476 



Fishery Bulletin 92|2). 1994 



where Lt = total length, and t - age in years. Fork 

 length data were converted to total length by the 

 regression of Hazin et al. (1991): 



FL = 11.27 + 0.78 TL 



where FL - fork length, and TL = total length. 



Depths of longline hooks were estimated using the 

 equations of Yoshihara (1952, 1954, a and b). Verti- 

 cal distribution of males and females was studied 

 through the relative distribution of mean catches on 

 longline hooks, during February to June and July to 

 December 1990. Differences in mean catch of males 

 and females on longline hooks were evaluated by chi- 

 square analysis (df=6). Sea water temperature from 

 to 300 m was surveyed in 35 DBT (digital bathy- 

 thermograph) profiles: 6 in May 1990, 3 in May 1991, 

 5 in June 1991, 13 in November 1990, 2 in Novem- 

 ber 1991, and 6 in December 1990. From January to 

 December 1990, sea surface temperature was mea- 

 sured by a mercury thermometer. 



Results 



From August to December 1987 and from February 

 1990 to December 1991, the blue shark mean CPUE 

 by quadrates increased eastward, being particularly 

 high east of long. 35°W (Fig. 2). Of the 325 sets, 260 

 (nearly 80%) were over bottom depths greater than 



1,000 m. In these areas, the mean CPUE of blue 

 shark was 0.50. The remaining 65 sets were in areas 

 with depths shallower than 1,000 m, close to oceanic 

 banks (Aracati, Sirius, and Guara banks), and west 

 of 35°W. In these areas, the mean CPUE of blue shark 

 was only 0.05. 



The fluctuation of the monthly mean CPUE of blue 

 shark in the area east of 35°W, over ocean depth of 

 1,000 m, for 1987, 1990 and 1991, were similar (Fig. 

 3). In 1990 and 1991 the CPUE was low until May, 

 increased during June and July, decreased again in 

 August, increased during September and October, 

 and decreased once more in November and Decem- 

 ber. In 1987 the CPUE was low in August, increased 

 in October and decreased during November and De- 

 cember. Differences in mean CPUE among months 

 were significant (ANOVA; P<0.0001; Table 3). 



The fluctuation of the monthly mean CPUE of male 

 and female blue sharks was distinct (Fig. 4). CPUE 

 for females was highest during March. From July to 

 December, CPUE for females was low in this fishing 

 ground. CPUE for males, however, was lowest dur- 

 ing March, after which abundance increased and 

 peaked during September and October. The sea sur- 

 face temperature in 1990 was highest in May and 

 lowest in September. During this year, in general, 

 the CPUE of males tended to decrease with an in- 

 crease in the sea surface temperature, whereas the 

 CPUE of females tended to increase. 



Jan . Feb . Mar . Apr . May Jun . Jul . Aug . Sep . Oct . Nov . Dec . 

 Month 



1987 



1991 



Figure 3 



Monthly mean catch per 100 hooks (CPUE) of blue 

 shark, Prionace glauca, in the southwestern equato- 

 rial Atlantic Ocean, in the area east of long. 35°W and 

 with ocean depth over 1,000 m, from August to Decem- 

 ber 1987 and from February 1990 to December 1991. 



