VARIABILITY STUDIES ON 

 PAIR TRAWL DATA 



Because the data records of Japanese groundfish 

 catch-effort statistics that were available through 

 INPFC were summaries of many daily trawls, variances 

 about these statistics cannot be estimated. Therefore, we 

 estimated variances of the fundamental daily catch, 

 effort, and CPUE data for statistical interpretation 

 of CPUE values. We selected the 120-219 GRT class of 

 pair trawlers attached to the Japanese surimi mother- 

 ship fishery for the analysis because these trawlers 

 were probably the most representative vessels used in 

 the fishery for monitoring pollock CPUE trends (Table 

 2). 



The analysis by monthly and annual periods shows 



that standard deviations of mean catch, effort, and 

 CPUE values decreased substantially after 1973 when 

 the number of daily catch-effort records (sample size) 

 were increased from earlier years (Table 3). Variances 

 about the monthly and annual CPUE values calculated 

 from daily catch-effort records were small after 1965. Co- 

 efficient of variation (CV) for the 1964 and 1965 data was 

 19''c and d'^'c. when the sample sizes were small (Table 4). 

 When the sample size approached 300 (1966-69), CV 

 generally was below 4'^i. As the sample size increased and 

 approached 5,000 (1970-76) the CV decreased to the 1-2% 

 range. 



A comparison of annual mean CPUEs at the 95% con- 

 tidence level shows that pollock CPUE increased signifi- 

 cantly each year from 1963 to 1969, followed by a sig- 

 nificant decrease from 1969 to 1971 (Table 4). Another 



Table 3.— Statistics on variability studies on catch-effort data for pair trawlers of the 120-219 gross 

 registered ton class in the Japanese surimi (a minced fish product) mothership fishery in the eastern 

 Bering Sea, 1964-76. 



