Mathews et at: Effects of vessel disturbance on Phoca vituhna on glacial ice in Tracy Arm, Alaska 
195 
h a m h m h m a m H m 
> >> >> >> >> >> 
2001 2002 2003 2004 2005 2006 
( 33 , 0 ) ( 12 , 10 ) ( 14 , 6 ) ( 5 , 2 ) ( 8 , 5 ) ( 6 , 6 ) 
2012 & 2013 
( 30 , 15 ) 
ESS3 Kayak 
cznSkiff 
E23 Inflatable 
CC3 Power boat 
i 1 Tour boat 
■■■Cruise ship 
-•-TA + EA 
Figure 4 
Mean daily counts of vessels, by type, for the period 2001-2006 in Tracy Arm (TA) and 
Endicott Arm (EA), Alaska, with counts for 2012 and 2013 (average for 2 years) included 
for comparison. The numbers of survey days in each inlet are listed (in parentheses) 
below year. 
warmer temperatures, and clear skies; more nonpups 
hauled out when there was no precipitation, and more 
pups hauled out when wind speed was intermediate. 
Other studies also have found that harbor seal counts 
are affected by environmental factors, but such fac- 
tors can be site specific or might not be detected in a 
study because of the limited range of conditions un- 
der which surveys were conducted (e.g., Boveng et ah, 
2003; Simpkins et ah, 2003; Jemison et ah, 2006, Jan- 
sen et al., 2015 ). Although harbor seal haul-out pat- 
terns on glacial ice typically are not affected by tide 
stage (Calambokidis et ah, 1987; Boveng et al., 2003; 
Mathews and Pendleton, 2006; but see Hoover-Miller 
et al., 2011), they are affected by the availability of 
floating ice. Calambokidis et al. (1987) and Young et al. 
(2014), both working in GBNP, found that seal counts 
were positively related to percent ice cover, the propor- 
tion of seals counted in the water was negatively re- 
lated to percent ice cover, and, when ice cover was low, 
hauled-out seals were concentrated on the few remain- 
ing icebergs. Jansen et al. (2015) reported that seals 
were most abundant with intermediate ice densities 
(5-7 tenths coverage). 
As with other harbor seal studies in glacial ljords 
in southeastern Alaska, we found peak pup counts in 
late June (Mathews and Pendleton, 2006, Jansen et al., 
2015). However, Hoover-Miller et al. (2011) reported 
peak pup counts in Aialik Bay, KFNP, in early- to mid- 
June, indicating regional variation in pup birth dates. 
The maximum proportion of pups from our counts (30- 
36%) was also similar to the high values reported for 
seals in other tidewater glacial fjords (KFNP: 21-34%, 
Hoover, 1983, Hoover-Miller et al., 2011; GBNP: 34- 
40%, Calambokidis et al., 1987, Mathews and Pendle- 
ton, 2006), but ~3 times greater than the proportion 
of pups in Disenchantment Bay (Jansen et al., 2015). 
Note also that the proportion of pups can be affected 
by immigration or emigration of nonpups in addition to 
changes in productivity. 
Seal haul-out patterns 
The number of nonpups that were hauled out peaked 
at around 1300 (Fig. 3) — a number similar to patterns 
seen with other ice-associated harbor seals (Calam- 
bokidis et al., 1987; Mathews and Pendleton, 2006; 
Hoover-Miller et al., 2011, Blundell and Pendleton, 
2015). Using telemetry data, Blundell and Pendleton 
(2015) found that seals in Tracy Arm and Endicott Arm 
had the highest haul-out probabilities in the middle 
of the day, especially during the pupping season. Our 
randomized observations indicated that the probability 
of a seal entering the water was not related to TOD, 
but the time span of our observations might have been 
too narrow to detect a pattern if the afternoon peak 
was broad. Blundell and Pendleton (2015) found seals 
more likely to end haul-out bouts (i.e. , enter the water) 
later in the day. 
In contrast to the pattern observed for nonpups, 
the predicted number of pups hauled out in Tracy 
Arm, after accounting for other predictors, showed no 
distinct diurnal pattern (Fig. 3), although the observed 
counts indicate a slight peak. In Aialik Bay, mother- 
pup pairs also were less influenced by environmental 
