Responses of competitive understorey species to spatial environmental gradients inaccurately explain temporal changes

Understorey plant communities play a key role in the functioning of forest ecosystems. Under favourable environmentalconditions, competitive understorey species may develop high abundances and influence important ecosystem processes suchas tree regeneration. Thus, understanding and predicting the response of competitive understorey species as a function ofchanging environmental conditions is important for forest managers. In the absence of sufficient temporal data to quantifyactual vegetation changes, space-for-time (SFT) substitution is often used, i.e. studies that use environmental gradients acrossspace to infer vegetation responses to environmental change over time. Here we assess the validity of such SFT approaches andanalysed 36 resurvey studies from ancient forests with low levels of recent disturbances across temperate Europe to assess howsix competitive understorey plant species respond to gradients of overstorey cover, soil conditions, atmospheric N depositionand climatic conditions over space and time. The combination of historical and contemporary surveys allows (i) to test ifobserved contemporary patterns across space are consistent at the time of the historical survey, and, crucially, (ii) to assesswhether changes in abundance over time given recorded environmental change match expectations from patterns recordedalong environmental gradients in space. We found consistent spatial relationships at the two periods: local variation in soilvariables and overstorey cover were the best predictors of individual species’ cover while interregional variation in coarse-scalevariables, i.e. N deposition and climate, was less important. However, we found that our SFT approach could not accuratelyexplain the large variation in abundance changes over time. We thus recommend to be cautious when using SFT substitution toinfer species responses to temporal changes.