552 



ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM 



alternating with bursts of acceleration. There is a general but 

 far from universal trend for bottom dwelling fishes to become 

 elongate; the eel-like body is widely distributed taxonomically. 

 Bottom-living fishes often use crevices for shelter and forage in 

 interstices, and may burrow. Elongation of the body accom- 

 panied by an increase in the number of vertebrae produces the 

 flexibility necessary for these activities. The elongate body form 

 requires either anguilliform swimming or undulation of the me- 

 dian fins. In either case the role of the caudal fin is reduced. 

 The pectorals are used in short darts or lunges, and their fan 

 shape is associated with accelerating a large amount of water 

 per thrust. This function is important even in relatively elongate 

 forms in accelerating the head in feeding strikes, and pectorals 

 are reduced or lost in only a few lineages. 



The pelvics of bottom living forms no longer have a hydro- 

 dynamic function as brakes or rudders. Instead they may func- 

 tion as props which hold the head ofl^ the bottom (as in the 

 Cottidae and Gobiidae as well). A reduction in the number of 

 rays is also seen in the Cottidae. 



That morphological features are functional does not mean 

 that their joint possession cannot be taken to demonstrate com- 

 mon ancestry. However, it does indicate caution. The only one 

 of Gosline's characters for the Blennioidei that is not clearly 

 functional is the 1:1 relationship of median fin rays and ver- 

 tebrae. However, the reduction in the number of fin rays per 

 segment to one is the culmination of a functional trend begun 

 in the Paleozoic, and we cannot yet be sure that it happened 

 but once. 



Although Gosline regarded his classification as owing more 

 to that of Jordan than Regan, his mam characters of pelvic 

 position and median fin ray arrangement are exactly those given 

 by Regan in his diagnosis. Gosline's concept of the Blennioidei 

 and its superfamilies, although not completely accepted (see 

 Nelson, 1976), has not been superseded, except that his Con- 

 grogadoidae is no longer included; the Congrogadidae is now 

 placed in the Percoidea (Winterbottom, 1982) and the Pero- 

 nedysidae has been synonymized with the Clinidae (George and 

 Springer, 1980). 



According to Gosline the Blennioidei (without the Congro- 

 gadoidae) may be divided into four superfamilies. The first of 

 these, the Notothenioidae, is clearly the most heterogeneous. In 

 addition to the Antarctic and sub-Antarctic families (Bovich- 

 thyidae, Nototheniidae, Harpagiferidae, Bathydraconidae and 

 Channichthyidae) usually placed in this group (Berg, 1940), the 

 tropical Mugiloididae (=Parapercidae) Trichonotidae and Chei- 



marrhichthyidae were included although they do not share with 

 them the specialized features of a smgle nostril and a loss of 

 one pectoral actinost. There appears to be no reason to regard 

 the two groups of families as closely related. 



The Trachinoidae was said to be comprised of the Trachin- 

 idae, Leptoscopidae, Uranoscopidae and Dactyloscopidae. All 

 are adapted for lying buried in the substrate, and it is likely that 

 their structural similarities are related to this habit. The Dac- 

 tyloscopidae has recently been placed in the Blennioidae (George 

 and Springer, 1982). 



The superfamily Blennioidae was regarded as composed by 

 the families Tripterygiidae, Clinidae, Chaenopsidae, and Blen- 

 niidae. Subsequently the subfamily Labrisominae of the family 

 Clinidae was raised to family status and the Dactyloscopidae 

 transferred from the Trachinoidei (George and Springer, 1980). 

 Within the Blennioidei, the superfamily may be characterized 

 by the combination of two nostrils on each side, pelvic soft- 

 rays four or fewer, prootic excluded from orbital rim (that is, 

 ascending wing of paraspheroid meets frontal), and basisphe- 

 noid present. 



The remaining superfamily, the Zoarceoidae, was regarded as 

 composed of 1 1 families, some poorly understood. Anderson 

 (1983, this volume) recognized 8 families in the group: Bathy- 

 masteridae, Stichaeidae, Pholididae, Anarhichadidae, Ptil- 

 ichthyidae, Zaproridae, Scytalinidae and Zoarcidae. Although 

 composed of forms differing greatly in morphology, the super- 

 family may be diagnosed as blennioids with a single nostril on 

 either side of the head, prootic excluded from rim of orbit, and 

 basisphenoid absent. There is no merit in the removal of the 

 Zoarcidae to the Gadiformes (see also Anderson, this volume). 



It should be clear from the foregoing that no satisfactory def- 

 inition of the Blennioidei has as yet been framed. Perhaps lines 

 of relationships would best be recognized by restricting the Blen- 

 nioidei to the Blennioidae and Zoarceoidae of Gosline, and 

 returning his other two superfamilies to the Percoidei. It appears 

 that ontogeny and larval characters have as yet little to con- 

 tribute to questions of suprafamilial and subordinal relation- 

 ships between and among these fishes. 



Perhaps it is fitting to end with a quote from Jordan (1923), 

 addressing issues such as this: "I may repeat a warning as old 

 as science itself: that we must not expect a degree of accuracy 

 which the subject in question does not permit." 



ScRipps Institution of Oceanography, University of Cal- 

 ifornia, San Diego, La Jolla, California 92093. 



Schindlerioidei: Development and Relationships 

 W. Watson, E. G. Stevens and A. C. Matarese 



THIS suborder contains a single paedomorphic family com- 

 posed of two species of the genus Schindlena. Both species 

 inhabit neritic surface waters of the subtropical and tropical 

 Indian and Pacific Oceans (Bruun, 1940; Schindler, 1932; R. J. 

 Lavenberg, pers. comm.). Their early life histories are known 



from the work of Watson and Leis (1974), Miller et al. (1979), 

 and Ozawa and Matsui (1979). Classification of Schindlerioidei 

 is speculative, and its placement here by Nelson (1976) follows 

 Gosline (1971), who tentatively considered this taxon a percoid 

 derivative, possibly related to Ammodytoidei. 



