Aquacultural Engineering最新文献

{"title":"Effects of biofloc-electrochemical system on water quality, fish performance, and microbial diversity","authors":"Wenfei Su ,&nbsp;Jincheng Lin ,&nbsp;Wenchang Liu ,&nbsp;Yanshuo Guo ,&nbsp;Jiahui Wang ,&nbsp;Guozhi Luo ,&nbsp;Hongxin Tan","doi":"10.1016/j.aquaeng.2025.102567","DOIUrl":"10.1016/j.aquaeng.2025.102567","url":null,"abstract":"<div><div>This study investigated the effects of the in-situ biofloc-electrochemical system (BES) on nitrogen (N) and phosphorus (P) removal, growth performance, health status of Nile tilapia (<em>Oreochromis niloticus</em>), and microbial composition in both biofloc and the intestinal tract. Tilapia were cultured in a biofloc system with microcurrent direct current electric fields at 0, 100, and 200 mA for 36 days. Results showed that electrolysis significantly enhanced the degradation of ammonia nitrogen (TAN), nitrite nitrogen (NO₂^--N), phosphate (PO₄^3--P), and total phosphorus (TP) (<em>p</em> &lt; 0.05), but had no significant effect on nitrate nitrogen (NO₃^--N) removal (<em>p</em> &gt; 0.05). In the 100 mA group, no significant differences were observed in feed conversion ratio (FCR) and specific growth rate (SGR) compared to the control group (<em>p</em> &gt; 0.05). However, in the 200 mA group, FCR was significantly increased, and SGR was significantly decreased (<em>p</em> &lt; 0.05), indicating that higher current levels suppressed tilapia growth performance. Enhanced digestive capacity in the intestinal tract was observed, with minimal oxidative stress effects in the gills and liver. Electrolysis also significantly altered the microbial composition, abundance, and diversity in both biofloc and intestinal microbiota (<em>p</em> &lt; 0.05). Notably, increases in Actinobacteriota and Bacteroidota may improve tilapia immunity and nutrient cycling, influencing both growth and water quality. These findings suggest that while BES effectively improves water quality, further optimization is needed to enhance fish growth performance and nitrogen removal efficiency.</div></div>","PeriodicalId":8120,"journal":{"name":"Aquacultural Engineering","volume":"111 ","pages":"Article 102567"},"PeriodicalIF":3.6,"publicationDate":"2025-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144116406","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Application of generative artificial intelligence in the aquacultural sector","authors":"Chiara Fini ,&nbsp;Simone Gaetano Amato ,&nbsp;Daniela Scutaru ,&nbsp;Sara Biancardi ,&nbsp;Francesca Antonucci ,&nbsp;Simona Violino ,&nbsp;Luciano Ortenzi ,&nbsp;Eugenio Nerio Nemmi ,&nbsp;Alessandro Mei ,&nbsp;Federico Pallottino ,&nbsp;Simone Figorilli ,&nbsp;Corrado Costa","doi":"10.1016/j.aquaeng.2025.102568","DOIUrl":"10.1016/j.aquaeng.2025.102568","url":null,"abstract":"<div><div>Artificial Intelligence (AI) applications in aquaculture have recently attracted growing attention, as these technologies are becoming vital for data analysis, improving production processes, and optimizing the use of natural resources. Among the different AI approaches, Generative Artificial Intelligence (GAI) emerges as one of the most innovative and promising.This paper explores the use of Generative Artificial Intelligence (GAI) in aquaculture, evaluating its advantages and challenges within the industry's unique context. It begins with an overview of generative model architectures, then delves into their potential contributions to aquatic resource management, the improvement of farming practices, and the promotion of environmental sustainability.To obtain the most up-to-date insights, research was carried out using database such as SCOPUS, Google Scholar, and Web of Science databases. GAI holds significant promise for aquaculture, with applications that include enhanced water quality management, fish stock health prediction, and automated feeding supervision. When applied responsibly, GAI can streamline routine operations while contributing to a more resilient and sustainable future for the aquaculture industry. With its ability to extract valuable insights from big data, GAI proves to be a powerful tool for tackling future challenges, ensuring food production that meets the demands of a growing global population while addressing increasingly urgent environmental concerns.</div></div>","PeriodicalId":8120,"journal":{"name":"Aquacultural Engineering","volume":"111 ","pages":"Article 102568"},"PeriodicalIF":3.6,"publicationDate":"2025-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144070485","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reluctance of farmed Atlantic salmon to feed in cold water revealed during automated hydroacoustic feeding control","authors":"Ole Folkedal ,&nbsp;Georgia Macaulay ,&nbsp;Jan Erik Fosseidengen ,&nbsp;Joakim Myrland ,&nbsp;Bendik Søvegjarto ,&nbsp;Tim Dempster ,&nbsp;Frode Oppedal ,&nbsp;Lars Helge Stien","doi":"10.1016/j.aquaeng.2025.102565","DOIUrl":"10.1016/j.aquaeng.2025.102565","url":null,"abstract":"<div><div>Waste feed remains a major issue in open sea-cage Atlantic salmon aquaculture. “Echofeeding” is an appetite-led feeding method that stops meals based on fish biomass detected by an echo sounder. The method reduced waste feed and upheld fish growth in a relatively vertically unstratified coastal farming environment. Here, we tested echofeeding at a commercially relevant scale over an 8-month period in a fjord environment with seasonal vertical temperature and salinity gradients. We compared fish behaviour and growth between echofed fish, fed at high intensity and near surface, and control fish, with feeding regulated by pellet detection without surface feeding restriction (conventional practice). Growth (SGR&gt;1.81) and FCR (&lt;0.87) were excellent and similar for three months after sea-transfer in August. However, a strong halocline in late November (&lt;5°C surface water) led echofed fish to avoid surface feeding, resulting in underfeeding. Following the setting of a deeper depth interval for triggering feeding, the echofed fish fed more, and fed at similar levels to control fish when feeding intensity was reduced. Echofeeding underperformed in early spring as rising surface temperatures attracted salmon, making it difficult for the system to distinguish between feeding and routine behaviours. Both groups contracted salmonid alphavirus during winter, reducing appetite and promoting early harvest. Results highlight the need for echofeeding to take environmental changes into account. Further, as fish grew, a gradual decline in the echo signal measured during feeding suggests a method for refining meal termination threshold to minimize waste feed while maintaining good fish growth.</div></div>","PeriodicalId":8120,"journal":{"name":"Aquacultural Engineering","volume":"111 ","pages":"Article 102565"},"PeriodicalIF":3.6,"publicationDate":"2025-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144089396","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Behavioral response of farmed Cod to environmental drivers and interaction with feeding practice","authors":"Peygham Ghaffari ,&nbsp;Thor Magne Jonassen ,&nbsp;Johannes Kvam ,&nbsp;Fredrik Staven","doi":"10.1016/j.aquaeng.2025.102560","DOIUrl":"10.1016/j.aquaeng.2025.102560","url":null,"abstract":"<div><div>This study investigated how environmental drivers and feeding practices influence the vertical behavior of farmed Atlantic Cod (<em>Gadus morhua</em>) in a commercial sea cage environment. A 70-day continuous observation was conducted at a farm in northern Norway, utilizing a dual-echosounder setup combined with temperature loggers and model-derived environmental data, including current velocity, salinity, and oxygen profiles. Cod behavior was compared between submerged (5 m) and surface feeding regimes across four sea cages. Fish consistently avoided surface layers above 15^∘C, aggregating at cooler depths, and shifted upward as thermal stratification weakened into autumn. Submerged feeding supported tighter aggregation, stronger rhythmicity, and vertical cohesion, while surface feeding induced anticipatory surfacing, behavioral fragmentation, and increased thermal exposure. These patterns were amplified by Cod’s physoclistic physiology and anatomical predispositions to strain from repeated vertical movement. Signal processing of depth-use patterns revealed stronger behavioral synchronization under submerged feeding, particularly aligned with the 48-hour feeding cycle. Aggregation dynamics were more closely aligned with residual current patterns and deeper water mass structure than with tidal forcing, and air temperature emerged as a reliable and practical proxy for estimating the vertical center of biomass. This study provides critical insight into farmed Cod’s depth preferences and environmental rhythms, which may be essential for enhancing fish welfare, reducing mortality, and improving the sustainability of Cod aquaculture.</div></div>","PeriodicalId":8120,"journal":{"name":"Aquacultural Engineering","volume":"111 ","pages":"Article 102560"},"PeriodicalIF":3.6,"publicationDate":"2025-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143942716","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Effective nitrification and removal of particulate organic matter in biofilters with compressible polyurethane foam","authors":"Bence Dániel Kovács,&nbsp;Kim João de Jesus Gregersen,&nbsp;Lars-Flemming Pedersen","doi":"10.1016/j.aquaeng.2025.102563","DOIUrl":"10.1016/j.aquaeng.2025.102563","url":null,"abstract":"<div><div>This study evaluated the treatment efficiency and backwashing capabilities of a novel fixed-bed biofilter utilizing compressible polyurethane foam media. The novelty of the biofilter system lies in its custom-designed backwashing mechanism, in which the foam media is mechanically compressed to remove retained solids. The system was tested at two hydraulic retention times (HRTs), 2.4 min (low HRT) and 16 min (high HRT), within a pilot-scale freshwater recirculating aquaculture system (RAS) stocked with rainbow trout over an 18-week period. Assessment included single-pass removal efficiencies and volume-specific removal rates based on inlet–outlet concentrations, as well as particulate matter retention rates derived from sludge collected during backwashing. Results demonstrated immediate treatment effectiveness, operational simplicity, and practical applicability. High HRT reactors consistently achieved higher single-pass removal efficiencies, with peak reductions of 90 % for total ammonia nitrogen (TAN), 66 % for nitrite-nitrogen (NO₂-N), 64 % (Bactiquant) and 48 % (H₂O₂ degradation assay) for bacterial activity, 44 % for turbidity, and a 10 percentage point improvement in ultraviolet transmittance (UVT) — approximately 1.5–3.5 times higher than values observed in low HRT reactors. Conversely, low HRT reactors outperformed in terms of volume-specific TAN and NO₂-N removal rates (67.9 ± 18.6 g TAN/m³/day; 22.5 ± 12.5 g NO₂-N/m³/day), due to increased substrate loading. Low HRT reactors also retained more particulate organic matter in sludge, with average retention rates of 38.2 ± 1.1 g/week for dry matter and 25.7 ± 1.2 g/week for total chemical oxygen demand (COD), compared to 24.8 ± 0.7 g/week and 7.7 ± 0.7 g/week under high HRT, respectively. Backwashing enhanced nitrification without compromising the system stability, however, a temporary reduction in NO₂-N removal was observed following a four-week operational pause.</div><div>Overall, this type of fixed-bed biofilter concept demonstrates strong potential for aquaculture applications, offering rapid effectiveness, operational resilience, and modular flexibility for continuous or intermittent treatment of nitrogenous compounds and particulate matter.</div></div>","PeriodicalId":8120,"journal":{"name":"Aquacultural Engineering","volume":"111 ","pages":"Article 102563"},"PeriodicalIF":3.6,"publicationDate":"2025-05-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144131406","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Energy optimization in large-scale recirculating aquaculture systems: Implementation and performance analysis of a hybrid deep learning approach","authors":"Ashwaq M. Alnemari ,&nbsp;Wael M. Elmessery ,&nbsp;Farahat S. Moghanm ,&nbsp;Víctor Espinosa ,&nbsp;Mahmoud Y. Shams ,&nbsp;Abdallah Elshawadfy Elwakeel ,&nbsp;Omar Saeed ,&nbsp;Mohamed Hamdy Eid ,&nbsp;Sadeq K. Alhag ,&nbsp;Laila A. Al-Shuraym ,&nbsp;Lamya Ahmed Alkeridis ,&nbsp;A.E. El-Namas","doi":"10.1016/j.aquaeng.2025.102561","DOIUrl":"10.1016/j.aquaeng.2025.102561","url":null,"abstract":"<div><div>Recirculating Aquaculture Systems (RAS) represent an increasingly important solution for sustainable fish production, yet their high energy consumption remains a significant operational challenge. This study extends our previous work on using Deep Deterministic Policy Gradient (DDPG) for optimizing feeding rates in Recirculating Aquaculture Systems (RAS) by developing a hybrid Long Short-Term Memory (LSTM)-DDPG approach for energy optimization in a large-scale commercial RAS facility. The system, comprising 108 tanks with a total water volume of 3132 m³ , was monitored over a complete annual cycle, collecting 8760 hourly observations of environmental, biological, and operational parameters. The hybrid model achieved high predictive accuracy for energy consumption patterns, with R² values exceeding 0.91 for key components. Implementation resulted in a 15–20 % reduction in daily energy consumption while maintaining optimal water quality. Economic analysis revealed a 17 % decrease in energy costs per kilogram of fish production. The system's performance was validated under varying fish biomass densities (80–120 kg/m³) and seasonal temperature profiles. These findings demonstrate the effectiveness of integrating deep learning techniques for energy optimization in RAS, offering a scalable solution for enhancing the economic and environmental sustainability of intensive aquaculture operations.</div></div>","PeriodicalId":8120,"journal":{"name":"Aquacultural Engineering","volume":"111 ","pages":"Article 102561"},"PeriodicalIF":3.6,"publicationDate":"2025-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143929045","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Exploring new frontiers: Current status and future research directions for AIoT application in shrimp farming in the Vietnamese Mekong delta","authors":"Tan Duy Le ,&nbsp;Huynh Phuong Thanh Nguyen ,&nbsp;Minh Tu Nguyen ,&nbsp;Ba Nhat Minh Le ,&nbsp;Kim Khoi Dang ,&nbsp;Phuc Quang Ha ,&nbsp;Tan Viet Tuyen Nguyen ,&nbsp;Hong Quan Nguyen","doi":"10.1016/j.aquaeng.2025.102559","DOIUrl":"10.1016/j.aquaeng.2025.102559","url":null,"abstract":"<div><div>With the proliferation of Internet of Things (IoT) devices for data sensing, communication, collection, exchange, and, accordingly, a huge amount of data being generated, the emerging artificial intelligence (AI) stands out as an excellent tool to provide learning capabilities for those interconnected devices. Together with high-speed mobile networks and big data, the mixture of AI and IoT, namely Artificial Intelligence of Things (AIoT), enables data analytics to optimize and enhance the performance of IoT systems. AIoT can potentially transform many aspects of human activities, especially agriculture applications. Shrimp farming, an essential sector of the aquacultural industry that provides a significant source of income and food for many communities worldwide, is expected to benefit most from AIoT. It is noticed that traditional shrimp farming methods are often labor-intensive and environmentally damaging. By integrating AIoT into the shrimp farming process, significant improvements can be achieved across various domains, including monitoring, disease prevention, feeding optimization, and sustainability. This study aims to serve as a comprehensive literature survey and a fieldwork carried out in the Vietnamese Mekong Delta (VMD). We explore the promising application of AIoT, its drivers, and barriers in shrimp farming globally, and specifically in the VMD. Our findings indicated that although the adoption of AIoT in this domain is still limited, the IoT technology has been widely used for monitoring and managing shrimp farming systems. This includes tracking essential environmental parameters such as temperature, pH, dissolved oxygen, and gas emissions. Furthermore, automatic control systems have been implemented to ensure optimal shrimp growth and survival of the shrimps. Those results were verified through interviews with local authorities and shrimp farmers. Despite discrepancies in the perception and level of promising AIoT applications, efforts have been made by shrimp farmers to implement basic IoT systems for environmental monitoring and farm management towards optimizing farming time and lowering labour demand. However, the application of continuous environmental monitoring and reporting using AI technologies is still limited. Owing to the advantages of learning capability and data analytics, AI integration into IoT for shrimp farming can substantially enhance the efficiency, sustainability, and cost-effectiveness while lowering labour demand and environmental impacts. Further research is, therefore, necessary to reach the full potential of AIoT in other critical areas of shrimp farming, such as disease detection and prevention, as well as supporting traceability and food safety monitoring in the whole production chain.</div></div>","PeriodicalId":8120,"journal":{"name":"Aquacultural Engineering","volume":"111 ","pages":"Article 102559"},"PeriodicalIF":3.6,"publicationDate":"2025-05-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143937558","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The impact of indoor biofloc-based system on water quality, growth, and disease resistance of black tiger shrimp","authors":"Leobert D. de la Peña ,&nbsp;Erish G. Estante-Superio ,&nbsp;Therese Marie M. Geanga ,&nbsp;Jose Louis A. Castellano ,&nbsp;Christian P. Cordero ,&nbsp;Sharmen C. Berlin ,&nbsp;Carlo C. Lazado","doi":"10.1016/j.aquaeng.2025.102564","DOIUrl":"10.1016/j.aquaeng.2025.102564","url":null,"abstract":"<div><div>The present study investigates the efficiency of an indoor biofloc-based system for the intensive culture of black tiger shrimp (<em>Penaeus monodon)</em>. Water quality, growth performance, and disease resistance of black tiger shrimp (0.35 ± 0.07 g) were evaluated after 90 days of rearing in a zero-water exchange system. Shrimp were stocked at 180 individuals/m³ in nine 5-ton concrete tanks under three treatments: biofloc with molasses (M), biofloc with wheat flour (WF), and a control group reared in a conventional clear water (CW) flow-through system. The WF treatment resulted in significantly higher final weight and improved FCR compared to CW (<em>p</em> &lt; 0.05). Survival was also significantly higher in both biofloc-based treatments despite elevated concentrations of total ammonia nitrogen, nitrite-nitrogen, and nitrate-nitrogen levels in the WF group (<em>p</em> &lt; 0.05). Additionally, Kaplan–Meier survival analysis indicated that shrimp reared and challenged in the biofloc system exhibited better resistance to <em>Vibrio parahaemolyticus</em> (VP_AHPND1212), as evidenced by higher survival rates. Although total hemocyte count, prophenoloxidase, and respiratory burst activities were not significantly different from the control group (<em>p</em> &gt; 0.05), they were consistently elevated in the biofloc-based treatments. Overall, biofloc-based treatments, specifically WF, proved effective in biofloc formation, contributing to improved water quality, enhanced immune response and disease resistance, and superior growth performance in <em>P. monodon</em>. This study highlights the potential of indoor biofloc systems as a sustainable strategy for intensive <em>P. monodon</em> culture in indoor biofloc tanks.</div></div>","PeriodicalId":8120,"journal":{"name":"Aquacultural Engineering","volume":"111 ","pages":"Article 102564"},"PeriodicalIF":3.6,"publicationDate":"2025-05-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143935836","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development of a novel reference feed intake model for Atlantic salmon (Salmo salar) based on temperature and body weight","authors":"Marina Linhares Azevedo ,&nbsp;Tomé Silva ,&nbsp;Filipe Soares ,&nbsp;Sergey Budaev ,&nbsp;Luis E.C. Conceição ,&nbsp;Ivar Rønnestad","doi":"10.1016/j.aquaeng.2025.102562","DOIUrl":"10.1016/j.aquaeng.2025.102562","url":null,"abstract":"<div><div>Precision aquaculture requires the use of advanced technologies to optimize fish management. Substantial progress has been achieved in modeling Atlantic salmon (<em>Salmon salar</em>) growth and feeding behavior. However, there is still no simple mathematical model to predict feed intake that can be used for reference and benchmarking. This study aims to infer basic parametrized equations for Atlantic salmon feed intake using a minimal number of predictors: body weight and temperature. We used data from 64 previous publications and 25 commercial feeding tables. Various mathematical models were tested, incorporating different temperature functions and fitting methods. The following model provided the most accurate predictions under a wider range of temperatures and fish body weights:</div><div><span><math><mrow><mrow><mi>FI</mi></mrow><mo>=</mo><mn>0.006</mn><mo>×</mo><msup><mrow><mi>BW</mi></mrow><mrow><mn>0.80</mn></mrow></msup><mo>×</mo><msup><mrow><mi>e</mi></mrow><mrow><mo>(</mo><mrow><mn>0.287</mn><mspace></mspace><mo>×</mo><mspace></mspace><mrow><mi>temperature</mi></mrow><mo>−</mo><mn>0.012</mn><mspace></mspace><mo>×</mo><mspace></mspace><msup><mrow><mi>temperature</mi></mrow><mrow><mn>2</mn></mrow></msup></mrow><mo>)</mo></mrow></msup></mrow></math></span></div><div>where FI is the feed intake (g/fish/day), BW the average body weight of fish (g) and temperature the water temperature (°C). Using simple least squares and robust fitting methods yields better prediction capacities, while data from commercial feeding tables does not significantly enhance model accuracy. This basic reference model developed on this study can be readily used as an applied tool, e.g. estimating the feed amount required for production or trials, determining baseline feed intake when building more complex models accounting for other factors, developing growth models taking feed intake as input, evaluating current feeding practices. Its simplicity, adaptability, and broad applicability represent a valuable contribution to the field, providing a practical foundation for future model development and decision-making in Atlantic salmon farming management.</div></div>","PeriodicalId":8120,"journal":{"name":"Aquacultural Engineering","volume":"111 ","pages":"Article 102562"},"PeriodicalIF":3.6,"publicationDate":"2025-05-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143935837","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"PUFFER-DETR: Tiger puffer similar abnormal behavior recognition based on transformer","authors":"Yixi Zhang ,&nbsp;Zeyuan Hu ,&nbsp;Jihang Liu ,&nbsp;Yinjia Li ,&nbsp;Jianjian Lin ,&nbsp;Yue Wang ,&nbsp;Hong Yu","doi":"10.1016/j.aquaeng.2025.102552","DOIUrl":"10.1016/j.aquaeng.2025.102552","url":null,"abstract":"<div><div>Fish behavior monitoring is crucial for fish farmers to obtain growth information, improve aquatic product quality, and adjust aquaculture strategies. However, the small size, severe occlusion, and similar behavior of fish pose challenges for identifying abnormal behavior. Therefore, this study proposes an abnormal behavior detection method based on PUFFER-DETR. Using the TripletAttention backbone network, the ability of the model to extract features of fish swarm behavior and individual fish behavior in turbid water has been improved. Furthermore, weight calculation is performed on the similar behavioral characteristics between individual fish and the behavioral characteristics of the fish groups to obtain a relationship feature map of similar behavior. Cross-scale feature fusion is performed using SHS-FPN, and the similarity behavior features of individual fish are adjusted to avoid the loss of similarity behavior features during the feature fusion process. Experimental results indicate that PUFFER-DETR achieved the best fusion accuracy at a speed of 127.9 frames per second, with an average accuracy of 92.8 %. This method can accurately detect abnormal behavior of fish and assist aquaculture personnel in judging the growth status of fish. Source code is available at <span><span>https://github.com/DLOU-FishBehavior/PUFFER-DETR</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":8120,"journal":{"name":"Aquacultural Engineering","volume":"111 ","pages":"Article 102552"},"PeriodicalIF":3.6,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143937518","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}