Key messageDuring the past decades, a multitude of oak stands have spontaneously established across the pine-dominated landscapes of the French Landes de Gascogne. Yet their future performance under modern climate change is unknown. We show that coppiced, dominant trees are most prepared to cope with drought episodes, displaying higher basal area increment and lower sensitivity to extreme events.ContextForest stands dominated by pedunculate oak (Quercus robur L.) have spontaneously established across the pine-dominated landscapes of the French Landes de Gascogne. These oak stands are typically unmanaged and unsystematically coppiced, resulting in mixtures of single- and multi-stemmed (coppiced) trees.AimsTo determine the ability of spontaneous oak forest stands to face climate change–related hazards, by analysing differences in growth (tree-ring width and basal area increment—BAI), wood density and climate sensitivity depending on their tree architecture (single- vs multi-stemmed trees) and their social status in the forest.MethodsWe exhaustively cored 15 oak stands (n = 657 trees). We compared stand characteristics and climate sensitivity between tree architectures considering two sampling designs, either all sampled trees (the exhaustive sampling) or those with a dominant status (dominant sampling). At the tree level, we used linear mixed effects models to compare wood density and growth between tree architectures and the trees’ social status within the canopy layer (dominant- vs non-dominant trees).ResultsMulti-stemmed trees exhibited higher wood density than single-stemmed trees for diameters > 30 cm. Dominant multi-stemmed trees showed lower sensitivity to extreme events (pointer years), higher BAI but lower annual growth rates than dominant single-stemmed trees.ConclusionDominant multi-stemmed trees are potentially the most prepared ones to cope with increasing soil water deficit following drought episodes, at least during the first 60 years of the life of the tree. The vulnerability to face harsher climate conditions for Q. robur stands can be misled when using a dominant sampling design.