced stress and persistent disturbance to the rearing environment. To address this issue, we developed a deployable stereovision system for automated individual-level length estimation within an RAS. The system was built and validated using data collected from 17 fry batches over four years, ensuring robust performance under operational hatchery conditions. By integrating a region of interest based detection pipeline with 3D reconstruction, the system delivers stable real-time body length measurements throughout the weaning-to-juvenile transition. The efficacy of the system was validated through comparative trials across seven batches (three automated and four manual). Real-time data suppressed size heterogeneity significantly better than did conventional visual estimation. The results of this study demonstrate that the developed stereovision technology fills a critical void in fry length monitoring. The proposed monitoring system offers a practical digital tool for enhancing production stability and reducing environmental disturbance in grouper hatcheries.
Passive acoustic monitoring (PAM) has proven effective for feeding management in penaeid shrimp aquaculture, but its extension to crab aquaculture requires species-specific acoustic characterization to inform system design. We aimed to provide this acoustic foundation for the Chinese mitten crab (Eriocheir sinensis), a major freshwater aquaculture species. We recorded feeding sounds from 38 crabs (23–85 g) using a calibrated hydrophone in a sound-insulated tank and examined three acoustic dimensions relevant to PAM system design: event classification, body weight effects, and sex effects. From 20,877 detected acoustic events, we identified three acoustically distinct feeding sound types: single-pulse crushing clicks (84.2% of events) produced by discrete mandible closures, composite crushing bursts (2.0%) with a dominant main pulse and weak secondary pulses, and chewing sequences (13.8%) reflecting sustained rhythmic grinding. Body weight correlated significantly with total feeding event rate (r = 0.674, p < 0.001) but not with peak frequency, establishing event rate as size-sensitive and peak frequency as size-invariant. Sex independently affected event count — males exhibited approximately 1.9-fold higher feeding event counts than females — while spectral features remained indistinguishable between sexes. Pulse width showed moderate between-individual consistency (ICC = 0.527), whereas peak frequency and pulse amplitude were dominated by within-individual variability. These findings establish the design specifications for PAM-based feeding systems in crab aquaculture: event rate as a proxy metric for feeding activity, peak frequency as a species-specific detection marker robust to size and sex variation, and the need for multi-class event detection algorithms.
The sustainability and productivity of modern aquaculture are strongly influenced by the continuous monitoring of water quality. Conventional water quality monitoring relies on fixed sensing stations where operational costs increase. To overcome this limitation, continuous real-time monitoring of aquatic environments is essential for sustained and efficient fish production. Rapid development in the field of Internet of Things (IoT) and mobile sensor nodes for real-time water quality monitoring has the potential to improve and modernize aquaculture practices. This article presents a cloud-based IoT framework with integration of an underwater mobile sensor node for real-time water-quality data collection across diverse aquatic environments. The proposed IoT-based water-quality monitoring framework aims for a more user-friendly approach through reduced operational complexity and cost, with increased operational lifetime. The developed monitoring framework utilizes different environmental sensors, such as pH, temperature, total dissolved solids, turbidity, and dissolved oxygen sensors, with a Raspberry Pi 5 microprocessor to form a complete embedded system. The collected water-quality data is then transmitted to a cloud monitoring platform using a LoRa module for further remote analysis. The development process of the underwater mobile sensor node includes an efficient approach for internal space allocation, optimized sensor workflow, and an intelligent navigation framework. The developed mobile sensor node is then deployed for collecting data from five different water bodies with varying environmental conditions and dimensions. For more reliable water-quality analysis, the data have been collected at different times of day. The collected data is then compared with standard reference values of different water-quality parameters to determine whether the water body is suitable or unsuitable for aquaculture.
In this study, we explored some ecological stoichiometry theory (ES theory) and plankton ecology group model (PEG model) concepts that may assist regenerative mechanisms in hybrid pond aquaculture that are gaining popularity worldwide. Using fish-free pond-water mesocosms, we examined the response of algal growth to additions of fish feces and carbonaceous material under realistic or challenging dilution scenarios during both dynamic and steady states (Experiments I-II). Using pond living labs, we examined uncertainties in zooplankton body composition under variable bottom-up and top-down proxies, mediated by nutrient stoichiometry and fish stock index (FSI), respectively (Experiment-III). Experiment I, which explored graded levels of N and P at the expense of cellulose-lignin type organic C, showed P-enrichment explaining treatment differences, with elevated C-inputs partially suppressing algal proliferation. Experiment-II, which explored stoichiometric effects of molar C:P (at fixed N) and N:P (at fixed C) ratios, indicated that solids with lower C:P and N:P ratios may favor algal response, with N:P exerting influence likely over C:P under the experimental conditions tested. At high C but low P, brownification or dissolved oxygen risks were hinted. Experiment-III revealed that zooplankton (>90% cladocerans) may maintain their amino acids under fluctuating bottom-up/top-down pressures, except tyrosine (for storage proteins). The fatty acids 20:4n-6, 20:5n-3, and 22:6n-3 (for reproduction) in zooplankton exhibit greater variability under variable bottom-up/top-down pressures and irrespective of season. Such sensitives in cladoceran body nutrient reserves may be concomitant with shifts in population fitness (condition factor). Under the experimental conditions tested, our findings hint that by better linking the ES theory (C-amendment, P-reduction, but carefully avoiding brownification risk) and PEG model (fish−−scenario), regenerative aquaculture could be explored in future hybrid ponds.
Integrating hydroponics into Recirculating Aquaculture Systems (RAS) enhances nutrient reuse and eco-efficiency. This study evaluated a coupled RAS–NFT aquaponic system cultivating rainbow trout (Oncorhynchus mykiss) and lettuce (Lactuca sativa) under different NFT pipe diameters (75, 90, and 110 mm) and hydraulic flow rates (20, 30, 46, and 60 L h⁻¹), integrating a nutrient mass flow rate modeling approach. A 28-day trial compared a fish-only traditional RAS against a biointegrated RAS–NFT system. Physicochemical water quality parameters and dissolved nutrients were monitored regularly. To alter standard hydrodynamics, the discharge end of each main NFT pipeline was fitted with a 50 mm PVC reducer, creating a hydraulic backwater effect that modified the pipelines into a water-retaining, flooded-channel system operating under a full-pipe cross-sectional regime. To evaluate nutrient mass delivery under different hydraulic conditions, nutrient mass flow rate (Jm, mg min⁻¹) delivered to the rhizosphere were estimated using a mass transport model based on volumetric flow rates and mean dissolved nutrient concentrations. Weekly plant growth and final biomass production indices were evaluated. Biointegration maintained physicochemical water quality within similar ranges in both production systems. Among the dissolved elements, potassium exhibited higher mean concentrations in the coupled aquaponic system, due to routine alkalinity corrections, whereas the remaining nutrients showed comparable descriptive values. Mass transport modeling indicated that estimated nutrient mass flow rates (Jm) increased proportionally with volumetric inflow rates (Q), as expected from the modeled hydraulic scenarios. Under the 20 L h⁻¹ configuration, the mass transport model estimated nutrient mass flow rates of 39.33 mg min⁻¹ for NO₃⁻ and 25.67 mg min⁻¹ for PO₄³ ⁻ delivered to the hydroponic NFT channels. Variations in pipe diameters and flow rates yielded no significant differences in final shoot fresh weight, dry mass, or leaf count (p > 0.05), suggesting that nutrient delivery was sufficient to support crop development across the tested operational range. The flooded-channel hydroponic NFT configuration operated under ultra-low fluid velocity conditions (4.53–4.84 m h⁻¹; 0.126–0.134 cm s−1 within the proportional core). Mass transport modeling indicated that estimated nutrient delivery remained sufficient across the evaluated hydraulic scenarios, and no reductions in lettuce growth were observed under the tested conditions. These findings suggest that, under the evaluated hydraulic conditions, engineering designs for coupled aquaponic systems may be optimized for structural, spatial, and economic considerations while maintaining plant growth.
Aquaponics integrates recirculating aquaculture systems (RAS) with hydroponic plant cultivation with the goal of creating a more resource-efficient food production strategy. A central question in aquaponic design is whether system performance is improved when aquaculture and hydroponic units are hydraulically coupled in one loop or decoupled into independently managed modules. This review evaluates that question through a systems and microbial-ecological lens, with emphasis on productivity, nutrient-use efficiency, and resilience outcomes. Current evidence does not support a consistent productivity advantage for either coupled or decoupled configurations. Reported differences are frequently confounded by nutrient loading, solids management, hydraulic retention time, species selection, environmental control, and site-specific operating conditions. Microbial effects are likely important, but most studies remain taxonomic rather than functional, limiting mechanistic interpretation. Overall, configuration should be treated as one design variable within a broader operational framework rather than as a stand-alone predictor of yield or robustness.
Waterless low-temperature transportation efficiently preserves live aquatic products but may induce multi-scale physiological stress responses, leading to reduced vitality, quality deterioration, and economic losses. Conventional stress-detection methods are destructive and unsuitable for real-time monitoring, whereas flexible sensing enables continuous non-invasive stress assessment. This review summarises advances in multi-scale flexible sensing and intelligent modelling for nondestructive stress assessment of live aquatic products. It outlines multi-scale fish stress-response mechanisms and key parameters, introduces flexible-sensor design and applications for nondestructive monitoring at cellular, tissue/organ, and individual scales, and examines flexible sensing and intelligent modelling for stress and vitality assessment. Current technical limitations and future research directions are discussed to guide further research on nondestructive stress assessment of live aquatic products.The integration of multimodal flexible sensing and intelligent modelling has considerable potential to advance stress assessment of live aquatic products. Flexible sensors can acquire multi-scale stress-related indicators nondestructively, while intelligent models can reveal complex relationships among multisource information and predict changes in stress and vitality, enabling early warning and timely intervention. These technologies may promote more intelligent and precise nondestructive stress monitoring during waterless low-temperature transportation of live aquatic products.
Two consecutive case studies assessed the feasibility of composting dewatered sludge (DS; fecal matter and uneaten feed) obtained from recirculating aquaculture systems. Utilizing a commercially available rotary drum composter (3.15 m3 working volume; 14-d retention time), DS was combined with wood flakes (WF) bulking agent and co-composted with either spent mushroom compost (SMC) in case study 1 or fish mortalities (FM) in case study 2. The case studies characterized feedstocks, evaluated potential recipes, and assessed process performance and finished product quality. In case study 1, co-composting DS with SMC (DS-SMC) reached thermophilic temperatures (average hottest-zone temperature of 58.8 ± 0.6 °C), thereby accelerating feedstock breakdown. Compost was considered stable and mature post-drum, as indicated by a low CO2 respiration rate of 0.2 ± 0.0 mg CO2-C/g OM/day, with a final C:N ratio of 19 ± 2 after a 6-month post-drum curing period. In case study 2, co-composting DS with FM (DS-FM) resulted in low hottest-zone temperatures (averaging 51.9 ± 1.0 °C), likely due to the heterogeneous nature of the combined feedstocks and increased drum rotations to minimize ammonia volatilization. High ammonium concentrations [571 ppm or 0.06%; 10% of total nitrogen (TN)] in the post-drum samples indicated incomplete decomposition. Six months post-drum, ammonia accounted for 1% of TN, but the C:N ratio remained elevated (28 ± 4). Additional analyses from both studies highlighted low concentrations of heavy metals and human pathogens. These two case studies suggest that co-composting with additional substrates may be an effective strategy for stabilizing aquacultural biosolids.
The shrimp Penaeus vannamei is the most reared crustacean species in the world, with production gradually increasing over time. This growth is linked to the intensification of production, which increases yield results and nutrient-rich residues, such as nitrogen. Over time, many strategies for the control of nitrogenous compounds have been developed and used in the intensive culture of P. vannamei, consistently aiming to reduce water and land use and effluent generation, as well as to increase productivity. The great effort of the academy to develop strategies such as recirculating aquaculture systems (RAS), photoautotrophic, heterotrophic and chemoautotrophic (e.g., biofloc system), synbiotic systems, use of probiotic bacteria, water reuse, use of artificial substrates, denitrification phases, and integrated multitrophic systems met the need for the management of nitrogenous compounds and the creation of environmentally friendly systems. Although several studies have demonstrated the effects of these strategies on water quality, few studies have comprehensively synthesized this information by describing their key characteristics, mechanisms of action, and the overall trends reported in the literature. Therefore, this review aimed to critically address the main strategies for the control of nitrogenous compounds in intensive cultures of P. vannamei.
As marine cage aquaculture continues to expand and the culture period extends, biofouling on cage nets has become an increasingly serious issue. Biofouling poses serious threats to cage structural integrity and the welfare of farmed fish. Traditional antifouling technologies for nets can be destructive to the netting material and the aquatic environment. To explore more efficient and environmentally friendly antifouling strategies for fouled nets, this study conducted a test to evaluate antifouling performance in July 2023.The test was conducted in a marine aquaculture area using a cage model fitted with two materials of netting (Polyethylene-PE, Polyamide-PA) and equipped with UV-C. The results indicate net biofouling and biofilm community was substantially reduced under the UV-C irradiation. After 10 days of UV-C irradiation, PA net fouled occlusion decreased by 24.5% and biomass gain decreased by 203.5 g·m-², while PE fouled net occlusion decreased by 5.7% and biomass gain decreased by 170.5 g·m-². Microbial α-diversity in PA and PE nets showed a significant reduction, with the most pronounced effects observed on day 4 and day 7. Moreover, the microbial communities conducive to the maturation of fouling biofilms have significantly declined. Desulfobulbia and Rhodobacteraceae significantly decreased in PA nets. In contrast, in PE nets, only Desulfobulbia decreased significantly, whereas Flavobacteriaceae remained unaffected. This study demonstrates the antifouling potential of UV-C technology as a promising method to control net biofouling, supporting its future application in aquaculture engineering.
Water-quality deterioration caused by feed waste and dissolved inorganic nitrogen (DIN) accumulation remains a major challenge in super-intensive shrimp farming. This study developed and field-tested an integrated smart pond engineering system for Pacific white shrimp (Litopenaeus vannamei) culture in southern Taiwan. The system combined continuous water-quality monitoring, underwater video-based feeding assessment, remotely controlled pond components, automated feeding support, and a funnel-bottom circular pond with central bottom discharge. A 114-day trial was conducted using a round concrete smart pond, while a rectangular pond stocked at the same density served as a conventional control. The funnel-bottom design helped concentrate feces, uneaten feed, and nitrogen-rich sludge toward the central discharge pipe for removal. Compared with the traditional pond, the smart pond showed lower mean nitrate, nitrite, and ammonium concentrations of 19.0, 4.8, and 6.9 μM, respectively, with a total DIN of 30.6 μM, whereas the traditional pond reached a total DIN of 272.6 μM. Shrimp in the smart pond also showed improved growth performance, suggesting more stable rearing conditions. Although the system improved monitoring and reduced some routine management demands, it required higher equipment investment and did not reduce unit production cost in this pilot-scale trial. These results suggest that integrated pond engineering, sensor-based monitoring, video-assisted feeding assessment, and targeted bottom discharge can improve nitrogen-waste management in intensive shrimp culture.
Fish school motion can substantially influence the hydrodynamic characteristics and waste discharge performance of aquaculture tanks. Herein, a computational fluid dynamics (CFD)-based fluid-structure interaction model was coupled with a discrete phase model to investigate waste transport and discharge under the disturbance induced by bionic fish schools. The model was applied to three tank configurations, with the no-fish condition serving as the control case. In the circular tank, the presence of fish schools reduced the mean flow velocity by approximately 23%-32% and decrease the particle discharge rate at 60 s by approximately 37%-50% relative to the no-fish case. In the square arc angle tanks, fish schools caused a 2%-9% reduction in mean velocity, and the overall flow structure remained comparatively stable. However, in the square tank, fish-induced motion enhanced mixing with localized low-velocity regions, increasing the mean velocity by up to approximately 42% and increasing the particle discharge rate at 110 s from 3.6% to 20.9%. These findings indicate that, in tanks with relatively weak baseline flow patterns, turbulence generated by fish school motion can serve as an important driving mechanism for particle transport. Overall, this study elucid ates the role of fish school motion in shaping tank hydrodynamics and waste discharge efficiency and provides insights for aquaculture system design.
Geosmin contamination causes substantial economic losses in recirculating aquaculture systems, yet tissue-specific accumulation patterns remain poorly characterized. This study examined geosmin distribution across Atlantic salmon (Salmo salar) tissues to identify factors influencing depuration efficacy. Market-size salmon (n = 101, 4.16 +/- 1.5 kg) were exposed to geosmin (622 ng L-& sup1; nominal) in a recirculating aquaculture system (RAS) and sampled over 168 h. Blood geosmin concentrations strongly correlated with all tissues (R & sup2;>0.89), demonstrating potential for non-lethal monitoring of tissue burdens. Among tissues examined, ovaries exhibited exceptional concentration-based enrichment (6-7 & times; relative to mass fraction), with peak concentrations (1764 ng kg(-)& sup1;) exceeding fillet (537 ng kg(-)& sup1;) and liver (277 ng kg(-)& sup1;) by 3.3- and 6.4-fold, respectively. However, all sampled fish had minimal ovarian development (gonadosomatic index, GSI <1.2%), such that ovaries contributed only 6.6% of whole-body geosmin burden at peak despite high tissue-specific concentrations. Projection modeling indicates that if observed tissue/blood ratios persist in sexually maturing fish, ovarys at GSI= 10% would contain approximately 43% of total body burden and could substantially influence depuration kinetics. With an estimated average of 34% of commercial salmon experiencing early maturation, this preliminary study establishes a quantitative hypothesis warranting experimental validation: that advanced sexual maturation may require extended depuration periods in RAS facilities. Blood-based monitoring protocols could enable non-lethal assessment of depuration progress and identification of high-burden individuals.
Automated underwater net-cleaning methods based on net-cleaning equipment (NCE) has effectively alleviated the prominent problems of high labor intensity and low cleaning efficiency inherent to conventional manual cleaning of deep-sea aquaculture net, thereby garnering extensive attention in the aquaculture industry. This paper systematically reviews and analyzes the structural designs and functional characteristics of NCE. Initially, the adverse impacts of marine biofouling organisms on the structure of deep-sea aquaculture cage systems and the aquaculture environment are summarized and analyzed. Subsequently, the performance indicators of currently existing NCE are compiled, the design features of two typical types of NCE are summarized, and each type is further reviewed and analyzed. Considering the operational characteristics of underwater net cleaning for deep‑sea aquaculture, this study further discusses core technologies for addressing key challenges, including insufficient underwater cleaning efficiency, poor adhesion to flexible net surface, and incomplete cleaning coverage. The involved technologies cover high-efficiency cleaning strategies, net-adhesive locomotion mechanisms, positioning and navigation systems, and motion control algorithms. Additionally, future developmental trends of NCE are prospected. This review provides theoretical fundamentals and practical engineering guidelines for the research, development, and industrial promotion of NCE applied in deep-sea aquaculture.
