Luthy et al. Identification of larval sailfish, white marlm, and blue marlin in the western North Atlantic Ocean 



589 



vae may also be identified by the presence of a complex 

 lateral line. Ueyanagi (1964) found this character in 

 Pacific blue marlin of 20 mm standard length (SL), but 

 the smallest SL of an Atlantic blue marlin from a recent 

 collection in which a complex lateral line was visible 

 was 26.9 mm. At lengths <20 mm, specific identifica- 

 tion of istiophorids is even more uncertain. Ueyanagi 

 (1963; 1964) based the identification of Indo-Pacific 

 istiophorids <5 mm SL on four characters: 1) anterior 

 projection of the eye orbit; 2) the position of the tip of 

 the snout in relation to the middle of the eye; 3) pres- 

 ence of pigments on the branchiostegal and gular mem- 

 branes; and 4) whether the pectoral fins are rigid — a 

 character that applies to larval black marlin tMakaira 

 indica), a species not known to spawn in the Atlantic 

 Ocean. For fish >5 mm SL, the characters of relative 

 snout length and eye size are used. Ueyanagi (1964) 

 described sailfish, striped marlin (Tetrapturus audax, 

 the Pacific counterpart to white marlin), and shortbill 

 spearfish {Tetrapturus angustirostris) between 10 and 

 20 mm SL as having long snouts. The short snout group 

 comprised blue marlin and black marlin. The angles at 

 which the pterotic and preopercular spines protrude 

 from the body have also been useful in identifying Indo- 

 Pacific specimens (Ueyanagi, 1974a). 



A troubling aspect of current larval istiophorid iden- 

 tification methods is the difficulty in using some of the 

 above characters. If a specimen is fixed with its mouth 

 open, snout position with respect to eye is an unread- 

 able character (Richards, 1974), and misidentifications 

 can occur (Ueyanagi, 1974a). Evaluation of certain char- 

 acters (e.g., whether the eye orbit projects anteriorly) 

 can be highly subjective. The lack of confirming identi- 

 fication characters compounds the problem; if just one 

 character cannot be assessed, identification may not 

 be possible (Richards, 1974). An additional problem is 

 the apparently high variability in characters such as 

 pigment locations and head spine angles in Atlantic 

 istiophorids (Richards, 1974). 



Most of the larval specimens examined by Ueyanagi 

 came from the Indo-Pacific; he assumed that the same 

 identification characters would apply to their Atlantic 

 counterparts (Ueyanagi, 1963, 1974a). Although recent 

 genetic evidence supports Morrow and Harbo's (1969) 

 opinion that Atlantic and Indo-Pacific sailfish are actu- 

 ally populations of a global species (Finnerty and Block, 

 1995; Graves and McDowell, 1995), morphological dif- 

 ferences have been noted in sailfish, especially at 90 

 cm. Specifically, the pectoral fin is longer, in relation 

 to the body, in Atlantic sailfish than in Indo-Pacific 

 sailfish. Differences in the spread of the caudal fin and 

 maximum total length have also been observed. These 

 characters were the impetus behind the separation of 

 sailfish, at least to subspecies, by ocean basin (Naka- 

 mura, 1974). Regardless of the taxonomic status of the 

 Atlantic and Indo-Pacific billfishes, physical attributes 

 of istiophorid species may vary by region. Therefore, the 

 assumption that the larvae of Atlantic istiophorids can 

 be identified by using the same characters attributed to 

 Indo-Pacific istiophorids may not be valid. 



Billfishes are not the only group whose larval iden- 

 tification has proven difficult. Species of the genus 

 Sebastes, the rockfishes, have some morphological and 

 pigmentation differences as larvae, but identification 

 was difficult and uncertain until genetic methods were 

 employed (Rocha-Olivares et al., 2000). Fulford and 

 Rutherford (2000) solved a similar problem by combin- 

 ing allozyme analysis of larval tissues with landmark- 

 based morphometries to distinguish between species of 

 the genus Morone. In each study, a molecular technique 

 was used to confirm larval species identity, facilitat- 

 ing the development of morphometric identification 

 techniques. 



Several molecular methods for identifying adult 

 billfishes have been developed (Chow, 1993; Innes et 

 al., 1998; McDowell and Graves, 2002). In the present 

 study, larval istiophorids from Atlantic waters were 

 identified to species using restriction fragment length 

 polymorphism (RFLP) analysis of a 1.2-kb segment 

 of nuclear DNA, as described for adult billfishes by 

 McDowell and Graves (2002). In this article we pres- 

 ent data for genetically identified istiophorid larvae, 

 analyses of morphometric and qualitative characters, 

 and a key for the identification of larval istiophorids of 

 the Straits of Florida and the Bahamas. 



Materials and methods 



Larval material 



Larval istiophorids were collected between June 1998 

 and April 2002 from the Straits of Florida and Exuma 

 Sound, Bahamas. Several preservation fluids were used, 

 but the majority of the larvae (-1000) were preserved 

 in 70-95% ethanol. Butylated hydroxytoluene (BHT) 

 saturated ethanol was used to preserve 150 larvae. 

 Approximately 300 larvae were fixed in 10% unbuffered 

 formalin and then transferred to 70% ethanol. In the 

 laboratory, each fish was assigned a unique identification 

 number and stored separately. 



Molecular identification 



Total DNA was extracted from the right eyeball of each 

 larva, using either a quick-digest method (Ruzzante et 

 al., 1996) or a standard high-molecular weight DNA 

 extraction protocol (Sambrook et al., 1989). Larval 

 identification was achieved by PCR amplification of 

 the nuclear locus MN32-2 (Buonaccorsi et al., 1999), 

 and subsequent RFLP analysis (restriction endonucle- 

 ases Dra I and Dde I, Life Technologies, Bethesda, 

 MD). If the restriction fragment pattern (Fig. 1) of a 

 larva matched one of those described for a known-iden- 

 tity adult, the larva was assigned to that species. See 

 McDowell and Graves (2002) for detailed protocols and 

 reaction parameters. Preliminary attempts to amplify 

 DNA from formalin-fixed larvae failed; only ethanol- 

 preserved specimens were used in subsequent molecular 

 work. 



