156 
Fishery Bulletin 117(3) 
TL (cm) 
700 n B 
600 H 
500 
>, 400 H 
! 
“■ 200 
100 
0 
-Tk. 
I 
III 
□ Female, n =882 
■ Male, n =2920 
Sex not determined, n=14 
jLi-ik 
,I~L i.. 
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 
TL (cm) 
Figure 2 
Length-frequency distribution of rainbow smelt (Osmerus mordax) caught 
(A) during a trawl survey off the coast of Massachusetts in 2004-2007 and 
(B) at a fyke net station in the Fore River in 2005-2007. Samples with sex not 
determined were immature rainbow smelt that had gonadal development that 
was insufficient for identifying sex. TL=total length. 
average water temperature of 3.6°C at the surface (range: 
0.7-8.9°C) and 3.5°C (range: 0.7-7.2°C) on the bottom. 
The highest catch rates for rainbow smelt occurred in 
Cape Cod Bay and Massachusetts Bay at an average 
depth of 26 m (range: 19-30 m). Of the total sample, 255 
rainbow smelt were male (3 immature), 319 rainbow 
smelt were female (1 immature), and the sex could not be 
determined for 60 fish (all immature) (Fig. 2A). All sam¬ 
ples used in fecundity analysis (n=110) were aged (except 
3 fish for which the scale samples had degraded). Sex, TL, 
and TW were recorded for all captured rainbow smelt, 
except for fish caught during 3 tows of the trawl in 2007, 
when females in unfilled TL bins for fecundity analysis 
were collected and all other rainbow smelt had their TL 
measured and sex determined. 
Fecundity Mean total fecundity in rainbow smelt 
increased 270% from age 1 to age 2 and 170% from age 
2 to age 3 (Table 2). The highest variability in fecun¬ 
dity was observed for age-2 rainbow smelt (Fig. 3A) as 
a consequence of the wide range of size represented 
for this age (Fig. 2A). Mean and maximum total fecun¬ 
dity of age-3 and age-4 fish were nearly the same, and 
rainbow smelt of both ages had much higher minimum 
fecundity than fish in the younger age classes (Fig. 3A). 
In regression analysis, total fecundity of rainbow smelt 
increased significantly with each measure of size and 
age (Table 3). The best-fitting relationship was that 
of fecundity and TW, closely followed by that of fecun¬ 
dity and TL (Table 3, Fig. 4). Although the regression of 
fecundity versus age was significant, the occurrence of 
