Article

pubs.acs.org/est

Pacific Ciguatoxins in Food Web Components of Coral Reef Systems in the Republic of Kiribati † § † § † ‡ † § † § Yim Ling Mak, , Tak-Cheung Wai, , Margaret B. Murphy, , Wing Hei Chan, , Jia Jun Wu, , † § † § † ‡ § James C. W. Lam, , Leo L. Chan,*, , and Paul K. S. Lam*, , , † ‡ State Key Laboratory in Marine Pollution and Department of Biology and Chemistry, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong Special Administrative Region, China § Research Centre for the Oceans and Human Health, City University of Hong Kong Shenzhen Research Institute, Shenzhen 518057, China

*S Supporting Information

ABSTRACT: Ciguatera fish poisoning (CFP) is a foodborne illness caused by consumption of coral reef fishes contaminated by ciguatoxins (CTXs); of the known CTX congeners, the Pacific ciguatoxins (P-CTXs) are the most toxic. Little is known about the trophodynamics of P-CTXs in coral reef systems. The present study explores the distribution, transfer, and trophic magnification of P-CTX-1, -2, and -3 in coral reef systems with high (ciguatoxic) and low (reference) ciguatoxicity in a CFP-endemic nation by use of liquid chromatography-tandem mass spectrometry (LC-MS/ MS). In ciguatoxic coral reef systems, P-CTXs were detected in 54% of herbivorous fishes [total P-CTXs <0.500−1670 pg/g wet weight (ww)], 72% of omnivorous fishes (<0.500−1810 pg/g ww), and 76% of carnivorous fishes (<0.500−69 500 pg/g ww), as well as a lobster (Panulirus penicillatus; 2.36 pg/g ww) and an octopus (Octopodidae; 2.56 pg/g ww). The dominant P-CTXs in grazers and piscivorous fishes were P-CTX-2 and -1, respectively. No significant correlation between P-CTX levels and lipid content in three target predatory fishes indicated that accumulation of P-CTXs does not depend on fat content. A weak but significant positive relationship was observed between δ15N and P-CTX-1 levels, but further investigation is required to confirm its biomagnification potential.

■ INTRODUCTION CFP-endemic regions, CTXs now represent a global threat to human health.2 Ciguatoxins (CTXs) are a group of natural marine biotoxins; 6 fi The food web hypothesis was proposed by Randall in 1958 more than 30 congeners or isomers have been identi ed in fi fl fi 1 to explain the sources of CTXs in shes. Epibenthic dino agellates and shes. They are highly oxygenated fl fi − dino agellates such as Gambierdiscus spp. have been identi ed lipophilic cyclic polyethers consisting of 13 14 rings fused as causative agents that produced gambiertoxins (such as P- into a rigid ladderlike structure. On the basis of geographic fi CTX-4A/4B) and a suite of CTX intermediate products (such origins and structural variants, ciguatoxins are classi ed as as P-CTX-2 and -3, also known as 52-epi-54-deoxy-CTX-1B fi 2 Paci c (P-CTXs), Caribbean (C-CTXs), or Indian (I-CTXs). and 54-deoxy-CTX-1B, respectively).7 These dinoflagellates not ff CTXs are of great concern, as exposure may a ect the survival only may be attached to macroalgae and turf but also may 3 of invertebrates and marine mammals such as endangered be found in association with sand, coral rubble, and 4 − Hawaiian monk seals (Monachus schauinslandi). In addition, seagrasses.8 10 Herbivorous fishes including surgeonfish and CTXs are present in coral reef fishes on the international parrotfish that graze on macroalgae or other substrates also market, and consumption of fishes containing more than 100 ingest the associated epiphytic dinoflagellates and their toxins, pg of P-CTX-1 equiv/g can cause ciguatera fish poisoning and these fishes are then in turn preyed upon by carnivorous (CFP) in humans.2 Worldwide public health institutions have fishes, which transform CTX precursors and intermediates to ranked CFP as one of the most common food-borne illnesses.1 relatively oxidized CTXs (such as P-CTX-1, also known as P- In the United States, CTXs were viewed as the second most CTX-1B). The lipophilic CTXs accumulate in fishes and then dangerous seafood-borne hazard and were estimated to account 5 for 21% of all seafood-related outbreaks from 1990 to 2006. Received: July 18, 2013 Annually, more than 50 000 people around the world are Revised: November 12, 2013 estimated to be adversely affected by CTXs.2 Because of the Accepted: November 14, 2013 expansion of the international trade of coral reef fishes into Published: November 14, 2013

© 2013 American Chemical Society 14070 dx.doi.org/10.1021/es403175d | Environ. Sci. Technol. 2013, 47, 14070−14079 Environmental Science & Technology Article Table 1. List of and Invertebrates Collected from Two Ciguatoxic Sites (T1 and T2) and a Reference Site (R) in Coral Reef System of Marakei in the Republic of Kiribati

and common name family sampling site n Herbivorous Fishes Acanthurus glaucopareius whitecheek surgeonfish surgeonfishes T1, T2 2 Acanthurus lineatus lined surgeonfish Acanthuridae surgeonfishes T1, T2 4 Acanthurus maculiceps white-freckled surgeonfish Acanthuridae surgeonfishes T1, R 3 Calotomus carolinus Carolines Scaridae T1 1 marginatus striped-fin surgeonfish Scaridae parrotfishes R 1 Kyphosus cinerascens blue sea chub Kyphosidae sea chubs T1 1 Naso lituratus orange-spine unicornfish Acanthuridae surgeonfishes T1 1 ghobban blue-barred parrotfish Scaridae parrotfishes T1, R 5 Scarus globiceps Scaridae parrotfishes R 1 Scarus rubroviolaceus ember parrotfish Scaridae parrotfishes T2 1 Scarus russelii eclipse parrotfish Scaridae parrotfishes T2 1 Scarus schlegeli yellowband parrotfish Scaridae parrotfishes T1 1 three-color parrotfish Scaridae parrotfishes T1 1 Forsten’s parrotfish Scaridae parrotfishes T2 1 Siganus argenteus streamlined spinefoot Siganidae rabbitfishes T1, R 3 Omnivorous Fishes Acanthurus gahhm black surgeonfish Acanthuridae surgeonfishes T1 4 Acanthurus mata elongate surgeonfish Acanthuridae surgeonfishes T1 1 Acanthurus xanthopterus yellowfin surgeonfish Acanthuridae surgeonfishes T1 3 Aluterus scriptus scribbled leatherjacket filefish Monacanthidae filefishes T1 1 Arothron nigropunctatus black-spotted puffer pufferfishes T1 1 Balistapus undulatus orange-lined Balistidae T2 1 Bolbometopon muricatum green humphead parrotfish Scaridae parrotfishes T1 1 ulietensis Pacific double-saddle Chaetodontidae R 1 Chaetodon auriga threadfin butterflyfish Chaetodontidae butterflyfishes T1 1 Cheilopogon atrisignis glider flyingfish Exocoetidae flyingfishes T1, R 3 Ctenochaetus striatus striated surgeonfish Acanthuridae surgeonfishes T1, R 4 Hipposcarus longiceps Pacific longnose parrotfish Scaridae parrotfishes T1 1 niger black triggerfish Balistidae triggerfishes R 1 Myripristis murdjan pinecone soldierfish Holocentridae soldierfishes R 1 Carnivorous Fishes Caesio lunaris lunar fusilier Caesionidae fusiliers R 1 Caesio xanthonotus yellow-back fusilier Caesionidae fusiliers R 1 Cantherhines paradalis honeycomb filefish Monacanthidae filefishes R 1 argus blue-spotted T1, T2, R 24 Cephalopholis urodeta darkfin hind Serranidae groupers T1, R 3 Cephalopholis sexmaculata sixblotch hind Serranidae groupers T2 1 Chaetodon lunula raccoon butterflyfish Chaetodontidae butterflyfishes T1 1 Chaetodon meyeri scrawled butterflyfish Chaetodontidae butterflyfishes T1, R 7 Chaetodon reticulatus mailed butterflyfish Chaetodontidae butterflyfishes T2, R 2 Cheilopogon antoncich Exocoetidae flyingfishes T1 10 Coris aygula clown coris Labridae wrasses T1 1 Diodon hystrix spot-fin porcupinefish Diodontidae ballonfishes T2 1 Diodon liturosus black-blotched porcupinefish Diodontidae ballonfishes T1 1 Epibulus insidiator slingjaw wrasse Labridae wrasses T1 1 coeruleopunctatus white-spotted grouper Serranidae groupers T1 1 Epinephelus fuscoguattus brown- Serranidae groupers T1 2 Epinephelus merra Serranidae groupers T1 2 Epinephelus multinotatus white-blotched grouper Serranidae groupers T1 1 Epinephelus polyphekadion Serranidae groupers T1 3 foursaddle grouper Serranidae groupers T1 4 Epinephelus tauvina Serranidae groupers T1 3 Forcipiger longirostris longnose butterflyfish Chaetodontidae butterflyfishes T1, T2 2 Gomphosus varius bird wrasse Labridae wrasses T1 1 Gracila albomarginata masked grouper Serranidae groupers T1 1 Gymnothorax flavimarginatus yellow-edged moray Muraenidae moray eels T1, T2, R 26 Gymnothorax javanicus giant moray Muraenidae moray eels T1, T2, R 22 Halichoeres hortulanus checkerboard wrasse Labridae wrasses T1 1 Lutjanus bohar blacktail snapper Lutjanidae snappers T1 3

14071 dx.doi.org/10.1021/es403175d | Environ. Sci. Technol. 2013, 47, 14070−14079 Environmental Science & Technology Article Table 1. continued

genus and species common name family type sampling site n Carnivorous Fishes Lutjanus fulvus blacktail snapper Lutjanidae snappers T1 1 Macolor niger black and white snapper Lutjanidae snappers T1 1 Monotaxis grandoculis humpnose big-eye bream Lethrinidae emperors T1, R 4 Myripristis berndti blotcheye soldierfish Holocentridae soldierfishes T1 1 Paracirrhites hemistictus whitespot hawkfish Cirrhitidae hawkfishes T2, R 3 Pomacanthus imperator emperor angelfish angelfishes T2, R 2 Parupeneus bifasciatus doublebar goatfish Mullidae goatfishes T1 1 Pygoplites diacanthus royal angelfish Pomacanthidae angelfishes T1, R 4 Sargocentron tiere blue-lined squirrelfish Holocentridae squirrelfishes T1, T2, R 4 Saurida gracilis gracile lizardfish Synodontidae lizardfishes T1 1 louti yellow-edged lyretail Serranidae groupers T1 1 Zanclus cornutus Moorish idol Zanclidae moorish idol T1, T2, R 4 Invertebrates Carpilius maculatus spotted reef crab Carpiliidae crabs T1 2 Carpilius convexns convex reef crab Carpiliidae crabs R 2 Charybdis paucidentata red swimming crab Portunidae crabs T1, R 3 Octopodidae octopus T2, R 4 Panulirus penicillatus green spiny lobster Palinuridae lobsters T1, R 4

Figure 1. Locations of (a) the Republic of Kiribati, (b) Marakei, and (c) ciguatoxic sites (T) and reference site (R) in Marakei, which is situated at the northern region of Gilbert Islands of the Republic of Kiribati. biomagnify up the food chain. Accordingly, size and trophic the basis of tests or cell-based assays, which are level have been suggested to be important parameters nonspecific and provide little information on CTX concen- governing the ciguatoxicity of fishes.6,10,11 trations and composition. To date, transfer of CTXs from prey (toxic muscle from a The Republic of Kiribati is composed of 33 atolls in the carnivorous fish) to predator (an omnivorous fish) has been central Pacific Ocean and lies on the Equator. Its fishery demonstrated via feeding experiments.12 Several studies have products not only support the local population but also are examined the occurrence of CTXs in dinoflagellates, grazers, exported to other countries such as Australia, Europe, and − and predators at various trophic levels.7,9,13 17 However, there Southeast Asia.17 It was recognized as a high-risk CFP region as are limited studies focusing on biotransformation and early as 1978.17 A recent study demonstrated that distribution biomagnification of these toxins in organisms in a food web of ciguatoxic fishes in one of these atolls, Marakei, was context. Detection and quantification of CTXs has also been localized: ciguatoxicities of blue-spotted grouper (Cephalopholis problematic, as ciguatoxicity in fishes is usually determined on argus), undulated moray (Gymnothorax undulatus), and yellow-

14072 dx.doi.org/10.1021/es403175d | Environ. Sci. Technol. 2013, 47, 14070−14079 Environmental Science & Technology Article edged moray (Gymnothorax flavimarginatus) collected from the lipid content was determined gravimetrically (Supporting western region of Marakei were 30−100 times greater than Information, Table S-1). those of individuals of the same species sampled from the Extraction and Cleanup. Whole bodies of invertebrates − southern region.17 19 In the present study, food web (excluding the internal organs) and dorsal muscle of fishes were components, including invertebrates and fishes (herbivorous, used for quantification of P-CTXs. Extraction of P-CTXs in omnivorous, and carnivorous), were collected from these two samples was carried out on an ASE 200 system (Dionex, regions of the coral reef system in Marakei with the aim to Sunnyvale, CA) as reported previously20 (Supporting Informa- examine (1) the distribution of P-CTXs in food web tion). In brief, 2.5−5 g freeze-dried samples were mixed with components by both analytical methods and cell-based assay, 3.5 g of diatomaceous earth (DE) and transferred into a 22-mL (2) the relationships of body size (i.e., total length and body stainless steel extraction cell (Dionex, Sunnyvale, CA). weight) and lipid content of fishes to CTX accumulation, (3) Extraction of P-CTXs was performed at 75 °C and 1500 psi the biotransformation of P-CTXs at each trophic level, and (4) with two 5-min static extractions with methanol. The extraction the trophic magnification potential of P-CTXs in the ciguatoxic cell was flushed with 60% methanol with a purge time of 100 s. coral reef system. Authentic standards and analytical methods The crude extract was concentrated to dryness via rotary for P-CTXs are limited; in this study, three target P-CTXs with evaporation and redissolved in 50% aqueous methanol before an oxepane backbone, including a biotransformation end being subjected to solid-phase extraction as described product, P-CTX-1, and two biotransformation intermediates, previously.20 Each crude extract was loaded on a C18 cartridge P-CTX-2 and -3, were quantified (Supporting Information, (Agilent BondElut, 500 mg, 6 mL) that was successively Figure S-1). Trophic transfer of these toxins was characterized preconditioned with acetonitrile, methanol, and Milli-Q water. with the aid of stable nitrogen isotope analysis. This approach The cartridge was washed with 6.5 mL of 65% aqueous provides essential field data on the flux, dynamics, and fate of methanol, and P-CTXs were eluted by 12 mL of 65% aqueous CTXs in CFP-endemic coral reef systems in order to assess acetonitrile. The eluent was then mixed with 6 mL of ecological and human health risk and manage fishery resources chloroform and 1 M sodium chloride. The lower organic for control of CFP. layer was collected and dried under a stream of high-purity N2. The residue was then redissolved in 2.5 mL of 2% methanol in ■ MATERIALS AND METHODS chloroform and passed through a silica cartridge (Waters Sep- Chemicals and Reagents. P-CTX-1 (purity ≥95%, Pak, 500 mg, 6 mL) that was preconditioned with chloroform. molecular mass = 1110 g/mol) was isolated and purified P-CTXs were eluted with 8 mL of 10% methanol in from viscera of undulated moray and yellow-edged moray chloroform. The eluent was dried under a stream of high- μ collected from the Republic of Kiribati according to methods purity N2 and resuspended in 100 L of methanol before described previously.13,20 P-CTX-2 and -3 (purity >95%, injection. molecular mass = 1094 g/mol) were purchased from Professor Determination of Pacific Ciguatoxin Levels and Total R. Lewis (The University of Queensland, Australia). Details of Ciguatoxicity. The reported total P-CTX and P-CTX-1, -2, other chemicals and solvents are provided in Supporting and -3 levels were determined via liquid chromatography- Information. tandem mass spectrometry (LC-MS/MS). Separation and Sample Collections. Invertebrates including crabs (n =7, quantification of P-CTXs were performed on an Agilent 1290 three species), lobsters (n = 4, one species) and octopi (n =4, ultraperformance liquid chromatography (UPLC) system one genus) and a total of 205 coral reef fishes of 70 species in (Agilent, Palo Alto, CA) interfaced with a 5500 QTRAP (AB 24 families including angelfish, butterflyfish, emperor, filefish, Sciex, Foster City, CA) with a turbo-ion spray operating in the flyingfish, fusilier, goatfish, grouper, hawkish, lizardfish, moorish positive ion and multiple-reaction-monitoring (MRM) mode. A idol, moray eel, parrotfish, porcupinefish, pufferfish, rabbitfish, 10 μL aliquot of the extract was injected onto a Phenomenex sea chub, snapper, soldierfish, squirrelfish, surgeonfish, trigger- Luna C18(2) column (250 × 2.0 mm i.d., 5 μm). Gradient fish, tuna, and wrasse, representing herbivorous, omnivorous, elution at a flow rate of 200 μL/min was performed with (A) and carnivorous fishes (Table 1) were sampled from two Milli-Q water and (B) 95% acetonitrile in Milli-Q water, both ciguatoxic sites (02°02′ N, 173°15′ E and 02°01′ N, 173°15′ E) of which contained 0.1% formic acid and 2 mM ammonium and a reference site (01°58′ N, 173°15′ E) of Marakei Atoll in formate. The initial gradient condition was set at 63% B and the Republic of Kiribati (Figure 1). Invertebrates were collected maintained for 3 min. It was then ramped to 100% B in 1 min from both sites by scuba divers. Fishes were speared or caught and held for 6 min before returning to 63% B in 0.1 min. The in storing nets, square nets, round nets, and long square nets, column was equilibrated at initial gradient conditions for 4.9 except for moray eels, which were collected in a locally made min prior to the next injection. The three precursor-to- trap known as “TeUu” with bait such as octopus and fragment transitions monitored for P-CTX-1 were m/z 1128.4 surgeonfish. All samples were randomly collected from each → 1075.5 (quantitation ion) and m/z 1128.4 → 1057.5 and location, except for blue-spotted grouper, yellow-edged moray, 1128.4 → 1039.5 (confirmation ions). For P-CTX-2 and -3, m/ and giant moray (Gymnothorax javanicus), which were the z 1112.7 → 1077.8 (quantitation ion) and m/z 1112.7 → designated fish species sampled for assessment of the 1059.8 and 1112.7 → 1041.8 (confirmation ions) were importance of size and lipid content to P-CTX levels because monitored. Mass spectrometer operating parameters are of their relatively high ciguatoxicities.17 All samples were kept at shown in Supporting Information, Table S-2. −20 °C during transportation and before analysis. Measure- Muscle extracts were also screened for toxicity using the ment of total length (centimeters), body weight (grams), and standardized mouse neuroblastoma (Neuro-2a) cells (ATCC, lipid content (percent) of blue-spotted grouper, yellow-edged CCL131; American Type Culture Collection, Manassas, VA) moray, and giant moray were conducted in the laboratory. according to methods described previously17,20 for assessing the Lipids were extracted from aliquots of muscle samples relationship between ciguatoxicity of fishes determined by LC- according to the method described by Folch et al.,21 and MS/MS and this standardized cell-based assay. The ciguatox-

14073 dx.doi.org/10.1021/es403175d | Environ. Sci. Technol. 2013, 47, 14070−14079 Environmental Science & Technology Article

Figure 2. Concentrations (picograms per gram wet weight) of P-CTX-1, -2, and -3 and total P-CTXs in herbivorous, omnivorous, and carnivorous fishes collected from ciguatoxic sites. Numbers above the bars indicate mean concentrations for those species. icities of fish samples were determined from a P-CTX-1 concentrations, as there are no commercially available isotopi- standard curve (Supporting Information, Figure S-2) with a cally based P-CTX standard or other chemically similar limit of quantification (LOQ) of 1.95 pg of P-CTX-1 equiv/g. compounds that can be used as an internal standard. P-CTX- Quality control of the assay was performed by testing each 1 (1 ng/mL), P-CTX-2 (5 ng/mL), and P-CTX-3 (5 ng/mL) mouse neuroblastoma assay (MNA) with additional dosages of standards were injected every two samples in order to check the P-CTX-1 standard of 39.1 pg/mL (n = 3 per plate) to cells. The performance of the instrument. The limits of quantification variations between each MNA performed on the same day (LOQs) were set as the lowest P-CTX concentrations required (interplate) and on different days (interassay) were expressed to produce a S:N ratio of 10:1 with accuracy within 80−120%, as the relative standard deviation (RSD) of all replicates. The and a concentration factor of 50 was applied for the enriched interplate relative standard deviation was 12.0%, and interassay sample volume. The LOQs for P-CTX-1, -2, and -3 were 0.500, relative standard deviation was 23.3%. The assays were 5.00, and 5.00 pg/g wet weight (ww), respectively (Supporting conducted twice and the toxicity values are reported as mean Information, Figure S-3). P-CTX-1 equivalent between two assays (mean of six wells). Stable Nitrogen Isotope Analysis. The use of δ15Nis Details of the MNA are given in Supporting Information. increasingly common as a tracer to provide continuous measure Quality Assurance and Quality Control. The accuracy of of trophic position and also trophic relationship of organisms. the present analytical method was assessed via matrix spike The close correspondence between estimates of trophic level recoveries of P-CTXs by spiking 100 pg of P-CTX-1, 500 pg of based on δ15N and literature dietary data confirms and supports P-CTX-2, and 500 pg of P-CTX-3 to freeze-dried muscle of the validity of the present approach to study trophic structures nontoxic moray eel (n = 2), blue-spotted grouper (n = 2), two- in the present coral reef ecosystem.22 As predators are enriched spot red snapper (Lutjanus bohar; n = 2), striped-fin in δ15N relative to their prey, the top predators usually possess surgeonfish (Ctenochaetus marginatus; n = 2), blue-barred the highest value of δ15N.23 Freeze-dried muscle samples (1.5 ± parrotfish (Scarus ghobban; n = 2), scrawled butterflyfish 0.1 mg) were weighed (Mettler Toledo, 0.00001 g readability) (Chartodon meyeri; n = 1), green spiny lobster (Panulirus in tin capsules and analyzed on a PDZ Europa ANCA-GSL penicillatus; n = 2) and octopus (Octopodidae; n = 2). Matrix elemental analyzer interfaced to a PDZ Europa 20−20 isotope spike recoveries were 73−87% for P-CTX-1, 68−83% for P- ratio mass spectrometer (Sercon Ltd., Cheshire, U.K.) by the CTX-2, and 61−78% for P-CTX-3, with RSDs less than 13% UC Davis Stable Isotope Facility (Department of Plant (Supporting Information, Table S-3). Sciences, University of California at Davis, Davis, California). Calibration curves were made from standard solutions based The stable isotope abundance was expressed as standard delta on 10 calibration levels for P-CTX-1, with concentrations notation (δ15N), defined as the parts-per-thousand (‰) ranging from 0.025 to 100 ng/mL, and eight calibration levels deviation from air for nitrogen according to for P-CTX-2 and -3, with concentrations ranging from 0.25 to 15 100 ng/mL. The correlation coefficients (r) of calibration δ N=−× [(RRsample / standard )] 1] 1000 (1) curves by use of 1/x weighed linear regression analysis were greater than 0.999 for all measured P-CTXs. Quantification of where R is the ratio 15N/14N. The analytical precision (as P-CTXs was based on a calibration curve constructed from the standard deviation for repeated measurements of the internal − + ‰ peak area of [M + H 2H2O] versus the respective P-CTX standards, n = 10) for the measurement was 0.3 .

14074 dx.doi.org/10.1021/es403175d | Environ. Sci. Technol. 2013, 47, 14070−14079 Environmental Science & Technology Article

Statistical Analysis. Statistical analysis was performed with Acanthurus have close-set, denticulate teeth and feed on SigmaStat 3.5 (Aspire Software International, Washburn, VA). filamentous algae, whereas those of the genus Naso have rigid Normality of data was evaluated by Kolmogorov−Smirnov teeth and usually feed on coarse or leafy algae.6 Food choice has tests. Correlation analysis was performed via Pearson product been suggested to be an important factor determining their moment test if the data were normally distributed. Non- ciguatoxicities.6 Since the sample size was small for species in parametric Spearman’s rank order test was used if the data did the genus of Acanthurus and Naso, further study will be not pass normality tests. Linear regression analysis was required to characterize the relationship between feeding performed to examine the relationships between groups if the strategies and P-CTX levels/ciguatoxicity of fishes. data were normally distributed. Statistical significance was Omnivorous Fishes. P-CTXs were detected in omnivorous accepted at p < 0.05. fishes, including black surgeonfish (Acanthurus gahhm), elongate surgeonfish (Acanthurus mata), yellowfin surgeonfish ■ RESULTS AND DISCUSSION (Acanthurus xanthopterus), black-spotted puffer (Arothron fi P-CTX Levels in Invertebrates. Among the three P-CTXs nigropunctatus), striated surgeon sh (Ctenochaetus striatus), fi fi measured in the invertebrates, only P-CTX-1 was present in Paci c longnose parrot sh (Hipposcarus longiceps), orange- fi fi fl one octopus (2.36 pg/g ww) and one lobster (2.56 pg/g ww), lined trigger sh (Balistapus undulatus), and thread n butter y- whereas P-CTXs were not detectable in crabs (n = 3) collected fish (Chaetodon auriga) from ciguatoxic sites. Total P-CTX from ciguatoxic sites (Supporting Information, Table S-4). concentrations in the omnivorous fishes ranged from <0.500 to Octopus and lobster primarily prey on crustaceans, and fish 1810 pg/g ww and their P-CTX-1, -2, and -3 levels were − represent a minor or incidental food source to them.24 26 In <0.500−1210, <5.00−1190, and <5.00−406 pg/g ww, contrast, the diet of crab consists mainly of bivalves, with other respectively, which were comparable to those measured in crustaceans being a minor food source.27 Differences in dietary the herbivorous fishes. Total P-CTX and P-CTX-2 and -3 levels composition could be an important reason explaining the in omnivorous parrotfish (635 ± 898, 432 ± 611, and 203 ± presence of trace P-CTX-1 levels in octopus and lobster but not 287 pg/g ww, respectively; n = 2) were found to be 2−3-fold in crab. In-depth monitoring of P-CTX concentrations in prey greater than those in omnivorous surgeonfish (347 ± 477, 254 of octopus, lobster, and crab may help to clarify sources and ± 370, and 58.6 ± 73.7 pg/g ww, respectively; n = 11). Trace exposure pathways of P-CTXs in these invertebrates. Although level of P-CTX-1 was detected in omnivorous surgeonfish (34.4 the P-CTX-1 levels in the octopus and lobster were ± 55.6 pg/g ww) but not in omnivorous parrotfish. A relatively approximately 5-fold lower than the safety level, its presence high total P-CTX level was detected in orange-lined triggerfish. demonstrated that they could be potential vectors for transfer This could be due to its relatively high trophic level (as of P-CTX-1 in ciguatoxic coral reef systems. No P-CTX was indicated by its δ15N value, 15.54), suggesting that it tends to found in invertebrates collected from the reference site (data feed on higher-trophic-level prey that may have already 11 not shown). accumulated greater amounts of P-CTXs. It is noteworthy P-CTX Levels in Coral Reef Fishes. P-CTX-1, -2, and -3 that P-CTX-1, -2, and -3 were found in an omnivorous black- were detectable in herbivorous (6 out of 11 species, 54%), spotted pufferfish at levels of 11.2, 83.4, and 43.9 pg/g ww, omnivorous (8 out of 11 species, 72%), and carnivorous (28 respectively. The total P-CTX level in this pufferfish was out of 37 species, 76%) coral reef fishes collected from determined to be 139 pg/g ww, which was above the threshold ciguatoxic sites (Figure 2, Supporting Information, Table S-4). level (10 pg of P-CTX-1 equiv/g) for acute induction of CFP. Herbivorous Fishes. Among herbivorous fishes, three species Pufferfish has been generally regarded as a major vector for of surgeonfish [whitecheek surgeonfish (Acanthurus glaucopar- human exposure to tetrodotoxin28 and saxitoxin,29 both of eius), lined surgeonfish (Acanthurus lineatus) and white-freckled which are sodium channel blockers. The occurrence of CTXs in surgeonfish (Acanthurus maculiceps)], two species of parrotfish black-spotted pufferfish reflects that they can also carry sodium [blue-barred parrotfish (Scarus ghobban) and eclipse parrotfish channel activators. These results indicate that it is essential to (Scarus russelii)], and a species of rabbitfish [streamlined develop an accurate and specific method for identifying spinefoot (Siganus argenteus)] collected from ciguatoxic sites multiple biotoxins present in fishes so as to better manage contained total P-CTX levels ranging from <0.500 to 1670 pg/ human health risks associated with seafood consumption. g ww. P-CTX-1, -2, and -3 levels were <0.500−180, <5.00−900, Carnivorous Fishes. Twenty-five species of carnivorous and <5.00−594 pg/g ww. Total P-CTX concentration in fishes collected from ciguatoxic sites (blue-spotted grouper, herbivorous parrotfish (347 ± 685 pg/g ww, n = 9) was yellow-edged moray, and giant moray were not included to generally 5 times higher than in herbivorous surgeonfish (77.0 prevent bias, as they were not sampled randomly) contained ± 141 pg/g ww, n = 9). This was attributed to the greater P- total P-CTX levels that ranged from <0.500 to 5300 pg/g ww. CTX-2 and -3 concentrations detected in these herbivorous Total P-CTX concentrations exhibited the following order: parrotfish (214 ± 426 vs 40.6 ± 64.8 pg/g ww for P-CTX-2 and snapper > grouper > hawkfish > wrasse > emperors > 112 ± 223 vs 11.2 ± 22.2 pg/g ww for P-CTX-3); their P- porcupinefishes > angelfish > butterflyfish > goatfish, soldier- CTX-1 concentrations were found to be comparable. Although fish, squirrelfish, Moorish idol > lizardfish. P-CTX-1, -2, and -3 some previously published reviews suggested that surgeonfish levels in these carnivorous fishes were <0.500−4000, <5.00− and parrotfish are generally ciguatoxic, no P-CTX was 1030, and <5.00 to 466 pg/g ww, respectively. The most toxic detectable in a species of surgeonfish [orange-spine unicornfish carnivorous fish species collected from the ciguatoxic sites (Naso lituratus)] and five species of parrotfish [Carolines included two-spot red snapper (total P-CTX <0.500−5300 pg/ parrotfish (Calotomus carolinus), ember parrotfish (Scarus g ww, n = 3), brown-marbled grouper (Epinephelus fuscoguattus; rubroviolaceus), yellowband parrotfish (Scarus schlegeli), three- total P-CTX 129−3100 pg/g ww, n = 2), and camouflage color parrotfish (Scarus tricolor) and Forsten’s parrotfish (Scarus grouper (Epinephelus polyphekadion; total P-CTX 820−4290 forsteni)], indicating that there were genera/species variations in pg/g ww, n = 3). These fishes have previously been deemed CTX exposure and accumulation. Species of the genus unsafe for human consumption in Tahiti, French and

14075 dx.doi.org/10.1021/es403175d | Environ. Sci. Technol. 2013, 47, 14070−14079 Environmental Science & Technology Article

Figure 3. Composition profiles of P-CTXs in herbivorous, omnivorous, and carnivorous fishes with at least two detectable CTXs collected from ciguatoxic sites.

− − Fanning Island, Republic of Kiribati.15 18,30 32 The heaviest ranged from 96.8 to 2980 pg/g ww (n = 17) for blue-spotted two-spot red snapper, brown-marbled grouper, and camouflage groupers, from 58.0 to 69 500 pg/g ww (n = 15) for yellow- grouper contained the greatest concentrations of total P-CTXs edged moray, and from 51.9 to 34 500 pg/g ww (n = 20) for (i.e., 5300 pg/g ww for a 4.40-kg two-spot red snapper, 3100 giant moray. P-CTX-1, -2, and -3 concentrations were 9.56− pg/g ww for a 8.20-kg brown-marbled grouper, and 4290 pg/g 1710, 28.4−676, and 96.8−2980 pg/g ww for blue-spotted ww for a 2.62-kg camouflage grouper), which was consistent grouper; 46.6−3920, 7.04−24 400, and 4.32−5940 pg/g ww for with other studies suggesting that there was a positive yellow-edged moray; and 34.2−17 000, 17.7−14 900, and 6,11,17 relationship between ciguatoxicity and body size. In the 4.82−4720 pg/g ww for giant moray (Supporting Information, U.S. Virgin Islands, porcupinefish have been designated as low- 11 Table S-1). Although the reference site was only around 4 km risk species for CFP. However, P-CTX-1 was found in a spot- away from the ciguatoxic sites, total P-CTX levels in blue- fin porcupinefish (Diodon hystrix) (46.0 pg/g ww) and a black- fi spotted grouper (n = 7), yellow-edged moray (n = 11), and blotched porcupine sh (Diodon liturosus) (12.3 pg/g ww) at giant moray (n = 2) collected at the reference site were 85-, 93-, levels above the threshold (10 pg/g ww) for acute CFP fi fi and 1240-fold lower than those collected from the ciguatoxic induction. Spot- n porcupine sh and black-blotched porcupi- sites (Supporting Information, Figure S-4). At the reference nefish were top predators at the ciguatoxic sites (δ15N values site, P-CTX-1, -2, and/or -3 were detected in four individual 16.82 and 16.21, respectively), but their levels of total P-CTXs streamlined spinefoot, striated surgeonfish, pinecone soliderfish were much lower than those measured in groupers (median fi δ15 δ15 (Myripristis murdjan), and blue-lined squirrel sh (Sargocentron N value 15.38) and snappers (median N value 15.37). fi ff tiere) but not in other shes or invertebrates. P-CTX-1, -2, and These di erences in P-CTX levels may be due to their dietary fi − preferences: porcupinefish consume mollusks, crustaceans, and -3 concentrations in these sh samples were <0.500 2.87, 33,34 <5.00−26.8, and <5.00−27.8 pg/g ww, respectively. The low or sea urchins as their major prey, whereas groupers and fi snappers are mainly piscivorous.6,11,35,36 Flyingfish usually rely nondetectable P-CTX levels in invertebrates and shes on pelagic food items,7 and P-CTXs were not detected in the collected from the reference site could be a result of variations glider flyingfish (Cheilopogon atrisignis; n = 1) and Cheilopogon in the composition, abundance, and ciguatoxicity of CTX- antoncichi (n = 5) in the present study. These results further producing agents at the ciguatoxic and reference sites, causing a confirmed that dietary and habitat preferences are important spatial shift in the dietary exposure of CTXs in grazers. determinants influencing P-CTX concentration in food web Gambierdiscus spp. were found to be the dominant benthic components of the present coral reef system. Besides, previous dinoflagellates in Marakei, and their abundance at the studies suggested higher levels of P-CTXs may cause behavioral ciguatoxic sites was 43−121-fold larger than at the reference and morphological changes in sensitive fishes,2 which can thus site. Besides, the ciguatoxicities of Gambierdiscus spp. measured be preferentially preyed upon and/or killed directly by P-CTXs by MNA showed that the dinoflagellates from the ciguatoxic and thus limited the amount of P-CTXs accumulated in their sites were up to 400-fold more toxic than those collected from bodies. Total P-CTX concentrations of the target species the reference site (unpublished data).

14076 dx.doi.org/10.1021/es403175d | Environ. Sci. Technol. 2013, 47, 14070−14079 Environmental Science & Technology Article

Relationships between Body Size, Lipid Content, and algae and invertebrates, such as tunicate, sponge, mollusk and Pacific Ciguatoxin Levels. Correlation analysis was per- echinoderm, but also fish were their food sources.38 formed to determine the relationships of body size (in terms of Carnivorous fishes collected at ciguatoxic sites exhibited total length and body weight) and lipid content of blue-spotted diverse P-CTX composition profiles. Carnivorous scrawled grouper, yellow-edged moray, and giant moray to P-CTX levels butterflyfish and emperor angelfish (Pomacanthus imperator) (Supporting Information, Table S-5). Significant positive occupy relatively high trophic levels (median δ15N values 15.20 correlations were consistently observed between total length and 14.31), but their P-CTX composition profiles were and P-CTX concentrations and between body weight and P- comparable to those of herbivorous and omnivorous fishes. CTX concentrations in giant moray. These results are in good P-CTX-2 was the dominant P-CTX, contributing 54% and 57% agreement with those of previous studies speculating that of the total P-CTXs in these two species, respectively. ciguatoxicity of muscle and liver was significantly correlated Butterflyfish and angelfish have been reported to be micro- with the body weight of moray eels.11,17 However, significant invertivores or carnivorous browsers that feed on such positive relationships were found only between total length and as sponges, tunicates, and polychaetes39,40 and consume total P-CTX/P-CTX-2 concentrations in yellow-edged moray, microalgae as incidental food items.41 Their analogous P- and this significant relationship was absent for P-CTX-1 and -3 CTX composition profiles suggested that (1) consumption of levels when considered individually. Additionally, body weight microalgae, instead of microinvertebrates, was the major P- of yellow-edged moray was not significantly correlated with P- CTX exposure pathway to butterflyfish/angelfish and/or (2) CTX-1, -2, or -3 or total P-CTX levels. In blue-spotted grouper, microinvertebrates such as sponge, tunicate, and polychaete no significant relationship was observed between total length or share similar P-CTX composition with microalgae. It is body weight and total P-CTX or P-CTX-1, -2, or -3 levels. necessary to monitor P-CTXs in the lower trophic food web These results indicated that the relationship between body size components in order to elucidate the flux of P-CTXs in a whole and P-CTX levels was highly variable among fish species, and food web. thus avoidance of consuming larger fishes (e.g., usually those Grouper, snapper, wrasse, and moray eel are mainly exceeding 2.5 kg in body weight), a risk management measure piscivorous and may also feed on macroinvertebrates such as recommended in some countries such as Australia,2 may not be crustaceans, holothurians, and cephalopods to different an effective way to control CFP.2,15,16 extents.34 Their median δ15N values were 15.38, 15.37, 14.96, Even though P-CTXs have been proposed to be lipophilic in and 16.17, respectively, reflecting that they occupied higher nature,2 no apparent trend was found between P-CTX levels trophic levels and could be the top predators in the present and muscle lipid content of blue-spotted grouper, yellow-edged coral reef system. As the trophic levels increased, there was an moray, and giant moray, reflecting that accumulation of P- apparent change in patterns of P-CTX composition profiles, CTXs does not depend on fat content alone. Because of the with P-CTX-1 being the predominant P-CTX in these species, absence of correlation, it was not necessary to normalize contributing 23−78% (median 55%) to total P-CTXs, whereas concentrations of P-CTX-1, -2, and -3 and total P-CTXs to P-CTX-2 and -3 accounted for 13−76% (median 36%) and 0− lipid content.37 Instead, P-CTX levels in fishes were reported 28% (median 11%) of total P-CTXs in these fishes. The shift in on a wet weight basis only. P-CTX composition profile could be due to (1) the presence of P-CTX Composition Profiles in Fishes at Ciguatoxic higher P-CTX-1 levels in the prey of piscivorous carnivorous Sites. The composition profiles of P-CTXs in herbivorous, fishes, (2) higher bioaccumulation potential of P-CTX-1 in omnivorous and carnivorous fishes with at least two detectable piscivorous carnivorous fishes, and/or (3) higher capacity of P-CTXs collected from the ciguatoxic sites are summarized in piscivorous carnivorous fishes to biotransform precursors/ Figure 3. The proportions of two intermediates, P-CTX-2 and intermediates of P-CTXs (such as P-CTX-4A/4B, -2, and -3) to -3, were relatively high in the herbivorous fishes. P-CTX-2 P-CTX-1. P-CTX-1 was 3−20 times more potent than its accounted for 46−71% (median 54%) of total P-CTXs, and P- precursors/intermediates.1 The greater levels of P-CTX-1 CTX-3 contributed 25−35% (median 29%) to total P-CTXs. present in these piscivorous carnivorous fishes suggested their Similar composition profiles were observed in omnivorous consumptions can pose a relatively higher risk to seafood fishes, in which P-CTX-2 and -3 were the dominant P-CTXs consumers. with contributions of 44−80% (median 66%) and 10−53% Trophic Magnification of Pacific Ciguatoxins in a (median 23%), respectively. The median δ15Nvaluesof Ciguatoxic Marine Food Web. In this study, the herbivorous and omnivorous fishes were 13.41 and 13.16, quantification of P-CTXs was carried out for invertebrates indicating that they shared comparable trophic levels in the and fishes from two ciguatoxic sites (T1 and T2), and therefore present coral reef system. P-CTX-1 has been proposed to be the examination of biomagnification of P-CTXs was only the biotranformation end product of P-CTX-4A/4B;9 in the conducted for a part of the food web. Significant differences in present study, this CTX accounted for <1−28% (median 11%) δ15N values and P-CTX concentrations were observed in blue- and 0−67% (median 9%) of the total P-CTXs in herbivorous spotted groupers collected from T1 and T2 (Supporting and omnivorous fishes, respectively. In a recently published Information, Table S-6), indicating that there could be variation study, low levels of P-CTX-2 and -3, but no P-CTX-1, were in food sources and also P-CTX exposure levels in food web detected in an extract of a RG-1 strain of Gambierdiscus toxicus components at T1 and T2. To minimize bias, the culture.7 The presence of P-CTX-1 in these grazers suggested bioaccumulation potential of P-CTXs was assessed only on that biotransformation and bioaccumulation of P-CTX-1 from the basis of food web components collected from T1, which P-CTX-4A, -4B, -2, and/or -3 may occur once CTX-producing included food web components of relatively large sample size. dinoflagellates are consumed by these organisms. In contrast to A weak but significant positive relationship was observed the results for the grazers, the P-CTX profile of the omnivorous between δ15N values and ln P-CTX-1 concentration (n = 92, r2 orange-lined triggerfish was dominated by P-CTX-1 (67%), = 0.314, p < 0.001) of individual invertebrate and fish samples probably because of a difference in its food choice: not only collected from T1. This significant positive relationship was

14077 dx.doi.org/10.1021/es403175d | Environ. Sci. Technol. 2013, 47, 14070−14079 Environmental Science & Technology Article also found between δ15N values and ln P-CTX-1 concentrations help to interpret patterns of P-CTXs. Last but not least, the (n = 20, r2 = 0.396, p < 0.01) when fish samples were grouped ultratrace P-CTX levels in the low trophic organisms restrict by their corresponding families. However, there was no their measurement by the LC-MS/MS method, and a more significant relationship between δ15N values and ln concen- sensitive analytical method may be required to assess transfer of trations of P-CTX-2 or -3. δ15N values was only weakly P-CTXs in a complete food web. correlated with ln total P-CTX concentration (n = 94, r2 = 0.154, p < 0.001) of individual invertebrate and fish samples, ■ ASSOCIATED CONTENT fi fi and no signi cant correlation was found when sh samples *S Supporting Information were grouped by their corresponding families (Supporting Additional text with details on chemicals and reagents, Information, Table S-7). These results indicated that P-CTX-1 fi extraction and clean-up procedures, MNA, and trophic possessed a relatively high biomagni cation tendency and it magnification; five figures showing chemical structures, stand- could be a consequence of biotransformation of P-CTX-2 and ard curve and chromatograms of standards, concentrations of -3, resulting in an increase of P-CTX-1 levels along the food toxins in three target species, and correlation between chain. Further study may help to confirm trophic magnification fi ciguatoxicity determined by LC-MS/MS and MNA; and eight factor (TMF) and hence biomagni cation potential of P-CTXs tables listing properties of three target species, MS operating (Supporting Information). parameters, matrix spike recoveries, δ15N values and concen- Correlation between Instrumental and Bioassay trations of P-CTXs, correlations between properties and P- Results. The muscle extracts of herbivorous, omnivorous, fi CTX concentrations, statistical analysis, linear regression, and and carnivorous shes used for LC-MS/MS analysis were also ciguatoxicity determined by LC-MS/MS and MNA. This tested in the mouse neuroblastoma assay (MNA). Total material is available free of charge via the Internet at http:// ciguatoxicity (P-CTX-1 equivalents) was expressed as the sum pubs.acs.org. of the concentrations of P-CTX-1, -2, and -3 in muscle determined by LC-MS/MS multiplied by their respective ■ AUTHOR INFORMATION toxicity equivalent factors (TEFs) derived from acute intra- peritoneal injection LD values in mice.1 The TEFs were 1, Corresponding Authors 50 * 0.3, and 0.3 for P-CTX-1, -2, and -3, respectively. Total (L.L.C.) Mailing address: Room B1705, Academic Building, ciguatoxicity of muscle extracts for herbivorous, omnivorous, City University of Hong Kong, Kowloon Tong, Hong Kong; and carnivorous fishes (n = 112, r=0.959, p < 0.001) phone: +852 3442 4125; fax: +852 2194 2281; e-mail: determined by LC-MS/MS was well-correlated with data [email protected]. * obtained from the MNA (Supporting Information, Figure S-5 (P.K.S.L.) Mailing address: Room B6705, Academic Building, and Table S-8). P-CTX-1, -2, and -3 measured by LC-MS/MS City University of Hong Kong, Kowloon Tong, Hong Kong; in these fishes contributed to 16.8−19.8% (i.e., a median of phone: +852 3442 6828; fax: +852 3442 0303; e-mail: bhpksl@ 5.57-fold lower, 95% confidence interval 4.95−6.18) of the total cityu.edu.hk. ciguatoxicities obtained by MNA, and the contributions were Notes generally comparable along the food chain. Even though the The authors declare no competing financial interest. present LC-MS/MS method can quantify only three individual P-CTXs (i.e., P-CTX-1, -2, and -3) instead of a complete suite, ■ ACKNOWLEDGMENTS these results indicated that a factor of 5.57, with application of This work was fully supported by grants from the National the above equation, can be potentially applied for interconver- Natural Science Foundation of China (41276110), China. We sion of ciguatoxicity determined by these two methods. fi thank Mr. Being Yeeting (Senior Fisheries Scientist, Secretariat However, further studies on the quanti cation of other CTXs of the Pacific Community, ) and the Fisheries and sodium channel activators/inhibitors (such as brevetoxins, Division (Ministry of Fisheries and Marine Resource Develop- saxitoxins, and tetrodotoxins), and the interaction of these ment, Republic of Kiribati) for facilitating sample collection. toxins (synergistic/antagonistic effects) may help to confirm its application and may assist in characterizating the metabolic ■ REFERENCES pathways of these toxins in food web components at various trophic levels. (1) EFSA Panel on Contaminants in Food Chain.. Scientific opinion To the best of our knowledge, this is the first attempt to on marine biotoxins in shellfish - Emerging toxins: Ciguatoxin group. EFSA J. 2010, 8, 1627−1665. clarify distribution, transfer, and bioaccumulation of individual (2) Lehane, L.; Lewis, R. J. Ciguatera: recent advances but the risk P-CTX-1, -2, and -3 in a ciguatoxic coral reef ecosystem by use remains. Int. J. Food Microbiol. 2000, 61,91−125. of LC-MS/MS. However, the biodiversity of the present coral (3) Kelly, A. M.; Kohler, C. C.; Tindall, D. R. Are crustaceans linked reef ecosystem is high and hence some important food web to the ciguatera food chain? Environ. Biol. . 1992, 33, 275−286. components such as benthic producers, microinvertebrates, (4) Bottein Dechraoui, M. Y.; Kashinsky, L.; Wang, Z. H.; Littnan, very large coral reef fishes (such as humphead wrasses), sea C.; Ramsdell, J. S. Identification of ciguatoxins in Hawaiian monk seals turtles, dolphins, and sharks could not be collected for the Monachus schauinslandi from the northwestern and main Hawaiian present study. It is necessary to conduct a comprehensive Islands. Environ. Sci. Technol. 2011, 45, 5403−5409. sampling scheme to confirm the sources, flux, and bioaccumu- (5) Meujo, D. A. F.; Hamann, M. T. Science, policy, and risk lation of individual P-CTXs in the ciguatoxic food web. As management. In Environmental Toxicology: Selected Entries from the Encyclopedia of Sustainability Science and Technology; Laws, E. A., Ed.; there is evidence indicating that food strategy and habitat use Springer Science + Business Media: New York, 2013; pp 461−501. are important parameters governing P-CTX levels in food web (6) Randall, J. E. A review of ciguatera, tropical fish poisoning, with a components, further study to characterize food web structures tentative explanation of its cause. Bull. Mar. Sci. 1958, 8, 236−267. (biodiversity, δ15N values of the baseline organisms, isotope (7) Yogi, K.; Oshiro, N.; Inafuku, Y.; Hirama, M.; Yasumoto, T. fractionation factor, etc.) of ciguatoxic and reference sites may Detailed LC-MS/MS analysis of ciguatoxins revealing distinct regional

14078 dx.doi.org/10.1021/es403175d | Environ. Sci. Technol. 2013, 47, 14070−14079 Environmental Science & Technology Article and species characteristics in fish and causative alga from the Pacific. (30) Ross, S. G. Preliminary report on fish poisoning at Fanning Anal. Chem. 2011, 83, 8886−8891. Island. Med. J. Aust. 1947, 2, 617−621. (8) Cruz-Rivera, E.; Villareal, T. A. Macroalgal palatability and the (31) Belotte, J. Les poissons “empoisonnes” du lagon Tahitien. Med. flux of ciguatera toxins through marine food webs. Harmful Algae Trop. 1955, 15, 232−236. 2006, 5, 497−525. (32) Harry, R. R. Ichthyological field data of Raroia Atoll, Tuamotu (9) Lewis, R. J.; Holmes, M. J. Origin and transfer of toxins involved Archipelago. Atoll Res. Bull. 1953, 18, 190. in ciguatera. Comp. Biochem. Physiol. 1993, 106C, 615−628. (33) Sano, M.; Shimizu, M., Nose, Y., Eds. Food Habits of Teleostean (10) Caillaud, A.; de la Iglesia, P.; Darius, H. T.; Pauillac, S.; Aligizaki, Reef Fishes in Okinawa Island, Southern Japan, University of Tokyo K.; Fraga, S.; Chinain, M.; Diogene, J. Update on methodologies Press: Tokyo, 1984. available for ciguatoxin determination: Perspectives to confront the (34) Leis, J. M. Diodontidae. Porcupine fishes (burrfishes). In FAO onset of ciguatera fish poisoning in Europe. Mar. Drugs 2010, 8, Species Identification Guide for Fishery Purposes. The Living Marine 1838−1907. Resources of the Western Central Pacific, 6th ed.; Carpenter, K. E., Niem, (11) Olsen, D. A.; Nellis, D. W.; Wood, R. S. Ciguatera in the eastern V., Eds.; FAO: Rome, 2001. caribbean. Mar. Fish. Rev. 1984, 46,13−18. (35) Kulbicki, M.; Bozec, Y. M.; Labrosse, P.; Letourneur, Y.; Mou- (12) Helfrich, P.; Banner, A. H. Experimental induction of ciguatera Tham, G.; Wantiez, L. Diet composition of carnivorous fishes from toxicity in fish through diet. Nature 1963, 197, 1025−1026. coral reef lagoons of New Caledonia. Aquat. Living Resour. 2005, 18, (13) Lewis, R. J.; Sellin, M.; Poli, M. A.; Norton, R. S.; MacLeod, J. 231−250. K.; Sheil, M. M. Purification and characterization of ciguatoxins from (36) Hiatt, R. W.; Strasburg, D. W. Ecological relationships of the fish moray eel (Lycodontis javanicus, Muraenidae). Toxicon 1991, 29, fauna on coral reefs of the Marshall Islands. Ecol. Monogr. 1960, 30, 1115−1127. 65−127. (14) Yasumoto, T.; Nakajima, I.; Bagnis, R.; Adachi, R. Finding a (37) Carro, N.; Carcia, I.; Ignacio, M.; Mouteira, A. Possible as a likely culprit of ciguatera. Bull. Jpn. Soc. Sci. Fish. influence of lipid content on levels of organochlonine compounds in 1977, 43, 1021−1026. mussels from Galicia coast (Northwestern, Spain). Spatial and (15) Chinain, M.; Darius, H. T.; Ung, A.; Fouc, M. T.; Revel, T.; temporal distribution patterns. Environ. Int. 2004, 30, 457−466. Cruchet, P.; Pauillac, S.; Laurent, D. Ciguatera risk management in (38) Myers, R. F. Micronesian Reef Fishes, 2nd ed.; Coral Graphics: French Polynesia: The case study of Raivavae Island (Australes Barrigada, Guam, 1991. Archipelago). Toxicon 2010, 56, 674−690. (39) Bohnsack, J. A.; Harper, D. E.; McClellan, D. B.; Sutherland, D. fi (16) Darius, H. T.; Ponton, D.; Revel, T.; Cruchet, P.; Ung, A.; L.; White, M. Resource survey of shes within Looe Key National Tchou Fouc, M.; Chinain, M. Ciguatera risk assessment in two toxic Marine Sanctuary. In NOAA Technical Memorandum Series NOS/ sites of French Polynesia using the receptor-binding assay. Toxicon MEMD, National Oceanic and Atmospheric Administration; U.S. 2007, 50, 612−626. Department of Commerce: Washington, 1987 (17) Chan, W. H.; Mak, Y. L.; Wu, J. J.; Jin, L.; Sit, W. H.; Lam, J. C. (40) Carassou, L.; Kulbicki, M.; Nicola, T. J. R.; Polunin, N. V. C. W.; Sadovy de Mitcheson, Y.; Chan, L. L.; Lam, P. K. S.; Murphy, M. Assessment of fish trophic status and relationships by stable isotope B. Spatial distribution of ciguateric fish in the Republic of Kiribati. data in the coral reef lagoon of New Caledonia, southwest Pacific. − 2011 − Aquat. Living Resour. 2008, 21,1 12. Chemosphere , 84, 117 123. fi (18) Dickey, R. W.; Plakas, S. M. Ciguatera: A public health (41) Roese, R.; Pauly, D. FishBase Website; http://www. shbase.org. perspective. Toxicon 2010, 56, 123−136. (19) Lewis, R. J. Ciguatera: Australian perspectives on a global problem. Toxicon 2006, 48, 799−809. (20) Mak, Y. L.; Wu, J. J.; Chan, W. H.; Murphy, M. B.; Lam, J. C. W.; Chan, L. L.; Lam, P. K. S. Simultaneous quantification of Pacific- ciguatoxins in fish blood using liquid chromatography tandem mass spectrometry. Anal. Bioanal. Chem. (in press). (21) Folch, J.; Lees, M.; Sloane Stanley, G. H. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 1957, 226, 497−509. (22) Jake Vander Zanden, M.; Cabana, G.; Rasmussen, J. B. Comparing trophic position of freshwater fish calculated using stable nitrogen isotope ratios (δ15N) and literature dietary data. Can. J. Fish. Aquat. Sci. 1997, 54, 1142−1158. (23) Cabana, G.; Rasmussen, J. Modelling food chain structure and contaminant bioaccumulation using stable nitrogen isotopes. Nature 1994, 372, 255−257. (24) Smith, C. D. Diet of Octopus vulgaris in False Bay, South . Mar. Biol. 2003, 143, 1127−1133. (25) Grubert, M. A.; Wadley, V. A.; White, R. W. G. Diet and feeding strategy of Octopus maorum in southeast Tasmania. Bull. Mar. Sci. 1999, 65, 441−451. (26) Hearn, A.; Murillo, F. C. Life history of the red spiny lobster, Panulirus penicillatus (Decapoda: Palinuridae), in the Galapagos Marine Reserve, Ecuador. Pac. Sci. 2008, 62, 191−204. (27) Jiang, W.; Meng, T.; Chen, R.; Wei, S. Diet of Charybdis japonica (A. Miline-Dedwards) and Portunus trituberculatus (Miers) in the Bohai Sea. Mar. Fish. Res. 1998, 19,53−59. (28) Saoudi, M.; Abdelmouleh, A.; El Feki, A. Tetrodotoxin: a potent marine toxin. Toxin Rev. 2010, 29,60−70. (29) Ngy, L.; Yu, C. F.; Taniyama, S.; Tajatani, T.; Arakawa, O. Co- occurrence of tetrodotoxin and saxitoxin in Cambodian marine pufferfish Takifugu oblongus. Afr. J. Libr. Arch. Info. 2009, 31, 349−354.

14079 dx.doi.org/10.1021/es403175d | Environ. Sci. Technol. 2013, 47, 14070−14079