KENDALL: SERRANIDAE 



509 



preopercular 



interopercular 



preopercu 



interopercular 



B 



preopercu 



nteropercular 



D 



preopercular 



interopercular 



Fig. 269. Representative preopercular and interopercular bones from larval serranids (from Kendall, 1979); (A) Serraninae: Serranus sp.; (B) 

 Anthiinae: Anthias sp. Type 1; (C) Epinephelinae: Epinephelini, Epmephelus nivealus: and (D) Epinephelinae: Grammistini. Pseudogramma 

 gregoryi. 



scribed. At present, however, some statements can be made 

 concerning serranid systematics from what is known about the 

 larvae. 



The serranid subfamilies are clearly distinct as larvae. In fact, 

 it is not possible to characterize the Serranidae based on larval 

 morphology, because no characters unite the subfamilies while 

 separating them from larvae of all other families. Serraninae 

 larvae seem to be the least specialized and are more similar to 

 percoid genera thought to represent the basal stock from which 

 serranids arose (e.g., Morone. Lateolahrax. and Dicentrarchus). 

 The serranine genera can be distinguished from each other and 

 ordered in a rough progression of divergence from the supposed 

 ancestral larval form (as exemplified by Morone), as follows: 

 Serraniculus. Centroprislis-Paralabrax, Diplectrum Type 1, 

 Serranus (see Kendall, 1979). Characters that lead to this as- 

 sessment include pigment, body shape, sequence of dorsal spine- 

 soft ray development, and dorsal fin spine elongation. 



Based on larval and other evidence, it appears that two major 

 radiations from the ancestral serranines arose leading to the 

 anthiine and the epinepheline lineages. The anthiines form a 

 fairly cohesive group of fish which are at the same time quite 

 speciose. The generic alignment of many anthiines is unclear 

 and in some cases larval evidence is in conflict with that based 

 on adults. Anthiine larvae, like the adults, share several char- 

 acters that unite them, yet they are quite diverse and will prob- 

 ably prove to be excellent subjects for phylogenetic investiga- 

 tions. Larvae of only about half of the presently understood 

 anthiine genera are known to any extent, some of them only 

 from one transforming larva. Thus the lack of generic revisions 

 and incomplete knowledge of larval development makes it pres- 

 ently unreasonable to attempt a thorough systematic assessment 

 that would include larvae. Within the group, a progression of 

 increasing spinyness and armature is apparent. Among the lar- 

 vae described to date, armature seems to be added as follows: 



