Future Science Brief N° 11 June 2024

Marine habitat
mapping
 EMB FUTURE SCIENCE BRIEF

European Marine Board IVZW

The European Marine Board provides a pan-European platform for its Member organisations to develop common
priorities, advance marine research, and to bridge the gap between science and policy in order to meet future marine
science challenges and opportunities.

The European Marine Board (EMB) is an independent and self-sustaining science policy interface organisation that
currently represents 38 Member organisations from 19 European countries. It was established in 1995 to facilitate
enhanced cooperation between European marine science organisations 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 organisations 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.

www.marineboard.eu

European Marine Board Member Organisations

2
 N° 11 2024

European Marine Board IVZW Future Science Brief N° 11

This Future Science Brief is a result of the work of the European Marine Board Expert Working
Group on Marine Habitat Mapping. See Annex 1 for the list and affiliations of the Working Group
Members.

Working Group Chairs

Simonetta Fraschetti, James Strong

Contributing Authors
Lene Buhl-Mortensen, Federica Foglini, Jorge M. S. Gonçalves, José Manuel González-Irusta, Helen Lillis, Mats
Lindegarth, Georg Martin, Lenaick Menot, Eimear O’Keeffe, António Pascoal, Maria Salomidi, Timm Schoening

Series Editor
Sheila J. J. Heymans

Publication Editors
Britt Alexander, Ángel Muñiz Piniella, Paula Kellett, Ana Rodriguez Perez, Konstantina Bairaktari, Fernanda Bayo Ruiz,
Sheila J. J. Heymans

External Reviewers
Carlo Cerrano, David Connor, Ellen Kenchington, Juan Ronco Zapatero

Internal review process

The content of this document has been subject to internal review, editorial support and approval by the European
Marine Board Member Organisations

Suggested reference
Fraschetti, S., Strong, J., Buhl-Mortensen, L., Foglini, F., Gonçalves, J. M. S., González-Irusta, J. M., Lillis, H., Lindegarth, M.,
Martin, G., Menot, L., O’Keeffe, E., Pascoal, A., Salomidi, M., Schoening, T. (2024) Marine habitat mapping. Alexander, B.,
Rodriguez Perez, A., Kellett, P., Muñiz Piniella, A., Bayo Ruiz, F., Bairaktari, K., Heymans, J. J. [Eds.] Future Science Brief
N°. 11 of the European Marine Board, Ostend, Belgium. ISSN: 25593-5232. ISBN: 9789464206234. DOI: 10.5281/
zenodo.11203128

www.marineboard.eu
info@marineboard.eu

Design
Zoeck

Front cover image credit:

Roberto Rinaldi

First edition, June 2024

3
 EMB FUTURE SCIENCE BRIEF

Foreword

Beneath the surface of our Ocean lies a complex tapestry of marine

habitats, intricately connected and fundamental to the health of our
Ocean and the life it sustains. The need for comprehensive marine
habitat maps to help make decisions on activities at sea has never been
more pressing. High-quality marine habitat maps are required to realise
the ambitions of the European Union's Mission to Restore Our Ocean and
Waters by 2030 and to navigate the challenges of the European Green
Deal to simultaneously protect and restore marine ecosystems, and to
scale-up offshore renewable energy and other Blue Economy sectors.

The European Marine Board's Future Science Brief on marine habitat

mapping emphasises the crucial role of accurate and extensive marine
habitat maps for achieving European and international goals for
biodiversity, conservation, restoration and climate action. It outlines
the science and policy needs to advance our understanding and
documentation of marine habitats, from increasing the resolution of
biological information to strengthening coordination mechanisms for interdisciplinary mapping efforts to fill critical gaps.
International initiatives such as the Nippon Foundation-GEBCO Seabed 2030 project have placed a heavy and important
emphasis on traditional hydrographic mapping. However, by going this important step further towards the advancement
of novel cost-effective mapping technologies, mapping of ecosystems in three dimensions, supporting repeat mapping
surveys to document changes over time, promoting standardisation and dissemination of mapping methods and
products, and advancing habitat classification schemes, we can unlock the full potential of marine habitat mapping as a
tool for informed decision-making. This Future Science Brief aims to inform policymakers, programme managers, research
funders and the wider science-policy and scientific communities in the advancement of next-generation marine habitat
mapping efforts.

The European Marine Board selected the topic of marine habitat mapping for a new activity in spring 2021. The Working
Group kicked-off with a hybrid meeting at the Stazione Zoologica Anton Dohrn (Naples, Italy) hosted by the Working
Group Chair, Prof. Simonetta Fraschetti. On behalf of the European Marine Board, I extend my gratitude to the Working
Group Members for their collaborative effort in writing this document, bringing together diverse perspectives and
approaches to marine habitat mapping. I would also like to thank the external reviewers for their constructive comments
and Leonardo Tunesi (ISPRA) for valuable comments and insights on habitat classification schemes. Finally, as always, I
would like to thank the EMB Secretariat, in particular Britt Alexander, for the coordination of the Working Group and for
shepherding the writing, editing and reviewing of this document through to publication.

Fiona Grant
Chair, European Marine Board
June 2024

4
 N° 11 2024

Table of Contents

Executive Summary 6
1. Introduction 9
1.1 What is marine habitat mapping? 9
1.2 Why is marine habitat mapping important? 9
1.3 Challenges facing marine habitat mapping 14
2. Collecting data for marine habitat mapping 15
2.1 Collecting remotely-sensed data 15
2.2 Collecting in situ observations for marine habitat mapping 20
2.3 Integrating artificial intelligence within marine habitat mapping 23
2.4 Recommendations 25
3. Combining data to produce marine habitat maps 26
3.1 Physical and biological habitat maps 27
3.2 Distribution models in marine habitat mapping 27
3.3 Are European habitat classification schemes fit-for-purpose? 30
3.4 Assessing and communicating accuracy and confidence 36
3.5 Recommendations 38
4. What and where to map 39
4.1 What has been mapped? 39
4.2 What are the gaps in marine habitat mapping? 44
4.3 Where to map first? The need for spatial prioritisation 45
4.4 Who uses marine habitat maps and for what purpose? 46
4.5 Bespoke, fit-for-purpose marine habitat maps 47
4.6 Recommendations 48
5. Communication and dissemination 49
5.1. Data dissemination: increasing the value of each map 49
5.2 Using marine habitat maps to improve public understanding of the Ocean 50
5.3 Recommendations 54
6. Overarching recommendations to advance marine habitat mapping 55
List of abbreviations and acronyms 57
Glossary 59
References 64
Annex 1. Members of the European Marine Board Working Group on Marine Habitat Mapping 72
Annex 2. Data driven approaches to the use of distribution models in marine habitat mapping 73

5
 EMB FUTURE SCIENCE BRIEF

Executive summary

Accurate and extensive marine habitat maps are fundamental to support a wide variety of marine policies and ambitions.
These include the European Union’s Marine Strategy Framework Directive and policies to deliver the ambitious plans of
the European Green Deal. The simultaneous scaling-up of sustainable Blue Economy activities, while protecting and
restoring marine ecosystems as part of the EU 2030 Biodiversity Strategy and the proposed Nature Restoration Law
will require increased knowledge of marine habitats. Marine habitat mapping aims to create a holistic representation
of the distribution of marine habitats in space and time, and provide insight into associated biological communities,
ecological status and condition, and physical properties. Habitat maps are valuable spatial decision-support tools that
inform the sustainable use of marine space when using an ecosystem-based approach. They can be used to assess the
impact of anthropogenic pressures on marine resources and ecosystem services, to identify and plan new networks of
marine protected areas and areas for restoration, and to inform maritime spatial planning. However, large areas remain
unmapped and current maps predominantly focus on physical aspects of marine habitats and lack sufficient biological
resolution, such as species and communities. Higher resolution maps are needed to better represent the linkages between
the seabed and water column in three dimensions and to enable an ecosystem approach to mapping that considers the
marine environment in the fourth dimension, capturing the timing of important ecological processes.

This Future Science Brief highlights science and policy needs and recommendations to advance marine habitat
mapping in order to fulfil European and international ambitions for biodiversity, conservation, restoration and climate.
It primarily targets policymakers, programme managers, research funders and the wider science-policy and scientific
communities. It highlights current methods and future trends in the acquisition of data from the seabed and water
column via remote sensing and direct, in situ techniques, combining data to produce maps using modelling approaches,
and recommendations for adopting fit-for-purpose classification schemes. It provides an overview of what has been
mapped and where within the European sea-basins, highlights the need to increase the quality and resolution of marine
habitat maps, and identifies critical gaps in habitat types and geographic extent, including the deep sea, Natura 2000
sites and other Marine Protected Areas across all regional seas. Finally, it describes the need to improve the assessment
and communication of uncertainty and confidence in maps, and for maps to be more easily accessible to a variety of
stakeholders to increase their value for end-users and to the public for Ocean literacy.

To address policy needs and increase the capacity for the production and dissemination of accurate marine habitat maps,
we recommend scientists/map producers and research funders to:

• Support multidisciplinary national and EU research projects to advance novel methods to increase the
resolution of biological information within marine habitat mapping;
• Support national and EU research programmes that focus on repeat mapping to understand temporal
change, particularly of ecologically significant spatial units, i.e. hot spots of ecosystem functioning
where high rates of change are expected;
• Promote the standardisation of mapping methods and outputs in research and mapping programmes;
• Promote and incentivise research and mapping programmes to publish marine habitat mapping data
according to the Findable, Accessible, Interoperable and Reusable principles and to submit data to
centralised data services;
• Support public-private research collaboration for the development of cost-effective mapping tools; and
• Support dedicated mapping projects focusing on citizen science and reformatting mapping products
that promote Ocean Literacy.

6
 N° 11 2024

In addition, we recommend policymakers to:

• Strengthen national, regional, European and international coordination mechanisms for interdisciplinary
mapping efforts to ensure effective use of mapping resources and identification of gaps;
• Establish an international effort to identify priority areas in need of mapping, with a focus on areas of
the largely unmapped deep sea and coastal areas, which are under the greatest pressure from human
activities;
• Require map producers (e.g. ICES Working Group on Marine Habitat Mapping, EMODnet, large mapping
projects) or map users (e.g. the European Environment Agency, Joint Nature Conservation Committee)
to produce best practice and reporting templates for the standardised assessment and reporting of map
accuracy and confidence; and
• Advance habitat classification schemes, which lie at the heart of all marine habitat maps, to include
quantitative characterisation of habitats to support the assessment of their condition. Habitat maps
will be enriched further if these classification schemes link to other sources of information such as
sensitivity to pressures and ecosystem services provision.

Credit: Ben James © NatureScot.

The advancement of novel methods to increase the resolution of biological information within marine habitat mapping is needed.

The European Marine Board acknowledges that while the Working Group members who wrote this document and
its recommendations represent diversity in terms of European geographical location (see Annex 1), professional
background, gender and career level, their views may not represent ideas from all forms of diversity. This document
has a European focus, but its messages and recommendations are relevant to stakeholders globally. The diversity in
scientific expertise in the Working Group has been crucial in highlighting different views and perspectives in marine
habitat mapping from different communities (e.g. geologists vs biologists, coastal vs deep-sea researchers, modellers
vs data collectors) and to address the complexity of the topic, adopting a common voice in this document.

7
 EMB FUTURE SCIENCE BRIEF

Contribution to the UN Ocean Decade

Challenges and Outcomes

This Future Science Brief and its recommendations support the

UN Decade of Ocean Science for Sustainable Development’s
(Ocean Decade) societal outcomes (O1 – O7) and challenges
(C1 – C10) in the following ways:

• A healthy and resilient Ocean’ (O2) where marine ecosystems are understood, protected, restored and managed
and ‘Protect and restore ecosystems and biodiversity’ (C2) by highlighting advances needed in marine habitat
mapping to best plan and monitor ecosystem conservation and restoration activities.

• ‘A productive Ocean’ (O3) supporting sustainable food supply and a sustainable Ocean economy, ‘Sustainably
feed the global population’ (C3) and ‘Develop a sustainable and equitable Ocean economy’ (C4) by providing
information on the distribution of vulnerable habitats within fishing fleet’s activity areas in order to minimise
fishing impacts and to select suitable sites for aquaculture.

• ‘A predicted Ocean’ (O4) where society understands and can respond to changing Ocean conditions by providing
recommendations on filling gaps in mapping habitat types and geographic areas to gain baseline information
on which to base management decisions and recommendations to implement repeat mapping to detect change
over time.

• ‘An accessible Ocean’ (O6) with open and accessible access to data, information, technology, and innovation by
highlighting the need for scientists and wider stakeholders to share maps and mapping data to increase uptake,
dissemination and value.

• ‘An inspiring and engaging Ocean’ (O7) where society understands and values the Ocean in relation to human
wellbeing and sustainable development and ‘Change humanity’s relationship with the Ocean’ (C10) by
providing recommendations on the use of marine habitat maps to increase public understanding of the Ocean.

Contribution to the EU Mission:

Restore our Ocean and Waters

This Future Science Brief and its recommendations support

the objectives of the EU Mission: Restore our Ocean and
Waters in the following ways:

• ‘Protect and restore marine and freshwater ecosystems and biodiversity’ by highlighting advances needed in
marine habitat mapping to best plan and monitor ecosystem conservation and restoration activities.

And the cross-cutting enabling actions:

• ‘Broad public mobilisation and engagement’ by providing recommendations on the use of marine habitat maps
to increase public understanding of the Ocean.

• ‘A digital Ocean and water knowledge system’ by highlighting the need for scientists and wider stakeholders to
share maps and mapping data to increase uptake, dissemination and value. Marine habitat maps form the basis
of spatial ecosystem models that are needed for digital twins of the Ocean.

8
 N° 11 2024

1 Introduction
1.1 What is marine habitat mapping?

The term “habitat” has various meanings in different contexts and scales (Fraschetti et al., 2018;
Montefalcone et al., 2021). Within the context of habitat mapping, a habitat refers to “a recognisable
space which can be distinguished by its abiotic (i.e. physical) characteristics and associated biological
assemblages1, assessed at particular spatial and temporal scales” (ICES, 2005) (see Figure 1.1 for examples
of different components of a habitat).

Marine Habitat Mapping (MHM) aims to gain a holistic representation be mapped in MHM initiatives depend on the aims, management
of the distribution of marine habitats in space and time. Marine needs, scale and context. MHM mainly refers to activities to produce
Habitat Maps (MHMs), in combination with other data (e.g. sensitivity maps that completely cover a specified geographical area using a
matrices, spatial data on anthropogenic pressures, repeat mapping combination of remotely-sensed techniques that collect data at a
over time) may also provide insight into changes in ecological distance from the mapped area, direct2 in situ (in water) observations
vulnerability and potential human impacts. The characteristics to (also referred to as ground truthing) and/or modelled data.

1.2 Why is marine habitat mapping important?

The marine environment hosts a wide variety of habitats. Benthic and how much). Such maps also contribute to the assessment
habitats (i.e. those associated with or occurring at the seafloor) of the economic value of ecosystem services, and promote the
are underpinned by various bottom types e.g. sandy and muddy integration of these values into accounting and reporting systems
seabeds, or hard bottoms, and include seagrass meadows, at European and national levels (Galparsoro et al., 2014).
coralligenous formations, cold-water coral reefs, mussel beds, kelp
and macroalgal forests and sponge aggregations. There are also MHMs are important for gaining basic knowledge of marine
habitats associated with the water column (i.e. pelagic habitats in habitats and to provide comprehensive advice on marine habitat
the open Ocean). Pelagic habitats support commercially important conservation and restoration agendas to achieve the objectives
fish species as well as vital processes maintaining ecosystem set by European Union (EU) and international policies. There are,
functions (e.g. photosynthesis from phytoplankton) and are tightly however, substantial gaps in MHMs in terms of biological detail,
connected to benthic seascapes. Habitats can be geophysical, geographic coverage and coverage of different habitat types
i.e. primarily shaped by geology and physical processes, and/or (Gerovasileiou et al., 2019), and the majority of maps do not
biogenic i.e. formed by living organisms and provide a habitat for accurately reflect the ecological role and importance of marine
other organisms3. habitats. Worldwide, widespread habitat loss and degradation
in coastal and marine systems have been observed as a result of
Europe’s citizens and economy depend on these habitats and multiple human pressures and a lack of efficient conservation
their associated species to deliver critical provisioning (e.g. food), measures at large scales. There is, however, very limited, mostly
regulating (e.g. heat and carbon storage) and cultural (e.g. recreation qualitative information on degraded habitats and their recovery
and tourism) ecosystem services. MHMs showing the distribution and restoration potential. As a result, current EU regulations
of species and habitat types, and their functional diversity provide underestimate ecosystem change due to human pressures, and
an initial inventory of these important resources (i.e. what, where the consequent reduction in habitat diversity and complexity.

1
A biological assemblage is a group of species that coexist in a specific habitat
2
Direct observations are those collected close to the object of interest
3
Examples of biogenic habitats include some species of seagrass (e.g. Posidonia oceanica), coralligenous formations, cold-water coral reefs and mussel beds.

9
 EMB FUTURE SCIENCE BRIEF

Credit: Mariacristina Prampolini (CNR-ISMAR).

Figure 1.1. Maps showing different components of habitats in a submarine canyon: (A) physical shape of the seabed (i.e. geomorphology);
(B) substrate (i.e. bottom type); and (C) biological assemblage (classified according to the CoCoNet4 Habitat Mapping Scheme).

4
https://cordis.europa.eu/project/id/287844/fr

10
 N° 11 2024

1.2.1 Marine habitat maps are needed to

fulfil policy objectives ecological processes. Currently, most conservation actions,
MHMs enable spatial management of human activities and are management decisions and policies are based on a two-dimensional
critical for providing reliable information to support various policies approach and do not explicitly incorporate the three-dimensional
and management tools (Table 1.1). The main EU policy driver for nature of the Ocean (i.e. linking benthic and pelagic systems) or
MHMs is the Marine Strategy Framework Directive (MSFD), which the fourth dimension (i.e. including time). In addition, mapping
requires Good Environmental Status (GES) to be achieved across and assessment of ecosystems and their services within Member
the entire seabed (i.e. the 22 Benthic Broad Habitat Types, BBHT), State’s national territories is one of the key approaches of the EU
with agreed policy goals to have 75% of each BBHT in GES. MHMs Biodiversity Strategy, which aims to support their maintenance and
are required to facilitate the reporting of these status assessments restoration.
via the indicators that are used to evaluate the spatial coverage
or extent of certain features or habitats. In addition, the coverage To plan and spatially prioritise active restoration interventions, we
of Marine Protected Areas (MPAs) needs to increase to achieve need to be able to document and monitor the location and extent
the target of protecting at least 30% of European seas by 2030 of degraded habitats (Gerovasileiou et al., 2019). The proposed EU
and building a network of MPAs with improved connectivity5 as Nature Restoration Law has set ambitious targets that demand a
part of the EU Biodiversity Strategy, and for the representative profound knowledge of the distribution and extent of European
implementation of the Natura 2000 network. To plan, design marine habitats to assess the percentage of each habitat that is
and monitor networks of coherent and effective MPAs we need in poor condition and therefore suitable for restoration measures
to know where habitats are, their connectivity, how they will (Hering et al., 2023). To manage and sustainably exploit fish and
change under climate change and ultimately, their ecological other commercial stocks that may see changes in their distribution
status (i.e. whether they are degraded or not). This knowledge with climate change, MHM over spatial and temporal scales can
will enable MPA networks to be designed to accurately represent also help to plan for the future.

Credit: Mišo Pavičić, IZOR (top left), Graham Saunders © Crown copright (top right), George Stoyle © NatureScot (bottom right).

The marine environment hosts a wide variety of habitats including sponge aggregations (top left), maërl beds, (top right), deep-sea coral gardens
(bottom left) and pelagic habitats associated with the water column (bottom right).

5
Connectivity is the extent to which populations in different parts of the species’ range are linked by the movement of eggs, larvae or other propagules, juveniles or adults.

11
 EMB FUTURE SCIENCE BRIEF

Table 1.1 Examples of international and EU policies, directives and conventions benefiting from MHMs, and objectives and activities were MHMs of
the physical environment, habitats and species composition are of critical importance for their implementation.

POLICY/DIRECTIVE/CONVENTION OBJECTIVES AND ACTIVITIES FOR WHICH MHMS ARE OF CRITICAL IMPORTANCE

International Level
· Developing ecosystem-based management of fisheries and marine biodiversity, which requires
UN Convention on Biodiversity (CBD, 1992)
spatial information on ecological values from local, regional and global scales.

· Mapping and monitoring biodiversity and ecosystem services for SDG14 ("Life below water").
UN Sustainable Development Goals (SDG's;
· Planning for sustainable aquaculture and conservation of fish habitats for SDG2 ("Zero hunger").
United Nations, 2015)
· Monitoring and planning carbon sequestration and energy production for SDG13 ("Climate action").

· Providing guidance for the following targets:

- Target 1: Plan and manage all areas to reduce biodiversity loss.
UN CBD Kunming-Montreal Global - Target 2: Restore 30% of all degraded ecosystems.
Biodiversity Framework - Target 3: Conserve 30% of land, waters and seas.
(CBD/COP/DEC/15/4, 2022) - Target 8: Minimise the impacts of climate change on biodiversity and build resilience.
- Target 11: Restore, maintain and enhance nature’s contributions to people.
- Target 14: Integrate biodiversity in decision-making at every level.

EU Level
· Representative and sufficient implementation of the Natura 2000 network of protected areas,
Habitats Directive (Council Directive 92/43/ which are regionally designated.
EEC, 1992) and Birds Directive
· Assessment of favourable conservation status to account for changes in areal extent and condition
(Directive 2009/147/EC, 2009)
of habitats and species.
· Guiding the assessment of the extent of good ecological status within water bodies as an integral
Water Framework Directive (Directive part of indicators and communicating to wider stakeholders.
2000/60/EC, 2000) · Supporting working materials before, during and after assessment phases. MHMs help policy-
makers to navigate and interpret the data, draw conclusions, and find knowledge gaps.
· Status assessments and programmes of measures for GES for MSFD descriptors, e.g. “Biodiversity”
Marine Strategy Framework Directive (MSFD; (Descriptor 1) requires estimates of the extent and quality of selected habitats and species in EU
Directive 2008/56/EC, 2008) marine waters and “Seafloor Integrity” (Descriptor 6) involves assessing the extent and condition of
22 ‘Benthic Broad Habitat Types’ (BBHT) which together cover the entire seabed of EU marine waters.
· Member States must map habitat extent, classify habitats and assign monetary value to mapped
European environmental economic accounts
classes in order to produce ecosystem accounts for terrestrial and marine ecosystems. The
(Regulation 691/2011, 2011)
upcoming 2024 report by EUROSTAT will include data on the marine environment.
· Understanding the spatial structure and connectivity among fish stocks, the location of Essential
Fish Habitats, and the sensitivity of benthic habitats in order to manage stocks and support the
Common Fisheries Policy environmental, economic and social dimensions of fisheries.
(Regulation EU 1380/2013)
· Site selection and expansion of sustainable aquaculture, which requires spatially explicit
knowledge about benthic biodiversity, physical properties and competing human activities.
· Planning, resolution of spatial conflicts and identification of synergies among sites for activities
Maritime Spatial Planning Directive
such as renewable energy, tourism, aquaculture, and fisheries, and in relation to achieving GES
(Directive 2014/89/EU, 2014), included
under MSFD and Favourable Conservation Status under the Habitats Directive. These decisions
within the Integrated Maritime Policy
must be based on sound knowledge of the spatial distribution of habitats and their ecological and
(COM/2007/574 final, 2007)
physical characteristics.
· Achieving the aims within the targeted policy areas for the marine environment, e.g. “Clean
European Green Deal (COM/2019/640 final, energy” (marine spatial planning of renewable energy), “Farm to fork” (site selection for
2019) aquaculture facilities and minimising fishing impacts) and “Biodiversity” (assessing status,
identifying suitable areas for restoration and for designation of MPAs).
· Identifying where habitats are, how big they are, their connectivity, and ultimately, their ecological
EU Biodiversity Strategy for 2030 status to enable MPAs and Natura 2000 site networks to be designed to accurately represent
(COM/2020/380 final, 2020) ecological processes in order to achieve the target of protecting at least 30% of European seas by
2030 (30x30 target).
Sustainable Blue Economy Strategy
· Planning the sustainable growth of the marine and maritime sectors.
(COM/2021/240 final, 2021)
· Knowledge on the distribution and extent of the European habitats, including marine habitats,
is required to assess the percentage of each habitat in poor condition and therefore in need of
EU Nature Restoration Law (proposed;
restoration measures. In its current form it will require an unprecedented effort to map the current
COM(2022) 304 final, 2022)
(and in some cases past) distribution and extent of marine habitats, which are already severely
affected by decades or even centuries of human impacts.

12
 N° 11 2024

The scaling-up of offshore renewable energy and other Blue and stakeholders to test the outcomes of different management
Economy activities are planned as part of the European Green Deal decisions using a virtual representation of the Ocean, also rely heavily
and the EU Sustainable Blue Economy Strategy, for which MHMs on the availability of high-resolution MHMs.
will aid spatial planning (Danovaro et al., 2024). On a global level,
the successful implementation of Sustainable Development Goal
14: Life Below Water (SDG14), also requires detailed knowledge of 1.2.3 Marine habitat maps support industry
the distribution of marine biodiversity and ecosystem services. MHMs are useful for some industry-led applications and are
valuable for many different economic developments (see Box 1
for further information on the economic return on investment
1.2.2 Marine habitat maps are essential tools
of MHM). MHMs should be prioritised in environmental impact
for informed management decisions assessment studies for new Blue Economy activities such as the
Ecosystem-based management is urgently needed but rarely siting of offshore wind farms, aquaculture facilities and underwater
implemented effectively due to a substantial lack of knowledge about pipelines that are expected to occupy large areas. A further example
biodiversity distribution and status, and ecological processes occurring of the importance of MHMs, is the Marine Stewardship Council’s
in space and time. MHMs can enable the successful application of (MSC) ecolabeling for seafood, where fisheries must demonstrate
ecosystem-based management through simultaneously visualising that they are carefully managed and do not affect the structure,
various types of information (e.g. of human activities, species, productivity, function, and diversity of the marine ecosystems.
ecosystem services), which can help to prioritise areas to be restored Local MHMs that include information about the distribution and
and protected. Other management tools, such as the development status of important ecosystems are used for this purpose (e.g.
and application of “digital twins”, which aim to allow decision-makers Morris et al., 2023).

Box 1 The financial return of marine habitat mapping

In 2008, PricewaterhouseCoopers (PwC) were commissioned to undertake a detailed appraisal of Ireland’s national marine
mapping programme, INFOMAR6, which delivers comprehensive marine datasets for Irish waters to multi-sectoral end-users
(see Table 4.1 for more information). PwC evaluated the costs and benefits, with benefits being identified and categorised
as: commercial/resource; knowledge economy; legislative requirements and obligations; and environmental. The analysis
estimated a four to six times return on investment based on economic maritime activity of interest for policymakers and
private operators involved in offshore renewable energy, fishing and aquaculture (PricewaterhouseCoopers, 2008).

Credit: Alberto Colletti

To plan and spatially prioritise active restoration interventions, we need to be able to document and monitor the location and extent of
degraded habitats.

6
www.infomar.ie

13
 EMB FUTURE SCIENCE BRIEF

1.3 Challenges facing marine habitat mapping

Effective management and conservation of marine habitats is more predict the presence of species in unsampled areas based on
difficult than terrestrial habitats, due to limited knowledge of their known physical features from depth data (i.e. bathymetry) and
conservation status. Approximately 27% of marine habitats in Europe environmental conditions (see Chapter 3). Reliable predictive
are classified as ‘unknown’ (Maes et al., 2020) and only a very small models require good data on environmental conditions that are
fraction of the seabed has been mapped at comparable resolution to correlated to in situ observations to ground truth the distribution
that on land (Wright & Heyman, 2008). This is in part because MHM of species and biological habitats (Stevens & Connolly, 2004).
is more complex and technically demanding than terrestrial habitat However, environmental data are not available in many areas
mapping due to difficulty in accessing the vast Ocean. There are or are not detailed enough to be useful for predicting species
various technical difficulties of mapping through water of varying distribution. The use of proxies can therefore limit the accuracy
depths, including the lack of light penetration, which greatly limits of MHMs, and their value for use in management, conservation
optical remote sensing techniques such as satellite imaging. and restoration.

This means that the extent and resolution of seabed mapping is Despite the critical need to refine distribution maps of many
highly variable, with 24.9%7 of the seabed currently mapped using species and biological communities, direct, in situ observations
bathymetric data. Typically for chartering purposes, the same area of marine life on the seafloor and in the water column (e.g. taken
hasn’t been converted into MHMs due to the lack of ground truthing by SCUBA diving and Remotely Operated Vehicles, ROVs) are less
data necessary for modelling habitats. MHM is therefore more abundant across space and time than physical mapping data,
challenging than mapping of other remote locations in our solar limiting the ability to accurately map their distribution. In addition,
system which do not have liquid surface water and the resolution the identification of species is difficult and typically requires the
of seabed data is significantly poorer than similar surface mapping collection of physical samples, as many species cannot be identified
of other planets. For example, NASA’s MErcury Surface Space from images alone. Taxonomists capable of identifying species
ENvironment, GEochemistry, and Ranging (MESSENGER) has are becoming increasingly rare, while new technologies such as
mapped the entire surface of Mercury at 166m resolution (Ernst et environmental DNA (eDNA11; Thomsen & Willerslev, 2015) have
al., 2022); NASA’s Magellan spacecraft mapped 98% of the surface limitations, including incomplete reference databases. For these
of Venus at a resolution of around 100m (Sauders et al., 1992; and reasons over 90% of marine species are estimated to be unknown
NASA’s Mars Reconnaissance Orbiter8 has imaged the entire surface to science (Mora et al., 2011). The importance of marine habitats
of Mars at 100m resolution, and over 60% of Mars has now been for biodiversity lies in the small-scale complexities of patterns and
mapped at approximately 20m resolution (Sidiropoulos et al., 2015). In processes, which require high spatial, temporal and taxonomic
comparison, satellite altimetry9 has mapped the entire seabed but resolution. New technologies that increase the spatial coverage
only at a resolution of 5900m on average (Tozer et al., 2019). of high-resolution direct observations are emerging, and show
promise for improving the quality and resolution of MHMs (see
Key challenges to mapping the Ocean floor include: (i) the need Chapter 2).
to rely on acoustic rather than optical techniques in aquatic
environments; (ii) the demanding engineering required for working A further challenge is the tendency for MHM efforts to focus
in deep, high-pressure environments; and (iii) the high cost of mostly on benthic habitats. Currently, the water column is mostly
mapping expeditions including the need for ships and specialised mapped according to single physical variables (e.g. salinity,
equipment (see Table 2.1 in Chapter 2). The reliance on acoustic temperature or physical currents). However, three-dimensional
techniques (i.e. remote sensing techniques using sound rather than mapping of large ecosystem patterns and processes, such as
light) for seafloor mapping in all but the shallowest water restricts ecological connections including life cycles, food webs and
the detection of the seabed to predominantly physical features and biogeochemical cycles (Boero et al., 2019), still largely do not exist
properties (Brown et al., 2011) rather than biological information. due to significant data gaps. These gaps can greatly limit MHM
This process is slow and it is estimated that it would take almost ambitions and are important to overcome for the production of
125 years to fully map the seafloor using acoustics10. predictive models of species distributions. Combining data from
different scales and collected using different techniques also
Due to the difficulty in accessibility to collect direct, in situ poses problems and often results in uncertainties with the use of
observations, MHM to date has heavily relied on models to proxies, and ultimately, end products.

7
https://seabed2030.org/our-mission/
8
https://mars.nasa.gov/resources/8333/a-decade-of-compiling-the-sharpest-mars-map/
9
Satellite altimetry is a technique used to measure the height of the Ocean’s surface from space, which varies depending on bathymetry therefore indirectly providing
information about the seabed.
10
https://www.nist.gov/how-do-you-measure-it/how-do-you-measure-depth-ocean#:~:text=Despite%20the%20advantages%20of%20using,world%27s%20
oceans%20have%20been%20mapped
11
eDNA is genetic material collected directly from environmental samples such as sediments or seawater.

14
 N° 11 2024

2 Collecting data for

marine habitat mapping
Data collection to create MHMs can be carried out via remote sensing and/or through direct observations
(i.e. in situ data) (Figure 2.1; see Glossary for definitions of technical methods referred to in this Chapter).
Remote sensing techniques typically collect data at a distance from the mapped area (e.g. satellite imaging
from space or sonar data for the seabed collected from a surface ship or mid-water autonomous platform)
and generate continuous spatial data. In shallow areas satellites, Light Detection and Ranging (LiDAR)
and drones (i.e. Unmanned Aerial Vehicles, UAVs) are emerging as viable remote sensing techniques for
data collection. In situ data are collected via direct observations from the seabed and water column
using various types of samplers, cameras and/or SCUBA divers and provide information on physical and
biological properties, which is required to ground truth remotely-sensed observations. SCUBA diving is
used to collect ground truthing data particularly in environments that are challenging to reach using other
techniques, such as submarine caves. In general, public-private partnerships can help to advance novel
methods to collect MHM data.

Both remotely-sensed and in situ data can be used in isolation to one metre, collecting precise and detailed measurements, enabling
produce maps. Remotely-sensed data have extensive coverage, more accurate mapping of coastal waters. In addition, UAVs can be
but low biological resolution, while in situ data are highly localised systematically adopted to facilitate mapping of coastal habitats
observations containing a high level of biological detail of habitat across areas that are tidally restricted to other mapping equipment,
types. If habitats cannot be directly detected in remotely-sensed and can also collect ground truthing data.
data, models can be used (see Chapter 3) to link and extrapolate
in situ observations with remotely-sensed data to produce maps Multibeam Echosounders (MBES) are widely used both in the
at resolutions and spatial extents appropriate for the aims of the deep sea and shallow water, to collect contiguous (i.e. continuous
specific MHM initiative and including relevant physical and/or across a geographic area) data over large areas about bathymetry
biological variables (e.g. Angeletti et al., 2019). (depth and shape), softness (an indicator of substrate type), and
roughness/rugosity of the seabed (an indicator of substrate type
and habitat complexity) (Lurton, 2010). The resolution and coverage
2.1 Collecting remotely-sensed data of MBES bathymetry and backscatter12 data are governed by water
depth and by the technical specificities of the MBES device such as
2.1.1 Current methods frequency. In areas down to a depth of 200m, higher frequencies
Remote sensing techniques vary in resolution, spatial coverage are ideal for achieving greater resolution. MBES coverage is directly
and scope of application. For the sea-surface and shallow-water proportional to water depth, therefore MBES is more cost-effective
benthic and pelagic habitats (in clear waters), satellite and airborne in terms of acquisition time in deeper waters. Higher-resolution data
methods are cost-effective ways of collecting data over large areas. can be acquired by bringing the MBES closer to the target, e.g. using
Depending on water properties, these methods can look into the platforms that operate closer to the seabed such as Autonomous
top 0-100m of the Ocean, and have been particularly effective at Underwater Vehicles (AUVs). These methods provide contiguous
mapping seagrass and kelp habitats at shallower depths (e.g. Casal three-dimensional data (i.e. benthic and water column) that serve
et al., 2011). For the seabed, remote sensing can be used to directly as the basis for the production of MHMs. These data, together with
detect features and can generate satellite-derived bathymetry seabed ground truthing using direct samples obtained from grab
data. An example of an airborne method is bathymetric LiDAR, sampling and/or photographs, allows for geological interpretation
which relies on infrared and blue-green laser pulses to measure and detection of biogenic habitats due to the differences in their
down to 40-70m below sea level, although resolution and accuracy signals. Bathymetric LiDAR, where feasible (i.e. in clear or turbid
are impacted by visibility, with reduced efficiency in turbid waters. water), has increased survey efficiency when compared to shallow-
They can measure the depth of the seabed to an accuracy within water MBES (Prampolini et al., 2020).

12
Backscatter data measures the intensity of sound waves released from Multibeam Echosounder devices reflected back from the seabed and are used to measure
substrate softness and texture.

15
 EMB FUTURE SCIENCE BRIEF

3
2 5
2

Credit: G. Castellan (CNR-ISMAR).

Figure 2.1 Overview of MHM approaches over a depth continuum, which can be used according to the aim of the mapping initiative. MHM from
intertidal to shallow depths is performed by satellites (1), Unmanned Aerial Vehicles (AUVs) or drones (2), snorkelling (3) and scuba diving (4).
Multibeam Echosounder systems and side-scan sonar on board oceanographic vessels (5), unmanned underwater technologies e.g. Remotely
Operated Vehicles (ROVs) (6), and Autonomous Underwater Vehicles (AUVs) (7) are used for MHM in deeper water. Note that multibeam echosounder
and side-scan sonar can be used for MHM in both shallow and deep water.

Side-Scan Sonar (SSS) is also widely used for MHM in shallow and 2.1.2 Future trends
deep water. It operates by emitting acoustic pulses to the sides Future aims for remotely-sensed data collection include efforts
of a survey vehicle or sonar tow-fish (i.e. an object carrying sonar to increase resolution, spatial coverage, information content,
equipment that is towed behind a vessel). The returning echoes operational efficiency and cost-effectiveness. Data should
provide detailed imagery of the seafloor and of objects. It provides increasingly be processed within the sensor itself or at a local
higher resolution images compared to MBES backscatter, particularly collection point to allow building of virtual Ocean environments
in deeper water where the tow-fish is closer to the seabed. Both SSS for applications such as situational awareness (e.g. for pilots to
and MBES backscatter technologies are valuable and serve different be able to comprehend the environment around their robot),
purposes, with SSS focusing on detailed imaging and being more virtual research environments (e.g. immersive virtual reality
suitable for object and feature detection, while MBES backscatter displays of complex data streams) and digital twins (i.e. coupled
emphasises bathymetric data and seafloor characterisation. observation and simulation data frameworks for human and
Artificial Intelligence (AI)-based scenario interpretation). Mapping
Hydro-acoustic sensors are also widely used by industrial with MBES at angles other than straight-down (e.g. by mounting
hydrographic surveyors and commercial fishermen to produce sonar heads at an angle) allows mapping of previously inaccessible
seabed maps that increase efficiency and reduce the environmental habitats like vertical walls, which although rare, are often inhabited
impact of their activities. This offers collaborative opportunities by diverse, sessile organisms (Zapata-Ramírez et al., 2016). The cost-
with scientists. effective collection of MBES data can also be supported through

16
 N° 11 2024

the continued collection of ‘underway’ data (i.e. opportunistic change and interact (see Figure 2.2). In parallel to collecting seabed
data collected during transits or non-mapping voyages). Such data, MBES should be used for water-column imaging (see Section
opportunistic data collection is currently being undertaken by Ships 2.1.3) to collect additional backscatter data on the pelagic habitat. In
of Opportunity13 and Seabed203014 Finally, the increased use of addition, time is also a highly relevant fourth dimension (European
autonomous underwater and surface platforms offers numerous Marine Board, 2019) that complements dynamic three-dimensional
advantages including: (i) reduced survey costs; (ii) access to remote MHM approaches and should be taken into account to assess the
and challenging locations; (iii) improved data resolution through functioning of ecosystems, also reflecting the rapid turnover of
greater proximity to the seabed; and (iv) a greater potential to life forms and seasonality. High resolution temporal data can
dramatically de-carbonise data collection. be collected from Ocean observatories and can provide valuable
data on the stability of habitats. However, this is still challenging
Irrespective of the depth and the instrument to be used, future considering that we are very far from having mapped the Ocean
trends in the collection of MHM data should recognise that the even once at sufficient resolution. Including the time dimension
Ocean is an interconnected three-dimensional volume where in mapping can further support an ecosystem-based management
physical, geological, biogeochemical and biological characteristics approach for marine ecosystems.

Credit: George Stoyle © NatureScot.

Mapping with MBES by mounting sonar heads at an angle allows mapping of habitats like vertical walls, which are inhabited by diverse, sessile
organisms.

13
https://community.wmo.int/en/ship-opportunity-programme
14
https://seabed2030.org/

17
 EMB FUTURE SCIENCE BRIEF

Figure 2.2 Graphical representation showing the integration of benthic habitat and water column data to show the three-dimensional structure of
a deep-sea ecosystem. Three-dimensional marine habitat mapping includes multiple depth ranges of the distribution of biodiversity and includes
species distributions by incorporating their life cycle, trophic interactions and exchanges between the water column and the seafloor (Levin et al.,
2018; CC BY 4.0 DEED).

To compare large-scale maps over time, the repeated collection dependent (Lurton et al., 2015; Figure 2.3). These ambiguities can be
of MBES backscatter data is particularly valuable. The global addressed by proper design of ground truthing and the acquisition
consortium “GeoHab Multibeam Backscatter Working Group”15 is of data with multispectral MBES, where the sensors acquire several
working towards improving technology and standards for this type MBES data using different acoustic frequencies simultaneously.
of data collection. In theory, repeat mapping over time (using remote This can result in increased contrast between seabed features and
sensing and in situ data) would help to understand: the longevity substrate types (e.g. mud, coarse sediment, rock), and thus increase
of maps and mapping data (i.e. how long they remain relevant the predictive power of the data for MHM applications. However,
determined by rate of change of mapped features and human multispectral mapping is challenging due to extensive data being
pressures); mobility of features and likely seasonal influences on required from multiple frequencies, which demands robust ground
marine communities. It would also help with monitoring condition truth support. Limited and inaccurate data constrain utility,
and/or recovery in designated areas and early warning for tipping hindering correlation between in situ substrate observations and
points16, which has not been widely studied (see Rindi et al. 2024 high-resolution acoustic data. This restricts the achievable detail
for an example). However, mapping vast and unknown areas of the in multispectral mapping. Although more research is essential in
Ocean for the first time is still the priority. this field, and additional testing and validation of the methods are
necessary, the efficacy of distinguishing seabed features through
Some seabed types, such as coralligenous formations and maërl beds, the application of a multispectral mapping approach has been
show a similar acoustic signal in backscatter data caused by subtle proven across a diverse range of seabed sediment types (Brown et
variations within the habitat itself, that are scale- and resolution- al., 2019).

15
https://geohab.org/backscatter-working-group/
16
A tipping point is a critical point at which a rapid and unexpected shift is triggered and an ecosystem transitions to a new state with altered composition and
functioning

18
 N° 11 2024

Credit: CNR-ISMAR.
Figure 2.3 Backscatter data showing coralligenous formations (red polygons) and maërl beds (violet polygons) in different areas. The acoustic
backscatter signal is similar for the two habitats, and interpretation was only possible due to the ROV images collected in the areas.

2.1.3 Current methods and future trends for water

front, and time), revealing high spatial and temporal variabilities
column mapping (Zhang et al., 2015).
There has been less focus to date on mapping the water column
compared to benthic habitats. Current methods for water column Marine habitat mappers are interested in how pelagic
mapping include those used in modern oceanography: satellites processes influence benthic habitats, and equally, how seabed
for collecting surface oceanographic data (e.g. on seasonal geomorphology modifies pelagic habitats (see Figure 2.4. for an
climatology, chlorophyll-a concentration) that can be used in example). Modern MBES is able to collect water column data,
mapping; drifting “Argo”17 floats that collect data on temperature, which provides information on the distribution of acoustically
salinity and changes in climate and hydrological cycles down to a reflective features (e.g. fish, plankton, gas bubbles) within the
depth of 2,000m; underwater gliders equipped with a variety of water column above the benthic area surveyed. However, this
sensors; and other forms of AUVs that are becoming increasingly information alone is insufficient to map pelagic habitats. As
important tools for cost-effective data collection on both such, multidisciplinary groups (i.e. including marine benthic
environmental and biological variables from the water column and plankton mappers, and oceanographers) need to work
(e.g. plankton). An example is the use of an AUV to autonomously more closely at large spatial scales, to gather the required
detect an upwelling front and track its movement on a fixed environmental information to characterise pelagic habitats and
latitude in four dimensions (i.e. vertical, across the front, along the processes connected to the seabed.

Credit: Federica Foglini (CNR-ISMAR) and Fabio Di Giovanna (UNINA).

Figure 2.4 Image of the water column derived from multibeam data showing (A) fish and (B) a gas plume.

17
https://argo.ucsd.edu/

19
 EMB FUTURE SCIENCE BRIEF

Credit: Alberto Colletti (right).

In situ data collection is currently mostly conducted using cameras mounted on moving platforms such as remotely operated vehicles (left). However,
imaging techniques do not provide the same level of taxonomic identification as physical samples taken using grabs (middle) and SCUBA diving (right).

2.2 Collecting in situ observations for marine habitat mapping

The cost-effectiveness of AUVs is also improving due to
2.2.1 Current methods miniaturisation, allowing deployment from smaller vessels or even
Traditional in situ data collection techniques include the use of from the shore. The greater duration of time for which modern units
grabs, corers, dredges, trawls and SCUBA divers to collect samples are able to operate is also helping to lower costs. Furthermore, AUVs
and observations. Observations are currently mostly conducted are able to operate much closer to the seabed when compared with
with cameras, which is a non-destructive method providing surface platforms. This proximity facilitates the collection of higher
high resolution data on the water column, seabed substrate and resolution products such as micro-bathymetry (i.e. capturing fine-
epibenthic communities (i.e. on or just above benthic habitats), scale details such as ridges and valleys) from in situ lasers and full
often at depths that are inaccessible to divers (Prampolini et al., coverage photo mosaics from onboard cameras.
2020). Although imaging techniques do not provide the same
level of taxonomic discrimination as physical samples taken using Making these vehicles intelligent enough to analyse raw data and
destructive methods, they do provide information on substrate, implementing communication within an underwater network of
habitat structure, biological community composition, resource robots can guide the selection of optimal observation and sampling
abundance, biomass, etc. Video footage provides more information locations, and steer vehicles efficiently and safely. This will help to
about the extent and patchiness of habitats compared to still images, implement efficient and cost-effective mapping, which maximises
while still images tend to have higher resolution, allowing finer level information obtained while minimising the resources and efforts
of taxonomic identification. Cameras are typically mounted on required. The suggested approach involves the initial production
moving platforms such as ROVs, AUVs, towed underwater platforms of an overview map by intelligent sensor platforms that reduce
or drop cameras suspended from a cable directly below a ship. and pre-classify data using AI, which speeds up later data analysis.
Extracting information from images has typically been a laborious Keeping these platforms in the Ocean, without the need for
task, mostly done by manual image annotation (Castellan et al., monitoring by a nearby surface vessel, substantially increases the
2020) resulting in a serious bottleneck in the production of MHMs. area from which data can be collected. Such persistent autonomous
Current approaches in using AI for this task have shown promising systems are well suited to map and monitor seasonal changes in
results, yet need to be further developed and standardised to allow habitats (including both the seabed and the water column) and the
wide adoption and application in monitoring tasks (see Section 2.3). identification of thresholds of changes and species invasions.

Despite still being very challenging, eDNA is a promising approach

2.2.2 Future trends to support broad scale MHM as a ground truthing technique.
In the future, in situ data collection should aim to decrease costs eDNA methods offer a number of important advantages over
(Table 2.1) through the greater use of autonomous platforms traditional techniques, including non-invasive sampling, and lower
(i.e. AUVs) and AI for image processing. This will mean moving cost and effort. Enabling the characterisation of biodiversity across
away from single, all-purpose, massive and complex vehicles broad taxonomic groups, eDNA can provide valuable insights
that serve all science demands towards multiple, low-cost, low- into biodiversity patterns and processes, including shedding light
operation profile vehicles with one, or just a few sensors that on the consequences of anthropogenic pressures and informing
operate in parallel. This also increases mapping efficiency, as management actions. However, there are still important limitations,
tasks can be done in parallel. Operating AUVs together with including the lack of standard protocols for eDNA sampling in the
other robotic platforms such as benthic crawlers (i.e. robots field and analysis in the lab, and gaps in taxon coverage in reference
that move independently carrying scientific instrumentation libraries, which means that many sequences derived from eDNA
for scanning a continuous track of the seabed for periods analysis cannot be assigned to their source taxon. This currently
longer than one month), or landers (i.e. robots able to provide limits the value of eDNA for high-resolution map production. For
high-resolution time-series data at fixed locations), can build more information on eDNA, see the Ocean and Biodiversity Chapter
a network, or swarm, of sensors to map an area efficiently. of Navigating the Future VI18.

18
https://www.marineboard.eu/navigating-future-vi

20
 N° 11 2024

Credit: PlanBlue.
Hyperspectral imaging technology offers the potential to collect more detailed and efficient in situ mapping data.

Hyperspectral imaging can be used to collect in situ imagery in the Underwater photogrammetry has increasingly been used (e.g.
visible portion of the electromagnetic spectrum (390-700nm) at up to Figure 2.5) to accurately measure the three dimensions of benthic
1nm resolution (Montes-Herrera et al., 2021). It captures images using habitats to represent their complexity. Photogrammetry uses
many wavelengths, resulting in finer resolution and more detailed multiple overlapping photographs to determine the size, shape
information. This additional optical resolution is useful for broader- and position of features. Recent developments in hardware and
scale automated classification of species and seabed features, each of image processing have made the reconstruction of high-resolution
which has a distinct pattern of electromagnetic light that is reflected three-dimensional models of relatively large areas (1ha / 0.01km2)
across different wavelengths. To collect more detailed and efficient possible (Pulido Mantas et al., 2023). The value of photogrammetry
in situ mapping data, additional light sources can be used that induce derived from imagery is that expensive sonars are not required. The
fluorescence within a species (e.g. Teague et al., 2019). However, due limitation is that it is dependent on optical imagery, which typically
to their cost, these cameras are not currently widely used. Equally, the limits data collection to very close to the seabed (limiting the spatial
database of spectral signatures required to match (i.e. cross-reference) extent of mapping) or only mapping in very shallow waters, using
and identify species and features, must be greatly expanded to enable diving or aerial platforms.
reliable and broader application of this technique.

A B C

Figure 2.5 3D models of a sea anemone generated using photogrammetry: (A) in situ image, (B) retracted anemone, (C) fully extended anemone
(Marlow et al. 2024; CC BY 4.0 DEED).

21
 EMB FUTURE SCIENCE BRIEF

Table 2.1 Overview of MHM data collection platforms, data types, achievable European Union Nature Information System (EUNIS) level
(see Section 3.3), associated use cases, spatial scale and resolution, and estimated acquisition costs. Costs are intended to provide a very broad
indication of the order of magnitude of the economic effort according to the different aims, resolution, platforms and research frameworks of
MHM initiatives.

POTENTIAL
EUNIS LEVEL RESOLUTION COST
PLATFORM DATA TYPE FIT-FOR PURPOSE USES EXTENT
ACHIEVABLE [m] [€k/km²]
[km2]

Gravity data Large-scale/low-resolution

to measure inferences on physical/
Satellite Up to 4 10 trillion 1000 0.5
bathymetry/ biological habitats, predictive
Remote images modelling
sensing Large-scale/low-resolution
Ship
inferences on physical
>1000m MBES/SSS Up to 3 1000 100 5.1
habitats, predictive
water depth
modelling

Small-scale/high-resolution
MBES/SSS inferences on physical
AUV Up to 3 10 1 3.2
Remote and images habitats, predictive
sensing modelling
and in situ Mapping of biological
observations UAV/drone
habitats, predictive
<10 m water Images Up to 6 0.1 1 0.5
modelling, in situ
depth
observations

Mapping of biological
ROV Images Up to 6 0.1 <0.01 4.3
habitats

Small boat Mapping of biological

Images from
<20 m water Up to 6 habitats, predictive 1 10 1.6
drop cameras
depth modelling

Images
and in situ Mapping of biological
In situ SCUBA diving observations Up to 6 habitats, direct mapping of 1 10 4
observations (e.g. species features
lists)

Sampling
(grabs,
Sediment Mapping of biological
dredges
and faunal Up to 6 habitats, predictive 0.1 1 8
cores, etc.
sample modelling
using small/
big vessels

22
 N° 11 2024

2.3
A
Integrating artificial intelligence withinB marine habitat mapping
The application of AI techniques, such as machine learning and The taxonomic ‘identification’ of species is the greatest challenge
deep learning (i.e. learning patterns directly from data), have for AI. The automation of biodiversity recognition through analysis
the potential to contribute to and revolutionise many aspects of of videos and photographs has a high potential for the development
MHM, from the acquisition of data through to the production and of biodiversity monitoring programmes. An increasing number
interpretation of end products, including automated image analysis of online platforms now automate species identification e.g.
(Figure 2.6). The recent migration to autonomous survey vehicles has iNaturalist19, Ocean Vision AI20, Squidle+21, Video and Image
generated opportunities for AI to plan missions and/or respond to Analytics for Marine Environments (VIAME22), Ecotaxa23, CoralNet24
the detection of features, for single devices and swarms of multiple and Linne Lens25. Most of these identification systems still have
devices using onboard autonomous decision-making. The future limited capabilities to identify multiple species in the same image
proliferation of autonomous vehicles and the advancement of and the size of species that they can ‘see’. This limits their ability
digital cameras could drive an exponential increase in the quantity, to provide quantitative estimates of abundance, which hinders the
quality and complexity of habitat imagery. Traditionally, the number possibility of deriving meaningful biodiversity indicators (e.g. GES
of days at sea limited the number of images that could be collected. for MSFD) compared to techniques that detect many more species
The availability of human expertise to examine and annotate these per sample, such as grabs, cores and SCUBA. Furthermore, many
images is also a key limiting factor. AI techniques could increasingly species can only be identified through dissection and examination
drive the development of machine vision, where machines are able of internal structures (e.g. most sponges), impairing the ability to
to autonomously perceive, interpret and understand visual data. identify species from such data.
This approach, in combination with expert opinion, is reducing the
effort needed to manually annotate the presence and location of
species in large datasets (e.g. Piechaud & Howell, 2022).

SCUBA, corers,
grabs

Sampling for eDNA

Low-cost
biological/physical/
Data resolution (e.g. observations or pixels per area)

cameras &
chemical variables hyperspectral
imaging
AI
/ Su
Ma p er-
ROV / AUV Photogrammetry ri ne res
ha olu
bi t tio
at n
ma
pp
i ng
LiDAR / Drones Robotic swarms

Multispectral
AUV MBES
MBES

Ship MBES Water-column

(Shallow water) imaging

Ship MBES Gravity

(Deep sea) data/bathymetry

Satellites

Spatial data coverage efficiency (e.g. area mapped or observed per time)

Figure 2.6 AI can be used across the various techniques for data collection, enabling super-resolution (i.e. enhanced resolution) at scale.
Green boxes indicate examples of data obtained from remote sensing techniques; pink boxes indicate data from in situ techniques;
and blue boxes indicate data obtained from both remote sensing and in situ techniques.

19
https://www.inaturalist.org/
20
https://www.mbari.org/news/ocean-vision-ai-uses-the-power-of-artificial-intelligence-to-process-ocean-imagery/
21
https://squidle.org/
22
https://www.viametoolkit.org/
23
https://ecotaxa.obs-vlfr.fr/
24
https://coralnet.ucsd.edu
25
https://lens.linne.ai/en/

23
 EMB FUTURE SCIENCE BRIEF

Credit: Luca Licciardi.

The taxonomic ‘identification’ of species is the greatest challenge for artificial intelligence.

These systems are also limited in their capacity to identify objects of size and compatibility of training data needed by the AI community.
interest and to classify them due to the variability in the underwater In the absence of high-quality data for training and validating AI
imagery (e.g. natural variability of individual samples, lighting, fields models, there is a risk of inaccurate identification or quantification
of view, changes in orientation, background habitats, visibility) and of species or features on the seabed, thereby inflating map error.
the shortage of annotated imagery that is manually classified by
humans for training AI models. In addition, many of the existing Demand for machine vision and AI methods for the assessment
annotated datasets are not produced or stored in a format that of marine ecosystems is growing rapidly, driven in part by greater
is immediately available for AI training. The application of deep- access to autonomous Ocean observing systems (Durden et al.,
learning techniques, such as deep convolutional neural networks, 2021) and the number of applications using this technology (e.g.
can partially overcome some of the background variation common in deep-sea mining exploration, monitoring of MPAs and restoration
seabed imagery (Salman et al., 2016) and overcome issues associated sites). Thus, the development of AI techniques for extraction and
with shortages in training data (Malde et al., 2020). The wider use of ecological data will require an ever-closer collaboration
implementation of standardised annotation systems for seabed between computer scientists, marine ecologists and environmental
imagery with data in the correct format, will greatly improve the policy specialists (Guidi et al., 2020).

24
 N° 11 2024

2.4 Recommendations
To advance data collection for MHM, we recommend scientists/ 2d) Implement cost-effective mapping
maps producers and research funders to: Remote sensing and direct, in situ data collection need to be cost-
effective. To achieve this, the continued development, adoption
2a) Further integrate biological data and coordination of autonomous platforms, such as AUVs and UAVs
Scientific attention and funding should increasingly be directed should be financially supported by EU, national and international
towards improving knowledge of the distribution of marine sources. Strategic coordination towards interoperability of
species and habitats, which is still extremely limited. More maps robotic platforms should also be supported. Important areas for
that characterise and represent the distribution and extent of the development include enhancing platform reliability, endurance,
biological components of marine habitats are critical for increased capability, reducing unit cost, and developing multiple networked
understanding of the ecosystem patterns and processes needed vehicles equipped with complementary sensors capable of
to promote scientifically-sound conservation, restoration and operating cooperatively while adapting their spatial and temporal
management decisions. The adoption of innovative technologies sampling strategies in real time using onboard decision making.
to collect more high-resolution biological data and to improve the It is recommended that sufficient resources are allocated to train
spatial and temporal scale of cost-effective mapping is a priority. and maintain sustainably sized teams of multidisciplinary experts,
including taxonomists. Repositories are needed that facilitate
Findable, Accessible, Interoperable, and Reusable (FAIR) data
2b) Further integrate water column data publication (see recommendation 5a) to support the collection and
Three-dimensional MHM that integrates benthic habitats and the distribution of information from autonomous sources, which will
water column should be supported. This will enable more high- in turn improve cost-effectiveness. Cost-effective mapping can
resolution data on species distributions to be included in mapping, also be supported through the continued collection of ‘underway’
taking into account life cycle, trophic interactions and exchanges data (i.e. opportunistic data collected during transits or non-
between the water column and seabed. Three-dimensional MHM mapping voyages). Stronger national and regional coordination for
that includes connectivity in ecological systems can be incorporated shared resources and facilities is also required e.g. via European
into 3D systematic conservation planning, fundamental for infrastructures such as the European Marine Biological Resource
underpinning design and management of conservation areas. Centre (EMBRC26).

2c) Support temporal/repeat surveys 2e) Further integrate AI

The majority of seabed habitats are relatively stable (e.g. rock, Efforts to promote the complete integration of AI into MHM should
mud). However, some geomorphological and hydrological features be supported in order to handle the exponential increase in the
are highly mobile (e.g. megaripple bedforms, the water column). volume of data collected from the combined use of autonomous
Equally, habitat condition and species composition can change platforms and modern sensors, and to enhance data acquisition.
significantly over time, meaning that repeat mapping surveys are The methods and infrastructure to embed AI in routine data
often necessary for monitoring change over time (i.e. the fourth- collection and processing should be made widely and freely available
dimension) e.g. every six years for the MSFD. Repeated mapping within the MHM community. AI-derived information needs to be
exercises are particularly helpful for the monitoring of naturally aggregated centrally, using standardised formats and metadata, to
dynamic ecosystems with strong seasonality and sensitive increase its availability. Lastly, efforts to promote the development
biogenic habitats, such as cold-water corals, seagrass and reef and retention of machine-learning expertise in the marine sciences
building organisms, which are vulnerable to human pressures. is necessary in the face of industrial demand.
This will require the development of cost-effective methods
(recommendation 2d), specifically the greater use of autonomy and
AI (recommendation 2e) to reduce costs and standardised methods 2f) Develop and apply standards for data collection
(recommendation 2f) to enable the reliable detection of change. and processing
Standardisation is essential to make MHM consistent across
borders and over time. Existing best practices should be collated
to produce standardised methodologies for the operation of
survey platforms, and remote sensing and ground truthing
techniques. Standardisation will, in part, provide provenance of all
data collection and processing steps. This information should be
provided in human and machine-readable formats and needs to be
published in a standardised manner, alongside MHMs i.e. in FAIR
data repositories and services (recommendation 5a).

26
https://www.embrc.eu/

25
 EMB FUTURE SCIENCE BRIEF

3 Combining data to produce

marine habitat maps
Although remote sensing and in situ ground truthing survey methods and technologies have improved
dramatically, it is challenging to provide full coverage MHMs of both the seabed and the water column
through these methods alone. Remotely-sensed data and models (e.g. biological models and Ocean
models of temperature and salinity) are therefore often used to extrapolate ground truthing observations
across the mapped area (Figure 3.1).

A physical habitat map delineates the environmental characteristics are included in maps, these offer understanding of both the physical
and features of a given area, such as substrate type (e.g. mud, sand, environment and the associated biological communities in an entire
rock, minerals), depth, seafloor morphology (including inclination/ ecosystem. In the early days of MHM, interpolation between samples
slope) and water flow. Physical habitat maps provide full coverage (i.e. the process of deducing the relationship between two points in a
at broad spatial scales and are created by segmenting the seabed dataset) and boundaries between habitats were determined by hand.
according to remotely-sensed abiotic variables, e.g. depth, seabed This was resource intensive and the availability of physical samples of
reflectivity (i.e. the acoustic energy reflected from the seabed), Ocean the seabed limited their spatial coverage. Increasingly, physical MHM
colour, and physical and chemical variables that can be modelled approaches and ground truthing, in situ observations are merged
globally. In contrast, a biological map illustrates the spatial distribution using distribution models, where mathematical relationships
of species, or communities of species, within that area, providing between physical attributes and biological units are used to predict
insights into biodiversity, species composition and potential for the distribution of habitats. These approaches are described in more
ecological interactions. When both pelagic and the benthic habitats detail on the next page.

Remote sensing e.g. acoustic

(MBES) survey of the physical
environment.

Additional environmental variables such

Photo credits: Elisabetta Campiani, CNR-ISMAR (top); Christoffer Engstrom, Unsplash (middle); Mišo Pavičić, IZOR (bottom)
as those derived from remotely sensed
data (e.g. backscatter, terrain) and
modelled products (e.g. hydrodynamics,
biogeochemistry).

Ground truthing (in situ, direct

observations) e.g. camera observations
of biology and substrate.

Geo-statistical model i.e. ground truthing

data ‘trains’ a model to predict species/
habitats across the environmental
variables.

Final marine habitat map with

uncertainty estimate.

Figure 3.1 Current workflow for the production of benthic habitat maps that combines remote sensing and derived variables from remotely-sensed
data or modelled products, with in situ observations within a geo-spatial model.

26
 N° 11 2024

3.1 Physical and biological habitat maps

A physical habitat map identifies areas with distinct physical narrow, distinct distributions, whilst others can tolerate a wide
conditions that are deemed to be suitable for certain groups of range of conditions, which makes it difficult to use physical proxies.
species. Among others, substrate type is an important physical Therefore, depending on the types of species and/or assemblages
driver of the distribution of species that live on or in the seabed, of interest and the region of interest, the usefulness of physical
and therefore physical habitat maps are often in the form of a habitat maps as proxies for biological habitats can vary.
substrate map. These are often created using a combination of
remote sensing data from Multibeam Backscatter (MBES) or Side- Biological habitat maps rely on the same data as a physical habitat
Scan Sonar (SSS), supplemented with photographs or physical map, plus biological data obtained via direct observations (e.g.
samples of the seabed. Additional information may be layered on underwater photography, grab sampling, MBES). Biological samples
top of a substrate map based on other physical variables known take more effort and time to collect, store and process than
to influence the distribution of biological assemblages, such as physical samples and are more expensive. It can also be difficult
depth, geological features, currents and light attenuation in the to use the biological information to inform the classification of the
water column. Together, this combination of physical information acoustic data derived from MBES or SSS because not all biological
can result in a more informative map that classifies the seafloor assemblages have a distinct signature that can be observed
according to narrower, biologically-relevant categories, which acoustically. The main exception to this is biogenic habitats, which
are sometimes created in the absence of any biological data. This tend to have acoustic signatures that distinguish themselves from
is the approach taken in the production of the European Marine the surrounding substrate.
Observation and Data Network (EMODnet) broad-scale seabed
habitat map for Europe, known as “EUSeaMap” (Vasquez et al., As a result of the difficulty in biological habitat mapping, there are
2023; see Figure 3.7). Whilst relying heavily on physical proxies, far more substrate maps (i.e. focusing on physical characteristics
sometimes at a coarse resolution, its complete coverage of European of the seabed) than biological habitat maps in Europe (see Figure
seas makes it useful, although not sufficient, to inform regional- 4.3. showing coverage of substrate and biological habitat maps in
scale and national marine ecosystem assessments such as for the European sea basins). However, the development of reliable biological
MSFD, Regional Sea Convention assessments and the integrated habitat maps covering large spatial scales is a prerequisite to assist
ecosystem assessments coordinated by the International Council decision-making processes and for environmental assessments such
for the Exploration of the Sea (ICES)27. as the MSFD. Mapping spatial patterns of marine biodiversity can be
carried out at a range of scales and the choice of scale depends on
Maps built using physico-chemical variables are often used as the aim of the mapping activity. For example, fine spatial resolution
proxies for habitats and consequently biological assemblages. data are required for the proper management of MPAs, including
However, this association is only applicable if the following criteria Natura 2000 sites, while broad-scale analyses are more suited for the
are met: (i) there is a sufficient understanding and training data of broad identification of candidate areas for MPA designation and the
the typical assemblages of species present; (ii) the environmental allocation of other human uses at sea using marine spatial planning.
requirements of the assemblages are understood; and (iii) the There is a need to develop guidance on appropriate scales of mapping
environmental requirements do not overlap too much. Some have required for different applications.

3.2 Distribution models in marine habitat mapping

Although some biological features can be detected directly from include: correlative models (i.e. relating known probabilities of
remote sensing data (e.g. biogenic habitats such as seagrass beds species presence to environmental variables); mechanistic models
from satellite imagery and cold-water coral structures from MBES), (i.e. relating physiological information about a species gained from
most current MHMs that contain biological information are the literature or laboratory experiences to environmental variables for
product of models. The most commonly used are distribution assessing their fitness at specific locations); and process-oriented
models, also known as habitat suitability models or species models (i.e. estimating species distribution based on processes
distribution models (Elith & Leathwick, 2009). These models such as ability to disperse and biotic interactions) (Melo-Merino
typically predict the probability of the presence, or the habitat et al., 2020). AI methods offer additional predictive power due to
suitability, for a given species, or selection of species when applying their ability to incorporate more complex interactions, and could
joint species distribution modelling (see Figure 3.2 for an example). lead to better representation of marine habitats’ multifaceted
They can use traditional statistical methods or AI methods, such as nature (e.g. Effrosynidis et al., 2018), however, increased model
machine learning, to determine the relationships between biological complexity could lead to challenges in interpretation of model
assemblages and environmental variables. Traditional methods outputs.

27
https://www.ices.dk/community/groups/Pages/IEASG.aspx

27
 EMB FUTURE SCIENCE BRIEF

Credit: Mariacristina Prampolini (CNR-ISMAR).

Figure 3.2 Examples of maps made using models. (A) Occurrences of cold-water corals (CWC) in the Bari Canyon. (B) Distribution of
CWC habitat inferred from geophysical data (bathymetry and side-scan sonar) and CWC occurrences (data from Prampolini et al., 2021).
(C) Modelled habitat suitability for a CWC (data from Bargain et al., 2018). Red areas indicate a high probability of CWC presence.

28
 N° 11 2024

Different approaches are used to model the distribution of species In principle, habitats and species distribution models based on the
and habitats, depending on the type of data available to fit the best available biological information can guide management and
model e.g. single species models using presence-only or presence restoration efforts: i.e. predictions of species or habitat presence/
with absence/pseudo-absence data (i.e. observations of where the absence, combined with information about human pressures
species of interest is not present or similar observations where (Fabbrizzi et al., 2020). The modelling process provides additional
proxies suggest the same species is highly unlikely to be present), information on the environmental variables potentially driving the
multi-species data (i.e. for joint species distribution modelling) or distribution of the modelled species or habitats (e.g. current speeds,
quantitative predictions using abundance, density or biomass (see depth, temperature and dissolved oxygen). Although modelled
Annex 2). However, these models do not predict the distribution outputs currently provide the best available evidence for the
of the ‘occupied’ habitat itself, but only the existence of a suitable ‘’potential” distribution of most marine species and habitats (i.e.
habitat or the probability of presence. They are, therefore, proxies non-biogenic habitats and species that cannot be directly observed
for the real distribution of a species or habitat, which will usually in remote sensing data), greater efforts need to be made to improve
occupy a smaller fraction of this space. Models can also be trained the accuracy, transferability and repeatability of these models.
on data on present conditions together with climate change Species distribution models need to be improved by the collection
projections to provide predictions of distribution under differing of higher quality environmental data at finer resolution and better
climate scenarios (Figure 3.3). species occurrence datasets.

Figure 3.3 Maps of the distribution of predicted habitat suitability (using salinity and temperature) for the seagrass Halophila decipiens in the
Mediterranean Sea under present conditions and future scenarios of climate change based on two contrasting carbon emission projections:
Representative Concentration Pathway (RCP) 2.6 (top) and 8.5 (bottom) by 2050 and 2100. Yellow shows additional habitat by 2050 in relation to
present distribution and red shows additional habitat in 2100 in relation to 2050 (Beca-Carretero et al., 2020; CC BY 4.0 DEED).

29
 EMB FUTURE SCIENCE BRIEF

Credit: George Stoyle © NatureScot.

EUNIS level 4 defines individual communities (or biocenosis) e.g. macroalgal communities dominated by kelp species.

3.3 Are European habitat classification schemes fit-for-purpose?

Habitat Classification Schemes (HCSs) are sets of instructions in the larger EUNIS information system29, which is a database
that identify, delimit, and describe the habitats of distinct species that brings together European data from several databases
and communities by categorising them into “classes” (Robinson & and organisations and contributes to the knowledge base for
Levings, 1995). Well-defined and comprehensive HCSs are central implementing the Biodiversity Strategy 2030. It is managed by the
to the production of MHMs. Once MHM data have been collected, European Environment Agency (EEA) and aims to cover all terrestrial,
HCSs facilitate the classification of discrete data (i.e. categorical data freshwater and marine habitats in Europe in a hierarchy that allows
such as substrate) and continuous data (e.g. salinity) into ecologically users to define habitats at different levels of detail (see Figure 3.4
relevant spatial units (Strong et al., 2019). These units are comparable for an example of a map classified using EUNIS).
between maps produced by different scientists, in different areas
and at different times because they use the same system to label For marine benthic habitats, EUNIS facilitates the comparable
areas hosting similar benthic assemblages. The use of a standard reporting of habitats for environmental management under
HCS adds significant value to a map because it allows the map to be several pieces of legislation at the national level. EU-wide, it
combined with other maps and/or translated into other HCSs and facilitates the six-yearly reporting of the MSFD, which includes
used for multiple purposes. For a comprehensive overview of how reporting the extent of Benthic Broad Habitat Types (BBHTs) and
different aspects of HCSs can influence the information content and Article 17 reporting of the Habitats Directive, which includes
format of MHMs see Strong et al. (2019). reporting the extent of Annex I habitats that Member States must
designate, protect and manage. For some HCS, listed habitats
There are several HCSs used to define marine habitats worldwide are not quantitatively defined in terms of species abundance, or
(Montefalcone et al., 2021). In Europe, the European Nature condition, which makes the use of these schemes less effective
Information System (EUNIS28) HCS is the most comprehensive and for the consistent production of maps and the subsequent use of
the latest version was published in 2022. The EUNIS HCS is included those maps for management.

28
https://www.eea.europa.eu/data-and-maps/data/eunis-habitat-classification-1
29
https://eunis.eea.europa.eu/

30
 N° 11 2024

The benthic marine section of the 2022 version of EUNIS divides the • Levels 4-6 add information about the distinct species and
classes into the following levels (note that level 1 separates terrestrial communities that can be observed within each level 3
and marine habitats):
class i.e. habitat classes defined by their species (e.g. the
• Level 2 is based only on substrate type (e.g. rock, biogenic octocoral Virgularia mirabilis and the sea star Ophiura
habitat) and broad biological zones related to depth (e.g. spp. with the bivalve Pecten maximus on circalittoral
littoral, infralittoral, circalittoral). These broad terms are sandy or shelly mud). They also include additional abiotic
applicable across all biogeographic regions of Europe and factors where relevant, such as substrate, depth and
translate directly to the BBHTs that EU Member States light. Level 4 defines individual biocenosis30/communities
must refer to for the MSFD (European Commission, 2017). (e.g. the Mediterranean photophilic algae biocenosis or
• Level 3 adds a qualifier that refers to the main macroalgal communities dominated by kelp species).
biogeographic regions of European seas (Arctic, Atlantic, Level 5 defines assemblages characterised by specific
Baltic Sea, Black Sea and Mediterranean Sea). This is species (e.g. the barnacle Chthamalus spp. on exposed
important because the biological character of habitats upper eulittoral rock) and level 6 has the greatest level
varies geographically, such that the same functional of biological and physical specificity (e.g. the barnacles
habitat (e.g. a surf beach, an exposed rocky shore) hosts Chthamalus montagui and Chthamalus stellatus on
different species and communities depending on its exposed upper eulittoral rock).
geographic location. This is caused by variation in abiotic
variables, particularly temperature and salinity, and
species origins via larvae transport from prevailing Ocean
currents.

Credit: EMODnet Seabed Habitats.

Figure 3.4 EUNIS (v2022) habitat map of the Wash Estuary, East Anglia, United Kingdom classed to level 5.

30
Biocenosis is a synonym for biological communities used in the Barcelona Convention.

31
 EMB FUTURE SCIENCE BRIEF

The adequacy of the previous version of EUNIS system was evaluated substrate classes. At small scales, different species assemblages can
by Galparsoro et al. (2012) and many of their recommendations inhabit the same habitat type, representing natural variability that,
were incorporated within the subsequent EUNIS updates, so that should not play a role in the identification of habitat distribution at
the 2022 version better reflects the main biogeographical regions large scale. This is a fundamental issue, which has the potential to
of Europe’s seas based on their distinct combinations of salinity and undermine comparisons across regions under the MSFD and wasn’t
temperature regimes. It also includes habitat cross-reference tables fully addressed in the 2022 EUNIS update.
to other regional HCSs such as HELCOM HUB31 and UNEP MAP-
RAC/SPA32 to support harmonisation of ecosystem definitions and Habitat classification levels (i.e. resolution of biological information)
mapping. However, limitations with the system remain, which are required by map producers and users need to be aligned. This
discussed further below. has been taken into account by EUNIS since its early versions and
addressed by adopting an approach with different levels that are
nested, and that can be chosen according to the cartographic detail
3.3.1 Standardisation of terms within
required. Nevertheless, an assessment is needed of how to close
classification schemes the gap between efforts to describe habitats for: GES assessment
For the consistent use of a HCS, the named habitat classes need for the MSFD; the Habitats and Birds Directives; the 30x30 target
to be clearly defined. This ideally requires the use of quantitative under the EU Biodiversity Strategy 2030; and prioritisation under
thresholds for defining habitats in terms of their spatial and the proposed EU Nature Restoration Law. The Habitats Directive
compositional properties. Furthermore, the variables used to Annex I habitat types include very broad and not very detailed
define the habitats, and ideally the condition of these habitats, typologies and particular effort is needed to align these with HCSs.
should be closely aligned to the variables reported by standard In addition, relevant Essential Ocean Variables (EOVs33) should
mapping techniques such as sonar (depth and intensity values), also be clearly correlated with EUNIS classes to avoid generating
cameras and particle size parameters delivered by cores and grabs. alternative typology and datasets for Europe.
The development of quantitative definitions of HCS classes is of
interest predominantly for the Atlantic in order to more precisely
3.3.2 Completeness and update mechanisms
define its benthic communities and annual changes should also
be considered. The development of quantitative definitions of for classification schemes
HCS classes is not considered a priority for the Mediterranean Sea, EUNIS is an important Europe-wide HCS, and the basis for EUSeaMap,
where a more qualitative approach is preferred due to differences which is the only Europe-wide habitat map. However, there are some
compared to the Atlantic in seafloor characteristics and biodiversity limitations. The original version of the marine section of EUNIS was
levels, which do not require in-depth definitions. based on the marine HCS for Britain and Ireland (Connor et al., 2004)
where most information was available at that time. The EUNIS
Both Galparsoro et al. (2012) and Strong et al. (2019) recommended system is currently widely used on the Atlantic coasts of Europe. Since
the inclusion of quantitative definitions of classes within HCSs 2004, EUNIS has expanded gradually to include classifications for the
to improve consistency in their application, particularly for soft Baltic and Mediterranean Seas, and newly-developed classifications
bottom sediment types. This would provide a more robust basis for: for pelagic habitats and the deep sea, advancing the system’s
(i) initial classification of habitats; (ii) the estimation of how well an comprehensiveness in terms of its geographical coverage of European
observation fits an assigned class; and (iii) greater certainty about seas. Recently, in the Mediterranean Sea, the Barcelona Convention
the detection of change in habitat condition, extent and spatial classification was revised to include new habitats discovered in the
configuration over time during repeat mapping. last 30 years (Montefalcone et al., 2021). It was conducted in parallel
to the update of EUNIS to ensure that the two systems are as aligned
The challenge for the 2022 version of EUNIS (and associated as possible. It would be desirable that other regional classifications
BBHTs under the MSFD), which uses common terms at level follow a similar alignment process in the future.
2 in an endeavour for consistency for all regions, is to strike a
balance between consistent definitions and biologically-relevant In the 2022 revision of EUNIS, some improvements were made
definitions across regions. These are sometimes in conflict due regarding the Atlantic region, however many other areas remain
to both regional differences in the predominant conditions that underrepresented, i.e. the Black Sea, Bay of Biscay and Azores
drive the distribution of biological communities and historical (Galparsoro et al., 2012), since only a small fraction of Europe’s
approaches to defining habitats in each region. Consequently, seas are well studied. The current update mechanism for EUNIS is
benthic species assemblages do not always fit neatly into the BBHT ad hoc and relies on a small number of experts from the European
defined combinations of substrate classes and depth (e.g. Cooper & Topic Centre on Biological Diversity34. The marine section of EUNIS
Barry, 2020). In some cases, the use of substrate classes as proxies requires an increase in resources in order to improve its update
for habitats is sufficient. However, there is large variability within mechanism.

31
https://helcom.fi/baltic-sea-trends/biodiversity/helcom-hub/
32
https://www.rac-spa.org/
33
https://goosocean.org/what-we-do/framework/essential-ocean-variables/
34
https://www.eea.europa.eu/data-and-maps/data-providers-and-partners/european-topic-centre-on-nature-protection-and-biodiversity

32
 N° 11 2024

It is widely accepted that HCSs require an element of generalisation an integrated, multi-scale and hierarchical approach to classify
so as to make the habitat classes more widely applicable. This can habitats from coastal waters to the deep sea (Boero et al., 2016). The
lead to the broader schemes having reduced applicability in areas challenge for a broad, unifying HCSs is to draw upon these bespoke
beyond where they have been developed. The poor fit of some schemes during update iterations without compromising their
classes in generic schemes continues to lead to the development generality or consistency of classification. Improvements should also
of alternative classifications. One example is the EU FP7 project be made to the way new biological habitats are proposed, reviewed,
CoCoNet, which developed the “CoCoNet Habitat Mapping Scheme”, accepted and published as part of EUNIS.

Credit: Guido Villani (top).

EUNIS has expanded gradually to include classifications for the Baltic and Mediterranean Seas, and newly developed classifications for pelagic
habitats and the deep sea.

33
 EMB FUTURE SCIENCE BRIEF

3.3.3 Additional attributes: human pressures

ecosystem-based management. This information is relevant for all
and ecosystem services EU and international environmental legislation to achieve targets
Habitat cross-reference tables (e.g. JNCC, 201835) enable the for the implementation of restoration measures and for monitoring
translation of a map into various HCSs, which can be useful of the marine environment.
for assessments and reporting for different legislation. It is
recommended that custodians of HCSs update their habitat Habitat sensitivity matrices assign categories of habitat sensitivity
descriptions to include additional attributes such as sensitivity to (in terms of resistance and resilience) to various human pressures e.g.
human pressures, conservation value (e.g. IUCN Red List36) , habitat using the Marine-Evidence-based Sensitivity Assessment (MarESA)
condition and ecosystem service provision (Strong et al., 2019). tool (Tyler-Walters et al., 2023). Linking a sensitivity matrix to a map
allows the creation of a map showing the sensitivity of habitats to
The increase of human activities is causing unprecedented changes a pressure and overlaying this with a map of human pressures can
to marine ecosystems. In some cases, the extent of these changes is indicate areas at the highest risk of impact. Maps of cumulative
so large that the structure and function of habitats and ecosystems human impacts have been produced for various areas of Europe
have no historical analogues (i.e. they are novel ecosystems) (Bulleri (Korpinen et al., 2021; Figure 3.5) and at multiple scales in order
et al., 2020), generating further issues for classification. Thus, habitat to combine multiple pressures into a single comparable estimate
condition should become a priority to include within MHM efforts of cumulative human impacts revealing relevant gaps (Bevilacqua
and EUNIS levels 4 and 5 (and 6 for the Atlantic, which is the only et al., 2018). However, a key limitation is the need for research on
area that this level is present). For most coastal marine ecosystems the most basic information, such as distribution of habitat types
there is little understanding of the impacts of multiple stressors, and whether and how different anthropogenic pressures interact
which is considered one of the most challenging questions for (Halpern et al., 2008).

Figure 3.5 Combined effects of anthropogenic pressures in Europe’s seas. See Korpin et al. 2021 for an explanation of the development of the index
values (Korpinen et al., 2021; CC BY 4.0 DEED).

35
https://mhc.jncc.gov.uk/resources#correlationtables
36
https://www.iucnredlist.org/

34
 N° 11 2024

The mapping and assessment of the ecosystem services provided 90% of the mapped area in European waters provides biodiversity
by marine habitats is also a highly valuable source of information maintenance and food provision services, while nursery grounds
for understanding their current and potential benefits to society. providing reproductive and nursery services are limited to half of
Galparsoro et al. (2014) showed that ecosystem services can be the mapped area. Benthic habitats generally provide more known
attributed to habitat classes, which allow a habitat map to be services closer to shore and in shallower waters, compared with
transformed into a map of ecosystem services, facilitating the deeper offshore habitats. This gradient is likely to be explained
valuation of the seabed and water column for natural capital by difficult access (i.e. distance and depth) and lack of scientific
accounting (Figure 3.6). Their results indicated that more than knowledge for most of the services provided by offshore habitats.

A B

C D

Figure 3.6 Maps of ecosystem services: (A) provisioning services; (B) regulating services; (C) cultural services; and (D) total ecosystem services
(Galparsoro et al., 2014; CC BY 3.0 DEED).

35
 EMB FUTURE SCIENCE BRIEF

3.4 Assessing and communicating accuracy and confidence

The availability of online maps and open source data has promoted Due to the complex nature and multiple sources of error in MHMs,
the use and adoption of maps for various purposes. However, it it is difficult to enforce a single, quantitative approach to describe
is impossible to produce maps that are completely accurate and confidence. However, it is necessary to consistently and accurately
satisfy all needs. Highly accurate terrestrial maps have led to the assess and communicate the confidence of MHMs so they can be
assumption that maps accurately show the location and name of used effectively by end-users who should be able to understand
objects, therefore maps are often used without considering their their limitations (see Figure 3.7 for EUSeaMap v2023 and Figure 3.8
accuracy (i.e. certainty within a map in terms of location and the for its associated confidence map). Failing to report accuracy may
quality of labelling for mapped units). Compared to terrestrial mean that too much trust and confidence are attached to maps,
habitat maps, MHMs rely more on proxies than direct observations. which may subsequently fail to support the desired purpose of
This is especially true for broad-scale maps where species and the end-user. There is often a discrepancy between map accuracy
habitats are modelled rather than being directly detected by remote and end-user expectation, which can undermine the mapping
sensing (i.e. fine-scale maps), meaning that the accuracy of MHMs is process and the products generated. It is, therefore, essential
more variable than that of terrestrial maps. that map producers report informative, standardised, and ideally
spatially explicit, measures of map accuracy and assumptions
MHMs are produced by interpreting multiple and varying types of used to generate maps. Equally, map users need to specify their
data (see Chapter 2), each with their own sources of error. There are requirements, in terms of accuracy for specific purposes, so that
also errors associated with analysis and interpretation techniques. map confidences can be estimated. There are several studies
In addition, many habitat classification definitions are imprecise that have sought to standardise the methodology for assessing
and one person’s understanding of a habitat may be different to accuracy when determining the value of and understanding the
another’s. Map accuracy is reduced by the accumulation of these source of error in MHMs (e.g. Mitchell et al., 2018). A consistent
errors within the map (Strong, 2020). Map error rates are estimated and widely adopted accuracy assessment will benefit both the
by using an independent dataset to test the predicted map classes. development of mapping methodologies as well as those needing
Map confidence relates to its fitness for a specific use and is to use the maps produced.
determined by its accuracy and its intended purpose by the end-user.

Credit: EMODnet Seabed Habitats.

Figure 3.7 EUSeaMap (v2023) broad-scale seabed habitat map for Europe.

36
 N° 11 2024

Figure 3.8 The 'confidence' map associated with EUSeaMap (v2023) giving an idication of the quality of the data sources and methods used to
create the map. Red = low, orange = moderate and green = high confidence (Vasquez et al., 2023).

37
 EMB FUTURE SCIENCE BRIEF

3.5 Recommendations
To advance distribution modelling and HCSs for MHM, we In addition, we recommend custodians of HCSs (e.g. the European
recommend scientists/map producers and research funders to: Environment Agency for EUNIS) to:

3a) Improve the datasets used in spatial models 3d) Develop quantitative definitions of HCS classes, and
Improving model outputs through high-quality data at finer regional definitions of broad habitats and
resolution on environmental variables and better species biogeographic regions
occurrences datasets is a priority. This would allow for more reliable Habitats need to be defined quantitatively (i.e. using variables and
predictions, and help to identify suitable and unsuitable areas for scales appropriate for mapping methodologies). This is important
species and communities. The use of more ecologically relevant to consistently classify and represent habitats in maps and for using
variables at a higher resolution will make models more sensitive to maps to monitor change in habitat condition, extent and spatial
differences in multiple ecological preferences among species. configuration over time. This will require compromises between
biological relevance per region and consistency across regions.
Definitions should be published, followed by strategic outreach
3b) Standardise the production and validation and communication to ensure they filter down to practitioners
of spatial models across their regions of application. Regional working groups may be
Best practice documents to improve standardisation must be required to establish how the levels and habitat classes are defined,
generated and applied to geospatial modelling techniques used to potentially facilitated through the Regional Sea Conventions, as
merge data and generate map data. This is particularly important already done by the Barcelona Convention for the Mediterranean.
as these models are complex statistical tools that need to meet a Furthermore, these regional working groups should seek to
series of robustness requirements throughout their development. establish quantitative definitions of habitats at the lower levels
We therefore recommend that: (i) best practice guidance is of the classification hierarchy to further improve consistency in
developed on the selection of modelling approaches to improve recording.
consistency; and (ii) best practice documents are generated for
implementing the recommended models e.g. providing advice on
sourcing suitable observations where species and habitats are not 3e) Improve the process of revision and further
present, i.e. absence data, which is informative for model training, development of HCSs
environmental predictors, model resolution, model parameters A simple online tool should be developed to allow scientists to
and assessing their predictive performance. The ICES Working submit proposals for revisions to a HCS. A mechanism is then
Group on MHM (WGMHM37) could be well placed to develop required to ensure that suggested revisions undergo an appropriate
such documentation, also involving the institutes part of the level of peer review, which could be coordinated in collaboration
EMODnet Benthic Habitats consortium. Better understanding and with user groups such as the Regional Sea Conventions. Custodians
communication of results and limitations of distribution models to of HCSs, including EUNIS, are encouraged to add additional
managers and policymakers is also a priority, and one of the most attributes to habitat descriptions such as sensitivity to human
important challenges in the field of MHM. pressures, conservation value, habitat condition, ecosystem service
provision and correspondence to habitats in other HCSs and lists,
using existing information and facilitated by tools such as the
3c) Better assess, communicate and standardise map Marine Evidence-based Sensitivity Assessment (MarESA).
accuracy and confidence
The consistent calculation and presentation of the accuracy estimates
associated with MHMs will facilitate a better understanding of their
value and use for specific tasks. It is recommended that standardised
accuracy assessments are produced and widely communicated
amongst the MHM community. Where possible, these standardised
accuracy assessments should include: (i) an overall (global) value
of map accuracy; (ii) accuracy information on specific classes or
subsets depicted on the map; and (iii) a spatially explicit depiction
of model performance, model agreement (i.e. when multiple
models are available within the same area) or map accuracy. It is
also recommended that this information be presented within a
standardised reporting template using consistent, well-referenced
and easy-to-understand terminology. Accessible guides are needed
for map users to interpret these map accuracy reports and establish
whether products are appropriate to use for the specific purpose
they require, as different end-users need different assessment
values and map products with different resolutions.

37
https://www.ices.dk/community/groups/pages/wgmhm.aspx

38
 N° 11 2024

4 What and where

to map
4.1 What has been mapped?
EU Directives and international legislation have generated a large number of EU programmes and funding
frameworks that have contributed to national and regional MHM e.g. BALANCE (Al-Hamdani & Reker,
2007), MESH38, MESHAtlantic39, iAtlantic40, ATLAS41, CoCoNet and BENTHIS42 (Andersen et al., 2018). For
most countries, inter-institutional cooperation at national level is very weak and they lack coordinated
national MHM programmes. Notable exceptions with ongoing national programmes seeking to map their
entire seabed are Norway (MAREANO43) and the Republic of Ireland (INFOMAR) (Table 4.1). These long-
term and systematic programmes are unparalleled in their ambition, and eventual rewards, in terms of
scientific achievement, economic return and efficiency of marine management. However, few countries
are prepared to commit the resources required to sustain ongoing national mapping programmes, with
punctuated, but valuable projects being favoured instead e.g. PNRR-MER44 (Italy), INTEMARS45 (Spain),
SedAWZ (Germany), REBENT46 (France), and Mapping of coastal and demersal marine habitats in the
Adriatic Sea under national jurisdiction47 (Croatia).

Table 4.1 Examples of national programmes with ongoing efforts to map their entire seabed.

NATIONAL MAPPING
COUNTRY OBJECTIVES
PROGRAMME
· To map bathymetry, sediment composition, contaminants, biological assemblages and
habitats in Norwegian waters.
· To provide data to assess the consequences of human activities.
MAREANO Norway
· To provide data to implement ecosystem-based management plans in different parts of the
Norwegian Exclusive Economic Zone (EEZ).
· To have full coverage of the Norwegian EEZ upon completion.

· To produce integrated mapping products covering the physical, chemical and biological
INFOMAR
features of the seabed.
A joint venture between · To provide comprehensive and freely accessible marine datasets for Irish waters via a
Geological Survey Ireland, dedicated web mapping portal.
Ireland
the Marine Institute and its
· To provide data to sustainably manage Ireland's marine resources.
predecessor, the Irish National
Seabed Survey (INSS) · To have full coverage (bathymetry and backscatter data) of the Irish designated shelf area
upon completion.

38
https://maritime-spatial-planning.ec.europa.eu/practices/mesh-survey-
scoping-tool 44
https://www.isprambiente.gov.it/en/projects/sea/pnrr-mer-marine-
39
https://keep.eu/projects/395/Mapping-Atlantic-Area-seabed--EN/ ecosystem-restoration
40
https://www.iatlantic.eu/ 45
https://intemares.es/en/
41
https://www.eu-atlas.org/ 46
https://rebent.ifremer.fr/
42
https://www.benthis.eu/en/benthis.htm 47
https://galijula.izor.hr/en/lansirana-je-jedinstvena-nacionalna-karta-morskih-
43
https://mareano.no/en stanista/

39
 EMB FUTURE SCIENCE BRIEF

In addition, several Regional Sea Conventions (e.g. HELCOM48, OSPAR49, systems58 and two regional classifications59 (Figure 3.7). EMODnet
Barcelona50 and Black Sea51) and the ICES Working Group on Marine has the most comprehensive collation of seabed habitats in Europe
Habitat Mapping were, and presently are, active in the coordination to-date and its products are used by many stakeholders, including
of MHM. At an international level, the Nippon Foundation52 and the national bodies and Regional Sea Conventions in quality status
General Bathymetric Chart of the Oceans (GEBCO53) came together reports and assessments for EU Directives e.g. the MSFD (see
in 2017 to identify how Ocean mapping might support SDG14. They Section 4.4).
launched the ambitious Seabed 203054 project to build the necessary
technical, scientific and management framework to compile all EMODnet Seabed Habitats have collated maps from individual
available seabed mapping information into a seamless digital map surveys across six regional seas in Europe: North-East Atlantic,
of the Global Ocean floor by 2030 (Mayer et al., 2018). Seabed 2030 Arctic, Baltic, North Sea, Mediterranean, and Black Sea (Table
are also partnering with private companies to map marine habitats55. 4.2). These maps vary in scale, biological detail, classification
There is however still improvement to be made in national, regional, system and modelling method used. Most are translated to
European and international coordination of mapping activities, the EUNIS classification (58%); one third are of marine habitats
including of public and private institutes carrying out geological, listed in Annex I of the Habitats Directive whose conservation
hydrographic, environmental and biological mapping. requires the designation of Special Areas of Conservation as part
of the Natura 2000 network (32%); and a small proportion adopt
In the past, MHMs were mostly published in the grey literature or other classification systems (10%). The North-East Atlantic (50%),
as technical reports by public research institutes. Over the past two Mediterranean (25%) and North Sea (18%) are the best mapped
decades, the main outlet has shifted to scientific peer reviewed regional seas in terms of numbers of available maps.
journals as the appreciation of, and level of sophistication in MHM
has evolved. To make these data available to a wide range of end- A large number (68%) of EMODnet Seabed Habitat maps collated
users, EMODnet Seabed Habitats56 collate and publish MHM of from surveys describe biology at a species or community level,
European waters on the EMODnet Portal57. Since 2009, almost 1,000 however it is important to note that this does not equate to
MHMs have been made publicly available as separate data layers spatial coverage. This suggests that there are considerable data
and as part of composite products that combine the information available on marine habitats. However, if we compare the extent
from the entire collection of maps in order to display the best of maps displaying substrate only with those showing a biological
estimate of the distribution of key habitats (see Figures 4.1 and 4.2 component (Figure 4.3), we can clearly see that biological habitat
for examples). Another key product of EMODnet Seabed Habitats maps generally cover small areas, mostly confined to coastal
is ‘EUSeaMap’, the predictive broad-scale seabed habitat map regions. This is because it is more difficult and time-consuming to
for Europe, which is available in three Europe-wide classification collect biological data in larger offshore areas.

Table 4.2 Overview of the number of maps from surveys by region and classification system that are available in the EMODnet Seabed Habitats portal60.

NUMBER OF MAPS BY SEA AREA

CLASSIFICATION
TOTAL
SYSTEM BALTIC BLACK MEDITERRANEAN NORTH
ARCTIC ATLANTIC
SEA SEA SEA SEA

EUNIS (v2007-11 and v2022) 2 350 0 7 114 86 559

Habitats Directive Annex 1 0 100 76 0 90 32 298

Other Classification Systems 5 16 2 3 28 46 100

48
https://helcom.fi/
49
https://www.ospar.org/convention
50
https://www.unep.org/unepmap/who-we-are/barcelona-convention-and-protocols 56
https://emodnet.ec.europa.eu/en/seabed-habitats
51
http://www.blacksea-commission.org/_convention.asp 57
emodnet.ec.europa.eu
52
https://www.nippon-foundation.or.jp/en 58
EUNIS habitat classification v2007-11, EUNIS habitat classification v2022 and
53
https://www.gebco.net/ MSFD Benthic Broad Habitat Types
54
https://seabed2030.org/ 59
HELCOM Underwater Biotopes in the Baltic and Barcelona Convention habitat
55
https://seabed2030.org/2024/04/11/seabed-2030-announces-new- types in the Mediterranean
partnership-with-ocean-ledger-in-boost-to-coastal-mapping-and-ecosystems/ 60
https://emodnet.ec.europa.eu/geoviewer/

40
 N° 11 2024

Credit: EMODnet Seabed Habitats.

Figure 4.1 Examples of maps in EMODnet Seabed habitats. (A) OSPAR threatened and/or declining habitats in the north-east Atlantic61;
(B) HELCOM underwater biotopes in the Baltic Sea62; (C) coralligenous and other calcareous bioconcretions in the Mediterranean63 (see Martin et al.,
2014 and Ingrosso et al., 2018 for updated data on calcareous bioconcretions not included in this map).

61
https://bit.ly/emodnet-ospar-t-and-d-habitats
62
https://bit.ly/emodnet-euseamap-hub 63
https://bit.ly/emodnet-coralligenous

41
 EMB FUTURE SCIENCE BRIEF

Credit: EMODnet Seabed Habitats 2023.

Figure 4.2 Maps showing the distribution of Essential Ocean Variables (EOVs) determined from polygon and point observations. Top: live hard coral
cover; middle: seagrass cover; bottom: macroalgal canopy cover (i.e. kelp forests). These maps were created by interrogating and combining the marine
habitat maps in EMODnet Seabed Habitats into new, composite data products.

42
 N° 11 2024

Figure 4.3 Areal extent of survey maps describing substrate only (green areas) and habitat classes with biological information (yellow areas), in
Regional Sea Convention areas. Note that areas labelled as predicted substrate (purple) include maps at a range of scales, including some very
coarse-scale substrate maps based on interpolation of sparse ground truthing samples and no MBES data (such as much of the deep waters of the
UK) and some depth-only mapping from MBES surveys (e.g. Porcupine Abyssal Plain off the coast of Ireland). Data sourced from EMODnet Seabed
Habitats64. Note that some data may be missing from this map (e.g. habitat map of Croatian EEZ) due to new data being published after the data in
the map were collated. Future iterations will include data published from 2023 onwards.

Maps showing only substrate-level data have greater coverage. leaving large parts of some sea basins and Exclusive Economic Zones
Table 4.3 provides an overview of the extent of Regional Sea (EEZs) unmapped (or at least with unpublished maps). Offshore
Convention areas mapped for substrate and biological information. and deeper areas that have not been surveyed rely on modelling to
In the North Sea (OSPAR region 2) and the Celtic Sea (OSPAR region 3) predict habitats.
over 40% of the areas have detailed substrate maps. However,
maps displaying data on biology account for only 10% and 5%, The regional differences are driven by a remarkable heterogeneity
respectively. The Mediterranean Sea (especially large areas off North among EU countries in the compliance with targets, Directives
Africa) and Black Sea maps have very low biological coverage (0.5%), and private uses of maps. A good example is the designation of
however, not all survey data have been collated from all bordering Natura 2000 sites and nationally designated MPAs, which have
countries, so these figures may improve once all data has been different management plans, monitoring approaches, reporting
collated. The Norwegian Sea and parts of the Barents Sea (Arctic, and threat assessments across Europe (Mazaris et al., 2019).
OSPAR region 1) have the largest area of seabed mapped with Another key driver for these differences is the funding of national
detailed biological information (>200,000km2). Countries that have seabed mapping programmes. For example, the OSPAR area has the
dedicated biological sampling programmes or large repositories of highest percentage map coverage in Europe due to the two state-
biological sample data, such as Norway and Germany, generally funded national mapping programmes: MAREANO in Norway and
have the most detailed MHMs. In general, there are more substrate INFOMAR, and its predecessor, the Irish National Seabed Survey
and habitat maps available in coastal areas than in offshore areas, (INSS) in Ireland (see Table 4.1).

64
https://emodnet.ec.europa.eu/geoviewer/?layers=12493:1:1,9985:1:1,12615:1:1,13017:1:1,12702:1:1,12616:1:1,12701:1:1,12618:1:1&basemap=esri-gray&ac
tive=12493&bounds=-14594115.96520002,3451984.993129384,13599742.612924984,16746172.475453604&filters=&projection=EPSG:3857

43
 EMB FUTURE SCIENCE BRIEF

Table 4.3 Percentages of Regional Sea Convention areas that have predicted seafloor substrate, mapped substrate and mapped habitat coverage.
Predicted substrate includes maps at a range of scales, including some very broad-scale substrate maps based on interpolation of sparse ground truthing
samples and no acoustic data. Data from EMODnet Seabed Habitats65.

PERCENTAGE OF REGIONAL SEA CONVENTION AREA MAPPED

MAP TYPE NORTH BISCAY/ WIDER
ARCTIC CELTIC SEA BALTIC BLACK MEDITERRA-
SEA IBERIA ATLANTIC
(OSPAR I) (OSPAR III) SEA SEA NEAN SEA
(OSPAR II) (OSPAR IV) (OSPAR V)

Predicted substrate 32 100 100 51 35 100 100 100

Mapped substrate 6 52 43 19 11 48 14 11

Mapped habitats
(including biological 4 10 5 4 0.3 10 0.5 0.5
information)

4.2 What are the gaps in marine habitat mapping?

There are several gaps in mapping of key marine benthic habitats Italian coasts and approximately 80 seamounts to support the
in European sea basins (even in the Norwegian and Barents Seas, creation of a network of Natura 2000 deep sea sites as part of the
which have large areas of their seabed mapped to a detailed PNRR-MER project. Several studies and projects (e.g. Interreg Med
biological level) and international waters of the North-East AMAre68) show that even in MPAs (both fully protected, where
Atlantic. These gaps include spatial coverage (especially in the all extractive uses are forbidden, and partially protected, where
deep sea), the consistency and resolution of coverage for both some extractive uses such as fishing are permitted), knowledge
common habitats and those of conservation importance, and about biodiversity distribution and status is often incomplete and
habitat condition, as concluded by EMODnet Seabed Habitats and should be updated.
others (e.g. Matear et al., 2023).
Despite the EU MSFD obligation to reach GES for seabed habitats,
In general, the North-East Atlantic and the Baltic Sea appear to the level of habitat degradation and loss is often not included
have better mapping coverage than the Mediterranean and Black in map records, and a low percentage of MHMs are from within
Seas. Most specific habitats (e.g. seagrass meadows, coralligenous MPAs (Gerovasileiou et al., 2019; see Figure 4.4 for two examples).
formations, maërl beds, macroalgal forests, coral gardens, sponge In addition, spatial information on degraded habitats is needed
aggregations, seamounts, submarine canyons, mud volcanoes to plan where restoration efforts are most needed. High-quality
and hydrothermal vents) need more mapping efforts. Specific data (i.e. with a high level of spatial resolution and classification
initiatives have been carried out on the distribution of Essential accuracy) on the distribution of habitats that need to be restored
Fish Habitats (i.e. areas or volumes of waters or bottom substrates and the distribution of human pressures are important to
where fish spawn, breed, feed and grow) to minimise adverse demonstrate the feasibility of restoration actions, to inform
effects from fishing activities. Some examples exist in Scotland66, prioritisation and to guide the allocation of the restoration
in the Baltic Sea (HELCOM, 2021) and in the Mediterranean Sea targets (Fabbrizzi et al., 2023) included in the proposed EU Nature
(Farrag, 2022), where these critical habitats have already been Restoration Law. Planning large-scale restoration interventions in
mapped. the absence of high-resolution information can compromise their
efficacy.
There is a substantial lack of maps providing complete and up-to-
date spatial distribution of many habitats in Natura 2000 sites. Finally, deep-sea ecosystems (i.e. below 200m depth) are the last
However, in Greece, where Natura 2000 sites were established large unknowns on our planet (Amon et al., 2022). However, many
mainly for protecting Posidonia seagrass meadows (Giakoumi resources crucial to society originate from these remote parts of
et al., 2013), recently produced seagrass coverage maps have the Ocean, such as oil and gas. In addition, deep-sea mining could
quantified for the first time their extent and spatial distribution become a new activity in the deepest parts of our Ocean. More
(Panayotidis et al., 2022). In Italy, 400 million euro allocated biological mapping is therefore needed in the deep sea to increase
from the European Green Deal will go to the Italian Institute knowledge and guide management actions. For more information
for Environmental Protection and Research (ISPRA67) to map the on knowledge gaps and research priorities in the deep sea, see the
distribution and condition of all species of seagrasses along all EMB Working Group on Deep Sea and Ocean Health69.

65
https://emodnet.ec.europa.eu/geoviewer/
66
https://www.gov.scot/publications/developing-essential-fish-habitat-maps-fish-shellfish-species-scotland-report/pages/4/
67
https://www.isprambiente.gov.it/en
68
https://maritime-spatial-planning.ec.europa.eu/projects/amare-actions-marine-protected-areas
69
https://www.marineboard.eu/deep-sea-and-ocean-health

44
 N° 11 2024

Credit: Mariacristina Prampolini & Simonetta Fraschetti (AMAre project).

Figure 4.4 Marine habitat maps are essential for management, especially within MPAs. All MPAs and Natura 2000 sites should have up-to-date,
fine-scale maps. Two examples are provided above of maps from Torre Guaceto and Porto Cesareo MPAs (Italy).

4.3 Where to map first? The need for spatial prioritisation

Spatial prioritisation is required to guide the selection of priority widely used open-source spatial prioritisation tools (Fabbrizzi et
areas for MHM. This is currently done on a use-case specific basis. al., 2023), and others, are particularly effective in spatial planning
Spatial prioritisation is particularly important due to the high cost thanks to Geographic Information System (GIS) based input of
of MHM activities, with the final aim to map the entire seabed. information and flexible interface that includes the costs of needed
This should be considered in stakeholder discussions to mitigate interventions. These may also be useful to prioritise areas for future
preconceived ideas on which particular places should, or should not, restoration and MHM activities.
be prioritised for MHM efforts.
Regardless of the software used, fine-scale MHM (i.e. with high
Chapter 1 describes how MHM is fundamental for the successful spatial resolution and classification accuracy) is a prerequisite
application of EU policies and Directives e.g. to meet the ambitious for the solid background needed to design, site and start active
targets of the MSFD to achieve GES of all BBHT, and to meet the or passive restoration interventions. Gaps in MHM have been
objectives of the Maritime Spatial Planning Directive and the identified in Section 4.2 and protected areas (MPAs and Natura
proposed EU Nature Restoration Law, among others. The proposed 2000 sites) are a priority starting point for MHM. Knowledge
EU Nature Restoration Law sets ambitious quantitative targets derived from MHM is particularly important to assess the criteria for
for areas and habitats to restore, with targets for 2030, 2040 and further implementing coherent networks of MPAs. These include
2050. However, while criteria for the allocation of conservation criteria such as MPA connectivity and representativity (i.e. the need
targets have been identified (Zhao et al., 2020), deciding how to for MPAs to represent, or sample, the full variety of biodiversity,
allocate restoration targets still requires considerable clarification ideally at all levels of organisation). Priority should also be given
(Fabbrizzi et al., 2023). The use of decision-support tools based on to ecologically significant spatial units, such as Cells of Ecosystem
scientific knowledge and considering socio-economic constraints Functioning (Boero et al., 2019) and hot spots of ecosystem
can support this process. These are currently being used in marine functioning (e.g. canyons, gyres, upwelling fronts). These require the
spatial planning and systematic conservation planning, i.e. a multi- definition and mapping of the significant ecological connections
component, stage-wise approach to identifying conservation areas that define marine ecosystems, which in turn requires a holistic
and devising management policies, with feedback, revision and approach to MHM, including benthic and pelagic components, their
reiteration, where needed, at any stage (Kukkala & Moilanen, 2013). hydrological and functional connections, and a holistic approach
The aim is to allocate human use and identify networks of MPAs to their management. This process would support moving MHM
at minimum cost. Software such as Marxan70, one of the most towards three-dimensional volumes rather than areas (e.g. marine

70
https://marxansolutions.org/about-marxan/

45
 EMB FUTURE SCIENCE BRIEF

protected volumes). Areas known to be affected by the most macroalgal forests (e.g. Verdura et al., 2023), shellfish beds (e.g.
extensive damage to the seabed (e.g. from bottom fishing) are also Pouvreau et al., 2021), maërl beds (e.g. Illa-López et al., 2023), sponge,
priorities. Sites for the installation of Blue Economy activities, such coral and coralligenous beds (Ingrosso et al., 2018), and vents and
as renewable energy and aquaculture, must be properly mapped seeps (Taviani, 2014). Suitable unprocessed data can also be collated
and managed to avoid significant repercussions at ecosystem level. into MHMs and collated full coverage maps should be transformed
Areas of the deep sea of interest for mining are also priority areas to into a common format and typology (e.g. EUNIS). More resources
be mapped in order to inform decision-making. Guidelines for the should be made available to support the collection of data that
prioritisation of mapping activities at an international level could would allow the condition of habitats to be captured within maps.
help aid in the selection of priority areas, including the consideration An example is the consultancy contract by the Specially Protected
of both active and passive restoration criteria (Fabbrizzi et al., 2023). Areas Regional Activity Centre (SPA/RAC) to produce updated,
standardised maps for three Mediterranean habitats (coralligenous
In parallel, collation of already existing information should be assemblages, Posidonia meadows and marine caves) within the
supported in order to assess what has already been mapped. Efforts regional project “Empowering the legacy: Scaling up co-managed
already exist in terms of seagrass beds (e.g. Traganos et al., 2022), and financially sustainable no-take zones/MPAs71”.

4.4 Who uses marine habitat maps and for what purpose?
The closest proxy for the distribution of users of MHMs in Seabed Habitats maps are mainly used for the following
Europe is the download statistics published by EMODnet applications: academic investigations (29%), implementation
Seabed Habitats72. The latest published report (EMODnet of the MSFD (16%), marine spatial planning (13%), studies for
Seabed Habitats, 2022) states the following users: researchers marine biodiversity conservation purposes (12%), research related
and academics (65%), private sector (16%), government/public to MPAs (11%), and baseline studies for implementing coastal
administration (11%), non-governmental organisations (5%) management including: environmental impact assessments
and other (4%)73 (Figure 4.5). The most downloaded product in (9%); Blue Economy private sectors (3%); and marine ecosystem
EMODnet Seabed Habitats is the EUSeaMap (>80%). EMODnet service assessments (3%) (Figure 4.5).

5
4 3 3
9
11 29

11

14

65
12
16
13

Researchers and academics Academic investigations

Private sector Implementation of the MSFD
Government/public administration Marine spatial planning
Non-government organisations Studies for marine biodiversity conservation
Other Research related to MPAs
Environmental impact assessments
Blue Economy private sectors
Marine Ecosystem service assessments

Figure 4.5 Users (left) and applications (right) for EMODnet Seabed Habitats maps. Values indicate percentages.

71
https://www.rac-spa.org/node/2023
72
https://emodnet.ec.europa.eu/en/seabed-habitats
73
Note that it was not mandatory for users to supply an organisation type, so these figures are the proportions of the 79% of downloads for which an organisation type
was supplied.

46
 N° 11 2024

4.5 Bespoke, fit-for-purpose marine habitat maps

It is difficult to produce a ‘one size fits all’ map. Nuances in available MHMs from surveys that were compiled into a single
individual maps cannot always be included when carrying out layer, with gaps filled using EUSeaMap. Data on fishing activity was
large-scale analyses, especially without good understanding of then overlaid to produce an estimate of the areas at highest risk
the input datasets. An example of the need for bespoke maps is to of physical disturbance. Subsequently, it was translated into MSFD
evaluate the extent and distribution of Sabellaria spinulosa worm BBHTs and provided to EU Member States in the OSPAR region as
reefs (protected under the Habitats Directive Annex I and OSPAR) an optional base layer to use for their six-yearly MSFD reporting.
in the Southern North Sea. It may be tempting to assume that a The key benefit of EUSeaMap, and other centrally-built composite
single authoritative data product could be used for a variety of data products, is that they are produced using consistent, and often
purposes. However, different mapping approaches may be required repeatable, procedures across Member State boundaries. However,
depending on the intended objective, particularly as S. spinulosa Member States often have access to additional data and expert
reefs are notorious for their short-lived nature. To understand local knowledge that can result in more accurate end products.
their natural range and identify areas for protection or restoration,
a compilation of all historic records and maps of the habitat may It would be beneficial for EMODnet Seabed Habitats to work
be the best approach. However, to inform management measures towards producing and promoting best practice guidelines and
within MPAs, knowledge on the current extent of their habitat is tools to support Member States in producing their own bespoke
best suited. Each of these use-cases would require a different map maps and composite data products using MHMs. This would allow
and a different estimate of the habitat’s extent, all of which are Member States to use their own local expert knowledge of habitats
equally valid. and additional available data, and to build maps at the scale, extent
and level of confidence that they require for different purposes. The
A further example of bespoke map production is the 2023 OSPAR creation of new maps could be automated on-demand, as much of
Quality Status Report, where the EMODnet Seabed Habitats the data used to create MHMs is stored centrally, and the models
consortium produced a bespoke map for the ‘Extent of Physical used to combine this information to produce new maps, typically
Damage’ indicator (Matear et al., 2023), which used the best use the same techniques.

Credit: © OCEANA.

Bespoke marine habitat maps are useful to evaluate the extent and distribution of Sabellaria spinulosa reefs, particularly given their short-lived nature.

47
 EMB FUTURE SCIENCE BRIEF

4.6 Recommendations
We recommend scientists/map producers and research funders to: 4d) Capture habitat condition in MHMs
Given the necessity of understanding condition (i.e. health), habitat
4a) Increase the spatial extent and resolution of mappers should collect and collate information (i.e. metrics or
biological information in marine habitat maps indicators) on habitat degradation and incorporate aspects of
Future MHM efforts should increase the resolution of biological habitat condition within maps. This should be supported by the
information so that maps progress from predominantly presenting quantitative definition of condition levels within mapped classes
coarse environmental and substrate information, or broad-scale (recommendation 3d) and by enabling revision and development
habitats, to representing greater amounts of biological information within habitat classification schemes (recommendation 3e).
(e.g. EUNIS classes 4 – 5, and 6 for the Atlantic). This will require the
use of map standards for data and modelling (see recommendations
2f and 3b), technological development of platforms and sensors 4e) Co-develop guidelines for the prioritisation of MHM
(recommendations 2d and 2e) and better resourced MHM activities at an international level
programmes. Map producers need to be tasked and resourced to This is needed to help in the selection of priority areas of where
tackle this critical gap, as habitat and species distribution maps are to map first given the various priorities for achieving GES for
the most compatible with management and policy requirements. degraded MSFD habitats, active and passive restoration, and
maritime spatial planning for Blue Economy activities. Guidelines
need to be co-developed in collaboration with a wide range of
4b) Develop specifications for specific types of MHMs stakeholders. Decision-support tools should be employed to assist
Map specifications should be developed for each map type intended in the prioritisation of MHM activities, which should be taken into
for various purposes e.g. habitat inventories, monitoring, advice for account in stakeholder discussions.
designations, ecosystem service assessments, ecological coherence
assessments. These specifications should clearly state the required
spatial resolution and extent (scale), habitat types and resolution In addition, we recommend policymakers to:
of biological information, as well as accuracy and reporting format.
These need to be co-developed by scientists (i.e. advising on what 4f) Strengthen national, regional, European and
is possible) and end-users (i.e. to determine what is needed and fit- international strategic coordination mechanisms for
for-purpose). interdisciplinary mapping efforts and resourcing
There is a need for enhanced cooperation on the delivery
of coordinated mapping programmes. Improving strategic
4c) Produce bespoke MHMs that are fit-for-purpose to coordination will help mapping efforts to adhere to defined
better answer stakeholders’ needs standards (recommendation 2f), the submission of data into
Different mapping purposes often require different formats and national and European data centres and services (recommendation
mapped classes (i.e. potentially using different HCSs or attribution 5a) and the prioritisation of mapping efforts (recommendation 4e).
other than habitat ‘identity’ e.g. ecosystem services), even while
using the same underlying data. The creation of new, bespoke maps
that are fit-for-purpose could be automated on-demand, using 4g) Increase and improve map coverage of habitat types
criteria set by the map user and hosted on existing map platforms, and spatial extent through national mapping
such as EMODnet, using the most up-to-date information available. programmes
A best practice methodology on how to create such bespoke maps Task national bodies (e.g. ISPRA or the UK Centre for Seabed
using the EMODnet Seabed Habitats resource could be provided as Mapping74), which coordinate the collection, management and
part of future project deliverables. access of seabed mapping data, and regional (e.g. ICES) and European
(e.g. EMODnet) bodies and initiatives, with maintaining oversight of
MHM coverage and gap analyses for important habitats or areas.
They should also update the MHM community on priority gaps
annually, and should be linked to national mapping programmes
that can coordinate and commission priority gap-filling surveys.

74
https://www.admiralty.co.uk/uk-centre-for-seabed-mapping

48
 N° 11 2024

5
5.1
Communication and
dissemination
Data dissemination: increasing the value of each map
The typical format for most MHMs remains digital, to be used in specialist software packages. Efforts
to collate and host these files (e.g. by EMODnet Seabed Habitats) have greatly improved their collective
extent, visibility and overall value. However, beyond online browsing and the ability to download map
images, downloadable files remain in specialist software formats and are generally inaccessible to
the public. In addition, most available maps are images of maps and not accessible as open-access
georeferenced data i.e. GIS. This should be changed. It is very expensive to produce MHMs and their
value can be greatly increased by making them and the data on which they are based more easily
accessible to a wider range of stakeholders, so they can be used in many different applications.

The extent to which a dataset is made accessible can be measured An effective way to close these gaps would be to facilitate
against the widely used FAIR data principles for scientific data and incentivise map producers to publish MHMs in common
management and stewardship (Wilkinson et al., 2016), which state repositories. The first step, however, is for map producers to know
that a dataset should be: where the repositories are and what the process is.

• Findable – easy to find for both humans and machines;

5.1.1 Current situation
• Accessible – once found, it should be clear how to For some types of marine data such as geological/biological surveys
access it; and hydrographic data, national data centres are the authoritative
data sources. They aggregate data on a suite of themes at a national
• Interoperable – able to be combined with other datasets
level and set data standards. Data from these centres are then
and/or integrated into a workflow for analysis, storage or
ingested by EU data infrastructures (e.g SeaDataNet77, EurOBIS78),
processing; and and subsequently the data is aggregated and standardised at the
European level by EMODnet’s thematic groups (e.g. EMODnet
• Reusable – requiring well-described metadata about
Physics, Geology and Bathymetry).
where the data came from.
Unfortunately, it is uncommon for European countries to have
The ‘collect once and use many times’ philosophy of EMODnet national data centres dedicated to MHMs. Without the national
benefits all marine data users, including policymakers, scientists, data centres aggregating national data, EMODnet Seabed Habitats
private industry and the public. It has been estimated that an (unlike most of the other EMODnet thematic groups) must first
integrated marine data policy will save offshore operators at least undertake this role before it can aggregate the data at European
one billion euro per year, and will open new opportunities for scale. This leads to bias in the MHMs that are aggregated, with
innovation and growth75. However, there is still a big gap between the majority of MHMs coming from countries who are, or have
maps that are published in scientific papers and those available previously been, partners in the EMODnet Seabed Habitats
in portals. The ICES workshop76 on the use of predictive habitat consortium, with data gaps elsewhere. Improving the process of
models in ICES advice found that MHMs are not yet used much in getting data into EMODnet is of great importance considering the
management compared to the number of available maps due to deadlines of the proposed EU Nature Restoration Law, and the need
the slow uptake of the open data concept, issues of map confidence to estimate the state of play with mapping of the Habitats Directive
and knowing which map to use, lack of data and maps submitted Annex I habitats and MSFD BBHT, and assessment of their condition
to EMODnet, data management issues (e.g. formatting maps), and across EU seas.
complex mechanisms to submit data (ICES, 2021).

75
https://emodnet.ec.europa.eu/sites/emodnet.ec.europa.eu/files/public/Brochure/EMODnet_brochure_updated_11-Jan-18_Vweb.pdf
76
https://www.ices.dk/community/groups/Archive%20for%20Community%20pages/WKPHM.aspx
77
https://www.seadatanet.org/
78
https://www.eurobis.org/

49
 EMB FUTURE SCIENCE BRIEF

5.1.2 Ensuring maps produced for policy and

services e.g. EMODnet, so that data are more accessible to the
research are submitted to EMODnet public and policy communities.
There are repositories and portals that bring together mapping
data on several marine themes for specific purposes, such
5.1.3 Industry data
as monitoring, marine spatial planning or conservation (e.g.
ISPRA79 portal and the National Biodiversity Future Centre80 for Data acquired by industry provides an under-utilised source
Italy, CoCoNet spatial geoportal81, Adriplan82), which have been of information. To incorporate this data into MHM-based
produced by and for government agencies and departments. decision processes, it needs to adhere to the same standards as
Map producers may perceive no added benefit in supplying environmental assessment and research data and needs to be
data to EMODnet, despite EMODnet offering users data and published in comparable FAIR archives e.g. in the Ocean Biodiversity
web services that are interoperable with global Ocean data Information System (OBIS85) or EMODnet. Quality Assurance/Quality
initiatives. In addition, the custodians of national and regional Control plans and procedures need to be implemented to raise
repositories do not always take responsibility for passing the information content to the level required for informative and
the data on further to ensure wider use. As a result, there robust MHM and for subsequent publication. To facilitate access to
are a number of different map producers and, for each map, industry data, national licensing bodies for Blue Economy activities
different end-points e.g. marine planning portals, MPA portals, should require data from site investigations and monitoring to be
aggregated datasets for environmental status reporting such submitted to central data repositories as a licensing condition, and
as for Regional Sea Conventions. What is missing is a pipeline of agreements for mutually beneficial partnerships and data exchange
named organisations and repositories who are responsible for formats should be activated. These data should come into national
aggregation, standardisation and publication of national or sub- data centres first and then to EMODnet. Data need to be provided
regional MHMs, that can feed MHMs into EMODnet. to specified standards and governments should oblige industry
to provide data from environmental impact assessments and
Increasingly, researchers are encouraged to publish their data and subsequent environmental monitoring in formats compatible with
results along with any written publications. PANGAEA83 provides a national databases. This might entail reduced resolution datasets
free repository for researchers to publish spatial datasets related to for commercially sensitive areas or providing public data in formats
Earth and environmental science and Zenodo84 is a multidisciplinary, that are compatible with industry software, e.g. bathymetric
open repository where data can be submitted. However, more must datasets that can be imported into charter plotters used by the
be done to connect repositories like PANGAEA and EC marine data fishing community (e.g. OLEX86 and TimeZero87).

5.2 Using marine habitat maps to improve public understanding of the Ocean
Although the Ocean covers more than 70% of the Earth's surface, and when developing accessible mapping products (e.g. atlases,
more than 90% of its volume and supports an estimated 90% of apps, posters and digital products) that promote public interest
the life forms on our planet, marine habitats, and the species they and knowledge of the Ocean. An example is the DONIA app89 that
support, remain largely inaccessible to humanity. Maps, as visual targets boaters who use MHMs to avoid anchoring on sensitive
tools, provide important foundational information on marine seabed habitats. Ireland’s national seabed mapping programme,
habitats in an intuitive and recognisable format that facilitate INFOMAR, also produces story maps for bays of interest around the
several of the principle messages of Ocean literacy88, namely: Irish coastline. These story maps document the natural and cultural
(i) the Earth has one big Ocean with many features; (ii) the Ocean heritage found in the area and link to the importance of managing
supports a great diversity of life and ecosystems; (iii) the Ocean and monitoring these features through detailed seabed maps.
and humans are inextricably interconnected; and (iv) the Ocean INFOMAR has also produced high-resolution bathymetric maps of
is largely unexplored. An increase in Ocean literacy will stimulate Ireland’s coastal waters (Blue Scale Map Series90) and the map of its
continued interest for new or updated mapping products. MHM offshore territory, “The Real Map of Ireland”, is used as a teaching
will undoubtedly play a major role in Ocean Literacy initiatives that resource in primary schools91 (Figure 5.1).
bring the challenges the Ocean faces to the attention of society

79
http://www.db-strategiamarina.isprambiente.it/app/#/
80
http://gismargrey.bo.ismar.cnr.it:8080/mokaApp/apps/pnrrb/index.html 86
https://olex.no/products/olex_software_en.html
81
http://coconetgis.ismar.cnr.it/ 87
https://mytimezero.com/
82
http://adriplan.eu/ 88
https://oceanliteracy.unesco.org/principles/
83
https://www.pangaea.de/ 89
https://donia.fr/en/home/
84
https://zenodo.org/ 90
https://www.infomar.ie/galleries/node/565
85
https://obis.org/ 91
https://www.scoilnet.ie/go-to-post-primary/geography/infomar/landscape/

50
 N° 11 2024

Credit: INFOMAR.
Figure 5.1 Outreach products from INFOMAR used as teaching resources in Irish primary schools: “The Real Map of Ireland” (left) displays
Ireland’s offshore bathymetry in shaded relief to highlight geomorphological features and the blue scale bathymetric map of Dublin Bay (right)
highlights complex sandbanks in the Irish Sea.

Rapidly evolving robotics, information and communication using specially designed interfaces and the robots become the
technologies hold considerable potential to help unravel the extended “arms” of the spectators, who can collectively select what
mysteries of the Ocean and allow citizens virtual access to otherwise type of data/information they wish to have access to or visualise.
inaccessible underwater regions. Google Streetview Underwater92 Explorers can play a double role of a general mission planner with
is a good example of how citizens can view and “explore” selected a say on what data to acquire and where, and a mission visualiser/
areas of the underwater world. Considerable effort is being analyser having access to selected data acquired in almost real-
made worldwide to virtually explore the Ocean, also known as time (e.g. temperature, salinity, turbidity as function of depth)
‘telepresence’ or Virtual Ocean Exploration Local Area Networks and/or underwater and seabed images (e.g. photographs and
(Figure 5.2), which will unleash a paradigm shift in the access that acoustic-based mosaics) during and after a mission has taken place.
citizens have of the Ocean. The concept involves virtual explorers In practice, the USA’s National Oceanographic and Atmospheric
within the comfort of a museum or science centre auditorium being Administration (NOAA) Ocean Exploration programme has made
‘transported’ to a given site via a satellite-based or underwater pioneering contributions in this area since the early 2000s. For
cable communications link to a support ship. The ship acts as a example, NOAA research vessel Okeanos Explorer uses satellite
command unit from which robotic-based vehicles are launched technology to transmit data and video in real time from the ship to
to explore the underwater environment, which maintain contact a shore-based hub and then to other sites via the internet (Figure
with the ship via acoustic links. The movement and data gathering 5.2) and NOAA Microbial Stowaways93 expedition aboard Point Sur
actions of the robots can be programmed by a virtual explorer allowed students to telecommunicate from shore.

92
https://www.underwater.earth/
93
https://oceanexplorer.noaa.gov/explorations/19microbial-stowaways/background/plan/plan.html

51
 EMB FUTURE SCIENCE BRIEF

Command and
Support ship data link
(interface unit) (satellite,

Credit: Antonio Pascoal & Pere Ridao (left), NOAA Ocean Exploration (right).
radio, etc)

Surface-Underwater
communications (comms)

Virtual explorer/
Cabled marine assets link
Acoustic comms via
comms umbilical

AUV AUV ROV

Figure 5.2 Left: Conceptual diagram of a Virtual Ocean Exploration Local Area Network (LAN; i.e. group of computers and linked devices) or
‘telepresence’ system, enabling virtual Ocean exploration by bringing together robotics, sensing and communications. Right: NOAA Ocean
Exploration uses satellite technology to deliver data from sensors on their ship Okeanos Explorer back to shore94.

An example of the coupling between marine robotic systems and Citizen science is an approach which involves members of the public
Internet of Things (IoT95) as an affordable tool for Ocean Literacy is in gathering scientific data and is a way to enhance awareness of
the EU project Blue Robotics for Sustainable Eco-friendly Services the marine environment. An example is the National Biodiversity
for innovative marinas and leisure boats (Blue RoSES96), where an Data Centre97 in Ireland that hosts a mapping and data portal, and
affordable system was developed to allow non-scientific users runs a series of citizen science projects which allow individuals
access to the underwater world following a scientific mission as it to submit their records. Other examples are the Hidden Deserts
unfolds (see Figure 5.3). Citizens were able to “connect” remotely to a initiative98, which includes citizen science data in mapping shallow
station on board a small support ship to which an ROV was tethered underwater habitats, and Reef Check Mediterranean initiatives
and could issue high level instructions using a specialised application that provide citizen science data of more than 40 species collected
and access images. Data were transferred between the support by trained snorkelers, free divers and SCUBA divers (Turicchia et al.,
ship and the shore station(s) using a standard telecommunication 2021a; see Figure 5.4). Crowd sourced image annotations is another
network (4G/LTE) enabling the rapid transmission of images. The example of the contribution of citizen science to MHM. However,
ROV was able to manoeuvre in response to high level commands Assurance/Quality Control plans and procedures are required to
while keeping itself within safe vicinity of the support ship, thus raise citizen science data to the level required for informative and
removing the need for an expensive dynamic positioning system. robust MHM or subsequent publication, as is done in OBIS and the
Ultimately, it is hoped that this will become a two-way system that Reef Check Mediterranean Underwater Coastal Environmental
will allow the control or programming of underwater assets from Monitoring Protocol (Turicchia et al., 2021b).
remote locations (i.e. on shore rather than ship-based control).

94
https://oceanexplorer.noaa.gov/technology/telepresence/telepresence.html#
95
The internet of things is a world-wide network of smart interconnected objects with a digital entity
96
https://bluerosesproject.wixsite.com/home/about
97
https://biodiversityireland.ie/
98
https://hiddendeserts.com/

52
 N° 11 2024

Credit: Blue RoSES project.

Figure 5.3 Ocean Literacy and virtual underwater exploration. Left: the ROV deployed in Portugal, near a shipwreck. Right: operator in Genova
controlling the ROV while observing the shipwrecks' cannon.

RC observations REEF CHECK ITALY

Mediterranean Underwater Coastal
Environmental Monitoring protocol

2,160 6,432
volunteers monitorings

MedSense
index

62,000 42
orbservations species
Credit: Carlo Cerrano.

FigureValidated records
5.4 Synthesis since 2008
of the activities (https://zenodo.org/records/6330628)
carried out in Reef Check (RC) Mediterranean to map shallow underwater habitats.

Published materials:
• Cerrano, C., Milanese, M., & Ponti, M. (2017). Diving for science-science for diving: volunteer scuba divers support science and conservation
in the Mediterranean Sea. Aquatic Conservation: Marine and Freshwater Ecosystems, 27(2), 303-323. https://doi.org/10.1002/aqc.2663
• Turicchia, E., Ponti, M., Rossi, G., Milanese, M., Di Camillo, C. G., & Cerrano, C. (2021). The reef check Mediterranean underwater coastal
environment monitoring protocol. Frontiers in Marine Science, 8, 620368. https://doi.org/10.3389/fmars.2021.620368
• Turicchia, E., Cerrano, C., Ghetta, M., Abbiati, M., & Ponti, M. (2021). MedSens index: The bridge between marine citizen science and coastal
management. Ecological Indicators, 122, 107296. https://doi.org/10.1016/j.ecolind.2020.107296

53
 EMB FUTURE SCIENCE BRIEF

5.3 Recommendations
We recommend research funders to:

5a) Facilitate and incentivise map producers to publish acquired for environmental assessments and research purposes
their maps according to the FAIR principles and (see recommendation 2f) and needs to be published in comparable
submit data to EMODnet FAIR archives (see recommendation 5a). Quality Assurance/Quality
MHM data should not solely be published in scientific journals or Control plans and procedures need to be implemented. To facilitate
reports as PDFs, but also as FAIR data layers that allow reuse for access to industrial data, mutually beneficial partnerships or
further studies and decision-support processes. This requires open exchange formats need to be established and national licensing
and interoperable data repositories and archives that are sustainably bodies for Blue Economy activities should require data from site
maintained in the long-term. No specific repositories are needed for investigations and monitoring to be submitted to central data
MHM data as existing solutions like PANGAEA and Zenodo can be repositories as a licensing condition.
used. Submitted data need to adhere to FAIR standards. In addition,
map producers should be incentivised and/or obliged by funding
bodies to submit data and maps to EMODnet. More data repositories In addition, we recommend scientists/map producers, with the
should be linked to EMODnet and a pipeline should be established of support of research funders to:
named organisations (e.g. national data centres) who are responsible
for aggregation, standardisation and publication of national or sub- 5c) Develop accessible mapping products for Ocean
regional MHMs, that can feed into EMODnet. literacy and support citizen science initiatives
Developing mapping products that promote Ocean Literacy will
stimulate a continued interest in new and updated mapping
We recommend policymakers to: products. Improvement and extension of projects that couple
marine robotic systems and the Internet of Things towards the
5b) Develop partnerships with wider stakeholders development of new affordable tools for Ocean literacy will pave
on open data the way for the development of virtual Ocean exploration missions,
To incorporate industry and citizen science data into MHM-based which are key for Ocean literacy. Additionally, citizen science
decision processes it needs to adhere to the same standards as data projects that collect mapping data should be supported.

Credit: N Credit: NOAA Ocean Exploration.

Students telecommunicating directly with Research Vessel Point Sur during NOAA’s Microbial Stowaways expedition in the Gulf of Mexico99.

99
https://oceanexplorer.noaa.gov/explorations/19microbial-stowaways/welcome.html

54
 N° 11 2024

6
Overarching recommendations to
advance marine habitat mapping

Marine habitat maps provide fundamental information on the ‘what, where and how much’ for marine
habitats. Although modern sensors and autonomous platforms are revolutionising the mapping of both
benthic and pelagic habitats, this requires significant resources and increased awareness of the urgent
need to complete the mapping of our Ocean. Many of the gaps in mapping activities, such as substantially
increasing biological information, especially in the deep sea, and mapping habitat condition, require
focused investments and dedicated projects, including new mapping programmes with an ongoing
duration. Meeting this need will provide a step change in improving the understanding of ecosystem
patterns and processes, and will inform decision-making in areas such as marine resource management,
environmental change, and Ocean conservation and restoration.

To meet these needs, we recommend scientists/map temporal change within marine habitat mapping is the
producers and research funders to: only way to support a holistic, ecosystem approach to
marine ecosystem management.
• Support multidisciplinary national and EU research projects
to advance novel methods to increase the resolution of
biological information within marine habitat mapping. • Promote the standardisation of mapping methods
and outputs in research and mapping programmes.
This will enable the step change needed to improve
mapping of biological communities and species, and Guidance should cover the required spatial extent and
mapping of both the seafloor and water column as three- resolution of biological information and reporting format
dimensional maps, rather than only physical habitats and of maps for specific purposes. This includes standards for
substrates. This links to the increasing need to represent data collection and processing, and best practices on the
species and habitat distribution within marine habitat choice, selection and parameterisation of models used in
maps so they can be compatible with the new challenges marine habitat mapping. This will help with transparency
posed by the spatial management of increasing human in model selection and development, and assist managers
uses and policy activities. In addition, this will improve the in evaluating whether a model is suitable for providing
information used in spatial models, making their outputs advice for spatial management. It should also include best
more robust. practice methodology on the creation of bespoke maps
for specific purposes.

• Support national and EU research programmes that

focus on repeat mapping for capturing temporal change • Promote and incentivise research and mapping
in patterns and processes, particularly of ecologically programmes to publish marine habitat mapping data
significant spatial units, i.e. hot spots of ecosystem according to the Findable, Accessible, Interoperable and
functioning where high rates of change are expected. Reusable principles and to submit data to centralised
data services.
Repeat mapping surveys for addressing changes in
habitats over time (i.e four-dimensional maps) focusing on This will improve data sharing and includes data acquired
habitat seasonality, human impacts (including long-term by researchers and from a wider range of stakeholders
changes induced by climate change), recovery trajectories, (e.g. industry, citizen science), which are largely
and the identification of early warning signals for tipping untapped sources of information. The value of data
points, will translate into scientific knowledge to support from stakeholders should be promoted and these data
management activities. Existing mapping activities incorporated into marine habitat mapping-based decision
typically provide ‘static’ products and incorporating processes. Open and interoperable FAIR data repositories,

55
 EMB FUTURE SCIENCE BRIEF

services and portals should be sustainably maintained in support reporting for the Marine Strategy Framework
the long-term to handle increasing volumes of data and Directive, Habitats and Birds Directives, 2030 Biodiversity
stakeholders incentivised to submit data. Strategy, and the proposed EU Nature Restoration Law.
This will also support the establishment and review of
spatial plans, as required by the Maritime Spatial Planning
• Support for public-private research collaboration is needed Directive. National and regional coordination for shared
for the development of cost-effective mapping tools. resources and facilities is also required, as is the funding
of new mapping programmes.
This can assist in the development of innovative
technologies and the collection and processing of mapping
data at larger spatial scales. This includes advanced • Establish an international effort to identify priority areas
autonomous and interoperable marine robotic platforms in need of mapping, with a focus on areas of the largely
equipped with suites of complimentary sensors for data unmapped deep sea and coastal areas, which are under
collection “underway”. Scaling-up the use of artificial the greatest pressure from human activities.
intelligence can assist with cost-effective data acquisition
and data analyses to deal with the large volumes of data Guidelines for the prioritisation of marine habitat
generated by new mapping technologies. mapping activities, including the use of decision-support
tools, should be developed.

• Support dedicated mapping projects focusing on citizen

science and reformatting mapping products that promote • Require map producers (e.g. ICES Working Group on
Ocean literacy. Marine Habitat Mapping, EMODnet, large mapping
projects) or map users (e.g. the European Environment
This will stimulate continued interest for new or updated Agency, Joint Nature Conservation Committee) to
mapping products, thereby promoting the support of produce best practice and reporting templates for the
the public. Improvement and extension of projects that standardised assessment and reporting of map accuracy
couple marine robotic systems and the Internet of Things and confidence.
towards the development of new affordable tools for
Ocean literacy will pave the way for the development of These should be widely communicated within the
advanced, yet affordable systems capable of executing marine habitat mapping community to facilitate better
virtual Ocean exploration missions. Additionally, citizen understanding of the value of maps and their use for
science projects that collect mapping data should be specific purposes. Additional guidance should be produced
supported. that assists map users in how to assess certainty and
establish what is fit-for-purpose.

In addition, we recommend policymakers to:

• Advance habitat classification schemes, which lie at the
• Strengthen national, regional, European and international heart of all marine habitat maps, to include quantitative
coordination mechanisms for interdisciplinary mapping characterisation of habitats to support the assessment
efforts to ensure effective use of mapping resources and of their condition. Habitat maps will be enriched further
identification of gaps. if these classification schemes link to other sources
of information such as sensitivity to pressures and
This will help ensure adherence to defined standards, ecosystem services provision.
the submission of data into national and European data
centres and services, and the prioritisation of mapping This will facilitate building a common framework and
efforts. Coordination bodies should identify gaps in the easily understood terminology for the description of
mapping of various habitats and the spatial coverage of habitats to be consistently adopted in initiatives aimed at
marine habitat maps, and then coordinate, or commission, marine habitat mapping, monitoring and data collation.
mapping studies to close priority gaps and to include In addition, an online tool should be developed to enable
aspects of habitat condition in the collection of mapping the submission of suggestions by scientists of revisions to
data and its presentation in final maps. Such efforts will a habitat classification system.

56
 N° 11 2024

List of abbreviations and acronyms

3D Three-dimensional

4G/LTE Fourth generation long-term evolution

AI Artificial Intelligence

AUV Autonomous Underwater Vehicle

BBHTs Benthic Broad Habitat Types

CBD Convention on Biological Diversity

DNA Deoxyribose Nucleic Acid

eDNA Environmental DNA

EEA European Environment Agency

EEZ Exclusive Economic Zone

EMBRC European Marine Biological Resource Centre

EMODnet European Marine Observation and Data Network

EOV Essential Ocean Variables

EU European Union

EUNIS European Nature Information System

EurOBIS European Ocean Biodiversity Information System

FAIR Findable, Accessible, Interoperable, Reusable

FP7 Seventh Framework Programme, European Union research and development funding programme

GEBCO The General Bathymetric Chart of the Oceans

GES Good Environmental Status

GIS Geographic Information System

GOOS The Global Ocean Observing System

GSI Geological Survey Ireland

Ha Hectare

HELCOM The Baltic Marine Environment Protection Commission

HELCOM HUB The HELCOM underwater biotope and habitat classification system

HCS(s) Habitat Classification Scheme(s)

ICES International Council for the Exploration of the Sea

INSS Irish National Seabed Survey

ISPRA Italian Institute for Environmental Protection and Research

LAN Local Area Network

57
 EMB FUTURE SCIENCE BRIEF

LiDAR Light Detection and Ranging

MarESA Marine Evidence-based Sensitivity Assessment

MBES Multibeam Echosounder

MESH Mapping European Seabed Habitats

MESSENGER MErcury Surface, Space ENvironment, GEochemistry, and Ranging

MSC Marine Stewardship Council

MSFD Marine Strategy Framework Directive

MHM Marine Habitat Mapping

MHMs Marine Habitat Maps

MPA(s) Marine Protected Area(s)

NASA National Aeronautics and Space Administration

Nm Nanometre

NOAA National Oceanographic and Atmospheric Administration

OBIS Ocean Biodiversity Information System

OSPAR Oslo and Paris Conventions

PDF Portable Document Format

PNRR MER National Recovery and Resilience Plan, Marine Ecosystem Restoration

PwC PricewaterhouseCoopers

RC Reef Check

RCP Representative Concentration Pathway

ROV Remotely Operated Vehicles

SCUBA Self Contained Underwater Breathing Apparatus

SDG(s) Sustainable Development Goal(s)

SPA/RAC Specially Protected Areas Regional Activity Centre

SSS Side-Scan Sonar

UAV Unmanned Aerial Vehicle

UK United Kingdom

USA United States of America

UN United Nations

UNEP MAP- RAC/SPA United Nations Environment Programme Mediterranean Action Plan Specially Protected Areas
Regional Activity Centre

VIAME Video and Image Analytics for Marine Environments

WGMHM Working Group Marine Habitat Mapping

58
 N° 11 2024

Glossary

Abiotic - Non-living components of an ecosystem or environment.

Absence data - Observational data on where a species of interest and habitats are not present.

Accuracy - Certainty within a map in terms of location and the quality of labelling for mapped units. Accuracy is
diminished by the cumulative influence of all errors (total error) within a map. Estimates of map error rate are derived
from cross-validation between observed versus predicted classes.

Artificial Intelligence - The theory and development of computer systems that are able to perform tasks or exhibit
behaviour normally requiring human intelligence.

Autonomous surface platforms - Uncrewed vehicles designed to operate on the surface of water without direct human
intervention.

Autonomous underwater vehicle - An underwater sensor platform that undertakes a programmed survey without
input from an operator and without a cabled connection to the surface. AUVs can carry a variety of sensors and
cameras to collect data over large spatial extents and at relatively low cost.

Backscatter data - Data on the intensity of sound waves released from Multibeam Echosounder (MBES) devices
reflected back from the seabed, used to measure substrate softness and texture.

Bathymetry - Underwater topography and physical features derived from depth data.

Benthic crawler - A robot that moves independently, carrying scientific instrumentation for scanning a continuous track
of the seabed for prolonged periods.

Benthic habitat - A habitat associated with or occurring at the seafloor.

Benthic landers - Static seabed platforms containing sensors able to provide high-resolution time-series data at fixed
locations.

Biocenosis - A group of living organisms that, through their composition, number of species and individuals, reflects
the average conditions of their environment. These organisms are interconnected through mutual dependence and
permanently live and reproduce in a specific location. The term is synonymous with a biological community.

Bioconcretion - Hardened biological structures formed by the accumulation and cementation of mineral materials
within a biological environment.

Biogenic habitat - Habitats formed by living organisms, which provide a habitat for other organisms. Typical examples
include mussel beds, coral reefs, coralligenous concretions and Posidonia oceanica meadows and algal-animal forests.

Biological assemblage - A group of species that coexist in a specific habitat.

Biological community - A group of interacting organisms coexisting in a specific habitat.

Biological habitat map - A map that illustrates the spatial distribution of living organisms within that area, providing
insights into biodiversity, species composition or ecological interactions.

Biotope - A distinct habitat or environment where particular types of organisms live.

Chlorophyll-a - A green pigment found in plants, algae and some cyanobacteria. It plays a crucial role in photosynthesis.

Circalittoral - The region extending from the low tide mark to the maximum depth at which photosynthesis is possible.

59
 EMB FUTURE SCIENCE BRIEF

Condition - The ecological health of a mapped unit in terms of environmental conditions (e.g. anthropogenic
modification of key environmental properties or concentrations) and biological disturbance (i.e. impacts on the
structure or functioning of a community of species within a habitat).

Confidence - The fitness of a map for a specific use. Confidence is determined both by the accuracy of a map and the
intended purpose of the map by the end-user.

Connectivity - The extent to which populations in different parts of the species’ range are linked by the movement of
eggs, larvae or other propagules, juveniles or adults.

Contiguous data - Data that is continuous across a geographic area.

Continuous data - Data with variables that can take on an infinite number of values within a certain range e.g. salinity.

Convolutional neural networks - A type of artificial neural network that is well-suited for analysing visual data such as
images and videos. They are designed to automatically and adaptively learn spatial hierarchies of features from input data.

Coralligenous formations - A hard surface made mostly from the buildup of calcareous coralline algae.

Correlative models - Models that relate known probabilities of species presence to environmental variables.

Corer - A device that retrieves a physical sample of the uppermost layers of the seabed.

Cultural ecosystem services - Ecosystem services that provide non-material benefits derived from nature such as
recreation and tourism, beauty, as well as spiritual, intellectual and cultural benefits.

Deep learning - Capable of learning patterns directly from data.

Digital twins - Coupled observation and simulation data frameworks for human and AI-based scenario interpretation.

Direct observations - Data collected close to the object of interest.

Discrete data - Categorical data with values that are separate, with no possible values in between e.g. substrate.

Distribution models / habitat suitability models / species distribution models - Models that typically predict the
probability of the presence, or the habitat suitability, for a given species, or selection of species when applying joint
species distribution modelling.

Dredge - A tool used for collecting samples from the seabed. It typically consists of a metal frame with an attached net
or basket, which is dragged along the seabed to scoop up sediment, rocks and biological organisms.

Drop camera - A type of underwater camera that is lowered into the water column from a boat, buoy, or other platform
and attached to a cable.

Ecosystem-based approach - An approach to management where all interactions within an ecosystem, including
human interactions, are considered holistically.

Ecosystem services - The social and economic benefits obtained by society from its use of the ecological functions of
ecosystems.

Environmental DNA - Genetic material collected directly from environmental samples such as sediments or seawater.

Epibenthic communities - Biological communities on or just above benthic habitats.

Essential Fish Habitats - Areas or volumes of water or bottom substrate that are crucial for fish life stages i.e. areas
where they spawn, breed, feed and mature.

Essential Ocean Variables - A series of variables to monitor and map the Ocean consistently and cost-effectively.

60
 N° 11 2024

Eulittoral - The area of the shore between the highest and lowest tides.

Geomorphology - The shape of the seabed.

Georeferenced data - Data that are associated with a location or physical space.

Gliders - A type of autonomous underwater vehicle that is deployed from vessels for survey missions at remote
distances from the vessel. They typically do not have an engine, and instead use changes in buoyancy to move up and
down through the water.

Grab - A device that retrieves a physical sample of the uppermost layers of the seabed.

Ground truthing - The process of validating or verifying data collected via remote sensing methods or from modelling.
This is done by direct, in situ observations.

Habitat - A recognisable space which can be distinguished by its abiotic characteristics and associated biological
assemblage, assessed at particular spatial and temporal scales.

Habitat classification scheme - A set of instructions that identify, delimit, and describe the habitats of distinct species
and communities by categorising them into “classes”.

Hyperspectral imaging - Optical camera technology that records in the visible portion of the electromagnetic spectrum
(390-700nm) at up to 1nm resolution.

Infralittoral - A specific depth range within the marine environment extending from the lowest tide limit to the the
limit at which enough light penetrates to support photosynthetic organisms.

In situ observations - Samples and observations collected in the water, close to the object of interest.

Internet of Things - A world-wide network of smart interconnected objects with a digital entity.

Light detection and ranging (LiDAR) - A method for determining distances by targeting an object or a surface with a
laser and measuring the time for the reflected light to return to the receiver.

Littoral - The part of a sea or Ocean that is close to the shore.

Local area network - A network of interconnected computers and devices within a limited geographical area. They allow
computers and devices to communicate and share resources.

Machine learning - Algorithms that automatically learn to recognise complex patterns in new datasets, improve their
performance from experience and produce models that have predictive power.

Machine vision - Machines that are able to autonomously perceive, interpret and understand visual data.

Maërl bed - A biogenic structure composed of unattached calcareous red algae living on sedimentary bottoms.

Mechanistic models - Models that relate physiological information about a species gained from literature or laboratory
experiences to environmental variables for assessing their fitness at specific locations.

Megaripple bedforms - Large wave-like features on the seabed typically formed by the interaction between strong
currents and mobile sediments.

Mosaic - A representation of the seabed composed of multiple habitat types or classes arranged in a spatially
contiguous manner.

Multibeam Echosounder (MBES) - An acoustic device that uses sonar to map seabed bathymetry, morphological
characteristics and substrate types.

61
 EMB FUTURE SCIENCE BRIEF

Multispectral MBES - Sensors that acquire several MBES data using different acoustic frequencies simultaneously.

Pelagic habitats - Habitats associated with the water column.

Photogrammetry - A technique using multiple overlapping photographs to determine the size, shape and position of
objects.

Photophilic - Algae or plants that grow best in strong light.

Physical habitat map - A map delineating the environmental characteristics and features of a given area, such as
substrate type, depth, seafloor morphology and water flow.

Phytoplankton - Microscopic algae that live in the water column.

Process-oriented models - Models used to estimate species distribution based on processes such as ability to disperse
and biotic interactions.

Provisioning ecosystem services - Ecosystem services that provide tangible, harvestable goods such as fish, shellfish and
seaweed for food, raw materials, algae and minerals.

Proxy - An observable variable that is used as a substitute or indicator for a specific habitat type or ecological feature.
Proxies are often derived from remotely-sensed data and are used to infer the presence or characteristics of different
habitat types across the marine environment.

Pseudo-absence data - Proxy observations suggesting that a species of interest is highly unlikely to be present e.g.
observations of other species that are known not to co-occur with the target species.

Reflectivity - The acoustic energy reflected from the seabed or an object in the water column.

Regulating ecosystem services - Ecosystem services that regulate natural processes and maintain ecological balance,
such as coastal protection, prevention of erosion, water purification and carbon storage.

Remotely operated vehicle - An underwater platform equipped with sensors, cameras and/or manipulator arms
remotely controlled from the surface via a cable.

Remote sensing - Collecting data at a distance from the mapped area.

Resilience - The capacity of systems to persist, adapt or transform when faced with disturbances whilst maintaining
their essential functions.

Resistance - A system’s ability to actively change while retaining its identity or to passively maintain system
performance following one or more adverse events.

Satellite altimetry - A technique used to measure the height of the Ocean’s surface from space, which varies depending
on bathymetry therefore indirectly providing information about the seabed.

Sensitivity - The degree to which marine features respond to stressors, which are deviations of environmental
conditions beyond the expected range.

Sessile - An organism that is fixed in one place i.e. immobile.

Side-scan sonar - A type of sonar that emits acoustic pulses across a wide angle perpendicular to the path of the sensor
through the water. Stacking the responses along the track line creates an image of reflection strength. It is used to
create images of large areas of the seabed and bathymetric features.

Status - A broad, composite assessment of various aspects of multiple habitats, used by marine managers to capture
overall ecosystem health.

62
 N° 11 2024

Substrate - Bottom type, also known as substrata/substratum.

Substrate map - A map depicting the sediment and rock type of the seabed with little or no information on the
biological communities present.

Super-resolution - Enhanced resolution.

Systematic conservation planning - A multi-component, stage-wise approach to identifying conservation areas and
devising management policies, with feedback, revision and reiteration, where needed, at any stage.

Three-dimensional marine habitat mapping - Mapping that includes multiple depth ranges of the distribution of
biodiversity and includes species distributions by incorporating their life cycle, trophic interactions and exchanges
between the water column and the seafloor.

Tipping points - The critical point at which a rapid and unexpected shift is triggered and an ecosystem transitions to a
new state with altered composition and functioning.

Towfish - An underwater vehicle, usually carrying instrumentation such as a side-scan sonar, that is towed behind a
surface vessel.

Trawl - A type of sampling device used to collect benthic samples. Trawls are dragged along the seabed by a vessel,
scooping up sediment, rocks and benthic organisms.

Underway data - Opportunistic data collected during transits or non-mapping voyages.

Unmanned aerial vehicle - An aircraft without any human pilot, crew, or passengers on board, commonly referred to as
a drone.

Virtual research environments - Immersive virtual reality displays of complex data streams.

Vulnerability - The probability that a feature will be exposed to a stressor to which it is sensitive.

Water column - The vertical column of water extending from the surface of the Ocean to the seabed.

63
 EMB FUTURE SCIENCE BRIEF

References

Al-Hamdani, Z., & Reker, J. (2007). Towards marine landscapes in the Baltic Sea. BALANCE interim report #10.
http://balance-eu.org/

Amon, D. J., Gollner, S., Morato, T., Smith, C. R., Chen, C., Christiansen, S., Currie, B., Drazen, J. C., Fukushima, T., Gianni, M.,
Gjerde, K. M., Gooday, A. J., Grillo, G. G., Haeckel, M., Joyini, T., Ju, S.-J., Levin, L. A., Metaxas, A., Mianowicz, K., … Pickens, C.
(2022). Assessment of scientific gaps related to the effective environmental management of deep-seabed mining. Marine
Policy, 138, 105006. https://doi.org/https://doi.org/10.1016/j.marpol.2022.105006

Andersen, J. H., Manca, E., Agnesi, S., Al-Hamdani, Z., Lillis, H., Mo, G., Populus, J., Reker, J., Tunesi, L., & Vasquez, M. (2018).
European Broad-Scale Seabed Habitat Maps Support Implementation of Ecosystem-Based Management. Open Journal of
Ecology, 08(02), 86–103. https://doi.org/10.4236/oje.2018.82007

Angeletti, L., Bargain, A., Campiani, E., Foglini, F., Grande, V., Leidi, E., Mercorella, A., Prampolini, M., & Taviani, M. (2019). 16
Cold-Water Coral Habitat Mapping in the Mediterranean Sea: Methodologies and Perspectives. In C. Orejas & C. Jiménez
(Eds.), Mediterranean Cold-Water Corals: Past, Present and Future: Understanding the Deep-Sea Realms of Coral (pp. 173–
189). Springer International Publishing. https://doi.org/10.1007/978-3-319-91608-8_16

Asjes, A., González-Irusta, J., & Wright, P. (2016). Age-related and seasonal changes in haddock Melanogrammus aeglefinus
distribution: Implications for spatial management. Marine Ecology Progress Series, 553. https://doi.org/10.3354/meps11754

Bargain, A., Foglini, F., Pairaud, I., Bonaldo, D., Carniel, S., Angeletti, L., Taviani, M., Rochette, S., & Fabri, M. C. (2018).
Predictive habitat modeling in two Mediterranean canyons including hydrodynamic variables. Progress in Oceanography,
169, 151–168. https://doi.org/10.1016/j.pocean.2018.02.015

Beca-Carretero, P., Teichberg, M., Winters, G., Procaccini, G., & Reuter, H. (2020). Projected Rapid Habitat Expansion
of Tropical Seagrass Species in the Mediterranean Sea as Climate Change Progresses. Frontiers in Plant Science, 11.
https://doi.org/10.3389/fpls.2020.555376

Bevilacqua, S., Guarnieri, G., Farella, G., Terlizzi, A., & Fraschetti, S. (2018). A regional assessment of cumulative impact mapping
on Mediterranean coralligenous outcrops. Scientific Reports, 8(1), 1757. https://doi.org/10.1038/s41598-018-20297-1

Boero, F., De Leo, F., Fraschetti, S., & Ingrosso, G. (2019). Chapter Four - The Cells of Ecosystem Functioning: Towards a
holistic vision of marine space. In C. Sheppard (Ed.), Advances in Marine Biology (Vol. 82, pp. 129–153). Academic Press.
https://doi.org/10.1016/bs.amb.2019.03.001

Boero, F., Foglini, F., Fraschetti, S., Goriup, P. D., Macpherson, E., Planes, S., & Soukissian, T. H. (2016). CoCoNet: Towards
coast to coast networks of marine protected areas (From the shore to the high and deep sea), coupled with sea-based
wind energy potential. SCIRES-IT: SCIentific RESearch and Information Technology, 6, 1–95. https://api.semanticscholar.
org/CorpusID:54783314

Brown, C. J., Beaudoin, J., Brissette, M., & Gazzola, V. (2019). Multispectral Multibeam Echo Sounder Backscatter as a Tool
for Improved Seafloor Characterization. Geosciences, 9(3). https://doi.org/10.3390/geosciences9030126

Brown, C. J., Smith, S. J., Lawton, P., & Anderson, J. T. (2011). Benthic habitat mapping: A review of progress towards
improved understanding of the spatial ecology of the seafloor using acoustic techniques. Estuarine, Coastal and Shelf
Science, 92(3), 502–520. https://doi.org/10.1016/j.ecss.2011.02.007

Bulleri, F., Batten, S., Connell, S., Benedetti-Cecchi, L., Gibbons, M., Nugues, M., & Gribben, P. (2020). Human pressures and
the emergence of novel marine ecosystems (pp. 441–493). https://doi.org/10.1201/9780429351495-9

Burgos, J. M., Buhl-Mortensen, L., Buhl-Mortensen, P., Ólafsdóttir, S. H., Steingrund, P., Ragnarsson, S. Á., & Skagseth, Ø.
(2020). Predicting the Distribution of Indicator Taxa of Vulnerable Marine Ecosystems in the Arctic and Sub-arctic Waters
of the Nordic Seas. Frontiers in Marine Science, 7. https://doi.org/10.3389/fmars.2020.00131

64
 N° 11 2024

Casal, G., Sánchez-Carnero, N., Sánchez-Rodríguez, E., & Freire, J. (2011). Remote sensing with SPOT-4 for mapping
kelp forests in turbid waters on the south European Atlantic shelf. Estuarine, Coastal and Shelf Science, 91(3), 371–378.
https://doi.org/10.1016/j.ecss.2010.10.024

Castellan, G., Angeletti, L., Correggiari, A., Foglini, F., Grande, V., & Taviani, M. (2020). Visual Methods for Monitoring
Mesophotic-to-Deep Reefs and Animal Forests: Finding a Compromise Between Analytical Effort and Result Quality. In
S. Rossi & L. Bramanti (Eds.), Perspectives on the Marine Animal Forests of the World (pp. 487–514). Springer International
Publishing. https://doi.org/10.1007/978-3-030-57054-5_15

CBD. (1992). Convention on Biological Diversity. https://www.cbd.int/convention/text/

CBD/COP/DEC/15/4. (2022). Kunming-Montreal Global Biodiversity Framework. https://www.cbd.int/doc/decisions/cop-

15/cop-15-dec-04-en.pdf

COM/2007/574 final. (2007). Communication from the Commission to the European Parliament, the Council, the European
Economic and Social Committee and the Committee of the Regions - An Integrated Maritime Policy for the European Union
(COM/2019/6 final). http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52007DC0575

COM/2019/640 final. (2019). Communication from the Commission to the European Parliament, the European Council,
the Council, the European Economic and Social Committee and the Committee of the Regions: The European Green Deal.
https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=COM%3A2019%3A640%3AFIN

COM/2020/380 final. (2020). Communication from the Commission to the European Parliament, the Council, the European
Economic and Social Committee and the Committee of the Regions: EU Biodiversity Strategy for 2030 Bringing nature back
into our lives. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A52020DC0380

COM/2021/240 final. (2021). COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT, THE COUNCIL,
THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE AND THE COMMITTEE OF THE REGIONS on a new approach for a
sustainable blue economy in the EU Transforming the EU’s Blue Economy for a Sustainable Future. https://eur-lex.europa.
eu/legal-content/EN/TXT/?uri=CELEX%3A52021DC0240

COM/2022/304 final. (2022). Proposal for a Regulation of the European Parliament and of the Council on nature restoration.
https://environment.ec.europa.eu/publications/nature-restoration-law_en

Cooper, K. M., & Barry, J. (2020). A new machine learning approach to seabed biotope classification. Ocean & Coastal
Management, 198, 105361. https://doi.org/10.1016/j.ocecoaman.2020.105361

Council Directive 92/43/EEC. (1992). Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats
and of wild fauna and flora. Official Journal of the European Union, 11, 0114. https://eur-lex.europa.eu/legal-content/EN/
TXT/?uri=celex%3A31992L0043

Danovaro, R., Bianchelli, S., Brambilla, P., Brussa, G., Corinaldesi, C., Del Borghi, A., Dell’Anno, A., Fraschetti, S., Greco, S.,
Grosso, M., Nepote, E., Rigamonti, L., & Boero, F. (2024). Making eco-sustainable floating offshore wind farms: Siting,
mitigations, and compensations. Renewable and Sustainable Energy Reviews, 197, 114386. https://doi.org/10.1016/j.
rser.2024.114386

de la Torriente, A., González-Irusta, J. M., Aguilar, R., Fernández-Salas, L. M., Punzón, A., & Serrano, A. (2019). Benthic habitat
modelling and mapping as a conservation tool for marine protected areas: A seamount in the western Mediterranean.
Aquatic Conservation: Marine and Freshwater Ecosystems, 29(5), 732–750. https://doi.org/https://doi.org/10.1002/
aqc.3075

Directive 2000/60/EC. (2000). Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000
establishing a framework for Community action in the field of water policy. Official Journal of the European Union, 327,
0001–0037. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A32000L0060

Directive 2008/56/EC. (2008). Directive 2008/56/EC of the European Parliament and of the Council of 17 June 2008
establishing a framework for community action in the field of marine environmental policy (Marine Strategy Framework
Directive). Official Journal of the European Union, L164/19. http://data.europa.eu/eli/dir/2008/56/oj

65
 EMB FUTURE SCIENCE BRIEF

Directive 2009/147/EC. (2009). Directive 2009/147/EC of the European Parliament and of the Council of 30 November
2009 on the conservation of wild birds (Codified version). Official Journal of the European Union, 20, 7. https://eur-lex.
europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32009L0147

Directive 2014/89/EU. (2014). Directive 2014/89/EU of the European Parliament and of the Council of 23 July 2014
establishing a framework for maritime spatial planning. Official Journal of the European Union, 257, 135. https://eur-lex.
europa.eu/legal-content/EN/TXT/?uri=celex%3A32014L0089

Durden, J. M., Hosking, B., Bett, B. J., Cline, D., & Ruhl, H. A. (2021). Automated classification of fauna in seabed photographs:
The impact of training and validation dataset size, with considerations for the class imbalance. Progress in Oceanography,
196, 102612. https://doi.org/10.1016/j.pocean.2021.102612

Effrosynidis, D., Arampatzis, A., & Sylaios, G. (2018). Seagrass detection in the mediterranean: A supervised learning
approach. Ecological Informatics, 48, 158–170. https://doi.org/10.1016/j.ecoinf.2018.09.004

Elith, J., & Leathwick, J. R. (2009). Species Distribution Models: Ecological Explanation and Prediction Across Space
and Time. Annual Review of Ecology, Evolution, and Systematics, 40(1), 677–697. https://doi.org/10.1146/annurev.
ecolsys.110308.120159

Ernst, C. M., Chabot, N. L., Klima, R. L., Kubota, S., Rogers, G., Byrne, P. K., Hauck, S. A., Vander Kaaden, K. E., Vervack, R. J.,
Besse, S. & Blewett, D. T. (2022). Science goals and mission concept for a landed investigation of Mercury. The Planetary
Science Journal, 3(3), 68. https://doi/10.3847/PSJ/ac1c0f

European Commission. (2017). Commission Decision (EU) 2017/848 of 17 May 2017 laying down criteria and methodological
standards on good environmental status of marine waters and specifications and standardised methods for monitoring and
assessment, and repealing Decision 2010/477/EU. https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32017
D0848&from=EN

European Marine Board. (2019). Navigating the Future V: Marine Science for a Sustainable Future. In J. J. Heymans, B.
Alexander, A. Muniz Piniella, P. Kellett, J. Coopman, & K. Larkin (Eds.), EMB Position Paper 24. https://doi.org/10.5281/
zenodo.2809392

Fabbrizzi, E., Giakoumi, S., De Leo, F., Tamburello, L., Chiarore, A., Colletti, A., Coppola, M., Munari, M., Musco, L., Rindi, F.,
Rizzo, L., Savinelli, B., Franzitta, G., Grech, D., Cebrian, E., Verdura, J., Bianchelli, S., Mangialajo, L., Nasto, I., … Fraschetti, S.
(2023). The challenge of setting restoration targets for macroalgal forests under climate changes. Journal of Environmental
Management, 326, 116834. https://doi.org/10.1016/j.jenvman.2022.116834

Fabbrizzi, E., Scardi, M., Ballesteros, E., Benedetti-Cecchi, L., Cebrian, E., Ceccherelli, G., De Leo, F., Deidun, A., Guarnieri, G.,
Falace, A., Fraissinet, S., Giommi, C., Mačić, V., Mangialajo, L., Mannino, A. M., Piazzi, L., Ramdani, M., Rilov, G., Rindi, L., …
Fraschetti, S. (2020). Modeling Macroalgal Forest Distribution at Mediterranean Scale: Present Status, Drivers of Changes
and Insights for Conservation and Management. Frontiers in Marine Science, 7. https://doi.org/10.3389/fmars.2020.00020

Farrag, M. M. S. (2022). Towards the identification of essential fish habitats for commercial deep-water species. In M.
Otero & C. Mytilineou (Eds.), Deep-sea Atlas of the Eastern Mediterranean Sea. IUCN. https://www.semanticscholar.org/
paper/Towards-the-identification-of-essential-fish-for/547c8c34b73fc5a20628ec57a96f6fcdf047f2dc

Ferrier, S., & Guisan, A. (2006). Spatial modelling of biodiversity at the community level. Journal of Applied Ecology, 43,
393–404. https://api.semanticscholar.org/CorpusID:85000606

Fraschetti, S., Pipitone, C., Mazaris, A. D., Rilov, G., Badalamenti, F., Bevilacqua, S., Claudet, J., Carić, H., Dahl, K., D’Anna,
G., Daunys, D., Frost, M., Gissi, E., Göke, C., Goriup, P., Guarnieri, G., Holcer, D., Lazar, B., Mackelworth, P., … Katsanevakis,
S. (2018). Light and Shade in Marine Conservation Across European and Contiguous Seas. Frontiers in Marine Science, 5.
https://doi.org/10.3389/fmars.2018.00420

Galparsoro, I., Borja, A., & Uyarra, M. C. (2014). Mapping ecosystem services provided by benthic habitats in the European
North Atlantic Ocean. Frontiers in Marine Science, 1. https://doi.org/10.3389/fmars.2014.00023

66
 N° 11 2024

Galparsoro, I., Connor, D. W., Borja, Á., Aish, A., Amorim, P., Bajjouk, T., Chambers, C., Coggan, R., Dirberg, G., Ellwood, H.,
Evans, D., Goodin, K. L., Grehan, A., Haldin, J., Howell, K., Jenkins, C., Michez, N., Mo, G., Buhl-Mortensen, P., … Vasquez,
M. (2012). Using EUNIS habitat classification for benthic mapping in European seas: Present concerns and future needs.
Marine Pollution Bulletin, 64(12), 2630–2638. https://doi.org/10.1016/j.marpolbul.2012.10.010

Gerovasileiou, V., Smith, C. J., Sevastou, K., Papadopoulou, N., Dailianis, T., Bekkby, T., Fiorentino, D., McOwen, C. J., Amaro,
T., Bengil, E. G. T., Bilan, M., Boström, C., Carreiro-Silva, M., Cebrian, E., Cerrano, C., Danovaro, R., Fraschetti, S., Gagnon, K.,
Gambi, C., … Scrimgeour, R. (2019). Habitat mapping in the European Seas - is it fit for purpose in the marine restoration
agenda? Marine Policy, 106, 103521. https://doi.org/10.1016/j.marpol.2019.103521

Giakoumi, S., Sini, M., Gerovasileiou, V., Mazor, T., Beher, J., Possingham, H. P., Abdulla, A., Çınar, M. E., Dendrinos, P., Gucu, A.
C., Karamanlidis, A. A., Rodić, P., Panayotidis, P., Taşkın, E., Jaklin, A., Voultsiadou, E., Webster, C., Zenetos, A., & Katsanevakis,
S. (2013). Ecoregion-Based Conservation Planning in the Mediterranean: Dealing with Large-Scale Heterogeneity. PLoS
ONE, 8. https://api.semanticscholar.org/CorpusID:10861874

Greathead, C., González-Irusta, J., Boulcott, P., Blackadder, L., Weetman, A., & Wright, P. (2014). Environmental
requirements for three sea pen species: Relevance to distribution and conservation. ICES Journal of Marine Science, 72.
https://doi.org/10.1093/icesjms/fsu129

Guidi, L., Guerra, A. F., Bakker, D. C. E., Canchaya, C., Curry, E., Foglini, F., Irisson, J.-O., Malde, K., Marshall, C. T., Obst, M.,
Ribeiro, R. P., & Tjiputra, J. (2020). Big Data in Marine Science. In J. J. Heymans, B. Alexander, A. Muniz Piniella, P. Kellett, &
J. Coopman (Eds.), EMB Future Science Brief 6. https://doi.org/10.5281/zenodo.3755793

Halpern, B. S., Walbridge, S., Selkoe, K. A., Kappel, C. V, Micheli, F., D’Agrosa, C., Bruno, J. F., Casey, K. S., Ebert, C., Fox, H.
E., Fujita, R., Heinemann, D., Lenihan, H. S., Madin, E. M. P., Perry, M. T., Selig, E. R., Spalding, M., Steneck, R., & Watson, R.
(2008). A Global Map of Human Impact on Marine Ecosystems. Science, 319(5865), 948–952. https://doi.org/10.1126/
science.1149345

HELCOM. (2021). Essential fish habitats in the Baltic Sea - Identification of potential spawning, recruitment and nursery
areas. https://helcom.fi/wp-content/uploads/2021/09/Essential-fish-habitats-in-the-Baltic-Sea.pdf

Hering, D., Schürings, C., Wenskus, F., Blackstock, K., Borja, A., Birk, S., Bullock, C., Carvalho, L., Dagher-Kharrat, M. B., Lakner,
S., Lovrić, N., McGuinness, S., Nabuurs, G.-J., Sánchez-Arcilla, A., Settele, J., & Pe’er, G. (2023). Securing success for the
Nature Restoration Law. Science, 382(6676), 1248–1250. https://doi.org/10.1126/science.adk1658

Hill, N. A., Foster, S. D., Duhamel, G., Welsford, D., Koubbi, P., & Johnson, C. R. (2017). Model-based mapping of assemblages
for ecology and conservation management: A case study of demersal fish on the Kerguelen Plateau. Diversity and
Distributions, 23(10), 1216–1230. https://doi.org/10.1111/ddi.12613

Howell, K., Holt, R., Endrino, P., & Stewart, H. (2011). When the species is also a habitat: Comparing the predictively modelled
distributions of Lophelia pertusa and the reef habitat it forms. Biological Conservation, 144. https://doi.org/10.1016/j.
biocon.2011.07.025

ICES. (2005). Report of the Working Group on Habitat Mapping (WGMHM), 5–8 April, Bremerhaven, Germany.
https://www.ices.dk/sites/pub/CM%20Doccuments/2005/E/WGMHM05.pdf

ICES. (2021). Workshop on the Use of Predictive Habitat Models in ICES Advice (WKPHM). https://doi.org/10.17895/ices.
pub.8213

Illa-López, L., Cabrito, A., de Juan, S., Maynou, F., & Demestre, M. (2023). Distribution of rhodolith beds and their functional
biodiversity characterisation using ROV images in the western Mediterranean Sea. Science of The Total Environment, 905,
167270. https://doi.org/10.1016/j.scitotenv.2023.167270

Ingrosso, G., Abbiati, M., Badalamenti, F., Bavestrello, G., Belmonte, G., Cannas, R., Benedetti-Cecchi, L., Bertolino, M.,
Bevilacqua, S., Bianchi, C. N., Bo, M., Boscari, E., Cardone, F., Cattaneo-Vietti, R., Cau, A., Cerrano, C., Chemello, R., Chimienti,
G., Congiu, L., … Boero, F. (2018). Chapter Three - Mediterranean Bioconstructions Along the Italian Coast. In C. Sheppard
(Ed.), Advances in Marine Biology (Vol. 79, pp. 61–136). Academic Press. https://doi.org/10.1016/bs.amb.2018.05.001

67
 EMB FUTURE SCIENCE BRIEF

Korpinen, S., Laamanen, L., Bergström, L., Nurmi, M., Andersen, J. H., Haapaniemi, J., Harvey, E. T., Murray, C. J., Peterlin, M.,
Kallenbach, E., Klančnik, K., Stein, U., Tunesi, L., Vaughan, D., & Reker, J. (2021). Combined effects of human pressures on
Europe’s marine ecosystems. Ambio, 50(7), 1325–1336. https://doi.org/10.1007/s13280-020-01482-x

Kukkala, A. S., & Moilanen, A. (2013). Core concepts of spatial prioritisation in systematic conservation planning. Biological
Reviews, 88(2), 443–464. https://doi.org/10.1111/brv.12008

Levin, N., Kark, S., & Danovaro, R. (2018). Adding the Third Dimension to Marine Conservation. Conservation Letters, 11(3),
e12408. https://doi.org/10.1111/conl.12408

Lurton, X. (2010). An introduction to Underwater Acoustics: Principles and Applications. 2nd edition. Springer Praxis Books &
Praxis Publishing.

Lurton, X., Lamarche, G., Brown, C. J., Lucieer, V., Rice, G., Schimel, A. C. G., & Weber, T. C. (2015). Backscatter measurements
by seafloor‐mapping sonars: guidelines and recommendations. https://api.semanticscholar.org/CorpusID:129508402

Maes, J., Teller, A., Erhard, M., Condé, S., Vallecillo, S., Barredo, J. I., Paracchini, M. L., Abdul Malak, D., Trombetti, M.,
Vigiak, O., Zulian, G., Addamo, A. M., Grizzetti, B., Somma, F., Hagyo, A., Vogt, P., Polce, C., Jones, A., Marin, A. I., … Santos-
-Martín, F. (2020). Mapping and Assessment of Ecosystems and their Services: An EU ecosystem assessment, EUR 30161 EN.
https://doi.org/10.2760/757183

Malde, K., Handegard, N. O., Eikvil, L., & Salberg, A.-B. (2020). Machine intelligence and the data-driven future of marine
science. ICES Journal of Marine Science, 77(4), 1274–1285. https://doi.org/10.1093/icesjms/fsz057

Marlow, J., Halpin, J. E., & Wilding, T. A. (2024). 3D photogrammetry and deep-learning deliver accurate estimates of
epibenthic biomass. Methods in Ecology and Evolution, n/a(n/a). https://doi.org/10.1111/2041-210X.14313

Martin, C., Giannoulaki, M., De Leo, F., Scardi, M., Salomidi, M., Knittweis, L., Pace, M. L., Garofalo, G., Gristina, M., Ballesteros,
E., Bavestrello, G., Belluscio, A., Cebrian, E., Gerakaris, V., Pergent, G., Pergent-Martini, C., Schembri, P. J., Terribile, K., Rizzo, L.,
... Fraschetti, S. (2014). Coralligenous and maërl habitats: predictive modelling to identify their spatial distributions across
the Mediterranean Sea. Scientific Reports, 4, 5073. https://doi.org/10.1038/srep05073

Matear, L., Vina-Herbon, C., Woodcock, K. A., Duncombe-Smith, S. W., Smith, A. P., Schmitt, P., Kreutle, A., Marra, S., Curtis,
E. J., & Baigent, H. N. (2023). Extent of Physical Disturbance to Benthic Habitats: Fisheries. In: OSPAR, 2023: The 2023 Quality
Status Report for the Northeast Atlantic. https://oap.ospar.org/en/ospar-assessments/quality-status-reports/qsr-2023/
indicator-assessments/phys-dist-habs-fisheries/

Mayer, L., Jakobsson, M., Allen, G., Dorschel, B., Falconer, R., Ferrini, V., Lamarche, G., Snaith, H., & Weatherall, P. (2018). The
Nippon Foundation—GEBCO Seabed 2030 Project: The Quest to See the World’s Oceans Completely Mapped by 2030.
Geosciences, 8(2). https://doi.org/10.3390/geosciences8020063

Mazaris, A. D., Kallimanis, A., Gissi, E., Pipitone, C., Danovaro, R., Claudet, J., Rilov, G., Badalamenti, F., Stelzenmüller, V.,
Thiault, L., Benedetti-Cecchi, L., Goriup, P., Katsanevakis, S., & Fraschetti, S. (2019). Threats to marine biodiversity in
European protected areas. Science of The Total Environment, 677, 418–426. https://doi.org/10.1016/j.scitotenv.2019.04.333

Melo-Merino, S. M., Reyes-Bonilla, H., & Lira-Noriega, A. (2020). Ecological niche models and species distribution
models in marine environments: A literature review and spatial analysis of evidence. Ecological Modelling, 415, 108837.
https://doi.org/10.1016/j.ecolmodel.2019.108837

Miesner, A. K., & Payne, M. R. (2018). Oceanographic variability shapes the spawning distribution of blue whiting
(Micromesistius poutassou). Fisheries Oceanography, 27(6), 623–638. https://doi.org/10.1111/fog.12382

Mitchell, P. J., Downie, A.-L., & Diesing, M. (2018). How good is my map? A tool for semi-automated thematic mapping and
spatially explicit confidence assessment. Environmental Modelling & Software, 108, 111–122. https://doi.org/10.1016/j.
envsoft.2018.07.014

Montefalcone, M., Tunesi, L., & Ouerghi, A. (2021). A review of the classification systems for marine benthic habitats
and the new updated Barcelona Convention classification for the Mediterranean. Marine Environmental Research, 169,
105387. https://doi.org/10.1016/j.marenvres.2021.105387

68
 N° 11 2024

Montes-Herrera, J. C., Cimoli, E., Cummings, V., Hill, N., Lucieer, A., & Lucieer, V. (2021). Underwater Hyperspectral
Imaging (UHI): A Review of Systems and Applications for Proximal Seafloor Ecosystem Studies. Remote Sensing, 13(17).
https://doi.org/10.3390/rs13173451

Mora, C., Tittensor, D., Adl, S., Simpson, A., & Worm, B. (2011). How Many Species Are There on Earth and in the Ocean?
PLoS Biology, 9, e1001127. https://doi.org/10.1371/journal.pbio.1001127

Moritz, C., Lévesque, M., Gravel, D., Vaz, S., Archambault, D., & Archambault, P. (2013). Modelling spatial distribution of
epibenthic communities in the Gulf of St. Lawrence (Canada). Journal of Sea Research, 78, 75–84. https://doi.org/10.1016/j.
seares.2012.10.009

Morris, K., Epstein, G., Kaiser, M. J., Porter, J., & Johnson, A. F. (2023). Adapting the marine stewardship council’s risk-based
framework to assess the impact of towed bottom fishing gear on blue carbon habitats. PLOS ONE, 18(11), e0288484-.
https://doi.org/10.1371/journal.pone.0288484

Murillo, F. J., Weigel, B., Bouchard Marmen, M., & Kenchington, E. (2020). Marine epibenthic functional diversity on Flemish
Cap (north-west Atlantic)—Identifying trait responses to the environment and mapping ecosystem functions. Diversity
and Distributions, 26(4), 460–478. https://doi.org/10.1111/ddi.13026

Panayotidis, P., Papathanasiou, V., Gerakaris, V., Fakiris, E., Orfanidis, S., Papatheodorou, G., Kosmidou, M., Georgiou,
N., Drakopoulou, P., & Loukaidi, V. (2022). Seagrass Meadows in The Greek Seas. In SEANOE (Ed.), SEANOE.
https://doi.org/10.17882/87740

Piechaud, N., & Howell, K. L. (2022). Fast and accurate mapping of fine scale abundance of a VME in the deep sea with
computer vision. Ecological Informatics, 71, 101786. https://doi.org/10.1016/j.ecoinf.2022.101786

Pouvreau, S., Cochet, H., Bargat, F., Petton, S., Le Roy, V., Guillet, T., & Potet, M. (2021). Current distribution of the residual
flat oysters beds (Ostrea edulis) along the west coast of France. In SEANOE. https://doi.org/10.17882/79821

Prampolini, M., Angeletti, L., Castellan, G., Grande, V., Le Bas, T., Taviani, M., & Foglini, F. (2021). Benthic Habitat Map of the
Southern Adriatic Sea (Mediterranean Sea) from Object-Based Image Analysis of Multi-Source Acoustic Backscatter Data.
Remote Sensing, 13(15). https://doi.org/10.3390/rs13152913

Prampolini, M., Savini, A., Foglini, F., & Soldati, M. (2020). Seven Good Reasons for Integrating Terrestrial and Marine Spatial
Datasets in Changing Environments. Water, 12(8). https://doi.org/10.3390/w12082221

PricewaterhouseCoopers. (2008). INFOMAR Marine Mapping Study. Options Appraisal Report: Final Report 30 June 2008.
Marine Institute. http://hdl.handle.net/10793/1652

Pulido Mantas, T., Roveta, C., Calcinai, B., di Camillo, C. G., Gambardella, C., Gregorin, C., Coppari, M., Marrocco, T., Puce,
S., Riccardi, A., & Cerrano, C. (2023). Photogrammetry, from the Land to the Sea and Beyond: A Unifying Approach to
Study Terrestrial and Marine Environments. Journal of Marine Science and Engineering, 11(4). https://doi.org/10.3390/
jmse11040759

Ramiro-Sánchez, B., González-Irusta, J. M., Henry, L.-A., Cleland, J., Yeo, I., Xavier, J. R., Carreiro-Silva, M., Sampaio, Í.,
Spearman, J., Victorero, L., Messing, C. G., Kazanidis, G., Roberts, J. M., & Murton, B. (2019). Characterization and Mapping
of a Deep-Sea Sponge Ground on the Tropic Seamount (Northeast Tropical Atlantic): Implications for Spatial Management
in the High Seas. Frontiers in Marine Science, 6. https://doi.org/10.3389/fmars.2019.00278

Regulation 691/2011. (2011). European environmental economics accounts. Official Journal of the European Union, L
192/1. https://eur-lex.europa.eu/legal-content/EN/ALL/?uri=celex%3A32011R0691

Regulation EU 1380/2013. (2013). Regulation (EU) No 1380/2013 of the European Parliament and of the Council
of 11 December 2013 on the Common Fisheries Policy (2013). Official Journal of the European Union, 354, 22.
http://data.europa.eu/eli/reg/2013/1380/oj

Rindi, L., Mintrone, C., Ravaglioli, C., & Benedetti-Cecchi, L. (2024). Spatial signatures of an approaching regime
shift in Posidonia oceanica meadows. Marine Environmental Research, 198, 106499. https://doi.org/10.1016/j.
marenvres.2024.106499

69
 EMB FUTURE SCIENCE BRIEF

Robinson, C. L. K., & Levings, C. D. (1995). An overview of habitat classification systems, ecological models, and geographic
information systems applied to shallow foreshore marine habitats. Canadian Manuscript Report of Fisheries and Aquatic
Sciences 2322. https://publications.gc.ca/site/fra/464405/publication.html

Rodríguez-Basalo, A., Ríos, P., Arrese, B., Abad Uribarren, A., Cristobo, J., Patrocinio Ibarrola, T., Ballesteros, M., Prado, E., &
Sánchez, F. (2022). Mapping the habitats of a complex circalittoral rocky shelf in the Cantabrian Sea (south Bay of Biscay).
Estuarine, Coastal and Shelf Science, 273, 107912. https://doi.org/10.1016/j.ecss.2022.107912

Salman, A., Jalal, A., Shafait, F., Mian, A., Shortis, M., Seager, J., & Harvey, E. (2016). Fish species classification in
unconstrained underwater environments based on deep learning. Limnology and Oceanography: Methods, 14(9), 570–585.
https://doi.org/10.1002/lom3.10113

Saunders, R. S., Spear, A. J., Allin, P. C., Austin, R. S., Berman, A. L., Chandlee, R. C., Clark, J., Decharon, A. V., De Jong, E. M.,
Griffith, D. G. & Gunn, J. M. (1992). Magellan mission summary. Journal of Geophysical Research: Planets, 97(E8), 13067-
13090. https://doi.org/10.1029/92JE01397

Sidiropoulos, P. & Muller, J. P. (2015). On the status of orbital high-resolution repeat imaging of Mars for the observation of
dynamic surface processes. Planetary and Space Science, 117, 207-222. https://doi.org/10.1016/j.pss.2015.06.017

Stevens, T., & Connolly, R. M. (2004). Testing the utility of abiotic surrogates for marine habitat mapping at scales relevant
to management. Biological Conservation, 119(3), 351–362. https://doi.org/10.1016/J.BIOCON.2003.12.001

Strong, J. A. (2020). An error analysis of marine habitat mapping methods and prioritised work packages required to
reduce errors and improve consistency. Estuarine, Coastal and Shelf Science, 240, 106684. https://doi.org/10.1016/j.
ecss.2020.106684

Strong, J. A., Clements, A., Lillis, H., Galparsoro, I., Bildstein, T., & Pesch, R. (2019). A review of the influence of marine
habitat classification schemes on mapping studies: inherent assumptions, influence on end products, and suggestions for
future developments. ICES Journal of Marine Science, 76(1), 10–22. https://doi.org/10.1093/icesjms/fsy161

Taviani, M. (2014). Marine Chemosynthesis in the Mediterranean Sea. In Z. Goffredo Stefano and Dubinsky (Ed.), The
Mediterranean Sea: Its history and present challenges (pp. 69–83). Springer Netherlands. https://doi.org/10.1007/978-94-
007-6704-1_5

Teague, J., Willans, J., Allen, M., Scott, T., & Day, J. (2019). Hyperspectral imaging as a tool for assessing coral health utilising
natural fluorescence. 8. https://doi.org/10.1255/jsi.2019.a7

Thomsen, P. F., & Willerslev, E. (2015). Environmental DNA – An emerging tool in conservation for monitoring past and
present biodiversity. Biological Conservation, 183, 4–18. https://doi.org/10.1016/j.biocon.2014.11.019

Tozer, B., Sandwell, D.T., Smith, W.H., Olson, C., Beale, J.R. & Wessel, P., (2019). Global bathymetry and topography at 15 arc
sec: SRTM15+. Earth and Space Science, 6(10), 1847-1864. https://doi.org/10.1029/2019EA000658

Traganos, D., Lee, C. B., Blume, A., Poursanidis, D., Čižmek, H., Deter, J., Mačić, V., Montefalcone, M., Pergent, G., Pergent-
Martini, C., Ricart, A. M., & Reinartz, P. (2022). Spatially Explicit Seagrass Extent Mapping Across the Entire Mediterranean.
Frontiers in Marine Science, 9. https://doi.org/10.3389/fmars.2022.871799

Turicchia, E., Ponti, M., Rossi, G., & Cerrano, C. (2021a). The Reef Check Med Dataset on Key Mediterranean Marine Species
2001–2020. Frontiers in Marine Science, 8. https://doi.org/10.3389/fmars.2021.675574

Turicchia, E., Ponti, M., Rossi, G., Milanese, M., Di Camillo, C. G., & Cerrano, C. (2021b). The Reef Check Mediterranean
Underwater Coastal Environment Monitoring Protocol. Frontiers in Marine Science, 8. https://doi.org/10.3389/
fmars.2021.620368

Tyler-Walters, H., Tillin, H. M., d’Avack, E. A. S., Perry, F., & Stamp, T. (2023). Marine Evidencebased Sensitivity Assessment
(MarESA) – Guidance Manual. Marine Life Information Network (MarLIN). https://www.marlin.ac.uk/publications

United Nations. (2015). Transforming Our World: The 2030 Agenda for Sustainable Development. In A/RES/70/1.
https://doi.org/10.1007/s13398-014-0173-7.2

70
 N° 11 2024

Vasquez, M., Ségeat, B., Condingley, A., Tilby, E., Wikstrom, S., Ehrnsten, E., Al Hamdani, Z., Agnesi, S., Andresen, M. S.,
Annunziatellis, A., Askew, N., Bekkby, T., Bentes, L., Daniels, E., Doncheva, V., Drakopoulou, V., Ernsten Verner, B., Goncalves,
J., Karavinen, V., … Woods, H. (2023). EUSeaMap 2023, A European broad-scale seabed habitat map, Technical Report.
https://archimer.ifremer.fr/doc/00859/97116/

Verdura, J., Rehues, L., Mangialajo, L., Fraschetti, S., Belattmania, Z., Bianchelli, S., Blanfuné, A., Sabour, B., Chiarore, A.,
Danovaro, R., Fabbrizzi, E., Giakoumi, S., Iveša, L., Katsanevakis, S., Kytinou, E., Nasto, I., Nikolaou, A., Orfanidis, S., Rilov, G.,
… Cebrian, E. (2023). Distribution, health and threats to Mediterranean macroalgal forests: defining the baselines for their
conservation and restoration. Frontiers in Marine Science, 10. https://doi.org/10.3389/fmars.2023.1258842

Wilkinson, M. D., Dumontier, M., Aalbersberg, Ij. J., Appleton, G., Axton, M., Baak, A., Blomberg, N., Boiten, J.-W., da Silva
Santos, L. B., Bourne, P. E., Bouwman, J., Brookes, A. J., Clark, T., Crosas, M., Dillo, I., Dumon, O., Edmunds, S., Evelo, C.
T., Finkers, R., … Mons, B. (2016). The FAIR Guiding Principles for scientific data management and stewardship. Nature
Scientific Data, 3. https://www.nature.com/articles/sdata201618

Wright, D., & Heyman, W. (2008). Introduction to the Special Issue: Marine and Coastal GIS for Geomorphology, Habitat
Mapping, and Marine Reserves. Marine Geodesy, 31, 223–230. https://doi.org/10.1080/01490410802466306

Zapata-Ramírez, P. A., Huete-Stauffer, C., Scaradozzi, D., Marconi, M., & Cerrano, C. (2016). Testing methods to support
management decisions in coralligenous and cave environments. A case study at Portofino MPA. Marine Environmental
Research, 118, 45–56. https://doi.org/10.1016/j.marenvres.2016.04.010

Zhang, Y., Bellingham, J. G., Ryan, J. P., Kieft, B., & Stanway, M. J. (2015). Autonomous Four-Dimensional Mapping and
Tracking of a Coastal Upwelling Front by an Autonomous Underwater Vehicle. Journal of Field Robotics, 33(1), 67–81.
https://doi.org/https://doi.org/10.1002/rob.21617

Zhao, Q., Stephenson, F., Lundquist, C., Kaschner, K., Jayathilake, D., & Costello, M. J. (2020). Where Marine Protected
Areas would best represent 30% of ocean biodiversity. Biological Conservation, 244, 108536. https://doi.org/10.1016/j.
biocon.2020.108536

71
 EMB FUTURE SCIENCE BRIEF

Annex 1

Members of the European Marine Board Working Group

on Marine Habitat Mapping

NAME INSTITUTION COUNTRY

Working Group Chairs

University of Naples Federico II
Simonetta Fraschetti Stazione Zoologica Anton Dohrn (SZN) Italy
National Biodiversity Future Centre

James Strong National Oceanography Centre (NOC) United Kingdom

Working Group Members

Lene Buhl-Mortensen Institute of Marine Research (IMR) Norway

Federica Foglini Institute of Marine Science (CNR-ISMAR) Italy

Jorge M.S. Gonçalves* Centre of Marine Sciences of the Algarve (CCMAR) Portugal

José Manuel González-Irusta Spanish Institute of Oceanography (IEO-CSIC) Spain

Helen Lillis* Joint Nature Conservation Committee (JNCC) United Kingdom

Mats Lindegarth University of Gothenburg Sweden

Georg Martin Estonian Marine Institute, University of Tartu Estonia

The French Research Institute for Exploitation of the Sea

Lenaick Menot France
(Ifremer)

Eimear O’Keeffe* Marine Institute (MI) Ireland

António Pascoal Instituto Superior Técnico (IST), University of Lisbon Portugal

Maria Salomidi Hellenic Centre for Marine Research (HCMR) Greece

Timm Schoening GEOMAR Helmholtz Centre for Ocean Research Kiel Germany

*Member of EMODnet Seabed Habitats Consortium

72
 N° 11 2024

Annex 2

Data driven approaches to the use of distribution models

in marine habitat mapping

TYPE OF OUTPUT AND

DESCRIPTION ADVANTAGES DISADVANTAGES
OBSERVATION EXAMPLES
Presence Models use presence Can be used with Do not offer Frequently used
only data records and pseudo- data from public information on as a proxy of the
absences randomly repositories (e.g. OBIS). prevalence or density. distribution of
generated to replace biogenic habitats such
the lack of directly Some algorithms are Limited proxy to as Desmophyllum
available absences. user friendly biogenic habitat pertusum coral reefs
(i.e. have specific distribution. (Howell et al., 2011)
Predict the suitability software provided). or sea pen fields
of the area for the Sensitive to sampling (Greathead et al.,
target species or bias. 2014).
habitat by measuring
how similar the area
is to the area with the
presence records.

Presence and Correlative models use Provide information on No information on Used as a proxy to the
absence data both presences and prevalence and is less density. distribution of biogenic
absences obtained sensitive to sampling habitats e.g. deep-
from sampling data. bias. Limited proxy to sea sponge grounds,
biogenic habitats (Ramiro-Sánchez et al.,
Predict the probability distribution. 2019).
of finding the species
or community in space. Sometimes real
absences are not
available (e.g. public
repositories).

Abundance of the Abundance models Used to model Data demanding and Usually applied to
target species, measured as biomass essential fish habitats more complex than identify essential
measured as or number of with information on presence/absence fish habitats such as
individuals. aggregations. Good models. spawning grounds
density or biomass
proxy to biogenic (Miesner & Payne,
habitats formed by 2018) or nursery areas
only one species. (Asjes et al., 2016).
Recently used to
model the distribution
of biogenic habitats
(Rodríguez-Basalo et
al., 2022).

Community data, Predict first - Can be applied using The outputs can Burgos et al., (2020)
ranging from only assemblage later: only presence-data generate assemblages used MAXENT and only
presence records First the distribution (thus using data from that do not occur. presence data to model
of indicator species public repositories). the distribution of
of several taxa
(habitat forming Accuracy is not 44 vulnerable marine
to abundance species) is predicted computed for the ecosystem indicator
matrices of several using Community whole process only for species. In a second
species models and presence- each step separately step they analysed
only or presence- (prediction and the co-occurrence of
absence models. assemblages). these species using a
Then the assemblages cluster analysis of the
are computed using predicted maps.
the prediction maps of
these models.
The analysis provided
the predicted
distribution of stacked
species.

73
 EMB FUTURE SCIENCE BRIEF

TYPE OF OUTPUT AND

DESCRIPTION ADVANTAGES DISADVANTAGES
OBSERVATION EXAMPLES
Community data, Assemblage first - Good proxy to Accuracy cannot be Used to model
ranging from only Predict later: biogenic habitats. computed for the the distribution
presence records First, the assemblages Allows modification of whole process only for of epibenthic
are defined by assemblage definition each step separately communities in the
of several taxa
using multivariate to better cover specific (prediction and Gulf of St Lawrence,
to abundance techniques (e.g. cluster biogenic habitats. assemblages). Canada (Moritz et al.,
matrices of several analysis) to analyse the 2013).
species biological samples. The sum of the
probability of all Deep-sea biogenic
Then, the assemblage assemblages does not habitats in the Galicia
distribution is necessarily equal 1. and Seco de los Olivos
modelled using a Banks, Spain (de la
presence/absence Torriente et al., 2019)
approach.

The analysis provides

the distribution of
biological communities
(assemblages)
previously defined
using multivariate
techniques.

Assemblage and Powerful models which Very data demanding Relatively new in
predict together: overcome most of the and technically very marine ecosystems,
Describes assemblages limitations of previous complex. but have been
and predicts their approaches. used extensively in
distribution within terrestrial ecosystems
the same model (Ferrier & Guisan,
framework. 2006).

The output type Joint species

differs slightly distribution models
depending on the type used in the marine
of model used: Joint ecosystems (Murillo et
species distribution al., 2020).
models include co-
occurrence matrices Region common profile
as latent variables to models (Hill et al.,
model community 2017).
data. Region common
profiles delineate
geographic areas
where the probabilities
of observing a group
of species remains
approximately
constant.

74
European Marine Board IVZW
Belgian Enterprise Number: 0650.608.890

Jacobsenstraat 1 I 8400 Ostend I Belgium

Tel: +32 (0)59 56 98 00
E-mail: info@marineboard.eu
www.marineboard.eu