ABSTRACT
The family Arecaceae is distributed throughout tropical and subtropical regions of the world. Among the five subfamilies, Arecoideae is the most species-rich and still contains some ambiguous inter-generic relationships, such as those within subtribes Attaleinae and Bactridineae. The hypervariable regions of plastid genomes (plastomes) are interesting tools to clarify unresolved phylogenetic relationships. We sequenced and characterized the plastome of Bactris gasipaes (Bactridinae) and compared it with eight species from the three Cocoseae sub-tribes (Attaleinae, Bactridinae, and Elaeidinae) to perform comparative analysis and to identify hypervariable regions. The Bactris gasipaes plastome has 156,646 bp, with 113 unique genes. Among them, four genes have an alternative start codon (cemA, rps19, rpl2, and ndhD). Plastomes are highly conserved within tribe Cocoseae: 97.3% identity, length variation of ~2 kb, and a single ~4.5 kb inversion in Astrocaryum plastomes. The LSC/IR and IR/SSC junctions vary among the subtribes: in Bactridinae and Elaeidinae the rps19 gene is completely contained in the IR region; in the subtribe Attaleinae the rps19 gene is only partially contained in the IRs. The hypervariable regions selected according to sequence variation (SV%) and frequency of parsimony informative sites (PIS%) revealed plastome regions with great potential for molecular analysis. The ten regions with greatest SV% showed higher variation than the plastid molecular markers commonly used for phylogenetic analysis in palms. The phylogenetic trees based on the plastomes and the hypervariable regions (SV%) datasets had well-resolved relationships, with consistent topologies within tribe Cocoseae, and confirm the monophyly of the subtribes Bactridinae and Attaleinae.