Mathews et ai.: Effects of vessel disturbance on Phoco vitulino on glacial ice in Tracy Arm, Alaska 
199 
CO 
CD 
(/) 
03 
-Q 
>> 
C 
CD 
CD 
"5 
£ 
O 
-Q 
03 
_Q 
0 
01 
1 
0 8 
0.6 
0.4 
0.2 
0 
Focal observations 
during 
vessel approaches 
1 Cruise ship 
Tour boat 
Power boat 
Inflatable 
- - Skiff 
Kayak 
NMFS guideline distance 
* ■<- 
0 50 100 150 200 250 300 350 400 450 500 
Seal-vessel distance (m) 
Figure 7 
Estimated probabilities of hauled-out harbor seals ( Phoca vitulina ) entering the water 
when approached by a vessel as a function of seal-vessel distance and vessel type. Bars 
span the widths of the categorical distance bands used in this study. The National Ma- 
rine Fisheries Service recommends that vessels remain beyond 100 yds (=91 m) from 
seals on haulouts (NMFS guideline avoidance area is blocked out in gray). Confidence 
intervals are presented in Table 2. Data were collected in Tracy Arm fjord in Alaska in 
2001 . 
study area. Although we do not have a direct explana- 
tion for this pattern, possibilities include the predict- 
able nature of tour boats in terms of arrival times and 
vessel maneuvering, the skill of experienced vessel cap- 
tains, and seal habituation to specific boats. Some tour 
boats typically approach more slowly than other vessel 
types and then deploy motorized inflatables or skiffs to 
dispatch passengers, after which they continue slowly 
toward the glacier. Thus, another possible explanation 
is that the auxiliary craft preceding the tour boat may 
displace seals, functionally buffering the mother ship 
from producing disturbance. In addition, this could be 
a spurious result in which there was no actual tour 
boat effect, but because tour boats were always pres- 
ent at midday when seal haul-out probability was also 
highest, these two variables could end up being corre- 
lated without any cause and effect relationship, as sug- 
gested by Blundell and Pendleton (2015) and Jansen et 
al. (2015). 
The number of vessel-caused disturbances is a func- 
tion of the number of vessels of each type, how they 
are distributed, and the probability of disturbance at 
specific distances. Tour and private power vessels were 
the majority of vessels observed in our study; there- 
fore, they almost certainly caused more of the total 
disturbances, a function of both disturbance risk and 
the number of encounters, than did the less commonly 
observed cruise ships or kayaks, even though the prob- 
ability that tour and private power vessels caused a 
disturbance at a given distance was considerably low- 
er. Consistent and predictable vessel speed after ves- 
sels pass harbor seals is also important in minimiz- 
ing disturbance. Hoover-Miller et al. (2013) found that, 
although vessel captains were careful in approaching 
and passing seals, once the seals were behind the boat 
and out of view, some vessels accelerated, alerting the 
seals and occasionally causing them to flee into the wa- 
ter. We found that vessels that were stopped, often at 
their closest approach to the glacier — where there are 
typically lots of icebergs and seals — were more likely 
to cause disturbance than were vessels in motion. This 
pattern is consistent with the observations of Johnson 
and Acevedo-Gutierrez (2007), where stopped power 
boats caused disturbances of hauled out harbor seals at 
distances of up to 10 times those of boats that passed 
by at steady speeds, even if the speed was fast. 
Habituation and tolerance There were no days of our 
study at Tracy Arm without vessels. We recorded up to 
33 vessels in one day and the minimum number of ves- 
sels we recorded was close to, or exceeded, the average 
numbers from other studies in glacial fjords (Calam- 
bokidis et al., 1987; Jansen et al., 2010; Young et al, 
2014). Tracy Arm is only ~1 km wide (Fig. 1), which 
is 2-3 times narrower than Disenchantment Bay and 
Johns Hopkins Inlet, and most seals are clustered near 
the glacier face where vessels typically stop to view 
the glacier. Thus, it is likely that most seals in Tracy 
Arm were exposed to vessel traffic because they would 
likely have been within the diameter of a potential dis- 
