Leaf Area Index (LAI) is the total leaf area per unit of land surface area and is a crucial parameter for assessing vegetation growth and productivity. Machine learning regression algorithms are widely applied for LAI estimation. Due to spectral response variations among sensors and susceptibility of mangrove-derived variables to environmental noise suppression, obtaining sensitivity indices and optimal machine learning regression models is essential for retrieving mangrove LAI at the population scale. This study proposes a novel approach to processing and retrieving mangrove LAI data by integrating multispectral indices with machine learning methods. Box–Cox transformation and CatBoost-based feature selection were employed to obtain the optimal dataset. Random Forest (RF), Gradient Boosting Regression Trees (GBRT), and Categorical Boosting (CatBoost) algorithms were used to evaluate the accuracy of LAI retrieval from Unmanned Aerial Vehicle (UAV) and Gaofen-6 (GF-6) data. Results indicate that when LAI > 3, LAI does not immediately saturate as CVI, MTVI 2, and other indices increase, demonstrating higher sensitivity. UAV data outperformed GF-6 data in retrieving LAI for diverse mangrove populations; during model training, RF proved more suitable for small-sample datasets, while CatBoost effectively suppressed environmental noise. Both RF and CatBoost demonstrated higher robustness in estimating Avicennia marina (AM) (RF: R² = 0.704) and Aegiceras corniculatum (AC) (R² = 0.766), respectively. Spatial distribution analysis of LAI indicates that healthy AM and AC cover 85.36% and 96.67% of the area, respectively. Spartina alterniflora and aquaculture wastewater may be among the factors affecting the health of mangrove forests in the study area. LAI retrieval holds significant importance for mangrove health monitoring and risk early warning.

Candidatus Hepatincolaceae is a poorly characterized family of obligate Alphaproteobacterial symbionts that are widely detected in ecdysozoans. They were previously assumed to play a nutrient-scavenging role in the gut lumen. In this study, two high-quality metagenome-assembled genomes (MAGs, 1.39 Mb and 1.48 Mb in size) were recovered from the gut of the snapping shrimp Alpheus brevicristatus via metagenomic sequencing. Phylogenetic and whole-genome similarity analyses confirm that these two MAGs represent two novel, undescribed genera within the family Ca. Hepatincolaceae. Metabolic reconstruction reveals that they not only retain the canonical nutrient-scavenging pathways conserved across all Hepatincolaceae members, but also encode previously undocumented functional modules for antioxidant defense, vitamin B₁ and B₂ biosynthesis, and short-chain fatty acid production. They maintain a high oxygen-affinity cytochrome bd terminal oxidase to thrive in the anoxic gut microenvironment. Consistent with their symbiotic lifestyle, their genomes exhibit typical signatures of reductive evolution, such as reduced genome size, low GC content, and gene loss in amino acid and nucleotide de novo biosynthesis pathways. This study presents the first reported high-quality genomes of marine Ca. Hepatincolaceae symbionts, which are predicted to possess multiple metabolic functions extending beyond nutritional mutualism.

Effective coastal management of seagrass systems requires identifying both extant meadows and degraded areas suitable for recovery. However, these decisions often rely on distribution maps that provide limited information on ecological condition or carbon potential. Here, we develop an ensemble habitat modeling (EHM) framework for seagrass meadows around Guanglu Island, China, to test whether a habitat suitability index (HSI) can serve as a proxy for ecological performance and carbon storage potential. By integrating multiple single-algorithm predictions, the ensemble model reduced algorithm-dependent variation and achieved strong predictive performance (ROC = 0.967, TSS = 0.876), with predictions capturing 92.59% to 98.98% of the 2023 field-mapped meadow area. HSI was positively associated with aboveground biomass and shoot density. A predictive model fitted directly with biomass as the response showed that aboveground biomass increased with HSI, with a leave-one-out cross-validated RMSE of 77.68 g DW m⁻². High HSI samples also exhibited smaller relative deviations from the theoretical biomass–density envelope and a greater realized fraction of theoretical biomass potential. Modeled sediment carbon densities were consistent with the 2023 sediment comparison, thereby supporting conditional carbon stock screening. Our findings support the use of ensemble-derived HSI as a local spatial proxy for biomass-related ecological potential, rather than as a direct measure of habitat quality, actual biomass, or realized carbon stock. By separating predicted suitable habitat, observed meadow distribution, and model-based potential carbon stock, the framework provides a spatial screening tool for conservation prioritization, restoration opportunity assessment, and blue carbon planning under explicitly stated ecological and modeling assumptions.