Climate change, plant functional traits and Diversification
High diversification rates are the result of the right combination of plant functional traits and environments. This implies that we should find a correlation in shifts in diversification rate, plant traits and environments. This is part of a larger research program in which we seek to understand the diversification of Angiosperms during the Cenozoic. The research involves building large phylogenetic trees, measure plant traits (both in the Gardens and in the Herbarium), correlating these to past climate changes as well as spatial shifts, and matching the patterns to the fossil record. The project integrates phylogenetic, palaeobotanical and ecological research: it is classical macro-evolutionary / macro-ecology project. The Masters project involves testing which characters allowed the temperate diversification of Salix from its tropical ancestors. Additional factors include a co-evolution with sawflies and other herbivores.
Pampas Grass Systematics
The Pampas grasses (Cortaderia) are widespread in South America, and the genus includes both very range restricted species along the Atlantic coast, as well as very widespread species along the Pacific coast. The species delimitations are very difficult, and there is no key to all the species. This taxonomic project will be heavily based on herbarium collections, and will investigate the delimitation fo the species, their geographical ranges and ecology, and where possibly also their breeding systems. It may be possible to build in a field-trip to South America.
The grasslands in Madagascar are reputed to be recent, the consequence of humans bringing fire into the system. However, there are many indications that this is wrong, and that at least some of the grasslands are ancient, presumably dating to the Miocene. This project is in collaboration with researchers at the Royal Botanic Gardens in Kew. We will model the niches of selected species, and interpret niche evolution from the models and the phylogeny. The project should also include extensive fieldwork in Madagascar. Students doing this project will learn grass systematics, ecological niche modelling, phylogenetic analysis, and mapping diversity patterns.
Alpha, Beta and Gamma Diversity patterns in the Afroalpine flora
The spatial and ecological patterns in biodiversity can be separated into three components. Alpha-diversity measures the richness in one community; beta-diversity measures the change in diversity along an ecological gradient (thus how much diversity is explained by niche specialization); and gamma diversity measures the replacement along a spatial gradient (thus how much diversity is explained by allopatry, species occupying the same niche, but in different places). Recently these measures have been linked to evolutionary patterns, thus giving insight into the environmental parameters that might drive in situ diversification.
The Afroalpine ecosystem is a superb place to explore these parameters. The whole flora is relative small (360 species) and well known. The vegetation communities can be readily classified into five types. And finally, the ecosystem is spatially fragmented into eight mountain blocks, separated by hundred to thousands of kilometers. It is therefore easy to separate the spatial (gamma) and ecological (beta) components.
The ecosystem has been intensively studied over the past five years, resulting in a major dataset of plot data, which can be used to explore the alpha, beta and gamma components of this diversity. For some clades, phylogenies are also available, making it possible to add an evolutionary component.
Questions / hypotheses:
- Of the total diversity, what proportion is alpha, beta and gamma. Alpha is the richness within each site; beta is turnover between habitats on a mountain, gamma is turnover between mountains. Turnover is calculated using Beta-sym. Since the data are presence / absence, we can only calculate turnover. Jaccard can be used to test whether sites are nested. Note that the plots are not in a transect, so it is more a chi-square than a regression approach. We predict that turnover should be smaller between habitats on one mountain, than between mountains for the same habitat, due to ecologically widespread species.
- By adding a phylogenetic component to Beta-sym (weighting the similarity by the taxonomic affinity), we test whether phylogenetically related taxa are more likely to be within one habitat, than on one mountain. The prediction is that the within-habitat relationship will be closer, than the within mountain relationship. This is consistent with phylogenetic niche conservatism.
- Where are the sister-species of eco-geographic endemics (species restricted to one habitat type on one mountain or mountain system)? This links to the island speciation models: is speciation largely sympatric (sisters in different habitats on the same island) or allopatric (sister in the same habitat, but on different islands). Not sure how many cases we will find. This is the same as "species replacement"
- Is there evidence of filtering for particular habitats? This can be measured by testing whether all habitats contain either a random sample of the growth-forms in the Afroalpine, or indeed all of them. If filtering occurs, then the growth-spectrum should be non-random. These data can be readily obtained from the floras, and also from our DELTA dataset.
- Is there evidence of phylogenetic structuring of the communities? We have two species pools: the lists for the mountain, and the list for the Afroalpine. The structuring can be tested at community and at mountain level. Possibly at Afroalpine spatial scale community, filtering could be expected. As we will not yet have a full phylogeny at species level, we can assemble a phylogeny at generic level, which should still give us adequate resolution.