Axis 1 : Spatial and temporal dependencies between environment, biodiversity, ecosystem functioning and nature’s contributions to people

We aim at building integrative datasets ...
The availability of biodiversity and ecosystem data is fast increasing, but this data is not yet well integrated. We collect experimental and observational data and adapt state of the art technologies to improve quantitatively and qualitatively the characterization of our study systems at different organisational (from individual to biome) and spatio-temporal scales (from local to remote sensing measures), including long-term observatories (e.g. Orchamp, Spipol, VigiNature). We combine multi-trophic and multi-functional biodiversity surveys (citizen-science data, metabarcoding and metagenomics, multispectral data, microbial activity analyses) with detailed inventories of abiotic variables (e.g. high resolution earth observation data, downscaled climate data) and ecosystem functions and nature´s contributions to people. We also assess how biodiversity and ecosystems, in particular through nature-based solutions, can contribute to adaptation to climate change. To be able to profit from these novel data we strongly invest in bio- and eco-informatics in relation with Big Data Science and Artificial Intelligence, and we also use social science research methods such as questionnaires, interviews and participatory workshops.

We use and further develop state-of-the-art modelling for these novel data …
We account for the inherent uncertainty of these data (e.g. environmental DNA based community descriptions, animal behavior assessed from high-frequency biologgers), and the breadth and complexity of information that they comprise. We advance models that (i) consider responses of multiple species, traits, ecosystem functions or nature’s contributions to people at the same time and model their dependencies (e.g. joint species distribution models, graphical models) to understand synergies and trade-offs between responses (e.g. identification of life-strategies through functional trait co-variations or of ecosystem service bundles) ; (ii) allow for dimension reduction when modelling multiple entities (e.g. plant functional group approach, latent variable modelling, graph embeddings) to represent the full system in a parsimonious way ; (iii) contribute to assess to what extent biodiversity and ecosystems can reduce climate change risks and thus enhance adaptation to climate change. Any relationships between environment and biodiversity or functions are scale dependent. We study which scales are relevant for which relationships to improve projections but also for building mechanistic hypotheses (e.g. from infra-second animal behavior inferred from new biologgers to individual habitat selection at broader spatio-temporal scales, to individual demographic performance and population distribution).