382 
Fishery Bulletin 106(4) 
the two older category-2 specimens, the results 
are less definitive because the A 14 C values fall in 
the stable pre-bomb era. Figure 2 is an example 
of how a category-2 otolith may be aged given the 
different possible interpretive options. 
Discussion 
Age validation 
This study is the first published age validation for 
GOA Dover sole. Ages in the range of 8 to 47 years 
were validated. Accurate ages were indicated by 
both the unstandardized and standardized A i4 C 
results for category 1 when a purposeful age bias 
of -1 or -2 years produced the lowest SSR s. The 
shift of approximately 1 year (due to core size, as 
explained earlier) was expected, and when con- 
sidered with the purposeful age bias and lowest 
SSi?s, indicated that ages estimated by the break- 
and-burn method were accurate. All category- 
1 specimens were within the 99% simultaneous 
confidence intervals on the reference chronology 
(Fig. 4B) and this result also provides strong evi- 
dence for accurate ages. 
The difference between the purposeful age bias 
of -1 and -2 was not resolved. The resolution of this 
bomb 14 C age validation study was limited by the ap- 
proximate nature of the expected 1-year shift, variabil- 
ity in 14 C due to geographic location, and variability in 
14 C measurement, but the general accuracy of the ages 
was validated. In using bomb Z\ 14 C for age validation, 
assumption 1 is the largest contributor to any concerns 
about resolution due to variability in A 14 C because of 
depth or geographic area (Nydal, 1993; Kalish, 1995; 
Andrews et al., 2007). 
Our method of separating the specimens into two 
categories addressed specific objectives. Use of only the 
clearest specimens to validate the age estimates for 
category-1 specimens is common practice (Piner and 
Wischniowski, 2004; Kerr et al., 2005; Kastelle et al., 
2008). In a long-lived species such as Dover sole, age 
determination is difficult (Chilton and Beamish, 1982; 
Kimura and Anderl, 2005) and subjective interpreta- 
tions of growth patterns (Fig. 2) must be made. When 
determining the age of a species on a routine basis, a 
set of species-specific interpretive rules or “aging cri- 
teria” are applied to all specimens, i.e., to specimens 
that are easy to interpret and to those that are dif- 
ficult to interpret (Kastelle et al., 2008). By choosing 
validation samples like category 1 where there is little 
ambiguity, the basic methods and aging criteria can 
be validated and then applied to all samples, includ- 
ing ones like category 2. If the specimens were chosen 
randomly, without consideration for the difficulty in 
estimating age or interpreting otoliths, the spread of 
validation sample points around the reference chronol- 
ogy would increase. This spread would provide less in- 
formative results regarding the basic aging criteria, as 
150 
100 
50 
9 0 
-50 
-100 
-150 - 
1950 1960 1970 1980 
Birth year 
Figure 5 
A 14 C %c in otolith cores of Dover sole ( Microstomus pacificus) 
plotted against birth year, by category 1 (•) and category 
2 (A). Horizontal lines on category-2 specimens represent- 
ing the minimum and maximum postmeasurement age, and 
a loess smoothed curve ( ) has been fitted to category-1 
data points. 
shown by the spread in the category-2 specimens. The 
types of specimens represented by category 2 are more 
common in Dover sole; hence this sample design was 
selected to provide the most informative results. The 
exact percentage represented by the two categories was 
not determined in this study. A further reason for us- 
ing different categories was to provide a tool to develop 
aging criteria for the hard-to-age specimens. Following 
the validation, the less common category-1 specimens 
were used to provide the loess fit in Figure 5 to which 
the category-2 specimens were compared. 
The A 14 C results displayed a gap in specimens at a 
birth year of 1961. Similar gaps have been seen in other 
studies (Kerr et al., 2004; Piner and Wischniowski, 
2004; Piner et al., 2005; Kastelle et al., 2008) and are 
likely due to two reasons. First, in the early- to mid- 
1960s, marine 14 C was likely increasing so quickly that 
even if otoliths were accurately aged, a gap would likely 
be present because of the limited time range when these 
mid-range values of A 14 C existed. Second, in our sample 
selection process, otoliths that could be categorized 
unambiguously as category 1 and aged such that they 
represented the 1961 birth year were not present. 
The utility of A 14 C standardization is apparent in this 
study. Without the standardization, most of the valida- 
tion sample h 14 C points before about 1968 are above the 
reference chronology, or to its left, and below the refer- 
ence chronology after 1969. With the A 14 C standard- 
ization and the purposeful bias of -2 years applied to 
the category-1 ages, it is clear that the Dover sole 4 14 C 
points and Pacific halibut reference chronology are in 
synchrony and that ages are accurate. As previously ex- 
plained, the shift to the left is due to the core size. The 
consistently low Dover sole h 14 C points after 1969 are 
