Archetype analysis is a key tool in landscape and sustainability research to organize social-ecological complexity and to identify social-ecological systems (SESs). While inductive archetype analysis can characterize the diversity of SESs within a region, deductively derived archetypes have greater interpretative power to compare across regions. Here, we developed a novel archetype approach that combines the strengths of both perspectives. We applied inductive clustering to an integrative dataset to map 15 typical SESs for 2016 and 12 social-ecological changes (1999–2016) in Andalusia region (Spain). We linked these types to deductive types of human-nature connectedness, resulting in a nested archetype classification. Our analyses revealed combinations of typical SESs and social-ecological changes that shape them, such as agricultural intensification and peri-urbanization in agricultural SESs, declining agriculture in natural SESs or population de-concentration (counter-urbanization) in urban SESs. Likewise, we identified a gradient of human-nature connectedness across SESs and typical social-ecological changes fostering this gradient. This allowed us to map areas that face specific sustainability challenges linked to ongoing regime shifts (e.g., from rural to urbanized systems) and trajectories towards social-ecological traps (e.g., cropland intensification in drylands) associated with decreasing human-nature connectedness. This provides spatial templates for targeting policy responses related to the sustainable intensification of agricultural systems, the disappearance of traditional cropping systems and abandonment of rural lands, or the reconnection of urban population with the local environment, among others. Generally, our approach allows for different levels of abstraction, keeping regional context-specificity while linking to globally recognisable archetypes, and thus to generalization and theory-building efforts.

原型分析是景观和可持续性研究中组织社会生态复杂性和识别社会生态系统 (SES) 的关键工具。虽然归纳原型分析可以表征一个地区内 SES 的多样性，但演绎衍生的原型在跨地区比较时具有更大的解释力。在这里，我们开发了一种新颖的原型方法，结合了两种观点的优势。我们将归纳聚类应用于综合数据集，以绘制 2016 年的 15 个典型 SES 和安达卢西亚地区（西班牙）的 12 个社会生态变化（1999-2016）。我们将这些类型与人与自然联系的演绎类型联系起来，从而产生了嵌套的原型分类。我们的分析揭示了典型的 SES 和塑造它们的社会生态变化的组合，围- 农业 SES 中的城市化、自然 SES 中农业的衰退或城市 SES 中的人口分散（逆城市化）。同样，我们确定了跨 SES 的人与自然联系的梯度以及促进这种梯度的典型社会生态变化。这使我们能够绘制出面临与正在进行的制度转变（例如，从农村到城市化系统）相关的特定可持续性挑战的区域，以及与人类与自然联系减少相关的社会生态陷阱（例如，旱地的农田集约化）的轨迹。这为与农业系统的可持续集约化、传统耕作系统的消失和农村土地的废弃或城市人口与当地环境的重新联系相关的政策响应提供了空间模板，其中。一般来说，我们的方法允许不同级别的抽象，保持区域上下文特异性，同时链接到全球可识别的原型，从而进行概括和理论构建工作。