Context. What are the proximate mechanisms – at the individual level – underlying the assembly and dynamics of ecological communities? Despite its apparent universality, this question is still puzzling scientists – and challenges their methodological limits. Global changes impact community diversity and structure, but how these changes are rooted in individual dynamics remain poorly understood. Modeling the dynamics of ecological communities is difficult due to the complex dynamics of interacting species, and the need to integrate information across several biological levels (individuals – populations – communities). Yet, to shed lights on the processes underlying observed community modifications, knowing to what extent each species contribute to the direction and intensity of these modifications is of paramount relevance.
Objective. In the DEMOCOM project, we will develop an integrated statistical framework for estimation and inference about community dynamics via the development of a multispecies demography, bringing together concepts, predictions and tools from both fields of demography and community ecology. Demography aims at capturing variation in size, ages and ontogenic stages within a population. Although well developed for single-species dynamics, demography has received little attention in community ecology.
Work package 1 - DEMOCOM with pain and tears: modeling framework. We will develop a statistical framework to fit multispecies stage-structured models to data and partition the contribution of intra-specific and inter-specific demographic variations to changes in community composition. In particular, i) we will devise community sensitivity analyses to explore how small changes in the demographic rates of species and in the strength of interaction within and between species may affect the structure and resilience of communities, ii) we will extend our deterministic multispecies demographic models to incorporate stochasticity and fully account for environmental variability and iii) we will integrate mathematical structured multispecies models with statistical approaches to properly integrate the complexity of community dynamics and accommodate the various sources of uncertainty at play.
Work package 2 - DEMOCOM in practice: motivating case studies. Motivated by case studies on wild populations, we will investigate the role of inter-specific interactions in shaping the community-level response to climate change and management interventions in bird, fish and mammal communities. The data are owned and collected by the partners of the consortium. Besides, these case studies have been carefully selected to start working with relatively simple systems in which we have a few species and explicit interactions (seabirds and tunas), then to proceed with more complex systems in which we have more species with many possible interactions (common birds and large carnivores and their ungulate preys).
Work package 3 - DEMOCOM for dummies: software development. Third, we will develop in concert with end-users an original, generic and user-friendly software (in the R statistical platform) to assist biologists in investigating communities’ structure and dynamics through multispecies demography modeling. Overall, we will adopt a multidisciplinary approach in which important questions in community ecology will be addressed with robust and modern methods from demography, statistics and computing science.