Search for: All records

Abstract PremiseThe ~140 species ofLoniceraare characterized by variously fused leaves, bracteoles, and ovaries, making it a model system for studying the evolution and development of organ fusion. However, previous phylogenetic analyses, based mainly on chloroplast DNA markers, have yielded uncertain and conflicting results. A well‐supported phylogeny ofLonicerawill allow us to trace the evolutionary history of organ fusion. MethodsWe inferred the phylogeny ofLonicerausing restriction site–associated DNA sequencing (RADSeq), sampling all major clades and 18 of the 23 subsections. This provided the basis for inferring the evolution of five fusion‐related traits. ResultsRADSeq data yielded a well‐resolved and well‐supported phylogeny. The two traditionally recognized subgenera (PericlymenumandChamaecerasus), three of the four sections (Isoxylosteum,Coeloxylosteum, andNintooa), and half of the subsections sampled were recovered as monophyletic. However, the large and heterogeneous sectionIsikawas strongly supported as paraphyletic.Nintooa, a clade of ~22 mostly vine‐forming species, includingL. japonica, was recovered in a novel position, raising the possibility of cytonuclear discordance. We document the parallel evolution of fused leaves, bracteoles, and ovaries, with rare reversals. Most strikingly, complete cupules, in which four fused bracteoles completely enclose two unfused ovaries, arose at least three times. Surprisingly, these appear to have evolved directly from ancestors with free bracteoles instead of partial cupules. ConclusionsWe provide the most comprehensive and well‐supported phylogeny ofLonicerato date. Our inference of multiple evolutionary shifts in organ fusion provides a solid foundation for in depth developmental and functional analyses. 

PremisePhylogenetic relationships within major angiosperm clades are increasingly well resolved, but largely informed by plastid data. Areas of poor resolution persist within the Dipsacales, including placement ofHeptacodiumandZabelia, and relationships within the Caprifolieae and Linnaeeae, hindering our interpretation of morphological evolution. Here, we sampled a significant number of nuclear loci using a Hyb‐Seq approach and used these data to infer the Dipsacales phylogeny and estimate divergence times. MethodsSampling all major clades within the Dipsacales, we applied the Angiosperms353 probe set to 96 species. Data were filtered based on locus completeness and taxon recovery per locus, and trees were inferred using RAxML and ASTRAL. Plastid loci were assembled from off‐target reads, and 10 fossils were used to calibrate dated trees. ResultsVarying numbers of targeted loci and off‐target plastomes were recovered from most taxa. Nuclear and plastid data confidently placeHeptacodiumwith Caprifolieae, implying homoplasy in calyx morphology, ovary development, and fruit type. Placement ofZabelia, and relationships within the Caprifolieae and Linnaeeae, remain uncertain. Dipsacales diversification began earlier than suggested by previous angiosperm‐wide dating analyses, but many major splitting events date to the Eocene. ConclusionsThe Angiosperms353 probe set facilitated the assembly of a large, single‐copy nuclear dataset for the Dipsacales. Nevertheless, many relationships remain unresolved, and resolution was poor for woody clades with low rates of molecular evolution. We favor expanding the Angiosperms353 probe set to include more variable loci and loci of special interest, such as developmental genes, within particular clades.