The distribution of plant species around the Mediterranean basin is a product of the influence of both geographical barriers and of climatic changes experienced during the Tertiary, with the transition from a warm to cool periods. Several species, once largely distributed across the Northern Hemisphere, retracted to refugial areas in southern Europe where they are described as Tertiary relicts. Platanus orientalis is a typical representative of Tertiary flora in southwest Eurasia; its distribution spreads from the Caucasus to the Mediterranean, with its western border in Southern Italy and Sicily. We analysed genetic diversity and differentiation in the centre and western part of its distribution range using nuclear microsatellites and compared genetic parameters between core and western populations. We found an overall decrease in genetic diversity estimates (heterozygosity, private allelic richness) from central towards western populations, with those from Southern Italy and Sicily showing the lowest values. The low level of genetic diversity probably results from historic range fragmentation experienced by P. orientalis in its westernmost distribution as confirmed by high level genetic isolation of these populations. Ornamental hybrids were genetically distinguished from P. orientalis as contained private alleles, indicating that gene flow with natural populations is rare. Population assignment and neighbour-joining (NJ) analysis of populations identified four groups belonging to two main phyletic groups (the Southern Italian-Sicilian-Balkan and Cretan-Bulgarian-Turkish lineages) that seem to have different biogeographic origin and also excluded an artificial origin for southern Italian and Sicilian populations as previously suggested. These results show that quantifying the genetic variation of a Tertiary relict in a geographical context, and the potential effect of hybridization with introduced ornamental hybrids, can provide useful insights on factors influencing population genetic structure. Such information is crucial to predict how such taxa will respond to increasing anthropogenic influence on the environment.

Tetraploid lineages are typically reproductively isolated from their diploid ancestors by post-zygotic isolation via triploid sterility. Nevertheless, polyploids often also exhibit ecological divergence that could contribute to reproductive isolation from diploid ancestors. In this study, we disentangled the contribution of different forms of reproductive isolation between sympatric diploid and autotetraploid individuals of the food-deceptive orchid Anacamptis pyramidalis by quantifying the strength of seven reproductive barriers: three prepollination, one post-pollination prezygotic and three post-zygotic. The overall reproductive isolation between the two cytotypes was found very high, with a preponderant contribution of two prepollination barriers, that is phenological and microhabitat differences. Although the contribution of post-zygotic isolation (triploid sterility) is confirmed in our study, these results highlight that prepollination isolation, not necessarily involving pollinator preference, can represent a strong component of reproductive isolation between different cytotypes. Thus, in the context of polyploidy as quantum speciation, that generates reproductive isolation via triploid sterility, ecological divergence can strengthen the reproductive isolation between cytotypes, reducing the waste of gametes in low fitness interploidy crosses and thus favouring the initial establishment of the polyploid lineage. Under this light, speciation by polyploidy involves ecological processes and should not be strictly considered as a nonecological form of speciation.

Platanthera bifolia and P. chlorantha are terrestrial and rewarding orchids with a wide Eurasian distribution. Although genetically closely related, they exhibit significant morphological, phenological and ecological differences that maintain reproductive isolation between the species. However, where both species co-occur, individuals with intermediate phenotypic traits, often considered as hybrids, are frequently observed. Here, we combined neutral genetic markers (AFLPs), morphometrics and floral scent analysis (GC-MS) to investigate two mixed Platanthera populations where morphologically intermediate plants were found. Self-pollination experiments revealed a low level of autogamy and artificial crossings combined with assessments of fruit set and seed viability, showed compatibility between the two species. The results of the genetic analyses showed that morphologically intermediate plants had similar genetic patterns as the P. bifolia group. These results are corroborated also by floral scent analyses, which confirmed a strong similarity in floral scent composition between intermediate morphotypes and P. bifolia. Therefore, this study provided a much more detailed picture of the genetic structure of a sympatric zone between two closely allied species and supports the hypothesis that intermediate morphotypes in sympatry could reflect an adaptive evolution in response to local pollinator-mediated selection.

Platanthera bifolia and P. chlorantha are terrestrial and rewarding orchids with a wide Eurasian distribution. Although genetically closely related, they exhibit significant morphological, phenological and ecological differences that maintain reproductive isolation between the species. However, where both species co-occur, individuals with intermediate These results are corroborated also by floral scent analyses, which confirmed a strong similarity in floral scent composition between intermediate morphotypes and P. bifolia. Therefore, this study provided a much more detailed picture of the genetic structure of a sympatric zone between two closely allied species and supports the hypothesis that intermediate morphotypes in sympatry could reflect an adaptive evolution in response to local pollinator-mediated selection. (Nilsson 1983;1985). These two closely related species exhibit not only 67 morphological differences, but also distinct ecological preferences (P. chlorantha favouring dry, 68 calcareous grasslands, while P. bifolia will be typically found in fresh to wet meadows on acidic soil). 80 Recently, a study on some Western-European Platanthera populations composed almost exclusively of 81 intermediate looking individuals, based on morphology and molecular markers, concluded that such 82 individuals were not hybrids, but constitute an independent lineage, distinct from both widespread species 83 (Durka et al. 2017). 84The level of geitonogamy was observed to be higher in P. bifolia than in P. chlorantha because 85 the latter has a pollinarium-bending mechanism that prevents deposition of the pollinia directly after 86 removal (Maad & Nilsson 2004;Maad & Reinhammar 2004). This process may also affect the probability 87 of hybrid formation (Ishizaki et al. 2013). An allopatric P. bifolia population with a high degree (i.e. 88 almost 60%) of self-pollination was found by Brzosko (2003), although self-pollination in Platanthera 89 species is considered generally rare (Nilsson 1983;Maad 2002). 90In the genus Platanthera, floral scent plays a crucial role in guiding pollinators to the flowers 91 (Nilsson 1983;1985;Tollsten & Bergström 1993). A strong fragrance is emitted after dusk, when 92 pollinators (nocturnal moths) are most active (Nilsson 1983;1985;Tollsten & Bergström 1993; Hapeman 93 & Inoue 1997;Plepys et al. 2002a;2002b). Floral fragrances of Platanthera have been classified into 94 linaloolic, lilac, geraniolic and benzenoic chemotypes depending on the main class of compounds present 95 in the blend (Tollsten & Bergström 1993;Plepys et al. 2002a;2002b (Plepys et al. 2002a;2002b). Furthermore, a change in floral scent composition has been suggested 98 by Nilsson (1983;1985) to prevent effective cross-pollination between both species (Nilsson 1978; 99 Tollsten & Bergström 1993), by acting as a reproductive barrier via ethological mechanisms. Tollsten & 100 Bergström (1993) discovered that important inter-individual and inter-population variation in floral scent 101 exists, and may act as an adapta...

No abstract