296 
Fishery Bulletin 11 7(4) 
Table 3 
Modifications specific to petrale sole (Eopsetta jordani) for oocyte stages described by Wallace and Selman (1981) and Low- 
erre-Barbieri et al. (2011). Oocytes were measured from prepared histological slides, except for hydrated oocytes, which were 
measured from whole mount samples. n=number of samples measured. 
Oocyte stage 
Petrale sole modification 
Size (pm) 
n 
Primary growth 
Include oogonia and perinuclear oocytes. 
<150 
150 
Cortical alveolar 
A distinct cortical alveolar stage is rare in petrale sole: cortical 
alveoli mostly appear concurrently with vitellogenin. 
140-180 
7 
Primary 
vitellogenic (Vtgl) 
Initial accumulation of small yolk granules occurs peripherally and 
mostly concurrently with cortical alveoli. Little to no oil droplet 
accumulation. 
140-325 
81 
Secondary 
vitellogenic (Vtg2) 
Small yolk granules completely fill cellular space. Little to no oil 
droplet accumulation. 
250-650 
234 
Tertiary 
vitellogenic (Vtg3) 
Oil droplets begin to appear and increase in size and numbers 
throughout this stage. 
450-775 
121 
Germinal vesicle 
migration 
Migration of the nucleus occurs concurrently with initial yolk 
coalescence (larger yolk granules). 
625-775 
27 
Germinal vesicle 
breakdown 
The nuclear membrane breaks down as yolk continues to coalesce, 
forming pale-pink “plates” within the oocyte. Oil begins to coalesce 
into a single droplet. 
625-850 
12 
Hydrated 
Yolk and lipids are completely coalesced. 
>1200 
20 
with length and weight and is best described as a power 
function: 
PAF = aL h or (1) 
PAF = aW h , (2) 
where L is the total length (in millimeters) and W is the 
somatic weight (in grams). The parameters a and b for 
each equation are the intercept and slope, respectively, of 
the linear natural-log-transformed least-squares regres¬ 
sion fit to the data: 
log(PAF) = log(a) + (b x log(D) or (3) 
log(PAF) = log(a) + (b x log(Wj). (4) 
The value for a was reported after back-transformation 
by using the bias correction term exp(o 2 /2). Values for P 
and the coefficient of determination (r 2 ) were reported for 
the log-transformed equation. Relative PAF (PAF rel ) was 
described as a linear function of female L or W: 
PAF rel = c + (d x L) or (5) 
PAF re] - c + (d x W), (6) 
where c and d are the intercept and slope, respectively, of 
the least-squares regression fit to the data. To understand 
potential geographic effects, we compared the performance 
of models relating fecundity (PAF or relative PAF) to mater¬ 
nal size in the following ways: 1) without considering region 
(California or Pacific Northwest) as a covariate, 2) allowing 
for region-specific intercepts along with a shared effect of 
size (i.e., classic analysis of covariance), or 3) allowing for 
region-specific effects of size (i.e., modeling an interaction 
between region and maternal size). We used the difference 
between Akaike information criterion (AIC) values for 
2 nested models (AAIC) to select the most parsimonious 
models for each combination of size and fecundity predictors. 
Although in all cases the use of AAIC resulted in selection of 
the model with an interaction between region and size as the 
best description of the data, we also present results for the 
size-only model because the use of this model could be more 
appropriate for stock assessments that do not distinguish 
among regions. 
All statistical analyses were conducted by using R (vers. 
3.4.3; R Core Team, 2017) with a significance level of 0.05. 
Results 
Female petrale sole collected off California ranged in size 
from 305 to 600 mm TL (mean: 472 mm TL [standard devi¬ 
ation (SD) 48]) and in age from 4 to 20 years (mean: 9.4 
years [SD 3.3]) (Table 1). Fish collected off Washington 
and Oregon measured 280-520 mm TL (mean: 436 mm 
TL [SD 63]) and had ages from 3 to 15 years (mean: 7.7 
years [SD 2.8]) (Table 1). All macroscopic stages of ovar¬ 
ian development were observed in fish collected from 
California (Fig. 3); the only immature fish were collected 
in NWFSC bottom-trawl surveys. At the earliest collec¬ 
tions, the majority of females had stage-2 ovaries. Spawn¬ 
ing was first evident in September, and peak spawning 
occurred in December, as determined by the percentage 
of females with stage-3 ovaries. By January, the majority 
of females had completed spawning and had stage-4 ova¬ 
ries. The reproductive season appeared to start later in 
Pacific Northwest collections, with the majority of ovaries 
found at stage 2 in December and the first stage-3 ovaries 
collected in January. 
