Chang et al. : The Antarctic Oscillation index as an environmental parameter for predicting catches of lllex argentinus 
207 
Table 1 
The coefficient of correlation (r) between the log-transformed catch per unit of effort (log U) of Argentine shortfin squid 
(lllex argentinus) in the southwest Atlantic during 1986-2010 and the environmental variables in the same fish- 
ing season and with time lags of the previous 1 and 2 years. Variables with P-values <0.05 are shown, includ- 
ing subsurface seawater temperature on the northern and southern Patagonian shelf and Antarctic Oscillation 
(AAO) indices. Note that N_5m and S_5m stand for the subsurface seawater temperature at a depth of 5 m at the 
northern (36°S, 53°W) and southern (5Q°S, 60°W) Patagonian shelf, respectively. The fishing season lasts from the 
November of the previous year to the June of the current fishing season. 
Environmental 
factors 
Previous 2 years 
Previous 1 year 
Fishing season 
Month 
r 
P 
Month 
r 
P 
Month 
r 
P 
N_5m 
_ 
_ 
_ 
March 
-0.452 
0.030 
_ 
_ 
_ 
S_5m 
- 
- 
- 
March 
-0.477 
0.021 
February 
-0.573 
0.005 
- 
- 
- 
- 
- 
- 
March 
-0.596 
0.003 
- 
- 
- 
- 
- 
- 
April 
-0.573 
0.005 
AAO 
November 
-0.478 
0.016 
- 
- 
- 
December 
0.401 
0.047 
December 
-0.564 
0.003 
- 
- 
- 
- 
- 
- 
March 
0.565 
0.003 
- 
- 
- 
- 
- 
- 
May 
0.436 
0.030 
- 
- 
- 
- 
- 
- 
Table 2 
List of coefficients of generalized linear models used for analyses of the influence of environmental factors on 
catch per unit of effort (CPUE) for Argentine shortfin squid (lllex argentinus) in the southwest Atlantic, P-values, 
coefficients of multiple determination (R 2 ), and Akaike’s information criteria (AIC) that resulted from backward 
stepwise procedure for predicting the squid’s abundance. The dependent variable is the log-transformed annual 
CPUE (logCT) from 1986 through 2007. Note that PPANov, PPAMar, and PPADec represent 2-year-lagged Antarc- 
tic Oscillation indices in November, March, and December, respectively, before the fishing season; S_5m_Mar and 
PS_5m_Mar stand for subsurface seawater temperatures at a depth of 5 m at the southern Patagonian shelf in the 
concurrent and previous March, respectively; and PN_5m_Mar represents the subsurface seawater temperatures 
at a depth of 5 m at the northern Patagonian shelf in the previous March. 
Environmental variables 
Model 
Intercept 
S_5m _ Mar 
PPANov 
PPAMar 
PS_5m _ Mar 
PPADec PN_5m _ Mar 
R 2 (%) 
AIC 
1 
12.186 
-0.278 
(0.003) 
35.6 
27.61 
2 
12.010 
-0.265 
-0.295 
63.7 
16.97 
(0.006) 
(0.001) 
3 
11.336 
-0.202 
-0.249 
0.309 
76.4 
9.548 
(0.002) 
(0.001) 
(0.006) 
4 
12.975 
-0.176 
-0.273 
0.302 
-0.176 
83.3 
3.927 
(0.003) 
(0.000) 
(0.003) 
(0.017) 
5 
12.799 
-0.154 
-0.213 
0.332 
-0.178 
-0.083 
85.6 
2.617 
(0.007) 
(0.005) 
(0.001) 
(0.012) 
(0.127) 
6 
12.019 
-0.163 
-0.216 
0.354 
-0.175 
-0.086 0.038 
86.0 
4.001 
(0.007) 
(0.006) 
(0.002) 
(0.016) 
(0.126) (0.524) 
fin squid, as shown by the CPUE values estimated 
from data from Taiwanese jigger fishing vessels in the 
southwest Atlantic. We found that the subsurface sea- 
water temperature (at a depth of 5 m) and regional 
atmospheric forcing, represented by the AAO, may be 
important indicators of annual variation in Argentine 
shortfin squid abundance in the southwest Atlantic. 
The subsurface seawater temperature in the southwest 
Atlantic during the main fishing season (February, 
March, and April) and the previous year of the fish- 
ing season were correlated negatively with the logZ7 
of this species. The AAO had high correlation with a 
2-year-lagged logt/ of this squid. Our predictive model 
included 4 parameters and explained 83.3% of the an- 
