Kastelle et at: Age validation of Microstomus pacificus by means of bomb radiocarbon 
383 
likely due to different environmental regimes or biologi- 
cal differences of the two species (Kalish, 1993, 1995; 
Nydal, 1993; Andrews et al., 2007). The standardization 
procedure is ideal for correcting this type of bias, where 
a difference in range of A 14 C exists. Previously, this 
procedure was used by Kastelle et al. (2008) to re-ana- 
lyze validation data for black drum (Pogonias cromis), 
originally presented by Campana and Jones (1998). For 
the re-analysis, the black drum A 14 C values were stan- 
dardized to a Northern Hemisphere atmospheric A u C 
reference chronology in a comparison where they were 
dramatically different in scale but similar in timing 
(Kastelle et al., 2008). If little difference in range ex- 
ists when the standardization is applied, the estimated 
values of p and a will be close to 0 and 1, respectively, 
provided the overall fit is good. This situation would 
indicate that the standardization had little effect and 
that the correct evaluation of any aging error will still 
be made by considering the SSRs. This was the case 
for Pacific ocean perch analyzed previously with this 
method (Kastelle et al., 2008). Therefore, we feel this 
standardization method can be applied generally. 
We chose the Pacific halibut reference chronology for 
several reasons. First, this reference chronology is based 
mostly on juvenile fish (Piner and Wischniowski, 2004). 
It also represents a wide geographic area in the GOA, 
similar to that for Dover sole. Finally, although the 
early life history of Dover sole in the GOA is not well 
understood, the pelagic larvae are found in the upper 30 
m of the water column, and immature fish are known to 
concentrate in nearshore areas and shallow waters over 
the continental shelf (Abookire et al., 2001; Abookire 
and Bailey, 2007). Juvenile Pacific halibut are typically 
found in shallow nearshore areas (Norcross et al., 1995; 
Abookire et al., 2001); therefore comparisons with Dover 
sole for this age validation were reasonable. 
The otolith cores from the validation samples were 
smaller than the measured guide otoliths from 3-year- 
olds. Some of this difference may be explained by the 
presence of newly deposited opaque material beyond the 
third translucent zone in the measured 3-year-olds. In 
the cored otoliths, this same material was often ground 
away to expose the third translucent zone, thereby pro- 
ducing a size difference. Also, a few of the cores may 
have been incorrectly centered during the grinding pro- 
cess, and therefore may have incorporated a little mate- 
rial from beyond the third year. Conversely, too much 
material could have been removed, down to the second 
year’s growth zone. The latter is more likely the case 
as evidenced by the small core weights. On the proxi- 
mal side, the coring process may have inadvertently 
removed some material belonging to the third year in 
an effort to remove all material from later years in the 
region of the sulcus groove. We considered the average 
age of the material represented by the cores to be ap- 
proximately 1.5 years. However, if the cores were too 
small and some material inside the third translucent 
zone was removed, than the average age of the mate- 
rial may have been closer to 1 year, indicating less of a 
required shift. The difference was only 0.5 years; hence 
we considered any error from this consideration to be 
negligible. Less of a required shift was also indicated if 
the otoliths were accumulating more mass during their 
first year than in subsequent years (see Materials and 
methods where linear otolith growth is assumed). As 
mentioned previously, this age validation method can 
not resolve either type of error when potentially very 
small. However, it is probably not coincidental that the 
purposeful age bias of -1 year for the unstandardized 
A 14 C results was the best fit. 
Dover sole otoliths often display a transition in growth 
rate typically seen as a pattern of decreased spacing 
between presumed annual growth zones. This occurs 
when the fish is estimated to be about 6- to 8-years-old, 
with growth zones deposited prior to the transition rep- 
resenting younger and faster growth and post-transition 
zones representing slower growth (Figs. 1 and 2). An 
association between the transition timing and maturity 
has not been documented in Dover sole, but Abookire 
and Macewicz (2003) reported that 50% maturity occurs 
at 6.7 years which roughly coincides with the observed 
transition. They used specimens aged by the same ex- 
perienced age readers as in this study; hence some level 
of circularity exists. However, our studies’ results lead 
us to believe that the timing of the transition pattern 
is likely associated with the onset of maturation. In 
other species such as orange roughy ( Hoplostethus at- 
lanticus) a decrease in the annual growth zone width 
is documented to correspond to the onset of maturity 
(Francis and Horn, 1997). 
Considerations concerning category-2 results 
Category-2 A 14 C results confirm that Dover sole are often 
a difficult-to-age species. This difficulty was exemplified 
by the range in possible ages of even the youngest cat- 
egory-2 specimens and was especially evident in the two 
older category-2 specimens. The CV of 4.21% for these 
hand-picked specimens, where the majority (38 out of 
43) of otoliths were deemed to be clear, although better 
than the typical CV of 9.64% for Dover sole, was high in 
comparison to that for many other flatfish species aged 
at the AFSC (Kimura and Anderl, 2005). The category- 
2 otoliths are typical of many Dover sole samples aged 
at the AFSC where subjective decisions are made by 
necessity in the age reading process. 
Results from the five difficult-to-age specimens in 
category 2 did not indicate consistent over-aging or 
under-aging. Correct decisions in how to interpret the 
growth zone patterns were made for specimens 225 
and 188, and a reasonably good choice was made in 
specimen 95 especially when its high age was consid- 
ered. It is clear that the choice of a mid-range or older 
age was the most accurate for specimens 95 and 188 
when compared to the loess-smoothed curve of cat- 
egory-1 specimens. However, specimens 146 and 317 
were demonstrated to be over-aged in the comparison 
to category-1 specimens. The underlying difficulty in 
aging and interpreting Dover sole otoliths lies in the 
framework of splits that can make up a single trans- 
