The multifunctionality of forest systems calls for appropriately complex modelling approaches to capture social and ecosystem dynamics. Using a social-ecological systems framework, we review the functionality of 31 existing agent-based models applied to managed forests. Several applications include advanced cognitive and emotional decision-making, crucial for understanding complex sustainability challenges. However, far from all demonstrate representation of key elements in a social-ecological system like direct interactions, and dynamic representations of social and ecological processes. We conclude that agent-based approaches are adequately complex for simulating both social and ecological subsystems, but highlight three main avenues for further development: i) robust methodological standards for calibration and validation of agent-based approaches; ii) modelling of agent learning, adaptive governance and feedback loops; iii) coupling to ecological models such as dynamic vegetation models or species distribution models. We round-off by providing a set of questions to support social-ecological systems modelling choices.

中文翻译：

---

将森林建模为社会生态系统：基于主体的方法的系统比较

森林系统的多功能性需要适当复杂的建模方法来捕捉社会和生态系统动态。使用社会生态系统框架，我们回顾了 31 个应用于管理森林的现有基于代理的模型的功能。多种应用包括高级认知和情感决策，这对于理解复杂的可持续发展挑战至关重要。然而，远非所有这些都展示了社会生态系统中关键要素的表征，例如直接相互作用以及社会和生态过程的动态表征。我们的结论是，基于代理的方法对于模拟社会和生态子系统来说足够复杂，但强调了进一步发展的三个主要途径：i）用于校准和验证基于代理的方法的稳健方法标准；ii) 代理学习、自适应治理和反馈循环的建模；iii) 与生态模型的耦合，例如动态植被模型或物种分布模型。最后，我们提供了一组问题来支持社会生态系统建模选择。