Properties of abundance indices obtained from acoustic data collected by inshore herring gillnet boats
Acoustic data collection during fishing activities can be used to obtain an abundance index. A simulation, calibrated against an experiment conducted during the Pictou, Nova Scotia, Canada, 1997 inshore herring fishery, is used to understand how survey design affects the properties of abundance indices derived from these data. Two fishing survey protocols and random and systematic transect surveys were simulated. During the complete fishing survey protocol, the simulated survey boat collected acoustic data before and after a management-imposed nightly boat limit was caught. In contrast, during the incomplete fishing survey protocol, data collection was terminated when the boat limit was caught. Properties of abundance indices derived from the fishing and transect surveys were examined over five levels of fish dispersion, two conditions of fish mobility, and in the presence and absence of concurrent fleet fishing. All indices were subject to change caused by changing fish dispersion, but only the incomplete fishing survey index was highly unsatisfactory. The complete fishing survey index is more susceptible to change than the transect indices but displays a lower sampling variation across conditions than the transect indices. We conclude that the complete fishing survey index is a viable alternative to the transect indices.