A Deep-Learning Based Pipeline for Estimating the Abundance and Size of Aquatic Organisms in an Unconstrained Underwater Environment from Continuously Captured Stereo Video.

The utilization of stationary underwater cameras is a modern and well-adapted approach to provide a continuous and cost-effective long-term solution to monitor underwater habitats of particular interest. A common goal of such monitoring systems is to gain better insight into the dynamics and condition of populations of various marine organisms, such as migratory or commercially relevant fish taxa. This paper describes a complete processing pipeline to automatically determine the abundance, type and estimate the size of biological taxa from stereoscopic video data captured by the stereo camera of a stationary Underwater Fish Observatory (UFO). A calibration of the recording system was carried out in situ and, afterward, validated using the synchronously recorded sonar data. The video data were recorded continuously for nearly one year in the Kiel Fjord, an inlet of the Baltic Sea in northern Germany. It shows underwater organisms in their natural behavior, as passive low-light cameras were used instead of active lighting to dampen attraction effects and allow for the least invasive recording possible. The recorded raw data are pre-filtered by an adaptive background estimation to extract sequences with activity, which are then processed by a deep detection network, i.e., Yolov5. This provides the location and type of organisms detected in each video frame of both cameras, which are used to calculate stereo correspondences following a basic matching scheme. In a subsequent step, the size and distance of the depicted organisms are approximated using the corner coordinates of the matched bounding boxes. The Yolov5 model employed in this study was trained on a novel dataset comprising 73,144 images and 92,899 bounding box annotations for 10 categories of marine animals. The model achieved a mean detection accuracy of 92.4%, a mean average precision (mAP) of 94.8% and an F1 score of 93%.

Extensive oceanic mesopelagic habitat use of a migratory continental shark species.

The identification of movement and behaviour patterns, as well as inter- and intra-population connectivity is crucial in order to implement effective and functional management and conservation measures for threatened migratory species such as tope (Galeorhinus galeus). Yet, previous studies struggled to elucidate clear and consistent movement and depth usage patterns of adult tope in the Northeast Atlantic, suggesting a high plasticity in the migration and behaviour. We deployed pop-up satellite archival tags on adult tope during their seasonal summer aggregations in the inner German Bight of the south-eastern North Sea and near a presumed mating site in southwest Scotland. Depth distribution and migration pathways were derived from time series data with location processing. Four individuals followed migration trajectories leaving coastal areas and crossed the European shelf slope into oceanic areas of the Northeast Atlantic, remaining fully pelagic for the rest of the deployment duration. These sharks showed far-ranging migration trajectories and undertook regular and frequent diel vertical migrations, reaching daytime depths of over 700 m. Vertical migration patterns closely overlapped with biological mesopelagic habitat structures and closely tracked the diel migration of organisms from deep scattering layers derived from hydroacoustic recordings. It is hypothesized that adult tope regularly utilize oceanic habitats, foraging on mesopelagic layers in an environment generally considered of low prey density.

Sea-ice derived meltwater stratification slows the biological carbon pump: results from continuous observations.

The ocean moderates the world's climate through absorption of heat and carbon, but how much carbon the ocean will continue to absorb remains unknown. The North Atlantic Ocean west (Baffin Bay/Labrador Sea) and east (Fram Strait/Greenland Sea) of Greenland features the most intense absorption of anthropogenic carbon globally; the biological carbon pump (BCP) contributes substantially. As Arctic sea-ice melts, the BCP changes, impacting global climate and other critical ocean attributes (e.g. biodiversity). Full understanding requires year-round observations across a range of ice conditions. Here we present such observations: autonomously collected Eulerian continuous 24-month time-series in Fram Strait. We show that, compared to ice-unaffected conditions, sea-ice derived meltwater stratification slows the BCP by 4 months, a shift from an export to a retention system, with measurable impacts on benthic communities. This has implications for ecosystem dynamics in the future warmer Arctic where the seasonal ice zone is expected to expand.

Seasonal and diel cycles of fin whale acoustic occurrence near Elephant Island, Antarctica.

This study investigates the relevance of the Elephant Island (EI) region for Southern Hemisphere fin whales (Balaenoptera physalus) in their annual life cycle. We collected 3 years of passive acoustic recordings (January 2013 to February 2016) northwest of EI to calculate time series of fin whale acoustic indices, daily acoustic occurrence, spectrograms, as well as the abundance of their 20 Hz pulses. Acoustic backscatter strength, sea ice concentration and chlorophyll-a composites provided concurrent environmental information for graphic comparisons. Acoustic interannual, seasonal and diel patterns together with visual information and literature resources were used to define the period of occupancy and to infer potential drivers for their behaviour. Spectral results suggest that these fin whales migrate annually to and from offshore central Chile. Acoustic data and visual information reveal their arrival at EI in December to feed without producing their typical 20 Hz pulse. For all 3 years, acoustic activity commences in February, peaks in May and decreases in August, in phase with the onset of their breeding season. Our results emphasize the importance of EI for fin whales throughout most of the year. Our recommendation is to consider EI for establishing a marine protected area to expedite the recovery of this vulnerable species.

Thick-shelled, grazer-protected diatoms decouple ocean carbon and silicon cycles in the iron-limited Antarctic Circumpolar Current.

Diatoms of the iron-replete continental margins and North Atlantic are key exporters of organic carbon. In contrast, diatoms of the iron-limited Antarctic Circumpolar Current sequester silicon, but comparatively little carbon, in the underlying deep ocean and sediments. Because the Southern Ocean is the major hub of oceanic nutrient distribution, selective silicon sequestration there limits diatom blooms elsewhere and consequently the biotic carbon sequestration potential of the entire ocean. We investigated this paradox in an in situ iron fertilization experiment by comparing accumulation and sinking of diatom populations inside and outside the iron-fertilized patch over 5 wk. A bloom comprising various thin- and thick-shelled diatom species developed inside the patch despite the presence of large grazer populations. After the third week, most of the thinner-shelled diatom species underwent mass mortality, formed large, mucous aggregates, and sank out en masse (carbon sinkers). In contrast, thicker-shelled species, in particular Fragilariopsis kerguelensis, persisted in the surface layers, sank mainly empty shells continuously, and reduced silicate concentrations to similar levels both inside and outside the patch (silica sinkers). These patterns imply that thick-shelled, hence grazer-protected, diatom species evolved in response to heavy copepod grazing pressure in the presence of an abundant silicate supply. The ecology of these silica-sinking species decouples silicon and carbon cycles in the iron-limited Southern Ocean, whereas carbon-sinking species, when stimulated by iron fertilization, export more carbon per silicon. Our results suggest that large-scale iron fertilization of the silicate-rich Southern Ocean will not change silicon sequestration but will add carbon to the sinking silica flux.

Deep carbon export from a Southern Ocean iron-fertilized diatom bloom.

Fertilization of the ocean by adding iron compounds has induced diatom-dominated phytoplankton blooms accompanied by considerable carbon dioxide drawdown in the ocean surface layer. However, because the fate of bloom biomass could not be adequately resolved in these experiments, the timescales of carbon sequestration from the atmosphere are uncertain. Here we report the results of a five-week experiment carried out in the closed core of a vertically coherent, mesoscale eddy of the Antarctic Circumpolar Current, during which we tracked sinking particles from the surface to the deep-sea floor. A large diatom bloom peaked in the fourth week after fertilization. This was followed by mass mortality of several diatom species that formed rapidly sinking, mucilaginous aggregates of entangled cells and chains. Taken together, multiple lines of evidence-although each with important uncertainties-lead us to conclude that at least half the bloom biomass sank far below a depth of 1,000 metres and that a substantial portion is likely to have reached the sea floor. Thus, iron-fertilized diatom blooms may sequester carbon for timescales of centuries in ocean bottom water and for longer in the sediments.