438 
Fishery Bulletin 99(3) 
< 
o o o 
o o o 
O 00 CD 

400 --- 
- - - Purse seine fit 
Gill net fit 
200 % 
X Purse seine observations 
Gill net observations 
1000 
800 
600 
400 
200 
0 
B 
12 
1000 
800 
600 
400 
200 
0 
0 
c 
2 4 6 8 10 
Age (years) 
12 
Figure 5 
Growth curves for bluefish caught by purse seines and gill nets calculated by 
(A) method 1 (mean lengths — last annulus only) (B) method 2 (mean lengths — 
all annuli) and (C) method 3 (individual lengths — last annulus only). 
tempt to fit the von Bertalanffy growth curve by method 4 
for the gill net failed because no feasible solution could be 
found for the parameters of the curve. 
For all methods there were significant differences in 
growth parameter estimates between the gill-net and the 
purse-seine samples (Hotelling’s t-test, P<0.05). 
The comparison between mean back-calculated (espe- 
cially criterion Ha) and mean observed length-at-age 
showed that the latter is systematically larger than the 
former (Table 3). This difference was expected because 
the mean back-calculated lengths-at-age were obtained 
for length at time of annulus formation (where spawning 
season and time of annulus formation are considered uni- 
form for all fish). The differences between both means 
may be attributed to growth after the growth mark for- 
mation. Moreover, the mean observed length-at-age cal- 
culated from sampled fish was obtained from individuals 
that may have spawned at different times of the year 
or whose ring formation may have been distinctly visible 
in time, resulting in fish with distinct real ages. Hence, 
growth parameters obtained from mean observed lengths- 
at-age of purse-seine catches during the fishing season 
(L„= 985 mm, £=0.12 and t 0 = 2.17) were significantly dif- 
ferent from the growth parameters derived from back- 
calculated lengths-at-age (also from purse-seine catches) 
(Hotelling’s t-test, PcO.001 for all methods). 
