NOTt Somarakis et al.: Catchability and retention of larval Engraulis encrasicolus 
919 
Table 1 
Cruise data. Subregions A, B, C (northern Aegean Sea), and D (central-southern Aegean Sea) are indicated in Figure 1 . n = 
number of stations. n p = number of positive stations. D = daytime stations. N = nighttime stations. T = twilight stations. 
Cruise 
Date 
Subregion 
n 
n P 
D 
N 
T 
Mesh size 
(mm) 
Larval 
abundance/m 2 
92ANC1 
16 Jul-3 Aug 1992 
A-B-C 
117 
117 
54 
44 
19 
0.250, 0.500 
97.25 
92ANC2 
3-9 Sep 1992 
D 
80 
70 
30 
26 
14 
0.250, 0.500 
21.02 
93ANC3 
7-14 Jul 1993 
A-B-C 
110 
110 
62 
30 
18 
0.250, 0.500 
233.91 
94ANC4 
19-24 Jan 1994 
A-C 
46 
43 
16 
18 
9 
0.250, 0.335 
114.55 
95EP1 
15-30 Jun 1995 
A-B-C 
121 
117 
65 
28 
24 
0.250, 0.335 
69.17 
Total 
474 
457 
227 
146 
84 
when the posterior margin of the upper hypural plate 
is at 90° from the notochord axis (Moser, 1996). 
Additionally, to determine whether the inferences 
drawn regarding retention rates were plausible in 
terms of the “diagonal rule,” we plotted maximum 
head width values against standard length based on 
measurements of 74 staged anchovy larvae (for a 
justification of using head width, see Colton et al., 
1980). In postyolksac larvae, maximum head width 
was the maximum body width and was greater than 
any body-depth measurement (Somarakis, unpubl. 
data). This is not the case for yolksac larvae because 
of the bulk of the yolk sac. Yolk sacs, however, may 
be easily compressed or crushed, hence, we consid- 
ered head width to be the significant dimension. 
Correction for net avoidance and 
mortality estimates 
In studies on mortality of European anchovy the 
usual practice has been to correct catches for net 
avoidance by using daymight catch ratios and the 
methods available for northern anchovy (e.g. Palomera 
and Lleonart, 1989) because catchability of European 
anchovy has been unknown. As a final step of this study 
we estimated mortality during 1994 and 1995 cruises 
without correcting catches for net avoidance, as well 
as after correction using three different methods: 
Method 1 The correction for avoidance of the net 
during day-light was calculated by the 
sinusoidal function (Hewitt and Methot, 
1982): 
( 1 + DN L ) (1 -DN l ) (2nt\ 
f 1= — + — cos , 
2 2 l 24 J 
where DN L = the midday-to-midnight catch ratio of 
L-length larvae; and 
t = the hour of the tow. 
DN j data used were those available for E. mordax. 1 
Method 2 The same as method 1, DN r data used 
were those calculated for E.encrasicolus 
in the present study. 
Method 3 The sinusoidal function was not used. 
Length-specific ratios of daymight and 
twilightmight catches calculated in the 
present study were used to correct 
catches of day and twilight stations. 
The correction for duration of each size class 
(Hewitt and Methot, 1982; Lo et al., 1989) was cal- 
culated from growth of postyolksac larvae in the sea 
measured by daily increments in otoliths (see also 
Somarakis et al., 1997a). Age-( micro-increment 
count )-at-length data and respective growth curves 
were available from 336 postyolksac larvae collected 
during the 1994 cruise, and 294 postyolksac larvae 
from the 1995 cruise. Significant spatial or inter- 
annual differences in growth of the European an- 
chovy larvae in the Aegean Sea can be found, which 
cannot be explained by temperature differences 
(Somarakis et al., 1997a; Somarakis et al., 1997b). 
Thus, we estimated mortality for only the 1994 and 
1995 cruises for which otolith data were available. 
Mean corrected catches at length (i.e. length-spe- 
cific daily production of larvae per m 2 ) were calcu- 
lated based on the lognormal model (delta distribu- 
tion — see above). Mortality was estimated for larvae 
>4- mm and <10-mm. The 3-3.99 size class was not 
used because it includes both yolksac and feeding 
larvae (the yolksac stage is characterized by differ- 
ent mortality rates than the feeding stage [Lo, 1985a; 
Lo, 1985b]). Mean length-specific daily production 
of larvae per m 2 with its age (calculated from the 
1 Lo, N. C. H. 1996. Southwest Fisheries Center, National Ma- 
rine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 
92038. Personal commun. 
