917 
Catchability and retention of 
larval European anchovy, 
Engraulis encrasicolus, 
with bongo nets 
Stylianos Somarakis*' ** 
Barbara Catalano* 
Nikolaos Tsimenides*' ** 
* Institute of Marine Biology of Crete 
PO. Box 2214, 7 1 0 03 Iraklion, Crete, Greece 
** University of Crete, Department of Biology 
PO. Box 1470, 711 10 Iraklion, Crete, Greece 
E-mail address (for S. Somarakis): somarak@talos.cc.uch.gr 
Whenever data on larval abun- 
dance are used in producing esti- 
mates or indices of stock size or re- 
cruitment, accurate length distri- 
butions are required. Factors that 
might bias sample distributions 
must be taken into account and, in 
broadscale ichthyoplankton sur- 
veys, the effect of environmental 
and behavioral factors on the con- 
tent of the samples or collections 
must be considered (Zweifel and 
Smith, 1981). The objective is to 
standardize sampling gear by ap- 
plying correction factors to make 
samples comparable (Smith and 
Richardson, 1977). 
A factor contributing to a possi- 
bly serious source of bias in larval 
fish collections is net avoidance. 
Larvae may be agile and capable of 
avoiding nets, with the general ef- 
fect of inaccurate abundance and 
mortality estimates (Clutter and 
Anraku, 1968). Catchability varies 
mainly with light regime and lar- 
val length. The visual stimulus of 
the sampling device is believed to 
be of crucial importance. This is 
indicated by numerous investiga- 
tions showing diel variation in 
avoidance, catches being signifi- 
cantly smaller during daylight than 
at night (Morse, 1989, and refer- 
ences therein). 
A second factor affecting larval 
fish collections is extrusion of cap- 
tured material through the net 
mesh. According to the “diagonal 
rule” (Saville, 1958; Smith et al., 
1968), the maximum cross-sec- 
tional diameter of an organism 
must be greater than the mesh di- 
agonal if it is to be fully retained. 
However, the “diagonal rule” is of- 
ten too conservative (Lenarz, 1972; 
Colton et al., 1980). 
We examined the effects of time 
of day (day-night-twilight), fish 
length, and ontogeny on the catcha- 
bility of European anchovy (En- 
graulis encrasicolus), by using a 60- 
cm bongo net. We also investigated 
biases resulting from differential re- 
tention of larvae of different lengths, 
and, further, we examined the effect 
of correction of catch for net avoid- 
ance on estimation of mortality. 
Information on catchability and 
retention of anchovy with plankton 
nets is very limited and results are 
often contrasting. Catchability of 
European anchovy with bongo nets 
is unknown. Regarding retention, 
there is just a single study indicat- 
ing full retention of European an- 
chovy larvae with the 0.333-mm net 
(Aldebert et al., 1975), whose data 
are contradictory to Lo ( 1983) who 
has estimated a 0.63 retention rate 
for northern anchovy, Engraulis 
mordax, larvae <4.0-mm with a 
0.333-mm net. 
Materials and methods 
From 1992 through 1995, five 
ichthyoplankton surveys were car- 
ried out in the Aegean Sea. A total 
of 474 stations were occupied in 
continental shelf and slope waters 
and areas characteristic of larval 
anchovy habitat (Fig. 1). Plankton 
and hydrographic sampling were 
performed at each station and cruise 
information is given in Table 1. Most 
stations were positive (i.e. at least 
one larva was captured). 
A 60-cm bongo net sampler (Hy- 
drobios) was used during all cruises. 
Mesh sizes on the sampler were 
0.250-mm, and 0.500- or 0.335-mm, 
depending on the cruise (Table 1). 
The 0.250-mm mesh net is consid- 
ered to retain clupeoid eggs and 
larvae completely (Aldebert et al., 
1975; Colton et al., 1980; Leslie and 
Timmins, 1989). 
Tows at two knots were double- 
oblique, within 5 m of the bottom 
to the surface, or from 120 m depth 
to the surface at deep stations. All 
tows consisted of a wire released to 
the desired depth and retrieved to 
the surface at standard speeds. The 
depth of the sampler could be moni- 
tored onboard at any time during 
the tow by means of a recording 
depthmeter attached to the sam- 
pler. Plastic codend buckets, with 
side windows of 155 cur covered 
with net gauze, were used in an ef- 
fort to minimize damage to larvae. 
Volumes filtered were calculated 
from a calibrated flowmeter in the 
mouth of each net. 
All samples were sorted in the 
laboratory and larvae were identi- 
fied to the lowest possible taxo- 
nomic level. Larval anchovies were 
counted and measured (notochord 
Manuscript accepted 14 January 1998. 
Fish. Bull. 96: 917-925 ( 1998). ' 
