• Home
• News
• Organisation ▼
  • Who are we?
  • Research themes
  • Scientific committee
• Program ▼
  • Program
  • Downloadable program
  • Presentations - Download
• Pratical Information ▼
  • Important dates
  • Oral comunications
  • Venue
  • Accomodation in Grenoble
• Submit
• Registration
• Contacts

1. Macroecology, macroevolution and spatio-temporal patterns of biodiversity

Macroecology and macroevolution are two disciplines that focus on the searching and understanding the types, distributions, abundances and diversity of species and biomes, from the local to the global scale, as well as developing and testing theoretical explanations for the observed patterns.

The two modelling approaches most commonly used to investigate these issues – the modelling of spatial patterns of biodiversity and the modelling of evolutionary, or temporal, patterns of biodiversity – are still poorly integrated. However, the biodiversity patterns we can observe today are the result of both ecological and historical processes; thus, a common theoretical framework is needed to integrate these two approaches and enable testing more robust hypotheses to explain the phenomena originating biodiversity patterns.

A better integration of current knowledge on ecology on macroevolutionary models, and of concepts, models and techniques stemming from macroevolution studies on ecological models is, therefore, needed. The objective of the GDR is then to coordinate and accelerate research already taking place in France that allows integrating macroevolutionary and ecological community models, in order to develop tools for modelling of biodiversity that better take into account historical and ecological factors shaping present biodiversity patterns.

2. Interaction networks

Ecological networks allow describing interactions between species in ecosystems, being at the basis of numerous fundamental questions in ecology, be it in respect to understanding the mechanisms that determine interspecific interactions, or to predict the consequences of these interactions for ecosystem dynamics and their response to perturbations (species loss, etc.). The study of networks is also strongly linked to the debates revolving around the relationship between the complexity and stability of communities.

The majority of studies on ecological networks focuses on particular types of interactions, studied independently from each other: trophic networks (predatory interactions), host-parasite networks, mutualistic plant-pollinator networks, etc. Also, certain types of interactions and groups of species have been ignored in most empirical and theoretical approaches. Therefore, most of studies of trophic networks do not consider interactions with abiotic components of ecosystems, like nutrient recycling and interactions between green (based on primary producers) and brown resources (based on decomposers). Network approaches have also systematically ignored the most biodiverse group of organisms, microorganisms, particularly prokaryotes (bacteria and Archaea), and their associations with larger organisms.

It is thus presently necessary to develop a more integrative approach for the study of ecological networks that allows a better consideration of the diversity of organisms and their interactions within an ecosystem. To do this, research on ecological networks requires not only new and more complete datasets, but also new theory that allows capturing the ecological and evolutionary consequences of the diversity of interactions within a network.

Considering this research theme, the objective of the GDR is to coordinate and develop research already underway in France on emergent questions concerning ecological networks. In particular, this working group aims to develop research towards a more integrative approach to the study of networks considering the diversity of interactions in ecosystems.

3. Diversity and stability of ecological systems

A large number of theoretical and experimental studies have unequivocally demonstrated that losses in biodiversity have important consequences for the functioning of ecosystems and the services they provide to societies, such as the reduction of the capacity of communities to use available resources, convert them to biomass and decompose and recycle nutrients.

A growing number of experimental results further shows that biodiversity increases the stability of ecosystem properties. This stabilising effect is particularly interesting because it suggests that biodiversity plays a key role in the sustainability of ecosystems and, consequently, of societies that depend on them, in the face of environmental change.

While the currently dominant ecological theory, developed in the 1970s, predicted that the diversity and complexity of communities leads instability, rather than stability, recent experiments have shown that species diversity often has opposite effects at the population (where the effect is often destabilising) and ecosystem levels (where the effect is usually stabilising), a result that classical theory does not explain.

It is therefore necessary to build new theory that can explain and predict experimental results, but also allows better understanding and predicting the capacity of ecosystems to resist disturbance. A theory of ecosystem stability centred on the temporal and spatial variability is necessary. The aim of the GDR on this theme is to coordinate and accelerate research already underway in France for developing new theory on the stability of ecological systems and its relationship with biodiversity that is, at the same time, able to clarify and predict empirical data and experimental results.

4. Spatial dynamics, metacommunities and metaecosystems

Concepts and models linking local and regional dynamics are defined using the prefix “meta”, by analogy with the first model developed, the metapopulation model (a population of populations). A metapopulation is a spatial network where each node can host a population of an organism. By extension, a metacommunity is a spatial network where each node can host a community, that is to say a set of different species sharing limiting factors. The approach “meta” aims to answer common questions such as the stability or coexistence of species in communities, but also to more applied issues such as the functioning of ecosystems, the emerging topology of interaction networks and conservation biology.

Conversion of ecosystems and the destruction of natural habitats have considerably increased during the last three centuries. These environmental changes are the major contributors for the biodiversity current crisis and the degradation of ecosystem services. The framework proposed by the spatially implicit “meta” approaches can address the dynamics of such changes in fragmented habitats and thus produce methodological tools adapted to conservation issues.

“Meta” approaches have greatly diversified behind these years, exploring food web spatial ecology, ecosystem functioning and biogeography. For example, considering the dynamics of nutrient recycling no longer a the a local scale, but at the metaecosystem scale, while taking into account the flow of detritus and nutrients, can lead to more complete theory and predictions on the dynamics of these systems. Integrating trophic networks in a spatially structured system also allows developing a better theory and taking into account the spatial and temporal variability of the observed food webs. Finally, metacommunity approaches are also beginning to incorporate genetic evolution.

The aim of the GDR in respect to this theme is to coordinate and stimulate theoretical research carried in France in the fields of metacommunities and metaecosystems, and to improve the understanding of the spatial dynamics of biodiversity, food webs, ecosystem functioning, species evolution and the effects of global change.

5. Predictive models of biodiversity changes

Making realistic predictions on a set of environmental issues is one of the biggest issues in ecological modelling. Two practical examples of these problems have been widely discussed in the literature over the past decade, illustrating the current limits of ecology models.

The first example concerns species distribution patterns, which have been widely studied to attempt at predicting future species distributions under climate change. Most approaches are statistical, meaning that they are based on the correlation of current occurrences of a species and climatic conditions, which combined, can create a bioclimatic envelope. These models are not based on a mechanistic understanding of the processes underlying the distribution range of changes, such as physiological processes, biotic interactions, or those related to the dispersal potential. Despite being powerful, these statistical approaches fail to make robust predictions for poorly known species or to consider species evolutionary responses.

A second example concerns predictive models of ecosystem response to climate change. These models have been developed mainly by climate modellers and take into account the physiological response of plants to changing water cycles, carbon and nutrients (e.g. Orchid, LPJ, ...), which could radically alter the future dynamics at the global scale. However, responses of decomposer groups are oversimplified and, more generally, effects of trophic levels other than plants (e.g. microbial decomposers and arthropods or vertebrate and arthropod herbivores) are generally ignored in these models. Such a simplification may be acceptable as a first approximation, but as soon as carbon, water, nitrogen or phosphorus cycles are considered, ignoring heterotrophic organisms is ignoring a large and essential part of the functioning of these cycles.

The current limits of the two approaches outlined above are partly inevitable. Yet, the essential problem is that of predicting population dynamics (in the first case) or ecosystems (in the second) at large spatial scales, while avoiding losing detail at the individual level, which allows the organism to ensure its reproduction, dispersal, or acclimatisation to environmental conditions. The GDR group working on this topic aims to facilitate a discussion within the French research community to perform several actions. Firstly, it will focus on developing general theoretical background to derive macroscopic models from individual-based or population-based models. Second, it will examine key issues for which the research teams involved have known expertise, including dynamic models of species distributions and forest dynamics models, and explore how these approaches can be implemented. Finally, together with the group from theme 6 and the GDR Écologie Statistique, it will explore the inference methods for setting and validating these models.

6. Linking data and models

There is a variety of ecological perspectives on the role of data in the modelling process. At one extreme of the gradient, data can be a source of inspiration. They are used to extract general patterns that theoretical studies will then aim to explain. At the other end of the gradient, data are seen as objects that models must reproduce finely. Along this continuum, a variety of practices exists, which use a different data and generate a different methods of analysis of these data, such as identifying characteristic patterns, the qualitative comparison between model predictions and data, or quantitative statistical inference (and validation) from potentially heterogeneous, noisy and/or partially informative data on the processes under investigation.

When fitting model predictions to data is the goal, we can question what degree of difference between observed and predicted data is acceptable. Although statistical methods such as “model checking” techniques can be mobilized to test the fit between models and data, the question on the how tolerant comparisons between models and data should be remains. Another open question is the optimal complexity of ecological models. Is it possible to formalise implicit arguments made by ecologists during the construction of their models? Investigating this question could help developing useful tools for ecologists and anticipate mistakes and errors when designing models linked to real data.

Finally, the use of computer simulations is an increasing trend among ecologists, which can be both very exciting, because it allows efficiently exploring hypothesis that have been understudied, but also a potential source of confusion if algorithmic hypothesis and details are not rigorously presented and if the robustness of the results to these algorithmic choices is not analysed. This poses new theoretical questions on how to steer the use of computer simulations in theoretical ecology.

The thematic working group on this topic proposes to carry out a series of review papers addressing a number of issues surrounding modelling in theoretical ecology. By bringing together ecologists with different appreciations on the use of data in theoretical ecology, this GDR group will seek consensus among ecologists, rather than seeking to oppose complementary approaches, as has been done recently in the literature.