Mathews et al.: Effects of vessel disturbance on Phoca vitulina on glacial ice in Tracy Arm, Alaska 
193 
Table 1 
Analyses of counts of harbor seals (Phoca vitulina) recorded in Tracy Arm, Alaska in 2001. Means, adjusted for other vari- 
ables in the model ( i.e. , least-squares means), are presented for categorical variables. For continuous variables, “%change” 
is the percentage increase in counts for a 1-unit increase in the predictor. Analyses were conducted separately for nonpups, 
pups, and total counts. For means and change percentages, 95% confidence intervals are given in parentheses. 
Nonpups 
Pups 
All seals 
Variable 
P 
Mean or %change 
P 
Mean or %change 
P 
Mean 1 or %change 
Count quality 
<0.01 
<0.01 
<0.01 
1.5 
350.6 (301.0-408.5) 
106.5 (90.8-128.2) 
466.1 
(401.4—541.2) 
2 
355.5 (303.4-416.6) 
111.0 (94.4-130.6) 
477.4 
(408.6-557.7) 
2.5 
399.9 (348.7-458.6) 
130.8 (114.0-150.0) 
543.2 
(474.9-621 .4) 
3 
389.1 (342.3-442.3) 
127.5 (113.4-144.4) 
524.1 
(462.0-594.4) 
3.5 
486.3 (408.9-578.4) 
167.7 (142.5-197.1) 
675.6 
(570.6-800.0) 
4 
476.3 (417.7-543.0) 
154.2 (136.6-174.0) 
645.4 
(567.4-723.3) 
4.5 
466.4 (352.7-616.9) 
176.9 (134.3-233.1) 
652.7 
(497.5-856.3) 
5 
414.3 (318.9-538.4) 
125.4 (92.0-171.6) 
528.3 
(404.9-689.2) 
6 
409.2 (301.6-555.2) 
125.9 (88.7-179.1) 
532.9 
(391.2-725.8) 
Day of year ( DOY ) 
<0.01 
2 
<0.01 
2 
<0.01 
2 
DOY'DOY 
0.28 
<0.01 
2 
0.58 
Time of day (TOD) 
0.17 
0.16 
0.19 
TOD'TOD 
<0.01 
3 
<0.01 
3 
<0.01 
3 
Ice cover (%) 
<0.01 
0.80% (0.43-1.18) 
<0.01 
1.22% (0.80-1.63) 
<0.01 
0.93% (0.57-1.31) 
Sky 
<0.01 
0.04 
<0.01 
Clear 
516.8(426.1-626.7) 
164.7 (134.0-202.3) 
698.9 
(577.7-845.5) 
Partly cloudy 
369.4 (326.2-418.4) 
119.1 (104.8-135.6) 
492.9 
(436.2-556.9) 
Overcast 
371.0 (335.2-410.7) 
123.5 (109.7-139.8) 
498.9 
(451.7-551.0) 
Temperature 
0.05 
1.27% (-0.01-2.57) 
0.01 
1.84% (0.45-3.24) 
0.03 
1.35% (0.08-2.63) 
Precipitation 
0.01 
0.41 
0.05 
None 
470.5 (435.3-508.5) 
612.8 
(567.0-662.4) 
Mist or light rain 
391.6 (337.1-454.9) 
523.6 
(451.2-607.8) 
Heavy rain 
384.4 (303.1-487.6) 
535.5 
(426.7-672.1) 
Wind speed 
0.16 
0.05 
0.10 
Beaufort 1 
127.6 (114.1-142.8) 
Beaufort 2 
147.2 (127.9-169.8) 
Beaufort 3 
129.0 (103.9-160.7) 
Wind direction 
0.55 
0.76 
0.66 
Vessels 
0.65 
0.30 
0.55 
because models were fitted separately for nonpups, pups, and all seals, the effect of the predictors differs, and the means for 
“all seals” are not the sum of the means for nonpups and pups. 
2 See Figure 2. 
3 See Figure 3. 
than when they were not (Fig. 5, vessel type pooled). 
Randomly selected seals were more likely to enter the 
water when inflatables or kayaks were present in the 
study area (Fig. 5, vessel type). However, seals entered 
the water less often when there was a tour boat in the 
inlet than when there were none. The estimated effects 
of other vessel types were imprecise, either because of 
small sample sizes or variable seal responses for seals 
on randomly selected patches of ices. When all vessel 
types were pooled, both the number of vessels and the 
presence of at least one vessel were associated with 
an increased probability of a seal entering the water 
in relation to the probability of entry when no vessels 
were present (Fig. 5, vessel type pooled). 
The odds of a seal from a randomly selected fo- 
cal group entering the water were 3.7 times (95% Cl 
:2.6-5.4) greater when vessels were in the 2 shortest 
distance classes (0-50 m or 51-100 m) than when they 
were within the longest distance class (> 300 m). The 
estimate for the remaining distance class (101-300 m), 
although imprecise, also indicated an increased likeli- 
hood of a seal entering the water than when boats were 
>300 m away (Fig. 5, vessel distance). 
Vessel approaches In 2001, vessels entered the inlet 
on all 34 days when vessel entries were tabulated. We 
monitored 141 vessels (of 348 seen) and 1199 icebergs 
with 1755 harbor seals for disturbances of seals as ves- 
