個體施為決策之土地利用變遷研究－以雲林沿海魚塭養殖為例

Agent-based Approach of Aquaculture Land Use Change Study-A Case of Yunlin County, Taiwan

陳超群(Chao-Chun Chen) ； 蔡博文(Bor-Wen Tsai)

《環境與世界》 12期 (2005/12) Pp. 93-119

https://doi.org/10.6304/EAW.2005.12.4

摘要

土地利用變遷一直以來都是「環境變遷」的重要議題，瞭解驅使土地利用變遷背後的作用力(driving forces)是目前「土地利用變遷」研究的核心。本研究藉由解析個人土地利用決策，探討造成土地利用變遷的主要作用力。在過去十年當中，雲林縣沿海四鄉－麥寮、台西、四湖、與口湖，許多養殖魚塭發生棄養情形。透過問卷調查，獲得魚塭棄養的影響因素（包括個人屬性、自然環境及社會作用力），並且以決策樹分析法找出棄養的規則。研究結果共發現有10組棄養的規則組合，其中在魚塭面積小於兩甲、有業外收入、年齡40歲以上、鹹水取得容易等四個作用力的組合，對魚塭棄養有決定性的影響。此外，研究結果顯示，決策樹分析方法是分析土地利用變遷作用力的有效方法。

關鍵字

土地利用變遷 ； 作用力 ； 決策樹分析

並列摘要

Land use and land cover change (LUCC) has long been a critical issue on the research of environmental change. The identification of driving forces becomes the core theme of LUCC research. This research employs agent-based approach to explore agents' decision-making process. Questionnaire interview is used to identify driving factors of abandoned aquaculture in the coastal area of Yunlin County, Taiwan. Decision-tree analysis is proceeded to delineate the decision process by the combination of driving factors. Ten decision processes are identified by the decision-tree analysis. The combination of: 1) agents whose raising area less than 2 acres, whose age over 40, and who have non-farming income; and 2) fish ponds that has high accessibility of salt water is the most significant decision process of abandoned aquaculture. The integration of questionnaire interview and the decision-tree analysis has shown to be a useful method for the LUCC research.

並列關鍵字

Land use and land cover change ； decision tree analysis ； driving force