Background. In my previous SNF project (MODIPLANT; grant nr 110000), niche-based species distribution models were successfully developed for predicting the fate of nearly 300 mountain plants in face of severe climate change. Scenarios revealed a great sensitivity of the alpine flora, with many high elevation species at severe risk of extinction. However, biotic interactions were not taken explicitly into account in these projections. It is currently debated whether changed biotic interactions may also change – and if so, to which extent - the outcome of such projections. To test this hypothesis, biotic interactions need to be more explicitly incorporated into the modelling process. In particular, future plant communities need to be predicted by selecting those species potentially co-occurring from a larger pool of candidate species predicted at a suitable site. Interactions with other organisms, such as pollinator insects, also need to be considered.
Specific aims. In this follow-up project, we aim at incorporating (1) plant-plant interactions (assembly rules) and (2) plant-insect interactions into niche-based statistical models of species distribution, and test whether their inclusion can change the outcome of projections in a warmer future. The two subprojects will run in parallel and share data.
Methods. A large field survey involving four teams will be conducted throughout the study area to complement existing vegetation plots and additionally sample soils and the entomofauna. In the first subproject, we will look for patterns of plant co-occurrences and test if and how identified interactions can be used to filter predictions – using both bottom-up assembly and top-down controls (e.g. species-energy) - of plant communities, and modify their future composition and structure. In the second subproject, we will look for patterns of plant-insect co-occurrences, and similarly assess whether a change in their respective distributions may lead to limited matching of plant and specialist insects in a warmer future, and as a result to disruption of some communities or ecosystems.
Funding requested. Funding is requested for only one PhD student (in subproject 2), three year of GIS technician (at 50%) and some months of field and lab technicians. The second PhD student (in subproject 1) will be granted by UNIL as matching funds. A postdoc from the ECOCHANGE EU-project will also actively collaborate on the project.
Expected value of the proposed project. With this project, we aim at providing: (i) a better understanding of biotic interactions (especially between plants and insects, and within each group) and how they shape species distribution, (ii) an improved approach to modelling biodiversity, that considers both top-down controls on and bottom-up assembly of communities, and (iii) ecologically more realistic projections of future plant distributions. Although not its primary aim, this project will also contribute indirectly to (iv) build a comprehensive inventory of plant and insect species in the Western Swiss Alps, and (v) as all plots will be marked in the field with buried metallic bars, our sampling will also contribute to the set-up of an impressive biomonitoring network in this area, which will be available in the future to test model predictions.
Alpine ecosystems were identified as potentially very sensitive to climate change. For instance, it has been hypothesized that alpine plants with narrow niche should be at greatest risk of extinction. However, we hypothesize here that projections made for these species at the European scale can differ greatly from those obtained at finer scale, e.g. due to the presence of micro-topographic refugias, and thus, species' turnover calculated on large scale may be entailed with errors. We also hypothesize that lack of consideration for validation and uncertainty prevents proper interpretation of model projections, and that some biological traits and dispersal ability may explain species' vulnerability.
This project aimed primarily at providing more informative predictions of climate change impact on alpine plant species distribution. By doing so, it contributed to improve our fundamental knowledge of species distributions and related ecological processes. The project had five main aims: 1-compare projections at various scales, 2-assess model robustness, 3-take dispersal into account, 4-estimate uncertainty, 5- assess species' sensitivities.
A large field survey had already been conducted during the summers 2002-2004 and 550 nested vegetation plots (including all vascular plant species, nested surfaces of 1, 4, 16 and 64 sq-m) are now available between 450 and 3200 m over a study area of 700 sq-km (Swiss Western Alps). These were used in conjunction with a GIS environmental data base to fit our models and test our hypotheses.
BIOASSEMBLE: A. Guisan, A. Dubuis, L. Pellissier, P. Vittoz, with help from L. Maiorano
MODIPLANT: C. Randin, P. Pearman, R. Engler, Y. Hautier, R. Milleret, G. Vuissoz, H. Jaccard