Do environmental conditions (temperature and food composition) affect otolith shape during fish early-juvenile phase? An experimental approach applied to European Seabass (Dicentrarchus labrax)

Otolith shape is an efficient tool for fish stock discrimination. Therefore, it is crucial to identify the environmental parameters that could influence otolith morphogenesis and growth. Current global ocean warming negatively affects the production of polyunsaturated fatty acids (EPA and DHA) by phytoplankton and thus their availability in marine food webs. As EPA and DHA are essential nutrients for most fish species, their deficiency could affect many aspects of fish physiology, notably otolith morphogenesis. The goal of this study was to assess experimentally the combined influences of temperature (T) and dietary polyunsaturated fatty acid (PUFA) content on European seabass (Dicentrarchus labrax) otolith morphogenesis during early life stages. 300 individuals were reared from 94 to 200 days post hatch (dph) at 15 or 20 °C and fed a high (1.65% PUFA) or low (0.73% PUFA) nutrient diet. To analyze the temperature effect, Growing Degree Days (GDD) were used. Otolith morphometric measurements, shape factors and normalized Elliptical Fourier coefficients describing outline shape were extracted using image analysis. The directional asymmetry (DA) between left and right otolith shapes was not significantly correlated to the GDD or food nutrient level. The effects of T and PUFA on otolith shape were estimated using four linear mixed effects models and, for all descriptors of otolith outline, the best model included GDD, PUFA and T and their interaction as predictors. This temperature effect was described by GDD value and the interaction between GDD and T. Consequently, temperature has an effect on the speed but also on the morphogenesis trajectory of the otolith. Conversely, the concentration of PUFA in the diet had no significant effect on otolith shape. The otolith shape was first modified by a decrease in the width/length ratio during the first period of experiment and by the distance between rostrum and antirostrum during the second period of experiment. Our results i