spaced gill rakers are more likely to feed on 

 plankton than those with more widely spaced 

 rakers (Suyehiro, 1942; Yasuda, 1960a; Brooks 

 and Dodson, 1965; Kliewer, 1970). 



This paper (1) quantitatively describes the 

 gill raker apparatus of certain scombrids and 

 coryphaenids with respect to the gap between 

 gill rakers and the filtering area of the first gill 

 arch, (2) compares differences in gill raker gap 

 among species and lengths of fish, and (3) con- 

 siders the proposition that observed inter- and 

 intraspecific variations in the diet are associated 

 functionally with the morphometries of the gill 

 raker apparatus. 



MORPHOMETRY OF GILL RAKER 

 APPARATUS 



Gill raker morphometry and fork length were 

 measured from 411 fish, representing eight spe- 

 cies of scombrids and two species of coryphae- 

 nids. Albacore, Thunniis alalunc/a (Bonna- 

 terre), were fi'om the commercial longline 

 fishery operated from American Samoa, the 

 Pacific bonito, Sarda chiliensis (Cuvier), were 

 from waters off Palos Verdes, Calif., and chub 

 mackerel. Scomber japonk.iui Houttuyn, were 

 from the Honolulu fish market. All other spe- 

 cimens were from Hawaiian waters and were 

 caught with pole and line or longline by com- 

 mercial fishermen or on numerous cruises of 

 the research vessel Charles H. Gilbert of the 

 Bureau of Commercial Fisheries Biological Lab- 

 oratory, Honolulu (now National Marine Fish- 

 eries Service Hawaii Area Fishery Research 

 Center) . 



Measurements were from the first right gill 

 arch of fresh or thawed specimens. The arch 

 was removed from the fish and extended by pull- 

 ing the upper and lower branches apart until 

 the rakers were stiffly erect. Gaps between ad- 

 jacent rakers (Figure 1) were measured at the 

 base of the rakers by expanding a vernier cal- 

 iper until the two gill rakers began to spread 

 apart. Arch length and gill raker length were 

 also measured with the caliper (Figure 1). De- 

 pending on the species, six to nine gaps and six 

 to eight gill raker lengths spaced along the arch 



FISHERY BULLETIN: VOL. 69, NO. 2 



-LENGTH OF GILL RAKER 



GILL RAKER GAP 



LENGTH j/^s 

 UPPER '  

 ARCH 



FiGlTRE 1. — Diagram of the first right gill arch of a 

 scombrid as viewed from oral chamber showing the 

 morphometric measurements. Numbers indicate par- 

 ticular rakers. 



were obtained from scombrids and three gaps 

 and five gill raker lengths from coryphaenids. 

 Mean gap was the average of those measured 

 along the arch. A gap near the middle of the 

 lower arch was also used to represent gap width 

 in the primary filtering area. Filtering area 

 was calculated from average length of gill rakers 

 and length of the arch. Lower and upper arch 

 filtering areas were computed separately and 

 summed. 



DESCRIPTION 



Gill rakers of the first arch of most scombrids 

 were conspicuous and well developed. Inner 

 edges of the rakers of most species were covered 

 with numerous short, spiny protuberances. For 

 S. japoniciis, these spines were thin, about as 

 long as the gill raker gap, and evenly spaced to 

 form a finer sieve between adjacent gill rakers. 

 The other three arches of scombrids lacked gill 

 rakers, but smaller rakerlike processes on the 

 inner faces of the all arches projected posteri- 

 orly to the adjacent arch forming a sieve. Inner 

 edges of these processes had short, «piny pro- 

 tuberances similar to the gill rakers. 



Rakers were articulated so that they became 

 stiffly erect forming a parallel row of blade- 

 shaped rakers when the acute angle between the 

 upper and lower arch was expanded toward 90 

 degrees. In the branchial chamber the tips of 

 the rakers extended to the inner surface of the 

 flared gill cover. 



The wahoo, Acanthocybinm solanderi (Cuv- 

 ier) , has no gill rakers, but most scombrids have 

 more than 20 elongated rakers — K. pelamis in 



362 



