Milton et al.: Ageing of Lutjanus erythropterus. L malabancus. and L sebae 



105 



The method of radioanalysis of the otoliths is de- 

 tailed in Fenton et al. (1990, 1991). It involves mea- 

 suring the specific activity of 226 Ra: 210 Pb by alpha- 

 spectrometry. Because of the extremely small spe- 

 cific activities measured (0.01-0.1 dprn-g" 1 for Polo- 

 nium-210 [ 210 Po]), cleanliness is of the utmost im- 

 portance in the analytical procedure. Every item of 

 laboratory ware that contacted the otolith solutions 

 and otoliths was chemically decontaminated in al- 

 kaline 0.05M Na 4 EDTA(pH 10.5). The otoliths were 

 washed and rinsed several times in this solution, then 

 washed several times in 0.1M HC1 (<10 s) and fi- 

 nally washed twice in water. 



Our analyses of 210 Pb, via its short-lived daugh- 

 ter-proxy 210 Po, and 226 Ra were made with high-reso- 

 lution alpha-spectrometers according to the meth- 

 ods of Fenton et al. (1990). The mean 210 Po reagent 

 blank was 0.0071 ± 0.0012 dpm. Recovery of 210 Po 

 was always at least 90% and instrument background 

 counts (for 208 Po and 210 Po) were less than one 

 countd -1 . 226 Ra was analyzed by a direct alpha-spec- 

 trometry method and chemical yield was measured 

 by gamma spectrometry of a Barium-133 ( 133 Ba) 

 tracer (Fenton et al., 1990, 1991). Mean activity of 

 the 226 Ra blanks was 0.0174 ± 0.0026 dpm, which 

 was lower than in previous studies (e.g. Bennett et 

 al., 1982; Fenton et al., 1991) owing to careful con- 

 trol of reagents. Recovery of 226 Ra (as estimated by 

 the recovery of 133 Ba tracer) was greater than 85% 

 for all samples. 



(l-e-' 



--). 



XT\\ 



l-(l-R) 



(l-e-) 



XT 



\ 



-X(t-T) 



(2) 



where all parameters are the same as in the previ- 

 ous model, except T, which is the estimated age of 

 the otolith core. A linear mass growth model was 

 assumed only up to the age of the core. The initial 

 uptake 210 Pb/ 226 Ra activity ratio was generally as- 

 sumed to be i?=0.0. This is the most conservative 

 value, and so radiometric age estimates derived with 

 this value must overestimate the maximum possible 

 age of the sample. The above equations were solved 

 numerically by a Newton-Raphson iteration method 

 (Fenton et al., 1991). 



Stable element analysis 



The levels of lead and barium in otoliths are pre- 

 sumed to act as stable equivalents of 210 Pb and 226 Ra 

 and so can be used to assess the uptake of the radio- 

 active isotopes and to normalize the radiometric data 

 (Fenton and Short, 1992). Therefore, the concentra- 

 tions of stable lead, barium, strontium (Sr), and cal- 

 cium (Ca) in each otolith sample were measured for an 

 aliquot of each dissolved otolith solution used in the 

 radiometric analysis. Each solution was analyzed by 

 inductively coupled plasma mass spectrometry for lead 

 and barium and by inductively coupled plasma atomic 

 emission spectrometry for strontium and calcium. 



Data analysis 



The ages of whole otoliths were calculated on the 

 basis of a single constant (linear) growth rate by the 

 equation originally derived by Bennett et al. (1982): 



A=l-(1-R) 



1 



It 



(1) 



where A=the ratio of the activity of 210 Pb to 226 Ra 

 activity at time t ( 210 Pb/ 226 Ra) t ; i?=ratio of 210 Pb to 

 226 Ra at the time of deposition [( 210 Pb/ 226 Ra) ]; and 

 >.=decay constant for 210 Pb (0.03114 yr _1 ). Assump- 

 tion of a single linear mass growth rate produces 

 radiometric ages that are greater than those that 

 would result from assumption of an exponential (non- 

 linear) rate, the bias always favoring a higher value 

 (Campana et al., 1993). It should be understood that 

 using this assumption (linear mass growth of the 

 otolith) will produce age estimates that always over- 

 estimate the real age. 



For otolith cores, ages were calculated from Smith 

 et al.'s (1991) equation: 



Otolith ageing 



Pairs of otoliths from each fish were cleaned of ex- 

 cess tissue, dried at 60°C for 24 h, weighed (± 0.1 

 mg) and measured along the longitudinal axis with 

 dial calipers (± 0.05 mm). One otolith of each pair 

 was embedded in polyester resin and cross-sectioned 

 with a diamond saw (Augustine and Kenchington, 

 1987). Thin sections (approximately 200 |im) of each 

 otolith were bonded to microscope slides with thermo- 

 plastic cement. Each section was polished on both 

 faces with 800-grit wet-and-dry carborundum paper 

 before being examined with a video-enhanced light 

 microscope attached to a microcomputer with pre- 

 cise distance-measuring software. The rings (pre- 

 sumed annuli) were counted and the distance be- 

 tween them measured along the dorsal axis adjacent 

 to the sulcus, where they were most clearly distin- 

 guishable. In whole otoliths, the rings were counted 

 against a strong background point light source. 



Counts of rings in all whole and sectioned otoliths 

 were made independently by two readers. When the 

 ring counts differed, the otoliths were reexamined 

 by both readers. If the counts still differed by more 



