The effect of Holocene treeline fluctuations on the sediment chemistry of Lake Kilpisjarvi, Finland

The sediments of Lake Kilpisjarvi were described and analysed for element chemistry and pollen to study the effects of treeline fluctuations in the catchment. Lake Kilpisjarvi is one of the largest lakes in Finnish Lapland, with its catchment partly above the treeline and partly covered with mountain birch woodland. Although the presence of subfossil pine shows that the catchment was previously covered with mountain birch woodland during the Holocene, the present pine treeline has receded 70 km from the lake. Pollen analysis results show that pine immigrated to the area during the Atlantic chrone and that similar to 7000 BP pine forests occupied much of the catchment. Pine started to decline around 3500 BP and vegetation in the catchment became more open. Alkaline and alkaline earth metals and some transition metals document the change from glaciolacustrine clay to more organic sediment. However, these geochemical trends give no indication of changes in erosion rate resulting from changes in catchment vegetation. These changes were detected by plotting suitable element ratios. In addition to the conventional Si/Al and Na/K ratios, the Ca-labile/Si ratio and especially the ratio of labile Ca to K were found to be useful. Of all the elements analysed, potassium showed the strongest reaction to changes in the balance between weathering and erosion. During the phase of denser forests, chemical weathering was dominant, whereas during phases of open catchment, physical erosion prevailed. The effects of changing climate and catchment vegetation were distinguished from other signals. For instance, iron and manganese were enriched at the top of the core due to diffusion and, at the same time, old precipitate layers persisted after burial to deeper levels in the sediment. These iron and manganese rich layers had an effect on the distributions of cobalt, zinc, and vanadium, showing increased concentrations of these elements. Other effects that made the interpretation of chemical records difficult were the effect of ongoing mineralization of organic matter in the top layers of sediment and the effect of biogenic silicon. Owing to the stable conditions of the lake, the desired chemical signals were detected, despite the masking trends.