Abstract

The human dimension is one major source of uncertainty in the management of social-ecological systems such as fisheries. Agent-based models (ABMs) can help to reduce these uncertainties by making it possible to model and simulate human behavior. To understand how ABMs can be applied in fisheries science, a classification scheme was developed based on reviews in other social-ecological domains, theoretical frameworks, and documentation standards. This classification scheme was subsequently used to review agent-based simulation studies that modeled human decision-making in a fisheries context to identify trends and knowledge gaps. Applying the classification scheme revealed that the existing fisheries-related ABMs employ a variety of decision theories, policies, social interactions, agent memories, and data sources, and revealed a wide potential for applications of ABMs to a broad range of research questions and management recommendations. Nevertheless, it turned out that it is, so far, virtually unexplored how environmental factors influence fishing decisions or how social norms and learning influence fishing behavior. It also became clear that the documentation and provenance information of ABMs need to be improved – e.g., by applying standardized documentation procedures, such as ODD + D and TRACE – to enhance the credibility, transparency, and reusability of ABMs in fisheries science.

摘要

人类因素是渔业等社会生态系统管理中不确定性的主要来源之一。基于代理的模型（ABM）可以通过对人类行为进行建模和模拟来帮助减少这些不确定性。为了了解 ABM 如何应用于渔业科学，根据其他社会生态领域的评论、理论框架和文件标准制定了分类方案。该分类方案随后用于审查基于代理的模拟研究，该研究模拟了渔业背景下的人类决策，以识别趋势和知识差距。应用分类方案表明，现有的与渔业相关的 ABM 采用了各种决策理论、政策、社会互动、代理记忆和数据源，并揭示了 ABM 在广泛的研究问题和管理建议中应用的广泛潜力。然而，事实证明，到目前为止，实际上尚未探索环境因素如何影响捕捞决策或社会规范和学习如何影响捕捞行为。很明显，ABM 的记录和来源信息需要改进，例如，通过应用 ODD + D 和 TRACE 等标准化记录程序，以提高 ABM 在渔业科学中的可信度、透明度和可重复使用性。