Long-term shifts in flowering phenology and floral abundance in a sub-alpine plant community

Background/Question/Methods

Several long-term datasets have shown that first flowering dates of many temperate plant species are advancing in association with recent climate change, with more rapid advancements in spring compared to summer.  However, it is unclear whether other aspects of flowering phenology, such as peak flowering and duration of flowering (last minus first date of flowering), mirror changes in flowering onset.  Understanding how entire phenological curves are changing, as opposed to distributional extremes like first flowering, will allow for more accurate predictions of ecological responses to climate change.  We use a 36-year record of flowering phenology collected at the Rocky Mountain Biological Laboratory (2900m a.s.l.) to describe how multiple aspects of flowering phenology and peak floral abundance are changing through time in a subalpine plant community. Recent studies have shown that snowmelt is occurring earlier and temperatures are increasing at our study site.  The number of flowers on all plant species present was counted in permanent 2 x 2m plots every other day throughout the growing season, spanning the years 1974-2011. We examined 71 of the 120 plant species present in our long-term record (species present in at least half of our sample years).

Results/Conclusions

First and peak flowering are generally advancing (44% and 27% of species show significant shifts, respectively), but duration of flowering and peak floral abundance are changing in unpredictable ways.  Only for first flowering are phenological shifts more pronounced in spring vs. summer, while changes in peak flowering, last flowering, and duration do not differ between spring and summer.  Shifts in spring flowering, however, are consistently less variable.  Various changes in last flowering dates relative to flowering onset are causing changes and shifts in the flowering duration of individual species (significant changes in 35% species): changes only in duration (flowering extended in 13 species, shortened in 2 species), advancement of entire flowering curves (6 species), and simultaneous changes and shifts (flowering extended plus entire flowering period advanced in 3 species, flowering shortened plus entire flowering period advanced in 1 species).  Peak floral abundance shows significant changes in 28% of species, with half increasing and half decreasing.  Our results present a more complicated picture than conclusions drawn from studies of flowering onset.  These idiosyncratic changes in phenology and floral abundance could have major implications for ecological interactions among plant species and between plants and their mutualists and herbivores.