Search

Abstract To address the issues of high demand for efficiently recognizing fish species in marine scientific research, such as impact assessments on biodiversity and monitoring, an automated hierarchical image classification web‐based platform, named FishAI, was developed. Trained with marine fish images collected from the World Register of Marine Species, FishAI used the Vision Transformer (ViT) model, to classify fish. The model considers hierarchy levels, covering 3 classes, 38 orders, 154 families, 438 genera, and 808 species. The FishAI achieved accuracies of 0.975 (Class), 0.798 (Order), 0.743 (Family), 0.638 (Genus), and 0.626 (Species) on test images, respectively, by using the hyperparameter optimization. Comparison between ViT and other baseline backbones proves its superiority by capturing long‐distance dependency. In addition, FishAI yields the top‐5 prediction accuracies of 1.000 (Class), 0.887 (Order), 0.816 (Family), 0.729 (Genus), and 0.727 (Species), respectively. In order to further enhance the practicality of FishAI, the user‐friendly graphic interface (http://www.csbio.sjtu.edu.cn/bioinf/FishAI/) facilitates its easy‐to‐use application. Furthermore, interpretability analysis by Grad‐CAM provides a visual explanation of the highlighted regions on the images for FishAI's prediction among different hierarchies.

Mudflats and sandflats are two common types of coastal tidal flats, the structure and function differences of microbial communities between them are still underappreciated. Beibu Gulf is a diurnal tidal regime located in China, the differences between the two type of tidal flats could be more distinct. In this study, we collected a total of 6 samples from Beibu Gulf, consisting of 3 sandflats samples and 3 mudflats samples, classified based on clay and silt content. Generally, the mudflats samples exhibited higher levels of NH4+_N and TOC, but lower in ORP and pH. The microbial diversity of the two types of tidal flats was investigated, revealing great differences existed and sandflats had higher microbial richness and diversity than mudflats. Furthermore, we analyzed the association between microbial communities and environmental factors, finding NH4+_N to have the highest contribution to the total variation in microbial community structure, and microbial groups such as Desulfobacterota, Campilobacterota, Chloroflexota, Calditrichota, Spirochaetota, Zixibacteria, Latescibacterota and Sva0485 group in mudflats were positively associated with NH4+_N. The functions of microbial community were predicted using metagenomic sequences and metagenome assembled genome (MAG). Mudflats contained more genes for carbon fixation. Nitrate and nitrite reduction were widely existed in mudflats and sandflats, but nitrogen fixation was only existed in mudflats, and Campilobacterota, Desulfobacterota and Gammaproteobacteria MAGs were mainly responsible for it. Sandflats composed more genes for ammonium oxidation, but no MAG was found whether in sandflats or mudflats. Microbes in mudflats exhibited a greater abundance of genes related to sulfur cycling, especially in reduction process, unique MAGs in mudflats such as Calditrichota, Chloroflexota, Desulfobacterota and Zixibacteria MAGs are responsible for sulfate and sulfite reduction. Finally, we predicted functions of ammonium related microbes in mudflats based on MAGs and found Campilobacterota and Desulfobacterota MAGs were important for high accumulation of ammonium in mudflats. This study illuminated the structural and functional differences of microbial communities in mudflats and sandflats, providing new insights into the relationship of microbial communities and environment in the tidal flat.

Special geographical location and abundant organic matter profiles in tidal flats have resulted in great microbial diversity, in which Bacteroidota strains are considered as one of the primary degraders of polysaccharides, playing a crucial role in the carbon cycle. In this study, we collected sediment or sand samples from 34 bare tidal flats in China and investigated the profile of culturable bacteria, selected three Bacteroidota for polyphasic taxonomic analysis and revealed their polysaccharide metabolic potential. Totally, we isolated 352 pure cultured bacteria and they mainly distributed in Bacteroidota, Pseudomonadota, Bacillota, and Actinomycetota. It is shown that the bare tidal flats contained a large number of potential novel species, mainly distributed in Flavobacteriales and Cytophagales within Bacteroidota. Three Bacteroidota strains, M17T, M82T, and M415T, isolated from mudflat were selected for polyphasic taxonomic analysis. The 16S rRNA gene sequence similarity between strain M17T and Mangrovivirga cuniculi KCTC 72349T was 99.28%, and less than 90.09% with other species; strain M82T shared the highest 16S rRNA gene sequence similarity of 97.85% with Pontibacter litorisediminis KCTC 52252T, and less than 97.43% with other species; strain M415T had higher 16S rRNA gene sequence similarities with type species of genera Eudoraea (92.62-93.68%), Zeaxanthinibacter (92.02-92.91%), and Muriicola (92.21-92.83%). Phylogenetic analysis based on 16S rRNA gene sequences and single-copy orthologous clusters showed that strains M17T and M82T represent novel species within the genus Mangrovivirga and Pontibacter, respectively, and strain M415T represents a novel species of a novel genus within the family Flavobacteriaceae. The potential in polysaccharide metabolism of all these three strains was analyzed by genomes. The analysis revealed that glycoside hydrolases and glycosyltransferases account for more than 70% of the total CAZymes. Additionally, the numbers of polysaccharide utilization loci (PULs) and annotated CAZymes in Cytophagales spp. M17T and M82T were found to be higher than those in Flavobacteriales sp. M415T. Highly specialized saccharolytic systems and the presence of numerous diversified CAZymes for obtaining energy through polysaccharide metabolism were speculated to help the three novel strains adapt to the utilization of both terrestrial and marine polysaccharides.

IntroductionIn response to the need for automated classification in global marine biological studies, deep learning is applied to image-based classification of marine echinoderms.MethodsImages of marine echinoderms are collected and classified according to their systematic taxonomy. The images belong to 5 classes, 38 orders, 145 families, 459 genera, and 1021 species, respectively. The deep learning model, EfficientNetV2, outperforms the competing model and is chosen for developing the automated classification tool, EchoAI. Then, the EfficientNetV2-based tool, EchoAI is applied to each taxonomic level.ResultsThe accuracy for the test dataset was 0.980 (class), 0.876 (order), 0.738 (family), 0.612 (genus), and 0.469 (species), respectively. Online prediction service is provided.DiscussionThe EchoAI model and results are facilitated for investigating the diversity, abundance and distribution of species at the global scale, and the methodological strategy can also be applied to image classification of other categories of marine organisms, which is of great significance for global marine studies. EchoAI is freely available at http://www.csbio.sjtu.edu.cn/bioinf/EchoAI/ for academic use.

The members of the phylum Bacteroidetes are recognized as some of the most important specialists for the degradation of polysaccharides. However, in contrast to research on Bacteroidetes in the human gut, research on polysaccharide degradation by marine Bacteroidetes is still rare. The genus Algibacter belongs to the Flavobacteriaceae family of the Bacteroidetes, and most species in this genus are isolated from or near the habitat of algae, indicating a preference for the complex polysaccharides of algae. In this work, a novel brown-seaweed-degrading strain designated HZ22 was isolated from the surface of a brown seaweed (Laminaria japonica). On the basis of its physiological, chemotaxonomic, and genotypic characteristics, it is proposed that strain HZ22 represents a novel species in the genus Algibacter with the proposed name Algibacter alginolytica sp. nov. The genome of strain HZ22, the type strain of this species, harbors 3,371 coding sequences (CDSs) and 255 carbohydrate-active enzymes (CAZymes), including 104 glycoside hydrolases (GHs) and 18 polysaccharide lyases (PLs); this appears to be the highest proportion of CAZymes (∼7.5%) among the reported strains in the class Flavobacteria. Seventeen polysaccharide utilization loci (PUL) are predicted to be specific for marine polysaccharides, especially algal polysaccharides from red, green, and brown seaweeds. In particular, PUL N is predicted to be specific for alginate. Taking these findings together with the results of assays of crude alginate lyases, we prove that strain HZ22T can completely degrade alginate. This work reveals that strain HZ22T has good potential for the degradation of algal polysaccharides and that the structure and related mechanism of PUL in strain HZ22T are worth further research. [ABSTRACT FROM AUTHOR]