from istiophorids since the former lack the strong 

 pterotic and preopercular spines which are so prom- 

 inent in the latter in the early stages. In sizes over 20 

 mm, the young are very dissimilar in appearance. 

 The identification problems lie within the is- 

 tiophorids. Ueyanagi (1964) has summarized the 

 present methods used to identify young stages of 

 istiophorids from the Indo-Pacific Ocean. No pa- 

 pers have appeared as yet distinguishing all the 

 species of the Atlantic from one another. One major 

 problem with this group is that meristic characters 

 are not particularly useful. The full complement of 

 fin rays does not appear until the young are at least 20 

 mm in length and, as I have shown in Table 1, the 

 counts exhibit little interspecific differences with 

 overlap in range of nearly every character. Only the 

 swordfish is separable on vertebral numbers (26 ver- 

 tebrae compared with 24 for istiophorids). The genus 

 Makiiira has 11 precaudal and 13 caudal vertebrae, 

 whereas Istiophonis and Tetiaptiinis have 12 pre- 

 caudal and 12 caudal vertebrae. This character is 

 difficult to use with specimens less than 20 mm in 

 length. The only other meristic character (with the 

 obvious exception of the pelvic rays, since they are 

 lacking in swordfish) of any use is the numberof first 

 dorsal rays. This will separate some species from 

 each other, but there is sufficient overlap so that the 

 numberofrays alone cannot be used. Forexample, a 

 specimen with a count of 42 could not be T . angiis- 

 tirostris or T. pfluegeri, but it could be any of the 

 others. Therefore, first dorsal counts are only useful 

 to eliminate some species. 



I have reproduced here Ueyanagi's methods for 

 separating the Indo-Pacific species of istiophorids as 

 follows (I have changed his names to conform with 

 present practices): 



"It is not easy to identify the larvae of different 

 istiophorid species, because of their close resem- 

 blance with each other and of marked difference 

 from their respective adults, generally speaking, 

 in their morphological characteristics. This is 

 particularly true with those of very early stage 

 before the snout develops its specific characteris- 

 tics. However, the specific separation of the lar- 

 vae is possible throughout their entire range 

 mainly on the basis of their head profile. 



"Following are the criteria for identification: 



"(1) Larvae under 5 mm in length: The characters, 

 as shown in Table [2], can be used for specific 

 separation, although snout length does not pro- 

 vide a useful clue. 



"(2) Larvae between 5 and 10 mm in length: Be- 

 sides the criteria given in Table [2], snout length 

 and size of eyes can be used. [M . mazara] larvae 

 are recognized by their short snout. The ratio of 

 snout length to diameter of orbit is largest in [/. 

 pUilyptenis]. smallest in [M. nuizara], and is be- 

 tween in [7". angiistirostris]. More precisely, the 

 ratio tends to be > 1 in [I . platypterus], < 1 in [M. 

 nuiztirci], and = 1 in [T. angustirostris] in speci- 

 mens 7-8 mm length. 



"(3) Larvae between 10 and 20 mm in length: They 

 are grouped into two on the basis of their snout 

 length; the long snout group with [T . 

 angtistirostris], [I. platypterus]. and [T. audax], 

 and the short snout group with [M. mazura] and 

 [M. indica]. In the former, the snout length ex- 

 ceeds V.i of their body length, while in the latter, it 

 does not. For the specific separation of the former 

 group. Table [2] applies: [T. ani>ti.stirostris] is dis- 

 tinguishable by black chromatophores on bran- 

 chiostegal membrane, while [/. platypterus] is 

 separated from [T . auda.\] by the difference of 

 their head profile: Unlike [7". audax] with a 

 straight snout, [/. platypterus] has a beak-like 

 snout. And because of this difference in the shape 

 of the snout, they are separable by the difference 



Table 1. — Meristic characters of adult billfishes based on data compiled from Nakamura 



at al. (1968) and Merrett (1971). 



65 



