Consilience across multiple, independent genomic data sets reveals species in a complex with limited phenotypic variation

Species delimitation in the genomic era has focused predominantly on the
application of multiple analytical methodologies to a single massive
parallel sequencing (MPS) data set, rather than leveraging the unique but
complementary insights provided by different classes of MPS data. In this
study we demonstrate how the use of two independent MPS data sets, a
sequence capture data set and a single nucleotide polymorphism (SNP) data
set generated via genotyping-by-sequencing, enables the resolution of
species in three complexes belonging to the grass genus Ehrharta, whose
strong population structure and subtle morphological variation limit the
effectiveness of traditional species delimitation approaches. Sequence
capture data are used to construct a comprehensive phylogenetic tree of
Ehrharta and to resolve population relationships within the focal clades,
while SNP data are used to detect patterns of gene pool sharing across
populations, using a novel approach that visualises multiple values of K.
Given that the two genomic data sets are fully independent, the strong
congruence in the clusters they resolve provides powerful ratification of
species boundaries in all three complexes studied. Our approach is also
able to resolve a number of single-population species and a probable
hybrid species, both which would be difficult to detect and characterize
using a single MPS data set. Overall, the data reveal the existence of 11
and five species in the E. setacea and E. rehmannii complexes, with the E.
ramosa complex requiring further sampling before species limits are
finalized. Despite phenotypic differentiation being generally subtle, true
crypsis is limited to just a few species pairs and triplets. We conclude
that, in the absence of strong morphological differentiation, the use of
multiple, independent genomic data sets is necessary in order to provide
the cross-data set corroboration that is foundational to an integrative
taxonomic approach.