Status of pelagic habitats within the EU-Marine Strategy Framework Directive: Proposals for improving consistency and representativeness of the assessment

Abstract

3 2017/848 have set criteria and methodological standards for the assessment and determination of Good Environmental 4 Status (GES) for pelagic habitats in EU waters, there is strong evidence that Member States have not yet harmonized the 5 pelagic GES assessment across EU marine waters. Today, pelagic habitats are assessed by evaluating whether good status 6 is achieved by each of the pelagic indicators, but this approach fails to observe the high variability of the pelagic 7 environment. To this end, GES is not estimated at pelagic habitats scale but only for each individual indicator. This paper 8 synthesises the latest developments on pelagic habitats assessment and identifies the main factors limiting the 9 consistency of the assessment across Member States: i) coarse spatial and temporal scales of sampling effort as regards 10 to the pelagic habitat dynamics, ii) little consideration of the whole range of plankton (and, to some extent, of 11 zooplankton) size and trophic spectra, iii) lack of integrated hydro-biogeochemical and biological studies and 12 collaboration among experts from different scientific fields, iv) limited availability of pressure-based indicators, and v) 13 lack of integration methods of the pelagic indicators' status for the GES determination. This analysis demonstrates the 14 importance of maintaining a consistent sampling frequency and a spatially extensive network of stations across the 15 gradient of anthropogenic pressures, where spatial environmental data can help objectively extrapolating field data.
consider changes in the spatial heterogeneity of regional sea characteristics [10], and therefore depends on the spatial 23 and temporal characteristics of the monitored and assessed marine area. As a recommendation, the GES assessment 24 of pelagic habitats in the MSFD needs common determination of GES, evaluation criteria, and consistent methods 25 across the European Seas (e.g., data type and frequency, indicators and analysis) to achieve comparable results for 26 the policy requirements [9,11].

27
Previous studies have highlighted the challenges of determining GES for pelagic habitats (e.g., [12,13]), and how 28 pelagic abiotic and biotic characteristics change over time in response to different factors (e.g., climate change, [14]).

29
For example, long-term (centennial to decadal) changes in hydro-meteorological conditions (e.g., water mixing, 30 precipitation, temperature) have been identified as dominant drivers of pelagic processes such as primary production 31 [15]. Planktonic organisms respond to changes in the characteristics of water masses and hydrographic processes 32 according to their lifespan, growth rate and size [16].

33
Research has shown that plankton dynamics (e.g., changes in biomass, diversity) in the North Sea and Northeast 34 Atlantic have strongly changed compared to decades ago due to large-scale climatic drivers (e.g., sea surface 35 temperature, water transparency, salinity [14,17]. Other types of changes also occur over seasonal and short 36 (episodic) timescales as a result of nutrient dynamics in coastal systems with high levels of land-based sources [18,19].

42
These changes, which occur over different timescales, are difficult to capture through specific driving factors and processes in the context of the reporting obligation of the MSFD. Yet, often because of lack of data spanning multiple 44 timescales, the pressure-state relationships between human activities and pelagic dynamics are difficult to establish, 45 as well as the importance of the short-term (days to months) relative to long term changes (years to decades). affected in square kilometres or as a percentage of the total extent of the habitat type [23], assuming that available 48 data are fully representative of the pelagic habitat and that the assessment will not be biased by the selected 49 indicators nor the sampling strategy. The MSFD requires MS to update their marine strategy every six years (MSFD 50 Article 17(2)), which could lead MS to establish the six-years cycle as the timescale for the pelagic assessment (GES 51 Decision). As the observed changes in pelagic habitats are strongly short time-dependent, limiting the MSFD pelagic 52 assessment to a timescale that is not relevant to capture this variability can generate biased results, with long-term 53 measurements at low frequency underestimating the total change that occurs over shorter timescales [24].

55
A series of indicators that address interacting pelagic processes have been adopted by MS to monitor the pelagic 56 habitat [24-26]. These indicators include both general (i.e., abundance, biomass) and plankton-specific metrics (i.e., 57 taxonomy diversity) that are based on different sampling strategies and methodologies, to address changes in the 58 status of pelagic habitats [10]. These indicators typically target groups of pelagic communities that are associated 59 with specific spatio-temporal scales, and anthropogenic pressures that are mostly not captured with traditionally 60 applied sampling strategies [27]. Also, to date, the data collection for pelagic indicators is limited by the extent of the 61 pelagic habitats. Therefore, there is a need for inter-regional consistency and explicit consideration of the relevant 62 timescales for each indicator to capture and evaluate the spatio-temporal extent of human impacts on pelagic 63 habitats [24]. To achieve this goal in a conceptually harmonised and coordinated way at the EU level, the MSFD 64 Common Implementation Strategy [28] has developed a step-by-step approach that ranges from the selection of 65 habitat characteristics to indicator identification and the setting of relevant thresholds. Given the differences in 66 pelagic habitat affecting the physical and biological characteristics in space and time across regional seas, the 67 development of indicators can be region-and subregion-specific (GES Decision). Accordingly, the GES Decision 68 highlights the need for MS to cooperate at EU, regional or subregional level for selecting indicators that ensure the 69 assessment is based on reliable data and functionally comparable methods. So far, MS assess pelagic habitat status 70 by evaluating whether good status is achieved at a pelagic indicator level without integrating the indicator results for 71 the GES assessment at habitat scale [11]. As a consequence, the assessment is fragmented, and the GES status is not 72 consistent across MS or inconclusive. To this end, the selection of common abiotic and biotic characteristics of pelagic 73 habitats across regional seas would facilitate the adoption of common indicators and their integration in the GES 74 determination and assessment [27].

75
This paper presents the progresses and challenges of the MSFD pelagic habitat assessment and discusses the potential 76 contribution to the harmonised evaluation of pelagic habitat GES, while taking into consideration the differences and 77 similarities between and within the marine regions. This analysis particularly focuses on three issues that emerged from 78 the latest MS reports [11]. First, the four broad habitat types (i.e., variable salinity, coastal, shelf, and oceanic beyond 79 shelf) defined by the GES Decision require revision to depict the spatio-temporal variability of pelagic habitats and to 80 consider the large extent of the assessment units, the broad oceanographic characteristics, and the low data sampling 81 (i.e., scaling effect of data collection) (Sections 2 and 3). Second, there is a general lack of agreed indicators and GES-82 related thresholds at sub-regional and regional scale to ensure comparable and harmonised assessments (Sections 3 83 and 4). Third, the paper recommends a more ecologically relevant and adaptive process for the assessment and 84 monitoring of the impacts and pressures of pelagic habitats (Section 4). 87   According to the GES Decision preamble, MS need to "build upon standards stemming from Union legislation or, where   88 do not exist, upon standards set by Regional Sea Conventions (RSCs) or other international agreements". For example, 89 when threshold values are not yet established for GES, MS can refer to existing ones of the RSCs (e.g., [29,30]). To this 90 end, the following section summarizes the pelagic indicators that are currently used by the MS for the pelagic habitat 91 GES, defined at national level or in the framework of RSCs. Also, it illustrates the data collection frequency for the main 92 indicators' parameters across Marine Regions.

93
A thorough understanding of the effect of pressures and their interactions in the marine realm is key to building robust 94 pressure indicators and ensuring consistency of MSFD assessment between marine regions. However, the current 95 indicators that have an EU-wide scale of applicability have regionally-specific thresholds, when thresholds exist (Table   96 1). None of the indicators in Table 1 quantify alone D1C6 as "extent of habitat adversely affected in square kilometres 97 (km 2 ) and as a proportion (percentage) of the total extent of the habitat type [9]", nor to direct anthropogenic pressures 98 (    3 REGIONAL inexistent REGIONAL BLK, MED [48] 11 In the Northeast Atlantic region (OSPAR area), a suite of complimentary plankton indicators, providing insight into 12 different aspects of the plankton community, are used for MSFD assessment and reporting [32,33]. The indicators are 13 informed by data covering a spectrum of taxonomic information, from bulk information such as chlorophyll 14 concentration to species abundance data (Figures 1, 2). This flexible approach makes best use of the wide variety of 15 plankton data available (e.g., Table 2), regardless of their sampling and analysis methodologies, with varying 16 taxonomic specificity, in the OSPAR region.

57
Since then, changes in some lifeforms in indicator PH1/FW5 have been linked to climate change [40]. No thresholds 58 exist for any of these pelagic indicators, so these indicators may reflect longer-term changes for the wider ecosystem 59 than changes induced by the human pressures that are monitored by the MSFD.

58
The definition of good status for each indicator is so far depending on regional characteristics and data availability 59 (Section 2). However, this definition must be consistent with the overall need to track the condition of the pelagic 60 habitats relative to the main pressures adopting an approach that is functionally similar and consistent at the EU level 61 (GES Decision). The GES determination is given by the assessment of each broad habitat type and MS must provide environment. National assessments refer to or reuse regional assessments as they are, and complement them with 35 additional elements, whilst seeking harmonization with neighbouring countries.

37
The indicators examples illustrate that long-term observations are essential to define reference conditions of pelagic 38 habitats (e.g., Black Sea). Long-term data on species-specific metrics can indicate tipping points and/or trends within 39 the sampled area. However, and besides the varying spatial and temporal sampling issue, the interpretation of their  Units. This GES definition for pelagic habitats is a challenge since, as these water bodies are fluid in movement that 50 are characterised by much higher spatial and temporal variabilities than the Reporting Units and the in-situ sampling 51 strategy can address. The approaches to defining Marine Reporting Units also vary between regional seas, MS and 52 descriptors. This section details why the use of Marine Reporting Units and broad habitat types, as defined in the GES 53 Decision, shows challenges for a pragmatic and effective spatio-temporal assessment of pelagic habitats.

The time scales 55
One of the main challenges for assessing pelagic habitat GES is to include processes that may act at weekly to seasonal 56 time scales (e.g., eutrophication event after a river flood) and at multi-decadal time scales (e.g., phytoplankton 57 community composition due to climate change or deep layer anoxia in permanent halocline areas). Mixing these two 58 time scales in the GES assessment is difficult because the longer time scale processes influence the shorter ones and    Figures 1, 3, 4), which is rarely 90 adapted to the local variability. This is key to detect the relevant natural and anthropogenic changes and their impacts on pelagic habitat ( Figure 6). The understanding of the area monitored and its pressures affecting GES would 92 substantially be improved using a grid-based approach dividing the assessment units of broad habitats into smaller 93 units of regular sizes. A regular distribution of sampling sites with weekly effort within each of the broad habitat type 94 would be ideal but, in practice, sampling sites are restricted to specific areas with often a much lower frequency (e.g., 95 river plume areas, Figure 6). It is thus unlikely that the monitoring results are indicative of the whole assessment unit.

96
In order to unlock most of this major limitation, data from satellite observations (e.g., surface chlorophyll-a -this 97 paper, harmful algal blooms -https://www.s3eurohab.eu/node/1) and operational models (e.g., nutrient 98 distributions from Copernicus Marine Services related to the risk for harmful algal blooms) at daily time scale can be 99 used to extrapolate the in-situ observations/indicators within the gridded approach to better depict the spatio-00 temporal variability of the pelagic habitat (e.g., algal bloom events in the Bay of Vilaine, France, Figure 6) (45 days   29 for the timeseries, 15 days for the maps) generated by the Vilaine river from about the end of the previous event on June 26, to the 30 peak on July 5 and 7, and the end on July 12. Note that a substantial part of the difference between in-situ and satellite-derived 31 chlorophyll-a levels may arise from the sampled size, about 1 litre versus 3.4 km by 4.6 km, respectively. The satellite information 32 may efficiently be used to reasonably extrapolate the extent and duration of the in-situ-derived GES estimate of pelagic habitat. 35 Monitoring data remain expensive and limited in space and time so that the Marine Strategy should optimize sampling 36 to best explore the pressure-response relationships and spatial representativeness of GES assessment. The current 37 sampling network raises two main problems: sampling gaps due to regional habitat variability and specificities, and 38 lack of structural organization for the monitoring of pelagic habitats.

39
The first issue is the coarse interpolation of the GES assessment for broad habitat types (Figures 1-5)

52
The second issue relates to the lack of agreed indicators at sub-regional and regional scale and characterized by 53 sampling bias on biological communities. Over the last years, monitoring of pelagic communities has shifted from 54 classic sampling technologies to approaches combining optical-image-molecular data that allow improving the 55 taxonomical resolution and to consider the whole size-trophic spectra of biological communities [16]. However, these 56 methods have often been applied to research projects at regional scale and not yet to improve the spatial and 57 temporal resolution of data sampling for the MSFD GES assessment. Collaboration with these scientific fields (e.g., 58 molecular biology, satellite remote sensing, optical/imaging automated techniques, biogeochemical modelling) is 59 recommended to increase the volume of relevant data for the GES assessment.