2MODELS · Allochronic and ecological models of sympatric speciation of plam trees on a small isolated island

Allochronic and ecological models of sympatric speciation of plam trees on a small isolated island

In this project, we aim at testing two competing scenarios to explain the emergence with no geographic isolation of two palm species on a remote pacific island (Lord Howe Island), i.e. the sympatric speciation of Howea belmoreana and H. forsteriana: one ecological scenario in which assortative mating results from adaptation to divergent ecological niches and one allochronic scenario in which assortative mating results from temporal segregation of reproduction among individuals. Test of the two scenarios relies on both experimental and theoretical work. Using the models built during this project, we will also reveal processes that promote or hinder sympatric speciation in flowering plants in general. The main experimental goals of the project were to estimate some key parameters involved in the speciation of the palm trees: (i) pollen and seed dispersal to gauge gene flow and colonisation abilities of the species, (ii) hybridisation rate between species to establish at which stage isolating mechanisms occur, (iii) fitness and germination rates of each palm species in different environments to measure current adaptation, (iv) current densities of the two palm species to find their ecological niches. The main theoretical goals were to find the effect of the above parameters as well as the evolution of flowering time on the likelihood of speciation; it first necessitated building a model designed for the model species. Both experimental and theoretical progress has been made and results have been obtained during the first year of the project. Two field trips to Lord Howe Island allowed measuring fitness on ∼400 adult trees, setting-up a seed transplant experiment, and collecting palm material to estimate hybridisation rate; an experimental protocol is now ready to be performed to estimate jointly pollen and seed dispersal for each palm species. The generation of a huge data base containing the position of all adult trees of both H. belmoreana and H. forsteriana, information on their ecology and the soil type on which they grow across the entire Lord Howe Island allowed identifying that H. forsteriana is not specialised on calcareous soils, as previously thought, whereas that H. belmoreana is indeed restricted to volcanic soils. Fitness measures also confirm that H. forsteriana is as fit on calcareous soils as on volcanic soils. Further, statistical analyses of the current distribution of the two species across the island will soon help determining the ecological niches of the two palm species and thus the (ecological) dimension(s) along which the species diverged. Knowing this diverging ecological dimension is crucial for building realistic ecological scenarios of sympatric speciation investigated by the theoretical part of the project. Experimental data on flowering phenology of both palm species collected by the hosting group were also analysed during the first year of the project. They show variation among sites, and potentially along an altitudinal gradient, and large variation among individuals within species with both the existence and absence of differences between male and female flowering period, a result that has implications for refining the allochronic model tested in this project. All the above data and results helped guiding the scenarios to be investigated in the theoretical study of this project. A model has now been built, using fast calculation algorithm supporting large population sizes and exploration of a big parameter space. This model is inspired by one published by collaborators but now incorporates new components that better account for the evolution of flowering time and for gene flow via pollen and seed dispersal and that will soon better account of the topography of Lord Howe Island.

Data: CORDIS, © European Union