Reconstructing oxygen deficiency in the glacial Gulf of Alaska: Combining biomarkers and trace metals as paleo-redox proxies

Abstract Marine anaerobic oxidation of ammonium (anammox) plays a central role in the nitrogen cycle of modern Oxygen Deficient Zones (ODZs). The newly developed bacteriohopanetetrol stereoisomer (BHT-x) biomarker for anammox, which is largely unaffected by early diagenesis, allows for the reconstruction of the presence and dynamics of past ODZs from the sedimentary record of continental margins. In this study, we investigate the development and dynamics of the ODZ in the Gulf of Alaska (GOA) between 60 and 15 cal ka BP using records of redox sensitive trace metals (TM) and the BHT-x anammox biomarker from IODP Site U1419 (~700 m water depth). The biomarker record indicates that the ODZ in the GOA was in concert with global climate fluctuations in the late Pleistocene. Anammox was more pronounced during warmer periods and diminished during cooler periods, as indicated by correlation with the δ18O signal obtained by the North Greenland Ice core Project (NGRIP). Trace metal enrichments, however, do not match the trend in BHT-x. Systematic metal enrichments in intervals where biomarkers point to more intense water column deoxygenation are not observed. We suggest that this proxy discrepancy was caused by environmental factors, other than water column redox conditions, with opposing effects on the TM and biomarker records. Two of the most widely used redox indicators, Mo and U, are not significantly enriched throughout the sediment record at Site U1419. Site U1419 experienced some of the highest sedimentation rates (100–1000 cm ka−1) ever reported for late Pleistocene continental margin sediments, leading to a continuous and rapid upward migration of the sediment-water interface. We suggest that despite water column and seafloor oxygen depletion, significant sedimentary enrichments of these redox sensitive trace metals were prevented by a limited time for their diffusion across the sediment-water interface and subsequent enrichment as authigenic phases. Thus, depositional conditions were ideal for biomarker preservation but prevented significant authigenic trace metal accumulations. Similar discrepancies between organic and inorganic redox proxies could exist in other high sedimentation rate environments, potentially putting constraints on paleo-redox interpretations in such settings if they are based on trace metal enrichments alone.