Nature 404, 35– 36 (2000)

Our title has led some readers to misinterpret our conclusions. Questions have also been raised regarding the speed at which the observations were carried out.

Autosub-1, the Autonomous Underwater Vehicle that we compared with the fisheries research vessel Scotia, had an operating speed of 3 knots. Vessel noise is usually thought to increase with speed, but the noise signature of Scotia (Fig. 1) shows that she makes a similar amount of noise at a low speed and at the normal survey speed of 10 knots. Our conclusions are, therefore, valid at the usual survey speed for this vessel.

Figure 1
figure 1

Noise levels of the Scotia at different speeds compared with the noise specification recommended by the International Council for the Exploration of the Sea.

Our paper's title has, in some cases, been taken to imply that avoidance is not a problem for any research vessel, without considering the qualifications we thought had been made explicit in the text. The Scotia is very quiet: she was the first vessel to be built to the International Council for the Exploration of the Sea's specifications for limiting noise1 (Fig. 1). We believe that our findings, valid only for such noise-reduced vessels, endorse this objective and justify the 5–10% increase in build cost. Our research shows that fisheries research vessels should be built to the same noise specification as Scotia in order to ensure that avoidance is not a source of bias.