Ostorhinchini Whitley 1959

Mabuchi, Kohji, Fraser, Thomas H., Song, Hayeun, Azuma, Yoichiro & Nishida, Mutsumi, 2014, Revision of the systematics of the cardinalfishes (Percomorpha: Apogonidae) based on molecular analyses and comparative reevaluation of morphological characters, Zootaxa 3846 (2), pp. 151-203 : 183-184

Tribe Ostorhinchini Whitley 1959

Type genus Ostorhinchus Lacepède 1802

Diagnosis. Members of the Apogoninae : dorsal fin VI+I,9 or VII+I,9; anal fin II,8–9; head and body with ctenoid scales; pored lateral-line scales 6–26; preopercle ridge smooth, edges serrate; 2–3 supraneurals; supramaxilla absent; basisphenoid present; one pair of reduced uroneurals present or absent; three epurals; five free hypurals or 1–2 fused; parhypural separate; caudal fin forked; head and/or body with one or more light or dark stripes, bars rarely present, basicaudal spot or broad mark present or absent.

Other characteristics. one or two supernumerary dorsal spines; branched first segmented fin ray in second dorsal-fin; first anal-fin ray branched and segmented; ctenoid scales on predorsal, cheek, breast, two pelvic scales, and body; ctenoid scale on opercle and onto base of caudal fin; pored lateral-line scales with multiple pores; pectoral fin-rays 11–16; three supraneurals; 9+8 segmented principal caudal rays, 15 branched, upper and lower unbranched; unbranched procurrent rays, longest segmented; teeth in one or multiple rows on premaxilla, dentary, vomer, palatine, all villiform or absent on palatine, sometimes enlarged, but not canine on side of dentary; six infraorbitals, bony shelf on third infraorbital; anterior ceratohyal smooth or notched; developed gill rakers 10–26; 10+14 vertebrae; 8 ribs; 8 epipleurals; blackish stomachs and intestines; low crest on PU2.

Distribution. The restricted Ostorhinchus is widespread throughout the Indo-Pacific from the Red Sea, East Africa to Japan, Hawaii, Easter Island and French Polynesia.

Remarks. Monophyly of recent Ostorhinchus (species indicated by blue in Figs. 2–6 ) was significantly rejected by AU test (H 16 in Table 6 ). This tribe, however, corresponds to one of the five lineages of the recent Ostorhinchus , clade XII. It included most species of the recent Ostorhinchus : many striped species of Ostorhinchus (species of Ostorhinchus -5) and Brephamia parvula Smith & Radcliffe in Radcliffe 1912 ( Figs. 2–6 , Table 4 ). Other species of the recent Ostorhinchus were divided into four lineages, Ostorhinchus -1, 2, 3 and 4. They were all placed far from the clade XII. Except for Ostorhinchus -3 ( O. margaritophorus ), monophyly between this tribe (clade XII) and each of the remaining lineages was rejected by AU tests (H07, 08 and 11 in Table 6 ). Although the bootstrap values supporting the clade XII were not so high (<50%) both in the ML ( Fig. 2 ) and MP ( Figs. 4 and 5 ) trees ( Table 4 ), and further the clade was divided into two separate clades in the BA tree ( Fig. 3 ), we are treating this group as a single genus, based on the typical color pattern (stripes on body and/or head) shared by most of the included species. This group corresponds to the "striped group" of Ostorhinchus ( Ostorhinchus III ) in Mabuchi et al. (2006). One of the other recent Ostorhinchus lineages, O. margaritophorus , has long stripes, short bars between two long stripes on body and fused hypurals 1+2. Phylogenetic position of this species remains unsettled in the molecular trees ( Table 5 ), but monophyly between this species and the members of this tribe was not rejected by AU test (H01 in Table 6 ). Based on its featured color pattern, we tentatively included this species in this tribe. Apogon (Brephamia) parvula has a very similar species within this tribe. It is Ostorhinchus neotes . Although they were not sister to each other in the molecular trees, their monophyly was not rejected by AU test (H 12 in Table 6 ). While recognition of Brephamia Jordan in Jordan & Jordan (1922) as a subgenus or genus needs further evaluation, we synonymized it under the restricted Ostorhinchus for the present. Likely synapomorphies may include color pattern groupings, probably at the subgenus level. At the genus level, with the exception of the loss of tiny first dorsal-fin spines, all species have 7 visible first dorsal spines, a serrated preopercular edge and most have blackish stomachs and intestines.

There are two available names ( Gronovichthys Whitley 1929 ; Lovamia Whitley 1930 ) which could be used in the future. There are at least 93 species in the group. We have sampled 31 species. Whitley (1959) first used the present name at family level Ostorhinchidae. The name appeared once. We use the name at the tribal level.

Figures

FIGURE 2. Phylogenetic tree from the partitioned maximum-likelihood (ML) analysis. Numbers besides internal branches indicate bootstrap probabilities (BPs) from 1000 replicates (only those of> 50% shown). Species names of recent Ostorhinchus [we included here Apogonichthyoides melas as a possible Ostorhinchus, following Mabuchi et al. (2006)] were blue, those of Apogon sensu lato [genus Apogon in Fraser (1972)] excepting the Ostorhinchus red, and those of genus Rhabdamia in Fraser (1972) green. The names of the twelve major clades (clades I to XII) within the Apogoninae were indicated on the left of the corresponding clades. Four species, which phylogenetic positions were largely different among the ML, Bayesian (BA: Fig. 3), and most parsimonious (MP: Figs. 4 and 5) tress, were indicated by arrows with asterisks.

FIGURE 3. Phylogenetic tree (50% majority-rule consensus tree) from the partitioned Bayesian (BA) analysis. Numbers besides internal branches indicate Bayesian posterior probabilities (PPs: shown as percentages, only those of> 50% shown). Species names were colored as in Figure 2. The names of the twelve major clades were indicated as in Figure 2. Four species, which phylogenetic positions were largely different among the ML (Fig. 2), BA (Fig. 3), and most parsimonious (MP: Figs. 4 and 5) tress, were indicated by arrows with asterisks.

FIGURE 4. Strict consensus tree of two of the five most parsimonious (MP) trees (MP tree-A). Numbers besides internal branches indicate bootstrap probabilities (BPs) from 1000 replicates (only those of> 50% shown). Topological incongruity between the two MP trees denoted by an arrowhead. Four species, which phylogenetic positions were largely different among the ML (Fig. 2), BA (Fig. 3), and MP (Figs. 4 and 5) tress, were indicated by arrows with asterisks. Species names were colored as in Figure 2. The names of the twelve clades were indicated as in Figure 2.

FIGURE 5. Strict consensus tree of the remaining three of the five most parsimonious (MP) trees (MP tree-B). Topological incongruities between the three MP trees denoted by arrowheads. Four species, which phylogenetic positions were largely different among the ML (Fig. 2), BA (Fig. 3), and MP (Figs. 4 and 5) tress, were indicated by arrows with asterisks. This tree was largely different from the MP tree-A (Fig. 4) in the positions of Pterapogon kauderni and Vincentia novaehollandiae.

FIGURE 6. Strict consensus tree of the ML, BA and MP trees shown in Figures 2-5. Species names were colored as in Figures 2–5. The names of the twelve major clades (clades I to XII) within the Apogoninae were indicated besides the corresponding clades, while the names of the thirteen tribes (defined based on morphological characters) on the right of corresponding major and minor clades. The tribe Lepidamiini is not in the tree. Tribes with solid bars include species of Apogon sensu lato, and those with open bars do not include it. Geographic ranges were indicated on the right of the clades only for the species of clade II (Apogonini), all the other apogonids occurring on Indo-Pacific Basin. For the revised genus names of some species, see Appendix A.