Microplastics in Fish Species and Their Environments

Abstract

Microplastics are a global concern, but Philippine studies are lacking.
The study investigates microplastic ingestion in mullet (detritus-feeding) vs. rabbitfish (herbivorous), and freshwater vs. marine fishes.
Herbivores ingested more microplastics (58.57%) than detritivores (30.0%).
A weak correlation (0.06) exists between fish weight and microplastic amount.
Marine fishes ingested more microplastics (66.0%) than freshwater fishes (45.0%).
A very weak correlation was observed between fish weight and amount of MPs ingested.
Estuary fish ingested more microplastics than those in other stations.
No significant differences (p = 0.23) were found between microplastics in water samples from each sampling station.

1. Introduction

Plastics have been used for over a century and accumulate in the environment due to being non-biodegradable.
Plastic pollution has gained attention in recent decades (Andrady, 2015).
Non-biodegradable plastics take centuries or millennia to degrade (K¨arrman et al., 2016), breaking down into microplastics.
Microplastics are < 5 mm in size (Rochman et al., 2016) and easily disperse in aquatic ecosystems (Cole et al., 2011; Wright et al., 2013).
Microplastics are the most abundant size of plastic in marine environments (Lusher et al., 2014; Zhao et al., 2014).
Aquatic organisms readily ingest microplastics.
Microplastics' composition and surface area allow them to adhere to waterborne organic pollutants (Cole et al., 2011).
These toxins enter the food web, potentially leading to bioaccumulation (Teuten, 2009).
Marine pollution research focuses on microplastics due to their threat to marine biota.
Microplastics cause changes in marine habitats (Ivar do Sul et al., 2015) and physiological disorders in marine biota (Tanaka et al., 2013).
Microplastic ingestion can cause digestion blockage and contaminant transfer in fish and invertebrates (Rochman et al., 2013; Wright et al., 2013).
Determining the extent of microplastic ingestion is vital for toxicology testing (Rochman et al., 2015).
The Philippines ranked third in mismanaged plastic waste production in 2010 (Jambeck et al., 2015), but microplastic ingestion studies are few.
Most microplastic studies are in the North Sea, Caribbean Sea, Mediterranean Sea, and China Sea.
Freshwater ecosystems have received less attention than marine environments.
This study aims to determine and compare microplastic amounts in fish guts from Cancabato Bay, differing in feeding guilds, size, and weight.
The objectives include determining and comparing microplastic amounts in detritus-feeding mullet (Valamugil speigleri) and herbivorous rabbitfish (Siganus canaliculatus), identifying microplastic types (fiber or fragment), and examining the interconnectedness of microplastic concentration among fishes and their surrounding waters in the northern part of Leyte Gulf.

2. Materials and Methods

2.1. Study Sites

Fish samples for the first section were collected from Cancabato Bay (Fig. 1A) in Tacloban City, Leyte.
Valamugil speigleri (mullets) were sampled in the southern mid portion of the bay.
Siganus canaliculatus (rabbitfishes) were collected along the eastern portion near the coast where seagrasses thrive.
For the second part of the study, water and fish samples representing the freshwater ecosystem were collected from the Lawaan River in Lawaan, Samar (Fig. 1B).
Marine fish species and seawater samples were collected from the adjoining estuarine of this river system, along the northern coastal part of the Leyte Gulf.
Two sampling points (S1, farthest from shore, and S2 in the estuarine area) were established in the coastal waters, and another two, up and along the connecting river system (S3, near a sewage outlet, and S4 located much upstream near households).

2.2. Sample Collection and Quality Control

Fish samples were placed in sealed containers, labeled, iced, and frozen within 2-3 hours of capture and thawed at 20-25°C before examination.
Materials for dissection, digestion, and processing were made of glass or ceramic to minimize contamination.
Laboratory materials were washed thoroughly, rinsed with distilled water, air-dried in an inverted position, and kept inverted when not in use (Lusher et al., 2015).
Fish species were dissected, and their removed guts were immediately placed in clean graduated cylinders for volume measurement.
A blank experiment was performed without any gut tissues (Jabeen et al., 2017).
A control was performed every time laboratory analysis was done where 20 li of tap water was analyzed.
Part one used 70 fish from Cancabato Bay, equally representing both species. Valamugil speigleri and S. canaliculatus were caught by fishermen using gillnets at daytime on February 10 and 28, 2018, respectively.
Part two samples were collected between February 17 and March 7, 2018, along the Lawaan River and its adjoining portion of the Leyte Gulf.
Fifty fish samples were collected from each of the four stations, i.e. S. canaliculatus from S1 and S2, and K. rupestris from S3 and S4. Water and fish sample collections were done almost simultaneously in each of these stations.
Basic measurements were recorded for each fish, including total length (mm) and wet body weight (g).
Digestive tracts were removed by dissection, from the start of the esophagus until just before the cloaca (Lusher et al., 2013).
The entire digestive tract was dissolved in three times the volume of each GI tract in 10% KOH and was incubated at 60 ◦C for 12–24 h (Rochman et al., 2015).
The volume of the gastrointestinal tract was measured using the displacement method.
After it was placed in the oven, the remaining material was filtered using a series of sieves: 4000 μm, 250 μm, and 150 μm.
The materials retained were transferred to a petri dish by inverting the sieve and flushing with minimal amounts of distilled water as recommended by the modified method used by Masura et al. (2015).
The remaining materials were then filtered using a 125 μm filter paper, then transferred to clean petri dishes.
The dishes were covered to prevent air contamination and to avoid any microplastics from being blown away, then labeled, after which the samples were allowed to dry for microscopic examination.
Water sampling stations were simultaneously established <50 m within each fish sampling station.
For each station, both vertical hauls and horizontal tows were done to define the difference in the MPs abundance between the vertical column and the surface water.
Water samples were collected using a 20 μm zooplankton net.
The plankton net was towed horizontally along the sub-surface layer for 20 min (Isobe et al., 2014).

2.3. Microplastics Identification and Data Analyses

The particles retained were visually inspected using a Wild Heer-Brugg M5A stereomicroscope.
Examination was done methodologically to avoid double counting. The shape and color of the microplastics were determined by visual inspection.
Several criteria (Filella, 2015; Hidalgo-Ruz et al., 2012; Lusher et al., 2014; Nor´en, 2008) were used during the visual classification of the particles.
Suspected microplastics were verified by carefully touching each with a hot needle tip. If the particle deforms due to contact with the hot needle, it is considered plastic (De Witte et al., 2014).
The sizes of the microplastics were measured digitally using the software ImageJ (Rochman et al., 2015).
The microplastics were then classified as either fibers, fragments, etc.
Statistical analyses were conducted using Microsoft Excel 2016 Data Analysis tool pack and the JASP statistical program.
Mann-Whitney U test was used to determine differences in the amount of plastic items in fish that had ingested MPs based upon species and type (Murphy et al., 2017).
A Pearson moment correlation was also conducted on the gastrointestinal (gut) weight and the number of plastic items in the fish with microplastic.
Independent-Samples t-test was performed on the total number of microplastics in the fish species as well as the two sampling sites, at 95% confidence level.
Horizontal tows and vertical hauls data were also subjected to this same test.
Additionally, for the second part of this study, significant differences in the microplastic amount among fishes and water samples collected among all four sampling sites were observed through one-way analysis of variance (ANOVA), followed by Tukey test’s HSD test.
Significant differences were recorded at p < 0.05.
Finally, Pearson’s correlation test was also performed.

3. Results and Discussion

Part one of the study dealt with microplastic ingestion of the S. canaliculatus which is primarily herbivorous, as opposed to the detritivorous V. speigleri.
Out of the 70 collected and dissected individuals for each species, 41 rabbitfishes were positive of microplastics, while only 21 mullets had MPs. This variability in ingestion may be related to the feeding habits of the species (Lusher et al., 2015).
According to Woodland (1990), rabbitfish feed mainly on benthic algae, and to some extent, on seagrasses.
Gutow et al. (2016) observed that MPs attached on leaf surfaces of the seaweed, Fucus vesiculosus. Hence, it is possible for S. canaliculatus to ingest microplastics while feeding on the plant material of algae.
El-Sayed (1994) reported significant amounts of zooplankton, e.g. copepods and amphipods, in the guts of S. canaliculatus, though these fish are primarily herbivores. These zooplankters may have ingested the microplastics while filter-feeding (Cole et al., 2013), and can be deduced as the vectors for MPs entering the rabbitfish’s gut. They could absorb organic pollutants, transport them through the marine environment, and release these pollutants in living organisms (Remy et al., 2015).
In contrast, V. speigleri feed on small algae, diatoms, and other organic matter, both living and detrital, taken in with sand and mud.
Microplastics can be trapped in detritus, thus are commonly ingested by the detritivores (Van Sebille et al., 2015).
Naidoo et al. (2016) emphasized that high incidence of plastic ingestion by another mullet, Mugil cephalus, may be due to their mode of indiscriminate benthic feeding, MPs tend to accumulate in the intertidal sediments.
Exposure of fish to MPs occurred primarily through ingestion from their targeted food, or taken in as incidental capture of items mistakenly identified as prey, or they may be ingested with prey items that already contain microplastics (Lusher et al., 2015).
The larger number of microplastics were in rabbitfish (85 MPs), in contrast to the mullets (33 MPs) (Fig. 2).
From the S. canaliculatus individuals, 33 had ingested fibers while 20 had fragments in their gut. For V. speigleri, 15 of them ingested fibers, whereas 11 had fragments.
Fibers dominated over fragments in the gut of both S. canaliculatus $(0.86 \pm 1.39)$ and V. speigleri $(0.30 \pm 0.69)$ .
Smaller plastics, i.e. fibers, are more prone to vertical transport than fragments (Reisser et al., 2015), and more open to ingestion than that of the other types.
Possible source of ‘fibers’ could be waste water from washing machines since the dominant color of fibers found was blue, which could have come from clothing, especially jeans. The rest could come from furnishing, hygiene products, and nappies.
V. speigleri had higher amount of the fragment type than S. canaliculatus.
The p-values for both species were larger than the α-value (p > 0.05), implying no significant difference between fish weight and amount of MPs in the system of both species.
Moreover, correlation coefficients for S. canaliculatus (r = 0.04) and V. speigleri (r = 0.08) suggest very weak relationship between fish weight and amount of microplastics. It means that the quantity of MPs being ingested do not entirely rely on size or weight of the fish.
Naidoo et al. (2016) also produced the same results, showing no differences in plastic ingestion relative to fish size.
In terms of average weight and length, the 100 marine species S. canaliculatus samples showed up as longer and heavier $(127.8 \pm 8.29 mm, 65.5 \pm 10.8 g)$ than the K. rupestris samples $(126.5 \pm 12.5 mm, 56.1 \pm 10.1 g)$ .
Further results show a correlation coefficient of 0.117 (p-value = 0.1), indicating weak relationship between amount of microplastics in each fish and its weight. The same result is reported in fishes in the North Sea (Foekema et al., 2013). However, opposite findings were established in the eastern coast of Hong Kong (Cheung et al., 2018), South Western Atlantic estuaries (Possatto et al., 2011). Aside from fish weight, some of the more important factors that affect MP ingestion are the feeding strategies of fish and their habitat.
Siganus canaliculatus collected from the two sampling stations had a total of 101 (59%) microplastics found in their gut, (S1: 39 MPs; S2: 62 MPs). This total MPs was far higher than the 29% observed by Rochman et al. (2015) in Indonesia. But this present study’s number is exceeded by that of van der Hal et al. (2018) where MP vulnerability of rabbitfishes reached 92%.
On the other hand, K. rupestris (Rock flagtails) from S3 (39 MPs) and S4 (31 MPs) revealed only a total of 70 (41%) microplastics.
The bulk of MP ingestion studies had focused on marine organisms, but this study provided evidence that K. rupestris, a typically freshwater species, was also prone to MP ingestion. A higher MP occurrence (41%) was established by this study, compared to Sanchez et al. (2014) that sampled Gobio gobio (12–26%; n = 186) in the French rivers. Taken all studies into consideration, microplastics occurrence may reach up to 40%.
Total number of microplastics inside the S1 rabbitfishes was found to significantly differ from those found in the S2 rabbitfishes (p = 0.042, α = 0.05) and the S4 Rock flagtails (p = 0.036, α = 0.05) (Fig. 3). MPs from S3 Rock flagtails also significantly differed from the rabbitfishes (p = 0.047, α = 0.05).
Amount of MPs found in the guts of S1 rabbitfishes significantly differed from the S4 Rock flagtails samples (p = 0.036, α = 0.05). The highest amount of MPs (1.24 MP/fish) were from the S2 rabbitfishes, while the lowest (0.62 MP/fish) was recovered from the S4 Rock flagtails. The difference in amount of retrieved MPs may be attributed to S2’s location which was nearest the estuary where riverine inputs mostly end up. The very few households in S4 (upstream) may have contributed these MPs into the river.
Fish are known to ingest microplastics of various types (Possatto et al., 2011; Wagner et al., 2014). The dominant MP type for S. canaliculatus was fiber (41%) followed by fragments (38%). The measurement of microfiber is usually 11 μm in diameter and 15 μm in length. Such small sizes make them available for ingestion by fish species (Frias et al., 2010). In the microscopic examination, some fibers appeared in clumps, which must have prevented the fish from egesting these particles. Fibers usually originate from synthetic clothing, ropes or even small remains from fishnets (Wieczorek et al., 2018). The remaining 21% consisted of pellets (17%), foam (3%), and microbeads (2%). As with the rabbitfishes, most of the Rock flagtails ingested fibers (42%), followed by fragments (36%), foams (9%), pellets (7%), and microbeads (7%). Other microplastic types isolated were pellets, microbeads and foams, constituting only 10%, 4% and 3% respectively of the total plastic debris isolated from all fish samples.
World consumption of plastic pellets has reached up to 257,000 kt per year (Boucher and Friot, 2017). These pellets are lost in the environment during manufacturing, processing or transport from land. The pellets may then be transferred through wind or washed off by rainfall, ending up in the sewer and into the water bodies (Lassen et al., 2015). Many field studies also report the occurrence of plastic pellets in the soil. Microbeads originate from various kinds of personal care products such as toothpaste, shower gels, or exfoliants. Microbeads were obtained from fish samples in S2 (3 items), S3 (2 items) and S4 (2 items). No microbeads were recovered from S1, for the probable reason that it is the station farthest away from all the households. Microbeads are small items that could have settled into the sediments through time. Such vertical movement of the microbeads also depends on its polymer density relative to the water density (Hidalgo-Ruz et al., 2012). Microbeads and pellets are more often than not released into the environment as primary microplastics. They pose more risks to aquatic organisms that ingest them. Foams have also been found in digestive tracts of fishes. The study of Rochman et al. (2015) recovered 33% of foam coming from Indonesia and 3.33% of foam from USA. Both of these aforementioned studies have a larger recovery percentage compared to the present study (3%). For now, it is difficult to discuss sources of MP with certainty since there is no information on polymer type.
No significant differences were found among and within the different types of microplastics found in the gut samples of S. canaliculatus (p = 0.173, α = 0.05) and K. rupestris (p-value = 0.199, α = 0.05). In terms of frequency however, it was easily observed that fibers were the most abundant type of MP: 26% and 16% is observed in rabbitfishes and Rock flagtails, respectively. Several studies (Rochman et al., 2015; Murray and Cowie, 2011) also reported the dominance of fiber among the other morphotypes. Fragments were found to be 41% of all the MPs collected from all fish samples, wherein 41 items where isolated from S. canaliculatus and 29 items from K. rupestris. Alarmingly, these values were much higher percentages than those fragments recovered from fishes in China (15.4%) (Jabeen et al., 2017) and California (3.33%) (Rochman et al., 2015). Incidentally, fragments can reach as high as 60%, like those from sampled fishes from an Indonesian market (Rochman et al., 2015). The stark contrast in these results may be because of the differences in waste management strategies in the sampling areas. The least abundant microplastic type is foam at 1% for S. canaliculatus, and K. rupestris with 2%.
The highest number of microplastics in the water was collected from S2 (0.96 MPs/m3). Among all the sampling stations, S2 situated relatively nearest the estuary, the most vulnerable sampling station to land-based plastic pollution (Fig. 4). Plastic wastes that end up in the aquatic environment are directly discarded along the seacoasts and river. These items degrade into smaller fragment over time. This result indicated inputs through rivers (Cheung et al., 2016) have relative importance. It could be the result of relatively longer distance covered during sampling (892 m) compared to other water sampling stations. Although S4 is a relatively remote sampling station compared to other, the present study still recovered MP items from the water, providing evidence for the widespread nature of this issue.
There is a positive correlation in the number of microplastics found in the gut of S. canaliculatus and the number of microplastic in S2 (r = 0.89) and S1 (r = 0.67). Also, a positive but weak correlation (r = 0.34) was found in the K. rupestris samples and water samples from S3 while a negative correlation was found in S4 (r = − 0.45). Fiber was the dominant type of microplastic found in all fish (72 items) and water samples (88 items).

4. Conclusion and Recommendations

The contamination of aquatic ecosystems with microplastic (MP) is a global ecological problem of increasing scientific concern. This has stimulated a great deal of research on the occurrence, interaction and uptake by aquatic organisms.
Part one of the study was able to establish that herbivores, represented here by the S. canaliculatus, could ingest large amounts of plastic debris than detritivores, like mullets. Fibers, as MPs morphotype, dominated the gut of both species. However, data for both species failed to strongly relate amount of MPs to fish weight.
In part two, more microplastics were found to be ingested by S. canaliculatus than that of K. rupestris. The present study confirmed that K. rupestris, a freshwater fish representative, is also vulnerable to ingest microplastics. This may be considered as a preliminary report for this fish species. Again, no correlation was found between fish weight and microplastic counts, maybe because there are other factors that affect level of ingestion such as feeding habit of fish species and trophic transfer through its prey.
Future studies may consider the sampling and accounting of the amount of microplastics on the food sources: seagrasses for Siganus species, decaying matter for Valamugil species, and smaller fishes and arthropods for the Kuhlia species. This may be beneficial in future works as it could be used to determine whether more plastics are found in these food sources than those found in the fishes’ guts.
Lastly, the widespread occurrence and ingestion of microplastic indicates that future research across a wider range of species and habitats should be considered in order to fully establish the potential effects of microplastics in both the marine and freshwater environments.