ИСТИНА

Lotus includes ca. 130 species of herbs, semishrubs and shrubs native to the Old World. Previous nrITS-based phylogenies were incongruent with all taxonomic systems of the genus. Members of the former genus Dorycnium were placed near species of L. sect. Lotus. The aims of the present study was to conduct a global phylogenetic study of Lotus using both plastid and nuclear markers with special attention to relationships between Lotus and Dorycnium. Materials: widely sampled genus Lotus (incl. Dorycnium and Tetragonolobus): 167 accessions ITS 1-2 nrDNA, 74 accessions 5’ETS nrDNA, 65 accessions plastid set (psbA-trnH, rps16 intron, trnL intron, trnL-F); outgroup including species of Cytisopsis, Hammatolobium and Tripodion. Eight nuclear SSR loci were studied in the L. corniculatus complex. Methods: Bayesian (MrBayes 3.2.2) and ML (MEGA X) phylogenetic analyses, S-DIVA analysis. Genetic structure of L. corniculatus complex (Structure, TCS and Arlequin programs). Results and Conclusions: I. Sequences variability (total length/variable sites/parsimony informative sites): ITS: 680/280/212; ETS: 589/254/212; plastid set: 2378/305/153. Among plastid markers, trnL-F region plays important role in tracking of Lotus evolution, especially due to the presence of long indels. Lotus schimperi has 50 bp insertion in trnL-F IGS, a group L. creticus - L. pseudocreticus has 91 bp deletion in trnL intron, and Lotus corniculatus complex is characterized by 73 bp repeats covering parts of trnF and trnL-F IGS. II. Taxonomy and Biogeography: Southern lineage of Lotus includes the majority of sections and supported by most markers. Northern lineage includes L. sections Lotus, Dorycnium and Bonjeanea (the last two were members of the former gen. Dorycnium sensu Lassen) and supported by plastid data only. Phylogenetic relationships within Northern lineage are the most controversial and its evolution is essentially homoplastic. Sections in traditional view are not supported by most markers, except for Bayesian analysis of ETS, which confirms L. sect. Lotus and gen. Dorycnium sensu Lassen. S-DIVA analysis of the ETS data set suggests that plants of Europe or N. Asia (area A) were ancestral for two Northern lineages and plants of Africa or S. Asia (area B) - for Southern lineage of Lotus. A group corresponding to gen. Dorycnium sensu Lassen may have ancestral forms in area A. Sect. Bonjeanea contains three unrelated species, combining characters of Lotus and Dorycnium. L.rectus is a widely distributed Mediterranean species, and L.strictus occurs from E.Mediterranean to W.China. L.hirsutus demonstrates a tendency towards subdivision into west and east lineages and close connections with L. sect. Dorycnium, that is inconsistent with morphology. Sect. Dorycnium is not monophyletic by all markers. L. dorycnium complex is widely distributed and geographically differentiated in Mediterranean. It is supported by ITS marker, but reveals introgression with L. hirsutus by plastid data. L.graecus from East Mediterranean and Black sea region is a monophyletic, geographically non-structured species, with two closely related Turkish endemics. Sect. Lotus apparently has ancestral forms in area A. It is polyphyletic by ITS and plastid data, but its core group, L. corniculatus complex, is well supported by all markers. L. conimbricensis and a cryptic species L. carpetanus form sister group of the complex. We suppose the European origin of L.corniculatus complex and its geographic expansion parallel to the changes in cp haplotypes from A group in W. Europe to D group in C. Asia. Within the complex, a correlation between population genetic structure and species reproductive strategy was found. The study is supported by grant 19-04-00883 from the Russian Foundation for Basic Research.