Introduction
In the context of climate change, sea surface temperature has already increased by 0.04°C per year between 1985 and 2006 in the Mediterranean Sea and, according to projections by the Intergovernmental Panel on Climate Change, it could rise by 1 to 2°C in the Atlantic Ocean and by 2.2 to 3.4°C in the Mediterranean Sea by 2080. Such thermal elevation is likely to strongly affect sea bass populations and, consequently, aquaculture production. However, these effects may vary depending on the genetic origin of the populations: Atlantic (AT), Western Mediterranean (WM), and Eastern Mediterranean (EM), which have evolved under different environmental conditions.
In response to changes in their thermal environment, fish must adapt in order to limit the negative effects of these variations. A key component of this adaptation is robustness, defined as the ability to maintain production potential across a wide range of environmental conditions without compromising health and welfare (Knap, 2005). In this context, the stability of growth performance within a population exposed to different temperatures is an important criterion of robustness, since a fish exhibiting stable growth across temperatures will maintain production potential over a broad range of environments. This stability can be studied using the "LnVar" approach, an indicator that describes the variability of individual performance over time to assess organism sensitivity to external fluctuations. This indicator has been shown to exhibit genetic variability in several terrestrial and aquatic species , opening perspectives for selective breeding of more robust animals.
Materials and Methods
European seabass from the AT, WM, and EM populations were reared in mixed groups under three distinct thermal regimes representative of the seasonal variations in their areas of origin: a relatively cold regime for the Atlantic (rAT), an intermediate regime for the Western Mediterranean (rWM), and a warmer regime for the Eastern Mediterranean (rEM) (Crestel et al., 2025).
At an average body weight of 10 g, 5,148 fish were individually tagged, allowing individual monitoring of performance traits (weight, length, growth, etc.) over three years through biometric measurements taken every six weeks. For each individual, the stability of the studied traits, LnVar, was calculated as follows :
, where x������ is the deviation of measurement i relative to the cohort of individual j for the studied trait, and x����� is the mean deviation of individual j for that trait. Two additional indicators, skewness and autocorrelation, were calculated to complement this study (Berghof et al., 2019). These indicators were calculated using body weight (BW) of Daily Growth Coefficient (DGC) as base traits.
Individuals were genotyped using the Axiom��� 57K SNP DlabChip array in order to estimate the genetic parameters of the measured traits and their stability.
Results
The lnVar classified fish as more or less stable in response to temperature variations, but it does not allow discrimination between high- and low-performing fish. Autocorrelation and skewness helped to characterize the temporal dynamics of fluctuations when individuals were not stable. Heritability was 0.14 ± 0.03 for both BW lnVar and BW skewness , while it was 0.00 ± 0.00 for BW autocorrelation. The genetic correlations between BW lnVar and BW at two years was 0.72 ± 0.15 and -0.29 ± 0.14 between BW lnVar and BW skewness. With DGC as base trait, h2 was 0.34 ± 0.03 for lnVar, 0.15 ± 0.03 for skewness and 0.0±0.0Y for BW autocorrelation. Other analysis are currently running taking into accounts other effects.
Discussion
These combined indicators help better understand fish growth profiles under thermal variation. lnVar appears to be the most relevant indicator, as it both effectively discriminates fish according to their stability and can be used in selective breedin, given its moderate heritability, especially with DGC as the base trait.
This study, currently in progress, will also deterrmine how many measurements are required to obtain a reliable estimate of individual lnVar. Such information is important for breeding programs, to assess how lnVar can be meaningfully estimated At a later stage, we will seek to establish correlations of lnVar with production traits and operational welfare indicators.
Acknowledgment
This study was conducted as part of the ANR FishNess project (ANR-21-CE20-0043).
References
T. V. L. Berghof, M. Poppe, and H. A. Mulder, 'Opportunities to Improve Resilience in Animal Breeding Programs', Front. Genet., vol. 9, Jan. 2019, doi: 10.3389/fgene.2018.00692.
D. Crestel et al., 'Do European Seabass Larvae Grow Better in Their Natural Temperature Regime?', Evolutionary Applications, vol. 18, no. 2, p. e70083, Feb. 2025, doi: 10.1111/eva.70083.
P. W. Knap, 'Breeding robust pigs', Aust. J. Exp. Agric., vol. 45, no. 8, pp. 763–773, Aug. 2005, doi: 10.1071/EA05041.