Nitrogen Cycling in the Late Paleoproterozoic Redox Stratified

17 Nitrogen cycling has been evaluated across a depth transect in the late Paleoproterozoic 18 Animikie Basin, spanning the end of Earth’s final period of global iron precipitation and a major 19 transition to euxinic conditions in areas of high productivity. Sediments from near shore where 20 productivity was highest have N compositions up to ~3‰ higher than at more distal sites. 21 This suggests that as NH4 + mixed vertically upwards into the photic zone, it was either 22 assimilated by organisms or oxidized. Subsequent enhanced production of N2 by denitrification 23 or anammox led to the observed increase in N close to shore. Any deficit in biologically 24 available N was overcome by N2-fixing organisms, but the input of N with low N from this 25 process did not overwhelm the increase in N from denitrification. Since there is no evidence 26 for conditions of severe N-stress arising from trace metal limitation (particularly Mo) of N 27 fixation during the transition to euxinic conditions, losses of N were either very small 28 (potentially because low O2 levels limited NH4 + oxidation), or alternative pathways which 29 retained N were important. The fact that Mo appears to have remained bioavailable for N 30 fixation, either suggests that the extent or severity of sulfidic water column conditions was not 31 sufficient to quantitatively sequester Mo on a global scale, or that rivers directly delivered Mo to 32 surface waters on the inner shelf. The effects of N2 fixation on N increased to more distal 33 parts of the shelf, consistent with models invoked for modern upwelling zones over broad 34 continental margins. 35 INTRODUCTION AND SETTING 36 Around 1840 Myr ago, a widespread transition to euxinia occurred along productive 37 continental margins (Canfield, 1998; Poulton et al., 2004; 2010; Kendall et al., 2011). These 38 anoxic sulfidic conditions extended outwards from near shore to the mid-shelf, and were overlain 39 by oxic surface waters, with deeper waters that were anoxic and Fe-rich (ferruginous) (Poulton et 40 al., 2010; Kendall et al., 2011; Planavsky et al., 2011). Widespread sulfidic conditions 41 throughout the mid-Proterozoic have been implicated in the protracted oxygenation of the 42 atmosphere and slow rates of eukaryotic evolution (Anbar and Knoll, 2002). However, the 43 continuation of ferruginous deep water conditions through the Mesoproterozoic, with sulfidic 44 conditions limited to areas of high organic C production (Scott et al., 2008; Kendall et al., 2009, 45 2001; Poulton et al., 2010; Poulton and Canfield, 2011; Planavsky et al., 2011), highlights that 46 the controls on productivity, and oxygenic productivity in particular, remain poorly understood. 47 One major question concerns feedbacks associated with the N cycle, and is the focus of 48 this study. The classical view of the N cycle is that as anoxic, NH4 + bearing deep water vertically 49 mixed into oxic surface water, nitrification-denitrification and anammox reactions would have 50 severely depleted the ocean inventory of dissolved fixed N (Fennel et al., 2005). Furthermore, it 51 has also been suggested that expansive euxinia may have placed additional stress on the N cycle 52 by the precipitation of chalcophiles such as Mo and Fe, which are essential to N2 fixation (Anbar 53 and Knoll, 2002). However, the ocean inventory of these metals had to become extremely 54 depleted in order to restrict N-fixation (e.g., Zerkle et al., 2006). Recent evidence for the 55 persistence of ferruginous deep ocean conditions through the mid-Proterozoic (Poulton et al., 56 2010; Planavsky et al., 2011), coupled with the likelihood that Fe was present in solution even 57 under euxinic conditions (potentially as FeS(aq) clusters; Luther et al., 1996), suggests that Fe was 58 unlikely to be biolimiting at this time. In terms of Mo, estimates based on Mo suggest that the 59 extent of sulfidic water column conditions may have been an order of magnitude higher than at 60 present (Kendall et al., 2011), which may not have been sufficiently depleted to restrict N2 61 fixation on a global scale. Furthermore, while sediment Mo and reactive Fe measurements 62 address overlying water, they do not account for the whole of water column, including the upper 63 ocean where the N cycle is most active. Consequently a direct evaluation of the operation of the 64 N cycle is needed to determine its state during this major transition in ocean chemistry. 65 We report new N and C isotope data for sediments from six drill cores that provide a 350 66 km long oblique transect to the paleo-coastline of the Animikie Basin on the margin of Superior 67 Province (Figure 1; Poulton et al., 2010). The Gunflint and Biwabik iron formations, which 68 accumulated in a back-arc basin setting 1878±1.3 Myr ago (Fralick et al., 2002), are overlain by 69 ejecta material from the Sudbury impact event, deposited sub-aerially during regression of the 70 ocean at ~1850±1 Myr ago (Krogh et al., 1984). Later flooding from the south led to deposition 71 of fine grained siliciclastic material of the Rove and Virginia Formations ~1836±5 Myr ago 72 (Addison et al., 2005). The lower 100 m of these formations is dominated by black shales, with 73 turbidites, derived from the north, gaining importance above this zone. Estimated water depths 74 are between ~100 to 400 m, with good connectivity to the world ocean (Rasmussen et al., 2012). 75 Iron speciation measurements suggest that a transition to persistently sulfidic mid-depth water in 76 near shore environments occurred in the Rove and Virginia Formations, while deeper waters 77 remained ferruginous (Poulton et al., 2010), providing an ideal setting in which to study the N 78 cycle during this important transition in ocean chemistry. 79