354 



Fishery Bulletin 104(3) 



W = aL'', 



where W = the total fish weight in grams; and 

 L = total length in cm. 



This relationship was also determined for ovary-free 

 weight. Because fresh ovarian weights were not mea- 

 sured for samples collected at sea, the fresh weight was 

 estimated from a linear regression between fresh and 

 preserved weights from the samples collected at pro- 

 cessing plants. The linear regression OW=1.127 (OWA- 

 0.024 was used to convert formalin-preserved ovary 

 weights, OWr, to estimates of fresh ovary weights, 

 OW (0=446, ^2=0.997, P<0.0001). When fresh and 

 preserved ovary weights differed; fresh ovaries usu- 

 ally weighed more than ovaries preserved in formalin, 

 and the maximum difference between fresh ovaries 

 and formalin-preserved ovaries was 20%. 



Parameters of the von Bertalanffy growth model 

 (L^, k, and t^) were estimated for female rex sole in 

 the GOA by using nonlinear least squares regression 

 with the equation 



L, = L,a- 

 where L, is length at age t. 



-k (/ - l„) 



All length and age data were used for the GOA 

 growth model. Statistical comparison of growth rates 

 for female rex sole in the GOA with those from the 

 Oregon coast were less rigorous because Hosie and 

 Horton (1977) used mean values of length at age to 

 calculate the von Bertalanffy growth model, thereby 

 decreasing the variability in length at age, which was 

 not presented. Therefore, the parameters L^, k. and 

 tg in the GOA growth model were each statistically 

 compared with the parameters in the Oregon growth 

 model for females age 1-15 years (data from Hosie and 

 Horton, 1977) with three separate Z-tests (Zar, 1999). 

 Alpha was set at 0.05 for all tests of significance. 



Results 



Maturity 



Oocyte development and maturity stage were deter- 

 mined for 594 female rex sole. Histological examination 

 revealed that 46% of the collected females were imma- 

 ture and 54% were mature. Both immature and mature 

 females were collected in every month; however, no fish 

 were sampled in August or September (Table 2>. Year- 

 round histological analysis of ovaries confirmed that rex 

 sole are batch spawners, as seen in the presence of POFs, 

 features associated with recent spawning, concurrent 

 with advanced vitellogenic oocytes. 



The spawning season of rex sole in the GOA was 

 protracted; postovulatory follicles first appeared in 

 October and hydrated oocytes occurred as late as May, 

 indicating a spawning season that spans at least eight 



Female rex sole TL = 287-420 mm 



1.0 

 08 

 06 

 04 

 S 0-2 



X 



0) 



to 



E 



2 10 



g 



^ 



 j 



Jan Feb Mar Apr May Jun Jul Oct Nov Dec 



Female rex sole TL >420 mm 



ill 



Jan Feb Mar Apr May Jun Jul Oct Nov Dec 

 D Nonspawning  Poslspawmng H Spawning D Immature 



Figure 3 



Monthly proportion of female rex sole (Gtyptocephalus zachi- 

 rus) maturity stages graphed separately for fish greater 

 than size at first maturity (287-420 mm in upper graph) 

 and fish greater than size at lOC/r maturity (>420 mm in 

 lower graph). No data were collected in August or Septem- 

 ber. All mature rex sole were in the postspawning stage 

 in June and Julv. 



months. Active females were present from October 

 through May and were absent in June and July (Ta- 

 ble 2). Mature postspawning females with no healthy 

 AY oocytes first appeared in April and increased in 

 June and July; thus, the spawning season ended in 

 April for some individuals, and by June all rex sole 

 had completed spawning (Fig. 3). The high number 

 of postspawning fish combined with the absence of 

 spawning fish in June and July indicated that the du- 

 ration of POFs in an ovary was several weeks. Active 

 females had at least one of the following oocytes (AY, 

 MN, or HY) and could have any combination of these 

 advanced stage oocytes, or all three oocyte stages at 

 once. These oocytes were large and stretched the ovary 

 wall quite thin, whereas inactive postspawning females 

 had shrunken ovaries and thicker ovary walls (Mad- 

 dock and Burton, 1999). In May and July ovary walls 

 in mature females were significantly thicker than 

 in all other months sampled (ANOVA: Fg .^-^ = 28.64, 

 P<0.0001; Fig. 4), corresponding to the high fraction 

 of postspawning females present from May to July 

 (Table 2, Fig. 3). 



Female rex sole sampled for length and maturity 

 ranged from 166 to 552 mm TL (Fig. 5). The smallest 



