Lefebvre et al.: Reproductive ecology and size-dependent fecundity in Eopsetta jordani 
297 
12 % - 
10 % - 
8 % - 
6 % - 
4% - 
2 % - 
0 % 111 I 111 Mllll 
o 
O 
O 
O 
o 
O 
o 
o 
o 
o 
o 
o 
o 
o 
o 
CO 
CO 
CM 
O 
CO 
CO 
■'sf 
CM 
o 
CO 
CO 
CM 
CM 
CO 
LO 
CO 
CO 
h- 
CO 
(J) 
o 
o 
1— 
CM 
Oocyte diameter (pm) 
Figure 2 
Representative examples of oocyte size-frequency distri¬ 
bution diagrams for ovarian samples taken from petrale 
sole (Eposetta jordani) collected off California, Oregon, 
and Washington in 2014-2017. Primary growth oocytes, 
which were <150 um, are not included, and all oocytes 
are at the primary vitellogenic oocyte stage or further 
advanced. (A) An example of when recruitment of oocytes 
for the year was nearing completion, from a macroscop¬ 
ic-stage-2 ovary sampled in August. (B) An example of 
oocytes fully recruited before spawning has commenced, 
from a macroscopic-stage-2 ovary sampled in October. 
(C) An example of a female that had initiated spawning, 
from a macroscopic-stage-3 ovary sampled in October. 
The gap between primary and secondary growth oocytes 
evident in panels B and C indicates a determinate fecun¬ 
dity pattern for petrale sole. 
Histological examination of ovarian tissues confirmed 
macroscopic assignment of developmental stages (Fig. 4) 
and provided evidence of the fecundity pattern and spawn¬ 
ing strategy of petrale sole. Increasing size of perinucleolar 
stage oocytes signaled the earliest development. Interest¬ 
ingly, there appeared to be no distinct cortical alveolar 
oocyte stage, with cortical alveoli appearing concomitantly 
with vitellogenin (yolk protein). Oocytes developed some¬ 
what asynchronously until a mid-vitellogenic state (Vtg2) 
because, presumably, recruitment from primary growth 
reserves was ongoing. Oocytes then developed synchro¬ 
nously to the tertiary vitellogenic stage (Vtg3), after which 
individual clutches underwent final maturation sequen¬ 
tially (observed histologically as ovaries with primary 
growth and Vtg3 oocytes with groups of oocytes at various 
stages of final maturation [e.g., germinal vesicle migra¬ 
tion, germinal vesicle membrane breakdown, or hydra¬ 
tion]). Once the leading cohort reached the Vtg3 stage, 
earlier vitellogenic stages were not observed (indicating 
recruitment from the primary growth stage was complete). 
This oogenesis pattern indicates that petrale sole exhib¬ 
ited a determinate fecundity pattern. The fecundity pat¬ 
tern was corroborated with examination of diagrams of 
oocyte size-frequency distribution that show a distinct gap 
between primary and secondary growth oocytes (Fig. 2). 
The presence of POFs (from prior spawning events) with 
hydrating oocytes (destined for imminent spawning) and 
late vitellogenic oocytes (for future spawning activity) 
demonstrated that petrale sole exhibited batch spawning 
behavior (Fig. 4D) (Murua and Saborido-Rey, 2003). The 
exact number of spawning events could not be determined 
because the duration of POFs were unknown in this spe¬ 
cies; however, 2 or 3 distinct stages of POFs were found in 
many histological sections from ovaries with both vitello¬ 
genic and hydrating oocytes, indicating that at least 4 or 
5 spawning events for an individual were common and that 
more were possible. Atresia of secondary growth oocytes 
appeared to be rare during the reproductive season, indi¬ 
cating minimal fecundity down regulation. A small per¬ 
centage (14%, or 1 of 7) of ovaries in the regressing phase 
appeared to be resorbing the final batch of oocytes at the 
end of the season. 
Fecundity models including the interaction term of 
length or weight and region were the best fit on the basis 
of AIC (Table 4). Total PAF of the fish examined in this 
study ranged from 458,000 to over 3 million eggs and was 
significantly related to maternal size (Table 5, Fig. 5). Rel¬ 
ative PAF values ranged from 481 to 1639 eggs per gram 
of somatic weight and was significantly related to mater¬ 
nal size in the combined and Pacific Northwest regional 
models (Table 5, Fig. 5); however, the slope parameter in 
the California regional model was not significantly dif¬ 
ferent from zero (Table 4). The difference in the regional 
slopes for relative PAF means that a 56-cm-TL fish pro¬ 
duced 1.25 times as many eggs per gram as a 40-cm-TL 
fish in California but produced 2.37 times as many eggs 
per gram as a 40-cm-TL fish off Washington and Oregon. 
This regional difference in the fecundity relationship 
appears to be driven by lower relative PAF for smaller 
