654 



Fishery Bulletin 91|4|, 1993 



it 7S 



Figure 2 (Continued) 



(Fig. 1). Altogether, 494 vertebrae were collected, of 

 which 66% were from troll-caught albacore. All alba- 

 core sampled were first measured according to the pro- 

 cedure described previously. Vertebrae were removed 

 by severing the caudal peduncle immediately anterior 

 to the large bony keels. Vertebrae in the caudal pe- 

 duncle were selected because they were easy to collect 

 and their removal did not affect the market value of 

 the fish. The caudal fins were trimmed, and the entire 

 peduncle was placed in a plastic bag and frozen. Sex 

 was recorded as often as possible (128 cases), but lo- 

 gistic difficulties prevented internal examination of the 

 majority of the albacore sampled. 



All vertebrae were processed in the laboratory ac- 

 cording to the methods described by Berry et al. ( 1977), 

 Berry (1978), and Lee et al. (1983). Caudal peduncles 

 were thawed and the flesh trimmed from around the 

 vertebrae. The 35th and 36th vertebrae were sepa- 

 rated from the others with a knife and a fine-toothed 

 hand saw. Each vertebra was then clamped in a vise 

 by the keel, and sawed in half along the sagittal (dorso- 

 ventral) plane. Remnants of the cone jelly were re- 

 moved and the anterior portion of the cone was lightly 

 scrubbed. 



The left lateral section was normally used for stain- 

 ing unless it had been damaged by cutting. Sections 

 were soaked for 5-20 minutes in Alizarin red-S stain 

 with Berry's (1978) "school tuna" solution, and then 

 rinsed in fresh water. Sections stained too darkly were 

 soaked in a 5% hydrogen peroxide solution until the 

 stain faded sufficiently. Stained vertebrae examined 

 under high intensity visible light displayed distinct 

 band couplets on the anterior cone (see Johnson, 1983). 

 Within each couplet, the bands are uniformly spaced 

 around the cone and separated by a thin translucent 

 zone. Each couplet is separated from adjacent ones by 

 a wide opaque zone that becomes somewhat depressed 

 towards the outside edge of the cone. Each couplet of 

 twin bands was considered to be a distinct growth 

 ring. The vertebral-ring-count was given by the num- 

 ber of complete couplet bands observed. 



The investigators examined 20 stained vertebrae to 

 derive criteria to be used in interpreting the ring struc- 

 tures. Vertebrae were then processed by the team: one 

 biologist read 20 vertebrae while the other recorded 

 data, and vice versa. At least 25% of the samples pro- 

 cessed each day were examined by both persons to 

 check consistency. Identical counts were obtained in 



