226 
Fishery Bulletin 107(2) 
Table 1 
Excursion duration and distance away from the bottom for two northern rock sole ( Lepidopsetta polyxystra) released with archi- 
val tags and recaptured in the eastern Bering Sea between 2003 and 2005. Minimum durations were limited to the collection 
frequencies of the tags which were 0.5 hour and 1.0 hour, respectively, for the northern and southern fish. Mean distances were 
weighted by the sampling frequency for the tags. Numbers in parentheses indicate the average maximum distance off bottom 
among excursions. Not all excursions could be identified with certainty for the southern fish because of the variable bathymetric 
terrain in the area where the southern fish resided. 
Duration away from 
bottom (h) Distance away from bottom (m) 
Identified Time 
Min. 
Max. 
Mean 
Min. 
Max. 
Mean 
excursions 
recordings 
Northern fish 
0.5 
6.5 
2.6 
2.2 
54.2 
10.4(13.8) 
78 
316 
Southern fish 
1.0 
7.0 
2.1 
3.0 
63.7 
12.2(13.4) 
154 
332 
Table 2 
Significance of tidal current direction and speed on the probability of a northern rock sole ( Lepidopsetta polyxystra) being off the 
bottom. Data are from one fish tagged in the more northerly area of the eastern Bering Sea shelf. Generalized additive modeling 
(GAM) with a binomial error term (0=on bottom, l=off bottom) was used to test probabilities for each month, during nighttime. 
n = the total number of timed tag depth recordings per month. Number in parentheses indicates the number of recordings when 
the fish was away from the bottom. Months of fewer than 10 nighttime off-bottom observations are excluded. 
Month 
Chi-square 
P-value 
n 
Direction 
Speed 
Direction 
Speed 
Oct. 03 
15.0 
1.3 
0.0018 
0.7310 
855 (28) 
Nov. 03 
9.2 
8.6 
0.0236 
0.0329 
969(12) 
Dec. 03 
10.9 
2.7 
0.0106 
0.4057 
1079(12) 
Jan. 04 
75.9 
2.0 
<0.0001 
0.5364 
1041 (126) 
Feb. 04 
40.7 
3.4 
<0.0001 
0.3313 
587(36) 
Mar. 04 
14.5 
2.8 
0.0022 
0.3991 
373 (22) 
Apr. 04 
17.6 
4.8 
0.0005 
0.1794 
288(39) 
location (Fig. 5). A seasonal migration is apparent in 
this path. After its release in July, the fish remained in 
the general vicinity of the release location for about 5 
months, then abruptly in January and early February 
2004, it migrated south approximately 200 km (straight 
line) from the release location to the southern most 
point of the path. In March and April, the fish nearly 
reversed direction and migrated to the north where it 
was recaptured. During the migration, the fish traveled 
an average of 6.4 km and maximum of 17.3 km per verti- 
cal excursion. The swimming speed which minimized the 
distance between the final migration path position and 
the reported capture location was 47 cm/s or 1.4 body 
lengths per second (BL/s) for the 34-cm fish. 
The bottom depths predicted at the locations along 
the migration path were very similar to the bottom 
depths (maximum depth within each 24-h period) mea- 
sured by the archival tag (Fig. 6) and had a mean ab- 
solute difference of only 2.5 m, thus corroborating the 
predicted migration path. The maximum depth during 
this migration occurred at the beginning of March when 
the fish abruptly changed its migration direction from 
the south to the northeast. 
Although a migration path for the southern fish could 
not be formulated, depth data from the archival tag 
(Fig. 2B), indicated that the fish must have remained 
along the Alaska Peninsula and did not migrate west 
toward the continental slope, or into the central Bering 
Sea shelf. Because predicted bottom depths gradually 
decreased from 90 m to 10 m over the first 10 months 
at liberty, the fish could not have migrated toward the 
slope. In addition, abrupt changes in bottom depth such 
as the 10-m to 40-m increase from 16 through 17 June 
2004 (Fig. 2B), indicated that the fish remained in an 
area of relatively steep bathymetry — an area that does 
not exist on the central shelf. 
Accuracy of tidal current prediction 
Overall, current direction was more accurately pre- 
dicted at the northern mooring site, where direction 
errors were less than 40 degrees during periods of the 
