Phylogenetic patterns in phenological response to climate in a suite of closely related species
Phenology is a powerful indicator of plant responses to climate change. Earlier spring flowering and leafing have been widely reported as spring temperatures increase, but studies of late season phenological changes are still uncommon. Before phenology can effectively be used to forecast plant responses to climate change, researchers must better understand the factors driving variation in plant responsiveness to climate, identify measurable traits that are correlated with the type and magnitude of response, and clearly characterize autumn phenological responses. We conducted a six-year phenological monitoring program for 35 species of Viburnum in a common garden setting at the Arnold Arboretum, Jamaica Plain, Massachusetts. This monitoring period included normal weather years and two of the warmest years on record for the area (2010 and 2012). In order to document fall phenological changes in addition to expected spring advancement, we monitored plants from bud break to senescence, and conducted detailed leaf lifespan surveys. We determined the phenological cues, phylogenetic lineages, and plant traits most correlated with phenological responses. We noted spring and fall growing season extensions, and additional phenological abnormalities in the fall: certain species initiated a second flush of growth late in the season.
All species shifted their spring phenology earlier in warmer years as compared with normal weather years. A subset of species shifted fall senescence later. The most important cues for phenological phases, even those that occurred in fall, were spring temperatures. A secondary suite of cues related to summer precipitation affected the timing of the start of senescence, but not the 50% or 95% senescence dates. Among these closely related species, there is a great deal of variability in degree of responsiveness to elevated temperatures. A significant portion of this variability is explained by the morphological character of naked buds, present in three lineages representing independent transitions from the tropics to the temperate zone. Species with naked buds adjusted their spring phenology more markedly than scale-bud species, were significantly more likely to initiate a late-season second flush of growth, and were more variable in their autumn response to inter-annual climatic variability. Their late-opened buds were also disproportionately likely to survive through the winter and provide limited spring reproductive potential. The naked-bud trait may therefore be extremely valuable for predicting plant response to increased growing season temperatures, improving our ability to assess species risk in the face of climate change.