CLARKE: SPAWNING. EGG SIZE, AND DEVELOPMENT OF HAWAIIAN ANCHOVY 



nearly fully developed. In the night samples, all 

 early "2" larvae appeared developed to feeding 

 stage and some had traces of food in the gut; "2" 

 larvae in morning samples were feeding. Thus 

 some of the June larvae may have reached feed- 

 ing status just before sunset or ca. 47 hours since 

 hatching, but many apparently did not develop 

 to this point until after dark and did not start 

 feeding until the next morning or about 57 hours 

 since hatching (Table 1). 



Egg Size 



Dry weight and volume estimates indicated 

 that egg size was about the same in September 

 and June, but ca. 25% and 30% larger in Decem- 

 ber and March, respectively (Table 1). The ratios 

 (Sept. :Dec.:Mar. June) of estimated dry weight 

 per egg were 1:1:21:1.32:1.01. Similar ratios for 

 average yolk volume (1:1.21:1.32:0.99) were 

 closer to those for weights than were the ratios 

 for whole egg volume (1:1.27:1.34:1.08). Yolk 

 volume averaged 79-86% of whole egg volume 

 with no evidence of a trend with egg size. The 

 ratios of egg to yolk for both width and length 

 ranged between 1.05 and 1.09. Except for a 

 value of 2.07 for March yolk length to width 

 ratio, the other length to width ratios for both 

 egg and yolk ranged from 2.16 to 2.26. 



Available data on larval size are few, but indi- 

 cate a positive relationship with egg size. In all 

 months, the newly hatched larvae were about 

 the same length (2.0-2.2 mm), but larvae at or 

 near first feeding status were 3.0-3.7 mm long in 

 March and December as opposed to 2.8-3.0 mm 

 in September and 3.0-3.5 mm in June. Mean dry 

 weights of late "0" and early "1" larvae from 

 September and March were 63% and 65%, re- 

 spectively, of mean egg weights for the same 

 periods, indicating commensurately heavier 

 larvae from the larger March eggs. 



The average weight per egg of hydrated ova 

 taken from 10 females 39-61 mm standard length 

 was 17.8 p-g. The estimates from different indi- 

 viduals ranged between 14.4 and 19.2 ji,g/egg. 

 The value for the smallest female was well below 

 that of the other nine (42.5-61 mm SL); the next 

 lowest value was 16.3 jjig/egg. There was, how- 

 ever, no correlation between average weight per 

 egg and female length for the whole series (P > 

 0.20, Spearman rank correlation coefficient). 



Discussion 



The observed seasonal differences in nehu re- 



productive parameters are not likely owing to 

 genetic differences between seasonal subpopula- 

 tions. Nehu reach spawning size at an age of 3-4 

 months and rarely live as long as six months 

 (Struhsaker and Uchiyama 1976). Thus the pro- 

 geny of, e.g., March spawners would be spawn- 

 ing in July-September rather than the following 

 March. Although annual changes in environ- 

 mental factors in Hawaii are fewer than at 

 higher latitudes, nehu spawn throughout the 

 entire year rather than over a short season, and 

 the observed differences in spawning time, egg 

 size, etc., are most likely responses to changes in 

 temperature, day length, light levels, etc., en- 

 countered over the entire annual cycle. 



The movement of adult nehu to spawning 

 areas and their near synchronous release of eggs 

 are probably stimulated by decreasing light 

 levels in the afternoon and evening, but the tim- 

 ing is not simply related to seasonal changes in 

 day length and time of sunset. The delay be- 

 tween sunset and spawning was less in the win- 

 ter, possibly because light levels in the water 

 column decrease earlier, relative to sunset, in 

 the winter than in the summer owing to differ- 

 ences in solar elevation. It is also possible that 

 responses to light are modified by some other 

 seasonal factor, e.g., temperature. 



Seasonal changes in egg size have been re- 

 ported from apparently the same stock for many 

 other species of fishes. For example, Ware's 

 (1977) data on egg diameters of Scomber 

 scombrus in the Gulf of St. Lawrence indicate 

 that egg volume at the beginning of the spawn- 

 ing season (early June) is about twice that at the 

 end (mid-August). The central population of the 

 northern anchovy, Engraulis mordax, is more 

 similar to nehu in that it spawns year round, and 

 maximum egg volume, which occurs in March, is 

 about 20% greater than the minimum in Septem- 

 ber-October (Hunter and Leong 1981). 



Several mechanisms for within-stock, seasonal 

 changes in egg size have been suggested. Egg 

 size may be related to size of the spawning 

 females, and the seasonal trend in egg size due to 

 the larger females' spawning early in the season 

 and the smaller ones later (Bagenal 1971). The 

 data on ovarian egg weights from spawning nehu 

 indicate no relation between egg size and female 

 size; furthermore, there is no evidence of 

 seasonal changes in size composition of spawning 

 nehu. Clarke (1987) found no difference in size 

 composition between winter and summer spawn- 

 ers examined for fecundity; unpublished data 

 from that study show that size composition of 



597 



