Tef (Eragrostis tef; Poaceae) is an allotetraploid (2n = 4x = 40) cereal crop whose origin within the large genus Eragrostis is unknown. Previous studies have suggested a total of 14 wild Eragrostis species as potential progenitors. Phylogenetic analysis of sequence data from the nuclear gene waxy and the plastid locus rps16 strongly supports the widely held hypothesis of a close relationship between tef and E. pilosa, a wild allotetraploid. Eragrostis heteromera, another previously proposed progenitor, is shown by the waxy data to be a close relative of one of the tef genomes. Other putative progenitors included in the taxon sample are not supported as closely related to tef. Plastid sequences from five varieties of tef and four E. pilosa accessions are identical and therefore are uninformative with respect to the question of multiple origins of these polyploids. The waxy phylogeny also resolves the relationships among other allopolyploids, supporting a close relationship between the morphologically similar allotetraploids E. macilenta, E. minor, and E. mexicana. Eragrostis cilianensis, another morphologically similar allopolyploid, appears to have shared one diploid progenitor with these species but derived its other genome from an unrelated diploid.

No abstract

No abstract

Technological advances have allowed phylogenomic studies of plants, such as full chloroplast genome (plastome) analysis, to become increasingly popular and economically feasible. Although next–generation short–read sequencing allows for full plastomes to be sequenced relatively rapidly, it requires additional attention using software to assemble these reads into comprehensive sequences. Here we compare the use of three de novo assemblers combined with three contig assembly methods. Seven plastome sequences were analyzed. Three of these were Sanger–sequenced. The other four were assembled from short, single–end read files generated from next–generation libraries. These plastomes represented a total of six grass species (Poaceae), one of which was sequenced in duplicate by the two methods to allow direct comparisons for accuracy. Enumeration of missing sequence and ambiguities allowed for assessments of completeness and efficiency. All methods that used de Bruijn–based de novo assemblers were shown to produce assemblies comparable to the Sanger–sequenced plastomes but were not equally efficient. Contig assembly methods that utilized automatable and repeatable processes were generally more efficient and advantageous when applied to larger scale projects with many full plastomes. However, contig assembly methods that were less automatable and required more manual attention did show utility in determining plastomes with lower read depth that were not able to be assembled when automatable procedures were implemented. Although the methods here were used exclusively to generate grass plastomes, these could be applied to other taxonomic groups if previously sequenced plastomes were available. In addition to comparing sequencing methods, a supplemental guide for short–read plastome assembly and applicable scripts were generated for this study.