Stable isotope analysis of a newly established macrofaunal food web 1.5 years after the Hebei Spirit oil spill

Marine Pollution Bulletin 90 (2015) 167–180

Contents lists available at ScienceDirect

Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul

Stable isotope analysis of a newly established macrofaunal food web 1.5 years after the Hebei Spirit oil spill Eunah Han a,b, Hyun Je Park a,b, Leandro Bergamino a, Kwang-Sik Choi c, Eun Jung Choy d, Ok Hwan Yu e, Tae Won Lee f, Heung-Sik Park e, Won Joon Shim g, Chang-Keun Kang a,⇑ a

School of Environmental Science & Engineering, Gwangju Institute of Science and Technology, Gwangju 500-712, Republic of Korea Ocean Science and Technology Institute, Pohang University of Sciences and Technology, Pohang 790-784, Republic of Korea School of Marine Biomedical Science, Jeju National University, Jeju 690-756, Republic of Korea d Korea Polar Research Institute, Korea Institute of Ocean Science and Technology, Incheon 406-840, Republic of Korea e Marine Ecosystem Research Division, Korea Institute of Ocean Science and Technology, Ansan 426-744, Republic of Korea f Department of Oceanography and Ocean Environmental Sciences, Chungnam National University, Daejeon 305-764, Republic of Korea g Oil and POPs Research Group, Korea Institute of Ocean Science and Technology, Geoje-shi 656-834, Republic of Korea b c

a r t i c l e

i n f o

Article history: Available online 20 November 2014 Keywords: Circular statistics Feeding guild Omnivory The Herbei Spirit oil spill Trophic plasticity Stable isotope

a b s t r a c t We examined trophic relationships in a newly established community 1.5 years after the Hebei Spirit oil spill on the west coast of Korea. Carbon and nitrogen stable isotope ratios in consumers and their potential food sources were compared between the oil-spill site and reference site, located 13.5 km from the oil-spill spot. The isotopic mixing model and a novel circular statistics rejected the influx of petrogenic carbon into the community and identified spatial consistencies such as the high contributions of microphytobenthos, food-chain length, and the isotopic niche of each feeding guild between sites. We suggested that high level of trophic plasticity and the prevalence of omnivory of consumers may promote the robustness of food web against the oil contamination. Furthermore, we highlighted the need of holistic approaches including different functional groups to quantify changes in the food web structure and assess the influence of different perturbations including oil spill. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Coastal ecosystem provides several services such as erosion control, recreation, habitat and refugia for a diversity of life (Costanza et al., 1997). Coastal areas receive a variety of anthropogenic activities due to the increasing human population toward these areas (Barbier et al., 2008; Barbier, 2011). For example, toxic chemicals, radioactive disposals, sewage discharges, and garbage from human habitat near coastal zones have threatened biodiversity of coastal life (Kennish, 2002). Particularly, oil pollution by wrecked tanker or natural seepage become more problematic as industrial demand for petroleum increased (Peterson et al., 2003). Oil spill can affect the survival of aquatic organism through the direct contact causing death or incorporation of sub-lethal amount that may reduce resistance to infection (Gin et al., 2001). In this context, it is increasingly critical to assess the ecosystem response to different perturbations including oil spill in order to predict future changes in the ecosystem structure and suggest management and conservation policies that promote the ecosystem sustainability. ⇑ Corresponding author. Tel.: +82 62 715 2834; fax: +82 62 715 2343. E-mail address: [email protected] (C.-K. Kang). http://dx.doi.org/10.1016/j.marpolbul.2014.10.054 0025-326X/Ó 2014 Elsevier Ltd. All rights reserved.

Stable isotope (SI) ratios have been used to examine trophic relationships in different aquatic food webs because they provide time-integrated information on consumers’ diet (Fry, 2006). The tissues of aquatic fauna including macrobenthos and fish species usually have turnover periods that range from weeks to months, and thus SI measurements can compensate for the deficiencies of long-term time-series datasets when assessing an ecosystem’s status (Buchheister and Latour, 2010). The stable carbon isotope ratio (d13C) in the tissues of a consumer reflects that of its prey with a slight modification (61‰), thereby indicating the dietary composition of consumers (McCutchan et al., 2003). The stable nitrogen isotope ratio (d15N) in the tissues of a consumer generally increases by 3–4‰ compared with that of its prey, which can delineate the trophic position (TP) of the consumer (Post, 2002). In particular, SI ratios of petroleum have a small carbon fractionation factor (0.5‰) even through evaporation, microbial decomposition, or physical weathering process (Macko et al., 1981; Chapelle, 2001; Jeffrey, 2006). Furthermore, previous works showed that the isotopic signatures of spilled petroleum persist with small fractionation even 2 years after an oil spill accident (Macko et al., 1981). This fact together with the well-known reference isotopic values of fresh oil (approximately 30‰ to 20‰) allowed us to use SI ratios in

Taean peninsula 100 km

37˚00´N

Korea

Dec. 9 PM 4

Dec. 7 AM 10

36˚50´

126˚E

Dec. 8 AM 10

●

Mallipo (Oil-spill site)

Dec. 8 PM 6

Dec. 9 PM 4

●

Mongsanpo (Reference site)

Dec. 9 PM 8

36˚40´

oil-contaminated regions for evaluating the incorporation of petroleum-derived carbon by the aquatic biota, as observed in other coastal ecosystems (e.g. Fuex, 1977; Hartman and Hammond, 1981; Mazeas and Budzinski, 2002a,b; Griebler et al., 2004; Reddy et al., 2012). Previous works using SIs showed the small incorporation of oil by pelagic consumers such as herbivorous fish Pacific herring, walleye pollock, filter feeding barnacles and mussels can occur in oil-contaminated regions (Kline and Thomas, 1999; Fry and Anderson, 2014). Despite this information, we still have much to learn about the consequences of an oil spill accident on coastal benthic food web including all trophic levels (Gin et al., 2001; Peterson et al., 2003). In the present study, we evaluated the SI values of the resident biota and diverse sources of organic matter to examine the consequences of oil contamination on an intertidal food web structure in the west coast of Korea 1.5 years after the Hebei Spirit oil spill (HSOS). Previous studies identified the effects of the HSOS in various ecosystem components separately including algal, invertebrate and fish species from individuals to community level (Ji et al., 2011; Lee et al., 2011). For example, at the community level the abundance of phytoplankton within two weeks after the HSOS was only half of the 10-year winter average, whereas the magnitude of the following spring blooms was three times greater than previous years (Lee et al., 2009). Massive accumulation of PAHs and the induction of hepatic PAH catabolism was observed in both demersal and pelagic fish at heavily oiled sites (Jung et al., 2011). Furthermore, the density of the dominant species of macrobenthic communities significantly decreased in the oil-affected sandy tidal flats (Yu et al., 2013). In addition, new macrobenthic and fish communities established immediately after beach clean-up efforts in 2008 (MLTM, 2009; Kim et al., 2013). However, the effects of the HSOS have not been assessed in the fauna using a more holistic approach and including different trophic levels. In this work we examined: (1) if petrogenic carbon was incorporated into the coastal food web, and (2) if the trophic niches of the consumers in the rebuilt community were similar to those in the unaffected community. Given that previous works showed that feeding habits may change after the oil spill accident, we tested the hypothesis that isotopic niche of benthic feeding groups at the impacted site changed owing to the effect of the oil contamination.

37˚N

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180

35˚

168

5 km

126˚10´

126˚20´

Fig. 1. Map of study site in the west coast of Korea showing the sampling sites for consumers and food sources: Mallipo and Mongsanpo in the Taean peninsula. Data for tracking the Hebei Spirit oil spill accident were obtained from MLTM (2009) and are shown in the map. Star represents the oil-spill spot by the wrecked Hebei Spirit oil tanker. The gray-colored area with date and time indicates the whole stretched area of the oil spill for a short period of time.

2. Materials and methods 2.1. Hebei Spirit oil spill On December 7, 2007, the oil tanker Hebei Spirit was wrecked off the Taean coast of South Korea (36°530 34.9700 N; 126°030 31.7000 E). The tanker released a total of 12,547 kL of crude oil, a large fraction of which washed along a 375 km stretch of shoreline (Kim et al., 2013). The magnitude of the HSOS was quite extensive, its total volume being roughly one-third of that of the Exxon Valdez oil spill (Peterson et al., 2003), and the overall extent of the affected coastline being close to one-fifth. However, as a consequence of immediate beach clean-up efforts, the petroleum hydrocarbon concentrations in the water column off the Taean coast decreased rapidly from an average of 732–31.0 lg L1 after one month, and the concentration on the affected beaches dropped below the environmental standard (10 lg L1) within 10 months after the accident (Kim et al., 2013). 2.2. Site description Mallipo beach (36°470 2900 N, 126°080 2700 E; Fig. 1) which was located 13.5 km from the oil-spill spot was selected as the oil-spill site for this study because it was heavily affected by the HSOS (Kim

et al., 2010). The oil concentration in the water column at the oilspill site ranged from 8.84 to 2700 lg L1 immediately after the HSOS. The concentration dropped abruptly to 9.16 lg L1 in the following month and fluctuated between 1.38 and 69.0 lg L1 until June 2008. Since July 2008, the oils spill concentration was approximately 1.0 lg L1 in the aquatic environment. By contrast, the oil concentration at Monsanpo beach (36°400 1400 N, 126°170 1200 E), which was selected as a reference site and located 32 km from the oil-spill spot, ranged from 0.34 to 2.27 lg L1, which is less than the values that are generally observed in natural coastal conditions (10 lg L1). When the HSOS occurred, northwesterly or northeasterly winds prevailed and the tidal currents were reversing (Lee et al., 2009). These physical forces spread the spilled oil north-eastwards or south-westwards. However, natural environment of two study sites (i.e., the oil-spill and reference sites) were similar as shown in their nearly identical values in salinity (30.9– 31.8 vs. 30.7–31.6) and temperature (6.1–22.2 vs. 6.2–22.3 °C) of surface water, texture of sediment (median grain size = around phi 2 at both sites), and morphodynamics of beach during the sampling period (MLTM, 2009; Yu et al., 2013). The species richness (S), diversity (H0 ), and evenness (J0 ) of macrofauna and fish species were estimated in Taean beach 1.5 years after the HSOS (MLTM, 2009; Yu et al., 2013). The invertebrate

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180

and fish communities at the oil-spill site exhibited low degrees of species richness, diversity, and evenness (S = 40 species, H0 = 1.13, and J0 = 0.31 for invertebrates; 31, 2.22, and 0.65 for fish) compared with those at the reference site (S = 83 species, H0 = 3.05, and J0 = 0.69 for invertebrates; 38 species, 2.65, and 0.73 for fish). Interestingly, the invertebrate and fish abundances were higher at the oil-spill site (18,509 and 2096 individuals, respectively) than those at the reference site (15,115 and 1104, respectively). The biomass of invertebrates, as the bulk wet weight (WWt), was higher at the oil-spill site (505 WWt m2) than that at the reference site (340 WWt m2), while the fish biomass at the oil-spill site (17,760 WWt m2) was similar to that at the reference site (17,961 WWt m2). Phytoplanktonic and microphytobenthic biomasses fluctuated greatly with season at both sites. Phytoplankton were dominated by Coscinodicus sp., Melosira sp., Actinoptychus sp., Nitzschia sp, Prorocentrum sp., and Pleurosigma sp. (MLTM, 2009). Chlorophyll a concentrations in water column varied from 1.0– 1.1 and 0.9–1.2 lg L1 in February to 4.8–6.2 and 3.2–6.9 lg L1 in August at the oil-spill and reference sites, respectively. Microphytobenthos were mainly composed of benthic diatoms such as Amphora sp., Cocconeis sp., Navicula sp., Paralia sp., and Pleurosigma sp. (this study). Chlorophyll a concentrations in sediments (61 m depth) were estimated from 3.5–9.1 and 3.9–12.7 mg m2 in January to 17.8–42.6 and 16.3–38.6 mg m2 in August at the oil-spill and reference sites, respectively. 2.3. Sample collection and processing All samples (fauna and basal sources) were collected in different replicates (one to seventeen) according to their abundance within the coastal area during each season (Table 1). The d13C and d15N levels of microphytobenthos (MPB) and suspended particulate organic matter (SPOM) were determined to define the trophic base that supported the coastal food web. Samples were monthly collected between July 2009 and June 2010 from both the oil-spill and reference sites (Fig. 1). MPB samples were collected by gently scraping the upper 1 mm of the sediment surface where visible mats of benthic diatoms were found at each intertidal flat site. A pure sample of benthic diatoms was extracted for isotope analysis using a procedure that was modified slightly from that of Couch (1989). Sediment containing MPB was spread to a depth of 2 cm in plastic trays. A nylon screen with a mesh of 63 lm was placed on the sediment surface and covered with a 2 mm layer of silica granules. Fluorescent lamps were used to illuminate the trays, which were kept wet with in situ filtered seawater. The motile benthic microalgae migrated onto the silica granule layer through the nylon screen. Next, the silica granules were collected onto 63 lm mesh screens and rinsed with distilled water. The MPB collected in glass tubes was centrifuged at 3293g for 15 min, and the concentrated MPB in the precipitates was lyophilized before isotopic analysis. To collect SPOM, a known volume (10 L) of seawater was pumped from a depth of about 1 m below the water surface using a van Dorn water sampler from each intertidal flat site during slack water at the high tide. The water samples were immediately passed through a 180 lm Nitex mesh to remove zooplankton and large particles, and were collected in acid-washed plastic bottles. The water was then filtered again through precombusted (450 °C for 4 h) Whatman 47 mm GF/F filters (0.70 lm nominal pore size). The GF/F filters were oven-dried at 60 °C for 48 h before isotopic analysis. Macrobenthic invertebrates were monthly collected in June– August 2009, January–February 2010, and March–May 2010 from both the oil-spill and reference sites. They were sampled by hand using a rectangular aluminum corer with a total surface area of 0.25 m2 at a sampling depth of 0.2 m during low tide, and the

169

sediments were sieved gently with a 1-mm mesh net. Only live and intact organisms were collected. The specimens obtained were washed with filtered seawater and transported immediately to the laboratory. The sorted invertebrates were identified to the species level if possible and kept alive overnight in filtered seawater from the sampling site to evacuate their gut contents. Bivalves and large gastropods were dissected carefully, and only muscle tissues were used for analysis. Whole tissues of polychaetes were used for analysis. All of the treated specimens were kept frozen (20 °C) until subsequent treatment. Fish were collected during June–August 2009 and 2010 from both sites. To collect fish, a 3 m high and 30 m long three-layer gill net was anchored overnight on the bottom sediment at the lower part of the intertidal flat of each site, and fish were collected in the early morning. After transportation to the laboratory, white muscle tissue was separated from the dorsal regions of the fish. Fish samples were rinsed with distilled water and kept frozen (20 °C). 2.4. SI analysis The SPOM samples were immediately acid-soaked with several drops of 1 N HCl, rinsed with distilled water, and kept frozen (20 °C) until later treatment. The samples used for d15N analysis were not acid treated. Small individual crustaceans and gastropods were treated with 10% HCl solution until the bubbling stopped and then rinsed with distilled water to remove the carbonates. To avoid the isotopic variation caused by differences in the contents of isotopically lighter lipids among species (Focken and Becker, 1998), lipids were extracted from the muscle tissues using a mixed solution of methanol, chloroform, and water (2:1:0.8) (Bligh and Dyer, 1959). Animal tissue samples were lyophilized, pulverized to a fine powder with a ball mill (MM200; Retsch GmBH, Haan, Germany), stored in glass vials, and kept in vacuum desiccators until the isotopic analysis. For the d13C and d15N values of animal, and microalgae, 0.5– 1.5 mg of powdered and homogenized samples were transferred to tin capsules, and the GF/F filters were wrapped with tin plate. Samples were combusted in an automated elemental analyzer (Vario MICRO Cube; Elementar Analysensysteme GmbH, Hanau, Germany) and oxidized at high temperature (1030 °C). The resulting CO2 and N2 gases were then analyzed using an interfaced continuous-flow isotope-ratio mass spectrometer (CF-IRMS, IsoPrime 100; IsoPrime Ltd, Cheadle Hulme, UK). The SI abundances were expressed using the delta notation (d) as standard deviations (SDs) in parts per thousand (‰) relative to Pee Dee Belemnite for carbon and atmospheric N2 standards for nitrogen. The ratios were derived using the equation: dX (‰) = [(Rsample/Rstandard)  1]  103, where X is 13C or 15N, and R is the corresponding ratio of 13C/12C or 15 N/14N. International Atomic Energy Agency (IAEA) CH-6 (sucrose, 13 d C = 10.45 ± 0.07‰; National Institute of Standards and Technology [NIST], Gaithersburg, MD, USA) and IAEA-N1 (ammonium sulfate, d15N = 0.4 ± 0.2‰; NIST) having known relationships with the international standard were used as reference materials and analyzed with every 10th unknown sample. The SDs of the means for repeated measurements of internal peptone and urea standards were 0.15‰ for d13C and 0.20‰ for d15N. SI values of a species were measured for each individual and the values were averaged. 2.5. Bayesian mixing model and the trophic positions of fish The relative contributions of the two end members of the trophic base (benthic and pelagic microalgae) to the diets of consumers were estimated using SI analysis in R (SIAR, Parnell et al., 2010). We used averaged trophic fractionation for aquatic consumers in McCutchan et al. (2003) and assumed to be 0.4 ± 0.1‰ for 13C

170

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180

Table 1 Carbon (d13C) and nitrogen (d15N) stable isotope signature (‰) of macrobenthos collected at the oil-spill (OS) and reference (RS) site. B stands for bivalvia; C, crustacea; E, echinodermata; G, gastropoda; P, polychaeta. Data represent mean ± 1SD. Numbers in the parentheses indicate the number of individuals for isotopic measurement. The values without parenthesized number mean that only one individual was collected and analyzed. Species

Taxon

OS

RS

d13C

d15N

17.0 18.4 ± 0.9 (12) 17.7

6.3 7.1 8.8

19.1 ± 0.7 (8)

8.1

18.3 ± 0.4 (7)

8.2

17.5 ± 0.7 (2) 16.3 ± 1.2 (4)

7.3 5.9 ± 2.2 (6)

2009 Summer Suspension feeders Balanus albicostatus (Acorn barnacle) Crassostrea gigas (Pacific oyster) Diogenens sp. Mactra veneriformis (Surf clam) Mytilus galloprovincialis (Mediterranean mussel) Phacosoma japonica (Japanese dosinia) Ruditapes phillipinarum (Manila clam) Solen strictus (Razor clam) Theora fragilis (Semele) Umbonium costatum (Sand snail)

C B C B B B B B B G

Deposit feeders Ampharetidae unid. (Bristle worm) Capitella sp. Ceratonereis erythraensis Cirratulus cirratus Cirriformia tentaculata Lagis bocki (Bock’s pectinated-worm) Pagurus sp. (Hermit crab) Protankyra bidentata (Sea cucumber) Thelepus setosus (Spaghetti worm)

P P P P P P C E P

17.4

6.7 ± 2.2 (17)

14.0 18.0 ± 0.4 (2) 14.2 17.5 14.7 17.2

10.3 8.7 9.5 9.8 10.2 ± 1.7 (2) 9.6

Grazers Acmaeidae unid. (Snowy limpet) Littorina brevicula (Asian periwinkle) Lottiidae unid. (True limpets) Lunella coronata coreensis (Coronate moon turban) Macrophthalmus dilatatusus (Granulate-hand ghost crab)

G G G G C

13.7 ± 0.7 (2) 13.4 14.2

7.5 10.0 7.8

Omnivores Asterina pectinifera (Bat sea star) Diopatra sugokai (Sugokai tube worm) Hemigrapsus penicillatus (Harry-clawed shore crab) Macoma sp. Nereis pelagica Perinereis nuntia Philomycidae unid.

E P C B P P G

16.8

12.3

14.5 ± 1.5 (3)

8.6

16.8 15.3 17.0

7.6 8.0 ± 2.1 (3) 11.0

Carnivores Asterias amurensis (Flat-bottom sea star) Charybdis japonica (Japanese swimming crab) Glossaulax didyma (Real bladder moon snail) Glycera chirori Lumbrineris sp. Nodilittorina (Tectarius) granularis (Granulated periwinkle) Philyra pisum (Pea pebble crab) Phyollodoce (Anaitides) sp. Portunus trituberculatus (Swimming crab) Rapana venosa (Asian rapa whelk)

E C G P P G C P C G

16.4

10.2

16.8 17.9 ± 1.6 (2) 12.2

13.9 8.7 ± 0.3 (3) 8.8

19.2 ± 1.3 (2)

10.9 ± 2.5 (2)

14.8

14.0

2009/2010 Winter Suspension feeders Crassostrea gigas (Pacific oyster) Mytilus galloprovincialis (Mediterranean mussel) Ruditapes phillipinarum (Manila clam) Solen strictus (Razor clam) Umbonium costatum (Sand snail)

B B B B G

17.9 ± 0.3 (10) 15.6 ± 0.4 (2) 17.6 ± 0.9 (9)

7.0 7.0 ± 0.8 (4) 8.4

G P C P

12.4 15.3 13.5

9.8 6.1 9.7 ± 2.0 (2)

G G

15.4

5.4

C P P

14.5

10.0

13.9

8.4

Deposit feeders Batillaria cumingii (Cuming’s false cerith) Cirriformia tentaculata Pagurus sp. (Hermit crab) Thelepus setosus (Spaghetti worm) Grazers Littorina brevicula (Asian periwinkle) Lunella coronata coreensis (Coronate moon turban) Omnivores Hemigrapsus penicillatus (Harry-clawed shore crab) Nereis pelagica Perinereis nuntia

d13C

d15N

18.0 ± 0.6 (5) 18.0 19.6

8.5 7.5 8.2

16.8 17.6 ± 0.6 (5) 18.3 ± 0.1 (4) 18.2 ± 0.3 (2)

7.6 ± 1.4 (6) 7.9 9.1 7.4

17.5 14.9 ± 0.2 (2)

10.3 11.5

17.6 ± 2.3 (2) 15.8 ± 2.0 (4) 16.7

10.6 10.3 8.7 ± 1.3 (4)

13.3 12.9 ± 0.9 (2)

9.9 8.4 ± 0.7 (2)

16.3 ± 1.0 (3) 16.8 15.0 16.0 15.0

12.8 11.6 12.4 7.7 ± 1.1 (14) 10.1

16.4

10.3

16.9 16.6 ± 0.4 (3) 16.2 ± 0.6 (7)

9.6 13.5 7.3

15.2

10.0

16.5 15.5

8.9 11.7

19.3 ± 0.8 (8)

4.5

19.2 ± 0.6 (8) 16.2 16.6

5.5 6.8 7.7

15.4 ± 1.0 (2) 14.9

5.5 ± 1.5 (5) 7.6 ± 2.8 (2)

16.0

12.1

14.8 16.8 14.7

6.7 ± 2.8 (5) 6.1 ± 2.5 (2) 10.6

171

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180 Table 1 (continued) Species

Taxon

OS

RS

d13C

d15N

d13C

d15N

Carnivores Crangon affinis (Japanese sand shrimp) Lumbrineris sp. Thais clavigera (Rock snail)

C P G

13.0 16.2 15.2

10.7 11.7 ± 2.4 (2) 7.5

2010 Spring Suspension feeders Balanidae unid. (Barnacles) Columbellidae unid. (Dove shell) Crassostrea gigas (Pacific oyster) Diogenens sp. Glauconome chinensis (Chinese glauconome) Musculista senhousia (Asian date mussel) Mytilus galloprovincialis (Mediterranean mussel) Ruditapes phillipinarum (Manila clam)

C G B C B B B B

17.2 18.1 19.4 ± 0.9 (27)

10.8 9.8 8.6

19.5 19.1 ± 0.3 (3) 18.5 ± 0.5 (27)

8.7 6.4 7.4

Deposit feeders Batillaria cumingii (Cuming’s false cerith) Bullacta exarata (Korean mud snail) Ceratonereis erythraensis Cirratulus cirratus (Bristle worm) Cirriformia tentaculata Lagis bocki (Bock’s pectinated-worm) Pagurus sp. (Hermit crab) Protankyra bidentata (Sea cucumber) Terebellidae unid.

G G P P P P C E P

15.5

10.6

14.2 ± 2.1 (2) 18.1 16.6 16.4 ± 0.7 (5) 15.7 ± 1.4 (2) 18.1

8.3 10.6 7.4 ± 0.8 (4) 8.3 ± 2.2 (7) 8.9 10.1

Grazers Acmaeidae unid. (Snowy limpet) Littorina brevicula (Asian periwinkle) Lunella coronata coreensis (Coronate moon turban) Monodonta labio confusa (Lipped periwinkle) Nerita (Heminerita) japonica (Japanese nerite) Nipponacmaea schrenckii Patella sp.

G G G G G G G

15.3 ± 0.9 (2)

7.3

17.3

6.5

15.0

9.6 ± 0.6 (4)

Omnivores Asterina pectinifera (Bat sea star) Hemigrapsus penicillatus (Harry-clawed shore crab) Nereis pelagica Nereis sp. Perinereis cultrifera (Rag worm) Perinereis nuntia

E C P P P P

17.3 16.4 15.9 ± 1.7 (4)

10.1 ± 2.0 (2) 13.0 9.1 ± 0.9 (2)

16.6

12.3

Carnivores Asterias amurensis (Flat-bottom sea star) Crangon affinis (Japanese sand shrimp) Glossaulax didyma (Real bladder moon snail) Glycera chirori Glyceridae unid. Halosydna sp. (Polynoid worm) Lumbrineris sp. Nassarius (Zeuxis) siquijorensis (Plaited nassa) Portunus trituberculatus (Swimming crab) Thais clavigera (Rock snail)

E C G P P P P G C G

15.3 18.0 15.7 16.7 16.9 16.3 14.8 15.9 ± 1.3 (2) 16.7

13.0 12.4 7.4 ± 2.4 (9) 9.2 ± 2.0 (6) 12.9 11.3 ± 2.5 (3) 13.7 9.8 10.4 ± 2.3 (2)

(D13C) and 2.3 ± 0.2‰ for 15N (D15N). The numbers of iterations run for the model and to discard were defined as 5,000,000 and 50,000, respectively. We employed averaged isotopic values of MPB and phytoplankton for 12 months while not pooled the values of consumers between seasons. We did not consider grazers when conducting mixing model calculations because they were aggregated on, and primarily consumed, macroalgae, and thus they appeared to be quantitatively less important in the diets of other macrobenthic feeding groups than particulate organic matter (i.e., MPB and SPOM) at the study site. The TP of each fish was calculated as follows: TPfish = (d15Nfish  d15Nbase)/Dd15N + 1. There was no significant difference in the d15N values of MPB and SPOM, and thus d15Nbase is the mean value (6.1‰) of the two end members of the primary consumers.

18.2 ± 0.6 (15) 16.3 ± 0.6 (2) 15.1 ± 0.6 (2)

6.9 7.3 ± 0.7 (2) 10.2 ± 0.3 (2)

17.7 ± 1.1 (3) 17.3 ± 0.3 (12)

6.1 9.7 ± 1.9 (2)

13.2 ± 0.6 (2)

11.4 ± 0.6 (2)

14.0 ± 0.6 (2)

11.5 ± 1.2 (4)

14.1 ± 0.6 (4)

11.2 ± 0.5 (4)

12.1 13.8 ± 0.5 14.2 ± 0.6 13.6 ± 0.6 13.2 ± 0.6

8.8 8.7 11.7 ± 1.6 (4) 6.9 ± 1.8 (5) 4.7

(3) (2) (2) (2)

9.6 ± 0.1 (2)

14.5

15.1 ± 2.0 15.6 ± 0.8 17.4 ± 1.3 15.7 ± 0.7 16.7 ± 0.6

(4) (3) (2) (3) (3)

18.6

12.1 9.8 10.8 ± 1.2 (3) 12.0 11.0 11.0 ± 1.7 (2)

16.1 ± 2.0 (2)

9.3 ± 0.9 (10)

17.5

9.0 ± 5.1 (2)

2.6. Linear statistics Linear statistical analysis was performed using SPSS (version 12.0.1; IBM SPSS Statistics, Armonk, NY). The Shapiro–Wilk test for normality and Levene’s test for homogeneity of variance were applied to all the isotope data without prior transformation. A one sample t-test was used to confirm differences between the isotopic values of the organic matter sources, macrobenthic invertebrates, and fish, as well as the petroleum levels. In our case we considered the d13C values of oils spill gathered from published data including a review paper with data from severals regions (Fuex, 1977) and also experimental study that measured d13C values of bulk crude oil or PAHs such as methylphenanthrenes (Mazeas and Budzinski, 2002a,b; Reddy et al., 2012). These d13C values recorded a narrow range: 26.3‰ to 28.9‰ and therefore

172

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180

represent well constrained information on isotopic values of crude oil. For statistical analysis, we set reference d13C and d15N value for crude oil as 28.5‰ and 1.8‰, respectively (Rumolo et al., 2011). Paired t-tests were used to compare: (1) the isotopic values for MPB and those for SPOM during each month, and (2) the values of the commonly collected macrobenthos or fish from the oil-spill and reference site. Two sample t-tests were performed to analyze the spatial differences in: (1) the mean isotopic values of macrobenthos, or (2) the mean trophic positions of fish. We tested whether the macrobenthos feeding guilds except grazers, which had much higher d13C ranges than the others, could be discriminated from one another based on their isotopic values using the Kruskal–Wallis test because the feeding guilds were not sufficiently large to conduct parametric tests. If there was a significant difference in the median values, a Mann–Whitney U test was performed as a post hoc test. The Kolmogorov–Smirnov test was used to compare the median isotopic values of each feeding guild at the oil-spill site with those at the reference site. 2.7. Circular statistics We performed a circular statistical analysis using the dual SI data according to a previously reported method (Schmidt et al., 2007, 2009) to determine the directionality of the trophic niche shift in the SI dual plot. The trophic niche shift of a trophospecies or a feeding guild comprises two components: magnitude (r) and direction (h, where 0 6 h < 360°). An arrow with a length of r and an angle of h represents a species or a feeding guild as a Euclidean vector from its point in the d15N vs. d13C dual plot for the oil-spill site relative to another point in the plot for the reference site. The magnitude r was estimated as follows: 2

2 1=2

r ¼ ½ðd13 COS  d13 CRS Þ þ ðd15 NOS  d15 NRS Þ 

:

The direction h was calculated as:

h ¼ arccosðDd15 N=rÞ for d13 COS  d13 CRS P 0; or h ¼ 360  arccosðDd15 N=rÞ for d13 COS  d13 CRS < 0: However, h had to be corrected according to the Dd13C and Dd15N values because higher-trophic-level consumers had more positive d13C values than lower-trophic-level consumers because a similar reliance on a specific food chain emerged from a trophic fractionation effect. Therefore, we corrected the h as follows:

hcorr ¼ h  arctanðDd13 C=Dd15 NÞ: The directional movement is represented as an arrow with a length r and an angle h. The uniformity in the direction of trophic niche changes for all trophospecies within a feeding guild can be tested using circular statistics. The test of uniformity is used to check whether there is a mean direction; i.e., to confirm whether all the arrows point in the same specific direction. The smallest sample size in of our dataset was five, and thus the results of the circular statistics were significant to two decimal places (Zar, 1999). The uncorrected values for the consumers were used in the circular statistical analysis because there were no differences in the d13C and d15N values of the organic matter sources. The feeding guilds of the consumers analyzed in the present study were determined according to previous reports (Fauchald, 1979; Kent and Day, 1983; Hong et al., 2006). To analyze whether the trophic niche changes between sites at the species and feeding guild level were oriented in a specific direction, Rayleigh’s test for circular uniformity was performed, and arrow diagrams were constructed using the ‘circular’ library in the R computing environment (version 2.15.2, www.r-project.org) and Sigma Plot 10.0 (Systat Software, San Jose, CA). The significance level was set to p < 0.05 for all statistical tests.

Table 2 Carbon (d13C) and nitrogen (d15N) stable isotope signature (‰) of fish collected at the oil-spill (OS) and reference site (RS) in summer 2009 and 2010. Data represent mean ± 1SD. Numbers in the parentheses indicate the number of individuals for isotopic measurement. The values without parenthesized number mean that only one individual was collected and analyzed. Species

2009 Summer Hexagrammos agrammus (Spotty-belly greenling) Hexagrammos otakii (Greenling) Limanda yokohamae (Marbled flounder) Muraenesox cinereus (Pike conger) Paralichthys olivaceus (Olive flounder) Repomucenus beniteguri Repomucenus koreanus (Stink fish) Sebastes schlegelii (Korean rockfish) Verasper variegatus (Spotted halibut) 2010 Summer Acanthogobius hasta (Javeline goby) Acanthopagrus schlegeli (Black porgy) Chelon haematocheilus (So-iny mullet) Cynoglossus joyneri (Red tonguesole) Hexagrammos otakii (Greenling) Konosirus punctatus (Dotted gizzard shad) Lateolabrax japonicus (Sea bass) Limanda yokohamae (Marbled flounder) Nibea albiflora Paralichthys olivaceus (Olive flounder) Paraplagusia japonica (Japanese spanish mackerel) Repomucenus beniteguri Repomucenus koreanus (Stink fish) Salangichthys microdon (Japanese ice fish) Scomberomorus niphonius Sebastes schlegelii (Korean rock fish) Takifugu niphobles (Grass puffer) Verasper variegatus (Spotted halibut) Zebrias fasciatus (Many-banded sole)

OS

RS

d13C

d15N

d13C

d15N

15.6 16.8 16 16.7 ± 1.5 (2) 15.9 15.7

14.0 13.6 13.2 15.7 ± 1.1 (2) 13.2 13.5

16.9 16.4 17.2 ± 0.0 (2) 15.2 ± 1.0 (2)

12.9 12.7 13.5 15.0

15.2

13.7

15.3

13.7 16.7

12.3

15.9 18.8 ± 1.5 (2) 15.3

12.1 11 ± 1.7 (2) 12.9

16.1 ± 1.1 (2)

12.4 ± 0.8 (2)

14.8 15.6 ± 0.2 (2) 15.1 ± 0.4 (2)

12.9 13.5 ± 0.0 (2) 12.6 ± 1.2 (2)

16.9 ± 2.0 (2) 15.1 ± 0.6 (2)

12.8 ± 2.4 (2) 13.1 ± 0.6 (2)

15.6

11.8

13.2 ± 0.3 14.9 ± 0.2 16.9 ± 1.7 20.2 ± 0.3 16.6 ± 0.2 18.2 ± 0.1 15.9 16.5 ± 0.4

(3) (2) (5) (2) (3) (2) (4)

12.5 ± 0.3 12.4 ± 0.9 11.8 ± 1.4 12.4 ± 0.1 11.4 11.2 12.8 11.5 ± 1.1

(3) (2) (5) (2)

(4)

15.9 ± 0.7 (5)

12.5 ± 0.6 (5)

15.7 ± 0.6 15.4 ± 0.4 19.5 ± 0.2 18.3 16.8 ± 1.4 15.6 ± 0.2 16.6 ± 0.2 16.4 ± 0.7

12.4 ± 1.0 13.2 10.8 ± 0.4 13.1 12.1 ± 1.1 13.9 ± 0.9 12.3 ± 0.1 11.1 ± 0.0

(6) (3) (2) (6) (3) (2) (2)

(4) (2) (6) (3) (2) (2)

173

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180

-14

much higher than the known values for crude oil (one sample t-test, t19, 0.05 = 95.47 and 31.012, p < 0.001 for d13C; t19, 0.05 = 13.97 and 17.53, p < 0.001 for d15N).

A

-16

δ13C (‰)

-18

3.2. Macrobenthic invertebrates

-20 -22

The SI values of the macrobenthos were determined for 28, 9, and 22 species from the intertidal zone of the oil-spill site during summer, winter, and spring, which were compared with those for 26, 13, and 22 species collected simultaneously from the reference site, respectively (Table 1). First, we compared the SI values of the consumers collected from both sites with the values of crude oil to confirm whether petrogenic carbon was assimilated directly into the tissue of the consumers at the oil-spill site. The mean isotopic values for the macrobenthos from each site and season were much higher than the previously reported value for crude oil (one sample t-test, p < 0.001 for both sites and all three seasons for d13C and d15N; mean differences: >10‰ for d13C and >6‰ for d15N). Next, we performed between-site comparisons of the d13C and 15 d N values for the consumers from three seasons at the whole community (for both entire species and common species) and trophic guild levels to determine whether the values at the oil-spill site remained consistently high (i.e., similar to the values for the consumers at the reference site) or whether they shifted toward the values for crude oil. First, the mean d13C and d15N values of the macrobenthos at the oil-spill site were not significantly different from those at the reference site (two sample t-test, p > 0.100 for 11 cases), except for the d13C values in spring (t49 = 2.91, p = 0.003, mean difference = 1.3‰). The macrobenthic consumers collected at the oil-spill and reference sites during three seasons had similar d13C ranges of 19.2‰ to 12.1‰ and 19.6‰ to 12.9‰, respectively, and their d15N values also had almost identical ranges of 4.7–14.0‰ and 4.5–13.7‰, respectively (Fig. 3). Common species, which were collected at both sites, tended to have slightly higher d13C values at the oil-spill site than those at the reference site during winter (paired t-test, t4 = 4.17, p = 0.014, mean difference = 1.2‰) and spring (t10 = 2.52, p = 0.031, mean difference = 1.0‰) but not during summer (t12 = 0.33, p = 0.7441, Fig. 4). The 13C-enrichment was most notable in the tissues of Pagurus sp. (Hermit crab, 2.3‰) and Monodonta labio confusa (Lipped periwinkle, 3.7‰) in spring. In addition, there were no differences in the d15N values among all seasons (t12 = 1.09, p = 0.298 for summer; t4 = 1.91, p = 0.129 for winter; t10 = 0.13, p = 0.899 for spring). Second, the results of the circular statistical analysis of common species and major feeding guilds showed that the isotopic differences between consumers from the oil-spill and reference sites were not oriented in specific directions at both the species and

-24 Range for petroleum and its component (-26.3 to -28.9 ‰)

-26 Crude oil (-28.5‰)

-28 -30 10

B

δ15N (‰)

8 6 4 2

MPB at OS

SPOM at OS

MPB at RS

SPOM at RS

0

J

J

A

S

O N D J

2009

F

M A M

2010 Month

Fig. 2. Stable isotope ratios of monthly collected microphytobenthos (square) and suspended particulate organic matter (SPOM, circle) from the oil-spill site (solid) and the reference site (open). (A) d13C values; (B) d15N values. Four plots of the right side indicate annual means (n = 10) and ±1SD (vertical bars). The d13C values of fresh oil spill obtained from literature survey (Fuex, 1977; Mazeas and Budzinski, 2002a,b; Reddy et al., 2012) were not shown. Source of the values was bulk crude oil or PAHs such as methylphenanthrenes. Reference d13C value for crude oil was 28.5‰ (Rumolo et al., 2011).

3. Results 3.1. Primary producers There were no spatial differences in the d13C and d15N values for MPB (paired t-test, t9, 0.05 = 0.746 and 0.375; p = 0.475 and 0.716, respectively) and SPOM (t9, 0.05 = 1.462 and 0.744; p = 0.178 and 0.476, respectively; Fig. 2). The d13C value was 5.0‰ higher for MPB than that for SPOM (t19, 0.05 = 16.24; p < 0.001), whereas there was no difference between their d15N values (t19, 0.05 = 1.05; p = 0.307). The pooled d13C and d15N values for MPB (mean ± 1SD, 15.7 ± 0.6‰ and 6.3 ± 1.2‰, respectively) and phytoplankton (20.7 ± 1.1‰ and 6.0 ± 1.3‰, respectively) were 16

A Summer

B Winter

14

C Spring

δ15N (‰)

12 10 8 OS RS Suspension feeders Deposit feeders Grazers Omnivores Carnivores

6 4 2 -22

-20

-18

-16

-14

δ13C (‰)

-12

-10 -22

-20

-18

-16

-14

δ13C (‰)

-12

-10 -22

-20

-18

-16

-14

-12

-10

δ13C (‰)

Fig. 3. d15N vs. d13C dual plot of macrobenthic invertebrates at the oil-spill site (solid; OS) and at the reference site (open; RS). Five major feeding guilds, suspension feeders (circle), deposit feeders (reverse triangle), grazers (square), omnivores (diamond), and carnivores (triangle), were distinguished by different symbols. Error bar indicates ± 1SD (for n, see Table 1). (A) 2009 summer; (B) 2009/2010 winter; (C) 2010 spring.

174

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180 16

-13

A δ13C (‰)

14 13

Reference site

-15

Reference site

B δ15N (‰)

15

-14

-16 -17

12 11 10 9

-18

8 Summer Winter Spring

-19

7 6

-20

5 -20

-19

-18

-17

-16

-15

-14

-13

5

6

7

8

9

Oil-spill site

10 11 12 13 14 15 16

Oil-spill site

Fig. 4. 1:1 comparison on the isotopic values of common macrobenthic species between the reference site vs. the oil-spill site. Three sampling periods, summer (solid), winter (gray), and spring (open), were distinguished by different colors of symbol. The dotted lines indicate 1:1 diagonals. (A) Comparison on d13C values; (B) comparison on d15N values.

A Species, Summer

0° Higher TP

B Species, Winter

12 6

16 4

8 270° Pelagic

5

17

2

15

4.8 4.0

3

3.2

17

2.4

12 17

1.6 11

12 18 0.8 1.6 2.4 3.2 4.0 4.8 Magnitude of change

90° 270° Benthic Pelagic

O 0.8 1.6 2.4 3.2 4.0 4.8

90° 270° Benthic Pelagic

14 13

8

180° Lower TP

E Feeding guild, Winter

180° Lower TP

F Feeding guild, Spring

0° Higher TP

0° Higher TP D

D

270° Pelagic

S O Magnitude of change O 0.9 1.8 2.7 3.6

S G O

90° Benthic

6 3

6

0° Higher TP

C

9

10

O

13 1

7

0.8

180° Lower TP

D Feeding guild, Summer

0° Higher TP

O

3

11

C Species, Spring

0° Higher TP

O

90° 270° Benthic Pelagic

0.9

G 1.8

2.7

7.0

90° Benthic

270° Pelagic

O

O 0.9 1.8 S

D

2.7

3.6

90° Benthic

C G

180° Lower TP

180° Lower TP

180° Lower TP

Fig. 5. Arrow diagram for spatial isotopic differences of common species and feeding guilds from the values at the oil-spill site to that at the reference site. Concentric lines are reference line for change of magnitudes which was set based on d13C trophic fractionation factor. (A)–(C) Common species collected in summer, winter, and spring, respectively. The number indicates the code for species (see Table 3). (D)–(F) Five major feeding guilds collected in summer, winter, and spring, respectively. S stands for suspension feeder; D deposit feeder; O omnivore; B benthivore; P piscivore.

functional group levels, thereby indicating interspecific and interguild variations (Rayleigh’s test for circular uniformity, p > 0.001 for both levels in three seasons, Fig. 5). Moreover, the changes in directions varied among/within each of the high (carnivores), middle (omnivores), and low (suspension feeders, deposit feeders, and grazers) trophic levels of consumer in all three seasons. In addition, intraspecific variations in the spatial isotopic differences among the three seasons were found in four macrobenthic species, which were collected at both sites in all three seasons: Crassostrea gigas, Hemigrapsus penicillatus, Pagurus sp., and Ruditapes phillipinarum. However, the spatial isotopic differences in each species had different magnitudes or angles between any two seasons.

We confirmed that there were clear intraguild variations in the changing isotopic patterns of the consumers based on both sets of results. For example, the spatial niche changes in four suspensionfeeding species were orthogonal to one another in summer (Fig. 5A, Table 3). The d13C value of R. phillipinarum (283.3°) was higher at the oil-spill site than that at the reference site, whereas the value of Theora fragilis (88.3°) was lower at the oil-spill site. The d15N value of Diogenes sp. (3.1°) was higher at the oil-spill site compared with that at the reference site, whereas the value of C. gigas (186.1°) was lower at the oil-spill site. However, the vector value for the feeding guild level niche change (magnitude = 0.7, angle = 143.6°) of the trophic group (i.e., suspension feeders in

175

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180

Table 3 Vectors of spatial niche shift of macrobenthic invertebrates between OS and RS with magnitude of r and angle of h (°). The first character in the parenthesis for common species represents feeding guild of the species. S stands for suspension feeders; D, deposit feeders; G, grazers; O, omnivores; C, carnivores. The second character indicates taxon of the species. B denotes bivalvia; C, crustacea; E, echinodermata; G, gastropoda; P, polychaeta. Code

Common species Ampharetidae unid. (Bristle worm; D, P) Asterina pectinifera (Bat sea star; D, E) Crassostrea gigas (Pacific oyster; S, B) Diogenens sp. (S, C) Glycera chirori (C, P) Hemigrapsus penicillatus (Harry-clawed shore crab; O, C) Littorina brevicula (Asian periwinkle; G, G) Lumbrineris sp. (C, P) Monodonta labio confusa (Lipped periwinkle; G, G) Mytilus galloprovincialis (Mediterranean mussel; S, B) Nereis pelagica (O, P) Pagurus sp. (Hermit crab; D, C) Perinereis nuntia (O, P) Portunus trituberculatus (Swimming crab; C, C) Protankyra bidentata (Sea cucumber; D, E) Rapana venosa (Asian rapa whelk; C, G) Ruditapes phillipinarum (Manila clam; S, B) Theora fragilis (Semele; S, B)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Feeding guild Suspension feeders Deposit feeders Grazers Omnivores Carnivores

S D G O C

2009 Summer

Winter

r

h

3.6 0.7 1.5 1.3 0.4 3.8

168.5 215.1 186.1 3.1 323.6 162.6

2.2

299.6

3.1 2.9

205.9 88.0

1.6 2.4 0.8 0.7

233.6 7.1 283.3 88.3

0.7 1.1 1.1 1.4 0.9

143.6 148.3 211.3 182.2 350.1

2010 Spring

r

h

r

h

2.9

19.3

2.1

134.9

3.3

355.3

4.6 2.3

14.4 150.1

1.6 2.1 2.0 3.7 1.4 0.8 3.7 1.3 1.8

114.8 37.1 164.4 73.9 92.2 13.3 28.5 174.5 233.5

3.3

19.0

2.6

17.7

1.6 2.4 6.7 1.8

19.9 28.4 165.0 30.7

1.8 3.4 2.4 0.5 1.7

99.6 39.9 65.5 80.1 210.4

summer) was different from the net vector (magnitude = 0.1, angle = 316.4°) of the four species (Fig. 5D). These intraguild variations were also observed among seasons. The spatial isotopic changes between any two different seasons lacked significant identical directions or magnitudes for any functional groups. Finally, the results of the SIAR mixing model calculations confirmed that most of the consumers primarily utilized benthic carbon instead of pelagic sources of organic matter at both the oil-spill and reference sites (Fig. 6). The deposit feeders exhibited distinctly greater reliance on MPB than other functional groups (95% confidence intervals for the dietary contribution of MPB to each functional group relative to phytoplankton, 79–81% and 73– 74% at the oil-spill and reference site, respectively, in summer; 93–96% and 95–96% in winter; 95–96% and 76–77% in spring). Omnivores exhibited a similar dependence on benthic food sources compared with deposit feeders, but they also had an unusually wide confidence interval at the oil-spill site in winter (82–83% and 89–90% at the oil-spill and reference site, respectively, in summer; 50–96% and 93–95% in winter; 83–84% and 73–74% in spring). Carnivores were also highly reliant on MPB, although to a slightly lower extent than deposit feeders and omnivores at the oil-spill site (74–76% and 81–82% at the oil-spill and reference site, respectively, in summer; 52–54% and 80–81% in spring). By contrast, the suspension feeders utilized more pelagic carbon sources than the other feeding groups, with an almost 1:1 mixture of MPB and phytoplankton at both sites during all three seasons (49–50% and 46–47% at the oil-spill and reference site, respectively, in summer; 57–59% and 49–50% in winter; 64–65% and 34–35% in spring).

cases, Fig. 7). The mean d13C and d15N values of the fish sampled from the oil-spill site were significantly different from those collected from the reference site during both 2009 (Student’s t-test, t11, 0.05 = 0.67, p = 0.517 for d13C; t11, 0.05 = 0.97, p = 0.351 for d15N) and 2010 (t25, 0.05 = 1.16, p = 0.258 for d13C; t25, 0.05 = 0.99, P = 0.333 for d15N). Common fish species, which were collected at both the oil-spill and reference site, exhibited insignificant spatial differences in their mean isotopic values during 2009 (paired t-test, t3, 0.05 = 0.28, p = 0.801 for d13C; t3, 0.05 = 0.48, p = 0.661 for d15N) and 2010 (t7, 0.05 = 0.95, p = 0.372 for d13C; t7, 0.05 = 0.88, p = 0.361 for d15N, Fig. 8). The fish were primarily dependent (>70%) on MPB rather than phytoplankton-derived carbon at both sites and during both years (95% confidence interval by SIAR, Fig. 9), which showed that most fish were involved in the benthic food chain rather than the pelagic food chain. The mean trophic positions of fish were similar at the oil-spill and reference sites (Fig. 10) during 2009 (two sample t-test, t11, 0.05 = 0.92, p = 0.376) and 2010 (t17, 0.05 = 1.00, p = 0.326). The apex predator occupied a similar trophic position at the oil-spill site (5.2 in 2009; 4.4 in 2010) and at the reference site (4.9 in 2009; 4.2 in 2010) in both years, thereby indicating that the food chain length was almost identical at the oil-spill and reference sites. In particular, the apex predators at the oil-spill site were Muraenesox cinereus (Pike conger) and Takifugu niphobles (Grass puffer) in 2009 and 2010, respectively, whereas Paralichthys olivaceus (Olive flounder) was the top predator at the reference site in both years. In 2009, the trophic position of M. cinereus was 1.0 higher at the oil-spill site while that of P. olivaceus was 0.8 lower at the oil-spill site.

3.3. Fish

4. Discussion

As found in the macrobenthic invertebrates, the fish species sampled from the oil-spill and reference sites in summer had distinctly higher d13C and d15N values (>10‰) compared with the values for petroleum (one sample t-test, p < 0.001 for all eight

Our investigation showed different isotopic data among consumers from the impacted area and crude oil, revealing that the assimilation of petroleum-derived carbon by the fauna is unimportant and rejecting our initial hypothesis. Furthermore, mixing

176

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180

Deposit feeders

Omnivores

Carnivores

20

40

60

80

100

Feeding guild Suspension feeders

80 60 40 20

Benthic contribution (%)

100 0

A Summer

20

40

60

80

100 0

B Winter

0

C Spring OS

RS

OS

RS

OS

RS

OS

RS

Site Fig. 6. SIAR (stable isotope analysis in R) results for relative trophic contribution (%) of microphytobenthos to the diet of macrobenthic invertebrates collected at the oil-spill and reference sites. Microphytobenthos and suspended particulate organic matter were assumed as two end members. (A) Summer; (B) winter; (C) spring. The boxes indicate 95%, 75%, 25%, and 5% confidence intervals, respectively.

18

A 2009

B 2010

δ15N (‰)

16

14

12

10 Oil-spill site Reference site

8 -22

-20

-18

-16

δ13C (‰)

-14

-12

-22

-20

-18

-16

-14

-12

δ13C (‰)

Fig. 7. d15N vs. d13C dual plot of fish at the oil-spill site (solid) and at the reference site (open). Error bar indicates ± 1SD (for n, see Table 2). (A) 2009 summer; (B) 2010 summer.

177

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180 -14

16

A δ13C (‰)

B δ15N (‰) 15

-16

Reference site

Reference site

-15

-17 -18 -19 -20

2009 2010

-21

14 13 12 11 10

-21

-20

-19

-18

-17

-16

-15

-14

Oil-spill site

10

11

12

13

14

15

16

Oil-spill site

100

was found when we included crude oil as possible food source. Furthermore, SI signatures of consumer feeding guilds had no significant differences between the oil-spill and reference sites although isotopic niche of each species was not consistent between sites. Mean d13C of the macrobenthos at the oil-spill site was rather 1.3‰ higher than that at the reference site in spring. This was likely to reflect slightly higher benthic contribution to the nutrition of suspension feeders, deposit feeders, and omnivores as indicated in Fig. 6. Therefore, we suggested that benthic community of the oil-spill site was reestablished 1.5 years after the HSOS. Probably cleaning effort together with bacterial breakdown that metabolize the toxic hydrocarbons (Blumer et al., 1971) may explain the little influence of hydrocarbon in the local fauna.

80

80

60

60

40

40

20

20

Benthic contribution (%)

100

Fig. 8. 1:1 comparison on the isotopic values of common fish species between the reference site vs. the oil-spill site. Two sampling periods, 2009 (solid) and 2010 (open), were distinguished by different colors of symbol. The dotted lines indicate 1:1 diagonals. (A) Comparison on d13C values; (B) comparison on d15N values.

A 2009

B 2010 0

0

OS

OS

RS Site

RS Site

Fig. 9. SIAR (stable isotope analysis in R) results for relative trophic contribution of microphytobenthos to the diet of fish collected at the oil-spill and reference sites compared to suspended particulate organic matter. (A) 2009 summer; (B) 2010 summer. The boxes indicate 95%, 75%, 25%, and 5% confidence intervals, respectively.

Trophic position

5.0

5.0

A 2009

4.5

4.5

4.0

4.0

3.5

3.5

3.0

3.0

2.5

B 2010

2.5

OS

RS Site

OS

RS Site

Fig. 10. Box–whisker plot for trophic position of fish at the oil-spill site (gray) and the reference site (open) in summer 2009 (A) and 2010 (B). The box boundaries indicate 25th and 75th percentiles. The horizontal line in the middle of the box is the median, and the whiskers mark the 10th and 90th percentiles. Outliers were distinguished with X symbols.

models support this suggestion and identified MPB as a main component in the diet of consumers within the impacted coastal area. It should be noted that no reliable solution in the mixing models

4.1. Integration of petrogenic carbon The d13C evidence from consumers rejected the hypothesis that carbon from the HSOS was incorporated into the intertidal food web of Taean beach. We examined this hypothesis by postulating two possible pathways that may allow petroleum-derived carbon to enter the community. One pathway may involve the influx of petrogenic carbon via autotrophic carbon fixation in the community. Based on the present study, this hypothesis may be rejected because of the much higher d13C values for MPB and SPOM at the oil-spill site compared with those for crude oil. Moreover, the spatial consistency of the values also indicates that the microalgae at the oil-spill site utilized similar inorganic carbon sources to those in the unaffected environment (i.e., the reference site). In addition, the d13C values for MPB and SPOM at the oil-spill site were within the ranges reported by previous investigations of the natural coastal food web of the Korean peninsula (16‰ to 13‰ and 24‰ to 20‰, respectively), indicating the incorporation of natural inorganic carbon sources rather than petrogenic ones (Kang et al., 2003, 2007, 2008). Most of the petrogenic hydrocarbon in the water column or sediment can be decomposed into inorganic carbon by microbial degradation rather than being metabolized directly by autotrophs (Atlas, 1981; Leahy and Colwell, 1990). Furthermore, Graham et al. (2010) showed that when SPOM is contamined by crude oil, the d13C value of SPOM decreased by 1–4‰ relative to the natural value non affected by crude oil. In our case, the isotopic values in MPB and SPOM at the affected area by oil were similar to those at the reference site, suggesting that a very low deposition of oil. The second possible pathway for the incorporation of petrogenic carbon into food webs is the direct ingestion of crude oil by consumers at higher trophic level. PAHs from petroleum have been detected frequently in the tissues of benthic invertebrates such as

178

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180

mussels, but they are hardly metabolized by benthic invertebrates and fish (James, 1989; Meador et al., 1995; Booth et al., 2006). By contrast, benthic invertebrates located near natural petroleum seepage in the Santa Barbara Channel had approximately 1‰ lower d13C values than those collected far from the petroleum-affected site (Spies and DesMarais, 1983). Fry and Anderson (2014) showed no assimilation of carbon derived from an oil spill by filter feeding invertebrates in the region of the Deepwater Horizon, in the Gulf of Mexico. In the present study, the different d13C values between consumers and those of petroleum together with the spatial consistency in the d13C levels of the consumers between impacted and non-impacted sites suggested that little petrogenic carbon remained in the tissue of the consumers at the oil-spill site. Furthermore, our d13C range of the consumers were within the ranges of the values obtained for MPB and SPOM which clearly indicates that these two carbon sources comprised the isotopic end members of the consumers, but there was no major contribution of petrogenic carbon. The circular statistic approach also support our suggestion of non-assimilation of petrogenic carbon into the food web since the direction of the isotopic differences in consumers between sites were random and showing no general trend (Fig 4). If consumers were assimilating petrogenic carbon, the consumers at the oil-spill site would show lower d13C and d15N values than those at the reference site. In this case, the vectors of the isotopic differences from the oil-spill site to the reference site should have been oriented toward 270°. However, the results of our circular statistical analysis showed that the arrows did not converge to 270°, but instead they pointed in almost all directions on the circle, and thus there was a lack of evidence for influx of petrogenic carbon from the trophic base. 4.2. Trophic niche in the newly established community The SIAR mixing model calculations demonstrated the same high importance of benthic sources of organic matter (i.e., MPB) in the macrofaunal food webs. Indeed, the SIAR calculations indicated that suspension feeders utilize a mixture of both sources, whereas deposit feeders exclusively use MPB. The provision of MPB to a pelagic space by a mechanism such as tide- or windinduced resuspension may increase the trophic contribution of benthic food sources, even to the pelagic food chain and thus the entire food web. This is highly consistent with the features of the trophic base observed in the natural tidal flat ecosystems of the Korean peninsula (Kang et al., 2003, 2007). Furthermore, this results showing a relative high importance of MPB as food source compared to the SPOM, suggesting that the benthic pathways via deposit feeders seems to be a major energy flow toward higher trophic levels compared to the pelagic pathways via suspension feeders. Despite slight interannual variations, the d15N value and the trophic position of the apex predator at the oil-spill site were almost identical to those at the reference site (Fig. 10). The trophic positions of the apex predators were approximately 5 at both sites in 2009 and 4 in 2010. This spatial consistency in the trophic position of the top predators indicated that the newly established community at the oil-spill site had a food chain length similar to that of the reference community at the reference site. Although the taxonomic composition of the communities differed between sites (Yu et al., 2013), the identical food chain lengths at the oil-spill and reference sites suggested that the intermediate trophospecies composition of the newly established community was similar to that of the reference community. This suggest that the oil spill accident did not influence the number of energy transfers from basal sources to top predators and therefore the energetic efficiencies within the food web (Yodzis and Winemiller, 1999).

We suggest that two critical processes may promote the recovery of trophic relationship within the impacted area: trophic plasticity and the prevalence of omnivory (Belgrano et al., 2005; Thompson et al., 2007). The dramatic changes in the d15N values of apex predators and some carnivorous invertebrate species indicated their apostatic feeding behavior and trophic plasticity. For example, during summer 2009, the apex predators switched from P. olivaceus at the reference site (mean d15N = 15.0‰, TP = 4.9) to M. cinereus at the oil-spill site (mean d15N = 15.7‰, TP = 5.2), although both species were found in both communities. The d15N values of M. cinereus (mean d15N = 13.5‰, TP = 4.2) at the reference site was similar to that of P. olivaceus (mean d15N = 13.2‰, TP = 4.1) at the oil-spill site, which suggested that there was an interchange in the trophic positions of the two species. Furthermore, previous work showed that the abundance and biomass of P. olivaceus was lower at the oil-spill site (109 individuals km3 and 47.6 g individual1) than at the reference site (187 individuals km3 and 16.1 g individual1; MLTM, 2009). This changes in the biometric and abundance of top predator between sites may help to explain the different trophic niche of this species within each site. The mean size of P. olivaceus was greater at the oil-spill site than that at the reference site, but its trophic level was lower at the oil-spill site. By contrast, although the mean size of M. cinereus was smaller at the oil-spill site (weight = 42.0 g individual1) than that at the reference site (weight = 94.0 g individual1), its trophic level was higher at the oil-spill site compared with that at the reference site. This intraspecific trophic niche shift is likely to reflect their apostatic feeding behavior (rather than an ontogenetic shift) in the newly established community. Benthic fauna can also display a considerable trophic plasticity. At the oil-spill site, H. penicillatus had similar d13C levels to those at the reference site, but the d15N levels were 3.4‰ higher than those at the reference site (changing magnitude = 3.8, angle = 162.6°) in summer. These results demonstrated that H. penicillatus was considered a deposit feeder, which primarily consumes MPB, at the oil-spill site, while it displayed carnivorous feeding behavior, which probably practices cannibalism (Kurihara and Okamoto, 1987), at the reference site. The niche shift of R. venosa (magnitude = 2.4, angle = 7.1°) also demonstrates its carnivorous characteristics, where it primarily predated omnivores or deposit feeders at the oil-spill site, and its omnivorous characteristics, where it consumed organic matter with high d13C levels but low d15N levels (i.e., MPB) at the reference site. In this context, the trophic plasticity that some organism can display highlighted the problems assuming same trophic habits for a particular species in different regions (Riera, 2010). Recent articles demonstrated that isotopic variation could be used to detect individual specialization or generalism (Romanuk et al., 2006; Araújo et al., 2011). In our study, omnivores had wide d15N ranges that ranged from primary consumers to carnivorous levels, and their d13C levels were located in the middle of the isotopic baseline values (i.e., suspension feeders and deposit feeders). This isotopic distribution support the suggestion that omnivory represents a major feeding feature at both sites (Fig. 3). In addition, the differences in the mean d15N levels of primary consumers and carnivores in the present study (about 2.3‰) were lower than the generally accepted Dd15N levels (3–4‰). This indicates that carnivores predated both primary consumers and low-trophic-level invertebrates, which suggests the prevalence of generalists and/ or omnivory in both food webs (Deegan and Garritt, 1997). This is also supported by the noninteger trophic position characteristics of fish, which indicates that there was a trophic continuum in both communities (Fig. 10). The high degree of trophic plasticity, the prevalence of omnivory in the food webs might enhance the buffering capacity of the trophic cascade and thus the ecosystem resistance (Polis and Strong, 1996; Pace et al., 1999; Antonio et al., 2006; Vandermeer, 2006; Thompson et al., 2007).

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180

In summary, our results showed similar food web structure between the impacted area by the oil-spill and the reference area. Furthermore, based on our isotopic data in consumers and food sources we suggest no assimilation of the oil spill, while MPB seems to be highly utilized by consumers within the coastal area of the Tean peninsula. We suggest that some consumers can display a considerable trophic plasticity and omnivory that may explain the recovery of the food web structure within the impacted area increasing the robustness of the food web structure (Vandermeer, 2006; Thompson et al., 2007; Ramos-Jiliberto et al., 2011). Our results reinforce the idea that circular statistic approach could be useful to identify differences in food web structure using stable isotopes data (Schmidt et al., 2007), with the incorporation of terms of enrichment factors for the consumers. Acknowledgements This research was a part of the project titled ‘‘Oil Spill Environmental Impact Assessment and Restoration (PM58000)’’, funded by the Ministry of Oceans and Fisheries, Korea. This research was partly supported by the project titled ‘‘Long-term change of structure and function in marine ecosystems of Korea’’, funded by the Ministry of Oceans and Fisheries, Korea. References Antonio, B., Alvarez-ossorio, M.T., Varela, M., 2006. Phytoplankton and macrophyte contributions to littoral food webs in the Galician upwelling estimated from stable isotopes. Mar. Ecol. Prog. Ser. 318, 89–102. Araújo, M.S., Bolnick, D.I., Layman, C.A., 2011. The ecological causes of individual specialisation. Ecol. Lett. 14, 948–958. Atlas, R.M., 1981. Microbial degradation of petroleum hydrocarbons: an environmental perspective. Microbiol. Rev. 45, 180–209. Barbier, E.B., 2011. The value of estuarine and coastal ecosystem services. Ecol. Monogr. 81, 169–193. Barbier, E.B., Koch, E.W., Silliman, B.R., Hacker, S.D., Wolanski, E., Primavera, J., Granek, E.F., Polasky, S., Aswani, S., Cramer, L.A., Stoms, D.M., Kennedy, C.J., Bael, D., Kappel, C.V., Perillo, G.M.E., Reed, D.J., 2008. Coastal ecosystem-based management with nonlinear ecological functions and values. Science 319, 321– 323. Belgrano, A., Scharler, U.M., Dunne, J., Ulanowicz, R.E., 2005. Aquatic Food Webs: An Ecosystem Approach. Oxford University Press, NY, pp. 120–126. Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37, 911–917. Blumer, M., Guillard, R.R.L., Chase, T., 1971. Hydrocarbons of marine phytoplankton. Mar. Biol. 8, 183–189. Booth, A.M., Sutton, P.A., Lewis, C.A., Lewis, A.C., Scarlett, A., Chau, W., Widdows, J., Rowland, S.J., 2006. Unresolved complex mixtures of aromatic hydrocarbons: thousands of overlooked persistent, bioaccumulative, and toxic contaminants in mussels. Environ. Sci. Technol. 41, 457–464. Buchheister, A., Latour, R.J., 2010. Turnover and fractionation of carbon and nitrogen stable isotopes in tissues of a migratory coastal predator, summer flounder (Paralichthys dentatus). Can. J. Fish. Aquat. Sci. 67, 445–461. Chapelle, F., 2001. Ground-water Microbiology and Geochemistry. John Wiley & Sons, NY, pp. 356–378. Costanza, R., d’Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K., Naeem, S., O’Neill, R.V., Paruelo, J., Raskin, R.G., Sutton, P., van den Belt, M., 1997. The value of the world’s ecosystem services and natural capital. Nature 387, 253–260. Couch, C.A., 1989. Carbon and nitrogen stable isotopes of meiobenthos and their food resources. Estuar. Coast. Shelf Sci. 28, 433–441. Deegan, L.A., Garritt, R.H., 1997. Evidence for spatial variability in estuarine food webs. Mar. Ecol. Prog. Ser. 147, 31–47. Fauchald, K., 1979. The diet of worms: a study of polychaete feeding guilds. Oceanogr. Mar. Biol. Annu. Rev. 17, 193–284. Focken, U., Becker, K., 1998. Metabolic fractionation of stable carbon isotopes: implications of different proximate compositions for studies of the aquatic food webs using d13C data. Oecologia 115, 337–343. Fry, B., 2006. Stable Isotope Ecology. Springer, NY, p. 46. Fry, B., Anderson, L.C., 2014. Minimal incorporation of Deepwater Horizon oil by estuarine filter feeders. Mar. Pollut. Bull. 80, 282–287. Fuex, A., 1977. The use of stable carbon isotopes in hydrocarbon exploration. J. Geochem. Explor. 7, 155–188. Gin, K.Y.H., Huda, M.K., Lim, W.K., Tkalich, P., 2001. An oil spill-food chain interaction model for coastal waters. Mar. Pollut. Bull. 42, 590–597. Graham, W.M., Condon, R.H., Carmichael, R.H., D’Ambra, I., Patterson, H.K., Linn, L.J., Hernandez, F.J., 2010. Oil carbon entered the coastal planktonic food web during the Deepwater Horizon oil spill. Environ. Res. Lett. 5, 045301.

179

Griebler, C., Safinowski, M., Vieth, A., Richnow, H.H., Meckenstock, R.U., 2004. Combined application of stable carbon isotope analysis and specific metabolites determination for assessing in situ degradation of aromatic hydrocarbons in a tar oil-contaminated aquifer. Environ. Sci. Technol. 38, 617–631. Hartman, B., Hammond, D.E., 1981. The use of carbon and sulfur isotopes as correlation parameters for the source identification of beach tar in the southern California borderland. Geochim. Cosmochim. Acta 45, 309–319. Hong, S., Park, K., Park, C., Han, C., Suh, H., Yun, S., Song, C., Jo, S., Lim, H., Kang, Y., 2006. Marine Invertebrates in Korean Coasts. Academy Publishing Company Inc., Seoul, Korea, p. 479. James, M.O., 1989. Biotransformation and disposition of PAH in aquatic invertebrates. In: Varanasi, U. (Ed.), Metabolism of Polycylic Aromatic Hydrocarbons in the Aquatic Environment. CRC, Boca Raton, FL, pp. 69–91. Jeffrey, A.W., 2006. Application of Stable Isotope Ratios in Spilled Oil Identification. Oil Spill Forensics: Fingerprinting and Source Identification. Academic Press, NY, pp. 207–227. Ji, K., Seo, J., Liu, X., Lee, J., Lee, S., Lee, W., Park, J., Khim, J.S., Hong, S., Choi, Y., Shim, W.J., Takeda, S., Giesy, J.P., Choi, K., 2011. Genotoxicity and endocrinedisruption potentials of sediment near an oil spill site: two years after the Hebei Spirit oil spill. Environ. Sci. Technol. 45, 7481–7488. Jung, J.H., Kim, M., Yim, U.H., Ha, S.Y., An, J.G., Won, J.H., Han, G.M., Kim, N.S., Addison, R.F., Shim, W.J., 2011. Biomarker responses in pelagic and benthic fish over 1 year following the Hebei Spirit oil spill (Taean, Korea). Mar. Pollut. Bull. 62, 1859–1866. Kang, C.-K., Kim, J.B., Lee, K.-S., Kim, J.B., Lee, P.-Y., Hong, J.-S., 2003. Trophic importance of benthic microalgae to macrozoobenthos in coastal bay systems in Korea: dual stable C and N isotope analyses. Mar. Ecol. Prog. Ser. 259, 79–92. Kang, C.K., Choy, E.J., Paik, S.K., Park, H.J., Lee, K.S., An, S., 2007. Contributions of primary organic matter sources to macroinvertebrate production in an intertidal salt marsh (Scirpus triqueter) ecosystem. Mar. Ecol. Prog. Ser. 334, 131–143. Kang, C.K., Choy, E.J., Son, Y., Lee, J.Y., Kim, J., Kim, Y., Lee, K.S., 2008. Food web structure of a restored macroalgal bed in the eastern Korean peninsula determined by C and N stable isotope analyses. Mar. Biol. 153, 1181–1198. Kennish, M.J., 2002. Environmental threats and environmental future of estuaries. Environ. Conserv. 29, 78–107. Kent, A.C., Day, R.W., 1983. Population dynamics of an infaunal polychaete: the effect of predators and an adult–recruit interaction. J. Exp. Mar. Biol. Ecol. 73, 185–203. Kim, M., Yim, U.H., Hong, S.H., Jung, J.-H., Choi, H.-W., An, J., Won, J., Shim, W.J., 2010. Hebei Spirit oil spill monitored on site by fluorometric detection of residual oil in coastal waters off Taean, Korea. Mar. Pollut. Bull. 60, 383–389. Kim, M., Hong, S.H., Won, J., Yim, U.H., Jung, J.H., Ha, S.Y., An, J.G., Joo, C., Kim, E., Han, G.M., Baek, S., Choi, H.W., Shim, W.J., 2013. Petroleum hydrocarbon contaminations in the intertidal seawater after the Hebei Spirit oil spill–effect of tidal cycle on the TPH concentrations and the chromatographic characterization of seawater extracts. Water Res. 47, 758–768. Kline, J., Thomas, C., 1999. Temporal and spatial variability of 13C/12C and 15N/14N in pelagic biota of Prince William Sound, Alaska. Can. J. Fish. Aquat. Sci. 56, 94– 117. Kurihara, Y., Okamoto, K., 1987. Cannibalism in a grapsid crab, Hemigrapsus penicillatus. Mar. Ecol. Prog. Ser. 41, 123–127. Leahy, J.G., Colwell, R.R., 1990. Microbial degradation of hydrocarbons in the environment. Microbiol. Rev. 54, 305–315. Lee, C.I., Kim, M.C., Kim, H.C., 2009. Temporal variation of chlorophyll a concentration in the coastal waters affected by the Hebei Spirit oil spill in the West Sea of Korea. Mar. Pollut. Bull. 58, 496–502. Lee, H.J., Shim, W.J., Lee, J., Kim, G.B., 2011. Temporal and geographical trends in the genotoxic effects of marine sediments after accidental oil spill on the blood cells of striped beakperch (Oplegnathus fasciatus). Mar. Pollut. Bull. 62, 2264–2268. Macko, S.A., Parker, P.L., Botello, A.V., 1981. Persistence of spilled oil in a Texas salt marsh. Environ. Pollut. Ser. B, Chem. Phys. 2, 119–128. Mazeas, L., Budzinski, H., 2002a. Improved accuracy of GC-MS quantification of aliphatic and aromatic hydrocarbons in marine sediments and petroleums. Validation on reference matrices and application to the Erika oil spill. Int. J. Environ. Anal. Chem. 82, 157–173. Mazeas, L., Budzinski, H., 2002b. Molecular and stable carbon isotopic source identification of oil residues and oiled bird feathers sampled along the Atlantic coast of France after the Erika oil spill. Environ. Sci. Technol. 36, 130–137. McCutchan, J.H., Lewis, W.M., Kendall, C., McGrath, C.C., 2003. Variation in trophic shift for stable isotope ratios of carbon, nitrogen, and sulfur. Oikos 102, 378– 390. Meador, J.P., Stein, J.E., Reichert, W.L., Varanasi, U., 1995. Bioaccumulation of polycyclic aromatic hydrocarbons by marine organisms. In: Reviews of Environmental Contamination and Toxicology. Springer, NY, pp. 79–165. MLTM (Ministry of Land, Transport and Maritime Affairs, Korea), 2009. Environmental Impact Assessment and Environmental Restoration of Hebei Spirit Oil Spill. Technical Report. MLTM, Seoul, Korea. Pace, M.L., Cole, J.J., Carpenter, S.R., Kitchell, J.F., 1999. Trophic cascades revealed in diverse ecosystems. Trends Ecol. Evol. 14, 483–488. Parnell, A.C., Inger, R., Bearhop, S., Jackson, A.L., 2010. Source partitioning using stable isotopes: coping with too much variation. PLoS ONE 5, e9672. Peterson, C.H., Rice, S.D., Short, J.W., Esler, D., Bodkin, J.L., Ballachey, B.E., Irons, D.B., 2003. Long-term ecosystem response to the Exxon Valdez oil spill. Science 302, 2082–2086.

180

E. Han et al. / Marine Pollution Bulletin 90 (2015) 167–180

Polis, G.A., Strong, D., 1996. Food web complexity and community dynamics. Am. Nat. 147, 813–846. Post, D.M., 2002. Using stable isotopes to estimate trophic position: models, methods, and assumptions. Ecology 83, 703–718. Ramos-Jiliberto, R., Valdovinos, F.S., Arias, J., Alcaraz, C., García-Berthou, E., 2011. A network-based approach to the analysis of ontogenetic diet shifts: an example with an endangered, small-sized fish. Ecol. Complex. 8, 123–129. Reddy, C.M., Arey, J.S., Seewald, J.S., Sylva, S.P., Lemkau, K.L., Nelson, R.K., Carmichael, C.A., McIntyre, C.P., Fenwick, J., Ventura, G.T., Van Mooy, B.A.S., Camilli, R., 2012. Composition and fate of gas and oil released to the water column during the Deepwater Horizon oil spill. Proc. Natl. Acad. Sci. U.S.A. 109, 20229–20234. Riera, P., 2010. Trophic plasticity of the gastropod Hydrobia ulvae within an intertidal bay (Roscoff, France): a stable isotope evidence. J. Sea Res. 63, 78–83. Romanuk, T.N., Beisner, B.E., Martinez, N.D., Kolasa, J., 2006. Non-omnivorous generality promotes population stability. Biol. Lett. 2, 374–377. Rumolo, P., Barra, M., Gherardi, S., Marsella, E., Sprovieri, M., 2011. Stable isotopes and C/N ratios in marine sediments as a tool for discriminating anthropogenic impact. J. Environ. Monit. 13, 3399–3408.

Schmidt, S.N., Olden, J.D., Solomon, C.T., Vander Zanden, M.J., 2007. Quantitative approaches to the analysis of stable isotope food web data. Ecology 88, 2793– 2802. Schmidt, S.N., Vander Zanden, M.J., Kitchell, J.F., 2009. Long-term food web change in Lake Superior. Can. J. Fish. Aquat. Sci. 66, 2118–2129. Spies, R.B., DesMarais, D.J., 1983. Natural isotope study of trophic enrichment of marine benthic communities by petroleum seepage. Mar. Biol. 73, 67–71. Thompson, R.M., Hemberg, M., Starzomski, B.M., Shurin, J.B., 2007. Trophic levels and trophic tangles: the prevalence of omnivory in real food webs. Ecology 88, 612–617. Vandermeer, J., 2006. Omnivory and the stability of food webs. J. Theor. Biol. 238, 497–504. Yodzis, P., Winemiller, K.O., 1999. In search of operational trophospecies in a tropical aquatic food web. Oikos 87, 327–340. Yu, O.H., Lee, H.G., Shim, W.J., Kim, M., Park, H.S., 2013. Initial impacts of the Hebei Spirit oil spill on the sandy beach macrobenthic community west coast of Korea. Mar. Pollut. Bull. 70, 189–196. Zar, J.H., 1999. Biostatistical Analysis. Prentice-Hall, NJ, pp. 592–663.