The Lombard effect in fishes: How boat noise impacts oyster toadfish vocalization amplitudes in natural experiments

Abstract

The Lombard effect (an involuntary increase in vocal levels in noisy environments) has been shown for humans, birds, and mammals. Here, we use experimental playbacks of vessel noise and other natural sounds in the normal soundscape of the oyster toadfish Opsanus tau to test if the Lombard effect occurred. Experiments were conducted at a noisy site adjacent to a port with high vessel traffic and in a quiet embayment. We played back vessel noise (inboard and outboard motor noise), predator sounds (bottlenose dolphins) and snapping shrimp sounds for 600 s and recorded the vocalizations made by toadfish in experimental dens during 600 s periods before, during, and after the playback period. Average call power of vocalizations increased by 6.8 dB during and 8.7 dB re 1 Pa 2 after playbacks of noise relative to pre-period levels, demonstrating the Lombard effect in toadfish. Fish at the noisy site had higher average call power relative to the quiet site. There was no change in the fundamental frequency of calls in after noise playback (average frequency was 224-233 Hz). Communication signals by oyster toadfish males may be masked by very loud vessels, and the Lombard effect is an attempt to overcome masking.