Excluding the di-unsaturated alkenone in the U37K index strengthens temperature correlation for the common lacustrine and brackish-water haptophytes

Alkenone unsaturation indices, represented by either U-37(K) or U-37(K'), are important tools for paleoclimate studies. The U-37(K) index is a reflection of the average number of double bonds from di-, tri-, tetra-unsaturated alkenones, but U-37(K') excludes the C-37:4 in the calculation. Extensive studies indicate U-37(K') provides better regressions against in situ Sea Surface Temperatures (SST) than U-37(K) and is the most widely used SST proxy. However, recent studies have shown that including C-37:4 alkenones for lacustrine or brackish water haptophytes such as Ruttnera (Chrysotila) lamellosa and Isochrysis galbana improves temperature correlations although there are still significant deviations at the extreme high and low temperatures. In this study, we use new culture-based calibration experiments alongside published culture data and in situ water column or surface sediment data, to demonstrate that a further improved temperature calibration for alkenones is, in fact, achieved when the di- unsaturated alkenone is excluded from the computation of the unsaturation index. We propose new indices, termed U-K '' [U-37(K '') = C-37:3/(C-37:3 + C-37:4) or U-38(K '') = C-38:3/(C-38:3 + C-38:4)], for lacustrine, brackish and estuarine waters. Our observation suggests that di-unsaturated alkenones play a less important role than triand tetra-unsaturated alkenones in regulating cell functions to temperature variations in lacustrine and brackish waters. We suggest using U-K '' indices for paleotemperature reconstructions in the lacustrine and estuarine settings. (C) 2015 Elsevier Ltd. All rights reserved.