The Northwestern Mediterranean Sea is undergoing rapid environmental changes driven by climate variability and intense anthropogenic pressure. To monitor and understand the long-term impacts on marine ecosystems, the Mediterranean Ocean Observing System for the Environment (MOOSE) program combines multidisciplinary observations, including physical, chemical, and biological data across temporal and spatial scales. This study presents a holistic assessment of planktonic communities across the Northwestern Mediterranean using integrated approaches—environmental genomics and high-resolution imaging—spanning all plankton size fractions and depths. Data collected during three MOOSE-GE cruises in 2017, 2018, and 2019 were analyzed to explore plankton diversity patterns in relation to oceanographic features. Plankton assemblages were primarily shaped by organism size and water column depth, with fractions of 0.2–3 and 3–180 µm in the surface and deep chlorophyll maximum layers showing the highest alpha diversity. Fractions > 64 µm were dominated by metazoans, particularly Arthropoda, whereas size classes collected by Niskin bottles were dominated by protists such as Syndiniales and Rhizaria. Differences among cruises and sampling periods were detected in Niskin bottle samples, especially for diatoms and dinoflagellates, while plankton tow samples exhibited less pronounced temporal variability. Physical clustering of stations revealed clear cross-shelf and basin-scale gradients, which aligned more closely with community structure at fine taxonomic resolution (OTU level) for small plankton. Integrating imaging with environmental genomic data enhanced the characterization of key taxa like Copepoda and Rhizaria, demonstrating the complementary strengths of each method. While imaging provided quantitative data, environmental genomics captured cryptic and morphologically indistinct taxa, emphasizing the value of molecular approaches for microbial plankton. This study highlights the critical importance of combining high-resolution molecular and imaging tools with detailed environmental context to unravel plankton biodiversity patterns. It demonstrates that depth, size, and taxonomic resolution are key dimensions for understanding community structure over time. The MOOSE program proves effective for ecosystem-scale monitoring, providing an essential foundation for future assessments of biogeochemical processes and ecosystem responses to climate change and human-induced alterations in the Mediterranean Sea.

The iMagine project is an EU-funded initiative led by the EGI Foundation. One of the objectives of this project is to provide an AI platform that leverages AI-powered tools to improve the processing and analysis of imaging data from marine and freshwater ecosystems, contributing to the study of the health of oceans, seas, coasts, and inland waters. Connected to the European Open Science Cloud (EOSC), iMagine supports the development, training, and deployment of AI models by collaborating with twelve use cases across diverse aquatic science fields. This collaboration fosters valuable insights and accelerates scientific progress by refining existing solutions in data acquisition, preprocessing, and model deployment. The platform offers trained models as a service, integrating AI tools for image annotation, ensuring the creation of high-quality datasets that comply with FAIR principles. Through these methodologies, iMagine enhances consistency, enabling researchers to efficiently publish and share data in repositories. Beyond its AI tools, iMagine places a strong emphasis on deep learning models, such as convolutional neural networks, for tasks like image classification, object detection, and segmentation, tailored to the unique requirements of aquatic sciences. It also provides robust performance evaluation tools, including experiment tracking, while tackling challenges such as AI model drift and data biases to ensure research reproducibility and transparency. The platform enables users to develop, train, share, and deploy AI models within a flexible environment that integrates with federated cloud and high-performance computing infrastructures, using Docker containers for smooth execution. Additionally, iMagine fosters collaboration with projects like AI4EOSC and Blue-Cloud, and Research Infrastructures such as EMSO and SeaDataNet, expanding its impact on the scientific community. This paper summarizes the key lessons and best practices learned in the iMagine project through the full process of AI-based aquatic image analysis, from data preparation and annotation to model deployment and evaluation. The paper therefore helps aquatic scientists advance their AI-driven image analysis approaches.

Plankton was sampled with a Continuous Underway Fish Egg Sampler (CUFES, 315µm mesh size) at 4 m below the surface, and a WP2 net (200µm mesh size) from 100m to the surface, or 5 m above the sea floor to the surface when the depth was < 100 m, in the Bay of Biscay. The full images were processed with the ZooCAM software and the embedded Matrox Imaging Library (Colas et a., 2018) which generated regions of interest (ROIs) around each individual object and a set of features measured on the object. The same objects were re-processed to compute features with the scikit-image library http://scikit-image.org. The 1, 286, 590 resulting objects were sorted by a limited number of operators, following a common taxonomic guide, into 93 taxa, using the web application EcoTaxa http://ecotaxa.obs-vlfr.fr. For the purpose of training machine learning classifiers, the images in each class were split into training, validation, and test sets, with proportions 70%, 15% and 15%. The archive contains : taxa.csv.gz Table of the classification of each object in the dataset, with columns : objid : unique object identifier in EcoTaxa (integer number). taxon_level1 : taxonomic name corresponding to the level 1 classification lineage_level1 : taxonomic lineage corresponding to the level 1 classification taxon_level2 : name of the taxon corresponding to the level 2 classification plankton : if the object is a plankton or not (boolean) set : class of the image corresponding to the taxon (train : training, val : validation, or test) img_path : local path of the image corresponding to the taxon (of level 1), named according to the object id features_native.csv.gz Table of morphological features computed by ZooCAM. All features are computed on the object only, not the background. All area/length measures are in pixels. All grey levels are in encoded in 8 bits (0=black, 255=white). With columns : area : object's surface area_exc : object surface excluding white pixels area_based_diameter : object's Area Based Diameter: 2 * (object_area/pi)^(1/2) meangreyobjet : mean image grey level modegreyobjet : modal object grey level sigmagrey : object grey level standard deviation mingrey : minimum object grey level maxgrey : maximum object grey level sumgrey : object grey level integrated density: object_mean*object_area breadth : breadth of the object along the best fitting ellipsoid minor axis length : breadth of the object along the best fitting ellipsoid majorr axis elongation : elongation index: object_length/object_breadth perim : object's perimeter minferetdiam : minimum object's feret diameter maxferetdiam : maximum object's feret diameter meanferetdiam : average object's feret diameter feretelongation : elongation index: object_maxferetdiam/object_minferetdiam compactness : Isoperimetric quotient: the ration of the object's area to the area of a circle having the same perimeter intercept0, intercept45 , intercept90, intercept135 : the number of times that a transition from background to foreground occurs a the angle 0ø, 45ø, 90ø and 135ø for the entire object convexhullarea : area of the convex hull of the object convexhullfillratio : ratio object_area/convexhullarea convexperimeter : perimeter of the convex hull of the object n_number_of_runs : number of horizontal strings of consecutive foreground pixels in the object n_chained_pixels : number of chained pixels in the object n_convex_hull_points : number of summits of the object's convex hull polygon n_number_of_holes : number of holes (as closed white pixel area) in the object roughness : measure of small scale variations of amplitude in the object's grey levels rectangularity : ratio of the object's area over its best bounding rectangle's area skewness : skewness of the object's grey level distribution kurtosis : kurtosis of the object's grey level distribution fractal_box : fractal dimension of the object's perimeter hist25, hist50, hist75 : grey level value at quantile 0.25, 0.5 and 0.75 of the object's grey levels normalized cumulative histogram valhist25, valhist50, valhist75 : sum of grey levels at quantile 0.25, 0.5 and 0.75 of the object's grey levels normalized cumulative histogram nobj25, nobj50, nobj75 : number of objects after thresholding at the object_valhist25, object_valhist50 and object_valhist75 grey level symetrieh :index of horizontal symmetry symetriev : index of vertical symmetry skelarea : area of the object skeleton thick_r : maximum object's thickness/mean object's thickness cdist : distance between the mass and the grey level object's centroids features_skimage.csv.gz Table of morphological features recomputed with skimage.measure.regionprops on the ROIs produced by ZooCAM. See http://scikit-image.org/docs/dev/api/skimage.measure.html#skimage.measure.regionprops for documentation. inventory.tsv Tree view of the taxonomy and number of images in each taxon, displayed as text. With columns : lineage_level1 : taxonomic lineage corresponding to the level 1 classification taxon_level1 : name of the taxon corresponding to the level 1 classification n : number of objects in each taxon group map.png Map of the sampling locations, to give an idea of the diversity sampled in this dataset. imgs Directory containing images of each object, named according to the object id objid and sorted in subdirectories according to their taxon.

The effect of mesoscale features on the distribution of planktonic organisms are well documented. Yet, the interaction between these spatial features and the temporal scale, which can result in sudden increases of the planktonic biomass, is less known and not described at high resolution. A permanent mesoscale front in the Ligurian Sea (north‐western Mediterranean) was repeatedly sampled between January and June 2021 using a SeaExplorer glider equipped with an Underwater Vision Profiler 6 (UVP6), a versatile in situ imager. Both plankton and particle distributions were resolved throughout the spring bloom to assess whether the front was a location of increased zooplankton concentration and whether it constrained particle distribution. Over the 5 months, the glider performed more than 5000 dives and the UVP6 collected 1.1 million images. We focused our analysis on shallow (300 m) transects, which gave a horizontal resolution of 900 m. About 13,000 images of planktonic organisms were retained. Ordination methods applied to particles and plankton concentrations revealed strong temporal variations during the bloom, with a succession of various zooplankton communities. Changes in particle abundance and size could be explained by changes in the plankton community. The front had a strong influence on particle distribution, while the signal was not as clear for plankton, probably because of the relatively small number of imaged organisms. This work confirms the need to sample both plankton and particles at fine scale to understand their interactions, a task for which automated in situ imaging is particularly adapted.

Abstract. In marine ecosystems, most physiological, ecological, or physical processes are size dependent. These include metabolic rates, the uptake of carbon and other nutrients, swimming and sinking velocities, and trophic interactions, which eventually determine the stocks of commercial species, as well as biogeochemical cycles and carbon sequestration. As such, broad-scale observations of plankton size distribution are important indicators of the general functioning and state of pelagic ecosystems under anthropogenic pressures. Here, we present the first global datasets of the Pelagic Size Structure database (PSSdb), generated from plankton imaging devices. This release includes the bulk particle normalized biovolume size spectrum (NBSS) and the bulk particle size distribution (PSD), along with their related parameters (slope, intercept, and R2) measured within the epipelagic layer (0–200 m) by three imaging sensors: the Imaging FlowCytobot (IFCB), the Underwater Vision Profiler (UVP), and benchtop scanners. Collectively, these instruments effectively image organisms and detrital material in the 7–10 000 µm size range. A total of 92 472 IFCB samples, 3068 UVP profiles, and 2411 scans passed our quality control and were standardized to produce consistent instrument-specific size spectra averaged to 1° × 1° latitude and longitude and by year and month. Our instrument-specific datasets span most major ocean basins, except for the IFCB datasets we have ingested, which were exclusively collected in northern latitudes, and cover decadal time periods (2013–2022 for IFCB, 2008–2021 for UVP, and 1996–2022 for scanners), allowing for a further assessment of the pelagic size spectrum in space and time. The datasets that constitute PSSdb's first release are available at https://doi.org/10.5281/zenodo.11050013 (Dugenne et al., 2024b). In addition, future updates to these data products can be accessed at https://doi.org/10.5281/zenodo.7998799.

Fast environmental changes and high coastal human pressures and impacts threaten the Mediterranean Sea. Over the last decade, recurrent blooms of the harmful dinoflagellate Ostreopsis cf. ovata have been recorded in many Mediterranean beaches. These microalgae produce toxins that affect marine organisms and human health. Understanding the environmental conditions that influence the appearance and magnitude of O. cf. ovata blooms, as well as how climate change will modify its future distribution and dynamics, is crucial for predicting and managing their effects. This study investigates whether the spatio-temporal distribution of this microalga and the frequency of its blooms could be altered in future climate change scenarios in the Mediterranean Western basin. For the first time, an ecological habitat model (EHM) is forced by physico-chemical climate change simulations at high-resolution, under the strong greenhouse gas emission trajectory (RCP8.5). It allows to characterize how O. cf. ovata may respond to projected conditions and how its distribution could shift over a wide spatial scale, in this plausible future. Before being applied to the EHM, future climate simulations are further refined by using a statistical adaptation method (Cumulative Distribution Function transform) to improve the predictions robustness. Temperature (optimum 23–26 °C), high salinity (>38 psu) and high inorganic nutrient concentrations (nitrate >0.25 mmol N·m$^{−3}$ and phosphate >0.035 mmol P·m−3) drive O. cf. ovata abundances. High spatial disparities in future abundances are observed. Namely, O. cf. ovata abundances could increase on the Mediterranean coasts of France, Spain and the Adriatic Sea while a decrease is expected in the Tyrrhenian Sea. The bloom period could be extended, starting earlier and continuing later in the year. From a methodological point of view, this study highlights best practices of EHMs in the context of climate change to identify sensitive areas for current and future harmful algal blooms.

Carbon fixation is a key metabolic function shaping marine life, but the underlying taxonomic and functional diversity involved is only partially understood. Using metagenomic resources targeted at marine piconanoplankton, we provide a reproducible machine learning framework to derive the potential biogeography of genomic functions through the multi-output regression of gene read counts on environmental climatologies. Leveraging the Marine Atlas of Tara Oceans Unigenes, we investigate the genomic potential of primary production in the global ocean. The latter is performed by ribulose-1,5-bisphosphate carboxylase/oxygenase (RUBISCO) and is often associated with carbon concentration mechanisms in piconanoplankton, major marine unicellular photosynthetic organisms. We show that the genomic potential supporting C 4 enzymes and RUBISCO exhibits strong functional redundancy and important affinity toward tropical oligotrophic waters. This redundancy is taxonomically structured by the dominance of Mamiellophyceae and Prymnesiophyceae in mid and high latitudes. These findings enhance our understanding of the relationship between functional and taxonomic diversity of microorganisms and environmental drivers of key biogeochemical cycles.

Abstract Primary production, performed by RUBISCO, and often associated with carbon concentration mechanisms, is of major importance in the oceans. Thanks to growing metagenomic resources (e.g., eukaryotic Metagenome-Assembled-Genomes; MAGs), we provide the first reproducible machine-learning-based framework to derive the potential biogeography of a given function, through the multi-output regression of the standardized number of reads of the associated genes on environmental climatologies. We use it to study the genomic potential of C4-photosynthesis of picoeukaryotes, a diverse and abundant group of marine unicellular photosynthetic organisms. We show that the genomic potential supporting C4-enzymes and RUBISCO exhibit strong functional redundancy and an important affinity towards tropical oligotrophic waters. This redundancy is then structured taxonomically by the dominance of Mamiellophyceae and Prymnesiophyceae in mid and high latitudes. Finally, unlike the genomic potential related to most C4-enzymes, the one of RUBISCO showed a clear pattern affinity for temperate waters.

Biodiversity is a measure of interest in many studies of global change. Depending on the question addressed, it takes a taxonomic, genetic, phylogenetic or functional perspective. Although aspects of morphology might be part of functional metrics, organism morphology is rarely considered explicitly in this context. We describe the changes in morphological diversity of marine zooplankton on the seasonal and long-term timescale. We digitalized weekly plankton samples collected from 2009-2020 in the NW Mediterranean Sea and automatically extracted 45 morphological features on > 800,000 individuals. Applying dimensional reduction (PCA) followed by clustering, we synthesized these features into four main morphological traits (describing size, transparency, circularity and shape complexity) and defined morphological groups (or ”morphs”). Based on these morphs, we computed time-series of morphological diversity indices (richness, divergence and evenness) relaying to metrics originally developed for functional diversity. Over the 12-year period, morphological – but not taxonomic – diversity increased while zooplankton concentration decreased. The environmental changes during this period consisted mainly in an increase in temperature and salinity and an overall impoverishment of surface waters, suggesting that niche specialization under low productivity increased the morphological divergence. Notably, an increased variability in the four morphological traits and higher proportions of extreme individuals caused this trend, with likely consequences for the ecosystem.

Biodiversity is studied notably because of its reciprocal relationship with ecosystem functions such as production. Diversity is traditionally described from a taxonomic, genetic or functional point of view but the diversity in organism morphology is seldom explicitly considered, except for body size. We describe morphological diversity of marine zooplankton seasonally and over 12 years using quantitative imaging of weekly plankton samples, in the northwestern Mediterranean Sea. We extract 45 morphological features on greater than 800 000 individuals, which we summarize into four main morphological traits (size, transparency, circularity and shape complexity). In this morphological space, we define objective morphological groups and, from those, compute morphological diversity indices (richness, evenness and divergence) using metrics originally defined for functional diversity. On both time scales, morphological diversity increased when nutritive resources and plankton concentrations were low, thus matching the theoretical reciprocal relationship. Over the long term at least, this diversity increase was not fully attributable to taxonomic diversity changes. The decline in the most common plankton forms and the increase in morphological variance and in extreme morphologies suggest a mechanism akin to specialization under low production, with likely consequences for trophic structure and carbon flux.

Machine learning covers a large set of algorithms that can be trained to identify patterns in data. Thanks to the increase in the amount of data and computing power available, it has become pervasive across scientific disciplines. We first highlight why machine learning is needed in marine ecology. Then we provide a quick primer on machine learning techniques and vocabulary. We built a database of ∼1000 publications that implement such techniques to analyse marine ecology data. For various data types (images, optical spectra, acoustics, omics, geolocations, biogeochemical profiles, and satellite imagery), we present a historical perspective on applications that proved influential, can serve as templates for new work, or represent the diversity of approaches. Then, we illustrate how machine learning can be used to better understand ecological systems, by combining various sources of marine data. Through this coverage of the literature, we demonstrate an increase in the proportion of marine ecology studies that use machine learning, the pervasiveness of images as a data source, the dominance of machine learning for classification-type problems, and a shift towards deep learning for all data types. This overview is meant to guide researchers who wish to apply machine learning methods to their marine datasets.

Floating plastic debris is a pervasive pollutant in seas and oceans, affecting a wide range of animals. In particular, microplastics (<5 mm in size) increase the possibility that marine species consume plastic and enter the food chain. The present study investigates this potential mistake between plastic debris and zooplankton by calculating the plastic debris to zooplankton ratio over the whole Mediterranean Sea. To this aim, in situ data from the Tara Mediterranean Expedition are combined with environmental and Lagrangian diagnostics in a machine learning approach to produce spatially-explicit maps of plastic debris and zooplankton abundance. We then analyse the plastic to zooplankton ratio in regions with high abundances of pelagic fish. Two of the major hotspots of pelagic fish, located in the Gulf of Gabes and Cilician basin, were associated with high ratio values. Finally, we compare the plastic to zooplankton ratio values in the Pelagos Sanctuary, an important hotspot for marine mammals, with other Geographical Sub-Areas, and find that they were among the larger of the Western Mediterranean Sea. Our results indicate a high potential risk of contamina-tion of marine fauna by plastic and advocate for novel integrated modelling approaches which account for potential trophic transfer within the food chain.

Autonomous and cabled platforms are revolutionizing our understanding of ocean systems by providing 4D monitoring of the water column, thus going beyond the reach of ship-based surveys and increasing the depth of remotely sensed observations. However, very few commercially available sensors for such platforms are capable of monitoring large particulate matter (100-2000 μm) and plankton despite their important roles in the biological carbon pump and as trophic links from phytoplankton to fish. Here, we provide details of a new, commercially available scientific camera-based particle counter, specifically designed to be deployed on autonomous and cabled platforms: the Underwater Vision Profiler 6 (UVP6). Indeed, the UVP6 camera-and-lighting and processing system, while small in size and requiring low power, provides data of quality comparable to that of previous much larger UVPs deployed from ships. We detail the UVP6 camera settings, its performance when acquiring data on aquatic particles and plankton, their quality control, analysis of its recordings, and streaming from in situ acquisition to users. In addition, we explain how the UVP6 has already been integrated into platforms such as BGC-Argo floats, gliders and long-term mooring systems (autonomous platforms). Finally, we use results from actual deployments to illustrate how UVP6 data can contribute to addressing longstanding questions in marine science, and also suggest new avenues that can be explored using UVP6-equipped autonomous platforms.

Plankton imaging systems supported by automated classification and analysis have improved ecologists' ability to observe aquatic ecosystems. Today, we are on the cusp of reliably tracking plankton populations with a suite of lab-based and in situ tools, collecting imaging data at unprecedentedly fine spatial and temporal scales. But these data streams have potential well beyond examining the abundances of different taxa; the individual images themselves contain a wealth of information on functional traits. Here we outline traits that could be measured from image data, suggest computer vision approaches to extract functional trait information from the images, and discuss promising avenues for novel studies. The approaches we discuss are data agnostic and are broadly applicable to other aquatic or terrestrial organisms.

Accurate plankton biomass estimations are essential to study marine ecological processes and biogeochemical cycles. This is particularly truefor copepods, which dominate mesozooplankton. Such estimations can efficiently be computed from organism volume estimated from images.However, imaging devices only provide 2D projections of 3D objects. The classical procedures to retrieve volumes, based on the Equivalent Spherical Diameter (ESD) or the best-fitting ellipse, are biased. Here, we present a method to correct these biases. First a new method aims to measure body area and fit an ellipse. Then, the body of copepods is modeled as an ellipsoid whose 2D silhouette is mathematically derived. Samples of copepod bodies are simulated with realistic shapes/sizes and random orientations. Their total volume is estimated from their silhouettes using the two classical methods and a correction factor is computed, relative to the known, total, volume. On real data, individual orientations and volumes are unknown but the correction factor still holds for the total volume of a large number of organisms. The correction is around -20% for the ESD method and +10% for the ellipse method. When applied to a database of ∼ 150 000 images of copepods captured by the Underwater Vision Profiler, the corrections decreased the gap between the two methods by a factor of 54. Additionally, the same procedure is used to evaluate the consequence of the bias in theestimation of individual volumes on the slopes of Normalised Biovolume Size Spectra and show that they are, fortunately, not sensitive to the bias.

Vertical variations in physical and chemical conditions drive changes in marine zooplankton community composition. In turn, zooplankton communities play a critical role in regulating the transfer of organic matter produced in the surface ocean to deeper layers. Yet, the links between zooplankton community composition and the strength of vertical fluxes of particles remain elusive, especially on a global scale. Here, we provide a comprehensive analysis of variations in zooplankton community composition and vertical particle flux in the upper kilometer of the global ocean. Zooplankton samples were collected across five depth layers and vertical particle fluxes were assessed using continuous profiles of the Underwater Vision Profiler (UVP5) at 57 stations covering seven ocean basins. Zooplankton samples were analysed using a Zooscan and individual organisms were classified into 19 groups for the quantitative analyses. Zooplankton abundance, biomass and vertical particle flux decreased from the surface to 1000 m depth at all latitudes. The zooplankton abundance decrease rate was stronger at sites characterised by oxygen minima (<5µmol O₂.kg⁻¹) where most zooplankton groups showed a marked decline in abundance, except the jellyfishes, molluscs, annelids, large protists and a few copepod families. The attenuation rate of vertical particle fluxes was weaker at such oxygen-depleted sites. Canonical redundancy analyses showed that the epipelagic zooplankton community composition depended on the temperature, on the phytoplankton size distribution and the surface large particulate organic matter while oxygen was an additional important factor for structuring zooplankton in the mesopelagic. Our results further suggest that future changes in surface phytoplankton size and taxa composition and mesopelagic oxygen loss might lead to profound shift in zooplankton abundance and community structure in both the euphotic and mesopelagic ocean. These changes may affect the vertical export and hereby the strength of the biological carbon pump.

Zooplankton plays a major role in ocean food webs and biogeochemical cycles, and provides major ecosystem services as a main driver of the biological carbon pump and in sustaining fish communities. Zooplankton is also sensitive to its environment and reacts to its changes. To better understand the importance of zooplankton, and to inform prognostic models that try to represent them, spatially-resolved biomass estimates of key plankton taxa are desirable. In this study we predict, for the first time, the global biomass distribution of 19 zooplankton taxa (1-50 mm Equivalent Spherical Diameter) using observations with the Underwater Vision Profiler 5, a quantitative in situ imaging instrument. After classification of 466,872 organisms from more than 3,549 profiles (0-500 m) obtained between 2008 and 2019 throughout the globe, we estimated their individual biovolumes and converted them to biomass using taxa-specific conversion factors. We then associated these biomass estimates with climatologies of environmental variables (temperature, salinity, oxygen, etc.), to build habitat models using boosted regression trees. The results reveal maximal zooplankton biomass values around 60°N and 55°S as well as minimal values around the oceanic gyres. An increased zooplankton biomass is also predicted for the equator. Global integrated biomass (0-500 m) was estimated at 0.403 PgC. It was largely dominated by Copepoda (35.7%, mostly in polar regions), followed by Eumalacostraca (26.6%) Rhizaria (16.4%, mostly in the intertropical convergence zone). The machine learning approach used here is sensitive to the size of the training set and generates reliable predictions for abundant groups such as Copepoda (R2 ≈ 20-66%) but not for rare ones (Ctenophora, Cnidaria, R2 < 5%). Still, this study offers a first protocol to estimate global, spatially resolved zooplankton biomass and community composition from in situ imaging observations of individual organisms. The underlying dataset covers a period of 10 years while approaches that rely on net samples utilized datasets gathered since the 1960s. Increased use of digital imaging approaches should enable us to obtain zooplankton biomass distribution estimates at basin to global scales in shorter time frames in the future.

Quantitative imaging instruments produce a large number of images of plankton and marine snow, acquired in a controlled manner, from which the visual characteristics of individual objects and their in situ concentrations can be computed. To exploit this wealth of information, machine learning is necessary to automate tasks such as taxonomic classification. Through a review of the literature, we highlight the progress of those machine classifiers and what they can and still cannot be trusted for. Several examples showcase how the combination of quantitative imaging with machine learning has brought insights on pelagic ecology. They also highlight what is still missing and how images could be exploited further through trait-based approaches. In the future, we suggest deeper interactions with the computer sciences community, the adoption of data standards, and the more systematic sharing of databases to build a global community of pelagic image providers and users. Expected final online publication date for the Annual Review of Marine Science, Volume 14 is January 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

A major challenge in characterizing plankton communities is the collection, identification and quantification of samples in a time-efficient way. The classical manual microscopy counts are gradually being replaced by high throughput imaging and nucleic acid sequencing. DNA sequencing allows deep taxonomic resolution (including cryptic species) as well as high detection power (detecting rare species), while RNA provides insights on function and potential activity. However, these methods are affected by database limitations, PCR bias, and copy number variability across taxa. Recent developments in high-throughput imaging applied in situ or on collected samples (high-throughput microscopy, Underwater Vision Profiler, FlowCam, ZooScan, etc) has enabled a rapid enumeration of morphologically-distinguished plankton populations, estimates of biovolume/biomass, and provides additional valuable phenotypic information. Although machine learning classifiers generate encouraging results to classify marine plankton images in a time efficient way, there is still a need for large training datasets of manually annotated images. Here we provide workflow examples that couple nucleic acid sequencing with high-throughput imaging for a more complete and robust analysis of microbial communities. We also describe the publicly available and collaborative web application EcoTaxa, which offers tools for the rapid validation of plankton by specialists with the help of automatic recognition algorithms. Finally, we describe how the field is moving with citizen science programs, unmanned autonomous platforms with in situ sensors, and sequencing and digitalization of historical plankton samples.

Images are increasingly used as a means to collect data in all fields of science, and Ecology is no exception. In the underwater realm, where direct observation of the organisms in their environment is difficult for humans, automated cameras provide invaluable insights. This partly explains the flourish of camera-based instruments specialised for taking pictures of plankton. Because most of them image a controlled volume in a systematic manner, they are coined "quantitative imaging" instruments; they allow computing concentrations and making replicable morphological measurements on the many thousands of images they collect. This also opens the avenue for the automation of their classification. EcoTaxa was designed as a platform to upload images, together with rich metadata, and sort them taxonomically in an efficient way. This efficiency is partly provided by machine learning: users can train models based on previous identifications in the database to suggest labels for newly uploaded images. By combining deep-learning feature extractors, a fast-to-train classifier, and enough flexibility to train models customised to the task at hand, EcoTaxa achieves classification performance similar to that of state of the art deep-learning networks while being usable in a matter of minutes by taxonomists with no computer science knowledge. The efficacy is also provided by the web-based graphical user interface: several users can collaborate on the classification of a dataset and each can rapidly review and classify hundreds of images at a time. As a result, trained operators routinely sort 5,000 to 10,000 per working day, within ~100 taxonomic groups. In the application as a whole, over 200 million images have been uploaded and over 90 million have been sorted by human operators, in its 6 years of existence. We will review the principle, functioning and potential for generalisation of the approach implemented in EcoTaxa.

Recent metagenomic studies have revealed that marine plankton is far more diverse than previously thought (Carradec et al. 2018, Salazar et al. 2019, Duarte et al. 2020), with hundreds of thousands of genetically distinct taxa and more than 116 million genes documented for eukaryotic plankton and 47 million genes for prokaryotes. However, the taxonomy and/or function of more than half of the planktonic ‘omic’ sequences is still unknown. These unprecedented amounts of data on planktonic communities call for innovative, data-driven approaches to quantify and observe their biogeographic importance (Faure et al. 2021). Marine plankton play a fundamental role in the global biogeochemical cycles and marine food webs. They are also a sentinel of environmental changes. Gathering more information about their genomics can help us better describe plankton distributions at global scale and further understand their response to environmental changes. The Blue-Cloud demonstrator Plankton Genomics responds to this challenge by mining the rich metagenomic and metatranscriptomic data collected during the Tara Oceans mission and combining it with in situ or climatological environmental information to infer the function, taxonomy and distribution of the large portion of unknown sequences. In this article, we are going to explore the main results of the demonstrator and its intended evolution. The demonstrator is led by the European Bioinformatics Institute (EMBL-EBI) and created by the Faculty of Sciences at Sorbonne University.

CNNs (Convolutional Neural Networks) are widely used for supervised classification. Although the networks themselves are designated as classifiers, they are in fact regressors trained to approximate the relationship between raw data and p predefined vectors of Rp playing the role of class representatives, where p is the number of classes. The actual classification decisions are taken by a nearest-neighbor classifier applied to the network outputs. Despite their usually impressive classification accuracies, ANNs (Artificial Neural Networks) are not always as straightforward to use as classical classifiers since they typically require large amounts of data, a high computational effort, and sometimes a solid experience to be trained. Yet, the principle of ANNs (input transformation into Rp, then basic nearest-neighbor classification) is interesting. In this work, we propose a simple, easily interpretable, and low on computational requirements alternative following the same principle. It relies on a weighted combination of ideal translations from the learning samples to some predefined targets. Because of its simplicity, it cannot directly deal with raw data as the ANNs do. Instead, it works with extracted features. Our experimental results, including on a realworld database of Plankton images, show classification accuracies on par with some classical classifiers.

As the basis of oceanic food webs and a key component of the biological carbon pump, planktonic organisms play major roles in the oceans. Their study benefited from the development of in situ imaging instruments, which provide higher spatio-temporal resolution than previous tools. But these instruments collect huge quantities of images, the vast majority of which are of marine snow particles or imaging artifacts. Among them, the In Situ Ichthyoplankton Imaging System (ISIIS) samples the largest water volumes (> 100 L s-1) and thus produces particularly large datasets. To extract manageable amounts of ecological information from in situ images, we propose to focus on planktonic organisms early in the data processing pipeline: at the segmentation stage. We compared three segmentation methods, particularly for smaller targets, in which plankton represents less than 1% of the objects: (i) a traditional thresholding over the background, (ii) an object detector based on maximally stable extremal regions (MSER), and (iii) a content-aware object detector, based on a Convolutional Neural Network (CNN). These methods were assessed on a subset of ISIIS data collected in the Mediterranean Sea, from which a ground truth dataset of > 3,000 manually delineated organisms is extracted. The naive thresholding method captured 97.3% of those but produced ~340,000 segments, 99.1% of which were therefore not plankton (i.e. recall = 97.3%, precision = 0.9%). Combining thresholding with a CNN missed a few more planktonic organisms (recall = 91.8%) but the number of segments decreased 18-fold (precision increased to 16.3%). The MSER detector produced four times fewer segments than thresholding (precision = 3.5%), missed more organisms (recall = 85.4%), but was considerably faster. Because naive thresholding produces ~525,000 objects from 1 minute of ISIIS deployment, the more advanced segmentation methods significantly improve ISIIS data handling and ease the subsequent taxonomic classification of segmented objects. The cost in terms of recall is limited, particularly for the CNN object detector. These approaches are now standard in computer vision and could be applicable to other plankton imaging devices, the majority of which pose a data management problem.

Abstract In the Gulf of Lions, small pelagic fish have shown reduced body size and body condition after 2007 that would result from changes in zooplankton community. We therefore examined zooplankton density, body size, and taxonomic composition at the closest long-term monitoring station (1995–2019): the coastal Point-B. To cover a broader spectrum of zooplankton community, samples obtained from two nets, the WP2 (200 µm mesh size) and the Regent (690 µm), were analysed with the imaging Zooscan method. One important result was the high stability through time of the zooplankton community. No long-term monotonous trends in density, size, and taxonomic composition were detected. Interannual variations in zooplankton size and density were not significantly correlated to any environmental variable, suggesting the possible importance of biotic interactions. Still, an increase in temperature was followed by a sharp decrease of zooplankton density in 2015, after which only gelatinous groups recovered. No change in the zooplankton community was detected around 2007 to support bottom-up control on small pelagic fish. Whether this derives from different local processes between the Gulf of Lions and the Ligurian Sea cannot be excluded, highlighting the need for simultaneous monitoring of different ecosystem compartments to fully understand the impact of climate change.

Abstract The larval stage is the main dispersive process of most marine teleost species. The degree to which larval behavior controls dispersal has been a subject of debate. Here, we apply a cross-species meta-analysis, focusing on the fundamental question of whether larval fish use external cues for directional movement (i.e., directed movement). Under the assumption that directed movement results in straighter paths (i.e., higher mean vector lengths) compared to undirected, we compare observed patterns to those expected under undirected pattern of Correlated Random Walk (CRW). We find that the bulk of larvae exhibit higher mean vector lengths than those expected under CRW, suggesting the use of external cues for directional movement. We discuss special cases which diverge from our assumptions. Our results highlight the potential contribution of orientation to larval dispersal outcomes. This finding can improve the accuracy of larval dispersal models, and promote a sustainable management of marine resources.

Marine particles of different nature are found throughout the global ocean. The term “marine particles” describes detritus aggregates and fecal pellets as well as bacterioplankton, phytoplankton, zooplankton and nekton. Here, we present a global particle size distribution dataset obtained with several Underwater Vision Profiler 5 (UVP5) camera systems. Overall, within the 64 µm to about 50 mm size range covered by the UVP5, detrital particles are the most abundant component of all marine particles; thus, measurements of the particle size distribution with the UVP5 can yield important information on detrital particle dynamics. During deployment, which is possible down to 6000 m depth, the UVP5 images a volume of about 1 L at a frequency of 6 to 20 Hz. Each image is segmented in real time, and size measurements of particles are automatically stored. All UVP5 units used to generate the dataset presented here were inter-calibrated using a UVP5 high-definition unit as reference. Our consistent particle size distribution dataset contains 8805 vertical profiles collected between 19 June 2008 and 23 November 2020. All major ocean basins, as well as the Mediterranean Sea and the Baltic Sea, were sampled. A total of 19 % of all profiles had a maximum sampling depth shallower than 200 dbar, 38 % sampled at least the upper 1000 dbar depth range and 11 % went down to at least 3000 dbar depth. First analysis of the particle size distribution dataset shows that particle abundance is found to be high at high latitudes and in coastal areas where surface productivity or continental inputs are elevated. The lowest values are found in the deep ocean and in the oceanic gyres. Our dataset should be valuable for more in-depth studies that focus on the analysis of regional, temporal and global patterns of particle size distribution and flux as well as for the development and adjustment of regional and global biogeochemical models. The marine particle size distribution dataset (Kiko et al., 2021) is available at https://doi.org/10.1594/PANGAEA.924375.

Plankton imaging instruments generate an ever increasing volume of data which is mostly processed through machine learning algorithms. However, classifying plankton images is a challenging computer science task in its own right: datasets are strongly unbalanced; the dominant classes are often not biologically interesting (artefacts, bubbles) and/or very heterogeneous looking (marine snow); and images span a large size range. Despite a wealth of reports on the performance of automatic plankton images classifiers, we still do not have a definitive idea regarding how methods compare with each other and where they can systematically be trusted. This is mostly because those reports rely on rather small unpublished datasets, not necessarily representative of real-life biological samples in terms of size, number of categories and proportions. Here we report the performance of a classic classification method (Random Forest on handcrafted image features) and a more recent one (a Convoluted Neural Network) on large publicly released datasets, from five widely used plankton imaging instruments. We show that CNN improve classification performance but only noticeably on poorly represented (a few hundred images) classes. Finally, we showcase the difference between the predictions of the two classifiers and a human-checked truth on several real-world datasets, to give insights regarding which ecological questions can or cannot be studied from computer-generated classifications only.

Plankton ecologists are increasingly looking toward functional trait-based approaches to analyze pelagic marine ecosystems. Within this framework, an individual organism’s fitness is assessed as a function of its particular combination of traits. Likewise, the distribution of organisms in the ocean can be understood as an ecosystem structured by functional traits. Lots of progress has been made from a modelling and theoretical perspective, but observational studies remain limited. Many traits of interest, however, are either directly measurable or inferable from data collected by in situ imaging systems. Here we discuss several such traits and the methods one might use to extract them from plankton image data. We particularly focus on quantifying the mass of egg sacs on ovigerous copepods directly from a globally distributed set of images collected by the Underwater Vision Profiler. Early results drawing on modern object detection, segmentation, and deep regression methods will be discussed. While these techniques are applied to a very particular set of data, they are quite general and could be broadly applied in image-based studies of other pelagic organisms.

A large majority of in situ imaging systems produce black and white images. From human eyes or image descriptors, we find that grey levels differences often represent a large source of variations within these dataset. For living organisms, these variations could bring insights on plankton ecology because there are signatures of pigmented structures (gut content, gonads and eggs) and camouflage strategies (transparency, pigmentation). This brings us to an important interest of behaviour and functional ecology : animal color. In this work, we will discuss variations of copepod color, firstly inferred from underwater images of arctic copepods of the Calanus genus. Then, from a small meta-analysis based on ~180 carotenoid quantifications from marine and freshwater ecosystems, we will try to understand how copepod color varies at a global scale and show how carotenoid pigmentation influences their fitness (survival, growth, and reproduction). Finally, we will attempt to bring perspectives about color in plankton imagery and the potentials it could have to monitor the sustainability of key pelagic ecosystems such as arctic and subarctic waters.

The organic carbon produced in the ocean's surface by phytoplankton is either passed through the food web or exported to the ocean interior as marine snow. The rate and efficiency of such vertical export strongly depend on the size, structure and shape of individual particles, but apart from size, other morphological properties are still not quantitatively monitored. With the growing number of in situ imaging technologies, there is now a great possibility to analyze the morphology of individual marine snow. Thus, automated methods for their classification are urgently needed. Consequently, here we present a simple, objective categorization method of marine snow into a few ecologically meaningful functional morphotypes using field data from successive phases of the Arctic phytoplankton bloom. The proposed approach is a promising tool for future studies aiming to integrate the diversity, composition and morphology of marine snow into our understanding of the biological carbon pump.

The development of a new generation of biological sensors has drastically changed the studies of marine plankton from lab bench work to in vivo and real-time observations. Target organisms, from bacteria to plankton, can now be optically characterized and/or photographed and archived. Consequently, scientists are facing the difficult challenge to handle a large amount of data which need to be processed rapidly and harmonised before being stored in databases to be accessible to scientific/environmental management communities. In the Joint European Research Infrastructure for Coastal Observatories (JERICO-RI), we intend to integrate the monitoring of physical, biogeochemical and biological variables to better understand coastal ecosystems. Focusing primarily on plankton communities measured at high spatial and temporal resolution, we aim to provide a framework for the data flow following the Findability, Accessibility, Interoperability, and Reuse (FAIR). To achieve this, we need to draw up best practices in data management to be followed by users and experts operating the sensors such as standardised protocol, minimal technical metadata elements for effective re-use, identify and extend appropriate vocabularies, identify tools for data integration and platforms for trust-worthy long-term archival, standardised data formats to be ingested by European data infrastructures. In this presentation, we will present some of our achievements regarding in vivo flow cytometry, imagery analysis, and multispectral fluorimetry.

Abstract Ocean plankton comprise organisms from viruses to fish larvae that are fundamental to ecosystem functioning and the provision of marine services such as fisheries and CO 2 sequestration. The latter services are partly governed by variations in plankton community composition and the expression of traits such as body size at community-level. While community assembly has been thoroughly studied for the smaller end of the plankton size spectrum, the larger end comprises ectotherms that are often studied at the species, or group-level, rather than as communities. The body size of marine ectotherms decreases with temperature, but controls on community-level traits remain elusive, hindering the predictability of marine services provision. Here, we leverage Tara Oceans datasets to determine how zooplankton community composition and size structure varies with latitude, temperature and productivity-related covariates in the global surface ocean. Zooplankton abundance and median size decreased towards warmer and less productive environments, as a result of changes in copepod composition. However, some clades displayed the opposite relationships, which may be ascribed to alternative feeding strategies. Given that climate models predict increasingly warmed and stratified oceans, our findings suggest that zooplankton communities will shift towards smaller organisms which might weaken their contribution to the biological carbon pump.

As phytoplankton is essential in supporting food webs and mediating biogeochemical cycles, it became crucial to complement discrete sampling and laboratory methods to address this compartment at its finest temporal and spatial scale. Within the “Joint European Research Infrastructure for Coastal Observatories – Novel EXperTise” (JERICO-NEXT), scientists proceeded to the exploration of novel phytoplankton automated in vivo/in situ techniques, based on single cell/particle or bulk optical characteristics. These techniques were implemented in a variety of monitoring platforms (fixed stations, moorings, research vessels, ships of opportunity), from oligotrophic (West Mediterranean) to mesotrophic and eutrophic marine coastal waters (English Channel, North Sea and Baltic Sea). These approaches could represent warning systems in water quality through phytoplankton total abundance and biomass, composition changes and the occurrence of blooms, including harmful algal blooms HAB, of special interest in areas of fishing, aquaculture and tourism. At the dawn of the UN Decade of the Ocean, the JERICO-S3 (Science, Services, Sustainability) project tends to generalise these approaches and seeks to standardise as much as possible their operability and capacity in addressing phytoplankton diversity and productivity, moving towards a new approach of Essential Ocean Variables (EOVs). Moreover, the development of automated tools for data analysis, using machine-learning and deep learning and will allow integrating these data into European marine data infrastructures available for research and marine management.

Imaging techniques are increasingly used in ecology studies, producing vast quantities of data. Inferring functional traits from individual images can provide original insights on ecosystem processes. Morphological traits are, as other functional traits, individual characteristics influencing an organism's fitness. We measured them from in situ image data to study an Arctic zooplankton community during sea ice break‐up. Morphological descriptors (e.g., area, lightness, complexity) were automatically measured on ∼ 28,000 individual copepod images from a high‐resolution underwater camera deployed at more than 150 sampling sites across the ice‐edge. A statistically‐defined morphological space allowed synthesizing morphological information into interpretable and continuous traits (size, opacity, and appendages visibility). This novel approach provides theoretical and methodological advantages because it gives access to both inter‐ and intra‐specific variability by automatically analyzing a large dataset of individual images. The spatial distribution of morphological traits revealed that large copepods are associated with ice‐covered waters, while open waters host smaller individuals. In those ice‐free waters, copepods also seem to feed more actively, as suggested by the increased visibility of their appendages. These traits distributions are likely explained by bottom‐up control: high phytoplankton concentrations in the well‐lit open waters encourages individuals to actively feed and stimulates the development of small copepod stages. Furthermore, copepods located at the ice edge were opaquer, presumably because of full guts or an increase in red pigmentation. Our morphological trait‐based approach revealed ecological patterns that would have been inaccessible otherwise, including color and posture variations of copepods associated with ice‐edge environments in Arctic ecosystems.

In low-nutrient low-chlorophyll areas, such as the Mediterranean Sea, atmospheric fluxes represent a considerable external source of nutrients likely supporting primary production, especially during periods of stratification. These areas are expected to expand in the future due to lower nutrient supply from sub-surface waters caused by climate-driven enhanced stratification, likely further increasing the role of atmospheric deposition as a source of new nutrients to surface waters. Whether plankton communities will react differently to dust deposition in a warmer and acidified environment remains; however, an open question. The potential impact of dust deposition both in present and future climate conditions was investigated in three perturbation experiments in the open Mediterranean Sea. Climate reactors (300 L) were filled with surface water collected in the Tyrrhenian Sea, Ionian Sea and in the Algerian basin during a cruise conducted in the frame of the PEACETIME project in May–June 2017. The experiments comprised two unmodified control tanks, two tanks enriched with a Saharan dust analogue and two tanks enriched with the dust analogue and maintained under warmer (+3 ∘C) and acidified (−0.3 pH unit) conditions. Samples for the analysis of an extensive number of biogeochemical parameters and processes were taken over the duration (3–4 d) of the experiments. Dust addition led to a rapid release of nitrate and phosphate, however, nitrate inputs were much higher than phosphate. Our results showed that the impacts of Saharan dust deposition in three different basins of the open northwestern Mediterranean Sea are at least as strong as those observed previously, all performed in coastal waters. The effects of dust deposition on biological stocks were different for the three investigated stations and could not be attributed to differences in their degree of oligotrophy but rather to the initial metabolic state of the community. Ocean acidification and warming did not drastically modify the composition of the autotrophic assemblage, with all groups positively impacted by warming and acidification. Although autotrophic biomass was more positively impacted than heterotrophic biomass under future environmental conditions, a stronger impact of warming and acidification on mineralization processes suggests a decreased capacity of Mediterranean surface plankton communities to sequester atmospheric CO2 following the deposition of atmospheric particles.

Plankton images not only allow to classify organisms taxonomically but also to measure their size, which is then used to estimate their biomass. Accurate plankton biomass estimations are needed to quantify their role in ecological processes and biogeochemical cycles; this is particularly true for copepods, which dominate mesozooplankton. However, imaging devices only provide a 2D projection of 3D objects. The assumptions made to retrieve a 3D volume and the diverse orientations of the organisms relative to the camera <i>in situ</i> induce a bias in our estimation of volume. Here we present a method to correct this bias. The body of copepods is modeled as an ellipsoid of known volume, whose projected silhouette is mathematically derived. Samples of copepod bodies are simulated with a realistic size distribution and random orientations, their volume is estimated from their silhouette using two classic methods (Equivalent Spherical Diameter and best-fitting ellipse), and a correction factor is computed for each, by comparing with the actual, known, volume. On real data, individual orientations and volumes are unknown but the correction factors still hold for the <b>total</b> volume of a <b>large</b> number of organisms. The correction is -22\% for the spherical diameter method and +12\% for the ellipse method. When applied to a database of ~150,000 real <i>in situ</i> images of copepods captured by the Underwater Vision Profiler, the corrections decreased the gap between the two methods by a factor of 20. Incidentally, we also propose more accurate techniques for surface estimation and ellipse fitting.

Optical particle measurements are emerging as an important technique for understanding the ocean carbon cycle, including contributions to estimates of their downward flux, which sequesters carbon dioxide (CO2) in the deep sea. Optical instruments can be used from ships or installed on autonomous platforms, delivering much greater spatial and temporal coverage of particles in the mesopelagic zone of the ocean than traditional techniques, such as sediment traps. Technologies to image particles have advanced greatly over the last two decades, but the quantitative translation of these immense datasets into biogeochemical properties remains a challenge. In particular, advances are needed to enable the optimal translation of imaged objects into carbon content and sinking velocities. In addition, different devices often measure different optical properties, leading to difficulties in comparing results. Here we provide a practical overview of the challenges and potential of using these instruments, as a step toward improvement and expansion of their applications.

Biological diversity encompasses all the variations of life, from genes to ecosystems, and is typically described from a taxonomic, genetic, phylogenetic, or functional point of view. Here we describe the changes in morphological diversity of zooplankton along a coastal time series. Morphology is of course characteristic of taxa but also has functional consequences; size, for example, is often considered as a dominant functional trait for zooplankton. Using high-throughput imaging of weekly plankton samples, collected from 2009 to 2017 in the Mediterranean Sea, we automatically measured ~40 morphological traits on ~542,000 individuals. A reduced morphological space was defined through Principal Component Analysis and individuals were regrouped in 200 "morphs" through clustering in that space. In this morphological space, time series of indices of morphological richness, divergence, and evenness were computed using the same metrics as the ones usually defined for functional diversity. Size, circularity, and opacity were the characteristics of the organisms that varied the most along the series. The morphs were homogeneous in appearance and recognisable, but often comprised more than one taxon. All morphological diversity indices were lower in the spring. This seemed related to the dominance of copepods during this season, which are quite homogeneous in appearance. But this could also be caused by selective disadvantages of extreme (and specialised) morphological types in this relatively eutrophic time of the year. Over the nine years, morphological divergence increased significantly while overall plankton concentrations decreased; since 2001, surface waters became significantly warmer and more oligotrophic. Overall, this fits the theory that oligotrophy leads to niche specialisation which, here, translates into morphological divergence.

Most coastal fish species spend their early life stages in the pelagic environment, before settling in coastal habitats. The variability in the arrival of larvae to coastal habitats provide information on the species' biology and recruitment potential. To explore the dynamics of larval fish supply to coastal habitats in the NW Mediterranean Sea, 13 sites were monitored using light-traps, from July 2012 to December 2015. Most variation in Catches Per Unit Effort (CPUE) among topographic basins and species were statistically significant for high (quantile 75%) and very high (quantile 90%) catches only. At year scale, CPUE displayed strong seasonality, and three main species assemblages were detected in late spring-early summer, summer and late autumn-early winter. At month scale, CPUE were higher around the new moon for all quantiles and temporally autocorrelated at a lag of about 28 days. Larval supply also varied spatially with site-specific associations, and with riverine influence. All together, these results confirm that the previously described patterns of larval supply observed in tropical and subtropical environments (e.g., the high variability at all spatial and temporal scales and the strong influence of the moon) also apply to Mediterranean fish assemblages. The larval supply in the NW Mediterranean Sea comes out as a solid candidate for monitoring the state of the marine ecosystems, highlighting the need to continue such time series.

Functional traits are individual characteristics that influence an organism's fitness and ecological functions. Thanks to technological progress, these traits can be measured at an individual level, completing with quantitative information the usual taxonomic approach to assess the structure and functioning of ecosystems. We studied the surface waters of Baffin Bay, an Arctic marginal sea located between Greenland and Canada, at the moment of sea ice break-up. Strong environmental gradients are created by the confrontation of two water masses, sea ice melting, and the increase in temperature and irradiance. We focused on copepods that overwhelmingly dominate zooplankton communities there. We measured morphological descriptors of images (area, darkness, complexity, etc.) on about 28,000 copepod images taken by the Underwater Vision Profiler (UVP). A statistically-defined multidimensional morphological space allows to synthesize individual images into interpretable continuous traits (size, transparency, appendages, etc.). The spatial distribution of these traits revealed that large copepods are associated with ice-covered waters in the West while smaller are present in open waters in the East. Copepods of the eastern part also seem to have higher feeding activity, as inferred by appendage visibility. High phytoplankton concentrations and probable strong visual predation pressure on big copepods in well-lit open waters could be responsible for these traits distributions. Furthermore, copepods located right at the ice edge appeared more opaque on images, suggesting that these individuals have a strong red pigmentation (the UVP light source is red). The combination of in situ imaging and individual trait-based approach revealed important ecological patterns that would have been inaccessible otherwise, including the role of copepod behaviour and ecological interactions on zooplankton ecosystem dynamics in the Arctic.

The European Marine Board is an independent and self-sustaining science policy interface organisation that currently represents 34 Member Organizations from 18 European countries. It was established in 1995 to facilitate enhanced cooperation between European marine science organizations towards the development of a common vision on the strategic research priorities for marine science in Europe. The EMB promotes and supports knowledge transfer for improved leadership in European marine research. Its membership includes major national marine or oceanographic institutes, research funding agencies and national consortia of universities with a strong marine research focus. Adopting a strategic role, the European Marine Board serves its member organizations by providing a forum within which marine research policy advice is developed and conveyed to national agencies and to the European Commission, with the objective of promoting the need for, and quality of, European marine research.

In recent years, autonomous platforms have been improved with the objective of observing the deep ocean. Many prototypes have reached 6,000 m, but only a few are commercially available. Imaging sensors have been developed to study the numbers, sizes and shapes of particles and plankton. These have been deployed independently or mounted on CTD rosettes (e.g. the UVP5 Underwater Vision Profiler). In some areas, they observed high concentrations of large particles down to the bottom of the ocean, showing intense carbon export. They also documented the presence of fragile organisms, such as rhizarians, whose key ecological roles in the ocean was unknown until then. The new miniaturized UVP6-LP (Low Power) sensor, developed to be mounted on autonomous platforms, is complementary to the larger sensors deployed on CTD rosettes. It also records and identifies particles and plankton, using imagery. It counts and sizes particles \textgreater80 µM ESD and identifies large aggregates and plankton \textgreater700µM ESD. This, at low cost and with very low power. Six prototypes of the sensor have been inter-calibrated with the reference UVP5. The instrument can be deployed down to 6,000 m, which corresponds to 97% of the ocean surface. A UVP6-LP was used to quantify particles released from a deep-sea mining experiment at 4,300 m. Others performed transects on gliders and vertical profiles on a float. A UVP6-LP was moored for one year at 50 m at 82°N. Other UVP6-LP will be deployed on the a cabled observatory in the Mediterranean Sea at 2,200 m, and on the SeaCycler mooring in the Labrador Sea. The very low power required for the operation of the UVP6-LP allows to optimize its use on profiling floats and gliders. Autonomous platforms cannot transmit images due to the limitation of satellite bandwidth or acoustic telemetry. To overcome this limitation, the UVP6-LP includes an embedded algorithm for the automatic classification of large aggregates and plankton images, which provides data that are accurate enough for monitoring programs and scientific studies. Because scientists, policy makers and the public require easy access to data, a complete software ecosystem is used to pilot the instrument, record the data, and make them freely available to the scientists and the public. When the instruments are recovered after deployment, their data include classified images.

“Imagery data” can be referred as qualitative and quantitative information from a collection of images. Imaging systems are used more and more frequently in the marine domain to generate huge amounts of imagery data. For example, automatic image classification is used to determine the abundance, size and biomass of plankton communities. In addition, the recent advances of imaging sensors and the growing datasets, highlight the importance of the management and storage capacity of these data. Thus, establishing data standards, optimized data flows and quality control procedures will promote the ability to make these datasets findable, accessible, interoperable and reusable (FAIR principles).At the moment, there are a number of online open-access databases that collect marine biodiversity data, such as the Ocean Biodiversity Information System (OBIS), and more specialized in plankton observations such as the COPEPOD for plankton biomass and ECOTAXA for taxonomic annotation of plankton images. However, they currently lack relevant standards to link metadata of the images.International biodiversity data standards, such as Darwin Core (DwC), are already widely used in OBIS and the Global Biodiversity Information Facility (GBIF). OBIS, has recently adopted the OBIS-ENV-DATA, a format that follows the DwC-Archive (DwC-A) standard, consisting of a DwC Event table in combination with an Occurrence an extended Measurement or Fact table. This structure enables the linkage of quantitative and qualitative properties to both sampling events and species occurrences. It also includes additional fields for property standardization, such as the BODC (British Oceanographic Data Centre) controlled vocabularies, the World Register of Marine Species (WoRMS), which provides crucial quality control support for taxonomic data, and Marine Regions, a database that provides standardized marine georeferenced place names and areas. However, there are no clear guidelines on how to include imagery metadata and derived data in OBIS-ENV-DATA.During this TDWG 2020 symposium, we would like to present and discuss our ongoing work to establish best practices and standardized protocols for imaging data and metadata acquired by a large spectrum of bio-optic sensors. Furthermore, we highlight how the integration of the metadata will flow to existing biodiversity data portals, the European EMODnet Biology portal, (EurOBIS) and OBIS.

Ma carrière scientifique a commencé par l'étude du comportement des larves de poissons. J'ai prouvé qu'elles nagent à des vitesses non-négligeables, s'orientent, notamment en fonction de l'azimut solaire, et s'orientent mieux en groupe que seules. Quand ces capacités comportementales sont introduites dans des modèles Lagrangiens de dispersion, elles influencent les patrons spatiaux de recrutement, réduisant souvent les distances de dispersion. Je me suis également intéressé à la distribution tridimensionnelle de ces larves et du plancton associé. Ces études ont montré que, dans les structures à mésoéchelle, une très haute résolution spatio-temporelle est nécessaire pour parvenir à détecter des patrons clairs dans la distribution des organismes ou les associations entre taxa. À plus large échelle, je me suis ensuite intéressé à la biogéographie d'espèces pélagiques. Dans un atlas de l'océan Austral, la modélisation de la niche écologique d'espèces de poissons et de krill a souligné la forte influence des fronts circum-antarctiques. Les travaux de régionalisation de la Méditerranée ont montré que des communautés spécifiques d'organismes pélagiques exploitent la zone de gradient entre les bassins Est et Ouest. La comparaisons entre diverses régionalisations de cette région a permis d'identifier des zones, de stabilité ou de fort gradient, cohérentes quelles que soit les caractéristiques des masses d'eau considérées. Enfin, j'ai contribué à l'effort d'observation à long terme mené à Villefranche, notamment en mettant en place un système de centralisation et visualisation des données récoltées. L'exploitation de ces données a confirmé que la Méditerranée est une zone où le changement climatique est prononcé, que l'oligotrophisation associée résulte en une diversification des organismes zooplanctoniques et que le recrutement des poissons côtiers est cohérent spatialement mais extrêmement sporadique temporellement. Le point commun entre ces activités apparemment disparates est l'utilisation d'outils numériques pour répondre à des questions écologiques. Mon projet de recherche est également focalisé sur des approches ``computationnelles'', telles que l'apprentissage machine, qui combine de nombreux calculs unitaires pour reconnaitre des structures dans de grandes masses de données. Ces données sont de maintenant disponibles pour l'étude du plancton, notamment grâce à de nouveaux instruments d'imagerie. Je commencerai par poursuivre mes efforts actuels vers l'automatisation de l'identification d'organismes planctoniques sur des images. En utilisant ces nouveaux outils instrumentaux et logiciels, il deviendra possible de décrire la distribution spatio-temporelle du plancton à la même, haute, résolution que son environnement biogéochimique. Cela permettra d'étudier leurs intéractions dans les structures à submésoéchelle, qui sont primordiales pour la dynamique de ces organismes à plus grande échelle. Enfin, à cette grande échelle, les satellites permettent d'avoir une vision synoptique de la couche de surface des océans. Je propose d'utiliser les corrélations spatio-temporelles locales dans données qu'ils récoltent pour décrire, et prédire à l'échelle globale, la structure des communautés de zooplancton.

Mortality is very high during the pelagic larval phase of fishes but the factors that determine recruitment success remain unclear and hard to predict. Because of their bipartite life history, larvae of coastal species have to head back to the shore at the end of their pelagic episode, to settle. These settlement-stage larvae are known to display strong sensory and motile abilities, but most work has been focused on tropical, insular environments and on the influence of coast-related cues on orientation. In this study we quantified the in situ orientation behavior of settlement- stage larvae in a temperate region, with a continuous coast and a dominant along- shore current, and inspected both coast-dependent and independent cues. We tested six species: one Pomacentridae, Chromis chromis, and five Sparidae, Boops boops, Diplodus annularis, Oblada melanura, Spicara smaris and Spondyliosoma cantharus. Over 85% of larvae were highly capable of keeping a bearing, which is comparable to the orientation abilities of tropical species. Sun-related cues influenced the precision of bearing-keeping at individual level. Three species, out of the four tested in sufficient numbers, oriented significantly relative to the sun position. These are the first in situ observations demonstrating the use of a sun compass for orientation by wild-caught settlement-stage larvae. This mechanism has potential for large-scale orientation of fish larvae globally.

Marine Protected Areas have become a major tool for the conservation of marine biodiversity and resources. Yet our understanding of their efficacy is often limited because it is measured for a few biological components, typically top predators or species of commercial interest. To achieve conservation targets, marine protected areas can benefit from ecosystem-based approaches. Within such an approach, documenting the variation of plankton indicators and their covariation with climate is crucial as plankton represent the base of the food webs. With this perspective, we sought to document the variations in the emerging properties of the plankton to better understand the dynamics of the pelagic fishes, mammals and seabirds that inhabit the region. For the first time, we analyze the temporal variations of the entire plankton community of one of the widest European protected areas, the Parc Naturel Marin de la Mer d’Iroise. We used data from several sampling transects carried out in the Iroise Sea from 2011 to 2015 to explore the seasonal and inter-annual variations of phytoplankton and mesozooplankton abundance, composition and size, as well as their covariation with abiotic variables, through multiple multivariate analyses. Overall, our observations are coherent with the plankton dynamics that have been observed in other regions of the North-East Atlantic. We found that both phytoplankton and zooplankton show consistent seasonal patterns in taxonomic composition and size structure but also display inter-annual variations. The spring bloom was associated with a higher contribution of large chain-forming diatoms compared to nanoflagellates, the latter dominating in fall and summer. Dinoflagellates show marked inter-annual variations in their relative contribution. The community composition of phytoplankton has a large impact on the mesozooplankton together with the distance to the coast. The size structure of the mesozooplankton community, examined through the ratio of small to large copepods, also displays marked seasonal patterns. We found that larger copepods (members of the Calanidae) are more abundant in spring than in summer and fall. We propose several hypotheses to explain the observed temporal patterns and we underline their importance for understanding the dynamics of other components of the food-web (such as sardines). Our study is a first step toward the inclusion of the planktonic compartment into the planning of the resources and diversity conservation within the Marine Protected Area.

Plankton plays a key role in the biological pump and has a big impact on marine living resources. However, plankton is difficult to observe in a consistent manner across its extended size range and by the multiple observers that uses protocols that are not inter calibrated. Imaging sensors have the potential to provide key ``ecosystem essential ocean variables'' eEOVs (plankton biodiversity, morphological traits) that complement other sensors such as optical ones. Lab and in-situ imaging sensors have been deployed the 10 last years to provide insights into local dynamics in the frame of time series programs (from daily to decadal scales) and during oceanographic surveys across ocean basins. Combining observations from the different programmes has sometimes allowed to detect concomitant changes in different areas or provide a better spatial distribution of plankton communities. For example, such efforts were supported by the European FP7 JERICO, H2020 JERICO-NEXT, BRIDGES, EURO-BASIN programs. Most of the observation efforts were performed independently and hundreds of millions of images have been collected (and billions to come as sensors are getting more available). All those sparsely distributed images are usually not available for the users because of limited development in software solutions for identification, archiving and distribution, which are in a current improving process. Several attempts for developing web based services for image recognition, distribution and archiving have been performed (ecotaxa.obs-vlfr.fr) but only a fraction of the existing and future data can be treated by them. Based on the past ten years of effort, we will present a synthesis of successful developments in using imaging systems to provide information on plankton community at local, regional and ultimately global scales. These examples will show how relevant they are for ecosystem monitoring (e.g. detection of ecosystem changes and regime-shifts) and services (e.g. aquaculture, fisheries, biological carbon pump). We will then build on these examples to discuss future developments with the aim of, better observing, harmonizing practices and developing state of the art marine data and information management in order to increase the connection with the relevant stakeholders and community of users among researchers, conservation managers and private companies.

To identify the spatial and temporal patterns of larval fish supply to coastal habitats in the NW Mediterranean Sea, 13 sites were monitored throughout the French Mediterranean coast using light-traps, from July 2012 to December 2015. A total of 27,800 coastal fish larvae belonging to 72 species were caught over 1073 sampling nights. Three species assemblages were detected in late spring-early summer, summer, and winter, with high consistency among years at each site. Catches Per Unit Effort (CPUE) were most often very low (median CPUE=0), with rare high catches events. Statistical analyses of CPUE were therefore conducted with quantile regressions, focusing on low (q25), medium (q50), high (q75), and extraordinary catches (q90). Most variations in CPUE among years, topographic basins and species were statistically significant for high and very high catches only. The year-to-year variability followed the climatic variability in the NW Mediterranean basin, while at year scale, CPUE displayed strong seasonality, with a major peak in early summer accounting for most of the larval supply and a weaker peak in autumn. An influence of the lunar cycle was detected for all catch levels, with higher CPUE around the new moon and significant temporal autocorrelations at a lag of about 28 days. Spatial patterns were also detected, with, among others: a strong influence of the Rhône river's triggering a lower and later larval supply in western sites compared to eastern sites, a lower species richness at sand-dominated sites compared to rock-dominated ones, and higher diversity and abundance at sites located within or close to a marine protected. These results confirm the high variability of larval supply at all temporal scales and the foremost influence of the moon, already observed in tropical environments, also apply to Mediterranean fish assemblages, and indicate that larval fish supply is a solid candidate for monitoring the state of the NW Mediterranean ecosystems.

The size of current plankton image datasets renders manual classification virtually infeasible. The training of models for machine classification is complicated by the fact that a large number of classes consist of only a few examples. We employ the recently introduced weight imprinting technique in order to use the available training data to train accurate classifiers in absence of enough examples for some classes. The model architecture used in this work succeeds in the identification of plankton using machine learning with its unique challenges, i.e. a limited number of training examples and a severely skewed class size distribution. Weight imprinting enables a neural network to recognize small classes immediately without retraining. This permits the mining of examples for novel classes.

In this paper we review the technologies available to make globally quantitative observations of particles in general—and plankton in particular—in the world oceans, and for sizes varying from sub-microns to centimeters. Some of these technologies have been available for years while others have only recently emerged. Use of these technologies is critical to improve understanding of the processes that control abundances, distributions and composition of plankton, provide data necessary to constrain and improve ecosystem and biogeochemical models, and forecast changes in marine ecosystems in light of climate change. In this paper we begin by providing the motivation for plankton observations, quantification and diversity qualification on a global scale. We then expand on the state-of-the-art, detailing a variety of relevant and (mostly) mature technologies and measurements, including bulk measurements of plankton, pigment composition, uses of genomic, optical and acoustical methods as well as analysis using particle counters, flow cytometers and quantitative imaging devices. We follow by highlighting the requirements necessary for a plankton observing system, the approach to achieve it and associated challenges. We conclude with ranked action-item recommendations for the next 10 years to move toward our vision of a holistic ocean-wide plankton observing system. Particularly, we suggest to begin with a demonstration project on a GO-SHIP line and/or a long-term observation site and expand from there, ensuring that issues associated with methods, observation tools, data analysis, quality assessment and curation are addressed early in the implementation. Global coordination is key for the success of this vision and will bring new insights on processes associated with nutrient regeneration, ocean production, fisheries and carbon sequestration.

Aim: To assess the impact of climate change on the functional diversity of marine zooplankton communities.Location: The Mediterranean Sea.Methods: We used the functional traits and geographic distributions of 106 copepod species to estimate the zooplankton functional diversity of Mediterranean surface assemblages for the 1965–1994 and 2069–2098 periods. Multiple environmental niche models were trained at the global scale to project the species habitat suitability in the Mediterranean Sea and assess their sensitivity to climate change predicted by several scenarios. Simultaneously, the species traits were used to compute a functional dendrogram from which we identified seven functional groups and estimated functional diversity through Faith's index. We compared the measured functional diversity to the one originated from null models to test if changes in functional diversity were solely driven by changes in species richness.Results: All but three of the 106 species presented range contractions of varying intensity. A relatively low decrease of species richness (−7.42 on average) is predicted for 97% of the basin, with higher losses in the eastern regions. Relative sensitivity to climate change is not clustered in functional space and does not significantly vary across the seven copepod functional groups defined. Changes in functional diversity follow the same pattern and are not different from those that can be expected from changes in richness alone.Main conclusions: Climate change is not expected to alter copepod functional traits distribution in the Mediterranean Sea, as the most and the least sensitive species are functionally redundant. Such redundancy should buffer the loss of ecosystem functions in Mediterranean zooplankton assemblages induced by climate change. Because the most negatively impacted species are affiliated to temperate regimes and share Atlantic biogeographic origins, our results are in line with the hypothesis of increasingly more tropical Mediterranean communities.

A recommendation – based on reviews by Hervé Capra and one anonymous reviewer – of the article: Nevoux M, Marchand F, Forget G, Huteau D, Tremblay J, and Destouches J-P. (2019) Field assessment of precocious maturation in salmon parr using ultrasound imaging. bioRxiv 425561, ver. 3 peer-reviewed and recommended by PCI Ecology. doi: 10.1101/425561

Held the 24th-26th of April 2019, in Villefranche-sur-Mer, France, this workshop gathered researchers interested in applying machine learning to identify and quantify functional traits of aquatic organisms from individual images.

Regionalisation aims at delimiting provinces within which physical conditions, chemical properties, and biological communities are reasonably homogeneous. This article proposes a synthesis of the many recent regionalisations of the open-sea regions of the Mediterranean Sea. The nine studies considered here defined regions based on different, and sometimes complementary, criteria: dynamics of surface chlorophyll concentration, ocean currents, three-dimensional hydrological and biogeochemical properties, or the distribution of organisms. Although they identified different numbers and patterns of homogeneous regions, their compilation in the epipelagic zone identifies nine consensus frontiers, eleven consensus regions with relatively homogeneous conditions, and four heterogeneous regions with highly dynamical conditions. The consensus frontiers and regions are in agreement with well-known hydrodynamical features of the Mediterranean Sea, which constrain the distribution of hydrological and ecological variables. The heterogeneous regions are rather defined by intense mesoscale activity. The synthesis proposed here could constitute a reference step for management actions and spatial planning, such as the application of the European Marine Strategy Framework Directive, and for future biogeochemical and ecological studies in the Mediterranean Sea.

Historically, the mortality of early‐life stages of marine fishes was supposed to be mostly caused by poor feeding during a critical period and aberrant drifting away from favorable recruitment areas. While fish larvae may display remarkable swimming abilities, Hjort's aberrant drift hypothesis has rarely been tested. In this study, we measured critical swimming speed (Ucrit) of settlement‐stage larvae of six coastal, warm temperate Mediterranean fish species, for which no data were previously available (Sparidae: Boops boops, Diplodus annularis, Spicara smaris, Spondyliosoma cantharus; Pomacentridae: Chromis chromis; Mugilidae sp.). Their swimming speeds were comparable with those of other temperate species, but also with the speeds of tropical species, which are considered as very fast swimmers. Mugilidae were the fastest (29.2 cm s−1), followed by Pomacentridae (22.8 cm s−1) and Sparidae (11.6 cm s−1). Most larvae swam in an inertial regime (Reynolds number > 1000). Those swimming speeds were then implemented in a Lagrangian model of the competency period of these species, set in the same area (the Ligurian Sea) and at the same time (June 2014) as the observations. In this modeling experiment, directional swimming strongly increased the proportion of successful settlers, independent of mesoscale hydrological structures. Fish larvae could settle on the coast from as far as tens of kilometers offshore, in just 4 d. These findings suggest that aberrant drift is unlikely to occur for strong swimming temperate larvae and show that larval behavior should be considered on equal footing with ocean currents when assessing larval fish dispersal.

Most demersal fishes undergo a dispersal phase as larvae, which strongly influences the connectivity among adult populations and, consequently, their genetic structure and replenishment opportunities. Because this phase is difficult to observe directly, it is frequently simulated through numerical models, most of which consider larvae as passive or only vertically migrating. However, in several locations, including the Mediterranean Sea, many species have been shown to swim fast and orient. Here we use a Lagrangian model to study connectivity patterns among three Mediterranean Marine Protected Areas (MPAs) and compare simulations in which virtual larvae are passive to simulations in which oriented swimming is implemented. The parameterization of behavior is based on observations for two groups of species of the family Sparidae: species with small larvae (i.e., 9-11 mm), displaying a maximum swimming speed of 6cm s(-1) and a pelagic larval duration of 13-19 days (e.g., Diplodus annularis L., Oblade melanura L.) and species with large larvae (i.e., 14-16 mm), displaying a maximum swimming speed of 10 cm s(-1) and a PLD of 28-38 days (e.g., Spondyliosoma cantharus L.). Including larval behavior in the model (i) increased the overall proportion of successful settlers, (ii) enhanced self-recruitment within the MPAs, but also (iii) increased the intensity, and (iv) widened the export of eggs and larvae (recruitment subsidy) from the MPAs; overall, it significantly changed connectivity patterns. These results highlight the need to gather the observational data that are required to correctly parameterize connectivity models.

The Paris Agreement target of limiting global surface warming to 1.5–2∘C compared to pre-industrial levels by 2100 will still heavily impact the ocean. While ambitious mitigation and adaptation are both needed, the ocean provides major opportunities for action to reduce climate change globally and its impacts on vital ecosystems and ecosystem services. A comprehensive and systematic assessment of 13 global- and local-scale, ocean-based measures was performed to help steer the development and implementation of technologies and actions toward a sustainable outcome. We show that (1) all measures have tradeoffs and multiple criteria must be used for a comprehensive assessment of their potential, (2) greatest benefit is derived by combining global and local solutions, some of which could be implemented or scaled-up immediately, (3) some measures are too uncertain to be recommended yet, (4) political consistency must be achieved through effective cross-scale governance mechanisms, (5) scientific effort must focus on effectiveness, co-benefits, disbenefits, and costs of poorly tested as well as new and emerging measures.

Ontogenetic Vertical Migration (OVM) is described by the changes in the depth of center of mass between early pre-flexion and more developed post-flexion larval stages. This behavior is mainly driven by taxa and ontogeny and is one of the prominent trait retaining larvae near their birth place. When incorporated into biophysical models of larval transport, this particular behavior contributes to change the outcome of the dispersal within sheered flows, decreasing dispersal kernels and changing connectivity patterns. Yet, only a few studies have reported the OVM of fishes and none has so far focused on investigating environmental forcing mechanisms. Here we test the hypothesis that fish larvae are responding to changes in environmental conditions by modifying the center of mass and range of their vertical distribution in the water column. We use an extensive ichthyoplankton survey around the island of Barbados to demonstrate that OVM changes not only between day and night times, but also with water chemistry. This study provides a basis for evaluating expected OVM and dispersal changes in a changing ocean.

The rise of in situ plankton imaging systems, particularly high-volume imagers such as the In Situ Ichthyoplankton Imaging System, has increased the need for fast processing and accurate classification tools that can identify a high diversity of organisms and nonliving particles of biological origin. Previous methods for automated classification have yielded moderate results that either can resolve few groups at high accuracy or many groups at relatively low accuracy. However, with the advent of new deep learning tools such as convolutional neural networks (CNNs), the automated identification of plankton images can be vastly improved. Here, we describe an image processing procedure that includes preprocessing, segmentation, classification, and postprocessing for the accurate identification of 108 classes of plankton using spatially sparse CNNs. Following a filtering process to remove images with low classification scores, a fully random evaluation of the classification showed that average precision was 84% and recall was 40% for all groups. Reliably classifying rare biological classes was difficult, so after excluding the 12 rarest taxa, classification accuracy for the remaining biological groups became > 90%. This method provides proof of concept for the effectiveness of an automated classification scheme using deep-learning methods, which can be applied to a range of plankton or biological imaging systems, with the eventual application in a variety of ecological monitoring and fisheries management contexts.

Most demersal fishes undergo a dispersal phase as larvae, which strongly influences the connectivity among adult populations and, consequently, their genetic structure and replenishment opportunities. Because this phase is difficult to observe directly, it is frequently simulated through numerical models, most of which consider larvae as passive or only vertically migrating. However, in several locations, including the Mediterranean Sea, several larval fish species have been shown to swim fast and orient. Here we use a Lagrangian model to study connectivity patterns among three Mediterranean Marine Protected Areas (MPAs) and compare simulations in which virtual larvae are passive to simulations in which oriented swimming is implemented. The parameterization of behavior is based on empirical data for two groups of species of the economically and ecologically important family Sparidae: species with small larvae (i.e., 9-11 mm), displaying a maximum swimming speed of 6 cm s−1 and a pelagic larval duration of 13-19 days (e.g., Diplodus annularis L., Oblada melanura L.) and species with large larvae (i.e., 14-16 cm), displaying a maximum swimming speed of 10cm s−1 and a PLD of 28-38 days (e.g., Spondyliosoma cantharus L.). Including larval behavior in the model (i) increased the overall proportion of successful settlers, (ii) enhanced self-recruitment within the MPAs, but also (iii) increased the intensity, and (iv) widened the export of eggs and larvae (recruitment subsidy) from the MPAs; overall, it significantly changed connectivity patterns. These results highlight the need to gather the observational data that are required to correctly parameterize connectivity models and efficiently manage marine resources.

Objectives: Pyogenic vertebral osteomyelitis (PVO) is a rare disease with possible severe complications (e.g. sepsis and spinal cord injury). In the 1990s, diagnostic delay (DD) was often extensive as PVO has a non-specific clinical spectrum, mostly afebrile with back pain, and access to magnetic resonance imaging (MRI) was not straightforward. Our aim was to perform a new study focusing on the clinical spectrum and DD of PVO and its associated factors.Method: This study examined a prospective cohort of 88 patients having PVO with microbiological identification between 15 November 2006 and 15 November 2010.Results: The 88 patients included in the study (female:male ratio 1:8) had a mean age of 64.1 years. The mean (sd) DD was 45.5 (50.4) days (range 2-280), and 46 patients (52.2%) were febrile at diagnosis. The main microorganism involved was Staphylococcus (n=45; 51.1%). In univariate and multivariate analyses, age > 75 years, antecedent back pain, involvement of bacteria, topography of PVO, and anti-inflammatory drug intake did not affect the DD, unlike a C-reactive protein (CRP) value > 63mg/L or a positive blood culture (DD lowered from 73to 17 days and from 90to 30 days, respectively). Conversely, X-ray investigation was associated with a longer DD (from 14 to 34.7days). Severity at diagnosis was not significantly different depending on the intake of anti-inflammatory drugs.Conclusions: Despite easier access to MRI, the DD for PVO remains long. One shortening factor is a high CRP value, which could be a useful diagnostic tool in case of back pain. Anti-inflammatory drugs seem to have no impact on DD and severity at diagnosis.

Just within the past two decades, studies on the early-life history stages of marine organisms have led to new paradigms in population dynamics. Unlike passive plant seeds that are transported by the wind or by animals, marine larvae have motor and sensory capabilities. As a result, marine larvae have a tremendous capacity to actively influence their dispersal. This is continuously revealed as we develop new techniques to observe larvae in their natural environment and begin to understand their ability to detect cues throughout ontogeny, process the information, and use it to ride ocean currents and navigate their way back home, or to a place like home. We present innovative in situ and numerical modeling approaches developed to understand the underlying mechanisms of larval transport in the ocean. We describe a novel concept of a Lagrangian platform, the Drifting In Situ Chamber (DISC), designed to observe and quantify complex larval behaviors and their interactions with the pelagic environment. We give a brief history of larval ecology research with the DISC, showing that swimming is directional in most species, guided by cues as diverse as the position of the sun or the underwater soundscape, and even that (unlike humans!) larvae orient better and swim faster when moving as a group. The observed Lagrangian behavior of individual larvae are directly implemented in the Connectivity Modeling System (CMS), an open source Lagrangian tracking application. Simulations help demonstrate the impact that larval behavior has compared to passive Lagrangian trajectories. These methodologies are already the base of exciting findings and are promising tools for documenting and simulating the behavior of other small pelagic organisms, forecasting their migration in a changing ocean.

When dividing the ocean, the aim is generally to summarise a complex system into a representative number of units, each representing a specific environment, a biological community or a socio-economical specificity. Recently, several geographical partitions of the global ocean have been proposed using statistical approaches applied to remote sensing or observations gathered during oceanographic cruises. Such geographical frameworks defined at a macroscale appear hardly applicable to characterise the biogeochemical features of semi-enclosed seas that are driven by smaller-scale chemical and physical processes. Following the Longhurst’s biogeochemical partitioning of the pelagic realm, this study investigates the environmental divisions of the Mediterranean Sea using a large set of environmental parameters. These parameters were informed in the horizontal and the vertical dimensions to provide a 3D spatial framework for environmental management (12 regions found for the epipelagic, 12 for the mesopelagic, 13 for the bathypelagic and 26 for the seafloor). We show that: (1) the contribution of the longitudinal environmental gradient to the biogeochemical partitions decreases with depth; (2) the partition of the surface layer cannot be extrapolated to other vertical layers as the partition is driven by a different set of environmental variables. This new partitioning of the Mediterranean Sea has strong implications for conservation as it highlights that management must account for the differences in zoning with depth at a regional scale.

With global climate change altering marine ecosystems, research on plankton ecology is likely to navigate uncharted seas. Yet, a staggering wealth of new plankton observations, integrated with recent advances in marine ecosystem modelling, may shed light on marine ecosystem structure and functioning. A EuroMarine foresight workshop on the “Impact of climate change on the distribution of plankton functional and phylogenetic diversity” (PlankDiv) identified five grand challenges for future plankton diversity and macroecology research: 1) What can we learn about plankton communities from the new wealth of high-throughput ‘omics’ data? 2) What is the link between plankton diversity and ecosystem function? 3) How can species distribution models be adapted to represent plankton biogeography? 4) How will plankton biogeography be altered due to anthropogenic climate change? and 5) Can a new unifying theory of macroecology be developed based on plankton ecology studies? In this review, we discuss potential future avenues to address these questions, and challenges that need to be tackled along the way.

Although future increases in water temperature and future changes in regional circulation are expected to have great impacts on the pelagic food-web, estimates focusing on community-level shifts are still lacking for the planktonic compartment. By combining statistical niche models (or species distribution models) with projections from a regional circulation model, the impact of climate change on copepod epipelagic communities is assessed for the Mediterranean Sea. Habitat suitability maps are generated for 106 of the most abundant copepod species to analyze emerging patterns of diversity at the community level. Using variance analysis, we also quantified the uncertainties associated to our modeling strategy (niche model choice, CO2 emission scenario, boundary forcings of the circulation model). Comparing present and future projections, changes in species richness (alpha diversity) and in community composition (beta diversity, decomposed into turnover and nestedness component) are calculated. Average projections show that copepod communities will mainly experience turn-over processes, with little changes in species richness. Species gains are mainly located in the Gulf of Lions, the Northern Adriatic and the Northern Aegean seas. However, projections are highly variable, especially in the Eastern Mediterranean basin. We show that such variability is mainly driven by the choice of the niche model, through interactions with the CO2 emission scenario or the boundary forcing of the circulation model can be locally important. Finally, the possible impact of the estimated community changes on zooplanktonic functional and phylogenetic diversity is also assessed. We encourage the enlargement of this type of study to other components of the pelagic food-web, and argue that niche models' outputs should always be given along with a measure of uncertainty, and explained in light of a strong theoretical background.

Worldwide increase in seawater temperature represents one of the major threats affecting corals, which experience bleaching, and thereafter a significant decrease in photosynthesis and calcification. The impact of bleaching on coral physiology may be exacerbated when coupled with eutrophication, i.e., increasing plankton, inorganic nutrient concentrations, sedimentation and turbidity due to coastal urbanization. Whereas zooplankton provision (heterotrophy) may alleviate the negative consequences of thermal stress, inorganic nutrient supply may exacerbate them, which creates a paradox. Our experimental study aims to disentangle the effects of these two components of eutrophication on the physiological response of Turbinaria reniformis subject to normal and to a short-term temperature increase. Additionally, three different inorganic nutrient ratios were tested to assess the influence of nutrient stoichiometry on coral physiology: control (ambient SW 0.5 mu M N and 0.1 mu M P), N only (ambient + 2 mu M N) and N+P (ambient + 2 mu M N and + 0.5 mu M P). Our results show a deleterious effect of a 2 mu M nitrate enrichment alone (N) on coral photosynthetic processes under thermal stress as well as on calcification rates when associated with heterotrophy. On the contrary, a coupled nitrate and phosphorus enrichment (N+P) maintained coral metabolism and calcification during thermal stress and enhanced them when combined with heterotrophy. Broadly, our results shed light on the tight relationship existing between inorganic nutrient availability and heterotrophy. Moreover, it assesses the relevance of N: P stoichiometry as a determining factor for the health of the holobiont that may be adapted to specific nutrient ratios in its surrounding environment.

Imaging systems were developed to explore the fine scale distributions of plankton (<10 m), but they generate huge datasets that are still a challenge to handle rapidly and accurately. So far, imaged organisms have been either classified manually or pre-classified by a computer program and later verified by human operators. In this paper, we post-process a computer-generated classification, obtained with the common ZooProcess and PlanktonIdentifier toolchain developed for the ZooScan, and test whether the same ecological conclusions can be reached with this fully automatic dataset and with a reference, manually sorted, dataset. The Random Forest classifier outputs the probabilities that each object belongs in each class and we discard the objects with uncertain predictions, i.e. under a probability threshold defined based on a 1% error rate in a self-prediction of the learning set. Keeping only well-predicted objects enabled considerable improvements in average precision, 84% for biological groups, at the cost of diminishing recall (by 39% on average). Overall, it increased accuracy by 16%. For most groups, the automatically-predicted distributions were comparable to the reference distributions and resulted in the same size-spectra. Automatically-predicted distributions also resolved ecologically-relevant patterns, such as differences in abundance across a mesoscale front or fine-scale vertical shifts between day and night. This post-processing method is tested on the classification of plankton images through Random Forest here, but is based on basic features shared by all machine learning methods and could thus be used in a broad range of applications.

Our understanding of the internal dynamics of the Earth is largely based on images of seismic velocity variations in the mantle obtained with global tomography. However, our ability to image the mantle is severely hampered by a lack of seismic data collected in marine areas. Here we report observations made under different noise conditions (in the Mediterranean Sea, the Indian and Pacific Oceans) by a submarine floating seismograph, and show that such floats are able to fill the oceanic data gap. Depending on the ambient noise level, the floats can record between 35 and 63% of distant earthquakes with a moment magnitude MZ6.5. Even magnitudes o6.0 can be successfully observed under favourable noise conditions. The serendipitous recording of an earthquake swarm near the Indian Ocean triple junction enabled us to establish a threshold magnitude between 2.7 and 3.4 for local earthquakes in the noisiest of the three environments.

Mortality is very high during the pelagic larval phase of fishes but the factors that determine recruitment success remain unclear and hard to predict. Because of their bipartite life history , larvae of coastal species have to head back to the shore at the end of their pelagic episode , to settle. These settlement-stage larvae are known to display strong sensory and motile abilities, but most work has been focused on tropical, insular environments and on the influence of coast-related cues on orientation. In this study we quantified the in situ orientation behavior of settlement-stage larvae in a temperate region, with a continuous coast and a dominant along-shore current, and inspected both coast-dependent and independent cues. We tested six species: one Pomacentridae, Chromis chromis, and five Sparidae, Boops boops, Diplodus annularis, Oblada melanura, Spicara smaris and Spondyliosoma cantharus. Over 85% of larvae were highly capable of keeping a bearing, which is comparable to the orientation abilities of tropical species. Sun-related cues influenced the precision of bearing-keeping at individual level. Three species, out of the four tested in sufficient numbers , oriented significantly relative to the sun position. These are the first in situ observations demonstrating the use of a sun compass for orientation by wild-caught settlement-stage larvae. This mechanism has potential for large-scale orientation of fish larvae globally.

Theory and some empirical evidence suggest that groups of animals orient better than isolated individuals. We present the first test of this hypothesis for pelagic marine larvae, at the stage of settlement, when orientation is critical to find a habitat. We compare the in situ behaviour of individuals and groups of 10–12 Chromis atripectoralis (reef fish of the family Pomacentridae), off Lizard Island, Great Barrier Reef. Larvae are observed by divers or with a drifting image recording device. With both methods, groups orient cardinally while isolated individuals do not display significant orientation. Groups also swim on a 15% straighter course (i.e. are better at keeping a bearing) and 7% faster than individuals. A body of observations collected in this study suggest that enhanced group orientation emerges from simple group dynamics rather than from the presence of more skilful leaders.

PERSEUS project aims to identify the most relevant pressures exerted on the ecosystems of the Southern European Seas (SES), highlighting knowledge and data gaps that endanger the achievement of SES Good Environmental Status (GES) as mandated by the Marine Strategy Framework Directive (MSFD). A complementary approach has been adopted, by a meta-analysis of existing literature on pressure/impact/knowledge gaps summarized in tables related to the MSFD descriptors, discriminating open waters from coastal areas. A comparative assessment of the Initial Assessments (IAs) for five SES countries has been also independently performed. The comparison between meta-analysis results and IAs shows similarities for coastal areas only. Major knowledge gaps have been detected for the biodiversity, marine food web, marine litter and underwater noise descriptors. The meta-analysis also allowed the identification of additional research themes targeting research topics that are requested to the achievement of GES.

Environmental changes resulting from anthropogenic CO2 emissions occur at global and local levels and have potentially harmful effects, particularly for calcifying taxa in the marine environment. A time series of pteropod abundance covering the period 1967-2003 was isolated from the Point B (northwestern Ligurian Sea) zooplankton time series. Inter-and intra-annual changes in the abundance of 3 families (Limacinidae, Cavoliniidae and Creseidae) were compared with the copepod time series to identify any differential effects driven by ocean acidification and temperature. pH values were hind-cast from total alkalinity estimated from local temperature and salinity measurements, and atmospheric CO2 taken from the Mauna Loa time series. Although surface waters were supersaturated with respect to aragonite throughout the study period, it is estimated that pH declined by 0.05 units. All pteropod groups displayed a trend of increasing abundance, suggesting that any deleterious effect of declining pH(T) in the range of 0.05 units has not caused sufficient reductions in fitness as to decrease local abundances between 1967-2003. Pteropod populations are influenced by inter-annual changes in summer temperatures. Spectral analysis identified a similar to 14 yr periodic oscillation in sea surface temperature. Similarly timed oscillations in abundance are present for all pteropod families but not for copepods, indicating a possible influence of the North Atlantic quasi-decadal mode on pteropod populations. While laboratory studies have shown pteropods to be sensitive to changes in pH, this analysis suggests that local and regional scale drivers have had a greater effect on pteropod populations in the northwestern Mediterranean Sea in recent decades. It should be noted that pH changes in laboratory studies exceed 0.05 pH units and that the saturation state with respect to aragonite (Oar) is usually much lower than that of the Mediterranean.

A network of moored hydrophones is an effective way of monitoring seismicity of oceanic ridges since it allows detection and localization of underwater events by recording generated T waves. The high cost of ship time necessitates long periods (normally a year) of autonomous functioning of the hydrophones, which results in very large data sets. The preliminary but indispensable part of the data analysis consists of identifying all T wave signals. This process is extremely time consuming if it is done by a human operator who visually examines the entire database. We propose a new method for automatic signal discrimination based on the Gradient Boosted Decision Trees technique that uses the distribution of signal spectral power among different frequency bands as the discriminating characteristic. We have applied this method to automatically identify the types of acoustic signals in data collected by two moored hydrophones in the North Atlantic. We show that the method is capable of efficiently resolving the signals of seismic origin with a small percentage of wrong identifications and missed events: 1.2% and 0.5% for T waves and 14.5% and 2.8% for teleseismic P waves, respectively. In addition, good identification rates for signals of other types (iceberg and ship generated) are obtained. Our results indicate that the method can be successfully applied to automate the analysis of other (not necessarily acoustic) databases provided that enough information is available to describe statistical properties of the signals to be identified.

Mesoscale fronts occur frequently in many coastal areas and often are sites of elevated productivity; however, knowledge of the fine-scale distribution of zooplankton at these fronts is lacking, particularly within the mid-trophic levels. Furthermore, small (<13 cm) gelatinous zooplankton are ubiquitous, but are under-studied, and their abundances underestimated due to inadequate sampling technology. Using the In Situ Ichthyoplankton Imaging System (ISIIS), we describe the fine-scale distribution of small gelatinous zooplankton at a sharp salinity-driven front in the Southern California Bight. Between 15 and 17 October 2010, over 129000 hydromedusae, ctenophores, and siphonophores within 44 taxa, and nearly 650000 pelagic tunicates were imaged in 5450 m(3) of water. Organisms were separated into 4 major assemblages which were largely associated with depth-related factors. Species distribution modeling using boosted regression trees revealed that hydromedusae and tunicates were primarily associated with temperature and depth, siphonophores with dissolved oxygen (DO) and chlorophyll a fluorescence, and ctenophores with DO. The front was the least influential out of all environmental variables modeled. Additionally, except for 6 taxa, all other taxa were not aggregated at the front. Results provide new insights into the biophysical drivers of gelatinous zooplankton distributions and the varying influence of mesoscale fronts in structuring zooplankton communities.

Understanding larval dispersal requires knowledge of whether larvae in situ have orientated swimming, and how this varies temporally and spatially. Orientation of >300 settlement-stage larvae of Chromis atripectoralis (Pomacentridae) measured over 1998-2008 by divers near Lizard Island, Great Barrier Reef was consistent. All 10 data sets had southerly orientation at all locations; 94% of larvae swam directionally. Median bearings east and west of Lizard Island were 166 degrees and 170 degrees, respectively. Orientation precision was significantly higher under sunny than cloudy skies. Similar mean bearings were obtained in 2008 with more than 125 larvae observed in a drifting in situ chamber (DISC). Orientation varied with time of day. In sunny conditions, precision was weakly, significantly correlated with time of day, but not solar elevation; however, a greater proportion of larvae was significantly directional at low (<50 degrees) than at high (>50 degrees) solar elevation. Mean bearing and time of day were weakly, but significantly correlated. Bearings changed from SE during most of the day to SSW in the late afternoon, with distribution of bearings significantly different. Location-independent but diurnally-dependent orientation implies that larvae used celestial cues for orientation. Of 91 Pomacentrus lepidogenys larvae that were followed by divers, 89% swam directionally, but orientation differed among locations and years. DISC results with 20 larvae were similar. The similarity of orientation returned by different methods used on 2 fish species corroborates previous results using diver following. Both methods are useful for the study of larval-fish orientation in situ: each has advantages and limitations, and their use is complementary.

The behavior of reef fish larvae, equipped with a complex toolbox of sensory apparatus, has become a central issue in understanding their transport in the ocean. In this study pelagic reef fish larvae were monitored using an unmanned open-ocean tracking device, the drifting in-situ chamber (DISC), deployed sequentially in oceanic waters and in reef-born odor plumes propagating offshore with the ebb flow. A total of 83 larvae of two taxonomic groups of the families Pomacentridae and Apogonidae were observed in the two water masses around One Tree Island, southern Great Barrier Reef. The study provides the first in-situ evidence that pelagic reef fish larvae discriminate reef odor and respond by changing their swimming speed and direction. It concludes that reef fish larvae smell the presence of coral reefs from several kilometers offshore and this odor is a primary component of their navigational system and activates other directional sensory cues. The two families expressed differences in their response that could be adapted to maintain a position close to the reef. In particular, damselfish larvae embedded in the odor plume detected the location of the reef crest and swam westward and parallel to shore on both sides of the island. This study underlines the critical importance of in situ Lagrangian observations to provide unique information on larval fish behavioral decisions. From an ecological perspective the central role of olfactory signals in marine population connectivity raises concerns about the effects of pollution and acidification of oceans, which can alter chemical cues and olfactory responses.

Pleuragramma antarcticum is a key component of the neritic assemblages in the Antarctic coastal waters. Larvae of this species were sampled from 2008 to 2011 in the Dumont d'Urville Sea (East Antarctica). The lipid class composition [triacylglycerols (TAG), cholesterol (Chol) and polar lipids (PL)] of larvae was measured to assess the larval condition. The total amount of lipids was linearly related to the quantity of structural polar lipids, suggesting that growth is favored over lipid storage. The TAG:Chol ratio showed interannual variability in the condition of fish larvae, probably related to prey availability. Nevertheless, the essential fatty acids composition of polar lipids illustrates that larvae with low levels of TAG:Chol could be either growing or under starvation. Only the combination of a low TAG:Chol ratio and low polar lipids content, which can also be mobilized during starvation periods, allowed identification of larvae in poor condition. This lipid condition index should be of great assistance to evaluate the probability of survival of P. antarcticum larvae in long-term monitoring. It has widespread applicability and should also be useful in the diagnosis of nutritional condition in other species.

Experiments were performed at lab scale in order to test the possibility to grow microalgae with CO2 from gaseous effluent of cement industry. Four microalgal species (Dunaliella tertiolecta, Chlorella vulgaris, Thalassiosira weissflogii, and Isochrysis galbana), representing four different phyla were grown with CO2 enriched air or with a mixture of gasses mimicking the composition of a typical cement flue gas (CFG). In a second stage, the culture submitted to the CFG received an increasing concentration of dust characteristic of cement industry. Results show that growth for the four species is not affected by the CFG. Dust added at realistic concentrations do not have any impact on growth. For dust concentrations in two ranges of magnitude higher, microalgae growth was inhibited.

The CCAMLR MPA Workshop on Marine Protected Areas, held in Brest (France) in August 2011, recommended that the Scientific Committee considers supporting three technical workshops including one specific to Planning Domain 5. Planning Domain 5 includes Marion and Prince Edward Islands, the Del Cano Rise and the Crozet Archipelago in the north. It also includes the Ob and Lena seamounts. The workshop focusing on Planning Domain 5 was held in St Pierre, La Réunion, France from 15th May to 18th May, 2012 at the headquarters of TAAF (French Southern and Antarctic Territories). It followed a meeting on the northern part of Planning Domain 5, which was held in South Africa in 2008, organized and funded by WWF South Africa and known as Del Cano 1. The intention of the CCAMLR workshop was to study the ecological values and the use of the marine environment and to identify possible threats that might occur in this area. It extended the Del Cano 1 study spatially and also ecologically to include the benthic and pelagic realms. Identification of objectives for Conservation Planning and future research were discussed in relation to national and international projects. Depending on the availability of data, the approach was based on mapping species distributions (either observed data or predictions for species or community presence/abundance based on environmental factors). Various national and international datasets were used including data from CCAMLR. However, South African and French data relevant to the Planning Domain 5 and surrounding domains were a major focus in the workshop because these CCAMLR member nations are the major scientific actors in this region. Species distributions were visualized by the mean of a Geographic Information System. Available Norwegian data from the Bouvetøya region were also discussed, but this region is less studied compared to the Planning Domain 5. The workshop provided benthic and pelagic abiotic classifications of the Planning Domain using geographic and oceanographic features. Distributions of plankton, mesopelagic fish and top predators were consistent with the abiotic regionalization showing latitudinal patterns of communities for the pelagic species. The importance of frontal zones such as the Antarctic Polar Front and especially the Subantarctic Front were highlighted. North of the CCAMLR area, the Agulhas Return Current has a strong influence on this region. The latitudinal zonation of bioregions according to frontal zones may be influenced by climate change. This will have consequences for marine bird and mammal populations as it will change the habitat of their main pelagic prey species (e.g. euphausiids, squids, mesopelagic fish, etc.). The working group concluded that ecoregionalisation has to be conducted at the scale of plateaus which includes Prince Edward Islands, Del Cano Rise and Crozet Islands i.e. a more detailed level than what has been done to date. High productive pelagic areas must be considered in relation to the bathymetry, iron enrichment, fronts and island mass effects, which contrast with high nutrient low chlorophyll areas farther south. Ichtyofauna and benthos were described as being characteristic of the subantarctic zone with some species being endemic. However, cryptic benthic species have not yet been studied. The French and South African islands support substantial colonies of seabirds and seals, which for several species have global importance. For example, the Crozet and Prince Edward Islands together host the entire population of Crozet shag, about 70% of the world population of wandering albatross, 54% of king penguin, 33% of Indian yellow-nosed albatross, 33% of subantarctic fur seal, 27% of sooty albatross and 21% of the world’s southern rockhopper penguin. The high productivity in the vicinity of the islands, together with the large aggregations of seabirds and seals found at the islands, attract various other animals, e.g. several cetaceans, to their vicinity. The populations of several seabirds that breed at the islands have decreased. There is accumulating evidence that decreases of albatrosses and petrels have been substantially influenced by by-catch mortality in fisheries, whereas decreases in some penguins are probably attributable to decreased availability of prey that may have been caused by environmental change. Although the islands themselves enjoy a protected status and fishing is at present excluded within 12 nautical miles of the islands, providing some protection to inshore-foraging species, many of the seabirds and seals range well beyond the immediate precincts of the islands. Some circumnavigate Antarctica and others move to the north well beyond the CCAMLR convention area. Hence, many seabirds and seals are affected by human activities, and almost certainly environmental change, in other CCAMLR domains as well as in regions of the high seas that are administered by other Regional Fisheries Management Organisations. In particular the CCAMLR Planning domains that neighbour Domain 5 are of importance, as is the southern region of the Indian Ocean Tuna Commission (IOTC, FAO Area 51). Human activities in these other areas adversely influence the conservation status of animals from Planning Domain 5, as is the case with by-catch mortality, it will be necessary for CCAMLR to work in close association with other Regional Fisheries Management Organisations (e.g. IOTC), treaties (e.g. Agreement on the Conservation of Albatrosses and Petrels, ACAP) and conservation organisations (e.g. BirdLife International) to achieve a favourable conservation status for species that are at present Threatened or Near Threatened. The problem is sometimes compounded by albatrosses and petrels segregating their at-sea distributions by sex or age or both, so that components of populations may suffer particularly high mortality leading, e.g., to sex imbalances or inadequate recruitment into breeding populations. Preliminary models suggest that both topographical (e.g. plateau and rises) and oceanographic (e.g. locations of fronts) features play important roles in defining good foraging grounds for some wide-ranging predators. Whereas topographical features are permanent, the locations of oceanographic features may be changing, thereby presenting a greater challenge for spatial conservation planning. For some albatrosses (notably Thalassarche spp.) and penguins (notably Eudyptes spp.) there is accumulating evidence that populations and species may segregate their feeding grounds, which also will need to be accounted for in any form of spatial conservation planning. The Working group noted set of preliminary strategic points essential to Systematic Conservation Planning, which include accounting for ecological relationships with surrounding areas (Bouvet to the West, Kerguelen to the East and East Antarctica to the South). The working group concluded that subtropical areas north of the Planning Domain 5 should be included in the planning, because of the spatial range covered by top predators, and also because the limit of the CCAMLR area cuts across the EEZs of both the Prince Edward and the Crozet Islands, as well as the Del Cano Rise. Strategic points include: First of all, biodiversity features needed to be mapped. The workshop concentrated on this objective most of the time. Second, biodiversity targets need to be determined. France will see how to adapt the ones that were used by Lombard et al. (2007) for the EEZ of the Prince Edward Islands, to the Crozet Islands. For the high seas, the working group recommended that the definition of targets should be discussed by EMM and the MPA circumpolar workshop. Third, the Working Group started to evaluate pressures and areas of research that need to be defined (spawning and nursery areas, bycatch, interactions with killer whales, etc.). The involvement of stakeholders was discussed, including fishing industries, NGOs, other relevant treaties and CCAMLR members for the high sea areas. This general theme has to be more specifically discussed during the circumpolar workshop and at the SC. The workshop also considered MPAs existing under national jurisdiction in the Planning Domain, procedures to extend them, and the need for strengthened cooperation between CCAMLR and other relevant legal instruments, organizations and initiatives. To achieve these goals, research and monitoring were discussed under three headings: (1) census of biodiversity, (2) ecoregionalisation classification and (3) monitoring. Such research would make up for a current lack of data, e.g. with regard to the benthos (deep and shallow), the mesopelagic zone and plankton. With regard to monitoring ecological processes, especially at the northern limits of the CCAMLR area, CCAMLR may wish to broaden the lists of species and environmental parameters that are monitored, to consider parameters of species that may best reflect change associated with global warming and, if necessary, develop protocols for any new parameters to be monitored. This would naturally include the use of the Continuous Plankton Recorder and tracking for birds, seals and mammals.

Climate may act on the dispersal and connectivity of marine populations through changes in the oceanic circulation and temperature, and by modifying species' prey and predator distributions. As dispersal and connectivity remain difficult to assess in situ, a first step in studying the effects of climate change can be achieved using biophysical models. To date, only a few biophysical models have been used for this purpose. Here we review these studies and also include results from other recent modelling efforts. We show that increased sea temperature, a major change expected under climate warming, may impact dispersal and connectivity patterns via changes in reproductive phenology (e.g., shift in the spawning season), transport (e.g., reduced pelagic larval duration under faster development rates), mortality (e.g., changes in the exposure to lethal temperatures), and behaviour (e.g.. increased larval swimming speed). Projected changes in circulation are also shown to have large effects on the simulated dispersal and connectivity patterns. Although these biophysical modelling studies are useful preliminary approaches to project the potential effects of climate change, we highlight their current limitations and discuss the way forward, in particular the need for adequate coupled hydrodynamic-biogeochemical simulations using atmospheric forcing from realistic climate change scenarios. (C) 2010 Elsevier Ltd. All rights reserved.

A reference collection of COI barcode (650 bp) for the Pacific Society Islands has been constituted for 22 species of Acanthuridae and 16 species of Holocentridae. Divergence between congeneric species was on average 20-fold to 87-fold higher than divergence between conspecific sequences and this set of DNA-identifiers was used to identify 40 larvae of both families. All larvae sequenced could be identified to species using DNA-barcodes. Pools of larvae constitute multi-specific assemblages and no additional species compared to adult reef communities were sampled in larval pools, suggesting that the larval assemblages originated from adult communities on neighboring reefs.

All early models of the pelagic phase of coastal organisms made the simplifying assumption that larvae could be treated as passive particles in a flow. This was justified in the early nineties because of both computational limitations and lack of knowledge about the behavioral ecology of marine larvae. However, for more than a decade now, the swimming abilities of coral reef fishes, in particular, as well as their vertical distribution, have been investigated using several methods. All agree on the tremendous swimming speeds and endurance displayed by fish larvae, on their ability to influence their vertical and horizontal distributions and on the potential impact of these factors on connectivity. Yet, no numerical model integrates this kind of behavior more extensively than for the last instants of larval life. We present two numerical models of the whole larval phase which both feature larval behavior. One explicitly integrates larval swimming in a mesoscale environment, around an island, and estimates the impact of swimming on self-recruitment. Because we still know very little about the orientation behavior of larvae in oceanic waters, we deduce their swimming decisions from an optimization method with realistic constraints and a biologically sensible objective (i.e. recruitment). Swimming along these optimal trajectories substantially enhances the possibility of self-recruitment compared to a passive scenario. The other model examines the influence of statistically modeled vertical distributions of larvae in the context of many inter-connected fish populations in the Caribbean. The distribution is shown to markedly modify the connectivity matrix. Overall, both models show that larval behavior has a great influence during the pelagic phase. Many studies show that larvae probably use their swimming abilities to reduce rather than to enhance dispersal, and our findings demonstrate that they can do so very efficiently.

La plupart des organismes marins démersaux présentent une phase larvaire pélagique avant le recrutement dans la population adulte. Cet épisode pélagique est souvent la seule opportunité de dispersion au cours du cycle de vie. De ce fait, il structure les connections entre populations, qui régissent la dynamique et la composition génétique des métapopulations benthiques. Cependant, ces "larves" ne sont pas de simples ébauches des adultes, dispersées au gré des courants en attendant leur métamorphose. Ce sont des organismes souvent très spécifiquement adaptés à leur milieu. Dans cette thèse nous nous sommes efforcés d'évaluer l'impact du comportement des larves lors de la phase pélagique. Nous nous sommes focalisés sur les larves de poissons (coralliens plus spécifiquement) dont les capacités sensorielles et motrices sont particulièrement élevées. Des approches expérimentales ont été développées afin de quantifier leur orientation et leur nage in situ. Grâce à une observation synchrone des caractéristiques physiques du milieu et de la distribution des larves lors d'une campagne océanographique, nous avons tenté de caractériser leur distribution en trois dimensions dans le milieu pélagique, afin de comprendre les interactions physico-biologiques déterminant le recrutement. Enfin, une approche de modélisation novatrice, faisant appel à des concepts de minimisation des coûts et de maximisation des bénéfices habituellement utilisés en économie ou en théorie de l'approvisionnement optimal, a permis d'intégrer le comportement des larves aux modèles Lagrangiens de dispersion

The swimming behaviour of 534 coral reef fish larvae from 27 species was explored at Moorea Island (French Polynesia) while they searched for a suitable settlement habitat, on the first night of their lagoon life. Most larvae swam actively (74%) and avoided the bottom (77%). A significant relationship was highlighted between the vertical position of larvae in the water column and the distance they travelled from lagoon entrance to settlement habitat: larvae swimming close to the surface settled farther away on the reef than bottom-dwelling larvae.

A central question in marine ecology today is to understand the spatial scales over which populations are connected by larval dispersal. Although coral reef fish larvae develop strong behavioral capabilities during the processes of dispersal (e.g., vertical migration, swimming), the influence of these capabilities on survival depends on the ability of larvae to orient in the open ocean. Yet, behavioral mechanisms by which reef fish larvae achieve successful recruitment from blue waters to coral reefs are still unknown. We describe a novel system designed to detect and quantify the orientation of larval coral reef fish in the pelagic environment, where they have no apparent frame of reference to navigate. The Orientation With No Frame Of Reference (OWNFOR) system is deployed at sea and drifts while videotaping the movement of a larva placed within a clear, circular arena. The data are then treated with a combination of open-source programs that track particles and analyze orientation using circular statistics. We demonstrate that (1) this system successfully detects orientation behavior and (2) the shape of the behavioral arena and appropriate statistical treatments minimize the impact of the semi-enclosure on quantifying larval bearing. This observational approach provides a means of assessing the abilities of larvae to orientate during ontogeny and of testing cues, representing a breakthrough in the field of larval ecology. Such behavioral data will provide critical inputs to a new generation of biophysical larval dispersal models that are vital to achieve a better understanding of larval connectivity in marine systems.