Shipping noise induces changes in fin whale songs

Criterion 11.2 of the Marine Strategy Framework Directive addresses continuous noise, including shipping. Its implementation requires setting thresholds to describe favorable conditions of habitats. Hence, the level of onset of behavioral effects on indicator species must be determined. Fin whales are such candidates for thresholds on continuous noise. We analyzed 645 days of recordings (1999-2002) from the Ligurian Sea. The whole dataset was used to assess presence/absence of fin whale songs, while a subset was used for song analysis, selecting only files with high signal-to-noise ratio. One-hour recordings for each day were randomly selected, obtaining a final subset consisting of 134.5 hours. Our fin whale songs lasted 6 minutes on average (SD=246.83 s) and were separated from each other by rest periods (max duration= 948.80 s, mean=156.82 s, SD=125.98 s). Two types of pulses prevailed, respectively “classic” and “backbeat”. These were separated by stereotyped inter-pulse intervals (IPI, mean=16.33 s, SD=4.08 s). Novel findings included: a) the year-round presence of fin whales in the Ligurian Sea; b) acoustic presence is highest during autumn, when births of fin whale calves in the Mediterranean Sea peak, confirming the hypothesis that songs may have a role in reproduction; c) acoustic presence is lowest during summer coinciding with males spending most of their time feeding, in view of the reproductive season; d) shipping noise significantly affects songs with respect to the following parameters: IPI, rest, maximum- and minimum frequency, peak- and center frequency of the notes; e) Markov-chains and sequence analysis show that shipping noise affects the probability of transitioning from one type of note to another. Our findings are crucial to understand how fin whale songs may have changed over the years in response to increasing shipping noise. Changes usually come at higher energetic cost, likely reducing fitness.