Introduction
T he most important marine fish species farmed in the Mediterranean are European seabass (Dicentrarchus labrax) and gilthead seabream ( Sparus aurata) and for both species, selective breeding programmes have been initiated in recent years. In these programmes, the control of inbreeding is crucial to achieve sustainable production. Monitoring in breeding is also important when dealing with wild populations.
Recently , a combined ~ 60K SNP array for both species has been developed (Peñaloza et al. 2021). The use of this tool ( the MedFish SNP array ) allows to obtain more accurate estimates of inbreeding coefficients than those obtained using pedigree information and also enables the evaluation of inbreeding patterns across the genome. Villanueva et al . (2022) compared inbreeding of wild and farmed populations for both species using the aforementioned array but their analysis was limited to avera ge genome values. This study aims to compare patterns of inbreeding throughout the genome of wild (W) and farmed (F ) populations for European seabass and gilthead seabream.
Material and Methods
S amples of both species we re the same as those used for the SNP array development and were collected from W and F populations located from East to West Mediterranean. Three Atlantic populations of seabream were also sampled. For seabream, SNP genotypes were available for 462 fish from 14 W and 12 F populations . For seabass, SNP genotypes were available for 516 fish from 9 W and 15 F populations. After quality control, the total number of SNPs was 24,548 and 21,797 for seabream and seabass, respectively.
The p roportion of genome-wide homozygosity and the molecular inbreeding coefficient were obtained using the software PLINK (Purcell et al., 2007) and an R script. The measure of molecular inbreeding (Fmol ) used was that based on deviations from Hardy-Weinberg proportions (Li and Horvitz, 1953 ). Patterns of genomic homozygosity and inbreeding were calculated using a sliding window approach. For both species, the length of the windows was ~ 1 Mb and they were moved one SNP at a time. For each window, homozygosity and inbreeding were estimated by averaging the values for all SNPs lying in that window. Afterwards, values were averaged across individuals.
Results
For seabream, patterns of observed homozygosity were very similar across the genom e in both W and F populations (Fig. 1) . Also, differences between W and F populations were negligible. V alues for homozygosity ranged from 0.485 to 0.725 . In general, seabass populations showed higher variation among them than seabream populations. In fact , some distinct patterns were observed in seabass populations (Fig. 1). For instance, t he Moroccan wild population (purple line in Fig. 1C) showed high peaks of inbreeding across the genome while one of the Greek farmed populations, exhibited lower levels of homozygosity (green line in Fig. 1D) . Values for homozygosity ranged from 0.415 to 0.880. Values for Fmol ranged from −0.386 to 0.193 in seabream and from −0.463 to 0.366 in seabass. Negative values appear because current population frequencies were used when computing Fmol and then its values are centered around zero.
Discussion
This study shows that differences between W and F populations in terms of inbreeding are small in both species. Results are in agreement with previous population structure results where W seabream populations showed lower levels of differentiation than W seabass populations and where the Moroccan seabass population presented a higher genetic differentiation when compared to other populations (Villanueva et al. 2022) . Here, the measures used to evaluate inbreeding have been the observed homozygosity and Fmol . Many other measures have been proposed, but they can lead to inconsistent results in terms of loss or gain of genetic variability ( Villanueva et al. 2021). In conclusion, in most cases the selection pressure exerted on farmed populations for a number of generations (7-8 for the older breeding programmes) did not drastically increase homozygosity.
Acknowledgements
This research was funded by MCIN/ AEI /10.13039/501100011033 (Project PID2020- 114426GB-C22).
References
Li CC, Horvitz DG (1953). Some methods of estimating the inbreeding coefficient. Am. J. Hum. Genet. 5: 107-117.
Peñaloza C, et al . (2021). Development and testing of a combined species SNP array for the European seabass (Dicentrarchus labrax) and gilthead seabream ( Sparus aurata). Genomics 113: 2096-2107.
Purcell S, et al. (2007). PLINK: A t ool s et for whole-g enome a ssociation and population-b ased l inkage analyses Am. J. Hum. Genet. 81: 559-75.
Villanueva B, et al . (2021). The value of genomic relationship matrices to estimate levels of inbreeding. Genet. Sel. Evol. 53: 42.
Villanueva B, et al . (2022). Population structure and genetic variability in wild and farmed Mediterranean populations of gilthead seabream and European seabass inferred from a 60K combined species SNP array. Aquac. Rep. 24: 101145.