International Journal of Aquaculture, 2024, Vol.14, No.4, 195-210 http://www.aquapublisher.com/index.php/ija 205 context of climate change adaptation to aquaculture and fisheries. A 3-D environment is created to analyze mass fishes swimming along different directions and would use sensors to have a good head count. Rowan (2023) noted that alongside the development of these digital tools (Gorbunova, 2020), this could be simultaneously synchronized to provide a real-time analysis (Saberioon et al., 2016), and while using machine learning and artificial intelligence, it could easily manage large amounts of data. Machine learning algorithms could predict fish populations, identify fishing patterns, and detect anomalies in ocean data (Er-Rousse and Qafas, 2024). Data visualization tools could help visualize complex data, identify trends, and inform decision-making. Predictive modelling may forecast fish migrations and optimize fishing operations. Rowan (2023) described that aquaculture production in Central Luzon and Calabarzon in the Philippines had increased consistently, indicating successful operations. However, some locations experienced production variations, requiring targeted interventions. Machine learning techniques, such as Linear Regression, Support Vector Machine, and Multi-Layer Perceptron, could improve forecasting accuracy, aiding in long-term fisheries management. Likewise, Saha et al. (2018) focused on the aspect of the Internet of Things which could be used to monitor the inland aquaculture system at a low cost. The IoT would involve capturing images (Sung et al., 2014.) and performing processing to control heaters, aeration facilities, actuators or feeding systems through the physiological component of the breeds. Since this is a very cheap and easily understandable technology, the farmers and fish breeders may have sufficient control over their production. A fish farm in Norway adopted IoT sensors to monitor water quality, track fish growth, and detect disease outbreaks. The sensors provided real-time data, enabling the farm to optimize feeding, reduce waste, and improve fish welfare (Zhang and Gui, 2023). As a result, the farm increased its production by 20% and reduced its environmental impact by 15%. Technological breakthroughs have been noticed in the use of submersible aquatic farming cages such as Aquapods (Connolly, 2018), vertical aquaponics and genetics modifications. Likewise, Wei et al. (2020) described the fact that using open-sea close cages could be a possible solution to maintain a mode of marine aquaculture. The use of different sensors such as water quality sensors, intelligent monitoring sensors (Ibrahin et al., 2017) and automatic level sensors have been described by the author as devices that could be used in the open sea cage aquaculture system. Nevertheless, the author again noticed that there had been problems with the wirings which could not transmit the required information effectively. Due to failed data conveyance, there had been less glitch in real-time monitoring of the water quality. The use of machine vision, camera sensors and sonar technologies failed to provide adequate information due to an unknown seabed environment and the limitations of the vision sensors (Zhou et al., 2018). Another digital challenge depicted by Rowan (2023) is the use of blockchain in the fishery sector where a safe model of the fish business model can be mapped reducing frauds and improving traceability of the farmed fish to fork while avoiding wastages and disease proliferation. Mobile apps and online platforms could facilitate real-time communication among fishermen, fisheries managers and researchers. Virtual reality could enhance training and education. Autonomous underwater vehicles would monitor ocean health, fish populations and pollution. Zhao et al. (2021) reviewed the use of machine learning for the aquaculture system, as compared to the traditional machine system, the development of deep learning and neural network systems have a high scope in breeding intelligently (Yang et al., 2020) concerning the climate differences. The machine learning system would add the component of data processing, information manipulation, real-time monitoring and decision-making process to facilitate the production of fisheries products. Nevertheless, the high complexity of this system together with difficulty in conjugation climate modelling, geographical mapping and biological control, proved that this system is not reliable for maintaining climate-smart production. Connolly (2018), described extensively the use of sensors that could be used to regulate the physiological environment of farmed fish through the regulation of water pH level, monitoring salinity of water, control of oxygen and carbon dioxide levels, and handling turbidity and pollutants. The author also emphasized that through
RkJQdWJsaXNoZXIy MjQ4ODYzNA==