Smale and Goosen: Reproduction and feeding of Triakis megalopterus 



989 



Diet was quantified by 1 ) frequency of occurrence 

 ( ' ; F ). the ratio of stomachs containing a particular prey 

 to stomachs containing any prey, expressed as a per- 

 centage; 2) numerical importance (%N), the number of 

 each prey expressed as a percentage of the total num- 

 ber of prey items; and 3) gi'avimetric importance (%M ), 

 the wet mass of a prey category as a percentage of the 

 total weight of the stomach contents (Hyslop, 1980). 

 By using all three methods of analysis, we avoided bias 

 associated with the use of any one method (Hynes, 1950; 

 Windell, 1968; Hyslop, 1980). No combination of meth- 

 ods was used because this may have resulted in com- 

 liining sources of error ( Berg, 1979 ). Reconstituted prey 

 weights were not used because bias may have been 

 introduced as a result of the different digestion and 

 accumulation rates of fish otoliths and cephalopod 

 beaks ( Smale, 1983). Consequently, the actual wet mass 

 of each item in stomachs was used to investigate prey 

 composition in this study (Smale, 1991). 



Preliminary analyses revealed considerable differ- 

 ences in prey between sharks of different sizes, and 

 samples were subsequently divided into three arbi- 

 trary size classes: less than <999 mm, 1000-1399 

 mm, and >1400 mm TL. To investigate the relation 

 between sizes of predator and prey, reconstructed 

 prey lengths and masses were calculated with otolith 

 and beak measurements. Recorded wet masses and 

 carapace widths and lengths had to be used for crus- 

 taceans, and only intact crustaceans were included 

 in the predator-prey size analysis. Common names 

 used for crustaceans are after Holthuis (1991). 



Results 



Depth range 



Material collected in this study was derived from fish- 

 ermen exploiting a wide range of depths. Both shore- 

 based and boat-based catches were sampled, al- 

 though the proportion of effort in different depth 

 ranges was not quantifiable. Nevertheless, T. mega- 

 lopterus was taken mainly from shallow water. Of 

 125 specimens where collection depth was known, 

 89.69^ were taken in shallow waters of 10 m and less. 

 Only S% were collected at 11-20 m and 2.4'7f at >20 m. 

 The gi-eatest capture depth recorded in this study was 

 approximately 30 m, at an offshore bank. No obvi- 

 ous size-based habitat segregation was evident in this 

 relatively small sample. 



Size range and maturation 



The overall male-to-female ratio in sharks sampled 

 was 1:2.5. The smallest male and female sharks cap- 



tured were 576 mm and 725 mm, respectively. Fe- 

 males attain larger sizes than males; the largest male 

 sampled was 1520 mm, compared with the largest 

 female of 2075 mm. The clasper length (CLI) of the 

 smallest male was 43 mm and that of the largest (a 

 male of 1402 mm TL) was 239 mm (Fig. 2A). Based 

 on clasper size, degree of calcification, and presence 

 of sperm in the seminal vesicle, maturation in these 

 sharks begins at about 1210 mm and is complete at 

 about 1369 mm (Fig. 2A). Claspers of one specimen 

 of 1345 mm were not quite rigid but sperm presence 

 was noted. The smallest mature male was 1250 mm 

 and the largest immature male was 1196 mm. The 

 claspers of adults measured, on average, 225 mm 

 (range 205-239 mm, « = 16). 



Increase in the diameter of ovarian eggs indicates 

 the beginning of maturation. Small eggs (<4 mm di- 

 ameter) could be seen in the ovaries of females of 

 977 mm and larger. Egg diameter increased notice- 

 ably in specimens larger than 1460 mm (Fig. 2B). 

 Mature females had yellow yolk-filled ova larger than 

 4 mm in diameter. The developing uteri appear as 

 thin strips of translucent tissue with diameters of 

 up to 2.5 mm in females smaller than 1365 mm. As 

 maturation proceeds, the uterus widens first at its 

 posterior end, becoming bottle-shaped at lengths of 

 1391-1405 mm, with a diameter of 4-16 mm at its 

 widest part (Fig. 2C). One female of 1490 mm TL 

 had a uterus width of 16 mm. All females over 1460 

 mm, except the previously mentioned individual, had 

 uterus widths wider than 20 mm, and all adult fe- 

 males had uterus widths measuring 20-140 mm 

 ( .V =82.81, n=21). The oviducal gland was difficult to 

 distinguish in the smallest females but usually mea- 

 sured between 3 and 7 mm ( .v=5.64, 7; = 14) in females 

 of less than 1365 mm, and gi-ows little until females 

 reach 1460 mm (Fig. 2D). All females over 1460 mm 

 had oviducal glands wider than 20 mm and all adult 

 females, both pregnant and non-pregnant (resting), had 

 oviducal glands measuring 20-51 mm ( .v=31.7,/!=29). 

 The smallest pregnant female measured 1465 mm TL. 



Reproduction 



Of 48 sexually mature females sampled, 81'7f were 

 pregnant. The number of embryos per litter ranged 

 from 5 to 15 ( .v=9.7, n=38). One female had 16 egg 

 cases, but of these only 13 had normal embryos, one 

 was empty, and two contained retarded embryos 

 (length 6.57^ of the next smallest embryo length). 

 Considering the size of the female ( 1650 mm TL). it 

 is possible that litters of up to 16 pups may be re- 

 corded. There is no significant difference in the num- 

 ber of embryos between left and the right uteri 



