94 



Fishery Bulletin 93(1), 1995 



and Trichiuridae (Cuvier and Valenciennes, 1832; 

 Gosline, 1968; Potthoffet al., 1986). 



How can these inferences from molecular data be 

 reconciled with the morphological data? We believe 

 that this is an instance where molecular data comple- 

 ment morphological data well. Cytochrome b provides 

 an unambiguous phylogenetic signal that billfishes 

 are genetically distant from other scombroids. In 

 contrast, the existing morphological data does not 

 clearly discriminate between a number of hypoth- 

 eses. The number of character reversals in morpho- 

 logical phylogenies that classify billfishes as scom- 

 broids indicates that there have been many ho- 

 moplastic changes in the billfish lineage. According 

 to the morphological evidence, either billfishes are 

 scombroids and have undergone several reversals to 

 the primitive state, such as their low number of ver- 

 tebrae, or billfishes are not scombroids but have 

 evolved many convergent similarities to scombroids, 

 such as their paired lateral caudal keels. 



Many of the morphological characters that unite 

 billfishes to other scombroids, particularly to Scom- 

 bridae, may be adaptations for continuous swimming, 

 and are therefore of questionable phylogenetic value. 

 These include hypurostegy, the projection of the cau- 

 dal fin-ray bases anteriorly to cover the hypurals (Col- 

 lette et al., 1984; Johnson, 1986), fusion of the hypurals 

 (Collette et al., 1984; Johnson, 1986), and inter- 

 filamentar gill fusion (Johnson, 1986). Hypurostegy 

 and interfilamentary gill fusion are known to have 

 evolved convergently in nonscombroid taxa (Luvarus 

 imperialis [Leis and Richards, 1984]; and Amia calva 

 [Bevelander, 1934]). The molecular data presented 

 here provide a phylogenetic signal that is indepen- 

 dent of convergent morphological adaptations that 

 might confound phylogenetic analysis. There has been 

 convergent evolution in the molecular characters, but 

 unlike many of the morphological characters men- 

 tioned, this convergent evolution does not appear to be 

 the result of strong selection: most amino acid substi- 

 tutions exchange amino acids with similar size, charge, 

 and degree of polarity. Therefore, when compared with 

 the existing morphological data, the phylogenetic sig- 

 nal in the molecular data is less likely to have been 

 obscured by similar selective pressures acting upon 

 distantly related lineages. 



Istiophorid phylogeny 



Historically, there have been numerous disagree- 

 ments over the number of species within the 

 Istiophoridae and their interrelationships (Goode, 

 1882; Jordan and Evermann, 1926; LaMonte and 

 Marcy, 1941; Nakamura, 1983). This is evidenced by 

 the synonymies for many istiophorids, e.g. the Medi- 



terranean spearfish, Tetrapturus belone, has also 

 been assigned to Istiophorus (Ben-Tuvia, 1953) and 

 Makaira (Tortonese, 1958). The most thorough treat- 

 ment of billfish systematics to date is a phenetic 

 analysis conducted by Nakamura (1983). Nakamura 

 recognized 11 species of istiophorid billfishes in three 

 genera, including the designation of separate Atlan- 

 tic and Indo-Pacific species for blue marlin (Makaira 

 nigricans and M. mazara) and sailfish (Istiophorus 

 albicans and /. platypterus). 



The molecular evidence presented here agrees with 

 Nakamura ( 1983) in supporting the monophyly of the 

 genus Tetrapturus, and within this genus, clades con- 

 sisting of audax + albidus and pfluegeri + angus- 

 tirostris + belone. Cytochrome b does not support the 

 recognition of separate Atlantic and Pacific species 

 of blue marlin and sailfish. Previous results (Finnerty 

 and Block, 1992) identified substantial overlap in the 

 cytochrome b haplotypes found among Atlantic and 

 Pacific populations of blue marlin. The sailfish 

 sample in this study includes one Atlantic specimen 

 and one Pacific specimen that differ at only two sites 

 among 590 (0.3%). We infer from the cytochrome b 

 data (Block et al., 1993; and this study) a nonmono- 

 phyletic Makaira and support for a clade consisting 

 of the blue marlin (Makaira nigricans) and the sail- 

 fish (Istiophorus platypterus). 



Based on the cytochrome b data, istiophorid tax- 

 onomy at the generic level is not concordant with 

 phylogeny. It is premature to suggest taxonomic re- 

 vision of istiophorid genera, but we believe it is im- 

 perative to obtain more molecular data, particularly 

 from nuclear genes, to determine whether the infer- 

 ences presented here can be corroborated. Further- 

 more, we recognize the need for an extensive cladis- 

 tic analysis of istiophorid relationships based on ad- 

 ditional morphological data. Another taxonomic is- 

 sue raised by this study concerns the number of valid 

 Tetrapturus species. An extensive genetic survey of 

 several populations from each species is required to 

 determine the number of evolutionarily independent 

 or reproductively isolated lineages within this genus. 



Relationships within the genus Thunnus 



The systematics of the genus Thunnus have been well 

 studied owing to the commercial importance of tu- 

 nas and interest in physiological specializations as- 

 sociated with the evolution of endothermy Collette 

 (1978) suggested a taxonomic subdivision of the ge- 

 nus reflecting a split between tropical species (sub- 

 genus Neothunnus: blackfin tuna, Thunnus atlan- 

 ticus, longtail tuna, Thunnus tonggol, andyellowfin 

 tuna, Thunnus albacares) and species that inhabit 

 cooler waters (subgenus Thunnus: bluefin tuna, 



