Disadvantages: 
1. This is the most time-consuming of methods considered. It 
is very laborious to count food-diving otters when many are 
in a limited area. 
2. Interruption of observations by changing weather causes the 
results to be unreliable; movements of otters from exposed 
areas to more sheltered areas are influenced by weather 
conditions. 
3. Some areas cannot be viewed because offshore rocks limit 
visibility and rugged precipitous coastal areas are in- 
accessible to the observer; an unknown number of otters are 
missed. 
Our best estimate of sea otter populations is indicated by con- 
sidering data obtained by each method. The comprehensive aerial 
survey's greatest value is that it reveals otter distribution and 
relative magnitudes of different island populations. At the same time 
counts are obtained that provide a basis for estimates of total popula- 
tion. In 1959, in 30 hours of survey flight time, an area more than 
500 miles in length and containing more than 50 islands was covered. 
If such a survey had been attempted by boat the undertaking would have 
required an estimated 2 or 3 years, many times the expense, and we 
believe, would have yielded less reliable data. At the present apparent 
rate of otter population spread and increase, parts of a boat survey 
would be out of date before the operation was completed. Counts made 
from a dory and from the shore in sample areas furnish data that may 
be used as a rough check on aerial counts. They also yield other 
information which cannot be obtained from the air. 
RESULTS AND EVALUATION OF THE SURVEY 
Dory counts compared with aerial counts 
In a sample area 9.5 nautical miles in length along the north 
coast of Amchitka, bounded by Crown Reefer Point and Ivakin Point, 
counts were made from a dory on 5 different days (see Appendix A). 
The number of otters counted in the area varied. The mean of all 
counts was 157 otters. If the best counts are chosen (surface counts 
made under the most favorable weather conditions) the total for the 
area was 193. The aerial count for this area was 166. These figures 
indicate that the aerial count is usually as reliable as the dory 
counts. From the best boat counts (total 193) and the aerial count 
(total 166) a correction factor of 1.163 was obtained. When this is 
applied to the total Amchitka aerial count (1,560) the corrected count 
becomes 1,814. Treated similarly, the total outer Aleutian aerial count 
(9,507) becomes 11,057. 
Shore counts compared with aerial counts 
In a sample area, Kirilof Point, having a shoreline of approxi- 
mately 3 nautical miles, five careful counts were made from the shore 
4 
