Mollet et al : Reproductive biology of Isurus oxyrinchus 



313 



1992; Francis and Stevens, 2000) and lamnoids (Mat- 

 thews, 1950; Gilmore et al., 1983; Chen et al, 1997; 

 Yano et al., 1999). The report of a 50-mm diameter 

 egg in the ovary of a 3.37-m-TL shortfin mako with 

 an empty uterus (Applegate, 1966) is inexplicable. 



Uterus width was the most suitable reproductive 

 parameter for distinguishing between immature and 

 mature female shortfin makos and also between dif- 

 ferent reproductive stages of mature females. Our 

 data indicated that immature shortfin mako have 

 UWIs between 0.11% and 2.19^, whereas mature 

 shortfin makos have UWIs between 1.7% and 7.5%, 

 with the largest values occurring at parturition and 

 in early postpartum females (Fig. 3B). Little com- 

 parative data for other lamnids were available. Two 

 immature great white sharks of 4.8 m and 4.9 m 

 had a UWI of 2.1% and 1.6%, respectively. 12. le A 

 5.2-m-TL gi-eat white shark had a UWI of 7.7%; this 

 fish may have recently given birth (Stevens, unpubl. 

 data). However, a 5.2-m, 1520 kg female reported 

 by Bruce (1992) to be mature, and bearing possible 

 mating scars had UWIs of only 0.88-1.0% (left) and 

 0.44-0.56% (right); these values appear to be far too 

 small for a mature shark of this size. 



Our study failed to confirm the potential of HSI as 

 a good indicator of reproductive status. Our sample 

 size was large (immature, 7i = 125; mature, n=35); 

 however, we had few data for pregnant fish. Cliff et 

 al. (1990) observed the highest HSIs in two recently 

 fertilized females, which also had high GSIs. Their 

 observations suggested that HSI might be useful as 

 an indicator of reproductive status: however, their 

 sample size was small (n = 12). They also noted that 

 the HSIs of males were just as variable and covered 

 about the same range (2.9-13.7%). For sharks with 

 long gestation period, the HSI may not be tied closely 

 to the reproductive cycle. Because the shortfin mako 

 has a high metabolic rate (Graham et al., 1990), 

 reserves in the liver cannot be expected to last longer 

 than a few weeks. 



Gestation period 



A small number of early and mid-term litters from 

 the western North Atlantic and the Northern Hemi- 

 sphere supported a gestation period in excess of 12 

 months. By contrast, the absence of measured mid- 

 term litters from the Southern Hemisphere made it 

 possible also to fit the data with a 6-7 month ges- 

 tation period. It is highly unlikely that a regional 

 difference of this magnitude could exist in a widely 

 distributed species such as the shortfin mako. It is 



therefore most probable that the duration of gesta- 

 tion is 15-18 months in both hemispheres. 



Is a 6-7 month gestation possible for the North- 

 ern Hemisphere? Reproductive asynchrony (where 

 individuals in an accessory population are out of 

 phase with the principal population) is a possibility 

 for wide-ranging oceanic sharks, and was suggested 

 for the great white shark (Lineaweaver and Backus, 

 1970). For our shortfin mako data, we would have 

 to propose that the early-term (no. 3 in Table 1) 

 and all five mid-term litters (nos. 4-8) belong to 

 the accessory population(s). However, the gestation 

 period becomes indeterminate when these litters are 

 excluded from the calculations. It is far more prob- 

 able that these pregnant females belong to the prin- 

 cipal populations, should this theory apply to the 

 shortfin mako, and that the gestation period is 15-18 

 months. Additional data would be needed to test the 

 hypothesis of reproductive asynchrony for elasmo- 

 branchs. 



No embryo length data were available for an early 

 fall litter from the Southern Hemisphere (no. 9 in 

 Table 1). If the gestation period were 6-7 months, 

 this litter would have been at the blastodisc stage 

 (as were litters no. 1 and no. 2 in Table 2) and might 

 have been overlooked without detailed examination 

 of the uterus. If the gestation period was 18 months, 

 the embryos of this litter would have been 35-45 cm 

 long and would have been observed (as they were) 

 during gutting of the pregnant fish on the boat. 



Body temperature may account for some of the dif- 

 ference in the gestation between the shortfin mako 

 (15-18 months) and the porbeagle (9 months in 

 both hemispheres; Francis and Stevens, 2000). The 

 porbeagle and the salmon shark possess a sizable 

 kidney rete, which is not found in the shortfin mako, 

 and body temperatures are in general higher in 

 Lamna than in Isurus.^'' Cloacal temperature mea- 

 surements from 21 salmon sharks had a mean of 

 23. 3^0 (SD=1.1°) which is probably most represen- 

 tative of the uterus area.''^ This temperature could 

 be 5"C higher than the temperature of the uterus 

 in the shortfin mako, which prefers water tempera- 

 tures around 18°C and is known for vertical move- 

 ments into colder water (Casey and Kohler, 1992; 

 Carey and Scharold 1990; Holts and Bedford, 1993). 



Our litter of 3.0-cm-TL embi-yos confiiTned slow 

 shortfin mako embryonic development, at least during 

 the early stage. These very early-term embryos were 

 found next to more than 40 nutritive egg cases in each 

 uterus ( Fig. IB, Table 2 ). Gilmore ( 1993 ) reported that 



I'' Bruce, B. 1998. CSIRO Manne Re.search, P.O. Box 1.538, 

 Hobart, Tasmania 7001. Personal commun. 



'" Goldman, K. 1998. Virginia Institute of Marine Science 

 (VIMSl, RO. Box 1346, Gloucester Pt.. VA 23062. Personal 

 commun. 



