Evaluation of Dual Frequency Data to Differentiate Species in the Great Lakes

Fisheries acoustic assessments in the Laurentian Great Lakes have typically been conducted with single-frequency systems operating at 120 kHz.  The use of a single frequency tends to limit the ability to distinguish species using scattering characteristics.  Recent efforts have revealed the utility of using a combination of frequencies to distinguish mysids from fish in the Great Lakes.  In addition to this ability, predictions from scattering models indicate that adding certain frequencies (e.g. 38 kHz) can provide target strength data that is less sensitive to organism orientation.  Only two Great Lakes species have been examined using a combination of 38 and 120 kHz (mysids and alewife).  We sought to characterize and compare scattering at these two frequencies for additional species and to compare estimated fish densities derived from the two frequencies.  We used traditional sampling (mysid netting and midwater trawling for fish) to identify areas within lakes Michigan and Huron that were dominated by single species.  Areas dominated by single fish species were identified using recursive partitioning, while areas dominated by mysids were identified visually using expert opinion and expectations from a scattering model.  Using lakewide acoustic survey data from 2010 and 2011, we measured the distribution of dB differences for 38 and 120 kHz for areas dominated by single species and compared these distributions to determine if the two frequencies could be used to differentiate species.