NOTE Fey et al.: Effect of type of otolith and preparation technique on age estimation of Leiostomus xanthurus 



545 



tank at constant temperature (20°C) and salinity (30%<->), 

 under 12 h light:12 h dark photoperiod. These conditions 

 were maintained throughout the rearing period. Hatch- 

 ing occurred three days after spawning. Larvae were fed 

 rotifers throughout the experiment and supplemented 

 with enriched Artemia from day 20 through day 30. 

 Larvae were collected 4 days (n = 5), 12 days (n=7), and 

 27 days (n = 5) after hatching, and live total length (L T ) 

 measurements were made. Larvae were then preserved 

 in 95% ethanol. 



Sagittae and lapilli were dissected with fine-tipped 

 forceps and embedded on microscope slides. The incre- 

 ments were clearly visible and otoliths did not require 

 any additional preparation. All increment counts were 

 conducted three times by one person on different occa- 

 sions with a lOOx oil objective and a Nikon E600 micro- 

 scope with transmitted light. The light was polarized 

 to obtain better visibility. The reader did not know the 

 ages of the fish. 



Known fish age and the number of observed incre- 

 ments were used to determine the time of first incre- 

 ment formation on both the sagittae and lapilli. The 

 number of increments deposited between sampling dates 

 divided by elapsed days indicated periodicity of incre- 

 ment formation. 



Accuracy and precision 



The experimental protocol and conditions were the same 

 as in the previous examination of first increment for- 

 mation, except that fish were reared for 53 days and 

 artificial diet was added after day 30. Larvae (?? =24, 

 8.8-16.1 mm L T , mean=11.8 mm L T ) were collected 34 

 days after hatching, and juveniles (rc=34, 19.4-28.1 mm 

 L T , mean=24.3 mm L T ) were collected 53 days after 

 hatching. 



Sagittae and lapilli were dissected from fish with 

 fine-tipped forceps and embedded for sectioning on the 

 transverse plane (right sagitta) or polishing on the sag- 

 ittal plane (left whole sagitta and lapillus). Priority was 

 given to transverse sections, and if the right sagitta 

 was damaged during preparation, the left sagitta was 

 used (n = 8). Otoliths were sectioned with a slow-speed 

 saw with dual diamond wafering blades. Sections were 

 then ground on one side with 1000-grit sandpaper and 

 polished with 0.3-|i<m alumina paste. After increments 

 were counted on the proximal side of sections that were 

 polished on one side (see below for details), sections 

 were flipped over, ground, and polished to the core to 

 provide a section that was polished on two sides. The 

 left whole sagitta and lapillus were ground and polished 

 in the sagittal plane with 0.3-fjm alumina paste. One 

 person made all the increment counts three times for 

 each preparation technique on different occasions with 

 a lOOx oil objective on a Nikon E600 microscope with 

 transmitted light. The reader knew the study design, 

 but not the ages of the fish. 



The mean number of increments counted from sagit- 

 tae and lapilli prepared with different techniques were 

 compared with known ages to determine the accuracy of 



the different aging methods. The statistical significance 

 of differences in increment counts (accuracy) was evalu- 

 ated with a one-way ANOVA. Increment formation rate 

 was determined by comparing the number of increments 

 counted to known age, and by comparing the difference 

 in the number of increments between 34- and 54-day- 

 old fish and the number of actual days between these 

 increments (20 days). 



Precision of increment counts from different otoliths 

 and preparation techniques was determined with the 

 coefficient of variation (CV), calculated by using the 

 three increment counts made for each individual type 

 of otolith and preparation technique (Chang, 1982). The 

 differences in CV values among the four age estima- 

 tion methods were analyzed by using a Kruskal-Wallis 

 ANOVA. The statistical significance of observed differ- 

 ences were estimated with a post hoc Tukey HSD for 

 unequal n test. All the statistical data analyses were 

 performed with Statistica 6.0 software (StatSoft Inc., 

 Tusla, OK). 



Results 



First-increment formation 



First-increment formation on the sagitta occurred at 

 hatching, but there may be problems in resolving incre- 

 ments near the core. Increment counts on sagittae were 

 variable for 4-day-old larvae. Four increments were 

 visible on the sagittae of one individual. The first incre- 

 ment was more pronounced than the others and was 

 interpreted as a hatching check. This increment was 

 approximately 8 f<m from the core. On the sagittae of 

 the remaining four 4-day-old larvae, only one increment 

 was visible corresponding to the location of the perceived 

 hatching check. Despite the apparent nondaily increment 

 formation in 4-day-old larvae, an average of 12.3 (range 

 12-13) increments were visible on the sagittae of 12-day- 

 old larvae, and an average of 26.5 (range 26-27) were 

 visible on the sagittae of 27-day-old larvae. The first 

 increment observed on the sagittae of 12- and 27-day-old 

 larvae corresponded to the location of the first increment 

 observed in the sagittae of 4-day-old larvae (Fig. 1). 



First increment formation on the lapillus occurred 

 6-7 days after hatching. No increments were visible 

 on the lapilli of 4-day-old larvae. In older larvae, an 

 average of 6.4 (range 6-7) increments were observed 

 on 12-day-old larvae and an average of 20.3 (range 

 20-21) increments were observed on the lapilli of 27- 

 day-old larvae. Additionally, lapilli of 12 and 27-day 

 old larvae exhibited two checks in the area between 

 the otolith core and the first increment, but it was dif- 

 ficult to distinguish which check, if either, was formed 

 at hatching (Fig. 1). 



Accuracy and precision 



Increments were clearly visible regardless of otolith prepa- 

 ration technique (Fig. 2). Increment width increased from 



