182 



Fishery Bulletin 94(1), 1996 



sion and accuracy of the counts, respectively, were 

 demonstrated by the small CV's between multiple 

 counts of seals at a large rookery (Table 4) and be- 

 tween counts made from photographs and those by 

 persons on the ground (Fig. 1; Tables 2 and 3). 



This study was the first to use a large-format cam- 

 era with fine-grain, color transparency film and a 

 camera equipped with IMC. Image interpretation is 

 improved because colors that print as the same shade 

 of gray in a black-and-white photograph can be dif- 

 ferentiated in a color photograph. Large-format cam- 

 eras have better resolution than medium-format (70- 

 mm) or small-format (35-mm) cameras (Glegg and 

 Scherz, 1975) that are commonly used for aerial pho- 

 tographic surveys of pinnipeds (e.g. Stewart, 1989; 

 Hanan et al. 3 ). The IMC mechanism improves the 

 resolution in photographs taken with a large-format 



' Hanan. D. A , E. S. Konno, and M. J. Beeson. 1991. Harbor 

 • i itulina richardsi, census in California, May-June 

 1990 Southwest Fisheries Science Center, Natl. Mar. Fish. 

 Serv., NOAA. La Jolla, CA. Admin. Rep. LJ -91-05, 68 p. 



camera (compared with one not equipped with IMC) 

 by eliminating the forward motion of the aircraft 

 when the photograph is taken. The result is that color 

 photographs taken with a large-format camera 

 equipped with IMC record more detail in the photo- 

 graph and, thus, make it possible to interpret what 

 is on the ground precisely and accurately. 



Elephant seal births are used as a relative index 

 to measure population growth and to estimate popu- 

 lation size (Laws, 1994; Stewart et al., 1994). It is 

 not possible to determine the absolute number of 

 births for large rookeries because pup mortality prior 

 to the census can only be estimated. Stewart et al. 

 ( 1994) describes various methods to estimate north- 

 ern elephant seal births. Where comparisons in esti- 

 mates are possible with the various methods de- 

 scribed by Stewart et al. (1994), large within-year 

 discrepancies exist (Table 5). These differences can 

 be attributed to several factors: 1) the large variance 

 associated with ground counts of pups and with phe- 

 nological studies; 2) differences between the dates 

 each census was conducted (storms are a known 



