A comparison of Holocene temperature reconstructions from vegetation proxies and brGDGTs in northern Fennoscandia

Studies on Holocene climate conditions commonly show discrepancies between different proxy-based reconstructions, calling for in-depth analysis of both the proxies and the derived climate reconstructions. Here, we conducted multi-proxy Holocene warm-season temperature reconstructions for a sedimentary record from Lake Tippakuru in northeastern Finland, using pollen, plant macrofossils, and the novel branched glycerol dialkyl glycerol tetraethers (brGDGTs) as climate proxies. In addition, we prepared a regional climate reconstruction by integrating pollen-based reconstructions from nearby sites. Mainly, we evaluated the potential of using brGDGTs as a climate proxy and compared the reconstructed climate shifts with those reconstructed from classical vegetation-based proxies (pollen and plant macrofossils). We see significant discrepancies between the three proxies, most notably during the early Holocene where plant macrofossils and brGDGTs indicate warm conditions while the pollen-based reconstructions are equivocal and strongly depend on the quantitative reconstruction method used. Additionally, the brGDGT records indicate an abrupt cooling at ca. 5.8 cal ka BP which is not seen in the classical proxies at the study site, but only in neighbouring localities. The results highlight the need for modern-day studies on seasonal variability and bacterial sources of brGDGTs to disentangle non-thermal biases and evaluate seasonal imprints in temperature reconstructions. In general, our results imply that not only the choice of climate proxies but also the methodological choices in the proxy-based quantitative reconstructions affect the climatic interpretation, especially during the early Holocene transitional period. We highlight the need for both multi-proxy and multi-method approaches in paleoclimate studies, as well as the need to expand the knowledge of brGDGT responses to climatic and other environmental variables. • Inconsistent early Holocene warm-season temperature conditions across proxies. • Mid-Holocene proxy-based temperature anomalies between +1 °C and +1.5 °C. • Proxies indicate long-term late Holocene cooling, but no short-term fluctuations.