Effects of vessel size and trawling on estimates of pelagic fish backscatter in Lake Huron

摘要

Integrated acoustic and netting surveys are conducted in marine and freshwater environments to assess the status of fisheries resources and study fish population dynamics. Vessel avoidance by fish, due to noise generated by the survey vessel or from simultaneous trawl operations, can bias acoustic estimates of fish density. Unlike in marine systems, few studies have assessed vessel avoidance in freshwaters, including the Laurentian Great Lakes. Annual acoustic and netting surveys have been conducted in Parry Sound, a coastal embayment of Georgian Bay, Lake Huron (Ontario, Canada). Parry Sound supports one of the lake's few rehabilitated native lake trout Salvelinus namaycush populations, as well as abundant pelagic prey fish. In 2014-2017, night-time fisheries acoustic surveys were conducted with a 120 kHz transducer while intermittently deploying a pelagic trawl from an 18.9 m long vessel, a typical size of vessel used for acoustic surveying on the Laurentian Great Lakes. In 2017, a simultaneous acoustic survey was conducted from a 7 m long vessel with a 120 kHz transducer. To assess potential bias associated with trawling, fish backscatter was compared before, during, and after trawl operation. Furthermore, to assess potential bias associated with vessel size, fish backscatter was compared between the large and small vessel surveys. Fish backscatter was significantly lower and deeper during and after trawl deployment compared to immediately before, suggesting fish avoided the trawling vessel. In contrast, there was no significant difference in fish backscatter between the large and small vessel, although mean backscatter depth was slightly deeper from the large vessel. These results show bias from avoidance is possible in the fish density estimates obtained from hydroacoustic and trawling surveys in large freshwater ecosystems.