A Diurnal Carbon Engine Explains 13C-Enriched Carbonates

Downloaded by guest on September 27, 2021 i ie natmospheric O in of However, rises amount tic burial. proportional carbon a carbon organic releases production, organic burial primary of global interpreted stoichiometry carbonate the elevated are through of of frequently records volumes deposits as significant These shelves. years, continental billion (δ carbon-isotopic Denmark, 3 high M., Odense with past Denmark, Southern of the 2019) University 21, Evolution, In May Earth review for for Center (received Nordic 2019 and 8, Biology October of approved Institute and Canfield, E. Donald by Edited a Geyman C. Emily oxygen of production global the increasing without carbonates explains engine carbon diurnal A www.pnas.org/cgi/doi/10.1073/pnas.1908783116 the that assumes paradox carbon–oxygen This positive )pooytei/eprto.Terltv otiuin fthese carbon- of and contributions 1) relative exchange, are The gas pool photosynthesis/respiration. 3) atmospheric DIC the 2) of precipitation/dissolution, out ate and into at carbon respired exchanging is day each depletion nutrient produced and zero. burial matter is organic-matter net organic that such the pho- night, of whereby organic-carbon all high mechanism significant if drives alternative of still an idea tosynthesis propose the have We refuting also (6) export. (5), slope periplatformal TOC the to low on shelf exported produced the and carbonates GBB explanation off the However, carbon burial nutrients. organic (TOC) depleting organic of carbon without export canonical organic wholesale the total require would low Thus, car- have (5). which GBB (∼0.3%) (from the waters on in residual car- sediments enriched the to precipitated) leave relative are to bonates carbon order organic in of bonate export/burial the increasing carbonates or matter organic burying without O carbonate, generating shallow waters and mecha- mixed simple DIC a sluggishly observed provides the shallow, and for photosynthesis, in nism of amplified rates high most with is (Ω that saturation sis inorganic carbonate and in organic changes and between coupled carbon forces of reservoirs transfer elevated calcite daily explain for where We expected seawater. than from δ precipitating higher aragonite is or sediment the carbonate However, water (DIC). carbon inorganic the reflects carbonates water mns h otcmo xlnto o h rdcinof production the the for of explanation half common high explain most only The and iments. (∼ 2) (1, calcite rate elevated inorganic precipitation and to ture (∼ enrich- bicarbonate Carbon-isotope aragonite for 1). but and (Fig. salinities, factors waters and marine ment times open sediment of residence values carbonate seawater produce of all range wide a A eateto esine,PictnUiest,Pictn J08544 NJ Princeton, University, Princeton Geosciences, of Department 13 nsalwbns h otipratmcaim for mechanisms important most 3 the banks, shallow On nsalwcroae ihadunlcro yl engine, cycle carbon diurnal a with carbonates shallow in C δ Bhms,adSakBy(etr utai)represent Australia) Salvador (Western San Bay [GBB]), Shark Bank and Bahama (Bahamas), (Great Island ndros δ 13 13 fDC hsegn anan abnccehystere- carbon-cycle a maintains engine This DIC. of C abnt nagoa 3 rlcl()saeinvolves scale (4) local or (3) global a on carbonate C δ δ | 13 13 abnisotopes carbon xusosta ucut h elgclrecord. geological the punctuate that excursions C inlosre nteesalwcroaesed- carbonate shallow these in observed signal C 2 . r eaieyisniiet tempera- to insensitive relatively are 2.7h) a,1 n dmC Maloof C. Adam and δ pO 13 | -eopigbtengoa seawater global between C-decoupling 2 chemostratigraphy uigte1t 0 ilo year-long million 100 to 1 the during 13 δ )vle cuuae nshallow on accumulated values C) δ 13 13 nsalwcroae,even carbonates, shallow in C fgoa seawater-dissolved global of C ∼ 5h 13 δ 13 .Hwvr carbonate However, C. ihrta the than higher fmdr shallow- modern of C 2 | rdcigunrealis- predicting , a paleoclimate δ 13 fshallow of C 1.h) δ 13 A C ) doi:10.1073/pnas.1908783116/-/DCSupplemental at online information supporting contains article This 1 is ulse oebr8 2019. 8, November published First the under Published Submission. Direct PNAS research, a is performed article This interest.y research, competing no designed declare authors The A.C.M. paper.y the and wrote and data, E.C.G. analyzed contributions: Author hnpeitdfraaoiei qiiru ihaeaelocal produce average to has with equilibrium is that in DIC. hysteresis record aragonite seawater for cycle sediment predicted carbon carbonate than diurnal shallow low this a the of depresses recording effect also production but bonate DIC of in of tion Ω consumes synthesis a mits up covers respiration but night. day, In each the tracks (12–15). to the during environments perturbations system carbonate substantial carbonate cycle drives the shallow over (24-h) photosynthesis in by diurnal words, 20 fluxes other single carbonate of a and factor and of atmospheric matter a gross course organic the the between exceed carbon over can of matter pool flux organic DIC the of the zero, burial to net close the and car- is Although 37, net 62, 2). the be (Fig. of to sizes respectively fluxes relative organic vs. the and DIC constrain gas-exchange, and GBB bonate, (5) the TOC from of (7) since Observations TA 2B), 1). total (Table (Fig. and DIC (TA) (9) in alkalinity change diagram characteristic Deffeyes a induces a mechanism each in evolution carbon seawater the banktop visualizing of by constrained be can processes δ owo orsodnemyb drse.Eal [email protected] Email: addressed. be may correspondence whom To 13 epeetsal abniooe(δ isotope carbon stable present We Significance a eoda lblrognztoso h abncceand cycle controls. paleoecological carbon and the of link reorganizations instead global as the record positive alleviate cal results interpret Our to record. geological need the to in positive model and waters anomalously today, shallow simple produce in to a cycling sufficient carbon and is diurnal observations ordinary how modern show use of carbon- We 97% shallow of ates. first interpretations on and the rely paleoclimate chemistry global for seawater about because, inferences many shelves. history, important carbonate Earth is shallow versus carbon realization ocean the This global of the decoupling in a cycles require that sediment carbonate A h ira yl fpooytei n eprto per- respiration and photosynthesis of cycle diurnal The frsda I,laigt abnt eietenriched sediment carbonate to leading DIC, residual of C h aiu aeo abnt rcptto afunc- (a precipitation carbonate of rate maximum The . 13 .A ih al,arbcrsiainlwr the lowers respiration aerobic falls, night As C. Ω δ 13 A PNAS sahee spa htsnhssicessthe increases photosynthesis peak as achieved is ) ytrss(i.3.A h u ie,photo- rises, sun the As 3). (Fig. hysteresis C 13 δ NSlicense.y PNAS 13 | C-enriched olclado lblpaleoenvironmental global and/or local to C eebr3 2019 3, December CO 2 rmtewtrclm,diigup driving column, water the from y δ 13 Ω A xusosi h geologi- the in excursions C | rvnigsgicn car- significant preventing , . y o.116 vol. https://www.pnas.org/lookup/suppl/ 13 )dt rmmodern from data C) δ 13 | aus h net The values. C o 49 no. δ 13 nshelves on C δ | 13 24433–24439 higher C <1%, δ 13 C

EARTH, ATMOSPHERIC, AND PLANETARY SCIENCES Fig. 1. Evidence in support of a diurnal organic car- A bon engine driving elevated δ13C in shallow carbonates. (A) Three modern carbonate shelves produce bulk sediment with δ13C that is too high to have precipitated directly from open marine waters. (B) The δ13C values of GBB sediments rise abruptly near the shelf edge and remain high (>4.5 ) in the bank interior. The number of samples ish denoted as this study/ref. 4 (1,159/167), and the banktop water ages in B are computed from a spatial interpolation of the 14C ages in ref. 7 (Fig. 2). The diurnal organic carbon engine hypothesis predicts a number of observations consistent with the dataset: 1 peak δ13C of DIC should be high enough to precipitate aragonite with the dominant value of δ13C∼ 5 ; 2 1.8 variability in DIC is the same range predictedh h during B one diurnal photosynthesis–respiration cycle (Fig. 3); 3 banktop waters have a δ13C of DIC composition roughly equal to or higher than that of offshore waters; 4 sediments at the shelf edge have low δ13C, consistent with open marine DIC; 5 the rapid rise in sediment δ13C near the platform edge cap- tures the transition from the weak photosynthetic forcing (κp) characteristic of the open ocean to the strong photosynthetic forcing observed in shallow waters (Fig. 4); and 6 the lack of a signiﬁcant cross- shelf δ13C gradient reﬂects the balancing isotopic effects of carbonate precipitation (depletes residual DIC) and CO2 gas escape (enriches residual DIC) (Fig. 7).

Carbon Model diurnal engine returns 100% of the carbon sequestered in organic To quantify the extent to which the temporal coupling between matter during the day back to the seawater each night (Fig. 3). photosynthesis and carbonate precipitation increases the δ13C of The only prescribed forcing in the engine model is the daily pho- carbonate, we develop a simple mass balance model of the diur- tosynthetic transfer of carbon between the inorganic and organic nal carbon engine (Fig. 3). To emphasize how the diurnal engine reservoirs (κp ). The gas exchange and carbonate ﬂuxes are com- mechanism is distinct from the canonical treatment whereby net puted independently based on values of pCO2 and ΩA obtained primary production drives elevated δ13C in carbonates (3, 4), our in each time step using the carbonate-system calculation software

250 A Broecker & Takahashi (1966) C 37% B 26.0 14

C banktop water ages Eq/kg) Atm. Depth (m) 200 TA ( CaCO3

2000 Corg 25.6 -5 62% < 1%5 20 6 22 6 150 48 10 8 1800 0 = 6 12 A

71 5 25.2 100 56 1600 -10 4 73 66 Banktop water age (days) 96 56 3 Broecker & Takahashi (1966) 133 50 255 1400 (aragonite) 24.8 245 142 180 Atm. equilibrium: m = 1.26 150 1 Linear fit: m = 1.24 -15 0 -79.0 -78.6 -78.2 -77.8 1000 1200 1400 1600 1800 2000 CO ( mol/kg) 2

14 Fig. 2. Measurements of C, ΣCO2, and TA in seawater from the GBB (7) help to constrain the carbon fluxes in our model (Fig. 3). (A) Shallow waters and sluggish mixing on the GBB afford banktop residence times >250 days. Bathymetry from ref. 8. (B) ΣCO2 and TA measurements produce a slope of m = 1.24 in the Deffeyes diagram (9) (Fig. 3C). The line representing atmospheric equilibrium was derived by allowing TA to vary from 1,200 to 2,375 µEq/kg (the y-axis range) and then using CO2SYS (10) to calculate ΣCO2 if [CO2 (aq)] = 317 ppm (i.e., equal to atmospheric pCO2 at the time of the TA and ΣCO2 measurements) (11). (C) We use the data in B and the observation that the mean TOC (weight %) of modern GBB sediment is 0.32±0.3% (1σ) (5) to constrain the sizes of the net carbonate precipitation, gas-escape, and Corg fluxes to be 62, 37, and <1%, respectively. We are able to compute the relative magnitudes of these 3 fluxes (2 unknowns, since Fcarb + Forg + Fgas = 1) because: 1) TOC measurements (5) link organic carbon and carbonate burial: Forg = 0.0107Fcarb; and 2) the slope in B reflects the relative contributions of carbonate precipitation, photosynthesis, and gas exchange, as each process induces a characteristic ∆TA:∆ΣCO2 ratio (Fig. 3C and Table 1).

24434 | www.pnas.org/cgi/doi/10.1073/pnas.1908783116 Geyman and Maloof Downloaded by guest on September 27, 2021 Downloaded by guest on September 27, 2021 emnadMaloof and Geyman (A ronments. 4. Fig. of composition cumulative The starting (D) a 2B). (Fig. with the 1.24 Water of makes slope simulation. that a model follows exchange (7) 24-h gas GBB the the air–sea on during of δ seawater Average produced rate 1). sediment (Table slow DIC bulk the and TA of is on it levers 0) that = (slope Note CO equilibration S14). that of Fig. fact Appendix, function the (SI (a possible; (C parameterization engine production chosen carbon carbonate the organic of to diurnal rates insensitive simulation. the are model modulate results engine 24-h photosynthesis model one by for the induced DIC of and chemistry TA function water in Changes in (B) changes photosynthesis/respiration. and The precipitation/dissolution, carbonate exchange, gas air–sea 3. Fig. eouintm eiso ira abnt ytmvariability 4). (Fig. system world the carbonate around locations diurnal shelf shallow of 21 from series of time value the resolution constrain To (10). (CO2SYS) mean the than rather instantaneous cycle, the diurnal the pool, no during DIC of sediment achieved photosynthesis bulk the composition DIC thus integrated, of to starting and the back a legend), rates, carbon 4 with precipitation organic Fig. water carbonate the (see that of increases P predicts 100% or primarily model N returns lever engine no photosynthesis the contains the 1), matter (Table that organic TA banktop bulk in which with increase in aragonite modest scenario end-member its the induces model longer we if example, For 13 h efys()darmdpcstepooytei–eprto soe= (slope photosynthesis–respiration the depicts diagram (9) Deffeyes The ) of C oeta h ira rai abnegn tl elevates still engine carbon organic diurnal the that Note +4.51h. ueia asblnemdltak the tracks model balance mass numerical A (A) engine. carbon organic diurnal The nilsrto fqatfigtedunlognccro rnfr(κ transfer carbon organic diurnal the quantifying of illustration An 1800 2000 2200 2200 2300 2400 200 400 600 and pCO 21 24 18 12 6 0 esrmnso ira I n Avraiiya eo sadLgo GetBrirRe)(5 soeeapeo h 1shallow 21 the of example one as (15) Reef) Barrier (Great Lagoon Island Heron at variability TA and DIC diurnal of Measurements B) B C A A -1 -1 2

( atm) ( mol kg ) ( mol kg ) δ rdetbtenteamshr n ewtr.W s omnprmtrztosfrCO for parameterizations common 5 use We seawater). and atmosphere the between gradient 13 of C = +2.7 = -2 Hour ofday Smoothed interpolation Observations McMahon etal.(2013) h esnfrterltv nestvt fthe of insensitivity relative the for reason The +4.48h. Atmospheric = p CO 2 Ω -1 -1 -1 -1 κ

-1 -1 = +2.7

F ( mol kg hr ) F ( mol kg hr ) F ( mol kg hr ) = -10 CaCO org p

gas 3 -50 ydaigdown drawing by -10 = 0 50 ecmiehigh- compile we , 10 -1 0 1 0 0 21 24 18 12 6 0 2 F D E osnteulbaewt h topeeo 4hcce lospooytei odieup drive to photosynthesis allows cycles 24-h on atmosphere the with equilibrate not does κ δ C p 13 δ TA ( Eq/kg) 13 erae by decreases C Hour ofday 2350 2300 fDIC. of C B 15:00 Σ CO p 2 δ o yadn A() oiein Notice (C). TA adding by not , 616,croaepeiiaindsouin(lp /) n atmospheric and 2/1), = (slope precipitation–dissolution carbonate −16/106),

rmhg-eouinosrain ftecroaesse nmdr envi- modern in system carbonate the of observations high-resolution from ) 7 13 δ es htw onticueterglto fitra car- vital internal to of rise regulation give which the organisms, include calcifying not in systems do bonate we that sense < vni rai atri oecro-ihta yia eﬁl ratio. Redﬁeld typical a than carbon-rich more is matter organic if even C 13 18:00 C h rcpttdsdmn rfrnilyrcrstehighest the records preferentially sediment precipitated The 0.3h nFg ,w oe abnt rcptto saitci the in abiotic as precipitation carbonate model we 3, Fig. In DIC oee,snetelwstrto tt tngtlimits night at state saturation low the since However, 0h. to falls

δ Ω 13 5 A = 6 12:00 nihett h uretcmoiino rai atris matter organic of nutrient-composition the to enrichment C δ 13 DIC ( Normalized Frequency through transferred carbon of ﬂuxes mass and composition C D 21:00 4 5 0 0 0 0 0 600 500 400 300 200 100 0 H PNAS Normalized F 4 org mol/kg)

Diurnal organiccarbontransfer 13 0 1 :060 20 80 0:00 18:00 12:00 6:00 0:00 bulk δ 13 G 902000 1900 | C Median forcing DIC eebr3 2019 3, December 8:00 applied in +1h = model 24:00 D 5 Median of21shallowdatasets hta ih,we eoi respiration aerobic when night, at that

2 4 rdcsaaoiewt naverage an with aragonite produces eedcniudo olwn page following on continued Legend a xhne(72)adso that show and (17–21) exchange gas 5:00 5:00 Ω 3 A δ n i–e a xhne(a exchange gas air–sea and ) 13 | C o.116 vol. DIC 0 1 2 3 +1h =

(21 sites,51days) 13 Shallow reefs (5 sites,2125days) Open ocean ( κ κ Instantaneous C DIC κ = 247 = 19 p | , δ 13 o 49 no. ol produce would mol kg composition C mol kg mol kg | -1 -1 ) -1 24435 RESPIRATION PHOTOSYNTHESIS δ Ω 13 A C .

EARTH, ATMOSPHERIC, AND PLANETARY SCIENCES environments collected in our database. (C) We use CO2SYS (10) to estimate pCO2 from the TA and DIC data. (D–F) We compute the carbon fluxes due to photosynthesis/respiration (D), carbonate precipitation/dissolution (E), and gas exchange (F) from the TA, DIC, and pCO2 observations (SI Appendix, Fig. S3). The calculated photosynthetic carbon flux (D) becomes the forcing of the engine model (Fig. 3). (G) A compilation of the 21 diurnal photosynthesis/respiration time series, normalized to the same amplitude to facilitate comparison of the timing of photosynthesis vs. respiration. (H) The diurnal organic-carbon transfer (κp), which represents the mass of carbon (per kilogram of seawater) that is sequestered in organic matter in the afternoon before aerobic respiration outpaces photosynthesis, computed for all 21 shallow reef and 5 open ocean settings in our dataset (SI Appendix, Table S3). Notice that κp typically is >10× larger in shallow reef environments than in open-ocean settings (22). We recognize that upwelling seawater often has insufficient phosphate to drive such large values of κp. For example, the average [PO4] of water 200- to 600-m-deep upwelling onto the GBB is ∼0.75 µmol/kg (SI Appendix, Fig. S7), which would support a κp of just 80 µmol/kg assuming the standard Redfield C:P = 106:1 (16). The large observed κp values likely reflect a combination of 1) active nutrient trapping on the shallow shelves (23–25) and 2) high carbon-to-nutrient ratios in banktop organic matter due to the abundance of microbial mats and their production of carbohydrate-rich extracellular polymeric substances (26–28). We take a simple empirical approach and drive our engine model with the median κp forcing observed in shallow environments (H).

effects (29). However, our mechanism is still important for bio- model forced by changes in TA and DIC measured in the same logically mediated precipitation because it explains how the δ13C waters (Fig. 6B). composition of banktop DIC—the starting material for biogenic carbonate—changes over the day. In other words, an active diur- Drivers of Banktop δ13C Variability 13 nal carbon cycle could elevate δ C in both abiotic and biotic The extent to which δ13C is elevated in carbonate sediment com- carbonate. pared to mean seawater DIC is dominated by two factors: 1) Model Validation

The engine model (Fig. 3), forced by the median κp from carbonate reef environments (Fig. 4), along with the median values A of water depth (30) and wind speed (8) on the GBB, suc- -30 10 cessfully reproduces measured rates of carbonate production 9 and air–sea gas exchange (7) (Fig. 5). In addition, the model produces carbonate with the same bulk δ13C values we have -25 8 measured in GBB sediment (Fig. 1A). The model also suc- 7 cessfully explains the rapid rise in δ13C at the platform edge (Fig. 1B), where waters transition from the weak photosyn- -20 6 13 p thetic forcing (low κp ) characteristic of the open ocean to the 5 strong forcing (high κp ) observed in shallow waters (Fig. 4H). 13 Finally, the large diurnal range in δ C of DIC observed in coral -15 4 reef environments can be predicted accurately with the engine 3

-10 2 0 100 200 300 400 500 600 Banktop water age (days) Shallow-water photosynthetic forcing ( , mol kg-1) 0 50 100 150 200 250 p

-1 4 Model prdiction A B 500 Broecker & Takahashi (1966) mol kg Carbonate Production 400 2 F = k n CaCO3 rate 300 Fractional f B 0 Carbonate 1.0 200 Flux 13 0.8 100 Final value: f = 0.62 -2 Measured τ ≈ 0.6 0 Predicted Cumulative carbonate flux ( ) 0 50 100 150 200 days -4 0:00 6:00 12:00 18:00 0:00 Fig. 5. A comparison between predictions from the carbon engine model (Fig. 3) and observational data from ref. 7 (Fig. 2). (A) The diurnal carbon Hour of day engine model, run through 250 days of simulation, effectively describes 13 rates of carbonate precipitation as a function of the days water has Fig. 6. An exploration of the δ C enrichments expected from the diurnal spent on the bank. Carbonate precipitation is parameterized as = organic carbon mechanism (A) and model comparison to observations of FCaCO3 n diurnal δ13C variability ( ). ( ) The magnitude of carbonate δ13C enrichment krate(Ω − 1) (29, 31, 32), where n ≈ 1.7 (31) and krate, which is poorly con- B A 13 strained for natural environments, is empirically optimized to fit the data relative to open-ocean DIC (∆δ C) is sensitive to the diurnal photosynthetic from ref. 7 and is 9.0 × 10−9 mol m−2 s−1.(B) Fractional carbonate flux transfer of carbon between inorganic and organic reservoirs (κp) and the FCaCO average carbon-isotope fractionation factor between organic matter and f = 3 vs. banktop water age. The asymptotic fall of the F +Fgas+Forg 13 CaCO3 local DIC (p). The δ C enrichment also is sensitive to the saturation state— 13 curve in B reflects the slow rate of air–sea CO2 gas exchange and equi- as ΩA increases, δ C enrichment falls because higher precipitation rates libration. Note that only krate, one of many parameters in the model (SI across the entire 24-h cycle cause the nighttime CaCO3 precipitation to con- Appendix, Tables S1 and S2), is tuned to data from ref. 7. Then, when the stitute a larger proportion of the total CaCO3 flux (SI Appendix, Figs. S16 model is forced with independent observation-based estimates of κp and and S17). (B) Forced by diurnal TA and DIC measurements from the same 13 p, and allowed to undergo air–sea gas exchange as a function of pCO2, the water, the engine model accurately predicts changes in the δ C of DIC from model predicts f = 0.62, which is identical to the estimate derived from the a shallow coral-reef environment in O’ahu (33). The diurnal cycles in δ13C of data in ref. 7 (Fig. 2C) and serves as a test of self-consistency in our model DIC predicted by the engine model also have been observed in other shallow system. carbonate environments around the world (34) (SI Appendix, Fig. S20).

24436 | www.pnas.org/cgi/doi/10.1073/pnas.1908783116 Geyman and Maloof Downloaded by guest on September 27, 2021 Downloaded by guest on September 27, 2021 t uili curn nsalwsevs(χ shelves δ shallow on occurring is carbon- burial global ate most if However, carbonates. shallow dissolved biology, rate, growth of function ( a matter organic is and DIC between factor fractionation vni h hneeetal sgoa nntr.Idvda hle xeinigcagsin changes experiencing shelves Individual nature. in global 3 in is the eventually Thus, change 0 the matter). a if organic even to of leads because burial matter, (the zero organic carbon to and of carbonate way sink the second a of remains all there damped not is spike low very is area shelf (χ global If duration. 20-kyr a over prompt- the change other ing or evolutionary, eustatic, the 10 is carbon of δ values organic (37), 0.23 diurnal global to active Modern 0.20 an (1). equilibrium of engine absence to carbon the corresponds in which calcite the +1.1 h, where flux, is deep ation a and 1A, Fig. observed in enrichment the empirical where the flux, +4h, is shallow tionation a carbonate into the flux divide a we burial case, canonical adopt our mod- In the 37). we a (3, use equation of (A), and 36 in version ref. curves in ified that the to similar for framework solve To ance. car- diurnal active χ the with and engines, shelves bon shallow on burial the between and Relation- ship (A) balance shelves. carbonate mass individual of global records on engine bon 8. Fig. the 2) and water, local with physical water the marine and open productivity primary of (κ both mixing of cycle function diurnal a each is during which between reservoirs carbon organic of and transfer inorganic photosynthetic the of magnitude the can which engine, carbon diurnal the of strength produce the to relative modest are of function a average the and cycling average the between difference 7. Fig. emnadMaloof and Geyman local to due rising before is system ocean–atmosphere (F the steady-state in in carbon of mass assume we Here, world. around shelves the shallow on carbonate feedback precipitated of global the illustrate by To implied (B) today. 0.2 and 13 13

erse lblDCt anangoa asbal- mass global maintain to DIC global depresses 13 δ ≈ nihetsga sdme to damped is signal enrichment C C C 13 1 2 DIC ) h hl 3h whole the 0), fDCwudproduce would DIC of C ≈ 13 (daily mean) rs-hl rdet nsediment in gradients Cross-shelf (A) δ

δ p mlctoso h ira rai car- organic diurnal the of Implications 13 0.4h 13 -10 -15 -20 -25 -30 C ofDIC esmlt 3h a simulate we A, niheto hlo hle occurs shelves shallow on enrichment C nihdb pto up by enriched C 5 κ 0 0 0 0 0 600 500 400 300 200 100 0 ie,temdr au)() n that and (3), value) modern the (i.e., y p h rcino lblcarbonate global of fraction the χ, Diurnal organiccarbontransfer( 3,3)sgetthat suggest 39) (38, and in = p −23h ≈ F p out δ h rdce rs-hl rdet of gradients cross-shelf predicted The . 13 δ niheti xrse in expressed is enrichment ,ta h eiec ieof time residence the that ), 13 δ fgoa I.Increased DIC. global of C 100 days 13 ,as (B), cycling diurnal of day 1 after produced C f enrichment. C org δ to 13 B ∼10h = eod htete pk rt eoetegoa eaiefebc depresses feedback negative global the before first, spike either that records C δ − nihetin enrichment δ 13 ,ta h total the that 0.2, 13 29h asbalance mass C The <1h. rdcdatr10dy fdiurnal of days 100 after produced C χ δ Fg 6A). (Fig. sbten0 between is 13 3 7,and 37), (3, fraction- C .2 δ ≈ 13 × frac- C ) the 1), δ f 3h 13 org δ δ δ ,wihw en sthe as define we which C, 13 13 13 p ≈ C C C = , 0.6h mol kg CO hnei lblDC (C DIC. global in change -1 −2h 2 A ) n sur- and , p A ,which ), o+1h to δ -2 -1 0 1 13 p C ), Cross-shelf 13 ,bttedmnn rgno ohita n intershelf and intra- both of origin dominant the but 7B), (Fig. for choices cross-shelf Reasonable the lish zero. the on to of escape close effect is gas DIC isotopic and net precipitation The carbonate 2C). coupled chem- and (Fig. water (7) from DIC, inferred measurements values banktop istry the to in identical is reduction estimate This the of 62% CO for accounts tion of function a seawater of of degassing function the to tied −9. inextricably is (1) Reaction CO air–sea precipitation, and carbonate (16), of photosynthesis Stoichiometry 1. Table δ isotope high carbon induce 4) +7h (Fig. to +4 environments of enrichments reef modern in of choices Reasonable (35). area/volume face 3C Fig. in htsnhss16CO 106 Ca Photosynthesis Precipitation a exchange Gas h 13 tmyse oneitiieta otne rdcinof production continued that counterintuitive seem may It m niheto hlo hle,wihicessthe increases which shelves, shallow on enrichment .Hwvr h rcptto of precipitation the However, C. hnein change A ) 2 7h eoe the denotes δ a saeacut o h remaining the for accounts escape gas 13 . abnt osntdieabetwtr oadlower toward waters ambient drive not does carbonate C 4) u niemdl negiggsecag sa as exchange gas undergoing model, engine Our (40). δ 13 nihetbfr ratrtegoal nue depression induced globally the after or before enrichment C Ω κ I fterato,adcrepnst h slope the to corresponds and reaction, the of ∆TA:∆DIC PNAS δ p 3,3,4) rdcsta abnt precipita- carbonate that predicts 41), 32, (31, 13 2+ and (CH rdet nDCo pt e e mil per few a to up of DIC in gradients C HCO 2 + | p 2 p 2 )0 (NH O)106 CO eebr3 2019 3, December a o tr ttesm oeteverywhere, moment same the at start not may 6HNO 16 + CO 2 δ 3 Fg 6A). (Fig. 1)adpeiiaigcroaeas carbonate precipitating and (17) 2 13 − (aq) ,o lnefis rmtegoa signal, global the from first plunge or C, → Equation 3 3 1 (H )16 CaCO → H + 2 CO a exchange gas CaCO 3 PO 3 3 2 | PO (atm) CO + 4 o.116 vol. 4 2 H 122 + 3 O 138 + ) 3 2 CO a)+H + (aq) with κ p p 7 Fg 5B). (Fig. ∼37% δ and 13 2 or | 2 fstbetween offset C O with 2 o 49 no. B arag → κ p 2 O p a estab- can +2 = C observed sea−air δ 16:−106 −2: | 13 :1 0: 24437 m of C .7h −1 ∼

EARTH, ATMOSPHERIC, AND PLANETARY SCIENCES A unrealistic changes in global carbon cycling (51), and instead may hint at previously unrecognized global forcings to κp and/or p . Certainly not all δ13C excursions in the geologic record of shallow carbonates were driven by synchronous changes in κp or p . However, the diurnal organic carbon engine (Fig. 3A) can suc- cessfully 1) modulate global DIC, forced by tectonic or eustatic controls on shelf area or biologically induced changes in vital B effects, p , and κp ; and 2) explain how a global forcing may be variably expressed with different peak δ13C values (48, 52). Conclusions Much of our understanding of ancient Earth history comes from shallow carbonates, as open marine records tend to be subducted at plate margins. However, it is time to recognize that the δ13C record of ancient shallow carbonates may not always directly reflect global carbon cycling. Recent work on marine (46), mete- oric (53), and burial (54) diagenesis has illuminated the role of postdepositional alteration. Here, we present a mechanism to Fig. 9. (A) δ13C record through the Cambrian, based on the compilations explain how the primary δ13C values of the shallow carbonate of refs. 39 and 43. (B) The δ13C variability, computed as the SD (1σ) of all sink can be partially decoupled from global DIC. Specifically, we 13 δ C values in a moving window of size τ = 3.5 Ma (43), is strongly anti- have shown how ordinary diurnal carbon cycling is sufficient to correlated with the volume of carbonates preserved in the geologic record 13 13 produce the most anomalously positive δ C on shelves today— of North America (44). Although δ C variability is a function of the chosen and perhaps in the geological record—without any net produc- window size, a wide range of window sizes from τ = 0 − 10 Myr yield strong tion of oxygen. Although the diurnal carbon engine and subse- Pearson coefficients ρ(x, y) = cov(x,y) with magnitudes greater than 0.6 (SI σx σy quent diagenesis complicates interpretations of carbon-isotope Appendix, Fig. S13). chemostratigraphy, frequent comparison of modern analogue studies to ancient records will help resolve shallow carbonate 13 variability in carbonate δ13C values might result from spatiotem- δ C into a more refined chronicle of paleoenvironment and poral variability in κp and/or p . For example, the full range Earth history. of κp values depicted in Fig. 4H is observed in different envi- Materials and Methods ronments from a single coral reef (42) (SI Appendix, Fig. S15), Carbonate surface sediments from North Andros, San Salvador, and Shark and p may vary from −5 to −25 depending on the relative Bay (SI Appendix, Fig. S1) were rinsed 3 times in deionized water, dried, abundance of photoautotrophsh suchh as cyanobacteria, algae, and homogenized using mortar and pestle, and then placed in individual borosil- seagrass (5). icate glass reaction vessels. The samples were heated to 110◦C to remove volatiles, capped and flushed with helium to remove atmospheric gas Implications for the Global Carbon Cycle and the from the reaction vessels, and finally reacted at 72◦C in a GasBench II Interpretation of the δ13C Record preparation device coupled to a Sercon continuous-flow isotope-ratio mass 13 δ13 spectrometer. The precision and accuracy of δ C measurements are moni- The C of shallow carbonates is a crucially important tool tored through analysis of 15 standards for every 57 measured samples. δ13C for global chemostratigraphic correlation, especially before the data are reported in the standard delta notation relative to Vienna Pee appearance of index animal fossils (39, 43, 45). If κp and p vary Dee Belemnite. Average precision is <0.1 (1σ). All sediment δ13C mea- 13 within or between shelves, their control on primary δ C would surements from North Andros, San Salvador,h and Shark Bay (Fig. 1A) are compound δ13C gradients driven by diagenesis (46) or the influx provided in a spreadsheet (Dataset S1). of terrestrial organic matter (47, 48), adding to the challenges ACKNOWLEDGMENTS. Thank you to J. Birch at Small Hope Bay Lodge for of chemostratigraphic correlation. In contrast, if κp and p are making work possible on Andros Island. Also, thank you to A. Cartwright, controlled by evolutionary, eustatic, and/or climate changes, car- R. Coakley, N. Hinsey, A. Mackey, A. Marshall, S. Martin, B. Neymor, bonate shelves might respond synchronously with similar shifts G. Thompson, L. Whyms, local customs and immigration, and the Bahamas in δ13C (49). As the area of shallow shelves increases, the burial Environment, Science & Technology Commission. C. Allen at Air Flight Char- 13 13 ters and D. Reading at Princeton provided logistical support. We thank of high δ C carbonates will lower the δ C of seawater DIC to L. O’Connor and T. Humes for field assistance in the Bahamas. We thank satisfy global isotope mass balance (Fig. 8A). This negative feed- B. Dyer and B. D’Andrea for assistance in San Salvador and D. Holley back means that, if a global forcing increases the strength of the and the Fenny family in Shark Bay. A. Gagnon inspired us to adapt mod- diurnal carbon engine on all shelves, the magnitude of the posi- els for carbonate system dynamics within individual corals to the scale 13 of an entire carbonate bank. Conversations with B. Dyer, M. Bender, tive δ C shift in shallow carbonate will be damped as a function W. Broecker, D. Sigman, W. Fischer, J. Grotzinger, J. Husson, J. Dunne, of global carbonate shelf area (Fig. 8B). The “falling amplitude J. Higgins, A.-S. Cruger¨ Ahm, and J. Strauss improved the manuscript. of carbon isotopic oscillations” through the Cambrian (Fig. 9A), We thank P. Swart and A. Oehlert for generously sharing their data observed by ref. 50, might be one geologic example of this nega- and for thought-provoking discussions. Feedback from Don Canfield, Brad Rosenheim, and 2 anonymous reviewers significantly improved the tive shelf-area feedback; as the volume of preserved carbonates manuscript. This material is based on work supported by NSF Division of in the geologic record of North America—an imperfect approx- Earth Sciences Grant 1410317 and by the Princeton Environmental Institute imation of global shallow carbonate area—increases during the at Princeton University through the Smith-Newton Scholars Program. This Cambrian (44), the magnitude of δ13C oscillations falls (Fig. 9B). work also was supported by the Geological Society of America Stephen G. Pollock Student Research Grant, the Evolving Earth Foundation, the High Ascribing the 1- to 3-Myr positive excursions in Fig. 9A to diurnal Meadows Foundation, the Sigma Xi Research Society, and the Princeton carbon-engine phenomena could alleviate the need to call upon Geosciences Student Research Fund.

1. M. Rubinson, R. Clayton, Carbon-13 fractionation between aragonite and calcite. 3. L. R. Kump, M. A. Arthur, Interpreting carbon-isotope excursions: Carbonates and Geochem. Cosmochim. Acta 33, 997–1002 (1969). organic matter. Chem. Geol. 161, 181–198 (1999). 2. C. Romanek, E. Grossman, J. Morse, Carbon isotopic fractionation in synthetic arago- 4. P. K. Swart, J. J. G. Reijmer, R. Otto, A re-evaluation of facies on Great Bahama Bank nite and calcite: Effects of temperature and precipitation rate. Geochem. Cosmochim. II: Variations in the δ13C, δ18O and mineralogy of surface sediments. Int. Assoc. Acta 56, 419–430 (1992). Sedimentol. Spec. Publ. 41, 47–59 (2009).

24438 | www.pnas.org/cgi/doi/10.1073/pnas.1908783116 Geyman and Maloof Downloaded by guest on September 27, 2021 Downloaded by guest on September 27, 2021 2 .C hw .R hn,B ibok .See,Ntrli iurltosissuggest relationships situ in Natural Steven, A. Tilbrook, B. Phinn, R. S. Shaw, C. E. 12. Keeling D. C. 11. 3 .C hw .I cel .Tlro,Ipcso ca cdfiaini aual variable naturally in acidification ocean of Impacts Tilbrook, B. McNeil, I. B. Shaw, C. E. 13. CO for developed Program Allison, J. L. Wallace, D. Lewis, E. 10. 5 .S esn .I ile,M .Kc,I eVcne,Popou ees ihcarbon- with release Phosphorus Vicentea, de I. Koch, S. M. Nielsen, and I. O. cycling review Jensen, biogeochemical S. marine H. Literature the in 25. seagrasses: margins continental on of Importance nutrients Walsh, J. sediment J. 24. of Effects Short, T. F. 23. Sutton A. 22. Nightingale D. P. 21. ocean the over exchange gas and speed wind between Relationship ocean. Wanninkhof, the R. over exchange 18. gas and speed wind between Relationship Wanninkhof, R. relation 17. their and water sea in derivatives organic of proportions the On Redfield, A. calcifi- 16. reef coral between Hysteresis Eyre, D. B. Cyronak, T. Santos, R. I. McMahon, A. metabolism 15. reef Coral Anthony, K. Caldeira, K. Cantin, N. Benthuysen, J. Albright, R. 14. 9 .Ce,A ann .Akn,Croi nyrs,crlclicto n new a and calcification coral anhydrase, Carbonic Adkins, microbial J. of fingerprints Gagnon, Isotopic A. Swart, Chen, K. P. S. Reid, P. 29. R. Sumner, Y. D. Andres, S. M. 28. B S. Veuger, B. Cook, P. 27. Pannard A. 26. in synthesis” and Introduction rates: exchange gas “Air-sea Merlivat, L. Liss, S. P. 20. Ho T. D. 19. 0 .C emn .C aof ipemto o xrcigwtrdphfo mul- from depth water extracting for method simple A Maloof, C. A. Geyman, C. E. 30. emnadMaloof and Geyman .P wr,G bri h aueo the of nature The Eberli, G. Swart, P. 6. Oehlert M. A. 5. .K efys abnt qiira rpi n leri approach. algebraic and graphic bathymetry A Mapping equilibria: Reijmer, G. J. Carbonate J. Swart, Deffeyes, K. P. K. Ellis, J. 9. Purkis, J. banks. S. Bahama Harris, the M. M. on P. precipitation 8. carbonate Calcium Takahashi, T. Broecker, W. 7. Hawaii. cycle. carbon global the reconstructing Sedimentology for Implications sediments: derived platform oa efclimcroaepouto ildciewt ca acidification. ocean with decline Oceanogr. will production carbonate calcium reef coral Ridge Oak 1998). Center, Laboratory, Analysis National Information Dioxide Carbon ORNL/CDIAC-105, Rep. (Tech. Oceanogr. Bank. Bahama Great on facies depositional and Res. Geophys. J. cycle. carbon global the and stratigraphy oa effltecosystems. flat reef coral t islto ope osld xdto nFoiaBysars sediments. seagrass Bay Florida in Oceanogr. oxidation sulfide to coupled dissolution ate nitrogen. and carbon of bounds. experiment. preindustrial mesocosm outside chemistry carbonate ocean Biogeosciences surface forces change tracers. volatile (2000). and conservative novel using Buat-M P. Cycling, Geochemical 113–127. pp. in 1986), Company, Exchange Publishing Air-Sea of Role revisited. Oceans Res. Geophys. J. 176–192. pp. 1934), Liverpool, of plankton. of composition the to state. (2013). saturation aragonite seawater the and cation flat. reef coral a (2015). of dynamics chemistry carbon and oe fsal stp ia effects. vital isotope stable of model stromatolites. Bahamian from aragonite in respiration near-shore in bacteria and algae benthic sediment. through sandy flows and incorporation nitrogen matter. organic particulate the (2016). of ratio carbon-to-nutrient parameterizations. (2006). global L16611 for Implications Ocean: Southern ipcrlstlieiaeyi ein fvral otmtype. bottom variable of regions (2019). 527–537 in imagery satellite tispectral Tellus esrmnso i-e a xhnea ihwn pesi the in speeds wind high at exchange gas air-sea of Measurements al., et inlOeng.Methods Oceanogr. Limnol 7–8 (2015). 777–788 60, (1965). 412–426 10, 7316 (2009). 1753–1764 54, sn rsn-a bevtost eetwe anthropogenic when detect to observations present-day Using al., et rdcino xplmr ES ycaoatra mato the on Impact cyanobacteria: by (EPS) exopolymers of Production al., et topei abndoievrain tMuaLaobservatory, Loa Mauna at variations dioxide carbon Atmospheric al., et 3–5 (1976). 538–551 28, h tbecro stpccmoiino rai aeilin material organic of composition isotopic carbon stable The al., et 0558 (2006). 5065–5083 13, 1–3 (2012). 319–335 59, 5510 (1966). 1575–1602 71, qa.Mco.Ecol. Microb. Aquat. nst vlaino i-e a xhneparameterizations exchange gas air-sea of evaluation situ In al., et e,J idlug feto uretaalblt ncro and carbon on availability nutrient of Effect Middelburg, J. oer, ¨ qa.Bot. Aquat. 3378 (1992). 7373–7382 97, Nature .Gohs e.Oceans Res. Geophys. J. 35 (1991). 53–55 350, ae ontn eoilVolume Memorial Johnstone James 15 (1987). 41–57 27, ece.Csohm Acta Cosmochim. Geochem. δ 5–6 (2014). 351–362 12, 6–8 (2007). 165–180 49, 13 fprpafr eiet:Ipiain for Implications sediments: periplatform of C eiet Geol. Sediment. lb igohm Cycles Biogeochem. Glob. Sedimentology 008(2012). C03038 117, epy.Rs Lett. Res. Geophys. epy.Rs Lett. Res. Geophys. Geology 1–2 (2005). 115–129 175, epy.Rs Lett. Res. Geophys. qa.Ecol. Aquat. nr,E.(.Reidel (D. Ed. enard, ´ 6–8 (2015). 566–589 62, 7–7 (2006). 973–976 34, 2 7–9 (2018). 179–197 236, at pc Sci. Space Earth ytmcalculation system Uiest Press (University 3980–3988 42, 4675–4679 40, 373–387 14, 29–44 50, Limnol. Limnol. Limnol. The 33, 6, 0 .Seetae,K unc,“ Munnich, K. Siegenthaler, U. in 40. stratigraphy” isotope “Carbon Thomas, E. Saltzman, R. M. 39. 4 .A er,Abra ignssoii o h daaa hrmWnk carbon Shuram–Wonoka Ediacaran the for origin diagenesis burial A altered Derry, meteorically A. of L. analysis Freeman, 54. probabilistic A H. Maloof, K. C. A. Pancost, Higgins, A. J. D. R. Dyer, B. Arthur, 53. A. M. Slupik, M. L. Patzkowsky, E. M. 52. oscillations isotopic carbon Maloof of C. A. amplitude 51. falling The Sukhov, S. S. Brasier, car- D. M. of composition 50. isotopic carbon the iso- in carbon changes synchronous sea Global epeiric Swart, the K. P. of 49. Sensitivity Kump, R. L. Holmden, C. Panchuk, M. K. carbonate 48. modern on DIC seawater in 13C of Depletion Walter, M. L. Patterson, P. W. 47. Higgins J. 46. Halverson data P. geological G. for platform 45. A Macrostrat: Czaplewski, J. Husson, M. J. Peters, E. S. 44. Maloof C. A. 43. a at rates calcification solutions. and dilute photosynthesis in of Estimation kinetics Nakamori, Aragonite T. Grossman, Nakamura, T. L. E. 42. Morse, W. J. Romanek, S. C. 41. of distribution The Kroopnick, P. 38. of abundance The Kaufman, J. A. Strauss, and H. carbonate Hayes, Authigenic M. Johnston, J. T. D. 37. Macdonald, A. F. Higgins, A. J. Schrag, P. CO D. rate, 36. growth of Effects Zondervan, I. Riebsell, U. Burkhardt, S. 35. 4 .Wbr .Woha,Dunlvrain nteiooi opsto fdissolved of composition isotopic the in variations Diurnal Woodhead, P. Weber, J. groundwater 34. Submarine cratonic Fackrell, J. and Ruttenberg, state K. Popp, saturation B. Dulai, mineral H. Richardson, Carbonate C. Wilkinson, cal- 33. H. Mg B. and Opdyke, aragonite N. of B. rates 32. precipitation Relative Walter, M. L. Burton, A. E. 31. .M rdti,J .Og .Shiz .Og d.(leir 02,pp. 2012), (Elsevier, Eds. Ogg, G. Schmitz, M. Ogg, G. J. Gradstein, 207–232. M. F. 2012, stp anomaly. isotope (2017). 135–138 45, patterns? δ the Cambrian of of Harbinger continuation extinction: or and (1997). change ordovician environmental late ordovician middle Late (2010). 623–626 boundary Nemakit-Daldynian-Tommotian the of data. Siberia Northern Cambrian: (1998). Middle 353–373 to lower the through cycle. carbon global the U.S.A. in Sci. changes to Sea. unrelated Mohawkian sediments bonate the from Tales processes: cycle Palaeoecol. carbon Palaeoclimatol. local-scale Palaeogeogr. to record tope carbonates. (1994). of 888 record isotopic carbon the for Significance platforms: sediments. carbonate water Bull. Am. Soc. Geol. research. crust (2018). earth 1393–1409 deep-time and integration Island, Ishigaki reef, Shiraho the change. for study case A slope: com- reef and Japan. coral southwest flat of zonation reef the on and variations munities diurnal for accounting by reef fringing Geochem. Aquat. 249–257. in sea” Pap. Res. Oceanogr. the A during carbon of cycle biogeochemical Ma. global 800 the past in fractionation isotopic and cycle. carbon global the phytoplankton. of history marine the in fractionation isotope carbon stable Acta the Cosmochim. Geochem. on size cell and nrai abni ewtrfo oa efenvironments. reef coral from Acta seawater in carbon inorganic (2017). reefs. 5363 Hawaiian two in biogeochemistry drives discharge accumulation. limestone control? ion carbonate or Temperature (1987). seawater: from cite 13 hmsrtgah rmlt aezi c g abnt platforms. carbonate age ice Paleozoic late from chemostratigraphy C 9–0 (1971). 891–902 35, abnCceMdln:SOE16 SCOPE Modeling: Cycle Carbon el o.A.Bull. Am. Soc. Geol. 34–34 (2008). 13741–13745 105, ieaoy al aiedaeei,adteceityo shallow- of chemistry the and diagenesis, marine early Mineralogy, al., et h aletCmra eodo nml n ca geochemical ocean and animals of record Cambrian earliest The al., et osrit neryCmra abnccigfo h duration the from cycling carbon Cambrian early on Constraints al., et hm Geol. Chem. oadaNortrzi opst abniooerecord. carbon-isotope composite Neoproterozoic a Toward al., et at lnt c.Lett. Sci. Planet. Earth oa Reefs Coral 3–5 (2011). 339–356 17, PNAS 1110 (2005). 1181–1207 117, 78 (1985). 57–84 32, m .Sci. J. Am. | 0–2 (1999). 103–125 161, 7934 (1999). 3729–3741 63, ece.Csohm Acta Cosmochim. Geochem. 7117 (2010). 1731–1774 122, eebr3 2019 3, December 2–5 (2009). 229–250 28, 13 13 C/ of C 12 1–3 (1993). 217–234 293, rcinto uigCO during fractionation C Science ΣCO 5–6 (2010). 152–162 294, .Bln d Jh ie os 91,pp. 1981), Sons, & Wiley (John Ed. Bolin, B. , 2–3 (2005). 320–337 228, 2 ntewrdoceans. world the in 4–4 (2013). 540–543 339, | δ ece.Gohs Geosyst. Geophys. Geochem. 13 o.116 vol. hf,Morocco. shift, C 13 inl Oceanogr. Limnol. 1–3 (2017). 512–534 220, nmrn rai matter organic marine in C Geology ece.Cosmochim. Geochem. Geology 2 | elgcTm Scale Time Geologic a.J at Sci. Earth J. Can. rnfrfo i to air from transfer epSaRs Part Res. Sea Deep o 49 no. Geology rc al Acad. Natl. Proc. 2 concentration, 911–914 25, Geology 111–114 15, 5348– 62, | Geology 885– 22, 24439 38, 35, 19,

EARTH, ATMOSPHERIC, AND PLANETARY SCIENCES