102 



Fishery Bulletin 90(1), 1992 



spawning because the standing stocl< of advanced 

 yolked oocytes is not replaced during the spawning 

 season. 



Indeterminate annual fecundity Annual fecundity 

 is indeterminate when the potential annual fecundity 

 of a female is not fixed prior to the onset of spawning 

 and unyolked oocytes continue to be matured and 

 spawned during the spawning season. 



Batch fecundity Number of hydrated oocytes re- 

 leased in one spawning; usually determined by count- 

 ing the number of hydrated oocytes in the ovary. 



Relative fecundity Fecundity divided by female 

 weight. 



Reproductive states 



Active Females capable of spawning at the time of 

 capture or in the near future (by the end of the survey 

 or of a season, or other temporal end point). Ovaries 

 of active females contain sufficient number of yolked 

 oocytes for a spawning. 



Inactive Females not capable of spawning at the time 

 of capture nor in the near future, although some may 

 have been mature in the past. 



IViature Females that have spawned in the current 

 reproductive season or can be expected to do so. 



Immature Females that have not spawned in the ciu-- 

 rent reproductive season nor can be expected to do so. 



The central methodological issue in fishes with deter- 

 minate fecundity (Hunter and Macewicz 1985a, Hor- 

 wood and Greer Walker 1990) is to establish that poten- 

 tial annual fecundity is an unbiased estimate of annual 

 fecundity. For this to be true in Dover sole requires 

 four key assumptions. The first and most important 

 assumption is that fecundity is determinate in Dover 

 sole. This means that potential annual fecundity 

 becomes fixed before spawning begins. Estimation of 

 the standing stock of advanced oocytes (total fecundity) 

 is meaningless if, during the spawning season, oocytes 

 are added to that stock. 



The second assumption is that the potential annual 

 fecundity is equivalent to annual fecundity. Strictly 

 speaking, this probably never happens because in any 

 fish population some of the females resorb some of their 

 advanced yolked oocytes rather than spawn them, a 

 process known as atresia. If many females resorbed 

 many of their advanced oocytes, potential annual fecun- 

 dity would be a serious overestimate of annual fecun- 

 dity in the population. In addition, not all ovulated 

 oocytes are spawned; a few remain in the ovigerous 

 folds of the ovary after spawning and are later re- 



sorbed. Retention of ovulated oocytes is probably 

 seldom a serious bias. 



The third assumption is that the females used to 

 estimate potential annual fecundity have not spawned 

 during the current reproductive season. Dover sole 

 females that have spawned some of their stock of ad- 

 vanced oocjftes cannot always be distinguished from 

 those that have not begun spawning. Inclusion of par- 

 tially spawned females in an estimate of potential 

 annual fecundity of the population could be a signifi- 

 cant bias. 



The fourth assumption is that one is able to identify 

 with certainty the oocytes that constitute the poten- 

 tial annual fecundity. An ovary may not be sufficient- 

 ly developed to identify all of the oocytes destined to 

 be spawned. On the other hand, if the ovary is highly 

 advanced, spawning may have begun and some ad- 

 vanced oocytes lost. Clearly, an optimal range of 

 ovarian development exists where these risks are 

 minimized. 



In addition to evaluating the above four assumptions 

 (determinate fecundity, atresia, spawning, and im- 

 maturity) we consider several other methodological 

 issues related to assessment of fecundity and female 

 sexual maturity. These issues are (1) validation of our 

 gross anatomical and histological classification of 

 ovaries into active or inactive and mature or immature 

 states; (2) four precision issues related to total fecun- 

 dity estimates (number of tissue samples per ovary, 

 number of females, location of ovarian tissue samples, 

 and within-trawl and between-trawl variance); and 

 (3) an evaluation of bias in the assessment of female 

 sexual maturity. 



Methods 



Collections and shipboard measurements 



Dover sole were collected along the central California 

 coast (Point Conception to San Francisco Bay) during 

 six research trawl cruises (Table 1). Dover sole were 

 taken off the Oregon coast between Cape Lookout and 

 Heceta Head during two cruises in 1988-89; miscel- 

 laneous collections provided by E. Pikitch off the 

 Oregon coast in 1985 and 1986 were also used. Re- 

 search trawls were one-half hour or one hour long, 

 depending on depth. In central California waters, we 

 used a 400-mesh Eastern trawl (mouth opening ~15m 

 wide and 1.5 m high; Wathne 1977). In Oregon waters, 

 either an Alaska Fisheries Science Center (AFSC) 

 modified 5-inch mesh, 90/120, high-rise "poly 

 Nor'Eastern" trawl (fishing dimensions ~4.6m high 

 and 13.5 m wide at wing tips), a 5-inch mesh, 92/83, poly 

 Nor'Eastern trawl, or a 5V2-inch mesh, 75/90, high-rise 

 Aberdeen trawl was used. Up to 100 Dover sole from 



