全球变化知识资源中心

Aims Forests worldwide are subjected to increasing pressures from altered disturbance regimes, climate change, and their interactions. We resampled previously established vegetation plots to directly assess long-term vegetation dynamics in the forest understory of mixed aspen forests, including species diversity, distribution, and composition. Study site Rocky Mountain National Park, Colorado, USA. Methods We resampled 89 vegetation plots that contained aspen (Populus tremuloides) in the original sampling in 1972/1973 in Rocky Mountain National Park to assess changes in understory diversity, non-native species abundance, community composition, and elevation ranges. Analyses were performed at three spatial scales: landscape (all plots), ecotone (montane vs. subalpine), and seven forest "series" according to the classification presented in the original publication. Results Understory vegetation diversity did not significantly change at the landscape scale but increased in forests of the mesic montane series. Changes in diversity varied with elevation, with predominantly increases at lower elevations and decreases at higher elevations. Furthermore, species turnover and upward shifts were more pronounced at lower elevations. The proportion of plots containing non-native species was similar between data sets, with 48% in 2012/2013 and 46% in 1972/1973. However, the number of non-native species per plot increased considerably, especially in Pinus contorta forests. Significant shifts in understory community composition occurred in mesic montane forests and Pinus flexilis forests. Generally, higher floristic overlap was evident between forest types, indicating homogenization between 1972/1973 and 2012/2013 understory plant communities. Conclusions Although our results indicate overall little change in understory communities between 1972/1973 and 2012/2013 in Rocky Mountain National Park, they also suggest that observed changes differ by elevation, possibly due to the interactions between elevation and changes in local climate. Our study underscores the importance of analyzing long-term vegetation dynamics at different spatial scales and provides data from direct observations to improve the predictive power of vegetation models.