Oxygenated Mesoproterozoic lake revealed through magnetic mineralogy Sarah P. Slotznicka,1, Nicholas L. Swanson-Hysella, and Erik A. Sperlingb aDepartment of Earth and Planetary Science, University of California, Berkeley, CA 94720; and bDepartment of Geological Sciences, Stanford University, Stanford, CA 94305 Edited by Paul F. Hoffman, University of Victoria, Victoria, BC, Canada, and approved October 29, 2018 (received for review August 4, 2018) Terrestrial environments have been suggested as an oxic haven Formation has been further interpreted to indicate the pres- for eukaryotic life and diversification during portions of the Pro- ence of more than 50 different species (4). This record is terozoic Eon when the ocean was dominantly anoxic. However, argued to be more diverse than similar-aged marine assemblages, iron speciation and Fe/Al data from the ca. 1.1-billion-year-old which leads to the interpretation that lacustrine environments Nonesuch Formation, deposited in a large lake and bearing a with stable oxygenated waters may have been more hospitable diverse assemblage of early eukaryotes, are interpreted to indi- to eukaryotic evolution than marine ones (4). Early oxygena- cate persistently anoxic conditions. To shed light on these distinct tion of lacustrine environments during the Mesoproterozoic hypotheses, we analyzed two drill cores spanning the trans- has also been proposed based on large sulfur isotope frac- gression into the lake and its subsequent shallowing. While the tionations from sedimentary rocks of the Stoer and Torridon proportion of highly reactive to total iron (FeHR/FeT) is consis- groups that were interpreted to have resulted from oxidative tent through the sediments and typically in the range taken sulfur cycling (15). However, this interpretation is equivocal to be equivocal between anoxic and oxic conditions, magnetic given that such fractionation can arise without oxidative cycling experiments and petrographic data reveal that iron exists in (16, 17). three distinct mineral assemblages resulting from an oxycline. In The chemistry and mineralogy of iron and oxygen in the the deepest waters, reductive dissolution of iron oxides records environment are tightly interwoven, and iron-based geochem- an anoxic environment. However, the remainder of the sedi- ical proxies are among the most mature available for gaining mentary succession has iron oxide assemblages indicative of an insight into local redox conditions (18). Iron speciation measure- EARTH, ATMOSPHERIC, oxygenated environment. At intermediate water depths, a mixed- ments, combined with total iron to aluminum ratios (FeT/Al), AND PLANETARY SCIENCES phase facies with hematite and magnetite indicates low oxygen have been performed on the Nonesuch Formation in the Presque conditions. In the shallowest waters of the lake, nearly every iron Isle Syncline and used to infer persistent water-column anoxia oxide has been oxidized to its most oxidized form, hematite. Com- throughout Paleolake Nonesuch (11). This finding was extrapo- bining magnetics and textural analyses results in a more nuanced lated to terrestrial environments as a whole in the Mesoprotero- understanding of ambiguous geochemical signals and indicates zoic, thereby challenging the interpretation of such environments that for much of its temporal duration, and throughout much of as a potential locus of aerobic Proterozoic eukaryotic evolution its water column, there was oxygen in the waters of Paleolake Nonesuch. Significance Proterozoic j oxygen j iron speciation j lacustrine environments j Constraining oxygen levels and redox chemistry of Protero- eukaryotic evolution zoic oceans and lakes is vital for placing environmental con- straints on early aerobic eukaryotic evolution. Recent work ollowing the origin of eukaryotic life in the Paleoproterozoic has used iron-based geochemical proxies—however, interpre- FEra (2,500–1,600 Ma), eukaryotic diversity is interpreted to tation of such measurements can be difficult due to uncertain- have remained relatively low in marine environments through- ties related to baselines for lake sediments and equivocal zones out the Mesoproterozoic Era (1,600–1,000 Ma) until ca. 800 Ma associated with empirically calibrated proxies. We integrate during the Neoproterozoic Era (1, 2). A hypothesis to explain magnetic, geochemical, and microscale imaging techniques delayed eukaryotic diversification is that marine environments to analyze the iron mineralogy of the 1.1-billion-year-old in a relatively low-oxygen world were prone to the upwelling Paleolake Nonesuch, one of the few lacustrine records of this of anoxic, and sometimes sulfidic, waters from widespread oxy- era. With these methods, we resolve ambiguous geochemical gen minimum zones (refs. 1, 3–5, but see ref. 6). The inhibitory signals and document an oxycline with oxygenated shallow effect of low-oxygen waters on aerobic eukaryotic life holds true waters and decreasing oxygen with depth. These results indi- whether hypoxic conditions were caused by low atmospheric cate a stable oxygenated environment in the terrestrial realm oxygen—as commonly assumed—or by shallow remineraliza- 1.1 billion years ago. tion of sinking organic matter (7). This potential challenge for Author contributions: S.P.S. and N.L.S.-H. designed research; S.P.S. and N.L.S.-H. studied eukaryotic life in the marine realm has led to the suggestion that and sampled the rock cores; S.P.S. and N.L.S.-H. conducted and analyzed rock magnetic oxygenated terrestrial environments may have been cradles of experiments; S.P.S. performed petrography and electron microscopy; E.A.S. developed eukaryotic diversification (4, 8). geochemical data; and S.P.S., N.L.S.-H., and E.A.S. wrote the paper.y Microfossils recovered from the Torridonian sequence of The authors declare no conflict of interest.y Scotland and the Nonesuch Formation of North America have This article is a PNAS Direct Submission.y been interpreted to indicate that by ca. 1.1 billion years ago Published under the PNAS license.y freshwater habitats were colonized by eukaryotes as well as Data deposition: The rock magnetic data are summarized in Dataset S1. The raw rock cyanobacteria (4, 8, 13). Recovered specimens from the None- magnetic measurements have been deposited in the Magnetics Information Consortium such Formation include Valeria lophostriata (4), which is con- (MagIC) Database (https://www.earthref.org/MagIC/doi/10.1073/PNAS.1813493115).y sidered to be diagnostically eukaryotic as the complex wall 1 To whom correspondence should be addressed. Email: [email protected] morphologies and microstructures could not be generated by This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. an organism that does not have a cytoskeleton and endomem- 1073/pnas.1813493115/-/DCSupplemental.y brane system (14). The microfossil record of the Nonesuch www.pnas.org/cgi/doi/10.1073/pnas.1813493115 PNAS Latest Articles j 1 of 6 Downloaded by guest on September 24, 2021 (11). However, published bulk-rock iron speciation data are drill core as far south as Iowa has led to an interpretation not entirely straightforward to interpret. The strength of the that Paleolake Nonesuch was >800 km long (24), although iron speciation proxy lies in its empirical calibration in modern the extent of these lithofacies could be due to multiple lakes marine sediments, allowing for the identification of authigenic along the rift axis as in the modern East African Rift. Regard- reactive iron enrichments (resulting from anoxic water column less, the lake in northern Wisconsin and Michigan was large processes) above an oxic baseline (the reactive iron delivered and persistent with lacustrine sedimentation continuing until through detrital processes). There are few baseline data from after the transition into the overlying Freda Formation. The lacustrine settings, and the delivery of iron to different lakes can Freda Formation is a >4-km-thick succession that is domi- be highly variable (19). Further, most of the existing iron specia- nantly composed of channelized sandstone and overbank silt- tion values from the Nonesuch Formation fall not in the clearly stone deposits representing a prolonged terrestrial fluvial envi- defined oxic or anoxic fields but in the “possibly anoxic” area ronment (25). The Nonesuch Formation has been directly of iron speciation interpretive space (18). Given these ambigui- dated using Re-Os geochronology with a preferred date of ties, new approaches to harness the redox information contained 1,078 ± 24 Ma (11). Paleomagnetic data from the Nonesuch in the sedimentary iron record will have high utility in lacus- Formation (26) suggest deposition in the tropics at a latitude trine rocks and other sediments where established proxies like of 3◦ ± 3◦. iron speciation are ambiguous. This study pairs rock magnet- Five drill cores from northern Wisconsin were used by ref. ics, geochemistry, and microscopy to develop a more detailed 23 to develop a sequence stratigraphic framework for the picture of assemblages of iron oxides and sulfides in Paleolake Nonesuch Formation. Our work focuses on two of these cores: Nonesuch. These data reveal distinct depth-dependent miner- DO-8 and WC-9 (Figs. 1 and 2). In this region, the transgression alogical facies associated with the oxycline of this 1.1-billion- that marks the flooding surface where alluvial facies of the Cop- year-old lake. These facies are also seen within iron speciation per Harbor Conglomerate transition to