Particules plastiques dans l’estomac du fulmar en mer du Nord

Introduction

Many marine organisms, including seabirds, turtles, marine mammals, fish, crustaceans, shellfish and zooplankton ingest man-made debris that they encounter in their environment (Kühn et al., 2015). The quantity of litter ingested and found in animal stomachs or intestines, in particular that of persistent materials such as plastics reflects the abundance of marine litter, the associated harm to wildlife and the marine ecosystem, and socio-economic harm.

Policy

Within its system of Common Indicators, OSPAR agreed to the monitoring of plastic abundance in stomachs of seabirds as an indicator for levels and trends in marine litter floating at the surface of the North Sea. The indicator has been implemented through long-term monitoring of plastic abundance in stomach contents of the northern fulmar Fulmarus glacialis which has been developed and applied by OSPAR as an Ecological Quality Objective (EcoQO) for the North Sea (OSPAR, 2009, 2010a,b). The EcoQO approach has been used as the basis of the OSPAR Common Indicator Assessment for the Intermediate Assessment 2017.

The EU Marine Strategy Framework Directive 2008/56/EC (MSFD) refers to indicators related to litter in the water column (including floating at the surface) and to litter ingested by marine animals.

Monitoring

The OSPAR Common Indicator on plastic particles in fulmar stomachs has been presented as a monitoring approach in the MSFD (MSFD Descriptor 10 Indicator 2; EC 2008, 2010; Galgani et al., 2010), which is suitable for implementation in the Greater North Sea, Arctic Waters and Celtic Seas. The methodology for undertaking assessments of ingested plastic has also been used for other biota indicators for marine litter in other MSFD regions where no fulmars occur.

The purpose of monitoring plastic abundance in stomachs of fulmars is:

• to obtain a quantitative measure for spatial and temporal patterns in the abundance and composition of marine litter, in particular plastics, mainly floating at the surface
• to provide an indication of ecological harm caused by such litter.

Monitoring methods and results have been published in regular reports and peer-reviewed scientific literature (Van Franeker et al., 2011; Van Franeker and Law 2015). In 2015 OSPAR adopted Guidelines for Monitoring and Assessment of Plastic Particles in Stomachs of Fulmars in the North Sea Area (OSPAR Agreement 2015-03) to guarantee consistent monitoring methods and uniform submission of data by the relevant OSPAR Contracting Parties.

While plastic ingested by the fulmar was selected as an ‘ecological indicator’ for trends in marine litter by OSPAR (EcoQO 3.3; OSPAR, 2009) and as an indicator for ‘impact on biota’ in the MSFD (MSFD Indicator 10.2.1; EC, 2010), the results of fulmar monitoring have a broader additional significance:

• both size range and location of debris monitored by fulmars is complementary to macro-debris monitored on beaches (MSFD Indicator 10.1.1)
• fulmar ingestion has a focus on floating marine debris and addresses MSFD Indicator 10.1.2 (trends in litter in the water column)

• the size range of plastics ingested by fulmars could be used as a functional instrument for monitoring microplastics at the upper end of the their size scale (MSFD Indicator 10.1.3 trends in micro-particles)

Suitability of fulmars for monitoring marine litter

Fulmars are pelagic (open sea) seabirds that belong to the large group of the tubenoses (Procellariiformes) of which the albatrosses are the best known representatives. These birds forage exclusively at sea and never on land, and rarely forage close to shore. The fulmar is a poor diver, and thus feeds on what is available at or within a few metres of the water surface. Like most tubenosed seabirds, fulmars regularly ingest a variety of marine debris, probably mostly taken directly and intentionally because resembling prey, or unintentionally when mixed with attractive food wastes. But indirect ingestion, for example, through preying on fish with ingested plastics or scavenging on guts of other dead animals, will also occur. A preliminary survey suggests that approximately 90% of ingested plastic items (not threads or soft sheets) found in the first glandular stomach of fulmars is 10 mm or less in diameter, and over 50% is 5 mm or less (Bravo Rebolledo, 2011). Thus, litter ingested by fulmars is mostly in the micro (<5 mm) and meso (5–25 mm) size range. Unlike most gulls, fulmars do not normally regurgitate indigestible components of their diet, but gradually grind these in their muscular stomach (gizzard) until particles are worn or broken into sizes small enough to pass into the intestines and be excreted (which appears to happen at particle sizes of approximately 2–3 mm; Bravo Rebolledo, 2011). Fulmar stomach contents integrate litter abundance encountered during feeding over periods of days to weeks (Van Franeker and Law, 2015).

Previous assessment

OSPAR’s Quality Status Report (QSR 2010) included an assessment of the North Sea EcoQO on plastic particles in seabird stomachs. The percentage of fulmars in the Greater North Sea with more than 0.1 g of plastic in the stomach ranged from 45% to over 60%. The English Channel area was the most heavily polluted area while the Scottish Islands were the ‘cleanest’ with a mean mass for plastics in fulmars of about a third of the level encountered in the English Channel. Data from the Faroe Islands (Arctic Waters) were included for comparison. The EcoQO was probably only achieved in Arctic populations. A long monitoring series from the Netherlands showed a significant reduction in plastic abundance from 1997 to 2006, mainly through a reduction in raw industrial plastics.

The Fulmar EcoQO methodology is also being used outside the OSPAR Maritime Area, in the North Atlantic and North Pacific areas (e.g. Provencher et al., 2009; Avery-Gomm et al., 2012; Kühn and Van Franeker, 2012; Bond et al., 2014; Trevail et al., 2015; Herzke et al., 2016) allowing wide spatial comparisons of marine litter distribution.

In the Dutch sector of the North Sea, changes in stomach contents of fulmars have been monitored since the 1980s. This monitoring provides contextual information that is valuable when interpreting OSPAR’s Indicator at the Greater North Sea scale. Although the overall abundance of plastics in the stomachs of fulmars from Dutch beaches has shown unexplained changes over time (Figure a.A), rapid reductions in abundance of industrial plastics within one to two decades (Figure a.B) have shown that fulmar stomach contents rapidly reflect changes in source-specific plastic abundances in their environment (Van Franeker et al., 2011; Van Franeker and Law, 2015), and are thus an effective way to assess the success of policy measures, reflecting the improved environmental quality for marine organisms and the pelagic marine environment.

At this point in time, OSPAR IA 2017 indicator assessment values are not necessarily to be considered as equivalent to proposed EU MSFD criteria threshold values, but can be used for the purposes of their MSFD obligations by those Contracting Parties that wish to do so.

Full details of methods are provided in the OSPAR Guidelines for Monitoring and Assessment of Plastic Particles in Stomachs of Fulmars in the North Sea Area (OSPAR Agreement 2014-03) and are repeated here only in a descriptive summary.

Corpses of dead beached birds or accidentally killed specimens are mostly collected by volunteer networks, but processed at experienced laboratories. At dissection, in addition to the date, the finding location is specified by a system of area codes and geographical coordinates for the area or more detailed location. Based on several internal and external anatomical characteristics, birds are classified as either adult or non-adult age group. The pilot study for fulmar monitoring (Van Franeker and Meijboom, 2002) showed that age is a relevant variable because younger birds generally have more plastics in the stomach than adults. Thus, in cases where samples to be compared have strongly different age compositions, analyses may need to be specified for separate age groups. Because age characteristics are gender specific, data recording includes sex, although there is currently no evidence for a relevant influence of gender on plastic abundance in stomachs.

Stomach contents are carefully rinsed in a sieve with a 1 mm mesh and then transferred to a petri dish for sorting under a binocular microscope. This size of mesh is used because smaller meshes become easily clogged with mucus from the stomach wall and with food remains. Analyses using smaller meshes were found to be extremely time consuming and particles smaller than 1 mm are very rare in the stomachs (Bravo Rebolledo, 2011) and thus contribute very little to numerical abundance of particles and even less to plastic mass.

Two types of plastic categories are distinguished in the OSPAR Common Indicator. Industrial plastic pellets are separated from consumer debris such as sheets, foams, threadlike materials, and hard fragments. For each of these categories the number of particles and mass (in grams and to the fourth decimal place) is recorded. The final assessment is based only on the total weight of plastics in stomachs, but industrial and consumer waste plastics have different sources and as such provide very useful information for interpreting the monitoring data.

Data collected in this way make it possible to determine:

• frequency of occurrence (’incidence’, the proportion of birds having plastic in the stomach)
• arithmetic average and standard error of the mean for number or mass of plastic
• geometric means (from transformed data to reduce impacts of data outliers)
• the indicator performance (i.e. the percentage of birds exceeding the threshold level of 0.1 g of ingested plastic as defined in the OSPAR long term goal)

The reference level for the presence of plastics in stomachs of northern fulmars (or any marine organism) is zero, because synthetic materials are solely man-made, and were only introduced into the marine environment in about the mid-1900s.

However, accepting that incidental losses are unavoidable, OSPAR (2008, 2009) has defined an (undated)long-term goalfor the fulmar indicator in the North Sea as:

“There should be less than 10% of northern fulmars (Fulmarus glacialis) having more than 0.1 g plastic particles in the stomach in samples of 50 to 100 beach-washed fulmars from each of 4 to 5 areas of the North Sea over a period of at least five years”.

Therefore the basic monitoring information required is the total mass of plastic in individual stomachs, and the percentage of stomachs exceeding the 0.1g threshold level (referred to as ‘Indicator performance’).

The OSPAR Assessment for abundance of plastics in stomachs of northern fulmars is therefore calculated as the percentage of investigated birds exceeding the 0.1 g threshold level of plastics in the stomach over the most recent five-year period of available data.

In this assessment, the most recent five-year period is 2010–2014. It is important to emphasise that all data concerning average ingested debris or indicator performance are so called ‘population averages’, meaning that clean birds without any plastic in the stomachs are included in all calculations. Analyses in the pilot study by Van Franeker and Meijboom (2002) suggest that about 40 stomachs is the minimum sample size to obtain a reliable figure for plastic ingestion representative of a selected area and period of time. This recommendation should be taken into account when spatial aggregations of data are being made.

For the fulmar indicator, OSPAR has sub-divided the Greater North Sea into five sub-regions (Figure 3): Scottish Islands (Orkney and Shetland), East England (includes incidental data for eastern Scotland), English Channel, South-eastern North Sea (Belgium, Netherlands, Germany), and Skagerrak (Denmark, Norway, Sweden). Significant spatial differences occurred, with highest level of pollution in the English Channel area, gradually decreasing to more northern regions within the North Sea and beyond (Van Franeker et al., 2011; Van Franeker and Law, 2015). In some regions, such as the Swedish Kattegat, there are insufficient numbers of beach washed birds to undertake monitoring.

OSPAR has set the same long-term goal for all these North Sea areas, however the timeline for reaching this long-term goal is not specified, but is certainly relevant for policies under the EU MSFD in Descriptor 10 Indicator 2. Good Environmental Status (GES) must be achieved by 2020 (Commission Decision, 2010). However OSPAR has not yet identified regional or overall targets to be achieved by 2020. Power analyses of Dutch data in the pilot study by Van Franeker and Meijboom (2002) indicate that fulmar monitoring data may be expected to detect statistically significant trends (p<0.05) over periods of at least 4–8 years depending on the type of plastics considered: indeed, periods of significant change have been observed in early monitoring years for both consumer plastics and industrial plastics, especially the latter (Van Franeker et al., 2011; Van Franeker and Law, 2015). Trends in quantities of plastics ingested can be shown using either annual updates of the five-year running average in the mass of ingested plastic or for the indicator performance. In such graphs, each data point thus overlaps with four years of data from the previous data point. This ‘smoothes’ most ad hoc variability in the data and emphasises longer-term trends. However, these are only visual representations of the data and have no statistical meaning. As agreed by OSPAR and published in the scientific peer-reviewed literature, the method to statistically evaluate trends in plastic ingestion, use linear regression analysis of log transformed mass of plastics in individual birds against the year of sampling over a period of the most recent ten years.

For evaluating regional differences, plastic data were fitted in a negative binomial generalised linear model with region included as a factor, and the test statistic is a t-score based on residual variance for the region (Van Franeker et al., 2011). Data for the current analyses were stored in Oracle. Graphs were made up in Microsoft Excel and statistical analyses for time trends or regional differences were conducted in Genstat (18th Edition).