An integrative perspective on fish health: Environmental and anthropogenic pathways affecting fish stress.

Multifactorial studies assessing the cumulative effects of natural and anthropogenic stressors on individual stress response are crucial to understand how organisms and populations cope with environmental change. We tested direct and indirect causal pathways through which environmental stressors affect the stress response of wild gilthead seabream in Mediterranean costal lagoons using an integrative PLS-PM approach. We integrated information on 10 environmental variables and 36 physiological variables into seven latent variables reflecting lagoons features and fish health. These variables concerned fish lipid reserves, somatic structure, inorganic contaminant loads, and individual trophic and stress response levels. This modelling approach allowed explaining 30 % of the variance within these 46 variables considered. More importantly, 54 % of fish stress response was explained by the dependent lagoon features, fish age, fish diet, fish reserve, fish structure and fish contaminant load latent variables included in our model. This integrative study sheds light on how individuals deal with contrasting environments and multiple ecological pressures.

Temperature-dependent metabolic consequences of food deprivation in the European sardine.

Aquatic ecosystems can exhibit seasonal variation in resource availability and animals have evolved to cope with the associated caloric restriction. During winter in the NW Mediterranean Sea, the European sardine Sardina pilchardus naturally experiences caloric restriction owing to a decrease in the diversity and quantity of plankton. However, ongoing global warming has had deleterious effects on plankton communities such that food shortages may occur throughout the year, especially under warm conditions in the summer. We investigated the interactive effects of temperature and food availability on sardine metabolism by continuously monitoring whole-animal respiration of groups of control (fed) and food-deprived sardines over a 60-day experiment in winter (12°C) or summer (20°C) conditions under natural photoperiod. In addition, we measured mitochondrial respiration of red muscle fibres, biometric variables and energy reserves of individuals sampled at 30 and 60 days. This revealed that winter food deprivation elicits energy saving mechanisms at whole animal and cellular levels by maintaining a low metabolism to preserve energy reserves, allowing high levels of survival. By contrast, despite energy saving mechanisms at the mitochondrial level, whole animal metabolic rate was high during food deprivation in summer, causing increased consumption of energy reserves at the muscular level and high mortality after 60 days. Furthermore, a 5-day re-feeding did not improve survival, and mortalities continued, suggesting that long-term food deprivation at high temperatures causes profound stress in sardines that potentially impairs nutrient absorption.

Metabarcoding confirms the opportunistic foraging behaviour of Atlantic bluefin tuna and reveals the importance of gelatinous prey.

Studies of the diet, feeding habits and trophic activity of top marine predators are essential for understanding their trophodynamics. The main direct method used for such studies thus far has been morphological inventories of stomach contents. This approach presents limitations such as missing gelatinous prey, which are usually digested too quickly to be detectable. Here, we analysed the stomachs of 48 Atlantic bluefin tuna (Thunnus thynnus, approximately 15 to 60 kg, including juveniles and adult fishes) collected from the Mediterranean Sea through the metabarcoding of two gene regions (cytochrome c oxidase subunit I (COI) and the ribosomal 18S-V1V2 region). The identified prey taxa and their relative read abundances (RRAs) estimated using COI results were in line with the findings of morphologically based inventories simultaneously performed on the same set of tuna samples. In both cases (and with the same rankings), the prey taxa included anchovy (Engraulis encrasicolus, here detected in more than 80% of samples, RRA = 43%), sardine (Sardina pilchardus, also approximately 80%, RRA = 30%), sprat (Sprattus sprattus, approximately 66%, RRA = 8%), mackerel (Scomber colias, approximately 44%, RRA = 7%) and cephalopods (approximately 15%, RRA = 1.4%). Another striking result was the detection, based on 18S (with which vertebrates were detected as the most abundant group, RRA = 61.6%), of a high prevalence and diversity of gelatinous organisms (RRA = 27.1%), including cnidarians (6.7%), salps (11.7%), and ctenophores (8.7%), the latter increasing with the size of the predator. These results thus support the hypothesis of the role of gelatinous prey in the diet of Atlantic bluefin tuna, suggesting that this species is even more generalist and opportunistic than previously thought. This study further confirms that DNA metabarcoding can be a powerful tool for assessing the diet and trophodynamics of top marine predators.

Effects of extraction method and storage of dry tissue on marine lipids and fatty acids.

Various protocols are currently used to study marine lipids, but there is a growing interest in working on dry samples that are easier to transport. However, reference protocols are still lacking for dry samples. In order to make recommendations on this use, lipid classes and fatty acids (FA) obtained from six analytical protocols using two different tissue states (dry vs wet) and three extraction methods (automat vs manual potter vs leaving the solvent to work on tissue) were compared. Three dry storage modes of tissue (freezer vs gas nitrogen vs dry room) during one and three months were also compared. These comparisons were made on seven marine species with different lipid profiles, including fishes, crustaceans and mollusks. Lipid classes and FA obtained from wet and dry tissues were similar, but they were affected by the extraction methods. Regardless of tissue state, "Leave to work" methods obtained the highest lipid quantities, followed by manual potter and automat methods (ca. 90% and 80% of "Leave to work" methods, respectively). Linear relationships allowed correction for lipid classes and FA concentrations obtained from different protocols. The repeatability of all protocols still needs to be improved, especially for fish species. Increasing the replicate number for each sample might be an indirect way to improve lipid quantification. Our results show that storing dry tissues in the freezer for more than one month was associated with a decrease in lipids, which is also observed for other storage methods. For qualitative studies of FA (expressed in %), a three-month storage of dry tissue in freezer did not affect the relative composition of species/tissues with a lipid content below 20% of dry weight.