The Influence of Resurvey Plot Location on Observed Vegetation Change

Aim Long‐term vegetation study through resurveys is essential for quantifying how biodiversity changes over time. The accuracy of resurvey data is influenced by plot relocation errors, arising from the spatial uncertainty of historical plot locations. Preferential location of resurvey plots potentially underestimates vegetation change in time, whereas random location may overestimate it. To address these methodological issues, we compared preferential and random relocation approaches to assess how plot location error influences observed temporal vegetation changes, focusing on diversity patterns across two common grassland types in Central Europe. Location Northeastern Austria and the southern Czech Republic. Methods We compared two methods for locating plots in resurveys at 50 sites along the Austrian‐Czech border: (a) resurvey plots located preferentially by the surveyor in the field in a place with vegetation similar to that described in the historical plot and (b) resurvey plots located randomly within a 300‐m spatial buffer around the probable location of the historical plot, within the same land‐cover type. We then compared changes in species richness, composition, and Ellenberg‐type indicator values between the two resurvey plot location methods. Results Across the entire dataset, species richness change and the Jaccard dissimilarity index were lower in the preferentially located than in the randomly located plots, while other metrics (species richness, number of species lost and gained, species composition, top winners and losers, Ellenberg‐type indicator values) showed no significant differences. In mesic grasslands, the resurvey plot location approach had the greatest effect on species richness. In contrast, in dry grasslands, it introduced uncertainty into the detection of compositional changes. Conclusions While both approaches yielded overall similar results, resurvey plots located preferentially indicated smaller changes in species richness and composition change than random plots. Results in resurvey studies based on quasi‐permanent plots should therefore be interpreted with caution, as plot location may affect temporal change detection in long‐term vegetation assessments.