982 



Fishery Bulletin 97(4), 1999 



200 300 400 500 600 700 

 Total length (mm) 



800 



25 

 20 

 15 

 10 

 5 



B 



i 



1 



II 



Age class (years) 



Figure 2 



(A) Length and (B) age-frequency distributions of 

 male (open bars) and female (hatched bars) Rhizo- 

 prionodon taylon from the waters of Townsville, 

 Queensland. The arrow indicates size at birth. 



population surviving to the beginning of an age class 

 (J) rather than to theenci (I). The value of r increased 

 from 0.271 to 0.600, doubling time decreased from 

 2.544 years to 1.155 years, and the maximum value 

 of F sustainable by the population (F^) increased from 

 0.179 to 0.600 (Table 3) when the beginning of an 

 age class was used. Doubling the value of M in the 

 first year of life (I vs. K) caused a reduction in the 

 value of r from 0.271 to 0.001, and population dou- 

 bling time increased from 2.544 to 529.2 years (Table 

 3). The maximum level of fishing mortality that the 

 population could sustain under such a scenario de- 

 creased from 0.179 to 0.001. Doubling first-year M 

 and also calculating age-specific natality from the 

 number surviving at the beginning of age classes (L) 

 produced results identical to the base case (Table 3). 

 Increasing the maximum age to 10 years resulted 

 in a small increase in /; from 0.271 to 0.276, and de- 

 creased population doubling time from 2.544 to 2.516 



years (Table 3). Setting age at maturity at two years 

 instead of one resulted in an r of 0.035, and a popu- 

 lation doubling time of 19.82 years (Table 3). 



Values of r varied considerably with changes in F 

 and AAFC (Fig. 4). With an AAFC of 0, F^. := 0.18; 

 with an AAFC of one, F = 0.27; and with an AAFC of 

 two, F = 0.67. When AAFC was three or more the popu- 

 lation was sustainable at F values up to at least 1.0. 



Discussion 



Mortality estimates 



In most studies where natural mortality has been 

 estimated for shark populations, particularly for use 

 in demographic analysis, indirect estimation meth- 

 ods have been used (e.g. Hoenig and Gruber, 1990; 

 Cortes, 1995; Sminkey and Musick, 1996; Au and 

 Smith, 1997). The results of the current study indi- 

 cate that these indirect methods may provide a wide 



am 

 res 

 tul 



trui 



Jof 



teal 



