Fig. 5 Porpoise behaviour before (phase 1) and during (phase 2) boat exposure, measured over a period of one minute. The values on the y-axis of a1, a2, b1, and b2 were transformed back from their cube root form.

Time before resuming pre-disturbance behaviour

Porpoise speeds varied in the course of the experiments (10 knots: p <0.001, R2 =0.083; 20 knots: p <0.001, R2 =0.17). This was particularly evident in the 20 knots experiments, where animals tended to move faster when the boat approached, then slowing down <50 seconds after the boat had passed. When the boat approached at 10 knots, this response trend seemed not as clear (Fig. 6). However, many animals moved as fast when the boat was not in motion as they did when the boat was nearby. Additionally, porpoises were more likely to dive deep as the boat approached at 10 knots, decreasing rapidly after it passed (p <0.001, R2 =0.11). But no similar trend was observed at 20 knots (p =0.09). Although the results presented above did not suggest that porpoises turned more steeply or that they were more likely to dive deep when approached by boats, the GAM analyses indicated that the porpoises’ horizontal movements (10 knots: p =0.003, R2 =0.029; 20 knots: p <0.001, R2 =0.028) changed significantly in the course of the experiments. It is, however, worth noting that time relative to CPA explained a very small proportion of the variation in this behaviour. The probability of breathing remained unchanged throughout the experiments for both speeds (p >0.05 for both boat speeds).