大阪湾における背伸長鰭条を2本有するウシノシタ科仔魚の形態と識別

抄録

We investigated the morphology of cynoglossid larvae with elongated dorsal double rays collected in Osaka Bay. Larvae were divided into three types based on the distribution of melanophores on the lateral surface of the tail. Type A had melanophores along the midline of the tail. Type B had no melanophores along the midline, but had melanophores scattered near the center of the tail. Type C had no melanophores on the lateral surface of the tail. Type A, B, and C were collected in August–November, August–October and July–October, respectively. The range of the body length (BL) of each type were as follows, A: 4.0–12.4 mm; B: 3.4–7.8 mm; C: 2.1–9.6 mm. There were no individuals larger than 8 mm BL in type B. In type C, all but two individuals larger than 9.5 mm BL, were smaller than 5.2 mm BL. Types A and B had similar numbers of dorsal, anal and caudal fin rays (D ≤ 115; A ≤ 88; C 10) and myomeres (M 48–50) but could be distinguished by ratios of the body depth and preanal length to BL. In type A, notochord flexion had not begun at 7 mm BL but in type B, notochord flexion was complete at ca 7 mm BL. The numbers of fin rays and myomeres of type C larger than 9.5 mm BL (D 125–130, A 101–102, C 10, M 58–59) differed from type A and B; however, type C smaller than 5.2 mm BL had similar numbers of myomeres and proportion to type A and B. Our analysis of cynoglossid larvae and the characteristics of cynoglossid fishes occurring in Osaka Bay i.e., Arelia bilineata, Cynoglossus abbreviates, C. interruptus, C. joyneri, C. robustus and Paraplagusia japonica, suggest the following; type A: C. interruptus; type B: C. joyneri; type C (larger than 9.5 mm BL)：C. robustus. Type C smaller than 5.2 mm BL were likely to be C. joyneri or C. interruptus. Results of our study provide a first attempt at a description of C. interruptus larvae. Our results differ from some previous reports, which described the morphology of C. joyneri and C. robustus larvae and it is possible that individuals identified as C. joyneri larva in previous studies included C. interruptus or C. robustus larva. Discrepancies between the results of the present and previous studies suggest that further study will be required before the cynoglossid larvae can be positively identified.