Publications des agents du Cirad


Contribution of spatial statistical models in the study of land use change and regional development trajectories : the case of São Félix do Xingù in Southern Parà (Brazil)

Mertens B., Poccard Chapuis R., Piketty M.G., Venturieri A., Thales M.C., Tourrand J.F., Boutonnet J.P., Laques A.E.. 2001. In : International Conference on Land Use and Cover Change Dynamics (LUCCD'2001): Land Use and Cover Change project (LUCC), Beijing, China, August 26 to 30, 2001. s.l. : s.n., 17 p.. International Conference on Land Use and Cover Change Dynamics (LUCCD'2001), 2001-08-26/2001-08-30, Pékin (Chine).

The Amazon is the largest tropical forest area on earth, and is undergoing rapid deforestation since the last four decades. Large scale pasture for cattle ranching and soybean production are the main land uses, and are currently leading to a yearly deforestation of 0.5 percent of the forests. In Brazil, these conversions are mostly located in frontier areas distributed along the so-called "arc of deforestation" from the western State of Acre to the eastern State of Parà (through the States of Rondonia and Mato Grosso). This arc of deforestation is a large zone rather than a line: within this zone, various land use change processes are interacting between two modes of land valorisation and organisation. Thence, it is a large transition area where the observed landspace structure and dynamics are driven on the one hand by the heritage of the past and, on the other hand by the influence of the expanding anthropisation. From several case studies in the state of Parà (Brazil), along a gradient of anthropization, the current project aims at analysing how these landscape dynamics in contrasted frontier areas are related to the infrastructure development (roads and markets accessibility), the ecological conditions, the zoning policies and to the evolution and the organisation of the production, consumption and marketing chains of meat and dairy products. First, the project is trying to identify and model the land use change trajectories as measured through the use of satellite remote sensing (Landsat TM and ETM+) in combination with relevant spatial determinants integrated within a geo-referenced database (logistical multiple regression models): it aims at analysing whether the relation between deforestation and a set of explanatory variables varies depending on geographical scale and time period analysed. The interpretation of the results is carried out with regards to the main driving forces which have lead to specific land use dynamics (résumé de l'auteur)

Mots-clés : modèle; utilisation des terres; déboisement; développement agricole; impact sur l'environnement; changement structurel; télédétection; brésil; para; amazonie; modèle de régression multiple

Communication de congrès

Agents Cirad, auteurs de cette publication :