282 



Fishery Bulletin 104(2) 



MPCL = -alb (a=intercept, 6 = slope) (Mollet 

 et al., 2000; Neer and Cailliet, 2001; Con- 

 rath and Musick, 2002; Conrath, 2004). To 

 estimate percent maturity by length for the 

 ENP salmon shark population, the 'a' and 

 '6' estimates (from the "GLM" model) were 

 substituted into the equation; percent ma- 

 ture = 100x[(e'"+*'<'''"^"")/(l+e"'+*'<'''"«"")], and 

 plotted against PCL. A nonparametric boot- 

 strap (n = 1000 replications) was conducted 

 with S-Plus and 95*^ confidence limits were 

 obtained from the 2.5th and 97.5th empirical 

 percentiles. 



Sperm storage in the oviducal gland has 

 been documented in several shark species 

 (Pratt, 1993). Oviducal glands were taken 

 from six mature females, ranging in size 

 from 180 to 192 cm PCL, caught in PWS 

 waters in late July and late August, to de- 

 termine the presence or absence of sperm. 

 Samples were initially fixed in 10% forma- 

 lin. Samples were rinsed extensively to re- 

 move fixative and then stored in 70% etha- 

 nol. Cross-sections along the entire length 

 of each gland were trimmed, dehydrated in 

 a graded series of ethanol, cleared in a limonene-based 

 solvent, infiltrated with paraffin, and embedded in 

 paraffin blocks. Serial sections (5 um) were prepared 

 by using a rotary microtome, mounted onto poly-L-ly- 

 sine-coated slides, dried, deparaffinized, rehydrated, 

 and stained by using standard Harris hemotoxylin and 

 eosin in order to examine general cellular detail (Hin- 

 ton, 1990). Sections were examined for the presence of 

 sperm by using a compound microscope at magnifica- 

 tions ranging from 100 x to 1000 x. 



Results 



Length and weight equations 



One hundred sixty-six of the 182 salmon sharks in this 

 study were female, resulting in a F:M sex ratio of 10.4/1. 

 Although females dominated our sample, the following 

 equations for converting PCL into FL and TL appear to 

 work equally well for both sexes. 



FL = 1.08 X PCL + 6.91 

 TL = 1.15 xPCL + 15.19 



ir- 

 (r- 



= 0.99; n-- 

 = 0.97; n-- 



:138) 



133). 



Weight-length relationships for female and male 

 salmon sharks we sampled in the ENP were W= 

 8.2xl0-05xL2 "9 (r-' = 0.99, ??=20) and W=3.2x 

 10-06xL3383 (;.2 = o.99, n=l) respectively, and 4.4x 

 10-°^xL2875 (^2_o.99) for the sexes combined. 

 A likelihood ratio test showed that separate equa- 

 tions for each sex describe the data better than a 

 single equation for the sexes combined (x^ = 12.1; df=3; 

 P<0,01). 



Vertebral analysis 



There was a slight curvilinear relationship between cen- 

 trum radius (CR) and shark PCL (Fig. 2). A linear regres- 

 sion gave a significant fit to the data (PCL = ( 10.553 x 

 Ci?)-h20.964; r- = 0.90; P<0.0001); however, a quadratic 

 equation produced a slightly better fit ( PCL = -63. 944 -i- 

 (25.189xCR)-(0.583xCi?2); ^-' = 0.94; P<0.0001) and a 

 t-tesi showed the third parameter to be statistically 

 significantly different from zero (/ = 10.32; df=181, 

 P<0.0001). However, it was necessary to compare the 

 mean back-calculated results from Equations 1 through 

 4 with the mean sample PCL data to see if the slightly 

 better statistical fit of the quadratic equation translated 

 into better biological accuracy for modeling growth. 



Percent agreement (PA) among readers was 68.1% for 

 the first set of blind reads and 72.0% for the second set, 

 and the within-reader PA was 79.7%f for reader one (the 

 lead author) and 75.3%c for reader two. Percent agree- 

 ment (±one yr) was >95%f for all reader comparisons. 

 Agreement between and within reader age assessments 

 was 100% until age 7 or 8, depending on the reader 

 comparison. When grouped by 10-cm length increments, 

 percent agreement was 100%^ for sharks <160 cm PCL, 

 and 100% ±1 yr for sharks ^180 cm PCL (Table 1). (PA 

 and PA ±1 for length groups were virtually the same 

 for both sets of between-reader comparisons, but slight 

 differences occurred above 160 and 180 cm PCL, re- 

 spectively.) The chi-square tests of symmetry gave no 

 indication that differences between and within readers 

 were systematic rather than due to random error (x^ 

 test, P>0.05 in all cases). 



Relative marginal increment (RMI) analysis showed 

 that postnatal rings form annually between January 

 and March. The smallest relative margins occurred in 



