AbstractThis study addresses the mesopelagic fish, Benthosema glaciale focusing on feeding, vertical distributions and diel migration pattern. During a cruise with the vessel “Trygve Braarud” to Masfjorden, ~50 km north of Bergen (60◦52’30 N, 5◦24’70 E) from the 2nd to the 6th October 2006 fish were sampled by oblique and horizontal trawls. A bottom deployed 38 kHz split beam echosounder, placed at 400 m depth, provided continuous acoustic data on both population and individual level. Mesozooplankton (potential prey) were sampled by net tows and the physical oceanography by CTD and a current meter.The main population of B. glaciale occurred at daytime in a weak back-scattering layer at 180-230 m. At night, peaks in concentrations were found at 0-100 m and in the scattering layer located at 230-270 m. Solitary individuals in a state of low activity were observed at 300-400 m at all times. No evidence of asynchronous diel migration pattern was found and it appears that individuals of B. glaciale are switching between performing a diel vertical migration or remain at depth even at night. During daytime young individuals occurred higher up in the water column with older individuals staying deeper. This was suggested both by oblique trawl sampling, and by sequential horizontal tows made in upper layers at sunset, showing that small specimens were the first to appear in near-surface waters during diel vertical migration. Benthosema glaciale were mainly feeding of copepods and in particular Calanus sp. but also on euphausiids, sergestiid shrimps and ostracods. Evidence of both day and night feeding was found but most stomachs examined had very little content, often much digested. No difference was observed in fullness and digestive state between migrants and non-migrants.Acoustic target tracking showed that the target strength varied from -70 dB to -50 dB for the group ascribed as mesopelagic fish, with a peak at -60 dB corresponding to fish with a length of 6 cm. The majority of the tracks were long lasting and had little or no speed. Four tracks were studied in more detail. The individual swimming speed was low, yet with frequent changes in speed usually including a rapid shift in vertical position. Such intermittent changes in vertical position were also seen in echograms at 300-400 m. The behaviour may represent a saltatory search behaviour, in which pauses to search for prey alternate with reposition of the predator for scanning new territory.During the study, the hydrography was vertically homogeneous below the sill and no currents were recorded at depth, suggesting that the results on vertical distribution, swimming speeds and behaviour in the deep part of the water column were not affected by these physical variables.