ABSTRACT
This, the largest suborder with 40 families represented from our area, is the most challenging to identify ELH stages. Eggs are nearly impossible unless one rears in the laboratory from known parents or rears to identifiable stages. For larvae and juveniles many families share similar features that for some indicate relationships. Most have 2426 myomeres (vertebrae) and some head spination (as do many others). The individual family accounts provide useful information about similar families, but great care must be taken to avoid reaching incorrect identifications. It is critical to develop good size series especially for this suborder as so many are very similar. It is extremely difficult to identify gerreids and haemulids when they are small, whereas some with a lot of pigment or unusual morphological feature can be placed in a family rather easily. For example the tholichthys stage of chaetodontids is very distinctive, but the small pomacanthids closely resemble some carangids. Even in a single family differences can be striking as the larvae of the serranid subfamilies are very different as are the larvae of the two malacanthid subfamilies. The principally freshwater cichlids are included because larvae of one species have been found in marine waters of South Florida where it is introduced and its native range in Mexico. The cichlids and pomacentrids have been placed in the Labroidei by some, but have now been removed to the Percoidei as discussed by Jones in the Labroidei chapter. The moronids (Morone saxatilus) and centrarchids are excluded as their larvae would only be rarely found in very low salinity waters. Cepolidae are not included either as their presence at Bermuda is doubtful (Smith-Vaniz et al. 1999). Johnson (1984, 1993) and Nelson (1994) provide excellent reviews of the systematics of the suborder; and family accounts in Carpenter (2003) provide identification keys and distributions for adults. Some of the family composition differs in this book in that we have more closely followed Johnson (1993) and Roberts (1993).
