354 



Fishery Bulletin 105(3) 



or dirty ice, although this advantage probably di- 

 minished with distance. Stationary seals could still 

 be difficult to distinguish at any distance, although 

 their characteristic profiles and the reflectivity of 

 their fur helped to distinguish seals from rocks or 

 dirt of similar size. As was the case for the aerial es- 

 timates, no correction was available for seals missed 

 or misidentified by shore-based observers, and some 

 unknown number of seals were in the water during 

 the observation periods. 



Advantages of each method 



Shore-based counts The main benefit associated with 

 counting harbor seals from land is the ability to obtain 

 multiple counts throughout the day, and on successive 

 days, relatively inexpensively.' Counts can also be made 

 from land under suboptimal weather conditions, when 

 aerial surveys are impossible or when the resulting 

 photographs would be of poor quality. Repeated surveys 

 allow an assessment of changes in seal counts related 

 to covariates such as time of day, ice conditions, and 

 weather (Mathews and Pendleton, 2006). The number 

 of harbor seals counted on ice in Johns Hopkins Inlet 

 varies from day to day and thus increases the probabil- 

 ity that a small number of aerial surveys could yield 

 misleading results. For example, on 10 consecutive days 

 in August 1999 the number of seals counted from shore 

 near midday ranged from 1465 to 2534. If surveys are 

 conducted when pups are nursing (generally during 

 June in Alaskan waters), shore observers can identify 

 seal pups based on size, shape, and relative position 

 of seals within a group hauled out on ice. By August, 

 however, almost all pups are weaned and the number 

 of seals in groups on ice is much larger than in June, 

 making it much more difficult to distinguish weaned 

 pups or juveniles from adults except at very close range. 

 With the aerial surveys there were also difficulties in 

 distinguishing pups and juveniles (i.e., the resolution 

 of the aerial photographs in the August study was not 

 high enough to distinguish weaned pups, and no large- 

 format aerial surveys were conducted during June when 

 dependent pups are more likely to be distinguishable 

 from adults). 



Aerial photography Large-format aerial photography 

 allows investigators to count seals from a set of images 

 taken at a consistent distance (altitude) from the seals 

 without potential "blind spots" caused by land or ice 

 features. Photographs can be taken with overlapping 

 images so that a mosaic of the complete study area 

 can be obtained for each sampling event, and so that 

 ice movement can be taken into account. The ability 

 to view seals from a vertical perspective, rather than 

 obliquely from a shore-based observation site, removes 

 many of the potential biases associated with sight- 

 ing seals at variable distances from the shoreline. The 

 photographs also represent a permanent record of the 

 distribution of the seals within a fjord — a record that 

 allows recounts or re-analyses of images. For example. 



the primary analyst was able to count seals in each 

 image independently three times to estimate variance 

 in the number of seals recorded; a secondary analyst 

 was also able to count seals in a subsample of the same 

 images to provide an independent verification of the final 

 estimates. Aerial photography also offers one the abil- 

 ity to evaluate the spatial distribution of seals within a 

 study area in relation to other seals (e.g., social interac- 

 tions) and to environmental features (e.g., ice types or 

 shifting ice patterns). A final advantage of using aerial 

 photography is that researchers are not required to 

 establish and maintain a remote field camp throughout 

 the study period. 



Future surveys of harbor seals in glacial fjords 



The development of reliable methods for surveying 

 harbor seal abundance in glacial ice habitats is a funda- 

 mental requirement for estimating the population size 

 of these seals in Alaska. Conventional aerial surveys 

 of harbor seals at terrestrial haul-out sites indicate 

 that approximately 180,000 seals may be found at ter- 

 restrial sites. Preliminary counts of harbor seals from 

 large-format photographs taken in glacial ice habitats 

 throughout Alaska indicate that as many as 20,000 to 

 25,000 additional harbor seals may be using glacial ice 

 habitats (J. L. Bengtson, unpubl. data). If 10% or more 

 of Alaska's harbor seal population are using glacial 

 ice habitats at various times of the year, monitoring 

 trends in seal abundance in these areas will be very 

 important to resource managers and to subsistence 

 hunters in the Alaska Native community. In some 

 regions, a much larger proportion of harbor seals may 

 use glacial ice habitats. Within Glacier Bay, an aver- 

 age of 72% of harbor seals surveyed between 1992 and 

 2001 (2400-4700 seals per year) were found within 

 glacial fjords during the breeding season (Mathews 

 and Pendleton, 2006). At present, there are about 20 

 sites in Alaska where harbor seals are known to haul 

 out in glacial ice habitats. Several of these fjords are of 

 special interest to resource managers because 1) some 

 local seal populations may be declining, 2) the fjords 

 are important hunting areas to Alaska Natives, and 

 3) logistical difficulties have hampered past efforts 

 to monitor changes in seal abundance with standard 

 survey methods. 



In the future, which survey methods seem most ap- 

 propriate for monitoring the abundance of harbor seals 

 in glacial ice habitats? Both shore-based counts and 

 aerial photography are valuable methods for monitoring 

 seals in glacial fjords, and each method has different 

 limitations and potential applications. Unlike Johns 

 Hopkins Inlet, many glacial fjords in Alaska do not 

 have an overlook with such a full view of seal habi- 

 tat, and thus large-format aerial photography may be 

 the only option for surveying seals in these important 

 breeding areas. The present study demonstrates that 

 large-format aerial photography is a promising method 

 for surveying the abundance of harbor seals using gla- 

 cial ice habitats in Alaska. 



