There are numerous studies showing that there is a constant increase in the ocean ambient noise level and the ever-growing demand for developing algorithms for detecting weak signals in ambient noise. In this study, we utilize dynamical and statistical complexity to detect the presence of weak ship noise embedded in ambient noise. The ambient noise and ship noise were recorded in the South China Sea. The multiscale entropy (MSE) method and the complexity-entropy causality plane (C-H plane) were used to quantify the dynamical and statistical complexity of the measured time series, respectively. We generated signals with varying signal-to-noise ratio (SNR) by varying the amplification of a ship signal. The simulation results indicate that the complexity is sensitive to change in the information in the ambient noise and the change in SNR, a finding that enables the detection of weak ship signals in strong background ambient noise. The simulation results also illustrate that complexity is better than the traditional spectrogram method, particularly effective for detecting low SNR signals in ambient noise. In addition, complexity-based MSE and C-H plane methods are simple, robust and do not assume any underlying dynamics in time series. Hence, complexity should be used in practical situations.

Offshore windfarms have recently emerged as a renewable energy solution. The effects of pile‐driving and long‐term impacts of operational noises on fish chorusing still, however, remain largely unknown. In this study, we investigated the variations of fish chorusing intensity and duration during the construction (2016) and operational phases (2017–2018) of a windfarm at the Eastern Taiwan Strait (ETS). At the ETS, two types of fish choruses (Types 1 and 2) were found to repeat over a diurnal pattern. In the 2 days after the pile driving, Type 1 chorusing showed lower intensity and longer duration, while Type 2 chorusing exhibited higher intensity and no changes in its duration. During the operational phases in 2017 and 2018, both choruses were longer in duration (2–3 h for Type 1; 0.5–1 h for Type 2). The intensity of Type 1 chorus increased by 5–10 dB, but no significant variation was recorded for Type 2. Our findings document, for the first time, different behavioral responses of two chorusing types exposed to pile‐driving and windfarm noise pressure. Fish choruses have been associated with several behavioral functions. Deviation from a regular pattern might affect fish reproductive success, cause migration, augmented predation, or physiological alterations. Because offshore windfarms installations are growing in magnitude and capacity worldwide, this study provides essential insights for policymakers and constitutes an important reference for assessing the effects of noise from pile driving and windfarms on fishes.

Automated acoustic indices to infer biological sounds from marine recordings have produced mixed levels of success. The use of such indices in complex marine environments, dominated by several anthropogenic and geophonic sources, have yet to be understood fully. In this study, we introduce a noise resilient method based on complexity-entropy (hereafter named C-H) for the detection of biophonic sounds originating from fish choruses. The C-H method was tested on data collected in Changhua and Miaoli (Taiwan) during the spring in both 2016 and 2017. Miaoli was exposed to continual shipping activity, which led to an increase of ~10 dB in low frequency ambient noise levels (5–500 Hz). The acoustic dataset was successively analyzed via the acoustic complexity index, the acoustic diversity index and the bioacoustic index. The C-H method was found to be strongly correlated with fish chorusing (Pearson correlation: r_H < −0.9; r_C > 0.89), and robust to noise originating from shipping activity or natural sources, such as wind and tides (r_H and r_C were between 0.22 and −0.19). Other indices produced lower or null correlations with fish chorusing due to missed identification of the choruses or sensitivity to other sound sources. In contrast to most acoustic indices, the C-H method does not require a prior setting of frequency and amplitude thresholds, and is therefore, more user friendly to untrained technicians. We conclude that the use of the C-H method has potential implications in the efficient detection of fish choruses for management or conservation purposes and could help with overcoming the limitations of acoustic indices in noisy marine environments.

Passive acoustic monitoring technology can depict underwater soundscapes, yet phenological and seasonal patterns and shifts in soundscapes are still poorly understood. Here we analyse the seasonal soundscape in the Eastern Taiwan Strait in 2017 using nonlinear dynamical complexity modelling. We find an annual phenology of fish chorusing that begins in spring, and peaks in summer before beginning to subside in autumn and becoming silent in winter. During spring and summer, the soundscape exhibited significantly higher complexity and predictability than in autumn and winter, due to the presence of regular fish chorusing. Soundscape dynamics shifted from being nonlinear in spring and summer to being linear and stochastic in autumn and winter. Our findings suggest that soundscapes could be used to measure phenological patterns and seasonal shifts in marine species behaviour. We propose that monitoring soundscapes could help assess the long-term health of marine ecosystems under environmental and climatic change.

Offshore wind farms have recently emerged as a renewable energy solution. However, the long-term impacts of wind turbine noise on fish chorusing phenology are largely unknown. We deployed a hydrophone 10 m from a foremost turbine in Taiwan situated at the Miaoli offshore wind farm (Taiwan Strait) for two years to investigate sound levels and assess the potential influence of turbine noise on seasonal fish chorusing patterns during 2017 and 2018. Wind turbine noise (measured in the 20–250 Hz frequency band) was significantly higher in autumn and winter (mean SPL: 138–143 dB re 1 μPa) and was highly correlated with wind speed (r = 0.76, P < 0.001). During both years, fish chorusing exhibited a consistent trend, that is, beginning in spring, peaking in summer, decreasing in autumn, and absent in winter. Our results show the noise from a single turbine during the two-year monitoring period did not influence the seasonal fish chorusing (r = −0.17, P ≈ 1). Since the offshore wind farm installations are growing in magnitude and capacity across the Taiwan Strait, this study for the first time provides baseline operational sound levels and an understanding of the fish seasonal vocalization behavior at the foremost turbine of the first wind farm in Taiwan. The results presented here provide useful insights for policymakers and constitute a reference starting point for advancing knowledge on the possible effects of wind turbines on fish chorusing in the studied area.

Recent offshore windfarm development has led to increased vessel traffic in the Eastern Taiwan Strait, which is part of the habitat of the critically endangered Taiwanese humpback dolphin (Sousa chinensis taiwanensis). However, data on possible effects on the behavior of this endemic subspecies are lacking to date. In this study, we observed Taiwanese humpback dolphins’ acoustic behavior associated with shipping noise and analyzed their whistles and clicks before, during, and after vessel transit. Before vessel transit, the median rate of dolphin whistles and clicks was 100 and 1550 counts per minute, which significantly reduced to less than 8 and 170 counts per minute during and after vessel transit. Dolphins produced significantly shorter whistles during (0.07 s) and after (0.15 s) vessel transit. The vocalizing behavior of the Taiwanese humpback dolphin may be affected by vessel transit, which, if sustained, could possibly influence the individual communication and feeding success of the population. Implementing measures such as re-routing of the vessel lanes and regulating the speed of the vessel traffic in the habitat are proposed to overcome the influence of vessel noise on Taiwanese humpback dolphins.