Contrasting effects of warming and increased snowfall on Arctic tundra plant phenology over the past two decades

Abstract

Recent changes in climate have led to significant shifts in phenology, with many studies demonstrating advanced phenology in response to warming temperatures. The rate of temperature change is especially high in the Arctic, but this is also where we have relatively little data on phenological changes and the processes driving these changes. In order to understand how Arctic plant species are likely to respond to future changes in climate, we monitored flowering phenology in response to both experimental and ambient warming for four widespread species in two habitat types over 21 years. We additionally used long-term environmental records to disentangle the effects of temperature increase and changes in snowmelt date on phenological patterns. While flowering occurred earlier in response to experimental warming, plants in unmanipulated plots showed no change or a delay in flowering over the 21-year period, despite more than 1 °C of ambient warming during that time. This counterintuitive result was likely due to significantly delayed snowmelt over the study period (0.05-0.2 days/yr) due to increased winter snowfall. The timing of snowmelt was a strong driver of flowering phenology for all species - especially for early-flowering species - while spring temperature was significantly related to flowering time only for later-flowering species. Despite significantly delayed flowering phenology, the timing of seed maturation showed no significant change over time, suggesting that warmer temperatures may promote more rapid seed development. The results of this study highlight the importance of understanding the specific environmental cues that drive species' phenological responses as well as the complex interactions between temperature and precipitation when forecasting phenology over the coming decades. As demonstrated here, the effects of altered snowmelt patterns can counter the effects of warmer temperatures, even to the point of generating phenological responses opposite to those predicted by warming alone.

Ecotypic differences in the phenology of the tundra species Eriophorum vaginatum reflect sites of origin

“…A long‐term warming experiment also found that phenology of high arctic plants was not responsive to warming, but year‐to‐year variation in phenology could be explained by snowmelt date (Bjorkman, Elmendorf, Beamish, Vellend, & Henry, 2015). Temperatures are warming in arctic ecosystems at unprecedented rates (Serreze & Barry, 2011), but our data suggest that warming within the range of our OTC treatment (2.4°C on sunny days) may not delay the onset of senescence even if conditions are still conducive for continued photosynthesis.…”

“…Melt of the winter snowpack occurred at approximately the same time in both years, but after this the 2 years diverged substantially. Other work has shown that that timing of snowmelt influences phenological stages (Bjorkman et al., 2015; Rosa et al., 2015; Semenchuk et al., 2016), but on an annual basis, photoperiod and warmer temperatures are the best predictors of bud break (Rosa et al., 2015). In the present study, the snow melted at a similar time, but temperatures stayed lower later in 2016 compared to 2015.…”

“…In this case, the alleviation of cold temperatures may be the cue for growth initiation. The current understanding of the start of the growing season and subsequent phenological phases is that the timing of snowmelt is important (Bjorkman et al., 2015; Rosa et al., 2015), but temperature has a significant predictive role (Rosa et al., 2015); therefore, air temperature succeeding snowmelt merits further study.…”

“…The endogenous rhythm hypothesis is supported by phenological studies in the Arctic, which show that snowmelt timing is a predictor of senescence across a range of PFTs (Bjorkman et al., 2015; Rosa et al., 2015; Semenchuk et al., 2016). Under this hypothesis, once growth is initiated, it proceeds for a fixed number of days until senescence is initiated.…”

TerraSAR-X Time Series Fill a Gap in Spaceborne Snowmelt Monitoring of Small Arctic Catchments—A Case Study on Qikiqtaruk (Herschel Island), Canada

“…In small Arctic catchments, snowmelt is often the most important hydrological driver and generates the majority of annual discharge [17]. The timing of snowmelt, as opposed to temperature, also drives the onset of vegetation phenology and influences subsequent phenological phases and overall fitness of individual plants [18,19]. On a regional scale, spring snow cover in May and June in the Northern Hemisphere has decreased drastically in the last 30 years following trends of increasing air temperatures and reductions in sea ice extent and duration [20].…”