Ice age-driven range shifts of diploids and expanding autotetraploids of Biscutella laevigata within a conserved niche

Abstract

Introduction

The rise of barriers to gene flow eventually promoting reproductive isolation between previously interbreeding populations at the origin of new species typically requires thousands of generations (Rieseberg & Willis, 2007; Sobel et al., 2010). Although selection likely speeds up speciation, its interactions with recombination and the rise of reproductive isolation under environmental changes are insufficiently understood (Schluter & Rieseberg, 2022). Accordingly, to what extent climate-induced range shifts due to cold and warm phases of the Quaternary ice ages have hindered speciation because allopatric differentiation was repeatedly counteracted by homogenising gene flow remains debated (Willis & Niklas, 2004; Kadereit & Abbott, 2021). In contrast to homoploid divergence, whole-genome duplication (WGD) immediately confers strong reproductive isolation from progenitor species (Levin, 1975; Ramsey & Schemske, 1998; Barker et al., 2016) and supports polyploidy as a major driver of plant speciation (Wood et al., 2009). However, to what extent polyploidy promotes the origin of new plant species during periods of climate changes remains elusive (Levin, 2019; Van de Peer et al., 2021).

Polyploid speciation couples WGD with the combination of more or less divergent gene sets either within or between species at the origin of autopolyploid or allopolyploid species, respectively (Parisod et al., 2010). At one end of the spectrum, autopolyploids derived from homologous chromosomes are characterised by tetrasomic inheritance, whereas the merging of divergent genomes results in allopolyploids displaying disomic inheritance and fixed heterozygosity. As natural polyploids can arise anywhere between these endpoints and be described as intervarietal autopolyploids or segmental allopolyploids derived from different taxa, whose homoploid hybrids are (semi-)sterile and increase their fertility through WGD (Stebbins, 1950), it is crucial to infer the exact mechanisms at their origin to address the consequences of polyploidy (Tayalé & Parisod, 2013). According to the secondary contact hypothesis (Stebbins, 1984), climate-induced range shifts typical of the Quaternary likely promoted admixture between locally adapted diploid lineages favouring the emergence of new hybrid and polyploid species, as reported for several narrow neoendemic species of allopolyploid origin (e.g. Grünig et al., 2021). This may partially explain the origin of polyploids around the last glacial maximum (LGM, c. 22 000 yr ago; Seguinot et al., 2018), as documented in for example the Arabidopsis genus (Novikova et al., 2018). Additionally, the expression of new phenotypes such as those related to physiological and ecological novelties resulting chiefly from the merging of divergent gene combinations has been considered as key in promoting rapid adaptation to novel environments, conferring allopolyploids with advantages in times of climatic instability (Sobel et al., 2010; Van de Peer et al., 2021). However, to what extent autopolyploidy may foster necessary phenotypic novelties beyond divergence facilitated by WGD remains a largely open question (Maherali et al., 2009; Parisod, 2024).

Quaternary climatic oscillations in the European Alps provide a great system to study underpinnings of plant diversification (Kadereit, 2024). In particular, cycles of climate-induced range shifts triggering allopatric differentiation and secondary gene flow (Hewitt, 1996; Kadereit et al., 2004; Boucher et al., 2016; Tomasello et al., 2020; Smyčka et al., 2022) are likely to have promoted hybrid speciation (Parisod, 2022) and fostered the origin of new allopolyploid species in the Alps (e.g. Widmer & Baltisberger, 1999; Berthouzoz et al., 2013), although the possible role of autopolyploidy in Quaternary speciation contrastingly remains elusive. The Biscutella laevigata species complex is a typical textbook example of a mixed-ploidy system associated with ice ages in the Alps that deserves further investigation. Diploids (2n = 2x = 18) were indeed early shown as mostly restricted to never-glaciated areas across lowlands around the Alps and interpreted as glacial relicts, whereas tetraploids (2n = 4x = 36) chiefly occurring across previously glaciated areas at high elevation were interpreted as postglacial recolonisers (Manton, 1934). Following the description of meiotic multivalents typical of autopolyploids (Manton, 1937), Stebbins (1950) classified tetraploids of B. laevigata as intervarietal autopolyploids to insist on their putative origin through hybridisation between geographically, morphologically, and/or ecologically divergent diploids following secondary contact. The extensive polymorphism within the strictly outcrossing B. laevigata, coupled with the fragmented distribution of diploid populations across central Europe, indeed led to the description of > 20 species, subspecies or varieties in this species complex (Machatschki-Laurich, 1926; Olowokudejo & Heywood, 1984; Raffaelli & Baldoin, 1997). As previous genetic studies focused on either only part of the distribution range (e.g. the western Alps; Parisod & Besnard, 2007) or used genetic markers with poor resolution (e.g. allozymes; Tremetsberger et al., 2002), the spatiotemporal differentiation of diploids and the origin of tetraploids of B. laevigata in the Alps remain elusive. In this study, we thus combined ddRAD-seq (double-digest restriction site-associated DNA sequencing) to depict genome-wide patterns of genetic variation and niche modelling to revisit the evolutionary history of B. laevigata in the European Alps. Based on a comprehensive sampling of diploid populations of B. laevigata around and in the Alps and dense sampling of tetraploids across the Alps, we assessed their genetic and ecological differentiation in space and time to specifically address (1) the role of hybridisation in the origin of the tetraploids of the Alps, (2) whether tetraploids evolved through a single vs multiple origins, and finally (3) whether the ploidy shift was accompanied by a climate niche shift.