Epsilon Archive for Student Projects

Abstract

Since the beginning of industrialization, emissions of mercury (Hg) from human activities in excess of natural levels have increased depo-sition rates to ecosystems, storage in soils and loading to aquatic envi-ronments. Toxicity to animals, subject to this accumulation, as well as to humans consuming them, are the major concerns driving research on this subject. Peatlands play a key role in Hg cycling as hotspots for Hg methylation, methyl mercury (MeHg) being a particularly mobile, bioavailable form of Hg that is prone to bioaccumulation. Underlying geography is fundamental in shaping the hydrology of a given area and, therefore, the locations of points of accumulation and methylation of Hg.

In this study, potential relationships between geographic parame-ters, elucidated via GIS analysis were investigated with the aim of iden-tifying which parameters were relevant as explanatory variables in the prediction of Hg concentrations in the study area. Elevation was ex-pected to strongly predict MeHg concentrations due to the presence of a local chronosequence, created by land rise. The land’s age since emergence from the sea ranges from years to thousands of years within a span of 10 km, enabled this investigation in an environment in which climate is controlled for.

With 13 of the 15 watershed areas less than 1 ha and 9 less than 500m2, little of meaning could be concluded from statistical analysis with certainty. Linear regression and PLS pointed to Elevation’s rela-tionship with THg, PLS implicated Watershed Area as being associated with MeHg, and PCA hinted at the relevance of Area as well as a clus-ter of Slope, Downslope Index, Curvature, and % Forest for sample sites with extreme values of Hg and other metals. Our results indicate that elevation alone is not a strong predictor of MeHg concentration along this peatland chronosequence.