Descriptive analysis of mesopelagic backscatter from acoustic data collected in the Ross Sea

This report presents results from a pilot study to investigate the use of acoustics to estimate middle trophic level prey organisms (e.g., krill and mesopelagic fish) in the Ross Sea. Single frequency (38 kHz) acoustic data were available from New Zealand longline vessels participating in the exploratory fishery for toothfish in 2002–03 and 2005–06, and multifrequency (12, 38, and 120 kHz) acoustic data were collected from the research vessel Tangaroa during a voyage to the western Ross Sea in February–March 2006. Analyses were carried out to assess data quality, describe different mark types, and quantify spatial and vertical distribution of acoustic backscatter in the upper 1000 m of the water column.
There were clear spatial patterns in the amount and type of mesopelagic backscatter observed. There was much more backscatter in the upper 1000 m and a wider variety of mark types north of 67° S. Common mark types in the northern region included a surface layer at less than 50 m depth, schools and layers centred on about 200 m and 400 m depth, and a diffuse deep scattering layer centred at 750 m depth. Average acoustic density was lower south of 70º S, and most of the backscatter was from schools and layers shallower than 100 m. Near bottom marks were associated with areas shallower than 1000 m on the Ross Sea shelf edge. In general, the amount of backscatter observed in the Ross Sea was much lower than that observed in shelf areas off New Zealand (Chatham Rise and Campbell Plateau).
Little direct information is available on the species composition of different mark types in the Ross Sea. However, different acoustic responses across the three frequencies available on Tangaroa provided some clues about the likely identity of the key scatterers. Marks shallower than 100 m depth were typically much stronger on 120 kHz than on 38 kHz, and weak on 12 kHz. This type of acoustic response is typical of krill or other large zooplankton. Schools and layers at 200–400 m depth showed a more consistent response across all three frequencies and may have been associated with small fish. This study identified key areas and mark types for further research, including directed sampling, and showed how fishing vessels could be used to opportunistically collect acoustic data.