312 



Fishery Bulletin 98(2) 



estimate agrees with the observation that a sandtiger 

 pup born in captivity at Sea World of Orlando did 

 not eat until 25 d after birth (Gilmore et al., 1983). 

 By comparison, HSI values of 13-19% were recorded 

 in three great white shark full-term embryos with 

 stomach yolk mass between 0% and 4% (Francis, 

 1996; Uchida et al., 1996). A near-term 97-cm-TL 

 longfin mako, Isurus paucus, embryo had a large 

 liver but surprisingly small amounts of yolk in the 

 cardiac stomach (0.549^) (Gilmore, 1983). 



There is no convincing evidence^^ to confirm that 

 older shortfin mako embryos eat smaller ones while 

 in the uterus, as claimed by Tricas et al. ( 1997). They 

 possibly inferred this conclusion from a misleading 

 statement in Compagno ( 1988, p. 83 ).>- Costa et al.i'^ 

 suggested that uterine cannibalism occurs on the 

 basis of a piece of lower jaw found in the stomach of 

 a term embryo. This behavior is difficult to explain, 

 and Francis (1996) observed only shed teeth in the 

 stomach of a great white shark full-term embryo. 

 We expected to but did not find any shed teeth in 

 the stomach of near-term embryos. Embryonic can- 

 nibalism has only been documented in the sandtiger 

 shark (Gilmore et al., 1983) and seems unlikely in 

 the shortfin mako with a litter size as high as 25- 

 30. 



Litter size 



Data on litter size of shortfin makos were scarce 

 and possibly biased by low values due to abortion 

 during capture and the reported large value of 25-30 

 in Sanzo (1912). The lowest values of 4-6 were for 

 near- term litters (TL~60 cm ) reported anecdotally by 

 game fishermen. Although we are sure they are reli- 

 able, they should be taken as minimum estimates 

 because such low values suggest abortion during 

 capture, especially for near-term litters. Abortion is 

 unlikely for mid-term embryos with their huge yolk 

 stomachs." A means model test was not conclusive 

 because litter size also depends on maternal size, 

 which was not available for five near-term litters. 

 The number of embryos in the Sanzo (1912) litter 

 was also based on information from a fisherman; 

 howevei", a study of the original Italian text has sug- 



" Gilmore, R. G. 1999. Dynamic Corp., Kennedy Space Center, 

 FL 32899. Pensonal commun. 



'■■^ Compagno. L. V. C. 1998 and 1999. Shark Research Centre, 

 P.O. Box 61. 8000 Cape Town, South Africa. Personal commun. 



" Costa, F. E. S.. F M. S. Braga, A. F Amorim. and C. A. Arfelli. 

 199.5. Reproductive biology of shortfin mako, Isiinis oxyrin- 

 chus, Rafinesque 1809. Resumos VII Reuniao do Grupo de Ti'a- 

 balho sobre Pesca e Pesquisa de Tubaroes e Raias do Brasi' Rio 

 Grande, November 20-24, 199.5. Fundacao Universidade do 

 Rio Grancc, FURG Rio Grande, RS Brasil. 



gested that the value of 25-30 is reliable.^ Although 

 western North Atlantic females are slightly heavier 

 than females of similar size from South Africa and 

 Australia, there were no regional differences in litter 

 size between the two regions. 



The mean litter size of 12.5 is larger than that for 

 other lamnids, although data for some species are 

 scarce. The longfin mako has a mean litter size of 

 four (range 2-8, /?=3, Compagno, 1984; Killam and 

 Parsons, 1986; Casey^^). The mean litter size of the 

 gi-eat white shark is 8.9 (range: 4-14, 11 = 11; Fran- 

 cis, 1996; Uchida et al., 1996). A single litter of four 

 was reported for the salmon shark (Otake, 1990). 

 The litter size of the porbeagle is 2-5, mostly 4 (e.g. 

 Swenander, 1907; Templeman, 1963; Francis and 

 Stevens. 2000). A weak relationship between litter 

 size and maternal length in the shortfin mako also 

 exists for the great white shark (based on data in 

 Francis, 1996). A positive relation between litter size 

 and maternal size has been observed for many car- 

 charhinids (Compagno, 1988). 



Indicators of sexual maturity 



The GSI was not conclusive in determining sexual 

 maturity in lamnids because near-term and post- 

 partum females both have low values. In oophagous 

 sharks the ovary continues production to nourish 

 the embryos well into gestation, thus producing high 

 GSIs. The large variations in ovary size in mature 

 shortfin makos (0.2-12 kg; GSI ~0.1-5.07( ) were as 

 expected. Data from other lamnids are scarce. A 

 longfin mako with near-term embryos had a spent 

 ovary with 081=0.469^ (Gilmore, 1983). Active ova- 

 ries in mid-term porbeagles weighed 2.75 kg (GSI 

 2.35%) and 6.3 kg (GSI 3.6Vr ) (Francis and Stevens, 

 2000; Swenander, 1907). Arfelli and de Amorim''^ 

 observed a 12-kg ovary (GSI 0.489^ ) in a nonpreg- 

 nant 5.3-m-TL great white shark. 



We found that MOD was not useful for determin- 

 ing maturity in shortfin makos, whereas it is in 

 sharks without oophagy (e.g. Pratt, 1979; Peres and 

 Vooren, 1991; Parsons, 1993). The values near the 

 upper limit of the observed range (1-8 mm. Fig. 2B) 

 were similar to values (5-11 mm) reported for other 

 lamnids (Swenander, 1907; Gilmore, 1983; Bruce, 



" Casey, J. 19(S(i. Disti'lbution of the longfin mako tlminiy 

 paucux\ in the northwest Atlantic. Program and Abstracts 

 ASIH and AES Annual Meeting, Victoria BC, Canada, 15-21 

 June 1986, no page numbers. 



1^ Arfelli, C. A., and A. F. de Aniorini, 199.3. Notes on the 

 white shark ( Carcharndon carc/iancis ) caught off Cananeia. Sao 

 Paulo-Brazil. Program and Abstracts ASIH and AES Annual 

 Meeting. The flniversitv of Texas at .Austin, 27 May-2 .lune 

 199.3, 348 p. 



