Mesoscale features and micronekton in the Mozambique Channel: An acoustic approach

Abstract

The oceanic circulation in the Mozambique Channel (MZC) is dominated by mesoscale cyclonic and anticyclonic eddies that are known to play a key role in biological processes of less-productive deep-sea ecosystems by converting physical energy into trophic energy and by restructuring and concentrating biomass across the eddy field. In this study, hydroacoustics was used to investigate the spatial distribution of micronekton according to four classes of mesoscale features and assess whether cyclonic eddies, anticyclonic eddies or eddy edges (divergence and frontal regions) impact the density of micronekton. Acoustic data were collected continuously with a Simrad EK60 split-beam echosounder during three surveys carried out in the MZC within the framework of the MESOBIO programme. First, micronekton ascents and descents during the crepuscular periods (dusk and dawn, respectively) were similar to the well-known process of diel vertical migration, with the largest changes in the shallow layer, much smaller in the deep layer, and almost non-existent in the intermediate layer. Additionally, the acoustic densities for the total water column were greater at night than during the day, suggesting that organisms migrate from layers deeper than the water column that was sampled (740m). Second, there was evidence of differences in the acoustic responses of micronekton to mesoscale features During two of the three surveys, cyclonic eddies exhibited greater micronekton density than anticyclonic eddies for day and night. In contrast, during the last survey, the greatest micronekton density was observed in anticyclonic eddies. To explain this discrepancy, several hypotheses are proposed, including the eddy generation site and trajectory throughout the life of the eddy, eddy-eddy interactions, seasonality and difference in monsoon wind regime, the depth of influence of eddies and a low dependence of movements of larger micronektonic organisms on the mesoscale gradients. Furthermore, this study demonstrated that mesoscale features could be predicted using acoustic responses at several acoustic frequencies.