Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Sep 22;3(1):101.
doi: 10.1038/s43705-023-00308-7.

Predicting global distributions of eukaryotic plankton communities from satellite data

Hiroto Kaneko  1 , Hisashi Endo  1 , Nicolas Henry  2   3 , Cédric Berney  2   4 , Frédéric Mahé  5   6 , Julie Poulain  7 , Karine Labadie  8 , Odette Beluche  8 , Roy El Hourany  9   10 ; Tara Oceans Coordinators; Samuel Chaffron  3   11 , Patrick Wincker  7 , Ryosuke Nakamura  12 , Lee Karp-Boss  13 , Emmanuel Boss  13 , Chris Bowler  3   10 , Colomban de Vargas  2   4 , Kentaro Tomii  14 , Hiroyuki Ogata  15
Collaborators, Affiliations

Predicting global distributions of eukaryotic plankton communities from satellite data

Hiroto Kaneko et al. ISME Commun. .

Abstract

Satellite remote sensing is a powerful tool to monitor the global dynamics of marine plankton. Previous research has focused on developing models to predict the size or taxonomic groups of phytoplankton. Here, we present an approach to identify community types from a global plankton network that includes phytoplankton and heterotrophic protists and to predict their biogeography using global satellite observations. Six plankton community types were identified from a co-occurrence network inferred using a novel rDNA 18 S V4 planetary-scale eukaryotic metabarcoding dataset. Machine learning techniques were then applied to construct a model that predicted these community types from satellite data. The model showed an overall 67% accuracy in the prediction of the community types. The prediction using 17 satellite-derived parameters showed better performance than that using only temperature and/or the concentration of chlorophyll a. The constructed model predicted the global spatiotemporal distribution of community types over 19 years. The predicted distributions exhibited strong seasonal changes in community types in the subarctic-subtropical boundary regions, which were consistent with previous field observations. The model also identified the long-term trends in the distribution of community types, which suggested responses to ocean warming.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1. Two-dimensional map of satellite-derived parameter space.
Points associated with metabarcoding samples used to train predictive models are projected on the parameter space map (large green points). Small points are randomly selected grid cells, which were used to train a UMAP projection, colored by the Longhurst biomes (see Fig. S7).
Fig. 2
Fig. 2. Plankton network inferred using metabarcoding data.
A Force-directed representation of the network. Nodes (plankton OTUs) are colored by the module they belong to. B Connections between modules in the network. The edge width is proportional to the number of inter-module edges.
Fig. 3
Fig. 3. Taxonomic breakdown of modules in the plankton network.
The breakdown of taxa annotated to OTUs belonging to each module. The taxonomic level is “taxogroup 2” in the EukRibo.
Fig. 4
Fig. 4. Assigned community types of samples.
A Heatmap of the edge satisfaction index. The rows are samples ordered by their latitude and the columns are modules. The leftmost column shows the community type of each sample by color. Community types were assigned using the module with the highest edge satisfaction index. B Geographic distribution of community types. The community type assigned for each sample is shown in the color of the sampling site on the map.
Fig. 5
Fig. 5. Performance of Support Vector Machine (SVM) on community type prediction using satellite-derived parameters.
Performance of SVM using all 17 satellite-derived parameters. A, B The confusion matrix (A) and the ROC curve (B) in the leave-one-out cross-validation. C, D The confusion matrix (C) and the ROC curve (D) in the buffered cross-validation.
Fig. 6
Fig. 6. Spatiotemporal distribution of community types predicted from satellite-derived parameters.
Community type distribution in February (A), May (B), August (C), and November (D), 2021, predicted from satellite-derived parameters. When multiple community types were predicted to the same point, the community type with the highest probability is shown in transparent color. Gray points mean that no community type was predicted.
See this image and copyright information in PMC

References

    1. Ibarbalz FM, Henry N, Brandão MC, Martini S, Busseni G, Byrne H, et al. Global trends in marine plankton diversity across kingdoms of life. Cell. 2019;179:1084–97. - PMC - PubMed
    1. Mouw CB, Hardman-Mountford NJ, Alvain S, Bracher A, Brewin RJW, Bricaud A, et al. A consumer’s guide to satellite remote sensing of multiple phytoplankton groups in the global ocean. Front Mar Sci. 2017;4:41.
    1. Bracher A, Bouman HA, Brewin RJW, Bricaud A, Brotas V, Ciotti AM, et al. Obtaining phytoplankton diversity from ocean color: a scientific roadmap for future development. Front Mar Sci. 2017;4:55.
    1. Mirkovic T, Ostroumov EE, Anna JM, van Grondelle R, Govindjee, Scholes GD. Light absorption and energy transfer in the antenna complexes of photosynthetic organisms. Chem Rev. 2017;117:249–93. - PubMed
    1. Hirata T, Hardman-Mountford NJ, Brewin RJW, Aiken J, Barlow R, Suzuki K, et al. Synoptic relationships between surface Chlorophyll- a and diagnostic pigments specific to phytoplankton functional types. Biogeosciences. 2011;8:311–27.