Search

Lichens collected worldwide for centuries have resulted in millions of specimens deposited in herbaria that offer the potential to assess species boundaries, phenotypic diversification, ecology, and distribution. The application of molecular approaches to historical collections has been limited due to DNA fragmentation, but high-throughput sequencing offers an opportunity to overcome this barrier. Here, we combined a large dataset of ITS sequences from recently collected material and historical collections, obtained through Sanger, 454, or Illumina Sequencing, to test the performance of ITS barcoding in two genera of lichenized Basidiomycota: Cora and Corella. We attempted to generate new sequence data for 62 fresh specimens (from 2016) and 274 historical collections (collected between 1888 and 1998), for a final dataset of 1325 sequences. We compared various quantitative approaches to delimit species (GMYC, bPTP, ASAP, ABGD) and tested the resolution and accuracy of the ITS fungal barcoding marker by comparison with a six-marker dataset. Finally, we quantitatively compared phylogenetic and phenotypic species delimitation for 87 selected Cora species that have been formally described. Our HTS approach successfully generated ITS sequences for 76% of the historical collections, and our results show that an integrative approach is the gold-standard for understanding diversity in this group. [ABSTRACT FROM AUTHOR]

The early radiation of Neoaves has been hypothesized to be an intractable "hard polytomy". We explore the fundamental properties of insertion/deletion alleles (indels), an under-utilized form of genomic data with the potential to help solve this. We scored >5 million indels from >7000 pan-genomic intronic and ultraconserved element (UCE) loci in 48 representatives of all neoavian orders. We found that intronic and UCE indels exhibited less homoplasy than nucleotide (nt) data. Gene trees estimated using indel data were less resolved than those estimated using nt data. Nevertheless, Accurate Species TRee Algorithm (ASTRAL) species trees estimated using indels were generally similar to nt-based ASTRAL trees, albeit with lower support. However, the power of indel gene trees became clear when we combined them with nt gene trees, including a striking result for UCEs. The individual UCE indel and nt ASTRAL trees were incongruent with each other and with the intron ASTRAL trees; however, the combined indel+nt ASTRAL tree was much more congruent with the intronic trees. Finally, combining indel and nt data for both introns and UCEs provided sufficient power to reduce the scope of the polytomy that was previously proposed for several supraordinal lineages of Neoaves. [ABSTRACT FROM AUTHOR]