Chase et al.: Reproductive life history of anadromous Osmerus mordax in Massachusetts 
157 
Table 2 
Summary statistics for rainbow smelt (Osmerus mordax) sampled during a trawl survey in marine waters off Massa¬ 
chusetts in 2004-2007 and for analysis of their fecundity. Data provided are sample size (n), mean length, measured in 
millimeters in total length (TL), mean total weight (TW), mean gonadosomatic index (GSI), mean ovarian weight (OW), 
mean egg density, relative fecundity, measured as the number of oocytes divided by body weight (BW), and fecundity, 
measured as the number of mature oocytes prior to spawning. 
Age 
n 
Mean 
length 
(mm TL) 
Mean 
TW (g) 
Mean 
GSI 
Mean 
OW (g) 
Mean egg 
density (no. 
oocytes/OW) 
Mean relative 
fecundity (no. 
oocytes/BW) 
Fecundity 
(no. of 
oocytes) 
Mean 
fecundity 
(no. of oocytes) 
1 
19 
137 
12.93 
6.53 
0.80 
13,234 
767 
4880-14,102 
9026 
2 
76 
184 
36.79 
11.23 
3.96 
8529 
747 
7885-50,001 
24,624 
3 
8 
215 
65.45 
15.10 
8.65 
6118 
789 
25,559-51,651 
42,957 
4 
4 
231 
88.34 
25.39 
17.42 
2652 
633 
36,721-47,328 
43,927 
50 
_ 40 
w 
o 
o 
o 
1 o 
c 
§20 
LI- 
10 
1 
2 3 
Age 
4 
B • 
1200 
.S’ 
=6 1000 
c 
3 
O 
0) 
JS 
<13 
DC 
800 
600 
400- 
12 3 4 
Age 
Figure 3 
Boxplots (A) of total fecundity, measured as the number 
of mature oocytes prior to spawning, and (B) of relative 
fecundity, measured as the number of oocytes divided by 
body weight, by age class, for rainbow smelt (Osmerus mor¬ 
dax) sampled during a trawl survey in marine waters off 
Massachusetts in 2004-2007. The lower and upper hinges 
of the box plots correspond to the 25 th and 75 th percentiles. 
The whiskers extend from the hinge to the farthest value 
no more than 1.5 times the interquartile range (distance 
between the 25 th and 75 th percentiles). Data beyond the 
end of the whiskers, called outlying points, are plotted 
individually. 
only 4 age classes and low sample sizes for age-3 and 
age-4 rainbow smelt resulted in the lowest coefficient of 
multiple determination CR 2 ) among predictor variables 
(Table 3). Mean oocyte density (number of eggs per gram) 
declined sharply with age (Table 2). Mean oocyte den¬ 
sity was 5 times higher in age-1 rainbow smelt than in 
age-4 rainbow smelt. Overall, relative fecundity was sim¬ 
ilar among age classes (Fig. 3B) with declines in oocyte 
density with age compensated by increases in body and 
ovary size. 
Maturity ogive A multiyear maturity ogive for both sexes 
combined was estimated from the sample of 634 rainbow 
smelt collected with a trawl in marine waters, with 64 imma¬ 
ture fish included (Fig. 5). The GLM was significant for the 
intercept and TL (P<0.001) and estimated L 50 at 126.2 mm 
TL. Maturity ogive estimates were not possible for the sam¬ 
ples collected in coastal rivers with a fyke net during spawn¬ 
ing runs because of the near absence of immature rainbow 
smelt. However, mean length of age-1 rainbow smelt (range: 
140-147 mm TL) was greater at the locations of fyke net 
sampling than the L 50 , indicating that not all age-1 rainbow 
smelt are recruited to spawning runs (Table 4). Our data are 
not sufficient to produce annual ogives or separate ogives for 
each sex. A slight difference in maturity ogive between sexes 
may occur given the differences in mean length. 
Oocyte diameter The diameters of oocytes from the ova¬ 
ries of a subsample of 4 rainbow smelt were measured 
before samples were fixed in formalin to determine if 
oocyte diameter and maturity differed among locations 
in the ovary where oocytes were removed. There was no 
variation in oocyte appearance or stage among samples: 
all had dense, dark yellow yolk, consistent with maturing 
oocytes of later vitellogenic stages (Crim and Glebe, 1990). 
No significant differences were found among the 4 subsa¬ 
mples of rainbow smelt by using the Kruskal-Wallis rank 
sum test fy 2 =3.91, 5.78, 11.07, and 6.72; degrees of free¬ 
dom (df)=6; P>0.05). No further oocyte diameter measure¬ 
ments were made given the confirmation of the reports by 
Chen (1970) and Clayton (1976) that oocyte maturity of 
rainbow smelt was evenly distributed along ovaries and 
the subsequent finding that fixing the remaining samples 
in 3% formalin reduced oocyte size and distorted shape. 
Size and age analyses 
Fyke net samples During spawning runs in the spring 
of 2005-2007, 13,428 rainbow smelt were caught with 
fyke nets at locations in 4 coastal rivers (Table 5). Rain¬ 
bow smelt were caught routinely with relatively high fre¬ 
quency of occurrence in hauls at each location. The fyke 
