The evidence from the 73% SL dorsal-spine length in
Eospinus indicates relationship with spinacanthids, while the
evidence from the number of spines indicates relationship with
balistids. If Eospinus is proposed as the sister group of
spinacanthids on the basis of dorsal-spine length, then the
number of spines had to be reduced from the ancestral six to
three or fewer independently in Eospinus and balistids (two
steps). If Eospinus is proposed as the sister group of balistids on
the basis of the number of spines, then the size of the spines had
to be increased from the ancestral moderate size to large
independently in Eospinus and spinacanthids (two steps). None
of the other characteristics of Eospinus supports one versus the
other of these equally parsimonious alternatives.
PELVIC FIN.—Eospinus has the pelvic fin reduced to a pair of
partially fused and relatively large rudimentary spines placed
posteriorly at the end of the pelvis. The plesiomorphic
condition for tetraodontiforms is a pelvic fin of one spine and
two (triacanthoids) to four (eoplectids) rays placed thoracically
along the middle of the length of the pelvis below the pectoral
fin (condition of pelvis unknown in eoplectids but pelvic fin
thoracic). Among the alternate outgroups for tetraodontiforms,
most acanthuroids and some zeiforms have a pelvic fin with a
spine and five well-developed rays (in some zeiforms up to
seven rays if a spine is present or up to 10 rays if the spine is
absent; rays reduced to three or four in some specialized
acanthuroids). The pelvic fin in zeiforms and acanthuroids is
like that of morphologically primitive tetraodontiforms in
being placed thoracically along the side of the pelvis, which
usually has a prominent posterior process behind the fin; i.e.,
the fin is never at the extreme posterior end of the pelvis as in
balistids.
The pelvic fin of tetraodontiforms has become much
reduced, apparently independently, in both lineages of the
Tetraodontoidei. In the Tetraodontoidea the pelvic fin is
completely lost in all groups except the Eoplectidae, although
the Triodontidae retain the pelvis. In the Balistoidea the pelvic
fin is either reduced to a complex rudimentary structure at the
end of the pelvis in front of the anus or absent altogether. In the
Balistidae the pelvic fin rudiment is thought to be the partial
fusion product of the pelvic spines from both sides of the body.
Except at its distal end, this rudimentary fused spine is
surrounded by a series of enlarged encasing scales (in balistins
and most monacanthins, while the rudiment and its encasing
scales are secondarily further reduced or completely lost in
some of the more specialized monacanthins, even though the
pelvis is always present: Tyler, 1962; Matsuura, 1979). In the
Ostraciidae both the pelvis and pelvic fin are completely lost. In
the Eocene Spinacanthidae there is no clear evidence in either
of the single specimens of the two species of a pelvic fin or
pelvis, and the pelvic fin must be considered to have been either
extremely rudimentary or entirely lost.
The spiny structure at the end of the pelvis in Eospinus is less
specialized than that of balistids because of its larger or less
apomorphically reduced size and its lack of development of the
specialized series of encasing scales that surround the rudiment
in the more primitive balistids (subsequently lost by some
specialized monacanthins). The spiny processes along the
lateral edges of the reduced and partially fused pelvic spines in
Eospinus also can be interpreted as plesiomorphic relative to
the smooth edges of the smaller rudiment in balistids because
such spiny processes are present on the pelvic spines of both the
more generalized triacanthoid (the triacanthodids) and eoplectid-
based outgroups.