FISHERY BULLETIN: VOL. 85, NO. 1 



samples at approximately the same times on other 

 days; consequently, the data from these "enclose 

 and hold" samples were pooled with the others for 

 all analyses. 



Time of collection was recorded as the beginning 

 of the set of the net; usually 15-30 min elapsed 

 before the sample was actually preserved. For sam- 

 ples taken from skipjack tuna vessels, the time of 

 collection was often only known within ±15 min and 

 the delay between collection and preservation of the 

 sample was often somewhat longer than 30 min. 



For analyses of oocyte development rate and 

 spawning frequency, collection time was adjusted 

 to hours since the most recent spawning. Data on 

 appearance of newly spawned eggs in the plankton 

 (Clarke unpubl. data) indicate that spawning begins 

 1-2 h after sunset and is nearly over in about an 

 hour; the delay after sunset is greatest during the 

 summer. For samples considered here, spawning 

 time was assumed to be 1 h after sunset for dates 

 between mid-October and the end of April and 2 h 

 after sunset for the remainder of the year. Given 

 the frequent uncertainty in actual time of capture, 

 this crude correction for spawning time was satis- 

 factory for the purposes of the present study. 



All specimens were preserved and held in ca. 4% 

 formaldehyde/seawater solution. The recently col- 

 lected samples were held at least 1 wk before mea- 

 surement and further analyses; by this time most 

 shrinkage in length had occurred. Although the 

 older samples had been in preservative for several 

 years, there was no evidence that long-term storage 

 had affected any parameters considered here, e.g., 

 length-weight relationships were similar for both 

 recent and older samples. 



For each sample, standard length (SL) of all or 

 a subsample of ca. 100 specimens was measured to 

 the nearest mm. Individuals for further examina- 

 tion were selected from throughout the size range 

 of nehu >35 mm SL in the sample. The selected in- 

 dividuals were measured to the nearest 0.5 mm, 

 opened, and the gonads examined under a dissect- 

 ing microscope. Females were classed as immature 

 —ovaries translucent and maximum oocyte length 

 <0.40 mm; mature— ovaries mostly opaque, oocytes 

 visibly yolked and over 0.40 mm; or hydrated— 

 mature and at least some oocytes with translucent, 

 globular yolk and the peri vitelline space visible. For 

 mature females the length of the apparent largest 

 oocyte was estimated to the nearest 0.1 mm using 

 an ocular micrometer. 



To determine oocyte size frequency of mature 

 females, a portion of the ovary was teased apart on 

 a glass slide, placed under a compound microscope 



at 100 X, and the lengths of oocytes over 0.40 mm 

 measured to the nearest 0.01 mm until 20-30 of the 

 largest oocytes were measured. Spawned nehu eggs 

 are ellipsoidal with the length about twice the width 

 (Yamashita 1951). Oocytes >0.3-0.4 mm are also 

 elongate but are more variable in shape. "Length" 

 as used here refers to the maximum dimension. Ex- 

 tremely elongate (length to width ca. 3 or more) and 

 nearly round (length to width less than ca. 1.5) 

 oocytes were noted as was the relative opacity of 

 each oocyte measured. These observations were 

 necessary in many cases to separate nearly round, 

 heavily yolked oocytes that belonged to an advanced 

 mode from very elongate, more nearly translucent 

 oocytes of the same "length" that clearly belonged 

 with a less developed mode. 



As reported by Leary et al. (1975), mature female 

 nehu may carry 0-2 separate size-frequency modes 

 of oocytes. If a distinct advanced mode of oocytes 

 was evident from the measurements and associated 

 notes, the maximum, minimum, and median lengths 

 of oocytes in this "largest" mode were used for sub- 

 sequent analyses. These parameters will be abbre- 

 viated as LMX, LMN, and LMD, respectively. If all 

 ova in the most advanced mode were hydrated, all 

 lengths were arbitrarily assigned a value of 1 mm. 

 If the largest mode was incompletely separated from 

 smaller oocytes, only LMX and an estimate of LMD 

 were recorded; if there was no separating mode evi- 

 dent, LMX (the largest oocyte in the subsample) was 

 the only datum recorded. If an advanced mode was 

 present and a second or "next" mode was also sep- 

 arated from yet smaller oocytes; the maximum, 

 minimum, and median lengths of oocytes in the next 

 mode will be abbreviated NMX, NMN, and NMD. 

 In most females, however, the next mode was either 

 only partially separated or not evident and, similarly 

 to the case for unseparated advanced modes, only 

 NMX and an estimate of NMD or only NMX, the 

 largest oocyte not in the advanced mode, could be 

 recorded. 



For 107 specimens for which size-frequency mea- 

 surements were made from a sample of the right 

 ovary, the left ovary was prepared, sectioned, and 

 stained with eosin/hemotoxylin as described by 

 Hunter and Goldberg (1980). The slides, identified 

 by only a code number, were examined for presence 

 of postovulatory follicles (POF). 



For determination of batch fecundity and dry 

 weight, the fish was first rinsed with distilled water. 

 The ovaries were removed and placed on a clean 

 glass slide. Oocyte size frequency was determined 

 as described above. If a distinct mode of advanced 

 oocytes was present and oocytes in this mode could 



128 



