KEENE AND TIGHE: BERYCIFORMES 



389 



Fig. 207. (A) Preflexion larva of Scopeloberyxsp.. 4.4 mm XL; (B) Postflexion larva of Scopeloberyx opisthopterus. 9.1 mm SL; (C) Postflexion 

 larva of Scopeloberyx robuslus. 1 3.0 mm SL; (D) Preflexion larva of Poromitra crassiceps complex, 7.9 mm SL; (E) Postflexion larva of Melamphaes 

 lepnis. 19.5 mm SL; all drawn by B. Washington. 



veioping. The dorsal, anal and pectoral fins are already devel- 

 oped, while pelvic fin buds are present. The pattern of head 

 spination described below is already well fisrmed. A 6.0 mm 

 postflexion specimen (Fig. 208B) has all fins completely devel- 

 oped except for the pelvics and procurrent caudal elements. 

 There is pigmentation on the head, lateral surface of the body 

 and caudal peduncle, while the abdominal area is pale with 

 scattered melanophores. A small pigmented area occurs on the 

 pectoral bases. A serrate frontal ridge bordenng the anterior of 

 each eye terminates in a short stout supraocular spine. Ridges 

 continuing posteriodorsad on the cranium terminate in long 

 serrate spines probably arising from the parietals. The pre- 

 opercles end in strong serrate spines directed posterioventrad. 

 By 9.0 mm SL, the pelvics have become well-developed and 

 the head spination is still strong. A small dense patch of me- 

 lanophores occurs on the ventral surface of the body justantenor 

 to the origin of the pelvic fins. With increasing growth (28 mm 

 SL), this pigmentation darkens and expands, extending forward 

 in a continuous band to the tip of the isthmus. Additional pig- 

 mentation occurs at the ongin of the pelvic fins, around the 

 vent, just posterior to the anal fin on the caudal peduncle, and 

 in a transverse bar on the abdomen midway between the pelvic 

 origin and the vent. The increase in dark pigmentation and the 

 reduction in cranial and preopercular spines in larger juveniles 

 is described by Woods and Sonoda (1973). 



Crossland (1981) illustrated a trachichthyid larva, probably 

 of OpliYus I'longatus. taken off northeastern New Zealand (Fig. 

 208C). Larger larvae of the same species had the skin on the 

 dorsal surface of the head and body covered with tiny spines. 

 Ahlstrom (notes) sketched early Trachichthys mento larvae that 

 are fairly deep-bodied at 3.5 to 4.5 mm, with the pectoral fin 

 showing precocious development. A dark spiny pigmented band 

 extending from the region of the anal to the dorsal occurs in 3.5 

 mm TL specimens. This spination covers areas on both sides 

 of the dorsal fin, parts of the head, thoracic region and jaws. In 

 a preflexion 6.4 mm specimen, the fin rays are mostly developed, 

 and the body is stockier, approaching the shape of the adult and 

 is covered with minute spines. The holotype of Korogaster nanus 

 Parr 1933, synonomized by Woods and Sonoda (1973) in Ho- 

 plostethus. is 19 mm long (Fig. 208D), possesses unbranched 



rays in the pectoral, pelvic and caudal fins, and has dermal 

 papillae and small spines all over its body. This specimen and 

 the second specimen of Korsogaster Tepone:<i by Johnson (1970) 

 (Fig. 208E) are juveniles of the family Trachichthyidae. 



The most striking early life history of any beryciform is ex- 

 hibited by the prejuvenile kasidoron stage of gibberichthyids 

 (Figs. 209, 210). This stage is characterized by a long trailing 

 pelvic appendage which is part of a modified third pelvic ray 

 and is present in specimens from at least 7.5 to 21 mm TL. It 

 is lost by 30 mm SL (de Sylva and Eschmeyer, 1977). Dunng 

 early growth, this trailing appendage becomes more ornate and 

 resembles the trailing tentacles of siphonophores or Sargassum 

 weed at about 15 mm SL. Up until about at least 20 mm, the 

 prejuveniles inhabit epipelagic waters but by 30 mm individuals 

 have lost the pelvic appendage and taken up a mesopelagic to 

 upper bathypelagic existence. The anterior dorsal and anal fin 

 elements are soft rays during the kasidoron stage, but develop 

 into strong fin spines in the adult. There is also a marked de- 

 velopment of bony head ridges in the adults, that is not found 

 in the stages 20 mm and smaller (de Sylva and Eschmeyer, 

 1977). 



Relationships 



Rosen and Patterson (1969) and Rosen (1973) emphasized 

 the futility of the present classification of the Beryciformes and 

 the rest of the Acanthopterygii, because it relies on grouping of 

 primitive characters to express relationships. Realizing this. 

 Zehren (1979) did a phylogenetic analysis of the Beryciformes 

 to attempt to determine whether or not the order is monophy- 

 letic(Fig. 211). 



Besides supporting Rosen and Patterson's removal of the 

 Polymixiidae from the Beryciformes, Zehren's analysis super- 

 ficially suggests that the remaining ten families form a mono- 

 phyletic group. However, he cautions that since none of the 

 derived character states that he uses is unique to the ten families, 

 their monophyly is uncertain. 



Zehren's results and discussion suggest that the Holocentridae 

 do not appear to be closely related to the other nine families 

 studied. Woods and Sonoda (1973) felt that the Holocentridae 

 were very different from the other Beryciformes and Rosen (1973) 



