University of Southern California Dissertations and Theses

Environmental controls on alkalinity generation from mineral dissolution: from the mineral surface to the global ocean

(USC Thesis Other)

Environmental controls on alkalinity generation from mineral dissolution: from the mineral surface to the global ocean

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content ENVIRONMENTAL CONTROLS ON ALKALINITY GENERATION FROM MINERAL DISSOLUTION: FROM THE MINERAL SURFACE TO THE GLOBAL OCEAN by Abby Michelle Lunstrum A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the degree DOCTOR OF PHILOSOPHY (GEOLOGICAL SCIENCES) August, 2022 Copyright 2022 Abby Michelle Lunstrum ii Acknowledgments Many years ago, I left my undergraduate education with little understanding of how science “works”, and—despite a Bachelor of Science degree in Ecology—little knowledge of biogeochemistry or elemental cycles. I feel quite lucky to have been introduced to these topics, which I love, later by several people. Margaret Palmer and her graduate students at the University of Maryland welcomed me into their lab as a volunteer and first introduced me to biogeochemistry. Their willingness to tutor a stranger set me on my current path, and I strive to pass on the favor to other curious minds. Luzhen Chen at Xiamen University encouraged me to write my first scientific paper when I truly had no idea that was something I could do. Karen McGlathery and my excellent labmates at UVA—Lillian Aoki and Matthew Oreska—taught me everything about coastal ecology. Suzanne Bricker showed me how to conduct society-relevant science (and be a super friendly federal scientist at the same time). And finally, John Harrison and Bridget Deemer at WSU Vancouver were instrumental in getting me excited about aquatic biogeochemistry. Thanks to their patience, generosity, and great spirits, I’m still asking questions about greenhouse gases! At USC, I am eternally grateful to Will Berelson for roping me into the carbonate system. Thank you for taking a chance on an environmental scientist interested in fish farming, and turning me into a passable carbonate geochemist. Your passion for science and sense of humor made all the difference in creating a positive work environment…which I know is no small feat in academia. Thanks to the program you’ve built, I’ll be leaving USC with not only a degree, but also genuine enthusiasm about the science that lies ahead. Thanks also to Josh West, who spent time guiding me through the terrestrial and silicate sides of weathering. I truly appreciate your openness, thoughtfulness on scientific and professional matters, and friendly conversation. Special thanks iii also to Jim Moffett for his dogged and effective efforts to highlight the merits of USC’s Earth Science program. Collaborations with colleagues were the most productive and enjoyable experiences of the past five years. Thanks to Seth John and Hengdi Liang for their positive attitudes and dedication to building the alkalinity model. Thanks to Xiaopeng Bian for being an excellent trace metal chemist, and just an all-around stand-up person. And thanks to Martin Van Den Berghe for your patience with microbes, curiosity about big questions, and for validating my thoughts about the similarities between academia and pyramid schemes. Many, many other colleagues offered various forms of help over the past few years. Of course, the stellar members of the Berelson lab: Nick Rollins for being a genius problem solver; Sijia Dong for infinite knowledge of carbonate mineralogy and contagious calm; and Jaclyn Pittman for unending kindness and willingness to help. Thanks to Doug Hammond for help with all things hydrological and frisbee-related. Thanks to Ken Nealson for his visionary ideas and warmth of character. Thanks to Abra Atwood for her thoughtful perspectives on academia and hiking. Thanks to Pratixa Savalia, Heidi Aronson, Didi Bojanova, and Jordan Coelho for their gracious help with O2 issues. And in general, thanks to everyone who’s helped build the USC Geology department into the warm and friendly environment I was fortunate enough to be part of. iv Table of Contents Acknowledgments........................................................................................................................... ii List of Tables ................................................................................................................................. vi List of Figures ............................................................................................................................... vii Abstract ........................................................................................................................................ viii Chapter 1: Introduction ................................................................................................................... 1 References ................................................................................................................................... 7 Chapter 2: Shewanella ondeidensis magnifies the effect of siderophores to increase olivine dissolution rates ............................................................................................................................ 10 Abstract ..................................................................................................................................... 10 2.1 Introduction ......................................................................................................................... 11 2.2 Methods ............................................................................................................................... 14 Materials preparation ............................................................................................................. 15 Batch reactor experiments ..................................................................................................... 16 Elemental analysis of experimental solution ......................................................................... 17 2.3 Results ................................................................................................................................. 20 Olivine elemental composition .............................................................................................. 20 Growth curves........................................................................................................................ 20 Silicon and metal cation release rates .................................................................................... 21 2.4 Discussion ........................................................................................................................... 26 Higher dissolution rates in biotic treatments ......................................................................... 26 Siderophores are required for biotically-enhanced olivine dissolution ................................. 28 Biotic dissolution rates exceed abiotic rates at the same siderophore concentration ............ 29 Possible mechanisms for magnification of siderophore effect .............................................. 30 Biotically-enhanced dissolution persists with DFOB addition, but slows with wild type .... 34 2.5 Conclusions and further considerations .............................................................................. 35 2.6 Acknowledgements ............................................................................................................. 36 2.7 References ........................................................................................................................... 37 2.8 Supplemental material ......................................................................................................... 42 Chapter 3: CaCO3 dissolution in carbonate-poor shelf sands increases with ocean acidification and porewater residence time .............................................................................................................. 43 Abstract ..................................................................................................................................... 43 3.1. Introduction ........................................................................................................................ 44 3.2. Methods .............................................................................................................................. 48 3.2.1 Experimental setup ....................................................................................................... 48 3.2.2 Isotope spike experiments ............................................................................................ 51 3.2.3 Sample analysis – liquid phase ..................................................................................... 52 3.2.4 Sample analysis – solid phase ...................................................................................... 54 3.2.5 Calculations .................................................................................................................. 55 v 3.3. Results ................................................................................................................................ 57 3.3.1 Sediment characterization ............................................................................................. 57 3.3.2 Consumption or production of alkalinity, DIC, and O2, and CaCO3 dissolution ......... 57 3.4. Discussion .......................................................................................................................... 63 3.4.1 Methodological considerations ..................................................................................... 63 3.4.2 CaCO3 dissolution rates vary with porewater residence time ...................................... 67 3.4.3 CaCO3 dissolution occurs at bulk porewater Ω >1 ....................................................... 70 3.4.4 CaCO3 dissolution increases in near-future ocean acidification conditions ................. 71 3.4.5 Similarity between carbonate-poor and carbonate-rich sands ...................................... 72 3.4.6 Carbonate-poor shelf sands may transition from acid source to buffering source ....... 74 3.4.7 Low-carbonate sands and ocean acidification: global considerations .......................... 76 3.5. Summary ............................................................................................................................ 78 3.6 Acknowledgements ............................................................................................................. 79 3.7. References .......................................................................................................................... 81 3.8 Supplemental Material ........................................................................................................ 89 Chapter 4: Constraining CaCO3 export and dissolution with an ocean alkalinity inverse model 95 Abstract ..................................................................................................................................... 95 4.1. Introduction ........................................................................................................................ 96 4.2. Methods ............................................................................................................................ 100 4.2.1 Model framework ....................................................................................................... 100 4.2.2 CaCO3 dissolution scenarios .......................................................................................... 105 4.2.3 Sensitivity experiments .................................................................................................. 109 4.2.4 Model optimization ........................................................................................................ 111 4.3. Results .................................................................................................................................. 116 4.3.1 Modeled vs observed TA0 and TA* ............................................................................... 116 4.3.2 CaCO3 export ................................................................................................................. 120 4.3.3 CaCO3 dissolution .......................................................................................................... 120 4.3.4 Sensitivity experiments .................................................................................................. 123 4.4. Discussion ............................................................................................................................ 124 4.4.1. Mixing alone cannot replicate TA* observations ......................................................... 124 4.4.2. Calcite and aragonite dissolution at bulk seawater Ω cannot explain observed TA* ... 125 4.4.3. CaCO3 dissolution linked to OM respiration recycles alkalinity in the upper ocean .... 126 4.4.4. Constant dissolution model similar to Ω + respiration model at depth ......................... 128 4.4.5. Potential mechanisms driving dissolution above the saturation horizons ..................... 129 4.4.6. The importance of calcite vs aragonite ......................................................................... 132 4.5. Conclusions ...................................................................................................................... 133 4.6 Acknowledgements ........................................................................................................... 134 4.7. References ........................................................................................................................ 135 4.8 Supplemental Material ...................................................................................................... 142 5. Conclusions ............................................................................................................................. 152 References ............................................................................................................................... 155 References ................................................................................................................................... 156 vi List of Tables Table 2.1. Experimental details and elemental release rates for Si and Fe. .................................. 19 Table 3.1. Measured Alkalinity, DIC and O2 fluxes, and calculated DIC added ......................... 58 Table 3S.1. Alkalinity production rates. ....................................................................................... 92 Table 3S.2. DIC production rates ................................................................................................. 93 Table 3S.3. Oxygen measurements and consumption rates .......................................................... 94 Table 4.1. Kinetic dissolution parameters for calcite and aragonite ........................................... 106 Table 4.2. Optimized parameters, model performance, and CaCO3 export and dissolution ...... 109 Table 4.S1. Optimized parameter values for the four base models ............................................ 142 vii List of Figures Figure 2.1. Growth curves for all biotic treatments ...................................................................... 21 Figure 2.2. Dissolved Si concentrations ...................................................................................... 22 Figure 2.3. Dissolved and total Mg and Fe release ....................................................................... 24 Figure 2.4. Elemental ratios of dissolution: Mg/Si and Fe/Si ....................................................... 25 Figure 2.5. Olivine dissolution rates as a function of DFOB addition ......................................... 27 Figure 2.6. Olivine dissolution compared to abiotic rates ............................................................ 31 Figure 2.S1. Dissolved and total Ni release .................................................................................. 42 Figure 3.1. Schematic of flow-through sand column .................................................................... 49 Figure 3.2. Alkalinity, DIC, CaCO3 dissolution, and porewater Ω (D) ........................................ 59 Figure 3.3. CaCO3 dissolution for three different advection rates ................................................ 60 Figure 3.4. CaCO3 dissolution rates as a function of Ωinlet ........................................................... 61 Figure 3.5. Evolution of δ 13 C for the 13 C spike experiments ........................................................ 63 Figure 3.6. Alkalinity from CaCO3 dissolution versus total alkalinity production ..................... 64 Figure 3.7. Alkalinity and DIC vectors ........................................................................................ 69 Figure 3.8. ΔAlk:ΔDIC as a function of Ωinlet ............................................................................. 75 Figure 3S.1. Breakthrough curves ............................................................................................... 91 Figure 4.1. Schematic diagram of the alkalinity model .............................................................. 101 Figure 4.2. Differences from various approaches to estimate preformed alkalinity ................... 114 Figure 4.3. Example of TA0 and TA* for M3 ........................................................................... 117 Figure 4.4. Difference between modeled and observed TA0 and TA* ...................................... 119 Figure 4.5. CaCO3 dissolution for all four base models ............................................................. 121 Figure 4.6. Global profiles of calcite and aragonite dissolution ................................................ 123 Figure 4.7. CaCO3 export for the three best-performing export sensitivity tests ...................... 129 Figure 4.S1. Data inputs for deriving CaCO3 export .................................................................. 143 Figure 4.S2. Observed and modeled TA0 and TA* for all four base models ............................. 144 Figure 4.S3. Export sensitivity tests for M1 (benthic only model) ............................................ 145 Figure 4.S4. Export sensitivity tests for M2 (Ω-dependent model) ............................................ 146 Figure 4.S5. Export sensitivity tests for M3 (Ω + respiration model) ........................................ 147 Figure 4.S6. Export sensitivity tests for M4 (constant dissolution model) ................................ 148 Figure 4.S7. Sinking speed sensitivity tests for M2 (Ω-dependent model) ................................ 149 Figure 4.S8. Sinking speed sensitivity tests for M3 (Ω + respiration model) ............................ 150 Figure 4.S9. Combined sinking speed and calcite/aragonite ratio sensitivity tests for M2 ........ 151 viii Abstract The titration of carbon dioxide (CO2) by silicate and carbonate minerals is one of the major regulators of earth’s climate. When silicate and carbonate minerals dissolve, they produce alkalinity, effectively converting CO2 into non-volatile bicarbonate and carbonate ions (HCO3 - and CO3 2- ) that remain dissolved in seawater on the timescale of 100 ky. This process, commonly called weathering, has regulated atmospheric CO2 concentration for much of earth history. Recently, however, anthropogenic CO2 emissions exceed the pace of natural mineral weathering, and CO2 is accumulating in the atmosphere at geologically unprecedented speed. The resulting climate crisis has expedited our need to better understand the details of mineral weathering -- both for predicting the weathering feedback in a higher CO2 world, and for optimizing engineered CO2 capture by “enhanced weathering”. This dissertation compiles the results of collaborative research, conducted by myself with contribution from several colleagues, exploring how various environmental factors control the dissolution rates of both carbonate and silicate minerals. Starting at the mineral surface, Chapter 2 considers how microbes use siderophores to extract Fe from olivine, effectively increasing dissolution and producing alkalinity. We find that Shewanella oneidensis increases dissolution rates by an order of magnitude above abiotic rates, and that siderophores are required for this enhancement. Furthermore, S. oneidensis appears to use siderophores synergistically with other mechanisms to achieve even higher dissolution rates than expected: in experiments with a mutant strain of S. oneidensis incapable of producing siderophores but “fed” exogenous siderophores, dissolution rates were 8-fold higher than abiotic experiments with the same siderophore concentration. These results shed light on fundamental geobiological ix mechanisms, as well as provide direction for future research on biotechnological approaches to enhanced weathering. Next, Chapter 3 considers the controls on CaCO3 dissolution rates in ocean sediments, specifically focusing on detrital sands. Such sands are among the most common sediment types on the shelf, yet almost completely unstudied with respect to carbonate chemistry. Using flow- through reactors and 13 C isotope mass balance, we show that dissolution is a function of both seawater saturation state (Ω), and porewater residence time. At decreased Ω, simulating ocean acidification, dissolution rates increased significantly and initiated sooner upon seawater advection into the sand. This response to acidification was surprisingly similar to that in higher carbonate- content sands, suggesting that detrital sediments have the potential to support enhanced dissolution in an acidifying ocean, and may be an increasing source of alkalinity to coastal waters. Finally, expanding perspective to the global ocean, Chapter 4 uses an ocean circulation inverse model (OCIM) to constrain the marine CaCO3 cycle. CaCO3 export from the surface ocean and dissolution in the water column are key controls on ocean alkalinity distribution (and thus air- sea CO2 flux), yet are poorly constrained. Testing models with different CaCO3 dissolution mechanisms against global observations of ocean alkalinity, we show that dissolution must occur above the calcite and aragonite saturation horizons, and that dissolution rates exceed those expected from known kinetics. We also show that a range of CaCO3 export values (1.1 to 1.8 Gt C y -1 ) can match observations, but that modeled exports all converge below 300 m, indicating significant recycling in the upper ocean. Collectively, these results show that dissolution is not only thermodynamically-driven, but that other mechanisms, e.g., microbial respiration in sinking particles and/or production of more soluble CaCO3, drive dissolution throughout the water column. x While not explicitly focusing on biological processes, all three chapters highlight the important role that biology—e.g., organic ligands, microbial respiration in sediments, and the coupling of organic matter with CaCO3 ballast in sinking particles—plays in controlling mineral dissolution rates. Mineral weathering is often considered a “geochemical” process, but this dissertation highlights the inextricable link between biology and geochemistry in driving global weathering processes. These results shed light on fundamental earth system processes and inform engineering solutions to CO2 sequestration by “enhanced weathering.” 1 Chapter 1: Introduction On the day I submitted this dissertation, atmospheric carbon dioxide (CO2) levels recorded at the Mauna Loa Observatory were 421 ppm, just inching past the ominous milestone of a 50% increase since pre-industrial times. CO2 is not a benign molecule, so this increase has consequences. Scientists have understood for more than 150 years that CO2 traps solar radiation as heat, insulating the earth from the cold of space and fundamentally making earth habitable (Arrhenius, 1897; Foote, 1856; Tyndall, 1861). But there can be too much of a good thing, and the rapid rise in CO2 —caused by fossil fuel emissions and land use change—is quickly becoming a bad thing. Detailed scientific studies have documented that the climate is changing rapidly, resulting in ever-more obvious loss of biodiversity, climate-related economic losses, and political instability (Bellard et al., 2012; Carleton & Hsiang, 2016; Hsiang et al., 2013; Pecl et al., 2017). This change in atmospheric CO2 concentrations is unprecedented in geologic history--save for perhaps the dinosaur-killing meteorite impact 66 million years ago—and we still do not understand many details about how the earth will respond (Gingerich, 2019; Lear et al., 2021). We do not know, for example, how the biosphere will respond to continued increases in CO2; we do not know how quickly the earth system will naturally clear the extra CO2 from the atmosphere; and, importantly, we do not know how to clean it up ourselves. To answer these questions, this dissertation explores facets of earth’s long-term carbon (C) cycle relevant to the geologic drawdown of CO2 from the atmosphere. To answer questions about atmospheric CO2, we must consider how C cycles through the earth’s various C reservoirs. Carbon can be found throughout the earth, tucked deep inside the core 2 and mantle, locked into rocks in the lithosphere, or rapidly exchanging between various pools at the earth’s surface. These pools range in size, with massive stores in the earth’s mantle (10 9 Pg C) and lithospheric rock reservoir (10 8 Pg C), to the much smaller surface (or exogenic) reservoirs, which cumulatively account for 4*10 4 Pg C (C.-T. A. Lee et al., 2019). Among these surface pools, dissolved C in the ocean accounts for approximately 90% of the total (3.8*10 4 Pg C), and gaseous C in the atmosphere (namely CO2) accounts for a tiny (albeit growing) fraction at 875 Pg C (Friedlingstein et al., 2022). Notwithstanding its place dead last in the size of earth’s C pools, atmospheric CO2 is the most critical for defining the habitability of the planet. Atmospheric CO2 is linked to the other C reservoirs by physical, chemical, and biological processes occurring at a range of timescales. For example, the rapid exchange of C between the atmosphere and biosphere by photosynthesis and respiration occurs at a manic rate of over 100 Pg C y -1 , while at the extreme other end of the spectrum, a slow trickle (<0.1 Pg C y -1 ) leaks from the lithosphere into the atmosphere from volcanism and metamorphism (and is balanced by an equally slow trickle of C returning back to earth’s depths by subduction of C-containing rocks) (C.-T. A. Lee et al., 2019). Given these large fluxes relative to the size of the atmospheric CO2 reservoir, one would expect frequent disequilibrium. However, in a classic box model experiment, Berner and Caldeira (1997) demonstrated that sources and sinks of atmospheric CO2 must be in tight equilibrium, otherwise small imbalances would lead to runaway greenhouse/icehouse conditions on geologically rapid (10 6 y) timescales. Such runaway conditions are not observed in the geologic record, however, and the earth quickly re-equilibrates following perturbations. In fact, geologic evidence indicates that earth’s climate has stayed within a relatively narrow, habitable window for most of its lifetime, over four billion years, despite changes in solar luminosity and tectonic activity 3 (Dasgupta, 2013; Sagan & Mullen, 1972). There must, then, be some kind of stabilizing feedback regulating atmospheric CO2. A geologic connection between atmospheric CO2 and the lithosphere was proposed as early as 1845 (Berner, 2012; Ebelmen, 1845). Subsequent research more than a century later clarified that atmospheric CO2 stimulates the dissolution of silicate minerals by increasing temperature and precipitation, and by decreasing the pH of rain- and seawater. When silicate minerals dissolve, proton transfer between carbonic acid and silicate anions effectively converts CO2 into non- volatile bicarbonate (HCO3 - ), thus removing it from the atmosphere (Berner et al., 1983; J. C. G. Walker et al., 1981). Using wollastonite as a stoichiometrically simple model silicate, the reaction proceeds as: CaSiO3 + 2CO2 + H2O  Ca 2+ + 2HCO3 - + SiO2 (Eq. 1) At steady state ocean chemistry, the dissolution products cannot accumulate in seawater indefinitely, so the metal and bicarbonate anions eventually precipitate in carbonate minerals: Ca 2+ + 2HCO3 -  CaCO3 + CO2 + H2O (Eq. 2) The net reaction thus transfers CO2 from the atmosphere to the lithosphere, and regulates climate on 10 5 to 10 6 y timescales (Berner & Caldeira, 1997; Broecker & Sanyal, 1998; Colbourn et al., 2015): CaSiO3 + CO2  SiO2 + CaCO3 (Eq. 3) The discussion of the weathering-climate feedback often stops here, as geologists focused on “long-term” or “slow” C cycling consider this burial of carbonate as the only geologic sink of atmospheric CO2. Considering shorter timescales, however, the silicate feedback system is too 4 slow to regulate climate (Uchikawa & Zeebe, 2008), and carbonate minerals are the major climate mediator. Considering Eq. 2 in reverse, CaCO3 dissolution shifts one mol each of solid carbonate and gaseous CO2 into the dissolved phase, thus acting as an additional atmospheric CO2 sink. Carbonate minerals are thermodynamically oversaturated in most of the surface ocean, but dissolution increases in response to ocean acidification (a direct consequence of rising atmospheric CO2) as well as to increased organic matter respiration (driven by increased primary productivity, which increases in warmer, high CO2 climates) (Archer & Maier‐Reimer, 1994; Brady & Carroll, 1994; Broecker & Peng, 1987). Because carbonate dissolution kinetics are orders of magnitude faster than silicates, carbonate dissolution effectively regulates climate on shorter timescales of 10 3 to 10 4 y, although it may also be significant at even longer timescales (Archer et al., 1998; Z. Liu et al., 2011; Zeebe, 2012; Zeebe & Westbroek, 2003). In simplest terms, weathering of both carbonates and silicates can be considered a global acid-base titration, the rate of which is controlled by many, often interlinking, variables. While there is general scientific consensus that the supply of acidic CO2, and its effect on dissolution kinetics via global temperature, is a fundamental control on the long-term climate-weathering feedback, weathering can also be driven by increased supply of mineral base through tectonics or erosion (Raymo & Ruddiman, 1992; West et al., 2005); by the co-location of CO2 and minerals in soils and sediments (Emerson & Bender, 1981); and by other factors that increase mineral dissolution kinetics such as silicate mineralogy, its geographic location in warmer climates, and organic catalysts (Amiotte-Suchet et al., 2003; Brady & Carroll, 1994; Drever & Stillings, 1997; Macdonald et al., 2019). All of these factors may interact in complex ways (Beerling et al., 2012; Hartmann et al., 2014). Because these controls are not well-defined, the timescale on which 5 weathering reacts to changes in atmospheric CO2, and thus the response to anthropogenic emissions, is still poorly constrained (Penman et al., 2020). Chapter summaries This dissertation investigates several such environmental controls on carbonate and silicate dissolution rates. Beginning at the mineral surface, Chapter 2 assesses dissolution rates of the silicate mineral olivine in the presence of the bacteria, Shewanella oneidensis. Previous research has shown that microbes can alternatively inhibit or enhance silicate dissolution rates, yet the mechanisms driving this variability are poorly understood. I focus on one specific mechanism that has been associated with dissolution enhancement: siderophores -- low molecular weight, organic molecules with extremely high affinity for iron (Fe). Laboratory studies have shown that siderophores can increase dissolution rates of olivine in abiotic conditions (Torres et al., 2019). However, how siderophores are used by microbes to affect dissolution in biotic conditions is largely unknown. Chapter 2 quantifies this effect, improving our understanding of fundamental geobiological processes, as well as providing valuable information for nascent biotechnological approaches to carbon dioxide removal. In Chapter 3, I move up in scale to consider rates of CaCO3 dissolution in a natural sand environment. While CaCO3 is oversaturated in most of the surface ocean, it nonetheless dissolves to some extent within sediments due to the spatial coupling of minerals and acids produced by the decomposition of organic matter. Furthermore, ocean acidification is driving carbonates closer to their dissolution threshold. While the effect of acidification has been quantified in coral reef sands, there are, to date, no measurements in low-carbonate, detrital sands, which cover much of the global continental shelf. Using manipulative, lab-based experiments, I provide the first 6 measurements of CaCO3 dissolution in such sediments, quantifying the effect of both ocean acidification and porewater residence time—i.e., the time seawater remains in close contact with sedimentary acids. Given their vast global area, the response of these sands to acidification may be a globally significant geologic pathway for CO2 neutralization. Moving to the global scale, in Chapter 4 I focus on global ocean CaCO3 cycling. The production and export of CaCO3 from the surface ocean effectively removes alkalinity and increases CO2 at the ocean-atmosphere interface (via Eq. 2). Thus, quantifying the amount of CaCO3 exported and where in the water column it dissolves to re-entrain CO2 as dissolved bicarbonate is important for atmospheric CO2 concentrations. Given the vast area of the ocean and seasonal dynamics, however, these processes cannot be feasibly measured and remain poorly constrained. Using an inverse model tested against global alkalinity measurements, I test various models of CaCO3 dissolution to constrain both export and dissolution patterns. All three chapters provide fundamental information on the controls of atmospheric CO2 by earth’s geologic acid-base titration system. Specifically, these chapters provide information on previously unconstrained rates of C cycling processes (Ch 3 and 4); how they are likely to change with continued increases in atmospheric CO2 and ocean acidification (Ch 3); and identifies the physical and biological mechanisms controlling these rates (Ch 2, 3, 4). These mechanisms provide insight not only about earth system processes, but also indicate pathways for designing CO2 removal strategies. 7 References Amiotte-Suchet, P., Probst, J., & Ludwig, W. (2003). Worldwide distribution of continental rock lithology: Implications for the atmospheric/soil CO2 uptake by continental weathering and alkalinity river transport to the oceans. Global Biogeochemical Cycles, 17(2). Archer, D., Kheshgi, H., & Maier‐Reimer, E. (1998). Dynamics of fossil fuel CO2 neutralization by marine CaCO3. Global Biogeochemical Cycles, 12(2), 259–276. Archer, D., & Maier‐Reimer, E. (1994). Effect of deep-sea sedimentary calcite preservation on atmospheric CO 2 concentration. Nature, 367(6460), 260–263. Arrhenius, S. (1897). On the Influence of Carbonic Acid in the Air upon the Temperature of the Earth. Publications of the Astronomical Society of the Pacific, 9, 14. https://doi.org/10.1086/121158 Beerling, D. J., Taylor, L. L., Bradshaw, C. D. C., Lunt, D. J., Valdes, P. J., Banwart, S. A., Pagani, M., & Leake, J. R. (2012). Ecosystem CO2 starvation and terrestrial silicate weathering: mechanisms and global‐scale quantification during the late Miocene. Journal of Ecology, 100(1), 31–41. Bellard, C., Bertelsmeier, C., Leadley, P., Thuiller, W., & Courchamp, F. (2012). Impacts of climate change on the future of biodiversity. Ecology Letters, 15(4), 365–377. Berner, R. A. (2012). Jacques-Joseph Ébelmen, the founder of earth system science. Comptes Rendus Geoscience, 344(11), 544–548. https://doi.org/https://doi.org/10.1016/j.crte.2012.08.001 Berner, R. A., & Caldeira, K. (1997). The need for mass balance and feedback in the geochemical carbon cycle. Geology, 25(10), 955–956. Berner, R. A., Lasaga, A. C., & Garrels, R. M. (1983). Carbonate-silicate geochemical cycle and its effect on atmospheric carbon dioxide over the past 100 million years. Am. J. Sci.;(United States), 283(7). Brady, P. V, & Carroll, S. A. (1994). Direct effects of CO2 and temperature on silicate weathering: Possible implications for climate control. Geochimica et Cosmochimica Acta, 58(7), 1853– 1856. Broecker, W. S., & Peng, T. (1987). The role of CaCO3 compensation in the glacial to interglacial atmospheric CO2 change. Global Biogeochemical Cycles, 1(1), 15–29. Broecker, W. S., & Sanyal, A. (1998). Does atmospheric CO2 police the rate of chemical weathering? Global Biogeochemical Cycles, 12(3), 403–408. Carleton, T. A., & Hsiang, S. M. (2016). Social and economic impacts of climate. Science, 353(6304), aad9837. Dasgupta, R. (2013). Ingassing, storage, and outgassing of terrestrial carbon through geologic time. Reviews in Mineralogy and Geochemistry, 75(1), 183–229. 8 Drever, J. I., & Stillings, L. L. (1997). The role of organic acids in mineral weathering. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 120(1–3), 167–181. Ebelmen, J.-J. (1845). Sur les produits de la décomposition des espèces minérales de la famille des silicates. Annales Des Mines, 7(3), 66. Emerson, S., & Bender, M. L. (1981). Carbon fluxes at the sediment-water interface: Calcium carbonate preservation. Journal of Marine Research, 39, 139–162. Foote, E. (1856). Circumstances affecting the Heat of the Sun’s Rays. The American Journal of Science and Arts, 22, 382–383. Friedlingstein, P., Jones, M. W., O’Sullivan, M., Andrew, R. M., Bakker, D. C. E., Hauck, J., Le Quéré, C., Peters, G. P., Peters, W., & Pongratz, J. (2022). Global carbon budget 2021. Earth System Science Data, 14(4), 1917–2005. Gingerich, P. D. (2019). Temporal scaling of carbon emission and accumulation rates: modern anthropogenic emissions compared to estimates of PETM onset accumulation. Paleoceanography and Paleoclimatology, 34(3), 329–335. Hartmann, J., Moosdorf, N., Lauerwald, R., Hinderer, M., & West, A. J. (2014). Global chemical weathering and associated P-release—The role of lithology, temperature and soil properties. Chemical Geology, 363, 145–163. Hsiang, S. M., Burke, M., & Miguel, E. (2013). Quantifying the influence of climate on human conflict. Science, 341(6151), 1235367. Lear, C. H., Anand, P., Blenkinsop, T., Foster, G. L., Gagen, M., Hoogakker, B., Larter, R. D., Lunt, D. J., McCave, I. N., & McClymont, E. (2021). Geological Society of London Scientific Statement: what the geological record tells us about our present and future climate. In Journal of the Geological Society (Vol. 178, Issue 1). Geological Society of London. Lee, C.-T. A., Jiang, H., Dasgupta, R., & Torres, M. (2019). A Framework for Understanding Whole-Earth Carbon Cycling. In B. N. Orcutt, I. Daniel, & R. Dasgupta (Eds.), Deep Carbon: Past to Present (pp. 313–357). Cambridge University Press. Liu, Z., Dreybrodt, W., & Liu, H. (2011). Atmospheric CO2 sink: silicate weathering or carbonate weathering? Applied Geochemistry, 26, S292–S294. Macdonald, F. A., Swanson-Hysell, N. L., Park, Y., Lisiecki, L., & Jagoutz, O. (2019). Arc- continent collisions in the tropics set Earth’s climate state. Science, 364(6436), 181–184. Pecl, G. T., Araújo, M. B., Bell, J. D., Blanchard, J., Bonebrake, T. C., Chen, I.-C., Clark, T. D., Colwell, R. K., Danielsen, F., & Evengård, B. (2017). Biodiversity redistribution under climate change: Impacts on ecosystems and human well-being. Science, 355(6332), eaai9214. Penman, D. E., Rugenstein, J. K. C., Ibarra, D. E., & Winnick, M. J. (2020). Silicate weathering as a feedback and forcing in Earth’s climate and carbon cycle. Earth-Science Reviews, 209, 103298. 9 Raymo, M. E., & Ruddiman, W. F. (1992). Tectonic forcing of late Cenozoic climate. Nature, 359(6391), 117–122. Sagan, C., & Mullen, G. (1972). Earth and Mars: Evolution of atmospheres and surface temperatures. Science, 177(4043), 52–56. Torres, M. A., Dong, S., Nealson, K. H., & West, A. J. (2019). The kinetics of siderophore‐ mediated olivine dissolution. Geobiology, 17(4), 401–416. Tyndall, J. (1861). I. The Bakerian Lecture.—On the absorption and radiation of heat by gases and vapours, and on the physical connexion of radiation, absorption, and conduction. Philosophical Transactions of the Royal Society of London, 151, 1–36. https://doi.org/https://doi.org/10.1098/rstl.1861.0001 Uchikawa, J., & Zeebe, R. E. (2008). Influence of terrestrial weathering on ocean acidification and the next glacial inception. Geophysical Research Letters, 35(23). Walker, J. C. G., Hays, P. B., & Kasting, J. F. (1981). A negative feedback mechanism for the long‐term stabilization of Earth’s surface temperature. Journal of Geophysical Research: Oceans, 86(C10), 9776–9782. West, A. J., Galy, A., & Bickle, M. (2005). Tectonic and climatic controls on silicate weathering. Earth and Planetary Science Letters, 235(1–2), 211–228. Zeebe, R. E. (2012). History of seawater carbonate chemistry, atmospheric CO2, and ocean acidification. Annual Review of Earth and Planetary Sciences, 40, 141–165. Zeebe, R. E., & Westbroek, P. (2003). A simple model for the CaCO3 saturation state of the ocean: The “Strangelove,” the “Neritan,” and the “Cretan” Ocean. Geochemistry, Geophysics, Geosystems, 4(12). 10 Chapter 2: Shewanella ondeidensis magnifies the effect of siderophores to increase olivine dissolution rates Contributors: Martin Van Den Berghe, Xiaopeng Bian, Seth John, Ken Nealson, Josh West Abstract Dissolution of mafic minerals like olivine is an important source of both metals and alkalinity to Earth’s surface. Environmental controls on dissolution rates remain poorly understood, however. Here, we conducted experiments that isolated the effect of microbial siderophores on olivine dissolution rates in both biotic and abiotic conditions. We find that Shewanella oneidensis increases dissolution of olivine above abiotic rates by an order of magnitude, and that siderophores are required for this enhancement. In solutions with a mutant strain of S. oneidensis incapable of producing siderophores, the microbial population did not grow, and olivine dissolution was below detection. With the addition of the siderophore deferoxamine B (DFOB) at concentrations >0.2 μM, however, both microbial growth and olivine dissolution rates increased. With sufficient DFOB to facilitate growth, dissolution rates with S. oneidensis exceeded abiotic rates at the same DFOB concentration by 8-fold, implying a synergy between siderophores and other biotic mechanisms. After reaching stationary phase, dissolution rates in the wild type treatment slowed, implying a reduction in siderophore production and little to no excess siderophores in solution. In contrast, dissolution in the mutant + DFOB treatments continued at high rates. These results 11 suggest that while siderophores are necessary for biotically-enhanced olivine dissolution, other microbe-related mechanisms magnify their effect. 2.1 Introduction The chemical weathering of silicate minerals is an important control on global biogeochemical cycling and climate via the release of metal cations and alkalinity. Biologically critical metals, in particular iron (Fe), are limited in the environment due to low solubility of Fe oxide minerals, so replenishment from silicate dissolution can be a significant control on biological productivity. Simultaneously, alkalinity from silicate dissolution is thought to be the major control on atmospheric carbon dioxide (CO2) levels at the >10 5 y timescale (Berner et al., 1983; J. C. G. Walker et al., 1981). Recently, however, geologically rapid CO2 emissions from anthropogenic activities (11 Gt C y -1 ) far exceed natural rates of silicate weathering (estimated at 0.2 Gt C y -1 ), leading to atmospheric CO2 accumulation and global warming (Friedlingstein et al., 2020). In response, researchers are actively seeking ways to remove excess CO2 from the atmosphere, including approaches using “enhanced weathering”, which seek to stimulate silicate dissolution in terrestrial and/or ocean environments (Hartmann et al., 2013; Köhler et al., 2013; Meysman & Montserrat, 2017; Montserrat et al., 2017). Thus, improving our understanding of the environmental controls on silicate dissolution is important, both for quantifying the climate- weathering feedback, and for improving the efficiency of engineered enhanced weathering efforts. In studies of silicate weathering, olivine ((Mg 2+ , Fe 2+ )2SiO4) is often studied as a model mineral due to its relatively fast dissolution rates. As an orthosilicate, olivine’s tetrahedral silicate anions are bound only to surrounding metals, enabling complete dissolution via breaking of 12 relatively weak metal-oxygen ionic bonds. Furthermore, olivine is a common component of mid- ocean ridge basalt and other volcanic rocks. As a result of both its prevalence and high solubility, olivine-rich rocks (such as basalt) are disproportionately significant in earth’s silicate weathering flux and supply of Mg and Fe to the biosphere (Dessert et al., 2003; Hartmann et al., 2009). For the same reasons, olivine and basalt are the most common focus of enhanced weathering studies. At earth’s surface, olivine dissolution kinetics can vary by orders of magnitude as a function of both abiotic and biotic environmental factors. The effect of abiotic conditions such as temperature, solution pH, and solution chemistry are relatively well constrained (Oelkers et al., 2018; Rimstidt et al., 2012). The effect of biology on dissolution rates, on the other hand, are less clear. While the net effect of biology on mineral dissolution in general is typically assumed to be positive (Schwartzman & Volk, 1989), some lab-based olivine studies have shown no significant effect (Shirokova et al., 2012), or significant inhibition of dissolution by microbes (Garcia et al., 2013; Oelkers et al., 2015; Santelli et al., 2001). Conversely, enhancement of dissolution has been documented at the field and laboratory scale, associated with microbes, lichen, vegetation, and even insects (Brady et al., 1999; Dorn, 2014; Gerrits et al., 2020; Lamérand et al., 2020; Pokharel et al., 2019; Wild et al., 2018, 2021). In these cases, dissolution can exceed abiotic rates by more than an order of magnitude, suggesting that biological processes have at least the potential to significantly enhance dissolution, even if enhancement is not universal. The mechanisms driving this enhancement remain poorly understood, but Al- and Fe-bearing minerals like olivine may be particularly susceptible to dissolution enhancement by the metal-binding action of organic ligands (Berner, 2010; Pokrovsky et al., 2009). In this respect, low molecular weight, multi-dentate ligands known as siderophores may be particularly important drivers of microbially-enhanced olivine dissolution. Siderophores are 13 secreted by many bacteria, fungi, and grasses in response to Fe limitation (Hider & Kong, 2010). While they are structurally diverse, siderophores typically contain multiple metal-binding ligands that collectively result in an exceptionally high binding affinity with Fe 3+ (logKf >30), and to a lesser degree, Al 3+ (Hider & Kong, 2010; Hofmann et al., 2020). In contrast, other low molecular weight organic acids also bind Fe 3+ , but with a much weaker affinity (e.g., logKf for oxalic acid = 8.8). As a result, siderophores are an extremely effective biological tool for retaining Fe in solution, preventing its loss by Fe-oxide precipitation. Significant research has shown that purified siderophores can abiotically enhance dissolution rates for a range of Fe- and Al-containing minerals, including Fe-oxides (Cheah et al., 2003; Kraemer, 2004; Reichard et al., 2007), Fe- containing phyllosilicates (Bray et al., 2015; Ferret et al., 2014; Haack et al., 2008; Rosenberg & Maurice, 2003; Shirvani & Nourbakhsh, 2010), and hornblende (Buss et al., 2007; Kalinowski et al., 2000; Liermann et al., 2000). These minerals all contain predominantly Fe 3+ and/or Al 3+ , both of which are chelated by siderophores. Olivine, in contrast, contains Fe almost exclusively in the +2 oxidation state, which is more weakly bound by siderophores (e.g., for the siderophore deferoxamine B (DFOB), logKf = 30 for Fe 3+ , vs 10 for Fe 2+ ) (Dhungana & Crumbliss, 2005). Indeed, the reduction of Fe 3+ to Fe 2+ is the mechanism by which microbes liberate Fe bound by hydroxamate siderophores like DFOB. Nonetheless, recent research has shown that, similar to Fe 3+ -oxides and silicates, siderophores can also increase olivine dissolution rates by nearly an order of magnitude (Torres et al., 2019). An important distinction, however, is that these experiments use purified siderophores at high micromolar concentrations, whereas environmental concentrations are consistently much lower (pico- to nanomolar) (Kraemer, 2004; Page & Huyer, 1984; A. Perez et al., 2016). Thus, there remains some uncertainty as to whether siderophores are 14 ultimately responsible for the correlated biotic dissolution enhancement, or if other mechanisms, such as lowered pH or other organic ligands, are more important. This study attempts to clarify whether siderophores are required for biotically enhanced dissolution, and furthermore, if they are the key mechanism of dissolution enhancement. Or, alternatively, if they act in tandem with other microbial processes. Along these lines, are sub- micromolar concentrations of siderophores—which do not significantly increase dissolution abiotically—sufficient to increase dissolution in biotic conditions? While previous studies have shown that siderophores are indeed critical for microbes to access mineral-bound Fe, they have not linked this dependency quantitatively to mineral dissolution rates (Dehner et al., 2010; Ferret et al., 2014; Van Den Berghe et al., 2021). Here, we focus specifically on this link between microbial use of siderophores and rates of mineral dissolution. 2.2 Methods The experiment consisted of batch reactors containing olivine grains, Fe-deplete growth medium, and the following treatments: wild type Shewanella oneidensis (MR-1); a gene-deletion mutant of MR-1 incapable of producing siderophores (ΔMR-1); ΔMR-1 with added deferoxamine B (∆MR-1 +DFOB, ranging from 0 to 50 μM); ΔMR-1 killed control; and abiotic treatments with no microbes but added DFOB (ranging from 0 to 50 μM). All treatments, listed in Table 2.1, were conducted in triplicate. Henceforth, the live MR-1 and ΔMR-1 treatments are collectively referred to as “biotic”, whereas the solutions with no microbial addition and the killed control are referred to as “abiotic”. Note that ΔMR-1 (a.k.a. ΔSO3031) was previously characterized as being incapable of producing siderophores but capable of utilizing them through the hydroxamate 15 reductase pathway (Fennessey et al., 2010). Furthermore, MR-1 is capable of taking up a range of siderophores, including the tris-hydroxamate DFOB, in addition to its native di-hydroxamate siderophore, putrebactin (L. Liu et al., 2018; Van Den Berghe et al., 2021). Materials preparation Olivine grains were collected from the University of Southern California mineral collection and were hand-crushed with a clean mortar and pestle. Crushed material was sieved to 150 – 300 μm diameter, ultrasonicated and rinsed 7 times (5 min sonication per rinse) in 200-proof ethanol, then air-dried in an oven at 130° C overnight. Immediately before the experiments, olivine grains were UV-sterilized for 30 min. Elemental composition of the olivine was analyzed by x-ray fluorescence (XRF; Bruker S8 Tiger). By assuming a spherical shape and binning into six grain sizes with a normal distribution, the grain size distribution yielded a mean mineral surface area of 10,640 mm 2 per g. All experiments and sample analyses were performed in a laminar flow hood to prevent microbial and/or metal contamination. Furthermore, experiments and all media preparation were performed in acid-washed polycarbonate or polypropylene containers to avoid potential metal contamination from biotically induced glass dissolution (Aouad et al., 2006; Gorbushina & Palinska, 1999). The pH 7.2 growth medium was based on the M-1 minimal medium (Gorby et al., 2006), slightly modified by using MOPS buffer (50 mM) and N-acetyl glucosamine (18 mM) as the carbon source and electron donor, in place of carboxylic acids that are known to enhance mineral dissolution abiotically (Neaman et al., 2005; Olsen & Rimstidt, 2008). Furthermore, Fe was omitted from the growth medium so that olivine was the only available Fe source. A detailed description of the growth medium composition is provided in Van Den Berghe et al (2021). pH was confirmed at the beginning of the experiment, and given the high buffering capacity and 16 relatively short duration of the experiment, was assumed to not change significantly over the course of the experiment. For the biotic experiments, inoculation cultures were grown from individual isolated colonies and conditioned prior to the experiments in the same minimal medium (though Fe-replete, with 3.6 μM FeSO4), without olivine. Microbes were extracted from these growth solutions by filtering (<0.2 μm), then triple-rinsed and concentrated in Fe-free medium prior to inoculation. Batch reactor experiments Batch reactor experiments were performed in Erlenmeyer flasks exposed to atmospheric conditions via porous caps, maintained in the dark at 30 C, and shaken continuously at 120 rpm. 100 mL of the Fe-deplete medium was added to each flask, followed by inoculation cultures (for biotic experiments), and siderophore solution (for DFOB addition experiments). 100 mg olivine was then added to each flask. Flasks for the killed ΔMR-1 control treatment were autoclaved after inoculation, but prior to adding olivine. For all biotic experiments, flasks were inoculated with approximately 5 × 10 10 cells (for a starting cell concentration of approximately 5 × 10 8 cells mL -1 ). This initial cell density was selected to rapidly achieve maximum cell concentration. At each sampling time, 1 mL samples were extracted to monitor growth by optical density (OD) measurements, using pre-determined OD vs cell count relationships (600 nm wavelength on Shimadzu UV-2600 spectrophotometer). Additional information on cell count methods is presented in Van Den Berghe et al (2021). For the ΔMR-1 +DFOB and abiotic DFOB treatments, a filter-sterilized stock solution of deferoxamine mesylate (Sigma Aldrich, US) was added to reach final concentrations ranging from 0.05 to 50 μM (Table 2.1). A maximum of 50 μM was used based on prior findings that >5 μM DFOB addition was sufficient to stimulate maximum ΔMR-1 growth (Van Den Berghe et al., 17 2021). While DFOB is a tris-hydroxamate, structurally different from the cyclic di-hydroxamate putrebactin, it is known to be readily bioavailable to ΔMR-1. Most experiments were run for approximately 48 h, just long enough for the biotic treatments to complete exponential growth and maintain stationary phase for 12+ hours. A subset of experiments was run for 79 h to assess dissolution rates over a longer time period. Over the course of the experiments, samples were extracted (without volume replacement) for elemental analyses, ensuring homogenous sampling of suspended material by gently swirling bottles. Elemental analysis of experimental solution For experiments marked “OES only” in Table 2.1, dissolved Si was measured by Inductively Coupled Plasma - Optical Emissions Spectrometry (ICP-OES, model 5110, Agilent, US), at emission wavelengths of 251.611 nm. Because dissolved metal measurements in these experiments were either below ICP-OES detection (Fe) or lower than the medium initial values (Mg), we ran additional experiments for analysis of dissolved and solid phase metals by inductively coupled plasma – mass spectrometry (ICP-MS, Thermofisher Scientific Element2); these are marked “ICPMS, OES” in Table 2.1. For the OES only experiments, ~5 mL samples were taken at each timepoint, then filtered, acidified to ~0.5% HNO3, and preserved in the dark at 4° C until analysis. Because Si does not have any biological function in Shewanella and is not a component of the minimal medium, dissolved Si was assumed to represent total dissolution flux from the mineral. For the ICP-MS, OES experiments, ~5 mL samples were taken at each timepoint, and separated into unfiltered and filtered (0.2 μm) aliquots. All samples were then acidified to 2% HNO3 and stored in the dark at room temperature until analysis. Prior to analysis, 100 μL aliquots were digested to prevent analytical interference by the organic matrix. Aliquots were added to PFA vials, acidified with 1 18 mL each of concentrated HCl and HNO3, and dried at 120 C overnight. Subsequently, dried samples were digested once more by adding 1 mL concentrated HNO3, heating at 120 C for 2-3 h, then drying at 120 C overnight. Finally, samples were reconstituted with 1 mL 0.1 N HNO3, with the addition of 10 ppb indium (In) as the internal standard to monitor analytical drift and correct for matrix effects. Measurements were taken for all metals present in the mineral (Table 2.1). This digestion procedure resulted in significant Si loss, so additional aliquots of undigested, filtered samples were analyzed for Si on the OES. For the OES analysis, 0.5 to 1 mL aliquots were reconstituted to 2.5 mL with Si-free growth medium, acidified to 2% HNO3, then analyzed on the OES as described above. Data analysis Erroneous measurements were removed from the dataset via an automated quality control procedure, with a few additional erroneous points removed manually. Measured Si and metal concentrations at each timepoint (Ci) were corrected (Ci, correct) to consider volume extracted at prior timepoints: 𝐶𝐶 𝑖𝑖 , 𝑐𝑐 𝑐𝑐 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑐𝑐 = (𝐶𝐶 𝑖𝑖 𝑉𝑉 𝑖𝑖 + ∑ 𝑣𝑣 ( 𝑖𝑖 − 1) 𝑗𝑗 = 1 𝐶𝐶 𝑗𝑗 )/𝑉𝑉 0 where Vi is the experimental volume remaining prior to sampling time i, v is the sample volume, Cj is the measured concentration at each prior timepoint, and V0 is the initial volume (100 mL). Dissolution rates were inferred from changes in Si concentrations because Si, unlike Mg or Fe, is not used by Shewanella and is not a component of the minimal medium. Furthermore, Mg is highly exchangeable with protons at the initiation of dissolution experiments, and Fe easily precipitates as insoluble Fe-oxide secondary minerals, making both metals potentially misleading for calculation of net dissolution rate (Oelkers et al., 2018; Reichard et al., 2007). Dissolution rates were calculated based on the slope of corrected Si concentration vs time (∆Si/∆t), considering only 19 timepoints between 23 and 53 h, so as to exclude early timepoints with little microbial activity and later timepoints with significant accumulation of dead cells. This window captured the transition from exponential growth to stationary phase. These rates were normalized by either mineral mass (g -1 ) or mineral surface area (m -2 ). Rates were calculated for each bottle individually, then replicate bottles for the same treatment were averaged to yield a single dissolution value for each treatment. The dissolution rate for each treatment was considered statistically significant if at least two of the three replicate bottles had p values <0.05. Table 2.1. Experimental details and elemental release rates for Si and Fe. Release rates were calculated using data between 23 to 53 h. Values shown are mean (± standard deviation) of triplicate bottles. n.s. indicates not significantly different from 0; -- indicates not measured Category Treatment DFOB (μM) Experiment Length (h) Analysis Si release (μmol g -1 h -1 ) Fe release (μmol g -1 h -1 ) Biotic MR-1 0 79 ICPMS, OES 0.23 ±0.01 0.04 ±0.01 MR-1 0 46 OES 0.25 ±0.02 -- ∆MR-1 0 79 ICPMS, OES n.s. n.s. ∆MR-1 0 53 OES n.s. -- ∆MR-1 +DFOB 0.05 53 OES n.s. -- ∆MR-1 +DFOB 0.2 79 ICPMS, OES n.s. n.s. ∆MR-1 +DFOB 1 53 OES 0.06 ±0.01 -- ∆MR-1 +DFOB 50 79 ICPMS, OES 0.23 ±0.01 0.05 ±0.01 ∆MR-1 +DFOB 50 46 OES 0.27 ±0.01 -- Abiotic Killed ∆MR-1 0 79 ICPMS, OES n.s. n.s. Killed ∆MR-1 0 46 OES n.s. -- Control 0 48 OES n.s. -- DFOB 0.05 48 OES 0.02 ±0.01 -- DFOB 1 48 OES 0.02 ±0.01 -- DFOB 50 48 OES 0.04 ±0.01 -- 20 2.3 Results Olivine elemental composition The cation composition of the ground olivine was 87.1% Mg, 11.9% Fe, and 0.5% each of Ca and Ni, with trace amounts of Mn, Cr, Co, Zn, and Cu (Table 2.S1). The overall chemical composition is summarized as Mg1.76Fe0.24SiO4 (i.e., Fo88; predominantly forsterite with approximately 12% Fe-containing fayalite in the solid solution). Growth curves Both MR-1 and ∆MR-1 with added DFOB of at least 0.2 μM exhibited exponential cell growth, reaching stationary phase between 24 to 40 h (Fig 2.1). The mutant ∆MR-1 scaled with DFOB addition, with no growth at DFOB <0.2 μM, significant but inhibited growth with added DFOB between 0.2 and 1 μM, and maximum growth reached at 50 μM DFOB addition. Previous research showed that cell density for ∆MR-1 plateaus at approximately 5 μM DFOB addition, so ∆MR-1+DFOB(50 μM) was not siderophore-limited (Van Den Berghe et al., 2021). Wild type MR-1 stationary cell density (8 × 10 9 cells mL -1 ) was nearly double that of the ∆MR-1 +DFOB(50 μM) treatment (4 × 10 9 cells mL -1 ). 21 Figure 2.1. Growth curves for all biotic treatments. For ∆MR-1 treatments, darker blue shades indicate increasing DFOB addition. All experiments performed in triplicate (error bars are standard deviation). The MR-1 and ∆MR-1 +DFOB(50 μM) experiments were conducted twice; each experimental batch is shown as a separate line in the figure. Silicon and metal cation release rates Dissolved Si was measurable in all samples, but the release rate during the stationary phase window considered was only significant for some treatments (Fig 2.2A, 2.2B; Table 2.1). In the abiotic treatments, Si release scaled with DFOB, ranging from 0.02 to 0.04 μmol g -1 h -1 , and was significant for all DFOB additions, including the lowest concentration of 0.05 μM DFOB. In contrast, Si release in the ∆MR-1 + DFOB treatments had higher variability, and was not significant until at least 1 μM DFOB addition. Once detectable, Si release rates in the biotic treatments were higher than the abiotic treatments at the same DFOB concentration, reaching 0.27 μmol g -1 h -1 in the ∆MR-1 +DFOB(50 μM) treatment. Si release in wild type MR-1 was comparably high, reaching 0.25 μmol g -1 h -1 . The change in Si release rate over time also varied between biotic and abiotic experiments. In abiotic conditions, Si release was fastest between 0 and 22 24 hours, followed by relatively slower rates, similar to other abiotic experiments with siderophores (Torres et al., 2019). In contrast, the biotic experiments typically had relatively constant release rates during the measurement period, and if anything, slower initial rates. However, wild type MR-1 uniquely exhibited a slowing of Si release rates shortly after reaching stationary phase, after approximately 40 h. Figure 2.2. Dissolved Si concentration over time for all biotic (A) and abiotic (B) experiments. Colors in both figures correspond to the same concentration of DFOB addition. At the same DFOB concentration, Si concentrations are lower in the abiotic treatments. All experiments performed in triplicate (error bars are standard deviation). In contrast to Si, total Mg was initially stable or decreased slightly, followed by an increase after approximately 24 h (Fig. 2.3A). Only the ∆MR-1 +DFOB(50 μM) treatment showed sustained Mg release through the end of the experiment, whereas the other treatments stabilized after about 40 h. Replicate variability for these latter timepoints was relatively high, however, A B 23 precluding clear trends. Considering dissolved Mg, both biotic treatments with high growth rates (MR-1 and MR-1∆ +DFOB(50 μM)), showed a significant reduction in dissolved Mg during the exponential growth phase, with an increase in dissolved Mg thereafter. Total Fe more closely mirrored Si, with increasing Fe in all biotic treatments with siderophore addition (Fig 2.3B). For the time period considered (23-53 h), however, release rates were only significant in the MR-1 and ∆MR-1 +DFOB(50 μM) treatments. Considering the later timepoint, ∆MR-1 +DFOB(0.2 μM) also had a significant Fe release rate. The ∆MR-1 +DFOB(50 μM) treatment showed a consistent release rate of 0.05 μmol g -1 h -1 , whereas the release rate for MR-1 wild type was initially similar, but slowed after stationary phase was achieved at 40 h. Prior to 40 h, there was no significant dissolved Fe in any treatment, implying the total Fe increase was entirely shunted to biomass. (In this study, we assume particulate Fe and Mg to be in biomass. While it is possible that metals may adsorb to cell exteriors, a distinction between internal or adsorbed Fe does not affect our analysis.) In the MR-1 treatment, dissolved Fe remained near-0 for the duration for the experiment. However, in all ∆MR-1 treatments, dissolved Fe increased after 40 h, following the total Fe trend, implying accumulation of Fe in dissolved phase but no change in the particulate Fe pool. 24 Figure 2.3. Dissolved and total Mg (A) and Fe (B) concentrations versus time for the subset of experiments analyzed by ICP-MS. Solid lines represent total (unfiltered) elemental concentrations, and dashed lines represent dissolved (filtered) concentrations. All experiments performed in triplicate (error bars are standard deviation). Compared to the known mineral stoichiometry, elemental ratios of Si, Mg, and Fe showed disproportionate Si release for the first ~30 h of the experiment (Fig 2.4A, 2.4B). After 40 h, Mg release increased for all treatments and Mg/Si appeared to remain congruent for the duration of the A B 25 experiment, although sample error was high. Total Fe followed a similar pattern for the ∆MR-1 treatments, with congruent Fe/Si after approximately 30 h. For the MR-1 and control treatments however, Fe/Si remained depressed at about 60% of expected stoichiometry. Figure 2.4. Elemental ratios of total Mg to dissolved Si release (A), and total Fe to dissolved Si release (B) versus time. Dissolved Si is assumed equivalent to total Si. Dashed lines represent stoichiometry of the source olivine. All experiments performed in triplicate (error bars are standard deviation). A B 26 For trace metals, only Ni exhibited any measurable release, and then only for the ∆MR-1 +DFOB(50 μM) treatment, with a release rate of 0.004 (±0.001) μmols Ni g olivine -1 h -1 (Supplemental Fig. 2.S1). Dissolved and total Ni concentrations were similar, implying that Ni was not taken up by, or adsorbed to biomass. All other metals showed no trends over time that exceeded analytical uncertainty. 2.4 Discussion Higher dissolution rates in biotic treatments All biotic treatments with microbial growth had significantly higher olivine dissolution rates than the abiotic treatments (Fig. 2.5). In the abiotic control with 0 μM DFOB addition, dissolution was similar to that expected from the pH- and temperature-dependent regression derived by Rimstidt et al (2012): 2.8 × 10 -10 mol m -2 s -1 , compared to the expected rate of 6.2 × 10 - 10 mol m -2 s -1 . (Note that including earlier timepoints, not only those during our stationary phase window >23 h, results in a slightly higher dissolution rate of 4.5 × 10 -10 mol m -2 s -1 .) In contrast, both the MR-1 and ∆MR-1 +DFOB(50 μM) treatments had dissolution rates an order of magnitude higher, reaching 6.2 × 10 -9 mol m -2 s -1 and 6.6 × 10 -9 mol m -2 s -1 , respectively. Such enhancement, on the scale of an order of magnitude, has been observed in other lab-based studies comparing biotic (including bacteria and fungi) and abiotic olivine dissolution rates (Gerrits et al., 2020; Lamérand et al., 2020; Pokharel et al., 2019). This enhancement lies in contrast, however, to other lab-based studies showing suppressed dissolution by bacteria (Garcia et al., 2013) and mixed microbial communities (Oelkers et al., 2015). Biotic inhibition is thought to be caused by the formation of organic passivating layers, consisting of live or dead cells and/or extracellular material on the mineral surface. For example, Lamérand et al (2020) found that olivine dissolution 27 rates were an order of magnitude lower when incubated in direct contact with Pseudomonas and Synechococcus, versus when grains were protected from colonization by dialysis membrane. In our experiments, however, SEM images indicated that MR-1 mostly remained in planktonic form, with only a small number of individuals attached to the mineral surface. This difference between permanent surface colonization leading to dissolution inhibition vs temporary colonization leading to enhancement may be an important distinction in the effect of microbes on dissolution rates. Thus, the significantly enhanced rates in our experiment may be a unique product of short-term, ideal conditions, wherein organic passivation of the mineral surface does not have time to develop. Perez et al (2016) saw similar enhanced dissolution of Fe-containing basaltic glass by Pseudomonas during short incubations, in contrast to inhibited dissolution in similar, but longer- term experiments (Stockmann et al., 2012). Figure 2.5. Olivine dissolution rates as a function of DFOB addition. Dissolution rates in biotic treatments (∆MR-1) exceed abiotic treatments at the same DFOB addition for DFOB>1 μM. At 50 μM DFOB addition, biotic dissolution rates exceed abiotic rates by 8-fold. Dashed line represents dissolution rate for wild type MR-1, with no exogenous siderophore addition. 28 Siderophores are required for biotically-enhanced olivine dissolution The observed dissolution enhancement in the biotic treatments was contingent on siderophore availability. The growth data presented here (Fig. 2.1), as well as additional data in a prior study (Van Den Berghe et al., 2021) show clearly that, without siderophores, ∆MR-1 cannot access olivine-bound Fe at rates fast enough to support growth, and correspondingly cannot enhance mineral dissolution. DFOB addition of at least 0.2 μM was required for measurable microbial growth and statistically significant Si release, although both at lower levels than observed in MR-1 wild type. Studies using mutant varieties of Pseudomonas sp. grown with Fe 3+ - containing oxides and clays as the only Fe source similarly showed reduced growth for siderophore-lacking mutants (Dehner et al., 2010; Ferret et al., 2014; Kuhn et al., 2013). However, in these studies, mutants were also able to access mineral-bound Fe in the absence of siderophores, albeit at apparently lower rates, by using other ligands or reductants associated with biofilm. Like Pseudomonas, Shewanella has the capacity to take up Fe through alternative pathways not involving siderophores, including direct import of Fe 2+ , assimilatory reduction of Fe 3+ to Fe 2+ , and/or ligand-bound Fe 3+ transported into the cell via the Feo system (Hersman et al., 2000; Liu et al., 2018). Thus, a dependence on siderophores for microbial growth is not obvious a priori, as long as Fe can somehow be liberated from the mineral at sufficient rates. Other processes associated with microbial growth, including pH- or ligand-promoted dissolution could facilitate such dissolution (Olsen & Rimstidt, 2008). For example, low molecular weight organic acids (e.g., oxalate) secreted by cells are widely known to increase dissolution rates of Fe-containing minerals, including olivine, at circum-neutral pH (Hausrath et al., 2009; Y. Liu et al., 2006; Neaman et al., 2005; Olsen & Rimstidt, 2008; Wogelius & Walther, 1991). However, this enhancement occurs only at high ligand concentrations (at least 10 μM, and often up to 1 mM), which may not be 29 achieved by slow-growing microbes (Cheah et al., 2003; Reichard et al., 2007). Similarly, pH gradients may develop in microbial biofilms, but they may not be severe enough to significantly affect dissolution (Gerrits et al., 2020; Liermann et al., 2000). While these or similar processes appear to be sufficient to facilitate microbial growth on Fe-oxides and clays (as cited above), our ∆MR-1 treatment, which had no significant growth or measurable dissolution, indicates that siderophores are required for access of olivine-bound Fe. This difference suggests that Fe 2+ bound within the olivine structure may be more difficult for microbes to access without siderophores than Fe 3+ in oxides or clays. Biotic dissolution rates exceed abiotic rates at the same siderophore concentration Purified siderophores, independent of live culture, have been shown to stimulate the dissolution of many Fe-containing minerals, including olivine (Torres et al., 2019). We show here, however, that live bacteria enhance this dissolution even further, beyond the rates expected for siderophore-associated dissolution alone. In other words, S. oneidensis does not simply benefit from passive use of siderophores, but rather, somehow magnifies their efficiency. In all ∆MR-1 treatments where microbial growth was detectable (DFOB > 0.2 μM), dissolution rates exceeded those in abiotic conditions at the same DFOB concentration. Even at 1 μM DFOB addition, at which concentration the effect of siderophores in abiotic solution would be miniscule, dissolution was measurably enhanced in the presence of Shewanella. At 50 μM DFOB addition, biotic dissolution exceeded abiotic rates by 8-fold. While this enhancement clearly relies on microbial activity, dissolution rates do not scale directly with cell concentrations; rather, rates are approximately constant, and do not follow the exponential pattern of cell growth. Furthermore, dissolution rates in the MR-1 wild type and ∆MR-1 +DFOB(50 μM) treatments are similar, despite 30 a 2-fold difference in cell density. Thus, the rate-limiting step in the biotic dissolution process is not apparently a function of cellular activity, but likely related instead to mineral surface processes. Possible mechanisms for magnification of siderophore effect One possible mechanism explaining higher rates in biotic vs abiotic conditions is simply the microbial recycling of siderophores. In abiotic conditions, siderophores in solution can become saturated with chelated Fe, lowering the free siderophore concentration available to interact with the mineral surface. Active microbial populations, however, can transfer Fe to biomass, thus freeing siderophores for reuse. Our total vs dissolved Fe data would support this mechanism, as total Fe accumulates, but chelated Fe in solution does not, at least for the first 40 h of the ∆MR-1 experiments, and for the entirety of the experiment with MR-1. Thus, all released Fe is efficiently transferred to particulate matter, which we assume to be biomass. Even at 50 μM DFOB addition, which exceeds the minimum concentration known to facilitate maximum growth of ∆MR-1 (approximately 5 μM; Van Den Berghe et al., 2021), chelated Fe does not accumulate in solution until well after stationary phase is reached, suggesting highly efficient siderophore use and shuttling of siderophore-bound Fe into biomass. However, previous work has shown that mineral dissolution does not necessarily scale with siderophore concentration. Rather, dissolution is generally a linear function of siderophores adsorbed to the mineral surface, which plateaus with respect to siderophore concentration according to an adsorption isotherm (Furrer and Stumm; Kraemer 1999; Cheah, 2003). As a result, dissolution rates similarly plateau with increasing siderophore concentration. For DFOB, this has been well-documented for olivine, Fe-oxides and Fe-containing phyllosilicates, with maximum dissolution rates reached as low as 10 μM DFOB, although sorption is highly dependent on 31 mineralogy and solution pH (Cervini-Silva & Sposito, 2002; Cheah et al., 2003; Haack et al., 2008; Kraemer, 2004; Shirvani & Nourbakhsh, 2010; Torres et al., 2019). Considering the relationship between olivine dissolution and DFOB concentration published by Torres et al (2019)—which was derived at experimental conditions similar to ours (30 C, pH 7.5)—dissolution rates plateau at >100 μM DFOB, suggesting that our 50 μM DFOB treatment may benefit from siderophore recycling (Fig. 2.6). However, our measured dissolution rates for MR-1 and ∆MR-1 +DFOB(50 μM) are well above the bounds of the isotherm, suggesting that siderophore recycling alone does not fully explain biotic dissolution rates. Some other biotic mechanism must facilitate increased siderophore efficiency, for example by increasing adsorption capacity, desorption efficiency, or otherwise making Fe more labile for uptake by siderophores. Figure 2.6. Olivine dissolution rates as a function of DFOB addition, compared to abiotic rates documented by Torres et al (2019). Filled circles are data from this study; hollow triangles are from Torres et al (2019), adjusted to pH 7.2 per the abiotic rate equations in Rimstidt et al (2012). At 50 μM DFOB addition, our abiotic rate is reasonably characterized by, although somewhat lower than, the abiotic adsorption isotherm, given differences in experimental conditions. In contrast, the MR-1 and ∆MR-1 rates are significantly above the isotherm. 32 Siderophore adsorption to mineral surfaces is thought to be limited by both steric hindrance and charge repulsion (Cocozza et al., 2002). Thus, one possible mechanism by which microbes may increase siderophore efficiency is through the secretion of surfactants or other charged molecules that facilitate adsorption. Carrasco et al. (2007) showed that DFOB adsorption onto goethite increased by up to 3-fold in the presence of the surfactant sodium dodecyl sulfate. Given the linear dependence of dissolution on adsorbed siderophores, such an increase would lead to an equal enhancement of dissolution rates. Stewart et al (2013) similarly showed that fulvic acid, which contains carboxylate ligands, increased goethite dissolution by reducing the mineral’s positive surface charge and allowing higher concentrations of cationic siderophore adsorption. In our experiment, the 8-fold biotic vs abiotic enhancement at 50 μM DFOB addition could potentially be explained by microbial surfactants or negatively charged ligands, in combination with siderophore recycling. Smaller organic ligands like oxalate have also been shown to interact synergistically with siderophores to enhance the dissolution rates of goethite and Fe-silicate glass (Buss et al., 2007; Cervini-Silva & Sposito, 2002; Cheah et al., 2003; Reichard et al., 2007). These ligands do not, however, increase siderophore adsorption (Cheah et al., 2003; Reichard et al., 2007). Rather they appear to increase dissolution with siderophores via a two-step Fe labilization process: first, the ligands bind with surface Fe, creating a labile pool of Fe at the mineral surface. Given their smaller size (e.g., 56 Da for oxalate, vs 562 Da for DFOB), ligands are more effective at binding to surface Fe, if not fully extracting it from the mineral surface (Cheah, 2003). Subsequently, this labile Fe pool is liberated only when the saturation state of the surrounding solution is lowered by siderophore-mediated chelation of dissolved Fe (Cheah et al., 2003; Reichard et al., 2007). Olivine, however, is highly undersaturated in most natural solutions, and saturation state does not 33 significantly affect dissolution rates (Oelkers et al., 2018); thus, this mechanism does not initially seem pertinent. However, if accumulation of Fe 3+ on the olivine surface inhibits dissolution, as initially proposed by Schott and Berner (1983), ligand-siderophore synergy could operate similarly. In this case, ligands may labilize Fe 3+ in the passivating surface layer, and siderophores subsequently promote its dissolution by lowering the solution’s saturation state with respect to relevant Fe 3+ -oxide minerals (Gerrits et al., 2020; Torres et al., 2019; Van Den Berghe et al., 2021). These ligands are typically present at much higher concentrations than siderophores in environmental solutions, highlighting their high production rates by microbes, and potential relevance to siderophore-mediated mineral dissolution (Kalinowski et al., 2000). Of course, several biotic mechanisms may conspire to collectively enhance dissolution rates, especially when considering the scale of the microbe-mineral interface. Microbes do not just passively facilitate reactions in solution; rather, they concentrate biochemical processes and products at the mineral surface, especially within extracellular biofilms (Aouad et al., 2006; Buss et al., 2007; Gerrits et al., 2020; Liermann et al., 2000). The active concentration of dissolution- enhancing agents (e.g., protons, ligands, and siderophores) in combination with microbial recycling would logically lead to faster effective reaction kinetics than would lower concentrations measured in bulk solution. Indeed, microscopic evidence has shown that siderophore-producing bacteria produce larger and more clustered etch pits on Fe-silicate glass, resulting in faster dissolution, than observed from purified siderophores alone (Buss et al., 2007). As a related consideration, bacteria may also selectively target high energy, more easily dissolvable sites on the mineral surface (Hutchens, 2009; Lüttge & Conrad, 2004; Oelkers et al., 2015). While this selective colonization may slow dissolution via organic passivation in the long run, it could potentially fuel faster dissolution in the short term. 34 Biotically-enhanced dissolution persists with DFOB addition, but slows with wild type When comparing the dissolution trends of MR-1 wild type and ∆MR-1 +DFOB(50 μM), three key differences are noted. First, dissolution rates are similar, despite ∆MR-1 achieving only 50% of the wild type cell density. Second, dissolution in the wild type treatment slows after stationary phase is reached, whereas dissolution rates are constant in the ∆MR-1 treatments. And third, dissolved Fe begins to accumulate in the ∆MR-1 treatments during stationary phase, whereas dissolved Fe is completely absent in wild type (Fig. 2.3B). The fact that dissolution slows in MR- 1 wild type can be explained by downregulation of siderophore production in response to Fe availability -- a well-established bio-feedback mechanism in siderophore producers (Page & Huyer, 1984; A. Perez et al., 2016). This downregulation ensures that siderophores are used efficiently, and—as evidenced by the lack of dissolved Fe—little are lost to solution. In contrast, the unabated dissolution and accumulation of dissolved Fe in ∆MR-1 + DFOB indicate that exogenous siderophores stimulate dissolution in excess of bacterial nutritional needs. Because dissolution continues at rates much higher than the abiotic DFOB rate, it can be inferred that whatever mechanism is responsible for siderophore “magnification” (e.g., other ligands) is not limited by the same processes that down-regulate siderophore production upon reaching stationary phase. Note that by the end of the experiment, dissolved Fe in the ∆MR-1 +DFOB(50 μM) treatment is only 3 μM. If dissolution continues to be unrestrained by other mechanisms, it could be assumed that enhanced dissolution would continue until siderophores are saturated with Fe, i.e., up to nearly 50 μM. Indeed, Reichard et al. (2007) showed that adding DFOB to solutions with goethite and oxalate resulted in dissolved Fe concentrations up to the total DFOB added. The fact that dissolved Fe does not accumulate in the MR-1 wild type treatment suggests that it uses siderophores extremely efficiently, shunting all chelated Fe into biomass. Considering 35 the stoichiometry of total metal release, Fe recovery in MR-1 was only 60% of expected at the end of the experiment, likely indicating loss to secondary mineral precipitation as siderophore production slowed (Fig. 2.4B). In contrast, the ∆MR-1 + DFOB treatments had full Fe recovery by the end of the experiment. The reduced Fe recovery and lack of dissolved Fe in MR-1 wild type both suggest that siderophore production is low and/or localized at the microbe-mineral interface, with little siderophore loss to solution. Similar studies with Psuedomonas have shown that siderophores can indeed be concentrated in biofilm given sufficient mineral substrate; however even in these cases, siderophores were simultaneously found in solution, exceeding 10 μM (Dehner et al., 2010; Ferret et al., 2014). Furthermore, a conundrum persists in that our previous research confirmed the presence of siderophores in MR-1 solution (using chrome-azurol S), and ∆MR-1 fed filtrate from MR-1 exhibited normal growth patterns (Van Den Berghe et al., 2021). It is possible that extremely low concentrations of Shewanella’s native siderophore putrebactin are sufficient to maintain the bacteria’s growth and mineral dissolution needs, and that structural differences between putrebactin (a cyclic dihydroxamate) and DFOB (a tris-hydroxamate) allow for more efficient use in the former. The fact that MR-1 grows to stationary phase cell densities double those of ∆MR-1, yet dissolution rates are lower, further suggests that MR-1 uses available Fe more efficiently with putrebactin. 2.5 Conclusions and further considerations In short incubations, S. oneidensis increased olivine dissolution rates by an order of magnitude above abiotic rates. This enhancement was not observed with the non-siderophore producing mutant ∆MR-1 unless significant exogenous DFOB (>0.2 μM) was added, illustrating 36 that siderophores are required for biotically-enhanced dissolution. Furthermore, biotic dissolution was 8-fold faster than that stimulated by DFOB alone in abiotic conditions, suggesting that Shewanella uses additional mechanisms to magnify the effect of siderophores. In other words, Shewanella does not use siderophores passively to enhance dissolution, but somehow increases their efficiency to access mineral-bound Fe. This biotically-enhanced dissolution persisted throughout the experiment in experiments with exogenous DFOB addition, but declined shortly after reaching stationary phase in the wild type experiment with no DFOB addition. Thus, in the environment, enhanced dissolution may occur during non-stationary conditions, e.g., when fresh mineral faces are exposed and/or microbial populations are growing exponentially following environmental perturbation (Reichard et al., 2007). Although dissolution slowed in the wild type experiment, high dissolution rates persisted in the ∆MR-1 treatments wherein dissolved Fe accumulated in excess of microbial needs. Thus, the addition of exogenous siderophores may help maintain high dissolution rates. These findings may be useful for engineering biotically-mediated enhanced weathering systems, a nascent but potentially useful approach to enhanced mineral weathering. If dissolution rates can be maintained at 10x above abiotic rates, with minimal energy inputs, enhanced mineral weathering may be a promising avenue for effective carbon dioxide removal. 2.6 Acknowledgements Thanks to Will Berelson for helpful comments during manuscript preparation. This research was funded by the National Science Foundation, Division of Earth Sciences (EAR-1324929) and by the University of Southern California Dornsife College of Letters, Arts, and Sciences. 37 2.7 References Aouad, G., Crovisier, J.-L., Geoffroy, V. A., Meyer, J.-M., & Stille, P. (2006). Microbially- mediated glass dissolution and sorption of metals by Pseudomonas aeruginosa cells and biofilm. Journal of Hazardous Materials, 136(3), 889–895. Berner, R. A. (2010). Comment: effect of organic ligands and heterotrophic bacteria on Wollastonite dissolution kinetics. American Journal of Science, 310(5), 424. Berner, R. A., Lasaga, A. C., & Garrels, R. M. (1983). Carbonate-silicate geochemical cycle and its effect on atmospheric carbon dioxide over the past 100 million years. Am. J. Sci.;(United States), 283(7). Brady, P. V, Dorn, R. I., Brazel, A. J., Clark, J., Moore, R. B., & Glidewell, T. (1999). Direct measurement of the combined effects of lichen, rainfall, and temperature onsilicate weathering. Geochimica et Cosmochimica Acta, 63(19–20), 3293–3300. Bray, A. W., Oelkers, E. H., Bonneville, S., Wolff-Boenisch, D., Potts, N. J., Fones, G., & Benning, L. G. (2015). The effect of pH, grain size, and organic ligands on biotite weathering rates. Geochimica et Cosmochimica Acta, 164, 127–145. Buss, H. L., Lüttge, A., & Brantley, S. L. (2007). Etch pit formation on iron silicate surfaces during siderophore-promoted dissolution. Chemical Geology, 240(3–4), 326–342. Carrasco, N., Kretzschmar, R., Pesch, M.-L., & Kraemer, S. M. (2007). Low concentrations of surfactants enhance siderophore-promoted dissolution of goethite. Environmental Science & Technology, 41(10), 3633–3638. Cervini-Silva, J., & Sposito, G. (2002). Steady-state dissolution kinetics of aluminum-goethite in the presence of desferrioxamine-B and oxalate ligands. Environmental Science & Technology, 36(3), 337–342. Cheah, S.-F., Kraemer, S. M., Cervini-Silva, J., & Sposito, G. (2003). Steady-state dissolution kinetics of goethite in the presence of desferrioxamine B and oxalate ligands: implications for the microbial acquisition of iron. Chemical Geology, 198(1–2), 63–75. Cocozza, C., Tsao, C. C. G., Cheah, S.-F., Kraemer, S. M., Raymond, K. N., Miano, T. M., & Sposito, G. (2002). Temperature dependence of goethite dissolution promoted by trihydroxamate siderophores. Geochimica et Cosmochimica Acta, 66(3), 431–438. Dehner, C. A., Awaya, J. D., Maurice, P. A., & DuBois, J. L. (2010). Roles of siderophores, oxalate, and ascorbate in mobilization of iron from hematite by the aerobic bacterium Pseudomonas mendocina. Applied and Environmental Microbiology, 76(7), 2041–2048. Dessert, C., Dupré, B., Gaillardet, J., François, L. M., & Allegre, C. J. (2003). Basalt weathering laws and the impact of basalt weathering on the global carbon cycle. Chemical Geology, 202(3–4), 257–273. Dhungana, S., & Crumbliss, A. L. (2005). Coordination chemistry and redox processes in siderophore-mediated iron transport. Geomicrobiology Journal, 22(3–4), 87–98. 38 Dorn, R. I. (2014). Ants as a powerful biotic agent of olivine and plagioclase dissolution. Geology, 42(9), 771–774. Fennessey, C. M., Jones, M. E., Taillefert, M., & DiChristina, T. J. (2010). Siderophores are not involved in Fe (III) solubilization during anaerobic Fe (III) respiration by Shewanella oneidensis MR-1. Applied and Environmental Microbiology, 76(8), 2425–2432. Ferret, C., Sterckeman, T., Cornu, J., Gangloff, S., Schalk, I. J., & Geoffroy, V. A. (2014). Siderophore‐promoted dissolution of smectite by fluorescent P seudomonas. Environmental Microbiology Reports, 6(5), 459–467. Friedlingstein, P., O’Sullivan, M., Jones, M. W., Andrew, R. M., Hauck, J., Olsen, A., Peters, G. P., Peters, W., Pongratz, J., & Sitch, S. (2020). Global carbon budget 2020. Earth System Science Data, 12(4), 3269–3340. Garcia, B., Lemelle, L., Rose-Koga, E., Perriat, P., Basset, R., Gillet, P., & Albarède, F. (2013). An experimental model approach of biologically-assisted silicate dissolution with olivine and Escherichia coli–Impact on chemical weathering of mafic rocks and atmospheric CO2 drawdown. Applied Geochemistry, 31, 216–227. Gerrits, R., Pokharel, R., Breitenbach, R., Radnik, J., Feldmann, I., Schuessler, J. A., Von Blanckenburg, F., Gorbushina, A. A., & Schott, J. (2020). How the rock-inhabiting fungus K. petricola A95 enhances olivine dissolution through attachment. Geochimica et Cosmochimica Acta, 282, 76–97. Gorbushina, A. A., & Palinska, K. A. (1999). Biodeteriorative processes on glass: experimental proof of the role of fungi and cyanobacteria. Aerobiologia, 15(3), 183–192. Gorby, Y. A., Yanina, S., McLean, J. S., Rosso, K. M., Moyles, D., Dohnalkova, A., Beveridge, T. J., Chang, I. S., Kim, B. H., & Kim, K. S. (2006). Electrically conductive bacterial nanowires produced by Shewanella oneidensis strain MR-1 and other microorganisms. Proceedings of the National Academy of Sciences, 103(30), 11358–11363. Haack, E. A., Johnston, C. T., & Maurice, P. A. (2008). Mechanisms of siderophore sorption to smectite and siderophore-enhanced release of structural Fe3+. Geochimica et Cosmochimica Acta, 72(14), 3381–3397. Hartmann, J., Jansen, N., Dürr, H. H., Kempe, S., & Köhler, P. (2009). Global CO2-consumption by chemical weathering: What is the contribution of highly active weathering regions? Global and Planetary Change, 69(4), 185–194. Hartmann, J., West, A. J., Renforth, P., Köhler, P., De La Rocha, C. L., Wolf‐Gladrow, D. A., Dürr, H. H., & Scheffran, J. (2013). Enhanced chemical weathering as a geoengineering strategy to reduce atmospheric carbon dioxide, supply nutrients, and mitigate ocean acidification. Reviews of Geophysics, 51(2), 113–149. Hausrath, E. M., Neaman, A., & Brantley, S. L. (2009). Elemental release rates from dissolving basalt and granite with and without organic ligands. American Journal of Science, 309(8), 633–660. 39 Hersman, L. E., Huang, A., Maurice, P. A., & Forsythe, J. H. (2000). Siderophore production and iron reduction by Pseudomonas mendocina in response to iron deprivation. Geomicrobiology Journal, 17(4), 261–273. Hider, R. C., & Kong, X. (2010). Chemistry and biology of siderophores. Natural Product Reports, 27(5), 637–657. Hofmann, M., Retamal-Morales, G., & Tischler, D. (2020). Metal binding ability of microbial natural metal chelators and potential applications. Natural Product Reports, 37(9), 1262– 1283. Hutchens, E. (2009). Microbial selectivity on mineral surfaces: possible implications for weathering processes. Fungal Biology Reviews, 23(4), 115–121. Kalinowski, B. E., Liermann, L. J., Givens, S., & Brantley, S. L. (2000). Rates of bacteria- promoted solubilization of Fe from minerals: a review of problems and approaches. Chemical Geology, 169(3–4), 357–370. Köhler, P., Abrams, J. F., Völker, C., Hauck, J., & Wolf-Gladrow, D. A. (2013). Geoengineering impact of open ocean dissolution of olivine on atmospheric CO2, surface ocean pH and marine biology. Environmental Research Letters, 8(1), 14009. Kraemer, S. M. (2004). Iron oxide dissolution and solubility in the presence of siderophores. Aquatic Sciences, 66(1), 3–18. Kuhn, K. M., DuBois, J. L., & Maurice, P. A. (2013). Strategies of aerobic microbial Fe acquisition from Fe-bearing montmorillonite clay. Geochimica et Cosmochimica Acta, 117, 191–202. Lamérand, C., Shirokova, L. S., Bénézeth, P., Rols, J.-L., & Pokrovsky, O. S. (2020). Olivine dissolution and hydrous Mg carbonate and silicate precipitation in the presence of microbial consortium of photo-autotrophic and heterotrophic bacteria. Geochimica et Cosmochimica Acta, 268, 123–141. Liermann, L. J., Kalinowski, B. E., Brantley, S. L., & Ferry, J. G. (2000). Role of bacterial siderophores in dissolution of hornblende. Geochimica et Cosmochimica Acta, 64(4), 587– 602. Liu, L., Li, S., Wang, S., Dong, Z., & Gao, H. (2018). Complex iron uptake by the putrebactin- mediated and Feo systems in Shewanella oneidensis. Applied and Environmental Microbiology, 84(20), e01752-18. Liu, Y., Olsen, A. A., & Rimstidt, J. D. (2006). Mechanism for the dissolution of olivine series minerals in acidic solutions. American Mineralogist, 91(2–3), 455–458. Lüttge, A., & Conrad, P. G. (2004). Direct observation of microbial inhibition of calcite dissolution. Applied and Environmental Microbiology, 70(3), 1627–1632. Meysman, F. J. R., & Montserrat, F. (2017). Negative CO2 emissions via enhanced silicate weathering in coastal environments. Biology Letters, 13(4), 20160905. 40 Montserrat, F., Renforth, P., Hartmann, J., Leermakers, M., Knops, P., & Meysman, F. J. R. (2017). Olivine dissolution in seawater: implications for CO2 sequestration through enhanced weathering in coastal environments. Environmental Science & Technology, 51(7), 3960– 3972. Neaman, A., Chorover, J., & Brantley, S. L. (2005). Implications of the evolution of organic acid moieties for basalt weathering over geological time. American Journal of Science, 305(2), 147–185. Oelkers, E. H., Benning, L. G., Lutz, S., Mavromatis, V., Pearce, C. R., & Plümper, O. (2015). The efficient long-term inhibition of forsterite dissolution by common soil bacteria and fungi at Earth surface conditions. Geochimica et Cosmochimica Acta, 168, 222–235. Oelkers, E. H., Declercq, J., Saldi, G. D., Gislason, S. R., & Schott, J. (2018). Olivine dissolution rates: A critical review. Chemical Geology, 500, 1–19. Olsen, A. A., & Rimstidt, J. D. (2008). Oxalate-promoted forsterite dissolution at low pH. Geochimica et Cosmochimica Acta, 72(7), 1758–1766. Page, W., & Huyer, M. (1984). Derepression of the Azotobacter vinelandii siderophore system, using iron-containing minerals to limit iron repletion. Journal of Bacteriology, 158(2), 496– 502. Perez, A., Rossano, S., Trcera, N., Huguenot, D., Fourdrin, C., Verney-Carron, A., van Hullebusch, E. D., & Guyot, F. (2016). Bioalteration of synthetic Fe (III)-, Fe (II)-bearing basaltic glasses and Fe-free glass in the presence of the heterotrophic bacteria strain Pseudomonas aeruginosa: Impact of siderophores. Geochimica et Cosmochimica Acta, 188, 147–162. Pokharel, R., Gerrits, R., Schuessler, J. A., & von Blanckenburg, F. (2019). Mechanisms of olivine dissolution by rock-inhabiting fungi explored using magnesium stable isotopes. Chemical Geology, 525, 18–27. Pokrovsky, O. S., Shirokova, L. S., Bénézeth, P., Schott, J., & Golubev, S. V. (2009). Effect of organic ligands and heterotrophic bacteria on wollastonite dissolution kinetics. American Journal of Science, 309(8), 731–772. Reichard, P. U., Kretzschmar, R., & Kraemer, S. M. (2007). Dissolution mechanisms of goethite in the presence of siderophores and organic acids. Geochimica et Cosmochimica Acta, 71(23), 5635–5650. Rimstidt, J. D., Brantley, S. L., & Olsen, A. A. (2012). Systematic review of forsterite dissolution rate data. Geochimica et Cosmochimica Acta, 99, 159–178. Rosenberg, D. R., & Maurice, P. A. (2003). Siderophore adsorption to and dissolution of kaolinite at pH 3 to 7 and 22° C. Geochimica et Cosmochimica Acta, 67(2), 223–229. Santelli, C. M., Welch, S. A., Westrich, H. R., & Banfield, J. F. (2001). The effect of Fe-oxidizing bacteria on Fe-silicate mineral dissolution. Chemical Geology, 180(1–4), 99–115. 41 Schott, J., & Berner, R. A. (1983). X-ray photoelectron studies of the mechanism of iron silicate dissolution during weathering. Geochimica et Cosmochimica Acta, 47(12), 2233–2240. Schwartzman, D. W., & Volk, T. (1989). Biotic enhancement of weathering and the habitability of Earth. Nature, 340(6233), 457–460. Shirokova, L. S., Bénézeth, P., Pokrovsky, O. S., Gerard, E., Ménez, B., & Alfredsson, H. (2012). Effect of the heterotrophic bacterium Pseudomonas reactans on olivine dissolution kinetics and implications for CO2 storage in basalts. Geochimica et Cosmochimica Acta, 80, 30–50. Shirvani, M., & Nourbakhsh, F. (2010). Desferrioxamine-B adsorption to and iron dissolution from palygorskite and sepiolite. Applied Clay Science, 48(3), 393–397. Stewart, A. G., Hudson-Edwards, K. A., & Dubbin, W. E. (2013). Mechanisms of goethite dissolution in the presence of desferrioxamine B and Suwannee River fulvic acid at pH 6.5. Geochimica et Cosmochimica Acta, 115, 1–14. Stockmann, G. J., Shirokova, L. S., Pokrovsky, O. S., Bénézeth, P., Bovet, N., Gislason, S. R., & Oelkers, E. H. (2012). Does the presence of heterotrophic bacterium Pseudomonas reactans affect basaltic glass dissolution rates? Chemical Geology, 296, 1–18. Torres, M. A., Dong, S., Nealson, K. H., & West, A. J. (2019). The kinetics of siderophore‐ mediated olivine dissolution. Geobiology, 17(4), 401–416. Van Den Berghe, M., Merino, N., Nealson, K. H., & West, A. J. (2021). Silicate minerals as a direct source of limiting nutrients: Siderophore synthesis and uptake promote ferric iron bioavailability from olivine and microbial growth. Geobiology, 19(6), 618–630. Walker, J. C. G., Hays, P. B., & Kasting, J. F. (1981). A negative feedback mechanism for the long‐term stabilization of Earth’s surface temperature. Journal of Geophysical Research: Oceans, 86(C10), 9776–9782. Wild, B., Imfeld, G., & Daval, D. (2021). Direct measurement of fungal contribution to silicate weathering rates in soil. Geology, 49(9), 1055–1058. Wild, B., Imfeld, G., Guyot, F., & Daval, D. (2018). Early stages of bacterial community adaptation to silicate aging. Geology, 46(6), 555–558. Wogelius, R. A., & Walther, J. V. (1991). Olivine dissolution at 25 C: Effects of pH, CO2, and organic acids. Geochimica et Cosmochimica Acta, 55(4), 943–954. 42 2.8 Supplemental material Figure 2.S1. Dissolved and total Ni concentrations versus time for the subset of experiments analyzed by ICP-MS. Solid lines represent total (unfiltered) elemental release, and dashed lines represent the dissolved (filtered) component. Only the ∆MR-1 +DFOB(50 μM) had a statistically significant Ni release rate. All experiments performed in triplicate (error bars are standard deviation). 43 Chapter 3: CaCO 3 dissolution in carbonate-poor shelf sands increases with ocean acidification and porewater residence time Contributors: Will Berelson Abstract Carbonate-poor sandy sediments comprise much of the shelf area, and—despite their low CaCO3 content—contain a significant pool of CaCO3 base available to neutralize ocean acid. Here, we conducted flow-through column experiments on permeable, carbonate-poor sand obtained from Catalina Island, CA, to quantify CaCO3 dissolution across a range of current and future seawater conditions. Using 13 C isotope mass balance, we show that dissolution depends both on the CaCO3 saturation state (Ω) of the inflowing seawater, as well as porewater residence time. At current ocean conditions (Ωaragonite =2.4 and Ωcalcite =3.7 at our field site), dissolution was negligible for porewater residence times <1.8 h, but increased thereafter, following sufficient production of CO2 from aerobic respiration. As Ω of inlet water was lowered, simulating future ocean conditions, dissolution began earlier and rates increased. The response to acidification was similar to previously reported observations in carbonate-rich shelf environments, suggesting that carbonate- poor sediments have the potential to support enhanced dissolution in an acidifying ocean, given sufficient CaCO3 substrate. With continued acidification projected to occur this century, these sediments could transition from a net source of acid to the overlying seawater (production of alkalinity to dissolved inorganic carbon, ΔAlk/ΔDIC<1) to net source of buffering capacity 44 (ΔAlk/ΔDIC>1) when overlying seawater Ωaragonite reaches 0.96 to 0.69 (Ωcalcite = 1.50 and 1.07), depending on porewater residence time. In some areas with naturally acidic water, this threshold has already been reached. 3.1. Introduction The ocean has absorbed approximately one quarter of anthropogenic carbon dioxide (CO2) emissions since preindustrial times, causing a decrease in global average seawater pH by 0.1 unit (Caldeira & Wickett, 2005; Friedlingstein et al., 2020). As CO2 emissions continue, average global surface ocean pH is expected to fall by up to 0.5 unit by 2100 (Caldeira & Wickett, 2005). In a more acidic ocean, the dissolved inorganic carbon (DIC) pool is thermodynamically shifted away from carbonate ions (CO3 2- ), and the stability of calcium carbonate (CaCO3) minerals decreases. The resulting dissolution of CaCO3 releases alkalinity and consumes CO2, constituting an important negative feedback on atmospheric CO2 levels (as reviewed in Zeebe, 2012). While full sequestration of anthropogenic CO2 is thought to require silicate weathering on the timescale of millions of years, CaCO3 dissolution is relatively more important on shorter timescales (<10,000 y). In deep sea sediments, increased CaCO3 dissolution in response to rising CO2—termed “carbonate compensation”—depends on ocean circulation delivering high-CO2 surface water to depth, and thus occurs on the >1,000-y scale (Broecker & Peng, 1987). We are beginning to see signs of this process in the deep North Atlantic (F. F. Perez et al., 2018; Sulpis et al., 2018). Sediments on the continental shelves are not typically considered in the discussion of carbonate compensation, as they are bathed in surface seawater with higher CaCO3 saturation states (Zeebe & Westbroek, 2003). However, in theory, these sediments may respond quickly to 45 acidifying surface seawater due to several factors, including: high concentrations of more soluble carbonate forms (aragonite and high-Mg calcite), high rates of organic matter (OM) deposition and respiration, which lower porewater pH; naturally acidic water from upwelling and/or terrestrial nutrient input; and, in the case of coarse-grained sediments, enhanced mineral-seawater interaction via porewater advection. Despite this potential, we currently have little understanding of how CaCO3 in sandy sediments on the continental shelf respond to ocean acidification. The continental shelf comprises just 7% of global ocean area (Emery, 1968), but is disproportionately important in ocean carbon cycling. For example, 80% of marine organic matter (OM) burial and 90% of benthic respiration are estimated to occur in shelf sediments (Gattuso et al., 1998). Rapid mixing with deeper waters can then effectively shuttle respired CO2 to depth, making the shelf area a more efficient CO2 sink than surface waters of the open ocean (Cai, 2011; Thomas et al., 2004; Tsunogai et al., 1999). Furthermore, it is estimated that half of oceanic CaCO3 is buried on the shelves, while the remaining half is distributed across the slopes and vast deep ocean (Iglesias‐Rodriguez et al., 2002; Milliman, 1993). Despite this, CaCO3 dynamics in shelf sediments remain poorly understood, and little has changed since Milliman and Droxler’s (1996) assessment that “Of all the carbonate environments, we probably know the least about carbonate production, export, and accumulation on continental shelves.” In studies focused on CaCO3 cycling, the global shelf area is often divided into four environmental categories, including coral reefs, banks/bays, carbonate shelves, and carbonate- poor shelves. The latter two groups are easily divided as sediments are bimodal in terms of either high- or low-carbonate content (Milliman, 1993). While a comprehensive quantification of global shelf CaCO3 content is beyond the scope of this study, it is clear from existing data that carbonate- 46 poor sediments, composed primarily of detrital or relict silicates with minor amounts of biogenic CaCO3, make up a significant fraction—by some estimates more than half—of the global shelf area (Jenkins, 2021; Milliman, 1993). Despite their large extent, they are not commonly studied for CaCO3 dynamics, presumably due to their low CaCO3 content (<1%) (Krumins et al., 2013). As a result, recent estimates of global CaCO3 accumulation on the carbonate-poor shelf vary by two orders of magnitude (Iglesias‐Rodriguez et al., 2002; O’Mara & Dunne, 2019). In contrast, there are many measurements of CaCO3 dissolution in shallow-water, coral reef sands (Andreas J. Andersson, 2015; Andreas J Andersson et al., 2007; Andreas J Andersson & Gledhill, 2013; Boucher et al., 1998; Cyronak, Santos, & Eyre, 2013; Cyronak, Santos, McMahon, et al., 2013; Cyronak & Eyre, 2016; Eyre et al., 2014, 2018; Fink et al., 2017; Rao et al., 2012; Stoltenberg et al., 2020, 2021; Trnovsky et al., 2016; K K Yates & Halley, 2003; Kimberly K Yates & Halley, 2006). In addition to ecosystem type, shelf sediments can be further divided into fine-grained sediments, in which dissolved components in porewater move primarily through diffusion, and coarse-grained sediments (average grain size equal to or larger than very fine sand, 62.5 μm), in which porewater movement is driven by advective transport. In the latter, solute transport can exceed diffusive flows by several orders of magnitude (Precht & Huettel, 2004). Approximately half of the continental shelf is thought to be covered by coarse-grained sediments (Hall, 2002; Hayes, 1967). Porewater flow through these sands occurs on globally relevant scales, with the entire ocean volume filtering through sediments every 3,000 to 14,000 years (Moore et al., 2008; Riedl et al., 1972; Santos et al., 2012). This largescale movement of seawater acts as a “subtidal pump”, stimulating biogeochemical processes via the delivery of oxygen (O2) and OM to particle- attached microbes, fluctuating redox conditions, and removal of diagenetic byproducts (Riedl et 47 al., 1972; Rusch et al., 2006; Shum & Sundby, 1996). As a result, benthic CaCO3 dissolution can be higher under advective versus diffusive porewater conditions (Cyronak, Santos, & Eyre, 2013; Cyronak, Santos, McMahon, et al., 2013; Rao et al., 2012). Unfortunately, complex advection dynamics in permeable sands also makes quantifying processes and fluxes difficult; techniques commonly used in diffusion-dominated sediments (e.g., porewater profiles, core incubations, etc.) are not useful because they do not consider nor accurately reproduce in situ advection (Boudreau et al., 2001; Huettel et al., 2014). Instead, alternative techniques such as advective chambers, flumes, and flow-through columns are used in permeable sediments. Because these approaches are arguably more challenging to implement and/or interpret, measurements of CaCO3 dissolution in permeable sediments are less common, and to our knowledge there are no direct measurements in carbonate-poor sands. Furthermore, few dissolution studies in sands exist outside of the tropics, although colder waters are already less saturated with respect to CaCO3 minerals, and expected to be more sensitive to ocean acidification (Andreas J. Andersson et al., 2005). This study attempts to fill this gap by measuring CaCO3 dissolution in carbonate-poor sands collected from the California coast. We used flow-through columns with alkalinity, DIC, and C isotopic measurements to quantify CaCO3 dissolution rates across a range of CaCO3 saturation states predicted for the coming century. The flow-through column technique also allowed us to control porewater residence time and advection rate to assess the response to acidification in different advection conditions. In this way, we aim to improve our understanding of how sediments on the continental shelves respond to ongoing ocean acidification and other global changes (e.g., wave and current energy) that may affect porewater advection dynamics. 48 3.2. Methods 3.2.1 Experimental setup Flow-through sand column experiments, commonly referred to as flow-through reactor (FTR) experiments, were conducted in June-July 2018 and March-May 2019. Sand was collected from the low intertidal by surface grab samples from Fisherman’s Cove on Catalina Island, off the coast of Los Angeles, CA, USA (33.4450289 N, 118.4838903 W (WGS84)). This sand was considered a reasonable representation of coarse-grained, detrital shelf sediments. Although seabed properties on the shelf are extremely varied, a global, quantitative analysis of grain size, OM content, etc. is beyond the scope of this paper (Jenkins, 2021). Following collection, sand was sieved through a 2-mm sieve to remove rocks and macrofauna. In all cases, very little, if any, material was retained by the sieve (i.e., gravel fraction was negligible). Sand was subsequently packed into columns, or stored under running seawater in the laboratory prior to packing (within 24 hours of collection). Columns were constructed from 5.7 cm inner diameter by 30 cm tall polycarbonate tubes, with each end enclosed by a mesh screen and rubber bungs connected to inlet/outlet tubing (Fig. 3.1). Sample ports were drilled into the columns at two intermediate heights, thus providing four total sampling ports at 0 (inlet), 8.5 (bottom), 18.5 (middle), and 26 cm (top) sand height. Columns were placed in a temperature-controlled room set to 19° C (June-July 2018) or 16° C (July 2018, and March-May 2019). The two selected temperatures were not tested as explicit experimental treatments, but instead represent spring (16° C) or annual (19° C) mean water temperatures at the research site. 49 direction of flow Inlet seawater reservoir Peristaltic pump Figure 3.1. Schematic of flow-through sand column. Water is peristaltically pumped from a gas-impermeable bag reservoir upward through the column. Sample ports are located below the column (0 cm sand height), at 8.5 cm, 18.5 cm, and above the column (26 cm sand height). Packed columns were then connected via the lower (0 cm) port to approximately 50 cm of 1/8” ID tubing (Tygon F-4040-A) connected via a peristaltic pump to a 5 or 10 L gas-impermeable, foil bag (Supel-Inert multi-layer foil with screwcap valve). Gas-proof bags were used instead of open bottles to eliminate the loss of CO2 and thus maintain the carbonate saturation state (Ω) of the inlet seawater. Evacuated bags were pre-filled with 0.2 μm filtered seawater, also collected from Catalina Island. For some treatments, henceforth referred to as “acidified”, inlet Ω was lowered by injecting HCl into the bags prior to the experiment. In this case, the bags were shaken for several hours prior to connecting to the columns to ensure thorough mixing of the acid. It should be noted that HCl lowers Ω by reducing alkalinity at a constant DIC, whereas in reality, ocean acidification increases DIC, but maintains alkalinity. Methodological studies have found no 50 difference between these two manipulations on biological processes or CaCO3 dissolution at a given Ω (A J Andersson & Mackenzie, 2012). Water was peristaltically pumped from the bags, upwards through the columns at specified pump rates. Most experiments were run at 1 mL min -1 , but a subset was also run at 0.5 or 2 mL min -1 to test different advection rates and longer/shorter residence times. The 1 mL min -1 pump rate yielded porewater residence times of approximately 1.8, 3.6, and 5.4 h (for bottom, middle, and top ports, respectively), based on a measured porosity of 0.5. These nominal times are used throughout the discussion, but actual porewater residence times varied slightly for each column. Pump rates were verified by weighing outlet water after several hours, and adjusted as necessary to reach the target pump rate. By visual inspection during and after the experiment, there was no evidence of channeling or flow along the column sides. Breakthrough curve experiments using fluoride tracer also indicated no channeling, and the apparent dispersivity suggested an average variation in porewater residence time of approximately ±10% (Supplemental Material). Pump rates were selected to replicate realistic advection rates (also commonly referred to in the literature as “flushing rates”). Our advection rates ranged from 280 to 1130 L m -2 d -1 , similar to those measured in situ in medium-coarse sand by Reimers et al (2004) (300 to 1,300 L m -2 d -1 ). In situ porewater advection measurements are scarce, but several studies have estimated average flushing rates that are slower than those used in our study, on the order of 100 L m -2 d -1 (McGinnis et al., 2014; Precht & Huettel, 2004; Santos et al., 2012). However, higher rates, up to 8,000 L m - 2 d -1 , have been documented for high energy environments (McLachlan, 1989). A related, but distinct, metric for advection is porewater advective velocity. Our velocities—2.3 to 9.2 cm h -1 — were similar to those measured by Reimers et. al. for sediments deeper than 1 cm. Like advection 51 rate, porewater velocities also vary widely depending on sediment depth and hydrodynamic forcing, and can range from <1 to >300 cm h -1 , although most hydrodynamic processes result in velocities of <20 cm h -1 (Oberdorfer & Buddemeier, 1986; Precht & Huettel, 2004). Thus, our advection rates were selected to represent a range of realistic conditions. Pumping continued for at least 48 hours (equivalent to flushing the column volume >8 times) prior to initial sample collection in order to remove respiratory byproducts accumulated during column preparation. During this time, the top outlet water was sampled and analyzed regularly to ensure that alkalinity and DIC were stable prior to taking samples for data analysis. Final samples were collected approximately 48-72 hours after pumping began for 1 and 2 mL min - 1 experiments, and up to 146 hours after pumping began for the 0.5 mL min -1 experiments. Samples were collected using glass syringes attached to one outlet port at a time, filling passively by pump pressure. Top port samples were collected first, then the lower ports collected in sequence (e.g., middle, then bottom) to avoid sampling stagnant porewater in the upper column following sampling at lower levels. For each sample, at least 40 mL water was collected for DIC and δ 13 C (24 mL) and alkalinity (14 mL) analysis, as well as for O2 for a subset of the 1 mL min -1 columns. 3.2.2 Isotope spike experiments Two sets of column experiments were run with the addition of a small amount (50 mg) of either 13 C-labeled calcite (April 2019) or 13 C-labeled aragonite (May 2019) added to the sediment. These experiments served as a test to confirm the presence or absence of dissolution observed in the natural abundance experiments, and to constrain the Ω at which aragonite and calcite dissolution begins (Ωarag and Ωcalc, respectively). (Dissolution rates were not calculated in these 52 columns.) The 13 C-calcite (>99% 13 C) was purchased from Aldrich, and the 13 C-aragonite (100 ±5% 13 C) was grown in our lab using a gel-diffusion method (Dong et al., 2019; Nickl & Henisch, 1969). Each 13 C spike was sieved to grain size fractions similar to the sand (between 0.05 mm and 2 mm), weighed, then thoroughly mixed into a bucket of drained sand, with just enough sand volume to fill the necessary columns. The mass of labeled carbonate added was approximately 0.5% of the total PIC present in the natural sand. This low spike addition was used to minimally alter the CaCO3 content of the sand, yet provide a clear dissolution signal in δ 13 C. The amended sand had a ~50% increase in 13 F (the ratio of 13 C to total C content) compared to the natural sand, and the resulting δ 13 CPIC was 702‰ (±15) for the calcite experiment, and 457‰ (±164) for the aragonite experiment. For both the calcite and aragonite spike experiments, four columns were incubated: two sets of paired spike/non-spike columns, one pair which was exposed to natural seawater (Ωarag = 2.3 to 2.4), and the other pair which was exposed to acidified seawater (Ωarag = 1.2 to 1.3). These four columns were incubated simultaneously to a) confirm no significant treatment effect of the isotope spike on alkalinity or DIC production, and b) to calculate dissolution of the spike using the isotopic difference between the paired spiked and control (non-spiked) columns. 3.2.3 Sample analysis – liquid phase Dissolved O2 concentrations were measured immediately upon sample collection. Measurements were taken by microsensor electrode (A/V Unisense) in 2018, or by optode (NeoFox, Ocean Optics) in 2019, in both cases standardizing between air-saturated and anoxic (N2-purged) seawater. Error on O2 measurements was assumed to be 5%. 53 For alkalinity, DIC and δ 13 C, water samples were filtered (0.45 μm pore size), refrigerated and analyzed within one week of collection. Alkalinity samples were collected in 15 mL Falcon tubes, and measured by open-system Gran titration on a custom-built instrument consisting of an electrode connected to a pH meter (Metrohm Ecotrode and Mettler Toledo SevenCompact, respectively) and electronic titrator (Metrohm 876 Dosimat Plus). Weighed samples were equilibrated to 21 C, stirred, and aerated during titration with 0.05–0.1 M HCl, as controlled by a custom MATLAB script that adds acid in 0.01 mL increments and measures pH in 20 s intervals. After 12 time points, the alkalinity endpoint was determined using a nonlinear least-squares approach. A standard reference (Dickson seawater) was run at the beginning of each analytical run to ensure consistency. Long-term alkalinity precision (standard deviation) for this analysis is about 4 μeq kg -1 over several months. Aliquots for DIC and δ 13 C analysis were collected in duplicate by injecting 7 mL sample into pre-evacuated 12 mL exetainer vials (Labco). Samples were transferred directly from the glass syringe, through a filter, into the vials, thereby avoiding DIC loss from outgassing. Samples were then measured with a cavity ringdown spectrometer (CRDS; Picarro Model G-2131) coupled to a Picarro Liaison interface and a modified Automate autosampler. Approximately 3 mL of 10% H3PO4 was added to each sample, and the CO2 carried with N2 carrier gas through a Nafion desolvating line, to the CRDS. Standard references (OPT calcite calibrated against measurements at the UC Davis Stable Isotope Lab and Dickson seawater) were included in each analytical run. Both DIC and δ 13 C were normalized to multiple standards and corrected for blanks, water content, and machine drift, as appropriate. The precision (standard deviation) for this method was <12 μmol kg -1 for DIC and <0.15‰ for δ 13 C. 54 3.2.4 Sample analysis – solid phase Sand used in the incubations was analyzed for solid phase parameters, including grain size, porosity, percent total particulate C, and particulate inorganic C (TC and PIC, respectively), as well as the δ 13 C of TC and PIC. For a subset of incubations, percent particulate organic C (POC) and its associated δ 13 C were also measured. %TC and δ 13 CTC were analyzed on 50 mg sand aliquots which were dried, ground, and packed in 6x10 mm tin capsules, then combusted on a Costech ECS 4010 Elemental Analyzer, coupled to the CRDS. %PIC and δ 13 CPIC were analyzed by weighing 150 mg aliquots into 12 mL exetainer vials, which were subsequently evacuated, injected with 5 mL 10% H3PO4, vortexed to convert PIC to CO2 gas, then run through the CRDS. For %POC and δ 13 CPOC, dried and ground aliquots were weighed into 6x10 mm silver capsules, then acidified with H3PO4 until bubbling ceased. After drying, the silver capsules were packed into larger (9x10mm) tins, and analyzed on the Elemental Analyzer-CRDS system. For experiments where POC was not measured directly, %POC was calculated as the difference between %TC and %PIC, and δ 13 CPOC calculated from the PIC, TC, and OC isotope mass balance. This approach was confirmed to be reasonable by comparing measured and calculated POC and δ 13 CPOC values for experiments with both data available: OC differed by <0.01 %C and δ 13 COC by <0.5‰. Carbonate mineralogy was determined both by Raman spectroscopy and x-ray diffraction (XRD) at the Natural History Museum of Los Angeles County. Sand porosity was measured by mass difference between wet and dry samples (60 C oven until constant weight), assuming a mineral density of 2.6 g cm -3 and correcting for seawater salt content. The sand grain size distribution was measured by wet, then dry sifting with metal sieves ranging in mesh size from 2 mm to 62.5 μm. 55 3.2.5 Calculations Alkalinity and DIC data were used to calculate Ωarag and Ωcalc using either the CO2SYS Excel macro (v 2.05) or Matlab code (v 1.1) (Lewis & Wallace, 1998; van Heuven et al., 2011), with options set as: total pH scale, KHSO4 from Dickson (1990), and CO2 constants from Mehrbach et al (1973), refit by Dickson and Millero (1987). Uncertainties were calculated using analytical errors and default errors for thermodynamic constants defined in Orr et al. (2018). Production (or consumption) of DIC, alkalinity, and O2 was assessed as the difference between inlet and outlet concentrations at each port (for example for DIC): DICadded = DICout - DICin (1) Where DICout and DICin are the concentrations of DIC measured at each outlet or the inlet, respectively (µmol kg -1 ). Areal production (or consumption) rates were then calculated at each outlet port, on a time-normalized basis as (for example for DIC): Production rate = (DICadded) (1.025) (Q) / (A) (2) Where 1.025 kg L -1 considers the density of seawater; Q is the flow rate of the experiments (in L h -1 ); and A is the cross-sectional area of the sand column (0.0026 m 2 ). A similar procedure was followed to calculate alkalinity production (Alkadded) and O2 consumption (using Eq. 1) and their respective areal production rates (using Eq. 2). CaCO3 dissolution was calculated based on the isotope mass balance difference between inlet and outlet water, per the following equations: (DICadded)(δ 13 Cadded) = (DICout)(δ 13 Cout) - (DICin)(δ 13 Cin) (3) (DICadded)(δ 13 Cadded) = (DICPIC)(δ 13 CPIC) + (DICOM)(δ 13 COM) (4) 56 Where δ 13 Cadded is the mean value of seawater δ 13 C added between the inlet and outlet port, and DICPIC and DICOM refer to DIC production from PIC and OM, respectively. The only two unknowns—DICPIC and DICOM—can be solved for using Eq.s 1, 3, and 4, and the reasonable assumption that added DIC originates only from PIC and OM. For comparison and validation of the isotope budget approach, DICPIC was also calculated independently for a subset of columns (those for which O2 was measured) using the DIC budget, as DICadded minus DICOM derived from O2 measurements (Table 3.1). For this calculation, DICOM production was calculated using O2 consumption and a respiration quotient (-O2 consumed:DICOM respired) of 1.1 (Moreno et al., 2020). Errors for all derived values were calculated by propagation of average analytical errors. For the 13 C-spike experiments, CaCO3 dissolution was not explicitly calculated, as the kinetics of biogenic and abiotic CaCO3 dissolution can vary (Subhas et al., 2018). Rather, any increase in δ 13 C in the spiked column samples was taken as evidence of spike dissolution. This is presented hereafter as Δδ 13 C, the difference between the δ 13 C of the spike (δ 13 Cout_spike) and control (δ 13 Cout_control) column outlet water. This approach assumes identical δ 13 C for the inlet water of both columns, which was confirmed within analytical error. Because a small amount of spike dissolution results in a large increase in δ 13 C, the spiked experiments provide a highly sensitive method for confirming dissolution of the different CaCO3 crystal forms under different inlet Ω values. 57 3.3. Results 3.3.1 Sediment characterization Porosity averaged 0.51 (±0.02), and grain size was dominated by medium-coarse sand (grain size distribution: 19%, 36%, 38%, 7%, and 0% for >1 mm, 0.5 to 1 mm, 0.25 to 0.5 mm, 0.06 to 0.25 mm, <0.06 mm, respectively). Carbon content was consistently low across collection dates: TC averaged 0.25% (±0.01%), with PIC 0.10% (±0.01%) and POC 0.14% (±0.01%). δ 13 C values varied slightly across experimental dates, with δ 13 CTC ranging from -11.2 to -14.5‰; δ 13 CPIC from -0.9 to -4.1‰; and δ 13 COC from -19.5 to -20.1‰. Given the small PIC content, quantitative mineralogical analysis was not possible. Raman spectroscopy and XRD both confirmed the presence of calcite and aragonite, but the counts of each were too low for statistically significant quantification. 3.3.2 Consumption or production of alkalinity, DIC, and O 2 , and CaCO 3 dissolution Alkalinity was generated in all columns, and there was a clear pattern of higher cumulative alkalinity production at lower inlet Ω (Fig. 3.2A; Table 3.S1). In the high Ω (non-acidified) columns, alkalinity production was negligible prior to 1.8 h, but net positive by the end of the experiment at 5.4 h. In the acidified columns (inlet Ωarag <2) however, alkalinity production was positive over all sampling points. The timing of peak alkalinity production also varied by inlet Ω: columns with inlet Ωarag <1 typically had peak alkalinity production (as indicated by the slope of lines in Fig. 3.2A) between 0 and 1.8 h, whereas those with Ωarag >1 had peak production after 1.8 h. 58 Table 3.1. Measured Alkalinity, DIC and O 2 fluxes, and calculated DIC added from OM (DIC OM ) and PIC (DIC PIC ) from a representative sample of columns. DIC OM values were calculated using O 2 data and a respiration quotient of 1.1. DIC PIC values were calculated from two different methods— 13 C mass balance, or DIC added minus DIC OM —and are within analytical error for most columns. Approximate errors (standard deviation) are as shown for calculated values; 7 µmol kg -1 for measured Alk added ; 17 µmol kg -1 for measured DIC added ; and 12 µmol kg -1 for O 2 consumed. Inlet Ωarag Porewater residence time (h) Alkadded (µmol kg -1 ) DICadded (µmol kg -1 ) O2 consumed (µmol kg -1 ) DICOM (µmol kg -1 ) DICPIC from 13 C budget (µmol kg -1 ) DICPIC from DIC budget (µmol kg -1 ) 0.12 1.8 605 400 -120 133 (±10) 326 (±52) 268 (±53) 3.6 827 493 -160 176 (±10) 347 (±39) 317 (±40) 5.4 920 574 -170 187 (±10) 397 (±38) 387 (±40) 1.04 1.8 61 130 -110 121 (±9) 12 (±3) 10 (±10) 3.6 250 224 -152 167 (±9) 58 (±9) 58 (±13) 5.4 332 271 -151 166 (±9) 101 (±15) 105 (±17) 1.29 1.8 80 189 -148 162 (±14) 22 (±4) 27 (±14) 3.6 211 326 -225 248 (±13) 96 (±11) 78 (±17) 5.4 273 337 -236 259 (±13) 84 (±9) 78 (±16) 2.42 1.8 9 121 -132 146 (±10) -17 (±5) -25 (±11) 3.6 50 195 -152 167 (±10) 22 (±4) 28 (±10) 5.4 141 263 -156 171 (±10) 65 (±9) 91 (±13) 2.43 1.8 -24 138 -148 162 (±14) -33 (±7) -24 (±16) 3.6 25 231 -227 249 (±13) 6 (±1) -18 (±13) 5.4 73 249 -233 257 (±13) 3 (±0) -8 (±13) Total DIC production was not well correlated with inlet Ω, apart from generally lower production in the unacidified columns and very high DIC production in the most acidic column (inlet Ωarag = 0.11) (Fig. 3.2B; Table 3.S2). This reflected the combined variability in DICPIC (DIC from CaCO3 dissolution) at a given Ω (as in Fig. 3.2C) and DICOM between columns. Rates of DICOM production likely varied in part due to variable inlet O2; in the ten columns for which inlet O2 was measured, inlet O2 was typically >215 μmol kg -1 , although three had lower O2, between 161-175 μmol kg -1 , suggesting incomplete air equilibration prior to initiating some experiments (Table 3.S3). However, this variability in inlet O2 was small in comparison to the range of inlet 59 Ω’s tested, and did not have a systematic effect on CaCO3 dissolution with respect to inlet Ω. O2 consumption was highest during the first 1.8 h, after which time porewater became hypoxic (<60 μM). While O2 consumption rates slowed over time, O2 concentrations continued to decrease, with a small amount remaining after 3.6 h, and most columns reaching <5% saturation by 5.4 h. Figure 3.2. Alkalinity added (A), DIC added (B), CaCO 3 dissolution determined by 13 C isotope budget (C), and porewater Ω (D) with respect to porewater residence time. Each line represents a single column experiment at 1 ml min -1 advection rate, with line color indicating Ω aragonite of the inlet seawater. Analytical errors (SD) are 5.6 μmol m -2 h -1 for alkalinity added; 17 μmol m -2 h -1 for DIC added; and as shown for CaCO 3 dissolved and Ω. 0 2 4 6 0 200 400 600 800 1000 Alkalinity added ( mol kg -1 ) 0 2 4 6 0 100 200 300 400 Carbonate dissolved ( mol kg -1 ) Porewater residence time (h) 0 2 4 6 0 100 200 300 400 500 600 DIC added ( mol kg -1 ) 0.5 1 1.5 2 Inlet arag 0 2 4 6 Porewater residence time (h) 0 1 2 3 Porewater arag arag =1 calc =1 0.5 1 1.5 2 Inlet arag A C B D 60 Figure 3.3. CaCO 3 dissolution with respect to porewater residence time for three different advection rates. Dashed line: 2 ml min -1 ; solid line: 1 ml min -1 ; dotted line: 0.5 ml min -1 . CaCO3 dissolution calculated with the 13 C isotope budget (Eq.s 1, 3, and 4) and with the DIC budget (DICadded minus DICOM calculated from O2 consumption) were within analytical error for most measurements (Table 3.1). Henceforth, dissolution will refer to values calculated with the 13 C isotope budget unless otherwise noted. CaCO3 dissolution was observed in all columns, and, like alkalinity production, was inversely correlated with inlet Ω (Fig. 3.2C). Dissolution rates also varied over porewater residence time, with different patterns for low- and high-Ω columns: in the high-Ω (unacidified) columns, CaCO3 dissolution was negative (i.e., net precipitation) or near 0 at porewater residence times <1.8 h, but positive thereafter. In the acidified columns, however, CaCO3 dissolution was already positive by the first sampling point at 1.8 h. Most of the acidified columns had peak dissolution rates between 0-1.8 h or 1.8-3.6 h, and decreased thereafter, while some of the unacidified cores had increasing dissolution rates through the end of the experiment. 0 2 4 6 8 10 Porewater residence time (h) -40 0 40 80 120 160 200 240 Carbonate dissolved ( µmol kg -1 ) 0.6 0.8 1 1.2 1.4 1.6 1.8 Inlet Ω arag 61 Columns treated with different pumping rates had similar CaCO3 dissolution rates when normalized for porewater residence time, indicating no clear effect of advection rate across the range tested (Fig. 3.3). Porewater Ω changed over time in all columns (Fig. 3.2D). For columns with inlet water Ωarag >0.65 (Ωcalc >1), Ω initially fell before rising again after 1.8 h. Most of these columns fell to a minimum porewater Ω between calcite and aragonite saturation. For columns with inlet Ωarag <0.65, Ω steadily increased over the entire experiment. Regardless of initial Ω, all columns had stable or increasing Ω by the end of the experiment, with many columns reaching Ωarag > 1 by the last sampling point. Figure 3.4. CaCO 3 dissolution rates, normalized to sediment surface area, as a function of Ω inlet . Linear relationships are shown for all three residence times: r 2 = 0.711, 0.619, and 0.547; p<0.001, <0.001, and 0.002, for 1.8-, 3.6-, and 5.4-h porewater residence time, respectively. All data are from 1 ml min -1 advection rate experiments. (Figure and linear fit exclude data for the inlet Ω aragonite = 0.1.) 0 0.5 1 1.5 2 2.5 Inlet Ω aragonite -20 0 20 40 60 80 100 120 1.8-h 3.6-h 5.4-h Cumulative CaCO3 dissolution rate (mmol m - ² d -1 ) y 1.8h = -23.09x + 41.30 y 3.6h = -27.31x + 71.66 y 5.4h = -25.59x + 80.72 Inlet Ω calcite 0 1 2 3 4 62 The relationship between inlet seawater Ω and cumulative CaCO3 dissolution was approximated well by linear models for all three porewater residence times (1.8-, 3.6-, and 5.4-h, respectively; r 2 = 0.711, 0.619, and 0.547; p<0.001, <0.001, and 0.002; Fig. 3.4). All three regressions had similar slopes. Although mechanistic studies of dissolution kinetics have shown non-linear dissolution rates with respect to Ω for single CaCO3 morphotypes in undersaturated water, we assume our sand contains a mixture of morphotypes, and thus use a linear model as an approximation (Naviaux, Subhas, Rollins, et al., 2019; Subhas et al., 2017). Alkalinity and DIC production for the columns with 13 C-labeled CaCO3 were the same (within error) as the unspiked column for all sampling points, indicating that the small 13 C spike did not affect bulk CaCO3 dissolution or respiration rates. However, δ 13 C increased in the spiked columns relative to the controls (i.e., Δδ 13 C>0), indicating dissolution of the 13 C-spiked mineral (Fig. 3.5). In the high-Ω (non-acidified) seawater columns, Δδ 13 C was significantly, yet only slightly>0 (aragonite), or insignificant (calcite) at 1.8 h, indicating little or no dissolution of the spike under short residence time conditions (Fig. 3.5A, and 3.5B, respectively). However, after 1.8 h, both high-Ω experiments had Δδ 13 C>0, suggesting dissolution of both CaCO3 species. In the acidified columns, Δδ 13 C was positive by the first sampling point at 1.8 h for both mineral species, indicating a more rapid onset of dissolution than in the high-Ω conditions. 63 Figure 3.5. Evolution of δ 13 C for the 13 C-aragonite (A) and 13 C-calcite (B) spike experiments. The y axis (Δδ 13 C) represents the difference in outlet water δ 13 C of the paired 13 C-spiked and natural cores: a positive value indicates higher outlet δ 13 C in the spiked cores, resulting from dissolution of the spike Ca 13 CO 3 . Numbers next to each point indicate porewater Ω for the respective mineral species at each sampling point. “High-Ω” treatments use unaltered seawater, whereas “low-Ω” treatments are acidified to near-future conditions. Error (SD) are shown for Δδ 13 C (smaller than point size for aragonite), and approximately 0.1 for Ω. 3.4. Discussion 3.4.1 Methodological considerations Studies of CaCO3 dissolution in permeable sediments typically use the Alkalinity Anomaly Technique, which assumes that dissolution is the sole alkalinity source during an incubation (i.e., total alkalinity = CaCO3 dissolution * 2). This assumption has been shown to be reasonable for 64 carbonate-rich coral reef sands where anaerobic respiration is minimal (Chisholm & Gattuso, 1991; Rao et al., 2012). However, in our natural abundance (i.e., non- 13 C-spiked) experiments, comparing the dissolution rates calculated by 13 C isotope mass balance and total alkalinity production suggests that sources other than dissolution account for a variable but significant fraction of total alkalinity, averaging at least 20% (Fig. 3.6). The estimated global average for non-carbonate alkalinity on the carbonate-poor shelf is even higher, at 58% (Krumins et al., 2013). In these cases, the 13 C method is a better option than the Alkalinity Anomaly Technique, as it is not affected by non-carbonate alkalinity. Figure 3.6. Alkalinity from CaCO 3 dissolution (calculated from DIC and δ 13 C mass balance) versus total alkalinity production. The regression slope (dotted line) suggests average alkalinity produced by CaCO 3 dissolution accounts for on average 80% of the total. Other alkalinity sources include anaerobic OM respiration (see text). Solid line is 1:1, shown for reference. 65 Two additional considerations of the 13 C isotope budget method merit discussion. First, fermentation in temporarily anoxic sands can produce DIC with extremely positive δ 13 C (> +20 ‰) (Bourke et al., 2017; Kessler et al., 2019, 2020). Addition of such positive δ 13 C would increase δ 13 Cadded, overestimating CaCO3 dissolution in a two-endmember isotope mixing model (since CaCO3 is the more positive δ 13 C endmember, relative to OM). In our experiments, however, all data points except one—including the least oxic water between the top port and outlet—had δ 13 Cadded values between the OM and PIC endmembers. We thus assume that fermentation does not interfere significantly with our two-endmember analysis. The second potential issue with the 13 C method is CaCO3 precipitation. The two-endmember mixing model ignores precipitation (Eq. 4), because the system is underconstrained for a potential third endmember. Precipitation cannot simply be considered “negative dissolution” because the δ 13 C of porewater (which precipitates) may be distinct from that of PIC (which dissolves); however, because the δ 13 C of porewater (typically around +1‰) is more similar to that of PIC (-1 to -3‰) than OM (-20‰), precipitation is approximately equal to “negative dissolution”. Thus, calculated CaCO3 dissolution rates are approximately equal to net dissolution. These considerations, and the fact that CaCO3 dissolution calculated using two different approaches—by isotope mass balance and by the DIC budget)— lead us to conclude that our isotope mass balance-derived dissolution rates are reasonable estimates of net dissolution (Table 3.1). Other methodological considerations relate to how our experimental setup differs from in situ conditions, including: no photosynthetically active radiation (PAR; i.e., dark-only conditions), constant temperature, use of filtered water, repacked sand, and constant advection rate and temperature. In shallow environments, photosynthesis at the sediment surface can increase Ω and depress CaCO3 dissolution (Cyronak, Santos, McMahon, et al., 2013; Fink et al., 2017; Lantz, 66 Carpenter, et al., 2017; Rao et al., 2012; Kimberly K Yates & Halley, 2006). Our data should thus be considered representative of dissolution from deeper sites beyond the reach of PAR and/or night-time rates. Approximately two-thirds of the shelf area is below the effective photic zone, so dark rates are applicable to a significant area (Gattuso et al., 2006). Similarly, our choice of a constant, annual average temperature represents only a subset of in situ climate conditions. Both OM respiration rates and CaCO3 kinetics vary with temperature, and recent studies in coral reef sands have shown a clear effect of temperature on benthic CaCO3 dissolution (Stoltenberg et al., 2020; Trnovsky et al., 2016). Future studies should assess this effect in carbonate-poor sands. Because our primary variable of interest was Ω, we used filtered seawater to maintain a stable inlet Ω. Some studies have shown that particulate OM (POM) can be respired quickly in permeable sands, contributing to benthic metabolism and enhanced CaCO3 dissolution (Bühring et al., 2006; Franke et al., 2006; Lantz et al., 2020; Lantz, Schulz, et al., 2017). However, other studies have shown that POM may be highly recalcitrant (Megens et al., 2001), and may not have a significant effect on benthic O2 demand in sands (C. Wild et al., 2005). The suspended POC concentration in seawater along the California coast is typically 10 to 20 µM (Amos et al., 2019). If 10 µM of POC were completely respired in the sand column, the porewater Ωarag would drop by a maximum of 0.09, which could slightly increase CaCO3 dissolution rates. Thus, at worst, the use of filtered water in our experiment leads to conservative CaCO3 dissolution rates; and at best, it has no impact. Our columns also used repacked sand, which may have altered the in situ geochemical and/or biological profile. We believe this effect to be minor, however, since the sand was collected from a high-energy intertidal zone, where microbial gradients in surface sediment are unlikely. 67 Finally, our flow-through column setup uses constant, unidirectional advection. In reality, porewater advection in continental shelf sediments varies drastically over episodic, diurnal, and seasonal timescales, driven by dynamic processes like bio-irrigation, tidal and wave pumping, ripple migration, and flow interaction with sediment topography (Huettel et al., 2014; Precht & Huettel, 2003; Santos et al., 2012). Thus, a constant advection rate may not accurately reflect the dynamic redox and pH conditions experienced by much of the shelf sediment volume, especially at the sediment-water interface. However, because we have framed our analysis in terms of porewater residence time, we presume that our results can be applied to future in situ analyses of varying advection regimes. 3.4.2 CaCO 3 dissolution rates vary with porewater residence time Our results clearly show that decreased Ω results in increased CaCO3 dissolution; however, the magnitude of this effect depends on porewater residence time (Fig. 3.4). Thus, we first discuss the progression of porewater carbonate chemistry over time before discussing overall effects related to inlet seawater Ω. Dissolution rates progressed differently over time, according to inlet Ω. In the columns exposed to the highest Ω (unacidified) seawater, dissolution was initially negative or near-0 and then increased toward peak rates near the end of the experiment (Fig. 3.2C). In contrast, dissolution in the acidified columns peaked earlier, at or prior to 1.8 h for the most acidified (Ωarag <0.6) columns. After this initial period of differentiation, however, similar dissolution rates prevailed across all columns as porewater Ω‘s converged toward more similar values (Fig. 3.2D). The consequence of this effect can be seen in Fig. 3.4, wherein the dissolution vs. Ω slope is established 68 by 1.8 h, and additional porewater residence time simply adds a constant dissolution rate to all columns, resulting in an unchanged slope, but higher x- and y-intercepts. This time-dependence was driven by two different processes: 1) in low-Ω columns, a pulse of dissolution occurred as an immediate result of inlet undersaturation; and 2) in high-Ω columns, dissolution was delayed until OM respiration sufficiently lowered porewater Ω. These phenomena are illustrated by the Alkalinity:DIC vectors in Fig. 3.7. For the first 1.8 h porewater residence time, the low-Ω column vectors approach a slope of 1.8, indicating substantial CaCO3 dissolution (since each mol of CaCO3 dissolved produces two mol of alkalinity and one mol DIC). In contrast, the high-Ω columns initially have a slope near -0.2, indicative of aerobic OM respiration (since alkalinity is slightly decreased by the respiratory regeneration of alkalinity-consuming nutrients) and little CaCO3 dissolution (Wolf-Gladrow et al., 2007). It is not until after 1.8 h that the high-Ω vectors turn toward the dissolution trajectory. After 3.6 h, all columns have relatively similar trajectories with positive alkalinity production, as they have more similar porewater Ω, and aerobic respiration is consistently low due to hypoxic porewater. Benthic CaCO3 dissolution is often considered to be tightly coupled with OM respiration via “metabolic dissolution”, following the net reaction: CaCO3 + CH2O + O2  Ca 2+ + 2HCO3 − (5) (Emerson & Bender, 1981; Jahnke et al., 1994; John W Morse, 1985). However, our high-Ω columns show a lag between OM respiration and the onset of CaCO3 dissolution in oversaturated water. Thus, the two processes are not necessarily coupled over short porewater residence times. 69 Figure 3.7. Vectors for 1 mL min -1 experiments in alkalinity-DIC space, with background color indicating Ω aragonite . Each column is represented by three vectors, with porewater residence time represented by vector color: 0-1.8 h (black), 1.8-3.6 h (gray), and 3.6-5.4 h (white). Black circles represent inlet water, white triangles outlet water, and x’s for intermediate points. Calcite and aragonite Ω = 1 lines are shown as dashed and dotted lines, respectively. The key shows the vector direction (with arbitrary magnitude) for CaCO 3 dissolution and aerobic OM respiration. Approximate uncertainties are 5.6 μmol kg -1 for alkalinity, and 17 μmol kg -1 for DIC. (Figure excludes data for the Ω aragonite = 0.1 experiment.) This decoupling between OM respiration and CaCO3 dissolution has been described by Andersson (2015), in that coupled metabolic dissolution occurs only after a critical Ω threshold is reached. In our experiments, we do not have the resolution to see the exact porewater Ω at which dissolution initiates, however the regression for our shortest (1.8-h) porewater residence time indicates net dissolution starts to occur at inlet Ωarag =1.75 (Fig. 3.4). This dissolution threshold likely depends on CaCO3 mineralogy and OM respiration rate, among other factors, but we note that Anderson et al (2015) observed this same threshold in a coral reef sand in Bermuda, to our 2000 2050 2100 2150 2200 2250 2300 2350 2400 DIC( mol kg-1) 2000 2050 2100 2150 2200 2250 2300 2350 2400 2450 Alkalinity ( mol kg-1) 1 2 3 4 5 arag 70 knowledge the only other such published value. For ambient seawater (Ωarag =2.4), reaching this dissolution threshold (Ωarag =1.75) would require an addition of approximately 150 µmol kg -1 DIC from OM respiration. In our experiments, OM respiration during the first 1.8 h porewater residence time yielded between 100 and 160 µmol kg -1 DIC, suggesting that the dissolution threshold was reached by or shortly after 1.8 h. Indeed, the undersaturated columns did show positive net dissolution between 1.8 to 3.6 h. After this time, dissolution was relatively small for most columns, as respiration-driven dissolution became limited in the hypoxic porewater. Thus, in current, oversaturated seawater conditions, a minimum porewater residence time on the scale of 1.8 h (depending on respiration rates) is required before net dissolution occurs. This delay would presumably be longer in winter when respiration rates are lower. As ocean acidification progresses and the saturation state of overlying water moves closer to the critical Ω, however, this lag time will decrease, resulting in a faster onset of CaCO3 dissolution. Ocean acidification is not expected to affect OM respiration rates, but warming is expected to do so, which could further expedite the onset of CaCO3 dissolution (Fink et al., 2017; Hancke & Glud, 2004; Rassmann et al., 2018; Simone et al., 2021). 3.4.3 CaCO 3 dissolution occurs at bulk porewater Ω >1 CaCO3 dissolution that occurs at porewater Ωarag>1 is typically attributed to dissolution of more soluble forms of CaCO3 (e.g., high-Mg calcite), based on the assumption that any given CaCO3 species will not dissolve prior to thermodynamic undersaturation of bulk porewater (Andreas J. Andersson, 2015). However, both our aragonite and calcite 13 C spike experiments showed clearly that dissolution can occur at porewater Ω >1 (Fig. 3.5). The aragonite spike experiment had significant dissolution at Ωarag >1 by 1.8 h in the high-Ω treatment (Fig. 3.5A), and the calcite spike experiment showed dissolution at Ωcalc >1 in both the high- and low-Ω treatments, 71 even though porewater remained oversaturated for the entirety of the experiments (Fig. 3.5B). Calcite dissolution was apparent at 1.3<Ωcalc<2.2, in alignment with the critical dissolution threshold Ω=1.75 suggested in Fig. 3.4. Such dissolution in supersaturated porewater may occur in acidic micro-niches, caused by either microbial activity embedded within highly porous, biogenic carbonate, or by locally-accumulated CO2 in high tortuosity flow paths (Freiwald, 1995; Kessler et al., 2014). In our case, using abiotic CaCO3 with low intragranular porosity, the latter is most likely responsible. Although high-Mg calcite may indeed play a role in higher than expected dissolution threshold Ω’s, especially in coastal areas where more soluble CaCO3 species are common (Compere Jr & Bates, 1973; Lebrato et al., 2016; Lowenstam, 1954; McClintock et al., 2011), our results suggest that less soluble CaCO3 morphologies can dissolve well before measured porewater Ω becomes undersaturated, and thus may also contribute to seawater acid titration at higher saturation states. 3.4.4 CaCO 3 dissolution significantly increases in near-future ocean acidification conditions At all porewater residence times considered, cumulative CaCO3 dissolution increased as inlet Ω decreased. For current seawater saturation (Ωarag = 2.4), dissolution averaged 24.5 ± 23.8 mmol m -2 d -1 after 5.4 h porewater residence time. As a rough estimate for in situ rates, we divide this value by two—to account for the fact that our experiment’s unidirectional flow of seawater out of sediment requires an approximate equal surface area for flow into sediment—yielding 12.3 ± 11.9 mmol m -2 d -1 . With acidification, this rate increased by 25.6 mmol CaCO3 m -2 d -1 , per unit Ωarag decrease (for 5.4-h residence time). As an estimate for in situ conditions—again, dividing the value by two to account for inflow and outflow—the response is approximately 12.8 mmol CaCO3 m -2 d -1 , per unit Ωarag decrease. In the California Current system, water column Ωarag is projected to be consistently below 1.5, and bottom water below 1.0 by 2050 (Gruber et al., 2012). 72 Considering these projections, and our 5.4-h residence time regression, benthic CaCO3 dissolution could more than double by mid-century. Although our study is for nighttime rates, photosynthesis in sandy sediments does not appear to change with ocean acidification (although data suggests it may increase in cohesive sediments) (Fink et al., 2017; Simone et al., 2021; Vopel et al., 2018); thus considering daytime processes should not alter the slope of the dissolution:Ωarag relationship. Hence, our observed slope may be a reasonable estimate for the benthic response to ocean acidification in photic areas as well. An important consideration for applying this rate to future acidification, however, is whether or not the delivery of CaCO3 to shelf sands will keep pace with increased dissolution. 3.4.5 Similarity between carbonate-poor and carbonate-rich sands Both the current rate of CaCO3 dissolution, and response to acidification in our low- carbonate sand are similar to measurements in high-carbonate sands from coral reefs. Our sediment area-normalized rate—12.3 ± 11.9 mmol m -2 d -1 for current seawater Ω (5.4 h porewater residence time)—falls within the range of nighttime rates measured in coral reef sands, typically between 7 and 24 mmol m -2 d -1 (Andreas J Andersson et al., 2007; Cyronak, Santos, & Eyre, 2013; Cyronak, Santos, McMahon, et al., 2013; Cyronak & Eyre, 2016; Rao et al., 2012; Stoltenberg et al., 2020). However, global estimates for the carbonate-poor shelf typically assume dissolution rates more than an order of magnitude lower than this (Krumins et al., 2013; Milliman, 1993; O’Mara & Dunne, 2019). Furthermore, a comprehensive study of alkalinity release from sandy shelf sediments in the North Sea estimated average CaCO3 dissolution at 2.7 mmol m -2 d -1 (Brenner et al., 2016). Thus, it is possible that our dissolution rates, measured over a relatively long porewater residence time, are too high, and that shorter porewater residence times may better reflect in situ conditions (Fig. 3.4). However, our dissolution rate for 5.4 h porewater residence time matches 73 well with the average CaCO3 production rate (equivalent to 11 mmol m -2 d -1 ) for the California shelf, and thus may be reasonable for our field site and similar near-shore, carbonate-poor sands with little CaCO3 accumulation (S. V Smith, 1972). The response of our sand to acidification was relatively consistent across porewater residence times (-11.5 to -13.7 mmol CaCO3 dissolution m -2 d -1 , per unit Ωarag), and was also similar to that measured in high-carbonate sands. Benthic chamber studies in coral reefs have measured acidification responses between -7.7 mmol CaCO3 m -2 d -1 , per unit Ωarag for nighttime rates (Cyronak & Eyre, 2016), and -11.5 mmol CaCO3 m -2 d -1 , per unit Ωarag for a daily average across five reef ecosystems (Eyre et al., 2018). Furthermore, the point at which sands transition from net precipitating to net dissolving was also similar between sand types. For a short residence time, our transition point approached the critical porewater dissolution threshold, Ωarag = 1.75, while at longer residence times, our sand was already net dissolving at current seawater saturation (Ωarag = 2.4) (Fig. 3.4). Extrapolating the linear model for 5.4-h porewater residence time, however, yields a “transition point” of Ωarag = 3.1, which is similar to measurements in coral reef sands, between Ωarag = 2.9 to 3.0 (Cyronak & Eyre, 2016; Eyre et al., 2018). Kessler et al (2020) found a lower transition point, around Ωarag = 2.2 using short FTR’s (1.2 cm height) with high-carbonate, but this is likely due to a shorter porewater residence time. Many coral reef environments have yet to reach this transition threshold, but our site has naturally lower Ω and thus—given sufficient in situ porewater residence time—is likely already net dissolving. Because dissolution rates vary not only with Ω but also with porewater residence time, this transition point is also sensitive to changing porewater advection conditions (as driven by e.g., changing wind and current patterns). 74 The similar dissolution rates and responses to acidification in high- and low-carbonate sands, despite the vast difference in CaCO3 content—i.e., <1% in our sand vs. nearly 100% in coral reef sands—suggest that the small amount of CaCO3 in our sand was not a significant limit on dissolution, at least on the timescale of our experiments. Instead, factors related to porewater acidity, including Ω of the overlying water, and O2 and OM available for aerobic respiration (and subsequent production of CO2) are more important drivers of dissolution (Higgins et al., 2009). In other words, our carbonate-poor sand behaves much like carbonate-replete sands, in that they are acid (CO2)-limited rather than base (CaCO3)-limited. An acidifying ocean will deliver more acid to these sediments, and thus directly increase dissolution, as long as CaCO3 does not become limiting. Given our measured dissolution rates for current seawater conditions, the mass of CaCO3 in our column volume would last approximately 1 year for a 5.4-h porewater residence time, or longer for shorter residence times. With continued acidification, it is thus possible that the available CaCO3 base could be exhausted, depending on porewater residence time and rate of replenishment by sedimentation. 3.4.6 Carbonate-poor shelf sands may transition from net source of acid to net source of buffering If CaCO3 does not become limiting, increased dissolution could fundamentally change the sediment’s net impact on overlying water column chemistry. The ratio of benthic alkalinity production to DIC production (ΔAlk:ΔDIC, i.e., the flux ratio) is a useful metric for quantifying the effect of sediment processes on water column Ω and pH (see Middelburg et al., 2020 for detailed explanation). ΔAlk:ΔDIC values <1 imply net production of CO2, which lowers seawater pH and Ω, whereas values >1 imply net production of CO3 2- , which increases pH and Ω. In the present ocean, the carbonate-poor sands in our study have ΔAlk:ΔDIC well below 1 across all 75 measured porewater residence times (Fig. 3.8). Although CaCO3 dissolution does occur after 1.8 h, it produces insufficient alkalinity to negate the CO2 produced by aerobic respiration. As ocean acidification decreases seawater Ω, however, CaCO3 dissolution increases to the point at which production of dissolution-derived CO3 2- , along with alkalinity from anaerobic respiration, exceeds the production of CO2. At that point, ΔAlk:ΔDIC >1, and the net effect on overlying water is buffering, instead of acidifying. Figure 3.8. Ratio of alkalinity to DIC production (ΔAlk:ΔDIC) as a function of Ω inlet . Linear relationships are shown for all three porewater residence times: r 2 = 0.88, 0.87, and 0.83 for 1.8-, 3.6-, and 5.4-h porewater residence time, respectively, and p<0.001 for all three. The sand transitions from acidifying to buffering between Ω aragonite = 0.69 and 0.96, depending on porewater residence time. (Linear fits exclude data for the inlet Ω aragonite = 0.1 experiment.) The point at which this transition occurs depends on porewater residence time, since longer residence times allow for more CaCO3 dissolution and anaerobic respiration, while aerobic OM 76 respiration rates fall as porewater O2 is depleted. In our study, the sand transitioned to net buffering when inlet water Ωarag = 0.96 and 0.69 (Ωcalc = 1.50 and 1.07), for 1.8-h to 5.4-h porewater residence times, respectively. Longer porewater residence times would likely yield transition points at even higher Ω. While the mean global ocean is expected to remain oversaturated with respect to aragonite over the next century, areas with naturally acidic water, including the Arctic and upwelled water along the California coast, are expected to fall well below Ωarag <1, and indeed in some areas already have (Feely et al., 2018; Gruber et al., 2012; Steinacher et al., 2009). Thus, permeable shelf sediments in these areas may cross the ΔAlk:ΔDIC >1 threshold this century, at which point they become a buffer to water column acidification, given sufficient stock or input of CaCO3. In this case, the effect of sediments on the overlying water fundamentally changes. 3.4.7 Low-carbonate sands and ocean acidification: global considerations Given its large global area, the carbonate-poor shelf may be an important component of the ocean CaCO3 cycle. For example, data suggest that seawater overlying the continental shelves has recently transitioned from a net source to net sink of atmospheric CO2 (Andreas J. Andersson et al., 2005; Andreas J Andersson & Mackenzie, 2004; Laruelle et al., 2018; Mackenzie et al., 2005). While this is largely attributed to changes in both the physical and biological pump, the contribution of increased CaCO3 dissolution in shelf sediments remains poorly quantified. Because our experiments impose a constant porewater residence time and cannot account for spatial and temporal heterogeneity, we do not attempt to extrapolate absolute, measured dissolution rates to the entire low-carbonate shelf; constraining absolute rates will require more in situ data from a range of environments and advective conditions. Regardless of the absolute dissolution rate currently, however, our data clearly show that total dissolution on the low-carbonate shelf may increase significantly as the ocean becomes more acidic (Fig. 3.4). Our results suggest that a 77 decrease in Ωarag of 0.5 units (e.g., to Ωarag = 1.9 at our site)—which is expected for surface water by 2100 (Caldeira & Wickett, 2005; Gruber et al., 2012)–would increase current dissolution rates by approximately 12 mmol m -2 d -1 (consistent across all porewater residence times tested). Extrapolating this to the global low-carbonate shelf area (1.5*10 7 km 2 ; Milliman, 1993) yields a dissolution increase of 0.79 Gt C y -1 , or for only the area below the effective photic zone, 0.53 Gt C y -1 , which is more than an order of magnitude higher than the estimated total CaCO3 production for the low-carbonate shelf area (Iglesias‐Rodriguez et al., 2002; Krumins et al., 2013; Milliman, 1993). This projected increase may be unreasonably high if actual in situ advection rates are beyond the bounds (in particular, shorter) of those tested here. Alternatively, this may imply that increased dissolution in sediments may exhaust annual CaCO3 production. By definition, the carbonate-poor shelf has little to no CaCO3 accumulation. Milliman (1993) estimated that an average of 25% of CaCO3 produced in this zone accumulates, but other authors have estimated accumulation closer to zero (O’Mara & Dunne, 2019). A small CaCO3 surplus could fuel all, or a fraction of the enhanced dissolution anticipated with ocean acidification. However, in areas that currently have no CaCO3 accumulation, sediments are already CaCO3- limited, and dissolution therefore cannot increase at all. Thus, the annual production of CaCO3 in much of the shelf may be insufficient to support the potential dissolution enhancement, and benthic CaCO3 could be exhausted. If this occurs, ΔAlk:ΔDIC would remain below 1 and the CaCO3-poor shelf would continue to be a source of acid to the overlying water column. The potential for carbonate-poor sands to support enhanced dissolution with ocean acidification thus likely relies on consumption of legacy CaCO3, and/or delivery of allochthonous CaCO3 to the sediment. A variety of global change processes could contribute to both. For 78 example, legacy CaCO3 already present in continental shelf sands may be increasingly exposed to acidifying seawater by increased turbulence from storm activity, and/or deeper porewater advection from increased bottom water current speeds (Hu et al., 2020). Similarly, changes in wave climate, sea level rise, increased storm activity, precipitation and river discharge all conspire to increase erosion, especially along sandy coastlines (Ranasinghe, 2016; Vousdoukas et al., 2020). Currently, a quarter of the world’s sandy beaches are eroding at >0.5 m y -1 (Luijendijk et al., 2018). In areas with carbonate lithologies, such erosion could deliver CaCO3 from land to coastal sediments, feeding enhanced dissolution rates. Alternatively, a reduction in CaCO3 deposition (i.e., from acidification-related impacts on calcifier physiology) could limit benthic CaCO3 dissolution and/or accumulation (Simeone et al., 2018). If carbonate-poor sands do become CaCO3-limited as ocean acidification progresses, addition of CaCO3 or other basic minerals to the shelf or coastal zone via “enhanced weathering” efforts could be considered as a tool to produce ocean alkalinity and sequester dissolved CO2 via the anticipated increase in dissolution (Harvey, 2008; Ilyina et al., 2013; Meysman & Montserrat, 2017). 3.5. Summary Our study provides clear evidence that CaCO3 dissolution in carbonate-poor sands a) occurs prior to undersaturation of the overlying water; b) increases as seawater Ω decreases; and c) increases with porewater residence time. In current seawater conditions, net dissolution occurred after 1.8 h porewater residence time as a result of aerobic OM respiration decreasing porewater Ω. Dissolution occurred sooner, and at higher rates, in lower-Ω treatments. 79 The sand in our study is currently a source of acid to the overlying water (i.e., ΔAlk/ΔDIC <1), but may transition to a net buffer (ΔAlk/ΔDIC >1) between Ωarag = 0.69 and 0.96 (Ωcalc = 1.07 and 1.50), depending on porewater residence time. The higher end of this threshold will be reached in many parts of the ocean with naturally acidic water, including the California Current, by the middle of this century, so sediments could fundamentally transition to a net water column buffer, provided sufficient CaCO3 substrate. Compared to carbonate-rich, coral reef sands, the carbonate-poor sands in our study had similar rates of CaCO3 dissolution, and respond similarly to decreasing seawater Ω. The slope of the CaCO3 dissolution vs water column Ω regression in our study is similar to those from coral reef sands, suggesting that carbonate-poor sands have the potential to respond similarly to ocean acidification. A key question for future study is whether or not the pool of existing CaCO3 in carbonate-poor sands is sufficient to support increased dissolution. Global change processes such as higher storm frequency and enhanced erosion, or even intentional delivery of CaCO3 to the coastal zone may increase the availability of CaCO3 to support enhanced dissolution. 3.6 Acknowledgements This research was supported by the USC Department of Earth Sciences, the Wrigley Institute for Environmental Studies, and the Victoria J. Bertics Graduate Fellowship. The authors would like to acknowledge Nick Rollins for assistance with technical and analytical issues; Jaclyn Pittman, Liana Kaye-Lew and Rabia Ghulam-Ali for help collecting and analyzing samples; Aaron Celestian for help with mineralogy analysis; and Doug Hammond, Jess Adkins and Josh West for 80 helpful discussions. We are also grateful to Filip Meysman, Andreas Andersson, and two anonymous reviewers for their constructive feedback during the review process. 81 3.7. References Amos C. M., Castelao R. M. and Medeiros P. M. (2019) Offshore transport of particulate organic carbon in the California Current System by mesoscale eddies. Nat. Commun. 10, 1–8. Andersson A. J. (2015) A fundamental paradigm for coral reef carbonate sediment dissolution. Front. Mar. Sci. 2, 52. Andersson A. J., Bates N. R. and Mackenzie F. T. (2007) Dissolution of carbonate sediments under rising pCO 2 and ocean acidification: observations from Devil’s Hole, Bermuda. Aquat. Geochemistry 13, 237–264. Andersson A. J. and Gledhill D. (2013) Ocean acidification and coral reefs: effects on breakdown, dissolution, and net ecosystem calcification. Ann. Rev. Mar. Sci. 5, 321–348. Andersson A. J. and Mackenzie F. T. (2012) Revisiting four scientific debates in ocean acidification research. Biogeosciences 9, 893–905. Andersson A. J. and Mackenzie F. T. (2004) Shallow‐water oceans: a source or sink of atmospheric CO2? Front. Ecol. Environ. 2, 348–353. Andersson A. J., MacKenzie F. T. and Lerman A. (2005) Coastal ocean and carbonate systems in the high CO2 world of the anthropocene. Am. J. Sci. 305, 875–918. Boucher G., Clavier J., Hily C. and Gattuso J.-P. (1998) Contribution of soft-bottoms to the community metabolism (primary production and calcification) of a barrier reef flat (Moorea, French Polynesia). J. Exp. Mar. Bio. Ecol. 225, 269–283. Boudreau B. P., Huettel M., Forster S., Jahnke R., McLachlan A., Middelburg J. J., Nielsen P., Sansone F., Taghon G. and Van Raaphorst W. (2001) Permeable marine sediments: overturning an old paradigm. EOS, Trans. Am. Geophys. Union 82, 133–136. Bourke M. F., Marriott P. J., Glud R. N., Hasler-Sheetal H., Kamalanathan M., Beardall J., Greening C. and Cook P. L. M. (2017) Metabolism in anoxic permeable sediments is dominated by eukaryotic dark fermentation. Nat. Geosci. 10, 30–35. Brenner H., Braeckman U., Le Guitton M. and Meysman F. J. R. (2016) The impact of sedimentary alkalinity release on the water column CO2system in the North Sea. Biogeosciences 13, 841– 863. Broecker W. S. and Peng T. (1987) The role of CaCO3 compensation in the glacial to interglacial atmospheric CO2 change. Global Biogeochem. Cycles 1, 15–29. Bühring S. I., Ehrenhauss S., Kamp A., Moodley L. and Witte U. (2006) Enhanced benthic activity in sandy sublittoral sediments: evidence from 13C tracer experiments. Mar. Biol. Res. 2, 120– 129. Cai W. J. (2011) Estuarine and coastal ocean carbon paradox: CO2 sinks or sites of terrestrial carbon incineration? Ann. Rev. Mar. Sci. 3, 123–145. 82 Caldeira K. and Wickett M. E. (2005) Ocean model predictions of chemistry changes from carbon dioxide emissions to the atmosphere and ocean. J. Geophys. Res. Ocean. 110. Chisholm J. R. M. and Gattuso J. (1991) Validation of the alkalinity anomaly technique for investigating calcification of photosynthesis in coral reef communities. Limnol. Oceanogr. 36, 1232–1239. Compere Jr E. L. and Bates J. M. (1973) Determination of calcite: aragonite ratios in mollusc shells by infrared spectra 1. Limnol. Oceanogr. 18, 326–331. Cyronak T. and Eyre B. D. (2016) The synergistic effects of ocean acidification and organic metabolism on calcium carbonate (CaCO3) dissolution in coral reef sediments. Mar. Chem. 183, 1–12. Cyronak T., Santos I. R. and Eyre B. D. (2013a) Permeable coral reef sediment dissolution driven by elevated pCO2 and pore water advection. Geophys. Res. Lett. 40, 4876–4881. Cyronak T., Santos I. R., McMahon A. and Eyre B. D. (2013b) Carbon cycling hysteresis in permeable carbonate sands over a diel cycle: Implications for ocean acidification. Limnol. Oceanogr. 58, 131–143. Dong S., Berelson W. M., Rollins N. E., Subhas A. V, Naviaux J. D., Celestian A. J., Liu X., Turaga N., Kemnitz N. J. and Byrne R. H. (2019) Aragonite dissolution kinetics and calcite/aragonite ratios in sinking and suspended particles in the North Pacific. Earth Planet. Sci. Lett. 515, 1–12. Emerson S. and Bender M. L. (1981) Carbon fluxes at the sediment-water interface: Calcium carbonate preservation. J. Mar. Res. 39, 139–162. Emery K. O. (1968) Relict sediments on continental shelves of world. Am. Assoc. Pet. Geol. Bull. 52, 445–464. Eyre B. D., Andersson A. J. and Cyronak T. (2014) Benthic coral reef calcium carbonate dissolution in an acidifying ocean. Nat. Clim. Chang. 4, 969. Eyre B. D., Cyronak T., Drupp P., De Carlo E. H., Sachs J. P. and Andersson A. J. (2018) Coral reefs will transition to net dissolving before end of century. Science 359, 908–911. Feely R. A., Okazaki R. R., Cai W.-J., Bednaršek N., Alin S. R., Byrne R. H. and Fassbender A. (2018) The combined effects of acidification and hypoxia on pH and aragonite saturation in the coastal waters of the California current ecosystem and the northern Gulf of Mexico. Cont. Shelf Res. 152, 50–60. Fink A., den Haan J., Chennu A., Uthicke S. and de Beer D. (2017) Ocean Acidification Changes Abiotic Processes but Not Biotic Processes in Coral Reef Sediments. Front. Mar. Sci. 4, 73. Franke U., Polerecky L., Precht E. and Huettel M. (2006) Wave tank study of particulate organic matter degradation in permeable sediments. Limnol. Oceanogr. 51, 1084–1096. Freiwald A. (1995) Bacteria-induced carbonate degradation: a taphonomic case study of Cibicides lobatulus from a high-boreal carbonate setting. Palaios, 337–346. 83 Friedlingstein P., O’Sullivan M., Jones M. W., Andrew R. M., Hauck J., Olsen A., Peters G. P., Peters W., Pongratz J. and Sitch S. (2020) Global carbon budget 2020. Earth Syst. Sci. Data 12, 3269–3340. Gattuso J.-P., Frankignoulle M. and Wollast R. (1998) Carbon and carbonate metabolism in coastal aquatic ecosystems. Annu. Rev. Ecol. Syst. 29, 405–434. Gattuso J.-P., Gentili B., Duarte C. M., Kleypas J. A., Middelburg J. J. and Antoine D. (2006) Light availability in the coastal ocean: impact on the distribution of benthic photosynthetic organisms and their contribution to primary production. Biogeosciences. Gruber N., Hauri C., Lachkar Z., Loher D., Frölicher T. L. and Plattner G.-K. (2012) Rapid progression of ocean acidification in the California Current System. Science 337, 220–223. Hall S. J. (2002) The continental shelf benthic ecosystem: current status, agents for change and future prospects. Environ. Conserv. 29, 350–374. Hancke K. and Glud R. N. (2004) Temperature effects on respiration and photosynthesis in three diatom-dominated benthic communities. Aquat. Microb. Ecol. 37, 265–281. Harvey L. D. D. (2008) Mitigating the atmospheric CO2 increase and ocean acidification by adding limestone powder to upwelling regions. J. Geophys. Res. Ocean. 113. Hayes M. O. (1967) Relationship between coastal climate and bottom sediment type on the inner continental shelf. Mar. Geol. 5, 111–132. van Heuven S., Pierrot D., Rae J. W. B., Lewis E. and Wallace D. W. R. (2011) MATLAB Program Developed for CO2 System Calculations. ORNL/CDIAC-105b. Available at: https://doi.org/10.3334/CDIAC/otg.CO2SYS_MATLAB_v1.1. Higgins J. A., Fischer W. W. and Schrag D. P. (2009) Oxygenation of the ocean and sediments: consequences for the seafloor carbonate factory. Earth Planet. Sci. Lett. 284, 25–33. Hu S., Sprintall J., Guan C., McPhaden M. J., Wang F., Hu D. and Cai W. (2020) Deep-reaching acceleration of global mean ocean circulation over the past two decades. Sci. Adv. 6. Huettel M., Berg P. and Kostka J. E. (2014) Benthic Exchange and Biogeochemical Cycling in Permeable Sediments. Ann. Rev. Mar. Sci. 6, 23–51. Iglesias‐Rodriguez M. D., Armstrong R., Feely R., Hood R., Kleypas J., Milliman J. D., Sabine C. and Sarmiento J. (2002) Progress made in study of ocean’s calcium carbonate budget. Eos, Trans. Am. Geophys. Union 83, 365–375. Ilyina T., Wolf‐Gladrow D., Munhoven G. and Heinze C. (2013) Assessing the potential of calcium‐based artificial ocean alkalinization to mitigate rising atmospheric CO2 and ocean acidification. Geophys. Res. Lett. 40, 5909–5914. Jahnke R. A., Craven D. B. and Gaillard J.-F. (1994) The influence of organic matter diagenesis on CaCO3 dissolution at the deep-sea floor. Geochim. Cosmochim. Acta 58, 2799–2809. Jenkins C. (2021) dbSEABED. INSTAAR, Univ. Color. Boulder. Available at: http://instaar.colorado.edu/~jenkinsc/dbseabed/. 84 Kessler A. J., Cardenas M. B., Santos I. R. and Cook P. L. M. (2014) Enhancement of denitrification in permeable carbonate sediment due to intra-granular porosity: A multi-scale modelling analysis. Geochim. Cosmochim. Acta 141, 440–453. Kessler A. J., Chen Y.-J., Waite D. W., Hutchinson T., Koh S., Popa M. E., Beardall J., Hugenholtz P., Cook P. L. M. and Greening C. (2019) Bacterial fermentation and respiration processes are uncoupled in anoxic permeable sediments. Nat. Microbiol. 4, 1014. Kessler A. J., Rogers A., Cyronak T., Bourke M. F., Hasler-Sheetal H., Glud R. N., Greening C., Meysman F. J. R., Eyre B. D. and Cook P. L. M. (2020) Pore water conditions driving calcium carbonate dissolution in reef sands. Geochim. Cosmochim. Acta 279, 16–28. Krumins V., Gehlen M., Arndt S., Van Cappellen P. and Regnier P. (2013) Dissolved inorganic carbon and alkalinity fluxes from coastal marine sediments: Model estimates for different shelf environments and sensitivity to global change. Biogeosciences 10, 371–398. Lantz C. A., Carpenter R. C. and Edmunds P. J. (2017a) Calcium carbonate (CaCO3) sediment dissolution under elevated concentrations of carbon dioxide (CO2) and nitrate (NO3−). J. Exp. Mar. Bio. Ecol. 495, 48–56. Lantz C. A., Schulz K. G. and Eyre B. D. (2020) Ocean Acidification and Short‐Term Organic Matter Enrichment Alter Coral Reef Sediment Metabolism Through Different Pathways. J. Geophys. Res. Biogeosciences 125, e2019JG005453. Lantz C. A., Schulz K. G., Stoltenberg L. and Eyre B. D. (2017b) The short-term combined effects of temperature and organic matter enrichment on permeable coral reef carbonate sediment metabolism and dissolution. Biogeosciences 14, 5377–5391. Laruelle G. G., Cai W.-J., Hu X., Gruber N., Mackenzie F. T. and Regnier P. (2018) Continental shelves as a variable but increasing global sink for atmospheric carbon dioxide. Nat. Commun. 9, 454. Lebrato M., Andersson A. J., Ries J. B., Aronson R. B., Lamare M. D., Koeve W., Oschlies A., Iglesias‐Rodriguez M. D., Thatje S. and Amsler M. (2016) Benthic marine calcifiers coexist with CaCO3‐undersaturated seawater worldwide. Global Biogeochem. Cycles 30, 1038– 1053. Lewis E. and Wallace D. W. R. (1998) Program Developed for CO2 System Calculations, ORNL/CDIAC-105. Available at: https://salish-sea.pnnl.gov/media/ORNL-CDIAC-105.pdf. Lowenstam H. A. (1954) Factors affecting the aragonite: calcite ratios in carbonate-secreting marine organisms. J. Geol. 62, 284–322. Luijendijk A., Hagenaars G., Ranasinghe R., Baart F., Donchyts G. and Aarninkhof S. (2018) The state of the world’s beaches. Sci. Rep. 8, 6641. Mackenzie F. T., Andersson A., Lerman A. and Ver L. M. (2005) Boundary exchanges in the global coastal margin: implications for the organic and inorganic carbon cycles. In The Sea, Volume 13: The Global Coastal Ocean (ed. A. R. R. and K. H. Brink). Harvard University Press, Cambridge, MA. pp. 193–225. 85 McClintock J. B., Amsler M. O., Angus R. A., Challener R. C., Schram J. B., Amsler C. D., Mah C. L., Cuce J. and Baker B. J. (2011) The Mg-calcite composition of Antarctic echinoderms: Important implications for predicting the impacts of ocean acidification. J. Geol. 119, 457– 466. McGinnis D. F., Sommer S., Lorke A., Glud R. N. and Linke P. (2014) Quantifying tidally driven benthic oxygen exchange across permeable sediments: An aquatic eddy correlation study. J. Geophys. Res. Ocean. 119, 6918–6932. McLachlan A. (1989) Water filtration by dissipative beaches. Limnol. Oceanogr. 34, 774–780. Megens L., Van der Plicht J. and De Leeuw J. W. (2001) Temporal variations in 13C and 14C concentrations in particulate organic matter from the southern North Sea. Geochim. Cosmochim. Acta 65, 2899–2911. Meysman F. J. R. and Montserrat F. (2017) Negative CO2 emissions via enhanced silicate weathering in coastal environments. Biol. Lett. 13, 20160905. Middelburg J. J., Soetaert K. and Hagens M. (2020) Ocean alkalinity, buffering and biogeochemical processes. Rev. Geophys. 58, e2019RG000681. Milliman J. (1993) Production and accumulation of calcium carbonate in the Ocean: Budget of a non-steady state. Global Biogeochem. Cycles 7, 927–957. Milliman J. and Droxler A. (1996) Neritic and pelagic carbonate sedimentation in the marine environment: ignorance is not bliss. Geol. Rundschau 85, 496–504. Moore W. S., Sarmiento J. L. and Key R. M. (2008) Submarine groundwater discharge revealed by 228 Ra distribution in the upper Atlantic Ocean. Nat. Geosci. 1, 309–311. Moreno A. R., Garcia C. A., Larkin A. A., Lee J. A., Wang W.-L., Moore J. K., Primeau F. W. and Martiny A. C. (2020) Latitudinal gradient in the respiration quotient and the implications for ocean oxygen availability. Proc. Natl. Acad. Sci. 117, 22866–22872. Morse J. W. (1985) Chemistry of calcium carbonate-rich shallow water sediments in the Bahamas. Am. J. Sci. 285, 147–185. Naviaux J. D., Subhas A. V, Rollins N. E., Dong S., Berelson W. M. and Adkins J. F. (2019) Temperature dependence of calcite dissolution kinetics in seawater. Geochim. Cosmochim. Acta 246, 363–384. Nickl H. J. and Henisch H. K. (1969) Growth of calcite crystals in gels. J. Electrochem. Soc. 116, 1258–1260. O’Mara N. A. and Dunne J. P. (2019) Hot spots of carbon and alkalinity cycling in the coastal oceans. Sci. Rep. 9, 1–8. Oberdorfer J. A. and Buddemeier and R. W. (1986) Coral-reef hydrology: field studies of water movement within a barrier reef. Coral Reefs 5, 7–12. Orr J. C., Epitalon J. M., Dickson A. G. and Gattuso J. P. (2018) Routine uncertainty propagation for the marine carbon dioxide system. Mar. Chem. 207, 84–107. 86 Perez F. F., Fontela M., García-Ibáñez M. I., Mercier H., Velo A., Lherminier P., Zunino P., De La Paz M., Alonso-Pérez F. and Guallart E. F. (2018) Meridional overturning circulation conveys fast acidification to the deep Atlantic Ocean. Nature 554, 515–518. Precht E. and Huettel M. (2003) Advective pore‐water exchange driven by surface gravity waves and its ecological implications. Limnol. Oceanogr. 48, 1674–1684. Precht E. and Huettel M. (2004) Rapid wave-driven advective pore water exchange in a permeable coastal sediment. J. Sea Res. 51, 93–107. Ranasinghe R. (2016) Assessing climate change impacts on open sandy coasts: A review. Earth- science Rev. 160, 320–332. Rao A. M. F., Polerecky L., Ionescu D., Meysman F. J. R. and De Beer D. (2012) The influence of pore‐water advection, benthic photosynthesis, and respiration on calcium carbonate dynamics in reef sands. Limnol. Oceanogr. 57, 809–825. Rassmann J., Lansard B., Gazeau F., Guidi-Guilvard L., Pozzato L., Alliouane S., Grenz C. and Rabouille C. (2018) Impact of ocean acidification on the biogeochemistry and meiofaunal assemblage of carbonate-rich sediments: Results from core incubations (Bay of Villefranche, NW Mediterranean Sea). Mar. Chem. 203, 102–119. Reimers C. E., Stecher III H. A., Taghon G. L., Fuller C. M., Huettel M., Rusch A., Ryckelynck N. and Wild C. (2004) In situ measurements of advective solute transport in permeable shelf sands. Cont. Shelf Res. 24, 183–201. Riedl R. J., Huang N. and Machan R. (1972) The subtidal pump: a mechanism of interstitial water exchange by wave action. Mar. Biol. 13, 210–221. Rusch A., Huettel M., Wild C. and Reimers C. E. (2006) Benthic oxygen consumption and organic matter turnover in organic-poor, permeable shelf sands. Aquat. Geochemistry 12, 1–19. Santos I. R., Eyre B. and Huettel M. (2012) The driving forces of porewater and groundwater flow in permeable coastal sediments : A review. Estuar. Coast. Shelf Sci. 98, 1–15. Shum K. T. and Sundby B. (1996) Organic matter processing in continental shelf sediments—the subtidal pump revisited. Mar. Chem. 53, 81–87. Simeone S., Molinaroli E., Conforti A. and De Falco G. (2018) Impact of ocean acidification on the carbonate sediment budget of a temperate mixed beach. Clim. Change 150, 227–242. Simone M. N., Schulz K. G., Eyre B. D. and Oakes J. M. (2021) Ocean acidification may mitigate negative effects of warming on carbon burial potential in subtidal unvegetated estuarine sediments. Limnol. Oceanogr. 66, 2953–2966. Smith S. V (1972) Production of calcium carbonate on the mainland shelf of Southern California. Limnol. Oceanogr. 17, 28–41. Steinacher M., Joos F., Frölicher T. L., Plattner G.-K. and Doney S. C. (2009) Imminent ocean acidification in the Arctic projected with the NCAR global coupled carbon cycle-climate model. Biogeosciences 6, 515–533. 87 Stoltenberg L., Schulz K. G., Cyronak T. and Eyre B. D. (2020) Seasonal variability of calcium carbonate precipitation and dissolution in shallow coral reef sediments. Limnol. Oceanogr. 65, 876–891. Stoltenberg L., Schulz K. G., Lantz C. A., Cyronak T. and Eyre B. D. (2021) Late Afternoon Seasonal Transition to Dissolution in a Coral Reef: An Early Warning of a Net Dissolving Ecosystem? Geophys. Res. Lett. 48, e2020GL090811. Subhas A. V., Adkins J. F., Rollins N. E., Naviaux J., Erez J. and Berelson W. M. (2017) Catalysis and chemical mechanisms of calcite dissolution in seawater. Proc. Natl. Acad. Sci. 114, 8175– 8180. Available at: http://www.pnas.org/lookup/doi/10.1073/pnas.1703604114. Subhas A. V, Rollins N. E., Berelson W. M., Erez J., Ziveri P., Langer G. and Adkins J. F. (2018) The dissolution behavior of biogenic calcites in seawater and a possible role for magnesium and organic carbon. Mar. Chem. 205, 100–112. Sulpis O., Boudreau B. P., Mucci A., Jenkins C., Trossman D. S., Arbic B. K. and Key R. M. (2018) Current CaCO3 dissolution at the seafloor caused by anthropogenic CO2. Proc. Natl. Acad. Sci. 115, 11700–11705. Thomas H., Bozec Y., Elkalay K. and De Baar H. J. W. (2004) Enhanced Open Ocean Storage of CO2 from Shelf Sea Pumping. Science 304, 1005–1008. Trnovsky D., Stoltenberg L., Cyronak T. and Eyre B. D. (2016) Antagonistic effects of ocean acidification and rising sea surface temperature on the dissolution of coral reef carbonate sediments. Front. Mar. Sci. 3, 211. Tsunogai S., Watanabe S. and Sato T. (1999) Is there a “continental shelf pump” for the absorption of atmospheric CO2? Tellus B Chem. Phys. Meteorol. 51, 701–712. Vopel K., Del-Río C. and Pilditch C. A. (2018) Effects of CO 2 enrichment on benthic primary production and inorganic nitrogen fluxes in two coastal sediments. Sci. Rep. 8, 1–15. Vousdoukas M. I., Ranasinghe R., Mentaschi L., Plomaritis T. A., Athanasiou P., Luijendijk A. and Feyen L. (2020) Sandy coastlines under threat of erosion. Nat. Clim. Chang. 10, 260– 263. Wild C., Rasheed M., Jantzen C., Cook P., Struck U., Huettel M. and Boetius A. (2005) Benthic metabolism and degradation of natural particulate organic matter in carbonate and silicate reef sands of the northern Red Sea. Mar. Ecol. Prog. Ser. 298, 69–78. Wolf-Gladrow D. A., Zeebe R. E., Klaas C., Körtzinger A. and Dickson A. G. (2007) Total alkalinity: The explicit conservative expression and its application to biogeochemical processes. Mar. Chem. 106, 287–300. Yates K. K. and Halley R. B. (2006) Diurnal variation in rates of calcification and carbonate sediment dissolution in Florida Bay. Estuaries and Coasts 29, 24–39. Yates K. K. and Halley R. B. (2003) Measuring coral reef community metabolism using new benthic chamber technology. Coral reefs 22, 247–255. 88 Zeebe R. E. (2012) History of seawater carbonate chemistry, atmospheric CO2, and ocean acidification. Annu. Rev. Earth Planet. Sci. 40, 141–165. Zeebe R. E. and Westbroek P. (2003) A simple model for the CaCO3 saturation state of the ocean: The “Strangelove,” the “Neritan,” and the “Cretan” Ocean. Geochemistry, Geophys. Geosystems 4. 89 3.8 Supplemental Material Summary of breakthrough curve experiments We flushed two columns with deionized water for >24 hours, then added 10 ppm NaF tracer, at pump rates of 0.8 mL min-1 (30-cm port experiment) and 1.3 mL min-1 (10-cm port experiment). (Deionized water was used to avoid CaF2 precipitation in seawater.) Outlet samples were measured every 30-60 minutes on a ThermoFisher Scientific Orion F- electrode, using TISAB II buffer. In both experiments, there was no increase in F - prior to the expected breakthrough time, and breakthrough occurred later than expected. Breakthrough midpoint was ~11 h vs the expected 6.9 h for the 30-cm experiment and ~2.7 h vs the expected 1.4 h for the 10- cm experiment. We consider this delay a possible effect of adsorption; anionic tracers like Br- and F-, although often considered “conservative” are susceptible to adsorption in non-clay substrates, similar to other commonly used tracers (Korom, 2000). Adsorption appeared likely in our columns as outlet water did not reach the full inlet F - tracer concentration after >24 hours. Nonetheless, the lack of any measurable F - prior to the expected breakthrough time, and the trailing tail of the distribution suggest no significant channeling and the absence of dead zones in the columns. In the plots below (Fig. 3.S1), the expected plug flow breakthrough time for a non-adsorbed solute is shown with a dashed red line, and the observed breakthrough midpoint is shown with a solid red line. Best fit curves were defined using Solver in Microsoft Excel, optimizing the k and a values for a logistic fit, and optimizing porosity and dispersivity for the complementary error function (cerf) fit, as in Rao et al (2007). The cerf function could not be solved using our measured porosity of 0.5, again suggesting adsorption. 90 The variation in breakthrough time due to the apparent dispersivity was calculated as the difference between the observed breakthrough time (when C = 0.5*C0), and the time at which C = 1/e, which equaled 9% and 11% for the 10- and 30-cm ports, respectively. This relative value is not expected to differ for adsorbed vs. non-adsorbed solutes, and is a good estimate of the variation in porewater residence time at a given sampling port, centered around the mean breakthrough time. Overall, these experiments show no early breakthrough suggestive of channeling, and suggest instead plug flow with reasonable dispersivity. References Korom, S. F. (2000). An adsorption isotherm for bromide. Water Resources Research, 36(7), 1969-1974. Rao, A. M., McCarthy, M. J., Gardner, W. S., & Jahnke, R. A. (2007). Respiration and denitrification in permeable continental shelf deposits on the South Atlantic Bight: Rates of carbon and nitrogen cycling from sediment column experiments. Continental Shelf Research, 27(13), 1801-1819. 91 Figure 3S.1. Breakthrough curves showing tracer breakthrough at the 10-cm (A) and 30-cm port (B). The expected plug flow breakthrough time for a non-adsorbed solute is shown with a dashed red line, and the observed breakthrough midpoint is shown with a solid red line. 0 0.2 0.4 0.6 0.8 1 0 2 4 6 8 C/C0 Time (h) 10 cm port Measured C/C0 Logistic fit cerf fit 0 0.2 0.4 0.6 0.8 1 0 10 20 30 C/C0 Time (h) 30 cm port Measured C/C0 Logistic fit cerf fit 92 Table 3S.1. Alkalinity production rates. Calculated production rates of total alkalinity, alkalinity from CaCO 3 dissolution (Alk PIC ), and alkalinity from other sources, assumed to be from anaerobic respiration (Alk OM ) for the 1 mL min -1 flow rate data set. Alk PIC was calculated as DIC PIC *2, and Alk OM as total Alk minus Alk PIC . Values are shown for all three nominal porewater residence times. Estimated error (standard deviation) is 6 μmol m -2 h -1 for total alkalinity, and as shown for Alk PIC , and Alk OM . Inlet Ωarag Total Alk production rate (μmol m -2 h -1 ) AlkPIC production rate (μmol m -2 h -1 ) AlkOM production rate (μmol m -2 h -1 ) 1.8-h 3.6-h 5.4-h 1.8-h 3.6-h 5.4-h 1.8-h 3.6-h 5.4-h 0.55 (0.05) 4777 7695 9287 2772 (683) 3876 (634) 5672 (792) 2005 (683) 3818 (634) 3615 (792) 0.54 (0.05) 4440 6900 9383 3838 (1253) 4667 (934) 4898 (729) 602 (1253) 2233 (934) 4485 (729) 0.49 (0.04) 4948 7434 9846 2795 (699) 3830 (613) 4678 (682) 2153 (699) 3604 (613) 5167 (682) 0.81 (0.07) 0.52 (0.04) 4911 7262 8630 2523 (515) 4785 (754) 5131 (717) 2389 (515) 2478 (754) 3499 (717) 0.5 (0.04) 4850 6328 9428 1308 (375) 4469 (745) 5071 (649) 3542 (375) 1858 (745) 4356 (649) 0.65 (0.06) 3647 6678 8430 3625 (767) 8279 (1127) 7903 (943) 23 (767) -1602 (1127) 527 (943) 1.81 (0.12) 268 1234 1807 -966 (224) -49 (38) 335 (96) 1234 (224) 1283 (39) 1472 (96) 0.89 (0.07) 2393 5544 6801 828 (213) 2935 (421) 6084 (770) 1565 (213) 2609 (421) 718 (770) 0.12 (0.01) 15886 21733 24171 17154 (2509) 18219 (1872) 20876 (1852) -1268 (2509) 3514 (1872) 3295 (1852) 1.03 (0.08) 1820 7399 9842 729 (155) 3418 (457) 5962 (706) 1091 (155) 3982 (457) 3880 (706) 2.41 (0.14) 295 1563 4405 -1075 (219) 1398 (191) 4086 (442) 1369 (219) 165 (191) 319 (442) 2.33 (0.14) -747 -999 465 434 (127) 641 (97) 1923 (253) -1181 (127) -1640 (97) -1458 (253) 1.17 (0.09) 2624 3648 112 (45) 2941 (488) 3119 (442) -112 (45) -318 (488) 529 (442) 2.42 (0.14) -578 598 1766 -1591 (352) 299 (42) 121 (16) 1014 (352) 299 (43) 1645 (17) 1.28 (0.09) 1857 4880 6316 1005 (184) 4456 (517) 3907 (430) 852 (184) 424 (517) 2409 (430) 93 Table 3S.2. DIC production rates. Calculated production rates of total DIC, DIC from CaCO 3 dissolution (DIC PIC ), and DIC from organic matter (DIC OM ) for the 1 mL min -1 flow rate data set. DIC PIC was calculated by 13 C isotope mass balance, and DIC OM as total DIC minus DIC PIC . Values are shown for all three nominal porewater residence times. Estimated error (standard deviation) is 17 μmol m -2 h -1 for total DIC, and as shown for DIC PIC , and DIC OM . Inlet Ωarag Total DIC production rate (μmol m -2 h -1 ) DICPIC production rate (μmol m -2 h -1 ) DICOM production rate (μmol m -2 h -1 ) 1.8-h 3.6-h 5.4-h 1.8-h 3.6-h 5.4-h 1.8-h 3.6-h 5.4-h 0.55 (0.05) 7311 1935 1357 1386 (342) 1938 (317) 2836 (396) 2557 (342) 3978 (317) 4475 (396) 0.54 (0.05) 6723 1700 1459 1919 (626) 2334 (467) 2449 (364) 1625 (627) 2946 (467) 4274 (365) 0.49 (0.04) 6776 2066 701 1398 (350) 1915 (306) 2339 (341) 2516 (350) 4112 (307) 4437 (341) 0.81 (0.07) 0.52 (0.04) 7181 1762 698 1261 (257) 2392 (377) 2566 (358) 3309 (258) 4015 (377) 4615 (359) 0.5 (0.04) 7698 2833 1591 654 (187) 2235 (373) 2536 (325) 2455 (188) 3790 (373) 5162 (325) 0.65 (0.06) 8337 3119 770 1812 (383) 4140 (563) 3951 (471) 2636 (384) 3428 (564) 4386 (472) 1.81 (0.12) 5275 1845 757 -483 (112) -24 (19) 167 (48) 3061 (113) 4487 (26) 5108 (51) 0.89 (0.07) 7657 2566 1039 414 (106) 1468 (210) 3042 (385) 3528 (108) 5086 (211) 4616 (386) 0.12 (0.01) 15080 2434 2129 8577 (1255) 9110 (936) 10438 (926) 1941 (1255) 3842 (936) 4642 (926) 1.03 (0.08) 8027 2785 1380 364 (78) 1709 (228) 2981 (353) 3497 (79) 4938 (229) 5046 (354) 2.41 (0.14) 8229 2317 2125 -537 (109) 699 (96) 2043 (221) 4324 (111) 5405 (97) 6186 (222) 2.33 (0.14) 6156 2507 975 217 (63) 321 (49) 961 (127) 2458 (66) 4861 (51) 5195 (128) 1.17 (0.09) 5956 3207 833 56 (22) 1471 (244) 1560 (221) 1860 (28) 3652 (244) 4397 (222) 2.42 (0.14) 5990 2232 421 -796 (176) 149 (21) 60 (8) 4133 (177) 5420 (27) 5930 (19) 1.28 (0.09) 7801 3174 255 502 (92) 2228 (258) 1953 (215) 3870 (94) 5319 (259) 5848 (216) 94 Table 3S.3. Oxygen measurements and consumption rates. Measured O 2 data and calculated O 2 consumption for the 1 mL min -1 flow rate data set, for all three nominal porewater residence times. Columns for which complete O 2 data (measurements taken at all ports) are represented in Table 3.1 in the manuscript. The six columns with no inlet O 2 data were not treated differently from the other columns, and expected to have similar inlet O 2 as the range shown here. Inlet Ωarag Measured O2 (μM) O2 consumption rate (μmol m -2 h -1 ) Inlet 1.8-h 3.6-h 5.4-h 1.8-h 3.6-h 5.4-h 0.55 (0.05) 0.54 (0.05) 0.49 (0.04) 0.81 (0.07) 0.52 (0.04) 0.5 (0.04) 0.65 (0.06) 215 16 -5271 1.81 (0.12) 226 5 -5962 0.89 (0.07) 232 10 -5943 0.12 (0.01) 174 54 14 5 -3165 -4198 -4457 1.03 (0.08) 161 51 10 10 -3249 -4491 -4476 2.41 (0.14) 175 43 23 19 -4146 -4763 -4879 2.33 (0.14) 236 1.17 (0.09) 236 2.42 (0.14) 236 88 9 3 -3558 -5462 -5620 1.28 (0.09) 236 88 10 0 -3416 -5217 -5453 95 Chapter 4: Constraining CaCO 3 export and dissolution with an ocean alkalinity inverse model Contributors: Hengdi Liang, Sijia Dong, Will Berelson, Seth John Abstract Ocean alkalinity plays a fundamental role in the apportionment of CO2 between the atmosphere and the ocean. The primary driver of the ocean’s vertical alkalinity distribution is the formation of calcium carbonate (CaCO3) by marine organisms at the ocean’s surface, and its dissolution at depth. This so-called “CaCO3 counterpump” is poorly constrained in two key ways, however. Specifically: How much CaCO3 is exported from the ocean’s surface? And at what depth does CaCO3 dissolve to return alkalinity to the water column? Here, we created a steady-state model of global ocean alkalinity using Ocean Circulation Inverse Model (OCIM) transport, biogeochemical cycling, and field-tested calcite and aragonite dissolution kinetics. We find that limiting CaCO3 dissolution to below the aragonite and calcite saturation horizons cannot explain excess alkalinity in the upper ocean, and that models allowing dissolution above the saturation horizons best match observations. Our best performing models require export between 1.1 and 1.8 Gt PIC y -1 (from 73 m), but all converge to 1.0 Gt PIC y -1 export at 279 m, indicating that both high- and low-export scenarios can match observations, as long as high export is coupled to high dissolution in the upper ocean. This upper ocean dissolution does not result in substantial alkalinity loss from the oceans’ surface, and can thus be considered recycling. Models that either link dissolution to organic matter 96 respiration, or impose a constant dissolution rate, both yield similar dissolution profiles below ~500 m, with dissolution exceeding that expected from CaCO3 undersaturation alone throughout the entire water column. Considering the two dominant CaCO3 mineralogies, we find that both calcite and aragonite dissolve significantly in the upper ocean. These results demonstrate that dissolution is not a simple function of seawater CaCO3 saturation (Ω) and calcite or aragonite solubility, and that other mechanisms, likely related to the biology and ecology of calcifiers, must drive significant dissolution throughout the entire water column. 4.1. Introduction The global ocean is the largest reservoir of carbon (C) on Earth’s surface, storing nearly 50 times more C than the atmosphere, and acting as a significant sink of anthropogenic CO2 emissions (Friedlingstein et al., 2022). The magnitude of ocean C storage is largely controlled by the distribution of alkalinity, which is mediated by the so-called “CaCO3 counterpump” or “hard tissue pump”. The CaCO3 counterpump is driven by calcifying planktonic organisms such as coccolithophores, foraminifera, and pteropods that remove alkalinity in the form of CaCO3 from the surface ocean, and transport it to depth upon sinking. The process is termed a counterpump— in contrast to the organic C pump—because it increases CO2 and decreases alkalinity at the ocean surface, and conversely decreases CO2 at depth when alkalinity is regenerated upon CaCO3 dissolution. This reduction of alkalinity at the ocean surface reduces the ocean’s buffering capacity with respect to the atmosphere. Thus, the amount of CaCO3 formed and exported from the surface, and the depth at which alkalinity is regenerated by dissolution, is an important control on atmospheric CO2. 97 The CaCO3 counterpump remains poorly constrained, especially with respect to a) how much and what form of CaCO3 is exported from the surface; b) where in the water column or sediment it dissolves; and c) what mechanisms drive dissolution. Estimates of global CaCO3 export range between 0.4 to 1.8 Gt PIC y -1 (Berelson et al., 2007; Sulpis et al., 2021), supported by studies using a range of techniques, including satellite data combined with sinking estimates (Dunne et al., 2007; Sarmiento & Gruber, 2006); seasonal variation in surface alkalinity (K. Lee, 2001); water column alkalinity analysis (Sulpis et al., 2021); and global circulation and earth system models (Battaglia et al., 2016; Gangstø et al., 2008; Jin et al., 2006; Ridgwell et al., 2007). Canonical understanding posits that approximately three quarters of this export dissolves and the remainder is buried in sediments, however the uncertainty in export mass translates directly into uncertainty in the amount dissolved. Historically, export was thought to be dominated by the more thermodynamically stable form of CaCO3, calcite, produced by coccolithophores and foraminifera. Recent studies, however, have considered the importance of more soluble morphotypes, including aragonite (produced primarily by pteropods) and amorphous and high-Mg calcite (produced by teleosts (finned fish)) that may dissolve higher in the water column (Bednaršek et al., 2012; Buitenhuis et al., 2019; Gangstø et al., 2008; Wilson et al., 2009; Woosley et al., 2012). A growing database of planktonic calcifier observations is helping quantify the spatial distribution of calcite and aragonite production, but it remains spatially and temporally limited, and there is not yet any direct evidence to quantify fish CaCO3 production (Buitenhuis et al., 2013). Because alkalinity in the surface ocean is so important in mediating atmospheric CO2 concentrations, the second point outlined above—where CaCO3 dissolves—is often the primary focus of CaCO3 counterpump studies. Dissolution deep in the water column effectively sequesters alkalinity on the >1,000-year timescale of ocean overturning, whereas shallower dissolution may 98 return alkalinity to the surface by isopycnal or diapycnal mixing. CaCO3 dissolution is relatively slow, so in situ measurements are challenging and sparse, leaving mostly models and analyses of water column chemistry to quantify dissolution rates. Historically, dissolution was thought to occur only below the lysocline, for example at the sediment-water interface, where CaCO3 is exposed to seawater with a low mineral saturation index (Ω), and can sit for thousands of years (Sverdrup et al., 1941). More modern studies, however, have considered dissolution higher in the water column once particles fall below the calcite and aragonite saturation horizons where increased pressure and decreased carbonate ion concentration (resulting from organic matter (OM) remineralization) decrease mineral stability. Some studies have also indicated the possibility of dissolution even higher in the water column, above the saturation horizons, based on observed excess alkalinity (Chung et al., 2003; Feely et al., 2002; Sabine et al., 2002), in situ dissolution rates and C isotopic measurements (Subhas et al., 2022), and decreased sinking fluxes of CaCO3 particles in the upper water column (Berelson et al., 2007; Dong et al., 2019; Milliman & Droxler, 1996). There are currently three primary mechanisms to explain alkalinity accumulation high in the water column, above thermodynamic saturation horizons. The most simple explanation is that previous analyses do not fully consider accurate ocean mixing, and that isopycnal mixing (and to a lesser extent diapycnal mixing) between low-alkalinity, shallow waters and higher-alkalinity deeper waters can explain much of the observed pattern (Friis et al., 2006). However, this explanation ignores the observed decrease in sinking CaCO3 particles above the saturation horizons. Another potential mechanism is that CaCO3 particles in microenvironments with aerobic OM respiration (e.g., in organic particles, fecal pellets, and/or zooplankton guts) experience localized undersaturation (Alldredge & Cohen, 1987; Freiwald, 1995). Alternatively, if export of 99 more soluble CaCO3 morphotypes (e.g., aragonite or amorphous CaCO3 from teleosts) is indeed significant, they may contribute to upper ocean dissolution (Wilson et al., 2009). These different hypotheses illustrate the difficulty of using models to constrain the CaCO3 cycle, as the amount and type of mineral exported, mechanisms driving dissolution (including uncertainty in kinetic rate parameters), and ocean circulation and mixing must all be accounted for accurately. Previous modeling efforts have taken various approaches to this challenge, ranging from simple 1D approaches with no circulation (Jansen et al., 2002) to more complex models using 3D global circulation (Battaglia et al., 2016; Friis et al., 2006; Gangstø et al., 2008; Gehlen et al., 2007; Koeve et al., 2014). We describe a few here to illustrate the range of approaches. Using the MIT global circulation model, Friis et al. (Friis et al., 2006) considered dissolution driven by thermodynamic undersaturation of calcite alone (i.e., no aragonite and no respiration-driven dissolution), with a reaction rate order of 4.5. Taking a similar approach, Gehlen et al. (Gehlen et al., 2007) used the PISCES biogeochemical model with reaction order of 1; Gangstø et al. (Gangstø et al., 2008) subsequently added aragonite dissolution to the model, but–given the information available at that time–used the same reaction rate for both minerals. Koeve et al. (Koeve et al., 2014), using a transport matrix model extracted from the MIT general circulation model and applied a mechanistically-agnostic exponential decay function to represent CaCO3 dissolution. Finally, Battaglia et al. (Battaglia et al., 2016) used a Monte Carlo approach with the Bern3D Earth System Model, varying CaCO3 export, calcite:aragonite ratios, and dissolution dependence on Ω for both mineral types. While all of these studies have provided valuable insight into the ocean CaCO3 cycle, Battaglia et al. found that different model parameterizations can perform well, highlighting the fact that our understanding of dissolution mechanisms remains poorly constrained. 100 Here, we use Ocean Circulation Inverse Model (OCIM) transport coupled with biogeochemistry implemented in the AWESOME OCIM (AO; John et al., 2020) to test several hypotheses of CaCO3 export and dissolution. We also use the most recent formulations of calcite and aragonite dissolution kinetics available, derived from in situ experiments (Dong et al., 2019; Naviaux, Subhas, Dong, et al., 2019; Subhas et al., 2022). Because OCIM-based models are less computationally intensive than other earth system or global circulation models, we are able to test several hypotheses, in each case optimizing model parameters over thousands of individual simulations in order to match observations. Specifically, we test the following four hypotheses: 1) CaCO3 dissolution occurs only at the seafloor; 2) dissolution in the water column is driven only by water column Ω, at rates defined by calcite and aragonite dissolution kinetics; 3) water column dissolution also depends on OM respiration; and 4) water column dissolution is constant and uncorrelated with Ω. 4.2. Methods 4.2.1 Model framework Model overview The AWESOME OCIM (AO) is a MATLAB-based modeling tool to simulate ocean tracer distributions through biogeochemical processes. Global ocean circulation is represented by a water transport matrix generated by the Ocean Circulation Inverse Model (OCIM 1.0; DeVries & Primeau, 2011; DeVries, 2014), which is constrained by global observational data, including temperature, salinity, CFCs, and radiocarbon. An advantage of AO is its computational efficiency, with each solution of a 3-D tracer distribution taking approximately 8 seconds on a laptop under 101 the conditions used in this study, allowing multiple parameters to be optimized for each model configuration of hypothetical biogeochemical processes. To constrain the global ocean alkalinity cycle, we incorporated into the AO several spatially-explicit observational and modeled datasets as well as biogeochemical process rates (Fig. 4.1). These inputs are described in more detail in the following sections. The spatial distribution of CaCO3 export from the ocean surface, and dissolution within the ocean interior by several different hypothetical mechanisms was then modeled, with specific parameter values optimized to best match the observed ocean alkalinity distribution from the GLODAP database. Figure 4.1. Schematic diagram of the alkalinity model. Constraints based on external datasets (fixed values) and variable parameters (optimized in each model run; marked in blue) are fed into the Awesome OCIM framework. For each model iteration, the resulting TA0 and TA* are compared to observed values based on GLODAP data. The optimized parameters are adjusted, and the model rerun until a minimum cost is found. The four different models of water column CaCO 3 dissolution used different parameters, as noted by the superscripts (M2, M3, and M4) (see Section 2.2). 102 Sources and sinks Alkalinity supply to the ocean is calculated by multiplying river alkalinity concentrations with river discharge. Our model includes the world’s 34 major rivers with the largest discharge and/or those which have reliable measurements from previous reports (Amiotte-Suchet et al., 2003). This approach ignores smaller rivers, and thus underestimates global fluvial alkalinity discharge. Therefore rivers entering each ocean basin (Pacific, Atlantic, Indian, Arctic, Mediterranean) are multiplied by a separate scaling factor so that the alkalinity influx to each basin matches reported estimates according to the procedure of Amiotte-Suchet et al. (Amiotte-Suchet et al., 2003). Modern ocean alkalinity is assumed to be in steady state, with alkalinity supply exactly balancing alkalinity burial in sediments (described in the next paragraph). Published riverine alkalinity fluxes, however, are deficient compared to estimated CaCO3 burial, potentially due to the contribution of non-riverine alkalinity input (e.g., anaerobic respiration on continental shelves) and/or alkalinity in riverine particulate matter, which has traditionally not been considered in riverine alkalinity flux measurements (Middelburg et al., 2020). Thus, assuming that riverine alkalinity sources may be underestimated, we applied an additional scaling factor, increasing each river’s alkalinity flux so that total river alkalinity flux equaled total burial flux; this required a global scaling factor of 1.5. Benthic alkalinity sinks via burial are based on previous model-based estimates of CaCO3 burial in both coastal and pelagic ocean sediments (Dunne et al., 2012; O’Mara & Dunne, 2019). It is important to note that our model only simulates planktonic CaCO3 production, and does not explicitly model coral or other benthic CaCO3 formation in coastal areas. For grid cells in which 103 high rates of coastal CaCO3 burial exceed surface ocean CaCO3 production, alkalinity is removed from the overlying model grid cell, effectively simulating benthic CaCO3 production. Because of our model’s coarser grid than the original studies, some areas with high coral CaCO3 production may be missed. However, CaCO3 burial attributed to coral formation is small on the global scale, and does not significantly influence either the distribution of alkalinity within the ocean or the global balance of sources and sinks. Other sources and sinks of alkalinity such as submarine groundwater discharge, hydrothermal vents, anaerobic respiration, marine silicate weathering and reverse weathering were not considered due to their small and/or uncertain magnitudes (Middelburg et al., 2020). In addition to balancing external sources and sinks, our steady-state model also requires a constraint on the total ocean inventory of alkalinity. This is accomplished by imposing a very slow input of alkalinity everywhere in the ocean, with the concentration of this input being the global average alkalinity concentration, which was balanced by a slow, first-order loss of alkalinity from everywhere in the ocean. Both the input and loss processes occur on a timescale of 1 million years, which is slow enough to have insignificant influence on alkalinity distributions while setting the magnitude of concentration, as outlined as the conc function described in John et al. (John et al., 2020). The global total flux of this in situ source and sink of alkalinity is 13 times smaller than global riverine fluxes and burial, and thus does not noticeably affect the distribution of alkalinity within the oceans. CaCO3 export from the euphotic zone In the AO, the top two layers (sea surface to 73 m) are considered the euphotic zone, where calcifying organisms precipitate CaCO3. CaCO3 export is calculated based on net phosphorus 104 uptake (Puptake) in the euphotic zone, multiplied by the C:P ratio and the ratio of particulate inorganic carbon (PIC) to particulate organic carbon (POC) (PIC:POC ratio, also known as the rain ratio): CaCO3 export = (POC) (PIC:POC) = (Puptake) (C:P) (PIC:POC) (Eq. 1) Puptake is an output from a previous phosphorus model under the OCIM framework (Weber et al., 2018). Instead of using the canonical Redfield C:P ratio 106, we adopted the C:P ratios that vary geographically as a function of NO3 concentrations, with the highest C:P ratios in the subtropical gyres (Martiny et al., 2013) (Fig. 4.S1). We used spatially-varying PIC:POC ratios from GFDL- ESM2M modeling results. The GFDL-ESM2M model has a full ecosystem of plankton community, and also splits PIC into calcite and aragonite components, which we utilize in a subset of our models, as described below (Dunne et al., 2013). Using this approach, the total modeled CaCO3 export is 2.3 Gt PIC y -1 , exceeding most current estimates. Therefore, we introduced an optimizable scaling factor into each model, constraining total export between estimates of 0.3 to 1.8 Gt PIC y -1 (Berelson et al., 2007). The uncertainties in CaCO3 export estimates largely lie in the PIC:POC ratio estimates, and the scaling factor serves as an adjustment to allow for uncertainty in global CaCO3 export (Balch et al., 2005; Sarmiento et al., 2002). Then for each mole of CaCO3 produced, two moles of alkalinity are consumed due to the loss of a CO3 2- ion. This process is represented as alkalinity loss in each grid cell by multiplying each cell’s CaCO3 net production by two. In addition to the optimized export for each model, we also performed sensitivity tests in which we enforced low, medium or high export (0.5, 1.1, and 1.8 Gt PIC y -1 , respectively), as described in Section 2.4. 105 4.2.2 CaCO 3 dissolution scenarios Exported CaCO3 may dissolve either in the water column or upon reaching the sediments, returning alkalinity to seawater. Here, we tested four scenarios of CaCO3 dissolution: 1) dissolution only at the seafloor, 2) dissolution based on the saturation state (Ω), 3) dissolution based upon Ω with an additional contribution to dissolution associated with OM remineralization, and 4) constant dissolution of CaCO3 particles throughout the water column, unrelated to any specific biogeochemical mechanism. Each of the four models have different optimized adjustable parameters. In all four models, the CaCO3 export scaling factor and global mean alkalinity concentration, as described above, were optimized. In Models 1 and 2, dissolution is prescribed, whereas Models 3 and 4 have additional optimized parameters related to dissolution mechanisms, as described below. The optimized parameters for each model are listed in Table 4.2. Model 1 (M1): Benthic only As particulate CaCO3 reaches the seafloor in our model, all or a fraction of that CaCO3 may be buried, as set by the burial data described above. Any CaCO3 reaching the seafloor in excess of burial is assumed to dissolve, releasing alkalinity to the overlying model grid cell. Within M1, all CaCO3 produced in the euphotic zone reaches the sediments and is either buried or dissolves into the overlying ocean-bottom grid cell. Model 2 (M2): Ω-dependent dissolution Our second model considers benthic dissolution, as above, as well as dissolution within the water column as driven by thermodynamic undersaturation of calcite and aragonite. We obtained the saturation states (Ω) of calcite and aragonite from the GLODAP mapped product (Lauvset et 106 al., 2016). The kinetic dissolution of the two minerals is described as a function of the saturation states using an empirical formulation (Keir, 1980; J W Morse et al., 1979): Rkinetic = k (1-Ω) n (Eq. 2) where k (g g -1 d -1 ) is the rate constant and n is the reaction order. The k and n values for both calcite and aragonite species were based on dissolution studies in seawater across a wide range of saturation states, which were adjusted to the temperature and pressure conditions where the majority of water column dissolution occurs (5 °C and below 700 m) (Dong et al., 2018, 2019; Naviaux, Subhas, Rollins, et al., 2019; Subhas et al., 2018) (Table 4.1). We note that this empirical function is not mechanistically accurate (Dong et al., 2020; Naviaux, Subhas, Dong, et al., 2019), as dissolution rates transition occurs at Ωcritical values. However, the kinetic formulation provides a simple approximation and is thus adopted here. Table 4.1. Kinetic dissolution parameters for calcite and aragonite Mineralogy 0.85 ≤ Ω < 1 0 < Ω < 0.85 k (g g -1 d -1 ) n k (g g -1 d -1 ) n calcite 0.016 0.33 0.50 2.2 aragonite 0.0026 0.13 0.028 1.5 Here we adopted the terminology of previous literature, but implemented Eq. 2 with Rkinetic being a dissolution rate constant, which controls the magnitude of particle flux attenuation through: Fi = Fi-1 exp(- 𝑅𝑅 𝑘𝑘 𝑖𝑖 𝑘𝑘 𝑐𝑐 𝑐𝑐 𝑖𝑖 𝑐𝑐 𝑘𝑘 ℎ 𝑖𝑖 𝑣𝑣 ) (Eq. 3) 107 where Fi is the CaCO3 particle flux entering the i th model layer from the top of the grid, hi is the height of the i th model layer, and v is the particle sinking speed in m d -1 . Since we assume net production of CaCO3 in the top two layers, the particle flux entering the top of the 3 rd layer is the CaCO3 export from the top 73 m. M2 uses Eq. 2 and 3, combined with a canonical CaCO3 sinking rate (100 m d -1 ), to calculate CaCO3 dissolution below the euphotic zone. This model thus allows for CaCO3 dissolution in undersaturated seawater, as well as in sediments. As we discuss later and explore in more detail in our sensitivity analyses (Section 2.3), the sinking velocity of PIC is a poorly constrained yet important parameter in all model formulations. Model 3 (M3): Ω + respiration based dissolution Our third model includes benthic and Ω-driven CaCO3 dissolution as in M2, and adds an additional driver of dissolution coupled to the remineralization of OM. Respiratory acids from aerobic OM respiration have been hypothesized to stimulate CaCO3 dissolution, either within sinking particles or in the guts or feces of grazers, although a global-scale formulation of the dependence of dissolution on respiration remains elusive (Milliman et al., 1999). In M3, we parameterize this process by including a CaCO3 dissolution component linked to OM respiration (Rrespiration) specified by: Rrespiration = x Prem m (Eq. 4) where Prem is the remineralization rate of particulate organic phosphorus within each grid cell, and x and m are model-optimized scaling parameters. The vertical Prem profile built into the AO follows the classic Martin curve, wherein the exponent b in the Martin curve has been optimized by Weber et al. (2018) to fit observed phosphate data. Particulate organic phosphorus remineralization is 108 used here, as it is assumed to be proportional to organic carbon remineralization. Then we use Rkinetic + Rrespiration, instead of Rkinetic alone, as the dissolution rate constant and implemented it into Eq. 3 to model the CaCO3 particle flux. In this model, x and m for calcite and aragonite are optimized separately to simulate the respiration effect on the dissolution of each mineral. Model 4 (M4): Constant dissolution Our fourth model does not consider seawater Ω or OM respiration, and instead assumes a fixed length-scale for remineralization throughout the water column. As our models assume a single fixed sinking rate for carbonate (100 m d -1 ), this is equivalent to specifying a fixed dissolution rate constant Rconst, which is optimized by the model, to substitute for Rkinetic in Eq. 3. As with the other three models, CaCO3 reaching the seafloor in excess of prescribed burial is returned to the water column as alkalinity by benthic dissolution. This mechanism is independent of Ω, and has no specific biogeochemical interpretation, though it has been previously employed in other models and produces a reasonable match to observations (Battaglia et al., 2016). 109 Table 4.2. Optimized parameters, model performance, and CaCO 3 export and dissolution for the four base models. All fluxes are in units of Gt PIC y -1 . Model Benthic only (M1) Ω-dependent (M2) Ω + respiration (M3) Constant (M4) Optimized parameters export, mean TA 1 export, mean TA export, mean TA, x arag , m arag , x calc , m calc export, mean TA, R const R 2 (TA0; TA*) 0.75; 0.68 0.75; 0.71 0.62; 0.87 0.56; 0.88 CaCO 3 export from 73 m 0.68 0.71 1.84 1.12 Aragonite 0.19 0.20 0.52 0.32 Calcite 0.49 0.51 1.32 0.80 CaCO 3 export from 279 m 0.66 0.69 0.95 1.02 Total water column dissolution (% of water column dissolution above 1700 m) 0 0.09 (14.5%) 1.38 (85.4%) 0.77 (57.7%) Aragonite water column dissolution (% of aragonite water column dissolution above 1700 m) 0 0.03 (23.3%) 0.40 (63.4%) 0.21 (58.4%) Calcite water column dissolution (% of calcite water column dissolution above 1700 m) 0 0.06 (10.7%) 0.98 (94.5%) 0.56 (57.4%) Benthic dissolution 0.42 0.36 0.20 0.09 1 mean TA is the global mean concentration of salinity-normalized TA (TA / salinity x 35) 4.2.3 Sensitivity experiments Three sets of sensitivity tests were conducted to evaluate the model sensitivity to key processes including S1) the magnitude of surface CaCO3 export, S2) the effect of particle sinking speed, and S3) the calcite/aragonite ratio. 110 The total magnitude of CaCO3 export from the surface ocean in the four base models is optimized. To test whether our models are sensitive to surface export, however, a separate set of experiments was performed for each model in which we specified ‘low’, ‘medium’ and ‘high’ export levels of 0.5, 1.1, 1.8 Gt PIC yr -1 , respectively. In subsequent discussions, the optimized export models are referred to as “base” models, in contrast to the “export sensitivity” models. Our base models assume a fixed particle sinking speed of 100 m d -1 , based on previous observations and modeling efforts (Battaglia et al., 2016). However, real-ocean particle sinking speeds are related to both particle size and density, and a wide range of sinking speeds can be generally consistent with ocean observations (Cael et al., 2021). Thus, to examine the extent to which sinking speed impacts CaCO3 dissolution profiles, we varied sinking speeds between 1 - 1,000 m d -1 for M2 and 50 - 200 m d -1 for M3. Note that M1 is not affected by sinking speed, since no particles dissolve before settling at the seafloor. M4 is similarly unaffected because the model- optimized rate constant Rconst is inversely related to sinking velocity, so an increase in sinking rate has the same effect as a decrease in the optimized dissolution constant. The two mineral forms of CaCO3 considered in this study, calcite and aragonite, have distinct properties in terms of morphology, solubility and dissolution rates. We rely on the global distributions of surface calcite and aragonite from an ecosystem model that includes various planktons types (Dunne et al., 2013). However, the actual global distribution of calcite and aragonite export is poorly constrained. Thus, to investigate how the relative export of calcite vs. aragonite influences model performance, end-member scenarios were tested assuming that all CaCO3 export is calcite, or that it is entirely aragonite. Again this has no impact on the M1 and M4 models, but it affects both other models as the dissolution kinetics and respiration sensitivity 111 of aragonite and calcite dissolution differ (Eq. 2 and Eq. 4). We performed this end-member test based on M2, since changing the calcite:aragonite ratio could significantly modify the dissolution patterns of this model, where the kinetic dissolution of these two minerals is the only mechanism that drives water column CaCO3 dissolution. 4.2.4 Model optimization Alkalinity data and metric definitions While our model simulates the cycling of total alkalinity (TA) in the global oceans, the specific quantities it seeks to match to observations is not TA, but rather two components of TA: preformed alkalinity (TA0) and alkalinity from CaCO3 dissolution (TA*), both of which are calculated as described below. CaCO3 dissolution cannot feasibly be measured directly, so researchers instead use metrics for calculating accumulated alkalinity from dissolution since a water mass leaves the surface ocean. There are two similar but distinct approaches commonly in use: 1) potential alkalinity (PALK) (Brewer et al., 1975; Carter et al., 2014; Sarmiento et al., 2002); and 2) TA* (Feely et al., 2002). Essential to both methods is the understanding that TA can be deconstructed into three components: 1) preformed alkalinity, or the alkalinity of a water mass at the time of last ventilation; 2) addition of alkalinity through CaCO3 dissolution; and 3) consumption of alkalinity by the release of inorganic acids following OM respiration (Brewer et al., 1975; Wolf-Gladrow et al., 2007). Other sources and sinks of TA are considered trivial. Preformed alkalinity accounts for at least 95% of alkalinity in all water masses, followed by CaCO3 dissolution, and to a smaller degree consumption of alkalinity by respiratory acids. The small amount of alkalinity added by CaCO3 112 dissolution relative to the large background of preformed alkalinity necessitates that we quantify both components accurately. By this formulation, we define TA as: TA = TA0 + TA* - TAr (Eq. 5) where TA0 is the preformed component, TA* is accumulated alkalinity from CaCO3 dissolution, and TAr is alkalinity consumed by respiration-derived acids. TA0 has typically been quantified by linear regression using salinity (S), PO, NO (measured plus estimated consumed O2 from aerobic respiration, as measured by phosphate or nitrate concentrations, respectively), and/or potential temperature; and TAr has been estimated with PO or apparent O2 utilization (Feely et al., 2002; Koeve et al., 2014). Improved surface observations and circulation models have enabled better modeling of both TA0 and preformed nutrients, the latter enabling calculation of TAr by subtracting preformed from measured nutrients, then multiplying by an alkalinity scaling factor (Carter et al., 2021). Given the AO’s capability to calculate TA0 and preformed nutrients using a circulation matrix, we use this latter approach. It is worth briefly explaining the other common metrics PALK and Alk*, both for comparison and to justify our choice of using TA*. Whereas the TA* approach uses explicit values for each alkalinity component, PALK is a composite metric useful for analysis of water mass chemistry when preformed values are unknown. PALK combines total measured alkalinity and estimated TAr (TA + 1.26 [NO3 - ]0). Carbonate alkalinity can then be calculated by subtracting an estimated preformed potential alkalinity value (Brewer et al., 1975; Carter et al., 2014; Fry et al., 2015). For example, the Alk* metric used by Carter et al. (Carter et al., 2014) is calculated as PALK minus preformed potential alkalinity, defined = 66.4 S. This approach assumes that: 1) 113 preformed alkalinity varies only with S; and 2) preformed nutrients are negligible and globally constant. In some areas however (e.g., the Southern Ocean), upwelled water with high alkalinity and nutrients is subducted before sufficiently mixing with other waters, leading to both higher than expected TA0 and non-negligible preformed nutrients (Carter et al., 2021; Koeve et al., 2014). These deviations would be interpreted as erroneously high Alk* (combined effect in Fig. 4.2a, and isolated preformed alkalinity effect in Fig. 4.2b). Given our model’s ability to incorporate preformed values, we have avoided this issue by using the TA* approach instead. Using the TA* approach, there are still decisions to be made about how to calculate TA0 and preformed nutrients. We adopted the method described in Carter et al. (2021) using concentrations at the deepest monthly-mean mixed layer depths as boundary conditions in the OCIM circulation matrix. Concentrations at the base of the mixed layer can be calculated either by regression against other variables (e.g., T and S) or by linear interpolation of gridded alkalinity datasets (Carter et al., 2021). We adopted the latter approach, interpolating TA concentrations from the GLODAP mapped dataset (Lauvset et al., 2016). For comparison, our calculated TA0 data show only minor deviations from the previously published TA0 in Carter et al. (2021), e.g. in the deep Pacific Ocean compared to their regression-based TA0 (Fig. 4.2c) and little difference compared to their interpolation-based TA0 (Fig. 4.2d). To calculate TAr, we multiplied excess PO4 3- (i.e., PO4 3- above the preformed value, where the preformed PO4 3- is calculated in a similar approach as TA0) by an alkalinity:PO4 3- ratio of 21.8 (Wolf-Gladrow et al., 2007). 114 Figure 4.2. Differences resulting from various approaches to estimating preformed alkalinity, compared to TA0 AO . Negative values represent deficits in preformed alkalinity, which would be interpreted as excess CaCO 3 dissolution in TA analysis. The Alk* approach yields relatively low preformed alkalinity, especially in the Southern Ocean, resulting from assumptions about both preformed nutrients and alkalinity (combined in a, and only alkalinity component in b). For approaches using circulation matrices to reconstruct TA0, results are sensitive to the method of averaging the mixed layer depth alkalinity, for example either by regression (c) or by interpolation (d). In (c) and (d), TA0 Carter is the preformed alkalinity data from Carter et al. (2021), which used the interpolation approach. Optimization criteria and genetic algorithm Each model is optimized towards a fit with a global database of alkalinity observations, starting with TA as taken from the GLODAP mapped product (Lauvset et al., 2016), re-gridded to 115 fit the AO with a data density of 2°x2° horizontally and 24 vertical depth levels. For each model run, the modeled TA0 and TA* (TA0model and TA*model) for each grid cell were compared to their respective “observed” values (TA0obs and TA*obs). TA0obs was calculated from GLODAP alkalinity observations at the base of the mixed layer as a constraint within our circulation model (Section 2.4.1). For each grid cell, TA*obs was calculated using Eq. 5, where TA is from the GLODAP dataset, and TA0obs and TArobs are as described in Section 2.4.1. Note that TA, TA0obs, and TArobs, and hence TA*obs, are all derived entirely from the GLODAP dataset and our OCIM circulation matrix, and do not depend on other model parameters. During the optimization process, grid cells at high latitudes (north of 70°N and south of 70°S) are excluded, since sparse temporal and spatial sampling introduces uncertainty in observational data. The data in the top two model layers are also excluded from optimization, as a small change in the export scaling factor could lead to a large change in surface alkalinity, resulting in the overweighting of the surface ocean at the expense of the ocean interior where CaCO3 dissolution occurs. A genetic algorithm (GA) was applied to tune the free model parameters and seek the best model fit to observational data. The natural selection process of this algorithm helps the model start from random initial parameter guesses within the given ranges, and is therefore less likely to be trapped in a local minimum. We utilized the MATLAB Global Optimization Toolbox which offers the function ‘ga’ to facilitate the use of this optimization solver. The GA tries to find the minimum of a function, in this case the misfit between model simulation and observational data, weighted by a volume factor w (w = volume in each model grid / global ocean volume), according to the cost function: 𝑐𝑐𝑐𝑐 𝑐𝑐𝑐𝑐 = ∑ (𝑇𝑇𝑇𝑇 0 𝑚𝑚 𝑐𝑐 𝑚𝑚 𝑐𝑐 𝑚𝑚 , 𝑖𝑖 − 𝑇𝑇𝑇𝑇 0 𝑐𝑐 𝑜𝑜𝑘𝑘 , 𝑖𝑖 ) 2 𝑤𝑤 𝑖𝑖 𝑁𝑁 𝑜𝑜 𝑜𝑜𝑜𝑜 𝑖𝑖 = 1 + ∑ (𝑇𝑇𝑇𝑇 𝑚𝑚 𝑐𝑐 𝑚𝑚 𝑐𝑐 𝑚𝑚 , 𝑖𝑖 ∗ − 𝑇𝑇𝑇𝑇 𝑐𝑐 𝑜𝑜 𝑘𝑘 , 𝑖𝑖 ∗ ) 2 𝑤𝑤 𝑖𝑖 𝑁𝑁 𝑜𝑜 𝑜𝑜𝑜𝑜 𝑖𝑖 = 1 (Eq. 6) 116 We defined the cost with both TA0 and TA*, instead of just TA, as TA on its own conceals error in its components (e.g., positive TA0 misfit and negative TA* misfit would cancel each other, resulting in a “good” TA fit). It is important to note that the modeled TA0 and TA* are not completely independent, as surface CaCO3 export, which feeds TA* in the ocean interior, reduces surface alkalinity, thus impacting TA0. Thus, a good match between modeled and observed TA0 alone is not sufficient to conclude that the model is accurately describing alkalinity cycling, if the model fit to observed TA* is not good, and vice versa. Hence our equal weighting of both quantities in the cost function puts equal emphasis on TA0 and TA* to constrain each component of alkalinity separately. 4.3. Results 4.3.1 Modeled vs observed TA 0 and TA* The four dissolution mechanisms produce distinct results for both TA0 and TA*, although some similar characteristics are observed across models. An example of optimized model results for M3 is shown in Fig. 4.3. (All base models are shown in Fig. 4.S2.) Modeled TA0 exhibits small error relative to TA0obs, with a significant misfit concentrated in the low-latitude surface mixed layer and in the North Atlantic Deep Water (NADW), reflecting ocean circulation patterns (Table 4.2; Fig. 4.3a, 4.3c, 4.3e). The error in TA* is similar in magnitude to TA0, but with larger horizontal and vertical spatial variations due to the uncertainties associated with CaCO3 export and vertical dissolution profiles (Fig. 4.3b, 4.3d, 4.3f). Specifically, this dissolution mechanism allows for high dissolution rates in the upper ocean driven by OM respiration and lower rates of dissolution in the deeper waters driven by undersaturation. 117 Figure 4.3. Example of two modeled transects (Pacific at 161°W and Atlantic at 29°W) compared to observations for M3 (Ω + respiration model). Observed TA0 and TA* based on the GLODAP dataset (a and b); Modeled TA0 and TA* (c and d); and the difference between model and observation (model - observation) (e and f). Considering TA0, the two simpler models, M1 and M2, match observations relatively well below the euphotic zone (Fig. 4.4a-b and Fig. 4.4e-f), whereas the two models that allow dissolution in oversaturated water in the upper water column (M3 and M4), have a worse model- data fit, especially in the North Atlantic Ocean (Fig. 4.4i-j and Fig. 4.4m-n). The reduced performance in the two latter models is mostly a result of higher CaCO3 export and thus reduced TA0 in surface water prior to subduction. All models yield a deficit in TA0 compared to TA0obs at low- and mid-latitudes in the upper ocean, potentially due to overproduction of CaCO3 in the surface. However, since these areas are not significant downwelling areas, this does not greatly affect TA0 in the abyssal ocean. The four models also produce distinct patterns in TA*. M1 has the worst fit to TA*obs (Fig. 4.4c-d). Because all CaCO3 dissolves at the seafloor, modeled TA* exceeds observations near the 118 sediment-water interface and has a deficit in shallower water. M2 shows slightly better model performance, because it allows for calcite and aragonite to dissolve in undersaturated waters (Fig. 4.4g-h). However, the dissolution rate constants and sinking speed (100 m d -1 ) used in the base model do not facilitate significant water column dissolution, thus the majority of particles (80%) still dissolve at the seafloor. M3 improves the model-data fit by shifting dissolution upward in the water column (Fig. 4.4k-l). In this scenario, most of the exported CaCO3 (77%) dissolves in the upper 15 levels of the model, above 1,700 meters, primarily due to respiration-driven dissolution. As a result of the upward shift in dissolution, the excess deep ocean TA* observed in M1 and M2 is minimized, but not entirely resolved. M4 also reproduces TA* well, with errors spread more evenly throughout the ocean, and a slightly better fit than M3 (Fig 4.4o-p). All four models do relatively well in reproducing TA* in the North Atlantic, where water mass age is young and observed TA* is low. In contrast, because TA* is cumulative, model biases accumulate as a water mass ages, so errors are largest in the oldest waters in the North Pacific Ocean, as seen most obviously in M1 and M2. The model performance of TA* relative to water age is also apparent in the joint probability density function plots for M1 and M2, where TA* is modeled well for low TA* waters, then becomes too low at intermediate TA* concentrations (Southern Ocean, upper Pacific Ocean) and too high at the highest TA* sites (deep North Pacific Ocean) (Fig. 4.4d and 4.4h). All four models consistently underestimate TA* in the Southern Ocean. 119 Figure 4.4. Difference between modeled and observed TA0 and TA* (TA0: a, e, i, m; TA*: c, g, k, o) in representative transects (Pacific at 161°W and Atlantic at 29°W). Joint probability density function plots of modeled vs observed data in each grid cell show the relative performance of each model (TA0: b, f, j, n; TA*: d, h, l, p). 120 4.3.2 CaCO 3 export Optimized export fluxes differ across the four base models. M1 and M2 optimize at lower CaCO3 export since particles have no or little chance of dissolving in the water column, both with 0.7 Gt PIC y -1 export out of the top 73 meters (Table 4.2). In contrast, M3 has the highest optimized CaCO3 export from the surface, reaching the upper bound of the constrained range, 1.8 Gt PIC y - 1 . This high export is necessary to satisfy high rates of dissolution linked to OM respiration below the mixed layer. We note that maximum export was constrained to 1.8 Gt PIC y -1 ; given a higher constraint, it is possible that the model could optimize to an even higher export value. M4 has an intermediate CaCO3 export compared to the other three models, resulting in 1.1 Gt PIC y -1 export out of the top 73 m. 4.3.3 CaCO 3 dissolution The four models also produce distinct CaCO3 dissolution vs. depth patterns (Fig. 4.5). In M1, constraining dissolution to the bottom grid cell naturally results in high rates of dissolution at depth and no water-column dissolution (Fig. 4.5a-b). M2 introduces dissolution to the undersaturated waters, but the slow rates result in small dissolution in the water column, which increases gradually with depth as the saturation state decreases, with most dissolution again occurring at the seafloor (Fig. 4.5c-d). As a result, M2 yields significant benthic dissolution both in the abyssal ocean and in shallow waters on the continental shelves, where export is high and the water column is oversaturated with respect to both calcite and aragonite. Linking CaCO3 dissolution to OM respiration (M3) results in significant water column dissolution, characterized by two distinct dissolution peaks: one at shallow depths just below the surface and another near the seafloor, with less dissolution occurring at intermediate depths (Fig. 4.5e-f). Extremely high rates of water-column dissolution in the upper 500 m dominate the total dissolution pattern, while 121 water-column dissolution below 500 m decreases continuously with depth since the respiration driver of dissolution outweighs any potential increase driven by undersaturation. Benthic dissolution is significant at shallow shelves and below 2,500 m at seafloor. The constant dissolution model (M4) spreads dissolution relatively evenly throughout the water column (Fig. 4.5g-h). M4 does not produce a peak in upper ocean dissolution such as observed in M3, however below about 500 meters, it exhibits a relatively similar trend, with a gradual decrease in water- column dissolution (Fig. 4.5f, 4.5h). Both M3 and M4 have high dissolution rates throughout the water column in the highly productive North Atlantic compared to the rest of the oceans. Figure 4.5. (a, c, e, g) CaCO 3 dissolution (µmol kg -1 y -1 ) in all four base models for representative transects (Pacific at 161°W and Atlantic at 29°W). Black lines represent the aragonite (dashed) and calcite (solid) saturation horizons. (b, d, f, h) Globally-integrated vertical profiles of total water column and benthic dissolution for each model (Gmol m -1 y -1 ). Black lines indicate water column dissolution, and red lines indicate benthic dissolution. M1 and M2 (benthic-only and Ω-dependent models) are dominated by benthic dissolution at depth (b, d). M3 (Ω + respiration model) has an extreme dissolution peak in the upper ocean (f inset), below which total dissolution is minor in comparison (f), and below about 500 m, relatively similar to M4 (constant dissolution model) (h). 122 In the two models considering mineral dissolution kinetics (M2 and M3), the relative magnitudes and profiles of calcite and aragonite dissolution differ. In M2, there is no or little dissolution of either mineral in the upper ocean where most of the seawater is supersaturated, and aragonite and calcite dissolution rates are approximately equal until about 2,500 m (Fig. 4.6a). Although much of the ocean volume above this depth is undersaturated with respect to aragonite, but not calcite, the kinetics of aragonite dissolution are relatively slow, resulting in similar dissolution for both minerals. Similarly, the slow aragonite dissolution kinetics prevent any obvious peak in dissolution rates near the aragonite saturation horizon. Below 2,500 m, calcite dissolution dominates due to its higher export and higher kinetic rate parameters. In contrast, M3 exhibits very different calcite vs aragonite dissolution profiles in the upper ocean, with an extreme surface peak dominated by calcite (Fig. 4.6b). The model optimizes such that calcite is preferentially dissolved in the upper ocean by OM respiration, apparently in order to maximize the export and dissolution of aragonite at intermediate depths, above the calcite saturation horizon. Below the surface peak, calcite dissolution decreases rapidly, following the decrease in OM respiration according to the adjusted power-law Martin curve, and only starts rising again below about 2,500 m due to the onset of undersaturation-driven dissolution. Aragonite dissolution also decreases with depth, but less dramatically than calcite, given its higher saturation horizon. 123 Figure 4.6. Global average profiles of calcite and aragonite dissolution for the two models that consider mineral kinetics (M2 and M3). The Ω-dependent model (M2) results in disproportionate calcite dissolution at depth due to faster dissolution kinetics (a), while the Ω + respiration model (M3) results in disproportionate calcite dissolution in the upper ocean (b inset). 4.3.4 Sensitivity experiments The high/mid/low export sensitivity tests revealed that M1 and M2 optimized best at low export (0.5 Gt C y -1 ); M3 optimized well for both mid and high export (1.1 and 1.8 Gt C y -1 , respectively); and M4 optimized well only for mid export (Figs. 4.S3-S6). Of all scenarios tested, the TA* fit was best for M3 mid, M3 high export, and M4 mid export. In the sinking speed sensitivity test for M2, scenarios with sinking speeds >=100 m d -1 all yielded excess TA* at depth, and the best TA* model fit was achieved for sinking speeds between 5 and 10 m d -1 (Fig. 4.S7). However, at slower sinking speeds, TA* accumulation at depth shifts away from the North Pacific, and into the North and central Atlantic. Limited sensitivity tests for sinking speed in M3 revealed a similar trend (Fig. 4.S8). 124 Varying calcite and aragonite endmember fractions in the M2 model made little difference for the base sinking speed of 100 m d -1 . A combined sensitivity analysis with sinking speed showed that slower sinking speeds again improved model performance slightly for both the calcite and aragonite scenario, but TA* nonetheless accumulated in the Atlantic with either endmember (Fig. 4.S9). 4.4. Discussion 4.4.1. Mixing alone cannot replicate TA* observations The well-documented presence of positive TA* above the calcite and aragonite saturation horizons is typically interpreted as evidence that CaCO3 must dissolve high in the upper water column, prior to thermodynamic expectations based on seawater Ω (Feely et al., 2002). Friis et al. (2006) suggested an alternative explanation, that water mass mixing is sufficient to explain much of the observed TA*. The poor performance of M1 and M2 however, shows that restricting CaCO3 dissolution to depths below the calcite and aragonite saturation horizons cannot adequately explain observed TA* distributions (Fig. 4.4c-d, 4.4g-h). These models do show—similar to Friis et al.’s model—that non-negligible TA* indeed accumulates above the saturation horizons as a result of mixing; however this mixing is insufficient to replicate the overall TA* distribution. Instead, alkalinity deficiency persists in the upper ocean, and excess alkalinity accumulates at depth. This is most prominent in the North Pacific, where water masses are oldest, and thus error in modeled TA* accumulates. Excess TA* at depth in the North Pacific, relative to GLODAP observations, is also evident in Friis et al.’s model. The circulation component of our model includes diapycnal and isopycnal mixing, and is one of the most accurate representations of ocean circulation 125 available, so we doubt that circulation deficiencies could be significant enough to explain the error in TA* in M2. Rather, limiting dissolution to below the saturation horizon clearly concentrates TA* at depth and suggests a missing source of alkalinity in the upper ocean. 4.4.2. Calcite and aragonite dissolution at bulk seawater Ω cannot explain observed TA* profiles The addition of water column dissolution in M2 shifted only a small amount of alkalinity regeneration upwards compared to M1 (Fig. 4.4g-h). However, this model is highly sensitive to the relative dissolution rate, which varies as a function of both sinking speed and kinetics (Fig. 4.S7). At our default sinking speed of 100 m d -1 , the TA* distribution is similar to M1, suggesting that dissolution kinetics are too slow and/or sinking speed too fast to support significant water column dissolution. The kinetic parameters for CaCO3 dissolution have been debated at length (reviewed in Sarmiento and Gruber), but those we used have been extensively tested, including field-verified at depth in the North Pacific (Dong et al., 2019; Naviaux, Subhas, Dong, et al., 2019; Subhas et al., 2022). That said, kinetic parameters are known to vary as a function of calcifier morphology and mineralogy, seawater DOC content, and pressure, among other factors, and this variability is not accounted for in our model (Dong et al., 2018; Naviaux, Subhas, Dong, et al., 2019). Sinking speed also introduces significant uncertainty, as it may vary by many orders of magnitude depending on particle size and morphology. Measured sinking speeds range from <1 m d -1 for single coccoliths, to 5,000 m d -1 for large foraminifera, with most observations for individual coccolithophores, foraminifera, and pteropods falling between 100 to 1,000 m d -1 (Bergan et al., 2017; Engel et al., 2009; Schiebel & Hemleben, 2017; M. Walker et al., 2021; Zhang et al., 2018). The actual sinking speed profile of CaCO3 particles is most certainly a distribution across these values. For modeling simplicity, we assumed a single sinking speed of 100 m d -1 in our base 126 models, but also considered sinking speeds between 1 and 1,000 m d -1 in our sensitivity analysis. Because both sinking speed and dissolution rate affect dissolution linearly in our model, this range captures three orders of magnitude of uncertainty in either sinking speed or dissolution rate. The optimized sinking speed for a kinetics-only scenario (a sensitivity test based on M2) was 5-10 m d -1 , which is likely too slow for realistic CaCO3 particles. Furthermore, as sinking speed decreased, excess TA* at depth simply shifted away from the North Pacific, and into the North and central Atlantic. This pattern arises because the relatively deep saturation horizons in the Atlantic (>2 km depth) allow dissolution of slowly sinking particles, and accumulation of alkalinity in the NADW. In the North Pacific, however, the shallow saturation horizons (<1 km depth) allow dissolution near the surface, where alkalinity becomes entrained into the TA0 pool, leaving little CaCO3 export to feed TA* at depth. Ultimately, both fast and slow sinking speed scenarios (or equivalently, slow and fast dissolution kinetics) result in excess TA* at depth: either concentrated in the Atlantic (slower sinking speeds), or in both oceans (faster sinking speeds). Thus, there must be a significant source of dissolution of CaCO3 above the calcite and aragonite saturation horizons. 4.4.3. CaCO 3 dissolution linked to aerobic OM respiration recycles alkalinity in the upper ocean The two models that allowed dissolution above the saturation horizons – M3 and M4 – both yielded TA* distributions that better match observations, compared to the models restricting dissolution to depth (Fig. 4.4). Because the oxidation rate of sinking organic particles is highest near the surface ocean, linking dissolution directly to OM respiration (M3) resulted in an extreme dissolution peak in the upper ocean, and consequently required more CaCO3 export (1.8 Gt PIC y - 127 1 ) to supply sufficient TA* at depth. The sensitivity test imposing a lower export (1.1 Gt PIC y -1 ) also performed well, with a weaker model-optimized dependence on OM respiration, and therefore less total CaCO3 dissolution (Fig. 4.S5). These observations match the findings of Battaglia et al. (Battaglia et al., 2016), that both high and low CaCO3 export scenarios can produce realistic alkalinity distributions, when balanced by higher or lower dissolution in the upper ocean, respectively. Indeed, across the four base models, the model-optimized CaCO3 export is correlated to the amount of dissolution in the upper ocean (Table 4.2). Interestingly, high export from the surface ocean (73 m) in M3 does not result in decreased TA0 relative to M4, which has lower optimized export (Fig. 4.4i and 4.4m). This suggests that high export, coupled to high dissolution in the upper ocean, does not significantly alter surface alkalinity, and that regenerated alkalinity is retained within the upper ocean. In other words, surplus “export” in high-export scenarios can effectively be considered upper ocean recycling. The optimized export for M3 (1.8 Gt PIC y -1 ) is on the high end of published values (Berelson et al., 2007; Sulpis et al., 2021). A confounding issue with published export values, however, is that export is defined at different depths. If dissolution is indeed concentrated in the upper water column, then the export gradient will be steep across shallow depths, and the defined export will vary significantly depending on the reference depth chosen, ranging from 73 m (our study, as well as Battaglia et al., 2016; Jin et al., 2006), 100 m (Gangstø et al., 2008; Sarmiento et al., 2002), 200 m (Berelson et al., 2007), or 300 m (Sulpis et al., 2021). Because we defined a shallow export horizon, our export in M3 may therefore not be anomalously high, but rather it may include a large amount of CaCO3 that is dissolved between 73 and 300 m. For example, our modeled export at 279 m is 1.0 Gt PIC y -1 for the base model and 0.9 Gt PIC y -1 for the lower- export sensitivity test, strikingly close to each other, and to other recent estimates (0.9 Gt PIC y -1 128 from 300 m; Sulpis et al., 2021) (1.1 Gt PIC y -1 from the mixed layer; Jin et al., 2006; Lee, 2001). Considering the geographically-variable, maximum monthly-mean mixed layer depth built into our model, instead of a constant 279-m horizon, yields an intermediate M3 export value of 1.5 Gt PIC y -1 . While dissolution in M3 is concentrated in the upper water column, dissolution throughout the water column is still higher than that driven by calcite and aragonite undersaturation alone (Fig. 4.5d, 4.5f), indicating that OM respiration drives dissolution throughout the entire water column, not only in the upper ocean where respiration rates are high. Below 2,500 m, dissolution at the sediment-water interface also becomes significant (0.2 Gt PIC y -1 ). This general pattern of two prominent dissolution zones--one in the upper ocean and one at the sediment-water interface- -was recently suggested by Sulpis et al. (Sulpis et al., 2021), using a completely different approach, combining Alk* analysis for the water column and a coupled geochemical-hydrodynamic sediment model. Our modeled results also show this pattern, however we see a more gradual decrease in dissolution with depth, as opposed to a dissolution minima at ~600 m in their study (Fig. 4.5f). Our model also predicts more total water column dissolution, and less benthic dissolution than their analysis, which could be a result of their Alk* analysis, which cannot differentiate benthic and pelagic sources of alkalinity. 4.4.4. Constant dissolution model similar to Ω + respiration model at depth Optimized CaCO3 export in M4 is significantly less than in the optimized M3, however as noted above, both models converge to ~1.0 Gt PIC y -1 at 279 m (Fig. 4.7). Furthermore, they are surprisingly similar at depth (> 500 m), despite their different “mechanisms” (Fig. 4.5f, 4.5h). For example, total pelagic dissolution below 1,700 m is 0.3 and 0.2 Gt PIC y -1 for M4 and M3, 129 respectively. The lower dissolution in the latter model is compensated for by higher benthic dissolution. The similarities in model performance at depth reinforce the observation that alkalinity regenerated from upper ocean dissolution in M3 is largely retained and recycled in the upper ocean, and does not impact the water column TA* inventory. Figure 4.7. CaCO 3 export profiles for the three best-performing models in the export sensitivity tests. “M3 (mid)”: Ω + respiration model with 1.1 Gt PIC y -1 export. “M3 (high)”: Ω + respiration model with 1.8 Gt PIC y -1 export. “M4 (mid)”: constant dissolution model with 1.1 Gt PIC y -1 export. 4.4.5. Potential mechanisms driving dissolution above the saturation horizons The potential for significant upper ocean dissolution has been discussed for decades, and is supported by several lines of observational evidence, including excess alkalinity and decreased sinking fluxes of calcitic particles (Feely et al., 2002, 2004; Milliman & Droxler, 1996; Sabine et al., 2002). More recent trap data has specifically shown upper water column disappearance of 130 aragonite (Dong et al., 2019) and specific calcifying groups, including pteropods (Oakes & Sessa, 2020) and coccolithophores (Ziveri et al., 2022). The mechanistic link between dissolution and OM respiration has been proposed as either CaCO3 experiencing low-Ω microniches caused by microbial respiration within sinking particles, or low-Ω environments in the guts of grazers. Our models do not attempt to determine which mechanism is more important, as concentrations of OM and grazers are closely linked. With either mechanism, however, bulk water column Ω may not accurately reflect the Ω experienced by CaCO3 particles. Subhas et al. (Subhas et al., 2022) used water column pH, nutrient, and C isotope measurements to show that respiration and CaCO3 dissolution are in fact tightly coupled in waters oversaturated for calcite in the N. Pacific. A constant dissolution scenario, in which dissolution is completely independent of seawater Ω and OM respiration, has been previously proposed for modeling simplicity, sidestepping the question of mechanisms (Battaglia et al., 2016). In theory, such a dissolution profile could be produced by a combination of respiration-driven dissolution and variable sinking speeds that effectively spread regenerated alkalinity throughout the water column. For example, particles with very slow sinking speeds following seasonal phytoplankton blooms could theoretically provide a nearly constant TA* source to the water column. Such ephemeral dissolution sources may be missed in observations, yet may be important in defining vertical alkalinity distribution (Schiebel, 2002). Similarly, rapidly sinking particles may also be missed in export measurements (Schiebel, 2002; C. R. Smith et al., 1996). Another potential explanation for significant dissolution in the upper ocean, not associated with OM respiration, is the presence of more soluble forms of CaCO3, including those produced by teleosts. Using liberal production and dissolution assumptions, teleost-produced crystallites 131 may constitute as much as 45% of ocean CaCO3 production (Wilson et al., 2009). However, we still know relatively little about teleost spatial and depth distribution, and essentially nothing about depth of crystallite production and excretion, crystallite production rates, and mineralogy/solubility in the open ocean (Irigoien et al., 2014; Lam & Pauly, 2005; Perry et al., 2011; Salter et al., 2017, 2018, 2019). Teleosts typically migrate between 200 to 1,000 m depth during the day, and the surface where they feed at night. The uptake of alkalinity (via CaCO3 crystallite formation) and release of alkalinity (via crystallite excretion and subsequent dissolution) may therefore occur at various depths within this range, depending on where the CaCO3 is formed, where it’s excreted, and how quickly it sinks and dissolves. For example, an upward alkalinity flux could theoretically occur if crystallites are formed at depth and excreted at the surface where they dissolve quickly. Alternatively, crystallites which are produced, excreted, and dissolved at the same depth would have no net effect on alkalinity. Given the right combination of formation vs excretion depth and mineral solubility, it is possible that teleost-produced CaCO3 could contribute to a “constant” dissolution profile. More research is needed to better constrain the magnitude of teleost CaCO3 production, as well as where in the water column it is produced, excreted, and dissolves. Considering the similar dissolution profiles for the constant and respiration models below ~500 m, one can imagine a combination of processes contributing to either one, or a hybrid, of the two models. Processes of potential importance include strong coupling of dissolution with OM respiration, variable particle sinking rates, CaCO3 consumption and excretion by grazers, and highly soluble “fish CaCO3”. Note that these are all fundamentally connected to the biology and ecology of calcifiers, and cannot be constrained by current modeling capabilities or chemical oceanographic analyses alone. Continued observations in the geographic and seasonal distribution 132 of planktonic calcifiers (Buitenhuis et al., 2013), as well as related modeling efforts, will therefore be useful in constraining these factors (Bednaršek et al., 2012; Buitenhuis et al., 2013; O’Brien et al., 2013; Schiebel & Movellan, 2012; Zaric et al., 2006) 4.4.6. The importance of calcite vs aragonite The relative contribution of calcite vs aragonite to alkalinity regeneration is another important component linking ecological structure and the CaCO3 pump, especially in light of changing ocean conditions that may alter planktonic ecosystem dynamics. The contribution of aragonite to total CaCO3 export is still poorly constrained, with observations ranging from 1 to 98% (Anglada-Ortiz et al., 2021; Berner & Honjo, 1981; Dong et al., 2019; Fabry & Deuser, 1991; Ziveri et al., 2022). However, a recent global compilation of available data suggests that aragonite constitutes at least one third of CaCO3 export (Buitenhuis et al., 2019). The relative importance of aragonite vs calcite dissolution depends greatly on the mechanisms driving dissolution. If only considering seawater Ω and dissolution kinetics, aragonite is indeed disproportionately important to upper water column dissolution due to its shallow saturation horizon. For example, in M2, aragonite accounted for only 28% of export, but ~50% of pelagic dissolution above 1700 m. (Our sinking speed sensitivity test showed this was true for all sinking speeds >5 m d -1 ). Gangstø et al. (Gangstø et al., 2008) came to similar results in their model, despite using different kinetic parameters. An important caveat, however, is that M2 is dominated by dissolution deeper in the water column and at the sediments, so upper water column dissolution is miniscule by comparison. At depth, calcite dissolution is much more significant than aragonite due to its higher export and faster kinetics (Fig 4.6a). 133 If dissolution is linked to OM respiration (as in M3), however, upper ocean alkalinity regeneration can be driven by both calcite (78%) and aragonite (22%), in fractions similar to their export ratios (Fig. 4.6b). Although calcite is more thermodynamically stable in upper ocean conditions, observations support our finding that both calcitic and aragonitic particles dissolve significantly in the upper ocean (Manno et al., 2007; Oakes et al., 2019; Oakes & Sessa, 2020; Schiebel, 2002; Subhas et al., 2022; Ziveri et al., 2022). Furthermore, experiments have shown that aragonite production and dissolution alone is insufficient to explain upper ocean alkalinity (Fabry, 1990). These studies and our modeled results indicate that bulk seawater Ω and mineral solubility may be less important than factors related to respiration-driven dissolution, such as the extent of mineral-associated labile OM, sinking speed, and grazing pressure. For example, pteropods have been shown to dissolve from the inside-out due to respiration of internal OM, irrespective of seawater Ω (Oakes et al., 2019). Foraminifera, on the other hand, often sink as empty tests following reproduction (Schiebel & Hemleben, 2017). Finally, coccolithophores do not sink quickly unless aggregated in organic particles, effectively “waiting” until they associate with organic matter and the potential for OM respiration (Biermann & Engel, 2010; Engel et al., 2009). These examples illustrate that the importance of aragonite vs calcite to upper ocean dissolution is ecosystem- and species-dependent, and will also benefit from continued progress in plankton biogeographic surveys. 4.5. Conclusions Testing four different models of CaCO3 dissolution (benthic only, Ω-dependent, Ω + respiration, and constant dissolution), we have shown that circulation alone cannot explain 134 observed TA* distributions, and that dissolution must occur above the calcite and aragonite saturation horizons. In our best-performing models (Ω + respiration and constant dissolution), both of which allow significant dissolution above the saturation horizons, we found a range of optimized CaCO3 export values from 1.1 to 1.8 Gt PIC y -1 just below the model euphotic zone at 73 m. However, considering a deeper export depth (279 m), export for both models converged at 1.0 Gt PIC y -1 . The surplus dissolution in the higher export model was effectively recycled within the upper ocean, suggesting that a range of “export” values can be accommodated in CaCO3 cycling scenarios, as long as recycling is proportionately increased in the uppermost ocean. Below ~500 m, both the Ω + respiration and constant dissolution models had similar dissolution profiles, supporting higher dissolution rates than what would occur due to mineral undersaturation alone. While our models cannot conclusively determine mechanisms driving CaCO3 dissolution, we clearly show that dissolution is not simply a function of seawater Ω and calcite and aragonite dissolution kinetics. Rather, potential dissolution mechanisms are all intrinsically related to the biology and ecology of calcifiers inhabiting the ocean’s surface. 4.6 Acknowledgements This research was supported by funding from the National Science Foundation to W.B. (OCE 1834475) and S.J. (OCE 2049639); by the Simons Foundation to S.J. (SCOPE Award 329108); and by the University of Southern California Dornsife College of Letters, Arts, and Sciences, and Provost Fellowship to A.L. We wish to thank Brendan Carter for sharing preformed alkalinity data, and Thomas Weber for sharing PIC:POC data. 135 4.7. References Alldredge, A. L., & Cohen, Y. (1987). Can microscale chemical patches persist in the sea? Microelectrode study of marine snow, fecal pellets. Science, 235(4789), 689–691. Amiotte-Suchet, P., Probst, J., & Ludwig, W. (2003). Worldwide distribution of continental rock lithology: Implications for the atmospheric/soil CO2 uptake by continental weathering and alkalinity river transport to the oceans. Global Biogeochemical Cycles, 17(2). Anglada-Ortiz, G., Zamelczyk, K., Meilland, J., Ziveri, P., Chierici, M., Fransson, A., & Rasmussen, T. L. (2021). Planktic foraminiferal and pteropod contributions to carbon dynamics in the Arctic Ocean (north Svalbard margin). Frontiers in Marine Science, 8, 636. Balch, W. M., Gordon, H. R., Bowler, B. C., Drapeau, D. T., & Booth, E. S. (2005). Calcium carbonate measurements in the surface global ocean based on Moderate‐Resolution Imaging Spectroradiometer data. Journal of Geophysical Research: Oceans, 110(C7). Battaglia, G., Steinacher, M., & Joos, F. (2016). A probabilistic assessment of calcium carbonate export and dissolution in the modern ocean. Biogeosciences, 13(9), 2823–2848. Bednaršek, N., Možina, J., Vogt, M., O’Brien, C., & Tarling, G. A. (2012). The global distribution of pteropods and their contribution to carbonate and carbon biomass in the modern ocean. Earth System Science Data, 4(1), 167–186. Berelson, W. M., Balch, W. M., Najjar, R., Feely, R. A., Sabine, C., & Lee, K. (2007). Relating estimates of CaCO3 production, export, and dissolution in the water column to measurements of CaCO3 rain into sediment traps and dissolution on the sea floor: A revised global carbonate budget. Global Biogeochemical Cycles, 21(1). Bergan, A. J., Lawson, G. L., Maas, A. E., & Wang, Z. A. (2017). The effect of elevated carbon dioxide on the sinking and swimming of the shelled pteropod Limacina retroversa. ICES Journal of Marine Science, 74(7), 1893–1905. Berner, R. A., & Honjo, S. (1981). Pelagic sedimentation of aragonite: its geochemical significance. Science, 211(4485), 940–942. Biermann, A., & Engel, A. (2010). Effect of CO 2 on the properties and sinking velocity of aggregates of the coccolithophore Emiliania huxleyi. Biogeosciences, 7(3), 1017–1029. Brewer, P. G., Wong, G. T. F., Bacon, M. P., & Spencer, D. W. (1975). An oceanic calcium problem? Earth and Planetary Science Letters, 26(1), 81–87. Buitenhuis, E. T., Le Quere, C., Bednaršek, N., & Schiebel, R. (2019). Large contribution of pteropods to shallow CaCO3 export. Global Biogeochemical Cycles, 33(3), 458–468. 136 Buitenhuis, E. T., Vogt, M., Moriarty, R., Bednaršek, N., Doney, S. C., Leblanc, K., Le Quéré, C., Luo, Y.-W., O’Brien, C., & O’Brien, T. (2013). MAREDAT: towards a world atlas of MARine Ecosystem DATa. Earth System Science Data, 5(2), 227–239. Cael, B. B., Cavan, E. L., & Britten, G. L. (2021). Reconciling the size‐dependence of marine particle sinking speed. Geophysical Research Letters, 48(5), e2020GL091771. Carter, B. R., Feely, R. A., Lauvset, S. K., Olsen, A., DeVries, T., & Sonnerup, R. (2021). Preformed properties for marine organic matter and carbonate mineral cycling quantification. Global Biogeochemical Cycles, 35(1), e2020GB006623. Carter, B. R., Toggweiler, J. R., Key, R. M., & Sarmiento, J. L. (2014). Processes determining the marine alkalinity and calcium carbonate saturation state distributions. Biogeosciences, 11(24), 7349–7362. Chung, S., Lee, K., Feely, R. A., Sabine, C. L., Millero, F. J., Wanninkhof, R., Bullister, J. L., Key, R. M., & Peng, T. (2003). Calcium carbonate budget in the Atlantic Ocean based on water column inorganic carbon chemistry. Global Biogeochemical Cycles, 17(4). DeVries, T. (2014). The oceanic anthropogenic CO2 sink: Storage, air‐sea fluxes, and transports over the industrial era. Global Biogeochemical Cycles, 28(7), 631–647. DeVries, T., & Primeau, F. (2011). Dynamically and observationally constrained estimates of water-mass distributions and ages in the global ocean. Journal of Physical Oceanography, 41(12), 2381–2401. Dong, S., Berelson, W. M., Adkins, J. F., Rollins, N. E., Naviaux, J. D., Pirbadian, S., El-Naggar, M. Y., & Teng, H. H. (2020). An atomic force microscopy study of calcite dissolution in seawater. Geochimica et Cosmochimica Acta, 283, 40–53. Dong, S., Berelson, W. M., Rollins, N. E., Subhas, A. V, Naviaux, J. D., Celestian, A. J., Liu, X., Turaga, N., Kemnitz, N. J., & Byrne, R. H. (2019). Aragonite dissolution kinetics and calcite/aragonite ratios in sinking and suspended particles in the North Pacific. Earth and Planetary Science Letters, 515, 1–12. Dong, S., Subhas, A. V, Rollins, N. E., Naviaux, J. D., Adkins, J. F., & Berelson, W. M. (2018). A kinetic pressure effect on calcite dissolution in seawater. Geochimica et Cosmochimica Acta, 238, 411–423. Dunne, J. P., Hales, B., & Toggweiler, J. R. (2012). Global calcite cycling constrained by sediment preservation controls. Global Biogeochemical Cycles, 26(3). Dunne, J. P., John, J. G., Shevliakova, E., Stouffer, R. J., Krasting, J. P., Malyshev, S. L., Milly, P. C. D., Sentman, L. T., Adcroft, A. J., & Cooke, W. (2013). GFDL’s ESM2 global coupled climate–carbon earth system models. Part II: carbon system formulation and baseline simulation characteristics. Journal of Climate, 26(7), 2247–2267. 137 Dunne, J. P., Sarmiento, J. L., & Gnanadesikan, A. (2007). A synthesis of global particle export from the surface ocean and cycling through the ocean interior and on the seafloor. Global Biogeochemical Cycles, 21(4). Duteil, O., Koeve, W., Oschlies, A., Aumont, O., Bianchi, D., Bopp, L., Galbraith, E., Matear, R., Moore, J. K., & Sarmiento, J. L. (2012). Preformed and regenerated phosphate in ocean general circulation models: can right total concentrations be wrong? Biogeosciences, 9(5), 1797–1807. Engel, A., Szlosek, J., Abramson, L., Liu, Z., & Lee, C. (2009). Investigating the effect of ballasting by CaCO3 in Emiliania huxleyi: I. Formation, settling velocities and physical properties of aggregates. Deep Sea Research Part II: Topical Studies in Oceanography, 56(18), 1396–1407. Fabry, V. J. (1990). Shell growth rates of pteropod and heteropod molluscs and aragonite production in the open ocean: implications for the marine carbonate system. Journal of Marine Research, 48(1), 209–222. Fabry, V. J., & Deuser, W. G. (1991). Aragonite and magnesian calcite fluxes to the deep Sargasso Sea. Deep Sea Research Part A. Oceanographic Research Papers, 38(6), 713–728. Feely, R. A., Sabine, C. L., Lee, K., Berelson, W., Kleypas, J., Fabry, V. J., & Millero, F. J. (2004). Impact of anthropogenic CO2 on the CaCO3 system in the oceans. Science, 305(5682), 362– 366. https://doi.org/10.1126/science.1097329 Feely, R. A., Sabine, C. L., Lee, K., Millero, F. J., Lamb, M. F., Greeley, D., Bullister, J. L., Key, R. M., Peng, T., & Kozyr, A. (2002). In situ calcium carbonate dissolution in the Pacific Ocean. Global Biogeochemical Cycles, 16(4), 91. Freiwald, A. (1995). Bacteria-induced carbonate degradation: a taphonomic case study of Cibicides lobatulus from a high-boreal carbonate setting. Palaios, 337–346. Friedlingstein, P., Jones, M. W., O’Sullivan, M., Andrew, R. M., Bakker, D. C. E., Hauck, J., Le Quéré, C., Peters, G. P., Peters, W., & Pongratz, J. (2022). Global carbon budget 2021. Earth System Science Data, 14(4), 1917–2005. Friis, K., Najjar, R. G., Follows, M. J., & Dutkiewicz, S. (2006). Possible overestimation of shallow‐depth calcium carbonate dissolution in the ocean. Global Biogeochemical Cycles, 20(4). Fry, C. H., Tyrrell, T., Hain, M. P., Bates, N. R., & Achterberg, E. P. (2015). Analysis of global surface ocean alkalinity to determine controlling processes. Marine Chemistry, 174, 46–57. Gangstø, R., Gehlen, M., Schneider, B., Bopp, L., Aumont, O., & Joos, F. (2008). Modeling the marine aragonite cycle: changes under rising carbon dioxide and its role in shallow water CaCO 3 dissolution. Biogeosciences, 5(4), 1057–1072. 138 Gehlen, M., Gangstø, R., Schneider, B., Bopp, L., Aumont, O., & Éthé, C. (2007). The fate of pelagic CaCO 3 production in a high CO 2 ocean: a model study. Biogeosciences, 4(4), 505– 519. Irigoien, X., Klevjer, T. A., Røstad, A., Martinez, U., Boyra, G., Acuña, J. L., Bode, A., Echevarria, F., Gonzalez-Gordillo, J. I., & Hernandez-Leon, S. (2014). Large mesopelagic fishes biomass and trophic efficiency in the open ocean. Nature Communications, 5(1), 1–10. Jansen, H., Zeebe, R. E., & Wolf‐Gladrow, D. A. (2002). Modeling the dissolution of settling CaCO3 in the ocean. Global Biogeochemical Cycles, 16(2), 11. Jin, X., Gruber, N., Dunne, J. P., Sarmiento, J. L., & Armstrong, R. A. (2006). Diagnosing the contribution of phytoplankton functional groups to the production and export of particulate organic carbon, CaCO3, and opal from global nutrient and alkalinity distributions. Global Biogeochemical Cycles, 20(2). John, S. G., Liang, H., Weber, T., DeVries, T., Primeau, F., Moore, K., Holzer, M., Mahowald, N., Gardner, W., & Mishonov, A. (2020). AWESOME OCIM: A simple, flexible, and powerful tool for modeling elemental cycling in the oceans. Chemical Geology, 533, 119403. Keir, R. S. (1980). The dissolution kinetics of biogenic calcium carbonates in seawater. Geochimica et Cosmochimica Acta, 44(2), 241–252. Koeve, W., Duteil, O., Oschlies, A., Kähler, P., & Segschneider, J. (2014). Methods to evaluate CaCO 3 cycle modules in coupled global biogeochemical ocean models. Geoscientific Model Development, 7(5), 2393–2408. Lam, V., & Pauly, D. (2005). Mapping the global biomass of mesopelagic fishes. Sea Around Us Project Newsletter, 30(4), 4. Lauvset, S. K., Key, R. M., Olsen, A., Van Heuven, S., Velo, A., Lin, X., Schirnick, C., Kozyr, A., Tanhua, T., & Hoppema, M. (2016). A new global interior ocean mapped climatology: The 1× 1 GLODAP version 2. Earth System Science Data, 8(2), 325–340. Lee, K. (2001). Global net community production estimated from the annual cycle of surface water total dissolved inorganic carbon. Limnology and Oceanography, 46(6), 1287–1297. Manno, C., Sandrini, S., Tositti, L., & Accornero, A. (2007). First stages of degradation of Limacina helicina shells observed above the aragonite chemical lysocline in Terra Nova Bay (Antarctica). Journal of Marine Systems, 68(1–2), 91–102. Martiny, A. C., Pham, C. T. A., Primeau, F. W., Vrugt, J. A., Moore, J. K., Levin, S. A., & Lomas, M. W. (2013). Strong latitudinal patterns in the elemental ratios of marine plankton and organic matter. Nature Geoscience, 6(4), 279–283. Middelburg, J. J., Soetaert, K., & Hagens, M. (2020). Ocean alkalinity, buffering and biogeochemical processes. Reviews of Geophysics, 58(3), e2019RG000681. 139 Milliman, J. D., & Droxler, A. (1996). Neritic and pelagic carbonate sedimentation in the marine environment: ignorance is not bliss. Geologische Rundschau, 85(3), 496–504. Milliman, J. D., Troy, P. J., Balch, W. M., Adams, A. K., Li, Y.-H., & Mackenzie, F. T. (1999). Biologically mediated dissolution of calcium carbonate above the chemical lysocline? Deep Sea Research Part I: Oceanographic Research Papers, 46(10), 1653–1669. Morse, J. W., de Kanel, J., & Harris, K. (1979). Dissolution kinetics of calcium carbonate in seawater; VII, The dissolution kinetics of synthetic aragonite and pteropod tests. American Journal of Science, 279(5), 488–502. Naviaux, J. D., Subhas, A. V, Dong, S., Rollins, N. E., Liu, X., Byrne, R. H., Berelson, W. M., & Adkins, J. F. (2019). Calcite dissolution rates in seawater: Lab vs. in-situ measurements and inhibition by organic matter. Marine Chemistry, 103684. Naviaux, J. D., Subhas, A. V, Rollins, N. E., Dong, S., Berelson, W. M., & Adkins, J. F. (2019). Temperature dependence of calcite dissolution kinetics in seawater. Geochimica et Cosmochimica Acta, 246, 363–384. O’Brien, C. J., Peloquin, J. A., Vogt, M., Heinle, M., Gruber, N., Ajani, P., Andruleit, H., Arístegui, J., Beaufort, L., & Estrada, M. (2013). Global marine plankton functional type biomass distributions: coccolithophores. Earth System Science Data, 5(2), 259–276. O’Mara, N. A., & Dunne, J. P. (2019). Hot spots of carbon and alkalinity cycling in the coastal oceans. Scientific Reports, 9(1), 1–8. Oakes, R. L., Peck, V. L., Manno, C., & Bralower, T. J. (2019). Degradation of internal organic matter is the main control on pteropod shell dissolution after death. Global Biogeochemical Cycles, 33(6), 749–760. Oakes, R. L., & Sessa, J. A. (2020). Determining how biotic and abiotic variables affect the shell condition and parameters of Heliconoides inflatus pteropods from a sediment trap in the Cariaco Basin. Biogeosciences, 17(7), 1975–1990. Perry, C. T., Salter, M. A., Harborne, A. R., Crowley, S. F., Jelks, H. L., & Wilson, R. W. (2011). Fish as major carbonate mud producers and missing components of the tropical carbonate factory. Proceedings of the National Academy of Sciences, 108(10), 3865–3869. Ridgwell, A., Hargreaves, J. C., Edwards, N. R., Annan, J. D., Lenton, T. M., Marsh, R., Yool, A., & Watson, A. (2007). Marine geochemical data assimilation in an efficient Earth System Model of global biogeochemical cycling. Biogeosciences, 4(1), 87–104. Sabine, C. L., Key, R. M., Feely, R. A., & Greeley, D. (2002). Inorganic carbon in the Indian Ocean: Distribution and dissolution processes. Global Biogeochemical Cycles, 16(4), 11–15. 140 Salter, M. A., Harborne, A. R., Perry, C. T., & Wilson, R. W. (2017). Phase heterogeneity in carbonate production by marine fish influences their roles in sediment generation and the inorganic carbon cycle. Scientific Reports, 7(1), 1–15. Salter, M. A., Perry, C. T., & Smith, A. M. (2019). Calcium carbonate production by fish in temperate marine environments. Limnology and Oceanography, 64(6), 2755–2770. Salter, M. A., Perry, C. T., Stuart-Smith, R. D., Edgar, G. J., Wilson, R. W., & Harborne, A. R. (2018). Reef fish carbonate production assessments highlight regional variation in sedimentary significance. Geology, 46(8), 699–702. Sarmiento, J. L., Dunne, J., Gnanadesikan, A., Key, R. M., Matsumoto, K., & Slater, R. (2002). A new estimate of the CaCO3 to organic carbon export ratio. Global Biogeochemical Cycles, 16(4), 51–54. Sarmiento, J. L., & Gruber, N. (2006). Ocean Biogeochemical Dynamics. Princeton University Press. Schiebel, R. (2002). Planktic foraminiferal sedimentation and the marine calcite budget. Global Biogeochemical Cycles, 16(4), 1–3. Schiebel, R., & Hemleben, C. (2017). Planktic foraminifers in the modern ocean. Springer. Schiebel, R., & Movellan, A. (2012). First-order estimate of the planktic foraminifer biomass in the modern ocean. Earth System Science Data, 4(1), 75–89. Smith, C. R., Hoover, D. J., Doan, S. E., Pope, R. H., Demaster, D. J., Dobbs, F. C., & Altabet, M. A. (1996). Phytodetritus at the abyssal seafloor across 10 of latitude in the central equatorial Pacific. Deep Sea Research Part II: Topical Studies in Oceanography, 43(4–6), 1309–1338. Subhas, A. V, Dong, S., Naviaux, J. D., Rollins, N. E., Ziveri, P., Gray, W., Rae, J. W. B., Liu, X., Byrne, R. H., & Chen, S. (2022). Shallow Calcium Carbonate Cycling in the North Pacific Ocean. Global Biogeochemical Cycles, e2022GB007388. Subhas, A. V, Rollins, N. E., Berelson, W. M., Erez, J., Ziveri, P., Langer, G., & Adkins, J. F. (2018). The dissolution behavior of biogenic calcites in seawater and a possible role for magnesium and organic carbon. Marine Chemistry, 205, 100–112. Sulpis, O., Jeansson, E., Dinauer, A., Lauvset, S. K., & Middelburg, J. J. (2021). Calcium carbonate dissolution patterns in the ocean. Nature Geoscience, 14(6), 423–428. Sverdrup, H. U., Johnson, N. W., & Fleming, R. H. (1941). The Oceans. Prentice Hall. Walker, M., Hammel, J. U., Wilde, F., Hoehfurtner, T., Humphries, S., & Schuech, R. (2021). Estimation of sinking velocity using free-falling dynamically scaled models: Foraminifera as a test case. Journal of Experimental Biology, 224(2), jeb230961. 141 Weber, T., John, S., Tagliabue, A., & DeVries, T. (2018). Biological uptake and reversible scavenging of zinc in the global ocean. Science, 361(6397), 72 LP – 76. Wilson, R. W., Millero, F. J., Taylor, J. R., Walsh, P. J., Christensen, V., Jennings, S., & Grosell, M. (2009). Contribution of fish to the marine inorganic carbon cycle. Science, 323(5912), 359–362. Wolf-Gladrow, D. A., Zeebe, R. E., Klaas, C., Körtzinger, A., & Dickson, A. G. (2007). Total alkalinity: The explicit conservative expression and its application to biogeochemical processes. Marine Chemistry, 106(1–2), 287–300. Woosley, R. J., Millero, F. J., & Grosell, M. (2012). The solubility of fish‐produced high magnesium calcite in seawater. Journal of Geophysical Research: Oceans, 117(C4). Zaric, S., Schulz, M., & Mulitza, S. (2006). Global prediction of planktonic foraminiferal fluxes from hydrographic and productivity data. Biogeosciences, 3, 187–207. https://doi.org/10.5194/bg-3-187-2006 Zhang, H., Stoll, H., Bolton, C., Jin, X., & Liu, C. (2018). A refinement of coccolith separation methods: measuring the sinking characteristics of coccoliths. Biogeosciences, 15(15), 4759– 4775. Ziveri, P., Gray, W., Ortiz, G., Manno, C., Grelaud, M., Incarbona, A., Rae, J., Subhas, A., Pallacks, S., White, A., Adkins, J., & Berelson, W. (2022). Pelagic carbonate production in the North Pacific Ocean. In Revision at Nature Communications. 142 4.8 Supplemental Material Table 4.S1. Optimized parameter values for the four base models Model Benthic only (M1) Ω-dependent (M2) Ω + respiration (M3) Constant (M4) Optimized parameters export, mean TA export, mean TA export, mean TA, x arag , m arag , x calc , m calc export, mean TA, R const CaCO 3 export scaling factor 0.30 0.31 0.80 0.49 Mean TA (TA = c × salinity / 35) 2389.2 2390.3 2394.8 2395.4 x arag , x calc 0.04, 4.57 m arag , m calc 0.00032, 0.97 R const 0.032 143 Figure 4.S1. Data inputs for deriving CaCO 3 export: P uptake (a) × C:P ratio (b) × PIC:POC ratio (c) = CaCO 3 export (d). In addition, a scaling factor was optimized for each model to adjust total CaCO 3 export. 144 Figure 4.S2. Observed and modeled TA0 and TA* for representative transects (Pacific at 161°W and Atlantic at 29°W). Observed (a, b); M1, benthic only model (c, d); M2, Ω-dependent model (e, f); M3, Ω + respiration model (g, h); M4, constant dissolution model (i, j). The transects for M3 (g, h) are shown in the main text in Fig. 3. 145 Figure 4.S3. Observed and modeled TA0 and TA*, and model-observation mismatch for the export sensitivity tests for M1 (benthic only model). Tests were run for low (0.5 Gt C y -1 ), mid (1.1 Gt C y -1 ), and high (1.8 Gt C y -1 ) export scenarios (export from 73 m). (a, b) Observed TA0 and TA*, respectively. (c, g, k) Modeled TA0. (d, h, l) TA0 model-observation mismatch. (e, i, m) Modeled TA*. (f, j, n) TA* model-observation mismatch. 146 Figure 4.S4. Observed and modeled TA0 and TA*, and model-observation mismatch for the export sensitivity tests for M2 (Ω-dependent model). Tests were run for low (0.5 Gt C y -1 ), mid (1.1 Gt C y -1 ), and high (1.8 Gt C y -1 ) export scenarios (export from 73 m). (a, b) Observed TA0 and TA*, respectively. (c, g, k) Modeled TA0. (d, h, l) TA0 model-observation mismatch. (e, i, m) Modeled TA*. (f, j, n) TA* model-observation mismatch. 147 Figure 4.S5. Observed and modeled TA0 and TA*, and model-observation mismatch for the export sensitivity tests for M3 (Ω + respiration model). Tests were run for low (0.5 Gt C y -1 ), mid (1.1 Gt C y -1 ), and high (1.8 Gt C y -1 ) export scenarios (export from 73 m). (a, b) Observed TA0 and TA*, respectively. (c, g, k) Modeled TA0. (d, h, l) TA0 model-observation mismatch. (e, i, m) Modeled TA*. (f, j, n) TA* model-observation mismatch. 148 Figure 4.S6. Observed and modeled TA0 and TA*, and model-observation mismatch for the export sensitivity tests for M4 (constant dissolution model). Tests were run for low (0.5 Gt C y -1 ), mid (1.1 Gt C y -1 ), and high (1.8 Gt C y -1 ) export scenarios (export from 73 m). (a, b) Observed TA0 and TA*, respectively. (c, g, k) Modeled TA0. (d, h, l) TA0 model-observation mismatch. (e, i, m) Modeled TA*. (f, j, n) TA* model-observation mismatch. 149 Figure 4.S7. Model-observation mismatch for the sinking speed sensitivity tests for M2 (Ω-dependent model). Differences between modeled and observed (model – observed) are shown for both TA0 and TA*, at a range of sinking speeds: 1 m d -1 (a, b); 10 m d -1 (c, d); 100 m d -1 , the default sinking speed in the base models (e, f); and 1,000 m d -1 (g, h). 150 Figure 4.S8. Model-observation mismatch for the sinking speed sensitivity tests for M3 (Ω + respiration model). Differences between modeled and observed (model – observed) are shown for both TA0 and TA*, for sinking speeds above and below the default sinking speed: 50 m d -1 (a, b); 100 m d -1 , the default sinking speed in the base models (c, d); and 200 m d -1 (e, f). 151 Figure 4.S9. Model-observation mismatch for the combined sinking speed and calcite/aragonite ratio sensitivity tests for M2 (Ω-dependent model). Differences between modeled and observed (model – observed) are shown TA*, at a range of sinking speeds: 1 m d -1 (a, b, c); 10 m d -1 (d, e, f); 100 m d -1 , the default sinking speed in the base models (g, h, i); and 150 m d -1 (j, k, l). The center column uses the spatially variable calcite/aragonite ratio from the GFDL dataset, as used in all base models. 152 5. Conclusions Scientific progress takes the form of individual observations, built into detailed studies, that feed larger bodies of knowledge within often disjointed fields. While individual scientific findings are often used to maximize our health and happiness, science is also motivated by a fundamental curiosity about the universe. The 18 th -19 th century explorer, and arguable founder of ecology, Alexander von Humboldt, described this distillation of scientific fact into something more profound: “An edifice cannot produce a striking effect until the scaffolding is removed, that had of necessity been used during is erection.” Within this context, I won’t reiterate here the data and individual conclusions which scaffold the larger theme of this dissertation; these are already summarized in each chapter and in the Abstract. While each study is—I hope—interesting to the reader in its own right, the picture they collectively provide may indeed be more striking. This dissertation ostensibly focuses on a geologic topic – rates of alkalinity generation by the dissolution of minerals – however all three studies ultimately point to the importance of biology in driving dissolution. In both laboratory and natural settings, the dissolution rates of olivine (Ch 2) and CaCO3 (Ch 3 and 4) were faster than expected given measurements of environmental conditions and known kinetics. In sediments (Ch 3), CaCO3 dissolution in supersaturated water was causally linked to its co-location with aerobic respiration of OM in porewater. Additional research conducted in support of this dissertation (data not shown) revealed porewater also contained significant carbonic anhydrase, an enzyme also known to facilitate dissolution. In the water column (Ch 4), this same mechanism –the coupling of minerals with aerobic microbial respiration—is a good candidate to explain global ocean alkalinity patterns. Considering olivine (Ch 2), dissolution was enhanced not by respiratory CO2, but by microbially-produced molecules 153 (siderophores, potentially in concert with other molecules) interacting with the mineral surface to lower the activation energy of dissolution. Of course, none of these findings suggest that biology is doing anything magical, or bypassing the laws of thermodynamics. Rather, microbes realize faster dissolution kinetics either by lowering activation energies of rate-limiting steps, or by moving the mineral-water system closer to thermodynamic equilibrium at microscales that evade “environmental” measurements (e.g., within organic particles or porewater). The energy for these microbial tricks ultimately comes from solar radiation and its transformation into high energy electrons by photosynthesis. Although solar radiation is the obvious source of climate warmth, it is also the major driver of CO2 drawdown by weathering: solar radiation fuels the hydrologic cycle that weathers exposed minerals in terrestrial settings. When photosynthetic life evolved more than three billion years ago, it conspired with the sun to further extract atmospheric CO2 and store it in organic matter. The chapters of this dissertation provide examples of additional mechanisms by which solar radiation, translated into organic matter, draw CO2 out of the atmosphere by enhancing rates of mineral dissolution. This microbe- mineral weathering phenomena is interesting to consider in the context of the Gaia hypothesis, which is, albeit controversial and arguably untestable, theoretically useful (Lovelock & Margulis, 1974). In this context, biological enhancement of weathering is a contributing but underappreciated process by which life maintains planetary homeostasis (Kleidon, 2004). To end on a more practical note, I return to climate change, the topic which opened this dissertation, and has framed each subsequent chapter. Climate change is an existential threat to global civilization, as evidenced by recent ecological disruptions and related political upheaval (Bellard et al., 2012; Hsiang et al., 2013). From the starting position that civilization and human 154 and other life are valued at all, we must reverse climate change, and it is increasingly clear that pulling CO2 from the atmosphere is “unavoidable” (IPCC, 2022). Abiotic approaches to CO2 removal require high energy inputs to pull and concentrate CO2 from air or water using electricity and/or heat; these are all currently very expensive. As an alternative, taking advantage of geologic weathering pathways may be less energy-intensive. Given geologically slow rates, however, researchers must find ways to expedite natural processes via “enhanced weathering”. This dissertation provides three examples of natural “enhanced weathering”, mediated by microbiological processes, that may serve as useful models for developing engineered solutions. Chapter 1 illustrates how microbes, and/or specifically their molecular products like siderophores, are effective catalysts for olivine dissolution. The field of microbially-enhanced mineral weathering is in its infancy, and mechanistic information like this is critical to push the field toward effective solutions. Chapters 2 and 3 both illustrate how the coupling of aerobic organic matter respiration with CaCO3 efficiently stimulates dissolution in initially oversaturated seawater. Via the physical co-location of CaCO3 and organic matter, either due to the physiology of planktonic calcifiers (Ch 3), and/or to gravitational coalescence on the seafloor (Ch 2), CaCO3 dissolves significantly in seawater where it otherwise would not. This simple concentration of basic minerals in naturally acidic environments (e.g., sands, oceanic upwelling regions, or tropical soils) may be a simple strategy to increase dissolution rates and alkalinity generation. The scale of the problem is huge: there are now nearly 600 Gt of excess C in the atmosphere, and that number grows annually (Friedlingstein et al., 2022). Avoiding further climate change will require removing at least 3 Gt C y -1 by 2050, small in comparison to the total, but huge by our current capabilities (NASEM, 2019). Natural mechanisms such as those explored here, potentially deployed in engineered systems, may be useful in reaching that goal. 155 References Bellard, C., Bertelsmeier, C., Leadley, P., Thuiller, W., & Courchamp, F. (2012). Impacts of climate change on the future of biodiversity. Ecology Letters, 15(4), 365–377. Friedlingstein, P., Jones, M. W., O’Sullivan, M., Andrew, R. M., Bakker, D. C. E., Hauck, J., Le Quéré, C., Peters, G. P., Peters, W., & Pongratz, J. (2022). Global carbon budget 2021. Earth System Science Data, 14(4), 1917–2005. Hsiang, S. M., Burke, M., & Miguel, E. (2013). Quantifying the influence of climate on human conflict. Science, 341(6151), 1235367. IPCC. (2022). Climate Change 2022: Mitigation of Climate Change. Contribution of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. https://doi.org/10.1017/9781009157926 Kleidon, A. (2004). Beyond Gaia: Thermodynamics of life and Earth system functioning. Climatic Change, 66(3), 271–319. Lovelock, J., & Margulis, L. (1974). Atmospheric homeostasis by and for the biosphere: the Gaia hypothesis. Tellus, 26(1–2), 2–10. NASEM. (2019). Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. The National Academies Press. https://doi.org/10.17226/25259 156 References Alldredge, A. L., & Cohen, Y. (1987). Can microscale chemical patches persist in the sea? Microelectrode study of marine snow, fecal pellets. Science, 235(4789), 689–691. Amiotte-Suchet, P., Probst, J., & Ludwig, W. (2003). Worldwide distribution of continental rock lithology: Implications for the atmospheric/soil CO2 uptake by continental weathering and alkalinity river transport to the oceans. Global Biogeochemical Cycles, 17(2). Amos, C. M., Castelao, R. M., & Medeiros, P. M. (2019). Offshore transport of particulate organic carbon in the California Current System by mesoscale eddies. Nature Communications, 10(1), 1–8. Andersson, A J, & Mackenzie, F. T. (2012). Revisiting four scientific debates in ocean acidification research. Biogeosciences, 9(3), 893–905. Andersson, Andreas J. (2015). A fundamental paradigm for coral reef carbonate sediment dissolution. Frontiers in Marine Science, 2(July), 52. https://doi.org/10.3389/fmars.2015.00052 Andersson, Andreas J., MacKenzie, F. T., & Lerman, A. (2005). Coastal ocean and carbonate systems in the high CO2 world of the anthropocene. American Journal of Science, 305(9), 875–918. https://doi.org/10.2475/ajs.305.9.875 Andersson, Andreas J, Bates, N. R., & Mackenzie, F. T. (2007). Dissolution of carbonate sediments under rising pCO 2 and ocean acidification: observations from Devil’s Hole, Bermuda. Aquatic Geochemistry, 13(3), 237–264. Andersson, Andreas J, & Gledhill, D. (2013). Ocean acidification and coral reefs: effects on breakdown, dissolution, and net ecosystem calcification. Annual Review of Marine Science, 5, 321–348. Andersson, Andreas J, & Mackenzie, F. T. (2004). Shallow‐water oceans: a source or sink of atmospheric CO2? Frontiers in Ecology and the Environment, 2(7), 348–353. Anglada-Ortiz, G., Zamelczyk, K., Meilland, J., Ziveri, P., Chierici, M., Fransson, A., & Rasmussen, T. L. (2021). Planktic foraminiferal and pteropod contributions to carbon dynamics in the Arctic Ocean (north Svalbard margin). Frontiers in Marine Science, 8, 636. Aouad, G., Crovisier, J.-L., Geoffroy, V. A., Meyer, J.-M., & Stille, P. (2006). Microbially- mediated glass dissolution and sorption of metals by Pseudomonas aeruginosa cells and biofilm. Journal of Hazardous Materials, 136(3), 889–895. Archer, D., Kheshgi, H., & Maier‐Reimer, E. (1998). Dynamics of fossil fuel CO2 neutralization by marine CaCO3. Global Biogeochemical Cycles, 12(2), 259–276. 157 Archer, D., & Maier‐Reimer, E. (1994). Effect of deep-sea sedimentary calcite preservation on atmospheric CO 2 concentration. Nature, 367(6460), 260–263. Arrhenius, S. (1897). On the Influence of Carbonic Acid in the Air upon the Temperature of the Earth. Publications of the Astronomical Society of the Pacific, 9, 14. https://doi.org/10.1086/121158 Balch, W. M., Gordon, H. R., Bowler, B. C., Drapeau, D. T., & Booth, E. S. (2005). Calcium carbonate measurements in the surface global ocean based on Moderate‐Resolution Imaging Spectroradiometer data. Journal of Geophysical Research: Oceans, 110(C7). Battaglia, G., Steinacher, M., & Joos, F. (2016). A probabilistic assessment of calcium carbonate export and dissolution in the modern ocean. Biogeosciences, 13(9), 2823–2848. Bednaršek, N., Možina, J., Vogt, M., O’Brien, C., & Tarling, G. A. (2012). The global distribution of pteropods and their contribution to carbonate and carbon biomass in the modern ocean. Earth System Science Data, 4(1), 167–186. Beerling, D. J., Taylor, L. L., Bradshaw, C. D. C., Lunt, D. J., Valdes, P. J., Banwart, S. A., Pagani, M., & Leake, J. R. (2012). Ecosystem CO2 starvation and terrestrial silicate weathering: mechanisms and global‐scale quantification during the late Miocene. Journal of Ecology, 100(1), 31–41. Bellard, C., Bertelsmeier, C., Leadley, P., Thuiller, W., & Courchamp, F. (2012). Impacts of climate change on the future of biodiversity. Ecology Letters, 15(4), 365–377. Berelson, W. M., Balch, W. M., Najjar, R., Feely, R. A., Sabine, C., & Lee, K. (2007). Relating estimates of CaCO3 production, export, and dissolution in the water column to measurements of CaCO3 rain into sediment traps and dissolution on the sea floor: A revised global carbonate budget. Global Biogeochemical Cycles, 21(1). Bergan, A. J., Lawson, G. L., Maas, A. E., & Wang, Z. A. (2017). The effect of elevated carbon dioxide on the sinking and swimming of the shelled pteropod Limacina retroversa. ICES Journal of Marine Science, 74(7), 1893–1905. Berner, R. A. (2010). Comment: effect of organic ligands and heterotrophic bacteria on Wollastonite dissolution kinetics. American Journal of Science, 310(5), 424. Berner, R. A. (2012). Jacques-Joseph Ébelmen, the founder of earth system science. Comptes Rendus Geoscience, 344(11), 544–548. https://doi.org/https://doi.org/10.1016/j.crte.2012.08.001 Berner, R. A., & Caldeira, K. (1997). The need for mass balance and feedback in the geochemical carbon cycle. Geology, 25(10), 955–956. Berner, R. A., & Honjo, S. (1981). Pelagic sedimentation of aragonite: its geochemical significance. Science, 211(4485), 940–942. 158 Berner, R. A., Lasaga, A. C., & Garrels, R. M. (1983). Carbonate-silicate geochemical cycle and its effect on atmospheric carbon dioxide over the past 100 million years. Am. J. Sci.;(United States), 283(7). Biermann, A., & Engel, A. (2010). Effect of CO 2 on the properties and sinking velocity of aggregates of the coccolithophore Emiliania huxleyi. Biogeosciences, 7(3), 1017–1029. Boucher, G., Clavier, J., Hily, C., & Gattuso, J.-P. (1998). Contribution of soft-bottoms to the community metabolism (primary production and calcification) of a barrier reef flat (Moorea, French Polynesia). Journal of Experimental Marine Biology and Ecology, 225(2), 269–283. Boudreau, B. P., Huettel, M., Forster, S., Jahnke, R., McLachlan, A., Middelburg, J. J., Nielsen, P., Sansone, F., Taghon, G., & Van Raaphorst, W. (2001). Permeable marine sediments: overturning an old paradigm. EOS, Transactions American Geophysical Union, 82(11), 133– 136. Bourke, M. F., Marriott, P. J., Glud, R. N., Hasler-Sheetal, H., Kamalanathan, M., Beardall, J., Greening, C., & Cook, P. L. M. (2017). Metabolism in anoxic permeable sediments is dominated by eukaryotic dark fermentation. Nature Geoscience, 10(1), 30–35. https://doi.org/10.1038/ngeo2843 Brady, P. V, & Carroll, S. A. (1994). Direct effects of CO2 and temperature on silicate weathering: Possible implications for climate control. Geochimica et Cosmochimica Acta, 58(7), 1853– 1856. Brady, P. V, Dorn, R. I., Brazel, A. J., Clark, J., Moore, R. B., & Glidewell, T. (1999). Direct measurement of the combined effects of lichen, rainfall, and temperature onsilicate weathering. Geochimica et Cosmochimica Acta, 63(19–20), 3293–3300. Bray, A. W., Oelkers, E. H., Bonneville, S., Wolff-Boenisch, D., Potts, N. J., Fones, G., & Benning, L. G. (2015). The effect of pH, grain size, and organic ligands on biotite weathering rates. Geochimica et Cosmochimica Acta, 164, 127–145. Brenner, H., Braeckman, U., Le Guitton, M., & Meysman, F. J. R. (2016). The impact of sedimentary alkalinity release on the water column CO2system in the North Sea. Biogeosciences, 13(3), 841–863. https://doi.org/10.5194/bg-13-841-2016 Brewer, P. G., Wong, G. T. F., Bacon, M. P., & Spencer, D. W. (1975). An oceanic calcium problem? Earth and Planetary Science Letters, 26(1), 81–87. Broecker, W. S., & Peng, T. (1987). The role of CaCO3 compensation in the glacial to interglacial atmospheric CO2 change. Global Biogeochemical Cycles, 1(1), 15–29. Broecker, W. S., & Sanyal, A. (1998). Does atmospheric CO2 police the rate of chemical weathering? Global Biogeochemical Cycles, 12(3), 403–408. 159 Bühring, S. I., Ehrenhauss, S., Kamp, A., Moodley, L., & Witte, U. (2006). Enhanced benthic activity in sandy sublittoral sediments: evidence from 13C tracer experiments. Marine Biology Research, 2(2), 120–129. Buitenhuis, E. T., Le Quere, C., Bednaršek, N., & Schiebel, R. (2019). Large contribution of pteropods to shallow CaCO3 export. Global Biogeochemical Cycles, 33(3), 458–468. Buitenhuis, E. T., Vogt, M., Moriarty, R., Bednaršek, N., Doney, S. C., Leblanc, K., Le Quéré, C., Luo, Y.-W., O’Brien, C., & O’Brien, T. (2013). MAREDAT: towards a world atlas of MARine Ecosystem DATa. Earth System Science Data, 5(2), 227–239. Buss, H. L., Lüttge, A., & Brantley, S. L. (2007). Etch pit formation on iron silicate surfaces during siderophore-promoted dissolution. Chemical Geology, 240(3–4), 326–342. Cael, B. B., Cavan, E. L., & Britten, G. L. (2021). Reconciling the size‐dependence of marine particle sinking speed. Geophysical Research Letters, 48(5), e2020GL091771. Cai, W. J. (2011). Estuarine and coastal ocean carbon paradox: CO2 sinks or sites of terrestrial carbon incineration? Annual Review of Marine Science, 3, 123–145. Caldeira, K., & Wickett, M. E. (2005). Ocean model predictions of chemistry changes from carbon dioxide emissions to the atmosphere and ocean. Journal of Geophysical Research: Oceans, 110(C9). Carleton, T. A., & Hsiang, S. M. (2016). Social and economic impacts of climate. Science, 353(6304), aad9837. Carrasco, N., Kretzschmar, R., Pesch, M.-L., & Kraemer, S. M. (2007). Low concentrations of surfactants enhance siderophore-promoted dissolution of goethite. Environmental Science & Technology, 41(10), 3633–3638. Carter, B. R., Feely, R. A., Lauvset, S. K., Olsen, A., DeVries, T., & Sonnerup, R. (2021). Preformed properties for marine organic matter and carbonate mineral cycling quantification. Global Biogeochemical Cycles, 35(1), e2020GB006623. Carter, B. R., Toggweiler, J. R., Key, R. M., & Sarmiento, J. L. (2014). Processes determining the marine alkalinity and calcium carbonate saturation state distributions. Biogeosciences, 11(24), 7349–7362. Cervini-Silva, J., & Sposito, G. (2002). Steady-state dissolution kinetics of aluminum-goethite in the presence of desferrioxamine-B and oxalate ligands. Environmental Science & Technology, 36(3), 337–342. Cheah, S.-F., Kraemer, S. M., Cervini-Silva, J., & Sposito, G. (2003). Steady-state dissolution kinetics of goethite in the presence of desferrioxamine B and oxalate ligands: implications for the microbial acquisition of iron. Chemical Geology, 198(1–2), 63–75. 160 Chisholm, J. R. M., & Gattuso, J. (1991). Validation of the alkalinity anomaly technique for investigating calcification of photosynthesis in coral reef communities. Limnology and Oceanography, 36(6), 1232–1239. Chung, S., Lee, K., Feely, R. A., Sabine, C. L., Millero, F. J., Wanninkhof, R., Bullister, J. L., Key, R. M., & Peng, T. (2003). Calcium carbonate budget in the Atlantic Ocean based on water column inorganic carbon chemistry. Global Biogeochemical Cycles, 17(4). Cocozza, C., Tsao, C. C. G., Cheah, S.-F., Kraemer, S. M., Raymond, K. N., Miano, T. M., & Sposito, G. (2002). Temperature dependence of goethite dissolution promoted by trihydroxamate siderophores. Geochimica et Cosmochimica Acta, 66(3), 431–438. Colbourn, G., Ridgwell, A., & Lenton, T. M. (2015). The time scale of the silicate weathering negative feedback on atmospheric CO2. Global Biogeochemical Cycles, 29(5), 583–596. Compere Jr, E. L., & Bates, J. M. (1973). Determination of calcite: aragonite ratios in mollusc shells by infrared spectra 1. Limnology and Oceanography, 18(2), 326–331. Cyronak, T., & Eyre, B. D. (2016). The synergistic effects of ocean acidification and organic metabolism on calcium carbonate (CaCO3) dissolution in coral reef sediments. Marine Chemistry, 183, 1–12. Cyronak, T., Santos, I. R., & Eyre, B. D. (2013). Permeable coral reef sediment dissolution driven by elevated pCO2 and pore water advection. Geophysical Research Letters, 40(18), 4876– 4881. Cyronak, T., Santos, I. R., McMahon, A., & Eyre, B. D. (2013). Carbon cycling hysteresis in permeable carbonate sands over a diel cycle: Implications for ocean acidification. Limnology and Oceanography, 58(1), 131–143. Dasgupta, R. (2013). Ingassing, storage, and outgassing of terrestrial carbon through geologic time. Reviews in Mineralogy and Geochemistry, 75(1), 183–229. Dehner, C. A., Awaya, J. D., Maurice, P. A., & DuBois, J. L. (2010). Roles of siderophores, oxalate, and ascorbate in mobilization of iron from hematite by the aerobic bacterium Pseudomonas mendocina. Applied and Environmental Microbiology, 76(7), 2041–2048. Dessert, C., Dupré, B., Gaillardet, J., François, L. M., & Allegre, C. J. (2003). Basalt weathering laws and the impact of basalt weathering on the global carbon cycle. Chemical Geology, 202(3–4), 257–273. DeVries, T. (2014). The oceanic anthropogenic CO2 sink: Storage, air‐sea fluxes, and transports over the industrial era. Global Biogeochemical Cycles, 28(7), 631–647. DeVries, T., & Primeau, F. (2011). Dynamically and observationally constrained estimates of water-mass distributions and ages in the global ocean. Journal of Physical Oceanography, 41(12), 2381–2401. 161 Dhungana, S., & Crumbliss, A. L. (2005). Coordination chemistry and redox processes in siderophore-mediated iron transport. Geomicrobiology Journal, 22(3–4), 87–98. Dong, S., Berelson, W. M., Adkins, J. F., Rollins, N. E., Naviaux, J. D., Pirbadian, S., El-Naggar, M. Y., & Teng, H. H. (2020). An atomic force microscopy study of calcite dissolution in seawater. Geochimica et Cosmochimica Acta, 283, 40–53. Dong, S., Berelson, W. M., Rollins, N. E., Subhas, A. V, Naviaux, J. D., Celestian, A. J., Liu, X., Turaga, N., Kemnitz, N. J., & Byrne, R. H. (2019). Aragonite dissolution kinetics and calcite/aragonite ratios in sinking and suspended particles in the North Pacific. Earth and Planetary Science Letters, 515, 1–12. Dong, S., Subhas, A. V, Rollins, N. E., Naviaux, J. D., Adkins, J. F., & Berelson, W. M. (2018). A kinetic pressure effect on calcite dissolution in seawater. Geochimica et Cosmochimica Acta, 238, 411–423. Dorn, R. I. (2014). Ants as a powerful biotic agent of olivine and plagioclase dissolution. Geology, 42(9), 771–774. Drever, J. I., & Stillings, L. L. (1997). The role of organic acids in mineral weathering. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 120(1–3), 167–181. Dunne, J. P., Hales, B., & Toggweiler, J. R. (2012). Global calcite cycling constrained by sediment preservation controls. Global Biogeochemical Cycles, 26(3). Dunne, J. P., John, J. G., Shevliakova, E., Stouffer, R. J., Krasting, J. P., Malyshev, S. L., Milly, P. C. D., Sentman, L. T., Adcroft, A. J., & Cooke, W. (2013). GFDL’s ESM2 global coupled climate–carbon earth system models. Part II: carbon system formulation and baseline simulation characteristics. Journal of Climate, 26(7), 2247–2267. Dunne, J. P., Sarmiento, J. L., & Gnanadesikan, A. (2007). A synthesis of global particle export from the surface ocean and cycling through the ocean interior and on the seafloor. Global Biogeochemical Cycles, 21(4). Ebelmen, J.-J. (1845). Sur les produits de la décomposition des espèces minérales de la famille des silicates. Annales Des Mines, 7(3), 66. Emerson, S., & Bender, M. L. (1981). Carbon fluxes at the sediment-water interface: Calcium carbonate preservation. Journal of Marine Research, 39, 139–162. Emery, K. O. (1968). Relict sediments on continental shelves of world. AAPG Bulletin, 52(3), 445–464. Engel, A., Szlosek, J., Abramson, L., Liu, Z., & Lee, C. (2009). Investigating the effect of ballasting by CaCO3 in Emiliania huxleyi: I. Formation, settling velocities and physical properties of aggregates. Deep Sea Research Part II: Topical Studies in Oceanography, 56(18), 1396–1407. 162 Eyre, B. D., Andersson, A. J., & Cyronak, T. (2014). Benthic coral reef calcium carbonate dissolution in an acidifying ocean. Nature Climate Change, 4(11), 969. https://doi.org/10.1038/nclimate2380 Eyre, B. D., Cyronak, T., Drupp, P., De Carlo, E. H., Sachs, J. P., & Andersson, A. J. (2018). Coral reefs will transition to net dissolving before end of century. Science, 359(6378), 908–911. Fabry, V. J. (1990). Shell growth rates of pteropod and heteropod molluscs and aragonite production in the open ocean: implications for the marine carbonate system. Journal of Marine Research, 48(1), 209–222. Fabry, V. J., & Deuser, W. G. (1991). Aragonite and magnesian calcite fluxes to the deep Sargasso Sea. Deep Sea Research Part A. Oceanographic Research Papers, 38(6), 713–728. Feely, R. A., Okazaki, R. R., Cai, W.-J., Bednaršek, N., Alin, S. R., Byrne, R. H., & Fassbender, A. (2018). The combined effects of acidification and hypoxia on pH and aragonite saturation in the coastal waters of the California current ecosystem and the northern Gulf of Mexico. Continental Shelf Research, 152, 50–60. Feely, R. A., Sabine, C. L., Lee, K., Berelson, W., Kleypas, J., Fabry, V. J., & Millero, F. J. (2004). Impact of anthropogenic CO2 on the CaCO3 system in the oceans. Science, 305(5682), 362– 366. https://doi.org/10.1126/science.1097329 Feely, R. A., Sabine, C. L., Lee, K., Millero, F. J., Lamb, M. F., Greeley, D., Bullister, J. L., Key, R. M., Peng, T., & Kozyr, A. (2002). In situ calcium carbonate dissolution in the Pacific Ocean. Global Biogeochemical Cycles, 16(4), 91. Fennessey, C. M., Jones, M. E., Taillefert, M., & DiChristina, T. J. (2010). Siderophores are not involved in Fe (III) solubilization during anaerobic Fe (III) respiration by Shewanella oneidensis MR-1. Applied and Environmental Microbiology, 76(8), 2425–2432. Ferret, C., Sterckeman, T., Cornu, J., Gangloff, S., Schalk, I. J., & Geoffroy, V. A. (2014). Siderophore‐promoted dissolution of smectite by fluorescent P seudomonas. Environmental Microbiology Reports, 6(5), 459–467. Fink, A., den Haan, J., Chennu, A., Uthicke, S., & de Beer, D. (2017). Ocean Acidification Changes Abiotic Processes but Not Biotic Processes in Coral Reef Sediments. Frontiers in Marine Science, 4, 73. Foote, E. (1856). Circumstances affecting the Heat of the Sun’s Rays. The American Journal of Science and Arts, 22, 382–383. Franke, U., Polerecky, L., Precht, E., & Huettel, M. (2006). Wave tank study of particulate organic matter degradation in permeable sediments. Limnology and Oceanography, 51(2), 1084– 1096. https://doi.org/10.4319/lo.2006.51.2.1084 163 Freiwald, A. (1995). Bacteria-induced carbonate degradation: a taphonomic case study of Cibicides lobatulus from a high-boreal carbonate setting. Palaios, 337–346. Friedlingstein, P., Jones, M. W., O’Sullivan, M., Andrew, R. M., Bakker, D. C. E., Hauck, J., Le Quéré, C., Peters, G. P., Peters, W., & Pongratz, J. (2022). Global carbon budget 2021. Earth System Science Data, 14(4), 1917–2005. Friedlingstein, P., O’Sullivan, M., Jones, M. W., Andrew, R. M., Hauck, J., Olsen, A., Peters, G. P., Peters, W., Pongratz, J., & Sitch, S. (2020). Global carbon budget 2020. Earth System Science Data, 12(4), 3269–3340. Friis, K., Najjar, R. G., Follows, M. J., & Dutkiewicz, S. (2006). Possible overestimation of shallow‐depth calcium carbonate dissolution in the ocean. Global Biogeochemical Cycles, 20(4). Fry, C. H., Tyrrell, T., Hain, M. P., Bates, N. R., & Achterberg, E. P. (2015). Analysis of global surface ocean alkalinity to determine controlling processes. Marine Chemistry, 174, 46–57. Gangstø, R., Gehlen, M., Schneider, B., Bopp, L., Aumont, O., & Joos, F. (2008). Modeling the marine aragonite cycle: changes under rising carbon dioxide and its role in shallow water CaCO 3 dissolution. Biogeosciences, 5(4), 1057–1072. Garcia, B., Lemelle, L., Rose-Koga, E., Perriat, P., Basset, R., Gillet, P., & Albarède, F. (2013). An experimental model approach of biologically-assisted silicate dissolution with olivine and Escherichia coli–Impact on chemical weathering of mafic rocks and atmospheric CO2 drawdown. Applied Geochemistry, 31, 216–227. Gattuso, J.-P., Frankignoulle, M., & Wollast, R. (1998). Carbon and carbonate metabolism in coastal aquatic ecosystems. Annual Review of Ecology and Systematics, 29(1), 405–434. Gattuso, J.-P., Gentili, B., Duarte, C. M., Kleypas, J. A., Middelburg, J. J., & Antoine, D. (2006). Light availability in the coastal ocean: impact on the distribution of benthic photosynthetic organisms and their contribution to primary production. Biogeosciences. Gehlen, M., Gangstø, R., Schneider, B., Bopp, L., Aumont, O., & Éthé, C. (2007). The fate of pelagic CaCO 3 production in a high CO 2 ocean: a model study. Biogeosciences, 4(4), 505– 519. Gerrits, R., Pokharel, R., Breitenbach, R., Radnik, J., Feldmann, I., Schuessler, J. A., Von Blanckenburg, F., Gorbushina, A. A., & Schott, J. (2020). How the rock-inhabiting fungus K. petricola A95 enhances olivine dissolution through attachment. Geochimica et Cosmochimica Acta, 282, 76–97. Gingerich, P. D. (2019). Temporal scaling of carbon emission and accumulation rates: modern anthropogenic emissions compared to estimates of PETM onset accumulation. Paleoceanography and Paleoclimatology, 34(3), 329–335. 164 Gorbushina, A. A., & Palinska, K. A. (1999). Biodeteriorative processes on glass: experimental proof of the role of fungi and cyanobacteria. Aerobiologia, 15(3), 183–192. Gorby, Y. A., Yanina, S., McLean, J. S., Rosso, K. M., Moyles, D., Dohnalkova, A., Beveridge, T. J., Chang, I. S., Kim, B. H., & Kim, K. S. (2006). Electrically conductive bacterial nanowires produced by Shewanella oneidensis strain MR-1 and other microorganisms. Proceedings of the National Academy of Sciences, 103(30), 11358–11363. Gruber, N., Hauri, C., Lachkar, Z., Loher, D., Frölicher, T. L., & Plattner, G.-K. (2012). Rapid progression of ocean acidification in the California Current System. Science, 337(6091), 220– 223. Haack, E. A., Johnston, C. T., & Maurice, P. A. (2008). Mechanisms of siderophore sorption to smectite and siderophore-enhanced release of structural Fe3+. Geochimica et Cosmochimica Acta, 72(14), 3381–3397. Hall, S. J. (2002). The continental shelf benthic ecosystem: current status, agents for change and future prospects. Environmental Conservation, 29(3), 350–374. Hancke, K., & Glud, R. N. (2004). Temperature effects on respiration and photosynthesis in three diatom-dominated benthic communities. Aquatic Microbial Ecology, 37(3), 265–281. Hartmann, J., Jansen, N., Dürr, H. H., Kempe, S., & Köhler, P. (2009). Global CO2-consumption by chemical weathering: What is the contribution of highly active weathering regions? Global and Planetary Change, 69(4), 185–194. Hartmann, J., Moosdorf, N., Lauerwald, R., Hinderer, M., & West, A. J. (2014). Global chemical weathering and associated P-release—The role of lithology, temperature and soil properties. Chemical Geology, 363, 145–163. Hartmann, J., West, A. J., Renforth, P., Köhler, P., De La Rocha, C. L., Wolf‐Gladrow, D. A., Dürr, H. H., & Scheffran, J. (2013). Enhanced chemical weathering as a geoengineering strategy to reduce atmospheric carbon dioxide, supply nutrients, and mitigate ocean acidification. Reviews of Geophysics, 51(2), 113–149. Harvey, L. D. D. (2008). Mitigating the atmospheric CO2 increase and ocean acidification by adding limestone powder to upwelling regions. Journal of Geophysical Research: Oceans, 113(C4). Hausrath, E. M., Neaman, A., & Brantley, S. L. (2009). Elemental release rates from dissolving basalt and granite with and without organic ligands. American Journal of Science, 309(8), 633–660. Hayes, M. O. (1967). Relationship between coastal climate and bottom sediment type on the inner continental shelf. Marine Geology, 5(2), 111–132. 165 Hersman, L. E., Huang, A., Maurice, P. A., & Forsythe, J. H. (2000). Siderophore production and iron reduction by Pseudomonas mendocina in response to iron deprivation. Geomicrobiology Journal, 17(4), 261–273. Hider, R. C., & Kong, X. (2010). Chemistry and biology of siderophores. Natural Product Reports, 27(5), 637–657. Higgins, J. A., Fischer, W. W., & Schrag, D. P. (2009). Oxygenation of the ocean and sediments: consequences for the seafloor carbonate factory. Earth and Planetary Science Letters, 284(1– 2), 25–33. Hofmann, M., Retamal-Morales, G., & Tischler, D. (2020). Metal binding ability of microbial natural metal chelators and potential applications. Natural Product Reports, 37(9), 1262– 1283. Hsiang, S. M., Burke, M., & Miguel, E. (2013). Quantifying the influence of climate on human conflict. Science, 341(6151), 1235367. Hu, S., Sprintall, J., Guan, C., McPhaden, M. J., Wang, F., Hu, D., & Cai, W. (2020). Deep- reaching acceleration of global mean ocean circulation over the past two decades. Science Advances, 6(6). https://doi.org/10.1126/sciadv.aax7727 Huettel, M., Berg, P., & Kostka, J. E. (2014). Benthic Exchange and Biogeochemical Cycling in Permeable Sediments. Annual Review of Marine Science, 6(1), 23–51. https://doi.org/10.1146/annurev-marine-051413-012706 Hutchens, E. (2009). Microbial selectivity on mineral surfaces: possible implications for weathering processes. Fungal Biology Reviews, 23(4), 115–121. Iglesias‐Rodriguez, M. D., Armstrong, R., Feely, R., Hood, R., Kleypas, J., Milliman, J. D., Sabine, C., & Sarmiento, J. (2002). Progress made in study of ocean’s calcium carbonate budget. Eos, Transactions American Geophysical Union, 83(34), 365–375. Ilyina, T., Wolf‐Gladrow, D., Munhoven, G., & Heinze, C. (2013). Assessing the potential of calcium‐based artificial ocean alkalinization to mitigate rising atmospheric CO2 and ocean acidification. Geophysical Research Letters, 40(22), 5909–5914. IPCC. (2022). Climate Change 2022: Mitigation of Climate Change. Contribution of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. https://doi.org/10.1017/9781009157926 Irigoien, X., Klevjer, T. A., Røstad, A., Martinez, U., Boyra, G., Acuña, J. L., Bode, A., Echevarria, F., Gonzalez-Gordillo, J. I., & Hernandez-Leon, S. (2014). Large mesopelagic fishes biomass and trophic efficiency in the open ocean. Nature Communications, 5(1), 1–10. 166 Jahnke, R. A., Craven, D. B., & Gaillard, J.-F. (1994). The influence of organic matter diagenesis on CaCO3 dissolution at the deep-sea floor. Geochimica et Cosmochimica Acta, 58(13), 2799–2809. Jansen, H., Zeebe, R. E., & Wolf‐Gladrow, D. A. (2002). Modeling the dissolution of settling CaCO3 in the ocean. Global Biogeochemical Cycles, 16(2), 11. Jenkins, C. (2021). dbSEABED. INSTAAR, University of Colorado, Boulder. http://instaar.colorado.edu/~jenkinsc/dbseabed/ Jin, X., Gruber, N., Dunne, J. P., Sarmiento, J. L., & Armstrong, R. A. (2006). Diagnosing the contribution of phytoplankton functional groups to the production and export of particulate organic carbon, CaCO3, and opal from global nutrient and alkalinity distributions. Global Biogeochemical Cycles, 20(2). John, S. G., Liang, H., Weber, T., DeVries, T., Primeau, F., Moore, K., Holzer, M., Mahowald, N., Gardner, W., & Mishonov, A. (2020). AWESOME OCIM: A simple, flexible, and powerful tool for modeling elemental cycling in the oceans. Chemical Geology, 533, 119403. Kalinowski, B. E., Liermann, L. J., Givens, S., & Brantley, S. L. (2000). Rates of bacteria- promoted solubilization of Fe from minerals: a review of problems and approaches. Chemical Geology, 169(3–4), 357–370. Keir, R. S. (1980). The dissolution kinetics of biogenic calcium carbonates in seawater. Geochimica et Cosmochimica Acta, 44(2), 241–252. Kessler, A. J., Cardenas, M. B., Santos, I. R., & Cook, P. L. M. (2014). Enhancement of denitrification in permeable carbonate sediment due to intra-granular porosity: A multi-scale modelling analysis. Geochimica et Cosmochimica Acta, 141, 440–453. Kessler, A. J., Chen, Y.-J., Waite, D. W., Hutchinson, T., Koh, S., Popa, M. E., Beardall, J., Hugenholtz, P., Cook, P. L. M., & Greening, C. (2019). Bacterial fermentation and respiration processes are uncoupled in anoxic permeable sediments. Nature Microbiology, 4(6), 1014. Kessler, A. J., Rogers, A., Cyronak, T., Bourke, M. F., Hasler-Sheetal, H., Glud, R. N., Greening, C., Meysman, F. J. R., Eyre, B. D., & Cook, P. L. M. (2020). Pore water conditions driving calcium carbonate dissolution in reef sands. Geochimica et Cosmochimica Acta, 279, 16–28. Kleidon, A. (2004). Beyond Gaia: Thermodynamics of life and Earth system functioning. Climatic Change, 66(3), 271–319. Koeve, W., Duteil, O., Oschlies, A., Kähler, P., & Segschneider, J. (2014). Methods to evaluate CaCO 3 cycle modules in coupled global biogeochemical ocean models. Geoscientific Model Development, 7(5), 2393–2408. 167 Köhler, P., Abrams, J. F., Völker, C., Hauck, J., & Wolf-Gladrow, D. A. (2013). Geoengineering impact of open ocean dissolution of olivine on atmospheric CO2, surface ocean pH and marine biology. Environmental Research Letters, 8(1), 14009. Kraemer, S. M. (2004). Iron oxide dissolution and solubility in the presence of siderophores. Aquatic Sciences, 66(1), 3–18. Krumins, V., Gehlen, M., Arndt, S., Van Cappellen, P., & Regnier, P. (2013). Dissolved inorganic carbon and alkalinity fluxes from coastal marine sediments: Model estimates for different shelf environments and sensitivity to global change. Biogeosciences, 10(1), 371–398. https://doi.org/10.5194/bg-10-371-2013 Kuhn, K. M., DuBois, J. L., & Maurice, P. A. (2013). Strategies of aerobic microbial Fe acquisition from Fe-bearing montmorillonite clay. Geochimica et Cosmochimica Acta, 117, 191–202. Lam, V., & Pauly, D. (2005). Mapping the global biomass of mesopelagic fishes. Sea Around Us Project Newsletter, 30(4), 4. Lamérand, C., Shirokova, L. S., Bénézeth, P., Rols, J.-L., & Pokrovsky, O. S. (2020). Olivine dissolution and hydrous Mg carbonate and silicate precipitation in the presence of microbial consortium of photo-autotrophic and heterotrophic bacteria. Geochimica et Cosmochimica Acta, 268, 123–141. Lantz, C. A., Carpenter, R. C., & Edmunds, P. J. (2017). Calcium carbonate (CaCO3) sediment dissolution under elevated concentrations of carbon dioxide (CO2) and nitrate (NO3−). Journal of Experimental Marine Biology and Ecology, 495, 48–56. Lantz, C. A., Schulz, K. G., & Eyre, B. D. (2020). Ocean Acidification and Short‐Term Organic Matter Enrichment Alter Coral Reef Sediment Metabolism Through Different Pathways. Journal of Geophysical Research: Biogeosciences, 125(12), e2019JG005453. Lantz, C. A., Schulz, K. G., Stoltenberg, L., & Eyre, B. D. (2017). The short-term combined effects of temperature and organic matter enrichment on permeable coral reef carbonate sediment metabolism and dissolution. Biogeosciences, 14(23), 5377–5391. Laruelle, G. G., Cai, W.-J., Hu, X., Gruber, N., Mackenzie, F. T., & Regnier, P. (2018). Continental shelves as a variable but increasing global sink for atmospheric carbon dioxide. Nature Communications, 9(1), 454. Lauvset, S. K., Key, R. M., Olsen, A., Van Heuven, S., Velo, A., Lin, X., Schirnick, C., Kozyr, A., Tanhua, T., & Hoppema, M. (2016). A new global interior ocean mapped climatology: The 1× 1 GLODAP version 2. Earth System Science Data, 8(2), 325–340. Lear, C. H., Anand, P., Blenkinsop, T., Foster, G. L., Gagen, M., Hoogakker, B., Larter, R. D., Lunt, D. J., McCave, I. N., & McClymont, E. (2021). Geological Society of London Scientific Statement: what the geological record tells us about our present and future climate. In Journal of the Geological Society (Vol. 178, Issue 1). Geological Society of London. 168 Lebrato, M., Andersson, A. J., Ries, J. B., Aronson, R. B., Lamare, M. D., Koeve, W., Oschlies, A., Iglesias‐Rodriguez, M. D., Thatje, S., & Amsler, M. (2016). Benthic marine calcifiers coexist with CaCO3‐undersaturated seawater worldwide. Global Biogeochemical Cycles, 30(7), 1038–1053. Lee, C.-T. A., Jiang, H., Dasgupta, R., & Torres, M. (2019). A Framework for Understanding Whole-Earth Carbon Cycling. In B. N. Orcutt, I. Daniel, & R. Dasgupta (Eds.), Deep Carbon: Past to Present (pp. 313–357). Cambridge University Press. Lee, K. (2001). Global net community production estimated from the annual cycle of surface water total dissolved inorganic carbon. Limnology and Oceanography, 46(6), 1287–1297. Lewis, E., & Wallace, D. W. R. (1998). Program Developed for CO2 System Calculations, ORNL/CDIAC-105. https://salish-sea.pnnl.gov/media/ORNL-CDIAC-105.pdf Liermann, L. J., Kalinowski, B. E., Brantley, S. L., & Ferry, J. G. (2000). Role of bacterial siderophores in dissolution of hornblende. Geochimica et Cosmochimica Acta, 64(4), 587– 602. Liu, L., Li, S., Wang, S., Dong, Z., & Gao, H. (2018). Complex iron uptake by the putrebactin- mediated and Feo systems in Shewanella oneidensis. Applied and Environmental Microbiology, 84(20), e01752-18. Liu, Y., Olsen, A. A., & Rimstidt, J. D. (2006). Mechanism for the dissolution of olivine series minerals in acidic solutions. American Mineralogist, 91(2–3), 455–458. Liu, Z., Dreybrodt, W., & Liu, H. (2011). Atmospheric CO2 sink: silicate weathering or carbonate weathering? Applied Geochemistry, 26, S292–S294. Lovelock, J., & Margulis, L. (1974). Atmospheric homeostasis by and for the biosphere: the Gaia hypothesis. Tellus, 26(1–2), 2–10. Lowenstam, H. A. (1954). Factors affecting the aragonite: calcite ratios in carbonate-secreting marine organisms. The Journal of Geology, 62(3), 284–322. Luijendijk, A., Hagenaars, G., Ranasinghe, R., Baart, F., Donchyts, G., & Aarninkhof, S. (2018). The state of the world’s beaches. Scientific Reports, 8(1), 6641. Lüttge, A., & Conrad, P. G. (2004). Direct observation of microbial inhibition of calcite dissolution. Applied and Environmental Microbiology, 70(3), 1627–1632. Macdonald, F. A., Swanson-Hysell, N. L., Park, Y., Lisiecki, L., & Jagoutz, O. (2019). Arc- continent collisions in the tropics set Earth’s climate state. Science, 364(6436), 181–184. 169 Mackenzie, F. T., Andersson, A., Lerman, A., & Ver, L. M. (2005). Boundary exchanges in the global coastal margin: implications for the organic and inorganic carbon cycles. In A. R. R. and K. H. Brink (Ed.), The Sea, Volume 13: The Global Coastal Ocean (pp. 193–225). Harvard University Press. Manno, C., Sandrini, S., Tositti, L., & Accornero, A. (2007). First stages of degradation of Limacina helicina shells observed above the aragonite chemical lysocline in Terra Nova Bay (Antarctica). Journal of Marine Systems, 68(1–2), 91–102. Martiny, A. C., Pham, C. T. A., Primeau, F. W., Vrugt, J. A., Moore, J. K., Levin, S. A., & Lomas, M. W. (2013). Strong latitudinal patterns in the elemental ratios of marine plankton and organic matter. Nature Geoscience, 6(4), 279–283. McClintock, J. B., Amsler, M. O., Angus, R. A., Challener, R. C., Schram, J. B., Amsler, C. D., Mah, C. L., Cuce, J., & Baker, B. J. (2011). The Mg-calcite composition of Antarctic echinoderms: Important implications for predicting the impacts of ocean acidification. The Journal of Geology, 119(5), 457–466. McGinnis, D. F., Sommer, S., Lorke, A., Glud, R. N., & Linke, P. (2014). Quantifying tidally driven benthic oxygen exchange across permeable sediments: An aquatic eddy correlation study. Journal of Geophysical Research: Oceans, 119(10), 6918–6932. McLachlan, A. (1989). Water filtration by dissipative beaches. Limnology and Oceanography, 34(4), 774–780. Megens, L., Van der Plicht, J., & De Leeuw, J. W. (2001). Temporal variations in 13C and 14C concentrations in particulate organic matter from the southern North Sea. Geochimica et Cosmochimica Acta, 65(17), 2899–2911. Meysman, F. J. R., & Montserrat, F. (2017). Negative CO2 emissions via enhanced silicate weathering in coastal environments. Biology Letters, 13(4), 20160905. Middelburg, J. J., Soetaert, K., & Hagens, M. (2020). Ocean alkalinity, buffering and biogeochemical processes. Reviews of Geophysics, 58(3), e2019RG000681. Milliman, J. D. (1993). Production and accumulation of calcium carbonate in the Ocean: Budget of a non-steady state. Global Biogeochemical Cycles, 7(4), 927–957. Milliman, J. D., & Droxler, A. (1996). Neritic and pelagic carbonate sedimentation in the marine environment: ignorance is not bliss. Geologische Rundschau, 85(3), 496–504. Milliman, J. D., Troy, P. J., Balch, W. M., Adams, A. K., Li, Y.-H., & Mackenzie, F. T. (1999). Biologically mediated dissolution of calcium carbonate above the chemical lysocline? Deep Sea Research Part I: Oceanographic Research Papers, 46(10), 1653–1669. 170 Montserrat, F., Renforth, P., Hartmann, J., Leermakers, M., Knops, P., & Meysman, F. J. R. (2017). Olivine dissolution in seawater: implications for CO2 sequestration through enhanced weathering in coastal environments. Environmental Science & Technology, 51(7), 3960– 3972. Moore, W. S., Sarmiento, J. L., & Key, R. M. (2008). Submarine groundwater discharge revealed by 228 Ra distribution in the upper Atlantic Ocean. Nature Geoscience, 1(5), 309–311. Moreno, A. R., Garcia, C. A., Larkin, A. A., Lee, J. A., Wang, W.-L., Moore, J. K., Primeau, F. W., & Martiny, A. C. (2020). Latitudinal gradient in the respiration quotient and the implications for ocean oxygen availability. Proceedings of the National Academy of Sciences, 117(37), 22866–22872. Morse, J W, de Kanel, J., & Harris, K. (1979). Dissolution kinetics of calcium carbonate in seawater; VII, The dissolution kinetics of synthetic aragonite and pteropod tests. American Journal of Science, 279(5), 488–502. Morse, John W. (1985). Chemistry of calcium carbonate-rich shallow water sediments in the Bahamas. In American Journal of Science (Vol. 285, Issue 2, pp. 147–185). https://doi.org/10.2475/ajs.285.2.147 NASEM. (2019). Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. The National Academies Press. https://doi.org/10.17226/25259 Naviaux, J. D., Subhas, A. V, Dong, S., Rollins, N. E., Liu, X., Byrne, R. H., Berelson, W. M., & Adkins, J. F. (2019). Calcite dissolution rates in seawater: Lab vs. in-situ measurements and inhibition by organic matter. Marine Chemistry, 103684. Naviaux, J. D., Subhas, A. V, Rollins, N. E., Dong, S., Berelson, W. M., & Adkins, J. F. (2019). Temperature dependence of calcite dissolution kinetics in seawater. Geochimica et Cosmochimica Acta, 246, 363–384. Neaman, A., Chorover, J., & Brantley, S. L. (2005). Implications of the evolution of organic acid moieties for basalt weathering over geological time. American Journal of Science, 305(2), 147–185. Nickl, H. J., & Henisch, H. K. (1969). Growth of calcite crystals in gels. Journal of the Electrochemical Society, 116(9), 1258–1260. O’Brien, C. J., Peloquin, J. A., Vogt, M., Heinle, M., Gruber, N., Ajani, P., Andruleit, H., Arístegui, J., Beaufort, L., & Estrada, M. (2013). Global marine plankton functional type biomass distributions: coccolithophores. Earth System Science Data, 5(2), 259–276. O’Mara, N. A., & Dunne, J. P. (2019). Hot spots of carbon and alkalinity cycling in the coastal oceans. Scientific Reports, 9(1), 1–8. 171 Oakes, R. L., Peck, V. L., Manno, C., & Bralower, T. J. (2019). Degradation of internal organic matter is the main control on pteropod shell dissolution after death. Global Biogeochemical Cycles, 33(6), 749–760. Oakes, R. L., & Sessa, J. A. (2020). Determining how biotic and abiotic variables affect the shell condition and parameters of Heliconoides inflatus pteropods from a sediment trap in the Cariaco Basin. Biogeosciences, 17(7), 1975–1990. Oberdorfer, J. A., & Buddemeier, and R. W. (1986). Coral-reef hydrology: field studies of water movement within a barrier reef. Coral Reefs, 5(1), 7–12. Oelkers, E. H., Benning, L. G., Lutz, S., Mavromatis, V., Pearce, C. R., & Plümper, O. (2015). The efficient long-term inhibition of forsterite dissolution by common soil bacteria and fungi at Earth surface conditions. Geochimica et Cosmochimica Acta, 168, 222–235. Oelkers, E. H., Declercq, J., Saldi, G. D., Gislason, S. R., & Schott, J. (2018). Olivine dissolution rates: A critical review. Chemical Geology, 500, 1–19. Olsen, A. A., & Rimstidt, J. D. (2008). Oxalate-promoted forsterite dissolution at low pH. Geochimica et Cosmochimica Acta, 72(7), 1758–1766. Orr, J. C., Epitalon, J. M., Dickson, A. G., & Gattuso, J. P. (2018). Routine uncertainty propagation for the marine carbon dioxide system. Marine Chemistry, 207, 84–107. https://doi.org/10.1016/j.marchem.2018.10.006 Page, W., & Huyer, M. (1984). Derepression of the Azotobacter vinelandii siderophore system, using iron-containing minerals to limit iron repletion. Journal of Bacteriology, 158(2), 496– 502. Pecl, G. T., Araújo, M. B., Bell, J. D., Blanchard, J., Bonebrake, T. C., Chen, I.-C., Clark, T. D., Colwell, R. K., Danielsen, F., & Evengård, B. (2017). Biodiversity redistribution under climate change: Impacts on ecosystems and human well-being. Science, 355(6332), eaai9214. Penman, D. E., Rugenstein, J. K. C., Ibarra, D. E., & Winnick, M. J. (2020). Silicate weathering as a feedback and forcing in Earth’s climate and carbon cycle. Earth-Science Reviews, 209, 103298. Perez, A., Rossano, S., Trcera, N., Huguenot, D., Fourdrin, C., Verney-Carron, A., van Hullebusch, E. D., & Guyot, F. (2016). Bioalteration of synthetic Fe (III)-, Fe (II)-bearing basaltic glasses and Fe-free glass in the presence of the heterotrophic bacteria strain Pseudomonas aeruginosa: Impact of siderophores. Geochimica et Cosmochimica Acta, 188, 147–162. Perez, F. F., Fontela, M., García-Ibáñez, M. I., Mercier, H., Velo, A., Lherminier, P., Zunino, P., De La Paz, M., Alonso-Pérez, F., & Guallart, E. F. (2018). Meridional overturning circulation conveys fast acidification to the deep Atlantic Ocean. Nature, 554(7693), 515–518. 172 Perry, C. T., Salter, M. A., Harborne, A. R., Crowley, S. F., Jelks, H. L., & Wilson, R. W. (2011). Fish as major carbonate mud producers and missing components of the tropical carbonate factory. Proceedings of the National Academy of Sciences, 108(10), 3865–3869. Pokharel, R., Gerrits, R., Schuessler, J. A., & von Blanckenburg, F. (2019). Mechanisms of olivine dissolution by rock-inhabiting fungi explored using magnesium stable isotopes. Chemical Geology, 525, 18–27. Pokrovsky, O. S., Shirokova, L. S., Bénézeth, P., Schott, J., & Golubev, S. V. (2009). Effect of organic ligands and heterotrophic bacteria on wollastonite dissolution kinetics. American Journal of Science, 309(8), 731–772. Precht, E., & Huettel, M. (2003). Advective pore‐water exchange driven by surface gravity waves and its ecological implications. Limnology and Oceanography, 48(4), 1674–1684. Precht, E., & Huettel, M. (2004). Rapid wave-driven advective pore water exchange in a permeable coastal sediment. Journal of Sea Research, 51(2), 93–107. Ranasinghe, R. (2016). Assessing climate change impacts on open sandy coasts: A review. Earth- Science Reviews, 160, 320–332. Rao, A. M. F., Polerecky, L., Ionescu, D., Meysman, F. J. R., & De Beer, D. (2012). The influence of pore‐water advection, benthic photosynthesis, and respiration on calcium carbonate dynamics in reef sands. Limnology and Oceanography, 57(3), 809–825. Rassmann, J., Lansard, B., Gazeau, F., Guidi-Guilvard, L., Pozzato, L., Alliouane, S., Grenz, C., & Rabouille, C. (2018). Impact of ocean acidification on the biogeochemistry and meiofaunal assemblage of carbonate-rich sediments: Results from core incubations (Bay of Villefranche, NW Mediterranean Sea). Marine Chemistry, 203, 102–119. Raymo, M. E., & Ruddiman, W. F. (1992). Tectonic forcing of late Cenozoic climate. Nature, 359(6391), 117–122. Reichard, P. U., Kretzschmar, R., & Kraemer, S. M. (2007). Dissolution mechanisms of goethite in the presence of siderophores and organic acids. Geochimica et Cosmochimica Acta, 71(23), 5635–5650. Reimers, C. E., Stecher III, H. A., Taghon, G. L., Fuller, C. M., Huettel, M., Rusch, A., Ryckelynck, N., & Wild, C. (2004). In situ measurements of advective solute transport in permeable shelf sands. Continental Shelf Research, 24(2), 183–201. Ridgwell, A., Hargreaves, J. C., Edwards, N. R., Annan, J. D., Lenton, T. M., Marsh, R., Yool, A., & Watson, A. (2007). Marine geochemical data assimilation in an efficient Earth System Model of global biogeochemical cycling. Biogeosciences, 4(1), 87–104. Riedl, R. J., Huang, N., & Machan, R. (1972). The subtidal pump: a mechanism of interstitial water exchange by wave action. Marine Biology, 13(3), 210–221. 173 Rimstidt, J. D., Brantley, S. L., & Olsen, A. A. (2012). Systematic review of forsterite dissolution rate data. Geochimica et Cosmochimica Acta, 99, 159–178. Rosenberg, D. R., & Maurice, P. A. (2003). Siderophore adsorption to and dissolution of kaolinite at pH 3 to 7 and 22° C. Geochimica et Cosmochimica Acta, 67(2), 223–229. Rusch, A., Huettel, M., Wild, C., & Reimers, C. E. (2006). Benthic oxygen consumption and organic matter turnover in organic-poor, permeable shelf sands. Aquatic Geochemistry, 12(1), 1–19. Sabine, C. L., Key, R. M., Feely, R. A., & Greeley, D. (2002). Inorganic carbon in the Indian Ocean: Distribution and dissolution processes. Global Biogeochemical Cycles, 16(4), 11–15. Sagan, C., & Mullen, G. (1972). Earth and Mars: Evolution of atmospheres and surface temperatures. Science, 177(4043), 52–56. Salter, M. A., Harborne, A. R., Perry, C. T., & Wilson, R. W. (2017). Phase heterogeneity in carbonate production by marine fish influences their roles in sediment generation and the inorganic carbon cycle. Scientific Reports, 7(1), 1–15. Salter, M. A., Perry, C. T., & Smith, A. M. (2019). Calcium carbonate production by fish in temperate marine environments. Limnology and Oceanography, 64(6), 2755–2770. Salter, M. A., Perry, C. T., Stuart-Smith, R. D., Edgar, G. J., Wilson, R. W., & Harborne, A. R. (2018). Reef fish carbonate production assessments highlight regional variation in sedimentary significance. Geology, 46(8), 699–702. Santelli, C. M., Welch, S. A., Westrich, H. R., & Banfield, J. F. (2001). The effect of Fe-oxidizing bacteria on Fe-silicate mineral dissolution. Chemical Geology, 180(1–4), 99–115. Santos, I. R., Eyre, B., & Huettel, M. (2012). The driving forces of porewater and groundwater flow in permeable coastal sediments : A review. Estuarine, Coastal and Shelf Science, 98(February), 1–15. https://doi.org/10.1016/j.ecss.2011.10.024 Sarmiento, J. L., Dunne, J., Gnanadesikan, A., Key, R. M., Matsumoto, K., & Slater, R. (2002). A new estimate of the CaCO3 to organic carbon export ratio. Global Biogeochemical Cycles, 16(4), 51–54. Sarmiento, J. L., & Gruber, N. (2006). Ocean Biogeochemical Dynamics. Princeton University Press. Schiebel, R. (2002). Planktic foraminiferal sedimentation and the marine calcite budget. Global Biogeochemical Cycles, 16(4), 1–3. Schiebel, R., & Hemleben, C. (2017). Planktic foraminifers in the modern ocean. Springer. 174 Schiebel, R., & Movellan, A. (2012). First-order estimate of the planktic foraminifer biomass in the modern ocean. Earth System Science Data, 4(1), 75–89. Schott, J., & Berner, R. A. (1983). X-ray photoelectron studies of the mechanism of iron silicate dissolution during weathering. Geochimica et Cosmochimica Acta, 47(12), 2233–2240. Schwartzman, D. W., & Volk, T. (1989). Biotic enhancement of weathering and the habitability of Earth. Nature, 340(6233), 457–460. Shirokova, L. S., Bénézeth, P., Pokrovsky, O. S., Gerard, E., Ménez, B., & Alfredsson, H. (2012). Effect of the heterotrophic bacterium Pseudomonas reactans on olivine dissolution kinetics and implications for CO2 storage in basalts. Geochimica et Cosmochimica Acta, 80, 30–50. Shirvani, M., & Nourbakhsh, F. (2010). Desferrioxamine-B adsorption to and iron dissolution from palygorskite and sepiolite. Applied Clay Science, 48(3), 393–397. Shum, K. T., & Sundby, B. (1996). Organic matter processing in continental shelf sediments—the subtidal pump revisited. Marine Chemistry, 53(1–2), 81–87. Simeone, S., Molinaroli, E., Conforti, A., & De Falco, G. (2018). Impact of ocean acidification on the carbonate sediment budget of a temperate mixed beach. Climatic Change, 150(3), 227– 242. Simone, M. N., Schulz, K. G., Eyre, B. D., & Oakes, J. M. (2021). Ocean acidification may mitigate negative effects of warming on carbon burial potential in subtidal unvegetated estuarine sediments. Limnology and Oceanography, 66(8), 2953–2966. Smith, C. R., Hoover, D. J., Doan, S. E., Pope, R. H., Demaster, D. J., Dobbs, F. C., & Altabet, M. A. (1996). Phytodetritus at the abyssal seafloor across 10 of latitude in the central equatorial Pacific. Deep Sea Research Part II: Topical Studies in Oceanography, 43(4–6), 1309–1338. Smith, S. V. (1972). Production of calcium carbonate on the mainland shelf of Southern California. Limnology and Oceanography, 17(1), 28–41. Steinacher, M., Joos, F., Frölicher, T. L., Plattner, G.-K., & Doney, S. C. (2009). Imminent ocean acidification in the Arctic projected with the NCAR global coupled carbon cycle-climate model. Biogeosciences, 6(4), 515–533. Stewart, A. G., Hudson-Edwards, K. A., & Dubbin, W. E. (2013). Mechanisms of goethite dissolution in the presence of desferrioxamine B and Suwannee River fulvic acid at pH 6.5. Geochimica et Cosmochimica Acta, 115, 1–14. Stockmann, G. J., Shirokova, L. S., Pokrovsky, O. S., Bénézeth, P., Bovet, N., Gislason, S. R., & Oelkers, E. H. (2012). Does the presence of heterotrophic bacterium Pseudomonas reactans affect basaltic glass dissolution rates? Chemical Geology, 296, 1–18. 175 Stoltenberg, L., Schulz, K. G., Cyronak, T., & Eyre, B. D. (2020). Seasonal variability of calcium carbonate precipitation and dissolution in shallow coral reef sediments. Limnology and Oceanography, 65(4), 876–891. Stoltenberg, L., Schulz, K. G., Lantz, C. A., Cyronak, T., & Eyre, B. D. (2021). Late Afternoon Seasonal Transition to Dissolution in a Coral Reef: An Early Warning of a Net Dissolving Ecosystem? In Geophysical Research Letters (Vol. 48, Issue 6, p. e2020GL090811). American Geophysical Union (AGU). https://doi.org/10.1029/2020GL090811 Subhas, A. V., Adkins, J. F., Rollins, N. E., Naviaux, J., Erez, J., & Berelson, W. M. (2017). Catalysis and chemical mechanisms of calcite dissolution in seawater. Proceedings of the National Academy of Sciences, 114(31), 8175–8180. https://doi.org/10.1073/pnas.1703604114 Subhas, A. V, Dong, S., Naviaux, J. D., Rollins, N. E., Ziveri, P., Gray, W., Rae, J. W. B., Liu, X., Byrne, R. H., & Chen, S. (2022). Shallow Calcium Carbonate Cycling in the North Pacific Ocean. Global Biogeochemical Cycles, e2022GB007388. Subhas, A. V, Rollins, N. E., Berelson, W. M., Erez, J., Ziveri, P., Langer, G., & Adkins, J. F. (2018). The dissolution behavior of biogenic calcites in seawater and a possible role for magnesium and organic carbon. Marine Chemistry, 205, 100–112. Sulpis, O., Boudreau, B. P., Mucci, A., Jenkins, C., Trossman, D. S., Arbic, B. K., & Key, R. M. (2018). Current CaCO3 dissolution at the seafloor caused by anthropogenic CO2. Proceedings of the National Academy of Sciences, 115(46), 11700–11705. Sulpis, O., Jeansson, E., Dinauer, A., Lauvset, S. K., & Middelburg, J. J. (2021). Calcium carbonate dissolution patterns in the ocean. Nature Geoscience, 14(6), 423–428. Sverdrup, H. U., Johnson, N. W., & Fleming, R. H. (1941). The Oceans. Prentice Hall. Thomas, H., Bozec, Y., Elkalay, K., & De Baar, H. J. W. (2004). Enhanced Open Ocean Storage of CO2 from Shelf Sea Pumping. Science, 304(5673), 1005–1008. Torres, M. A., Dong, S., Nealson, K. H., & West, A. J. (2019). The kinetics of siderophore‐ mediated olivine dissolution. Geobiology, 17(4), 401–416. Trnovsky, D., Stoltenberg, L., Cyronak, T., & Eyre, B. D. (2016). Antagonistic effects of ocean acidification and rising sea surface temperature on the dissolution of coral reef carbonate sediments. Frontiers in Marine Science, 3, 211. Tsunogai, S., Watanabe, S., & Sato, T. (1999). Is there a “continental shelf pump” for the absorption of atmospheric CO2? Tellus B: Chemical and Physical Meteorology, 51(3), 701– 712. 176 Tyndall, J. (1861). I. The Bakerian Lecture.—On the absorption and radiation of heat by gases and vapours, and on the physical connexion of radiation, absorption, and conduction. Philosophical Transactions of the Royal Society of London, 151, 1–36. https://doi.org/https://doi.org/10.1098/rstl.1861.0001 Uchikawa, J., & Zeebe, R. E. (2008). Influence of terrestrial weathering on ocean acidification and the next glacial inception. Geophysical Research Letters, 35(23). Van Den Berghe, M., Merino, N., Nealson, K. H., & West, A. J. (2021). Silicate minerals as a direct source of limiting nutrients: Siderophore synthesis and uptake promote ferric iron bioavailability from olivine and microbial growth. Geobiology, 19(6), 618–630. van Heuven, S., Pierrot, D., Rae, J. W. B., Lewis, E., & Wallace, D. W. R. (2011). MATLAB Program Developed for CO2 System Calculations. ORNL/CDIAC-105b. https://doi.org/10.3334/CDIAC/otg.CO2SYS_MATLAB_v1.1 Vopel, K., Del-Río, C., & Pilditch, C. A. (2018). Effects of CO 2 enrichment on benthic primary production and inorganic nitrogen fluxes in two coastal sediments. Scientific Reports, 8(1), 1–15. Vousdoukas, M. I., Ranasinghe, R., Mentaschi, L., Plomaritis, T. A., Athanasiou, P., Luijendijk, A., & Feyen, L. (2020). Sandy coastlines under threat of erosion. Nature Climate Change, 10(3), 260–263. Walker, J. C. G., Hays, P. B., & Kasting, J. F. (1981). A negative feedback mechanism for the long‐term stabilization of Earth’s surface temperature. Journal of Geophysical Research: Oceans, 86(C10), 9776–9782. Walker, M., Hammel, J. U., Wilde, F., Hoehfurtner, T., Humphries, S., & Schuech, R. (2021). Estimation of sinking velocity using free-falling dynamically scaled models: Foraminifera as a test case. Journal of Experimental Biology, 224(2), jeb230961. Weber, T., John, S., Tagliabue, A., & DeVries, T. (2018). Biological uptake and reversible scavenging of zinc in the global ocean. Science, 361(6397), 72 LP – 76. West, A. J., Galy, A., & Bickle, M. (2005). Tectonic and climatic controls on silicate weathering. Earth and Planetary Science Letters, 235(1–2), 211–228. Wild, B., Imfeld, G., & Daval, D. (2021). Direct measurement of fungal contribution to silicate weathering rates in soil. Geology, 49(9), 1055–1058. Wild, B., Imfeld, G., Guyot, F., & Daval, D. (2018). Early stages of bacterial community adaptation to silicate aging. Geology, 46(6), 555–558. Wild, C., Rasheed, M., Jantzen, C., Cook, P., Struck, U., Huettel, M., & Boetius, A. (2005). Benthic metabolism and degradation of natural particulate organic matter in carbonate and silicate reef sands of the northern Red Sea. Marine Ecology Progress Series, 298, 69–78. 177 Wilson, R. W., Millero, F. J., Taylor, J. R., Walsh, P. J., Christensen, V., Jennings, S., & Grosell, M. (2009). Contribution of fish to the marine inorganic carbon cycle. Science, 323(5912), 359–362. Wogelius, R. A., & Walther, J. V. (1991). Olivine dissolution at 25 C: Effects of pH, CO2, and organic acids. Geochimica et Cosmochimica Acta, 55(4), 943–954. Wolf-Gladrow, D. A., Zeebe, R. E., Klaas, C., Körtzinger, A., & Dickson, A. G. (2007). Total alkalinity: The explicit conservative expression and its application to biogeochemical processes. Marine Chemistry, 106(1–2), 287–300. Woosley, R. J., Millero, F. J., & Grosell, M. (2012). The solubility of fish‐produced high magnesium calcite in seawater. Journal of Geophysical Research: Oceans, 117(C4). Yates, K K, & Halley, R. B. (2003). Measuring coral reef community metabolism using new benthic chamber technology. Coral Reefs, 22(3), 247–255. Yates, Kimberly K, & Halley, R. B. (2006). Diurnal variation in rates of calcification and carbonate sediment dissolution in Florida Bay. Estuaries and Coasts, 29(1), 24–39. Zaric, S., Schulz, M., & Mulitza, S. (2006). Global prediction of planktonic foraminiferal fluxes from hydrographic and productivity data. Biogeosciences, 3, 187–207. https://doi.org/10.5194/bg-3-187-2006 Zeebe, R. E. (2012). History of seawater carbonate chemistry, atmospheric CO2, and ocean acidification. Annual Review of Earth and Planetary Sciences, 40, 141–165. Zeebe, R. E., & Westbroek, P. (2003). A simple model for the CaCO3 saturation state of the ocean: The “Strangelove,” the “Neritan,” and the “Cretan” Ocean. Geochemistry, Geophysics, Geosystems, 4(12). Zhang, H., Stoll, H., Bolton, C., Jin, X., & Liu, C. (2018). A refinement of coccolith separation methods: measuring the sinking characteristics of coccoliths. Biogeosciences, 15(15), 4759– 4775. Ziveri, P., Gray, W., Ortiz, G., Manno, C., Grelaud, M., Incarbona, A., Rae, J., Subhas, A., Pallacks, S., White, A., Adkins, J., & Berelson, W. (2022). Pelagic carbonate production in the North Pacific Ocean. In Revision at Nature Communications.

Abstract (if available)

Abstract The titration of carbon dioxide (CO2) by silicate and carbonate minerals is one of the major regulators of earth’s climate. When silicate and carbonate minerals dissolve, they produce alkalinity, effectively converting CO2 into non-volatile bicarbonate and carbonate ions (HCO3- and CO32-) that remain dissolved in seawater on the timescale of 100 ky. This process, commonly called weathering, has regulated atmospheric CO2 concentration for much of earth history. Recently, however, anthropogenic CO2 emissions exceed the pace of natural mineral weathering, and CO2 is accumulating in the atmosphere at geologically unprecedented speed. The resulting climate crisis has expedited our need to better understand the details of mineral weathering -- both for predicting the weathering feedback in a higher CO2 world, and for optimizing engineered CO2 capture by “enhanced weathering”. This dissertation compiles the results of collaborative research, conducted by myself with contribution from several colleagues, exploring how various environmental factors control the dissolution rates of both carbonate and silicate minerals.
Starting at the mineral surface, Chapter 2 considers how microbes use siderophores to extract Fe from olivine, effectively increasing dissolution and producing alkalinity. We find that Shewanella oneidensis increases dissolution rates by an order of magnitude above abiotic rates, and that siderophores are required for this enhancement. Furthermore, S. oneidensis appears to use siderophores synergistically with other mechanisms to achieve even higher dissolution rates than expected: in experiments with a mutant strain of S. oneidensis incapable of producing siderophores but “fed” exogenous siderophores, dissolution rates were 8-fold higher than abiotic experiments with the same siderophore concentration. These results shed light on fundamental geobiological mechanisms, as well as provide direction for future research on biotechnological approaches to enhanced weathering.
Next, Chapter 3 considers the controls on CaCO3 dissolution rates in ocean sediments, specifically focusing on detrital sands. Such sands are among the most common sediment types on the shelf, yet almost completely unstudied with respect to carbonate chemistry. Using flow-through reactors and 13C isotope mass balance, we show that dissolution is a function of both seawater saturation state (Ω), and porewater residence time. At decreased Ω, simulating ocean acidification, dissolution rates increased significantly and initiated sooner upon seawater advection into the sand. This response to acidification was surprisingly similar to that in higher carbonate-content sands, suggesting that detrital sediments have the potential to support enhanced dissolution in an acidifying ocean, and may be an increasing source of alkalinity to coastal waters.
Finally, expanding perspective to the global ocean, Chapter 4 uses an ocean circulation inverse model (OCIM) to constrain the marine CaCO3 cycle. CaCO3 export from the surface ocean and dissolution in the water column are key controls on ocean alkalinity distribution (and thus air-sea CO2 flux), yet are poorly constrained. Testing models with different CaCO3 dissolution mechanisms against global observations of ocean alkalinity, we show that dissolution must occur above the calcite and aragonite saturation horizons, and that dissolution rates exceed those expected from known kinetics. We also show that a range of CaCO3 export values (1.1 to 1.8 Gt C y-1) can match observations, but that modeled exports all converge below 300 m, indicating significant recycling in the upper ocean. Collectively, these results show that dissolution is not only thermodynamically-driven, but that other mechanisms, e.g., microbial respiration in sinking particles and/or production of more soluble CaCO3, drive dissolution throughout the water column.
While not explicitly focusing on biological processes, all three chapters highlight the important role that biology—e.g., organic ligands, microbial respiration in sediments, and the coupling of organic matter with CaCO3 ballast in sinking particles—plays in controlling mineral dissolution rates. Mineral weathering is often considered a “geochemical” process, but this dissertation highlights the inextricable link between biology and geochemistry in driving global weathering processes. These results shed light on fundamental earth system processes and inform engineering solutions to CO2 sequestration by “enhanced weathering.”

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Calcite and aragonite dissolution in seawater: kinetics, mechanisms, and fluxes in the North Pacific

PDF

Carbonate dissolution at the seafloor: fluxes and drivers from a novel in situ porewater sampler

PDF

Exploring bacteria-mineral interactions

PDF

Chemical weathering across spatial and temporal scales: from laboratory experiments to global models

PDF

Investigating the global ocean biogeochemical cycling of alkalinity, barium, and copper using data-constrained inverse models

PDF

Exploring temporal changes in surface species on weathered feldspar mineral surfaces using solid-state NMR spectroscopy

PDF

Concentration and size partitioning of trace metals in surface waters of the global ocean and storm runoff

PDF

Germanium and silicon isotope geochemistry in terrestrial and marine low-temperature environments

PDF

Investigations on marine metal cycling through a global expedition, a wildfire survey, and a viral infection

PDF

Effects of global change on the physiology and biogeochemistry of the N₂-fixing cyanobacteria Trichodesmium erythraeum and Crocosphaera watsonii

PDF

The impact of Lagrangian environmental variability on the growth of phytoplankton

PDF

Paleoecology of Upper Triassic reef ecosystems and their demise at the Triassic-Jurassic extinction, a potential ocean acidification event

PDF

How open ocean calcifiers broke the link between large igneous provinces and mass extinctions

PDF

Particulate organic carbon flux calculated from ²³⁴Th measurements and sediment traps in the eastern tropical south Pacific

PDF

Iron-dependent response mechanisms of the nitrogen-fixing cyanobacterium Crocosphaera to climate change

PDF

Modeling deep ocean water and sediment dynamics in the eastern Pacific Ocean using actinium-227 and other naturally occurring radioisotopes

PDF

Investigating microbial biofilm community mediated processes on surfaces: from single cell genomics to community meta-omics

PDF

The molecular adaptation of Trichodesmium to long-term CO₂-selection under multiple nutrient limitation regimes

$Germanium-silicon fractionation in a continental shelf environment: insights from the northern Gulf of Mexico$

PDF

Germanium-silicon fractionation in a continental shelf environment: insights from the northern Gulf of Mexico

PDF

The geobiology of fluvial, lacustrine, and marginal marine carbonate microbialites (Pleistocene, Miocene, and Late Triassic) and their environmental significance

Asset Metadata

Creator Lunstrum, Abby Michelle (author)

Core Title Environmental controls on alkalinity generation from mineral dissolution: from the mineral surface to the global ocean

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Geological Sciences

Degree Conferral Date 2022-08

Publication Date 07/26/2022

Defense Date 06/15/2022

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag alkalinity,calcium carbonate,carbonate counterpump,carbonates,continental shelf,enhanced weathering,geological carbon sequestration,hard tissue pump,mineral dissolution,OAI-PMH Harvest,ocean acidification,olivine,Sand,siderophores,weathering

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Berelson, William (committee chair), Hutchins, David (committee member), John, Seth (committee member), West, Joshua (committee member)

Creator Email abbylunstrum@gmail.com,lunstrum@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC111375215

Unique identifier UC111375215

Legacy Identifier etd-LunstrumAb-11003

Document Type Dissertation

Format application/pdf (imt)

Rights Lunstrum, Abby Michelle

Type texts

Source 20220728-usctheses-batch-962 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu