4.1 Species boundary delimitation of Androsace cantabricausing phylogenomics
Clarifying the taxonomic status of A. cantabrica has both scientific and conservation implications, as distinguishing it from closely related taxa is crucial for understanding its threatened status and prioritizing conservation measures (Godfray et al., 2004; Ottewell et al., 2016; Kress et al., 2017). We implemented a phylogenomic approach to resolve whether A. cantabrica is a valid species and, therefore, threatened or if it is part of a wider and non-threatened taxon (A. adfinis subsp. adfinis ). Phylogenetic reconstructions based on hundreds of nuclear loci and plastid sequences obtained from Angiosperms353 target capture data consistently resolvedA. cantabrica as a monophyletic clade, confirming its phylogenetic distinctiveness in alignment with previous morphological (Kress, 1997) and karyotypic data (Kress, 1984). Our phylogenomic approach improved our understanding of A. cantabrica and allowed us to investigate its distribution, genetic diversity and evolutionary origin and supported its categorization as a valid species in need of conservation.
The placement of A. cantabrica as sister to the /halleri clade in the nuclear Angiosperms353 phylogenetic trees, rather than to A. adfinis, aligns with the morphological groupings of Androsace sect. Aretia as per Smith & Lowe (1997). Interestingly, while nuclear Angiosperms353 data positioned A. cantabrica as a sister to the /halleri clade, plastid data and previous studies suggested a closer phylogenetic relation to A. adfinis . This topological incongruence may reflect a divergence in evolutionary paths between the plastid and nuclear genomes, a phenomenon observed across many plant species (e.g., Galbany-Casals et al., 2014; Viruel et al., 2018 Favre et al., 2022; Liu et al., 2023).
The evolutionary origin and taxonomic identity of A. cantabrica have been subjects of significant debate. Initially, it was hypothesized that A. cantabrica was an allopolyploid species resulting from hybridization between A. laggeri (2n = 38, localized in the central Pyrenees) and A. halleri (2n = 38, distributed in the Cantabrian Mountains, Pyrenees, Massif Central, and Vosges; Kress, 1984). However, this hypothesis was challenged by Dixon et al. (2008), who used amplified fragment length polymorphism (AFLP) data to refute a close relationship between A. cantabrica and eitherA. halleri or A. laggeri . Instead, they suggested thatA. cantabrica was an autopolyploid related to the southwestern Alps’ A. adfinis s.l. , which includes A. adfinis subsp.adfinis , A. adfinis subsp. puberula (Jord. & Fourr.) Kress, and A. adfinis subsp. brigantiaca . Our data indicate that A. cantabrica is a polyploid and likely has a complex evolutionary history. While Dixon et al. (2008) proposed an autopolyploid origin closely related to A. adfinis , our analysis suggests a different scenario, albeit without conclusively identifying the exact polyploidy type. Although paralogous genes were limited in our analysis, which included A. cantabrica samples, definitive proof of an autopolyploid origin is absent. For example, A. adfinis subsp. brigantiaca is suspected of being a recent hybrid (Boucher et al., 2016) with tetraploid features, yet only exhibited six paralogous genes in HybPiper analyses. Possible scenarios for A. cantabrica origin include an ancient homoploid hybridization event involving the ancestors of A. cantabrica and the /halleri clade, followed by local polyploidization, or an initial allopolyploidization. In any case, given its phylogenetic, morphological, and karyological uniqueness, A. cantabrica should be considered a valid species with an evolutionary trajectory shaped by rapid speciation, introgression, and possibly hybridization in alpine environments (Hibbins et al., 2020; Smyčka et al., 2022).