Species interactions within communities are governed by phenology. Which species interact, and how they interact, are determined in part by the overlap in timing of their life cycles. Extensive research has documented how phenologies have been shifting in response to changing climate, and how species vary in their responses. Most phenological studies have focused on individual phenophases, such as timing of leaf production or flowering. However, species interact across multiple phenophases through the year, and it's largely unknown whether different these phenophases are shifting synchronously or idiosyncratically. In this study, we performed a field experiment that manipulated snow-melt dates in two sub-alpine meadows to test how species’ responses to changing growing season length vary among phenophases. We monitored the phenology of all plant species over 4 years, creating a unique dataset of how multiple phenophases (leaf production, flowering, seed set, and senescence) respond to changing snowmelt date. We used continuous development models, fit with hierarchical Bayesian models, to quantify how species' development from first leaf-out through senescence was affected by growing season length, and test whether phenological covariance was altered within and among species.
Results of the study reveal complex changes to species and community-level phenology patterns. In particular, we show how the covariance among different phenophases is remarkably species-specific and can be changed by altered snowmelt date. We examined first leaf production, different stages of flowering, seed set, and senescence for closely interacting species. Although many species shifted each phenophase by similar amounts in response to the changes in snowmelt date, many other species had very different responses among phenophases. The result of this variation in responses was a significant reshuffling of which species' phenologies were overlapping through the season. For example, both co-flowering and early season co-occurrence patterns were altered by changing snowmelt date. However, some phenological characteristics of the whole community were relatively buffered by the large differences among species. These results suggest that while individual species interactions will likely change, community-level phenological properties may be more robust to changing climate.