Biosonar activity of the Indo-Pacific humpback dolphin (Sousa chinensis) near the tunnel section of the world’s longest cross-sea bridge—the Hong Kong-Zhuhai-Macao Bridge—is negatively correlated with underwater noise

Xue An, Peng‐Xiang Duan, Weilun Li, Jing Yuan, Yuwei Chen, Fei Fan, Xiaojun Deng, Yousheng Xiao, Xi Chen, Ding Wang, Zhitao Wang, Kexiong WangFrontiers in Marine Science 2023 · DOI

TL;DR

Researchers used passive acoustic monitoring to track Indo-Pacific humpback dolphins near the world’s longest cross-sea bridge (HZMB). They found that dolphin biosonar activity peaks at night and during high tide, but significantly drops when underwater noise from bridge traffic and shipping increases—particularly during the winter months.

What Problem Does This Paper Address?

The Hong Kong-Zhuhai-Macao Bridge (HZMB) is a feat of engineering, but it cuts directly through the Pearl River Estuary, home to the world’s largest population of Indo-Pacific humpback dolphins (Sousa chinensis). While the impact of the bridge’s construction (e.g., pile-driving noise) was heavily studied, the impact of its operation—the constant hum of vehicular traffic and associated shipping—is less understood.

Cetaceans rely on biosonar for navigating, foraging, and communicating. Anthropogenic noise can "mask" these signals or cause auditory stress, forcing dolphins to abandon prime habitats. This paper investigates how the presence of the HZMB tunnel section and its resulting noise pollution correlates with dolphin activity patterns.

Methods

The team employed Static Passive Acoustic Monitoring (PAM). They deployed a SoundTrap 300 HF hydrophone at a depth of 5 meters near a fairway beacon adjacent to the HZMB tunnel.

  • Data Collection: The device recorded continuously at a 288 kHz sampling rate for 24-hour cycles. Monitoring occurred in four distinct two-week blocks between July 2021 and March 2022 to capture seasonal variations.
  • Signal Processing: Using custom MATLAB scripts, the researchers identified "click trains" (navigation/searching) and "buzzes" (high-frequency attempts to capture prey).
  • Noise Analysis: They calculated the root mean square sound pressure level ($SPL_{rms}$) in 10-minute bins to quantify the ambient and anthropogenic noise environment.
  • Statistical Framework: The study used Kruskal-Wallis tests for seasonal/tidal variance and Spearman correlation to link noise levels directly to dolphin detections.

Datasets

The study generated a primary dataset consisting of 1,209 hours of acoustic data, partitioned into 7,254 ten-minute bins.

  • Dolphin Detections: 8,992 echolocation signals were monitored across 46 days (88.5% of the total monitoring period).
  • Environmental Data: Tidal and diel (day/night) data were synced with the acoustic recordings to provide context.

Key Insights and Ideas

  • The "Night Shift": There is a clear temporal offset between humans and dolphins. Noise levels were significantly higher during the day due to vessel traffic, while dolphin biosonar activity peaked at night. This suggests dolphins may be utilizing the area more heavily when acoustic interference is minimized.
  • Tidal Foraging: The highest acoustic activity occurred during high tide. This is likely a biological response to prey movement; as the tide rises, fish move into the estuary, and dolphins follow.
  • Seasonal Sensitivity: While dolphins were present year-round, the negative correlation between noise and activity was only statistically significant in winter. The authors hypothesize that during winter, dolphins may face higher survival pressures or that the specific frequency of winter noise is more disruptive to their foraging.
  • Avoidance vs. Masking: The researchers suggest that the drop in detections during high-noise periods isn't just "masking" (where the noise is so loud we can't hear the dolphin); rather, the dolphins may be physically avoiding the noisy waters for short periods to escape auditory stress.

Key Results

  • Presence: Dolphins were detected on 88.5% of the monitored days, confirming the area near the HZMB tunnel remains a critical habitat.
  • Noise Levels: The median $SPL_{rms}$ was 136.98 ± 9.70 dB. For context, noise levels were significantly higher in winter compared to summer.
  • Correlation: In winter, a significant negative correlation (Spearman’s rho = −0.073, p < 0.01) was found between noise levels and the number of click trains.
  • Diel Patterns: In winter, the number of click trains per 10 minutes was nearly double at night (1.96) compared to the day (1.03).

Limitations and Future Work

The authors acknowledge several constraints:

  • Spatial Scope: Data was collected from a single monitoring station. This makes it difficult to determine if dolphins are leaving the estuary entirely or just moving a few hundred meters away.
  • Duration: The two-week seasonal snapshots provide a baseline but may miss long-term trends or sporadic high-traffic events.
  • Causality: While the correlation is clear, it is difficult to decouple whether dolphins are avoiding noise or if their prey (fish) are avoiding the noise, leading the dolphins to follow.

Future Work: The team suggests expanding the PAM network to multiple sites and developing a comprehensive noise calculation model that accounts for boat types, weather, and tunnel operation noise to better inform conservation strategies.

Summary updated

Abstract

Underwater noise pollution from the world’s longest cross-sea bridge—the Hong Kong-Zhuhai-Macao Bridge (HZMB)—which stretches across the Chinese White Dolphin National Nature Reserve (of the People’s Republic of China, PRC) in the Pearl River Estuary may affect the distribution of local humpback dolphins. In this study, static passive acoustic monitoring was applied to monitor biosonar activity of humpback dolphins and underwater noise adjacent to the tunnel section of the HZMB for more than one year. During the monitoring period, dolphin biosonar signals were detected on 88.5% of days. A significant temporal pattern occurred in dolphin biosonar activity and in anthropogenic noise. Biosonar activity was significantly higher at night than during the day, whereas underwater noise occurred more in the day than at night. Significantly more echolocation signals were detected in winter-spring than in summer-autumn, and highest acoustic activity occurred significantly more during high tide than at other tidal periods. In addition, the negative correlation between elevated underwater noise and dolphin sonar activity in winter suggests that dolphins may avoid noisy waters for short periods, perhaps due to auditory stress, but fish prey movement cannot be ruled out. These findings facilitate understanding activity patterns of humpback dolphins in the Pearl River Estuary and may contribute to conservation efforts.
BibTeX
@inproceedings{an2023biosonar,
  author = {Xue An and Peng‐Xiang Duan and Weilun Li and Jing Yuan and Yuwei Chen and Fei Fan and Xiaojun Deng and Yousheng Xiao and Xi Chen and Ding Wang and Zhitao Wang and Kexiong Wang},
  title = {{Biosonar activity of the Indo-Pacific humpback dolphin (Sousa chinensis) near the tunnel section of the world’s longest cross-sea bridge—the Hong Kong-Zhuhai-Macao Bridge—is negatively correlated with underwater noise}},
  booktitle = {Frontiers in Marine Science},
  year = {2023},
  doi = {10.3389/fmars.2023.1171709},
}
Figure