Integrative DNA barcoding and multimethod species delimitation reveal marker performance in the genus Ficus L. (Moraceae)
TL;DR
A comparative framework demonstrates that ITS, alone or in multilocus schemes, provides superior resolution for DNA-based species delimitation in Ficus and highlights the need for integrative taxonomic approaches to resolve complex clusters and obtain robust species boundaries.
Abstract
Accurate delimitation of species is fundamental for understanding biodiversity, evolutionary history, and conservation priorities, particularly in taxonomically challenging genera such as Ficus. This study systematically evaluates the performance of three single locus barcodes (matK, rbcL, ITS) and four multilocus combinations (matK + rbcL, matK + ITS, rbcL + ITS, matK + rbcL + ITS) for resolving species limits in Ficus. We apply four widely used analytical frameworks, Automatic Barcode Gap Discovery (ABGD), Assemble Species by Automatic Partitioning (ASAP), Bayesian Poisson Tree Processes (bPTP), and the Generalized Mixed Yule Coalescent (GMYC) model, to compare their discriminatory power. Across distance based analyses (ABGD, ASAP), ITS consistently emerges as the most informative marker, recovering the largest number of well-supported species partitions and the clearest barcode gaps, whereas plastid markers (matK, rbcL) show extensive species lumping and low resolution. Multilocus datasets, especially those including ITS, provide incremental gains in cluster recovery and stability. Coalescent approaches (bPTP, GMYC) generate biologically plausible groupings with ITS and multilocus data but tend to oversplit when applied to plastid loci, revealing persistent ambiguity in some lineages caused by both biological and methodological factors. Several species complexes (e.g. Ficus abelii/erecta, Ficus hirta/fulva, Ficus variegata/fistulosa, Ficus superba/hispida) remain problematic, exhibiting either merging or splitting across methods and pointing to cryptic diversity, recent divergence, or hybridization. Overall, our comparative framework demonstrates that ITS, alone or in multilocus schemes, provides superior resolution for DNA-based species delimitation in Ficus and highlights the need for integrative taxonomic approaches to resolve complex clusters and obtain robust species boundaries.