924 
Fishery Bulletin 96(4), 1 998 
and Coyne, 1982) have shown that, at about 11 mm, 
swimbladder inflation and schooling begins, where- 
upon patchiness increases rapidly. Our field obser- 
vations of European anchovy indicate initiation of 
swimbladder inflation at flexion. It is therefore pos- 
sible that schooling behavior also begins during this 
stage in E. enci'asicolus. 
School formation during the day and dispersion of 
schools during the night could result in low daymight 
catches for two main reasons: 1) larvae are less vul- 
nerable to plankton nets when in a school than when 
reacting individually, owing to the reduction in the 
reaction distance of the organisms in the school to 
the approaching net; and 2) increased patchiness 
resulting from schooling during the day produces a 
larger number of negative tows. 
In summary, field data on the European anchovy 
suggest the existence of an ontogenetic change in 
catchability with slowly towed plankton nets, which 
can be attributed to a respective change in the swim- 
ming ability, or a change in behavior (i.e. onset of 
schooling), or both. 
To our knowledge, the present study is the first to 
examine retention and catchability in a larval fish 
not only in terms of length but also of ontogeny. Sur- 
prisingly, this study also shows that, when sampling 
is undertaken on a 24-h basis, bias that results from 
light-induced avoidance of the net may not be great 
enough to affect significantly estimates of daily lar- 
val production or larval mortality. However, when 
only daytime sampling is undertaken, mean catches 
of postflexion larvae may well be biased and correc- 
tion factors may have to be applied. 
Size selectivity due to net avoidance by larvae is 
expected to lead to an overestimation of mortality rate, 
because older larvae are underrepresented in relation 
to younger larvae. In a recent simulation study, 
Somerton and Kobayashi (1992) have shown that 
approaches usually taken to eliminate the selection 
bias from sampled length frequencies (i.e. division 
of sampled length frequencies by length-specific es- 
timates of capture probability or elimination of the 
biased portion of the length distribution) may be only 
partially effective in reducing the bias in estimated 
mortality rates. The latter, as well as the present 
paper, suggests that the effectiveness of methods that 
attempt to correct for net avoidance is largely un- 
known. Their effectiveness has to be fully addressed. 
Acknowledgments 
This study was partially funded by an EU Study 
Project DG XIV (MED/91/011). We gratefully ac- 
knowledge the useful comments of three anonymous 
referees which substantially improved an initial ver- 
sion of this manuscript. We also thank. B. Nafpaktitis 
for his valued help and discussions and N. Lo for 
providing daymight ratios for Engraulis mordax. 
Literature cited 
Aldebert, Y., A. J. Dicenta, Y. Marinaro, and C. Piccinetti. 
1975 . Engins de peches pour F ichthyoplancton: essais 
comparatifs. Rev. Trav. Peches Marit. 39 (3):261— 277. 
Batty, R. S. 
1984 . Development of the swimming movements and mus- 
culature of larval herring (Clupea harengus). J. Exp. Biol. 
110:217-229. 
Chen, Y., D. A. Jackson, and H. H. Harvey. 
1992 . A comparison of von Bertalanffy and polynomial func- 
tions in modeling fish growth data. Can. J. Fish. Aquat. 
Sci. 49:1228-1235. 
Clarke, T. A. 
1989 . Seasonal differences in spawning, egg size, and early 
development time of the Hawaiian anchovy or nehu, 
Encrasicholina purpurea. Fish Bull. 87:593-600. 
Clutter, R.I., and M. Anraku. 
1968 . Avoidance of samplers. In D. J. Tranter (ed. ), Zoop- 
lankton sampling, p. 57-76. UNESCO Monogr. Oceanogr. 
Methodol., vol. 2. 
Colton J. B., Jr., J. R. Green, R. R. Byron, and 
J .L. Frisella. 
1980 . Bongo net retention rates as affected by towing speed 
and mesh size. Can. J. Fish. Aquat. Sci. 37:606-623. 
Heath, M., and J. Dunn. 
1990 . Avoidance of a midwater frame trawl by herring 
larvae. J. Cons. Int. Explor. Mer 47:140-147. 
Hewitt, R. 
1981 . The value of pattern in the distribution of young fish. 
Rapp. P.-V. Reun. Cons. Int. Explor. Mer 178:229-236. 
Hewitt R. P., and R. D. Methot. 
1982 . Distribution and mortality of northern anchovy lar- 
vae in 1978 and 1979. CalCOFI Rep. 23:226-245. 
Houde, E. D. 
1977 . Abundance and potential yield of the round herring 
Etrumeus teres and aspects of its life history in the east- 
ern Gulf of Mexico. Fish. Bull. 75:61-89. 
Hunter, J. R. 
1976 . Behavior and survival of northern anchovy Engraulis 
mordax larvae. CalCOFI Rep. 19:138-146. 
Hunter, J. R., and K. M. Coyne. 
1982 . The onset of schooling in northern anchovy 
larvae. CalCOFI Rep. 23:246-251. 
Hunter, J. R., and C. Sanchez. 
1976 . Diel changes in swim bladder inflation of the larvae 
of the northern anchovy, Engraulis mordax. Fish. Bull. 
74:847-855. 
Lenarz, W.H. 
1972 . Mesh retention of Sardinops caerulea and Engraulis 
mordax by plankton nets. Fish. Bull. 70:839-848. 
Leslie, J. K., and C. A. Timmins. 
1989 . Double nets for mesh aperture selection and sam- 
pling in ichthyoplankton studies. Fish. Res. 7:225-232. 
Lo, N. C. H. 
1983 . Re-estimation of three parameters associated with 
anchovy egg and larval abundance: Temperature depen- 
dent incubation time, yolk-sac growth rate and egg and 
