Landscape Ecology 18: 303–314, 2003. © 2003 Kluwer Academic Publishers. Printed in the Netherlands. 303 Research article Temporal variability of connectivity in agricultural landscapes: do farming activities help? Jacques Baudry 1,* , Françoise Burel 2 , St´ ephanie Aviron 2 , Manuel Martin 2 , Annie Ouin 2,3 , Guillaume Pain 1,4 & Claudine Thenail 1 1 INRA – SAD Armorique, CS 84215, 65 Rue de Saint Brieuc, 35042 Rennes C´ edex, France 2 UMR 6553 Ecobio, CNRS – Universit´ e de Rennes 1, Campus de Beaulieu, 35042 Rennes Cedex, France 3 INP-ENSAT Avenue de l’Agrobiopole, B.P. 107, Auzeville Tolosane 31326, Castanet, Tolosan Cedex, France 4 Ecole Sup´ erieure d’Agriculture, 55 rue Rabelais, BP 748, 49007 Angers Cedex 01, France * Corresponding author: (Tel.: 00 33 (0)223485621; Fax: 00 33 (0)223485620; E-mail: jbaudry@roazhon.inra.fr) Key words: connectivity, farming system, landscape structure, simulations, temporal variability Abstract In landscapes where natural habitats have been severely fragmented by intensive farming, survival of many species depends on connectivity among habitat patches. Spatio-temporal structure of agricultural landscapes depends on interactions between the physical environment and farming systems, within a socio-economic and historical background. The question is how incentives in agricultural policies may influence connectivity? May they be used to manage the land for biodiversity conservation? We used simulations based on property field maps to compare connectivity on the same landscape during seven years of crop succession for two dairy farming systems, one being representative of conventional systems of western France, the second one representative of systems undergoing intensification of production. Connectivity is a measure of landscape structure and species characteristics based on individual area requirements and dispersal distance. Models used are based on weighed distances, considering differential viscosity for different land uses. The results show that, for a given farming system, physical and field patterns constraints are such that landscape connectivity remains the same over years, while it is significantly different between the two farming systems. This is consistent with the recent input of policies to promote environ- mentally friendly farming systems, and confirms that policies must encounter the landscape level. The analysis also demonstrates that the localisation of forest patches, resulting from long term land cover changes, plays a central role in connectivity and overrides changes in agricultural land uses. Introduction The main characteristic of temperate agricultural land- scapes is the expansion of farmland at the expense of forest. In western Europe this process has been taking place for centuries (Duby and Wallon 1975). For plants and animal living in these fragmented and heterogeneous landscapes, movement is a key process for survival (Wiens et al. 1993). For mobile species, daily movements are for food search and predator avoidance. For all species, individuals or propagules disperse among local populations or to colonise new habitats, at a time scale of a year or more. Before the period of land use intensification grew from the 1950s on, patches and linear elements of semi-natural habitats facilitated the movements of some species. Since then, these habitats have decreased dramatically in intensively farmed regions (Leonard and Cobham 1977; Agger and Brandt 1988; Meeus 1990). Nev- ertheless, numerous forest species continue to thrive in these environments (Burel 1996). Landscape con- nectivity, defined in this paper as ‘the degree to which the landscape facilitates or impedes movement among