Population genetic structure of iberian white poplar (populus alba l.)the role of mating system, hybridization and demographical history

Resumen

Populus represents an interesting tree genus for breeding system and evolutionary studies. It combines dioecy and clonality as reproductive modes, possesses an incipient sex chromosome (XIX) but, at the same time, shows occasional sex inconstancy and hybridises frequently, albeit not among some Sections. Moreover, many of its species grow exclusively in riparian areas and wetlands (often in apparently isolated populations), but with subtly different ecological optima. Finally, it has a relatively small genome with elevated synteny that has been successfully sequenced. White and black poplars (Populus alba and P. nigra, respectively) grow along many water courses in the Iberian Peninsula, sometimes in sympatry, whereas European aspen (P. tremula) inhabits upland hillsides in parapatry. Neither poplar can interbreed naturally, although, in occasions, white poplar and European aspen do (Populus × canescens). Furthermore, some Iberian white poplar populations possess high clonality rates or hermaphroditic individuals. In order to gain insight into the evolutionary and ecological significance of the reproductive system in Iberian Populus, this PhD work aimed at (i) studying its spatial genetic structure at the local and regional scale; (ii) evaluating the degree of local adaptation in white poplar through different approaches; (iii) researching the extent, causes and implications of clonality; (iv) appraising the ecological and evolutionary issues of hybridisation between white poplar and European aspen; and (v) enhancing the knowledge of the incipient Populus sex chromosome. To achieve these goals, chloroplast DNA of numerous Iberian populations of poplars was characterised and quantitative genetic differentiation of several adaptive traits was evaluated among contrasting populations. Clonality was investigated in several Iberian populations, with a higher focus in one of them (the hybrid zone located in the proximity of Aranda de Duero), using nuclear microsatellites. These markers were also used to genotype European populations of P. tremula, including one from central Spain (Sierra de Gredos), as well as individuals from common garden trials ¿ for which some phenotypic traits were also measured ¿, and an artificial backcross from P. × canescens to P. alba (BC1), in the latter case with a higher focus on chromosome XIX. Additionally, several open-pollinated progenies from an Andalusian subdioecious (i.e. comprising males, females and some hermaphrodites) white poplar population were genotyped. Finally, several genes involved in wood formation were sequenced in Iberian poplar populations that were representative of different gene pools. White and black poplars displayed contrasting phylogeographic patterns in the Iberian Peninsula, reflecting that differences in ecological requirements (e.g. tolerance to cold temperatures) impose different regional genetic structures. In particular, white poplar confinement to river and drainage basins was higher than black poplar¿s. Southern populations of white poplar were more genetically diverse, and migration between the northern and southern regions was higher than between the eastern and western ones. These patterns suggest the effect of Holocene glaciations on poplars, despite their dependence of water courses. The presence of numerous private haplotypes indicated that no regional extinctions have taken place during glaciations in any of the studied river basins. In white poplar, local adaptation was detected for two quantitative traits (total height and stem form at age 3), but at a wider spatial scale than in other forest trees (among river basins but not among populations within river basins). Molecular genetic adaptation imprints in wood formation genes were controversial: several outlier loci were found using one method, but none was detected using a more stringent analysis. This probably reflects sample size issues. In European aspen, local adaptation was also detected in quantitative traits. In addition, clear signatures on patterns of standing genetic variation, probably created by the admixture of two European distinct lineages, were uncovered for this species. In white poplar, levels of clonality averaged 4.2 ramets per genet, in populations with 32-49 sampled ramets, genets reaching sizes of ca. 100 m. However, in two populations much larger clones (>10 km) were found. One of these populations, Aranda de Duero in north-central Spain, was more intensively studied. In Aranda de Duero, I found two large and ancient clones, one of them a hybrid with aspen, which spread for over a hundred kilometres. While genotypic richness and evenness was lower than in other European populations, genetic diversity was similar in Aranda de Duero and there was no apparent risk of genetic depauperation. Coalescence modelling showed that overall effective population size was declining, although large genets were probably increasing in ramet number. Besides, larger genets were generally parents of more individuals. Punctual genomic introgression from P. tremula and heterozygosity could be involved in the accelerated clonal spread in this population. Further insides on the role of interspecific heterozygosity in adaptation were obtained from the BC1 cross. I found segregation distortion towards P. tremula, probably due to cyto-nuclear interactions or heterozygote advantage. This supports the idea that heterozygosity could play an important role in adaptation, in particular when enhanced through hybridisation. Some findings agreed with expectations for a sex chromosome in chromosome XIX, while other disagreed. No recombination was detected in backcrossed progeny among four markers of chromosome XIX, as expected for sex chromosomes. However, the distribution of genetic diversity at this chromosome of separated male and female groups and its reduced divergence in natural populations were against sex chromosome expectations. Overall, my research supports the hypothesis that chromosome XIX is at a very early step of degeneration into a mature sex chromosome. Finally, the paternity analyses done in the Jimena population revealed that the effective number of fathers involved in pollination ranged from 1.58 to 8.33. Clone size and distance were apparently involved in pollination rate. Regarding hermaphrodites, one of them effectively sired offspring. In addition, a correlation between feminity (average proportion of female flowers per catkin, considering hermaphrodites flowers as half-female flowers) and germination rate was detected. Based on the results of this PhD thesis ¿ that local adaptation and regional spatial genetic structure occur among Iberian poplar populations ¿, it is advisable to take provenance origin into account in restoration programmes and to consider a river-basin basis when implementing in situ conservation plans. The conspicuous extent of clonality makes it necessary to consider natural clone replicates in any management measure concerning these trees, although it is equally important to remark that such degree of clonality is not producing detrimental genetic effects in the studied population. Hybridisation and lineage admixture have proven to be valuable ways of genetic enrichment, with repercussions on fitness and local adaptation. Forest managers could put this fact to good use by adopting simple mitigation measures in response to future climate change threats. Finally, chromosome XIX operates like a sex chromosome at an initial stage of development, which makes room for the occasional existence of hermaphrodites. A deeper view on the interplay among hermaphroditism, sex chromosomes and clonality would allow a better understanding of chromosome XIX¿s relative ecological significance, and of white poplar¿s potential evolutionary pathways, as the species is an important dioecious riparian tree.