Investigating, measuring and characterising whistles of Indian Ocean humpback dolphins Sousa plumbeain Richards Bay, KwaZulu Natal

Abstract

Some dolphin species produce individually distinctive vocalisations, called signature whistles, that broadcast the signal sender’s identity. Signature whistles are behaviourally important, and are used for recognition and group cohesion. Previous chance encounters and opportunistic behavioural observations of Australian humpback dolphins Sousa sahulensis and Indo-Pacific humpback dolphins S. chinensis have suggested signature whistle use in the genus but empirical evidence is absent. Signature whistles may be identified by analysing the temporal patterns of whistle production to identify bouts of stereotyped whistles separated by gaps lasting 1 –10s, an approach called SIGnature IDentification (SIGID). I investigated signature whistle occurrence in the endangered Indian Ocean humpback dolphins S. plumbeaat Richards Bay, South Africa. I collected acoustic data to examine the temporal patterns of the whistles of humpback dolphins to evaluate whether similar whistle contours were repeated in quick succession, and if so, to describe the pattern of repeats. I also assessed at what rate new signature whistles were discovered and recaptured. Acoustic data were collected using a hydrophone deployed at the Richards Bay harbour mouth continuously from May 2017 to September 2019, and during March 2020. Occasions when only humpback dolphins were present were verified photographically. Spectrograms of whistle contours were categorised visually by quality and similarity in shape, in a sound window with a minimum frequency of 0 kHz and a maximum of 24 kHz. I collected 30 encounters, of which potential signature whistle types were identified in 24 encounters. In total, 19 of 22 potential signature whistle types fulfilled the SIGID criteria using a sequential bout analysis, with 26.09% repeated at intervals between 0 and 1s, 60.53% between 1 and 10 s, 30.66% between 10 s and 1 min. The inter-whistle-interval population average was 11.4±14.19 s standard deviation (SD)and the median inter-whistle-interval was 5.30s. There were 667 signature whistle contours occurring in bouts and 672 occurring out of bout. Daily recaptures ranged from 2 to 18 days, and annual recaptures were high with 8 signature whistle types captured in 2 years, 7 captured in 3 years, and 1 captured in all 4 years. In conclusion, I have shown that humpback dolphins in Richards Bay use signature whistle types in a natural context by visually categorising whistle types and thereafter conducting the SIGID test on potential signature whistle types using acoustic data collected from free swimming humpback dolphins. However, without identifying the whistling individual, I cannot speculate on the function of signature whistles. The high boat presence, calf presence and muddy coastal waters at the Richards Bay harbour mouth suggests that the humpback dolphin population here relies heavily on acoustic signals to maintain group cohesion. Future studies could incorporate focal follows to collect photo-identification data to investigate potential matches between signature whistle types and photo-identified dorsal fins in order to determine whether the number of photo-identified individuals in a group correspond to the number of signature whistle types classified or use acoustic tags to assess signature whistle use in Sousa and formulate species-specific bout-interval criteria. In the process, this could yield larger signature whistle type criteria. In the process, this could yield larger signature whistle type catalogues which could serve in mark-recapture studies to assess humpback dolphin population demography and aid in the conservation of this genus