FISHERY BULLETIN: VOL. 81, NO. 3 



O 









10 



6 8^ 



CO 

 Q 



o 

 o 



4H 







MALES 



F 



r 



l . r 



X 



T I 



-i 1 1 r - 



FEMALES 



"I 1 1 1 1 1 1 1 1 1 1 1 — 



JFMAMJJASOND 

 MONTH 



FIGURE 3. — Relationship between gonad/somatic weight and month 

 of capture for adult (>80 mm SL) male and female Herklotsichthys 

 quadrimaculatus, Kaneohe Bay, Oahu. 



quesan sardine (see their figure 5). In addition to 

 transparent ova less than ca. 0.10-0.15 mm in 

 diameter, the ovaries of most adult female herring 

 also contained partially yolked, semi-opaque ova ca. 

 0.16-0.30 mm and completely yolked, opaque ova 

 >0. 30-0. 35 mm. Distinctly separated size-frequency 

 modes of opaque ova were found from 0.30-0.40 mm 

 to 0.60-0.90 mm. No larger, hydrated ova were found. 

 The small, partially yolked ova were usually con- 

 tinuous with the smaller transparent ova; but in some 

 fish, they formed a partially separated mode — often 

 with opaque ova at the large end. Such modes, which 

 were never completely separated from the smaller 

 ova, were found in fish both with and without a 

 separate mode of larger, opaque ova. Although a few 

 females without a separate advanced mode had 

 somewhat flaccid ovaries, we found no atretic ova nor 

 any other evidence that these or any of the fish had 

 already spawned. There was no indication of syn- 

 chronous spawning or any short-term cycle; the large 

 samples from the middle of the spawning season in- 

 cluded females with ova in a wide range of sizes and 

 stages of development. 



In the 46 females (80-121 mm SL) with a separate 

 advanced mode, the batch fecundity or number of 

 ova in that mode ranged from 1,155 to 6,296 (Fig. 4). 

 The relationships of batch fecundity to length or 

 weight as determined by least squares linear regres- 

 sion were 



in Kaneohe Bay in all months. The juveniles, 

 however, were decidedly more abundant and more 

 frequently recorded during the summer and fall as 

 would be predicted by a summer peak in spawning. 

 Our findings on ova development and size frequen- 

 cy in the gold spot herring are essentially the same as 

 those of Nakamura and Wilson (1970) for the Mar- 



F =7,518.0 + 110.8 SL(mm) 

 F = 172.0 + 218.4 ww (g) 

 F = -10.2 + 795.6 dw,(g) 

 F = 21.8 + 842.1 dw s (g) 



(r 2 = 0.84), 

 (r 2 = 0.79), 

 (r 2 = 0.80), 

 {r 2 = 0.77). 



Fecundity was clearly correlated with some measure 

 of size, but not very precisely with any of them. The 

 appropriate equations predicted relative fecundities 



7,000 



6.000 



5.000 



| 4.000 



3.000 



2.000 



1.000 



7000 

 •DIRECT COUNTS OF TOTAL OVARY 



o ESTIMATES BASED ON OVARY , 

 SUBSAMPLE 6 000 



5.000 



4,000 - 



3.000 



2,000 



1,000 



80 90 100 110 120 



STANDARD LENGTH (mm) 



7 9 11 13 15 17 19 21 23 25 27 

 WET WEIGHT (g) 



FIGURE 4. — Relationship between batch fe- 

 cundity and length and wet weight of female 

 Herklotsichthys quadrimaculatus, Kaneohe 

 Bay, Oahu. Straight lines are drawn from 

 regression equations (see text). 



592 



