Albacore— Otsu 
261 
Fig. 4. Net movements of albacore tagged by POFI. 
(The lines have no significance other than connecting 
the points of release and recovery of each fish. ) 
The parameter a, or the upper asymptote of 
the growth curve (118 cm.), had already been 
estimated by Walford’s method. These param- 
eters were fitted into the Gompertz equation 
as follows: 
y = (118) (0.177 ) (1 - 43) ~ x+(l 
which resulted in the asymmetric sigmoid curve 
shown in Figure 7. Inspection showed that if 
this curve is shifted 0.23 year to the right, it 
will present an adequate fit for the data; thus, 
an estimate of d is 0.23. 
The curve derived by the Walford method 
(Fig. 6) is superimposed on the Gompertz 
curve. There is fair agreement between the two 
in the portions representing lengths between 
40 and 85 cm., and the observed data do not 
deviate much from either curve. If albacore 
growth follows a typical Gompertz curve, then 
this represents a possible entire growth curve 
of the albacore, including that portion not pos- 
sible to determine by the Walford method. It 
should now be possible to assign specific ages. 
It can be seen by inspection that the curve 
approaches the abscissa reasonably closely at N 
— 2 years, suggesting that hatching is in this 
time vicinity. While the size indicated at hatch- 
ing deviates from the expected size to the order 
of 2 cm., this error is not large considering the 
assumedly random deviations of the observed 
data from the estimated growth curve and the 
small sample size. With N — 2 as the arbitrary 
origin, the adjusted ages are given below the 
original designations. The resulting length for 
each year beginning with this origin is pre- 
sented in Table 3 along with values obtained 
by the Walford method. While there is every 
indication that at least a portion of each curve 
reliably represents the albacore growth, those 
portions for which there are no observed data 
should be considered tentative. The initial 
growth by the Gompertz curve appears to be 
unreasonably slow since it requires about 3 years 
for a fish to attain a weight of 1 pound (about 
30 cm.). The early growth is slow, relative to 
the result obtained by the Walford method. The 
results suggest a possible error in assigned ages 
of 1 or 2 years. 
This disagreement and others may result from 
one or a combination of the following debilities: 
(a) the assumption that albacore growth fol- 
lows the Gompertz curve may be unwarranted; 
(b) the number of observations may be in- 
adequate; (c) there is the possibility that there 
is differential growth between the sexes after 
the onset of sexual maturity, causing an un- 
balanced sex ratio in favor of males among the 
larger albacore ( Otsu and Uchida, 1959 b) . 
If this is true, then the growth curves are sub- 
ject to inaccuracies in the upper portions rep- 
resenting fish larger than 90 cm., the approxi- 
mate size at which the albacore is believed to 
attain sexual maturity. A differential growth 
rate may have been responsible for the slightly 
Fig. 5. Growth shown by tagged fish plotted by Wal- 
ford’s (1946) method and fitted with a straight line. 
The intersection of this line and the line Ln = L» + 1 
indicates the upper asymptote of the albacore’s length. 
Fish No. 2 was excluded from calculation of the re- 
gression line. 
