4.3 Species-specific responses of leaf senescence and last flower dates
Warming-induced changes in leaf senescence and last flower dates were highly variable among species. First, unlike leaf out and first flower dates, leaf senescence and last flower dates are more likely jointly controlled by the complex interactions among multiple factors, such as soil temperature, soil moisture, soil nutrient availability, and photoperiod (Ernakovich et al.2014; Estiarte & PeƱuelas 2015; Gill et al. 2015). Changes in these factors may advance or delay the last dates of plant phenology, depending on their balanced effects. Second, the underlying driving mechanisms for plant growth and maintenance of plant metabolic activities are also highly variable among species (Myers-Smith et al. 2015). For example, some legumes are reported to be sensitive to reductions in soil moisture, while forbs are expected to be more responsive to warming-induced changes in soil N availability (White et al. 2000; Engelbrecht et al. 2007). Third, even species within the same plant functional group were also highly varied in their responses of leaf senescence and last flower dates to experimental warming. For example, warming advanced leaf senescence and last flower dates for Gentiana squarrosa but delayed them forHeteropappus altaicus and Artemisia scoparia , despite the fact that they are all forbs. The advancement of leaf senescence and last flower dates of Gentiana squarrosa in the early-growing could confer a competitive advantage by allowing this species to avoid the shading effects induced by other taller species later in the growing season. In addition, plant functional traits (e.g., leaf morphology and plant height) are also reported to have critical impacts on plant phenology even within the same plant function group (Guerin et al. 2012; Dorji et al. 2013). These results call for caution when using plant functional group as a predictor of the effects of warming on last dates of plant phenology.
Our results from the cold Tibetan Plateau stress that warming-induced shifts in phenological firsts are inadequate to predict the ecological consequences of climatic warming on plant phenology and plant community turnover. Rather, we find considerable variability among species in how warming impacts the first and last dates of their phenological patterns. As a consequence, it is warming-induced changes in full phenological periods and phenological lasts, and how this varies among species that drive how warming reshapes species dominance. Our results provide novel insights for understanding the effects of climate warming on plant phenology and plant community turnover in a climatically sensitive ecosystem, and underscore the need to assess how climatic warming will impact the phenological lasts and the full phenological periods.