Marine reef soundscape monitoring with fiber-optic distributed acoustic sensing

Abstract

Coral reefs are essential marine ecosystems that support a vast array of biodiversity and provide numerous benefits, including fisheries, tourism, and coastal protection. However, these ecosystems are increasingly threatened by various factors, including anthropogenic noise from activities such as shipping and coastal development. Traditional acoustic methods of monitoring reef health, such as hydrophones, are limited by their point-based sensing, reliance on batteries, and need for manual data retrieval, which can be labor-intensive and costly. In this study, we explore the application of fiber-optic distributed acoustic sensing (DAS) for real-time marine reef monitoring, a new application compared to its previous use in deep-sea soundscape monitoring. We deployed a fiber-optic DAS system in a reef area on the coast of the Central Red Sea, alongside a conventional hydrophone for comparison. The experiment was conducted in a degraded inshore reef near the KAUST shoreline, characterized by sand, macroalgae, scattered boulders, and encrusting sponges. This site was selected as a proxy for coral reef monitoring due to its biological activity, including snapping shrimp and the presence of reef-related fish species. Our observations revealed significant acoustic activity within the 1.5 to 5 kHz range, with snapping shrimp sounds increasing after the onshore lights were switched off, consistent with known behavioral patterns of increased acoustic activity during low-light conditions. Additionally, we detected various fish vocalizations, including drums and impulses, within the 100 to 1000 Hz range. The DAS system also successfully tracked the timing and trajectory of scuba diver movements along the reef. These findings demonstrate the potential of fiber-optic DAS technology to provide high-resolution spatial mapping of reef soundscapes, offering a comprehensive and cost-effective solution for continuous reef monitoring, thereby demonstrating the feasibility of DAS for real-time acoustic monitoring in reef environments.

Introduction

Coral reefs are among the most biodiverse marine habitats, often referred to as “rainforests of the sea” due to their ecological richness¹. While the analogy highlights biodiversity, it should not be interpreted as implying functional equivalence between terrestrial and marine ecosystems. Despite occupying less than 1% of the ocean floor, coral reefs support approximately 25% of all marine species². Additionally, coral reefs play a pivotal role in the global economy. Fishing and tourism are strongly supported by marine and coastal biodiversity. Coral reefs provide habitat and spawning grounds for commercially important fish species, sustaining both artisanal and industrial fisheries. Additionally, the visual richness and ecological integrity of reef ecosystems attract millions of tourists annually, making biodiversity a key driver of reef-based tourism³. These relationships are increasingly recognized in international frameworks such as the United Nations Sustainable Development Goals (SDGs), particularly SDG 14 (“Life Below Water”), which includes indicators for sustainable fisheries, biodiversity protection, and ecosystem resilience. The Global Fund for Coral Reefs, led by UNEP (United Nations Environment Programme), exemplifies how blended finance mechanisms are being mobilized to support coral reef conservation and restoration in alignment with SDG 14⁴.

The Red Sea hosts one of the largest reef ecosystems in the world and has been studied for its unique thermal resilience and biodiversity⁵. Red Sea coral reefs exhibit greater resilience to heat stress compared to those in other regions due to their adaptation to high temperatures^6,7. However, increased stresses, such as the rapid rise in sea temperatures and anthropogenic noises from shipping, resource exploitation, and infrastructure development, are jeopardizing their integrity. This mirrors the challenges faced by many other coral reefs worldwide⁸.

Various symptoms of gradual decay in reef health are currently visible in the Red Sea, reflecting a broader problem affecting coral reefs worldwide. One major issue is the alarming spread of mass coral bleaching, which is depleting the living coral cover across the world’s tropical coral reefs due to human activity and global warming⁸. Additionally, changes in the reef’s soundscape, comprising biophony (biological sounds), geophony (natural non-biological sounds), and anthrophony (human-generated sounds), serve as another indicator of ecological decline⁹. Among these, anthrophony has been increasing due to expanding human activity in marine environments, while biophony and geophony vary depending on environmental and ecological conditions. A resilient reef soundscape, characterized by high biological activity and acoustic diversity, is distinguishable from an acoustically impoverished reef^{10, 11, 12–13} and is fundamental in attracting marine life, from larval to adult stages, through phonotaxis¹⁴. As sound propagates further and attenuates less underwater compared to light, it has become a primary sensory channel for many marine organisms. This long-range propagation enables biological communication, navigation, and habitat selection in environments where visual cues are limited. Reef biophony, mainly dominated by broadband frequency snaps of Alpheidae sp. (snapping shrimp), may be detected tens of kilometers away by marine organisms offshore¹⁵. However, anthropogenic noise, such as shipping, construction, and other activities, can mask soundscapes and negatively impact marine life. Effects of anthropogenic noise on the behavior of fish and invertebrates have been well documented since the 1970s^16,17. The main sources of noise pollution typically fall within a frequency range of 100 to 10 kHz, overlapping the hearing ranges of reef fishes and invertebrates^13,17,18, disrupting their natural behaviors and reaction mechanisms for foraging or hunting, escaping predators, courtship and habitat selection¹³. Therefore, monitoring reef soundscapes is vital to enable a deeper understanding of changes in reef biodiversity and health while facilitating proactive measures to mitigate the harmful effects of anthropogenic noise¹¹. Snapping shrimp, in particular, are known to exhibit increased acoustic activity during low-light conditions such as dusk and nighttime. This behavior is relevant to our study, as we observed a clear increase in shrimp sound intensity after the onshore lights were turned off¹⁹.

In this context, traditional passive acoustic monitoring (PAM) units, such as hydrophones, are commonly used by ecologists to monitor the acoustic complexity of underwater ecosystems^10,14,20. PAM refers to the method of recording underwater sound without introducing artificial signals. Hydrophones are underwater acoustic sensors that convert pressure fluctuations from sound waves into electrical signals. They come in various forms and orientations, including piezoelectric and capacitive types²¹. Once manually deployed, these devices can be programmed to record intermittently, allowing extended operation²². While many hydrophones rely on batteries, which require periodic replacement and manual retrieval, more advanced systems use cabled connections to surface buoys or shore-based stations, enabling continuous power and real-time data access. Diver-based retrieval is a common and manageable practice, but DAS offers an alternative approach by enabling continuous data access without physical intervention, which may be advantageous in large-scale or remote deployments. Nevertheless, the most significant limitation lies in their pointwise sensing capacity. Hydrophones monitor only their immediate surroundings, and accurate placement is essential given the high spatial and temporal variability of reef soundscapes, which depend on factors such as reef morphology, depth, and species presence^23,24. Achieving representative coverage requires dense PAM arrays, which are often economically unfeasible, especially in resource-limited regions²⁵. Additionally, hydrophones, like other electronic devices, may be affected by harsh reef conditions, including high temperatures and salinity²⁶. In regions such as the Red Sea during summer months, water temperatures can exceed $35_{}^{\circ }\text{C}$ ²⁷, which presents operational challenges for any electronic-based device deployed over extended periods.

In response to these challenges, researchers have been searching for alternatives to hydrophones to enhance the economic viability and reliability of traditional passive acoustic monitoring technology while enabling hands-off data retrieval and risk mitigation by avoiding the operational challenges associated with battery-powered devices, such as finite lifespan, replacement logistics, and potential reactivity when exposed²⁸. While cold-water environments do not universally degrade battery performance, shallow deployments are often constrained by biofouling, which can hinder sensitivity within days to weeks depending on biological activity. Optical fibers are well-suited for harsh environments and have been reliably used in oil and gas wells for years, where they endure high temperatures and pressures²⁹. This makes fiber-optic distributed acoustic sensing (DAS) a robust alternative for long-term monitoring in thermally challenging marine settings. As a result, DAS has emerged as a scalable and environmentally sustainable solution for underwater soundscape monitoring. DAS enables autonomous data collection and real-time coverage along the entire length of a standard single-mode optical fiber, which acts as a distributed array of sensing points. The DAS system operates as a distributed sensing platform based on phase-sensitive optical time-domain reflectometry ( $\varphi$ -OTDR), where each spatial segment of the fiber acts as a virtual sensing point. This enables high-resolution acoustic detection along the fiber length without requiring discrete sensors. For further details on DAS principles, readers are referred to^29,30. While the interrogator unit represents a significant capital investment, the fiber itself is inexpensive and widely available. Furthermore, a single interrogator can be shared across multiple monitoring sites using optical switches and time-division multiplexing³¹, allowing efficient monitoring of multiple areas from a central unit. This architecture reduces the cost per sensing point and simplifies deployment logistics, particularly when existing fiber infrastructure is available.

It is important to note that DAS deployment geometry must be adapted to the target ecosystem for ecological studies. The current deployment was designed for technical validation, and future applications in coral reef environments would require customized layouts, based on reef depth and morphology, to match ecological monitoring goals. The greatest advantage for integrating DAS as a potential cornerstone of oceanic observational infrastructure lies in the abundance of deployed optical fibers in the ocean, facilitating implementation costs and extended sensing coverage. While existing submarine cables rarely intersect shallow coral reef ecosystems, recent studies have demonstrated successful DAS deployments in nearshore environments. For example, Harmon et al.³² demonstrated DAS monitoring using a shallow-water energy cable deployed at the European Marine Energy Center (EMEC) in Orkney, UK, which includes tidal and wind-powered infrastructure in nearshore settings. These examples highlight the feasibility of DAS in shallow water and suggest future opportunities for reef-specific deployments, contingent on ecological compatibility and tailored deployment strategies. Notably, submarine optical fiber cables form the foundation of international and intercontinental telecommunications, with their deployed length steadily increasing to meet the rising need for internet and mobile services, now exceeding a million kilometers^33,34. Additionally, optical fibers are highly reliable in harsh environmental conditions³⁵, making them well-suited for long-term deployment in these challenging settings.

Hence, existing works in the literature have explored repurposing existing subsea telecommunication cables to observe various oceanic events. These include tracing the routes followed by blue whales traveling through Nordic fjords³⁶, as well as overseeing shipping activities³⁷, geological events like earthquakes^38,39, and meteorological phenomena such as storms and typhoons^37,40. In addition to this, fiber-optic sensing technologies like DAS are highly versatile and allow room for integrating various concurrent functionalities. For that reason, researchers have leveraged existing multiplexing technologies, originally envisioned to enlarge the capacity of fiber-optic communication networks⁴¹, to enable the coexistence of various signals corresponding to different fiber-optic applications. Some of the works done in this field include simultaneous distributed vibration and temperature monitoring over the same fiber⁴². Additionally, integrated sensing and communication (ISAC) systems incorporating DAS and optical communications have been developed^{43, 44–45}. Furthermore, comprehensive fiber-optic miscellaneous networks have been created, incorporating a wide range of functionalities such as vibration and temperature sensing, wired and wireless optical communications, power delivery, and energy harvesting from untapped optical signals^46,47. In terms of ocean soundscape monitoring, fiber-optic DAS has mainly been used in the literature for deep-sea applications, while DAS for reef monitoring has not yet been explored.

Fig. 1 [Images not available. See PDF.]

Illustration of fiber-optic DAS and hydrophones deployed in a reef ecosystem. This conceptual diagram applies to both reef and coral reef environments. The current study was conducted in a degraded inshore reef near the shoreline of KAUST (Thuwal, Saudi Arabia), with scattered boulders and encrusting sponges. Inset: Optical fiber stripped of its protective jacket. Acoustic signals from reef inhabitants cause perturbations in the fiber, resulting in variations in the backscattered Rayleigh signal. Note: This figure is schematic and intended to illustrate the concept of DAS deployment in reef environments. It does not represent the exact geometry of the deployed cable.

Given the ecological significance of reef environments and the advantages of fiber-optic DAS in terms of performance and cost-effectiveness, this study demonstrates the viability of DAS as a key tool for marine reef health monitoring. We aim to demonstrate DAS’s capability as a distributed soundscape monitoring system that can provide high-resolution one-dimensional spatial mapping of reef activities along the fiber axis. Figure 1 illustrates our deployment concept and operational principles of fiber-optic DAS in a marine reef ecosystem. Unlike hydrophones, which function as point sensors requiring manual intervention for data retrieval and battery maintenance, optical fibers form a continuous, distributed sensor array, enabling large-area reef monitoring at lower costs. Acoustic signals from reef inhabitants, such as snapping shrimp and fish vocalizations, and anthropogenic activities, like diver movements, induce perturbations in the fiber deployed along the reef. These perturbations modulate the backscattered Rayleigh signal (as shown in the inset of Fig. 1), enabling real-time, spatially resolved acoustic monitoring with data accessible onshore for post-processing. We deployed an optical fiber cable in a reef area in the Red Sea alongside a battery-powered hydrophone as a reference to validate this concept. This dual-sensor configuration allowed a systematic comparison between DAS-recorded soundscapes and conventional hydrophone measurements, focusing on key reef acoustic events, including snapping shrimp sounds, fish vocalizations, and diver activity. This study does not focus on biodiversity metrics, cataloging, species discovery, or conservation outcomes. Instead, it introduces a new application of fiber-optic DAS as an acoustic sensor for monitoring reef environments, demonstrated through three case studies: snapping shrimp acoustic activity, fish vocalizations, and diver tracking. These examples validate the system’s real-time monitoring capabilities, with future relevance to conservation applications.

Results

Following our experimental approach described in Fig. 1, we immersed a 50-m section of a 1-km-long fiber-optic cable in a reef region located in the central Red Sea along the western coast of Saudi Arabia. The reef site is located near the shoreline of KAUST (Thuwal, Saudi Arabia) and is characterized by sand, macroalgae, scattered boulders, and encrusting sponges. While not a dense coral reef, the habitat supports snapping shrimp, other invertebrates, and abundant fish, making it a suitable proxy for coral reef monitoring. The cable descended from the shore to a depth of almost 15 m at its far end. This section functioned as a hydroacoustic sensor for reef monitoring. The fiber-optic cable remained submerged in the reef area, with one end connected to a DAS interrogator unit near the shore, which continuously monitored acoustic signals along the fiber with a spatial sampling of $\sim$ 0.5 m. For comparison, we also deployed a hydrophone in the same reef area, positioned approximately 1 m from the immersed fiber section. Further details on the georeferencing of the cable and the equipment utilized are provided in the “Materials and methods” section. Here, we present the acoustic signals recorded by the hydrophone and fiber-optic DAS, comparing their capabilities to detect, locate, and differentiate various acoustic signals in the reef environment.

Hydrophone sound collection and analysis

Fig. 2 [Images not available. See PDF.]

(a) Long-term averaged spectrogram of the pressure spectral density recorded from the Red Sea reef using a hydrophone over 48 hours. A photoperiod plot shows night hours between sunset and sunrise, and midnight when shore lights were switched off. (b) Average pressure spectral density filtered over 1.5 to 5 kHz, showing cyclic snapping shrimp activity with peaks at 06:00 and dips at 16:00.

Figure 2a presents the long-term averaged spectrogram of the pressure spectral density. The results were recorded from the Red Sea reef using the battery-powered hydrophone over 48 hours and processed with the CHORUS toolbox⁴⁸. The averaging time was set at 0.08333 hours. A photoperiod bar indicating the night (7 PM–6 AM) and day (6 AM–7 PM) periods is shown above the spectrogram. In this experiment, midnight corresponds to the time of switching off the lights around the reef area (i.e., onshore). Observing the light intensity within the reef is important since it directly impacts the activities of snapping shrimp and fish within the reef⁴⁹. The results of Fig. 2a show the presence of acoustic signals across different frequency components, revealing the diversity of marine life activity, including vocalizations, movements, feeding behaviors, and diurnal variation patterns over time¹⁰. Notably, two distinct frequency bands are observed: (i) 1.5–20 kHz, associated with shrimp sounds and movements⁴⁹, and (ii) 100 Hz - 1000 Hz, associated with fish communication and schooling behavior^{50, 51–52}. Shrimp acoustic activity exhibits a cyclic pattern, with peak activity around sunrise (06:00). In contrast, fish acoustic activity follows an inverse trend to that of shrimp, starting acoustic activities at sunrise (06:00) and continuing until midnight (00:00), coinciding with turning off the shore lights (Fig. 2a). This anti-correlation implies light-driven behavioral dynamics, where both natural light cycles and anthropogenic illumination influence fish activity, allowing shrimp-generated acoustics to dominate during dark-adapted periods⁴⁹. Since the sound of snapping shrimp is a feature that distinguishes a healthy reef^{10, 11, 12–13} and is unique and easily recognizable, we provide additional analysis of it. Snapping shrimp produce broadband impulsive sounds with significant energy dominating the frequency range (>1.5 kHz), with peak frequencies typically between 2 kHz and 5 kHz^19,49. Figure 2b presents the average pressure spectral density filtered over the 1.5 to 5 kHz range, where the snaps’ peak frequencies exist, over the same 48-hour period as Fig. 2a. The results of Fig. 2b highlight the cyclic behavior of the sounds of the snapping shrimp, with activity increasing from 18:00 and reaching a peak around 06:00, followed by low activity between 11:00 and 16:00. This cyclic behavior can be attributed to their response to light availability and diurnal patterns. Snapping shrimp exhibit increased acoustic activity during periods of low light, such as dawn and dusk, which aligns with their crepuscular behavior. This pattern is likely driven by their need to avoid predators and optimize foraging during times when they are less visible¹⁹. For comparison with the DAS system, we focused on the 1.5-5 kHz range, which contains the dominant energy of snapping shrimp sounds and is well-covered by both systems.

Fig. 3 [Images not available. See PDF.]

Spectrograms of DAS recordings over a 2-hour period. (a) Underwater section of the fiber near the hydrophone, showing snapping shrimp activity. (b) Onshore section of the fiber, insulated from environmental noise, showing minimal acoustic activity. (c) and (d) Zoomed-in views of the first and last 20 min of the underwater DAS spectrogram, respectively. The DAS spectrogram appears quieter than the hydrophone due to differences in sensitivity. (e) Averaged and normalized power spectral density (PSD) of DAS recordings every 20 min for underwater (blue) and onshore (orange) sections.

Detection of snapping shrimp sounds using fiber-optic DAS

Building on the hydrophone-observed patterns of shrimp sounds, we further explored the capability of the fiber-optic DAS to observe the same behavior. Figure 3a shows the spectrogram of a 2-hour DAS recording from a point on the deployed fiber located at a depth of $\sim$ 5 m underwater, near the hydrophone that recorded the data in Fig. 2. This DAS recording spans approximately one hour before and one hour after midnight, encompassing the period when the lights were turned off around the reef area. Additionally, Fig. 3b presents the spectrogram of the same 2-hour DAS recording but from a point on the fiber positioned deep within the spool located onshore. This point was selected to intentionally minimize the impact of environmental noise; however, we cannot perfectly isolate it. Some random noise onshore, such as wind gusts, vehicle engines, and other human activities, can still couple into the fiber and affect the recorded signal. This onshore recording serves as a reference, showing quiet acoustic activity compared to the underwater recording in Fig. 3a. Both graphs are plotted within the 1.5 to 5 kHz range, which corresponds to the peak frequencies of the snaps, and we applied a time-adaptive spectral notch filter to discard harmonic noises produced by the electronics of the DAS system at such high frequencies ("Materials and methods” section). The results of Fig. 3a confirm the ability of fiber-optic DAS to record the distinctive snapping shrimp sounds, as compared to the quiet environment onshore (Fig. 3b).

Figures 3c and d are zoomed-in views of the first and last 20 min of the underwater DAS spectrogram data, respectively (labeled with red and white rectangles in Fig. 3a). These figures provide detailed insights into the behavior of snapping shrimp when the light was on and off, correspondingly. Comparing the spectrogram data in Fig. 3c and d, the snapping sounds become more prominent in the absence of light, which aligns with the hydrophone results. To further quantify and highlight this observation, we averaged the power spectral density (PSD) of the DAS measurements presented in Figs. 3a and b over 20-min intervals for the points along the fiber located underwater and onshore, respectively. Figure 3e presents the averaged and normalized PSD for the underwater point (blue bars). The data reveals a gradual increase, particularly after the lights are turned off, with an obvious rise during the final 20 min of the two-hour recording session. This trend of increasing shrimp sounds and activities post lights-off aligns with the hydrophone data analysis. Conversely, the averaged and normalized PSD for the onshore fiber point (orange bars) fluctuates randomly with lower values compared to the underwater point, indicating minimal sound activity onshore throughout the recorded period.

Fig. 4 [Images not available. See PDF.]

Examples of snapping shrimp activity detected by fiber-optic DAS. (a) Temporal representation of snapping shrimp sounds at a single location. (b) Corresponding frequency-domain representation, showing snapping shrimp signatures in the 1.5 to 8 kHz range. (c) Time-domain signals of two distinct acoustic events detected simultaneously at different locations along the underwater fiber section. (d) Corresponding spectral representation for the two locations.

We provided two representative examples, depicted as Ex. 1 and Ex. 2, to illustrate the temporal and spectral signals of the snapping shrimp’s sound as recorded by the fiber-optic DAS (Fig. 4). Unlike hydrophones, which capture localized sounds or aggregate sounds over a broad area without precise localization, DAS can spatially distinguish sounds along the submerged fiber. In Fig. 4, the distance is measured from the fiber’s entry into the water. We observed that starting at a distance of $\sim$ 5 m, near the shore and close to the base of the rocks, shrimp were more active in the reef. Figure 4a (Ex. 1) shows the temporal sound detected at a single location, while Fig. 4b displays the fast Fourier transform (FFT) of this sound, revealing snapping shrimp signatures in the 1.5–8 kHz range at a distance of $\sim$ 6–7 m. Additionally, Fig. 4c (Ex. 2) shows the time-domain signals of two distinct acoustic events detected simultaneously at different locations along the underwater fiber section: one at $\sim$ 6–7 m and the other at $\sim$ 8–9 m. Figure 4d presents the FFT of these sounds, which clearly shows two separate snapping shrimp signatures.

Both examples correspond to the same section of fiber (5–10 m), with Ex. 1 and Ex. 2 representing snapshots separated by only a short time interval. In Ex. 1, shrimp sounds are detected between 6-7 m, while a few seconds later in Ex. 2, shrimp sounds are detected both between 6-7 m and 8-9 m. This demonstrates the DAS system’s ability to detect snapping shrimp sounds and resolve their one-dimensional spatial distribution along the fiber axis. While the DAS can identify the location of peak acoustic energy along the cable, this does not constitute full localization in the directional or three-dimensional sense. Rather, it provides one-dimensional spatial mapping along the fiber axis. Full localization, including direction of arrival, typically requires advanced techniques such as beamforming⁵³, triangulation, or parallel fibers⁵⁴, which were not applied in this study. Supplementary Video V1https://drive.google.com/file/d/1wq-ckVDXBD70DssA787q2rkTc6o9eRie/view?usp=sharing, recorded near the fiber deployment site, shows snapping shrimp behavior near its burrow and helps contextualize the impulsive acoustic signals observed in the DAS recordings.

Detection and monitoring of fish activity using fiber-optic DAS

While fiber-optic DAS effectively resolved the spatiotemporal dynamics of high-frequency shrimp sounds, its utility can extend to characterizing low-frequency acoustic signatures associated with fish behavior. Building on the hydrophone-derived evidence of fish vocalizations and schooling patterns (Fig. 2a), DAS is leveraged to localize and temporally dissect fish-generated sounds within the same nearshore habitat. Figure 5a–d presents four representative spectrograms, labeled as Ex. 3-Ex. 6, of fish sounds captured at different times from the fiber section near the shore and rocks (i.e., recorded at a distance of 7 m), each covering a 1.5-s period. All spectrograms include some of the common acoustic signatures of fish vocalizations and behavior.

Some of the more commonly observed fish vocalizations in the low-frequency range are loosely classified into four general categories: drums, impulses, roars, and quacks, though these labels are subjective and don’t capture the full diversity of sounds across species. In the literature, drums are defined as periodic pulse trains, impulses as short-duration pulses, roars as wideband longer-duration signals, and quacks as short harmonic signals^55,56. As shown in Figs. 5a–d, a drum is a low-frequency acoustic signal that consists of periodic pulse trains, produced by fish when vibrating their swim bladder or by contracting specialized muscles against the swim bladder^55,57. Additionally, drums may be preceded or followed by knocks or impulses, which often have slightly higher frequencies than the primary drum signal (e.g., in Figs. 5a, 5b, and 5d)⁵⁵. Despite these acoustic observations, identifying the exact fish species responsible for these sounds remains challenging, as fish vocalizations in the Red Sea are an understudied topic. However, based on similarities to spectrographic patterns reported in the literature^55,56,58, two possible candidates for these vocalizations are the Lutjanus bohar (Bohar snapper) and Lutjanus argentimaculatus (Mangrove red snapper). Both species are common in the Red Sea and were frequently observed near the deployed fiber and photographed by scuba divers (see the picture (S1)). Nevertheless, species identification was not confirmed through direct visual methods and should therefore be considered likely but not definitive. As with many PAM studies, detailed information on sound production specific to the species and region is not yet available⁵⁹.

To highlight the comprehensive monitoring capabilities of the fiber-optic DAS in detecting and localizing fish activities, we provide more details on the drumming period of the first example of fish sound (Ex. 3, Fig. 5a). Over the length of the deployed fiber section near the shore (i.e., distance = 5 m to 10 m), Fig. 5e shows the spatiotemporal signal, while Fig. 5f presents the corresponding FFT of the drumming acoustic event of Ex. 3. Both figures show significant variation in the fish sounds along the fiber section, with acoustic energy concentrated between $\sim$ 6 m and $\sim$ 8 m. This pattern is interpreted as simultaneous or slightly delayed signals of varying strength originating from different points within that region. While the DAS system enables spatial mapping along the fiber, it does not resolve lateral or depth coordinates from this data.

Fig. 5 [Images not available. See PDF.]

Representative spectrograms of fish sounds captured by fiber-optic DAS. (a–d) Spectrograms labeled as Ex. 3-Ex. 6, showing various fish vocalizations and behaviors. (e) Spatiotemporal signal of the drumming acoustic event from Ex. 3 along the fiber section near the shore. (f) Corresponding FFT of the drumming acoustic event. Note: This figure focuses on low-frequency fish vocalizations (<1 kHz). The high-frequency snapping shrimp sounds are presented in Figs. 2, 3 and 4.

Detection and tracking of diver activity using fiber-optic DAS

Following the detection of snapping shrimp and fish vocalizations, this section shows the capability of the DAS system to detect and track human activity in underwater reef environments. Figure 6a illustrates the trajectory (black arrows) of a diver along the 50-m deployed fiber section. During the DAS recording, the diver’s movement, trajectory, and speed were designed to mimic common human activities in reef environments, such as snorkeling, scuba diving, and underwater photography. The diver completed two full laps, swimming away from the starting point near the shore twice (i.e., from Point 1 to Point 2 and from Point 3 to Point 4) and returning twice (i.e., from Point 2 to Point 3 and from Point 4 to Point 5).

Figure 6a also includes the location of the hydrophone, positioned near the shore. The diver passed three times over the hydrophone’s region, specifically in the vicinity of Points 1, 3, and 5. Figure 6b presents the spectrogram of the hydrophone recording, capturing the acoustic signals generated during the diver’s activity as the diver passed by the hydrophone. The spectrogram reveals three distinct time intervals during the diver’s movement where the acoustic activity captured by the hydrophone is most pronounced, particularly in the frequency range below 1 kHz. These periods of heightened acoustic signals correspond precisely to the moments when the diver was near the shore, in close proximity to the hydrophone, specifically at Points 1, 3, and 5 (marked in Fig. 6b).

However, in contrast to the hydrophone, which records the diver’s movement signatures only at fixed locations (i.e., within its coverage area), the DAS system delivers a comprehensive spatiotemporal representation of the diver’s movement dynamics along the entire deployed fiber. This allows for continuous tracking of the diver’s trajectory. Figure 6c displays a 40-60 Hz bandpass-filtered spatiotemporal graph generated from DAS recordings during the diver’s trip, depicting the movement pattern along the immersed fiber section. The DAS system precisely maps the diver’s trip, capturing the complete trajectory during the two full laps along the fiber and providing information on the diver’s movement pattern and speed. In particular, the diver completed the first lap in $\sim$ 7 min, while the second lap took $\sim$ 8-9 min, highlighting the advantages of DAS over hydrophone. While the DAS system enables continuous trajectory tracking along the fiber, it does not provide directional or depth information. In contrast, a hydrophone can detect diver-generated sounds with higher sensitivity when the diver is nearby, but lacks distributed spatial coverage. Each system offers complementary strengths depending on the monitoring objective.

Fig. 6 [Images not available. See PDF.]

Diver detection and tracking using fiber-optic DAS. (a) Diver’s trajectory along the 50-m fiber section, including hydrophone location. (b) Hydrophone spectrogram showing acoustic activity at Points 1, 3, and 5. (c) 40-60 Hz bandpass-filtered spatiotemporal graph from DAS recordings, depicting the diver’s movement pattern.

Discussion

This study presents a technical proof-of-concept for using fiber-optic DAS to monitor underwater acoustic activity in shallow reef environments. The primary objective was to validate the feasibility of DAS for distributed soundscape detection using available equipment. The hydrophone served as a reference to confirm DAS signal acquisition, not as part of a comprehensive sensor comparison. While the study references coral reef applications, the deployment was conducted in a reef area with limited coral presence, and conservation relevance is discussed as a future direction. The results of our study demonstrate the potential of fiber-optic DAS for real-time monitoring in shallow reef environments. While the current deployment was conducted in a degraded inshore reef, the demonstrated capabilities are directly applicable to coral reef ecosystems, which represent a critical future direction for this technology.

However, this study does not assess coral reef health or ecological resilience directly. The deployment was conducted in a coastal reef habitat with limited coral presence, and the acoustic measurements presented here serve as a technical demonstration of DAS capabilities. Future ecological applications would require integration with biological indicators and conservation frameworks to evaluate reef health. The DAS system effectively detected a range of acoustic activities within the reef environment, including snapping shrimp sounds, which were consistent with the data obtained from the conventional hydrophone. For fish vocalizations, DAS recordings were cross-referenced with the FishSounds database to aid in species-specific identification. However, species attribution based solely on acoustic signatures has inherent limitations, and future work should integrate acoustic data with visual observations or tagging-based validation to strengthen identification confidence. This validation underscores the reliability of DAS technology in capturing detailed acoustic profiles of reef ecosystems. Notably, the DAS system’s ability to track diver movements further highlights its versatility and applicability in monitoring human activities within reef areas.

Fiber-optic DAS technology offers several advantages over traditional hydrophones. Its capacity for high-resolution spatial mapping of reef soundscapes allows for a comprehensive understanding of the spatial distribution of acoustic events, which is crucial for assessing reef health. DAS can be used to monitor the acoustic health of a large reef ecosystem and easily regulate anthropogenic stressors. Additionally, DAS is cost-effective and environmentally sustainable, as it eliminates the need for batteries and manual data retrieval, reducing both operational costs and environmental risks. The real-time monitoring capability of DAS facilitates timely responses to changes in reef health, making it a valuable tool for conservation efforts. While the demonstrated capabilities are relevant to conservation monitoring, this study does not evaluate ecological indicators or conservation outcomes. Future work may build on this technical foundation to support biodiversity assessments and habitat resilience studies.

However, DAS technology also has its disadvantages. Hydrophones are an established technology with highly developed software, so they retain their usefulness. DAS systems generate large amounts of data, requiring substantial storage space for recording, which can be a logistical challenge. The complexity in processing and analyzing these large datasets also poses a significant challenge, requiring advanced data management and processing capabilities. Despite these drawbacks, the advantages of DAS in providing extensive coverage and real-time data collection make it a promising tool for coral reef monitoring.

In addition to data volume and processing complexity, DAS systems exhibit sensitivity limitations that vary with frequency, deployment geometry, and environmental conditions. These include possible sensitivity degradation due to biofouling in shallow deployments, and potential mechanical coupling challenges, which require detailed future analysis and investigation. In high-biodiversity reef environments, biofouling can begin within days and become substantial within a week, leading to dampened acoustic signals and reduced sensitivity. Cleaning procedures and anti-biofouling coatings are potential mitigation strategies, but further research is needed to quantify their effectiveness. DAS systems are also susceptible to electronic noise, especially at higher frequencies⁶⁰. Compared to hydrophones, this can reduce the detectability of certain acoustic events. Although DAS systems offer distributed sensing and continuous data access, their cost-effectiveness depends on the deployment scenario. The interrogator unit involves a high initial cost, which may be prohibitive for small-scale or short-term monitoring. However, when deployed over long distances or across multiple sites using optical switching and time-division multiplexing, DAS becomes economically viable. The use of standard, low-cost optical fibers and the ability to repurpose existing infrastructure further enhance its practicality for large-scale environmental monitoring. Unlike point-based sensors that require manual retrieval for data access or battery replacement, DAS enables remote and continuous monitoring along the entire fiber length. This distinction becomes more relevant in deployments where diver access is logistically intensive or where long-range sensing is needed. While diver-based retrieval is feasible and widely practiced, DAS offers a scalable alternative for distributed sensing in complex or extended reef environments.

Additionally, DAS deployment geometry plays a critical role in ecological monitoring. The current study used a straight-line layout for technical validation, but future ecological applications would benefit from tailored configurations that match reef morphology and investigate biological activity hot spots. In addition, ecological risks such as coral entanglement must be considered. Fiber-optic cables, if laid without spatial planning, may interact with soft coral structures in ways similar to fishing lines or anchors. Future deployments should incorporate ecological mapping and mechanical safeguards (e.g., weights or substrate attachments) to minimize physical disturbance. Besides, DAS sensitivity variation along the fiber length is a known limitation in DAS systems. Despite these fluctuations, it is generally recommended to maintain a signal-to-noise ratio (SNR) greater than 2 dB, which can be reliably achieved over kilometer-scale distances with proper system design⁶⁰. Additionally, environmental dependencies such as depth and salinity, as well as jacket thickness were not modeled in this study and should be considered in future ecological deployments.

Regarding the SNR, reported values from the literature highlight the contrast between fiber-optic DAS and PAM systems, including hydrophones. Underwater DAS deployments have demonstrated highly variable SNR performance, with improvements of up to  40 dB achievable through the use of modified fibers, interrogators, or sensing structures. However, these outcomes are strongly dependent on deployment depth, cable design, and interrogator configuration^{61, 62, 63–64}. PAM units, by comparison, consistently achieve SNR values in the range of  30-60 dB, depending on sensor design and ambient noise conditions^{65, 66–67}. Direct numerical comparisons between DAS and PAM systems are inherently difficult due to their fundamentally different sensing mechanisms and the complexity of reef environments. In our deployment, the “background” snapping shrimp soundscape was itself the biological signal of interest, present continuously and at unknown distances. This makes conventional SNR calculations less meaningful and prevents a one-to-one comparison between our colocated hydrophone and DAS recordings.

The work of distributed fiber-optic sensing to monitor the health of coral reefs can be extended to address other critical environmental factors. While coral reef environments typically maintain relatively stable temperature profiles, environmental temperature fluctuations, such as those associated with marine heatwaves⁶⁸, are a primary cause of coral bleaching. DAS-DTS systems may be useful for detecting subtle thermal anomalies or long-term trends that could signal ecological stress. Optical fibers can offer distributed temperature sensing (DTS)⁶⁹ and hybrid DAS-DTS systems⁷⁰, providing high reliability in harsh environments. This capability allows for comprehensive monitoring of both acoustic and thermal conditions, enhancing our ability to detect and respond to threats to coral health. Additionally, identifying the marine species responsible for various sounds remains a challenge. Integrating artificial intelligence (AI) with DAS technology can significantly improve the accuracy and efficiency of species identification. AI algorithms can analyze the complex acoustic data collected by DAS⁶⁰, enabling automated classification and monitoring of marine life. These advancements in DAS technology, combined with AI, hold great promise for expanding the scope and effectiveness of coral reef monitoring and conservation efforts.

In conclusion, our study demonstrates the effectiveness of fiber-optic DAS technology for coral reef monitoring. The high-resolution spatial mapping and real-time data collection capabilities of DAS offer a comprehensive and cost-effective solution for continuous reef monitoring. In summary, this work demonstrates the technical viability of DAS for spatially resolved acoustic monitoring in reef environments. The system successfully detected biological and anthropogenic signals, including snapping shrimp activity, fish vocalizations, and diver movement. Future studies should expand on ecological validation, species attribution, and stakeholder engagement to support conservation-oriented deployments. These findings significantly contribute to coral reef conservation efforts by providing a reliable method for assessing reef health and facilitating proactive measures to mitigate the harmful effects of anthropogenic noise. Continued research and innovation in this field are essential for advancing coral reef monitoring technologies and supporting global conservation initiatives.

Fig. 7 [Images not available. See PDF.]

Experimental area and deployment setup. (a) Geographic location of the reef site near the KAUST monument in the central Red Sea. Satellite imagery was created in QGIS version 3.36.0-Maidenhead, licensed under the GNU General Public License URL. (b) Positioning of the hydrophone and fiber-optic cable in the reef environment. (c) Observed marine biodiversity, including snapping shrimp and goby fish, near the deployed fiber.

Materials and methods

Experimental area geography and deployment methodology

Acoustic reef soundscape data were collected in the central Red Sea, near the shoreline along the western coast of Saudi Arabia, as shown in Fig. 7a. The study site was a reef located close to the KAUST monument (see the Map https://www.google.com/maps/place/King+Abdullah+Monument/%4022.3424548,39.0898705,13z/data=%214m10%211m2%212m1%211skaust+monuments%213m6%211s0x15c11dea8ff44b3d:0x3a675f89d9ad0e30%218m2%213d22.3423029%214d39.0898889%2115sCg9rYXVzdCBtb251bWVudHNaESIPa2F1c3QgbW9udW1lbnRzkgEGbXVzZXVtmgEjQ2haRFNVaE5NRzluUzBWSlEwRm5TVVF6ZWs5MVJWSkJFQUXgAQD6AQQIABA5%2116s%2Fg%2F11c37hvx51?entry=ttu%20&g_ep=EgoyMDI1MDIxMi4wIKXMDSoJLDEwMjExNDU1SAFQAw%3D%3D), where water depth is within a 15 m range. The selected site is a degraded inshore reef with limited coral coverage, including scattered boulders and encrusting sponges. The habitat supports snapping shrimp, and several reef-associated fish species (e.g., blackfin barracuda (Sphyraena qenie), mangrove red snapper (Lutjanus argentimaculatus), diamondfish (Monodactylus argenteus), as well as several species of Gobidae, Acanthuridae, Scaridae). These species are known, or likely to emit active or passive underwater sounds, according to the global sonifery index⁵⁶, making this site a suitable proxy for coral reef monitoring. We deployed a 70-m standard single-mode fiber (SMF) (GTFJU-2G652D, YOFC), with 50 m laid out in a straight line across the shallow reef area and the remaining 20 m coiled into loops at the cable’s distal end. However, slight deviations from a perfectly taut geometry may occur due to underwater conditions, which can introduce uncertainty in localization accuracy. To ensure high-quality data recordings, we excluded the DAS data from the 20-m loops to avoid noise caused by high-intensity reflections at the fiber end facet (i.e., Fresnel reflections)⁷¹. To stabilize the fiber against underwater currents, weights were spaced approximately every meter along its length. The fiber was tightly buffered and encased in an inner jacket made of polyvinyl chloride (PVC) and an outer jacket made of thermoplastic polyurethane (TPU), providing enhanced protection and durability in the reef environment.

During deployment, care was taken to avoid contact with live coral structures. In coral reef environments, especially those with soft coral formations, fiber-optic cables must be carefully managed to prevent entanglement or abrasion. Future DAS deployments should consider using cable guides, elevated suspension systems, or mapped pathways that avoid sensitive benthic habitats. Onshore, the portion of the fiber at the shoreline and the remaining unwound section from the spool were shielded with a PVC cover. This protection is guarded against mechanical damage and environmental noise near the rocky shoreline, such as wind, waves, and human activity, ensuring reliable data collection. The DAS system provides one-dimensional spatial mapping along the fiber axis. It does not resolve lateral or angular directionality, as the deployment geometry does not support multi-dimensional localization. Broader area-based spatial mapping would require multiple fibers or a defined deployment geometry, which was not part of this study.

To evaluate the performance of our fiber-optic DAS, we compared its detection fidelity with that of a conventional battery-powered hydrophone (SoundTrap ST300 HF, Ocean Instruments NZ). The DAS cable and hydrophone were deployed simultaneously in the same reef area, with the hydrophone positioned approximately 1 m from the submerged section of the fiber. This ensured that both systems recorded acoustic activity under identical environmental conditions, allowing for a valid comparison of their detection capabilities. Positioned $\sim$ 10 m from the fiber’s entry point into the water and about 1 m beside the fiber-optic cable, the hydrophone was anchored with weights and floated with a buoy (Fig. 7b). The selected reef site was rich in marine biodiversity, hosting species such as snapping shrimp and goby fish, which were observed and photographed near the deployed fiber (Fig. 7c).

Fig. 8 [Images not available. See PDF.]

Schematic diagram of the experimental DAS setup for underwater monitoring of reef soundscapes. A narrow-linewidth continuous-wave (CW) laser (green arrows) was modulated into 50 ns pulses at a 50 kHz repetition rate using an acousto-optic modulator (AOM). The pulses were amplified by a pulsed erbium-doped fiber amplifier (EDFA) and launched into a 1 km optical fiber via Circulator 1 (Cir. 1). The final 70 m of the fiber were deployed in the reef environment, while the remaining length remained onshore as a reference for background noise. Each pulse generated Rayleigh backscatter (red arrows) that returned toward the interrogator and was routed by Cir. 1. Because the backscattered signal is weak, it was amplified by a CW EDFA and filtered using a fiber Bragg grating (FBG) with Circulator 2 (Cir. 2) to suppress amplified spontaneous emission (ASE) noise. The cleaned signal was converted to an electrical signal by a photodetector operating at 200 MS/s, then digitized for analysis. Timing and intensity of the backscatter enabled precise localization of acoustic events along the fiber.

Design and characterization of the fiber-optic DAS system

In principle, a fiber-optic DAS system monitors the temporal dynamics of Rayleigh backscattering signal intensity along the entire length of an optical fiber. These signals arise from multiple light reflections caused by inhomogeneities within the optical fiber core, which are inherent to the manufacturing process²⁹. In Rayleigh-based backscattering, energy is not transferred to the glass, resulting in frequency matching between the incident and scattered signals (elastic backscattering). When the optical fiber is undisturbed by external mechanical sources (e.g., acoustic events or vibrations), the intensity of the Rayleigh signal remains ideally constant over time. However, when an acoustic event, such as marine wildlife biophony or human activity, occurs near the optical fiber, the phase of the Rayleigh signal is modulated in response to the mechanical waves produced by the interaction of sound with the fiber⁷². To detect and identify acoustic signals, we use our in-house DAS system based on direct detection²⁹. This approach involves real-time monitoring of intensity fluctuations in the Rayleigh backscattering signals. These fluctuations result from phase modulations that encode the temporal and spectral patterns of detected sounds along the fiber under test⁷³.

The fiber-optic DAS unit was designed using phase-sensitive optical time-domain reflectometry ( $\mathrm{\Phi }$ –OTDR)⁷³. The experimental setup of the DAS is illustrated in Fig. 8, where a narrow linewidth laser generated continuous wave (CW) light at a wavelength of 1550.12 nm and an optical power of 40 mW. Using an acousto-optic modulator (AOM), the CW light was converted into optical pulses with a duration of 50 ns (i.e., a 5-m spatial resolution) and a repetition rate of 50 kHz. We selected a relatively high repetition rate to effectively record the high-frequency components of the snapping shrimp sounds. The pulses were then amplified using a pulsed erbium-doped fiber amplifier (EDFA) and launched through a circulator (Cir. 1) into a 1 km long optical fiber. The last 70 m of the optical fiber was deployed in the reef, as mentioned earlier, while the remaining fiber was onshore, serving as a reference for measuring the background noise of the DAS system. Each injected pulse induced the formation of a Rayleigh backscattering trace (Fig. 8), which propagated back toward the interrogation side. These backscattered signals were directed through Cir. 1 to separate them from the incoming pulses. Due to their inherently weak intensity, the backscattered signals underwent a second amplification stage using a CW EDFA. Next, amplified spontaneous emission (ASE) noise associated with the EDFAs was filtered out using a fiber Bragg grating (FBG) and a second circulator (Cir. 2). Finally, the cleaned signal was converted into the electrical domain using a photodetector. The resulting electrical signal was then logged and processed by a digitizer for further analysis. This process enables precise localization of acoustic events along the fiber by analyzing the timing and differential intensity of the backscattered signals^69,73.

Before installing our DAS system at the reef deployment site, we performed a brief characterization experiment using a standard 1-km-long SMF as the fiber under test. In this experiment, we wound a 10-m fiber section in air, located at a $\sim$ 890-m distance from the fiber’s input port, around a piezoelectric transducer (PZT). The vibration frequency of the PZT was controlled by electrical sinusoidal signals emitted by a waveform generator. We tested the detection accuracy of our DAS system with two different frequencies: 500 Hz and 800 Hz. Figure 9 shows the resulting detections achieved by the DAS interrogation unit. Specifically, Figs. 9a and 9c depict the spatiotemporal heatmaps produced through the implemented signal processing framework for the 500 Hz and 800 Hz vibrations, respectively. As expected, the entire optical fiber remains calm except for the vibrations detected at the PZT’s location ( $\sim$ 890 m). Additionally, Fig. 9b and d highlight the detected frequencies of the PZT vibration signals, which match precisely with the frequencies produced by the waveform generator. These results confirm the accuracy and reliability of our DAS system for detecting specific acoustic events. The peak observed at 600 Hz in the 800 Hz test signal in Fig. 9d is attributed to harmonic distortion arising from the direct detection method used in our DAS system²⁹. More advanced DAS systems, such as those based on phase demodulation or coherent detection, can suppress such harmonics effectively⁷⁴.

Fig. 9 [Images not available. See PDF.]

Characterization results of the DAS system using a PZT at two different frequencies. (a) and (c) Spatiotemporal heatmaps for 500 Hz and 800 Hz vibrations, respectively. (b) and (d) Detected frequency spectra matching the waveform generator’s output for 500 Hz and 800 Hz, respectively.

Signal processing and data acquisition in DAS

The digitizer in the DAS sensing unit operated in streaming mode, acquiring the Rayleigh signal continuously over time without gaps. The acquired Rayleigh traces needed to be precisely aligned and segmented to ensure accurate signal processing and data interpretation. This can be facilitated by calculating the total number of data points generated for each optical pulse. This is estimated by the ratio of the digitizer’s sampling rate to the pulse repetition rate. In our DAS unit, the digitizer operated at a 200 MS/s sampling rate, while the pulse repetition rate was 50 kHz. Thus, for each optical pulse, the digitizer acquired 4000 data points, representing the spatial dimension in the DAS data. For the temporal dimension, the DAS system collects 50,000 Rayleigh traces per second, given the 50 kHz repetition rate. Considering the calculated spatiotemporal dimensions, the digitizer’s streamed data can be reshaped and become ready for signal processing.

Considering the Nyquist sampling theorem and the pulse repetition rate used, the fiber-optic DAS can detect sounds with a broad frequency range of up to 25 kHz, which is convenient for monitoring various marine biological activities, such as shrimp snaps, acoustically visualized as short-time pulses over a broadband frequency range⁷⁵. Thus, besides the significant advantage of DAS for offering distributed sensing, the frequency range the DAS can detect is not far from what conventional hydrophones offer. Particularly, the hydrophone utilized in this experiment had a sampling rate of 72 kHz, allowing the detection of sounds with frequencies of up to 36 kHz.

Under normal conditions, without disturbances along the optical fiber, the temporal intensity of the Rayleigh backscattered signals remains stable. However, when an acoustic event affects a specific point on the fiber, it induces intensity fluctuations in the backscattered light at that location, reflecting the temporal patterns of the acoustic signal. This causes variations in the Rayleigh signal intensity at that particular point along the fiber. To accurately locate the acoustic events along the fiber and produce the spatiotemporal DAS map, we used the normalized differential method described in⁷⁶. This normalization technique not only facilitates spatiotemporal data representation but also enables the study of frequency components of any particular signal captured along the optical fiber by applying FFT to the normalized differential traces acquired within a given time period.

It is worth highlighting that when we pushed the DAS unit to record the high-frequency components of the snapping shrimp sounds, a high-frequency harmonic noise appeared in the PSD. This noise was produced by an electronic component within the unit and, fortunately, occurred at a single frequency that slightly shifted over time with changes in the unit’s internal temperature. This noise was evident throughout the entire fiber length, encompassing sections both onshore and underwater. To resolve this challenge, we apply a time-adaptive spectral notch filter to continuously discard the harmonic noise that slowly shifts within the 4050 to 4575 Hz frequency range. Removing such narrow spectral harmonic noise does not impact our analysis of the broad-band spectrum snapping shrimp sound. Besides, DAS sensitivity is influenced by several factors, including the type of optical fiber, the jacket material, the mechanical coupling to the environment, and the design of the interrogator unit. A comprehensive sensitivity characterization across different configurations was beyond the scope of this proof-of-concept study. Future work should include sensitivity curve measurements to support biodiversity-focused deployments and improve system calibration.

Acknowledgements

The authors would like to thank the divers from KAUST Coastal and Marine Resources for the installation of the hydrophone and fiber-optic cable. This work was supported by the NEOM Ocean Science and Solutions Applied Research Institute (OSSARI) (RGC/3/6001-01-01), KAUST Coral Restoration Initiative (REI/1/6081-01-01), Transition Award in Semiconductors (FCC/1/5939-04-01), and KAUST baseline funding (BAS/1/1614-01-01).

Author contributions

T.A. and M.N.H. are equally contributing first authors of this paper. I.A., C.H.K., A.R., T.A., and M.N.H. designed the experiments and prepared the field setup. I.A., C.H.K., T.A., and A.R. collected the data. I.A., C.H.K., T.A., A.R., and M.N.H. contributed to the data interpretation and discussion of the results. I.A., C.H.K., T.A., A.R., M.N.H., and J.M.M. contributed to the writing and revision of the manuscript. T.K.N., C.M.D., and B.S.O. supervised the project. Correspondence and requests for data should be addressed to I.A., C.M.D, or B.S.O.

Funding

This work was supported by the NEOM Ocean Science and Solutions Applied Research Institute (OSSARI) (RGC/3/6001-01-01), KAUST Coral Restoration Initiative (REI/1/6081-01-01), Transition Award in Semiconductors (FCC/1/5939-04-01), and KAUST baseline funding (BAS/1/1614-01-01).

Data availability

The data supporting the findings of this study are available from the corresponding authors upon reasonable request

Declarations

Competing interests

The authors declare no competing interests.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1038/s41598-025-30200-4.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Sobha, T., Vibija, C., & Fahima, P. Coral reef: A hot spot of marine biodiversity. In: Conservation and Sustainable Utilization of Bioresources, pp. 171–194. Springer, Singapore (2023)

2. Programme, U.E. Coral Reefs. Accessed: 17 September 2025 (2020). https://www.unep.org/topics/ocean-seas-and-coasts/blue-ecosystems/coral-reefs

3. Lachs, L. & Oñate-Casado, J. Fisheries and tourism: Social, economic, and ecological trade-offs in coral reef systems. In Youmares 9-the Oceans: Our Research 243–260 (Springer, Oldenburg, Germany, 2020).

4. Küfeoğlu, S. SDG-14: Life Below Water, pp. 453–468. Springer, Cham (2022). https://doi.org/10.1007/978-3-031-07127-0_16

5. Souter, D., Planes, S., Wicquart, J., Logan, M., Obura, D., & Staub, F. Status of coral reefs of the world: 2020: Executive summary. Global Coral Reef Monitoring network (GCRMN) and International Coral Reef Initiative (2021)

6. Lin, Y-J et al. Coral reefs in the northeastern saudi arabian red sea are resilient to mass coral mortality events. Mar. Pollut. Bull.; 2023; 197, [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37890317][DOI: https://dx.doi.org/10.1016/j.marpolbul.2023.115693] 115693.

7. Kleinhaus, K et al. Science, diplomacy, and the red sea’s unique coral reef: It’s time for action. Front. Mar. Sci.; 2020; 7, 90.2020FrMat..7..90K [DOI: https://dx.doi.org/10.3389/fmars.2020.00090]

8. Monroe, AA et al. In situ observations of coral bleaching in the central saudi arabian red sea during the 2015/2016 global coral bleaching event. PLoS One; 2018; 13, 4 0195814. [DOI: https://dx.doi.org/10.1371/journal.pone.0195814]

9. ISO: Underwater Acoustics-Terminology. International Organization for Standardization Geneva, Switzerland (2017)

10. Lin, T-H; Akamatsu, T; Sinniger, F; Harii, S. Exploring coral reef biodiversity via underwater soundscapes. Biol. Cons.; 2021; 253, [DOI: https://dx.doi.org/10.1016/j.biocon.2020.108901] 108901.

11. Lamont, TA et al. The sound of recovery: Coral reef restoration success is detectable in the soundscape. J. Appl. Ecol.; 2022; 59, 3 pp. 742-756. [DOI: https://dx.doi.org/10.1111/1365-2664.14089]

12. Gordon, TA et al. Habitat degradation negatively affects auditory settlement behavior of coral reef fishes. Proc. Natl. Acad. Sci.; 2018; 115, 20 pp. 5193-5198.2018PNAS.115.5193G [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29712839][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5960293][DOI: https://dx.doi.org/10.1073/pnas.1719291115]

13. Ferrier-Pagès, C et al. Noise pollution on coral reefs?—A yet underestimated threat to coral reef communities. Mar. Pollut. Bull.; 2021; 165, [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33588103][DOI: https://dx.doi.org/10.1016/j.marpolbul.2021.112129] 112129.

14. Piercy, J.J.B. The Relevance of Coral Reef Soundscapes to Iarval Fish Responses. PhD thesis, University of Essex (2015)

15. Raick, X., Di Iorio, L., Gervaise, C., Lossent, J., Lecchini, D., & Parmentier, E. From the reef to the ocean: Revealing the acoustic range of the biophony of a coral reef (moorea island, french polynesia). J. Mar. Sci. Eng.9(4) (2021https://doi.org/10.3390/jmse9040420

16. Payne, R; Webb, D. Orientation by means of long range acoustic signaling in baleen whales. Ann. N. Y. Acad. Sci.; 1971; 188, 1 pp. 110-141.1971NYASA.188.110P [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/5288850][DOI: https://dx.doi.org/10.1111/j.1749-6632.1971.tb13093.x]

17. Duarte, CM et al. The soundscape of the anthropocene ocean. Science; 2021; 371, 6529 4658. [DOI: https://dx.doi.org/10.1126/science.aba4658]

18. Slabbekoorn, H et al. A noisy spring: The impact of globally rising underwater sound levels on fish. Trends Ecol. Evolut.; 2010; 25, 7 pp. 419-427. [DOI: https://dx.doi.org/10.1016/j.tree.2010.04.005]

19. Bohnenstiehl, DR; Lillis, A; Eggleston, DB. The curious acoustic behavior of estuarine snapping shrimp: Temporal patterns of snapping shrimp sound in sub-tidal oyster reef habitat. PLoS One; 2016; 11, 1 0143691. [DOI: https://dx.doi.org/10.1371/journal.pone.0143691]

20. Kaplan, MB; Mooney, TA; Partan, J; Solow, AR. Coral reef species assemblages are associated with ambient soundscapes. Mar. Ecol. Prog. Ser.; 2015; 533, pp. 93-107.2015MEPS.533..93K [DOI: https://dx.doi.org/10.3354/meps11382]

21. Saheban, H; Kordrostami, Z. Hydrophones, fundamental features, design considerations, and various structures: A review. Sens. Actuators, A; 2021; 329, [DOI: https://dx.doi.org/10.1016/j.sna.2021.112790] 112790.

22. Desjonquères, C; Gifford, T; Linke, S. Passive acoustic monitoring as a potential tool to survey animal and ecosystem processes in freshwater environments. Freshw. Biol.; 2020; 65, 1 pp. 7-19. [DOI: https://dx.doi.org/10.1111/fwb.13356]

23. Nedelec, SL et al. Soundscapes and living communities in coral reefs: Temporal and spatial variation. Mar. Ecol. Prog. Ser.; 2015; 524, pp. 125-135.2015MEPS.524.125N [DOI: https://dx.doi.org/10.3354/meps11175]

24. Azofeifa-Solano, JC et al. Distance and orientation of hydrophones influence the received soundscape in shallow coral reefs. Front. Remote Sens.; 2025; 6, 1527988. [DOI: https://dx.doi.org/10.3389/frsen.2025.1527988]

25. Howell, KL et al. A decade to study deep-sea life. Nat. Ecol. Evolut.; 2021; 5, 3 pp. 265-267.4264522 [DOI: https://dx.doi.org/10.1038/s41559-020-01352-5]

26. Ford, B., Robinson, S., & Ablitt, J. A study of the stability exhibited by hydrophones when exposed to variation in temperature and hydrostatic pressure. In: Proc. of Meetings on Acoustics, vol. 44 (2021)

27. Burkholz, C; Duarte, C; Garcias-Bonet, N. Thermal dependence of seagrass ecosystem metabolism in the red sea. Mar. Ecol. Prog. Ser.; 2019; [DOI: https://dx.doi.org/10.3354/MEPS12912]

28. Reif, RH; Liffers, M; Forrester, N; Peal, K. Lithium battery safety: A look at woods hole oceanographic institution’s program. Prof. Saf.; 2010; 55, 02 pp. 32-37.

29. Ashry, I et al. A review of distributed fiber-optic sensing in the oil and gas industry. J. Lightwave Technol.; 2022; 40, 5 pp. 1407-1431.2022JLwT..40.1407A [DOI: https://dx.doi.org/10.1109/JLT.2021.3135653]

30. Juškaitis, R; Mamedov, A; Potapov, V; Shatalin, S. Distributed interferometric fiber sensor system. Opt. Lett.; 1992; 17, 22 pp. 1623-1625.1992OptL..17.1623J [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19798266][DOI: https://dx.doi.org/10.1364/OL.17.001623]

31. Tucker, RS; Eisenstein, G; Korotky, SK. Optical time-division multiplexing for very high bit-rate transmission. J. Lightwave Technol.; 2002; 6, 11 pp. 1737-1749.1988JLwT..6.1737T [DOI: https://dx.doi.org/10.1109/50.9991]

32. Harmon, N., Belal, M., Mangriotis, M.-D., Spingys, C., & Rychert, C.A. Distributed acoustic sensing along a shallow water energy cable. IEEE J. Oceanic Eng. 2025)

33. Marra, G et al. Ultrastable laser interferometry for earthquake detection with terrestrial and submarine cables. Science; 2018; 361, 6401 pp. 486-490.2018Sci..361.486M [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29903881][DOI: https://dx.doi.org/10.1126/science.aat4458]

34. Winzer, PJ; Neilson, DT. From scaling disparities to integrated parallelism: A decathlon for a decade. J. Lightwave Technol.; 2017; 35, 5 pp. 1099-1115.2017JLwT..35.1099W [DOI: https://dx.doi.org/10.1109/JLT.2017.2662082]

35. Miele, P., Snead, K., Zakhireh, N., Homa, D., Pickrell, G., & Risch, B.G. Optical fiber reliability in harsh environments. In: Int. Wire & Cable Symp (2020)

36. Bouffaut, L et al. Eavesdropping at the speed of light: Distributed acoustic sensing of baleen whales in the arctic. Front. Mar. Sci.; 2022; 9, [DOI: https://dx.doi.org/10.3389/fmars.2022.901348] 901348.

37. Landrø, M et al. Sensing whales, storms, ships and earthquakes using an arctic fibre optic cable. Sci. Rep.; 2022; 12, 1 19226.2022NatSR.1219226L [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36357493][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9649797][DOI: https://dx.doi.org/10.1038/s41598-022-23606-x]

38. Sladen, A et al. Distributed sensing of earthquakes and ocean-solid earth interactions on seafloor telecom cables. Nat. Commun.; 2019; 10, 1 5777.2019NatCo.10.5777S [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31852896][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6920424][DOI: https://dx.doi.org/10.1038/s41467-019-13793-z]

39. Rørstadbotnen, R.A., Landrø, M., Taweesintananon, K., Bouffaut, L., Potter, J.R., Johansen, S.E., Kriesell, H.J., Brenne, J.K., Haukanes, A., Schjelderup, O., & Storvik, F. Analysis of a local earthquake in the arctic using a 120 km long fibre-optic cable 2022(1), 1–5 (2022) https://doi.org/10.3997/2214-4609.202210404

40. Lin, J et al. Monitoring ocean currents during the passage of typhoon muifa using optical-fiber distributed acoustic sensing. Nat. Commun.; 2024; 15, 1 1111.2024NatCo.15.1111L [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38321090][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10847141][DOI: https://dx.doi.org/10.1038/s41467-024-45412-x]

41. Richardson, DJ; Fini, JM; Nelson, LE. Space-division multiplexing in optical fibres. Nat. Photonics; 2013; 7, 5 pp. 354-362.2013NaPho..7.354R [DOI: https://dx.doi.org/10.1038/nphoton.2013.94]

42. Mao, Y et al. Simultaneous distributed acoustic and temperature sensing using a multimode fiber. IEEE J. Sel. Top. Quantum Electron.; 2020; 26, 4 pp. 1-7. [DOI: https://dx.doi.org/10.1109/JSTQE.2020.2964398]

43. Huang, M-F et al. First field trial of distributed fiber optical sensing and high-speed communication over an operational telecom network. J. Lightwave Technol.; 2019; 38, 1 pp. 75-81.2020JLwT..38..75H [DOI: https://dx.doi.org/10.1109/JLT.2019.2935422]

44. Marin, JM et al. Simultaneous distributed acoustic sensing and communication over a two-mode fiber. Opt. Lett.; 2022; 47, 24 pp. 6321-6324.2022OptL..47.6321M [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36538428][DOI: https://dx.doi.org/10.1364/OL.473502]

45. Hu, Z et al. Enabling cost-effective high-performance vibration sensing in digital subcarrier multiplexing systems. Opt. Express; 2023; 31, 20 pp. 32114-32125.2023OExpr.3132114H [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37859021][DOI: https://dx.doi.org/10.1364/OE.497616]

46. Guo, Y et al. Submarine optical fiber communication provides an unrealized deep-sea observation network. Sci. Rep.; 2023; 13, 1 15412.2023NatSR.1315412G [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37723196][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10507058][DOI: https://dx.doi.org/10.1038/s41598-023-42748-0]

47. Gunawan, W.H., Marin, J.M., Rjeb, A., Kang, C.H., Ashry, I., Ng, T.K., & Ooi, B.S. Energy harvesting over fiber from amplified spontaneous emission in optical sensing and communication systems. J. Lightwave Technol. 2024)

48. Gavrilov, A.N., & Parsons, M.J. A matlab tool for the characterisation of recorded underwater sound (chorus). Acoustics Australia42(3) (2014)

49. Song, Z et al. Sounds of snapping shrimp (alpheidae) as important input to the soundscape in the southeast china coastal sea. Front. Mar. Sci.; 2023; 10, 1029003.2023sra.book...S [DOI: https://dx.doi.org/10.3389/fmars.2023.1029003]

50. Amorim, MCP. Diversity of sound production in fish. Commun. Fishes; 2006; 1, pp. 71-104.

51. Ladich, F. Ecology of sound communication in fishes. Fish Fish.; 2019; 20, 3 pp. 552-563. [DOI: https://dx.doi.org/10.1111/faf.12368]

52. Ladich, F; Bass, A; Farrell, A. Vocal behavior of fishes: Anatomy and physiology. Encyclopedia of Fish Physiology: From Genome to Environment; 2011; 1, pp. 321-329. [DOI: https://dx.doi.org/10.1016/B978-0-12-374553-8.00018-6]

53. Xing, C., Tan, G., & Ran, Y. Enhanced off-grid underwater acoustic signals direction estimation using toeplitz covariance reconstruction and subspace fitting. Circuits, Systems, and Signal Processing, 1–29 (2025)

54. Rørstadbotnen, RA et al. Simultaneous tracking of multiple whales using two fiber-optic cables in the arctic. Front. Mar. Sci.; 2023; 10, 1130898. [DOI: https://dx.doi.org/10.3389/fmars.2023.1130898]

55. Malfante, M; Mars, JI; Dalla Mura, M; Gervaise, C. Automatic fish sounds classification. J. Acoust. Soc. Am.; 2018; 143, 5 pp. 2834-2846.2018ASAJ.143.2834M [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29857733][DOI: https://dx.doi.org/10.1121/1.5036628]

56. Looby, A et al. A quantitative inventory of global soniferous fish diversity. Rev. Fish Biol. Fisheries; 2022; 32, 2 pp. 581-595. [DOI: https://dx.doi.org/10.1007/s11160-022-09702-1]

57. Ladich, F; Fine, ML. Sound-generating mechanisms in fishes: A unique diversity in vertebrates. Commun. Fishes; 2006; 1, pp. 3-43.

58. Staaterman, E; Paris, CB; Kough, AS. First evidence of fish larvae producing sounds. Biol. Let.; 2014; 10, 10 20140643. [DOI: https://dx.doi.org/10.1098/rsbl.2014.0643]

59. Parsons, MJ et al. Sounding the call for a global library of underwater biological sounds. Front. Ecol. Evol.; 2022; 10, [DOI: https://dx.doi.org/10.3389/fevo.2022.810156] 810156.

60. Ashry, I et al. Cnn-aided optical fiber distributed acoustic sensing for early detection of red palm weevil: A field experiment. Sensors; 2022; 22, 17 6491.2022Senso.22.6491A [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36080949][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9459888][DOI: https://dx.doi.org/10.3390/s22176491]

61. Rivet, D; Cacqueray, B; Sladen, A; Roques, A; Calbris, G. Preliminary assessment of ship detection and trajectory evaluation using distributed acoustic sensing on an optical fiber telecom cable. J. Acoust. Soc. Am.; 2021; 149, 4 pp. 2615-2627.2021ASAJ.149.2615R [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33940881][DOI: https://dx.doi.org/10.1121/10.0004129]

62. Liu, Z., Zhang, L., Liu, H., Qiu, Z., Xiao, Z., Chen, Z., Wang, T., & Pang, F. 3d printing technology-enhanced phase-sensitive otdr for underwater acoustic wave detection. Optical Fiber Sensors Conference 2020 Special Edition (2021) https://doi.org/10.1364/ofs.2020.t3.26

63. Zhu, S., Chen, J., Ai, K., Fan, C., Li, H., Yan, Z., & Sun, Q. Fully distributed fiber-optic hydrophone cable for acoustic source azimuth estimation. 2024 OES China Ocean Acoustics (COA), 1–5 (2024) https://doi.org/10.1109/COA58979.2024.10723668

64. Zhang, C; Yang, S; Wang, X. Dual pulse heterodyne distributed acoustic sensor system employing soa-based fiber ring laser. Front. Phys.; 2023; 11, 1196067. [DOI: https://dx.doi.org/10.3389/fphy.2023.1196067]

65. Zhang, Y; Yang, H; Chen, Z; Sun, F; Mao, B. Design and analysis of mems piezoelectric hydrophone based on signal-to-noise ratio. IEEE Sens. J.; 2025; 25, pp. 11314-11322.2025ISenJ.2511314Z [DOI: https://dx.doi.org/10.1109/JSEN.2025.3540307]

66. Lamont, T.A.C., Chapuis, L., Williams, B., Dines, S., Gridley, T., Frainer, G., Fearey, J., Maulana, P.B., Prasetya, M.E., Jompa, J., Smith, D.J., & Simpson, S. Hydromoth: Testing a prototype low?cost acoustic recorder for aquatic environments. Remote Sens. Ecol. Conserv.8 (2022) https://doi.org/10.1002/rse2.249

67. Dahl, P; Miller, JH; Cato, D; Andrew, R. Underwater ambient noise. Acoustics Today; 2007; 3, 23. [DOI: https://dx.doi.org/10.1121/1.2961145]

68. Dalton, SJ et al. Successive marine heatwaves cause disproportionate coral bleaching during a fast phase transition from el niño to la niña. Sci. Total Environ.; 2020; 715, [DOI: https://dx.doi.org/10.1016/j.scitotenv.2020.136951] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32014776]136951.

69. Lu, P., Lalam, N., Badar, M., Liu, B., Chorpening, B.T., Buric, M.P., & Ohodnicki, P.R. Distributed optical fiber sensing: Review and perspective. Appl. Phys. Rev.6(4) (2019)

70. Mao, Y et al. Simultaneous distributed acoustic and temperature sensing using a multimode fiber. IEEE J. Sel. Top. Quantum Electron.; 2020; 26, 4 pp. 1-7. [DOI: https://dx.doi.org/10.1109/JSTQE.2020.2964398]

71. Mao, Y et al. Sensing within the otdr dead-zone using a two-mode fiber. Opt. Lett.; 2020; 45, 11 pp. 2969-2972.2020OptL..45.2969M [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32479435][DOI: https://dx.doi.org/10.1364/OL.393539]

72. Lu, Y; Zhu, T; Chen, L; Bao, X. Distributed vibration sensor based on coherent detection of phase-otdr. J. Lightwave Technol.; 2010; 28, 22 pp. 3243-3249.2010JLwT..28.3243L

73. Bao, X; Zhou, D-P; Baker, C; Chen, L. Recent development in the distributed fiber optic acoustic and ultrasonic detection. J. Lightwave Technol.; 2017; 35, 16 pp. 3256-3267.2017JLwT..35.3256B [DOI: https://dx.doi.org/10.1109/JLT.2016.2612060]

74. Posey, R, Jr; Johnson, G; Vohra, S. Strain sensing based on coherent rayleigh scattering in an optical fibre. Electron. Lett.; 2000; 36, 20 pp. 1688-1689.2000ElL..36.1688P [DOI: https://dx.doi.org/10.1049/el:20001200]

75. Lillis, A; Mooney, TA. Snapping shrimp sound production patterns on caribbean coral reefs: Relationships with celestial cycles and environmental variables. Coral Reefs; 2018; 37, 2 pp. 597-607.2018CorRe.37.597L [DOI: https://dx.doi.org/10.1007/s00338-018-1684-z]

76. Ashry, I et al. Normalized differential method for improving the signal-to-noise ratio of a distributed acoustic sensor. Appl. Opt.; 2019; 58, 18 pp. 4933-4938.2019ApOpt.58.4933A [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31503813][DOI: https://dx.doi.org/10.1364/AO.58.004933]