304 



Fishery Bulletin 98(2) 



These results justified combining shortfin mako data 

 from different regions and the use of a Hnear regres- 

 sion. Embryonic development may, however, deviate 

 from a straight line and be more S-shaped; therefore 

 a straight line may underestimate time of fertiliza- 

 tion and overestimate parturition. These deviations 

 cancel each other and a straight line still provides a 

 good estimate of gestation. 



Reproductive cycle 



We lacked sufficient numbers of females in all repro- 

 ductive stages to determine the reproductive cycle 

 from a stage-frequency distribution. For many litters, 

 little or no maternal data were available, which pre- 

 cluded the use of a more rigorous multivariate statis- 

 tical analysis in determining the reproductive cycle. 

 We used temporal uterus width index data (UWI, 

 uterus width as % of TL) of pregnant and postpartum 

 females with a graphical comparison of 2- and 3-year 

 reproductive cycles. The utenas of postpartum short- 

 fin mako sharks shrinks after parturition. We chose 

 the reproductive cycle compatible with this, given that 

 our results suggested an extended gestation period of 

 15-18 months. For easier graphic interpretation, we 

 plotted the 19 March (1988) Southern Hemisphere 

 specimen a few days later. This procedure assured 

 that the specimen appeared at the beginning of the 

 gestation period (fall) on the far left-hand side of the 

 temporal graphs. The use of polar coordinates did not 

 provide a better presentation. 



Results 



Pregnant females and details of selected litters 



There were 35 pregnant females on record (Table 

 1), with mean length of 3.1 m (range 2.63-3.85 m, 

 /!=24) and mass range of 153-547 kg. Capture dates 

 ranged from pre-1832 (Vaillant, 1889; no. 14, Table 

 1) to 27 August 1997 (no. 20, Table 1 ). Capture loca- 

 tions were distributed worldwide and corresponding 

 sea surface temperatures, where available, ranged 

 from 18" to 28.5 C. Detailed information was avail- 

 able for six females, three in very early pregnancy 

 and three with well-developed embryos (Table 2). 



Two recently fertilized South African sharks of 

 2.92 and 2.88 m were caught in late summer and 

 late fall of 1987 and 1986 (nos. 1 and 2 in Tables 

 1 and 2, respectively). A 3.09-m female was caught 

 near Puerto Rico in January (early winter) with a 

 litter of nine early-term embryos (no. 3 in Tables 1 

 and 2). The 2.6-3.3 cm embryos, each still inside an 

 egg case, had external gills and a.large yolk sac (Fig. 



lA). The size of the yolk sac was estimated from Fig. 

 lA to be 0.9 X 0.6 cm, assuming the embryo in the 

 center of the figure was 3.0 cm. There were over 40 

 nutritive egg cases in each uterus (Fig. IB). 



A3.25-m female with a near-term litter was caught 

 in the same month and location (January and Puerto 

 Rico, respectively) by a commercial swordfish long- 

 liner (no. 16 in Table 1, no. 4 in Table 2). This 

 female carried 15 embryos of around 52 cm, each 

 with a similarly large yolk-filled stomach (Fig. IC). 

 No uterine compartments were observed and no egg 

 cases remained in the uterus. A3. 34-m female shark 

 stranded on the beach at Umhlanga, South Africa, 

 in late winter (no. 20 in Table 1; no. 5 in Table 2). 

 This female carried a litter comprising nine embryos 

 of about 60 cm. Emerging adultlike teeth were pres- 

 ent in both jaws of the embryo examined. The stom- 

 achs of four embryos that were cut open had no 

 shed teeth but had large amounts of yolk (Fig. ID). 

 No uterine compartments were obsei"\'ed and no egg 

 cases remained in the uterus. About one liter of clear, 

 viscous fluid (not seawater) was found in each uterus 

 but was not analyzed. A Japanese longliner caught a 

 3.14-m female in the southern Pacific in midwinter 

 (no. 30 in Table 1, no. 6 in Table 2). The 13 full-term 

 embryos had a length of about 69 cm and a total mass 

 of 36.8 kg, 14% of the estimated matei'nal mass of 266 

 kg. The embryo stomachs were slightly distended and 

 contained small but variable amounts of yolk. 



Litter size 



The majority of the litters examined for this study 

 were near-term. The mean number of embryos per 

 litter (litter size) was 12.5 (range: 4 to 27.5, ;7=30). 

 The largest litter comprised 25-30 embryos (we 

 used 27.5 in our calculations) in a female caught 

 in the Mediterranean Sea in about August of 1903 

 (Sanzo. 1912). The capture date was inferred fi'om 

 the observation that it occurred during a long holi- 

 day period ( summer) at the University of Messina. ' A 

 summer date agrees with the dates of the traditional 

 swordfish fishery, in which this shark was probably 

 caught.'^ One of these embryos was described by 

 Sanzo (1912) as being a great white shark, Carcha- 

 rodon carcharias. but a recent re-examination has 

 shown it to be a shortfin mako shark.' The male to 

 female ratio of 10 litters was 57:68 and is not signifi- 

 cantly different from 1:1 (pooled ;^-=0.80, Yates cor- 

 rected, P=0.37). 



' Notarbartolo di Sciara, G. 1997. Institute Centrale pi>r la 

 Riceica Applicata al Mare ( ICRAM I, via di Casalotti 300. 0()l(i6 

 Rome, Italy. Personal commun. 



" Fergu.'ison, I. 1998. The Shark Ti-u.st. 36 Kingfisher Court, 

 Hambridge Road, Newbury, UK RG14 5SJ. Personal commun. 



