Evolution in the Model Genus Antirrhinum Based on Phylogenomics of Topotypic Material

Researchers in phylogenetic systematics typically choose a few individual representatives of every species for sequencing based on convenience (neighboring populations, herbarium specimens, samples provided by experts, garden plants). However, few studies are based on original material, type material or topotypic material (living specimens from the locality where the type material was collected). The use of type or topotypic material in phylogenetic studies is paramount particularly when taxonomy is complex, such as that of Antirrhinum (Plantaginaceae). In this paper, we used topotypic materials of Antirrhinum at the species level (34 species proposed by previous authors), 87 specimens representing the species distributions and >50,000 informative nucleotide characters (from ∼4,000 loci) generated by the genotyping-by-sequencing (GBS) technique: (i) to test two explicit taxonomic hypotheses widely followed by local taxonomic treatments; (ii) to robustly estimate phylogenetic relationships; (iii) to investigate the evolution of key morphological characters and biogeographic centers of differentiation. Two GBS phylogenies based on two datasets (87 localities and 34 topotypic specimens) revealed that: (1) Sutton’s (1988) taxonomic account is the most congruent with phylogenetic results, whereas division of Antirrhinum into three major clades disagrees with Rothmaler’s (1956) infrageneric classification; (2) monophyly of populations currently included in the same species is primarily supported; (3) the historically recognized Antirrhinum majus group is not monophyletic; (4) sister-group relationships are robust for eight species pairs; (5) the evolutionary radiation of 26 species since the Pliocene is underpinned given a high rate of diversification (0.54 spp. Myr(–1)); (6) a geographic pattern of speciation is reconstructed, with northern Iberia as the center of early diversification followed by more recent speciation in southeastern Iberia; and (7) multiple acquisitions of key taxonomic characters in the course of Antirrhinum diversification are strongly supported, with no evidence of hybridization between major clades. Our results also suggest incipient speciation in some geographic areas and point to future avenues of research in evolution and systematics of Antirrhinum.