Finding northernmost baselines: high variability of above-ground biomass on Eurasian polar desert islands

Abstract

As the Arctic rapidly warms, a major change in its vegetation and biomass is expected. Understanding the current state of Arctic plant biomass is crucial due to its role in the surface energy budget and ecosystem carbon storage yet challenging due to logistical and methodological limitations. Arctic polar deserts are one of the most vulnerable terrestrial biomes on Earth, highly sensitive to climate change, and likely also the most understudied. During the 2021 Arctic Century expedition, we performed vegetation surveys and collected aboveground plant and lichen biomass samples at 8 sites on the Severnaya Zemlya archipelago, Franz Josef Land, Vize, and Uedineniya Islands, filling an important spatial gap in biomass measurements in the Arctic. For these study sites, we explored three different methods for estimating plant and lichen biomass: using (1) in-situ species richness, (2) in-situ cover, and (3) remotely sensed normalized difference vegetation index (NDVI) and plant cover. We found no relationship between total in-situ species richness and in-situ biomass, but in-situ lichen species richness significantly predicted lichen biomass. Remotely sensed NDVI had a limited explanatory power for in-situ biomass. However, drone-derived plant cover predicted in-situ biomass measurements well and could thus be used to effectively estimate landscape-level biomass of Arctic polar deserts. Our findings reveal that biomass varies widely among the sites, with an almost complete absence of biomass on Graham-Bell and Komsomolets islands, intermediate levels on October Revolution Inland and Pioneer islands, and maximum biomass found on Vize Island, although the results for this site are highly uncertain. Our findings could be used as a baseline to document future biomass changes in polar deserts. We propose our cover-based approach as an alternative to NDVI-based estimates of polar desert plant and lichen biomass and discuss its uncertainties and limitations.