A picit jeu: Agent-based modelling with serious gaming for a fire-resilient landscape.

Wildfire governance requires addressing driving physical, biological and socio-economic processes, by promoting the development of fire-resistant and resilient landscapes. These landscapes can best be achieved by strategies that integrate fuel management for direct prevention with allied socio-economic activities, through the collaboration of stakeholders with different and sometimes conflicting interests. This work aims to address the need for new approaches supporting the participatory process of collective decision-making, helping stakeholders explore land management strategies for landscape fire resilience. We present and discuss a methodology combining agent-based modelling with a role-playing game. It was tested in a valley of the Italian Alps, involving 23 local stakeholders in forest and pasture management in three game sessions. Evaluation was based on observation of game sessions, collection of feedback via immediate post-session debriefing and questionnaires, and long-term (multi-year) assessment carried out through semi-structured interviews. We found the methodology valuable for facilitating discussion among different stakeholders, who were able to identify context-related challenges (land fragmentation and land abandonment, stakeholders' limited collaboration, controversial drives of European funding) and possible strategies for producing a fire-resilient landscape (community management forms of pastoralists activities for maintaining land cover diversity). The approach also triggered a positive process for longer-term change. By analysing the outcomes, we are able to identify four key recommendations for future work using serious gaming for sustainable landscapes: 1) aim for an even composition of session groups, 2) consider the multiple levels of organisation in the area, 3) use the allocation of game roles to disrupt power dynamics, and 4) seek to involve the broadest stakeholder spectrum in developing the game itself.