Global Distribution of Mercury in Foliage Predicted by Machine Learning.

ABSTRACT

Foliar assimilation of elemental mercury (Hg0) from the atmosphere plays a critical role in the global Hg biogeochemical cycle, leading to atmospheric Hg removal and soil Hg insertion. Recent studies have estimated global foliar Hg assimilation; however, large uncertainties remained due to coarse accounting of observed foliar Hg concentrations, posing a substantial challenge in constraining the global Hg budget. Here, we integrated a comprehensive observation database of foliar Hg concentrations and machine learning algorithms to predict the first spatial distribution of foliar Hg concentrations on a global scale, contributing to the first estimate of global Hg pools in foliage. The global average of foliar Hg concentrations was estimated to be 24.0 ng g-1 (7.5-56.5 ng g-1), and the global total in foliar Hg pools reached 4561.3 Mg (1455.2-9062.8 Mg). The spatial distribution showed the hotspots in tropical regions, including the Amazon, Central Africa, and Southeast Asia. A range of 2268.5-2727.0 Mg yr-1 was estimated for annual foliar Hg assimilation accounting for the perennial continuous assimilation by evergreen vegetation foliage. The first spatial maps of foliar Hg concentrations and Hg pools may aid in understanding the global biogeochemical cycling of Hg, especially in the context of climate change and global vegetation greening.