Generating Advances in Modeling Biovidersity And ecosystem Services: statistical improvements and ecological relevance of joint species distribution models – GAMBAS

Global changes are causing deep modifications of ecosystems and their functioning, leading directly to species extinctions and functional shifts and indirectly to alterations of the ecosystem services they deliver. In the last decades, modelling the impact of global changes on biodiversity and ecosystem services has been an active area of research with strong expectations from conservation agencies, national and international programs, as well as biodiversity managers and conservation associations. However, most existing models still ignore basic mechanisms including biotic and abiotic interactions. Biodiversity is not merely the sum of species, but is also the result of interacting species that form assemblages, at various spatial and temporal resolutions, with different functional characteristics and evolutionary histories. Models ignoring these mechanisms are prone to provide erroneous predictions of how global changes will impact biodiversity or ecosystem services. Addressing these gaps requires a multi-scale approach that spans different fields of ecology and statistics with input from stakeholders and policy makers in order to ensure that the results from the theoretical research can realistically be implemented. Recently, tremendous progress has been made in extending species distribution models to Joint Species Distribution Models (JSDMs). These models predict species distributions based on environmental and spatial variables, but they also allow species to share information on their responses. JSDMs hierarchically model biodiversity from the species level to the functional group level, taking into account potential dependencies between species - for example based on their phylogenetic distances. JSDMSs expand and unify classical methods used in community ecology or in conservation biology and are currently used to model synergies and tradeoffs between ecosystem services.

However, as promising as they are, the large and extensive use of JSDMs is still hampered by several limitations both ecologically and statistically that GAMBAS proposes to address. First, there is a lack of a clear classification and nomenclature associated with the ecological and statistical assumptions on which they are based. Second, JSDMs provide estimates of correlation between species. But do these correlations indicate true interactions between species? Third, mathematical developments are required to elucidate links between statistical and ecological assumptions or to insure a more robust and efficient use. Finally, JSDMs, which currently remain an academic research topic, have rarely been confronted to practical problems such as the identification of conservation zones. In GAMBAS, we will focus on the analysis of five datasets, which will both be an inspiration and will constitute the first concrete case studies for our developments.

GAMBAS gathers a collective comprised of quantitative ecologists and mathematicians with aspirations in ecology. These investigators will be supported by an advisory committee composed of representatives of the French government and civil societies. Together, they have set the goal of expanding JSDMs and to promote their use in community ecology, biogeography, and conservation ecology.