Transient fertilization of a post-Sturtian Snowball ocean margin with dissolved phosphate by clay minerals

The study investigates how bioavailable phosphate (PO4) was sourced from continents and transported to the ocean during and after Cryogenian Snowball Earth glaciations, a process hypothesized to have stimulated primary productivity and oxygenation. Prior inferences of elevated seawater PO4 during these intervals come from Fe-rich deposits and shales with high P contents, but the specific delivery mechanisms remain unclear. The authors hypothesize that extensive glacial erosion and deglacial weathering at the end of the Sturtian Snowball Earth enhanced production and transport of fine-grained detrital clays and highly reactive iron oxyhydroxides to continental margins, acting as vectors that adsorbed and then released PO4 to seawater. They test this by analyzing sedimentary sequences in the Dalradian Supergroup (Islay and Garvellachs, Scotland) spanning pre-glacial, glacial, and post-glacial phases, integrating mineralogical, bulk geochemical, Fe-speciation, P-speciation, stable isotope, and numerical modeling approaches to reconstruct PO4 dynamics and redox conditions across the Sturtian deglaciation.

• Elevated P in Cryogenian marine sediments has been linked to post-Snowball deglaciation, enhanced PO4 supply, productivity, and oxygenation, ultimately enabling metazoan evolution. Fe-rich deposits and shales show higher P relative to older counterparts, suggesting higher dissolved PO4 during glaciations.
• Modern systems indicate most PO4 is delivered to oceans adsorbed to riverine clays and metal oxides rather than in solution; up to ~97% PO4 in Greenland glacial meltwater is particle-associated. During the Great Oxygenation Event, clay minerals in more acidic rivers likely transported PO4 to seawater.
• Detrital clay minerals (illite, muscovite) and Fe-oxyhydroxides possess large surface areas and positively charged surfaces at river pH, making them efficient PO4 sorbents. Prior global stratigraphic correlations of the Port Askaig Tillite Formation and Bonahaven Dolomite Formation contextualize the study area within Sturtian glaciation deposits.
• High Cr/Ti ratios have elsewhere been linked to acid rock drainage under oxic atmospheres, but such processes are not generally supported here, except for a brief, non-diagnostic upper BDF anomaly. Overall, literature supports a key role for particulate phases in P transfer and for Fe minerals (ferrihydrite, green rust, magnetite) in nutrient scavenging and preservation.

• Field sampling: Carefully collected outcrop samples from the Dalradian Supergroup (Islay and Garvellachs), spanning pre-Snowball Lossit Limestone Formation (LLF), glacial Port Askaig Tillite Formation (PATF), and post-Snowball Bonahaven Dolomite Formation (BDF). Weathered/metamorphosed lithologies and those near metabasaltic sills were excluded. Unweathered interiors were milled for analysis; selected slices were thin-sectioned for SEM-EDS.
• Mineralogy: Bulk X-ray diffraction (PANalytical X’Pert Pro, Bruker D8 ADVANCE) for semi-quantitative mineralogy; peak area integration using Fityk for relative abundances. SEM-EDS (Oxford FEI-XL30 ESEM) for mineral mapping and point analyses to identify clay minerals and Fe phases.
• Bulk geochemistry: Whole-rock digestion (HF-HCl-HNO3; HF neutralized with H3BO3) and ICP-AES quantification. Major oxides (e.g., P2O5, Al2O3, TiO2, Fe2O3, MnO) and trace elements (Cr, V, Ni, Co). Loss on ignition (LOI) used to assess organic/carbonate contents. Normalization to Ti used to assess detrital versus authigenic enrichment and elemental ratios (P/Ti, Fe/Ti, Mn/Ti, Cr/Ti, etc.).
• Iron speciation: Sequential extraction (Poulton & Canfield protocol) to quantify Fe phases: FeT, poorly reactive sheet silicate Fe (FeSS), Fe-oxyhydroxides (goethite, akaganeite, hematite), magnetite (FeMag), carbonate-associated Fe (FeCarb), pyrite Fe (FePy), and unreactive silicate Fe (FeURS = FeT – sum of others). Fe measured via ferrozine.
• Extractable Fe-bound P: Modified SEDEX P speciation (Thompson et al.) co-extracting P and Fe for Fe-Pcarb, Fe-OXHR-P (oxyhydroxides), PMag (magnetite-bound), and PSS (sheet silicate-associated). Residual P assigned to unreactive silicates (PURS = PT – sum of extracted pools). Used modern linear relationships between dissolved PO4 and Fe-OXHR-bound P to infer paleo-dissolved PO4 dynamics.
• Stable isotopes: δ13Ccarb and δ18Ocarb via phosphoric acid digestion at 60 °C; δ13Corg on acid-decarbonated residues with size-dependent calibrations; Fe isotopes (δ56Febulk) following established protocols.
• Redox proxies: FeHR/FeT and FePy/FeT thresholds and Fe/Al ratios to infer bottom-water redox; evaluation of potential biases by carbonate content and sedimentation rates.
• Numerical modeling: Four-box global ocean PO4-oxygen model (after Alcott et al., Slomp & Van Cappellen) initialized at low-oxygen steady state, subjected to pulses of reactive P to proximal margin waters and/or ocean interior. Tracked atmosphere/deep ocean O2, fraction of oxic seafloor (proximal/distal shelves), P burial in depositional environments, and total marine P concentration. Compared scenarios of increased riverine P (40–60% above modern for 5 Myr) versus direct interior ocean P input.

• Mineralogical and detrital signatures: XRD and SEM-EDS show dominance of sheet silicates (illite, muscovite) and quartz across facies, with chlorite mainly in tillites and albite throughout, indicating significant detrital input from physically eroded felsic-intermediate continental sources. Kaolinite is rare.
• Bulk geochemistry: Elevated P2O5, TiO2, Al2O3, Fe2O3 and trace metals (Cr, V, Ni) in PATF and lower BDF; lower, more variable concentrations in upper BDF (PS-16). P/Ti and Fe/Ti co-vary, implying coupled detrital enrichment; Mn shows decoupled behavior. Al2O3/TiO2 ratios are near-static, pointing to stable UCC-like detrital sources. LOI inversely correlates with P2O5 and Fe2O3, indicating inorganic P and Fe preservation rather than association with organics/carbonates.
• P host phases: Across the sequence, bulk P is predominantly associated with non-calcium-bearing phases (Fe-oxides, unreactive silicates), with limited sedimentary apatite except in some immediate post-Snowball samples. CaO and P2O5 are inversely correlated (up to ~79% inverse correlation in tillites), consistent with limited Ca-phosphate preservation.
• Highly reactive Fe dynamics: Fe-OXHR reservoir declines by about a factor of two from glacial to immediate post-Snowball greenhouse conditions, paralleling high P2O5 in the immediate post-Snowball interval and low P2O5 before and after.
• Leachable P distribution: Sheet-silicate-bound P is highest in tillites (consistent with glacial rock flour), then declines as putative magnetite-bound P rises in the immediate post-Snowball interval. Measured leached P partitions: ~0% with Fe carbonates; ~9.1% with combined goethite/akaganeite/hematite (post-Snowball); ~45.5% with magnetite (42.5% post-Snowball, 3.0% pre-Snowball); ~6.1% with hematite (pre-Snowball); ~39.4% with sheet silicates (tillites and post-Snowball).
• Magnetite as a PO4 recorder: Magnetite-bound P and P/Fe ratios increase at the tillite–post-Snowball boundary and upward, consistent with co-precipitation from PO4-enriched waters. The up-section pattern suggests that as Fe-OXHR reservoir shrank, its capacity to scavenge PO4 decreased, enabling dissolved PO4 buildup captured by magnetite.
• Quantified bioavailability increase: Using the linear relation between magnetite-bound and dissolved PO4, PO4 bioavailability in immediate post-Snowball waters increased by at least 20-fold relative to pre-Snowball conditions.
• Redox reconstruction: Average FeHR/FeT = 0.55±0.21 (pre-Snowball), 0.16±0.04 (immediate post-Snowball), 0.68±0.30 (late post-Snowball). FePy/FeT averages 0.21±0.22, 0.31±0.34, 0.42±0.31, respectively. FeHR/FeT <0.22 in the immediate post-Snowball indicates oxygenated conditions; >0.38 elsewhere indicates anoxia. Magnetite-bound P negatively correlates with FeHR/FeT across the peak post-glacial P interval, linking higher dissolved PO4 with more oxic waters.
• Modeling: A transient increase in riverine PO4 input (40–60% above modern for ~5 Myr) reproduces a brief eutrophication-induced deoxygenation followed by enhanced PO4 burial and transient oxygenation, and then a decline as inputs wane. Interior ocean P addition maintains higher marine P inventory by avoiding proximal burial, contrasting with observed proximal signals.
• Global context: Data align with a Cryogenian spike in P in fine-grained siliciclastic rocks, supporting a detrital-clay-driven P pulse following Sturtian deglaciation.

The findings support the hypothesis that intense glacial erosion during Sturtian deglaciation generated abundant fine-grained clay minerals and Fe-oxyhydroxide particles that adsorbed PO4 on land and during transport. Upon entering higher pH and salinity seawater at continental margins, these particles released PO4, transiently increasing dissolved PO4 availability. Simultaneously, a reduction in the Fe-OXHR reservoir decreased particulate scavenging efficiency, further permitting PO4 accumulation. Magnetite precipitating from the water column captured the elevated dissolved PO4 signal, evidenced by increases in magnetite-bound P and P/Fe. Fe-speciation indicates that this PO4 pulse coincided with transient oxygenation of margin waters (FeHR/FeT <0.22), linking nutrient fertilization to redox improvement. The subsequent decline in detrital P supply and re-expansion of Fe-OXHR sinks likely reduced dissolved PO4, returning to anoxic conditions. These results clarify a key mechanism for post-Snowball fertilization and oxygenation, consistent with modern analogs (e.g., Greenland glacial melt systems) and global Cryogenian P enrichments, and refine our understanding of how particulate vectors controlled nutrient inventories and redox states critical to early animal evolution.

This study demonstrates a causal link between deglacially produced clays and transient fertilization of post-Sturtian continental margin seas, increasing dissolved PO4 bioavailability by at least 20-fold and driving a brief oxygenation event. Mineralogical, bulk geochemical, Fe- and P-speciation, and modeling collectively indicate that detrital inputs dominated P enrichment, with magnetite faithfully recording dissolved PO4 dynamics. The work resolves a key mechanism for nutrient delivery during Snowball Earth termination and its redox consequences. Future research should (i) quantify spatial and temporal variability of the clay-mediated PO4 flux across multiple basins, (ii) develop higher-resolution time constraints on the duration and pacing of the PO4 pulse and oxygenation, (iii) refine mineral-specific PO4 partitioning and adsorption-desorption kinetics under varying pH-salinity-redox regimes, and (iv) integrate additional proxies (e.g., P isotopes, REEs) and coupled Earth system models to better constrain feedbacks among weathering, nutrient cycling, and oxygenation.

• Extractable Fe phase designations can be ambiguous (e.g., potential overlap of goethite in magnetite extracts), introducing uncertainty in partitioning Fe- and P-host phases.
• The extent of Fe(II) oxidation during deglaciation cannot be quantitatively resolved; δ56Febulk suggests partial oxidation but lacks precise constraints on pathways.
• Low-grade metamorphism and diagenesis are minimized but cannot be completely excluded; potential transformations (e.g., kaolinite to illite, chlorite to muscovite) may alter primary mineral assemblages.
• Sampling represents specific localities within the Dalradian Supergroup; broader geographic representativeness remains to be established.
• Proxies such as FeHR/FeT and magnetite-bound P rely on modern analog relationships that may vary under Precambrian conditions; inferred absolute dissolved PO4 increases (≥20-fold) are conservative and subject to proxy uncertainties.
• Brief Cr/Ti anomalies suggest possible localized acidification events, but data indicate they did not control P behavior; causality cannot be fully resolved.