Chironomid-inferred air temperatures from Lateglacial and Holocene sites in north-west Europe: progress and problems

Abstract Non-biting midges (Insecta: Diptera: Chironomidae) are now widely recognised as powerful biological proxies for inferring past climatic change. In this paper the response of chironomid assemblages to Lateglacial and Holocene climatic change from sites in Scotland and Norway is discussed. Analyses have shown that midge assemblages have responded to major and minor climatic fluctuations during the last 15,000 yr, such as the Aegelsee Oscillation, Gerzensee Oscillation and Younger Dryas in the Lateglacial, and the thermal optimum, “Little Ice Age” and cooling events at 8.2 and 2.0 cal ka BP in the Holocene. In an attempt to quantify these and other climatic fluctuations a chironomid-mean July air temperature inference model, based on a modern calibration set of 109 lakes in Norway and Svalbard, has been developed. The 3-component weighted averaging partial least squares inference model performs well with a root mean squared error of prediction of 0.90°C and a maximum bias of 0.59°C (both based on leave-one-out cross-validation) which is lower than chironomid-temperature inference models developed in Canada, Switzerland and Finland. However, while temperature-inference models perform well in the Lateglacial, the more subtle climatic changes of the Holocene are more difficult to reconstruct because the likely temperature changes are close to the limits of the prediction errors of the existing models. To address this problem, future work is required to refine and develop the calibration set by further improving taxonomic resolution, by adding more lakes to the calibration set and by investigating the influence of other environmental variables on chironomid distributions and abundances.