JOHNSON: PERCOIDEI 



477 



Table 120. Extended. 



5/3/?/2 



3/3/2/2 



1-2; 3-4 



5/3/1-2/2 



from as few as 3 in some acanthoclinids to 89 in some cepolids. 

 Within families, the range of dorsal fin ray counts may be rel- 

 atively restncted as in the Lutjanidae (X-XII, 10-17) or quite 

 broad, as in the Sciaenidae (VII-XV — 1, 17-46). 



In most percoids the spinous and soft portions of the dorsal 

 fin are continuous, but gradual shortening of the posteriormost 

 spines results in a variously developed cleft or apparent sepa- 

 ration. Where this cleft is present, the ultimate spine is notably 

 longer than the penultimate and is considered to form the first 

 element of the soft dorsal portion of the fin. Some groups (e.g., 

 Pseudochromidae, Grammatidae, Plesiopidae, Priacanthidae, 

 Cepolidae) do not develop this cleft. Others, such as the Ac- 

 ropomatidae, Ambassidae, Apogonidae, Emmelichthyidae, En- 

 oplosidae and Epigonidae have such a well-developed cleft that 

 the spinous and soft portions of the fin appear completely sep- 

 arate. Pterygiophores usually continue beneath the resultant gap 

 and may even bear minute spines. The Mullidae and Echenei- 

 didae are exceptional in having no pterygiophores below this 

 gap. Extreme separation of the spinous and soft dorsal fins oc- 

 curs only in the Echeneididae, where the spinous dorsal fin has 

 been modified as an attachment disc and has moved far forward 

 to cover the flattened cranium. 



The anal fin of percoids is less variable in form and com- 

 position than the dorsal fin. The most common, and apparently 

 primitive condition is three anal spines. The first anal pteryg- 

 iophore is larger than the succeeding pterygiophore and bears 

 the first two spines in supernumerary (non-serial) association 

 and the third in serial association (see Johnson, 1980). Scato- 

 phagids and some chaetodontids and pomacanthids have four 

 spines, the first two supernumerary. Centrarchids have from 

 two to eight spines, pentacerotids from two to six and gerreids 

 from three to five, all with the first two supernumerary. The 

 only other percoids characterized by more than three anal spines 

 (eight to eleven) are the Acanthoclinidae, where one or two may 

 be supernumerary. Several groups have fewer than three anal 

 spines, and, as in the dorsal fin, it is important to understand 

 the nature of this reduction. Apogonids, for example, have only 

 two spines and only one of these is supernumerary, suggesting 

 that the anteriormost spine was lost. The mesopelagic epigon- 

 ines (sensii Eraser, 1 972a) have usually been treated as a subfam- 

 ily of the Apogonidae, for they also have only two anal spines. 

 The epigonine anal spines, however, are both supernumerary 

 (as are those of the Sciaenidae), suggesting that the usually spi- 

 nous third (serial) element has not transformed into a spine. 

 Hence, the two anal spine conditions of epigonines and apo- 

 gonids are not homologous. In bathyclupeids, the single anal 

 spine is serially associated with the first pterygiophore, sug- 

 gesting that the first two spines have been lost. Only a few groups, 

 Bramidae. Caristiidae, Congrogadidae, Coryphaenidae and some 

 cepolids and grammistine serranids, lack anal spines. The pres- 

 ence of 1-3 supernumerary elements on the first pterygiophore 

 in all these groups indicates transformation rather than loss of 

 the pre-existing spines. Anal soft rays range in number from 4 

 in the Acanthoclinidae to 101 in the Cepolidae and, with some 

 exception, the range of variability within families is comparable 

 (frequently within two or three rays) to thai of the dorsal soft 

 rays. 



4/3/1/2 

 1-2 



Ct 



Predorsal bones. — In most percoids, one to three strut-like bones 

 precede the anteriormost pterygiophores of the dorsal fin. It has 

 been proposed (Smith and Bailey, 1961), but never conclusively 

 demonstrated, that these predorsal bones were derived from 

 true pterygiophores that once bore spines or rays, but Eraser 



