Greening responses to long-term experimental warming in High Arctic tundra communities using plot-level remote sensing

Alongside significant climate warming, a widespread 30-year greening trend has been observed in high northern latitudes. Indicative of increases in vegetation productivity and biomass accumulation, this trend has been associated with shifts in phenology, longer growing seasons, and shrub expansion. Vegetation changes, which can drive feedbacks that can both amplify and dampen warming, have thus been monitored across the tundra biome to better understand the underlying controls, consequences, and future dynamics. Recent studies have shown considerable variability in patterns of warming-induced vegetation change, but efforts remain hindered by the logistical difficulties and spatiotemporal insufficiencies of common monitoring methods. We extracted the Greenness Excess Index (GEI) from a ten-year archive of digital photographs of plots in a set of long-term warming experiments in five High Arctic tundra plant communities. We determined the phenological patterns of greenness and response to warming treatments and climate variables and found that both GEI and phenology showed expected variability across plant communities, warming treatments, and time. Experimental warming significantly increased GEI overall, but the treatment effect varied by individual plant community. Soil moisture was the most important driver of seasonal GEI phenology across all plant communities, but when examined within communities, the most common driver was temperature. We also found significant plot-level variability in the timing of peak greenness across and within plant communities. Long-term trends in GEI and shifts in phenology at the site were difficult to infer using the available image archive, but the observed patterns suggest that – at least over the short time span of this study – GEI has remained stable. GEI was able to detect differences in vegetative vigor and phenology across warming treatments and plant communities and correlated well with common biomass proxies, demonstrating the potential for this simple and cost-effective method of proximal remote sensing for monitoring Arctic vegetation change. Overall, the results of this study are in line with many others from across the Arctic, showing that vegetation responses can vary significantly by time and location, plant community type, and temperature and soil moisture regime, even at small spatial scales.Arts, Faculty ofGeography, Department ofGraduat