Controls on the abundance, provenance and age of organic carbon buried in continental margin sediments.

Ea h and Plane a y Science Le e s 558 (2021) 116759 Con en s lis s a ailable a ScienceDi ec Ea h and Plane a y Science Le e s www.else ie .com/loca e/epsl Con ols on he abundance, p o enance and age o o ganic ca bon bu ied in con inen al ma gin sedimen s Blanca Ausína,b,∗, Elena B unia, Nega Haghipou a,c, Ca oline Wel ea,c, S e ano M. Be nasconia, Timo hy I. Eglin ona aGeological Ins i u e, ETH Zu ich, Swi ze land bGeology Depa men , Salamanca Uni e si y, Salamanca, Spain cLabo a o y o Ion Beam Physics, ETH Zu ich, Swi ze land a i c l e i n o a b s a c A icle his o y: Recei ed 20 Ma ch 2020 Recei ed in e ised o m 7 Janua y 2021 Accep ed 9 Janua y 2021 A ailable online xxxx Edi o : Y. Asme om Keywo ds: o ganic ca bon adioca bon age hyd odynamic so ing g ain size con inen al ma gins Con inen al ma gins play a undamen al ole in he ca bon cycle as p ima y oceanic loca ions o o ganic ca bon (OC) bu ial. Howe e , gaps emain in ou unde s anding o ac o s con olling he dis ibu ion and p ese a ion o o ganic ma e (OM) in hese he e ogeneous and dynamic sys ems. In pa icula , he impac o hyd odynamic p ocesses on he age, abundance, and s able iso opic composi ion o sedimen a y OC is poo ly cons ained. He e, we cha ac e ize he OC p esen in bulk and g ain-size sedimen ac ions om se en con inen al ma gin se ings. Ou esul s e eal ha hyd odynamic pa icle so ing p ocesses exe a ubiqui ous influence on he adioca bon age o OC. Bo h, hyd odynamic cha ac e is ics o mine al pa icles and he na u e o hei in e ac ions wi h OM influence sedimen a y OC con en , whe eas no significan influence o ei he e ec is mani es ed in co esponding δ13C alues. Since OC p e e en ially esides wi hin he fine sil ac ion (2-8 μm), and his ac ion accoun s o a subs an ial ac ion o he bulk sedimen mass, ansloca ion and subsequen e-deposi ion o dis an fine sil has he g ea es po en ial o dis o local OC signa u es ela i e o hose associa ed wi h clay o coa se sil ac ions. We sugges ha he magni ude o di e ences in 14C-age and OC con en among g ain-size ac ions, de e mined by he in e play o hyd odynamic so ing and o he si e-specific p ocesses, allow h ee di e en ca ego ies o deposi ional en i onmen o be defined: ini ial, s able, and ma u e. Each domain is cha ac e ized by di e en deg ees o e ical and la e al OC supply ha eflec influences o local biological p oduc i i y and ca bon expo om o e lying su ace wa e s and physical o cing ha d i e hyd odynamic p ocesses. This gene ic amewo k may se e as a guide o efine assessmen o OC bu ial and o cons ain he magni ude o po en ial aliasing among co-e al p oxy signals in con inen al ma gin sedimen a y sequences. ©2021 The Au ho (s). Published by Else ie B.V. This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/by/4.0/). 1. In oduc ion Al hough con inen al ma gins accoun o only 10-20% o he global ocean floo , hese egions s o e ca. 90% o he o ganic ca - bon (OC) p ese ed in ma ine sedimen s (Hedges and Keil, 1995; P emuzic e al., 1982), and hus play a key ole in he global ca - bon cycle as majo OC sinks. Mo eo e , con inen al ma gins a e cha ac e ized by apid sedimen accumula ion, and hus o ganic signa u es p ese ed in ma ine sedimen sequences e ie ed om hese spa ially he e ogeneous and highly dynamic egions a e o s a egic alue in s udies o con inen -ocean in e ac ions and pa- *Co esponding au ho a : Geology Depa men , Salamanca Uni e si y, Sala- manca, Spain. E-mail add ess: [email p o ec ed] (B. Ausín). leoclima e a ia ions a cen ennial o e en highe empo al es- olu ion. A de ailed unde s anding o he p ocesses a ec ing OC deposi ion and bu ial in hese con inen -ocean bounda y en i on- men s is a p e equisi e bo h om he pe spec i e o unde s anding he C cycle and in e p e a ion o sedimen a y eco ds. Di e se, and equen ly coupled, ac o s con ibu e o he e - ficiency o emine aliza ion and na u e o o ganic ma e (OM) p ese ed in ma ine se ings. Sedimen a ion a e, oxygen exposu e ime (OET) and mic obial ac i i y, o name a ew, ha e been ound o influence OC bu ial efficiency (e.g., A nd e al., 2013; Mülle and Suess, 1979). The obse a ion ha OC con en is b oadly co - ela ed wi h mine al-specific su ace a ea o con inen al ma gin sedimen s led Maye (1993) o p opose OM so p ion on mine al su aces as a mechanism o i s long- e m physical p o ec ion. Fu he wo k highligh ed he p e e en ial associa ion o OM wi h fine-g ained, la ge-su ace-a ea mine als, as hese p o ide s onge h ps://doi.o g/10.1016/j.epsl.2021.116759 0012-821X/©2021 The Au ho (s). Published by Else ie B.V. This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/by/4.0/). B. Ausín, E. B uni, N. Haghipou e al. Ea h and Plane a y Science Le e s 558 (2021) 116759 Fig. 1. Loca ions o he su ace sedimen samples analyzed in his s udy and sample SHAK06-5K; 2-3 cm om Magill e al. (2018). The co esponding ac onym o each si e is shown wi hin b acke s. p o ec ion agains deg ada ion (Hedges and Keil, 1995; Keil and Maye , 2014; Keil e al., 1994b; Maye , 1994a,b). Howe e , such associa ions wi h fine-g ained mine als may expose OC o u he emine aliza ion i s ong hyd odynamic o cing comes in o play (Bao e al., 2016, 2018a,c). In addi ion o con olling OC con en and composi ion (Be gamaschi e al., 1997), di e en mine al pa - icles may exhibi cohesi e beha io o may be p one o selec i e mobiliza ion depending on hei size, and on p e ailing hyd ody- namic condi ions (e.g., shea s ess). Besides, he po en ial impac s o hyd odynamic so ing p o- cesses on OM dispe sal and bu ial o e con inen al ma gins, hese p ocesses aise conce ns wi h espec o he fideli y o o ganic sig- na u es p ese ed in con inen al ma gin sedimen eco ds (In ho n e al., 2006). Such p ocesses may induce spa ial and empo al o - se s be ween sedimen a y cons i uen s esiding in di e en g ain size ac ions. Indeed, la e al sedimen anspo p ocesses ha e been in oked o explain olde 14C ages o bulk OC ela i e o co- deposi ed plank onic o amini e a (e.g., Ausín e al., 2019; Kusch e al., 2010; Mollenhaue e al., 2005; Ohkouchi e al., 2002). Defini- i e e idence o his hypo hesis equi es in-dep h in es iga ions a he mine al g ain-size le el in o de o a ibu e e ec s o hy- d odynamic so ing. Recen s udies ha e e ealed he la ge ex en o which OC anspo , ( e)dis ibu ion and p ese a ion a e influ- enced by hyd odynamic so ing p ocesses in he shallow Chinese ma ginal seas and on he Washing on ma gin (Bao e al., 2019a, 2016, 2018b). The impac s o hese p ocesses on eco ded paleo- clima e signals in a sedimen co e om he Ibe ian ma gin ha e also been documen ed (Magill e al., 2018). Howe e , hese s ud- ies emain limi ed o ew specific egions, and we cu en ly lack knowledge o he b oade ele ance o hese p ocesses wi h e- spec o sedimen a ion and OC bu ial on con inen al ma gins. He e, we examine geochemical cha ac e is ics (OC-con en , - adioca bon age, -ca bon iso opes, C/N a io, mine al su ace a ea and g ain size) o disc e e sedimen g ain-size ac ions collec ed om a ange o con inen al ma gin se ings and one deep-sea sedimen d i in o de o assess he p e alence and na u e o hyd odynamic mine al-so ing influences on sedimen a y OC. 2. Ma e ial and me hods 2.1. Sample desc ip ion We selec ed sedimen samples om con inen al ma gin sys- ems spanning a ange o hyd og aphic and deposi ional se ings ha a y wi h espec o su ace ocean p oduc i i y, sedimen a- ion a e, and oxygen le els. All si es a e dis al om he influence o majo i e s o diminish he po en ial con ibu ion o e es ial OC and cha ac e ized by apid sedimen a ion (>20 cm ky −1) o minimize bio u ba ion e ec s (Fig. 1). The Pe u ian ma gin in he equa o ial Pacific is one o he mos p oduc i e egions o he Wo ld Oceans (100–400 g C m−2y −1, (Suess, 1973)) as a esul o wind-d i en upwelling. Sample KNR 182/9 MC13 (abb e ia ed he e o “PER”) was e ie ed om he ou e edge o he con inen al shel , whe e la ge fluxes o labile ca bon and esul ing oxygen-deficien wa e s esul in e y high bulk OC con en s o su ace sedimen s (up o 16%) while OM de- composi ion hinde s ca bona e p ese a ion (A hu e al., 1998). The Namibian ma gin in he sou heas A lan ic Ocean also sus- ains high biological p oduc i i y and an in ense oxygen minimum zone (OMZ) o e he shel (100-400 m dep h) due o he pe - manen Benguela Upwelling Sys em (Moffi e al., 2015; Schim- melmann e al., 2016). Sample Mi abilis 2017 26-90 (“NAM”) was e ie ed in oxic wa e s om he mid-slope, whe e hemipelagic sedimen a ion, including subs an ial la e al anspo o aged OM has p e iously been documen ed (In ho n e al., 2006). The NW A ican ma gin in he sub opical eas e n A lan ic is pa o he Cana y Upwelling Sys em, cha ac e ized by high o mode a e a es o biological p oduc i i y. Sample OC437 MC31 (“NAF”) was e ie ed no h o he OMZ ha ex ends sou h o 21◦N (Moffi e al., 2015) and selec ed conside ing his egion may ex- hibi minimal hyd odynamic biases gi en synch onous signals ha e been obse ed be ween OC and plank onic o amini e a in unde - lying sedimen s (Mollenhaue e al., 2005), con a y o modeling s udies (Fische e al., 2009). Sample OCE326 MC2 (“NAT”) was eco e ed om a shallow de- p ession loca ed on he sou he n New England Mud Pa ch, in he Mid-A lan ic Bigh con inen al shel . This egion sus ains high o mode a e p oduc i i y and is cha ac e ized by high a es o fine- g ained ac i e deposi ion (Twichell e al., 1981). The San a Ba ba a Basin and San a Monica Basin a e loca ed nea he coas on he Cali o nia ma gin in he no heas Pacific Ocean, wi hin a majo eas e n bounda y upwelling sys em. High p oduc i i y and pa ially es ic ed ci cula ion esul ing in oxygen- deple ed bo om wa e s p omo e deposi ion o a ed sedimen s wi h minimum bio u ba ion in hese bo de land basins (Schim- melmann e al., 2016). Sample SBB4 MC43 (“SBB”) was eco e ed om he pa ially isola ed and silled San a Ba ba a Basin, cha ac- e ized by sub-oxic o anoxic bo om wa e s and whe e e idence o sedimen ocusing and deposi ion o p e-aged OC has p e iously been documen ed (Mollenhaue and Eglin on, 2007). Sample SMB1 MC49 (“SMB”), loca ed in he San a Monica Basin, was eco e ed om he well-defined OMZ be ween 500 and 900 m dep h (Mo - fi e al., 2015). The Be muda Rise is a deep-ocean sedimen d i in he sub- opical no hwes A lan ic Ocean. Al hough his si e is no loca ed on a con inen al ma gin, sample OCE326 BC9 (“BER”) was selec ed o his s udy gi en ha he la ges 14C age disc epancies be ween OC and co-deposi ed plank onic o amini e a ha e been epo ed he e (Ohkouchi e al., 2002), implying s ong hyd odynamic con ol on sedimen p ope ies in his egion. Resul s p e iously epo ed om sample SHAK06-5K; 2-3 cm (“IBE”), e ie ed om he so-called “Shackle on Si es” o he sou hwes Ibe ian Ma gin (Magill e al., 2018) we e included in his s udy. This si e is loca ed on he lowe pa o he slope and ea u es high sedimen a ion a es due o high o mode a e p ima y p oduc i i y and sedimen ocusing. Excep o he Ibe ian ma gin, which was sampled using a kas- en co e , he sedimen -wa e in e ace was p ese ed a all si es, sampled using a mul i-co e (PER, NAM, NAF, NAT, SBB and SMB) and a box-co e (BER). 2.2. G ain-size ac iona ion Sedimen co es we e sec ioned on-boa d e e y 1 cm, and sealed and s o ed in glass o plas ic con aine s a −20 ◦C un il hey we e 2 B. Ausín, E. B uni, N. Haghipou e al. Ea h and Plane a y Science Le e s 558 (2021) 116759 u he p ocessed o analyses. Be ween 2 o 5 op cm om each co e (Table 1) we e eeze-d ied, homogenized, and ac iona ed in o ou di e en g ain size ac ions, namely sand (>300-63 μm); coa se sil (63-10 μm [CS]); fine sil (10-2 μm [FS]); and clay (<2μm). Addi ionally, plank onic o amini e a we e picked om a sub- ac ion (250-300 μm) o he sand ac ion. The g ain-size ac- ions we e ob ained ollowing he p o ocol in oduced by Magill e al. (2018). Only Milli-Q wa e was used o all s eps du ing sample p epa- a ion. P io o use, all glasswa e was hea - ea ed o e nigh a 450◦C, whe eas sie es and o he me al ools we e cleaned wi h o - ganic sol en s. Be ween 40-200 g o d y sedimen pe sample we e we -sie ed wi h a sie ing owe using sie e mesh-sizes o 20, 63, 250, 300 μm. The sand ac ion con aining mos ly o amini e a and he CS (20-63 μm) ac ion we e ans e ed o glass ja s. Ma e- ial passing he 20 μm was hen ans e ed o cen i uge ials and cen i uged a 160 RCF 2 min a maximum speed and 2 min o slow down. The supe na an con aining he clay ac ion (<2μm) was eco e ed and ans e ed in o glass ja s. This cycle was e- pea ed un il he supe na an was clea . To sepa a e he fine sil ac ion (2-10 μm) om he lowe end o he CS ac ion (10- 20 μm), he emaining ac ion (2-20 μm) was se led in a olu- me ic flask. Re e ence se ling eloci ies o he 10 μm cu o we e adop ed om Gibbs e al. (1971). Acco dingly, he 2-10 μm ac- ion was pipe ed and ma e ial be ween 10-20 μm was combined wi h he 20-63 μm ac ion o ob ain he CS ac ion (10-63 μm). All ac ions we e hen immedia ely ans e ed o glass ja s and ozen a −20◦C p io o eeze-d ying o u he analyses. The sand ac ion was excluded om he s udy as mic oscopic analy- sis showed ha his ac ion con ained disc e e pa icula e o ganic deb is no bound o mine al su aces (Keil e al., 1994b; Ransom e al., 1998) as well as OM bound wi hin plank ic o amini e a. As he la e a e assumed o se le apidly wi hou subs an ial la e al ad ec ion, hei p esence may mask he impac o hyd odynamic so ing on mine al associa ed OM wi hin his ac ion. 2.3. G ain size and mine al-specific su ace a ea (SA) analyses Be ween 100-200 mg o eeze-d ied bulk sedimen we e hea ed in an o en a 350 ◦C o 12 h and cooled down o 40 ◦C a 50◦C pe hou in o de o emo e OM. Samples we e subsequen ly suspended in a solu ion o 4 mL o sodium hexame aphospha e (1 g L−1) in Mili-Q wa e and kep in a shaking pla o m un- il g ain-size analyses we e pe o med wi h a Mas e size 2000. Polys y ene la ex and sphe ical glass beads o 300 nm-100 μm we e used as e e ence ma e ials. Samples we e measu ed in ip- lica es unde condi ions o epea abili y and he a e age and p e- cision a e epo ed. Aliquo s o sol en -ex ac ed sedimen s we e hea ed o 150 ◦C o 24 h o ensu e OM emo al and cooled o oom empe a u e o ca. 45 min. Ni ogen-based BET (B unaue –Emme –Telle ) su - ace a ea de e mina ions we e made on sample aliquo s o ∼30 mg on a Quan ach ome NOVA 4000e. Degassing was pe o med wi h a Quan ach ome FLOVAC degasse a 150 ◦C unde acuum o e nigh . P ecision was >±0.04 m2/g and <±1.00 m2/g de e - mined om eplica e measu emen s o low- and high-su ace a ea Quan ach ome ins umen s s anda ds, espec i ely. 2.4. Radioca bon analyses o plank onic o amini e a and OC All adioca bon measu emen s we e pe o med using gas ion sou ce accele a o mass spec ome y (AMS) on he compac Mini Ca bon Da ing Sys em (MICADAS) a he Labo a o y o Ion Beam Physics, ETH Zü ich. An aliquo (12-70 mg) o eeze-d ied sedimen was umiga ed wi h concen a ed HCl (37%, 72 h) o emo e ino ganic ca bon and Table 1 Sample loca ions. Mean annual ne p ima y p oduc i i y (NPP) was based on chlo ophyll-a, SST, and PAR sa elli e measu emen s o he MODIS mission (2003-2016) da a. A e age sedimen a ion a e (SR) is adop ed om p e ious wo ks conduc ed a iden ical o nea by loca ions and wa e dep hs. Loca ion [ac onym] Sample name C uise/yea Longi ude La i ude Wa e dep h [m] NPP [mg C m−2day−1] A e age SR [cm ky −1] Re e ence Deposi ional se ing/Oxygen condi ions Pe u ian ma gin [PER] KNR 182/9 MC13 0-3 cm KNR 182/9 2005 −78.17 −11.00 326 2773 20.8 We e e al. (1990)Ou e con inen al shel /Anoxic (OMZ impingemen ) Namibian ma gin [NAM] 2017 26-90 0-3 cm Mi abilis May 2016 13.3 −26 1277 1431 22 In ho n e al. (2006)Mid-slope/Oxic NW A ican ma gin [NAF] OC437 MC31 0-3 cm Chee a C uise 2007 −13.74 27.54 1090 1377 20.3 Mollenhaue e al. (2005)Uppe con inen al slope/Oxic NW A lan ic ma gin [NAT] OCE326 MC2 0-3 cm Be muda Rise 1998 −70.54 40.46 80 1276 32 Bo hne e al. (1981)Shel depocen e /Oxic San a Ba ba a Basin [SBB] SBB4 MC43 0-2 cm New Ho izon 2001 −119.87 34.33 340 1172 200 Schaa and Thu ow (1995)Lowe flank o he basin/Sub-oxic (OMZ impingemen ) San a Monica Basin [SMB] SMB1 MC49 1-2 cm New Ho izon 2001 −119.22 33.90 765 1055 200 Bales a e al. (2018)Sligh ly sloping basin floo /Anoxic (OMZ impingemen ) Be muda Rise [BER] OCE326 BC9 2-5 cm Be muda Rise 1998 −57.61 33.69 4517 374 27 Ohkouchi e al. (2002)D i deposi /Oxic Ibe ian ma gin [IBE] SHAK06-5K 2-3 cm JC089 2013 −10.09 37.34 2646 1030 22 Magill e al. (2018)D i deposi /Oxic 3 B. Ausín, E. B uni, N. Haghipou e al. Ea h and Plane a y Science Le e s 558 (2021) 116759 subsequen ly kep unde basic a mosphe e (NaOH pelle s, 72 h) in a desicca o a 60 ◦C. Samples we e hen w apped in in cap- sules and measu ed by elemen al analyze (EA) – AMS. Fossil (in house shale) and mode n (in house sedimen ) e e ence ma e ials we e p epa ed ollowing he same p ocedu e and used as p ocess- ing blanks. Oxalic acid e e ence ma e ial (NIST SRM 4990C) was used as no maliza ion s anda d. Capsule con ibu ion and cons an con amina ion in oduced du ing umiga ion and EA – AMS analy- ses was assessed and co ec ed o acco ding o Wel e e al. (2018). Co ec ion pa ame e s we e a ca bon mass o 8 ±4μg wi h a F14C o 0.6 ±0.2. Monospecific samples o Globige ina bulloides o Globige inoides ube (40-100 es s; 0.5-1 mg o ca bona e) we e picked om he 250-300 μm ac ion in six ou o he eigh samples. Mixed plank- onic o amini e a we e picked om SBB and none o PER due o he sca ci y o o amini e a. An au oma ed me hod o acid di- ges ion was applied o leach he ou e shell and emo e su ace con aminan s (Wacke e al., 2014). In sho , samples we e placed in sep um-sealed ials and he ou e shell was leached by adding 100 μl o 0.06 M HCl. Subsequen ly, he emaining sample was acidified a e he addi ion o 100 μl o ul apu e H3PO4(85%). The CO2o bo h, he leacha e and he main ac ion, was succes- si ely collec ed on a zeoli e ap o he gas in e ace sys em and ans e ed in o a sy inge, whe e i was dilu ed wi h He o a 5% CO2mix u e and measu ed as wo sepa a e samples in he AMS. Fossil and mode n e e ence ma e ials, i.e. ma ble (IAEA-C1, nomi- nal alue 0.0020 ±0.0010) and an in-house co al s anda d (CSTD, nominal alue 0.9447 ±0.0002, G. Dos San os, pe sonal communi- ca ion), we e p epa ed in he same way as he samples and used, espec i ely, o blank co ec ion and alida ion pu poses. All a- dioca bon esul s a e epo ed as ac ion mode n (F14C), which co esponds o he ac i i y a io o he sample ela i e o a mode n e e ence ma e ial (Reime e al., 2004). The adioca bon con en o he OC (14C) con ained wi hin g ain size samples is also epo ed and calcula ed conside ing F14C and yea o sample collec ion (S u- i e and Polach, 1977). 2.5. OC con en and p ope ies The OC con en was measu ed simul aneously wi h OC-14C us- ing he EA o an accu acy o be e han 0.1% based on s anda ds. Fo s able ca bon iso opic analyses, aliquo s o up o 200 mg o eeze-d ied sedimen we e acidified by means o acid insing wi h 5 mL o 6M HCl o emo e ca bona es, subsequen ly neu alized wi h deionized wa e and d ied in an o en a 50 ◦C. A e sample homogeniza ion, an aliquo (0.5-100 mg) was measu ed o s able ca bon iso opic composi ion (δ13COC) on an FlashEA elemen al ana- lyze coupled in con inuous-flow wi h a Del a V iso ope a io mass spec ome e (all The moFische Scien ific, B emen, Ge many). Ca - bon iso ope alues a e epo ed in pe mille no a ion () ela i e o Vienna Pee Dee Belemni e (VPDB). The p ecision was be e han 0.1(1σ) based on eplica e measu emen s o s anda ds. Sim- ila ly, 25-30 mg o non-deca bona ed eeze-d ied sedimen was used o measu e o al ni ogen concen a ion (N o ) and calcula e he a omic C/N a io. 3. Resul s 3.1. Bulk sedimen cha ac e is ics The majo i y o sedimen mass is ound in he CS and FS ac- ions, in ha o de , oge he accoun ing o >80% o he bulk sed- imen in all samples (Table S1). O e all, sand is he leas abundan ac ion, ollowed by clay. Mean pa icle size is wi hin he lowe ange o CS (∼10.5 μm in PER, ∼30 μm in NAM and ∼20 μm in all he o he si es; Fig. S1A). Excep o BER, which shows a bi- modal size dis ibu ion cen e ed a 7 and 30 μm, all samples show a unimodal pa icle diame e anging om 10 o 30 μm (Fig. S1B). OC con en is highes in sedimen s om he Pe u ian (13.2%) and Namibian (6.3%) ma gins si es, mode a e in sedimen s om San a Ba ba a and San a Monica Basins (∼2.25%) and low (<1%) in sedimen s a he o he si es, eaching minimum alues in Be muda ise (0.4%) (Table S1). These esul s a e in e y good ag eemen wi h p e ious wo ks a simila loca ions (A hu e al., 1998; In ho n e al., 2006; Mollenhaue and Eglin on, 2007; Mollenhaue e al., 2005; Ohkouchi e al., 2005; Rowe e al., 1988). The δ13COC alues ange om 19.8 o 23.2and, oge he wi h C/N a ios be ween 8.7 and 6.0, indica e OM is p edominan ly o ma ine o i- gin (Meye s, 1994). The nea by basins o San a Ba ba a and San a Monica show equi alen TOC, δ13COC and C/N alues, e idencing hei geog aphical p oximi y and simila deposi ional se ings. 3.2. Su ace a ea (SA) The highes su ace a ea is obse ed o clay and/o FS ac ions, while CS shows he lowes alues (Fig. 2A). G ain-size ac ions and bulk sedimen samples om he Pe u ian and he Namibian ma gin exhibi an OC/SA a io >1 mg OC/m2while samples om he NW A ican and Cali o nian ma gins show alues be ween 1 and 0.4 mg OC/m2, and hose om he NW A lan ic ma gin and Be muda ise a e below he 0.4 mg OC/m2 h eshold. 3.3. OC p ope ies: con en and s able iso opic composi ion The Pe u ian ma gin shows he highes OC con en wi h la ge a iabili y among g ain-size ac ions in absolu e e ms (5-18%) (Table S1, Fig. 3A). The opposi e is ue o Be muda Rise, which has he lowes OC con en and li le a iabili y among ac ions (0.1-0.5%). O e all, FS hos s he highes p opo ion o OC in all he si es ollowed by he clay ac ion in h ee o hem. The CS ac ion is usually he poo es ac ion, while bulk sedimen eflec s in e - media e OC con en alues wi hin he ange displayed by ela ed g ain-size ac ions. Clay shows a g ea e a ia ion in OC con en in ela ion o he o he ac ions. Co esponding δ13COC alues a y be ween −23.5 and −19.5 (Table S1, Fig. 3B and C). F ac ions show ela i ely homogeneous alues wi hin each si e, a y wi hin a 1.5 ange o δ13COC, and no sys ema ic pa e n is e iden among g ain-size sedimen sam- ples. Wi h he excep ion o he Pe u ian and he San a Monica Basin clay ac ions, which show C/N a ios o 14.9 and 12.1, espec i ely, all samples exhibi a C/N a io be ween 4 and 10 (Table S1, Fig. 3C). Al hough no clea pa e ns among g ain-sizes a e obse ed, C/N alues o FS and bulk sedimen a e simila (i no iden ical) o mos si es. 3.4. Radioca bon con en and age o OC and plank onic o amini e a Plank onic o amini e a 14C ages a e sys ema ically younge han co-deposi ed OC o all si es (Fig. 4). Samples om San a Ba - ba a and San a Monica basins imply inco po a ion o bomb 14C (Table S2) (Mollenhaue and Eglin on, 2007), meaning he analyzed o amini e a o igina ed some ime a e he he monuclea weapon es ing (1960’s) and canno be used o an accu a e assessmen o o amini e a-g ain size 14C age ela ionships. Replica e 14C measu emen s on OC om bulk sedimen samples o 5 ou o 8 si es yielded i ually iden ical esul s (Table S2). Radioca bon ages o bulk sedimen samples gene ally ep esen a e age alues wi hin he ange eco ded by ela ed ac ions. Ex- cep ions a e he wo eplica es o he Be muda Rise bulk sedimen , which lie in he uppe ange, bu wi hin he e o ma gin o CS, 4 B. Ausín, E. B uni, N. Haghipou e al. Ea h and Plane a y Science Le e s 558 (2021) 116759 Fig. 2. A) Mine al su ace a ea (SA) and B) OC con en s SA o bulk sedimen and g ain-size ac ions. OC loadings (OC/SA a io) h esholds disc imina e be ween highly p oduc i e en i onmen s cha ac e ized by sho OET and la ge OC inpu (>1 mg OC m−2) and long OET and efficien OM emine aliza ion ela i e o supply (<0.4 mg OC m−2), wi h in e media e alues cha ac e izing sedimen s om ypical con inen al shel se ings (Bianchi e al., 2018; Blai and Alle , 2012). Fig. 3. O ganic ca bon con en and p ope ies in bulk sedimen and co esponding g ain-size ac ions (uppe panels) and g ain-size ac ion alues no malized o he bulk sedimen alue (lowe panel) o (A) o ganic ca bon con en (w %), (B) s able ca bon iso opes o OC, and (C) a omic C/N a io. he mos abundan ac ion a his si e, and he NW A ican ma - gin, whe e bulk sedimen is younge han co esponding g ain-size classes. Un o una ely, no eplica e is a ailable o his sample. Excep o BER, he abundance weigh ed a e age F14C and OC con en o he analyzed g ain-size ac ions (i.e., all ac ions bu sand) is lowe han alues ob ained o bulk sedimen s, a disc ep- ancy ha canno be solely accoun ed o he ela i e con ibu ion o he sand ac ion (Table S2). Sample p epa a ion in ol ing MilliQ wa e was minimized o a oid OM loss. Ye , p e e en ial solubiliza- ion o a small ac ion o labile ( ypically younge ) OC in MilliQ wa e is possible. Likewise, emo al o ca bona es by means o acid umiga ion can also lead o dec eased F14C in sedimen samples as a unc ion o ino ganic ca bon con en , en i onmen al ma ix, and p opo ions o labile and e ac o y OC in he sample (Bao e al., 2019a,b). The po en ial impac o ei he p ocesses on F14C al- ues and OC con en canno be disca ded. Ne e heless, ou samples ep esen a wide a ie y in e ms o he ac o s ha de e mine he magni ude o he influence o hese p ocesses, and ye , a s ong 14C age-g ain size dependence is appa en in sedimen s om all si es, whe e clay is he younges ac ion and CS he oldes . 14C age ela ionships be ween he oldes (CS) and younges (clay) ac- ions a y om 515 y s o Pe u o 2335 y s o Be muda Rise, wi h o he si es exhibi ing an in e media e sp ead in ages, mos commonly wi hin he 1040-1620 y s ange (Table S2). 4. Discussion 4.1. The e ec o hyd odynamics on he cha ac e is ics o bulk OC The s udied si es eflec a wide a ay o deposi ional se ings, and ye a s ong and sys ema ic 14C age-mine al g ain-size de- pendence is appa en in all o hem (Figs. 4and 5). Specifically, OC in he clay ac ion is he younges , wi h 14C ages inc eas- ing in he FS and CS, in ha o de . We a ibu e his pa e n o he ubiqui ous influence o hyd odynamic so ing. McCa e e al. (1995)de e mined ha pa icles wi hin he 10-63 μm size ange (i.e., CS, o “so able sil ”) a e non-cohesi e and he e o e exhibi g ea e p opensi y o esuspension and edis ibu ion dispe sal. Ac- 5 B. Ausín, E. B uni, N. Haghipou e al. Ea h and Plane a y Science Le e s 558 (2021) 116759 Fig. 4. Radioca bon ages and age ela ionships in bulk sedimen s and g ain-size classes. A) 14C ages o plank onic o amini e a, OC o bulk sedimen and ela ed eplica es, and OC esiding wi hin each g ain-size sedimen ac ion. Open diamonds indica e o amini e a ha inco po a e bomb 14C. B) Age disc epancy be ween g ain-size ac ions, bulk sedimen , and o amini e a. co dingly, OM esiding wi hin CS is mo e likely o expe ience p o- ac ed ansloca ion, being exposed o aging and selec i e deg a- da ion, and leading o olde OC and lowe OC con en (Figs. 3A and 4). FS (2-10 μm) beha es in a mo e cohesi e manne (McCa e e al., 1995), bu can be also winnowed by s onge cu en s (>10-15 cm s−1) (McCa e and Hall, 2006) and deposi ed in dis al loca ions a e flow educ ion o in e e ence o cohesi e o ces. In con- as , clays a e cohesi e and show a lowe p opensi y o mobiliza- ion and dispe sal, which explains he obse a ion o he younges OC 14C ages o his ac ion (Figs. 4and 5). Whea c o (1992) demons a ed ha he e ec o bio u ba ion can be pa icle-size dependen as fine pa icles pene a e deepe in o he sedimen . High sedimen a ion a es cha ac e is ic o all s udy si es diminish hei suscep ibili y o bio u ba ion (Table 1). Mo eo e , sub-oxic o anoxic condi ions (e.g., PER, SBB and SMB) and he p esence o fine lamina ion in some si es indica e ha bio u ba ion e ec is negli- gible o absen (Schimmelmann e al., 2016) and ye , he same 14C age- mine al-size dependence is e iden a all si es. A compa able OC con en -g ain-size dependence only applies o FS and CS ac ions, as clay shows he lowes OC con en a PER, NAM and SMB, in e media e alues in ela ion o FS and CS a SBB and IBE, and he highes OC con en , compa able o ha obse ed wi hin FS, a NAF, NAT and BER (Table S1). Such dependence may be d i en by bo h hyd odynamically-d i en so ing p ocesses and by o gano-mine al associa ions. The o me ac s h ough p e e en- ial dispe sal o CS in ela ion o FS and clay ac ions (McCa e and Hall, 2006; McCa e e al., 1995) wi h subsequen emine al- iza ion o associa ed OM (Bao e al., 2016), whe eas fine g ain sizes ha e he la ge so p i e capaci y (g ea e SA) han coa se g ain sizes (Keil e al., 1994a,b; Maye , 1993; Maye e al., 1988; P emuzic e al., 1982). The lowe and a iable deg ees o OC as- socia ion wi h he clay-size ac ion con as wi h p io s udies whe e clays a e ypically obse ed o hos g ea e p opo ions o OC ela i e o o he size classes, an obse a ion a ibu ed o he la ge su ace a ea o clay-size mine als a ailable o OM asso- cia ion (Be gamaschi e al., 1997; Keil e al., 1994a,b; P emuzic e al., 1982). Howe e , he e a e a ious po en ial modes o OM- mine al in e ac ion (Maye , 1993). Fo ins ance, su ace oughness ea u es like nanopo es may inc ease he amoun o OC embedded wi hin mine al su aces by one o wo o de s o magni ude (Maye , Fig. 5. Radioca bon age-mine al g ain-size ela ionship wi h OC con en . Maximum and minimum bubble size o OC con en a e adjus ed o he OC con en ange co e ed by he g ain-size ac ions o each s udy si e. 1993, 1994a). This is pa icula ly so o coa se ac ions, whe eas clay mine als exhibi li le po en ial o gene a ing su ace ough- ness (Ransom e al., 1998). Likewise, clay mine alogy (i.e., densi y, chemis y and floccula ion beha io ) has p o en o play a key ole in OM-clay mine al associa ions (Maye e al., 2004; Ransom e al., 1998). Such ac o s may accoun o lowe - han-expec ed OC con- en s o clay ac ions om some o he samples. In gene al, δ13COC and C/N alues o bulk sedimen s and g ain size classes a e consis en wi h ma ine inpu s as he p ima y OM sou ce (Meye s, 1994) (Table S1, Fig. 3A and B). Highe δ13COC alues a NAM, NAF and PER ag ee wi h la ge dus annual deposi- ion om adjacen dese s in hese egions (Jickells e al., 2005), whe eas mo e nega i e δ13COC alues a e obse ed a BER, he 6 B. Ausín, E. B uni, N. Haghipou e al. Ea h and Plane a y Science Le e s 558 (2021) 116759 Fig. 6. Rela ionship be ween 14C and SA wi h OC con en (A) and δ13COC (B). mos o sho e loca ion. Despi e significan a iabili y in adioca - bon composi ion among g ain-size ac ions a each si e, a ia ions in C/N a ios and s able iso opes a e ela i ely small, and do no allow o a mo e de ailed disc imina ion o sedimen a y OM o i- gin. Mo eo e , such pa ame e s may be influenced by selec i e deg ada ion o specific compounds as a consequence o di e en ial esuspension and dispe sal o di e en g ain-size classes. Selec i e emo al o labile sedimen a y OM is expec ed o lead o lowe 14COC (olde 14C ages) and δ13COC ( ia p e e en ial deg ada ion o 13C-en iched, labile o ganic componen s) alues (Blai and Alle , 2012; Wang and D u el, 1996). Howe e , he SA-δ13COC ela ion- ship is less clea han ha o SA wi h 14COC, indica ing ha he possible influence o hyd odynamic so ing on OC s able iso opic composi ion h ough selec i e OC deg ada ion wi hin specific ac- ions does no mani es i sel in δ13COC-g ain-size/SA ela ionships (Fig. 6). These esul s migh also eflec he influence o sample ea men (i.e., acid insing) on δ13COC alues. Remo al o ino - ganic ca bon by means o acid insing can lead o a la ge loss o OC in ela ion o acid umiga ion (Bao e al., 2019b), he la e em- ployed he e p io 14C and OC con en analyses. I is possible ha di e ences in he amoun o OC ha can po en ially be los due o he applica ion o each o hese acidifica ion ea men s led o he p ese a ion o appa en pa e ns in ela ion o hyd odynamic con- ols only on 14C age and OC con en . O e all, he iso opic con as s be ween mo e and less labile OM appea insufficien o disce n he possible influence o hyd odynamically d i en-so ing p ocesses. 4.2. The key ole o fine sil in de e mining bulk OC signa u es The OC-14C age o bulk sedimen samples in es iga ed in his s udy sys ema ically exceeds ha o co-deposi ed plank onic o amini e a (Fig. 4A), implying supply o OM ha is olde han ha deli e ed om o e lying su ace wa e p oduc i i y. Gi en an absence o significan inpu s o (p e-aged) e es ial OM, hese e- sul s p o ide e idence o a pe asi e influence o hyd odynamic so ing on bulk sedimen a y OC 14C (see sec ion 4.1). We a gue ha di e en ial con ibu ions om g ain-size classes ca ying OM wi h a ying OC con en s and 14C signa u es a e esponsible o hese disc epancies. The ela i e impac o each g ain-size ac ion depends on he in e play be ween hei ac ional con ibu ion o he sedimen mass, he amoun o OC hey hos , and he specific p ope ies o he la e (e.g., 14C and δ13C). Clay con ibu es a neg- ligible/mino amoun o he bulk sedimen mass, shows a iable OC con en in compa ison o ela ed CS and FS ac ions wi hin each si e, and his ac ion always con ains he younges OC (Ta- ble S1, Fig. 5). OC associa ed wi h FS is sys ema ically olde han clay and comp ises 34% o bulk sedimen on a e age, while CS is gene ally he mos abundan ac ion (50% on a e age) hos ing he oldes OC. Howe e , OC con en s o CS a e subs an ially lowe , and he majo i y o OC esides wi hin he FS (Fig. 5). Co espond- ingly, he la e ac ion exhibi s he g ea es influence on bulk OC cha ac e is ics. Mo eo e , gi en he p opensi y o his ac ion o unde go mobiliza ion and ansloca ion, FS has he po en ial o o e w i e local/con empo aneous signals eco ded in ma ine sed- imen s. Indeed, 14C ages and C/N a ios o FS a e compa able o hose o bulk sedimen o mos si es (Fig. 3D, E, F, Table S1). Ou esul s sugges , he e o e, ha age o se s p e iously obse ed be ween co-deposi ed bulk OC and plank onic o amini e a (e.g. Ausín e al., 2019; Kusch e al., 2010; Mollenhaue and Eglin on, 2007; Mollenhaue e al., 2005; Ohkouchi e al., 2002) we e mos likely due o he influence o p e-aged OC associa ed wi h FS ha was ad ec ed o he s udy si es. Olde - han- o amini e al 14C ages o FS and CS sugges he occu ence o p ocesses such as pos deposi ional al e a ion o sed- imen a y OC ia bio u ba ion and/o deli e y o alloch honous OC ia i e ine and ae osol anspo , emobiliza ion o in-si u sedimen s, and/o ad ec ion o dis an ma e ial. The influence o bio u ba ion is expec ed o be minimal a all si es. Mo eo e , pa icle-size dependen bio u ba ion ypically esul s in younge - han- o amini e al 14C ages o smalle pa icles (Ba d, 2001; Whea c o , 1992). The C/N and δ13COC alues sugges ha con- ibu ions o e es ial OC o FS and CS ac ions (supplied ia flu ial o aeolian anspo ) a e negligible. The measu emen s pe - o med in he p esen s udy do no allow us o disce n be ween emobiliza ion/ e-suspension o p oximal e sus dis al sedimen s. Howe e , p io s udies specially designed o iden i y hese p o- cesses sugges ha long-dis ance la e al sedimen anspo is a majo ansloca ion mechanism on di e se con inen al ma gin sys- ems including o Namibia (In ho n e al., 2006), Ibe ia (Magill e al., 2018), Sibe ia (B öde e al., 2018), China and he no hwes e n US ma gins (Bao e al., 2018b). Deposi ion o la e ally anspo ed OM associa ed wi h fine-g ained mine als implies he en ainmen 7 B. Ausín, E. B uni, N. Haghipou e al. Ea h and Plane a y Science Le e s 558 (2021) 116759 o alloch honous and p e-aged OC ha migh pe u b he o igi- nal en i onmen al signals o igina ing om o e lying wa e s (Ausín e al., 2019). Fu he conside a ion o hese issues, bo h om he pe spec i e o unde s anding ca bon cycling and in e p e a ion o con inen al ma gin sedimen a y eco ds is he e o e wa an ed. 4.3. A amewo k o assessing si e-specific a iabili y Al hough all s udy egions show a s ong dependence o SA, 14C age, and OC con en on g ain-size, he amoun o p ese ed OC in compa able ac ions and he magni ude o 14C disc epancies among g ain-size ac ions (Figs. 3A and 4B) a e highly a iable be- ween si es. Fo ins ance, he g ain-size classes om he Pe u ian ma gin show simila and ela i ely young 14C ages, bu he la ges OC con en a iabili y among ac ions o all si es (Figs. 3and 4). Con e sely, ac ions om Be muda Rise show la ge 14C age a i- abili y bu uni o mly low alues o OC con en . This obse a ion is asc ibed o di e ences in he egional condi ions ha cha ac- e ize each si e, such as p ima y p oduc i i y, sedimen a ion a e, OET, hyd odynamic egime, and deposi ional se ing, among o he s (A na son and Keil, 2007; Mülle and Suess, 1979). Indeed, co - esponding OC loadings (OC/SA a ios; Fig. 2B) indica e ha esh OM supply p edomina es on he high p oduc i e Pe u ian ma - gin despi e s ong coas al cu en s, whe eas he opposi e occu s in Be muda Rise, cha ac e ized by oligo ophic o e lying wa e s, long OETs and s ong ad ec i e cu en s and sedimen ocusing (Mc- Ca e, 2002). These wo si es can be conside ed as end-membe s wi hin he s udied deposi ional se ing ange (Fig. 7), and p o ide dis inc in o ma ion abou he ole o OM-mine al associa ions and hyd odynamic p ocesses on he a e o OC p ese ed in ma ine sedimen s. A si es like he Pe u ian ma gin (and NAM o a lesse ex en ), he e ec o hyd odynamic so ing – while no iceable in he 14C age-g ain size ela ionship – is bu e ed in he bulk sedi- men by he la ge and con inuous e ical flux o esh OM. He e, 14C age-g ain size ela ionship p o ides in o ma ion on he po en- ial o each ac ion o associa e wi h OC and hos i a an ea ly s age. In con as , he majo i y o deposi ed OM a si es such as he Be muda Rise (and IBE o a lesse ex en ) has p e iously been exposed o s ong hyd odynamic o cing ( ia en ainmen wi hin u bidi y cu en s o in e media e/bo om nepheloid laye s, and/o esuspension cycles), while labile biological deb is se ling om he o e lying wa e column has expe ienced ex ensi e deg ada- ion du ing e ical ansi and sedimen a ion. Long- e m oxygen exposu e may also ha e played a ole hough he des uc ion o o ganic-mine al agg ega es (A na son and Keil, 2007). The la e condi ions, which may lead o simila OC con en in each o he ac ions ega dless o hei ini ial po en ial o hos OM, ep esen a “ma u e” sys em ha may in o m abou he impac o p olonged and s ong hyd odynamic p ocesses on sedimen a y OC. Si es ex- hibi ing in e media e 14C age and OC con en disc epancies be- ween g ain-size classes ep esen sys ems ha in o m abou he po en ial o each ac ion o p o ec OC om he e ec s o hyd o- dynamic p ocesses. These loca ions also show in e media e mine al su ace a ea-no malized OC loadings (0.4-1 mg OC m−1) (Fig. 2B) co esponding o ela i ely s able OM-mine al associa ions (Blai and Alle , 2012), implying p o ec ion om emine aliza ion (Keil e al., 1994b). Howe e , we a gue ha hese si es ep esen a s a- ble si ua ion ha is no necessa ily pe manen . Ashi owa ds s onge hyd odynamic o cing may emobilize hese sedimen s, exposing associa ed OM o u he decomposi ion and shi ing esidual ma e ials owa ds a mo e ma u e end-membe . S udy si es a e classified wi hin one o hese sys ems acco ding o OC/SA al- ues and so ed as a unc ion o co esponding PP, OC con en - and 14C-disc epancy alues. In e es ingly, wa e dep h inc eases con in- uously om s able o ma u e sys ems, in ag eemen wi h p e ious wo k in he Chinese Ma ginal Seas ha e idence inc easing 14C age Fig. 7. Concep ual model summa izing he cha ac e is ics o he “Ini ial”, “S able”, and “Ma u e” sys ems as explained in sec ion 4.3. disc epancies be ween g ain-size sedimen ac ions wi h inc eas- ing dep h (Bao e al., 2019a). Ou esul s confi m such ela ionship in a b oade con ex o s able and ma u e sys ems, whe eas o ini ial sys ems, high PP and e ical inpu o esh OM likely mask he impac o any pos -deposi ional sedimen a ion p ocess on 14C age disc epancies. Mo eo e , i is likely ha he ela i e con ibu- ion o e ac o y OC inc eases p og essi ely om ini ial o ma u e sys ems, due o lowe e ical fluxes o labile OC and selec i e p ese a ion o e igenous and elic OM du ing la e al anspo (Mollenhaue and Eglin on, 2007). In his con ex , ca bon fluxes (p oduc i i y), wa e dep h, OET, and mine al g ain size – he la - e influencing bo h he capaci y o hos OM and p opensi y o ( e)mobiliza ion and la e al anspo – each play a key ole in dic- a ing he amoun , p o enance, composi ion, and in age o o ganic ca bon signa u es p ese ed in con inen al ma gin sedimen s. 5. Conclusions We find ha hyd odynamically-d i en p ocesses exe a pe a- si e influence on he OM con en and composi ion o con inen al ma gin sedimen s, and a e mani es ed in he 14C age a iabili y o OC esiding in bulk sedimen s and co esponding g ain-size ac- ions. Mo eo e , o gano-mine al in e ac ions exe a dual influence on sedimen a y OC, wi h espec o bo h OM p o ec ion and i s p opensi y o mobiliza ion and edis ibu ion. OC is p e e en ially associa ed wi h he fine sil ac ion in all si es. Due o i s highe OC con en and con ibu ion o bulk sedi- men mass, his ac ion s ongly influences bulk sedimen OC 14C age and C/N o OC. Gi en i s p opensi y o esuspension and ad- ec ion unde s ong cu en s, ansloca ed (alloch honous) OC as- socia ed wi h fine sil may supp ess and/o dis o p ima y signals o igina ing om o e lying su ace wa e s p ese ed wi hin con i- nen al ma gin sedimen s. Based on obse ed ela ionships be ween p oduc i i y, oxygen exposu e, wa e dep h, and mine al g ain size, we sugges ha di - e en deposi ional en i onmen s can be ca ego ized as “ini ial”, “s able” (quasi-s eady s a e), and “ma u e” phases. These a e dis- inguished based on 14C-age and OC con en ela ionships among g ain-size classes ha eflec he in e play o hyd odynamic so - ing and o he egional condi ions. Deposi ional sys ems eflec ing he “ini ial” phase include hose wi h high p oduc i i y, low oxy- gen exposu e and local deposi ion (i.e., Pe u ian and Namibian ma gins), and a e cha ac e ized by small 14C age- bu la ge OC con en -disc epancies among g ain-size ac ions. “S able” phases o deposi ion occu in sys ems ha a e cha ac e ized by high o mode a e p oduc i i y and sedimen ocussing (e.g. NW A lan ic, and NW A ican ma gins), and ea u e mode a e 14C age- and OC con en -disc epancies among g ain-size ac ions. Addi ional ac- o s, such as a iabili y in oxygen exposu e (e.g., SBB, SMB) con- ibu e o composi ional he e ogenei y wi hin s able deposi ional se ings. “Ma u e” sys ems, which exhibi la ge 14C age disc epan- cies bu smalle a ia ions in OC con en among g ain-size classes (i.e., Be muda Rise and SW Ibe ian ma gin o a lesse ex en ), a e 8 B. Ausín, E. B uni, N. Haghipou e al. Ea h and Plane a y Science Le e s 558 (2021) 116759 cha ac e ized by weak e ical OC expo and s ong ad ec i e cu - en s and la e al anspo , eflec ing he impac o p olonged ex- posu e o hyd odynamic p ocesses on sedimen a y OC. This simple in e p e a ional amewo k may help o assess he ex en o which down-co e a ia ions in con inen al ma gin sedimen a y eco ds eflec hyd odynamic e sus o he en i onmen al changes. CRediT au ho ship con ibu ion s a emen B.A. and T.I.E. planned his in es iga ion. N.H. and C.W. assis ed wi h adioca bon analyses. E.B. assis ed wi h g ain-size and su - ace a ea analyses. S.M.B. assis ed wi h s able ca bon iso ope and ni ogen de e mina ions. B.A. p epa ed he samples, analyzed he esul s, and w o e he manusc ip wi h con ibu ions by all coau- ho s. Decla a ion o compe ing in e es The e a e no in e es s o decla e. Acknowledgemen s We would like o hank Madalina Jaggi and Daniel Mon luçon o hei assis ance du ing s able iso ope and su ace a ea analyses. We acknowledge he Regional G adua e Ne wo k o Oceanog aphy (RGNO) Disco e y Camps, suppo ed by he Agou on Ins i u e, he Simons Founda ion, he Scien ific Commi ee o Oceanog aphic Resea ch (SCOR), he Minis y o Fishe ies and Ma ine Resou ces (MFMR), he Na ional Ma ine In o ma ion and Resea ch Cen e (Na MIRC), he Uni e si y o Namibia (UNAM), ETH Zu ich and he Swiss i- esea ch & aining ins i u e, as well as scien is s and c ew o he R/V Mi abilis o ealiza ion o Namibian ma gin sam- pling. We g ea ly app ecia e he anonymous e iewe o hei ca e ul eading o ou manusc ip and hei many insigh ul com- men s. This s udy was suppo ed by he p ojec “TRAMPOLINE” (200021_175823) unded by he Swiss Na ional Science Founda- ion, g an ed o T.I.E. and B.A. Appendix A. Supplemen a y ma e ial Supplemen a y ma e ial ela ed o his a icle can be ound on- line a h ps://doi .o g /10 .1016 /j .epsl .2021.116759. Re e ences A na son, T.S., Keil, R.G., 2007. Changes in o ganic ma e –mine al in e ac ions o ma ine sedimen s wi h a ying oxygen exposu e imes. Geochim. Cosmochim. Ac a 71, 3545–3556. A nd , S., Jø gensen, B.B., LaRowe, D.E., Middelbu g, J.J., Pancos , R.D., Regnie , P., 2013. Quan i ying he deg ada ion o o ganic ma e in ma ine sedimen s: a e iew and syn hesis. Ea h-Sci. Re . 123, 53–86. A hu , M.A., Dean, W.E., Laa kamp, K., 1998. O ganic ca bon accumula ion and p ese a ion in su ace sedimen s on he Pe u ma gin. Chem. Geol. 152, 273–286. Ausín, B., Magill, C., Haghipou , N., Fe nández, Á., Wacke , L., Hodell, D., Baumann, K.-H., Eglin on, T.I., 2019. (In)cohe en mul ip oxy signals in ma ine sedimen s: implica ions o high- esolu ion paleoclima e econs uc ion. Ea h Plane . Sci. Le . 515, 38–46. Bales a, B., Quin ana K upinski, N.B., E ohina, T., Fessenden-Rahn, J., Rahn, T., Pay- an, A., 2018. Bo om-wa e oxygena ion and en i onmen al change in San a Monica Basin, Sou he n Cali o nia du ing he las 23ky . Palaeogeog . Palaeo- clima ol. Palaeoecol. 490, 17–37. Bao, R., Bla mann, T.M., McIn y e, C., Zhao, M., Eglin on, T.I., 2019a. Rela ionships be ween g ain size and o ganic ca bon 14C he e ogenei y in con inen al ma gin sedimen s. Ea h Plane . Sci. Le . 505, 76–85. Bao, R., McIn y e, C., Zhao, M., Zhu, C., Kao, S.-J., Eglin on, T.I., 2016. Widesp ead dispe sal and aging o o ganic ca bon in shallow ma ginal seas. Geology. Bao, R., McNichol, A.P., Hemingway, J.D., La die Gaylo d, M.C., Eglin on, T.I., 2019b. Influence o di e en acid ea men s on he adioca bon con en spec um o sedimen a y o ganic ma e de e mined by RPO/accele a o mass spec ome y. Radioca bon 61, 395–413. Bao, R., Uchida, M., Zhao, M., Haghipou , N., Mon lucon, D., McNichol, A., Wacke , L., Hayes, J.M., Eglin on, T.I., 2018a. O ganic ca bon aging du ing ac oss-shel anspo . Geophys. Res. Le . 45, 8425–8434. Bao, R., Uchida, M., Zhao, M., Haghipou , N., Mon lucon, D., McNichol, A., Wacke , L., Hayes, J.M., Eglin on, T.I., 2018b. O ganic ca bon aging du ing ac oss-shel anspo . Geophys. Res. Le . Bao, R., an de Voo , T.S., Zhao, M., Guo, X., Mon luçon, D.B., McIn y e, C., Eglin on, T.I., 2018c. Influence o hyd odynamic p ocesses on he a e o sedimen a y o - ganic ma e on con inen al ma gins. Glob. Biogeochem. Cycles 32, 1420–1432. Ba d, E., 2001. Paleoceanog aphic implica ions o he di e ence in deep-sea sedi- men mixing be ween la ge and fine pa icles. Paleoceanog aphy 16, 235–239. Be gamaschi, B.A., Tsamakis, E., Keil, R.G., Eglin on, T.I., Mon luçon, D.B., Hedges, J.I., 1997. The e ec o g ain size and su ace a ea on o ganic ma e , lignin and ca bohyd a e concen a ion, and molecula composi ions in Pe u Ma gin sedi- men s. Geochim. Cosmochim. Ac a 61, 1247–1260. Bianchi, T.S., Cui, X., Blai , N.E., Bu dige, D.J., Eglin on, T.I., Galy, V., 2018. Cen- e s o o ganic ca bon bu ial and oxida ion a he land-ocean in e ace. O g. Geochem. 115, 138–155. Blai , N.E., Alle , R.C., 2012. The a e o e es ial o ganic ca bon in he ma ine en- i onmen . Annu. Re . Ma . Sci. 4, 401–423. Bo hne , M.H., Spike , E.C., Johnson, P.P., Rendigs, R.R., A usca age, P.J., 1981. Geo- chemical e idence o mode n sedimen accumula ion on he con inen al shel o sou he n New England. J. Sedimen . Res. 51, 281–292. B öde , L., Tesi, T., Ande sson, A., Semile o , I., Gus a sson, Ö., 2018. Bounding c oss- shel anspo ime and deg ada ion in Sibe ian-A c ic land-ocean ca bon ans- e . Na . Commun. 9, 806. Fische , G., Ka akas, G., Blaas, M., Ra meye , V., Nowald, N., Schli ze , R., Helmke, P., Da enpo , R., Donne , B., Neue , S., We e , G., 2009. Mine al ballas and pa i- cle se ling a es in he coas al upwelling sys em o NW A ica and he Sou h A lan ic. In . J. Ea h Sci. 98, 281–298. Gibbs, R.J., Ma hews, M.D., Link, D.A., 1971. The ela ionship be ween sphe e size and se ling eloci y. J. Sedimen . Res. 41, 7–18. Hedges, J.I., Keil, R.G., 1995. Sedimen a y o ganic ma e p ese a ion: an assess- men and specula i e syn hesis. Ma . Chem. 49, 81–115. In ho n, M., Wagne , T., Scheede , G., Zabel, M., 2006. La e al anspo con ols dis ibu ion, quali y, and bu ial o o ganic ma e along con inen al slopes in high-p oduc i i y a eas. Geology 34, 205–208. Jickells, T.D., An, Z.S., Ande sen, K.K., Bake , A.R., Be game i, G., B ooks, N., Cao, J.J., Boyd, P.W., Duce, R.A., Hun e , K.A., Kawaha a, H., Kubilay, N., laRoche, J., Liss, P.S., Mahowald, N., P ospe o, J.M., Ridgwell, A.J., Tegen, I., To es, R., 2005. Global i on connec ions be ween dese dus , ocean biogeochemis y, and clima e. Sci- ence 308, 67–71. Keil, R.G., Maye , L.M., 2014. Mine al ma ices and o ganic ma e . In: T ea ise on Geochemis y, 2nd ed. Else ie , pp. 337–359. Keil, R.G., Mon lucon, D.B., P ahl, F.G., Hedges, J.I., 1994a. So p i e p ese a ion o labile o ganic ma e in ma ine sedimen s. Na u e 370, 549–552. Keil, R.G., Tsamakis, E., Fuh, C.B., Giddings, J.C., Hedges, J.I., 1994b. Mine alogi- cal and ex u al con ols on he o ganic composi ion o coas al ma ine sed- imen s: hyd odynamic sepa a ion using SPLITT- ac iona ion. Geochim. Cos- mochim. Ac a 58, 879–893. Kusch, S., Eglin on, T.I., Mix, A.C., Mollenhaue , G., 2010. Timescales o la e al sedimen anspo in he Panama Basin as e ealed by adioca bon ages o alkenones, o al o ganic ca bon and o amini e a. Ea h Plane . Sci. Le . 290, 340–350. Magill, C.R., Ausín, B., Wenk, P., McIn y e, C., Skinne , L., Ma ínez-Ga cía, A., Hodell, D.A., Haug, G.H., Kenney, W., Eglin on, T.I., 2018. T ansien hyd odynamic e ec s influence o ganic ca bon signa u es in ma ine sedimen s. Na . Commun. 9, 4690. Maye , L.M., 1993. O ganic ma e a he sedimen -wa e in e ace. In: Engel, M.H., Macko, S.A. (Eds.), O ganic Geochemis y. P inciples and Applica ions. Plenum, New Yo k, pp. 171–184. Maye , L.M., 1994a. Rela ionships be ween mine al su aces and o ganic ca bon con- cen a ions in soils and sedimen s. Chem. Geol. 114, 347–363. Maye , L.M., 1994b. Su ace a ea con ol o o ganic ca bon accumula ion in con i- nen al shel sedimen s. Geochim. Cosmochim. Ac a 58, 1271–1284. Maye , L.M., Macko, S.A., Cammen, L., 1988. P o enance, concen a ions and na u e o sedimen a y o ganic ni ogen in he gul o Maine. Ma . Chem. 25, 291–304. Maye , L.M., Schick, L.L., Ha dy, K.R., Wagai, R., McCa hy, J., 2004. O ganic ma - e in small mesopo es in sedimen s and soils. Geochim. Cosmochim. Ac a 68, 3863–3872. McCa e, I.N., 2002. A poisoned chalice? Science 298, 1186–1187. McCa e, I.N., Hall, I.R., 2006. Size so ing in ma ine muds: p ocesses, pi alls, and p ospec s o paleoflow-speed p oxies. Geochem. Geophys. Geosys . 7. McCa e, I.N., Manighe i, B., Robinson, S.G., 1995. So able sil and fine sedimen size/composi ion slicing: pa ame e s o palaeocu en speed and palaeoceanog- aphy. Paleoceanog aphy 10, 593–610. Meye s, P.A., 1994. P ese a ion o elemen al and iso opic sou ce iden ifica ion o sedimen a y o ganic ma e . Chem. Geol. 114, 289–302. Moffi , S.E., Moffi , R.A., Sau ho , W., Da is, C.V., Hewe , K., Hill, T.M., 2015. Pale- oceanog aphic insigh s on ecen oxygen minimum zone expansion: lessons o mode n oceanog aphy. PLoS ONE 10, e0115246. 9