Important ecological processes happen over long periods of time and at the landscape scale. Effective conservation of biodiversity and management of natural resources and ecosystem services requires an understanding of these processes. Unfortunately, it is often impractical to conduct appropriate long-term, landscape-scale studies. Modelling offers an alternative approach. Complete ecosystems are too complex to model practically, but simulations of simplified systems provide useful insights of practical value. LandBaSE-P is an individual-based model for Puya hamata that provides information about impacts of fire on ecological processes in the páramo of the Reserva Ecológica El Ángel, Ecuador. Puya hamata is a flagship plant affected by fires and plays a key role in a number of ecological processes. This research found Puya hamata germinated much more frequently after fires, can form large aggregations of single recruitment cohorts, suffers very low mortality (with and without fires) once established, and lives up to 28 years. The spatial aggregation of Puya hamata plants reduced effective reproductive output, consistent with the theory that pollinator behaviour around large groups of Puya plants reduces cross-pollination, leading to inbreeding depression and poorer seed viability and germination. Puya hamata’s population structure can be an indicator of recent fire regime. LandBaSE-P simulations showed that population size is not affected by rare, long-distance seed dispersal. However, in the simulations of páramo grasslands, Puya relative germination is maintained in high numbers by burning. Puya hamata has an important role in ecology and biodiversity. The model LandBaSE-P is a complementary tool for conservation and sustainable land management. This thesis shows how fieldwork combined with laboratory studies and modelling, can provide a good understanding of complex dynamics of real-world populations, and generate ideas for management and future research.