290
Wei-Jen Chen et al.
bootstrap proportions: one node over 13 showed a bootstrap proportion
below 50% and 9 nodes above 70%.
100
98
84
89
17.2
Gymnodraco acuticeps
Champsocephalus gunnari
r - - - Neopagetopsis ionah - - - - ,
Pseudochaenichthys
georgianus
Pagetopsis macropterus
Pagetopsis maculatus
Dacodraco hunteri
Chaenocephalus aceratus
Channichthys rhinoceratus
Chionobathyscus dewitti
Cryodraco antarcticu
Chaenodraco wilsoni
Chionodraco myersi
C. rastrospinosus
C. hamatus
A
C
D
18.2
Fig. 1. Two equiparsimonious trees obtained from a Branch and Bound search without
weighting schemes (P AUP 3.1.1) on 1135 sites of two mitochondrial genes (cytochrome b
and D-loop). Each tree has a length of 659 steps, CI of 0.665 and RI of 0.570. Branch
lengths are proportional to the number of steps affected to the branch (ACCTRAN). The
numbers below branches refer to bootstrap proportions (500 iterations). Asterisks refer to
bootstrap proportions inferior to 50%. The difference between these two trees is the relative
position of Chaenocephalus aceratus
Considering the three species that were previously instable, we now
confinn the position of Chaenodraco wilsoni as the sister-group of
Chionodraco species (bootstrap proportion of 67%) and Channichthys as
the sister-group of the clade Chionobathyscus + Cryodraco. The total
evidence approach fails to clearly identify the relationships of
Chaenocephalus. This tree from the total evidence (Fig. 1) is strikingly
similar to the one proposed by Iwami [2], excepting the relationships of
Channichthys.