How do farmers’ representations influence landscapes? A multi-scale approach combining mental modelling and forest monitoring in South-western France

In French agricultural landscapes, rural forest (including small private woodlots, hedgerows, and isolated trees) is crucial for landscape functioning and related ecosystem services. Farmers own and manage rural forests according to their individual objectives and rationality, which is more complex than the sole economic profitability. Social networks, socio-cultural norms and legislative limitations are also at play. Rural forest management, characteristics and evolution, at the farm and landscape scales, may then be linked to farmers’ representations of this complex social-ecological system. We used mental models, elicited during individual interviews in South-western France, to assess how farmers perceive ecosystem services and disservices provided by rural forests, and the social and legislative entities influencing their management. On a focus area, a monitoring of rural forest evolution between 1962 and 2010 was performed through photo-interpretation of diachronic images. Twenty individual mental models were elicited during interviews with farmers and a classification of ecosystem services, disservices and stakeholders was operated to allow comparisons and analyses. A total of 31 services were reported, with an average of 7.2 (±1.9 SD) per farmer, including 10 provisioning, 9 regulating, 10 cultural and 2 support services. Fourteen (14) disservices were identified, with an average of 3.2 (±0.5 SD), essentially related to agricultural, economic and legislative issues. Hedgerows and isolated trees were the main sources of ecosystem disservices, while woodlots were mainly associated with services. This result supports the trends obtained by photo-interpretation, i.e., a decline of hedgerows and isolated trees and a maintenance of woodlots over the last half-century. In complement, farmers with mixed crop-livestock systems listed more services and had more complex mental models (more entities and more links) than farmers who have only crops. Rural forest perception then seemed to vary with farming systems. Diverse social actors were reported, including farmers and their family, local and regional forestry companies, and environmental agencies. However the implications of these actors were highly variable between farmers, according to their needs in firewood, the amount of hedgerows or the woodlot areas they owned. These preliminary results show the potential of mental models to assess farmers’ representations and their impact on rural forest management at farms’ scale. An on-going work will permit to assess the influence of these contrasted management practices on rural forests itself, through on-farm forest measurements. Finally, the aggregation of individual mental models and of forest measurements may allow to scale-up our results to landscape scale. The semi-quantitative analyses allowed by mental modelling may be of great interest to link social and ecological processes, a crucial challenge for scientists working in managed landscapes.