Genetic species concept as applied to multilocus nuclear data

Abstract

Using the example of multiple species groups and species complexes, it is empirically shown that the resolving power of a set of 10–20 nuclear loci is insufficient for the study of their genetic differentiation. For the adequate analysis of phylogenetic relationships (including reticulation events) and assessment of divergence levels within such groups of closely related species, it is necessary to multiply the number of nuclear loci and transit to high-throughput next-generation deep sequencing. However, the question of the genomic coverage required for the purposes of such study remains open. In other words, exactly how many genes in a set are needed to measure the genetic distance resolving the relationships between branches on the phylogenetic tree reconstruction of a complex of closely related species? We chose protein-coding sequences as a standard set of markers. Distances for 160–180 genes with a combined length of 270 000–300 000 bp were used to estimate the divergence levels of closely related mammalian species. In most cases, the interspecific distances are within the 0.15–0.75% range (median of 0.33%). The range of distances between semi-species is 0.12–0.28% (median of 0.14%). Intraspecific distances are always lower than 0.11%. Rodents show higher values of interspecific distances, 0.25–2.3% (median of 0.72%); distances between half-species range from 0.1 to 0.35%. To determine the number of nuclear loci and their combined length sufficient for the calculation of a genetic distance, we use simulations based on a model that included the following parameters: the average rate of gene evolution, its dispersion, and the level of polymorphism in the modern and ancestral populations. We performed a preliminary analysis of the distribution of loci evolution rates among mammalian taxa based on the data on ~50 thousand nuclear markers. It is shown that a relative error of 10–15%, comparable to the same value for mitochondrial distances between close species based on individual genes (approximately 1000 bp in length), is achieved using approximately 100 loci of 300 bp in length. Based on these data, we propose the following working hypothesis: the threshold of interspecific/intraspecific genetic distance calculated on the basis of the exons most frequently used in mammalian molecular phylogenetics is approximately 0.15%. This hypothesis assumes a species rank for the forms in statu nascendi. It should be noted that not all “good” species have significant genetic distances, since in some cases reproductive barriers form faster than substitutions accumulate in the genome.