Assessing the potential of acoustic telemetry to underpin the regional management of basking sharks ( Cetorhinus maximus )

Acoustic telemetry can provide valuable space-use data for a range of marine species. Yet the deployment of species-specific arrays over vast areas to gather data on highly migratory vertebrates poses formidable challenges, often rendering it impractical. To address this issue, we pioneered the use of acoustic telemetry on basking sharks ( Cetorhi-nus maximus ) to test the feasibility of using broadscale, multi-project acoustic receiver arrays to track the movements of this species of high conservation concern through the coastal waters of Ireland, Northern Ireland, and Scotland. Throughout 2021 and 2022, we tagged 35 basking sharks with acoustic transmitters off the west coast of Ireland; 27 of these were detected by 96 receiver stations throughout the study area ( n = 9 arrays) with up to 216 detections of an individual shark (mean = 84, s.d. 65). On average, sharks spent ~ 1 day at each acoustic array, with discrete residency periods of up to nine days. Twenty-one sharks were detected at multiple arrays with evidence of inter-annual site fidelity, with the same individuals returning to the same locations in Ireland and Scotland over 2 years. Eight pairs of sharks were detected within 24 h of each other at consecutive arrays, suggesting some level of social coordination and synchronised movement. These findings demonstrate how multi-project acoustic telemetry can support international, cost-effective monitoring of basking sharks and other highly mobile species. Decision support tools such as these can consolidate cross-border management strategies, but to achieve this goal, collaborative efforts across jurisdictions are necessary to establish the required infrastructure and secure ongoing support.


Introduction
Our understanding of large marine animal movement and behaviour gained significant momentum following the advent of satellite telemetry [5,22,23,29,49].This technology has been used to investigate the ocean-scale movements of migratory species (e.g., [22,29,39]), identify critical habitats [5,24,40], reveal foraging behaviours (e.g., [53]), and identify hotspots of human/wildlife conflicts (e.g., [47,48]).For managers, such information can underpin the implementation of effective conservation measures [12,24].While satellite telemetry has the potential to track animals over multiple years [49], studies using this technology typically focus on providing powerful snapshots of behaviour and movement over periods of less than one year [49], rather than acquiring long-term data sets.However, for long-lived species, spatial protection must be reinforced by monitoring over multiple years to understand behavioural plasticity and animal responses to changing environments [1,2,9,11,40].Acoustic telemetry offers the potential to monitor individuals for up to 10 years [1,29,40].This longevity has driven a proliferation of acoustic telemetry projects over recent decades [15,40], deepening our understanding of animal movement patterns on a global scale [24].
Acoustic telemetry determines the presence and location of the study species via a two-component system: (i) static or mobile receivers, which can detect (ii) transmitters (tags), either internally or externally attached to the animals, which emit an encoded acoustic signal.In its simplest form, when a tagged individual moves into the detection range of a receiver (typically less than 1,000 m), its unique ID code is recorded, along with a date and time stamp.Arrays of receivers are usually deployed at a given location depending on the research question [1,2,40] and the scale of the animal movements to be tracked, which can range from high-resolution movements over small spatio-temporal scales to pan-oceanic migrations [1,2,17,29,35,40,51].For example, the long battery life of acoustic tags [1] enables studies of inter-year seasonal site fidelity, ontogenetic variation in movement patterns and overall space use (e.g., [19,46]).However, for highly migratory marine vertebrates, the logistical and financial challenges of deploying species-specific acoustic telemetry arrays over large geographic ranges are frequently beyond the capacity and/or resources of a single research team.To address this issue, a host of collaborations and infrastructure and data-sharing agreements have emerged so that receivers deployed to study a given species can add value to other projects by detecting and sharing information on non-target species [1,11].However, for this approach to work, a formalised network needs to be in place to connect acoustic telemetry users and streamline the transfer of data [1,11].To meet this global need, the acoustic telemetry community has been cultivating partnership platforms, such as the Ocean Tracking Network (OTN) [11] and the European Tracking Network [1], to create a global formalised network.These platforms and collaborations greatly expand the coverage of effective acoustic telemetry arrays with minimised additional resource requirements, to the benefit of all concerned [1].
The SeaMonitor project (2019-2023) was developed based on this collaborative approach, spanning the territorial waters of the Republic of Ireland, Northern Ireland (UK), and Scotland (UK).A key aim of the project was to use acoustic telemetry to produce regionally coherent management recommendations for a range of migratory species that habitually exhibit trans-boundary movements across multiple jurisdictional boundaries.The centrepiece of SeaMonitor was an acoustic telemetry array (N = 108 receiver stations across 65 km) spanning the Malin-Islay front that linked the north coast of Ireland with the west coast of Scotland.While this array was primarily deployed to estimate the survivorship of Atlantic salmon (Salmo salar) smolts at sea, we wanted to assess its value for studies of other highly mobile species.Basking sharks (Cetorhinus maximus) were selected as a target species given their conservation priority status (e.g., an EC-wide moratorium on target fishing, Republic of Ireland (ROI) Wildlife Act 1976, Wildlife (Northern Ireland) Order 1985, Nature Conservation (Scotland) Act 2004, International Appendices I and II of the Convention on Migratory Species (CMS), Appendix II of the CITES, Endangered on the IUCN Red List) and seasonal occurrence in all three jurisdictions [14-16, 21, 31, 32, 54, 58-61, 70].
To date, spatial-based management for basking sharks has been based primarily on protecting established hotspots identified through effort-corrected visual observation and satellite telemetry, leading to the designation of the Sea of Hebrides Marine Protected Area in Scotland [46] and a network of marine reserves in the Isle of Man.However, less attention has been paid to basking shark conservation at a multi-national scale which is pertinent given their frequent, wide-ranging movements across borders [14-16, 21, 31, 32, 54, 58-61, 70].To meet this need, we capitalised on the extensive acoustic telemetry network around the island of Ireland and off western Scotland, including the Malin-Islay array, to establish a basking shark acoustic tracking programme in 2020.
The rationale for acoustically tracking basking sharks was to assess the technology's efficacy as a long-term monitoring tool to support the management of this protected species.Thus, we posed five primary questions: (1) What proportion of the acoustic telemetry receivers deployed in the Ireland/Western Scotland region detected basking sharks?(2) What proportion of tagged animals were subsequently detected at arrays beyond the tagging location within the same year?(3) Was there any evidence of inter-annual site fidelity to particular areas?(4) How long were individual sharks detectable within given locations, providing estimates of individual residency?(5) Was tag retention adequate to record interannual movements?

Array
All acoustic receivers were deployed sub-surface using acoustic release units with no surface markers.The location of all receivers was marked via GPS positioning at the site of deployment.This study used two types of receivers: Innovasea VR2W units, deployed with separate Innovasea Ascent Acoustic Releases (Ascent AR) units, and Innovasea VR2AR units, which had built-in acoustic releases.The VR2AR and Ascent AR units were connected to 70 kg of sacrificial ballast via a 1-m-long, 15-mm-diameter sea-steel leash.For both VR2AR and Ascent AR units, two × 2 m lengths of 10-mm sea steel were attached to the bolt holes at the top of the units.These were plaited together approximately 30 cm above the units, and two 12-in.(30 cm) hard plastic buoys were threaded onto the opposite end of the rope and knotted in place.For Ascent AR units, a VR2W unit was cable tied to the plaited 10-mm riser rope between the Ascent AR and the floats.The SeaMonitor project deployed a central main receiver "curtain", linking Malin Head to the Isle of Islay (Fig. 1).Additional arrays were deployed to monitor the presence of tagged basking sharks at known Irish hotspots around Malin Head, Tory Island and Achill Island (Fig. 1).The SeaMonitor arrays were complemented by networks of receivers deployed by the West Coast Tracking Project [63], MEFS [64], COMPASS [10], and SAMOSAS [68] in the marine waters of western Ireland, north-western England, and west Scotland (Fig. 1).Combined, the deployed arrays consisted of 637 receiver stations, spanning a latitudinal distance of ca.480 km and a longitudinal distance of ca.470 km.All receivers operated at 69 kHz.

Deployment of acoustic tags
Tagging was undertaken on free-swimming sharks in May (n = 11), August (n = 3) and September (n = 9) 2021, and April 2022 (n = 12).All tagging was conducted from either an Orkney 4.87 m (16 foot) fibreglass boat with a 50-hp outboard or a 6 m XS 600 RHIB with a 115-hp outboard.Basking sharks were tagged with externally mounted Innovasea V16-4 × ID tags inside the manufacturer's external casing (length = 87 mm, diameter = 18 mm, weight in air = ~ 12 g).All tags transmitted at 69 kHz with a nominal tag delay of 120-240 s and an estimated tag life of 3650 days.Tag cases were attached to a 5-cm stainless steel aviation wire tether.A 5-cm Wildlife Computers titanium anchor was used for each tag to secure the device in the dorsal musculature at the rear of the shark's dorsal fin using a 2-m fibreglass pole with an applicator pin attached to one end to dart the tag into place.The tagger was stationed at the bow of the vessel.The total length of each tagged animal was estimated with reference to the boat [6]

Data processing and analysis
The detection data were sorted to remove all detections of tags not associated with this study and then further filtered to remove possible false detections, identified as a single detection within a 24-h period.These were identified using the false_detections function in the R package GLATOS [27].Receiver arrays were grouped based on geographic location (see Fig. 2), and summary data were calculated, including the total number of detections, detections per shark, detections per receiver, and detections per array.Discrete residency periods within an array were determined using the detection_events function from the GLATOS package.A break duration of one week was employed to determine a residency event whereby a break between detections exceeding one week was deemed to represent separate residency events.To investigate the potential of acoustic telemetry data to reveal whether basking sharks travel through coastal waters in groups, we identified "group detection events" per array.These were defined as detections of two sharks or more at a single array with less than a 24-h break between detections.

Results
In total, 35 basking sharks were tagged at two sites off the west coast of Ireland during the SeaMonitor project, 23 in 2021 (Achill Island, County Mayo and Loop Head, County Clare) and 12 in 2022 (Achill Island).The sharks ranged from 4 to > 8 m in estimated total length (see Appendix Table S1).Of these, 27 were subsequently detected (77%).In total, there were 2304 detections between 14 May 2021 and 06 Oct 2022.Of these, 47 (1.9%) detections from 23 sharks (ranging from 1 (n = 14 sharks) to 7 for shark #11) were removed as possible false detections.The SeaMonitor main array had the greatest number of possible false detections (n = 17), with 12 other arrays throughout the study region also having detections that were removed as possible false detections (mean number of possible false detections per array = 3.61, s.d.4.17).Of the 2,257 filtered detections, 1462 were from the SeaMonitor receivers, and 795 were from additional arrays.Basking sharks were detected by 96 receivers Overall, 21 (60%) sharks were detected by multiple arrays.For example, eight sharks were detected at more than one location during 2021, two of which were also detected at more than one location during 2022 (Fig. 3), with a further 12 sharks detected at multiple locations during that year (Fig. 3).Of the 23 sharks tagged in 2021, four (#5, #6, #9, and #10) showed inter-annual site fidelity between 2021 and 2022 off Achill Island, Malin Head, the Hebrides (West of the Hebrides, Coll to Small Isles, Barra, Uist, and Skye arrays), and the Sea-Monitor main array (e.g., #5 Fig. 4).The location of the detections showed some seasonal patterns, with sharks Group detection events occurred at Achill Island, Tory Island and Malin Head in Ireland, the SeaMonitor Main Array, South Barra, and Coll to Small Isles in Scotland (Table 1).Achill Island recorded two large group detection events comprising 11 (May 2021) and 12 (April 2022) sharks (Table 1).Each of these group events comprised different sharks and it is worth noting that the sharks had recently been tagged near Achill in the same month.The Malin Head array recorded several group detection events in 2021 involving six sharks in total.There was evidence of synchronous movements of 10 shark pairs detected within 24 h of each other at two or more arrays over different months.In addition, one pair (#5 and #9) was detected within 24 h of each other four times at three arrays in different months and years (Table 1), with paired detections continuing to occur after nearly a full year at liberty for this pair.
The average residency period for sharks at an array was 0.94 days (s.d.1.64).However, some sharks were detected for more extended periods, up to 9.3 days between 20 May 2022 and 29 May 2022 (detected on May 20th, 21st, 22nd, 23rd, 24th, 27th, and the 29th) at the Coll to Small Isles array off the west coast of Scotland (shark #11, Fig. 5).Other locations with notable individual residency periods included Achill Island (5.6 days), around Tory Island (5.6 days) and off Malin Head (4 days) (Fig. 5).The most extended residency times (> 4 days) were all in April and May, with residencies of greater than 1 day recorded frequently in these months.Residency times of approximately three days were also observed during September and October.Sharks displaying residency of over three days (seven occurrences in total by seven different sharks) ranged from relatively small (4-6 m size class; 9.3 days) to very large > 8 m (3 and 5.6 days).Two sharks (#5 and #10) were detected 506 days post-tagging off Achill Island, both by the SeaMonitor main array, demonstrating tags can be retained for long periods.

Discussion
This study is the first to demonstrate that acoustic telemetry can reveal multi-year, multi-spatial scale information on space use in coastal waters for basking sharks.Understanding a species' movement and space use is of particular importance for mobile species as connectivity [3,7] and time spent within certain areas [7,13] are highly relevant when designing spatially referenced conservation measures [12].
The ability to undertake relatively straightforward residency analysis with acoustic telemetry data can help identify areas where tagged sharks spend a proportionately high amount of time.The acoustic telemetry data revealed basking sharks spent periods of up to 9.3 days in the vicinity of arrays within the Sea of the Hebrides Marine Protected Area (MPA) and visited multiple times in one year.In some instances, this repeated site use occurred over multiple years, with four of the sharks tagged in 2021 also detected in 2022, showing interannual fidelity to Achill Island, Malin Head, the Hebrides (Coll to Small Isles, Uist, Skye, and Barra), and Malin-Islay front (covered by the SeaMonitor main array), indicating these areas may represent important habitats for the species.While previous research has shown interannual fidelity for basking sharks at the Hebrides [15], longer-term acoustic studies will provide an opportunity to demonstrate this over multiple years and should be supported by additional research that could provide information on what these sites are being used for.The consistent use of a particular site is useful for establishing protected areas, for example the Sea of the Hebrides MPA was designated in part for the protection of basking sharks from effort-corrected sightings data [46] and satellite tracking [15].The present study highlights the potential for acoustic telemetry to provide longer-term data that would complement these other methods, monitoring individuals in relation to MPAs or other effective conservation measures to support effective management in the face of a changing ocean.For instance, the interannual connectivity between Achill Island and the Sea of the Hebrides MPA by shark #5 demonstrates current repeated connectivity between the MPA and external arrays, which could contribute to the development of a coherent network of protection for these mobile marine species of high conservation priority [3,7].Furthermore, evidence of inter-annual fidelity highlights how acoustic telemetry can contribute valuable insight into basking shark behaviour over protracted timescales.
Prior to this study, our understanding of basking shark movements through marine waters in the UK and Ireland predominantly came from data obtained through a combination of archival and transmitting satellite tags and population genetic studies [14-16, 31, 35, 36, 54, 55, 57, 58, 62, 70].These highlighted the extensive movements that basking sharks make annually [21,32].While data from satellite telemetry provides valuable insight into the movements of a species, the fine-scale movement data that is possible from acoustic telemetry can complement these technologies [40].If sharks are double tagged with satellite and acoustic tags, acoustic detections can help validate estimated positions from satellite tags.However, acoustic telemetry data are reliant on a network of acoustic receivers to be maintained and the spatial resolution of the data depends on the number of receivers in the network.While mark and recapture data are fundamental for the conservation of many mobile species (e.g., [43,69]), it has been noted that sightings data for sharks, including basking sharks, has several inherent biases [8,19], largely related to depth use [31,58].In whale sharks (Rhincodon typus), for example, seasonal variation in depth use resulted in sightings data incorrectly suggesting the animals only had a seasonal presence, while a concurrent acoustic telemetry project demonstrated year-round residency [8].The acoustic telemetry data presented here could undoubtedly support mark and recapture studies for basking sharks; collectively helping us to investigate long-standing questions regarding site fidelity, habitat use, and abundance (e.g., [8,33,34,43]).
Currently available acoustic tags have the potential to track animals for up to ten years, but only if the devices stay attached.In this study, the detection of 77% of tagged sharks, and detections up to 506 days after tagging provides confidence that the externally attached acoustic tags could be retained in the long-term, supporting their use as part of a long-term monitoring tool.This capacity to monitor movements over long periods is especially relevant for species' with typically long-life cycles, such as elasmobranchs, as ontogenetic variation in space used has been documented for several species [66,67].
The acoustic telemetry data not only revealed the movement and space use of individual basking sharks, but provided tentative insights into group behaviour.Our ability to tag sharks within aggregations allowed us to explore the consistency of groups over time.Our data suggested that some sharks tagged in the same location on the same day tended to visit other coastal areas simultaneously.Ten pairs of sharks Table 1 Groups of individual sharks (Shark ID) that were detected at the same array within a 24-h period Each row shows a group detection event, defined as groups of detections with less than a 24-h break.'Array' shows the arrays the group detections occurred at.'Shark IDs' shows which sharks were part of the grouped detection event; a solidus splits discrete group detections that occurred within the same month.'Month and year' shows when each group event took place detected within 24 h of each other at two or three arrays throughout the study area up to one-year posttagging, suggesting shared routes and/or the possibility of social behaviour.While basking sharks have been observed in groups [19,52,55,56], the long-term, multi-site nature of these groups has not been explored with tracking technology.Acoustic telemetry data have been previously used to explore the social interactions within other shark species [30,45] and using long-term acoustics telemetry would allow further investigation of the social interactions in basking sharks populations by tracking pairs or groups of sharks across multiple locations over successive years.Further understanding of drivers behind grouping behaviours is crucial for effective conservation management [30,41]; evidence from molecular studies suggests that basking sharks may have some form of social cohesion, moving throughout the NE Atlantic in genetically discrete kin groups [37] which has implications for management when considering preserving genetic diversity [65].However, the capacity of acoustic telemetry to study this behaviour more extensively relies on the continued deployment of large-scale, regionally coherent receiver arrays.The detections of sharks across 96 receiver stations with a maximum straight-line travel distance of 1,236 km between detections reinforces the value of coordinating monitoring efforts and data sharing across organisations and national boundaries from projects dedicated to different target species to maximise array coverage.Innovative partnerships such as the Ocean Tracking Network [11] and the European Tracking Networks [50] are advancing the use of acoustic telemetry to facilitate global-scale studies, investigating a wealth of pressing questions about aquatic animals the answers to which help inform management and policy decisions [1,2,11,24,29,40].Future acoustic telemetry approaches for basking sharks could involve the use of real-time acoustic telemetry data collection systems [71] which would underpin rapid response adaptive management decisions.For example, tools that that notify marine users of time and areas where due care and attention needs to be paid (e.g., proximate to shark aggregations).Such approaches have been adopted with great success for other marine species, including real-time fisheries closures for salmon species (Oncorhynchus tshawytscha) [38], spurdog (Squalus acanthias) [25], bluefin tuna (Thunnus thynnus) [4,26], and cod (Gadus morhua) [28].
In conclusion, this study demonstrated that acoustic telemetry offers a cost-effective, and adaptive tool that has strategic benefit in the study of highly mobile species, such as basking sharks.Acoustic telemetry can provide complementary data to augment mark and recapture and satellite tagging studies, providing information on space use, residency behaviour, inter-annual fidelity, connectivity among political jurisdictions, and may, over time, provide insight into group dynamics.Although the current study only considered results comprising detections across two years, these data indicate that acoustic telemetry may provide long-term value for monitoring basking sharks.The initial evidence of residency, connectivity, and inter-annual site fidelity that the acoustic data showed in relation to the Sea of the Hebrides MPA suggests that acoustic telemetry could form part of the monitoring toolbox for this site, and any future areas of spatial management for basking sharks.As demonstrated in this study, the use of acoustic telemetry as a monitoring tool requires wide-scale acoustic telemetry infrastructure to be in place.The continued growth and networking of acoustic telemetry capabilities in the NE Atlantic is providing an increased capability to track the movements of basking sharks and other species.Yet, in order to fully exploit acoustic telemetry as a tool, there needs to be a trans-boundary collaborative effort [2] to support longterm deployments of acoustic telemetry array networks.However, aside from an obvious funding requirement, such aspirations rely on dialogue, collaboration, and planning among researchers, jurisdictions, government agencies and departments to ensure access to long-term infrastructure and investment.

Fig. 1
Fig. 1 Map of all compatible acoustic receivers in the water from when the first basking shark was tagged.Red points show receivers deployed as part of SeaMonitor and blue points show receivers deployed as part of other projects.The 12 nm territorial waters of Ireland, Northern Ireland, the UK and the Isle of Man are shown.The Sea of the Hebrides Marine Protected Area (of which basking sharks are one of the protected species) is shown in yellow

Fig. 2
Fig. 2 All basking shark detections per array station.The point size indicates the number of total detections (all sharks) at each receiver station, and the colour of the detection point shows the array this detection was from.The green triangle shows tagging location off Achill Island, and the red triangle shows the tagging location off County Clare

Fig. 3
Fig. 3 Detection plot for all basking sharks (Study ID on the y-axis).Detections are colour-coded based on the array that detected the shark (see Fig. 2 for more information).Green triangles denote the date of tagging at Achill Island, and red triangles show the date of tagging at County Clare

Fig. 4
Fig. 4 Detections of basking shark #5 showing annual site fidelity to Achill Island and the Hebrides (Skye to Uist and Barra to Uist arrays).The green triangle shows the tagging location.Coloured points show the receiver stations the shark was detected on.Detections are linked by straight lines with arrows to aid visualisation.These lines are coloured based on the year the detections occurred.Arrows indicate direction of travel

Fig. 5
Fig. 5 Residency periods of basking sharks.Residency is estimated per array based on a maximum break of one week between detections.The point is at the mean location of residence within the array, and the size and the colour of the points represent the number of days of residency