Assessing the effect of landscape change on fauna by agent-based model simulation

Authors and Affiliations: 

Laurence Jolivet, IGN, Université Paris 1, France

Marianne Cohen, LADYSS, Université Paris 7, France

Anne Ruas, IFSTTAR, France


Analysing landscape changes and their effects requires retrospective studies that have been carried out in the past but cannot be repeated in the future. Planning authorities need to be able to assess the impact of their decisions on the current landscape. In France, the political measure called Trame verte et bleue incites the authorities to take into account the effects of planning on biodiversity1. Landscape ecology and modelling contribute to these objectives. We have tested a simulation method for fauna movements at a local spatial scale.


This abstract is about one of our studied cases i.e. the effect of spatial planning in a peri-urban area on movements of red fox that is a generalist species with a large dispersion capacity. The tested developments are: a) the construction of a highway b) the construction of a highway with wildlife crossings c) the construction of a highway with wildlife crossings connected to an ecological corridor. Movement data were provided by E. Robardet5 and mapped in a GIS with RGE® geographical data. We determined the features of the landscape around the trajectories based on GPS points. Then we integrated the knowledge on the landscape characterisation in an agent-based model. Movement simulation was made for one agent i.e. one fox, with regard to its behaviour and the elements that can favour or hinder movementsa.


The constructed highway limits the total distance covered, from 6.9 km to 6.5 km per day on average (crossing probability set at 10 %b). The animal movement decreased, illustrating a barrier effect: we used as an index the number of the simulated localizations on the same side of the road divided by the total number of localizations. In the initial space, the index was equal to 0.42, due to the spatial distribution of the animal’s areas of interest. The index increased to 0.55 with the highwayc. When the road benefited from wildlife crossing structures, the results were quite the same. In the third scenario, the ecological corridor alleviated the barrier effect: the index remained around 0.41. Simulation shows that the distance covered decreased less from 6.9 km to 6.8 km.


Agent-based models are often used for modelling the interaction between species and space3 as well as human-animal conflicts2. We use it to model the influence that landscape elements have on fauna movements at a fine spatial scale. This method allows for considering the effect of the road as being separate from its spatial environment6, contrary to the cost path methods that aggregate land use which can be better adapted to habitat studies4. Our results reveal that the protection of landscape quality through ecological corridors seems quite effective to preserve movement capacity of species like fox. Simulations launched for deers allows for assessing the effects of roads in a forest area. On the basis of all the results, we propose a general model for considering planning effects, which can be a helping tool for planning.


1Amsallem J., Deshayes M., Bonnevialle M. (2010). Analyse comparative de méthodes d’élaboration de trames vertes et bleues nationales et régionales. Irstea, Sciences, Eaux et Territoires, Volume 3, p. 40-45


2Anwar S. M., Jeanneret C. A., Parrott L., Marceau D. J. (2007). Conceptualization and implementation of a multi-agent model to simulate whale-watching tours in the St. Lawrence Estuary in Quebec, Canada. Environmental Modelling & Software, Volume 22, p. 1775-1787


3Hooten M. B., Johnson D. S., Hanks E., Lowry J. H. (2010). Agent-based inference for animal movement and selection. Journal of Agricultural, Biological, and Environment Statistics, Volume 15 (4), p. 523-538


4La Morgia V., Malenotti E., Badino G., Bona F. (2011). Where do we go from here? Dispersal simulations shed light on the role of landscape structure in determining animal redistribution after reintroduction. Landscape Ecology, Volume 26, p. 969-981


5Robardet E. (2007). Étude de la transmission d’Échinococcus multilocularis dans une grande agglomération : influence du comportement alimentaire et de l’utilisation de l’espace par le renard rous (Vulpes vulpes) sur la contamination de l’environnement. Mémoire de thèse en Sciences de la Vie et de la Santé, Université de Franche-Comté, 173 p.


6Vuillemier S., Metzger R. (2005). Animal dispersal modelling: Handling landscape features and related animal choices. Ecological Modelling, Volume 190, p. 159-170