Prey availability and fine-scale oceanography drive reef manta ray aggregations at the world’s largest hotspot

Prey availability and fine-scale oceanography drive reef manta ray aggregations at the world’s largest hotspot | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Prey availability and fine-scale oceanography drive reef manta ray aggregations at the world’s largest hotspot Hannah M Moloney, Guy M.W Stevens, Asia O Armstrong, Alvise Dabala, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9455345/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 4 You are reading this latest preprint version Abstract Understanding the drivers of megafauna aggregations is critical for managing threatened species in a changing climate. We analysed an 18-year dataset (2007–2024) of reef manta ray ( Mobula alfredi ) abundance and zooplankton density at Hanifaru Bay, Maldives (5.1733°N, 73.145°E) —the world’s largest known manta ray aggregation. Generalised Linear Models were used to evaluate sightings against local environmental variables and broad-scale climatic indices. Results reveal that interannual abundance was modulated by the Indian Ocean Dipole, with sightings and prey density reaching decadal highs during the record Negative Indian Ocean Dipole and La Niña of 2021–2022. Locally, zooplankton density was the strongest predictor of M. alfredi abundance, explaining > 60% of the observed variation. Sightings were 25x higher during peak prey availability than during prey absence. Aggregations were timed with spring tides (peaking 2.7 hours post-high tide), new and full moon phases, and 25–35 km/h north-northwest winds. We also identified a critical thermal threshold at 28.6°C, beyond which sightings declined by 15.7% per degree of sea surface temperature warming. This sequence confirms that the local bathymetric retention mechanisms are dependent on specific regional oceanographic conditions. Our findings suggest that while Hanifaru Bay consistently concentrates zooplankton biomass, its functional utility as a foraging hotspot is vulnerable to climate-driven shifts in the Indian Ocean Dipole and rising sea surface temperatures, which may decouple the predator from its primary nutrient supply. Life Below Water Climate Action Elasmobranch Planktivores Seasonality Environmental variability Conservation Mobula alfredi Endangered species Marine protected area Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Understanding drivers of spatial distribution and habitat selection of large, highly mobile species is essential for effective management (Graham et al. 2012 ). In dynamic marine environments, megaplanktivores – large marine animals that feed on plankton – aggregate at specific sites to exploit prey patches that exceed energetic thresholds required to sustain their large body size (Stephens and Krebs 1986 ). Foraging for zooplankton presents unique energy challenges, including high search costs and the requirement for a minimum prey density to offset the metabolic demands of bulk-filter feeding (Hildebrand et al. 2022 ). However, locating high concentrations of zooplankton is difficult, as zooplankton distributions are often characterised by spatiotemporal unpredictability across multiple scales (Hays et al. 2006 ). Consequently, the movement and habitat use of these species are highly coupled to oceanographic variables, such as currents, tides, winds, and upwelling, that dictate prey availability. Disentangling these drivers remains complex, as the relative influence of foraging ecology, species-specific physiologies, and reproductive requirements vary across taxa and geographic regions (Sims et al. 2005 ; Zerbini et al. 2006 ; Skomal et al. 2009 ; Anderson et al. 2011 ; Harris and Stevens 2021 ; Cullain et al. 2025 ). With increasing anthropogenic pressures and climatic shifts predicted to alter reef-wide ocean dynamics and productivity (Hays et al. 2005 ; Kwiatkowski et al. 2019 ; Heneghan et al. 2023 ), identifying the specific environmental triggers of megaplanktivore aggregations is critical to evaluate the adequacy of current and future management. The coupling between oceanography and biology is often mediated by environmental drivers operating across multiple spatio-temporal scales (Sims et al. 2006 ; Jaine et al. 2012 ). At the regional scale, broad climatic indices such as the El Niño-Southern Oscillation (ENSO) and the Indian Ocean Dipole (IOD) regulate primary productivity by modulating sea surface temperatures (SSTs) and thermocline depth (Roxy et al. 2016 ; Dalpadado et al. 2024 ; Abeywickrama et al. under review). At the mesoscale, features such as fronts and eddies act as productive hotspots by physically concentrating biomass into predictable foraging patches (Sims et al. 2003 ; Jaine et al. 2012 ). Finally, at the local scale, interactions between wind-driven currents, tidal flux, and complex bathymetry create high-density retention zones that facilitate the aggregation of both prey and predators (Wolanski et al. 1996 ; Chenoweth et al. 2011 ). The influence of these scales is evident in the movement and abundance of various megaplanktivores. For example, in Australia, regional-scale Southern Oscillation Index influences the abundance of whale sharks ( Rhincodon typus ) on the Ningaloo Reef by altering the strength of the Leeuwin Current (Sleeman et al. 2010 ); while on the local scale, abundance of reef manta rays ( Mobula alfredi ) at Lady Elliot Island is modulated by the tidal cycle, lower wind speeds, and the moon cycle (Jaine et al. 2012 ). Despite the broad-scale predictability of these aggregations for megaplanktivores, the specific fine-scale mechanisms that drive the formation, maintenance, and predictability of these high-quality dense zooplankton patches remain poorly resolved. Being obligate ram-feeders, M. alfredi are sensitive to these dynamics as their foraging efficiency is dependent on prey density (Paig-Tran et al. 2013 ; Stevens 2016 ). Unlike other elasmobranchs capable of stationary feeding (e.g., vertical/or suction feeding; (Rohner and Prebble 2021 ), M. alfredi must maintain a minimum swimming speed to move water over their branchial filters (Paig‐Tran et al. 2013). Consequently, they are energetically tethered to specific oceanographic retention zones that can concentrate prey above a critical threshold (Armstrong et al. 2016 ; Armstrong et al. 2021b ). As large, highly mobile planktivores residing in oligotrophic waters, M. alfredi rely on ephemeral productivity peaks to locate dense zooplankton patches (Armstrong et al. 2016 ; Armstrong et al. 2021b ). Their conservative life history is characterised by slow growth, late maturity, and low fecundity (Dulvy et al. 2014 ; Stevens 2016 ; Lawson et al. 2017 ; Marshall et al. 2022 ; Froman et al. 2023 ), making M. alfredi highly vulnerable to anthropogenic pressures, including climate-induced prey shifts, targeted fisheries, bycatch, and unsustainable tourism (Richardson 2008 ; Stewart et al. 2018 ; Murray et al. 2020 ; Jabado et al. 2024 ; Venables et al. 2024 ). While population declines have been reported globally (for review, see Laglbauer et al. 2026 ), long-term monitoring of aggregation dynamics remains scarce. Critically, few studies have described the underlying zooplankton dynamics alongside direct observations on the manta rays themselves, leaving a gap in our understanding of the environmental thresholds sustaining these threatened populations. The Republic of Maldives supports the world’s largest known population of M. alfredi (Kitchen-Wheeler et al., 2012; Stevens, 2016 ), with Hanifaru Bay, located within eastern Baa Atoll, hosting the largest feeding aggregation. Indeed, aggregations regularly exceed 50 individuals and occasionally exceed 200 individuals concurrently (Harris et al., 2020 ). Beyond its ecological importance (Harris et al., 2020 ; Harris and Stevens, 2021 ), the site is a major economic asset for the Maldives’ tourism sector (Murray et al. 2020 ; Moloney et al. in press). However, the highly predictable nature of these aggregations increases their susceptibility to localised threats, including habitat degradation (Stevens and Froman 2018 ) and tourism pressure (Murray et al. 2020 ). Further, climate projections suggest warming and weakening monsoonal winds may shift the Indian Ocean toward an increasingly oligotrophic state (Roxy et al. 2015 ; Dalpadado et al. 2024 ), potentially leading to a decline in prey biomass in the Maldives (Roxy et al. 2016 ). The climate of the Maldives archipelago is regulated by the South Asian Monsoon system, which strongly influences the seasonal distribution of the resident M. alfredi population (Anderson et al. 2011 ; Harris et al. 2020 ). During the Southwest Monsoon (SW; Hulhangu , May–November), strong winds induce localised deep-water upwelling on the leeward side of the archipelago, enhancing chlorophyll- a concentrations and productivity (Sasamal 2006 ). On a finer scale, lunar-driven spring tides transport nutrient- and plankton-rich oceanic waters onto the atolls and into shallow lagoons and embayments, facilitating the retention of dense zooplankton concentrations (Stevens 2016 ). Although these macro- and meso-scale processes are recognised, the oceanographic and environmental conditions that concentrate the zooplankton required to trigger these mass-foraging events remain poorly understood. Here, we use an 18-year dataset to investigate the fine-scale environmental and oceanographic drivers of M. alfredi habitat use and zooplankton prey dynamics at Hanifaru Bay. Long-term datasets have proven valuable for investigating the spatial ecology of species with high site fidelity, such as manta rays (Rohner et al. 2025 ). Using this dataset, we quantify the influence of multi-scale physical variables on prey availability and predator recruitment using Generalised Linear Models (GLMs) at Hanifaru Bay. Specifically, we ask: (1) What are the M. alfredi sighting trends over the 18-year period, and do they correspond with temporal changes in zooplankton biomass? (2) Which specific environmental and oceanographic predictors (including lunar phase, tidal state, and wind conditions) drive mass aggregation events of M. alfredi ? (3) To what extent do these physical conditions also drive the localised concentration of zooplankton prey? By resolving these questions, we aim to improve the understanding of the ecological triggers behind these globally significant foraging events, ultimately informing targeted conservation strategies for critical habitats facing increasing anthropogenic pressure and climatic instability. Materials and Methods Study location The Republic of Maldives is an archipelago of 26 geographical atolls situated south of India in the Indian Ocean, extending 870 km latitudinally to just below the equator (Harris et al. 2020 ) (Fig. 1 ; inset). This study was conducted in Baa Atoll, a UNESCO Biosphere Reserve located in the central double-chain atoll zone of the archipelago (Murray et al. 2020 ; Baa Atoll Biosphere Reserve 2021 ). Between 2007 and 2024, across both the SW and Northwest (NW) Monsoons, a total of 13,109 surveys were conducted across 59 sites within Baa Atoll. The data set was analysed to identify key sites, defined as having recorded ≥ 50 M. alfredi sightings during the study. These sites were distributed along a 44 km stretch of coastal waters. The primary study site was Hanifaru Bay (5.1733°N, 73.145°E), located within the Hanifaru Marine Protected Area (MPA), a small cul-de-sac-shaped reef system that ephemerally concentrates zooplankton (Armstrong et al. 2021b ) (Fig. 1 ). Hanifaru Bay was selected for intensive monitoring as it supports the highest recorded density of M. alfredi sightings within the Maldives (Harris et al. 2020 ). During the SW Monsoon (May–November), the site hosts large aggregations of foraging megaplanktivores, including M. alfredi (Stevens 2016 ; Harris et al. 2020 ; Armstrong et al. 2021b ; The Manta Trust 2024 ) and R. typus (Neves 2009 ). This study used sighting and prey data collected at this site to investigate drivers of M. alfredi habitat use and behaviour at this location. Study design Manta ray sightings and behaviour To assess long-term trends in M. alfredi abundance, we used an 18-year logbook dataset curated by the Maldives Manta Conservation Programme and The Manta Trust ( https://www.mantatrust.org/maldives ; The Manta Trust 2024 ). Long-term logbook data (hereafter referred to as logbook data) offer a cost-effective approach to amassing extensive long-term datasets. They are a common method used to investigate trends in sightings, movement, and behaviour of megaplanktivores (see Watson et al. 2001 ; Jaine et al. 2012 ; Rohner et al. 2013 ; Armstrong et al. 2016 ; Venables et al. 2024 ). When integrated with in situ or remotely sourced environmental data, these datasets can provide insights into decadal-scale patterns of occurrence, habitat use, and species behaviour, which are critical for effective, long-term conservation (Rohner et al. 2013 ). This dataset includes both M. alfredi presence and absence data. While surveys were conducted year-round, effort was intensified during the SW Monsoon to coincide with seasonal peaks in M. alfredi occurrence (Harris et al. 2020 ). Trained field researchers conducted 13,109 surveys to collect this information over 18 years in eastern Baa Atoll using three primary methods: snorkelling, scuba diving, and boat-based spotter surveys. During surveys, searches over the reef consistently followed the same track, with boat-based surveys kept to a speed of 6–9 knots and observers monitoring the surrounding water. A cumulative 8,995 survey hours were expended, representing 24,209 total observer hours. Effort was standardised in the GLMs by including the duration of each survey in Baa Atoll. The mean survey duration varied across different survey types: snorkelling surveys = 128.0 ± 2.02 minutes, n = (± standard error); scuba diving surveys = 83.2 ± 13.1 minutes, n = ; and boat-based surveys = 45.7 ± 7.44 minutes, n =. Note that all surveys were capped at a maximum of 300 minutes and the number of observers (median = 3 people) was capped at a maximum of 5 people. There was interannual and monthly variability in survey effort, with a mean of 17.8 ± 0.8 surveys per month (ranging between 1–37 surveys) and 115.8 ± 10.5 surveys per year (ranging between 29–185 surveys). At Hanifaru Bay, the median number of observers was three (ranging from 1–4) and the mean survey duration was 125.0 ± 1.98 minutes (varying between 2–300 minutes). To minimise double counting of individuals at the same event, individual M. alfredi were identified using ventral photo-identification (photo-ID), a reliable method for distinguishing individuals based on unique pigment patterns that remain largely unchanged throughout their lives (Marshall et al. 2011 ). During each survey, an “estimated count of sightings” (total individuals observed) and a “confirmed count of sightings” (unique individuals verified via photo-ID) were recorded. In mass aggregation events where observer limitations, environmental conditions, or animal behaviour might limit photo-ID efficiency (e.g., difficult to ID all manta rays individually), we used the confirmed count of sightings unless it deviated by ≥ 50% from the estimated count of sightings, in which the estimated count was adopted to avoid under-representation of abundance when there were many manta rays present and photo-ID of all individuals was difficult (see Online Resource. 1). This is referred to as the abundance of M. alfredi . Primary behaviour was categorised as (1) feeding, (2) cleaning, (3) cruising, or (4) courtship, according to the dominant behaviour during the encounter recorded. Environmental data To investigate the physical drivers of M. alfredi habitat use and prey availability (zooplankton), a suite of predictor variables was recorded in situ concurrently with each survey. These included sea state, weather conditions, current strength, and in-water visibility (Table 1 ). To ensure data quality and observer safety, surveys were restricted to conditions below a Beaufort scale of 7 (i.e., near gale conditions). These field observations were supplemented with remotely sensed data obtained from online databases (e.g., Copernicus, Meteoblue, NOAA and FES2014) such as wind speed/strength, SST, high tide (time and size), lunar phase and IOD index (Table 1 ). We selected predictors for the statistical models based on their hypothesised influence on tropical productivity or previously documented relationships with tropical megaplanktivore aggregations globally (Table 1 ). Zooplankton density Local prey availability was assessed during surveys using the Zooplankton Visual Index (ZVI), a semi-quantitative scale used to categorise plankton density in situ (Moloney et al. in press). The ZVI provides a rapid, reliable proxy for prey density that is standardised and validated (Moloney et al. in press). Additionally, horizontal in-water visibility was recorded as a proxy for turbidity and phytoplankton blooms, which may influence filter-feeding efficiency or predator detection (Table 1 ). Table 1 Description and data source of predictor variables that were used in the Generalised Linear Model Variable Type Explanation Grouping Source Levels (for categorical variables) Rationale Mobula alfredi sightings Response (count) Number of individual M. alfredi per survey Biological Identification (ID) photos (confirmed) and logbook count (estimated) Sightings from photo-identification (ID) and logbook data to investigate trends in sightings and movement (Watson et al. 2001 ; Jaine et al. 2012 ; Rohner et al. 2013 ; Armstrong et al. 2016 ; Venables et al. 2024 ). Zooplankton density (binary, 0 or 1) Response (categorical) In-water assessment of Zooplankton Visual Index (ZVI) Biological Logbook Low (0) : Levels (0) complete absence of zooplankton or (1) thin layer or small patch High (1) : Levels (2) multiple layers or patches, (3) water is thick and cloudy, felt on skin or (4) water is dense and ‘soup-like’ Zooplankton density (low or high) as an index for food availability (Moloney et al. in press). Zooplankton Visual Index (ZVI) Predictor (categorical) In-water assessment of Zooplankton Visual Index Biological Logbook (0) complete absence of zooplankton, (1) thin layer or small patch, (2) multiple layers or patches, (3) water is thick and cloudy, felt on skin, (4) water is dense and ‘soup-like’ Prey availability influences manta ray abundance and behaviour (Rohner et al. 2013 ; Armstrong et al. 2016 ; Armstrong et al. 2021b ; Venables et al. 2024 ; Moloney et al. in press). Water visibility (m) Predictor (categorical) The underwater visibility was estimated horizontally at 5 m increments Biological Logbook 0–5, 5–10, 10–15, 15–20, 25+ High water visibility shown to influence devil ray sightings, likely due to higher probability of being sighted (Venables et al. 2024 ). Weather Predictor (categorical) The weather was assigned to a category during each survey Environmental Logbook (1) clear skies and sunshine, (2) partial cloud cover, (3) overcast, (4) light rain, (5) heavy rain, (6) torrential rain Sunny weather conditions can influence sighting numbers of manta rays (Venables et al. 2024 ). Sea state (Beaufort scale) Predictor (categorical) The sea state was recorded during each survey using the Beaufort scale (0–7) Environmental Logbook (0) dead calm (sea like mirror), (1) light air (ripples without crests), (2) light breeze (small wavelets), (3) gentle breeze (large wavelets), (4) moderate breeze (small waves), (5) fresh breeze (moderate waves), (6) strong breeze (large waves), (7) near gale (sea heaps up) Sightings of M. alfredi shown to be higher during a medium swell (Venables et al. 2024 ). Sea state is a combination of swell and wind. Wind direction (0–360°) Predictor (continuous) Wind direction at 10 m elevation measured hourly. Environmental Meteoblue (Meteoblue 2025 ) Wind direction shown to influence M. alfredi abundance (Anderson et al. 2011 ; Setyawan 2016 ; Couturier et al. 2018 ; Harris et al. 2020 ) and behaviour (Jaine et al. 2012 ), including at Hanifaru Bay (Harris and Stevens 2021 ). Wind speed (km/h) Predictor (continuous) Wind strength at 10 m elevation measured hourly Environmental Meteoblue (Meteoblue 2025 ) Wind speed can influence M. alfredi abundance (Couturier et al. 2018 ; Harris et al. 2020 ) and behaviour (Jaine et al. 2012 ), including at Hanifaru Bay (Harris and Stevens 2021 ). Current direction Predictor (categorical) In-water assessment visually in relation to the reef Oceanographic Logbook (In) into the channel, (out) out of the channel, (slack) no direction Current direction shown to influence M. alfredi abundance and behaviour (Jaine et al. 2012 ; Rohner et al. 2013 ; Harris et al. 2020 ; Venables et al. 2024 ) but was removed from the models due to confounding variables. Current strength Predictor (categorical) The current strength was assessed underwater based on effort for kicking to remain in the same place Oceanographic Logbook (0) none = able to hold position without finning, (1) slight = able to hold position with little effort, (2) moderate = able to hold position with strong finning, (3) strong = unable to hold position Current strength influences manta ray abundance and behaviour (Rohner et al. 2013 ; Harris et al. 2020 ; Ahsin et al. 2022 ; Venables et al. 2024 ). Sea Surface Temperature (SST) (° degrees Celsius) Predictor (continuous) Water temperature on the reef. Filtered Reprocessed data: 2007–2021. Filtered NRT: 2022–2024. Oceanographic Copernicus (Worsfold et al. 2024 ) SST shown to influence manta ray sightings (Jaine et al. 2012 ; Weeks et al. 2015 ; Setyawan 2016 ; Beale et al. 2019 ; Ahsin et al. 2022 ; Fonseca-Ponce et al. 2022 ). Indian Ocean Dipole (IOD) Index (-1 + + 1 Predictor (continuous) Index value of warmer or cooler water in the Indian Ocean Oceanographic NOAA (NOAA Physical Sciences Laboratory 2024 ) IOD index shown to affect M. kuhlii sightings in Mozambique (Venables et al. 2024 ). Time from High Tide (± 6 hours) Predictor (continuous) The time difference between the observation (middle of the survey) and closest high tide Oceanographic FES2014 (Lyard et al. 2021 ) and oce R package (Kelley and Richards 2024 ) Tide shown to influence zooplankton prey (Cairns et al. in press), and M. alfredi abundance and behaviour (Dewar et al. 2008 ; Jaine et al. 2012 ; Armstrong et al. 2016 ; Couturier et al. 2018 ), including at Hanifaru Bay (Armstrong et al. 2021b ). Tide data obtained using a harmonic tidal model with oce R package (Kelley and Richards 2024 ). Tidal constituents (i.e., m2, s2, n2, k2, k1, o1, p1, q1, m4, ms4, m6, l2, t2, s4, mn4) obtained from FES2014 (Lyard et al. 2021 ) and used to predict high/low tide and time. Tide epoch was set at: 1992-01-01. Tide size (m) Predictor (continuous) Tide size difference between high and low tides on the day (from 0 m) Oceanographic FES2014 (Lyard et al. 2021 ) and oce R package (Kelley and Richards 2024 ) Greater tidal ranges may interact with complex nearshore bathymetry, increasing mixing in the water column and potentially promoting foraging opportunities for M. birostris (Fonseca-Ponce et al. 2022 ). Lunar illumination (Proportion 0–1) Predictor (proportion) Proportion of moon disk illuminated (0) new – (1) full Oceanographic Package R Lunar package (Lazaridis 2022 ) Moon phase/lunar illumination known to influence M. alfredi abundance (Dewar et al. 2008 ; Jaine et al. 2012 ; Setyawan 2016 ; Couturier et al. 2018 ; Andrzejaczek et al. 2020 ; Harris and Stevens 2021 ; Fonseca-Ponce et al. 2022 ). Month Predictor (categorical) Month of observation Temporal Logbook For Baa Atoll : Jan–Dec For Hanifaru Bay : May–Nov Month used to examine seasonality of abundance due to changes in monsoon over the season (Harris et al. 2020 ). Year Predictor (categorical) Year of the observation Temporal Logbook Year used to examine trends in M. alfredi abundance over time (Rohner et al. 2013 ; Venables et al. 2024 ). M. alfredi behaviour Predictor (variable not in model; categorical) Primary behaviour of individual M. alfredi per survey Biological Logbook (1) feeding; (2) cleaning; (3) cruising; and (4) courtship Behavioural observations from logbook data used to investigate habitat use (Jaine et al. 2012 ; Rohner et al. 2013 ; Harris et al. 2020 ). Survey duration (minutes) Offset variable (continuous) Duration of the survey NA Logbook Capped at 300 minutes To account for survey effort Observers Offset variable (count) Number of in-water observers NA Logbook Capped at 4 observers To account for the survey effort Statistical analysis To assess the trends and drivers of M. alfredi abundance and zooplankton density in Hanifaru Bay, we analysed the researcher-collected logbook survey data spanning 2007 to 2024. Generalised Linear Models (GLMs) were built using the statistical software R version 4.4.2 (R Core Team 2024 ) to model the data from the logbook and remote sources, including a suite of temporal (month, year), environmental (weather, sea state, wind direction and strength), oceanographic (current strength, SST, time to high tide, tide size, lunar illumination, IOD index), and biological (ZVI, water visibility) predictors (Table 1 ). During the SW Monsoon, there was commonly one survey per day (occasionally two), six days per week. Opportunistic surveys conducted between December and April were excluded to maintain consistency in search effort. The same researcher-collected dataset (2010 to 2024) was used to investigate zooplankton density trends and drivers; the shorter temporal range for zooplankton reflects the lack of records prior to 2010. Four separate GLMs were built: “ Manta trends ” and “ Manta drivers ” (response: M. alfredi abundance modelled as a count variable), and “ Zooplankton trends ” and “ Zooplankton drivers ” (response: binary zooplankton abundance modelled as a binary variable). For all four GLMs, model residuals were visually assessed for normality and homogeneity of variance. Trend models To investigate the annual trends and seasonality at Hanifaru Bay, we constructed two models, one for " Manta trends " and the other for " Zooplankton trends " using temporal predictors (Month, Year). " Manta trends " model : glm.nb(Manta count ~ Year + Month + Year:Month) " Zooplankton trends " model : glm(Zooplankton density ~ Year + Month + Year:Month, family = binomial) For the " Manta trends " model, M. alfredi abundance was treated as a continuous count variable. To account for overdispersion common to ecological count data (evident from an initial model using a Poisson error structure), the model was fitted with a negative binomial error structure using the glm.nb() function in the MASS R package (Venables and Ripley 2022 ). To account for sampling biases inherent in logbook data, we standardised for survey effort by including survey duration and the number of observers as offsets in the “Manta trends” and the “Manta drivers” models. For the " Zooplankton trends " model, zooplankton density was treated as a binary response (0 = ZVI levels 0,1, and 1 = ZVI levels 2,3,4) and thus modelled as a binomial GLM with a logit link function. To define the degrees of freedom (df) for smoothing terms in the GLM, it was conservatively set at df = 4. Driver models Two additional GLMs were constructed to identify the primary influences on M. alfredi abundance (“ Manta drivers ”) and high zooplankton density (“ Zooplankton drivers ”) “ Manta drivers ” model glm.nb(Manta count ~ ZVI + Water visibility + Lunar illumination + SST + Sea state + Wind speed + Wind direction + Hours from high tide + Current strength + IOD index). “ Zooplankton drivers ” model glm(Zooplankton density ~ Hours from high tide + Current strength + IOD index, family = binomial). Both driver models initially considered the same global suite of environmental and oceanographic predictors (Table 1 ). We employed a hierarchical backward-selection approach to determine the most parsimonious model, beginning with a full model containing all predictors. Predictors were removed one-by-one, and model fit was reassessed at each step by identifying the candidate model with the lowest Bayesian Information Criterion (BIC) using step(), ensuring an optimal balance between explanatory power and model complexity. The significance of each model term in the final model selection was then confirmed by using anova() and likelihood ratio tests to ensure each remaining variable significantly improved the model fit. The “ Manta drivers ” model included ZVI with 5 levels, but ZVI was excluded as a predictor in the “ Zooplankton drivers ” model. To define the degrees of freedom (df) to smooth SST and wind speed in the GLM, it was conservatively set at df = 3, unless the GLM indicated that df = 4 more appropriately defined a relationship. A harmonic function ( k = 1) was applied to circular predictors (hour from high tide, wind direction, and lunar illumination) to account for their cyclic nature. Nagelkerke pseudo- R² values were calculated for all four final models to assess the goodness-of-fit (Nagelkerke 1991 ). Results Abundance and behaviour of Mobula alfredi in Baa Atoll From the total sightings of M. alfredi in Baa Atoll ( n = 71,341), 21 key sites were identified. However, most of these sightings (72.3%; n = 51,555) were recorded at one location, Hanifaru Bay (Fig. 2 ), and the remaining 27.7% of the sightings were spread across 58 additional sites. Consequently, subsequent analyses focus exclusively on Hanifaru Bay, as the disproportionately high sightings frequency and high survey replication at this site provide the statistical power necessary to minimise modelling variability. Most sightings at Hanifaru Bay were during the SW Monsoon between May–November (99.4%), where the most frequently observed primary behaviour was feeding (70.5%), with cleaning being the second most common (22.9%), and cruising (8.3%) and courtship (2.3%) behaviours observed less frequently (Fig. 3 ). Mobula alfredi were seen on 82.5% of surveys, with observers counting a mean of 24.6 ± 0.8 SE and up to 244 individual M. alfredi per survey. Monthly and inter-annual “ Manta trends ” at Hanifaru Bay At Hanifaru Bay, M. alfredi abundance – derived from the “ Manta trends ” model for surveys conducted between May–November (2007–2024) and standardised for the number of observers and survey duration – was significantly related to Month and Year (p < 0.001). A significant interaction between month and year (p < 0.001) indicated that annual peaks in abundance fluctuated seasonally. While the model identified a historical peak in M. alfredi abundance in 2007 (driven largely by the high September sightings), analysis of the 2010–2024 period revealed distinct peaks in 2021 and 2022 (predicted mean 38.6 and 37.7 individuals). In these years, specific months, including May, June, October, and November, reached their highest predicted abundances (Fig. 4 a). Conversely, the lowest abundance in this period was in 2017. For monthly trends, the model predicted a peak in M. alfredi sightings generally during August (35.5 individuals), followed by September (33.5 individuals). Monthly and inter-annual “ Zooplankton trends ” At Hanifaru Bay, binary zooplankton density from the “ Zooplankton trends ” model spanning May–November in the years between 2010–2024 was significantly related to Year (p < 0.001) but not to Month (p = 0.595). There was, however, a significant interaction between Month and Year (p < 0.001). Like the M. alfredi models, zooplankton density fluctuated seasonally, with a predicted peak in 2022 (Fig. 4 b). Almost all months – excluding July and August – reached their maximum values between 2021 and 2022. The model also identified the lowest density of zooplankton to be in 2017. There were also monthly trends with peaks in November (0.431), followed by August (0.426). Model for “ Manta drivers ” The model of best fit for “Manta drivers” included ten predictors (zooplankton density, water visibility, lunar illumination, hours from high tide, current strength, SST, sea state, wind speed and direction, and IOD index). This model explained 95.2% of the variance for the predictors, with the dominant predictor, zooplankton density (ZVI), explaining 61.1% (using Nagelkerke R²). While the remaining variance was attributed to secondary environmental and oceanographic drivers, such as sea state, followed by SST, and wind direction and speed. The ZVI was a strongly significant predictor of M. alfredi abundance (p < 0.001), with an increasing trend in sightings with higher zooplankton density (Fig. 5 ). Comparing the predicted count for the highest recorded zooplankton density (ZVI 4) to the complete absence of zooplankton (ZVI 0), predicted manta abundance was 2,590.5% higher during ZVI 4 conditions. Further, predicted abundance was 311.4% higher during ZVI 4 conditions compared to minimal zooplankton presence (ZVI 1). Water visibility significantly influenced abundance (p 20 m) and by 13.3% in highly turbid conditions (0–10 m, i.e., when there was extensive phytoplankton and/or sediment in the water column). The predicted M. alfredi abundance was significantly influenced by the lunar illumination cycle (p < 0.001). The abundance rate varied throughout the lunar month, with predicted abundance peaking around the new moon (0.1 proportion), followed by the full moon (1.0). Abundance during a new moon was predicted to be 22.2% higher compared to a full moon. The number of predicted M. alfredi abundance varied significantly with the tidal cycle (hours from high tide; p < 0.001). Abundance peaked 2.7 hours after high tide and reached a minimum 3.3 hours before high tide. Predicted abundance was 33.4% higher during the optimal tidal state compared to the minimal tidal state. Abundance was also predicted to be 54.8% higher in a stronger current than in no current and 84.6% higher than when there was a weak current (p < 0.001). Sea state was a significant predictor of M. alfredi abundance (p < 0.001), with abundance predicted to increase by 564.9% when there was a light breeze (Beaufort scale 1) compared to calm waters (Beaufort scale 0). The magnitude of the positive effect generally declined as the sea state increased to a strong breeze. Wind speed had a non-linear significant influence on predicted M. alfredi abundance (p < 0.001), with sightings peaking at 0 km/h and again at 30.3 km/h. Sightings dropped when the wind speed was lower (13.3 km/h). Predicted abundance was 43.2% higher during the optimal wind conditions compared to the minimal wind conditions. The predicted abundance of M. alfredi was also significantly influenced by the wind direction (p < 0.001). The abundance peaked when the wind was north-northwest (NNW) in direction (345.0°) and decreased when the wind came from the south-southeast (SSW; 165.5°). Predicted abundance was 66.1% higher during the optimal wind direction compared to the minimal wind direction. Warmer waters had a significant negative effect on M. alfredi abundance (p < 0.001). While sightings peaked around 28.6°C, predicted abundance declined sharply at warmer temperatures, declining by 24.4% for the first 1.0°C and 15.7% per degree after this. When comparing the optimal conditions to the highest observed temperature of 31.1°C, the predicted abundance of M. alfredi was 45.0% lower. The IOD index also significantly impacted the abundance of M. alfredi (p < 0.001), with a 26.0% increase in sightings when it moved from a negative IOD phase to a positive IOD phase ( ß =0.175). While the univariate model showed a negative correlation between IOD and M. alfredi abundance ( ß =-0.147). Model for “Zooplankton drivers” The “ Zooplankton drivers ” model with a binary response (0 = low ZVI and 1 = high ZVI) densities included three predictors (hours from high tide, current strength, and IOD index). This model explained 8.5% of the variance for the predictors (using Nagelkerke R²). The strongest predictor was current strength explaining 4.6% of the variance, while the remaining variance was attributed to secondary oceanographic drivers such as hours from high tide, followed by the IOD index. Current strength was positively correlated with high zooplankton density (p < 0.001; Fig. 6 ), with densities predicted to be 335.9% higher when there was a stronger current than in no current and 32.8% higher than when there was a weak current. Zooplankton density varied significantly with the tidal cycle (hours from high tide; p < 0.001). The odds of observing high density peaked at 1.6 hours after high tide and reached a minimum 4.4 hours before high tide. The predicted odds of high-density zooplankton were 130.3% higher during the optimal tidal state compared to the minimal tidal state. The IOD index significantly impacted zooplankton density (p < 0.001), with a 72.8% increase in density when it moved from a strong positive IOD phase to a negative IOD phase. Discussion Our 18-year dataset (2007–2024) confirms that Hanifaru Bay functions as a critical foraging aggregation site during the SW Monsoon, accounting for 71% of all feeding-related sightings recorded in eastern Baa Atoll. The high degree of synchrony between M. alfredi abundance and prey density — both peaking in the later months of the season and reaching a decadal maximum in 2021–2022 — underscores a tight coupling between biological activity and environmental forcing. Our models suggest the utility of this hotspot is regulated by two primary tiers of influence: (1) local-scale biophysical drivers, including lunar and tidal phases, wind-driven currents, and SST; and (2) broad-scale climatic indices, such as the IOD. While the bathymetry of Hanifaru Bay is a constant feature, its retention of zooplankton fluctuates in response to these local and regional indices which modulate the local environmental conditions (e.g., wind and SST) required to trigger mass-foraging events. Given the reliance of M. alfredi on these high-density, ephemeral zooplankton patches, climate-driven shifts in oceanographic regimes may fundamentally alter the functional productivity of this site, impacting the energetic fitness of this vulnerable population (Stewart et al. 2018 ). Long-term trends reveal interannual and seasonal variability This nearly two-decade study reveals interannual and seasonal fluctuations in M. alfredi abundance, reflecting the dynamic nature of the Maldivian preyscape, though annual sightings remained relatively steady. Seasonally, the August/September peaks aligns with the maturation of the SW Monsoon, a period characterised by sustained upwelling and productivity (Anderson et al. 2011 ; Harris et al. 2020 ). While similar seasonal pulses are documented in other megaplanktivores globally (Witt et al. 2012 ; Rohner et al. 2013 ; Klotz et al. 2017 ; Couto et al. 2017 ; Harvey-Carroll et al. 2021 ; McInturf et al. 2022 ), the magnitude of the Hanifaru Bay aggregations remains globally unparalleled. The predictable recurrence of these aggregations is underpinned by high site fidelity, as photo-ID records confirm individuals returning to the same reef system across many years (Marshall et al. 2011 ; Couturier et al. 2014 ; Harris et al. 2020 ). While site fidelity represents phenotypically plastic responses to habitat quality, and niche specialisation, the consistent return of M. alfredi individuals to Hanifaru Bay suggests a high-fidelity reliance on the site’s unique foraging opportunities (Armstrong et al. 2021a ; Knochel et al. 2022 ). This behavioural strategy allows M. alfredi – and other large megaplanktivores such as M. birostris and R. typus – to maximise the exploitation of high quality foraging niches within a patchy marine environment (Rohner et al. 2013 ; Guzman et al. 2022 ; Rubin et al. 2025 ). Consequently, local abundance at Hanifaru Bay serves as a high-fidelity proxy for the efficiency of the site’s physical mechanisms. While morphology and bathymetry remains constant, the formation of this retention zone is governed by broader Indian Ocean productivity (Roxy et al. 2016 ). Interannual variability likely corresponds to large-scale climatic indices, specifically the IOD and the ENSO, which can modulate regional sea levels, rainfall, and thermal profiles (Saji et al. 1999 ; Carruthers et al. 2024 ). The 2021–2022 peak coincided with a record-breaking Negative IOD and a triple-dip La Niña event (starting mid-2020 and continuing till 2022, with 2022 being the strongest on record), which intensified the westerly monsoonal winds (Srivastava et al. 2026 ) and maximised the wind-driven concentration of prey. The Negative IOD, persisting from mid-2021 through 2022, was the first ever multi-year of its kind since the 1960s, occurring alongside wetter conditions and higher sea levels, with wind anomalies also mainly remaining westerly (Chowdhury et al. 2025 ; Srivastava et al. 2026 ). In contrast, the 2017 trough highlights ecosystem vulnerability following a strong El Niño event (Cowburn et al. 2019 ; Chowdhury et al. 2025 ) and a Positive IOD phase (NOAA Physical Sciences Laboratory 2024 ). Notably, 2016 was also associated with mass coral bleaching and thermal anomalies in the Maldives (Cowburn et al. 2019 ). During this period, the Maldives experienced dry spells and lower than average sea levels (Chowdhury et al. 2025 ), resulting in extreme surface warming which likely suppressed vertical mixing and disrupted prey patch formation required for mass-foraging (Min and Noh 2004 ; Roxy et al. 2016 ). As climatic events increasingly squeeze the regional preyscape, ensuring that mobile planktivores can forage with limited anthropogenic disturbances at a hotspot like Hanifaru Bay is essential for population resilience. Zooplankton density is the primary predictor of Mobula alfredi abundance High zooplankton density (using ZVI) was the dominant predictor of M. alfredi abundance, explaining > 60% of the observed variation in sightings. Sightings were > 2,500% higher during peak prey availability compared to periods of apparent prey absence. This strong correlation aligns with global observations of megaplanktivores targeting localised productivity and dense zooplankton patches (Jaine et al. 2012 ; Rohner et al. 2013 ; Armstrong et al. 2016 ; Armstrong et al. 2021b ; Venables et al. 2024 ). However, our findings suggest the attraction to Hanifaru Bay may extend beyond prey volume, with this site likely functioning as a location for metabolic optimisation. Foraging in the deep, cold offshore waters imposes thermoregulatory costs, requiring specialised counter-current heat exchange systems in the cranial and branchial regions to maintain core temperatures (Arostegui 2025 ). Hanifaru Bay is unique because the physical mechanisms transport and concentrate deep-water prey into shallow, sun-warmed waters. This allows M. alfredi to exploit high density prey while simultaneously reducing the metabolic cost of thermoregulation (Arostegui 2025 ). By mitigating the thermal stress typically associated with deep-water foraging (Beale et al. 2025 ), the site provides an energetic surplus that likely facilitates the high levels of social and secondary behaviours observed, such as cleaning and courtship (Stevens 2016 ), while also reducing their exposure to predators. Further, M. alfredi presence was negatively impacted by high content of phytoplankton and/or suspended sediment–similar to observations in Mozambique (Rohner et al. 2013 ), suggesting water quality is a critical constraint. High sediment loads may impose significant physiological costs, potentially impeding gill ventilation and reducing filter-feeding efficiency, thereby driving mobile planktivores to seek clearer, prey-rich environments (Rohner et al. 2013 ). Consequently, rather than searching wide areas for diffuse prey, M. alfredi appears energetically tethered to Hanifaru Bay. The site offers a rare intersection of high-density prey, clear water, and thermal refuge, allowing for maximising efficient foraging and high calorific intake, while minimising the metabolic and physiological tolls associated with search effort and environmental stress (Armstrong et al. 2016 ). Mobula alfredi abundance: Lunar, tidal, and wind-driven convergence Beyond direct prey density, M. alfredi abundance is modulated by an interplay of lunar, tidal, and wind-driven mechanisms that facilitate the retention of zooplankton. We observed a > 30% increase in sightings during the new and full moon phases; a pattern consistent with foraging aggregations on the Great Barrier Reef in Australia and in Komodo, Indonesia, where activity is timed with spring tides (Dewar et al. 2008 ; Jaine et al. 2012 ). At Hanifaru Bay, tidal currents typically shift 180 degrees between phases, transitioning from an incoming flood (east–west) and an outgoing ebb (west–east). This lunar influence is intrinsically linked to current velocity; sightings at Hanifaru Bay were ~ 50% higher during high-velocity currents, which likely facilitate the transport of oceanic zooplankton into the reef system. Abundance peaked 2.7 hours after high tide, when intensified outgoing lunar and high tide currents appear to overcome prevailing monsoonal flows, forcing plankton-rich oceanic water into the shallow reef systems of the atoll (Stevens 2016 ). Optimal foraging conditions were associated with NNW winds (25–35 km/h), which trigger the development of high-energy Langmuir Circulation (LC) – a series of shallow, wind-driven, counter-rotating cells that concentrate buoyant zooplankton into visible surface slicks (Li 2000 ; Smith 2001 ; Thorpe 2004 ). At ~ 30 km/h, such as what we have reported in Hanifaru Bay, the LC cells can extend 5–10 m deep, concentrating a larger volume of the water column, and bringing prey to the surface (Thorpe 2004 ). Because Hanifaru Bay opens to the NNW, these LC convergence zones align with the Bay’s primary axis, effectively accumulating prey into a bathymetric cul-de-sac where reef morphology prevents downwind dispersal and maximises prey density (Li 2000 ). Thermal thresholds and climatic sensitivity Despite the efficiency of the Hanifaru Bay physical retention zone, we identified clear environmental thresholds that constrain its functional utility. Mobula alfredi abundance was highly sensitive to thermal shifts, declining by 24.4% for the first 1.0°C of warming above the 28.6°C optimum. Across the observed warming range, the average decline of 15.7% per degree suggests that even minor increases in SST may lead to a disproportionate reduction in the abundance at this site. While the thermal limit in Komodo, Indonesia, is thought to be 29°C (Dewar et al. 2008 ), the M. alfredi population in the Red Sea routinely forage in higher temperatures (Stevens Pers. Obs.). This suggests that the Maldives population may be acclimatised to a localised thermal window, beyond which metabolic costs outweigh the caloric gains (Dewar et al. 2008 ; Couturier et al. 2018 ). As M. alfredi use regional endothermy and counter-current heat exchange to regulate vital organ temperatures (Arostegui 2025 ), hot SSTs may inhibit foraging by exceeding physiological preferences or impairing the ability to shed metabolic heat generated during high-intensity feeding (Pistevos et al. 2015 ). Beyond physiology, SST likely serves as a proxy for broader thermocline dynamics. The presence of oceanic-derived zooplankton within Hanifaru Bay (Armstrong et al. 2021b ) is dependent on the transport of cool, nutrient-rich waters onto the atoll’s shallow shelf. As the Indian Ocean warms and stratifies, a deepening thermocline may effectively sink this prey-laden water below the depth of the atoll’s shelf (Roxy et al. 2016 ). If the thermocline depth exceeds the bathymetric shelf of the atoll, the influx of zooplankton-rich water could be physically obstructed, potentially limiting the availability of oceanic prey at inshore sites. Consequently, while the retention zone remains constant, its function as a productive hotspot may be increasingly vulnerable to a warming oceanographic regime that threatens to decouple the predator from a primary nutrient supply chain. Oceanographic drivers of prey resources Our zooplankton models indicate that current strength, tidal phase, and the IOD index are the strongest predictors of prey density. Zooplankton density peaked ~ 1.6 hours after high tide (a 130.3% increase), supporting the hypothesis of a physical retention mechanism that transports offshore species, such as Undinula vulgaris onto the atoll (Armstrong et al. 2021b ). Once within Hanifaru Bay, unique bathymetry generates eddies that concentrate biomass (Cairns et al. in press). These eddies form midway through the flood tide and dissipate ~ 2 hours after high tide, acting as the dominant mechanism driving feeding aggregations (Cairns et al. in press). The models reveal a distinct temporal lag between prey availability and peak predator sightings; while zooplankton density peaked ~ 1.6 hours after high tide, M. alfredi abundance peaked ~ 1 hour later (~ 2.7 hours post high tide). This sequence suggests a high-fidelity behavioural response as individuals aggregate specifically as the ebbing tide bottlenecks prey within the bathymetry (Armstrong et al. 2021b ). Tidal forcing is a well-documented driver of zooplankton enrichment in other topographically complex regions, such as the Komodo National Park, Indonesia (Dewar et al., 2008 ), and the Great Barrier Reef, Australia (Armstrong et al. 2016 ). Beyond localised tidal forcing, the interannual availability of prey at Hanifaru Bay is significantly modulated by the IOD. Our models demonstrate that zooplankton densities significantly increase during Negative IOD phases, which align with intensified westerly monsoonal winds, cooler than average SSTs, and enhanced chlorophyll- a concentrations in the central Maldives (Srivastava et al. 2026 ; Abeywickrama et al. under review). This enrichment is driven by a shallower thermocline and wind-stress conditions that facilitate the upwelling of nutrient-rich subsurface waters (Roxy et al. 2016 ; Thushara and Vinayachandran 2020 ; Abeywickrama et al. under review). Conversely, during a Positive IOD, the thermocline deepens, and weaker westerly winds induce downwelling, suppressing the nitrate and phosphate transport essential for phytoplankton growth (Roxy et al. 2016 ; Abeywickrama et al. under review). Interestingly, while zooplankton increased during Negative IOD events, our multivariate model showed a decline in M. alfredi abundance—a divergence from the univariate results. This suggests that while a Negative IOD creates an optimal climatic window for prey production, predator recruitment is simultaneously constrained by local-scale factors, despite trends for the highest individual years of abundance occurring during Negative IOD phase. Although research on the interactions between ENSO and the IOD in the Maldives remains limited (Chowdhury et al. 2025 ), our data underscores the vulnerability of this hotpot to broader climatic shifts that regulate the productivity of the Indian Ocean. Implications for a changing Indian Ocean The long-term stability of this hotspot is tied to the SW Monsoon, which drives the deep-water upwelling essential for regional productivity (Sasamal 2006 ; Anderson et al. 2011 ; Harris et al. 2020 ). However, the South Asian Monsoon is increasingly influenced by anthropogenic climate change, with a documented 20% reduction in marine phytoplankton over the past six decades (Joseph and Simon 2005 ; Roxy et al. 2016 ). Rapid surface warming has led to enhanced ocean stratification that prevents nutrient-rich subsurface waters from reaching the euphotic zone (Roxy et al. 2016 ). As warming intensifies, there is an acute risk of reduced prey biomass reaching the atoll due to inhibited upwelling (Roxy et al. 2016 ; Singh 2018 ), and the breakdown of the concentration mechanisms required to aggregate prey into forageable patches (Min and Noh 2004 ). As M. alfredi are energetically tethered to these physical drivers, they serve as effective bio-indicators of regional oceanographic health. Their presence at Hanifaru Bay acts as a signal that the physical retention zone is functioning; conversely, their absence during historically productive windows may signal a shift toward a more oligotrophic state. Rising SSTs may cause retention zones to fail even during periods of stable regional productivity (Roxy et al. 2016 ), as the thermal buoyancy of the surface layer overcomes the wind-driven mixing required to form surface slicks (Min and Noh 2004 ). Climate-induced changes may force this resident population to travel further or forage deeper, increasing metabolic search costs and predation risks, while potentially impacting individual fitness and reproductive output (Stewart et al. 2018 ). Conservation relevance: spatial-temporal bottleneck As Hanifaru Bay hosts > 70% of M. alfredi sightings recorded in eastern Baa Atoll in the SW Monsoon, it represents a non-substitutable habitat. The tendency to aggregate in these confined areas increases vulnerability to vessel strikes, discarded fishing gear, pollution, and concentrated tourist groups, as a single threat may impact many individuals simultaneously (Croll et al. 2016 ; Stewart et al. 2018 ; Strike et al. 2022 ; Palacios et al. 2023 ). Designated as an MPA in 2009, Hanifaru Bay serves as the core protected zone within the Baa Atoll Biosphere Reserve (Murray et al. 2020 ; Baa Atoll Biosphere Reserve 2021 ). While management regulations (e.g., prohibition on scuba diving, mandatory licensed guides, a code-of-conduct for interacting with wildlife, and strict caps on the number of concurrent visitors and vessels) enforced by on-site rangers provide a robust framework for managing direct interactions (Murray et al. 2020 ; Baa Atoll Biosphere Reserve 2021 ), the environmental cues that trigger mass foraging also create a spatial and temporal bottleneck of anthropogenic exposure. This cumulative behavioural stress of high-density wildlife activity and intense human interest mirrors dynamics seen in terrestrial systems, such as Yellowstone National Park, where predictable seasonal movements concentrate human-wildlife conflict into localised high-risk zones (Soulsbury and White 2016 ; Middleton et al. 2020 ; Fortin et al. 2020 ). Protecting the ecosystems surrounding Hanifaru Bay may provide umbrella benefits, safeguarding the preyscapes used by other vulnerable megaplanktivores, including R. typus and other mobulid species (Stevens and Froman 2018 ; The Manta Trust 2024 ). Caveats While the dataset provides an insight into the trends and drivers of M. alfredi sightings and their prey, several limitations must be acknowledged. First, the study relies on logbook data which are subject to observer bias, including variations in survey effort and the difficulty of accuracy quantifying individuals during mass-aggregation events. As well as non-uniform sampling effort with survey frequency often fluctuating with monsoonal weather and seasonal accessibility. However, the sheer volume of sightings and the nearly two-decade duration of the study provide a robust overview that effectively smooths these individual survey anomalies and temporal variations. Second, zooplankton density was assessed using ZVI rather than quantitative biomass sampling (Moloney et al. in press). While this introduces a level of subjectivity, the ZVI is a validated and reliable proxy for the relative abundance of prey (Moloney et al. in press). Third, the use of GLMs identifies potential associations but cannot establish biological causation, particularly for the zooplankton drivers model where only a small amount of variation was explained. Hence, future work could introduce structural causal models to help disentangle the effect of current velocity from the effect of the lunar cycle (Arif and MacNeil 2023 ). Fourth, certain environmental predictors (e.g., currents, tides) were recorded at qualitatively or coarse resolution, and other potential drivers (e.g., nutrient concentrations, fine-scale oceanography) were not included. Future work could consider including these variables at a fine-scale resolution. Lastly, a notable divergence was observed where zooplankton density increased during a Netative IOD phase, while M. alfredi abundance (only in the multivariate model) declined. We are unsure why this has occurred. Conclusion Our 18-year investigation into the world’s largest aggregation of M. alfredi represents one of the longest continuous datasets of its kind, revealing a multi-scale interaction of localised oceanography and broad-scale climatic drivers. While previous research established that site-specific environmental conditions are required to trigger mass-foraging in megaplanktivores (Sleeman et al. 2010 ; Jaine et al. 2012 ; Rohner et al. 2013 ; Weeks et al. 2015 ; Harris and Stevens 2021 ), our results elucidate the specific environmental configuration of high-velocity tides, NNW winds, and spring lunar phases which drive these events at Hanifaru Bay. These factors activate a bathymetric prey retention zone where wind-induced LC and tidal eddies concentrate zooplankton into hyper-dense surface patches required for high-energy feeding (Harris and Stevens 2021 ; Armstrong et al. 2021b ; Cairns et al. in press). Crucially, our findings suggest that Hanifaru Bay functions as a site for both foraging and metabolic optimisation. By concentrating deep-water prey into shallow, sun-warmed waters, Hanifaru Bay allows M. alfredi to maximise caloric intake while minimising the thermoregulatory costs associated with deep-water foraging. However, this energetic advantage is governed by a thermal threshold; the observed 28.6°C tipping point suggests that while Hanifaru Bay’s bathymetry is a constant, its functional utility is highly vulnerable to a warming and more stratified Indian Ocean (Roxy et al. 2016 ). Rising temperatures and a deepening thermocline threaten to decouple the M. alfredi population from a primary nutrient supply, even if regional productivity remains stable. As the Indian Ocean trends towards a more oligotrophic state (Roxy et al. 2016 ), the importance of reliable foraging hotspots like Hanifaru Bay — which hosts 71% of regional sightings — will only intensify. The site represents a non-substitutable habitat and a critical spatial-temporal bottleneck where both ecological activity and anthropogenic pressure converge. By defining these oceanographic boundaries, this study provides a framework for management strategies to evolve alongside a changing climate, safeguarding the reproductive fitness and long-term resilience of this vulnerable population. Declarations Competing Interests There are no competing interests to declare from authors. Ethics Approval Approval to undertake this research was received from the Maldives Ministry of Fisheries, Marine Resources and Agriculture (annually renewable permit: [ (FRM)30-D/PRIV/2022/54 ]) and the Maldives Environmental Protection Agency (annually renewable permit: [ EPA/2023/PSR-M04 ]). This research was in accordance with the University of the Sunshine Coast Animal Ethics [ ANS23101 ]. Documentary evidence available on request. Funding Hannah M. Moloney was funded by the University of the Sunshine Coast Research Training Program scholarship. This study was made possible due to funding from the Save Our Seas Foundation, Carl F. Bucherer and logistical support and funding from the Four Seasons Resort Maldives at Landaa Giraavaru. The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Author Contributions Hannah M. Moloney, Guy M.W. Stevens, Asia O. Armstrong, Christine l. Dudgeon, Kathy, A. Townsend and Anthony J. Richardson made substantial contributions to the study conception and design. Data collection was performed by Hannah M. Moloney, Guy M.W. Stevens, Elspeth M. Strike, and Tamaryn J. Sawers. Data analysis and interpretation was performed by Hannah M. Moloney, Alvise Dabala and Anthony J. Richardson, with Kathy, A. Townsend and Guy M.W. Stevens also contributing to the interpretation of the data. The first draft of the manuscript was written by Hannah M. Moloney and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Acknowledgements The authors acknowledge the extensive data management, and logistical and field assistance from the team at the Maldives Manta Conservation Programme and The Manta Trust , specifically, Bethany Faulkner, Tiff Bond, Niv Froman, Moosa Mohamed, Annie Murrray, Simon Hilbourne, Katie Lee-Brooks, Ibrahim Lirar, Hussain Rasheed, and Yoosuf Abdul Haadhee. We thank the Four Seasons Resort Maldives team based at Landaa Giraavaru for providing ongoing support to this research project, particularly the General Manager, Armando Kraenzlin. Authors would like to thank the Maldives’ Ministry of Fisheries, Marine Resources and Agriculture and the Maldives Environmental Protection Agency (now formally known as the Environmental Regulatory Authority ), who granted permission to undertake this research. We also acknowledge the dedicated support of the Biosphere Reserve Office , particularly the sea rangers. Thank you to Save Our Seas Foundation and Carl F. Bucherer for funding the field operations. Data Availability The datasets generated and analysed during the current study – logbook data (including manta ray abundance and ZVI observations) – are available on reasonable request from the corresponding author. References Abeywickrama T, Pathirana G, Noh K-M, Dissanayake K, Wang D Lee D-G (under review) Asymmetric effects of Indian Ocean dipole on surface chlorophyll variability in the Indian Ocean Ahsin A, Hartati R, Sitorus ED, Azizah H, Endrawati H (2022) Oceanographic factors on coastal aggregation of Reef Manta ( Mobula alfredi ) in the Manta Sandy, Raja Ampat, Indonesia. ILMU Kelaut Indones J Mar Sci 27(4):330–340. https://doi.org/10.14710/ik.ijms.27.4.330-340 Anderson RC, Adam MS, Goes JI (2011) From monsoons to mantas: seasonal distribution of ( Mobula alfredi ) in the Maldives. Fish Oceanogr 20(2):104–113. https://doi.org/10.1111/j.1365-2419.2011.00571.x Andrzejaczek S, Chapple T, Curnick D, Carlisle A, Castleton M, Jacoby D, Peel L, Schallert R, Tickler D, Block B (2020) Individual variation in residency and regional movements of reef manta rays Mobula alfredi in a large marine protected area. Mar Ecol Prog Ser 639:137–153. https://doi.org/10.3354/meps13270 Arif S, MacNeil MA (2023) Applying the structural causal model framework for observational causal inference in ecology. Ecol Monogr 93(1):e1554. https://doi.org/10.1002/ecm.1554 Armstrong AO, Armstrong AJ, Bennett MB, Richardson AJ, Townsend KA, Everett JD, Hays GC, Pederson H, Dudgeon CL (2021a) Mutualism promotes site selection in a large marine planktivore. Ecol Evol 11(10):5606–5623. https://doi.org/10.1002/ece3.7464 Armstrong AO, Armstrong AJ, Jaine FRA, Couturier LIE, Fiora K, Uribe-Palomino J, Weeks SJ, Townsend KA, Bennett MB, Richardson AJ (2016) Prey density threshold and tidal influence on reef manta ray foraging at an aggregation site on the Great Barrier Reef. PLoS ONE 11(5):e0153393. https://doi.org/10.1371/journal.pone.0153393 Armstrong AO, Stevens GMW, Townsend KA, Murray A, Bennett MB, Armstrong AJ, Uribe-Palomino J, Hosegood P, Dudgeon CL, Richardson AJ (2021b) Reef manta rays forage on tidally driven, high density zooplankton patches in Hanifaru Bay. Maldives PeerJ 9:e11992. https://doi.org/10.7717/peerj.11992 Arostegui MC (2025) Cranial endothermy in mobulid rays: Evolutionary and ecological implications of a thermogenic brain. J Anim Ecol 94(1):11–19. https://doi.org/10.1111/1365-2656.14200 Baa Atoll Biosphere Reserve (2021) UNESCO - Man and the Biosphere (MAB) Programme - Biosphere reserve periodic review. Ministry of Environment Climate Change and Technology, UNESCO, Division of Ecological and Earth Sciences Beale CS, Runtuboy F, Sianipar AB, Beer AJE, Kadarusman, Erdmann MV, Setyawan E, Green L, Duffy CAJ, Andrzejaczek S, Block BA, Forsberg K, Meekan M, Gleiss AC (2025) Deep diving behaviour in oceanic manta rays and its potential function. Front Mar Sci 12:1630451. https://doi.org/10.3389/fmars.2025.1630451 Beale CS, Stewart JD, Setyawan E, Sianipar AB, Erdmann MV (2019) Population dynamics of oceanic manta rays ( Mobula birostris ) in the Raja Ampat Archipelago, West Papua, Indonesia, and the impacts of the El Niño–Southern Oscillation on their movement ecology. Divers Distrib 25(9):1472–1487. https://doi.org/10.1111/ddi.12962 Cairns H, Conley D, Stokes C, Scott T, Hosegood P (in press) Tidally Driven Topographic Eddies Exploited by Manta Feeding Aggregations Carruthers L, Ersek V, Maher D, Sanders C, Tait D, Soares J, Floyd M, Hashim AS, Helber S, Garnett M, East H, Johnson JA, Ponta G, Sippo JZ (2024) Sea-level rise and extreme Indian Ocean Dipole explain mangrove dieback in the Maldives. Sci Rep 14(1):27012. https://doi.org/10.1038/s41598-024-73776-z Chenoweth E, Gabriele C, Hill D (2011) Tidal influences on humpback whale habitat selection near headlands. Mar Ecol Prog Ser 423:279–289. https://doi.org/10.3354/meps08891 Chowdhury MR, Ahmed A, Moosa S, Mohamed S (2025) ENSO and seasonal climate variability in the Maldives: an analysis of early warning opportunities. Theor Appl Climatol 156(11):615. https://doi.org/10.1007/s00704-025-05851-y Copernicus Data Space Ecosystem (2026) Copernicus Sentinel data 2026 Couto A, Queiroz N, Relvas P, Baptista M, Furtado M, Castro J, Nunes M, Morikawa H, Rosa R (2017) Occurrence of basking shark Cetorhinus maximus in southern Portuguese waters: a two-decade survey. Mar Ecol Prog Ser 564:77–86. https://doi.org/10.3354/meps12007 Couturier L, Newman P, Jaine F, Bennett M, Venables W, Cagua E, Townsend K, Weeks S, Richardson A (2018) Variation in occupancy and habitat use of Mobula alfredi at a major aggregation site. Mar Ecol Prog Ser 599:125–145. https://doi.org/10.3354/meps12610 Couturier LIE, Dudgeon CL, Pollock KH, Jaine FRA, Bennett MB, Townsend KA, Weeks SJ, Richardson AJ (2014) Population dynamics of the reef manta ray Manta alfredi in eastern Australia. Coral Reefs 33(2):329–342. https://doi.org/10.1007/s00338-014-1126-5 Cowburn B, Moritz C, Grimsditch G, Solandt J (2019) Evidence of coral bleaching avoidance, resistance and recovery in the Maldives during the 2016 mass-bleaching event. Mar Ecol Prog Ser 626:53–67. https://doi.org/10.3354/meps13044 Croll DA, Dewar H, Dulvy NK, Fernando D, Francis MP, Galván-Magaña F, Hall M, Heinrichs S, Marshall A, Mccauley D, Newton KM, Notarbartolo‐Di‐Sciara G, O’Malley M, O’Sullivan J, Poortvliet M, Roman M, Stevens G, Tershy BR, White WT (2016) Vulnerabilities and fisheries impacts: the uncertain future of manta and devil rays. Aquat Conserv Mar Freshw Ecosyst 26(3):562–575. https://doi.org/10.1002/aqc.2591 Cullain N, Tibiriçá Y, Venables SK, Rohner CA, Tittensor DP, Lotze HK (2025) Declines in sightings and changing visitation patterns of reef manta rays at an important aggregation site in Mozambique. Environ Biol Fishes 108(9):1361–1377. https://doi.org/10.1007/s10641-025-01729-0 Dalpadado P, Roxy MK, Arrigo KR, Van Dijken GL, Chierici M, Ostrowski M, Skern-Mauritzen R, Bakke G, Richardson AJ, Sperfeld E (2024) Rapid climate change alters the environment and biological production of the Indian Ocean. Sci Total Environ 906:167342. https://doi.org/10.1016/j.scitotenv.2023.167342 Dewar H, Mous P, Domeier M, Muljadi A, Pet J, Whitty J (2008) Movements and site fidelity of the giant manta ray, Manta birostris , in the Komodo Marine Park, Indonesia. Mar Biol 155(2):121–133. https://doi.org/10.1007/s00227-008-0988-x Dulvy NK, Fowler SL, Musick JA, Cavanagh RD, Kyne PM, Harrison LR, Carlson JK, Davidson LN, Fordham SV, Francis MP, Pollock CM, Simpfendorfer CA, Burgess GH, Carpenter KE, Compagno LJ, Ebert DA, Gibson C, Heupel MR, Livingstone SR, Sanciangco JC, Stevens JD, Valenti S, White WT (2014) Extinction risk and conservation of the world’s sharks and rays. eLife 3:e00590. https://doi.org/10.7554/eLife.00590 Fonseca-Ponce I, Zavala-Jiménez A, Aburto-Oropeza O, Maldonado-Gasca A, Galván-Magaña F, González-Armas R, Stewart J (2022) Physical and environmental drivers of oceanic manta ray Mobula birostris sightings at an aggregation site in Bahía de Banderas, Mexico. Mar Ecol Prog Ser 694:133–148. https://doi.org/10.3354/meps14106 Fortin D, Brooke CF, Lamirande P, Fritz H, McLoughlin PD, Pays O (2020) Quantitative spatial ecology to promote human-wildlife coexistence: A tool for integrated landscape management. Front Sustain Food Syst 4:600363. https://doi.org/10.3389/fsufs.2020.600363 Froman N, Genain M, Stevens GMW, Pearce GP (2023) Use of underwater contactless ultrasonography to elucidate the internal anatomy and reproductive activity of manta and devil rays (family: Mobulidae). J Fish Biol 103(2):305–323. https://doi.org/10.1111/jfb.15423 Graham RT, Witt MJ, Castellanos DW, Remolina F, Maxwell S, Godley BJ, Hawkes LA (2012) Satellite tracking of manta rays highlights challenges to their conservation. PLoS ONE 7(5):e36834. https://doi.org/10.1371/journal.pone.0036834 Guzman HM, Collatos CM, Gomez CG (2022) Movement, behavior, and habitat use of whale sharks ( Rhincodon typus ) in the tropical eastern Pacific Ocean. Front Mar Sci 9:793248. https://doi.org/10.3389/fmars.2022.793248 Harris JL, McGregor PK, Oates Y, Stevens GMW (2020) Gone with the wind: seasonal distribution and habitat use by the reef manta ray ( Mobula alfredi ) in the Maldives, implications for conservation. Aquat Conserv Mar Freshw Ecosyst 30(8):1649–1664. https://doi.org/10.1002/aqc.3350 Harris JL, Stevens GMW (2021) Environmental drivers of reef manta ray ( Mobula alfredi ) visitation patterns to key aggregation habitats in the Maldives. PLoS ONE 16(6):e0252470. https://doi.org/10.1371/journal.pone.0252470 Harvey-Carroll J, Stewart JD, Carroll D, Mohamed B, Shameel I, Zareer IH, Araujo G, Rees R (2021) The impact of injury on apparent survival of whale sharks ( Rhincodon typus ) in South Ari Atoll Marine Protected Area, Maldives. Sci Rep 11(1):937. https://doi.org/10.1038/s41598-020-79101-8 Hays GC, Hobson VJ, Metcalfe JD, Righton D, Sims DW (2006) Flexible foraging movements of leatherback turtles across the North Atlantic Ocean. Ecology 87(10):2647–2656. https://doi.org/10.1890/0012-9658 87%255B2647:FFMOLT%255D2.0.CO;2 Hays GC, Richardson A, Robinson C (2005) Climate change and marine plankton. Trends Ecol Evol 20(6):337–344. https://doi.org/10.1016/j.tree.2005.03.004 Heneghan RF, Everett JD, Blanchard JL, Sykes P, Richardson AJ (2023) Climate-driven zooplankton shifts cause large-scale declines in food quality for fish. Nat Clim Change 13(5):470–477. https://doi.org/10.1038/s41558-023-01630-7 Hildebrand L, Sullivan F, Orben R, Derville S, Torres L (2022) Trade-offs in prey quantity and quality in gray whale foraging. Mar Ecol Prog Ser 695:189–201. https://doi.org/10.3354/meps14115 Jabado RW, Morata AZA, Bennett RH, Finucci B, Ellis JR, Fowler S, Grant MI, Barbosa MAP, Sinclair SL (2024) The global status of sharks, rays, and chimaeras. IUCN, Gland, Switzerland: IUCN Jaine FRA, Couturier LIE, Weeks SJ, Townsend KA, Bennett MB, Fiora K, Richardson AJ (2012) When giants turn up: sighting trends, environmental influences and habitat use of the manta ray ( Mobula alfredi ) at a coral reef. PLoS ONE 7(10):e46170. https://doi.org/10.1371/journal.pone.0046170 Joseph PV, Simon A (2005) Weakening trend of the Southwest Monsoon current through peninsular India from 1950 to the present. JSTOR 89(4):687–694 Kelley D, Richards C (2024) oce: Analysis of Oceanographic Data Klotz L, Fernández R, Rasmussen MH (2017) Annual and monthly fluctuations in humpback whale ( Megaptera novaeangliae ) presence in Skjálfandi Bay, Iceland, during the feeding season (April–October). J Cetacean Res Manage 16:9–16. https://doi.org/10.47536/jcrm.v16i%60.433 Knochel AM, Hussey NE, Kessel ST, Braun CD, Cochran JEM, Hill G, Klaus R, Checkchak T, Elamin El Hassen NM, Younnis M, Berumen ML (2022) Home sweet home: spatiotemporal distribution and site fidelity of the reef manta ray ( Mobula alfredi ) in Dungonab Bay, Sudan. Mov Ecol 10(1):22. https://doi.org/10.1186/s40462-022-00314-9 Kwiatkowski L, Aumont O, Bopp L (2019) Consistent trophic amplification of marine biomass declines under climate change. Glob Change Biol 25(1):218–229. https://doi.org/10.1111/gcb.14468 Laglbauer BJL, D’Costa NG, Stewart JD, Palacios MD, Cronin M, Fernando D, Lezama-Ochoa N, Armstrong AO, Jabado RW, Fowler S, Lawson JM, Koubrak O, Murua J, Ko Gyi T, Karnad D, Chopra M, Notarbartolo-di-Sciara G, Rambahiniarison J, Croll D, Rojas S, Fahmi, Harris JL, Binthe Haque A, Murua H, Pérez-Jiménez JC, Humble E, Barrowclift E, Salim MG, De Bruyne G, Seidu I, Zambrano-Vizquel LA, Davies K, Moazzam Khan M, Bucair N, Johnson J, Labyedh G, Takoukam Kamla A, Fuentes K, Carter R, Barros N, Stevens GMW (2026) Global manta and devil ray population declines: Closing policy and management gaps to reduce fisheries mortality. Biol Conserv 313:111589. https://doi.org/10.1016/j.biocon.2025.111589 Lawson JM, Fordham SV, O’Malley MP, Davidson LNK, Walls RHL, Heupel MR, Stevens G, Fernando D, Budziak A, Simpfendorfer CA, Ender I, Francis MP, Di Notarbartolo G, Dulvy NK (2017) Sympathy for the devil: a conservation strategy for devil and manta rays. PeerJ 5:e3027. https://doi.org/10.7717/peerj.3027 Lazaridis E (2022) lunar: Calculate Lunar Phase & Distance, Seasons and Related Environmental Factors Li M (2000) Estimating horizontal dispersion of floating particles in wind-driven upper ocean. Spill Sci Technol Bull 6(3–4):255–261. https://doi.org/10.1016/S1353-2561(01)00044-5 Lyard FH, Allain DJ, Cancet M, Carrere L, Picot N (2021) FES2014 global ocean tide atlas. design and performance Marshall AD, Barreto R, Carlson J, Fernando D, Fordham S, Francis MP, Herman K, Jabado RW, Liu KM, Pacoureau N, Rigby CL, Romanov E, Sherley RB (2022) Mobula alfredi (amended version of 2019 assessment). The IUCN Red List of Threatened Species 2022 Marshall AD, Dudgeon CL, Bennett MB (2011) Size and structure of a photographically identified population of manta rays Manta alfredi in southern Mozambique. Mar Biol 158(5):1111–1124. https://doi.org/10.1007/s00227-011-1634-6 McInturf AG, Muhling B, Bizzarro JJ, Fangue NA, Ebert DA, Caillaud D, Dewar H (2022) Spatial distribution, temporal changes, and knowledge gaps in basking shark ( Cetorhinus maximus ) sightings in the California current ecosystem. Front Mar Sci 9:818670. https://doi.org/10.3389/fmars.2022.818670 Meteoblue (2025) Historical weather data for Hanifarurah (10 m wind speed and direction) Middleton AD, Sawyer H, Merkle JA, Kauffman MJ, Cole EK, Dewey SR, Gude JA, Gustine DD, McWhirter DE, Proffitt KM, White P (2020) Conserving transboundary wildlife migrations: recent insights from the Greater Yellowstone Ecosystem. Front Ecol Environ 18(2):83–91. https://doi.org/10.1002/fee.2145 Min HS, Noh Y (2004) Influence of the surface heating on Langmuir Circulation. J Phys Oceanogr 34:2630–2641 Moloney HM, Armstrong AO, Stevens GMW, Dudgeon CL, Harris JL, Townsend KA, Richardson AJ (in press) A new visual index for assessing zooplankton biomass and its utility in assessing prey availability for megaplanktivores Moloney HM, Garcia Rojas MI, Rothe N, Armstrong AO, Ballard K, Barraud F, Hamdan F, Richardson AJ, Ryad EM, Sawers T, Townsend KA, Stevens GMW (in press) Valuing conservation and natural wealth: The blue economy of manta ray watching in the Maldives Murray A, Garrud E, Ender I, Lee-Brooks K, Atkins R, Lynam R, Arnold K, Roberts C, Hawkins J, Stevens G (2020) Protecting the million-dollar mantas; creating an evidence-based code of conduct for manta ray tourism interactions. J Ecotourism 19(2):132–147. https://doi.org/10.1080/14724049.2019.1659802 Nagelkerke NJD (1991) A More General Definition of the Coefficient of Determination. Biometrika 78(3). https://doi.org/10.1093/biomet/78.3.691 Neves L (2009) Investigating anthropogenic impacts on the manta rays and whale sharks of Hanifaru, Maldives. MSc Thesis, University of York NOAA Physical Sciences Laboratory (2024) Monthly Indian Ocean Dipole Index (DMI) based on HadISST1 Paig-Tran EWM, Kleinteich T, Summers AP (2013) The filter pads and filtration mechanisms of the devil rays: Variation at macro and microscopic scales. J Morphol 274(9):1026–1043. https://doi.org/10.1002/jmor.20160 Palacios MD, Stewart JD, Croll DA, Cronin MR, Trejo-Ramírez A, Stevens GMW, Lezama-Ochoa N, Zilliacus KM, González–Armas R, Di Notarbartolo G, Galván–Magaña F (2023) Manta and devil ray aggregations: conservation challenges and developments in the field. Front Mar Sci 10:1148234. https://doi.org/10.3389/fmars.2023.1148234 Pistevos JCA, Nagelkerken I, Rossi T, Olmos M, Connell SD (2015) Ocean acidification and global warming impair shark hunting behaviour and growth. Sci Rep 5(1):16293. https://doi.org/10.1038/srep16293 R Core Team (2024) R: a language and environment for statistical computing Richardson AJ (2008) In hot water: zooplankton and climate change. ICES J Mar Sci 65(3):279–295. https://doi.org/10.1093/icesjms/fsn028 Rohner CA, Pierce S, Marshall A, Weeks S, Bennett M, Richardson A (2013) Trends in sightings and environmental influences on a coastal aggregation of manta rays and whale sharks. Mar Ecol Prog Ser 482:153–168. https://doi.org/10.3354/meps10290 Rohner CA, Prebble CEM (2021) Whale shark foraging, feeding, and diet. Whale Sharks, 1st edn. CRC, Boca Raton, pp 153–180 Rohner CA, Venables SK, Knochel AM, Rambahiniarison JM, Marillac V, Cardon C, Scholten N, Pierce SJ, Kiszka JJ (2025) Movements and habitat use of reef manta rays around the Mozambique Channel Island of Mayotte, Southwestern Indian Ocean. Environ Biol Fishes 108(6):937–955. https://doi.org/10.1007/s10641-025-01695-7 Roxy MK, Modi A, Murtugudde R, Valsala V, Panickal S, Prasanna Kumar S, Ravichandran M, Vichi M, Lévy M (2016) A reduction in marine primary productivity driven by rapid warming over the tropical Indian Ocean. Geophys Res Lett 43(2):826–833. https://doi.org/10.1002/2015GL066979 Roxy MK, Ritika K, Terray P, Murtugudde R, Ashok K, Goswami BN (2015) Drying of Indian subcontinent by rapid Indian Ocean warming and a weakening land-sea thermal gradient. Nat Commun 6(1):7423. https://doi.org/10.1038/ncomms8423 Rubin RD, Kumli KR, Klimley AP, Stewart JD, Ketchum JT, Hoyos-Padilla EM, Galván-Magaña F, Zavala-Jiménez AA, Fonseca-Ponce IA, Saunders M, Dominguez-Sanchez PS, Ahuja P, Nevels CR, González PAP, Corgos A, Diemer SJ (2025) Insular and mainland interconnectivity in the movements of oceanic manta rays ( Mobula birostris ) off Mexico in the Eastern Tropical Pacific. Environ Biol Fishes 108(4):555–568. https://doi.org/10.1007/s10641-024-01622-2 Saji NH, Goswami BN, Vinayachandran PN, Yamagata T (1999) A dipole mode in the tropical Indian Ocean. Nature 401(6751):360–363. https://doi.org/10.1038/43854 Sasamal SK (2006) Island mass effect around the Maldives during the winter months of 2003 and 2004. Int J Remote Sens 27(22):5087–5093. https://doi.org/10.1080/01431160500177562 Setyawan E (2016) Correlating environmental parameters with the movement patterns and site fidelity of reef manta ray ( Manta alfredi , Krefft 1868) using acoustic telemetry and remote sensing in Raja Ampat, Indonesia. Thesis, University of Tasmania Sims DW, Southall E, Richardson A, Reid P, Metcalfe J (2003) Seasonal movements and behaviour of basking sharks from archival tagging: no evidence of winter hibernation. Mar Ecol Prog Ser 248:187–196. https://doi.org/10.3354/meps248187 Sims DW, Southall EJ, Tarling GA, Metcalfe JD (2005) Habitat-specific normal and reverse diel vertical migration in the plankton‐feeding basking shark. J Anim Ecol 74(4):755–761. https://doi.org/10.1111/j.1365-2656.2005.00971.x Sims DW, Witt MJ, Richardson AJ, Southall EJ, Julian D, Metcalfe (2006) Encounter success of free-ranging marine predator movements across a dynamic prey landscape. Proc R Soc B. https://doi.org/10.1098/rspb.2005.3444 Singh AD (2018) Rapid switch in monsoon-wind induced surface hydrographic conditions of the eastern Arabian Sea during the last deglaciation. Quat Int 479:3–11. https://doi.org/10.1016/j.quaint.2018.03.027 Skomal GB, Zeeman SI, Chisholm JH, Summers EL, Walsh HJ, McMahon KW, Thorrold SR (2009) Transequatorial migrations by basking sharks in the Western Atlantic Ocean. Curr Biol 19(12):1019–1022. https://doi.org/10.1016/j.cub.2009.04.019 Sleeman JC, Meekan MG, Fitzpatrick BJ, Steinberg CR, Ancel R, Bradshaw CJA (2010) Oceanographic and atmospheric phenomena influence the abundance of whale sharks at Ningaloo Reef, Western Australia. J Exp Mar Biol Ecol 382(2):77–81. https://doi.org/10.1016/j.jembe.2009.10.015 Smith JA (2001) Observations and theories of Langmuir Circulation: a story of mixing. In: Lumley JL (ed) Fluid Mechanics and the Environment: Dynamical Approaches. Springer Berlin Heidelberg, Berlin, Heidelberg, pp 295–314 Soulsbury CD, White PCL (2016) Human–wildlife interactions in urban areas: a review of conflicts, benefits and opportunities. Wildl Res 42(7):541–553. https://doi.org/10.1071/WR14229 Srivastava A, Martin GM, Pradhan M, Rao SA, Ineson S (2026) The multi-year negative Indian Ocean Dipole of 2021–2022. Weather Clim Dyn 7(1):1–15. https://doi.org/10.5194/wcd-7-1-2026 Stephens DW, Krebs JR (1986) Foraging theory. Princeton University Press Stevens GMW (2016) Conservation and population ecology of manta rays in the Maldives. PhD Thesis, University of York Stevens GMW, Froman N (2018) The Maldives Archipelago. World Seas: an environmental evaluation. Elsevier, pp 211–236 Stewart JD, Jaine FRA, Armstrong AJ, Armstrong AO, Bennett MB, Burgess KB, Couturier LIE, Croll DA, Cronin MR, Deakos MH, Dudgeon CL, Fernando D, Froman N, Germanov ES, Hall MA, Hinojosa-Alvarez S, Hosegood JE, Kashiwagi T, Laglbauer BJL, Lezama-Ochoa N, Marshall AD, McGregor F, Di Sciara N, Palacios G, Peel MD, Richardson LR, Rubin AJ, Townsend RD, Venables KA, Stevens SK GMW (2018) Research priorities to support effective manta and devil ray conservation. Front Mar Sci 5:314. https://doi.org/10.3389/fmars.2018.00314 Strike EM, Harris JL, Ballard KL, Hawkins JP, Crockett J, Stevens GMW (2022) Sublethal injuries and physical abnormalities in Maldives manta rays, Mobula alfredi and Mobula birostris . Front Mar Sci 9:773897. https://doi.org/10.3389/fmars.2022.773897 The Manta Trust (2024) Maldives Manta Conservation Programme database Thorpe SA (2004) Langmuir circulation. Annu Rev Fluid Mech 36(1):55–79. https://doi.org/10.1146/annur%20ev.fluid.36.052203.071431 Thushara V, Vinayachandran PN (2020) Unprecedented surface Chlorophyll blooms in the southeastern Arabian Sea during an extreme Negative Indian Ocean Dipole. Geophys Res Lett 47(13). https://doi.org/10.1029/2019GL085026 . e2019GL085026 UNEP-WCMC (2020) Ocean+ Habitats [On-line] Venables SK, Rohner CA, Flam AL, Pierce SJ, Marshall AD (2024) Persistent declines in sightings of manta and devil rays (Mobulidae) at a global hotspot in southern Mozambique. Environ Biol Fishes. https://doi.org/10.1007/s10641-024-01576-5 Venables WN, Ripley BD (2022) Modern applied statistics with S (4th ed.) Watson M, Stamation K, Charlton C, Bannister J (2001) Calving intervals, long range movements and site fidelity of southern right whales ( Eubalaena australis ) in south­eastern Australia. J Cetacean Res Manag 22 Weeks S, Magno-Canto M, Jaine F, Brodie J, Richardson A (2015) Unique sequence of events triggers manta ray feeding frenzy in the southern Great Barrier Reef, Australia. Remote Sens 7(3):3138–3152. https://doi.org/10.3390/rs70303138 Witt M, Hardy T, Johnson L, McClellan C, Pikesley S, Ranger S, Richardson P, Solandt J, Speedie C, Williams R, Godley B (2012) Basking sharks in the northeast Atlantic: spatio-temporal trends from sightings in UK waters. Mar Ecol Prog Ser 459:121–134. https://doi.org/10.3354/meps09737 Wolanski E, Asaeda T, Tanaka A, Deleersnijder E (1996) Three-dimensional island wakes in the field, laboratory experiments and numerical models. Cont Shelf Res 16(11):1437–1452. https://doi.org/10.1016/0278-4343(95)00087-9 Worsfold M, Good S, Atkinson C, Embury O (2024) Presenting a long-term, reprocessed dataset of global sea surface temperature produced using the OSTIA system Zerbini A, Andriolo A, Heide-Jørgensen M, Pizzorno J, Maia Y, VanBlaricom G, DeMaster D, Simões-Lopes P, Moreira S, Bethlem C (2006) Satellite-monitored movements of humpback whales Megaptera novaeangliae in the southwest Atlantic Ocean. Mar Ecol Prog Ser 313:295–304. https://doi.org/10.3354/meps313295 Supplementary Files SupplementaryInformation.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 12 May, 2026 Reviewers invited by journal 06 May, 2026 Editor assigned by journal 19 Apr, 2026 First submitted to journal 17 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9455345","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":635387291,"identity":"d43d1ee6-3b46-4236-8a66-68bb44e26dc5","order_by":0,"name":"Hannah M Moloney","email":"data:image/png;base64,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","orcid":"https://orcid.org/0009-0003-8509-8325","institution":"University of the Sunshine Coast School of Science Technology and Engineering","correspondingAuthor":true,"prefix":"","firstName":"Hannah","middleName":"M","lastName":"Moloney","suffix":""},{"id":635387295,"identity":"d539e233-f9e0-477a-bcc0-57b2def4b4a4","order_by":1,"name":"Guy M.W Stevens","email":"","orcid":"","institution":"The Manta Trust","correspondingAuthor":false,"prefix":"","firstName":"Guy","middleName":"M.W","lastName":"Stevens","suffix":""},{"id":635387296,"identity":"7c48650f-e70c-4ec3-983a-57948bfc440f","order_by":2,"name":"Asia O Armstrong","email":"","orcid":"","institution":"University of the Sunshine Coast School of Science Technology and Engineering","correspondingAuthor":false,"prefix":"","firstName":"Asia","middleName":"O","lastName":"Armstrong","suffix":""},{"id":635387297,"identity":"b3d81588-f5a5-4935-b365-7c4e820b6257","order_by":3,"name":"Alvise Dabala","email":"","orcid":"","institution":"University of Queensland - Centre for Biodiversity and Conservation Science","correspondingAuthor":false,"prefix":"","firstName":"Alvise","middleName":"","lastName":"Dabala","suffix":""},{"id":635387298,"identity":"5a49beab-a07f-43f2-8bd0-04df504d252c","order_by":4,"name":"Christine L Dudgeon","email":"","orcid":"","institution":"University of the Sunshine Coast School of Science Technology and Engineering","correspondingAuthor":false,"prefix":"","firstName":"Christine","middleName":"L","lastName":"Dudgeon","suffix":""},{"id":635387299,"identity":"599f8ed9-7df7-404c-b127-073879daf888","order_by":5,"name":"Elspeth M Strike","email":"","orcid":"","institution":"The Manta Trust","correspondingAuthor":false,"prefix":"","firstName":"Elspeth","middleName":"M","lastName":"Strike","suffix":""},{"id":635387300,"identity":"330f4de3-e2f3-448d-b317-7e5ee98e5053","order_by":6,"name":"Tam J Sawers","email":"","orcid":"","institution":"The Manta Trust","correspondingAuthor":false,"prefix":"","firstName":"Tam","middleName":"J","lastName":"Sawers","suffix":""},{"id":635387301,"identity":"4b3ed066-7fec-4407-b00a-7e1821155bc6","order_by":7,"name":"Kathy A Townsend","email":"","orcid":"","institution":"University of the Sunshine Coast School of Science Technology and Engineering","correspondingAuthor":false,"prefix":"","firstName":"Kathy","middleName":"A","lastName":"Townsend","suffix":""},{"id":635387302,"identity":"c9639e7e-46f1-4b40-94e3-c122bbf3140e","order_by":8,"name":"Anthony J Richardson","email":"","orcid":"","institution":"University of Queensland - School of the Environment","correspondingAuthor":false,"prefix":"","firstName":"Anthony","middleName":"J","lastName":"Richardson","suffix":""}],"badges":[],"createdAt":"2026-04-18 07:56:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9455345/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9455345/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109297582,"identity":"ab5c79d9-7895-4b7f-8d6c-27e18f77e92f","added_by":"auto","created_at":"2026-05-15 09:00:19","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1032316,"visible":true,"origin":"","legend":"\u003cp\u003eThe study site in the Republic of Maldives (shown in the top right inset). Bathymetry and morphology of Hanifaru Bay, within Hanifaru Marine Protected Area in eastern Baa Atoll (red box), the focal site for investigating the fine-scale oceanographic drivers, and temporal sightings, of \u003cem\u003eMobula alfredi\u003c/em\u003eabundance. This map was created in R Studio using Copernicus Sentinel data (Copernicus Data Space Ecosystem 2026)\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9455345/v1/254dca77bcf845e80afc0c05.png"},{"id":109297584,"identity":"37c0949c-55ef-4893-b787-a1ecf3b31f0a","added_by":"auto","created_at":"2026-05-15 09:00:20","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":524983,"visible":true,"origin":"","legend":"\u003cp\u003eThe 21 sites identified as key \u003cem\u003eMobula alfredi\u003c/em\u003e habitats (≥50 sightings) during the Southwest Monsoon (May–November) between 2007–2024 in eastern Baa Atoll, Maldives. The (a) total number of \u003cem\u003eM. alfredi\u003c/em\u003e sightings per site shown on the map and (b) in the table. Latitude and longitude are given in decimal degrees (N, E). This map was created in R Studio using reef features from Millennium Coral Reef Mapping Project (MCRMP; \u003ca href=\"https://habitats.oceanplus.org/\"\u003ehttps://habitats.oceanplus.org\u003c/a\u003e); MCRMP validated maps provided by the Institute for Marine Remote Sensing, University of South Florida (IMaRS/USF) and Institut de Recherche pour le D\u003cem\u003eé\u003c/em\u003eveloppement (IRD, Centre de Noum\u003cem\u003eéa)\u003c/em\u003e, with support from NASA. IRD does not endorse these products (UNEP-WCMC 2020)\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9455345/v1/a37a214c6608e5a8675218d4.png"},{"id":109297610,"identity":"a88f89cf-3c0d-4b1b-a677-c21e4fefc283","added_by":"auto","created_at":"2026-05-15 09:00:33","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":235993,"visible":true,"origin":"","legend":"\u003cp\u003eMean\u003cem\u003e Mobula alfredi\u003c/em\u003e annual counts (± standard error) during the Southwest Monsoon (May–November) between 2007–2024 at Hanifaru Bay, Baa Atoll, Maldives\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-9455345/v1/3b31daad4fbe8485bbf90c02.png"},{"id":109297622,"identity":"b9e17d46-442f-4ca1-bd72-5a45985e6633","added_by":"auto","created_at":"2026-05-15 09:00:50","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":453751,"visible":true,"origin":"","legend":"\u003cp\u003ePredicted trends models during the months of the Southwest Monsoon (May–November) at Hanifaru Bay. The “\u003cem\u003eManta trends\u003c/em\u003e” models with the (a) annual abundance of \u003cem\u003eMobula alfredi\u003c/em\u003e(counts are from 2007–2024). The “\u003cem\u003eZooplankton trends\u003c/em\u003e” models with the (b) annual high zooplankton density (zooplankton density is from 2010–2024). The 95% confidence interval is shown\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-9455345/v1/bd3eec13e491acb311a38335.png"},{"id":109297608,"identity":"952d7d1c-b4da-43db-96ee-e8939d2da8c6","added_by":"auto","created_at":"2026-05-15 09:00:32","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":310655,"visible":true,"origin":"","legend":"\u003cp\u003e“\u003cem\u003eManta drivers”\u003c/em\u003e model outputs showing the relationship between \u003cem\u003eMobula alfredi\u003c/em\u003e abundance and all significant predictors during the Southwest Monsoon at Hanifaru Bay (2007–2024). The 95% confidence interval is shown\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-9455345/v1/11f3231b62a7f04e13a12770.png"},{"id":109297588,"identity":"4f6b9bb3-c435-4e6a-ab7a-270ad6be4f2a","added_by":"auto","created_at":"2026-05-15 09:00:25","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":95807,"visible":true,"origin":"","legend":"\u003cp\u003e“\u003cem\u003eZooplankton drivers”\u003c/em\u003e model outputs showing the relationship between high density of zooplankton and all significant predictors during the Southwest Monsoon at Hanifaru Bay (2010–2024). The 95% confidence interval is shown\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-9455345/v1/f15cf2a7dd4fb53633b3f3c2.png"},{"id":109298593,"identity":"d54f0357-f7a6-4b91-a4c5-645b35e5a6be","added_by":"auto","created_at":"2026-05-15 09:15:04","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3325544,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9455345/v1/248ade41-c6c6-48d0-a80d-c1b33c0320a5.pdf"},{"id":109297569,"identity":"6b7b8b9a-a44a-46df-870e-86c3b7a950fd","added_by":"auto","created_at":"2026-05-15 08:59:55","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":77115,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-9455345/v1/d4f14b44c76dfbd463a964a7.docx"}],"financialInterests":"","formattedTitle":"Prey availability and fine-scale oceanography drive reef manta ray aggregations at the world’s largest hotspot","fulltext":[{"header":"Introduction","content":"\u003cp\u003eUnderstanding drivers of spatial distribution and habitat selection of large, highly mobile species is essential for effective management (Graham et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). In dynamic marine environments, megaplanktivores \u0026ndash; large marine animals that feed on plankton \u0026ndash; aggregate at specific sites to exploit prey patches that exceed energetic thresholds required to sustain their large body size (Stephens and Krebs \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e1986\u003c/span\u003e). Foraging for zooplankton presents unique energy challenges, including high search costs and the requirement for a minimum prey density to offset the metabolic demands of bulk-filter feeding (Hildebrand et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, locating high concentrations of zooplankton is difficult, as zooplankton distributions are often characterised by spatiotemporal unpredictability across multiple scales (Hays et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Consequently, the movement and habitat use of these species are highly coupled to oceanographic variables, such as currents, tides, winds, and upwelling, that dictate prey availability. Disentangling these drivers remains complex, as the relative influence of foraging ecology, species-specific physiologies, and reproductive requirements vary across taxa and geographic regions (Sims et al. \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Zerbini et al. \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Skomal et al. \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Anderson et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Harris and Stevens \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Cullain et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). With increasing anthropogenic pressures and climatic shifts predicted to alter reef-wide ocean dynamics and productivity (Hays et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Kwiatkowski et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Heneghan et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), identifying the specific environmental triggers of megaplanktivore aggregations is critical to evaluate the adequacy of current and future management.\u003c/p\u003e \u003cp\u003eThe coupling between oceanography and biology is often mediated by environmental drivers operating across multiple spatio-temporal scales (Sims et al. \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). At the regional scale, broad climatic indices such as the El Ni\u0026ntilde;o-Southern Oscillation (ENSO) and the Indian Ocean Dipole (IOD) regulate primary productivity by modulating sea surface temperatures (SSTs) and thermocline depth (Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Dalpadado et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Abeywickrama et al. under review). At the mesoscale, features such as fronts and eddies act as productive hotspots by physically concentrating biomass into predictable foraging patches (Sims et al. \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Finally, at the local scale, interactions between wind-driven currents, tidal flux, and complex bathymetry create high-density retention zones that facilitate the aggregation of both prey and predators (Wolanski et al. \u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e1996\u003c/span\u003e; Chenoweth et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). The influence of these scales is evident in the movement and abundance of various megaplanktivores. For example, in Australia, regional-scale Southern Oscillation Index influences the abundance of whale sharks (\u003cem\u003eRhincodon typus\u003c/em\u003e) on the Ningaloo Reef by altering the strength of the Leeuwin Current (Sleeman et al. \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2010\u003c/span\u003e); while on the local scale, abundance of reef manta rays (\u003cem\u003eMobula alfredi\u003c/em\u003e) at Lady Elliot Island is modulated by the tidal cycle, lower wind speeds, and the moon cycle (Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Despite the broad-scale predictability of these aggregations for megaplanktivores, the specific fine-scale mechanisms that drive the formation, maintenance, and predictability of these high-quality dense zooplankton patches remain poorly resolved. Being obligate ram-feeders, \u003cem\u003eM. alfredi\u003c/em\u003e are sensitive to these dynamics as their foraging efficiency is dependent on prey density (Paig-Tran et al. \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Stevens \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Unlike other elasmobranchs capable of stationary feeding (e.g., vertical/or suction feeding; (Rohner and Prebble \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), \u003cem\u003eM. alfredi\u003c/em\u003e must maintain a minimum swimming speed to move water over their branchial filters (Paig‐Tran et al. 2013). Consequently, they are energetically tethered to specific oceanographic retention zones that can concentrate prey above a critical threshold (Armstrong et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Armstrong et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAs large, highly mobile planktivores residing in oligotrophic waters, \u003cem\u003eM. alfredi\u003c/em\u003e rely on ephemeral productivity peaks to locate dense zooplankton patches (Armstrong et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Armstrong et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e). Their conservative life history is characterised by slow growth, late maturity, and low fecundity (Dulvy et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Stevens \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Lawson et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Marshall et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Froman et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), making \u003cem\u003eM. alfredi\u003c/em\u003e highly vulnerable to anthropogenic pressures, including climate-induced prey shifts, targeted fisheries, bycatch, and unsustainable tourism (Richardson \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Stewart et al. \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Murray et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Jabado et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Venables et al. \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). While population declines have been reported globally (for review, see Laglbauer et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2026\u003c/span\u003e), long-term monitoring of aggregation dynamics remains scarce. Critically, few studies have described the underlying zooplankton dynamics alongside direct observations on the manta rays themselves, leaving a gap in our understanding of the environmental thresholds sustaining these threatened populations.\u003c/p\u003e \u003cp\u003eThe Republic of Maldives supports the world\u0026rsquo;s largest known population of \u003cem\u003eM. alfredi\u003c/em\u003e (Kitchen-Wheeler et al., 2012; Stevens, \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), with Hanifaru Bay, located within eastern Baa Atoll, hosting the largest feeding aggregation. Indeed, aggregations regularly exceed 50 individuals and occasionally exceed 200 individuals concurrently (Harris et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Beyond its ecological importance (Harris et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Harris and Stevens, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), the site is a major economic asset for the Maldives\u0026rsquo; tourism sector (Murray et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Moloney et al. in press). However, the highly predictable nature of these aggregations increases their susceptibility to localised threats, including habitat degradation (Stevens and Froman \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) and tourism pressure (Murray et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Further, climate projections suggest warming and weakening monsoonal winds may shift the Indian Ocean toward an increasingly oligotrophic state (Roxy et al. \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Dalpadado et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), potentially leading to a decline in prey biomass in the Maldives (Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe climate of the Maldives archipelago is regulated by the South Asian Monsoon system, which strongly influences the seasonal distribution of the resident \u003cem\u003eM. alfredi\u003c/em\u003e population (Anderson et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). During the Southwest Monsoon (SW; \u003cem\u003eHulhangu\u003c/em\u003e, May\u0026ndash;November), strong winds induce localised deep-water upwelling on the leeward side of the archipelago, enhancing chlorophyll-\u003cem\u003ea\u003c/em\u003e concentrations and productivity (Sasamal \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). On a finer scale, lunar-driven spring tides transport nutrient- and plankton-rich oceanic waters onto the atolls and into shallow lagoons and embayments, facilitating the retention of dense zooplankton concentrations (Stevens \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Although these macro- and meso-scale processes are recognised, the oceanographic and environmental conditions that concentrate the zooplankton required to trigger these mass-foraging events remain poorly understood.\u003c/p\u003e \u003cp\u003eHere, we use an 18-year dataset to investigate the fine-scale environmental and oceanographic drivers of \u003cem\u003eM. alfredi\u003c/em\u003e habitat use and zooplankton prey dynamics at Hanifaru Bay. Long-term datasets have proven valuable for investigating the spatial ecology of species with high site fidelity, such as manta rays (Rohner et al. \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Using this dataset, we quantify the influence of multi-scale physical variables on prey availability and predator recruitment using Generalised Linear Models (GLMs) at Hanifaru Bay. Specifically, we ask: (1) What are the \u003cem\u003eM. alfredi\u003c/em\u003e sighting trends over the 18-year period, and do they correspond with temporal changes in zooplankton biomass? (2) Which specific environmental and oceanographic predictors (including lunar phase, tidal state, and wind conditions) drive mass aggregation events of \u003cem\u003eM. alfredi\u003c/em\u003e? (3) To what extent do these physical conditions also drive the localised concentration of zooplankton prey? By resolving these questions, we aim to improve the understanding of the ecological triggers behind these globally significant foraging events, ultimately informing targeted conservation strategies for critical habitats facing increasing anthropogenic pressure and climatic instability.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003eStudy location\u003c/p\u003e \u003cp\u003eThe Republic of Maldives is an archipelago of 26 geographical atolls situated south of India in the Indian Ocean, extending 870 km latitudinally to just below the equator (Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e; inset). This study was conducted in Baa Atoll, a UNESCO Biosphere Reserve located in the central double-chain atoll zone of the archipelago (Murray et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Baa Atoll Biosphere Reserve \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Between 2007 and 2024, across both the SW and Northwest (NW) Monsoons, a total of 13,109 surveys were conducted across 59 sites within Baa Atoll. The data set was analysed to identify key sites, defined as having recorded\u0026thinsp;\u0026ge;\u0026thinsp;50 \u003cem\u003eM. alfredi\u003c/em\u003e sightings during the study. These sites were distributed along a 44 km stretch of coastal waters.\u003c/p\u003e \u003cp\u003eThe primary study site was Hanifaru Bay (5.1733\u0026deg;N, 73.145\u0026deg;E), located within the Hanifaru Marine Protected Area (MPA), a small cul-de-sac-shaped reef system that ephemerally concentrates zooplankton (Armstrong et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Hanifaru Bay was selected for intensive monitoring as it supports the highest recorded density of \u003cem\u003eM. alfredi\u003c/em\u003e sightings within the Maldives (Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). During the SW Monsoon (May\u0026ndash;November), the site hosts large aggregations of foraging megaplanktivores, including \u003cem\u003eM. alfredi\u003c/em\u003e (Stevens \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Armstrong et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e; The Manta Trust \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and \u003cem\u003eR. typus\u003c/em\u003e (Neves \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). This study used sighting and prey data collected at this site to investigate drivers of \u003cem\u003eM. alfredi\u003c/em\u003e habitat use and behaviour at this location.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eStudy design\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eManta ray sightings and behaviour\u003c/h2\u003e \u003cp\u003eTo assess long-term trends in \u003cem\u003eM. alfredi\u003c/em\u003e abundance, we used an 18-year logbook dataset curated by the \u003cem\u003eMaldives Manta Conservation Programme\u003c/em\u003e and \u003cem\u003eThe Manta Trust\u003c/em\u003e (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.mantatrust.org/maldives\u003c/span\u003e\u003cspan address=\"https://www.mantatrust.org/maldives\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e; The Manta Trust \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Long-term logbook data (hereafter referred to as logbook data) offer a cost-effective approach to amassing extensive long-term datasets. They are a common method used to investigate trends in sightings, movement, and behaviour of megaplanktivores (see Watson et al. \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Armstrong et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Venables et al. \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). When integrated with \u003cem\u003ein situ\u003c/em\u003e or remotely sourced environmental data, these datasets can provide insights into decadal-scale patterns of occurrence, habitat use, and species behaviour, which are critical for effective, long-term conservation (Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThis dataset includes both \u003cem\u003eM. alfredi\u003c/em\u003e presence and absence data. While surveys were conducted year-round, effort was intensified during the SW Monsoon to coincide with seasonal peaks in \u003cem\u003eM. alfredi\u003c/em\u003e occurrence (Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Trained field researchers conducted 13,109 surveys to collect this information over 18 years in eastern Baa Atoll using three primary methods: snorkelling, scuba diving, and boat-based spotter surveys. During surveys, searches over the reef consistently followed the same track, with boat-based surveys kept to a speed of 6\u0026ndash;9 knots and observers monitoring the surrounding water. A cumulative 8,995 survey hours were expended, representing 24,209 total observer hours. Effort was standardised in the GLMs by including the duration of each survey in Baa Atoll. The mean survey duration varied across different survey types: snorkelling surveys\u0026thinsp;=\u0026thinsp;128.0\u0026thinsp;\u0026plusmn;\u0026thinsp;2.02 minutes, \u003cem\u003en\u003c/em\u003e = (\u0026plusmn;\u0026thinsp;standard error); scuba diving surveys\u0026thinsp;=\u0026thinsp;83.2\u0026thinsp;\u0026plusmn;\u0026thinsp;13.1 minutes, \u003cem\u003en =\u003c/em\u003e ; and boat-based surveys\u0026thinsp;=\u0026thinsp;45.7\u0026thinsp;\u0026plusmn;\u0026thinsp;7.44 minutes, \u003cem\u003en\u003c/em\u003e =. Note that all surveys were capped at a maximum of 300 minutes and the number of observers (median\u0026thinsp;=\u0026thinsp;3 people) was capped at a maximum of 5 people. There was interannual and monthly variability in survey effort, with a mean of 17.8\u0026thinsp;\u0026plusmn;\u0026thinsp;0.8 surveys per month (ranging between 1\u0026ndash;37 surveys) and 115.8\u0026thinsp;\u0026plusmn;\u0026thinsp;10.5 surveys per year (ranging between 29\u0026ndash;185 surveys). At Hanifaru Bay, the median number of observers was three (ranging from 1\u0026ndash;4) and the mean survey duration was 125.0\u0026thinsp;\u0026plusmn;\u0026thinsp;1.98 minutes (varying between 2\u0026ndash;300 minutes).\u003c/p\u003e \u003cp\u003eTo minimise double counting of individuals at the same event, individual \u003cem\u003eM. alfredi\u003c/em\u003e were identified using ventral photo-identification (photo-ID), a reliable method for distinguishing individuals based on unique pigment patterns that remain largely unchanged throughout their lives (Marshall et al. \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). During each survey, an \u0026ldquo;estimated count of sightings\u0026rdquo; (total individuals observed) and a \u0026ldquo;confirmed count of sightings\u0026rdquo; (unique individuals verified via photo-ID) were recorded. In mass aggregation events where observer limitations, environmental conditions, or animal behaviour might limit photo-ID efficiency (e.g., difficult to ID all manta rays individually), we used the confirmed count of sightings unless it deviated by \u0026ge;\u0026thinsp;50% from the estimated count of sightings, in which the estimated count was adopted to avoid under-representation of abundance when there were many manta rays present and photo-ID of all individuals was difficult (see Online Resource. 1). This is referred to as the abundance of \u003cem\u003eM. alfredi\u003c/em\u003e. Primary behaviour was categorised as (1) feeding, (2) cleaning, (3) cruising, or (4) courtship, according to the dominant behaviour during the encounter recorded.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eEnvironmental data\u003c/h3\u003e\n\u003cp\u003eTo investigate the physical drivers of \u003cem\u003eM. alfredi\u003c/em\u003e habitat use and prey availability (zooplankton), a suite of predictor variables was recorded \u003cem\u003ein situ\u003c/em\u003e concurrently with each survey. These included sea state, weather conditions, current strength, and in-water visibility (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). To ensure data quality and observer safety, surveys were restricted to conditions below a Beaufort scale of 7 (i.e., near gale conditions). These field observations were supplemented with remotely sensed data obtained from online databases (e.g., Copernicus, Meteoblue, NOAA and FES2014) such as wind speed/strength, SST, high tide (time and size), lunar phase and IOD index (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). We selected predictors for the statistical models based on their hypothesised influence on tropical productivity or previously documented relationships with tropical megaplanktivore aggregations globally (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\n\u003ch3\u003eZooplankton density\u003c/h3\u003e\n\u003cp\u003eLocal prey availability was assessed during surveys using the Zooplankton Visual Index (ZVI), a semi-quantitative scale used to categorise plankton density \u003cem\u003ein situ\u003c/em\u003e (Moloney et al. in press). The ZVI provides a rapid, reliable proxy for prey density that is standardised and validated (Moloney et al. in press). Additionally, horizontal in-water visibility was recorded as a proxy for turbidity and phytoplankton blooms, which may influence filter-feeding efficiency or predator detection (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDescription and data source of predictor variables that were used in the Generalised Linear Model\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eType\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eExplanation\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eGrouping\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSource\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eLevels (for categorical variables)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eRationale\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eMobula alfredi\u003c/em\u003e sightings\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eResponse (count)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNumber of individual \u003cem\u003eM. alfredi\u003c/em\u003e per survey\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBiological\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eIdentification (ID) photos (confirmed) and logbook count (estimated)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSightings from photo-identification (ID) and logbook data to investigate trends in sightings and movement (Watson et al. \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Armstrong et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Venables et al. \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eZooplankton density (binary, 0 or 1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eResponse (categorical)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIn-water assessment of Zooplankton Visual Index (ZVI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBiological\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLogbook\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003eLow (0)\u003c/b\u003e: Levels (0) complete absence of zooplankton or (1) thin layer or small patch\u003c/p\u003e \u003cp\u003e\u003cb\u003eHigh (1)\u003c/b\u003e: Levels (2) multiple layers or patches, (3) water is thick and cloudy, felt on skin or (4) water is dense and \u0026lsquo;soup-like\u0026rsquo;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eZooplankton density (low or high) as an index for food availability (Moloney et al. in press).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eZooplankton Visual Index (ZVI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (categorical)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIn-water assessment of Zooplankton Visual Index\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBiological\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLogbook\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0) complete absence of zooplankton, (1) thin layer or small patch, (2) multiple layers or patches, (3) water is thick and cloudy, felt on skin, (4) water is dense and \u0026lsquo;soup-like\u0026rsquo;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003ePrey availability influences manta ray abundance and behaviour (Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Armstrong et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Armstrong et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e; Venables et al. \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Moloney et al. in press).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater visibility (m)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (categorical)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThe underwater visibility was estimated horizontally at 5 m increments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBiological\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLogbook\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u0026ndash;5, 5\u0026ndash;10, 10\u0026ndash;15, 15\u0026ndash;20, 25+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eHigh water visibility shown to influence devil ray sightings, likely due to higher probability of being sighted (Venables et al. \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWeather\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (categorical)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThe weather was assigned to a category during each survey\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEnvironmental\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLogbook\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(1) clear skies and sunshine, (2) partial cloud cover, (3) overcast, (4) light rain, (5) heavy rain, (6) torrential rain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSunny weather conditions can influence sighting numbers of manta rays (Venables et al. \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSea state (Beaufort scale)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (categorical)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThe sea state was recorded during each survey using the Beaufort scale (0\u0026ndash;7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEnvironmental\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLogbook\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0) dead calm (sea like mirror), (1) light air (ripples without crests), (2) light breeze (small wavelets), (3) gentle breeze (large wavelets), (4) moderate breeze (small waves), (5) fresh breeze (moderate waves), (6) strong breeze (large waves), (7) near gale (sea heaps up)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSightings of \u003cem\u003eM. alfredi\u003c/em\u003e shown to be higher during a medium swell (Venables et al. \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Sea state is a combination of swell and wind.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWind direction (0\u0026ndash;360\u0026deg;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (continuous)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eWind direction at 10 m elevation measured hourly.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEnvironmental\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMeteoblue\u003c/p\u003e \u003cp\u003e(Meteoblue \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2025\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eWind direction shown to influence \u003cem\u003eM. alfredi\u003c/em\u003e abundance (Anderson et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Setyawan \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Couturier et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e)\u003c/p\u003e \u003cp\u003eand behaviour (Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e), including at Hanifaru Bay (Harris and Stevens \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWind speed (km/h)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (continuous)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eWind strength at 10 m elevation measured hourly\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEnvironmental\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMeteoblue (Meteoblue \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2025\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eWind speed can influence \u003cem\u003eM. alfredi\u003c/em\u003e abundance\u003c/p\u003e \u003cp\u003e(Couturier et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e)\u003c/p\u003e \u003cp\u003eand behaviour (Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e), including at Hanifaru Bay (Harris and Stevens \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCurrent direction\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (categorical)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIn-water assessment visually in relation to the reef\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOceanographic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLogbook\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(In) into the channel, (out) out of the channel, (slack) no direction\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eCurrent direction shown to influence \u003cem\u003eM. alfredi\u003c/em\u003e abundance and behaviour (Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Venables et al. \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) but was removed from the models due to confounding variables.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCurrent strength\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (categorical)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThe current strength was assessed underwater based on effort for kicking to remain in the same place\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOceanographic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLogbook\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0) none\u0026thinsp;=\u0026thinsp;able to hold position without finning, (1) slight\u0026thinsp;=\u0026thinsp;able to hold position with little effort, (2) moderate\u0026thinsp;=\u0026thinsp;able to hold position with strong finning, (3) strong\u0026thinsp;=\u0026thinsp;unable to hold position\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eCurrent strength influences manta ray abundance and behaviour (Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Ahsin et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Venables et al. \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSea Surface Temperature (SST) (\u0026deg; degrees Celsius)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (continuous)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eWater temperature on the reef.\u003c/p\u003e \u003cp\u003eFiltered Reprocessed data: 2007\u0026ndash;2021.\u003c/p\u003e \u003cp\u003eFiltered NRT: 2022\u0026ndash;2024.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOceanographic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCopernicus\u003c/p\u003e \u003cp\u003e(Worsfold et al. \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSST shown to influence manta ray sightings\u003c/p\u003e \u003cp\u003e(Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Weeks et al. \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Setyawan \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Beale et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Ahsin et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Fonseca-Ponce et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIndian Ocean Dipole (IOD) Index (-1\u0026thinsp;+\u0026thinsp;+\u0026thinsp;1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (continuous)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIndex value of warmer or cooler water in the Indian Ocean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOceanographic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eNOAA (NOAA Physical Sciences Laboratory \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eIOD index shown to affect \u003cem\u003eM. kuhlii\u003c/em\u003e sightings in Mozambique (Venables et al. \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTime from High Tide (\u0026plusmn;\u0026thinsp;6 hours)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (continuous)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThe time difference between the observation (middle of the survey) and closest high tide\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOceanographic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eFES2014 (Lyard et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) and \u003cem\u003eoce\u003c/em\u003e R package (Kelley and Richards \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eTide shown to influence zooplankton prey\u003c/p\u003e \u003cp\u003e(Cairns et al. in press), and \u003cem\u003eM. alfredi\u003c/em\u003e abundance and behaviour (Dewar et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Armstrong et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Couturier et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), including at Hanifaru Bay (Armstrong et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTide data obtained using a harmonic tidal model with \u003cem\u003eoce\u003c/em\u003e R package (Kelley and Richards \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Tidal constituents (i.e., m2, s2, n2, k2, k1, o1, p1, q1, m4, ms4, m6, l2, t2, s4, mn4) obtained from FES2014 (Lyard et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) and used to predict high/low tide and time. Tide epoch was set at: 1992-01-01.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTide size (m)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (continuous)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTide size difference between high and low tides on the day (from 0 m)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOceanographic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eFES2014 (Lyard et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) and \u003cem\u003eoce\u003c/em\u003e R package (Kelley and Richards \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eGreater tidal ranges may interact with complex nearshore bathymetry, increasing mixing in the water column and potentially promoting foraging opportunities for \u003cem\u003eM. birostris\u003c/em\u003e (Fonseca-Ponce et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLunar illumination (Proportion 0\u0026ndash;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (proportion)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eProportion of moon disk illuminated (0) new \u0026ndash; (1) full\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOceanographic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ePackage R\u003c/p\u003e \u003cp\u003eLunar package (Lazaridis \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eMoon phase/lunar illumination known to influence \u003cem\u003eM. alfredi\u003c/em\u003e abundance (Dewar et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Setyawan \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Couturier et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Andrzejaczek et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Harris and Stevens \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Fonseca-Ponce et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMonth\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (categorical)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMonth of observation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTemporal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLogbook\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003eFor Baa Atoll\u003c/b\u003e: Jan\u0026ndash;Dec\u003c/p\u003e \u003cp\u003e\u003cb\u003eFor Hanifaru Bay\u003c/b\u003e: May\u0026ndash;Nov\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eMonth used to examine seasonality of abundance due to changes in monsoon over the season (Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYear\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (categorical)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eYear of the observation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTemporal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLogbook\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eYear used to examine trends in \u003cem\u003eM. alfredi\u003c/em\u003e abundance over time (Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Venables et al. \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eM. alfredi\u003c/em\u003e behaviour\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePredictor (variable not in model; categorical)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePrimary behaviour of individual \u003cem\u003eM. alfredi\u003c/em\u003e per survey\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBiological\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLogbook\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(1) feeding; (2) cleaning; (3) cruising; and (4) courtship\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBehavioural observations from logbook data used to investigate habitat use (Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSurvey duration (minutes)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOffset variable (continuous)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDuration of the survey\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLogbook\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCapped at 300 minutes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eTo account for survey effort\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eObservers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOffset variable (count)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNumber of in-water observers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLogbook\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCapped at 4 observers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eTo account for the survey effort\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eTo assess the trends and drivers of \u003cem\u003eM. alfredi\u003c/em\u003e abundance and zooplankton density in Hanifaru Bay, we analysed the researcher-collected logbook survey data spanning 2007 to 2024. Generalised Linear Models (GLMs) were built using the statistical software R version 4.4.2 (R Core Team \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) to model the data from the logbook and remote sources, including a suite of temporal (month, year), environmental (weather, sea state, wind direction and strength), oceanographic (current strength, SST, time to high tide, tide size, lunar illumination, IOD index), and biological (ZVI, water visibility) predictors (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). During the SW Monsoon, there was commonly one survey per day (occasionally two), six days per week. Opportunistic surveys conducted between December and April were excluded to maintain consistency in search effort. The same researcher-collected dataset (2010 to 2024) was used to investigate zooplankton density trends and drivers; the shorter temporal range for zooplankton reflects the lack of records prior to 2010.\u003c/p\u003e \u003cp\u003eFour separate GLMs were built: \u0026ldquo;\u003cem\u003eManta trends\u003c/em\u003e\u0026rdquo; and \u0026ldquo;\u003cem\u003eManta drivers\u003c/em\u003e\u0026rdquo; (response: \u003cem\u003eM. alfredi\u003c/em\u003e abundance modelled as a count variable), and \u0026ldquo;\u003cem\u003eZooplankton trends\u003c/em\u003e\u0026rdquo; and \u0026ldquo;\u003cem\u003eZooplankton drivers\u003c/em\u003e\u0026rdquo; (response: binary zooplankton abundance modelled as a binary variable). For all four GLMs, model residuals were visually assessed for normality and homogeneity of variance.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eTrend models\u003c/strong\u003e \u003cp\u003eTo investigate the annual trends and seasonality at Hanifaru Bay, we constructed two models, one for \"\u003cem\u003eManta trends\u003c/em\u003e\" and the other for \"\u003cem\u003eZooplankton trends\u003c/em\u003e\" using temporal predictors (Month, Year).\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003e\"\u003c/b\u003e \u003cb\u003eManta trends\u003c/b\u003e \u003cb\u003e\" model\u003c/b\u003e: glm.nb(Manta count\u0026thinsp;~\u0026thinsp;Year\u0026thinsp;+\u0026thinsp;Month\u0026thinsp;+\u0026thinsp;Year:Month)\u003c/p\u003e \u003cp\u003e \u003cb\u003e\"\u003c/b\u003e \u003cb\u003eZooplankton trends\u003c/b\u003e \u003cb\u003e\" model\u003c/b\u003e: glm(Zooplankton density\u0026thinsp;~\u0026thinsp;Year\u0026thinsp;+\u0026thinsp;Month\u0026thinsp;+\u0026thinsp;Year:Month, family\u0026thinsp;=\u0026thinsp;binomial)\u003c/p\u003e \u003cp\u003eFor the \"\u003cem\u003eManta trends\u003c/em\u003e\" model, \u003cem\u003eM. alfredi\u003c/em\u003e abundance was treated as a continuous count variable. To account for overdispersion common to ecological count data (evident from an initial model using a Poisson error structure), the model was fitted with a negative binomial error structure using the glm.nb() function in the MASS R package (Venables and Ripley \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). To account for sampling biases inherent in logbook data, we standardised for survey effort by including survey duration and the number of observers as offsets in the \u0026ldquo;Manta trends\u0026rdquo; and the \u0026ldquo;Manta drivers\u0026rdquo; models.\u003c/p\u003e \u003cp\u003eFor the \"\u003cem\u003eZooplankton trends\u003c/em\u003e\" model, zooplankton density was treated as a binary response (0\u0026thinsp;=\u0026thinsp;ZVI levels 0,1, and 1\u0026thinsp;=\u0026thinsp;ZVI levels 2,3,4) and thus modelled as a binomial GLM with a logit link function. To define the degrees of freedom (df) for smoothing terms in the GLM, it was conservatively set at df\u0026thinsp;=\u0026thinsp;4.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eDriver models\u003c/strong\u003e \u003cp\u003eTwo additional GLMs were constructed to identify the primary influences on \u003cem\u003eM. alfredi\u003c/em\u003e abundance (\u0026ldquo;\u003cem\u003eManta drivers\u003c/em\u003e\u0026rdquo;) and high zooplankton density (\u0026ldquo;\u003cem\u003eZooplankton drivers\u003c/em\u003e\u0026rdquo;)\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003e\u0026ldquo;\u003cem\u003eManta drivers\u003c/em\u003e\u0026rdquo; model\u003c/strong\u003e \u003cp\u003eglm.nb(Manta count\u0026thinsp;~\u0026thinsp;ZVI\u0026thinsp;+\u0026thinsp;Water visibility\u0026thinsp;+\u0026thinsp;Lunar illumination\u0026thinsp;+\u0026thinsp;SST\u0026thinsp;+\u0026thinsp;Sea state\u0026thinsp;+\u0026thinsp;Wind speed\u0026thinsp;+\u0026thinsp;Wind direction\u0026thinsp;+\u0026thinsp;Hours from high tide\u0026thinsp;+\u0026thinsp;Current strength\u0026thinsp;+\u0026thinsp;IOD index).\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003e\u0026ldquo;\u003cem\u003eZooplankton drivers\u003c/em\u003e\u0026rdquo; model\u003c/strong\u003e \u003cp\u003eglm(Zooplankton density\u0026thinsp;~\u0026thinsp;Hours from high tide\u0026thinsp;+\u0026thinsp;Current strength\u0026thinsp;+\u0026thinsp;IOD index, family\u0026thinsp;=\u0026thinsp;binomial).\u003c/p\u003e \u003c/p\u003e \u003cp\u003eBoth driver models initially considered the same global suite of environmental and oceanographic predictors (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). We employed a hierarchical backward-selection approach to determine the most parsimonious model, beginning with a full model containing all predictors. Predictors were removed one-by-one, and model fit was reassessed at each step by identifying the candidate model with the lowest Bayesian Information Criterion (BIC) using step(), ensuring an optimal balance between explanatory power and model complexity. The significance of each model term in the final model selection was then confirmed by using anova() and likelihood ratio tests to ensure each remaining variable significantly improved the model fit.\u003c/p\u003e \u003cp\u003eThe \u0026ldquo;\u003cem\u003eManta drivers\u003c/em\u003e\u0026rdquo; model included ZVI with 5 levels, but ZVI was excluded as a predictor in the \u0026ldquo;\u003cem\u003eZooplankton drivers\u003c/em\u003e\u0026rdquo; model. To define the degrees of freedom (df) to smooth SST and wind speed in the GLM, it was conservatively set at df\u0026thinsp;=\u0026thinsp;3, unless the GLM indicated that df\u0026thinsp;=\u0026thinsp;4 more appropriately defined a relationship. A harmonic function (\u003cem\u003ek\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1) was applied to circular predictors (hour from high tide, wind direction, and lunar illumination) to account for their cyclic nature. Nagelkerke pseudo-\u003cem\u003eR\u0026sup2;\u003c/em\u003e values were calculated for all four final models to assess the goodness-of-fit (Nagelkerke \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e1991\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003eAbundance and behaviour of \u003cem\u003eMobula alfredi\u003c/em\u003e in Baa Atoll\u003c/p\u003e \u003cp\u003eFrom the total sightings of \u003cem\u003eM. alfredi\u003c/em\u003e in Baa Atoll (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;71,341), 21 key sites were identified. However, most of these sightings (72.3%; \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;51,555) were recorded at one location, Hanifaru Bay (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), and the remaining 27.7% of the sightings were spread across 58 additional sites. Consequently, subsequent analyses focus exclusively on Hanifaru Bay, as the disproportionately high sightings frequency and high survey replication at this site provide the statistical power necessary to minimise modelling variability. Most sightings at Hanifaru Bay were during the SW Monsoon between May\u0026ndash;November (99.4%), where the most frequently observed primary behaviour was feeding (70.5%), with cleaning being the second most common (22.9%), and cruising (8.3%) and courtship (2.3%) behaviours observed less frequently (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). \u003cem\u003eMobula alfredi\u003c/em\u003e were seen on 82.5% of surveys, with observers counting a mean of 24.6\u0026thinsp;\u0026plusmn;\u0026thinsp;0.8 \u003cem\u003eSE\u003c/em\u003e and up to 244 individual \u003cem\u003eM. alfredi\u003c/em\u003e per survey.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eMonthly and inter-annual \u0026ldquo;\u003cem\u003eManta trends\u003c/em\u003e\u0026rdquo; at Hanifaru Bay\u003c/p\u003e \u003cp\u003eAt Hanifaru Bay, \u003cem\u003eM. alfredi\u003c/em\u003e abundance \u0026ndash; derived from the \u0026ldquo;\u003cem\u003eManta trends\u003c/em\u003e\u0026rdquo; model for surveys conducted between May\u0026ndash;November (2007\u0026ndash;2024) and standardised for the number of observers and survey duration \u0026ndash; was significantly related to Month and Year (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). A significant interaction between month and year (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) indicated that annual peaks in abundance fluctuated seasonally. While the model identified a historical peak in \u003cem\u003eM. alfredi\u003c/em\u003e abundance in 2007 (driven largely by the high September sightings), analysis of the 2010\u0026ndash;2024 period revealed distinct peaks in 2021 and 2022 (predicted mean 38.6 and 37.7 individuals). In these years, specific months, including May, June, October, and November, reached their highest predicted abundances (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). Conversely, the lowest abundance in this period was in 2017. For monthly trends, the model predicted a peak in \u003cem\u003eM. alfredi\u003c/em\u003e sightings generally during August (35.5 individuals), followed by September (33.5 individuals).\u003c/p\u003e \u003cp\u003eMonthly and inter-annual \u0026ldquo;\u003cem\u003eZooplankton trends\u003c/em\u003e\u0026rdquo;\u003c/p\u003e \u003cp\u003eAt Hanifaru Bay, binary zooplankton density from the \u0026ldquo;\u003cem\u003eZooplankton trends\u003c/em\u003e\u0026rdquo; model spanning May\u0026ndash;November in the years between 2010\u0026ndash;2024 was significantly related to Year (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) but not to Month (p\u0026thinsp;=\u0026thinsp;0.595). There was, however, a significant interaction between Month and Year (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Like the \u003cem\u003eM. alfredi\u003c/em\u003e models, zooplankton density fluctuated seasonally, with a predicted peak in 2022 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb). Almost all months \u0026ndash; excluding July and August \u0026ndash; reached their maximum values between 2021 and 2022. The model also identified the lowest density of zooplankton to be in 2017. There were also monthly trends with peaks in November (0.431), followed by August (0.426).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eModel for \u0026ldquo;\u003cem\u003eManta drivers\u003c/em\u003e\u0026rdquo;\u003c/p\u003e \u003cp\u003eThe model of best fit for \u003cem\u003e\u0026ldquo;Manta drivers\u0026rdquo;\u003c/em\u003e included ten predictors (zooplankton density, water visibility, lunar illumination, hours from high tide, current strength, SST, sea state, wind speed and direction, and IOD index). This model explained 95.2% of the variance for the predictors, with the dominant predictor, zooplankton density (ZVI), explaining 61.1% (using Nagelkerke R\u0026sup2;). While the remaining variance was attributed to secondary environmental and oceanographic drivers, such as sea state, followed by SST, and wind direction and speed. The ZVI was a strongly significant predictor of \u003cem\u003eM. alfredi\u003c/em\u003e abundance (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), with an increasing trend in sightings with higher zooplankton density (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Comparing the predicted count for the highest recorded zooplankton density (ZVI 4) to the complete absence of zooplankton (ZVI 0), predicted manta abundance was 2,590.5% higher during ZVI 4 conditions. Further, predicted abundance was 311.4% higher during ZVI 4 conditions compared to minimal zooplankton presence (ZVI 1). Water visibility significantly influenced abundance (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), with sightings reaching a peak at 10\u0026ndash;20 m. Abundance declined by 54% in clear water conditions (\u0026gt;\u0026thinsp;20 m) and by 13.3% in highly turbid conditions (0\u0026ndash;10 m, i.e., when there was extensive phytoplankton and/or sediment in the water column).\u003c/p\u003e \u003cp\u003eThe predicted \u003cem\u003eM. alfredi\u003c/em\u003e abundance was significantly influenced by the lunar illumination cycle (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The abundance rate varied throughout the lunar month, with predicted abundance peaking around the new moon (0.1 proportion), followed by the full moon (1.0). Abundance during a new moon was predicted to be 22.2% higher compared to a full moon. The number of predicted \u003cem\u003eM. alfredi\u003c/em\u003e abundance varied significantly with the tidal cycle (hours from high tide; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Abundance peaked 2.7 hours after high tide and reached a minimum 3.3 hours before high tide. Predicted abundance was 33.4% higher during the optimal tidal state compared to the minimal tidal state. Abundance was also predicted to be 54.8% higher in a stronger current than in no current and 84.6% higher than when there was a weak current (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e \u003cp\u003eSea state was a significant predictor of \u003cem\u003eM. alfredi\u003c/em\u003e abundance (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), with abundance predicted to increase by 564.9% when there was a light breeze (Beaufort scale 1) compared to calm waters (Beaufort scale 0). The magnitude of the positive effect generally declined as the sea state increased to a strong breeze. Wind speed had a non-linear significant influence on predicted \u003cem\u003eM. alfredi\u003c/em\u003e abundance (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), with sightings peaking at 0 km/h and again at 30.3 km/h. Sightings dropped when the wind speed was lower (13.3 km/h). Predicted abundance was 43.2% higher during the optimal wind conditions compared to the minimal wind conditions. The predicted abundance of \u003cem\u003eM. alfredi\u003c/em\u003e was also significantly influenced by the wind direction (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The abundance peaked when the wind was north-northwest (NNW) in direction (345.0\u0026deg;) and decreased when the wind came from the south-southeast (SSW; 165.5\u0026deg;). Predicted abundance was 66.1% higher during the optimal wind direction compared to the minimal wind direction.\u003c/p\u003e \u003cp\u003eWarmer waters had a significant negative effect on \u003cem\u003eM. alfredi\u003c/em\u003e abundance (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). While sightings peaked around 28.6\u0026deg;C, predicted abundance declined sharply at warmer temperatures, declining by 24.4% for the first 1.0\u0026deg;C and 15.7% per degree after this. When comparing the optimal conditions to the highest observed temperature of 31.1\u0026deg;C, the predicted abundance of \u003cem\u003eM. alfredi\u003c/em\u003e was 45.0% lower. The IOD index also significantly impacted the abundance of \u003cem\u003eM. alfredi\u003c/em\u003e (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), with a 26.0% increase in sightings when it moved from a negative IOD phase to a positive IOD phase (\u003cem\u003e\u0026szlig;\u003c/em\u003e=0.175). While the univariate model showed a negative correlation between IOD and \u003cem\u003eM. alfredi\u003c/em\u003e abundance (\u003cem\u003e\u0026szlig;\u003c/em\u003e=-0.147).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eModel for \u0026ldquo;Zooplankton drivers\u0026rdquo;\u003c/p\u003e \u003cp\u003eThe \u0026ldquo;\u003cem\u003eZooplankton drivers\u003c/em\u003e\u0026rdquo; model with a binary response (0\u0026thinsp;=\u0026thinsp;low ZVI and 1\u0026thinsp;=\u0026thinsp;high ZVI) densities included three predictors (hours from high tide, current strength, and IOD index). This model explained 8.5% of the variance for the predictors (using Nagelkerke R\u0026sup2;). The strongest predictor was current strength explaining 4.6% of the variance, while the remaining variance was attributed to secondary oceanographic drivers such as hours from high tide, followed by the IOD index.\u003c/p\u003e \u003cp\u003eCurrent strength was positively correlated with high zooplankton density (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e), with densities predicted to be 335.9% higher when there was a stronger current than in no current and 32.8% higher than when there was a weak current. Zooplankton density varied significantly with the tidal cycle (hours from high tide; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The odds of observing high density peaked at 1.6 hours after high tide and reached a minimum 4.4 hours before high tide. The predicted odds of high-density zooplankton were 130.3% higher during the optimal tidal state compared to the minimal tidal state. The IOD index significantly impacted zooplankton density (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), with a 72.8% increase in density when it moved from a strong positive IOD phase to a negative IOD phase.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur 18-year dataset (2007\u0026ndash;2024) confirms that Hanifaru Bay functions as a critical foraging aggregation site during the SW Monsoon, accounting for 71% of all feeding-related sightings recorded in eastern Baa Atoll. The high degree of synchrony between \u003cem\u003eM. alfredi\u003c/em\u003e abundance and prey density \u0026mdash; both peaking in the later months of the season and reaching a decadal maximum in 2021\u0026ndash;2022 \u0026mdash; underscores a tight coupling between biological activity and environmental forcing. Our models suggest the utility of this hotspot is regulated by two primary tiers of influence: (1) local-scale biophysical drivers, including lunar and tidal phases, wind-driven currents, and SST; and (2) broad-scale climatic indices, such as the IOD. While the bathymetry of Hanifaru Bay is a constant feature, its retention of zooplankton fluctuates in response to these local and regional indices which modulate the local environmental conditions (e.g., wind and SST) required to trigger mass-foraging events. Given the reliance of \u003cem\u003eM. alfredi\u003c/em\u003e on these high-density, ephemeral zooplankton patches, climate-driven shifts in oceanographic regimes may fundamentally alter the functional productivity of this site, impacting the energetic fitness of this vulnerable population (Stewart et al. \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e\n\u003ch3\u003eLong-term trends reveal interannual and seasonal variability\u003c/h3\u003e\n\u003cp\u003eThis nearly two-decade study reveals interannual and seasonal fluctuations in \u003cem\u003eM. alfredi\u003c/em\u003e abundance, reflecting the dynamic nature of the Maldivian preyscape, though annual sightings remained relatively steady. Seasonally, the August/September peaks aligns with the maturation of the SW Monsoon, a period characterised by sustained upwelling and productivity (Anderson et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). While similar seasonal pulses are documented in other megaplanktivores globally (Witt et al. \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Klotz et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Couto et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Harvey-Carroll et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; McInturf et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), the magnitude of the Hanifaru Bay aggregations remains globally unparalleled.\u003c/p\u003e \u003cp\u003eThe predictable recurrence of these aggregations is underpinned by high site fidelity, as photo-ID records confirm individuals returning to the same reef system across many years (Marshall et al. \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Couturier et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). While site fidelity represents phenotypically plastic responses to habitat quality, and niche specialisation, the consistent return of \u003cem\u003eM. alfredi\u003c/em\u003e individuals to Hanifaru Bay suggests a high-fidelity reliance on the site\u0026rsquo;s unique foraging opportunities (Armstrong et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2021a\u003c/span\u003e; Knochel et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). This behavioural strategy allows \u003cem\u003eM. alfredi\u003c/em\u003e \u0026ndash; and other large megaplanktivores such as \u003cem\u003eM. birostris\u003c/em\u003e and \u003cem\u003eR. typus\u003c/em\u003e \u0026ndash; to maximise the exploitation of high quality foraging niches within a patchy marine environment (Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Guzman et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Rubin et al. \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Consequently, local abundance at Hanifaru Bay serves as a high-fidelity proxy for the efficiency of the site\u0026rsquo;s physical mechanisms. While morphology and bathymetry remains constant, the formation of this retention zone is governed by broader Indian Ocean productivity (Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Interannual variability likely corresponds to large-scale climatic indices, specifically the IOD and the ENSO, which can modulate regional sea levels, rainfall, and thermal profiles (Saji et al. \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Carruthers et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe 2021\u0026ndash;2022 peak coincided with a record-breaking Negative IOD and a triple-dip La Ni\u0026ntilde;a event (starting mid-2020 and continuing till 2022, with 2022 being the strongest on record), which intensified the westerly monsoonal winds (Srivastava et al. \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2026\u003c/span\u003e) and maximised the wind-driven concentration of prey. The Negative IOD, persisting from mid-2021 through 2022, was the first ever multi-year of its kind since the 1960s, occurring alongside wetter conditions and higher sea levels, with wind anomalies also mainly remaining westerly (Chowdhury et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Srivastava et al. \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2026\u003c/span\u003e). In contrast, the 2017 trough highlights ecosystem vulnerability following a strong El Ni\u0026ntilde;o event (Cowburn et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Chowdhury et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and a Positive IOD phase (NOAA Physical Sciences Laboratory \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Notably, 2016 was also associated with mass coral bleaching and thermal anomalies in the Maldives (Cowburn et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). During this period, the Maldives experienced dry spells and lower than average sea levels (Chowdhury et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), resulting in extreme surface warming which likely suppressed vertical mixing and disrupted prey patch formation required for mass-foraging (Min and Noh \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). As climatic events increasingly squeeze the regional preyscape, ensuring that mobile planktivores can forage with limited anthropogenic disturbances at a hotspot like Hanifaru Bay is essential for population resilience.\u003c/p\u003e \u003cp\u003e \u003cb\u003eZooplankton density is the primary predictor of\u003c/b\u003e \u003cb\u003eMobula alfredi\u003c/b\u003e \u003cb\u003eabundance\u003c/b\u003e\u003c/p\u003e \u003cp\u003eHigh zooplankton density (using ZVI) was the dominant predictor of \u003cem\u003eM. alfredi\u003c/em\u003e abundance, explaining\u0026thinsp;\u0026gt;\u0026thinsp;60% of the observed variation in sightings. Sightings were \u0026gt;\u0026thinsp;2,500% higher during peak prey availability compared to periods of apparent prey absence. This strong correlation aligns with global observations of megaplanktivores targeting localised productivity and dense zooplankton patches (Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Armstrong et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Armstrong et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e; Venables et al. \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). However, our findings suggest the attraction to Hanifaru Bay may extend beyond prey volume, with this site likely functioning as a location for metabolic optimisation. Foraging in the deep, cold offshore waters imposes thermoregulatory costs, requiring specialised counter-current heat exchange systems in the cranial and branchial regions to maintain core temperatures (Arostegui \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Hanifaru Bay is unique because the physical mechanisms transport and concentrate deep-water prey into shallow, sun-warmed waters. This allows \u003cem\u003eM. alfredi\u003c/em\u003e to exploit high density prey while simultaneously reducing the metabolic cost of thermoregulation (Arostegui \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). By mitigating the thermal stress typically associated with deep-water foraging (Beale et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), the site provides an energetic surplus that likely facilitates the high levels of social and secondary behaviours observed, such as cleaning and courtship (Stevens \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), while also reducing their exposure to predators.\u003c/p\u003e \u003cp\u003eFurther, \u003cem\u003eM. alfredi\u003c/em\u003e presence was negatively impacted by high content of phytoplankton and/or suspended sediment\u0026ndash;similar to observations in Mozambique (Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), suggesting water quality is a critical constraint. High sediment loads may impose significant physiological costs, potentially impeding gill ventilation and reducing filter-feeding efficiency, thereby driving mobile planktivores to seek clearer, prey-rich environments (Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Consequently, rather than searching wide areas for diffuse prey, \u003cem\u003eM. alfredi\u003c/em\u003e appears energetically tethered to Hanifaru Bay. The site offers a rare intersection of high-density prey, clear water, and thermal refuge, allowing for maximising efficient foraging and high calorific intake, while minimising the metabolic and physiological tolls associated with search effort and environmental stress (Armstrong et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cb\u003eMobula alfredi\u003c/b\u003e \u003cb\u003eabundance: Lunar, tidal, and wind-driven convergence\u003c/b\u003e\u003c/p\u003e \u003cp\u003eBeyond direct prey density, \u003cem\u003eM. alfredi\u003c/em\u003e abundance is modulated by an interplay of lunar, tidal, and wind-driven mechanisms that facilitate the retention of zooplankton. We observed a\u0026thinsp;\u0026gt;\u0026thinsp;30% increase in sightings during the new and full moon phases; a pattern consistent with foraging aggregations on the Great Barrier Reef in Australia and in Komodo, Indonesia, where activity is timed with spring tides (Dewar et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). At Hanifaru Bay, tidal currents typically shift 180 degrees between phases, transitioning from an incoming flood (east\u0026ndash;west) and an outgoing ebb (west\u0026ndash;east). This lunar influence is intrinsically linked to current velocity; sightings at Hanifaru Bay were ~\u0026thinsp;50% higher during high-velocity currents, which likely facilitate the transport of oceanic zooplankton into the reef system. Abundance peaked 2.7 hours after high tide, when intensified outgoing lunar and high tide currents appear to overcome prevailing monsoonal flows, forcing plankton-rich oceanic water into the shallow reef systems of the atoll (Stevens \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOptimal foraging conditions were associated with NNW winds (25\u0026ndash;35 km/h), which trigger the development of high-energy Langmuir Circulation (LC) \u0026ndash; a series of shallow, wind-driven, counter-rotating cells that concentrate buoyant zooplankton into visible surface slicks (Li \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Smith \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Thorpe \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). At ~\u0026thinsp;30 km/h, such as what we have reported in Hanifaru Bay, the LC cells can extend 5\u0026ndash;10 m deep, concentrating a larger volume of the water column, and bringing prey to the surface (Thorpe \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Because Hanifaru Bay opens to the NNW, these LC convergence zones align with the Bay\u0026rsquo;s primary axis, effectively accumulating prey into a bathymetric cul-de-sac where reef morphology prevents downwind dispersal and maximises prey density (Li \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2000\u003c/span\u003e).\u003c/p\u003e\n\u003ch3\u003eThermal thresholds and climatic sensitivity\u003c/h3\u003e\n\u003cp\u003eDespite the efficiency of the Hanifaru Bay physical retention zone, we identified clear environmental thresholds that constrain its functional utility. \u003cem\u003eMobula alfredi\u003c/em\u003e abundance was highly sensitive to thermal shifts, declining by 24.4% for the first 1.0\u0026deg;C of warming above the 28.6\u0026deg;C optimum. Across the observed warming range, the average decline of 15.7% per degree suggests that even minor increases in SST may lead to a disproportionate reduction in the abundance at this site. While the thermal limit in Komodo, Indonesia, is thought to be 29\u0026deg;C (Dewar et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2008\u003c/span\u003e), the \u003cem\u003eM. alfredi\u003c/em\u003e population in the Red Sea routinely forage in higher temperatures (Stevens Pers. Obs.). This suggests that the Maldives population may be acclimatised to a localised thermal window, beyond which metabolic costs outweigh the caloric gains (Dewar et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Couturier et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). As \u003cem\u003eM. alfredi\u003c/em\u003e use regional endothermy and counter-current heat exchange to regulate vital organ temperatures (Arostegui \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), hot SSTs may inhibit foraging by exceeding physiological preferences or impairing the ability to shed metabolic heat generated during high-intensity feeding (Pistevos et al. \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2015\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eBeyond physiology, SST likely serves as a proxy for broader thermocline dynamics. The presence of oceanic-derived zooplankton within Hanifaru Bay (Armstrong et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e) is dependent on the transport of cool, nutrient-rich waters onto the atoll\u0026rsquo;s shallow shelf. As the Indian Ocean warms and stratifies, a deepening thermocline may effectively sink this prey-laden water below the depth of the atoll\u0026rsquo;s shelf (Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). If the thermocline depth exceeds the bathymetric shelf of the atoll, the influx of zooplankton-rich water could be physically obstructed, potentially limiting the availability of oceanic prey at inshore sites. Consequently, while the retention zone remains constant, its function as a productive hotspot may be increasingly vulnerable to a warming oceanographic regime that threatens to decouple the predator from a primary nutrient supply chain.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eOceanographic drivers of prey resources\u003c/h2\u003e \u003cp\u003eOur zooplankton models indicate that current strength, tidal phase, and the IOD index are the strongest predictors of prey density. Zooplankton density peaked\u0026thinsp;~\u0026thinsp;1.6 hours after high tide (a 130.3% increase), supporting the hypothesis of a physical retention mechanism that transports offshore species, such as \u003cem\u003eUndinula vulgaris\u003c/em\u003e onto the atoll (Armstrong et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e). Once within Hanifaru Bay, unique bathymetry generates eddies that concentrate biomass (Cairns et al. in press). These eddies form midway through the flood tide and dissipate\u0026thinsp;~\u0026thinsp;2 hours after high tide, acting as the dominant mechanism driving feeding aggregations (Cairns et al. in press).\u003c/p\u003e \u003cp\u003eThe models reveal a distinct temporal lag between prey availability and peak predator sightings; while zooplankton density peaked\u0026thinsp;~\u0026thinsp;1.6 hours after high tide, \u003cem\u003eM. alfredi\u003c/em\u003e abundance peaked\u0026thinsp;~\u0026thinsp;1 hour later (~\u0026thinsp;2.7 hours post high tide). This sequence suggests a high-fidelity behavioural response as individuals aggregate specifically as the ebbing tide bottlenecks prey within the bathymetry (Armstrong et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e). Tidal forcing is a well-documented driver of zooplankton enrichment in other topographically complex regions, such as the Komodo National Park, Indonesia (Dewar et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2008\u003c/span\u003e), and the Great Barrier Reef, Australia (Armstrong et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Beyond localised tidal forcing, the interannual availability of prey at Hanifaru Bay is significantly modulated by the IOD. Our models demonstrate that zooplankton densities significantly increase during Negative IOD phases, which align with intensified westerly monsoonal winds, cooler than average SSTs, and enhanced chlorophyll-\u003cem\u003ea\u003c/em\u003e concentrations in the central Maldives (Srivastava et al. \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2026\u003c/span\u003e; Abeywickrama et al. under review). This enrichment is driven by a shallower thermocline and wind-stress conditions that facilitate the upwelling of nutrient-rich subsurface waters (Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Thushara and Vinayachandran \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Abeywickrama et al. under review). Conversely, during a Positive IOD, the thermocline deepens, and weaker westerly winds induce downwelling, suppressing the nitrate and phosphate transport essential for phytoplankton growth (Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Abeywickrama et al. under review).\u003c/p\u003e \u003cp\u003eInterestingly, while zooplankton increased during Negative IOD events, our multivariate model showed a decline in \u003cem\u003eM. alfredi\u003c/em\u003e abundance\u0026mdash;a divergence from the univariate results. This suggests that while a Negative IOD creates an optimal climatic window for prey production, predator recruitment is simultaneously constrained by local-scale factors, despite trends for the highest individual years of abundance occurring during Negative IOD phase. Although research on the interactions between ENSO and the IOD in the Maldives remains limited (Chowdhury et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), our data underscores the vulnerability of this hotpot to broader climatic shifts that regulate the productivity of the Indian Ocean.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eImplications for a changing Indian Ocean\u003c/h2\u003e \u003cp\u003eThe long-term stability of this hotspot is tied to the SW Monsoon, which drives the deep-water upwelling essential for regional productivity (Sasamal \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Anderson et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Harris et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). However, the South Asian Monsoon is increasingly influenced by anthropogenic climate change, with a documented 20% reduction in marine phytoplankton over the past six decades (Joseph and Simon \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Rapid surface warming has led to enhanced ocean stratification that prevents nutrient-rich subsurface waters from reaching the euphotic zone (Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). As warming intensifies, there is an acute risk of reduced prey biomass reaching the atoll due to inhibited upwelling (Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Singh \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), and the breakdown of the concentration mechanisms required to aggregate prey into forageable patches (Min and Noh \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2004\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAs \u003cem\u003eM. alfredi\u003c/em\u003e are energetically tethered to these physical drivers, they serve as effective bio-indicators of regional oceanographic health. Their presence at Hanifaru Bay acts as a signal that the physical retention zone is functioning; conversely, their absence during historically productive windows may signal a shift toward a more oligotrophic state. Rising SSTs may cause retention zones to fail even during periods of stable regional productivity (Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), as the thermal buoyancy of the surface layer overcomes the wind-driven mixing required to form surface slicks (Min and Noh \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Climate-induced changes may force this resident population to travel further or forage deeper, increasing metabolic search costs and predation risks, while potentially impacting individual fitness and reproductive output (Stewart et al. \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eConservation relevance: spatial-temporal bottleneck\u003c/h2\u003e \u003cp\u003eAs Hanifaru Bay hosts\u0026thinsp;\u0026gt;\u0026thinsp;70% of \u003cem\u003eM. alfredi\u003c/em\u003e sightings recorded in eastern Baa Atoll in the SW Monsoon, it represents a non-substitutable habitat. The tendency to aggregate in these confined areas increases vulnerability to vessel strikes, discarded fishing gear, pollution, and concentrated tourist groups, as a single threat may impact many individuals simultaneously (Croll et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Stewart et al. \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Strike et al. \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Palacios et al. \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Designated as an MPA in 2009, Hanifaru Bay serves as the core protected zone within the Baa Atoll Biosphere Reserve (Murray et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Baa Atoll Biosphere Reserve \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). While management regulations (e.g., prohibition on scuba diving, mandatory licensed guides, a code-of-conduct for interacting with wildlife, and strict caps on the number of concurrent visitors and vessels) enforced by on-site rangers provide a robust framework for managing direct interactions (Murray et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Baa Atoll Biosphere Reserve \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), the environmental cues that trigger mass foraging also create a spatial and temporal bottleneck of anthropogenic exposure. This cumulative behavioural stress of high-density wildlife activity and intense human interest mirrors dynamics seen in terrestrial systems, such as Yellowstone National Park, where predictable seasonal movements concentrate human-wildlife conflict into localised high-risk zones (Soulsbury and White \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Middleton et al. \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Fortin et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Protecting the ecosystems surrounding Hanifaru Bay may provide umbrella benefits, safeguarding the preyscapes used by other vulnerable megaplanktivores, including \u003cem\u003eR. typus\u003c/em\u003e and other mobulid species (Stevens and Froman \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; The Manta Trust \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eCaveats\u003c/h2\u003e \u003cp\u003eWhile the dataset provides an insight into the trends and drivers of \u003cem\u003eM. alfredi\u003c/em\u003e sightings and their prey, several limitations must be acknowledged. First, the study relies on logbook data which are subject to observer bias, including variations in survey effort and the difficulty of accuracy quantifying individuals during mass-aggregation events. As well as non-uniform sampling effort with survey frequency often fluctuating with monsoonal weather and seasonal accessibility. However, the sheer volume of sightings and the nearly two-decade duration of the study provide a robust overview that effectively smooths these individual survey anomalies and temporal variations. Second, zooplankton density was assessed using ZVI rather than quantitative biomass sampling (Moloney et al. in press). While this introduces a level of subjectivity, the ZVI is a validated and reliable proxy for the relative abundance of prey (Moloney et al. in press). Third, the use of GLMs identifies potential associations but cannot establish biological causation, particularly for the zooplankton drivers model where only a small amount of variation was explained. Hence, future work could introduce structural causal models to help disentangle the effect of current velocity from the effect of the lunar cycle (Arif and MacNeil \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Fourth, certain environmental predictors (e.g., currents, tides) were recorded at qualitatively or coarse resolution, and other potential drivers (e.g., nutrient concentrations, fine-scale oceanography) were not included. Future work could consider including these variables at a fine-scale resolution. Lastly, a notable divergence was observed where zooplankton density increased during a Netative IOD phase, while \u003cem\u003eM. alfredi\u003c/em\u003e abundance (only in the multivariate model) declined. We are unsure why this has occurred.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eOur 18-year investigation into the world\u0026rsquo;s largest aggregation of \u003cem\u003eM. alfredi\u003c/em\u003e represents one of the longest continuous datasets of its kind, revealing a multi-scale interaction of localised oceanography and broad-scale climatic drivers. While previous research established that site-specific environmental conditions are required to trigger mass-foraging in megaplanktivores (Sleeman et al. \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Jaine et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Rohner et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Weeks et al. \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Harris and Stevens \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), our results elucidate the specific environmental configuration of high-velocity tides, NNW winds, and spring lunar phases which drive these events at Hanifaru Bay. These factors activate a bathymetric prey retention zone where wind-induced LC and tidal eddies concentrate zooplankton into hyper-dense surface patches required for high-energy feeding (Harris and Stevens \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Armstrong et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e; Cairns et al. in press).\u003c/p\u003e \u003cp\u003eCrucially, our findings suggest that Hanifaru Bay functions as a site for both foraging and metabolic optimisation. By concentrating deep-water prey into shallow, sun-warmed waters, Hanifaru Bay allows \u003cem\u003eM. alfredi\u003c/em\u003e to maximise caloric intake while minimising the thermoregulatory costs associated with deep-water foraging. However, this energetic advantage is governed by a thermal threshold; the observed 28.6\u0026deg;C tipping point suggests that while Hanifaru Bay\u0026rsquo;s bathymetry is a constant, its functional utility is highly vulnerable to a warming and more stratified Indian Ocean (Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Rising temperatures and a deepening thermocline threaten to decouple the \u003cem\u003eM. alfredi\u003c/em\u003e population from a primary nutrient supply, even if regional productivity remains stable. As the Indian Ocean trends towards a more oligotrophic state (Roxy et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), the importance of reliable foraging hotspots like Hanifaru Bay \u0026mdash; which hosts 71% of regional sightings \u0026mdash; will only intensify. The site represents a non-substitutable habitat and a critical spatial-temporal bottleneck where both ecological activity and anthropogenic pressure converge. By defining these oceanographic boundaries, this study provides a framework for management strategies to evolve alongside a changing climate, safeguarding the reproductive fitness and long-term resilience of this vulnerable population.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eCompeting Interests\u003c/h2\u003e \u003cp\u003eThere are no competing interests to declare from authors.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eEthics Approval\u003c/h2\u003e \u003cp\u003eApproval to undertake this research was received from the Maldives Ministry of Fisheries, Marine Resources and Agriculture (annually renewable permit: [\u003cem\u003e(FRM)30-D/PRIV/2022/54\u003c/em\u003e]) and the Maldives Environmental Protection Agency (annually renewable permit: [\u003cem\u003eEPA/2023/PSR-M04\u003c/em\u003e]). This research was in accordance with the University of the Sunshine Coast Animal Ethics [\u003cem\u003eANS23101\u003c/em\u003e]. Documentary evidence available on request.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eHannah M. Moloney was funded by the University of the Sunshine Coast Research Training Program scholarship. This study was made possible due to funding from the Save Our Seas Foundation, Carl F. Bucherer and logistical support and funding from the Four Seasons Resort Maldives at Landaa Giraavaru. The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.\u003c/p\u003e\u003ch2\u003eAuthor Contributions\u003c/h2\u003e \u003cp\u003eHannah M. Moloney, Guy M.W. Stevens, Asia O. Armstrong, Christine l. Dudgeon, Kathy, A. Townsend and Anthony J. Richardson made substantial contributions to the study conception and design. Data collection was performed by Hannah M. Moloney, Guy M.W. Stevens, Elspeth M. Strike, and Tamaryn J. Sawers. Data analysis and interpretation was performed by Hannah M. Moloney, Alvise Dabala and Anthony J. Richardson, with Kathy, A. Townsend and Guy M.W. Stevens also contributing to the interpretation of the data. The first draft of the manuscript was written by Hannah M. Moloney and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eThe authors acknowledge the extensive data management, and logistical and field assistance from the team at the \u003cem\u003eMaldives Manta Conservation Programme\u003c/em\u003e and \u003cem\u003eThe Manta Trust\u003c/em\u003e, specifically, Bethany Faulkner, Tiff Bond, Niv Froman, Moosa Mohamed, Annie Murrray, Simon Hilbourne, Katie Lee-Brooks, Ibrahim Lirar, Hussain Rasheed, and Yoosuf Abdul Haadhee. We thank the \u003cem\u003eFour Seasons Resort Maldives\u003c/em\u003e team based at Landaa Giraavaru for providing ongoing support to this research project, particularly the General Manager, Armando Kraenzlin. Authors would like to thank the \u003cem\u003eMaldives\u0026rsquo; Ministry of Fisheries, Marine Resources and Agriculture\u003c/em\u003e and the \u003cem\u003eMaldives Environmental Protection Agency\u003c/em\u003e (now formally known as the \u003cem\u003eEnvironmental Regulatory Authority\u003c/em\u003e), who granted permission to undertake this research. We also acknowledge the dedicated support of the \u003cem\u003eBiosphere Reserve Office\u003c/em\u003e, particularly the sea rangers. Thank you to \u003cem\u003eSave Our Seas Foundation\u003c/em\u003e and \u003cem\u003eCarl F. Bucherer\u003c/em\u003e for funding the field operations.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e \u003cp\u003eThe datasets generated and analysed during the current study \u0026ndash; logbook data (including manta ray abundance and ZVI observations) \u0026ndash; are available on reasonable request from the corresponding author.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAbeywickrama T, Pathirana G, Noh K-M, Dissanayake K, Wang D Lee D-G (under review) Asymmetric effects of Indian Ocean dipole on surface chlorophyll variability in the Indian Ocean\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAhsin A, Hartati R, Sitorus ED, Azizah H, Endrawati H (2022) Oceanographic factors on coastal aggregation of Reef Manta (\u003cem\u003eMobula alfredi\u003c/em\u003e) in the Manta Sandy, Raja Ampat, Indonesia. ILMU Kelaut Indones J Mar Sci 27(4):330\u0026ndash;340. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.14710/ik.ijms.27.4.330-340\u003c/span\u003e\u003cspan address=\"10.14710/ik.ijms.27.4.330-340\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnderson RC, Adam MS, Goes JI (2011) From monsoons to mantas: seasonal distribution of (\u003cem\u003eMobula alfredi\u003c/em\u003e) in the Maldives. Fish Oceanogr 20(2):104\u0026ndash;113. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1365-2419.2011.00571.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1365-2419.2011.00571.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndrzejaczek S, Chapple T, Curnick D, Carlisle A, Castleton M, Jacoby D, Peel L, Schallert R, Tickler D, Block B (2020) Individual variation in residency and regional movements of reef manta rays \u003cem\u003eMobula alfredi\u003c/em\u003e in a large marine protected area. Mar Ecol Prog Ser 639:137\u0026ndash;153. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3354/meps13270\u003c/span\u003e\u003cspan address=\"10.3354/meps13270\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArif S, MacNeil MA (2023) Applying the structural causal model framework for observational causal inference in ecology. Ecol Monogr 93(1):e1554. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ecm.1554\u003c/span\u003e\u003cspan address=\"10.1002/ecm.1554\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArmstrong AO, Armstrong AJ, Bennett MB, Richardson AJ, Townsend KA, Everett JD, Hays GC, Pederson H, Dudgeon CL (2021a) Mutualism promotes site selection in a large marine planktivore. Ecol Evol 11(10):5606\u0026ndash;5623. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ece3.7464\u003c/span\u003e\u003cspan address=\"10.1002/ece3.7464\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArmstrong AO, Armstrong AJ, Jaine FRA, Couturier LIE, Fiora K, Uribe-Palomino J, Weeks SJ, Townsend KA, Bennett MB, Richardson AJ (2016) Prey density threshold and tidal influence on reef manta ray foraging at an aggregation site on the Great Barrier Reef. PLoS ONE 11(5):e0153393. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0153393\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0153393\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArmstrong AO, Stevens GMW, Townsend KA, Murray A, Bennett MB, Armstrong AJ, Uribe-Palomino J, Hosegood P, Dudgeon CL, Richardson AJ (2021b) Reef manta rays forage on tidally driven, high density zooplankton patches in Hanifaru Bay. Maldives PeerJ 9:e11992. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7717/peerj.11992\u003c/span\u003e\u003cspan address=\"10.7717/peerj.11992\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArostegui MC (2025) Cranial endothermy in mobulid rays: Evolutionary and ecological implications of a thermogenic brain. J Anim Ecol 94(1):11\u0026ndash;19. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1365-2656.14200\u003c/span\u003e\u003cspan address=\"10.1111/1365-2656.14200\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaa Atoll Biosphere Reserve (2021) UNESCO - Man and the Biosphere (MAB) Programme - Biosphere reserve periodic review. Ministry of Environment Climate Change and Technology, UNESCO, Division of Ecological and Earth Sciences\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeale CS, Runtuboy F, Sianipar AB, Beer AJE, Kadarusman, Erdmann MV, Setyawan E, Green L, Duffy CAJ, Andrzejaczek S, Block BA, Forsberg K, Meekan M, Gleiss AC (2025) Deep diving behaviour in oceanic manta rays and its potential function. Front Mar Sci 12:1630451. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fmars.2025.1630451\u003c/span\u003e\u003cspan address=\"10.3389/fmars.2025.1630451\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeale CS, Stewart JD, Setyawan E, Sianipar AB, Erdmann MV (2019) Population dynamics of oceanic manta rays (\u003cem\u003eMobula birostris\u003c/em\u003e) in the Raja Ampat Archipelago, West Papua, Indonesia, and the impacts of the El Ni\u0026ntilde;o\u0026ndash;Southern Oscillation on their movement ecology. Divers Distrib 25(9):1472\u0026ndash;1487. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/ddi.12962\u003c/span\u003e\u003cspan address=\"10.1111/ddi.12962\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCairns H, Conley D, Stokes C, Scott T, Hosegood P (in press) Tidally Driven Topographic Eddies Exploited by Manta Feeding Aggregations\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarruthers L, Ersek V, Maher D, Sanders C, Tait D, Soares J, Floyd M, Hashim AS, Helber S, Garnett M, East H, Johnson JA, Ponta G, Sippo JZ (2024) Sea-level rise and extreme Indian Ocean Dipole explain mangrove dieback in the Maldives. Sci Rep 14(1):27012. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-024-73776-z\u003c/span\u003e\u003cspan address=\"10.1038/s41598-024-73776-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChenoweth E, Gabriele C, Hill D (2011) Tidal influences on humpback whale habitat selection near headlands. Mar Ecol Prog Ser 423:279\u0026ndash;289. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3354/meps08891\u003c/span\u003e\u003cspan address=\"10.3354/meps08891\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChowdhury MR, Ahmed A, Moosa S, Mohamed S (2025) ENSO and seasonal climate variability in the Maldives: an analysis of early warning opportunities. Theor Appl Climatol 156(11):615. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00704-025-05851-y\u003c/span\u003e\u003cspan address=\"10.1007/s00704-025-05851-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCopernicus Data Space Ecosystem (2026) Copernicus Sentinel data 2026\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCouto A, Queiroz N, Relvas P, Baptista M, Furtado M, Castro J, Nunes M, Morikawa H, Rosa R (2017) Occurrence of basking shark \u003cem\u003eCetorhinus maximus\u003c/em\u003e in southern Portuguese waters: a two-decade survey. Mar Ecol Prog Ser 564:77\u0026ndash;86. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3354/meps12007\u003c/span\u003e\u003cspan address=\"10.3354/meps12007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCouturier L, Newman P, Jaine F, Bennett M, Venables W, Cagua E, Townsend K, Weeks S, Richardson A (2018) Variation in occupancy and habitat use of \u003cem\u003eMobula alfredi\u003c/em\u003e at a major aggregation site. Mar Ecol Prog Ser 599:125\u0026ndash;145. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3354/meps12610\u003c/span\u003e\u003cspan address=\"10.3354/meps12610\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCouturier LIE, Dudgeon CL, Pollock KH, Jaine FRA, Bennett MB, Townsend KA, Weeks SJ, Richardson AJ (2014) Population dynamics of the reef manta ray \u003cem\u003eManta alfredi\u003c/em\u003e in eastern Australia. Coral Reefs 33(2):329\u0026ndash;342. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00338-014-1126-5\u003c/span\u003e\u003cspan address=\"10.1007/s00338-014-1126-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCowburn B, Moritz C, Grimsditch G, Solandt J (2019) Evidence of coral bleaching avoidance, resistance and recovery in the Maldives during the 2016 mass-bleaching event. Mar Ecol Prog Ser 626:53\u0026ndash;67. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3354/meps13044\u003c/span\u003e\u003cspan address=\"10.3354/meps13044\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCroll DA, Dewar H, Dulvy NK, Fernando D, Francis MP, Galv\u0026aacute;n-Maga\u0026ntilde;a F, Hall M, Heinrichs S, Marshall A, Mccauley D, Newton KM, Notarbartolo‐Di‐Sciara G, O\u0026rsquo;Malley M, O\u0026rsquo;Sullivan J, Poortvliet M, Roman M, Stevens G, Tershy BR, White WT (2016) Vulnerabilities and fisheries impacts: the uncertain future of manta and devil rays. Aquat Conserv Mar Freshw Ecosyst 26(3):562\u0026ndash;575. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/aqc.2591\u003c/span\u003e\u003cspan address=\"10.1002/aqc.2591\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCullain N, Tibiri\u0026ccedil;\u0026aacute; Y, Venables SK, Rohner CA, Tittensor DP, Lotze HK (2025) Declines in sightings and changing visitation patterns of reef manta rays at an important aggregation site in Mozambique. Environ Biol Fishes 108(9):1361\u0026ndash;1377. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10641-025-01729-0\u003c/span\u003e\u003cspan address=\"10.1007/s10641-025-01729-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDalpadado P, Roxy MK, Arrigo KR, Van Dijken GL, Chierici M, Ostrowski M, Skern-Mauritzen R, Bakke G, Richardson AJ, Sperfeld E (2024) Rapid climate change alters the environment and biological production of the Indian Ocean. Sci Total Environ 906:167342. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.scitotenv.2023.167342\u003c/span\u003e\u003cspan address=\"10.1016/j.scitotenv.2023.167342\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDewar H, Mous P, Domeier M, Muljadi A, Pet J, Whitty J (2008) Movements and site fidelity of the giant manta ray, \u003cem\u003eManta birostris\u003c/em\u003e, in the Komodo Marine Park, Indonesia. Mar Biol 155(2):121\u0026ndash;133. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00227-008-0988-x\u003c/span\u003e\u003cspan address=\"10.1007/s00227-008-0988-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDulvy NK, Fowler SL, Musick JA, Cavanagh RD, Kyne PM, Harrison LR, Carlson JK, Davidson LN, Fordham SV, Francis MP, Pollock CM, Simpfendorfer CA, Burgess GH, Carpenter KE, Compagno LJ, Ebert DA, Gibson C, Heupel MR, Livingstone SR, Sanciangco JC, Stevens JD, Valenti S, White WT (2014) Extinction risk and conservation of the world\u0026rsquo;s sharks and rays. eLife 3:e00590. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7554/eLife.00590\u003c/span\u003e\u003cspan address=\"10.7554/eLife.00590\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFonseca-Ponce I, Zavala-Jim\u0026eacute;nez A, Aburto-Oropeza O, Maldonado-Gasca A, Galv\u0026aacute;n-Maga\u0026ntilde;a F, Gonz\u0026aacute;lez-Armas R, Stewart J (2022) Physical and environmental drivers of oceanic manta ray \u003cem\u003eMobula birostris\u003c/em\u003e sightings at an aggregation site in Bah\u0026iacute;a de Banderas, Mexico. Mar Ecol Prog Ser 694:133\u0026ndash;148. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3354/meps14106\u003c/span\u003e\u003cspan address=\"10.3354/meps14106\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFortin D, Brooke CF, Lamirande P, Fritz H, McLoughlin PD, Pays O (2020) Quantitative spatial ecology to promote human-wildlife coexistence: A tool for integrated landscape management. Front Sustain Food Syst 4:600363. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fsufs.2020.600363\u003c/span\u003e\u003cspan address=\"10.3389/fsufs.2020.600363\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFroman N, Genain M, Stevens GMW, Pearce GP (2023) Use of underwater contactless ultrasonography to elucidate the internal anatomy and reproductive activity of manta and devil rays (family: Mobulidae). J Fish Biol 103(2):305\u0026ndash;323. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/jfb.15423\u003c/span\u003e\u003cspan address=\"10.1111/jfb.15423\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGraham RT, Witt MJ, Castellanos DW, Remolina F, Maxwell S, Godley BJ, Hawkes LA (2012) Satellite tracking of manta rays highlights challenges to their conservation. PLoS ONE 7(5):e36834. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0036834\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0036834\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuzman HM, Collatos CM, Gomez CG (2022) Movement, behavior, and habitat use of whale sharks (\u003cem\u003eRhincodon typus\u003c/em\u003e) in the tropical eastern Pacific Ocean. Front Mar Sci 9:793248. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fmars.2022.793248\u003c/span\u003e\u003cspan address=\"10.3389/fmars.2022.793248\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarris JL, McGregor PK, Oates Y, Stevens GMW (2020) Gone with the wind: seasonal distribution and habitat use by the reef manta ray (\u003cem\u003eMobula alfredi\u003c/em\u003e) in the Maldives, implications for conservation. Aquat Conserv Mar Freshw Ecosyst 30(8):1649\u0026ndash;1664. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/aqc.3350\u003c/span\u003e\u003cspan address=\"10.1002/aqc.3350\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarris JL, Stevens GMW (2021) Environmental drivers of reef manta ray (\u003cem\u003eMobula alfredi\u003c/em\u003e) visitation patterns to key aggregation habitats in the Maldives. PLoS ONE 16(6):e0252470. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0252470\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0252470\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarvey-Carroll J, Stewart JD, Carroll D, Mohamed B, Shameel I, Zareer IH, Araujo G, Rees R (2021) The impact of injury on apparent survival of whale sharks (\u003cem\u003eRhincodon typus\u003c/em\u003e) in South Ari Atoll Marine Protected Area, Maldives. Sci Rep 11(1):937. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-020-79101-8\u003c/span\u003e\u003cspan address=\"10.1038/s41598-020-79101-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHays GC, Hobson VJ, Metcalfe JD, Righton D, Sims DW (2006) Flexible foraging movements of leatherback turtles across the North Atlantic Ocean. Ecology 87(10):2647\u0026ndash;2656. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1890/0012-9658\u003c/span\u003e\u003cspan address=\"10.1890/0012-9658\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e87%255B2647:FFMOLT%255D2.0.CO;2\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHays GC, Richardson A, Robinson C (2005) Climate change and marine plankton. Trends Ecol Evol 20(6):337\u0026ndash;344. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.tree.2005.03.004\u003c/span\u003e\u003cspan address=\"10.1016/j.tree.2005.03.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHeneghan RF, Everett JD, Blanchard JL, Sykes P, Richardson AJ (2023) Climate-driven zooplankton shifts cause large-scale declines in food quality for fish. Nat Clim Change 13(5):470\u0026ndash;477. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41558-023-01630-7\u003c/span\u003e\u003cspan address=\"10.1038/s41558-023-01630-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHildebrand L, Sullivan F, Orben R, Derville S, Torres L (2022) Trade-offs in prey quantity and quality in gray whale foraging. Mar Ecol Prog Ser 695:189\u0026ndash;201. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3354/meps14115\u003c/span\u003e\u003cspan address=\"10.3354/meps14115\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJabado RW, Morata AZA, Bennett RH, Finucci B, Ellis JR, Fowler S, Grant MI, Barbosa MAP, Sinclair SL (2024) The global status of sharks, rays, and chimaeras. IUCN, Gland, Switzerland: IUCN\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJaine FRA, Couturier LIE, Weeks SJ, Townsend KA, Bennett MB, Fiora K, Richardson AJ (2012) When giants turn up: sighting trends, environmental influences and habitat use of the manta ray (\u003cem\u003eMobula alfredi\u003c/em\u003e) at a coral reef. PLoS ONE 7(10):e46170. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0046170\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0046170\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJoseph PV, Simon A (2005) Weakening trend of the Southwest Monsoon current through peninsular India from 1950 to the present. JSTOR 89(4):687\u0026ndash;694\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKelley D, Richards C (2024) oce: Analysis of Oceanographic Data\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlotz L, Fern\u0026aacute;ndez R, Rasmussen MH (2017) Annual and monthly fluctuations in humpback whale (\u003cem\u003eMegaptera novaeangliae\u003c/em\u003e) presence in Skj\u0026aacute;lfandi Bay, Iceland, during the feeding season (April\u0026ndash;October). J Cetacean Res Manage 16:9\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.47536/jcrm.v16i%60.433\u003c/span\u003e\u003cspan address=\"10.47536/jcrm.v16i%60.433\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKnochel AM, Hussey NE, Kessel ST, Braun CD, Cochran JEM, Hill G, Klaus R, Checkchak T, Elamin El Hassen NM, Younnis M, Berumen ML (2022) Home sweet home: spatiotemporal distribution and site fidelity of the reef manta ray (\u003cem\u003eMobula alfredi\u003c/em\u003e) in Dungonab Bay, Sudan. Mov Ecol 10(1):22. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40462-022-00314-9\u003c/span\u003e\u003cspan address=\"10.1186/s40462-022-00314-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKwiatkowski L, Aumont O, Bopp L (2019) Consistent trophic amplification of marine biomass declines under climate change. Glob Change Biol 25(1):218\u0026ndash;229. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/gcb.14468\u003c/span\u003e\u003cspan address=\"10.1111/gcb.14468\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLaglbauer BJL, D\u0026rsquo;Costa NG, Stewart JD, Palacios MD, Cronin M, Fernando D, Lezama-Ochoa N, Armstrong AO, Jabado RW, Fowler S, Lawson JM, Koubrak O, Murua J, Ko Gyi T, Karnad D, Chopra M, Notarbartolo-di-Sciara G, Rambahiniarison J, Croll D, Rojas S, Fahmi, Harris JL, Binthe Haque A, Murua H, P\u0026eacute;rez-Jim\u0026eacute;nez JC, Humble E, Barrowclift E, Salim MG, De Bruyne G, Seidu I, Zambrano-Vizquel LA, Davies K, Moazzam Khan M, Bucair N, Johnson J, Labyedh G, Takoukam Kamla A, Fuentes K, Carter R, Barros N, Stevens GMW (2026) Global manta and devil ray population declines: Closing policy and management gaps to reduce fisheries mortality. Biol Conserv 313:111589. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.biocon.2025.111589\u003c/span\u003e\u003cspan address=\"10.1016/j.biocon.2025.111589\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLawson JM, Fordham SV, O\u0026rsquo;Malley MP, Davidson LNK, Walls RHL, Heupel MR, Stevens G, Fernando D, Budziak A, Simpfendorfer CA, Ender I, Francis MP, Di Notarbartolo G, Dulvy NK (2017) Sympathy for the devil: a conservation strategy for devil and manta rays. PeerJ 5:e3027. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7717/peerj.3027\u003c/span\u003e\u003cspan address=\"10.7717/peerj.3027\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLazaridis E (2022) lunar: Calculate Lunar Phase \u0026amp; Distance, Seasons and Related Environmental Factors\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi M (2000) Estimating horizontal dispersion of floating particles in wind-driven upper ocean. Spill Sci Technol Bull 6(3\u0026ndash;4):255\u0026ndash;261. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S1353-2561(01)00044-5\u003c/span\u003e\u003cspan address=\"10.1016/S1353-2561(01)00044-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLyard FH, Allain DJ, Cancet M, Carrere L, Picot N (2021) FES2014 global ocean tide atlas. design and performance\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarshall AD, Barreto R, Carlson J, Fernando D, Fordham S, Francis MP, Herman K, Jabado RW, Liu KM, Pacoureau N, Rigby CL, Romanov E, Sherley RB (2022) \u003cem\u003eMobula alfredi\u003c/em\u003e (amended version of 2019 assessment). The IUCN Red List of Threatened Species 2022\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarshall AD, Dudgeon CL, Bennett MB (2011) Size and structure of a photographically identified population of manta rays \u003cem\u003eManta alfredi\u003c/em\u003e in southern Mozambique. Mar Biol 158(5):1111\u0026ndash;1124. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00227-011-1634-6\u003c/span\u003e\u003cspan address=\"10.1007/s00227-011-1634-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcInturf AG, Muhling B, Bizzarro JJ, Fangue NA, Ebert DA, Caillaud D, Dewar H (2022) Spatial distribution, temporal changes, and knowledge gaps in basking shark (\u003cem\u003eCetorhinus maximus\u003c/em\u003e) sightings in the California current ecosystem. Front Mar Sci 9:818670. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fmars.2022.818670\u003c/span\u003e\u003cspan address=\"10.3389/fmars.2022.818670\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeteoblue (2025) Historical weather data for Hanifarurah (10 m wind speed and direction)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiddleton AD, Sawyer H, Merkle JA, Kauffman MJ, Cole EK, Dewey SR, Gude JA, Gustine DD, McWhirter DE, Proffitt KM, White P (2020) Conserving transboundary wildlife migrations: recent insights from the Greater Yellowstone Ecosystem. Front Ecol Environ 18(2):83\u0026ndash;91. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/fee.2145\u003c/span\u003e\u003cspan address=\"10.1002/fee.2145\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMin HS, Noh Y (2004) Influence of the surface heating on Langmuir Circulation. J Phys Oceanogr 34:2630\u0026ndash;2641\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoloney HM, Armstrong AO, Stevens GMW, Dudgeon CL, Harris JL, Townsend KA, Richardson AJ (in press) A new visual index for assessing zooplankton biomass and its utility in assessing prey availability for megaplanktivores\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoloney HM, Garcia Rojas MI, Rothe N, Armstrong AO, Ballard K, Barraud F, Hamdan F, Richardson AJ, Ryad EM, Sawers T, Townsend KA, Stevens GMW (in press) Valuing conservation and natural wealth: The blue economy of manta ray watching in the Maldives\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurray A, Garrud E, Ender I, Lee-Brooks K, Atkins R, Lynam R, Arnold K, Roberts C, Hawkins J, Stevens G (2020) Protecting the million-dollar mantas; creating an evidence-based code of conduct for manta ray tourism interactions. J Ecotourism 19(2):132\u0026ndash;147. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/14724049.2019.1659802\u003c/span\u003e\u003cspan address=\"10.1080/14724049.2019.1659802\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNagelkerke NJD (1991) A More General Definition of the Coefficient of Determination. Biometrika 78(3). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/biomet/78.3.691\u003c/span\u003e\u003cspan address=\"10.1093/biomet/78.3.691\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNeves L (2009) Investigating anthropogenic impacts on the manta rays and whale sharks of Hanifaru, Maldives. MSc Thesis, University of York\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNOAA Physical Sciences Laboratory (2024) Monthly Indian Ocean Dipole Index (DMI) based on HadISST1\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePaig-Tran EWM, Kleinteich T, Summers AP (2013) The filter pads and filtration mechanisms of the devil rays: Variation at macro and microscopic scales. J Morphol 274(9):1026\u0026ndash;1043. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/jmor.20160\u003c/span\u003e\u003cspan address=\"10.1002/jmor.20160\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePalacios MD, Stewart JD, Croll DA, Cronin MR, Trejo-Ram\u0026iacute;rez A, Stevens GMW, Lezama-Ochoa N, Zilliacus KM, Gonz\u0026aacute;lez\u0026ndash;Armas R, Di Notarbartolo G, Galv\u0026aacute;n\u0026ndash;Maga\u0026ntilde;a F (2023) Manta and devil ray aggregations: conservation challenges and developments in the field. Front Mar Sci 10:1148234. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fmars.2023.1148234\u003c/span\u003e\u003cspan address=\"10.3389/fmars.2023.1148234\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePistevos JCA, Nagelkerken I, Rossi T, Olmos M, Connell SD (2015) Ocean acidification and global warming impair shark hunting behaviour and growth. Sci Rep 5(1):16293. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/srep16293\u003c/span\u003e\u003cspan address=\"10.1038/srep16293\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eR Core Team (2024) R: a language and environment for statistical computing\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRichardson AJ (2008) In hot water: zooplankton and climate change. ICES J Mar Sci 65(3):279\u0026ndash;295. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/icesjms/fsn028\u003c/span\u003e\u003cspan address=\"10.1093/icesjms/fsn028\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRohner CA, Pierce S, Marshall A, Weeks S, Bennett M, Richardson A (2013) Trends in sightings and environmental influences on a coastal aggregation of manta rays and whale sharks. Mar Ecol Prog Ser 482:153\u0026ndash;168. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3354/meps10290\u003c/span\u003e\u003cspan address=\"10.3354/meps10290\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRohner CA, Prebble CEM (2021) Whale shark foraging, feeding, and diet. Whale Sharks, 1st edn. CRC, Boca Raton, pp 153\u0026ndash;180\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRohner CA, Venables SK, Knochel AM, Rambahiniarison JM, Marillac V, Cardon C, Scholten N, Pierce SJ, Kiszka JJ (2025) Movements and habitat use of reef manta rays around the Mozambique Channel Island of Mayotte, Southwestern Indian Ocean. Environ Biol Fishes 108(6):937\u0026ndash;955. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10641-025-01695-7\u003c/span\u003e\u003cspan address=\"10.1007/s10641-025-01695-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoxy MK, Modi A, Murtugudde R, Valsala V, Panickal S, Prasanna Kumar S, Ravichandran M, Vichi M, L\u0026eacute;vy M (2016) A reduction in marine primary productivity driven by rapid warming over the tropical Indian Ocean. Geophys Res Lett 43(2):826\u0026ndash;833. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/2015GL066979\u003c/span\u003e\u003cspan address=\"10.1002/2015GL066979\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoxy MK, Ritika K, Terray P, Murtugudde R, Ashok K, Goswami BN (2015) Drying of Indian subcontinent by rapid Indian Ocean warming and a weakening land-sea thermal gradient. Nat Commun 6(1):7423. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/ncomms8423\u003c/span\u003e\u003cspan address=\"10.1038/ncomms8423\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRubin RD, Kumli KR, Klimley AP, Stewart JD, Ketchum JT, Hoyos-Padilla EM, Galv\u0026aacute;n-Maga\u0026ntilde;a F, Zavala-Jim\u0026eacute;nez AA, Fonseca-Ponce IA, Saunders M, Dominguez-Sanchez PS, Ahuja P, Nevels CR, Gonz\u0026aacute;lez PAP, Corgos A, Diemer SJ (2025) Insular and mainland interconnectivity in the movements of oceanic manta rays (\u003cem\u003eMobula birostris\u003c/em\u003e) off Mexico in the Eastern Tropical Pacific. Environ Biol Fishes 108(4):555\u0026ndash;568. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10641-024-01622-2\u003c/span\u003e\u003cspan address=\"10.1007/s10641-024-01622-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSaji NH, Goswami BN, Vinayachandran PN, Yamagata T (1999) A dipole mode in the tropical Indian Ocean. Nature 401(6751):360\u0026ndash;363. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/43854\u003c/span\u003e\u003cspan address=\"10.1038/43854\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSasamal SK (2006) Island mass effect around the Maldives during the winter months of 2003 and 2004. Int J Remote Sens 27(22):5087\u0026ndash;5093. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/01431160500177562\u003c/span\u003e\u003cspan address=\"10.1080/01431160500177562\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSetyawan E (2016) Correlating environmental parameters with the movement patterns and site fidelity of reef manta ray (\u003cem\u003eManta alfredi\u003c/em\u003e, Krefft 1868) using acoustic telemetry and remote sensing in Raja Ampat, Indonesia. Thesis, University of Tasmania\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSims DW, Southall E, Richardson A, Reid P, Metcalfe J (2003) Seasonal movements and behaviour of basking sharks from archival tagging: no evidence of winter hibernation. Mar Ecol Prog Ser 248:187\u0026ndash;196. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3354/meps248187\u003c/span\u003e\u003cspan address=\"10.3354/meps248187\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSims DW, Southall EJ, Tarling GA, Metcalfe JD (2005) Habitat-specific normal and reverse diel vertical migration in the plankton‐feeding basking shark. J Anim Ecol 74(4):755\u0026ndash;761. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1365-2656.2005.00971.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1365-2656.2005.00971.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSims DW, Witt MJ, Richardson AJ, Southall EJ, Julian D, Metcalfe (2006) Encounter success of free-ranging marine predator movements across a dynamic prey landscape. Proc R Soc B. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1098/rspb.2005.3444\u003c/span\u003e\u003cspan address=\"10.1098/rspb.2005.3444\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSingh AD (2018) Rapid switch in monsoon-wind induced surface hydrographic conditions of the eastern Arabian Sea during the last deglaciation. Quat Int 479:3\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.quaint.2018.03.027\u003c/span\u003e\u003cspan address=\"10.1016/j.quaint.2018.03.027\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSkomal GB, Zeeman SI, Chisholm JH, Summers EL, Walsh HJ, McMahon KW, Thorrold SR (2009) Transequatorial migrations by basking sharks in the Western Atlantic Ocean. Curr Biol 19(12):1019\u0026ndash;1022. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cub.2009.04.019\u003c/span\u003e\u003cspan address=\"10.1016/j.cub.2009.04.019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSleeman JC, Meekan MG, Fitzpatrick BJ, Steinberg CR, Ancel R, Bradshaw CJA (2010) Oceanographic and atmospheric phenomena influence the abundance of whale sharks at Ningaloo Reef, Western Australia. J Exp Mar Biol Ecol 382(2):77\u0026ndash;81. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jembe.2009.10.015\u003c/span\u003e\u003cspan address=\"10.1016/j.jembe.2009.10.015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith JA (2001) Observations and theories of Langmuir Circulation: a story of mixing. In: Lumley JL (ed) Fluid Mechanics and the Environment: Dynamical Approaches. Springer Berlin Heidelberg, Berlin, Heidelberg, pp 295\u0026ndash;314\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoulsbury CD, White PCL (2016) Human\u0026ndash;wildlife interactions in urban areas: a review of conflicts, benefits and opportunities. Wildl Res 42(7):541\u0026ndash;553. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1071/WR14229\u003c/span\u003e\u003cspan address=\"10.1071/WR14229\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSrivastava A, Martin GM, Pradhan M, Rao SA, Ineson S (2026) The multi-year negative Indian Ocean Dipole of 2021\u0026ndash;2022. Weather Clim Dyn 7(1):1\u0026ndash;15. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5194/wcd-7-1-2026\u003c/span\u003e\u003cspan address=\"10.5194/wcd-7-1-2026\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStephens DW, Krebs JR (1986) Foraging theory. Princeton University Press\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStevens GMW (2016) Conservation and population ecology of manta rays in the Maldives. PhD Thesis, University of York\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStevens GMW, Froman N (2018) The Maldives Archipelago. World Seas: an environmental evaluation. Elsevier, pp 211\u0026ndash;236\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStewart JD, Jaine FRA, Armstrong AJ, Armstrong AO, Bennett MB, Burgess KB, Couturier LIE, Croll DA, Cronin MR, Deakos MH, Dudgeon CL, Fernando D, Froman N, Germanov ES, Hall MA, Hinojosa-Alvarez S, Hosegood JE, Kashiwagi T, Laglbauer BJL, Lezama-Ochoa N, Marshall AD, McGregor F, Di Sciara N, Palacios G, Peel MD, Richardson LR, Rubin AJ, Townsend RD, Venables KA, Stevens SK GMW (2018) Research priorities to support effective manta and devil ray conservation. Front Mar Sci 5:314. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fmars.2018.00314\u003c/span\u003e\u003cspan address=\"10.3389/fmars.2018.00314\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStrike EM, Harris JL, Ballard KL, Hawkins JP, Crockett J, Stevens GMW (2022) Sublethal injuries and physical abnormalities in Maldives manta rays, \u003cem\u003eMobula alfredi\u003c/em\u003e and \u003cem\u003eMobula birostris\u003c/em\u003e. Front Mar Sci 9:773897. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fmars.2022.773897\u003c/span\u003e\u003cspan address=\"10.3389/fmars.2022.773897\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThe Manta Trust (2024) Maldives Manta Conservation Programme database\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThorpe SA (2004) Langmuir circulation. Annu Rev Fluid Mech 36(1):55\u0026ndash;79. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1146/annur%20ev.fluid.36.052203.071431\u003c/span\u003e\u003cspan address=\"10.1146/annur%20ev.fluid.36.052203.071431\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThushara V, Vinayachandran PN (2020) Unprecedented surface Chlorophyll blooms in the southeastern Arabian Sea during an extreme Negative Indian Ocean Dipole. Geophys Res Lett 47(13). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2019GL085026\u003c/span\u003e\u003cspan address=\"10.1029/2019GL085026\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. e2019GL085026\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUNEP-WCMC (2020) Ocean+ Habitats [On-line]\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVenables SK, Rohner CA, Flam AL, Pierce SJ, Marshall AD (2024) Persistent declines in sightings of manta and devil rays (Mobulidae) at a global hotspot in southern Mozambique. Environ Biol Fishes. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10641-024-01576-5\u003c/span\u003e\u003cspan address=\"10.1007/s10641-024-01576-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVenables WN, Ripley BD (2022) Modern applied statistics with S (4th ed.)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWatson M, Stamation K, Charlton C, Bannister J (2001) Calving intervals, long range movements and site fidelity of southern right whales (\u003cem\u003eEubalaena australis\u003c/em\u003e) in south\u0026shy;eastern Australia. J Cetacean Res Manag 22\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeeks S, Magno-Canto M, Jaine F, Brodie J, Richardson A (2015) Unique sequence of events triggers manta ray feeding frenzy in the southern Great Barrier Reef, Australia. Remote Sens 7(3):3138\u0026ndash;3152. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs70303138\u003c/span\u003e\u003cspan address=\"10.3390/rs70303138\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWitt M, Hardy T, Johnson L, McClellan C, Pikesley S, Ranger S, Richardson P, Solandt J, Speedie C, Williams R, Godley B (2012) Basking sharks in the northeast Atlantic: spatio-temporal trends from sightings in UK waters. Mar Ecol Prog Ser 459:121\u0026ndash;134. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3354/meps09737\u003c/span\u003e\u003cspan address=\"10.3354/meps09737\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWolanski E, Asaeda T, Tanaka A, Deleersnijder E (1996) Three-dimensional island wakes in the field, laboratory experiments and numerical models. Cont Shelf Res 16(11):1437\u0026ndash;1452. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/0278-4343(95)00087-9\u003c/span\u003e\u003cspan address=\"10.1016/0278-4343(95)00087-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWorsfold M, Good S, Atkinson C, Embury O (2024) Presenting a long-term, reprocessed dataset of global sea surface temperature produced using the OSTIA system\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZerbini A, Andriolo A, Heide-J\u0026oslash;rgensen M, Pizzorno J, Maia Y, VanBlaricom G, DeMaster D, Sim\u0026otilde;es-Lopes P, Moreira S, Bethlem C (2006) Satellite-monitored movements of humpback whales \u003cem\u003eMegaptera novaeangliae\u003c/em\u003e in the southwest Atlantic Ocean. Mar Ecol Prog Ser 313:295\u0026ndash;304. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3354/meps313295\u003c/span\u003e\u003cspan address=\"10.3354/meps313295\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"marine-biology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mabi","sideBox":"Learn more about [Marine Biology](https://www.springer.com/journal/227)","snPcode":"227","submissionUrl":"https://submission.nature.com/new-submission/227/3","title":"Marine Biology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Life Below Water, Climate Action, Elasmobranch, Planktivores, Seasonality, Environmental variability, Conservation, Mobula alfredi, Endangered species, Marine protected area","lastPublishedDoi":"10.21203/rs.3.rs-9455345/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9455345/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eUnderstanding the drivers of megafauna aggregations is critical for managing threatened species in a changing climate. We analysed an 18-year dataset (2007\u0026ndash;2024) of reef manta ray (\u003cem\u003eMobula alfredi\u003c/em\u003e) abundance and zooplankton density at Hanifaru Bay, Maldives (5.1733\u0026deg;N, 73.145\u0026deg;E) \u0026mdash;the world\u0026rsquo;s largest known manta ray aggregation. Generalised Linear Models were used to evaluate sightings against local environmental variables and broad-scale climatic indices. Results reveal that interannual abundance was modulated by the Indian Ocean Dipole, with sightings and prey density reaching decadal highs during the record Negative Indian Ocean Dipole and La Ni\u0026ntilde;a of 2021\u0026ndash;2022. Locally, zooplankton density was the strongest predictor of \u003cem\u003eM. alfredi\u003c/em\u003e abundance, explaining\u0026thinsp;\u0026gt;\u0026thinsp;60% of the observed variation. Sightings were 25x higher during peak prey availability than during prey absence. Aggregations were timed with spring tides (peaking 2.7 hours post-high tide), new and full moon phases, and 25\u0026ndash;35 km/h north-northwest winds. We also identified a critical thermal threshold at 28.6\u0026deg;C, beyond which sightings declined by 15.7% per degree of sea surface temperature warming. This sequence confirms that the local bathymetric retention mechanisms are dependent on specific regional oceanographic conditions. Our findings suggest that while Hanifaru Bay consistently concentrates zooplankton biomass, its functional utility as a foraging hotspot is vulnerable to climate-driven shifts in the Indian Ocean Dipole and rising sea surface temperatures, which may decouple the predator from its primary nutrient supply.\u003c/p\u003e","manuscriptTitle":"Prey availability and fine-scale oceanography drive reef manta ray aggregations at the world’s largest hotspot","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-15 08:49:38","doi":"10.21203/rs.3.rs-9455345/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"","date":"2026-05-12T12:36:01+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-05-06T09:11:26+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-20T03:54:44+00:00","index":"","fulltext":""},{"type":"submitted","content":"Marine Biology","date":"2026-04-18T03:56:03+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"marine-biology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mabi","sideBox":"Learn more about [Marine Biology](https://www.springer.com/journal/227)","snPcode":"227","submissionUrl":"https://submission.nature.com/new-submission/227/3","title":"Marine Biology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"792806a1-ccea-4c08-94b5-976abbd631b9","owner":[],"postedDate":"May 15th, 2026","published":true,"recentEditorialEvents":[{"type":"reviewerAgreed","content":"","date":"2026-05-12T12:36:01+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-05-06T09:11:26+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-15T08:49:39+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-15 08:49:38","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9455345","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9455345","identity":"rs-9455345","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}