This study investigated the vertical and temporal distribution of Thaumarchaeota derived core isoprenoid glycerol dialkyl glycerol tetraether (GDGT) lipids through sampling and analysis of both suspended particulate matter from the water column at different times in the annual cycle and a 3year long record of settling particles in two sediment traps at different depths at an open lake location in Lake Superior. Results from these analyses suggest that Thaumarchaeota were present throughout the water column during times of overturning, but mainly resided below the depth of the thermocline (20-40m) during the period of thermal stratification. Fluxes of thaumarchaeotal produced GDGTs were highly periodic and mainly occurred during two periods of the annual cycle (winter and late spring/early summer). A covariance of both branched and isoprenoid GDGT fluxes with the mass accumulation flux combined with the observation that those periods of maximum fluxes were associated with increased BIT index values, however, suggest that these two periods of elevated fluxes may be related to an influx of resuspended particles transported from shallower near shore regions of Lake Superior. During all sampling periods TEX 86 inferred temperatures from SPM were in good agreement with in situ water temperatures of the depths at which the SPM was sampled. The observed range of TEX 86 inferred temperatures in 3years of settling particles is relatively small and does not show significantly higher inferred temperatures during the thermally stratified period, indicating that the sedimentary TEX 86 signal during the summer thermally stratified period mainly originated from depths below the relatively shallow thermocline. Additionally, TEX 86 values during the winter period of increased fluxes did not capture the decrease in water temperatures observed throughout the water column during this period, and thus may be a further indication that the thaumarchaeotal lipid flux was the result of sediment focusing. Flux-weighted TEX 86 inferred temperatures from both sediment traps were in good agreement with TEX 86 temperatures from surface sediments from the same location in Lake Superior. Both flux weighted TEX 86 temperatures from the sediment traps and average TEX 86 temperatures from surface sediments were similar to averaged measured water temperatures at below ~40m depth within the error of the lacustrine TEX 86 calibration. Based on the observed depths of Thaumarchaeota in the water column, TEX 86 values in sediments of Lake Superior likely reflect a combination of mixed-season and sub-thermocline temperatures. This is effectively the same as the annual averaged water temperature observed at depths below 40m in Lake Superior. Thus, trends in TEX 86 inferred temperatures in sediment records of Lake Superior, and similar lakes, are likely to reflect subsurface temperature variability rather than that of surface temperatures.
Bibliographical noteFunding Information:
Two anonymous reviewers are acknowledged for providing helpful comments that have improved the earlier version of our manuscript. We would like to thank the captain and crew of the R/V Blue Heron. We also thank Sarah Grosshuesch Brent Dalzell, Clarissa Booth, Ben Marsh, Jenna Bergin, Matt Stuart, Brittany Kruger, Melissa Berke, Liz Minor, Brandon Stephens and Jeff Strom for sampling assistance, and Ellen Hopmans and Jort Ossebaar for analytical support. This work was supported by NSF Grant #OCE-0452927 to J.P.W. and R.H. J.P.W. acknowledges support from the Gledden Fellowship. J.S.S.D. and S.S. received funding from the ERC project Pacemaker . This work forms contribution 2401-JW at the Centre for Water Research, The University of Western Australia.