Carabid beetle assembly along an elevational gradient: Spatial heterogeneity dominates over direct altitudinal effects

Elevational gradients offer natural settings for evaluating biodiversity patterns and identifying the ecological processes that shape insect communities. Ground beetles (Carabidae) are widely used as ecological indicators, yet trait-based elevational studies from Central Europe remain scarce. 2. This study examines how species richness, community composition, and a multidimensional set of traits, including functional, life-history, and biogeographical attributes, vary along an elevational gradient in the eastern Ore Mountains and adjacent Děčín Sandstone Uplands in the Czech Republic. Carabids were sampled using pitfall traps across 20 sites spanning 390–820 m a.s.l. over four consecutive years (2003–2006). 3. A total of 19,932 individuals representing 88 species were collected. Species richness declined with increasing altitude, and high-elevation sites supported distinct, cold-tolerant assemblages compared with low-elevation stands dominated by large, predatory forest species. 4. Community-weighted mean traits exhibited strong and consistent non-linear elevational responses after accounting for geometric mid-domain constraints. Dispersal ability showed a shallow mid-elevation minimum, increasing sharply toward higher altitudes where macropterous species became more prevalent. Biogeographical affinity increased with elevation, indicating the growing dominance of species with broad northern distributions at the higher sites. Moisture preference formed an independent trait axis with weak raw structure but displayed a pronounced positive deviation from the geometric null, revealing that high-elevation assemblages host disproportionately moisture-associated species. 5. These results demonstrate that elevation acts as a driver of carabid community structure in Central European mountain forests. Incorporating multiple trait dimensions and geometric null expectations reveals ecological filtering that is not detectable from raw patterns alone. This trait-based perspective enhances understanding of community assembly along environmental gradient and provides insights relevant to biodiversity assessment and forest management.