Oxygen isotope ratios of chironomids, aquatic macrophytes and ostracods for lake-water isotopic reconstructions – Results of a calibration study in Patagonia

• The applicability of biogenic proxies for host-water isotope reconstructions in Patagonia was tested.
• Oxygen isotope ratios of aquatic macrophytes, chironomids and ostracods were compared.
• Aquatic macrophytes best reflected host-water isotopic composition.
• No clear temperature dependency of isotopic fractionation was found for macrophytes and chironomids.

SummaryOxygen stable isotopes (δ18O) of biogenic sedimentary components from lake sediment archives, such as ostracod valves, are frequently used for palaeoclimatic reconstructions. However, the effects of host water isotope changes, temperature-dependent fractionations, and vital effects on the isotopic ratios of such biogenic proxies often cannot be disentangled. Here, δ18O values of ostracod valves, head capsules of chironomid larvae and aquatic macrophyte cellulose from various lake and stream sites in southern Patagonia (41–52°S) are compared with δ18O values of their host waters. The best correlation between δ18O values of bioproxy and host water was obtained for aquatic cellulose exhibiting a slope of the regression of almost unity. Chironomid head capsules showed a strong correlation, but the data set was rather small (seven sites). The δ18O values of ostracod valves showed strong inter-specific vital effects. Strong offsets of ostracod values from isotopic equilibrium are possibly explained by re-working of subfossil valves, seasonality effects or different chemical composition of host waters. A weak, but significant temperature dependency of oxygen isotope fractionation was observed for cellulose of one aquatic taxon (Myriophyllum) and for ostracod calcite, but not for chironomids and aquatic moss cellulose. The present dataset suggests that δ18O values of aquatic cellulose are the most reliable proxy for host water isotope reconstructions, but clearly more investigations are needed to substantiate this finding.