How integration of socio-ecological data can shape regional environmental management decisions: an example from Australia

In this article, we explore how sensitive recommendations that guide transferability of environmental management solutions are to the type of data used. Working with an integrated Australian data set containing over 200 variables, we use clustering techniques to identify similar regions. We find that variables that drive cluster membership come from all three data domains: biophysical (climate, extreme events, type of vegetation, community and species richness, habitat condition), social (political orientation, personal, household, economic characteristics, built infrastructure, Indigenous governance, land tenure) and interactions (adaptive capacity, disaster resilience, land use, grant value and ecosystem services). We demonstrate that regions cluster differently when only biophysical or only interaction or only social data are used in the analysis. We therefore argue that policy-makers need to be clear about what their policy seeks to achieve and by which mechanisms, before exploring data to find regions to where the policy can successfully be transferred. To use information that only describes one part of the interconnected system is to risk overlooking or misunderstanding other parts of it. Policy makers require information about social and ecological systems and about interactions between systems in order to better understand, analyse and take action to improve the state of environment.本文探究了指导环境管理解决方案跨区域迁移适配性的推荐方案对所采用数据类型的敏感性。本研究依托包含200余个变量的整合型澳大利亚数据集，运用聚类技术识别相似区域。研究发现，决定聚类归属的变量涵盖三大数据范畴：生物物理（气候、极端事件、植被类型、群落与物种丰富度、栖息地状况）、社会（政治倾向、个人与家庭经济特征、建成基础设施、原住民治理、土地权属）以及交互作用（适应能力、灾害韧性、土地利用、资助金额与生态系统服务）。研究表明，若分析中仅采用生物物理数据、仅采用交互作用数据或仅采用社会数据，区域聚类结果会存在显著差异。因此本文提出，政策制定者在通过数据筛选可成功适配该政策的区域之前，需明确政策的目标与实现机制。仅采用描述互联系统单一维度的信息，可能会忽视甚至误判系统的其他组成部分。政策制定者需要掌握社会与生态系统及其交互作用的相关信息，才能更好地理解、分析并采取行动改善环境状况。