634 
Fishery Bulletin 99(4) 
Although right and left sectioned otoliths generally gave 
the same age, presumed annual marks were usually clear- 
er and easier to interpret on the right otolith. Right sec- 
tioned otoliths had a much longer counting path and were 
therefore easier to age than left sectioned otoliths, where 
the marks were more crowded and less clearly defined 
(Fig. IB). This difference was also reflected in higher con- 
fidence scores and lower reading times for the right sec- 
tioned otolith than for the left (Table 1). 
Comparison of different calcified 
structures from the same fish 
Whole and sectioned otoliths generally gave the same age 
readings. The number of presumed annual marks on whole 
and sectioned otoliths showed high agreement (95%), with 
100% agreement for fish under age 4 (Fig. 4A). In addition, 
the null hypothesis that the slope of the line equals one 
was not rejected (P=0.901). 
Although whole and sectioned otoliths generally provid- 
ed the same age, presumed annual marks were often clear- 
er on sectioned otoliths than on whole ones, especially in 
older fish, where crowding of marks at the edge of whole 
otoliths became a problem. This observation is supported 
by the much higher confidence scores for sectioned otoliths 
(Table 1). As a specific example, the oldest fish in the com- 
parison showed very clear marks and was aged 10 in ev- 
ery reading using both right and left sectioned otoliths 
(Fig. 5A), and all confidence scores were 5. Marks were less 
clear on the whole otolith (Fig. 5B), however, with between 
8 and 10 marks counted in different readings, and 
an average confidence score of only 2.5. In general, 
the use of sectioned otoliths appeared to greatly in- 
crease mark clarity in fish over age 4 or 5. 
Scales and sectioned otoliths often did not give 
the same age readings. Agreement in the number 
of presumed annual marks on scales and sectioned 
otoliths was undesirably low, at only 80% (Fig. 4B). 
In addition, the null hypothesis that the slope of 
the line equals one was rejected (P=0.047). Scales 
tended to overage compared with sectioned otoliths 
in fish age 4 and younger, but to underage in fish 
older than age 4. Agreement between scales and 
sectioned otoliths was fairly high for ages 0 to 4 
(86%, u=56) but decreased to only 65% in fish over 
age 4 (ti=23). 
Scales and whole otoliths often did not give the 
same age readings. Agreement in the number of 
presumed annual marks on scales and whole oto- 
liths was also undesirably low, at only 76% (Fig. 6). 
In addition, the null hypothesis that the slope of 
the line equals one was again rejected (P=0.039). 
As with sectioned otoliths, scales tended to over- 
age compared with whole otoliths in fish age 4 and 
younger and to underage in fish older than age 4. 
Agreement between whole otoliths and scales was 
fairly high for ages 0 to 4 (85%, 77=53) but decreased 
to only 56% in fish over age 4 (77=25). 
Increase in number of marks with 
structure size and fish size 
Mark counts on calcified structures increased as 
structure size and fish size increased, indicating 
that each structure tested could be useful in age 
determination. All regressions of mark counts on 
structure size were significant at P<0.001, and all 
slopes were positive (Table 4). Regressions were 
generally strong and explained much of the vari- 
ation in mark counts because 100 r 2 values were 
high, generally from 80% to 86%. Values for 100 
r 2 were lowest for left sectioned otolith radius 
and scale radius, at 67% and 73%, respectively. 
Likewise, all regressions of mark counts on fish 
size were significant at P < 0.001, and all slopes 
