Phenology-mediated feedbacks between vegetation and climate are not well-understood, and the role of floral albedo in these feedbacks is unstudied. Flowers could play an important role in such feedbacks if they reflect more or less sunlight from the Earth’s surface than green leaves. In some temperate habitats, earlier flowering under climate change leads to frost damage to developing flower buds, thereby altering the reflective surface of the habitat to be more green (fewer flowers). At our study site in the Colorado Rocky Mountains, climate change is resulting in warmer air temperatures, less winter snowfall, earlier spring snowmelt, and drier soils; this early snowmelt is associated with reduced floral abundance via frost damage in a subalpine sunflower species – Helianthella quinquenervis– with up to 100% floral loss in some years. Here we asked whether such a reduction of flowers results in a different reflective surface and enhanced sunlight absorbance and heating. We conducted a field experiment in which we compared NDVI (normalized difference vegetation index), soil temperature (at 8cm depth), and volumetric soil moisture between 2´2m plots in which we removed the sunflowers and control plots that contained sunflowers.
Plots in which flowers were removed were significantly greener (higher NDVI) relative to controls that contained open sunflowers (0.86 ± 0.005; 0.84 ± 0.005; t = 4.39, p = 0.0018, n = 20 plots). There was no effect of treatment on soil temperature or moisture before flower removal (not shown). We analyzed soil temperatures on sunny days (> 20,000 lux) during the flowering period of H. quinquenervis. Mean soil temperature across all hours of the day was on average 0.24˚C higher in removal vs. control plots (F = 15.7; p < 0.0001). As expected, the treatment effect was larger during the warmest two hours of the day, when mean soil temperature was 0.48˚C higher in removal vs. control plots (F = 10.9; p = 0.0011). Volumetric soil moisture was on average 2.4% lower in removal vs. control plots during peak flowering (t = 1.99; p = 0.077). Although we were unable to directly measure albedo, soils were drier and warmer in flower removals relative to controls, consistent with the hypothesis that green leaves absorb more sunlight than yellow flowers. Our results show that floral albedo may play an important role in feedbacks between vegetation and climate and deserves further study.