Underwater noise from pile-driving and its impact on Hector's dolphins in Lyttelton Harbour, New Zealand
Noise levels were measured in Lyttelton Harbour in order to study pile-driving noise produced during wharf reconstruction. Sound recordings were made throughout the harbour, using several moored and mobile recording systems. In addition, an autonomous system recorded sound over a one-month period. Ambient noise in Lyttelton was heavily influenced by anthropogenic sources such as large and small vessel traffic, particularly in the low frequency range, as well as natural sources such as wind, rain and snapping shrimp in the mid-to-high frequency range. Measured noise levels were highly variable in time and space, with an overall RMS broadband level of 118 dB re 1 μPa near the channel. Recordings made over a month-long period showed higher levels during the day across a broad frequency range. Compared to other places heavily influenced by anthropogenic activities, noise levels in Lyttelton harbour were similar, although some very busy ports show much higher levels. Repairs to the port of Lyttelton involved 15 months of pile-driving. At a range of 100 m, 1/3 octave-band levels were raised by up to 45 dB across a wide frequency range due to pile-driving noise, exceeding background levels over an area of up to 16.3 square km. The maximum source Sound Exposure Level was estimated to be 194 dB re 1 μPa2s @ 1 m (average 182 dB). Most of the energy was within the 100-1000 Hz frequency range, but with significant energy well above 100 kHz at close range. An empirically based propagation model was fitted to estimate the loss in dB with range, and to allow visualisation of how the noise spread throughout the harbour. The bathymetry of the harbour and the breakwater significantly influenced propagation of pile-driving noise. Levels measured in this study tended to be lower than in other studies of pile-driving noise, due mainly to smaller pile drivers and softer substrate in Lyttelton. The impact of this noise on Hector’s dolphins was investigated using passive acoustic monitoring devices (T-PODs). T-PODs were moored in the inner, mid and outer harbour for three months. Statistical analysis of dolphin positive minutes per day and per hour, and how these detection rates were influenced by pile-driving noise as well as environmental variables, was carried out using Generalised Additive Models. Hector’s dolphins showed a clear avoidance reaction to pile-driving noise. A decrease in the rate of detections was evident on days with piling. The detection rates recovered to pre-piling levels after 50-83 hours. A simultaneous increase in detections at the mid-harbour T-POD suggests that the animals disturbed by the noise were displaced toward the mid harbour. Based on hearing studies of harbour porpoise, pile-driving noise levels in Lyttelton could cause temporary hearing damage to Hector’s dolphins. The shallowness and form of the harbour restricted noise propagation and therefore reduced the potential zone of impact on hearing. Hector’s dolphin show avoidance reactions at slightly lower levels than estimated for harbour porpoise, indicating Hector’s dolphin may be more sensitive to the disturbance of pile-driving.
Advisor: Dawson, Stephen Michael; Rayment, William
Degree Name: Master of Science
Degree Discipline: Marine Science
Publisher: University of Otago
Keywords: Cephalorhynchus; hectori; pile-driving; noise; construction; anthropogenic; ambient; soundscape; harbour; Hector's dolphin; underwater noise
Research Type: Thesis