NOTE Cooper et al : Natural mortality rate, annual fecundity, and maturity at length for Reinhardtius hippog/ossoides 



299 



oocytes touching the lines of the sampling grid were 

 measured until 50 nonvitellogenic and 50 vitellogenic 

 oocytes had been measured. Nonvitellogenic oocytes 

 were reported as percent frequency, whereas vitellogenic 

 oocytes were reported as number of vitellogenic oocytes 

 per gram of ovary tissue. Because oocyte diameter may 

 vary by ovary location (see "Results" section), all oocyte 

 measurements were taken from the middle location of 

 the ovaries. 



To select ovaries suitable for fecundity estimates, the 

 diameters of six vitellogenic oocytes sectioned through 

 the nucleus were measured from the histological sec- 

 tions by using a microscopic image analysis system. 

 Because oocytes were not perfectly round, oocyte diam- 

 eter was calculated from oocyte area with the following 

 equation: 



Diameter = 2 



Oocyte Area 



Females with the mean diameter of vitellogenic oocytes 

 smaller than 1000 jim were excluded from the fecundity 

 estimate because not all oocytes to be spawned in the 

 current year could be identified (see "Results" section). 

 To exclude fish that had already begun spawning, sam- 

 ples containing POFs or ova were excluded. 



Fecundity was estimated gravimetrically by count- 

 ing vitellogenic oocytes from weighed subsamples of 

 ovarian tissue. The weight of the thick ovarian wall 

 could not be discounted; therefore subsamples were cut 

 with a proportionately weighted piece of ovarian wall 

 attached. This procedure was accomplished by using 

 either an entire cross section of the ovary, or a wedge- 

 shaped sample cut from an entire cross section. Because 

 small oocyte density differences were found between the 

 posterior location and the anterior-i-middle locations, 

 and between the eyed and blind lobes (see "Results" 

 section), four subsamples were taken for each fish. The 

 posterior region of each lobe and the area between the 

 middle and anterior locations of each lobe were sampled. 

 Weighting factors for the posterior and for the middle 

 and anterior locations of the ovary were determined by 

 calculating the proportional mass of these locations of 

 the ovaries for 26 fish. The average proportional mass 

 for the posterior and for the middle and anterior loca- 

 tions was 0.11 and 0.89 (SE: 0.0048), respectively. When 

 divided by two (to account for the two ovary lobes), the 

 average weighting factors for the posterior blind and 

 eyed lobes = 0.06, and the average weighting factor of 

 the anterior and middle locations for the blind and eyed 

 lobes = 0.44. Fecundity of each female was estimated 

 from the four subsamples by using the fecundity equa- 

 tion with weighting factors from Nichol and Acuna 

 (2001) to account for the smaller proportional mass of 

 the posterior sections. 



The fraction of vitellogenic oocytes undergoing atre- 

 sia was estimated for 174 fish from histological cross 

 sections. The total number of vitellogenic oocytes and 

 the number of atretic vitellogenic oocytes (Hunter and 

 Macewicz, 1985) for each histological section were 



counted. The fraction of vitellogenic oocytes undergo- 

 ing atresia was estimated as the number of atretic 

 vitellogenic oocytes divided by the total number of vi- 

 tellogenic oocytes. 



Length at maturity 



Maturity was determined histologically, by using all fish 

 40 cm and larger, and a subsample (n = 6) of females less 

 than 40 cm. Maturity stages were determined according 

 to the method of Gundersen (2003). 



Results 



Length-GSI relationship and 

 Instantaneous rate of natural mortality 



The conversion factor between ovary fresh weight and 

 weight after fixation and storage in formalin was: 



Fresh = (1.437 x Formalin) + 6.8276, 



(71=28, r2 = o.9903), 



where Fresh = fresh ovary weight (g); and 



Formalin = ovary weight after storage in formalin (g). 



Maturity stages increased from spring (March) to 

 summer (June and July) and again to autumn (Sep- 

 tember and October) (Fig. 3). In March, mature females 

 were spawning, spent, or beginning vitellogenesis for 

 the next spawning season. Most fish were in the vi- 

 tellogenesis-1 stage. By June-July, the ovaries had 

 advanced until most were in the vitellogenesis-2 stage, 

 although some were also in stages vitellogenesis 1 and 

 3. By September-October, the majority of mature ova- 

 ries were in vitellogenesis-3 stage, although some had 

 progressed to vitellogenesis 4 and a small percentage 

 (4%) of the mature females collected in September were 

 ready to spawn (presence of hydrated oocytes) or showed 

 signs of recent spawning (presence of ova or POFs). 

 The largest mean oocyte diameter (« = 6 for histological 

 examination) for any female with oocytes before hydra- 

 tion was 1720 ftm. 



The mean length of mature females was 79.2 cm dur- 

 ing 1977-87, and the GSI data were derived from fe- 

 males smaller and greater than 79.2 cm. The GSI was 

 very poorly correlated with length (^^=0.04), and we 

 used the mean GSI (0.063, 0^=0.000018) in our sample 

 to obtain a natural mortality estimate of M = 0.112. 

 The variance of M (VarM, 0.0002) was estimated ac- 

 cording to the method of Gunderson et al. (2003) with 

 the following equation: 



VarM = {GSI fVar(k) + k^Var{GSiX 



where Va r M = the variance of mortality; 

 GSI = mean value of GSI; and 



/fe= constant from the GSI-M regression 

 (Gunderson, 1997). 



