Marine seismic surveys: analysis and propagation of air-gun signals; and effects of exposure on humpback whales, sea turtles, fishes and squid.

An experimental program was run by the Centre for Marine Science and Technology of Curtin University between March 1996 and October 1999 to study the environmental implications of offshore seismic survey noise. This work was initiated and sponsored by the Australian Petroleum Production Exploration Association. The program: characterised air-gun signal measurements; modelled air-gun array sources and horizontal air-gun signal propagation; developed an 'exposure model' to predict the scale of potential biological effects for a given seismic survey over its duration; made observations of humpback whales traversing a 3D seismic survey; carried out experiments of approaching humpback whales with a single operating air-gun; carried out trials with an air-gun approaching a cage containing sea turtles, fishes or squid; and modelled the response of fish hearing systems to air-gun signals. The generalised response of migrating humpback whales to a 3D seismic vessel was to take some avoidance manoeuvre at > 4 km then to allow the seismic vessel to pass no closer than 3 km. Humpback pods containing cows which were involved in resting behaviour in key habitat types, as opposed to migrating animals, were more sensitive and showed an avoidance response estimated at 7-12 km from a large seismic source. Male humpbacks were attracted to a single operating air-gun due to what was believed the similarity of an air-gun signal and a whale breaching event (leaping clear of the water and slamming back in). Based on the response of captive animals in cold water to an approaching single air-gun and scaling these results, indicated sea turtles displayed a general 'alarm' response at an estimated 2 km range from an operating seismic vessel and behaviour indicative of avoidance estimated at 1 km. Similar trials with captive fishes showed a common fish 'alarm' response of swimming faster, swimming to the bottom, tightening school structure, or all three, at an estimated 2-5 km from a seismic source. Modelling the fish ear predicted that at ranges < 2 km from a seismic source the ear would begin a rapid increase in displacement parameters. Captive fish exposed to short range air-gun signals were seen to have some damaged hearing structures, but showed no evidence of increased stress. Captive squid showed a strong startle responses to nearby air-gun start up and evidence that they would significantly alter their behaviour at an estimated 2-5 km from an approaching large seismic source