Spatial distribution and activity patterns as welfare indicators in response to water quality changes in European sea bass, Dicentrarchus labrax

Highlights

• Groups of fish were exposed to acute stressors (Hypoxia, Hyperoxia, High ammonia).

• Behavior was recorded both during stress condition and following stress.

• All stress conditions induce a decrease in activity, thigmotaxis and group cohesion.

• Novel tank diving test was validated for welfare assessment in European sea bass.

• Activity, thigmotaxis and group cohesion are sensitive welfare indicators.

Abstract

In aquaculture, fish are exposed to unavoidable stressors that can be detrimental for their health and welfare. However, welfare in farmed fish can be difficult to assess, and, so far, no standardized test has been universally accepted as a welfare indicator. This work contributes to the establishment of behavioural welfare indicators in a marine teleost in response to different water quality acute stressors. Groups of ten fish were exposed to high Total Ammonia Nitrogen concentration (High TAN, 18 mg.L^-1), Hyperoxia (200 % O₂ saturation), Hypoxia (20 % O₂ saturation), or control water quality (100% O₂ saturation and TAN < 2.5 mg.L^-1) over 1 hour. Fish were then transferred in a novel environment for a group behaviour test under the same water quality conditions over 2 hours. Videos were recorded to assess thigmotaxis, activity and group cohesion. After this challenge, plasma cortisol concentration was measured in a subsample, while individual behavioural response was measured in the other fish using novel tank diving test. Prior to this study, the novel tank diving test was validated as a behavioural challenge indicative of anxiety state, by using nicotine as anxiolytic drug. Overall, all stress conditions induced a decrease in activity and thigmotaxis and changes in group cohesion while only fish exposed to Hypoxia and High TAN conditions displayed elevated plasma cortisol concentrations. In post-stress condition, activity was still affected but normal behaviour was recovered within the 25 minutes of the test duration. Our work suggests that the activity, thigmotaxis and group cohesion are good behavioural indicators of exposure to degraded water quality, and could be used as standardized measures to assess fish welfare.

Introduction

Fish production has expanded importantly during the last decades, both because of the world’s diminishing natural wild resources and the increase in demand for fish products (FAO, 2018). Aquaculture represented 53 % of the total fish production (including non-food uses) in 2016 (FAO, 2018) and is now recognized as a major food production industry. Thus, as well as for terrestrial farming industry, concerns about sustainability, environmental issues and animal welfare in aquaculture are increasing (Conte, 2004; Ashley, 2007; Martins et al., 2010; Martins et al., 2012; Hixson, 2014; FAO, 2018; Lembo et al., 2019). It is common that under aquaculture conditions, and in every fish husbandry system, variations of water quality variables such as temperature, pH, oxygen (O₂), carbon dioxide (CO₂) or Total Ammonia Nitrogen (TAN) concentrations occur. Such variations when they reach a certain threshold, depending on the species preferendum, could be considered as stress factors (stressor) and therefore deleterious for fish health and welfare. The prolonged exposure to stressors, such as degraded water quality may mobilize fish energy for coping with the stressor hereby decreasing the available energy allocated to growth and reproduction, or directly causes death if the magnitude of stress is too high (Barton, 2002; Sneddon et al., 2016). Therefore, it can finally affect fish production and has economic consequences for farmers (Conte, 2004; Lembo et al., 2019).

Exposure to stress factors triggers a cascade of biological events within an organism to cope with these factors. In fish, the hypothalamo-pituitary-interrenal axis (HPI) is involved in the production and release of cortisol into circulation acting as an activator of the physiological and behavioural responses (Sumpter, 1997; Sadoul and Vijayan, 2016; Schreck and Tort, 2016). Among the previously cited variables, oxygen and ammonia concentrations are known to activate the HPI axis when they vary, leading to the stimulation of cortisol release and the triggering of behavioural adaptive responses (Knoph and Olsen, 1994; van Raaij et al., 1996; Espmark and Baeverfjord, 2009).

Thus, behavioural measurements have proven to be sensitive indicators of the complex existing biochemical and physiological changes that occur in response to stress (Schreck, 1990; Scherer, 1992; Schreck et al., 1997; Martins et al., 2012). Behaviours, such as changes in food-anticipatory activity, feed intake, ventilation rate, individual and group swimming activities are commonly used as welfare indicators (Huntingford et al., 2006; Martins et al., 2012; Huntingford and Kadri, 2014; Carbonara et al., 2015; Carbonara et al., 2019). Group swimming behaviour is defined as the spatial distribution and swimming activity of the group of fish held within an aquaculture production unit and it covers shoal structure, the horizontal and vertical distribution of the group, their swimming speed and direction (Martins et al., 2012). For instance, exposure to negative stimuli, such as poor water quality, is known to lead to rapid escape movements (Stien et al., 2007; Bratland et al., 2010) or to alter group cohesion (Domenici et al., 2002; Espmark and Baeverfjord, 2009; Sadoul et al., 2014; Sadoul et al., 2017). Thus, group swimming behaviour appears to be a sensitive welfare indicator even if it is still lacking calibration efforts to be precisely translated into an operational welfare indicator; nevertheless some examples exist (Papandroulakis et al., 2014; Pettersen et al., 2014). Moreover, the appraisal of negative or positive stimuli and, hence, the psychological dimension of stress as defined for fish by Galhardo and Oliveira (2009) is seldom tackled in welfare research. There exists however a complementary measure which is the individual behavioural responses to novel environment and in particular the novel tank diving test which is worldwide used along with the measure of stereotypies, such as thigmotaxis to assess anxiety in zebrafish (Danio rerio) in ecotoxicology and pharmacology research (Levin et al., 2007; Egan et al., 2009; Vignet et al., 2014; Macaulay et al., 2015; Alfonso et al., 2019a). In further details, the novel tank diving test was validated as a tool for evaluating anxiety by using drugs, such as nicotine. Short exposure to nicotine is known to reduce anxiety in fish, through its action on nicotinic acetylcholine receptors as demonstrated by the use of specific inhibitors (Levin et al., 2007; Bencan & Levin, 2008). In the context of novel tank diving test, nicotine-exposure (bathing) has been shown to be anxiolytic by triggering change in fish space utilization, such as higher time spent in the top area of the novel tank which translate a relief from bottom dwelling behaviour that fish would express under predator threat for example. The novel tank diving test could thus be a helpful non-invasive tool to monitor farmed fish anxiety state post stress exposure hereby assessing psychological stress and contributing to the assessment of positive or negative emotions and, hence better welfare state determination.

Overall, the objectives of the present study were to further contribute to the establishment of behavioural welfare indicators including the psychological dimension of stress in a model marine teleost in response to different water quality stressors. Firstly, the novel tank diving test outcome was validated as a behavioural indicator of anxiety in European sea bass Dicentrarchus labrax, using nicotine as an anxiolytic reference drug. Secondly, behavioural responses of fish group in response to a novel environment under acute and severe water quality deterioration, including Total Ammonia Nitrogen (High TAN) increase (18 mg.L^-1), Hyperoxia (200 % O₂ saturation) and Hypoxia (20 % O₂ saturation) were evaluated along with cortisol measurement. Finally, individual behaviour expressed following the same water quality exposures were assessed using the novel tank diving test translated from ecotoxicology studies.