154 



Fishery Bulletin 97(1), 1999 



Materials and methods 



Sampling 



Sharks were sampled in NSB nets from 1984 to 1995. 

 Each net was 214 m long, 6 m deep, had a 50-cm 

 stretched mesh, and was set in water 10-14 m deep, 

 parallel to and 300-400 m from shore. For additional 

 details of the netting operation see Cliff etal. (19881. 

 Precaudal length (PCD was measured in a straight 

 line from the snout tip to the precaudal notch and is 

 used throughout this study, unless indicated other- 

 wise. To compare our findings with those reported in 

 the literature, the following equations were used to 

 convert lengths: 



Total length (TL) = 1.251 PCL + 5.207 



(n =36; range 131-307 cm PCL; 95% confidence lim- 

 its on slope: 1.233 and 1.268; r-=0.9984) (Cliff et al., 

 1996a); and 



PCL = 0.8550 TL - 0.0955 



in =58; range 96—447 cm PCL; 95'^ confidence limits 

 on intercept: -0.130 and -0.061; r-=0.996) (Mollet 

 and Cailliet, 1996). 



Lengths were converted by means of the method con- 

 sidered most appropriate. 



Mass was determined by weighing each shark and 

 subtracting the mass of gut contents where they ex- 

 ceeded 1 kg. Maturity was assessed by the criteria of 

 Bass et al. (1973), where males were considered 

 mature only if their claspers were fully calcified. In 

 females, maturity was based on the presence of dis- 

 tinct ova in the ovary and uteri, which had expanded 

 from a thin, tubelike condition to form loose sacs 

 (Bass et al., 1973; Cliff et al., 1988). Vertebral samples 

 were taken anterior to the origin of the first dorsal 

 fin from 61 females ( 128-297 cm) and 53 males ( 142- 

 373 cm). Mass range was 42-442 kg for females 

 («=60) and 46-882 kg for males («=53). Vertebrae 

 were stored frozen (60%) or in 70% isopropyl alcohol 

 (31%), or dried (9%). Individual centra were cleaned 

 by removing the connective tissue from the corpus 

 calcareum with forceps. Dried samples needed addi- 

 tional soaking in a 5% solution of sodium hypochlo- 

 rite for 20-40 minutes. 



Ring counts 



X-radiography was used to enhance the visibility of 

 the growth rings. X-radiographs of whole centra were 

 prepared on an Odel Pollux 700 generator with a 



Comet tube by using Agfa Ortho (Extremity) film and 

 were processed with an Agfa Curix 160 processor 

 We x-rayed, using mostly oblique exposure, all cen- 

 tra with the corpus calcareum (Figs. lA and 2A) fac- 

 ing the tube at a set distance of 100 cm. At 50 mA, 

 exposure times ranged from 0.3 to 0.8 seconds and 

 voltage from 28 to 34 kV. The x-radiographs were 

 then scanned with an Agfa Arcus II scanner and 

 Adobe photoshop software. Ulead ImagePals 2 GO! 

 (Ulead Systems, 1992-94) and CorelDRAW! (Corel 

 Corp., 1993) were used to enhance and work with 

 the images. 



A growth ring (GR) was defined as a band pair, 

 composed of one calcified (opaque) and one less-cal- 

 cified (translucent) band (Fig. lA). The finer, nar- 

 rower rings (circuli), also observed by Cailliet et al. 

 ( 1985 ), were not used for ageing purposes. The angle 

 change on the centrum face (Fig. 2A), a result of the 

 difference between fast intrauterine and slower post- 

 natal growth (Walter and Ebert, 1991), was regarded 

 as the birth mark. 



Counts were made directly from the x-radiographs 

 (XRs) by one reader (A) and from the scanned im- 

 ages (SCs) by two readers (A and B). These three 

 ways of viewing the vertebrae will be called meth- 

 ods XR-A, SC-A, and SC-B, for brevity Each reader 

 made three nonconsecutive GR counts, without 

 knowledge of the shark's length and previous counts. 

 Count reproducibility was determined by using the 

 following four methods: 



1 The average percentage error (APE) as described 

 by Beamish and Fournier ( 1981) in which an up- 

 per limit in the APE was arbitrarily set at 20% 

 for each vertebra (samples were discarded if, af- 

 ter a recount, they were still above this limit) and 

 a final APE index was recalculated and an 

 intrareader comparison (XR-A vs. SC-A) and 

 interreader comparison (SC-A vs. SC-B) of APE 

 values were then conducted; 



2 The index of precision D ( Chang, 1982); 



3 The percentage agreement among the three 

 counts for each method; and 



4 The percentage agreement in paired GR counts 

 between the methods. 



Centrum analyses 



Dorsal centrum diameter and dorsal "birth diameter" 

 were measured in a transverse plane along a straight 

 line through the focus of each vertebra ( Fig. 2B ). The 

 dorsal "birth diameter" was marked on the x-radio- 

 graphs and screen images. Distance from the focus 

 to the outer edge of each GR (Fig. 2B) was measured 

 on the scanned images by using CorelDRAW! 



