Species identification of pelagic fish schools using acoustic descriptors and ancillary information

Abstract

It is shown that schools of similarly sized and behaving pelagic fish species can be accurately identified to species over a large spatial area, on the order of the broad spatial-scales typical of standard assessment surveys, and over a time-span of three years. Acoustic measurements of morphometric, energetic, and bathymetric features of anchovy, sardine, and round herring schools were extracted using commercially available software from acoustic data collected by a conventional single-frequency narrow-band echosounder during trawling operations of pelagic stock surveys of the South African continental shelf (November 1997, 1998 and 1999). Discriminant function analysis of 18 descriptors of the 214 schools for which the species composition was satisfactorily determined by trawl samples indicated that schools could be correctly identified to species in 88.3% of all cases (94.9% for anchovy, 82.6% sardine, 82.6% round herring). The three species were differentiated primarily on the basis of school bathymetric position and backscattered energy. Including ancillary information (latitude, longitude, sea surface temperature, bottom depth, and time of day) in analysis improved the accuracy of school identification to 94.9%, since such variables allowed discrimination on the basis of interspecific differences in habitat use, as well as in schooling behavior. a)Currently at MIT/Woods Hole Joint Program.