Oresitrophe and
Mukdenia (Saxifragaceae) are epilithic sister genera used in traditional Chinese medicine. The taxonomy of
Mukdenia, especially of
M. acanthifolia, has been controversial. To address this, we produced plastid and mitochondrial data using genome skimming for
Mukdenia acanthifolia and
Mukdenia rossii, including three individuals of each species. We assembled complete plastomes, mitochondrial CDS and nuclear ribosomal ETS/ITS sequences using these data. Comparative analysis shows that the plastomes of
Mukdenia and
Oresitrophe are relatively conservative in terms of genome size, structure, gene content, RNA editing sites and codon usage. Five plastid regions that represent hotspots of change (
trnH-
psbA,
psbC-
trnS,
trnM-
atpE,
petA-
psbJ and
ccsA-
ndhD) are identified within
Mukdenia, and six regions (
trnH-
psbA,
petN-
psbM,
trnM-
atpE,
rps16-
trnQ,
ycf1 and
ndhF) contain a higher number of species-specific parsimony-informative sites that may serve as potential DNA barcodes for species identification. To infer phylogenetic relationships between
Mukdenia and
Oresitrophe, we combined our data with published data based on three different datasets. The monophyly of each species (
Oresitrophe rupifraga,
M. acanthifolia and
M. rossii) and the inferred topology ((
M. rossii,
M. acanthifolia),
O. rupifraga) are well supported in trees reconstructed using the complete plastome sequences, but
M. acanthifolia and
M. rossii did not form a separate clade in the trees based on ETS + ITS data, while the mitochondrial CDS trees are not well-resolved. We found low recovery of genes in the Angiosperms353 target enrichment panel from our unenriched genome skimming data. Hybridization or incomplete lineage sorting may be the cause of discordance between trees reconstructed from organellar and nuclear data. Considering its morphological distinctiveness and our molecular phylogenetic results, we strongly recommend that
M. acanthifolia be treated as a distinct species.
相似文献