Genetic structure, phylogeography, and migration routes of Bouteloua gracilis (Kunth) Lag. ex Griffiths (Poaceae:Chloridoideae)

Abstract

Blue grama grass (Bouteloua gracilis) populations are found in widely variable environments, tolerating drought, alkaline soils and different levels of grazing. Many ploidy levels have been reported for this species that is also considered to be phenotypically plastic and morphologically variable. Recently a decline in its cover and frequency in the North American shortgrass steppe and central Mexico has been reported although much about the biology of the species is unknown, including genetic diversity throughout its distribution. Genetic and phylogeographic structure and phylogenetic relationships among B. gracilis were estimated employing next generation sequencing of a high number of SNPs and loci. Population genetics and Structure analyses were performed. We compared the marginal likelihoods of different migration models using MIGRATE and obtained the best population model of migration for our data. Demographic expansion of B. gracilis was observed graphically with a mismatch distribution obtained in DNAsp. Bayesian and Maximum Likelihood methods were used to resolve phylogenetic relationships among B. gracilis and its closely related species as well as within B. gracilis populations. B. gracilis is sister to the B. chasei and B. herrera arrietae clade. Among the populations of the species two highly supported clades were resolved, grouping samples from Mexico and USA respectively. Allele frequencies determined three population clusters: CUSA from the Great Plains, MEX from central and southern Mexico, and WUSA-NMEX from northern Mexico and the western mountainous region of USA, the latter of which contains an allele admixture of the other two clusters. The haplotype network revealed an ancestral haplotype originating in Mexico, from which the rest of the haplotypes diversified to the north. Both evidence of gene flow and isolation among populations was observed. Genetic clusters are not genetically structured and variation is higher among populations. The genetic and morphological data do not support recognition of ecotypes or infraespecific taxa. However, the Great Plains populations are least diverse, making them most vulnerable to environmental change.