FISHERY BULLETIN: VOL. 87, NO. 3, 1989 



recent regular egg survey, one to three stations 

 were sampled close to the solstices and equi- 

 noxes: 30 September 1984; 27 December 1984; 21 

 March 1985; and 26 June 1985. The hours sam- 

 pled also included expected hatching time for the 

 previous night's eggs for all dates except March, 

 in which case samples were also taken at 0300- 

 0430. The stations were also sampled in the 

 afternoon, 1-2 days after the sunset to midnight 

 series. Times of sunset and sunrise for the above 

 dates were taken from astronomical tables. 

 Water temperatures for each period were taken 

 from the data log at the Hawaii Institute of 

 Marine Biology laboratory at Coconut Island. 

 Hourly surface temperatures for a 2 wk period 

 bracketing the samphng were averaged. 



Nehu eggs and larvae are easily identified. 

 Eggs were usually very numerous (lOO's per 

 sample) and were either all at very nearly the 

 same stage or, for certain times of the day in 

 December and March, readily separated into two 

 age gi'oups: eggs from the most recent spawning 

 and much more developed eggs from the pre- 

 vious night's spawning. There were many fewer 

 larvae in the samples (typically lO's per sample); 

 in several cases larvae from 2 to 4 different sam- 

 ples taken within an hour of each other were 

 combined in order to assess development at a 

 given time of day. Larvae less than 4 mm noto- 

 chord length could be separated into either two 

 or three age gi'oups based primarily upon pres- 

 ence or absence of yolk, pectoral fins and rays, 

 and eye pigmentation. Between appearance of 

 the pectoral fin buds and development of pec- 

 toral rays, age gi'oups could be further discrim- 

 inated using the diameter of the roughly semicir- 

 cular pectoral fin bases relative to that of the 

 pupil or the orbit (neither of which appeared to 

 increase significantly after the pectoral buds ap- 

 pear). Other characters such as the development 

 of the mouth or the gut were correlated with the 

 principal characters used, but were not as useful 

 for qualitative separation. For any given time of 

 day, the least developed larvae were considered 

 "0" group larvae (0-24 h past hatching); the next 

 most developed, "1" group (24-48 h past hatch- 

 ing); and the third, if present, "2" group (48-72 h 

 past hatching). 



Except for the late afternoon samples from 

 June, there was no difficulty in separating small 

 larvae into age/development groups, i.e., larvae 

 in each of the two or three groups present were 

 similar to each other and there were no inter- 

 mediates. In the June series, the apparent oldest 

 group of afternoon larvae showed a broader 



range of several features than was evident for 

 similar stages at other seasons. There was, how- 

 ever, still no overlap with younger stages, and 

 after examination of more larvae from other 

 afternoon samples taken at the same time of the 

 year in 1984 and 1985, it was concluded that 

 these larvae represented only one variably 

 developed age group (see Results section). 

 Larvae larger than 4 mm notochord length were 

 taken infrequently, and it was not possible to 

 estimate age groups among these on the basis of 

 either size of development. 



Volumes and dry weights were determined for 

 undamaged newly spawned eggs from each 

 season. Nehu eggs are ellipsoidal with the major 

 axis about twice the minor. The yolk mass is 

 similarly shaped and separated from the chorion 

 by an obvious pervitelline space. For 40-50 

 freshly spawned eggs from each season (except 

 June when only 29 eggs with no visible embryo 

 were available), the major and minor diameters 

 of both the entire egg and the yolk within were 

 measured to the nearest 0.01 mm using an ocular 

 micrometer at 100 x magnification. Egg and yolk 

 volumes were calculated using the formula for 

 a prolate spheroid. Replicate samples (3 per 

 season) of 20 eggs each were rinsed with distilled 

 water, placed in a preweighed aluminum pan and 

 dried at 60°C. Weights of empty and full pans 

 were determined to the nearest 0.002 mg; weight 

 per egg was calculated by subtraction and divi- 

 sion by the number of eggs in the sample. 



Dry weights were also determined for samples 

 of 10-20 hydrated ovarian ova from spawning 

 nehu taken by purse seine. Preliminary studies 

 indicated that unless ova had already ovulated, 

 they could not be reliably and completely sep- 

 arated from folhcular tissue. The time between 

 ovulation and spawning is apparently very short 

 (Clarke 1987), and few samples of adult nehu 

 contained many fish in this condition. Conse- 

 quently, the only data reported here were from 

 10 females taken from a purse seine collection on 

 1 June 1983. This collection contained the widest 

 size range (39-61 mm SL) of females with ova 

 free in the oviducts. 



Notochord lengths of larvae were measured to 

 the nearest 0.1 mm. There were, however, 

 usually less than 10 larvae of each day class at 

 each different time of day, and statistical com- 

 parisons of average length between seasons for 

 the same stages were not possible. The numbers 

 of undamaged larvae suitable for dry weight de- 

 terminations were even lower. Dry weight de- 

 terminations were made for only three samples 



594 



