As plant phenology shifts in response to climate change, the rates and directions of shifts are likely to differ across species’ distributions. Understanding these differences and their drivers will be important for predicting the composition of novel communities in time and space. We use data from two elevational gradients, one semi-arid and one subalpine, to explore how plant phenology differs for the lower and upper limits of species’ distributions. For the semi-arid gradient, located in the Santa Catalina Mountains in Arizona, the data comprise a 30-year record of flowering phenology for all plants in the community. For the subalpine gradient, located in the Rocky Mountains in Colorado, the data come from reciprocal seed transplants in the field and common garden experiments in the greenhouse for four wildflowers. With the long-term data from the semi-arid gradient, we investigated whether flowering phenologies have shifted to different extents or in different directions at the upslope vs. downslope edges of species’ ranges. With the experimental data from the subalpine gradient, we tested whether germination phenologies are plastic and environmentally determined or whether they reflect local adaptation to different elevations.
We found that both the magnitude and the direction of shifts in flowering time differed at the lower vs. upper limits of species’ distributions. Further, we found evidence for local adaptation in the timing of germination, with seeds from high elevation germinating on average several days earlier than seeds from low elevation under common field and greenhouse conditions, indicating accelerated development. Together, these findings suggest phenological responses differ across the elevational distributions of species. Pooling responses across gradients can therefore obscure important population-level differences. Furthermore, local adaptation of phenological traits may hinder upslope migration of plants under climate change.