A drift study of vertical distribution and mortality of Engraulis anchoita eggs and larvae.
de Ciechomski, J.D.
MetadataShow full item record
Lagrangian changes in distribution and survival within an anchovy egg and larvae patch were studied by means of multiple opening/closing sampling following a free drifting buoy. The buoy was equipped with a radar reflector and flashing light, and it was connected to a current cross drogue floating at 15 m depth, which corresponded to the main egg and larvae concentration within the patch. The drifter was launched on the Buenos Aires continental shelf, off Necochea during the spring spawning peak of 1983. Three discrete depths (within, below and above the thermocline)were repeatedly sampled close to the drogue with a Motoda sampler. Egg and larval abundances in the vicinity of the drogue were monitored at 3/4 hour intervals for a period of three days. Each plankton station included a set of hydrographic and climatological data registered simultaneously. The efficiency of the drogue as a Lagrangian marker was examined by means of the application of multifactorial ANOVA to test the significance of variability between days, light regime, sampling depths, population structure and their interactions. Wind drag on the surface unit did not significantly affect the trajectory of the drogue. Patterns of embryonic and larval vertical distribution are discussed by means of the variations in the centre of mass, in the vertical dispersion and patchiness indexes for each developmental stage analyzed. There is evidence of diel vertical migration in larvae larger than 8 mm SL. Although the incidence of feeding based on observations of the presence of undigested food in the gut content of larvae was low, a daily feeding pattern over a 14-hour period starting after sunrise, was evident. Evening ascent of larger larvae was not related to feeding activity. A single equation two-stage model, assuming age dependent mortality in the embryonic and post-larval period is derived and compared to standard models based on constant exponential decay during each developmental phase.