computed using Formula #3. This formula seems to combine the advantages 

 of a line transect estimator with that of a strip survey. It gave these 

 investigators some degree of confidence that computed estimates generally 

 agreed with our feel for the numbers and densities of the animals seen. 



There remains a basic schism between choosing a conservative but 

 realistic formula for computation of density and area-wide population 

 estimates on one hand, and, on the other, the sure knowledge that we are 

 underestimating actual population numbers with existent survey methods. 

 There is no reason to believe that new improved or more precise survey 

 methodologies will become available in the foreseeable future. There- 

 fore, we prefer to maintain a more conservative reproducible data analysis 

 technique that will allow subsequent investigators to reasonably compare 

 their data base and those data collected during this three-year study. 



Total population numbers would be of considerable interest, but we 

 believe they are unobtainable at this time. To monitor the health of the 

 cetacean community and to look at future population trends, seasonal 

 sampling will probably be the method of choice. Given that actual popula- 

 tion numbers are unknowable, a rigorous and reproducible sampling tech- 

 nique would appear to be the method of choice." 



5.5 THE TAPE 



The tape supplied with this report is an IBM Standard Label Tape. It is 

 written at 6,250 BPI. It contains three files whose DCBs are; 



LRECL = 80 

 BLKSIZE = 3120 

 RECFM = FB 



File One contains BLM data described in the previous section. The data set 

 name is BLMDATA. It is 400 blocks in length. 



File Two contains the SAS program to analyze data from sighting cards. The 

 data set name is SIGHT. CARD. PGM. The data set is 19 blocks long. 



File Three contains the SAS program to analyze data from in-person interviews, 

 The data set name is INTER. VIEW. PGM. The data set is 8 blocks long. 



B-74 



e@ 



