Search

Technological developments have facilitated the collection of large amounts of imagery from isolated deep-sea ecosystems such as abyssal nodule fields. Application of imagery as a monitoring tool in these areas of interest for deep-sea exploitation is extremely valuable. However, in order to collect a comprehensive number of species observations, thousands of images need to be analysed, especially if a high diversity is combined with low abundances such is the case in the abyssal nodule fields. As the visual interpretation of large volumes of imagery and the manual extraction of quantitative information is time-consuming and error-prone, computational detection tools may play a key role to lessen this burden. Yet, there is still no established workflow for efficient marine image analysis using deep learning–based computer vision systems for the task of fauna detection and classification. Methods In this case study, a dataset of 2100 images from the deep-sea polymetallic nodule fields of the eastern Clarion-Clipperton Fracture zone from the SO268 expedition (2019) was selected to investigate the potential of machine learning–assisted marine image annotation workflows. The Machine Learning Assisted Image Annotation method (MAIA), provided by the BIIGLE system, was applied to different set-ups trained with manually annotated fauna data. The results computed with the different set-ups were compared to those obtained by trained marine biologists regarding accuracy (i.e. recall and precision) and time. Results Our results show that MAIA can be applied for a general object (i.e. species) detection with satisfactory accuracy (90.1% recall and 13.4% precision), when considered as one intermediate step in a comprehensive annotation workflow. We also investigated the performance for different volumes of training data, MAIA performance tuned for individual morphological groups and the impact of sediment coverage in the training data. Discussion We conclude that: a) steps must b

Trait-based approaches that complement taxonomy-based studies have increased in popularity among the scientific community over the last decades. The collection of biological and ecological characteristics of species (i.e., traits) provides insight into species and ecosystem vulnerability to environmental and anthropogenic changes, as well as ecosystem functioning. Here, we present the FUN Azores trait database, describe our approach, evaluate its scope, compare it to other marine trait databases, and explore the spatial distribution of its traits with “functional maps.” While most of the available trait databases to date contain essential information to understand the functional diversity of a taxonomic or functional group, our ecosystem-based approach provides a comprehensive assessment of diverse fauna (i.e., meio-, macro-, and megafauna) from benthic and pelagic environments in the Azores Marine Park; including ridges, seamounts, hydrothermal vents, and the overlying water column. We used a collaborative approach involving 30 researchers with different expertise to develop the FUN Azores database, which contains compiled data on 14 traits representing morphological, behavioral, and life history characteristics for 1,210 species across 10 phyla. The “functional maps” show a distinct distribution of the two most common size classes, suggesting different communities with different functionalities. The following traits had the best scoring coverage (i.e., >95% of the species scored): maximum body size, body form, skeleton material, feeding structure, motility, environmental position, substratum affinity, distribution, and depth range; while traits related to species behavior (e.g., sociability or aggregation tendencies) and life history (e.g., developmental mechanism) had lower scoring coverage, highlighting the need for further research to fill these knowledge gaps. We found a larger number of species in the benthic compared to the pelagic environment and differing

Trait-based approaches that complement taxonomic-based studies have increased in popularity among the scientific community over the last decades. The collection of biological and ecological characteristics of species (i.e., traits) provides insight into species and ecosystem vulnerability to environmental and anthropogenic changes, as well as ecosystem functioning. While most of the available trait databases to date contain essential information to understand the functional diversity of a taxonomic group or functional group based on size, the FUN Azores trait database has an ecosystem-based approach that provides a comprehensive assessment of diverse fauna (meio-, macro-, and megafauna) from benthic and pelagic environments in the Azores Marine Park; including ridges, seamounts, and hydrothermal vents. We used a collaborative approach involving 30 researchers with different expertise to develop the trait database; which contains compiled data on 14 traits representing morphological, behavioral, and life history characteristics for 1210 species, across 10 phyla

With the mining of polymetallic nodules from the deep-sea seafloor once more evoking commercial interest, decisions must be taken on how to most efficiently regulate and monitor physical and community disturbance in these remote ecosystems. Image-based approaches allow non-destructive assessment of the abundance of larger fauna to be derived from survey data, with repeat surveys of areas possible to allow time series data collection. At the time of writing, key underwater imaging platforms commonly used to map seafloor fauna abundances are autonomous underwater vehicles (AUVs), remotely operated vehicles (ROVs) and towed camera “ocean floor observation systems” (OFOSs). These systems are highly customisable, with cameras, illumination sources and deployment protocols changing rapidly, even during a survey cruise. In this study, eight image datasets were collected from a discrete area of polymetallic-nodule-rich seafloor by an AUV and several OFOSs deployed at various altitudes above the seafloor. A fauna identification catalogue was used by five annotators to estimate the abundances of 20 fauna categories from the different datasets. Results show that, for many categories of megafauna, differences in image resolution greatly influenced the estimations of fauna abundance determined by the annotators. This is an important finding for the development of future monitoring legislation for these areas. When and if commercial exploitation of these marine resources commences, robust and verifiable standards which incorporate developing technological advances in camera-based monitoring surveys should be key to developing appropriate management regulations for these regions.

© The Author(s), 2019. This article is distributed under the terms of the Creative Commons Attribution License. The definitive version was published in Chapman, A. S. A., Beaulieu, S. E., Colaco, A., Gebruk, A. V., Hilario, A., Kihara, T. C., Ramirez-Llodra, E., Sarrazin, J., Tunnicliffe, V., Amon, D. J., Baker, M. C., Boschen-Rose, R. E., Chen, C., Cooper, I. J., Copley, J. T., Corbari, L., Cordes, E. E., Cuvelier, D., Duperron, S., Du Preez, C., Gollner, S., Horton, T., Hourdez, S., Krylova, E. M., Linse, K., LokaBharathi, P. A., Marsh, L., Matabos, M., Mills, S. W., Mullineaux, L. S., Rapp, H. T., Reid, W. D. K., Rybakova (Goroslavskaya), E., Thomas, T. R. A., Southgate, S. J., Stohr, S., Turner, P. J., Watanabe, H. K., Yasuhara, M., & Bates, A. E. sFDvent: a global trait database for deep-sea hydrothermal-vent fauna. Global Ecology and Biogeography, 28(11), (2019): 1538-1551, doi: 10.1111/geb.12975., Motivation Traits are increasingly being used to quantify global biodiversity patterns, with trait databases growing in size and number, across diverse taxa. Despite growing interest in a trait‐based approach to the biodiversity of the deep sea, where the impacts of human activities (including seabed mining) accelerate, there is no single repository for species traits for deep‐sea chemosynthesis‐based ecosystems, including hydrothermal vents. Using an international, collaborative approach, we have compiled the first global‐scale trait database for deep‐sea hydrothermal‐vent fauna – sFDvent (sDiv‐funded trait database for the Functional Diversity of vents). We formed a funded working group to select traits appropriate to: (a) capture the performance of vent species and their influence on ecosystem processes, and (b) compare trait‐based diversity in different ecosystems. Forty contributors, representing expertise across most known hydrothermal‐vent systems and taxa, scored species traits using online collaborative tools and shared workspaces. Here, we characterise the sFDvent database, describe our approach, and evaluate its scope. Finally, we compare the sFDvent database to similar databases from shallow‐marine and terrestrial ecosystems to highlight how the sFDvent database can inform cross‐ecosystem comparisons. We also make the sFDvent database publicly available online by assigning a persistent, unique DOI. Main types of variable contained Six hundred and forty‐six vent species names, associated location information (33 regions), and scores for 13 traits (in categories: community structure, generalist/specialist, geographic distribution, habitat use, life history, mobility, species associations, symbiont, and trophic structure). Contributor IDs, certainty scores, and references are also provided. Spatial location and grain Global coverage (grain size: ocean basin), spanning eight ocean basins, including vents on 12 mid‐ocean ridges and 6 back‐arc spreading centre, We would like to thank the following experts, who are not authors on this publication but made contributions to the sFDvent database: Anna Metaxas, Alexander Mironov, Jianwen Qiu (seep species contributions, to be added to a future version of the database) and Anders Warén. We would also like to thank Robert Cooke for his advice, time, and assistance in processing the raw data contributions to the sFDvent database using R. Thanks also to members of iDiv and its synthesis centre – sDiv – for much‐valued advice, support, and assistance during working‐group meetings: Doreen Brückner, Jes Hines, Borja Jiménez‐Alfaro, Ingolf Kühn and Marten Winter. We would also like to thank the following supporters of the database who contributed indirectly via early design meetings or members of their research groups: Malcolm Clark, Charles Fisher, Adrian Glover, Ashley Rowden and Cindy Lee Van Dover. Finally, thanks to the families of sFDvent working group members for their support while they were participating in meetings at iDiv in Germany. Financial support for sFDvent working group meetings was gratefully received from sDiv, the Synthesis Centre of iDiv (DFG FZT 118). ASAC was a PhD candidate funded by the SPITFIRE Doctoral Training Partnership (supported by the Natural Environmental Research Council, grant number: NE/L002531/1) and the University of Southampton at the time of submission. ASAC also thanks Dominic, Lesley, Lettice and Simon Chapman for their support throughout this project. AEB and VT are sponsored through the Canada Research Chair Programme. SEB received support from National Science Foundation Division of Environmental Biology Award #1558904 and The Joint Initiative Awards Fund from the Andrew W. Mellon Foundation. AC is supported by Program Investigador (IF/00029/2014/CP1230/CT0002) from Fundação para a Ciência e a Tecnologia (FCT). This study also had the support of Fundação para a Ciência e a Tecnologia, through the strategic project UID/MAR/04292/2013 grante

Seamounts are abundant and prominent features on the deep-sea floor and intersperse with the nodule fields of the Clarion-Clipperton Fracture Zone (CCZ). There is a particular interest in characterising the fauna inhabiting seamounts in the CCZ because they are the only other ecosystem in the region to provide hard substrata besides the abundant nodules on the soft-sediment abyssal plains. It has been hypothesised that seamounts could provide refuge for organisms during deep-sea mining actions or that they could play a role in the (re-)colonisation of the disturbed nodule fields. This hypothesis is tested by analysing video transects in both ecosystems, assessing megafauna composition and abundance. Nine video transects (ROV dives) from two different license areas and one Area of Particular Environmental Interest in the eastern CCZ were analysed. Four of these transects were carried out as exploratory dives on four different seamounts in order to gain first insights into megafauna composition. The five other dives were carried out in the neighbouring nodule fields in the same areas. Variation in community composition observed among and along the video transects was high, with little morphospecies overlap along intra-ecosystem transects. Despite the observation of considerable faunal variations within each ecosystem, differences between seamounts and nodule fields prevailed, showing significantly different species associations characterising them, thus calling into question their use as a possible refuge area.

Mining impacts will affect local populations to different degrees. Impacts range from removal of habitats and possible energy sources to pollution and smaller-scale alterations in local habitats that, depending on the degree of disturbance, can lead to extinction of local communities. While there is a shortage or even lack of studies investigating impacts that resemble those caused by actual mining activity, the information available on the potential long-lasting impacts of seabed mining emphasise the need for effective environmental management plans. These plans should include efforts to mitigate deep-sea mining impact such as avoidance, minimisation and potentially restoration actions, to maintain or encourage reinstatement of a resilient ecosystem. A wide range of mitigation and restoration actions for deep-sea ecosystems at risk were addressed. From an ecological point of view, the designation of set-aside areas (refuges) is of utmost importance as it appears to be the most comprehensive and precautionary approach, both for well-known and lesser studied areas. Other actions range from the deployment of artificial substrates to enhance faunal colonisation and survival to habitat recreation, artificial eutrophication, but also spatial and temporal management of mining operations, as well as optimising mining machine construction to minimise plume size on the sea floor, toxicity of the return plume and sediment compression. No single action will suffice to allow an ecosystem to recover, instead combined mitigation/restoration actions need to be considered, which will depend on the specific characteristics of the different mining habitats and the resources hosted (polymetallic sulphides, polymetallic nodules and cobalt-rich ferromanganese crusts). However, there is a lack of practical experience regarding mitigation and restoration actions following mining impacts, which severely hamper their predictability and estimation of their possible effect and success. We pro

During 2011, two deep-sea observatories focusing on hydrothermal vent ecology were up and running in the Atlantic (Eiffel Tower, Lucky Strike vent field) and the Northeast Pacific Ocean (NEP) (Grotto, Main Endeavour Field). Both ecological modules recorded imagery and environmental variables jointly for a time span of 23 days (7–30 October 2011) and environmental variables for up to 9 months (October 2011–June 2012). Community dynamics were assessed based on imagery analysis and rhythms in temporal variation for both fauna and environment were revealed. Tidal rhythms were found to be at play in the two settings and were most visible in temperature and tubeworm appearances (at NEP). A ∼ 6 h lag in tidal rhythm occurrence was observed between Pacific and Atlantic hydrothermal vents, which corresponds to the geographical distance and time delay between the two sites.

The NEPTUNE cabled observatory network hosts an ecological module called TEMPO-mini that focuses on hydrothermal vent ecology and time series, granting us real-time access to data originating from the deep sea. In 2011-2012, during TEMPO-mini's first deployment on the NEPTUNE network, the module recorded high-resolution imagery, temperature, iron (Fe) and oxygen on a hydrothermal assemblage at 2186 m depth at Main Endeavour Field (North East Pacific). 23 days of continuous imagery were analysed with an hourly frequency. Community dynamics were analysed in detail for Ridgeia piscesae tubeworms, Polynoidae, Pycnogonida and Buccinidae, documenting faunal variations, natural change and biotic interactions in the filmed tubeworm assemblage as well as links with the local environment. Semi-diurnal and diurnal periods were identified both in fauna and environment, revealing the influence of tidal cycles. Species interactions were described and distribution patterns were indicative of possible microhabitat preference. The importance of high-resolution frequencies (< 1 h) to fully comprehend rhythms in fauna and environment was emphasised, as well as the need for the development of automated or semi-automated imagery analysis tools.

Although the spatial distribution of hydrothermal vent assemblages in relation to environmental conditions has been assessed in several studies, there is little documented data on the temporal variation of the fauna and corresponding abiotic factors in a vent community. Here, we present one of the longest integrated (faunal and environmental data) time series ever obtained in a hydrothermal ecosystem. The data were acquired using the TEMPO ecological module that was deployed between 2006 and 2008 on the Mid-Atlantic Ridge, providing the first insights into the day-to-day variations in a Bathymodiolus azoricus mussel assemblage from the Lucky Strike vent field for the 48 days during which the video camera operated. The time-series yielded additional valuable information on longer-term variation in faunal distribution (comparing∼2 years), temperature (11.7 months) and iron concentrations (3.8 months). Results from daily observations showed that the vent mussel assemblage was quite stable over the 48 days of the study, reflecting the relative stability of environmental conditions during this period. Bathymodiolus azoricus mussels appeared to thrive in areas of very limited hydrothermal fluid input in habitats that are, as in other deep-sea ecosystems, significantly influenced by ocean tidal signals. Variation in species abundance was observed but, with the exception of Mirocaris fortunata shrimp, no links could be established with measured environmental variables. Although we did not observe any clear tidal influence on vent fauna, it is likely that physiological processes and species’ activities are influenced by these periodic variations. Longer time series are currently being acquired by different experiments deployed on the EMSO-Açores MoMAR observatory (2010–2013 and still recording). They should further improve our knowledge of the dynamics of hydrothermal systems and their associated faunal communities.

The Eiffel Tower edifice is situated in the Lucky Strike hydrothermal vent field at a mean depth of 1690 m on the Mid-Atlantic Ridge (MAR). At this 11-m-high hydrothermal structure, different faunal assemblages, varying in visibly dominant species (mussels and shrimp), in mussel size and in density of mussel coverage, were sampled biologically and chemically. Temperature and sulphide (Sigma S) were measured on the different types of mussel-based assemblages and on a shrimp-dominated assemblage. Temperature was used as a proxy for calculating total concentrations of CH4. Based on the physico-chemical measurements, two microhabitats were identified, corresponding to (i) a more variable habitat featuring the greatest fluctuations in environmental variables and (ii) a second, more stable, habitat. The highest temperature variability and the highest maximum recorded temperatures were found in the assemblages visibly inhabited by alvinocaridid shrimp and dense mussel beds of large Bathymodiolus azoricus, whereas the less variable habitats were inhabited by smaller-sized mussels with increasing bare surface in between. Larger mussels appeared to consume more Sigma S compared with smaller-sized (< 1 cm) individuals and thus had a greater influence on the local chemistry. In addition, the mussel size was shown to be significantly positively correlated to temperature and negatively to the richness of the associated macrofauna. The presence of microbial mats was not linked to specific environmental conditions, but had a negative effect on the presence and abundance of macro-fauna, notably gastropods. Whereas some taxa or species are found in only one of the two microhabitats, others, such as polychaetes and Mirocaris shrimp, cross the different microhabitats. Temperature was proposed to be a more limiting factor in species distribution than Sigma S.