WILLIAMS and CLARKE: BIOLOGY OF THE GOLD SPOT HERRING 



of 225-242 ova/g ww and 791-794 ova/g dw, for the 

 observed ranges of fish weights, while actual values 

 ranged from 160-311 ova/gww and 502-1,102 ova/g 

 dw t . There was no trend between relative fecundity 

 and size. 



Age and Growth 



The holding experiments (Fig. 5, Table 1) provided 

 strong evidence that growth increments on the 

 otoliths are deposited daily. In the first experiment, 

 there was a wide range of size and otolith ring counts 

 in the initial and subsequent subsamples. Probably 

 because of this, the increase in average number of 

 rings agreed closely with the numbers of days elapsed 

 for only two of the subsamples; otherwise, only a 

 general trend for increase in average ring count was 

 evident. In the second experiment, the fish were 

 smaller and consisted primarily of two distinct 

 groups separated by 3 mm SL and five otolith rings. 

 These two groups were apparent in all the subsam- 

 ples, and the increase for each in number of rings cor- 

 responded closely with the number of days elapsed. 

 Since each group was about equally represented, the 

 average difference in number of rings for the whole 

 experiment also correlated closely with the number 

 of days. 



Both the fish and their otoliths grew faster in the 

 third and fourth experiments, probably due to the 

 higher feeding rate. The distance between rings was 

 markedly greater and counts much easier to make 

 than in the first two experiments. Fish in the initial 

 subsample for the third experiment had a wide range 

 of sizes and ring numbers, but a single group of fish 

 with 26 rings dominated the initial subsample (16 of 

 41 fish). This group was apparent in most of the rest 

 of the subsamples. The increase in number of rings in 

 this group and differences in means for the subsam- 

 ples corresponded closely with number of days. The 

 results of the fourth experiment were similar. The 

 subsample which deviated most from the predicted 

 increase had a much narrower range of ring counts 

 and was apparently made up mostly of the younger 

 fish in the experiment. 



Initial collections for the third and fourth ex- 

 periments were made on consecutive days (31 July 

 and 1 August 1979) from the same location and from 

 what appeared to be the same school of transforming 

 juveniles. In the initial subsample for the third ex- 

 periment, the dominant group had 26 rings, while in 

 that for the fourth, 13 of 39 had 27 rings (Fig. 5). 

 Similarly, the lowest number of rings in the initial 

 subsample for the third experiment was 23 (5 fish), 

 and that for the fourth experiment was 24 (6 fish). 



55 



45 



35 



25- 



<S> 



z 



or 



Ll_ 



o 

 o 



OCT 16, 1978 



55 



45 



35- 



25- 



15 



55 



45 - 



35- 



25 



15 



-1 1 1 - 



—i 15 



NOV 17, 1978 



• »-2 



-1 1 I 1 1 1 1 



10 20 30 10 20 30 



JULY 31, 1979 • 



• -2 



: ,-\ - 3 



• 3-J «-2 



" «-3 

 16 -• 



10 20 30 10 20 30 



NO OF DAYS ELAPSED 



FIGURE 5. — Relationship between number of otolith rings and 

 elapsed time for subsamples of juvenile Herklotsichthys quad- 

 rimaculatus from holding experiments started on four dates, 

 1978-79 (see text for details). Numbers of specimens are given for 

 points representing more than one fish per subsample. 



Table 1. — Number of days elapsed and mean increase in the num- 

 ber of otolith rings in subsamples of Herklotsichthys quad- 

 rimaculatus from holding experiments started on four different 

 dates, 1978-79, at Kaneohe Bay, Oahu, Hawaii. 



Starting date/ 

 No. days elapsed 



No. in subsample 

 (Size range, mm SL) 



Increase in mean 

 increment count 



16 October 1978 



Two fish in the initial subsample for the third experi- 

 ment had higher counts than the maximum for the 



593 



