(2003), whereby whistles from spinner dolphins regularly extend beyond 20 kHz. A detailed description of these results is presented in Table I.īased on the limited number of whistles recorded (n = 46) during this brief sighting, the fundamental frequency of some whistles extended beyond the human hearing range (20 kHz), which agrees with the findings of Lammers et al. The mean start and end whistle frequencies were 14.3 ± 3.4 and 16.3 ± 3.3 kHz, respectively. The whistle duration ranged from 36 to 977 ms, while the mean minimum and mean maximum frequencies were 13.2 ± 2.8 and 17.2 ± 3.0 kHz, respectively. The average signal-to-noise ratio (SNR) of 20 randomly selected whistles was 18.6 dB. Eight whistles had one harmonic and four whistles had two harmonics. Out of the 46 whistles, only 12 had harmonics. An example of whistle spectrograms from the recording is shown in Fig. Downsweep, constant, and concave whistle contours comprised 17.4% (n = 8), 15.2% (n = 7), and 13.0% (n = 6) of all whistles, respectively. The majority of the recorded whistles were upsweeps, which accounted for 54.4% (n = 25) of the whistles. Forty-six whistles were extracted only four contour types (upsweeps, downsweeps, concave, and constant contours) were detected. Water flow was detected in the recordings however, no other background sounds, such as snapping shrimps, were detected. The research vessel was on the move during the recordings to keep pace with the dolphins. The sighting lasted 30 min and a 20 min recording of the dolphin sounds was obtained. These parameters were chosen because they are consistent with studies on other spinner dolphin populations (e.g., Bazúa-Durán and Au, 2002 Bonato et al., 2015). Whistle parameters (duration, minimum frequency, maximum frequency, frequency range, and mean frequency) were calculated in RStudio (version 1.1.463 The R Foundation, Vienna, Austria). ![]() Whistles were categorized into six different contours as described in Bazúa-Durán (2004): (i) constant, (ii) downsweep, (iii) upsweep, (iv) concave, (v) convex, and (vi) sine. Only the fundamental frequency of the whistles was analysed in detail and the number of harmonics detected was noted. Each whistle contour (both fundamental frequency and harmonics) was manually drawn out using a custom-made graphical user interface (GUI) software in matlab. The recorded dolphin whistles were automatically detected using a custom-made code developed in matlab (R2019a The MathWorks, Inc., Natick, MA, USA) by the fourth author of this paper (KI) and modified slightly for the purpose of this research (see Ichikawa et al., 2006). Inspection of the recordings was first visually and aurally carried out using Adobe Audition software (Fast Fourier transform (FFT) size: 256 points Hamming window overlap 50% San Jose, CA, USA).
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |