Geochemical assessment of an annually laminated lake sediment record from northern Sweden: a multi-core, multi-element approach

Abstract

We used seven annually laminated (varved) sediment cores from Nylandssjon, a lake in northern Sweden, to assess between-core variation and diagenetic changes at annual resolution. By using several cores, multiple elements and employing principal components analysis (PCA), we also studied how the geochemical composition changed over time, and assessed to what extent these changes were related to variations in the weather. There are between-core differences for aluminum, silica, lead, titanium, zirconium and dry-mass accumulation rate. Diagenesis causes a decrease in bromine, as well as carbon, nitrogen and varve thickness, as reported in previous publications. In spite of anoxic bottom waters phosphorus is not lost from the sediment. In fact, there is an increase in phosphorus concentrations with time. The PCA identified four principal components (PCs). PC-1 accounts for the relative content of mineral and organic material; PC-2 represents mineral-particle grain size; PC-3 reflects phosphorus loading and PC-4 reflects atmospheric pollution. Variations in the weather partly explain the temporal patterns in PC-1 and PC-2: cold winters, i.e. more accumulation of snow, resulted in more mineral than organic matter (i.e. higher PC-1 scores), and increased the relative amount of coarse-grained mineral particles in the sediment (i.e. lower PC-2 scores). Increased spring precipitation had a weak positive effect on the PC-2 scores by promoting the transport of fine-grained material. However, the influence of weather is weak, explaining at most 30 % of the variance, and hence, other factors, e.g. land use and its effect on soil erosion, seem to be more important for the sediment geochemical composition. The importance of land use is also exemplified by an increase in PC-3 scores in the late 1970s, which can be attributed to a shift in agricultural practices that resulted in increased phosphorus loading to the lake. In summary, our findings show that down-core trends are reproducible between cores, but between-core variability and diagenesis need to be considered when interpreting some elements. We also conclude that there is a need to constrain temporal changes in land use before using lake sediments to study changes in weather or climate.