Research activities

The main focus of the group concerns the understanding of the interplay of population structure, mating systems and selection. For this, we use different approaches, from theory and the development of statistical tools to field observations via experimental work. The main biological models are different species of freshwater snails and a gynodioecious perenial plant, Silene vulgaris. On the theoretical side, we investigate the dynamics of multilocus genetic system under the influence of selection and drift, and develop statistical methods to infer mating systems and population structure.

Mating systems and population structure

Multi locus interactions, population bottleneck and genetic variance

Statistical methods & Software developments

Jérôme Goudet

Mating systems and population structure

Evolutionary biologists struggle to explain why so many species reproduce sexually. Among these sexually reproducing organisms, many reproductive systems exist, ranging from separate sex to full hermaphroditism via some intermediate stages, where some individuals harbor the two sexes while others are unisexual. One characteristic of reproductive systems where at least some individuals harbor simultaneously the two sexes is the ability to self-fertilize. Selfing can be advantageous, since it alleviates the cost of producing males and of finding a mate, provides a reproductive insurance, and could preserve successful genotypes. But it can also be detrimental since one of the consequences of selfing is a higher homozygosity, that can translate into inbreeding depression. Our goal is to gain a better understanding of the factors promoting selfing in natural populations. Species where the proportion of selfing varies among individuals and populations are particularly well suited to study the relative importance of these different factors.

But this goal can only be attained if we have access to sufficient genetic information. The increasing availability of genetic markers, inherited either uniparentally (e.g. the cytoplasmic genome), or biparentally (most nuclear DNA), has opened new avenues to the evolutionary ecologist. Not only critical parameters such as relatedness, coancestry or selfing levels can be readily estimated from field data, but models of population structure have been greatly refined since the insightful ideas of Wright for partitioning genetic variation among components reflecting population structure. In particular equations for gene correlations have been derived, that include different elements of the social structure found in vertebrates (polygamy, multiple mating, preferential dispersal of one sex, variance in reproductive success). One goal of the research group is to adapt this model to understand the fine scale population structure of plants and hermaphroditic animals.

Galba truncatula

Collaborators: Sandrine Trouvé, Elodie Chapuis, Loïc Degen and François Renaud (Montpellier, France).

Lymnaea truncatula is a self-fertile pulmonate snail found on most continents. It is an intermediate host of many trematode parasites that castrate infested individuals. We are investigating the population structure of this snail in Switzerland, using microsatellite markers. We are in particular interested in the following questions:
· Does parasite prevalence affect the level of selfing in populations?
· Is selfing rate affected by the temporal stability of the habitat, or its size?
· How much genetic exchange is there between the populations?

Silene vulgaris

Collaborators: Mélanie Glaettli, Daniel Croll, Nicolas Juillet, Luca Pescatore.

Silene vulgaris is a common weed that originates from Eurasia. It is classified as a poor competitor, which might be unable to resist invasion by species coming later in the succession. It has been classified as gynodioecious (some plants are hermaphrodites, other are females) by most authors, although a small percentage of gynomonoecy (hermaphrodite and female flowers on the same individuals) has been found. It is allogamous, but autocompatible and the many flowers per plant allow for some selfing. In large populations, the proportion of hermaphrodite is high, while smaller populations could be mainly females. Mode of inheritance of sex is complex with one cytoplasmic and up to four nuclear genes involved. Genetic structure in S. vulgaris have been investigated using nuclear allozyme markers, which prove polymorphic and inherited in a Mendelian fashion, and cytoplasmic chloroplastic markers, which also proved to be polymorphic. The project goal is to identify how at a very fine geographical scale, the different forces involved in shaping the genetic structure of S. vulgaris patches interact. To this end, we will use both field studies (2 populations with contrasted sex-ratio in each of 3 valleys), green-house experiments and modeling.

Multi locus interactions, population bottleneck and genetic variance

Collaborators: Yamama Naciri-Graven (Geneva) and Alex Ding.

Both theoretical and empirical studies have pointed to the puzzling result that while bottlenecks (the reduction in size of a population) diminish genetic variability, in some cases they lead to an increase in additive genetic variance. Epistasis (the non-additive action of genes at different loci on one character) seems to play a major role. Using multilocus models, we are trying to understand

• What type of genotypic values lead to stable polymorphic equilibrium?
  
• Can multilocus epistasis help in maintaining additive genetic variance following bottlenecks?

We've answered partially the last question (yes, it can!), but we need to add selection to the model. Also, matrices of genotypic values are generated at random for the moment, but eventually, we want to link our work with metabolic control theory.

Statistical methods & Software developments

Collaborators: Nicolas Perrin, Peter Waser (Purdue, USA) , François Balloux (Cambridge, UK), Guillaume Evanno, Sébastien Regnault

Molecular markers are plentifull nowadays, and of different types. The type of information that these markers can provide for understanding different aspects of the ecology of species is enormous, but largely unexplored. For instance, we were recently able to show that sex-biased dispersal can be detected using genetic markers. Statistical tools for analysing genetic data are often complex, and require specific packages. To this end, I developed the package FSTAT, which estimates and tests gene diversities and F-statistics, among other things. Another package on which I am working is PCAGEN , which carries out Multivariate Analysis on genetic data. These softwares are under constant development.

When applied to individuals, multivariate analysis allows the identification of clusters of individuals. I am currently investigating how a coupling of Multivariate analysis and clustering can help in identifying clusters of individuals.

Once the tools have been developed, they need testing. We are also investigating using computer simulations how these tools perform

Sandrine Trouvé

EVOLUTIONARY CONSEQUENCES OF SELFING IN THE FRESHWATER SNAIL GALBA TRUNCATULA

Collaborators:Elodie Chapuis

Mating system evolution remains a hotly debated topic in evolutionary biology. Species that sometimes realized uniparental reproduction, sometimes biparental reproduction are of great interest to investigate this exciting thematic. In this context, self-compatible hermaphroditic organisms offer the opportunity to investigate the evolution of selfing versus outcrossing. The cost of outcrossing and inbreeding depression are thought to be the two major factors affecting the evolution of selfing. In populations of simultaneous hermaphrodites, alleles for self-fertilization enjoy an inherent transmission advantage over alleles for outcrossing. Consequently, a mutation that increases the selfing rate is expected to spread in outcrossing (or partially outcrossing) populations.

Despite this inherent transmission advantage, self-fertilization is relatively rare. Some evolutionists still wrestle with arguments concerning the adaptiveness of this mating system and argue that selfers are in an evolutionary dead end, doomed to extinction. The basic ideas of this argument are that selfers present a reduced genetic variability, carry more deleterious mutations and might loose the ability to revert to cross-fertilization. On the contrary some authors predicted that a little sex is sufficient to enjoy all its advantages (increased rate of evolution, reduced genetic load…). Despite decades of theoretical and empirical work on the evolution of sex, it is difficult to determine whether selfing is viable as a long-term reproductive strategy.

If many studies considered self-fertilization and its evolutionary consequences in plant species, in animals, where hermaphroditism occurs in most metazoans groups, information only starts to accumulate and much more work remains to be done. These studies are particularly interesting in snail populations since numerous genera of gastropoda are concerned by hermaphroditism.

We investigate these contradictory hypotheses on the hermaphroditic freshwater snail Galba truncatula. In a recently submitted paper, we demonstrated that this species is a preferential selfer, although among population variations in selfing rate were detected. The actual study investigates (1) if inbred populations of G. truncatula presents a reduce variability, (2) if inbred populations of G. truncatula presents higher genetic load, and (3) if inbreeding depression is reduced in this species allowing selfing to be favourably selected.

Keywords: Mating system, hermaphrodite, genetic variability, genetic load, inbreeding depression, microsatellites, freshwater snail, Galba truncatula

Elodie Chapuis

Many plants and some animals are hermaphrodites and can therefore self-fertilzed. Selfing presents avantages like reproductive assurance and not the cost of outcrossing. Despite those, selfing is not the major mating system choosen by organisms. The reason often argued for this state is the presence of inbreeding depression. Inbreeding depression is, in general, viewed as the loss of fitness of inbred individuals compared to outbred individuals. The freshwater snail Galba (Lymnaea) truncatula represents a good model for studying the evolution of mating system. Indeed, this specie is a preferential selfer but selfing rates can vary between populations. In this regard, the purpose of this study is to determine the evolutionary consequences of selfing in Galba truncatula:

1. do high-selfing populations have reduce variability (neutral and adaptative characters) ?
2. do high-selfing populations have a larger genetic load ?
3. does inbreeding depression occurs in this specie ? Which magnitude ? Small amout of inbreeding depression could explain the existence/persistence of high selfing rates.

Alex Ding

The number of loci coding for a trait, their individual effects, and their interactions compose the genetic architecture of a trait. It influences adaptive processes. I am interested in the influence of the genetic architecture on the behaviour of adaptive processes. What are the equilibrium and dynamical properties of a biological system in function of the genetic architecture? I use theoretical approaches to gain some insights.

Two systems are used: (i) gametic frequencies under selection and (ii) known metabolic chains. The behaviour of such system is complex and not well understood (except in particular cases). When a trait is composed of one biallelic locus, heterosis is the only architecture that maintains polymorphism at equilibrium. It is not the case for less simple systems. What kinds of architecture maintain polymorphism? How does the architecture influence the rate of adaptation? How many equilibria exist? How far from reality are both systems?
Data on natural of artificial selection using real biological organisms can shed some light on the whole problem.

Keywords: multiloci system, metabolic chain, adaptation, natural & artificial selection, discrete system, non-linear system

Guillaume Evanno

My Phd aims at studying biodiversity at both genetic and specific levels. Patterns of biodiversity are more often considered at the community level and commonly addressed questions include : How the relationships between species (i.e. competition, predation, niche complementarity…) affect community stability? How does specie loss affect ecosystem functioning ? However, within the community each species is represented by a population whose genetic properties may greatly influence the maintenance of the species and therefore its role in the community. Populations with high genetic diversity may better respond to environmental changes than less diverse ones, which are ultimately more prone to extinction.

The knowledge of the genetic diversity of all species constituting an assemblage is hardly achieved but one can investigate this issue in a significant sample within a community of few species.
I focus on a guild of freshwater snails (about 12 species) living in an alluvial habitat. We are developing AFLP (Amplified Fragment Length Polymorphism) markers for 3 species: Radix balthica, Bithynia tentaculata and Planorbis carinatus. Several natural disturbances occur in our study site including floods and temporary dryness. As a result, there is a mosaic of habitats characterised by differences in the structure of their snails communities. I want to answer the following questions:

- Are there any relationships between species and genetic diversities of a community?

-What environmental factors influence populations' genetic diversity? I focus on the role of disturbances and parasites but confounding factors include population size, mating system, historical determinants…

Mélanie Glaettli

Spatial and temporal variations of population demography and their effects on the overall genetic structure and fitness are parts of my research interest areas.

For my master degree (directed by J. Goudet), I used individual based computer simulations to infer whether past bottleneck could be detected on the basis of allele frequency data only. The main idea was to test whether subsequent population growth would erase any signature of bottlenecks in the allele frequency data.

I am now involved in a project dealing with the demography and genetic structure of Silene vulgaris. This common weed is gynodioecious: populations bear both hermaphrodite and female individuals. Sex ratio in natural populations varies between 100% hermaphrodites up to 90% females. Geitonogamy can occur and some components of gender fitness are known to be frequency dependent.
In order to investigate the genetic structure, we sampled 30 populations within 3 valleys. We also sampled seeds from these population and are currently doing a greenhouse experiment (with controlled pollinations) to investigate the magnitude of inbreeding depression throughout the life cycle of hermaphrodites from these valleys.

Daniel Croll

CCOMPARING GENE-FLOW AT THE CYTOPLASMIC AND NUCLEAR LEVEL IN POPULATIONS OF SILENE VULGARIS

Sex determination in several gynodioecious species has been shown to depend on cytoplasmic male sterility (CMS) factors and nuclear restorer genes. The studied populations of Silene vulgaris vary considerably in their sex ratio, as it has been observed in other gynodioecious species. A potential explanation for the variable sex ratio is variation at sex-determining loci. Therefore, population substructure and its role in the maintenance of genetic polymorphisms may constitute an important factor in the dynamics of the gynodioecious system. Furthermore, given the fact that sex determination may depend both on cytoplasmic and nuclear loci, gene flow at these two different levels needs to be considered. Nuclear genes are potentially spread through pollen as well as seeds, while cytoplasmic genes depend entirely on seed dispersal. This asymmetry in the dispersal ability of nuclear and cytoplasmic genes may contribute to the dynamics of sex-determination in the populations of Silene vulgaris.

The aim of my research project is to develop chloroplastic markers for the studied Silene vulgaris populations. The focus will be on the PCR-RFLP technique to detect polymorphisms in non-coding regions of the chloroplast. Primer pairs developed for the study of intraspecific variation will be tested. Chloroplastic marker may then be used to assess the degree of genetic structuration in the populations and to compare gene flow at the cytoplasmic level with the nuclear level (estimated by microsatellite markers, Nicolas Juillet). Furthermore, the association of chloroplast haplotypes and gender may be tested for the different populations and eventually be used to conduct controlled crossing experiments.

Nicolas Juillet

Chez les espèces gynodioïques, les individus hermaphrodites et femelles coexistent dans les mêmes populations. Le maintien des individus femelles dans les populations implique en théorie un avantage sélectif important pour les femelles, pour compenser la perte de la fonction mâle. Des études récentes montrent que même avec un avantage sélectif faible, les femelles peuvent être maintenues dans les populations, à conditions que les flux de gènes entre les populations soient faibles. Pour tester cette hypothèse, nous avons étudié la structure génétique de 36 sous-populations réparties dans trois vallées de l'Ouest suisse, à l'aide de 4 marqueurs microsatellites.
Cette analyse montre une structure faible entre sous-populations (Fst = 0.03), et très faible entre populations ou entre vallée. Ceci suggère des échanges génétiques importants entre les unités d'échantillonnage, probablement dû aux mouvements de pollen. Ces échanges pourraient permettre la propagation des facteurs restaurateurs de la fertilité mâle entre les populations, et donc tendraient à faire disparaître les femelles.
En revanche, à cause de la faible dispersion des graines et du comportement des pollinisateurs, il semble que l'autofécondation et les croisements entre individus apparentés soient fréquents dans nos populations (Fis = 0.315, taux d'autofécondation estimé = 50%). Ceci pourrait être une des causes du maintien des femelles dans nos populations, puisqu'une dépression de consanguinité marquée est visible sur la germination des graines issues d'autofécondation (expérimentations en serre).
Pour poursuivre ce travail, nous lançons actuellement un travail de module visant à estimer directement le taux d'autofécondation dans les populations naturelles, par l'analyse de tableaux de descendants (génotypage de plantes et de quelques unes de leurs graines, produites par fécondation libre sur le terrain). Une autre partie des graines de chaque capsule sera mise en culture sous serre puis génotypée au stade plantule et / ou adulte, donnant des indications sur la dépression de consanguinité aux différents stades de vie des plantes en fonction du type de croisement dont elles sont issues.

Luca Pescatore

Compétition pollinique chez Silene vulgaris : incompatibilité cryptique et dépression de consanguinité (introduction)

Mon travail s'insère dans le cadre d'une étude (celle de Mélanie) qui vise à mettre en évidence de la dépression de consanguinité à chaque étape du cycle de vie de la plante. Le cycle de vie d'une plante peut être divisé en deux étapes : une phase diploïde (sporophyte) qui est prédominante chez les angiospermes (c'est la plante elle-même) et une phase haploïde (gamétophyte) qui, chez les angiospermes, est très éphémère mais pourtant pas moins importante du point de vue de la fitness. Le gamétophyte mâle (pollen) a le rôle de transporter le gamète mâle vers le gamète femelle, qui lui se trouve dans l'ovule (gamétophyte femelle).

Fertilisation : de nombreux grains de pollen sont déposés sur le stigmate (par un pollinisateur). Ceux qui arrivent à s'y accrocher pourront ensuite germer -> formation du tube pollinique. Le tube pollinique croit le long du style, jusqu'à l'ovaire, et permet au gamète mâle de passer dans un ovule. Or, le nombre d'ovules est limité. Si un excès de pollen est déposé sur le stigmate, seuls les tubes polliniques les plus performants pourront rejoindre la base du style : il y a compétition pollinique.

L'importance évolutive de la compétition pollinique est assez évidente : lorsque plusieurs types de pollen sont présents sur le même stigmate, le résultat de cette compétition déterminera qui pourra féconder des ovules. Le but de mon travail est de mieux comprendre qui va gagner, et ceci grâce à deux expériences.

1. Incompatibilité cryptique

La première expérience vise à mettre en évidence une relation (positive, j'espère) entre la distance de croisement de deux plantes et la performance des tubes polliniques. L'idée est que dans un même " receveur " (mère) le pollen provenant d'une plante géographiquement éloignée (>10m) croît mieux que le pollen d'une plante géographiquement proche (<3m), et que ce dernier croît mieux que le pollen provenant du receveur même (self-pollen). Si un tel phénomène est mis en évidence, on peut supposer que la cause est une relation positive entre la distance géographique qui sépare deux plantes et la distance génétique entre elles (Nicolas a déjà mis en évidence que d'autres populations de S. vulgaris possèdent une structure à petite échelle). Pour tester cette dernière hypothèse, je génotype les individus à l'aide de quatre marqueurs microsatellites. Si l'hypothèse est confirmée, on aura mis en évidence une forme d'incompatibilité qui peut diminuer la proportion de croisements consanguins. Or, cette espèce ne présente à la base pas de systèmes incompatibilité (si seul du self-pollen est déposé sur le stigmate il pourra féconder les ovules sans problèmes). Donc, " incompatibilité cryptique " car elle n'est mise en évidence que lors de la compétition pollinique.

2. Dépression de consanguinité

Il s'agit de mettre en évidence la dépression de consanguinité au tout dernier stade de vie de la plante : le gamétophyte. Les questions posées sont : (1) est-ce que le pollen produit par une plante issue d'autofécondation est moins performant que le pollen produit par une plante issue d'allofécondation ? (i.e. est-ce que la dépression de consanguinité affecte la croissance du tube pollinique ?) et (2) Est-ce qu'un tube pollinique croit mieux dans le style d'une plante issue d' allofécondation que dans le style d'une plante issue d' autofécondation? (i.e. est-ce que la dépression de consanguinité affecte la performance du style ?)
N.B. : on ne parlera pas de dépression de consanguinité pour le pollen (car son génome est haploïde, et donc ne peut être consanguin) mais pour la plante dont il est issu. Une partie du génome diploïde de la plante " père " est exprimé lors de la croissance du tube pollinique. Le style, lui, il est diploïde.