The higher, the closer, the better? Influence of sampling frequency and distance on the acoustic properties of short‐beaked common dolphins burst pulses in the Mediterranean Sea. (22nd October 2019)

• Record Type: Journal Article
• Authors: Papale, Elena; Alonge, Giuseppe; Caruso, Francesco; Grammauta, Rosario; Mazzola, Salvatore; Mussi, Barbara; Pace, Daniela S.; Buscaino, Giuseppa
• Other Names: Pace Daniela S. guestEditor.; Mussi Barbara guestEditor.; Vella Joseph G. guestEditor.; Vella Adriana guestEditor.
• Is Part Of: Aquatic conservation. Volume 31(2021)Supplement 1
• Page Start: 51
• Page End: 60
• Publication Date: 2019-10-22
• Subjects: burst‐pulses -- clicks -- sampling frequency -- short‐beaked common dolphin -- spectral properties -- transmission loss; Aquatic ecology -- Periodicals; Conservation of natural resources -- Periodicals; Aquatic resources -- Periodicals; 333.95216
• DOI: 10.1002/aqc.3158
• Languages: English
• ISSNs: 1052-7613

Abstract

Abstract: When implementing conservation strategies involving passive acoustic monitoring, an accurate description of signal properties is crucial, especially when studying threatened and elusive vocalizing species. Misrepresentation of detected sounds can result in information loss or misclassification of species, thus compromising the management process. Distortion of the time–frequency features depends on the characteristics of acoustic propagation through sea water, sensitivity and sampling frequency of the acquisition system and distance between cetacean source and the receiving sensor (hydrophone). This study examined how the sampling frequency and the distance from the source affect the properties of burst pulsed sounds emitted by short‐beaked common dolphins in the Mediterranean Sea and how the signal is modified in relation to the recording scenarios. Burst pulses were collected at 300 000 samples/s, and down‐sampled at 192 000, 96 000, 48 000 and 44 100. Fifteen acoustic parameters were automatically measured. The median values were measured and compared among the signals sampled at different rates. Finally, a transmission loss model was applied to explore the variation of the power spectrum density during click propagation through seawater. The results showed that, even if higher sampling rates are a more accurate match with the original signal, some parameters (number of clicks, inter‐click‐interval and maximum peak) measured at 96 000 and 192 000 Hz did not display significant differences compared with 300 000 Hz, except for the frequency parameters. The transmission loss model also showed that the highest frequency components strongly attenuate within 800 m from the source. These results suggest that high recording distances and lower sampling rates do not allow an accurate analysis of burst pulsed signals. However, taking into account data storage limitations, recording at the highest frequency is not always required for an accurate description of signals and the outcomes of this study improve the accuracy of passive acoustic monitoring systems, thus enhancing the potential for conservation of short‐beaked common dolphins.