96 



Fishery Bulletin 88(1), 1990 



Figure 1 



Distriliution ami movenit'iits of hakes as proposed by Inada (1981) 

 (—) and Kabata and Ho (1981) ( «). 



body parts and counts for nine meristic characters. 

 Moreover, he added osteological information for 14 

 groups of bones, and discussed the distribution, biol- 

 ogy, and fishery of ail the species recognized. It is the 

 most complete revision ever published on the genus 

 Merluccius. 



The morphological differences and diagnostic char- 

 acters of the 12 species of hake were succinctly listed 

 by Inada (1981, tables 37 and 38). Because quantitative 

 variables are difficult to analyze cladistically (Pimentel 

 and Riggins 1987), data in table 37 of Inada (1981) were 

 not included in the present analysis. Thus, only the 

 osteological information summarized in his table 38 was 

 used. 



Steindachneria is a problematical gadiform genus 

 that has been treated as a merluciid or as a monotypic 

 family with uncertain affinities. Recently, Fahay (1989) 

 presented evidence from ontogenetic and osteological 

 grounds to support the removal of the genus from the 

 Merlucciidae. Nevertheless, in his report on the sys- 

 tematics of Merlucciidae, Inada (1989) concluded that 

 the family contains four genera in two subfamilies, 

 Steindachneriinae (Steindachneria) and Merlucciinae 

 (Lyconus, Macruronus, Merluccius). This treatment is 

 in line with Nolf and Steurbaut's (1989) result based 

 on analysis of otolith features. Therefore, I included 

 Steindachneria in my analysis of outgroup genera to 

 determine the sister group of Merluccius. 



The osteological information given in Inada's (1989) 

 Table 1 was used in my search for outgroup genera. 

 Lotinae was selected as the outgroup to polarize the 

 character states of the four merlucciid genera, because 

 both the works of Dunn (1989) and Nolf and Steurbaut 

 (1989) indicate that it is the nearest sister group of 

 Merlucciidae. Since six of the thirteen enumerated 

 characters in table 1 of Inada (1989: 199) show no dif- 



ferences in their states among the five genera under 

 consideration, they were excluded from the present 

 analysis. The BRANCH AND BOUND algorithm from 

 the phylogenetic computer package PAUP (version 

 2.4.1, David L. Swofford, 111. State Nat. Hist. Surv., 

 Urbana 61820), which guarantees the finding of all the 

 most parsimonious trees, was used in this analysis. Two 

 trees were obtained, both with a consistency index of 

 0.667. The one with the lower F value (= 0.174) is 

 selected and reproduced in Figure 2. Unexpectedly, it 

 shows that Merluccius is a sister group of the rest of 

 the merlucciid genera. This disagrees with the phylog- 

 eny proposed by Inada (1989) and Okamura (1989). 

 According to the outgroup procedure as proposed by 

 Maddison et al. (1984), all three genera— Steindach- 

 neira, Macrouronus, and Lyconus— were included in 

 the analysis of the "outgroup node" for polarization 

 of character states of the hake. Dr. Tadashi Inada 

 (Tohoku Reg. Fish. Res. Lab., Same, Hachinohe 031, 

 Japan, pers. commun., Dec. 1988) has kindly provided 

 osteological information for these three genera. Appen- 

 dix 1 gives the coding of osteological characters (Inada 

 1981) for the hake species, and Appendix 2 shows the 

 matrix of the character states in the twelve extant 

 species of hakes. A new species of hake, Merluccius 

 hernandezi, has been reported from the Gulf of Califor- 

 nia (Mathews 1985) since revision of Inada. However, 

 it is not included in the present analysis due to the lack 

 of osteological information. 



