Interactive effects of land‐use change and distance‐to‐edge on the distribution of species in plant communities at the forest–grassland interface

QUESTIONS: What are the patterns of plant communities in terms of richness and floristic composition along stable forest–grassland interfaces? Are there specific species at the edge and/or specialist species transgressions between adjacent habitats? How does edge displacement following land‐use change alter distance‐to‐edge plant diversity patterns? What are the relative influences of the edge effect and the edge displacement in forests and grasslands? LOCATION: Forests and grasslands in northeast France. METHODS: Floristic surveys were conducted in three types of forest–grassland interface, which can be stable or the result of edge displacements due to afforestation or deforestation. The sample comprised a total of 132 plots in 22 forest–grassland interfaces. Point‐biserial correlation coefficients were used to classify the species into several groups in stable interfaces: forest, edge or grassland specialists and generalists. Using LMMs and Tukey's HSD tests, the total species richness and that of each species group were compared between the plot positions (i.e. distance‐to‐edge) and between the stable interfaces and cases of afforestation and deforestation. To evaluate the relative influence of the edge effect (plot position) and edge displacement (afforestation/deforestation) on plant community composition, CCAs and variation partitionings were performed. RESULTS: The largest difference in both plant community richness and composition was found between the forest and grassland edges. The highest species richness was found at the grassland edges. In addition to a high number of generalist species, a mixture of forest and grassland specialist species were predominant at the edges, and the forest specialists were almost the only transgressive species. Afforestation interfaces showed an extinction debt of grassland and edge specialist species and a colonization credit of forest specialists in recent forests. The effects of edge displacement were larger than those of the edge effect in forests, and the opposite was observed in the grasslands. CONCLUSIONS: Forest and grassland plant communities differ strongly even over very short distances. However, the edges host a large number of forest and grassland specialist species, which explain higher species richness than in adjacent habitats. This results from interactive effects of distance‐to‐edge and edge displacement following land‐use change.