Context: Understanding the complex interdependencies between nature and people is a major challenge for the sustainable management of social-ecological systems. Spatially explicit identification of these interdependencies is particularly relevant for managing biodiversity hotspots, such as Tropical Dry Forests (TDF). Objectives: We provided a methodology to spatially identify the components of social-ecological systems that have been shaped by both environmental conditions and management practices at three relevant decision-making scales: plots owned by individuals, plot owners, and governance units. To do so, we identified and characterized: (1) ecological clusters (EC), (2) social-management clusters (SC), and (3) social-ecological systems units (SESU) in a TDF in western Mexico. Methods: We used multivariate analysis to identify and characterize the ECs, SCs, and SESU at the respective decision-making scales. Results: We found four EC, SC, and SESU clusters. Differences between ECs were based on their elevation and land cover type. The SC differed according to the management intensity of cattle and forests. Differences between SESU were based on land management regime (individual vs collective), plot sizes, and time under private schemes. Conclusions: Our findings suggested that decision-makers (ejidatarios) are bounded by the topographical characteristics and the public policies that determine communal (or private) governance, also by the number of resources available to them. The methodology can be applied to other contexts and nested decision-making scales. The spatial identification of these interdependencies is critical for landscape planning since it can contribute to reconciling productive activities and biodiversity conservation.