Multiple genotypes of Phelipanche ramosa indicate repeated introductions to the Americas: Sequence alignments and phylogenetic trees

Premise: Phelipanche ramosa is an economically damaging parasitic plant that has been reported in North America since the late 1800s. While this species comprises a variety of genetically distinct host races in its native range, the genetic composition of adventive populations in the New World remains unexplored. On the basis of morphological and ecological variation, some have suggested that the closely related P. nana may also be present.

Methods: Genome skimming was used to assess the relationships of 30 populations of Phelipanche spanning the geographic and host ranges in North and South America, plus one P. nana reference population from Lebanon.

Results: Phylogenetic analysis indicated four distinct genetic groups, though plastome and nrDNA data supported conflicting signals of relationships among them. First, specimens from Chilean tomato fields were nearly indistinguishable genetically from the reference P. nana. Second, a pair of samples from Virginia showed similar nrDNA as the first group, but divergent plastomes. The remaining 24 samples sorted into two groups, one which parasitizes cultivated plants, especially tomato, and the other on roadside weeds in different parts of the United States.

Conclusions: The geographic and ecological cohesiveness of four distinct genetic groups supports a hypothesis of multiple introductions to the Americas, presumably from Eurasia, followed by little to no subsequent gene flow among them. However, such groups do not align with existing morphological or ecological species concepts for P. ramosa and P. nana. In practice, threat assessment of Phelipanche populations to agricultural settings should be evaluated regionally given the phylogeographic and ecological heterogeneity.

Methods
Thirty-one specimens of Phelipanche were collected across the known range of the species. For 26 of these, dried flowers or an inflorescence tip from single individuals were ground, and DNA extracted using a DNeasy Plant Pro kit. DNA extractions of the remaining five, all from Texas, were shared by Chris Randle (Sam Houston State University, Texas, USA). Genomic DNA was enzymatically fragmented and sequenced using 2 x 150 paired end Illumina sequencing.

From each set of reads, the nuclear ribosomal repeat and plastome were assembled de novo using GetOrganelle v.1.7.4.1. When multiple contigs were returned instead of one complete sequence, these were joined by comparison with congeneric reference sequences. Multiple sequence alignments were generated using MAFFT v.7.505, and phylogenetic analysis of the respective nrDNA and plastome alignments were performed using RAXML-NG v.1.2.2.

ITS and trnLF alignments were generated by extracting the respective regions from the nrDNA and plastome alignments and adding existing sequences from three sources: the curated alignments of Piwowarczyk et al. (2021), 22 vouchered samples from cultivated tomato fields in California (GenBank # OR690545–OR690569; voucher specimens at CDA), and 6 specimens from tobacco fields in Bulgaria (GenBank # MK024283–MK024290). Phylogenetic trees were generated as above.