The Age of Aquariums: Managing Elasmobranch Behaviour in a Mixed-Species Habitat

The Age of Aquariums: Managing Elasmobranch Behaviour in a Mixed-Species Habitat | 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 Article The Age of Aquariums: Managing Elasmobranch Behaviour in a Mixed-Species Habitat Claudia Tay, Todd J. McWhorter, Clement Yu Wei Koh, Kai Le Leong, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9208737/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 6 You are reading this latest preprint version Abstract Managing elasmobranchs in mixed-species aquarium exhibits presents unique welfare challenges. With many species facing conservation threats, optimising behaviour in managed care is critical. This study investigated the effects of modified feeding events and positive reinforcement training on sharks, rays, and bony fish in Singapore Oceanarium's Open Ocean Habitat. Experiment 1 employed a reversal design, transitioning from single daily feeding to three feedings a day while maintaining the total food quantity. Sharks showed a significant reduction in undesired responses; however, rays showed an increase in such behaviours. Bony fish showed minimal effects due to the manipulations. Experiment 2 targeted undesired responses in the rays through a target training intervention. Javanese cownose rays ( Rhinoptera javanica ) received positive reinforcement training using stretcher conditioning, while spotted eagle rays ( Aetobatus narinari) received fixed location feeding. Training significantly reduced the cownose ray’s undesired responses. Eagle rays showed improvements, though not statistically significant. Results demonstrated that live-feeding modifications effectively reduced undesired responses in sharks but required additional reward-focused interventions for rays. Positive reinforcement training successfully addressed welfare challenges, providing practical management tools for aquarium professionals working with diverse elasmobranch collections. Biological sciences/Ecology Earth and environmental sciences/Ecology Earth and environmental sciences/Ocean sciences Biological sciences/Zoology feeding management positive reinforcement training aquarium welfare elasmobranchs Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Elasmobranchs have become flagship species in modern aquariums, captivating millions of visitors annually while serving critical roles in conservation education and species preservation. However, the management of captive elasmobranchs in mixed-species exhibits presents unique welfare challenges. These challenges necessitate changes to husbandry that address species-specific needs while maintaining a positive visitor experience. Currently, 102 chondrichthyan species are maintained across European facilities alone (Janse et al., 2017 ), many of which face escalating threats in their native ecosystems. As public aquariums increasingly position themselves as conservation institutions, the imperative to optimise animal welfare while maintaining educational value has never been more pressing. Despite growing recognition of the importance of welfare assessment in aquarium settings, significant knowledge gaps persist regarding fish welfare relative to that of terrestrial vertebrates. Huntingford et al. ( 2006 ) highlighted that welfare assessments remain more challenging for aquatic animals than for mammals, primarily due to methodological difficulties, the diversity of fish taxa, and historical research biases. This disparity is particularly concerning given that fish possess complex behavioural repertoires, sophisticated sensory systems, and physiological needs that differ substantially from terrestrial species. The lack of standardised welfare indicators for elasmobranchs specifically has left aquarium professionals with limited empirical guidance for husbandry decisions that directly impact animal wellbeing. A review of animal welfare publications in zoos and aquariums found that only 2 of 310 (0.6%) studies focused on fish, compared to 232 on mammals (Binding et al., 2020 ). Animal-visitor interactions (AVIs) represent a critical yet understudied aspect of aquarium animal management (Fernandez & Sherwen, 2024 ; Lin et al., 2025 ). Nonetheless, research on AVIs in aquariums remains limited. D’Cruze et al. ( 2019 ) conducted a global review of AVIs in zoos and aquariums, revealing that Chondrichthyes and Actinopterygii accounted for only 5% and 1.8% of published interaction studies, despite their large numbers in aquatic collections. Sherwen and Hemsworth ( 2019 ) noted that there were no studies published on the visitor effect on fish then and it was only in 2020 that the first paper on fish was published (Williams et al., 2023 ). This lack of research on aquatic AVIs, combined with the prevalence of visitor-related stressors in high-traffic exhibits, underscores the urgent need for empirical investigations into how visitor presence and feeding interactions affect the behaviour and welfare of elasmobranchs. Feeding management represents a fundamental factor of captive elasmobranch welfare, yet current practices often fail to reflect species-specific ecological requirements. In wild populations, elasmobranchs exhibit remarkable diversity in foraging strategies, from benthic ambush predators to active pelagic hunters, with many species displaying opportunistic or seasonally variable feeding patterns (Munroe et al., 2022 ). Under human care, standardised feeding routines typically constrain this natural flexibility, often imposing across species with different metabolic demands and activity patterns and behaviour (Morimura & Ueno, 1999 ). Even at the individual level, there may be high variability in feeding patterns (López-Olmeda et al., 2012 ). Fish research demonstrates that feeding events can intensify competition, elevate stress-related cortisol levels, and increase the risk of aggressive encounters and physical injury (Mushtaq, 2024 ; Ward et al., 2006 ). These findings suggest that strategic modifications to feeding schedules and methodologies may be effective non-invasive interventions to improve welfare outcomes across diverse elasmobranch species in aquarium settings. In addition to feeding strategies, the use of operant conditioning and training programmes has become an integral component of animal care in aquariums and zoological facilities. Animal training, particularly using positive reinforcement techniques, can promote cooperative participation in husbandry, reduce stress during medical procedures and provide a form of cognitive enrichment that enhances welfare (Boundey et al., 2025 ; Melfi, 2013 ). Positive reinforcement training has also been shown to reduce stress in animals (Bassett et al., 2003 ; Whitham & Wielebnowski, 2013 ). Brown and Schluessel ( 2023 ) demonstrated that sharks and rays are capable of discrimination learning, habituation, and long-term memory, providing a cognitive foundation for training. The application of training to elasmobranchs has shown considerable promise, although fewer studies have been conducted than for other taxonomic groups, such as mammals and birds. Corwin ( 2012 ) documented the successful application of positive reinforcement training with eels to reduce aggression during feeding. In another study, implementation of visual target training at two designated locations within a multispecies exhibit resulted in an 80% reduction in behavioural problems associated with feeding events (Laterveer, 2006 ). Aquarists introduced a visual target as a conditioned stimulus for food delivery, significantly reducing aggressive behaviour and facilitating safer, more controlled feeding events in spotted eagle rays ( Aetobatus narinari ) (Muraco & Stamper, 2003 ). However, a lack of peer-reviewed publications persists, with few studies examining the effects of training on fish behaviour compared with other taxa (Brando & Norman, 2023 ). Given the complex interactions among feeding events, AVIs, and behavioural modification, this study aimed to evaluate two management interventions for improving fish welfare at Singapore Oceanarium, Resorts World Sentosa, Singapore. The study was structured as two complementary experiments. The first experiment examined the efficacy of modified feeding events in mitigating undesired responses across shark, ray, and bony fish species, testing the hypothesis that strategic alterations in feeding presentation and frequency would reduce anticipatory aggression and improve overall behavioural patterns. The second experiment focused specifically on target training interventions for Javanese cownose rays ( Rhinoptera javanica ) and fixed location feeding for spotted eagle rays ( Aetobatus narinari ). This experiment tested whether positive reinforcement training could reshape specific behavioural patterns while maintaining appropriate feeding responses. Together, these investigations provide a comprehensive assessment of two critical management tools, feeding event and training modifications, for improving elasmobranch behaviour in managed care environments. Materials and Methods General Materials and Methods The study was conducted at Open Ocean Habitat (OOH) (Appendix 1), located at Singapore Oceanarium within Resorts World Sentosa, Singapore. OOH is a habitat that contains 18 million litres of water and is home to 80 species of animals. The fish were divided into three main taxonomic groups: “rays,” representing all batoids from Class Chondrichthyes ; “sharks,” representing all Selachii; and “bony fish”, representing all Class Osteichthyes . Some examples of rays in OOH included the Javanese cownose ray ( Rhinoptera javanica ), spotted eagle ray ( Aetobatus narinari ), leopard whipray ( Himantura leoparda ), and bowmouth guitarfish ( Rhina ancylostoma ). Some examples of sharks in OOH were scalloped hammerhead ( Sphyrna lewini ), tawny nurse shark ( Nebrius ferrugineus ), blacktip reef shark ( Carcharhinus melanopterus ), and Indo-Pacific leopard shark ( Stegostoma tigrinum ). Some examples of bony fish in OOH were giant grouper ( Epinephelus lanceolatus ), golden pompano ( Trachinotus blochii) , and yellowtail scad ( Atule mate ). The animals in OOH were fed once a day, five times a week, at 1600 hours (Tuesdays, Wednesdays, Thursdays, Saturdays, Sundays), which was when observations were conducted for both experiments. The animals were only fed on Mondays via surface scatter feeding, and not at all on Fridays. All behavioural observations were systematically recorded utilising ZooMonitor software (version 1.7.140; Lincoln Park Zoo, Chicago, IL, USA). Two distinct classes of responses were identified among all animals observed: Desired and Undesired. Desired responses were defined as positive or neutral interactions, with the fish taking the food from divers at the correct position and without rushing. Conversely, undesired responses included any unwanted contact with divers or their equipment, such as nibbling, pushing, blanketing, or grabbing. The responses were mutually exclusive. Data collection was terminated upon completion of each 20-minute dive feeding session. Experiment 1 In the first experiment, feeding events were modified in two ways: the timing of delivery (schedule) and what occurred. The schedule was increased from one to three times a day, with scatter feeding during the first two feedings and dive target feeding during the last. Dive scatter feeding was discontinued. Materials and Methods The study was conducted from 1 May 2024 to 14 July 2024. A total of 40 sessions were recorded. There were some days when observations could not be conducted because a lightning warning prevented divers from entering the water. During the baseline phase, feed distribution occurred daily at 1600h using three distinct methodologies: Surface scatter feeding - Feed comprising a variety of fish, crustaceans and molluscs was scattered outwards into the habitat at two feeding points via a suspension bridge hanging above the habitat. Dive scatter feeding - Three divers (two feeders, one safety) descended into the habitat with feeding tubs containing a variety of fish, crustaceans, and molluscs, and the feed was scattered along the habitat floor. Dive target feeding - Six divers (four feeders, two safety) descended into the habitat, each feeder carrying a specific amount of large and small feeds separated by size. Feeders were divided into two groups: one fed all the rays, while the other fed all the other fish in the habitat. The feed composition consisted of whole fish and fish segments. Each dive feeding session, which incorporated both dive scatter and dive target feeding techniques, lasted approximately 20 minutes. The observer was positioned at a standardised location in the gallery to count visitors to OOH habitat. They then relocated to a predetermined position directly facing OOH to measure ambient crowd noise five minutes before the start of dive feeding at 1600h. Crowd noise was quantified as sound pressure levels (reference value: 20 µPa) in decibels (dBA) using A-weighting and a frequency response range of 31.5-8,000Hz, measured with a Mengshen Sound Level Meter. The feeding protocol involved six divers, divided into two equal groups, stationed in designated areas within OOH. Each group consisted of two divers conducting target feeding and one safety diver performing scatter feeding simultaneously. Throughout the dive feeding session, the observer alternated between two fixed observation points in front of OOH at one-minute intervals, continuing this pattern until the conclusion of the feeding session. In the multiple feeding event condition, all experimental parameters were maintained at baseline levels, including the total quantity of feed provided and the frequency of dive-feeding days. However, the modification was implemented, halving the food allocated for surface scatter feeding and dive scatter feeding and providing it at 0900h and 1300h, respectively (Fig. 1 ). Consequently, at 1600h, only dive target feeding was conducted during the scheduled dive feeding session. The animals were categorised into three taxonomic groups (sharks, rays, or bony fish), and the frequency of both desired and undesired responses was counted. Upon completion of the 20-minute feeding session, data collection was terminated, and sound pressure levels and crowd size were immediately reassessed at the exact locations as before the dive feed. The experimental protocol consisted of 10 consecutive baseline sessions, followed by 10 sessions under the modified feeding schedule intervention. Subsequently, a return to baseline conditions was implemented, and ten additional sessions were recorded. Then, the experimental feeding schedule intervention was reintroduced for a final sequence of ten sessions. This comprehensive ABAB experimental design (Fig. 2 ) allowed a comparative analysis between the intervention and control conditions, whilst controlling for potential temporal effects and enhancing the reliability of observed behavioural responses to the modified feeding schedule. Interobserver Agreement (IOA) To assess interobserver reliability, a secondary observer independently collected data during 25% of all observation sessions. Both observers conducted simultaneous in-person observations from an identical vantage point. Interobserver agreement (IOA) was calculated using the total agreement method (Poling et al., 1995 ), with agreement scores derived for each behavioural category. There was an overall mean IOA of 94.15% across all behavioural classes (Bony Fish, Desired Response: 100%; Bony Fish, Undesired Response: 100%; Ray, Desired Response: 91.84%; Ray, Undesired Response: 96.67%; Shark, Desired Response: 88.89%; Shark, Undesired Response: 87.50%). Statistical Analysis Statistical analyses were performed using RStudio (version 2024.12.1 + 563). A generalised linear mixed-effects model (GLMM) with a negative binomial error distribution was used for the responses count dataset. Treatment was coded with two levels (Control, Modified Feed) and entered as a fixed effect. Session ID was included as a random intercept to account for repeated sampling and session-level heterogeneity. For the GLMM, models were fit using the package ‘glmmTMB’. Overdispersion and residual diagnostics were assessed using the ‘DHARMa’ package, which simulated residuals (QQ plots, residual vs. fitted plots, and dispersion tests). The results did not indicate significant deviations from model assumptions for the fitted models. AIC/BIC and dispersion values are consistent with an adequate fit. The undesired contact for bony fish exhibited complete zeros across treatments, and the treatment effect was not interpreted for this outcome. Statistical significance was set at α = 0.05. Regarding the crowd size and sound pressure level dataset, normal distribution was evaluated using the Shapiro-Wilk test. Levene's test for homogeneity of variances was applied to assess equality of variances; all tests failed to reject the null hypothesis, confirming that the assumption of equal variances was met. For datasets violating normality assumptions, the Mann-Whitney U test was used, while independent-samples t-tests were used for datasets that satisfied the normality assumptions. Statistical significance was established at α = 0.05 across all analyses. Results Sharks displayed significantly reduced undesired contact response under the modified feed treatment compared to the control. The average counts of undesired contact were 18.60 under control conditions and 10.25 under modified feed conditions (β = -0.596, SE = 0.157, p < 0.001; Fig. 3 a). In contrast, desired contact response in sharks did not differ significantly between treatments, with mean counts of 10.45 in the control group and 11.73 for the modified feed (β = 0.116, SE = 0.114, p = 0.312). For the rays, the LMM indicated a significant increase (64.55 to 98.8) in undesired response after the feed conditions were modified (t = 4.59, df = 58.08, p < 0.0001; Fig. 3 b). For desired contact in rays, there was an increase (50.5 to 61.5), although not statistically significant (t = 1.47, df = 58.08, p = 0.15). Bony fish demonstrated minimal behavioural interactions overall. Desired responses were observed at low frequencies in both conditions, with a slight reduction from baseline (0.40) to multiple feeding conditions (0.20), although this difference did not reach statistical significance (β = -0.693, SE = 0.694, p = 0.318; Fig. 3 c). Notably, no undesired responses were documented in either feeding condition (0.00 in both conditions). Pre-dive feed crowd sizes remained consistent between baseline (206.45) and experimental condition (207.15), with no significant difference observed (t = -0.044604, df = 38, p = 0.965; Fig. 4 ). However, post-dive feed crowd size showed a non-significant increase (U = 139, p = 0.102) in the experimental condition (197.35) compared to the baseline (175.45; Fig. 4 ). Pre-dive feed sound levels were comparable between baseline (70.58) and experimental condition (70.76), with minimal variation observed (t = -0.19357, df = 38, p = 0.848; Fig. 5 ). Similarly, post-dive feed sound levels showed negligible differences between baseline (70.53) and experimental condition (71.31) (U = 158, p = 0.262). The consistency in sound pressure levels across both time points and experimental conditions suggests that the modified feeding schedule did not substantially alter the acoustic environment within OOH. Experiment 1 resulted in a significant decrease in undesired responses for the sharks. However, the experiment also resulted in a significant increase in both desired and undesired responses among the rays. Observers noted that this response was primarily exhibited by two species: the Javanese cownose ray ( Rhinoptera javanica ) and the spotted eagle ray ( Aetobatus narinari ). This increase may be attributed to reduced food availability during the targeted feeding sessions compared to baseline conditions. Since these species require time to process their food, the reduced feeding opportunities may have left them less satiated, potentially triggering undesired responses. This led to the next experiment, which focused on interventions for the two ray species: training for cownose rays and fixed location feeding for spotted eagle rays. Experiment 2 In the second experiment, the feeding protocols remained unchanged for the other fish species; however, the Javanese cownose rays received a reward-based target-training procedure, while the spotted eagle rays received fixed-location feeding. Observations were limited to the rays. Materials and Methods This study was conducted from 22 October 2024 to 23 January 2025, encompassing 40 observational sessions. Certain scheduled observations were omitted due to lightning warnings, which prevented divers from entering the water for safety reasons. The experimental subjects were exclusively rays from the initial experiment, categorised into three groups: 42 individual Javanese cownose rays, 12 spotted eagle rays, and a combined group of all other ray species. The baseline protocol for this second experiment replicated the feeding intervention established in the previous experiment, consisting of two surface scatter feedings at 0930h and 1300h, and dive target feeding at 1600h. Baseline data collection comprised 20 sessions, concluding on 15 November 2024. This phase focused solely on behavioural observations; crowd size and sound pressure level measurements were not included in the experimental design. Following the completion of these initial sessions, a reward-based target training programme involving contact with (“targeting”) a surface water stretcher (Appendix 2) was initiated for the Javanese cownose rays, whilst the eagle rays commenced surface feeding at a designated location. Rewards consisted of fish, constituting their standard dietary regimen. Training sessions were conducted at 1530h for 20 minutes on each dive feeding day (Tuesdays, Wednesdays, Thursdays, Saturdays and Sundays). Figure 6 details the baseline and experimental conditions for both rays. Only the rays were observed for this experiment, and they were categorised into three groups: “Cownose ray” for the Javanese cownose ray, “Eagle ray” for the spotted eagle ray, and “Other ray” for all other ray species. The subsequent phase of data collection was contingent upon the Javanese cownose rays demonstrating proficiency in target training to the stretcher, specifically requiring more than 42 instances of complete stretcher transit (entry to exit) across three consecutive sessions. This criterion was met on 15 December 2024, allowing for the commencement of the final data collection phase on 17 December 2024. An additional 20 observational sessions were documented during this phase, with training protocols continuing concurrently on all dive feeding days. Interobserver Agreement To assess interobserver reliability, two additional observers independently collected data during 18% and 10% of all observation sessions. All three observers conducted simultaneous in-person observations from an identical vantage point. IOA was calculated using the total agreement method (Poling et al., 1995 ), with agreement scores derived for each behavioural category. Between the primary and first additional observer, there was an overall mean IOA of 91.98% across all behavioural classes (Cownose Ray, Desired Response: 92.31%; Cownose Ray, Undesired Response: 91.3%; Spotted Eagle Ray, Desired Response: 92.31%; Spotted Eagle Ray, Undesired Response: 93.33%; Other Ray, Desired Response: 88.89%; Other Ray, Undesired Response: 93.75%). Between the primary and second additional observer, there was an overall mean IOA of 92.21% across all behavioural classes (Cownose Ray, Desired Response: 92.31%; Cownose Ray, Undesired Response: 92%; Spotted Eagle Ray, Desired Response: 92.86%; Spotted Eagle Ray, Undesired Response: 88.24%; Other Ray, Desired Response: 93.75%; Other Ray, Undesired Response: 94.12%). Statistical Analysis Generalised linear mixed models (GLMMs) with a Poisson distribution and log link function were fitted for each species-response combination to account for repeated measures across sessions and the count nature of the behavioural data. The model structure was Response Count ~ Treatment + (1 | SessionID) where Treatment was a fixed effect (Pre vs. Post) and SessionID was included as a random intercept to account for session-level variation. Models were fitted using the ‘glmmTMB’ package in R. Model fit was assessed using Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). Statistical significance was determined at α = 0.05. Results Desired contact responses in cownose rays did not change significantly (15.75 to 13.54) following the stretcher training intervention (β = -0.152, SE = 0.114, z = -1.32, p = 0.185; Fig. 7 a). In contrast, undesired contact responses decreased significantly post-training, from 42.25 to 23.1 (β = -0.665, SE = 0.103, z = -6.47, p < 0.001; Fig. 7 a). This represents a substantial reduction in undesired contact responses following the training intervention. For spotted eagle rays, desired contact responses showed no significant change post-training, 14.35 to 14.55 (β = 0.018, SE = 0.102, z = 0.18, p = 0.857; Fig. 7 b). Undesired contact responses in spotted eagle rays showed a marginally non-significant trend toward reduction after the training, 27.75 to 22, (β = -0.214, SE = 0.117, z = -1.83, p = 0.067; Fig. 7 b). Other ray species showed no significant changes in either desired or undesired contact responses following training. Desired contact responses remained stable at 23.55 to 24.55 (β = 0.032, SE = 0.107, z = 0.30, p = 0.766; Fig. 7 c). Similarly, undesired contact response in other ray species showed no significant change at 23.8 to 20.6 (β = -0.165, SE = 0.157, z = -1.06, p = 0.291; Fig. 7 c). The training intervention resulted in a significant reduction in undesired contact responses, specifically in cownose rays, while desired contact responses remained unchanged across all species. Spotted eagle rays showed a trend toward fewer undesired contacts, which approached but did not reach statistical significance. Other ray species showed no significant behavioural changes in response to the training intervention. Stretcher target training was effective in reducing undesired response in Javanese cownose rays. The fixed-location feeding training for spotted eagle rays resulted in a reduction in undesired responses, although this was not significant. Discussion This study offers insights into modifying feeding events (Exp 1) and implementing behaviour modification procedures (Exp 2) in a large-scale, mixed-species oceanarium habitat. The differential responses observed among sharks, rays, and bony fish in the first experiment highlighted the need for tailored management approaches that account for the unique behavioural and physiological characteristics of ray species. Specifics of each experiment are detailed below. Experiment 1 The significant reduction in undesired responses among sharks following the implementation of multiple daily feedings aligns with documented environmental sensitivity in elasmobranch species. For instance, Lawrence et al. ( 2021 ) found that Port Jackson ( Heterodontus portusjackson) sharks significantly reduced abnormal behaviours following environmental design change, indicating that sharks are responsive to husbandry modifications. This finding is also consistent with research on feeding behaviour in aquaculture settings, where Georgopoulou et al. ( 2024 ) demonstrated that a single daily feeding in European seabass ( Dicentrarchus labrax ) produced higher activity levels and more symmetrical behaviour around feeding times than multiple feedings, which may be stressful for the fish. The authors noted that reduced feeding led to prolonged increased activity before and after feeding, directly paralleling our observations of reduced problematic behaviours when food was distributed across multiple sessions rather than concentrated in a single feeding event. Undesired responses during feeding represent a significant challenge in aquarium management, particularly in mixed-species exhibits where competition for resources can lead to aggressive interactions and stress-related behaviours. Martins et al. ( 2012 ) reported that social stress during feeding leads dominant individuals to show increased aggression, with underfeeding increasing aggression between individuals when food becomes scarce and defendable. These behavioural patterns can compromise both animal welfare and visitor experience, necessitating targeted interventions. Furthermore, certain feeding practices may inadvertently create problematic behavioural associations. For instance, during the dive scatter feeding, where divers enter exhibits to distribute food, the animals may have been conditioned to associate human presence with food availability, potentially intensifying undesired responses. The contrasting response observed in rays, with significant increases in both desired responses and undesired responses, reflects the complex social and feeding dynamics characteristic of batoid species. Sasko et al. ( 2006 ) documented the specialised hydraulic excavation-feeding behaviour of the Atlantic cownose rays ( Rhinoptera bonasus ), which involves complex jaw movements and winnowing, and which may be disrupted by frequent feeding interruptions. Additionally, Atlantic cownose rays maintain complex social structures with individual personality differences, suggesting that feeding modifications may differentially impact individuals within the population (Rogers, 2024 ). The increased undesired responses in rays during multiple feeding conditions may reflect frustration related to reduced food availability per feeding session or disruption of natural feeding sequences. Elasmobranchs have demonstrated species-specific feeding under human care, with some species showing distinct feeding time preferences that may conflict with imposed feeding schedules (Costa et al., 2023 ). The increase in undesired responses among rays necessitated the targeted interventions implemented in Experiment 2. Additionally, the aquarists could have inadvertently rewarded undesired responses during feeding (for instance, by feeding the animals more food to encourage them to move away), which may have negatively reinforced undesired responses in the fish. Experiment 2 The successful implementation of positive reinforcement training for Javanese cownose rays demonstrates the trainability of batoid species. The significant reduction in undesired responses following target training to stretchers validates the effectiveness of targeted behavioural interventions. This finding aligns with Heinrich et al. ( 2020 ), who demonstrated that Port Jackson sharks achieved learning criteria in 13–18 sessions, with high reinforcement frequency producing faster learning rates and higher success rates. Marranzino ( 2013 ) who applied positive reinforcement techniques to target training Indo-Pacific leopard sharks ( Stegostoma tigrinum ) on a stretcher, enabling cooperative participation in medical procedures and reducing stress-related behaviours. Muraco and Stamper ( 2003 ) were also able to train one of the same species as this study, the spotted eagle rays. The stretcher target training procedure employed in this study builds on established training methodologies documented in the Elasmobranch Husbandry Manual II, in which 88.8% of surveyed institutions reported successful training programs (Janssen et al., 2017 ). The cognitive abilities demonstrated by elasmobranchs in training programs are supported by Brown and Schluessel ( 2023 ), who concluded that sharks and rays are on par with most other vertebrates in their learning and memory capabilities. The fixed-location feeding intervention for spotted eagle rays showed improvements, although statistical significance was not reached. This finding suggests that environmental modifications may be effective for some species but may require more extended implementation periods or larger effect sizes to achieve statistical significance. Studies on cownose ray enrichment have shown that behavioural changes can vary in magnitude and timeline, with some modifications requiring extended observation periods to demonstrate a significant effect (Harris et al., 2024 ). The success of positive reinforcement training programs also provides welfare benefits that extend beyond behavioural modification. Carlstead et al. ( 2019 ) demonstrated that positive keeper-animal relationships are associated with lower stress indicators and improved welfare outcomes in elephants. This also worked in reverse, with keepers’ job satisfaction related to the strength of their bond with the elephants. Positive reinforcement training is an opportunity to increase positive interactions between keepers (or, in this case, aquarists) and animals (Melfi et al., 2024 ). Cooperative participation in husbandry procedures achieved through training programmes reduces the need for restraint and sedation, minimising stress and health risks associated with routine animal care. Additionally, training can have enriching benefits, which increase positive experiences for animals in human care (Fernandez, 2022 ). Limitations The study had some limitations. First, the research was conducted at a single institution, and the specific environmental conditions, species composition, and established management protocols may not be representative of other aquarium facilities, thereby limiting the generalisability of our findings. Second, the aquarium exhibit setting inherently constrained experimental control; variables such as visitor presence and behaviour, ambient environmental conditions, water quality parameters, and social dynamics among individuals could not be fully controlled or systematically manipulated as they might be in a laboratory setting. Third, the sample sizes for some species were relatively small, limiting statistical power and reducing confidence in extrapolating these results to broader elasmobranch populations or other aquarium contexts. Finally, this study primarily focused on behavioural observations, and the inclusion of physiological stress indicators, such as cortisol concentrations, alongside additional behavioural metrics could have provided a more comprehensive assessment of welfare impacts. Future Direction Future research directions should include additional welfare measures, both physiological and physical, to complement behavioural observations, investigate individual variation in response to interventions, and long-term monitoring to assess the persistence of behavioural changes. Additionally, comparative studies across multiple institutions would strengthen the results and identify institutional factors that influence the success of interventions. It was worth noting how little has been done to study the welfare and behaviour of fish, including their responsiveness to training. This should be explored further. Although the results of the spotted eagle ray interventions were not statistically significant, future research should explore alternative environmental modifications or extended implementation periods. Converting fixed location feeding to positive reinforcement training protocols may yield more substantial behavioural improvements, as suggested by the success observed in cownose ray training. Conclusion This research demonstrates that modifying feeding events and using reward-based behaviour modification procedures represent valuable tools for improving elasmobranch behaviour in mixed-species aquarium exhibits. The observed undesired and desired responses highlight the importance of tailored management approaches that account for the unique behavioural and physiological characteristics of different taxonomic groups. The success of reward-based interventions particularly validates elasmobranchs' cognitive capabilities and supports the expanded implementation of training programs in aquarium settings. These findings contribute to the growing body of evidence supporting evidence-based management practices that enhance animal welfare in managed care environments. Declarations Author Contributions Conceptualisation, C.T., T.J.M., E.J.F.; methodology, C.T. and E.J.F.; data collection, C.T., C.K.Y.W. and L.K.L.; data analysis, C.T.; original draft, C.T.; reviewing and editing, T.J.M., E.J.F., C.K.Y.W. and L.K.L. All authors agreed to the submission of the final copy of the manuscript. Funding This research received no external funding. Conflicts of Interest The authors declare that they have no conflict of interest. Ethics approval and consent to participate This study received ethics approval from Adelaide University’s Animal Ethics Committee (Ethics No.: S-2024-030). The Singapore Oceanarium Research Panel approved all procedures, which were in accordance with all applicable national laws and/or guidelines (Singapore). Data Availability Statement The datasets generated for this study are available from the corresponding author upon request. Acknowledgements We want to thank Ramil Bentillo for designing and building the stretcher for the cownose rays. We also want to thank Nicholas Derbyshire, Raja S/O Murugesan, and Goh Zhe Zuan for their guidance and support. Lastly, we would like to thank all the team members at Large Habitats for their assistance with feeding and training the animals. References Bassett, L., Buchanan-Smith, H. M., McKinley, J. & Smith, T. E. Effects of Training on Stress-Related Behavior of the Common Marmoset ( Callithrix jacchus ) in Relation to Coping With Routine Husbandry Procedures. J. Appl. Anim. Welfare Sci. 6 (3), 221–233. https://doi.org/10.1207/S15327604JAWS0603_07 (2003). Binding, S., Farmer, H., Krusin, L. & Cronin, K. Status of animal welfare research in zoos and aquariums: Where are we, where to next? J. Zoo Aquarium Res. 8 (3), 166–174. https://doi.org/10.19227/jzar.v8i3.505 (2020). Boundey, A., Nikitins, F. & Fernandez, E. J. An examination of husbandry training in zoos: A scoping review. Appl. Anim. Behav. Sci. 292 , 106805. https://doi.org/10.1016/j.applanim.2025.106805 (2025). Brando, S. & Norman, M. Handling and Training of Wild Animals: Evidence and Ethics-Based Approaches and Best Practices in the Modern Zoo. Animals 13 (14), 2247 (2023). Brown, C. & Schluessel, V. Smart sharks: a review of chondrichthyan cognition. Anim. Cogn. 26 (1), 175–188. https://doi.org/10.1007/s10071-022-01708-3 (2023). Carlstead, K., Paris, S. & Brown, J. L. Good keeper-elephant relationships in North American zoos are mutually beneficial to welfare. Appl. Anim. Behav. Sci. 211 , 103–111. https://doi.org/10.1016/j.applanim.2018.11.003 (2019). Corwin, A. L. Training Fish and Aquatic Invertebrates for Husbandry and Medical Behaviors. Veterinary Clin. North. America: Exotic Anim. Pract. 15 (3), 455–467. https://doi.org/10.1016/j.cvex.2012.06.009 (2012). Costa, S., Neves, J., Tirá, G. & Andrade, J. P. Predatory Responses and Feeding Behaviour of Three Elasmobranch Species in an Aquarium Setting. J. Zoological Bot. Gardens . 4 (4), 775–787. https://doi.org/10.3390/jzbg4040055 (2023). D’Cruze, N. et al. A Global Review of Animal–Visitor Interactions in Modern Zoos and Aquariums and Their Implications for Wild Animal Welfare. Animals 9 (6), 332 (2019). Fernandez, E. J. Training as enrichment: A critical review. Anim Welf. 31 (1), 1–12. https://doi.org/10.7120/09627286.31.1.001 (2022). Fernandez, E. J. & Sherwen, S. L. Human-Animal Interactions in Zoos: Integrating Science and Practice (CABI, 2024). Georgopoulou, D. G., Vouidaskis, C. & Papandroulakis, N. Swimming behavior as a potential metric to detect satiation levels of European seabass in marine cages [Original Research]. Frontiers in Marine Science , Volume 11–2024 . (2024). https://doi.org/10.3389/fmars.2024.1350385 Harris, M. C. Y., Frazier, H., Mayall, S., Frey, A. D. & Boyle, S. A. Novel Food-Based Enrichment Increases Captive Cownose Stingray ( Rhinoptera bonasus ) Engagement with Enrichment Item. J. Zoological Bot. Gardens . 5 (4), 552–562. https://doi.org/10.3390/jzbg5040037 (2024). Heinrich, D. D. U., Pouca, V., Brown, C., Huveneers, C. & C., & Effects of reward magnitude and training frequency on the learning rates and memory retention of the Port Jackson shark Heterodontus portusjacksoni . Anim. Cogn. 23 (5), 939–949. https://doi.org/10.1007/s10071-020-01402-2 (2020). Huntingford, F. A. et al. Current issues in fish welfare. J. Fish Biol. 68 (2), 332–372. https://doi.org/10.1111/j.0022-1112.2006.001046.x (2006). Janse, M., Zimmerman, B., Geerlings, L., Brown, C. & Nagelkerke, L. A. J. Sustainable species management of the elasmobranch populations within European aquariums: a conservation challenge. J. Zoo Aquarium Res. 5 (4), 172–181. https://doi.org/10.19227/jzar.v5i4.313 (2017). Janssen, J. D., Kidd, A., Ferreira, A. & Snowden, S. Training and conditioning of elasmobranch in aquaria. The elasmobranch husbandry manual II: recent advances in the care of sharks, rays and their relatives. Ohio Bological Survey, Columbus , 209–222. (2017). Laterveer, M. Changing the feeding regime in a multi-species open-ocean tank by means of target training pelagic sharks (Drum and Croaker, 2006). Lawrence, K., Sherwen, S. L. & Larsen, H. Natural Habitat Design for Zoo-Housed Elasmobranch and Teleost Fish Species Improves Behavioural Repertoire and Space Use in a Visitor Facing Exhibit. Animals 11 (10), 2979. https://doi.org/10.3390/ani11102979 (2021). Lin, G. Y., Ng, K. C. H. & Fernandez, E. J. Animal–Visitor Interactions in Zoos and Aquariums: A Systematic Review. Animals , 15 (13), 1924. (2025). https://doi.org/10.3390/ani15131924 López-Olmeda, J., Noble, C. & Sánchez-Vázquez, F. Does feeding time affect fish welfare? Fish Physiol. Biochem. 38 (1), 143–152. https://doi.org/10.1007/s10695-011-9523-y (2012). Marranzino, A. The Use of Positive Reinforcement in Training Zebra Sharks ( Stegostoma fasciatum ). J. Appl. Anim. Welfare Sci. 16 (3), 239–253. https://doi.org/10.1080/10888705.2013.798555 (2013). Martins, C. I. M. et al. Behavioural indicators of welfare in farmed fish. Fish Physiol. Biochem. 38 (1), 17–41. https://doi.org/10.1007/s10695-011-9518-8 (2012). Melfi, V. Is training zoo animals enriching? Appl. Anim. Behav. Sci. 147 (3), 299–305. https://doi.org/10.1016/j.applanim.2013.04.011 (2013). Melfi, V., Ward, S., Pawson, C. & Hosey, G. Animal–Staff Interactions. In Human-Animal Interactions in Zoos: Integrating Science and Practice (33–45). CABI GB. (2024). Morimura, N. & Ueno, Y. Influences on the Feeding Behavior of Three Mammals in the Maruyama Zoo: Bears, Elephants, and Chimpanzees. J. Appl. Anim. Welfare Sci. 2 (3), 169–186. https://doi.org/10.1207/s15327604jaws0203_1 (1999). Munroe, S., Meyer, L. & Heithaus, M. R. Elasmobranch foraging strategies and tactics. In Biology of Sharks and Their Relatives (323–355). CRC. (2022). Muraco, H. & Stamper, A. Training spotted eagle rays ( Aetobatus narinari (Euphrasen) ) to decrease aggressive behaviors towards divers. (2003). Mushtaq, S. T. Aggression in aquatic environments and its relevance in aquaculture and conservation efforts. Discover Anim. 1 (1), 28. https://doi.org/10.1007/s44338-024-00026-x (2024). Poling, A., Methot, L. L. & LeSage, M. G. Fundamentals of behavior analytic research (Springer Science & Business Media, 1995). Rogers, K. N. Exploring the welfare of a captive population of Cownose rays (Rhinoptera bonasus) through social and resource interactions [Masters Thesis, University of North Florida]. Florida, USA. (2024). Sasko, D. E., Dean, M. N., Motta, P. J. & Hueter, R. E. Prey capture behavior and kinematics of the Atlantic cownose ray, Rhinoptera bonasus . Zoology 109 (3), 171–181. https://doi.org/10.1016/j.zool.2005.12.005 (2006). Sherwen, S. L. & Hemsworth, P. H. The Visitor Effect on Zoo Animals: Implications and Opportunities for Zoo Animal Welfare. Animals 9 (6), 366. https://doi.org/10.3390/ani9060366 (2019). Ward, A. J. W., Webster, M. M. & Hart, P. J. B. Intraspecific food competition in fishes. Fish Fish. 7 (4), 231–261. https://doi.org/10.1111/j.1467-2979.2006.00224.x (2006). Whitham, J. C. & Wielebnowski, N. New directions for zoo animal welfare science. Appl. Anim. Behav. Sci. 147 (3), 247–260. https://doi.org/10.1016/j.applanim.2013.02.004 (2013). Williams, E., Hunton, V., Hosey, G. & Ward, S. J. The Impact of Visitors on Non-Primate Species in Zoos: A Quantitative Review. Animals 13 (7), 1178. https://doi.org/10.3390/ani13071178 (2023). Additional Declarations No competing interests reported. Supplementary Files Appendices.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 02 May, 2026 Reviewers invited by journal 01 Apr, 2026 Editor invited by journal 31 Mar, 2026 Editor assigned by journal 25 Mar, 2026 Submission checks completed at journal 25 Mar, 2026 First submitted to journal 24 Mar, 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-9208737","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":617361939,"identity":"d9501319-0773-4847-a307-98c72876b0e1","order_by":0,"name":"Claudia Tay","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABCElEQVRIiWNgGAWjYDACdiDmMQAzGR8Qp4UZqoUHyIToZCNKCwSzSRClhb+Z+ZnEmwI7Bnv2M2bVlW11cubyzYc//GCwk9NtwK5F4jCbmeQcg2QGHp4cs5tn2w4bW7axpUn2MCQbmx3ArsWAmcFMmgdI8jAAtTRuO5C44RiPGZB3IHEbTi3s34Ba6hl4+N+YFTZuq6sHajH++AevFh6QLYcZeCRyzBgbtzEnGBzjMZDGZ4vEYZ5iyzkGx3l4bjwrlmz8d9hww7G0NGkZA9x+4W9v33jjzZ9qOfb+5I0fG87UyRscPnz445sKOzlcWmAAGC0cBsgOxq8cCtgfEKVsFIyCUTAKRh4AAJwGUOu6nvKkAAAAAElFTkSuQmCC","orcid":"","institution":"Adelaide University","correspondingAuthor":true,"prefix":"","firstName":"Claudia","middleName":"","lastName":"Tay","suffix":""},{"id":617361940,"identity":"593a76bd-03d8-4fdf-b7aa-36434dcdf5aa","order_by":1,"name":"Todd J. McWhorter","email":"","orcid":"","institution":"Adelaide University","correspondingAuthor":false,"prefix":"","firstName":"Todd","middleName":"J.","lastName":"McWhorter","suffix":""},{"id":617361941,"identity":"69c7efbd-4d59-45d0-bb43-abe43284e7dc","order_by":2,"name":"Clement Yu Wei Koh","email":"","orcid":"","institution":"Singapore Oceanarium, Resorts World Sentosa","correspondingAuthor":false,"prefix":"","firstName":"Clement","middleName":"Yu Wei","lastName":"Koh","suffix":""},{"id":617361946,"identity":"da8802fe-eb52-4a11-bc6b-c040995c164e","order_by":3,"name":"Kai Le Leong","email":"","orcid":"","institution":"Singapore Oceanarium, Resorts World Sentosa","correspondingAuthor":false,"prefix":"","firstName":"Kai","middleName":"Le","lastName":"Leong","suffix":""},{"id":617361947,"identity":"e032cfed-50aa-4cc3-9870-0586d81ba913","order_by":4,"name":"Eduardo J. Fernandez","email":"","orcid":"","institution":"Adelaide University","correspondingAuthor":false,"prefix":"","firstName":"Eduardo","middleName":"J.","lastName":"Fernandez","suffix":""}],"badges":[],"createdAt":"2026-03-24 08:10:45","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9208737/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9208737/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106403865,"identity":"fdf4a39a-d98e-426d-a0e0-2d3a022fa7dc","added_by":"auto","created_at":"2026-04-08 09:15:06","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":136391,"visible":true,"origin":"","legend":"\u003cp\u003eExperimental set-up for Experiment 1. Baseline: Surface scatter, dive scatter and dive target feed at 1600h. Multiple Feeding Intervention: Surface scatter at 0900h and 1300h, dive target feed at 1600h.\u003c/p\u003e","description":"","filename":"Fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9208737/v1/b47f4744e56ab1b084a90e05.jpg"},{"id":106402793,"identity":"f6b6142c-2495-408b-8859-c6c81cecbfcd","added_by":"auto","created_at":"2026-04-08 09:12:52","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":91850,"visible":true,"origin":"","legend":"\u003cp\u003eOrder in which the study was carried out and the sequence of data collection in each session\u003c/p\u003e","description":"","filename":"Fig2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9208737/v1/a43b1d44dc7f2612a0867dd8.jpg"},{"id":106402782,"identity":"0ea63a07-af05-40cd-aa39-1422e8a3dfe8","added_by":"auto","created_at":"2026-04-08 09:12:50","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":994269,"visible":true,"origin":"","legend":"\u003cp\u003eThe mean count of desired and undesired responses exhibited by sharks (a), rays (b) and bony fish (c) under control (single feeding at 1600h) and modified feeding (feedings at 0930, 1300, and 1600h) conditions. n = 40. Error bars represent the standard error of the mean. Asterisks denote statistical significance: * p \u0026lt; 0.05, ** p \u0026lt; 0.01, *** p \u0026lt; 0.001, **** p\u0026lt;0.0001.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9208737/v1/d006f5ee106f952206b84754.jpg"},{"id":106241626,"identity":"cb29d6ef-5ef5-4a20-823c-1ff322fda2d4","added_by":"auto","created_at":"2026-04-06 15:08:03","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":669162,"visible":true,"origin":"","legend":"\u003cp\u003eAverage crowd size before and after dive feeding sessions under baseline and experimental feeding conditions. n = 40. Error bars represent the standard error of the mean.\u003c/p\u003e","description":"","filename":"Fig4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9208737/v1/eb6c29a5255abceac488cac0.jpg"},{"id":106241628,"identity":"9d725e7b-066c-470e-a6b1-fe9441754e4d","added_by":"auto","created_at":"2026-04-06 15:08:03","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":801235,"visible":true,"origin":"","legend":"\u003cp\u003eAverage sound pressure levels (dBA) before and after dive feeding sessions under baseline and experimental feeding conditions, n =40. Error bars represent the standard error of the mean.\u003c/p\u003e","description":"","filename":"Fig5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9208737/v1/11eb0f00fc7f7c43cf8ba305.jpg"},{"id":106414834,"identity":"408f7a16-1141-496d-9550-f21221685d75","added_by":"auto","created_at":"2026-04-08 10:26:37","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":154406,"visible":true,"origin":"","legend":"\u003cp\u003eExperimental setup for Experiment 2. Multiple Feeding Intervention: Surface scatter at 0900h and 1300h, dive target feed at 1600h. Post-training Intervention: Surface scatter at 0900h and 1300h, dive target feed at 1600h, positive reinforcement training for cownose rays and fixed location feeding for eagle rays at 1530h\u003c/p\u003e","description":"","filename":"Fig6..jpg","url":"https://assets-eu.researchsquare.com/files/rs-9208737/v1/ae9c3c72e915bf2bac6d11ae.jpg"},{"id":106241629,"identity":"3567e39f-e0df-44ce-b1a9-6250e73f8053","added_by":"auto","created_at":"2026-04-06 15:08:03","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":1291622,"visible":true,"origin":"","legend":"\u003cp\u003eAverage count of desired and undesired responses exhibited by (a) Javanese cownose ray, (b) spotted eagle ray, (c) other ray under baseline and training conditions. Error bars represent the standard error of the mean. Asterisks denote statistical significance: *** p \u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9208737/v1/e458b98a2b5d9277494471ea.jpg"},{"id":106417485,"identity":"b1a9367f-acb6-4bdd-bd78-932398271db7","added_by":"auto","created_at":"2026-04-08 10:50:39","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4783569,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9208737/v1/2dbb26be-05b3-4ff0-a301-78c47cf93bbb.pdf"},{"id":106241623,"identity":"d9aac798-707d-41e4-a84e-0e07bd856972","added_by":"auto","created_at":"2026-04-06 15:08:03","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1477302,"visible":true,"origin":"","legend":"","description":"","filename":"Appendices.docx","url":"https://assets-eu.researchsquare.com/files/rs-9208737/v1/5c23355489a2e621547958f9.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"The Age of Aquariums: Managing Elasmobranch Behaviour in a Mixed-Species Habitat","fulltext":[{"header":"Introduction","content":"\u003cp\u003eElasmobranchs have become flagship species in modern aquariums, captivating millions of visitors annually while serving critical roles in conservation education and species preservation. However, the management of captive elasmobranchs in mixed-species exhibits presents unique welfare challenges. These challenges necessitate changes to husbandry that address species-specific needs while maintaining a positive visitor experience. Currently, 102 chondrichthyan species are maintained across European facilities alone (Janse et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), many of which face escalating threats in their native ecosystems. As public aquariums increasingly position themselves as conservation institutions, the imperative to optimise animal welfare while maintaining educational value has never been more pressing.\u003c/p\u003e \u003cp\u003eDespite growing recognition of the importance of welfare assessment in aquarium settings, significant knowledge gaps persist regarding fish welfare relative to that of terrestrial vertebrates. Huntingford et al. (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2006\u003c/span\u003e) highlighted that welfare assessments remain more challenging for aquatic animals than for mammals, primarily due to methodological difficulties, the diversity of fish taxa, and historical research biases. This disparity is particularly concerning given that fish possess complex behavioural repertoires, sophisticated sensory systems, and physiological needs that differ substantially from terrestrial species. The lack of standardised welfare indicators for elasmobranchs specifically has left aquarium professionals with limited empirical guidance for husbandry decisions that directly impact animal wellbeing. A review of animal welfare publications in zoos and aquariums found that only 2 of 310 (0.6%) studies focused on fish, compared to 232 on mammals (Binding et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAnimal-visitor interactions (AVIs) represent a critical yet understudied aspect of aquarium animal management (Fernandez \u0026amp; Sherwen, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Lin et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Nonetheless, research on AVIs in aquariums remains limited. D\u0026rsquo;Cruze et al. (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) conducted a global review of AVIs in zoos and aquariums, revealing that Chondrichthyes and Actinopterygii accounted for only 5% and 1.8% of published interaction studies, despite their large numbers in aquatic collections. Sherwen and Hemsworth (\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) noted that there were no studies published on the visitor effect on fish then and it was only in 2020 that the first paper on fish was published (Williams et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This lack of research on aquatic AVIs, combined with the prevalence of visitor-related stressors in high-traffic exhibits, underscores the urgent need for empirical investigations into how visitor presence and feeding interactions affect the behaviour and welfare of elasmobranchs.\u003c/p\u003e \u003cp\u003eFeeding management represents a fundamental factor of captive elasmobranch welfare, yet current practices often fail to reflect species-specific ecological requirements. In wild populations, elasmobranchs exhibit remarkable diversity in foraging strategies, from benthic ambush predators to active pelagic hunters, with many species displaying opportunistic or seasonally variable feeding patterns (Munroe et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Under human care, standardised feeding routines typically constrain this natural flexibility, often imposing across species with different metabolic demands and activity patterns and behaviour (Morimura \u0026amp; Ueno, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e1999\u003c/span\u003e). Even at the individual level, there may be high variability in feeding patterns (L\u0026oacute;pez-Olmeda et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Fish research demonstrates that feeding events can intensify competition, elevate stress-related cortisol levels, and increase the risk of aggressive encounters and physical injury (Mushtaq, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Ward et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). These findings suggest that strategic modifications to feeding schedules and methodologies may be effective non-invasive interventions to improve welfare outcomes across diverse elasmobranch species in aquarium settings.\u003c/p\u003e \u003cp\u003e In addition to feeding strategies, the use of operant conditioning and training programmes has become an integral component of animal care in aquariums and zoological facilities. Animal training, particularly using positive reinforcement techniques, can promote cooperative participation in husbandry, reduce stress during medical procedures and provide a form of cognitive enrichment that enhances welfare (Boundey et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Melfi, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Positive reinforcement training has also been shown to reduce stress in animals (Bassett et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Whitham \u0026amp; Wielebnowski, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Brown and Schluessel (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) demonstrated that sharks and rays are capable of discrimination learning, habituation, and long-term memory, providing a cognitive foundation for training. The application of training to elasmobranchs has shown considerable promise, although fewer studies have been conducted than for other taxonomic groups, such as mammals and birds. Corwin (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2012\u003c/span\u003e) documented the successful application of positive reinforcement training with eels to reduce aggression during feeding. In another study, implementation of visual target training at two designated locations within a multispecies exhibit resulted in an 80% reduction in behavioural problems associated with feeding events (Laterveer, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Aquarists introduced a visual target as a conditioned stimulus for food delivery, significantly reducing aggressive behaviour and facilitating safer, more controlled feeding events in spotted eagle rays (\u003cem\u003eAetobatus narinari\u003c/em\u003e) (Muraco \u0026amp; Stamper, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). However, a lack of peer-reviewed publications persists, with few studies examining the effects of training on fish behaviour compared with other taxa (Brando \u0026amp; Norman, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eGiven the complex interactions among feeding events, AVIs, and behavioural modification, this study aimed to evaluate two management interventions for improving fish welfare at Singapore Oceanarium, Resorts World Sentosa, Singapore. The study was structured as two complementary experiments. The first experiment examined the efficacy of modified feeding events in mitigating undesired responses across shark, ray, and bony fish species, testing the hypothesis that strategic alterations in feeding presentation and frequency would reduce anticipatory aggression and improve overall behavioural patterns. The second experiment focused specifically on target training interventions for Javanese cownose rays (\u003cem\u003eRhinoptera javanica\u003c/em\u003e) and fixed location feeding for spotted eagle rays (\u003cem\u003eAetobatus narinari\u003c/em\u003e). This experiment tested whether positive reinforcement training could reshape specific behavioural patterns while maintaining appropriate feeding responses. Together, these investigations provide a comprehensive assessment of two critical management tools, feeding event and training modifications, for improving elasmobranch behaviour in managed care environments.\u003c/p\u003e"},{"header":"Materials and Methods ","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eGeneral Materials and Methods\u003c/h2\u003e \u003cp\u003eThe study was conducted at Open Ocean Habitat (OOH) (Appendix 1), located at Singapore Oceanarium within Resorts World Sentosa, Singapore. OOH is a habitat that contains 18\u0026nbsp;million litres of water and is home to 80 species of animals. The fish were divided into three main taxonomic groups: \u0026ldquo;rays,\u0026rdquo; representing all batoids from Class \u003cem\u003eChondrichthyes\u003c/em\u003e; \u0026ldquo;sharks,\u0026rdquo; representing all \u003cem\u003eSelachii;\u003c/em\u003e and \u0026ldquo;bony fish\u0026rdquo;, representing all Class \u003cem\u003eOsteichthyes\u003c/em\u003e. Some examples of rays in OOH included the Javanese cownose ray (\u003cem\u003eRhinoptera javanica\u003c/em\u003e), spotted eagle ray (\u003cem\u003eAetobatus narinari\u003c/em\u003e), leopard whipray (\u003cem\u003eHimantura leoparda\u003c/em\u003e), and bowmouth guitarfish (\u003cem\u003eRhina ancylostoma\u003c/em\u003e). Some examples of sharks in OOH were scalloped hammerhead (\u003cem\u003eSphyrna lewini\u003c/em\u003e), tawny nurse shark (\u003cem\u003eNebrius ferrugineus\u003c/em\u003e), blacktip reef shark (\u003cem\u003eCarcharhinus melanopterus\u003c/em\u003e), and Indo-Pacific leopard shark (\u003cem\u003eStegostoma tigrinum\u003c/em\u003e). Some examples of bony fish in OOH were giant grouper (\u003cem\u003eEpinephelus lanceolatus\u003c/em\u003e), golden pompano (\u003cem\u003eTrachinotus blochii)\u003c/em\u003e, and yellowtail scad (\u003cem\u003eAtule mate\u003c/em\u003e). The animals in OOH were fed once a day, five times a week, at 1600 hours (Tuesdays, Wednesdays, Thursdays, Saturdays, Sundays), which was when observations were conducted for both experiments. The animals were only fed on Mondays via surface scatter feeding, and not at all on Fridays.\u003c/p\u003e \u003cp\u003eAll behavioural observations were systematically recorded utilising ZooMonitor software (version 1.7.140; Lincoln Park Zoo, Chicago, IL, USA). Two distinct classes of responses were identified among all animals observed: Desired and Undesired. Desired responses were defined as positive or neutral interactions, with the fish taking the food from divers at the correct position and without rushing. Conversely, undesired responses included any unwanted contact with divers or their equipment, such as nibbling, pushing, blanketing, or grabbing. The responses were mutually exclusive. Data collection was terminated upon completion of each 20-minute dive feeding session.\u003c/p\u003e \u003c/div\u003e"},{"header":"Experiment 1","content":"\u003cp\u003eIn the first experiment, feeding events were modified in two ways: the timing of delivery (schedule) and what occurred. The schedule was increased from one to three times a day, with scatter feeding during the first two feedings and dive target feeding during the last. Dive scatter feeding was discontinued.\u003c/p\u003e\n\u003ch3\u003eMaterials and Methods\u003c/h3\u003e\n\u003cp\u003eThe study was conducted from 1 May 2024 to 14 July 2024. A total of 40 sessions were recorded. There were some days when observations could not be conducted because a lightning warning prevented divers from entering the water. During the baseline phase, feed distribution occurred daily at 1600h using three distinct methodologies:\u003c/p\u003e \u003cp\u003e \u003cem\u003eSurface scatter feeding\u003c/em\u003e - Feed comprising a variety of fish, crustaceans and molluscs was scattered outwards into the habitat at two feeding points via a suspension bridge hanging above the habitat.\u003c/p\u003e \u003cp\u003e \u003cem\u003eDive scatter feeding\u003c/em\u003e - Three divers (two feeders, one safety) descended into the habitat with feeding tubs containing a variety of fish, crustaceans, and molluscs, and the feed was scattered along the habitat floor.\u003c/p\u003e \u003cp\u003e \u003cem\u003eDive target feeding\u003c/em\u003e - Six divers (four feeders, two safety) descended into the habitat, each feeder carrying a specific amount of large and small feeds separated by size. Feeders were divided into two groups: one fed all the rays, while the other fed all the other fish in the habitat.\u003c/p\u003e \u003cp\u003eThe feed composition consisted of whole fish and fish segments. Each dive feeding session, which incorporated both dive scatter and dive target feeding techniques, lasted approximately 20 minutes. The observer was positioned at a standardised location in the gallery to count visitors to OOH habitat. They then relocated to a predetermined position directly facing OOH to measure ambient crowd noise five minutes before the start of dive feeding at 1600h. Crowd noise was quantified as sound pressure levels (reference value: 20 \u0026micro;Pa) in decibels (dBA) using A-weighting and a frequency response range of 31.5-8,000Hz, measured with a Mengshen Sound Level Meter. The feeding protocol involved six divers, divided into two equal groups, stationed in designated areas within OOH. Each group consisted of two divers conducting target feeding and one safety diver performing scatter feeding simultaneously. Throughout the dive feeding session, the observer alternated between two fixed observation points in front of OOH at one-minute intervals, continuing this pattern until the conclusion of the feeding session.\u003c/p\u003e \u003cp\u003eIn the multiple feeding event condition, all experimental parameters were maintained at baseline levels, including the total quantity of feed provided and the frequency of dive-feeding days. However, the modification was implemented, halving the food allocated for surface scatter feeding and dive scatter feeding and providing it at 0900h and 1300h, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Consequently, at 1600h, only dive target feeding was conducted during the scheduled dive feeding session.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe animals were categorised into three taxonomic groups (sharks, rays, or bony fish), and the frequency of both desired and undesired responses was counted. Upon completion of the 20-minute feeding session, data collection was terminated, and sound pressure levels and crowd size were immediately reassessed at the exact locations as before the dive feed. The experimental protocol consisted of 10 consecutive baseline sessions, followed by 10 sessions under the modified feeding schedule intervention. Subsequently, a return to baseline conditions was implemented, and ten additional sessions were recorded. Then, the experimental feeding schedule intervention was reintroduced for a final sequence of ten sessions. This comprehensive ABAB experimental design (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) allowed a comparative analysis between the intervention and control conditions, whilst controlling for potential temporal effects and enhancing the reliability of observed behavioural responses to the modified feeding schedule.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eInterobserver Agreement (IOA)\u003c/h3\u003e\n\u003cp\u003eTo assess interobserver reliability, a secondary observer independently collected data during 25% of all observation sessions. Both observers conducted simultaneous in-person observations from an identical vantage point. Interobserver agreement (IOA) was calculated using the total agreement method (Poling et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e1995\u003c/span\u003e), with agreement scores derived for each behavioural category. There was an overall mean IOA of 94.15% across all behavioural classes (Bony Fish, Desired Response: 100%; Bony Fish, Undesired Response: 100%; Ray, Desired Response: 91.84%; Ray, Undesired Response: 96.67%; Shark, Desired Response: 88.89%; Shark, Undesired Response: 87.50%).\u003c/p\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eStatistical analyses were performed using RStudio (version 2024.12.1\u0026thinsp;+\u0026thinsp;563). A generalised linear mixed-effects model (GLMM) with a negative binomial error distribution was used for the responses count dataset. Treatment was coded with two levels (Control, Modified Feed) and entered as a fixed effect. Session ID was included as a random intercept to account for repeated sampling and session-level heterogeneity. For the GLMM, models were fit using the package \u0026lsquo;glmmTMB\u0026rsquo;. Overdispersion and residual diagnostics were assessed using the \u0026lsquo;DHARMa\u0026rsquo; package, which simulated residuals (QQ plots, residual vs. fitted plots, and dispersion tests). The results did not indicate significant deviations from model assumptions for the fitted models. AIC/BIC and dispersion values are consistent with an adequate fit. The undesired contact for bony fish exhibited complete zeros across treatments, and the treatment effect was not interpreted for this outcome. Statistical significance was set at α\u0026thinsp;=\u0026thinsp;0.05.\u003c/p\u003e \u003cp\u003eRegarding the crowd size and sound pressure level dataset, normal distribution was evaluated using the Shapiro-Wilk test. Levene's test for homogeneity of variances was applied to assess equality of variances; all tests failed to reject the null hypothesis, confirming that the assumption of equal variances was met. For datasets violating normality assumptions, the Mann-Whitney U test was used, while independent-samples t-tests were used for datasets that satisfied the normality assumptions. Statistical significance was established at α\u0026thinsp;=\u0026thinsp;0.05 across all analyses.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eResults\u003c/h3\u003e\n\u003cp\u003eSharks displayed significantly reduced undesired contact response under the modified feed treatment compared to the control. The average counts of undesired contact were 18.60 under control conditions and 10.25 under modified feed conditions (β = -0.596, SE\u0026thinsp;=\u0026thinsp;0.157, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). In contrast, desired contact response in sharks did not differ significantly between treatments, with mean counts of 10.45 in the control group and 11.73 for the modified feed (β\u0026thinsp;=\u0026thinsp;0.116, SE\u0026thinsp;=\u0026thinsp;0.114, p\u0026thinsp;=\u0026thinsp;0.312). For the rays, the LMM indicated a significant increase (64.55 to 98.8) in undesired response after the feed conditions were modified (t\u0026thinsp;=\u0026thinsp;4.59, df\u0026thinsp;=\u0026thinsp;58.08, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb). For desired contact in rays, there was an increase (50.5 to 61.5), although not statistically significant (t\u0026thinsp;=\u0026thinsp;1.47, df\u0026thinsp;=\u0026thinsp;58.08, p\u0026thinsp;=\u0026thinsp;0.15). Bony fish demonstrated minimal behavioural interactions overall. Desired responses were observed at low frequencies in both conditions, with a slight reduction from baseline (0.40) to multiple feeding conditions (0.20), although this difference did not reach statistical significance (β = -0.693, SE\u0026thinsp;=\u0026thinsp;0.694, p\u0026thinsp;=\u0026thinsp;0.318; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). Notably, no undesired responses were documented in either feeding condition (0.00 in both conditions).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003ePre-dive feed crowd sizes remained consistent between baseline (206.45) and experimental condition (207.15), with no significant difference observed (t = -0.044604, df\u0026thinsp;=\u0026thinsp;38, p\u0026thinsp;=\u0026thinsp;0.965; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). However, post-dive feed crowd size showed a non-significant increase (U\u0026thinsp;=\u0026thinsp;139, p\u0026thinsp;=\u0026thinsp;0.102) in the experimental condition (197.35) compared to the baseline (175.45; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003ePre-dive feed sound levels were comparable between baseline (70.58) and experimental condition (70.76), with minimal variation observed (t = -0.19357, df\u0026thinsp;=\u0026thinsp;38, p\u0026thinsp;=\u0026thinsp;0.848; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Similarly, post-dive feed sound levels showed negligible differences between baseline (70.53) and experimental condition (71.31) (U\u0026thinsp;=\u0026thinsp;158, p\u0026thinsp;=\u0026thinsp;0.262). The consistency in sound pressure levels across both time points and experimental conditions suggests that the modified feeding schedule did not substantially alter the acoustic environment within OOH.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eExperiment 1 resulted in a significant decrease in undesired responses for the sharks. However, the experiment also resulted in a significant increase in both desired and undesired responses among the rays. Observers noted that this response was primarily exhibited by two species: the Javanese cownose ray (\u003cem\u003eRhinoptera javanica\u003c/em\u003e) and the spotted eagle ray (\u003cem\u003eAetobatus narinari\u003c/em\u003e). This increase may be attributed to reduced food availability during the targeted feeding sessions compared to baseline conditions. Since these species require time to process their food, the reduced feeding opportunities may have left them less satiated, potentially triggering undesired responses. This led to the next experiment, which focused on interventions for the two ray species: training for cownose rays and fixed location feeding for spotted eagle rays.\u003c/p\u003e"},{"header":"Experiment 2","content":"\u003cp\u003eIn the second experiment, the feeding protocols remained unchanged for the other fish species; however, the Javanese cownose rays received a reward-based target-training procedure, while the spotted eagle rays received fixed-location feeding. Observations were limited to the rays.\u003c/p\u003e\n\u003ch3\u003eMaterials and Methods\u003c/h3\u003e\n\u003cp\u003eThis study was conducted from 22 October 2024 to 23 January 2025, encompassing 40 observational sessions. Certain scheduled observations were omitted due to lightning warnings, which prevented divers from entering the water for safety reasons. The experimental subjects were exclusively rays from the initial experiment, categorised into three groups: 42 individual Javanese cownose rays, 12 spotted eagle rays, and a combined group of all other ray species. The baseline protocol for this second experiment replicated the feeding intervention established in the previous experiment, consisting of two surface scatter feedings at 0930h and 1300h, and dive target feeding at 1600h.\u003c/p\u003e \u003cp\u003eBaseline data collection comprised 20 sessions, concluding on 15 November 2024. This phase focused solely on behavioural observations; crowd size and sound pressure level measurements were not included in the experimental design. Following the completion of these initial sessions, a reward-based target training programme involving contact with (\u0026ldquo;targeting\u0026rdquo;) a surface water stretcher (Appendix 2) was initiated for the Javanese cownose rays, whilst the eagle rays commenced surface feeding at a designated location. Rewards consisted of fish, constituting their standard dietary regimen. Training sessions were conducted at 1530h for 20 minutes on each dive feeding day (Tuesdays, Wednesdays, Thursdays, Saturdays and Sundays). Figure\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e details the baseline and experimental conditions for both rays.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eOnly the rays were observed for this experiment, and they were categorised into three groups: \u0026ldquo;Cownose ray\u0026rdquo; for the Javanese cownose ray, \u0026ldquo;Eagle ray\u0026rdquo; for the spotted eagle ray, and \u0026ldquo;Other ray\u0026rdquo; for all other ray species. The subsequent phase of data collection was contingent upon the Javanese cownose rays demonstrating proficiency in target training to the stretcher, specifically requiring more than 42 instances of complete stretcher transit (entry to exit) across three consecutive sessions. This criterion was met on 15 December 2024, allowing for the commencement of the final data collection phase on 17 December 2024. An additional 20 observational sessions were documented during this phase, with training protocols continuing concurrently on all dive feeding days.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eInterobserver Agreement\u003c/h2\u003e \u003cp\u003eTo assess interobserver reliability, two additional observers independently collected data during 18% and 10% of all observation sessions. All three observers conducted simultaneous in-person observations from an identical vantage point. IOA was calculated using the total agreement method (Poling et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e1995\u003c/span\u003e), with agreement scores derived for each behavioural category. Between the primary and first additional observer, there was an overall mean IOA of 91.98% across all behavioural classes (Cownose Ray, Desired Response: 92.31%; Cownose Ray, Undesired Response: 91.3%; Spotted Eagle Ray, Desired Response: 92.31%; Spotted Eagle Ray, Undesired Response: 93.33%; Other Ray, Desired Response: 88.89%; Other Ray, Undesired Response: 93.75%). Between the primary and second additional observer, there was an overall mean IOA of 92.21% across all behavioural classes (Cownose Ray, Desired Response: 92.31%; Cownose Ray, Undesired Response: 92%; Spotted Eagle Ray, Desired Response: 92.86%; Spotted Eagle Ray, Undesired Response: 88.24%; Other Ray, Desired Response: 93.75%; Other Ray, Undesired Response: 94.12%).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eGeneralised linear mixed models (GLMMs) with a Poisson distribution and log link function were fitted for each species-response combination to account for repeated measures across sessions and the count nature of the behavioural data. The model structure was Response Count\u0026thinsp;~\u0026thinsp;Treatment + (1 | SessionID) where Treatment was a fixed effect (Pre vs. Post) and SessionID was included as a random intercept to account for session-level variation. Models were fitted using the \u0026lsquo;glmmTMB\u0026rsquo; package in R. Model fit was assessed using Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). Statistical significance was determined at α\u0026thinsp;=\u0026thinsp;0.05.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eResults\u003c/h3\u003e\n\u003cp\u003eDesired contact responses in cownose rays did not change significantly (15.75 to 13.54) following the stretcher training intervention (β = -0.152, SE\u0026thinsp;=\u0026thinsp;0.114, z = -1.32, p\u0026thinsp;=\u0026thinsp;0.185; Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ea). In contrast, undesired contact responses decreased significantly post-training, from 42.25 to 23.1 (β = -0.665, SE\u0026thinsp;=\u0026thinsp;0.103, z = -6.47, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ea). This represents a substantial reduction in undesired contact responses following the training intervention. For spotted eagle rays, desired contact responses showed no significant change post-training, 14.35 to 14.55 (β\u0026thinsp;=\u0026thinsp;0.018, SE\u0026thinsp;=\u0026thinsp;0.102, z\u0026thinsp;=\u0026thinsp;0.18, p\u0026thinsp;=\u0026thinsp;0.857; Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eb). Undesired contact responses in spotted eagle rays showed a marginally non-significant trend toward reduction after the training, 27.75 to 22, (β = -0.214, SE\u0026thinsp;=\u0026thinsp;0.117, z = -1.83, p\u0026thinsp;=\u0026thinsp;0.067; Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eb). Other ray species showed no significant changes in either desired or undesired contact responses following training. Desired contact responses remained stable at 23.55 to 24.55 (β\u0026thinsp;=\u0026thinsp;0.032, SE\u0026thinsp;=\u0026thinsp;0.107, z\u0026thinsp;=\u0026thinsp;0.30, p\u0026thinsp;=\u0026thinsp;0.766; Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ec). Similarly, undesired contact response in other ray species showed no significant change at 23.8 to 20.6 (β = -0.165, SE\u0026thinsp;=\u0026thinsp;0.157, z = -1.06, p\u0026thinsp;=\u0026thinsp;0.291; Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ec).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe training intervention resulted in a significant reduction in undesired contact responses, specifically in cownose rays, while desired contact responses remained unchanged across all species. Spotted eagle rays showed a trend toward fewer undesired contacts, which approached but did not reach statistical significance. Other ray species showed no significant behavioural changes in response to the training intervention.\u003c/p\u003e \u003cp\u003eStretcher target training was effective in reducing undesired response in Javanese cownose rays. The fixed-location feeding training for spotted eagle rays resulted in a reduction in undesired responses, although this was not significant.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study offers insights into modifying feeding events (Exp 1) and implementing behaviour modification procedures (Exp 2) in a large-scale, mixed-species oceanarium habitat. The differential responses observed among sharks, rays, and bony fish in the first experiment highlighted the need for tailored management approaches that account for the unique behavioural and physiological characteristics of ray species. Specifics of each experiment are detailed below.\u003c/p\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eExperiment 1\u003c/h2\u003e \u003cp\u003eThe significant reduction in undesired responses among sharks following the implementation of multiple daily feedings aligns with documented environmental sensitivity in elasmobranch species. For instance, Lawrence et al. (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) found that Port Jackson (\u003cem\u003eHeterodontus portusjackson)\u003c/em\u003e sharks significantly reduced abnormal behaviours following environmental design change, indicating that sharks are responsive to husbandry modifications. This finding is also consistent with research on feeding behaviour in aquaculture settings, where Georgopoulou et al. (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) demonstrated that a single daily feeding in European seabass (\u003cem\u003eDicentrarchus labrax\u003c/em\u003e) produced higher activity levels and more symmetrical behaviour around feeding times than multiple feedings, which may be stressful for the fish. The authors noted that reduced feeding led to prolonged increased activity before and after feeding, directly paralleling our observations of reduced problematic behaviours when food was distributed across multiple sessions rather than concentrated in a single feeding event.\u003c/p\u003e \u003cp\u003eUndesired responses during feeding represent a significant challenge in aquarium management, particularly in mixed-species exhibits where competition for resources can lead to aggressive interactions and stress-related behaviours. Martins et al. (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2012\u003c/span\u003e) reported that social stress during feeding leads dominant individuals to show increased aggression, with underfeeding increasing aggression between individuals when food becomes scarce and defendable. These behavioural patterns can compromise both animal welfare and visitor experience, necessitating targeted interventions. Furthermore, certain feeding practices may inadvertently create problematic behavioural associations. For instance, during the dive scatter feeding, where divers enter exhibits to distribute food, the animals may have been conditioned to associate human presence with food availability, potentially intensifying undesired responses.\u003c/p\u003e \u003cp\u003eThe contrasting response observed in rays, with significant increases in both desired responses and undesired responses, reflects the complex social and feeding dynamics characteristic of batoid species. Sasko et al. (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2006\u003c/span\u003e) documented the specialised hydraulic excavation-feeding behaviour of the Atlantic cownose rays (\u003cem\u003eRhinoptera bonasus\u003c/em\u003e), which involves complex jaw movements and winnowing, and which may be disrupted by frequent feeding interruptions. Additionally, Atlantic cownose rays maintain complex social structures with individual personality differences, suggesting that feeding modifications may differentially impact individuals within the population (Rogers, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe increased undesired responses in rays during multiple feeding conditions may reflect frustration related to reduced food availability per feeding session or disruption of natural feeding sequences. Elasmobranchs have demonstrated species-specific feeding under human care, with some species showing distinct feeding time preferences that may conflict with imposed feeding schedules (Costa et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The increase in undesired responses among rays necessitated the targeted interventions implemented in Experiment 2. Additionally, the aquarists could have inadvertently rewarded undesired responses during feeding (for instance, by feeding the animals more food to encourage them to move away), which may have negatively reinforced undesired responses in the fish.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eExperiment 2\u003c/h2\u003e \u003cp\u003eThe successful implementation of positive reinforcement training for Javanese cownose rays demonstrates the trainability of batoid species. The significant reduction in undesired responses following target training to stretchers validates the effectiveness of targeted behavioural interventions. This finding aligns with Heinrich et al. (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), who demonstrated that Port Jackson sharks achieved learning criteria in 13\u0026ndash;18 sessions, with high reinforcement frequency producing faster learning rates and higher success rates. Marranzino (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) who applied positive reinforcement techniques to target training Indo-Pacific leopard sharks (\u003cem\u003eStegostoma tigrinum\u003c/em\u003e) on a stretcher, enabling cooperative participation in medical procedures and reducing stress-related behaviours. Muraco and Stamper (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2003\u003c/span\u003e) were also able to train one of the same species as this study, the spotted eagle rays. The stretcher target training procedure employed in this study builds on established training methodologies documented in the Elasmobranch Husbandry Manual II, in which 88.8% of surveyed institutions reported successful training programs (Janssen et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The cognitive abilities demonstrated by elasmobranchs in training programs are supported by Brown and Schluessel (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), who concluded that sharks and rays are on par with most other vertebrates in their learning and memory capabilities.\u003c/p\u003e \u003cp\u003eThe fixed-location feeding intervention for spotted eagle rays showed improvements, although statistical significance was not reached. This finding suggests that environmental modifications may be effective for some species but may require more extended implementation periods or larger effect sizes to achieve statistical significance. Studies on cownose ray enrichment have shown that behavioural changes can vary in magnitude and timeline, with some modifications requiring extended observation periods to demonstrate a significant effect (Harris et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe success of positive reinforcement training programs also provides welfare benefits that extend beyond behavioural modification. Carlstead et al. (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) demonstrated that positive keeper-animal relationships are associated with lower stress indicators and improved welfare outcomes in elephants. This also worked in reverse, with keepers\u0026rsquo; job satisfaction related to the strength of their bond with the elephants. Positive reinforcement training is an opportunity to increase positive interactions between keepers (or, in this case, aquarists) and animals (Melfi et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Cooperative participation in husbandry procedures achieved through training programmes reduces the need for restraint and sedation, minimising stress and health risks associated with routine animal care. Additionally, training can have enriching benefits, which increase positive experiences for animals in human care (Fernandez, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eLimitations\u003c/h2\u003e \u003cp\u003eThe study had some limitations. First, the research was conducted at a single institution, and the specific environmental conditions, species composition, and established management protocols may not be representative of other aquarium facilities, thereby limiting the generalisability of our findings. Second, the aquarium exhibit setting inherently constrained experimental control; variables such as visitor presence and behaviour, ambient environmental conditions, water quality parameters, and social dynamics among individuals could not be fully controlled or systematically manipulated as they might be in a laboratory setting. Third, the sample sizes for some species were relatively small, limiting statistical power and reducing confidence in extrapolating these results to broader elasmobranch populations or other aquarium contexts. Finally, this study primarily focused on behavioural observations, and the inclusion of physiological stress indicators, such as cortisol concentrations, alongside additional behavioural metrics could have provided a more comprehensive assessment of welfare impacts.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eFuture Direction\u003c/h2\u003e \u003cp\u003eFuture research directions should include additional welfare measures, both physiological and physical, to complement behavioural observations, investigate individual variation in response to interventions, and long-term monitoring to assess the persistence of behavioural changes. Additionally, comparative studies across multiple institutions would strengthen the results and identify institutional factors that influence the success of interventions. It was worth noting how little has been done to study the welfare and behaviour of fish, including their responsiveness to training. This should be explored further.\u003c/p\u003e \u003cp\u003eAlthough the results of the spotted eagle ray interventions were not statistically significant, future research should explore alternative environmental modifications or extended implementation periods. Converting fixed location feeding to positive reinforcement training protocols may yield more substantial behavioural improvements, as suggested by the success observed in cownose ray training.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis research demonstrates that modifying feeding events and using reward-based behaviour modification procedures represent valuable tools for improving elasmobranch behaviour in mixed-species aquarium exhibits. The observed undesired and desired responses highlight the importance of tailored management approaches that account for the unique behavioural and physiological characteristics of different taxonomic groups. The success of reward-based interventions particularly validates elasmobranchs' cognitive capabilities and supports the expanded implementation of training programs in aquarium settings. These findings contribute to the growing body of evidence supporting evidence-based management practices that enhance animal welfare in managed care environments.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualisation, C.T., T.J.M., E.J.F.; methodology, C.T. and E.J.F.; data collection, C.T., C.K.Y.W. and L.K.L.; data analysis, C.T.; original draft, C.T.; reviewing and editing, T.J.M., E.J.F., C.K.Y.W. and L.K.L. All authors agreed to the submission of the final copy of the manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research received no external funding.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eConflicts of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study received ethics approval from Adelaide University\u0026rsquo;s Animal Ethics Committee (Ethics No.: S-2024-030). The Singapore Oceanarium Research Panel approved all procedures, which were in accordance with all applicable national laws and/or guidelines (Singapore).\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eData Availability Statement\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated for this study are available from the corresponding author upon request.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe want to thank Ramil Bentillo for designing and building the stretcher for the cownose rays. We also want to thank Nicholas Derbyshire, Raja S/O Murugesan, and Goh Zhe Zuan for their guidance and support. Lastly, we would like to thank all the team members at Large Habitats for their assistance with feeding and training the animals.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBassett, L., Buchanan-Smith, H. M., McKinley, J. \u0026amp; Smith, T. E. Effects of Training on Stress-Related Behavior of the Common Marmoset (\u003cem\u003eCallithrix jacchus\u003c/em\u003e) in Relation to Coping With Routine Husbandry Procedures. \u003cem\u003eJ. Appl. Anim. Welfare Sci.\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e (3), 221\u0026ndash;233. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1207/S15327604JAWS0603_07\u003c/span\u003e\u003cspan address=\"10.1207/S15327604JAWS0603_07\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBinding, S., Farmer, H., Krusin, L. \u0026amp; Cronin, K. Status of animal welfare research in zoos and aquariums: Where are we, where to next? \u003cem\u003eJ. Zoo Aquarium Res.\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e (3), 166\u0026ndash;174. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.19227/jzar.v8i3.505\u003c/span\u003e\u003cspan address=\"10.19227/jzar.v8i3.505\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoundey, A., Nikitins, F. \u0026amp; Fernandez, E. J. An examination of husbandry training in zoos: A scoping review. \u003cem\u003eAppl. Anim. Behav. Sci.\u003c/em\u003e \u003cb\u003e292\u003c/b\u003e, 106805. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.applanim.2025.106805\u003c/span\u003e\u003cspan address=\"10.1016/j.applanim.2025.106805\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrando, S. \u0026amp; Norman, M. Handling and Training of Wild Animals: Evidence and Ethics-Based Approaches and Best Practices in the Modern Zoo. \u003cem\u003eAnimals\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e (14), 2247 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown, C. \u0026amp; Schluessel, V. Smart sharks: a review of chondrichthyan cognition. \u003cem\u003eAnim. Cogn.\u003c/em\u003e \u003cb\u003e26\u003c/b\u003e (1), 175\u0026ndash;188. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10071-022-01708-3\u003c/span\u003e\u003cspan address=\"10.1007/s10071-022-01708-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarlstead, K., Paris, S. \u0026amp; Brown, J. L. Good keeper-elephant relationships in North American zoos are mutually beneficial to welfare. \u003cem\u003eAppl. Anim. Behav. Sci.\u003c/em\u003e \u003cb\u003e211\u003c/b\u003e, 103\u0026ndash;111. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.applanim.2018.11.003\u003c/span\u003e\u003cspan address=\"10.1016/j.applanim.2018.11.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCorwin, A. L. Training Fish and Aquatic Invertebrates for Husbandry and Medical Behaviors. \u003cem\u003eVeterinary Clin. North. America: Exotic Anim. Pract.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e (3), 455\u0026ndash;467. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cvex.2012.06.009\u003c/span\u003e\u003cspan address=\"10.1016/j.cvex.2012.06.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCosta, S., Neves, J., Tir\u0026aacute;, G. \u0026amp; Andrade, J. P. Predatory Responses and Feeding Behaviour of Three Elasmobranch Species in an Aquarium Setting. \u003cem\u003eJ. Zoological Bot. Gardens\u003c/em\u003e. \u003cb\u003e4\u003c/b\u003e (4), 775\u0026ndash;787. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/jzbg4040055\u003c/span\u003e\u003cspan address=\"10.3390/jzbg4040055\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eD\u0026rsquo;Cruze, N. et al. A Global Review of Animal\u0026ndash;Visitor Interactions in Modern Zoos and Aquariums and Their Implications for Wild Animal Welfare. \u003cem\u003eAnimals\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e (6), 332 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFernandez, E. J. Training as enrichment: A critical review. \u003cem\u003eAnim Welf.\u003c/em\u003e \u003cb\u003e31\u003c/b\u003e (1), 1\u0026ndash;12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7120/09627286.31.1.001\u003c/span\u003e\u003cspan address=\"10.7120/09627286.31.1.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFernandez, E. J. \u0026amp; Sherwen, S. L. \u003cem\u003eHuman-Animal Interactions in Zoos: Integrating Science and Practice\u003c/em\u003e (CABI, 2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeorgopoulou, D. G., Vouidaskis, C. \u0026amp; Papandroulakis, N. Swimming behavior as a potential metric to detect satiation levels of European seabass in marine cages [Original Research]. \u003cem\u003eFrontiers in Marine Science\u003c/em\u003e, \u003cem\u003eVolume 11\u0026ndash;2024\u003c/em\u003e. (2024). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fmars.2024.1350385\u003c/span\u003e\u003cspan address=\"10.3389/fmars.2024.1350385\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarris, M. C. Y., Frazier, H., Mayall, S., Frey, A. D. \u0026amp; Boyle, S. A. Novel Food-Based Enrichment Increases Captive Cownose Stingray (\u003cem\u003eRhinoptera bonasus\u003c/em\u003e) Engagement with Enrichment Item. \u003cem\u003eJ. Zoological Bot. Gardens\u003c/em\u003e. \u003cb\u003e5\u003c/b\u003e (4), 552\u0026ndash;562. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/jzbg5040037\u003c/span\u003e\u003cspan address=\"10.3390/jzbg5040037\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHeinrich, D. D. U., Pouca, V., Brown, C., Huveneers, C. \u0026amp; C., \u0026amp; Effects of reward magnitude and training frequency on the learning rates and memory retention of the Port Jackson shark \u003cem\u003eHeterodontus portusjacksoni\u003c/em\u003e. \u003cem\u003eAnim. Cogn.\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e (5), 939\u0026ndash;949. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10071-020-01402-2\u003c/span\u003e\u003cspan address=\"10.1007/s10071-020-01402-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuntingford, F. A. et al. Current issues in fish welfare. \u003cem\u003eJ. Fish Biol.\u003c/em\u003e \u003cb\u003e68\u003c/b\u003e (2), 332\u0026ndash;372. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.0022-1112.2006.001046.x\u003c/span\u003e\u003cspan address=\"10.1111/j.0022-1112.2006.001046.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJanse, M., Zimmerman, B., Geerlings, L., Brown, C. \u0026amp; Nagelkerke, L. A. J. Sustainable species management of the elasmobranch populations within European aquariums: a conservation challenge. \u003cem\u003eJ. Zoo Aquarium Res.\u003c/em\u003e \u003cb\u003e5\u003c/b\u003e (4), 172\u0026ndash;181. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.19227/jzar.v5i4.313\u003c/span\u003e\u003cspan address=\"10.19227/jzar.v5i4.313\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJanssen, J. D., Kidd, A., Ferreira, A. \u0026amp; Snowden, S. Training and conditioning of elasmobranch in aquaria. \u003cem\u003eThe elasmobranch husbandry manual II: recent advances in the care of sharks, rays and their relatives. Ohio Bological Survey, Columbus\u003c/em\u003e, 209\u0026ndash;222. (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLaterveer, M. \u003cem\u003eChanging the feeding regime in a multi-species open-ocean tank by means of target training pelagic sharks\u003c/em\u003e (Drum and Croaker, 2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLawrence, K., Sherwen, S. L. \u0026amp; Larsen, H. Natural Habitat Design for Zoo-Housed Elasmobranch and Teleost Fish Species Improves Behavioural Repertoire and Space Use in a Visitor Facing Exhibit. \u003cem\u003eAnimals\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e (10), 2979. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ani11102979\u003c/span\u003e\u003cspan address=\"10.3390/ani11102979\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin, G. Y., Ng, K. C. H. \u0026amp; Fernandez, E. J. Animal\u0026ndash;Visitor Interactions in Zoos and Aquariums: A Systematic Review. \u003cem\u003eAnimals\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(13), 1924. (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ani15131924\u003c/span\u003e\u003cspan address=\"10.3390/ani15131924\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eL\u0026oacute;pez-Olmeda, J., Noble, C. \u0026amp; S\u0026aacute;nchez-V\u0026aacute;zquez, F. Does feeding time affect fish welfare? \u003cem\u003eFish Physiol. Biochem.\u003c/em\u003e \u003cb\u003e38\u003c/b\u003e (1), 143\u0026ndash;152. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10695-011-9523-y\u003c/span\u003e\u003cspan address=\"10.1007/s10695-011-9523-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarranzino, A. The Use of Positive Reinforcement in Training Zebra Sharks (\u003cem\u003eStegostoma fasciatum\u003c/em\u003e). \u003cem\u003eJ. Appl. Anim. Welfare Sci.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e (3), 239\u0026ndash;253. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/10888705.2013.798555\u003c/span\u003e\u003cspan address=\"10.1080/10888705.2013.798555\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartins, C. I. M. et al. Behavioural indicators of welfare in farmed fish. \u003cem\u003eFish Physiol. Biochem.\u003c/em\u003e \u003cb\u003e38\u003c/b\u003e (1), 17\u0026ndash;41. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10695-011-9518-8\u003c/span\u003e\u003cspan address=\"10.1007/s10695-011-9518-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMelfi, V. Is training zoo animals enriching? \u003cem\u003eAppl. Anim. Behav. Sci.\u003c/em\u003e \u003cb\u003e147\u003c/b\u003e (3), 299\u0026ndash;305. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.applanim.2013.04.011\u003c/span\u003e\u003cspan address=\"10.1016/j.applanim.2013.04.011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMelfi, V., Ward, S., Pawson, C. \u0026amp; Hosey, G. Animal\u0026ndash;Staff Interactions. In Human-Animal Interactions in Zoos: Integrating Science and Practice (33\u0026ndash;45). CABI GB. (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMorimura, N. \u0026amp; Ueno, Y. Influences on the Feeding Behavior of Three Mammals in the Maruyama Zoo: Bears, Elephants, and Chimpanzees. \u003cem\u003eJ. Appl. Anim. Welfare Sci.\u003c/em\u003e \u003cb\u003e2\u003c/b\u003e (3), 169\u0026ndash;186. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1207/s15327604jaws0203_1\u003c/span\u003e\u003cspan address=\"10.1207/s15327604jaws0203_1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1999).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMunroe, S., Meyer, L. \u0026amp; Heithaus, M. R. Elasmobranch foraging strategies and tactics. In Biology of Sharks and Their Relatives (323\u0026ndash;355). CRC. (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMuraco, H. \u0026amp; Stamper, A. Training spotted eagle rays (\u003cem\u003eAetobatus narinari (Euphrasen)\u003c/em\u003e) to decrease aggressive behaviors towards divers. (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMushtaq, S. T. Aggression in aquatic environments and its relevance in aquaculture and conservation efforts. \u003cem\u003eDiscover Anim.\u003c/em\u003e \u003cb\u003e1\u003c/b\u003e (1), 28. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s44338-024-00026-x\u003c/span\u003e\u003cspan address=\"10.1007/s44338-024-00026-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePoling, A., Methot, L. L. \u0026amp; LeSage, M. G. \u003cem\u003eFundamentals of behavior analytic research\u003c/em\u003e (Springer Science \u0026amp; Business Media, 1995).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRogers, K. N. \u003cem\u003eExploring the welfare of a captive population of Cownose rays (Rhinoptera bonasus) through social and resource interactions\u003c/em\u003e [Masters Thesis, University of North Florida]. Florida, USA. (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSasko, D. E., Dean, M. N., Motta, P. J. \u0026amp; Hueter, R. E. Prey capture behavior and kinematics of the Atlantic cownose ray, \u003cem\u003eRhinoptera bonasus\u003c/em\u003e. \u003cem\u003eZoology\u003c/em\u003e \u003cb\u003e109\u003c/b\u003e (3), 171\u0026ndash;181. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.zool.2005.12.005\u003c/span\u003e\u003cspan address=\"10.1016/j.zool.2005.12.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSherwen, S. L. \u0026amp; Hemsworth, P. H. The Visitor Effect on Zoo Animals: Implications and Opportunities for Zoo Animal Welfare. \u003cem\u003eAnimals\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e (6), 366. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ani9060366\u003c/span\u003e\u003cspan address=\"10.3390/ani9060366\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWard, A. J. W., Webster, M. M. \u0026amp; Hart, P. J. B. Intraspecific food competition in fishes. \u003cem\u003eFish Fish.\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e (4), 231\u0026ndash;261. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1467-2979.2006.00224.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1467-2979.2006.00224.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWhitham, J. C. \u0026amp; Wielebnowski, N. New directions for zoo animal welfare science. \u003cem\u003eAppl. Anim. Behav. Sci.\u003c/em\u003e \u003cb\u003e147\u003c/b\u003e (3), 247\u0026ndash;260. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.applanim.2013.02.004\u003c/span\u003e\u003cspan address=\"10.1016/j.applanim.2013.02.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilliams, E., Hunton, V., Hosey, G. \u0026amp; Ward, S. J. The Impact of Visitors on Non-Primate Species in Zoos: A Quantitative Review. \u003cem\u003eAnimals\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e (7), 1178. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ani13071178\u003c/span\u003e\u003cspan address=\"10.3390/ani13071178\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\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":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"feeding management, positive reinforcement training, aquarium welfare, elasmobranchs","lastPublishedDoi":"10.21203/rs.3.rs-9208737/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9208737/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eManaging elasmobranchs in mixed-species aquarium exhibits presents unique welfare challenges. With many species facing conservation threats, optimising behaviour in managed care is critical. This study investigated the effects of modified feeding events and positive reinforcement training on sharks, rays, and bony fish in Singapore Oceanarium's Open Ocean Habitat. Experiment 1 employed a reversal design, transitioning from single daily feeding to three feedings a day while maintaining the total food quantity. Sharks showed a significant reduction in undesired responses; however, rays showed an increase in such behaviours. Bony fish showed minimal effects due to the manipulations. Experiment 2 targeted undesired responses in the rays through a target training intervention. Javanese cownose rays (\u003cem\u003eRhinoptera javanica\u003c/em\u003e) received positive reinforcement training using stretcher conditioning, while spotted eagle rays (\u003cem\u003eAetobatus narinari)\u003c/em\u003e received fixed location feeding. Training significantly reduced the cownose ray\u0026rsquo;s undesired responses. Eagle rays showed improvements, though not statistically significant. Results demonstrated that live-feeding modifications effectively reduced undesired responses in sharks but required additional reward-focused interventions for rays. Positive reinforcement training successfully addressed welfare challenges, providing practical management tools for aquarium professionals working with diverse elasmobranch collections.\u003c/p\u003e","manuscriptTitle":"The Age of Aquariums: Managing Elasmobranch Behaviour in a Mixed-Species Habitat","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-06 15:07:58","doi":"10.21203/rs.3.rs-9208737/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"70982990534123386608267246124111404333","date":"2026-05-02T16:01:50+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-01T14:54:12+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-31T16:42:46+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-25T10:52:44+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-25T10:52:41+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-03-24T08:02:09+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"744552d0-a6da-4ef6-a1d7-0e3d3818d02c","owner":[],"postedDate":"April 6th, 2026","published":true,"recentEditorialEvents":[{"type":"reviewerAgreed","content":"70982990534123386608267246124111404333","date":"2026-05-02T16:01:50+00:00","index":57,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":65701644,"name":"Biological sciences/Ecology"},{"id":65701645,"name":"Earth and environmental sciences/Ecology"},{"id":65701646,"name":"Earth and environmental sciences/Ocean sciences"},{"id":65701647,"name":"Biological sciences/Zoology"}],"tags":[],"updatedAt":"2026-04-06T15:07:58+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-06 15:07:58","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9208737","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9208737","identity":"rs-9208737","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}