The French National Research Agency Projects for science

Voir cette page en français

ANR funded project

Comprendre et prévoir les évolutions de l'environnement (DS0101)
Edition 2014


Adaptation and Resilience of Spatial Ecological Networks to human-Induced Changes

Adaptation and Resilience of Spatial Ecological Networks to human-Induced Changes (ARSENIC)
Most studies in predictive ecology rely on a static view of the ecological niche. ARSENIC aims at understanding how ecological systems react to global changes when ecological niches are considered as dynamic, varying with community composition and evolutionary processes.

Objectives of the project
An evolutionary objective is to understand how evolution affects trait associations among species, and how such variation in multi-trait phenotypes affects the characteristics of the ecological networks, population densities and interaction strengths. This relies on the development of evolutionary models based on trophic or mutualistic interactions that will explicitly incorporate various stressors (e.g. climate change, eutrophication, metal pollution). An ecological objective is to understand the resilience and robustness of ecological networks to external disturbances, and to investigate the effects of keystone species extinction in coevolved networks. We also intend to develop spatially explicit versions of models to evaluate how species distributions vary with global changes, by considering the coevolution of ecological networks along environmental clines affected by external disturbances. A final objective will be to go beyond theoretical developments and to provide an integrative picture that incorporates empirical aspects. To do so, we will compare predictions on trait distributions with empirical observations from large databases and from newly field-collected network and phenotypic data. We propose to study coevolved networks and the phenotypes of interacting species on two types of gradients: along a 1000 km long climatic gradient, we will study the interactions and traits of calcareous grassland plant-pollinator communities; along local heavy metal pollution gradients close to mines, we will study the interactions of metallicolous and non-metallicolous plant populations with their herbivores and pollinators.

Methods and work packages
ARSENIC is a collaborative project involving four main institutions (EEP, IEES, ISEM and CEFE) as well as several satellite groups. The management is shared by François Massol (EEP, Lille) and Nicolas Loeuille (IEES, Paris). The project is divided in two work packages that are balanced in terms of importance and time schedule. Work package 1 focuses on the necessary theoretical developments associated with the project. Work package 2 focuses on the collection and analysis of relevant data sets and on their use to explicitly test the results obtained in WP 1. Because the two work packages are intimately linked within the complete project, the two coordinators of the whole project will co-manage each. Each of two work packages is divided in four tasks. In WP1, the first two tasks aim at understanding the consequences of eco-evolutionary dynamics on trait distributions and network architectures. The first task focuses on food webs, while the second one tackles pollination networks. The third task focuses on the understanding of the dynamical properties of coevolved networks. The fourth task extends these models to incorporate spatial aspects to understand species’ range shifts in the face of global changes, given network and evolutionary constraints. In WP2, the first two tasks are explicitly devoted to confronting model results to empirical patterns as observed in large, existing databases. The first task focuses on food webs, while the second one focuses on pollination networks. Tasks three and four organize the collection of new data on two types of environmental gradients, with task three comparing pollination networks and plant and pollinator traits in calcareous grasslands along a latitudinal climatic gradient, and task four comparing plant, pollinators and herbivores traits and interaction networks along a heavy metal pollution gradient.


The theoretical model developed to understand the evolution of food webs under different temperature has yielded potentially interesting results, but yet to be confirmed in the coming months. To sum up what we have found so far, the model shows that: (i) species size in food webs tend to decrease with increasing temperature, even in the absence of direct effects of temperature on the rates of species interaction. This effect, which agrees with many empirical observations, is only the consequence of the link between temperature and mortality/respiration rates and carrying capacities of primary producers, as well as the size-based structure of predator-prey interactions; (ii) under the same relatively general assumptions, food webs are expected to be shorter (fewer trophic levels) and wider (more species in the first trophic levels) with increasing temperature; (iii) food web instability, as measured by the coefficient of variation of total biomass, tends to increase with temperature. Overall, this still unpublished work bears conclusions of potentially very general scope, which might revolutionize the understanding of Bergmann’s rule in food webs and modify predictions pertaining to the persistence of interaction networks under global changes.


A short term perspective of current work is to extend the model of food web evolution to other types of interactions. Using this new model, we will be able to predict how plant-pollinator networks evolve on gradients of primary productivity and under different climates. Another immediate modeling perspective relies on making current model spatially explicit using diffusion equations, so that we can use it to predict how interacting species ranges shift in response to climate change. On the empirical side, our immediate current perspectives revolve around the use of existing databases to identify trait correlations within food webs and plant-pollinator networks, and relate these correlations with environmental variables such as climate or primary productivity. We are also currently in the process of obtaining data on plant-pollination networks in calcareous grasslands. Based on this data, we will be able to test predictions of the evolutionary model that is going to be developed in the months to come.

Scientific outputs and patents

Massol, F., Dubart, M., Calcagno, V., Cazelles, K., Jacquet, C., Kéfi, S., & Gravel, D. (soumis) Island biogeography of spatially structured food webs. Advances in Ecological Research, 56 – papier soumis pour un numéro invité sur les réseaux d’interaction, à paraître fin 2016/début 2017 Kamenova, S., Bartley, T., Bohan, D., Boutain, J. R., Colautti, R. I., Domaizon, I., Fontaine, C., Lemainque, A., LeViol, I., Mollot, G., Perga, M.-E., Ravigné, V., & Massol, F. (soumis) Invasions toolkit: current methods for tracking the spread and impact of invasive species. Advances in Ecological Research, 56 – papier soumis pour un numéro invité sur les réseaux d’interaction, à paraître fin 2016/début 2017 Pantel, J. et al. (soumis) 14 Questions for Invasion in Ecological Networks. Advances in Ecological Research, 56 – papier soumis pour un numéro invité sur les réseaux d’interaction, à paraître fin 2016/début 2017 Geslin, B., Gauzens, B., Baude, M., Dajoz, I., Fontaine, C., Henry, M., Rollin, O., Thébault, E., & Vereecken, N. J. (soumis) Massively Introduced Managed Species and their consequences for plant-pollinator interactions. Advances in Ecological Research, 56 – papier soumis pour un numéro invité sur les réseaux d’interaction, à paraître fin 2016/début 2017 Allhoff, K. T. & Drossel, B. (2016) Biodiversity and ecosystem functioning in evolving food webs. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 371. David, P. Thébault, E., Anneville, O., Duyck, P.-F., Chapuis, E., & Loeuille, N. (soumis) Impacts of invasive species on food webs: a review of empirical data. Advances in Ecological Research, 56 – papier soumis pour un numéro invité sur les réseaux d’interaction, à paraître fin 2016/début 2017 Romanuk, T., Loeuille, N. et al. (soumis) XX [titre non définitif]. Advances in Ecological Research, 56 – papier soumis pour un numéro invité sur les réseaux d’interaction, à paraître fin 2016/début 2017


CEFE Centre d'Ecologie Fonctionnelle et Evolutive

CESCO Centre des sciences de la conservation

GEPV Génétique et Evolution des Populations Végétales

HYAX Hydrobiologie

IEES Paris Institut d'Ecologie et des Sciences de l'Environnement de Paris

ISA Institut Sophia Agrobiotech

ISEM Institut des sciences de l'évolution - Montpellier

UT Austin College of Natural Sciences

ANR grant: 498 682 euros
Beginning and duration: octobre 2014 - 48 mois

Submission abstract

Anthropogenic environmental changes increasingly threaten biodiversity and ecosystem services, thus kindling a societal demand for predictions that ecology as a science has yet to answer. Available models are poorly suited to predicting the ecological effects of such changes because they ignore variation in species’ niche due to ecological interactions and evolution. Without understanding the functioning of ecological networks and how they are shaped by evolution, it is indeed difficult to predict how changes of the environment will cascade through ecosystems and make species traits evolve. Understanding the dynamics of ecological networks is a dual goal, both for fundamental research and for building informed programs on sustainable ecosystem services and species conservation. Accounting for species interactions and evolution to understand the consequences of global changes is the critical question we want to tackle through an integrative approach that combines coevolutionary models and analyses of empirical datasets, existing or to be obtained through the project actions. We will tackle these challenges by capitalizing on the strengths of a pluridisciplinary consortium with highly complementary skills, from theoretical modelling to expert field work and taxonomical identification.
First, we will build evolutionary models of spatially structured antagonistic and mutualistic networks to understand how evolution affects (i) the association of traits in interacting species, (ii) the dynamical properties of ecological networks, and (iii) the dynamics of species’ ranges when embedded in a network of interactions. These models will help understand how climate change, eutrophication and pollution affect the properties of interaction networks through the evolution of interaction strengths and species dispersal abilities. These models will hint at how and why some specialization traits can be evolutionarily associated with higher dispersal, and thus will suggest trait associations within and across networks that can then be tested. Effects of natural selection on the stability of feasible ecological equilibria will also be studied in the context of May’s diversity-stability paradox. Finally, studying the dynamics of species’ ranges in networks will allow us to make predictions of how trait evolution shapes the geographical distribution of mutualistic or trophic partners.
Second, we will test these models (i) on existing databases on traits and interactions and (ii) using new observations. Existing databases on food web and plant-pollinator networks, on species traits and on geographical distributions will be tapped to uncover correlations between species traits and their position within networks, and among traits in species with different trophic niches or different degrees of specialization. We will also perform field surveys to test predictions of our models. The first survey will assess plant-pollinator interactions and their associated traits along a thousand-kilometre long latitudinal gradient of calcareous grasslands. Focusing on a few entomophilous, widely distributed plant species, we will identify and measure relevant traits of their associated pollinating fauna. Plant traits will be measured along the gradient to uncover associations between trait values and breadths of the pollinator spectrum. The second survey will compare herbivore and pollinator communities associated with metallicolous vs. non-metallicolous genotypes of the plant species Noccaea caerulescens. We will assess the links between the abundances of pollinators and herbivores, differences in heavy metal accumulation abilities and variation in self-fertilization rates in plants.


ANR Programme: Comprendre et prévoir les évolutions de l'environnement (DS0101) 2014

Project ID: ANR-14-CE02-0012

Project coordinator:
Monsieur Francois Massol (Génétique et Evolution des Populations Végétales)

Project web site:


Back to the previous page


The project coordinator is the author of this abstract and is therefore responsible for the content of the summary. The ANR disclaims all responsibility in connection with its content.