Nichol and Somerton: Tidal stream transport of Lepidopsetta polyxystra in the eastern Bering Sea 
227 
most frequently observed current speeds 
(10-40 cm/s; Fig. 7A). By comparison, 
errors of over 40 degrees were common at 
the southern mooring site, particularly 
when current speeds were less than 20 
cm/s (Fig. 7B). Model prediction of cur- 
rent direction improved with increasing 
observed current speed at both north- 
ern and southern mooring sites (Fig. 7). 
Model estimates of current speed at the 
northern mooring displayed progressive 
underestimation of the observed speed 
with increasing speed (Fig. 8A). When 
the observed current speeds were 33 cm/ 
s, for example, the model underestimated 
these speeds by an average of 10 cm/s. At 
the southern mooring, observed speeds 
were underestimated at slow speeds and 
overestimated at faster speeds (Fig. 8B). 
Discussion 
Archival tag data from two northern rock 
sole released and recaptured in the east- 
ern Bering Sea provide evidence that 
selective tidal stream transport can be 
used to aid horizontal migration. Verti- 
cal excursions, although infrequent, were 
correlated to both diel and tidal factors. 
Not only did both fish undergo verti- 
cal excursions during mostly nighttime 
hours, the northern fish did so during 
select periods of the tidal cycle when 
tidal currents were moving in a particu- 
lar direction. The significance of current 
direction as a determinate of vertical 
movement for the northern fish indicated 
that the fish did not randomly leave the 
bottom, but did so only when the current 
was moving in a specific direction. The 
successful application of tidal current 
information to predict a migration path 
for the northern fish validates that at 
least some northern rock sole use tidal 
currents for transport, and also may 
indicate that their vertical excursions 
are conducted primarily for the purpose 
of horizontal migration. 
Our attempt to predict a migration 
path for the southern fish was unsuccessful for several 
reasons. First, the identification of vertical excursions 
was less certain because of the variable bathymetric 
contours in the area. Second, model estimates of current 
direction were considerably less accurate when com- 
pared to the measured current direction at the location 
of the southern fish than for the location of the northern 
fish. Finally, the assumption for the model was that 
all horizontal movements occurred only during periods 
when the fish left the bottom; however, for the southern 
Northern fish 
B 
Southern fish 
Day (2003) 
Figure 3 
Consecutive nightly vertical excursions (shown as peaks) recorded for two 
female northern rock sole ( Lepidopsetta polyxystra ) tagged with archival 
tags in the eastern Bering Sea. Examples for both northern (A) southern 
fish ( B ) are presented. Nighttime periods are shaded. Model estimates 
of tide height (heavy line) at a position midway between fish release 
and recovery positions are provided to highlight the tidal fluctuation 
recorded by the tag and show that the timing of vertical excursions is 
related to tidal factors in addition to diel factors. 
fish there was evidence of additional horizontal move- 
ment. Gradual decreases in fish depth sometimes oc- 
curred during periods in which tidal fluctuations could 
be recognized in the depth record, indicating that the 
fish migrated to some extent while it remained on or 
close to the seafloor. With no way to account for these 
movements in the model, we were unable to accurately 
calculate the projection of the path. 
The estimated average swim speed of 1.4 BL/s (47 
cm/s) that minimized the distance between the final 
