Using a newly developed chironomid transfer function for reconstructing mean annual air temperature at Lake Potrok Aike, Patagonia, Argentina

In the Southern Hemisphere, the lack of quantitative temperature records hampers the understanding of climate change since the Last Glaciation and refrains the comparison with the Northern Hemisphere records. To provide quantitative data, a 63-lake chironomid transfer functions was developed in Patagonia. Mean annual air temperature (MAT) was one of the most important factors explaining the distribution of chironomids while precipitation did not have any significant relationship with chironomid assemblages. The MAT model had a r2 of 0.64, a RMSE of 0.83 and a maximum bias of 1.81 °C, comparable to other transfer functions of this size. This model was applied to the Lake Potrok Aike (PTA) chironomid records which consisted of only four taxa (Phaenopsectra, Cricotopus, Smittia and Polypedilum). The chironomid-inferred air temperatures were colder-than-the-average (10.8 °C) during the Lateglacial with the coldest temperatures (9 °C in average) during the Antarctic Cold Reversal (ACR). Between ca. 8000 and 3500 cal. years BP, the chironomid-inferred air temperatures were warmer-than-the-average with a decreasing trend. From ca. 3500 cal. years BP to the present, the chironomid-inferred temperatures oscillated around the average. The difference between the chironomid-inferred air temperature in the surface sample and the climate normal (1961–1990) was 0.6 °C, suggesting that chironomids are sensitive enough to quantitatively reconstruct MAT at PTA. The general pattern of temperature changes reconstructed by the PTA chironomid record corresponded well to other quantitative records in the Southern Hemisphere. The results presented here show that investing in the development of chironomid transfer functions for quantitative climate research in the Southern Hemisphere is valuable.

► We developed a chironomid based temperature inference model for southern Patagonia.
► We quantitatively reconstructed temperature at Lake Potrok Aike during the Lateglacial/Holocene.
► The model inferred colder-than-average temperatures during the Lateglacial.
► During the Lateglacial–glacial cold temperatures were associated with low precipitation levels.