1. Plant interactions are fundamental processes for structuring plant communities and are an important mechanism governing the response of plant species and communities to environmental changes. Thus, understanding the role played by the interaction network in modulating the impact of environmental changes on plant community composition and diversity is crucial. Here, we aimed to develop a new analytical and conceptual framework to evaluate the responses of plant communities to environmental changes.
2. This framework uses functional traits as sensitivity measures for simulated environmental changes and assesses the consequences of microhabitat loss. We show here its application to an alpine plant community where we recorded functional traits (specific leaf area [SLA] and leaf dry matter content [LDMC) of all plants associated with three foundation species or the surrounding open areas. We then simulated primary species loss based on different scenarios of environmental change and explored community persistence to the loss of foundation species.
3. Generally, plant community responses differed among environmental change scenarios. In a scenario of increasing drought alone (i.e. species with lower LDMC were lost first) or increasing drought with increasing temperature (i.e. species with lower LDMC and higher SLA were lost first), the plant community resisted because drought-tolerant foundation species tolerated those deteriorating conditions. However, in scenarios with increasing
nitrogen input (i.e. species having lower SLA were lost earlier), foundation species accelerated species loss due to their early primary extinctions and the corresponding secondary extinctions of species associated to their microhabitat.
4. The resistance of a plant community depends on the driver of environmental change, meaning that the prediction of the fate of this system is depending on the knowledge of the main driver of environmental change. Our framework provides a mechanistic understanding of an ecosystem response to such environmental changes thanks to the integration of biologyinformed criteria of species sensitivities to environmental factors into a network of interacting