Rodgveller et al: Evidence of hook competition in longlme surveys 
369 
Figure 2 
Scatterplots of grenadier (giant grenadier [Albatrossia pectoralis ]) versus sablefish (Anoplo- 
poma fimbria) catch rates and rockfish (shortraker \Sebastes borealis] and rougheye [ Sebastes 
aleutianus ] rockfish) versus sablefish catch rates from the National Oceanic and Atmospheric 
Administration Alaska Fisheries Science Center annual longline surveys, 1979-2003. Only data 
from the East Yakutat and Southeast management area are shown as an example. Each data 
point represents the average number of fish caught per 45 hooks (one skate of gear) at preferred 
depths at each station per year. Ninety percent density ellipses are included to demonstrate the 
trend of the relationship. Catch per unit of effort is abbreviated as CPUE. Data transformations 
are also abbreviated: square root (SqRt) and natural logarithm (Ln). 
Table 5 
Correlation coefficients (coefficient) between giant grenadier ( Albatrossia pectoralis ) (grenadier) and sablefish ( Anoplopoma fim- 
bria) catch rates and between rockfish (shortraker [ Sebastes borealis] and rougheye [ Sebastes aleutianus] combined) and sable- 
fish catch rates in the Gulf of Alaska. Data were collected during the National Oceanic and Atmospheric Administration Alaska 
Fisheries Science Center Alaska longline and trawl surveys (1984, 1987, 1990, 1993, 1996, 1999, 2001, 2003, and 2005). The 
Pearson’s correlation coefficient of the relationship between sablefish and either grenadier or rockfish is listed. The coefficients for 
nonparametric tests (Spearman’s p and Kendall’s r) are listed for the rockfish and sablefish comparison in the trawl data because 
these data were not normally distributed. The P-value associated with the significance of the correlation tests are shown as well 
as the sample size in). The sampling gear is specified as either longline (LL) or trawl (Trawl). 
Grenadier-Sablefish Rockfish-Sablefish 
Gear 
Test 
Coefficient 
P 
n 
Coefficient 
P 
n 
LL 
Pearson’s 
-0.45 
<0.0001 
327 
-0.24 
<0.0001 
355 
Trawl 
Pearson’s 
0.51 
<0.0001 
249 
Trawl 
Spearman’s p 
-0.08 
0.1070 
419 
Trawl 
Kendall’s t 
-0.05 
0.1102 
419 
Longline versus trawl correlations 
When longline data were summarized in the same way 
as the trawl data, the correlations between sablefish 
and grenadier and sablefish and rockfish catch rates 
were significantly negative (Table 5), just as they 
were in the analyses that were stratified by sablefish 
management area. Scatter plots of the joint catch 
rates of sablefish and grenadier (Fig. 5A) and sable- 
fish and rockfish (Fig. 5B) showed that when catch 
rates were low for grenadier and rockfish, they were 
high for sablefish, and vice versa. The 90% density 
ellipses illustrated the negative trend in the joint 
catch rates. 
Correlation coefficients between sablefish and grena- 
dier catch rates from the Gulf of Alaska trawl surveys 
were significantly positive and the correlations between 
sablefish and rockfish catch rates were not significantly 
different from zero. Because sablefish and rockfish 
catch rates within the rockfish preferred depth range 
could not be transformed to fit a normal distribution, 
nonparametric methods and tests of significances were 
