Statistical dependence for detecting whale-watching effects on humpback whales

Whale-watching is one of the fastest growing ecotourism industries and involves the observation of endangered wild cetacean species. However, this growth has raised concerns because of the negative effects this activity may have on the behavior and survival of focal species. Hence, detecting the effects of this activity requires sensitive analytical methods allowing the implementation of regulations to protect cetacean welfare. We compared the performance of different hypothesis tests from classical and Bayesian approaches to detect whale-watching effects on humpback whale (Megaptera novaeangliae) behavior. From a cliff located 31 m above sea level in northern Peru, we measured breathing frequency, surface time, long dive duration, directness index (i.e., path linearity), and swimming speed of humpback whales before, during, and after encounters with whale-watching boats. During 167 hours of observation, we tracked 180 humpback whale groups; 43% of groups had calves and 57% did not. Inference by null-hypothesis testing indicated significant changes only in directness index after boat encounters in groups with a calf. Other methods of inference detected moderate behavior responses as increments in the number of adult breaths, swimming speed, and dive intervals for adults and calves. Whale-watching regulations must be implemented in Peru to regulate number of boats, distance to whales, approximate speed, and time observing humpback whales. Whale-watching of humpback whales with calves should be avoided.