SUMMARYThe auditory evoked potential technique has been used for the past 30years to evaluate the hearing ability of fish. The resulting audiograms are typically presented in terms of sound pressure (dB re. 1Pa) with the particle motion (dB re. 1ms -2 ) component largely ignored until recently. When audiograms have been presented in terms of particle acceleration, one of two approaches has been used for stimulus characterisation: measuring the pressure gradient between two hydrophones or using accelerometers. With rare exceptions these values are presented from experiments using a speaker as the stimulus, thus making it impossible to truly separate the contribution of direct particle motion and pressure detection in the response. Here, we compared the particle acceleration and pressure auditory thresholds of three species of fish with differing hearing specialisations, goldfish (Carassius auratus, weberian ossicles), bigeye (Pempheris adspersus, ligamentous hearing specialisation) and a third species with no swim bladder, the common triplefin (Forstergyian lappillum), using three different methods of determining particle acceleration. In terms of particle acceleration, all three fish species have similar hearing thresholds, but when expressed as pressure thresholds goldfish are the most sensitive, followed by bigeye, with triplefin the least sensitive. It is suggested here that all fish have a similar ability to detect the particle motion component of the sound field and it is their ability to transduce the pressure component of the sound field to the inner ear via ancillary hearing structures that provides the differences in hearing ability. Therefore, care is needed in stimuli presentation and measurement when determining hearing ability of fish and when interpreting comparative hearing abilities between species.

A growing body of evidence suggests that larval fish use sound as an orientation cue to remotely locate suitable settlement habitats. Several theoretical models have used hearing thresholds to estimate the distance at which fish larvae can detect reefs. However, researchers have often measured hearing thresholds from fish raised in aquaculture environments, or held them in closed-system tanks once caught, where equipment can significantly increase the noise levels in the tanks. The possibility arises that exposure to such noise negatively affects the hearing ability of the fish held there, such that distances predicted from these studies may be underestimated. The objective of this study was to compare the hearing ability of wild versus aquacultured snapper Pagrus auratus using auditory evoked potentials. Juvenile snapper from aquaculture tanks had significantly higher hearing thresholds at 100, 200 and 400 Hz than wild snapper. A controlled noise-exposure experiment, where snapper were ex posed for 2 wk to low-intensity noise at 120 dB re 1 µPa, confirmed these results. When plotted using an extended reef-based model, the distance at which aquaculture-raised snapper are predicted to detect a particular reef was half of that found for their wild counterparts. Clearly, the acoustic history of experimental subjects is an important consideration when using hearing thresholds to make ecological estimations, such as the distance at which larvae could detect reefs.

Atlantic cod Gadus morhua populations in the northeast USA have failed to recover since major declines in the 1970s and 1990s. To rebuild these stocks, managers need reliable information on spawning dynamics in order to design and implement control measures; discovering cost-effective and non-invasive monitoring techniques is also favorable. Atlantic cod form dense, site-fidelic spawning aggregations during which they vocalize, permitting acoustic detection of their presence at such times. The objective of this study was to detect spawning activity of Atlantic cod using multiple fixed-station passive acoustic recorders to sample across Massachusetts Bay during the winter spawning period. A generalized linear modeling approach was used to investigate spatio-temporal trends of cod vocalizing over 10 consecutive winter spawning seasons (2007-2016), the longest such timeline of any passive acoustic monitoring of a fish species. The vocal activity of Atlantic cod was associated with diel, lunar, and seasonal cycles, with a higher probability of occurrence at night, during the full moon, and near the end of November. Following 2009 and 2010, there was a general decline in acoustic activity. Furthermore, the northwest corner of Stellwagen Bank was identified as an important spawning location. This project demonstrated the utility of passive acoustic monitoring in determining the presence of an acoustically active fish species, and provides valuable data for informing the management of this commercially, culturally, and ecologically important species.