doi:10.1016/S0016-7037(00)00168-6 Asynchronous alkenone and foraminifera records from the Benguela Upwelling System G. MOLLENHAUER, 1, * T. I. EGLINTON, 2 N. OHKOUCHI, 2,† R. R. SCHNEIDER, 1 P. J. M¨ ULLER, 1 P. M. GROOTES, 3 and J. RULLK ¨ OTTER 4 1 Fachbereich Geowissenschaften, Universita ¨t Bremen, Bremen, Germany 2 Department of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA 3 Leibniz Labor fu ¨r Altersbestimmung und Isotopenforschung, Christian-Albrechts-Universita ¨t Kiel, Kiel, Germany 4 Institut fu ¨r Chemie und Biologie des Meeres, Carl von Ossietzky Universita ¨t, Oldenburg, Germany (Received September 6, 2002; accepted in revised form February 17, 2003) Abstract—Radiocarbon stratigraphy is an essential tool for high resolution paleoceanographic studies. Age models based on radiocarbon ages of foraminifera are commonly applied to a wide range of geochemical studies, including the investigation of temporal leads and lags. The critical assumption is that temporal coupling between foraminifera and other sediment constituents, including specific molecular organic com- pounds (biomarkers) of marine phytoplankton, e.g. alkenones, is maintained in the sediments. To test this critical assumption in the Benguela upwelling area, we have determined radiocarbon ages of total C 37 -C 39 alkenones in 20 samples from two gravity cores and three multicorer cores. The cores were retrieved from the continental shelf and slope off Namibia, and samples were taken from Holocene, deglacial and Last Glacial Maximum core sections. The alkenone radiocarbon ages were compared to those of planktic foraminifera, total organic carbon, fatty acids and fine grained carbonates from the same samples. Interest- ingly, the ages of alkenones were 1000 to 4500 yr older than those of foraminifera in all samples. Such age differences may be the result of different processes: Bioturbation associated with grain size effects, lateral advection of (recycled) material and redeposition of sediment on upper continental slopes due to currents or tidal movement are examples for such processes. Based on the results of this study, the age offsets between foraminifera and alkenones in sediments from the upper continental slope off Namibia most probably do not result from particle-selective bioturbation processes. Resuspension of organic particles in response to tidal movement of bottom waters with velocities up to 25 cm/s recorded near the core sites is the more likely explanation. Our results imply that age control established using radiocarbon measurements of foraminifera may be inadequate for the interpretation of alkenone-based proxy data. Observed temporal leads and lags between foraminifera based data and data derived from alkenone measurements may therefore be secondary signals, i.e. the result of processes associated with particle settling and biological activity. Copyright © 2003 Elsevier Science Ltd 1. INTRODUCTION Molecular level studies of organic compounds from marine sediments can provide useful information on past and present oceanic environments. In particular, one suite of marine bi- omarkers, the long-chain unsaturated ketones (“alkenones”) produced by certain haptophyte algae contain valuable infor- mation on paleo sea surface temperatures (SSTs) (e.g., Brassell et al., 1986; Mu ¨ller et al., 1997). Reconstructing SSTs by measuring the unsaturation ratio of C 37 alkenones is robust, rapid, and inexpensive, requires small samples and is applica- ble in most oceanographic settings (e.g., Prahl et al., 1988; Mu ¨ller et al., 1998). Thus the alkenone parameter is now widely used and largely established as a paleoceanographic proxy. Furthermore, combining proxy information derived from alkenones and isotopic composition of calcareous nano- and microfossils can help deducing more detailed information on past ocean surface conditions, such as estimates of paleosa- linities (Rostek et al., 1993; Lamy et al., 2002). Another appli- cation of a combination of such proxy records is to disentangle the timing of variations in ocean temperature relative to changes in sea-level and continental ice volume (Schneider et al., 1995; Kirst et al., 1999; Ru ¨ hlemann et al., 1999; Bard et al., 2000; Herbert et al., 2001; Kim et al., 2002). The underlying assumption for all these investigations is that temporal coupling between the different signal carriers is maintained upon depo- sition and burial in the sediments. This is a prerequisite which may not be met under certain conditions. In particular, physical processes associated with sedimentation and burial of different constituents after signal formation in the surface waters have been proposed to cause difficulties (e.g., Sachs et al., 2000). In contrast, chemical alteration of the alkenone paleotemperature signal through degradation is generally regarded to be minor (see review by Grimalt et al., 2000). Examples for physical processes are sediment mixing and advection, both of which can smooth the various proxy records. Through bioturbation, apparent temporal offsets between pa- leoclimate proxies may be produced (e.g., Hutson, 1980; Bard et al., 1987; Broecker et al., 1999; Bard, 2001). Sediment constituents that differ in size, shape, density and composition may be subject to differential mixing. The attenuation and apparent temporal offset of a signal is inversely proportional to the sedimentation rate (Guinasso and Schink, 1975; Bard et al., 1987; Brown et al., 2000; Anderson, 2001). Intensity and depth of bioturbation, however, are hard to measure directly, and * Author to whom correspondence should be addressed, at Department of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA (gmollenhauer@whoi.edu). † Present address: Institute for Frontier Research on Earth Evolution (IFREE); Natsushima-cho Yokosuka, Japan. Pergamon Geochimica et Cosmochimica Acta, Vol. 67, No. 12, pp. 2157–2171, 2003 Copyright © 2003 Elsevier Science Ltd Printed in the USA. All rights reserved 0016-7037/03 $30.00 + .00 2157