on the ability of the counter to distinguish a 

 sea lion from its surroundings. This requires 

 field experience with the animals that enables 

 the counter to visualize the size of an animal 

 in relation to the distance and angle from the 

 axis of the lens. Some sea lions will not be 

 detected at all because they may be partly 

 or wholly obscured by rocks or other animals. 

 Therefore, counts of the animals in the photo- 

 graphs will always be minimal. The greatest 

 deviation from the actual population may be 

 expected in heavily populated rookeries on 

 rough terrain. 



Consistency in interpretation is perhaps 

 the most important factor in analyzing 

 photographs. This has been tested by com- 

 paring the counts of three different persons 

 who made independent readings from duplicate 

 series of photographs on Chernabura Island. ^ 

 Ron Lopp, an experienced counter, was desig- 

 nated as counter A and read a series of photo- 

 graphs twice, using unmarked pictures in every 

 case. His deviations generally amounted to 

 only a few animals between first and second 

 readings, regardless of whether the photo- 

 graph included fewer than 10 or even when 

 there were as many as 250 animals (fig. 4). 



Counter B was an experienced aerial stream 

 surveyor with long training in estimating 

 salmon escapements. At his first attempt to 

 count sea lions on photographs, counter B 

 deviated from counter A, the experienced 

 counter, particularly on photographs with large 

 numbers of animals. On these, his estimates 

 were lower than those of the experienced 

 counter. 



Counter C was a clerical assistant without 

 any previous experience. Her deviations from 

 counter A were considerably larger than those 

 between counter A and counter Band generally 

 showed a positive bias. 



Lopp made all final readings on our ex- 

 periments with the exceptionof Prince William 

 Sound-Kenai district on the September survey 



' Baade, Robert T., Ole A. Mathisen, and Ron J. Lopp. 

 Studies on tlie Steller sea lion (Eumetnpias jubataj on 

 Qiernabura Island in the Sliumasiii area of Alaska dur- 

 ing the summer of 1958. Typed manuscript , on file. 

 Fisheries Research Institute. 



20 40 60 80 100 120 140 160 180 200 220 240 260 

 Number of seo lions on first count 

 by counter A 



Figure 4.- -Comparison of results of three counters in- 

 terpreting independently the same series of photo- 

 graphs of Chernabura Island sea lion rool<ery. 



and all districts on the December survey in 

 1957. These were read by Donald E. Rogers. 



Summary of Counts 



Summarized in tables 1 to 6 are counts 

 obtained during the years 1956-58. Each year 

 the surveys were numbered consecutively. 

 Three surveys were made in 1956, seven in 

 1957, and three in 1958. The surveyed region 

 was divided into six districts whose boundaries 

 are outlined on the map in figure. 1. These 

 districts are artificial units created for ease 

 of description and do not reflect groups or 

 races of sea lions. Within each district each 

 rookery or hauling ground has been given a 

 number, which is listed in tables 1 to 6 and 

 which corresponds to a number on the maps 

 in figure 1. Rookeries shown on figure 1 that 

 are not included in the tables had harbor seals 

 (Phoca viiuUnn) only. The counts of harbor 

 seals are given in appendix table 1. 



Counts of adults and pups are separated. 

 Counts of adults in the water and estimated 

 animals have been added to the counts of 

 adults on the rocks for the total figure of 

 adult animals. All photographs and detailed 

 basic counts are available for inspection in the 

 library of the Fisheries Research Institute. 



