KEENE AND TIGHE: BERYCIFORMES 



391 



Fig. 209. Kasidoron prejuvenile of Gibberichthys pumilus, 1 5.3 mm SL (source: de Sylva and Eschmeyer, 1977). 



considered the holocentrids to be a distinct major subgroup 

 within the order. Rosen (pers. comm. to Zehren) believes that 

 the Holocentridae should be placed within the Perciformes. 



Another result of Zehren's study is that the Berycidae appear 

 to be the primitive sister group to the other eight families and 

 should be placed in their own suborder, the Berycoidei. The 

 Trachichthyidae, Diretmidae. Anoplogasteridae, Anomalopi- 

 dae and Monocentridae are closely related and should be placed 

 in the suborder Trachichthyoidei, as suggested by Parr (1933). 

 The Gibberichthyidae, Stephanoberycidae and Melamphaidae 

 also appear closely related and form the suborder Slephanob- 

 erycoidei. 



Despite the efforts of Rosen and Zehren, there are still prob- 



lems with beryciform classification. Only adult characters have 

 presently been used, but early life history data is pertinent in 

 two instances. In the cladogram, a common ancestry is suggested 

 for the Diretmidae, Anoplogasteridae, Trachichthyidae. An- 

 omalopidae and Monocentridae with no character state to sep- 

 arate them. The larval head spine pattern in the Diretmidae and 

 Anoplogasteridae is similar and distinctive, and may help to 

 resolve the cladogram. Gibberichthys with its kasidoron stage 

 may appear to be vastly different from the Melamphaidae, but 

 the occurrence of very long branched pelvics in larval Poromitra 

 suggest a possible relationship (de Sylva and Eschmeyer, 1977). 

 In summary, further phylogenetic studies of the order Bery- 

 ciformes are needed in order to determine if the order is mono- 



