POPULATION GENOMICS, DISTRIBUTION AND REPRODUCTIVE ECOLOGY OF AN ITALIAN STENO-ENDEMIC CHASMOPHYTE, CAMPANULA RAINERI

Plant species distributions are the result of climatic and evolutionary histories. Range restriction in steno-endemic species often results from climate oscillations that have reduced the extent of suitable habitat or limited species dispersion across ecological barriers. Narrow endemics may also be of recent origin (yet to disperse) or relicts of ancient lineages. High-elevation environments tend to be rich in endemism as a result of complex climatic histories (underpinning speciation events) and the survival of relicts. Mountain species typically experience extreme conditions and significant temperature changes across short spatial gradients and timeframes, and are particularly sensitive to climate fluctuations, especially on exposed substrates such as rocky faces and screes. Consequently, mountain chasmophytes (i.e. rupestral plant species), are often range-restricted, characterised by small and fragmented populations. The current thesis aims to better understand the causes and consequences of rarity and predict the possibilities for future conservation of a model chasmophyte species, Campanula raineri Perp. (Campanulaceae), a tetraploid steno-endemic of the Italian Prealps, by addressing the following hypotheses: 1) Polyploidy (chromosome set duplication) is a principle driver for the emergence of new endemics. A global-scale analysis of 4210 species representing the principal clades of angiosperms addressed the relationships between chromosome number, geographic distribution and taxon age (Chapter 2). Polyploidy was shown to be a key driver of endemism in angiosperms, with a significant negative exponential relationship between the maximum number of chromosomes and taxon age (R2adj = 0.49, p < 10-4), and greater maximum chromosome numbers (2n = 164 vs. 111) declining more rapidly with taxon age (decay constant = 0.12 vs. 0.04) compared to non-endemics. 2) The genetic structure of C. raineri populations is consistent with changes in habitat suitability throughout the Quaternary. Demographic history (gene flows, bottlenecks, and reproductive isolation events), can be investigated through population genomics to infer species phylogeography. Phylogeographically informed Species Distribution Models (SDMs) can reconstruct and predict changes in the geographical range of different intraspecific genetic lineages according to their ecological requirements and climate projections. Hypothesis 2 was thus verified by comparing the genetic structure of 14 populations across the entire distributional range of C. raineri with the climatic history of the region during the Quaternary (Chapter 3). Genomic analysis of C. raineri populations revealed two main genetic lineages, probably becoming isolated during the Last Interglacial (~ 120,000 years bp) due to habitat recolonization dynamics and subsequently during the LGM due to the persistence of the Valsassina glacier barrier. Within these two metapopulations, a more recent site-specific differentiation was evident as a result of the upward migration of the species on isolated reliefs (posterior population assignment probability calculated by Markov chain Monte Carlo clustering algorithm always = 1). 3) An upward migration of C. raineri is expected in future decades as a consequence of sub-optimal conditions at lower elevation sites due to climate warming. Predictions based on phylogeography and SDMs can be informative for rare species conservation under current climate change scenarios. Understanding the optimal conditions for species fitness (i.e. survival and reproductive success) and where these could be found in the future is crucial to inform targeted conservation actions such as population reinforcements and assisted migrations. Hypothesis 3 was addressed both by building phylogeographically-informed future SDMs (investigating future decades until 2070; Chapter 3) and testing the impacts of elevation on the reproductive ecology of C. raineri, from pollinator guilds to pollen and seed quality (Chapter 4). Future SDMs suggest a decrease in habitat suitability at lower-elevation sites, where C. raineri populations exhibited reduced fitness (correlation coefficients = 0.62 and 0.93 for seed germination and pollen viability, respectively). 4) In-vitro propagation and plant production is feasible for the conservation of rare chasmophytes. Germination protocols for C. raineri and other rare species (4 chasmophytes and 1 orchid) of northern Italy were tested (Chapter 5): successful in-vitro propagation of target species was performed (mean germination rate of each species generally > 80%, lower for Linaria tonzigii and Saxifraga tombeanensis), with the production of hundreds of new plants. In conclusion, ecological barriers during the Quaternary resulted in the genetic differentiation of C. raineri into two metapopulations, the ‘lower’ lineage exhibiting genetic uniformity and reduced fitness probably as a consequence of sub-optimal conditions for the species with increasing temperatures. The survival of this lineage is further threathened by the lack of possibilities for spontaneous upward migration, entailing the risk of genetic depletion within the entire species. However, SDMs and successful in-vitro propagation suggest that informed conservation actions could be undertaken to counter this trend.