232 
Fishery Bulletin 107(2) 
conducted (Acuna and Lauth, 2008). From the data on 
the two fish examined here, northern rock sole remain 
on the bottom 99.8% of the time during summer (June 
and July) daylight hours. 
Although we could not predict a migration path for 
the southern fish, it was apparent that the migration 
pattern differed between the two fish. The northern 
fish clearly used tidal currents to facilitate a south- 
ern migration to deeper water during winter and a 
migration back north during spring. These movements 
are consistent with the seasonal spawning and post- 
spawning migrations suggested by Fadeev (1965) and 
Shubnikov and Lisovenko (1964). As with some Euro- 
pean plaice, which migrate south to warmer waters 
for spawning (Hunter et al., 2004a), northern rock sole 
that reside on the northern part of the eastern Ber- 
ing Sea shelf during summer (i.e., the northern fish) 
may require a migration to more southern or deeper 
waters to reach temperatures suitable for spawning. 
Adult rock sole (likely L. polyxystra ) from the western 
Bering Sea also undergo a migration to deeper water 
in winter, and do so presumably to avoid temperatures 
below 0°C (Shvetsov, 1979). Temperatures experienced 
by both fish decreased during winter months but sta- 
bilized to about 2°C in February and March. Had the 
northern fish stayed in the vicinity of its release, it 
would have experienced bottom temperatures below 
0°C in February, as recorded by instruments at the 
northern oceanographic mooring site. The southern 
fish also underwent nighttime vertical excursions that 
were tidal in nature, but unlike the northern fish, there 
was no indication of a spawning migration; excursion 
frequency did not increase before the known spawning 
season (winter-spring), and depth records indicated 
no repeatable pattern from one winter (2004) to the 
next (2005). It is logical to assume that the extent of 
migrations is dependent on the proximity of feeding 
and spawning locations. Thus, the northern fish may 
require a directed seasonal migration to reach a vi- 
able spawning location, whereas the southern fish can 
remain resident if suitable feeding and spawning loca- 
tions are within close proximity. 
If the northern fish migrated south for the purpose 
of spawning, the southern extent of the migration route 
may have been a spawning location. We can infer from 
the spatial distribution of the fishery for roe of northern 
rock sole — a fishery that operates in the eastern Bering 
Sea during February and March just before the spawn- 
ing season (Stark and Somerton, 2002; Wilderbuer and 
Nichol, 2007) — that spawning aggregations occur over 
a wide area of the central and outer continental shelf 
extending from Unimak Island to west of the Pribilof 
Islands. This distribution overlaps with the southern 
point of the migration path. 
The eastern Bering Sea shelf offers a multitude of 
possibilities for tidally assisted transport, and the dis- 
tribution range that individuals seasonally inhabit may 
partly depend on the nature of the tidal currents. Based 
on tidal current ellipses for the M 2 tidal constituent in 
the eastern Bering Sea, tidal currents are rotary in 
nature over the majority the shelf area south of lati- 
tude 60°N but become more bidirectional (i.e., 60- and 
240-degrees) close to the Alaska Peninsula and into 
Bristol Bay (Pearson et al., 1981; Kowalik, 1999). Be- 
cause the northern fish inhabited the central part of 
the eastern Bering Sea continental shelf, opportunities 
for selective tidal stream transport were available in 
all directions, thus enabling a round-trip migration. By 
comparison, the southern fish resided along the Alaska 
Peninsula; therefore opportunities for selective tidal 
stream transport were limited to northeasterly and 
southwesterly directions. Fish that undergo migrations 
in northeasterly and southwesterly directions could 
use selective tidal stream transport over much of the 
eastern Bering Sea shelf. Adult yellowfin sole ( Limanda 
aspera), for example, are known to migrate annually in 
a northeasterly direction more than 500 km from winter 
grounds west and southeast of the Pribilof Islands to 
nearshore summer spawning grounds in Kuskokwim 
and Bristol bays ( Wakabayashi, 1989). Given the extent 
of this migration and the availability of tidal currents, 
it is reasonable to assume that yellowfin sole also use 
selective tidal stream transport. 
Results presented here provide the first known evi- 
dence of selective tidal stream transport among aquatic 
animals in the eastern Bering Sea. Among larval flat- 
fish in the eastern Bering Sea, including northern rock 
sole, passive forms of transport involving wind-driven 
surface currents and geostrophic flow have been shown 
to contribute to their horizontal distribution and like- 
lihood of survival (Wilderbuer et al., 2002; Lanksbury 
et al., 2007). The contribution of more active forms 
of transport such as selective tidal stream transport 
may become evident as more is learned about the verti- 
cal migration behavior of larvae. Evidence that larval 
northern rock sole as small as 8 mm can regulate their 
depth in the water column (Lanksbury et al., 2007) 
indicates that selective tidal stream transport is a pos- 
sibility. As we learn about how adult, juvenile, and 
larval fishes use tidal currents for migration, the need 
becomes evident for more accurate tide-prediction mod- 
els that can be used for modeling fish migration. Such 
models should become available in the near future with 
the completion of a baroclinic tide model of the eastern 
Bering Sea. 
Acknowledgments 
We thank D. Rachel and P. Stabeno for providing tidal 
information at Pacific Marine Environmental Labora- 
tory (PMEL) mooring sites, and for providing insight 
concerning tides in the eastern Bering Sea. A. Greig, S. 
Kotwicki, and J. Benson provided technical support on 
use of ArcView and circular trigonometry. T. Wilderbuer 
and J. Gauvin provided productive discussion concern- 
ing flatfish spawning and migration. Finally, K. Baily, 
T. Hurst, T. Wilderbuer, J. Lee, S. Cadrin, and three 
anonymous reviewers improved the manuscript with 
insightful suggestions. 
