52 
Fishery Bulletin 116(1) 
Materials and methods 
Sampling and preparation 
D. gigas samples were collected during 
various commercial jigging vessels oper¬ 
ating in 2009, 2013, and 2014 in the wa¬ 
ters of northern (off the Costa Rica Dome, 
[CRD]) and southern hemisphere (off the 
Peru Exclusive Economic Zone, [PE]) and 
equatorial Pacific Ocean (offshore waters 
of the central eastern Pacific, [CEP]) (Fig. 
2). All squid were frozen on board and then 
transported to the Key Laboratory of Sus¬ 
tainable Exploitation of Oceanic Fisheries 
Resources at Shanghai Ocean University 
where they were defrosted under normal 
room temperature. Dorsal mantle length 
(ML), fin length (FL), and fin width (FW) 
were recorded to the nearest 1 mm, and 
body weight (BW) was measured to the 
nearest 1 g. Sex and maturity stage were 
determined on the basis of visual evalu¬ 
ation of the gonad morphometric charac¬ 
teristics defined by Lipiilski and Underhill 
(1995). The gladius was extracted from its 
dorsomedial site in the mantle cavity and 
cleaned in an ultrasonic cleaner with dis¬ 
tilled water. Gladii which appeared to be 
damaged were eliminated. Statoliths were 
extracted for age determination and the 
number of growth increments were counted 
under the assumption that the increment 
is deposited daily (Jackson and Forsythe, 2002). 
Morphometric measurements 
Figure 2 
Map of the major surface currents (adapted from Anderson and 
Rodhouse, 2001) and the locations where jumbo squid (Dosidicus 
gigas ) were sampled in 2009, 2013, and 2014 off the Costa Rica 
Dome (CRD). Sampling locations off the CRD are indicated with 
open circles, sampling locations off the Peruvian exclusive eco¬ 
nomic zone (PE), are indicated with black triangles, and those in 
offshore waters of the central eastern Pacific (CEP) are indicated 
with black squares. Morphological features were used to distin¬ 
guish between populations in different geographic zones 
r Y (ML 0 
' T ML, 
( 1 ) 
A total of 562 gladii from the three regions were used 
in the subsequent analyses, and detailed information 
on the samples can be seen in Table 1. Eight morpho¬ 
metric characteristics of the gladius (gladius length 
[GL], conus length [CL], maximum width of conus 
[CW], proostracum length [PL], maximum width of 
prostracum [PW], length from anterior tip of conus to 
the widest point of proostracum [PWL], length from 
anterior tip of conus to proximal end of lateral plates 
[LPL], the width between two anterior apices of lat¬ 
eral plates [LPW]) were measured from each sample to 
the nearest 1 mm (Fig. 1). All morphometric measure¬ 
ments were made by a single researcher throughout 
our study. 
Statistical analyses 
All morphometric data used for the analyses of sexual 
dimorphism and geographic variation were standard¬ 
ized to remove the allometric effects of body size and 
meet the assumptions of normality. To accomplish this, 
raw data were transformed by using the following for¬ 
mula (see Lleonart et ah, 2000 for details): 
where F - one of the morphometric variables of the 
gladius; 
F, = the standardized value for the individual /; 
F; and ML; = the observed values of F and ML 
for the individual /; 
ML 0 - the arithmetic mean for the study popula¬ 
tion; and 
b can be predicted from the following formula: 
ln(F) = ln(a) + b\n(ML) + ee~ M(0,a 2 ), (2) 
where a = a parameter to be estimated; and 
a 2 - the variance for the normally distributed un¬ 
explained errors e. 
The standardized variables are represented by sub¬ 
script “S”, i.e. CL S , CW S , PL S , PW S , PWL S , LPL S , 
LPW S , FL S and FW S . 
One-way analysis of variance (ANOVA) and stepwise 
discriminant analysis (SDA) were used to analyze sex¬ 
ual dimorphism. The SDA is a method of linear model¬ 
ing to estimate the proportion of correct classification 
with a leave-one out stepwise jack-knife procedure. 
Owing to the small sample size of males, samples from 
off the CRD was excluded from this analysis. 
