Investigating Microplastic Presence Amongst Grey Seals (Halichoerus Grypus) of the North Sea

Plastic pollution is of increasing concern to marine ecosystems worldwide. Specically, microplastics (<5mm) may interact with a variety of biota with potential to cause harm to organism health. Studies concerning microplastics are increasing, yet their occurrence within live marine mammals remains largely unexplored. Here, faecal samples collected from a haul-out site in the North Sea, were used to investigate microplastic pollution within grey seals (Halichoerus grypus). 71 microplastic particles, consisting of both bres and fragments in a variety of colours and sizes, were identied across 66 scat subsamples analysed. This indicates that marine mammals are ingesting microplastics and that faecal material can be used to indirectly and humanely record microplastic uptake data in pinnipeds. Since the current paper is the rst to document microplastic exposure amongst wild, living and free-ranging grey seals in the North Sea, further research is needed to begin to understand the biological signicance of these ndings.


Introduction
Plastic is a material of such societal bene t 1 that production has soared in recent decades -as has the quantity of marine plastic pollution 2 . With current estimates predicting that oceanic plastic litter could reach 250 million metric tonnes by 2025 3 , this anthropogenic debris is considered a major threat to marine biodiversity 4 . Recently, increased research interest has focussed on plastic debris at the micro- 5 , and nano-scale 6 , with particles less than 5mm in size being collectively referred to as microplastics 7,8 . Whether originally manufactured as microscopic in size (for instance, bres), or formed through the degradation of larger items of plastic debris 9 , microplastics have accumulated in our oceans for over 40 years 10 and now pervade even some of the most remote environments 5,11 . As such, these synthetic particles are believed to be ubiquitous within marine ecosystems worldwide 7,9,12 . Consequently, better understanding of the potential ecological rami cations of microplastic pollution is a globally recognised high-priority research area in Marine Biology 13 .
Owing to their small size, microplastics are bioavailable for ingestion by a range of marine fauna including; seabirds 14 , sh 15 , invertebrates 16 and marine mammals 17,18 . For mammalian species, microplastic ingestion has been hypothesised to occur unintentionally during prey capture in microplastic-polluted areas, particularly during lter feeding 19,20 . Furthermore, indirect microplastic consumption following the ingestion of microplastic-contaminated prey, termed trophic transfer, and this has been proven to occur under captive conditions 21 and is widely speculated to occur in natura 17,18,22 .
Microplastics are known to be associated with toxic substances including adsorbed Persistent Organic Pollutants (POP's) and/or harmful additives contained by the debris 23 . Following the ingestion of plastic particles by marine biota, [24][25][26][27] . For marine mammals, exposure to these lipophilic, microplastic-associated toxins may be related to adverse health effects, including cancer-associated mortality in California sea lions (Zalophus californianus) 28 and reproductive toxicity in dolphins 29 .
Although several papers report the presence of microplastics amongst marine mammals 5,17,19,30-32 , research typically relies on necropsies performed on stranded or by-caught individuals 33 -providing few occasions to study anthropogenic contaminants 32 and with relatively small sample sizes from individuals that have died, and may have been feeding abnormally pre-mortem 12 . Alternatively, scat analysis offers the opportunity to obtain data on microplastic exposure in live, wild and free-ranging pinnipeds through a non-invasive and humane manner 18,21 . This could permit the analysis of larger sample sizes, which may provide more representative estimates of microplastic pollution amongst wider populations 12 . Nonetheless, scat-based microplastic investigations on wild pinnipeds are sparse and none exist for grey seals inhabiting the North Sea.
In the current study, faecal material collected from 66 wild, free-ranging grey seals was analysed for microplastic pollution. We strove not only to gain preliminary data on microplastic ingestion within a grey seal population of the North Sea, but also to characterise the particles identi ed according to type ( bre or fragment), colour and size.

Results
Microplastic presence. A total of 71 microplastics were found in 43 of the 66 scat subsamples examined (supplementary table S1), with a mean of 1.08 ± 1.01 particles identi ed per subsample. Of the microplastics detected, bres (n=43; 61%) were more numerous than fragments (n=28; 39%). Amongst fragments, light blue was the main colour observed (n=10; 36%), then clear (n=8; 29%), blue (n=4; 14%), white (n=3; 11%), yellow, black and red (all three n=1; 3%; Figure 1. i.). Conversely, black and blue were the most common colours of bre (both n=18; 42%), followed by red (n=5; 11%) and light blue (n=2; 5%; Figure 1. ii.). Clear coloured bres were also seen, but were excluded from our analysis as consistent separation of these and the thread-like structures present on the lter was not possible. Therefore, the gures shown under present plastic occurrence, as clear bres were not included. Fragment sizes ranged from 30µm to 1400µm, with a mean of 248 ± 264µm. Whilst bres had a mean length of 1212 ± 811µm, the length varied between 100µm and 3400µm (Figure 1. iii.). Photographs of microplastics from the present study are included in Figure 1. iv. Data analysis. No signi cant difference was found between the incidence of one colour from any other. A statistical difference (p<0.0001) was shown, where fragments were signi cantly smaller than bres.
Laboratory controls. Microplastics were not identi ed in any of the procedural blanks or airborne contamination controls used, validating the e cacy of the measures employed to reduce the risks of contamination.

Discussion
Whilst contamination between defecation and sample collection cannot be excluded, microplastics were absent from all our blanks and controls, providing con dence that particles observed were endogenous to the scats and did not arise from laboratory-based sources. Conversely, controls were not exposed to the clear plastic sample collecting bags and this may have potentially contaminated the scats -although the range of bre/particle types observed and exclusion of clear coloured microplastics from our results excludes clear plastics derived from any source from the analysis. Given that each lter paper analysed represented 44.4 % of the initial mass of the faecal subsample, we estimate an average microplastic abundance within these individuals to be 0.81 particles per gram of dried scat. Contrary to other studies reporting both particle number per scat 18,22,33 and per gram of wet scat 18 , we consider particle number per gram of dried scat to be a robust 'relative measure' -an 'index' independent of original scat sample weight, and a useful way to present the detection rate data, as it accounts for variation in scat mass and water content. However, microplastic egestion rates in grey seals are unknown, and thus whether plastic particles would be intermittently or continually shed remains unknown.
Placing these ndings into context is problematic, since only a handful of studies have analysed wild pinniped faeces for microplastic presence 18,22,33,37,38 , and only two have examined grey seal scat, and none from Seals in the North Sea: Hudak and Sette 39 isolated two microplastic particles from 129 scats in south-eastern Massachusetts, USA, and Nelms et al. 12 found 17 microplastic particles, in 15 subsampled scats collected from Skomer Island, Wales. While caution should be applied when interpreting the preliminary work of Nelms et al. 12 due to the small sample size analysed (15 subsamples), the microplastic abundance (53%) was similar to that observed here (66%) and considerably higher than that reported by Hudak and Sette 39 (<2%). Although spatial variations may potentially in uence environmental and/or prey-based microplastic uptake in marine mammals 20 , along with disparities which may be caused by the different prey species ingested 40 , these differences in microplastic abundance may also be due to dissimilarities in the detection methods used in our study and these previous studies. Firstly, absence of an enzymatic digestion technique in the method adopted by Hudak and Sette 39 , but employed by Nelms et al. 12 and in our study, may have caused biological material to obscure microplastics during identi cation. Secondly, particles <500µm were excluded in the work of Hudak and Sette 39 . This size category constituted a great proportion (37/71) of all microplastics identi ed in our study, and the entirety of those identi ed by Nelms et al. 12 . The exclusion of certain sized particles -often through disregarding the uid collected after sieving in any analyses, which may contain small microparticles -is a procedural bias evident within other microplastic investigations on pinniped species 22,37,39 . For the present study, all material (including all uid) was collected after sieving and analysed for microplastics. While very small microplastics <500µm may still be under reported due to the limited detectability of these small particles and the presence of pores (20µm) within the lter papers used 17 , the analysis of all material collected gives con dence that the true number of particles present was recorded. Therefore, it is di cult to draw comparisons between the microplastic abundance observed here and in previous studies 22,37,39 , since these studies potentially under report the presence of small ( <500µm) particles. This emphasises the need for future studies to employ methods which can detect and report all grades of microplastic extracted, such as that utilised here and by Nelms et al. 12,17,21 , to facilitate increased comparability between datasets.
Investigations into microplastic presence amongst pinnipeds 12,18,21,31 , and marine mammals in general 17,30,32 , largely report a higher number of bre particles when compared to fragments, and this corresponds with our nding of 61% of our samples ndings being bres. Thus could be explained by two principal factors. Firstly, the bres could be disproportionately represented, as they have been suggested to have an increased predisposition for methodological contamination during sampling and analyses 19,41 . Yet, the strict contamination measures employed, and absence of any microplastics present on our controls, and the mix of colours and bre types seen in our samples, makes this seem an unlikely explanation. More likely, the nding of higher levels of bres is likely to be re ective of the high relative levels of environmental bres 7,42,43 . Domestic washing, and the textiles industry expel large quantities of micro bres into water, which often evade ltration systems and/or municipal water treatment plants, and enter the ocean 44 . It could however be the case that pinnipeds ingest these bres in higher proportions along with prey in water ingested at the time of prey capture 18 . However, what appears to be more probable is that the seals are ingesting bres indirectly via trophic transfer 21,45 . Previous studies have shown bres to have long body residence times in low-trophic level organisms, which would suggest an increased chance of secondary bre ingestion by predators 43,[46][47][48] . Microplastic bres may be transferred along food chains and re ected in greater proportions among high-trophic level species such as marine mammals 43,46,49 . Whether the higher relative number of bres compared to fragments observed here is representative of a biological process such as trophic transfer, or purely an incidental nding, is unknown.
Previous studies examining microplastic pollution amongst marine mammals commonly describe a variety of colours, including colours identi ed here 17,18,30,31 . Conversely, some investigations report a predominance of white 22,33 and blue microplastics 18,50 . Notably, after oxidation and density separation treatments, Donohue et al. 33 only recovered white microplastic fragments from northern fur seal (Callorhinus ursinus) scats. Hydrogen peroxide has been employed in some studies to digest organic debris, and this may have bleached the particles observed. Donohue et al. 33 hypothesised that the prey which seals were ingesting had disproportionally taken in white particles due to enhanced observability of white particles in the water column. Other authors have postulated that certain coloured microplastics may resemble the prey of predatory sh, thus causing visual confusion and a propensity for the ingestion of speci c particle colours by prey species 51,52 . Although the exact mechanism remains unclear, selective ingestion by prey species could result in a disparity in the microplastic colours subsequently found in seal scat 21,33 . Nevertheless, the range of particle colours in our study, combined with the lack of any signi cant difference in the frequency of any one colour against another, suggests that active selection (based on colour) by grey seal prey was not occurring in the sea area where the study was carried out.
Instead, microplastics could have been visually targeted based on size 22 or selected through olfactory signals, as is seen in foraging seabirds 53 . Alternatively, direct accidental microplastic uptake could have occurred -potentially during bottom-feeding, as suggested for harbour seals of the North Sea 37 -which would re ect a reasoning for the diversity of colours found in reported environmental microplastics 35 .
A signi cant difference was present between particle sizes -fragments were found to be signi cantly smaller than bres, with all but two being <500µm in size. This correlates with other reports on marine mammals that have employed similar protocols for the extraction of a wide size range of microplastics 5,17,33 and this may be biologically consequential. Smaller sized particles are assumed to be increasingly bioavailable, facilitating uptake by lower trophic level organisms 7,54 . Therefore, the predominance of fragments <500µm detected in our study, suggest that small synthetic particles may have a high bioavailability to ingestion throughout numerous trophic guilds 33 . Consequently, microplastics could be contaminating multiple trophic tiers of the grey seal food chain (as shown in Figure 2.), which may not only assist the widespread trophic transfer of microplastics to these top marine predators, but also to various compartments of the marine food web 7 . However, it is not possible to determine whether the microplastics recovered here originate from direct ingestion, or through trophic transfer from contaminated prey and/or organisms lower in the food chain.
The size and abundance of particles observed in the current study are unlikely to present a physical hazard (e.g. gastrointestinal obstruction) to grey seals in the same way they may do for lower trophic level organisms 9 . Equally, our microplastic recovery from faecal matter, alongside previous plastic particle identi cation from post mortems conducted on grey seal intestinal tracts 31 , indicates that microplastics are transitory, i.e.. not retained in the gut, for Grey seals. However, the potential for microplastics to cause localised harm to the digestive tract 5 cannot be excluded, since particle retention times and egestion rates in Grey seals are undetermined. Furthermore, while microplastics have been proposed to be transitory in marine mammals 5,17,55 , their presence in faecal matter is still a concern. Defecation releases particles back into the environment where they are made re-available for uptake by other organisms 18,54 , and this cycle may continue to very long periods, potentially for hundreds of years.
Microplastics are known to adsorb and concentrate hydrophobic toxins (e.g. POPs) present in the ocean 56,57 -perhaps due to their large surface area to volume ratio and the natural a nity of these compounds for the hydrophobic surface of plastic 23,26,58 . These substances, along with chemical additives used during plastic manufacture (e.g. phthalates and Bisphenol A), may leach into organisms tissues following microplastic ingestion [24][25][26][27] . Owing to the small size, and thus high bioavailability of this synthetic debris by low-trophic level species 33 , ingested microplastics could act as a carrier for transporting these substances throughout the aquatic food web 7,25,30 . The long-life spans, high trophic level status, and dense lipid stores of marine mammals, may leave them potentially vulnerable to the bioaccumulation of lipophilic contaminants 59 . Chronic microplastic ingestion by marine mammals, and/or elsewhere in the food web, may expose these megafauna to high concentrations of these chemicals with the potential to in uence population viability through immune, endocrine and/or reproductive system disruption 28,29 , but, studies demonstrating whether microplastics may facilitate increased contaminant uptake in marine mammals directly, or elsewhere in the food chain, remain absent.
Our identi cation of microplastics within Grey seal scat, collected non-invasively, and with no direct disturbance of seals, not only veri es that grey seals in the North Sea are ingesting these synthetic particles, but also indicates the pervasiveness of this issue. Currently, the health, animal welfare, and conservation implications of microplastic, and associated chemical ingestion, remain unclear and do appear to warrant further investigation. The present study shows that faecal material can be used to humanely (without disturbance) and indirectly record levels of microplastic exposure -particularly important since opportunities to conduct necropsy are infrequent, and collection of biopsies for phthalate analysis 20 would be challenging due to the diving behaviour of seals 31 , and inadvisable from an animal welfare perspective. Our method has produced results delivered as a relative measure 'particles per gram' of faecal material, and this 'metric' of environmental contamination could be used as a reference for comparison across different species and geographical areas. Plastics are relatively new agents present in the evolution of the wild animal world, which we may ignore or explore -time will probably illustrate whether monitoring of sentinel markers of animal plastic uptake should raise, or should have raised, concerns about human derived environmental change. Faecal matter retained by the sieve, along with all liquid contained after sieving, was then dried at 60 O C until no moisture remained.

Methods
Enzymatic digestion. To remove any biological material that may have concealed microplastics during later identi cation, a methodology involving enzymatic digestion was employed. This protocol was introduced by Lindeque and Smerdon 34 , modi ed by Cole et al. 35 and revised by Nelms et al. 21 for use in seal scats. only recorded if all the following criteria proposed by Norén 36 were met: 1.) No organic structures should be visible within the particle; 2.) if the present particle is a bre, there should be no tapering towards one/both ends and it must have three-dimensional bending; 3.) each particle should be homogeneously coloured.
Once identi ed, microplastics were classi ed according to their type (fragment or bre) and colour, which was visually determined. The length of each microplastic's longest axis (irrespective of type) was also measured using the eyepiece graticule. Limitations of funding, and time available to conduct the study, restricted our ability to conduct comprehensive Fourier transform infrared (FTIR) spectrometer analysis on any recovered microplastics which would have indicated the plastic polymer type present, and it is recommended that future studies of scat analysis apply methods to determine the polymer type of the plastic microparticles. This study is a proof of concept study -'could we detect particles in Grey seal scat from the North Sea'.
Data analysis. A Kruskal-Wallis test was carried out to determine if there was a signi cant difference in the incidence of any one particular microparticle colour. Secondly, a D'Agostino-Pearson omnibus normality test and Mann-Whitney U test were performed to establish if a signi cant difference was present between the particle sizes observed. Following this, a Chi-Squared test was completed to ascertain if there were signi cant differences in the frequency distribution of particle sizes overall, and then whether there were differences between individual size grade categories. All analyses were undertaken using GraphPad Prism 7®.
Contamination prevention. During laboratory procedures, contamination of samples by microplastics risks generation of inaccurate results 17 . Therefore, several measures were implemented throughout the present study to minimise, and assess the risk of, external microplastic contamination from equipment and/or atmospheric sources.
Prior to any work commencing, a cotton lab coat was worn over synthetic clothing and all work surfaces were cleaned using 70% ethanol and paper towels. Glass or metal equipment was used alternatively to plastic wherever possible and all equipment was washed meticulously with Milli-Q water before use. Likewise, the sieves used were soaked in detergent and hot water, cleaned, rinsed with Milli-Q water and then visually examined for any lasting debris and/or microplastics between samples.
Sample bags were closed following scat collection and remained closed throughout sample storage and thawing. Lids were placed on Duran bottles wherever possible and when required to be removed, such as during the addition of reagents, this was performed for as little time as possible and within a positive pressure laminar ow hood.
Our ability to minimise external microplastic contamination was appraised using two types of laboratory control. Four procedural blanks (comprising 50ml Milli-Q water) were processed identically and concomitantly to the scats, to reveal contamination by reagents, equipment and/or other procedural sources. Two lter papers soaked in Milli-Q water were placed within the laminar ow hood to serve as airborne controls. Filter papers from all blanks and controls were visually inspected under a microscope for any microplastic contamination.
The author would like to thank Professor Philip Hammond for collecting the scat samples and donating them to this research; Dr Laura Peachy for her input during the laboratory analysis; Katie Bull for giving advice in the laboratory and helping photograph the microplastics observed; and Sharon Holt for her assistance in arranging a workspace within the laboratory and completion of the relevant health and safety forms needed to undertake this research.
CRediT authorship contribution statement. Additional information.
Supplementary information. The data that support the ndings of this study are available in the supplementary information that accompanies this paper.
Competing interests. The authors declare no competing interests.