Marine environmental vulnerability and cumulative risk profiles to support ecosystem-based adaptive maritime spatial planning

Marine environmental vulnerability and cumulative risk profiles to support ecosystem-based... Abstract Human use of marine and coastal areas is increasing worldwide, resulting in conflicts between different interests for marine space, overexploitation of marine resources, and environmental degradation. In this study we developed a methodology that combines assessments of marine environmental vulnerability and cumulative human pressures to support the processes of ecosystem-based adaptive maritime spatial planning. The methodology is built on the spatially explicit marine environmental vulnerability profile (EVP) that is an aggregated product of the distribution of essential nature values (habitat-forming benthic macroalgal and invertebrate species, benthic species richness, birds and seals as top marine predators) and their sensitivities to disturbances. The marine environmental cumulative risk profile (ERP) combines the EVP and the HELCOM Baltic Sea Pressure Index (BSPI), the latter representing the spatial distribution of intensities of cumulative anthropogenic pressures. The ERP identifies areas where environmental risks are the highest due to both long recoveries of the biota and high intensities of human pressures. This methodology can be used in any other sea areas by modifying the list of nature values, their sensitivity to disturbances, and the intensities of human pressure. Introduction Human use of marine and coastal areas is increasing worldwide, resulting often in conflicts among different interests for the space, overexploitation of marine resources, and environmental degradation. As a consequence of such increasing and diversifying trends of human use, the marine environment, especially intensively used marine areas like the Baltic Sea, is becoming to a greater extent stressed and impacted (Korpinen et al., 2012). The multiple competing uses of marine and coastal areas have resulted in a rapid increase of maritime spatial planning (MSP) initiatives to safeguard sustainable use of marine resources as well as to mitigate cross-sectoral and transboundary conflicts over the use of sea space (Douvere and Ehler, 2010; Stelzenmüller et al., 2015). The MSP Directive 2014/89/EU establishes a framework for MSP aimed at promoting the sustainable growth of maritime economies, the sustainable development of marine areas, and the sustainable use of marine resources. The directive defines the MSP as a process, by which the relevant Member State’s competent authorities analyse and organize human activities in marine areas to achieve ecological, economic, and social objectives (EU, 2014). Cormier et al. (2015) states that “MSP should be seen as the practical implementation of the ecosystem approach to marine management through holistic and integrated analysis of all relevant human activities, pressures, and impacts within the planning area at the ecosystem scale”. In the frame of the MSP, assessments of environmental vulnerability and cumulative risks of human pressures are recognized as important management tools and a range of approaches have been described earlier (Villa and McLeod, 2002; Hiddink et al., 2007; Selkoe et al., 2009; Ardron et al., 2014; Stelzenmüller et al., 2015; La Rivière et al., 2016). Environmental vulnerability assessment is usually based on ecosystem sensitivity to pressures, expressed as resistance and/or recovery potential, i.e. resilience, and exposure to anthropogenic pressures (Bax and Williams, 2001; Hiddink et al., 2007; De Lange et al., 2010; Beroya-Eitner, 2016). Sensitivity has been expressed as the inverse of recovery time for a nature value (NV), e.g. community, habitat, or ecosystem, after being exposed to a pressure (Hiddink et al., 2007), or as a “capacity to tolerate a pressure (resistance) and the time needed to recover after an impact (resilience)” (La Rivière et al., 2016). The actual assessment or quantification of ecosystem sensitivity is difficult and usually it is aimed to produce the least disappointing vulnerability profile with the resources and time available (Villa and McLeod, 2002). The criteria that have been used to quantify the ecosystem sensitivity often include uniqueness or rarity of species or habitat, functional significance (e.g. habitat-forming species), fragility or susceptibility to degradation of species or habitat, and life history traits that are related to recovery potential (Ardron et al., 2014). Expert-derived ratings or rankings based on available scientific publications are widely used to assess intrinsic sensitivity of the NVs and their sensitivities to stressors (Villa and McLeod, 2002; De Lange et al., 2010; Okey et al., 2015; La Rivière et al., 2016). However, the quantitative assessment of ecosystem vulnerability is still challenging (Stelzenmüller et al., 2015) being calculated subjectively with little scientific justification (Villa and McLeod, 2002). Comprehensive environmental vulnerability assessment is pressure-driven and includes exposure, sensitivity to pressure, and recovery of a NV or ecosystem component (De Lange et al., 2010), and is based on best available knowledge (La Rivière et al., 2016). Moreover, most of vulnerability assessments that aim at contributing directly to MSP processes are either regional or national (e.g. Foley et al., 2013; La Rivière et al., 2016) and only seldom performed in a transboundary context (Martin et al., 2009). Risk assessment methods are widely used to assess and grade environmental problems (deFur et al., 2007) but current methods in general are not designed to address the risks of cumulative effects of environmental stressors. However, the U.S. Environmental Protection Agency has developed a general framework for cumulative risk assessment (EPA, 2003). According to this framework, a cumulative risk consists of the combined risks from aggregate exposures to multiple agents or stressors and the cumulative risk assessment means an analysis, characterization, and possible quantification of the combined risks to health or the environment from multiple agents or stressors. In order to develop spatially explicit marine environmental vulnerability and cumulative risk profiles, only the spatially “stable” ecosystem components (species or species groups with low mobility like benthic infauna) can be included. An inclusion of the characteristic habitat forming (e.g. large perennial seaweeds) and/or functionally important (e.g. suspension-feeding mussels) species ensures that the essential spatial proxies of “dynamic” species (highly mobile species like fish) are also included. In many coastal regions, several benthic plant and invertebrate species are considered as habitat engineers or habitat-forming species. They are capable of creating a specific local environment that facilitates colonization of other species that otherwise would not be present in the area (Martin et al., 2013; Koivisto and Westerbom, 2010, 2012). Large perennial seaweeds, mussels, and clams provide food, habitat (or both) to many macroalgae, invertebrates, fish, and birds (e.g. Beukema and Cadée, 1996). Moreover, suspension-feeding mussels clean the water by removing phytoplankton and organic matter and thereby controlling the water quality of the environment (Manganaro et al., 2009). Both seaweeds and benthic invertebrates generate spatial complexity and contribute to the stability of the environment. Consequently, a loss of such a complexity often causes drastic changes in the structure and functioning of the whole coastal ecosystem (Bertocci et al., 2010). Describing marine ecosystems can be challenging, mainly because sampling networks are generally sparse, which leaves most of the area not sampled and therefore with no information. Marine environments are more difficult to access and map compared with terrestrial ecosystems (Robinson et al., 2011; McArthur et al., 2010). Common seabed sampling methods such as grab sampler, trawls, scuba diving, and underwater video and photography yield point-wise data of seabed characteristics (Eleftheriou and McIntyre, 2005). Here, mathematical predictive modelling that is based on species–environment relationships provides a useful framework to extrapolate information from scattered samples into coherent seamless maps of distributions of species, habitats, ecological goods, and services (Guisan and Zimmermann, 2000; Guisan and Thuiller, 2005). The objective of this study was to develop a transparent, science-based and data-driven methodology that combines assessments of marine environmental vulnerability and cumulative human pressures into simple indices to support the processes of ecosystem-based adaptive MSP. Material and methods Study area The tideless Baltic Sea is characterized by a steep salinity gradient resulting in a variable fauna and flora, which tolerates well the prevailing environmental conditions. Material for the study originated from Estonian and Finnish marine areas, located in the north-eastern Baltic Sea (Figure 1). Figure 1. View largeDownload slide Study area and benthos sampling stations. Figure 1. View largeDownload slide Study area and benthos sampling stations. The area includes five major sub-basins of the Baltic Sea: Archipelago Sea, Bothnian Sea, the Northern Baltic Proper, Gulf of Finland, and Gulf of Riga. All of the sub-basins exhibit strong gradients of wave exposure, depth, and salinity. The sea areas west of the islands Saaremaa and Hiiumaa and the southwestern outer Archipelago Sea are exposed to the open Baltic Proper and have a wave fetch of hundreds of kilometres. In contrast, the inner reaches of the bays of the mainland are very sheltered both by the mainland and by islands. Salinity exceeds 7 PSU in the westernmost study area while it falls to almost 0 PSU in the inner parts of bays with riverine inflow (Kautsky and Kautsky, 2000; Karlson et al., 2002; Zettler et al., 2013; Alenius et al., 2016). The patterns of species distribution and species richness in the Baltic Sea follows many environmental gradients, with salinity appearing to be the most influential (Zettler et al., 2013; Snoeijs-Leijonmalm et al., 2017). Data Biological sampling Benthos data The data was compiled from different sources—the macrobenthos database of the Estonian Marine Institute (EMI) of the University of Tartu, the Environmental Information System database (HERTTA) of the Finnish Environment Institute, and the Finnish Inventory Programme for the Underwater Marine Environment (VELMU). The Estonian dataset covered the observations of benthic macrophytes and invertebrates, the HERTTA database contained the benthic invertebrate data and the VELMU database macrophyte data, respectively. Data from 11.523 benthos sampling stations in the Estonian waters and 2.818 benthic invertebrate sampling stations, 27.367 dive transect stations, and 93.363 drop-video stations in the Finnish waters were used as an input for species distribution and species richness models (Figure 1). All the samples were collected in Estonia between 2005 and 2015 and in Finland between 2000 and 2016. The sampling stations covered a depth range of 0.1−193 m in the Estonian waters and 0.1−286 m in the Finnish waters and spanned important environmental gradients like salinity, wave exposure, water transparency, and seabed type. Ekman and Van Veen type bottom grab samplers were used for benthic invertebrate samples on soft sediments in both countries. On hard surfaces, scuba divers collected all the fauna and flora inside a 0.04 m2 metal frame (Kautsky sampler). Sampling and sample analysis followed the guidelines developed for the HELCOM COMBINE programme (HELCOM, 2015). In Finland, macrophyte species composition was recorded by mapping a diver’s inspection square, typically 4 m2, or a video clip covering 20 m2 for every sampling station. Birds and seals Spatial data on birds was based on an aerial mapping and modelling by Luigujõe and Auniņš (2016) and integrated information on the density of benthos feeders, fish feeders, gulls, and swans during winter season in Estonia (Table 1). No bird data were available for Finnish area. Table 1. Summary of bird species/groups included in the study, from the Luigujõe and Auniņš (2016) models. Benthos feeders Fish feeders Gulls and swans Clangula hyemalis Gavia sp. Larus sp. Bucephala clangula Gavia stellate Larus minutus Somateria mollissima Mergus serrator Larus canus Polysticta stelleri Mergus merganser Larus argentatus Melanitta nigra Mergus albellus Cygnus sp. Melanitta fusca Phalacrocorax carbo Aythya fuligula Aythya marila Benthos feeders Fish feeders Gulls and swans Clangula hyemalis Gavia sp. Larus sp. Bucephala clangula Gavia stellate Larus minutus Somateria mollissima Mergus serrator Larus canus Polysticta stelleri Mergus merganser Larus argentatus Melanitta nigra Mergus albellus Cygnus sp. Melanitta fusca Phalacrocorax carbo Aythya fuligula Aythya marila Table 1. Summary of bird species/groups included in the study, from the Luigujõe and Auniņš (2016) models. Benthos feeders Fish feeders Gulls and swans Clangula hyemalis Gavia sp. Larus sp. Bucephala clangula Gavia stellate Larus minutus Somateria mollissima Mergus serrator Larus canus Polysticta stelleri Mergus merganser Larus argentatus Melanitta nigra Mergus albellus Cygnus sp. Melanitta fusca Phalacrocorax carbo Aythya fuligula Aythya marila Benthos feeders Fish feeders Gulls and swans Clangula hyemalis Gavia sp. Larus sp. Bucephala clangula Gavia stellate Larus minutus Somateria mollissima Mergus serrator Larus canus Polysticta stelleri Mergus merganser Larus argentatus Melanitta nigra Mergus albellus Cygnus sp. Melanitta fusca Phalacrocorax carbo Aythya fuligula Aythya marila Information provided by the Estonian Nature Information System (EELIS, 2017) on nationally protected moulting, resting, or breeding areas of seals was used as an additional input layer when calculating the EVP and ERP values at the Estonian side of the study area. There are two seal species in the study area: grey seal (Halichoerus grypus) and ringed seal (Pusa hispida). Sensitivity of nature values It is widely known that there exists a strong relationship between the health of natural environment and services these ecosystems provide to human wellbeing. Therefore, it becomes interesting and useful to develop a methodology that advises environmental managers how to minimize conflicts between different interests for marine space, reduce overexploitation of marine resources, and diminish environmental degradation. The sensitivity of the environment is defined by an interplay of natural settings (i.e. the biota) and the intensities and diversity of pressures. The biota represents the basement of the sensitivity assessments as the existing NVs largely define vulnerability of the environment to human pressures. In order to construct a holistic judgement on the sensitivities of NVs, a wide range of ecosystem elements should be incorporated into the analysis. As the MSP framework has a clear spatial dimension, it is practical to include among NVs only the spatially “stable” ecosystem components such as characteristic habitats (e.g. macroalgal beds) and/or functionally important species (e.g. ecoengineers). The inclusion of the latter ensures that the essential spatial proxies of “dynamic” ecosystem components (e.g. plankton, fish) are also included. Assessing pressure-specific sensitivity is very challenging because different pressures impact the marine environment simultaneously. Moreover, the magnitude of the cumulative impact is further modulated by different environmental variables such as salinity, depth, hydrodynamic activity, etc. To date there is a lack of such empirical knowledge to quantitatively formalize species sensitivity as functions of environmental variables. A practical approach to this complex problem can be the use of the recovery potential of an environmental value that is measured in time that is needed to recover from a destruction after its impact has ceased. For example, a reefs habitat type with ephemeral algae would recover (given that hard substrate is still present) within one year or a growing season because the spores of ephemeral algal species are produced in abundance and they disperse and colonize efficiently available habitats. However, the recovery of a reef habitat type with bladder wrack community would require 2−3 years. A ringed seal population would need more than 10 years to recover. Although the exact time needed for recovery is difficult to estimate and depends on the prevailing environmental conditions, generalizations based on known biological parameters of species or taxonomic groups can be utilized as a basis for such analyses. The Estonian and Finnish experts selected ten important benthic macroalgal, invertebrate, bird, and seal species (or groups of species) with different ecosystem functions and recovery potentials to represent NVs in this study (Table 2). The selection of NVs was based on their ecological importance (e.g. habitat forming species, top predators) and on the data availability. The recovery estimations were based on the earlier project results (e.g. Aps et al., 2011), expert opinions and on the literature (Appendix Table A3), combining relevant life history traits, observed time of recoveries and/or (re)colonization capacity of species in the Baltic Sea and/or similar areas. NVs were divided into five coefficients according to their recovery potential (time needed for recovery) to provide optimal differentiation between rapidly recovering annual filamentous algal species, more slowly recovering perennial algal species, benthic fauna and vascular plants, and very slowly recovering vertebrates. Table 2. Species and groups of species chosen to represent NVs with their recovery classes and coefficient for the further calculations according to the recovery class. Species/group Recovery class (years) Sensitivity coefficient Fucus vesiculosus 2–3 2 Furcellaria lumbricalis 5–10 4 Filamentous algae <2 1 Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3–5 3 Vascular plants excl. Zostera marina 3–5 3 Zostera marina >10 5 Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 Seals >10 5 Birds >10 5 Species/group Recovery class (years) Sensitivity coefficient Fucus vesiculosus 2–3 2 Furcellaria lumbricalis 5–10 4 Filamentous algae <2 1 Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3–5 3 Vascular plants excl. Zostera marina 3–5 3 Zostera marina >10 5 Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 Seals >10 5 Birds >10 5 Higher sensitivity coefficient indicates longer recovery time. For more detailed information see Appendix Table A3. Table 2. Species and groups of species chosen to represent NVs with their recovery classes and coefficient for the further calculations according to the recovery class. Species/group Recovery class (years) Sensitivity coefficient Fucus vesiculosus 2–3 2 Furcellaria lumbricalis 5–10 4 Filamentous algae <2 1 Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3–5 3 Vascular plants excl. Zostera marina 3–5 3 Zostera marina >10 5 Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 Seals >10 5 Birds >10 5 Species/group Recovery class (years) Sensitivity coefficient Fucus vesiculosus 2–3 2 Furcellaria lumbricalis 5–10 4 Filamentous algae <2 1 Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3–5 3 Vascular plants excl. Zostera marina 3–5 3 Zostera marina >10 5 Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 Seals >10 5 Birds >10 5 Higher sensitivity coefficient indicates longer recovery time. For more detailed information see Appendix Table A3. The selected benthic taxa can be considered as habitat-forming species that are a precondition or promote the existence of other species that otherwise would not be present in the area (Martin et al., 2013). The selected benthic algae serve as a spawning ground for economically important fish species like the Baltic herring (Rajasilta et al., 2006) and support a high biomass of invertebrates (Wikström and Kautsky, 2007). They can also be extensively consumed by waterfowl, thus forming a substantial component of the foodweb (Schmieder et al., 2006). The suspension feeding bivalves form a very important trophic link between pelagic and benthic systems (Lauringson et al., 2009; Koivisto and Westerbom, 2010, 2012) and maintain self-purification and water quality in marine coastal ecosystems. Fish and plankton were excluded in the current EVP and ERP assessments because of high spatial and temporal variability of these ecosystem components, and a lack of harmonized georeferenced data. NVs also included total species richness of benthic macroalgae and invertebrates calculated according to Peterson and Herkül (2017). Abiotic environment data The abiotic environmental variables included different bathymetric (depth, slope of seabed, topographic position), hydrodynamic (wave exposure, currents), geological (seabed substrate), and physical–chemical (temperature, salinity, transparency, nutrients, ammonium, ice conditions) variables. Altogether 18 Estonian and 23 Finnish environmental variables (Appendix Tables A1and A2) were used in the modelling. The resolution of the GIS layers was 100 m for Estonia and 20 m for Finland. Modelling Common marine monitoring methods yield point-wise data of seabed characteristics and in order to extrapolate information from such scattered samples into coherent seamless maps of distributions of species and habitats spatial modelling is needed. In this study the boosted regression trees (BRT) models were used to predict the spatial distribution of all benthic NVs, including benthic species richness, in the study area. BRT is an ensemble machine learning method that combines the strength of two algorithms: regression trees and boosting (Elith et al., 2008). In the BRT models the tree complexity was set to 5 for Estonian and 7 or 9 for Finnish models. The learning rates of Estonian and Finnish models were set to 0.005 and 0.01, respectively. The bag fraction was set at 0.5 which is the recommended default value (Elith et al., 2008). Modelling was done in the statistical software R version 3.3.1 (R Core Team, 2015) using the packages gbm (Ridgeway, 2007) and dismo (Elith and Leathwick, 2017). Finnish benthic species richness was calculated by combining invertebrate and macrophyte species richnesses modelled from taxa observed within every sampling station (invertebrates) or 200 m wide grid cell (macrophytes) using R package randomForest (Liaw and Wiener, 2002). Random forest (RF) is a machine learning method that similar to BRT modelling does not use any predefined data model but instead uses an algorithm to analyse the relationship between the biota and environmental variables. The RF generates a large number of regression trees, each calibrated on a bootstrap sample of the original data. Calibrated models were then used to predict benthic species richness and the probability of occurrence of NVs from the coastline to the outer border of the exclusive economic zone with a grid size of 100 m in Estonia and 20 m in Finland. Calculation of marine environmental vulnerability and cumulative risk profiles The HELCOM Baltic Sea Pressure Index (BSPI) (HELCOM, 2017) represented the intensity of cumulative anthropogenic pressure at the 1 × 1 km grid resolution in the study area. The BSPI incorporates multitude of human pressures weighed by their potential impacts on ecosystem. Environmental vulnerability profile (EVP) was calculated as a sum aggregation of all NVs that were first rescaled between 0 and 1 (by dividing with maximum value) and then weighed by NV-specific sensitivity coefficient (see Table 2). The ERP was a multiplication product of EVP and BSPI. The general scheme of the index calculation is shown in Figure 2. Figure 2. View largeDownload slide The general scheme of calculations of environmental vulnerability (EVP) and risk (ERP) profiles. Figure 2. View largeDownload slide The general scheme of calculations of environmental vulnerability (EVP) and risk (ERP) profiles. As the BSPI has been calculated at coarser spatial resolution than NVs used in this study, the spatial resolution of all NVs were upscaled to 1 × 1 km grid cells to match the BSPI resolution. When upscaling NVs, maximum values in each grid cell were used as it (1) underscored the most prominent NVs in grid cells (i.e. the precautionary principle not to mask the presence of high NVs by using mean values) and (2) better revealed spatial differences in NVs. The calculation of EVP and ERP included several steps (see Figure 2) all of which were proceeded in each 1 km grid cell. Benthic species richness was natural logarithm transformed to reduce the variation and divided by the maximum logarithmed richness value over all cells to make the values vary between 0 and 1. Then, to the product of logarithm a value of 1 was added to eliminate zero-values that would render further multiplication operations to zero (Formula 1). ln(richness+1)/max(ln(richness+1))+1 (1) The probability of occurrence of each benthic NV (excluding benthic richness) was multiplied by its respective sensitivity coefficient which was determined for each species or species group separately according to their recovery time (Table 2, Appendix A3). After that, all multiplication products within a grid cell were averaged. The averaged value was then multiplied by the transformed benthic species richness. Similar to benthic NVs, bird and seal NVs were also multiplied by their respective sensitivity coefficients. The resulting benthos, bird and seal products were averaged and then rescaled by dividing each value by its maximum value over all cells to make the values vary between 0 and 1. This rescaled product of the distribution of NVs and their sensitivities was termed as EVP-F. EVP-F was calculated for Estonian sea area only. Due to the lack of comparable data on birds in Finnish sea areas, an alternative index was developed that included only benthic NVs, hereafter termed as the EVP-B. The EVP-B was calculated for both Estonian and Finnish sea areas. The cumulative ERP calculation was based on the following basic assumptions: (1) the level of cumulative human pressure (BSPI) is proportional to the likelihood of potential environmental cumulative effect events and (2) the level of environmental vulnerability is proportional to potential consequences of environmental cumulative effect events. In order to calculate the ERP in the study area, the BSPI was divided by its maximum value over all cells to make the values vary between 0 and 1. Then the rescaled BSPI was multiplied with EVP and divided by the maximum value of such multiplication term over all grid cells to make the values vary between 0 and 1 (Formula 2). (EVP×BSPI/max(BSPI))/max(EVP×BSPI/max(BSPI)) (2) Similar to the EVP, when calculations were based on EVP-F the index was termed ERP-F and in case calculations were based on EVP-B, the index was termed ERP-B. Results Marine environmental vulnerability and risk profiles Two different versions of the EVP profiles are shown in Figure 3. EVP-B had the highest values in the Finnish Archipelago Sea and in the western coasts of Estonia and around its islands. In general, coastal areas had higher EVP-F values compared with offshore areas. The highest values of EVP-F coincided with nationally protected moulting, resting, or breeding areas of seals in the Estonian sea waters. This is because these areas are often characterized by high benthic richness and the presence of other NVs. Figure 3. View largeDownload slide (a) Environmental vulnerability profile based on benthic nature values (EVP-B) and (b) environmental vulnerability profile including benthic nature values, birds and seals (EVP-F) in Estonian sea waters. Values vary between 0 and 1, where 1 expresses the highest vulnerability. Figure 3. View largeDownload slide (a) Environmental vulnerability profile based on benthic nature values (EVP-B) and (b) environmental vulnerability profile including benthic nature values, birds and seals (EVP-F) in Estonian sea waters. Values vary between 0 and 1, where 1 expresses the highest vulnerability. The two different versions of ERP (ERP-B and ERP-F) are shown in Figure 4. ERP-B had the highest values in the Finnish Archipelago Sea and close to the City of Helsinki. In the Estonian waters, the highest environmental risk values were characteristic to the coastal sea areas around the islands and close to the city of Tallinn. Similar to EVP layers, higher values of ERP-F coincided with seal areas. In addition, ERP-F had higher values adjacent to the City of Tallinn and along the northern coasts of Estonia. Figure 4. View largeDownload slide (a) Environmental risk profile based on benthic nature values (ERP-B) and (b) environmental risk profile that including benthic nature values, birds and seals (ERP-F) in Estonian waters. Risk values vary between 0 and 1, where 1 expresses the highest risk value. Figure 4. View largeDownload slide (a) Environmental risk profile based on benthic nature values (ERP-B) and (b) environmental risk profile that including benthic nature values, birds and seals (ERP-F) in Estonian waters. Risk values vary between 0 and 1, where 1 expresses the highest risk value. Discussion Environmental vulnerability Here, we described a transparent, science-based, and data-driven methodology that summarizes vulnerabilities of different NVs and risks of different human pressures as a single map. Such map can be easily communicated to MSP experts and other interested stakeholders to jointly analyse and compare the potential environmental risk levels resulting from different planning solutions. While it is impossible to include all the species and habitats to the assessment of ecosystem sensitivity, acknowledging what is valuable in the environment is undoubtedly important in environmental management planning (Villa and McLeod, 2002). We included the spatially “stable” ecosystem components such as the characteristic habitat forming species (e.g. large perennial seaweeds), functionally important organisms (e.g. suspension-feeding mussels) and top predators (birds, seals) into our analyses. Due to strong interlinkage of different ecosystem components our choice of NVs ensures that the essential spatial proxies of “dynamic” species are also included. For example, the presence or absence of habitat forming benthic species impact all trophic levels from phytoplankton to seabirds either modifying seabed structure and/or food web (e.g. Beukema and Cadée, 1996) and therefore can be regarded as a proxy for other groups of organisms not used in the current assessment. Similarly, the distribution of seabed biota has a strong linkage to fish because there are certain essential fish habitats (spawning, feeding) that are primarily defined by the benthic macroalgae, vascular plant, and invertebrate communities. There is a generic lack of empirical knowledge to quantify the sensitivities of species and habitats to different human pressures along environmental variables (Villa and McLeod, 2002). Moreover, pressure-specific sensitivities are very challenging to be assessed because different pressures impact the environment simultaneously. A practical solution to this complex problem can be alleviated by using the expert-based and/or empirically derived recovery potential of NV that is measured in time that is needed to recover from a disturbance after its effect has ceased. Recovery time has been suggested also by Hiddink et al., (2007) as useful parameter in estimating habitat sensitivity and has been later used by Ardron et al., (2014) and Stelzenmüller et al., (2015). EVP is not meant to reflect sensitivity to any specific pressures; it only sums up the spatial distribution of NVs for which we have the distribution data and knowledge on their recovery rates. Different pressures may have specific impacts on different ecosystem components mediated by natural abiotic gradients (e.g. depth, seabed substrates, currents, salinity) and EVP was not expected to elucidate such details. Ideally, detailed impact assessments should be carried out during environmental impact assessments (EIA) of concrete projects. EVP is meant to help maritime spatial planners as a starting point for allocating marine areas that host many NVs. Environmental cumulative risk The BSPI (HELCOM, 2010, 2017) is defined as the “straightforward measure of the geographical distribution and intensity of anthropogenic pressure on the Baltic Sea marine environment”. The BSPI approach allows calculation of cumulative pressure by summing the pressure intensities in the area without including ecosystem data into the calculations. There exist diverse pathways of the pressure–biota relationships and only empirical data supporting cause-effect pathways enable a detailed and highly accurate estimation of the effects of specific anthropogenic pressures on ecosystem components. In the real world we face the situation where we lack such detailed knowledge but we still need generalizations for large-scale management tasks, e.g. for MSP. The BSPI is a product that has been jointly developed by the Baltic Sea states in order to provide a large-scale estimation of the cumulative human pressures and it summarizes the best available scientific and expert knowledge. When joining assessments of marine environmental vulnerability (EVP) and cumulative human pressures (BSPI), the marine cumulative ERP can be produced. The ERP identifies spatial areas of highest risk based on the likelihood and magnitude of environmental vulnerability and human-induced pressures. For example, high levels of environmental cumulative risk are situated near to the largest cities. Coastal sea usually supports the highest species richness (Peterson and Herkül, 2017), but at the same time sustain the strongest human pressure. Here, differences in the environmental cumulative risk value can guide planners toward the most environment-friendly planning solutions. However, establishment of environmental cumulative risk-related tolerability levels require extensive consultations with regulators, stakeholders, and the public in order to determine the level of risk that is acceptable to all stakeholders. It is stated (ICES, 2014) that “Given that a scientific assessment is objective and is based on facts, it would simply reflect likelihood and magnitude leaving the severity, tolerability or values to the governance decision-making processes and stakeholder constituency”. Conclusions In this study we developed a methodology that combines assessments of marine environmental vulnerability (EVP) and cumulative human pressures to produce the marine cumulative ERP. The marine EVP was calculated as a spatial data layer incorporating the distribution of essential NVs (habitat-forming benthic macroalgal and invertebrate species, benthic species richness, seals, and birds) and their recovery time to disturbances. The ERP was calculated as a georeferenced data layer by linking the EVP and the HELCOM BSPI, the latter representing the spatial distribution of the intensity of cumulative anthropogenic pressures. The developed spatially explicit profiles enable the development of environmentally sustainable MSP solutions. The EVPs visualize the spatial patterns of marine NVs and their sensitivities to different human pressures in terms of their recovery potential. The EVP and ERP enable maritime spatial planners to compare the environmental cumulative risk of different planning solutions and thereby to overcome the major environmental challenges faced by any highly impacted marine ecosystem. The suggested methodology can be used in any other sea areas by modifying the list of NVs, their sensitivity to disturbances and the level of human pressure. Moreover, the EVP and ERP are applicable at various scales—from a single bay to large marine areas. Acknowledgements This study is supported by European Regional Development Fund, INTERREG Central Baltic project Plan4Blue “Maritime Spatial Planning for Sustainable Blue Economies”, the Estonian Environmental Investment Centre, the Estonian Research Council (Institutional research funding, IUT02-20), and the European Union’s Seventh Programme for research, technological development, and demonstration BONUS project MARES “Multi-method Assessment for Resilient Ecosystem Services and human-nature system integration”. The authors thank the Working Group for Marine Planning and Coastal Zone Management (WGMPCZM) of the International Council for the Exploration of the Sea (ICES) for facilitating this research. References Alenius P. , Myrberg K. , Roiha P. , Lips U. , Tuomi L. , Pettersson H. , Raateoja M. 2016 . Gulf of Finland physics. In The Gulf of Finland Assessment, No. 27 . 42 pp. Ed. by Raateoja M. , Setälä O. . Reports of the Finnish Environment Institute , Helsinki, Finland . Aps R. , Ambjörn C. , Fetissov M. , Karjalainen M. , Kotta J. , Kuikka S. 2011 . OILRISK web: advanced tool for enhancing spill response decision-making in the Baltic Sea. In Oil Spill Risk Management , pp. 109 – 117 . Ed. by Bellefontaine N. , Lindén O. . The Proceedings of the International Conference on Oil Spill Risk Management, Mamlö, Sweden . Ardron J. A. , Clark M. R. , Penney A. J. , Hourigan T. F. , Rowden A. A. , Dunstan P. K. , Watling L. et al. 2014 . A systematic approach towards the identification and protection of vulnerable marine ecosystems . Marine Policy , 49 : 146 – 154 . Google Scholar CrossRef Search ADS Austin A. P. 1960 . Observations on the growth, fruiting and longevity of Furcellaria fastigiata (L.) Lamouroux . Hydrobiologia , 15 : 193 – 207 . Google Scholar CrossRef Search ADS Bax N. J. , Williams A. 2001 . Seabed habitat on the southeastern Australian continental shelf: context, vulnerability and monitoring . Marine and Freshwater Research , 52 : 491 – 512 . Google Scholar CrossRef Search ADS Bendtsen J. , Gustafsson K. E. , Söderkvist J. , Hansen J. L. S. 2009 . Ventilation of bottom water in the North Sea–Baltic Sea transition zone . Journal of Marine Systems , 75 : 138 – 149 . Google Scholar CrossRef Search ADS Beroya-Eitner M. A. 2016 . Ecological vulnerability indicators . Ecological Indicators , 60 : 329 – 334 . Google Scholar CrossRef Search ADS Bertocci I. , Arenas F. , Matias M. , Vaselli S. , Araújo R. , Abreu H. , Pereira R. et al. 2010 . Canopy-forming species mediate the effects of disturbance on macroalgal assemblages on Portuguese rocky shores . Marine Ecology Progress Series , 414 : 107 – 116 . Google Scholar CrossRef Search ADS Beukema J. J. , Cadée G. C. 1996 . Consequences of the sudden removal of nearly all mussels and cockles from the Dutch Wadden Sea . Marine Ecology , 17 : 279 – 289 . Google Scholar CrossRef Search ADS Beukema J. J. , Flach E. C. , Dekker R. , Starink M. 1999 . A long-term study of the recovery of the macrozoobenthos on large defaunated plots on a tidal flat in the Wadden Sea . Journal of Sea Research , 42 : 235 – 254 . Google Scholar CrossRef Search ADS Bird N. L. , Chen L.,C.-M. , McLachlan J. 1979 . Effects of temperature, light and salinity of growth in culture of Chondrus crispus, Furcellaria lumbricalis, Gracilaria tikvahiae (Gigartinales, Rhodophyta), and Fucus serratus (Fucales, Phaeophyta) . Botanica Marina , 22 : 521527. Google Scholar CrossRef Search ADS Boedeltje G. , Smolders A. J. P. , Roelofs J. G. M. , Van Groenendael J. M. 2001 . Constructed shallow zones along navigation canals: vegetation establishment and change in relation to environmental characteristics . Aquatic Conservation: Marine and Freshwater Ecosystems , 11 : 453 – 471 . Google Scholar CrossRef Search ADS Boström C. , Baden S. , Bockelmann A.-C. , Dromph K. , Fredriksen S. , Gustafsson C. , Krause-Jensen D. , et al. 2014 . Distribution, structure and function of Nordic eelgrass (Zostera marina) ecosystems: implications for coastal management and conservation . Aquatic Conservation: Marine and Freshwater Ecosystems , 24 : 410 – 434 . Google Scholar CrossRef Search ADS PubMed Cormier R. , Kannen A. , Elliott M. , Hall P. 2015 . Marine spatial planning quality management system. ICES Cooperative Research Report No. 327, 106 pp. deFur P. L. , Evans G. W. , Hubal E. A. C. , Kyle A. D. , Morello-Frosch R. A. , Williams D. R. 2007 . Vulnerability as a function of individual and group resources in cumulative risk assessment . Environ Health Perspect , 115 : 817 – 824 . Google Scholar CrossRef Search ADS PubMed De Lange H. J. , Sala S. , Vighi M. , Faber J. H. 2010 . Ecological vulnerability in risk assessment—A review and perspectives . Science of the Total Environment , 408 : 3871 – 3879 . Google Scholar CrossRef Search ADS PubMed Douvere F. , Ehler C. N. 2010 . The importance of monitoring and evaluation in adaptive maritime spatial planning . Journal of Coastal Conservation , 15 : 305 – 311 . Google Scholar CrossRef Search ADS EELIS . 2017 . Estonian nature information system —Environmental register. Environment Agency, Republic of Estonia. Accessed 19 January 2017. Eleftheriou A. , McIntyre A. D. 2005 . Methods for the Study of Marine Benthos , 3rd edn . Blackwell , Oxford . Elith J. , Leathwick J. R. , Hastie T. 2008 . A working guide to boosted regression trees . Journal of Animal Ecology , 77 : 802 – 813 . Google Scholar CrossRef Search ADS PubMed Elith J. , Leathwick J. R. 2017 . Boosted regression trees for ecological modeling. https://cran.r-project.org/web/packages/dismo/vignettes/brt.pdf. Accessed 18 August 2017 EPA . 2003 . Framework for cumulative risk assessment. EPA/600/P-02/001F. National Center for Environmental Assessment, Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. EU . 2014 . Establishing a framework for maritime spatial planning, Directive 2014/89/EU, European Council, Brussels. Foley M. M. , Armsby M. H. , Prahler E. E. , Caldwell M. R. , Erickson A. L. , Kittinger J. N. , Crowder L. B. et al. 2013 . Improving ocean management through the use of ecological principles and integrated ecosystem assessments . Bioscience , 63 : 619 – 631 . Google Scholar CrossRef Search ADS Green A. J. , Figuerola J. , Sánchez M. I. 2002 . Implications of waterbird ecology for the dispersal of aquatic organisms . Acta Oecologica , 23 : 177 – 189 . Google Scholar CrossRef Search ADS Greve T. M. , Krause-Jensen D. , Rasmussen M. B. , Christensen P. B. 2005 . Means of rapid eelgrass (Zostera marina L.) recolonisation in former dieback areas . Aquatic Botany , 82 : 143 – 156 . Google Scholar CrossRef Search ADS Guisan A. , Thuiller W. 2005 . Predicting species distribution: offering more than simple habitat models . Ecology Letters , 8 : 993 – 1009 . Google Scholar CrossRef Search ADS Guisan A. , Zimmermann N. E. 2000 . Predictive habitat distribution models in ecology . Ecological Modelling , 135 : 147 – 186 . Google Scholar CrossRef Search ADS HELCOM . 2010 . Towards a tool for quantifying anthropogenic pressures and potential impacts on the Baltic Sea marine environment. A background document on the method, data and testing of the Baltic Sea Pressure and Impact Indices. Baltic Sea Environment Proceedings No, 125. HELCOM . 2015 . Manual for marine monitoring in the COMBINE programme of HELCOM. http://www.helcom.fi/action-areas/monitoring-and-assessment/manuals-and-guidelines/combine-manual (last accessed 28 December 2017). HELCOM . 2017 . The assessment of cumulative impact using the BSPI and BSPII: Supplementary report to the first version of the ‘State of the Baltic Sea’ Report 2017. http://stateofthebalticsea.helcom.fi/wpcontent/uploads/2017/09/HELCOM_The_assessment_of_cumulative_impacts_Supplementary_report_first_version_2017.pdf (last accessed 2 January 2018) Hiddink J. G. , Jennings S. , Kaiser M. J. 2007 . Assessing and predicting the relative ecological impacts of disturbance on habitats with different sensitivities . Journal of Applied Ecology , 44 : 405 – 413 . Google Scholar CrossRef Search ADS Holt T. J. , Hartnoll R. G. , Hawkins S. J. 1995 . Sensitivity and vulnerability to man induced change of selected communities: intertidal brown algal shrubs Zostera beds and Sabellaria spinulosa reefs . English Nature. ENRR No. 234 , 97. pp. ICES . 2014 . Report of the Joint Rijkswaterstaat/DFO/ICES Workshop: Risk Assessment for Spatial Management (WKRASM), 24–28 February 2014, Amsterdam, the Netherlands. ICES CM 2014/SSGHIE: 01. 35 pp. Karlson K. , Rosenberg R. , Bonsdorff E. 2002 . Temporal and spatial large-scale effects of eutrophication and oxygen deficiency on benthic fauna in Scandinavian and Baltic waters – a review . Oceanography and Marine Biology, an Annual Review , 40 : 427 – 489 . Kautsky N. 1982 . Growth and Size Structure in a Baltic Mytilus edulis Population . Marine Biology , 68 : 117 – 133 . Google Scholar CrossRef Search ADS Kautsky N. 1982 . Quantitative Studies on Gonad Cycle, Fecundity, Reproductive Output and Recruitment in a Baltic Mytilus edulis Population . Marine Biology , 68 : 143 – 160 . Google Scholar CrossRef Search ADS Kautsky L. 1988 . Life strategies of soft bottom macrophytes . Oikos , 53 : 126 – 135 . Google Scholar CrossRef Search ADS Kautsky L. , Kautsky N. 2000 . The Baltic Sea, including Bothnian Sea and Bothnian Bay. In Seas at the Millennium: An Environmental Evaluation , pp. 121 – 133 . Ed. by Sheppard C. . Elsevier , Amsterdam . Koivisto M. E. , Westerbom M. 2010 . Habitat structure and complexity as determinants of biodiversity in blue mussel beds on sublittoral rocky shores . Marine Biology , 157 : 1463 – 1474 . Google Scholar CrossRef Search ADS Koivisto M. E. , Westerbom M. 2012 . Invertebrate communities associated with blue mussel beds in a patchy environment: a landscape ecology approach . Marine Ecology Progress Series , 471 : 101 – 110 . Google Scholar CrossRef Search ADS Korpinen S. , Meski L. , Andersen J. H. , Laamanen M. 2012 . Human pressures and their potential impact on the Baltic Sea ecosystem . Ecological Indicators , 15 : 105 – 114 . Google Scholar CrossRef Search ADS Kostamo K. , Mäkinen A. 2006 . Observations on the mode and seasonality of reproduction in Furcellaria lumbricalis (Gigartinales, Rhodophyta) populations in the northern Baltic Sea . Botanica Marina , 49 : 304 – 309 . Google Scholar CrossRef Search ADS Kraufvelin P. , Ruuskanen A. T. , Nappu N. , Kiirikki M. 2007 . Winter colonisation and succession of filamentous macroalgae on artificial substrates and possible relationships to Fucus vesiculosus settlement in early summer . Estuarine, Coastal and Shelf Science , 72 : 665 – 674 . Google Scholar CrossRef Search ADS La Rivière M. , Aish A. , Gauthier O. , Grall J. , Guérin L. , Janson A.-L. , Labrune C. et al. 2016 . Assessing benthic habitats’ sensitivity to human pressures: a methodological framework - Summary report. Rapport SPN 2016-87. MNHN. Paris, 42. pp. Lauridsen T. L. , Jeppesen E. , Sondergaard M. 1994 . Colonization and succession of submerged macrophytes in shallow Lake Vaeng during the first five years following fish manipulation . Hydrobiologia , 275-276 : 233 – 242 . Google Scholar CrossRef Search ADS Lauringson V. , Kotta J. , Orav-Kotta H. , Kotta I. , Herkül K. , Põllumäe A. 2009 . Comparison of benthic and pelagic suspension feeding in shallow water habitats of the Northeastern Baltic Sea . Marine Ecology , 30 : 43 – 55 . Google Scholar CrossRef Search ADS Liaw A. , Wiener M. 2002 . Classification and regression by random forest . R News , 2 : 18 – 22 . Luigujõe L. , Auniņš A. 2016 . Talvituvate lindude rahvusvaheline lennuloendus. Report. https://www.keskkonnaamet.ee/sites/default/files/luigujoeaunins_2016_talvituvate_veelindude-rahvusvaheline_lennuloendus_lopparuanne_rev01-2017.pdf (last accessed 7 December 2017). Mackie G. L. , Gibbons W. N. , Muncaster B. W. , Gray I. M. 1989 . The Zebra Mussel, Dreissena Polymorpha: A Synthesis of European Experiences and a Preview for North America . Report prepared for Water Resources Branch, Ontario Ministry of Environment , Canada . ISBN 0-7729-5647-2. Maljutenko I. , Raudsepp U. 2014 . Validation of GETM model simulated long-term salinity fields in the pathway of saltwater transport in response to the Major Baltic Inflows in the Baltic Sea. In 2014 IEEE/OES Baltic International Symposium (BALTIC) , 23 – 31 . Ed. by Carroll J. . Tallinn . Manganaro A. , Pulicanò G. , Reale A. , Sanfilippo M. , Sarà G. 2009 . Filtration pressure by bivalves affects the trophic conditions in Mediterranean shallow ecosystems . Chemistry and Ecology , 25 : 467 – 478 . Google Scholar CrossRef Search ADS Martin C. S. , Carpentier A. , Vaz S. , Coppin F. , Curet L. , Dauvin J.-C. , Delavenne J. et al. 2009 . The Channel habitat atlas for marine resource management (CHARM): an aid for planning and decision-making in an area under strong anthropogenic pressure . Aquatic Living Resources , 22 : 499 – 508 . Google Scholar CrossRef Search ADS Martin G. , Kotta J. , Möller T. , Herkül K. 2013 . Spatial distribution of marine benthic habitats in the Estonian coastal sea, northeastern Baltic Sea . Estonian Journal of Ecology , 62 : 165 – 191 . Google Scholar CrossRef Search ADS Martin G. , Paalme T. , Torn K. 2006 . Growth and production rates of loose-lying and attached forms of the red algae Furcellaria lumbricalis and Coccotylus truncatus in Kassari Bay, the West Estonian Archipelago Sea . Hydrobiologia , 554 : 107 – 115 . Google Scholar CrossRef Search ADS McArthur M. A. , Brooke B. P. , Przeslawsk I. R. , Ryan D. A. , Lucieer V. L. , Nichol S. , McCallum A. W. et al. 2010 . On the use of abiotic surrogates to describe marine benthic biodiversity . Estuarine, Coastal and Shelf Science, S , 88 : 21 – 32 . Google Scholar CrossRef Search ADS Minchinton T. E. , Scheibling R. E. , Hunt H. L. 1997 . Recovery of an intertidal assemblage following a rare occurrence of scouring by sea ice in Nova Scotia, Canada . Botanica Marina , 40 : 139148. Google Scholar CrossRef Search ADS Nikolopoulos A. , Isæus M. 2008 . Wave Exposure Calculations for the Estonian Coast . AquaBiota Water Research AB , Stockholm, Sweden . Okey T. A. , Agbayani S. , Hussein M. , Alidina H. M. 2015 . Mapping ecological vulnerability to recent climate change in Canada's Pacific marine ecosystems . Ocean and Coastal Management , 106 : 35 – 48 . Google Scholar CrossRef Search ADS Olsen J. L. , Stam W. T. , Coyer J. A. , Reusch T. B. , Billingham M. , Boström C. , Calvert E. , et al. 2004 . North Atlantic phylogeography and large-scale population differentiation of the seagrass Zostera marina L . Molecular Ecology , 13 : 1923 – 1941 . Google Scholar CrossRef Search ADS PubMed Orlova M. I. , Panov V. E. 2004 . Establishment of the zebra mussel, Dreissena polymorpha (Pallas), in the Neva Estuary (Gulf of Finland, Baltic Sea): distribution, population structure and possible impact on local un-ionid bivalves . Hydrobiologia , 514 : 207 – 217 . Google Scholar CrossRef Search ADS Peterson A. , Herkül K. 2017 . Mapping benthic biodiversity using georeferenced environmental data and predictive modeling . Marine Biodiversity , 1 – 16 . Plus M. , Deslous-Paoli J.-M. , Dagault F. 2003 . Seagrass (Zostera marina L.) bed recolonisation after anoxia-induced full mortality . Aquatic Botany , 77 : 121 – 134 . Google Scholar CrossRef Search ADS R Core Team . 2015 . R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org/ (last accessed 23 July 2018). Rajasilta M. , Laine P. , Eklund J. 2006 . Mortality of herring eggs on different algal substrates (Furcellaria spp. and Cladophora spp.) in the Baltic Sea—an experimental study . Hydrobiologia , 554 : 127 – 130 . Google Scholar CrossRef Search ADS Rayment W. J. 2008 . Furcellaria Lumbricalis a Red Seaweed. In Marine Life Information Network: Biology and Sensitivity Key Information Reviews, [Online] . Ed. by Tyler-Walters H. , Hiscock K. . Marine Biological Association of the United Kingdom , Plymouth . http://www.marlin.ac.uk/species/detail/1616 (last accessed 7 December 2017). Ridgeway G. 2007 . Generalized boosted models: a guide to the gbm package. http://www.saedsayad.com/docs/gbm2.pdf (last accessed 7 December 2017). Robinson L. M. , Elith J. , Hobday A. J. , Pearson R. G. , Kendal B. E. , Possingham H. P. , Richardson A. J. 2011 . Pushing the limits in marine species distribution modelling: lessons from the land present challenges and opportunities . Global Ecology and Biogeography , 20 : 789 – 802 . Google Scholar CrossRef Search ADS Rothäusler E. , Corell H. , Jormalainen V. 2015 . Abundance and dispersal trajectories of floating Fucus vesiculosus in the Northern Baltic Sea . Limnology and Oceanography , 60 : 2173 – 2184 . Google Scholar CrossRef Search ADS Sabol B. M. , Shafer D. J. , Lord E. 2005 . Dredging Effects on Eelgrass (Zostera marina) in a New England Small Boat Harbor . Journal of Marine Environmental Engineering , 8 : 57 – 81 . Schmieder K. , Werner S. , Bauer H. G. 2006 . Submersed macrophytes as a food source for wintering waterbirds at Lake Constance . Aquatic Botany , 84 : 245 – 250 . Google Scholar CrossRef Search ADS Selkoe K. A. , Halpern B. S. , Ebert C. M. , Franklin E. C. , Selig E. R. , Casey K. S. , Bruno J. et al. 2009 . A map of human impacts to a ‘‘pristine’’ coral reef ecosystem, the Papahānaumokuākea. Marine National Monument . Coral Reefs , 28 : 635 – 650 . Google Scholar CrossRef Search ADS Serrão E. A. , Brawley S. H. , Hedman J. , Kautsky L. , Samuelsson G. 1999 . Reproductive success of Fucus vesiculosus (Phaeophyceae) in the Baltic Sea . Journal of Phycology , 35 : 254 – 269 . Google Scholar CrossRef Search ADS Silberhorn G. M. , Dewing S. , Mason P. A. 1996 . Production of reproductive shoots, vegetative shoots, and seeds in populations of Ruppia maritima L. from the Chesapeake Bay, Virginia . Wetlands , 16 : 232 – 239 . Google Scholar CrossRef Search ADS Snoeijs-Leijonmalm P. , Schubert H. , Radziejewska T. 2017 . Biological Oceanography of the Baltic Sea . Springer , Netherlands . Steinhardt T. , Selig U. 2007 . Spatial distribution patterns and relationship between recent vegetation and diaspore bank of a brackish coastal lagoon on the southern Baltic Sea . Estuarine, Coastal and Shelf Science , 74 : 205 – 214 . Google Scholar CrossRef Search ADS Stelzenmüller V. , Fock H. O. , Gimpel A. , Rambo H. , Diekmann R. , Probst W. N. , Callies U. et al. 2015 . Quantitative environmental risk assessments in the context of marine spatial management: current approaches and some perspectives . ICES Journal of Marine Science , 72 : 1022 – 1042 . Google Scholar CrossRef Search ADS Suursaar Ü. , Alari V. , Tõnisson H. 2014 . Multi-scale analysis of wave conditions and coastal changes in the north-eastern Baltic Sea . Journal of Coastal Research , 70 : 223 – 228 . Google Scholar CrossRef Search ADS Torn K. , Martin G. , Kotta J. , Kupp M. 2010 . Effects of different types of mechanical disturbances on a charophyte dominated macrophyte community . Estuarine, Coastal and Shelf Science , 87 : 27 – 32 . Google Scholar CrossRef Search ADS Van Colen C. , Montserrat F. , Vincx M. , Herman P. M. J. , Ysebaert T. , Degraer S. 2008 . Macrobenthic recovery from hypoxia in an estuarine tidal mudflat . Marine Ecology Progress Series , 372 : 31 – 42 . Google Scholar CrossRef Search ADS Van den Berg M. S. , Coops H. , Simons J. 2001 . Propagule bank buildup of Chara aspera and its significance for colonization of a shallow lake . Hydrobiologia , 462 : 9 – 17 . Google Scholar CrossRef Search ADS Van Katwijk M. M. , Hermus D. C. R. , de Jong D. J. , Asmus R. M. , de Jonge V. R. 2000 . Habitat suitability of the Wadden Sea for restoration of Zostera marina beds . Helgoland Marine Research , 54 : 117 – 128 . Google Scholar CrossRef Search ADS Vári A. , Tóth V. R. 2017 . Quantifying macrophyte colonisation strategies—A field experiment in a shallow lake (Lake Balaton, Hungary) . Aquatic Botany , 136 : 56 – 60 . Google Scholar CrossRef Search ADS Villa F. , McLeod H. 2002 . Environmental vulnerability indicators for environmental planning and decision-making: guidelines and applications . Environmental Management , 29 : 335 – 348 . Google Scholar CrossRef Search ADS PubMed Wikström S. A. , Kautsky L. 2007 . Structure and diversity of invertebrate communities in the presence and absence of canopy-forming Fucus vesiculosus in the Baltic Sea . Estuarine, Coastal and Shelf Science , 72 : 168 – 176 . Google Scholar CrossRef Search ADS Zettler M. L. , Karlsson A. , Kontula T. , Gruszka P. , Laine A. O. , Herkül K. , Schiele K. S. et al. 2013 . Biodiversity gradient in the Baltic Sea: a comprehensive inventory of macrozoobenthos data . Helgoland Marine Research , 68 : 49 – 57 . Google Scholar CrossRef Search ADS Appendix Appendix 1 Table A1. List of georeferenced environmental variables that were used in the modelling of NVs and species richness in Estonia. Variable Abbreviation Source Water depth depth 1 Average water depth in 2000 m radius depth2 1 Slope of seabed slope 1 Slope of seabed in 2000 m radius slope2 1 Salinity salinity 2, 4 Wave exposure based on simplified wave model wave 5 Chlorophyll a content of sea surface based on satellite imagery chl 2 Water transparency estimated as attenuation coefficient based on satellite imagery attenuation 2 Ice coverage ice 6 Water temperature in cold season tempcold 3 Water temperature in warm season tempwarm 3 Current velocity current 3 Orbital speed of water movement at seabed induced by wind waves orbspeed 7 Proportion of soft sediment softsed 2 Secchi depth secchi 2 Concentration of ammonium ammonium 3 Concentration of nitrates nitrate 3 Concentration of phosphates phosphate 3 Sources: 1. Bathymetric data by Estonian Maritime Administration 2. Databases of the Estonian Marine Institute, University of Tartu 3. Hydrographic model developed by the Marine Systems Institute, Tallinn University of Technology (Maljutenko and Raudsepp, 2014) 4. COHERENS ocean circulation model (Bendtsen et al., 2009) 5. Simplified wave model based on fetch and wind data (Nikolopoulos and Isæus, 2008) 6. Finnish Meteorological Institute 7. SWAN hydrodynamic model (Suursaar et al., 2014) Variable Abbreviation Source Water depth depth 1 Average water depth in 2000 m radius depth2 1 Slope of seabed slope 1 Slope of seabed in 2000 m radius slope2 1 Salinity salinity 2, 4 Wave exposure based on simplified wave model wave 5 Chlorophyll a content of sea surface based on satellite imagery chl 2 Water transparency estimated as attenuation coefficient based on satellite imagery attenuation 2 Ice coverage ice 6 Water temperature in cold season tempcold 3 Water temperature in warm season tempwarm 3 Current velocity current 3 Orbital speed of water movement at seabed induced by wind waves orbspeed 7 Proportion of soft sediment softsed 2 Secchi depth secchi 2 Concentration of ammonium ammonium 3 Concentration of nitrates nitrate 3 Concentration of phosphates phosphate 3 Sources: 1. Bathymetric data by Estonian Maritime Administration 2. Databases of the Estonian Marine Institute, University of Tartu 3. Hydrographic model developed by the Marine Systems Institute, Tallinn University of Technology (Maljutenko and Raudsepp, 2014) 4. COHERENS ocean circulation model (Bendtsen et al., 2009) 5. Simplified wave model based on fetch and wind data (Nikolopoulos and Isæus, 2008) 6. Finnish Meteorological Institute 7. SWAN hydrodynamic model (Suursaar et al., 2014) View Large Table A1. List of georeferenced environmental variables that were used in the modelling of NVs and species richness in Estonia. Variable Abbreviation Source Water depth depth 1 Average water depth in 2000 m radius depth2 1 Slope of seabed slope 1 Slope of seabed in 2000 m radius slope2 1 Salinity salinity 2, 4 Wave exposure based on simplified wave model wave 5 Chlorophyll a content of sea surface based on satellite imagery chl 2 Water transparency estimated as attenuation coefficient based on satellite imagery attenuation 2 Ice coverage ice 6 Water temperature in cold season tempcold 3 Water temperature in warm season tempwarm 3 Current velocity current 3 Orbital speed of water movement at seabed induced by wind waves orbspeed 7 Proportion of soft sediment softsed 2 Secchi depth secchi 2 Concentration of ammonium ammonium 3 Concentration of nitrates nitrate 3 Concentration of phosphates phosphate 3 Sources: 1. Bathymetric data by Estonian Maritime Administration 2. Databases of the Estonian Marine Institute, University of Tartu 3. Hydrographic model developed by the Marine Systems Institute, Tallinn University of Technology (Maljutenko and Raudsepp, 2014) 4. COHERENS ocean circulation model (Bendtsen et al., 2009) 5. Simplified wave model based on fetch and wind data (Nikolopoulos and Isæus, 2008) 6. Finnish Meteorological Institute 7. SWAN hydrodynamic model (Suursaar et al., 2014) Variable Abbreviation Source Water depth depth 1 Average water depth in 2000 m radius depth2 1 Slope of seabed slope 1 Slope of seabed in 2000 m radius slope2 1 Salinity salinity 2, 4 Wave exposure based on simplified wave model wave 5 Chlorophyll a content of sea surface based on satellite imagery chl 2 Water transparency estimated as attenuation coefficient based on satellite imagery attenuation 2 Ice coverage ice 6 Water temperature in cold season tempcold 3 Water temperature in warm season tempwarm 3 Current velocity current 3 Orbital speed of water movement at seabed induced by wind waves orbspeed 7 Proportion of soft sediment softsed 2 Secchi depth secchi 2 Concentration of ammonium ammonium 3 Concentration of nitrates nitrate 3 Concentration of phosphates phosphate 3 Sources: 1. Bathymetric data by Estonian Maritime Administration 2. Databases of the Estonian Marine Institute, University of Tartu 3. Hydrographic model developed by the Marine Systems Institute, Tallinn University of Technology (Maljutenko and Raudsepp, 2014) 4. COHERENS ocean circulation model (Bendtsen et al., 2009) 5. Simplified wave model based on fetch and wind data (Nikolopoulos and Isæus, 2008) 6. Finnish Meteorological Institute 7. SWAN hydrodynamic model (Suursaar et al., 2014) View Large Appendix 2 Table A2. List of georeferenced environmental variables that were used in the modelling of NVs and species richness in Finland. Variable Abbreviation Source Water depth depth 1 Slope of seabed slope 1 Salinity on the surface sursal 1 Salinity on the bottom botsal 1 Euphotic depth Zeu_mean 1 Concentration of oxygen on the bottom oxygen 1 Concentration of nitrogen on the bottom botnit 1 Concentration of phosphorus on the bottom botphos 1 Minimum temperature on the bottom botTmin 1 Mean temperature on the bottom botTmean 1 Maximum temperature on the bottom botTmax 1 Depth attenuated wave exposure depth_expo 1 Potential reefs reefs 2 Distance to sandy shore distsand 1 Distance to river mouth distriver 1 Distance to harbour distharbor 1 Topographical position index 5 × 5 m TPI5x5 1 Probability of bottom being unstable/soft unstable 1 Hard/soft bottom substrate hardsoft 1 Bottom ruggedness ruggedbot 1 East-west position xc 1 North-south position yc 1 Sources: 1. Finnish Environment Institute SYKE 2. Geological Survey of Finland Variable Abbreviation Source Water depth depth 1 Slope of seabed slope 1 Salinity on the surface sursal 1 Salinity on the bottom botsal 1 Euphotic depth Zeu_mean 1 Concentration of oxygen on the bottom oxygen 1 Concentration of nitrogen on the bottom botnit 1 Concentration of phosphorus on the bottom botphos 1 Minimum temperature on the bottom botTmin 1 Mean temperature on the bottom botTmean 1 Maximum temperature on the bottom botTmax 1 Depth attenuated wave exposure depth_expo 1 Potential reefs reefs 2 Distance to sandy shore distsand 1 Distance to river mouth distriver 1 Distance to harbour distharbor 1 Topographical position index 5 × 5 m TPI5x5 1 Probability of bottom being unstable/soft unstable 1 Hard/soft bottom substrate hardsoft 1 Bottom ruggedness ruggedbot 1 East-west position xc 1 North-south position yc 1 Sources: 1. Finnish Environment Institute SYKE 2. Geological Survey of Finland View Large Table A2. List of georeferenced environmental variables that were used in the modelling of NVs and species richness in Finland. Variable Abbreviation Source Water depth depth 1 Slope of seabed slope 1 Salinity on the surface sursal 1 Salinity on the bottom botsal 1 Euphotic depth Zeu_mean 1 Concentration of oxygen on the bottom oxygen 1 Concentration of nitrogen on the bottom botnit 1 Concentration of phosphorus on the bottom botphos 1 Minimum temperature on the bottom botTmin 1 Mean temperature on the bottom botTmean 1 Maximum temperature on the bottom botTmax 1 Depth attenuated wave exposure depth_expo 1 Potential reefs reefs 2 Distance to sandy shore distsand 1 Distance to river mouth distriver 1 Distance to harbour distharbor 1 Topographical position index 5 × 5 m TPI5x5 1 Probability of bottom being unstable/soft unstable 1 Hard/soft bottom substrate hardsoft 1 Bottom ruggedness ruggedbot 1 East-west position xc 1 North-south position yc 1 Sources: 1. Finnish Environment Institute SYKE 2. Geological Survey of Finland Variable Abbreviation Source Water depth depth 1 Slope of seabed slope 1 Salinity on the surface sursal 1 Salinity on the bottom botsal 1 Euphotic depth Zeu_mean 1 Concentration of oxygen on the bottom oxygen 1 Concentration of nitrogen on the bottom botnit 1 Concentration of phosphorus on the bottom botphos 1 Minimum temperature on the bottom botTmin 1 Mean temperature on the bottom botTmean 1 Maximum temperature on the bottom botTmax 1 Depth attenuated wave exposure depth_expo 1 Potential reefs reefs 2 Distance to sandy shore distsand 1 Distance to river mouth distriver 1 Distance to harbour distharbor 1 Topographical position index 5 × 5 m TPI5x5 1 Probability of bottom being unstable/soft unstable 1 Hard/soft bottom substrate hardsoft 1 Bottom ruggedness ruggedbot 1 East-west position xc 1 North-south position yc 1 Sources: 1. Finnish Environment Institute SYKE 2. Geological Survey of Finland View Large Appendix 3 Table A3. The recovery classes and the respective sensitivity coefficients of NVs together with a short explanation and reference. Species/group Recovery class (years) Coefficient in calculations Rationale, references Fucus vesiculosus 2–3 2 As northern Baltic Sea littoral habitats are well connected (Rothäusler et al., 2015), recolonization will possibly occur during the next reproduction period (after complete removal, if conditions for growth are adequate and substrate present), but gamete dispersal is rather limited and can slow recovery (Serrão et al., 1999). Reaching to canopy state prior to removal, takes time. It took about 2 years (in Nova Scotia) after ice scouring completely eliminated fucoid assemblages (incl. F. vesiculosus) to re-colonize and fully recover (to canopy state similar to pre-scouring; recolonization took less than 7 months; Minchinton et al., 1997). Furcellaria lumbricalis 5–10 4 Recovers slowly due to low growth rate (Bird et al., 1979; Martin et al., 2006), long time to reach maturity (5 years in Wales; Austin, 1960) and recruitment that usually occurs in the vicinity of parent plant (Rayment, 2008). In addition, due to low salinity in the Gulf of Finland, vegetative reproduction prevails (Kostamo and Mäkinen, 2006) rendering the recovery even slower. Filamentous algae <2 1 After ice scouring that completely removed algae, it took less than few months to establish filamentous algae cover on rocks (including, among others: Pilayella littoralis, Polysiphonia sp., Ectocarpus sp., Ceramium virgatum, Cladophora sp.), in Nova Scotia (Minchinton et al., 1997). Artificial substrates became colonized by Pilayella littoralis in 3 months during winter and within one month in spring (Kraufvelin et al., 2007). Also other filamentous algae colonized artificial substrates within a year (during the first spring/summer; Kraufvelin et al., 2007). Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3-5 3 After ice scouring that completely removed bivalves, Mytilus spp. recolonized substrates in about 1-1.5 years; distribution similar to pre-scouring was observed within 5 years (Nova Scotia; Minchinton et al., 1997). While growth rates are lower in the Gulf of Finland than in optimal conditions (Kautsky, 1982a), recruitment is possible all year round in the Baltic Sea (Kautsky, 1982b).D. polymorpha: high fecundity, good dispersal abilities and rather fast growth rates (Mackie et al., 1989) support fast recovery. Becomes sexually mature in second year of life (Mackie et al., 1989). However, recovery may be slowed in the Gulf of Finland by limited or low spawning/recruitment in years with cold summers (Orlova and Panov, 2004). Vascular plants (excl. Zostera marina) 3-5 3 Ruppia maritima (and Najas marina)—annual species with short life cycle; high production rate (Kautsky, 1988) and high seed production (Silberhorn et al., 1996) indicates fast recovery. Stuckenia pectinata was successfully established in habitats created 3-5 years ago (Boedeltje et al., 2001).In Lake Balaton vascular plants (that occur also in the Baltic Sea/ Gulf of Finland) colonized rapidly de-vegetated areas via rhizomes, fragments of plants etc. (i.e. vegetative reproduction) from adjacent vegetated areas (Vári and Tóth, 2017), also in a Danish lake submerged macrophytes recolonized the lake within 5 years (up to 90% coverage; Lauridsen et al., 1994).Zannichellia palustris have a wide-ranging generative recolonization potential (Steinhardt and Selig, 2007).The dispersal and recolonization of aquatic plants and charophytes are encouraged by local propagule banks (Steinhardt and Selig, 2007) and waterbirds transporting the seeds (mostly in their guts) especially on local scale (Green et al., 2002). Zostera marina >10 5 Recolonization after total loss (i.e. no seed bank) can be extremely difficult (Holt et al., 1995). If there is adequate seed bank in sediments: recolonization (after anoxic event) was observed during next summer, but seedling mortality is huge (∼99%), so recovery takes definitely several years (Greve et al., 2005). In France it took less than 9 months till biomasses similar to pre-destruction and 2 years till flowering (Plus et al., 2003), but here, in colder Gulf of Finland eelgrass grows slower (Boström et al., 2014). In the northern Baltic Sea Z. marina commonly reproduces vegetatively (Olsen et al., 2004), thus there is no seed bank in sediments. Therefore, re-establishment will be very slow due to very limited vegetative dispersal (Holt et al., 1995) and possibly impoverished gene bank (Boström et al., 2014), and lack of suitable genotypes (as hypothesized for lower zones in Wadden Sea: Van Katwijk et al., 2000). No recovery was observed two years post-dredging in New England (seed bank removed together with sediment, but Z. marina growing in the area; Sabol et al., 2005). Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Chara vulgaris was successfully established in newly created habitats in less than 3 years (Boedeltje et al., 2001). If there is a sufficient oospore bank in sediment, it may greatly enhance recolonization (C. aspera in shallow lake; Van den Berg et al., 2001). However, charophyte recovery (by biomass) can take more than 2 years (Torn et al., 2010). C. tomentosa has not recolonized Byviken in Hanko, southwestern Finland since the dredging that took place many years ago. Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 In defaunated areas, infaunal bivalves (incl. Limecola balthica, Mya arenaria, and Cerastoderma edule) biomass recovery takes several years (Beukema et al., 1999), but abundance reached to similar values as in undisturbed areas within 1 year (at least 1 summer needed) or even faster (juvenile abundance; Van Colen et al., 2008). Seals >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). Birds >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). Species/group Recovery class (years) Coefficient in calculations Rationale, references Fucus vesiculosus 2–3 2 As northern Baltic Sea littoral habitats are well connected (Rothäusler et al., 2015), recolonization will possibly occur during the next reproduction period (after complete removal, if conditions for growth are adequate and substrate present), but gamete dispersal is rather limited and can slow recovery (Serrão et al., 1999). Reaching to canopy state prior to removal, takes time. It took about 2 years (in Nova Scotia) after ice scouring completely eliminated fucoid assemblages (incl. F. vesiculosus) to re-colonize and fully recover (to canopy state similar to pre-scouring; recolonization took less than 7 months; Minchinton et al., 1997). Furcellaria lumbricalis 5–10 4 Recovers slowly due to low growth rate (Bird et al., 1979; Martin et al., 2006), long time to reach maturity (5 years in Wales; Austin, 1960) and recruitment that usually occurs in the vicinity of parent plant (Rayment, 2008). In addition, due to low salinity in the Gulf of Finland, vegetative reproduction prevails (Kostamo and Mäkinen, 2006) rendering the recovery even slower. Filamentous algae <2 1 After ice scouring that completely removed algae, it took less than few months to establish filamentous algae cover on rocks (including, among others: Pilayella littoralis, Polysiphonia sp., Ectocarpus sp., Ceramium virgatum, Cladophora sp.), in Nova Scotia (Minchinton et al., 1997). Artificial substrates became colonized by Pilayella littoralis in 3 months during winter and within one month in spring (Kraufvelin et al., 2007). Also other filamentous algae colonized artificial substrates within a year (during the first spring/summer; Kraufvelin et al., 2007). Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3-5 3 After ice scouring that completely removed bivalves, Mytilus spp. recolonized substrates in about 1-1.5 years; distribution similar to pre-scouring was observed within 5 years (Nova Scotia; Minchinton et al., 1997). While growth rates are lower in the Gulf of Finland than in optimal conditions (Kautsky, 1982a), recruitment is possible all year round in the Baltic Sea (Kautsky, 1982b).D. polymorpha: high fecundity, good dispersal abilities and rather fast growth rates (Mackie et al., 1989) support fast recovery. Becomes sexually mature in second year of life (Mackie et al., 1989). However, recovery may be slowed in the Gulf of Finland by limited or low spawning/recruitment in years with cold summers (Orlova and Panov, 2004). Vascular plants (excl. Zostera marina) 3-5 3 Ruppia maritima (and Najas marina)—annual species with short life cycle; high production rate (Kautsky, 1988) and high seed production (Silberhorn et al., 1996) indicates fast recovery. Stuckenia pectinata was successfully established in habitats created 3-5 years ago (Boedeltje et al., 2001).In Lake Balaton vascular plants (that occur also in the Baltic Sea/ Gulf of Finland) colonized rapidly de-vegetated areas via rhizomes, fragments of plants etc. (i.e. vegetative reproduction) from adjacent vegetated areas (Vári and Tóth, 2017), also in a Danish lake submerged macrophytes recolonized the lake within 5 years (up to 90% coverage; Lauridsen et al., 1994).Zannichellia palustris have a wide-ranging generative recolonization potential (Steinhardt and Selig, 2007).The dispersal and recolonization of aquatic plants and charophytes are encouraged by local propagule banks (Steinhardt and Selig, 2007) and waterbirds transporting the seeds (mostly in their guts) especially on local scale (Green et al., 2002). Zostera marina >10 5 Recolonization after total loss (i.e. no seed bank) can be extremely difficult (Holt et al., 1995). If there is adequate seed bank in sediments: recolonization (after anoxic event) was observed during next summer, but seedling mortality is huge (∼99%), so recovery takes definitely several years (Greve et al., 2005). In France it took less than 9 months till biomasses similar to pre-destruction and 2 years till flowering (Plus et al., 2003), but here, in colder Gulf of Finland eelgrass grows slower (Boström et al., 2014). In the northern Baltic Sea Z. marina commonly reproduces vegetatively (Olsen et al., 2004), thus there is no seed bank in sediments. Therefore, re-establishment will be very slow due to very limited vegetative dispersal (Holt et al., 1995) and possibly impoverished gene bank (Boström et al., 2014), and lack of suitable genotypes (as hypothesized for lower zones in Wadden Sea: Van Katwijk et al., 2000). No recovery was observed two years post-dredging in New England (seed bank removed together with sediment, but Z. marina growing in the area; Sabol et al., 2005). Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Chara vulgaris was successfully established in newly created habitats in less than 3 years (Boedeltje et al., 2001). If there is a sufficient oospore bank in sediment, it may greatly enhance recolonization (C. aspera in shallow lake; Van den Berg et al., 2001). However, charophyte recovery (by biomass) can take more than 2 years (Torn et al., 2010). C. tomentosa has not recolonized Byviken in Hanko, southwestern Finland since the dredging that took place many years ago. Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 In defaunated areas, infaunal bivalves (incl. Limecola balthica, Mya arenaria, and Cerastoderma edule) biomass recovery takes several years (Beukema et al., 1999), but abundance reached to similar values as in undisturbed areas within 1 year (at least 1 summer needed) or even faster (juvenile abundance; Van Colen et al., 2008). Seals >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). Birds >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). View Large Table A3. The recovery classes and the respective sensitivity coefficients of NVs together with a short explanation and reference. Species/group Recovery class (years) Coefficient in calculations Rationale, references Fucus vesiculosus 2–3 2 As northern Baltic Sea littoral habitats are well connected (Rothäusler et al., 2015), recolonization will possibly occur during the next reproduction period (after complete removal, if conditions for growth are adequate and substrate present), but gamete dispersal is rather limited and can slow recovery (Serrão et al., 1999). Reaching to canopy state prior to removal, takes time. It took about 2 years (in Nova Scotia) after ice scouring completely eliminated fucoid assemblages (incl. F. vesiculosus) to re-colonize and fully recover (to canopy state similar to pre-scouring; recolonization took less than 7 months; Minchinton et al., 1997). Furcellaria lumbricalis 5–10 4 Recovers slowly due to low growth rate (Bird et al., 1979; Martin et al., 2006), long time to reach maturity (5 years in Wales; Austin, 1960) and recruitment that usually occurs in the vicinity of parent plant (Rayment, 2008). In addition, due to low salinity in the Gulf of Finland, vegetative reproduction prevails (Kostamo and Mäkinen, 2006) rendering the recovery even slower. Filamentous algae <2 1 After ice scouring that completely removed algae, it took less than few months to establish filamentous algae cover on rocks (including, among others: Pilayella littoralis, Polysiphonia sp., Ectocarpus sp., Ceramium virgatum, Cladophora sp.), in Nova Scotia (Minchinton et al., 1997). Artificial substrates became colonized by Pilayella littoralis in 3 months during winter and within one month in spring (Kraufvelin et al., 2007). Also other filamentous algae colonized artificial substrates within a year (during the first spring/summer; Kraufvelin et al., 2007). Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3-5 3 After ice scouring that completely removed bivalves, Mytilus spp. recolonized substrates in about 1-1.5 years; distribution similar to pre-scouring was observed within 5 years (Nova Scotia; Minchinton et al., 1997). While growth rates are lower in the Gulf of Finland than in optimal conditions (Kautsky, 1982a), recruitment is possible all year round in the Baltic Sea (Kautsky, 1982b).D. polymorpha: high fecundity, good dispersal abilities and rather fast growth rates (Mackie et al., 1989) support fast recovery. Becomes sexually mature in second year of life (Mackie et al., 1989). However, recovery may be slowed in the Gulf of Finland by limited or low spawning/recruitment in years with cold summers (Orlova and Panov, 2004). Vascular plants (excl. Zostera marina) 3-5 3 Ruppia maritima (and Najas marina)—annual species with short life cycle; high production rate (Kautsky, 1988) and high seed production (Silberhorn et al., 1996) indicates fast recovery. Stuckenia pectinata was successfully established in habitats created 3-5 years ago (Boedeltje et al., 2001).In Lake Balaton vascular plants (that occur also in the Baltic Sea/ Gulf of Finland) colonized rapidly de-vegetated areas via rhizomes, fragments of plants etc. (i.e. vegetative reproduction) from adjacent vegetated areas (Vári and Tóth, 2017), also in a Danish lake submerged macrophytes recolonized the lake within 5 years (up to 90% coverage; Lauridsen et al., 1994).Zannichellia palustris have a wide-ranging generative recolonization potential (Steinhardt and Selig, 2007).The dispersal and recolonization of aquatic plants and charophytes are encouraged by local propagule banks (Steinhardt and Selig, 2007) and waterbirds transporting the seeds (mostly in their guts) especially on local scale (Green et al., 2002). Zostera marina >10 5 Recolonization after total loss (i.e. no seed bank) can be extremely difficult (Holt et al., 1995). If there is adequate seed bank in sediments: recolonization (after anoxic event) was observed during next summer, but seedling mortality is huge (∼99%), so recovery takes definitely several years (Greve et al., 2005). In France it took less than 9 months till biomasses similar to pre-destruction and 2 years till flowering (Plus et al., 2003), but here, in colder Gulf of Finland eelgrass grows slower (Boström et al., 2014). In the northern Baltic Sea Z. marina commonly reproduces vegetatively (Olsen et al., 2004), thus there is no seed bank in sediments. Therefore, re-establishment will be very slow due to very limited vegetative dispersal (Holt et al., 1995) and possibly impoverished gene bank (Boström et al., 2014), and lack of suitable genotypes (as hypothesized for lower zones in Wadden Sea: Van Katwijk et al., 2000). No recovery was observed two years post-dredging in New England (seed bank removed together with sediment, but Z. marina growing in the area; Sabol et al., 2005). Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Chara vulgaris was successfully established in newly created habitats in less than 3 years (Boedeltje et al., 2001). If there is a sufficient oospore bank in sediment, it may greatly enhance recolonization (C. aspera in shallow lake; Van den Berg et al., 2001). However, charophyte recovery (by biomass) can take more than 2 years (Torn et al., 2010). C. tomentosa has not recolonized Byviken in Hanko, southwestern Finland since the dredging that took place many years ago. Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 In defaunated areas, infaunal bivalves (incl. Limecola balthica, Mya arenaria, and Cerastoderma edule) biomass recovery takes several years (Beukema et al., 1999), but abundance reached to similar values as in undisturbed areas within 1 year (at least 1 summer needed) or even faster (juvenile abundance; Van Colen et al., 2008). Seals >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). Birds >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). Species/group Recovery class (years) Coefficient in calculations Rationale, references Fucus vesiculosus 2–3 2 As northern Baltic Sea littoral habitats are well connected (Rothäusler et al., 2015), recolonization will possibly occur during the next reproduction period (after complete removal, if conditions for growth are adequate and substrate present), but gamete dispersal is rather limited and can slow recovery (Serrão et al., 1999). Reaching to canopy state prior to removal, takes time. It took about 2 years (in Nova Scotia) after ice scouring completely eliminated fucoid assemblages (incl. F. vesiculosus) to re-colonize and fully recover (to canopy state similar to pre-scouring; recolonization took less than 7 months; Minchinton et al., 1997). Furcellaria lumbricalis 5–10 4 Recovers slowly due to low growth rate (Bird et al., 1979; Martin et al., 2006), long time to reach maturity (5 years in Wales; Austin, 1960) and recruitment that usually occurs in the vicinity of parent plant (Rayment, 2008). In addition, due to low salinity in the Gulf of Finland, vegetative reproduction prevails (Kostamo and Mäkinen, 2006) rendering the recovery even slower. Filamentous algae <2 1 After ice scouring that completely removed algae, it took less than few months to establish filamentous algae cover on rocks (including, among others: Pilayella littoralis, Polysiphonia sp., Ectocarpus sp., Ceramium virgatum, Cladophora sp.), in Nova Scotia (Minchinton et al., 1997). Artificial substrates became colonized by Pilayella littoralis in 3 months during winter and within one month in spring (Kraufvelin et al., 2007). Also other filamentous algae colonized artificial substrates within a year (during the first spring/summer; Kraufvelin et al., 2007). Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3-5 3 After ice scouring that completely removed bivalves, Mytilus spp. recolonized substrates in about 1-1.5 years; distribution similar to pre-scouring was observed within 5 years (Nova Scotia; Minchinton et al., 1997). While growth rates are lower in the Gulf of Finland than in optimal conditions (Kautsky, 1982a), recruitment is possible all year round in the Baltic Sea (Kautsky, 1982b).D. polymorpha: high fecundity, good dispersal abilities and rather fast growth rates (Mackie et al., 1989) support fast recovery. Becomes sexually mature in second year of life (Mackie et al., 1989). However, recovery may be slowed in the Gulf of Finland by limited or low spawning/recruitment in years with cold summers (Orlova and Panov, 2004). Vascular plants (excl. Zostera marina) 3-5 3 Ruppia maritima (and Najas marina)—annual species with short life cycle; high production rate (Kautsky, 1988) and high seed production (Silberhorn et al., 1996) indicates fast recovery. Stuckenia pectinata was successfully established in habitats created 3-5 years ago (Boedeltje et al., 2001).In Lake Balaton vascular plants (that occur also in the Baltic Sea/ Gulf of Finland) colonized rapidly de-vegetated areas via rhizomes, fragments of plants etc. (i.e. vegetative reproduction) from adjacent vegetated areas (Vári and Tóth, 2017), also in a Danish lake submerged macrophytes recolonized the lake within 5 years (up to 90% coverage; Lauridsen et al., 1994).Zannichellia palustris have a wide-ranging generative recolonization potential (Steinhardt and Selig, 2007).The dispersal and recolonization of aquatic plants and charophytes are encouraged by local propagule banks (Steinhardt and Selig, 2007) and waterbirds transporting the seeds (mostly in their guts) especially on local scale (Green et al., 2002). Zostera marina >10 5 Recolonization after total loss (i.e. no seed bank) can be extremely difficult (Holt et al., 1995). If there is adequate seed bank in sediments: recolonization (after anoxic event) was observed during next summer, but seedling mortality is huge (∼99%), so recovery takes definitely several years (Greve et al., 2005). In France it took less than 9 months till biomasses similar to pre-destruction and 2 years till flowering (Plus et al., 2003), but here, in colder Gulf of Finland eelgrass grows slower (Boström et al., 2014). In the northern Baltic Sea Z. marina commonly reproduces vegetatively (Olsen et al., 2004), thus there is no seed bank in sediments. Therefore, re-establishment will be very slow due to very limited vegetative dispersal (Holt et al., 1995) and possibly impoverished gene bank (Boström et al., 2014), and lack of suitable genotypes (as hypothesized for lower zones in Wadden Sea: Van Katwijk et al., 2000). No recovery was observed two years post-dredging in New England (seed bank removed together with sediment, but Z. marina growing in the area; Sabol et al., 2005). Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Chara vulgaris was successfully established in newly created habitats in less than 3 years (Boedeltje et al., 2001). If there is a sufficient oospore bank in sediment, it may greatly enhance recolonization (C. aspera in shallow lake; Van den Berg et al., 2001). However, charophyte recovery (by biomass) can take more than 2 years (Torn et al., 2010). C. tomentosa has not recolonized Byviken in Hanko, southwestern Finland since the dredging that took place many years ago. Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 In defaunated areas, infaunal bivalves (incl. Limecola balthica, Mya arenaria, and Cerastoderma edule) biomass recovery takes several years (Beukema et al., 1999), but abundance reached to similar values as in undisturbed areas within 1 year (at least 1 summer needed) or even faster (juvenile abundance; Van Colen et al., 2008). Seals >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). Birds >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). View Large © International Council for the Exploration of the Sea 2018. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png ICES Journal of Marine Science Oxford University Press

Marine environmental vulnerability and cumulative risk profiles to support ecosystem-based adaptive maritime spatial planning

Loading next page...
 
/lp/oxford-university-press/marine-environmental-vulnerability-and-cumulative-risk-profiles-to-2EU7PdMtB2
Publisher
ICES/CIEM
Copyright
© International Council for the Exploration of the Sea 2018. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
1054-3139
eISSN
1095-9289
D.O.I.
10.1093/icesjms/fsy101
Publisher site
See Article on Publisher Site

Abstract

Abstract Human use of marine and coastal areas is increasing worldwide, resulting in conflicts between different interests for marine space, overexploitation of marine resources, and environmental degradation. In this study we developed a methodology that combines assessments of marine environmental vulnerability and cumulative human pressures to support the processes of ecosystem-based adaptive maritime spatial planning. The methodology is built on the spatially explicit marine environmental vulnerability profile (EVP) that is an aggregated product of the distribution of essential nature values (habitat-forming benthic macroalgal and invertebrate species, benthic species richness, birds and seals as top marine predators) and their sensitivities to disturbances. The marine environmental cumulative risk profile (ERP) combines the EVP and the HELCOM Baltic Sea Pressure Index (BSPI), the latter representing the spatial distribution of intensities of cumulative anthropogenic pressures. The ERP identifies areas where environmental risks are the highest due to both long recoveries of the biota and high intensities of human pressures. This methodology can be used in any other sea areas by modifying the list of nature values, their sensitivity to disturbances, and the intensities of human pressure. Introduction Human use of marine and coastal areas is increasing worldwide, resulting often in conflicts among different interests for the space, overexploitation of marine resources, and environmental degradation. As a consequence of such increasing and diversifying trends of human use, the marine environment, especially intensively used marine areas like the Baltic Sea, is becoming to a greater extent stressed and impacted (Korpinen et al., 2012). The multiple competing uses of marine and coastal areas have resulted in a rapid increase of maritime spatial planning (MSP) initiatives to safeguard sustainable use of marine resources as well as to mitigate cross-sectoral and transboundary conflicts over the use of sea space (Douvere and Ehler, 2010; Stelzenmüller et al., 2015). The MSP Directive 2014/89/EU establishes a framework for MSP aimed at promoting the sustainable growth of maritime economies, the sustainable development of marine areas, and the sustainable use of marine resources. The directive defines the MSP as a process, by which the relevant Member State’s competent authorities analyse and organize human activities in marine areas to achieve ecological, economic, and social objectives (EU, 2014). Cormier et al. (2015) states that “MSP should be seen as the practical implementation of the ecosystem approach to marine management through holistic and integrated analysis of all relevant human activities, pressures, and impacts within the planning area at the ecosystem scale”. In the frame of the MSP, assessments of environmental vulnerability and cumulative risks of human pressures are recognized as important management tools and a range of approaches have been described earlier (Villa and McLeod, 2002; Hiddink et al., 2007; Selkoe et al., 2009; Ardron et al., 2014; Stelzenmüller et al., 2015; La Rivière et al., 2016). Environmental vulnerability assessment is usually based on ecosystem sensitivity to pressures, expressed as resistance and/or recovery potential, i.e. resilience, and exposure to anthropogenic pressures (Bax and Williams, 2001; Hiddink et al., 2007; De Lange et al., 2010; Beroya-Eitner, 2016). Sensitivity has been expressed as the inverse of recovery time for a nature value (NV), e.g. community, habitat, or ecosystem, after being exposed to a pressure (Hiddink et al., 2007), or as a “capacity to tolerate a pressure (resistance) and the time needed to recover after an impact (resilience)” (La Rivière et al., 2016). The actual assessment or quantification of ecosystem sensitivity is difficult and usually it is aimed to produce the least disappointing vulnerability profile with the resources and time available (Villa and McLeod, 2002). The criteria that have been used to quantify the ecosystem sensitivity often include uniqueness or rarity of species or habitat, functional significance (e.g. habitat-forming species), fragility or susceptibility to degradation of species or habitat, and life history traits that are related to recovery potential (Ardron et al., 2014). Expert-derived ratings or rankings based on available scientific publications are widely used to assess intrinsic sensitivity of the NVs and their sensitivities to stressors (Villa and McLeod, 2002; De Lange et al., 2010; Okey et al., 2015; La Rivière et al., 2016). However, the quantitative assessment of ecosystem vulnerability is still challenging (Stelzenmüller et al., 2015) being calculated subjectively with little scientific justification (Villa and McLeod, 2002). Comprehensive environmental vulnerability assessment is pressure-driven and includes exposure, sensitivity to pressure, and recovery of a NV or ecosystem component (De Lange et al., 2010), and is based on best available knowledge (La Rivière et al., 2016). Moreover, most of vulnerability assessments that aim at contributing directly to MSP processes are either regional or national (e.g. Foley et al., 2013; La Rivière et al., 2016) and only seldom performed in a transboundary context (Martin et al., 2009). Risk assessment methods are widely used to assess and grade environmental problems (deFur et al., 2007) but current methods in general are not designed to address the risks of cumulative effects of environmental stressors. However, the U.S. Environmental Protection Agency has developed a general framework for cumulative risk assessment (EPA, 2003). According to this framework, a cumulative risk consists of the combined risks from aggregate exposures to multiple agents or stressors and the cumulative risk assessment means an analysis, characterization, and possible quantification of the combined risks to health or the environment from multiple agents or stressors. In order to develop spatially explicit marine environmental vulnerability and cumulative risk profiles, only the spatially “stable” ecosystem components (species or species groups with low mobility like benthic infauna) can be included. An inclusion of the characteristic habitat forming (e.g. large perennial seaweeds) and/or functionally important (e.g. suspension-feeding mussels) species ensures that the essential spatial proxies of “dynamic” species (highly mobile species like fish) are also included. In many coastal regions, several benthic plant and invertebrate species are considered as habitat engineers or habitat-forming species. They are capable of creating a specific local environment that facilitates colonization of other species that otherwise would not be present in the area (Martin et al., 2013; Koivisto and Westerbom, 2010, 2012). Large perennial seaweeds, mussels, and clams provide food, habitat (or both) to many macroalgae, invertebrates, fish, and birds (e.g. Beukema and Cadée, 1996). Moreover, suspension-feeding mussels clean the water by removing phytoplankton and organic matter and thereby controlling the water quality of the environment (Manganaro et al., 2009). Both seaweeds and benthic invertebrates generate spatial complexity and contribute to the stability of the environment. Consequently, a loss of such a complexity often causes drastic changes in the structure and functioning of the whole coastal ecosystem (Bertocci et al., 2010). Describing marine ecosystems can be challenging, mainly because sampling networks are generally sparse, which leaves most of the area not sampled and therefore with no information. Marine environments are more difficult to access and map compared with terrestrial ecosystems (Robinson et al., 2011; McArthur et al., 2010). Common seabed sampling methods such as grab sampler, trawls, scuba diving, and underwater video and photography yield point-wise data of seabed characteristics (Eleftheriou and McIntyre, 2005). Here, mathematical predictive modelling that is based on species–environment relationships provides a useful framework to extrapolate information from scattered samples into coherent seamless maps of distributions of species, habitats, ecological goods, and services (Guisan and Zimmermann, 2000; Guisan and Thuiller, 2005). The objective of this study was to develop a transparent, science-based and data-driven methodology that combines assessments of marine environmental vulnerability and cumulative human pressures into simple indices to support the processes of ecosystem-based adaptive MSP. Material and methods Study area The tideless Baltic Sea is characterized by a steep salinity gradient resulting in a variable fauna and flora, which tolerates well the prevailing environmental conditions. Material for the study originated from Estonian and Finnish marine areas, located in the north-eastern Baltic Sea (Figure 1). Figure 1. View largeDownload slide Study area and benthos sampling stations. Figure 1. View largeDownload slide Study area and benthos sampling stations. The area includes five major sub-basins of the Baltic Sea: Archipelago Sea, Bothnian Sea, the Northern Baltic Proper, Gulf of Finland, and Gulf of Riga. All of the sub-basins exhibit strong gradients of wave exposure, depth, and salinity. The sea areas west of the islands Saaremaa and Hiiumaa and the southwestern outer Archipelago Sea are exposed to the open Baltic Proper and have a wave fetch of hundreds of kilometres. In contrast, the inner reaches of the bays of the mainland are very sheltered both by the mainland and by islands. Salinity exceeds 7 PSU in the westernmost study area while it falls to almost 0 PSU in the inner parts of bays with riverine inflow (Kautsky and Kautsky, 2000; Karlson et al., 2002; Zettler et al., 2013; Alenius et al., 2016). The patterns of species distribution and species richness in the Baltic Sea follows many environmental gradients, with salinity appearing to be the most influential (Zettler et al., 2013; Snoeijs-Leijonmalm et al., 2017). Data Biological sampling Benthos data The data was compiled from different sources—the macrobenthos database of the Estonian Marine Institute (EMI) of the University of Tartu, the Environmental Information System database (HERTTA) of the Finnish Environment Institute, and the Finnish Inventory Programme for the Underwater Marine Environment (VELMU). The Estonian dataset covered the observations of benthic macrophytes and invertebrates, the HERTTA database contained the benthic invertebrate data and the VELMU database macrophyte data, respectively. Data from 11.523 benthos sampling stations in the Estonian waters and 2.818 benthic invertebrate sampling stations, 27.367 dive transect stations, and 93.363 drop-video stations in the Finnish waters were used as an input for species distribution and species richness models (Figure 1). All the samples were collected in Estonia between 2005 and 2015 and in Finland between 2000 and 2016. The sampling stations covered a depth range of 0.1−193 m in the Estonian waters and 0.1−286 m in the Finnish waters and spanned important environmental gradients like salinity, wave exposure, water transparency, and seabed type. Ekman and Van Veen type bottom grab samplers were used for benthic invertebrate samples on soft sediments in both countries. On hard surfaces, scuba divers collected all the fauna and flora inside a 0.04 m2 metal frame (Kautsky sampler). Sampling and sample analysis followed the guidelines developed for the HELCOM COMBINE programme (HELCOM, 2015). In Finland, macrophyte species composition was recorded by mapping a diver’s inspection square, typically 4 m2, or a video clip covering 20 m2 for every sampling station. Birds and seals Spatial data on birds was based on an aerial mapping and modelling by Luigujõe and Auniņš (2016) and integrated information on the density of benthos feeders, fish feeders, gulls, and swans during winter season in Estonia (Table 1). No bird data were available for Finnish area. Table 1. Summary of bird species/groups included in the study, from the Luigujõe and Auniņš (2016) models. Benthos feeders Fish feeders Gulls and swans Clangula hyemalis Gavia sp. Larus sp. Bucephala clangula Gavia stellate Larus minutus Somateria mollissima Mergus serrator Larus canus Polysticta stelleri Mergus merganser Larus argentatus Melanitta nigra Mergus albellus Cygnus sp. Melanitta fusca Phalacrocorax carbo Aythya fuligula Aythya marila Benthos feeders Fish feeders Gulls and swans Clangula hyemalis Gavia sp. Larus sp. Bucephala clangula Gavia stellate Larus minutus Somateria mollissima Mergus serrator Larus canus Polysticta stelleri Mergus merganser Larus argentatus Melanitta nigra Mergus albellus Cygnus sp. Melanitta fusca Phalacrocorax carbo Aythya fuligula Aythya marila Table 1. Summary of bird species/groups included in the study, from the Luigujõe and Auniņš (2016) models. Benthos feeders Fish feeders Gulls and swans Clangula hyemalis Gavia sp. Larus sp. Bucephala clangula Gavia stellate Larus minutus Somateria mollissima Mergus serrator Larus canus Polysticta stelleri Mergus merganser Larus argentatus Melanitta nigra Mergus albellus Cygnus sp. Melanitta fusca Phalacrocorax carbo Aythya fuligula Aythya marila Benthos feeders Fish feeders Gulls and swans Clangula hyemalis Gavia sp. Larus sp. Bucephala clangula Gavia stellate Larus minutus Somateria mollissima Mergus serrator Larus canus Polysticta stelleri Mergus merganser Larus argentatus Melanitta nigra Mergus albellus Cygnus sp. Melanitta fusca Phalacrocorax carbo Aythya fuligula Aythya marila Information provided by the Estonian Nature Information System (EELIS, 2017) on nationally protected moulting, resting, or breeding areas of seals was used as an additional input layer when calculating the EVP and ERP values at the Estonian side of the study area. There are two seal species in the study area: grey seal (Halichoerus grypus) and ringed seal (Pusa hispida). Sensitivity of nature values It is widely known that there exists a strong relationship between the health of natural environment and services these ecosystems provide to human wellbeing. Therefore, it becomes interesting and useful to develop a methodology that advises environmental managers how to minimize conflicts between different interests for marine space, reduce overexploitation of marine resources, and diminish environmental degradation. The sensitivity of the environment is defined by an interplay of natural settings (i.e. the biota) and the intensities and diversity of pressures. The biota represents the basement of the sensitivity assessments as the existing NVs largely define vulnerability of the environment to human pressures. In order to construct a holistic judgement on the sensitivities of NVs, a wide range of ecosystem elements should be incorporated into the analysis. As the MSP framework has a clear spatial dimension, it is practical to include among NVs only the spatially “stable” ecosystem components such as characteristic habitats (e.g. macroalgal beds) and/or functionally important species (e.g. ecoengineers). The inclusion of the latter ensures that the essential spatial proxies of “dynamic” ecosystem components (e.g. plankton, fish) are also included. Assessing pressure-specific sensitivity is very challenging because different pressures impact the marine environment simultaneously. Moreover, the magnitude of the cumulative impact is further modulated by different environmental variables such as salinity, depth, hydrodynamic activity, etc. To date there is a lack of such empirical knowledge to quantitatively formalize species sensitivity as functions of environmental variables. A practical approach to this complex problem can be the use of the recovery potential of an environmental value that is measured in time that is needed to recover from a destruction after its impact has ceased. For example, a reefs habitat type with ephemeral algae would recover (given that hard substrate is still present) within one year or a growing season because the spores of ephemeral algal species are produced in abundance and they disperse and colonize efficiently available habitats. However, the recovery of a reef habitat type with bladder wrack community would require 2−3 years. A ringed seal population would need more than 10 years to recover. Although the exact time needed for recovery is difficult to estimate and depends on the prevailing environmental conditions, generalizations based on known biological parameters of species or taxonomic groups can be utilized as a basis for such analyses. The Estonian and Finnish experts selected ten important benthic macroalgal, invertebrate, bird, and seal species (or groups of species) with different ecosystem functions and recovery potentials to represent NVs in this study (Table 2). The selection of NVs was based on their ecological importance (e.g. habitat forming species, top predators) and on the data availability. The recovery estimations were based on the earlier project results (e.g. Aps et al., 2011), expert opinions and on the literature (Appendix Table A3), combining relevant life history traits, observed time of recoveries and/or (re)colonization capacity of species in the Baltic Sea and/or similar areas. NVs were divided into five coefficients according to their recovery potential (time needed for recovery) to provide optimal differentiation between rapidly recovering annual filamentous algal species, more slowly recovering perennial algal species, benthic fauna and vascular plants, and very slowly recovering vertebrates. Table 2. Species and groups of species chosen to represent NVs with their recovery classes and coefficient for the further calculations according to the recovery class. Species/group Recovery class (years) Sensitivity coefficient Fucus vesiculosus 2–3 2 Furcellaria lumbricalis 5–10 4 Filamentous algae <2 1 Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3–5 3 Vascular plants excl. Zostera marina 3–5 3 Zostera marina >10 5 Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 Seals >10 5 Birds >10 5 Species/group Recovery class (years) Sensitivity coefficient Fucus vesiculosus 2–3 2 Furcellaria lumbricalis 5–10 4 Filamentous algae <2 1 Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3–5 3 Vascular plants excl. Zostera marina 3–5 3 Zostera marina >10 5 Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 Seals >10 5 Birds >10 5 Higher sensitivity coefficient indicates longer recovery time. For more detailed information see Appendix Table A3. Table 2. Species and groups of species chosen to represent NVs with their recovery classes and coefficient for the further calculations according to the recovery class. Species/group Recovery class (years) Sensitivity coefficient Fucus vesiculosus 2–3 2 Furcellaria lumbricalis 5–10 4 Filamentous algae <2 1 Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3–5 3 Vascular plants excl. Zostera marina 3–5 3 Zostera marina >10 5 Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 Seals >10 5 Birds >10 5 Species/group Recovery class (years) Sensitivity coefficient Fucus vesiculosus 2–3 2 Furcellaria lumbricalis 5–10 4 Filamentous algae <2 1 Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3–5 3 Vascular plants excl. Zostera marina 3–5 3 Zostera marina >10 5 Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 Seals >10 5 Birds >10 5 Higher sensitivity coefficient indicates longer recovery time. For more detailed information see Appendix Table A3. The selected benthic taxa can be considered as habitat-forming species that are a precondition or promote the existence of other species that otherwise would not be present in the area (Martin et al., 2013). The selected benthic algae serve as a spawning ground for economically important fish species like the Baltic herring (Rajasilta et al., 2006) and support a high biomass of invertebrates (Wikström and Kautsky, 2007). They can also be extensively consumed by waterfowl, thus forming a substantial component of the foodweb (Schmieder et al., 2006). The suspension feeding bivalves form a very important trophic link between pelagic and benthic systems (Lauringson et al., 2009; Koivisto and Westerbom, 2010, 2012) and maintain self-purification and water quality in marine coastal ecosystems. Fish and plankton were excluded in the current EVP and ERP assessments because of high spatial and temporal variability of these ecosystem components, and a lack of harmonized georeferenced data. NVs also included total species richness of benthic macroalgae and invertebrates calculated according to Peterson and Herkül (2017). Abiotic environment data The abiotic environmental variables included different bathymetric (depth, slope of seabed, topographic position), hydrodynamic (wave exposure, currents), geological (seabed substrate), and physical–chemical (temperature, salinity, transparency, nutrients, ammonium, ice conditions) variables. Altogether 18 Estonian and 23 Finnish environmental variables (Appendix Tables A1and A2) were used in the modelling. The resolution of the GIS layers was 100 m for Estonia and 20 m for Finland. Modelling Common marine monitoring methods yield point-wise data of seabed characteristics and in order to extrapolate information from such scattered samples into coherent seamless maps of distributions of species and habitats spatial modelling is needed. In this study the boosted regression trees (BRT) models were used to predict the spatial distribution of all benthic NVs, including benthic species richness, in the study area. BRT is an ensemble machine learning method that combines the strength of two algorithms: regression trees and boosting (Elith et al., 2008). In the BRT models the tree complexity was set to 5 for Estonian and 7 or 9 for Finnish models. The learning rates of Estonian and Finnish models were set to 0.005 and 0.01, respectively. The bag fraction was set at 0.5 which is the recommended default value (Elith et al., 2008). Modelling was done in the statistical software R version 3.3.1 (R Core Team, 2015) using the packages gbm (Ridgeway, 2007) and dismo (Elith and Leathwick, 2017). Finnish benthic species richness was calculated by combining invertebrate and macrophyte species richnesses modelled from taxa observed within every sampling station (invertebrates) or 200 m wide grid cell (macrophytes) using R package randomForest (Liaw and Wiener, 2002). Random forest (RF) is a machine learning method that similar to BRT modelling does not use any predefined data model but instead uses an algorithm to analyse the relationship between the biota and environmental variables. The RF generates a large number of regression trees, each calibrated on a bootstrap sample of the original data. Calibrated models were then used to predict benthic species richness and the probability of occurrence of NVs from the coastline to the outer border of the exclusive economic zone with a grid size of 100 m in Estonia and 20 m in Finland. Calculation of marine environmental vulnerability and cumulative risk profiles The HELCOM Baltic Sea Pressure Index (BSPI) (HELCOM, 2017) represented the intensity of cumulative anthropogenic pressure at the 1 × 1 km grid resolution in the study area. The BSPI incorporates multitude of human pressures weighed by their potential impacts on ecosystem. Environmental vulnerability profile (EVP) was calculated as a sum aggregation of all NVs that were first rescaled between 0 and 1 (by dividing with maximum value) and then weighed by NV-specific sensitivity coefficient (see Table 2). The ERP was a multiplication product of EVP and BSPI. The general scheme of the index calculation is shown in Figure 2. Figure 2. View largeDownload slide The general scheme of calculations of environmental vulnerability (EVP) and risk (ERP) profiles. Figure 2. View largeDownload slide The general scheme of calculations of environmental vulnerability (EVP) and risk (ERP) profiles. As the BSPI has been calculated at coarser spatial resolution than NVs used in this study, the spatial resolution of all NVs were upscaled to 1 × 1 km grid cells to match the BSPI resolution. When upscaling NVs, maximum values in each grid cell were used as it (1) underscored the most prominent NVs in grid cells (i.e. the precautionary principle not to mask the presence of high NVs by using mean values) and (2) better revealed spatial differences in NVs. The calculation of EVP and ERP included several steps (see Figure 2) all of which were proceeded in each 1 km grid cell. Benthic species richness was natural logarithm transformed to reduce the variation and divided by the maximum logarithmed richness value over all cells to make the values vary between 0 and 1. Then, to the product of logarithm a value of 1 was added to eliminate zero-values that would render further multiplication operations to zero (Formula 1). ln(richness+1)/max(ln(richness+1))+1 (1) The probability of occurrence of each benthic NV (excluding benthic richness) was multiplied by its respective sensitivity coefficient which was determined for each species or species group separately according to their recovery time (Table 2, Appendix A3). After that, all multiplication products within a grid cell were averaged. The averaged value was then multiplied by the transformed benthic species richness. Similar to benthic NVs, bird and seal NVs were also multiplied by their respective sensitivity coefficients. The resulting benthos, bird and seal products were averaged and then rescaled by dividing each value by its maximum value over all cells to make the values vary between 0 and 1. This rescaled product of the distribution of NVs and their sensitivities was termed as EVP-F. EVP-F was calculated for Estonian sea area only. Due to the lack of comparable data on birds in Finnish sea areas, an alternative index was developed that included only benthic NVs, hereafter termed as the EVP-B. The EVP-B was calculated for both Estonian and Finnish sea areas. The cumulative ERP calculation was based on the following basic assumptions: (1) the level of cumulative human pressure (BSPI) is proportional to the likelihood of potential environmental cumulative effect events and (2) the level of environmental vulnerability is proportional to potential consequences of environmental cumulative effect events. In order to calculate the ERP in the study area, the BSPI was divided by its maximum value over all cells to make the values vary between 0 and 1. Then the rescaled BSPI was multiplied with EVP and divided by the maximum value of such multiplication term over all grid cells to make the values vary between 0 and 1 (Formula 2). (EVP×BSPI/max(BSPI))/max(EVP×BSPI/max(BSPI)) (2) Similar to the EVP, when calculations were based on EVP-F the index was termed ERP-F and in case calculations were based on EVP-B, the index was termed ERP-B. Results Marine environmental vulnerability and risk profiles Two different versions of the EVP profiles are shown in Figure 3. EVP-B had the highest values in the Finnish Archipelago Sea and in the western coasts of Estonia and around its islands. In general, coastal areas had higher EVP-F values compared with offshore areas. The highest values of EVP-F coincided with nationally protected moulting, resting, or breeding areas of seals in the Estonian sea waters. This is because these areas are often characterized by high benthic richness and the presence of other NVs. Figure 3. View largeDownload slide (a) Environmental vulnerability profile based on benthic nature values (EVP-B) and (b) environmental vulnerability profile including benthic nature values, birds and seals (EVP-F) in Estonian sea waters. Values vary between 0 and 1, where 1 expresses the highest vulnerability. Figure 3. View largeDownload slide (a) Environmental vulnerability profile based on benthic nature values (EVP-B) and (b) environmental vulnerability profile including benthic nature values, birds and seals (EVP-F) in Estonian sea waters. Values vary between 0 and 1, where 1 expresses the highest vulnerability. The two different versions of ERP (ERP-B and ERP-F) are shown in Figure 4. ERP-B had the highest values in the Finnish Archipelago Sea and close to the City of Helsinki. In the Estonian waters, the highest environmental risk values were characteristic to the coastal sea areas around the islands and close to the city of Tallinn. Similar to EVP layers, higher values of ERP-F coincided with seal areas. In addition, ERP-F had higher values adjacent to the City of Tallinn and along the northern coasts of Estonia. Figure 4. View largeDownload slide (a) Environmental risk profile based on benthic nature values (ERP-B) and (b) environmental risk profile that including benthic nature values, birds and seals (ERP-F) in Estonian waters. Risk values vary between 0 and 1, where 1 expresses the highest risk value. Figure 4. View largeDownload slide (a) Environmental risk profile based on benthic nature values (ERP-B) and (b) environmental risk profile that including benthic nature values, birds and seals (ERP-F) in Estonian waters. Risk values vary between 0 and 1, where 1 expresses the highest risk value. Discussion Environmental vulnerability Here, we described a transparent, science-based, and data-driven methodology that summarizes vulnerabilities of different NVs and risks of different human pressures as a single map. Such map can be easily communicated to MSP experts and other interested stakeholders to jointly analyse and compare the potential environmental risk levels resulting from different planning solutions. While it is impossible to include all the species and habitats to the assessment of ecosystem sensitivity, acknowledging what is valuable in the environment is undoubtedly important in environmental management planning (Villa and McLeod, 2002). We included the spatially “stable” ecosystem components such as the characteristic habitat forming species (e.g. large perennial seaweeds), functionally important organisms (e.g. suspension-feeding mussels) and top predators (birds, seals) into our analyses. Due to strong interlinkage of different ecosystem components our choice of NVs ensures that the essential spatial proxies of “dynamic” species are also included. For example, the presence or absence of habitat forming benthic species impact all trophic levels from phytoplankton to seabirds either modifying seabed structure and/or food web (e.g. Beukema and Cadée, 1996) and therefore can be regarded as a proxy for other groups of organisms not used in the current assessment. Similarly, the distribution of seabed biota has a strong linkage to fish because there are certain essential fish habitats (spawning, feeding) that are primarily defined by the benthic macroalgae, vascular plant, and invertebrate communities. There is a generic lack of empirical knowledge to quantify the sensitivities of species and habitats to different human pressures along environmental variables (Villa and McLeod, 2002). Moreover, pressure-specific sensitivities are very challenging to be assessed because different pressures impact the environment simultaneously. A practical solution to this complex problem can be alleviated by using the expert-based and/or empirically derived recovery potential of NV that is measured in time that is needed to recover from a disturbance after its effect has ceased. Recovery time has been suggested also by Hiddink et al., (2007) as useful parameter in estimating habitat sensitivity and has been later used by Ardron et al., (2014) and Stelzenmüller et al., (2015). EVP is not meant to reflect sensitivity to any specific pressures; it only sums up the spatial distribution of NVs for which we have the distribution data and knowledge on their recovery rates. Different pressures may have specific impacts on different ecosystem components mediated by natural abiotic gradients (e.g. depth, seabed substrates, currents, salinity) and EVP was not expected to elucidate such details. Ideally, detailed impact assessments should be carried out during environmental impact assessments (EIA) of concrete projects. EVP is meant to help maritime spatial planners as a starting point for allocating marine areas that host many NVs. Environmental cumulative risk The BSPI (HELCOM, 2010, 2017) is defined as the “straightforward measure of the geographical distribution and intensity of anthropogenic pressure on the Baltic Sea marine environment”. The BSPI approach allows calculation of cumulative pressure by summing the pressure intensities in the area without including ecosystem data into the calculations. There exist diverse pathways of the pressure–biota relationships and only empirical data supporting cause-effect pathways enable a detailed and highly accurate estimation of the effects of specific anthropogenic pressures on ecosystem components. In the real world we face the situation where we lack such detailed knowledge but we still need generalizations for large-scale management tasks, e.g. for MSP. The BSPI is a product that has been jointly developed by the Baltic Sea states in order to provide a large-scale estimation of the cumulative human pressures and it summarizes the best available scientific and expert knowledge. When joining assessments of marine environmental vulnerability (EVP) and cumulative human pressures (BSPI), the marine cumulative ERP can be produced. The ERP identifies spatial areas of highest risk based on the likelihood and magnitude of environmental vulnerability and human-induced pressures. For example, high levels of environmental cumulative risk are situated near to the largest cities. Coastal sea usually supports the highest species richness (Peterson and Herkül, 2017), but at the same time sustain the strongest human pressure. Here, differences in the environmental cumulative risk value can guide planners toward the most environment-friendly planning solutions. However, establishment of environmental cumulative risk-related tolerability levels require extensive consultations with regulators, stakeholders, and the public in order to determine the level of risk that is acceptable to all stakeholders. It is stated (ICES, 2014) that “Given that a scientific assessment is objective and is based on facts, it would simply reflect likelihood and magnitude leaving the severity, tolerability or values to the governance decision-making processes and stakeholder constituency”. Conclusions In this study we developed a methodology that combines assessments of marine environmental vulnerability (EVP) and cumulative human pressures to produce the marine cumulative ERP. The marine EVP was calculated as a spatial data layer incorporating the distribution of essential NVs (habitat-forming benthic macroalgal and invertebrate species, benthic species richness, seals, and birds) and their recovery time to disturbances. The ERP was calculated as a georeferenced data layer by linking the EVP and the HELCOM BSPI, the latter representing the spatial distribution of the intensity of cumulative anthropogenic pressures. The developed spatially explicit profiles enable the development of environmentally sustainable MSP solutions. The EVPs visualize the spatial patterns of marine NVs and their sensitivities to different human pressures in terms of their recovery potential. The EVP and ERP enable maritime spatial planners to compare the environmental cumulative risk of different planning solutions and thereby to overcome the major environmental challenges faced by any highly impacted marine ecosystem. The suggested methodology can be used in any other sea areas by modifying the list of NVs, their sensitivity to disturbances and the level of human pressure. Moreover, the EVP and ERP are applicable at various scales—from a single bay to large marine areas. Acknowledgements This study is supported by European Regional Development Fund, INTERREG Central Baltic project Plan4Blue “Maritime Spatial Planning for Sustainable Blue Economies”, the Estonian Environmental Investment Centre, the Estonian Research Council (Institutional research funding, IUT02-20), and the European Union’s Seventh Programme for research, technological development, and demonstration BONUS project MARES “Multi-method Assessment for Resilient Ecosystem Services and human-nature system integration”. The authors thank the Working Group for Marine Planning and Coastal Zone Management (WGMPCZM) of the International Council for the Exploration of the Sea (ICES) for facilitating this research. References Alenius P. , Myrberg K. , Roiha P. , Lips U. , Tuomi L. , Pettersson H. , Raateoja M. 2016 . Gulf of Finland physics. In The Gulf of Finland Assessment, No. 27 . 42 pp. Ed. by Raateoja M. , Setälä O. . Reports of the Finnish Environment Institute , Helsinki, Finland . Aps R. , Ambjörn C. , Fetissov M. , Karjalainen M. , Kotta J. , Kuikka S. 2011 . OILRISK web: advanced tool for enhancing spill response decision-making in the Baltic Sea. In Oil Spill Risk Management , pp. 109 – 117 . Ed. by Bellefontaine N. , Lindén O. . The Proceedings of the International Conference on Oil Spill Risk Management, Mamlö, Sweden . Ardron J. A. , Clark M. R. , Penney A. J. , Hourigan T. F. , Rowden A. A. , Dunstan P. K. , Watling L. et al. 2014 . A systematic approach towards the identification and protection of vulnerable marine ecosystems . Marine Policy , 49 : 146 – 154 . Google Scholar CrossRef Search ADS Austin A. P. 1960 . Observations on the growth, fruiting and longevity of Furcellaria fastigiata (L.) Lamouroux . Hydrobiologia , 15 : 193 – 207 . Google Scholar CrossRef Search ADS Bax N. J. , Williams A. 2001 . Seabed habitat on the southeastern Australian continental shelf: context, vulnerability and monitoring . Marine and Freshwater Research , 52 : 491 – 512 . Google Scholar CrossRef Search ADS Bendtsen J. , Gustafsson K. E. , Söderkvist J. , Hansen J. L. S. 2009 . Ventilation of bottom water in the North Sea–Baltic Sea transition zone . Journal of Marine Systems , 75 : 138 – 149 . Google Scholar CrossRef Search ADS Beroya-Eitner M. A. 2016 . Ecological vulnerability indicators . Ecological Indicators , 60 : 329 – 334 . Google Scholar CrossRef Search ADS Bertocci I. , Arenas F. , Matias M. , Vaselli S. , Araújo R. , Abreu H. , Pereira R. et al. 2010 . Canopy-forming species mediate the effects of disturbance on macroalgal assemblages on Portuguese rocky shores . Marine Ecology Progress Series , 414 : 107 – 116 . Google Scholar CrossRef Search ADS Beukema J. J. , Cadée G. C. 1996 . Consequences of the sudden removal of nearly all mussels and cockles from the Dutch Wadden Sea . Marine Ecology , 17 : 279 – 289 . Google Scholar CrossRef Search ADS Beukema J. J. , Flach E. C. , Dekker R. , Starink M. 1999 . A long-term study of the recovery of the macrozoobenthos on large defaunated plots on a tidal flat in the Wadden Sea . Journal of Sea Research , 42 : 235 – 254 . Google Scholar CrossRef Search ADS Bird N. L. , Chen L.,C.-M. , McLachlan J. 1979 . Effects of temperature, light and salinity of growth in culture of Chondrus crispus, Furcellaria lumbricalis, Gracilaria tikvahiae (Gigartinales, Rhodophyta), and Fucus serratus (Fucales, Phaeophyta) . Botanica Marina , 22 : 521527. Google Scholar CrossRef Search ADS Boedeltje G. , Smolders A. J. P. , Roelofs J. G. M. , Van Groenendael J. M. 2001 . Constructed shallow zones along navigation canals: vegetation establishment and change in relation to environmental characteristics . Aquatic Conservation: Marine and Freshwater Ecosystems , 11 : 453 – 471 . Google Scholar CrossRef Search ADS Boström C. , Baden S. , Bockelmann A.-C. , Dromph K. , Fredriksen S. , Gustafsson C. , Krause-Jensen D. , et al. 2014 . Distribution, structure and function of Nordic eelgrass (Zostera marina) ecosystems: implications for coastal management and conservation . Aquatic Conservation: Marine and Freshwater Ecosystems , 24 : 410 – 434 . Google Scholar CrossRef Search ADS PubMed Cormier R. , Kannen A. , Elliott M. , Hall P. 2015 . Marine spatial planning quality management system. ICES Cooperative Research Report No. 327, 106 pp. deFur P. L. , Evans G. W. , Hubal E. A. C. , Kyle A. D. , Morello-Frosch R. A. , Williams D. R. 2007 . Vulnerability as a function of individual and group resources in cumulative risk assessment . Environ Health Perspect , 115 : 817 – 824 . Google Scholar CrossRef Search ADS PubMed De Lange H. J. , Sala S. , Vighi M. , Faber J. H. 2010 . Ecological vulnerability in risk assessment—A review and perspectives . Science of the Total Environment , 408 : 3871 – 3879 . Google Scholar CrossRef Search ADS PubMed Douvere F. , Ehler C. N. 2010 . The importance of monitoring and evaluation in adaptive maritime spatial planning . Journal of Coastal Conservation , 15 : 305 – 311 . Google Scholar CrossRef Search ADS EELIS . 2017 . Estonian nature information system —Environmental register. Environment Agency, Republic of Estonia. Accessed 19 January 2017. Eleftheriou A. , McIntyre A. D. 2005 . Methods for the Study of Marine Benthos , 3rd edn . Blackwell , Oxford . Elith J. , Leathwick J. R. , Hastie T. 2008 . A working guide to boosted regression trees . Journal of Animal Ecology , 77 : 802 – 813 . Google Scholar CrossRef Search ADS PubMed Elith J. , Leathwick J. R. 2017 . Boosted regression trees for ecological modeling. https://cran.r-project.org/web/packages/dismo/vignettes/brt.pdf. Accessed 18 August 2017 EPA . 2003 . Framework for cumulative risk assessment. EPA/600/P-02/001F. National Center for Environmental Assessment, Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. EU . 2014 . Establishing a framework for maritime spatial planning, Directive 2014/89/EU, European Council, Brussels. Foley M. M. , Armsby M. H. , Prahler E. E. , Caldwell M. R. , Erickson A. L. , Kittinger J. N. , Crowder L. B. et al. 2013 . Improving ocean management through the use of ecological principles and integrated ecosystem assessments . Bioscience , 63 : 619 – 631 . Google Scholar CrossRef Search ADS Green A. J. , Figuerola J. , Sánchez M. I. 2002 . Implications of waterbird ecology for the dispersal of aquatic organisms . Acta Oecologica , 23 : 177 – 189 . Google Scholar CrossRef Search ADS Greve T. M. , Krause-Jensen D. , Rasmussen M. B. , Christensen P. B. 2005 . Means of rapid eelgrass (Zostera marina L.) recolonisation in former dieback areas . Aquatic Botany , 82 : 143 – 156 . Google Scholar CrossRef Search ADS Guisan A. , Thuiller W. 2005 . Predicting species distribution: offering more than simple habitat models . Ecology Letters , 8 : 993 – 1009 . Google Scholar CrossRef Search ADS Guisan A. , Zimmermann N. E. 2000 . Predictive habitat distribution models in ecology . Ecological Modelling , 135 : 147 – 186 . Google Scholar CrossRef Search ADS HELCOM . 2010 . Towards a tool for quantifying anthropogenic pressures and potential impacts on the Baltic Sea marine environment. A background document on the method, data and testing of the Baltic Sea Pressure and Impact Indices. Baltic Sea Environment Proceedings No, 125. HELCOM . 2015 . Manual for marine monitoring in the COMBINE programme of HELCOM. http://www.helcom.fi/action-areas/monitoring-and-assessment/manuals-and-guidelines/combine-manual (last accessed 28 December 2017). HELCOM . 2017 . The assessment of cumulative impact using the BSPI and BSPII: Supplementary report to the first version of the ‘State of the Baltic Sea’ Report 2017. http://stateofthebalticsea.helcom.fi/wpcontent/uploads/2017/09/HELCOM_The_assessment_of_cumulative_impacts_Supplementary_report_first_version_2017.pdf (last accessed 2 January 2018) Hiddink J. G. , Jennings S. , Kaiser M. J. 2007 . Assessing and predicting the relative ecological impacts of disturbance on habitats with different sensitivities . Journal of Applied Ecology , 44 : 405 – 413 . Google Scholar CrossRef Search ADS Holt T. J. , Hartnoll R. G. , Hawkins S. J. 1995 . Sensitivity and vulnerability to man induced change of selected communities: intertidal brown algal shrubs Zostera beds and Sabellaria spinulosa reefs . English Nature. ENRR No. 234 , 97. pp. ICES . 2014 . Report of the Joint Rijkswaterstaat/DFO/ICES Workshop: Risk Assessment for Spatial Management (WKRASM), 24–28 February 2014, Amsterdam, the Netherlands. ICES CM 2014/SSGHIE: 01. 35 pp. Karlson K. , Rosenberg R. , Bonsdorff E. 2002 . Temporal and spatial large-scale effects of eutrophication and oxygen deficiency on benthic fauna in Scandinavian and Baltic waters – a review . Oceanography and Marine Biology, an Annual Review , 40 : 427 – 489 . Kautsky N. 1982 . Growth and Size Structure in a Baltic Mytilus edulis Population . Marine Biology , 68 : 117 – 133 . Google Scholar CrossRef Search ADS Kautsky N. 1982 . Quantitative Studies on Gonad Cycle, Fecundity, Reproductive Output and Recruitment in a Baltic Mytilus edulis Population . Marine Biology , 68 : 143 – 160 . Google Scholar CrossRef Search ADS Kautsky L. 1988 . Life strategies of soft bottom macrophytes . Oikos , 53 : 126 – 135 . Google Scholar CrossRef Search ADS Kautsky L. , Kautsky N. 2000 . The Baltic Sea, including Bothnian Sea and Bothnian Bay. In Seas at the Millennium: An Environmental Evaluation , pp. 121 – 133 . Ed. by Sheppard C. . Elsevier , Amsterdam . Koivisto M. E. , Westerbom M. 2010 . Habitat structure and complexity as determinants of biodiversity in blue mussel beds on sublittoral rocky shores . Marine Biology , 157 : 1463 – 1474 . Google Scholar CrossRef Search ADS Koivisto M. E. , Westerbom M. 2012 . Invertebrate communities associated with blue mussel beds in a patchy environment: a landscape ecology approach . Marine Ecology Progress Series , 471 : 101 – 110 . Google Scholar CrossRef Search ADS Korpinen S. , Meski L. , Andersen J. H. , Laamanen M. 2012 . Human pressures and their potential impact on the Baltic Sea ecosystem . Ecological Indicators , 15 : 105 – 114 . Google Scholar CrossRef Search ADS Kostamo K. , Mäkinen A. 2006 . Observations on the mode and seasonality of reproduction in Furcellaria lumbricalis (Gigartinales, Rhodophyta) populations in the northern Baltic Sea . Botanica Marina , 49 : 304 – 309 . Google Scholar CrossRef Search ADS Kraufvelin P. , Ruuskanen A. T. , Nappu N. , Kiirikki M. 2007 . Winter colonisation and succession of filamentous macroalgae on artificial substrates and possible relationships to Fucus vesiculosus settlement in early summer . Estuarine, Coastal and Shelf Science , 72 : 665 – 674 . Google Scholar CrossRef Search ADS La Rivière M. , Aish A. , Gauthier O. , Grall J. , Guérin L. , Janson A.-L. , Labrune C. et al. 2016 . Assessing benthic habitats’ sensitivity to human pressures: a methodological framework - Summary report. Rapport SPN 2016-87. MNHN. Paris, 42. pp. Lauridsen T. L. , Jeppesen E. , Sondergaard M. 1994 . Colonization and succession of submerged macrophytes in shallow Lake Vaeng during the first five years following fish manipulation . Hydrobiologia , 275-276 : 233 – 242 . Google Scholar CrossRef Search ADS Lauringson V. , Kotta J. , Orav-Kotta H. , Kotta I. , Herkül K. , Põllumäe A. 2009 . Comparison of benthic and pelagic suspension feeding in shallow water habitats of the Northeastern Baltic Sea . Marine Ecology , 30 : 43 – 55 . Google Scholar CrossRef Search ADS Liaw A. , Wiener M. 2002 . Classification and regression by random forest . R News , 2 : 18 – 22 . Luigujõe L. , Auniņš A. 2016 . Talvituvate lindude rahvusvaheline lennuloendus. Report. https://www.keskkonnaamet.ee/sites/default/files/luigujoeaunins_2016_talvituvate_veelindude-rahvusvaheline_lennuloendus_lopparuanne_rev01-2017.pdf (last accessed 7 December 2017). Mackie G. L. , Gibbons W. N. , Muncaster B. W. , Gray I. M. 1989 . The Zebra Mussel, Dreissena Polymorpha: A Synthesis of European Experiences and a Preview for North America . Report prepared for Water Resources Branch, Ontario Ministry of Environment , Canada . ISBN 0-7729-5647-2. Maljutenko I. , Raudsepp U. 2014 . Validation of GETM model simulated long-term salinity fields in the pathway of saltwater transport in response to the Major Baltic Inflows in the Baltic Sea. In 2014 IEEE/OES Baltic International Symposium (BALTIC) , 23 – 31 . Ed. by Carroll J. . Tallinn . Manganaro A. , Pulicanò G. , Reale A. , Sanfilippo M. , Sarà G. 2009 . Filtration pressure by bivalves affects the trophic conditions in Mediterranean shallow ecosystems . Chemistry and Ecology , 25 : 467 – 478 . Google Scholar CrossRef Search ADS Martin C. S. , Carpentier A. , Vaz S. , Coppin F. , Curet L. , Dauvin J.-C. , Delavenne J. et al. 2009 . The Channel habitat atlas for marine resource management (CHARM): an aid for planning and decision-making in an area under strong anthropogenic pressure . Aquatic Living Resources , 22 : 499 – 508 . Google Scholar CrossRef Search ADS Martin G. , Kotta J. , Möller T. , Herkül K. 2013 . Spatial distribution of marine benthic habitats in the Estonian coastal sea, northeastern Baltic Sea . Estonian Journal of Ecology , 62 : 165 – 191 . Google Scholar CrossRef Search ADS Martin G. , Paalme T. , Torn K. 2006 . Growth and production rates of loose-lying and attached forms of the red algae Furcellaria lumbricalis and Coccotylus truncatus in Kassari Bay, the West Estonian Archipelago Sea . Hydrobiologia , 554 : 107 – 115 . Google Scholar CrossRef Search ADS McArthur M. A. , Brooke B. P. , Przeslawsk I. R. , Ryan D. A. , Lucieer V. L. , Nichol S. , McCallum A. W. et al. 2010 . On the use of abiotic surrogates to describe marine benthic biodiversity . Estuarine, Coastal and Shelf Science, S , 88 : 21 – 32 . Google Scholar CrossRef Search ADS Minchinton T. E. , Scheibling R. E. , Hunt H. L. 1997 . Recovery of an intertidal assemblage following a rare occurrence of scouring by sea ice in Nova Scotia, Canada . Botanica Marina , 40 : 139148. Google Scholar CrossRef Search ADS Nikolopoulos A. , Isæus M. 2008 . Wave Exposure Calculations for the Estonian Coast . AquaBiota Water Research AB , Stockholm, Sweden . Okey T. A. , Agbayani S. , Hussein M. , Alidina H. M. 2015 . Mapping ecological vulnerability to recent climate change in Canada's Pacific marine ecosystems . Ocean and Coastal Management , 106 : 35 – 48 . Google Scholar CrossRef Search ADS Olsen J. L. , Stam W. T. , Coyer J. A. , Reusch T. B. , Billingham M. , Boström C. , Calvert E. , et al. 2004 . North Atlantic phylogeography and large-scale population differentiation of the seagrass Zostera marina L . Molecular Ecology , 13 : 1923 – 1941 . Google Scholar CrossRef Search ADS PubMed Orlova M. I. , Panov V. E. 2004 . Establishment of the zebra mussel, Dreissena polymorpha (Pallas), in the Neva Estuary (Gulf of Finland, Baltic Sea): distribution, population structure and possible impact on local un-ionid bivalves . Hydrobiologia , 514 : 207 – 217 . Google Scholar CrossRef Search ADS Peterson A. , Herkül K. 2017 . Mapping benthic biodiversity using georeferenced environmental data and predictive modeling . Marine Biodiversity , 1 – 16 . Plus M. , Deslous-Paoli J.-M. , Dagault F. 2003 . Seagrass (Zostera marina L.) bed recolonisation after anoxia-induced full mortality . Aquatic Botany , 77 : 121 – 134 . Google Scholar CrossRef Search ADS R Core Team . 2015 . R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org/ (last accessed 23 July 2018). Rajasilta M. , Laine P. , Eklund J. 2006 . Mortality of herring eggs on different algal substrates (Furcellaria spp. and Cladophora spp.) in the Baltic Sea—an experimental study . Hydrobiologia , 554 : 127 – 130 . Google Scholar CrossRef Search ADS Rayment W. J. 2008 . Furcellaria Lumbricalis a Red Seaweed. In Marine Life Information Network: Biology and Sensitivity Key Information Reviews, [Online] . Ed. by Tyler-Walters H. , Hiscock K. . Marine Biological Association of the United Kingdom , Plymouth . http://www.marlin.ac.uk/species/detail/1616 (last accessed 7 December 2017). Ridgeway G. 2007 . Generalized boosted models: a guide to the gbm package. http://www.saedsayad.com/docs/gbm2.pdf (last accessed 7 December 2017). Robinson L. M. , Elith J. , Hobday A. J. , Pearson R. G. , Kendal B. E. , Possingham H. P. , Richardson A. J. 2011 . Pushing the limits in marine species distribution modelling: lessons from the land present challenges and opportunities . Global Ecology and Biogeography , 20 : 789 – 802 . Google Scholar CrossRef Search ADS Rothäusler E. , Corell H. , Jormalainen V. 2015 . Abundance and dispersal trajectories of floating Fucus vesiculosus in the Northern Baltic Sea . Limnology and Oceanography , 60 : 2173 – 2184 . Google Scholar CrossRef Search ADS Sabol B. M. , Shafer D. J. , Lord E. 2005 . Dredging Effects on Eelgrass (Zostera marina) in a New England Small Boat Harbor . Journal of Marine Environmental Engineering , 8 : 57 – 81 . Schmieder K. , Werner S. , Bauer H. G. 2006 . Submersed macrophytes as a food source for wintering waterbirds at Lake Constance . Aquatic Botany , 84 : 245 – 250 . Google Scholar CrossRef Search ADS Selkoe K. A. , Halpern B. S. , Ebert C. M. , Franklin E. C. , Selig E. R. , Casey K. S. , Bruno J. et al. 2009 . A map of human impacts to a ‘‘pristine’’ coral reef ecosystem, the Papahānaumokuākea. Marine National Monument . Coral Reefs , 28 : 635 – 650 . Google Scholar CrossRef Search ADS Serrão E. A. , Brawley S. H. , Hedman J. , Kautsky L. , Samuelsson G. 1999 . Reproductive success of Fucus vesiculosus (Phaeophyceae) in the Baltic Sea . Journal of Phycology , 35 : 254 – 269 . Google Scholar CrossRef Search ADS Silberhorn G. M. , Dewing S. , Mason P. A. 1996 . Production of reproductive shoots, vegetative shoots, and seeds in populations of Ruppia maritima L. from the Chesapeake Bay, Virginia . Wetlands , 16 : 232 – 239 . Google Scholar CrossRef Search ADS Snoeijs-Leijonmalm P. , Schubert H. , Radziejewska T. 2017 . Biological Oceanography of the Baltic Sea . Springer , Netherlands . Steinhardt T. , Selig U. 2007 . Spatial distribution patterns and relationship between recent vegetation and diaspore bank of a brackish coastal lagoon on the southern Baltic Sea . Estuarine, Coastal and Shelf Science , 74 : 205 – 214 . Google Scholar CrossRef Search ADS Stelzenmüller V. , Fock H. O. , Gimpel A. , Rambo H. , Diekmann R. , Probst W. N. , Callies U. et al. 2015 . Quantitative environmental risk assessments in the context of marine spatial management: current approaches and some perspectives . ICES Journal of Marine Science , 72 : 1022 – 1042 . Google Scholar CrossRef Search ADS Suursaar Ü. , Alari V. , Tõnisson H. 2014 . Multi-scale analysis of wave conditions and coastal changes in the north-eastern Baltic Sea . Journal of Coastal Research , 70 : 223 – 228 . Google Scholar CrossRef Search ADS Torn K. , Martin G. , Kotta J. , Kupp M. 2010 . Effects of different types of mechanical disturbances on a charophyte dominated macrophyte community . Estuarine, Coastal and Shelf Science , 87 : 27 – 32 . Google Scholar CrossRef Search ADS Van Colen C. , Montserrat F. , Vincx M. , Herman P. M. J. , Ysebaert T. , Degraer S. 2008 . Macrobenthic recovery from hypoxia in an estuarine tidal mudflat . Marine Ecology Progress Series , 372 : 31 – 42 . Google Scholar CrossRef Search ADS Van den Berg M. S. , Coops H. , Simons J. 2001 . Propagule bank buildup of Chara aspera and its significance for colonization of a shallow lake . Hydrobiologia , 462 : 9 – 17 . Google Scholar CrossRef Search ADS Van Katwijk M. M. , Hermus D. C. R. , de Jong D. J. , Asmus R. M. , de Jonge V. R. 2000 . Habitat suitability of the Wadden Sea for restoration of Zostera marina beds . Helgoland Marine Research , 54 : 117 – 128 . Google Scholar CrossRef Search ADS Vári A. , Tóth V. R. 2017 . Quantifying macrophyte colonisation strategies—A field experiment in a shallow lake (Lake Balaton, Hungary) . Aquatic Botany , 136 : 56 – 60 . Google Scholar CrossRef Search ADS Villa F. , McLeod H. 2002 . Environmental vulnerability indicators for environmental planning and decision-making: guidelines and applications . Environmental Management , 29 : 335 – 348 . Google Scholar CrossRef Search ADS PubMed Wikström S. A. , Kautsky L. 2007 . Structure and diversity of invertebrate communities in the presence and absence of canopy-forming Fucus vesiculosus in the Baltic Sea . Estuarine, Coastal and Shelf Science , 72 : 168 – 176 . Google Scholar CrossRef Search ADS Zettler M. L. , Karlsson A. , Kontula T. , Gruszka P. , Laine A. O. , Herkül K. , Schiele K. S. et al. 2013 . Biodiversity gradient in the Baltic Sea: a comprehensive inventory of macrozoobenthos data . Helgoland Marine Research , 68 : 49 – 57 . Google Scholar CrossRef Search ADS Appendix Appendix 1 Table A1. List of georeferenced environmental variables that were used in the modelling of NVs and species richness in Estonia. Variable Abbreviation Source Water depth depth 1 Average water depth in 2000 m radius depth2 1 Slope of seabed slope 1 Slope of seabed in 2000 m radius slope2 1 Salinity salinity 2, 4 Wave exposure based on simplified wave model wave 5 Chlorophyll a content of sea surface based on satellite imagery chl 2 Water transparency estimated as attenuation coefficient based on satellite imagery attenuation 2 Ice coverage ice 6 Water temperature in cold season tempcold 3 Water temperature in warm season tempwarm 3 Current velocity current 3 Orbital speed of water movement at seabed induced by wind waves orbspeed 7 Proportion of soft sediment softsed 2 Secchi depth secchi 2 Concentration of ammonium ammonium 3 Concentration of nitrates nitrate 3 Concentration of phosphates phosphate 3 Sources: 1. Bathymetric data by Estonian Maritime Administration 2. Databases of the Estonian Marine Institute, University of Tartu 3. Hydrographic model developed by the Marine Systems Institute, Tallinn University of Technology (Maljutenko and Raudsepp, 2014) 4. COHERENS ocean circulation model (Bendtsen et al., 2009) 5. Simplified wave model based on fetch and wind data (Nikolopoulos and Isæus, 2008) 6. Finnish Meteorological Institute 7. SWAN hydrodynamic model (Suursaar et al., 2014) Variable Abbreviation Source Water depth depth 1 Average water depth in 2000 m radius depth2 1 Slope of seabed slope 1 Slope of seabed in 2000 m radius slope2 1 Salinity salinity 2, 4 Wave exposure based on simplified wave model wave 5 Chlorophyll a content of sea surface based on satellite imagery chl 2 Water transparency estimated as attenuation coefficient based on satellite imagery attenuation 2 Ice coverage ice 6 Water temperature in cold season tempcold 3 Water temperature in warm season tempwarm 3 Current velocity current 3 Orbital speed of water movement at seabed induced by wind waves orbspeed 7 Proportion of soft sediment softsed 2 Secchi depth secchi 2 Concentration of ammonium ammonium 3 Concentration of nitrates nitrate 3 Concentration of phosphates phosphate 3 Sources: 1. Bathymetric data by Estonian Maritime Administration 2. Databases of the Estonian Marine Institute, University of Tartu 3. Hydrographic model developed by the Marine Systems Institute, Tallinn University of Technology (Maljutenko and Raudsepp, 2014) 4. COHERENS ocean circulation model (Bendtsen et al., 2009) 5. Simplified wave model based on fetch and wind data (Nikolopoulos and Isæus, 2008) 6. Finnish Meteorological Institute 7. SWAN hydrodynamic model (Suursaar et al., 2014) View Large Table A1. List of georeferenced environmental variables that were used in the modelling of NVs and species richness in Estonia. Variable Abbreviation Source Water depth depth 1 Average water depth in 2000 m radius depth2 1 Slope of seabed slope 1 Slope of seabed in 2000 m radius slope2 1 Salinity salinity 2, 4 Wave exposure based on simplified wave model wave 5 Chlorophyll a content of sea surface based on satellite imagery chl 2 Water transparency estimated as attenuation coefficient based on satellite imagery attenuation 2 Ice coverage ice 6 Water temperature in cold season tempcold 3 Water temperature in warm season tempwarm 3 Current velocity current 3 Orbital speed of water movement at seabed induced by wind waves orbspeed 7 Proportion of soft sediment softsed 2 Secchi depth secchi 2 Concentration of ammonium ammonium 3 Concentration of nitrates nitrate 3 Concentration of phosphates phosphate 3 Sources: 1. Bathymetric data by Estonian Maritime Administration 2. Databases of the Estonian Marine Institute, University of Tartu 3. Hydrographic model developed by the Marine Systems Institute, Tallinn University of Technology (Maljutenko and Raudsepp, 2014) 4. COHERENS ocean circulation model (Bendtsen et al., 2009) 5. Simplified wave model based on fetch and wind data (Nikolopoulos and Isæus, 2008) 6. Finnish Meteorological Institute 7. SWAN hydrodynamic model (Suursaar et al., 2014) Variable Abbreviation Source Water depth depth 1 Average water depth in 2000 m radius depth2 1 Slope of seabed slope 1 Slope of seabed in 2000 m radius slope2 1 Salinity salinity 2, 4 Wave exposure based on simplified wave model wave 5 Chlorophyll a content of sea surface based on satellite imagery chl 2 Water transparency estimated as attenuation coefficient based on satellite imagery attenuation 2 Ice coverage ice 6 Water temperature in cold season tempcold 3 Water temperature in warm season tempwarm 3 Current velocity current 3 Orbital speed of water movement at seabed induced by wind waves orbspeed 7 Proportion of soft sediment softsed 2 Secchi depth secchi 2 Concentration of ammonium ammonium 3 Concentration of nitrates nitrate 3 Concentration of phosphates phosphate 3 Sources: 1. Bathymetric data by Estonian Maritime Administration 2. Databases of the Estonian Marine Institute, University of Tartu 3. Hydrographic model developed by the Marine Systems Institute, Tallinn University of Technology (Maljutenko and Raudsepp, 2014) 4. COHERENS ocean circulation model (Bendtsen et al., 2009) 5. Simplified wave model based on fetch and wind data (Nikolopoulos and Isæus, 2008) 6. Finnish Meteorological Institute 7. SWAN hydrodynamic model (Suursaar et al., 2014) View Large Appendix 2 Table A2. List of georeferenced environmental variables that were used in the modelling of NVs and species richness in Finland. Variable Abbreviation Source Water depth depth 1 Slope of seabed slope 1 Salinity on the surface sursal 1 Salinity on the bottom botsal 1 Euphotic depth Zeu_mean 1 Concentration of oxygen on the bottom oxygen 1 Concentration of nitrogen on the bottom botnit 1 Concentration of phosphorus on the bottom botphos 1 Minimum temperature on the bottom botTmin 1 Mean temperature on the bottom botTmean 1 Maximum temperature on the bottom botTmax 1 Depth attenuated wave exposure depth_expo 1 Potential reefs reefs 2 Distance to sandy shore distsand 1 Distance to river mouth distriver 1 Distance to harbour distharbor 1 Topographical position index 5 × 5 m TPI5x5 1 Probability of bottom being unstable/soft unstable 1 Hard/soft bottom substrate hardsoft 1 Bottom ruggedness ruggedbot 1 East-west position xc 1 North-south position yc 1 Sources: 1. Finnish Environment Institute SYKE 2. Geological Survey of Finland Variable Abbreviation Source Water depth depth 1 Slope of seabed slope 1 Salinity on the surface sursal 1 Salinity on the bottom botsal 1 Euphotic depth Zeu_mean 1 Concentration of oxygen on the bottom oxygen 1 Concentration of nitrogen on the bottom botnit 1 Concentration of phosphorus on the bottom botphos 1 Minimum temperature on the bottom botTmin 1 Mean temperature on the bottom botTmean 1 Maximum temperature on the bottom botTmax 1 Depth attenuated wave exposure depth_expo 1 Potential reefs reefs 2 Distance to sandy shore distsand 1 Distance to river mouth distriver 1 Distance to harbour distharbor 1 Topographical position index 5 × 5 m TPI5x5 1 Probability of bottom being unstable/soft unstable 1 Hard/soft bottom substrate hardsoft 1 Bottom ruggedness ruggedbot 1 East-west position xc 1 North-south position yc 1 Sources: 1. Finnish Environment Institute SYKE 2. Geological Survey of Finland View Large Table A2. List of georeferenced environmental variables that were used in the modelling of NVs and species richness in Finland. Variable Abbreviation Source Water depth depth 1 Slope of seabed slope 1 Salinity on the surface sursal 1 Salinity on the bottom botsal 1 Euphotic depth Zeu_mean 1 Concentration of oxygen on the bottom oxygen 1 Concentration of nitrogen on the bottom botnit 1 Concentration of phosphorus on the bottom botphos 1 Minimum temperature on the bottom botTmin 1 Mean temperature on the bottom botTmean 1 Maximum temperature on the bottom botTmax 1 Depth attenuated wave exposure depth_expo 1 Potential reefs reefs 2 Distance to sandy shore distsand 1 Distance to river mouth distriver 1 Distance to harbour distharbor 1 Topographical position index 5 × 5 m TPI5x5 1 Probability of bottom being unstable/soft unstable 1 Hard/soft bottom substrate hardsoft 1 Bottom ruggedness ruggedbot 1 East-west position xc 1 North-south position yc 1 Sources: 1. Finnish Environment Institute SYKE 2. Geological Survey of Finland Variable Abbreviation Source Water depth depth 1 Slope of seabed slope 1 Salinity on the surface sursal 1 Salinity on the bottom botsal 1 Euphotic depth Zeu_mean 1 Concentration of oxygen on the bottom oxygen 1 Concentration of nitrogen on the bottom botnit 1 Concentration of phosphorus on the bottom botphos 1 Minimum temperature on the bottom botTmin 1 Mean temperature on the bottom botTmean 1 Maximum temperature on the bottom botTmax 1 Depth attenuated wave exposure depth_expo 1 Potential reefs reefs 2 Distance to sandy shore distsand 1 Distance to river mouth distriver 1 Distance to harbour distharbor 1 Topographical position index 5 × 5 m TPI5x5 1 Probability of bottom being unstable/soft unstable 1 Hard/soft bottom substrate hardsoft 1 Bottom ruggedness ruggedbot 1 East-west position xc 1 North-south position yc 1 Sources: 1. Finnish Environment Institute SYKE 2. Geological Survey of Finland View Large Appendix 3 Table A3. The recovery classes and the respective sensitivity coefficients of NVs together with a short explanation and reference. Species/group Recovery class (years) Coefficient in calculations Rationale, references Fucus vesiculosus 2–3 2 As northern Baltic Sea littoral habitats are well connected (Rothäusler et al., 2015), recolonization will possibly occur during the next reproduction period (after complete removal, if conditions for growth are adequate and substrate present), but gamete dispersal is rather limited and can slow recovery (Serrão et al., 1999). Reaching to canopy state prior to removal, takes time. It took about 2 years (in Nova Scotia) after ice scouring completely eliminated fucoid assemblages (incl. F. vesiculosus) to re-colonize and fully recover (to canopy state similar to pre-scouring; recolonization took less than 7 months; Minchinton et al., 1997). Furcellaria lumbricalis 5–10 4 Recovers slowly due to low growth rate (Bird et al., 1979; Martin et al., 2006), long time to reach maturity (5 years in Wales; Austin, 1960) and recruitment that usually occurs in the vicinity of parent plant (Rayment, 2008). In addition, due to low salinity in the Gulf of Finland, vegetative reproduction prevails (Kostamo and Mäkinen, 2006) rendering the recovery even slower. Filamentous algae <2 1 After ice scouring that completely removed algae, it took less than few months to establish filamentous algae cover on rocks (including, among others: Pilayella littoralis, Polysiphonia sp., Ectocarpus sp., Ceramium virgatum, Cladophora sp.), in Nova Scotia (Minchinton et al., 1997). Artificial substrates became colonized by Pilayella littoralis in 3 months during winter and within one month in spring (Kraufvelin et al., 2007). Also other filamentous algae colonized artificial substrates within a year (during the first spring/summer; Kraufvelin et al., 2007). Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3-5 3 After ice scouring that completely removed bivalves, Mytilus spp. recolonized substrates in about 1-1.5 years; distribution similar to pre-scouring was observed within 5 years (Nova Scotia; Minchinton et al., 1997). While growth rates are lower in the Gulf of Finland than in optimal conditions (Kautsky, 1982a), recruitment is possible all year round in the Baltic Sea (Kautsky, 1982b).D. polymorpha: high fecundity, good dispersal abilities and rather fast growth rates (Mackie et al., 1989) support fast recovery. Becomes sexually mature in second year of life (Mackie et al., 1989). However, recovery may be slowed in the Gulf of Finland by limited or low spawning/recruitment in years with cold summers (Orlova and Panov, 2004). Vascular plants (excl. Zostera marina) 3-5 3 Ruppia maritima (and Najas marina)—annual species with short life cycle; high production rate (Kautsky, 1988) and high seed production (Silberhorn et al., 1996) indicates fast recovery. Stuckenia pectinata was successfully established in habitats created 3-5 years ago (Boedeltje et al., 2001).In Lake Balaton vascular plants (that occur also in the Baltic Sea/ Gulf of Finland) colonized rapidly de-vegetated areas via rhizomes, fragments of plants etc. (i.e. vegetative reproduction) from adjacent vegetated areas (Vári and Tóth, 2017), also in a Danish lake submerged macrophytes recolonized the lake within 5 years (up to 90% coverage; Lauridsen et al., 1994).Zannichellia palustris have a wide-ranging generative recolonization potential (Steinhardt and Selig, 2007).The dispersal and recolonization of aquatic plants and charophytes are encouraged by local propagule banks (Steinhardt and Selig, 2007) and waterbirds transporting the seeds (mostly in their guts) especially on local scale (Green et al., 2002). Zostera marina >10 5 Recolonization after total loss (i.e. no seed bank) can be extremely difficult (Holt et al., 1995). If there is adequate seed bank in sediments: recolonization (after anoxic event) was observed during next summer, but seedling mortality is huge (∼99%), so recovery takes definitely several years (Greve et al., 2005). In France it took less than 9 months till biomasses similar to pre-destruction and 2 years till flowering (Plus et al., 2003), but here, in colder Gulf of Finland eelgrass grows slower (Boström et al., 2014). In the northern Baltic Sea Z. marina commonly reproduces vegetatively (Olsen et al., 2004), thus there is no seed bank in sediments. Therefore, re-establishment will be very slow due to very limited vegetative dispersal (Holt et al., 1995) and possibly impoverished gene bank (Boström et al., 2014), and lack of suitable genotypes (as hypothesized for lower zones in Wadden Sea: Van Katwijk et al., 2000). No recovery was observed two years post-dredging in New England (seed bank removed together with sediment, but Z. marina growing in the area; Sabol et al., 2005). Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Chara vulgaris was successfully established in newly created habitats in less than 3 years (Boedeltje et al., 2001). If there is a sufficient oospore bank in sediment, it may greatly enhance recolonization (C. aspera in shallow lake; Van den Berg et al., 2001). However, charophyte recovery (by biomass) can take more than 2 years (Torn et al., 2010). C. tomentosa has not recolonized Byviken in Hanko, southwestern Finland since the dredging that took place many years ago. Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 In defaunated areas, infaunal bivalves (incl. Limecola balthica, Mya arenaria, and Cerastoderma edule) biomass recovery takes several years (Beukema et al., 1999), but abundance reached to similar values as in undisturbed areas within 1 year (at least 1 summer needed) or even faster (juvenile abundance; Van Colen et al., 2008). Seals >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). Birds >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). Species/group Recovery class (years) Coefficient in calculations Rationale, references Fucus vesiculosus 2–3 2 As northern Baltic Sea littoral habitats are well connected (Rothäusler et al., 2015), recolonization will possibly occur during the next reproduction period (after complete removal, if conditions for growth are adequate and substrate present), but gamete dispersal is rather limited and can slow recovery (Serrão et al., 1999). Reaching to canopy state prior to removal, takes time. It took about 2 years (in Nova Scotia) after ice scouring completely eliminated fucoid assemblages (incl. F. vesiculosus) to re-colonize and fully recover (to canopy state similar to pre-scouring; recolonization took less than 7 months; Minchinton et al., 1997). Furcellaria lumbricalis 5–10 4 Recovers slowly due to low growth rate (Bird et al., 1979; Martin et al., 2006), long time to reach maturity (5 years in Wales; Austin, 1960) and recruitment that usually occurs in the vicinity of parent plant (Rayment, 2008). In addition, due to low salinity in the Gulf of Finland, vegetative reproduction prevails (Kostamo and Mäkinen, 2006) rendering the recovery even slower. Filamentous algae <2 1 After ice scouring that completely removed algae, it took less than few months to establish filamentous algae cover on rocks (including, among others: Pilayella littoralis, Polysiphonia sp., Ectocarpus sp., Ceramium virgatum, Cladophora sp.), in Nova Scotia (Minchinton et al., 1997). Artificial substrates became colonized by Pilayella littoralis in 3 months during winter and within one month in spring (Kraufvelin et al., 2007). Also other filamentous algae colonized artificial substrates within a year (during the first spring/summer; Kraufvelin et al., 2007). Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3-5 3 After ice scouring that completely removed bivalves, Mytilus spp. recolonized substrates in about 1-1.5 years; distribution similar to pre-scouring was observed within 5 years (Nova Scotia; Minchinton et al., 1997). While growth rates are lower in the Gulf of Finland than in optimal conditions (Kautsky, 1982a), recruitment is possible all year round in the Baltic Sea (Kautsky, 1982b).D. polymorpha: high fecundity, good dispersal abilities and rather fast growth rates (Mackie et al., 1989) support fast recovery. Becomes sexually mature in second year of life (Mackie et al., 1989). However, recovery may be slowed in the Gulf of Finland by limited or low spawning/recruitment in years with cold summers (Orlova and Panov, 2004). Vascular plants (excl. Zostera marina) 3-5 3 Ruppia maritima (and Najas marina)—annual species with short life cycle; high production rate (Kautsky, 1988) and high seed production (Silberhorn et al., 1996) indicates fast recovery. Stuckenia pectinata was successfully established in habitats created 3-5 years ago (Boedeltje et al., 2001).In Lake Balaton vascular plants (that occur also in the Baltic Sea/ Gulf of Finland) colonized rapidly de-vegetated areas via rhizomes, fragments of plants etc. (i.e. vegetative reproduction) from adjacent vegetated areas (Vári and Tóth, 2017), also in a Danish lake submerged macrophytes recolonized the lake within 5 years (up to 90% coverage; Lauridsen et al., 1994).Zannichellia palustris have a wide-ranging generative recolonization potential (Steinhardt and Selig, 2007).The dispersal and recolonization of aquatic plants and charophytes are encouraged by local propagule banks (Steinhardt and Selig, 2007) and waterbirds transporting the seeds (mostly in their guts) especially on local scale (Green et al., 2002). Zostera marina >10 5 Recolonization after total loss (i.e. no seed bank) can be extremely difficult (Holt et al., 1995). If there is adequate seed bank in sediments: recolonization (after anoxic event) was observed during next summer, but seedling mortality is huge (∼99%), so recovery takes definitely several years (Greve et al., 2005). In France it took less than 9 months till biomasses similar to pre-destruction and 2 years till flowering (Plus et al., 2003), but here, in colder Gulf of Finland eelgrass grows slower (Boström et al., 2014). In the northern Baltic Sea Z. marina commonly reproduces vegetatively (Olsen et al., 2004), thus there is no seed bank in sediments. Therefore, re-establishment will be very slow due to very limited vegetative dispersal (Holt et al., 1995) and possibly impoverished gene bank (Boström et al., 2014), and lack of suitable genotypes (as hypothesized for lower zones in Wadden Sea: Van Katwijk et al., 2000). No recovery was observed two years post-dredging in New England (seed bank removed together with sediment, but Z. marina growing in the area; Sabol et al., 2005). Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Chara vulgaris was successfully established in newly created habitats in less than 3 years (Boedeltje et al., 2001). If there is a sufficient oospore bank in sediment, it may greatly enhance recolonization (C. aspera in shallow lake; Van den Berg et al., 2001). However, charophyte recovery (by biomass) can take more than 2 years (Torn et al., 2010). C. tomentosa has not recolonized Byviken in Hanko, southwestern Finland since the dredging that took place many years ago. Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 In defaunated areas, infaunal bivalves (incl. Limecola balthica, Mya arenaria, and Cerastoderma edule) biomass recovery takes several years (Beukema et al., 1999), but abundance reached to similar values as in undisturbed areas within 1 year (at least 1 summer needed) or even faster (juvenile abundance; Van Colen et al., 2008). Seals >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). Birds >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). View Large Table A3. The recovery classes and the respective sensitivity coefficients of NVs together with a short explanation and reference. Species/group Recovery class (years) Coefficient in calculations Rationale, references Fucus vesiculosus 2–3 2 As northern Baltic Sea littoral habitats are well connected (Rothäusler et al., 2015), recolonization will possibly occur during the next reproduction period (after complete removal, if conditions for growth are adequate and substrate present), but gamete dispersal is rather limited and can slow recovery (Serrão et al., 1999). Reaching to canopy state prior to removal, takes time. It took about 2 years (in Nova Scotia) after ice scouring completely eliminated fucoid assemblages (incl. F. vesiculosus) to re-colonize and fully recover (to canopy state similar to pre-scouring; recolonization took less than 7 months; Minchinton et al., 1997). Furcellaria lumbricalis 5–10 4 Recovers slowly due to low growth rate (Bird et al., 1979; Martin et al., 2006), long time to reach maturity (5 years in Wales; Austin, 1960) and recruitment that usually occurs in the vicinity of parent plant (Rayment, 2008). In addition, due to low salinity in the Gulf of Finland, vegetative reproduction prevails (Kostamo and Mäkinen, 2006) rendering the recovery even slower. Filamentous algae <2 1 After ice scouring that completely removed algae, it took less than few months to establish filamentous algae cover on rocks (including, among others: Pilayella littoralis, Polysiphonia sp., Ectocarpus sp., Ceramium virgatum, Cladophora sp.), in Nova Scotia (Minchinton et al., 1997). Artificial substrates became colonized by Pilayella littoralis in 3 months during winter and within one month in spring (Kraufvelin et al., 2007). Also other filamentous algae colonized artificial substrates within a year (during the first spring/summer; Kraufvelin et al., 2007). Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3-5 3 After ice scouring that completely removed bivalves, Mytilus spp. recolonized substrates in about 1-1.5 years; distribution similar to pre-scouring was observed within 5 years (Nova Scotia; Minchinton et al., 1997). While growth rates are lower in the Gulf of Finland than in optimal conditions (Kautsky, 1982a), recruitment is possible all year round in the Baltic Sea (Kautsky, 1982b).D. polymorpha: high fecundity, good dispersal abilities and rather fast growth rates (Mackie et al., 1989) support fast recovery. Becomes sexually mature in second year of life (Mackie et al., 1989). However, recovery may be slowed in the Gulf of Finland by limited or low spawning/recruitment in years with cold summers (Orlova and Panov, 2004). Vascular plants (excl. Zostera marina) 3-5 3 Ruppia maritima (and Najas marina)—annual species with short life cycle; high production rate (Kautsky, 1988) and high seed production (Silberhorn et al., 1996) indicates fast recovery. Stuckenia pectinata was successfully established in habitats created 3-5 years ago (Boedeltje et al., 2001).In Lake Balaton vascular plants (that occur also in the Baltic Sea/ Gulf of Finland) colonized rapidly de-vegetated areas via rhizomes, fragments of plants etc. (i.e. vegetative reproduction) from adjacent vegetated areas (Vári and Tóth, 2017), also in a Danish lake submerged macrophytes recolonized the lake within 5 years (up to 90% coverage; Lauridsen et al., 1994).Zannichellia palustris have a wide-ranging generative recolonization potential (Steinhardt and Selig, 2007).The dispersal and recolonization of aquatic plants and charophytes are encouraged by local propagule banks (Steinhardt and Selig, 2007) and waterbirds transporting the seeds (mostly in their guts) especially on local scale (Green et al., 2002). Zostera marina >10 5 Recolonization after total loss (i.e. no seed bank) can be extremely difficult (Holt et al., 1995). If there is adequate seed bank in sediments: recolonization (after anoxic event) was observed during next summer, but seedling mortality is huge (∼99%), so recovery takes definitely several years (Greve et al., 2005). In France it took less than 9 months till biomasses similar to pre-destruction and 2 years till flowering (Plus et al., 2003), but here, in colder Gulf of Finland eelgrass grows slower (Boström et al., 2014). In the northern Baltic Sea Z. marina commonly reproduces vegetatively (Olsen et al., 2004), thus there is no seed bank in sediments. Therefore, re-establishment will be very slow due to very limited vegetative dispersal (Holt et al., 1995) and possibly impoverished gene bank (Boström et al., 2014), and lack of suitable genotypes (as hypothesized for lower zones in Wadden Sea: Van Katwijk et al., 2000). No recovery was observed two years post-dredging in New England (seed bank removed together with sediment, but Z. marina growing in the area; Sabol et al., 2005). Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Chara vulgaris was successfully established in newly created habitats in less than 3 years (Boedeltje et al., 2001). If there is a sufficient oospore bank in sediment, it may greatly enhance recolonization (C. aspera in shallow lake; Van den Berg et al., 2001). However, charophyte recovery (by biomass) can take more than 2 years (Torn et al., 2010). C. tomentosa has not recolonized Byviken in Hanko, southwestern Finland since the dredging that took place many years ago. Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 In defaunated areas, infaunal bivalves (incl. Limecola balthica, Mya arenaria, and Cerastoderma edule) biomass recovery takes several years (Beukema et al., 1999), but abundance reached to similar values as in undisturbed areas within 1 year (at least 1 summer needed) or even faster (juvenile abundance; Van Colen et al., 2008). Seals >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). Birds >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). Species/group Recovery class (years) Coefficient in calculations Rationale, references Fucus vesiculosus 2–3 2 As northern Baltic Sea littoral habitats are well connected (Rothäusler et al., 2015), recolonization will possibly occur during the next reproduction period (after complete removal, if conditions for growth are adequate and substrate present), but gamete dispersal is rather limited and can slow recovery (Serrão et al., 1999). Reaching to canopy state prior to removal, takes time. It took about 2 years (in Nova Scotia) after ice scouring completely eliminated fucoid assemblages (incl. F. vesiculosus) to re-colonize and fully recover (to canopy state similar to pre-scouring; recolonization took less than 7 months; Minchinton et al., 1997). Furcellaria lumbricalis 5–10 4 Recovers slowly due to low growth rate (Bird et al., 1979; Martin et al., 2006), long time to reach maturity (5 years in Wales; Austin, 1960) and recruitment that usually occurs in the vicinity of parent plant (Rayment, 2008). In addition, due to low salinity in the Gulf of Finland, vegetative reproduction prevails (Kostamo and Mäkinen, 2006) rendering the recovery even slower. Filamentous algae <2 1 After ice scouring that completely removed algae, it took less than few months to establish filamentous algae cover on rocks (including, among others: Pilayella littoralis, Polysiphonia sp., Ectocarpus sp., Ceramium virgatum, Cladophora sp.), in Nova Scotia (Minchinton et al., 1997). Artificial substrates became colonized by Pilayella littoralis in 3 months during winter and within one month in spring (Kraufvelin et al., 2007). Also other filamentous algae colonized artificial substrates within a year (during the first spring/summer; Kraufvelin et al., 2007). Epibenthic bivalves (Mytilus trossulus, Dreissena polymorpha) 3-5 3 After ice scouring that completely removed bivalves, Mytilus spp. recolonized substrates in about 1-1.5 years; distribution similar to pre-scouring was observed within 5 years (Nova Scotia; Minchinton et al., 1997). While growth rates are lower in the Gulf of Finland than in optimal conditions (Kautsky, 1982a), recruitment is possible all year round in the Baltic Sea (Kautsky, 1982b).D. polymorpha: high fecundity, good dispersal abilities and rather fast growth rates (Mackie et al., 1989) support fast recovery. Becomes sexually mature in second year of life (Mackie et al., 1989). However, recovery may be slowed in the Gulf of Finland by limited or low spawning/recruitment in years with cold summers (Orlova and Panov, 2004). Vascular plants (excl. Zostera marina) 3-5 3 Ruppia maritima (and Najas marina)—annual species with short life cycle; high production rate (Kautsky, 1988) and high seed production (Silberhorn et al., 1996) indicates fast recovery. Stuckenia pectinata was successfully established in habitats created 3-5 years ago (Boedeltje et al., 2001).In Lake Balaton vascular plants (that occur also in the Baltic Sea/ Gulf of Finland) colonized rapidly de-vegetated areas via rhizomes, fragments of plants etc. (i.e. vegetative reproduction) from adjacent vegetated areas (Vári and Tóth, 2017), also in a Danish lake submerged macrophytes recolonized the lake within 5 years (up to 90% coverage; Lauridsen et al., 1994).Zannichellia palustris have a wide-ranging generative recolonization potential (Steinhardt and Selig, 2007).The dispersal and recolonization of aquatic plants and charophytes are encouraged by local propagule banks (Steinhardt and Selig, 2007) and waterbirds transporting the seeds (mostly in their guts) especially on local scale (Green et al., 2002). Zostera marina >10 5 Recolonization after total loss (i.e. no seed bank) can be extremely difficult (Holt et al., 1995). If there is adequate seed bank in sediments: recolonization (after anoxic event) was observed during next summer, but seedling mortality is huge (∼99%), so recovery takes definitely several years (Greve et al., 2005). In France it took less than 9 months till biomasses similar to pre-destruction and 2 years till flowering (Plus et al., 2003), but here, in colder Gulf of Finland eelgrass grows slower (Boström et al., 2014). In the northern Baltic Sea Z. marina commonly reproduces vegetatively (Olsen et al., 2004), thus there is no seed bank in sediments. Therefore, re-establishment will be very slow due to very limited vegetative dispersal (Holt et al., 1995) and possibly impoverished gene bank (Boström et al., 2014), and lack of suitable genotypes (as hypothesized for lower zones in Wadden Sea: Van Katwijk et al., 2000). No recovery was observed two years post-dredging in New England (seed bank removed together with sediment, but Z. marina growing in the area; Sabol et al., 2005). Charophytes (Chara spp., Tolypella nidifica) 2–3 2 Chara vulgaris was successfully established in newly created habitats in less than 3 years (Boedeltje et al., 2001). If there is a sufficient oospore bank in sediment, it may greatly enhance recolonization (C. aspera in shallow lake; Van den Berg et al., 2001). However, charophyte recovery (by biomass) can take more than 2 years (Torn et al., 2010). C. tomentosa has not recolonized Byviken in Hanko, southwestern Finland since the dredging that took place many years ago. Infaunal bivalves (Limecola balthica, Cerastoderma glaucum, Mya areanaria) 2–3 2 In defaunated areas, infaunal bivalves (incl. Limecola balthica, Mya arenaria, and Cerastoderma edule) biomass recovery takes several years (Beukema et al., 1999), but abundance reached to similar values as in undisturbed areas within 1 year (at least 1 summer needed) or even faster (juvenile abundance; Van Colen et al., 2008). Seals >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). Birds >10 5 Long-lived organisms with low fecundity and late reproductive maturity (compared with other organisms in the table). View Large © International Council for the Exploration of the Sea 2018. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

ICES Journal of Marine ScienceOxford University Press

Published: Aug 10, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off