Karlou-Riga et al.: Sex change and oscillating growth pattern of Spicara smaris in the Saronikos Gulf (Greece) 
353 
Table 2 
Growth parameters of picarel (Spicara smaris) calculated by the different runs (classic, sex specific, and seasonal) of a von 
Bertalanffy growth function (VBGF) model for different subsets of individuals (all, males, and females): L . =the mean asymp¬ 
totic total length; ff=the growth coefficient; C=the amplitude of the growth oscillation; and t s or summer point =the beginning 
of the sinusoid growth oscillation with respect to t= 0. Standard errors are given in parentheses. MSE=mean squared error; 
n=number of specimens in each subset. Lengths are total lengths. 
VBGF 
Data subset 
(cm) 
K 
t 0 (year) 
C 
ts 
MSE 
n 
Classic 
All fish 
19.91 (0.480) 
0.512 (0.059) 
-0.412 (0.145) 
0.015 
1694 
Males and females 
19.93 (0.697) 
0.433 (0.072) 
-0.843 (0.265) 
0.011 
Sex specific 
Females 
16.52 (0.339) 
0.973 (0.158) 
-0.040 (0.163) 
0.009 
951 
Males 
20.29 (0.656) 
0.455 (0.079) 
-0.745 (0.288) 
563 
Seasonal 
Females 
16.95 (0.461) 
0.757 (0.142) 
-0.312 (0.251) 
1.261 (0.569) 
-0.457 (0.052) 
0.018 
951 
Males 
20.94 (0.844) 
0.379 (0.072) 
-0.701 (0.179) 
1.261 (0.569) 
-0.572 (0.074) 
563 
Age interpretation criteria 
Because the picarel spawning period lasts from Feb¬ 
ruary to June (Suppl. Fig. 4), the fish spawned dur¬ 
ing this period constituted one cohort. Given that the 
annuli are completed in the 2 nd semester, specimens 
caught in the 1 st semester with n annuli and an opaque 
zone on their otolith edge were assigned to the ra+1 age 
group, whereas specimens with the n th annulus on the 
otolith edge caught either in the 1 st semester or the 2 nd 
semester, were assigned to the n th age group. 
The following otolith readings were the most fre¬ 
quent: OP, HI, Hl+OP, H2, H2+OP, H3, H3+OP, H4 
and H4+OP, where OP and H correspond to opaque and 
hyaline edges respectively, and the number specifies 
the successive annulus. The clusters of the length dis¬ 
tributions, classified according to their otolith readings, 
are shown in Figure 4. It was observed that the clus¬ 
ters corresponded with specimens of those age groups 
obtained according to age interpretation criteria. 
Growth 
The mean lengths at age showed a significant differ¬ 
ence between sexes (t-test, n- 39, P<0.05). Table 2 sum¬ 
marizes the VBGF parameters for the three versions of 
the model (classic, sex specific, and seasonal) and the 
various subsets investigated (all specimens and only 
specimens with identified sex). The growth curve that 
best fitted the data was the sex-specific VBGF (mean 
squared error [M8E]=0.Q09, see Table 2). Based on 
this curve, asymptotic length for females and males 
was 16.52 and 20.29 cm TL respectively (Fig. 5A). The 
two runs of the classic VBGF for all fish (Fig. 5B) and 
for only fish with identified sex gave similar estimates 
of (all fish: 19.91 cm TL; for those with identified 
sex: 19.93 cm TL), similar K values (all fish: 0.512; for 
those with identified sex: 0.433) and quite different t 0 
values (all fish: -0.412; for those with identified sex: 
-0.843). Residual diagnostic plots and histograms are 
A 
I X I 
B 
Figure 4 
A cluster diagram of length-frequency distributions for 
specimens of picarel (Spicara smaris ) collected during 
1998-1999 in the Saronikos Gulf, Greece, classified 
according to results of otolith readings during the (A) 
first and (B) second semester. Clusters were formed ac¬ 
cording to the following criteria: no annulus (OP), first 
annulus on the edge (HI), first annulus that was com¬ 
pleted and was opaque on the edge (Hl+OP) and so on 
for H2, H2+OP, H3, H3+OP, H4, and H4+OP. 
