NOTE Somarakis et a I.. Catchability and retention of larval Engraulis encrasicolus 
921 
Table 2 
Estimates of mean catch and their coefficients of variation (in parentheses), by length class. 
Length (mm) 
2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5 
0.500:0.250 
0.500 
0.419 
(0.079) 
0.378 
(0.067) 
0.250 
0.836 
(0.084) 
0.607 
(0.071) 
0.335:0.250 
0.335 
0.581 
(0.067) 
0.453 
(0.067) 
0.250 
0.614 
(0.068) 
0.498 
(0.076) 
day:night and twilight:night 
day 4.008 
(0.067) 
2.900 
(0.072) 
night 
4.002 
(0.085) 
2.775 
(0.087) 
twilight 
3.864 
(0.115) 
3.098 
(0.113) 
0.323 
0.264 
0.187 
0.136 
(0.062) 
(0.064) 
(0.072) 
(0.082) 
0.408 
0.329 
0.230 
0.139 
(0.069) 
(0.072) 
(0.080) 
(0.101) 
0.438 
0.325 
0.225 
0.182 
(0.068) 
(0.064) 
(0.077) 
(0.089) 
0.427 
0.332 
0.241 
0.184 
(0.068) 
(0.069) 
(0.074) 
(0.085) 
2.047 
1.582 
0.890 
0.587 
(0.075) 
(0.082) 
(0.093) 
(0.107) 
2.090 
1.512 
1.352 
1.003 
(0.088) 
(0.102) 
(0.101) 
(0.111) 
2.322 
1.683 
1.146 
0.748 
(0.114) 
(0.114) 
(0.129) 
(0.156) 
0.078 
0.054 
0.020 
0.012 
(0.115) 
(0.131) 
(0.200) 
(0.248) 
0.080 
0.056 
0.021 
0.012 
(0.128) 
(0.156) 
(0.233) 
(0.306) 
0.098 
0.062 
0.023 
0.013 
(0.109) 
(0.132) 
(0.201) 
(0.249) 
0.109 
0.082 
0.024 
0.015 
(0.110) 
(0.124) 
(0.232) 
(0.287) 
0.326 
0.201 
0.059 
0.043 
(0.132) 
(0.167) 
(0.297) 
(0.335) 
0.623 
0.496 
0.221 
0.099 
(0.129) 
(0.145) 
(0.191) 
(0.280) 
0.454 
0.319 
0.161 
0.101 
(0.186) 
(0.217) 
(0.319) 
(0.397) 
that catchability changes with varying light condi- 
tions and length. On the other hand, the “sudden” 
change in catchability at 6.5 mm coincides with the 
onset of the development of the tail, i.e. the flexion 
stage (Fig. 5). 
Correction for avoidance and 
its effect on mortality estimates 
Estimates of length-specific larval production for the 
1994 and 1995 cruises obtained by using the three 
different methods of adjusting data for net avoidance, 
or without any correction, were very similar (Table 
3). Method 1, which was based on length-specific 
day:night catch ratio values calculated for Engraulis 
mordax, gave slightly greater values for larvae 4-6 
mm, although differences were not statistically sig- 
nificant (overlapping confidence intervals). 
Consequently, analyses of the residual sum of 
squares showed that the resulting mortality curves 
(Table 4) were not statistically different (1994: F= 
0.361, P> 0.5; 1995: F=0.780, P> 0.5). 
1 1 1 1 1 r~ — i 1 1 1 1 
0 1 23456789 10 11 12 
Length (mm) 
Figure 3 
Relation between standard length and maximum head 
width for larval anchovy. The horizontal broken lines indi- 
cate the mesh diagonals of 0.500-, 0.335-, and 0.250-mm nets. 
Solid squares = yolksac larvae. Open squares = preflexion 
larvae. Solid circles = flexion larvae. Open circles = postflexion 
larvae. 
Discussion 
Retention of larvae 
The results of this study indicate that the “diagonal 
rule” (maximum head width being considered as the 
maximum cross-sectional diameter of the larva) may 
adequately “predict” retention of anchovy larvae in the 
0.500-mm mesh net but appears to be conservative in 
the case of the 0.335-mm mesh net. Specifically, yolksac 
larvae have a smaller head width than the diagonal of 
the 0.335-mm mesh net, but the latter seems to fully 
