Behavioural effects of naval sonars on fish and cetaceans

Abstract

Anthropogenic noise in the sea has increased during the past decades as a consequence of increased shipping traffic, oil and gas exploitation, and underwater construction work (Southall et al., 2007; Slabbekoorn et al., 2010). Such anthropogenic noise may affect marine animals by causing hearing injuries (Smith et al., 2004; Popper et al., 2007), masking of biological sounds (Richardson et al., 1995), or behavioural responses such as avoidance of the exposed habitat (e.g. Engås et al., 1996; Morton and Symons, 2002). In 2006, the Norwegian Navy began to operate a new class of frigates, equipped with high-power sonars transmitting at frequencies overlapping with the hearing range of several species of fish and marine mammals found in Norwegian waters (Enger, 1967; Richardson et al., 1995). As naval sonar operations often coincide with the geographic distribution of these species, an examination of potentially negative effects was needed. My PhD work includes three papers on the effect of such sonars have on herring (Clupea harengus), and one paper concerning the effect of sonars on the diving behaviour of three species of cetaceans; killer whales (Orcinus orca), pilot whales (Globicephala melas) and sperm whales (Physeter macrocephalus). In paper 1, herring were exposed to sonar signals at 1-2 kHz (Low Frequency Active Sonar, LFAS) and 6-7 kHz (Mid frequency Active Sonar, MFAS), and playbacks of recorded sounds produced by killer whales while feeding, during herring overwintering in the Lofoten area. Herring behaviour was monitored by two upward-looking echosounders, located 400 m apart, as the sonar source approached and passed the fish. No significant horizontal or vertical avoidance reactions were detected in response to the sonar transmissions, however, the killer whale feeding sounds induced both vertical and horizontal movements. The results indicate that sonar transmissions at 1-2 kHz and 6-7 kHz have negligible influence on herring at the source levels tested (197 and 209 dB re 1 Pa, for LFAS and MFAS, respectively). The behavioural response during playback of killer-whale feeding sounds demonstrates the natural avoidance reaction, and the ability of the experimental design to reveal it. In paper 2, herring schools were exposed to LFAS and MFAS sonar signals as well as playbacks of recorded sounds of feeding killer whales during herring summer-feeding migration in the Norwegian Sea. Herring behaviour was monitored by a high-frequency fishery sonar (110 kHz) following the moving schools. The schools did not dive nor change their packing density in response to the LFAS and MFAS signals at estimated maximum received sound pressure levels of 176 and 157 dB (re 1 Pa m) and sound exposure levels of 181 and 162 dB (re 1 μPa² s) for LFAS and MFAS, respectively. In contrast, killer-whalefeeding sounds induced diving responses by the herring. Based on these results combined with those from papers 1 and 3, and additional information from the literature, thresholds of behavioural responses and injury were estimated, and a simple model was constructed to estimate the accumulated effects of sonar on the herring population. It is concluded that naval sonars are unlikely to cause any behavioural changes or physiological injuries which would affect the herring population. In paper 3, captive herring in a net pen were exposed to sonar signals at 1.0-1.6 kHz frequency, transmitted by a naval sonar source on a naval frigate. Experiments were conducted in all seasons. No behavioural reactions were detected, with maximum received sound pressure levels (SPL) of 168 dB re 1μPa, at a distance of 500 m between the herring and the frigate. However, the fish did exhibit a significant diving reaction when exposed to noise from an outboard engine at a much lower SPL. This showed that captive herring exhibit avoidance responses similar to wild herring. Further, it demonstrated that the experimental setup allowed such responses to be detected when they occurred. In paper 4, diving behaviour of killer whales, pilot whales and sperm whales during exposure of MFAS and LFAS signals were studied during three field seasons of controlled exposure experiments. Diving behaviour were monitored before and during exposure by a digital tag attached on the animal, recording its vertical movement in addition to horizontal movement, vocalisation and received sound pressure level from the sonar. All three species tended to spend more time close to the surface during exposure, and less time diving. Changes in killer whale diving behaviour was strongly dependent on behavioural state, with whales in a feeding mode showing abrupt changes in diving behaviour, while no response were seen during travelling or socializing. Pilot whales showed only moderate responses at high exposure levels. Sperm whales conducted normal deep feeding dives during MFAS exposure, but reduced foraging activity during LFAS exposure.