308 



Fishery Bulletin 98(2) 



1.2 



1.0 - 



0.8 - 

 0.6 



0.4 - 



0.2 - 



O 



f- 0.0  



Fal ,Win , Spr .Sum, Fal .Win , Spr .Surrii F al , Wi^.^fr, ^""i 

 -Size at birth • w^ o o^ 



Fal Win Spr Sum Fal Win Spr Sum 



D 



-^- 



Spent ovary 



EC 



, EC 



large E. spent ovary 



□ o 



i 



Fal Win Spr Sum Fal Win Spr Sum Fal Win Spr Sur 

 Time (seasons) 



Figure 3 



Gestation, parturition, and reproductivt- cycle of Isurus o.xyrinchuK. (A) The 

 relationships between length of embryos and age-O-i- fish, and time of cap- 

 ture. Regression with 95'f confidence band for the line of combined embryo 

 data from Northern and Southern Hemispheres is shown. • = embryos; 

 O=age-0-^ fish. (B) Temporal uterus width index lUWI) of mature females 

 based on a 3-year reproductive cycle. The dotted line was fitted to the 

 data by eye. (C( Temporal L'WI of mature females based on a 2-year repro- 

 ductive cycle. (D) Temporal gonadosomatic index (GSIi based on a 3-year 

 reproductive cycle. The dotted line shows suggested decrease of GSI during 

 last third of gestation. i> = pregnant; n = mature, not pregnant; E=embryos 

 present; EC=egg cases with single ovum present i. 



ova (Table 2, nos. 1 and 2) are more than 

 one year apart from the near-term litters 

 (on the basis of the Northern Hemisphere 

 data), the regression predicted a 19-mo. 

 gestation period (n = 15,slope=3. 7 cm/mo., 

 SE=0.3 cm/mo., P<0.001, ^2=0.92). It is 

 also possible to fit these data with a 

 regi-ession that predicts a 6.7-mo. ges- 

 tation period (7! = 15, slope=10.4 cm/mo., 

 SE=1.8 cm/mo., P<0.001, /? = 15, r-=0.T3). 

 The intercept and slope of Northern 

 and Southern Hemisphere regressions, 

 assuming a gestation longer than 12 

 months, were not significantly different 

 (P=0.8 and 0.7, respectively) and the 

 combined regression was 



Einbryo TL (cm)=3.71 (SE-0.27)T, 



(P<0.001, r~0.89 7!=26) 



where T = the number of months since 

 midfall. 



It predicted a gestation period of 15-24 

 months ( 19 for the line) with 70 cm for 

 the size-at-birth (Fig. 3A). However, we 

 suggest the shorter estimate is more 

 realistic because parturition appears 

 to be a little earlier than what was 

 indicated by the regression. The GM 

 regression line (not shown) was slightly 

 steeper than the y-on-x regression and 

 predicted a gestation of 17.8 months. 

 The GM regression reflects the func- 

 tional relationship better because of the 

 uncertainties of several dates. There- 

 fore, we concluded that the best esti- 

 mate for the gestation is 15-18 months. 

 When the three early-term litters with 

 fairly high leverage (L=0.2-0.35) were 

 excluded, the gestation was still around 

 18 months based on a significant regres- 

 sion (P=0.001) with a slope of 3.8 

 cnVmonth (SE= 0.6, n=23}. 



Reproductive cycle 



A three-year reproductive cycle provided 

 the best fit to the temporal UWI data. 

 UWI increased in early-term females 

 and reached a maximum of 7-8*^^^ at par- 

 turition (Fig. 3B). The pregnant female 

 with the 52-cm embryos and UWI of 

 6.8'r would be expected to give birth 



