Abandoned management of wooded grasslands leads to development of secondary woodland. We investigated how this development affects the colonization–extinction dynamics and persistence of epiphytic lichens associated with old trees. We modelled colonization probability based on observed colonizations (turnover data) during four years of two old-oak-associated lichens on 1236 oaks. Persistence was assessed by projections of future dynamics. We also used the turnover models to validate models fitted to snapshot data (from one point in time). Epiphyte colonization probabilities were lower on trees in closed than in open conditions, and the probabilities increased with increasing connectivity to surrounding occupied trees. The additional four study species had too few colonizations to be modelled, and thus, very low colonization rates. Local extinctions occurred only deterministically through patch destruction processes. In projections of future metapopulation dynamics, when assuming that all trees were in closed conditions, the metapopulations decreased slowly; new equilibria had not been reached after 200 years. In contrast, when assuming open conditions for all trees, to test for effects of clearing vegetation around oaks in closed conditions, the metapopulations increased comparatively fast. The turnover models and the snapshot models, gave similar projections of metapopulation sizes, when assuming that the present level of secondary woodland remained constant over time. Development of secondary woodland in wooded grassland has negative impacts on epiphyte metapopulations. However, the slow metapopulation declines suggest that restoration will be successful. High priority should be given to resumed grazing and clearing vegetation around old trees, in particular close to dispersal sources.