The bacterial community in sediment is sensitive to artificial disturbance, and they respond differently to human disturbance, such as changing the nutrient cycling and energy flow in marine ecosystems. However, little is known about the dynamics and distribution of bacterial community structures in marine sediments and potential biogeochemical functions during the long-time succession in marine ranching. In the present study, we compared the dynamics of the bacterial composition and potential biogeochemical functions of sediment to ten years (TR) and one-year new artificial reef (NR) areas using metagenomic next-generation sequencing technology. Results revealed that NR reduces the Pielou’s evenness and Shannon index. Similarly, nonmetric multidimensional scaling showed that the beta diversity of sediment bacterial communities in NR significantly differed between TR and non-artificial reef areas. Previously, TR biomarkers were frequently associated with organic matter decomposing and assimilating in the organically enriched sediments (i.e., Acinetobacter). The soluble reactive phosphate (SRP) and total phosphorus (TP) concentrations were thought to be the primary driving forces in shaping the microbial community in sediment. Pseudomonas, Lactobacillus, and Ralstonia have a significant positive correlation with SRP, TP, nitrate, and TN, but a negative association with pH, Salinity, Hg, and depth. NR was found to have more negative correlation nodes, indicating that taxa face more competition or predation press. Vibrio served as the module-hubs in the network in all areas. In addition, chemoheterotrophy, aerobic chemoheterotrophy, and fermentation were the three most prominent functions of the three areas, accounting for 59.96% of the relative abundance of the functional annotation. Different bacteria in sediments may change the amount of biogeochemical cycle in the marine ranching ecosystem. These findings can increase our understanding of the succession of the microecosystem for the marine ranching sedimentary environment by revealing how artificial reefs affect the indigenous sediment bacterial community and their responses to environmental variation.

The selection of environmental variables with different spatial resolutions is a critical factor affecting the accuracy of machine learning-based fishery forecasting. In this study, spring-season survey data of Decapterus macarellus in the South China Sea from 2016 to 2024 were used to construct six machine learning models—decision tree (DT), extra trees (ETs), K-Nearest Neighbors (KNN), light gradient boosting machine (LGBM), random forest (RF), and extreme gradient boosting (XGB)—based on seven environmental variables (e.g., sea surface temperature (SST), chlorophyll-a concentration (CHL)) at four spatial resolutions (0.083°, 0.25°, 0.5°, and 1°), filtered using Pearson correlation analysis. Optimal models were selected under each resolution through performance comparison. SHapley Additive exPlanations (SHAP) values were employed to interpret the contribution of environmental predictors, and the maximum entropy (MaxEnt) model was used to perform habitat suitability mapping. Results showed that the XGB model at 0.083° resolution achieved the best performance, with the area under the receiver operating characteristic curve (ROC_AUC) = 0.836, accuracy = 0.793, and negative predictive value = 0.862, outperforming models at coarser resolutions. CHL was identified as the most influential variable, showing high importance in both the SHAP distribution and the cumulative area under the curve contribution. Predicted suitable habitats were mainly located in the northern and central-southern South China Sea, with the latter covering a broader area. This study is the first to systematically evaluate the impact of spatial resolution on environmental variable selection in machine learning models, integrating SHAP-based interpretability with MaxEnt modeling to achieve reliable habitat suitability prediction, offering valuable insights for fishery forecasting in the South China Sea.

This paper aims to explore the impact of marine ranching construction on water quality and fishery resources in the surrounding marine areas. Utilizing in situ water quality and fishery resource data collected before and after the establishment of marine ranching, the study analyzes changes in water quality parameters from both temporal and spatial perspectives. A quantitative evaluation of the water quality data is conducted using several models to assess the accuracy of different evaluation methods. By integrating the SHAP algorithm with physical significance, the study examines the differences between optically sensitive and non-optically sensitive water quality parameters during the machine learning evaluation process. Finally, based on the inverted water quality data, the potential impact range and resource output following the deployment of artificial reefs are investigated. The results indicate that in the marine area near Wailingding Island, Zhuhai, the deployment of artificial reefs with a volume of 38,048 cubic meters led to an increase in fishery resources by 318 kg/km² in spring and 660 kg/km² in autumn. Additionally, deployment had varying degrees of impact on the concentrations of chlorophyll a (Chla), dissolved oxygen (DO), chemical oxygen demand (COD), and phosphate (PO4-P) in the surface water within an approximate range of 10 km. This study provides a valuable reference for calculating input–output ratios, as well as for the management and evaluation of marine ranching.

Environmental physicochemical factors and microorganisms play critical roles in the health of oysters. However, the impact of high-density oyster farming—a highly efficient filter-feeding bivalve system—on environmental conditions and microbial community structure and function remains poorly understood. This study conducted four-season monitoring of the water and sediment parameters in a semi-enclosed bay commercial oyster aquaculture (OA) system and a control area (CT), coupled with 16S rRNA amplicon sequencing of the environmental microbiota. Oyster aquaculture caused negligible disruption to water column parameters but significantly increased the concentrations of total organic carbon (TOC, annual mean OA vs. CT:1.15% vs. 0.56%), sulfides (annual mean OA vs. CT:67.72 vs. 24.99 mg·kg⁻¹), and heavy metals (Cd, Pb, Cu, Zn, and Cr) in the sediment. α-diversity (Shannon and Chao indices) exhibited minimal overall perturbation, with significant inter-regional differences observed only in winter for both water and sediment. The bacterial community structure of the water column was significantly altered only in winter, whereas sediment communities showed structural shifts in spring, summer, and autumn. Water microbiota were primarily influenced by turbidity, dissolved oxygen, salinity, the Si/N ratio, and silicates. Sediment microbiota were correlated with Pb, Cu, Zn, TOC, Cr, and sediment particle size. Water bacterial functions displayed only four significantly divergent biogeochemical processes annually (sulfur compound respiration; OA vs. CT). In contrast, sediment bacteria exhibited 29 significantly disrupted functions annually, with the greatest seasonal divergence in winter (11/67 functions). Spring, summer, and autumn sediment functions showed distinct patterns. Understanding these environmental–microbial interactions is essential for sustainable oyster aquaculture and ecological optimization.

Overfishing and climate change have led to the depletion of fishery resources in the offshore South China Sea. The purpleback flying squid (Sthenoteuthis oualaniensis) has emerged as a promising alternative due to its ecological and economic value. However, information on its preferred habitat conditions remains scarce. This study integrates geostatistical and fisheries oceanographic approaches to explore optimal spatial–temporal scales for habitat modeling and to assess habitat changes under warming scenarios. Utilizing fishery data from 2013 to 2017, environmental variables including SST, sea surface temperature anomaly (SSTA), and chlorophyll-a concentration (CHL) were analyzed. Fishing effort data revealed significant seasonal differences, with the highest vessel numbers in summer and the lowest in autumn. Among the six modeling schemes, the combination of 0.5° × 0.5° spatial resolution and seasonal temporal resolution yielded the highest HSI model accuracy (84.02%). Optimal environmental ranges varied by season. Simulations of SST deviations (±0.2 °C, ±0.5 °C, and ±1 °C) showed that extreme warming or cooling could eliminate suitable habitats. These findings highlight the vulnerability of squid habitats to thermal shifts and support adaptive fishery strategies in the South China Sea.