Shifts in dry grassland vegetation depend on soil and micro-topography

Background and Aims: This study assesses how environmental gradients and land-use jointly shape taxonomic and functional changes in dry grassland vegetation in Central Germany. We tested whether plant functional types and traits explain these changes and reflect ecological strategies under land-use and climate change. Given the widespread abandonment of less fertile and intensification of more fertile grasslands, we hypothesized that disturbance-related variables have stronger impacts on vegetation change than abiotic factors. Methods: Vegetation and environmental data were collected from rocky outcrops differing in slope, aspect, soil properties and proximity to agricultural fields. Boosted Regression Tree (BRT) models were used to evaluate the relative influence of abiotic predictors on taxonomic turnover and functional change, including non-linear and interactive effects. Functional types were defined by life form, longevity, strategy type and leaf traits. Model performance was assessed via cross-validated deviance explained, root mean square error and correlation, while spatial autocorrelation tested for missing landscape gradients. Results: BRT models performed well, revealing that taxonomic turnover was mainly driven by microtopography (slope, aspect), with soil and edge effects being secondary. Turnover peaked on flatter, mesic slopes with low soil C/N ratios. Functional shifts were better captured by aggregated functional types than by individual traits, showing increases in ruderal, stress-ruderal and competitive-ruderal strategists and declines in perennial forms. Interactions (e.g. C/N x pH, aspect x C/N) amplified functional changes in transition zones. Conclusion: Microtopography dominates taxonomic shifts, functional types outperform single traits and mesic transition zones emerge as hotspots of vegetation turnover.