The importance of ecosystems and their services to human well-being and the economy is widely recognised (Burkhard et al. 2010, Haines-Young and Potschin 2010) and various international commitments, including the Kunming-Montreal Global Biodiversity Framework (CBD 2022) , advocate for a system capable of monitoring and quantifying ecosystem changes across spatial and temporal scales. The System of Environmental-Economic Accounting-Ecosystem Accounting (SEEA EA) is an integrated statistical framework for organising biophysical data, measuring ecosystem services, tracking changes in ecosystem extent and condition and linking this information to economic and other human activities (UN 2021).

According to SEEA EA, ecosystem assets are contiguous, mappable areas of a specific ecosystem type, defined by unique biotic and abiotic components and their interactions. An ecosystem accounting area (EAA) is the geographical region where an ecosystem account is compiled, determining which assets are included (UN 2021). Ecosystem assets should be geographically and conceptually exhaustive and exclusive; for example, within a marine EAA, each horizontal depth layer or area of the seafloor should be occupied by only one ecosystem asset, though multiple assets of the same type may exist. Together, extent and condition define the ecosystem´s capacity to provide ecosystem services (UN 2021). The asset´s value is the discounted future exchange of services between the ecosystem and society, linked to ecosystem capacity, which depends on its condition. In the SEEA EA framework, degradation occurs when there is a loss in the asset´s value due to changes in extent or condition. Well-defined extent and condition accounts are essential for accurately determining ecosystem service flows and their value, as they provide a solid basis for assessing the integrity and capacity of ecosystems, which is critical for effective ecosystem management and decision-making.

Studies on ecosystem extent and condition have slowly evolved with the aim to increase the transparency regarding the value ecosystems hold for people and, so far, 35 countries have produced SEEA EA accounts (UN 2023). For instance, Gomez Cardona et al. (2023) developed ecosystem accounts for a wetland ecosystem in Colombia, Gacutan et al. (2022) for a lake ecosystem in Australia, Bruzón et al. (2022) for Spanish terrestrial ecosystems and Alarcon Blazquez et al. (2023) for the North-East Atlantic regional sea. Compared to terrestrial environments where a wealth of data is easily available from Earth Observation and usable for assessing land use and cover, data for marine systems are more difficult to acquire. Concerns have also been raised about the quality of data used in ecosystem accounting (Navarro et al. 2024) and how the uncertainty inherent in marine data, resulting from the dynamic and highly variable nature of marine ecosystems, can undermine the reported accounts and the credibility of resulting management decisions. Accounts should always be based on best available knowledge, which most often is national data of high taxonomic and spatial accuracy, though not necessarily comparable with data from other countries. For instance, in Europe, various habitat classification systems have been developed for reporting requirements of various marine management and conservation policies. Thus, a considerable research gap exists on how ecosystem accounts differ depending on the ecosystem data used and how the reporting requirements of other policies can support or complement ecosystem accounting, including condition indicators.

Amendments to EU Regulation 691/2011 on environmental-economic accounts currently lack mandates for marine ecosystem service reporting, risking their exclusion from key policy and resource allocation decisions. To prevent marine ecosystems from being overlooked as critical societal assets, it is essential to advance accounting methods tailored for marine ecosystems and ecosystem services. This need is further underscored by, for example, the EU Nature Restoration Law, Marine Strategy Framework Directive (MSFD), Habitats Directive (HD), maritime spatial planning initiatives and the Kunming-Montreal Global Biodiversity Framework, all of which could benefit from and have synergies with accurate and comprehensive marine ecosystem accounts.

Here, we report the first pilot accounts of marine ecosystem extent and condition for Finland. Lai et al. (2018) previously linked some ecosystem service indicators to the Finnish ecosystem accounting framework, while Jernberg et al. (2024) compiled comprehensive information on ecosystem services provided by marine habitats in the northern Baltic Sea. Detailed marine habitat maps in Finland are available through comprehensive biological surveys (Forsblom et al. 2024), which are essential for assessing ecosystem services, such as blue carbon that involve complex processes operating at fine spatial scales (Asplund et al. 2021). The SEEA EA provides a useful framework that can support Finland´s national commitments to EU policies and also provide valuable input for national conservation status assessments and for evaluating environmental impacts of development. Similarly, existing EU and national policies can complement and support SEEA EA reporting, reducing the bureaucratic burden. We aim to provide a baseline on ecosystem extents for future studies and show how existing EU policies can support the implementation of the SEEA EA. We base our ecosystem extent evaluation on three ecosystem classifications, usually broadly available across the European Union that support various policy requirements: i) broad habitat types of the MSFD, ii) habitat types of the HD and iii) threatened habitat types of the IUCN Red List of Ecosystems (RLE). We discuss the compatibility of SEEA EA with EU policy reporting requirements, the spatial scale of reporting ecosystem extents, data requirements and challenges, as well as the suitability of indicators.

An account table for SEEA EA compiles spatially referenced data on the assets into a tabulated form and presents aggregated information by ecosystem type. This process necessarily leads to the loss of information on the details and distribution of the spatial dimensions and condition of the assets. However, aggregation is necessary to provide data consistent with the structure and measures used in the system of national accounts. The extent of marine ecosystems can be defined in several ways, depending on the taxonomic unit, spatial scale and ecosystem classification system in use.

Habitat data

Habitat types of the IUCN Red List of Ecosystems

The assessment methodology of IUCN Red List of Ecosystems (RLE) is a global standard for assessing the extinction risk of ecosystems and monitoring their status (Keith et al. 2015). Threatened status of both habitat types and species in Finland were assessed in 2018 and 2019, respectively, using the IUCN RLE criteria (Kontula and Raunio 2018, Hyvärinen et al. 2019 In the assessment, habitats are defined, based on dominant biota, such as habitat forming algae, seagrasses or invertebrates. As such, the RLE habitat types are more detailed than those of the MSFD or HD and include benthic habitat types from the littoral and deeper zones, as well as pelagic (open water) (Fig. 2). Data for these were available from species distribution models (Virtanen et al. 2018), based on comprehensive biological surveys (Forsblom et al. 2024). We include 19 habitats in the present assessment (see Suppl. material 1). By evaluating the extent and condition of habitats based on three classification systems presently in use, this allows a foundation for future ecosystem accounting at appropriate levels. Input data sources for calculating ecosystem extents and their spatial and temporal resolution are shown in Fig. 1.

Figure 1.

Input data sources for calculating ecosystem extents and their spatial and temporal resolution.

Figure 2.

Example maps of ecosystem classification used to report ecosystem extents: (A) Marine Strategy Framework Directive benthic broadscale habitats (MSFD habitats), (B) Habitats Directive habitats (HD habitats) and (C) habitat types used in national threatened status reporting, based on threatened status of Red List of Ecosystems (RLE habitats).

Ecosystem extents

Ecosystem condition indicators

The accounting tables in Suppl. materials 2, 3 provide data of the total ecosystem extent and condition of habitats, along with the proportion of the extent of habitats that could be assessed using the WFD indicators. To facilitate interpretation of results, we present aggregated versions of accounting tables in Tables 1, 2, 3 as well as in Figs 3, 4. SEEA EA offers two alternative presentation types for aggregated ecosystem condition indices and the tables presented here use the presentation method of the table as per paragraph 5.97 of the SEEA EA manual (UN 2021).

Figure 3.

Extent of (A) MSFD habitats, (B) HD habitats and (C) number of extent units for RLE habitats within Finnish marine areas. Each RLE extent unit has a maximum size of 400 m², determined by the used model´s resolution; it is unlikely that the habitat occupies the entire unit.

Figure 4.

Ecosystem condition expressed as the percentage of the total extent of ecosystem types including Marine Strategy Framework Directive broad-scale habitat types (MSFD habitats), Habitats Directive Annex I habitats (HD habitats) and Red List of Ecosystems habitat types (RLE habitats), based on the latest reporting period of Water Framework Directive. ND (Not Defined) refers to WFD areas where status was not assessed.

Shallow, circalittoral, mixed and muddy substrate habitats are the most common MSFD habitats in Finland, as their combined extent exceeds that of all other MSFD habitats. MSFD habitats also cover much larger areas compared to HD habitats, with extents approximately ten times greater (Fig. 3). The RLE habitats are not directly comparable with MSFD and HD habitats, as their extents are reported in extent units, i.e. the number of pixels where each RLE habitat type is predicted to occur. The RLE extent units can be used to calculate RLE extent accounts, but would require assumptions on how large area of the grid cells each habitat covers. The largest RLE habitat is the filamentous annual algae habitat, followed by the red algae habitat, which are predicted to occur in over 12 million and 9.5 million extent units, respectively (Fig. 3).

A large portion of the assessed habitats were in poor condition, regardless of the habitat classification used (Fig. 4). Habitats that occur in the inner archipelago (e.g. estuaries), or closer to the shore in shallow areas (e.g. lagoons or benthic habitats characterised by Hippuris), were found to be in poor condition. In contrast, habitats occurring deeper and in more exposed areas, such as reefs or benthic habitats characterised by red algae, were in better condition. Overall, MSFD habitats were generally in moderate condition, apart from shallow and circalittoral muddy areas.

In this study, we present ecosystem extents for Finland, based on three common classification systems typically available across EU Member States: MSFD benthic habitat types, HD habitats and RLE habitat types. As condition indicators, we used variables reported under the Water Framework Directive, which describe ecosystem condition through ecological status to which biological, ecological and physico-chemical quality assessments contribute.

We found that the majority of habitats in the inner archipelago and shallow nearshore areas were in poor condition, which reflects the eutrophication status of our coastal waters (Syke 2024). Agriculture is the primary source of nutrient load in Finland, except in the Bothnian Bay, where natural runoff and forestry dominate. Despite extensive water protection efforts, achieving good water quality within the next 30 years remains unlikely, partly also due to climate change, which is expected to increase nutrient runoff (Fleming et al. 2023). Across the Baltic Sea countries, eutrophication is the primary pressure affecting marine ecosystems (Gustafsson et al. 2012, Andersen et al. 2015, Meier et al. 2018). Therefore, indicators that monitor eutrophication status (e.g. those under the WFD), are crucial for meeting EU policy requirements, including SEEA EA. WFD indicators are also suggested to be used by the SEEA EA (UN 2021) and they are part of the indicator variables of the Habitats Directive reporting (structure and function parameters) used in Finland. Thus, WFD indicators are useful measures for describing ecosystem condition for SEEA EA. However, as WFD concentrates on coastal areas, the status of offshore areas is not assessed. Many marine habitats extend beyond Finnish territorial waters and require use of additional indicators. A similar situation most likely applies also to other countries. An alternative option could be to use MSFD indicators (i.e. descriptors), but a potential challenge is the extensive geographical scope of the MSFD reporting areas. Finland´s marine areas are divided into six MSFD reporting areas, in contrast to 276 WFD areas, which provide more detailed information on ecosystem condition. A critical aspect of assessing ecosystem condition with WFD indicators is the length of the WFD reporting period. Given the rapid warming of surface waters, which may amplify impacts of eutrophication (Safonova et al. 2024), the condition of marine ecosystems could change significantly over the six-year reporting cycle. Some WFD water quality variables are already monitored with methods that allow annual assessments of ecosystem condition, such as Earth Observation used for measuring chlorophyll a (Attila et al. 2018). These advancements could technically support more frequent reporting of ecosystem condition, even on an annual basis. Additionally, habitat extents may also improve in precision, not necessarily due to habitat degradation, but as new, higher resolution data become available (Misiuk and Brown 2023).

Our results also show that habitat extent can be defined in several ways, depending on habitat classification in use and data availability. This also means that ecosystem accounting results vary, depending on the chosen classification and its spatial scale (e.g. Rahayu et al. (2024)). As the scale of assessed habitats differs, so does the condition assessment; then the question is: what condition indicator should be used? If the definition of ecosystem extents is based on the coarsest taxonomic and spatial scale (in our example, MSFD habitats), then condition indicators will have much lower impact on defining the actual change in habitats. The SEEA EA endorses the IUCN Global Ecosystem Typology (GET) for reporting ecosystem extents in an internationally standardised manner (UN 2021, Keith et al. 2022). The GET comprises six hierarchical levels (Keith et al. 2022), consisting of core realms (e.g. marine), which divide into biomes (e.g. marine shelf biome) and functional groups (e.g. subtidal rocky reefs). While the typology is useful for ensuring comparability between countries, it is rather coarse in terms of taxonomic representation and not fully aligned with national and local reporting requirements, as set out, for instance, by various EU policies and directives. Additionally, it adds to habitat classifications already in use, which are not necessarily in alignment with the reporting standards of ecosystem accounts. Only a small part of marine habitats in Finland are compatible with GET (e.g. seagrass meadows and rocky reefs), missing the diversity of marine habitats. This highlights the need for complementary approaches for defining ecosystem extents and development of crosswalks between the GET and habitat classifications currently in use in various countries.

The SEEA EA framework emphasises both exclusivity and exhaustiveness, meaning each spatial unit should uniquely represent a habitat and collectively cover the entire area. Achieving this in marine environments is challenging, where data precision is limited and habitats often form patchy, overlapping distributions. This limitation complicates the creation of exhaustive ecosystem accounts, especially since only a subset of marine ecosystems may be fully represented, potentially leading to gaps in comprehensive regional or national ecosystem assessments. Given that the majority of habitat information of marine ecosystems relies on some type of modelling as mapping the seabed is expensive, the exclusivity requirement may be a severe hindrance to the development and implementation of SEEA EA in the marine context. This also applies to all habitats that are naturally patchy and biologically diverse, irrespective of the realm in question. In the Finnish context, many RLE habitats form belts at different depth zones, with much finer resolution than what current models can produce (Eriksson and Bergström 2005, Lappalainen et al. 2019). For example, filamentous algae, brown algae and red algae may co-exist within the same 400 m² area, making it difficult to justify selecting one over the other. Despite the limitations of the “extent units,” which are overlapping and non-exhaustive, they can better capture the complex and irregular boundaries of marine ecosystems. This flexibility allows for finer spatial detail, especially in probabilistic models where multiple habitat types may co-occupy an area. In cases of limited data or resources, this approach offers a practical means of describing marine ecosystems and supporting specific management needs.

The requirement for exclusivity and exhaustiveness also strain the functionality of current habitat classification systems, emphasising the need to ensure that they adequately represent all habitats (Virtanen et al. 2018, Rinne et al. 2021). For instance, HD habitats cover only a fraction of marine ecosystems and, being primarily abiotic habitats, they are poor surrogates for marine biodiversity (Virtanen et al. 2018). MSFD and HD reporting rely on exclusive data (MSFD and HD habitats), though some overlap occurs due to Finland´s interpretation of HD habitat definitions. In the national threatened status assessment, spatial data on RLE habitats together with expert knowledge are used and each habitat type is assessed independently; therefore, non-exclusive data are not problematic. However, the non-exhaustive and non-exclusive nature of "extent units" can create challenges for marine governance and management. Overlapping units may lead to double-counting, inflating ecosystem service values, while unrepresented areas cause data gaps that may undervalue certain ecosystem services. Since ecosystem accounting aims to provide a comprehensive and standardised approach across ecosystems (terrestrial and aquatic), these units may hinder comparability with other standardised accounts and complicate national or regional ecosystem account compilation, reducing their effectiveness in supporting integrated environmental policy. Further, as SEEA EA focuses on separate ecosystem types, it does not fully capture the diversity between ecosystems and species. To address this limitation, thematic biodiversity accounts on species and diversity have been proposed to complement the SEEA EA framework (King et al. 2021, Luisetti and Schratzberger 2022). Although it could be possible to evaluate species-based extents (e.g. Virtanen et al. (2023)), reporting their condition would be problematic and comparability between countries impossible due to ecosystem differences.

Our study covered only 19 RLE habitats, missing particularly habitats defined by their fauna. Ecosystem extents of fauna are more challenging to compile and model, as the habitat definition relies on biomass (Kontula and Raunio 2018), for which data are limited. Additionally shoreline habitats, such as reed belts that are known to function as carbon sinks (Gu et al. 2020), were excluded from the present study, as they do not belong to any of the used habitat classifications or assessed as part of habitats on shore. The extent of reeds could be assessed using Earth Observation (e.g. Koponen et al. (2022)), since they have substantial above-water growth. It should also be stressed that RLE habitat types reflect the mean distribution of habitats over the past ~ 20 years. As the models predicting habitat extent already include explanatory variables related to eutrophication status (see Suppl. material 1), certain habitats may not be predicted to occur in WFD areas with poor condition. They can instead be viewed as a baseline to use for comparison in the SEEA EA framework.

Extent and condition accounts form the foundation for defining ecosystem service supply and use accounts, but they do not directly reflect the value ecosystems provide to society. Clear, well-structured extent and condition accounts are essential for understanding the capacity of ecosystems to provide services. However, the supply and use of ecosystem services also depend on factors such as ownership and stewardship of ecosystem assets and the institutional context regulating and promoting the exchange of ecosystem services (Barton 2022). Without well-defined extent and condition accounts, it becomes challenging to accurately map how ecosystems contribute to human well-being. The global biodiversity loss (WWF 2024) and rapidly proceeding climate change (Ripple et al. 2024), threaten not only nature itself, but also world economies (World Economic Forum 2023). Demonstrating and quantifying the monetary value of ecosystem, habitat and population losses is crucial for enhancing acceptance of conservation and climate mitigation efforts between both decision-makers and the public. However, if the reliability of these assessments can be questioned, necessary actions may fall short of the urgency required. Therefore, clarifying the role of different habitat classification systems in evaluating extent and condition is essential for producing credible estimates of ecosystem value. The proposed amendments to the EU Regulation (EU) 691/2011 on environmental economic accounts do not include a mandatory requirement for reporting ecosystem service accounts for marine ecosystems. Consequently, marine ecosystems are overlooked in the methods and tools (European Commission 2023) designed to be used to allocate the supply and use of ecosystem services—such as global and local climate regulation and nature-based tourism—to different ecosystem types. The same also applies to studies on the Baltic Sea ecosystem services (Kuhn et al. 2021). Therefore, it is essential to improve methods related to marine accounting to ensure a sufficient data basis for including marine ecosystem services in mandatory ecosystem accounts in the future. Otherwise, marine ecosystems risk being neglected as critical sources of well-being for society (Storie et al. 2021).

Our work provides the first attempt for marine ecosystem accounting in Finland that is streamlined with other EU policies, thus simplifying and facilitating the reporting under the SEEA EA. The reported data on ecosystem extents are in line with other reporting requirements of the EU and with national policies, such as threatened status assessments. Building on our work, future steps could be to use complementary MSFD indicators on eutrophication for offshore areas. We concentrated only on eutrophication, while, in reality, also other human activities, such as coastal land use or offshore industrial activities, exert pressures on marine ecosystems, degrading their condition. Therefore, additional indicators that measure the integrity of seabed would be important to include for future studies. While SEEA EA aims to provide standardised methods for producing comparable accounts on ecosystems, the standard still offers a plethora of methods for compiling these accounts, which can result in ecosystem accounts that are difficult to compare. Although a large amount of observational data on the distribution of species and habitats exists in Finland, these may not be compatible with higher levels of ecosystem classification, because similar data and (or) methods for constructing extent units may not be available in the same breadth from other regions. The Finnish case serves as an example of relatively detailed level of classification that can be achieved if suitable data exist and it also demonstrates the importance of pursuing accuracy and realism in the accounting framework.

We thank Marco Nurmi for providing us the data of the broad habitat types used and compiled for the most recent MSFD assessment.