Comparisons of climate model hindcasts with independent proxy data are essential for assessing model performance in non-analogue situations. However, standardized paleoclimate datasets for assessing the spatial pattern of past climatic change across continents are lacking for some of the most dynamic episodes of Earth's recent past. Here we present a new chironomid-based paleotemperature dataset designed to assess climate model hindcasts of regional summer temperature change in Europe during the late-glacial and early Holocene. Latitudinal and longitudinal patterns of inferred temperature change are in excellent agreement with simulations by the ECHAM-4 model, implying that atmospheric general circulation models like ECHAM-4 can successfully predict regionally diverging temperature trends in Europe, even when conditions differ significantly from present. However, ECHAM-4 infers larger amplitudes of change and higher temperatures during warm phases than our paleotemperature estimates, suggesting that this and similar models may overestimate past and potentially also future summer temperature changes in Europe.

We used a series of experiments to determine whether stable carbon isotope analysis of modern and fossil larval head capsules of chironomids allowed identification of their dietary carbon source. Our main focus was to assess whether carbon from naturally 13 C-depleted methane-oxidizing bacteria (MOB) can be traced in chironomid cuticles using stable carbon isotope analysis. We first showed that a minimum sample weight of *20 lg was required for our equipment to determine head capsule d 13 C with a standard deviation of 0.5%. Such a small minimum sample weight allows taxon-specific d 13 C analyses at a precision sufficient to differentiate whether head capsules consist mainly of carbon derived from MOB or from other food sources commonly encountered in lake ecosystems. We then tested the effect of different chemical pre-treatments that are commonly used for sediment processing on d 13 C measurements on head capsules. Processing with 10% KOH (2 h), 10% HCl (2 h), or 40% HF (18 h) showed no detectable effect on d 13 C, whereas a combination of boiling, accelerated solvent extraction and heavy chemical oxidation resulted in a small (0.2%) but statistically significant decrease in d 13 C values. Using culturing experiments with MOB grown on 13 C-labelled methane, we demonstrated that methanogenic carbon is transferred not only into the larval tissue, but also into chironomid head capsules. Taxon-specific d 13 C of fossil chironomid head capsules from different lake sediments was analyzed. d 13 C of head capsules generally ranged from -28 to -25.8%, but in some instances we observed d 13 C values as low as -36.9 to -31.5%, suggesting that carbon from MOB is traceable in fossil and subfossil chironomid remains. We demonstrate that stable carbon isotope analyses of fossil chironomid head capsules can give insights into dietary links and carbon cycling in benthic food webs in the past and that the method has the potential to reconstruct the importance of MOB in the palaeo-diet of chironomid larvae and, indirectly, to infer past changes in methane flux at the sediment water interface in lakes.

Abstract

Stable oxygen isotope measurements on fossil chironomid head capsules from lake sediments show that these chitinous remains can be used to reconstruct past lake water d18 O and, indirectly, past climate change. We examined the impact of chemical pretreatment procedures on the chemical and stable oxygen isotope composition, and morphology of chironomid cuticles. Use of alkali, acids, and sodium chlorite alters the chemical composition and the morphological structure of chironomid cuticles by selective removal of chitin or proteins. Gas chromatograms of pyrolyzates show that NaClO 2 causes deproteination, whereas the combined use of HCl and HF results in partial chitin removal. Head capsules pretreated with KOH contained both chitin-and protein-derived moieties, although the concentration of protein was reduced, especially after KOH treatment at high concentration (28%) and temperature (100°C). Scanning electron microscopy confirmed that a proteinaceous matrix is still present in modern and fossil head capsules after KOH treatment. This matrix, however, is largely absent in head capsules pretreated with NaClO 2 . A change in the proportion of chitin and proteins in our samples was associated with differences in chironomid

No abstract