LARVAE OF 5 FISH FAMILIES WITH PATTERN 10 OF THE RAMUS-LATERALIS-ACESSORIUS NERVE (ARRIPIDAE, GIRELLIDAE, KYPHOSIDAE, MICROCANTHIDAE AND SCORPIDIDAE) - RELEVANCE TO RELATIONSHIPS
Fj. Neira et al., LARVAE OF 5 FISH FAMILIES WITH PATTERN 10 OF THE RAMUS-LATERALIS-ACESSORIUS NERVE (ARRIPIDAE, GIRELLIDAE, KYPHOSIDAE, MICROCANTHIDAE AND SCORPIDIDAE) - RELEVANCE TO RELATIONSHIPS, Bulletin of marine science, 60(1), 1997, pp. 117-138
We describe the larvae of five temperate Australian fishes from five p
ercoid families considered to form part of a monophyletic group becaus
e they share the uncommon pattern 10 of the facial nerve ramus lateral
is accessorius (RLA-10). An attempt was made to assess to what extent
the larval characters described for these species-Arripis trutta (Arri
pidae), Girella tricuspidata (Girellidae), Kyphosus sp. (Kyphosidae),
Atypichthys strigatus (Microcanthidae) and Scorpis (lineolata?) (Scorp
ididae)-support a close relationship among these RLA-10 families. Larv
ae of the five species share similarities in body form, head spination
, fin formation and pigment pattern, and generally fit the ''distincti
ve larval form'' previously described for ''girelloid'' RLA-10 familie
s. Although such similarities are consistent with a close relationship
among these families, most attributes of the distinctive larval form
are also present in larvae of a number of non RLA-10 percoid families,
notably the Pomatomidae. Based only on larval characters, one could h
ypothesize that most RLA-10 percoid families, including the Dichistiid
ae (Leis and van der Lingen, 1997), form part of a larger group that a
lso includes non RLA-10 families such as the Pomatomidae. Alternativel
y, the ''distinctive larval form'' in Pomatomus and some other non RLA
-10 families could have been acquired independently. However, since we
could not show that any of the attributes of the distinctive larval f
orm constitute synapomorphies, nor is the sister group of the RLA-10 f
amilies known, neither of these alternatives could be tested during th
is study. Larval characters alone cannot at present resolve the relati
onships of the RLA-10 families.