56 



Fishery Bulletin 98(1) 



dance declining above about 19°C. The preferred tem- 

 perature in the North Atlantic is less than 18°C 

 (Aasen, 1963). 



Porbeagles can maintain their body temperature up 

 to 11°C above that of the surrounding water (Carey 

 et al., 1985). Among the small group of endothermic 

 sharks, they are exceeded in this capacity only by 

 salmon sharks (Carey et al., 1985). The apparent pref- 

 erence for higher latitudes by large (Yatsu, 1995) and 

 pregnant porbeagles may indicate an increased ability 

 to thermoregulate in larger sharks. High body tem- 

 perature is probably necessary for lamnid sharks to 

 function as active predators of fast-moving prey in 

 cold water (Goldman, 1997). 



In the North Atlantic, porbeagle abundance varies 

 seasonally and spatially (Aasen, 1961, 1963; Temple- 

 man, 1963; Mejuto and Garces, 1984; Mejuto, 1985; 

 Gauld, 1989), and there are indications of seasonal 

 variability in their vertical distribution ( Bigelow and 

 Schroeder, 1948; Aasen, 1961, 1963). Limited tagging 

 results show that they are capable of movements up to 

 2370 km (Aasen, 1962; Stevens, 1976, 1990). In combi- 

 nation with evidence of seasonal latitudinal migration 

 in the South Pacific (Yatsu, 1995), this range of move- 

 ment suggests that Southern Hemisphere porbeagles 

 may exhibit complex seasonal, spatial, and length-re- 

 lated distribution patterns. Our data were collected 

 mainly during April-July, and therefore provide little 

 information on seasonality. 



Length, weight, and growth 



Several length-weight relationships have been pub- 

 lished for the North Atlantic (Aasen, 1961; Mejuto and 

 Garces, 1984; Gauld, 1989; Stevens, 1990; Ellis and 

 Shackley 1995; Kohler et al., 1995). All were based 

 on small samples, except those of Gauld (1989), who 

 found a significant difference between males and fe- 

 males above 180 cm. Our New Zealand sample com- 

 prised mainly sharks less than 150 cm, and compared 

 with the length-weight relationships of Mejuto and 

 Garces (1984), Gauld (1989) and Kohler et al. (1995), 

 our length-weight relationship rises too steeply be- 

 yond 150 cm. Our relationship should therefore not be 

 extrapolated beyond 150 cm. 



The largest porbeagles in our samples were 228 cm 

 (male) and 208 cm (female). The maximum length 

 reached in the North Atlantic is not clear. Maxima of 

 253 cm (288 cm TL) and 278 cm (317 cm TL) for Scot- 

 tish males and females, respectively, appear to be the 

 largest reliable measurements (Gauld, 1989). Lengths 

 of 294-325 cm ("estimated ... 11 feet", "ca 11-12 feet", 

 or "340-370" cm TL) reported by McKenzie ( 1959) and 

 Templeman (1963) were obviously not measured ac- 

 curately and may have been overestimated. A TL of 



12 feet equals 365.8 cm TL, which likely forms the 

 basis for the maximum length of 365 cm TL reported 

 by Pratt and Casey (1990). If Southern Hemisphere 

 porbeagles grow as large as those in the North Atlan- 

 tic, our samples do not include the larger size classes. 

 Tuna longlines catch shortfin makos that exceed 300 

 cm FL (senior author, pers. obs.); therefore, they should 

 be capable of retaining large porbeagles. The latter evi- 

 dently inhabit latitudinal or depth ranges outside our 

 sampling area, which is inhabited mainly by juveniles. 



Males and females were equally represented at 

 lengths up to 150 cm. At greater lengths, males sig- 

 nificantly outnumbered females by about 3:1 in both 

 New Zealand and Australia. Skewed sex ratios have 

 been reported fi-equently in the North Atlantic, in fa- 

 vor of males (Mejuto and Garces, 1984; Mejuto, 1985; 

 Ellis and Shackley 1995), females (Gauld, 1989), or 

 either sex depending on the length range or sample 

 (Aasen, 1963; O'Boyleetal., 1996). Aasen (1963) found 

 that the overall sex ratio in large samples was close to 

 1:1. These results indicate that juveniles do not segre- 

 gate by sex, but that larger sharks do. 



MIX analysis discriminated 3 and 5 length modes 

 respectively in southwest New Zealand and Australia, 

 which we interpret as age classes. In northeast New 

 Zealand, the mean sampling date was 25 June, over 

 seven weeks after the mean sampling date for south- 

 west New Zealand; therefore the first mode in the for- 

 mer could represent new-bom sharks. Alternatively, it 

 may represent slow-gi'owing one-year-olds. The latter 

 interpretation is supported by the similarity of the po- 

 sitions of modes 2 and 3 in both northeast and south- 

 west New Zealand. 



Juveniles gi'ow linearly and rapidly, reaching 110- 

 125 cm FL in three years. They may grow slightly 

 faster in southwest New Zealand (20 cm/year) than in 

 Australia ( 16 cm/year), but the standard errors were 

 large for New Zealand 2- and 3-year-olds, and the com- 

 parison could be biased by incorrect determination of 

 the number of modes in the length-fi-equency data. The 

 modal lengths for Australian juveniles agree well with 

 the first four modes for northwest Atlantic porbeagles 

 presented by Aasen (1963) (Fig. 4), providing that our 

 modes represent age classes. Aasen (1963) presented 

 two growth curves — one based on length-fi-equency 

 modes, and the other on back-calculated lengths-at-age 

 ft-om a single vertebra fi-om a 226-cm female. ^ His two 

 growth curves were practically identical. 



Our growth estimates also agi'ee with length-incre- 

 ment data for five tagged northeast Atlantic porbea- 



^ In a preliminary report, Aasen (1961) presented vertebral age 

 data for 50 porbeagles. However the mean lengths-at-age differed 

 sub.stantially from those presented later, and Aasen ( 1963) stated 

 that his earlier results "were not very accurate." 



