NOTE Beckmann et ai : Reproductive biology, growth, and natural mortality of Sebastes emphaeus 
353 
Table 1 
Date and sampling method for collections of Puget Sound rockfish older than age 0. 
Date 
Number 
collected 
Method 
Date 
Number 
collected 
Method 
26 Jun 1994 
6 
spear 
3 Aug 1994 
41 
hook and line 
28 Jun 1994 
4 
spear 
8 Aug 1994 
19 
hook and line 
29 Jun 1994 
12 
spear or hand net 
21 Aug 1994 
11 
spear or hook and line 
5 Jul 1994 
5 
hook and line 
26 Aug 1994 
8 
hook and line 
6 Jul 1994 
22 
spear 
29 Aug 1994 
8 
hook and line 
7 Jul 1994 
11 
spear 
30 Aug 1994 
10 
hook and line 
7 Jul 1994 
20 
hook and line 
2 Sep 1994 
21 
hook and line 
13 Jul 1994 
35 
hook and line 
6 Sep 1994 
7 
hook and line 
22 Jul 1994 
15 
hook and line 
10 Sep 1994 
13 
hook and line 
24 Jul 1994 
3 
hand net 
30 Dec 1994 
21 
hook and line 
26 Jul 1994 
6 
hand net 
12 Mar 1995 
37 
hook and line 
28 Jul 1994 
27 
hook and line 
act year) between independent readings, with no 
obvious bias. 
Age-length relationships were estimated from the 
mean length at age for age groups 1-13, derived from 
the otolith readings. The mean length of age-0 fish 
was estimated from the mean length of the YOY cap- 
tured with nets during February- August 1994. 
Lengths for YOY were originally recorded as stan- 
dard length, and were converted to fork length by 
using the relationship for shortbelly rockfish (S. 
jordani) reported by Echeverria and Lenarz (1984). 
The mean age of YOY (0.7 years) was estimated by 
assuming that parturition occurred on 1 September. 
Mean length at age for age groups 1-13 was esti- 
mated from fish collected during June-December, 
1994. Data were fitted to the von Bertalanfly growth 
model (Ricker, 1975) with nonlinear least squares re- 
gression. Male and female fish were treated separately 
owing to the difference in sex-specific growth rates. 
Mortality was estimated from the age composition 
of catches (281 fish aged successfully, sexes combined) 
made during June-December. Because Puget Sound 
rockfish are not the target of any fishery, the catch 
curve (Ricker, 1975) was used to estimate natural 
mortality (M) with linear regression of log (frequency) 
on age. Only fish 3 years and older were used to esti- 
mate M, because it was apparent (Fig. 1) that 1- and 
2-year-olds were not fully available to our sampling 
gear. 
Fork length (FL) to the nearest millimeter and wet 
gonad weight (GW) and wet somatic weight (SW) to 
the nearest milligram were measured for all adult 
females. Somatic weight was defined as the weight 
of the fish with the stomach emptied and the gonads 
removed. The gonadosomatic index ( GSI = GW/SW ) 
was used to determine relative reproductive effort 
for a mature female of average size. Allometric rela- 
tionships between GW-body length and SW-body 
length were determined by using log-log linear re- 
gression (Gunderson and Dygert, 1988). Only fully 
mature females in the later stages of vitellogenesis 
were used in estimating GSI. Histological examina- 
tion of a subsample of 12 individuals, representing a 
broad range of lengths and relative GSI values, showed 
that all of them were in the migratory nucleus stage of 
egg development just prior to ovulation (McDermott 1 ). 
Fecundity was estimated by counting all nonatretic 
oocytes in advanced vitellogenesis with subsamples 
from both ovaries of 16 fish ranging from 112 to 178 
mm FL. From preliminary measurements it could 
be seen that vitellogenetic oocytes were at least twice 
the size of the resting and immature primary oocytes. 
All oocytes with the longest axis exceeding 250 pm 
were counted. A gravimetric subsampling method 
(Nichol and Pikitch, 1994) was chosen to estimate 
oocyte numbers. 
Because of the nonuniform distribution of oocytes 
and connective tissue within the ovaries, ovarian sacs 
and ovarian stroma were removed prior to subsamp- 
ling. Thin sections of the remaining egg mass were 
taken from five different transects equally spaced 
along the longitudinal axis of each ovary. Each 
subsample was weighed, pipetted into a plastic grid 
dish, and the eggs of each subsample were counted. 
Final fecundity estimates were made by multiplying 
1 McDermott, S. 1995. School of Fisheries, Univ. Washington, 
Seattle, WA 98195. Personal commun. 
