DUNN AND MATARESE: GADIDAE 



299 



Phycinae 



Gadinae 



Raniceps 



/ 



Fig. 152. Proposed relationships of gadid subfamilies. 



The ventral branch of the posttemporal is shortest in lotines 

 and phycines (moderately long in Merluccius) and longest in 

 gadines. The gadids we examined all have four pectoral radials, 

 except for a single specimen of Ciliata with five radials on one 

 side. Phycines lack an expanded distal head on the postcleithra. 

 Raniceps. however, has two postcleithra; the upper is oblong in 

 shape, the lower is distally pointed as in phycines. Among the 

 lotines, the distal end of the postcleithrum is slightly expanded 

 in Brosme and Molva whereas the postcleithrum is distally 

 pointed in Lota. In Merluccius it is moderately expanded while 

 in gadines, the postcleithrum is considerably expanded at its 

 tip, which we infer is a derived condition. Predorsal bones are 

 present in some, but not all, phycines (Urophycis. Phycis. and 

 Raniceps). but are lacking in lotines, gadines, and Merluccius: 

 the loss is considered an advanced state. The posterior process 

 of the basipterygia is quite long in some phycines (and in Mer- 

 luccius), moderately long in lotines, and shortest in gadines, and 

 the latter state is considered derived. 



The shape of the neural and haemal spine on PU, varies 

 among genera. The neural spine in Raniceps is distally rounded 

 (a primitive condition), but this spine is flattened in all other 

 gadids, as it is in Merluccius. Raniceps. Brosme, and Molva 

 have a rounded haemal spine on PU,, in contrast to the flattened 

 tip on all other gadids (and Merluccius): x/y bones are present 

 in all phycines (usually present in Raniceps) and Merluccius, 

 but are absent {Brosme) or usually absent {Molva, Lota) in 

 lotines (Markle, 1982; this study) and are absent in gadines. All 

 gadids and Merluccius (Ahlstrom and Counts, 1955; Inada. 

 1981b) have three hypural bones (including the parhypural); 

 Raniceps alone, among the gadids examined by us, showed evi- 

 dence of ontogenetic reduction by fusion from six hypural bones 

 to three. As noted by Markle ( 1 982) and Fahay and Markle (this 

 volume), lotines and gadines have four or five primary caudal 

 fin rays, while phycines have five or six such rays. Merluccius 

 and Raniceps each have six primary rays (Inada, 1981b; this 

 study). 



We consider Raniceps a basal gadid considering the following 

 characters: eggs small with a single oil globule; larvae tadpole- 

 or liparid-shaped; one branchiostegal ray on the epihyal; two 

 postcleithra present; a predorsal bone present; the neural and 

 haemal spines on PU, distally rounded; six hypural bones which 

 fuse into three during ontogeny; x/y bones usually present; and 

 six primary rays on the superior hypural bone. 



We further consider phycines to be a more primitive group 

 than lotines based on the following characters: eggs small, with 

 multiple oil globules which coalesce into one during develop- 

 ment; larvae stocky and deep bodied (at the pectoral fin base); 

 elongate and precocious pelvic fins present; postcleithrum with- 

 out an expanded head; one or more predorsal bones present; 

 elongate pelvic process; and x/y bones present. Until the pres- 

 ence or absence of transient pterotic spines is established in all 

 phycine larvae, the most parsimonious explanation is that their 

 presence represents a derived character state. 



Lotines, as presently constituted, appear to us to possess a 

 number of primitive and intermediate characters, as well as 

 some rather specialized traits: eggs moderately large with a single 

 oil globule; larvae elongate, relatively shallow at the pectoral fin 

 base; pelvic fins precocious, elongate and with the posterior 

 process of the basipterygium moderately long; postcleithnam 

 with slightly expanded head; predorsal bones absent; x/y bones 

 usually absent; and three hypural bones present. Brosme has 

 both apparently primitive (e.g., all branchiostegal rays carried 

 on the outside surface of the ceratohyal and a rounded haemal 

 spine on PU,) and derived characters (e.g., x/y bones always 

 lacking); its single dorsal fin was considered primitive by Sve- 

 tovidov (1948) or derived (within Lotinae, sensu Svetovidov, 

 1948) by Mujib (1969). As noted by Markle (1982), high total 

 dorsal and anal fin ray counts may be primitive for the order 

 Gadiformes. 



Gadines seem to us a relatively homogenous group, charac- 

 terized by reductive (or lost) and apparently derived characters. 

 The former include: eggs without an oil globule; posterior pro- 

 cess of the pelvic bone reduced in length or wanting; predorsal 

 and x/y bones absent; and three hypural bones present. The 

 latter characters include: eggs moderate in size; larvae morpho- 

 logically uniform in appearance; lower branch of posttemporal 

 relatively long; postcleithrum with expanded head; and three 

 dorsal and two anal fins present, with the anal fins and dorsal 

 fins two and three internally continuous. 



Our hypothesis of relationships of gadid subfamilies is pre- 

 sented in Figure 152. The relationships of a number of genera, 

 such as Brosme and Raniceps. and the relationships of Phycis, 

 Gaidropsarus, and Ciliata to other phycines still remain con- 

 fused. Based on early life history characteristics and osteology, 

 we consider Merluccius a gadid related to, but more primitive 

 than. Gadinae and, following Svetovidov (1948, 1969), restrict 

 Merlucciinae to this genus. The relationship of all nominal gadid 

 subfamilies requires further study. 



Northwest and Alaska Fisheries Center, 2725 Montlake 

 Boulevard East, Seattle, Washington 981 15. 



