Work Package 3:
Coevolution, Invasion and Plant Conservation Biology
COORDINATORS Isabelle Olivieri (UMR 5554 CNRS Montpellier) and Bruce Anderson (Department of Botany and Zoology, University of Stellenbosch)
Project 1: Gene flow, life-histories and sexual dimorphism in plants of a fynbos endemic plant genus, Leucadendron (Proteaceae).
Participants : Isabelle Olivieri, Agnès Mignot, Ophélie Ronce, Sandrine Maurice, Emmanuel Douzery (UMR 5554 CNRS, Montpellier), Tony Rebelo (Kirstenbosch Botanical Garden, Rondebosh), Jeremy Midgley, Nicola Illing (University of Cape Town)
The genus Leucadendron is made of 96 taxa, half of which are threatened (two already extinct). This genus is highly diversified from the point of view of its ecology, mating sysem, dispersal structures, adaptation to fire. For instance some species have diaspores dispersed by ants, others by wind (wings or parachutes), others just roll on the soil thanks to hairs. Some species are polymorphic for seed coat or flower colour, others possess different types of dormancy (serotinous species which seed dispersal only occurs after a fire, or soil seed bank with germination induced by fire). Pollination is either insect or wind mediated. All species are dioecious, but they present various degrees of sexual dimorphism.
We now have collected several individuals of most taxa, and are about to sequence the whole transcriptome for three of them, plus two outgroup species. This will allow us to design ca. 30 molecular primers specific to the genus and related genera, that will be very useful to reconstruct the molecular phylogeny of the genus (the one presently available is not robust at all, and botanists strongly disagree with it), hence addressing questions about macroevolutionary patterns of life-history evolution.
Thanks to the molecular markers that we have now developed (microsatellites) we will address a number of questions relative to speciation process, and the ecological and evolutionary significance of sexual dimorphism. We will for instance determine both pollen and seed dispersal kernels in species differing by their amount of sexual dimorphism.
Using common garden experiments, we will compare various quantitative traits and molecular diversity of several populations of two species, closely related on the available phylogeny, but very different for life-histories, and for the amount of phenotypic diversity. This will allow us to address questions about the first step of potential ecological speciation
Finally, modelling the evolution of resource allocation to growth, survival and reproduction, we will ask about the role of the fire regime in determining the optimal life-history, and about the ecological and evolutionary consequences of changes in the fire regime.
Project 2 : Pollination-induced speciation in plants.
Participants : Isabelle Olivieri (UMR 5554 CNRS, Montpellier), Bruce Anderson, Anton Pauw (University of Stellenbosch).
One particularly attractive hypothesis to explain the extraordinary speciation rates in the south-african fynbos, is that soil substrate or climatic factors governs the distribution of pollinators, which then governs the amount of gene flow among populations. In the genus Leucadendron and Hemimeris, parapatric populations occur on different substrates. There is some evidence that in Leucadendron salignum, different flower colours, occuring mostly in parapatric situations, are associated to different pollinators, both soil types and climatic adaptation seem to be involved. Populations of Leudadendron platyspermum and of L. meridianum also occur on parapatry on very different soil types, and are both insect pollinated, but which pollinators occur is unknown. Could they help reinforce the local adaptation pattern likely to occur on such different substrates? In Hemimeris racemosa, there is also some evidence that populations are normally outcrossing, but that populations occuring on sand have lost self-incompatibility dimorphism and instead have evolved towards selfing. This could be due to a change in the type of pollinators. If such soil substrate differences is indeed at the origin of speciation mediated by pollinator types or availability, one expect such evolution of reproductive isolation to often occur. Therefore, evolution towards self-compatibility and selfing is expected to have evolved several times. This should translate into parapatric populations on different soil types to be more related than allopatric populations. In order to distinguish between current gene flow and recent common origin, we will compare genetic differentiation among populations of Hemimeris racemosa.
Project 3 : The Ficus-fig-wasp association: a model system to investigate coevolutionary processes.
Participants : Finn Kjellberg, Martine Hossaert-McKey (CEFE UMR 5175, Montpellier), Jean-Yves Rasplus (CBGP UMR 1062, Montpellier), Simon van Noort (IZIKO museum), Michael McLeish (University of Stellenbosch), Jaco Greeff (University of Pretoria).
Ficus (Moraceae) produce closed, urnshaped-inflorescences, called figs, the inside of which is lined by flowers. They are exlusively pollinated by Agaonid wasps (hymenoptera, chalcidoidea) that enter figs to oviposit. Each Ficus species is pollinated by one or a few generally species specific Agaonid wasps. Further a number of chalcid wasp species belonging to different families may parasitize the fig-pollinating wasp association. Some gall female flowers, some use galls initiated by other wasps, some are parasitoids and some may even consume seeds. Most of these wasps oviposit from outside figs, but some may enter the figs as pollinators. Most of these wasp species are assumed to specialize on a single host. There are over 750 species of Ficus distributed worldwide constituting a fantastic model to investigate the evolutionary history of 750 separate micro-hymenopteran communities.
The participants are involved in a long term joint research effort that will be facilitated by the GDRI. The study is aimed at understanding how and why the Ficus-fig-wasp system diversified. We are progressively revising the taxonomics of fig wasps and Ficus on a worldwide basis. We are also involved in a large joint long term phylogenetic project the aim of which is to understand how the communities diversified. We are studying the biology of the system in order to understand the ecological determinants of community structure.
The participants are involved in a global worldwide perspective but they also focus specifically on Africa. Indeed, in Africa there seem to be an exceptional diversity of cases of one wasp visiting several host Ficus and one Ficus being visited by several agaonid wasps. To unravel this situation we are building on a phylogeographic approach all over Africa and on more specific local studies on the biology of the interactions, especially in South Africa.
The participants are progressively developing the website, http://www.figweb.org/, which aims at providing worldwide information for the identification of all Ficus and fig wasp species. The aim is to show that it is feasible to provide complete identification assistance, including morphological and molecular data to identify all members of these communities, i.e. for about 10000 species. Figweb is hosted at Iziko museums, Cape Town, with a mirror site at CBGP, Montpellier.