Specific identification of western Atlantic Ocean scombrids using mitochondrial DNA cytochrome C oxidase subunit I (COI) gene region sequences

Identification of scombrids (tunas, mackerels, bonitos, etc.) is difficult when morphological characters are ambiguous or missing, such as with early life history stages or tissues found in the stomachs of predators. The mitochondrial cytochrome c oxidase subunit I (COI) gene region was evaluated as a molecular marker for the specific identification of the 17 members of the family Scombridae common to the western Atlantic Ocean. A 950 base pair region in the COI gene was sequenced from up to 20 individuals of each species, and suites of nucleotide polymorphisms that unambiguously distinguish among these scombrid species were identified. A shorter 250 base pair fragment of COI proved to be sufficient for species identification and was better suited for analyzing degraded tissue samples. Scombrid larvae collected in the Florida Straits and scombrid remains in the stomachs of large pelagic predators were used to demonstrate the utility of both the long and short COI fragments. Members of the family Scombridae (tunas, mackerels, bonitos, etc.) are important components of pelagic ecosystems, with several species supporting large commercial and recreational fisheries throughout the world’s oceans. proper identification of these species at all life stages and in various conditions, even as degraded stomach contents, is essential to better understand early life history characteristics and ecological relationships in the pelagic ecosystem, and to enable effective management. In addition, specific identification of processed tissues or fillets is necessary for enforcement of fisheries management regulations. while specific identification of adult scombrids is essentially unambiguous (Collette and Nauen, 1983), identification is problematic in situations where morphological characters are difficult to interpret (early life history stages) or missing (fillets, digested stomach contents). Identification of early life history stages of scombrids has been especially challenging. Scombrid eggs, larvae (especially those of the genus Thunnus), and juveniles generally cannot be distinguished unambiguously based solely on morphology (Richards et al., 1990). Molecular markers can provide a means for positive identification when morphological identification is uncertain. Various molecular markers have been used to identify fish eggs and larvae including allozymes (Morgan, 1975), polymerase chain reaction (pCR)/restriction fragment length polymorphism (RFLp) analysis (daniel and graves, 1994; Mcdowell and graves, 2002), multiplex pCR (Rocha-Olivares, 1998; hyde et al., 2005) and sequencing (hare et al., 1994; Kirby and Reid, 2001; perez et al., 2005). Many of these techniques have been used to identify scombrids. Allozymes have been successfully used to discriminate between early juveniles of bigeye tuna Thunnus obesus (Lowe, 1839) and yellowfin tuna Thunnus albacares (Bonnaterre, 1788) (graves et al., 1988) as well as between adult pacific northern bluefin tuna Thunnus thynnus orientalis Serventy, 1956 and southern bluefin tuna Thunnus maccoyii Castelnau, 1872 (ward, 1995). Several studies have used pCR/ BULLETIN OF MARINE SCIENCE, VOL. 80, NO. 2, 2007 354 RFLp analysis to identify species of the scombrid tribes Thunnini and Sardini (Chow et al., 2003) as well as eight species of the genus Thunnus (Chow and Inoue, 1993). In addition, sequencing of a mitochondrial gene region has been used to identify Thunnus species (Bartlett and davidson, 1991; Ram et al., 1996; quintero et al., 1998; Terol et al., 2002; ward et al., 2005). while each of these techniques has advantages and disadvantages, sequencing provides the highest level of resolution as it shows genetic differences at the nucleotide level. A few studies have used sequence analysis to identify scombrids, but these investigations were limited as they only distinguished between a few species, used a region that revealed considerable intraspecific variation, had limited sample sizes, or encountered problems with non-specific amplification (Bartlett and davidson, 1991; Ram et al., 1996; quintero et al., 1998; Terol et al., 2002; ward et al., 2005). Additionally, molecular techniques have only recently appeared promising as an answer for the difficulties associated with identifying degraded remains in stomachs (harper et al., 2005; gorokhova, 2006), and sequencing has been used to identify gut contents of marine invertebrates (Blankenship and yayanos, 2005) and large pelagic fishes (Smith et al., 2005). The mitochondrial genome has been preferred for analysis in many genetic studies as it has a high number of copies per cell, which facilitates pCR amplification, and because the presence of a single allele makes it possible to sequence products directly (Avise, 1994). Many mitochondrial gene regions (cytochrome b, Nd4, 16S, COI) have been successfully used for fish identification (Bartlett and davidson, 1991; Mcdowell and graves, 2002; hyde et al., 2005; Lopez and pardo, 2005). These gene regions display different levels of genetic variation as a result of different evolutionary rates. while variation is necessary to highlight interspecific differences, too much variation can hinder primer design. Because of this, the use of a conserved region is advantageous for effective amplification across many species. One of the most conserved protein coding genes in the mitochondrial (mt) genome is cytochrome c oxidase subunit I (COI) (Brown, 1985). COI is critical for cellular energy production and this functional importance constrains its evolution (Rawson and Burton, 2002). The high level of conservation of COI allows for the design of a unique primer pair that successfully amplifies the same fragment across the diverse members of the Scombridae. previous work has taken advantage of COI for broad taxonomic studies (11 invertebrate phyla, Folmer et al., 1994; 11 animal phyla, hebert et al., 2003), but COI has also been useful to distinguish closely related genera in species identification (three copepod genera, Bucklin et al., 1999). Efforts to use a segment of dNA as a barcode of identity have successfully employed COI to identify various taxa including fish species (Steinke et al., 2005; ward et al., 2005). Because COI is informative for distinguishing species across and within many different taxa, it is well suited for identification across a family as diverse as the Scombridae. In this study, a molecular key is developed based on the mitochondrial COI region for the specific identification of the 17 scombrids present in the western Atlantic Ocean. Materials and Methods Tissue samples were obtained from up to 20 specimens of each of the 17 scombrid species common to the western Atlantic Ocean: Acanthocybium solandri (Cuvier, 1831); Auxis rochei (Risso, 1810); Auxis thazard (Lacépède, 1800); Euthynnus alletteratus (Rafinesque, 1810); Katsuwonus pelamis (Linnaeus, 1758); Sarda sarda (Bloch, 1793); Scomber colias gmepAINE ET AL.: SCOMBRId IdENTIFICATION USINg COI SEqUENCES 355 lin, 1789; Scomber scombrus Linnaeus, 1758; Scomberomorus brasiliensis Collette, Russo and Zavalla-Camin, 1978; Scomberomorus cavalla (Cuvier, 1829); Scomberomorus maculatus (Mitchill, 1815); Scomberomorus regalis (Bloch, 1793); Thunnus alalunga (Bonnaterre, 1788); Thunnus albacares (Bonnaterre, 1788); Thunnus atlanticus (Lesson, 1830); Thunnus obesus (LOwE, 1834) and Thunnus thynnus (Linnaeus, 1758). All specimens were identified based on morphological characters (Collette and Nauen, 1983). Tissue samples were either stored in dMSO buffer (Seutin et al., 1991) or frozen. published COI sequences of A. thazard and A. rochei (Infante et al., 2004) were used to supplement the number of samples for these species. Collection information is provided in Table 1. To evaluate the efficacy of COI as a marker to identify scombrids, specimens of larval scombrids stored in ethanol were obtained from d. Richardson and R. Cowen, Rosenstiel School of Marine and Atmospheric Science, University of Miami. In addition, stomach content samples containing putative scombrids were collected from blue marlin and white marlin caught in recreational fishing operations out of Cape May, NJ, USA and La guaira, Venezuela. putative scombrids were removed from the marlin stomachs dockside and rinsed with water. Either a muscle sample was removed and placed in dMSO buffer (Seutin et al., 1991) or the whole fish was frozen until analysis. Total genomic dNA was extracted from adult tissues of known scombrid species using a standard phenol/chloroform isolation protocol (modified from Sambrook and Russell, 2001). A series of extractions was performed on each sample using equilibrated phenol, followed by phenol: chloroform: isoamyl alcohol (25:24:1) and finally, chloroform: isoamyl alcohol (24:1). Following extraction, dNA was precipitated with ethanol. For larval fishes, one eyeball (right eyeball when available) was removed and rinsed with distilled water. dNA was extracted from this tissue using proteinase-K and Chelex beads (Bio-Rad Laboratories, hercules, CA) (Estoup et al., 1996). Each larva was photographed using a digital camera attached to a stereomicroscope via a phototube, capturing as much detail as possible for future morphological or meristic analysis. primers that amplify the COI gene region across the scombrid family were designed using conserved regions of seven scombrid COI sequences (A. rochei, A. thazard, E. alletteratus, K. pelamis, S. scombrus, T. alalunga, and T. thynnus) available through genBank (National Center for Biotechnology Information). Two sets of primers were developed that amplify a ~950 base pair (bp) fragment (long fragment) of the COI gene and a ~250 bp fragment (short fragment) located within the 950 bp fragment: 950 bp fragment: LCOI 121 CTA AgC CAA CCA ggT gCC CTT CT hCOI 1199 AAT AgT ggg AAT CAg TgT ACg A 250 bp fragment: LCOI 646 AAT ACA ACC TTC TTC gAC C hCOI 947 gTT ggA ATT gCg ATA ATC (The number in the primer