Estimating Atlantic Bluefin Tuna number density using the second moment of intensity
Fish number density can be estimated from the normalized second moment of acoustic backscatter intensity [Denbigh et al., J. Acoust. Soc. Am. 90, 457–469 (1991)]. This method assumes that the distribution of fish scattering amplitudes is known and that the fish are randomly distributed following a Poisson volume distribution within regions of constant density. It is most useful at low fish densities, relative to the resolution of the acoustic device being used, since the estimators quickly become noisy as the number of fish per resolution cell increases. The method was applied to an acoustic assessment of juvenile Atlantic Bluefin Tuna, Thunnus thynnus. The data were collected using a 400 kHz multibeam echosounder during the summer month of 2009 in Cape Cod, MA. Due to the high resolution of the multibeam system used, the large size (approx. 1 m) of the tuna, and the spacing of the fish in the school, we expect there to be low fish densities relative to the resolution of the multibeam system. Results based on the normalized second moment of acoustic intensity are compared to fish packing density estimated using aerial imagery that was collected simultaneously.
Journal or Conference Title
Journal of the Acoustical Society of America
130, Issue 4
Acoustical Society of America
Digital Object Identifier (DOI)
M. L. Schroth-Miller, T. C. Weber, and M. Lutcavage, ‘Estimating Atlantic Bluefin tuna number density using the second moment of intensity’, The Journal of the Acoustical Society of America, vol. 130, no. 4, p. 2337, 2011.