Four Years in Notothenioid Systematics: A Molecular Perspective
263
cannot be considered to be representive of the family Bovichtidae. We
know, in fact, from both nuclear [10] and mitochondrial [16] ribosomal
DNA data that bovichtids are not a natural group. In both cases it has been
shown with good statistical support that bovichtids are paraphyletic, with
Pseudaphritis closer to the rest of the notothenioids than are Cottoperca
and Bovichtus. The two latter genera are so close to perciform outgroups
that it is difficult to recover monophyletic notothenioids (Fig. 2).
65
. . . . . . . . Trigla lueema
-- Perea tluviatils
Scomber scombrus
Trachinus draco
Labrus bergylta
Neolamprologus meeli
Austrolycus depressiceps
Pholis gunnellus
- Cottoperca gobio
. . . . . . . . . . . . . . . . . Bovichtus varicgatus
Pscudaphritis urvillii
I
Eleginops maclovinus
9~ .10Q<-1 - Dissostiehus mawsoni
! - - Artedidraco loennbergi
Fig. 2. Phylogenetic relationships of the three bovichtid genera as resulted from the
analysis of nuclear ribosomal DNA (28S) dataset [10]. This tree is a strict consensus tree of
six equiparsimonious trees. Numbers above branches are bootstrap proportions obtained
after 1000 replicates analyzed by the parsimony method as implemented in PAUP [11]
With regard to the remaining notothenioid families, all our analyses
suggest that Eleginops maclovinus should not be included in the family
Nototheniidae (Fig. 1), since it branches out before the divergence of all
the other species (excluding bovichtids). It has been recently proposed
[17] based on morphological data that this species should form a
monospecific family (Eleginopsidae), and should be considered the sister
group to all the other families (again excluding bovichtids). In our
263
cannot be considered to be representive of the family Bovichtidae. We
know, in fact, from both nuclear [10] and mitochondrial [16] ribosomal
DNA data that bovichtids are not a natural group. In both cases it has been
shown with good statistical support that bovichtids are paraphyletic, with
Pseudaphritis closer to the rest of the notothenioids than are Cottoperca
and Bovichtus. The two latter genera are so close to perciform outgroups
that it is difficult to recover monophyletic notothenioids (Fig. 2).
65
. . . . . . . . Trigla lueema
-- Perea tluviatils
Scomber scombrus
Trachinus draco
Labrus bergylta
Neolamprologus meeli
Austrolycus depressiceps
Pholis gunnellus
- Cottoperca gobio
. . . . . . . . . . . . . . . . . Bovichtus varicgatus
Pscudaphritis urvillii
I
Eleginops maclovinus
9~ .10Q<-1 - Dissostiehus mawsoni
! - - Artedidraco loennbergi
Fig. 2. Phylogenetic relationships of the three bovichtid genera as resulted from the
analysis of nuclear ribosomal DNA (28S) dataset [10]. This tree is a strict consensus tree of
six equiparsimonious trees. Numbers above branches are bootstrap proportions obtained
after 1000 replicates analyzed by the parsimony method as implemented in PAUP [11]
With regard to the remaining notothenioid families, all our analyses
suggest that Eleginops maclovinus should not be included in the family
Nototheniidae (Fig. 1), since it branches out before the divergence of all
the other species (excluding bovichtids). It has been recently proposed
[17] based on morphological data that this species should form a
monospecific family (Eleginopsidae), and should be considered the sister
group to all the other families (again excluding bovichtids). In our