688 
Fishery Bulletin 95(4), 1 997 
100 
80 
60 
40 
20 
£ 
2 
CO 
S' 
Figure 8 
Relative seasonal changes in five environmental variables in the study 
area (see text for definitions): water temperature, feeding hours, 
daylength, rainfall, and calm surf. Note different y-axes. 
ing to an annual loss of about 20%. Simi- 
lar values were obtained with Hoenig’s 
equation (Z=0.23) and Beverton and 
Holts’ relation between Z and mean 
length of fish in the catch (Z=0.19, 
K= 0.12, 1^=22.! cm, L c =19.6, L'=18 cm). 
Natural mortality (M) was estimated 
to be 0.2 and 0.45 with the empirical 
equations of Longhurst and Pauly ( 1987) 
and Pauly (1980), respectively, and with 
the adult values of K and L m . The lower 
value indicated that M is equivalent to 
Z; the latter value was spurious given 
that it exceeded Z. 
Biological characteristics of 
harvest 
The flattened growth curve exhibited by 
older A. lineatus limited analyses based 
on length-converted ages, but the conver- 
sion did indicate that most fish taken in 
the fishery were relatively young (Fig. 
14). During the 9-year period 1987-95, annual 
catches varied moderately in mean age (3. 6-6. 2 
years), mean length (17.5-19.3 cm FL), proportions 
of immature fish taken (29-60%), and total mortal- 
ity (0.16-0.31 XTable 2). No trends in these variables 
were apparent, with one exception: the maximum size 
of fish decreased. Maximum sizes of fish in 1987-88, 
however, seem unrealistically high (Fig. 14), and in 
any case, mean sizes in later years (1994-95) were 
significantly larger than in earlier years (1987- 
88)(<=9.7, df=5088, PcO.001). 
Although the more detailed 1994-95 market data 
spanned a relatively short period (Fig. 15), available 
data indicated no decrease in fish size (r=0.28, df=17, 
P>0.1) or CPUE over time (r=0.41, df=17, P=0.085), 
and no relation between fish size and CPUE at vari- 
ous island-wide fishing sites, i.e. sites with low CPUE 
did not have smaller fish (r=0.28, df=17 P>0.1). 
Figure 10 
Age-length relations for A. lineatus. The solid line indi- 
cates a two-phase von Bertalanffy curve for ages 0-3 and 
4-18; the dashed line indicates the growth of naturally 
marked fish based on visual estimates of fish size over time. 
