Food webs Thematic Assessment

Changes in state of marine food webs

Food webs have shown widespread changes in the OSPAR Maritime Area in recent decades. Fishing and climate change represent the main human pressures affecting trophic networks and OSPAR marine ecosystems. There has been a general pattern of decreasing phytoplankton and zooplankton abundance and / or biomass across the Greater North Sea, Celtic Seas, Bay of Biscay and Iberian Coast, which might have consequences for upper trophic levels. Demersal fish communities have not achieved good status in any of the assessed Regions. Due to widespread human pressures across multiple compartments of the ecosystem, the food web ecosystems in the four OSPAR Regions assessed in the present Thematic Assessment are indicatively considered not to be in good status according to the definition and categorisation of quality status described in McQuatters-Gollop et al., (2022).

Table S.1: Confidence assessment results across the five OSPAR Regions, determined using the methodology outlined in the QSR guidance

OSPAR acts as a coordination platform in the North-East Atlantic for the regional implementation of the EU Marine Strategy Framework Directive (MSFD) that aims to achieve a Good Environmental Status (GES) in European marine environments, as well as for the coordination of other national frameworks. The characteristics of GES are determined by the individual EU member states, based on criteria elements, threshold values and methodological standards set regionally or at EU level.  
Norwegian, Icelandic, United Kingdom, Greenlandic and Faroese marine areas are not covered by the MSFD.

Food webs are key to our understanding of changes in ecosystem structure and functioning. Changes in marine food web structure and functioning can affect the provision of ecosystem services, as outlined in the Ecosystem Services section.

The food web common and candidate indicators briefly described below address the EU Marine Strategy Framework Directive Descriptor 4 criteria, as outlined in Table S.3. In Arctic Waters (OSPAR Region I) a different set of information was used and therefore not included in the assessment overviews for indicators.

Food webs are key to our understanding of changes in ecosystem structure and functioning. Changes in marine food web structure and functioning can affect the provision of ecosystem services, as outlined in the Ecosystem Services section. 

The food web common and candidate indicators briefly described below address the EU Marine Strategy Framework Directive Descriptor 4 criteria, as outlined in Table S.3 (Norwegian, Icelandic and United Kingdom marine areas are not covered by the MSFD). In Arctic Waters (OSPAR Region I) a different set of information was used and therefore not included in the assessment overviews for indicators. 

The assessment of the state of food webs in OSPAR Regions was undertaken using a variety of indicators (common and candidate), methods and spatial scales. A set of indicators focusing on different trophic levels (e.g. primary production, zooplankton, fish, whole food web) was independently assessed. No integration method was used for the assessment results. Results are described for each ecosystem: variable salinity, coastal, shelf and oceanic / deep-sea (Figure S.1, Table S.3, Table S.4, Table S.5 and Table S.6). Assessment units represent OSPAR Regions or their sub-divisions (specified areas that reflect biogeographic / hydrological characteristics of the sub-region). This assessment provides a descriptive summary of the results of each indicator rather than an integrated assessment of food webs. The main results for each indicator in each OSPAR Region are summarised in Figure S.1. The status assessment is based on expert judgement, since most indicators have no agreed threshold value. 

Indicators representing differing trophic levels of the food web have been developed to create an overall perspective of change. The indicators for Primary production (FW2) as well as Phytoplankton and zooplankton abundance (PH1 / FW5 and FW6) capture change in bottom-up processes often linked to climate change and eutrophication. The Biomass of feeding guilds indicator (FW7) links the base of the food web to higher predators (e.g. through typically smaller fish that feed on plankton to typically larger fish that feed on fish) and to the seabed (through fish that feed on benthos) and can capture the cumulative impacts of human activities on essential fish habitats and those from the direct removal (harvesting), in addition to climate impacts from the base of the food web that may favour planktivory over benthivory or vice versa, depending on the status of the ecosystem. Potentially, fishing impacts on planktivores can also cascade down the food web and impact the planktonic community and cascade up the food web to marine birds (Lynam et al., 2017). 

Fishing also preferentially targets larger individuals and, over time, a sustained increase in the mortality of a population will lead to a reduction in the proportion of large individuals in the ecosystem. Furthermore, as species that can grow large are depleted, particularly those which are slow-growing such as demersal sharks, for example tope (Galeorhinus galeus), the biomass of smaller bodied species such as dab (Limanda limanda) can increase to fill the ecological niche. The indicators of these highly diverse demersal fish communities (FC2, FC3, FW3) can capture such wider impacts of fishing on the resultant size structure (FC2 and FW3) and composition of species (FC3). In the Greater North Sea, Celtic Seas, Bay of Biscay and Iberian Coast, the proportion of large fish (FC2) in a survey has been shown to track change in fishing mortality over long periods. This indicator (FC2) requires a specific threshold in order to determine a large fish with the tendency to reflect change in the biomass of larger species targeted by fisheries (e.g. cod, Gadus morhua, and saithe, Pollachius virens, in the North Sea). The indicator for typical length of fish (FW3) complements that for proportion of large fish (FC2), since typical length is a consistent metric over both time and space for a specific survey and can be used to demonstrate where changes in the fish community are leading to an overall failure in the proportion of large fish (FC2). This can help to determine which process is leading the change (e.g. climate, or fisheries or habitat change). 

Changes in average trophic level of marine predators (FW4) is an ecological indicator that characterises change in the trophic pyramid and can integrate the cumulative effects of fishing pressure and other drivers on the trophic structure of food webs. Conceptually, this indicator has aimed at highlighting unsustainable fishery practices in past decades. The lack of trends found in many scenarios with this assessment does not necessarily indicate lack of impact. Rather, this needs to be further investigated in order to determine whether such absence can be attributed to resilience of the ecosystem or other possible causes. Ecological Network Analysis (ENA) is a system-ecology oriented approach to analyse all trophic interactions (flows in energy or carbon) among all compartments of a food web. ENA and its related ENA indices (FW9) enable assessment of the structure and functioning of food webs based on the analysis of all interactions among the living and the non-living compartments (detritus carbon pools). The food web is represented by a network of quantified interactions (flows) between nodes (trophic compartments, gathering organisms having similar preys and predators and metabolic rates). This indicator can successfully capture the effects of cumulated pressures and has shown its capacity to detect changes in food webs and ecosystems due to climate change, eutrophication, fisheries, hypoxia events or invasive species.

These food web indicators used to assess the state of marine food webs (Figure S.1) are briefly described below, followed by a more extensive overview of the status assessments for each Region.

Figure S.1: Indicator results schematic for food web ecosystems (variable salinity, coastal, shelf, and oceanic / beyond shelf) within the OSPAR Regions. For the OSPAR food web biodiversity indicators there was variability among indicators, ecosystem habitats, and OSPAR Regions. Icons have been coloured according to indicator trends. Not all indicators were assessed in some Regions due to lack of data or because the habitat type was not present within the OSPAR Region. Grey background colour indicates that a candidate indicator was used and a pilot assessment conducted. The common indicator ‘proportion of large fish’ (FC2) has agreed threshold values in the North Sea and Celtic Sea and the icon is coloured accordingly. The candidate indicator ‘ecosystem network modelling’(FW9) is excluded from Figure S.1 since this pilot assessment in Regions II and V focused on specific areas and not on whole Regions. An explanation of the meaning of any observed trend or change for each indicator assessment in each habitat type and Region is given in table headings 3-6

Changes in Phytoplankton and Zooplankton Communities (PH1/FW5):
This Common Indicator includes phytoplankton and zooplankton lifeform abundance data. The main human pressures on these lifeforms are climate change and eutrophication.

Pilot Assessment of Primary Productivity (FW2):
This Candidate Indicator includes phytoplankton productivity data; the main human activities / pressures on phytoplankton productivity are climate change and eutrophication, even though hydrodynamical changes can be important by changing mixing/stratification as well as light conditions, especially in coastal and variable salinity habitats. 

Size composition and spatial distribution of zooplankton (FW6): This Candidate Indicator targets zooplankton biomass and body size. In the current OSPAR cycle, this indicator could not be assessed because of insufficient data. However, efforts have been made to automate the collection of size and biomass data from fisheries surveys and test the approach (e.g. Pitois et al., 2021).

Proportion of Large Fish (Large Fish Index) (FC2) and Size Composition in Fish Communities (FW3):
These Common Indicators include many species of demersal fish (including elasmobranchs) and summarise the change in the size structure of the fish community at the scale of regional seas and within spatial sub-divisions (FW3 only). The main human activities / pressures affecting demersal fish communities are trawl fisheries, since fishing mortality reduces biomass and the proportion of larger and older fish, but climate change may act to hinder the growth of typically larger cold-water species (e.g. Queirós et al., 2018). FC2 has agreed threshold values for the North Sea and Celtic Sea 

Mean Maximum Length of Fish (FC3):
This Candidate Indicator includes demersal fish (including elasmobranchs) and summarises change in the relative biomass of species, where species are weighted by their ability to grow large (using the stationary trait of maximum body length), since larger species are vulnerable to depletion through additional mortality. The main human activities / pressures affecting the assessed fish are fisheries (Spence et al., 2021) and climate change. 

Changes in Average Trophic Level of Marine Consumers (FW4):
This Common Indicator includes fish, cephalopods and benthic invertebrates dwelling in sediment bottoms of the circalittoral zone. The main human activities / pressures that were identified are fisheries and climate change. Eutrophication and invasive species can also be of relevance in specific areas. 

Feeding Guilds (FW7):
Due to data availability, this Candidate Indicator currently includes pelagic and demersal fish and elasmobranchs only, but in future it will include data on higher and lower trophic levels. This indicator tracks the change in biomass and diversity of taxonomic groups that share similar prey items (e.g. planktivores, benthivores and piscivores) and have similar functional roles in the food web (Thompson et al., 2020). The main human activities / pressures are fisheries and climate change, which can interact to alter the strength of energy pathways through the food web (Thorpe et al., 2022).

Ecological Network Analysis Indices (FW9):
This Candidate Indicator is a multitrophic level indicator. It includes pelagic elements (phytoplankton, zooplankton), benthic elements, fish, seabirds and mammals. Ecological Network Analysis (ENA) is a system-ecology oriented approach for analysing all trophic interactions (e.g. flows in energy or carbon) among all compartments of a food web (Ulanowicz, 2004; Safi et al., 2019). ENA and its related indices (FW9) enable assessment of the structure and functioning of food webs based on the analysis of all interactions among the living and the non-living compartments (detrital carbon pools).  The food web is represented by a network of quantified interactions (flows) between nodes (trophic compartments, gathering organisms having similar preys and predators and metabolic rates).

Table S.2: Food web indicators in relation to EU Marine Strategy Framework Directive Descriptor 4 criteria. 

OSPAR acts as a coordination platform in the North-East Atlantic for the regional implementation of the EU Marine Strategy Framework Directive (MSFD) that aims to achieve a Good Environmental Status (GES) in European marine environments, as well as for the coordination of other national frameworks. The characteristics of GES are determined by the individual EU member states, based on criteria elements, threshold values and methodological standards set regionally or at EU level.  
Norwegian, Icelandic, United Kingdom, Greenlandic and Faroese marine areas are not covered by the MSFD.

The current food web ecosystem status of the five OSPAR Regions is summarized in the section below from the key information for each indicator considered. Results and information are presented per ecosystem per OSPAR Region with the intention of providing a qualitative description of the parameters that have been assessed. It has not yet been possible to provide an integrated assessment of the state of the ecosystems. 

Overall, in the 38 assessments conducted over four Regions all the indicators showed either a failed threshold, trends associated with human activities or a unknown result. Moreover, none of the indicators assessed for food web currently has achieved a threshold value which would be an expression of the desired status of the particular parameter. No integration was done and none of the ecosystems can be assessed for good or not good environmental status. Overall, these results clearly show the need for further data collection and research to support the mitigation of food web effects arising from the impacts of human activities. The results indicate risks for both bottom-up and top-down effects. Trophic cascades affect the entire food web and thereby the functioning of the entire ecosystem.

The status of the Arctic Waters Region is illustrated through region-relevant peer-reviewed literature as well as expert knowledge since it was not possible to assess the region using OSPAR Common Indicators at this time.

Arctic Waters (Region I)

The assessments for the Arctic Waters ecosystems are based on third party sources (Arctic Council / (AMAP) / ICES regional sea reports). The assessments constitute a different set of information from the OSPAR Common Indicators. 

A novel assessment method is being developed in Norway for evaluating the ecological condition of ecosystems (Jepsen et al., 2020). It will be applied to all of Norway’s three seas: the Barents Sea, the Norwegian Sea and the Norwegian sector of the North Sea. The overall question that the assessment aims to answer is whether there has been a change from the defined reference condition (“intact nature”, i.e. a situation where the ecosystem is largely unimpacted by modern industrial activities) which can be attributed to anthropogenic impacts. This is done by selecting indicators for seven ecosystem characteristics: primary productivity, biomass distribution among trophic levels, functional groups within trophic levels, functionally important species and biophysical structures, landscape-ecological patterns, biological diversity and abiotic factors. Food webs are thus represented mainly by the indicators of primary productivity (e.g. annual net primary production, timing of spring bloom) and biomass distribution across trophic levels (e.g. in the Barents Sea Arctic and Sub-Arctic ecosystems, the biomass of low trophic level zooplankton, benthos suspensivores, pelagic planktivorous fish and high trophic level seabirds, among others). These indicators are calculated from satellite data (for primary productivity) and from stock assessment and surveys available in the Region. For benthic suspensivore biomass, for example, the indicator is represented by the sum of biomass (kg/km²) of megabenthic species (sampled with bottom trawl), weighted by a fuzzy coding for their degree of suspension feeding. Fuzzy coding is a method of indicating to what extent a taxon exhibits selected trait categories, in this case suspension feeding; some functional trait-related indicators such as the ratio of gelatinous zooplankton to krill, or fish size in the community, are also included in other characteristics of the assessment. This last indicator is represented by biomass-weighted community mean body length at maturity. Species-specific lengths at maturity were taken from the literature (Wiedmann et al., 2014). Climate change is considered currently to be a stronger driver of change than fishing for the food web in the Barents Sea, with substantially increasing air and water temperatures causing loss of sea ice and habitats for ice-associated species in the Arctic region. There is strong evidence of important changes in the upper and lower trophic levels, suggesting important changes in trophic structure, in the Norwegian sector of the Arctic Barents Sea. However, in the sub-Arctic part, there is low or intermediate evidence of changes in parts of the upper (high trophic level seabirds) and lower (suspension feeders) trophic levels. There are major uncertainties linked to the short time series available for most of the biological groups.

Following the assessment of the seven ecosystem characteristics (primary productivity, biomass distribution among trophic levels, functional groups within trophic levels, functionally important species and biophysical structures, landscape-ecological patterns, biological diversity and abiotic factors) a panel of fisheries, oceanographic and marine biology experts concluded that primary production, biomass distribution within trophic levels, landscape-ecological patterns and abiotic factors showed substantial deviations from the reference conditions. By contrast, the panel found more limited deviation for functional groups within trophic levels and biological diversity, and no deviation for functionally important species and biophysical structures. In the Atlantic region of the Barents Sea, substantial deviations were assessed for landscape-ecological patterns and abiotic factors, while the remainder were found to have no deviation from expected characteristics (Siwertson et al., in prep, 2022).

Greater North Sea (Region II)

In the Greater North Sea Region, one Common Indicator failed its threshold value, and the other Common Indicators and Candidate indicators show a trend that is associated with human activities (Table S.3). The two indicators for the first two layers of the food web (FW2 and PH1 / FW5) revealed trends within coastal and shelf ecosystems and unknown trend within the variable salinity ecosystem. All indicators for demersal fish communities either fail the threshold value or show a declining trend associated with human activities, indicating that this component of the food web is under pressure.

Table S.3: Trends for Common and Candidate Indicators within each ecosystem for Greater North Sea based on each indicator’s results. NA = Not Assessed; FC2 = threshold failed in shelf ecosystems; FW3 = downward trend (deteriorating state) in shelf ecosystems; FW5 = downward trends in holoplankton and upward trends in meroplankton in shelf ecosystems; FW2 = downward trends in variable salinity, coastal and shelf ecosystems; FC3 = downward trend (deteriorating state) in shelf ecosystems; FW7 = downward trend (deteriorating state) for planktivorous fish and no trend (unknown implications) for sub-apex demersal predators in shelf ecosystems; and FW9= no trend (unknown implications) in coastal and shelf ecosystems.
*This pilot assessment was done in specific areas and did not cover the whole region

Variable salinity ecosystems

Plankton community assessment

For variable salinity ecosystems, the Common Indicator Changes in Phytoplankton and Zooplankton Communities (PH1 / FW5) indicated that dinoflagellates and larval fish have been increasing in abundance according to long-term trends. Holoplankton, large copepods and small copepods indicated significant downward trends; however, plankton monitoring datasets were only available for a limited number of assessment units, with significant gaps in time series. Therefore, lower confidence must be assigned to these results, particularly for the zooplankton lifeform time-series. For this Common Indicator, no clear links could be made between changes in lifeform abundance and anthropogenic pressures, owing to the generally low internal confidence among the assessed time-series. Contrary to the PH1 / FW5 Common Indicator, it was possible to complete the Candidate Indicator Pilot Assessment of Primary Productivity  (FW2) for a large proportion of assessment units (89% of assessment units covered). The results revealed a significant downward trend in primary production, as also evidenced by the long-term evolution of primary production in the North Sea (e.g. Capuzzo et al., 2018). As with the decreasing phytoplankton biomass for the common indicator Changes in Phytoplankton Biomass and Zooplankton Abundance (PH2), the most important predictor of decreasing primary production was decreasing concentration of dissolved inorganic phosphate.

Coastal ecosystems

Plankton community assessment

For coastal ecosystems, the Changes in Phytoplankton and Zooplankton Communities (PH1/FW5) Common Indicator revealed significant upward trends in meroplankton and larval fish abundance, which may have been associated with rise in sea surface temperatures and increasing salinity, respectively. While the data used to assess trends exhibited a high degree of confidence and the majority of assessment units were well represented, except for the gelatinous zooplankton lifeform, a high degree of variability resulted from linking trends in lifeform abundance with environmental pressures, and thus the associations with change in sea surface temperature and salinity lacked confidence. The Candidate Indicator  Pilot Assessment of Primary Productivity (FW2) revealed a significant downward trend in primary production which was attributed to increasing sea surface temperature. Similar results were evidenced by the long-term evolution of primary production in the North Sea (e.g. Capuzzo et al., 2018). 

Multitrophic level assessment

Changes in trophic structure and functioning derived by Ecological Network Analysis (ENA) indices, ( Candidate Indicator Pilot Assessment of Ecological Network Analysis Indices , FW9) are very much linked to changes in primary productivity. In the coastal North Sea, primary production decreased from 2009 to 2014 but has increased since 2015. This has been accompanied by an increase in the biomass of benthic filter feeding invertebrates and benthic feeding birds (in 2016), which probably benefit from the increasing biomass of filter feeding invertebrates. However, an overall decrease in benthic filter feeding invertebrates was found over time. The Detritivory: Herbivory ratio has increased since 2015. Flow diversity is a measurement of the number of interactions and evenness of energy flows. As with the biodiversity index, a high value shows a highly diverse and well-developed system. Since all models have the same number of compartments, energy flows between trophic guilds seem to be more unevenly distributed. As in the case of the coastal North Sea food web, ENA indices in the Seine Bay model are linked to changes in primary production and to the biomass of phytoplankton. By contrast with the Elbe Plume, the Finn Cycling Index decreased from 2000 to 2014 while the Detritivory:Herbivory ratio increased from 2000 to 2006 / 2007 but then decreased until 2015. These changes mainly related to the pelagic parts of the food web, including a strong increase in the biomass of zooplankton, bacteria and zooplanktivorous fish.

Shelf ecosystems

Plankton community assessment

For shelf ecosystems, the Changes in Phytoplankton and Zooplankton Communities (PH1/FW5) Common Indicator revealed that diatoms, meroplankton, and larval fish underwent significant upward long-term trends, while dinoflagellates, holoplankton, and small copepods experienced significant downward trends. For meroplankton and holoplankton, the abundance trends were linked to rising sea surface temperatures with a high degree of consistency, while the increase in larval fish abundance was associated with decreasing light attenuation. While no data were available to assess trends for gelatinous zooplankton, there was high confidence and high representation of assessment units for all time-series assessed. The results revealed a significant downward trend in primary production, the most important predictor of this being an increase in the ratio of nitrogen to phosphorous (N/P ratio). In turn the latter is a consequence of efforts to combat eutrophication, with greater reductions achieved for phosphorous loading than for nitrogen. The results presented some deviation, as long-term decrease in primary production in the North Sea has not been linked to increasing N/P ratio (e.g. Capuzzo et al., 2018).

Demersal fish community assessment

The fish community indicators of size structure, namely Proportion of Large Fish (Large Fish Index) (FC2) and Size Composition in Fish Communities (FW3), and of species composition, namely the Candidate Indicator Pilot Assessment of Mean Maximum Length of Fish (FC3), demonstrated long-term decline of this ecosystem component in the Greater North Sea. The assessments show a common pattern of decreases between the 1980s and 2000 that was consistent between scientific surveys and indicators and that has been linked to high levels of fishing mortality in demersal fish stocks (Greenstreet et al., 2011). 

Within sub-divisions of the Greater North Sea, the Candidate Indicator on fish Species composition (FC3) showed overall patterns of decline that were driven by a loss of the species that grow large in the southern North Sea sub-divisions, with no change overall in the north. The typical length of fish (FW3) Common Indicator demonstrated an overall decline in the observed sizes of individual fish in all areas that was followed by a recovery in the northern North Sea (north-eastern and Orkney-Shetland sub-divisions) during the 2000s and a variable pattern with no clear trend in the 2010s. The depletion of species that can grow large in the south (FC3) suggests that the recovery in community size structure (FW3) back towards the previous high proportion of large fish in the early 1980s (FC2) may have been hindered. However, provided that all commercial species are fished under fishing mortality targets consistent with Maximum Sustainable Yield, increases towards a new higher equilibrium are expected within two decades (Spence et al., 2021).

Evidence of continued recoveries in the eastern part of the Channel sub-division were evident in the proportion of large fish (FC2) and species composition (FC3), with mixed patterns in typical length (FW3).

Feeding guild assessment

The spatial gradient for planktivore biomass and species richness showed an increasingly northward trend in the shelf ecosystem of the Greater North Sea, with high points in the north-western North Sea and Kattegat, in the Candidate Indicator Pilot Assessment of Feeding Guilds (FW7). Benthivore biomass was uniformly distributed across the Region, with higher species richness in the north-western North Sea and Kattegat. The spatial gradients for pisco-crustivore (i.e., fish that feed omnivorously, largely preying on crustaceans and fish) and piscivore biomass and species richness increasingly moved north and west across the Region. These patterns are probably heavily influenced by natural, large-scale gradients in environmental conditions. The temporal dynamics of planktivore biomass showed an increase in the Channel and a minor decrease in the north-western North Sea but no significant change elsewhere. Planktivore species richness was increasing across the northern North Sea, Kattegat and the Channel with no significant change elsewhere. Benthivore biomass showed increases in the southern North Sea and decreases in the Kattegat, with little change elsewhere. Benthivore species richness was increasing in the English Channel, decreasing in the north-eastern North Sea and Kattegat with limited change elsewhere. Pisco-crustivore biomass was increasing in the southern North Sea but decreasing across much of the north, with increases in species richness in the north-west, the English Channel and the south-east but decreases elsewhere. Piscivore biomass showed contrasting directions of change in the western North Sea, with decreases off the east coast of Scotland and northern England but increases further south, with limited change elsewhere. Piscivore species richness was increasing across the north and the Kattegat but decreasing in the west, with limited change in the southern North Sea.

Oceanic / beyond shelf

Although the Greater North Sea (Region II) intersects with a small portion of an Oceanic / beyond shelf assessment units, this ecosystem was not assessed for the Greater North Sea since the ecosystem shares far greater overlap with the Bay of Biscay and Iberian Coast (OSPAR Region IV).

Celtic Seas (OSPAR Region III)

In the Celtic Seas Region, one Common Indicator failed the threshold value and the other indicators show trends that are associated with human activities (Table S.4). Within coastal ecosystems, only the lower trophic levels of the pelagic food web were assessed and showed trends that were considered linked to negative impacts from human activities.

Table S.4: The trends for Common and Candidate indicators within each ecosystem for Region III are given based on each indicator results. NA = not assessed; FC2 = threshold failed in shelf ecosystems; FW3= downward trend (deteriorating state) in shelf ecosystems; FW5 = downward trends for four plankton lifeforms in coastal ecosystems, with downward trends in dinoflagellates and holoplankton and upward trends in meroplankton in shelf ecosystems; FW2= downward trends in variable salinity and shelf ecosystems and no trend in coastal ecosystems; FC3 = no trend (unknown implications) in shelf ecosystems; and FW7= no trend (unknown implications) for planktivorous fish and increasing trend (improving state) for sub-apex demersal predators in shelf ecosystems

Variable salinity ecosystem

Plankton community assessment

For variable salinity ecosystems in the Celtic Seas, the Common Indicator Assessment on Changes in Phytoplankton and Zooplankton Communities (PH1/FW5) revealed that dinoflagellates have been undergoing significant upward long-term trends, while all other planktonic lifeforms have undergone significant downward trends. However, there was very low representation of assessment units for all zooplankton lifeforms. For diatoms, downward abundance trends were linked to an increase in the ratio of nitrogen to phosphorous. For large copepods, which were only assessed at a single fixed-point monitoring station, the downward trend was linked to a rise in sea surface temperature. For the Candidate Indicator Pilot Assessment of Primary Productivity (FW2), the pilot assessment results revealed a significant decrease in primary production linked to a modelled decreasing trend in pH. Relevant literature on primary production trends in the area is scarce. While the present assessment identifying a decrease in primary production is congruent with Tilstone et al., (2022), Hernvann et al., (2020) have noted long-term stability in primary production in the Celtic Sea. In addition, the existing literature has not yet investigated the link between ocean acidification and primary production over long-term timescales. Changes in pH arise from concentration of dissolved inorganic carbon, which in turn depends on phytoplankton activity (resulting in the balance between production and respiration) and CO₂ level. Since phytoplankton activity within coastal ecosystems (including variable salinity and coastal habitats) is subject to large diel variability but is CO₂ neutral over long timescales (Duarte et al., 2013), it is unclear at this stage if the decrease in primary production in the variable salinity and coastal habitats assessed here is due to decreasing pH or if it is the cause of decreasing pH. Further investigations should untangle the links between phytoplankton activity (production and respiration) and long-term concentration of dissolved inorganic carbon and anthropogenic CO₂.

Coastal ecosystems

Plankton community assessment

For coastal ecosystems, the Common Indicator Assessment on Changes in Phytoplankton and Zooplankton Communities (PH1/FW5)  showed that dinoflagellates, holoplankton, large copepods and small copepods all underwent significant downward trends linked to rising sea surface temperatures, and for holoplankton, to decreasing modelled pH. No data were available for the gelatinous zooplankton lifeform. There was high confidence for all assessed time series and high spatial representation of assessment units. For the Candidate Indicator Pilot Assessment of Primary Productivity (FW2), the pilot assessment results revealed a significant downward trend in primary production probably linked to decreasing modelled pH. For the same reason as stated with variable salinity habitats, it is currently unclear whether the decrease in primary production detected in this assessment was the consequence of ocean acidification.

Shelf ecosystems

Plankton community assessment

For shelf ecosystems of the Celtic Seas, the  Common Indicator Assessment on Changes in Phytoplankton and Zooplankton Communities (PH1/FW5) revealed that meroplankton underwent significant upward trends, while dinoflagellates and holoplankton indicated significant downward trends. For dinoflagellates, the changes were linked to an increase in the ratio of nitrogen to phosphorous. For meroplankton, the changes were linked to increasing sea surface temperature, and for holoplankton the changes were linked to ocean acidification. No data were available for gelatinous zooplankton; however, there was high confidence for all time-series assessed and high spatial representation of assessment units. The Candidate Indicator Pilot Assessment of Primary Productivity (FW2) pilot assessment results revealed a significant downward trend in primary production linked to a decreasing trend in modelled pH in shelf ecosystems. While the present assessment identified a decrease in primary production congruent with Tilstone et al., (2022), Hernvann et al., (2020) noted long-term stability in primary production in the Celtic Sea. The decrease in primary production in this assessment is probably related to shallower mixed layer depth, which is predicted in the future for the Irish Sea (Olbert et al., 2012). 

Demersal fish community assessment

The fish community indicators of size-structure, Proportion of Large Fish (Large Fish Index) (FC2) and Size Composition in Fish Communities (FW3), demonstrate long term decline in status in the Region. Since the results are all at a lower level than during previous assessment they indicate a status that cannot be interpreted as good. No clear pattern was evident in the species composition Candidate Indicator Pilot Assessment of Mean Maximum Length of Fish (FC3).

Following the depletion observed in the 1980s, some surveys of the Proportion of Large Fish (Large Fish Index) (FC2) in the Region show evidence of ongoing recoveries during the 2000s and 2010s. At smaller spatial scales, mean maximum length (FC3) and size composition (FW3) found evidence of mixed patterns, with increases in some sub-divisions (including the Irish Sea, Bristol Channel and much of the area south of Ireland) and decreases elsewhere (including the deeper waters at the shelf edge).

Feeding guild assessment

Planktivore biomass was highest off the north-west coast of Scotland and around the north and west coast of Ireland, while their species richness was highest off the north-west coast of Scotland and in the north of the Celtic Sea in the Candidate Indicator Pilot Assessment of Feeding Guilds (FW7). Benthivore biomass was highest in the Irish Sea and off the west coast of Ireland, with their species richness highest in the Irish Sea and in all coastal areas across the Region. Pisco-crustivore and piscivore biomass and species richness tended to increase northwards and westwards across the region. 

Regarding temporal changes, planktivore biomass was increasing to the west of Scotland and north of the island of Ireland, with minor patchy increases in the north-eastern Celtic Sea and similarly patchy decreases in the north-western Celtic Sea and limited change elsewhere. Planktivore species richness was increasing in the southern Irish Sea and north-eastern Celtic Sea, but decreasing off the north-west coast of Ireland and remaining relatively stable elsewhere. Benthivore biomass showed updard trends, from the north-eastern Celtic Sea through the Irish Sea and to the West of Scotland, with one area south of Ireland showing a slight decrease and insignificant change elsewhere. Benthivore species richness showed contrasting directions of change, with increases in the south and decreases in the north, with insignificant change in the areas between them. Pisco-crustivore biomass was increasing over much of the study region, with one area between Scotland and Northern Ireland showing a slight decrease. Pisco-crustivore species richness was increasing in the south west of the region and to the north but decreasing around Northern Ireland and in the Bristol Channel. Piscivore biomass was also increasing over much of the study region, with a relatively small area of decline along the west coast of Ireland. Piscivore species richness was increasing in the south of the region, with limited change elsewhere.

Oceanic / beyond shelf

While the Celtic Seas (Region III) intersects a small portion of an Oceanic / beyond shelf assessment units, this ecosystem was not assessed for the Celtic Seas, since it shares far greater overlap with the Bay of Biscay and Iberian Coast (OSPAR Region IV).

Bay of Biscay and Iberian Coast (OSPAR Region IV)

In the Bay of Biscay and Iberian Coast Region, one Common Indicator fails the threshold value and three other showing trends that are associated with human activities. Plankton-related indicators (PH1 / FW5; FW2) were assessed in shelf and ocean / deep ecosystems showing trends associated with human activities in both ecosystem types. The status of fish-related indicators was classed as unknown for all habitat types assessed. An overall resilient bentho-demersal community due to persistent pressure over time could explain the stable trends observed in this specific ecosystem compartment.

Table S.5: The trends for each Common and Candidate indicator within each ecosystem for Region IV are given based on each indicator result. NA= not assessed; FC2 = no trend (unknown implications) in shelf ecosystems; FW4 = no trend (unknown implication) in the three habitat types assessed; FW3 = no trend (unknown implications) in shelf ecosystems; FW5 = downward trends for holoplankton, small copepods and large copepods in shelf ecosystems, and downward trends in small and large copepods in oceanic/beyond shelf ecosystems; FW2 = downward trend in shelf and oceanic/beyond shelf ecosystems; FC3 = no trend (unknown implications) in shelf ecosystems; and FW7 = downward trend (deteriorating state) for planktivorous fish and upward trend (improving state) for sub-apex demersal predators in shelf ecosystems
 

Variable salinity habitat

Plankton community assessment

For variable salinity habitats, there were no suitable plankton monitoring data available to evaluate the pelagic habitat indicators. For the Candidate Indicator Pilot Assessment of Primary Productivity (FW2), the assessment was also not possible because of lack of data.

Coastal ecosystems

Plankton community assessments

For coastal habitats, only phytoplankton data were available to inform Common Indicator Assessment on Changes in Phytoplankton and Zooplankton Communities (PH1/FW5). Further, trends in diatoms and dinoflagellates in this case had to be represented by a dataset which only recorded specimens from the genera Pseudo-nitzschia and Dinophysis. This indicator showed that dinoflagellates have undergone significant upward long-term trends. While the assessed time-series had moderate confidence, there was poor spatial representation of assessment units and no clear links could be drawn between changes in lifeform abundance and environmental pressures. For the Candidate Indicator Pilot Assessment of Primary Productivity (FW2), assessment was not possible due to lack of data.

Changes in average trophic level of marine consumers assessment

The Common Indicator Assessment on Changes in Average Trophic Level of Marine Consumers (FW4) was assessed in coastal ecosystems in only two sub-divisions of the Region: French shelf and Gulf of Cadiz. Regarding the upper trophic levels, the MTL indicator showed a significantly increasing trend in the southernmost areas of the French sub-division, whereas no significant trend was observed in Gulf of Cadiz. Taking into account other scenarios (including mesopredators), downward trends were detected in Gulf of Cadiz. Regarding the spatio-temporal approach at local scale, no significant changes were observed in coastal areas under any of the scenarios in the French shelf. In Gulf of Cadiz however, downward trends were detected in the temporal approach, concentrated in local coastal areas. The contrasting trends observed in coastal areas indicate an uncertain status for MTL in this ecosystem.

Shelf ecosystems 

Plankton community assessment

For shelf ecosystems in the Bay of Biscay and Iberian Coast, the Common Indicator Assessment on Changes in Phytoplankton and Zooplankton Communities (PH1/FW5) revealed that diatoms, holoplankton, meroplankton, large copepods, and small copepods all showed significant downward trends. No data were available to assess the gelatinous zooplankton lifeform from the Continuous Plankton Recorder; however, there was generally high confidence across the assessed time-series and moderate spatial representation of assessment units. The decreasing trends in holoplankton, large copepods, and small copepods were all linked to rising sea surface temperatures. Concerning the Candidate Indicator Pilot Assessment of Primary Productivity  (FW2), the pilot assessment results revealed a significant downward trend in primary production. Spatial assessment was limited to one area (representing 17% of all sub-divisions in the habitat), resulting in low confidence in those results. Primary production in the Bay of Biscay and Iberian Coast is subject to large uncertainty regarding their future projection (Holt et al., 2014). The decrease identified in primary production reported in the current assessment was associated with a decrease in wind speed, which is contradictory to the increasing wind speed in the Bay of Biscay shown by Chust et al., (2022). 

Demersal fish community assessment

The fish community indicators of size-structure, namely Proportion of Large Fish (Large Fish Index) (FC2), Size Composition in Fish Communities (FW3), and of species composition, namely the Candidate Indicator Pilot Assessment of Mean Maximum Length of Fish (FC3), demonstrated no clear pattern of change in the Bay of Biscay and Iberian Coast. This result has unclear implications. 

Changes in average trophic level of marine consumers assessment

Considering top predators alone, the Common Indicator Assessment on Changes in Average Trophic Level of Marine Consumers (FW4) showed significant upward trends in the French and Gulf of Cadiz shelves, in contrast to the downward trends observed in the North Iberian and Portuguese shelves. However, when including mesopredators in the calculations, the trends became positive or stable in most of scenarios. In the spatio-temporal approach, a decline of the indicator was observed in the westernmost area of the North Iberian sub-division, more pronounced in shelf ecosystems than in deeper waters. In the Portuguese sub-division, little evidence of MTL changes at the local scale was observed in any of the scenarios, with a very small number of significant trends scattered across the area. These results at the local scale mostly agree with the temporal trends obtained when analysing the sub-division as a whole, which highlights the greater stability/resilience of the Portuguese sub-division in relation to the adjacent areas. Gulf of Cadiz displayed significant positive trends for top predators and mesopredators in this ecosystem, scattered across the southern part of the sub-division. Overall, the contrasting trends observed in the different scenarios by sub-division indicate an uncertain status for this indicator in shelf ecosystems.

Feeding guild assessment

The Candidate Indicator Pilot Assessment of Feeding Guilds (FW7) results showed that planktivore biomass and species richness was highest in the Bay of Biscay, decreasing southwards. Benthivore biomass was also highest in the Bay of Biscay and decreased southwards, whereas benthivore species richness was highest in the north of the Bay of Biscay and along the coast of northern Spain. Pisco-crustivore biomass and species richness was highest in the north of the Bay of Biscay and along the coast of northern Spain. Piscivore biomass was highest in the Bay of Biscay and along the coast of Portugal, with species richness highest in the north of the Region. 

Temporal changes showed that planktivore biomass had decreased across the Bay of Biscay, increasing at the north-western tip of Spain but relatively constant elsewhere. Change in planktivore species richness and benthivore biomass and species richness was limited to patchy increases along the west coasts of Portugal and Spain. There were increases in pisco-crustivore and piscivore biomass and species richness in the Bay of Biscay and in patches along the west coasts of Portugal and Spain, with patchy decreases in piscivore species richness along the north coast of Spain.

Oceanic/beyond shelf

Plankton community assessment

For oceanic / beyond shelf ecosystems, the  Common Indicator Assessment on Changes in Phytoplankton and Zooplankton Communities (PH1/FW5) demonstrated that meroplankton underwent significant upward trends. All other lifeforms underwent significant downward trends except for gelatinous zooplankton, which were not assessed due to a lack of data. Downward trends in holoplankton, large copepods, and small copepods were all linked to rising sea surface temperature. Despite the low spatial representation of assessment units for informing PH1/FW5, the assessment units considered for this pelagic habitat were spatially extensive and incorporated a very large quantity of samples. Further, there was high confidence in the time-series used to assess PH1/FW5. Concerning the Candidate Indicator Pilot Assessment of Primary Productivity (FW2), the pilot assessment results revealed a significant downward trend in primary production. Spatial assessment was limited to one area representing 20% of all sub-divisions in the habitat, resulting in low confidence in those results. Primary production in the Bay of Biscay and Iberian Coast is subject to large uncertainty as to its future projection (Holt et al., 2014). The decrease in primary production identified in the current assessment was associated with a decrease of light in the water column. If phytoplankton productivity (FW2) is linked to the relevant pelagic and eutrophication assessments, the increase in light attenuation could be due to increase in chlorophyll-a concentration, which was evidenced by the Common Indicator Assessment Concentrations of Chlorophyll-a in the Greater North Sea, Celtic Seas and Bay of Biscay and Iberian Coast but not by the Common Indicator Assessment of Changes in Phytoplankton Biomass and Zooplankton Abundance (PH2) assessment. 

Changes in average trophic level of marine predators assessment

Contrasting trends were observed in the evolution of the indicator when looking at top predators, showing a decline in deeper areas of the North Iberian sub-division in contrast to the positive trends observed in Gulf Cadiz for the  Common Indicator Assessment on Changes in Average Trophic Level of Marine Consumers (FW4). When mesopredators are included, positive trends emerge in all sub-divisions except for Portuguese slope, which displayed stable trends. Regarding the spatio-temporal approach, trends appeared heterogeneously and patchily distributed in all sub-divisions. Again, the contrasting trends of the indicator observed in deeper areas seemed to indicate an unclear status for the average trophic level of the bentho-demersal communities in this habitat type.

Wider Atlantic (OSPAR Region V)

In the Wider Atlantic Region, the results of the assessment showed that the two Common Indicators that were assessed had an unknown status due to no trend being detected and the reference condition not being known. Although, this might seem more positive compared with the other Regions, the result is due to the lack of data, and this should be taken into account in the interpretation of the outcome.

Table S.6: The trends for Common and Candidate indicators within each ecosystem for Region V based on each indicator result. NA = Not Assessed; FC2 = no trend (unknown implications) in shelf ecosystems; FW3 = no trend (unknown implications) in shelf ecosystems; FC3 = no trend (unknown implications) in shelf ecosystems; FW7 = increasing trends (improving state) for planktivorous fish and sub-apex predators in shelf ecosystems; and FW9 = no trend (unknown implications) in oceanic/beyond shelf ecosystems.
* This pilot assessment was done on a specific area and did not cover the whole Region

Shelf ecosystems

Demersal fish community assessment

No long-term changes were evident in the Wider Atlantic for the size-structure and community composition indicators Proportion of Large Fish (Large Fish Index) (FC2), Size Composition in Fish Communities (FW3) and the Candidate Indicator Pilot Assessment of Mean Maximum Length of Fish (FC3). The implications are unknown. 

Feeding guild assessment

The Candidate Indicator Pilot Assessment of Feeding Guilds (FW7) showed that planktivore and piscivore biomass was high in the Wider Atlantic in the Porcupine Bank survey, and lower to the west of Scotland at the Rockall Bank. Benthivore biomass showed no clear gradient of change between the two surveyed areas in the Wider Atlantic, whereas pisco-crustivore biomass was highest in the Rockall Bank survey. The species richness of planktivores, benthivore and piscivores was higher to the west of Ireland and lower to the west of Scotland, with no clear gradient in pisco-crustivore species richness between the surveyed areas. 

Temporal changes of planktivores showed patchy increases in biomass and species richness across the surveyed areas. Benthivore biomass was increasing in the Rockall Bank survey and decreasing to the west in the Porcupine Bank survey, with limited change elsewhere. Benthivore species richness was increasing in the Porcupine Bank survey, with no clear change elsewhere. Pisco-crustivore biomass was increasing in the Rockall Bank survey, but there was limited change in biomass elsewhere and no change in species richness across either the Rockall Bank or the Porcupine Bank surveys. Piscivore biomass and species richness showed increases across both Rockall Bank and Porcupine Bank surveys.

Oceanic / beyond shelf

Multitrophic Level assessment

The Candidate Indicator Pilot Assessment of Ecological Network Analysis Indices (FW9) was assessed for a deep-sea ecosystem in the Wider Atlantic. A globally decreasing trend for biomass at all trophic levels was evident in the Azores food web model. An overview of the relative contribution by the low and mid-trophic level groups that shape the biomass spectra to the diets of top predators (TL>4) suggests two distinct energy pathways reaching the top of the chain. Model-derived ENA indicators were estimated between 1997 and 2018. The inter-annual variability of the indicators was strongly influenced by the availability of primary production in the system. No significant changes of ENA indices were found over time. The only exception was the resilience indicator, suggesting that the resilience of the ecosystem is decreasing.

Knowledge gaps and way forward towards an integrated assessment of food webs

The functioning of marine food webs, and particularly the assessment of their status, is a complex task owing to the intrinsic complexity of marine networks and trophic interactions. To better assess the state of food webs an integrated approach should be developed and implemented, taking into account all trophic guild compartments of the ecosystem. Rigorous, consistent and coordinated monitoring programmes are being implemented by Contracting Parties in some cases (e.g., groundfish surveys). However, many trophic guilds are still under-sampled (e.g., zooplankton biomass, which is used in the candidate indicator FW6) and are therefore underrepresented in indicators and models worldwide. Most of the food web indicators developed so far focus on specific functional groups (e.g., phytoplankton, fish). However, to assess a network where every node is interconnected, information on as many compartments as possible is needed. From an operational point of view, this is a difficult task which has emerged as one of the main challenges in the assessment of food web status. The complexity of assessing the structure and functioning of food webs is a common issue in all OSPAR Regions. 

Further development of food web indicators should be directed towards more integrative and process-based indicators, operationalising the ENA approach. Most of the existing monitoring data can be used to construct the network to be analysed through the model, and most food web, benthic and pelagic habitat indicators can benefit from monitoring of the parameters needed for ENA indices. Future efforts should focus on making this mutual benefit a reality. To do this, the spatial scale and resolution of the assessments (regional vs local) should be agreed among the Contracting Parties. Food web indicators should also be understandable for policy makers and the general public. How to convey the results of the set of indicators developed under ENA indices (FW9), turning complexity into simplicity, is an issue that needs to be addressed in the future.

In the next five years future work will explore the changes in food web indicators due to climate change impacts within IPCC scenarios. In order to capture the impacts of human activities and climate change on food webs, in terms of changes in trophic and non-trophic interactions and the supply of ecosystem services, there is a need to develop integration methods and cumulative effects assessments further (e.g., Piet et al., 2021). As an example, further development of ecosystem modelling approaches will allow the exploration of different scenarios addressing the impacts of fishing and climate change on food webs (Thorpe et al., 2022; Spence et al., 2021).