96 



Fishery Bulletin 102(1 



120° 



130° 



N 



4 



Australia 



Great 

 LG Barrier 15° 

 Reef 



PG 



Queensland 



35° 







Lizard Island group (LG) 



J\ 



Lizard 

 Island <\ 



V-— V^" ,25km, 

 Palfrey Q . ' ' 



Island _ Seabird Islet 



South Island 



Palm Island group (PG) 



Pelorus Island 



Brisk Island 

 <V,Havannah Island 



Figure 1 



Location of the Palm Island group (PG) and Lizard Island (LG) group within the Great Barrier Reef 

 off the coast of Queensland, Australia. 



oocytes and, in the case of inactive ovaries, the relative 

 thickness of the gonad wall and the compactness of the 

 ovarian lamellae (Samoilys and Roelofs, 2000). For those 

 samples processed histologically, macroscopic features 

 were compared within and between reproductive stages to 

 determine whether any macroscopic characteristics could 

 be used to accurately stage ovaries 



Age determination 



Ages offish were determined in order to estimate age-based 

 schedules of growth, mortality, and maturation. Otoliths 

 lacking broad, opaque macro-increments were processed 

 to enumerate finer presumed daily micro-increments. 

 These otoliths were ground by hand first from the anterior 

 end and then the posterior end through a progressively 

 finer series of P1200 sandpaper, 12-um lapping film, and 

 9-um lapping film until a thin section through the nucleus 

 remained. Micro-increments were enumerated on two inde- 

 pendent occasions. If the two counts for one specimen did 

 not deviate by more than 5% of their mean, the mean was 

 used as the age estimate. Otherwise, a third reading was 

 performed and the mean of this and the more similar of 

 the first two readings was used as the age estimate, again 

 provided that the counts differed by no more than 5' < of 

 their mean. 



Macro-increments in the otoliths of L. carponotatus have 

 been validated as annuli by tetracycline labeling (Cappo et 

 al., 2000). A pilot analysis indicated that age estimates did 



not differ between readings of whole left and right otoliths 

 (paired t-test: df=59; r=0.60; P=0.55); therefore one otolith 

 was randomly selected from each sample for age determi- 

 nation. All otoliths were initially read whole. A second pilot 

 analysis compared whole and sectioned age estimates for 

 a subsample of L. carponotatus otoliths. This comparison 

 suggested that whole readings began to drastically under- 

 estimate age beyond approximately sectioned age 12 (see 

 Kritzer, 2002 ). To capitalize on both the greater efficiency of 

 whole readings and the greater accuracy of sectioned read- 

 ings, whole readings were used for all fish except for those 

 for which any whole reading exceeded 10 or for which there 

 was not agreement in at least two out of three independent 

 whole readings. If at least two out of three independent 

 readings of either whole or sectioned otoliths (as appropri- 

 ate) agreed, then that value was used as the age estimate. 

 Ferreira and Russ (1994) have described the whole- and 

 sectioned-otolith preparation and reading methods used 

 in the present study. 



Sex-specific demography 



Early growth of L. carponotatus was estimated by linear 

 regression of FL on age for those samples processed to 

 read subannual micro-increments. Separate regressions 

 were performed for the Palm and Lizard island groups and 

 these were compared by analysis of covariance ( ANCOVA). 

 Because of the undeveloped nature of the gonads of the 

 smallest fish, early growth was estimated without refer- 



