Branquinho, CristinaPríncipe, Adriana2023-06-012023-06-012022-102022-04http://hdl.handle.net/10451/57804Forests in drylands are crucial to maintaining important ecosystem services. However, dryland forests are being degraded by multiple pressures, mainly climate change and land-use intensification, leading to an increase in tree mortality and decreasing natural regeneration. Therefore, is critical to understand the underlying environmental factors that are limiting tree cover and distribution in dryland forests. The aim of this thesis was to unravel the contribution of local-scale factors, mainly microclimatic conditions, to explain forest performance in different stages of tree development: productivity, mortality and natural regeneration in oak woodlands of Mediterranean drylands. Particularly, we aimed to i) compare the relative importance of local-scale factors with broad-scale factors in explaining tree productivity, ii) identify what are the most important local-scale factors explaining tree productivity, mortality and regeneration, and iii) understand how microclimate can influence tree productivity, mortality and regeneration. For that, we used broad- (>500 m) and local-scale (<500 m) factors as main predictors of tree productivity, mortality and natural regeneration with high spatial resolution. Our approach used geographic information, remote sensing, multivariate statistics, and field data. This thesis showed that: i) local-scale factors and their interaction with broad-scale factors explained more tree productivity variance than broad-scale factors alone, ii) the most important local-scale factors that limited tree productivity were related to microclimatic conditions, particularly water availability in the soil and Potential Solar Radiation, iii) tree mortality increases in areas where water availability in the soil derived from topographic indices was lower or when groundwater was more distant from the tree roots, iv) very dry microclimatic conditions are the main limiting factor for tree natural regeneration in the long term (more than 60 years), particularly in the tree establishment phase. These results highlight the importance of including variables with a high spatial resolution to model and map the potential of tree productivity, mortality and regeneration over space and time. Modelling dryland forests' response using local-scale factors improves the precision of reforestation and restoration plans, increasing management efficiency and the long-term sustainability of these forests.engsemiáridotopografiamicroclimaagro-silvopastorilmediterrâniosemi-aridtopographyHolm oakCork oakMediterranean ecosystemsdoctoral thesis101530790