NUMBER AND PLACEMENT OF DORSAL-FIN SPINES.—
Eospinus has three dorsal spines. The plesiomorphic condition
for tetraodontiforms is six spines, as found in all Triacanthoidei
and in Eoplectus, the most generalized representative of the
Tetraodontoidea. Within the Tetraodontoidei, reduction in the
number of dorsal spines has taken place independently in the
Balistoidea (5-6 spines in Spinacanthidae, 2-3 or rarely 1 in
Balistidae, and none in Ostraciidae) and Tetraodontoidea (6
spines in the Eocene Eoplectidae, 1-3 rudimentary spines or
none in Triodontidae, and none in the Eocene Zignoichthyidae,
Tetraodontidae, Diodontidae, and Molidae).
Winterbottom's (1974) phylogeny of the balistoids indicates
that an a posteriori synapomorphy of the ostraciids is the
posterior migration of the basal pterygiophore of the spiny
dorsal fin toward the origin of the soft dorsal fin and the
complete loss of the spiny dorsal fin, with the pterygiophores
becoming the prominent supraneural element found just in
front of the first basal pterygiophore of the soft dorsal fin in all
ostraciids. Conversely, the sister group relationship of balistids
and spinacanthids is supported by the anterior migration of the
dorsal spines and their pterygiophores to at least the rear of the
skull, which condition is apomorphic relative to the position
just behind the rear of the skull in the triacanthoid and
eoplectid-based outgroups. Spinacanthids retained five or six
dorsal spines from the ancestral triacanthoid-eoplectid condition
and these spines migrated even further forward onto the
top of the skull over the eye and greatly increased in length
relative to the moderate length in triacanthoids and eoplectids.
The reduction of the number of dorsal spines in balistids to
three in balistins and to two (rarely one) in monacanthins is
independent of the loss of spines in ostraciids. Balistins
retained the position of the spiny dorsal fin and their supports
at the rear of the skull, plesiomorphic for the balistidspinacanthid
clade, while monacanthins further specialized by
the migration of the spines further forward onto the top of the
skull, usually to the region of the eye or even in front of the eye.
The forward migration of the spiny dorsal fin in monacanthins
is here interpreted as independent of that in spinacanthids.
The presence of three dorsal spines in Eospinus relates it to
balistids within the spinacanthid-balistid clade, while the origin
of the spines at the rear of the skull is the ancestral condition for
that clade.
On the basis only of the number of dorsal spines and not of
their position (nor of the supraneural in their absence),
spinacanthids with five or six spines would be the sister group
of both balistids and ostraciids, with three or fewer spines being
a synapomorphy for a balistid-ostraciid clade and the complete
loss of the dorsal spines a synapomorphy of ostraciids.
It is clear that the position of the spiny dorsal fin was
anterior, toward the rear of the skull, in the ancestral balistoid.
One could hypothesize that the spines were lost in ostraciids
while still in this position and that only then did the basal
pterygiophores, having lost their original function of spine
support, migrate posteriorly to the origin of the soft dorsal fin
to become the supraneural. However, in triodontids, which are
the only tetraodontiforms with a rudimentary spiny dorsal fin,
the spiny dorsal fin and its pterygiophores have migrated
posteriorly to just in front of the soft dorsal fin. Moreover,
when the rudimentary spiny dorsal fin is absent (Indian Ocean
populations) in the only Recent species, Triodon macropterus
Lesson, the basal pterygiophores are entirely comparable to the
posteriorly placed supraneural of ostraciids. This analogous
positioning of the pterygial supports of the absent spiny dorsal
fin in triodontids and ostraciids is in conformity with
Winterbottom's (1974) contention of the sister group relationship
of spinacanthids and balistids, both of which have anterior
migration of the spiny dorsal fin, and the sister group
relationship of that clade with the ostraciids, which have
posterior migration of the pterygiophores and loss of the spines.
This scenario also is consistent with Rosen's (1984) hypothesis
of ostraciid relationships.