526 



Fishery Bulletin 92(3), 1994 



The importance of holocentrids as predators has 

 been well documented (Randall, 1967; Hobson, 1974; 

 Vivien and Peyrot-Clausade, 1974; Gladfelter and 

 Johnson, 1983). Gladfelter and Johnson (1983) found 

 that seven species of squirrelfishes made up >99% of 

 the nocturnally active, benthic crustacean-feeding 

 fishes at St. Croix, U.S. Virgin Islands. Randall ( 1967) 

 reported that holocentrids accounted for about 14% 

 by number and 11% by weight of all zooplankton con- 

 sumed as prey by the fish community at St. John, 

 U.S. Virgin Islands. In the NWHI, holocentrids, in- 

 cluding M. amaena, are among the most successful 

 families offish predators. Holocentrids accounted for 

 nearly 40% by number, over 60% by weight, and 

 about 50% by volume of the large crustacean com- 

 munity (crabs, shrimps, stomatopods, and lobsters) 

 taken as prey by the 78 fish species from 28 families 

 that contained large crustacean prey in our NWHI 

 diet studies. Holocentrids also were responsible for 

 about 2.5% of all the individual fish eaten. The frac- 

 tion of the complete food consumption (all prey in 

 the community combined) by this entire fish com- 

 munity that is eaten by holocentrids was about 13- 

 17% (Parrish, unpubl. data). 



Holocentrids also are an important element of the 

 community as prey for other fishes. In the NWHI, 

 4% of all identified fish prey individuals were 

 holocentrids (Norris and Parrish, 1988). In the west- 

 ern Atlantic, Randall (1967) found evidence that 

 seven species of fishes from four families had eaten 

 holocentrids; three species from three families had 

 eaten Myripristinae iMyripristis jacobus). Dragovich 

 ( 1970) also found that postlarval holocentrids (includ- 

 ing Myripristinae) were fairly common prey of skip- 

 jack tuna, Katsuwonus pelamis, and yellowfin tuna, 

 Thunnus albacares, in the western Atlantic. As a 

 widespread and abundant group that is an active 

 predator and vulnerable prey, holocentrids play a 

 major role in the trophic structure of tropical ma- 

 rine ecosystems. 



Reproduction 



For specimens from JA, the results from the three 

 independent analyses of gonads (histology, GSI, and 

 visual examination) indicated that sexual maturity 

 of M. amaena occurs at 153-156 mm SL for females 

 and at 149-156 mm SL for males. These results cor- 

 respond closely to SFR estimates from our specimens 

 collected at Puako: 145-160 mm SL, sexes combined 

 (Hayes et al., 1982). Data from both locations are 

 included in Figures 1 and 2. These values of SFR 

 correspond to about 75-80% of h m (as determined by 

 fitting data from length measurements and otolith 

 increment counts to a von Bertalanffy growth model ) 



and to an age of about six years (Fig. 8 in Dee and 

 Radtke, 1989). Dee and Radtke (1989) aged speci- 

 mens up to nearly 14 years old. Their oldest speci- 

 men (of many available for analysis) was somewhat 

 larger than the L m derived from the regression, so it 

 seems unlikely that many individuals live much 

 longer. Therefore, the age at first reproduction (AFR) 

 is probably about 40% (or a little less) of the maxi- 

 mum lifespan commonly attained, and some individu- 

 als may reproduce for as many as eight years. 



The relation between SFR and maximum body size 

 has been investigated by several workers in a num- 

 ber of locations. The only results reported for 

 squirrelfishes are estimates, based on large sample 

 sizes, for two species of Holocentrinae from the Car- 

 ibbean Sea (Wyatt, 1976). For Holocentrus adscen- 

 sionis, FL at sexual maturity was about 175 mm, 

 asymptotic (maximum) FL about 265 mm, and the 

 ratio about 0.66; for Holocentrus rufus, FL at sexual 

 maturity was about 130-135 mm, asymptotic FL 

 about 230 mm, and the ratio about 0.59. Both these 

 species reach considerably larger sizes than M. 

 amaena, and M. amaena has the largest SFR/L^ ra- 

 tio of the three species. The ratios for these 

 squirrelfishes seem to be in the high portion of the 

 range of published values for tropical fishes (Munro, 

 1974; Loubens, 1980). Myripristis amaena, in par- 

 ticular, matures at an advanced absolute age and at a 

 surprisingly large fraction of its maximum age and size. 



Spawning of M. amaena at JA seems to occur pri- 

 marily in April-May; a secondary peak probably 

 takes place in late September. All specimens collected 

 during the fall peak showed GSI values above the 

 inactive (off-season ) level, but considerably below the 

 mean value for the spring peak (Fig. 5). Although no 

 collections were possible in September, visual exami- 

 nation and GSI data from collections made throughout 

 October 1985 suggested the late stages of a spawning 

 period that probably peaked in late September. Back 

 calculation using the total number of otolith incre- 

 ments counted for the two smallest individuals aged 

 by scanning electron microscope examination (Dee 

 and Radtke, 1989) indicated that one individual was 

 spawned in late September and the other in early 

 October. A spawning peak also was observed in speci- 

 mens collected in January 1986, but not in January 

 1985. The 1986 event may have resulted from un- 

 seasonably calm conditions that occurred during that 

 period. Spawning also was recorded during January 

 1986 for Chaetodon trifascialis, an unusual time of year 

 for that species. Values of the GSI for M. amaena col- 

 lected in January 1986 were generally as high as those 

 of specimens collected during the April spawning peak. 



Data were not collected at Puako in a way that 

 would permit a comparable assessment of seasonal 



