Taxonomic consequences of inter-specific gene flow
Taxonomy in the genus Mentha is solely based on morphology (Harley & Brighton, 1977; Tucker et al., 1980; Tucker & Naczi, 2007; Vining et al., 2020). However, here we show that the results of genomic analyses are not fully congruent with a strictly morphology-based taxonomy (Figures 1 and 2). However, although our genomic analyses suggest that M. suaveolens , M. longifolia , M. ×rotundifolia , and M. spicata should be considered a coalescent complex where no taxa are completely isolated genetically, we also find that for most parts the classically recognized taxa are morphologically supported (Figures 1 and 2). Hence, the classical taxonomic units are still mostly useful although we suggest that an in-depth taxonomic revision including more taxa is made. The observed dissociations between the morphological and genomic clustering appear to be partially due to the continued segregation of genetic variants associated with characters of indumentum. Continued and on-going cultivation and escape into the wild of various mint species and varieties, including M. spicata , will likely continue to feed the native gene pools with alleles from non-native taxa. It is therefore likely that there will continue to be dissociations between morphological and genomic clustering of specimens.
Here, we have focused on mostly old herbarium specimens and hence we cannot be certain that we have included any of the currently widely cultivated varieties ofM. spicata . These varieties could possibly be genomically distinct and show a genomic admixture profile that is different from what we present here. There are numerous studies on the evolutionary origin of various mint hybrids and the phylogenetic relationships withinMentha (eg. Bunsawatt, Elliott, Hertweck, Sproles, & Alice, 2004; Gobert et al., 2002; Heylen et al., 2021; Panjeshahin, Sharifi-Sirchi, & Samsampour, 2018). Contradictions in conclusions between previous studies can partially be explained by biased inclusions of particular specimens and/or inconsistent delimitation of taxa, but as we, show here not including all sub-populations in such analyses can also bias the results. Inclusion of multiple samples per taxonomic unit is therefore of importance, especially for taxa, such as mints, where frequent inter-fertility is reported (Heylen et al., 2021).