A systemic approach to landscape ecology: towards dynamic frameworks for vegetation monitoring and resilience analysis

Ecosystems worldwide are undergoing accelerating transformation under the combined pressures of land-use change, fragmentation, and climate variability. These drivers alter not only ecosystem structure but also the feedbacks that sustain stability across scales. This thesis advances a systemic dynamic framework for landscape ecology, linking spatial configuration, functional processes, and temporal dynamics to assess vegetation resilience through time. The framework treats landscapes as self-organizing systems whose stability emerges from cross-scale interactions between pattern, process, and disturbance. The thesis is articulated in two complementary papers. The first, Landscape Context and Forest Persistence (Bardino et al., 2023, Landscape Ecology), explores the structural dimension of resilience, identifying how topography, hydrology, and accessibility shape the long-term persistence of regenerating forests in central Panama. The second, Hidden Instability in Protected Tropical Forests (in preparation), investigates the functional and dynamic dimensions, showing that even structurally intact forests may experience functional decoupling from climate after extreme droughts such as the 2015 2016 El Niño. Declining recovery rates and altered vegetation climate coupling reveal hidden instability beneath apparent stability. By combining spatial and temporal diagnostics derived from remote sensing, time-series modelling, and machine learning, this work reframes landscape ecology as a predictive science of adaptive systems. It translates resilience from an abstract concept into a quantifiable, spatially explicit property, offering practical tools for monitoring, restoration, and climate adaptation. Ultimately, it demonstrates that safeguarding forest structure alone is insufficient: sustaining the feedbacks that maintain ecological coherence is essential for long-term stability under global change.