Pacific sardine (Sardinops sagax) is a commercially important species and supports important fisheries in the Northwest Pacific Ocean (NPO). Understanding the habitat distribution patterns of Pacific sardine is of great significance for fishing ground prediction and stock management. In this study, both single-algorithm and ensemble distribution models were established through the Biomod2 package for Pacific sardine by combining the species occurrence data, sea surface temperature (SST), sea surface height (SSH), sea surface salinity (SSS) and chlorophyll-a concentration (Chla) in the NPO during the main fishing season (June–November) from 2015 to 2020. The results indicated that the key environmental variables affecting the habitat distribution of Pacific sardine were the SSH and SST. The suitable habitat area for Pacific sardine showed significant monthly changes: the suitable habitat range in June was larger than that in July and August, while the suitable habitat range gradually increased from September to November. Furthermore, the monthly geometric centers of habitat suitability index (HSI) for Pacific sardine presented a counterclockwise pattern, gradually moving to the northeast from June, and then turning back to the southwest from August. Compared with single-algorithm models, the ensemble model had higher evaluation metric values and better spatial correspondence between habitat prediction and occurrence records data, which indicated that the ensemble model can provide more accurate prediction and is a promising tool for potential habitat forecasting and resource management.

Understanding the spatial pattern of human fishing activity is very important for fisheries resource monitoring and spatial management. The environmental preferences of tropical tuna purse seine fleet in the Western and Central Pacific Ocean (WCPO) were constructed and compared at different spatial scales based on the fishing effort (FE) data from the available automatic identification system (AIS) and commercial fishery data compiled from the Western and Central Pacific Fisheries Commission (WCPFC), using maximum entropy (MaxEnt) methods. The MaxEnt models were fitted with FE and commercial fishery data and remote sensing environmental data. Our results showed that the area under the curve (AUC) value each month based on the commercial fishery data (1°) and FE at 0.25° and 0.5° spatial scales was greater than 0.8. The AUC values each month based on the FE data at a 1° scale ranged from 0.775 to 0.829. The AUC values based on commercial fishing data at the 1° scale were comparable to the model results based on FE data at the 0.5° scale and inferior to the model results based on FE data at the 0.25° scales. Overall, the sea surface temperature (SST), temperature at 100 metres (T100), oxygen concentration at 100 metres (O100) and total primary production (PP) had the greatest influence on the distribution of the purse seine tuna fleet. The oxygen concentration at 200 metres (O200), distance to shore (DSH), dissolved oxygen (Dox), EKE, mixed layer depth (Mld), sea surface salinity (SSS), salinity at 100 metres (S100) and salinity at 200 metres (S200) had moderate influences, and other environmental variables had little influence. The suitable habitat areas varied in response to environmental conditions. The purse seine tuna fleet was mostly present at locations where the SST, T100, O100, O200 and PP were 28–30 °C, 27–29 °C, 150–200 mmol/m³ and 5–10 mg/m⁻³, respectively. The MaxEnt models enable the integration of AIS data and high-resolution environmental data from satellite remote sensing to describe the spatiotemporal distribution of the tuna purse seine fishery and the influence of environmental variables on the distribution, and can provide forecasts for fishing ground distributions based on future remote sensing environmental data.

Artificial reefs are man-made structures submerged in the ocean, and the design of these structures plays a crucial role in determining their effectiveness. Precisely measuring the configuration of artificial reefs is vital for creating suitable habitats for marine organisms. This study presents a novel approach for automated detection of artificial reefs by recognizing their key features and key points. Two enhanced models, namely, YOLOv8n-PoseRFSA and YOLOv8n-PoseMSA, are introduced based on the YOLOv8n-Pose architecture. The YOLOv8n-PoseRFSA model exhibits a 2.3% increase in accuracy in pinpointing target key points compared to the baseline YOLOv8n-Pose model, showcasing notable enhancements in recall rate, mean average precision (mAP), and other evaluation metrics. In response to the demand for swift identification in mobile fishing scenarios, a YOLOv8n-PoseMSA model is proposed, leveraging MobileNetV3 to replace the backbone network structure. This model reduces the computational burden to 33% of the original model while preserving recognition accuracy and minimizing the accuracy drop. The methodology outlined in this research enables real-time monitoring of artificial reef deployments, allowing for the precise quantification of their structural characteristics, thereby significantly enhancing monitoring efficiency and convenience. By better assessing the layout of artificial reefs and their ecological impact, this approach offers valuable data support for the future planning and implementation of reef projects.

Fish body measurement is essential for monitoring fish farming and evaluating growth. Non-destructive underwater measurements play a significant role in aquaculture management. This study involved annotating images of fish in aquaculture settings and utilized a line laser for underwater distance calibration and fish body inclined-angle calculation. The YOLOv8 model was employed for fish identification and key-point detection, enabling the determination of actual body dimensions through a mathematical model. The results show a root-mean-square error of 6.8 pixels for underwater distance calibration using the line laser. The pre-training YOLOv8-n, with its lower parameter counts and higher MAP values, proved more effective for fish identification and key-point detection, considering speed and accuracy. Average body length measurements within 1.5 m of the camera showed a minor deviation of 2.46% compared to manual measurements. The average relative errors for body length and width were 2.46% and 5.11%, respectively, with corresponding average absolute errors. This study introduces innovative techniques for fish body measurement in aquaculture, promoting the digitization and informatization of aquaculture processes.