The Importance of Methods in Assessing Conservation Status of Abundant Fish Species through Genetic Diversity Estimates

This study compares the accuracy of two genomic approaches in estimating genetic diversity levels, which could be useful for informing species conservation assessments of abundant, exploited fish species. The first approach (SNP-calling-based) is the commonly used pipeline of SNP calling followed by SNP filtering at a determined Minor Allele Frequency (MAF). The second approach (genotype-likelihood-based) does not perform SNP calling but estimates the Site Spectrum Frequency (SFS) based on alignment quality and sample size. The results show up to two-fold differences in the magnitude of the estimated nucleotide diversities among the analyzed datasets. The SNP-calling-based approach produces overestimates when missing data are considered in the analysis and shows pronounced deviations of the SFS towards high-frequency SNPs when filtering by MAF > 5%. The genotype likelihood-based approach showed that nucleotide diversity estimates significantly deviated from neutral expectations, as expected based on the known history of the case-study fish population analyzed here, regardless of whether missing data were considered. In contrast, the SNP-calling-based approach only shows this expected difference when no missing data are included and no MAF filtering is performed. Overall, the results indicate that using the SNP-calling-based approach may hide the effects of population size declines in abundant exploited fish species, while genotype-likelihood-based estimates of nucleotide diversity can effectively contribute to informing conservation assessments.