The PAYSAGRI model: from agricultural plot to landscape

Depigny Sylvain, Poix Christophe, Chevillot B.. 2005. The PAYSAGRI model: from agricultural plot to landscape. In : Conference on Multi-agent modelling for environmental management, Bourg Saint Maurice - Les Arcs, France, 21-25 mars 2005. CABM-HEMA. Bourg-Saint-Maurice : s.n., 2 p. Colloque CABM-HEMA-SMAGET 2005, Bourg-Saint-Maurice, France, 21 Mars 2005/25 Mars 2005.

Matériel d'accompagnement : 1 poster

Résumé : Nowadays, social reorganization of rural territories is topical issue. More and more people, who don't belong to agricultural corporations, live on these territories. This phenomenon creates some disagreements. On the one hand, new residents think that landscape is only their living environment. According to them, landscape must be suitable for hobbies. New residents wish this landscape not to evolve. On the other hand, farmers don't understand these new expectations, which sometimes contradictory their activity. This situation results from a lack of knowledge about the processes of landscape evolution. For agronomist point of view, these processes are deeply linked to agricultural system changes. One cannot see landscape just as a picture of a territory, but more as the result of economic activities. The main objective of our model is to clarify and to explain relationships between public policies, agricultural system changes and landscape changes. Two stages are necessary for this modelling. Firstly, we must understand how public policies are applied and how these policies change strategies of agricultural systems. Secondly, we try to assess landscape evolutions, which are due to the reorganization of agricultural systems. Our modelling is based on medium altitude areas where the main part of land is managed by agricultural systems. The Paysagri model is based on three concepts: The landscape is the measured variable. Considering geoagronomy, landscape can be seen as a puzzle made of many pieces. According to this point of view, each piece is both a component and a consequence of a sociological and economical system. In our model, landscape is a puzzle made of all agricultural plots of the studied territory. The plot is the elementary object in the model. That's why it is assumed homogeneous; i.e. the model doesn't handle phenomena that occur inside the plot. We keep the only information that is essential for the farmer's decisions. Texture and age characterize a plot. The textures are “ natural meadow ”, “ artificial meadow ”, “ crop ” or “ unused ”. The transition for one texture to another is the result of agricultural practices on the plot during a year. The agricultural practices are the factors that modify the texture of the plots. In the present version of our model, we only implemented the dairy farming functioning. This choice is due to the fact that dairy farming is the most important activity of the studied areas (north of the Massif central). The model is based on the point of view of the breeder. The herd and its feeding are of great importance to the farmer's decision process. We modelled two kinds of farming activities: the usual ones and exceptional ones. The usual ones are fodder production and herd grazing. To do that, the breeder must carefully choose the plots for each activity, according to production capacities of each plot. The second ones correspond to the agricultural system adaptations. The breeder can give up some plots, for example because of their lack of production capacities. He can also use new plots that he recently inserted in the agricultural system or that were unused for some time. These exceptional practices are essential in the Paysagri model. They represent the most important variations of area used by the agricultural systems. In our model, a special attribute named “landscape perception” is associated with each breeder. That's why the breeder has a key role in his agricultural system changes. The “landscape perception” attribute guides the farmer's choice concerning the exceptional practices. This attribute is an original point of our model, because technical and economical facets of the farmers are not only aspects taken into account. The breeder's feeling about his occupation, his living environment, his way of life... are numerous elements considered in the model. Public policies are the first cause of agricultural system changes. They influence the breeder's strategy and the impact of the farming activities on landscape. Therefore, those policies often act indirectly on landscape. They are not included in the model, but there are outside factors that modify the behaviour of the model. To build Paysagri, we used an object oriented approach. Although our model is not a Multi Agent Model yet, our object-oriented approach will allow the implementation of agents as soon as it will be necessary. Presently, it is based on discrete event simulation method. We used the “Unified Modelling Language” to formalize the model. Then we chose the “Python” language for computer implementation, because it is a powerful object-oriented programming language with very clear syntax. Starting from a given landscape, the software performs simulations. Each year, it computes the choices of every breeder and their consequences on the territory. It produces a new landscape, which is the gathering of simulated plot textures. As an example, the simulation of 15 years, with 100 plots and 2 breeders takes about 3 minutes on a PIII 800 Mhz computer. We inserted the Paysagri model into a larger project that allows us to work with “Geographic Information System” data. We generate both maps or 3D realistic landscape images. We plan to use the data resulting from our model into two different ways. Firstly, we would like to use Paysagri as a research model. We think it could help to better understand the relationships between landscape production and agricultural system development. Secondly, results will be used as a basis for discussion in the frame of construction of shared territory projects. For example, our model could be used to put the emphasis on the farmer's role in landscape evolution.

Classification Agris : B10 - Géographie
A01 - Agriculture - Considérations générales
U30 - Méthodes de recherche
E14 - Économie et politique du développement
E90 - Structure agraire
L01 - Élevage - Considérations générales
L02 - Alimentation animale
P01 - Conservation de la nature et ressources foncières

Auteurs et affiliations

• Depigny Sylvain, ENITA (FRA) ORCID: 0000-0003-2207-9904
• Poix Christophe, ENITA (FRA)
• Chevillot B., ENITA (FRA)

Source : Cirad-Agritrop (https://agritrop.cirad.fr/582616/)

Métriques

[ Page générée et mise en cache le 2022-04-22 ]