Modeling of Water Quality Indicators in the Western Baltic Sea: Seasonal Oxygen Deficiency

Piehl, Sarah et al. Article — Published Version Modeling of Water Quality Indicators in the Western Baltic Sea: Seasonal Oxygen Deficiency Environmental Modeling & Assessment Provided in Cooperation with: Springer Nature Suggested Citation: Piehl, Sarah et al. (2022) : Modeling of Water Quality Indicators in the Western Baltic Sea: Seasonal Oxygen Deficiency, Environmental Modeling & Assessment, ISSN 1573-2967, Springer International Publishing, Cham, Vol. 28, Iss. 3, pp. 429-446, https://doi.org/10.1007/s10666-022-09866-x This Version is available at: https://hdl.handle.net/10419/308454 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Vol.:(0123456789) 1 3 Environmental Modeling & Assessment (2023) 28:429–446 https://doi.org/10.1007/s10666-022-09866-x Modeling ofWater Quality Indicators intheWestern Baltic Sea: Seasonal Oxygen Deficiency SarahPiehl1 · RenéFriedland1· BirgitHeyden2· WeraLeujak3· ThomasNeumann1· GeraldSchernewski1,4 Received: 13 December 2021 / Accepted: 30 October 2022 / Published online: 8 December 2022 © The Author(s) 2022 Abstract Hypoxia in coastal seas is a severe threat to marine ecosystems, with the Baltic Sea exhibiting the largest hypoxic areas worldwide. While perennial oxygen deficiency in the deep basins is a component of environmental assessments, seasonal oxygen deficiency in shallow areas is not routinely assessed. Current measurements alone cannot provide the spatio-temporal resolution needed for highly dynamic seasonal oxygen deficiency, making estimations on its duration and extent uncertain. Utilizing long-term 3D model simulations with a horizontal resolution of 3 nautical miles, we analyzed the development of seasonal oxygen deficiency in the western Baltic Sea. Different metrics (near-bottom area, water volume, duration, and frequency) and critical oxygen concentrations were analyzed for exemplary sub-basins as defined by the Helsinki Commission. Our results indicate that the extent of seasonal oxygen deficiency has continually increased in the second half of the twentieth century until the end of the 1980s and slightly decreased in the last two decades. In the 1950s, the spatial extent of oxygen deficiency was still at a low plateau before increasing, indicating that this period could be suitable as a reference period representing a good status, including naturally occurring oxygen deficiency. Overall, seasonal oxygen deficiency is a suitable indicator for describing the ecological status of the western Baltic Sea. For an application as eutrophication indicator in shallow areas, a sufficient spatio-temporal resolution of seasonal oxygen deficiency is needed which can be gained by a combination of model simulations and measurements. Further analysis is needed to integrate insitu measurements and model results to obtain the most reliable approach. Keywords Biogeochemical modeling· Western Baltic Sea· Seasonal hypoxia· HELCOM· MSFD· Oxygen indicator 1 Introduction The Baltic Sea exhibits the worldwide largest hypoxic areas (oxygen concentration < 2mg/l; [1] and its frequency and extent continually increased over the last century due to excess nutrient inputs [2]. Therefore, monitoring of hypoxia is of high importance and the development of an oxygen indicator current subject of the work of commissions on the protection of the Baltic Sea. Due to limited exchange with salty North Sea water, a vertical salinity stratification is characteristic for the intracontinental Baltic Sea [3]. Due to restricted ventilation with oxygen-rich surface waters, the Baltic Sea exhibits naturally occurring hypoxic areas [4]. But with its catchment area about four times larger than its surface area, the Baltic Sea is specifically influenced by eutrophication-induced hypoxia through excess nutrient inputs (nitrogen and phosphorus) from land and the atmosphere [5, 6]. Although hypoxia-stimulating nutrient inputs have decreased since the 1980s [6], significant improvements in oxygen concentrations are not yet evident [7–9]. The reasons for this are the long water residence time of about 30years [1] and feedback loops (phosphorus release under anoxic conditions,vicious circle [10]). The effects of climate change can further enhance hypoxic conditions by increasing temperature and stratification as well as salinity changes, resulting in lower solubility of oxygen and reduced vertical mixing [11–13]. * Sarah Piehl [email protected] 1 Leibniz Institute forBaltic Sea Research Warnemünde, Rostock, Germany 2 AquaEcology GmbH & Co. KG, Oldenburg, Germany 3 German Environment Agency, Dessau-Roßlau, Germany 4 Marine Research Institute ofKlaipeda University, Klaipeda, Lithuania 430 S.Piehl et al. 1 3 For the Baltic Sea, three types of hypoxia can be distinguished based on a temporal aspect: perennial hypoxia in the deep open and central part, seasonal hypoxic events during summer and autumn, and episodic hypoxic events at many shallow coastal sites [1]. Perennial hypoxia occurs in areas with a permanent halocline, where vertical mixing with oxygen-rich surface water is limited. It is suggested that perennial hypoxia in the Baltic Sea has reached its maximum areal extent [7, 8]. In contrast, seasonal and episodic hypoxia is increasingly observed in many places [14, 15]. Especially in the western Baltic Sea, phenomena of seasonal hypoxia were observed only occasionally until the 1970s, before annual observations of seasonal hypoxia were made, for example, in Kiel Bay and Bay of Mecklenburg [16]. Seasonal stratification of the water column due to decreased bottom water transport and seasonal temperature increase favors the occurrence of seasonal hypoxia [14]. In addition, specific local conditions, such as low oxygen input from the air, complex bottom topography, and rapid oxygen consumption by biological processes, are favorable for episodic hypoxic events to occur [1, 17]. Eutrophication and oxygen deficiency as an indirect effect are the most challenging environmental problems of the Baltic Sea. Negative impacts of hypoxia include altered distributions and abundances of animal populations and key changes in benthic community structures [18, 19], whereas most prominent are mass mortality events of fishes caused by hypoxia [16, 20]. In the Baltic Sea, not only spawning success of cod can be impaired by hypoxia [21]. Moreover, about 30% of total secondary production is missing due to persistent hypoxic zones [22] and therewith a potential benthic food energy for fisheries is lost. To counteract associated negative effects for the marine ecosystem, regional and (inter)national conventions (Helsinki Convention (HELCOM)) and legislations (EU Water Framework Directive (WFD, 2000/60/EC); Marine Strategy Framework Directive (MSFD, 2008/56/EC)) implemented actions against eutrophication and require their member states to reach a Good Environmental (or Ecological) Status (GES). As an indirect effect of eutrophication, dissolved oxygen in the bottom of the water column is included as a primary and thereby mandatory criterion to assess under Descriptor 5 “eutrophication” of the MSFD. For the Baltic Sea Action Plan, the achievement of “natural oxygen levels” is one of the ecological objectives in the eutrophication segment [23, 24]. More specifically, “oxygen debt” is implemented by HELCOM as core indicator to evaluate the average oxygen debt below the halocline [25], thus assessing perennial hypoxia in the deep basins. Due to increasing occurrences of seasonal hypoxic events [14], HELCOM already stated the need for an additional shallow water oxygen indicator, assessing oxygen deficiency in areas without a permanent halocline [26, 27]. Currently, the shallow water oxygen indicator has “pre-core” status but was not applied in the last holistic HELCOM assessment HOLAS II and the report on the state of the Baltic Sea due to the lack of common target values [5, 28]. Several Baltic States already implemented monitoring approaches to assess oxygen deficiency in their water bodies as demanded by EU Directives (MSFD, WFD). In contrast to the oxygen debt indicator, various oxygen thresholds from < 2 to < 6mg/l are used to describe the boundary between a good and moderate condition for the shallow water oxygen indicator [16, 29]. Volume and/or area are mostly used as metrics, whereas single-point oxygen measurements are extrapolated into space to evaluate oxygen deficiency. Currently lacking in the assessment of Baltic States, but of extreme importance to benthic communities, is the duration and frequency of hypoxia [19]. For example, in Chesapeake Bay duration of hypoxia has already been used as metric to quantify the bay-wide hypoxic areas [30]. Although the frequency and spatial coverage of oxygen measurements has improved since the 1960s [31], the assessment of highly dynamic seasonal oxygen deficiency stays uncertain due to the spatio-temporal limits of single-point measurements. Thus, the development of an adequate and comparable shallow water oxygen indicator, which targets seasonal oxygen deficiency, is still ongoing. As time and resources to conduct extensive cruises are limited, additional methods to assess seasonal oxygen deficiency are needed. Here, numerical models can help to provide high-resolution information on the frequency, the duration, and both the horizontal and vertical extent of oxygen deficiency areas. The German national working group explicitly mentioned hydrodynamic modeling as an important approach to gain information on the temporal and spatial extent of oxygen deficiency to support the MSFD criterion “dissolved oxygen in the bottom of the water column” (Commission Decision 2017/848/EU) in the future [16]. In our study, we used the coupled hydrodynamicbiogeochemical model MOM-ERGOM to gain knowledge on the spatio-temporal variability of seasonal oxygen deficiency in the western Baltic Sea and to assess various oxygen metrics. Specifically, we aim (I) to evaluate the model quality by comparing model data against insitu measurements for representative stations, (II) to analyze the spatio-temporal variability of oxygen concentrations in the western Baltic to draw conclusions for a more efficient oxygen monitoring, (III) to assess various oxygen metrics derivable from model simulation products to be used as oxygen indicators, and (IV) to analyze time-series of oxygen concentrations as a basis for the determination of reference values and related thresholds. 431Modeling ofWater Quality Indicators intheWestern Baltic Sea: Seasonal Oxygen Deficiency 1 3 2 Material andMethods 2.1 The Ecosystem Model Approach The model simulations in the Baltic Sea (Fig.1) were performed with the integrated biogeochemical model ERGOM (www. ergom. net) which is coupled to a 3D circulation model (MOM) [32, 33]. The model has previously been applied in the Baltic Sea to analyze the extent of hypoxic areas in conjunction with saltwater inflows from the North Sea, nutrient input scenarios as well as climate change scenarios [34–37]. The biogeochemical model simulates the marine nitrogen and phosphorus cycle: the three nutrients dissolved in water, ammonium, nitrate, and phosphate, which are the basis for primary production realized by three functional phytoplankton groups (large cells, small cells, and cyanobacteria). Grazing pressure on the phytoplankton is applied in the model via a dynamically developing bulk zooplankton variable. Dead organic material is considered by a detritus state variable. During sinking, part of the detritus is mineralized again into dissolved ammonium and phosphate. The portion that reaches the sea bottom accumulates and is partly buried permanently, or alternatively mineralized or resuspended when the velocity of near-bottom currents is sufficiently high. Coupled to the nitrogen and phosphorus cycle is a carbon cycle as described in Kuznetsov and Neumann [38]. Under oxic conditions, part of the mineralized phosphate is bound by iron oxides and is retained in the sediment. When conditions become anoxic, it is released to the water column. Oxygen development is coupled to biogeochemical processes via stoichiometric ratios, with oxygen levels in turn controlling processes such as denitrification and nitrification. The physical part of the model is based on the circulation model MOM (version 5.1; [39, 40]). It has been adapted to the Baltic Sea with an open boundary condition to the North Sea and riverine freshwater input. To estimate ice cover thickness and extent, the MOM model is complemented with a sea ice model [41]. The horizontal resolution of the model grid is three nautical miles, while vertically the model is Fig. 1 The Baltic Sea in northwestern Europe with sub-basin divisions according to HELCOM (black lines) and our study area including the sub-basins “Kiel Bay,” “Bay of Mecklenburg,” “Arkona Basin,” and “Pomeranian Bay” located in the shallow western Baltic Sea including measuring stations used for model validation (red dots) 432 S.Piehl et al. 1 3 resolved into 152 layers, with layer thicknesses from 0.5 to 2m. The model domain and bathymetry are shown in Fig.1. Atmospheric forcing is based on a dynamical downscaling provided by the coastDat data set [42, 43] with a grid resolution of about 25 × 25km. Nutrient loads to the Baltic Sea due to riverine discharge have been compiled based on data from HELCOM assessments (e.g., [44, 45]) and prior to 1995 based on Gustafsson etal. [46]. Atmospheric deposition of nitrogen inputs was provided by EMEP [47, 48] and prior to 1995 by Ruoho-Airola etal. [49]. 2.2 Model Performance As the model system ERGOM-MOM is largely validated against recent observations, focusing on key parameters and stations in the Baltic Sea [37, 50–53], the presented study is focused on the oxygen dynamics in the western Baltic Sea. Therefore, we compared the model data with insitu measurements for representative stations in the western Baltic Sea. Observational data was provided by local authorities (State Agency for Agriculture, Environment and Rural Areas (LLUR) and State Agency for Environment, Nature Conservation and Geology Mecklenburg-Vorpommern (LUNG)) and enhanced by data from the LeibnizInstitute for Baltic Sea Research Warnemünde (https:// odin2. iowarne muende. de/) and data from the ICES oceanographic database (https:// www. ices. dk/ data/ dataporta ls/ Pages/ ocean. aspx). Observations were checked for plausibility and only the deepest layer was selected (except for the vertical profile comparison). To analyze the seasonal cycle, four representative stations in the western Baltic Sea were selected (Fig.1), representing mostly the deeper basins (Kiel Bay, Bay of Mecklenburg, and Arkona Basin). While these stations are dominated by muddy sediments [54], for comparison a sandy station (O9, west of Hiddensee) was additionally selected. To compare the model behavior, observations from the last years (2010–2019) were condensed to multi-annual monthly means to verify that the model is capable to reproduce the seasonal cycle. To analyze the spatial agreement, the averaged annual minima from the period 2010 to 2019 of the model results and observations were compared. 2.3 Model Data Analysis The analysis of the model results was focused on the western Baltic Sea sub-basins depicted in Fig.1, following the HELCOM assessment units expanded by the newly defined Pomeranian Bay unit. A long-term model run from 1950 to 2019, with daily oxygen concentrations as output, served as basis for all analysis. The initial conditions of 1950 were taken from an earlier simulation starting in the 1850s, implying a model spin-up time of 100years. The period from 2011 to 2016 which coincides with the latest HELCOM assessment period (HOLASII) was selected to assess the spatio-temporal variability of oxygen concentrations (II) as well as to assess the various oxygen metrics (III). For the assessment of long-term changes in oxygen concentrations (IV), the full dataset from 1950 to 2019 was investigated. For the analysis of oxygen deficiency of the near-bottom area (II), we utilized the model results approximately 2.5m above the bottom. Extracted daily oxygen values were averaged either monthly, yearly, or over assessment periods of 6years. For the calculation of the standard deviation, hydrogen sulfide (as negative oxygen equivalents) was considered by subtracting twice the concentration of H2S from the oxygen values. For the analysis of the oxygen metrics (III), mean values were not based on daily oxygen concentrations but on whether oxygen concentrations did fall below critical levels (2, 4, or 6mg/l oxygen). In the following, all situations below 2mg/l will be referred to as hypoxia and all situations below 6mg/l to 2mg/l will be referred to as oxygen deficiency. For the metric “oxygen depleted area,” the sum of all horizontal model grid cells 2.5m above the bottom with oxygen values below critical levels was calculated. For the “oxygen depleted volume,” the size of all horizontal and vertical model grid cells (starting from about 2.5m above the bottom) with oxygen values below critical levels were summed up. To analyze the “frequency of oxygen deficiency,” the annual average number of days with oxygen concentrations below critical levels in the near-bottom area was calculated and subsequently averaged over the assessment period. For the “duration of oxygen deficiency,” we averaged the number of occurrences of consecutive days (> 2, > 7, > 14, or > 21days) within a year for oxygen concentrations below the specific critical oxygen levels. Analyzed oxygen concentration-period combinations were chosen based on the comprehensive study by Vaquer-Sunyer and Duarte [55] providing information on lethal and sublethal oxygen concentrations and periods for benthic organisms. The area of the near-bottom water layer and water volume used for calculations were computed from the model domain with the values for each sub-basin shown in Table1. Data aggregation and analysis was performed using CDO (version 1.9.8) and R (version 3.6.3 (2020–02-29)) with the tmap package (version 3.3) to produce all maps. Table 1 Near-bottom area and water volume calculated from the ERGOM model domain for the sub-basin divisions according to HELCOM Sub-basin Bottom area (km2) Water volume (km3) Kiel Bay 2747 66 Bay of Mecklenburg 3766 82 Arkona Basin 14,508 461 Pomeranian Bay 4044 68 433Modeling ofWater Quality Indicators intheWestern Baltic Sea: Seasonal Oxygen Deficiency 1 3 3 Results 3.1 Model Performance The model performance was evaluated focusing on the seasonal cycle (Fig.2) and the spatial fit with low-oxygen areas (Fig.3). The mean absolute error of the modeled oxygen concentrations over a year is comparably lower for the sandy station O9 (0.14) than for the central station in the Arkona Basin (1.75), Kiel Bay (1.51), and Bay of Mecklenburg (1.92; Fig.2). For the latter two stations, the model shows considerably lower values especially from October to February (mean absolute error of 3.54 and 3.85, respectively) while the spring and summer months are well represented. In particular, the increase in oxygen concentrations after October is comparatively not fast enough. This underestimation of the bottom oxygen concentrations in Kiel Bay and Bay of Mecklenburg is accompanied by a too strong vertical stratification, which can be seen by a comparison of modeled and observed salinities (SI 1). Since measurements directly above the seafloor are not possible, and rather taken within a range of about 1 to 4m above the seafloor, we also compared the observations with the values in the model layer 2.5m above the seafloor (purple dots in Fig.2). Here, the fit between modeled and observed oxygen concentration improved substantially for the stations in the Kiel Bay and Bay of Mecklenburg (Fig.2a, b). Although for stratified stations some differences between observed and modeled data still exist in the autumn and winter months, the model results 2.5m above the bottom capture the seasonal oxygen minima better. For our analysis of the bottom water layer, we thus utilized this model layer which is in the following referred to as “near-bottom layer.” Next, we compared the minima oxygen values from 2010 to 2019 of the modeled near-bottom oxygen concentrations with the deviation to observed values for stations in the entire western Baltic Sea (Fig.3). The average annual modeled near-bottom oxygen minima are mostly in good agreement with the average of the annual minima from the observations (Fig.3). A reasonable good correlation between modeled and observed values (r = 0.71, R2 = 0.5) indicates that the model system is well able to reproduce the spatial gradients in the western Baltic Sea. Higher modeled oxygen concentrations compared to the measurements occur mainly along the coastline. In some near-shore areas (like Bay of Wismar and the southern coast of Funen), the model tends to overestimate the oxygen concentrations substantially, as the steep bathymetry gradients are not resolved. In contrast, too low oxygen concentrations are mainly predicted by the model in the deep basins and the western part of Pomeranian Bay. Apart from that, in other areas the differences between model values and observations seem to vary strongly although some measuring stations are quite near to each other as, for example, in the Arkona Basin (Fig.3). This is maybe caused by unevenly distributed observations in time among measuring stations, as data from some measuring stations is less often available. For example, if Fig. 2 Comparison of multiannual monthly means of modeled oxygen concentrations from the deepest vertical layer (orange) and 2.5m above bottom (purple) with observed near-bottom data (gray and black) for four stations: a Kiel Bay, b Bay of Mecklenburg, c Station O9 in the southwestern Arkona Basin, and d Station TF113 in the central Arkona Basin in the western Baltic Sea (see Fig.1). Single measurements are indicated by gray points and aggregated to monthly means (black squares). Observations and model results were taken from the period 2010 to 2019 434 S.Piehl et al. 1 3 no observations for the time period with the lowest oxygen concentrations are available, the annual minima are hardly comparable between the monitoring stations. 3.2 Analysis ofSpatial andSeasonal Oxygen Deficiency According to model results, there is a large difference of the extent of areas affected by oxygen deficiency among the assessment units (Fig.4a and Table2). For the period 2011–2016, averaged annual dissolved oxygen concentrations do not show hypoxia in the western Baltic Sea (Fig.4a), but seasonally hypoxia occurs from July to September (Fig.5). Considering oxygen deficiency, the Bay of Mecklenburg exhibited the largest areas with critical oxygen concentrations below 6mg/l with on average 45% of the near-bottom water layer, followed by Kiel Bay with on average 11% (Fig.4a). In the Arkona Basin and the Pomeranian Bay, the average annual oxygen concentrations in the nearbottom water layer were above critical levels (Fig.4a). In most areas of the western Baltic Sea, the annual average variability of oxygen concentration is 0 to 4mg/l with highest variability observed in the south-western part of Bay of Mecklenburg (Fig.4b). Comparing average annual oxygen concentrations among single years (SI 7), the differences were not as pronounced as between the individual months (Fig.5); thus, seasonal variability seems to be more important for an assessment than inter-annual variability. Moreover, the lower model fit in the months of October through March is also reflected by a higher variability of oxygen concentrations in these months (SI 3). Therefore, only the results for the months April to September are shown in Fig.5 and Table2 and are considered for the seasonal analysis below. Generally, the spatial extent and time-span of oxygen deficiency differed among sub-basins (Figs.5 and 6, Table2). To account for naturally occurring hypoxia and oxygen deficiency in the area, we defined critical periods exemplary as periods in which at least 10% of the near-bottom area of a sub-basin exhibits concentrations below a critical oxygen level. Accordingly, critical periods of hypoxia can be observed in the Bay of Mecklenburg (17 to 21% during August and September) and the Arkona Basin (10% in September). According to model results, hypoxia was also observed in a small area during the summer in the Kiel Bay (3 to 7%) and the Pomeranian Bay (1 to 2%; Fig.5, Table2). For oxygen deficiency, critical periods are highly variable among the sub-basins ranging from about 1 to 5months (Fig.5, Table2). Likewise, the spatial extent exhibits a high variability among the sub-basins from about 10 to 74% (Fig.5, Table2). For example, the critical period in the Bay of Mecklenburg lasts 5months and affects an area from 16 to 74%, whereas in the Pomeranian Bay an area of 10% is affected for only 1month (Fig.5, Table2). Kiel Bay exhibits a near-bottom area between 17 and 74% affected by oxygen concentrations below 6mg/l, whereas the period is 1month less as compared to the Bay of Mecklenburg. In the Arkona Basin an average area from 16 to 36% is affected within a period of 3months. The high spatio-temporal variability between assessment units is also seen in the analysis of the water volume affected by hypoxia and oxygen deficiency (Fig.6). Within the period from April to September, the affected water volume by hypoxia reaches less than 2% in Kiel Bay, the Arkona Basin, and in the Pomeranian Bay. Only in the Bay of Mecklenburg the monthly average volume of water reaches up to 7% in September (Fig.6). Looking at the monthly average volume of water where oxygen concentration has dropped below 6mg/l, less than 10% of the water volume in the Fig. 3 The modeled near-bottom oxygen minima (colorcoded; average from 2010 to 2019) and the average difference from the observed annual minima (red and blue circles) 435Modeling ofWater Quality Indicators intheWestern Baltic Sea: Seasonal Oxygen Deficiency 1 3 Arkona Basin and Pomeranian Bay is on average affected over the year. Again, Bay of Mecklenburg is the most affected, with water volumes from 3 to 27% experiencing oxygen deficiency, followed closely by the Kiel Bay where water volumes from 2 to 22% experience oxygen deficiency during the period from April to September (Fig.6). Fig. 4 Modeled a average annual oxygen concentrations (DO [mg/l]) and b standard deviation (STD [mg/l]) in the near-bottom water layer in the period 2011 to 2016 in the western Baltic Sea. Black lines indicate sub-basin divisions according to HELCOM (KB, Kiel Bay; BM, Bay of Mecklenburg; AB, Arkona Basin; PB, Pomeranian Bay). Country borders in gray Table 2 Percentage (%) area affected by critical oxygen concentrations (mg/l) as average over the HOLASII assessment period (2011–2016) Kiel Bay Bay of Mecklenburg Arkona Basin Pomeranian Bay Oxygen threshold < 2 < 4 < 6 < 2 < 4 < 6 < 2 < 4 < 6 < 2 < 4 < 6 April 000003000000 May 0030316000001 June 0 1 17 0 19 43 0 0 3 1 3 7 July 0 7 31 7 43 63 2 7 16 2 5 10 August 3 19 51 17 56 71 7 17 30 1 4 9 September 7 34 74 21 64 74 10 21 36 0 1 2 436 S.Piehl et al. 1 3 3.3 Assessment ofEcological Relevant Oxygen Indicator Metrics To evaluate various oxygen metrics that can be derived from model simulation products and be used as oxygen indicators, we further looked more closely at the temporal component in addition to intensity and extent. Considering the duration and recurrence of oxygen deficiency situations is a critical step toward developing metrics relevant to the ecological health of benthic communities, as duration is most important along with the intensity of oxygen deficiency. Hypoxic conditions in the near-bottom water do frequently (at least half a year ± 1 to 2months) occur within 5% of the Bay of Mecklenburg. About 25% of the area shows hypoxia from a quarter to half a year and 50% of the area experiences hypoxic conditions for less than 2months (Fig.7). Within the Arkona Basin 91% and in the Pomeranian Bay even the whole area is less than twice a month affected by hypoxic conditions (Fig.7). If a threshold of 6mg/l is considered, the near-bottom water layer experiencing oxygen deficiency of at least half a year rises to 43% in the Bay of Mecklenburg (SI 4). In addition, 13% of the near-bottom water layer in Kiel Bay is also affected by critical oxygen levels below 6mg/l at this frequency (SI 4), while in the Arkona Basin 22% of the near-bottom water layer are affected from a quarter to half a year (SI 4). Additional information on the frequency of situations below critical oxygen levels of 4 and 6mg/l is provided within the supplementary information. The largest share of areas and highest incidences where hypoxia occurs for more than 2 as well as 7 consecutive days are in the Bay of Mecklenburg and Arkona Basin (for up to 6 times per year and 2 to 4 times a year, respectively), followed by Kiel and Pomeranian Bay (Fig.8). Hypoxic conditions for more than 14 consecutive days occur on average about 1 to 2 times a year for noticeable areas in the Arkona Basin and Bay of Mecklenburg, the latter showing areas with incidences of up to 4 times (Fig.8). However, Fig. 5 Modeled average monthly oxygen concentrations (DO [mg/l]) in the near-bottom water layer from April to September for the period 2011 to 2016 in the western Baltic Sea (for standard deviation see SI 3). Black lines indicate sub-basin divisions according to HELCOM (KB, Kiel Bay; BM, Bay of Mecklenburg; AB, Arkona Basin; PB, Pomeranian Bay). Country borders in gray 443Modeling ofWater Quality Indicators intheWestern Baltic Sea: Seasonal Oxygen Deficiency 1 3 of hypoxia [67]. Moreover, looking at the time-series of the annual average oxygen concentrations for the Arkona Basin (Fig.9) there seems to be no improvement. Thus, further research is needed to clarify if improvements in the hypoxic near-bottom water layer and water volume are a result of nutrient input reductions as agreed to in 1988 [68]. In addition, the effects of climate change on oxygen concentrations will also need to be studied in the future, which could counteract potential improvements. A proposed warmer Baltic Sea in the future is expected to favor hypoxic conditions due to increased water temperature, microbial activity, and river runoff [12, 69]. Nevertheless, our long-term analysis shows that the relatively stable situation in the 1950s could indicate a situation with non-eutrophication influenced oxygen concentrations and its natural fluctuations. In order to have a comparable approach as for the HELCOM oxygen debt indicator, a somewhat more distant period for the analysis would have been desirable. In the current discussions of the HELCOM working group on the oxygen indicator, later periods are also considered as reference periods, e.g., for the Bothnian Sea and Bothnian Bay (personal communication). In addition, other authors point to a still good ecological status in the 1950s and early 1960s for nutrients and chlorophyll-a [1, 33]. If the current model limitations mentioned above can be overcome in further model runs, then the analysis of past conditions from model runs will help to derive water quality targets for the oxygen indicator. Thereby, our current approach did not consider future climate scenarios, which would be of high importance considering the proposal of water quality targets for environmental assessments. 5 Conclusion Our analyses show that seasonal oxygen deficiency and the related metrics are indeed a suitable indicator for describing the ecological status of the western Baltic Sea, where seasonal and episodic rather than perennial hypoxia and oxygen deficiency occurs. Likewise to nutrients and chlorophyll-a, the situation in the 1950s may be regarded as “good status,” including naturally occurring oxygen deficits in the western Baltic Sea. Oxygen deficiency has a seasonal and a spatial dimension, which both should be considered in defining a suitable indicator. The use of model simulations allows us to add these dimensions. By providing further information on the duration and recurrence interval of oxygen deficiency, the proposed metrics serve as an interface to the most important ecological drivers for benthic biology, so that they can be related to the Baltic Sea Action Plan eutrophication goal of “natural distribution and occurrences of plants and animals.” However, model simulations always have some shortcomings and should not be used alone to derive thresholds. Instead, an integrated approach is needed to combine model results with observations in a suitable way so that seasonal oxygen deficiency can be used as a meaningful eutrophication indicator in future. Supplementary Information The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s1066602209866-x. Acknowledgements We like to thank Marina Carstens (LM-MV), Mario von Weber (LUNG-MV), Hannah Lutterbeck (LLUR-SH), Clarissa Vock, and the HELCOM IN-EUTRO working group for providing data and information. Author Contribution All authors contributed to the study conception and design. Model simulations were carried out by TN. Data collection and analysis were performed by SP and RF. The first draft of the manuscript was written by SP and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Funding Open Access funding enabled and organized by Projekt DEAL. This research was funded by the German Environment Agency (UBA) project “The Baltic Sea Action Plan—Modelling of Water Quality Indicators” (Grant number 3720252020). Supercomputing power was provided by HLRN (North-German Supercomputing Alliance). Data Availability The simulation datasets analyzed during the current study are available in the IOW THREDDS repository, https:// t hred dsio w. iowarne muende. de/ thred ds/ catal ogs/ proje cts/ integ r al/ catal og_ integ ral. html. Other datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. Code Availability Not applicable. Declarations Ethics Approval Not applicable. Consent to Participate Not applicable. Consent for Publication Not applicable. Competing Interests The authors declare no competing interests. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. 444 S.Piehl et al. 1 3 References 1. Carstensen, J., & Conley, D. J. (2019). Baltic Sea hypoxia takes many shapes and sizes.Limnology and Oceanography Bulletin, 28, 125–129. https:// doi. org/ 10. 1002/ lob. 10350 2. Meier, H. E. M., Eilola, K., Almroth-Rosell, E., etal. (2019). Disentangling the impact of nutrient load and climate changes on Baltic Sea hypoxia and eutrophication since 1850. Climate Dynamics, 53, 1145–1166. https:// doi. org/ 10. 1007/ s003820184296-y 3. Reissmann, J. H., Burchard, H., Feistel, R., etal. (2009). Progress in Oceanography Vertical mixing in the Baltic Sea and consequences for eutrophication – A review.Progress in Oceanography, 82, 47–80. https:// doi. org/ 10. 1016/j. pocean. 2007. 10. 004 4. Carstensen, J., Andersen, J. H., Gustafsson, B. G., & Conley, D. J. (2014). Deoxygenation of the Baltic Sea During the Last Century, 111, 5628–5633. https:// doi. org/ 10. 1073/ pnas. 13231 56111 5. HELCOM. (2018a). State of the Baltic Sea – Second HELCOM holistic assessment 2011–2016. Baltic Sea Environment Proceedings, 155. Retrieved May 19, 2022, fromhttps:// helcom. f i/ balticseatrends/ holis ticasses sments/ stateofthebalticsea2018/ repor tsandmater ials/ 6. HELCOM. (2018b). HELCOM thematic assessment of eutrophication 2011–2016. Baltic Sea Environment Proceedings no. 156. Available at: https:// helcom. fi/ balticseatrends/ holis ticasses sments/ stateofthebalticsea2018/ repor tsandmater ials/ 7. Almroth-Rosell, E., Wåhlström, I., Hansson, M., etal. (2021). A regime shift toward a more anoxic environment in a eutrophic sea in Northern Europe. Frontiers in Marine Science. https:// doi. org/ 10. 3389/ fmars. 2021. 799936 8. Kõuts, M., Maljutenko, I., Elken, J., etal. (2021). Recent regime of persistent hypoxia in the Baltic Sea. Environmental Research Communications, 3, 075004. https:// doi. org/ 10. 1088/ 25157620/ AC0CC4 9. Krapf, K., Naumann, M., Dutheil, C., & Meier, H. E. M. (2022). Investigating hypoxic and euxinic area changes based on various datasets from the Baltic Sea. Frontiers in Marine Science. https:// doi. org/ 10. 3389/ fmars. 2022. 823476 10. Vahtera, E., Conley, D. J., Gustafsson, B. G., etal. (2007). Internal ecosystem feedbacks enhance nitrogen-fixing cyanobacteria blooms and complicate management in the Baltic Sea. Ambio, 36, 186–194. https:// doi. org/ 10. 1579/ 00447447(2007) 36[186: IEFENC] 2.0. CO;2 11. Carstensen, J., Conley, D. J., Bonsdorff, E., etal. (2014). Hypoxia in the Baltic Sea: Biogeochemical cycles, benthic fauna, and management. Ambio, 43, 26–36. https:// doi. org/ 10. 1007/ s132800130474-7 12. HELCOM. (2021a). Climate change in the Baltic Sea. 2021 fact sheet. Baltic Sea environment proceedings n°180. HELCOM/Baltic Earth 2021. Retrieved May 19, 2022, from https:// helcom. fi/ balticseaclima techangefactsheetnewpubli cationshowslatestscien tificknowl edgeonclima techangeinthebalticsea/ 13. Liblik, T., & Lips, U. (2019). Stratification has strengthened in the Baltic Sea – An analysis of 35 years of observational data. Frontiers in Earth Science, 7, 1–15. https:// doi. org/ 10. 3389/ feart. 2019. 00174 14. Conley, D. J., Carstensen, J., Aigars, J., etal. (2011). Hypoxia is increasing in the coastal zone of the Baltic Sea. Environmental Science and Technology, 45, 6777–6783. https:// doi. org/ 10. 1021/ es201 212r 15. Reusch, T. B. H., Dierking, J., Andersson, H. C., etal. (2018). The Baltic Sea as a time machine for the future coastal ocean. Science Advances. https:// doi. org/ 10. 1126/ sciadv. aar81 95 16. BLANO. (2018). Zustand der deutschen Ostseegewässer 2018. Aktualisierung der Anfangsbewertung nach § 45c, der Beschreibung des guten Zustands der Meeresgewässer nach § 45d und der Festlegung von Zielen nach § 45e des Wasserhaushaltsgesetzes zur Umsetzung der Meeresstrategie-Rahmenrichtlinie. Anlage 1: Ergänzende nationale Indikatorblätter. Retrieved August 29,2022, fromhttps:// www. meere sschu tz. info/ beric hteart-810. html? file= f iles/ meere sschu tz/ beric hte/% 0bart 8910/ zyklu s18/ Zusta ndsbe richt_ Ostsee_ 2018. pdf. 17. Virtanen, E. A., Norkko, A., Nyström Sandman, A., & Viitasalo, M. (2019). Identifying areas prone to coastal hypoxia – The role of topography. Biogeosciences, 16, 3183–3195. https:// doi. or g/ 10. 5194/ bg1631832019 18. Diaz, R. J., & Rosenberg, R. (1995). Marine benthic hypoxia: A review of its ecological effects and the behavioural responses of benthic macrofauna.Oceanography and Marine Biology an Annual Review, 33(33), 245–303., 33(33), 245–303. 19. Zettler, M. L., Friedland, R., Gogina, M., & Darr, A. (2017). Variation in benthic long-term data of transitional waters: Is interpretation more than speculation? PLoS ONE, 12, e0175746. https:// doi. org/ 10. 1371/ journ al. pone. 01757 46 20. Garlo, E. V., Milstein, C. B., & Jahn, A. E. (1979). Impact of hypoxic conditions in the vicinity of Little Egg Inlet, New Jersey in summer 1976. Estuarine and Coastal Marine Science, 8, 421–432. https:// doi. org/ 10. 1016/ 03023524(79) 90059-8 21. Cardinale, M., & Modin, J. (1999). Changes in size-at-maturity of Baltic cod (Gadus morhua) during a period of large variations in stock size and environmental conditions. Fisheries Research, 41, 285–295. https:// doi. org/ 10. 1016/ S01657836(99) 00021-1 22. Diaz, R. J., & Rosenberg, R. (2008). Spreading dead zones and consequences for marine ecosystems. Science, (80-)321:926–929. https:// doi. org/ 10. 1126/ scien ce. 11564 01 23. HELCOM. (2021b). Baltic Sea Action Plan – 2021 update. Adopted on 20 October 2021 in Lübeck, Germany by the HELCOM Ministerial Meeting. Available at: https:// helcom. f i/ balticseaactionplan 24. HELCOM. (2007). HELCOM Baltic Sea Action Plan. Adopted on 15 November 2007 in Krakow, Poland by the HELCOM Extraordinary Ministerial Meeting. Retrieved May 19, 2022, from https:// www. helcom. fi/ wpconte nt/ uploa ds/ 2019/ 08/ BSAP_ Final. pdf 25. HELCOM. (2018c). HELCOM core indicator report – Oxygen debt. Retrieved May 19, 2022, from https:// helcom. fi/ balticseatrends/ indic ators/ 26. HELCOM. (2020). HELCOM indicator on near-bottom oxygen concentrations in shallow waters - Towards a Baltic Sea wide application IN-EUTROPHICATION 16–2020. Retrieved May 19, 2022, from https:// portal. helcom. fi/ meeti ngs/ INEUTRO PHICA TION% 20162020694/ Meeti ngDoc uments/ Forms/ AllIt ems. aspx 27. HELCOM. (2021c). German national indicator on near-bottom oxygen concentrations. IN-EUTROPHICATION 19A-2021. Retrieved August 29, 2022, from https:// portal. helcom. fi/ meeti ngs/ INEUTRO PHICA TION% 2019A2021887/ Meeti ngDoc uments/ 2-1% 20Ger man% 20nat ional% 20ind icator% 20on% 20nearbo ttom% 20oxy gen% 20con centr ations. pdf 28. HELCOM. (2016). Shallow-water oxygen candidate indicator – Draft description of methodology. State & Conservation 5–2016. Available at: https:// por tal. helcom. fi/ meeti ngs/ STATE% 20-% 20CON SERVA TION% 2052016363/ Meeti ngDoc uments/ 4J31% 20Sha llo wwatero xygen% 20can didate% 20ind icator% 20-% 20dr a ft% 20des cript ion% 20of% 20met hdolo gy. pdf 29. Sveegaard, S. (Ed.). (2020). Danish scientific appraisal of HELCOM’s and OSPAR’s indicators in relation to the monitoring programme under the Marine Strategy Framework Directive – Status 2020. Aarhus University, DCE – Danish Centre for Environment and Energy, 114 pp. Technical report no. 168. Retrieved August 29, 2022, from http:// dce2. au. dk/ pub/ TR168. pdf 30. Bever, A. J., Friedrichs, M. A. M., Friedrichs, C. T., etal. (2013). Combining observations and numerical model results to improve estimates of hypoxic volume within the Chesapeake Bay, USA.TheJournalofGeophysical Research, 118, 4924–4944. https:// doi. org/ 10. 1002/ jgrc. 20331 445Modeling ofWater Quality Indicators intheWestern Baltic Sea: Seasonal Oxygen Deficiency 1 3 31. Conley, D. J., Bjorck S, Bonsdorff E, etal. (2009). Hypoxiarelated processes in the Baltic Sea. Environmental Science and Technology, 43, 3412–3420. https:// doi. org/ 10. 1021/ es802 762a 32. Neumann, T., Koponen, S., Attila, J., etal. (2021). Optical model for the Baltic Sea with an explicit CDOM state variable: A case study with model ERGOM (version 1.2).Geoscientific Model Development, 14, 5049–5062. https:// doi. org/ 10. 5194/ gmd1450492021 33. Neumann, T., & Schernewski, G. (2008). Eutrophication in the Baltic Sea and shifts in nitrogen fixation analyzed with a 3D ecosystem model. Journal of Marine Systems, 74, 592–602. https:// doi. org/ 10. 1016/j. jmars ys. 2008. 05. 003 34. Friedland, R., Neumann, T., & Schernewski, G. (2012). Climate change and the Baltic Sea action plan: Model simulations on the future of the western Baltic Sea. Journal of Marine Systems, 105– 108, 175–186. https:// doi. org/ 10. 1016/j. jmars ys. 2012. 08. 002 35. Meier, H. E. M., Andersson, H. C., Arheimer, B., etal. (2012). Comparing reconstructed past variations and future projections of the Baltic Sea ecosystem—First results from multi-model ensemble simulations. Environmental Research Letters, 7, 034005. https:// doi. org/ 10. 1088/ 17489326/7/ 3/ 034005 36. Meier, H. E. M., Andersson, H. C., Eilola, K., etal. (2011). Hypoxia in future climates: A model ensemble study for the Baltic Sea.Geophysical Research Letters,38:n/a-n/a. https:// doi. org/ 10. 1029/ 2011G L0499 29 37. Neumann, T., Radtke, H., & Seifert, T. (2017). On the importance of major Baltic inflows for oxygenation of the central Baltic Sea. Journal of Geophysical Research Ocean, 122, 1090–1101. https:// doi. org/ 10. 1002/ 2016J C0125 25 38. Kuznetsov, I., & Neumann, T. (2013). Simulation of carbon dynamics in the Baltic Sea with a 3D model. Journal of Marine Systems, 111–112, 167–174. https:// doi. org/ 10. 1016/j. jmars ys. 2012. 10. 011 39. Griffies, S. M., Harrison, M. J., Pacanowski, R. C., & Rosati, A. (2004). A technical guide to MOM4. GFDL Ocean Group technical report no. 5. NOAA/Geophysical Fluid Dynamics Laboratory. Version prepared on August 23, 2004. Available online at: www. gfdl. noaa. gov 40. Griffies, S. M. (2012). Elements of the Modular Ocean Model (MOM) - 2012 release with updates. GFDL Ocean Group technical report no. 7. NOAA/Geophysical Fluid Dynamics Laboratory. 41. Winton, M. (2000). A reformulated three-layer sea ice model.The Journal of Atmospheric and Oceanic Technology, 17, 525–531. https:// doi. org/ 10. 1175/ 15200426(2000) 017% 3c0525: ARTLSI% 3e2.0. CO;2 42. Weisse, R., von Storch, H., Callies, U., etal. (2009). Regional meteorological–marine reanalyses and climate change projections. Bulletin of the American Meteorological Society, 90, 849–860. https:// doi. org/ 10. 1175/ 2008B AMS27 13.1 43. Geyer, B., & Rockel, B. (2013). coastDat-2 COSMO-CLM atmospheric reconstruction. World Data Center for Climate (WDCC) at DKRZ. Available at: https:// doi. org/ 10. 1594/ WDCC/ coast Dat-2_ COSMOCLM 44. HELCOM. (2015). Updated fifth Baltic Sea pollution load compilation (PLC-5.5). Baltic Sea environment proceedings no. 14. Retrieved May 19, 2022, from https:// helcom. fi/ balticseatrends/ pollu tionloadcompi latio ns/ 45. HELCOM (2021d) Input of nutrients by the seven biggest rivers in the Baltic Sea region 1995-2017. Baltic Sea Environment Proceedings No. 178. Available at:https:// helcom. fi/ wpconte nt/ uploa ds/ 2021/ 09/ Thesevenbigge striversintheBalticSeaRegion. pdf 46. Gustafsson, B. G., Schenk, F., Blenckner, T., etal. (2012). Reconstructing the development of Baltic Sea eutrophication 1850–2006. Ambio, 41, 534–548. https:// doi. org/ 10. 1007/ s132800120318-x 47. Gauss, M., Bartnicki, J., & Klein, H. (2019). Atmospheric nitrogen deposition to the Baltic Sea, Oslo. Retrieved May 19,2022,fromhttps:// emep. int/ publ/ helcom/ 2018/ index. html. 48. Gauss, M., Bartnicki, J., Jalkanen, J., etal. (2021). Airborne nitrogen deposition to the Baltic Sea: Past trends, source allocation and future projections. Atmospheric Environment, 253, 118377. https:// doi. org/ 10. 1016/j. atmos env. 2021. 118377 49. Ruoho-Airola, T., Eilola, K., Savchuk, O. P., etal. (2012). Atmospheric nutrient input to the Baltic Sea from 1850 to 2006: A reconstruction from modeling results and historical data. Ambio, 41, 549–557. https:// doi. org/ 10. 1007/ s132800120319-9 50. Eilola, K., Gustafsson, B. G., Kuznetsov, I., etal. (2011). Evaluation of biogeochemical cycles in an ensemble of three state-of-theart numerical models of the Baltic Sea. Journal of Marine Systems, 88, 267–284. https:// doi. org/ 10. 1016/j. jmars ys. 2011. 05. 004 51. Meier, H. E. M., Edman, M. K., Eilola, K. J., etal. (2018). Assessment of eutrophication abatement scenarios for the Baltic Sea by multi-model ensemble simulations. Frontiers in Marine Science, 5, 1–24. https:// doi. org/ 10. 3389/ fmars. 2018. 00440 52. Neumann, T., Siegel, H., & Gerth, M. (2015). A new radiation model for Baltic Sea ecosystem modelling. Journal of Marine Systems, 152, 83–91. https:// doi. org/ 10. 1016/j. jmars ys. 2015. 08. 001 53. Neumann, T., Siegel, H., Moros, M., etal. (2020). Ventilation of the northern Baltic Sea. Ocean Science, 16, 767–780. https:// doi. org/ 10. 5194/ os167672020 54. Leipe, T., Naumann, M., Tauber, F., etal. (2017). Regional distribution patterns of chemical parameters in surface sediments of the south-western Baltic Sea and their possible causes. Geo-Marine Letters, 37, 593–606. https:// doi. org/ 10. 1007/ s003670170514-6 55. Vaquer-Sunyer, R., & Duarte, C. M. (2008). Thresholds of hypoxia for marine biodiversity. Proceedings of the National Academy of Sciences, 105, 15452–15457. https:// doi. org/ 10. 1073/ pnas. 08038 33105 56. Radtke, H., Lipka, M., Bunke, D., etal. (2019). Ecological ReGional Ocean Model with vertically resolved sediments (ERGOM SED 1.0): Coupling benthic and pelagic biogeochemistry of the south-western Baltic Sea. Geosciences Model Development, 12, 275–320. https:// doi. org/ 10. 5194/ gmd122752019 57. Lipka, M., Woelfel, J., Gogina, M., etal. (2018). Solute reservoirs reflect variability of early diagenetic processes in temperate brackish surface sediments. Frontiers in Marine Science, 9, 1–20. https:// doi. org/ 10. 3389/ fmars. 2018. 00413 58. Gogina, M., Lipka, M., Woelfel, J., etal. (2018). In search of a field-based relationship between benthic macrofauna and biogeochemistry in a modern brackish coastal sea. Frontiers in Marine Science, 5, 1–18. https:// doi. org/ 10. 3389/ fmars. 2018. 00489 59. Schubert, P. R., Hukriede, W., Karez, R., etal. (2015). Mapping and modeling eelgrass Zostera marina distribution in the western Baltic Sea. Marine Ecology Progress Series, 522, 79–95. https:// doi. org/ 10. 3354/ meps1 1133 60. Löffler, A., Leipe, T., & Emeis, K. C. (2000). The “fluffy layer” in the Pomeranian Bight (western Baltic Sea): Geochemistry, mineralogy and environmental aspects. Meyniana, 52, 85–86. 61. LLUR. (2020). LLUR-Bericht zu den Sauerstoffuntersuchungen in der Westlichen Ostsee im Sept. 2020. Retrieved November 14, 2022, from https:// www. schle swigholst ein. de/ DE/ Fachi nhalte/ M/ meere sschu tz/ Downl oads/ Beric ht_ LLUR_ Sauer stoff_ 2020. pdf 62. Janas, U., Wocial, J., & Szaniawska, A. (2004). Seasonal and annual changes in the macrozoobenthic populations of the Gulf of Gdańsk with respect to hypoxia and hydrogen sulphide. Oceanologia, 46, 85–102. 446 S.Piehl et al. 1 3 63. Sinkko, H., Hepolehto, I., Lyra, C., etal. (2019). Increasing oxygen deficiency changes rare and moderately abundant bacterial communities in coastal soft sediments. Science and Reports, 9, 1–14. https:// doi. org/ 10. 1038/ s4159801951432-1 64. Villnäs, A., Norkko, J., Lukkari, K., etal. (2012). Consequences of increasing hypoxic disturbance on benthic communities and ecosystem functioning. PLoS ONE. https:// doi. org/ 10. 1371/ jour n al. pone. 00449 20 65. Conley, D. J., Carstensen, J., Ærtebjerg, G., etal. (2007). Longterm changes and impacts of hypoxia in Danish coastal waters. Ecological Applications, 17, 165–184. https:// doi. org/ 10. 1890/ 050766.1 66. Villnäs, A., Norkko, J., Hietanen, S., etal. (2013). The role of recurrent disturbances for ecosystem multifunctionality. Ecology, 94, 2275–2287. https:// doi. org/ 10. 1890/ 121716.1 67. Meier, H. E. M., Väli, G., Naumann, M., etal. (2018). Recently accelerated oxygen consumption rates amplify deoxygenation in the Baltic Sea.TheJournalofGeophysical Research Ocean, 123, 3227–3240. https:// doi. org/ 10. 1029/ 2017J C0136 86 68. HELCOM. (1988). Declaration on the protection of the marine environment of the Baltic Sea area. Adopted on 15 February 1988 in Helsinki, Finland by the Ministers responsible for the environmental protection in the Baltic Sea States. Retrieved May 19, 2022,from https:// helcom. fi/ helcomatwork/ minis terialmeeti ngs/ 1988helsi nki/ 69. Meier, H. E. M., Hordoir, R., Andersson, H. C., etal. (2012). Modeling the combined impact of changing climate and changing nutrient loads on the Baltic Sea environment in an ensemble of transient simulations for 1961–2099. Climate Dynamics, 39, 2421–2441. https:// doi. org/ 10. 1007/ s003820121339-7 Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.