Otimização da amostragem para estudos populacionais com locos de microssatélites em espécies Neotropicais de Drosophila (Diptera, Drosophilidae)

Abstract

The definition of the sample size of individuals from natural populations that can reliably represent genetic variability is an important factor for the study of population genetic structure. The low number of specimens sampled can lead to inaccurate conclusions, while a very high number can be a challenge to collect some groups of animals, in addition of being very laborious and could also represent a waste of financial resources. In this way, to determine the appropriate sample size that is representative of the natural variation, with the lowest possible cost and laboratory effort, becomes relevant. The objective of this work was to determine the sample size that represents the genetic variability of microsatellite loci for population studies of Drosophila species from the Neotropical region, with optimization of resources and collection and laboratory efforts. We used 100 individuals from natural populations genotyped for microsatellite loci for each of four Drosophila species: D. ornatifrons, D. antonietae, belonging to the Drosophila subgenus, and D. prosaltans and D. sturtevanti, of the Sophophora subgenus. From the original matrices, 50 subsamples were randomly obtained for the following population sizes (N): 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75 and 95 individuals. For all 50 replicates of each N, the percentage of detection of all non-rare alleles (frequency ≥ 5%) from the original sample, expected mean heterozygosity (He), and the genetic differentiation (Fst) to the original 100 individuals matrix were evaluated. original. The ideal N for each species was established using a non-parametric analysis of the non-significant values of He and Fst paired between the different population sizes, and for the detection of the non-rare alleles, the presence of these in at least 95% of the 50 subsamples in at least 70% of the loci was considered. After determining the ideal N, the rare alleles (frequency < 5%) of the original matrix were replaced by other genotypes with non-rare alleles (also from natural populations), and from this matrix without rare alleles, 50 subsamples with the ideal N were randomly selected. The genetic compositions of the 50 replicates of ideal N with rare alleles were compared with those of ideal N without rare alleles, using the STRUCTURE software. According to the established parameters, the sample sizes for D. ornatifrons, N = 25, D. antonietae, N = 30, D. prosaltans and D. sturtevanti, N = 35, were considered ideal for the analysis of the genetic structure. The presence of rare alleles produced inconspicuous differences in the distribution of the genetic clusters of individuals, obtained in the STRUCTURE software, within the grouping classes proposed. Species of the Drosophila subgenus probably presented lower ideal N due to lower genetic diversity, as a result of the specificity of their diets, more protein in the case of D. ornatifrons and cactophily for D. antonietae. On the other hand, the generalism of the Sophophora subgenus species, D. prosaltans and D. sturtevanti, results in greater genetic diversity and, consequently, greater ideal N for population analyzes using microsatellite loci. The results of this work also showed that the same ideal N can be used in case of microsatellite loci transferred between phylogenetically close species, as was the case of the loci described for D. sturtevanti and transferred to D. prosaltans.