COS 93-4 - Gone with the wind: Linking environmental, trait and phylogenetic data with spatial occurrences to build Joint Species Distribution Models in Subantarctic Marion Island

Background/Question/Methods

Species distribution modelling (SDMs) are powerful tools that can inform on ecosystems spatial patterns and processes. In order to better understand the functional structure of ecosystems, modelling techniques ought to incorporate information about species traits and interactions. The subantarctic region comprises island archipelagos with native flora exposed to extreme environmental conditions. As a result, its vegetation has adaptations to cope with low temperatures and strong winds. Marion Island (Prince Edwards – South Africa) is a remote island semi-isolated from the rest of the globe with a unique native flora that merits the highest status of conservation. Even this semi-pristine ecosystem pressures exist, namely changing climates, and the introduction of non-indigenous flora and fauna. It remains to be assessed how environmental factors relate to trait variation and whether these differ for native and alien species.

Results/Conclusions

Using joint species distribution models, we modelled the distribution of plant species on Marion with joint SDMs using environmental factors such as temperature, elevation, wind and geology, but also trait data, phylogenetic relationships and species co-occurrences. By incorporating trait data into the models, we could assess how mean community traits such as plant height and leaf area correspond to environmental predictors and how trait distribution in varies space. We can also infer biotic associations that go beyond similar environmental occupancies and lastly generate a bioregionalization classification from where to definite vegetation types. Overall, the hierarchical modelling of species communities provides a robust array of methodologies towards advancing ecological modelling into a more mechanistical understanding of macroecological responses when aligned with robust ecological data. In this regard, the characterization of Marion Island dynamic flora, singular traits and extreme environmental conditions (including hard-to-retrieve variables such as wind speed) sets an excellent study case to approach these questions.