Open access publications in the SLU publication database

Abstract

This study examines the changes of total organic carbon (TOC) concentrations in 20 Swedish lakes throughout Sweden using pre-industrial (1860) TOC0 inferred from near-infrared spectrometry (TOCNIRS) of lake sediments to determine if land use change over time could be a plausible explanation for changes in lake water TOC. The study also focuses on the importance of using inferred pre-industrial data of lake water pH and TOC for acidification assessment, in particular ANC0. Most lakes in this study show a long-term decreasing trend of TOC from 1860 up to -0.45 mg/l/yr-1 to an identified breaking point where the TOC turns from decreasing to increasing. Fifteen of the lakes have a breaking point in the mid to late 20th century (1950-1980) while five lakes do not display a clear breaking point. The magnitude of the increasing trend of TOC after the breaking point is up to 0.16 mg/l/yr-1 . Changes in land use were studied by comparing historical maps with present databases of land use. Land use changes in the catchment area show substantial differences in forest cultivation; for instance the coniferous forest has increased by 26% on average. This increase is due to removal of native forest (deciduous forest) and removal of wetlands. Two major conclusions can be drawn from the effects of land use change on TOC levels: (I) No direct correlation between land use change and long-term trends of TOC could be identified in this study. Previous studies have identified the effects of land use change on the carbon storage in the catchment area that corresponds well with the findings of this study. (II) the character of the forest land plays an important role when discussing the effects of land use change for long term TOC trends. The change from open-ended forests with large trees to intense managed forest is considered as an important driving force for TOC. To determine reference conditions there is a need to make good estimations of ANC0 for acidification assessment in Swedish lakes. This study examines the precision of MAGIC model ANC0 calculations (ANC0-MAGIC) against ANC0 calculated with TOCNIRS, diatom-pH and calculated pCO2 (ANC0, diatom-NIRS). ANC0, diatom-NIRS shows a mean difference of (-31µeq/l) when comparing it with ANC0-MAGIC. In comparison, when using contemporary TOC (TOCt) mean lake value 1990-2005 (ANC0,diatom-TOC) the results show a mean difference of (-0.45 µeq/l) in comparison with ANC0-MAGIC. A better fit is generated with TOCNIRS then TOCt. This could be an indication that ANC0- MAGIC overestimates the acidification of Swedish lakes. The European Union’s “Water 8 Framework Directive”, which Sweden has implemented, requires that all surface waters within the Union’s authority have achieved good ecological status by 2015. According to the ecological quality standard the differences between the pre-industrial pH and contemporary pH, i.e. ΔpH=pH0-pHt, should not be more than 0.4 units. This study shows that the long-term trends have to be accounted for when calculating reference conditions and ecological status for acidification