Hesp et al : Age and size composition, growth rate, reproductive biology, and habitats of Glaucosoma hebraicum 



219 



0.4 r 



02 



25 55 



08 



0.4 



2-5 opaque zones 



g 7 28 



16 '' 20 

 14 



6-8 opaque zones 



04 

 



9-11 opaque zones 



0.8 



0.4 



12-15 opaque zones g .j. 



>- 



> 16 opaque zones 



0.4 



JFMAMJJASOND 

 Month 



Figure 2 



Mean monthly marginal increments ±1SE for sag- 

 ittal otoliths of Glaucosoma hebraicum. Sample 

 size is given for each month. In this Figure and 

 Figure 5. the closed rectangles on the horizontal 

 axis refer to summer and winter months and the 

 open rectangles to autumn and spring months. 



10 - 



5 





 15 

 10 



15 

 10 



5 





 ■5 

 10 



5 



D 

 15 

 10 



5 





 15 

 10 



5 



'- 

 ' 5 - 

 ■0 - 



■0 



15 

 10 



■0 

 5 

 

 ■5 

 10 



[H] 



r"~^ 



April 

 n=3 



r\/1ay 

 n=8 



June 

 n=13 



July 

 n=3 



August 

 n=11 



October 

 n=51 



November 

 n=63 



December 

 Q n=60 



January 

 n=27 



February 

 n=27 



Marcti 

 n=45 



Total length (mm) 



Figure 3 



Length-fi-equency distributions for Glaucosoma 

 hebraicum caught by trawling along the lower west 

 coast of Australia by using data for corresponding 

 months in the period May 1996 to June 1999. 

 "denotes mean lengths of 0-t- and early 1+ fish. His- 

 tograms in gray refer to 0+ fish, and those in black 

 and white refer to 1+ and 2-1- fish, respectively. 



caught until October, presumably reflecting the time typi- 

 cally required for the 0-f age class to be recruited into these 

 areas from those in which spawning occurs. The mean 

 length of the O-i- age class had reached 95 mm by October, 

 in which month the first opaque zone became delineated 

 on the otoliths, and 108 mm by January, when fish were 

 approaching the end of their first year of life. The mean 

 length of the corresponding cohort, now early l-i-, was 127 

 mm in March, after which month the number of l-i- fish 

 caught in trawl samples declined markedly (Fig. 3). 



The best of the 4- and 5-parameter growth curves were 

 selected as the models with the largest log-hkelihood at 

 that level of model complexity. The best 4-parameter model 

 was the cui-ve that assumed different asymptotic lengths 

 for the sexes, and the best 5-parameter model was that 

 which assumed that the growth coefficients were equal. 



Comparisons between the curves demonstrated that the 

 model that assumed common growth coefficients for fe- 

 males and males was not significantly different {P>0.05) 

 from the more complex model, which assumed that all 



