Human disturbance in alpine habitats is expected to increase, and improved knowledge of short-term recovery  after disturbance events is necessary to interpret vegetation responses and formulate planning and mitigation efforts. The ability of a plant community to return to its original state after a disturbance (community resilience) depends on species composition and environmental conditions. The aim of this study is to analyze initial short-term effects of disturbance in alpine plant communities in contrasting climates (oceanic vs. continental; central Norway). We used a nested block-design to examine vegetative regrowth and seedling recruitment after experimental perturbation. Three plant community types along the snow pack gradient were exposed to (1) no disturbance, (2) clipping, and (3) clipping and uprooting. Slow vegetative regrowth and low seedling establishment rates were found in dry alpine ridges and late-melting oceanic snowbed communities. Leeside habitats with intermediate snow conditions were found more resilient. The difference was related to growth form and species diversity. Woody species, which dominated in ridges and oceanic snowbeds, showed the most negative response to disturbance. Species-rich plant communities dominated by graminoids and herbs showed higher rates of regrowth. Species richness seems to cause resilience to the plant communities through higher response diversity. Plant communities at the extreme ends of abiotic gradients, ridges and late-melting snowbeds, will be most sensitive to both disturbance and environmental change. In an up-scaled human-used landscape disturbance effects will be amplified and further limit recovery to a pre-disturbance state.

The data consists of floristic composition in 216 plots pre-disturbance and after disturbance, in three habitat types in two alpine sites. The experiment was performed using a nested design. In each site (1 km²), three transects were selected subjectively. All transects included three vegetation communities (habitats) along the snow pack gradient; ridge, leeside, and snowbed. Within the habitat, 12 permanent plots (0.5 x 0.5 m²) were randomly located, giving totally 36 plots per transect and 108 plots per site. The distance between plots was minimum 2 m. Within each habitat, plots were randomly allocated to one of three disturbance types: no disturbance (control), moderate disturbance (clipping and removing of field layer vegetation; simulating summer grazing by sheep and reindeer), or severe disturbance (removing of field layer and bottom layer vegetation, and roots that could be removed/pulled without digging; simulating anthropogenic disturbance such as trampling and motorized summer activities).

Prior to disturbance treatments (late 2007/early 2008), total vegetation cover and bare soil was recorded as % cover. Abundance of vascular plants was recorded as the presence/absence of each species in 16 subplots (12.5 x 12.5 cm²) per plot. Plots were reanalyzed at the end of the growing season in 2009. Additionally, a detailed post-growing season search for seedlings was performed in all plots in 2009.