Comparison of acoustic and Miller high-speed sampler estimates of larval fish abundance in Oneida Lake, New York

Abstract

We compare acoustic measures of target strength (TS) (at 70 kHz) and fish abundance with fish size and abundance obtained from high-speed Miller larval fish samplers in Oneida Lake, New York, collected in June and July, 1995–1999. High-speed Miller samplers showed two peaks in size distributions on all dates, 3–15 and 15–30 mm. Corresponding acoustic TS distributions also showed two peaks, one between −76 and −64 dB and the other between −64 and −52 dB. TS of small yellow perch ( Perca flavescens) 15–30 mm followed expectations from a widely used equation by Love, but this equation overestimated the TS of smaller larvae. Depth distribution in Miller samplers and acoustics were also similar. Abundance estimates obtained with the two methods were highly correlated ( r>0.8) for both fish groups. However, acoustics gave higher abundance estimates than Miller samplers for the smaller fish larvae when densities in the samplers were low. We conclude that reliable estimates of abundance of fish as small as 15 mm are obtainable with acoustics, but that it is difficult to exclude noise or invertebrate targets in abundance estimates of smaller fish larvae.