Background Quantifying variation of genetic traits over the geographical range of species is crucial for understanding the factors driving their range dynamics. The center-periphery hypothesis postulates, and many studies support, the idea that genetic diversity decreases and genetic differentiation increases toward the geographical periphery due to population isolation. The effects of environmental marginality on genetic variation has however received much less attention. Methods We tested the concordance between geographical and environmental gradients and the genetic predictions of center-periphery hypothesis for endemic Lilium pomponium in the southern Alps. Results We found little evidence for concordance between genetic variation and both geographical and environmental gradients. Although the prediction of increased differentiation at range limits is met, genetic diversity does not decrease towards the geographical periphery. Increased differentiation among peripheral populations, that are not ecologically marginal, may be explained by a decrease in habitat availability that reduces population connectivity. In contrast, a decrease of genetic diversity along environmental but not geographical gradients may be due to the presence of low quality habitats in the different parts of the range of a species that reduce effective population size or increase environmental constraints. As a result, environmental factors may affect population dynamics irrespective of distance from the geographical center of the range. In such situations of discordance between geographical and environmental gradients, the predictions of decreasing genetic diversity and increasing differentiation toward the geographical periphery may not be respected.

Climate change is known to have a profound influence on plant reproduction, mainly because it affects plant/pollinator interactions, sometimes driving plants to extinction. Starting from the Neogene, the European climate was subjected to severe alterations. Nevertheless, several genera, including Berardia, survived these climatic changes. Despite the numerous studies performed about the relationship between climate change and plant reproductive biology, equivalent studies on ancient species are lacking, even though they may furnish crucial information on the strategies that allowed them to survive drastic climatic fluctuations. We investigated floral and reproductive features in Berardia subacaulis (Asteraceae), describing pollen vectors, capitulum and florets phenology, evaluating reproductive efficiency and defining the reproductive mode of the plant with bagging experiments and test of apomixis. B. subacaulis grows in habitats with low pollination services; it is self-compatible, but many typical features favouring cross-pollination are still present: florets are characterized by incomplete protandry, capitulum protogyny and high pollen-ovule ratio. The plant is not apomictic and self-fertilization is allowed within each capitulum. Similarly to other European Alpine endemics supposed to belong to the Mediterranean ancient tropical flora, the reproductive mode observed in the monospecific genus Berardia assured reproduction also under a pollinator decline. Differently from the other endemics, it took advantage of its spontaneous self-pollination and compatibility and its generalist pollination service, common both among high altitude plants and in the Asteraceae.

Geographical limits of species’ distributions are assumed to be coincident with ecological margins, although this assumption might not always be true. Indeed, harsh environments such as Alpine and Mediterranean ecosystems may favour high phenotypic variability among populations, especially those in peripheral sites. Floral traits are often found to be less variable and less affected by environmental heterogeneity than vegetative traits because variation in the former may have negative effects on fitness. For this reason, it is important to quantify variation in floral traits and plant fecundity in study range limits. The objective of the study is to examine phenotypic variation and differences in reproduction in endemic Lilium pomponium in the Maritime and Ligurian Alps in relation to environmental variation across its distribution range. In this species, marginal climatic populations occur both in the peripheral and central geographical locations of the distribution range; hence, geographical and ecological gradients are not concordant. Floral trait variation is related to local environmental conditions with an array of interactions among resource availability, potential pollen limitation and population size that are differentially related to floral traits. Contrary to the general expectation, all central and peripheral populations had similar, moderate seed production with each group limited by different factors acting on different stages of the life-history strategy. Our results are in line with the idea that general expectations are confirmed only when its assumptions are met and that the differences in pollination environment along an environmental gradient may not be the main determinant of the distribution limit.

No abstract

No abstract