356 
Fishery Bulletin 11 6(3-4) 
TL (cm) 
Age (months) 
Figure 7 
Fitted probability ( P ) of picarel (Spicara smaris) being male with (A) total 
length and (B) age and accompanying proportions of males per 0.5-cm and 
1-month classes, respectively. The length at median (50%) female-to-male 
sex change was 15.34 cm in total length (TL). The age at median female- 
to-male sex change was 36.4 months. The solid line indicates the fitted 
curve, the dotted lines indicate the 95% confidence intervals, and the gray 
circles indicate the observed data. 
sex change. The presence of juveniles (see also Sup- 
pi. Fig. 1), as well as the high temperature (19-25°C) 
which usually appears during this period (Suppl. Fig. 
7), may play a role in the observed decrease in growth 
rate. It is noted however that Tsangridis and Filippou- 
sis (1994), who have tried a series of models for study¬ 
ing picarel growth, reported that the model that fitted 
best, showed that picarel growth changes when sex 
changes and the plotted curve also shows some kind 
of oscillation. 
According to the literature (Salekhova, 1979; Tsan¬ 
gridis and Filippousis, 1992), sex change occurs after 
spawning, which is common in protogynous species 
(Provost et ah, 2017). In the present study, the major¬ 
ity of spent females were observed from June through 
September (see Suppl. Fig. 4). A change in sex dur¬ 
ing nonbreeding periods is not surprising because this 
strategy is predicted to maximize seasonal reproductive 
output in individual fish as mentioned by Provost et al. 
(2017). In an attempt to explain this decrease of aver¬ 
age length, we believe that during sex transition, the 
longest females at each age (1, 2, and 3 years) change 
into males. Considering that the mean lengths at age 
of males are greater than those of females, this inver¬ 
sion results in lower mean lengths at age for each sex. 
The selection of larger females for inversion into males 
