Study of whistle spatio-temporal distribution and repertoire of a school of false killer whales, Pseudorca crassidens, in the eastern south pacifi c

Abstract

Eight recording sessions were performed with a Ca20 hydrophone of a school of false killer whales, Pseudorca crassidens, circling Chañaral Island, Chile, at 8.1 knots of average speed during almost 16 hours. From a total of 153 whistles, 148 were selected by their quality and analyzed with Linux-based Baudline DSP software. General whistles and new whistles were recorded at rates of 5.3 and 2.4 per minute, respectively. Four recording sessions, at three sites, resulted in a whistle repertoire composed of 69 distinct spectrograms with 6.9 kHz of average frequency (3.8 - 13.6 kHz) and 0.09 to 2.86 seconds of duration. Whistle abundance and diversity were correlated at all recording sites (Pearson corr.: 1, P-value: 0.008). Despite the less than four hours between sessions, only 11.6 % of the individual spectra were shared by two sites. Sessions at the same site presented a different vocal composition. On an intra-vocalization level, the number of inflection points (less than four) is slightly related to whistle duration (Pearson corr.: 0.207, P-value: 0.012), and whistle structure is determined by variations on frequency, duration and, also, energy. With distance, the lower intensity sections of six of the vocalizations faded-out, resembling different individual whistles. Therefore, vocal repertoires may be overestimated if this effect is not considered. Only 4.3 % of the assessed whistle repertoire was represented at all recording sites. These results suggest that neither the vocal repertoire nor its conservative section is likely to be able to be assessed by a single recording event.