516 



Fishery Bulletin 105(4) 



45°N - 



40°N 



35°N 



30°N 



45°N 



40°N 



35°N 



30°N - 



130'W 125°W 120 W 130°W 125°W 120°W 



Figure 2 

 Sighting locations (•) for the species most commonly seen on the 1991-2005 

 surveys. Light gray lines indicate transects surveyed, and black lines indicate 

 the four geographic regions. 



nose dolphins, and Pacific white-sided dolphins) were 

 found with other species more often than they were 

 found alone, indicating that these associations were 

 not coincidental. 



Calibration of group size 



Most regression coefficients for the indirect method of 

 group-size calibration were less than one, indicating that 

 observers were more likely to underestimate group size. 

 For all groups and all species in this study, the ratio of 

 the sum of all uncalibrated group sizes divided by the 

 sum of all calibrated group sizes was 0.79. The mean 

 ratio of calibrated to uncalibrated group size estimates 

 was 0.92. This difference implies that proportionately 

 larger corrections were applied to larger groups. 



Probability of detection along the trackline 



New trackline detection probabilities, giO). were esti- 

 mated from sightings that were made by the independent 

 observers but missed by the primary observers (Table 

 3). Beaufort sea state was not a significant factor in 

 the numbers of delphinids or large whales seen by the 

 independent observers (Fisher's exact test, P=0.60 and 

 0.87, respectively). Group size was a significant factor for 

 delphinids (P<0.0001), but not for large whales (P=0.79). 

 Consequently, estimates of giO) for delphinids were 

 stratified by group size. The number of Dalls porpoise 

 sightings by independent observers (n=12) was insuffi- 

 cient to stratify estimates ofg(O) for this species. Values 

 of^(O) from the literature that were used for other long- 

 diving species are given in Table 4. 



