Estimating contemporary pollen-flow and associated outbreeding effects on herbaceous plant populations along altitudinal gradients

Philippe Matter

Résumé:

Current rapid global warming is inducing changes in local environmental conditions. To cope with those changes plant populations might migrate, adapt genetically, or react with phenotypic plasticity. Genetic adaption requires that the genetic diversity available is large enough to undergo selection, or it may occur through outbreeding and integration of beneficial, adaptive alleles. For outbreeding to happen, gene ‐ flow must be able to connect populations. The effects of outbreeding on plant populations are difficult to predict because they depend on many factors, including the level of genetic differentiation, the possible occurrence of local adaptation, and the existence of adaptive alleles in the populations concerned. Depending on these factors, populations subject to outbreeding events can either experience outbreeding vigour, outbreeding depression, or remain unaffected. In mountainous ecosystems, because of the strong environmental gradients, plant populations are undergoing potentially divergent selection pressures which give rise to different local adaptations even within short distances. Altitude and the associated temperature change is one of the most important gradients in mountain ecosystems. Rising temperatures due to global warming have proven to induce a shift of plant species and communities towards higher altitudes. Species not able to migrate will have to rely on their plasticity or the possible integration of adaptive alleles from lower populations, enabling them to cope with e.g. warmer temperatures. In this research, I investigated both the extent of gene ‐ flow and the potential consequences of inter ‐ population outbreeding on mountain plant populations. I used two common and perennial herbaceous species, the mountain clover Trifolium montanum L. and the bulbous buttercup Ranunculus bulbosus L.. In order to simulate the 3K temperature increase expected in the next century in the European Alps, I investigated plant populations located 600m elevation apart. I focused on pollen ‐ flow because it is more efficient to disperse genes between existing populations than seed ‐ flow. Outbreeding assessment was done with various populations from Switzerland located at 1200 and 1800m a.s.l.. Investigation of pollen ‐ flow was based on molecular genetic methods, for which I first had to develop the necessary tools for fingerprinting. For each of the study species I developed a set of microsatellite markers. Ranunculus bulbosus samples were genotyped with seven microsatellites while T. montanum samples were genotyped with eleven microsatellites. In the first part (Chapter 2) I used a paternity analysis to explore patterns of contemporary pollen dispersal in an experimental population. Although not realized in a natural population because of the high density of conspecifics, this direct ‐ assignment method has the advantage of having a great resolution, and allows for investigating correlated paternity and mating system. I found that the majority of the dispersal occurred over very short distances, and that long distance dispersal events happened over 300m. Trifolium montanum proved to have slightly less extensive pollen dispersal than R. bulbosus even though R. bulbosus was displaying very variable and sometimes high selfing rates. Also, R. bulbosus proved to have higher level of correlated paternity compared to T. montanum . In the second part (Chapter 3), I used a pollen ‐ pool analysis, an indirect ‐ assignment method, which allows investigating pollen ‐ flow in large populations with limited sampling effort compared to a paternity analysis that would be realized at the same scale. This method was applied on a 1200 to 1800m a.s.l. altitudinal gradient, and the patterns of contemporary pollen dispersal were set in relation with flowering phenology along the gradient. Historic gene ‐ flow was also assessed and proved to be extensive, as revealed by very low F st values; genetic diversity was similar along the gradient. In T. montanum , contemporary pollen flow proved to be extensive and able to connect higher and lower populations. In R. bulbosus , pollen ‐ flow appeared to be slightly less extensive, which was consistent with the decreasing density of individuals along the gradient and the smaller flowering overlap compared to T. montanum. In the third part (Chapter 4), I used several populations of R. bulbosus from 1200 and 1800m a.s.l. to investigate the effects of outbreeding in regard to altitudinal and geographical origin of pollen (relative to the recipient population). I realized controlled crosses among population arranged in three geographical replicates and assessed seed ‐ set, germination, growth, flowering phenology and fitness of the offspring over two growing seasons. Outbreeding effects in terms of fitness were mainly driven by geographical origin, while growth was partially affected by the altitudinal origin of pollen, however only in the higher populations. Higher and lower populations differed in several traits, which were attributed to a potential home effect of the lower vs. higher populations in the greenhouse conditions. Taken together, these findings suggest that R. bulbosus and T. montanum are likely to be resilient to climate change, owing to populations’ very low genetic differentiation, high genetic diversity, and to their connectivity. The lack of strong outbreeding effects due to altitudinal origin of pollen suggests that local adaptation is not occurring in those populations. I conclude that alleles from other populations can be integrated easily which increases the reaction potential of plant populations in the face of global warming

Revue:

Thèse de l'ETH ZURICH

Lien:

e-collection.library.ethz.ch/eserv/eth:7590/eth-7590-02.pdf