264 



ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM 



Fig. 137. Euclichlhys polynemus, holotype. From McCulloch (1926). 



line after fusion of the neural spine and at least some of their 

 evolution is in parallel with the gadids. 



Palaeogadits is a well-known Eocene fossil genus in which the 

 V-shaped crest has been retained, but specializations include a 

 joined first neural spine and only two hypurals. It is, in fact, 

 very similar to modem Merluccius. DaniPchenko (1950), who 

 reviewed Palaeogadus. believed that it gave rise independently 

 to Lotinae and Gadinae as well as to Merluccius. 



Phycinae, as recently modified by Markle (1982), is presently 

 included in the family Gadidae. Fahay and Markle (this volume) 

 would like to escort it out. An early Oligocene fossil genus, 

 Eophycis (Jerzmanska, 1968) has been suggested as a precursor 

 of Phycts and Urophycis, and probably arose independently of 

 other gadid subfamilies, which supports Fahay and Markle's 

 position. 



Lotinae is a gadid subfamily that I mainly leave to Fahay and 

 Markle and Dunn and Matarese. I note, however, Mujib's ( 1 967) 

 conclusion based on cranial osteology that Lotinae could have 

 arisen from Merlucciinae. Lotines have no V-shaped crest but 

 retain X-Y bones. Hypurals are two, the first neural spine is 

 joined to the occipital crest, and there are more branched rays 

 than in any other gadid. 



Gadinae has about a dozen genera, all of which have three 

 external dorsal and two external anal fins and a large caudal, 

 even though there are only two hypurals. Derived characters 

 include fused frontals, absence of X-Y bones, and a joined neu- 

 ral spine; Fahay and Markle and Dunn and Matarese (this vol- 

 ume) give more. 



Merluccius, with about a dozen closely related species (Inada, 

 1981b), has been treated as the type of a separate family or as 

 a subfamily of Gadidae. Primitive characters include a V-shaped 

 ridge and X-Y bones. Advanced ones are the joined first neural 

 spine and the reduced number of hypurals. Merluccius appears 

 to be the modem representative of a lineage commencing with 

 "Protocodus" and extending through Rhinocephalus and Pa- 

 laeogadus. which it closely resembles (Rosen and Patterson, 

 1969). 



Macruronus, which has three nominal species found in tem- 

 perate waters of the southern hemisphere, is basically a Mer- 

 luccius wiih a much reduced caudal fin. I mention it here because 

 it has been referred incorrectly to Macrouridae and considered 

 by some to be a link between that family and Merluccius. 



Lyconus. with several pelagic oceanic species, is probably re- 



lated to Merluccius. It lacks a caudal fin and has a single dorsal 

 fin. 



Steindachneria, is a monotypic tropical western Atlantic ge- 

 nus with luminescent organs, a wide separation between the 

 anus and urogenital openings, and no caudal fin. It has been 

 placed in Macrouridae and also considered a separate family 

 (Marshall and Cohen, 1973). It may be closer to Merluccius 

 than to any other known gadiform. 



Classification 



How best to classify gadiforms for working purposes in a way 

 that approximates their possible phylogenetic relationships is 

 diflicult because the existence of fossils, which appears to help 

 indicate lineages, creates logical traps for the classifier. The fol- 

 lowing arrangement, unfortunately based on gaps for some groups 

 and on a continuum for others, is an interim suggestion for 

 further testing. 



Euclichthys is accorded family status for the first time because 

 it can not be placed in any gadiform family. Gadidae is restricted 

 to the gadines, and Lotidae and Phycidae are recognized at the 

 full family level (family group names for the latter two date at 

 least from Goode and Bean, 1 896), because available evidence 

 indicates an independent origin from Palaeogadus for each of 

 the three groups. If merlucciids were reduced to subfamily rank 

 and placed with gadines, lotines, and phycines in a more inclu- 

 sive family Gadidae, then consistency would require the inclu- 

 sion of at least two other well-defined apparent derivatives of 

 the Rhinocephalus- Palaeogadus- Merluccius stem, Moridae and 

 Melanonidae. In the present instance I believe that splitting is 

 more useful than lumping. 



Suborder Muraenolepoidei 

 Family Muraenolepidae 

 Suborder Bregmacerotoidei 

 Family Bregmacerotidae 

 Suborder Macrouroidei 

 Family Euclichthyidae 

 Family Macrouridae 

 Subfamily Macrouroididae 

 Subfamily Trachyrincinae 

 Subfamily Macrourinae 

 Subfamily Bathygadinae 

 Suborder Gadoidei 



