Page 58 - Demo

46Marine Research of the National Academy of Marine ResearchDOI: 10.29677/MR.202406_4(1).0004A Review of Passive Acoustics Monitoring Results in the CoastalWaters off Western Taiwan for the Past Decade During the operational period of turbines No. 28 and No. 21, as well as the construction period of other turbines, passive acoustic sound files captured ship noise, marine engineering noise, and vocalizations of whales and dolphins, as depicted in Figure 7. Ship noise and dredging operations at marine works are labeled as V1, V2, D1, and D2, respectively, while the click burst of dolphins is labeled as CB (Hu et al., 2022). Given that the energy of dredging noise typically falls below 1 kHz, it is less likely to overlap with the vocal frequency bands of dolphins (Todd et al., 2015) or Taiwanese humpback dolphins. Therefore, this study exclusively examines the impact of vessel noise on the Taiwanese humpback dolphin based on seven identified noise characteristics. Figure 8 illustrates the comparison of changes in the number of whistles and clicks of dolphins as vessels pass by. Each box plot's central red mark represents the median, while the top and bottom edges indicate the 25th and 75th percentiles. The whiskers denote the maximum and minimum values, with anomalous values indicated by \a decrease in the number of whistles and clicks, as well as the duration of whistles, during vessel passages. This suggests that dolphins respond to vessel passages by reducing their call rate and producing shorter calls with simpler frequency patterns. Choral singing among fish is a prevalent phenomenon observed in passive acoustic sound files, serving various behavioral functions such as social cohesion (Van Oosterom et al., 2016), territorial defense (Buscaino et al., 2015), foraging (Versluis et al., 2000), and spawning activity (Buscaino et al., 2016; Ladich & Fine, 2006; Siddagangaiah et al., 2021). These choruses exhibit cyclical patterns, influenced by lunar phases such as full or new moons (Buscaino et al., 2016; Ladich & Fine, 2006; Siddagangaiah et al., 2021). In a study conducted by Mok's team (2009) in the coastal waters of estuaries in Taiwan, vocal fish species belonging to the Sciaenidae family, such as the blackspotted croaker (Protonibea diacanthus), Japanese croaker,and big-snout croaker, were identified. These fish exhibit vocal characteristics consisting of pulses (Mok et al., 2009), as illustrated in Figure 9. In the present study, conducted at the Miaoli Demonstration Wind Farm between 2016 and 2017, the call spectra of two distinct types of fish with differing pulse patterns were identified. Please refer to Figure 10 for details. Type 1 exhibited a peak frequency of 1.8 kHz, enabling continuous singing for durations of 5-9 hours. Conversely, Type 2 displayed peak frequencies of 0.7 kHz and 1.7 kHz, allowing for singing periods of 1.5-3 hours. Ecologists utilize concepts such as nonlinear dynamics, predictability, complexity, and regularity (Dakos, 2020) to analyze ecosystems (Siddagangaiah et al., 2022a). These concepts can also be applied to passive acoustic sound files for analyzing their complexity, regularity, predictability, and dynamics (Siddagangaiah et al., 2021). Rosso et al. (2007) developed the complexity-entropy (C-H) method, employing statistical complexity (C) and permutation entropy (H), to distinguish between noise and chaos. This method, which quantifies complexity indices, has been practically applied to the Miaoli and Changhua demonstration wind farms for detecting fish choruses (Siddagangaiah et al., 2019). It enables the tracking of fish choruses even amidst noisy ship transportation (Siddagangaiah et al., 2021). Due to physiological constraints, fish possess a limited contraction rate of sonic swim bladder muscles, restricting the vocal frequency band to primary energy levels below 1000 Hz (Ladich & Fine, 2006).3.2 Fish soundscape