Interactions between farming systems and landscapes at various scales: a data-mining approach of international literature
Résumé
Agricultural activities play a major role to design rural landscapes, through interactions
between actors and resources. In this paper, we provide an outlook on the interactions
between farming systems, landscapes and land use among the international scientific
literature. We made the hypothesis that the majority of studies dealing with landscape
generally do not take into account agricultural practices and farm management. This result in a potential bias in future landscape designing. We proposed a three-step methodology: The first step consisted of a manual data-mining based on an ontology of key-words built by expert knowledge. A temporal analysis of the
key-words was also performed with some bibliometric tools. In a second step, we analysed each study case according to a grid composed of four axes: i) Framework of the study; ii) level of land use and farm management analysis; iii) spatial and temporal scales considered; iv) extension to a modelling approach. Finally, we used a Multiple Correspondence Analysis then an Ascendant Hierarchical Classification on the corpus of selected papers to build a typology of articles. 210 papers were selected and a typology was built. The combination of expert knowledge and statistical methods into data-mining approaches revealed that few studies connected the dynamic - in space and time - of crop distribution and management with landscape and resources management whereas the number of studies on landscape dynamic is increasing. Landscape designing requires that agronomists pay more attention to the articulation between landscape management, resources management and farming systems