Incongruence of morphological determinations and DNA barcode sequences: a case study in Cinnamomum (Lauraceae)

In the course of molecular systematic studies of Lauraceae we received a sample of a plant cultivated under the name Cinnamomum porrectum in the Botanical Garden München-Nymphenburg. Preliminary determinations, both morphologically based on the Flora of China (Li & al. 2008) and by chloroplast sequences (psbA-trnH spacer, trnK intron including matK gene, trnL intron, trnL-trnF spacer and trnQ-rps16 spacer) obtained by Sanger sequencing suggested that it was C. camphora, still the plant looked different from other individuals of C. camphora cultivated in the botanical gardens of Berlin, Hamburg, Mainz, Munich and Oldenburg. Attempts to sequence the more informative nuclear internal transcribed spacer repeatedly led to mixed signals. We therefore used Illumina sequencing on a set of pre-amplified molecular markers (ITS, trnK 3′ and 5′ intron, trnL intron, and the intergenic spacers psbA-trnH, trnL-trnF, as well as parts of the trnQ-rps16 spacer), and downloaded available sequences of C. camphora and C. parthenoxylon from GenBank for comparison. Considerable differences were found among these sequences, but the haplotype groups do not coincide with the current species determinations. Particularly the internal transcribed spacer sequences are rather diverse, suggesting possible misidentifications, contaminations, and/or a common gene pool that is larger than anticipated. Concerning the plant in question, our results suggest that it may be a hybrid, with C. camphora as the maternal and another species, possibly C. parthenoxylon, as the paternal parent.