香京julia种子在线播放

As a consequence of decades of marine litter entering our seas (Ryan, 2015), microplastics have been found in coastal and pelagic environments around the globe with an ever increasing distribution (Barnes et al., 2009; Galgani et al., 2015). Considering the prevalence of microplastics, there is now a substantial amount of research effort investigating their abundance in the gastrointestinal tracts of various organisms. For example, some early studies found 83% of Nephrops norvegicus had microplastics in their tracts (Murray and Cowie, 2011) and Lusher et al. (2013) found microplastics in the tracts of 35% of the pelagic and demersal fish species examined. Indeed, Gall and Thompson (2015) have reported that over 690 marine species are impacted by marine litter. More recent studies have moved from quantifying which animals have ingested microplastics to examining the physical and health implications of microplastic ingestion (Rochman et al., 2013; Cole et al., 2014; Pedà et al., 2016). For example, Wright et al. (2013) demonstrated how the ingestion of microplastics by the polychaete Arenicola marina (lugworm), an important ecosystem engineer of Northern Europe's intertidal zones, caused inflammation and decreased feeding and ultimately depleted energy reserves. Such studies have prompted researchers to investigate the impact on ecosystem processes. Indeed, Cole et al. (2016) found that microplastics alter the sinking rates of copepod faecal pellets if ingested and in consequence may affect the downward flux of carbon to the ocean floor. With the increasing evidence that microplastics represent an ecosystem and environmental health concern, UNEP and the EU Commission have established bodies and efforts to guide in decision making and legislation (Galgani et al., 2013; UNEP, 2016). Furthermore, several governments have taken legislative steps by introducing a ban on microbeads in cosmetics and detergents by 2020 (Sutherland et al., 2017).

Despite this substantial increase in studies investigating the ingestion of microplastics and their associated impacts, there are still important taxa playing key roles in ecosystem functioning that have not been well-studied. Mesopelagic fish inhabiting the disphotic zone of the pelagic realm (200–1,000 m depth) from the Arctic to the Antarctic (Gjøsaeter and Kawaguchi, 1980) are one of these understudied groups. Many species are known to undergo diurnal vertical migrations by residing at depth during the day before migrating to the surface at night to feed (Gjøsaeter and Kawaguchi, 1980). Smaller mesopelagic fish such as Myctophum punctatum and Benthosema glaciale feed by filtering zooplankton, predominantly copepods, euphausiids, amphipods, eggs, and larvae over their gill rakers (Scotto di Carlo et al., 1982; Roe and Badcock, 1984). Larger mesopelagic fish such as Stomias boa and Serrivomer beanii also actively target decapods and fish using their anterior vertebrae and branchial apparatus to swallow larger prey (Roe and Badcock, 1984; Bauchot, 1986). Thus, mesopelagic fish are exposed to microplastics either through the direct consumption of a microplastic mistakenly identified as prey item, or indirectly, through the consumption of a prey item (e.g., copepod or euphausiid) that had already consumed microplastics.

As mesopelagic fish undergo large vertical migrations, they are known to play a key role in the cycling of carbon and nutrients to the deep ocean (Radchenko, 2007; Davison et al., 2013). For instance, Radchenko (2007) has shown that such species in the Bering Sea transport 15,000 tonnes of carbon daily to the deep ocean. Therefore, the ingestion of microplastics by mesopelagic fish may disrupt carbon cycling and aid in the transport of microplastics to deeper waters, as suggested by Lusher et al. (2016).

The importance of mesopelagic fish was recently further highlighted in studies by Kaartvedt et al. (2012) and Irigoien et al. (2014) who found that the mesopelagic fish biomass in the global oceans may have previously been underestimated by at least one order of magnitude due to avoidance behaviour and mesh extrusion. Because they make up such a large biomass in the pelagic realm they provide an important food source for a variety of predatory fish and marine mammals which, through trophic transfer from their mesopelagic fish prey, may suffer from the impacts of microplastics and associated toxins (Lusher et al., 2016). Some of the species preying on mesopelagic fish such as tuna and swordfish (Scott and Tibbo, 1968; Varela et al., 2013) are commercially important food sources and thus toxins and microplastics transferred to these species may also pose a danger to human health. To date mesopelagic fish have not been exploited as a human food source due to the high levels of wax esters in their tissue (Gjøsaeter and Kawaguchi, 1980). This may change in the near future as the demand for fish protein increases and new policies (e.g., Blue Growth Strategy by the European Union) encourage sustainable exploitation of potential resources (St. John et al., 2016). Furthermore, the food safety issues concerned with microplastics and the associated toxin exposure through the consumption of commercially exploited fish have recently been outlined in an extensive report by the Food and Agriculture Organization of the United Nations drawing attention to the potential threat of microplastics to human health (Lusher et al., 2017).

However, to date, only a few studies have investigated microplastic ingestion by mesopelagic fish: one in the North Atlantic (Lusher et al., 2016) and two in the North Pacific Gyre region (Boerger et al., 2010; Davison and Ash, 2011). Since then, new and improved methodologies for microplastic extraction have been developed with an emphasis on ultra-clean techniques in order to prevent airborne contamination (Wesch et al., 2017).

This study set out to quantify microplastic ingestion by mesopelagic fish from an eddy region in the Northwest Atlantic, known to be a hot spot for mesopelagic fish (McKelvie, 1985; Fennell and Rose, 2015) and potentially microplastics (Yu et al., 2018). Specifically, this study investigated whether: (1) species, stomach fullness, and the depth at which fish were caught at had an effect on the amount of microplastics found in the gut contents of mesopelagic fish, and (2) how the type, shape, and size of microplastics found in the gut contents compared to those found in the surface waters. Importantly, we applied strict measures to prevent microplastic contamination during extraction and identified microplastic type using micro-FTIR spectroscopy.

A total of 280 fish was captured of which 233 were examined for the presence of microplastics in their gut contents. The most common species amongst the subsampled fish were the spotted lantern fish M. punctatum (with 86 individuals, or 37% of catch), the glacier lantern fish B. glaciale (69 indiv., 29%) and the white-spotted lantern fish Diaphus rafinesquii (34 indiv., 15%). The remaining species were the Rakery beaconlamp Lampanyctus macdonaldi (16 indiv., 7%), the stout sawpalate S. beanii (14 indiv., 6%), the scaly dragonfish S. boa (9 indiv., 4%) and Gonostoma denudatum (5 indiv., 2%). Where information on sexual maturity size exists (M. punctatum, B. glaciale, D. rafinesquii, G. denudatum), every sampled fish was assessed as being sexually mature.

Overall 73% of fish contained plastics in their stomachs with G. denudatum having the highest frequency of occurrence (100%), followed by S. beanii (93%) and L. macdonaldi (75%) (Table 1). In total, 452 microplastic fragments were extracted from the fish gut contents, with an average of 1.8 microplastic fragments per fish. The highest average number of microplastics in the gut contents was recorded in S. beanii (2.36), followed by M. punctatum (2.28), and G. denudatum (2.2) (Table 1).

There was no significant difference between the median stomach fullness indices of fishes which had microplastics in their stomachs and those that did not (W = 5253, P = 0.976). Furthermore, there was no significant difference in median microplastic counts among the seven different species (H = 10.904, d.f. = 6, P = 0.091), nor between fish caught in shallower and deeper waters (U = 5877, P = 0.389).

In total, 341 particles were found in the surface water samples (8 samples totalling 2,400 L of surface water) resulting in an estimated concentration of 14 microplastic fragments per 100 litres of water. Plastics identified from fish guts were very similar to those found in the surface waters (Figure 2). Ninety-eight percent of microplastics identified from the fish and 99% of those identified from the water samples were classed as fibres with the remainder being flattened fragments of plastics. Recorded microplastic colours included black, grey, blue, green, purple, red, yellow, and white. Black and blue were by far the most common colours, followed by red, making up 67, 18, and 6% and 65, 26, and 5% of sampled plastics from fish guts and surface waters, respectively. Polymers of other colours only made up a minor fraction of the particles in both cases (Figure 2). Likewise, sizes of extracted microplastics were very similar between those found in fish (mean ± SD: 969 ± 1,048) and in surface waters (mean ± SD: 985 ± 1,101). The smallest recorded polymer fragment had a length of 42 μm and the largest a length of 8,150 μm.

Micro-FTIR analysis was successfully carried out for 118 of the 191 microplastic fragments originating from 35 fish and from one surface water sample. The 73 particles which could not be assessed for their polymer nature either fractured when pressure was applied by the diamond of the micro-FTIR machine or did not show a significant fit with any of the materials within the FTIR spectra library and thus were excluded from analysis. Polymers identified from fish and water samples were of similar polymer nature with the majority being polyethylene, followed by methyl cellulose and a relatively small proportion were identified as alginic acids. An exception was nylon, which comprised 10% of the particles found in the fish guts, but was not identified amongst the particles extracted from the surface water (Figure 2). Scanning electron microscopy images of two fibres extracted from fish gut contents had visible signs of polymer fracturing (Figure 3).

No microplastics were found on the filters used as blanks to ensure no airborne contamination or any contamination from the filtration equipment and procedure.

Using forensic methods, this study assessed microplastic frequency of occurrence in mesopelagic fish gut contents from a warm-core eddy in the Northwest Atlantic. We detected a significantly higher occurrence rate of 73% in contrast to previous studies reporting occurrence rates of 11% in the North Atlantic and 9 and 35% in the North Pacific Gyre regions (Boerger et al., 2010; Davison and Ash, 2011). There are several reasons which may explain our much higher frequency of occurrence. Firstly, there are no standardised methods for the extraction of microplastics from gastrointestinal tracts of fish and so different research teams have used different protocols such as visual sorting of gut contents (Boerger et al., 2010), staining of organic materials (Davison and Ash, 2011), and extraction by the use of alkaline dissolution (Lusher et al., 2016). In the latter, the authors used a similar approach to this study but used a more concentrated caustic solution (1.8 vs. 1 M) and incubated samples for a longer time period (14 days vs. 1 day). Cole et al. (2014) assessed many different extraction methodologies and found that the hydrolysis of organic compounds using a caustic solution was an efficient and viable method. However, they noted that extractions using higher concentrations and longer incubation times than recommended damaged and discoloured pH sensitive polymers such as nylon, uPVC, and polyethylene and thus these may have been underestimated previously. Furthermore, we used fine-meshed borosilicate filters in contrast to Lusher et al. (2016) who used a 250 μm filter. As a result, our study detected much smaller microplastics (down to 42 μm) which made up 20% of all detected microplastics.

Another potential explanation for differences among reported microplastic occurrence rates may be due to differences in the abundances of microplastics found in the study areas. Samples for this study were collected around a warm core eddy ~1,200 km off the Newfoundland coast which is known to be an area of enhanced biomass for all trophic levels (Dufois et al., 2016), including mesopelagic fish (McKelvie, 1985; Fennell and Rose, 2015), and may also potentially aggregate microplastics (Yu et al., 2018). Surface water samples collected within this study indeed showed 10 times higher concentrations of microplastics than reported for other regions of the Atlantic (Lusher et al., 2014) where Lusher et al. (2016) collected their samples. The other two studies collected samples at the edge region of the North Pacific Gyre, which while potentially having slightly higher concentrations of plastics, were still not located close to the centre of the gyre, known to be a hot spot for microplastics (Eriksen et al., 2014). At this point it is also important to consider how mesopelagic fish may be exposed to microplastics. All of the seven investigated species migrate to the surface at night to feed and therefore ingestion could happen through the direct consumption of microplastics mistaken as prey items or through trophic transfer from their prey species. Indeed the most common prey of mesopelagic fish are copepods, euphausiids, amphipods, larvae, and decapods and all have been reported to ingest microplastics (Carpenter et al., 1972; Setälä et al., 2014; Desforges and Ross, 2015). Lusher et al. (2016) previously excluded trophic transfer as a likely route of exposure as the average size of microplastics they found in the gut contents was 1.9 mm. This was considered too large to be ingested by their prey species but they also noted that their study only targeted plastic particles over 250 μm in size. In this study the average size of identified microplastics from fish guts was 970 μm with 20% of all plastics being smaller than 250 μm and thus trophic transfer from planktivorous prey species may indeed be a likely route of exposure. In addition to size, it is also worth noting that the colour of any microplastics is unlikely to play an important role in the ingestion of microplastics by mesopelagic fish as the colours of the microplastics identified from the fish gut contents were similar to those identified from the surface waters.

Lastly, different microplastic abundances in the gut contents may be caused by some mesopelagic species being more selective or impacted than others. For example, it is well known that some bird species are more prone to microplastic and marine litter ingestion than others (e.g., petrels: Van Franeker and Bell, 1988). However, our study found no differences in microplastic occurrence rates between the seven mesopelagic fish species examined. Neither did depth seem to explain any variation in microplastic abundances amongst individuals caught at different depths. Therefore, we can conclude that the notably higher occurrence rates reported within this study are likely due to the differences in microplastic extraction methods as well as the fact that the present study was carried out in a hot spot for mesopelagic fish and microplastics alike. While this study reports one of the highest abundances of microplastics in the gut contents of fish, other studies have reported similar results for different species, particularly in polluted areas. For example, Tanaka and Takada (2016) report a 77% encounter rate of microplastics in Japanese anchovies (Engraulis japonicus) sampled from Tokyo Bay and Nadal et al. (2016) found microplastics to occur in 68% of seabream (Boops boops) sampled around the Balearic Islands. It is also noteworthy that while Lusher et al. (2016) and Davison and Ash (2011) reported a lower average microplastic count per individual fish of 0.13 and 0.11, respectively, Boerger et al. (2010) indeed found the average microplastic count per fish to be 2.1, higher than observed by us (1.8).

Such high numbers of microplastics in the gut contents of mesopelagic fish is of great concern. Microplastics have previously been shown to adversely impact invertebrate species such as lugworms, causing weight loss, reduced feeding activity and inflammation (Besseling et al., 2013; Wright et al., 2013), and detrimental effects on the intestinal functioning of seabass (Dicentrarchus labrax) have also been noticed (Pedà et al., 2016). Furthermore, there is growing concern about the effect of chemical pollutants sorbed to microplastics. For example, Mato et al. (2001) reported up to 10⁶ higher concentrations of PCBs on polypropylene pellets than in the surrounding sea water and recently, it has been shown that Japanese rice fish (Oryzias latipes) and rainbow fish (Melanotaenia fluviatilis) readily accumulate chemical pollutants from ingested microplastics (Rochman et al., 2013; Wardrop et al., 2016).

The ingestion of microplastics by mesopelagic fish may also have secondary implications for other species as well as the entire ecosystem. Mesopelagic fish are now known to make up a substantial biomass in the pelagic realm (Kaartvedt et al., 2012) and provide an important food source for many large predators such as dolphins, seals, and tuna as well as sea birds (Cherel et al., 2008; Danielsen et al., 2010; Spitz et al., 2010; Varela et al., 2013). These taxa consume large amounts of mesopelagic fish and consequently ingest the microplastics within them. More importantly, due to trophic transfer, predators of mesopelagic fish may also bioaccumulate chemical pollutants absorbed from ingested microplastics. As some of the species preying on mesopelagic fish are commercial exploited fish the transfer of microplastics and biaccumulated toxins may also pose a threat to human health (Lusher et al., 2017).

Mesopelagic fish are also responsible for a significant amount of carbon and nutrient cycling (Radchenko, 2007). Organic material released as faeces or from dead and decaying organisms, sink very slowly from the upper surface to the deep ocean. A large proportion of this organic material is recycled by other organisms and re-released before it can reach the ocean floor. Mesopelagic fish however, undergo diurnal migrations, quickly travelling long distances from the epipelagic layer where they feed, to the deeper ocean where they deposit their faeces. Therefore, they play a key role in speeding up the downward flux of carbon and nutrients to deeper depth and circumvent recycling by other organisms (Irigoien et al., 2014). As discussed above, we are now aware that microplastic ingestion can have substantial effects on fish health and in particular digestive functions. Therefore, reported microplastic abundances in the fish gut contents may have implications for the cycling of carbon and nutrients by these species. Moreover, as suggested by Lusher et al. (2016), mesopelagic fish may aid in the downward transport of microplastics to the deep-sea benthos and cause potential harm to organisms in this habitat.

In terms of our methods, the applied protocols have successfully extracted very small plastic particles. However, the micro-FTIR spectroscopy analysis identified a large proportion of the analysed microplastics as methyl cellulose and alginic acids. This, while seeming unusual at first, is very likely a consequence of insufficient cleaning of the microplastics after extraction. Methyl cellulose is produced synthetically by heating cellulose with a caustic solution. As we used a caustic solution to hydrolyse organic materials some of the sodium hydroxide seemed to have remained on the plastic particles and potentially skewed the absorbance spectrum. Similarly, alginic acids are likely to be a reading of an outer biofilm coating of the microplastic particles which had not been removed during extraction. For future studies we strongly recommend a more thorough cleaning of plastic particles with filtered, ultrapure water. Furthermore, it is interesting to note that despite taking a forensic approach during the extraction of microplastics (Wesch et al., 2017) we noted a large amount of fibres (98%) amongst the sampled microplastics. This is in agreement with other findings (Lusher et al., 2013, 2016; Neves et al., 2015; Rochman et al., 2015; Bellas et al., 2016). However, fibres are often considered to be a contaminant of airborne nature and are sometimes excluded from analysis (Foekema et al., 2013; Rummel et al., 2016; Tanaka and Takada, 2016). As we did not observe any fibres on the filters used as blanks, we argue that fibres do indeed make up a large proportion of microplastics and are not of airborne nature. In support of this, Rochman et al. (2015) found high numbers of fibres in fish sampled from USA fish markets, but not in those sampled from Indonesian fish markets. The authors suggest that this is due to the large amount of waste water effluents carrying synthetic fibres from washing machines as such machines are more common in developed areas. In fact, the microplastics we identified from the fish gut contents closely overlapped in colour, size, shape, and type with those sampled from the surface water (Figure 2) and we can thus assume that types of microplastics sampled from organisms are a reflection of those found in the environment they inhabit.

While large gyres have been a major focus of microplastic research, this study together with that of Yu et al. (2018) show that mesoscale features such as eddies may also be a hot spot for microplastics and should be further investigated. Furthermore, future studies quantifying microplastic ingestion by predatory fish species should also consider to sampling their putative prey to investigate trophic transfer of microplastics.

In conclusion, this study reports the highest ingestion rates of microplastics in the gastrointestinal tracts of mesopelagic fish. This has important consequences for the health of pelagic ecosystems and biogeochemical cycling in general. Additionally, using forensic techniques, we provide more evidence that fibres are found throughout our oceans rather than being an artefact of airborne contamination.

TD and EM conceptualised the study; EM and AA (chief scientist) participated in the research cruise and collected samples; TD and AW designed the experiments; AW and HB performed the experiments; HB and LM took SEM images; AW, OS, and LM carried out micro-FTIR analysis; AW analysed the data; AW and TD wrote the manuscript with contributions from PC, LM, AA, and EM.